Development of micellar novel drug carrier utilizing temperature-sensitive block copolymers containing cyclodextrin moieties

Article abstract of DOI:10.1021/ma2013964

A drug-delivery system for albendazole (ABZ) based on β-cyclodextrin has been synthesized. Well-defined statistical copolymers, composed of N-isopropylacrylamide (NIPAAM) and trimethylsilylpropargyl acrylate (TMSPA), have been prepared by reversible addition-fragmentation chain transfer (RAFT) polymerization. The reactivity ratios were determined to be r_TMSPA = 1.12 and r_NIPAAm = 0.49, in the absence of RAFT agent, and r _TMSPA = 1.35 and r_NIPAAm = 0.35, in the presence of RAFT agent using the average of different techniques. Block copolymers were prepared using a POEGMEA₄₀ macro-RAFT agent chain extended with NIPAAm and TMSPA in various feed ratios. After deprotection, the polymers were reacted with 6I-azido-6I-deoxy-β-cyclodextrin via Huisgen azide-alkyne 1,3-dipolar cycloaddition, resulting in thermo-responsive block copolymers with pendant β-cyclodextrin groups, which were then acetylated to modify the polarity and inclusion-complex formation of β-cyclodextrin with the drug albendazole (ABZ). Only block copolymers with small amounts of cyclodextrin were observed to have an LCST while the copolymers containing higher β-cyclodextrin fractions increased the LCST of PNIPAAm beyond measurable temperature ranges. Encapsulation of ABZ increased the LCST. The loading efficiency increased in the polymer β-cyclodextrin conjugate compared to native β-cyclodextrin with the highest loading observed in the block copolymer after all remaining cyclodextrin hydroxyl groups had been acetylated. While β-cyclodextrin is toxic, attachment of a polymer lowered the toxicity to nontoxic levels. The ABZ-loaded polymers were all observed to be highly toxic to OVCAR-3 ovarian cancer cell lines with the acetylated polymer showing the highest toxicity.

Full text of DOI:10.1021/ma2013964

ARTICLE
pubs.acs.org/Macromolecules
Development of Micellar Novel Drug Carrier Utilizing
Temperature-Sensitive Block Copolymers Containing Cyclodextrin
Moieties
†
,‡
†
†,§
†
†
Firdaus Yhaya, Johnny Lim, Yoseop Kim, Mingtao Liang, Andrew M. Gregory,
,†
and Martina H. Stenzel*
†
Centre for Advanced Macromolecular Design (CAMD), School of Chemical Engineering, University of New South Wales,
Sydney, NSW 2052, Australia
‡
School of Industrial Technology, University Sains Malaysia, 11800 Minden, Penang, Malaysia
§
Cancer Research Laboratories, Department of Surgery, St. George Hospital, University of New South Wales, Sydney,
NSW 2217, Australia
S Supporting Information
b
ABSTRACT: A drug-delivery system for albendazole (ABZ)
based on β-cyclodextrin has been synthesized. Well-defined sta-
tistical copolymers, composed of N-isopropylacrylamide (NIPAAM)
and trimethylsilylpropargyl acrylate (TMSPA), have been pre-
pared by reversible additionꢀfragmentation chain transfer
(
RAFT) polymerization. The reactivity ratios were determined
to be rTMSPA = 1.12 and rNIPAAm = 0.49, in the absence of RAFT
agent, and rTMSPA = 1.35 and rNIPAAm = 0.35, in the presence of
RAFT agent using the average of different techniques. Block
copolymers were prepared using a POEGMEA40 macro-RAFT
agent chain extended with NIPAAm and TMSPA in various
feed ratios. After deprotection, the polymers were reacted with 6I-azido-6I-deoxy-β-cyclodextrin via Huisgen azideꢀalkyne 1,3-
dipolar cycloaddition, resulting in thermo-responsive block copolymers with pendant β-cyclodextrin groups, which were then
acetylated to modify the polarity and inclusion-complex formation of β-cyclodextrin with the drug albendazole (ABZ). Only block
copolymers with small amounts of cyclodextrin were observed to have an LCST while the copolymers containing higher
β-cyclodextrin fractions increased the LCST of PNIPAAm beyond measurable temperature ranges. Encapsulation of ABZ increased
the LCST. The loading efficiency increased in the polymer β-cyclodextrin conjugate compared to native β-cyclodextrin with the
highest loading observed in the block copolymer after all remaining cyclodextrin hydroxyl groups had been acetylated. While
β-cyclodextrin is toxic, attachment of a polymer lowered the toxicity to nontoxic levels. The ABZ-loaded polymers were all observed
to be highly toxic to OVCAR-3 ovarian cancer cell lines with the acetylated polymer showing the highest toxicity.
7
1. INTRODUCTION
antiproliferative activity compared to native cyclodextrins alone.
Despite the increase in solubility, this technique still required huge
amount of cyclodextrins and the inclusion complexes were too
small in size for drug delivery systems.
Recently, albendazole [ABZ, methyl(5-propylthio-1H-benz-
imidazol-2-yl)carbamate, Figure 1] has been identified as a
potential systemic anticancer agent besides its normal use as an
1
ꢀ8
Nanoparticles based on cyclodextrin are widely investigated in
order to combine the features of solubility enhancer with the
anthelmintic drug against human and animal parasites.
How-
ever, ABZ has low aqueous solubility, which limits its use for the
treatment of cancer.
1
3ꢀ17
enhanced permeationꢀretention effect (EPR) of nanoparticles.
The EPR of nanoparticles allows passive targeting of drugs via
the preferred lodgment of nanoparticles in the tumor while the
lymphatic system of the tumor is not capable of clearing the
polymer therefore, the drug carrier remains trapped. However,
many of these “cyclodextrin polymers” are undefined since they
In order to enhance the aqueous solubility of ABZ and its
bioavailability, several strategies have been proposed in the litera-
ture. Cylodextrins can be used to help solubilize poorly water-
soluble species by the formation of “inclusion complexes” or
9
,10
“
hostꢀguest complexes” (Figure 1).
Among several types
of cyclodextrins, used to form inclusion complexes with ABZ,
2
-hydroxypropyl-β-cyclodextrin was found to be useful in terms of
Received: June 20, 2011
11,12
bioavailability and complexation potential.
The solubilization
Revised:
September 12, 2011
of ABZ in cyclodextrin polymers has been found to enhance
Published: October 05, 2011
r 2011 American Chemical Society
8433
dx.doi.org/10.1021/ma2013964
|Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
(AIBN) was purified by recrystallization twice in methanol. Triethyl-
amine was dried using molecular sieves overnight prior to use and
β-cyclodextrin was recrystallized from water. Only copper sulfate was
purchased from BDH. All solvents used were of analytical grade, except
acetone and ethanol. Distilled water from Ultrapure was used through-
out this work. All chemicals were used as received unless stated
otherwise.
Synthesis and Methods. Synthesis of Trimethylsilylpropargyl
Acrylate (TMSPA). The synthesis of TMSPA was carried out in a two-
3
8
stage process. Propargyl acrylate (2-propynyl propenoate) was synthe-
sized first and later reacted with chlorotrimethylsilane to obtain
Figure 1. Inclusion complex of ABZ in cyclodextrin. The aliphatic part
of ABZ was suggested to be responsible for the formation of the
39
ꢀ1
TMSPA. Initially, acryloyl chloride (8.42 mL, 1.03 ꢁ 10 mol) was
added dropwise to a stirred solution of propargyl alcohol (5 mL, 8.60 ꢁ
10
inclusion complex.
ꢀ2
ꢀ1
10
mol) and triethylamine (14.4 mL, 1.03 ꢁ 10 mol) in dichloro-
were made by polycondensation of the native cyclodextrins lead-
methane (400 mL) at 0 °C. The clear solution turned yellow. The
reaction mixture was allowed to reach room temperature where the color
darkened. The mixture was stirred overnight and then quenched with
saturated sodium hydrogen carbonate solution. The organic layer was
extracted with 10% hydrochloric acid (3 ꢁ 30 mL), saturated sodium
hydrogen carbonate solution (1 ꢁ 30 mL), and water (1 ꢁ 30 mL),
dried over magnesium sulfate, filtered through neutral alumina, con-
centrated in vacuo to obtain green/yellow propargyl acrylate (87%
yield). In the second stage, silver chloride (1.56 g, 1.06 ꢁ
18
ing to branching and often insoluble networks.
Well-defined polymers with pendant cyclodextrin groups
are rarely discussed in the literature, compared to the branched
1
8
“
cyclodextrin polymers”. The synthesis of cyclodextrin poly-
mers using cyclodextrin monomers was first reported by Furue
19
and co-workers in 1975. Although most of the polymers pro-
18
duced were homopolymers, in some cases copolymerizations
1
8,20ꢀ25
were examined.
