In vitro biocompatibility of 45S5 Bioglass ® -derived glass-ceramic scaffolds coated with...

JOURNAL OF TISSUE ENGINEERING AND REGENERATIVE MEDICINE R E S E A R C H A R T I C L EJ Tissue Eng Regen Med 2009; 3: 139–148.Published online 26 January 2009 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/term.150

In vitro biocompatibility of 45S5 Bioglass-derivedglass–ceramic scaffolds coated withpoly(3-hydroxybutyrate)

Oana Bretcanu1#, Superb Misra1, Ipsita Roy2, Chiara Renghini3, Fabrizio Fiori3, Aldo R. Boccaccini1*and Vehid Salih4*1Department of Materials, Imperial College London, Prince Consort Road, London SW7 2BP, UK2Department of Molecular and Applied Biosciences, University of Westminster, London W1W 6UW, UK3Department SAIFET, University Polythecnic of Marche and INBB, Istituto Nazionale Biostrutture e Biosistemi, Italy4Division of Biomaterials and Tissue Engineering, UCL Eastman Dental Institute, London WC1X 8LD, UK

Abstract

The aim of this work was to study the in vitro biocompatibility of glass–ceramic scaffolds basedon 45S5 Bioglass, using a human osteosarcoma cell line (HOS-TE85). The highly porous scaffoldswere produced by the foam replication technique. Two different types of scaffolds with differentporosities were analysed. They were coated with a biodegradable polymer, poly(3-hydroxybutyrate)(P(3HB)). The scaffold bioactivity was evaluated by soaking in a simulated body fluid (SBF)for different durations. Compression strength tests were performed before and after immersionin SBF. These experiments showed that the scaffolds are highly bioactive, as after a few daysof immersion in SBF a hydroxyapatite-like layer was formed on the scaffold’s surface. It was alsoobserved that P(3HB)-coated samples exhibited higher values of compression strength than uncoatedsamples. Biocompatibility assessment was carried out by qualitative evaluation of cell morphologyafter different culture periods, using scanning electron microscopy, while cell proliferation wasdetermined by using the AlamarBlue assay. Alkaline phosphatase (ALP) and osteocalcin (OC)assays were used as quantitative in vitro indicators of osteoblast function. Two different types ofmedium were used for ALP and OC tests: normal supplemented medium and osteogenic medium.HOS cells were seeded and cultured onto the scaffolds for up to 2 weeks. The AlamarBlue assayshowed that cells were able to proliferate and grow on the scaffold surface. After 7 days in culture,the P(3HB)-coated samples had a higher number of cells on their surfaces than the uncoated samples.Regarding ALP- and OC-specific activity, no significant differences were found between sampleswith different pore sizes. All scaffolds containing osteogenic medium seemed to have a slightlyhigher level of ALP and OC concentration. These experiments confirmed that Bioglass/P(3HB)scaffolds have potential as osteoconductive tissue engineering substrates for maintenance andnormal functioning of bone tissue. Copyright 2009 John Wiley & Sons, Ltd.

Received 11 August 2008; Revised 16 November 2008; Accepted 12 December 2008

Keywords scaffold; bioglass; composite material; bone tissue engineering; osteoblast cell; bioactivity

*Correspondence to: Aldo R. Boccaccini, Department of Mate-rials, Imperial College London, Prince Consort Road, LondonSW7 2BP, UK . E-mail: [email protected]: Vehid Salih, Division of Biomaterials and Tissue Engineer-ing, UCL Eastman Dental Institute, London WC1X 8LD, UK.E-mail: [email protected]# Current affiliation: Department of Materials Science andChemical Engineering, Politecnico di Torino, Italy.

1. Introduction

In recent decades many biomaterials for bone tissueengineering have been examined with respect to theirbiocompatibility (Ratner et al., 2004; Hench and Polak,2002). In order to best mimic bone tissue, materials

Copyright 2009 John Wiley & Sons, Ltd.

140 O. Bretcanu et al.

exhibiting high porosity (scaffolds) are being devel-oped. The scaffolds used in tissue engineering shouldhave three-dimensional (3D) interconnected porosity andappropriate pore size and morphology to allow cell migra-tion, fluid circulation and cell proliferation. Moreover, thescaffold surface should favour cell attachment, growth anddifferentiation (Lanza et al., 2000). Other requirementsconcern osteoconductivity, biocompatibility and appro-priate mechanical properties to allow optimal integrationwith the surrounding tissue (Karageorgiou and Kaplan,2005). If the scaffolds are biodegradable, their degrada-tion products should not be toxic and should not provokean adverse immune response (Rezwan et al., 2006).

Among the many biomaterials developed for bonetissue scaffolds, composite materials containing bothinorganic (glass, ceramic or glass–ceramic) and organic(biocompatible polymer) phases are considered anoptimal approach (Rezwan et al., 2006; Guarino et al.,2007). The bioactive ceramic phase is brittle but hasgood compressive strength as well as good bone-bondingability. The polymeric phase confers the toughness andplasticity. Therefore, by combining a bioactive ceramicphase with a polymeric one, composite materials with highmechanical properties and good biological behaviour canbe obtained (Rezwan et al., 2006; Guarino et al., 2007;Zhang et al., 2004).

