Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) composite ......proposed for use in bone tissue engineering1 owing to their favorable degradation rates2–4 and nontoxic de-gradation

Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) compositebiomaterials for bone tissue regeneration: In vitroperformance assessed by osteoblast proliferation,osteoclast adhesion and resorption, and macrophageproinflammatory response

S. M. Cool,1,2,3 B. Kenny,1,4,5 A. Wu,4,5 V. Nurcombe,1,2,3 M. Trau,4,6 A. I. Cassady,5 L. Grøndahl4,61School of Biomedical Sciences, University of Queensland, Queensland 4072, Australia2Laboratory of Stem Cells and Tissue Repair, Institute of Molecular and Cell Biology, Proteos, 61 Biopolis Drive,Singapore 1386733Department of Orthopaedic Surgery, Yong Loo Lin School of Medicine, National University of Singapore,Singapore 1190744Centre for Nanotechnology and Biomaterials, Australian Institute for Bioengineering and Nanotechnology,The University of Queensland, Queensland 4072, Australia5Institute for Molecular Bioscience and Co-operative Research Centre for Chronic Inflammatory Diseases,University of Queensland, Queensland 4072, Australia6School of Molecular and Microbial Sciences, University of Queensland, Queensland 4072, Australia

Received 11 May 2006; revised 27 July 2006; accepted 2 November 2006Published online 21 February 2007 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.a.31174

Abstract: The efficacy of composite materials for bone tis-sue engineering is dependent on the materials’ ability tosupport bone regeneration whilst inducing a minimalinflammatory response. In this study we examined thein vitro osteogenic and inflammatory properties of poly(3-hydroxybutyrate-co-3-valerate) (PHBV) with various cal-cium phosphate-reinforcing phases: nano-sized hydroxy-apatite (HA); submicron-sized calcined hydroxyapatite(cHA); and submicron-sized b-tricalcium phosphate (b-TCP), using bioassays of cultured osteoblasts, osteoclasts,and macrophages. Our study showed that the addition ofa nano-sized reinforcing phase to PHBV, whilst improvingosteogenic properties, also reduces the proinflammatoryresponse. Proinflammatory responses of RAW264.7/ELAM-eGFP macrophages to PHBV were shown to bemarkedly reduced by the introduction of a reinforcingphase, with HA/PHBV composites having the lowest

inflammatory response. Osteoclasts, whilst able to attach toall the materials, failed to form functional actin rings orresorption pits on any of the materials under investigation.Cultures of osteoblasts (MC3T3-E1) readily attached andmineralised on all the materials, with HA/PHBV inducingthe highest levels of mineralization. The improved biologi-cal performance of HA/PHBV composites when comparedwith cHA/PHBV and b-TCP/PHBV composites is mostlikely a result of the nano-sized reinforcing phase of HA/PHBV and the greater surface presentation of mineralin these composites. Our results provide a new strategyfor improving the suitability of PHBV-based materials forbone tissue regeneration. � 2007 Wiley Periodicals, Inc.J Biomed Mater Res 82A: 599–610, 2007

Key words: bone tissue engineering; PHBV; inflammatoryresponse; osteogenesis; biocompatibility

INTRODUCTION

Materials designed for bone tissue engineeringneed to have suitable mechanical properties to ensuretheir utility. In addition, the biological responsemust be optimal both with respect to encouragingbone remodeling, and with respect to inhibitingexhaustive proinflammation. Thus, the challenge isto design a material that will evoke a specific bio-logical response.

Correspondence to: S. M. Cool; e-mail: [email protected] grant sponsor: Wesley Research Institute,

Brisbane, Australia; contract grant number: 2000100294Contract grant sponsor: Australian Research Council;

contract grant numbers: DP0209873, DP0343547

' 2007 Wiley Periodicals, Inc.

The bacterially-derived biodegradable polyesterspoly(3-hydroxybutyrate) (PHB) and poly(3-hydroxy-butyrate-co-3-hydroxyvalarate) (PHBV) have beenproposed for use in bone tissue engineering1 owing totheir favorable degradation rates2–4 and nontoxic de-gradation products.5 PHB degrades in vivo forming3-hydroxybutyrate, a natural component of blood,2

although its usefulness is limited by its brittleness. Theaddition of varying ratios of polyhydroxyvalerate toform the copolymer PHBV, however, has been shownto improve both the ductility and processability6 andhence its biomedical applicability.2 Importantly, whenPHB(V) (PHB or PHBV) is implanted subcutaneouslyin mice, no acute inflammation, abscess formation, ortissue necrosis is observed.2

Addition of a reinforcing calcium phosphate phaseinto PHB(V) materials should potentially have thedual effect of improving both the bioactivity andmechanical properties. Bone integration of PHB(V)with the addition of micrometer-sized reinforcingphases has been demonstrated in vivo with a layer ofbone-like apatite forming at the HA/PHB(V) implantinterface after implantation into both bone tissue7,8

and in vitro for HA/PHBV and b-tricalcium phos-phate (b-TCP)/PPHBV composite materials.9 Nota-bly, none of these studies compared the bioactivityto that of the PHB(V) polymer alone. However, aseparate in vivo study on HA/PHB failed to detectimproved bioactivity when compared with PHB.10

Furthermore, PHB(V) matrices with no reinforcingphase have been shown to support the in vitroattachment, growth and differentiation of a range ofcell types, including fibroblasts, epithelium, endothe-lium, hepatocytes, and osteoblasts.1,11–15

Bone repair is an ongoing process that incorpo-rates the normal bone and mineral homeostaticresponse to tissue wear. The ideal implant for bonetissue engineering should be able to support thesame repair mechanisms. As such, it would need tobe resorbed by osteoclasts whilst at the same timestimulate osteoblastogenesis and matrix formation.In addition, the material’s proinflammatory responseshould be minimal so as to reduce the risk of local-ized osteolysis. Yet to date, little is known aboutthe responses of osteoblasts, osteoclasts, and macro-phages to particulate-reinforced PHBV.

We aimed to evaluate the in vitro biological res-ponse to composite materials of PHBV, in which thereinforcing mineral phase of calcium phosphateremains nano-sized after processing16 when com-pared with the composites where the reinforcingphase is submicron sized. In this study nano- andsubmicron-sized bioactive calcium phosphate miner-als, hydroxyapatite (HA), calcined hydroxyapatite(cHA), and b-TCP were incorporated into PHBV andthe in vitro responses of osteoblasts, osteoclasts, andmacrophages determined.