Alternatively, postfunctionalization of a
ꢀ2
reactive linear polymer with monofunctional cyclodextrin was
10 mol) was suspended in 154 mL of dry dichloromethane. Propargyl
1
8
26
ꢀ1
investigated. Seo et al. were among the first to use the two-
step polymer analogous reaction. An efficient route to polymers
with pendant cyclodextrin has been developed by Ritter and co-
workers using efficient Cu(I) catalyzed alkyne azide Huisgen
cycloaddition (click reaction). Despite all the recent advance-
ments in this area, the polymers were typically prepared via free
radical polymerization and more complex architectures such as
block copolymers are noticeably absent.
acrylate (12.4 mL, 1.12 ꢁ 10 mol) and DBU (21.4 mL, 1.43 ꢁ
ꢀ2
10 mol) were added to this suspension. A dark red color was observed.
The reaction mixture was then heated to 40 °C and chlorotrimethylsi-
ꢀ1
lane (21.2 mL, 1.59 ꢁ 10 mol) was added dropwise and stirring
continued for the next 24 h. The dark solution obtained was diluted with
n-hexane (400 mL) and the organic phase was washed successively with
saturated aqueous sodium hydrogen carbonate, hydrochloric acid (1%)
and water, dried over magnesium sulfate, filtered, and concentrated
in vacuo. The crude product obtained was purified by column chromatog-
raphy eluting with a 25:1 mixture of n-hexane and diethyl ether to obtain
RAFT polymerization allows the precise control over the macro-
molecular architecture with structures such as block copolymers
and other complex and elaborate architectures easily accessible
colorless trimethylsilylpropargyl acrylate liquid (36.7% yield). R = 0.71.
f
27ꢀ30
Synthesis of the PNIPAAM and PTMSPA Homopolymers. NIPAAM
through facile reactions.
The effortless design of amphiphi-
0
was recrystallized from n-hexane and 2,2 -azobis(isobutyronitrile)
lic block copolymers via RAFT opens the door to new self-
(
AIBN) was recrystallized twice from methanol. NIPAAM (3.39 g,
assembled structures including micelles, which are attractive
ꢀ
2
31ꢀ33
3.00 ꢁ 10 mol) was mixed with RAFT agent (3-BSPA) (32.6 mg, 1.20 ꢁ
nanomaterials for drug delivery.
Poly(ethylene glycol)
ꢀ
4
ꢀ5
ꢀ1
1
0
mol) inside a Schlenk tube. A stock solution of 1.20 ꢁ 10 mol L
(
PEG) is the polymer of choice to create the corona of the
of AIBN with N,N-dimethylacetamide (DMAc) was prepared and 10 mL
of this solution was added to the Schlenk tube. The tube was sealed and
degassed using five freezeꢀpumpꢀthaw cycle. The polymerization
mixture was immersed in an oil bath at 60 °C for different time intervals.
The polymers were analyzed using NMR and SEC analyses to determine
monomer conversion and molecular weight, respectively. For the syn-
micelle to enhance circulation time of the drug carrier, although
there are some indications that PEG might interact with
proteins. The design of the core of the micelle is inspired by
the work of Ritter and-co-workers who showed that the LCST of
NIPAAM can be influenced by having cyclodextrin as a polymer
building block. The changes in the LCST behavior is intrinsically
3
4
ꢀ2
thesis of PTMSPA homopolymer, TMSPA (5.46 g, 3.00 ꢁ 10 mol) of
24,35
linked to the type of guest.
Upon loading of ABZ into
was used, but other parameters and conditions were similar as of syn-
thesis of PNIPAAM homopolymer.
β-cyclodextrin and heating of the block copolymer above the
LCST of the block copolymer, micelles are expected to form.
Synthesis of P(NIPAAM-s-TMSPA) Statistical Copolymers. Statistical
copolymers P(NIPAAM-s-TMSPA) were prepared similar to the pro-
cedures above by varying the feed ratios between NIPAAM and TMSPA
monomers. The molar ratios were [NIPAAM]:[TMSPA] = 80:20,
[NIPAAM]:[TMSPA] = 90:10, and [NIPAAM]:[TMSPA] = 95:5.
The total monomer concentration for each copolymer system was kept
’
EXPERIMENTAL SECTION
Materials. The synthesis of RAFT agent, 3-(benzylsulfanylthio-
3
6,37
carbonylsulfanyl)propionic acid (3-BSPA), is described elsewhere.
ꢀ
1
3
-Mercaptopropionic acid, carbon disulfide, benzyl bromide, N-isopro-
pylacrylamide (NIPAAM), propargyl alcohol, acryloyl chloride, oligo-
ethylene glycol) methyl ether acrylate (OEGMEA) of M
at 3 mol L . For the synthesis of the copolymer with the feed ratio of
ꢀ
2
[NIPAAM]:[TMSPA] = 80:20, NIPAAM (2.71 g, 2.40 ꢁ 10 mol),
ꢀ
3
(
n
∼ 480 g
TMSPA (1.09 g, 6.00 ꢁ 10 mol), 3-BSPA RAFT agent (32.0 mg, 1.20
ꢀ4 ꢀ5
ꢀ
1
mol , chlorotrimethylsilane (CTMS), silver chloride, anhydrous mag-
nesium sulfate, 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), p-toluene-
sulfonyl chloride (PTSC), sodium azide, tetrabutylammonium fluo-
ride (TBAF), silica gel, sodium azide, ascorbic acid, albendazole, acetic
anhydride, pyridine, 4-dimethylaminopyridine (DMAP) were purchased
ꢁ 10 mol), and AIBN (1.97 mg, 1.20 ꢁ 10 mol) were added to
10 mL of DMAc. After the monomer conversion and molecular weight
had been determined, the polymer were purified by dialysis using tubular
membranes with a molecular weight cutoff (MWCO) of 3 500 Da for 2
days in ethanol to remove TMSPA and 2 days in water to remove
NIPAAM. The copolymers were then freeze-dried.
0
from Sigma-Aldrich and used as received. 2,2 -azobis(isobutyronitrile)
8
434
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
Synthesis of POEGMEA macro-RAFT Agent. In a 100 mL round-
study the feasibility of the reaction of β-cyclodextrin azide onto the co-
ꢀ
3
bottomed flask, OEGMEA (3.46 g, 7.20 ꢁ 10 mol) was polymerized
polymer, a model reaction was carried out. Propargyl alcohol was reacted
ꢀ
3
in toluene (2.4 mL) at 60 °C in the presence of the RAFT agent 3-BSPA
with β-cyclodextrin azide in DMF. 1.00 ꢁ 10 mol (5.61 g) of
ꢀ
4
ꢀ5
ꢀ3
(
32.6 mg, 1.20 ꢁ 10 mol) and AIBN (1.97 mg, 1.20 ꢁ 10 mol). The
molar ratio used was [OEGMEA]:[RAFT agent 3-BSPA]:[AIBN] =
0:1:0.1. The solution was purged with nitrogen for 45 min and the
propargyl alcohol and 1.00 ꢁ 10 mol (1.16 g) of β-cyclodextrin azide
were dissolved together in 10 mL of DMF. The reaction flask was sealed
with a rubber septum and the solution was stirred under a nitrogen atmo-
sphere for 24 h at 100 °C. The product was collected after precipitating
the solution into 100 mL of acetone followed by filtration. The liquid
6
polymerization was carried out for 3 h to obtain 44% conversion by
NMR. The POEGMEA macro-RAFT agent was purified by dialysis
against methanol to obtain a yellow viscous liquid.
phase was dialyzed for 4 days in water and freeze-dried. The dried
1
Theoretical number-average molecular weight calculated using conver-
product was analyzed by H NMR in DMSO-d
6
.
ꢀ
1
sion Mn(theo)= 19 200 g mol . Size exclusion chromatography (SEC)
Huisgen AzideꢀAlkyne 1,3-Dipolar Cycloaddition Reaction be-
tween β-Cyclodextrin Azide and Polymer. For the reaction of β-cyclo-
dextrin azide with various copolymers, two systems were tested: the
traditional Huisgen azideꢀalkyne 1,3-dipolar cycloaddition without
catalyst and the Cu(I) catalyzed approach (click chemistry), which was
ꢀ
1
determined number-average molecular weight Mn(SEC)= 20 600 g mol
,
PDI: 1.23 (polystyrene standards).
Synthesis of POEGMEA-b-P(NIPAAM-s-TMSPA) Block Copolymer.
The chain extension was carried out according to the method described
for P(NIPAAM-s-TMSPA) by replacing 3-BSPA with POEGMEA
macro-RAFT agent. The molar ratios [NIPAAM]:[TMSPA] = 80:20,
42
carried out according to the method described by Munteanu et al. The
noncopper system employed similar conditions as the model reaction.
1
[
NIPAAM]:[TMSPA] = 90:10, and [NIPAAM]:[TMSPA] = 95:5 were
H NMR (DMSO-d
ꢀN, COꢀOꢀCH
ꢀCH ꢀO, H4, H2), 1.5ꢀ2.3 (CH backbone), 1.0
6
): δ 5.73 (OH-3, OH-4), 4.84 (H1), 4.5
employed. For ratio of [NIPAAM]:[TMSPA] = 80:20, NIPAAM
(OH-6)), 4.1 (CH
3.5 (ꢀOꢀCH
(CH(CH ).
