Synthesis, characterization and cytotoxicity of photo-crosslinked maleic chitosan–polyethylene glycol diacrylate hybrid hydrogels

Acta Biomaterialia 6 (2010) 3908–3918

Contents lists available at ScienceDirect

Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat

Synthesis, characterization and cytotoxicity of photo-crosslinked maleic chitosan–polyethylene glycol diacrylate hybrid hydrogels Chao Zhong a,1, Jun Wu b, C.A. Reinhart-King b, C.C. Chu a,b,* a b

Fiber and Polymer Science Program, Department of Fiber Science and Apparel Design, Cornell University, Ithaca, NY 14853-4401, USA Biomedical Engineering Program, Cornell University, Ithaca, NY 14853-4401, USA

a r t i c l e

i n f o

Article history: Received 6 February 2010 Received in revised form 4 April 2010 Accepted 19 April 2010 Available online 21 April 2010 Keywords: Polysaccharide Maleic chitosan Polyethylene glycol Hydrogel Photo-crosslinking

a b s t r a c t Synthetic hydrogels are important biomaterials for many biomedical applications and hydrogels produced via photo-gelation have shown particular promise. In this paper, we describe a new family of biodegradable hybrid hydrogels fabricated in aqueous solution via long wavelength UV photo-crosslinking using maleic chitosan and polyethylene glycol diacrylate (PEGDA) as precursors. The maleic chitosan precursor was prepared by a simple one-step chemical modification of chitosan, with high yields, and characterized by Fourier transform infrared spectroscopy, 1H NMR and 13C NMR. Maleic chitosan and PEGDA precursors at a wide range of weight feed ratios were mixed in aqueous solution and directly photocrosslinked for 10 min under a long wavelength UV light (365 nm) using 4-(2-hydroxyethoxy) phenyl(2-hydroxy-2-propyl) ketone (Irgacure 2959) as photoinitiator. It was observed that as the weight feed ratio of maleic chitosan to PEGDA decreased the pore sizes of the hydrogel samples decreased, thereby increasing the densities of the hydrogel networks and producing a lower swelling ratio and a higher compressive modulus. The molecular weight of PEGDA had a similar effect. Preliminary cell cytotoxicity tests of both the maleic chitosan precursor and maleic chitosan/PEGDA hydrogels, based on the MTT assay and live–dead assay, respectively, showed that these new chitosan-based biodegradable biomaterials were relatively non-toxic to bovine aortic endothelial cells at low dosages. Ó 2010 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction Hydrogels are of keen interest in the field of tissue engineering and controlled drug delivery because of their high water content, permeability to oxygen and nutrients and tissue-like elastic properties [1–3]. Hydrogels can also be integrated with microdevices for various applications, including biosensors [4] and diagnostic imaging [5], using photolithographic, molding or other approaches [3]. Although natural physical hydrogels have been widely utilized for biomedical applications, chemically crosslinked hydrogels provide better control over their final physical, structural and mechanical properties [2,6]. Several chemical approaches, including photo and thermally initiated crosslinking, have been used to prepare synthetic covalent hydrogels. Among these approaches, photogelation has several advantages: it allows better spatial and temporal control over the reaction and more rapid entrapment of

* Corresponding author at: Fiber and Polymer Science Program, Department of Fiber Science and Apparel Design, Cornell University, Ithaca, NY 14853-4401, USA. Tel.: +1 607 255 1938; fax: +1 607 255 1093. E-mail address: [email protected] (C.C. Chu). 1 Present address: Department of Materials Science and Engineering, University of Washington, Seattle, WA 98105, USA.

cells with minimal cell death due to the fast curing rates, ranging from less than a second to a few minutes. In addition, some gelation reactions can be performed under very mild conditions, such as in aqueous medium, at room temperature, at body pH and even in situ in a minimally invasive manner [7]. Accordingly, photogelation is the preferred approach for the preparation of covalent hydrogels, especially for biomedical applications [8]. For many tissue engineering applications it is often desirable to encapsulate cells directly inside hydrogels. This situation requires that the polymer precursors are soluble in neutral pH aqueous medium. At first glance chitosan would seem to be a promising candidate for the preparation of photo-crosslinked hydrogels. Chitosan is a partially deacetylated product of chitin, the second most abundant polysaccharide in nature. It is also a linear polycationic polysaccharide comprising glucosamine and N-acetylglucosamine residues. Because of its biocompatibility, biodegradability, non-toxicity and antimicrobial properties [9–11] chitosan has been widely utilized for biomedical applications such as controlled drug and protein delivery [12,13], non-viral gene delivery [14] and tissue engineering [15–17]. Unfortunately, chitosan is insoluble in neutral pH aqueous medium, and so several types of watersoluble, photo-curable chitosan derivatives have been prepared to overcome this obstacle [18–21]. These include azidobenzoic

1742-7061/$ - see front matter Ó 2010 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.actbio.2010.04.011

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

acid-modified lactose chitosan (Az-CH-LA) [18,19], methacrylated glycol chitosan [20] and methacrylated O-carboxymethyl chitosan [21]. The synthesis of these derivatives requires the preparation of a water-soluble chitosan derivative, followed by the incorporation of photo-crosslinkable moieties into the derivative. This two-step process produces both relatively low yields of final products and randomly distributed photo-crosslinkable functional groups along the backbone, thereby making these derivatives less attractive than other photo-crosslinkable polysaccharides derived from dextran and hyaluronic acid. In this paper, we report a photo-crosslinkable, water-soluble chitosan precursor, maleic chitosan, based on a simple one-step modification of chitosan under mild and homogeneous reaction conditions using toluene sulfonic acid/formamide as mixed solvents. Our photo-curable precursor (maleic chitosan) prepared by this approach has higher yields and well-defined chemical structure in comparison with other water-soluble chitosan precursors usually synthesized by a two-step process as described above. The maleic chitosan precursor synthesized is of anionic nature due to the presence of carboxyl groups and hence may offer additional functionalities that some water-soluble chitosan precursors, like methacrylamide chitosan [22], do not have. For example, the carboxyl functionality in maleic chitosan can either be used to couple biologically active agents, like nitric oxide derivative [23], or serve as functional groups on hydrogel surfaces to promote the deposition of inorganic minerals for bone engineering applications [24]. For hybrid hydrogel fabrication we integrated maleic chitosan with the polyethylene glycol (PEG) derivative polyethylene glycol diacrylate (PEGDA) in aqueous medium via photo-crosslinking to form hybrid biodegradable hydrogels that are pH-sensitive. The specific use of PEGDA as the crosslinker was based on reported studies that PEG hydrogels are non-toxic, non-immunogenic and approved by the US Food and Drug Administration for various clinical uses [3]. PEG-based hydrogels are also one of the most widely used materials for biomedical applications [25]. By varying the weight feed ratio of maleic chitosan to PEGDA precursors and the molecular weight of PEGDA, we were able to tune the swelling ratios, mechanical properties and pore sizes of the resulting hybrid hydrogels. To preliminarily evaluate the cytotoxicity of both maleic chitosan and maleic chitosan/PEGDA hydrogels, we used the MTT assay and live–dead assay, respectively. Our results indicated that maleic chitosan and maleic chitosan/PEGDA hydrogels elicit overall good viability of endothelial cells.

