Cellular differentiation, bioactive and mechanicalproperties of experimental light-curing pulpprotection materials

Available

online

at

www.sciencedirect.com

ScienceDirect

j ou rn a l h om ep a ge :w w w . i n t l . e l s e v i e r h e a l t h . c o m / j o u r n a l s / d e m a

Cellular

differentiation,

bioactive

and

mechanical

properties

of

experimental

light-curing

pulp

protection

materials

Salvatore

Sauro

_,

_Ashvin

_Babbar

_,

_Borzo

_Gharibi

_,

Victor

Pinheiro

Feitosa

_,

_Ricardo

_Marins

_Carvalho

_,

Lidiany

Karla

Azevedo

Rodrigues

_,

_Avijit

_Banerjee

_,

_Timothy

_Watson

a_{DentalBiomaterialsandMinimallyInvasiveDentistry,DepartmentodeOdontologia,FacultaddeCienciasdela}

Salud,CEU-CardenalHerreraUniversity,AlfaradelPatriarca,46115Valencia,Spain

b_{TissueEngineeringandBiophotonicsResearchDivision,King’sCollegeLondonDentalInstitute,King’sHealth}

Partners,London,UK

c_{SchoolofDentistry,FacultyofPharmacy,DentistryandNursing,FederalUniversityofCeará,Fortaleza,Brazil} d_{PauloPicanc¸oSchoolofDentistry,Fortaleza,Brazil}

e_{DepartmentofOralBiologicalandMedicalSciences,DivisionofBiomaterials,FacultyofDentistry,TheUniversity}

ofBritishColumbia,Vancouver,BCV6T1Z3,Canada

f_{DepartmentofConservative&MIDentistry,King’sCollegeLondonDentalInstitute,King’sHealthPartners,}

London,UK

a

r

t

i

c

l

e

i

n

f

o

Received1November2017 Receivedinrevisedform 31December2017 Accepted27February2018 Keywords: Bioactivity Biocompatibility Fracturetoughness Flexuralstrength Celldifferentiation Pulpprotection

a

b

s

t

r

a

c

t

Objective.Materialsforpulpprotectionshouldhavetherapeuticpropertiesinorderto stim-ulateremineralizationandpulpreparativeprocesses.Theaimofthisstudywastoevaluate themechanicalproperties,biocompatibility,celldifferentiationandbioactivityof experi-mentallight-curableresin-basedmaterialscontainingbioactivemicro-fillers.

Methods.Fourcalcium-phosphosilicatemicro-fillerswerepreparedandincorporatedintoa resinblend:1)Bioglass45S5(BAG);2)zinc-dopedbioglass(BAG-Zn);3)␤TCP-modified cal-ciumsilicate(␤-CS);4)zinc-doped␤-CS(␤-CS-Zn).Theseexperimentalresinsweretested forflexuralstrength(FS)andfracturetoughness(FT)after24hand30-daystoragein sim-ulatedbodyfluid(SBF).CytotoxicitywasevaluatedusingMTTassay,whilebioactivitywas evaluatedusingmineralizationandgeneexpressionassays(Runx-2&ALP).

Results.ThelowestFSandFTat24hwasattainedwith␤-CSresin,whilealltheothertested materialsexhibitedadecreaseinFSafterprolongedstorageinSBF.␤-CS-Znmaintaineda stableFTafter30-daySBFaging.Incorporationofbioactivemicro-fillershadnonegative effectonthebiocompatibilityoftheexperimentalmaterialstestedinthisstudy.The inclu-sionofzinc-dopedfillerssignificantlyincreasedthecellularremineralizationpotentialand expressionoftheosteogenicgenesRunx2andALP(p<0.05).

∗ _{Correspondingauthorat:DentalBiomaterials,PreventiveandMinimallyInvasiveDentistry,DepartamentodeOdontología,Facultadde}

CienciasdelaSalud,UniversidadCEU-CardenalHerrera,C/DelPozos/n,AlfaradelPatriarca,46115Valencia,Spain. E-mailaddress:[email protected](S.Sauro).

https://doi.org/10.1016/j.dental.2018.02.008

Significance.Theinnovativematerialstestedinthisstudy,inparticularthosecontaining ␤-CS-ZnandBAG-Znmaypromotecelldifferentiationandmineralization.Thus,these mate-rialsmightrepresentsuitabletherapeuticpulpprotectionmaterialsforminimallyinvasive andatraumaticrestorativetreatments.

©2018TheAcademyofDentalMaterials.PublishedbyElsevierLtd.Allrightsreserved.

1.

Introduction

The operative treatment of deep carious lesions can be challenging especially when approaching the pulp as the increasedriskofpulpexposureduringexcavationcanreduce theprobabilityofpulpsurvival[1].Anidealpulpprotection materialforsuchscenariosshouldbehighlybiocompatible andbioactive[2,3].Thisisespeciallyrelevantwiththe con-temporaryminimallyinvasiverestorativephilosophywhere contaminated(caries-infected)dentinisremovedfromdeep cavity selectively,retaining mostof the demineralized but repairable(caries-affected)dentinforpotential remineraliza-tionaswellasavoidingpulpexposure[4,5].Indeed,clinicians areincreasinglyrelyingonrestorativeion-releasingmaterials suchcalciumsilicatecementstosealtherestoredinterfaces inordertohelpremineralizethecaries-affectedtissues[6].

Incertainclinicalcases, indirectand/ordirectpulp pro-tection may help maintain pulp sensibility by facilitating healing/repair.Materialsusedforthispurposeshouldinteract with the pulp cells to stimulate the formation of repar-ative dentin [7–9]. Calcium silicate and MTA-like cements have been used as they can encourage remineralization anddentinbridgeformationwithno/minimalinflammatory pulpresponse[9–11].However,theiruseasabio-interactive restorativematerialislimitedduetoshortcomingsintheir mechanical properties, setting time and dissolution rate

[12,13]. It is hypothesized that the formulation of

resin-modified bioactive cements might present a solution to combinethe best ofboth resin-based technologywith the bioactivityofsuchcements. Indeed,resin-modifiedcalcium silicatecement–basedmaterialssuchasTheracalLC(BISCO, Chicago,IL,USA),alight-curablematerialadvocatedfordirect andindirectpulpprotection,showedgreatercompressiveand flexuralstrengthscomparedtoconventionalcalciumsilicate cements,beingmoreabletoresistfractureduringimmediate placement ofa definitive overlying restoration [14]. Never-theless,sucharesin-modifiedcalciumsilicatecement–based materialdemonstratedareductionincellularmetabolismand proteinexpression.Itwasalsoexhibitedgreatercytotoxicity comparedtoconventionalcalciumsilicatecements[15].

Sodiumcalciumphosphosilicates(e.g.Bioglass45S5,BAG) have been used successfully in orthopedics as regenera-tive materialsfor bone [8,16]. In particular, BAG has been shown to induce calcium-phosphate precipitation, subse-quentlyconverting tohydroxyapatite-likecrystallites[8,17]. Within dentistry, BAG is used in toothpastes and also as powdersfordentalair-abrasion/polishingtoremineralizethe dentalhardtissuesaswellastreatingdentin

hypersensitiv-ity[18,19].Bioactiveglassesaresocalledduetotheirinloco

remineralizationoftissues,buttheycannotbeusedinasa

definitivedentalrestorativematerialunlessincorporatedinto aresin-basedmatrix[20,21].