2
2
), 3.8 (CH(CH , H5, H6),
3 2
)
ꢀ
3
ꢀ3
(
2.71 g, 2.40 ꢁ 10 mol), TMSPA (1.09 g, 6.00 ꢁ 10 mol),
2
2
ꢀ4
POEGMEA macro-RAFT agent (2.30 g, 1.20 ꢁ 10 mol), and AIBN
3 2
)
ꢀ
5
(
1.97 mg, 1.20 ꢁ 10 mol) were added to 10 mL of DMAc. The samples
Acetylation of Copolymer. 20 mg of copolymer was mixed together
with 1 mL of acetic anhydride, 2 mL of pyridine in a round-bottomed
flask and 1 mg of DMAP was added as catalyst. The solution was stirred
at 50 °C for 24 h and dialyzed in dialysis tube 6000ꢀ8000 MWCO for 4
were then treated as described above. The block copolymers were purified
by dialysis in ethanol and then water followed by freeze-drying.
Deprotection of Copolymers. The cleavage of the trimethylsilyl
protecting group was carried out according to the method by Ladmiral
1
days with frequent water changes, followed by freeze-drying for 48 h. H
40
NMR (DMSO-d ): δ 4.8 (H1), 4.1 (CH ꢀN, COꢀOꢀCH ), 3.8
et al. with modifications to the amount of acetic acid and TBAF used
and the time taken to complete the reaction. The trimethylsilyl
protected polymer (150 mg) and acetic acid (2.0 equiv mol/mol with
respect to the alkyneꢀtrimethylsilyl groups) were dissolved in THF
6
2
2
(
CH(CH ) , H5, H6), 3.5 (ꢀOꢀCH ꢀCH ꢀO, H4, H2), 2.1 (CH -
3 2
2
2
3
CdO), 1.5ꢀ2.3 (CH backbone), 1.0 (CH(CH
3 2
) )
Measurement of Drug Loading Efficiency on Polymeric Micelles.
Initially, 20 mg of the β-cyclodextrin/copolymer and ABZ (mole ratio of
β-cyclodextrin moieties:ABZ = 1:1) were dissolved in 4 mL of water and
(
10 mL). Nitrogen was purged through the solution (ca. 20 min) and
then cooled to 0 °C. A 1 M solution of tetrabutylammonium fluoride
TBAF 3H O) in THF (2.0 equiv mol/mol with respect to the
alkyneꢀtrimethylsilyl groups) was added slowly via syringe (ca.
4
mL of THF (or acetone) separately. Both solutions were then mixed.
(
2
3
Water was added slowly and stopped as soon as precipitate started
forming. The water addition step is crucial in creating an aqueous
environment for ABZ, by doing so the ABZ is pushed into the
β-cyclodextrin cavities, forming the inclusion complexes. Omitting this
2
ꢀ3 min). The resulting turbid mixture was stirred at this temperature
for 30 min and then warmed to ambient temperature. The deprotection
was complete after 48 h, and the reaction solution was passed through a
short silica pad in order to remove excess TBAF and the pad was
subsequently washed with additional THF. The resulting solution was
then concentrated under reduced pressure, diluted in chloroform, and the
polymer was precipitated in petroleum ether 40ꢀ60 °C before centrifuga-
tion. The copolymer was dissolved in DMAc and dialyzed in dialysis tube
1
step would result in zero loading, as proven by H NMR (result not
shown). After 24 h of stirring at room temperature, THF was slowly
removed by vacuum. The unloaded drug was removed by means of
passing the solution through 0.45 μm filter and subsequently the sample
1
was freeze-dried for 48 h. For loading analysis via H NMR, equimolar
ratio of adamantane methanol to β-cyclodextrin moieties was added, and
whole system was dissolved in deuterated DMSO (DMSO-d) with 0.5
μL of styrene added as an internal standard. Styrene of known
concentrations was used to determine the concentration of ABZ by
comparing the integrals of styrene with the ABZ integrals. The loading
3
500 MWCO for 4 days against water, followed by freeze-drying.
Synthesis of 6I-Azido-6I-deoxy-β-cyclodextrin. β-Cyclodextrin was
monotosylated before azidification to obtain 6I-azido-6I-deoxy-β-cyclo-
41
dextrin (β-CD azide). Recrystallized β-cyclodextrin (63.3 g, 5.00 ꢁ
ꢀ
2
1
0
mol) was suspended in 500 mL water and stirred. A 5.65 g sample of
was determined from the vinylic peak of styrene (ꢀCdCH
2
, δ =
NaOH was dissolved in 20 mL of water and added dropwise into the
β-cyclodextrin solution. Meanwhile, p-toluenesulphonyl chloride (PTSC)
5
.7ꢀ6.0) and the ABZ ꢀCH ꢀCH peak (δ = 2.8ꢀ3.2). The drug
2
2
ꢀ
2
loading efficiency (DLE) was calculated according to
amount of ABZ in micelle
(
9.50 g, 6.00 ꢁ 10 mol) was dissolved in 30 mL of acetonitrile, and later
also added dropwise into the β-cyclodextrin solution. A white precipitate
was observed immediately. The reaction was run overnight and the white
solid was filtered and dried in the vacuum oven at 30 °C overnight. A yield
of 4.43 g of mono-6-p-toluenesulfonyl-β-cyclodextrin was obtained. For
the azide formation, dried mono-6-p-toluenesulfonyl-β-cyclodextrin(4.21
DLE ð%Þ ¼
ꢁ 100
amount of ABZ added initially
Other internal standards can be applied as long as they do not overlap
with existing signals. Addition of admantane methanol is recommended
to compete with ABZ for the cavity. The released ABZ was found to have
ꢀ
3
g, 3.26 ꢁ 10 mol) was reacted with 5 equiv of sodium azide in 20 mL of
1
anhydrous DMF at 80 °C overnight. The solution was precipitated in
clearer and more intensive H NMR signals.
acetone and dried under vacuum at 30 °C to obtain 6I-azido-6I-deoxy-β-
In Vitro Cytotoxicity Tests. Human ovarian cancer OVCAR-3 cells
1
cyclodextrin (β-cyclodextrin azide). H NMR (DMSO-d
6
): δ 6.0ꢀ5.90
were seeded in 96-well plates (3 000 cells per well) with culture medium
ꢀ
3
ꢀ1
(14H, OH-2, OH-3), 4.97 (d, 1H, H1
I
), 4.95ꢀ4.88 (m, 6H, H1IIꢀVI),
10% RPMI-1640 [2 ꢁ 10
M
L-glutamine, 1.5 g L
sodium
4
.67ꢀ4.53 (m, 6H, OH-6)), 3.90ꢀ3.60 (m, 28H, H3, H5, H6), 3.5ꢀ3.35
bicarbonate, 0.010 M of 2-hydroxyethylpiperazinesulfonic acid (HEPES),
ꢀ
1
ꢀ3
(m, 14H, H4, H2).
4.5 g L glucose, 1.00 ꢁ 10 M sodium pyruvate at 37 °C in 5% CO
2
Huisgen AzideꢀAlkyne 1,3-Dipolar Cycloaddition Model Reaction
environment for 24 h. The medium was refreshed with 0.2 mL of a
solution consisting of 0.1 mL medium and 0.1 mL of micelle solution of
between β-Cyclodextrin Azide and Propargyl Alcohol. In order to
8
435
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
Scheme 1. Schematic Approach to Thermo-Responsive Micelles Based on Block Copolymers with Pendant β-Cyclodextrin
ARTICLE
P(NIPAAM116-s-PA-β-cyclodextrin43), POEGMEA40-b-P(NIPAAM150-s-
PA-β-cyclodextrin11) and acetylated POEGMEA40-b-P(NIPAAM150-s-
PA-β-cyclodextrin11) micelles with and without ABZ loading to reach a
methanol with concentration of 1 mg/mL and filtered with 0.45 μm
filter. Mass spectrometry analyses were undertaken with a Thermo
Finnigan LCQ Deca quadruple ion trap mass spectrometer (Thermo
Finnigan, San Jose, CA), equipped with an atmospheric pressure
ionization source operating in the nebulizer assisted electrospray mode
and was used in positive ion mode. Mass calibration was performed using
caffeine, MRFA and Ultramark 1621 (Aldrich) in the m/z range of
195ꢀ1822 Da. All spectra was acquired within the m/z range of
150ꢀ2000 Da, and typical instrumental parameters were a spray voltage
of 4.5 kV, a capillary voltage of 44 V, a capillary temperature of 275 °C
and flow rate of 5 μL/min. Nitrogen was used as sheath gas (flow: 50%
maximum) and helium was used as auxiliary gas (flow: 5% maximum).
ꢀ
1
final micelle concentration of 62.5, 125, 250, and 500 μg mL
,
respectively, followed by incubation at 37 °C in the incubator for 72 h.
Subsequently, the medium was removed and washed 5 times with tap
water and 5 times with 1% acetic acid. After drying overnight, 100 μg of
0.010 M Tris (pH = 10.5) was added to solubilize the dye. Absorbance
was measured at 570 nm using Σ960 platereader (Metertech, Taiwan).
