Collagen-embedded hydroxylapatite–beta-tricalcium phosphate–silicon dioxide

Collagen-embedded hydroxylapatite–beta-tricalcium phosphate–silicon dioxide bonesubstitute granules assist rapid vascularization and promote cell growth

This article has been downloaded from IOPscience. Please scroll down to see the full text article.2010 Biomed. Mater. 5 025004

(http://iopscience.iop.org/1748-605X/5/2/025004)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 162.105.247.124

The article was downloaded on 06/03/2011 at 07:38

Please note that terms and conditions apply.

IOPPUBLISHING

Biomed.Mater.5(2010)025004(11pp)

BIOMEDICALMATERIALS

doi:10.1088/1748-6041/5/2/025004

Collagen-embeddedhydroxylapatite–beta-tricalciumphosphate–silicondioxidebonesubstitutegranulesassistrapidvascularizationandpromotecellgrowth

ShahramMGhanaati1,3,4,BenjaminWThimm1,2,4,RonaldEUnger1,CarinaOrth1,ThomasKohler2,MikeBarbeck1,RalphM¨uller2andCJamesKirkpatrick1

12

InstituteofPathology,JohannesGutenberg-UniversityMainz,Langenbeckstr.1,55101Mainz,GermanyInstituteforBiomechanics,ETHZ¨urich,Wolfgang-Pauli-Str.10,8093Z¨urich,Switzerland

E-mail:ghanaati@uni-mainz.de

Received28December2009

Acceptedforpublication29January2010Published8March2010

Onlineatstacks.iop.org/BMM/5/025004

Abstract

Inthepresentstudyweassessedthebiocompatibilityinvitroandinvivoofalow-temperaturesol–gel-manufacturedSiO2-basedbonegraftsubstitute.Humanprimaryosteoblastsandtheosteoblasticcellline,MG63,culturedontheSiO2biomatrixinmonocultureretainedtheirosteoblasticmorphologyandcellularfunctionalityinvitro.TheeffectofthebiomaterialinvivoanditsvascularizationpotentialwastestedsubcutaneouslyinWistarratsand

demonstratedbothrapidvascularizationandgoodintegrationwithintheperi-implanttissue.Scaffolddegradationwasprogressiveduringthefirstmonthafterimplantation,with

tartrate-resistantacidphosphatase-positivemacrophagesbeingpresentandpromotingscaffolddegradationfromanearlystage.Thismanuscriptdescribessuccessfulosteoblasticgrowthpromotioninvitroandapromisingbiomaterialintegrationandvasculogenesisinvivoforapossibletherapeuticapplicationofthisbiomatrixinfutureclinicalstudies.(Somefiguresinthisarticleareincolouronlyintheelectronicversion)

1.Introduction

Autologousbonegraftingisstillconsideredtobethegoldstandardfornon-unionbonedefects[1].However,theincreasedriskofasecondsurgicaloperation[2,3]combinedwithcomplicationssuchasinfectionsandseverelypersistingpainnecessitatesasearchforalternativestrategies.Thedevelopmentofsyntheticallyderivedbonesubstitutematerialshasbeenpromotedtoavoidtheabove-mentionedcomplicationsassociatedwithbonetissueharvestingfromthepatient[4].Syntheticbonesubstitutematerialssofarextendfromallogenicandxenogenicbonegraftstoalloplasticmaterials[5–7].Alloplastsandxenoplastslackosteoinduction

34

Authortowhomanycorrespondenceshouldbeaddressed.Co-firstauthors.

andosteogenesisbyhostboneandfurthermorecarryariskofpathogentransmissionaswellasagraft-versus-hostreaction[4,8].Amongalloplasticbonesubstitutematerials,hydroxylapatites(HA)andbeta-tricalciumphosphates(ß-TCPs)arecurrentlythemostfrequentlyusedmaterials[9].Botharebiocompatibleandosteoconductiveandsupportregenerativebonegrowth[10].ß-TCPs,morepreciselyapatites,seemtobethenaturalchoicebecausetheyarenaturalconstituentsofboneinvivoandhavedemonstratedencouragingresultsinanimalstudies[11].ß-TCPs,however,undergoafastdegradation,whichitselfmayresultinconnectivetissueinfluxwithintheimplantationbed.Asaresult,thisinfluxmayinhibitadequateosteoblastfunctionality.Ontheotherhand,HA-basedscaffoldsareknowntobemorestablebutundergoslowdegradation.This

1

©2010IOPPublishingLtdPrintedintheUK

1748-6041/10/025004+11$30.00

Biomed.Mater.5(2010)025004SMGhanaatietal

(A)(B)

R

Figure1.Micro-computedtomography(μCT)3DimageofaspecimenofBONITmatrix󰀁-collagenscaffold(5mm3)showing

RR󰀁andcollagenBONITmatrixgranulesingray(B)anditsinterspersedcollagenmeshofthesamespecimeninyellow(A).BONITmatrix󰀁

havebeenmadevisibleforμCTmeasurementbyOsO4staining.

canhavenegativeinfluencesondenovoboneformationwithintheimplantationbed,asmacrophagesmaydifferentiateintomultinucleatedgiantcellsinordertofacilitatethedegradationofthebonesubstitute.Inturn,theencapsulatedbiomaterialmaybeconsideredasaforeignbody.Thismaychangetheinflammatorymilieuwithintheimplantationbedandenhancetheproductionofpro-inflammatorycytokines[12–15].Underthesecircumstances,thesurvivalofpre-culturedcellsisnotguaranteed.

Ideally,asyntheticbonebiomaterialshouldbedegradableandshouldcontaincomponentswhichatthesametimeinhibitarapiddegradationinvivo.Thebiomaterialused

R

inourstudy,BONITmatrix󰀁-collagen,abiphasiccalciumphosphatemixture(60%hydroxylapatiteand40%β-TCP)with13%SiO2,isproducedbyasol–gelprocedurewhichcreatesananoporousSiO2–calciumphosphatecompound,interconnectedbycollagenfibers(figure1).TheSiO2processedinthisscaffoldhasbeenproventoassistosteoblastcellproliferation,andthecollagenfibernetworkinterconnectingtheß-TCP/SiO2granulesisusedtoyieldarigid,fleece-likemold[16].Promisingresultshavebeenshowninlong-terminvivodentalinvestigationsforpure

R

BONITmatrix󰀁[17].Inarecentstudy,wewereabletoshowsuccessfulinvitrohumanendothelialcell(EC)adhesion

R

andgrowthimportantforvascularizationofBONITmatrix󰀁-collagen[18].However,asuitablebonesubstitutematerialshouldalsoenableprimarybonecellstogrowanddeveloptheirfunctionalityonitssurfacewhileinhibitinganegativeinflammatoryresponseduringitsdegradation.