Theaimofthisstudywastoevaluatethemechanical prop-erties,biocompatibility,celldifferentiationandbioactivityof experimentallight-curableresin-basedmaterialscontaining bioactivemicro-fillersfortheirpotentialuseasindirectpulp protection materials, after storage in simulated body fluid (SBF).Themechanicalpropertieswereassessedthroughthe evaluationoftheirflexuralstrength(FS)andfracture tough-ness(FT).Thecytotoxicityofthetestedmaterialswastested using MTTassay, whilecelldifferentiation and mineraliza-tionwereassessedusinggeneexpressionassays(Runx-2and ALP). Thebioactivity ofthetested materialswas evaluated usingRamanspectroscopyandscanningelectronmicroscopy (SEM).Thenullhypothesestestedinthisstudywerethatthe additionofbioactivemicro-fillers:1)wouldhavenoeffecton mechanicalpropertiesoftheexperimentalresin-based mate-rialstested;2)wouldneitherincreasecytotoxicitynorinduce differentiation in primaryhuman mesenchymal stem cells (MSCs).

2.

Materials

&

methods

2.1. Formulationoftheresin-basedbioactivematerials A control filler-free resin (RES) was made using three hydrophobic monomers (55wt% urethane dimethacry-late and, 4.5wt% bisphenol A diglycidildimethacrylate, 10.5wt% triethylene glycol dimethacrylate, Sigma–Aldrich, Gillingham, UK), 20wt% 2-hydroxyethyl methacrylate (Sigma–Aldrich), 8.5wt% absolute ethanol (Sigma–Aldrich) and a photo-initiating complex comprising 0.5wt% cam-phorquinone/1.0wt% ethyl 4-dimethylaminobenzoate (Sigma–Aldrich).Theexperimentalion-releasingresinswere formulatedusing40vol%micro-fillerand60vol%resinblend

[21]. Four experimental light-curable resin-based materials containing tailored bioactive micro-fillers were formulated asdescribedbySauroetal.[21].Inbrief,Bioglass45S5(BAG) micro-filler (<20␮m size) was sintered and incorporated within the composition of the control light-curable resin blend.Thesecondmicro-filler,BAG-Zn(<20␮m),wascreated bymodifyingthe compositionofthe BAGwith20wt% zinc oxide(ZnO:Sigma–Aldrich)anda10wt%polycarboxylicacid solution(PAA:Mw1800;Sigma–Aldrich)andfinally

incorpo-ratedintothelight-curableresinblend.Thethirdmicro-filler (␤-CS) was formulated by modifying the composition of a type I ordinary Portland cement (OPC: Italcementi Group, Cesena, Italy) by adding 10wt% ␤-tri-calcium phosphate [␤TCP:Ca3(PO4)2(Sigma–Aldrich)].Thecementwasmixedin

inanincubatorat37◦Cfor24h.Thesetcementwas subse-quentlygroundandsievedasdescribedpreviously[30]until <20␮m particle size was achieved. The fourth micro-filler (␤-CS-Zn)usedinthisstudy wascreated bymixing70wt% OPC, 20wt%ZnOand 10wt%␤TCP inH2O (ratio2:1).After

setting for 24h at 37◦C, the micro-fillerwas treated(ratio 1:1)inapolymerflaskwitha10vol%polyacrylicacid(PAA) solution for15min (45rpm) and incubated in afurnace at 40◦Cfor12h.Theresultingproductwasfinallyre-groundand sieved(<20␮m).Theselattermicro-fillerswereincorporated alsointothelight-curableresinblendaspreviouslydescribed. 2.2. Flexuralstrengthtest(FS)

Light curing resin-based specimens prepared in accor-dance with ISO 4049 (25mm×2mm×2mm) may receive a non-uniform distribution of energy during the photo-polymerization procedure, which may in turn, affect their flexural strength. Therefore, smaller specimen (10mm×2mm×2mm)wereusedinthisstudytoreducethe variableeffectofpolymerizationasdescribed inaprevious study [22]. The experimental resin-based materials were insertedintorectangularsiliconemolds.Aglassslidealong witha plastic stripwere placedon top toachieve parallel surfaceswithnoairbubbles.Thespecimenswerelight-cured (>1000mW/cm2_{)frombothsidesfor1minusingaLEDlight}

curingunit(Litex695,DentamericaInc.,Industry,CA,USA). Twelve specimens (n=12) were prepared for each experi-mentalresincementaswellasthecontrolresinblend.The specimens were stored for 24h at room temperature, half (n=6) were tested, and the other half (n=6) were further incubatedfor30dayat37◦Cinsimulatedbodyfluidsolution (SBF)[21].Allsamplesweretestedusingauniversaltesting machine(Instron5569)witha500Nloadcellandcrosshead speed of0.5mm/minusing a3-point bendingsystem. The maximum fracture load (Newton) for each specimen was recorded and the flexural strength in MPa was calculated accordingtotheequation:

Flexurestrength = 3FL/2bd2_{, whereW= widthand}

H= heightofthespecimens.

Two-way ANOVA statistical analysis with a significance levelof0.05andaTukey’spost-hoctestwereperformedwith theresinblendsasindependentvariablesandtheFSasthe dependentvariable(SigmaStat®,Version3.5,SystatSoftware Inc.,PointRichmond,USA).

2.3. Fracturetoughnesstest(FT)

TheFT(KIC)wasmeasuredaccordingtotheASTMstandards

(E399-83),usingsingle-edge notched-beamspecimens[23]. Twelve specimens (16mm×2mm×2mm) for each experi-mentalgroupwerecreatedusingbar-shapedsiliconemolds andbycompressingthematerialusingaglassslideandfinally light-curedasdescribed previously.Afterremovalfromthe mold,thespecimensweregroundwithsiliconcarbidesand paper(gritsize#1200/4000)toremoveanyexcessmaterialon

thespecimens’edges.Allspecimenswerethenstoredin dis-tilledwaterat37◦Cpriortotestingfor24h.Anotch(0.3mm wide,1mmdeep)wascreatedforeachspecimenwitha dia-mond saw under waterirrigation. Thewidth ofthe notch wasdetermined bythethicknessoftheblade,whereas the standardizationofthenotch’sdepthwasassuredbya slid-ing depth-gaugemechanism installedonthe saw,to1mm intrusion.Thedepthofthenotchwasmeasuredwitha dig-italmicroscope(CY-800B,Tokyo,Japan).Theresinspecimens weresubjectedtothesametwoenvironments,24h(n=6)at roomtemperatureandincubatedfor30days(n=6)at37◦Cin SBF.SpecimensweretestedusinganInstron5569machinein a3-pointbendingtest,wherethesupportinglength[(S)=4W] was 16mm.Thesampleswereloaded untilfailure,usinga crosshead speed of0.5mm/min. During testing, the speci-menswereimmersedindistilledwateratroomtemperature. Theforceduringbendingofthespecimenwasmeasuredas afunctionofitsdeflection.Loadversusdeflectionplotswere recordedandthemaximumload(P)beforefailurewas mea-sured.Theheight(B)andwidth(W) ofthespecimenswere measured withamicrometerand thenotchdepth (a)with a measuringmicroscope.TheKIC wascalculated according

totheequationshownbelow,frommeasurementswiththe single-edgenotched-beamspecimens(MPam0.5_).

