- Email: [email protected]
Thermal-crosslinked porous chitosan scaffolds for soft tissue engineering applications
Materials Science and Engineering C 33 (2013) 3780–3785
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Thermal-crosslinked porous chitosan scaffolds for soft tissue engineering applications Chengdong Ji ⁎, Jeffrey Shi School of Chemical and Biomolecular Engineering, The University of Sydney, Sydney, NSW 2006, Australia
a r t i c l e
i n f o
Article history: Received 28 November 2012 Received in revised form 28 March 2013 Accepted 3 May 2013 Available online 13 May 2013 Keywords: Chitosan porous scaffolds Steam sterilization Thermal-crosslinking Soft tissue engineering
a b s t r a c t The aim of this study was to demonstrate the feasibility of using a steam autoclave process for sterilization and simultaneously thermal-crosslinking of lyophilized chitosan scaffolds. This process is of great interest in biomaterial development due to its simplicity and low toxicity. The steam autoclave process had no significant effect on the average pore diameter (~ 70 μm) and overall porosity (> 80%) of the resultant chitosan scaffolds, while the sterilized scaffolds possessed more homogenous pore size distribution. The sterilized chitosan scaffolds exhibited an enhanced compressive modulus (109.8 kPa) and comparable equilibrium swelling ratio (23.3). The resultant chitosan scaffolds could be used directly for in vitro cell culture without extra sterilization. The data of in vitro studies demonstrated that the scaffolds facilitated cell attachment and proliferation, indicating great potential for soft tissue engineering applications. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Chitosan, a linear polysaccharide, holds great potential in tissue engineering applications due to its readily availability, unique physicochemical properties, biocompatibility and biodegradability [1–3]. Chitosan has been fabricated in the form of lyophilized porous scaffolds to support cell attachment and proliferation, and eventually to facilitate new tissue formation after chitosan degradation [3]. Sterilization of the porous scaffolds remains an issue for tissue engineering applications. Chitosan scaffolds are usually sterilized by ethanol rinse or by UV exposure before laboratory-level in vitro studies [4,5], while for in vivo and clinical studies, ethylene oxide (EtO) exposure or gamma ray (Cobalt-60) irradiation are compulsory and more efficient [6–8]. However, such sterilants are highly hazardous not only to the host tissues and organs, but also to the operators [9,10]. The complete evacuation of these reagents (especially for porous materials) usually takes weeks-to-months, which is time-intensive [7,11]. Steam autoclave is a common sterilization method for thermalinert medical devices or buffered solution [12]. Compared with EtO exposure and gamma-ray irradiation, this process is simpler and more environmentally-friendly, while it is rarely applied for tissue engineering scaffolds given that most biomaterials have low thermal resistance. Our previous study reported that steam autoclave was used to sterilize and simultaneously induce chitosan-based hydrogel formation, and the resultant hydrogels are potent for drug delivery and tissue engineering applications [13]. In this study, we evaluated the feasibility of using steam autoclave to sterilize lyophilized chitosan scaffolds. The effect of sterilization on ⁎ Corresponding author at: Chemical Engineering Building J01, The University of Sydney, Sydney, NSW 2006 Australia. Tel.: + 61 2 9351 3411; fax: + 61 2 9351 2854. E-mail address: [email protected] (C. Ji). 0928-4931/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2013.05.010
the scaffolds' performances such as porosity, swelling behavior and mechanical properties were demonstrated. Cells were statically seeded on the sterilized scaffolds and in vitro culture was subsequently undertaken. The cell attachment and proliferation within chitosan scaffolds were investigated. 2. Materials and methods 2.1. Materials Chitosan (Catalog No. 448877, medium molecular weight and deacetylation degree of 75–85%), fluorescein diacetate (FDA), propidium iodide (PI), and Dulbecco's modified eagle medium (DMEM), were purchased from Sigma-Aldrich. MTS assay kit, fetal bovine serum (FBS) and penicillin–streptomycin solution (pen–strep) were purchased from Invitrogen. A 0.2 M acetic acid solution was prepared by diluting glacial acetic acid (Ajax Fine Chem) in MilliQ water. Chitosan solution (1.5 wt.%) was prepared by dissolving chitosan powder in 0.2 M acetic acid solution. Phosphate buffered saline (PBS, pH 7.2–7.4) was prepared by dissolving PBS tablets (Sigma) in MilliQ water. 2.2. Fabrication and sterilization of chitosan porous scaffolds Chitosan solution as prepared above was poured into a custommade mold and lyophilized for 30 h. The lyophilized scaffolds were collected and rinsed extensively by PBS to remove acidic residue, and eventually lyophilized again. The steam sterilization was held in a steam autoclave (Shenan Medical Devices) chamber for 30 min at 121 °C (Fig. 1A). Given the knowledge that chitosan possessed a melting temperature around 130–140 °C [14], the use of high temperature incubation 121 °C would not change the thermal behavior of chitosan.
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
A
Lyophilized porous scaffold
Sterilized porous scaffold Steam sterilization
B
3781
2.4.4. Mechanical strength Uniaxial compression tests were performed in a hydrated state (in PBS) at room temperature by using an Instron (Model 5543) with a 500 N load cell. Prior to mechanical testing, the scaffolds were immersed in PBS for at least 2 h; The compression (mm) and load (N) were collected at a crosshead speed of 30 μm/s until 60% compression was achieved. The compressive moduli were then calculated as the tangent slope of the stress–strain curves within linear regions (10–20% strain rate). 2.4.5. Swelling behavior The swelling behaviors of the scaffolds were evaluated at 37 °C in PBS. The initial dry weights were recorded (W0). After immersion in excessive PBS for different time intervals (1–24 h), the swollen chitosan scaffolds were weighted (Wt). The swelling ratio was subsequently calculated as (Wt − W0) / W0.
Fig. 1. A: Schematic diagram of chitosan scaffolds' fabrication and sterilization; and B: image of chitosan scaffolds before (left) and after (right) steam sterilization.