In this work we evaluated the in vitro biocompatibilityof inorganic/organic composite scaffolds composed ofa glass–ceramic material and a biodegradable polymer.The starting inorganic phase was 45S5 Bioglass (Hench,2006), while for the organic phase the biodegradablepolymer poly(3-hydroxybutyrate) [P(3HB)] was selected.We hypothesized that addition of a polymer coatingto a base Bioglass-derived scaffold would increasethe mechanical properties and would allow the surfaceproperties and biocompatibility of these compositescaffolds to be tailored.

45S5 Bioglass is a commercially available materialthat has been used as bone replacement for more than15 years (Hench, 2006). This material is highly bioactiveand it is both osteoinductive and osteoconductive.Moreover, it has been shown that the ionic dissolutionproducts of 45S5 Bioglass may enhance new boneformation (osteogenesis) through direct control overgenes that regulate cell induction and proliferation (Xynoset al., 2000). There are also indications from previousreports of in vivo studies (Day et al., 2004) that bioactiveglass (45S5 Bioglass) enhances vascularization of tissueengineering constructs.

P(3HB) is a natural thermoplastic polymer producedby many types of microrganisms (Doi and Steinbuchel,2000). It is highly biocompatible, biodegradable and hasgood processability. The polymer retains its strengthunder long-term degradation and takes more than52 weeks for complete degradation in aqueous media(Doyle et al., 1991). In the present study, P(3HB) wasisolated from Bacillus cereus SPV by a fermentation processdescribed in detail elsewhere (Valappil et al., 2007).

In this investigation, the macrostructure and poremorphology of the scaffolds before and after P(3HB)coating, produced following a previously developedprotocol (Bretcanu et al., 2007), were investigated, whilethe bioactivity of the P(3HB)–Bioglass composites wasanalysed by immersion in simulated body fluid (SBF).Moreover, the compressive strength of the scaffold wasmeasured before and after soaking in SBF. In vitrobiocompatibility was studied by using the HOS-TE85human osteosarcoma cell line. Osteoblasts are bone-forming cells which generate the bone matrix. HOS-TE85cells are osteoblast-like cells that are widely used forin vitro tests, as they show several characteristics ofprimary osteoblasts and are often used as experimentalmodels for investigating osteoblast function in relationto growth and differentiated cell function (Clover andGowan, 1994). Cell proliferation and differentiatedfunction were investigated using AlamarBlue assay,alkaline phosphatase assay and Gla-type osteocalcinassay, respectively.

Previous cell proliferation studies on Bioglass-derivedscaffolds using the MG63 cell line have shown that thesescaffolds have excellent cell supporting ability (Chenet al., 2008). In the present study we investigated for thefirst time the HOS cell response to the newly developedBioglass –P(3HB) composite scaffolds.

2. Experimental

2.1. Materials

45S5 Bioglass-derived glass–ceramic scaffolds werefabricated by the foam replication technique (Chen et al.,2006), by impregnation of polyurethane (PU) foamswith a 45S5 Bioglass slurry (NovaMin-US; particle size<5 µm). Two types of fully reticulated polyester-basedpolyurethane foams (Recticel, UK) with 60 pores/inch(p.p.i.) and 45 p.p.i. were used as templates for thereplication method. The samples were sintered at 1100 ◦Cfor 1 h. During heat treatment the glass phase partiallycrystallized, obtaining a glass–ceramic structure withenhanced mechanical properties, as reported elsewhere(Chen and Boccaccini, 2006a). In order to distinguish thetwo types of glass–ceramic scaffolds with two differentporosities, the scaffolds obtained from the 45 p.p.i. foamswere labelled W, while the scaffolds obtained from the 60p.p.i. foams were labelled B. The two labels derive fromthe colour of the starting (i.e. sacrificial) PU foams: white(W) and black (B).

The polymer P(3HB) was isolated from Bacillus cereusSPV by a fermentation process, as described in detailelsewhere (Valappil et al., 2007). The P(3HB) coatingsolution was prepared by dissolving 5 wt% P(3HB) inchloroform at room temperature.

P(3HB)–Bioglass composites were obtained byimmersing the 45S5 Bioglass-derived scaffolds intoP(3HB) solution for 15 min in a sealed glass container.Subsequently, the coated scaffolds were left to dry in air

Copyright 2009 John Wiley & Sons, Ltd. J Tissue Eng Regen Med 2009; 3: 139–148.DOI: 10.1002/term

In vitro biocompatibility of Bioglass-derived glass–ceramic scaffolds coated with [P(3HB)] 141

at room temperature in a fume cupboard. The coated sam-ples were labelled B-P(3HB) and W-P(3HB), dependingon the starting scaffold type (B or W).

2.2. Microstructural characterization

Microstructural characterization of the Bioglass-derivedglass–ceramic scaffolds before and after coating wasperformed by using X-ray diffraction (XRD) and scanningelectron microscopy (SEM) techniques.

X-ray diffraction patterns were registered by aPhilips diffractometer with Cu Kα radiation, for the2θ range 10–75◦, at 40 kV and 40 mA. The crystallinephase identification was performed by X’Pert HighScoresoftware, using the PCPDFWIN database. XRD patternswere registered on powder samples, obtained by grindingthe scaffolds using a pestle and mortar.