KIC=

3







W +2.7(a/W)2





Two-way ANOVA statistical analysis with a significance levelof0.05andaTukey’spost-hoctestwereperformedwith theresinblendsasindependentvariablesandtheFSasthe dependentvariable(SigmaStat®,USA).

2.4. BioactivityevaluationthroughRaman spectroscopyandSEM

The specimens used in the FS test were alsoanalyzed at 24h and after 30 day of SBF storage using Raman spec-troscopyand SEM.Inbrief,threespecimensforeachgroup werescannedinwetconditionsusingacomputer-controlled confocallaser Raman spectro-microscope (HoribaScientific Xplora,Villeneuved’Ascq,France)equippedwithopticalX20 objective andCCD detectorattachedtoamodularresearch spectrograph.Anear-infrareddiodelaserspot-sizeof≤1␥m operatingat785nmwasusedtoinducetheRaman scatter-ingeffect.Ramansignalswereacquiredusinga600-lines/mm grating centeredbetween500–1000cm−1. Thecalibrationof wavelengthandintensitywasperformedaccordingto man-ufacturer’s specification using a silicon standard and the calibrationsystemintegratedwiththesoftware[17,24].One surfaceofeachspecimenwasscannedinthreedifferentareas (ROI:1mm×1mm),andthensubmittedtoK-meanscluster (KMC)analysisasdescribedbySauroetal.[2]using multivari-ate analysis,whichincludedstatisticalpatterningtoderive independentclusters.Thebiochemicalcontentofeach clus-terwasanalyzedusingtheaverageclusterspectra.Principal componentanalysis(PCA)dataweresetintoabilinearmodel oflinearindependentvariables,theso-calledprincipal com-ponents(PCS).Twotofourclusterswereidentifiedbutonlytwo representativeclusterssuchasresinandmineralization(peak

at961cm−1 asreference)wereobtainedindependentlyand consideredinthisstudy.AllthespecimensusedforRaman spectroscopyanalysisweredehydratedinincreasing concen-trationsofethanol(50%–20min;75%–20min;90%–30min; 95%twice–30min;100%twice–30min)anddesiccatedfor 24hat3◦C.Theywere mountedon aluminumstubsusing carbontape,goldsputtercoatedandfinallyanalyzedthrough FEG-SEM(S4000Hitachi,Tokyo,Japan)at3kV.

2.5. Evaluationofthealkalinizingactivity(pH)

Three resin-diskspecimens were preparedfor each tested materialusingsiliconmolds(Ø=6mm;h=1mm)and light-curedfor40susingaLEDcuringsystem(Litex695,USA)as previouslydescribed (seeSection 2.4). Thespecimenswere immersedin25mlofdeionizedH2O(pH6.8)in

polypropylene-sealedcontainersstoredat37◦C.ThepH/alkalinizingactivity was evaluated using a professional pH electrode (Mettler-Toledo,Leicester,UK)at37◦Cafter1day,15day,30day.The H2Owasreplacedovereachmeasurementperiod.

2.6. Cellculture

Human mesenchymal stem cells basal medium (MSCs: ref. PT3238, Lonza), (Lonza Group Ltd., Basel, Switzerland)

[25] were cultured in ␣-Minimal Essential Medium (MEM) containing penicillin (50U/ml), streptomycin (50␮g/ml) (all from Sigma–Aldrich, Poole, Dorset, UK), Glutamax (2mM) (Invitrogen, Paisley, UK) and 10% fetal bovine serum (FBS) (Sigma–Aldrich);cellswere maintainedat37◦Cina humid-ified5%CO2:95%airatmosphere.Threeresin-diskspecimens

werepreparedforeachtestedmaterialusingsiliconmolds (Ø=6mm;h=1mm)andlight-curedfor40susingaLED cur-ingsystem(Litex695,USA).Subsequently,thespecimenswere polished using SiC papersup to#1000-grit under continu-ousdistilledwater(DW)irrigation.Thespecimenswerethen decontaminated bysoakingin absoluteethanolfor10min. Cellmonolayerswereseededinto24-wellplateswith5×104

cells/ml and cultured inthe presenceof the experimental materialsfor14daysimmersedinanosteogenicmedia sup-plemented with 0.1␮M dexamethasone, 0.05mM Ascorbic Acid(AA)and10mMglycerophosphate(Sigma–Aldrich). 2.7. Biologicaltesting(MTT)andosteogenic

differentiation(Runx-2andALP)

Assessmentofthebiocompatibilityoftheexperimental bioac-tive resin-based materials was performed in accordance with the ISO 10993-5 using a direct cytotoxicitytest, MTT assay.Inbrief,after14daysthemediawasreplacedwitha MTTsolution(Methylthiazolyldiphenyl-tetrazoliumbromide; Sigma–Aldrich)andincubatedfor4hat37◦C.Subsequently, the MTT solution was replaced with DMSO (Dimethyl sul-foxide, 99.9%, Sigma–Aldrich), and kept under continuous agitation(5min)usinganautomaticchemicalshaker(Titertek, FlowLaboratories).Finally, theabsorbanceofthe MTTwas measuredonamicroplatereader(OpsysMR,Dynex)at570nm wavelengthwithreference wavelengthat630nm.One-way ANOVAandTukey’stest(p<0.05)wereperformedtoanalyze the data (Sigma Stat®, USA). Live/Dead fluorescent

stain-ingwasperformedtoassessvisuallytheeffectoftheresin blends on the MSCs. Here, solutions of 2uM calcein AM (Sigma–Aldrich)and4␮MethidiumhomodimerinDulbeccos’s PhosphateBufferedSaline(PBS,Sigma–Aldrich)wereused,the systemthenincubatedfor10minandassessedvisually qual-itativelyusinganOlympusX51microscopecoupledwithan OlympusV-RFL-Tcamera,10×magnification.

Toassessosteogenicdifferentiation,mRNAexpressionof themarkersofdifferentiation(i.e.Runx-2andALP)was deter-mined byquantitativereal-timepolymerase chainreaction (RT-qPCR),whileaccumulationofcalciumdepositswas visu-alizedbystainingusingAlizarinRed.Briefly,cellswerefixed (15min with 4% formaldehyde in PBS), stained for 10min withalizarinredS(1:100dilutioninH2O)andwashedin50%

ethanolandair-dried.

