Comparison of systematically functionalized heterogeneous and homogenous glycopolymers as toxin inhibitors

Article abstract of DOI:10.1002/pola.29279

Multivalent glycosylated polymers and particles display enhanced binding affinity toward lectins compared to individual glycans. The design of glycopolymers with selectivity toward pathogen-associated lectins (toxins) for sensing or in antiadhesion therap

Full text of DOI:10.1002/pola.29279

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ORIGINAL ARTICLE
Comparison of Systematically Functionalized Heterogeneous and
Homogenous Glycopolymers as Toxin Inhibitors
1,2
Benjamin Martyn,1 Caroline I. Biggs ,1 Matthew I. Gibson
1Department of Chemistry, University of Warwick, Gibbet Hill Road, Coventry CV4 7AL, United Kingdom
2Warwick Medical School, University of Warwick, Gibbet Hill Road, Coventry CV4 7AL, United Kingdom
Correspondence to: M. I. Gibson (E-mail: m.i.gibson@warwick.ac.uk)
Received 10 October 2018; Accepted 30 October 2018
DOI: 10.1002/pola.29279
ABSTRACT: Multivalent glycosylated polymers and particles dis-
play enhanced binding affinity toward lectins compared to indi-
vidual glycans. The design of glycopolymers with selectivity
toward pathogen-associated lectins (toxins) for sensing or in
antiadhesion therapy is complicated due to lectins having pro-
miscuous binding profiles and can be considered to be pattern
recognition “readers,” with the capability to bind to several dif-
ferent glycans. Here, heterogeneous glycopolymers bearing var-
iable densities of two different monosaccharides are
synthesized by a three-step postpolymerization modification
approach, enabling systematic control over composition. It is
found that heterogeneous polymers displayed increased inhibi-
tory activity, compared to homogeneous polymers, against a
RCA120 and the cholera toxin. This demonstrates that embracing
heterogeneity in glycomaterials could result in improved perfor-
mance or emergent properties. © 2018 The Authors. Journal of
Polymer Science Part A: Polymer Chemistry published by Wiley
Periodicals, Inc. J. Polym. Sci., Part A: Polym. Chem. 2018
KEYWORDS: biomaterials; biological applications of polymers;
carbohydrates; glycopolymer; infection; lectin; polymers; revers-
ible addition fragmentation chain transfer (RAFT)
inhibitor of DC-Sign/GP120 interactions,16 and inhibition of
the Ebola virus by a 120-valent mannose glycofullerene clus-
ter has been observed.17 High affinity sialic acid glycopolymer
arrays have been used by Godula and coworkers to dissect
how influenza engages with its host.18 While these multivalent
systems all show high avidity, they typically present monosac-
charides and hence do not have high specificity. Glycopoly-
mers have been developed by Kiick and Polizzotti19 and
Gibson and coworkers20,21 with high affinity for the cholera
toxin by modulation of the linker length to match the binding
pocket depth, increasing affinity without additional synthetic
complexity. In addition to the chemical complexity of oligosac-
charides, cell surfaces are heterogeneous and dynamically dis-
play many different glycans.22 Evidence is emerging that these
lower affinity glycans play a key role in the overall affinity,
and that reproduction of this heterogeneity in glycomaterials
can enhance avidity.23 García Fernández and coworkers have
shown that lectin binding to β-cyclodextrin-based heterogly-
coclusters exhibit strong synergistic binding effects, termed
the “heterocluster effect.”24 Percec and coworkers have intro-
duced nonbinding glycans into amphiphilic Janus glycodendri-
mers resulting in a 12-fold increase in agglutination activity
for the particles.25 Wu and coworkers reported that cholera
INTRODUCTION Many pathogenic bacteria secrete toxins as
their primary mode of pathogenicity, including Escherichia coli
O-157 Shiga toxins (food poisoning) and the toxin from Vibrio
cholerae (cholera).1–4 Plant toxins can also cause significant
harm, such as ricin (from Ricinus communis) which has a
lethal dose of ~20 μg kg−1 and is more toxic than cyanide.5 A
common feature is that these proteins contain a carbohydrate
binding domain to hijack host cell-surface glycans and gain
entry. Carbohydrate–protein interactions mediate a huge
range of recognition/signaling events including inflammation,6
immune-responses,7 and cell–cell signaling.8 The “reader” pro-
teins that mediate these interactions are known as lectins,
which typically have very weak binding affinities (mM).9
Nature therefore presents multiple copies of each glycan, giv-
ing a nonlinear increase in affinity, known as the “cluster gly-
coside effect.”10,11 This has inspired the use of glycomaterials,
such as polymers or particles, which use polyvalent presenta-
tion of carbohydrates to enhance affinity.12,13 This enhanced
affinity offers a route to prophylactic antiadhesion therapies
against pathogens/toxins, as an alternative to traditional anti-
biotics or small molecule drugs.14 “STARFISH” glycodendri-
mers have 106-fold enhanced activity against Shiga toxins
compared to the free glycan,15 linear poly(mannose) is a nM
Additional supporting information may be found in the online version of this article.
© 2018 The Authors. Journal of Polymer Science Part A: Polymer Chemistry published by Wiley Periodicals, Inc.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and
reproduction in any medium, provided the original work is properly cited.