Nontreated cells were used as controls. The absorbance was measured at
570 nm and the optical density (OD) was used to calculate cell viability
[cell viability = (test ꢀ blank)/(control ꢀ blank) x 100]:
30 microscans, with maximum inject time of 10 ms per microscan, were
cellviability ð%Þ ¼
performed. For each respective scan, approximately 35 scans were
averaged to obtain the final spectrum. The solvent used was a 3:1 mix-
ture of dichloromethane:methanol with sodium acetate concentration of
½ðOD570, sample ꢀ OD570, blankÞ=OD570, control ꢀ OD570, blankÞꢂ ꢁ 100
0.3 μM. Sodium acetate was added to the solvent prior to analyses to
Self-Assembly and Thermal Properties of Micelles. Solutions of both
ABZ loaded and unloaded copolymers were prepared at a concentration
ensure ionization and to suppress potassium salt peaks. All theoretical
molecular weights were calculated using the exact mass for the first peak
in any given isotopic pattern. The molecular weights of the most
ꢀ
1
of 1 mg mL in water. The solution were placed in a quartz cuvette after
being passed through a 0.45 μm filter to remove the particle impurities.
The cuvette was placed in a dynamic light scattering (DLS) particle size
analyzer. The temperature was increased slowly from 20 to 80 °C, with
a 5 min stabilization period before measurement at each temperature.
The change of the average particles diameters or mean count rate vs tem-
perature was then observed. The temperature where scattering intensity
drastically increased (cloud point) was taken as the LCST of the co-
polymers being tested. Transmission electron microscopy (TEM) was
also used to observe the formation of micelles.
12
abundant isotopes were calculated using the following values: C
12.000000; H = 1.007825; O = 15.994915; Na = 22.989768.
=
1
16
23
Size Exclusion Chromatography (SEC). Molecular weight distribu-
tions of the copolymer systems were determined by means of SEC
using a Shimadzu modular system, comprising an autoinjector, a Polymer
2
Laboratories (PL) 5.0 μm bead-size guard column (50 ꢁ 7.5 mm ),
5
4
3
followed by three linear PL columns (10 , 10 , 10 ) and a differential-
refractive-index detector. The eluent was DMAc (0.05% w/v LiBr, 0.05%
ꢀ
1
2,6-dibutyl-4-methylphenol) at 50 °C with a flow rate of 1 mL min . The
Analysis. NMR Spectroscopy. NMR spectra were recorded using a
system was calibrated using narrowly dispersed polystyrene standards
6
ꢀ1
Bruker 300 MHz spectrometer; samples were analyzed in CDCl and
ranging from 500 to 10 g mol . The polymer (5 mg) was dissolved in
2 mL DMAc, followed by filtration using a filter with a pore size of 0.45 μm.
Dynamic Light Scattering (DLS). Hydrodynamic diameters of copo-
lymers in water were obtained using a Brookhaven Zetaplus particle size
analyzer. Samples were filtered before analyzing and the temperature
range used was from 20 to 80 °C.
3
DMSO-d
6
at 25 °C. For the determination of reactivity ratios, testing of
solvent choice and optimization of the NMR experiment were vital
preliminary experiments. The solvent DMSO-d and an H NMR
6
1
relaxation time of 1 s were deemed suitable.
FT-IR Spectroscopy. The FT-IR spectra were recorded on a Perkin-
ꢀ
1
Elmer FT-IR spectrometer. The scanning range was 400ꢀ4000 cm
Transmission Electron Microscopy (TEM). The TEM micrographs
were obtained using a JEOL 1400 transmission electron microscope.
The instrument operates at an accelerating voltage of 100 kV. Samples
were negative stained with phosphotungstic acid (2 wt %). A Formvar-
coated grid was coated by casting a polymer aqueous solution for 1 min.
Excess solution was removed using filter paper. For staining, a drop of
ꢀ1
and the resolution was 1 cm
.
Electrospray Ionization Mass Spectrometry (ESI-MS). ESI-MS was
used to confirm the existence of mono-6-p-toluenesulfonyl-β-cyclodextrin
and 6I-azido-6I-deoxy-β-cyclodextrin. Each sample was freshly prepared
before analysis by dissolving the product in a 1:1 solution of water:
8
436
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
Scheme 2. Synthesis of P(NIPAAM-s-TMSPA) Statistical Copolymer
Figure 3. Molecular weight distribution obtained from SEC of the
Figure 2. Pseudo first order kinetic plot showing the consumption of
TMSPA (square) and NIPAAM (triangle) of the copolymerization of
ꢀ1
copolymerization of NIPAAM and TMSPA at 60 °C. ([M] = 3 mol L
,
ꢀ2
ꢀ1
ꢀ3
ꢀ1
[
RAFT groups] = 1.20 ꢁ 10 mol L ,[AIBN] = 1.20 ꢁ 10 mol L
both monomers at 60 °C in DMAc in the presence of RAFT agent
in N,N-dimethylacetamide). From left to right, the molar ratios between
TMSPA]:[NIPAAM] were 20:80, 10:90, and 5:95.
ꢀ
1
ꢀ3
ꢀ1
3
3
-BSPA ([M] = 3 mol L , [AIBN] = 1.2 ꢁ 10 mol L , [RAFT agent
2 ꢀ1
[
-BSPA] = 1.2 ꢁ 10ꢀ mol L , [TMSPA]:[NIPAAM] = 5:95).
phosphotungstic acid was gently applied onto the surface of the grid for
0 s. The stained grid was dried in air.
radical polymerization is observed indicating a sluggish addition
of the macroradical to the RAFT agent (Figure S1, Supporting
Information). The copolymerization of NIPAAM and TMSPA
was subsequently investigated in detail at various feed ratios
3
’
RESULTS AND DISCUSSION
(
fTMSPA= 5, 10 and 20%) to establish the distribution of both
RAFT polymerization was chosen to construct the copoly-
monomers along the polymer chain (Scheme 2). In general, the
rate of polymerization declined with increasing amounts of
TMSPA (Figures S2 and S3, Supporting Information). The
consumption of both monomers was monitored independently
via NMR showing a preference for TMSPA in all cases with a
typical example displayed in Figure 2 (see as Table S1, Figures S2
and S3, Supporting Information, for details).
mers due its simplicity and easy control of polymer architecture.
Concerns about the toxicity of thiocarbonylthio RAFT end group
should be dismissed as the deprotecting stage and dialysis re-
move this group irreversibly. Before any block copolymer synthe-
sis as outlined in Scheme 1 was attempted, a detailed study on the
RAFT polymerization for the homopolymerization of PNIPAAM
and TMSPA and their respective copolymerization was carried
out. N,N-Dimethylacetamide (DMAc) was chosen as the solvent
for the polymerization. It was found that the DMAc dried in
magnesium sulfate gave better reproducibility as compared to
the DMAc dried in molecular sieves. The polymerization of
The polymers obtained as a result of the copolymerization
were all observed to have PDIs of approximately 1.3. The dif-
ference between predicted and experimental molecular weights
was very likely due to the SEC calibration using polystyrene
standards (Figure 3, Table S1, Supporting Information). Figure 3
display typical molecular weight distributions obtained from
statistical polymers P(NIPAAM116-s-TMSPA ), P(NIPAAM -s-
43
NIPAAM using 3-BSPA was reported earlier. The comonomer,
TMSPA, was homopolymerized using the same RAFT agent
resulting in polymers with reasonably low molecular weight
distribution, although a hybrid behavior between RAFT and free
43
84
TMSPA ), and P(NIPAAM -s-TMSPA ), which were gener-
31
54
19
ated after a polymerization time of 20 h with molar feed ratios of
8
437
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
Table 1. Reactivity Ratios Obtained by Different Approaches
with polydispersity index (PDI) of 1.23 was achieved, again suggest-
ing a controlled living system. Purification was carried out via dialysis
against methanol as the trial precipitation with n-hexane failed.
As outlined in Scheme 3, POEGMEA40 macro-RAFT was
utilized in a mixture with NIPAAM and TMSPA, using similar
reaction conditions as employed in the copolymerization, to
generate block copolymers. From the SEC curves in Figure 4, the
absence of byproduct and narrow PDI between 1.16 and 1.36
suggested the polymerization proceeded in a living manner. The
purification processes used for these block copolymers were
similar to the statistical copolymers. Akin to the copolymeriza-
tions, the three mol ratios of fNIPAAM = 95, 90 and 80%
were employed leading after a reaction time of 20 h to the
block copolymers POEGMEA -b-P(NIPAAM -s-TMSPA )],
free radical polymerization
RAFT polymerization
reactivity ratio
r
1
r
2
r
1
r
2
Fineman Ross
Kelen T €u d €o s
Mayo Lewis
CONTOUR
average
1.22
0.95
1.19
1.12
1.12
0.56
0.45
0.51
0.46
0.47
1.48
1.22
1.30
1.40
1.35
0.41
0.31
0.31
0.36
0.35
fTMSPA = 0.05, 0.1 and 0.2, respectively (Table S1, Supporting
Information).