2. Materials and methods 2.1. Materials The starting material, chitosan (Sigma product), was 75% deacetylated, with a Brookfield viscosity of 20.0–200 cps (1% acetic acid) and a molecular weight of approximately 50,000 Da based on viscosity data. p-Tolune sulfonic acid monohydrate (Alfa Aesar, Ward Hill, MA), sodium bicarbonate (Sigma), maleic anhydride (Fluka), PEG [weight-average molecular weights (Mw) 2000, 4000 and 8000 g mol1] (Aldrich Chemical Co., Milwaukee, WI) and acryloyl chloride (Aldrich Chemical Co.) were all used without further purification. PEGDA (Mw 2000, 4000 and 8000) was synthesized in our own laboratory, while PEGDA (Mw 700) from Sigma was used as received. Triethylamine from Fisher Scientific (Fairlawn, NJ) was dried via refluxing with calcium hydride and then distilled before use. Formamide from EMD Chemicals (stored in a refrigerator) was unfrozen and used immediately. Other solvents, including benzene, hexane and acetone, were purchased from VWR Scientific (West Chester, PA) and used as received. 2-Hydro-

3909

xy-l-[4-(hydroxyethoxy) phenyl]-2-methyl-l-propanone (Irgacure 2959) was donated by Ciba Specialty Chemicals Corp. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) purchased from Sigma (St. Louis, MO) was used to evaluate the cell toxicity of maleic chitosan and the maleic chitosan/PEGDA hydrogels. Other chemicals and reagents were purchased from Sigma (St. Louis, MO) if not otherwise specified. 2.2. Synthesis of N,O-maleic chitosan and PEGDA macromer precursors Synthesis of maleic chitosan was based on a one-step chemical reaction between maleic anhydride and chitosan in which the chitosan was dissolved in toluene sulfonic acid/formamide mixed solvent. PEGDA was synthesized according to a procedure modified from a previously reported method [26,27]. The specific procedures for synthesizing maleic chitosan and PEGDA can be found in the Supporting information. The maleic chitosan used for the preparation of maleic chitosan/PEG hydrogels in this study had a yield 89%, a 1.17 degree of substitution (based on 1H NMR) and a molecular weight (Mr) = 52,000. 2.3. Preparation of maleic chitosan/PEGDA hybrid hydrogels Maleic chitosan solution at a concentration of 3.5% (w/v) was prepared by dissolving 0.350 g maleic chitosan in 10 ml of distilled water in a glass bottle. PEGDA (Mn 700–8000) with a pre-determined weight feed ratio to maleic chitosan (see Table S1) was then added to the prepared maleic chitosan solution. The photoinitiator 4-(2hydroxyethoxy) phenyl-(2-hydroxy-2-propyl) ketone (Irgacure 2959) was added to the precursor solution at a concentration of 0.1% (m/v). The mixed solution was then stirred for 10 min at 50 °C to ensure complete dissolution of the photoinitiator. The homogeneous, transparent solution (5 ml in total) was first transferred to a custom-made 20-well Teflon mold (250 lm volume per well) using a micropipette and then irradiated using a long wavelength UV lamp (365 nm and 8 W) at room temperature for 10 min. The photo-gelation procedure is shown in Scheme 1. After photo-gelation the hydrogel samples were immersed in distilled water at room temperature for at least 3 days to leach out the unreacted chemicals. During this period the distilled water was replaced every 12 h. 2.4. Characterization of maleic chitosan and PEGDA macromer precursors Both 1H and 13C NMR spectra of maleic chitosan were recorded on a Varian (Palo Alto, CA) Unity INOVA 500-MHz spectrometer operating at 500 and 125.7 MHz, respectively. Samples of maleic chitosan (20 mg) were dissolved in 600 ll of D2O. Chemical shifts (d) are reported in parts per million (ppm) using sodium 3-(trimethylsilyl) propionate-d4 as an internal standard. To obtain good signals, 13C NMR measurements were carried out at 343 K for 12 h. For PEGDA samples only 1H NMR spectra were recorded to confirm that the purified product was free of any contaminants. The molecular weight of maleic chitosan was determined from the intrinsic viscosity data using the Mark–Houwink equation. Specifically, we adopted the empirical equation [g] = 3.04  105  Mr1.26 to determine the molecular weight of maleic chitosan [28]. 2.5. Characterization and properties testing of maleic chitosan/PEGDA hybrid hydrogels 2.5.1. Fourier transform infrared (FTIR) spectroscopy For FTIR spectroscopic characterization dried maleic chitosan/ PEGDA hydrogels were ground to a powder, mixed with KBr (5 wt.% sample) and compressed into KBr pellets. FTIR spectra were

3910

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

Scheme 1. One of the possible photo-crosslinked structures between N, O-maleic chitosan and PEGDA

then obtained with a Perkin-Elmer (Madison, WI) Nicolet Magana 560 FTIR spectrometer with Omnic software for data acquisition and analysis. For comparison, both precursors (maleic chitosan and PEGDA) were also grounded to powers and compressed into pellets for FTIR analysis. 2.5.2. Morphology of maleic–PEGDA hybrid hydrogels The interior morphology of maleic chitosan/PEGDA hydrogels was probed by scanning electron microscopy (SEM). The swollen hydrogel samples, having reached their maximum swelling ratio in distilled water at room temperature after 24 h, were quickly frozen in liquid nitrogen and then freeze dried in a Virtis Freeze Drier (Gardiner, NY) under vacuum at 42 °C for 3 days until all water was sublimed. The freeze-dried hydrogel specimens were cut and fixed on aluminum stubs and then coated with gold for 30 s for interior morphology observation with a scanning electron microscope (Keck FE-SEM, LEO 1550). 2.5.3. Swelling ratio and swelling kinetics measurements in PBS and deionized water The swelling kinetics of the maleic chitosan/PEGD hydrogels were measured over a period of 5 days at room temperature. Dry

maleic chitosan/PEGDA gel samples were weighed and immersed in 20 ml of either phosphate-buffered saline (PBS) (pH 7.4) or deionized (DI) water at room temperature for pre-determined periods. The samples were removed from the immersion medium and blotted with filter papers to remove surface water. The samples were then weighed. The swelling ratio (Q) was calculated as follows [29,30]:

Q ¼ ½ðW s  W d Þ=W d   100% where Ws is the weight of swollen hydrogel at time t and Wd is the weight of the dry hydrogel at t = 0. All swelling ratio results were obtained from triplicate samples and data are expressed as means ± standard deviations. 2.5.4. Compressive modulus determined by dynamic mechanical analysis (DMA) The mechanical property of the maleic chitosan/PEDGA hybrid hydrogels was measured with a DMA 2980 Dynamic Mechanical Analyzer (TA Instruments Inc., New Castle, DE) in ‘‘controlled force” (CF) mode. The disc shaped swollen hydrogel samples were submerged in distilled water and mounted between a movable compression clamp (diameter 30 mm) and a fluid cup with a

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

0.1 N preloading force. A force ramp from 0.1 to 1.5, 2.0 or 2.5 N (depending on the gel strength) at a rate of 0.3 or 0.5 N min1 was applied. All measurements were carried out at room temperature. The compression elastic moduli (E) of the swollen hydrogels were extracted from the stress–strain curves. All compression elastic modulus data in this study were obtained from triplicate samples and data are expressed as means ± standard deviations. 2.6. Cytotoxicity of maleic chitosan and maleic chitosan/PEGDA hydrogel 2.6.1. Cell culture Bovine aortic endothelial cells (BAEC) were purchased from VEC Technologies. BAECs were maintained at 37 °C in 5% CO2 in Medium 199 (Invitrogen, Carlsbad, CA) supplemented with 10% Fetal Clone III (HyClone, Logan, UT) and 1% each of penicillin–streptomycin, minimal essential medium amino acids (Invitrogen, Carlsbad, CA) and minimal essential medium vitamins (Mediatech, Manassas, VA). BAECs were used between passages 8 and 12. The medium was changed every 2 days. Cells were grown to a minimum of 70% confluence before splitting or harvesting. Cell culture plates were treated with 2 wt.% gelatin aqueous solution before use. 2.6.2. Cytotoxicity of maleic chitosan Evaluation of the cytotoxicity of maleic chitosan was performed using the MTT assay. Maleic chitosan aqueous solution (5 wt.%) was obtained by dissolving maleic chitosan in PBS. Cultured BAECs at an appropriate cell density (5000 cells well1) were seeded into 96-well plates and incubated overnight. After 12 h the cells were treated with various volumes of freshly prepared aqueous maleic chitosan solution. Cells without maleic chitosan were used as the control. After 48 h treatment and incubation 20 ll of MTT solution (5 mg ml1) were added to each well, followed by 4 h incubation at 37 °C under a 5% CO2 atmosphere. The cell culture medium including polymer solution was then carefully removed and 200 ll of acidic isopropyl alcohol (with 0.1 M HCl) was added to dissolve the formazan crystals that had formed. The plate was gently shaken for 20 min to ensure that the purple crystals dissolved completely. Sample absorbance (OD) was measured at 570 nm and a background reading was taken at 690 nm using a microplate reader (VersaMax Tunable Microplate reader, Molecular Devices, USA). The relative cell proliferation (%) was calculated according to the equation:

relative cell proliferationð%Þ ¼ OD570ðsampleÞ  OD620ðsampleÞ





OD570ðcontrolÞ  OD620ðcontrolÞ  100%

where OD570(control) represents measurements from wells treated with medium only and OD570(sample) measurements from wells treated with various amounts of maleic chitosan solution. Dunnett’s test at P < 0.05 was used to determine any statistical significance of the MTT data. 2.6.3. Cytotoxicity of maleic chitosan/PEGDA hydrogels Evaluation of the cytotoxicity of maleic chitosan/PEGDA hydrogels was performed by Live–Dead assay (Invitrogen). Specifically, we selected maleic chitosan/PEGDA(8000) at a 1:3 weight feed ratio for this cytotoxicity study. The purified hydrogels were cut into round shapes (diameter 1 cm) and put into 24-well cell culture plates after 30 min UV sterilization. Cultured BAECs were seeded at an appropriate cell density (10,000 cells well1) and incubated overnight. After 24 h treatment and incubation the Live–Dead assay was performed according to the manufacturer’s protocol.

