microparticles for cartilage regeneration

Biomaterials 35 (2014) 9984e9994

Contents lists available at ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Intra-articular delivery of kartogenin-conjugated chitosan nano/ microparticles for cartilage regeneration Mi Lan Kang, Ji-Yun Ko, Ji Eun Kim, Gun-Il Im* Department of Orthopedics, Dongguk University Ilsan Hospital, Goyang, Republic of Korea

a r t i c l e i n f o

a b s t r a c t

Article history: Received 16 July 2014 Accepted 29 August 2014 Available online 17 September 2014

We developed an intra-articular (IA) drug delivery system to treat osteoarthritis (OA) that consisted of kartogenin conjugated chitosan (CHI-KGN). Kartogenin, which promotes the selective differentiation of mesenchymal stem cells (MSCs) into chondrocytes, was conjugated with low-molecular-weight chitosan (LMWCS) and medium-molecular-weight chitosan (MMWCS) by covalent coupling of kartogenin to each chitosan using an ethyl(dimethylaminopropyl) carbodiimide (EDC)/N-hydroxysuccinimide (NHS) catalyst. Nanoparticles (NPs, 150 ± 39 nm) or microparticles (MPs, 1.8 ± 0.54 mm) were fabricated from kartogenin conjugated-LMWCS and eMMWCS, respectively, by an ionic gelation using tripolyphosphate (TPP). The in vitro release profiles of kartogenin from the particles showed sustained release for 7 weeks. When the effects of the CHI-KGN NPs or CHI-KGN MPs were evaluated on the in vitro chondrogenic differentiation of human bone marrow MSCs (hBMMSCs), the CHI-KGN NPs and CHI-KGN MPs induced higher expression of chondrogenic markers from cultured hBMMSCs than unconjugated kartogenin. In particular, hBMMSCs treated with CHI-KGN NPs exhibited more distinct chondrogenic properties in the long-term pellet cultures than those treated with CHI-KGN MPs. The in vivo therapeutic effects of CHIKGN NPs or CHI-KGN MPs were investigated using a surgically-induced OA model in rats. The CHIKGN MPs showed longer retention time in the knee joint than the CHI-KGN NPs after IA injection in OA rats. The rats treated with CHI-KGN NPs or CHI-KGN MPs by IA injection showed much less degenerative changes than untreated control or rats treated with unconjugated kartogenin. In conclusion, CHIKGN NPs or CHI-KGN MPs can be useful polymer-drug conjugates as an IA drug delivery system to treat OA. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Polymer-drug conjugate Kartogenin Chitosan nanoparticles Chitosan microparticles Intra-articular injection Osteoarthritis

1. Introduction Osteoarthritis (OA), also known as degenerative arthritis or degenerative joint disease, affects millions of people around the world. Intra-articular (IA) drug delivery can be an useful modality in OA treatment, delivering a drug directly to the main focus of the disease. IA drug administration has several advantages, such as initial high local drug concentrations, lower total drug dose, avoidance of systemic side effects, and fewer drug interactions [1]. The therapeutic effect of IA drug depends mostly on the efficacy of the drug delivery system, due to the short retention time and rapid clearance of soluble drugs from the joint. Several IA drug delivery systems, including liposomes [2,3], hydrogel [4], nanoparticles

* Corresponding author. Department of Orthopedics, Dongguk University Ilsan Hospital, 814 Siksa-Dong, Goyang 411-773, Republic of Korea. Tel.: þ82 31 961 7315; fax: þ82 31 961 7314. E-mail address: [email protected] (G.-I. Im). http://dx.doi.org/10.1016/j.biomaterials.2014.08.042 0142-9612/© 2014 Elsevier Ltd. All rights reserved.

[5,6], and microparticles [7e9], have been used to achieve prolonged and sustained release of drugs in the joint. Glucocorticoids and sodium hyaluronate/hyaluronic acid (HA) are IA injection materials that have been used broadly for OA treatment. IA injection of such anti-inflammatory analgesic agents is an effective measure for alleviating the symptoms and preventing the progression of OA [10]. However, those drugs do not induce regeneration of damaged cartilage, which is crucial for obtaining ‘good’ long-term results in OA treatment. Only surgical treatment such microfracture, osteochondral autograft/allograft, or autologous chondrocyte implantation has provided limited regeneration of articular cartilage in OA. Recently, various studies are underway to evaluate stem cells as a regenerative medicine for OA, with some research being translated into clinical studies [11]. Kartogenin is a recently characterized material that promotes the selective differentiation of mesenchymal stem cells (MSCs) into chondrocytes, thus stimulating cartilage regeneration. Kartogenin frees core-binding factor (CBF)-b from the filament A. Freed CBF-b enters the nucleus, where it binds to the DNA-binding transcription

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994

factor RUNX1. The CBFb-RUNX1 complex then activates the transcription of proteins involved in components of the cartilage matrix, such as collagen type II, aggrecan, and tissue inhibitors of metalloproteinase [12]. Kartogenin also induces type-I collagen synthesis of human dermal fibroblasts by activating the smad4/ smad5 pathway [13]. Chitosan has been investigated extensively for drug delivery systems because of its biodegradability, biocompatibility, polycationic characteristics, good solubility at a pH value close to the physiological range, and the presence of amino groups along the chitosan chain that can be used for further functionalization [14]. Kartogenin, which is a hydrophobic and low-molecular-weight compound, has a carboxyl group that can couple covalently with the amine groups of chitosan. Drug conjugation to a hydrophilic polymer not only enhances the aqueous solubility of hydrophobic drugs, but can also change drug pharmacokinetics in the body [15,16]. Thus, the conjugation of kartogenin to chitosan may possibly enhance the solubility and permeability of kartogenin, promoting its therapeutic efficacy. In the present study, the kartogenin conjugated chitosan (CHIKGN) was synthesized to enhance the aqueous solubility and the biocompatibility of hydrophobic kartogenin. Two drug delivery systems according to size range, nanoparticles (CHI-KGN NPs) and microparticles (CHI-KGN MPs), were prepared by an ionic gelation of the CHI-KGN conjugate with tripolyphosphate (TPP) anion that can interact with cationic chitosan by electrostatic forces. Thus the aim of this study was to (1) characterize the CHI-KGN particles for sustained release and chondrogenic activity in vitro, (2) evaluate the CHI-KGN particles as IA drug delivery systems for cartilage regeneration in OA joint in vivo. 2. Materials and methods 2.1. Materials 2.1.1. Polymers and reagents Low-(50e190 kDa; deacetylation degree, ~85.0%) and medium-(190e310 kDa; deacetylation degree, ~85.0%) molecular weight chitosan powders purchased from SigmaeAldrich (St. Louis, MO, USA) were used. Kartogenin was obtained from Tocris Bioscience (Bristol, UK). TPP, N-(3-dimethylaminopropyl)-N0 -ethylcarbodiimide hydrochloride (EDC) and N-hydroxysuccinimide (NHS) were purchased from SigmaeAldrich. Leucine and 2,4,6-trinitrobenzene sulfonic acid (TNBSA) were obtained from Thermo Scientific (Waltham, MA, USA). All other chemicals used were of analytical reagent grade. Dulbecco's modified Eagle's medium/F-12 (DMEM/F-12) and Alpha-Minimum Essential Medium (Alpha-MEM) purchased from Welgene (Dalseodu, Daegu, Korea) were used for pellet culture of human bone marrow mesenchymal stem cells (hBMMSCs) and chondrocyte culture, respectively. Insulin/ transferrin/selenium (ITS) and bovine serum albumin (BSA) of cell culture grade were purchased from Gibco (Grand Island, NY, USA). Dexamethasone, ascorbate-2phosphate, L-proline and sodium pyruvate were obtained from SigmaeAldrich. 2.1.2. Cells and experimental animals Bone marrow samples, used to isolate hBMMSCs, were obtained from three patients (mean age: 64 years, range: 54e72 years) undergoing total hip replacements due to OA. The isolated hBMMSCs were characterized and cultured according to our previous report [17]. Chondrocytes were isolated from the fragments of human articular cartilage that were obtained during total knee arthroplasties [7]. The donors were three patients (age range 59e65 years) who had advanced OA of the knee joints. Informed consent was obtained from all donors. The isolated hBMMSCs or chondrocytes from each patient were separately stored in liquid nitrogen and all the experiments were performed with cells from single individuals. The animal experiments conducted in this study were approved by the Animal Research and Care Committee of our institution. Nine-week-old male Sprague Dawley rats (Orient Inc., Seoul, Korea) were used according to the policies and regulations for the care and use of laboratory animals (Laboratory Animal Center, Dongguk University Ilsan Hospital, Goyang, Korea). 2.2. Preparation of kartogenin conjugated chitosan (CHI-KGN) 2.2.1. Synthesis of kartogenin conjugated chitosan (CHI-KGN) Carbodiimide chemistry was utilized to mediate the formation of an amide linkage between terminal carboxylic group of kartogenin and amine group of chitosan. Briefly, EDC/NHS solution at the appropriate concentration and molar ratio