R

MostinvivostudiesonpureBONITmatrix󰀁haveexhibitedinsitucompatibilityinhumanandanimaldentalapplications[19–21].Thus,theaimofthisstudywastodeterminewhetherthisnovelcalciumphosphate/SiO2–xerogelcollagencompositeaffectshumanosteoblastcell(HOB)morphology/functioninvitroanditsbiocompatibilityinvivo.Forsuccessfulbiomaterial-basedbonetissueengineering,biomaterialsshouldserveassuitablescaffoldsforosteoblaststounfoldtheirfunctionalityinproducingbonematrix.Atthesametime,thebonesubstitutesshouldbewellvascularizedinordertoundergorapiddegradation.

Thegoalofthisprojectwastotestthebonesubstitution

R

material,BONITmatrix󰀁-collagen,inaninvitrostudyon

2

itsosteoblasticcellgrowthpotentialandgeneexpression.

Aninvivoexaminationinparallelexaminedindetailthebiomaterialwithrespecttoitsbiocompatibility,degradability,vascularizationandbioactivity.Theinvivostudywasplannedtobeexecutedinanothertissueratherthanboneinordertoexcludeeffectsofresidualboneontheinvivobiomaterialintegrationanddegradation.

2.Materialsandmethods

2.1.Scaffolds

R

BONITmatrix󰀁isafullysynthetic,highlyporousbiomaterialwhichisbasedonabiphasicmixtureof87%calciumphosphates(60%hydroxylapatite(∅90nm)/40%β-tricalciumphosphate(∅400nm))and13%SiO2interconnectedviaa1%freeze-driedcollagensolutionresultinginamacro-andmicroporouscollagenmesh(yellow

R

fibersshowninfigure1).ThenanoporousBONITmatrix󰀁granules(∅0.6mm,graygranulesinfigure1)areproducedviaanon-sinteringsol–gelprocedure(nanoporosityshowninmoredetailbyGerberetal[22]).Inthisprocess,calciumphosphateisputintoaSiO2solbasedonalkoxidesandisthenstirredhomogeneously.Duringgeltransformation,SiO2formsananoporousscaffoldwhichinterconnectsthecalciumphosphatecrystalsinalooseway.Dryingat200◦Cevaporatesthesolventandcreatesporesinthemicro-andnanoporousrange.Thescaffoldhasaporosityof60–80%withalowdensity(0.9gcm−3),achlorideconcentrationof39mgkg−1andisfurtherdescribedelsewhereinmoredetail[19,23].

2.2.Micro-computedtomography

R

A5mmcubicpieceofplainBONITmatrix󰀁-collagenwascontraststainedin1%OsO4for2daysanddriedatroomtemperaturefor3daysforthe3Dvisualizationof

R

-collagen.Themicro-tomographicimagingBONITmatrix󰀁

system(μCT40,ScancoMedicalAG,Br¨uttisellen,Switzerland)usedinthisstudyisequippedwitha5μmfocalspotx-raytubeasasource.Atwo-dimensionalcharge-coupleddevice,coupledtoathinscintillatorasadetector,permitsacquisitionof100tomographicimagesinparallel.Thex-ray

Biomed.Mater.5(2010)025004SMGhanaatietal

tubewasoperatedat70kVpand114μAwithanintegrationtimesetto200msandallprojectionframeswererecordedfivetimesandthenaveraged.Thescanwasperformedatanisotropic,nominalresolutionof6μm(highresolution

R

granulesweresegmentedbyamode).TheBONITmatrix󰀁

globalthresholdingprocedure[24]withathresholdvaluesetto10%ofthemaximumgrayscalevalue(figure1).2.3.Cellsandinvitrogrowthconditions

Thehumanosteoblast-likecellline,MG63(ATCCCRL-1427),derivedfromanosteosarcoma,wasused.ThecellswereculturedinEMEM(Sigma-Aldrich,Steinbach,Germany)+10%fetalcalfserum(LifeTechnologies,Karlsruhe,Germany)+2mMglutamaxI(LifeTechnologies)+1%(100UmL−1penicillinand100mgmL−1streptomycin)andincubatedat37◦Cina95%humidifiedairwith5%CO2atmosphere.Isolatedprimaryosteoblastsfromjawboneextractionswereusedinthisstudy.Primaryhumanosteoblasts(HOS)wereculturedinDMEM+10%fetalcalfserum+1%(100UmL−1penicillin+100μgmL−1streptomycin)+2mMglutamine/ascorbicacid+100nMdexamethasone.HOSwereuseduntilthesecondpassage.

R

2.4.PreparationandcellcultureonBONITmatrix󰀁-collagenRBONITmatrix󰀁-collagenscaffoldswerepreparedforcell

R

-collagenintopiecesofculturebycuttingBONITmatrix󰀁

3

approximately0.5cmcubes,andthesewereplacedintheindividualwellsofa24-wellplate.Thescaffoldsweresterilizedin70%ethanolfor15minfollowedbythreerinsesinsterilephosphatebufferedsaline(PBS)for5mineach.Thematriceswerecoatedwithpooledhumanbloodserumfor1hat37◦CandwashedinPBSoncemore[18].ConfluentHOSwereremovedfromcellcultureflasksbytrypsinization,and1.5mLofmediumcontaining0.5×105cellswasaddedperscaffoldina48-wellplate.Thecultureplatewasincubatedfor24hat37◦Candthescaffoldswerethentransferredtoanew12-wellplatecontaining3mLfreshculturemedium.Mediumwaschangedevery2–3days.Inallcases,controlswerecarriedoutwiththesamecellsgrownoncellculturepolystyrene.Allexperimentswererepeatedwithatleastthreedifferentprimarycelldonorsandwiththecelllineatleastthreetimes.

incubatedin1%osmiumtetroxidefor1h,dehydratedinincreasingconcentrationsofacetone,criticallypointdriedandsputter-coatedwithgold.

2.6.MolecularanalysisofosteoblastgeneexpressionOsteoblast-specificgeneexpressionexaminationwascarried

R-outaftera10dayincubationperiodofcellsonBONITmatrix󰀁collagenscaffoldsbymeansofthesemi-quantitativepolymerasechainreaction(RT-PCR)technique.Thegenesunderanalysiswerealkalinephosphatase,collagen-I,osteocalcin,osteonectin,osteopontin,TGF-β1,FGF-2andthehousekeepinggeneβ-actinasanendogenouscontrol.

R

TotalRNAwasisolatedfromosteoblastsonBONITmatrix󰀁-collagenusingtheRNeasyMiniKit(Qiagen,Hilden,Germany).Cell-seededscaffoldswereplaceddirectlyintolysisbuffer.AfterRNAextraction,RNAconcentrationwasdeterminedbymeansofaNanoDropmicrospectrophotometer(NanoDropTechnologies,Rockland,Delaware,USA)and1μgRNAwasreversetranscribedintocDNA(OmniscriptRTKit,Qiagen).cDNAamplificationwascarriedoutbyusingtheTaqDNAPolymeraseKit(Qiagen)togetherwithgene-specificprimersandtheirspecificamplificationconditionsasshownintable1.Denaturationat94◦Cfor2minwasinitiallycarriedout,followedby35cycles(30cyclesforosteonectin,40cyclesforosteocalcin)at94◦Cfor30s,annealingfor30s(variesfortherespectivegenes)and30schainelongationat72◦Cfollowedbyafinalextensionof10minat72◦C.ThePCRproductswereseparatedbyagarosegel(1%)electrophoresisin1×TBEbuffer(Sigma-Aldrich)andstainedwithethidiumbromide.