MATERIALS AND METHODS

Materials

Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) with a hy-droxyvalerate content of 8% (Goodfellow UK) was repreci-pitated from dimethylformamide (DMF) by the addition ofmethanol (MeOH) before use. Poly(acrylic acid) (PAA) ofa molecular weight of 2000 g mol�1 was purchased fromAldrich. b-TCP was purchased from Fulka, Ca(NO3)2�4H2O from Riedel de Haen, (NH4)2�HPO4 from Univar,and concentrated (25%) ammonia solution (AnalaR) fromMerck. Human soluble recombinant RANKL was obtainedfrom Peprotech and human recombinant M-CSF was a giftfrom Chiron.

Synthesis of HA and cHA

HA was synthesized using a modification of the methoddescribed previously.17 Briefly; Ca(NO3)2�4H2O (23.5 g)was dissolved in water (255 mL) at 408C. (NH4)2�HPO4

(7.9 g) was dissolved in water (390 mL) and added to thecalcium solution in two aliquots. Subsequently concentratedammonia solution was added resulting in a solution witha pH of 10–11. The resulting solution was stirred for 1 h at608C, allowed to cool to ambient temperature, and agedovernight without stirring before isolation and washing ofthe HA crystals by vacuum filtration. Carbon analysis(CarloErba elemental analyzer) was performed on allmaterials to calculate the w/w composition of the compos-ite materials. The HA powder yielded a value of 4.13%.cHA was prepared as previously described17 by heatingHA in a porcelain crucible to 9008C for 2 h in a Eurothermoven. Carbon analysis of the cHA and b-TCP powdersboth yielded 0.0%.

Synthesis of HA/PHBV composite material

A suspension of HA prepared as earlier without agingwas added to PAA (1.5 g, 15% by weight of HA) in water(100 mL). The resulting suspension was reduced to a vol-ume of 50 mL by heating to 1008C. PHBV (40 g) dissolvedin DMF (3 L) at 808C was slowly added the aqueousPAA/HA suspension (cooled to 808C) under stirring. ThePHBV/HA/PAA mixture was stirred energetically andcooled rapidly, using an ice bath until gelation occurred.Excess MeOH was added to the gel, which caused the pre-cipitation of a fine white composite powder that was iso-lated by vacuum filtration and washed with MeOH. Thecomposite was subsequently subjected to several MeOH-washing steps to remove any residual DMF. Carbon analy-sis of batch 1 of the HA/PHBV composite yielded 43.8%C, of batch 2 of HA/PHBV yielded 47.7%, and that ofpure PHBV yielded 55.57% C.

Synthesis of cHA/PHBV and b-TCP/PHBVcomposite materials

Powdered b-TCP (10 g) or cHA (10 g) were suspendedin water (100 mL) before a PAA (1.5 g, 15% by weight)

600 COOL ET AL.

Journal of Biomedical Materials Research Part A DOI 10.1002/jbm.a

solution (100 mL) was added. The volume of the solutionwas subsequently reduced to 50 mL and added to pre-dissolved PHBV (40 g) in DMF (3 L) at 808C. Powderformation was achieved as earlier. Carbon analysis ofthe b-TCP/PHBV composite yielded 44.2%, and of thecHA/PHBV composite yielded 44.1%.

Processing of materials

Two processing methods were employed; melt process-ing and solvent casting. For the melt processed samplesall composites (batch 1 of the HA/PHBV composite) andpure polymer was used. Polymer powder and compositepowders were processed under pressure in vacuum in acustom-designed metal mould and heating jacket in amechanical press. The temperature was controlled, usinga Eurotherm temperature controller attached to a thermo-couple. The samples were compressed at room tempera-ture under vacuum, then heated to 1708C, and held for2 min before being rapidly cooled. HA and b-TCP werepelleted under pressure in a standard IR press. Both proce-dures resulted in round material disks with a diameter of12.5 mm and a thickness of 1–2 mm. For the solvent castfilms, PHBV and HA/PHBV composite (batch 2) wereused. Powders were cast from chloroform to producethin films by dissolving/dispersing the powder (0.3 g) in10–15 mL of HPLC grade chloroform (LAB-SCAN, Ireland)at 408C, after which the solution/dispersion was pouredinto a 52-mm glass Petri dish, covered with a lid, and leftin a fume cupboard to allow the chloroform to evaporateslowly.

Characterization of material powders

The minerals (HA, cHA, and b-TCP), PHBV polymerand composite powders were analyzed by X-ray diffraction(XRD) on a Bruker AXS X-ray diffractometer, using a cop-per Ka radiation source. Samples were examined from 108to 758 2h at a scanning speed of 3 s per step increment of0.058 2h. Voltage and current were set at 40 kV and40 mA, respectively. Readouts were analyzed with a PDFMaintEx Library 2001 (minerals, version 7.0.108) searchengine, thereby identifying each mineral species both inpure mineral samples and in composite materials. Diluteethanol suspensions of HA, cHA, and b-TCP (1 mg/mL)were evaporated onto transmission electron microscopy(TEM) grids precoated with a thin Formvar-carbon supportfilm. The grids were left to air-dry and mounted onto aJEOL 1010 TEM with a setting of 80 KeV. Samples werethen imaged and particle size determined.

Characterization of processed materials

Evaluation of the chemical surface composition wasachieved with the use of a Perkin–Elmer PHI model 560X-ray photoelectron spectrometer with a double-pass cylin-drical mirror analyzer and a vacuum system giving a basepressure of *10–9 Torr. X-ray photoelectron spectroscopy(XPS) survey (0.5 eV resolution) and multiplex (0.1 eV

resolution) scans were recorded of an area with a diameterof 4 mm and to a depth of 40 A. X-rays were generatedfrom an Mg Ka source (1253.6 eV).

Cell culture: Macrophages

The RAW264.7/ELAM-eGFP cell line stably transfectedwith the ELAM-eGFP expression plasmid was obtainedfrom Dr. D. Sester (Institute for Molecular Bioscience, Uni-versity of Queensland). The E-selectin (ELAM) promoter isinduced by the activation of the NF-kappa B transcriptionfactor, that is, strongly associated with a proinflammatorystimuli. This results in the expression of the enhancedgreen fluorescent gene (eGFP).18 RAW264.7/ELAM-eGFPcells were used as a sensitive in vitro assay by fluores-cence-activated cell sorting (FACS) of the induction of aproinflammatory response in this macrophage cell line bythe biomaterials under investigation.19 The cells were cul-tured in RPMI-1640 media (Phenol Red-free) containing10% heat-inactivated fetal calf serum (FCS), 2 mM Gluta-max, 30 lg/mL penicillin, and 100 lg/mL streptomycin(complete RPMI medium).

