Chondroitin sulfate-based anti-inflammatory macromolecular prodrugs

Article abstract of DOI:10.1016/j.ejps.2006.05.010

Macromolecular prodrugs of three non-steroidal anti-inflammatory drugs (NSAIDs), ibuprofen, ketoprofen, and naproxen, were prepared by the covalent attachment of the drugs onto chondroitin sulfate (ChS) using PEG 1000 as a spacer. Drug-PEG adducts were synthesized using 1,1′-carbonyl diimidazole as a coupling agent in dimethyl sulfoxide, followed by the reaction with ChS in highly dilute aqueous solution at pH 6.8 via N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDAC) as a conjugation agent. The drug-ChS conjugates were confirmed by FTIR, ¹H NMR and ¹³C NMR and the molar percent of drug substitution onto ChS was characterized by ¹H NMR using the peak areas of the three protons of -ΦCHCH₃ on the drugs to those of -NHCOCH₃ on ChS. All drug-ChS conjugates are water-soluble. The release amounts of the free drugs from their corresponding drug-ChS conjugates were evaluated in the presence or absence of either esterase or chondroitinase, and the both enzymes in pH 7.4 Tris-buffer solutions at 37 °C by high performance liquid chromatography (HPLC). Keto-ChS conjugates released ～100% ketoprofen within 12 h in the presence of esterase, but the combination with chondroitinase did not accelerate the release rate. The degradation of Keto-ChS conjugates by chondroitinase was confirmed by gel permeation chromatography (GPC). The Keto-ChS conjugates still retained the enzymatic recognition even at the substitution of ketoprofen as high as 56 mol%. The inhibition percent of carrageenan-induced edema of Keto-ChS-56 was comparable to that of a simple blend of ChS and ketoprofen, suggesting that biologically active ChS and ketoprofen could be liberated from the conjugate.

Full text of DOI:10.1016/j.ejps.2006.05.010

european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/ejps
Chondroitin sulfate-based anti-inflammatory
macromolecular prodrugs
Yu-Shiang Penga, Shih-Chun Lina, Shih-Jer Huanga, Yu-Ming Wanga,
Ying-Jer Lina, Li-Fang Wanga,∗, Jenn-Shing Chenb
a
College of Life Science, Kaohsiung Medical University, Kaohsiung, Taiwan 807, ROC
Department of Applied Chemistry, National University of Kaohsiung, Kaohsiung, Taiwan 811, ROC
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Macromolecular prodrugs of three non-steroidal anti-inflammatory drugs (NSAIDs), ibupro-
fen, ketoprofen, and naproxen, were prepared by the covalent attachment of the drugs onto
chondroitin sulfate (ChS) using PEG 1000 as a spacer. Drug–PEG adducts were synthesized
using 1,1ꢀ-carbonyl diimidazole as a coupling agent in dimethyl sulfoxide, followed by the
reaction with ChS in highly dilute aqueous solution at pH 6.8 via N-(3-dimethylaminopropyl)-
Nꢀ-ethylcarbodiimide hydrochloride (EDAC) as a conjugation agent. The drug–ChS con-
jugates were confirmed by FTIR, 1H NMR and 13C NMR and the molar percent of drug
substitution onto ChS was characterized by 1H NMR using the peak areas of the three pro-
tons of –ꢀCHCH3 on the drugs to those of –NHCOCH3 on ChS. All drug–ChS conjugates are
water-soluble. The release amounts of the free drugs from their corresponding drug–ChS
conjugates were evaluated in the presence or absence of either esterase or chondroitinase,
and the both enzymes in pH 7.4 Tris-buffer solutions at 37 ◦C by high performance liquid
chromatography (HPLC). Keto–ChS conjugates released ∼100% ketoprofen within 12 h in
the presence of esterase, but the combination with chondroitinase did not accelerate the
release rate. The degradation of Keto–ChS conjugates by chondroitinase was confirmed by
gel permeation chromatography (GPC). The Keto–ChS conjugates still retained the enzy-
matic recognition even at the substitution of ketoprofen as high as 56 mol%. The inhibition
percent of carrageenan-induced edema of Keto–ChS-56 was comparable to that of a simple
blend of ChS and ketoprofen, suggesting that biologically active ChS and ketoprofen could
be liberated from the conjugate.
Received 7 March 2006
Received in revised form
24 April 2006
Accepted 29 May 2006
Published on line 3 June 2006
Keywords:
Ibuprofen
Ketoprofen
Naproxen
Chondroitin sulfate
Macromolecular prodrug
Non-steroidal anti-inflammatory
drug
© 2006 Elsevier B.V. All rights reserved.
1.
Introduction
the large intestine (Saylers, 1979; Salyers and O’Brien, 1980).
In the human colon, ChS is present in sloughed epithelial
Chondroitin sulfate (ChS) is a copolymer of d-glucuronic acid
and sulfated N-acetyl-d-galactosamine in C4 or C6, and is an
important structural component in connective tissues and
cartilage. ChS can be degraded by anaerobic bacteria, namely
Bacteroides thetaiotaomicron and B. ovatus, which are resident in
cells and dietary meat. This characteristic suggests that ChS
is a potentially good candidate for use as a colon-targeted
drug carrier (Rubinstein et al., 1992). ChS has been reported
as a good structure/disease modifying anti-osteoarthritis drug
(S/DMOAD). The orally administrated ChS reduces the pain
∗
Corresponding author at: College of Life Science, Kaohsiung Medical University, 100 Shih-Chuan 1st Road, Kaohsiung, Taiwan 807, ROC.
