Novel pH-sensitive chitosan-based hydrogel for encapsulating poorly

water-soluble drugs
Tse-Ying Liu
*
, Yi-Ling Lin
Institute of Biomedical Engineering, National Yang-Ming University, 155, Sec. 2, Lih-Nong St., Taipei, Taiwan, ROC
a r t i c l e i n f o
Article history:
Received 22 April 2009
Received in revised form 31 August 2009
Accepted 5 October 2009
Available online 9 October 2009
Keywords:
Chitosan derivatives
pH-sensitive hydrogel
Cell compatibility
Controlled release
a b s t r a c t
Carboxymethylhexanoyl chitosan (CHC) is an amphiphilic chitosan derivative with excellent swelling
ability and water solubility under natural conditions. In this work, the inuence of the degree of carboxy-
methyl and hexanoyl substitution on the pH-sensitive swelling behavior, drug release behavior, and
antiadhesion behavior of CHC hydrogels (cross-linked with genipin) were studied. It was found that
the pH sensitivity was more pronounced in CHC than in N,O-carboxymethyl chitosan because the hexa-
noyl group altered the state of water in CHC by inhibiting intermolecular hydrogen bonding. In addition,
greater pH sensitivity was observed in samples bearing longer hydrophobic chains (carboxymethyl
palmityl chitosan). Interestingly, when used with ibuprofen (a poorly water-soluble therapeutic agent
used here as a model drug), the bursting release of the drug was less prominent in the CHC samples
having a high degree of carboxymethyl substitution. The CHC hydrogel also demonstrated good cell com-
patibility and its antiadhesive ability after grafting was altered by changes in the degree of hexanoyl
substitution.
2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction
In recent years, considerable attention has been focused on
chitosan (CS) hydrogels and their use in tissue engineering scaf-
folds, controlled release and implants [13]. This is because of their
glycosaminoglycan-like structure and wide range of outstanding
characteristics such as biodegradability and availability. However,
their application is limited by their poor water solubility under
neutral physiological conditions, poor solubility in organic sol-
vents, and lack of amphipathicity. Moreover, it is known that an in-
crease in the hydrophobicity of a drug-loaded hydrogel, when
administered via the mucosal route, will not only improve drug
encapsulation efciency but also drug transport across the buccal
mucosa [4,5]. Therefore, several chitosan derivatives have been
developed over the years with improved properties for enhanced
applicability [68].
Recently, our group developed a novel chitosan derivative (car-
boxymethylhexanoyl chitosan, CHC) with excellent water solubil-
ity under neutral conditions [9]. In addition, we found that the
presence of both carboxymethyl (hydrophilic) and hexanoyl
(hydrophobic) groups affords an amphiphilic nature, which might
make CHC suitable for use as a drug-loaded implant material for
poorly water-soluble agents. In an earlier report, it was suggested
that a CHC monolithic drug-loaded membrane could efciently
encapsulate ibuprofen (a nonsteroidal anti-inammatory drug,
IBU) which is poorly water soluble in the neutral physiological
environment. In addition, recently, CHC micelles were also suc-
cessfully prepared and employed for encapsulation of an antitumor
agent with poor water solubility [10]. Moreover, it is expected that
the CHC hydrogel might exhibit pH-sensitive behavior since it
bears both acidic (COOH) and basic (NH
2
) functional groups. The
effects of the ionic functional groups on the nature of the pH
response of chitosan derivatives have been described by several
researchers [6,11]. However, little research has been done on the
inuence of hydrophobic substitution groups on this pH response.
In our previous study, we performed preliminary investigations on
the water absorption and water retention behavior of CHC [9].
However, it remains unclear whether the ligand substitutions
inuence the pH sensitivity of the resulting amphipathic hydrogel.
In general, the swelling (water absorption) behavior of pH-sen-
sitive hydrogels is determined by the ionization of the functional
groups of hydrogel and the intermolecular volume for water; the
latter depends on the macromolecular structure, the state of water,
the hydrophobic/hydrophilic characteristics, and the electric
charge [1214]. It is known that these factors also may govern cell
adhesion to materials as well as drug release [1517]. Hence, the
inuence of the hydrophobic substitution groups on the cell
adhesion and drug release behaviors of the CHC hydrogel were also
investigated in this study. The CHC hydrogel has a hyaluronan-like
structure (Scheme 1) with controlled hydrophilicity/hydrophobic-
