Facile synthesis of β-cyclodextrin-dextran polymers by "click" chemistry

Article abstract of DOI:10.1021/bm9013233

Three series of novel water-soluble β-cyclodextrin-dextran polymers have been prepared by "click" chemistry. The polymers were synthesized from alkyne-modified dextrans (AMDs) onto which mono-6-O-deoxy-monoazido- βCD (N₃βCD) was grafted by a copper(I)-catalyzed azide-alkyne cycloaddition (CuAAC). The polymers have been characterized by NMR spectroscopy and size exclusion chromatography (SEC). The binding properties have been characterized by isothermal titration calorimetry (ITC) and show excellent accessibility of the βCDs.

Full text of DOI:10.1021/bm9013233

1
710
Biomacromolecules 2010, 11, 1710–1715
Facile Synthesis of ꢀ-Cyclodextrin-Dextran Polymers by “Click”
Chemistry
†
‡
‡
†
Thorbjørn Terndrup Nielsen, V e´ ronique Wintgens, Catherine Amiel, Reinhard Wimmer,
,‡
and Kim Lambertsen Larsen*
Department of Biotechnology, Chemistry, and Environmental Engineering, Aalborg University,
Sohngaardsholmsvej 57, 9000 Aalborg, Denmark, and Syst e` mes Polym e` res Complexes, ICMPE, CNRS
and University Paris Est, 2 rue Henri Dunant, 94320 Thiais, France
Received November 22, 2009; Revised Manuscript Received May 8, 2010
Three series of novel water-soluble ꢀ-cyclodextrin-dextran polymers have been prepared by “click” chemistry.
The polymers were synthesized from alkyne-modified dextrans (AMDs) onto which mono-6-O-deoxy-monoazido-
ꢀ
3
CD (N ꢀCD) was grafted by a copper(I)-catalyzed azide-alkyne cycloaddition (CuAAC). The polymers have
been characterized by NMR spectroscopy and size exclusion chromatography (SEC). The binding properties have
been characterized by isothermal titration calorimetry (ITC) and show excellent accessibility of the ꢀCDs.
Introduction
Cyclodextrin (CD) polymers are of high interest in biomedical
Experimental Section
Materials. Amberlite GT73 resin (Aldrich, Steinheim, Germany),
L(+)-ascorbic acid sodium salt 99% (NaAsc, Fluka, Steinheim,
Germany), ꢀ-cyclodextrin (ꢀCD, Beta W7 Pharma, Wacker-Chemie,
Burghausen, Germany), copper(II) bromide 99% (Aldrich, Steinheim,
Germany), copper(I) iodide 99% (Aldrich, Steinheim, Germany),
science, due to their numerous applications, for example,
1
2,3
4-6
formation of nanoparticles, gels, DNA transfection,
and
7
drug carrier systems. In the last decades, different types of CD
polymers have been synthesized, ranging from branched poly-
mers by reacting CDs with, for example, epichlorohydrin, linear
polymers with pending CDs,
incorporated into the backbone. Because the discovery of the
copper(II) sulfate 99% (Fluka, Steinheim, Germany), dextran (D70,
8
4
M
w
7 × 10 Da, Pharmacosmos A/S, Holbaek, Denmark), dichlo-
9
,10
5
to linear polymers with CDs
romethane, glycidyl propargyl ether 90% (GPE, Fluka, Steinheim,
Germany), 5-hexynoic acid 97% (Aldrich, Steinheim, Germany), N,N-
isopropylethylamine 99% (DIPEA, Aldrich, Steinheim, Germany), 2,6-
lutidine 99% (Aldrich, Steinheim, Germany), oxalyl chloride 99%
11,12
copper(I)-catalyzed azide-alkyne cycloaddition (CuAAC),
this versatile “click” reaction has found applications in many
fields. In recent years, new ꢀCD polymers have been prepared
(
Fluka, Steinheim, Germany), 2-propanol, puriss (Aldrich, Steinheim,
Germany), and propargyl chloroformate 96% (Aldrich, Steinheim,
Germany) were used as received.
1
3,14
15
by “click” chemistry,
as well as ꢀCD conjugates and ꢀCD
1
6
analogs. In biomedical science, “click” chemistry has been
used in, for example, microcapsules, glycopolymers, DNA
conjugation, and in preparation of thermosensitive hydrogels.
1
7
18
24
Benzyl azide,
bromotris(triphenylphosphine)copper(I) (Cu-
19
20
25
26
(PPh
3 3
) Br), tris-(benzyltriazolylmethyl)amine (TBTA), and p-tolu-
2
7
CuAAC can be made in many solvents over a wide range of
temperatures and pH values and in the presence of numerous
functional groups. It proceeds fast and normally with very high
efficiency. The triazole is formed exclusively as the 1,4-isomer
ensulfonyl imidazole were prepared according to literature.