9985

Table 1 Kartogenin formulation in synthesis of kartogenin conjugated chitosan.

Amount of kartogenin to chitosan used in conjugation Conjugation efficiency of kartogenin

LMWCS-KGN

MMWCS-KGN

95 mol% 5% weight ratio 98.1 ± 1.6%

95 mol% 5% weight ratio 97.9 ± 1.9%

were prepared in ionized water according to the manufacturer's instructions. Kartogenin was immersed in an appropriate concentration mixture of EDC and NHS for 1 h at 25  C. Low-molecular-weight chitosan (LMWCS) and medium-molecularweight chitosan (MMWCS) in acetic acid solution (1% v/v) were reacted with the NHS-esterified kartogenin for 24 h with low-speed stirring. The amount of kartogenin used in the cross link formation was 95 mol% and 5% weight ratio to LMWCS and MMWCS, respectively. The two kinds of kartogenin conjugated chitosan (CHI-KGN) were then dialyzed against deionized water with Spectra/Por dialysis tube (molecular weight cut-off ¼ 20 kD, Spectrum Lab., Rancho Dominguez, CA, USA) for 1 day. The content of unconjugated kartogenin in deionized water used in dialysis was monitored using isocratic reversed-phase high performance liquid chromatography (HPLC: Ultimate 3000, Thermo Dionex, Sunnyvale, CA, USA) spectrum. Inno C-18 column (150  4.6 mm, 5u, Youngjinbiochrom, Seoul, Korea) was used for the separation. The analysis was carried out using a flow rate of 1.0 mL/ min and recorded at 274 nm with a run time of 10 min. Kartogenin was used as a standard in the range of 1e100 mg/L. The conjugates were then lyophilized for further use. The conjugation efficiency of kartogenin was calculated from the HPLC spectrum as follows: Conjugation efficiencyð%Þ ¼

Kartogenintotal  Kartogeninunconjugated  100% Kartogenintotal

2.2.2. Characterization of CHI-KGN conjugate Both FTIR (Fourier transform infrared) and 1H NMR (proton nuclear magnetic resonance) spectroscopy were used to characterize the surface chemistry of the synthesized CHI-KGN conjugate. The lyophilized powders of CHI-KGN conjugate were applied on the FTIR sample folder and recorded on Nicolet 6700 FTIR spectrometer (Thermo Scientific) from 650 to 4000 cm1 with a resolution 8 cm1 and at 32 scan. For the 1H NMR studies, deuterated water (D2O) or dimethyl sulfoxide (dDMSO) were used as the solvents. The chemical shifts were measured in parts per million (ppm, d) using D2O or dDMSO as the internal reference. The 1H NMR spectra were obtained using an AVANCE 600 NMR spectrometer (Bruker BioSpin, Rheinstetten, Germany) at room temperature. 2.3. Preparation of CHI-KGN nanoparticles and CHI-KGN microparticles 2.3.1. Ionic gelation method CHI-KGN nanoparticles (CHI-KGN NPs) and CHI-KGN microparticles (CHI-KGN MPs) were fabricated by an ionic gelation of TPP with kartogenin conjugated LMWCS and kartogenin conjugated MMWCS, respectively. Physical conditions used in the method of the CHI-KGN conjugates with TPP were as shown in Table 2. After the preparation, CHI-KGN NPs and CHI-KGN MPs were obtained by centrifugation (15,000 rpm, 20 min), and washed three times with deionized water. The particles were then lyophilized for further use. 2.3.2. Characterization of CHI-KGN NPs and CHI-KGN MPs The morphology of the CHI-KGN NPs and CHI-KGN MPs were studied using a field-emission scanning electron microscopy (FE-SEM: ZEISS SUPRA 55VP, Carl Zeiss AG, Oberkochen, Germany). One drop of aqueous CHI-KGN NPs or CHI-KGN MPs was

Table 2 Physical conditions in production of CHI-KGN NPs and CHI-KGN MPs. CHI-KGN NPs Chitosan molecular weight CHI-KGN concentration (w/v) CHI-KGN to TPP weight ratio CHI-KGN to TPP volume ratio pH Stirring speed (rpm) Acetic acid concentration (v/v) Elapsed time after TPP addition Average size on 1 day Average size on 14 day Zeta potential (mV)

CHI-KGN MPs

LMWCS: 50e190 kDa MMWCS: 190e310 kDa 0.05% 0.85% 2: 1 4: 1 3.3: 1 1: 3 5 5 700 700 1% 1% 10 min 10 min 150 ± 39 nm 1.84 ± 0.54 mm 162 ± 43 nm 1.92 ± 0.74 mm 11.84 ± 1.2 þ7.80 ± 1.1