2.3. In vitro cell culture In vitro cell culture was performed on sterilized chitosan scaffolds under a biosafety cabinet. Each scaffold (~10 mm diameter and 3 mm thickness) was immersed in culture media (DMEM, 10% FBS, 1% pen– strep) at 37 °C for at least 2 h. The cells (human skin fibroblast cells GM3348) were then seeded (pipetted) onto the scaffold at a concentration of 1 × 10 5 cells/sample (for MTS assay, a cell concentration of 1 × 10 4 cells/sample was used). The cell-seeded scaffolds were kept in a CO2 incubator (Thermo Fisher HERAcell 150i) at 37 °C for further characterizations. The media was refreshed every two days. 2.4. Characterizations 2.4.1. Fourier transform infrared (FTIR) spectroscopy The molecular interactions of chitosan scaffolds upon steam sterilization were determined by using Fourier transform infrared (FTIR) spectroscopy (Varian 660-IR) with a resolution of 4 cm −1, averaging for 32 scans. 2.4.2. Gel permeation chromatography (GPC) The molecular weight of fresh and steam sterilized lyophilized chitosan scaffold was determined by using GPC as previously described [15]. In brief, each chitosan scaffold was dissolved in an acetic acid (0.3 M)/sodium acetate (0.2 M) buffer with a pH of 4.5 to achieve a concentration 0.5 mg/mL. Magnetically stirring and a moderate heating (~40 °C) were helpful for complete dissolution of sterilized chitosan. The chitosan solution was eventually filtered through a 0.45 μm membrane (Millipore) before GPC (Perkin Elmer Series 200) analysis. The molecular weight was calculated from a calibration curve using a series of poly ethylene glycol (PEG) with known molecular weights. 2.4.3. Surface morphology The porous structure on the surface of resultant scaffolds was visualized by using a phase contrast microscope (Leica). The porous structure on the cross section of resultant scaffolds was examined after cryo-fracture as previously described [16]. In brief, each fabricated scaffold was immersed in liquid nitrogen for 45 s, and cut by using a pre-cooled single edge razor blade. The fractured cross section was then mounted on glass slide for observation under a light microscope (Leica). The equivalent circle diameter (ECD) was then calculated by using ImageJ software. At least 300 pores were analyzed for each condition.
2.4.6. Overall porosity The overall porosities of resultant chitosan scaffolds were evaluated as previously described [17]. In brief, each scaffold was submerged in absolute ethanol of a known volume (V1), and a series of vacuumrelease cycles was performed to force the liquid into the pores of the scaffold. After these cycles, the volume of the liquid and liquidimpregnated scaffold was recorded as V2. When the liquid-impregnated scaffold was removed, the remaining liquid volume was recorded as V3. The overall porosity was given as [(V1 − V3) / (V2 − V3)] × 100% (n = 9). 2.4.7. Live/dead staining Cell proliferation in the resultant scaffolds was examined by live/dead staining. The cell-seeded scaffolds were stained with FDA and PI (both 1 μg/mL in PBS) for 5 min. The stained samples were assessed by using a fluorescence microscope (Eclipse E800, Nikon). Live cells were stained with fluorescent green due to intracellular esterase activity that de-acetylated FDA to a green florescent product. Dead cells were stained with fluorescent red as their compromised membranes were permeable to nucleic acid stain (PI). Percent cell viability values were calculated by counting the number of live (green) cells and the number of dead (red) cells on images (10× magnification). The values were obtained by dividing the number of live cells by the number of total cells (live cells + dead cells). A statistical significance level of 99.5% (p b 0.005) was considered to avoid potential human error in cell counting (n = 9). 2.4.8. In vitro proliferation assay In vitro cell proliferation was examined by using MTS assay. The cell-seeded scaffolds (1 × 104 cells/sample) were immersed in culture medium within 48 well-plates (n = 10). At different time intervals (3 h, 1 day, 4 and 7 days), the samples were rinsed by PBS three times; 250 μL fresh medium and 50 μL MTS was subsequently added into each well. The samples were then kept in a CO2 (5% CO2 and 95% humidity) incubator at 37 °C for 1 h; allowing MTS to react with metabolically active cells and subsequently result in water-soluble formazan product quantifiable by the optical density (OD) at 490 nm by using a microplate reader (Bio Rad 680). 2.5. Statistical analysis Each test was repeated three times except elsewhere mentioned. The statistical significance was determined at each condition by an independent Student's t-test for two groups of data using SPSS statistical software (PASW Statistics 18). Data are represented as mean ± standard deviation (SD). Confidence level of 95% (p b 0.05) was considered as statistically significant except elsewhere mentioned.
3782
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
3. Results and discussion
Table 1 Molecular weight of chitosan scaffolds.
3.1. General appearance White and rigid scaffolds were obtained after lyophilization, while the sterilized scaffolds exhibited about 10% shrinkage from 15.8 ± 0.2 mm to 14.4 ± 0.1 mm in diameter, and turned to be yellowish in color as shown in Fig. 1B (measured by using a digital caliper (J.B.S)). The sterilization process had negligible effect on the thickness of the scaffolds (~2.5 mm). 3.2. FTIR analysis Shrinkage in size and color change are common symbols of crosslinking of chitosan [5]. We thus used FTIR spectroscopy to characterize the effect of steam sterilization on chitosan scaffolds at a molecular level. As expected, unsterilized chitosan scaffold exhibited a characteristic peak at 1636 cm − 1, corresponding to the C_O stretch for amide I, which indicated that chitosan was not completely deacetylated, and another peak was detected at 1547 cm − 1 (Fig. 2a), which corresponds to either primary amines or amide II (N\H) [18]. After steam sterilization, peak shifts were observed on both characteristic peaks (from 1637 to 1651 cm−1, and 1547 to 1556 cm−1, respectively) indicating the molecular interaction, (Fig. 2b). The exact mechanism during steam sterilization is still under investigation, while we anticipated that possible mechanism is the molecular interaction between amino groups and either the primary hydroxyl groups or the carbonyl groups under high temperature [19]. As an evidence, the sterilized chitosan scaffolds turned to be yellowish (Fig. 1B), which is a common phenomenon of crosslinking based on amino groups [5]. 3.3. Molecular weight determination
R ela tiv e A b so rb a n ce (a .u .)
The molecular weight of chitosan scaffolds was listed in Table 1. Fresh lyophilized chitosan scaffolds showed a weight average molecular weight (Mw) of 230 ± 24 kDa and number average molecular weight (Mn) of 131 ± 7 kDa, with an average polydispersity index (PDI) of 1.76. After steam sterilization, the Mw, Mn and average PDI of the chitosan scaffolds was 272 ± 19 kDa, 158 ± 4 kDa and 1.72, respectively. Statistical analysis revealed that steam sterilization resulted in enhanced Mw and Mn, indicating the presence of crosslinking. Moreover, we surprisingly found that the sterilized chitosan scaffolds showed enhanced acid-resistance. In practice, the unsterilized chitosan scaffolds were completely dissolved in acidic solution (0.1% HCl) within 2 h, while the sterilized chitosan scaffolds could maintain their integrity in acidic environment for over one month. These data are in well agreement with previous studies which corroborated that
1556 1651
b: Sterilized chitosan
a: Unsterilized chitosan
1800
C=O
1700
1637
N-H
1600
1547
1500
Wavenumbers (cm-1) Fig. 2. FTIR spectroscopy of a: unsterilized and b: sterilized chitosan scaffolds.
