Branched chitosans II: Effects of branching on degradation, protein adsorption and cell growth properties

Available online at www.sciencedirect.com

Acta Biomaterialia 5 (2009) 1575–1581 www.elsevier.com/locate/actabiomat

Branched chitosans II: Effects of branching on degradation, protein adsorption and cell growth properties Dinesh Aggarwal, Howard W.T. Matthew * Wayne State University, Chemical Engineering & Materials Science, 5050 Anthony Wayne Drive, Detroit, MI 48202, United States Received 8 April 2008; received in revised form 17 December 2008; accepted 4 January 2009 Available online 18 January 2009

Abstract The demand for biodegradable implant materials has fueled interest in chitosan as a biomaterial. In previous work, branched chitosans were synthesized and structurally characterized. In this study the biological properties of branched chitosans were explored. Branched chitosans were synthesized by grafting low molecular weight chitosan chains (1.6, 16 and 80 kDa) to high molecular weight (600 kDa) linear chitosans via reductive amination. Films of the branched materials were evaluated with regard to: lysozyme-mediated degradation; protein adsorption; cell adhesion and proliferation. Branched chitosan with a 1.6 kDa branch length exhibited higher degradation rates than either linear or higher branch length materials. Branched chitosans also exhibited reduced adsorption of bovine serum albumin that was more pronounced with thicker films. Branched chitosans supported proliferation of rat endothelial cells, but growth rates were significantly lower than on linear chitosan. The results of this study demonstrate that control of many aspects of chitosan’s physical and biological properties can be achieved by changes in molecular architecture. Ó 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Branched chitosan films; Biodegradation; Protein adsorption; Cell proliferation; Branched polymers

1. Introduction There is a growing need for advanced biomaterials that are biodegradable, can support tissue generation and have mechanical properties comparable to that of native tissue. Chitosan is a promising implantable material that derives potential from its gel-forming properties and cationic nature that allows it to form insoluble ionic complexes with a variety of anionic polymers. Primary amino groups in the chitosan structure can be easily derivatized with useful biological ligands, or modified with other entities to alter mechanical and degradation properties, as well as protein adsorption properties [1–3]. In addition, the material has been shown to promote wound healing, and exhibits a minimal foreign body response with accelerated angiogenesis [4–8]. Chitosan has been used in a variety of biomedical applications including: wound dressings [9–12], drug deliv*

Corresponding author. Tel.: +1 313 577 5238; fax: +1 313 578 5812. E-mail address: [email protected] (H.W.T. Matthew).

ery systems [13,14] and tissue engineered implants [15–19]. In most of these efforts, chitosan has been blended, crosslinked, or grafted with another molecule to bring about changes in properties as required for the specific application. Efforts have been made to enhance chitosan’s mechanical properties by incorporating polymers such as poly(ethylene glycol) [20], alginate [21] and silk fibroin [1]. However, addition of another polymer adds an extra level of complexity to the system and may result in adverse changes to other desirable properties. Given chitosan’s linear architecture and its semi-crystalline nature, manipulation of its molecular architecture offers another method of altering the material’s physical and biological properties. We previously reported on the synthesis and mechanical characterization of a family of branched chitosan materials [22]. In order to facilitate educated choices of these branched polymers for biomaterial applications, additional characterization is needed. Literature reports suggest that the low degradation rates of highly deacetylated chitosans are partly due to its semi-

1742-7061/$ - see front matter Ó 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.actbio.2009.01.003

1576

D. Aggarwal, H.W.T. Matthew / Acta Biomaterialia 5 (2009) 1575–1581

crystalline nature. Degradation kinetics (as measured by lysozyme-mediated cleavage) is mainly dependent upon two factors, namely molecular weight (MW) and degree of deacetylation (DD) [23–26]. However, changes in molecular architecture introduce material structure changes that can mimic changes in both MW and DD. Protein adsorption is one of the early events during the interaction of an implant material with a biological system. In addition to monomer chemistry, polymer architecture also can strongly influence chain organization and hence surface characteristics. Thus, inducing branching of linear chitosan may substantially alter some surface characteristics, and in doing so change aspects of protein adsorption properties that could substantially change cell and tissue responses to chitosan implants. Furthermore, the ability to effectively tune interactions between proteins and an implant material’s surface can be a useful tool in designing biomaterials specific to particular applications. In this study, branched chitosans were synthesized as previously reported [22], and various properties relevant to implant performance were evaluated. Specifically, vascular endothelial cell growth kinetics on cast films were characterized along with the adsorption of serum proteins, and lysozyme-mediated degradation kinetics.