Tel.: +886 7 3121101x2217; fax: +886 7 3125339.
E-mail address: lfwang@kmu.edu.tw (L.-F. Wang).
0928-0987/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.ejps.2006.05.010

european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
61
in osteoarthritis patients especially over long periods com-
pared to the commonly used diclofenac sodium (Morreale
et al., 1996; Bourgeois et al., 1998; Volpi, 2004). Although the
mechanism of action to alleviate symptoms of osteoarthri-
tis remains unclear, the acceptable mechanism is that ChS
inhibits the activities of cartilage degradation enzymes such
as collagenase, elastase, and proteoglycanase, etc., and pre-
vents cartilage damage from nitric oxide (NO) free radical (Park
et al., 2000).
model NSAIDs, ibuprofen, ketoprofen, or naproxen, onto ChS
with the use of a long chain PEG as a spacer. As mentioned
before, the anti-inflammatory activity of ChS is due to the inhi-
bition of the NO synthesis and that of NSAIDs is attributed to
the inhibition of the COX enzymes. The combination of ChS
and NSAIDs as one conjugate may be a promising design in
the treatment of degenerative disease in arthritic joints. The
temporary masking of free carboxylic groups of NSAIDs via
ester linkage in drug–ChS conjugates may prohibit the topical
irritancy in GI.
Non-steroidal anti-inflammatory drugs (NSAIDs) are
widely used to treat pain, fever, and inflammation. The
pharmacological activity of NSAIDs is related to the inhi-
bition of cyclooxygenases (COX) which are responsible for
prostaglandin synthesis, and regulate pain and inflammation
(Heyneman et al., 2000). The long-term use of NSAIDs may
cause gastrointestinal irritation, bleeding and ulceration
with side effects including nausea, vomiting, abdominal
pain, dyspepsia and diarrhea (Lombardino, 1985). Therefore,
it is preferable to temporarily mask the carboxylic group
of the NSAID and hence prohibit its direct effect on the
gastric mucosa. The “prodrug approach”, whereby a drug is
linked to a biocompatible polymer via hydrolysable bond, is
a widely used method for the controlled and targeted deliv-
ery of drugs, while masking side groups that elicit adverse
effects of drugs. This methodology also has advantage in
the solubilization of drugs, especially to help hydrophobic
drugs traverse compartmental barriers. The NSAIDs-linked
prodrugs with short chain PEGs have been well studied for
dermal applications (Bonina et al., 2001, 2002). All prodrugs
are readily hydrolyzed by human plasma and no significant
hydrolysis rate is observed with the increase in the length
of the oligoethylene chain. These ester prodrugs show the
anti-inflammatory activity similar to that of their respective
parent drug, but irritate the gastric mucosa to a significantly
lesser extent.
2.
Materials and methods
2.1.
Materials
Sodium chondroitin sulfate (ChS, oral grade, lot no. OC-
97112) was obtained from Tohoku Miyagi Pharmaceutical Co.
Ltd. (Tokyo, Japan). Chondroitinase ABC, 1,1ꢀ-carbonyl diim-
idazole (CDI), esterase, 1-hydroxybenzotriazole (HOBT), and
N-(3-dimethylaminopropyl)-Nꢀ-ethylcarbodiimide hydrochlo-
ride (EDAC) were purchased from Aldrich–Sigma Co. (St.
Louis, MO) and used as received. Poly(ethylene glycol) PEG
with the molecular weight of 1000, 2000, and 4000 was pur-
chased from Showa (Tokyo, Japan). Ibuprofen, ketoprofen, and
naproxen (ICI Chemical Co.) were recrystallized from 1:1 v/v
of acetone:hexane. Acetonitrile (CH3CN), chloroform (CHCl3),
dimethyl sulfoxide (DMSO), methanol (CH3OH), and triethy-
lamine were of HPLC grade and were obtained from Tedia
(Fairfield, OH). Trizma base and glycine, sodium hydroxide,
sodium sulfate were obtained from Fluka (Buchs, Switzerland)
and used as received.
2.2.
Synthesis of drug–PEG
Poly(ethylene glycol) esters of ibuprofen, ketoprofen, and
naproxen were modified from a procedure reported by Bonina
et al. (2002) for synthesizing oligoethylene esters of ketopro-
fen, naproxen and diclofenac. Briefly, 4 mmol of ibuprofen,
ketoprofen, or naproxen was separately dissolved in 50 ml of
chloroform and then 4.4 mmol of CDI was added portion-wise
into each solution and stirred at 0 ◦C for 4 h. To each solu-
tion, 16 mmol of PEG pre-dried overnight in a vacuum, was
added. The reaction mixtures were carried out at room tem-
perature for 1 day and then extracted twice sequentially with
50 ml of double deionized (DD) water, 50 ml of 0.1N HCl, 50 ml
of DD water, 50 ml of 0.1N NaOH, and 50 ml of DD water. The
extracted organic portions were dried over anhydrous sodium
sulfate and filtered. The organic solvent was removed by rotary
evaporator (Buchs, Switzerland) to yield white solid products.