FACS analysis of eGFP expression

RAW264.7/ELAM-eGFP cells were plated at 13 105 cells/mL in 1.0 mL of complete RPMI medium in the wells ofa 24-well tissue culture plate containing the test materials(n ¼ 6 per material). Bacterial lipopolysaccharide (LPS), at10 ng/mL (Salmonella minnesota, Sigma), was included inthe cultures as a positive control of proinflammatorystimulus. Cells were incubated for 24 h in a 5% CO2/airatmosphere at 378C. After incubation the cells wereallowed to detach for 5 min in 500 lL of ice cold PBS,1 mM EDTA, 0.1% NaN3, and were aspirated off and ana-lyzed with a FACS-Calibur flow cytometer, using theCELLQuest software package (Becton Dickinson). Datawas collected from 1 3 104 events with FL1 emission at550 nm.

Osteoclasts

Osteoclast differentiation was stimulated from bonemarrow progenitor cells in vitro, using the osteoclastogeniccytokines RANKL and M-CSF. Bone marrow cells wereharvested by flushing out the femurs from 6-week-oldmale Balb/c mice. The cells were washed with DMEMcontaining 10% heat-inactivated FCS, 2 mM Glutamax,30 lg/mL penicillin, and 100 lg/mL streptomycin (com-plete DMEM medium) to remove endogenous cytokines.The cells were plated onto 10 cm tissue culture plasticdishes containing a 1 mm thick plug of 0.2% collagen,20 mM HEPES, 1 3 complete DMEM, and were culturedfor 7 days in complete DMEM in the presence of 20 ng/mLrecombinant human soluble RANKL and 1 3 104 U/mL re-combinant human M-CSF. Osteoclasts were allowed todifferentiate for 7 days with replacement of the mediumon days 3 and 6.20,21

PHBV COMPOSITE BIOMATERIALS FOR BONE TISSUE REGENERATION 601


Incubation of osteoclasts with biomaterials

Osteoclasts were harvested following treatment withRANKL and M-CSF for 7 days. The culture medium wasremoved and the collagen was solubilized, using 0.1%collagenase/0.2% dispase in PBS, at 378C for 30 min. Thecells were pelleted by centrifugation at 1000 rpm, washedwith fresh complete DMEM and resuspended in 5 mL ofcomplete DMEM, whereupon they were placed onto12.5 mm diameters solvent cast discs of the test materials(n ¼ 6 per material) in 6-well dishes and incubated for48 h. Solvent cast discs were chosen for this experiment asit was not possible to assess actin ring formation on melt-processed disks due to their thickness.

Tartrate-resistant acid phosphatase staining

Adherent osteoclasts on the materials (after incubationfor 48 h) were fixed and stained for tartrate-resistant acidphosphatase (TRAP) activity, using a commercial kit(Leukocyte Acid phosphatase kit; Sigma) according to themanufacturer’s instructions. Cells were photographed undera dissection microscope (Olympus SZX 12), using SPOTDiagnostic Instrumentation (Camera and Programming).

Immunofluorescence and confocal laser microscopy

Osteoclast cells cultured on the solvent cast biomaterialsand cortical bone slices were fixed with TRAP fixative(0.25% citrate, 0.65% Acetone, 0.1% formaldehyde) for5 min and washed with several changes of PBS. The cellson the materials were permeabilized with 0.5% TritonX-100 (Banksia Scientific, Australia) for 10 min followed byseveral washes with PBS. Cells were then incubated withTRITC Phalloidin conjugates (Sigma, Australia), diluted inPBS 1:200 for 15 min. This step was performed in the darkto prevent bleaching. The materials were then washedwith several changes of PBS and mounted with Fluores-cent Mounting medium (DakoCytomation) on a glass slide.Investigation of the actin ring on the materials wasperformed with a Zeiss Meta inverted confocal lasermicroscope, using an Argon/2 laser at 560 and 580 nmwavelengths.

Visualization of resorption pits

At the osteoclast culture end-points, materials were soni-cated to remove any cellular debris (Sonics and Materials,VibraCell with a microprobe) for 60 sec, with a duty cycleof 50%, and power output of 3. Resorption pit formationwas examined by scanning electron microscopy (SEM) ona Jeol JSM 6400F electron microscope.

Osteoblasts

Stock cultures of murine calvarial preosteoblast cells,MC3T3-E1 S14 (ATCC), were cultured in a-MEM (mini-mum essential medium) supplemented with 10% heat-

inactivated FCS, 2 mM glutamine, 1 mM sodium pyruvate,100 U/mL of penicillin, 100 lg/mL of streptomycin(standard media) in a humidified atmosphere of 5% CO2.Until the time of study, cells were subcultured every 3–4days, using 0.001% (w/v) pronase to achieve cell detach-ment. To characterize the effects of the various materialson osteoblast growth and development, stock cultures ofMC3T3-E1 S14 cells were serum-starved overnight, lifted,counted, and statically seeded on to each melt-processedmaterial (2 3 103 cells/cm2 of material) in 12-well cultureplates (Nunc) or control polystyrene surface wells andcultured for 21 days in the presence of standard mediasupplemented with 5 mM b-glycerophosphate and 25 lg/mL of ascorbic acid (osteogenic media).

Measurement of calcium levels

Calcium levels were determined after 21 days growth,using a modified Calcium C test (Wako #272-21801) (n ¼6 per material). Briefly, the cultures were washed fourtimes in Ca2þ-free and Mg2þ-free phosphate-bufferedsaline and pelleted after 5 min incubation in 0.25% trypsincontaining 1 mM EDTA. Cells were then incubated with1 mL of 1% trichloroacetic acid for 5 min, vortexed, andcentrifuged for 20 min at 1600 g at 258C. The supernatant(100 lL) was then reacted with o-cresol–phthalein-complex,and the colorimetric reaction was read at 570 nm. Theabsolute calcium concentration was determined using astandard.

Mineralization in vitro

After 21 days in culture, matrix mineralization wasassessed by von Kossa histochemical staining (n ¼ 6 permaterial). Samples were washed three times with PBS,fixed with 10% neutral buffered (PBS) formalin for 12 h,and rinsed with deionized water. After the addition of 5%aqueous silver nitrate solution, the samples were exposedto UV light for 1 h. The reaction was stopped by treatmentwith 5% sodium thiosulfate and samples imaged with astereo microscope (Olympus SZX 12) and SPOT CCD cam-era (Diagnostic Instruments Incorporated) and intensitiesof the signals from four distinct area per material weremeasured, using NIH image version 1.62 (http://rsb.info.nih.gov/nih-image/) and expressed as mean 6 SD.