of primary amine in the chitosan backbone polymer, assuming 100% reaction. 2.2. Biodegradation studies Degradation rates of cast chitosan films were evaluated using the dry weight loss method. Films of branched or linear chitosan were cast from solution by air-drying 25 ml of 1.5 wt.% solutions in 10 cm, non-adhesive, polystyrene petri dishes in a fume hood. Residual acetic acid was extracted from the dried membranes by washing with absolute ethanol 4–5 times. The membranes were then rehydrated through an ethanol series (80%, 60% and 40%) and finally equilibrated with water. Rectangular film specimens (1  2 cm) were cut and their initial wet weights were noted. Specimens were then immersed in PBS containing 30 mg ml1 lysozyme and incubated at 37 °C in a 5% CO2 atmosphere. Degraded samples were collected at time zero and every 4 days up to day 20. Collected samples were washed thoroughly with water, blotted on filter paper and their wet weights were documented. Dry weight and water content were determined by drying membranes in an oven at 105 °C overnight. Degradation was calculated as reduction in dry weight as a function of time. Statistical comparisons were done using Student t-test (paired two samples for means).

2. Materials and methods 2.3. Evaluation of protein adsorption 2.1. Synthesis of branched chitosan Branched chitosan materials were synthesized as previously described [22] using a two-step procedure. The first step involved synthesis of low MW chitosan polymers by nitrous acid depolymerization [27,28] and the second step involved grafting the low MW chains (i.e. the branches) to high MW chitosan backbones using reductive amination [29]. Chitosan (90% deacetylated, 600 kDa, Fluka) was dissolved in 1% acetic acid to form a 1.5 wt.% solution. Portions of this solution were depolymerized by addition of NaNO2 solution [27,28] to give solutions with average MW of 1.6, 16 and 80 kDa. The nitrous acid depolymerization procedure generates a reactive aldehyde group at the reducing end of each new low MW chitosan molecule. To prepare branched chitosans, 50 ml volumes of high MW chitosan solution and methanol were mixed. To this solution was added an appropriate volume of a low MW (depolymerized) chitosan solution. The volume of low MW solution added was determined by the stoichiometry of the desired branch density. Finally, 0.1 g of NaCNBH4 was added and the mixture was stirred at room temperature for 4–5 h. The branched chitosan product was purified by precipitation with 30% ammonia solution, followed by water washing, then redissolved in 1% acetic acid. Using this procedure a range of branched products varying in branch density and branch length were synthesized [22]. In this work, branch density is defined as the moles of branch molecule (i.e. low MW chitosan) added per mole

2.3.1. Quantification of total adsorbed protein Protein adsorption studies were conducted using cast films of linear and branched chitosans, with a branch density of 0.1 and branch lengths 1.6, 16 and 80 kDa. Three different film thicknesses were employed (denoted as 1, 2 and 3 in Fig. 3). Films were prepared by casting and airdrying 1.5 wt.% chitosan solutions using 0.4, 0.8 and 1.2 ml/well in standard 24-well tissue culture plates. The dried films were neutralized using 10% ammonia solution, washed extensively with water and equilibrated with PBS for 6 h at 4 °C. After aspiration of the PBS, the films were overlayed with 1.2 ml/well of a protein solutions for 36 h at 4 °C. The protein solutions used were bovine serum albumin (BSA, 2 mg ml1 in PBS) and fetal bovine serum (5% in PBS). After incubation, wells with adsorbed protein were washed with PBS four times, for 1 h each with constant rotary shaking in order to remove loosely bound proteins and proteins passively entrapped within the hydrogel films. Adsorbed proteins on the films were quantified in situ using a modified Lowry assay (DC Protein Assay kit from Bio-Rad Laboratories). 2.3.2. Quantification of adsorbed fibronectin and vitronectin The relative adsorption of fibronectin and vitronectin to chitosan films was evaluated after film exposure to 10% FBS in PBS. Films of intermediate thickness were prepared as described above, with the exception that 48-well plates were employed with 0.4 ml of chitosan