The 3D-porous structure of the scaffolds before andafter P(3HB) coating was evaluated by X-ray computedmicrotomography (micro-CT). The micro-CT experimentswere performed at the beam line SYRMEP of theSynchrotron Elettra (Trieste, Italy).

SEM was carried out using a JEOL JSM 840A instrumentcoupled with energy dispersive spectroscopy (EDS).Sample pieces were gold-coated and observed at anaccelerating voltage of 15 kV. The porosity of the scaffoldswas calculated from density measurements. The finalvalue represents the mean of five samples.

2.3. Measurement of surface tension

Surface tension was measured on disc samples using aCAM200 Optical Contact Angle Meter (KSV InstrumentsLtd). The discs were made of sintered Bioglass andhad a diameter of 2 cm. The Bioglass powder waspressed in a cylindrical die, using a uniaxial press, andthe produced pellets were sintered at 1100 ◦C for 1 h(using the same heat treatment as for the scaffolds). Aftersintering, a part of the Bioglass discs was coated withP(3HB) solution [5 wt% P(3HB) in chloroform] at roomtemperature for 15 min. The contact angle was measuredfor distilled water and diiodomethane. Before the contactangle measurements, the coated and uncoated Bioglass

pellets were first sterilized for 30 min using a UV sterilizer.

2.4. Development of bioactive layers in SBF

The formation of a bioactive layer was determined usingthe standard acellular in vitro procedure described byKokubo et al. (1990), as follows. Small prismatic pieces(5 × 6 × 10 mm3) of Bioglass-based glass–ceramic scaf-folds were immersed each in 10 ml acellular simulatedbody fluid (SBF) in conical flasks. The SBF solution wasbuffered at pH 7.25 at 37 ◦C before soaking the sam-ples. The conical flasks were placed in an incubator ata controlled temperature of 37 ◦C. The SBF solution wasrefreshed twice a week. Samples were removed from the

SBF solution after 3, 7, 14, 21 and 28 days. After extrac-tion from the SBF, the samples were gently rinsed withdeionized water and left to dry at room temperature. Theformation of a hydroxyapatite (HA) layer on the sam-ple surfaces was examined by SEM and XRD, using theinstruments mentioned above.

2.5. Compression strength test

The compression strength of foams before and after28 days of soaking in SBF was measured by a Zwick/RoellZ010 instrument, using a cell load of 1 kN and a crossheadspeed of 0.5 mm/min. The load was applied until acompression strain of 70% was reached. For each typeof scaffold, five samples were analysed and the meanof the results obtained. The dimensions of the prismaticspecimens for these experiments were 5 × 6 × 10 mm3.

2.6. In vitro biocompatibility assessment

2.6.1. HOS cells

The in vitro biocompatibility of the scaffolds wasinvestigated by using the osteoblast-like cell line HOS-TE85. The cells were thawed at 37 ◦C for 1 min, thentransferred to a tissue culture flask with 10 ml normalsupplemented medium [Dulbecco’s modified Eagle’smedium (DMEM), low glucose, with L-glutamine, 10vol% fetal calf serum and 1 vol% penicillin/streptomycinsolution]. The flasks were then incubated at 37 ◦C (5%CO2 and 95% air in a humidified atmosphere).

Small scaffold samples (1 × 1 × 0.5 cm3) were steril-ized in a UV sterilizer for 30 min, then passivated byimmersion in 1.5 ml normal supplemented medium andincubated for 24 h. UV sterilization was selected becauseit is an environmentally friendly method that typicallydestroys any living structures present on the materialsurface (i.e. bacteria, etc.) and also renders the surfaceless receptive to further contact from these components.Moreover, passivation of biomaterial surfaces is carriedout as a routine procedure, since proteins that adsorb ontothe scaffold surface from the medium might be helpful forcell differentiation and maturation of the osteoblasts.

Each scaffold was subsequently seeded with 100 µl cellsuspension containing 10 000 cells and incubated for 2 h,then 1.5 ml new medium was added to each well andthe plates with seeded scaffolds were incubated further.For control measurements, the cells were seeded directlyin the plastic wells (without scaffolds) at the same celldensity (10 000 cells/well) and then incubated for 2 h,then 1.5 ml new medium was added to each well and theplates were incubated further. The culture medium wasrenewed twice a week for all samples.

2.6.2. Cell morphology

The cell morphology after different culture periods wasdetermined using SEM analysis, in which the attached



cells were fixed with a solution of 3 vol% glutaraldehyde in0.1 M sodium cacodylate and kept refrigerated for 1 day.In order to avoid cell shrinkage by room temperaturedrying, all the samples were soaked twice in a series ofethanol solutions (50%, 70%, 90% and 100%) for 5 min,followed by immersion into a hexamethyldisilazanesolution for 2 min. The samples were left to dry at roomtemperature and further cut into two parts in order toobserve the cell morphology on the inner part of thescaffolds. Before SEM examination, the scaffolds wereprocessed by dehydrating and gold-coating, as describedin Section 2.2.

