Bifunctionalized dextrans for surface PEGylation via multivalent host-guest interactions

Article abstract of DOI:10.1016/j.carbpol.2015.07.027

The main goal of this work was to develop a supramolecular chemistry strategy to decorate interfaces with polyethylene glycol (PEG) grafts. A series of novel bifunctionalized dextrans, bearing 40-60 PEG pending chains and 12-24 hydrophobic adamantyl group

Full text of DOI:10.1016/j.carbpol.2015.07.027

Carbohydrate Polymers 133 (2015) 473–481
Contents lists available at ScienceDirect
Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol
Bifunctionalized dextrans for surface PEGylation via multivalent
host–guest interactions
Iurii Antoniuka, Véronique Wintgensa, Gisèle Voleta, Thorbjørn T. Nielsenb,
Catherine Amiela,∗
a University Paris Est, ICMPE (UMR 7182), CNRS, UPEC, F-94320 Thiais, France
b Department of Chemistry and Bioscience, Aalborg University, Fredrik Bajers Vej 7H, 9220 Aalborg Ø, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
The main goal of this work was to develop a supramolecular chemistry strategy to decorate interfaces with
polyethylene glycol (PEG) grafts. A series of novel bifunctionalized dextrans, bearing 40–60 PEG pending
chains and 12–24 hydrophobic adamantyl groups, have been prepared by copper(I)-catalyzed azide-
alkyne cycloaddition. Their binding properties toward native ␤CD and ␤CD polymers were characterized
both in solution and at interface using isothermal titration microcalorimetry and surface plasmon reso-
nance. The polymers were found to form stable layers onto neutral and positively charged ␤CD-polymer
films pre-adsorbed on gold surfaces, due to multivalent interpolymer host–guest interactions. The thick-
ness and stability of the layers could be tuned by varying the ratio between the degrees of substitution
by PEG and adamantyl groups.
Received 29 April 2015
Received in revised form 15 June 2015
Accepted 8 July 2015
Available online 22 July 2015
Keywords:
Modified dextrans
Cyclodextrin polymers
Supramolecular association
Surface films
© 2015 Elsevier Ltd. All rights reserved.
PEGylation
1. Introduction
achieved in various ways, including preparation of PEG-containing
block copolymers for polymeric micelles formation (Gaucher et al.,
Two key requirements for polymeric nanocarriers to be suc-
cessfully applied in controlled delivery and release are high
sterical stability and robust protection from the uptake by the
blood mononuclear phagocyte system (MPS) (Gaucher, Dufresne,
Sant, Maysinger, & Leroux, 2005; Li & Huang, 2008; Vonarbourg,
Passirani, Saulnier, & Benoit, 2006). One of the preferred and well-
established ways to meet these consists in decorating the surface
of nanoparticles with polyethylene glycol (PEG), often referred
as PEGylation (Gaucher et al., 2005). The presence of hydrophilic
and highly flexible PEG chains in contact with physiological media
proved to significantly reduce the opsonin proteins adsorption,
thereby increasing the average blood circulation times of nanocar-
riers (Immordino, Dosio, & Cattel, 2006; Jokerst, Lobovkina, Zare,
& Gambhir, 2011; Loira-Pastoriza, Todoroff, & Vanbever, 2014).
Besides providing steric stabilization and stealth properties, PEG
“corona” usually leads to higher apparent sizes, thus increasing
the efficiency of a nanocarrier administration due to the enhanced
permeability and retention effect (EPR). The PEGylation might be
2005), covalent modification (Roberts, Bentley, & Harris, 2002;
Yoncheva, Lizarraga, & Irache, 2005) or physical adsorption by
“host–guest” interactions (David, Millot, Sebille, & Levy, 2003; Pun
& Davis, 2002; Volet & Amiel, 2012).
Cyclodextrins (CD) are cyclic (␣-1,4)-linked oligosaccharides
consisting of six, seven or eight ␣-d-glucopyranose in the case of
the most common ␣-, ␤- and ␥-cyclodextrins respectively. They
are characterized by toroidal shape with hydrophilic exterior and
hydrophobic internal cavity, which allows them to form inclu-
sion complexes with various hydrophobic substrates (Loftsson &
Brewster, 1996). This property has led to the widespread use
of cyclodextrin-based materials in biomedical and personal care
industry as excipients, drug carriers and solubility enhancers
(Brewster & Loftsson, 2007; Hedges, 1998; van de Manakker,
Vermonden, van Nostrum, & Hennink, 2009). During the last two
decades a high number of researches were aiming at diminishing
the phenomenon of initial “burst release” of actives from nanoscale
drug delivery carriers (Wu & Brazel, 2008). Along with superficial
cross-linking, adsorption or impregnation of drug carriers with CD-
containing materials was found to be an efficient strategy to address
this issue. For instance, it was reported that co-encapsulation of
hydroxypropyl-␤CD into poly(lactide-co-glycolide) (PLGA) micro-
∗
Corresponding author.
E-mail addresses: antoniuk@icmpe.cnrs.fr (I. Antoniuk), wintgens@icmpe.cnrs.fr
(V. Wintgens), volet@icmpe.cnrs.fr (G. Volet), ttn@bio.aau.dk (T.T. Nielsen),
amiel@icmpe.cnrs.fr (C. Amiel).
spheres led to
a significant decrease of the initial “burst”
of loaded insulin (Quaglia et al., 2003); benzophenone-loaded
http://dx.doi.org/10.1016/j.carbpol.2015.07.027
0144-8617/© 2015 Elsevier Ltd. All rights reserved.

474
I. Antoniuk et al. / Carbohydrate Polymers 133 (2015) 473–481
polylactic acid (PLA) nanoparticles were covered with neutral
poly-␤CD and shown to have improved kinetic profile of the ben-
zophenone release, governed by its “host–guest” interactions with
␤CDs in the shell (El Fagui & Amiel, 2012). In a more recent exam-
ple membrane fusion between liposomes and cyclodextrin vesicles
was used to prepare hybrid assemblies comprised of liposomal
core covered with layers of ␤CD amphiphiles (Versluis et al., 2014).
Another way for introducing ␤CD in nanoscale drug carriers con-
sists in assembling ␤CD polymers with “guest”-labeled amphiphilic
copolymers (Ren, Chen, & Jiang, 2009; Wintgens, Nielsen, Larsen, &
Amiel, 2011).
