Materials Science and Engineering C 52 (2015) 333–342

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

Hybrid scaffold bearing polymer-siloxane Schiff base linkage for bone tissue engineering Bindu P. Nair, Dhanya Gangadharan, Neethu Mohan, Babitha Sumathi, Prabha D. Nair ⁎ Division of Tissue Engineering and Regeneration Technologies, Biomedical Technology Wing, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Thiruvananthapuram 695012, Kerala, India

a r t i c l e

i n f o

Article history: Received 4 October 2014 Received in revised form 23 February 2015 Accepted 22 March 2015 Available online 25 March 2015 Keywords: Osteogenesis Bone tissue engineering Biomineralisation Siloxane Cell-clustering

a b s t r a c t Scaffolds that can provide the requisite biological cues for the fast regeneration of bone are highly relevant to the advances in tissue engineering and regenerative medicine. In the present article, we report the fabrication of a chitosan–gelatin–siloxane scaffold bearing interpolymer-siloxane Schiff base linkage, through a single-step dialdehyde cross-linking and freeze-drying method using 3-aminopropyltriethoxysilane as the siloxane precursor. Swelling of the scaffolds in phosphate buffered saline indicates enhancement with increase in siloxane concentration, whereas compressive moduli of the wet scaffolds reveal inverse dependence, owing to the presence of siloxane, rich in silanol groups. It is suggested that through the strategy of dialdehyde cross-linking, a limiting siloxane loading of 20 wt.% into a chitosan -gelatin matrix should be considered ideal for bone tissue engineering, because the scaffold made with 30 wt.% siloxane loading degrades by 48 wt.%, in 21 days. The hybrid scaffolds bearing Schiff base linkage between the polymer and siloxane, unlike the stable linkages in earlier reports, are expected to give a faster release of siloxanes and enhancement in osteogenesis. This is verified by the in vitro evaluation of the hybrid scaffolds using rabbit adipose mesenchymal stem cells, which revealed osteogenic cell-clusters on a polymer-siloxane scaffold, enhanced alkaline phosphatase activity and the expression of bone-specific genes, whereas the control scaffold without siloxane supported more of cell-proliferation than differentiation. A siloxane concentration dependent enhancement in osteogenic differentiation is also observed. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Silicon is an element which has been shown to promote osteogenic differentiation, alkaline phosphatase activity and gene expression [1–4]. Silicon plays an important role in calcification during bone formation and increased concentrations of silicon have been found in connective tissues and bone [5,6]. Epidemiological studies have proven that silica intake rate has a positive effect on bone mineral density and bone strength at the hip site of men and pre-menopausal women [7,8]. Consequently, ceramic and polymer-inorganic hybrid scaffolds containing silicon have been extensively explored for bone tissue engineering applications [9–12]. In the perspective of a homogeneous regeneration of bone, nano-hybrids with uniformly distributed silicon are favored as scaffolds for bone tissue engineering [13,14]. Siloxane, an oxidized form of silicon can be incorporated into polymer matrices through the sol–gel method for achieving a desired hybrid having uniformly distributed silicon content. The most commonly used precursor for the covalent incorporation of siloxane into the polymer matrix is glycidoxypropyltriethoxysilane, through a secondary amine ⁎ Corresponding author. E-mail addresses: [email protected] (B.P. Nair), [email protected] (P.D. Nair).

http://dx.doi.org/10.1016/j.msec.2015.03.040 0928-4931/© 2015 Elsevier B.V. All rights reserved.

linkage stable towards hydrolysis [10,11,15,16]. Other methods reported for the covalent incorporation of siloxane include ester linkage generated on the limited number of carboxylic groups on gelatin, and also through urethane and urea linkages [17,18]. The stability of urea and urethane linkages towards hydrolysis and the limited possibility of formation of ester linkages on gelatin can reduce the ready availability of siloxane to the cells for enhanced osteogenesis. It has been reported that the release of inorganic ions like calcium and silicon into the culture medium and their internalization can enhance osteogenesis significantly [5,19]. These inorganic ions can also serve as nucleating agents for the mineralization of osteoids, helping in the acceleration of bone nodule formation and maturation, which is a key factor for the success of bone tissue engineering [5,20,21]. Schiff base or an imine, which is produced through the reaction of an aldehyde or ketone with an amine, has been used for various bio-medical applications, including cross-linking of polymers for bone tissue engineering [22–24]. However, no report is available hitherto, which explored the cross-linking chemistry using a dialdehyde to covalently incorporate a siloxane into a polymer matrix for fabricating a hybrid scaffold for bone tissue engineering. It is expected that the incorporation of siloxane and simultaneous cross-linking of polymers through the Schiff base linkage, which is reactive towards hydrolysis can yield a scaffold that can provide enhanced osteogenesis [25]. In the present study,

334

B.P. Nair et al. / Materials Science and Engineering C 52 (2015) 333–342

covalent incorporation of amine-containing siloxane through Schiff base formation into gelatin, a partial derivative of collagen; and chitosan, an amine-bearing polysaccharide is attempted. Though supports cell adhesion and proliferation, chitosan's major drawback is its high brittleness and low mechanical strength and a combination of gelatin and chitosan was opted to compensate for the brittleness of chitosan [26,27]. Covalent incorporation of siloxane in to chitosan-gelatin matrices through Schiff base formation has not been reported yet. Structural and in vitro evaluation of the scaffold for bone tissue engineering applications using rabbit adipose mesenchymal stem cells (AD-MSC) is presented in the article. 2. Experimental section 2.1. Materials Gelatin (acid extracted from bovine skin with bloom number equal to 225, Mw ~ 50,000), chitosan (with a degree of deacetylation of ≥ 75%) and 3-aminopropyltriethoxysilane (AS) were purchased from Sigma (USA). Glutaraldehyde solution (25% w/w, aqueous), all the inorganic salts for simulated body fluid and hydrochloric acid were of analytical grade and purchased from Merck Specialties Pvt. Ltd., India. Dulbecco's Modified Eagles Medium, High Glucose (DMEM-HG), fetal bovine serum (FBS), and penicillin-streptomycin were procured from Gibco (USA).