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2018
1

JOURNAL OF
POLYMER SCIENCE
ORIGINAL ARTICLE
WWW.POLYMERCHEMISTRY.ORG
ꢀ
toxin has increased binding capacity toward mixtures of GM2
and fucosyl-GM1, despite homo fucosyl-GM1 having far lower
binding affinity compared to homo GM2.26 Multivalency is also
observed to result in altered binding affinities compared to
monovalent systems. Richards et al. have reported that man-
a BioTek Synergy plate reader set at 25 C. Ultraviolet–visible
(UV–vis) spectra were obtained an Agilent Cary spectrometer.
BLI was performed using a ForteBio Octet 96 RED interferom-
eter, with the indicated probes. DMF size exclusion chroma-
tography (SEC) was performed on a varian 390-LC MDS
system equipped with a PL-AS RT/MT autosampler, a PLgel
3 μm (50 × 7.5 mm) guard column, two PLgel 5 μm
(300 × 7.5 mm) mixed-D columns using DMF with 5 mM
nosylated gold nanoparticles have affinity toward RCA120
,
despite the monosaccharides having little or no affinity.27
Sequence-controlled glycooligomers have shown increased
affinity due to the nonbinding glycan providing steric shield-
ing rather than more binding28,29 and mixed-carbohydrate
particles have been shown to have increased cell uptake into
macrophages over the homogeneous equivalents.30 This pre-
sent work investigates the impact of sequentially modified
heterogeneous multivalent polymer scaffolds assembled by a
three-step postpolymerization methodology. It is found that
addition of the second glycan has a significant effect on the
overall inhibitory activity compared to homogeneous poly-
mers and shows that screening heterogeneous polymers may
enable the discovery of more active glycomaterials.
NH4BF4 at 50 C as eluent at a flow rate of 1.0 mL min−1. The
ꢀ
SEC system was equipped with UV/vis (set at 280 and
461 nm) and differential refractive index detectors. Narrow
molecular weight poly(methyl methacrylate) standard
(200–1.0 × 106 g mol−1) was used for calibration using a sec-
ond order polynomial fit.
Bilayer Interferometry
Amine reactive Second-Generation BLI sensors from ForteBio
were presoaked for 10 min in Milli-Q water before activation
with an EDC/NHS solution. After 10 min, the sensors were
moved to a solution containing the lectin at 25 μg mL−1 in
pH 5 HEPES. After 10 min, the sensors were quenched with
ethanolamine, placed in HEPES buffer at the corresponding
pH to baseline for 10 min, and tested against five serial dilu-
tions of polymer solution for 30 min. The sensors were then
placed into HEPES buffer for 10 min to measure dissociation.
The raw data were processed using ForteBio analysis software
heterogeneous ligand model. To measure the pseudo-steady-
state Dissociation constant (KDs), the dissociation steps only
where plotted and fit with an exponential decay in Origin to
extract the end point deflection. The end points were then
plotted versus both polymer and galactose concentration and
fit with logistic curves to extract a midpoint value.
EXPERIMENTAL
2-Chloro-1,3-dimethylimidazolinium chloride, and GM1 gangli-
oside were purchased from Carbosynth, (Newbury, UK). D-
(+)-Galactose was purchased from MP Biomedicals (Supplied
by Fisher Scientific UK, Loughborough). D-(+)-mannose, triet-
hylamine, sodium azide, pentafluorophenol, methacroyl chlo-
ride, 2-cyano-2-propyl benzodithioate, 4,40-azobis(4-cyanov-
aleric acid), 1,4-dioxane, dichloromethane, dibenzocyclooctyne
(DBCO)-amine, dimethylformamide (DMF), 2-aminoethan-1-ol,
Greiner 384-well high-binding microliter plates, phosphate-
buffered saline (PBS), HEPES, Fluorescein isothiocyanate
(FITC)-labeled Ctx B, and unlabeled CTx B, were purchased
from Sigma-Aldrich (Gillingham, UK). 2,6-Lutidine 99% was
purchased from Acros Organics. FITC-labeled RCA120 and
RCA120 were purchased from Vector Labs (Peterborough, UK).
Amine reactive 2nd generation biolayer interferometry (BLI)
sensors, 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC),
N-hydroxysuccinimide (NHS), from ForteBio (Supplied by VWR,
Lutterworth, UK). Ultrapure Milli-Q water was obtained from a
Merk Milli-Q water purifier at 18.2 MΩ cm−1 resistivity at
Competitive Binding Assays
Here, 384-well high-binding microliter plates were incubated
for 16 h with 50 μL of 1 mg mL−1 GM1 ganglioside dissolved
in PBS solution, per well. Unattached GM1 was removed by
washing extensively with PBS buffer.