9986

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994

placed on a stud. After air-drying at room temperature, the samples were goldcoated using an ion sputtering device (BAL-TEC Sputter Coater SCD 005, Capovani Brothers Inc., Scotia, NY, USA), and observed using the FE-SEM. The particle size distributions of the CHI-KGN NPs and CHI-KGN MPs were measured using a dynamic light scattering spectrophotometer (DLS: DLS-7000, Otsuka Electronics Ltd., Osaka, Japan) with an argon laser beam at 488 nm and 25  C. A 90 scattering angle was used. 2.3.3. In vitro release study The lyophilized CHI-KGN NPs and CHI-KGN MPs were placed in 1.5 mL phosphate buffered saline (PBS) solution, each tube containing 100 mg of particles. The release studies were performed at 37  C and 100 rpm in a shaking incubator (SI-600, JEIO TECH, Daejeon, Korea). For the detection of release kinetics of kartogenin from CHI-KGN NPs or CHI-KGN MPs, total volume of PBS was collected after centrifugation and replaced with the same volume of PBS at each sampling time. The amounts of kartogenin released from CHI-KGN NPs or CHI-KGN MPs were evaluated using HPLC (Ultimate 3000, Dionex) spectrum in the same way as described in 2.2.1. The initial amounts of CHI-KGN conjugates in CHI-KGN NPs or CHI-KGN MPs were determined by comparative quantification of the free amino groups of chitosan. Briefly, the free amino groups of CHI-KGN NPs or CHI-KGN MPs were measured by adding TNBSA solution, which can react with amines and form a highly chromogenic derivative. Quantitative determination of the number of amines contained within the conjugates was accomplished through comparison to a standard curve generated by the use of the amino-containing leucine, dissolved in a series of known concentrations. 2.3.4. Cytotoxicity and pro-inflammatory activity of CHI-KGN NPs and CHI-KGN MPs Long-term cell proliferation was studied to evaluate the cytotoxicity of the CHIKGN NPs or CHI-KGN MPs. Briefly, passage 3 chondrocytes (1  104 cells/well in 96 well plate) were cultured in Alpha-MEM and then treated with different amounts of the particles that can maximally release kartogenin to 1, 10, 100, and 1000 nM, respectively, over 7 days according to the release curve. After incubation for 7 days with the particles, MTT assay was performed using MTT solution (1 mg/mL, Sigma). The absorbance was measured with an automated spectrophotometric microtiter plate reader (SpectraMax 340; Molecular Devices, Sunnyvale, CA, USA) using a 570 nm filter. Proeinflammatory activity of the CHI-KGN NPs or CHI-KGN MPs was determined by measuring cytokines secreted from chondrocyte treated with the particles. Briefly, passage 3 chondrocytes (1  105 cells/well in 6 well plate) were cultured in Alpha-MEM and then stimulated with 1 mg/well of each stimulant including lipopolysaccharide (LPS), kartogenin, CHI-KGN NPs, and CHI-KGN MPs. After stimulation, the concentrations of secreted IL-6 in the culture medium were assessed with an enzyme-linked immunoabsorbent assay (ELISA), according to the manufacturer's instruction (Endogen Co., Commerce Way Woburn, MA, USA).

2.4. In vitro chondrogenic differentiation 2.4.1. Induction of chondrogenic differentiation in hBMMSC pellets To investigate in vitro chondrogenic differentiation, hBMMSC pellets were formed according to a previous report [18]. The cell suspension (2.5  105 cells, passage 3e5) was aliquoted into 15-mL polypropylene centrifuge tubes, and spun in a bench top centrifuge (500 g, 10 min). The pellets were incubated in a 5% CO2 atmosphere for 3 days and then transferred to lower well of a transwell plate (SPL Life Science Co., Seoul, Korea). The pellets were cultured in DMEM/F-12 supplemented with BSA (1% w/v), ITS (1% v/v), dexamethasone (107 M), ascorbate-2-phosphate (50 mM), L-proline (50 mM), and sodium pyruvate (1 mM) for chondrogenic differentiation. The CHI-KGN NPs (17.2 mg) or CHI-KGN MPs (191.2 mg), which can release kartogenin to 100 nM in the medium for 21 days, were added to transwell insert (upper well) which has membrane with 0.1 mm pore size (Merck Millipore, Billerica, MA, USA) (Fig. 1A). The CHI-KGN NPs and CHI-KGN MPs were dispersed in the medium and presented on the membrane of upper well in transwell plate. The culture medium of lower well in transwell plate was exchanged every 3 days. For a positive control, unconjugated kartogenin was treated in the same condition as the particles at the concentration (100 nM) reported in a previous study [12]. After 21 and 28 days, the pellets were harvested for analysis (Fig. 1A). 2.4.2. DNA quantitation and GAG contents analysis Genomic DNA from each pellet was prepared from a GeneAll Tissue SV mini Kit (GeneAll, Seoul, Korea) according to the manufacturer's protocol. DNA contents were determined using a nanophotometer (Implen, Inc., Westlake Village, CA, USA). GAG production was determined with a Blyscan kit (Biocolor, Carrickfergus, UK) based on the specific binding of the cationic dye 1,9-DMMB to the sulfated GAG (s-GAG) chains of proteoglycans and protein-free s-GAG chains. Briefly, pellet digested in papain solution were mixed with Blyscan dye reagent. s-GAG-dye complex was recovered by centrifugation and the complex was resuspended in dye dissociation buffer. Absorbance was measured at 656 nm in a Spectra max plus 340 apparatus (Molecular Devices). Quantitative determination of GAG was accomplished through comparison with a standard curve generated with bovine tracheal chondroitin 4sulfate dissolved at a series of known concentrations. GAG contents were expressed as mg of GAG per mg of DNA. 2.4.3. Reverse transcription and real-time polymerase chain reaction (PCR) analysis RNA extraction was performed using the TRIzol reagent (Invitrogen Co., Carlsbad, CA, USA) according to the manufacturer's instructions and quantified using a nanophotometer (Implen, Inc.). The total RNA was reverse-transcribed with Maxime RT preMix kit oligo(dT) primer (iNtRON Biotechnology, Gyunggido, Korea) according to the manufacturer's instructions. All PCR reactions were performed on the

Fig. 1. General scheme of the in vitro induction of chondrogenesis from hBMMSCs pellets using CHI-KGN particles (A) and in vivo experimental procedure in surgically-induced OA model (B).