ity and therefore has the potential to be employed as an
1742-7061/$ - see front matter 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.actbio.2009.10.010
* Corresponding author. Tel.: +886 2 28267923; fax: +886 2 28210847.
E-mail address: andyliudpum@yahoo.com.tw (T.-Y. Liu).
Acta Biomaterialia 6 (2010) 14231429
Contents lists available at ScienceDirect
Acta Biomaterialia
j our nal homepage: www. el sevi er . com/ l ocat e/ act abi omat
ntiadhesive membrane encapsulating therapeutically active agents
that prevent postoperative tissue adhesion.
The present study is focused on the inuence of the degree of
carboxymethyl and hexanoyl substitution on the pH-sensitive
swelling behavior. In addition, IBU was employed as a poorly
water-soluble model drug in order to investigate drug release
behavior. Furthermore, a preliminary investigation of the variation
in cytotoxicity and antiadhesive properties with the degree of
hexanoyl substitution was also carried out.
2. Experimental
2.1. Materials
Chitosan (M
w
= 215,000 g mol
1
, deacetylation degree = 80%,
insoluble impurity < 1%), hexanoic anhydride, palmitic anhydride,
and ibuprofen were purchased from SigmaAldrich. Genipin was
sourced from WAKO. Methanol was purchased from TEDIA.
2.2. Synthesis of CHC and carboxymethylpalmityl chitosan (CPC)
CHC was synthesized from N,O-carboxymethyl chitosan
(NOCC). The synthesis of NOCC and CHC with various degrees of
substitution has been reported in our previous work [9]. In brief,
the home-made NOCC with low and high degrees of carboxy-
methyl substitution were named NOCC-1 and NOCC-2, respec-
tively. The NOCC samples (2 g) were dissolved in distilled (DI)
water (50 ml) and stirred for 24 h. The resulting solutions were
mixed with methanol (50 ml), followed by the addition of hexanoic
anhydride at concentrations of 0.3 M (low degree of hexanoyl
substitution) and 0.5 M (high degree of hexanoyl substitution).
CPC, an amphiphilic NOCC derivative bearing a longer hydrophobic
chain, was produced by using palmitic anhydride instead of hexa-
noyl anhydride. After a reaction time of 12 h, the resulting solu-
tions were dialyzed against an ethanol solution (25% v/v) for
24 h. The obtained ethanol/water-soluble (volume ratio = 3:2)
chitosan derivatives with various degrees of carboxymethyl, hexa-
noyl, and palmityl substitution were named as shown in Table 1.
For the subsequent material characterization, a 1.3% (w/v) solution
of each chitosan derivative was prepared by dissolving the ob-
tained derivatives in DI water. In order to prepare hydrogels, these
solutions were then cross-linked with genipin solution (1% w/v;
molecular structure shown in Scheme 2a) at 50 C for 2 days
[18]. The molar ratio of genipin to chitosan derivative was xed
at 300 for all samples to obtain the hydrogels with identical
cross-linking density (i.e. effective number of cross-links per unit
volume) rather than identical cross-linking degree (i.e. effective
number of cross-links/total number of the D-glucosamine residues
available for cross-link), which means that the inuence of the
retractile force on the swelling behavior was almost equal for each
derivative. This is benecial to enhance and clarify the inuences
of carboxymethyl and hexanoyl groups on the pH-sensitive swell-
ing ratios and drug release behaviors of the CHC hydrogels.
2.3. Material characterization
Proton nuclear magnetic resonance (
1
H NMR) spectra were re-
corded by an NMR spectrometer (Varian UNITYINOVA 500) at
270 MHz to conrm the degree of substitution. Attenuated total
reectanceFourier transform infrared (ATRFTIR) spectra were
recorded on a spectrometer (Bomem DA8.3, Canada) using a
lm-type sample (4 cm 0.5 cm). The ATRFTIR spectra were re-
corded at a resolution of 2 cm
1
in the range 4000400 cm
1
.
The state of water was characterized by differential scanning calo-
rimetry (DSC; Perkin-Elmer Instruments) [19]. Each dried sample
was weighed in an aluminum pan to which different amounts of
DI water were then added. Prior to the DSC test, samples with var-
ious water absorption ratios (W
C
; W
C
= W
w
/W
d
, where W
w
and W
d
are the weights of the moist and dry samples, respectively) were
quenched from room temperature to 60 C and conditioned at
the same temperature for 10 min. The DSC curves were then ob-
tained by reheating to 300 K at a scanning rate of 10 K min
1
.
2.4. Characterization of the swelling ratio
The samples for the swelling test were dried in a vacuum cham-
ber with P
2
O
5
for 24 h prior to the experiment. The test was
Scheme 1. Molecular structures of CHC.
Table 1
Estimated substitution degree (by
1
H NMR) for each sample.
Carboxymethyl group Hexanoyl group D-Glucosamine residue