Due to the relatively low solubility of the polymers, polymers
were dissolved in HEPES buffered saline, and any undissolved
material removed by centrifugation. Exact concentrations
were determined by the DBCO absorbance at 292 nm, using a
ꢀ
25 C. 3500 MWCO “SnakeSkin” dialysis tubing was purchased
from Thermo Fisher (Loughborough, UK).
calculated value for the extinction coefficient
ε
as
2.00 mL mg−1 cm−1, shown in Table S1. To perform the com-
petitive binding assays, lectin binding assays were first per-
formed to find the optimal concentration. Serial dilutions of
FITC-labeled RCA and FITC-labeled CTx in HEPES buffered
saline were made and incubated in a GM1 coated 384 well
plates for 30 min at 37 ꢀC. After this time, the plates were
washed extensively with buffer to remove unbound lectin and
the fluorescence. The middle of the dose-dependent binding
curves were chosen, which gave concentrations of 0.13 and
0.05 mg mL−1 for RCA and CTx, respectively.
Physical and Analytical Methods
1H, nuclear magnetic resonance (13C NMR), and 19F spectra
were recorded on Bruker HD-300, HD-400, and AV-500
spectrometers using deuterated solvents purchased from
Sigma-Aldrich. Chemical shifts are reported relative to resid-
ual nondeuterated solvent. Mass spectrometry (MS) was per-
formed on an Agilent 6130B single Quad (electrospray
ionization [ESI]). Fourier transform infrared (FTIR) spectra
were acquired using a Bruker Vector 22 FTIR spectrometer
with a Golden Gate diamond attenuated total reflection cell.
Raman spectra were collected on a Reinshaw inVia Reflex
Raman using a 442 nm HeCd laser. Liquid handling was per-
formed by Gilson Pipette Max. 96-well plates were read using
Polymer solutions were made up as serial dilutions in HEPES
from saturated stock solutions to give a volume of 180 μL in
each well. Total of 120 μL FITC labeled lectin in HEPES was
2
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2018

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ORIGINAL ARTICLE
added to 180 μL of each polymer solution and incubated for
30 min at 37 ꢀC. Total of 45 μL of the polymer FITC-labeled
lectin solutions was then added to the GM1 surfaces and incu-
bated at 37 ꢀC for 30 min. Fluorescence was then measured at
excitation/emission wavelengths of 485/528 nm. All RCA
experiments were carried out 18 times, CTx B experiments
were carried out in triplicate. MIC50 values were calculated
using logistic fitting in Origin.
46.56 (NH-CH2-CH2-C(=O)-DBCO), 43.77 (Backbone CH2),
43.14 (Backbone CH2) 30.93 (backbone CH3).
Synthesis of Glycopolymers
Using stock solutions of
1
mg mL−1 sugar azide
(4.87 × 10−3 mmol mL−1), 2.75 mL of each of; [Gal]:
[Man],100:0, 75:25, 50:50 and 25:75 (v:v) solutions were pre-
pared. For each glycopolymer, 2.75 mL of the corresponding
azido-sugar solution was added to p(DBCO)15(HEMA)35 (5 mg,
357 nmol) in a vial. The reaction was left at room temperature
overnight. To remove excess sugar, the solutions were passed
through a 1000 MWCO centrifugal filter and resuspended in
water three times. The resulting solution was then freeze-
dried. Raman of final polymers showed no presence of alkyne
peak. N.B. To give a 2.5× excess of sugar azide to polymer
alkyne, the number of moles of polymer was multiplied by
15 (for each alkyne unit) and 2.5 (to give an excess of sugar).
Synthesis of Poly(pentafluorophenol methacrylate)
Pentafluorophenyl methacrylate (PFMA; 4.7 g, 18.6 mmol),
2-cyano-2-propyl benzodithioate (55.3 mg, 0.25 mmol) and
4,40-azobis(4-cyanovaleric acid) (35.0 mg, 0.12 mmol) were
dissolved in dioxane (9 mL). A sample was removed for NMR
analysis. The solution was degassed with N2 for 30 min. The
reaction was then heated to 90 ꢀC and left for 90 min. The
polymerization was quenched in liquid nitrogen and precipi-
tated three times from pentane into THF to give a pink solid,
2.3 g 50% yield. 62% Conversion by NMR. SEC (DMF):
Synthesis of Azido-Monosaccharides
M
w = 15,250 g mol−1, D = 1.7.
2-Chloro-1,3-dimethylimidazolinium
chloride
(2.82
g,
16.7 mmol) was added to a solution of galactose/mannose
(1.00 g, 5.6 mmol), triethylamine (7.7 mL, 55 mmol), and
sodium azide (3.61 g, 55.5 mmol) dissolved in ultrapure Milli-Q
water (20 mL), sitting on ice. The solution was stirred for
40 min on ice before removing the solvent in vacuo. Ethanol
(40 mL) was added to precipitate NaN3, filtered, and the solvent
removed (repeat to ensure complete removal of NaN3). The
resulting solid was then dissolved in ultrapure Milli-Q water
(10 mL) and washed three times with dichloromethane. The
water layer was freeze-dried to give a yellow solid. The product
was then purified on a silica column using 5:1 chloroform: meth-
anol (Rf = 0.3) to give an off-white product. Yield: 0.98 g, 86%.
1H NMR (CDCl3) 400 MHz, ppm: 2.42 (2H, br, CH2) 1.72 (NC-C
(CH3)2-) 1.54 (3H, br, CH3).
19F NMR (CDCl3) 376 MHz, ppm: −150.35 (1F, br s), −151.44
(1F, br s), −156.97 (1F, br s), −162.11 (1F, br s).