4
2-(1-Adamantyl)ethyl trimethylammonium bromide (AdaTMA) and
8
the ꢀCD/epichlorohydrin copolymer (pꢀCD) were synthesized as
5
previously described. pꢀCD has a molecular weight M
w
of 1.6 × 10
-
1
1
2
1
g·mol and a ꢀCD content of 59% (w/w), determined by H NMR.
Water used for synthesis and dialysis was of Milli-Q quality. When
anhydrous conditions were applied all glassware was dried at 130 °C
for 2 h and cooled in a desiccator prior to use. ꢀCD and dextran were
dried at 115 °C in vacuum overnight. AMDs were stored at -20 °C.
and is stable against oxidation, reduction, and hydrolysis.
Dextran is a linear natural glucose polymer with molecular
weights ranging from 1-2.000 kDa linked by R-1,6-glycosidic
bonds with few short R-1,4-linked branches. It is used exten-
sively in biomedical science due to its great biocompatibility
and high solubility in water. Despite the attractive biomedical
properties of dextran, very few ꢀCD-dextran polymers have been
synthesized. Villalonga and co-workers have described polymers
made from oxidatively cleaved dextran and amino derivatives
Methods. NMR analysis of the polymers were conducted in D
30 g/L) with a Bruker DRX600 spectrometer (5 mm TXI (H/C/N)
2
O
(
1
xyz-gradient probe). H spectra were calibrated using the residual water
2
8
signal in accordance with literature.
1
3
C chemical shifts were calibrated indirectly based on the 1H
calibration. Quantitative H NMR analyses were made with a delay
time (d1) set for 30 s. For assignment and structural analysis, double-
22,23
1
of ꢀCD by reductive amination,
but so far, no dextran-ꢀCD
polymers have been prepared from native dextran. Herein we
present a facile synthesis of three new series of ꢀCD-dextran
polymers prepared by “click” chemistry using alkyne-modified
dextrans (AMDs) and 6-O-monodeoxy-6-monoazido-ꢀCD
2
9
quantum filtered COSY, multiplicity edited HSQC, TOCSY (with
100 ms mixing time and a spin-lock field strength of 8.1 kHz), and
NOESY (with 150 ms mixing time) spectra were recorded. NMR
3
0
spectra were interpreted with CARA 1.8.2. Carbonyl carbon atoms
were not observed. In HSQC and DEPT spectra they do not yield
3
(N ꢀCD). The polymers consist of ether-, ester-, and carbonate-
linked ꢀCD, respectively. The ꢀCD content is tunable by
changing the degree of substitution (DS) of the AMDs and varies
from approximately 17-54% (w/w).
13
signals, and the signal-to-noise ratio in the C spectra was not sufficient
to observe those weak signals. TLC analyses were made with
ALUGRAM SIL G/UV254 TLC plates with 0.2 mm silica gel, visualized
2 4
SO
in ethanol. SEC was made in water with 0.1 mol ·L-
1
with 5% H
NaN
miniDawn light scattering detector and a Wyatt Optilab Rex. refractive
*
To whom correspondence should be addressed. E-mail: kll@bio.aau.dk.
Aalborg University.
CNRS and University Paris Est.
†
(columns TSK-gel type SW 4000-3000 and detection by a Wyatt
3
‡
10.1021/bm9013233
 2010 American Chemical Society
Published on Web 06/16/2010

Synthesis of ꢀ-Cyclodextrin-Dextran Polymers
Biomacromolecules, Vol. 11, No. 7, 2010 1711
1
index detector). Fourier transform infrared spectroscopy (FT-IR) was
made on a Bruker Tensor 27 FT-IR spectrometer. Matrix-assisted laser
desorption/ionization-time-of-flight mass spectrometry (MALDI-TOF
MS) was made with a Bruker Daltronics Reflex III in reflector mode,
using R-cyano-4-hydroxycinnamic acid as matrix. Dialysis was made
with Spectrum Spectra/Por, MWCO 6000-8000 25.5 mm diameter
and MWCO 500 15 mm diameter and freeze-drying on a Heto CT
product was stored under nitrogen at 0 °C until use. H NMR (CDCl
δ ) 2.65 (t, 1H), 5.0 (d, 2H), 7.1 (s, 1H), 7.5 (s, 1H), 8.2 (s, 1H).
3
)
D70 Propargyl Carbonate (D70PC). Dry dextran (2 g, 12.35 mmol
glucose units) was dissolved in 60 mL of DMSO under nitrogen. The
appropriate amount of propargyl imidazole formate (0.92-3.08 mmol)
was added. The solution was stirred under nitrogen at 50 °C for 24 h.
The product was precipitated in 600 mL of 2-PrOH, filtered, and washed
with 100 mL of 2-PrOH. After drying under nitrogen, the product was
6
0e. TGA was made with a NETZSCH STA 449C Jupiter using argon
as purge gas. A total of 10-12 mg of polymer in a platinum crucible
was heated to 160 °C using a heating rate of 5 °C/min. The water
content of the polymers was calculated from the weight loss.