Degree of substitution
NOCC-1 0.32 0 0.75
CHC-1A 0.32 0.26 0.49
CHC-1B 0.32 0.46 0.29
NOCC-2 0.50 0 0.73
CHC-2A 0.50 0.26 0.47
CHC-2B 0.50 0.48 0.25
Palmityl group
CPC 0.50 0.40 0.33
Degree of N-acetyl-D-glucosamine for all samples was around 0.2.
Scheme 2. Molecular structures of (a) genipin and (b) ibuprofen.
1424 T.-Y. Liu, Y.-L. Lin/ Acta Biomaterialia 6 (2010) 14231429
performed by immersing each sample into solutions of various pH
(pH 110) at 20 C for 48 h. The swelling ratio (W
S
) at equilibrium
was determined by Eq. (1):
W
S
W
wet
W
dry
=W
dry
1
where W
wet
and W
dry
are the weights of the sample after and before
the swelling test.
2.5. Cytotoxicity and cell adhesion tests
Cytotoxicity tests were performed by the elution method
according to ISO 10993-5. Fibroblast cells (L929) were cultured
in Dulbeccos modied Eagles medium (DMEM) containing 10% fe-
tal bovine serum and were plated at a density of 2 10
5
cells ml
1
in a 24-well plate at 37 C in 5% CO
2
atmosphere. After 24 h of cul-
ture, the medium was replaced with extract uids obtained by
placing the CHC hydrogel (0.2 g ml
1
of culture medium) in the cell
culture medium at 37 C for 24 h. The cells grown in the culture
medium and in that containing 5% dimethylsulphoxide (DMSO)
under the same conditions for 24 h acted as negative and positive
controls, respectively. The adherent cells were trypsinized, centri-
fuged, and resuspended for vital cell counting using a hemacytom-
eter combined with an inverted phase-contrast microscope. For the
cell adhesion test, broblast cultures were prepared from the
meninges covering the brains of Wistar rats that were purchased
from the Animal Center, National Taiwan University (Taipei, Tai-
wan). All animals used in the present study were anesthetized
and sacriced in strict adherence to guidelines from the Guide
for the Care and Use of Laboratory Animals (National Institutes of
Health Publication No. 85-23, revised 1985) under a protocol ap-
proved by the Animal Center Committee of National Defense Med-
ical Center. Cells were seeded at 5 10
5
cells ml
1
onto six-well
plates coated with the CHC hydrogel. Cells cultured on chitosan-
coated plates under the same conditions acted as controls. After
24 h of culture, the adherent cells were observed using an inverted
phase-contrast microscope.
2.6. Drug encapsulation efciency and release test
IBUwas employedas a partiallyhydrophobic model drug(molec-
ular structure shown in Scheme 2b). IBU-loaded monolithic lms
(matrix lms) were prepared from the chitosan derivatives using
the lm-casting method. In brief, a chitosan/acetic acid solution,
NOCC/DI water solutions, and CHC/ethanol/DI water solutions were
taken and their pH values were adjusted to 7.0 using a 1 M NaOH
solution. These polymeric solutions were then mixed with IBU to
form IBU-loaded polymeric solutions (IBU concentration =
1.2 mg ml
1
). Subsequently, the drug-loaded polymeric solutions
were poured into Petri dishes and cross-linked with genipin to form
drug-loaded matrix lms after drying at 50 C for 2 days. To calcu-
late the real encapsulation efciency, the cross-linked matrix lms
were rinsed with an ethanol solution for 30 s and the amount of
IBU rinsed out into the ethanol solution (L
2
) was determined by
UVvisible spectroscopy (Agilent 8453) at 264.4 nm using a prede-
terminedstandardconcentrationintensity curve. The IBUencapsu-
lation efciency (E) was determined by Eq. (2):
E L
1
L
2
=L
1
100% 2
where L
1
is the initial loading amount of IBU incorporated. The IBU
release test was performed by placing the above-mentioned IBU-
loaded membrane (50 mg) in a quartz cuvette containing a release
medium (50 ml) with constant rotary shaking at 37 C. The release
medium (phosphate-buffered solution; pH 7.4) was withdrawn at
specic time intervals and replaced with an equivalent volume of
fresh buffer. The drug release prole was obtained by plotting a
curve of M
t
/M against t, where M
t
is the amount of drug released
at time t and M is the amount of drug released once the equilibrium
state is reached.
3. Results and discussion
3.1. Synthesis of CHC
CHC was synthesized from NOCC. Upon ligand substitution, a
hydrogen atom of the amino group was replaced by a hexanoyl
group. The molecular structure is shown in Scheme 1 and the
degree of substitution conrmed by
1
H NMR spectra is shown in
Table 1.
3.2. pH-sensitive swelling behavior
The inuence of the carboxymethyl group on the pH sensitivity
is shown in Fig. 1. Both NOCC-1 and NOCC-2 showed similar pH
sensitivity with a transition (i.e. the point with the minimum
swelling ratio) in the range pH 78; this is thought to be due to
electrostatic attraction between the COO