Postpolymerization Modification of
Poly(pentafluorophenol methacrylate)
Poly(PFPMA) (0.260 g, 1.0 mmol by side chain) and DBCO-
amine (72 mg, 0.26 mmol) were dissolved in 3 mL DMF and
left at 50 ꢀC overnight under N2. Reaction completion was
confirmed by fluorine NMR, ratio of pentafluorophenol peaks
to polymeric pentafluorophenol ester peaks was 33%. With-
out further workup, a large excess of 2-aminoethan-1-ol
(0.5 mL, 8.3 mmol) was added, and left for a further 16 h at
50 ꢀC. Reaction completion was again confirmed by fluorine
NMR observation of only pentafluorophenol peaks. The reac-
tion was then diluted into ultrapure Milli-Q water and dia-
lyzed for 3 days. Total of 0.10 g of white polymer was
isolated. No fluorine was observed in the NMR of the final
product. DOSEY was carried out to further confirm the conju-
gation of the DBCO unit to the polymer.
1-Azido-1-Deoxy-β-D-Galactose
1H NMR (MeOD) 400 MHz, ppm: 5.57 (1H, d, J1-2 = 4.40 Hz,
H1, α anomer 23.7%), 4.67 (1H, d, J1-2 = 8.68 Hz, H1, β
anomer 76.3%), 3.96 (1H, d, J1-2 = 3.30 Hz, H5), 3.79–3.78
(1H, m, H4), 3.77–3.75 (2H, m, H6), 3.70 (1H, dd J1-2 = 3.42,
J
3-4 = 9.78 Hz, H3) 3.53 (1H, t, J1-2 = 9.78, H2).
13C NMR (MeOD) 100 MHz, ppm: 90.55 (β C1), 89.44 (α C1)
77.21 (β C4), 75.88, 75.13, 74.23, 73.06, 72.64 (β C3), 71.20
70.32 (β C2), 69.19, 68.51(β C5), 68.20, 64.20, 63.46, 61.17,
60.94 (β C6).
1H NMR (MeOD) 500 MHz, ppm: 7.5–7.0 (br, Benzyl) 4.42 (br,
cyclooctyne ring CH2) 3.65 (br, NH-CH2-CH2-), 3.28 (br, -NH-
CH2-CH2-), 2.5–1.5 (br, Backbone CH2) 1.5–1.0 (br, back-
bone Me).
MS (ESI+): Observed: 228.0 Expected: 228.17 [M + Na]+.
IR: 2107 cm−1 (-N3).
DOSEY NMR (MeOD) 500 MHz, log(m2 s−1) = −9.25, ppm:
7.5–7.0, (Benzyl) 4.42 (br, cyclooctyne ring CH2) 3.65 (br, NH-
CH2-CH2-), 3.28 (br, -NH-CH2-CH2-), 2.5–1.5 (br, Backbone
CH2) 1.5–1.0 (br, backbone Me).
1-Azido-1-Deoxy-α-D-Mannose
1H NMR (MeOD) 400 MHz, ppm: 5.46 (1H,d, J1-2 = 1.71 Hz,
alpha 100%) 3.92 (1H, d, J = 10.27 Hz) 3.86 (1H, dd,
J1-2 = 1.96, J3-4 = 3.18 Hz,H2) 3.78, (2H, m) 3.75–3.72 (2H, m),
3.64 (1H, t, J1-2 = 9.54 Hz).
13C NMR (MeOD) 500 MHz: 179.42 (br, C=O) 141.16 (Ar),
136.93 (Ar), 132–128 (Ar), 63.02, 61.31 (NH-CH2-CH2-OH),
61.08 (NH-CH2-CH2-OH), 60.09, 59.61, 56.93, 52.49
(Cyclooctyne Ring CH2), 46.95 (NH-CH2-CH2-C(=O)-DBCO),
13C NMR (MeOD) 100 MHz, ppm: Major Anomer (100%):
90.57 (C1), 78.47 (C5), 70.36 (C2), 70.37 (C3), 66.60 (C4),
61.10(C6).
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2018
3

JOURNAL OF
POLYMER SCIENCE
ORIGINAL ARTICLE
WWW.POLYMERCHEMISTRY.ORG
MS (ESI+): Observed: 228.0 Expected: 228.17 [M + Na]+.
IR: 2107 cm−1 (-N3).
13C NMR (CDCl3) 100 MHz, ppm: 163.04 (C4, C=O) 142.52,
140.67, 140.14, 139.14, 138.16 (C5-10, Aromatics), 133.68
(C3, Me-C (=CH2)-CO)), 129.91(C2, =CH2) 18.18 (C1, −Me).
MS (ESI +): Observed: 253.1 Expected 253.1 [M + H]+.
Synthesis of Pentafluorophenyl Methacrylate
Pentafluorophenol (5.4 g, 29.3 mmol) and lutidine (3.5 mL,
30.0 mmol) were added to a round-bottom flask of dichloro-
methane (50 mL) on ice. Methacroyl chloride (3.0 mL, 31.0 mmol)
was slowly added. The reaction was stirred for 3 h on ice, before
leaving at room temperature overnight. The lutidine HCl precipi-
tate was filtered and the filtrate was washed twice with water
(30 mL), dried with MgSO4, and the solvent removed in vacuo.