ITC measurements were carried out at 25 °C using a MicroCal VP-
ITC microcalorimeter. In each titration, injections of 5 µL of concen-
dissolved in 75 mL of water, dialyzed against water for 36 h, and freeze-
1
dried, yielding a white solid (1.84-2.00 g). H NMR (D
(t, 1H), 3.4-4.2 (m, 5H), 4.8 (s, 2H), 4.9-5.3 (m, 1H).
2
O) δ ) 3.05
D70 Propargyl Glycerol (D70GP). Dextran (1.5 g, 9.25 mmol
glucose units) was dissolved in 10 mL of 0.1 M NaOH. Glycidyl
propargyl ether (7.71-30.86 mmol) was added, the temperature was
increased to 35 °C, and the solution was stirred for 16 h. The product
was precipitated in 200 mL of 2-PrOH, filtered, and washed with 100
mL of PrOH. After drying under nitrogen, the product was dissolved
-
1
-1
trated AdaTMA solutions (0.1 mol ·L in 0.1 mol ·L NaCl) were
added from the computer-controlled 295 µL microsyringe at an interval
of 180 s into the cell (volume ) 1.4569 mL) containing the polymer
-
3
-1
solution (corresponding to 10 mol·L of ꢀCD), while stirring at
50 rpm. The raw experimental data were obtained as the amount of
in 50 mL of water, dialyzed against water for 36 h, and freeze-dried,
4
1
yielding a white solid (1.38-1.55 g). H NMR (D
2
O) δ ) 2.95 (s,
heat produced per second following each injection of adamantyl
derivative as a function of time. Integration of the heat flow peaks by
the instrument software (after taking the heat of dilution into account)
gives the amount of heat produced per injection. The experimental data
are fitted with a theoretical titration curve using the instrument software,
assuming a 1:1 complex. The enthalpy change, ∆H, the binding
constant, K, and the stoichiometry, n, are the adjustable parameters.
Synthesis. 6-O-Monotosyl-ꢀCD. ꢀCD (40 g, 35.2 mmol) was
dissolved in 900 mL of water by warming. The temperature was lowered
to 45 °C, tosyl imidazole (31.3 g, 141 mmol) was added, and the
suspension was stirred vigorously for 2 h while cooling to ambient
temperature. Sodium hydroxide (18 g, 45 mmol) in 50 mL of water
was added over 15 min. The mixture was stirred for 10 min and quickly
filtered through a sintered glass filter into a flask containing ammonium
chloride (48.2 g, 0.9 mol). After dissolving the ammonium chloride,
the mixture was poured into a large beaker and concentrated by blowing
air over the surface overnight. The precipitate was filtered off and
washed with 2 × 100 mL of ice cold water and 200 mL of acetone.
The product was dried in vacuo and recrystallized twice, first by
dissolving in approximately 15 mL of DMSO and precipitating in 200
mL of water. The isolated product was then recrystallized in 75 mL of
water brought to boiling. After cooling on ice, the product was filtered
off and washed with 2 × 50 mL of water and 100 mL of acetone. The
1
H), 3.4-4.2 (m, 5H), 4.2 (s, 2H), 4.9-5.3 (m, 1H).
-Hexynoyl Chloride. 5-Hexynoic acid (0.46-1.54 mmol) was
dissolved in 1 mL of CDCl , and a catalytic amount of DMF (10 µL
of 100 µL of DMF in 1 mL CDCl ) was added. The solution was cooled
5
3
3
in an ice bath and oxalyl chloride (0.46-1.54 mmol) was added dropwise
under nitrogen atmosphere. The solution was stirred at ambient temperature
and monitored by NMR. When NMR indicated nearly complete conversion
(typically within 3 h), the solution of hexynoyl chloride was used
1
immediately in the next step without purification. H NMR (CDCl
.9 (m, 2H), 2.0 (t, 1H), 2.3 (m, 2H), 3.1 (t, 2H).
D70-Hex-5-ynoate (D70HX). Lithium chloride (0.5 g) and dextran
2 g, 12.35 mmol glucose units) were dissolved in 50 mL of DMF by
warming. DMAP (0.69-2.3 mmol) was added. Freshly prepared
-hexynoyl chloride in CDCl (0.46-1.54 mmol) was dissolved in 3
3
) δ )
1
(
5
3
mL of DMF, and was added to the solution of dextran at ambient
temperature. The temperature was raised to 80 °C, and the solution
was stirred for 2 h under nitrogen. After cooling to ambient temperature,
the product was precipitated in 300 mL of 2-PrOH, filtered, and washed
with 100 mL of 2-PrOH. After drying under nitrogen, the product was
dissolved in 75 mL of water and dialyzed against water for 36 h. The
product was obtained by freeze-drying as white solid (1.88 - 2.13 g).