In this work we present a possible path for combination of
the benefits of PEGylation and ␤CD-containing materials. Although
several works have already reported the host–guest anchoring
of PEG chains onto ␤-cyclodextrin containing surfaces (David
et al., 2003), to the best of our knowledge, only few studies of
the multivalence effect, when multiple PEPO (Pluronics) grafts
and guest groups are attached to a polymer backbone, have
been done so far (Wintgens, Layre, Hourdet, & Amiel, 2012). We
report a synthesis of novel (PEG, Ada)-grafted dextrans, prepared
by copper-catalyzed azide-alkyne cycloaddition (CuAAC). These
bifunctionalized “guest” polymers are bearing various amounts of
5000 g/mol PEG chains and hydrophobic adamantyl (Ada) groups
at the same time. The ability of (PEG, Ada)-grafted dextrans to form
inclusion complexes in solution and on surface with native ␤CD
and ␤CD polymers are investigated with the purpose of their fur-
ther use for facile PEGylation of interfaces and nanoparticles via
multivalent host–guest interactions.
reactor was programmed to maintain a constant temperature by
adjusting the applied power; the desired temperature was reached
gradually using an initial 3 min temperature ramp. After the end
of the reaction, solutions were cooled to room temperature with
compressed air. When necessary, ultrafiltration was carried out
under N2 atmosphere using Amicon equipment and regenerated
cellulose membranes with MWCO 30,000. Dialysis was done with
Spectra/Por molecular porous tubing MWCO 6000–8000. The ultra-
filtrated and dialyzed solutions were freeze-dried on CHRIST ALPHA
1–2 LD Plus lyophilizator.
Short names of the polymers were attributed using the follow-
ing pattern: D40-XGP-YAda-ZPEG with X, Y and Z indicating the
molar degree of substitution (DS) of anhydroglucosidic rings (fur-
ther AHG) by different substituents.
2.2.1. Ada-N3
2.2.1.1. 1-Adamantaneethyltosylate. 1-Adamantaneethanol
(4.80 g, 26.6 mmol) was dissolved in 25 mL of pyridine, cooled to
0 ◦C on ice bath and p-toluenesulfonyl chloride (8.37 g, 43.9 mmol)
was added to the solution. The reaction mixture was allowed to
warm and stirred overnight at room temperature. The solvent
was evaporated under reduced pressure and the product was
redissolved in 100 mL of EtOAc. The organic layer was washed once
with saturated NaHCO3 and twice with brine, dried over Na2SO4,
filtered and evaporated under reduced pressure. The subsequent
filtration through
a silica pad (40 g, EtOAc/n-heptane = 1/25)
yielded the final product as a white solid. Thin layer chromatogra-
phy (TLC) was carried out using Silica gel 60 F254 TLC plates with
0.2 mm silica gel (EtOAc/heptane = 1/25). Yield: 89.4% (7.96 g). 1H
NMR (DMSO-d6) ı = 1.30–1.40 (m, 6H + 2H, Ada overlapped with
Ada-CH2), 1.45–1.70 (m, 6H, Ada), 1.90 (s, 3H, Ada), 2.42 (s, 3H,
CH3), 4.03 (t, 2H, CH2-OTs), 7.49 (d, 2H, Ts), 7.78 (d, 2H, Ts).
2. Materials and methods
2.1. Materials
Pyridine (99.9%, anhydrous), N,Nꢀ-dimethylformamide (DMF,
99.8%, anhydrous), tetrakis(acetonitrile)copper(I)tetrafluoroborate
(Cu(MeCN)4PF6, 97%), thionyl chloride (SOCl2, 99+%), 1-
adamantaneethanol (98%) and glycidyl propargyl ether (GPE,
90+%) were purchased from Sigma–Aldrich (Saint Quentin
Fallavier, France) and used as received. p-Toluenesulfonyl chloride
(98%), sodium azide (99%), l-ascorbic acid sodium salt (Na-Asc,
99%) and dimethyl sulfoxide (DMSO, 99.8%, anhydrous) were pur-
chased from Alfa Aesar and used as received. Sodium hydroxide
(98+%, anhydrous pellets) were purchased from Acros Organics
and used as received.
2.2.1.2. 1-Adamantaneethylazide (Ada-N3). 1-Adamantaneethyl-
tosylate (7.96 g, 23.8 mmol) was dissolved in 0.5 mol/L NaN3
solution in DMSO (167 mL, 83.3 mmol) and stirred at 80 ◦C for 50 h.
250 mL of water were added to quench the reaction, and after
cooling down 800 mL of EtOAc were added. The organic phase
was washed 3 times with water and one time with brine, dried
over Na2SO4 and filtered. Concentration under reduced pressure
and drying under vacuum gave yellowish oil. Yield: 74.6% (3.64 g).
The IR spectrum of the product showed a single N3 peak at
2094 cm−1 1H NMR (CDCl3) ı = 1.38 (t, 2H, Ada-CH2), 1.48 (s, 6H,
.
Ada), 1.57–1.72 (m, 6H, Ada), 1.94 (s, 3H, Ada), 3.24 (t, 2H, CH2N3).
Dextran (DT40, Mw = 4.3 · 104 g/mol, PDI = 1.5, Amersham Phar-
macia Sweden), lithium chloride (Alfa Aesar) and poly(ethylene
glycol) methyl ether of 5000 Da (MeOPEG-OH, Sigma–Aldrich)
were dried overnight in vacuum at 90 ◦C prior to use.
2.2.2. MeOPEG-N3
2.2.2.1. MeOPEG-Cl. MeOPEG-OH (10 g, 2 mmol) of 5000 g/mol was
dissolved in 15 mL (206.5 mmol) of SOCl2; the reaction mixture was
heated to 85 ◦C and refluxed for 40 h. Excess SOCl2 was removed
by bubbling argon at 65 ◦C. Then 100 mL of DCM were added and
the obtained solution was precipitated in 1 L of Et2O, filtered and
washed with Et2O. The subsequent drying in vacuum at 40 ◦C
yielded the product as a white solid. Yield: 94.6% (9.51 g).
Tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine
(TBTA)
(Chan, Hilgraf, Sharpless, & Fokin, 2004) was synthesized accord-
ing to the literature data. Neutral epichlorohydrin-branched
p␤CD polymer and positively charged p␤CDN+, modified with
quaternary ammonia, were prepared according to the previously
described procedures (Blomberg, Kumpulainen, David, & Amiel,
2004; Renard, Deratani, Volet, & Sebille, 1997).
2.2.2.2. MeOPEG-N3. MeOPEG-Cl (9.40 g, 1.87 mmol) was dissolved
in 100 mL of anhydrous DMF and treated with NaN3 (3.30 g,
50 mmol). The reaction mixture was stirred at 80 ◦C for 60 h, pre-
cipitated in 1 L of Et2O and filtered. The product was subsequently
purified by dissolving it in 100 mL dichloromethane, washing with
200 mL of water, concentrating the organic layer to 50 mL under
reduced pressure and its precipitation in 600 mL of 2-propanol.