The product was then passed through a silica column in petro-
leum ether 40–60 (Rf = 0.3) to give a colorless liquid, Yield 4.7 g.
IR: 1760 cm−1 (ester), 1517 cm−1 (unsaturated C=C),
1086 cm−1 (C-O), 994 cm−1 (C-F).
RESULTS AND DISCUSSION
To access heterogeneous glycopolymers, a three-step postpo-
lymerization modification strategy was developed to enable
the use of a single reactive precursor, which can be sequen-
tially modified ensuring all polymers had identical chain
length distributions, Fig 1.20 PFMA was polymerized by
reversible addition–fragmentation transfer polymerization to
obtain master polymers (Tables 1 and 2). The observed dis-
persities were higher than expected due to the elution behav-
ior of the PPFMA in SEC, but the use of a “masterbatch”
means the actual dispersity is not crucial here as each
1H NMR (CDCl3) 400 MHz, ppm: 6.37 (1H, s), 5.83(1H, s), 2.01
(3H, s).
19F NMR (CDCl3) 376 MHz, ppm: −152.89 (2F, dd,
J1-2 = 16.35 Hz, J3-4 = 6.81 Hz) -158.34 (1F, td, J1-2 = 21.80,
J
3-4 = 9.53 Hz) -162.63 (2F, m).
FIGURE 1 Synthesis of heterogeneous and homogenous glycopolymers. (i) DBCO (0.3 equiv/dioxane); (ii) ethanolamine (3 equiv/
dioxane); and (iii) sugar-N3 (2.5 equiv/DMF). DMC = 2-chloro-1,3-dimethylimidazolinium chloride. [Color figure can be viewed at
wileyonlinelibrary.com]
4
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2018

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ORIGINAL ARTICLE
TABLE 1 Heterogeneous Glycopolymers
Polymer
M
n (g mol−1
)
Mw/Mna (−)
Galb (%)
Galc (−)
Mand (−)
PPFPMA-1
PG25M75
PG50M50
PG75M25
PG100
8900a
1.7
–
–
–
–
12,900e
12,900e
12,900e
12,900e
25
50
75
100
4
11
7.5
4
–
7.5
11
15
–
–
0
c
PGxMy
refers
to
(galactose-functionalized
polymer)x
Average number of galactose residues/chain.
Average number of mannose residues/chain.
Calculated from PPFMA precursor using conversion values.
d
e
%(mannose-functionalized polymer)y%
.
a
From SEC in DMF.
b
Percentage of repeat units with galactose estimated from 1H NMR.
polymer in the series will have the same distirbution.31 19F
NMR and IR spectroscopy confirmed that the PFP group was
retained during polymerization (Supporting Information).
our previous investigations using glycosylated gold nanoparti-
cles revealed nonlinear behavior to galactose/mannose gradi-
ents in aggregation-based assays, suggested mannose could
enhance the overall avidity/binding.35 Fluorescently labeled
lectins were incubated with each polymer, before being
applied to a galactose-rich (GM1) microliter (96 well) plate.
To enable introduction of the glycans by strain promoted click,
DBCO-amine was incorporated into PPFMA-1 at 30 mol % fol-
lowed by excess ethanolamine in a one-pot, two-stage reaction,
to generate a hydrophilic, copper-free “clickable” template
polymer.32,33 Cyclooctyne incorporation was confirmed by 19F
NMR, IR, and Raman spectroscopy as well as 1H diffusion
ordered spectroscopy NMR (Supporting Information). With
this reactive precursor to hand, 1-azido-1-deoxy-D-galactose
(GalN3) and 1-azido-1-deoxy-D-mannose (ManN3) were synthe-
sized using 2-chloro-1,3-dimethylimidazolinium chloride in
a one-step procedure from the free monosaccharides.34 It
should be noted that as the DBCO unit is not symmetric, two
possible regioisomers can be formed. To obtain heterogeneous,
galactose-rich polymers GalN3 and ManN3 were mixed in the
indicated ratios (Table 1) and then applied to the “clickable”
precursor polymer such that the overall ratio [alkyne]:[N3]
was 1:2.5. Excess glycosyl azide was removed by dialysis
and the polymers isolated by freeze-drying. For the lectin
binding experiments (vide infra) it was necessary to have con-
trols of homogeneous galactose-only polymers at the same
overall galactose density. PPFMA-2 was modified in the same
manner, but with the DBCO content varied to enable the same
overall density of galactose to be introduced as in the hetero-
geneous library following click conjugation, Table 2.
After incubation and washing the total fluorescence was mea-
sured. Less plate-associated fluorescence means more toxin
inhibition. Example inhibitory curves for each set of polymers
are shown in Figure 2(B,C). For the heterogeneous polymers,
there is a clear dose-dependent response with increasing (poly-
mer) leading to less lectin binding, demonstrating these are
potent inhibitors of RCA120. Using galactose alone, no inhibition
was possible without resorting to very high concentrations,
demonstrating the cluster-glycoside enhancement. Interestingly,
PGal100 (which contains no mannose) which has the highest
number of expected binding units failed to fully inhibit
RCA120 binding in the concentration range tested. In contrast,
PGal75Man25 was the most active inhibitor. Polymers with inter-
mediate Gal/Man ratios showed intermediate activity, highlight-
ing the nonlinear effect between glycan densities and function.