1
H NMR (D O) δ ) 1.75 (s, 2H), 2.3 (s, 2H), 2.4 (s, 1H), 2.6 (s, 2H),
2
3
.4-4.2 (m, 5H), 4.2 (s, 2H), 4.9-5.3 (m, 1H).
pure 6-O-monotosyl-ꢀCD (TsOꢀCD) was dried overnight in vacuo at
General Synthesis of ꢀCD-Dextran Polymers. A total of 0.25 g
1
6
0 °C. Yield: 10.61 g. (23.4%). H NMR (DMSO-d
6
) δ (ppm) ) 2.43
alkyne-grafted dextran (0.072-0.305 mmol alkyne), N
3
ꢀCD (0.134-
(
4
(
s, 3 H), 3.39-3.20 (m, overlaps with DHO), 3.70-3.42 (m, 28H),
.52-4.17 (m, 6 H), 4.85-4.77 (m, 7 H), 5.83-5.63 (m, 14 H), 7.43
d, 2 H), 7.75 (d, 2 H).
-Monodeoxy-6-monoazido-ꢀCD. TsOꢀCD (6.5 g, 5 mmol) and
0
.610 mmol), and TBTA (0.011-0.456 mmol) were dissolved in 25
2
mL of degassed DMSO/H O 1:1 under nitrogen atmosphere. NaAsc
(
134-603 µL of a 20 mg/mL solution, 0.0144-0.061 mmol) was added
and the solution was degassed by ultrasound and subsequent bubbling
with nitrogen for 10 min while reaching 50 °C. CuSO (108-486 µL
6
sodium azide (1.64 g, 25 mmol) were dissolved in 40 mL of DMF.
The mixture was stirred at 75 °C for 24 h, precipitated in 800 mL of
acetone, and filtered. The crude product was recrystallized by dissolving
it in 50 mL of boiling water and precipitating in 800 mL of acetone.
This procedure was repeated until IR showed no peak from residual
4
of a 10 mg/mL solution, 0.0072-0.0305 mmol) was added and the
solution was stirred for 24 h and subsequently dialyzed against water
for 72 h. The polymers were obtained as white solids by freeze-drying.
Analytical samples were produced by stirring solutions of 250 mg
polymer in 10 mL of water with 2 g Amberlite GT73 resin for 12 h
before being percolated through a short column with 1 g of the same
-
1
sodium azide (2138 cm ). The product was dialyzed against water
MWCO 500) for 24 h and freeze-dried. The pure 6-monodeoxy-6-
monoazido-ꢀCD (N ꢀCD) was further dried in vacuo at 60 °C
overnight. Yield: 4.96 g. (84.8%). H NMR (DMSO-d
.41-3.26 (m, overlaps with DHO), 3.76-3.55 (m, 28 H), 4.53-4.46
(
3
resin and freeze-dried.
1
) δ (ppm) )
1
6
D70HXꢀCD. H NMR (D
O) δ ) 2.0 (s, 2H, hexynoic-CH
2
), 2.5
2
3
(
s, 2H, hexynoic-CH
2
), 2.7 (s, 2H, hexynoic-CH ), 2.8 (d, 1H, ꢀCD-
2
2-5
6
′
6′
2-5
(
m, 6 H), 4.88-4.82 (m, 7 H), 5.78-5.62 (m, 14 H).
C ), 3.1 (d, 1H, ꢀCD-C ), 3.4-4.2 (m, ꢀCD-H and dextran-H ),
1
1
Propargyl Imidazole Formate (PocIm). Propargyl chloroformate (5.0
4.5 (s, 2H), 4.9-5.3 (m, ꢀCD-H and dextran-H ), 7.85 (s, 1H, triazole).
1
3
g, 42.2 mmol) was dissolved in 25 mL of dry DMC and cooled to
10 °C. Imidazole (6.03 g, 88.62 mmol) was added in small portions
C NMR (D O) δ ) 26.4 (hexynoic-CH ), 26.6 (hexynoic-CH ), 35.7
2
2
2
6
6′
-
2
(hexynoic-CH ), 53.7 (ꢀCD-C triazole), 61.6 (ꢀCD-C ), 62.8 (ꢀCD-
C ), 68.0 (dextran-C ), 72.0-75.8 (several signals from ꢀCD-C
dextran-C ),83.6-85.6(ꢀCD-C ),97.3-100.0(dextran-C ),101.7-104.2
(ꢀCD-C ), 127.2 (triazole).
6
6
2,3,5
under stirring. After 1 h, the solution was allowed to warm to ambient
temperature and was stirred for an additional 2 h. The solution was
extracted four times with 5 mL of water. The organic phase was dried
and
2-5
4
1
1
1
6′
over MgSO
4
and filtered. DCM was removed in vacuo and propargyl
2
D70GPꢀCD. H NMR (D O) δ ) 2.8 (d, 1H, ꢀCD-C ), 3.1 (d, 1H,
6
′
2-5
2-5
imidazole formate was obtained as a white solid (5.65 g, 88%). The
ꢀCD-C ), 3.4-4.2 (m, ꢀCD-H
and dextran-H ), 4.7 (s, 2H,

1
712 Biomacromolecules, Vol. 11, No. 7, 2010
Nielsen et al.