the loss of dimensional stability. Briefly, to a 1.2 wt.% solution of chitosan in 0.1 N HCl, gelatin (4.8 wt.%) was added so that the total polymer content will be 6 wt.%, and the solution was heated at 40 °C for 1 h to ensure complete dissolution of gelatin. 3-aminopropyltriethoxysilane was hydrolysed by aging in 0.1 N HCl for 48 h to obtain the hydrolysed siloxane solution. To the polymer solution, the hydrolysed siloxane solution was added and stirred using a mechanical stirrer for 30 min. The viscous solution thus obtained was further cross-linked by mixing with glutaraldehyde at a polymer:glutaraldehyde weight ratio of 14:1, and the highly viscous solution thus obtained was immediately transferred in to cylindrical moulds and allowed to incubate at 4 °C for 6 hours, to obtain a gel. Subsequently, the cross-linked hybrid gel was frozen in cylindrical molds by being held at −80 °C for 5 h and then lyophilized by holding for 48 h to obtain the hybrid scaffold. The scaffold was washed twice with neutralizing buffer followed by washing four times with distilled water to remove the adsorbed acid and unreacted siloxane (each washing using a shaker took an average of 2 h). Scaffolds of the required dimension for various studies were cut from the bulk scaffold using a scalpel. Chitosan–gelatin scaffold without siloxane, was used as the control scaffold (CG). In order to confirm the incorporation of siloxane using glutaraldehyde cross-linking, the hydrolysed siloxane solution was precipitated by cross-linking with glutaraldehyde and the powder obtained (SIL) was vacuum dried at 80 °C. Fabrication of the scaffolds is schematically shown in Fig. 1. The hybrid scaffold CG30 showed a weight loss of 48 wt.% in 21 days, and was not opted for the in vitro evaluation (Fig. S1, Supplementary Information).

2.2. Methods 2.2.1. Fabrication of the scaffolds The hybrid scaffolds were prepared in two steps, cross-linking and freeze drying. The combination of gelatin and chitosan was opted to compensate for the brittleness of chitosan [25,26]. A chitosan:gelatin weight ratio of 1:4 was chosen to provide sufficient mechanical strength for the scaffold. Representative siloxane concentrations of 10 and 20% with respect to the total weight of the polymers were opted and the scaffolds obtained were denoted as CG10 and CG20, respectively. The maximum siloxane concentration was limited to 20 wt.% because a siloxane concentration of 30 wt.% (CG30) resulted in extensive degradation and shrinkage of the scaffold up on cross-linking with glutaraldehyde, leading to

2.2.2. Structural evaluation of the scaffolds The scaffolds were characterized using X-ray diffraction (XRD), Fourier-Transform Infrared Spectroscopy (FT-IR), thermogravimetry (TGA), differential scanning calorimetry (DSC), scanning electron microscopy (SEM), energy dispersive spectra (EDS), for its structure and morphology. XRD data for 2θ between 5° and 25° were collected on a Philips X'pert Pro X-ray diffractometer equipped with a graphite monochromator and X'celerator detector. FT-IR measurements were made on a Perkin-Elmer Spectrum one spectrophotometer in the range of 4000– 400 cm−1 using KBr pellets containing 2 wt.% samples. TGA was performed on a TGA-50 (Shimadzu) thermogravimetric analyzer employing a heating rate of 10 °C/min from 30 to 775 °C under a nitrogen flow of

Fig. 1. Scheme showing the steps involved in the fabrication of the hybrid scaffolds CG10 and CG20. Scaffold without siloxane (CG) was fabricated as the control and SIL was precipitated for confirming the incorporation of siloxane in to the polymer matrix.

B.P. Nair et al. / Materials Science and Engineering C 52 (2015) 333–342

20 mL/min. DSC was performed using a Waters differential scanning calorimeter with mass flow control calibrated using indium as standard. Samples were encapsulated in aluminium pans and heated at a rate of 10 °C/min from − 50 to 150 °C. SEM images and EDS were taken in a FEI Quanta 200 ESEM FEG scanning electron microscope. 2.2.3. Mechanical testing Mechanical properties of the scaffolds in the dry state and the wet state (in air, after hydrating in water for about 4 h and removing excess water using a tissue paper) were determined using compressive testing. Scaffolds were made into cylindrical samples with an aspect ratio of 2:1 (13 mm in diameter and 26 mm in height). Compressive testing was performed on a computer-controlled WDW-1 universal material testing machine with a cross head speed of 1 mm/min, by giving a maximum strain of 30%. The compressive elastic modulus was calculated from the initial slope of the stress–strain curve below a strain of 15%. The value reported is the average of the measurements on at least six samples (n = 6). 2.2.4. Swelling and degradation studies The swelling capacities of the scaffolds were determined in PBS buffer solution of pH 7.4 at room temperature. The samples were removed from the swelling medium at regular intervals, washed with distilled water and blotted with a piece of tissue paper to absorb excess water on the surfaces. Swelling was calculated by weighing the scaffold before and after soaking in PBS. The percentage swelling was calculated as [(Wh − W0) / W0] × 100, where Wh is the weight of the swollen scaffold at regular intervals and W0 is the weight of the dry scaffold. (PBS) and the weights of the scaffolds were recorded at regular intervals after washing two times with water, and drying the samples using a lyophilizer. The porosity of the scaffolds in the wet state was measured by the liquid displacement method. The scaffolds were immersed in water for 4 h for complete swelling. The swollen scaffolds were taken out and excess water filling the pores were removed using cotton tissue. The sample was immersed in known volume (V1) of hexane in a graduated cylinder. The sample was left in the hexane for approximately 10 min, in a shaker. The total volume of the hexane and the hexane filled scaffold was V2. The volume difference (V2 − V1) is the volume of the polymer scaffold. The scaffold was then removed from the cylinder and the residual hexane volume was recorded as V3. The quantity (V1 − V3), volume of hexane within the scaffold, was determined as the void volume of the scaffold. The porosity of the scaffold was obtained by (V1 − V3) / (V2 − V3). 2.2.5. In vitro evaluation 2.2.5.1. Biomimetic mineralization studies. Simulated body fluid (SBF) was prepared as per method reported in literature [28]. Scaffolds of dimension 5 × 5 × 1 mm3 were then soaked in 100 mL of SBF in an incubation box at 37 °C for 14 days. Finally, the scaffolds were washed once with de-ionized water, and lyophilized at − 80 °C for 6 h. EDS were taken in the FEI Quanta 200 ESEM FEG scanning electron microscope. 2.2.5.2. Isolation of AD-MSCs. The rabbit AD-MSCs used in this study were obtained from New Zealand white rabbits, aged between 6 months and 1 year, and weighing between 2 and 2.5 kg. All procedures were approved by the Institutional Animal Ethics Committee (IAEC). AD-MSCs were isolated from the subcutaneous site of quarantined rabbits. The tissue obtained was washed twice with PBS to remove blood. The adipose tissue was then minced finely using sterilized surgical scissors, followed by digestion for 15 min at 37 °C with collagenase type II (Sigma) in PBS (0.075% w/v) under constant shaking. After digestion, DMEM-HG containing 10% FBS, penicillin 100 U/mL, and streptomycin 100 mg/mL was added to inactivate enzymatic activity. Cells were pelletized by centrifugation at 2000 g for 5 min, re-suspended in