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994 LightCycler 480 system (Roche Diagnostics, Mannheim, Germany). Expression of the following genes was examined: collagen type II (COL2A1), aggrecan, collagen type I (COL1A1), and collagen type X (COL10A1). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control in the PCR amplification and the relative normalization ratio of PCR products derived from each target gene was calculated using the LightCycler System software (Roche, Nutley, NJ, USA). The primer pairs used in the reactions are shown in Table 3. The values obtained were normalized versus the negative control, and expressed as fold changes. All samples were assayed in triplicate. 2.4.4. Histological analysis After 21 days of chondrogenic differentiation, pellets were fixed in 4% paraformaldehyde solution and then embedded in paraffin wax. Then, 4-mm sections were cut from the paraffin wax blocks and placed on glass slides. The sections were deparaffinized with xylene and ethanol. For Safranin-O staining, aqueous Safranin-O (0.1% w/v in distilled water) was applied for 5 min followed by washing with running tap water. The section was stained in Alcian blue solution (1% w/v in 3% v/v acetic acid) for 30 min and then washed in running tap water. 2.5. In vivo cartilage regeneration 2.5.1. IA injection of CHI-KGN NPs and CHI-KGN MPs in OA rats OA was induced surgically using anterior cruciate ligament transection (ACLT), according to a previous report [7]. The rats were treated with CHI-KGN NPs and CHIKGN MPs by IA injection at weeks 6 and 9 after OA induction. Briefly, the CHI-KGN NPs (0.215 mg) or CHI-KGN MPs (2.39 mg), which can release kartogenin to 25 mM over 3 week, were suspended in 100 mL PBS and then injected into the knee joint. IA injection of vehicle (100 mL PBS) or 25 mM kartogenin in 100 mL PBS was performed in the same way to be used as controls (n ¼ 8 in each group). Rats were sacrificed for analysis 14 weeks after OA induction (Fig. 1B). 2.5.2. Retention time of CHI-KGN NPs and CHI-KGN MPs in OA joint The retention time of CHI-KGN NPs and CHI-KGN MPs was evaluated by IA injection in OA rats after fluorescence labeling. Briefly, the CHI-KGN NPs and CHI-KGN MPs were labeled with fluorescence dye (FCR-675-carboxylic acid, FlammaFluors series, Bioacts, Incheon, Korea) according to the manufacturer's instructions. After IA injection of the fluorescence dye-labeled CHI-KGN NPs and CHI-KGN MPs, each fluorescence spectrum in the OA rats (n ¼ 3) was scanned using an IVIS-spectrum measurement system (Xenogen, Hopkinton, MA, USA). Images were obtained on days 0, 2, 7, 14, and 24 after injection with three replicate images at each time point. A separate set of three replicates were imaged at all six time points. 2.5.3. Histology The distal femora in each group were dissected and fixed in 10% formaldehyde for 1 day. They were decalcified and embedded in paraffin wax or Tissue-Tek O.C.T. compound (Sakura Finetek, Inc., Torrance, CA, USA). The paraffin wax sections were stained with Safranin-O/fast green. The frozen sections were analyzed by immunohistochemistry of COL2 and aggrecan using a mouse anti-COL2A1 monoclonal antibody (Millipore; 1/200) and a rabbit anti-aggrecan polyclonal antibody (Abcam, Cambridge, UK; 1/200), respectively. The Osteoarthritis Research Society International (OARSI) Scoring System was used to grade the degenerative status of the repaired tissue [19]. The system takes into account the extent of joint involvement (stage: 0e4) as well as the depth of lesion (grade: 0e6). Multiplying grade by stage produces the OA score, with a range of 0e24 based on the most advanced grade and most extensive stage present. Necropsy of heart, liver, spleen and kidney was performed grossly and microscopically to examine any possible systemic toxicity. 2.6. Statistical analysis Descriptive statistics were used to determine group means and standard deviations. Statistical comparisons were made using two-way ANOVA with Bonferroni's post-hoc analysis when more than two groups are studied on more than one

Table 3 Primers used for real-time polymerase chain reaction. Gene symbol

Sequence (50 e30 )

Accession no.

COL2A1

F-AACCAGATTGAGAGCATCCG R-ACCTTCATGGCGTCCAAG F-TCGAGGACAGCGAGGCC R-TCGAGGGTGTAGCGTGTAGAGA F-CCCCTGGAAAGAATGGAGATG R-TCCAAACCACTGAAACCTCTG F-CAGTCATGCCTGAGGGTTTT R-GGGTCATAATGCTGTTGCCT F-CACATGGCCTCCAAGGAGTAA R-GTACATGACAAGGTGCGGCTC

NM_33150

Aggrecan COL1A1 COL10A1 GAPDH

NM_013227.3 NM_000088 NM_000493 NM_002046

9987

factor or one-way ANOVA when more than two groups are studied on one factor (SPSS 15.0; SPSS Inc., Chicago, IL, USA). P values <0.05 were considered to indicate statistical significance.

3. Results and discussion 3.1. Synthesis of CHI-KGN conjugate The chemistry in conjugating chitosan and kartogenin is shown in Fig. 2A. Amine coupling via reactive esters such as EDC/NHS is a common technique for linking ligands covalently [20]. During the cross-link formation, EDC/NHS was used to catalyze the reaction between the carboxyl group of kartogenin and the amine group of chitosan to form amide bonds. The conjugation efficiency of LMWCS and MMWCS with kartogenin was 98.1 ± 1.6% and 97.9 ± 1.9%, indicating almost complete conjugation of kartogenin in the preparation of CHI-KGN conjugates (Table 1). Successful preparation of the CHI-KGN conjugate was confirmed by FTIR and 1 H NMR. The FTIR spectrum of chitosan showed absorption bands at 1656 cm1 (C]O stretching in amide group, amide I vibration), 1598 cm1 (NH bending in nonacetylated 2-aminoglucose primary amine) and 1560 cm1 (NeH bending in amide group, amide II vibration) [21]. In the CHI-KGN conjugate, however, the amide II peak at 1560 cm1 was barely identifiable while the amide I peak at 1656 cm1 was readily apparent (Fig. 2B). The decreased amide II peak is probably due to the loss of primary amine (eNH2) groups to the secondary (eNHe) ones when chitosan is cross-linked with kartogenin, which, at the same time, strengthens the amide I peak as a result of the formation of more amide bonds [22]. Our findings indicate a successful conjugation of chitosan onto kartogenin via the formation of amide bonds during the EDC/NHS-catalyzed process. The 1H NMR spectra of chitosan, kartogenin, and CHI-KGN conjugate are given in Fig. 2C. The proton NMR chemical shift of chitosan is normally referenced to the internal standard; d ¼ 4.89 ppm for protons (H1) from unsubstituted D-glucosamine unit, d ¼ 3.27 ppm for eCHeNH2 protons (H2), d ¼ 3.92e3.72 ppm for protons (H3, H4, H5, and H6) of glucosamine ring and upfield d ¼ 2.04 ppm for acetamido (eNHCOeCH3) protons [23]. The 1H NMR spectrum of kartogenin showed prominent resonance peaks at d ¼ 7.3e7.9 ppm for protons of the benzene ring substituted with dicarboxylic acid. Compared with the peaks of chitosan and kartogenin, the 1H NMR spectrum of the CHI-KGN conjugate showed the major peaks of the benzene rings in kartogenin along with the resonance peaks at d ¼ ~1.9 and 2.7 ppm corresponding to the methyl (-CH3) and methylene proton at the C2 position of chitosan, respectively. This finding also confirms the conjugation of chitosan and kartogenin. 3.2. Characterization of CHI-KGN NPs and CHI-KGN MPs SEM and DLS were used to characterize the morphology and determine the size of the CHI-KGN NPs (Fig. 3A) and CHI-KGN MPs (Fig. 3B), respectively. The particles were spherical in shape, the CHI-KGN NPs had an average size of 150 ± 39 nm with a zeta potential of 11.84 ± 1.2 mV while the CHI-KGN MPs had an average size of 1.8 ± 0.54 mm with a zeta potential of þ7.80 ± 1.1 mV (Table 2). In an earlier study, reproducible chitosan particles were formulated by chemical cross-linking in a water/oil emulsion system [24]. While chemical cross-linking agents, such as glutaraldehyde, have negative effects on cell viability and the integrity of macromolecular drugs, the TPP ionic cross-linking polyanion has non-toxic properties and rapid gelling abilities [25]. Thus, the ionic cross-linking of cationic chitosan with polyanionic TPP has become