Glycan density has been shown to be a key modulator of lectin
affinity,36 so to rule out galactose density leading to the above
observations (rather than the mannose units), galactose
homopolymers of variable density were also tested, Figure 2
(B). The homogenous polymers had similar EC50 values (point
where activity reaches 50% of their maximum activity) to
what is seen for heterogeneous counterparts. However, the
total level of inhibition (i.e., how much RCA120 was stopped
from binding) was significantly less than was observed for the
heterogeneous polymers. A comparison is shown in Figure 2
(C) showing the total inhibition at the highest concentration
used. This clearly demonstrates that at all galactose composi-
tions, addition of mannose units resulted in significantly more
toxin inhibition, with the heterogeneous polymers appearing
to “capture” more toxin. The homopolymers all gave <50%
maximal inhibition, demonstrating them to be rather poor
inhibitors. These results may suggest affinity or a secondary
binding site for RCA120 and Man which could lead to this
enhancement, and explain why the IC50 of the polymers were
similar in terms of galactose density. It is important to note
that these glycopolymers had relatively poor solubility (see
With this panel of low density, heterogeneous/homogeneous
glycopolymers to hand, their function could be tested. RCA120
was selected as it is a substitute for the highly toxic ricin, and
TABLE 2 Homogenous Glycopolymers
Polymer
Mna (g mol−1
)
Mw/Mna (−)
Galb (%)
Galc (−)
PPFMA-2
PG25
8700a
1.6
–
0
0
25
50
75
4.9
8.0
10.7
PG50
–
PG75
–
a
From SEC.
b
c
Percentage of repeat units with galactose estimated from 1H NMR.
Total number of galactose residues/chain.
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2018
5

JOURNAL OF
POLYMER SCIENCE
ORIGINAL ARTICLE
WWW.POLYMERCHEMISTRY.ORG
units to simply ensure the binding units are being presented in
the correct three-dimensional location.37 To probe the binding
process in detail BLI was employed; a method related to sur-
face plasmon resonance enabling label-free protein binding to
be evaluated.38
RCA120 was immobilized onto BLI sensors by NHS coupling and
the polymer library was probed for binding as a function of
concentration [Fig. 3(A)]. Figure 3(B) shows an example BLI
binding curve for PGal25Man75 showing the characteristic asso-
ciation and dissociation phases and a clear dose-dependent
increase in binding. Estimating the affinity (dissociation con-
stant, Kd) of multivalent systems is challenging due to there
being a suite of individual interactions in any binding curve.
Therefore, a steady-state approximation was used which is
widely employed to enable comparisons to be made between
systems.39,40 The extracted Kd values are shown in terms of
both galactose concentration [Fig. 3(C)] and polymer concentra-
tion [Fig. 3(D)] enabling separation of the contributing factors.
On a per-polymer basis, reducing the galactose content lead to
lower Kd (higher affinity) with a sixfold difference between
100 and 25% galactose. This result is interesting as the most
active polymers only had four to five galactose residues per
chain, showing only a minimum number are required for sig-
nificant cluster glycoside enhancements (while acknowledging
these are not the most active inhibitors, as the aim is to probe
heterogeneity effects). Analyzing the above data on a per-
galactose basis shows the increase in affinity per galactose
unit is twofold between 100 and 25% density. It was
observed that PGal75Man25 (which has 11 galactose chain−1
)
had the highest affinity per sugar, with significant enhance-
ments over the homogenous system. This is in line with the
inhibitory data showing that 75% galactose was the most
potent in terms of EC50 and total inhibition achieved. This
rational for optimal binding agrees with previous reports that
~30% mannose is optimal for glyconanoparticles agglutinat-
27,35
ing RCA120
.
These observations could support an addi-
tional binding mode for mannose, and that binding is not only
due to the galactose ligands. In understanding these results,
chain flexibility differences seem to be unlikely to contribute
heterogeneous polymers have the same number and distribu-
tion of glycan side chains at all galactose densities due to the
postpolymerization strategy used, but the aromatic linker with
either β or α linkage could lead to stacking. By limiting the
degrees of freedom available to the glycan with the rigid
linker, we may be reducing the entropic penalty required for
binding overall. The overall affinities in this system
(~100 mM) are relatively weak for a glycopolymer, with
nanomolar affinities often reported.16 Spain and Cameron
have shown (by Surface plasmon resonance (SPR)) that
increasing valency of homogalactose polymers leads to
increased affinity to RCA120 which at first seems in contradic-
tion to what is seen here, but the lower density and linker
effects make comparisons difficult.39
FIGURE
2 Competitive inhibition of glycopolymers against
RCA120. (A) Inhibition curves for heterogeneous glycopolymers;
(B) inhibition curves for homogeneous glycopolymers; and
(C) comparison of total maximum inhibition achieved. Inhibitory
experiments were average from 18 repeats and error bars are
ꢁSD. [Color figure can be viewed at wileyonlinelibrary.com]
Supporting Information), but there was no evidence of aggre-
gation or self-assembly, which would impact the binding data.