and NH
3

groups. Below
and above the transition point, the swelling ratio increased due to
the electrostatic repulsion between the identical NH
3

groups
NH
3

. . . NH
3

and the COO

groups (COO

. . .COO

), respectively.
The inuence of the hexanoyl group on the swelling ratio at vari-
ous pH was also investigated. For the CHC and NOCC samples with
the identical degree of carboxymethyl substitution, such as CHC-
1A and NOCC-1, since the number of D-glucosamine residues avail-
able for cross-link of the CHC samples were reduced by hexanoyl
substitution, it is reasonable to believe that the cross-linking de-
gree of CHC-1A was supposed to be higher than that of NOCC-1
while the cross-linking density was xed. Interestingly, as shown
in Fig. 1, the swelling ratios of CHC-1A and CHC-2A were higher
than those of NOCC-1 and NOCC-2, respectively. This implies that
the roles of carboxymethyl and hexanoyl substitutions on the
swelling behaviors of the chitosan derivatives employed in the
present work need to be further explored. In general, besides the
consideration of cross-linking degree and cross-linking density,
the swelling behavior is determined by intermolecular interactions
such as hydrogen bonding, hydrophobic interaction, and electro-
static interaction, which depend on the macromolecular structure
and the state of water. Therefore, the inuence of hexanoyl substi-
tution on the macromolecular structure and the state of water in
the chitosan derivatives were characterized by ATRFTIR and DSC.
Fig. 1. Inuence of carboxymethyl and hexanoyl substitution on the swelling ratio
at various pH.
T.-Y. Liu, Y.-L. Lin/ Acta Biomaterialia 6 (2010) 14231429 1425
The ATRFTIR spectra are shown in Fig. 2. The characteristic
peaks of amide I bands (1635 cm
1
) of NOCC-1 and NOCC-2 exhib-
ited a signicant red shift (d = 20 cm
1
) compared with those of
unmodied chitosan, which can be attributed to the fact that the
intermolecular hydrogen bonds in NOCC (O@CNH
2
. . .O@COH) are
stronger than those in unmodied chitosan (amide resonance
H-bonding, O@CNH
2
. . .O@CNH
2
). This is because the dipole mo-
ment of an NAH bond is smaller than that of an OAH bond. How-
ever, the extent of the above-mentioned red shift decreased as
hexanoyl groups were introduced (CHC-1A and CHC-2A). More-
over, the characteristic peaks assigned to the carboxyl dimer
(O@COH. . .O@COH, at 1723 cm
1
) observed for the CHC samples
were broader then those observed for the NOCC samples, which
implies that the carboxyl groups in CHC are in the form of a
monomer (O@COH) rather than a dimer. These ndings suggest
that the formation of intermolecular hydrogen bonding (i.e.
O@CNH
2
. . .O@COH and/or O@COH. . .O@COH) in the CHC samples
might be inhibited by hexanoyl substitution to a certain extent. If
true, it would be reasonable to believe that CHC should have a dif-
ferent state of water than NOCC. As evidenced in Fig. 3a, all the DSC
endothermic peaks were noticeably lower than 4.8 C, indicating
that free water was not present in the NOCC-2 hydrogel even in
the fully swollen state (W
C
= 2, which is almost equal to the W
S
va-
lue of NOCC-2) [19,20]. However, as shown in Fig. 3b, peak III of
CHC-2A was very close to 4.8 C, indicating that the CHC-2A hydro-
gel contains free water when not fully swollen. This, together with
the ATRFTIR observations, demonstrates that the polymerpoly-
mer interaction was weakened and the volume of intermolecular
space for free water increased on the introduction of hexanoyl
groups. Thus, the swelling ratios of the CHC samples were signi-
cantly higher than those of the NOCC samples. Furthermore, once
the formation of intermolecular H-bonding (i.e. O@CNH
2
. . .O@COH
and/or O@COH. . .O@COH) was inhibited by hexanoyl substitution,
the electrostatic repulsion NH
3