9988

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994

Fig. 2. Proposed reaction scheme of CHI-KGN conjugate (A). Typical FTIR (B) and 1H NMR spectra (C) of CHI-KGN conjugate showing successful cross-link formation between chitosan and kartogenin. The mark (*) on 1H NMP spectra of CHI-KGN conjugate indicates the resonance peaks at d ¼ 1.9 and 2.7 ppm corresponding to the methyl (eCH3) and methylene proton at C2 position of chitosan, respectively. Other marks (i), (ii) and (iii) on 1H NMR spectra of CHI-KGN conjugate indicate the resonance peaks derived from chitosan [(i), (ii)] and kartogenin (iii), respectively.

a popular method in preparing chitosan particles for entrapping and delivering drugs. In this study, stable and reproducible CHIKGN NPs and CHI-KGN MPs were formulated successfully by the ionic gelation method using the CHI-KGN conjugate and TPP, anionic linkages being formed between TPP phosphates and unconjugated free amino groups of the CHI-KGN conjugate immediately on mixing. The size and surface charge of chitosan particles fabricated by ionic gelation methods are affected by various physical conditions, such as chitosan molecular weight, chitosan concentration, chitosan to TPP weight ratio, solution pH value [25], acetic acid concentration, cross-linking ambient temperature [26], gelation temperature [27], chitosan to TPP volume ratio [28], stirring speed during ionic gelation, time of ionic gelation reaction, and time elapsed for TPP addition [29]. In this study, while solution circumstances such as pH value, temperature, and acid concentration were the same during the ionic gelation in the production of CHIKGN NPs and CHI-KGN MPs, the different concentrations and quantities of CHI-KGN conjugates and TPP determined the sizes of the particles (Table 2). The CHI-TPP particle system is thermodynamically unstable because of the high surface energy, especially resulting from unfavorable solution pH conditions, high particle concentration, and

chitosan molecular weight [25]. The critical aggregation of chitosan microspheres made by ionic gelation occurred after 14 days in our previous study [30]. However, in this study, there was no critical aggregation observed from DLS analysis in size distributions of CHIKGN NPs and CHI-KGN MPs after 14 days (Table 2). We assume that the CHI-KGN NPs and CHI-KGN MPs did not aggregate because they had fewer available free amino groups than unconjugated chitosan particles although a direct comparison would be needed to validate this assumption. The particle sizes also appeared to increase slightly from day 1 to day 14. This increase might be due to saturation of nanoparticles with surface charge [31]. 3.3. In vitro release study Fig. 3C and D shows the sustained and continuous release of kartogenin from CHI-KGN NPs and CHI-KGN MPs in vitro. A larger amount of kartogenin was released from CHI-KGN MPs than from CHI-KGN NPs. It is suggested that the size difference of CHI-KGN particles affected the kartogenin release rate. A previous study demonstrated that the release rate of drug increased with increasing microparticle radius [32]. The study showed that large microparticles became more porous during drug release than small microparticles, leading to faster drug transport rates [32]. We also

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994

A

9989

B

2 μm

200nm

D

CHI-KGN NPs

100

Cumulative release (%)

Cumulative release (%)

C

80 60 40

20 0 0

10

20

30

40

50

DAYS

CHI-KGN MPs

100 80 60 40 20 0 0

10

20

30

40

50

DAYS

Fig. 3. Scanning electron micrographs of CHI-KGN NPs (A) and CHI-KGN MPs (B), bars represent 200 nm and 2 mm, respectively. In vitro release of kartogenin from CHI-KGN NPs (C) and CHI-KGN MPs (D) at 37  C (n ¼ 3).

assume that CHI-KGN MPs became more porous than CHI-KGN NPs in fluid environment, facilitating the hydrolysis of conjugated kartogenin. Previous studies showed that drugs entrapped in chitosan nano/ microparticles, based on an ionic interaction, were released rapidly from the particles, with a strong burst effect within 30 min regardless of the pH of the dissolution medium [33e35]. In contrast, the CHI-KGN NPs and CHI-KGN MPs showed sustained release profiles of kartogenin in the current study. This is because covalent linkages between amine group of chitosan and carboxyl group of kartogenin were stronger than ionic bonds. It is also considered that the entanglement of kartogenin chains with the chitosan molecules protected the packed and covalent bonding between chitosan and kartogenin, resulting in sustained kartogenin release. Although the in vitro release data of kartogenin from the particles showed sustained release profiles, only 30e50% of total kartogenin were detected until 7 weeks. It is surmised that greater amount of kartogenin have been released and degraded in the PBS with warm temperature and shaking condition.

morphology by electron microscopy (data not shown). Chondrocytes treated with the CHI-KGN NPs or CHI-KGN MPs below 100 nM showed normal cell proliferation profiles, which were comparable to untreated chondrocytes. A previous study reported that pro-inflammatory activity of chitosan MPs resulted in severe joint swelling and cellular infiltration after injection into healthy rabbit joints [36]. In contrast, another study reported that chitosan microspheres showed no inflammatory infiltrate or inflammatory change in the synovium after injection into the rat knee joint [37]. To confirm the possibility of inflammation by chitosan NPs and MPs, pro-inflammatory activities of the CHI-KGN NPs or CHI-KGN MPs to chondrocytes were evaluated. While pro-inflammatory cytokines are released from both chondrocytes and synoviocytes, human articular chondrocytes were used in this study because, with synthetic function, they are more important for structural integrity of articular cartilage [38e40]. As shown in Fig. 4C, IL-6 secretion from chondrocyte treated with LPS increased significantly while CHI-KGN NPs and CHI-KGN MPs did not induce significant greater IL-6 secretion compared with unconjugated kartogenin or no treatment.

3.4. Cytotoxicity and pro-inflammatory activity test

3.5. In vitro chondrogenic differentiation

The cytotoxicity of both CHI-KGN particles was evaluated in chondrocytes. The 7-day MTT data after being exposed to the CHIKGN NPs or CHI-KGN MPs with differing amounts of MPs or NPs that could maximally release kartogenin from 1 to 1000 nM during the test period are shown in Fig. 4A-B. Cytotoxicity was observed in chondrocytes exposed to both particles at the dose that could maximally release kartogenin to 1000 nM. Indeed, chondrocytes treated with the particles at the dose showed typical necrotic cell

In vitro chondrogenic differentiation was evaluated in pellet cultures of hBMMSCs. While DNA levels did not change significantly, GAG per DNA content increased significantly, up to two-fold (p < 0.05), when exposed to CHI-KGN NPs versus no treatment or unconjugated kartogenin. Although the GAG per DNA amount in pellets treated with CHI-KGN MPs was significantly lower than those treated with CHI-KGN NPs, it was still higher than in those treated with unconjugated kartogenin (Fig. 5A). Safranin-O and

9990

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994

0.8

B

CHI-KGN NPs

1 1DAY 4DAY 7DAY

O.D at 540nm

O.D at 540nm

A

0.6 0.4

*** *** ***

0.2 0

IL-6 concentration (ng)