After filtration and drying in vacuum at 55 ◦C, the final product was
obtained as white flakes. Yield: 70.9% (6.68 g). The IR of the product
confirmed the absence of residual NaN3, showing a single N3 peak
2.2. Synthesis
The syntheses under microwave irradiation (MW) were per-
formed using a Monowave-300 reactor from Anton Paar. The
microwave source was a magnetron with a 2.5 GHz frequency pow-
ered by a 900 W power generator, which could be operated at
different power levels. The reaction mixtures were loaded inside
quartz glass tubes; the tubes were closed with silicon caps using a
pressure monitor unit, and placed into the reactor. The microwave

I. Antoniuk et al. / Carbohydrate Polymers 133 (2015) 473–481
475
at 2117 cm−1 1H NMR (D2O) ı = 3.38 (s, 3H, OCH3), 3.45–3.95 (m,
.
PEG chain protons).
each injection of a host solution as a function of time. Integration of
the heat flow peaks by the instrument software (after taking into
account heat of dilution) provides the amount of heat produced
per injection. The experimental data were fitted with a theoret-
ical titration curve using the instrument software with a model
assuming a 1:1 stoichiometry for the adamantyl/␤CD complex. The
enthalpy change, |ꢀH|, the association constant, Ka, and the overall
stoichiometry were adjustable parameters.
2.2.3. D40-18GP-18PEG
D40-18GP (0.150 g, 0.155 mmol alkynes) and MeOPEG-N3
(1.32 g, 0.265 mmol) were dissolved in 16 mL of DMSO by warm-
ing at 50 ◦C. Subsequently TBTA (0.54 mL, 0.10 M in DMSO) and
Na-Asc (0.17 mL, 0.28 M in water) were added to the solution, fol-
lowed by bubbling with argon for 10 min. Finally, Cu(MeCN)4PF6
(23 mg, 0.062 mmol) was added to the reaction mixture, it was
bubbled with argon for another 5 min and azide-alkyne cycloaddi-
tion (CuAAC) was performed under microwave irradiation (85 ◦C,
40 min). The reaction mixture was dialyzed against DMSO for
20 h, diluted with 250 mL of water and ultrafiltrated (MWCO
30,000). The product was obtained by freeze-drying as white solid
(0.58 g). 1H NMR (DMSO-d6) ı = 2.90–3.90 (m, dextran glucosidic
and PEG protons, overlapped with HDO), 4.40–5.20 (m, 4H, dextran
hydroxyl + anomeric), 8.06 (s, 1H, triazole).
2.3.3. Surface plasmon resonance (SPR)
SPR measurements were carried out with a Spreeta SPR-EVM-
BT from Texas Instruments. The light of an infrared light-emitting
diode (LED; ꢁ = 840 nm) reaches the sensor surface at a range of
angles above the critical angle. A reflectivity-versus-angle spec-
trum is obtained, and an apparent refractive index (RI) is derived
from it by the Spreeta software. The sensor surface is a gold layer
(ca. 50 nm). In the case of adsorption on the golden surface the
RI changes provide information on the adsorbed layer parameters.
The sensor is integrated in a flow cell and the measurements were
carried out at 23 ◦C using a continuous water flow (3 mL/h) which
was delivered by a syringe pump (Kd Scientific), and a Rheodyne
injection valve to switch to the sample solution.
The sensor surface was cleaned with a 4% CrO3 solution in water
and rinsed with MeOH before each experiment. After settling the
sensor into the flow cell, an in situ cleaning was performed with one
injection (0.1 mL) of a solution of 0.1 M NaOH and 1% Triton X-100
in water followed by 3 injections (0.1 mL) of 10% EtOH in water.
The sensor was calibrated in pure water, assuming a RI value of
1.33300, and the RI variations after the different polymer injections
were monitored as a function of time.
Two types of experiments were run. In the first type, three con-
secutive injections of 0.1 mL of p␤CD in water or p␤CDN+ in 0.5 M
NaCl (both at 2 g/L) were carried out to saturate the gold surface
of the sensor. This was followed by 3–4 consecutive injections
(0.1 mL) of one of the PEG-containing dextrans in water with vari-
ous DS by adamantyls (1 g/L). Then the RI variation due to the 2nd
layer deposition was measured 20 min after the first injection.
In the second type, 1 mL of p␤CDN+ solution at 0.2 g/L was
injected to saturate the gold surface of the sensor; then an injection
of 1 mL of the D40-28GP-10Ada-18PEG solution at various con-
centrations (0.01–0.1 g/L) was performed, and the kinetics of the
adsorption were recorded and analyzed.
2.2.4. D40-28GP-(5,10)Ada-(23,18)PEG
D40-28GP (0.160 g, 0.225 mmol alkyne) and Ada-N3
(0.23–0.46 mL of a solution at 0.17 mol/L in DMSO) were dis-
solved in 17 mL of DMSO. TBTA (0.79 mL of a solution at 0.10 mol/L
in DMSO) and Na-Asc (0.20 mL of a solution at 0.28 mol/L in water)
were added, and the obtained solution was bubbled with argon for
10 min. Then, Cu(MeCN)4PF6 (50 mg, 0.132 mmol) was added, the
reaction mixture was bubbled with argon for 5 min under sonica-
tion, followed by CuAAC, performed under microwave irradiation
(70 ◦C, 20 min). Further, MeOPEG-N3 (1.58–1.25 g, 0.32–0.25 mmol)
was added to the reaction mixture and microwave irradiation
was proceeded (85 ◦C, 40 min). The reaction mixture was dialyzed
against DMSO for 20 h, diluted with 250 mL of water and ultrafil-
trated (MWCO 30,000). The product was obtained by freeze-drying
as white solid (0.69 g). 1H NMR (DMSO-d6) ı = 1.50 (s, 6H, Ada),
1.57–1.72 (m, 6H, Ada), 1.92 (s, 3H, Ada), 2.90–3.90 (m, dextran
glucosidic and PEG protons, overlapped with HDO), 4.40–5.20 (m,
4H, dextran hydroxyl + anomeric), 8.06 (s, 1H, triazole-PEG), 8.10
(s, 1H, triazole-Ada).
2.3. Methods
2.3.1. General methods
From the values of ꢀRI the adsorbed amounts of the polymers
may be estimated based on the following equation (Jung, Campbell,
Chinowsky, Mar, & Yee, 1998):
NMR measurements were performed in D2O or DMSO-d6, unless
other is indicated, on a Bruker Avance II Ultrashield Plus 400 MHz
NMR spectrometer. 1H NMR spectra were calibrated using the
chemical shifts of the residual solvents signals according to the
literature data (Fulmer et al., 2010). Fourier transform infrared
spectroscopy (FT-IR) measurements were made on a Bruker Ten-
sor 27 FT-IR spectrometer. Size exclusion chromatography coupled
to multi-angle laser light scattering (SEC-MALLS) was performed
in deionized water with 0.1 mol/L LiNO3 (0.05% NaN3) on TSK-gel
type SW4000-3000 columns and detection by a Wyatt Dawn 8+
light scattering detector and a Wyatt Optilab Rex refractive index
detector.