40
150
11
According to these initial results, the copolymer should have a
gradient structure with the enrichment of TMSPA in the begin-
ning of the polymerization. To quantify this observation, the
reactivity ratios of this copolymerization in the presence and
absence of the RAFT agent were determined. The monomer feed
ratios were varied from fNIPAAM= 0.1 to fNIPAAM= 0.9 in 0.1
increments, hence, resulting in a set of nine samples. To achieve a
monomer conversion below 5%, imperative for a reactivity ratio
study, the free radical polymerization was stopped at 1 h or less
while the presence of the RAFT agent required a reaction time of
POEGMEA -b-(NIPAAM -s-TMSPA )], and POEGMEA -b-
40 137 18 40
(NIPAAM115-s-TMSPA )], respectively. The molecular weights
24
were in good agreement with the theoretical value while the
SEC curves showed monomodal distributions. In contrast to
the copolymerization using a low molecular weight RAFT
agent, the rate of polymerization of TMSPA and NIPAAM in
the presence of the macro-RAFT agent was less affected by
the monomer composition (Table 2). With increasing amount
of TMSPA only a slight drop in the rate of polymerization has
been observed. The reason is the lowered consumption of
TMSPA, compared to the copolymerization. The influence
of the macro-RAFT agent on the reactivity ratio of the
subsequent copolymerization has been described earlier as
3
h due to an inhibition period (Figures S2 and S3, Supporting
Information). The composition of the polymers after purification
1
was determined using H NMR spectrum as outlined in Figure
4
7
S4, Supporting Information. Detailed calculations to determine
FNIPAAM, the composition of the polymer, can be found in the
Supporting Information.
“bootstrap” effect. The hydrophilic macro-RAFT agent has a
preferred accumulation of hydrophilic NIPAAM around the
active RAFT end-group leading to the increased consumption
of the faster propagating NIPAAM. Comparison of actual
compositions of the polymer with the calculated composition using
the reactivity ratios in Table 1 (Figure S5, Supporting Information)
show the incorporation of TMSPA has been delayed in the block
copolymerization.
The mole fractions gathered from the NMR spectrum was
used to calculate reactivity ratios. Of the four salient methods,
Fineman Ross, Kelen T €u d €o s, Mayo Lewis, and the program
4
4,45
CONTOUR,
the latter three methods yielded harmonious
results, despite having incomparable calculation approaches
(
Table 1). The program CONTOUR, implemented by van
Huisgen AzideꢀAlkyne 1,3-Dipolar Cycloaddition with 6I-
Azido-6I-deoxy-β-cyclodextrin. Prior to reaction of 6I-azido-
6I-deoxy-β-cyclodextrin, the polymers were deprotected using
tetrabutylammonium fluoride. The polymer lost its color due to
the cleavage of the RAFT group during the procedure, probably
replaced by a hydrogen as earlier mass spectroscopy analysis
46
Herk, is generally recommended by IUPAC.
The reactivity ratios suggest that in both polymerization sys-
tems, free radical and RAFT, TMSPA (r ) chain ends have a
slight tendency to homopropagation (r g 1). NIPAAM (r ) is
1
1
2
slightly inclined to cross propagate in RAFT, thus, the slight
gradient structure of the polymer. The differences between free
radical polymerization and RAFT polymerization are not sig-
nificant, but visible. Similar to the finding by Barner-Kowollik
and co-workers, the polymer mole fraction of the monomer with
the larger reactivity ratio is increased in RAFT polymerization
compared to the conventional copolymerization. In summary,
as observed by the consumption of the individual monomers, the
initial stages of the polymerization consume more TMSPA, there-
fore an enrichment of TMSPA can be found at the α-terminal of
the polymer chain, where the R-group is positioned. The
composition drift can be visualized by calculating cumulative
F using the measured reactivity ratios (Figure S6, Supporting
Information).
Subsequently, block copolymers were prepared using OEGMA
as the building block due to the biocompatible properties of
the resulting polymer (Scheme 3). Before chain extension of
P(OEGMEA) macro-RAFT agent with NIPAAM and TMSPA,
POEGMEA homopolymer synthesis was achieved using 3-BSPA
at 60 °C in DMAc as a solvent. After 3 h of reaction time,
P(OEGMEA) homopolymer with a theoretical molecular weight
48
indicated. NMR analysis confirmed the efficient removal of the
protective group and the loss of the RAFT end group yielding
POEGMEA-b-P(NIPAAM-s-PA) (Figure 5). A procedure from
the literature has been adopted for the Huisgen azideꢀalkyne
1,3-dipolar cycloaddition (Scheme 4) using a copper-catalyzed
click system using copper(II) sulfate pentahydrate/ascorbic acid
46
42
at 140 °C. However, full copper removal was deemed impos-
sible. Several procedures were attempted from removal via silica
gel, extensive dialysis, to using a thiol complex and to washing
with EDTA. After significant product loss, the seemingly color-
less product was found to have still traces of Cu(I) ions, which
were observed to be cytotoxic. Therefore, the traditional Huisgen
azideꢀalkyne 1,3-dipolar cycloaddition in the absence of any
catalyst was employed. A model reaction between 6I-azido-6I-
deoxy-β-cyclodextrin and propargyl alcohol at 100 °C resulted in
complete reaction after 24 h with, as expected, the formation of
two stereoisomers as evidenced via NMR (Figure S7).
After the successful model reaction, four polymers—one
statistical polymer, P(NIPAAM116-s-PA ) and the three block
43
copolymers listed in Table 2—were reacted with 6I-azido-6I-
deoxy-β-cyclodextrin. The modification process was confirmed
40
ꢀ
1
of 19 200 g mol and a narrow molecular weight distribution
8
438
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
Scheme 3. Synthesis of POEGMEA-b-P(NIPAAM-s-TMSPA) Block Copolymer
modification (4.7 ppm) shifted to 4.2 ppm after reaction
Figure 5). Even though the peaks assigned to the proton in
(
the triazole ring (7.9ꢀ8.4 ppm) appeared in the spectrum of
model reaction, these peaks were not detected in the polymer
peaks, possibly due to increased relaxation times. In order to
confirm that the cyclodextrin peaks were not from free
(
7
unreacted) β-cyclodextrin azide, the polymer was dialyzed for
days in water using a tubular dialysis membrane with a mol-
ecular weight cutoff of 6000ꢀ8000 Da. FT-IR analysis of 6I-
ꢀ
1
azido-6I-deoxy-β-cyclodextrin exhibited a sharp peak at 2000 cm
assigned to the azide functional group, which disappeared after
ꢀ
1
reaction. The broad peak at 3000ꢀ3500 cm is indicative for
the hydroxyl groups of cyclodextrin, which were absent before
reaction (Figure S8, Supporting Information). The SEC con-
firmed the increase of the molecular weight of monomodal peak
of POEGMEA-b-(NIPAAM-s-PA) to the higher molecular weight
of POEGMEA-b-P(NIPAAM-s-PA-β-cyclodextrin) (Figure 6).
Interestingly, the are no significant weight differences between
the product obtained via Huisgen azideꢀalkyne 1,3-dipolar
cycloaddition with and without Cu(I) catalyst.
Figure 4. Molecular weight distribution obtained from SEC of the
block copolymerization of NIPAAM and TMSPA (fNIPAAM= 95%) at
The statistical polymers prior to β-cyclodextrin modification
are insoluble in water while the block copolymers have clearly
amphiphilic properties. The presence of TMSPA or, after depro-
tection, propargyl acrylate clearly lowers the water solubility of
PNIPAAM. With conjugation of the hydrophilic β-cyclodextrin,
the polymers become fully water-soluble. The presence of a co-
monomer will ultimately affect the lower critical solution tem-
perature (LCST) of PNIPAAM. The LCST of all polymers was
determined via dynamic light scattering (DLS) using the scatter-
ing intensity as a means of determining the temperature. The
onset of the increase coincides with the formation of more and
bigger particle due to precipitation or micelle formation
(Figure 7). The hydrophobic comonomer TMSPA lowered the
LCST in all cases to such an extent that these polymers are
insoluble in water at room temperature. In contrast, after reaction
with 6I-azido-6I-deoxy-β-cyclodextrin, the LCST increased to
6
3
1
0 °C with the presence of POEGMEA macro-RAFT agent. ([M] =
ꢀ
1
ꢀ2
ꢀ1
mol L , [RAFT groups] = 1.20 ꢁ 10 mol L , [AIBN] = 1.20 ꢁ
ꢀ
3
ꢀ1
0
mol L in N,N-dimethylacetamide. The polymerization times of
the curves shown are 0, 4, 8, 12, 24, and 48 h.
1
by H NMR (Figure 5), FT-IR (Figure S8, Supporting In-
formation), and SEC (Figure 6). SEC traces from Figure 6 shows
that the SEC curves are monomodal and high molecular weight
shoulders are absent suggesting that no linkages formed between
polymer chains. This reinforces the purity of 6I-azido-6I-deoxy-
β-cyclodextrin and the presence of bifunctional cyclodextrin can
be assumed absent. It should be noted here that the results for
both block copolymers and statistical polymers are similar, unless
49
1
stated otherwise. From H NMR analysis, the modification
process with 6I-azido-6I-deoxy-β-cyclodextrin was observed to
be complete. The ꢀCH ꢀ signal of the propargyl ester before
2
8
439
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
1
Figure 5. H NMR analysis of POEGMEA40-b-P(NIPAAM115-s-TMSPA24) before deprotection (A) and after deprotection (B), 6I-azido-6I-deoxy-β-
1
cyclodextrin (C), and POEGMEA40-b-P(NIPAAM115-s-PA-β-cyclodextrin24) (D). H NMR spectra for 1 and 2 were obtained in CDCl
for 3 and 4 were obtained in deuterated DMSO.