TotalRNAwasextractedusingTRIreagent(Ambion, War-rington, UK) and Phase Lock Gel Heavy tubes (Five-prime VWR, Leicestershire, UK) according to the manufacturer’s instructions. After RNA purity and quantity was assessed bynanodrop(FisherScientific,London,UK)(A260/A280 1.8–2

was considered suitableforfurtheranalysis), possible con-taminatingDNAwasremovedandcDNApreparedfrom1␮g RNAusingQuantiTectReverseTranscriptionKit(Qiagen,West Sussex, UK) according to the manufacturer’s instructions. RT-qPCR was performed on a Mx3000P real time PCR sys-tem using Brilliant III Ultra-Fast SYBR Green qPCR Master mix (Stratagene, Agilent Technologies, Cheshire, UK) and thefollowingprimerpairs(5–3)Runx-2 (AATGGTTAATCTC-CGCAGGTC and TTCAGATAGAACTTGTACCCTCTGTT); ALP (AACACCACCCAGGGGAAC and TGGCATGGTTCACTCTCGT). PCRconditionsconsistedof1cycleof95◦Cfor3minand40 cyclesof95◦Cfor10sand60◦Cfor20s.RPL13awasusedas aninvarianthousekeepinggene.One-wayANOVAandTukey’s test(p<0.05)wereusedtoanalyzethedata(SigmaStat®,USA). Alizarinredstaining(ARS)stainingwasperformedto deter-minethemineralizationabilityofthetestedmaterialsonthe cellsineachwellplate(n=3).After14daysofincubation,the cellswererinsed,fixedwith10%formaldehydefor30min,and stained with40mMalizarinredS(pH4.2)for30min.After staining,themorphologywasobservedusinglightmicroscopy (OlympuslX71,Shinjuku,Tokyo,Japan).

3.

Results

3.1. Flexuralstrength(FS)andfracturetoughness(FT) Themeanandstandarddeviationoftheresultsobtainedwith theflexuralstrength(MPa)andfracturetoughness(MPam0.5₎

aredepictedinTable1.At24htherewasnosignificant differ-enceinFSbetweentheexperimentalresinsandthecontrol filler-freeresin(RES),(p>0.05).However,allthetested experi-mentalresinsshowedasignificantdropinFSafterASstorage for 30d (p<0.05); therewas no significant FSreductionin the RES specimens(p>0.05). MeanFS was affected by the materialtype(F=99.61;p<0.001),andbystoragetime(F=56.9; p<0.001).

At24htherewasnosignificantdifferenceinFTbetween the experimental resins and the RESgroup (p>0.05). Con-versely,onlytheRES,experimentalresin-basedmaterialsBAG

Table1–Flexuralstrength(FS)andfracturetoughness(FT)ofexperimentalresinsafter24hand30dayageinginASat 37◦C.

Flexuralstrength(FS;MPa) Fracturetoughness(FT;MPa.m0.5₎

24h 30day 24h 30day

RES 98.0±12.9[A1] 99.2±11.2[A1] 2.7±0.5[A1] 2.0±0.2[A2]

BAG 88.4±15.6[A1] 48.0±5.6[B2] 2.4±0.5[A1] 1.2±0.2[B2]

BAG-Zn 91.3±13.4[A1] 53.1±9.4[B2] 2.5±0.4[A1] 1.9±0.2[A1]

␤-CS 85.3±16.4[A1] 55.6±7.1[B2] 1.9±0.4[A1] 1.3±0.3[B2]

␤-CS-Zn 79.6±8.6[A1] 50.9±8.4[B2] 2.3±8.5[A1] 2.1±0.1[A1]

Similaruppercaseletterindicatesnosignificantdifferencesincolumnat24hor30day(p>0.05). Similarnumberindicatesnosignificantdifferencesinrowbetween24hand30day(p>0.05).

and␤-CSshowedasignificantdropinFTafterprolongedAS storage(p<0.05).Theothertwoexperimentalresins contain-ingthezinc-dopedfillers(␤-CS-Zn;BAG-Zn)hadnosignificant dropinFTafter30daysofASstorage(p>0.05).ThemeanFT wasaffectedbythematerialtype(F=79.61;p<0.001),andby storagetime(F=26.5;p<0.001).

3.2. Bioactivityandalkalinizingactivity(pH)

TheRamanclusterspectraofeachtestedmaterialobtained at24handafter30dayofSBFstorageareillustratedinFig.1

(A–H).Itwasobservedthatalltheexperimentalresins con-tainingthebioactivefillerspresented,afterclusteranalysis, a prominent peak at961cm−1 after 30 day ofAS storage, whichindicatedtheformationofapatite.Conversely,the rep-resentativeclusterofremineralizationwasneverobservedin specimensafter24h. These resultswere confirmed during the SEMassessmentofthe specimensafter30daysofSBF storage.Presenceofapatitewasdetectedinalltheresins con-tainingbioactivefillers.However,differencesinapatitecrystal morphologywereobservedbetweentheresinscontainingthe zinc-dopedbioactivefillers(BAG-Znand␤-CS-Zn)andthose containingBAG or ␤-TCS. Nodefinedcrystal structurewas observedinthe RESgroup. ThespecimenscontainingBAG and ␤-CS presented consistent precipitation of needle-like apatitecrystallites(Fig.1I).Theresins BAG-Znand␤-CS-Zn showedprecipitationofcrystallitesofapatitewitha globular-likestructure(Fig.1L).

Themean andstandard deviationofpHvaluesattained duringthealkalinizingactivity evaluationareillustrated in

Fig.2G.Theresinscontainingthebioactivemicro-fillersBAG or␤-CSinducedanincreaseofthepHofthestoragesolution at15dand30day.However,the␤-CSresinshowedthemost alkalinizingpotentialwithapHof12.6(±0.3)at15dayand12.7 (±0.2)after30dayofstorageinH2O.TheresincontainingBAG

fillerhadslightalkalizingpotential(pH∼10.0).Conversely,the resins containing the ZnO/polycarboxylated fillers (BAG-Zn and␤-CS-Zn)presentedamoreattenuatedalkalinizing activ-itybothat15day(pH∼8.6/8.8)andat30day(pH∼9/9.1).The resincontrol(RES)presentednochangeat15day(pH6.7±0.2), butthismaterialinducedslightacidificationofthemediaafter 30dofstorageinH2O(pH5.7±0.3).