2.7.Relativequantificationofalkalinephosphatasegeneexpressionbyreal-timePCR

Forreal-timeRT-PCRmeasurements,theAppliedBiosystems7300Real-TimePCRSystem(AppleraDeutschlandGmbH,Darmstadt,Germany)wasused.StandardcyclingconditionsfortheQiagenQuantiTectSYBRGreenKit(QiagenGmbH,Hilden,Germany)wereused,thatis,aninitialdenaturationandenzymeactivationphaseat95◦Cfor15min,followedby40cyclesat94◦Cfor15s,annealingat55◦Cfor30sand35schainelongationat72◦C.Thefollowingreal-timeprimerwaschosenforalkalinephosphatase(ALP)geneexpressionquantification(HsALPL1SGQuantiTectPrimerAssay(200),Kat.-Nr.:QT00012957).Thereactioncompositionwasasfollows:12.5μLof2×QuantiTectTMSYBRsGreenPCRMix,2.5μLof10×QuantiTectTMSYBRsGreenPrimerAssay,5μLofRNase-freewaterand5μLofcDNAprediluted1:100withRNase-freewater.TheanalysiswasassessedintriplicateandthespecificityofthePCRwasprovenbyitsdissociationcurves.TherelativequantificationofgeneexpressionwasassessedbyusingtheAppliedBiosystemsSequenceDetectionSoftwarev.1.2.2.Forreal-timePCRmeasurements,glyceraldehyde-3-phosphatedehydrogenase(GAPDH)wasusedasanendogenouscontrol.

3

2.5.Osteoblastimaging

R-Formorphologicalstudies,osteoblastsonBONITmatrix󰀁collagenwerevisualizedbyCalcein-AMstainingwithsubsequentCLSMimagingandfixationforscanningelectron

R-microscopy(SEM).Briefly,cell-seededBONITmatrix󰀁collagenscaffoldswereplacedinmediumcontaining0.1μMcalcein-acetoxymethylester(MoBiTec,G¨ottingen,Germany)

◦

andincubatedfor5minat37C.EndogenousesteraseshydrolyzecalceinAMtothestronglygreenfluorescentcalceinwhichisretainedandspreadinthecytoplasm.ThestainedcellscouldthenbeexaminedbyCLSM(LeicaTCSNT,Wetzlar,Germany).ForSEM,sampleswerefixedwith2.5%glutaraldehydein0.1Msodiumcacodylatebufferfor30min,

Biomed.Mater.5(2010)025004SMGhanaatietal

Table1.SpecificRT-PCRprimersandconditionsofamplification.

Gene(GenBankaccessionno)β-actin(AB004047)Osteonectin(NM003118)Osteocalcin(NM199173)Osteopontin(NM001040060)Alkalinephosphatase(NM000478)Collagen-I(NM000088)TGF-β1(NM000660)FGF-2(NM002006)

Annealingsize(bp)574304302348476600198236

Temperature(◦C)6556645655565556

Primerpairsequences

5󰀁-AGCATTTGCGGTGGACGATGGAG-3󰀁5󰀁-GACCTGACTGACTACCTCATGA-3󰀁5󰀁-TGCCTGATGAGACAGAGGTG-3󰀁5󰀁-AAGTGGCAGGAAGAGTCGAA-3󰀁5󰀁-CATGAGAGCCCTCACA-3󰀁5󰀁-AGAGCGACACCCTAGAC-3󰀁5󰀁-CCAAGTAAGTCCAACGAAAG-3󰀁5󰀁-GGTGATGTCCTCGTCTGTA-3󰀁5󰀁-CTGGTAGGCGATGTCCTTA-3󰀁5󰀁-ACGTGGCTAAGAATGTCATC-3󰀁5󰀁-TGACGAGACCAAGAACTG-3󰀁5󰀁-CCATCCAAACCACTGAAACC-3󰀁5󰀁-ACCAACTATTGCTTCAGCTC-3󰀁5󰀁-TTATGCTGGTTGTACAGG-3󰀁

5󰀁-CTGCCCAGTTCGTTTCAG-3󰀁

5󰀁󰀁-GAAGAGCGACCCTCACATCAAG-3󰀁

󰀁

2.8.Animals

AllexperimentswereperformedwiththeapprovaloftheCommitteeontheUseofLiveAnimalsinTeachingandResearch,Rhineland-Palatinate,Germany.Forthesestudies,femaleWistarrats(6-to8-weekold,90–120gbodyweight,CharlesRiverLaboratories,Sulzfeld,Germany)wereusedandhousedonepercage,keptwithwateradlibitum,withanartificiallight–darkregime,andfedwithregularmousepellets(LaboratoryRodentChow,Altromin,Germany)attheLaboratoryAnimalUnitoftheInstituteofPathology,JohannesGutenbergUniversity,Mainz,Germany.2.9.Subcutaneousimplantationmodel

R-ForassessmentofthetissuereactiontoBONITmatrix󰀁collagen,16animalswererandomlyassignedtofourgroups(n=4animals)foreachofthefollowingtimepoints:3,10,15and30days.Anesthesia(10mLketamine[50mgmL−1]with1.6mLxylazine[2%],wasadministeredbyintraperitonealinjection.Theskinoftherostralportion

R-oftheinterscapularregionwasshaved,andBONITmatrix󰀁

2

collagenspecimenswithasizeof10×10mmwereimplantedinthesubscapularregionundersterileconditions,followingapreviouslydescribedmethod[25].Allanimalssurvivedtheindicatedtimeperiodswithoutanycomplications.

paraffin.Fromeachsegment,threeconsecutive3–4μmthicksectionsweredeparaffinizedandrehydrated.ThesesectionswerestainedwithMayer’shematoxylinandeosin(H&E)andexaminedtoestablishwhichsegmentcontainedthemostrepresentativeviewofthematerials.Thiswasselectedforqualitativehistopathologicalevaluation.Fromthecorrespondinghistologicalblock,threeconsecutivesectionswerepreparedforfurtherhistochemicalanalysisofconnectivetissueingrowth.

2.11.HistochemistryforconnectivetissueingrowthTheremainingthreeslideswerepreparedtovisualizeconnectivetissueingrowthwithintheimplantationbedusinghistochemicalstainingproceduresforAzan,Movat’spentachromeandSiriusred[26,27].Thesestainsarecommonlyusedforthedetectionofreticularandcollagenfibers.Azanstainindicatescollagenandreticularfibersblue.Siriusredhighlightscollagenfibersinred.Movat’spentachromestainscollagengreen.AllchemicalswerepurchasedfromSigma-Aldrichandusedwithoutfurtherpurification.