. . . NH
3

and COO

. . . COO

probably became dominant among the intermolecular interactions

when the pH was less than the pK
a
of COO

and larger then the pK

a
of NH
3

. Furthermore, the electrostatic repulsion force did not

seem to be inhibited on hexanoyl substitution, probably because
electrostatic interactions form under different conditions than
hydrogen bonding interactions; the latter require a specic
distance and angle. Therefore, a more prominent level of pH sensi-
tivity was observed in the CHC samples than in the NOCC samples,
although they had the same degree of carboxymethyl substitution.
If the above argument is correct, it is then possible that the above-
mentioned phenomenon (i.e. the pH sensitivity being enhanced by
the presence of hexanoyl groups) will become more pronounced
when a longer side chain (CPC) is introduced. As evidenced in
Fig. 4, the swelling ratios of CPC at low and high pH were consid-
erably greater than for CHC-2A, which agrees with the above
hypothesis.
3.3. In vitro cytotoxicity and cell adhesion test
CHC might serve an antiadhesion function due to its hyaluro-
nan-like structure, thus making CHC a candidate implant material
with the potential for preventing postoperative tissue adhesion
[21,22]. Therefore, the in vitro cytotoxicity and antiadhesive prop-
erties of CHC were investigated in the present study. The cytotox-
icity of the newly synthesized CHC is shown in Fig. 5. No
statistically signicant difference (P > 0.05) in the population of
surviving cells was detected between the negative controls and
the CHC samples. In addition, signicant differences (P < 0.05) in
cell survival were observed between all samples and the positive
controls. In these preliminary tests, CHC did not show any signs
of cytotoxicity in vitro and the degree of hexanoyl substitution
did not affect the number of surviving cells. The inuence of the
degree of hexanoyl substitution on cell adhesion was also studied.
In general, hydrophobicity/hydrophilicity, electric charge, surface
morphology, and the nature of surface functional groups are
Fig. 2. ATRFTIR spectra of chitosan derivatives.
Fig. 3. DSC curves of the chitosan derivatives (carboxymethyl substitution
degree = 0.5) measured at W
C
= 2 under neutral conditions. The dashed lines
represent the tted curve (Lorentzian curve-tting).
Fig. 4. Swelling ratios of CHC-2A and CPC at various pH, recorded at 37 C.
1426 T.-Y. Liu, Y.-L. Lin/ Acta Biomaterialia 6 (2010) 14231429
considered to be the factors that signicantly affect cell adhesion
[16,23]. Fig. 6 shows the cell morphology on pristine chitosan-
immobilized plates (a), CHC-2A-immobilized plates (b), and CHC-
2B-immobilized plates (c) after cell seeding for 24 h. We found that
the cells spread extensively on the chitosan-modied plates and
had a spindle morphology; this contrasted with the cells on the
CHC-2A-immobilized plates, which retained a round morphology.
This difference can be attributed to the fact that the highly swollen
surface of CHC-2A, which has a dynamic hydration layer, retarded
the formation of cell anchorage. Moreover, the electrostatic attrac-
tion between the positively charged chitosan NH
3