D. Aggarwal, H.W.T. Matthew / Acta Biomaterialia 5 (2009) 1575–1581

solution per well. After equilibration with PBS, the coated wells were overlayed with 0.5 ml of 10% FBS in PBS overnight at 4 °C. 10% FBS was used to increase the assay signal, since fibronectin and vitronectin comprise only a small fraction of the total protein. Films were then washed 4 times for 15 min each in PBS with constant shaking. Wells were blocked with 0.5 ml/well of blocking buffer (5% non-fat dry milk in PBS) for 3 h at room temperature. Blocked wells were then washed three times with PBS for 15 min each. Wells were then treated with a rabbit anti-bovine fibronectin antibody (dilution 1:5000, 150 ll per well) for 2 h at room temperature. After three PBS washes (15 min each), wells were treated with detecting antibodies. For fibronectin, a goat anti-rabbit IgG-alkaline phosphatase conjugate (dilution 1:15000, 150 ll per well) for 2 h at room temperature was used. For vitronectin the primary antibody used was a mouse monoclonal anti-vitronectin clone VIT-2 (dilution 1:5000) and the secondary antibody was anti-mouse IgM (l-chain specific) alkaline phosphatase conjugate (dilution 1:15000). Following secondary antibody incubation, the surfaces were washed 3 times and incubated with p-nitrophenyl phosphate substrate solution (0.6 ml/well) for approximately 30 min, at room temperature. Color development was stopped by addition of 150 ll of 3 N NaOH to each well. Solution aliquots (200 ll) were transferred to 96-well plates and absorbances were read at 405 nm using a multiwell plate spectrophotometer. Blank readings were obtained from wells containing no adsorbed protein, but treated with blocking solution and primary and secondary antibodies. Statistical analyses (Student t-tests, paired two samples for means) were performed on data obtained from the total protein, BSA, fibronectin and vitronectin adsorption studies to assess the effects of branching and membrane thickness. 2.4. Cell culture studies For cell interaction studies, surfaces were prepared by air-drying 0.8 ml/well of linear or branched chitosan solutions (1.5 wt.% in 1% acetic acid) on 24-well tissue culture plates. The dried surfaces were rehydrated through an ethanol series (100%, 80%, 60% and 40%) and finally equilibrated with distilled water. After aspiration of the water, wells were sterilized by filling with 70% ethanol and incubating overnight at room temperature. After aspirating ethanol, the wells were washed twice with sterile PBS and incubated with Dulbecco’s modified Eagle’s medium (DMEM) for 24 h at 37 °C in a 5% CO2 atmosphere. The medium was then removed and rat aortic endothelial cells were seeded at a density of 5000 cells cm2. The culture medium employed was DMEM supplemented with 5% FBS, 2 ng ml1 hFGF-2, 0.5 ng ml1 hEGF, 5 lg ml1 insulin, 50 lg ml1 gentamycin and 50 lg ml1 amphotericin-B. Medium changes were performed after 24 h and every 48 h