The drug–PEG adducts were abbreviated as Ibu–PEG, Keto–PEG,
and Nap–PEG, respectively, for linking with ibuprofen, keto-
profen, and naproxen in the later section.
NSAIDs-conjugated polysaccharides specifically for colon
targeting have been studied on cyclodextrin (Kamada et al.,
2002; Yano et al., 2002), dextran (Harboe et al., 1989; Larsen
et al., 1991), and pectin (Cheng et al., 2005). Polysaccharide
prodrugs of 5-aminosalicylic acid (5-ASA) based on chitosan,
hydroxypropyl cellulose, and cyclodextrins have been pre-
pared to examine the effect of solubility of these polysaccha-
ride prodrugs on the release of 5-ASA in the gastrointestinal
contents of rats (Zou et al., 2005). Only cyclodextrins-5-ASA
releases significantly high amounts of 5-ASA because of its
good water solubility. In clinical trials, Rovetta et al. (2004)
traced a 2-year study in erosive osteoarthritis (EOA) of the
hands using 500 mg naproxen alone (group 1) or a combination
with 800 mg ChS (group 2). This study confirmed the partial
efficacy of oral ChS in improving some aspects of EOA. In a
light review of the above literature, the conjugation of NSAIDs
onto polysaccharides not only maintains their biological ther-
apeutic activities of drugs, but reduces the side effects caused
by carboxylic groups of NSAIDs as well.
In the present study, the merit of using chondroitin sulfate
as a matrix polysaccharide is because of its analgesic and high
water-soluble properties. We hypothesize that one conjugate
with two different pharmacological moieties would have syn-
ergetic anti-inflammatory activity and water solubility even at
a high degree of drug substitution. Thus, we conjugated three
2.2.1. Ibu–PEG
IR: 1104 cm−1 (C–O–C stretching); 1600 cm−1 (benzene ring
stretching); 1728 cm−1 (C
O
stretching); 2873 cm−1 (C–H
stretching); 3445 cm−1 (–OH stretching). 1H NMR (200 MHz,
CDCl3), ı (ppm): 0.86 (d, 6H, –CH(CH3)2); 1.46 (d, 3H, –ꢀCHCH3);
1.70–1.90 (m, 1H, –CH(CH3)2); 2.40 (d, 2H, ꢀCH2); 3.4–3.7 (m,

62
european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
–(OCH2CH2)n–); 3.75 (m, 1H, –ꢀCHCH3); 4.18 (m, 2H, CO2CH2);
7.0–7.3 (m, 4H, ArH). 13C NMR (200 MHz, CDCl3), ı (ppm): 174
(–OCO–); 76–70 (–OCH2CH2O–); 45 (–ꢀCHCH3); 30 (ꢀCH2); 22
(–CH(CH3)2); 18 (–ꢀCHCH3).
2.3.3. Nap–ChS
1H NMR (200 MHz, 1/1 v/v, D2O/DMSO-d6), ı (ppm): 1.37 (d,
3H, –ꢀCHCH3); 1.95 (s, 3H, NHCOCH3); 7.04–7.45 (m, 9H,
ArH). 13C NMR (200 MHz, 1/1 v/v, D2O/DMSO-d6), ı (ppm):
175.8 (–Drug–COO–); 174.6 (ChS–COO–); 174.1 (–NHCOCH3–);
69–67 (–OCH2CH2O–); 54.9 (–OCH3); 44.7 (–ꢀCHCH3); 22.3
(–NHCOCH3); 17.4 (–ꢀCHCH3). Yields: ∼47.0%.
2.2.2. Keto–PEG
IR: 1116 cm−1 (C–O–C stretching); 1600 cm−1 (benzene
ring stretching); 1731 cm−1 (C
O
stretching); 2874 cm−1
(C–H stretching); 3563 cm−1 (–OH stretching). 1H NMR
(200 MHz, CDCl3), ı (ppm): 1.52 (d, 3H, –ꢀCHCH3); 3.2–3.8
(m, –(OCH2CH2)n–); 3.93 (m, 1H, –ꢀCHCH3); 4.16 (m, 2H,
CO2CH2); 7.40–7.85 (m, 9H, ArH). 13C NMR (200 MHz, CDCl3), ı
(ppm): 196 (–ꢀC Oꢀ–); 173 (–OCO–); 76–70 (–OCH2CH2O–); 45
(–ꢀCHCH3); 18 (–ꢀCHCH3).
2.4.
Drug–ChS hydrolysis
Ten milligrams of each sample was dissolved in 1 ml of Tris-
buffer at pH 7.4 at 37 ◦C. At certain intervals, the solvent was
removed by freeze-drying in a vacuum. The free drug was
extracted from the polymer residue with 5 ml of acetone. The
supernatant acetone solution was then withdrawn and evap-
orated using a rotary evaporator to produce a white solid. The
resultant residue was dissolved in 1 ml of ethanol and ana-
lyzed by HPLC on a Hewlett-Packard 1100 system containing a
quaternary pump, an online-degassed auto sampler, a HP 1100
photodiode array detector, and a C18 column (HP Spherisob
ODS-2 column). Samples were filtered with 0.45 ␮m millipore
filters and eluted with the volume ratio of phosphoric acid
solution at pH 3.0:acetonitrile as 65%:35% at 1 ml/min. The
eluent was monitored at 254 nm for ketoprofen, 271 nm for
naproxen, and 263 nm for ibuprofen, respectively. The column
oven was set at 50 ◦C. Quantities of the hydrolysates in the
solution were determined by comparing the HPLC peak area
with the calibration intensities obtained from the pure drugs.