3911

3. Results and discussion 3.1. Synthesis and characterization of maleic chitosan The water-soluble precursor, maleic chitosan, was prepared based on a simple one-step chemical reaction between maleic anhydride and chitosan in toluene sulfonic acid/formamide mixed solvent. The yield of the final product could reach as high as 89% using this one-step process. In addition, the mixed solvent provided homogeneous mild reaction conditions under which polymer backbone degradation was not apparent, as indicated by retention of the molecular weight (52,000) of maleic chitosan rather than the original molecular weight (50,000) of chitosan. The chemical structure of maleic chitosan was relatively easy to determine because the photo-curable groups (double bonds) were always present in the same sites as carboxyl end groups. Several water-soluble chitosan derivatives, such as Az-CH-LA [18,19], methacrylated glycol chitosan [20] and methacrylated Ocarboxymethyl chitosan [21], have been reported before. However, the synthesis of those water-soluble chitosan precursors usually required a two-step process, including the preparation of a water-soluble chitosan derivative, followed by the incorporation of photo-crosslinkable moieties into the derivative, as chitosan itself is not soluble in neutral pH aqueous medium. Chitosan derivatives prepared via the two-step processes, therefore, usually had relatively lower yields, less defined chemical structures and randomly distributed photo-crosslinkable functional groups along the backbone. Unfortunately, complete information about the yields of the final products could not be found in the published reports. Identification of the chemical structures of those reported chitosan derivatives based on NMR spectra was also not given fully in the published reports. It was also noted that Yu et al. had reported a water-soluble, chitosan-based precursor, methacrylamide chitosan, prepared via a one-step reaction [22]. However, there are significant differences between their and our approaches. For example, in Yu’s report the reaction was carried out in a 2% acetic acid aqueous solution, which would result in significant polymer backbone degradation because of the acidic conditions (even though no molecular weight information was given in the report). Moreover, the degree of methacrylamide substitution on chitosan was kept low in order to obtain methacrylamide chitosan that was water-soluble. This lower degree of substitution also required a longer gelation reaction time (at least 2 h) compared with 10 min in our photo-induced gelation for hydrogel fabrication. In addition, the methacrylamide chitosan hydrogel was prepared by a thermal gelation process using ammonium persulfate (APS) and sodium metabisulfite (SMBS) as initiators. In addition to these differences in fabrication method, maleic chitosan prepared by our approach offers a completely different chemical functionality from that of methacrylamide chitosan. For example, the carboxyl functionality of maleic chitosan is anionic and can be used either to couple biologically active agents like nitric oxide derivatives [23] or serve as reactive sites to promote the deposition of inorganic minerals [24,31]. Specific chemical characterization of the maleic chitosan structure is discussed below. The incorporation of double bonds into the chitosan backbone was confirmed by 1H NMR, as indicated by the appearance of proton signals from ACH@CHA at d 5.95, 6.42 and 6.68 ppm (Fig. 1). This was confirmed by FTIR spectroscopy with the appearance of an absorption band at 808.6 cm1, corresponding to the unsaturated HAC@(ACH@CHA) out-of-plane deformation (Fig. 1a). Correspondingly, the presence of new peaks around d 120–150 ppm in 13C NMR spectrum was due to C signals from the double bonds, while the peaks around d 170–180 ppm in 13C NMR could be attributed to C signals from the carbonyl groups

3912

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

Fig. 2. (A) Optical images of maleic chitosan/PEGDA hybrid hydrogels at different weight feed ratios of maleic chitosan to PEGDA(700) after reaching swelling equilibrium: (a) maleic chitosan:PEGDA(700) = 1:1; (b) maleic chitosan:PEGDA(700) = 1:2; (c) maleic chitosan:PEGDA(700) = 1:3; (d) maleic chitosan:PEGDA(700) = 1:4; (e) maleic chitosan:PEGDA(700) = 1:5. (B) Typical optical image of maleic chitosan/PEGDA(8000) at a feed ratio of maleic chitosan to PEGDA(8000) of 1:5. When the weight feed ratio of maleic chitosan to PEGDA(700) increased beyond 1:1 no hydrogel could be formed.

Fig. 1. (a) 1H NMR and (b)

13

C NMR spectra of maleic chitosan.

because of the newly formed ester and amide bonds and carboxyl groups (Fig. 1b). The several signals from double bonds in both the 1H and 13C NMR spectra suggest that the substitution of maleic anhydride took place at more than one location in the chitosan structure. The total degree of substitution (DS) of maleic chitosan was around 1.17, with DS 0.42 and 0.75 at the C6 and C2 positions, respectively.

their backbone chain characteristics. Specifically, we believe the branched chain of maleic dextran (due to the branched nature of dextran) might be more favorable for forming a crosslinked network upon photo-crosslinking than linear maleic chitosan because branched maleic dextran can act as a crosslinker to itself. Therefore, in this study PEGDA was added as a crosslinker to facilitate photo-crosslinking of maleic chitosan, with Irgacure 2959 acting as the photoinitiator to initiate photo-crosslinking in an aqueous system with 10 min UV irradiation. The Irgacure 2959 initiator was previously reported to have minimal toxicity (cell death) over a broad range of mammalian cell types and species ranging from human fetal osteoblasts to bovine chondrocytes [33]. The hybrid hydrogels fabricated from low molecular weight PEGDA (Mn 700) usually became more opaque as the feed ratio of maleic chitosan to PEGDA precursors decreased beyond 1:2, as shown in Fig. 2A. In contrast, the hybrid hydrogels fabricated from high molecular weight PEGDA (Mn P 20,00) are almost all transparent, even those prepared from low weight feed maleic chitosan to PEGDA ratios (e.g. maleic chitosan/PEGDA(8000) hydrogel with a feed ratio of 1:5 (Fig. 2B)). The properties of these hybrid hydrogels, such as the swelling behavior, compression moduli and interior morphology were examined, as was the dependence of these properties on the precursor feed ratio and the molecular weight of PEGDA. The sample codes and the corresponding material parameters, including the feed ratios of the macromer precursors used, are summarized in Table S1 (supporting information).

3.2. Characterization and properties of maleic chitosan/PEGDA hybrid hydrogels

3.3. Characterization of maleic chitosan/PEGDA hydrogels

Theoretically, maleic chitosan should be able to undergo free radical crosslinking using a photoinitiator due to its pendant double bond moiety. In this study, however, we found that maleic chitosan itself at a low concentration (e.g. 3.5 wt.% solution) cannot form a hydrogel network. At high concentrations (e.g. 8.0 wt.% solution) maleic chitosan could form a hydrogel, but the resultant hydrogel easily dissociated in water after swelling. This dissociation could be partially due to the formation of a relatively loose hydrogel network because of the difficulty of self-crosslinking between double bonds hindered by carboxyl end groups. The photo-gelation reactivity of the pendant maleic acid was found to depend on the host macromolecules. Although maleic chitosan showed difficulty in forming a hydrogel by itself, maleic dextran (reported earlier by our group) can form hydrogels by itself under prolonged UV irradiation (40 min) [32]. The resultant dextran-based hydrogels were very stable, even after swelling. Thus, the differences in photo-gelation capability between maleic chitosan and maleic dextran could be ascribed to differences in

FTIR analysis was performed to ascertain the success of photocrosslinking between maleic chitosan and PEGDA. The IR spectra of the hydrogel samples were compared with those of the corresponding precursors. As shown in Fig. 3f, the vital peaks present in the IR spectrum of the PEGDA precursor (Mw 8000) were 2910 cm1 for alkylACH stretching, 1724.7 cm1 for >C@O stretching (acrylate) and 1100 cm1 for the ether (ACAOACA) group. The main peaks in the IR spectrum of the maleic chitosan precursor (Fig. 3a) were 3500–3200 cm1 due to AOH and ANH stretching, 1720 cm1 for >C@O ester stretching and 808.6 cm1 for the ACH< (>CH@CH<) out-of-plane deformation. The IR spectra of all the hydrogel samples were similar, although they differed slightly among each other due to different weight feed ratios of the macromer precursors. Briefly, the intensity of the peak around 3500– 3200 cm1 (AOH and ANH stretching of maleic chitosan) decreased as the weight feed ratio of maleic chitosan to PEGDA decreased (Fig. 3b–e). The characteristic absorption peaks of >CH@CH< at 808.6 cm1 (from maleic chitosan) almost disappeared in all

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

Fig. 3. IR spectra of (a) N,O-maleic chitosan and (b) maleic chitosan/PEGDA hydrogel at a weight feed ratio of 1/2; (c) maleic chitosan/PEGDA hydrogel at a weight feed ratio of 1/3; (d) Maleic chitosan/PEGDA hydrogel at a weight feed ratio of 1/4; (e) maleic chitosan/PEGDA hydrogel at a weight feed ratio of 1/5; (f) PEGDA precursor (Mn = 8000).