1577

thereafter. Phase contrast microscopy images of the attached cells were captured every 24 h for 5 days. 2.5. Assessment of cell growth The MTT-formazan conversion assay was used to assess the growth rate of cells on various chitosan surfaces. Wells with seeded endothelial cells were sacrificed for evaluation of growth kinetics after 3 and 6 days of culture. The medium was removed and the well surfaces were rinsed twice with PBS. 0.5 ml of MTT solution (thiazolyl blue tetrazolium bromide, 2 mg ml1 in Krebs–Ringer buffer) was then applied to each well. After 3 h incubation at 37 °C, the MTT solution was removed by aspiration. The accumulated formazan product was dissolved by addition of DMSO (0.5 ml/well) and mixing for 15 min. Aliquots (200 ll) were collected and the absorbance of the DMSOformazan solution was then read at 540 nm. Specific growth rates were then calculated by assuming exponential cell growth and the proportionality of formazan absorbance to cell number. Thus the following relationship may be applied: A ln½  ¼ l½t  t0  A0 Where A and A0 are the formazan absorbances at times t and t0 of culture, respectively, and l is the specific growth rate. 3. Results 3.1. Effects of branching on the degradation properties of chitosan films Grafting of chitosan branches produces large increases in the chitosan molecular weight. In enzyme studies, substrate molecular weight differences are often addressed by keeping the molar ratio of enzyme to substrate constant. However, in these studies the enzyme to chitosan mass ratio was kept constant. This was justified by the fact that lysozyme recognizes and cleaves trisaccharide motifs, which are randomly distributed at multiple locations within the polymer. As a result, maintenance of a constant enzyme to substrate molar ratio is best achieved by keeping the lysozyme to chitosan mass ratio (and hence the lysozyme to glucosamine molar ratio) constant. In addition, this approach further served to minimize the confounding effect of high polydispersity in typical chitosan samples. The branched chitosan material films having branch lengths of 16 and 80 kDa exhibited low rates of degradation similar to linear chitosan (Fig. 1). After 20 days, the dry weight reduction for the branched 16 and 80 kDa and linear chitosans was less than 2%. However, branched chitosan with the 1.6 kDa branch length demonstrated a significantly higher degradation rate, exhibiting approximately a 5% loss of dry weight after 20 days. The water content data (Fig. 2) showed that the shortest branch mate-

1578

D. Aggarwal, H.W.T. Matthew / Acta Biomaterialia 5 (2009) 1575–1581

reversed as additional degradation increased material permeability and reduced osmotic gradients. 3.2. Effects of branching on the protein adsorption properties of chitosan films

Fig. 1. Degradation of linear (control) and branched chitosans over 20 days in a lysozyme solution. The legend indicates molecular weight of the branches with branch density shown in parentheses. Data points are the means of at least 5 tested samples. Error bars represent the standard deviation.

Most of the branched chitosans when compared with the linear chitosan materials, demonstrated significantly lower (p < 0.05) levels of BSA adsorption as well as total serum (FBS) protein adsorption (Fig. 3, see Table 1 for statistical comparisons). In contrast, examination of fibronectin and vitronectin binding to a subset of surfaces (Fig. 4) showed significant differences (p < 0.05) only between the linear control and the 1.6 kDa branched chitosan (Table 2). Variations in film thickness produced no clear changes in adsorption of total serum protein on any surface. However, BSA adsorption on the branched chitosans showed significant decreases with increasing film thickness (Fig. 3, Table 1). 3.3. Cellular interaction studies

rial exhibited water content significantly higher than the control linear chitosan. Furthermore, mean water content appeared to decrease with increasing branch length. These results suggest that the higher degradation rates of the 1.6 kDa branched material was due in part to a looser molecular packing, lower levels of crystallinity and hence greater lysozyme access to cleavage sites in the polymer. It was also noted that water content at 4 days was slightly higher than at 20 days for all of the branched materials. This suggests that, similar to other hydrogel polymers, initial degradation of the material resulted in some swelling due to osmotic effects. This swelling may have subsequently

Fig. 2. Water content of linear (control) and branched chitosan materials during lysozyme-mediated degradation. Labels on the horizontal axis indicate the molecular weight of the branches with branch density shown in parentheses. Data points are the means of at least 5 tested samples. Error bars represent the standard deviation.

Endothelial cells attached and grew on all culture surfaces. However, cell spreading was lower on chitosan surfaces than on tissue culture plastic (Fig. 5). In addition, cells cultured on branched chitosan surfaces exhibited significantly lower growth rates compared to linear chitosan surfaces (Fig. 6). In fact, statistical analysis established that the specific growth rates were significantly different for all comparisons between linear chitosan, branched chitosans and tissue culture plastic, with the exception of the 16– 80 kDa branched chitosan comparison (Table 3).