and the neg-

atively charged cell membrane was decreased because of the
degree of hexanoyl substitution. However, as the degree of hexa-
noyl substitution increased further (CHC-2B), the number of adher-
ent cells increased slightly, probably because the hydrophobic
interaction between the hexanoyl groups and cell membrane
favored the formation of initial cell contact [16].
3.4. Drug encapsulation efciency and release behavior
Generally, postoperative tissue adhesion and peritonitis both
complications of abdominal surgery can be treated with IBU and
chloramphenicol (a broad spectrum antibiotic), respectively
[24,25]. These two pharmaceutical agents are poorly water soluble
and are therefore dissolved in water-miscible organic solvents and
then encapsulated in amphiphilic matrices [5,26]. IBU is used as an
antiadhesion agent after abdominal surgery because it reduces the
inammatory response and subsequent formation of brous tissue
during adhesion; this results in an overall reduction in tissue adhe-
sion after surgery [24,27]. The inuence of degree of both carboxy-
methyl and hexanoyl substitution on the IBU encapsulation
efciency of the NOCC and CHC hydrogels was investigated in
our previous work and it was found that the IBU encapsulation ef-
ciency was signicantly increased when some of the amino groups
in NOCC were substituted with hexanoyl groups. It is believed that
this occurred because of the increased hydrophobicity of the poly-
mer and the decreased strength of polymerpolymer interactions
[9]. In addition, an interesting phenomenon was further explored
in the present report, as shown in Fig. 7. For the samples without
hexanoyl groups (NOCC-1 and NOCC-2), the degree of carboxy-
methyl substitution did not affect the drug encapsulation ef-
ciency to any great extent. However, for the samples bearing
hexanoyl groups, the degree of carboxymethyl substitution did
signicantly affect the drug encapsulation efciency. This can be
explained on the basis of the fact that the presence of hexanoyl
groups probably reduces the strength of polymerpolymer interac-
tions, which is favorable to the formation of hydrogen bonds
between the drug molecules (in terms of the ACOOH moiety)
and the unionized groups on the polymer chains, i.e. the COOH
and NH
2
groups. Hence, the effect of the degree of carboxymethyl
substitution on the number of drugpolymer binding sites was
pronounced for the samples with hexanoyl groups. Accordingly,
the drug encapsulation efciencies of CHC-1A and CHC-1B (low de-
gree of COOH substitution) were lower than those of CHC-2A and
CHC-2B (high degree of COOH substitution), respectively. This sug-
gests that the carboxymethyl groups play an important role and af-
fect IBU encapsulation efciency if hexanoyl groups are present. In
other words, the effects of carboxymethyl and hexanoyl groups on
the drugpolymer bonding sites are not independent.
Fig. 8 illustrates the IBU release proles from CHC lms. CHC
samples with a carboxymethyl substitution degree of 0.3
Fig. 5. Cytotoxicity of CHC samples with a carboxymethyl substitution degree of
0.5.
Fig. 6. Cell morphology on chitosan-immobilized plates (a), CHC-2A-immobilized
plates (b), and CHC-2B-immobilized plates (c) after cell seeding for 24 h.
T.-Y. Liu, Y.-L. Lin/ Acta Biomaterialia 6 (2010) 14231429 1427
(CHC-1A and CHC-1B) showed a pronounced burst behavior in the
rst 2 min, during which 83% of the IBU was released; this was fol-
lowed by a near zero-order release behavior. In contrast, CHC sam-
ples with a carboxymethyl substitution degree of 0.5 (CHC-2A and
CHC-2B) showed a release behavior that followed the Higuchi
model at the initial stages; subsequently, near zero-order release
was observed. The IBU released in the rst stage was attributed
to IBU molecules that were not tightly bound by the polymer
chains. Therefore, a faster release rate was detected at the initial
stage with a carboxymethyl substitution degree of 0.3 than with
a carboxymethyl substitution degree of 0.5; this is because the
latter has a lower swelling ratio and slower swelling rate, as ob-
served in Figs. 1 and 9, respectively. However, the subsequent
near-zero-order release behavior was probably correlated with
IBU molecules that are tightly cross-linked with the polymer
chains via intermolecular forces, including hydrophobic interac-
tion, electrostatic attraction (IBUCOO