Unsterilized Sterilized
Mw (kDa)
Mn (kDa)
PDI
230 ± 24 272 ± 19⁎
131 ± 7 158 ± 4⁎
1.76 1.72
Student's t-tests were performed to demonstrate the difference of molecular weights between unsterilized and sterilized chitosan scaffolds.⁎p b 0.05.
saturated steam autoclave (>120 °C) of chitosan powder induced color change (from white to brownish) and reduction of aqueous solubility [20].
3.4. Surface morphology The microstructure of the resultant chitosan scaffolds were shown in Fig. 3(A–D). As expected, the lyophilized scaffolds exhibited highly porous structure with interconnected pores and thin walls, both on the surface and within internal structure. The pore diameter on the surface was 70.3 ± 18.3 μm with porosity of 82.8 ± 6.3%, sterilization treatment had no significant effect on average pore diameter (72.0 ± 11.8 μm) and overall porosity (80.9 ± 6.9%) as shown in Fig. 3E. However, we found an obvious difference in pore size distribution before and after steam sterilization (Fig. 3 F). The unsterilized chitosan scaffolds possessed heterogeneous pore size distribution. For example, the proportions of the pores under 50 μm and over 100 μm were 12.2% and 7.6%, respectively, while in the sterilized scaffolds, the equivalent percentages were 4.0% and 2.8%, respectively, and over 65% of the pores laid in the range from 60 μm to 80 μm. The exact reason for variation in pore size distribution is still under investigation, while we anticipated that it might be attributed to the re-organization of microstructure during steam sterilization. Porosity plays a significant role in tissue engineering scaffolds in controlling cell behavior upon seeding. In addition, pore size regulates processes such as angiogenesis within scaffold [21,22]. Yannas et al. demonstrated that lyophilized scaffolds with an average pore diameter from 20 μm to 125 μm showed optimal morphological active in an in vivo skin regeneration model [21]. In this study, the fabricated scaffolds possessed an average pore diameter over 70 μm with a reasonable porosity (> 80%), indicating great potential for tissue engineering applications.
3.5. Mechanical characterization The compressive stress–strain curves of the resultant scaffolds were shown in Fig. 4A. The thermal-sterilized scaffolds exhibited linear stress–strain relationship with up to 52.9 ± 4.4% strain rate (yield point), and after this strain rate, the stress–strain curve showed a fluctuated plateau, which might be resulted from the plastic deformation upon compression. The compressive modulus of sterilized chitosan scaffolds was determined to be 109.8 ± 11.2 kPa and the yield compressive strength was 65.0 ± 12.4 kPa. The unsterilized scaffolds showed linear region over 60% strain rate with a compressive modulus of 45.5 ± 4.4 kPa, which was significantly lower than that of sterilized scaffolds (p b 0.001). These results indicate that steam sterilization (thermal-crosslinking) had a beneficial effect on improving the mechanical properties of chitosan scaffolds. The compressive modulus of sterilized scaffolds is within the range of moduli for various soft natural tissues such as human carotid artery (84 ± 22 kPa) and human spinal cord (89 kPa) [23,24]. Although steam-sterilization induced an earlier yield compared with unsterilized scaffolds, the yield point over 50% with yield strength of over 65 kPa are comparable with those of other materials that were designed for soft tissue engineering applications [25].
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
A
3783
B
100 µm
100 µm
C
D
100 µm
Pore size
F
100
Porosity
80
80
60
60
40
40
20
20
0
0
Porosity (%)
Pore size (µm)
100
Numerical frequency (%)
E
100 µm
40
Unsterilized Sterilized
30 20 10 0
Unsterilized
<50
Sterilized
50 -60 60 -70 70 -80 80 -90 90 -100 >100
Pore size range (µm)
Fig. 3. Phase contrast microscopy images of A and B: unsterilized and sterilized chitosan scaffolds on a view from surface, respectively; C and D: unsterilized and sterilized chitosan scaffolds on a view from cross section, respectively; E: pore size and porosity and F: pore size distributions of unsterilized and sterilized chitosan scaffolds.
B Compressive modulus (kPa)
Compressive stress (kPa)
A 80 Sterilized
60
40
20
Unsterilized 0
150
***
100
50
0 0
20
40
60
80
Unsterilized
Sterilized
Strain (%) Fig. 4. A: Compressive stress–strain curves and B: compressive moduli of unsterilized and sterilized chitosan scaffolds. Student's t-test was used to compare the compressive modulus of unsterilized and sterilized chitosan scaffolds (n = 3). ***p b 0.001.
3784
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
A 50
B
* 40
Unsterilized
40 30
ESR
Swelling ratio
50
Sterilized
20
30 20 10
10 0
0 0
5
10
15
20
25
Unsterilized
Sterilized
Immersion time (hours) Fig. 5. A: Time course swelling behavior and B: equilibrium swelling ratio of unsterilized and sterilized chitosan scaffolds. Student's t-test was used to compare the equilibrium swelling ratio of unsterilized and sterilized chitosan scaffolds (n = 3). *p b 0.01.
sterilized chitosan scaffolds was comparable with the other scaffolds for tissue engineering applications [27–31].
3.6. Swelling behavior In vitro and in vivo nutrient and waste exchange relies on swelling properties of scaffolds. As shown in Fig. 5A, unsterilized chitosan scaffolds showed burst water absorption (swelling ratio of 35.3 ± 3) within 1 h immersion in PBS, and the swelling ratio reached a plateau (35.8 ± 3) over 24 h. While the sterilized scaffolds exhibited a more sustained swelling behavior. The swelling ratio was increased from 12.7 ± 2 (1 h) to 22.0 ± 5 (4 h) and approaching an equilibrium swelling ratio (ESR) of 23.3 ± 3 upon 24 h immersion, which was significantly lower than that of unsterilized samples (p b 0.01). The decline in ESR is considered to be attributed to the intermolecular crosslinking upon steam sterilization [26], which in turn reduced the hydrophilicity of chitosan. However, the swelling property of the
3.7. In vitro biocompatibility test The feasibility of using the resultant chitosan scaffolds for further tissue engineering applications was assessed by using live/dead staining and MTS assay on cell-seeded scaffolds. The fluorescent staining images indicated that cells were able to adhere on the top surface of fabricated scaffold and also penetrate within the 3D structure of the scaffolds after 1 day post-seeding (Fig. 6A). The number of live cells (green dots) increased dramatically at day 7 (Fig. 6B), indicating cell proliferation within the fabricated scaffolds.