Fig. 3. Effects of branching and film thickness on the adsorption of bovine albumin and serum proteins on linear (control) and branched chitosans. Branched chitosans with a branch density of 0.1 were evaluated. The numbers 1, 2 and 3 on the horizontal axis indicate the relative thicknesses of the linear and branched chitosan films. Branched chitosans are designated by the molecular weight of the branches (1.6, 16 and 80 kDa). Data are the means of at least 5 tested samples. Error bars represent the standard deviation.

D. Aggarwal, H.W.T. Matthew / Acta Biomaterialia 5 (2009) 1575–1581

1579

Table 1 Statistical significance chart for adsorption of albumin and total serum protein. Plastic

1.6-(0.1)

0.2% Bovine serum albumin Control 0.01434 0.0017 1.6-(0.1) 16-(0.1) 5% Fetal bovine serum Control 0.000002 1.6-(0.1) 16-(0.1)

0.00011

16-(0.1)

80-(0.1)

0.01237

0.00125

0.00062

0.0013

0.00052 0.0812

0.0258 0.0031

0.000432 0.0258

0.0607 0.0098 0.3880

0.02434 0.0879 0.7320

0.00009 0.0290 0.001706

0.0000

0.00278

0.0043

0.006837

0.00054 0.9230

0.00002 0.0932

0.000408 0.0873

0.0003 0.1909 0.2031

0.010836 0.6268 0.5804

0.0711 0.3756 0.9629

Note: The values in the table are the p values for statistical comparisons of differences between the column material and the row material using Student’s t-test (paired two sample for means). The number of samples was N = 6.

backbones. This was a departure from earlier approaches where branching was achieved by attaching non-chitosan saccharides or other entities to a linear chitosan chain [32–34]. The results presented clearly indicate that the branched chitosan having the shortest branch lengths (1.6 kDa) have significantly higher degradation rates compared to the linear and longer branch length chitosans (Fig. 1). This result is likely due to the crystallite-breaking effect of the short-branch architecture, illustrated by its higher water absorption (Fig. 2). This observation correlates well with the work by: (i) Kurita et al. on branched chitosan, where the introduction of short oligosaccharide branches significantly improved the hydrophilicity [34] and (ii) Tangpasuthadol et al. showing that lysozyme adsorption is enhanced on more hydrophilic chitosan surfaces [31]. Very low degradation rates for linear chitosan were expected due to the synergistic effects of high DD (>90%) and high crystallinity [24,26]. Furthermore, the control, linear chitosan (MW 600 kDa) is known to assume a ‘‘rigid-rod” conformation in solution because of its high charge density. As a result, the cast films posses a much higher level of crystallinity than would be seen with random-coil polymers of similar MW. We postulate that addition of ‘‘short” branches (1.6 kDa) interferes with packing of the rigid-rod backbone, thus producing a less crystalline, more enzyme-accessible material. In contrast, higher MW branches (16 or 80 kDa) are capable of forming crystallites with branches on adjacent molecules. These long-branched polymers can, therefore, be expected to be less degradable, since more of the polymer structure is protected within crystallites, compared to the shortbranch materials.

Fig. 4. Fibronectin and Vitronectin adsorption on linear (control) and branched chitosan films of intermediate thickness. Branched chitosans are designated by the molecular weight of the branches (1.6, 16 and 80 kDa). Branch density is shown in parentheses. Data are the means of at least 5 tested samples. Error bars represent the standard deviation.

4. Discussion Chitosan has been modified in a number of ways to tune the degradation kinetics. The various factors that have been shown to affect the degradation rate include molecular architecture [30], molecular weight (MW), crystallinity, water absorption [31] and degree of deacetylation (DD). Of these, DD and MW have been shown to be the most important [23,24,26]. In the work presented here, synthesis of branched chitosans was accomplished by grafting low MW chitosan chains to high MW chitosan

Table 2 Statistical significance chart for fibronectin (Fn) and vitronectin (Vn) adsorption. Plastic Fn

1.6 Vn

Fibronectin (Fn) and vitronectin (Vn) adsorption Control 0.00002 0.0004 1.6 16

16

80

Fn

Vn

Fn

Vn

Fn

Vn

0.0531

0.0005

0.2144 0.0510

0.1975 0.0031

0.1598 0.0798 0.6493

0.1758 0.02177 0.9663

Note: The values in the table are the p values for statistical comparisons of differences between the column material and the row material using Student’s t-test (paired two sample for means). The number of samples was N = 6.