Statistical analysis

Statistical significance was tested, using one-way ana-lysis of variance (ANOVA) followed by Tukey Post-hoctesting where appropriate (SPSS).

RESULTS

Biomaterials

The materials investigated in this study includedthe single component materials, HA, b-TCP, and

602 COOL ET AL.


poly(hydroxybutyrate-co-hydroxyvalerate) (PHBV) aswell as the composite materials HA/PHBV, cHA/PHBV, and b-TCP/PHBV. The HA was produced bya solution method and the calcined HA (cHA) bysubsequent heat treatment.17 Transmission electronmicrograms (Fig. 1) confirm that the mineral particleswere nano-sized. The HA particles consisted of round

ended prismatic crystals averaging *60 3 20 nm2 insize with a small crystal size distribution. In samplesof the cHA and b-TCP crystals, a large size distri-bution was observed. The smaller nano-sized crys-tals displayed spindle-shaped (*80 3 25 nm2) crys-tal morphology whilst the larger submicron-sizedcrystals were rounded (*100–400 nm). The phase

Figure 1. Material characterization. XRD patterns of HA (A), cHA (C), b-TCP (E) shows all the inorganic materials to beof high phase purity. Transmission electron micrograms of HA (G), cHA (H), b-TCP (I) shows that all materials are nano-sized; the cHA and b-TCP display a large size distribution. XRD patterns of HA/PHBV (B), cHA/PHBV (D), b-TCP/PHBV (F) confirm phase purity of the mineral component in the composite materials. Scale bar ¼ 150 nm.



purity of HA, cHA, and b-TCP were confirmed byXRD [Fig. 1(A,C,E)].

The HA/PHBV composite material was producedby a novel method,16 using a dual solvent system, inwhich the solution prepared inorganic particles werestabilized by added PAA.22 Thus, aqueous suspen-sions of the inorganic phase (HA, cHA or b-TCP)were added a PAA solution in a staggered mannersuch that the final suspension contained 15% PAAby weight with respect to the crystals.23 The PAAstabilized particle suspension was added to a solu-tion of PHBV in DMF, in which the inorganicparticles remain dispersed. The choice of DMF asthe polymer solvent was based on DMF and waterbeing miscible and PHBV undergoing a tempera-ture-controlled gelation process in DMF, even at lowconcentrations.24 Rapid gelation upon cooling pre-vented large scale phase separation. Subsequent addi-tion of MeOH caused the polymer gel to collapseresulting in a fine white powder of the compositematerial. This rapid gelation appears to trap thenanoparticles and reduces agglomeration in the com-posite materials. Since cHA and b-TCP were isolatedin the dry state prior to composite formation, amodified procedure involving suspension of theinorganic phases in water prior to PAA addition wasutilized. Composite material powders, HA/PHBV,cHA/PHBV, and b-TCP/PHBV, were analyzed byXRD [Fig. 1(B,D,F)] from which peaks from the inor-ganic component were observed at diffraction anglesdifferent to that of the polymer. Again, phase purityof the mineral component was confirmed. To deter-mine the amount of mineral component in the com-posite materials, carbon analysis was performed. Bycomparing the carbon content in the compositematerials with that of pure polymer and of HA,cHA, and b-TCP, it was found that HA/PHBV,cHA/PHBV and b-TCP/PHBV composites contained77, 79, and 80% PHBV, respectively, for the powdersused to prepare melt-processed disks. For thepowder used to prepare the solvent cast HA/PHBVfilm contained 85% PHBV.

To assess the chemical surface composition, XPSwas employed. This technique analyses the top 50–100 A of the material surface. It was found that 10%or less of the inorganic phase was presented on thesurface; typically, Ca atom% of 0.64% in the meltprocessed HA/PHBV sample, and 0.32% in thecHA/PHBV and b-TCP/PHBV samples, comparedwith 6.3% in the powders. For the solvent-cast filmsit was similarly found that again 10% or less of theinorganic phase was presented on the surface with a15% HA/PHBV composite yielding a Ca atom% of0.65%. Clearly, the processing techniques used forpreparing the samples do not allow for full pre-sentation of the inorganic component on the materialsurface. Importantly, the surface presentation is sig-

nificantly higher in the HA/PHBV than the cHA/PHBV and b-TCP/PHBV materials.

Macrophage assays

The induction of a proinflammatory response bymacrophages can be assayed by the production ofproinflammatory cytokines or through the activationof signal transduction pathways, including thosethat result in the activation of the transcription factorNF-kappa B.18 We have established the use of theNF-kappa B activation-responsive ELAM promoterdriving expression of a reporter gene as a readoutfor these pathways.19,25 The RAW264.7/ELAM-eGFPcell line was cultured for 24 h with the melt-pro-cessed materials under test and then analyzed foreGFP expression by FACS. Figure 2(A–D) shows thegating parameters and fluorescence profile of theFACS analysis of the RAW264.7/ELAM-eGFP cellline. The autofluorescence observed in the untreatedcells was used to set the M1/M2 cutoff such thatthe fluorescence level of 99.9% of untreated cells iswithin M1 [Fig. 2(B)]. Maximal induction of eGFPexpression by incubation with LPS resulted in anincrease in the number of cells exhibiting eGFP fluo-rescence detected in the M2 region [Fig. 2(D)].

All materials were incubated with RAW264.7/ELAM-eGFP cells for 24 h then analyzed by FACSfor eGFP fluorescence [Fig. 2(E)]. HA disks induceminimal eGFP expression, and while b-TCP disksshow apparently higher induction, it should be notedthat this predominantly results from one outlyingsample. Melt-processed PHBV was able to induceGFP expression such that 29% 6 3% of cells werein the M2 region but the incorporation of variouscalcium phosphate minerals to form PHBV compos-ite materials resulted in a significant reduction ineGFP expression. The HA/PHBV composite wasmost effective in reducing the proinflammatoryresponse achieving a level that was comparable tob-TCP alone. Statistical analysis of the induction ofeGFP expression established that the levels were sig-nificantly different between the materials [Fig. 2(E)].