Each data was averaged with three measurements.
2.2.3. Nap–PEG
IR: 1103 cm−1 (C–O–C stretching); 1600 cm−1 (benzene
ring stretching); 1728 cm−1 (C
O
stretching); 2875 cm−1
(C–H stretching); 3433 cm−1 (–OH stretching). 1H NMR
(200 MHz, CDCl3), ı (ppm): 1.53 (d, 3H, –ꢀCHCH3); 3.2–3.9
(m, –(OCH2CH2)n–, –OCH3); 3.97 (m, 1H, –ꢀCHCH3); 4.21 (m,
2H, CO2CH2); 7.11–7.72 (m, 6H, ArH). 13C NMR (200 MHz,
CDCl3), ı (ppm): 174 (–OCO–); 76–70 (–OCH2CH2O–); 55 (–OCH3);
45 (–ꢀCHCH3); 18 (–ꢀCHCH3).
2.3.
Synthesis of drug–PEG and chondroitin sulfate
conjugates (drug–ChS)
For each drug–PEG, three molar ratios relative to ChS as 2:1,
1:1, and 0.5:1 were prepared. The concentration of ChS was
fixed at 1.09 mmol in 150 ml of DD water and the various molar
concentrations of drug–PEG were added. After complete dis-
solution, 1.09 mmol of EDAC and 1.09 mmol of HOBT in 1 ml
of 1/1 v/v of H2O/DMSO were added drop-wise into the reac-
tion solutions, which were adjusted to pH 6.8 by 0.1N HCl or
0.1N NaOH. The reaction was carried out at room temperature
for 2 days. Then, the reaction mixture was precipitated into
ethanol to remove any unreacted residues. The precipitated
products were dried in a vacuum oven at 60 ◦C. The drug–PEG
and ChS conjugates are abbreviated as Ibu–ChS, Keto–ChS
and Nap–ChS for ibuprofen, ketoprofen, and naproxen
linked.
2.5.
Drug–ChS hydrolysis with enzymes
The enzymatic hydrolysis of drug–ChS was conducted using a
similar procedure as stated above except for adding enzymes
in Tris-buffer at pH 7.4. The amounts of enzymes are five units
for esterase and 0.5 units for chondroitinase ABC. The enzy-
matic hydrolysis was tested using esterase or chondroitinase
ABC individually or in combination.
2.6.
Degradability of Keto–ChS by chondroitinase ABC
The degradation of Keto–ChS was conducted by dissolving a
10 mg sample in 1 ml Tris-buffer at pH 7.4 with 0.5 units of
chondroitinase ABC at 37 ◦C. At certain intervals, 10 ␮l of solu-
tion was injected into gel permeation chromatography (GPC).
The molecular weight was measured using a Shodex sugar KS-
G, a KS-804 column equipped with a Water Model 501 pump
and a HP 1047 refractive index detector. An aqueous solution
of 0.05 M NaCl was used as a mobile phase at a flow rate of
1 ml/min at 50 ◦C. The column setting was calibrated with four
monodisperse dextran standards.
2.3.1. Ibu–ChS
1H NMR (200 MHz, 1/1 v/v, D2O/DMSO-d6), ı (ppm): 0.83 (d,
6H, –CH(CH3)2); 1.32 (d, 3H, –ꢀCHCH3); 1.60–1.80 (m, 1H,
–CH(CH3)2); 1.96 (s, 3H, NHCOCH3); 2.38 (d, 2H, ꢀCH2);
7.0–7.3 (m, 4H, ArH). 13C NMR (200 MHz, 1/1 v/v, D2O/DMSO-
d6), ı (ppm): 175.7 (–Drug–COO–); 174.8 (ChS–COO–); 174.3
(–NHCOCH3–); 69–67 (–OCH2CH2O–); 44.5 (–ꢀCHCH3); 29.6
(ꢀCH2); 22.4 (–NHCOCH3); 21.9 (–CH(CH3) 2); 17.9 (–ꢀCHCH3).
Yields: ∼44.3%
2.7.
Animal experiments
2.3.2. Keto–ChS
1H NMR (200 MHz, 1/1 v/v, D2O/DMSO-d6), ı (ppm): 1.36 (d,
3H, –ꢀCHCH3); 1.96 (s, 3H, NHCOCH3); 7.14–7.55 (m, 9H,
ArH). 13C NMR (200 MHz, 1/1 v/v, D2O/DMSO-d6), ı (ppm):
197.9 (–ꢀC Oꢀ–); 175.2 (–Drug–COO–); 174.6 (ChS–COO–); 174.1
(–NHCOCH3–); 69–67 (–OCH2CH2O–); 44.5 (–ꢀCHCH3); 22.3
(–NHCOCH3); 17.9 (–ꢀCHCH3). Yields: ∼51.9%.