1580

D. Aggarwal, H.W.T. Matthew / Acta Biomaterialia 5 (2009) 1575–1581

Fig. 5. Microscopic images of rat aortic endothelial cells on films of linear and branched chitosan. Images were captured after 96 h of culture. Branched chitosans are designated by the molecular weight of the branches (1.6, 16 and 80 kDa). Scale bars are 100 lm in length.

Protein adsorption is one of the early events that define the interaction of an implant material with a biological system. In this work we chose to examine the interaction of the chitosan materials with bovine albumin and bovine serum. Since binding is influenced by protein concentration, the serum was diluted to achieve total protein concentrations comparable to that of the BSA solution. The effects of chitosan film thickness were also examined, since film thickness during casting is known to influence polymer chain organization. Thinner cast films tend to be more crystalline than thicker films due to the phenomenon of surface-induced alignment [35]. The most notable observa-

tion of this study was the declining albumin adsorption with increasing film thickness. Since the thicker films tend to be more amorphous, the reduced binding suggests that BSA preferentially binds to more crystalline chitosan [36– 39]. This idea was supported by the fact that BSA binding was higher and exhibited no thickness effect on the linear chitosan films. The effects of film thickness are actually more pronounced than suggested in Fig. 3 since the absorbance data were not corrected for the increase in chitosan mass per well with increasing film thickness. Of the branched chitosans, only the polymer with 1.6 kDa branches exhibited a difference in binding of the adhesion-promoting proteins fibronectin and vitronectin (Fig. 4) compared to linear chitosan, and the difference was significant only for vitronectin. Examination of the endothelial cell growth rate on these surfaces showed that the proliferation rate on the 1.6 kDa branch chitosan was the lowest of the chitosan materials (Fig. 6). Furthermore, both of the materials with longer branches also exhibited growth rates significantly lower than the linear control. Since visual observation of the culture surfaces showed Table 3 Statistical significance chart for specific growth rate (lh1).

Control 1.6-(0.1) 16-(0.1) Fig. 6. Specific growth rates (l) for linear (control) and branched chitosan surfaces. Sample numbers indicate the branch lengths in kDa, with branch density in parentheses.

1.6-(0.1)

16-(0.1)

80-(0.1)

0.00008

0.017997 0.02627

0.005603 0.015507 0.987248

Note: The values in the table are the p values for statistical comparisons of differences between the column material and the row material using Student’s t-test (paired two sample for means). The number of samples was N = 3.