The increased inhibitory activity upon including mannose might
seem counterintuitive as mannose is not a strong binder of
RCA120 on its own. However, in nature, native glycan ligands
are often branched carbohydrates enabling presentation of mul-
tiple monosaccharides which undergo separate (or allosteric)
interactions26 or the glycans also act as structure directing
As a final test, the heterogeneous polymers were also evalu-
ated for binding to CTxB (Cholera toxin B subunit) using BLI
6
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2018

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ORIGINAL ARTICLE
RCA120 (and to cholera toxin) was enhanced by addition of
second glycan with a 3:1 ratio of galactose to mannose having
the highest activity. BLI studies enabled estimation of the
overall affinity, and in all cases lower galactose lead to lower
dissociation constants (higher affinity) even though the overall
affinity was rather low for a multivalent system, potentially
due to the sterically large linker. Similar results were seen for
another galactose binding lectin, the cholera toxin B subunit.
This shows that total inhibitory activity can be modulated by
using increasingly complex heterogeneous glycopolymers and
that mixing distinct glycans together could give rise to unique
binding modes to identify more active multivalent ligands,
especially in the context of high-throughput screening.
ACKNOWLEDGMENTS
M. I. Gibson acknowledges the ERC for a Starting Grant, CRYO-
MAT 638661. B. Martyn thanks EPSRC (EP/M506679/1) and
UoW for a studentship and the polymer characterization
Research Technology Platform (RTP) for polymer analysis.
DATA ACCESS STATEMENT
The research data supporting this publication can be accessed
at http://wrap.warwick.ac.uk.
REFERENCES AND NOTES
1 T. R. Branson, W. B. Turnbull, Chem. Soc. Rev. 2013, 42, 4613.
2 G. E. Soto, S. J. Hultgren, J. Bacteriol. 1999, 181, 1059.
3 J. D. Esko, N. Sharon, Microbial Lectins: Hemagglutinins,
Adhesins, and Toxins; Cold Spring Harbor Laboratory Press:
Cold Spring Harbor, 2009.
4 C. M. Thorpe, Clin. Infect. Dis. 2004, 38, 1298.
5 C. D. Hébert, Toxic. Rep. Ser. 1993, 37, 1.
6 L. A. Lasky, Annu. Rev. Biochem. 1995, 64, 113.
7 P. M. Rudd, T. Elliott, P. Cresswell, I. A. Wilson, R. A. Dwek,
Science 2001, 291, 2370.
FIGURE 3 BLI analysis of glycopolymers binding to RCA120
.
(A) Example BLI curves for serial dilution (4–64 μM) of
PGal25Man75; (B) KD from 1:1 binding approximation in terms
galactose concentration; and (C) polymer concentration. *Fitting
parameter did not enable error to be extracted. [Color figure can
be viewed at wileyonlinelibrary.com]
8 R. O. Hynes, Cell 1992, 69, 11.
9 M. Ambrosi, N. R. Cameron, B. G. Davis, Org. Biomol. Chem.
2005, 3, 1593.
10 T. K. Dam, C. F. Brewer, Biochemistry 2008, 47, 8470.
11 J. J. Lundquist, E. J. Toone, Chem. Rev. 2002, 102, 555.
12 A. Bernardi, J. Jiménez-Barbero, A. Casnati, C. De Castro,
T. Darbre, F. Fieschi, J. Finne, H. Funken, K.-E. Jaeger,
M. Lahmann, T. K. Lindhorst, M. Marradi, P. Messner,
A. Molinaro, P. V. Murphy, C. Nativi, S. Oscarson, S. Penadés,
F. Peri, R. J. Pieters, O. Renaudet, J.-L. Reymond, B. Richichi,
J. Rojo, F. Sansone, C. Schäffer, W. B. Turnbull, T. Velasco-
Torrijos, S. Vidal, S. Vincent, T. Wennekes, H. Zuilhof,
A. Imberty, Chem. Soc. Rev. 2013, 42, 4709.
(Supporting Information). In line with the RCA120 results, het-
erogeneous glycopolymers resulted in increased affinity on a
per polymer and per galactose basis. Due to the very high
affinity of CTx for the GM1 surfaces complementary inhibitor
experiments showed limited activity (Supporting Information),
which may support that heterogeneity effects dominate in the
lower affinity range, which has been probed in both lectins
used here.
13 S. Won, S.-J. Richards, M. Walker, M. I. Gibson, Nanoscale
Horiz. 2017, 3, 1593.
14 N. Sharon, Biochim. Biophys. Acta-Gen. Subj. 1760,
2006, 527.
CONCLUSIONS
15 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.
In conclusion, a series of heterogeneous glycopolymers with
varying ratios of galactose:mannose were synthesized using a
three-step postpolymerization modification strategy enabling
heterogeneity effects to be separated from density effects. It
was found that the inhibitory potency of the polymers against
16 C. R. Becer, M. I. Gibson, J. Geng, R. Ilyas, R. Wallis,
D. A. Mitchell, D. M. Haddleton, J. Am. Chem. Soc. 2010, 132,
15130.