ꢀ
ꢁ
ꢂꢃ
−2dpol
ld
ꢀRI = m(npol − ns) 1 − exp
(1)
where ns and npol are refractive indexes of the solvent and the poly-
mer respectively, dpol is the thickness of deposited polymer layer
and ld is the depth of penetration of the evanescent electromag-
netic field (typically 25–50% of the probing light wavelength). In
the present work it was roughly estimated as 37% (310.8 nm) (Jung
et al., 1998; Wintgens & Amiel, 2005). We assumed ld to be constant
upon each subsequent polymer adsorption, given the low density of
highly hydrated layers and their thickness being much lower than
310.8 nm. The calibration coefficient m of the sensor was estimated
at 1.
2.3.2. Isothermal titration microcalorimetry
Isothermal titration microcalorimetry (ITC) measurements
were performed using a MicroCal VPITC microcalorimeter. In each
titration, injections of 10 ␮L of native ␤CD, p␤CD or p␤CDN+
solutions (CˇCD = 5 × 10−3 mol/L) were added from the computer
controlled 295 ␮L microsyringe at an interval of 180 s into the
cell (V = 1.4569 mL) containing the investigated polymer solution
(5 × 10−4 mol/L of Ada-groups) while stirring at 450 rpm. The tem-
perature was fixed at 25 ◦C. The raw experimental data were
obtained as the amount of heat produced per second following
Since dpol ꢁ ld, ꢀRI is directly proportional to the layer thickness
and Eq. (1) may be rewritten as:
ꢁ
ꢂ
2dpol
ꢀRI = m(npol − ns)
(2)
ld

476
I. Antoniuk et al. / Carbohydrate Polymers 133 (2015) 473–481
The deposited polymer being highly swollen with the solvent
an excess of NaN3 in DMF. During the purification stage washing
leads npol − ns which means that:
with water proved to be crucial for the full elimination of unreacted
NaN3.
ꢁ
ꢂ
ꢄ
ꢅ
ꢄ
ꢅ
dn
dc
dn
dc
Qads
dpol
Dextran of 40 kDa was modified with alkynes using the basic
epoxide opening of glycidyl propargyl ether in water, according to
the procedure described in literature (Nielsen, Wintgens, Amiel,
Wimmer, & Larsen, 2010) (Fig. 1). The choice of the ether bond
is justified by its hydrolytic stability and relatively long resulting
spacer between the dextran backbone and grafted alkynes. The lat-
ter proved to be important for the preservation of good inclusion
properties of subsequently grafted host–guest functions. The 1H
NMR spectrum of D40-28GP may be found in Supplementary data
(Fig. S3).
npol − ns = Csurf
=
·
pol
(3)
pol
where Csurf is the surface concentration and Qads is the adsorbed
amount of the polymer; dn/dc is the RI increment of the polymer
with respect to water. In this work it was assumed as 0.137 mL/g
for both p␤CD and p␤CDN+ (Wintgens & Amiel, 2005). Combining
Eqs. (2) and (3) one obtains:
ꢀRI · ld
Qads
=
(4)
2m · (dn/dc)pol
In the next step D40-GP-Ada-PEG with various degrees of sub-
stitution were prepared by one-pot CuAAC (Fig. 1). Pure DMSO as a
solvent and Cu(MeCN)4PF6/Na-Asc as a catalytic system were cho-
sen due to poor solubility of Ada-N3 in water. A Cu(I)-stabilizing
polytriazole ligand TBTA was applied to prevent any undesirable
acetylenic cross-couplings often encountered when dealing with
alkyne polymeric scaffolds (Antoniuk, Volet, Wintgens, & Amiel,
2014; Nielsen et al., 2010; Siemsen, Livingston, & Diederich, 2000).
Typically in the first stage the aimed amount of Ada-N3 was
“clicked” quantitatively; then 1.6–1.8 equivalents of MeOPEG-
N3 were added to the reaction mixture to consume the residual
alkynes. When performed under classical heating, due to the bulk-
iness of MeOPEG-N3, the CuAAC required very long reaction times
(3 days) for the second stage completion. Microwave-assisted heat-
ing was reported to significantly increase the reaction rates of
CuAAC, especially when the reactants are bulky or polymeric units
(Hoogenboom, Moore, & Schubert, 2006; Kappe & Van der Eycken,
2010). Indeed, when the same one-pot procedures were performed
under microwave irradiation, the reaction times reduced dramat-
ically. Full conversion of the residual alkynes by the reaction with
MeOPEG-N3 was typically achieved within 40 min at 85 ◦C.
3. Results and discussion
3.1. Guest and host polymers
3.1.1. D40-XGP-YAda-ZPEG preparation
Copper-catalyzed azide-alkyne cycloaddition (CuAAC) was cho-
sen as an efficient tool to couple both adamantyl and MeOPEG
moieties to the dextran backbone (Dedola, Nepogodiev, & Field,
2007; Hein, Liu, & Wang, 2008). Azide-substituted adamantane
derivatives, azide-terminated MeOPEG and alkyne-substituted
dextrans were chosen as building blocks for CuAAC.
Azide-substituted adamantane derivative (Ada-N3, Fig. S1a in
Supplementary data) was prepared according to the modified 2-
step procedure used by Suzuki et al. (2010). It involves tosylation
of 1-adamantaneethanol in pyridine followed by nucleophilic sub-
stitution with NaN3 in DMSO. It should be noted that the obtained
1-adamantaneethyltosylate was not exposed to tedious column
chromatography purification and after filtration on silica pad it con-
tained some traces of native 1-adamantaneethanol. However it was
not a problem due to its inability to react with NaN3 in DMSO during
the second step.
By the MW-assisted heating approach,
a set of D40-GP-
Ada-PEG with various degrees of substitution by hydrophobic
Ada (0–10 mol%) and by hydrophilic flexible MeOPEG chains
(15–23 mol%) were obtained (Table 1). Whereas the 1H NMR
Azide-terminated MeOPEG (MeOPEG-N3) was synthesized by a
two-step procedure (Fig. S1b). The terminal hydroxyl of PEG was
first converted to chloride with SOCl2 according to a method previ-
ously described (Greenwald, Pendri, & Bolikal, 1995). The chloride
was subsequently converted into azide by SN2 substitution with
spectrum of D40-18GP-18PEG demonstrated
a single triaz-
ole proton signal (Fig. S4), in the case of bifunctionalized
Fig. 1. Reaction scheme for the MW-assisted one-pot CuAAC synthesis of D40-GP-Ada-PEG.

I. Antoniuk et al. / Carbohydrate Polymers 133 (2015) 473–481
477
Table 1
decrease of Ka is observed while increasing the substitution degree
by Ada-groups from 5 to 10 mol%, which could be attributed to a
gradual loss of their accessibility due to steric effects.
Characteristics of D40-XGP-YAda-ZPEG (guest) and p␤CD (host) polymers.