. . .NH
3

polymer), and
hydrogen bonding (IBUCOOH. . .O@COHpolymer and IBU
COOH. . .O@CNH
2
polymer). This favors the stabilization of the
polymerdrug network and thus leads to lower drug mobility.
Therefore, IBU release in the second stage was mainly dependent
on further swelling of the polymerdrug network, the rate of which
was relatively low when compared with that in the rst stage. In
addition, the diffusion rate in the second stage was considerably
higher than that in the rst stage because the highly swollen poly-
mer network had become loose and lled with water. Hence, the
IBU release in the second stage can be considered to be a swell-
ing-controlled mechanism.
CHC demonstrated cell compatibility and antiadhesion proper-
ties, which make it a candidate material for antiadhesion implants.
In addition, the pH-sensitive swelling behavior is suitable for the
absorption of excess body uids and local release of antibiotics
(such as gentamicin and chloramphenicol) when the pH of body
uids in abdomen deviate from pH 7.4, e.g. during serious postsur-
gical complications when there is uid leakage from the stomach
(acidic) and duodenum (weakly basic). However, the correlation
between the pH-sensitive swelling behavior and the pH-response
drug release behavior is not reported here because we found this
correlation to be strongly agent-dependent. The pH-sensitive re-
lease proles seem to depend on the hydrophobicity and amino
groups of the model agents. However, the results remain unclear
and require further investigation separately.
4. Conclusions
An amphiphilic pH-sensitive hydrogel was synthesized from
chitosan through carboxymethyl and hexanoyl substitution. The
pH sensitivity was dependent on the degree and nature of such
substitutions. Moreover, hexanoyl substitution signicantly
altered the antiadhesion and drug release behaviors. The investiga-
tion of the CHC hydrogel with good cell compatibility and con-
trolled antiadhesion behavior may lead to the development of a
new type of antiadhesive membrane to prevent postoperative
tissue adhesion. The resulting membrane can also encapsulate
therapeutically active agents with a wide range of hydrophilicity.
These characteristics together should exert a synergic effect in
postsurgical use.
Acknowledgement
The authors gratefully acknowledge the National Science Coun-
cil of the Republic of China for its nancial support through con-
tract NSC 97-2221-E-010-012-MY2.
Fig. 7. Inuence of the degree of carboxymethyl and hexanoyl substitution on
ibuprofen encapsulation efciency.
Fig. 8. Ibuprofen release proles of the CHC samples measured in phosphate-
buffered solution (pH 7.4) at 37 C.
Fig. 9. Swelling kinetic curves of the CHC samples measured in phosphate-buffered
solution (pH 7.4) at 37 C.
1428 T.-Y. Liu, Y.-L. Lin/ Acta Biomaterialia 6 (2010) 14231429
Appendix: Figures with essential colour discrimination
Certain gure in this article, particularly Figure 6, is difcult to
interpret in black and white. The full colour images can be found in
the on-line version, at doi: 10.1016/j.actbio.2009.10.010.
References
[1] Hsu SH, Whu SW, Hsieh SC, Tsai CL, Chen DC, Tan TS. Evaluation of chitosan-
alginate-hyaluronate complexes modied by an RGD-containing protein as
tissue-engineering scaffolds for cartilage regeneration. Artif Organs
2004;28:693703.
[2] Yuan Q, Venkatasubramanian R, Hein S, Misra RDK. A stimulus-responsive
magnetic nanoparticle drug carrier: magnetite encapsulated by chitosan-
grafted-copolymer. Acta Biomater 2008;4:102437.
[3] Liu H, Li H, Cheng WJ, Yang Y, Zhu MY, Zhou CR. Novel injectable calcium
phosphate/chitosan composites for bone substitute materials. Acta Biomater
2006;2:55765.
[4] Martin L, Wilson CG, Koosha F, Tetley L, Gray AI, Senel S, et al. The release of
model macromolecules may be controlled by the hydrophobicity of palmitoyl
glycol chitosan hydrogels. J Control Release 2002;80:87100.
[5] Martin L, Wilson CG, Koosha F, Uchegbu IF. Sustained buccal delivery of the
hydrophobic drug denbufylline using physically cross-linked palmitoyl glycol