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2018
7

JOURNAL OF
POLYMER SCIENCE
ORIGINAL ARTICLE
WWW.POLYMERCHEMISTRY.ORG
17 A. Muñoz, D. Sigwalt, B. M. Illescas, J. Luczkowiak,
L. Rodríguez-Pérez, I. Nierengarten, M. Holler, J.-S. Remy,
K. Buffet, S. P. Vincent, J. Rojo, R. Delgado, J.-F. Nierengarten,
N. Martín, Nat. Chem. 2015, 8, 50.
K. Kantner, G. Khadem-Saba, J.-M. Montenegro, J. Rejman,
T. Rojo, I. R. de Larramendi, R. Ufartes, A. Wenk, W. J. Parak,
Nat. Nanotechnol. 2015, 10, 619.
29 D. Ponader, P. Maffre, J. Aretz, D. Pussak, N. M. Ninnemann,
S. Schmidt, P. H. Seeberger, C. Rademacher, G. U. Nienhaus,
L. Hartmann, J. Am. Chem. Soc. 2014, 136, 2008.
18 M. L. Huang, M. Cohen, C. J. Fisher, R. T. Schooley,
P. Gagneux, K. Godula, Chem. Commun. 2015, 51, 5326.
30 L. Wu, Y. Zhang, Z. Li, G. Yang, Z. Kochovski, G. Chen,
M. Jiang, J. Am. Chem. Soc. 2017, 139, 14684.
19 B. D. Polizzotti, K. L. Kiick, Biomacromolecules 2006, 7, 483.
20 S.-J. Richards, M. W. Jones, M. Hunaban, D. M. Haddleton,
M. I. Gibson, Angew. Chem. Int. Ed. 2012, 51, 7812.
31 M. I. Gibson, E. Fröhlich, H.-A. Klok, J. Polym Sci. Part A:
Polym. Chem. 2009, 47, 4332.
21 M. W. Jones, L. Otten, S.-J. Richards, R. Lowery, D. J. Phillips,
D. M. Haddleton, M. I. Gibson, Chem. Sci. 2014, 5, 1611.
32 N. J. Agard, J. A. Prescher, C. R. Bertozzi, J. Am. Chem. Soc.
2004, 126, 15046.
22 K. W. Moremen, M. Tiemeyer, A. V. Nairn, Nat. Rev. Mol. Cell
Biol. 2012, 13, 448.
33 S. Agar, E. Baysak, G. Hizal, U. Tunca, H. Durmaz, J. Polym.
Sci. Part A: Polym. Chem. 2018, 56, 1181.
23 M. Ortega-Muñoz, F. Perez-Balderas, J. Morales-Sanfrutos,
F. Hernandez-Mateo, J. Isac-Garcia, F. Santoyo-Gonzalez, Eur.
J. Org. Chem. 2009, 2009, 2454.
34 N. Vinson, Y. Gou, C. R. Becer, D. M. Haddleton, M. I. Gibson,
Polym. Chem. 2011, 2, 107.
35 L. Otten, D. Vlachou, S.-J. Richards, M. I. Gibson, Analyst
2016, 141, 4305.
24 M. Gómez-García, J. M. Benito, D. Rodríguez-Lucena,
J. X. Yu, K. Chmurski, C. Ortiz Mellet, R. Gutiérrez Gallego,
A. Maestre, J. Defaye, J. M. García Fernández, J. Am. Chem.
Soc. 2005, 127, 7970.
36 K. Godula, C. R. Bertozzi, J. Am. Chem. Soc. 2012, 134,
15732.
25 S. Zhang, Q. Xiao, S. E. Sherman, A. Muncan,
A. D. M. Ramos Vicente, Z. Wang, D. A. Hammer,
D. Williams, Y. Chen, D. J. Pochan, S. Vértesy, S. André,
M. L. Klein, H. J. Gabius, V. Percec, J. Am. Chem. Soc. 2015,
137, 13334.
37 Z. Wang, Z. S. Chinoy, S. G. Ambre, W. Peng, R. McBride,
R. P. de Vries, J. Glushka, J. C. Paulson, G.-J. Boons, Science
2013, 341, 379.
38 G. Volkers, L. J. Worrall, D. H. Kwan, C.-C. Yu, L. Baumann,
E. Lameignere, G. A. Wasney, N. E. Scott, W. W. Wakarchuk,
L. J. Foster, S. G. Withers, N. C. J. Strynadka, Nat. Struct. Mol.
Biol. 2015, 22, 627.
26 N. C. Worstell, P. Krishnan, J. D. Weatherston, H. J. Wu, PLoS
One 2016, 11, e0153265.
27 S.-J. Richards, L. Otten, M. I. Gibson, J. Mater. Chem. B 2016,
4, 3046.
39 S. G. Spain, N. R. Cameron, Polym. Chem. 2011, 2, 1552.
40 S.-G. Lee, J. M. Brown, C. J. Rogers, J. B. Matson,
C. Krishnamurthy, M. Rawat, L. C. Hsieh-Wilson, Chem. Sci.
2010, 1, 322.
28 W. G. Kreyling, A. M. Abdelmonem, Z. Ali, F. Alves,
M. Geiser, N. Haberl, R. Hartmann, S. Hirn, D. J. de Aberasturi,
8
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2018