Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications

European Polymer Journal xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

European Polymer Journal journal homepage: www.elsevier.com/locate/europolj

Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications G.-J. Graulus a, A. Mignon a, S. Van Vlierberghe a,⇑, H. Declercq b, K. Fehér c, M. Cornelissen b, J.C. Martins c, P. Dubruel a,⇑ a Polymer Chemistry and Biomaterials Research Group, Department of Organic and Macromolecular Chemistry, Ghent University, Krijgslaan 281 (building S4), B-9000 Ghent, Belgium b Department of Basic Medical Sciences, Ghent University, De Pintelaan 185 6B3, B-9000 Ghent, Belgium c NMR and Structure Analysis Unit, Department of Organic and Macromolecular Chemistry, Ghent University, Krijgslaan 281 (building S4), B-9000 Ghent, Belgium

a r t i c l e

i n f o

Article history: Received 4 March 2015 Received in revised form 25 June 2015 Accepted 30 June 2015 Available online xxxx Keywords: Modified gelatin Hydrogel Alginate Cross-linking Tissue engineering

a b s t r a c t When it comes to failing or injured tissues and organs, patients often end up on waiting lists for tissue or even organ transplantation negatively affecting the patient’s quality of life. The multidisciplinary research field of tissue engineering may offer more innovative ways to replace or ideally regenerate failing tissues and organs. A widely used material in this research field is gelatin because of its biocompatibility and interesting hydrogel forming properties. However, at body temperature gelatin’s mechanical properties are greatly reduced due to the dissolution of collagen-like triple helices. With the aim to obtain materials that retain their mechanical properties at body temperature, we propose to combine sodium alginate and methacrylamide-modified gelatin (Gel-MOD) in the form of a graft copolymer to obtain a material that closely resembles the extracellular matrix. The obtained materials can be cross-linked via three distinct pathways including cation mediated, temperature mediated or via covalent bond formation after UV irradiation in the presence of a photo-initiator. The current contribution covers the synthesis of the above mentioned alginate-graft-gelatin copolymers and the characterization of the resulting hydrogels. The materials developed are highly hydrophilic, showing high gel fractions and satisfactory mechanical properties. Moreover the attained storage moduli were tunable by divalent cation addition (72–275% increase at 21 °C, 42–405% increase at 40 °C). One formulation was found to outperform Gel-MOD in terms of mechanical properties at 40 °C, thus indicating the proposed strategy can be used to improve the mechanical properties of gelatin-based hydrogels. Moreover, in vitro biocompatibility assays indicated that cell adhesion and proliferation improves with increasing gelatin content. The present paper illustrates that the developed triple cross-linkable materials are suitable cell carriers, promising to be applied for biomedical purposes. Ó 2015 Published by Elsevier Ltd.

1. Introduction Regenerative medicine is a growing interdisciplinary research field that may offer new treatments for patients with failing tissues due to disease or trauma. Regenerative medicine encompasses various strategies to repair or replace cells, tissues ⇑ Corresponding authors. E-mail addresses: [email protected] (S. Van Vlierberghe), [email protected] (P. Dubruel). http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033 0014-3057/Ó 2015 Published by Elsevier Ltd.

Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

2

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

and organs to restore the impaired function of the affected body parts. Since regenerative medicine aims to bring the patient back to normal health, it can be distinguished from organ transplantation as the inherent need for immunosuppressant medication cannot be considered normal health [1,2]. Tissue engineering has been defined by Langer and Vacanti as an interdisciplinary field that applies principles from engineering and life sciences toward the development of biological substitutes that restore, maintain or improve tissue function [3]. Since cells are always embedded in a three-dimensional polymer network, the extracellular matrix (ECM), one method applies polymer constructs onto which cells can be seeded. These scaffolds should be designed in such a way that they mimic the ECM’s mechanical and biological properties. A material often applied in this regard is gelatin, a single stranded protein obtained from collagen by hydrolytic degradation [4]. Gelatin has already been used in a large variety of applications including food industry applications, pharmaceutical formulations, photographic and other technical products. Gelatin is an interesting biopolymer for tissue engineering applications, since gelatin solutions readily form gel-like structures upon cooling. This gelation is driven by hydrogen bonding and van der Waals interactions, resulting in the aggregation of certain gelatin domains into collagen-like triple helices separated by random coil peptide residues [5–7]. However, these junction zones, being physical in nature, melt at temperatures around 30 °C [4]. This implies that chemical cross-linking is required to avoid dissolution of the scaffolds at body temperature. Despite applying chemically cross-linked gelatin-based hydrogels, the mechanical properties remain lower at body temperature. With the aim to overcome this limitation and develop scaffolds with mechanical properties that outperform modified gelatin (Gel-MOD), we propose the application of cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications. Applying these systems also implies a better mimic of the aqueous environment cells naturally reside in, since the ECM is composed of both polysaccharides and proteins. To the best of our knowledge, the herein proposed triple cross-linkable alginate-graft-gelatin copolymers have not been applied to date. Alginates are anionic polysaccharides derived from brown algae and consist of D-mannuronic acid (i.e. the M block) and L-guluronic

acid (i.e. the G block) units arranged in an irregular, block wise pattern of varying proportions of GG, MM and MG blocks [8]. Mannuronic acid forms b (1 ? 4) linkages, while guluronic acid forms a (1 ? 4) bonds resulting in steric hindrance around the carboxylic acid groups. As a result, M blocks form linear domains while G blocks introduce folded regions responsible for a more rigid structure [8,9]. Alginate was selected in this study since it is an interesting biopolymer for biomedical applications because of its ability to rapidly form gels upon addition of multivalent ions [4,10]. Alginate’s gelation mechanism is, however, hard to control and does not result in a uniform structure [11]. The formation of an ionotropic hydrogel starting from alginate upon Ca2+ addition mainly involves the GG blocks along the polymer backbone [9]. Grant et al. proposed a model in which the GG blocks were thought to combine with the Ca2+ ions forming structures resembling an egg-box [12]. In addition to its potential to form hydrogels in the presence of multivalent ions, alginate is mucoadhesive, biocompatible and non-immunogenic making it very suitable for tissue engineering applications [13]. In order to develop scaffold materials of which the mechanical properties remain unaffected at body temperature, gelatin type B was first modified with methacrylic anhydride in order to introduce methacrylamide pendant groups. Next, a fraction of alginate carboxylic acids was converted into reactive esters using carbodiimide chemistry in combination with N-Hydroxysuccinimide (NHS). Finally, unreacted amines present in the Gel-MOD were used for alginate conjugation. The obtained materials could be cross-linked via three distinct strategies including cation mediated, temperature mediated or via covalent bond formation after UV irradiation in the presence of a photo-initiator. The hydrogel materials were characterized via High Resolution-Magic Angle Spinning (HR-MAS) 1H NMR spectroscopy, rheology, and swelling experiments. Finally, the in vitro cell viability and – proliferation behaviour was assessed.