2.12.MorphologicalevaluationoftheinflammatoryresponseHistopathologicalevaluationwasconductedusingaNikonECLIPSE80imicroscope(Nikon,Japan)bytwoindependentexaminers(SG,CO)experiencedinhistomorphologicalanalysis,whowereblindedtotheexperimentalprotocol.Thehistologicalslidesoftheimplantationbed,aswellasoftheorgans,wereassessedqualitativelyforthefollowingcharacteristicfeatures:fibroticcapsulesaroundthebiomaterials,fibrosis,hemorrhage,necrosis,vascularization,neutrophils,lymphocytes,plasmacells,macrophages,giantcellsandscaffolddegradation.MicrophotographsweretakenusingaNikonDS-Fi1/Digitalcameraandadigitalsightcontrolunit(Nikon,Japan).

4

2.10.HistologicalpreparationoftheimplantationbedTheanimalswereeuthanizedbyanoverdoseofketamineandxylazinattheindicatedtimepoint.Immediatelyaftersacrifice,theimplantationbedwasexplantedtogetherwiththesurroundingperi-implantationtissueandfixedin4%bufferedformalinfor24hpriortohistologicalandhistochemicalanalyses.Thefixedtissuewascutintosevensegmentsofidenticaldimensionscoveringthemarginsandcenteroftheimplantationbed.Theyweredehydratedinaseriesofalcohols,transferredtoxyleneandembeddedin

Biomed.Mater.5(2010)025004SMGhanaatietal

(A)(B)(C)

(D)(E)(F)

Figure2.Confocallaserscanningmicroscopyimagesofcalcein-AM-stainedMG63(A)–(C)andhumanprimaryosteoblasts(D)–(F)on

R

-collagenscaffoldsafter3days(A),(D),7days(B),(E)and10days(C),(F).Theimagesshowthehumanserum-coatedBONITmatrix󰀁

excellentosteoblastadhesionandprogressivecolonizationoftheentirescaffold.Thelong-termcultures(C),(D),(E)demonstratesheet-likecellulargrowthonthebiomaterial.

2.13.Histomorphometry

ThehistomorphometricanalysiswasperformedusingthesoftwareNIS-Elements(Nikon,Tokyo,Japan)accordingtothemanufacturer’sinstructions.Briefly,imageswereobtainedwithaDS-Fi1/DigitalcameraconnectedtoanEclipse80ihistologicalmicroscope(Nikon,Tokyo,Japan),equippedwithanautomaticscanningtable(Prior,USA).Atotalscan(onelargeimageassembledfrom100–120imagesoftheregionofinterest,containingthebiomaterialanditscorrespondingperi-implanttissue)wastakenbyusinga100×magnificationandaresolutionof2500×1200pixels.Foreachanimal,theslidethatwasstainedwithH&Ewastakentodetectthehostvascularizationofthescaffold.UsingtheNIS-Elements‘annotationsandmeasurements’tool,thebloodvesselsoftheimplantationbedweremarkedseparatelybythe‘area’tool.Thetotalnumberofvesselsandtheirtotalareaoneachslideweredetermined.Thenumberofvesselspermm2wascalculatedfromthetotalvesselnumberfoundwithintheimplantationbed.Foreachtimepoint,ameannumberofvesselspermm2wasdetermined.Thetotalareaofvesselswasdeterminedandcalculatedasapercentageofthetotalareaoftheimplantationbed.Foreachtimepoint,ameanpercentageofvesselsarea/totalareawasdetermined.2.14.Statisticalanalysis

Quantitativedataarepresentedasmean±standarddeviation(SD)andaone-wayunivariateanalysisofvariance(ANOVA)followedbyLSDposthocassessmentwasappliedtocomparegroupsusingtheSPSS16.0.1software(SPSSInc.,Chicago,IL).DifferenceswereconsideredsignificantifP-valueswere

5

lessthan0.05(∗P<0.05),andhighlysignificantifP-valueswerelessthan0.01(∗∗P<0.01).TheSigmaPlot11.0software(SigmaPlot,SystatSoftwareInc.,Erkrath,Germany)wasusedforplottinggraphs.

3.Results

3.1.Invitroresults

Primaryhumanosteoblastsandthehumanosteosarcomacellline,MG63,wereaddedtohumanbloodserum-R

-collagenscaffoldstodeterminethecoatedBONITmatrix󰀁

cellbehavioronthebiomaterial.Previousexperimentswith

R

-collagenhaveshownendothelialcellsonBONITmatrix󰀁

thathumanbloodserumcoatingoptimizedendothelialcelladhesionandproliferationinshort-andlong-terminvestigationsonthisbiomaterial.Therefore,thiscoatingmethodwasalsousedinthisstudy[18].CellproliferationofHOSandMG63asshownbythetotalpopulationofcalcein-AM-stainedcellsrevealedexcellentgrowthandextensivecellspreadingafter10daysofcellcultivation(figure2).Osteoblastproliferationappearedtobemarkedsothatsinglecellscouldnotbedistinguishedafter10days.

SEMmicrographsrevealedregularcelladhesionandspreadingofindividualcells1dayafterseeding(figures3(A)and(D)),whichwascomparabletoosteoblastsgrownontissueculturepolystyrene.Duringthefollowing10days,regularcellgrowthandmorphologyofbothHOSandMG63on

R

-collagenweremaintained,withcellsshowingBONITmatrix󰀁

aflattened,spreadmorphologyandincreasinglyconfluentcellcoverage(figures3(B)and(E)).Completecoveragewasreachedafter10days(figures3(C)and(F)).

Biomed.Mater.5(2010)025004SMGhanaatietal

(A)(B)(C)

(D)(E)(F)

R

Figure3.ScanningelectronmicroscopeimagesofosteoblasticcellsonBONITmatrix󰀁-collagenscaffolds:(A)–(C)MG63and(D)–(E)

R󰀁primaryosteoblastsonhumanserum-coatedBONITmatrix-collagenafter1day(A),(D),7days(B),(E)and10daysofculture(C),(F).

Theultrastructuralimagesconfirmthetime-dependentformationofosteoblasticsheetsonthebiomaterial,withincreasinglossofindividualcelldemarcation,especiallyin(C),(E)and(F).