An experiment was devised to assess the cause ofthe low proinflammatory response to the HA/PHBVcomposite. PHBV dissolved in DMF was treated withammonia solution and subsequently precipitatedand washed with MeOH to simulate this part of theHA/PHBV fabrication process; however, the samplesproduced in this experiment showed a high proin-flammatory response similar to that unpurifiedPHBV (data not shown). In addition, HA pelletswere cultured with 10, 1, and 0.1 ng/mL of LPS for24 h. Conditioned media were harvested and cul-tured with the RAW264.7/ELAM-eGFP cells; how-ever, a similar proinflammatory response to that ofthe positive control was observed (data not shown).

604 COOL ET AL.


Osteoclast assays

When bone marrow-derived osteoclast-like (OCL)cells were cultured on the surface of the solvent castmaterials, TRAP staining revealed a population ofadherent mononuclear and multinuclear TRAPþ cellson the surface of the materials [Fig. 3(A–C)]. Cellscultured on control-bone were in general large andmultinucleated, while cells cultured on PHBV werepredominantly mononuclear with fewer apparentmultinucleated cells. Cells cultured on HA/PHBVwere mixed with both mononuclear and multi-nucleated cells evident, representing an intermediatebetween the cell populations on bone and PHBV.

The formation of an actin ring within the cellscultured on the materials was examined as a mea-sure of the functional resorptive response of the cellsto the materials [Fig. 3(D–F)]. Phalloidin staining of

actin rings was observed in cells cultured on eithercortical bone or bovine dentine (control) but not incells cultured on either PHBV or HA-PHBV compo-sites (cells cultured on these materials only showedcortical actin staining).

After removal of the cells, the materials were exa-mined for the presence of pits in the surface. OCLcells cultured on bovine cortical bone (control) con-sistently generated surface pits, indicating they werecapable of resorptive activity. While OCL cells attach-ed to PHBV, HA/PHBV, cHA/PHBV, and b-TCP/PHBV, they failed to form surface pits in the mate-rials, as determined by SEM [Fig. 3(G–L)].

Osteoblast assays

All materials investigated were found to have asignificant effect (p < 0.001) on osteogenic differentia-

Figure 2. FACS analysis of the production of eGFP by RAW264.7/ELAM-GFP cells incubated with different materials. A:Scatter plot of events showing the region gated for detection. SSC-H represents side scatter and FSC-H represents forwardscatter for RAW264.7/ELAM-eGFP cells incubated on tissue culture plastic alone. B: Histogram of the number of gatedevents plotted against the fluorescence intensity (FL1-H) for RAW264.7/ELAM-GFP cells incubated on tissue cultureplastic alone. M1 represents the events with a fluorescence intensity equivalent to 99.9% of unstimulated events and M2represents the number of events with a fluorescence intensity in excess of M1. C: Scatter plot for RAW264.7/ELAM-eGFPcells incubated on tissue culture plastic with 10 ng/mL LPS for 24 h. D: Histogram of the number of gated events plottedagainst the fluorescence intensity (FL1-H) for RAW264.7/ELAM-eGFP cells incubated on tissue culture plastic with 10 ng/mL LPS for 24 h. E: Histogram of the percentage of events in M2/(M1þM2) (%).



tion. Equal numbers (4 3 104 cells/well) of MC3T3-E1 S14 cells were grown on each material as de-scribed in the Material and Methods section, withan attachment efficiency of almost 90% (data notshown). Seven and twenty-one days after seeding,calcium accumulation and von Kossa histochemicalstaining was used to examine the effect of each ma-terial on osteogenic differentiation. No significantdifference in the morphology of the cells grown onthe materials was observed. On both the control andthe various materials, cells adopted a spindle shapeimmediately after seeding [Fig. 4(A)], and becamecuboidal upon reaching confluence (7 days) (datanot shown). By twenty-one days, all cultures formeda dense monolayer with obvious areas of mineraliza-tion [Fig. 4(B)].

Mineralization and calcium accumulation in osteo-genic cultures is known to follow an initial period of

cell proliferation lasting up to 16 days, with matrixmineralization involving growth of apatite mineralon and between collagen fibrils, and the formationof calcified nodules.26 Cultures of MC3T3-E1 S14cells grown on the various materials failed to accu-mulate calcium and mineralize prior to 16-days (datanot shown), however by 21-days, cultures varied sig-nificantly in their levels of calcium (p < 0.001), withcells grown on HA having a significantly higher con-centration of calcium (p < 0.001) when comparedwith the other materials tested [Fig. 4(C)]. Cellscultured on HA, PHBV and HA/PHBV materialshad up to a 5-fold increase in calcium when com-pared with tissue culture plastic, whilst b-TCP/PHBV and cHA/PHBV composites had similarcalcium levels to tissue culture plastic [Fig. 4(C)].

This pattern was repeated in the bone noduleformation assay, as determined by von Kossa histo-

Figure 3. Bone marrow cells obtained from 7-week-old male Balb/c mice and differentiated as described in the Materialsand Methods were grown on the materials for 7 days. TRAP staining of OCL on material surfaces (A) cortical bone; (B)HA/PHBV; (C) PHBV; the scale bar on the photomicrographs represents 50 lm. FITC-phalloidin actin staining of OCL onmaterial surfaces (D) cortical bone; (E) HA/PHBV; (F) PHBV. The scale bar on the photomicrographs represents 50 lm.SEM of materials without OCLs (G) cortical bone; (H) HA/PHBV; (I) PHBV and SEM of osteoclastic resorption pits; (J)cortical bone; (K) HA/PHBV; (L) PHBV. Scale bar ¼ 10 lm. [Color figure can be viewed in the online issue, which is avail-able at www.interscience.wiley.com.]

606 COOL ET AL.


chemistry [Fig. 4(D)]. Since von Kossa histochemistrystains phosphate formations, materials comprisinghigh levels of calcium phosphates (b-TCP and HA)stained entirely black and were removed from thisassay (data not shown). For the remaining materials(culture plastic, cHA/PHBV, b-TCP/PHBV, PHBV,HA/PHBV), no-cell controls produced low back-ground staining (data not shown). Osteogenic cul-tures grown on these materials for twenty-one days,showed clear evidence of mineralization with littlebackground [Fig. 4(D)]. Both PHBV and HA/PHBVhad significantly higher levels of mineralization com-

pared with the other materials (p < 0.007) [Fig. 4(E)].Of all the materials tested, cultures grown on HA/PHBV produced the highest levels of calcium[Fig. 4(C)] and mineralized matrix [Fig. 4(D,E)].