ꢀCD made
MALDI-TOF MS (Figure 2) shows that the pure N
from the TsOꢀCD contains some native ꢀCD and a very small
amount of 6,6′-dideoxy-6,6′-diazido-ꢀCD (diN ꢀCD). From the
MS intensities the amount of diN ꢀCD was estimated to be in
the range of 0.5-1%, while approximately 15% native ꢀCD
were present, depending on the batch. The N ꢀCD was used
3
3
3
3
without further purification because the CuAAC with AMDs
still should lead to a high degree of linearity, despite the presence
Figure 1. Reaction scheme for the synthesis of 6-monodeoxy-6-
monoazido-ꢀCD (N ꢀCD).
3
of a small amount of diN ꢀCD.
3
Alkyne-Modified Dextrans (AMDs). Three different series
of AMDs were prepared with carbonate, ester, and ether bonds,
respectively. Because the alkyne functions are assumed to be
randomly distributed in the dextran chains, some might be
difficult to access when already grafted ꢀCD is in the proximity.
Especially when dealing with AMDs with a high DS, a
pronounced steric effect of the grafted ꢀCD can be anticipated.
Recently, Ritter and co-workers described the preparation of a
1
1
propagyl-CH ), 4.9-5.3 (m, ꢀCD- H and dextran- H ), 8.1 (s, 1H,
2
1
3
6
triazole). C NMR (D O) δ ) 53.8 (ꢀCD-C triazole), 61.8 (ꢀCD-
2
6′
6
6
C ), 62.9 (ꢀCD-C ), 66.1 (propagyl-CH
2
), 68.1 (dextran-C ), 71.8-76.0
2
,3,5
2-5
(
several signals from ꢀCD-C
and dextran-C ), 83.7-85.7 (ꢀCD-
C ), 97.4-100.3 (dextran-C ), 101.8-104.5 (ꢀCD-C ), 129.1 (triazole).
D70PCꢀCD. H NMR (D O) δ ) 2.9 (d, 1H, ꢀCD-C ), 3.2 (d, 1H,
CD-C ), 3.4-4.2 (m, ꢀCD-H and dextran-H ), 4.9-5.3 (m, ꢀCD-
H and dextran- H ), 5.4 (s, 2H, propargyl-CH
4
1
1
1
6′
2
6′
2-5
2-5
ꢀ
1
1
2
), 8.1 (s, 1H, triazole).
1
3
6
6′
ꢀCD polymer by microwave-assisted CuAAC of N ꢀCD to
3
C NMR (D
2
O) δ ) 54.0 (ꢀCD-C triazole), 62.1 (ꢀCD-C ), 63.1
6
6
poly(propargyl methacrylate). The reaction proceeded very
slowly without microwave assistance, even at 140 °C, due to
the bulkiness of the ꢀCD, and their attempts to synthesize an
identical polymer by radical polymerization only led to oligo-
(
ꢀCD-C ), 63.6 (propargyl-CH
2
), 67.8 (dextran-C ), 71.8-75.7 (several
and dextran-C ), 83.8-85.6 (ꢀCD-C ),
7.9-100.2 (dextran-C ), 102.0-104.5 (ꢀCD-C ), 129.9 (triazole).
2
,3,5
1
2-5
4
signals from ꢀCD-C
9
1
1
4
mers. Furthermore, Cu(I) saturation can take place when
multiple alkynes are present in the near vicinity of each other
due to a chelating effect between the Cu(I)-acetylide complex and
surrounding alkynes. Hence, the azide is prevented from approach-
Results and Discussion
6
-Monodeoxy-6-monoazido-ꢀCD. The choice of alkyne-
grafted dextran and azide-substituted ꢀCD was more attractive
than the reverse due to the ease of obtaining a relatively high
31
ing the Cu(I)-acetylide complex and the CuAAC fails. For this
reason, the DS of the AMDs was chosen to be in the range of
approximately 5-20%, corresponding to a theoretical mass per-
centage of ꢀCD between 20 and 55% (w/w).
purity N
TsOꢀCD) by the route illustrated in Figure 1. TsOꢀCD was
synthesized by a slightly modified version of the procedure
3
ꢀCD. N ꢀCD was prepared from 6-O-monotosyl-ꢀCD
3
(
2
7
described in the literature. Applying this method ensures that
only a low content of 6-O-ditosyl-ꢀCD (diTsOꢀCD) is formed
compared to other procedures and excludes the need for tedious
and expensive chromatographic purification. The typical yield
is around 20%, which is lower than the one reported, but in
return, the content of diTsOꢀCD is assumed to be lower.
The ether series (D70GPxs) was synthesized by a basic
epoxide opening of glycidyl propargyl ether (GPE) in 0.5 M
NaOH. Water was applied as solvent to avoid homopolymer-
ization of the GPE. Due to a competing hydrolysis by hydroxide
approximately three times the amount of GPE was applied to
achieve the desired modification.