2. Experimental details 2.1. Materials Gelatin type B, isolated from bovine bone via an alkaline process, was obtained from Rousselot (Ghent, Belgium). Sodium alginate was purchased from Sigma–Aldrich Fine Chemicals (Bornem, Belgium). Potassium phosphate dibasic (K2HPO4), sodium phosphate monobasic (NaH2PO4), sodium borate (Na2B4O5), sodium chloride (NaCl) and potassium chloride (KCl) used to prepare the buffer solutions, were obtained from Acros Organics (Geel, Belgium). Sodium azide was purchased from Avocado Research Chemicals Ltd. (Karlsruhe, Germany). Methacrylic anhydride, 2-mercaptoethanol, n-butylamine, 1-ethy l-3-(3-dimethylamino-propyl)carbodiimide (EDC), N-Hydroxysuccinimide (NHS), o-phthalaldehyde (OPA), hydrochloric acid (HCl) and sodium hydroxide (NaOH) and calcium chloride (CaCl2) were acquired from Sigma–Aldrich (Bornem, Belgium) and were used as received. Ethanol was obtained from Chem-Lab (Zedelgem, Belgium). The photoinitiator Irgacure 2959 was received from BASF (Antwerp, Belgium). Dialysis membranes (MWCO 12,000–14,000 Da) were purchased from Polylab (Antwerp, Belgium). All NMR spectra were recorded in deuterated solvents obtained from Euriso-top (Saint-Aubin Cedex, France). Cell seeding experiments were conducted in cell culture media obtained from Gibco Invitrogen. Live/dead staining was performed using propidium iodide and calcein AM which were obtained from Sigma–Aldrich (Bornem, Belgium) and Anaspec (Fremont, USA) respectively. MTT (3-(4,5-dimethyldiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) obtained from Merck (Nottingham, UK), was applied to determine cell viability. Dulbecco’s Modified Eagle Medium (DMEM) Glutamax medium, foetal bovine serum (FBS), penicillin–streptomycin (10 U/ml–10 mg/ml) and sodium-pyruvate were acquired from Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

3

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

Gibco Invitrogen (Ghent, Belgium). Deionised water was obtained via Millipore water purification systems (Merck, Overijse, Belgium) giving respectively Milli-RO (27.2 lS/cm) and Milli-Q (54.9 nS/cm) water. 2.2. Synthesis procedures 2.2.1. Methacrylation of gelatin B Gelatin type B (100 g, 38.5 mmol of amines) was dissolved in 1 L of a 0.1 M phosphate buffer (pH 7.8) at 40 °C under mechanical stirring. Next, 2.5 equivalents methacrylic anhydride (14.34 ml; 96.25 mmol) were added after which the solution was stirred for 1 h. Following this, 1 L of Milli-Q was added to the reaction mixture and the resulting solution was dialyzed in Milli-RO water (MWCO 12,000–14,000) for 24 h at 40 °C. The solution was then transferred to Petri dishes and frozen at 20 °C enabling the subsequent removal of water via lyophilisation. 2.2.2. Synthesis of alginate-Gel-MOD (X%) Sodium alginate was dissolved overnight in Milli-Q (3% (w/v)) at 40 °C. The pH was adjusted to 3.4 using a 0.4 M HCl solution followed by dilution to 2% (w/v). Subsequently, EDC and NHS were added and the reaction was stirred for 5 h at 40 °C. Then, Gel-MOD in 200 ml of Milli-Q (40 °C) was added to the alginate solution, after which the mixture was stirred overnight at 40 °C. The solution was purified via dialysis (MWCO 12,000–14,000 Da) in Milli-RO (40 °C) for 24 h. The obtained solution was frozen at 20 °C and subsequently lyophilised. The procedure was repeated for various degrees of activation. The applied amounts of the above-mentioned products are summarized in Table 1. EDC and NHS were applied to activate 2%, 6% or 8% of the carboxylic acids present in alginate. 2.3. Methods 2.3.1. Preparation of hydrogel films Hydrogel films were prepared by dissolving 1 g of the various modified biopolymers in 10 ml Milli-Q at 40 °C. Next, Irgacure 2959 solution (0.8 w/v%) was added to the solution to obtain an initiator concentration of 2 or 4 mol% relative to the number of present methacrylamide double bonds. The solutions were degassed and injected between two parallel glass plates separated by a 1 mm thick silicone spacer. To enable easy hydrogel removal, both glass plates were first covered with a sheet of TeflonÒ release foil. UV-A irradiation was applied for 1 h to realize chemical cross-linking. Samples were placed between two Long Wave UV lamps model VL-400L (Vilber Lourmat, Marne La Vallée, France), with an intensity of 10 mW/cm2 and a wavelength range of 250–450 nm. The intensity was determined via an ACCU-CAL-50 UV dosimeter (Dymax Corporation, Wiesbaden, Germany). The distance between the samples and both lamps was approximately 4 cm. In order to study the effect of additional Ca2+ cross-linking a number of samples prepared using the above mentioned procedure, were incubated overnight in a 10% (w/v) CaCl2 solution to introduce additional cross-links between the alginate domains. 2.3.2. Freeze-drying Lyophilisation of the polymer derivatives occurred by means of a Christ freeze-dryer alpha 2-4-LSC, typically during 24 h. 2.3.3. Size exclusion chromatography The molecular weight for the various materials was determined by size exclusion chromatography on a set-up composed of a Waters 610 fluid unit and a Waters 600 control unit equipped with a Waters 410 RI detector (40 °C). The two Shodex SB806MHQ columns placed in series eluting 0.1 M phosphate buffer (pH 7.8) were kept at 80 °C using an external heating unit and were calibrated using dextran standards (103–4 ⁄ 106 Da range). Samples were dissolved in 0.1 M phosphate buffer (pH 7.8) at a concentration of 1 mg/ml.