D. Aggarwal, H.W.T. Matthew / Acta Biomaterialia 5 (2009) 1575–1581

no obvious differences in cell number or cell spreading, these results suggest that determination of the specific growth rate was a more sensitive assay of cell response than purely morphological parameters. References [1] Gobin AS, Froude VE, Mathur AB. Structural and mechanical characteristics of silk fibroin and chitosan blend scaffolds for tissue regeneration. J Biomed Mater Res A 2005;74A(3):465–73. [2] Aiping Z, Tian C. Blood compatibility of surface-engineered poly(ethylene terephthalate) via o-carboxymethylchitosan. Colloids Surf B Biointerfaces 2006;50(2):120–5. [3] Channasanon S, Graisuwan W, Kiatkamjornwong S, Hoven VP. Alternating bioactivity of multilayer thin films assembled from charged derivatives of chitosan. J Colloid Interface Sci 2007;316(2):331–43. [4] Chupa JM, Foster AM, Sumner SR, Madihally SV, Matthew HWT. Vascular cell responses to polysaccharide materials: in vitro and in vivo evaluations. Biomaterials 2000;21(22):2315. [5] VandeVord PJ, Matthew HWT, DeSilva SP, Mayton L, Wu B, Wooley PH. Evaluation of the biocompatibility of a chitosan scaffold in mice. J Biomed Mater Res 2002;59(3):585–90. [6] Fujita M, Ishihara M, Shimizu M, Obara K, Nakamura S, Kanatani Y, et al. Therapeutic angiogenesis induced by controlled release of fibroblast growth factor-2 from injectable chitosan/non-anticoagulant heparin hydrogel in a rat hindlimb ischemia model. Wound Repair Regen 2007;15(1):58–65. [7] Ishihara M, Fujita M, Obara K, Hattori H, Nakamura S, Nambu M, et al. Controlled releases of FGF-2 and paclitaxel from chitosan hydrogels and their subsequent effects on wound repair, angiogenesis, and tumor growth. Curr Drug Deliv 2006;3(4):351–8. [8] Ishihara M, Obara K, Nakamura S, Fujita M, Masuoka K, Kanatani Y, et al. Chitosan hydrogel as a drug delivery carrier to control angiogenesis. J Artif Organs 2006;9(1):8–16. [9] Denkbas EB, Ozturk E, Ozdemir N, Kececi K, Agalar C. Norfloxaxin-loaded chitosan sponges as wound dressing material. J Biomater Appl 2004;18(4):291–303. [10] Knill CJ, Kennedy JF, Mistry J, Miraftab M, Smart G, Groocock MR, et al. Alginate fibers modified with unhydrolyzed and hydrolyzed chitosans for wound dressings. Carbohydrate Polymers 2004;55(1):65–76. [11] Azad AK, Sermsintham N, Chandrkrachang S, Stevens WF. The effect of chitosan membrane dressing on wound healing at the skin graft donor site. Adv Chitin Sci 2002;5:620–3. [12] Gustafson SB, Fulkerson P, Bildfell R, Aguilera L, Hazzard TM. Chitosan dressing provides hemostasis in swine femoral arterial injury model. Prehosp Emerg Care 2007;11(2):172–8. [13] Baran ET, Reis RL. Use of chemically modified chitosan and other natural-origin polymers in tissue engineering and drug delivery. In: Reis RL, San Roma´n J, editors. Biodegradable Syst Tissue Eng Regerat Med. United Kingdom: CRC Press; 2005. p. 325–35. [14] Di Martino A, Sittinger M, Risbud MV. Chitosan: a versatile biopolymer for orthopaedic tissue-engineering. Biomaterials 2005;26(30):5983–90. [15] Hsieh C-Y, Tsai S-P, Wang D-M, Chang Y-N, Hsieh H-J. Preparation of g-PGA/chitosan composite tissue engineering matrices. Biomaterials 2005;26(28):5617–23. [16] Jiang T, Pilane CM, Laurencin CT. Fabrication of novel porous chitosan matrices as scaffolds for bone tissue engineering. Materials Research Society Symposium Proceedings 2005;845:187–192, (Nanoscale Materials Science in Biology and Medicine). [17] Li Z, Li Y, Yang A, Peng X, Wang X, Xiang Z. Preparation and in vitro investigation of chitosan/nano-hydroxyapatite composite used as bone substitute materials. J Mater Sci Mater Med 2005;16(3):213–9. [18] Xia W, Liu W, Cui L, Liu Y, Zhong W, Liu D, et al. Tissue engineering of cartilage with the use of Chitosan–Gelatin complex scaffolds. J Biomed Mater Res B Appl Biomater 2004;71B(2):373–80.