The anti-inflammatory activity was evaluated using
a
carrageenan-induced edema test in rat paws according to the
technique reported by Winter et al. (1962). Male Wistar rats
(250–300 g) were purchased from the National Laboratory Ani-
mal Breeding and Research Center (Taipei, Taiwan). Rats were
fasted for 24 h with free access to water and divided into III

european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
63
and a ester absorption at 1731 cm−1, which are absent in pure
ketoprofen. The ethylene oxide segments in drug–PEG adducts
appear as multiple absorption peaks around 3.2–3.8 ppm in
1H NMR and 70–76 ppm in 13C NMR besides the characteristic
absorption peaks of the drugs. The newly appeared peaks of
ester carbon were obtained at 173 or 174 ppm after the conju-
gation reaction between PEG and the drugs. These confirmed
drug–PEG adducts were further used to conjugate with ChS.
The FTIR spectra of drug–ChS were combined with the spectra
of drug–PEG and pure ChS. Fig. 1(b) illustrates the FTIR spec-
trum of Keto–ChS. The further supporting data supplied by
1H and 13C NMR have been indicated in Section 2. The peak
areas at 1.32–1.37 ppm due to the three protons of –ꢀCHCH3
on the drugs, and those at 1.95–1.96 ppm due to the three pro-
tons of –NHCOCH3 on ChS were used to calculate the drug
molar percent in drug–ChS conjugates. For example, the 1H
NMR spectrum of Keto–ChS in Fig. 2 is the 56 mol% substitution
of ketoprofen when the molar feed ratio between Keto–PEG
and ChS is 2:1. The 13C NMR spectrum of this conjugate was
illustrated in Fig. 3 to show that the four carbonyl signals were
observed at 197.9, 175.2, 174.6 and 174.1 ppm, corresponding
to –ꢀC Oꢀ–, –Drug–COO–, ChS–COO–, and –NHCOCH3, respec-
tively. The drug molar percents were organized in Table 1
using three different feed ratios between drug–PEG and ChS.
The numeral after the code indicates the drug molar per-
cent in a conjugate. The drug molar percent increases with
the increase in the feed ratio of drug–PEG/ChS and main-
tains a similar value in the conjugates with the same feed
ratio of 1/2 or 1/1 among three model drugs. When the molar
groups. Group I served as a control group without being fed.
Groups II and III were fed with ChS (7.14 mg) and ketoprofen
(1.00 mg), and with Keto–ChS-56 (8.14 mg), respectively. The
individual weight of ChS and ketoprofen used for the blend
was calculated based on 56 mol% substitution of ketoprofen
onto chondroitin sulfate. Each group had seven rats. Disc
pellets were orally administered. Ten minutes after admin-
istration, the rats were anaesthetized with an i.p. injection of
urethane (3 ml/kg). After 30 min each rat received in its right
hind paw a subplantar injection of a 0.1 wt.% carrageenan in
normal saline (␭-carrageenan, type IV, Sigma, 0.1 ml/rat). The
measurement of the hind paw volume was carried out using
an Ugo Basile Plethysmometer model 7150, before any treat-
ment (V0) and in any intervals (Vt) after the administration.
All results were expressed as means S.E.M. Statistical eval-
uations were performed using analysis of variance (ANOVA)
followed by the Newman–Keul’s test for subgroup compari-
son.
3.
Results and discussion
3.1.
Drug molar percent in drug–ChS
Scheme 1 illustrates the three-step reaction of drugs onto ChS.
Three NSAIDs, ibuprofen, ketoprofen, and naproxen linked to
PEG were confirmed by NMR and FTIR. As seen in Fig. 1(a), the
FTIR spectrum of Keto–PEG shows a broad stretching absorp-
tion around 3500 cm−1 attributed to the hydroxyl groups of PEG
Scheme 1 – The conjugation reactions of drugs and chondroitin sulfate using PEG as a spacer. (a) CDI, CHCl3, 0 ◦C, 4 h; (b)
PEG, RT, 1 d; (c) ChS, EDAC/HOBT, DDW, rt, 2 d.

64
european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
Table 1 – Molar percent of drug in drug–ChS with various
molar feed ratios of drug–PEG and ChS
Molar ratio in
feed
ChS:drug–PEG
Molar
Code name
percents
of druga
2:1
1:1
0.5:1
2:1
1:1
0.5:1
2:1
1:1
10.7
28.3
65.2
9.2
24.4
56.1
10.4
26.6
66.8
Ibu–ChS-10
Ibu–ChS-28
Ibu–ChS-65
Keto–ChS-9
Keto–ChS-24
Keto–ChS-56
Nap–ChS-10
Nap–ChS-26
Nap–ChS-66
Ibuprofen
Ketoprofen
Naproxen
0.5:1
a
Determined by NMR.
ratio of drug–PEG/ChS is 2/1, the difference of the 10 mol%
was observed between Keto–ChS-56 and Nap–ChS-66. Since
the minimum energies of trimers of these three drugs linked
with 2-hydroxyethyl methacrylate (HEMA) have been calcu-
lated from the molecular modeling (Wang et al., 2002), the
results show minimum energy increases in the sequence of
Ibu–HEMA, Nap–HEMA, and Keto–HEMA. The largest mini-
mization energy value indicates the largest steric hindrance to
free rotation along the polymer chain. Thus, the larger steric
hindrance may result in the lower degree of substitution.
3.2.
Hydrolysis of drug–ChS
Fig. 1 – (a) FTIR spectra of ketoprofen and Keto–PEG; (b)
chondroitin sulfate, Keto–PEG, and Keto–ChS.