3.3. MTTbiocompatibilityandLive&Deadassay Themeanandstandarddeviationoftheresultsobtainedwith MTTarepresentedinTable2,whilethequalitativeresultsLive &DeadaredepictedinFig.2.Inbrief,allexperimentalresins ledtoasignificantviabilitydecrease(p<0.01)comparedtothe cellcontrol.However,filler-freeresin(RES;Fig.2A)and␤-CS resin(Fig.2B)ledtocelldeath(p<0.05).Cellshada spindle-likemorphologyduringLive&Deadfluorescencemicroscopy, which is indicative of cell death. Similar cell density and aspectwasencounteredforgroups,BAG(Fig.2C)andthetwo zinc-dopedfillersBAG-ZNand␤-CS-Znresin(Fig.2D). 3.4. Geneexpression&mineralization

Themeanandstandarddeviationoftheresultsobtainedwith the gene expressionassayare presentedinTable2. Itwas observedthatMSCsgrowninpresenceoftheexperimental materialsdifferentiatedintoosteoblasts.Changesinmarkers ofosteogenesiswereanalyzedbyRT-qPCRandaccumulation ofmineralized matrix wasvisualized byalizarin red stain-ing.MSCsdifferentiatedinthepresenceofBAGshowedslight effectonexpressionofeitherosteogenicmarkers,Runx-2and ALP.Similarly,␤-CSandcontrolREShadnosignificanteffect (p>0.01)on expressionofosteogenicmaster switch (Runx-2) andALP expression.Interestingly,incorporationofZnto eitherBAGor␤-CShadpotentosteogeniceffectwith signifi-cantincreaseinexpressionofbothmarkers(p>0.01).Runx-2 expressionwasincreasedbymorethanfivefold(p<0.05)in bothBAG-Znand␤-CS-ZnandALPlevelbymorethan four-fold (p<0.05)when comparedtocells culturedinpresence ofpureresinalone.TheeffectseenonmRNAexpressionof osteogenicgenesbyBAG-Zn(Fig.2E)and␤-CS-Zn(Fig.2F)was alsoconfirmedbythealizarinstainingofcalciumdeposits.

4.

Discussion

Pulpprotectionmaterialsusedintheatraumaticrestorative treatment (ART) technique should be able to evoke rem-ineralization with no inflammatory pulp response [9–13]. Accordingly,thisstudyaimedatgeneratingadvanced light-curable resin-based materials containingdifferent typesof experimentalbioactivemicro-fillers:evaluatingtheirflexural strength,fracture toughness,biocompatibility,cell differen-tiationand bioactivityafterstorageinsimulatedbodyfluid (SBF).However,itisimportanttoanticipatethatthe

exper-Fig.1–Ramanspectroscopicclusteranalysisofprincipalcomponentsidentifiedbeforeandafterstorage[resinand remineralization].TheresinBAGat24hshowedaRamanresinclusterwithnosignofCa/Pmineralsontheoutersurface (A),whileafter30daysofSBFstorage,itwasabletoinduceapatiteprecipitation(notethepeakat961cm−1)(B).Thesame situationwasobservedwiththeresinBAG-Znat24h,whichshowednosignofmineralization(C),butclearapatite

Table2–Biocompatibility(MTT)andcelldifferentiation(geneexpressionRunx2andALP)ofexperimentalresinsafter 24hand30dayageinginASat37◦C.

MTT mRNAgeneexpression

14days Runx2 ALP

Cell-control 0.5±0.01[A] – – RES 0.2±0.03[B] 1.1±0.1[A] 1.0±0.1[1] BAG 0.4±0.04[C] 1.74±0.5[A] 1.3±0.3[1] BAG-Zn 0.4±0.04[C] 4.9±1.7[B] 4.1±1.2[2] ␤-CS 0.1±0.05[D] 1.7±0.2[A] 1.5±0.1[1] ␤-CS-Zn 0.4±0.05[C] 4.8±1.9[B] 4.7±1.9[2]

SimilaruppercaseletterindicatesnosignificantdifferencesincolumnforMTTandRunx2(p>0.01). SimilarnumberindicatesnosignificantdifferencesincolumnforALP(p>0.001).

imental light-curable resin-based materials with bioactive micro-fillersassessedinthisstudyhavenotbeendeveloped tobeusedasdirectpulp-cappingagents.

The flexural strength (FS) is representative of tensions developed in the dental cavities, as it encompasses ten-sile, compressive and shear stresses under bending loads. Although FS is a static single-movement mechanical test, it is usually the first choice to survey the preliminary physicalpropertiesofnewlydevelopedresin-based materi-als. It is well known that several internal and superficial defects/imperfectionsmaybepresentinspecimensand mate-rials[3],evenwhentheyseemflatandsmoothtothenaked eye.Therefore,undermasticatoryforces,materialsundergo notonlyflexuraltensions,but alsothe needtoresist frac-ture/crack propagation; this measure is known as fracture toughness(FT).Inviewoftheresultsobtainedwiththepresent FSand FTtests, thefirst nullhypothesis that theaddition ofbioactivemicro-fillerswouldhavenoeffectonmechanical propertiesoftheexperimentalresin-basedmaterialstested must be partially rejected. Indeed,this study showed that there was no significant difference in FS and FT at 24h betweentheresin-basedmaterialscontainingthe experimen-talmicro-fillersandthecontrolfiller-freeresin.However,all thetestedexperimentalmaterialsshowedasignificantdrop inFSafter30dstorageofSBFstorage,whilethespecimens createdwiththefiller-freeresin(RES)showednosignificant reduction.

ApossibleexplanationfortheFSresultsobtainedwiththe specimenscreatedusingtheresin-basedmaterialscontaining theexperimentalmicro-fillersisthatsuchareductionmaybe attributedtothehydrophilicityoftheincludedfillers,which absorb quantitiesofwater [20]to evokethe bioactive pro-cessesformineralprecipitation(Fig.1A–H).However,although these typesof bioactive resin-based materials may not be suitableasdefinitiverestorativematerialsduetotheirhigher watersorptionandsolubilitycomparedtoconventional den-talresincomposites,theycanbeusedasapulpprotectionor

temporary therapeuticmaterialsinthestepwiserestorative technique[20].

TheFTresultsshowedasignificantdropafter30day stor-ageofSBFstorageonlyinthespecimenscreated usingthe experimentalBAG and ␤-CS and the filler-freeresins. Con-versely,theexperimentalresins␤-CS-Zn;BAG-Znshowedno significantFTreduction(Table1).

Suchadropinmechanicalpropertiesobservedafter pro-longedaginginSBF(30day)canbeduetothehighopacity (low reflectionindex) ofsomeofthe micro-fillers (40vol%) usedinthisstudy.Thesemayhaveinterferedwiththe photo-polymerization due toanincomplete diffusionofthe light (energy)throughoutthebulkofmaterial.Suchaphenomenon waspotentiallyresponsibleforthedecreaseofmonomer con-version,whichcausedanaccelerateddegradationoftheresin matrix over time, especially in the experimental BAG and ␤-CS resins [26]. Conversely,inthe filler-free controlresin, thereductionoffracturetoughnessmaybeattributedtothe hydrolyticdegradationuponSBFimmersion,withhydrolysis andelutionofthemorelinearpolymerchainswithintheresin matrix[27].