HOS,althoughMG63showednovisiblealkalinephosphatase(ALP)andosteopontinexpression(figure4).MorepreciseexaminationoftheimportantcalcificationfactorALPinMG63byreal-timePCRmeasurementshowedthatthelowlevel-expressedALPwasinfactsignificantlyup-regulatedindirectcomparisontoMG63grownontissueculturepolystyrene(2.71-foldALPincrease,SD±1.38,n=5pergroup,P<0.05).3.2.Invivoresults

R-Onday3oftheinvivoscaffoldapplication,BONITmatrix󰀁collagenimplantationresultedinafibrin-rich,peri-implantationtissuereactionwithathincelllayeraroundeachgranule(figure5(A)).Predominantly,macrophagesandfibroblastswithfewgranulocytescouldbeobservedinthiscelllayer(figure6(A)).Thenumberofthesecellsincreasedbetweenday3andday10(figures5(B)and6(B))andresultedinacell-richandwell-vascularizedconnectivetissueatday15afterimplantation(figures5(C)and6(C)).Thevascularizationoftheimplantationbedsignificantlyincreasedbetweenday3andday10with8vesselsmm−2andremainedinaplateau-likeconditionwith10vesselsmm−2uptoday15(figure7(A)).However,thetotalvesselareaunderwentasignificantthree-foldincreasebetweenday10andday15(figure7(B)).Betweenday15andday30,theconnectivetissuepenetratedthroughthegranules(figures5(D)and6(D))andvesselnumberaswellastheirpercentalcontributiontothevascularizationoftheimplantationbedsignificantlyincreased(figures7(A)and

R-(B)).Withintheobservedstudytimepoints,BONITmatrix󰀁collagenunderwentcell-mediateddegradation,startingfromthegranuleperipherytowarditscenterandresultingingranule

Figure4.RT-PCRanalysisofosteoblasticgeneexpressionon

R

-collagenscaffolds.misamolecularserum-coatedBONITmatrix󰀁

basepair(bp)sizemarker,nisthenegativewatercontrolwithoutcDNAtemplate.Primaryosteoblastsgivethefullrangeofexpressedmarkergenes,whereasthepermanentcellline,MG63,failstoshowvisibletranscriptionofalkalinephosphataseandosteopontin.

RT-PCRanalysisofHOSandMG63culturedon

R

-collagenfor10dayswasperformedforBONITmatrix󰀁

osteoblast-specificgenes,suchasalkalinephosphatase,collagen-I,osteocalcin,osteonectin,osteopontin,TGF-β1andFGF-2(figure4).HOSshowedregularexpressionofallgenesofinterest10daysafterseeding,withcellsbeing

R

-collagen.inthestateoftotalconfluenceonBONITmatrix󰀁

GeneexpressionofMG63cellswassimilarcomparedto

6

Biomed.Mater.5(2010)025004SMGhanaatietal

(A)(B)

(C)(D)

RR

Figure5.TissueintegrationanddissolutionofBONITmatrix󰀁-collagen(BM󰀁)attheobservedtimeperiodof30days.(H&E,100×magnification)(A=day3;B=day10;C=day15andD=day30afterimplantation).Scalebar=10μm.

dissolution.Macrophagesandafewmulti-nucleatedgiant

R

cellswereinvolvedinBONITmatrix󰀁-collagendegradationandexpressedpredominantlyTRAP(tartrate-resistantacidphosphatase)(figure6(E)).

4.Discussion

Theaimofthisstudywastodeterminewhetherthisnovelcalciumphosphate/SiO2–xerogelcollagencompositeaffectsHOBmorphology/functioninvitroanditsbiocompatibilityinvivo.ThisincludedexaminingtheadherenceandspreadingofprimaryHOBsandtheosteoblast-likecellline,MG63,to

R

studytheabilityofBONITmatrix󰀁-collagentoserveasanappropriatesubstrateforosteoblasts.Furthermore,expressionofspecificosteoblasticgeneswasexaminedatthetranscriptionlevel.Ontheotherhand,aninvivostudyofupto30daystestedthescaffoldbiocompatibilityregardingvascularizationanddegradationdeterminedhistologically.Thetissueresponsetothebiomaterialwascharacterizedbyquantifyingvascularstructuresinandaroundtheimplantaswellasdemonstratingphagocyticcellactivityinvolvedinscaffolddegradation.

PreviousstudieswithhumanprimarymicrovascularandmacrovascularECsshowedastrongaffinityofECforhuman

R

bloodserum-coatedBONITmatrix󰀁-collagenscaffoldswithsuccessfulcellgrowth[18].Withoutserumcoating,cellsgenerallydetachedwithinashorttime,eventuallyleavingnocellsonthescaffoldsurface.Thisphenomenonhasoftenbeenobservedandindicatesthatcellsgenerallyneedanappropriatematrixtoadhereto[28].Althoughmanyotherbiomaterialsareoftencoatedwithfibronectin,collagen,

7

gelatinorlaminin[29–31],noneoftheseextracellularmatrix(ECM)proteinsweresufficienttoinducesuccessfuladhesion

R

-collagen.andproliferationofanyECtypeonBONITmatrix󰀁

Tocreatehomologousgrowthconditionsinourinvitrostudy,

R

-collagenwascoatedwithhumanbloodserumBONITmatrix󰀁

inthisstudyaswell.Coatingwithhumanbloodserumactually

R-mimicsthenaturalcoatingprocesswhenBONITmatrix󰀁collagenisimplantedintothebody.Mixingofthebiomaterialwithhostbloodbeforeimplantationisroutinelypracticedby

R

already[19].surgeonsforBONITmatrix󰀁

R󰀁

BONITmatrixisproducedviaasol–gelprocesswithsilicaasanadjuvant.Asoneofthefirstbonesubstitution

R

hascombinedtheprofitablematerials,BONITmatrix󰀁

propertiesofcalciumphosphatesandbioglasseswithinaSiO2networkasanexcellentstartingpointfortheformationofautogenousapatite[32].Unlikesilicateglassesorquartz,solublesilicateinducesbone-likeapatite[33].AccordingtoLehninger,siliconisascribedanosteoinductivefunctionwhichcouldstimulateneo-osteogenesisadditionally[34].Indeed,silicatesubstitutionforphosphateionsinhydroxylapatitehasbeenshowntoenhanceosteoblastcellactivity,boneappositionduringosseousintegrationandanincreasedboneremodelingatthebone/hydroxylapatiteinterface[35–37].On

R

,ithasbeendiscoveredthatSiO2promotesBONITmatrix󰀁

osteoblasticdifferentiationofhumanmesenchymalstemcellswithoutosteogenicdifferentiationadditivesasshownbyincreasedALPactivityandexpressionofosteogenicgenes

R

byBecker[38].MoredetailedstudiesonBONITmatrix󰀁

etalverifiedanupregulatedmRNAprofileofdifferentiationmarkerssuchasosteocalcin,osteopontinandcollagen-Iinthe

Biomed.Mater.5(2010)025004SMGhanaatietal

(A)(B)

(C)(D)

(E)

Figure6.Tissueintegrationdemonstratingthehighvessel(redarrowheads)densityandincreasedvessellumenareawithintheconnective

RR

tissuearoundBONITmatrix󰀁-collagen(BM󰀁)granulesandTRAP-positivemacrophages(blackarrowheads)involvedinbiomaterial

degradation(400×magnification)(A=day3;B=day10;C&E=day15andD=day30afterimplantation);(A)H&E-staining;(B)–(D)Movat’spentachromestaining;(E)tartrate-resistantacidphosphatasestaining).Scalebar=20μm.