Fig. 4. Swelling kinetics of maleic chitosan/PEGDA hydrogels at different weight feed ratios of maleic chitosan to PEGDA in DI water. (a) Maleic chitosan:PEGDA(8000) 1:2; (b) maleic chitosan:PEGDA(8000) 1:3; (c) maleic chitosan:PEGDA(8000) 1:4; (d) maleic chitosan:PEGDA(8000) 1:5; (e) pure PEGDA(8000) hydrogel (control).

hydrogel samples. This observation indicated that maleic chitosan and PEGDA successfully crosslinked to form hydrogels. Meanwhile, shifting of the characteristic absorption peaks of >C@O from both 1724 cm1 (ester groups of PEGDA due to conjugation with >C@C< before reaction) and 1719 cm1 (ester groups of maleic chitosan due to conjugation with >C@C< before reaction) to an overlapping normal (saturated) ester group (1736 cm1) in the hydrogel samples also suggested coupling of the two macromer precursors. 3.4. Equilibrated swelling ratio and swelling kinetics of maleic chitosan/PEGDA in PBS and deionized water The swelling kinetics of the maleic chitosan/PEGDA hydrogels were studied over a period of 5 days in DI water. As shown in Fig. 4, all the maleic chitosan/PEGDA hydrogels of varying weight feed ratios generally showed a high swelling rate during the initial 3 h, leveling off thereafter and finally reaching equilibrium in about 2 days. Moreover, the swelling rates of these hybrid hydrogels were generally higher than that of a pure PEGDA hydrogel,

3913

Fig. 5. Dependence of equilibrated swelling ratio on the weight feed ratio of maleic chitosan to PEGDA (Mn = 8000) in PBS (pH 7.4) solution and DI water. (a) Maleic chitosan:PEGDA(8000) 1:2; (b) maleic chitosan:PEGDA(8000) 1:3; (c) maleic chitosan:PEGDA(8000) 1:4; (d) maleic chitosan:PEGDA(8000) 1:5; (e) pure PEGDA hydrogel (control).

and decreased with a decrease in the weight feed ratio of maleic chitosan to PEGDA. The equilibrium swelling ratio was determined in both DI water and PBS. Fig. 5 shows the dependence of equilibrated swelling ratios at room temperature on the weight feed ratio of maleic chitosan to PEGDA (Mn 8000) hybrid hydrogels. The equilibrated swelling ratio of the hybrid hydrogels generally decreased with a decrease in the weight feed ratio of maleic chitosan to PEGDA in both DI and PBS due to the increase in crosslinking density. For example, the hybrid hydrogel with a weight feed ratio of maleic chitosan to PEGDA of 1:2 had the highest equilibrium swelling ratio (17,610 ± 450% and 15,880 ± 490% in DI water and PBS, respectively), while the hybrid hydrogel sample with a weight feed ratio of 1:5 had the smallest swelling ratio (2740 ± 30% and 2020 ± 46% in DI water and PBS, respectively). The swelling ratios of all maleic chitosan/PEGDA hybrid hydrogels were higher than that of pure PEGDA control hydrogel (1620 ± 21% and 1650 ± 34% in DI water and PBS, respectively), implying that the incorporation of carboxyl moieties into the maleic chitosan segment enhanced the hydrophilicity and hence swelling of the hybrid hydrogels. This relationship of the effect of precursor feed ratio on swelling ratio was also found to be consistent with the effect of precursor feed ratio on the compressive moduli of the hydrogels, as described later. The fast swelling kinetics and high water retention capability of these hybrid hydrogels are due to their hydrophilic nature. Not all hybrid hydrogels reported in the literature show such fast and high levels of swelling. For example, Guo et al. reported that a relatively hydrophobic hybrid hydrogel family containing amino acid-based poly(ester amide) (PEA)/PEGDA precursors had much slower swelling kinetics (12 h before the swelling rate leveled off) and lower equilibrium swelling ratios (between 1% and 2400% in water) [30]. The swelling ratios of maleic chitosan/PEGDA hybrid hydrogels in DI water were always higher than those of the corresponding hydrogels in PBS for all the feed ratios studied (Fig. 5a–d), however, the difference became smaller as the feed ratio of maleic chitosan to PEGDA decreased. No meaningful differences in swelling ratio between PBS and DI water were observed for the pure PEGDA hydrogel (Fig. 5e). This observation suggests that the electrolytes in PBS might interfere with interaction between water molecules and carboxyl groups in the hybrid hydrogels. To further understand the effect of ions on the swelling ratio of these hybrid hydrogels we studied the effect of pH (varying H+ concentration) on the swelling ratios of the maleic chitosan/PEGDA hybrid hydrogels. The results are shown in Fig. S1 (supporting information). As expected, the swelling ratios of the hybrid