1DAY 4DAY 7DAY

0.8 0.6 0.4

*** *** ***

0.2 0

n/t 1nM 10nM 100nM 1000nM Concentration of releasing KGN

C

CHI-KGN MPs

1

n/t 1nM 10nM 100nM 1000nM Concentration of releasing KGN

14 12

***

***

***

10 NS

8

NS

NS

6

LPS 1µg

4

n/t

2

CHI-KGN NPs 1µg

0

Kartogenin 1µg CHI-KGN MPs 1µg

0

24

48

72

Hours Fig. 4. Seven day MTT data of chondrocyte after incubation with serum-free medium containing various doses of CHI-KGN NPs (A) or CHI-KGN MPs (B) which can maximally release kartogenin to 1, 10, 100, and 1000 nM, respectively, over 7 days. 1000 nM kartogenin treated group showed significant lower cell numbers compared with other groups [***p < 0.001, n/t: no treatment (negative control), n ¼ 3]. IL-6 secretion from chondrocytes after incubation with CHI-KGN particles (C). LPS-treated group had significant elevated level of IL-6 compared with other groups [***p < 0.001, NS: not significant, n/t: no treatment (negative control), LPS: lipopolysaccharide-treated (positive control), Kartogenin: unconjugated kartogenin-treated, n ¼ 3].

Alcian blue staining, associated with proteoglycan synthesis, showed the greatest intensity in the pellets treated with CHI-KGN NPs (Fig. 5B). RT-qPCR with mRNA isolated from differentiated hBMMSC pellets confirmed expression of genes associated with chondrogenic differentiation, including COL2A1 and aggrecan (Fig. 5C). The gene expression of COL2A1 and aggrecan increased in hBMMSC pellets exposed to unconjugated kartogenin and both CHI-KGN particles for 21 days compared with those of untreated hBMMSCs. In particular, hBMMSC pellets treated with CHI-KGN NPs showed significant increases in both genes compared with other pellets treated with either unconjugated kartogenin or CHI-KGN MPs. On the other hand, there were only slight changes in the gene expression of COL1A1 and COL10A1 after 28 days, which were not statistically significant (Fig. 4C). These results suggest that while chondrogenic differentiation of hBMMSCs pellets was enhanced by exposure to unconjugated kartogenin and CHI-KGN particles, the CHI-KGN NPs were most effective in inducing chondrogenic differentiation of hBMMSCs. 3.6. Retention time in the joint Retention times of the CHI-KGN NPs and CHI-KGN MPs in the joint were investigated by fluorescence imaging after IA injection in the OA rat (Fig. 6). The fluorescence signals from both CHI-KGN particles were observed in the knee joint up to 24 days. In

particular, the CHI-KGN MPs showed significantly higher fluorescence intensity than CHI-KGN NPs on days 2 and 7. While the CHIKGN NPs showed larger areas of fluorescence intensity on day 24, the value was not significantly greater than CHI-KGN MPs. IA retention of a drug over a prolonged period of time is important for therapeutic efficiency [1]. Drugs of low molecular weight in aqueous formulations leak into the blood circulation immediately after IA injection, necessitating repeated injections and increased dosages of drugs [41]. In this study, the CHI-KGN particles showed long retention in the OA joint over 3 weeks. This increased retention time can potentially enhance therapeutic efficacy of kartogenin. 3.7. In vivo cartilage regeneration To evaluate the regeneration of degenerated cartilage, the CHI-KGN particles were injected through the knee joint 6 and 9 weeks after surgical induction of OA. At 14 weeks after OA induction, the rats were sacrificed for histological assessments. The histological findings according to Safranin-O staining are shown in Fig. 7. The results demonstrated distinct cartilage regeneration of CHI-KGN particle-treated rats in comparison with other groups. Vehicle-treated rats (negative control treated with PBS only) showed broad areas of cartilage destruction, with matrix loss and surface denudation. Kartogenin-treated rats showed lesser loss of cartilage with matrix vertical fissures and

C

CHI-KGN MPs

rto n/t I-K ge C GN nin H I-K N G Ps N M Ps

Ka

C

* *

*

rto t I-K ge n C GN in H I-K N G Ps N M Ps

b n/t

Kartogenin

Ka

H

C

Relative gene expression (fold to n/t)

n r H tog /t I-K e C GN nin H I-K N G Ps N M Ps

0

***

150 120 90 60 30 0

H

1

CHI-KGN NPs

Aggrecan

C

NS

Ka

Relative gene expression (fold to n/t)

COL1A1 2

Kartogenin

n/

rto n/t I-K ge C GN nin H I-K N G Ps N M Ps

* ** *

Relative gene expression (fold to n/t)

***

n/t

H

H

150 120 90 60 30 0

a

B

0

C

COL2A1

9991

20

Ka

Relative gene expression (fold to n/t)

C

** *

40

rto n/t I-K ge C GN nin H I-K N G Ps N M Ps

0

*

60

COL10A1 3

NS

2

0.5mm

0.5mm

CHI-KGN NPs

CHI-KGN MPs

0.5mm

0.5mm

1 0

C rto n/t H I-K ge C GN nin H I-K N G Ps N M Ps

0.1

GAG/DNA

Ka

NS

GAG/DNA (ug)

DNA

0.2

Ka

A

DNA contents (ug)

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994

Fig. 5. DNA content, GAG normalized by DNA content on day 21 (A) and Safranin-O (B-a) and Alcian-blue (B-b) staining on day 21, RT-qPCR (C) for chondrogenic genes of COL2A1 and aggrecan on day 21, and of COL1A1 and COL10A1 on day 28 from hBMMSCs that underwent chondrogenic induction in pellets with no treatment (n/t), 100 nM unconjugated kartogenin, CHI-KGN NPs or CHI-KGN MPs which can release kartogenin to 100 nM for 21 days. Fold-change relative to untreated control is shown (*p < 0.05, **p < 0.01, ***p < 0.001, NS: not significant, n ¼ 3).

delamination of the superficial layer as well as proteoglycan stain depletion into the deep zone of the cartilage. On the other hand, CHI-KGN particle-treated OA rats showed generally intact superficial surfaces, although with minor surface abrasion in focal areas. Notably, the OARSI scores, which grade histopathology of OA and reflect depth of the lesion and extent of OA over the joint surface [19], were significantly lower in CHI-KGN particles-treated rats than those of kartogenin or vehicletreated rats. There was no significant difference between CHIKGN NPs- and CHI-KGN MPs-treated rats. Biochemical changes in the composition of articular cartilage were also investigated by immunofluorescence for COL2 and aggrecan. There were notable decreases in both proteins in the cartilage matrix of vehicle- or kartogenin-treated rats while the decrease was less marked in CHI-KGN particle-treated rats. These data indicate that CHI-KGN particles regenerated articular cartilage and arrested the progression of the OA more effectively than kartogenin. Declined therapeutic effect by the rapid

clearance of kartogenin from the joint necessitates repeated injection and increased dosage. The dosage (25 mM) and injection number of kartogenin or CHI-KGN particles used in the current study are lower than in previous report which used 100 mM kartogenin [12]. Our results showed that effective cartilage regeneration can be achieved in lower dosage by using CHI-KGN particles through a longer retention in OA joint. Necropsy of heart, liver, spleen, and kidney showed no apparent histological difference in CHI-KGN particle-treated rats versus untreated rats (data not shown). In terms of formulation design issues for IA drug delivery, the size of particles is a significant matter in therapeutic effect because it determines their transport through the joint, tissue penetration, and cellular uptake. The most appropriate size of drug formulations for IA delivery is still controversial. Whitmire et al. suggested that self-assembling nanoparticles smaller than 60 nm could be transported through the dense collagen network in the extracellular matrix (ECM) [5]. On the other hand, Eswaramoorthy et al.