335

DMEM with FBS, seeded into 25 cm2 flasks at 6000 cells/mL, and incubated at 37 °C with 5% CO2. Non-adherent cells were removed by changing the culture medium after incubation overnight. After 10 days of primary culture, the cells were trypsinized at 80% confluence. Cells at the fifth passage were used for subsequent experiments. 2.2.5.3. In vitro osteoinduction on scaffolds. Scaffolds of dimensions 5 × 3 × 2 mm3 were placed in 24-well culture plates, submerged in 99% ethanol, and sterilized by allowing the ethanol to evaporate overnight. After ethanol had dried completely, scaffolds were next rinsed three times with excess sterile phosphate-buffered saline. The scaffolds were pre-wetted using 50 μL of DMEM-HG prior to cell-seeding and incubated for 3 h in 24-well plates. AD-MSCs were then seeded onto scaffolds as a concentrated droplet such that the density of cells was 4000 cells/mm3 of the scaffold. 1 mL of the medium was placed in each well containing the samples. The cells were supplemented with 1 mL DMEM-HG containing 10% FBS, 1% penicillin–streptomycin, 10 nM dexamethasone, 1 mM β-glycerol phosphate and 100 μM ascorbic acid 2-phosphate (Sigma-Aldrich). The medium was changed every 2 days, and the scaffolds were cultured for up to 7, 14 and 21 days. 2.2.5.4. SEM analysis of the cell-seeded scaffolds. The cell-seeded scaffolds after one, two and three weeks were fixed in paraformaldehyde solution (4% w/v) for 6 h. Thereafter, the scaffolds were serially dehydrated with increasing concentrations of methanol, coated by an ultra-thin layer of gold, and the morphology of the cells was examined by scanning electron microscopy (SEM; Hitachi S-2400, Tokyo, Japan). 2.2.5.5. Cell viability. At the end of days 2 and 21, the viability of the cells on the scaffolds was assessed using the Live/Dead staining kit (Molecular Probes, Eugene). The scaffolds were rinsed in PBS and incubated in 4 mM calcein-AM and 2 mM ethidium homodimer-1 for 45 min at room temperature. The samples were again rinsed in PBS, and images were obtained on a Nikon A1R confocal microscope. Green fluorescence of calcein-AM was detected by using a 488 nm Argon ion laser and a band pass 505–550 filter. Red fluorescence of ethidium homodimer-1 was detected by using a 543 nm helium–neon laser and a long pass 560 filter. Red, green and blue auto-fluorescence from the scaffolds was also measured and the overlapped image was used for distinguishing the cells and the scaffold. 2.2.5.6. Cell proliferation. Scaffolds were evaluated for cellularity on 7, 14 and 21 days, by determining the DNA content. The cells were first lysed by incubating in cell-lysis buffer containing 0.05% Triton X-100 (Sigma Aldrich) at 37 °C in a shaker. The DNA amount in the lysate was quantified using the Qubit dsDNA BR Assay Kit and a Qubit 2.0 Fluorometer. 2.2.5.7. Histology. The scaffolds on days 14 and 21 were sectioned and processed as explained below. Briefly, the cultured scaffolds were fixed in 4% paraformaldehyde solution for 6 h and dehydrated by passing through an increasing series of ethanol (50%, 70%, 80%, 90% and 100%) and then in xylene. Processed scaffolds were embedded in paraffin wax for 60 min. Sections of 5 μm thickness were cut using a microtome (Leica, Germany). The sections were stretched out in a water bath at 40 °C, mounted on poly-L-lysine coated microscope slides and left to dry in an air oven at 37 °C overnight. The sections were deparaffinized in xylene for 5 min and rehydrated by passing through a series of alcohol (100%, 90%, 80%, 70% and 50%) and finally in distilled water. Hematoxylin and eosin staining (H&E) was employed for visualizing, respectively the nuclei of the cells and the cytoplasm. A 1% Alizarin red S solution at pH 4.1 was used to visualize the deposition of calcium phosphate. The slides were immersed in alizarin solution for 10 min and the unfixed dye was removed by gently rinsing the glass slides with pure water.

336

B.P. Nair et al. / Materials Science and Engineering C 52 (2015) 333–342

Table 1 Details of primer sequences. No.

Gene

Primer sequences

1

Osteopontin

2

Runx2

3

ALP

4

GAPDH

F-5′ CACCATGAGAATCGCCGT-3′ R-5′ CGTGACTTTGGGTTTCTACGC-3′ F-5′ GCCGTAGAGAGCAGGGAAGAC-3′ R-5′ CTGGCTTGGATTAGGGAGTCAC-3′ F-5′ AGCGACACGGACAAGAAGC-3′ R-5′ GGCAAAGACCGCCACATC-3′ F-5′ ACCCATCACCATCTTCCAGGAG-3′ R-5′ GAAGGGGCGGAGATGATGAC-3′