DISCUSSION

The success of a bone substitute material is de-pendent upon both its structural and chemical prop-erties as well as the interactions it has with the host’s

Figure 4. The effect of HA/PHBV and b-TCP/PHBV composites on osteoblastic differentiation. Equal numbers(4 3 104 cells/well) of MC3T3-E1 S14 cells were cultured on a range of materials as described in the Material and Methodssection. A: Immediately following seeding, cells adopted a spindle shape with cytoplasmic extensions reaching out totouch neighboring cells. B: By twenty-one days, a dense monolayer of cells is formed on all materials, with obvious areasof mineralization. C: A significant material-dependent effect on calcium accumulation was observed. Cells cultured onHA, b-TCP, PHBV, and HA/PHBV materials had up to a 5-fold increase in calcium, when compared with tissue cultureplastic, whilst b-TCP/PHBV and cHA/PHBV materials had similar calcium levels to tissue culture plastic. D: Von Kossahistochemistry showed clear evidence of mineralization with little background on all but the pure HA and b-TCPmaterials. Significantly higher levels of mineralization were observed for PHBV and HA/PHBV when compared with theother materials. Of all the materials tested, cultures grown on HA/PHBV produced the highest levels of calcium andmineralized matrix.



tissues. In this study the in vitro biological responseof pure PHBV and mineral-reinforced PHBV compo-sites was assessed, using cultured osteoblasts, osteo-clasts, and macrophages.

Previous studies that have utilized a combinationof a bioactive calcium phosphate mineral and PHB(V)have been inconclusive. Implantation of a micron-HA/PHB(V) composite into a bone defect was foundto form a layer of bonelike apatite at the HA/PHB(V) implant interface, although no comparisonwas made to the pure polymer.7,8 In a separatestudy on HA/PHB composite materials no differencein bioactivity was seen between PHB and its com-posite.10 In the latter study, XPS of the materials wasobtained prior to implantation and revealed that theHA was not presented on the surface, thus explain-ing the lack of improved bioactivity. Unfortunately,the former studies did not investigate the surfacepresentation of the inorganic phase and so it is diffi-cult to compare the results.

Our approach to production of the HA/PHBVnanocomposite differs from previously publishedprocedures in that the HA particles were not isolatedin solid form prior to composite formation. This newapproach was taken to avoid irreversible agglomera-tion. The cHA and b-TCP composites were, how-ever, made from particles that had been isolated anddried prior to composite formation. In all the compo-sites, PAA was used to stabilize the aqueous suspen-sion of the nano- and submicron-sized particles.Polycarboxylates are known to bind to HA with thecarboxyl group possibly being incorporated into thelattice.27,28 Several authors have shown that PAA canbe used to prepare colloidally-stable suspensions ofnano-sized HA particles.23,29–31 We have previouslyshown using TEM16 that the HA/PHBV compositesindeed contain discrete nano-sized HA particles inthe polymer matrix. Few other studies on fabricationof HA nanocomposites have shown similar disper-sion of the particles.32–35

One determinant of tissue rejection is the inflam-matory response by host macrophages. Activationand nuclear translocation of the N-kappa B tran-scription factor is characteristic of the signaling path-ways associated with proinflammatory responses.The observation that PHBV alone could stimulatea proinflammatory signal suggests that the melt-processed material contains compounds that aredetected by RAW264.7 cells, which then respond byactivating intracellular signaling via NF-kappa B.Since the polymer PHBV is derived from gram nega-tive bacteria, it is likely that it contains remains fromthese bacteria, such as LPSs or bacteria DNA. Previ-ous studies by Sevastianov and colleagues36,37 havedemonstrated, using gas chromatography–massspectrometry, that PHB(V) samples contain bacteri-ally-derived impurities, such as long-chain fatty acids

of b-hydroxypalmitic and saturated b-hydroxytetra-decanoic acid. The presence of calcium phosphate inany of the PHBV composites ameliorated the proin-flammatory response. The composition of the calciumphosphate also affected the proinflammatory activityof PHBV and thus it was found that HA/PHBVcomposite materials yielded the lowest response ofall the composite materials studied. The mechanismof this process was investigated by a number ofexperiments. Firstly, since only the HA/PHBV com-posite has the majority of the particles on the nano-scale, a much higher surface area of inorganic phaseexists in this composite. LPS solutions added to HAparticles did not show a reduced proinflammatoryresponse, suggesting that LPS is not sequestered byHA. Secondly, the HA/PHBV composite was pro-duced by a different method to that of the othercomposites in that the HA/PAA suspension containsammonia. PHBV was subjected to ammonia treat-ment but showed a large proinflammatory responseand thus it cannot be this aspect of the fabricationprocess that causes the observed reduction. It ishowever possible that it is a simple barrier effect, asthe HA/PHBV composite displayed a significantlyhigher surface presentation of the inorganic phasecompared with the other composite materials.

Osteoclast cells differentiated in vitro using solubleRANKL and M-CSF have the capacity to resorb den-tine and form pits in the surface.38 Biomaterials stud-ies investigating OCL resorption are restricted toresorption of pure inorganic materials39,40 and to thebest of our knowledge, no study has previously beenperformed on resorption of polymeric or compositematerials. Using actin staining and SEM, we testedwhether murine in vitro-differentiated OCL cellshave the capacity to form actin rings and resorptionpits in PHBV and the PHBV composite materials.The observation that osteoclasts adhere to PHBV andthe composite materials, as well as being able todifferentiate on these surfaces, establishes that thesematerials are capable of acting as a supporting ma-trix for these cells. The attachment of OCL to thematerials presumably occurs through an integrin-independent process. Within the time frame of theexperiment (48 h) no resorption of PHBV or any ofthe composite materials was observed. In vitro differ-entiation of osteoclasts on the material over a periodof 7 days also failed to yield resorptive pits (data notshown). Further analyses of material resorptionmight be better studied in an in vivo model. In anycase, the failure of osteoclasts to form focal resorp-tion pits on the surface of PHBV may not necessarilyobviate hydrolysis of the material through the releaseof hydrolytic enzymes by macrophage lineage cells.Progressive hydrolysis accompanied by the observeddeposition of bone mineral by osteoblasts may resultin effective replacement of the PHBV with bone.