Guest polymers
GP, mol%
Ada, mol%
PEG, mol%
In the next step we applied ITC to study the interpolymer inter-
actions. The concentration of adamantyl groups in the solutions
of D40-GP-Ada-PEGs was fixed at 5 × 10−4 mol/L. The solution of
D40-18GP-18PEG, bearing no adamantyl groups, was used at the
mass concentration equal to that of D40-28GP-10Ada-18PEG with
CAda = 5 × 10−4 mol/L. The enthalpograms, shown in Fig. 2b and c
are plotted as a function of the ␤CD/Ada molar ratio. Except for
the case of D40-18GP-18PEG (free of adamantyl groups) where no
interaction could be detected (black circles in Fig. 2b and c), the
guest polymers display highly exothermic interactions with both
host polymers (p␤CD and p␤CDN+).
Likewise for monomeric ␤CD, the fits are done assuming a sim-
ple 1:1 interaction model. The imperfect matching between the
experimental points and the fitting curves could be attributed to
slight correlations between the interaction sites within a same
chain. Nevertheless the fits allowed us to estimate thermodynamic
parameters of the interactions, reported in Table 2. The interactions
strength is strongly dependent on the adamantyl content of the
guest polymers. In the case of titration by p␤CD, Ka becomes almost
one order of magnitude higher in favor of D40-28GP-10Ada-18PEG
– more substituted by adamantyl groups and less by PEG chains.
Such behavior might be ascribed to interplay between the coopera-
tivity effects of neighboring adamantyl groups and steric hindrance
arising from PEG grafts. One can suppose that D40-28GP-5Ada-
23PEG is less likely to approach and form stable complexes with
p␤CD due to a lower number of linking points (5 mol% Ada) and
higher density of bulky PEG grafts. The evolution of the TꢀS values
is in agreement with this assumption: significantly negative TꢀS
values at 5 mol% Ada, which could correspond to anti-cooperative
binding, switch to positive TꢀS values at 10 mol% Ada, giving an
indication of more cooperative binding.
D40-18GP-18PEG
D40-28GP-5Ada-23PEG
D40-28GP-10Ada-18PEG
18
28
28
0
5
10
18
23
18
Host polymers
CD, wt%
Mw, kDa
N+/CD
p␤CD
67
65
208
–
0
1.62
p␤CDN+a
a
Prepared from p␤CD.
D40-28GP-10Ada-18PEG one observes two distinct singlets at 8.06
and 8.10 ppm attributed to the PEG and Ada triazole protons respec-
tively (Fig. S5).
3.1.2. pˇCD host polymers
The “host” polymers, neutral epichlorhydrin-interconnected
p␤CD and positively charged p␤CDN+, modified with quaternary
ammonia, were prepared according to the previously reported pro-
cedure (Blomberg et al., 2004; Renard et al., 1997). These polymers
have branched structure where ␤CD moieties may be thought
of as balls interlinked by poly(2-hydroxypropyl)ether sequences
of different lengths (Fig. S2a). Due to the seven-time excess of
epichlorohydrin with respect to ␤CD, a part of it is consumed on the
formation of polytails along with polybridges. The sample used in
this work has a cyclodextrin weight ratio (g/g) of 65%, determined
by 1H NMR, a molecular weight Mw of 208 kDa, and a polydis-
persity index of 1.81, which were determined by size exclusion
chromatography (SEC). To obtain positively charged host polymer,
p␤CD was modified with quaternary ammonia by reaction with 2,3-
epoxypropyltrimethylammonium (Fig. S2b). The characteristics of
all the host and guest polymers are reported in Table 1.
The same tendency was observed in the case of titration by
charged p␤CDN+. Interestingly, the binding constants for the poly-
mer bearing 5 mol% of adamantyl groups are of the same order for
p␤CD and p␤CDN+; on the other hand for the 10 mol% polymer, Ka
is pronouncedly higher when titrated by p␤CDN+ (Table 2).
Given the results above, in the next step we were interested
in evaluation and comparison of the binding properties of (PEG,
adamantyl)-grafted dextrans on the surface by SPR.
3.2. ITC studies
The ability of D40-GP-Ada-PEG to form inclusion complexes
with ␤CD and ␤CD-polymers was examined by isothermal titra-
tion microcalorimetry (ITC). The interactions of the polymers with
monomeric ␤CD are highly exothermic (Table 2) as expected,
given the perfect matching in size between grafted Ada and ␤CD-
cavities leading to strong van der Waals interactions between
them. The fits of the titration curves (Fig. 2a) were done apply-
ing a simple 1:1 interaction model, where one assumes that each
of the adamantyl groups grafted on one chain (12–25 per chain)
acts independently from the others. The different thermodynamic
parameters derived from these experiments are given in Table 2.
The association constants are as high as 4.0–6.0 × 104 L/mol, which
is superior to previously described dextrans modified with Ada-
groups via simple esterification by 1-adamantanecarbonyl chloride
(Wintgens et al., 2011). It might be explained by the presence of rel-
atively long and hydrophilic spacer in D40-GP-Ada-PEG, the latter
making adamantyl groups more loosely attached and available for
complexation. As in the reference (Wintgens et al., 2011), a slight
3.3. SPR adsorption studies
To perform SPR binding experiments, ␤CD-polymers were
deposited as a 1st layer on the golden surface of the sensor. When
either p␤CD or p␤CDN+ solution is flowed over the sensor surface,
the RI increases due to the physical adsorption of CD-polymers,
which should be driven by non-specific and electrostatic inter-
actions between the polymers and the gold surface. Typically 3
injections of 2 g/L solutions were enough to reach the plateau in
RI, indicating the surface saturation, as shown in the first part of
the sensorgrams for neutral p␤CD (Fig. 3). It should be noted that in
Fig. 4 the transient increase in RI during the host polymer injections
Table 2
Thermodynamic parameters of interaction of the D40-GP-Ada-PEG polymers with monomeric ␤CD and ␤CD-polymers at 25 ◦C.
Titrant
Ada, mol%
Ka × 10−3, L/mol
ꢀH, kJ/mol
TꢀS, kJ/mol
ꢀG, kJ/mol
5
10
60.7
40.3
−44.4
−41.6
−16.9
−15.2
−27.4
−26.3
␤CD
5
10
3.34
21.9
−40.0
−21.7
−19.8
−20.2
−24.8
p␤CD
3.0
5
10
3.02
35.4
−35.6
−18.1
−15.7
−19.8
−26.0
p␤CDN+
7.8

478
I. Antoniuk et al. / Carbohydrate Polymers 133 (2015) 473–481
Fig. 2. Enthalpograms of the interaction of D40-GP-Ada-PEG polymers with monomeric ␤CD (a), p␤CD (b), and p␤CDN+ (c). The solid lines represent the fits of the data by
One Set of Sites model (Origin 7.0 ITC custom version).
is due to p␤CDN+ being dissolved in 0.5 M NaCl solution. The result-
ing 1st layer obtained after equilibration for 20 min gives a quite
stable ꢀRI in both cases (8.5 × 10−4 0.8 and 12.1 × 10−4 0.7 for
p␤CDN+ and p␤CD respectively).