2.2.5.8. Alkaline phosphatase activity. Osteogenic differentiation of the cells on the scaffolds was monitored by performing the alkaline phosphatase (ALP) activity assay. ALP activity was determined by colorimetric endpoint assay which measures the enzymatic conversion of p-nitrophenyl phosphate (pNPP) to the yellowish product p-nitrophenol (pNP) in the presence of ALP. The same lysate solution that was used to determine the DNA content was used for this purpose and the solution was centrifuged at 10,000 rpm for 10 min. To 50 μL of the supernatant, a 200 μL portion of pNPP solution (4 mg/mL) was added and incubated for 15 min. The reaction was stopped by adding 1 N NaOH, and the absorbance was measured at 405 nm on a plate reader. All samples were run in triplicate and compared to pnitrophenol standards. The ALP activity was normalized by the amount of DNA obtained from the Qubit assay. Alkaline specific phosphatase activity was expressed as micromoles of pNP/μg DNA/min. 2.2.5.9. Reverse transcription-polymerase chain reaction (RT-PCR) analysis. The relative gene expression of osteogenic markers ALP, osteopontin (OPN) and RunX2 on the hybrid scaffolds was analyzed on days 7, 14 and 21 using RT-PCR and compared with the gene expression on CG, according to standard protocols. Briefly, total RNA was extracted from the scaffolds using Trizol reagent (Invitrogen) according to the manufacturer's instructions. RNA concentration and purity were measured using a Nanodrop® ND-1000 UV–Vis Spectrophotometer

Table 2 Percentage porosity of the wet scaffolds and compressive moduli of the dry and wet scaffolds. Scaffold

% porosity

Compressive modulus (dry) (MPa)

Compressive modulus (wet) (kPa)

CG CG10 CG20

72.1 ± 1.5 69.4 ± 1.3 64.2 ± 1.2

7.3 ± 0.8 9.2 ± 1.4 11.4 ± 0.9

52.8 ± 1.8 45.5 ± 1.7 38.7 ± 2.6

and cDNA was synthesized from 1 μg of total RNA using Revert Aid H Minus first strand cDNA synthesis kit (Fermentas, Germany) according to the manufacturer's instructions. Quantitative real time PCR was performed with Applied Biosystems 7500 Real-Time system using Fast SYBR Green Master Mix (Applied Biosystems, California, USA) with primers listed in Table 1. After normalization to GAPDH, relative expression levels and fold induction of each target gene were calculated using a comparative CT method ΔCT, where ΔCT = CT of the gene of interest — CT of the reference gene (GAPDH), with Microsoft Excel 2007. 2.2.5.10. Statistical analysis. All measurements of three independent experiments were collected in triplicate and expressed as mean ± standard deviation. For all experiments, statistical significance of differences between groups was determined using t-test for two treatments and one-way ANOVA for more than two treatments with Tukey's post hoc test in Vassar Stats: Website for Statistical Computation. In all the cases, when the difference was significant, symbols are used to indicate the difference. 3. Results and discussion 3.1. Structural evaluation of the scaffolds The aim of the study was to fabricate a chitosan–gelatin–siloxane scaffold through a single step dialdehyde cross-linking. The literature overwhelmingly shows that the incorporation of siloxane can have a

Fig. 2. SEM images of the cross-sections of the scaffolds; (a) and (d) CG, (b) and (e) CG10, and (c) and (f) CG20, at two different magnifications.

B.P. Nair et al. / Materials Science and Engineering C 52 (2015) 333–342

337

Fig. 3. (a) FT-IR spectra of chitosan, gelatin, CG20 and SIL, (b) thermogram, (c) percentage swelling and (d) degradation profile in PBS (pH 7.4), of CG, CG10 and CG20. Values are expressed as mean ± standard deviation from three different replicates. * indicates statistical significance of p b 0.001 (CG vs. CG20).

positive effect on osteogenic lineage progression [10,11,15]. What makes this study relevant is that a scaffold bearing interpolymer-siloxane Schiff base linkage, which can provide siloxane at a faster rate, has not been attempted so far. For achieving the aim, the siloxane precursor AS was prehydrolysed under acidic condition to produce amine-containing siloxane having silanol (Si–OH) groups and random Si–O–Si linkage. The choice of prehydrolysis under acidic condition was based on the literature reports that the rate of hydrolysis of AS under acidic condition is slow when compared to that under basic conditions and hence extensive condensation of silanes and the precipitation of insoluble siloxane network can be prevented [29,30]. If the condensation reaction was extensive before or after the addition of siloxane into the polymer matrix, and before the commencement of cross-linking with glutaraldehyde, it would have led to a non-homogeneous distribution of siloxane within the polymer matrix and affected the aim of achieving a polymer-siloxane hybrid scaffold with uniformly distributed siloxane. The strategy was continued after the confirmation obtained through the clear nature of the hydrolysed siloxane solution. Fig. 2 shows the SEM images of the horizontal cross-sections of the control scaffold CG and the hybrid scaffolds, CG10 and CG20, at two different magnifications. Pore size of the scaffolds varied between 50-300 µm and the strut thickness between 10-20 µm. The SEM images confirmed the three dimensional inter-connected nature of the porosity. From the SEM images, it was evident that addition of siloxane resulted in more irregular pores for the hybrid scaffold CG10 when compared to the control scaffold. A plausible reason for this could be the siloxane-rich micro-domains formed within the polymer-matix due to hydrogen-bonding interaction with the polymers, which can behold more water and during lyophilization will lead to large and irregular distribution of pores. This irregualr distribution of pore sizes was found decreasing with further increase in siloxane concentration due to sufficient availabiliy of siloxane for interaction with high polymercontent. EDS spectra, confirmed the siloxane content in the hybrid scaffolds (Figure S2, Suplementary Information). The percentage porosity of the CG, CG10 and CG20 in the wet state were respectively, 72.1±1.5, 69.4±1.3, 64.2±1.2 (Table 2). The color of the scaffolds showed a gradual shift from yellow to orange with incorporation of and increase in