2.6.3. AlamarBlue assay

Cell proliferation was evaluated by using the AlamarBlue

assay (AbD Serotec: http://www.serotec.com), whichis based on an oxidation–reduction (REDOX) reaction,which indicates the cell’s metabolic activity. This blue dyechanges its colour to red (reduction reaction) when thereis intense cell growth activity. Reduction related to growthcauses the REDOX indicator to change from an oxidized(non-fluorescent, blue) form to a reduced (fluorescent,red) form. Continued growth maintains a reducedenvironment, while inhibition of growth maintains anoxidized environment (http://www.serotec.com).

Each measurement was performed in triplicate, usingthree different samples. The fluorescence was readafter 1, 3 and 7 days, using a Fluoroskan Ascentinstrument (Labsystems). Each reading was performedtwice. Fluorescence was monitored at 590 nm emissionwavelength (530 nm excitation wavelength). At each timepoint, 150 µl AlamarBlue dye (10%vol) was added to allwells. The well-plates were then put in the incubator.After 4 h, 200 µl medium (containing the blue dye) wastransferred to a 96-well plate for the fluorescence reading.Cells were returned to their culture environment and theculture medium was renewed twice a week and after eachreading.

2.6.4. Alkaline phosphatase assay

This assay is designed to measure ALP activitydirectly in biological samples without pre-treatment(http://www.bioassaysys.com/DALP.pdf). Osteoblastsare very rich in alkaline phosphatase, which participatesin the mineralization process. Thus, an increase in boneformation gives rise to an increase in ALP concentration.The ALP kit (http://www.bioassaysys.com/DALP.pdf) isbased on a colorimetric reaction. The ALP enzyme cataly-ses the hydrolysis of p-nitrophenyl phosphate (p-NPP) intoa p-nitrophenol (p-NP) and orthophosphate in an alkalineenvironment, with a change in colour from transparent toyellow (p-NP). The rate of the reaction is directly propor-tional to the enzyme activity. Increased enzyme activitygives rise to an increase in colour intensity. Nitrophenolexhibits a maximal absorbance at 405 nm, therefore therate of formation of the yellow colour of p-NP, produced

by hydrolysis of p-NPP in alkaline solution, was measuredspectrophotometrically at 405 nm and 37 ◦C, by using aDynex Technologies spectrophotometer.

Triplicate measurements for each sample were per-formed. Three different specimens per scaffold type wereextracted from each well after 1, 7 and 14 days, storedat −20 ◦C and subsequently analysed. Each reading wasperformed in duplicate. Two different types of mediumwere used for all samples: normal supplemented medium(1 g/l DMEM low glucose with L-glutamine, 10 vol% fetalcalf serum and 1 vol% penicillin/streptomycin solution)and osteogenic medium (normal supplemented mediumwith 50 µg/ml ascorbic acid, 10 mM β-glycerophosphateand 1 × 10−8 M dexamethasone).

2.6.5. Gla-type Osteocalcin assay

The Gla-type Osteocalcin (Gla-OC) EIA kit (Cambrex,UK) is an in vitro enzyme immunoassay (EIA) kit forquantitative determination of Gla-OC in cell culturesupernatants and other biological fluids. Osteocalcin(OC), also known as bone γ -carboxyglutamic acidprotein, is a vitamin K-dependent Ca2+-binding proteinwhich has three carboxylated glutamic acid (Gla)residues. These radicals mediate strong binding ofOC to hydroxyapatite. Because of this tissue-specificexpression, the level of OC could be considered as anindicator of the overall activity of cells operating inbone formation. OC constitutes about 15% of the bonematrix proteins and is produced exclusively in osteoblastsand odontoblasts (http://www.takara-bio.com). Thus, anincreased bone formation gives rise to an increase of theOC concentration.

The amount of Gla-OC present in the sampleswas determined by absorbance measurements of theunknown sample and standard solutions (with knownOC concentrations). The unknown OC concentration ofthe samples was calculated indirectly, using a standardcurve. All the measurements were performed in duplicate(using two different samples) using a Dynex Technologiesspectrophotometer. The absorbance was measured at450 nm wavelength and each reading was performedtwice. The two types of medium (normal, −; andosteogenic, +) were used for all samples. The absorbancewas read for samples from 1, 7 and 14 days. For the OCmeasurements, 100 µl medium was removed from eachwell, placed in the 96-well plate and finally analysed usingthe OC kit.

2.6.6. Statistical analysis

All data are expressed as mean ± standard deviation(SD). The data were compared using Student’s t-test anddifferences were considered significant at p < 0.05. A pvalue higher than 0.05 (p > 0.05) was taken as indicatingno significant difference.



Figure 1. XRD patterns of a W-type glass–ceramic scaffold,before and after P(3HB) coating. �, Na2Ca2Si3O9; ž, P(3HB).Similar crystallization behaviour was observed in B-typescaffolds (Bretcanu et al., 2007)

3. Results and discussion

3.1. Microstructural characterization

Typical XRD patterns of a sintered W scaffold (derivedfrom the 45 p.p.i. PU foam) before and after P(3HB)coating are shown in Figure 1. As can be seen, thetwo patterns are quite similar, which is probably dueto the thin polymeric coating layer deposited. Thediffraction patterns are characteristic of a glass–ceramicstructure. The main crystalline phase is identified asNa2Ca2Si3O9, in agreement with our previous results(Chen et al., 2006). Under optimum sintering conditions,nearly full densification of the foam struts occurredand fine crystals of mixed sodium calcium silicate

formed. These characteristics confer to the achievedscaffolds a reasonable compressive and flexural strength,as discussed elsewhere (Chen and Boccaccini, 2006a).Similar results were obtained for B-type scaffolds.