The adsorbed amounts of the polymers may be estimated based
on Eq. (4). Calculations showed that the adsorbed amount is roughly
1.4 times higher for the neutral p␤CD – 1.37 mg/m2 – than for the
charged p␤CDN+ – 0.97 mg/m2. The poorer adsorption character-
istics of p␤CDN+ might be related to strong hydration and spatial
extension of the charged molecules, which prevents them from
forming a strong linkage to the surface. In support of this it should
be noted that when p␤CDN+ solution in pure H2O is used, i.e. elec-
trostatic repulsions are not screened at all, only negligible amounts
are adsorbed.
The next step was to investigate the adsorption of the D40-GP-
Ada-PEG polymers on the preformed p␤CD-layer as a function of
their DS by adamantyl groups. The results obtained in the case of
neutral p␤CD as the 1st layer are shown in Fig. 3. One can observe
Fig. 3. Adsorption of D40-GP-Ada-PEG on neutral p␤CD layer. DS(Ada): 0 (a), 5 (b), and 10 (c) mol%.

I. Antoniuk et al. / Carbohydrate Polymers 133 (2015) 473–481
479
Fig. 4. Adsorption of D40-GP-Ada-PEG on p␤CDN+ with 1.6 N+ per CD. DS(Ada): 0 (a), 5 (b), and 10 (c) mol%. The consecutive injections are indicated by arrows.
that for the polymer bearing no adamantyl groups there is no depo-
sition of the 2nd layer with ꢀRI going back to zero after the flow
of water for ca. 15 min (Fig. 3a). This indicates that non-specific
adsorption onto the gold surface has been prevented by the first
layer of neutral p␤CD and that there is no detectable interaction
between the p␤CD layer and D40-18GP-18PEG. This finding corre-
lates with the solution ITC studies results (see Section 3.2).
For the polymer bearing 5 mol% of adamantyl groups, clear
formation of the 2nd layer was observed after the 3 successive
injections. However a gradual desorption takes place as water is
flowed for ∼30 min, indicating that the deposited polymer is rather
loosely attached (Fig. 3b). When the DS(Ada) is raised to 10 mol%
the deposited D40-GP28-Ada10-PEG18 layer becomes more firmly
anchored and the ꢀRI after equilibration in water is increased by
30% (Fig. 3c). The content of PEG in all investigated polymers was at
ca. 80–85 wt%. Therefore we assumed them to have a dn/dc equal
to that of pure MeOPEG-OH (0.135 mL/g) in order to calculate the
adsorbed amounts (after 30 min of equilibration in water).
of equilibration a part of it is washed out with the flow of water
(Fig. 4b). Finally, for the D40-GP28-Ada10-PEG18 the adsorbed
amount is further increased and the layer gains in stability (Fig. 4c).
Interestingly, the adsorption pattern on p␤CDN+ is changed as
well. No or little decrease in refractive index is observed straight
after each consecutive injection. One can assume that due to the
higher strength of interpolymer interaction with p␤CDN+, there
are less of loosely anchored D40-XGP-YAda-ZPEG molecules which
are washed out after the adsorption.
The obtained results (Fig. 5) show that the adsorption of the D40-
GP-Ada-PEG polymers onto non-charged p␤CD is predominantly
driven by host–guest interactions. Moreover, higher adamantyl
substitution degree appears to be responsible for the formation of
thicker and more stable 2nd layer.
The situation is quite different when p␤CDN+ is used as the
1st layer (Fig. 4). Even D40-18GP-18PEG with no Ada functions
leads to formation of the 2nd layer with an adsorbed amount of
0.63 mg/m2 (Fig. 5). It also shows no significant decrease in ꢀRI
during the equilibration period (20 min) in water, indicating strong
binding between the layers (Fig. 4a). For D40-GP28-Ada5-PEG23
the adsorbed amount is ca. 3 times higher, but at the initial stage
Fig. 5. The adsorbed amounts of D40-GP-Ada-PEG with various DS(Ada): on p␤CD
(black bars, horizontal pattern); on p␤CDN+ (red bars, vertical pattern). (For inter-
pretation of the references to color in this figure legend, the reader is referred to the
web version of this article.)

480
I. Antoniuk et al. / Carbohydrate Polymers 133 (2015) 473–481
interpolymer interaction. For the first part of the sensorgram (until
RImax is reached) the interaction might be schematically written as:
[R − Ada]sol. + [R − CD]surf. ꢀ [Complex]surf.
(5)
In this case the variation of RI versus time, due to complexation,
is expressed by the following equation (Oshannessy, Brighamburke,
Soneson, Hensley, & Brooks, 1993):
dRI
dt
= kaCAda(RImax − RI) + kd
(6)
where CAda is the concentration of adamantyl groups in the injected
solution. The ka and kd constants are obtained by fitting the exper-
imental curves for the polymer concentrations ≥0.025 g/L to the
integrated form of the dependence of RI versus time:
CAdaka[R]max − RI(0)[1 − exp (−CAdaka + kdt)]
RI(t) =
(7)
CAdaka + kd
Fig. 6. The evolution of RI versus time for various D40-28GP-10Ada-18PEG concen-
trations flowed over a p␤CDN+ layer.
in which RI(0) is the refractive index in the moment of injection
(t = 0).
The average values of ka and kd obtained from the sensor-
grams recorded at polymer concentrations varying in the range
0.08–0.20 g/L are of 7.0 × 102 mol−1 s−1 and 1.1 × 10−3 s−1 (more
details in Table S1); examples of the fits may be found in the Sup-
porting data (Fig. S6). From this one can calculate the association
constants Ka using the following equation:
Contrarily to the ITC study where no interaction could be
detected between D40-GP-PEG and p␤CDN+ in solution (Section
3.2), the same polymer (D40-GP-PEG) leads to a non-negligible
interaction onto the gold/p␤CDN+ surface. There is probably a
contribution of the gold surface which is not as fully saturated
with p␤CDN+ as in the case of p␤CD. However this cannot fully
explain the quite large adsorbed amount. As polysaccharides
and polysaccharide-based nanoassemblies are typically charac-
terized by slightly negative values of zeta potential (Wintgens
et al., 2011), non-specific electrostatic interactions could also be
at the origin of the enhanced interpolymer interaction. Given
that, the adsorption behavior of D40-GP-Ada-PEG on the posi-
tively charged p␤CDN+-layer might be related to interplay between
the host–guest and reinforced non-specific interactions in the sys-
tem. Accordingly, the adsorbed amounts of D40-GP-Ada-PEG on
p␤CDN+ are roughly 3 times higher compared to the neutral p␤CD
(Fig. 5).