A
B
Day 1
Day 7
D
120
**
100
O.D. value (490 nm)
Percent cell viability (%)
C
80 60 40 20
1.5
1
0.5
0
0 Day 1
Day 7
0
1
4
7
Time (Day) Fig. 6. Fluorescent microscopy images of cell-seeded chitosan scaffolds after A: 1 day and B: 7 days post-seeding, scale bars represent 200 μm; C: cell viabilities of cell-seeded chitosan scaffolds at different time post-seeding; Student's t-test was used to compare the cell viability at 1 day and 7 days post-seeding (n = 9). **p b 0.005; and D: MTS analysis on cell-seeded chitosan scaffolds at different time post-seeding (n = 10).
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
The percent cell viability was assessed by cell counting. As shown in Fig. 6C, the resultant chitosan scaffolds demonstrated a cell viability of 74.8 ± 8% after 1 day of in vitro culture. A statistically significant (p b 0.005) increase in viability (85.4 ± 5%) was observed at day 7 post-seeding, and this can be explained by cell proliferation (increase in live cells) and wash-out of dead cells during medium exchange. MTS analysis also confirmed cellular proliferation on the fabricated scaffolds (Fig. 6D). The OD values at 490 nm dramatically increased at day 4 and day 7. These results are consistent with cell viability data as previously presented. The results of in vitro studies demonstrate that the steam autoclave treatment on the lyophilized chitosan scaffolds is efficient for sterilization, which is fundamental for tissue engineering application, and the resultant scaffold is a promising candidate for in vitro cell attachment and growth. A preliminary in vitro incubation test was conducted on unsterilized chitosan scaffolds, and the medium became cloudy indicating bacterial contamination after 3–4 days. 4. Conclusions Steam sterilization treatment was efficient to sterilize and simultaneously induce intermolecular crosslinking of chitosan lyophilized scaffolds. Compared with unsterilized scaffolds, the sterilized scaffolds possessed similar average pore diameter (>70 μm) and overall porosity (>80%) but more homogenous pore size distribution. The use of steam sterilization resulted in enhanced mechanical strength of the chitosan scaffolds. This process eliminates the use of toxic sterilants, and the sterilized scaffolds can be directly used for in vitro cell culture studies. The results of in vitro cell culture suggest that these scaffolds are promising for soft tissue engineering applications. A long-term in vitro cell culture will be performed to ensure further sterility and biocompatibility of the chitosan scaffolds prior to in vivo studies. Such biomaterials with simple and environmentally-benign processing are of great interest in tissue engineering applications. Acknowledgments The authors would like to acknowledge Dr. X. Zhong for assistance in FTIR and XRD characterizations of samples, and Dr. L. Tan for assistance in GPC analysis.
3785
References [1] M. Rinaudo, A. Domard, in: G. Skjak-Braek, T. Anthonsen, P. Sandford (Eds.), Elsevier, London, 1989, pp. 71–86. [2] C.K.S. Pillai, C.P. Sharma, Trends Biomater. Artif. Organs 22 (2009) 179–201. [3] J.K. Suh, H.W. Matthew, Biomaterials 21 (2000) 2589–2598. [4] B. Hoffmann, D. Seitz, A. Mencke, A. Kokott, G. Ziegler, J. Mater. Sci. Mater. Med. 20 (2009) 1495–1503. [5] Y. Huang, S. Onyeri, M. Siewe, A. Moshfeghian, S.V. Madihally, Biomaterials 26 (2005) 7616–7627. [6] C. Chu, D. Williams, J. Biomed. Mater. Res. 17 (1983) 1029–1040. [7] K. Hooper, J. Cox, J. Kohn, J. Appl. Polym. Sci. 63 (1997) 1499–1510. [8] S. Bruck, J. Biomed. Mater. Res. 5 (1971) 139–158. [9] K. Steenland, L. Stayner, J. Deddens, Occup. Environ. Med. 61 (2004) 2–7. [10] J. Bousquet, F. Michel, Clin. Rev. Allergy 9 (1991) 357–370. [11] G. Abraham, P. Frontini, T. Cuadrado, J. Appl. Polym. Sci. 65 (1997) 1193–1203. [12] W. Hugo, J. Appl. Bacteriol. 71 (1991) 9–18. [13] C. Ji, J. Shi, Mater. Lett. 72 (2012) 110–112. [14] C. Ji, A. Barrett, A. Poole-Warren Laura, N.R. Foster, F. Dehghani, Int. J. Pharm. 391 (2010) 187–196. [15] M. Tiera, X. Qiu, S. Bechaouch, Q. Shi, J. Fernandes, F. Winnik, Biomacromolecules 7 (2006) 3151–3156. [16] T. Romeo, Aust. Electron. Microsc. Newsl. 52 (1996) 16–18. [17] V. Karageorgiou, D. Kaplan, Biomaterials 26 (2005) 5474–5491. [18] Z. Osman, A.K. Arof, Electrochim. Acta 48 (2003) 993–999. [19] L. Ma, J. Biomater. Sci. Polym. Ed. 14 (2003) 861–874. [20] L. Lim, E. Khor, C. Ling, J. Biomed. Mater. Res. Appl. Biomater. 48 (1999) 111–116. [21] I.V. Yannas, E. Lee, D.P. Orgill, E.M. Skrabut, G.F. Murphy, Proc. Natl. Acad. Sci. U. S. A. 86 (1989) 933–937. [22] N. Annabi, J.W. Nichol, X. Zhong, C. Ji, S. Koshy, A. Khademhosseini, F. Dehghani, Tissue Eng. B 16 (2010) 371–383. [23] B. Amsden, Soft Matter 3 (2007) 1335–1348. [24] I. Levental, P. Georges, P. Janmey, Soft Matter 3 (2007) 299–306. [25] H. Shin, J.W. Nichol, A. Khademhosseini, Acta Biomater. 7 (2011) 106–114. [26] W.E. Hennink, C.F. van Nostrum, Adv. Drug Deliv. Rev. 54 (2002) 13–36. [27] L. Ma, C. Gao, Z. Mao, J. Zhou, J. Shen, X. Hu, C. Han, Biomaterials 24 (2003) 4833–4841. [28] K.Y. Chen, W.J. Liao, S.M. Kou, F.J. Tsai, Y.S. Chen, C.Y. Huang, C.H. Yao, Biomacromolecules 10 (2009) 1642–1649. [29] J. Shang, Z.Z. Shao, X. Chen, Polymer 49 (2008) 5520–5525. [30] S.S. Silva, A. Motta, M.T. Rodrigues, A.F.M. Pinheiro, M.E. Gomes, J.F. Mano, R.L. Reis, C. Migliaresi, Biomacromolecules 9 (2008) 2764–2774. [31] H. Tan, R. Chu Constance, K.A. Payne, G. Marra Kacey, Biomaterials 30 (2009) 2499–2506.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Thermal-crosslinked porous chitosan scaffolds for soft tissue engineering applications Chengdong Ji ⁎, Jeffrey Shi School of Chemical and Biomolecular Engineering, The University of Sydney, Sydney, NSW 2006, Australia
a r t i c l e
i n f o