SEM micrographs of the internal structure of aB-type glass–ceramic scaffold, before and after coatingwith P(3HB), are presented in Figure 2a, b, respectively,while Figure 2c shows the 3D pore structure of anuncoated W-type scaffold. The samples show a highlyporous structure with interconnected porosity. It canbe observed that the interconnected porosity is alsomaintained after polymer coating. As can be seen inFigure 2b, the coating structure is not homogeneous.The presence of small uncoated regions exposingthe glass–ceramic surface can have an importantrole in influencing the bioactivity mechanism (asdiscussed below). Similar SEM images were obtained forP(3HB)-coated W-type scaffolds, as presented elsewhere(Bretcanu et al., 2007).

X-ray Micro-CT images of B-scaffolds before and aftercoating with P(3HB) solution are illustrated in Figure 3.The images show the scaffold material (green) and thecoating (blue). As can be observed, the coating seemsto be homogeneously distributed inside the scaffold.The P(3HB) coating did not change the pore structuresignificantly: the porosity is still open and interconnected.The porosity of the W-scaffolds is approximately 85 ±2% (mean pore size, 470 µm), while the porosity ofB-scaffolds (derived from the 60 p.p.i. PU foam) is80 ± 2% (mean pore size, 350 µm). The B-foam, havinga smaller pore size in comparison with the W-foam, alsodisplays a lower porosity. The adhesion strength of the

(c)

(a) (b)

Figure 2. SEM micrographs of: (a) B-type glass–ceramic scaffold before P(3HB) coating; (b) after P(3HB) coating (arrows indicateareas of uniform coating); (c) W-type scaffold before coating



(a) (b)

Figure 3. X-ray Micro-CT of B-type scaffolds, (a) before and (b) after coating with P(3HB)

0

10

20

30

40

50

60

70

80

uncoated P(3HB)-coated

Sur

face

tens

ion

(mN

/m)

Figure 4. Surface tension values of Bioglass pellets sinteredunder the same conditions of the scaffolds, before and aftercoating with P(3HB)

polymer to the scaffold substrate was not investigatedhere, as it was considered in a previous study (Bretcanuet al., 2007).

3.2. Measurement of the surface tension

The surface tension data of the Bioglass sintered pelletsbefore and after coating with P(3HB) solution are shownin Figure 4. It is evident that the uncoated pellets are morehydrophilic, exhibiting a relatively high surface tension,while the P(3HB)-coated pellets have a more hydrophobicsurface. The surface tension is likely to affect the cellattachment, migration and spreading on the scaffolds(Boyan et al., 1996).

3.3. Bioactivity assessment

SEM images of B-type composite scaffolds, before andafter coating with P(3HB) solution, after soaking for28 days in SBF are shown in Figure 5. These resultsare similar to those on W-type scaffolds. Results onW-type scaffolds and at lower immersion time pointsin SBF have been reporter elsewhere (Bretcanu et al.,

2007), where the formation of HA on the scaffolds wasidentified by EDS and XRD analysis. Figure 5 indicatesthat the scaffold surface is homogeneously coated by ahydroxyapatite (HA) layer, in agreement with previousresults (Chen et al., 2006). It is evident that after suchrelatively short time in SBF, P(3HB)-coated scaffolds stillretained a thin polymer layer on their surfaces (e.g.arrows in Figure 5b). It is suggested that the polymerstarts to resorb and the thickness of the polymeric layerdecreases. As mentioned above (see also Figure 2b), theP(3HB) coating is not homogeneous and the uncoatedregions of the composite scaffold can come into directcontact with the SBF solution. Probably HA nucleates firston the uncoated region of the struts and after that HAcrystals start growing, expanding also onto the P(3HB)-coated regions. It is assumed that the mechanism ofHA formation is based on the ionic exchange betweenthe scaffold surface and the SBF solution, as proposedfor bioactive glasses (Hench, 2006). In the first stage,a silica-rich layer is formed on the scaffold surface,followed by nucleation and growth of the HA crystalson the top of this silica-rich layer. In Figure 5a, the crackspresent on the glass–ceramic struts correspond to thedried silica-rich layer. The ionic exchange will initiallytake place between the uncoated glass–ceramic and SBF,leading subsequently to the formation of the HA-richlayer. Therefore, the presence of these small uncoatedregions of the scaffold struts favours the precipitationof HA and hence the bioactivity of the scaffolds is notimpaired by the P(3HB) coating.