chitosan hydrogels. Eur J Pharm Biopharm 2003;55:3545.
[6] Chen LY, Tian ZG, Du YM. Synthesis and pH sensitivity of carboxymethyl
chitosan-based polyampholyte hydrogels for protein carrier matrices.
Biomaterials 2004;25:372532.
[7] Lin YH, Liang HF, Chung CK, Chen MC, Sung HW. Physically crosslinked
alginate/N,O-carboxymethyl chitosan hydrogels with calcium for oral delivery
of protein drugs. Biomaterials 2005;26:210513.
[8] Sashiwa H, Kawasaki N, Nakayama A, Muraki E, Yamamoto N, Aiba S. Chemical
modication of chitosan. 14: Synthesis of water-soluble chitosan derivatives
by simple acetylation. Biomacromolecules 2002;3:11268.
[9] Liu TY, Chen SY, Lin YL, Liu DM. Synthesis and characterization of amphiphatic
carboxymethylhexanoyl chitosan hydrogel: water-retention ability and drug
encapsulation. Langmuir 2006;22:97405.
[10] Liu KH, Chen SY, Liu DM, Liu TY. Self-assembled hollow nanocapsule from
amphipathic carboxymethylhexanoyl chitosan as drug carrier. Macro-
molecules 2008;41:65116.
[11] Mi FL, Liang HF, Wu YC, Lin YS, Yang TF, Sung HW, et al. O-carboxymethyl
chitosan and alginate. J Biomater Sci Polym Ed 2005;16:133345.
[12] Guan YL, Shao L, Yao KD. A study on correlation between water state and
swelling kinetics of chitosan-based hydrogels. J Appl Polym Sci 1996;61:
232535.
[13] Kurkuri MD, Aminabhavi TM. Poly(vinyl alcohol) and poly(acrylic acid)
sequential interpenetrating network pH-sensitive microspheres for the
delivery of diclofenac sodium to the intestine. J Control Release
2004;96:920.
[14] Vashuk EV, Vorobieva EV, Basalyga II, Krutko NP. Water-absorbing
properties of hydrogels based on polymeric complexes. Mater Res Innov
2001;4:3502.
[15] Freier T, Koh HS, Kazazian K, Shoichet MS. Controlling cell adhesion and
degradation of chitosan lms by N-acetylation. Biomaterials 2005;26:58728.
[16] Zhu AP, Fang N. Adhesion dynamics, morphology, and organization of 3T3
broblast on chitosan and its derivative: the effect of O-carboxymethylation.
Biomacromolecules 2005;6:260714.
[17] Le Tien C, Lacroix M, Ispas-Szabo P, Mateescu MA. N-acylated chitosan:
hydrophobic matrices for controlled drug release. J Control Release
2003;93:113.
[18] Muzzarelli RAA. Genipin-crosslinked chitosan hydrogels as biomedical and
pharmaceutical aids. Carbohydr Polym 2009;77:19.
[19] Qu X, Wirsen A, Albertsson AC. Novel pH-sensitive chitosan hydrogels:
swelling behavior and states of water. Polymer 2000;41:458998.
[20] Goycoolea FM, Heras A, Aranaz I, Galed G, Fernandez-Valle ME, Arguelles-
Monal W. Effect of chemical crosslinking on the swelling and shrinking
properties of thermal and pH-responsive chitosan hydrogels. Macromol Biosci
2003;3:6129.
[21] Burns JM, Skinner K, Colt J, Sheidlin A, Bronson R, Yaacobi Y, et al. Prevention of
tissue injury and postsurgical adhesions by precoating tissues with hyaluronic
acid solutions. J Surg Res 1995;59:64452.
[22] Zhou J, Liwski RS, Elson C, Lee TDG. Reduction in postsurgical adhesion
formation after cardiac surgery in a rabbit model using N,O-
carboxymethyl chitosan to block cell adherence. J Thorac Cardiovasc
Surg 2008;135:77783.
[23] Fang N, Zhu AP, Chan-Park MB, Chan V. Adhesion contact dynamics of
broblasts on biomacromolecular surfaces. Macromol Biosci 2005;5:102231.
[24] Lee JH, Go AK, Oh SH, Lee KE, Yuk SH. Tissue anti-adhesion potential of
ibuprofen-loaded PLLA-PEG diblock copolymer lms. Biomaterials
2005;26:6718.
[25] Nomikos IN, Katsouyanni K, Papaioannou AN. Washing with or without
chloramphenicol in the treatment of peritonitis a prospective, clinical-trial.
Surgery 1986;99:205.
[26] Ridell A, Evertsson H, Nilsson S, Sundelof LO. Amphiphilic association of
ibuprofen and two nonionic cellulose derivatives in aqueous solution. J Pharm
Sci 1999;88:117581.
[27] Oh SH, Kim JK, Song KS, Noh SM, Ghil SH, Yuk SH, et al. Prevention of
postsurgical tissue adhesion by anti-inammatory drug-loaded pluronic
mixtures with solgel transition behavior. J Biomed Mater Res A
2005;72A:30616.
T.-Y. Liu, Y.-L. Lin/ Acta Biomaterialia 6 (2010) 14231429 1429