9992

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994

Day 0

A

High

Day 2

6e8 L

R

L

R

5e8

Day 7

Day 14

4e8

Day 24

3e8 L

R

L

R

R

L 2e8

Low

B

Relative fluorescence intensity (fold to day0)

L : CHI-KGN NPs, R : CHI-KGN MPs 4

3

CHI-KGN NPs CHI-KGN MPs

**

*

2

1

0

DAYS Fig. 6. In vivo fluorescence imaging (A) using fluorescence dye-labeled CHI-KGN NPs (left knee) and CHI-KGN MPs (right knee) at various time points from rats in which OA was induced surgically 6 weeks before. The scale bar range is 0e6  108 in fluorescence intensity. The relative fluorescence intensities are determined as foldechanges versus those of day 0 and shown as the mean with standard deviation. CHI-KGN MPs showed significantly higher fluorescence intensity than CHI-KGN NPs on days 2 and 7 (*p < 0.05, **p < 0.01, n ¼ 3).

suggested a larger size, 51e85 mm PLGA microspheres [9], as an appropriate size for IA drug delivery. Thus, we compared two CHIKGN particles with different size ranges to evaluate the optimal size in treating OA. We had expected that CHI-KGN NPs would be more effective in vivo because they could infiltrate into MSCs more readily than CHI-KGN MPs. However, there was no significant difference between CHI-KGN NP- and CHI-KGN MPs-treated rats in OARSI scores. MSCs exist in normal synovial fluid (SF), and they are increased in early human OA [42,43]. These synovial MSCs may provide a useful target in a regenerative strategy. The CHI-KGN particles can be retained for 3 weeks in OA joint and are expected to sustainably release kartogenin to the SF. The continuously released kartogenin could induce chondrogenic differentiation of SF MSCs, which regenerate the articular cartilage and arrest the progression of OA. Our in vivo results suggest that the small molecule therapy using

the CHI-KGN particles has a potential as a convenient regenerative strategy for OA treatment. 4. Conclusions CHI-KGN NPs and CHI-KGN MPs were prepared successfully by conjugation of chitosan and kartogenin using EDC/NHS catalysis and different formulations of an ionic gelation method with TPP. They had sustained release of kartogenin for 7 weeks in vitro. CHIKGN NPs and CHI-KGN MPs induced chondrogenic differentiation of hBMMSCs more effectively than unconjugated kartogenin. In particular, hBMMSCs treated with CHI-KGN NPs exhibited more distinct chondrogenic properties in long-term pellet cultures than those treated with CHI-KGN MPs. The rats in which OA was induced surgically showed much less degenerative changes when treated with CHI-KGN NPs or MPs versus untreated control or rats treated

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994

9993

Fig. 7. In vivo effects of CHI-KGN NPs and CHI-KGN MPs on cartilage regeneration of surgically-induced OA rats. Representative image of knee joints: Safranin O and immunohistochemistry for aggrecan and COL2 to evaluate the pathological and biochemical changes at 14 weeks after the surgical induction of OA by ACL transection. Vehicle (100 mL PBS), unconjugated kartogenin (25 mM in 100 mL PBS), CHI-KGN NPs or CHI-KGN MPs which can release kartogenin to 25 mM over 3 week in 100 mL PBS were injected into the knee joint at weeks 6 and 9 after OA induction. Normal articular cartilage from rats that did not undergo surgical procedures is also shown. The graph shows the OARSI scores from medial tibial plateau. The OARSI scores were significantly lower in CHI-KGN particles-treated rats than those of unconjugated kartogenin-treated or vehicle-treated rats (*p < 0.05, **p < 0.01, NS: not significant, n ¼ 8).

with unconjugated kartogenin. CHI-KGN NPs or MPs can be useful polymer-drug conjugates for IA drug delivery system to treat OA.

Acknowledgments This work was supported by a grant from the Bio & Medical Technology Development Program of the National Research Foundation (NRF) funded by the Korean Government (2012M3A9B4028566). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References [1] Kang ML, Im GI. Drug delivery systems for intra-articular treatment of osteoarthritis. Expert Opin Drug Deliv 2014;11:269e82. [2] Dong J, Jiang D, Wang Z, Wu G, Miao L, Huang L. Intra-articular delivery of liposomal celecoxib-hyaluronate combination for the treatment of osteoarthritis in rabbit model. Int J Pharm 2013;441:285e90. [3] Elron-Gross I, Glucksam Y, Margalit R. Liposomal dexamethasone-diclofenac combinations for local osteoarthritis treatment. Int J Pharm 2009;376:84e91.

[4] Hui JH, Chan SW, Li J, Goh JC, Li L, Ren XF, et al. Intra-articular delivery of chondroitin sulfate for the treatment of joint defects in rabbit model. J Mol Histol 2007;38:483e9. [5] Whitmire RE, Wilson DS, Singh A, Levenston ME, Murthy N, Garcia AJ. Selfassembling nanoparticles for intra-articular delivery of anti-inflammatory proteins. Biomaterials 2012;33:7665e75. [6] Rothenfluh DA, Bermudez H, O'Neil CP, Hubbell JA. Biofunctional polymer nanoparticles for intra-articular targeting and retention in cartilage. Nat Mater 2008;7:248e54. [7] Ko JY, Choi YJ, Jeong GJ, Im GI. Sulforaphane-PLGA microspheres for the intraarticular treatment of osteoarthritis. Biomaterials 2013;34:5359e68. [8] Zhang Z, Huang G. Intra-articular lornoxicam loaded PLGA microspheres: enhanced therapeutic efficiency and decreased systemic toxicity in the treatment of osteoarthritis. Drug Deliv 2012;19:255e63. [9] Eswaramoorthy R, Chang CC, Wu SC, Wang GJ, Chang JK, Ho ML. Sustained release of PTH(1-34) from PLGA microspheres suppresses osteoarthritis progression in rats. Acta Biomater 2012;8:2254e62. [10] Friedman DM, Moore ME. The efficacy of intraarticular steroids in osteoarthritis: a double-blind study. J Rheumatol 1980;7:850e6. [11] Keerthi N, Chimutengwende-Gordon M, Sanghani A, Khan W. The potential of stem cell therapy for osteoarthritis and rheumatoid arthritis. Curr Stem Cell Res Ther 2013;8:444e50. [12] Johnson K, Zhu S, Tremblay MS, Payette JN, Wang J, Bouchez LC, et al. A stem cell-based approach to cartilage repair. Science 2012;336:717e21. [13] Wang J, Zhou J, Zhang N, Zhang X, Li Q. A heterocyclic molecule kartogenin induces collagen synthesis of human dermal fibroblasts by activating the smad4/smad5 pathway. Biochem Biophys Res Commun 2014;450:568e74.