608 COOL ET AL.


Notably, all the materials investigated were foundto have a significant effect on osteogenic differentia-tion and varied in their levels of mineralization, withcells grown on HA having a significantly higherlevel of accumulated calcium when compared with theother materials tested. Specifically, when osteoblastswere cultured on HA, PHBV, and HA/PHBV materi-als, a 5-fold increase in calcium accumulation wasobserved, compared with tissue culture plastic, whilstb-TCP/PHBV and cHA/PHBV composites hadsimilar levels to tissue culture plastic. Furthermore,both PHBV and HA/PHBV had significantly higherlevels of mineralization (von Kossa staining) whencompared with the other composite materials. Impor-tantly, our study revealed that only the HA/PHBVcomposite material (and not the cHA/PHBV or b-TCP/PHBV materials) displayed improved bioactiv-ity (highest levels of accumulated calcium and amineralized matrix) when compared with the purepolymer. There are two factors to consider here:firstly, only the HA/PHBV composite material hasthe majority of the reinforcing phase on the nano-scale; secondly, a higher surface presentation of thereinforcing phase was seen for the HA/PHBVmaterial. A recent study looked at the effect of thereinforcing phase being nano-sized versus onmicron-sized in poly(propylene fumarate) compositematerials and found enhanced bioactivity for thenanocomposite.41 However, this study did notinvestigate the surface presentation and so it is notclear if this parameter was the same or different intheir two materials. Another recent study comparedtwo composite materials composed of HA and poly(lactic-co-glycolic acid) produced by different meth-ods resulting in different surface presentations ofHA.42 They found a correlation between surface pre-sentation (assessed by XPS) and in vitro and in vivobioactivity. Therefore it is likely that the parametersthat govern the osteoblasts response in our studyare a combination of the surface presentation ofthe calcium phosphate phase and the presence ofnano-sized HA particles in the HA/PHBV compositematerial.

CONCLUSION

This study investigated the in vitro osteogenic andinflammatory properties of poly(3-hydroxybutyrate-co-3-valerate) (PHBV) with nano-sized HA; submi-cron-sized cHA; and submicron-sized b-TCP ascalcium phosphate-reinforcing phases, using bioas-says of cultured osteoblasts, osteoclasts and macro-phages. The proinflammatory response to PHBV itselfwas very high but the introduction of a reinforcingphase vastly reduced this response, with the HA/

PHBV composite material yielding the greatest reduc-tion. Osteoclasts readily attached and differentiatedon all the materials but only formed resorption pitson cortical bone. In comparison, osteoblasts differen-tiated on all the materials tested, but differentiatedand mineralized most strongly on HA/PHBV. Thus,both the macrophage and osteoblast assays yieldedsimilar results on the HA/PHBV composite wherenonheat-treated HA nanocrystals were used as thereinforcing phase. We believe that a combinationthe nano-sized reinforcing phase of the HA/PHBVcomposite together with the surface presentation ofthe mineral accounts for this observation. Thereforeit is clear that the use of a carefully chosen reinforc-ing phase can vastly improve the in vitro cell re-sponse to PHBV biomaterials.

The authors would also like to thank the followingpeople for their assistance with data collection: Dr. GwenLawrie (Nanotechnology and Biomaterials Centre, Univer-sity of Queensland), Dr. Barry Wood and Anya Yago(Brisbane Surface Analysis Centre, University of Queens-land), and George Blazak (Microanalysis facility, Chemis-try Department, University of Queensland).

References

1. Kose GT, Korkusuz F, Korkusuz P, Purali N, Ozkul A,Hasirci V. Bone generation on PHBV matrices: An in vitrostudy. Biomaterials 2003;24:4999–5007.

2. Gogolewski S, Jovanovic M, Perren SM, Dillon JG, HughesMK. Tissue response and in vivo degradation of selected poly-hydroxyacids: Polylactides (PLA), poly(3-hydroxybutyrate)(PHB), and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHB/VA). J Biomed Mater Res 1993;27:1135–1148.

3. Holland SJ, Jolly AM, Yasin M, Tighe BJ. Polymers for biode-gradable medical devices. II. Hydroxybutyrate-hydroxyvaler-ate copolymers: Hydrolytic degradation studies. Biomaterials1987;8:289–295.

4. Taylor MS, Daniels AU, Andriano KP, Heller J. Six bioabsorb-able polymers: In vitro acute toxicity of accumulated degra-dation products. J Appl Biomater 1994;5:151–157.

5. Chen GQ, Wu Q. The application of polyhydroxyalkanoatesas tissue engineering materials. Biomaterials 2005;26:6565–6578.

6. Luzier WD. Materials derived from biomass/biodegradablematerials. Proc Natl Acad Sci USA 1992;89:839–842.

7. Luklinska ZB, Bonfield W. Morphology and ultrastructure ofthe interface between hydroxyapatite-polyhydroxybutyratecomposite implant and bone. J Mater Sci Mater Med 1997;8:379–383.

8. Luklinska ZB, Schluckwerder H. In vivo response to HA-polyhydroxybutyrate/polyhydroxyvalerate composite. J Microsc2003;211(Part 2):121–129.

9. Chen J, Davis SS. The release of diazepam from poly(hydroxy-butyrate-hydroxyvalerate) microspheres. J Microencapsul2002;19:191–201.

10. Doyle C, Tanner ET, Bonfield W. In vitro and in vivo evalua-tion of polyhydroxybutyrate and of polyhydroxybutyratereinforced with hydroxyapatite. Biomaterials 1991;12:841–847.

11. Kose GT, Kenar H, Hasirci N, Hasirci V. Macroporouspoly(3-hydroxybutyrate-co-3-hydroxyvalerate) matrices forbone tissue engineering. Biomaterials 2003;24:1949–1958.



12. Kumarasuriyar A, Jackson RA, Grondahl L, Trau M,Nurcombe V, Cool SM. Poly(b-hydroxybutyrate-co-b-hydroxy-valerate) supports in vitro osteogenesis. Tissue Eng 2005;11:1281–1295.

13. Nebe B, Forster C, Pommerenke H, Fulda G, Behrend D,Bernewski U, Schmitz KP, Rychly J. Structural alterations ofadhesion mediating components in cells cultured on poly-b-hydroxy butyric acid. Biomaterials 2001;22:2425–2434.

14. Saad B, Neuenschwander P, Uhlschmid GK, Suter UW. Newversatile, elastomeric, degradable polymeric materials formedicine. Int J Biol Macromol 1999;25:293–301.

15. Tezcaner A, Bugra K, Hasirci V. Retinal pigment epitheliumcell culture on surface modified poly(hydroxybutyrate-co-hydroxyvalerate) thin films. Biomaterials 2003;24:4573–4583.