2
7
3
Figure 2. MALDI-TOF MS spectrum of N ꢀCD.

Synthesis of ꢀ-Cyclodextrin-Dextran Polymers
Biomacromolecules, Vol. 11, No. 7, 2010 1713
Figure 3. Multiplicity-edited HSQC spectrum of D70PGꢀCD4.
The carbonate series (D70PCs) was made from propargyl
carbonyl imidazole (PocIm) in DMSO by a modified version
1
7
of the procedure described in the literature. Commercially
available propargyl chloroformate lead to gel formation, and
PocIm was used instead. Purification of the D70PCs by direct
dialysis of the solution as described in literature was not made.
Instead, the D70PCs were precipitated in 2-PrOH and filtered
prior to dialysis for 24 h, leading to a lower loss of alkyne
function due to hydrolysis and still in good yields.
The ester series (D70HXs) was made in DMF by esterification
of freshly prepared hexynoyl chloride using DMAP as base and
catalyst. This method is a modified version of the preparation
2
of hydrophobically modified dextran, previously described, and
leads to comparable grafting ratios and yields.
The degree of substitution of the AMDs was determined by
NMR from the ratio between the integral of the terminal alkyne
protons (at 2.4, 2.9, and 3.0 ppm for D70HX, D70GP, and
D70PC, respectively) and the anomeric proton of dextran. SEC
shows no significant change in molar weight or polydispersity
index (PDI) after alkyne modification in the case of D70GP
and D70HX. In the carbonate series, an increase of ap-
proximately 50% of molar weight was observed in the cases of
high modification degrees, most likely due to a few percent
carbonate cross-linking. DS and yields of the AMDs can be
found as Supporting Information.
Figure 4. Indication of all observed NOEs in D70GPꢀCD4.
hydrophobic character of the AMDs, a 1:1 DMSO/H
was preferred over pure water and proved appropriate. As a
copper(I) source, a CuSO /NaAsc system was applied. After
the CuAAC with N ꢀCD NMR analysis showed disappearance
2
O mixture
4
3
of the terminal alkyne proton and corresponding appearance of
a triazole proton at 7.8-8.1 ppm. Interestingly, two doublets
appear at 2.8 and 3.15 ppm for all three series of polymers. A
resonance assignment based on COSY, HSQC (Figure 3),
Copper(I)-Catalyzed Azide-Alkyne Cycloaddition (CuAAC).
Regarding solvents, the possibilities were limited by few
common solvents for the AMDs and N ꢀCD. Due to the partial
3

1
714 Biomacromolecules, Vol. 11, No. 7, 2010
Nielsen et al.
Figure 5. Reaction scheme for the syntheses of D70GPꢀCDs, D70PCꢀCDs, and D70HXꢀCDs.
TOCSY, and NOESY spectra proved these signals to come from
the C6 CH group from the glucose unit whose C4 is glyco-
sidically bound to the C1 of the glucose carrying the triazole
coupling have successfully been achieved by applying bromo-
18,33
2
tris(triphenylphosphine)copper(I) (Cu(PPh
3 3
) Br).
In this case,
the approach was not successful. The CuAAC proceeded very
slowly, even at elevated temperatures (140 °C for 24 h), and
only low conversion (∼20%) could be achieved, probably
caused by low solubility of the polymers in pure DMSO. Even
(
Figure 4).
In the case of the D70GPꢀCDs, a downfield shift of the CH
2
group in the R-position to the alkyne at 4.2-4.7 ppm was
observed. For the D70HXꢀCDs there was a slight downfield
shift of the aliphatic protons of the hexynyl chain.
though both the AMDs and the N ꢀCD are readily soluble in
3
DMSO, ꢀCD-dextran polymers made in a 1:1 mixture of
The ꢀCD content of the polymers was found from the ratio
between the integrals of the characteristic triazol signal at
DMSO/water turned out to have a very low solubility in DMSO.