Theexperimentalmicro-fillers,␤-CS-ZnandBAG-Zn,may havetriggedaprolongedcationicpolymerizationwithinthe polymermatrix.Itisknown thatacationicpolymerization maybeactivatedbypolyionicsilicateswhichformduringthe formulationoftheBAG-Znand␤-CS-Znmicro-fillerstreated withthepolycarboxylic acidicsolution[20,28,29].Moreover, ithasbeenstatedthatglasspolyalkenoatecements contain-ing zinc oxidemay react withPAA and set via a two-step setting mechanism: 1) zinc-polyacrylate species cause pri-mary hardening of the cement and later; 2) during water storage,formationofcalcium-polyacrylatemayinducea sec-ondpost-hardeningstep.Furthermore,calciumionsreleased from the experimentalmicro-fillersmayhavereacted with the carboxylategroupinthe polyacrylicacidand stabilized the mechanical propertiesresin BAG-Zn and ␤TCS-Zn dur-ingprolongedASstorage[30].However,itisalsolikelythat

precipitation(notethepeakat961cm−1)after30daysofSBFstorage(D).Theresin␤-CS(E)andthe␤-CS-Zn(G)showedat 24hshowedonlyaRamanclusterwithnosignofmineralization.Thesetworesins␤-CS(F)andthe␤-CS-Zn(H)wereable toinduceapatiteformation(notethepeakat961cm−1)after30daysofSBFstorage.FiguresIandLarerepresentativeSEM micrographsshowingrespectivelyaneedle-likeapatiteformation(pointer)inresinscontainingfillerthatwerenotdoped withzinc(BAGand␤-CS)andglobular-likenano-apatiteformation(arrow)inresinscontainingthezinc-dopedfillers (BAG-Znand␤-CS-Zn).

Fig.2–ImagesobtainedduringtheLive&Deadfluorescencemicroscopy.Itispossibletoseespindle-likecellmorphology (pointer),whichisindicativeofcelldeath,inthegroupofthefillerfreeresin(A)and␤-CSresin(B).ThegroupBAG,BAG-ZN and␤-CS-Znpresentedsimilarcelldensity.FigureCisarepresentativeimageobtainedinpresenceoftheBAGandfigureD isarepresentativeimageobtainedwiththeresincontainingthezinc-dopedfillers(BAG-ZNand␤-CS-Zn).Inallcases,itis possibletoseegreatcelldensityandacleardifferentiationoftheMSCsintofibroblasts(pointer).Fig.(E)and(F)areoptical micrographs(alizarinredstainingtechnique)thatshowthecalciumdeposits(darkareas)observedintheresinsBAG-Zn and␤-CS-Zn,respectively.Infigure(G)itisreportedthepHvalueofeachmaterialtestedinthisstudy.Itispossibletosee thatthemostalkalinizingresinwasthe␤-CS,whichwasabletomaintainastrongalkalinepHovertime[pH>12].While theotherexperimentalresinspresentedamilderpH,inparticularthosecontainingthezinc-dopedfillers.

bioactiveionleachingandthebioactivetransformationofthe fillersintoneedle-likeapatite(Fig.2I)atastronglyalkalinepH (Fig.2G)mayhavecreatedexcessivemicro-porositieswithin

theexperimentalBAGand␤-CSresins.Thismayhavecaused asignificantreductioninFTovertime[20].Theexperimental BAG-Znand␤-CS-Znresinsshowedglobular-likeapatite

pre-cipitation(Fig.2L)atmildlyalkalinepH(Fig.2G).Inthiscase,it ishypothesizedthatthereductioninFTwaslessevidentdue tothefewerporositieswithinthematerialcreatedbyamore compactmineralprecipitation.Alltheresultsdescribedsofar seemtobeinaccordancewiththosepreviouslyreportedby Sauroetal.[20]andbyLeetal.[31].Thelatterteam demon-stratedthat, inastronglyalkalineenvironment,(pH10–11) apatiteprecipitateswithamorphologycharacterizedby clus-tersofacicularcrystals(needle-likestructures).Atamildly alkalinepH (8–9), apatite precipitates as globulesof nano-particles(globular-likestructure)duetoafastnucleationrate sothatmoreparticlesareformedandthenucleiofcrystals growslowerduetothelowtemperature[32].

Itisimportanttoconsiderthatthe differencesobserved in these experimental resins in terms of pH, kinetics of precipitationandmorphologyoftheapatitemayhave influ-enced the biocompatibility and played a role in the cell growth/differentiation. Indeed, the present results showed thatalltheresin-basedmaterialstestedledtoasignificant cell viabilitydecrease (MTTtest) comparedtothe negative controlwherethecellswereseededintheabsenceof resin-basedmaterial.However,thefiller-freeresin(RES)and␤-CS resininducedcytotoxicityandthecellsshowedaspindle-like morphology,indicativeofcelldeath(Fig.2AandB).Theeffects ofthetestedmaterialsonthemineralizationpotentialwere investigatedatthemRNAlevel,enzymaticlevel(ALPonly)and extracellularmatrix mineralizationlevel(Alizarinred stain-ing).TheRESand␤-CSshowed nosignificanteffectonthe expressionofthe osteogenicmarkers, Runx-2and ALP. All theseoutcomesencounteredwiththe␤-CSresinmay have beenaconsequenceoftheabilityofsuchmaterialtomake themediumalkalineovertime(Fig.2G)[33].Moreover,itis speculatedthattheremay havebeen animportantelution ofunreacted monomersinall resins[27].Thisledtoa sig-nificantcellviabilityreductioncomparedtothecontrol(no resinmaterial).Previously,thepossibilityofalowdegreeof conversioninthe␤-CSduetoalackofphoto-polymerization wasdiscussed.Inthiscase,itmayhavecausedtheelutionof unreactedmonomersandinitiator/co-initiatoragents caus-inggreatercytotoxicitywiththe␤-CSresin[34,35].Leaching ofsuch components may accumulate over time to a toxic level,whichdepleteirreversiblythedefensemechanismsof cellsand resultincellapoptosis [36]. The␤-CSresin hasa compositionverysimilartothatofacommercialpulp pro-tectionmaterialknownasTheracalLC(containing40–50wt% ofPortland cementpowder,30–40wt%ofneutralor mildly acidichydrophilicresinmonomers,5–10wt%ofa hydropho-bicresinmonomer(BisGMA) and5–10wt%ofahydrophilic filler).Indeed,thepresentresultsseemtobeinaccordance withthoseofBortoluzzietal.[33],whoshowedthatTheracal LCcausedareductionincellmetabolismandprotein expres-sionwhenincontactwithpulpcells.However,apartfrom␤-CS resin,thedropinabsorbance(MTT)waslessthan50% com-paredtothecontrolcellsgroups,whichmayrepresentminor (reversible)cytotoxiceffects[37].

Similarcelldensityandmorphologywereencounteredfor resincontainingBAG(Fig.2C)andthosecontainingthe zinc-dopedpolycarboxylatefillers(BAG-ZNand␤-CS-Zn)(Fig.2D). These materials showed less cytotoxicity comparedto the controlRESand␤-CS.Moreover,theresinsBAG-Znand

␤-CS-Znwereabletoinduceanosteogeniceffectwithsignificant increase in expressionof both markers (Runx-2 and ALP). Calcium depositswere observedusingthe alizarinstaining techniqueintheBAG-Zn(Fig.2E)and␤-CS-Zn(Fig.2F)resins. Inviewoftheseresultsitispossibletoconfirmthatthe secondnullhypothesis,theadditionofbioactivemicro-fillers wouldcauseincreaseincytotoxicityandinduceno differen-tiationin primaryhumanmesenchymal stemcells (MSCs), should bepartially rejectedasthese factors areassociated withthechemicalpropertiesandcompositionofeachsingle testedbioactivefiller.