Article history: Received 28 November 2012 Received in revised form 28 March 2013 Accepted 3 May 2013 Available online 13 May 2013 Keywords: Chitosan porous scaffolds Steam sterilization Thermal-crosslinking Soft tissue engineering
a b s t r a c t The aim of this study was to demonstrate the feasibility of using a steam autoclave process for sterilization and simultaneously thermal-crosslinking of lyophilized chitosan scaffolds. This process is of great interest in biomaterial development due to its simplicity and low toxicity. The steam autoclave process had no significant effect on the average pore diameter (~ 70 μm) and overall porosity (> 80%) of the resultant chitosan scaffolds, while the sterilized scaffolds possessed more homogenous pore size distribution. The sterilized chitosan scaffolds exhibited an enhanced compressive modulus (109.8 kPa) and comparable equilibrium swelling ratio (23.3). The resultant chitosan scaffolds could be used directly for in vitro cell culture without extra sterilization. The data of in vitro studies demonstrated that the scaffolds facilitated cell attachment and proliferation, indicating great potential for soft tissue engineering applications. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Chitosan, a linear polysaccharide, holds great potential in tissue engineering applications due to its readily availability, unique physicochemical properties, biocompatibility and biodegradability [1–3]. Chitosan has been fabricated in the form of lyophilized porous scaffolds to support cell attachment and proliferation, and eventually to facilitate new tissue formation after chitosan degradation [3]. Sterilization of the porous scaffolds remains an issue for tissue engineering applications. Chitosan scaffolds are usually sterilized by ethanol rinse or by UV exposure before laboratory-level in vitro studies [4,5], while for in vivo and clinical studies, ethylene oxide (EtO) exposure or gamma ray (Cobalt-60) irradiation are compulsory and more efficient [6–8]. However, such sterilants are highly hazardous not only to the host tissues and organs, but also to the operators [9,10]. The complete evacuation of these reagents (especially for porous materials) usually takes weeks-to-months, which is time-intensive [7,11]. Steam autoclave is a common sterilization method for thermalinert medical devices or buffered solution [12]. Compared with EtO exposure and gamma-ray irradiation, this process is simpler and more environmentally-friendly, while it is rarely applied for tissue engineering scaffolds given that most biomaterials have low thermal resistance. Our previous study reported that steam autoclave was used to sterilize and simultaneously induce chitosan-based hydrogel formation, and the resultant hydrogels are potent for drug delivery and tissue engineering applications [13]. In this study, we evaluated the feasibility of using steam autoclave to sterilize lyophilized chitosan scaffolds. The effect of sterilization on ⁎ Corresponding author at: Chemical Engineering Building J01, The University of Sydney, Sydney, NSW 2006 Australia. Tel.: + 61 2 9351 3411; fax: + 61 2 9351 2854. E-mail address: [email protected] (C. Ji). 0928-4931/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2013.05.010
the scaffolds' performances such as porosity, swelling behavior and mechanical properties were demonstrated. Cells were statically seeded on the sterilized scaffolds and in vitro culture was subsequently undertaken. The cell attachment and proliferation within chitosan scaffolds were investigated. 2. Materials and methods 2.1. Materials Chitosan (Catalog No. 448877, medium molecular weight and deacetylation degree of 75–85%), fluorescein diacetate (FDA), propidium iodide (PI), and Dulbecco's modified eagle medium (DMEM), were purchased from Sigma-Aldrich. MTS assay kit, fetal bovine serum (FBS) and penicillin–streptomycin solution (pen–strep) were purchased from Invitrogen. A 0.2 M acetic acid solution was prepared by diluting glacial acetic acid (Ajax Fine Chem) in MilliQ water. Chitosan solution (1.5 wt.%) was prepared by dissolving chitosan powder in 0.2 M acetic acid solution. Phosphate buffered saline (PBS, pH 7.2–7.4) was prepared by dissolving PBS tablets (Sigma) in MilliQ water. 2.2. Fabrication and sterilization of chitosan porous scaffolds Chitosan solution as prepared above was poured into a custommade mold and lyophilized for 30 h. The lyophilized scaffolds were collected and rinsed extensively by PBS to remove acidic residue, and eventually lyophilized again. The steam sterilization was held in a steam autoclave (Shenan Medical Devices) chamber for 30 min at 121 °C (Fig. 1A). Given the knowledge that chitosan possessed a melting temperature around 130–140 °C [14], the use of high temperature incubation 121 °C would not change the thermal behavior of chitosan.
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
A
Lyophilized porous scaffold
Sterilized porous scaffold Steam sterilization
B
3781
2.4.4. Mechanical strength Uniaxial compression tests were performed in a hydrated state (in PBS) at room temperature by using an Instron (Model 5543) with a 500 N load cell. Prior to mechanical testing, the scaffolds were immersed in PBS for at least 2 h; The compression (mm) and load (N) were collected at a crosshead speed of 30 μm/s until 60% compression was achieved. The compressive moduli were then calculated as the tangent slope of the stress–strain curves within linear regions (10–20% strain rate). 2.4.5. Swelling behavior The swelling behaviors of the scaffolds were evaluated at 37 °C in PBS. The initial dry weights were recorded (W0). After immersion in excessive PBS for different time intervals (1–24 h), the swollen chitosan scaffolds were weighted (Wt). The swelling ratio was subsequently calculated as (Wt − W0) / W0.
Fig. 1. A: Schematic diagram of chitosan scaffolds' fabrication and sterilization; and B: image of chitosan scaffolds before (left) and after (right) steam sterilization.