3
, while spectra
Scheme 4. Huisgen AzideꢀAlkyne 1,3-Dipolar Cycloaddition with 6I-Azido-6I-deoxy-β-cyclodextrin and Subsequent
Acetylation
temperatures of well above the LCST of PNIPAAM. The LCST
of the statistical copolymer increases with increasing fraction of
Figure 6. The LCST is slightly lower than the statistical block
copolymer owing to the reduced presence of TMSPA, hence,
subsequently, yielding a lower β-cyclodextrin content. Block
copolymers prepared with a higher TMSPA feed radio, however,
did not possess a visible LCST and the scattering intensity
remain unaffected over the range of temperatures measured.
24,25,50
β-cyclodextrin in agreement with earlier studies.
Interest-
ing is the behavior of the prepared block copolymers. Block
copolymers with a low content of β-cyclodextrin still show
similar curves to the statistical block copolymer as depicted in
8
440
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
Table 2. Summary of Block Copolymers Obtained Using POEGMEA -macro-RAFT with Different Ratios of TMSPA and
ARTICLE
4
0
NIPAAM after Polymerizing for 20 h
before deprotection
after deprotection and CD modification
POEGMEA-b-P(NIPAAM-s-PA-β-
cyclodextrin)
POEGMEA-b-P(NIPAAM-
s-TMSPA)
feed
ratio
calculated composition
from conversion
cum FTMSPA(theor Mn,theo/g
M
n,GPC/g mol-1
M
mol‑1
n,theo/g
Mn,GPC/g
mol‑1 (PDI)c
a
b
mol‑1
c
% convn
cum FTMSPA
)
(PDI)
f
f
f
TMSPA= 0.05 64 POEGMEA40-b-P(NIPAAM150-s-TMSPA11
)
)
)
7% (7.4%)
11% (14.2%)
17% (29%)
38 150
38 000
36 570
35500 (1.65)
33000 (1.55)
27000 (1.51)
50 150
57 550
62 700
57500 (1.56)
65000 (1.51)
54000 (1.59)
TMSPA= 0.1
TMSPA= 0.2
62 POEGMEA40-b-P(NIPAAM137-s-TMSPA18
56 POEGMEA40-b-P(NIPAAM115-s-TMSPA24
a
b
overall conversion of both monomers. Cumulative composition of the P(TMSPA-s-NIPAAM) block at the listed conversion calculated using the
reactivity ratios in Table 1 (CONTOUR) for RAFT (Figure S6, Supporting Information, for graph). DMAc, measured against PS standards.
c
Figure 6. SEC traces of POEGMEA-b-P(NIPAAM-s-PA) before reac-
_
__
Figure 7. Scattering intensity vs temperature of POEGMEA40-b-P-
NIPAAM150-s-PA-β-cyclodextrin11)], before and after acetylation,
in water.
tion ( ), POEGMEA-b-P(NIPAAM-s-PA) clicked with 6I-azido-6I-
deoxy-β-cyclodextrin using copper catalyst at 140 °C for 30 min ( ),
(
3
3 3
and POEGMEA-b-P(NIPAAM-s-PA)] modified with β-cyclodextrin
azide at 100 °C for 24 h (----).
below the LCST may not only be caused by inclusion complex
formation, but also by the formation of strong hydrogen bonding
between two cyclodextrin molecules, a process that is especially
prevalent with polymers that contain higher concentrations of
This may be explained by the absence of phase separation
between both blocks. While it is expected that above the LCST,
POEGMA and PNIPAAM form two immiscible blocks, it also
needs to be considered that cyclodextrin undergo inclusion
complex formation with many polymers including poly(ethylene
oxide) (PEO), although α-cyclodextrin is more suitable for the
50
cyclodextrin moieties.
For further variation of the LCST, the hydroxyl groups of
β-cyclodextrin were acetylated using a standard procedure. As a
result, the hydrophilicity of the building block was lowered,
which is reflected by a LCST of 34 °C but also by a strong
tendency to form a high fraction of micelles with a hydrodynamic
diameter of 100 nm even at room temperature. At the LCST, the
core of the micelles collapses forming particles of 30 nm.
Albendazole Drug Loading and Cell Toxicity Tests. The
51ꢀ54
inclusion complex formation with PEO.
gels have been generated using this approach,
hydrogels from PEO and cyclodextrin only are not very stable
A range of hydro-
54,55
although
54
in the presence of high amount of water. This hostꢀguest
formation could force the hydrophilic PEO into the PNIPAAM
environment shifting the LCST to values above the measured
range. Evidence can be found in the aggregate formation below
the LCST. Theoretically, below the LCST, aggregate formation
should be absent due to the water-solubility of both blocks.
However, all block copolymers listed in Table 3 show small
round particles of diameters of 10ꢀ15 nm under the TEM
accompanied by a large fraction of undefined large particles with
ꢀ
4
ꢀ1
insolubility of albendazole in water (2.00 ꢁ 10 g L or 7.57 ꢁ
ꢀ
6
ꢀ1 56
10 mol L
)
requires the consumption of vast amounts of
drug to treat any diseases in a meaningful way. Cyclodextrins
10,57,58
have been chosen earlier to address the low water-solubility.
phase diagrams have been studied in detail confirming the for-
1
0,58
mation of 1:1 complexes between β-cyclodextrin
analysis suggest inclusion via the propyl group as depicted in
while NMR
1
00ꢀ500 nm (Figure 8). Above the LCST and the dehydration
of PNIPAAM, the formation of micelles with sizes of around
0 nm (TEM) or 50 nm (DLS) takes place. Aggregate formation
10
Figure 1. Although the ABZ-CD complex has been thoroughly
studied in literature with a range of techniques, NMR studies are
4
8
441
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
Table 3. LCST Values Determined Using DLS (Scattering Intensity) before and after ABZ Loading
ꢀ1
sample
polymer composition
F/%
LCST/°C
LCST (after ABZ loading)/°C
K 1:1/ L mol
1
2
3
4
5
6
β-cyclodextrin
P(NIPAAM116-s-PA-β-cyclodextrin43
-
27
7
-
42
38
-
-
46
1600 ( 5%
1600 ( 5%
3950 ( 5%
3950 ( 5%
4200 ( 5%
)
POEGMEA40-b-P(NIPAAM150-s-PA-β-cyclodextrin11)]
POEGMEA40-b-P(NIPAAM137-s-PA-β-cyclodextrin18)]
POEGMEA40-b-P(NIPAAM115-s-PA-β-cyclodextrin24)]
POEGMEA40-b-P(NIPAAM150-s-PA-β-acetyl-cyclodextrin11)]
42
11
17
7
>55
>55
48
-
34
37000 ( 5%
Figure 8. TEM images of block copolymer POEGMEA40-b-P(NIPAAM150-s-PA-β-cyclodextrin11)] particles using phosphotungstic acid negative
staining at 25 °C (left) and 55 °C (right).
ꢀ
1
limited. In most cases, the shift of β-cyclodextrin signals have
been monitored, but also the ABZ signals underwent some
changes. As seen in Figure S8, Supporting Information, confirm-
tested. The apparent binding constant of 1600 L mol was
obtained with the polymer listed in Table 3, but also with other
copolymers prepared with different feed ratios (Figure 3), there-
fore loading was not affected by the cyclodextrin content in the
polymer, in contrast to other work where the interaction was
1
ing various studies in literature, the H NMR signals correspond-
ing to cyclodextrin is dependent on the amount of albendazole
added with the largest changes occurring at a 1:1 molar ratio bet-
ween β-CD and ABZ (ESI, Figure S8, Supporting Information).
The structure of the complex as shown in Figure 1 has been
6
0
reduced when CD was conjugated to a polymer. On the
contrary, the block copolymers with POEGMA lead to double
the ABZ loading. This can be understood by the role of PEG
chains to contribute to the overall enhancement of the solubility
of ABZ delaying precipitation of ABZ when the organic solvent is
removed from aqueous system. A significant jump in the com-
plex formation can be observed after acetylation (Table 3). The
effect of the modification of hydroxyl groups on the complex
10
derived from the shift of the propyl signals of ABZ. The
aliphatic protons were shielded in the presence of β-CD, but it
is not clear to what extend the aromatic group may be involved.
However, so far no two-dimensional rotating frame NOE
59
spectroscopy (ROESY) has been carried out to elucidate the
structure of the hostꢀguest complex. However, even after many
attempts, we failed to obtain suitable cross peaks and the
formation of the complex was only characterized by establishing
the phase-solubility diagram. The apparent binding constant
between ABZ and β-cyclodextrin was calculated from this
6
1,62
formation has been discussed earlier.