3.4. Compression strength test

Typical compressive stress–strain curves of uncoatedand coated B-type scaffolds, before and after soakingin SBF solution for 28 days, are shown in Figure 6. Thearea under the stress–strain curve represents the workof fracture and quantifies the ability of a material toresist fracture. The mean values of the work of fracture(obtained as the average of five samples) are plottedin Figure 7. It is clear that the work of fracture of the



(a) (b)

Figure 5. SEM micrographs of B-type glass–ceramic scaffolds after 28 days in SBF, (a) before and (b) after P(3HB) coating (arrowsindicate areas coated by the polymer)

0.00

0.10

0.20

0.30

0.40

0.50

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Strain

Str

ess

(MP

a)

BB-P(3HB)B-SBFB-P(3HB)-SBF

Figure 6. Compressive stress–strain diagrams of uncoated andcoated B-type scaffolds, before and after soaking in SBF solutionfor 28 days

0

50

100

150

200

250

300

B B-P(3HB) B-SBF B-P(3HB)-SBF

Wor

k of

frac

ture

(kJ

/m3 )

Figure 7. Work of fracture (kJ/m3) of uncoated and coatedB-type scaffolds, before and after soaking in SBF solutionfor 28 days. Values obtained from stress–strain diagrams (seeFigure 6)

scaffolds is substantially increased after coating withP(3HB). It is suggested that the polymer layer coversand fills the microcracks situated on the strut surfaces,improving the mechanical stability of the scaffold (Yunoset al., 2008). After 28 days in SBF the work of fracturehas decreased, due to the degradation of the scaffoldassociated with the dissolution of the bioactive glasssurface layer and formation of HA. The concept of coatingbrittle bioceramics and bioactive glass scaffolds withbiodegradable polymers is being increasingly explored todevelop scaffolds with enhanced fracture properties and

desired structural integrity for bone tissue engineering(Yunos et al., 2008; Chen and Boccaccini, 2006b; Huangand Miao, 2007; Kim et al., 2004; Peroglio et al., 2007).This approach is inspired by the fact that nearly 60 wt%of bone is constituted of an inorganic phase (HA) andthe rest is the organic phase (collagen) and water. Thefracture behaviour of mineralized tissues such as bone isinfluenced by the optimal interaction of the inorganic andorganic phases, and the toughening mechanisms inducedby the presence of collagen fibrils in bone have beenstudied (Nalla et al., 2003). The addition of a polymerphase to a porous ceramic scaffold not only enhancesthe fracture toughness of the composite but also allowsthe functionalization of the scaffold surface to induceenhanced bioactivity, as reviewed elsewhere (Yunos et al.,2008).

3.5. In vitro biocompatibility assessment

SEM micrographs of the inner part of uncoated andP(3HB)-coated composite scaffolds (W-type) after 1, 3and 7 days of cell culture are illustrated in Figure 8a–f.Qualitatively similar results were obtained on B-typescaffolds, but the detailed SEM investigation was carriedout only on W-type scaffolds. It was anticipated thatdifferent scaffold porosity (since B- and W-type scaffoldshave exactly the same chemistry and surface topography)would not affect the surface interactions with cells.The HOS cells are seen to adhere and spread on thescaffold surfaces. With longer culture periods, the cellsgrow steadily on the scaffold. Clear images of thecells’ filopodia, which spread on sample surfaces, canbe seen in Figure 8b, f. In contact with the medium,calcium phosphate precipitates and forms small roundagglomerates (due to ionic exchange between the scaffoldsurface and the liquid medium). The precipitation is moreintense on the uncoated samples (see Figure 8b, c). In thecoated samples, the polymer acts like a buffer, possiblylowering the solution pH. It seems also that cells spreadmore on the coated samples (see Figure 8f).

The coated scaffolds have different surface topogra-phy/roughness and the cells seem to prefer the smoothersurface at the earlier stages of culture. As can be seen in



(a) W-1day (d) W-P(3HB)-1day

(b) W-3 days (e) W-P(3HB)-3 days

(c) W-7 days (f) W-P(3HB)-7 days

Figure 8. SEM images of HOS cells on uncoated and P(3HB)-coated [W-P(3HB)] scaffolds after 1, 3 and 7 days of cell culture

Figure 8d, the rougher surface of the polymer layer is notcompletely covered by cells after 3 days of incubation.The cells grow only on the smooth part of the polymer.However, after 7 days of incubation the cells also coverthe roughened areas (Figure 8f).

A histogram representing cell proliferation and growthon and inside the scaffolds is shown in Figure 9. Thecells proliferated and grew on all samples. In contrastwith the control surface (tissue culture plastic), thenumber of cells that proliferate on the scaffolds in allcases, e.g. W- and B-type, with and without coating,is lower. This fact can be partly explained in termsof pH. The scaffolds were not pre-conditioned in SBFbefore the cells were seeded and the ionic exchangebetween the scaffolds surface and the medium may haveinduced a pH increase, which could have affected cellproliferation and viability. Considering both the uncoatedand P(3HB)-coated samples, the fluorescence values arehigher for the coated samples (for which the presenceof the polymer should decrease the pH). Although thenumber of samples did not allow a robust statisticalassessment, the results do indicate that at day 7 the

coated samples exhibited enhanced cell proliferation andviability. Taking into consideration the different pore sizescaffolds (W- and B-types), the highest values of thefluorescence were obtained for the W-type scaffolds. TheB-type samples have a higher specific surface area thanthe W-type samples; it is therefore suggested that in thesescaffolds the ionic exchange is more intense and thisshould produce higher pH values.