ka
Ka
=
(8)
kd
The estimated average Ka = 6.4 × 105 L/mol is more than one
order of magnitude higher than the one determined for the inter-
action in solution (Ka = 3.5 × 104 L/mol, Table 2); it is also roughly
3 times higher than the Ka between native ␤CD and monomeric
adamantane derivatives (Nielsen et al., 2010). This large Ka increase
compared to the one determined in solution has been reported
previously for polymeric guests and is attributed to increased coop-
erative interactions, the host moieties being constrained in close
proximity on the surface (Wintgens & Amiel, 2005). Moreover, Ka
is one order of magnitude higher than the previously reported
Ka between p␤CD and adamantyl modified PNIPAMs (PNIPAM-
Ada) on gold surface (Wintgens & Amiel, 2005), thus indicating an
excellent affinity of D40-GP-Ada-PEG to p␤CDN+. More specifically,
association rate constants ka (700 mol−1 s−1) are around 3 times
higher than the ones determined for PNIPAM-Ada/p␤CD (Wintgens
& Amiel, 2005) or for adamantyl carboxylic acid/␤CD systems
(Busse, DePaoli, Wenz, & Mittler, 2001). At the same time the disso-
ciation rate constants kd (1.1 × 10−3 s−1) are more than one order
of magnitude lower than those for monomeric host–guest inter-
actions (adamantyl carboxylic acid/␤CD) and 3 times lower than
for PNIPAM-Ada/p␤CD multivalent interactions case. The larger ka
and the lower kd than expected can be related to the previously
mentioned additional interactions arising between D40-GP-Ada-
PEG and p␤CDN+ at interfaces, leading to larger adsorption and
slower desorption rates.
3.4. Reaction kinetics by SPR
The kinetics parameters of interaction of D40-28GP-10Ada-
18PEG with positively charged p␤CDN+ were evaluated given that
the latter has demonstrated the most interesting SPR adsorp-
tion profile (Fig. 5). First, the gold surface was saturated with
p␤CDN+, and then D40-28GP-10Ada-18PEG solutions at various
concentrations (0.01–0.20 g/L) were passed over the p␤CDN+ layer.
The results are represented in Fig. 6. At low concentrations (e.g.
0.01 g/L), the RI increases quasi-linearly with time, indicating that
every polymer molecule reaching the surface should be adsorbed
in this case. The adsorbed amount corresponding to the plateau
(0.59 mg/m2) is less than 30% of the saturation amount, obtained
from the SPR adsorption experiment (see Section 3.3). The sen-
sorgram recorded at 0.025 g/L changes its character drastically
(Fig. 6). One observes a steep initial growth with 1.65 mg/m2 of
the deposited polymer at the plateau, which is already 78% of the
saturation value. When increasing the concentration up to 0.1 g/L
and higher, the initial exponential growth becomes even steeper
and the amount of adsorbed polymer at the plateau gradually
approaches the saturation level from the experiment with multiple
injections of the 1 g/L solution (Fig. 5). The above is an indication
of strong affinity between the polymers. Indeed, at such low con-
centrations as 0.025 g/L one already observes a fast exponential
saturation of the surface sites of p␤CDN+.
4. Conclusion
A series of new (PEG, Ada)-grafted dextrans have been prepared
using copper(I)-catalyzed azide-alkyne cycloaddition. Due to the
presence of longer and more flexible spacer between the adamantyl
groups and the dextran backbone, the polymers show significant
improvement of their surface binding properties comparing to the
previously described hydrophobically modified PNIPAms. More-
over, their ability to form inclusion complexes with monomeric
␤CD, neutral poly-␤CD and cationic poly-␤CDN+, both in solution
and on surface, proves to be strongly dependent on the degree of
substitution by Ada-functions, giving indications of cooperativity
Since the obtained sensorgrams contain continuous real-time
data, they can be used for evaluation of the kinetics of the

I. Antoniuk et al. / Carbohydrate Polymers 133 (2015) 473–481
481
effects between them. Given the above, (PEG, Ada)-grafted dex-
trans should be auspicious candidates for biomedical applications
where fast and reversible PEGylation is required.
Jung, L. S., Campbell, C. T., Chinowsky, T. M., Mar, M. N., & Yee, S. S. (1998).
Quantitative interpretation of the response of surface plasmon resonance
sensors to adsorbed films. Langmuir, 14(19), 5636–5648.
Kappe, C. O., & Van der Eycken, E. (2010). Click chemistry under non-classical
reaction conditions. Chemical Society Reviews, 39(4), 1280–1290.
Li, S. D., & Huang, L. (2008). Pharmacokinetics and biodistribution of nanoparticles.
Molecular Pharmaceutics, 5(4), 496–504.
Acknowledgement
Loftsson, T., & Brewster, M. E. (1996). Pharmaceutical applications of cyclodextrins.
1. Drug solubilization and stabilization. Journal of Pharmaceutical Sciences,
85(10), 1017–1025.
Loira-Pastoriza, C., Todoroff, J., & Vanbever, R. (2014). Delivery strategies for
sustained drug release in the lungs. Advanced Drug Delivery Reviews, 75, 81–91.
Nielsen, T. T., Wintgens, V., Amiel, C., Wimmer, R., & Larsen, K. L. (2010). Facile
synthesis of beta-cyclodextrin-dextran polymers by “Click” chemistry.
Biomacromolecules, 11(7), 1710–1715.
This work has been supported by the European Commission for
Research in the Framework of NanoS3 Marie Curie Initial Training
Network (Grant agreement no.: 290251).
Appendix A. Supplementary data
Oshannessy, D. J., Brighamburke, M., Soneson, K. K., Hensley, P., & Brooks, I. (1993).
Determination of rate and equilibrium binding constants for macromolecular
interactions using surface plasmon resonance: Use of nonlinear least squares
analysis methods. Analytical Biochemistry, 212(2), 457–468.
Pun, S. H., & Davis, M. E. (2002). Development of a nonviral gene delivery vehicle
for systemic application. Bioconjugate Chemistry, 13(3), 630–639.
Quaglia, F., De Rosa, G., Granata, E., Ungaro, F., Fattal, E., & La Rotonda, M. I. (2003).
Feeding liquid, non-ionic surfactant and cyclodextrin affect the properties of
insulin-loaded poly(lactide-co-glycolide) microspheres prepared by
spray-drying. Journal of Controlled Release, 86(2–3), 267–278.
Ren, S. D., Chen, D. Y., & Jiang, M. (2009). Noncovalently connected micelles based
on a beta-cyclodextrin-containing polymer and adamantane end-capped
poly(epsilon-caprolactone) via host–guest interactions. Journal of Polymer
Science Part A: Polymer Chemistry, 47(17), 4267–4278.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.carbpol.2015.07.027
References
Antoniuk, I., Volet, G., Wintgens, V., & Amiel, C. (2014). Synthesis of a new
dextran-PEG-b-cyclodextrin host polymer using “Click” chemistry. Journal of
Inclusion Phenomena and Macrocyclic Chemistry, 80, 93–100.