siloxane content (Fig. 1). The SIL also appeared orange red in color, confirming its contribution towards the change in the colour of the hybrid scaffolds with increase in siloxane concentration. When the pre-hydrolysed AS was added to a solution of chitosan and gelatin, a slight increase in the viscosity of the mixture was observed, which could be due to gelling attained through hydrogen bonding interaction between the amine and the silanol groups on the hydrolysed siloxane and the amine and the hydroxyl groups on chitosan and gelatin [31]. The reactive group available in the chitosan–gelatin–siloxane mixture for further cross-linking with glutaraldehyde was amino-groups on chitosan, gelatin and siloxane through an imine-type Schiff base formation. Compared to gelatin for which the amino and hydroxyl groups are scarce from lysine, hydroxylysine, hydroxyproline etc., chitosan has one amino-group on its every structural repeating unit and therefore possesses more reactivity towards glutaraldehyde. A lower chitosan:gelatin weight ratio of 1:4 prevented a non-homogeneous distribution of chitosan during cross-linking with glutaraldehyde. For confirming the nonreactivity of silanol groups towards glutaraldehyde cross-linking, vinyltriethoxysilane not bearing amino groups was hydrolysed under identical condition of AS and reacted with glutaraldehyde, however no product was formed. The observation suggests the presence of free silanol groups in the hybrid scaffolds, which was supported by FT-IR spectra. Even though the incorporation of siloxane through Schiff base formation was the focus of the present study, the formation of Schiff base linkages between chitosan and gelatin, or among the individual polymers and also the formation of enamines through the secondary amino group on gelatin cannot be ruled out; however the method was expected to complement the overall integrity of the scaffold through extensive cross-linking of its all the three components. Fig. 3a shows the FT-IR spectra of chitosan, gelatin, and CG20; as a representative of the hybrid scaffolds and SIL in the scan range of 500– 2200 cm−1. For gelatin, the peaks can be assigned as 1636–1640 cm−1 (amide I), 1542–1544 cm−1 (N–Hbend of amide II and amine, and C– Hstr), and 1240 cm−1 amide III (C–Nstr and N–Hbend). For chitosan, the peak at 1500–1540 cm−1 was due to N–Hbend, 1094 cm−1 due to C–Ostr and 916 cm−1 due to typical saccharine structure. The SIL showed peaks due to Si–O–Sibend at 740 cm−1, Si–OH symmetric stretching from incompletely condensed siloxane at 835 cm−1, and Si–O–Sistr at

338

B.P. Nair et al. / Materials Science and Engineering C 52 (2015) 333–342

Fig. 4. SEM images of the cell seeded scaffolds CG, CG10 and CG20 on days 7, 14 and 21. Scale bar represents 30 μm.

950–1080 cm−1, imine stretching at 1663 cm−1 and C–Hbend at 1420 cm−1. The intensity of the peak corresponding to incompletely condensed siloxane at 835 cm−1 was distinguishably higher than that of the completely condensed Si–O–Si vibration at 950–1080 cm−1. The hybrid scaffold CG20 exhibited the characteristic peaks of imine and the siloxane and silanol groups exhibited by SIL. FT-IR spectra as a single separated figure is provided as Fig. S3, Supplementary Information. From the FT-IR spectra, it was confirmed that the siloxane was successfully incorporated into the polymer matrix through the Schiff base formation. FT-IR also supported the strategy of pre-hydrolysis and glutaraldehyde cross-linking; as it evidenced that the siloxane present in the hybrid scaffold contains mostly incompletely condensed silanol groups with a limited degree of Si–O–Si network formation. Fig. 3b shows the thermograms of CG and the hybrid scaffolds, indicating silica dry weight contents of 1.2 and 4 wt.% at 775 °C, respectively for CG10 and CG20, which was arising from the calcination of siloxanes present in the hybrid scaffolds. Although a higher siloxane loading; 10 and 20 wt.% with respect to the amount of polymers were adopted for the study, it was confirmed from thermogravimtery that the strategy was not leading to the complete incorporation of the siloxane loading given. From the silica dry weight, the actual siloxane content arising from the hydrolysed siloxanes for CG10 and CG20 was calculated as approximately equal to or greater than 3 and 10 wt.%, by using the approximation that two molecules of hydrolysed AS can yield one siloxane linkage. The plausible reason for this decrease in siloxane content could be the removal of a fraction of low molecular weight siloxanes from the scaffold due to its highly hydrophilic nature, during extensive washing with distilled water, post-fabrication. The hybrid scaffold CG20 was analyzed for the

presence of any localized ordering and crystallinity induced due to the presence of siloxane, using DSC, by comparing with pure chitosan and gelatin (Fig. S4, Supplementary Information). From the second heating curves in the DSC plots, the endothermic transition corresponding to the loss of crystallinity of chitosan was absent for CG20. In the X-ray diffractogram also, the peak corresponding to the crystallinity of chitosan was absent (Fig. S5, Supplementary Information). The loss of the crystalline peak of chitosan in the DSC plot and the X-ray diffractogram confirmed that chitosan exhibited a disordered structure in the hybrid scaffolds than in its pure state. These results confirm that the hybrid scaffolds are having covalently incorporated siloxane and a completely disordered network of its individual polymers: chitosan and gelatin. 3.2. Swelling, degradation profile and in vitro mineralization An important parameter that decides the usefulness of a scaffold for tissue engineering application is its swelling ability, which determines the transfer of cell nutrients and metabolites within the scaffold. The water absorption capacities of CG, CG10 and CG20 are shown in Fig. 3c. CG exhibited a water absorption capacity 4 times its dry weight in 6 h. The hybrid scaffolds exhibited water absorption capacity greater than that of CG, which was also found increasing with increase in siloxane concentration. The water absorption capacities of the control scaffold and the hybrid scaffolds were mainly attributed to its high gelatin content. Interestingly, the incorporation of siloxane through the present method was found to increase the water absorption ability of the hybrid scaffolds. The plausible reason for the enhancement in water uptake is the presence of hydrophilic silanol groups on siloxanes, which was

B.P. Nair et al. / Materials Science and Engineering C 52 (2015) 333–342

339

Fig. 5. Confocal micrographs of the cell seeded scaffolds CG, CG10 and CG20 on days 2 and 21. Live cells were stained using calcein AM, dead cells using ethidium homodimer-1 and merged image of red, green and blue autofluorescence was used for visualizing the scaffolds. Autofluorescent images are given in detail as supporting information. Scale bar represents 50 μm.