9994

M.L. Kang et al. / Biomaterials 35 (2014) 9984e9994

[14] Zhang J, Xia W, Liu P, Cheng Q, Tahirou T, Gu W, et al. Chitosan modification and pharmaceutical/biomedical applications. Mar Drugs 2010;8: 1962e87. [15] Canal F, Sanchis J, Vicent MJ. Polymer-drug conjugates as nano-sized medicines. Curr Opin Biotechnol 2011;22:894e900. [16] Duncan R, Spreafico F. Polymer conjugates. Pharmacokinetic considerations for design and development. Clin Pharmacokinet 1994;27:290e306. [17] Kim KI, Park S, Im GI. Osteogenic differentiation and angiogenesis with cocultured adipose-derived stromal cells and bone marrow stromal cells. Biomaterials 2014;35:4792e804. [18] Lee JM, Im GI. PTHrP isoforms have differing effect on chondrogenic differentiation and hypertrophy of mesenchymal stem cells. Biochem Biophys Res Commun 2012;421:819e24. [19] Moskowitz RW. Osteoarthritis cartilage histopathology: grading and staging. Osteoarthr Cartil 2006;14:1e2. [20] Wissink MJ, Beernink R, Pieper JS, Poot AA, Engbers GH, Beugeling T, et al. Immobilization of heparin to EDC/NHS-crosslinked collagen. Characterization and in vitro evaluation. Biomaterials 2001;22:151e63. [21] Archana D, Singh BK, Dutta J, Dutta PK. In vivo evaluation of chitosan-PVPtitanium dioxide nanocomposite as wound dressing material. Carbohydr Polym 2013;95:530e9. [22] Heacock RA, Marion L. The infrared spectra of secondary amines and their salts. Can J Chem 1956;34:1782e95. [23] Hoven VP, Tangpasuthadol V, Angkitpaiboon Y, Vallapa N, Kiatkamjornwong S. Surface-charged chitosan: preparation and protein adsorption. Carbohydr Polym 2007;68:44e53. [24] Ohya Y, Takei T, Kobayashi H, Ouchi T. Release behavior of 5-fluorouracil from chitosan-gel nanospheres immobilizing 5-fluorouracil coated with polysaccharides and their cell specific cytotoxicity. J Macromol Sci Pure Appl Chem 1994;31:629e42. [25] Gan Q, Wang T, Cochrane C, McCarron P. Modulation of surface charge, particle size and morphological properties of chitosan-TPP nanoparticles intended for gene delivery. Colloids Surf B Biointerfaces 2005;44:65e73. [26] Fan W, Yan W, Xu Z, Ni H. Formation mechanism of monodisperse, low molecular weight chitosan nanoparticles by ionic gelation technique. Colloids Surf B Biointerfaces 2012;90:21e7. [27] Morris GA, Castile J, Smith A, Adams GG, Harding SE. The effect of prolonged storage at different temperatures on the particle size distribution of tripolyphosphate (TPP) e chitosan nanoparticles. Carbohydr Polym 2011;84: 1430e4. [28] Abdel-Hafez SM, Hathout RM, Sammour OA. Towards better modeling of chitosan nanoparticles production: screening different factors and comparing two experimental designs. Int J Biol Macromol 2014;64:334e40. [29] Fabregas A, Minarro M, Garcia-Montoya E, Perez-Lozano P, Carrillo C, Sarrate R, et al. Impact of physical parameters on particle size and reaction

[30]

[31]

[32]

[33]

[34]

[35] [36]

[37]

[38]

[39] [40] [41]

[42]

[43]

yield when using the ionic gelation method to obtain cationic polymeric chitosan-tripolyphosphate nanoparticles. Int J Pharm 2013;446:199e204. Jiang HL, Park IK, Shin NR, Kang SG, Yoo HS, Kim SI, et al. In vitro study of the immune stimulating activity of an atrophic rhinitis vaccine associated to chitosan microspheres. Eur J Pharm Biopharm 2004;58:471e6. Gazori T, Khoshayand MR, Azizi E, Yazdizade P, Nomani A, Haririan I. Evaluation of alginate/chitosan nanoparticles as antisense delivery vector: formulation, optimization and in vitro characterization. Carbohydr Polym 2009;77: 599e606. Siepmann J, Faisant N, Akiki J, Richard J, Benoit JP. Effect of the size of biodegradable microparticles on drug release: experiment and theory. J Control Release 2004;96:123e34. Mi F, Shyu S, Kuan CY, Lee ST, Lu K, Jang SF. Chitosan-polyelectrolyte complexation for the preparation of gel beads and controlled release of anticancer drug. I. Effect of phosphorus polyelectrolyte complex and enzymatic hydrolysis of polymer. J Appl Polym Sci 1999;74:1868e79. Fernandez-Urrusuno R, Calvo P, Remunan-Lopez C, Vila-Jato JL, Alonso MJ. Enhancement of nasal absorption of insulin using chitosan nanoparticles. Pharm Res 1999;16:1576e81. Ma Z, Yeoh HH, Lim LY. Formulation pH modulates the interaction of insulin with chitosan nanoparticles. J Pharm Sci 2002;91:1396e404. Liggins RT, Cruz T, Min W, Liang L, Hunter WL, Burt HM. Intra-articular treatment of arthritis with microsphere formulations of paclitaxel: biocompatibility and efficacy determinations in rabbits. Inflamm Res 2004;53: 363e72. Thakkar H, Sharma RK, Mishra AK, Chuttani K, Murthy RS. Efficacy of chitosan microspheres for controlled intra-articular delivery of celecoxib in inflamed joints. J Pharm Pharmacol 2004;56:1091e9. Kapoor M, Martel-Pelletier J, Lajeunesse D, Pelletier JP, Fahmi H. Role of proinflammatory cytokines in the pathophysiology of osteoarthritis. Nat Rev Rheumatol 2011;7:33e42. Goldring MB, Otero M. Inflammation in osteoarthritis. Curr Opin Rheumatol 2011;23:471e8. Loeser RF, Goldring SR, Scanzello CR, Goldring MB. Osteoarthritis: a disease of the joint as an organ. Arthritis Rheum 2012;64:1697e707. Ratcliffe JH, Hunneyball IM, Smith A, Wilson CG, Davis SS. Preparation and evaluation of biodegradable polymeric systems for the intra-articular delivery of drugs. J Pharm Pharmacol 1984;36:431e6. Jones EA, Crawford A, English A, Henshaw K, Mundy J, Corscadden D, et al. Synovial fluid mesenchymal stem cells in health and early osteoarthritis: detection and functional evaluation at the single-cell level. Arthritis Rheum 2008;58:1731e40. Sekiya I, Ojima M, Suzuki S, Yamaga M, Horie M, Koga H, et al. Human mesenchymal stem cells in synovial fluid increase in the knee with degenerated cartilage and osteoarthritis. J Orthop Res 2012;30:943e9.