1581

[19] Shalaby SW, DuBose JA, Shalaby M. Chitosan-based systems. In: Shalaby W, Burg KJL, editors. Absorbable Biodegradable Polym. United Kingdom: CRC Press; 2004. p. 77–89. [20] Alexeev VL, Kelberg EA, Evmenenko GA, Bronnikov SV. Improvement of the mechanical properties of chitosan films by the addition of poly(ethylene oxide). Polym Eng Sci 2000;40(5):1211–5. [21] Majima T, Funakosi T, Iwasaki N, Yamane S-T, Harada K, Nonaka S, et al. Alginate and chitosan polyion complex hybrid fibers for scaffolds in ligament and tendon tissue engineering. J Orthop Sci 2005;10(3):302–7. [22] Aggarwal D, Matthew HW. Branched chitosans: effects of branching parameters on rheological and mechanical properties. J Biomed Mater Res A 2007;82(1):201–12. [23] Tsaih ML, Chen RH. Effect of degree of deacetylation of chitosan on the kinetics of ultrasonic degradation of chitosan. J Appl Polym Sci 2003;90(13):3526–31. [24] Freier T, Koh HS, Kazazian K, Shoichet MS. Controlling cell adhesion and degradation of chitosan films by N-acetylation. Biomaterials 2005;26(29):5872–8. [25] Hwang KT, Kim JT, Jung ST, Cho GS, Park HJ. Properties of chitosan based biopolymer film with various degree of deacetylation and molecular weight. Adv Chitin Sci 2002;5:216–20. [26] Zhang H, Neau SH. In vitro degradation of chitosan by a commercial enzyme preparation: effect of molecular weight and degree of deacetylation. Biomaterials 2001;22(12):1653–8. [27] Allan GG, Peyron M. Molecular weight manipulation of chitosan. II: Prediction and Control of extent of depolymerization by nitrous acid. Carbohydrate Res 1995;277(2):273–82. [28] Allan GG, Peyron M. Molecular weight manipulation of chitosan I: kinetics of depolymerization by nitrous acid. Carbohydrate Res 1995;277(2):257–72. [29] Yalpani M, Hall LD. Some chemical and analytical aspects of polysaccharide modifications. 3. Formation of branched-chain, soluble chitosan derivatives. Macromolecules 1984;17(3):272–81. [30] Tasaka F, Ohya Y, Ouchi T. Synthesis of novel comb-type polylactide and its biodegradability. Macromolecules 2001;34(16):5494–500. [31] Tangpasuthadol V, Pongchaisirikul N, Hoven VP. Surface modification of chitosan films. Effects of hydrophobicity on protein adsorption. Carbohydrate Res 2003;338(9):937–42. [32] Yalpani M. New approaches to the synthesis of branched polysaccharide derivatives for structure/function studies. In: Crescenzi V, Dea ICM, editors. New Developments in Industrial Polysaccharides: Conf Proc. New York: Taylor and Francis; 1985. p. 295–306. [33] Kurita K, Shimada K, Nishiyama Y, Shimojoh M, Nishimura S. Nonnatural branched polysaccharides: synthesis and properties of chitin and chitosan having alpha-mannoside branches. Macromolecules 1998;31(15):4764–9. [34] Kurita K, Akao H, Yang J, Shimojoh M. Nonnatural branched polysaccharides: synthesis and properties of chitin and chitosan having disaccharide maltose branches. Biomacromolecules 2003;4(5):1264–8. [35] Childs MA, Matlock DD, Dorgan JR, Ohno TR. Surface morphology of poly(caprolactone)-b-poly(dimethylsiloxane)-b-poly(caprolactone) copolymers: effects on protein adsorption. Biomacromolecules 2001;2(2):526–37. [36] Kawamoto N, Mori H, Yui N, Terano M. Mechanistic aspects of blood-contacting properties of polypropylene surfaces – from the viewpoint of macromolecular entanglement and hydrophobic interaction via water molecules. J Biomater Sci Polym Ed 1998;9(6):543–59. [37] Tanaka H, Mori H, Nitta KH, Terano M, Yui N. Improved blood compatibility of drawn polyamide sheets. J Biomater Sci Polym Ed 1996;8(3):211–24. [38] Chen S, Yu F, Yu Q, He Y, Jiang S. Strong resistance of a thin crystalline layer of balanced charged groups to protein adsorption. Langmuir 2006;22(19):8186–91. [39] Alcantar NA, Aydil ES, Israelachvili JN. Polyethylene glycol-coated biocompatible surfaces. J Biomed Mater Res 2000;51(3):343–51.