also found to favor initial cell adhesion on the hybrid scaffolds. The irregularity in the pore size of the hybrid scaffolds when compared to the control scaffold was found to have no significant effect on its water absorption and retention capacity, due to the high gelatin content and hydrophilicity of all the scaffolds. The degradation profile studied in PBS at pH 7.4 also showed a direct dependence on siloxane concentration (Fig. 3d). A total weight loss of 18, 22 and 28 wt.% was found, respectively for CG, CG10 and CG20, on day 21. The hybrid scaffolds exhibited a degradation rate greater than that of the control scaffold. Siloxane and silanol linkages are considered as strong bonds against hydrolysis. Therefore the enhanced degradation of hybrid scaffolds could be due to the easy penetration of water molecules into the scaffold facilitated by the hydrophilic silanol groups, which in turn can attack the reactive Schiff base linkages initiating hydrolysis. Mechanical properties of the scaffolds were compared by measuring the uni-axial compressive modulus, both in the dry and the wet states (Table 2). Compressive moduli of the scaffolds showed improvement with increase in siloxane concentration, in the dry state. The compressive moduli of CG, CG10 and CG20 in the dry state were, respectively 7.3 ± 0.8, 9.2 ± 1.4 and 11.4 ± 0.9 MPa. However in the wet state compressive moduli were inversely proportional to the siloxane concentration. The values were 52.8 ± 1.8, 45.5 ± 1.7 and 38.7 ± 2.6 kPa, respectively for CG, CG10 and CG20. It is presumed that the siloxane in dry state can reinforce the polymer since it is having inorganic nature in its nano-domains inside the polymer matrix. However, when solvated

the siloxane will behave like a hydrophilic gel, which can decrease the mechanical property of the polymers. In agreement with the literature reports, the ability of siloxanes to precipitate calcium phosphate was confirmed from the SEM images of the hybrid scaffolds immersed in SBF for 14 days (Fig. S6, Supplementary Information) [17]. A siloxane concentration-dependent enhancement in the deposition of calcium phosphate was evident, whereas CG failed in depositing the mineral. The increased concentration of siloxanes in CG20 caused the precipitation of larger particles of calcium phosphate when compared to CG10 of lower siloxane concentration. It is presumed that when the siloxane concentration is high, it can provide large nucleation sites for precipitating larger mineral particles. 3.3. SEM analysis of the cell-seeded scaffolds After confirming the structural aspects of the hybrid scaffolds and its capacity to support cell adhesion and growth, we proceeded to in vitro evaluation. SEM images of the cell-seeded scaffolds on days 7, 14 and 21 are shown in Fig. 4. On day 7, the cells were appeared spread throughout the surface of the scaffold in the case of CG. On CG10, the cells were well attached to the surface of the scaffold and also appeared to migrate into the pores. In contrast to CG and CG10, organization of cells into clusters was evident on CG20 on day 7. Several clusters were observed within the pores of CG20 and the high magnification image of a cell-cluster within a pore is shown in Fig. 4 (Low magnification images are given as Fig. S7, Supplementary Information.). Adhesion of cells on the surface

340

B.P. Nair et al. / Materials Science and Engineering C 52 (2015) 333–342

of osteogenic cell-clusters on a soft polymer-siloxane scaffold could be due to the faster availability of siloxane into the culture medium. 3.4. Cell viability

Fig. 6. (a) Total DNA content (* Statistical significance of p b 0.05 for CG vs. CG20, p b 0.05 for CG vs. CG10, #p b 0.001 for CG vs. CG10, **p b 0.001 for CG vs. CG20 and ¤p b 0.05 for day 14 vs. day 21), and (b) normalized ALP activity ( p b 0.05 for CG vs. CG10, *p b 0.001 for CG vs. CG20, #p b 0.05 for CG10 vs. CG20, ¤p b 0.05 for day 7 vs. day 14) on days 7, 14, and 21.

of the scaffold and diffusion of cells into the inner pores were also evident. As day proceeded to 14, clogging of the pores with cells was evident for CG, whereas the cell-clusters on CG20 were found to be more condensed. On day 21, complete clogging of the pores with a thick deposition of the cells was evident on CG. When compared to CG, matrix deposition to the inner pores of the scaffold was evident for the hybrid scaffolds. The cell-clusters on CG20 appeared more spread and enlarged in size (Fig. S7, Supplementary Information). This first time observation

To assess cell infiltration into the 3D scaffolds, cell distribution on and throughout the scaffolds was visualized by confocal microscopy on days 2 and 21 (Fig. 5). The scaffold itself was highly autofluorescent in all the three filters, emitting red, green and blue fluorescence (due to characteristic properties of gelatin and siloxane), however the autofluorescence was adjusted in such a way to visualize the live or dead cells and the scaffold. The fluorescence was found varying locally as day proceeds, especially with difference in composition of the scaffold due to degradation. The confocal images obtained using the three filters and the merged images, are given as Figs. S8–S9, Supplementary Information. In Fig. 5, each panel shows live cells on the scaffold stained using calcein AM and the merged image of calcein AM and ethidium homodimer-1 fluorescence (for dead cells) with the auto-fluorescence of the scaffold (a sum of red, blue and green fluorescence). A slight macroscopic difference in cell density was detectable on day 2; CG20 showed cells with spread morphology, compared to CG and CG10. On day 21, the cells on CG appeared randomly distributed in a monolayer, whereas those on CG10 appeared to exhibit clustering behavior, supporting the observations from SEM. The cell-clusters on CG20 also appeared to include mostly live cells, and the red color was due to autofluorescence from the scaffold (are given as Figs. S10–S11, Supplementary Information). A quantification of DNA showed that on day 7, the number of cells on CG20 showed significantly higher value than CG and a non-significant difference from CG10 (Fig. 6a). By day 14, the cell number on CG increased drastically indicating cell proliferation (p b 0.05, against CG10), whereas the increase was non-significant for CG10 and CG20 when compared to day 7. On day 21, the cell content for CG was even higher than that from day 14 (p b 0.001, against CG10 and CG20), whereas the hybrid scaffolds showed only a nonsignificant increase, in comparison with day 14. 3.5. Alkaline phosphatase activity A comparison between the normalized ALP activity of the control scaffold and the hybrid scaffolds, with respect to the DNA content indicated significant ALP activity for CG20 on day 7 (Fig. 6b). For the hybrid scaffolds, the value increased to a maximum at 14 days of culture, and thereafter remained nearly constant. On day 14 and day 21, the ALP activity of the hybrid scaffolds was significantly increased when compared to the control scaffold, but the increase was significant among CG10 and CG20, on day 21 only. A comparison on the total and normalized ALP activity is given as Fig. S12, Supplementary Information. 3.6. Histology and alizarin red staining

Fig. 7. Hematoxylin and eosin staining of the histology sections, showing cell clusters on the hybrid scaffolds CG10 and CG20, in comparison with the control scaffold CG, at days 14 and 21. Scale bar represents 50 μm.