Two hydrolysable ester groups are present in the drug–ChS
conjugates. In principle, the free drugs can be released using
two routes as illustrated in Scheme 2. The drug–ChS conju-
gates were hydrolyzed through a competing mechanism: to
Fig. 2 – 1H NMR spectrum of Keto–ChS-56.

european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
65
Fig. 3 – 13C NMR spectrum of Keto–ChS-56.
Scheme 2 – The hydrolysis routes of free drugs from drug–ChS conjugates.

66
european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
Fig. 4 – Hydrolysis profiles of drug–ChS conjugates at pH 7.4
Tris-buffer.
Fig. 6 – Hydrolysis profiles of Nap–ChS-66 in the different
enzyme conditions at pH 7.4 Tris-buffer.
release free drugs directly (k1) or drug–PEG adducts (k2), which
could be further hydrolyzed into free drugs and PEG (k3). How-
ever, due to the inability to separate the large molecules such
as drug–ChS and drug–PEG in the HPLC study, only the total
amounts of the free drugs were quantitatively measured in the
drug–ChS conjugates regardless of the hydrolysis routes. Fig. 4
displays the accumulated amounts of free drugs in Tris-buffer
at pH 7.4 using the highest drug contents in the drug–ChS
conjugates. As shown, the release amount of ibuprofen is
∼600 g/ml from Ibu–ChS-65 within 24 h but is unnoticeable
from Keto–ChS-56 and Nap–ChS-66. The hydrolysis studies
were further carried out in the presence of esterase or chon-
droitinase alone or in combination. The release amount of
ibuprofen is slightly higher in the presence of esterase but
no synergetic effect was observed using esterase and chon-
droitinase together. The same experiments were carried out
in Keto–ChS-56 and Nap–ChS-66 and the results are presented
in Figs. 5 and 6. Apparently, the amounts of the free drugs
released from Keto–ChS-56 or Nap–ChS-66 are much higher in
the presence of esterase. A combination of chondroitinase and
esterase does not increase the release amounts of ketoprofen
or naproxen. This insensitivity of the molecular weight to the
Fig. 7 – Drug release percentage based on the theoretical
drug concentrations determined by NMR. (᭹, ꢀ)
Ibu–ChS-65; (ꢀ, ꢁ) Keto–ChS-56; (ꢂ, ꢃ) Nap–ChS-66. The
filled legends represent the hydrolysis at pH 7.4 Tris-buffer
and the open ones are using esterase.
drug release from drug–ChS, is in agreement with the findings
in dextran–naproxen ester prodrugs (Harboe et al., 1989). All
exhibit 100% bioavailability to parent naproxen independent
of the molecular weights of dextran (molecular weight was
used from 10,000 to 500,000 g/mol). An alternative plot was
adopted to easily compare the catalytic efficiency of enzymes
to substrates among the three drug–ChS conjugates. The drug
release percentage (relative to the drug content determined
by NMR) was plotted in Fig. 7. The drugs released from the
drug–ChS conjugates appears massively at 2–3 h and slightly
increases thereafter. It is obvious that Keto–ChS-56 shows the
highest efficiency in the hydrolysis of ester bonds. Almost
100% of ketoprofen is released within 12 h under esterase
catalysis.
The release percentage of Ibu–ChS-65 in the presence and
absence of esterase is similar at the first 5 h, but subse-
quently increases in the presence of esterase. The catalytic
efficiency by esterase in Ibu–ChS-65 is not as promising as in
Keto–ChS-56 or Nap–ChS-66. Similar results have been found
Fig. 5 – Hydrolysis profiles of Keto–ChS-56 in the different
enzyme conditions at pH 7.4 Tris-buffer.

european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
67
in the copolymers of methacrylic acid and 2-hydroxyethyl
methacrylate linked with the three drugs (Wang et al., 2002).
The enzymatic hydrolysis of ester bonds is highly sensitive
to the hydrophilicity of the substitute itself. The solubility
values are 0.205, 0.0264 and 0.0111 g/l, respectively to ketopro-
fen, naproxen and ibuprofen in water at 37 ◦C (Kyuki, 1982).
Although the three drug–ChS conjugates are still highly water-
soluble, we believe that the characteristics of a pure drug itself
impact the hydrolysis behavior significantly.
Because ChS has been modified by the drug moieties at
the high degree of substitution (DS), it may alter the enzyme-
specific recognition. Thus, the molecular weight of Keto–ChS
was checked after enzymatic degradation. Fig. 8 shows the GPC
profiles of pure ChS and the Keto–ChS conjugates. The reten-
tion peak that appeared around 7.8 min does not shift with the
introduction of the ketoprofen in Keto–ChS-9 and Keto–ChS-
24, but a slight shoulder peak around 6.2 min appears in
Keto–ChS-56. The number average molecular weight calcu-
lated based on four mono-disperse dextran standards are 159,
145, 143 and 127 × 103 g/mol corresponding to Keto–ChS-56,
Keto–ChS-24, Keto–ChS-9, and ChS, respectively. The molec-
ular weight increases with the molar percent substitution
of ketoprofen onto the Keto–ChS conjugate. The degrada-
tion profiles of Keto–ChS-56 in the presence of chondroitinase
(Fig. 9) are similar to those of pure ChS (Tsai et al., 2005).