ALP was measured using a commercially available assaykit, which provides a convenient calorimetric assay fordetecting ALP with a phosphatase-conjugated secondaryantibody by using pNPP as a phosphatase substrate. Upondephosphorylation by phosphatase, pNPP turns yellowand can be detected at absorbance of 405 nm. The sampleresults are then compared against a standard, as providedwithin the assay kit. The level of ALP activity present in thesamples is shown in Figure 10. After 14 days of incubationthe ALP intensity increased in all samples. It seems thatthe samples with osteogenic medium (+) have a slightlyhigher level of ALP, as expected, compared to the sampleswhich had not been treated with osteogenic medium (−).It must be noted, however, that there is no significant



0

5

10

15

20

25

30

35

40

45

50

1 day 3 days 7 days

Flu

ores

cenc

e in

tens

ity (

a.u.

) W

W-P(3HB)

B

B-P(3HB)

CTRL

Figure 9. Cell proliferation study for 1, 7 and 14 days, using AlamarBlue assay on P(3HB)-coated and -uncoated samples ofdifferent porosities (W = 45 p.p.i.; B = 60 p.p.i.). The data are represented as mean ± SD

0.00

0.02

0.04

0.06

0.08

0.10

1 day 14 days

Abs

orba

nce

(at 4

05 n

m)

-W

+W

-W-P(3HB)

+W-P(3HB)

-B

+B

-B-P(3HB)

+B-P(3HB)

Figure 10. Alkaline phosphatase activity of cells cultured over 14 days on P(3HB)-coated and -uncoated samples of differentporosities. The data are represented as mean ± SD

statistical difference between the samples. The uncoatedsamples have ALP values higher than coated samplesafter 14 days of incubation. This fact can be explainedby surface reactivity. The uncoated samples have a moreintense ionic exchange between the glass–ceramic surfaceand the medium and therefore the precipitation of calciumphosphate species is increased. The uncoated samplesare richer in HA, in contrast to coated samples. Thehighest values of absorbance were obtained for W-typesamples. The B-type samples have a higher specific surfacearea relative to W-type samples and the ion exchange(and therefore the HA precipitation) is facilitated. Atthe same time, the higher pH levels should decreasethe cell proliferation. Therefore these two factors; i.e.precipitation of HA and increased pH levels, have anantagonistic effect on cell growth.

The OC activity determined for the composite scaffoldsis shown in Figure 11. It seems that the OC activity forall samples increases during the 2 weeks of incubation.These results are similar to those obtained for the ALPactivity. All samples containing osteogenic medium (+)had a slightly higher level of OC at 7 and 14 days ofincubation in comparison with non-treated samples (−).It is also interesting to note that at 14 days almost allscaffolds (except B-type) have OC intensity higher thanthe control samples. Probably the B-type scaffolds need

to be pre-conditioned in SBF solution before incubation,in order to avoid the higher increase of pH (due totheir higher specific surface area). The relative increaseof the level of OC for the scaffolds at day 14, incomparison to the control, might be related to thepresence of ionic dissolution products in the medium(silica, calcium ions), which is in agreement with resultsin the literature on osteoblast activity in contact withbioactive glass dissolution products (Xynos et al., 2000).It should be noted also that the 14 day culture period fordifferentiation is standard and widely accepted to notechanges in the differential expression of bone-associatedproteins such as ALP and OC as markers of differentiation.

4. Conclusions

3D, highly porous, composite scaffolds were preparedby coating 45S5 Bioglass-derived glass–ceramic foamswith bacteria-derived P(3HB) polymer. The glass–ceramicscaffolds were synthesized using the foam replicationtechnique. The polymer coating did not affect theinterconnectivity of the pore structure and had a positiveeffect on the compression strength and structural integrityof the scaffolds. Another advantage of the polymer coatingwas confirmed to be the decrease of the pH values of the



0

5

10

15

20

25

30

35

1 day 7 days 14 days

Gla

-Ost

eoca

lcin

(ng

)

W-

W+

W-P(3HB)-

W-P(3HB)+

B-

B+

B-P(3HB)-

B-P(3HB)+

CTRL-

CTRL+

Figure 11. Osteocalcin activity of cells cultured over 14 days on P(3HB)-coated and -uncoated samples of different porosities. Thedata are represented as mean ± SD

solution after cell culture, which resulted in increasingcell proliferation. In vitro experiments showed that thecomposite scaffolds were biocompatible for HOS cellsand that cells penetrated the 3D structure, althoughthis aspect was not investigated in detail. The resultsof this investigation contribute toward the establishmentof Bioglass-based glass–ceramic foams as the scaffolds ofchoice for bone tissue engineering when their mechanicalproperties are enhanced by polymer coatings. Furtherwork will encompass methods to investigate longer-termcultures and issues of nutrient and/or oxygen limitswill need to be addressed in the deeper areas of thescaffolds.

Acknowledgements

The authors thank Dr Alejandro Gorustovich (CONICET,Argentina) for helpful discussions and Dr Wojtek Chrzanowskifor the contact angle measurements. This work was fundedby an EU IntraEuropean Marie Curie Fellowship (No. 24248).Partial financial support from the EU Network of Excellence,‘Knowledge-based Multicomponent Materials for Durable andSafe Performance’ (No. KMM-NoE, NMP3-CT-2004-502243),is acknowledged. We also acknowledge the first TERMIS–EUSummer School for financial support.