Table 1 Overview of the chemicals used for the synthesis of Alg-Gel with various activation degrees. Product

Sodium alginate (g) Sodium alginate (mmola) EDC (mmol) NHS (mmol) Gel-MOD (mmolb) Mass % gelatin a b

% activated carboxylic acids in alginate (theoretical) 2%

4%

8%

4 20.19 0.40 0.40 0.46 70.55%

2 10.95 0.40 0.40 0.46 79.09%

1 5.05 0.40 0.40 0.46 88.32%

mmol available carboxylic acids in sodium alginate. mmol available amines in Gel-MOD.

Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

4

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

To assess the degradation of Gel-MOD at low pH values. 200 mg of gel-MOD was dissolved in 2 ml of acetate buffer (0.1 M, pH 3.4). The solution was stirred for 24 h at 40 °C in order to simulate the conditions experienced in reaction. Next, a drop of the resulting solution was added to 2 ml of phosphate buffer (0.1 M, pH 7.8) and analyzed as discussed above. 2.3.4. Fourier Transform Infrared spectroscopy Fourier Transform Infrared spectroscopy was performed on a Bio-Rad FT-IR spectrometer FTS 575C operating in Attenuated Total Reflection mode. The recorded spectra were analyzed using WIN-IR Pro software. 2.3.5. Thermogravimetric analysis Thermogravimetric analyses were performed on a TA Instruments Q50 device. Samples were placed on a platinum sample pan and subjected to a stepwise incremental heating program. After equilibrating the sample at 45 °C the sample was heated at a heating rate of 10 °C/min. When the mass loss transcended a value of 1% per minute the temperature was kept isothermal until the mass loss had decreased to a value below 0.1%/min. This procedure was repeated until the end temperature of 1000 °C. Results were analyzed using the TA Instruments Universal Analysis 2000 software. 2.3.6. Determination of amines present in gelatin The amount of amines present in Gel-MOD was determined spectroscopically. For this purpose, o-phthalaldehyde (OPA, 20 mg) was dissolved in 10 ml ethanol. Next, the mixture was diluted to 50 ml using Milli-Q water. A second stock solution, containing 25 ll 2-mercaptoethanol in 50 ml borate buffer (0.1 M, pH 10) was prepared. Samples were dissolved in Milli-Q (40 °C) at a concentration of 0.025 g/ml. To 50 ll of the heated sample solutions, 950 ll Milli-Q water, 1500 ll mercaptoethanol solution and 500 ll of the OPA solution were added subsequently, followed by vigorously mixing. Finally, the absorbance at 335 nm was measured compared to a blank (i.e. a mixture using 50 ll water instead of sample solution) on a UVIKON XL UV–VIS spectrometer running UVIKONXL software (Bio-Tek instruments) equipped with heated cuvette holders (40 °C). All measurements were performed in triplicate. Analogous measurements were performed with n-butylamine (0.002–0.01 M) standards to obtain a calibration curve. 2.3.7. HR-MAS 1H NMR spectroscopy HR-MAS analysis of the hydrogel films developed was performed on a Bruker Avance II 700 spectrometer (700.13 MHz) using a HR-MAS probe equipped with a 1H, 13C, 119Sn and gradient channel. The spinning rate was set to 6 kHz. HR-MAS samples were prepared by placing a small amount of the freeze-dried hydrogel inside a 4 mm zirconium oxide MAS rotor (50 ll). D2O (30 ll) was added to the rotor, allowing the samples to swell. The samples were homogenized by manual stirring prior to analysis. A TeflonÒ coated cap was used to close the rotor. 2.3.8. Rheology The mechanical properties of the gels were evaluated using a rheometer type physica MCR-301 using Rheoplus v3.40 software (Anton Paar, Sint-Martens-Latem, Belgium). Oscillation experiments were performed using two parallel plates (Ø = 25 mm). The spectra were obtained using a frequency of 1 Hz and a gap of approximately 1 mm. To ensure a close contact between the sample and both plates, the upper plate was lowered with increments of 0.01 mm, if required, until a normal force of 0.2–0.8 N was observed. The % strain was selected by determining the visco-elastic range. The storage modulus (G0 ) was measured by means of oscillation rheology. The oscillation can be described using the following equation:

cðtÞ ¼ c0 sinðxtÞ

ð1Þ

with c0 the amplitude and x the frequency. The shear stress corresponding to this deformation is then given by a phase shifted sine function:

sðtÞ ¼ s0 sinðxt þ dÞ

ð2Þ

with d the phase shift between the preset and the resulting curve. By comparing the preset oscillation (from Eq. (1)), with the resulting shear stress (see Eq. (2)), the storage modulus (G0 ) was derived:

G0 ðxÞ ¼

s0 cosðdÞ c0

ð3Þ

2.3.9. Determination of gel fraction of hydrogel films After weighing the lyophilised hydrogel films (Ø = 8 mm, h = 1 mm), the samples were submerged in Milli-Q at 37 °C for 24 h to dissolve the physically cross-linked polymer chains from the networks. After incubation, the excess water was removed by pressing the samples gently on a piece of paper. Next the samples were frozen at 20 °C to allow the removal of the remaining water via lyophilisation. The changes in mass were recorded for all samples. From these masses the gel fraction, i.e. the percentage of material that is chemically incorporated in the three-dimensional network, can be derived using the following formula:

Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

Gel fraction ð%Þ ¼ ðW df =W 0d Þ  100%

5

ð4Þ

with W 0d the dry weight before swelling and W df the dry weight after swelling. The gel fraction of the various hydrogel films was determined in triplicate by incubating the samples Milli-Q at 37 °C. The results were reported as mean values with corresponding standard deviations. 2.3.10. Swelling properties of hydrogel films developed PBS (0.1 M) was prepared by dissolving 80 g NaCl, 2 g KCl, 12.583 g Na2HPO4 and 4.025 g KH2PO4 in Milli-Q (1 l solution). The stock solution was diluted to 0.01 M before use and the pH was adjusted to 7.4 using NaOH (0.4 M) or HCl (0.4 M). Sodium azide was finally added to prevent bacterial growth during the swelling experiments. The hydrogel films (Ø = 8 mm, h = 1 mm) were submerged in PBS at 37 °C and the changes in mass were recorded as a function of time. The swelling degree (%) was derived from these results using the following equation:

Swelling degree ð%Þ ¼

ðW ts  W 0d Þ 1   100% q W 0d

ð5Þ

with W ts the mass of the sample at time t, W 0d the dry mass of the sample prior to immersion in PBS and q the mass density of the applied PBS buffer. The density of the medium is included in Eq. (5) to facilitate the comparison of the swelling behaviour in different media. In the present study the density of the incubation medium was determined to be 1.0039 g/ml. All experiments were performed in triplicate and the results were reported as mean values with corresponding standard deviations. 2.3.11. Cell seeding on hydrogel samples Human foreskin fibroblasts (HFF-1 cells) were cultured in DMEM Glutamax medium supplemented with 10% foetal bovine serum, penicillin–streptomycin (10 U/ml–10 mg/ml) and 100 mM sodium-pyruvate. Cells were cultured at 37 °C in a humidified atmosphere containing 5% CO2. Hydrogel films were placed in 24-well plates and sterilized by ethylene oxide-cold cycle (Maria Middelares Hospital, Ghent, Belgium). After the sterilization cycle of 4 days, cell culture experiments were started. HFF-1 cells were seeded at a density of 40,000 cells/0.5 ml culture medium/hydrogel in the cell proliferation assay. In the cell adhesion assay, HFF-1 cells were seeded at a density of 100,000 cells/0.5 ml culture medium/hydrogel. Cell adhesion and proliferation were evaluated after 1 and 7 days. Cells cultured on tissue culture polystyrene (TCPS) were taken as a positive control. 2.3.12. Fluorescence microscopy To visualize cell attachment and distribution on the hydrogels, the cells were evaluated by fluorescence microscopy after live/dead staining. After rinsing the hydrogels with PBS, the supernatant was replaced by 1 ml PBS solution supplemented with 2 ll (1 mg/ml) calcein AM (Anaspec, USA) and 2 ll (1 mg/ml) propidium iodide (Sigma). Cultures were incubated for 10 min at room temperature, washed twice with PBS solution and evaluated by fluorescence microscopy (Type U-RFL-T, CellTM software, Olympus, Aartselaar, Belgium). The fluorescence of calcein-AM and propidium iodide was monitored at 460–495 nm excitation/550 nm emission and 545–580 nm excitation/610 nm emission respectively. Evaluations were done 1 and 7 days post-seeding. 2.3.13. MTT-viability The colorimetric MTT assay, using a 3-(4,5-dimethyldiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Merck Promega) was performed to quantify cell viability and proliferation on the hydrogel films. The tetrazolium component is reduced in living cells by mitochondrial dehydrogenase enzymes into a water-soluble purple formazan product, which can be solubilized by the addition of lysis buffer and measured using spectrophotometry. The cell culture medium was replaced by 0.5 ml (0.5 mg/ml) MTT reagent and cells were incubated for 4 h at 37 °C. The MTT reagent was removed and replaced by 0.5 ml lysis buffer (0.1% Triton X-100 in isopropanol/0.04 N HCl) for 30 min. The dissolved formazan solution (200 ll) was transferred into a 96-well plate and measured spectrophotometrically at 580 nm (Universal microplate reader EL800, Biotek Instruments). Triplicate measurements were performed at the same time-points as the microscopic evaluation. 2.3.14. Statistical analysis Statistical analysis was performed using the student t-test. Two values were considered significantly different when p < 0.05. 3. Results and discussion 3.1. Material synthesis Since its introduction, Gel-MOD has gained increasing interest as a hydrogel precursor [14]. This can be attributed to the fact that upon photopolymerisation of the methacrylamide pendant groups the resulting gel no longer dissolves at body temperature. This opened up new application areas including controlled (drug) release and tissue engineering. However, despite Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

6

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

the chemical cross-linking, the material’s mechanical properties are negatively affected when applied at body temperature due to the melting of the collagen like triple helices. To address this issue, Gel-MOD has already been blended with other polymers, both synthetic as well as natural to improve their mechanical performance [15]. One commonly applied strategy aims at the formation of (semi-)interpenetrating networks (IPNs) by cross-linking the gelatin independently from the other polymer [16,17]. Alternatively, chemically oxidized alginate has already been applied of which the aldehyde functionalities readily bind to the free amines of gelatin [18–20]. The current work can be situated at the interface of both strategies. Since the amines present in gelatin are first substituted to a great extent by methacrylamides, the reaction with activated alginate does not result in an insoluble network. The latter enables to process the material using the same approach as one would apply for IPNs, while both biopolymers are linked together as is the case in the activated alginate strategy. In addition, the presence of UV cross-linkable methacrylamides enables the materials’ shape to be fixed after being processed into tailor-made implants via rapid prototyping techniques [21]. Gelatin B was modified with methacrylamide pendant groups according to a procedure developed previously by Van Den Bulcke et al. [14]. Gelatin B contains pendant amine groups under the form of lysine, hydroxylysine, ornithine, histidine and arginine amino acids. At pH 7.8, arginine will be protonated, while histidine, ornithine, lysine and hydroxylysine can be partially present under their deprotonated forms enabling them to act as nucleophiles. Histidine, however, will not result in the formation of stable products and can thus be omitted from the further calculations [22]. One hundred grams of gelatin B contains 0.0385 moles of free amines that can be converted to methacrylamide moieties by means of nucleophilic attack at a carbonyl functionality of methacrylic anhydride. The methacrylic acid formed during the reaction is removed via dialysis (MWCO: 12,000–14,000 Da). Finally, the Gel-MOD solution was frozen at 20 °C and subsequently lyophilised. Sodium alginate was dissolved in Milli-Q water (40 °C). EDC and NHS were subsequently added to convert a fraction of the carboxylic acids present in alginate into reactive succinimidyl esters. Next, a Gel-MOD solution (40 °C) was added enabling the reactive esters to undergo a nucleophilic attack by the unmodified amines present in Gel-MOD. The reaction mixture was stirred overnight and purified via dialysis (MWCO: 12,000–14,000 Da) in Milli-RO at 40 °C for 24 h. To evaluate the effect of the amount of alginate present, different materials were synthesized by varying the amount of activated carboxylic acids (2–8%) in alginate. The Gel-MOD added was adjusted accordingly, to 1.14 equivalents of free amines relative to the reactive esters. In order to effectively graft gelatin on alginate, the number of available amines present in gelatin should be kept low to reduce the probability of forming loops (i.e. gelatin forming multiple bonds with a single alginate chain) around and cross-links between alginate strands. Therefore Gel-MOD with a high degree of methacrylation (>80%) is required. 3.2. Characterization of hydrogel precursors To prove the proposed concept of gelatin grafts on an alginate core, the materials were subjected to various techniques in order to distinguish them from alginate-gelatin blends generally reported on in literature. When applying the EDC/NHS-mediated coupling reaction, a new amide bond is formed for each graft. This amide bond, however, cannot be distinguished in 1H NMR or FT-IR spectroscopy due to considerable overlap with the gelatin signals. The NMR spectra of the reported materials are presented in the supplementary information (Figs. S1–6). The determination of the grafting efficiency was therefore attempted via indirect methods. Thermogravimetric analyses showed no significant difference between the various Alg-Gel materials (see supplementary information, Fig. S7). Only for the unmodified alginate, a lower degradation temperature was observed, while in all other samples gelatin’s degradation profile dominated the thermograms. Because of the limitations of the above mentioned techniques, a spectroscopic technique using OPA was applied to determine the amount of reacted amines during the grafting procedure. This method is particularly powerful as OPA readily and selectively reacts with primary amines while the procedure is easily implemented [23]. Further details can be found in the supplementary information (Table S2). In short, for the analyzed sample masses, it is possible to calculate the mass of Gel-MOD via the mass percentage of Gel-MOD added. By measuring the amount of amines in Gel-MOD via the OPA method, it is possible to calculate the number of amines present in the various Alg-Gel samples prior to grafting. By analyzing the samples subsequently, with the same OPA protocol a considerable drop in absorbance was observed. Via an external calibration curve, the amount of unreacted amines was calculated. Finally by dividing the reacted amines with the original amount of available amines, reaction efficiencies of 71.60–82.11% were obtained. Considering the grafting efficiencies calculated in the previous section, an impact on the molecular weight was anticipated. Alg-Gel samples were thus subjected to size exclusion chromatography to determine their molecular weight and molecular weight distribution. Alginate, gelatin B and Gel-MOD were subjected to the same analysis. The results are listed in Table 2 and show both an increase in molecular weight and dispersity when gelatin is grafted on alginate. The SEC traces can be found in the supplementary information (Fig. S9). The results show that the molecular weight obtained for the various Alg-Gel samples is lower than the Gel-MOD used in their synthesis. This observation can be rationalized by the acidic alginate solution required in the grafting procedure. During the reaction and purification some hydrolysis can be expected giving rise to lower molecular weights than were expected. This hypothesis was confirmed by repeating the SEC analysis after Gel-MOD had been incubated in acetate buffer for 24 h (supporting information, Table S3). This experiment showed a considerable decrease in molecular weight accompanied by an increase in dispersity. This degradation process, being uncontrolled, prevented the calculation of the average number of gelatin grafts per alginate chain. However, the combination of Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