3914

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

hydrogels were very sensitive to changes in pH, and the pH sensitivity depended on the feed ratio of maleic chitosan to PEGDA. In general, all the four hybrid hydrogels showed an increase in swelling ratio with an increase in pH. The effect of pH on swelling ratio became more pronounced as the maleic chitosan to PEGDA feed ratio increased. For example, for the hybrid hydrogels with a weight feed ratio of 1:2 the swelling ratio at highly basic pH (pH 11) was almost three times that under acidic conditions (pH 3). This pH-dependent swelling ratio of maleic chitosan/PEGDA hybrid hydrogels has also been reported for other polymeric systems, including PEG-grafted poly(methacrylic acid) (PMAA) [34], maleic dextran hydrogels [32] and poly(vinyl alcohol)–co-maleic anhydride hydrogels [35]. It was suggested that hydrogen bond complexation between carboxyl groups and hydroxyl groups in hydrogel networks under acidic conditions leads to a reduction in swelling ratio. Similarly, hydrogen bonds may also form in maleic chitosan/PEGDA hybrid hydrogels due to the presence of both carboxyl (from the maleic acid) and hydroxyl (from the chitosan backbone) groups under acidic conditions, thus resulting in a lower swelling ratio. In a highly alkaline environment, however, swelling of the maleic chitosan/PEGDA hybrid hydrogels was different from pure maleic dextran hydrogels (no PEGDA). In the present maleic chitosan/PEGDA hybrid hydrogel case the swelling ratio continued to increase at pH 11, while the maleic dextran hydrogels showed a slight reduction in swelling ratio at the same pH [32]. We further investigated the effect of molecular weight of the co-precursor (PEGDA) on the swelling properties of the hybrid hydrogels at a constant molar ratio of maleic chitosan to PEGDA of 1:0.267 (see Table S1). As shown in Fig. S2 (supporting information), the swelling ratio decreased as the molecular weight of PEGDA increased. At this molar ratio (1:0.267) maleic chitosan/PEGDA hydrogels could not form when Mw 700 PEGDA was used, however, at other molar ratios (shown in Fig. 2) hybrid hydrogels could form with Mw 700 PEGDA. These observations suggest that lower molecular weight PEGDA might be less effective in crosslinking with maleic chitosan, probably due to relatively short root mean square distance between the molecules. 3.5. Mechanical property (compressive modulus) of maleic/PEGDA hybrid hydrogels Typical stress–strain curves for maleic chitosan/PEGDA hydrogels at different weight feed ratios of maleic chitosan to PEGDA(8000) are shown in Fig. 6A. The calculated compressive moduli based on the linear fit approach are shown in Fig. 6B. The average moduli of maleic chitosan/PEGDA hybrid hydrogels in-

creased with a decrease in maleic chitosan to PEGDA weight feed ratio; for example, for hybrid hydrogels at weight feed ratios of 1:2, 1:3, 1:4 and 1:5 the compressive moduli were 61 ± 1.9, 125.5 ± 11.1, 310.8 ± 23.7 and 560.4 ± 18.1 kPa, respectively. This trend is a result of the increased crosslinking densities of the hybrid hydrogel networks due to the decreased weight feed ratio of maleic chitosan to PEGDA. As described previously, the crosslinking density of these hybrid hydrogels also affected their swelling ratios, suggesting that the swelling of these hybrid hydrogels became more limited as the crosslinking density increased. A similar relationship between the swelling ratio, compressive modulus and crosslinking density was reported in a relatively hydrophobic hybrid hydrogel family fabricated from amino acid-based poly(ester amide) and PEGDA precursors [30]. We also tested the effect of the molecular weight of the crosslinker, PEGDA, on the mechanical properties of the hybrid hydrogels by maintaining a constant molar ratio of maleic chitosan to PEGDA (Table S1). As shown in Fig. S3, the compressive modulus appeared to increase as the molecular weight of PEGDA increased, with values of 60.9 ± 6.5, 141.0 ± 8.2 and 310.8 ± 23.7 for maleic chitosan/PEGDA(2000), maleic chitosan/PEGDA(4000) and maleic chitosan/PEGDA(8000), respectively. This trend could be directly relevant to the elastic chain length of PEGDA molecule. Each crosslinked chain of PEGDA can be considered a small spring at the molecular level, therefore, their capability to resist compression depends on the length of the spring, with a longer chain being more resistant to compression. 3.6. Interior morphology (SEM) of maleic chitosan/PEGDA hydrogels To understand the structure–function relationship of the maleic chitosan/PEGDA hybrid hydrogels we selectively examined their cross-sectional interior morphology. All the freeze-dried hydrogels displayed porous network structures (Fig. 7). The average pore size of these honeycomb-like porous structures decreased with an increase in the amounts of crosslinker, PEGDA. For example, MCPEGDA5 (Fig. 7g) had the smallest pore size, with an average diameter of 5 lm, while MC-PEGDA2 (Fig. 7a) had the largest, with an average diameter of 20 lm. Our interior morphological data shown in Fig. 7 are thus consistent with the effect of the precursor feed ratio on swelling ratios and compressive moduli data discussed above. In addition to the dependence of pore size on the molecular dimensions and hydrophilicity of the macromer, previous studies of other hydrogel systems suggested that the pore size of hydrogels also depends on the actual crosslinking density [30,36,37].

Fig. 6. (A) Typical stress–strain curves of maleic chitosan/PEGDA hydrogels at different weight feed ratios of maleic chitosan to PEGDA. (B) Calculated compression moduli of maleic chitosan/PEGDA hydrogels at different weight feed ratios of maleic chitosan to PEGDA. (a) Maleic chitosan:PEGDA(8000) 1:2; (b) Maleic chitosan:PEGDA(8000) 1:3; (c) Maleic chitosan:PEGDA(8000) 1:4; (d) Maleic chitosan:PEGDA(8000) 1:5.

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

3915

Fig. 7. SEM images of maleic chitosan/PEGDA hybrid hydrogels at different weight feed ratios of maleic chitosan to PEGDA(8000). (a and b) Maleic chitosan:PEGDA(8000) 1:2; (c and d) maleic chitosan:PEGDA(8000) 1:3; (e and f) maleic chitosan:PEGDA(8000) 1:4; (g and h) maleic chitosan:PEGDA(8000) 1:5. Note: in the discussion in Section 3.6 MC-PEGDA2 and MC-PEGDA5 stand for the hybrid hydrogels with weight feed ratios of maleic chitosan:PEGDA(8000) of 1:2 and 1:5, respectively.

Fig. 8. Representative micrographs of BAECs after 48 h culture. (a) Cells without any maleic chitosan solution (control); (b) cells treated with 60 ll of 5 wt.% maleic chitosan solution; (c) cells treated with 100 ll of 5 wt.% maleic chitosan solution.

3916

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

Fig. 9. BAEC viability (%) after incubation in various amounts of maleic chitosan solution for 4 and 48 h, respectively, based on results from a 3-(4,5-dimethylthiazol-2-yl)2,5-diphenyltetrazolium bromide (MTT) assay. Cells in the absence of maleic chitosans were used as control samples. Note: cell proliferation (viability) of the test samples on the y-axis is expressed as a percentage of the blank control. Error bars represent means ± SEM. *P < 0.05. Number indicates the polymer concentration.