osteoblasticcelllineMG63[39].AlthoughMG63arenativelylowintheexpressionoftheimportantcalcificationfactoralkalinephosphatase,wecouldshowsimilar,statisticallysignificantresultsfortheexpressionofALPwithinMG63on

R

-collagencomparedtoMG63grownontissueBONITmatrix󰀁

culturepolystyrenefor10days.Thelatterresultdemonstrates

R

-collagenonMG63.astimulatingeffectofBONITmatrix󰀁

R

-collagenOsteoblastsgrowingonBONITmatrix󰀁

exhibitednormalcellmorphologyafterseeding,asvisualizedbyCLSMandSEM.SimilarosteoblasticmorphologywasshownbyJonesetalonabioactiveglassscaffold[40].RT-PCRanalysisshowedregularexpressionofosteoblast-specificgenesofinterestinprimaryosteoblasts,mostimportantlyalkalinephosphataseandcollagen-I,bothbeingimportantformatrixmineralizationandbonestrength.Alkalinephosphataseandcollagen-Irepresenttwoparameterstypicallyusedasmarkersofosteoblasticdifferentiation.

8

TGF-β1andFGF-2,stronglyexpressedinbothHOSandMG63,representtwoabundantgrowthfactorsandstimulatorsforosteogenesisinhumanbone.TGF-β1andFGF-2influencenormalskeletaldevelopmentbyplayingacriticalroleininducingmesenchymalcelldifferentiationtoeitherchondrocytesorosteoblasts.Generally,osteoblastdifferentiationispromoted,withcorrespondingmatrixformationandmineralization[41,42].

MG63showedsimilarresultsintheoverallgeneexpressionpatterncomparedtoHOS.Theabilitytoexpressosteopontin,atransformation-associatedcelladhesionphosphoprotein,wasnotactivatedinMG63.MG63cellsarerepresentativesofearlyundifferentiatedosteoblastsandconstituteasubclassofosteoblastswhichexpressosteopontinlaterintheosteoblastdevelopmentalsequence[43,44].Osteocalcin,alsoexpressedinbothcelltypes,representsatypicalproteinwhichbindstothemineralphaseofbone.

Biomed.Mater.5(2010)025004SMGhanaatietal

(A)

(B)

Figure7.Histomorphometricmeasurements:(A)numberof

vesselspermm2and(B)distributionofvesselscalculatedfromthetotalareaoftheimplantationbed(in%)(∗∗P<0.01).

Theglycoprotein,osteonectin,isanotherveryabundantnon-collagenousproteinofdevelopingbonewhichhasahighaffinityforbindingcalcium,hydroxyapatite,collagenandthrombospondin[45–48]andisthereforeimportantforlinking

R

upwithBONITmatrix󰀁-collagen.

Theexpressionofthesemultipleproteinsisinducedduringosteoblasticdifferentiationinastepwisefashion,suggestiveofseveralregulatoryfactors.ItisassumedthatalkalinephosphataseandtypeIcollagenappearearlyduringthecommitmenttotheosteoblasticphenotype,whereasosteopontinandosteocalcinareexpressedlaterduringosteoblasticdifferentiation[41,49].Asallofthesegenesare

R

expressedinprimaryosteoblastsonBONITmatrix󰀁-collagen,thissuggeststhatthecellsonthescaffoldarepresentina

R

laterstateofdifferentiationandBONITmatrix󰀁-collagenfullysupportsosteoblasticgeneexpression.

Invivo,theimplantationofabiomaterialinducesaninflammatoryresponsewhichisspecificforthebiomaterial[12–15,50].Thisinflammatoryresponseisgenerallydependentonparticlesizeandparticlemorphologyandactivates,besidesothercascades,therecruitmentofmononuclearcells[50–57].Itisknownthatmultinucleatedgiantcellsareformedfrommacrophageswhenmononuclearcellscannotcontributetobiomaterialdegradation[52,58–61].Thefusionofmultiplemononuclearcellsintogiantcellsdemonstratestheeffortofthehosttissuetoadapttothespecificcharacteristicsofthebiomaterialparticleasaforeignbody[12,14,58].Consequently,theformationofmultinucleatedgiantcellsismaterialdependent[12].

9

Theinvivopartofthestudyshowedthatpredominantlymacrophagesandonlyafewmultinucleatedgiantcells

R

wereinvolvedintheBONITmatrix󰀁-collagendegradationprocessatalltimes.Theseresultshighlighttheability

R

ofBONITmatrix󰀁-collagen,asyntheticallyderivedbonesubstitute,toundergodegradationwithouttheinvolvementofmultinucleatedgiantcells.Tore-emphasizethis,multinucleatedgiantcellsdonotnecessarilyhavetobeinvolvedinthedegradationprocessofsyntheticallyderivedbonesubstitutes.Obviously,biomaterialslike

R󰀁

BONITmatrix-collagencanbedevelopedwhichcanundergotheobserveddissolutionintonumeroussmallparticlesenablingtissueingrowthbetweentheparticles.Accordingly,

R

BONITmatrix󰀁-collagencanbeconsideredasextremelybiocompatible,asitiswelltoleratedbythehostwithoutthenecessityofanexacerbatedinflammatoryresponse(i.e.theformationofmultinucleatedgiantcells)foritsdegradation.

However,thepresenceofmultinucleatedgiantcellsandespeciallytartrate-resistantacidphosphatase(TRAP)-positivemultinucleatedgiantcells,knownasosteoclast-likecells,mightindicatethebioactivityofthebonesubstitutegranuleanditsstabilitysimilartoboneultrastructure[62,63].Whenconsideringthegenerationofmultinucleatedgiantcells,especiallyofTRAP-positive,osteoclast-likecellsasanindicatorofthestabilityofabiomaterial,and,inthe

R

caseofaninductionofosteoclast-likecells,BONITmatrix󰀁-collagenwouldhavebeenstableandpotentiallyabletoserveasaplaceholderforlongerthan15daysasobservedinthisstudy.

Sufficientvascularizationiscrucialtoeverysuccessfulbiomaterialintegration.Thefastdegradationofthisbiomaterialresultedintheinfluxofavessel-richconnectivetissuewithcontinuousvascularizationwithintheimplantationbed.Theseresultshighlighttheabilityofthisbonesubstitutetocontributetotheregenerationofthedefectbyinducinganactivegranulationtissue,withitsvascularizationincreasingovertheobservedtimeperiod.Theincreaseofvesseldiametersreflectsthematurityofthenewlyformedconnectivetissue.Thesefindingsalsoindicatethatforastablevascularizationoftheimplantationbed,boththeincreaseinvesselnumbersandthematurityofthevesselswithintheimplantationbedareimportant.