Acetylation will affect
the size of the cavity and also the type of forces between CD and
63
drug.
The drug is encapsulated via the hydrophobic tail while the
more hydrophilic part is exposed to the outside. This exposed
part of the drug molecule can now manipulate the overall hydro-
10
diagram using a method developed by Higushi and Connors.
2
4,35
The apparent binding constant K1:1 = slope/S (1ꢀslope) was
philicity of the PNIPAAM copolymer influencing the LCST.
o
obtained from the slope of the linear relationship between
cyclodextrin concentration and the amount of encapsulated
ABZ and the initial solubility of ABZ in the absence of cyclodex-
trin S . Values of K for this inclusion complex formation was
The measured LCST values were observed to be slightly
increased owing to the presence of the hydrophilic carbamate
part of the drug ABZ (Table 3). The drug molecule also imparts
now stimuli-responsive properties to the block copolymers with
higher β-cyclodextrin content while a visible LCST was absent in
the absence of ABZ. Drug loading obviously facilitates phase
separation between PNIPPAM and the POEGMA block. The
changes in the LCST in the acetylated polymer were much more
pronounced due to the much higher loading of ABZ.
The ABZ loaded polymers were tested as drug delivery
carriers. The statistical copolymer and the block copolymer are
tested against β-cyclodextrin in their performance. The aim is to
have a nontoxic carrier while the drug-carrier system should show
0
1:1
ꢀ158
ꢀ1
reported as 1382 L mol
and 965 L mol (3.643 mL
ꢀ
1 10
mg ). Loading was usually achieved by mixing an aqueous
cyclodextrin solution with ABZ, which is dissolved in an organic
solvent such as THF or acetone. After incubation, the organic
solvent was removed under vacuum and excess ABZ was filtered
off. Using this approach, we obtained a slightly higher value for
ꢀ
1
the inclusion complex formation with K1:1 =1600 L mol
Table 3). The loading capacity was not affected when the
statistical copolymers P(NIPAAM-s-PA-β-cyclodextrin) were
(
8
442
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
Figure 9. Cell viability test using OVCAR-3 ovarian cancer cell lines with β-cyclodextrin, statistical copolymer P(NIPAAM116-s-PA-β-cyclodextrin43),
and the block copolymer POEGMEA -b-P(NIPAAM -s-PA-β-cyclodextrin ) and the acetylated block copolymer before ABZ loading (left) and
40
150
11
after ABZ loading (right).
high cytotoxicity. It should be noted here that the block
copolymer is not expected to form micelles under these condi-
tions. LCST values seem to suggest that the polymers are water-
soluble at the cell test temperature of 37 °C. In addition, the
stability of block copolymer micelles is highly challenged in cell
cell death. An interesting result was obtained when comparing
two polymers. Although the complex formation constant be-
tween ABZ and β-cyclodextrin in the block copolymer is slightly
higher, the overall amount of cyclodextrin is lower in this system.
As a consequence, the amount of ABZ encapsulated by the same
vector concentration is lower in POEGMEA40-b-P(NIPAAM150-s-
PA-β-cyclodextrin11). However, the cytotoxicity of the block
copolymer is significantly higher leading to more than 90% cell
death even at low concentration.
A remarkable effect is observed using the acetylated polymer.
Even very small concentrations lead to more than 90% of cell
death. This drug carrier is therefore highly effective owing to the
higher ABZ loading, but could also have other origins.
34
growth media resulting often in disaggregation. However, the
presence of PEG can increase biocompatibility and possibly
enhance cellular uptake of the drug carrier. Results from pre-
liminary studies have suggested that ABZ is an efficient antic-
8
ancer drug against ovarian cancer. OVCAR-3 cell lines were
chosen to test the activity of the ABZ loaded systems. Parent
(unmodified) β-cyclodextrin, statistical copolymer, and block
copolymer were tested for toxicity against cell lines before and
after ABZ loading. Initially, the biocompatibility of β-cyclodex-
trin and its polymers were tested on viable OVCAR-3 cells.
Figure 9 shows the percentage viability of OVCAR-3 cell relative
to the control sample after 72 h incubation at 37 °C with four
different vectors concentration of 62.5, 125, 250, and 500
To understand these effects further studies are warranted. The
increased toxicity may be the preferred cellular uptake of carriers
carrying PEG moieties by the cell, but further studies, including
drug-release rates, are needed to understand this effect.
ꢀ
1
μg mL , which is equivalent to different amountsofcyclodextrin.
Therefore, the cell viability was recorded against the concentration
of cyclodextrin (Figure 9). β-cyclodextrin alone was found to be
toxic to cells, even when unloaded with drug. This has been
’
CONCLUSION
A biocompatible drug delivery system has been created which
combines the hostꢀguest complexation of β-cyclodextrin with
the possibility of targeting a tumor passively by employing
polymers. The hostꢀguest complexation between the drug
albendazole and the cyclodextrin carrying polymer was highly
dependent on the fine-structure of the polymer. Loading effi-
ciencies affected the performance of the drug carrier but also the
LCST of the PNIPAAm block. Higher loading led to increased
cytotoxicity but also to increased LCST values since the hostꢀ
guest complex mainly buries the hydrophobic part of the drug
while the hydrophilic part is exposed. The cytotoxicity results can
potentially be correlated to the loading capacity but more studies
are needed to fully understand how the structure of the polymer
affects drug release and cellular uptake.
64ꢀ67
observed earlier
and the toxicity of β-cyclodextrin may be
attributed to its uptake by cells, leading to the disruption of
intracellular function or the extraction of lipid membrane compo-
65
nents such as cholesterol and phospholipids. The statistical
copolymer also showed its toxicity as the concentration increased
while the block copolymer, with its PEG chains counterbalanced
the toxicity of β-cyclodextrin.
The polymers were subsequently loaded with ABZ. Vectors
concentration, incubation time, and temperature were the same
as before. Even after ABZ loading, the statistical copolymer was
more toxic than the block copolymer. ABZ loaded β-cyclodextrin
is as cytotoxic as β-cyclodextrin alone resulting in more than 90%
8
443
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
’
ASSOCIATED CONTENT
(20) Liu, Y.-y.; Fan, X.-d.; Gao, L. Macromol. Biosci. 2003, 3 (12),
15–719.
(21) Carbonnier, B.; Janus, L.; Deratani, A.; Morcellet, M. J. Appl.
7
S
Supporting Information. Evolution of M vs the conver-
b
n
sion, pseudo-first order kinetic plots, characteristics of homopolymers
Polym. Sci. 2005, 97 (6), 2364–2374.
22) Gonzalez, H.; Hwang, S. J.; Davis, M. E. Bioconjugate Chem.
999, 10 (6), 1068–1074.
23) Cheng, J.; Khin, K. T.; Jensen, G. S.; Liu, A.; Davis, M. E.
Bioconjugate Chem. 2003, 14 (5), 1007–1017.
24) Amajjahe, S.; Choi, S.; Munteanu, M.; Ritter, H. Angew. Chem.,
Int. Ed. 2008, 47 (18), 3435–3437.
25) Amajjahe, S.; Munteanu, M.; Ritter, H. Macromol. Rapid
Commun. 2009, 30 (11), 904–908.
26) Seo, T.; Kajihara, T.; Iijima, T. Makromol. Chem. 1987, 188 (9),
071–82.
1
(
and random copolymers H NMR and FTIR spectra, calculation
1
of polymer composition FNIPAAM and FTMSPA from NMR,
molecular weight distribution obtained from SEC, and calculated
cumulative polymer composition. This material is available free
of charge via the Internet at http://pubs.acs.org.
(
(
(
’
AUTHOR INFORMATION
(
Corresponding Author
2
*Corresponding author. E-mail: m.stenzel@unsw.edu.au.
(27) Perrier, S.; Takolpuckdee, P. J. Polym. Sci., Part A: Polym. Chem.
2
005, 43 (22), 5347–5393.
(
28) Favier, A.; Charreyre, M. T. Macromol. Rapid Commun. 2006,
27 (9), 653–692.
29) Moad, G.; Rizzardo, E.; Thang, S. H. Aust. J. Chem. 2009, 62
11), 1402–1472.
’
ACKNOWLEDGMENT
(
M.H.S. thanks the ARC (Australian Research Council) for a
(
Future Fellowship (FT0991273) and funding for this project
DP110102409). The authors also thank the staff from the NMR
(30) Moad, G.; Thang, S. H. Aust. J. Chem. 2009, 62 (11), 1379–1381.
(31) Stenzel, M. H. Chem. Commun. 2008, 30, 3486–3503.
(32) Stenzel, M. H. Macromol. Rapid Commun. 2009, 30 (19),
(
facility (Jim Hook, Dr Donald Thomas) at UNSW for their help
and support. F.Y. also acknowledges Ministry of Higher Educa-
tion Malaysia for sponsoring his doctorate studies.
1603–1624.
(33) Gregory, A.; Stenzel, M. H. Expert Opin. Drug Delivery 2011, 8
(2), 237–269.
(
34) Kim, S.; Shi, Y. Z.; Kim, J. Y.; Park, K.; Cheng, J. X. Expert Opin.