Fig. 10. Fluorescent images from a Live/Dead Assay of BAECs attached on the maleic chitosan/PEGDA hybrid hydrogel surface (weight feed ratio maleic chitosan:PEGDA(8000) 1:3). Live cells were stained green, while dead cells were stained red. (a) Image showing viable and healthy BAECs attached to the hydrogel surface; (b) image showing dead or unhealthy cells on the hydrogel surface.

In the maleic chitosan/PEGDA hybrid hydrogel system as the maleic chitosan to PEGDA feed ratio decreased the crosslinking density of the hybrid hydrogel networks increased, thus resulting in a tighter network structure, reflected in the observed higher compressive moduli (Fig. 9) and lower swelling ratios (Figs. 4 and 5). Our results are also in agreement with previous studies of other types of gels regarding the effect of crosslinking density on the swelling ratio and mechanical properties. For example, Guo et al. reported that the elastic modulus of unsaturated poly(ester amide)/PEGDA hybrid hydrogels was directly proportional to their crosslinking density, while the swelling ratio had an inverse relationship with the crosslinking density [30]. Similarly, Ortega suggested that an increase in effective crosslink density with increasing crosslinking agent concentrations in tert-butyl acrylate–co-PEGDA gel networks led to decreasing equilibrium swelling ratios and increasing elastic moduli [38]. 3.7. Cytotoxicity of maleic chitosan and maleic chitosan/PEGDA hydrogels The cytotoxicity of maleic chitosan and maleic chitosan/PEGDA hydrogels was evaluated as cell morphology and proliferation. The cytotoxicity of maleic chitosan was assessed based on the MTT assay, while the maleic chitosan/PEGDA hydrogels were tested using a Live–Dead assay.

As shown in Fig. 8, no significant change in morphology of the BAEC cells was detected under phase contrast microscopy after 2 days treatment, with either 60 or 100 ll 5 wt.% maleic chitosan solution. There were no visible signs of cell rounding, increased vacuolization or membrane blebbing, which are typical indicators of cell death. The results of the MTT assay (Fig. 9) indicated that, when compared with the blank control, more than 80% of cells treated with maleic chitosan were viable up to 60 ll dosage at 4 h and to 10 ll dosage at 48 h. Even at the highest maleic chitosan dosage (i.e. 100 ll) more than 60% of the BAEC cells remained viable at 48 h when compared with the blank control. Statistical analysis of the 4 h MTT data indicated that all of the maleic chitosan treatment conditions, except the highest dosage (100 ll), were statistically similar to the blank control. At 48 h the four highest maleic chitosan dosages (20–100 ll) were statistically lower compared with the blank control at P < 0.05 (Dunnett’s test), however, the two lowest maleic chitosan dosages (3 and 10 ll) were not statistically significantly different from the blank control. The results of these statistical analyses indicate that the toxicity of maleic chitosan is dosage dependent and that low dosages would be advisable for biomedical applications. It is important to note that the MTT assay measures the relative number of cells compared with a control (no maleic chitosan solution present); it is not possible to use the MTT assay alone to

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

determine whether the assay either causes cell death or inhibits proliferation. However, given the BAEC cell morphological data in Fig. 8, where the maleic chitosan treated cells appeared as wellspread and healthy as the blank control, it is likely that very high dosages of maleic chitosan solution may slow proliferation rather than cause cell death. Because the doubling time of these cells is slightly less than 24 h, it does not appear to completely halt proliferation within the dose range tested. The results indicate that maleic chitosan is only minimally cytotoxic at high doses and that it may have potential as a new biomaterial for various biomedical applications. However, to fully understand the effects of maleic chitosan on long-term cell behavior, additional experiments that investigate viability and cell adhesion strength over a longer time scale should be performed. The MTT assay is unsuitable to evaluate the cytotoxicity of maleic chitosan/PEGDA hydrogels because of the limited number of BAEC cells attached to the hydrogel surface. Therefore, a Live– Dead assay was adopted to assess the cytotoxicity of maleic chitosan/PEGDA hydrogels. Because we aimed only to test the viability of cells attached to the surface of the hydrogels in this study, the Live–Dead assay could be used as a preliminary evaluation of hydrogel cytotoxicity. As shown in Fig. 10, most of the attached BAEC cells were viable (stained green) and very few cells were dead (stained red) on the maleic chitosan/PEGDA hybrid hydrogel (1:3 weight feed ratio) surface. While the maleic chitosan/PEGDA hydrogel surface was only minimally adhesive for cells, those cells that did attach remained viable, implying that maleic chitosan/ PEGDA hydrogels are not measurably toxic to BAECs. Because our data indicate that the material favors cell viability, future studies could include modifying that maleic chitosan/PEGDA hydrogel with cell-adhesive ligands to increase cell adhesion. 4. Conclusions We synthesized and characterized a new type of pH-sensitive and low cytotoxicity maleic chitosan/PEGDA hybrid hydrogels synthesized via UV photo-crosslinking in an aqueous solution. We demonstrated that by varying the feed ratio of maleic chitosan to PEGDA and the molecular weight of PEGDA we could tune the swelling, mechanical and morphological properties of the resulting hybrid hydrogels. This study contributes to an understanding of the structure–properties relationships of an acidic polysaccharide hydrogel-based biomaterial. Acknowledgements This project was partially supported by the graduate student thesis funds from the College of Human Ecology at Cornell University and the Morgan Tissue Engineering Program. We are grateful for Prof. Martha A. Mutschler for her assistance in statistical analysis of the MTT data. This work made use of the electron microscopy and DMA facilities of the Cornell Center for Materials Research with support from the National Science Foundation Materials Research Science and Engineering Centers program (DMR 0520404). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.actbio.2010.04.011. Appendix B. Figures with essential colour discrimination Certain figures in this article, particularly Figures 2, 4, 5, 6, 9 and 10, are difficult to interpret in black and white. The full colour

3917

images can be found in the on-line version, at doi:10.1016/ j.actbio.2010.04.011.