Blomberg, E., Kumpulainen, A., David, C., & Amiel, C. (2004). Polymer bilayer
formation due to specific interactions between beta-cyclodextrin and
adamantane: A surface force study. Langmuir, 20(24), 10449–10454.
Brewster, M. E., & Loftsson, T. (2007). Cyclodextrins as pharmaccutical solubilizers.
Advanced Drug Delivery Reviews, 59(7), 645–666.
Busse, S., DePaoli, M., Wenz, G., & Mittler, S. (2001). An integrated optical
Mach–Zehnder interferometer functionalized by beta-cyclodextrin to monitor
binding reactions. Sensors and Actuators B: Chemical, 80(2), 116–124.
Chan, T. R., Hilgraf, R., Sharpless, K. B., & Fokin, V. V. (2004). Polytriazoles as
copper(I)-stabilizing ligands in catalysis. Organic Letters, 6(17),
2853–2855.
Renard, E., Deratani, A., Volet, G., & Sebille, B. (1997). Preparation and
characterization of water soluble high molecular weight
beta-cyclodextrin-epichlorohydrin polymers. European Polymer Journal, 33(1),
49–57.
Roberts, M. J., Bentley, M. D., & Harris, J. M. (2002). Chemistry for peptide and
protein PEGylation. Advanced Drug Delivery Reviews, 54(4), 459–476.
Siemsen, P., Livingston, R. C., & Diederich, F. (2000). Acetylenic coupling: A
powerful tool in molecular construction. Angewandte Chemie International
Edition, 39(15), 2633–2657.
Suzuki, T., Ota, Y., Kasuya, Y., Mutsuga, M., Kawamura, Y., Tsumoto, H., et al. (2010).
An unexpected example of protein-templated click chemistry. Angewandte
Chemie International Edition, 49(38), 6817–6820.
van de Manakker, F., Vermonden, T., van Nostrum, C. F., & Hennink, W. E. (2009).
Cyclodextrin-based polymeric materials: Synthesis, properties, and
pharmaceutical/biomedical applications. Biomacromolecules, 10(12),
3157–3175.
Versluis, F., Voskuhl, J., Vos, J., Friedrich, H., Ravoo, B. J., Bomans, P. H. H., et al.
(2014). Coiled coil driven membrane fusion between cyclodextrin vesicles and
liposomes. Soft Matter, 10(48), 9746–9751.
Volet, G., & Amiel, C. (2012). Polyoxazoline adsorption on silica nanoparticles
mediated by host–guest interactions. Colloids and Surfaces B: Biointerfaces, 91,
269–273.
Vonarbourg, A., Passirani, C., Saulnier, P., & Benoit, J. P. (2006). Parameters
influencing the stealthiness of colloidal drug delivery systems. Biomaterials,
27(24), 4356–4373.
Wintgens, V., & Amiel, C. (2005). Surface plasmon resonance study of the
interaction of a beta-cyclodextrin polymer and hydrophobically modified
poly(N-isopropylacrylamide). Langmuir, 21(24), 11455–11461.
Wintgens, V., Layre, A. M., Hourdet, D., & Amiel, C. (2012). Cyclodextrin polymer
nanoassemblies: Strategies for stability improvement. Biomacromolecules,
13(2), 528–534.
Wintgens, V., Nielsen, T. T., Larsen, K. L., & Amiel, C. (2011). Size-controlled
nanoassemblies based on cyclodextrin-modified dextrans. Macromolecular
Bioscience, 11(9), 1254–1263.
David, C., Millot, M. C., Sebille, B., & Levy, Y. (2003). The reversible binding of
hydrophobically end-capped poly(ethylene glycol)s to
poly-beta-cyclodextrin-coated gold surfaces: A surface plasmon resonance
investigation. Sensors and Actuators B: Chemical, 90(1–3), 286–295.
Dedola, S., Nepogodiev, S. A., & Field, R. A. (2007). Recent applications of the
Cu-I-catalysed Huisgen azide-alkyne 1,3-dipolar cycloaddition reaction in
carbohydrate chemistry. Organic & Biomolecular Chemistry, 5(7),
1006–1017.
El Fagui, A., & Amiel, C. (2012). PLA nanoparticles coated with a beta-cyclodextrin
polymer shell: Preparation, characterization and release kinetics of a
hydrophobic compound. International Journal of Pharmaceutics, 436(1–2),
644–651.
Fulmer, G. R., Miller, A. J. M., Sherden, N. H., Gottlieb, H. E., Nudelman, A., Stoltz, B.
M., et al. (2010). NMR chemical shifts of trace impurities: Common laboratory
solvents, organics, and gases in deuterated solvents relevant to the
organometallic chemist. Organometallics, 29(9), 2176–2179.
Gaucher, G., Dufresne, M. H., Sant, V. P., Maysinger, D., & Leroux, J. C. (2005). Block
copolymer micelles: Preparation, characterization and application in drug
delivery. Journal of Controlled Release, 109(1–3), 169–188.
Greenwald, R. B., Pendri, A., & Bolikal, D. (1995). Highly water soluble taxol
derivatives: 7-Polyethylene glycol carbamates and carbonates. Journal of
Organic Chemistry, 60(2), 331–336.
Hedges, A. R. (1998). Industrial applications of cyclodextrins. Chemical Reviews,
98(5), 2035–2044.
Hein, C. D., Liu, X. M., & Wang, D. (2008). Click chemistry, a powerful tool for
pharmaceutical sciences. Pharmaceutical Research, 25(10), 2216–2230.
Hoogenboom, R., Moore, B. C., & Schubert, U. S. (2006). Synthesis of star-shaped
poly(epsilon-caprolactone) via ‘click’ chemistry and ‘supramolecular click’
chemistry. Chemical Communications, (38), 4010–4012.
Immordino, M. L., Dosio, F., & Cattel, L. (2006). Stealth liposomes: Review of the
basic science, rationale, and clinical applications, existing and potential.
International Journal of Nanomedicine, 1(3), 297–315.
Wu, L. F., & Brazel, C. S. (2008). Modifying the release of proxyphylline from PVA
hydrogels using surface crosslinking. International Journal of Pharmaceutics,
349(1–2), 144–151.
Yoncheva, K., Lizarraga, E., & Irache, J. M. (2005). Pegylated nanoparticles based on
poly(methyl vinyl ether-co-maleic anhydride): Preparation and evaluation of
their bioadhesive properties. European Journal of Pharmaceutical Sciences, 24(5),
411–419.
Jokerst, J. V., Lobovkina, T., Zare, R. N., & Gambhir, S. S. (2011). Nanoparticle
PEGylation for imaging and therapy. Nanomedicine, 6(4), 715–728.