That the retention peak appeared at 7.8 min shifts into the
longer time implies the recognition between chondroitinase
and the Keto–ChS substrate even at the 56 mol% of ketopro-
fen substitution. To our knowledge, no polysaccharides have
been reported to be degradable at this high DS. The flexible
PEG 1000 spacer definitely plays an important role because
in the drug–HEMA systems, the conjugates become insolu-
ble when the drug molar percents are higher than 30 mol%
(Wang et al., 2002). The improvement of therapeutic efficacy
of compounds has been reported in using long chain PEGs as
the spacer (Conforti et al., 1991; Greenwald et al., 2003) but in
NSAIDs-linked prodrugs with short chain PEGs (Bonina et al.,
2001, 2002), no significant hydrolysis rate was reported with
the increase in the length of the oligoethylene chain, therefore
the effect of the PEG length on the hydrolysis rate of drug–ChS
conjugates is most interesting. Fig. 10 shows the hydrolysis
Fig. 9 – Degradation profiles of Keto–ChS-56 vs. time in the
presence of chondroitinase by GPC.
Fig. 10 – Hydrolysis profiles of Ibu–ChS conjugates with the
various PEG molecular weight at pH 7.4 Tris-buffer.
profiles for three Ibu–ChS conjugates with the PEG molecu-
lar weights of 1000, 2000, and 4000 when the feed ratio of
Ibu–PEG/ChS was fixed at 1/2. The hydrolysis rate increases
with the PEG molecular weight in the Tris-buffer solutions at
pH 7.4. Using the long chain PEG as the spacer indeed acceler-
ates the release rate in this study.
3.3.
Anti-inflammatory test
The anti-inflammatory effect is represented by the percentage
of swelling and inhibition calculated using Eqs. (1) and (2):
Vt − V0
Swelling (%) =
Inhibition (%) =
× 100
(1)
(2)
V0
(Vt − V0)control − (Vt − V0)treat
(Vt − V0)control
× 100
where V0 is the average volume in the hind paws of rats
(n = 7) before any treatment and Vt the average volume
after the inflammatory agent injection. Pretreatment with
Fig. 8 – GPC diagrams of Keto–ChS with the various drug
contents.

68
european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
ChS + Ketoprofen and Keto–ChS-56 started to reduce swelling
percentages magnificently at 5 h but the high swelling percent-
age was maintained in the control group for all time points
(Table 2). The inhibition level of Keto–ChS-56 increased up to
13 h. The simple blend of ChS and ketoprofen reduced the
swelling promptly but the Keto–ChS conjugate resulted in the
reduction of edema swelling over a prolonged period of time.
The inhibition percent of free ketoprofen reached the maxi-
mum value of 67.6% at 2 h and decayed after 6 h (Wang et al.,
2002) but that of Keto–ChS-56 increased up to 83.3% within
13 h. The synergetic anti-inflammatory activity with the use
of NSAIDs and ChS has been found to be maintained over long
periods compared with use of an NSAID only (Morreale et al.,
1996; Tsai et al., 2005).
4.
Conclusions
Three NSAIDs (ibuprofen, ketoprofen, or naproxen)-linked
ChS prodrugs were successfully synthesized with the use of
PEG as a spacer. This design not only solved the low water-
solubility of NSAIDs but also temporarily masked the free
carboxylic groups of NSAIDs with ester linkage to prohibit
the topical irritancy in GI. Moreover, one conjugate of the
analgesic ChS and NSAIDs could bring about a synergetic anti-
inflammatory effect and the delayed release of the drugs. The
hydrolysis of labile ester linkages between the drugs and ChS
were accelerated with the use of esterase. However, no syn-
ergetic catalysis was observed in the presence of esterase
and chondroitinase in the three conjugates. The large amount
of ketoprofen was released from Keto–ChS-56 probably due
to its high water solubility. The degradation of ChS sugar
main chains did not accelerate the enzymatic hydrolysis of
the ester bonds between the drugs and ChS, and the enzy-
matic recognition was still retained in Keto–ChS-56, that is the
highest degree of substitution found in literature. The anti-
inflammatory activity of Keto–ChS-56 was better than pure
ketoprofen in the carrageenan-induced edema test. In prin-
ciple, any drugs containing carboxylic functional groups can
be linked onto ChS with PEG as a spacer using the same con-
jugation chemistry in this study.
Acknowledgement
We would like to thank the National Science Council of Taiwan
for the financial support under the grant number NSC-92-
2216-E037-001.
r e f e r e n c e s
Bonina, F., Puglia, C., Santagati, N.A., Saija, A., Tomaino, A.,
Tita, B., 2002. Oligoethylene ester derivatives of ketoprofen,
naproxen and diclofenac as oral prodrugs: a
pharmacological evaluation. Pharmazie 57, 552–555.
Bonina, F.P., Puglia, C., Barbuzzi, T., Caprariis, P., Palagiano, F.,
Rimoli, M.G., Saija, A., 2001. In vitro and in vivo evaluation
of polyoxyethylene esters as dermal prodrugs of ketoprofen,
naproxen and diclofenac. Eur. J. Pharm. Sci. 14, 123–134.
Bourgeois, P., Chales, G., Dehais, J., Delcambre, B., Kuntz, J.L.,
Rozenberg, S., 1998. Efficacy and tolerability of chondroitin

european journal of pharmaceutical sciences 2 9 ( 2 0 0 6 ) 60–69
69
sulfate 1200 mg/day vs. chondroitin sulfate 3 × 400 mg/day
Rovetta, G., Monteforte, P., Molfetta, G., Balestra, V., 2004. A
two-year study of chondroitin sulfate in erosive
osteoarthritis of the hands: behavior of erosions,
osteophytes, pain and hand dysfunction. Drugs Exp. Clin.