5.Conclusion

ThiscombinedinvitroandinvivostudydemonstratedthattheobservedSiO2-basedcalciumphosphate–collagenbonesubstitutematerialexhibitssufficientpropertiestobeconsideredasabiocompatiblebiomaterialforcell-basedtissueengineering.Theinvitroresultsshowedthatthephysico-chemicalcharacteristicsofthismaterialallowmesenchymalcellssuchasosteoblaststosurviveanddeploytheirtypicalcharacteristicsbyexpressingosteoblast-specificgenes.Theinvivopartofthisstudyrevealedthatbiomaterialdegradationisexecutedmainlybymononuclearcells.Itsfastdegradationresultedinavessel-richconnectivetissue,inwhichnotonlythevesselnumberbutalsothevesseldiameterincreasedinthecourseoftime.

Biomed.Mater.5(2010)025004SMGhanaatietal

Acknowledgments

ThisworkwassupportedbyaBMBFgrantno0313405C.TheauthorsthankDOTGmbH,Rostock,Germany,fortheir

R

generoussupplyofBONITmatrix󰀁-collagenscaffolds.ThanksalsogotoMrsASartoris,MrsBPavic,MrBBondarandMrRTsaryk,fortechnicalassistance.TheyarealsogratefultoMsMM¨ullerandMsKMolterfortheirdocumentationofthescanningelectronmicroscopemicrographs.

[19]

[20]

[21]

References

[1]AsahinaI,SetoI,OdaM,MarukawaE,ImranulA

andEnomotoS1999BoneEngineering(Toronto:EmSquared)p526

[2]YoungerEMandChapmanMW1989Morbidityatbone

graftdonorsitesJ.Orthop.Trauma3192–5

[3]ArringtonE,SmithW,ChambersH,BucknellAand

DavinoN1996ComplicationsofiliaccrestbonegraftharvestingClin.Orthop.Relat.Res.329300–9

[4]MischCandDietshF1993Bone-graftingmaterialsinimplant

dentistryImplant.Dent.2158–67

[5]KasperkC,EwersR,SimonsBandKasperkR1988

Algae-derived(phycogene)hydroxylapatite.Acomparativehistologicalstudy[correctedandissuedwithoriginalpaging]Int.J.OralMaxillofac.Surg.17319–24

[6]NystromE,LegrellPE,ForssellAandKahnbergK1995

Combineduseofbonegraftsandimplantsintheseverelyresorbedmaxilla.PostoperativeevaluationbycomputedtomographyInt.J.OralMaxillofac.Surg.2420–5[7]ParikhS2002Bonegraftsubstitutes:past,present,future

J.Postgrad.Med.48142–8

[8]ErbeE,MarxJ,ClineffTandBellincampiL2001Potentialof

anultraporousbeta-tricalciumphosphatesyntheticcancellousbonevoidfillerandbonemarrowaspiratecompositegraftEur.SpineJ.10(Suppl2)S141–6[9]McAndrewM,GormanPandLangeT1988Tricalcium

phosphateasabonegraftsubstituteintrauma:preliminaryreportJ.Orthop.Trauma2333–9

[10]MetsgerD,DriskellTandPaulsrudJ1982Tricalcium

phosphateceramic—aresorbableboneimplant:reviewandcurrentstatusJ.Am.Dent.Assoc.1051035–8

[11]YoshikawaHandMyouiA2005Bonetissueengineeringwith

poroushydroxyapatiteceramicsJ.Artif.Organs8131–6

[12]AndersonJM,RodriguezAandChangDT2008Foreign

bodyreactiontobiomaterialsSemin.Immunol.2086–100

[13]WilsonCJ,CleggRE,LeavesleyDIandPearcyMJ2005

Mediationofbiomaterial–cellinteractionsbyadsorbedproteins:areviewTissueEng111–18

[14]HuWJ,EatonJW,UgarovaTPandTangL2001Molecular

basisofbiomaterial-mediatedforeignbodyreactionsBlood981231–8

[15]JenneyCRandAndersonJM2000Adsorbedserumproteins

responsibleforsurfacedependenthumanmacrophagebehaviorJ.Biomed.Mater.Res.49435–47

[16]vonWilmowskyCetal2008Effectsofbioactiveglassand

beta-TCPcontainingthree-dimensionallasersinteredpolyetheretherketonecompositesonosteoblastsinvitroJ.Biomed.Mater.Res.A87896–902

[17]Bienengr¨aberV,GerberT,TrykovaT,KundtGandHenkelK

2004EineinnovativimSol-Gel-Prozeßhergestellte,hochpor¨oseSiliziumoxidkeramikzum

Knochenersatz-In-VivoLangzeitergebnisseMater.-Wiss.Werkst.Tech.35234–9

[18]ThimmB,UngerR,NeumannHandKirkpatrickC2008

Biocompatibilitystudiesofendothelialcellsonanovel

10[22]

[23]

[24][25]

[26][27][28]

[29][30]

[31]

[32][33]

[34][35]

[36]

calciumphosphate/SiO2–xerogelcompositeforbonetissueengineeringBiomed.Mater.315007

HoppM,RogaschewskiS,KrahlTandBiffarR2005

KlinischeErfahrungenundmaterialkundlicheAspektebeiderAnwendungeinesneuenxerogenen

AugmentationsmaterialsImplantologie13301–15TraykovaT,B¨ottcherR,NeumannH,HenkelK,Bienengr¨aberVandGerberT2004Silica/calcium

phosphatesol–gelderivedbonegraftingmaterial—fromanimalteststofirstclinicalexperienceKeyEng.Mater.254–56679–82

HenkelK,GerberT,DietrichWandBienengr¨aberV2004Novelcalciumphosphateformulaforfillingbonedefects.Initialinvivolong-termresultsMundKieferGesichtschir8277–81

GerberT,KnoblichB,TraykovaT,Holzh¨uterG,D¨orflingPandHenkelK2001Entwicklung,invitroundinvivotestseineshochpor¨osenKnochenersatzmaterialsOsteologie10175–83

HenkelK,GerberT,DietrichW,KundtGandBienengr¨aberV2004Sol–gelderivedcalciumphosphateceramicsintreatingbonedefects—abreakthroughtoartificialbonesubstitutes?AnanimalstudyOsteologie1357–64R¨uegseggerP,KollerBandM¨ullerR1996A

microtomographicsystemforthenondestructiveevaluationofbonearchitectureCalcif.TissueInt.5824–9FuchsS,GhanaatiS,OrthC,BarbeckM,KolbeM,HofmannA,EblenkampM,GomesM,ReisRandKirkpatrickC2009Contributionofoutgrowth

endothelialcellsfromhumanperipheralbloodoninvivovascularizationofbonetissueengineeredconstructsbasedonstarchpolycaprolactonescaffoldsBiomaterials30526–34

MovatH1955Demonstrationofallconnectivetissueelementsinasinglesection;pentachromestainsAMAArch.Pathol.60289–95

SweatF,PuchtlerHandRosenthalS1964SiriusredF3baasastainforconnectivetissueArch.Pathol.7869–72

ChuC,LuA,LiszkowskiMandSipehiaR1999EnhancedgrowthofanimalandhumanendothelialcellsonbiodegradablepolymersBiochim.Biophys.Acta1472479–85