7

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx Table 2 SEC results for the various materials. Alg-Gel samples show an increase in molecular weight and dispersity when grafted with gelatin. Material

Mn (Da) Mw (Da) Ð

Alginate

Alg-Gel 2%

Alg-Gel 4%

Alg-Gel 8%

Gel-MOD

Gel-B

42,710 52,630 1.23

56,560 95,570 1.69

64,740 138,690 2.14

62,600 119,890 1.92

81,680 235,500 2.88

72,350 136,220 1.88

both the increased molecular weight and the significant drop in the amount of amines present, demonstrate that the grafting procedure was successful. 3.3. Characterization of hydrogel films 3.3.1. Determination of gel fraction in cross-linked hydrogel films For all materials satisfactory gel fractions (>80%) were obtained (Fig. 1). The results show that both cross-linking methodologies (i.e. 2 or 4 mol% Irgacure 2959) result in stable hydrogel networks, although no significant difference could be observed between both methods. However, from gel fraction experiments it is not possible to accurately determine the cross-linking efficiency as some uncross-linked chains can be physically entrapped in the polymer matrix. Therefore these results were supplemented with HR-MAS NMR experiments which will be discussed in the next section. 3.3.2. HR-MAS 1H NMR spectroscopy The UV curing earlier described results in water-insoluble polymer networks, which cannot be characterized using conventional 1H NMR spectroscopy due to considerable line broadening in the dry (solid) state. These broad signals can be attributed to dipolar interactions, chemical shift anisotropy (CSA) and magnetic susceptibility effects [24–26]. The line width can, however, be decreased considerably when hydrogel samples are swollen in a deuterated solvent (e.g. deuterium oxide, D2O). As a result, the polymer becomes solvated and gains segmental motion, enabling the chains to more closely resemble the conditions they would experience in solution. The latter results in the removal of dipolar interactions and CSA effects resulting in a decrease in line width to a few kHz. Additionally, the line broadening due to magnetic susceptibility effects can be reduced by rapidly rotating the sample at an angle of 54.7° relative to the static magnetic field [25]. At this magic angle h, the contribution of the Hamiltonian disappears, effectively removing the line broadening effects. Magic angle spinning NMR spectroscopy (MAS-NMR) thus allows the recording of NMR spectra of solids, with more narrow signals. The applied spinning rates are generally in the order of a few kilohertz. The combination of the above-mentioned effects thus enables the application of 1H NMR spectroscopy on semi-solid samples. In the present study, HR-MAS 1H NMR spectroscopy has been applied to determine the cross-linking degree of the polymer materials developed. By comparing the integration of the signal corresponding to the methacrylamide double bonds present in Gel-MOD with the integration of a signal that remains chemically inert during the cross-linking procedure, the cross-linking efficiency can be determined [27]. The methacrylamide functionality corresponds with signals at 5.45 and 5.70 ppm. The signal at 0.95 ppm, which corresponds to the methyl groups of valine (Val), leucine (Leu) and isoleucine (Ile), was selected as reference. The signal corresponding with the chemically inert amino acids should integrate for 18 protons. One hundred grams of gelatin type B contain

Fig. 1. Overview of gel fractions (%) obtained for the various hydrogel films prepared (white: UV-A irradiated for 60 min in the presence of 2 mol% IrgacureÒ 2959; light grey: UV-A irradiated for 60 min in the presence of 4 mol% IrgacureÒ 2959). The graph shows that the applied cross-linking procedure is effective for obtaining water insoluble hydrogel networks. All materials are significantly different (p < 0.05) from each other with the exception of the pairs indicated by ‘X’.

Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

8

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

0.023 mol Val, 0.026 mol Leu and 0.015 mol Ile. If the integration of the NMR signal at 0.95 ppm is set to 18, it thus corresponds to 0.3836 mol of these methyl protons per 100 g gelatin B. In addition, 100 g gelatin B contains 0.0385 moles of free amines that can be converted to methacrylamide pendant groups. These figures together with the spectroscopic data enable the calculation of the degree of methacrylation (DM) [27,28].