To visualize the cell distribution and the extent of mineralization on the scaffolds, histological cross sections of the scaffolds were taken and stained using H&E and alizarin red, the specific marker for calcium deposition. H&E staining revealed that the cell distribution on CG was uniform, whereas the hybrid scaffolds showed increased localized cell density and clustering (Fig. 7), supporting the observation made from SEM and confocal images. Also, positive signal for calcium staining using alizarin red S was evident for the hybrid scaffolds indicating higher calcium deposition than the control scaffold. A comparison on alizarin red staining of the cell seeded and non-cell seeded scaffolds on day 14 and 21 is shown in Fig. 8. The cell clusters on the hybrid scaffolds were found to have higher concentration of calcium, than the control scaffold, however it was difficult to reach a conclusion on the difference in the quantity of mineral deposition for the hybrid scaffolds, from alizarin red staining alone.

B.P. Nair et al. / Materials Science and Engineering C 52 (2015) 333–342

341

Fig. 8. Alizarin red S staining of the histology sections of CG, CG10 and CG20 scaffolds at day 14 and 21, in comparison with the non-cell seeded scaffolds, showing calcium rich cell clusters on the hybrid scaffolds. Scale bar represents 100 μm.

3.7. RT-PCR The mRNA expression level of various marker genes relevant to osteogenic differentiation of AD-MSCs was evaluated by RT-PCR analysis. Significant differences in ALP and OPN gene expression were detected among the control and the hybrid scaffolds (Fig. 9). On day 7, the expression was significant for the hybrid scaffolds and by days 14 and 21, the expression of osteogenic markers was markedly up-regulated when compared to CG. Therefore, RT-PCR analysis was used to confirm cell-proliferation without significant differentiation on CG and osteogenic differentiation on the hybrid scaffolds. A siloxane concentration dependent enhancement of the bone specific genes was also evident by day 21. The expressions of Runx2, the master osteogenic transcription factor regulating the MSC differentiation into the osteoblasts were also significantly higher for the hybrid scaffolds and maintained high level at days 14 and 21. The relatively lower expression of bonespecific genes by the control scaffold and lower ALP activity, despite having higher cell content confirmed cell proliferation over differentiation. The lower cell content for the hybrid scaffolds in comparison with the control scaffold also could be due to this lag phase between osteogenic differentiation and a simple proliferation of stem cells. Mineralized nodule formation of human primary osteoblasts on a porous bioactive glass scaffold was reported in literature [21]. However; the formation of mineralized cell-clusters on an organopolymer-siloxane hybrid scaffold is reported for the first time. Such an observation is highly important with regard to developing materials for various bone tissue engineering applications. The finding that the control scaffold without siloxane failed to induce any cell-clustering and significant osteogenic differentiation, also demonstrates the remarkable ability of siloxane to cause osteogenic differentiation and osteoblast mineralization. 4. Conclusions

Fig. 9. Relative gene expression of osteogenic markers (a) ALP, (b) OCN, and (c) OPN on CG, CG10 and CG20. GAPDH was used as the reference gene. All values are expressed as mean ± standard deviation from three different replicates. * Statistical significance of p b 0.05 for CG vs. CG10, **p b 0.001 for CG vs. CG10, ^p b 0.05 for CG10 vs. CG20, and #p b 0.001 for CG vs. CG20.

In conclusion, chitosan–gelatin–siloxane hybrid scaffolds bearing interpolymer-siloxane Schiff base linkage were fabricated using 3aminopropyltriethoxysilane as the siloxane source and glutaraldehyde as the cross-linking agent, with a view to facilitate the faster availability of siloxanes for enhanced osteogenesis. The observation of osteogenic

342

B.P. Nair et al. / Materials Science and Engineering C 52 (2015) 333–342

cell clusters on an organopolymer-siloxane scaffold and enhancement in osteogenic lineage progression confirmed the success of the method. Author contributions The manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript. Acknowledgments

[11]

[12]

[13] [14] [15]

This work was supported by DST INSPIRE-Faculty fellowship and research grant (IFA-11-CH-19) to BPN from the Department of Science and Technology, India. Financial assistance from DBT Indo-Danish project-8054 is also acknowledged. We also thank Director, SCTIMST and Head, BMT Wing, SCTIMST for providing the facilities to carry out this work, and Dr. Roy Joseph for mechanical testing. Appendix A. Supplementary data Electronic supplementary information (ESI) available: DSC plots and X-ray diffractogram of CG20, chitosan and gelatin, confocal micrographs of cell-seeded scaffolds and low magnification SEM images of the cell seeded scaffolds. This material is available free of charge via the internet. Supplementary data to this article can be found online at http://dx.doi. org/10.1016/j.msec.2015.03.040. References [1] E.M. Carlisle, Silicon: a possible factor in bone calcification, Science 167 (1970) 279–280. [2] W. Zhaia, H. Lub, L. Chena, X. Lin, Y. Huang, K. Dai, K. Naoki, G. Chen, J. Chang, Silicate bioceramics induce angiogenesis during bone regeneration, Acta Biomater. 8 (2012) 341–349. [3] A.M. Pietak, J.W. Reid, M.J. Stott, M. Sayer, Silicon substitution in the calcium phosphate bioceramics, Biomaterials 28 (2007) 4023–4032. [4] M. Bohner, Silicon-substituted calcium phosphates — a critical view, Biomaterials 30 (2009) 6403–6406. [5] P. Valerioa, M.M. Pereirab, A.M. Goesc, M.F. Leitea, The effect of ionic products from bioactive glass dissolution on osteoblast proliferation and collagen production, Biomaterials 25 (2004) 2941–2948. [6] D.M. Reffitt, N. Ogston, R. Jugdaohsingh, H.F. Cheung, B.A. Evans, R.P. Thompson, J.J. Powell, G.N. Hampson, Orthosilicic acid stimulates collagen type 1 synthesis and osteoblastic differentiation in human osteoblast-like cells in vitro, Bone 32 (2003) 127–135. [7] C.T. Price, K.J. Koval, J.R. Langford, Silicon: a review of its potential role in the prevention and treatment of postmenopausal osteoporosis, Int. J. Endocrinol. 2013 (2013) 316783. [8] R. Jugdaohsingh, K.L. Tucker, N. Qiao, L.A. Cupples, D.P. Kiel, J.J. Powell, Dietary silicon intake is associated with bone mineral density in premenopausal women and postmenopausal women taking HRT, J. Bone Miner. Res. 19 (2004) 297–307. [9] M.Y. Shie, S.J. Ding, H.C. Chang, The role of silicon in osteoblast-like cell proliferation and apoptosis, Acta Biomater. 7 (2011) 2604–2614. [10] Y. Shirosaki, K. Tsuru, S. Hayakawa, A. Osaka, M.A. Lopes, J.D. Santos, M.A. Costa, M.H. Fernandes, Physical, chemical and in vitro biological profile of chitosan hybrid