2.3. In vitro cell culture In vitro cell culture was performed on sterilized chitosan scaffolds under a biosafety cabinet. Each scaffold (~10 mm diameter and 3 mm thickness) was immersed in culture media (DMEM, 10% FBS, 1% pen– strep) at 37 °C for at least 2 h. The cells (human skin fibroblast cells GM3348) were then seeded (pipetted) onto the scaffold at a concentration of 1 × 10 5 cells/sample (for MTS assay, a cell concentration of 1 × 10 4 cells/sample was used). The cell-seeded scaffolds were kept in a CO2 incubator (Thermo Fisher HERAcell 150i) at 37 °C for further characterizations. The media was refreshed every two days. 2.4. Characterizations 2.4.1. Fourier transform infrared (FTIR) spectroscopy The molecular interactions of chitosan scaffolds upon steam sterilization were determined by using Fourier transform infrared (FTIR) spectroscopy (Varian 660-IR) with a resolution of 4 cm −1, averaging for 32 scans. 2.4.2. Gel permeation chromatography (GPC) The molecular weight of fresh and steam sterilized lyophilized chitosan scaffold was determined by using GPC as previously described [15]. In brief, each chitosan scaffold was dissolved in an acetic acid (0.3 M)/sodium acetate (0.2 M) buffer with a pH of 4.5 to achieve a concentration 0.5 mg/mL. Magnetically stirring and a moderate heating (~40 °C) were helpful for complete dissolution of sterilized chitosan. The chitosan solution was eventually filtered through a 0.45 μm membrane (Millipore) before GPC (Perkin Elmer Series 200) analysis. The molecular weight was calculated from a calibration curve using a series of poly ethylene glycol (PEG) with known molecular weights. 2.4.3. Surface morphology The porous structure on the surface of resultant scaffolds was visualized by using a phase contrast microscope (Leica). The porous structure on the cross section of resultant scaffolds was examined after cryo-fracture as previously described [16]. In brief, each fabricated scaffold was immersed in liquid nitrogen for 45 s, and cut by using a pre-cooled single edge razor blade. The fractured cross section was then mounted on glass slide for observation under a light microscope (Leica). The equivalent circle diameter (ECD) was then calculated by using ImageJ software. At least 300 pores were analyzed for each condition.
2.4.6. Overall porosity The overall porosities of resultant chitosan scaffolds were evaluated as previously described [17]. In brief, each scaffold was submerged in absolute ethanol of a known volume (V1), and a series of vacuumrelease cycles was performed to force the liquid into the pores of the scaffold. After these cycles, the volume of the liquid and liquidimpregnated scaffold was recorded as V2. When the liquid-impregnated scaffold was removed, the remaining liquid volume was recorded as V3. The overall porosity was given as [(V1 − V3) / (V2 − V3)] × 100% (n = 9). 2.4.7. Live/dead staining Cell proliferation in the resultant scaffolds was examined by live/dead staining. The cell-seeded scaffolds were stained with FDA and PI (both 1 μg/mL in PBS) for 5 min. The stained samples were assessed by using a fluorescence microscope (Eclipse E800, Nikon). Live cells were stained with fluorescent green due to intracellular esterase activity that de-acetylated FDA to a green florescent product. Dead cells were stained with fluorescent red as their compromised membranes were permeable to nucleic acid stain (PI). Percent cell viability values were calculated by counting the number of live (green) cells and the number of dead (red) cells on images (10× magnification). The values were obtained by dividing the number of live cells by the number of total cells (live cells + dead cells). A statistical significance level of 99.5% (p b 0.005) was considered to avoid potential human error in cell counting (n = 9). 2.4.8. In vitro proliferation assay In vitro cell proliferation was examined by using MTS assay. The cell-seeded scaffolds (1 × 104 cells/sample) were immersed in culture medium within 48 well-plates (n = 10). At different time intervals (3 h, 1 day, 4 and 7 days), the samples were rinsed by PBS three times; 250 μL fresh medium and 50 μL MTS was subsequently added into each well. The samples were then kept in a CO2 (5% CO2 and 95% humidity) incubator at 37 °C for 1 h; allowing MTS to react with metabolically active cells and subsequently result in water-soluble formazan product quantifiable by the optical density (OD) at 490 nm by using a microplate reader (Bio Rad 680). 2.5. Statistical analysis Each test was repeated three times except elsewhere mentioned. The statistical significance was determined at each condition by an independent Student's t-test for two groups of data using SPSS statistical software (PASW Statistics 18). Data are represented as mean ± standard deviation (SD). Confidence level of 95% (p b 0.05) was considered as statistically significant except elsewhere mentioned.
3782
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
3. Results and discussion
Table 1 Molecular weight of chitosan scaffolds.
3.1. General appearance White and rigid scaffolds were obtained after lyophilization, while the sterilized scaffolds exhibited about 10% shrinkage from 15.8 ± 0.2 mm to 14.4 ± 0.1 mm in diameter, and turned to be yellowish in color as shown in Fig. 1B (measured by using a digital caliper (J.B.S)). The sterilization process had negligible effect on the thickness of the scaffolds (~2.5 mm). 3.2. FTIR analysis Shrinkage in size and color change are common symbols of crosslinking of chitosan [5]. We thus used FTIR spectroscopy to characterize the effect of steam sterilization on chitosan scaffolds at a molecular level. As expected, unsterilized chitosan scaffold exhibited a characteristic peak at 1636 cm − 1, corresponding to the C_O stretch for amide I, which indicated that chitosan was not completely deacetylated, and another peak was detected at 1547 cm − 1 (Fig. 2a), which corresponds to either primary amines or amide II (N\H) [18]. After steam sterilization, peak shifts were observed on both characteristic peaks (from 1637 to 1651 cm−1, and 1547 to 1556 cm−1, respectively) indicating the molecular interaction, (Fig. 2b). The exact mechanism during steam sterilization is still under investigation, while we anticipated that possible mechanism is the molecular interaction between amino groups and either the primary hydroxyl groups or the carbonyl groups under high temperature [19]. As an evidence, the sterilized chitosan scaffolds turned to be yellowish (Fig. 1B), which is a common phenomenon of crosslinking based on amino groups [5]. 3.3. Molecular weight determination
R ela tiv e A b so rb a n ce (a .u .)
The molecular weight of chitosan scaffolds was listed in Table 1. Fresh lyophilized chitosan scaffolds showed a weight average molecular weight (Mw) of 230 ± 24 kDa and number average molecular weight (Mn) of 131 ± 7 kDa, with an average polydispersity index (PDI) of 1.76. After steam sterilization, the Mw, Mn and average PDI of the chitosan scaffolds was 272 ± 19 kDa, 158 ± 4 kDa and 1.72, respectively. Statistical analysis revealed that steam sterilization resulted in enhanced Mw and Mn, indicating the presence of crosslinking. Moreover, we surprisingly found that the sterilized chitosan scaffolds showed enhanced acid-resistance. In practice, the unsterilized chitosan scaffolds were completely dissolved in acidic solution (0.1% HCl) within 2 h, while the sterilized chitosan scaffolds could maintain their integrity in acidic environment for over one month. These data are in well agreement with previous studies which corroborated that
1556 1651
b: Sterilized chitosan
a: Unsterilized chitosan
1800
C=O
1700
1637
N-H
1600
1547
1500
Wavenumbers (cm-1) Fig. 2. FTIR spectroscopy of a: unsterilized and b: sterilized chitosan scaffolds.