AndersonJ,PriceT,HansonSandHarkerL1987Invitroendothelializationofsmall-calibervasculargraftsSurgery101577–86

UngerR,HuangQ,PetersK,ProtzerD,PaulD

andKirkpatrickC2005Growthofhumancellsonpolyethersulfone(PES)hollowfibermembranesBiomaterials261877–84

KaehlerJ,ZillaP,FasolR,DeutschMandKadletzM1989PrecoatingsubstrateandsurfaceconfigurationdetermineadherenceandspreadingofseededendothelialcellsonpolytetrafluoroethylenegraftsJ.Vasc.Surg.9535–41B¨ottcherR,NeumannH-GandBienengr¨aberV2004

BONITmatrixalsoptimalerKnochenersatzIndustryReportDentalTribuneGermanEdition19

LiP,OhtsukiC,KokuboT,NakanishiK,SogaNand

deGrootK1994Theroleofhydratedsilica,titania,andaluminaininducingapatiteonimplantsJ.Biomed.Mater.Res.287–15

LehningerA1987PrinzipienderBiochemie(Berlin:deGruyter)pp298–301

PorterA,PatelN,SkepperJ,BestSandBonfieldW2003Comparisonofinvivodissolutionprocessesin

hydroxyapatiteandsilicon-substitutedhydroxyapatitebioceramicsBiomaterials244609–20

PatelN,BestS,BonfieldW,GibsonI,HingK,DamienEandRevellP2002Acomparativestudyontheinvivobehaviorofhydroxyapatiteandsiliconsubstituted

Biomed.Mater.5(2010)025004SMGhanaatietal

[37][38][39]

[40]

hydroxyapatitegranulesJ.Mater.Sci.Mater.Med.131199–206

GibsonI,BestSandBonfieldW1999Chemical

characterizationofsilicon-substitutedhydroxyapatiteJ.Biomed.Mater.Res.44422–8M¨ullerPetal2008Calciumphosphatesurfacespromote

osteogenicdifferentiationofmesenchymalstemcellsJ.CellMol.Med.12281–91BeckerP,TellerM,L¨uthenF,NebeB,BuhlnheimU,RychlyJandNeumannH2004BONITmatrix—bone

regenerationmaterialfordentalapplicationBIOmaterialien10–1

JonesJ,TsigkouO,CoatesE,StevensM,PolakJandHenchL2007Extracellularmatrixformationandmineralizationon[51]

[52][53][54]

inflammatorygeneexpressionbygiantcellsduringtheforeignbodyreactionJ.Biomed.Mater.Res.A83879–86ZhuXD,ZhangHJ,FanHS,LiWandZhangXD2009EffectofphasecompositionandmicrostructureofcalciumphosphateceramicparticlesonproteinadsorptionActaBiomater.(PMID:19857608)

SunJS,TsuangYH,ChangWH,LiJ,LiuHCandLinFH1997EffectofhydroxyapatiteparticlesizeonmyoblastsandfibroblastsBiomaterials18683–90

DumitriuS2001PolymericBiomaterials,RevisedandExpanded(BocaRaton,FL:CRCPress)pp430–2GhanaatiS,OrthC,DeisingerU,DetschR,BoomsP,UngerR,ZieglerGandKirkpatrickC2008The

compositionofbonesubstitutematerialsinfluencestheiraphosphate-freeporousbioactiveglassscaffoldusingprimaryhumanosteoblast(HOB)cellsBiomaterials281653–63

[41]KanaanRandKanaanL2006Transforminggrowthfactorbeta1,boneconnectionMed.Sci.Monit.12RA164–9[42]MarieP,DebiaisFandHayE2002RegulationofhumancranialosteoblastphenotypebyFGF-2,FGFR-2andBMP-2signalingHistol.Histopathol.17877–85[43]CarmelietG,NysGandBouillonR1997MicrogravityreducesthedifferentiationofhumanosteoblasticMG-63cellsJ.BoneMiner.Res.12786–94

[44]

SteinGandLian/J1993MolecularmechanismsmediatingproliferationdifferentiationinterrelationshipsduringprogressivedevelopmentoftheosteoblastphenotypeEndocr.Rev.14424–42

[45]RombergR,WernessP,LollarP,RiggsBandMannK1985IsolationandcharacterizationofnativeadultosteonectinJ.Biol.Chem.2602728–36

[46]

ClezardinP,MalavalL,EhrenspergerA,DelmasP,

DechavanneMandMcGregorJ1988ComplexformationofhumanthrombospondinwithosteonectinEur.J.Biochem.175275–84

[47]TermineJ1988Non-collagenproteinsinboneCibaFoundSymp.136178–202

[48]TermineJ,KleinmanH,WhitsonS,ConnK,McGarveyMandMartinG1981Osteonectin,abone-specificproteinlinkingmineraltocollagenCell2699–105

[49]RodanGandNodaM1991GeneexpressioninosteoblasticcellsCrit.Rev.Eukaryot.GeneExpr.185–98[50]

LuttikhuizenDT,DankersPY,HarmsenMCand

vanLuynMJ2007Materialdependentdifferencesin

bioactivityinvivoBIOmaterialien990

[55]

GhanaatiS,OrthC,HoffmannC,BarbeckM,UngerR,

SaderR,PetersFandKirkpatrickC2009Influenceofsize,structureandporosityonthebiologicalresponseof

b-tricalciumphosphatebiomaterialsinasubcutaneousratimplantmodelJ.RegenerativeMed.4182–3

[56]GhanaatiS,OrthCandKirkpatrickC2008Themorphologyofb-TCP-basedbiomaterialsinfluencestheirbioactivityinvivoTissueEng.A14917–8

[57]XiaZandTriffittJT2006AreviewonmacrophageresponsestobiomaterialsBiomed.Mater.1R1–9

[58]CannonGJandSwansonJA1992ThemacrophagecapacityforphagocytosisJ.CellSci.101Pt4907–13

[59]

JaySM,SkokosEA,ZengJ,KnoxKandKyriakidesTR2009MacrophagefusionleadingtoforeignbodygiantcellformationpersistsunderphagocyticstimulationbymicrospheresinvitroandinvivoinmousemodelsJ.Biomed.Mater.Res.A(PMID:19536825)

[60]

ShenM,GarciaI,MaierRVandHorbettTA2004Effectsofadsorbedproteinsandsurfacechemistryonforeignbodygiantcellformation,tumornecrosisfactoralphareleaseandprocoagulantactivityofmonocytesJ.Biomed.Mater.Res.A70533–41

[61]RatnerB1997BiomaterialsScience:AnIntroductiontoMaterialsinMedicine(Amsterdam:Elsevier)pp170–3[62]DeGrootK1980BioceramicsconsistingofcalciumphosphatesaltsBiomaterials147–50

[63]

XuW,HolzhuterG,SorgH,WolterD,LenzS,GerberTandVollmarB2009Earlymatrixchangeofa

nanostructuredbonegraftingsubstituteintheratJ.Biomed.Mater.Res.B91692–9

11