References [1] Hoffman AS. Hydrogels for biomedical applications. Adv Drug Deliv Rev 2002;54:3–12. [2] Lee KY, Mooney DJ. Hydrogels for tissue engineering. Chem Rev 2001;101:1869–79. [3] Peppas NA, Hilt JZ, Khademhosseini A, Langer R. Hydrogels in biology and medicine: from molecular principles to bionanotechnology. Adv Mater 2006;18:1345–60. [4] Ruan CM, Zeng KF, Varghese OK, Grimes CA. A magnetoelastic bioaffinitybased sensor for avidin. Biosens Bioelectron 2004;19:1695–701. [5] Kuang M, Wang DY, Bao HB, Gao MY, Mohwald H, Jiang M. Fabrication of multicolor-encoded microspheres by tagging semiconductor nanocrystals to hydrogel spheres. Adv Mater 2005;17:267–70. [6] Lutolf MP, Hubbell JA. Synthetic biomaterials as instructive extracellular microenvironments for morphogenesis in tissue engineering. Nat Biotechnol 2005;23:47–55. [7] Decker C. UV-curing chemistry – past, present, and future. J Coat Technol 1987;59:97–106. [8] Nguyen KT, West JL. Photopolymerizable hydrogels for tissue engineering applications. Biomaterials 2002;23:175–8. [9] Kumar M. A review of chitin and chitosan applications. React Funct Polym 2000;46:1–27. [10] Rabea EI, Badawy MET, Stevens CV, Smagghe G, Steurbaut W. Chitosan as antimicrobial agent: applications and mode of action. Biomacromolecules 2003;4:1457–65. [11] Rinaudo M. Chitin and chitosan: properties and applications. Prog Polym Sci 2006;31:603–32. [12] Singla AK, Chawla M. Chitosan: some pharmaceutical and biological aspects – an update. J Pharm Pharmacol 2001;53:1047–67. [13] Dodane V, Vilivalam VD. Pharmaceutical applications of chitosan. Pharma Sci Technol Today 1998;1:246–53. [14] Borchard G. Chitosans for gene delivery. Adv Drug Deliv Rev 2001;52:145–50. [15] Khor E, Lim LY. Implantable applications of chitin and chitosan. Biomaterials 2003;24:2339–49. [16] Madihally SV, Matthew HWT. Porous chitosan scaffolds for tissue engineering. Biomaterials 1999;20:1133–42. [17] Suh JKF, Matthew HWT. Application of chitosan-based polysaccharide biomaterials in cartilage tissue engineering: a review. Biomaterials 2000;21:2589–98. [18] Ono K, Saito Y, Yura H, Ishikawa K, Kurita A, Akaike T, et al. Photocrosslinkable chitosan as a biological adhesive. J Biomed Mater Res 2000;49:289–95. [19] Obara K, Ishihara M, Ishizuka T, Fujita M, Ozeki Y, Maehara T, et al. Photocrosslinkable chitosan hydrogel containing fibroblast growth factor-2 stimulates wound healing in healing-impaired db/db mice. Biomaterials 2003;24:3437–44. [20] Amsden BG, Sukarto A, Knight DK, Shapka SN. Methacrylated glycol chitosan as a photopolymerizable biomaterial. Biomacromolecules 2007;8: 3758–66. [21] Poon YF, Bin Zhu Y, Shen JY, Chan-Park MB, Ng SC. Cytocompatible hydrogels based on photocrosslinkable methacrylated O-carboxymethylchitosan with tunable charge: synthesis and characterization. Adv Funct Mater 2007;17: 2139–50. [22] Yu LM, Kazazian K, Shoichet MS. Peptide surface modification of methacrylamide chitosan for neural tissue engineering applications. J Biomed Mater Res A 2007;82:243–55. [23] Jokhadze G, Machaidze M, Panosyan H, Chu CC, Katsarava R. Synthesis and characterization of functional elastomeric poly(ester amide) co-polymers. J Biomater Sci Polym Ed 2007;18:411–38. [24] Zhong C, Chu CC. Biomimetic mineralization of acid polysaccharide-based hydrogels: towards porous 3-dimensional bonelike biocomposites, in preparation. [25] Varghese S, Elisseeff JH. Hydrogels for musculoskeletal tissue engineering. Polym Regen Med 2006:95–144. [26] Cruise GM, Hegre OD, Scharp DS, Hubbell JA. A sensitivity study of the key parameters in the interfacial photopolymerization of poly(ethylene glycol) diacrylate upon porcine islets. Biotechnol Bioeng 1998;57:655–65. [27] Cruise GM, Scharp DS, Hubbell JA. Characterization of permeability and network structure of interfacially photopolymerized poly(ethylene glycol) diacrylate hydrogels. Biomaterials 1998;19:1287–94. [28] Roberts GAF, Domszy JG. Determination of the viscometric constants for chitosan. Int J Biol Macromol 1982;4:374–7. [29] Li Q, Wang J, Shahani S, Sun DDN, Sharma B, Elisseeff JH, et al. Biodegradable and photocrosslinkable polyphosphoester hydrogel. Biomaterials 2006;27:1027–34. [30] Guo K, Chu CC. Synthesis and characterization of novel biodegradable unsaturated poly(ester amide)/poly(ethylene glycol) diacrylate hydrogels. J Polym Sci A Polym Chem 2005;43:3932–44. [31] Zhong C, Chu CC. Acid polysaccharide-induced amorphous calcium carbonate (ACC) films: colloidal nanoparticle self-organization process. Langmuir 2009;25:3045–9.

3918

C. Zhong et al. / Acta Biomaterialia 6 (2010) 3908–3918

[32] Kim SH, Won CY, Chu CC. Synthesis and characterization of dextran–maleic acid based hydrogel. J Biomed Mater Res 1999;46:160–70. [33] Williams CG, Malik AN, Kim TK, Manson PN, Elisseeff JH. Variable cytocompatibility of six cell lines with photoinitiators used for polymerizing hydrogels and cell encapsulation. Biomaterials 2005;26:1211–8. [34] Klier J, Scranton AB, Peppas NA. Self-associating networks of poly(methacrylic acid-g-ethylene glycol). Macromolecules 1990;23:4944–9. [35] Liou FJ, Wang YJ. Preparation and characterization of crosslinked and heattreated PVA–MA films. J Appl Polym Sci 1996;59:1395–403.

[36] Wang D-AE, Jennifer H. Photopolymerization. New York: Marcel Dekker; 2004. [37] Sperling LH. Introduction to physical polymer science. New York: Wiley; 1992. [38] Ortega AM, Kasprzak SE, Yakacki CM, Diani J, Greenberg AR, Gall K. Structure– property relationships in photopolymerizable polymer networks: effect of composition on the crosslinked structure and resulting thermomechanical properties of a (meth)acrylate-based system. J Appl Polym Sci 2008;110: 1559–72.