Because the Cu(PPh
mixture, the catalyst was not applicable. Instead, efforts were
made to optimize the CuSO /Asc system. An attempt to stabilize
3 3
) Br was insoluble in the DMSO/water
7
.8-8.1 ppm and the integrals of the overlap of the anomeric
protons from ꢀCD and dextran at 4.9-5.4 ppm. The amount of
grafted ꢀCD was in all cases lower than the theoretical (∼85%)
and SEC revealed that the molar weights of the polymers were
much higher than expected (2-3 times higher). This was especially
4
the Cu(I) species against oxidation or disproportionation was
made by the addition of tris-(benzyltriazolylmethyl)amine
(TBTA) and applying an inert atmosphere. TBTA is known to
2
6
the case when only a slight excess of N
two equivalents of N ꢀCD lowered the phenomenon. Though not
often, difficulties in CuAAC, especially when dealing with alkyne
3
ꢀCD was applied, whereas
stabilize Cu(I) as well as accelerate CuAAC. CuAAC in
DMSO/water mixtures have been made in very high yields when
3
3
4
catalyzed in the presence of TBTA. When these conditions
were applied in the CuAAC between Τ70GPs and N ꢀCD,
12,32
scaffolds, have been reported
and are most likely due to
3
competing alkyne cross-couplings by a Glaser, Straus, or Eglinton
mechanism. CuAAC on alkyne scaffolds without cross-
dramatic improvements were observed. For the D70GPꢀCDs,
virtually quantitative coupling was achieved (>97%). In the cases
32

Synthesis of ꢀ-Cyclodextrin-Dextran Polymers
Biomacromolecules, Vol. 11, No. 7, 2010 1715
of D70HXꢀCDs, a slightly lower content of ꢀCD was found
in the isolated polymers, most likely due to ester hydrolysis
during the CuAAC and the subsequent dialysis. For the
D70PCꢀCDs, the hydrolysis was even more pronounced and,
further, carbonate coupling lead to a large increase in molecular
weights in the cases of high DS. ꢀCD content, yields, molecular
weights, and PDI of the polymers can be found as Supporting
Information. The reaction scheme for the syntheses of the
polymers is illustrated in Figure 5.
dextran polymers. This material is available free of charge via
the Internet at http://pubs.acs.org.
References and Notes
(1) Gref, R.; Amiel, C.; Molinard, K.; Daoud-Mahammed, S.; Sebille,
B.; Gillet, B.; Beloeil, J. C.; Ringard, C.; Rosilio, V.; Poupaert, J.;
Couvreur, P. J. Controlled Release 2006, 111, 316–324.
(
2) Wintgens, V.; Daoud-Mahammed, S.; Gref, R.; Bouteiller, L.; Amiel,
C. Biomacromolecules 2008, 5, 1434–1442.
(
3) Daoud-Mahammed, S.; Couvreur, P.; Bouchemal, K.; Ch e´ ron, M.;
Lebas, G.; Amiel, C.; Gref, R. Biomacromolecules 2009, 10, 547–
Binding Properties. The ability of the polymers to form
inclusion complexes was investigated by isothermal titration
microcalorimetry (ITC) at 25 °C with 2-(1-adamantyl) ethyl
trimethylammonium bromide (AdaTMA) as a guest. The
interactions between the species are strongly exothermic, as
expected for a complex formation between ꢀCD cavities and
an adamantyl derivative. The |∆H| values are high, around
5
54.
(
4) Burckbuchler, V.; Wintgens, V.; Lecomte, S.; Percot, A.; Leborgne,
C.; Danos, O.; Kichler, A.; Amiel, C. Biopolymers 2006, 81, 360–
3
70.
(5) Gonzales, H.; Hwang, S. J.; Davis, M. E. Bioconjugate Chem. 1999,
10, 1068–1074.
(6) Pun, S. H.; Davis, M. E. Bioconjugate Chem. 2002, 13, 630–639.
(7) Li, J.; Xiao, H.; Li, J.; Zhong, Y. Int. J. Pharm. 2004, 278, 329–342.
(8) Renard, E.; Deratani, A.; Volet, G.; Sebille, B. Eur. Polym. J. 1997,
-1
3
0.8-32.6 kJ ·mol , and the complex formation is enthalpy
driven. Large negative enthalpy changes can partially be
attributed to van der Waals interactions caused by the precise
matching in size and shape between the guest and the host,
which is particularly the case of the adamantyl group and the
3
3, 49–57.
(9) Harada, A.; Furue, M.; Nozakura, S. Macromolecules 1976, 9, 701–
04.
7
(
(
10) Renard, E.; Volet, G.; Amiel, C. Polym. Int. 2005, 54, 594–599.
11) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew.
Chem., Int. Ed. 2002, 41, 2596–2599.
35
2,36
ꢀCDcavity. ComparedtoꢀCDpolymerspreviouslydescribed
the new polymers show excellent binding properties. The
association constant between the polymers and AdaTMA is in
(12) Torne, C. W.; Christensen, C.; Meldal, M. J. Org. Chem. 2002, 67,
3057–3064.
(13) Srinivasachari, S.; Reineke, T. M. Biomaterials 2009, 30, 928–938.
5
5
-1
the range of 1.2 × 10 - 2.05 × 10 mol·L . Hence it is more
(
(
(
(
14) Munteanu, M.; Choi, S. W.; Ritter, H. Macromolecules 2008, 41, 9619–
than 1 order of magnitude higher than the one for pꢀCD (1.3
9
623.
4
-1
×
10 mol·L ), and is comparable to the one for native ꢀCD
5
15) Liu, X.-M.; Lee, H.-T.; Reinhardt, R. A.; Marky, L. A.; Wang, D. J.
Controlled Release 2007, 122, 54–62.