Res. 30, 11–16.
Rubinstein, A., Nakar, D., Sintov, A., 1992. Colonic drug delivery:
enhanced release of indomethacin from cross-linked
chondroitin matrix in rat cecal content. Pharm. Res. 9,
276–280.
Salyers, A.A., O’Brien, M., 1980. Cellular location of enzymes
involved in chondroition sulfate breakdown by Bacteroides
thetaiotaomicron. J. Bateriol. 143, 772–780.
Saylers, A.A., 1979. Energy sources of major intestinal
fermentative anaerobes. Am. J. Clin. Nutr. 32, 158–163.
Tsai, M.F., Chiang, Y.L., Wang, L.F., Huang, G.W., Wu, P.C., 2005.
Oral sustained delivery of diclofenac sodium using calcium
chondroitin sulfate matrix. J. Biomater. Sci. Polym. Ed. 16,
1319–1331.
Volpi, N., 2004. The pathobiology of osteoarthritis and the
rationale for using the chondroitin sulfate for its treatment.
Curr. Drug Targets Immune Endocr. Metabol. Disord. 4,
119–127.
Wang, L.F., Chiang, H.N., Wu, P.C., 2002. Kinetics and hydrolysis
mechanism of polymeric prodrugs containing ibuprofen,
ketoprofen, and naproxen as pendent agents. J. Biomater.
Sci. Polym. Ed. 13, 287–299.
Winter, C., Risley, E., Nuss, G., 1962. Carrageenan-induced
oedema in hind paw of rats as an assay for
anti-inflammatory drug. Proc. Soc. Exp. Biol. Med. 111,
544–547.
Yano, H., Hirayama, F., Kamada, M., Arima, H., Uekama, K.,
2002. Colon-specific delivery of prednisolone-appended
alpha-cyclodextrin conjugate: alleviation of systemic side
effect after oral administration. J. Control. Release 79,
103–112.
vs. placebo. Osteoarthritis Cartilage 6 (Suppl. A), 25–30.
Cheng, H., Cao, X., Xian, M., Fang, L., Cai, T.B., Ji, J.J., Tunac, J.B.,
Sun, D., Wang, P.G., 2005. Synthesis and enzyme-specific
activation of carbohydrate-geldanamycin conjugates with
potent anticancer activity. J. Med. Chem. 48, 645–652.
Conforti, A., Caliceti, P., Sartore, L., Schiavon, O., Veronese, F.,
Velo, G.P., 1991. Anti-inflammatory activity of
monomethoxypolyethylene glycol superoxide dismutase on
adjuvant arthritis in rats. Pharmacol. Res. 23, 51–56.
Greenwald, R.B., Choe, Y.H., McGuire, J., Conover, C.D., 2003.
Effective drug delivery by PEGylated drug conjugates. Adv.
Drug Deliv. Rev. 55, 217–250.
Harboe, E., Larsen, C., Johansen, M., Olesen, H.P., 1989.
Macromolecular prodrugs. XV. Colon-targeted
delivery—bioavailability of naproxen from orally
administered dextran–naproxen ester prodrugs varying in
molecular size in the pig. Pharm. Res. 6, 919–923.
Heyneman, C.A., Lawless-Liday, C., Wall, G.C., 2000. Oral versus
topical NSAIDs in rheumatic disease: a comparison. Drugs
60, 555–574.
Kamada, M., Hirayama, F., Udo, K., Yano, H., Arima, H., Uekama,
K., 2002. Cyclodextrin conjugate-based controlled release
system: repeated- and prolonged-release of ketoprofen after
oral administration in rats. J. Control. Release 82, 407–416.
Kyuki, K., 1982. Evaluation of non-steroidal anti-inflammatory
drugs (NSAID) intended for use as topical anti-inflammatory
drugs. Nippon Yakurigaku Zasshi 79, 461–485.
Larsen, C., Jensen, B.H., Olesen, H.P., 1991. Bioavailability of
ketoprofen from orally administered ketoprofen–dextran
ester prodrugs in the pig. Acta. Pharm. Nordica 3, 71–76.
Lombardino, G., 1985. Non-steroidal Anti-inflammatory Drugs.
Wiley and Sons, New York.
Morreale, P., Manopulo, R., Galati, M., Boccanera, L., Saponati,
G., Bocchi, L., 1996. Comparison of the anti-inflammatory
efficacy of chondroitin sulfate and diclofenac sodium in
patients with knee osteoarthritis. J. Rheumatol. 23,
1385–1391.
Zou, M., Okamoto, H., Cheng, G., Hao, X., Sun, J., Cui, F., Danjo,
K., 2005. Synthesis and properties of polysaccharide
prodrugs of 5-aminosalicylic acid as potential colon-specific
delivery systems. Eur. J. Pharm. Biopharm. 59,
155–160.
Park, E.S., Chang, S.Y., Hahn, M., Chi, S.C., 2000. Enhancing
effect of polyoxyethylene alkyl ethers on the skin
permeation of ibuprofen. Int. J. Pharm. 209, 109–119.