Recently,Ca–Si-basedceramics dopedwithspecific ther-apeutic elements (Mg [38], Zn[39], and Sr [40]) have been designed toimprove their biological properties. In particu-lar, Znhasbeen demonstratedtohaveastimulatoryeffect on boneformation andan inhibitoryeffect onosteoclastic boneresorption[41].Indeed,ZnionsappeartoinduceALP activity,sinceALPisaZn-dependentenzyme[42].Zinc defi-ciencyinducesarrestofbonegrowth,bonedevelopment,and jeopardizes theoverall maintenanceofbonehealth[43,44]. Yuetal.[45]showedthatbioceramicsdopedwithzinc(e.g. Ca2ZnSi2O7)enhancesattachment,proliferation,

differentia-tion andup-regulated bonemarkergeneexpressionofALP and growth factor genes (e.g. IGF-I andTGF-␤1),compared tozinc-freesilicates(CaSiO3).Chesters[46]andCousins[46]

reported that zinc is a crucial trace element that induces several metabolic, cellular signaling pathways and gene expressionforboneformation.Moreover,theincorporation of zinc into calcium phosphate cement significantly pro-motespre-osteoblastproliferationanddifferentiationinvitro

[47–50]. Furthermore, it was also suggested that zinc is

involvedinregulatingthetranscriptionofpre-osteoblast dif-ferentiationgenes(e.g.Col-I,ALPandosteopontin)[51].Hence, zincisconsideredapromisingagentforenhancingthe bone-formingabilityofimplantmaterials,whichcanbeachieved bycontrollingthereleaseofZnions.

5.

Conclusions

Thepresentstudyhasdemonstratedthattheincorporation oftheexperimentalfillersintoresin-basedmaterialsmakes them“bioactive”enoughtoinduceapatiteprecipitation. How-ever,onlytheresinscontainingthetwozinc-dopedbioactive fillersareabletomaintainstablemechanicalpropertiesover time.Intermsofbiocompatibilityandcellsdifferentiation,the resinscontainingthezinc-dopedfillers(␤-CS-ZnandBAG-Zn) werethemostpromisingaspossiblealternativepulp protec-tionmaterialsforuseclinically.

Acknowledgements

ThisworkwassupportedbytheresearchgrantINDI– “Pro-gramadeConsolidacióndeIndicadores:FomentoPlanEstatal CEU-UCH”2014–2018toProf.Dr.SalvatoreSauro.Thisresearch was also in part supported byProject CNPq 407282/2013-0 (BolsaPVE–Linha2),UBCStart-UPfundstoRMC.

r

e

f

e

r

e

n

c

e

s

[1] InnesNPT,FrenckenJE,BjorndalL,BuchallaW,MaltzM,

MantonDJ,etal.Managingcarieslesions:consensus

recommendationsonterminology.AdvDentRes

2016;28:49–57.

[2] SauroS,OsorioR,WatsonTF,ToledanoM.Influenceof

phosphoproteins’biomimeticanalogsonremineralization

ofmineral-depletedresin–dentininterfacescreatedwith

ion-releasingresin-basedsystems.DentMater

2015;31:759–77.

[3] AbdullahD,FordTR,PapaioannouS,NicholsonJ,McDonald

F.AnevaluationofacceleratedPortlandcementasa

restorativematerial.Biomaterials2002;23:4001–10.

[4] SchwendickeF,FrenckenJE,BjorndalL,BuchallaW,MaltzM,

MantonDJ,etal.Managingcarieslesions:consensus

recommendationsoncarioustissueremoval.AdvDentRes

2016;28:58–67.

[5] BanerjeeA,FrenckenJE,SchwendickeF,InnesNPT.

Contemporaryoperativecariesmanagement:consensus

recommendationsonminimallyinvasivecariesremoval.Br

DentJ2017;223:215–22.

[6] HashemD,MannocciF,PatelS,AndiappanM,BrownJE,

WatsonTF,etal.Efficacyofcalciumsilicateindirectpulp

capping;arandomizedcontrolledclinicaltrial.JDentRes

2015;94:562–8.

[7] KimYK,MaiS,MazzoniA,LiuY,Tezvergil-MutluayA,

TakahashiK,etal.Biomimeticremineralizationasa

progressivedehydrationmechanismofcollagen

matrices—implicationsintheagingofresin–dentinbonds.

ActaBiomater2010;6:3729–39.

[8] HenchLL.Bioceramicsandtheoriginoflife.JBiomedMater

Res1989;23:685–93.

[9] KokuboT,TakadamaH.HowusefulisSBFinpredicting

invivobonebioactivity.Biomaterials2006;27:2907–15.

[10] CamilleriJ,MontesinFE,DiSilvioL,PittFordTR.The

chemicalconstitutionandbiocompatibilityofaccelerated

Portlandcementforendodonticuse.IntEndodJ

2005;38:834–42.

[11] CamilleriJ.Investigationofbiodentineasdentine

replacementmaterial.JDent2013;41:600–10.

[12] BogenG,KuttlerS.Mineraltrioxideaggregateobturation:a

reviewandcaseseries.JEndod2009;35:777–90.

[13] PorterML,BertoA,PrimusCM,WatanabeI.Physicaland

chemicalpropertiesofnew-generationendodontic

materials.JEndod2010;36:524–8.

[14] NielsenMJ,CaseyJA,VanderWeeleRA,VandewalleKS.

Mechanicalpropertiesofnewdentalpulp-capping

materials.GenDent2016;64:44–8.

[15] HeblingJ,LessaFC,NogueiraI,CarvalhoRM,CostaCA.

Cytotoxicityofresin-basedlight-curedliners.AmJDent

2009;22:137–42.

[16] HenchLL.Thestoryofbioglass.JMaterSciMaterMed

2006;17:967–78.

[17] SauroS,ThompsonI,WatsonTF.Effectsofcommondental

materialsusedinpreventiveoroperativedentistryon

dentinpermeabilityandremineralization.OperDent

2011;36:222–30.

[18] GendreauL,BarlowAP,MasonSC.Overviewoftheclinical

evidenceforuseofNovaMininprovidingrelieffromthe

painofdentinhypersensitivity.JClinDent2011;22:90–5.

[19] MillyH,FestyF,AndiappanM,WatsonTF,ThompsonI,

BanerjeeA.Surfacepre-conditioningwithbio-activeglass

air-abrasioncanenhanceenamelwhitespotlesion

remineralisation.DentMater2015;31:522–33.

[20] SauroS,OsorioR,WatsonTF,ToledanoM.Therapeutic

effectsofnovelresinbondingsystemscontainingbioactive

glassesonmineral-depletedareaswithinthe

bonded-dentineinterface.JMaterSciMaterMed

2012;23:1521–32.