DMð%Þ ¼

ðI5:7 ppm þI5:45 ppm Þ=2 0:0385 mol ðNH2 Þ=100 g gelatin I0:9 ppm 0:3836 mol ðrefÞ=100 g gelatin

 100%

ð6Þ

Since the double bonds present in the samples, are consumed during cross-linking, it is possible to calculate the degree of cross-linking (DC) by comparing the intensity of the signals at 5.7 and 5.45 ppm before (i) and after cross-linking (c). Since different samples have to be compared, it is required to normalize these intensities using the inert signal. The degree of cross-linking is given by the following formula [27]:

DCð%Þ ¼

ðIi5:7 ppm =Ii0:9 ppm Þ  ðIc5:7 ppm =Ic0:9 ppm Þ Ii5:7 ppm =Ii0:9 ppm

 100%

ð7Þ

It should be noted that all spectra were recorded with suppression of the water signal. Interestingly, in some samples a broad water signal was still observed. This was an indication of the presence of a different aqueous phase. Water molecules that strongly interact with the polymer chains will experience a different chemical environment and a slow exchange with the bulk water. As a result, both aqueous phases display a different resonance frequency. Since this phenomenon was only observed in samples with relatively high alginate contents, it was expected that the alginate domains were responsible for the observed signal. This was confirmed by recording a 1H NMR spectrum of sodium alginate, where the same signal (4.84 ppm) was observed. The results of the HR-MAS 1H NMR spectroscopy measurements based on Eqs. (6) and (7) are summarized in Fig. 2. Since the materials depicted in Fig. 2A were prepared using the same Gel-MOD, the relative amount of methacrylamide functionalities is very similar. After addition of IrgacureÒ 2959 (2 mol%) and subsequent UV-A irradiation, the number of double bonds within the polymer networks show a very sharp drop leading to values ranging from 6.09% to 21.03%. The latter clearly indicates the effectiveness of our cross-linking protocol. Increasing the amount of initiator (4 mol%) results in a further drop in the number of methacrylamide functionalities. HR-MAS 1H NMR spectroscopy is thus an excellent tool to investigate the chemical structure of (cross-linked) hydrogel materials. The degrees of methacrylation remaining after chemically cross-linking (Fig. 2A) were subsequently applied as starting value to calculate the cross-linking degrees using Eq. (7) (see Fig. 2B). The results are in close correspondence with the data shown in Fig. 2A. These results together with the results shown in Fig. 1, confirm the successful cross-linking in the presence of the photo-initiator Irgacure.

3.3.3. Rheological evaluation of hydrogel films The obtained G0 values for the various hydrogels developed are depicted in Fig. 3 on a logarithmic scale. G0 is a measure for the elastic behaviour of a sample. Given the elastic behaviour is correlated with the number of cross-links in the material higher values are associated with tougher gels.

Fig. 2. HR-MAS analyses of the studied materials. (A) Degree of methacrylation of the materials developed as a function of the amount of photo-initiator (white: uncross-linked material; light grey: UV-A irradiated for 60 min in the presence of 2 mol% IrgacureÒ 2959; dark grey: UV-A irradiated for 60 min in the presence of 4 mol% IrgacureÒ 2959). (B) Degree of cross-linking of the various hydrogel films as a function of photo-initiator concentration (white: UV-A irradiated for 60 min in the presence of 2 mol% IrgacureÒ 2959; grey: UV-A irradiated for 60 min in the presence of 4 mol% IrgacureÒ 2959). Interestingly, already at 2 mol% photo-initiator present, satisfactory cross-linking efficiencies were obtained. By increasing the initiator concentration, the results could be further improved.

Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

9

Fig. 3. Graphs showing the effect of initiator concentration and temperature on the mechanical properties: white bar: UV-A irradiated for 60 min in the presence of 2 mol% IrgacureÒ 2959; light grey bar UV-A irradiated for 60 min in the presence of 4 mol% IrgacureÒ 2959 dark grey bar UV-A irradiated for 60 min in the presence of 4 mol% IrgacureÒ 2959, analyzed at 40 °C.

Increasing the amount of photo-initiator results in a slight improvement of the mechanical properties. Despite the high cross-linking efficiency derived from HR-MAS 1H NMR spectroscopy, Alg-Gel samples show lower storage moduli compared to Gel-MOD. It has been indicated in literature that different polysaccharides including alginate, can act as radical scavengers [29]. This could lead to shorter poly(methacrylate) bridges and thus a lower amount of gelatin chains being linked together, effectively lowering the mechanical properties. Additionally, Fig. 3 shows the temperature influence on G0 . At 40 °C, the collagen-like triple helices are fully dissolved (i.e. removal of the physical cross-linking). As a result, G0 is only determined by the chemical cross-linking. For Gel-MOD, a temperature increase clearly reduces G’ values as already communicated earlier by Van Den Bulcke et al. and Billiet et al. [14,21]. This trend was observed as well for the Alg-Gel samples, although less pronounced (note the logarithmic scale in Fig. 3). Next, the incubation of the hydrogels in a 10% (w/v) CaCl2 solution after UV cross-linking clearly improves the mechanical properties (i.e. a 42–405% increase in G0 ) of the various Alg-Gel hydrogels (Fig. 4). Upon Ca2+ addition, the alginate domains present can form egg-box like structures which will reinforce the hydrogel [12]. Interestingly, the covalent grafting of gelatin onto the alginate backbone did not disturb the physical cross-linking potential of alginate using Ca2+. Since the applied modification strategy did not discriminate between guluronic or mannuronic carboxylic acids, the increase in G0 did not show a clear correlation with the alginate content of the various samples. As mentioned earlier the GG blocks are mainly responsible for the formation of the egg-box like structures [9,12]. The impact of this additional physical cross-linking on the mechanical properties is visualized in Fig. 4. These results clearly show that the additional calcium cross-linking counters the effect of an increase in temperature: in the presence of calcium ions, the storage modulus of Alg-gel 8% at 40 °C is in fact higher (42,500 Pa) than the storage modulus of Gel-MOD at the same temperature (36,250 Pa) illustrating the potential of this additional cross-linking.

Fig. 4. Graph showing the positive effect of divalent cations on the mechanical properties of Alg-Gel materials. At 21 °C, the addition of Ca2+ ions leads to an improvement of the mechanical properties for all Alg-Gel materials, while Gel-MOD shows a lower storage modulus under these conditions (white bar: 21 °C, no Ca2+; dark grey bar: 21 °C, additional Ca2+). At 40 °C, the same effects were observed (light grey bar: 40 °C, no Ca2+; black bar: 40 °C, additional Ca2+).

Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

10

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

Finally, rheology was performed on non-cross-linked blends of Gel-MOD and alginate showing the same chemical compositions. The result of these experiments is given in the supplementary information (Fig. S10). The mechanical properties of the blends were found to be considerably lower than was the case for the graft copolymers. This result further strengthens our statement that graft-copolymers were indeed synthesized. The blends showed a trend of improving mechanical properties obtained with increasing gelatin content, as anticipated since the gelatin methacrylamides are consumed to form cross-links. A higher amount of Gel-MOD thus also implies a higher amount of cross-links present within the network. 3.3.4. Swelling experiments Since the developed materials are to be applied in tissue engineering, the equilibrium swelling degree was determined. Hydrogel materials able to retain large quantities of water will indeed better resemble the aqueous surroundings cells reside in [4]. The various hydrogel materials were therefore submerged in PBS buffer at 37 °C. The mass change was monitored as a function of time. The results are shown in Fig. 5. The results revealed that the developed materials are capable of absorbing large quantities of water. No significant difference was observed between both cross-linking methodologies (2 or 4 mol% Irgacure 2959). Alg-Gel 4% cross-linked in the presence of 2 mol% Irgacure 2959 exhibited a significantly (p < 0.05) higher swelling degree than the other samples. This result is in agreement with the lower storage moduli obtained for Alg-Gel 4% as a lower amount of cross-links opposing water uptake, also results in poorer mechanical properties. This effect is further enhanced due to the lower gel fractions obtained for Alg-Gel 4%. Among the samples cross-linked with a photoinitiator concentration of 4 mol%, swelling degrees were again correlated with the obtained storage moduli, although no significant differences (p > 0.05) were revealed for this cross-linking methodology. Samples that were incubated in a 10% (w/v) CaCl2 solution and subsequently freeze-dried showed a strong decrease in equilibrium swelling as compared to the samples that lacked this additional cross-linking. This observation can, however, not completely be attributed to the increased cross-link density, since Ca2+ ions were able to leach out into the PBS buffer thus perturbing the results. 3.4. In vitro cell viability, adhesion and proliferation 3.4.1. Fluorescence microscopy analysis Fluorescence microscopy after live/dead staining was applied to visualize the cells on the hydrogel films (Fig. 6). One day after seeding, spherical non-attached cells were observed on the Alg-Gel hydrogels with 2% (Fig. 6A1 and B1) and 4% activation (Fig. 6A2 and B2). In contrast, cells were well spread with an elongated morphology on the Alg-Gel hydrogels for the 8% activation hydrogel (Fig. 6A3 and B3) comparable with the Gel-MOD hydrogel (Fig. 6A4 and B4). On all hydrogels, the cells were viable as detected by the live/dead staining, independent of the cell morphology (round or elongated). After 7 days, Alg-Gel hydrogels with 4% (Fig. 6C2 and D2) and 8% (Fig. 6C3 and D3) activation and Gel-MOD hydrogels (Fig. 6C4 and D4) were covered with a confluent cell layer. The cells were well spread and elongated with a morphology identical to the control (Fig. 6C5 and D5). On the Alg-Gel with 2% activation (Fig. 6C1 and D1), the cells were not confluent and predominantly cells with a rounded as well as super-stretched morphology, typical for non-adherent cells, were present. The cells were viable as shown by the live/dead staining. For most biomaterial applications cell spreading on hydrogels is

Fig. 5. Obtained degree of swelling after immersing the samples in PBS for 24 h (white bars: 2 mol% of Irgacure 2959; light grey bars: 4 mol% Irgacure 2959; dark grey bars: 2 mol% of Irgacure 2959 and subsequent incubation in 10% (w/v) CaCl2 solution; black bars: 4 mol% of Irgacure 2959 and subsequent incubation in 10% (w/v) CaCl2 solution. All experiments were performed in triplicate and the results are shown as mean values with corresponding standard deviations. Significant differences (p < 0.05) are marked by means of an asterisk (⁄).

Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

11

Fig. 6. Photographs of HFF-1 cells cultured on hydrogel films (Alg-Gel 2% (row 1); Alg-Gel 4% (row 2); Alg-Gel 8% (row 3); Gel-MOD (row 4) and TCPS (row 5) after 1 (Column A and B) and 7 (Column C and D) days). Scale bars represent 500 lm (columns A and C) or 200 lm (columns B and D). Fluorescent micrographs after live/dead staining.

preferred as this is an indication of satisfactory cell attachment, which will more closely resemble native tissue. In the present study a minimal amount of 4% activation was required to obtain good cell attachment after 7 days.

3.4.2. MTT-analysis The amount of viable cells attached to the hydrogel samples as determined by the MTT-assay is shown in Figs. 7 and S11 (supplementary information) for respectively the cell proliferation and cell adhesion assays. Data are shown for the analysis after 1 and 7 days of incubation, relative to the tissue culture plate control. After 1 day, the amount of viable cells on all the Alg-Gel hydrogels (i.e. percentage of viable cells in the range of 18.81–21.18%) was comparable to the Gel-MOD hydrogel (23.98% of viable cells) and no significant difference could be observed (p > 0.05). However, a significant difference (p < 0.01) was observed relative to the control (42.01%). After 7 days, the amount of viable and attached cells on Alg-Gel hydrogels (with 4% and 8% activation) (49.75–45.23%) was comparable with the Gel-MOD hydrogel (55.41%) and no significant difference (p > 0.05) could be observed. In contrast, the number of viable cells on the Alg-Gel hydrogel with 2% activation was only 25% relative to the control (100%) which also demonstrated a significant difference (p < 0.05) with the other co-polymers. With a higher alginate content, a lower amount of cells attaching to the hydrogel surface have been observed. This can easily be rationalized by the lack of cell-binding motives in alginate [30]. Alg-gel 4% and 8% which contained more Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

12

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx

Fig. 7. Percentage of viable HFF-1 cells cultured on hydrogel films (Alg-Gel 2%, 4% and 8% activation; Gel-MOD) for 1 (white bars) and 7 days (grey bars) relative to the control (tissue culture polystyrene) The amount of viable cells was quantified with the MTT-assay. Significant differences (p < 0.05) are indicated with an asterisk (⁄).