16. Lutton C, Read J, Trau M. Nanostructured biomaterials: Anovel approach to artificial bone implants. Aust J Chem 2001;54:621–623.

17. Wei M, Evans JH, Bostrom T, Grondahl L. Synthesis andcharacterization of hydroxyapatite, fluoride-substituted hydro-xyapatite and fluorapatite. J Mater Sci Mater Med 2003; 14:311–320.

18. Schindler U, Baichwal VR. Three NF-j B binding sites inthe human E-selectin gene required for maximal tumornecrosis factor a-induced expression. Mol Cell Biol 1994;14:5820–5831.

19. Stacey KJ, Young GR, Clark F, Sester DP, Roberts TL, Naik S,Sweet MJ, Hume DA. The molecular basis for the lack ofimmunostimulatory activity of vertebrate DNA. J Immunol2003;170:3614–3620.

20. Akatsu T, Tamura T, Takahashi N, Udagawa N, Tanaka S,Sasaki T, Yamaguchi A, Nagata N, Suda T. Preparation andcharacterization of a mouse osteoclast-like multinucleated cellpopulation. J Bone Miner Res 1992;7:1297–1306.

21. Takahashi N, Yamana H, Yoshiki S, Roodman GD, MundyGR, Jones SJ, Boyde A, Suda T. Osteoclast-like cell formationand its regulation by osteotropic hormones in mouse bonemarrow cultures. Endocrinology 1988;122:1373–1382.

22. Bertoni E, Bigi A, Cojazzi G, Gandolfi M, Panzavolta S,Roveri N. Nanocrystals of magnesium and fluoride substi-tuted hydroxyapatite. J Inorg Biochem 1998;72:29–35.

23. Bertoni E, Bigi A, Falini G, Panzavolta S, Roveri N. Hydroxy-apatite/polyacrylic acid nanocrystals. J Mater Chem 1999;9:779–782.

24. Cesaro A, Fabri D, Sussich F, Paradossi G. Structural andthermodynamic features of the polyhydroxybutyrate physicalgels. Macromol Symp 1999;138:165–174.

25. Hume DA, Underhill DM, Sweet MJ, Ozinsky AO, Liew FY,Aderem A. Macrophages exposed continuously to lipopoly-saccharide and other agonists that act via toll-like receptorsexhibit a sustained and additive activation state. BMC Immu-nol 2001;2:11.

26. Quarles LD, Yohay DA, Lever LW, Caton R, Wenstrup RJ.Distinct proliferative and differentiated stages of murineMC3T3-E1 cells in culture: An in vitro model of osteoblastdevelopment. J Bone Miner Res 1992;7:683–692.

27. Ellis J, Jackson AM, Scott RP, Wilson AD. Adhesion ofcarboxylate cements to hydroxyapatite. III. Adsorption ofpoly(alkenoic acids). Biomaterials 1990;11:379–384.

28. Misra DN. Adsorption on hydroxyapatite: Role of hydrogenbonding and interphase coupling. Langmuir 1988;4:953–958.

29. Brown PW. Hydroxyapatite-polymer composites. PhosphorusSulfur Silicon Relat Elem 1999;144–146:57–60.

30. Misra DN. Adsorption of polyacrylic acids and their sodiumsalts on hydroxyapatite: Effect of relative molar mass. J Col-loid Interface Sci 1996;181:289–296.

31. Zhang S, Hou Z, Gonsalves KE. Preparation and characteriza-tion of hydroxyapatite nanowhiskers. Polym Mater Sci Eng1995;73:300–301.

32. Deng X, Hao J, Wang C. Preparation and mechanical proper-ties of nanocomposites of poly(D,L-lactide) with Ca-deficienthydroxyapatite nanocrystals. Biomaterials 2001;22:2867–2873.

33. Hao J, Yuan M, Deng X. Biodegradable and biocompatiblenanocomposites of poly(e-caprolactone) with hydroxyapatitenanocrystals: Thermal and mechanical properties. J ApplPolym Sci 2002;86:676–683.

34. Jie W, Yubao L, Weiqun C, Yi Z. A study on nanocomposites ofhydroxyapatite and polyamide. J Mater Sci 2003;38:3303–3306.

35. Yamaguchi I, Tokuchi K, Fukuzaki H, Koyama Y, TakakudaK, Monma H, Tanaka J. Preparation and microstructure anal-ysis of chitosan/hydroxyapatite nanocomposites. J BiomedMater Res 2001;55:20–27.

36. Sevastianov VI, Perova NV, Shishatskaya EI, Kalacheva GS,Volova TG. Production of purified polyhydroxyalkanoates(PHAs) for applications in contact with blood. J Biomater SciPolym Ed 2003;14:1029–1042.

37. Shishatskaya EI, Volova TG, Puzyr AP, Mogilnaya OA, Efre-mov SN. Tissue response to the implantation of biodegrad-able polyhydroxyalkanoate sutures. J Mater Sci Mater Med2004;15:719–728.

38. Udagawa N, Takahashi N, Jimi E, Matsuzaki K, Tsurukai T,Itoh K, Nakagawa N, Yasuda H, Goto M, Tsuda E, HigashioK, Gillespie MT, Martin TJ, Suda T. Osteoblasts/stromal cellsstimulate osteoclast activation through expression of osteo-clast differentiation factor/RANKL but not macrophage col-ony-stimulating factor: Receptor activator of NF-j B ligand.Bone 1999;25:517–523.

39. Langstaff S, Sayer M, Smith TJ, Pugh SM. Resorbable biocer-amics based on stabilized calcium phosphates. II. Evaluationof biological response. Biomaterials 2001;22:135–150.

40. Monchau F, Lefevre A, Descamps M, Belquin-myrdycz A,Laffargue P, Hildebrand HF. In vitro studies of human andrat osteoclast activity on hydroxyapatite, b-tricalcium phos-phate, calcium carbonate. Biomol Eng 2002;19:143–152.

41. Lewandrowski KU, Bondre SP, Wise DL, Trantolo DJ.Enhanced bioactivity of a poly(propylene fumarate) bonegraft substitute by augmentation with nano-hydroxyapatite.Biomed Mater Eng 2003;13:115–124.

42. Kim S-S, Park MS, Jeon O, Choi CY, Kim B-S. Ploy(lactide-co-glocolide)/hydroxyapatite composite scaffolds for bone tissueengineering. Biomaterials 2006;27:1399–1409.

610 COOL ET AL.


Documents

Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) composite ......proposed for use in bone tissue engineering1 owing to their favorable degradation rates2–4 and nontoxic de-gradation