[21] SauroS,OsorioR,FulgencioR,WatsonTF,CamaG,

ThompsonI,etal.Remineralisationpropertiesofinnovative

light-curableresin-baseddentalmaterialscontaining

bioactivemicro-fillers.JMaterChemB2013;1:2624–38.

[22] El-AskaryFS,BotrosSA,AbdelazizNassifMF,ÖzcanM.

Flexuralstrengthofnano-hybridresincompositeasa

functionoflightattenuationdistanceandspecimen

dimension.JAdhesSciTechnol2017;31:520–9.

[23] IIlieN,HiltonTJ,HeintzeSD,HickelR,WattsDC,SilikasN,

etal.Academyofdentalmaterialsguidance-resin

composites:partI-mechanicalproperties.DentMater

2017;33:880–94.

[24] TsudaH,ArendsJ.Ramanspectroscopyindentalresearch:a

shortreviewofrecentstudies.AdvDentRes1997;11:539–47.

[25] Blázquez-PruneraA,DíezJM,GajardoR,GranchaS.Human

mesenchymalstemcellsmaintaintheirphenotype,

multipotentiality,andgeneticstabilitywhenculturedusing

adefinedxeno-freehumanplasmafraction.StemCellRes

Ther2017;8:103.

[26] SalgadoVE,CavalcanteLM,MoraesRR,DavisHB,Ferracane

JL,SchneiderLF.Degradationofopticalandsurface

propertiesofresin-basedcompositeswithdistinct

nanoparticlesizesbutequivalentsurfacearea.JDent

2017;59:48–53.

[27] FerracaneJL.Hygroscopicandhydrolyticeffectsindental

polymernetworks.DentMater2006;22:211–22.

[28] ShahwanT,ZünbülB,Tunuso ˘gluO,Ero ˘gluAE.AAS,XRPD,

SEM/EDS,andFTIRcharacterizationofZn2+_retentionby

calcite,calcite–kaolinite,andcalcite–clinoptiloliteminerals.

JColloidInterfaceSci2005;286:471–8.

[29] O’DonnellJNR,SchumacherGE,AntonucciJM,SkrticD.

Structure–composition–propertyrelationshipsinpolymeric

amorphouscalciumphosphate-baseddentalcomposites.

Materials2009;2:1929–54.

[30] KamitakaharaM,KawashitaM,KokuboT,NakamuraT.

Effectofpolyacrylicacidontheapatiteformationofa

bioactiveceramicinasimulatedbodyfluid:fundamental

examinationofthepossibilityofobtainingbioactive

glass-ionomercementsfororthopaedicuse.Biomaterials

2001;22:3191–6.

[31] LeHR,ChenKY,WangCA.EffectofpHandtemperatureon

themorphologyandphasesofco-precipitated

hydroxyapatite.JSolGelSciTechnol2012;61:592–9.

[32] LiuC,HuangY,ShenW,CuiJ.Kineticsofhydroxyapatite

precipitationatpH10–11.Biomaterials2001;22:301–6.

[33] BortoluzziEA,NiuLN,PalaniCD,El-AwadyAR,Hammond

BD,PeiDD,etal.Cytotoxicityandosteogenicpotentialof

silicatecalciumcementsaspotentialprotectivematerials

forpulpalrevascularization.DentMater2015;31:1510–22.

[34] VanLanduytKL,KrifkaS,HillerKA,BolayC,WahaC,Van

MeerbeekB,etal.Evaluationofcellresponsestoward

adhesiveswithdifferentphotoinitiatingsystems.Dent

Mater2015;31:916–27.

[35] SiguschBW,PflaumT,VölpelA,SchinkelM,JandtKD.The

influenceofvariouslightcuringunitsonthecytotoxicityof

dentaladhesives.DentMater2009;25:1446–52.

[36] EngelmannJ,JankeV,VolkJ,LeuhausenG,NeuhoffNV,

SchlegelbergerB,etal.EffectsofBisGMAonglutathione

metabolismandapoptosisinhumangingivalfibroblastsin

vitro.Biomaterials2004;25:4573–80.

[37] HeblingJ,BianchiL,BassoFG,ScheffelDL,SoaresDG,

CarrilhoMR,etal.Cytotoxicityofdimethylsulfoxide(DMSO)

indirectcontactwithodontoblast-likecells.DentMater

[38] WuCT,ChangJA,WangJY,NiSY,ZhaiWY.Preparationand

characteristicsofacalciummagnesiumsilicate(bredigite)

bioactiveceramic.Biomaterials2005;26:2925–31.

[39] RamaswamyY,WuCT,ZhouH,ZreiqatH.Biological

responseofhumanbonecellstozinc-modifiedCa–Si-based

ceramics.ActaBiomater2008;4:1487–97.

[40] WuCT,RamaswamyY,KwikD,ZreiqatH.Theeffectof

strontiumincorporationintoCaSiO3ceramicsontheir

physicalandbiologicalproperties.Biomaterials

2007;28:3171–81.

[41] YamaguchiM,OishiH,SuketaY.Stimulatoryeffectofzinc

onboneformationintissueculture.BiochemPharmacol

1987;36:4007–12.

[42] HabibovicP,BarraletJE.Bioinorganicsandbiomaterials.

Bonerepair.ActaBiomater2011;7:3013–26.

[43] HsiehHS,NaviaJM.Zincdeficiencyandboneformationin

guineapigalveolarimplants.JNutr1980;110:1581–8.

[44] OnerG,BhaumickB,BalaRM.Effectofzincdeficiencyon

serumsomatomedinlevelsandskeletalgrowthinyoung

rats.Endocrinology1984;114:1860–3.

[45] YuJ,LiK,ZhengX,HeD,YeX,WangM.Invitroandinvivo

evaluationofzinc-modifiedCa–Si-basedceramiccoatingfor

boneimplants.PLoSOne2013;8:e57564.

[46] ChestersJK.Molecularbiologicalapproachesto

nutrient–geneinteractions.ProcNutrSoc1998;57:249–56.

[47] CousinsRJ.Aroleofzincintheregulationofgene

expression.ProcNutrSoc1998;57:307–11.

[48] ItoA,KawamuraH,OtsukaM,IkeuchiM,OhgushiH,

IshikawaK,etal.Zinc-releasingcalciumphosphatefor

stimulatingboneformation.MaterSciEngCBiomim

SupramolSyst2002;22:21–5.

[49] KawamuraH,ItoA,MiyakawaS,LayrolleP,OjimaK,

IchinoseN,etal.Stimulatoryeffectofzinc-releasing

calciumphosphateimplantonboneformationinrabbit

femora.JBiomedMaterRes2000;50:184–90.

[50] ItoA,OtsukaM,KawamuraH,IkeuchiM,OhgushiH,SogoY,

etal.Zinc-containingtricalciumphosphateandrelated

materialsforpromotingboneformation.CurrApplPhys

2005;5:402–6.

[51] KwunIS,ChoYE,LomedaRA,ShinHI,ChoiJY,KangYH,

etal.Zincdeficiencysuppressesmatrixmineralizationand

retardsosteogenesistransientlywithcatch-uppossibly