’
REFERENCES
Drug Delivery 2010, 7 (1), 49–62.
(
1) Pourgholami, M. H.; Akhter, J.; Wang, L.; Lu, Y.; Morris, D. L.
Cancer Chemother. Pharmacol. 2005, 55 (5), 425–432.
2) Pourgholami, M. H.; Woon, L.; Almajd, R.; Akhter, J.; Bowery,
P.; Morris, D. L. Cancer Lett. (Shannon, Ireland) 2001, 165 (1), 43–49.
3) Morris, D. L.; Jourdan, J.-L.; Pourgholami, M. H. Oncology 2001,
1 (1), 42–46.
4) Hossein Pourgholami, M.; Cai, Z. Y.; Lu, Y.; Wang, L.; Lawson
Morris, D. Clin. Cancer Res. 2006, 12 (6), 1928–1935.
5) Khalilzadeh, A.; Wangoo, K. T.; Morris, D. L.; Pourgholami,
M. H. Biochem. Pharmacol. 2007, 74 (3), 407–414.
6) Jung, H.; Medina, L.; Garcia, L.; Fuentes, I.; Moreno-Esparza, R.
J. Pharm. Pharmacol. 1998, 50 (1), 43–48.
7) Pourgholami, M. H.; Wangoo, K. T.; Morris, D. L. Anticancer Res.
008, 28 (5A), 2775–2779.
8) Kim, Y.; Pourgholami, M. H.; Morris, D. L.; Stenzel, M. H.
Macromol. Bioscience 2011, 11 (2), 219–233.
(35) Maciollek, A.; Munteanu, M.; Ritter, H. Macromol. Chem. Phys.
2010, 211 (2), 245–249.
(36) Stenzel, M. H.; Davis, T. P. J. Polym. Sci., Part A: Polym. Chem.
(
2002, 40 (24), 4498–4512.
(
(37) Stenzel, M. H.; Cummins, L.; Roberts, G. E.; Davis, T. P.; Vana,
P.; Barner-Kowollik, C. Macromol. Chem. Phys. 2003, 204 (9), 1160–1168.
(38) Harvey, D. F.; Lund, K. P.; Neil, D. A. J. Am. Chem. Soc. 1992,
114 (22), 8424–34.
(39) Malkoch, M.; Thibault, R. J.; Drockenmuller, E.; Messerschmidt,
M.; Voit, B.; Russell, T. P.; Hawker, C. J. J. Am. Chem. Soc. 2005, 127 (42),
14942–14949.
(40) Ladmiral, V.; Mantovani, G.; Clarkson, G. J.; Cauet, S.; Irwin,
J. L.; Haddleton, D. M. J. Am. Chem. Soc. 2006, 128 (14), 4823–4830.
(41) Khan, A. R.; Forgo, P.; Stine, K. J.; D’Souza, V. T. Chem. Rev.
1998, 98 (5), 1977–1996.
6
(
(
(
(
2
(
(42) Munteanu, M.; Choi, S.; Ritter, H. Macromolecules 2008, 41
(24), 9619–9623.
(9) Szejtli, J. Pure Appl. Chem. 2004, 76 (10), 1825–1845.
(
10) Joudieh, S.; Bon, P.; Martel, B.; Skiba, M.; Lahiani-Skiba, M.
(43) Zhang, L.; Bernard, J.; Davis, T. P.; Barner-Kowollik, C.;
Stenzel, M. H. Macromol. Rapid Commun. 2008, 29 (2), 123–129.
(44) vanHerk, A. M. CONTOUR (v1.8).
(45) vanHerk, A. M.; Droge, T. Macromol. Theory Simul. 1997, 6 (6),
1263–1276.
(46) Feldermann, A.; Toy, A. A.; Phan, H.; Stenzel, M. H.; Davis,
T. P.; Barner-Kowollik, C. Polymer 2004, 45 (12), 3997–4007.
(47) Pai, T. S. C.; Barner-Kowollik, C.; Davis, T. P.; Stenzel, M. H.
Polymer 2004, 45 (13), 4383–4389.
(48) Quemener, D.; Le Hellaye, M.; Bissett, C.; Davis, T. P.; Barner-
Kowollik, C.; Stenzel, M. H. J. Polym. Sci. Part A: Polym. Chem. 2008, 46
(1), 155–173.
J. Nanosci. Nanotech. 2009, 9 (1), 132–140.
11) Uekama, K.; Hirayama, F.; Irie, T. Chem. Rev. 1998, 98 (5),
045–2076.
12) Evrard, B.; Chiap, P.; DeTullio, P.; Ghalmi, F.; Piel, G.; Van
Hees, T.; Crommen, J.; Losson, B.; Delattre, L. J. Controlled Release
002, 85 (1ꢀ3), 45–50.
13) Cheng, J.; Khin, K. T.; Davis, M. E. Mol. Pharmaceutics 2004, 1
3), 183–193.
14) Ramirez, H. L.; Valdivia, A.; Cao, R.; Torres-Labandeira, J. J.;
Fragoso, A.; Villalonga, R. Bioorg. Med. Chem. Lett. 2006, 16 (6),
499–1501.
15) Skiba, M.; Bounoure, F.; Barbot, C.; Arnaud, P.; Skiba, M.
J. Pharm. Pharmaceut. Sci. 2005, 8 (3), 409–418.
16) Pun, S. H.; Bellocq, N. C.; Liu, A.; Jensen, G.; Machemer, T.;
(
2
(
2
(
(
(
1
(
(49) Schild, H. G. Prog. Polym. Sci. 1992, 17 (2), 163–249.
(50) Choi, S.; Munteanu, M.; Ritter, H. J. Polym. Res. 2009, 16 (4),
389–394.
(
Quijano, E.; Schluep, T.; Wen, S.; Engler, H.; Heidel, J.; Davis, M. E.
Bioconjugate Chem. 2004, 15 (4), 831–840.
(51) Tonelli, A. E. Macromolecules 2008, 41 (11), 4058–4060.
(52) Tonelli, A. E. Polymer 2008, 49 (7), 1725–1736.
(53) Harada, A.; Kamachi, M. Macromolecules 1990, 23 (10),
2821–2823.
(54) van de Manakker, F.; Vermonden, T.; van Nostrum, C. F.;
Hennink, W. E. Biomacromolecules 2009, 10 (12), 3157–3175.
(55) Zhu, W.; Li, Y.; Liu, L.; Chen, Y.; Wang, C.; Xi, F. Biomacro-
molecules 2010, 11 (11), 3086–3092.
(
17) Zhang, J.-T.; Huang, S.-W.; Zhuo, R.-X. Macromol. Chem. Phys.
004, 205 (1), 107–113.
18) Yhaya, F.; Gregory, A. M.; Stenzel, M. H. Aust. J. Chem. 2010, 63
2), 195–210.
19) Furue, M.; Harada, A.; Nozakura, S. J. Polym. Sci., Polym. Lett.
Ed. 1975, 13 (6), 357–60.
2
(
(
(
8
444
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445

Macromolecules
ARTICLE
(56) Evrard, B.; Chiap, P.; DeTullio, P.; Ghalmi, F.; Piel, G.; Van
Hees, T.; Crommen, J.; Losson, B.; Delattre, L. J. Controlled Release
2
002, 85 (1ꢀ3), 45–50.
(
57) Kata, M.; Shauer, M. Acta Pharm Hung 1991, 61, 23.
(58) Moriwaki, C.; Costa, G. L.; Ferracini, C. N.; de Moraes, F. F.;
Zanin, G. M.; Pineda, E. A. G.; Matioli, G. Braz. J. Chem. Eng. 2008, 25
2), 255–267.
59) Rojas-Aguirre, Y.; Y ꢀe pez-Mulia, L.; Castillo, I.; L ꢀo pez-Vallejo,
(
(
F.; Soria-Arteche, O.; Hern ꢀa ndez-Campos, A.; Castillo, R.; Hern ꢀa ndez-
Luis, F. Bioorg. Med. Chem. 2011, 19 (2), 789–797.
(
60) Weickenmeier, M.; Wenz, G. Macromol. Rapid Commun. 1996,
7 (10), 731–736.
61) Gelb, R. I.; Schwartz, L. M. J. Inclusion Phen. 1989, 7 (5), 537–
43.
1
5
9
(
(
62) Zhang, B. L.; Breslow, R. J. Am. Chem. Soc. 1993, 115 (20),
353–9354.
(63) Rekharsky, M. V.; Inoue, Y. Chem. Rev. 1998, 98 (5), 1875–1917.
(64) Irie, T.; Uekama, K. J. Pharm. Sci. 1997, 86 (2), 147–162.
(65) Stella, V. J.; Rajewski, R. A. Pharm. Res. 1997, 14 (5), 556–567.
(66) Brewster, M. E.; Loftsson, T. Adv. Drug Delivery Rev. 2007, 59
(
7), 645–666.
67) Loftsson, T.; Brewster, M. E. J. Pharm. Pharmacol. 2010, 62
11), 1607–1621.
(
(
8
445
dx.doi.org/10.1021/ma2013964 |Macromolecules 2011, 44, 8433–8445