gelatin, were shown to closely resemble Gel-MOD’s biological properties while outperforming the mechanical properties of native Gel-MOD (see higher). These results illustrate the need for the inclusion of gelatin in alginate-based materials, as alginate itself is not bio-interactive [4,31,32]. 4. Conclusions With the aim to compensate for the reduction in gelatin’s mechanical properties at body temperature, a variety of alginate-graft-gelatin copolymers were successfully synthesized. These hydrogel building blocks were successfully processed into hydrogel films and subsequently characterized. Depending on the exact alginate-graft-gelatin composition as well as the applied cross-linking method (ionic versus physical versus covalent), the mechanical properties could be tuned. For all samples, additional Ca2+ cross-linking was shown to have a beneficial influence on the mechanical properties (i.e. their storage moduli), indicating that cation-induced cross-linking is not inhibited upon coupling both biopolymers. This effect allowed the resulting hydrogels to be reinforced in such a way that the reduction in mechanical properties at body temperature could be removed. The cross-linking efficiencies derived from HR-MAS 1H NMR spectroscopy were satisfactory for all samples studied, with all samples showing gel fractions of 80% or more. Preliminary biocompatibility work indicated that the materials support cell adhesion and growth. Increasing the gelatin content of the samples, improved the biological properties. Both cell adhesion and cell proliferation were negatively affected for the hydrogels with a higher alginate content. Future work will focus on the processing of these copolymers into tailor-made, structural supports using a variety of 3D printing and two-photon polymerization techniques to ensure an architectural mimic of the natural extracellular matrix. Acknowledgements The authors would like to acknowledge Dr. Bjorn Van Gasse and Mr. Tim Courtin for their assistance in the recording of the (HR-MAS) 1H NMR spectra. The authors would like to acknowledge Ghent University for financial support in the frameworks of the UGent-GOA project 2010–2015 (BOF10/GOA/005, Biomedical Engineering for Improved Diagnosis and Patient-Tailored Treatment of Aortic Aneurysms and Dissection) and the UGent Multidisciplinary Research Partnership Nano- and biophotonics (2010–2014). Sandra Van Vlierberghe would like to acknowledge the Research Foundation-Flanders (FWO, Belgium) for financial support under the form of a post-doctoral fellowship and a Research Grant (‘Development of the ideal tissue engineering scaffold by merging state-of-the-art processing techniques’, FWO Krediet aan Navorsers). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033. References [1] [2] [3] [4]

I.V. Yannas, Tissue and Organ Regeneration in Adults, Springer, 2001. C. Mason, P. Dunnill, Regen. Med. 3 (2008) 1–5. R. Langer, J. Vacanti, Science 260 (1993) 920–926. S. Van Vlierberghe, P. Dubruel, E. Schacht, Biomacromolecules 12 (2011) 1387–1408.

Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033

G.-J. Graulus et al. / European Polymer Journal xxx (2015) xxx–xxx [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32]

13

K.B. Djagny, Z. Wang, S. Xu, Crit. Rev. Food Sci. Nutr. 41 (2001) 481–492. S. Chatterjee, H.B. Bohidar, Int. J. Biol. Macromol. 35 (2005) 81–88. S. Van Vlierberghe, E. Schacht, P. Dubruel, Eur. Polym. J. 47 (2011) 1039–1047. J.S. Yang, Y.J. Xie, W. He, Carbohydr. Polym. 84 (2011) 33–39. Y.M. Qin, Polym. Int. 57 (2008) 171–180. Z. Ahmad, G.K. Khuller, Expert Opin. Drug Deliv. 5 (2008) 1323–1334. Peter X. Ma, Mater. Today 7 (2004) 30–40. G.T. Grant, E.R. Morris, D.A. Rees, P.J.C. Smith, D. Thom, FEBS Lett. 32 (1973) 195–198. G.G. d’Ayala, M. Malinconico, P. Laurienzo, Molecules 13 (2008) 2069–2106. A.I. Van Den Bulcke, B. Bogdanov, N. De Rooze, E.H. Schacht, M. Cornelissen, H. Berghmans, Biomacromolecules 1 (2000) 31–38. M.A. Daniele, A.A. Adams, J. Naciri, S.H. North, F.S. Ligler, Biomaterials 35 (2014) 1845–1856. A. Saarai, V. Kasparkova, T. Sedlacek, P. Saha, J. Mech. Behav. Biomed. Mater. 18 (2013) 152–166. X. Hu, L. Lu, C. Xu, X. Li, Int. J. Biol. Macromol. 72 (2015) 403–409. B. Balakrishnan, N. Joshi, A. Jayakrishnan, R. Banerjee, Acta Biomater. 10 (2014) 3650–3663. B. Sarker, R. Singh, R. Silva, J.A. Roether, J. Kaschta, R. Detsch, D.W. Schubert, I. Cicha, A.R. Boccaccini, PLoS ONE 9 (2014) e107952. P.R. Sarika, P.R. Anil Kumar, D.K. Raj, N.R. James, Carbohydr. Polym. 119 (2015) 118–125. T. Billiet, E. Gevaert, T. De Schryver, M. Cornelissen, P. Dubruel, Biomaterials 35 (2014) 49–62. S. Van Vlierberghe, Doctor of Sciences, Chemistry, Ghent University, 2008. Z. Dai, Z. Wu, S. Jia, G. Wu, J. Chromatogr. B 964 (2014) 116–127. M.J. Shapiro, J.S. Gounarides, Biotechnol. Bioeng. 71 (2000) 130–148. W. Li, Analyst 131 (2006) 777–781. A.D. Roy, K. Jayalakshmi, S. Dasgupta, R. Roy, B. Mukhopadhyay, Magn. Reson. Chem. 46 (2008) 1119–1126. S. Van Vlierberghe, B. Fritzinger, J.C. Martins, P. Dubruel, Appl. Spectrosc. 64 (2010) 1176–1180. A. Salamon, S. Van Vlierberghe, I. Van Nieuwenhove, F. Baudisch, G.-J. Graulus, V. Benecke, K. Alberti, H.-G. Neumann, J. Rychly, J.C. Martins, P. Dubruel, K. Peters, Materials (2014) 1–17. H. Tomida, T. Yasufuku, T. Fujii, Y. Kondo, T. Kai, M. Anraku, Carbohydr. Res. 345 (2010) 82–86. J.A. Rowley, G. Madlambayan, D.J. Mooney, Biomaterials 20 (1999) 45–53. G. Chan, D.J. Mooney, Trends Biotechnol. 26 (2008) 382–392. N.C. Hunt, L.M. Grover, Biotechnol. Lett. 32 (2010) 733–742.

Please cite this article in press as: G.-J. Graulus et al., Cross-linkable alginate-graft-gelatin copolymers for tissue engineering applications, Eur. Polym. J. (2015), http://dx.doi.org/10.1016/j.eurpolymj.2015.06.033