Unsterilized Sterilized
Mw (kDa)
Mn (kDa)
PDI
230 ± 24 272 ± 19⁎
131 ± 7 158 ± 4⁎
1.76 1.72
Student's t-tests were performed to demonstrate the difference of molecular weights between unsterilized and sterilized chitosan scaffolds.⁎p b 0.05.
saturated steam autoclave (>120 °C) of chitosan powder induced color change (from white to brownish) and reduction of aqueous solubility [20].
3.4. Surface morphology The microstructure of the resultant chitosan scaffolds were shown in Fig. 3(A–D). As expected, the lyophilized scaffolds exhibited highly porous structure with interconnected pores and thin walls, both on the surface and within internal structure. The pore diameter on the surface was 70.3 ± 18.3 μm with porosity of 82.8 ± 6.3%, sterilization treatment had no significant effect on average pore diameter (72.0 ± 11.8 μm) and overall porosity (80.9 ± 6.9%) as shown in Fig. 3E. However, we found an obvious difference in pore size distribution before and after steam sterilization (Fig. 3 F). The unsterilized chitosan scaffolds possessed heterogeneous pore size distribution. For example, the proportions of the pores under 50 μm and over 100 μm were 12.2% and 7.6%, respectively, while in the sterilized scaffolds, the equivalent percentages were 4.0% and 2.8%, respectively, and over 65% of the pores laid in the range from 60 μm to 80 μm. The exact reason for variation in pore size distribution is still under investigation, while we anticipated that it might be attributed to the re-organization of microstructure during steam sterilization. Porosity plays a significant role in tissue engineering scaffolds in controlling cell behavior upon seeding. In addition, pore size regulates processes such as angiogenesis within scaffold [21,22]. Yannas et al. demonstrated that lyophilized scaffolds with an average pore diameter from 20 μm to 125 μm showed optimal morphological active in an in vivo skin regeneration model [21]. In this study, the fabricated scaffolds possessed an average pore diameter over 70 μm with a reasonable porosity (> 80%), indicating great potential for tissue engineering applications.
3.5. Mechanical characterization The compressive stress–strain curves of the resultant scaffolds were shown in Fig. 4A. The thermal-sterilized scaffolds exhibited linear stress–strain relationship with up to 52.9 ± 4.4% strain rate (yield point), and after this strain rate, the stress–strain curve showed a fluctuated plateau, which might be resulted from the plastic deformation upon compression. The compressive modulus of sterilized chitosan scaffolds was determined to be 109.8 ± 11.2 kPa and the yield compressive strength was 65.0 ± 12.4 kPa. The unsterilized scaffolds showed linear region over 60% strain rate with a compressive modulus of 45.5 ± 4.4 kPa, which was significantly lower than that of sterilized scaffolds (p b 0.001). These results indicate that steam sterilization (thermal-crosslinking) had a beneficial effect on improving the mechanical properties of chitosan scaffolds. The compressive modulus of sterilized scaffolds is within the range of moduli for various soft natural tissues such as human carotid artery (84 ± 22 kPa) and human spinal cord (89 kPa) [23,24]. Although steam-sterilization induced an earlier yield compared with unsterilized scaffolds, the yield point over 50% with yield strength of over 65 kPa are comparable with those of other materials that were designed for soft tissue engineering applications [25].
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
A
3783
B
100 µm
100 µm
C
D
100 µm
Pore size
F
100
Porosity
80
80
60
60
40
40
20
20
0
0
Porosity (%)
Pore size (µm)
100
Numerical frequency (%)
E
100 µm
40
Unsterilized Sterilized
30 20 10 0
Unsterilized
<50
Sterilized
50 -60 60 -70 70 -80 80 -90 90 -100 >100
Pore size range (µm)
Fig. 3. Phase contrast microscopy images of A and B: unsterilized and sterilized chitosan scaffolds on a view from surface, respectively; C and D: unsterilized and sterilized chitosan scaffolds on a view from cross section, respectively; E: pore size and porosity and F: pore size distributions of unsterilized and sterilized chitosan scaffolds.
B Compressive modulus (kPa)
Compressive stress (kPa)
A 80 Sterilized
60
40
20
Unsterilized 0
150
***
100
50
0 0
20
40
60
80
Unsterilized
Sterilized
Strain (%) Fig. 4. A: Compressive stress–strain curves and B: compressive moduli of unsterilized and sterilized chitosan scaffolds. Student's t-test was used to compare the compressive modulus of unsterilized and sterilized chitosan scaffolds (n = 3). ***p b 0.001.
3784
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
A 50
B
* 40
Unsterilized
40 30
ESR
Swelling ratio
50
Sterilized
20
30 20 10
10 0
0 0
5
10
15
20
25
Unsterilized
Sterilized
Immersion time (hours) Fig. 5. A: Time course swelling behavior and B: equilibrium swelling ratio of unsterilized and sterilized chitosan scaffolds. Student's t-test was used to compare the equilibrium swelling ratio of unsterilized and sterilized chitosan scaffolds (n = 3). *p b 0.01.
sterilized chitosan scaffolds was comparable with the other scaffolds for tissue engineering applications [27–31].
3.6. Swelling behavior In vitro and in vivo nutrient and waste exchange relies on swelling properties of scaffolds. As shown in Fig. 5A, unsterilized chitosan scaffolds showed burst water absorption (swelling ratio of 35.3 ± 3) within 1 h immersion in PBS, and the swelling ratio reached a plateau (35.8 ± 3) over 24 h. While the sterilized scaffolds exhibited a more sustained swelling behavior. The swelling ratio was increased from 12.7 ± 2 (1 h) to 22.0 ± 5 (4 h) and approaching an equilibrium swelling ratio (ESR) of 23.3 ± 3 upon 24 h immersion, which was significantly lower than that of unsterilized samples (p b 0.01). The decline in ESR is considered to be attributed to the intermolecular crosslinking upon steam sterilization [26], which in turn reduced the hydrophilicity of chitosan. However, the swelling property of the
3.7. In vitro biocompatibility test The feasibility of using the resultant chitosan scaffolds for further tissue engineering applications was assessed by using live/dead staining and MTS assay on cell-seeded scaffolds. The fluorescent staining images indicated that cells were able to adhere on the top surface of fabricated scaffold and also penetrate within the 3D structure of the scaffolds after 1 day post-seeding (Fig. 6A). The number of live cells (green dots) increased dramatically at day 7 (Fig. 6B), indicating cell proliferation within the fabricated scaffolds.