References

Boyan BD, Hummert TW, Dean DD, et al. 1996; Role of materialsurfaces in regulating bone and cartilage cell response.Biomaterials 17: 137–146.

Bretcanu O, Chen QZ, Misra SK, et al. 2007; Biodegradable polymer-coated 45S5 Bioglass-derived glass–ceramic scaffolds for bonetissue engineering. Glass Technol Eur J Glass Sci Technol A 48:227–234.

Chen QZ, Boccaccini AR. 2006; Coupling mechanical competenceand bioresorbability in Bioglass-derived tissue engineeringscaffolds. Adv Eng Mater 8: 285–289.

Chen QZ, Boccaccini AR. 2006; Poly(D,L-lactic acid)-coated 45S5Bioglass-based scaffolds: processing and characterization. JBiomed Mater Res 77A: 445–457.

Chen QZ, Efthymiou A, Salih V, et al. 2008; Bioglass-derivedglass–ceramic scaffolds: study of cell proliferation and scaffolddegradation in vitro. J Biomed Mater Res 84A: 1049–1060.

Chen QZ, Thompson ID, Boccaccini AR. 2006; 45S5 Bioglass-derived glass–ceramic scaffolds for bone tissue engineering.Biomaterials 27: 2414–2425.

Clover J, Gowen M. 1994; Are MG-63 and HOS TE85 humanosteosarcoma cell lines representative models of the osteoblasticphenotype? Bone 15: 585–591.

Day RM, Boccaccini AR, Shurey S, et al. 2004; Assessment ofpolyglycolic acid mesh and bioactive glass for soft-tissueengineering scaffolds. Biomaterials 25: 5857–5866.

Doi Y, Steinbuchel A. 2000; Biopolymers, vol 4 – Polyesters III.Wiley–VCH: Weinheim.

Doyle C, Tanner ET, Bonfield W. 1991; In vitro and in vivo evaluationof polyhydroxybutyrate and of polyhydroxybutyrate reinforcedwith hydroxyapatite. Biomaterials 12: 841–847.

Guarino V, Causa F, Ambrosio L. 2007; Bioactive scaffolds for boneand ligament tissue. Expert Rev. Med Devices 4(3): 405–418.

Hench LL, Polak JM. 2002; Third-generation biomedical materials.Science 295: 1014–1017.

Hench LL. 2006; The story of Bioglass. J Mater Sci Mater Med 17:967–978.

Huang X, Miao X. 2007; Novel porous hydroxyapatite prepared bycombining H2O2 foaming with PU sponge and modified with PLGAand bioactive glass. J Biomater Appl 21(4): 351–375.

Lanza RP, Langer R, Vacanti J. 2000; Principles of Tissue Engineering.Academic Press: New York.

Karageorgiou V, Kaplan D. 2005; Porosity of 3D biomaterial scaffoldsand osteogenesis. Biomaterials 26: 5474–5491.

Kim HW, Knowles JC, Kim HE. 2004; Hydroxyapatite/poly(ε-caprolactone) composite coatings on hydroxyapatite porous bonescaffold for drug delivery. Biomaterials 25: 1279–1287.

Kokubo T, Ito S, Huang ZT, et al. 1990; Ca,P-rich layer formed onhigh-strength bioactive glass–ceramic A–W. J Biomed Mater Res24: 331–343.

Nalla RK, Kinney JH, Ritchie RO. 2003; Mechanistic fracture criteriafor failure of human cortical bone. Nat Mater 2: 164–169.

Peroglio M, Gremillard L, Chevalier J, et al. 2007; Toughening ofbioceramics scaffolds by polymer coating. J Eur Ceram Soc 27:2679–2685.

Ratner BD, Hoffman AS, Schoen FJ, et al. 2004; Biomaterials Science:An Introduction to Materials in Medicine. Elsevier Academic Press:Amsterdam.

Rezwan K, Chen QZ, Blaker JJ, et al. 2006; Biodegradable andbioactive porous polymer/inorganic composite scaffolds for bonetissue engineering. Biomaterials 27: 3413–3431.

Valappil SP, Peiris D, Langley GJ, et al. 2007; Polyhydroxyalkanoate(PHA) biosynthesis from structurally unrelated carbon sources bya newly characterized Bacillus sp. J Biotechnol 127: 475–487.

Xynos ID, Hukkanen MV, Batten JJ, et al. 2000; Bioglass 45S5stimulates osteoblast turnover and enhances bone formationin vitro: implications and applications for bone tissue engineering.Calc Tissue Int 67(4): 321–329.

Yunos DM, Bretcanu O, Boccaccini AR. 2008; Polymer–bioceramiccomposites for tissue engineering scaffolds. J Mater Sci 43:4433–4442.

Zhang K, Wang Y, Hillmayer MA, et al. 2004; Processing andproperties of porous poly(L-lactide)/bioactive glass composites.Biomaterials 25: 2489–2500.


In vitro biocompatibility of 45S5 Bioglass ® -derived glass-ceramic scaffolds coated with...

Documents

Transcript of In vitro biocompatibility of 45S5 Bioglass ® -derived glass-ceramic scaffolds coated with...