16) Bodine, K. D.; Gin, D. Y.; Gin, M. S. J. Am. Chem. Soc. 2004, 126,
-1
(
1.9 × 10 mol·L ). The difference can probably be ascribed
to the high degree of linearity in most of the polymers and thus
a high accessibility of the ꢀCD cavities. It should be stressed
that the ꢀCD cavities are linked to the main chain via spacers
of different lengths. The ꢀCD accessibility will be favored in
the order of the spacer length, D70PCꢀCD < D70HXꢀCD <
D70GPꢀCD. In addition, the accessibility is expected to decrease
at large DS. A slight decrease of the association constant is
observed for the D70PCꢀCD and D70HXꢀCD series with
increasing DS. In the case of D70PCꢀCD series, another
explanation of the largest decrease of K could be the more cross-
linked structure (higher molecular weights than expected). A
table with the thermodynamic data can be found as Supporting
Information.
1
638–1639.
17) De Geest, B. G.; Van Camp, W.; Du Prez, F. E.; De Smedt, S. C.;
Demeesterb, J.; Henninka, W. E. Chem. Commun. 2008, 19, 0–192.
(18) Ladmiral, V.; Mantovani, G.; Clarkson, G. J.; Cauet, S.; Irwin, J. L.;
Haddleton, D. M. J. Am. Chem. Soc. 2006, 128, 4823–4830.
(
(
19) Seela, F.; Sirivolu, V. R.; Chittepu, P. ; Bioconjugate Chem. 2008,
9, 211–224.
1
20) Zhang, J.; Xu, X.-D.; Wu, D.-Q.; Zhang, X.-Z.; Zhuo, R. Q.
Carbohydr. Polym. 2009, 77, 583–589.
(21) Meldal, M.; Tornøe, C. W. Chem. ReV. 2008, 108, 2952–3015.
(22) Ramirez, H. L.; Valdivia, A.; Cao, R.; Fragoso, A.; Labandeira, J. J. T.;
Ba n˜ os, M.; Villalonga, R. Polym. Bull. 2007, 59, 597–605.
(
23) Fern a´ ndez, M.; Villalonga, M. L.; Caballero, J.; Fragoso, A.; Cao,
R.; Villalonga, R. Biotechnol. Bioeng. 2003, 83, 743–747.
(24) Alvarez, S. G.; Alvarez, M. T. Synthesis 1997, 4, 412–413.
(25) Gujadhur, R.; Venkataraman, D.; Kintigh, J. T. Tetrahedron Lett. 2001,
42, 4791–4793.
Conclusion
(
26) Chan, T. R.; Hilgraf, R.; Sharpless, K. B.; Fokin, V. V. Org. Lett.
Three new series of ꢀCD-dextran polymers have been made
by “click chemistry”. To our knowledge these are the first ꢀCD
polymers based on native alkyne grafted dextran prepared by
2
004, 6, 2853–2855.
(27) Byun, H.-S.; Zhong, N.; Bittman, R. Org. Synth. 2004, 10, 690–692.
(
(
28) Gottlieb, H. E.; Kotlyar, V.; Nudelman, A. J. Org. Chem. 1997, 62,
512–7515.
7
“click” chemistry. The control of ꢀCD content, the high water
29) Willker, W.; Leibfritz, D.; Kerssebaum, R.; Bermel, W. Magn. Reson.
Chem. 1993, 31, 287–292.
(30) Keller, R. The Computer Aided Resonance Assignment Tutorial, 1st
ed.; CANTINA Verlag: Goldau, 2004.
31) Hein, C. D.; Liu, X.-M.; Wang, D. Pharm. Res. 2008, 25, 2216–2230.
solubility and the excellent binding properties should make them
very attractive in future biomedical and supramolecular studies,
such as DNA transfection or drug delivery. The bond variation
between the ꢀCD moieties and the dextran backbone of the
polymers leads to difference in chemical stability and the
biodegradability of the polymers is currently under investigation.
(
(
32) Bock, V. D.; Hiemstra, H.; van Maarseveen, J. H. Eur. J. Org. Chem.
2
006, 51–68.
(
33) Fleischmann, S.; Komber, H.; Appelhans, D.; Voit, B. I. Macromol.
Chem. Phys. 2007, 208, 1050–1060.
Acknowledgment. The authors would like to thank Phar-
macosmos A/S, Holbaek, Denmark, for generous supply of
dextran.
(34) Donnelly, P. S.; Zanatta, S. D.; Zammit, S. C.; White, J. M.; Williams,
S. J. Chem. Commun. 2008, 21, 2459–2461.
(
(
35) Rekharsky, M. V.; Inoue, Y. Chem. ReV. 1998, 98, 1875–1918.
36) Nielsen, T. T.; Wintgens, V.; Larsen, K. L.; Amiel, C. J. Inclusion
Phenom. Macrocyclic Chem. 2009, 65, 341–348.
Supporting Information Available. Various NMR spectra
and additional information regarding the AMDs and ꢀCD-
BM9013233