A
B
Day 1
Day 7
D
120
**
100
O.D. value (490 nm)
Percent cell viability (%)
C
80 60 40 20
1.5
1
0.5
0
0 Day 1
Day 7
0
1
4
7
Time (Day) Fig. 6. Fluorescent microscopy images of cell-seeded chitosan scaffolds after A: 1 day and B: 7 days post-seeding, scale bars represent 200 μm; C: cell viabilities of cell-seeded chitosan scaffolds at different time post-seeding; Student's t-test was used to compare the cell viability at 1 day and 7 days post-seeding (n = 9). **p b 0.005; and D: MTS analysis on cell-seeded chitosan scaffolds at different time post-seeding (n = 10).
C. Ji, J. Shi / Materials Science and Engineering C 33 (2013) 3780–3785
The percent cell viability was assessed by cell counting. As shown in Fig. 6C, the resultant chitosan scaffolds demonstrated a cell viability of 74.8 ± 8% after 1 day of in vitro culture. A statistically significant (p b 0.005) increase in viability (85.4 ± 5%) was observed at day 7 post-seeding, and this can be explained by cell proliferation (increase in live cells) and wash-out of dead cells during medium exchange. MTS analysis also confirmed cellular proliferation on the fabricated scaffolds (Fig. 6D). The OD values at 490 nm dramatically increased at day 4 and day 7. These results are consistent with cell viability data as previously presented. The results of in vitro studies demonstrate that the steam autoclave treatment on the lyophilized chitosan scaffolds is efficient for sterilization, which is fundamental for tissue engineering application, and the resultant scaffold is a promising candidate for in vitro cell attachment and growth. A preliminary in vitro incubation test was conducted on unsterilized chitosan scaffolds, and the medium became cloudy indicating bacterial contamination after 3–4 days. 4. Conclusions Steam sterilization treatment was efficient to sterilize and simultaneously induce intermolecular crosslinking of chitosan lyophilized scaffolds. Compared with unsterilized scaffolds, the sterilized scaffolds possessed similar average pore diameter (>70 μm) and overall porosity (>80%) but more homogenous pore size distribution. The use of steam sterilization resulted in enhanced mechanical strength of the chitosan scaffolds. This process eliminates the use of toxic sterilants, and the sterilized scaffolds can be directly used for in vitro cell culture studies. The results of in vitro cell culture suggest that these scaffolds are promising for soft tissue engineering applications. A long-term in vitro cell culture will be performed to ensure further sterility and biocompatibility of the chitosan scaffolds prior to in vivo studies. Such biomaterials with simple and environmentally-benign processing are of great interest in tissue engineering applications. Acknowledgments The authors would like to acknowledge Dr. X. Zhong for assistance in FTIR and XRD characterizations of samples, and Dr. L. Tan for assistance in GPC analysis.
3785
References [1] M. Rinaudo, A. Domard, in: G. Skjak-Braek, T. Anthonsen, P. Sandford (Eds.), Elsevier, London, 1989, pp. 71–86. [2] C.K.S. Pillai, C.P. Sharma, Trends Biomater. Artif. Organs 22 (2009) 179–201. [3] J.K. Suh, H.W. Matthew, Biomaterials 21 (2000) 2589–2598. [4] B. Hoffmann, D. Seitz, A. Mencke, A. Kokott, G. Ziegler, J. Mater. Sci. Mater. Med. 20 (2009) 1495–1503. [5] Y. Huang, S. Onyeri, M. Siewe, A. Moshfeghian, S.V. Madihally, Biomaterials 26 (2005) 7616–7627. [6] C. Chu, D. Williams, J. Biomed. Mater. Res. 17 (1983) 1029–1040. [7] K. Hooper, J. Cox, J. Kohn, J. Appl. Polym. Sci. 63 (1997) 1499–1510. [8] S. Bruck, J. Biomed. Mater. Res. 5 (1971) 139–158. [9] K. Steenland, L. Stayner, J. Deddens, Occup. Environ. Med. 61 (2004) 2–7. [10] J. Bousquet, F. Michel, Clin. Rev. Allergy 9 (1991) 357–370. [11] G. Abraham, P. Frontini, T. Cuadrado, J. Appl. Polym. Sci. 65 (1997) 1193–1203. [12] W. Hugo, J. Appl. Bacteriol. 71 (1991) 9–18. [13] C. Ji, J. Shi, Mater. Lett. 72 (2012) 110–112. [14] C. Ji, A. Barrett, A. Poole-Warren Laura, N.R. Foster, F. Dehghani, Int. J. Pharm. 391 (2010) 187–196. [15] M. Tiera, X. Qiu, S. Bechaouch, Q. Shi, J. Fernandes, F. Winnik, Biomacromolecules 7 (2006) 3151–3156. [16] T. Romeo, Aust. Electron. Microsc. Newsl. 52 (1996) 16–18. [17] V. Karageorgiou, D. Kaplan, Biomaterials 26 (2005) 5474–5491. [18] Z. Osman, A.K. Arof, Electrochim. Acta 48 (2003) 993–999. [19] L. Ma, J. Biomater. Sci. Polym. Ed. 14 (2003) 861–874. [20] L. Lim, E. Khor, C. Ling, J. Biomed. Mater. Res. Appl. Biomater. 48 (1999) 111–116. [21] I.V. Yannas, E. Lee, D.P. Orgill, E.M. Skrabut, G.F. Murphy, Proc. Natl. Acad. Sci. U. S. A. 86 (1989) 933–937. [22] N. Annabi, J.W. Nichol, X. Zhong, C. Ji, S. Koshy, A. Khademhosseini, F. Dehghani, Tissue Eng. B 16 (2010) 371–383. [23] B. Amsden, Soft Matter 3 (2007) 1335–1348. [24] I. Levental, P. Georges, P. Janmey, Soft Matter 3 (2007) 299–306. [25] H. Shin, J.W. Nichol, A. Khademhosseini, Acta Biomater. 7 (2011) 106–114. [26] W.E. Hennink, C.F. van Nostrum, Adv. Drug Deliv. Rev. 54 (2002) 13–36. [27] L. Ma, C. Gao, Z. Mao, J. Zhou, J. Shen, X. Hu, C. Han, Biomaterials 24 (2003) 4833–4841. [28] K.Y. Chen, W.J. Liao, S.M. Kou, F.J. Tsai, Y.S. Chen, C.Y. Huang, C.H. Yao, Biomacromolecules 10 (2009) 1642–1649. [29] J. Shang, Z.Z. Shao, X. Chen, Polymer 49 (2008) 5520–5525. [30] S.S. Silva, A. Motta, M.T. Rodrigues, A.F.M. Pinheiro, M.E. Gomes, J.F. Mano, R.L. Reis, C. Migliaresi, Biomacromolecules 9 (2008) 2764–2774. [31] H. Tan, R. Chu Constance, K.A. Payne, G. Marra Kacey, Biomaterials 30 (2009) 2499–2506.