[16]

[17] [18]

[19]

[20] [21]

[22]

[23]

[24]

[25]

[26]

[27] [28] [29]

[30]

[31]

membrane as a function of organosiloxane concentration, Acta Biomater. 5 (2009) 346–355. Y. Shirosaki, K. Tsuru, S. Hayakawa, A. Osaka, M.A. Lopes, J.D. Santos, M.H. Fernandes, In vitro cytocompatibility of MG63 cells on chitosan-organosiloxane hybrid membranes, Biomaterials 26 (2005) 485–493. K.A. Hing, P.A. Revell, N. Smith, T. Buckland, Effect of silicon level on rate, quality and progression of bone healing within silicate-substituted porous hydroxyapatite scaffolds, Biomaterials 27 (2006) 5014–5026. E.M. Valliant, J.R. Jones, Softening bioactive glass for bone regeneration: sol–gel hybrid materials, Soft Matter 7 (2011) 5083–5095. D. Eglin, D. Mortisen, M. Alini, Degradation of synthetic polymeric scaffolds for bone and cartilage tissue repairs, Soft Matter 5 (2009) 938–947. L. Ren, K. Tsuru, S. Hayakawa, A. Osaka, Novel approach to fabricate porous gelatin– siloxane hybrids for bone tissue engineering, Biomaterials 23 (2002) 4765–4773. L. Ren, J. Wang, F.Y. Yang, L. Wang, D. Wang, T.X. Wang, M.M. Tian, Fabrication of gelatin–siloxane fibrous mats via sol–gel and electrospinning procedure and its application for bone tissue engineering, Mater. Sci. Eng. C 30 (2010) 437–444. S.S. Silva, R.A.S. Ferreira, L. Fu, L.D. Carlos, J.F. Mano, R.L. Reis, J. Rocha, Functional nanostructured chitosan–siloxane hybrids, J. Mater. Chem. 15 (2005) 3952–3961. O. Mahony, O. Tsigkou, C. Ionescu, C. Minelli, L. Ling, R. Hanly, M.E. Smith, M.M. Stevens, J.R. Jones, Silica-gelatin hybrids with tailorable degradation and mechanical properties for tissue regeneration, Adv. Funct. Mater. 20 (2010) 3835–3845. I.D. Xynos, A.J. Edgar, L.D.K. Buttery, L.L. Hench, J.M. Polak, Ionic products of bioactive glass dissolution increase proliferation of human osteoblasts and induce insulin-like growth factor II mRNA expression and protein synthesis, Biochem. Biophys. Res. Commun. 276 (2000) 461–465. H.M. Aydin, B. Hu, J.S. Suso, A.E. Haja, Y. Yang, Study of tissue engineered bone nodules by Fourier transform infrared spectroscopy, Analyst 136 (2011) 775–780. J.E. Gougha, J.R. Jones, L.L. Hench, Nodule formation and mineralisation of human primary osteoblasts cultured on a porous bioactive glass scaffold, Biomaterials 25 (2004) 2039–2046. P.B. Malafaya, R.L. Reis, Bilayered chitosan-based scaffolds for osteochondral tissue engineering: influence of hydroxyapatite on in vitro cytotoxicity and dynamic bioactivity studies in a specific double-chamber bioreactor, Acta Biomater. 5 (2009) 644–660. M. Bhattacharjee, S. Chameettachal, S. Pahwa, A.R. Ray, S. Ghosh, Strategies for replicating anatomical cartilaginous tissue gradient in engineered intervertebral disc, ACS Appl. Mater. Interfaces 6 (2014) 183–193. S. Kuddannaya, Y.J. Chuah, M.H.A. Lee, N.V. Menon, Y. Kang, Y. Zhang, Surface chemical modification of poly(dimethylsiloxane) for the enhanced adhesion and proliferation of mesenchymal stem cells, ACS Appl. Mater. Interfaces 5 (2013) 9777–9784. J.H. Moon, J.W. Shin, S.Y. Kim, J.W. Park, Formation of uniform aminosilane thin layers: An imine formation to measure relative surface density of the amine group, Langmuir 12 (1996) 4621–4624. Y. Huang, S. Onyeri, M. Siewe, A. Moshfeghian, S.V. Madihally, In vitro characterization of chitosan -gelatin scaffolds for tissue engineering, Biomaterials 26 (2005) 7616–7627. J.S. Mao, L.G. Zhao, Y.J. Yin, K.D. Yao, Structure and properties of bilayer chitosan gelatin scaffolds, Biomaterials 24 (2003) 1067–1074. T. Kokubo, H. Takadama, How useful is SBF in predicting in vivo bone bioactivity? Biomaterials 27 (2006) 2907–2915. G.E. Kozerski, R.H. Gallavan, M.J. Ziemelis, Investigation of trialkoxysilane hydrolysis kinetics using liquid chromatography with inductively coupled plasma atomic emission spectrometric detection and non-linear regression modeling, Anal. Chim. Acta 489 (2003) 103–114. B.P. Nair, C. Pavithran, Bifunctionalized hybrid silica spheres by hydrolytic cocondensation of 3-aminopropyltriethoxysilane and vinyltriethoxysilane, Langmuir 26 (2010) 730–735. S.Sh. Rashidova, D.Sh. Shakarova, O.N. Ruzimuradov, D.T. Satabaldieva, S.V. Zalyalieva, O.A. Shpigun, V.P. Varlamov, B.D. Kabulov, Bionanocompositional chitosan -silica sorbent for liquid chromatography, J Chromatogr B 800 (2004) 49–53.