OpenAccessMesenchymalStemCell‐DerivedExosomesPromoteFractureHealinginaMouseModelTaisukeFuruta

CorrespondingAuthor

E-mailaddress:fu09100913@yahoo.co.jp

81-82-257-5233

DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,Hiroshima,Japan

Contributedequally.

Correspondence:TaisukeFuruta,M.D.,DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,1-2-3Kasumi,Minami-ku,Hiroshima,734-5234,Japan.SearchformorepapersbythisauthorShigeruMiyaki

DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,Hiroshima,Japan

DepartmentofRegenerativeMedicine,MedicalCenterforTranslationalandClinicalResearch,HiroshimaUniversityHospital,Hiroshima,Japan

Contributedequally.

SearchformorepapersbythisauthorHiroyukiIshitobi

DepartmentofRegenerativeMedicine,MedicalCenterforTranslationalandClinicalResearch,HiroshimaUniversityHospital,Hiroshima,Japan

SearchformorepapersbythisauthorToshihikoOgura

BiomedicalResearchInstitute,NationalInstituteofAdvancedIndustrialScienceandTechnology,Tsukuba,Japan

SearchformorepapersbythisauthorYoshioKato

BiomedicalResearchInstitute,NationalInstituteofAdvancedIndustrialScienceandTechnology,Tsukuba,Japan

SearchformorepapersbythisauthorNaosukeKamei

DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,Hiroshima,Japan

DepartmentofRegenerativeMedicine,MedicalCenterforTranslationalandClinicalResearch,HiroshimaUniversityHospital,Hiroshima,Japan

SearchformorepapersbythisauthorKenjiMiyado

DepartmentofReproductiveBiology,NationalCenterforChildHealthandDevelopment,Tokyo,Japan

SearchformorepapersbythisauthorYukihitoHigashi

DepartmentofRegenerativeMedicine,MedicalCenterforTranslationalandClinicalResearch,HiroshimaUniversityHospital,Hiroshima,Japan

SearchformorepapersbythisauthorMitsuoOchi

DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,Hiroshima,Japan

Searchformorepapersbythisauthor
TaisukeFuruta

CorrespondingAuthor

E-mailaddress:fu09100913@yahoo.co.jp

81-82-257-5233

DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,Hiroshima,Japan

Contributedequally.

Correspondence:TaisukeFuruta,M.D.,DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,1-2-3Kasumi,Minami-ku,Hiroshima,734-5234,Japan.SearchformorepapersbythisauthorShigeruMiyaki

DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,Hiroshima,Japan

DepartmentofRegenerativeMedicine,MedicalCenterforTranslationalandClinicalResearch,HiroshimaUniversityHospital,Hiroshima,Japan

Contributedequally.

SearchformorepapersbythisauthorHiroyukiIshitobi

DepartmentofRegenerativeMedicine,MedicalCenterforTranslationalandClinicalResearch,HiroshimaUniversityHospital,Hiroshima,Japan

SearchformorepapersbythisauthorToshihikoOgura

BiomedicalResearchInstitute,NationalInstituteofAdvancedIndustrialScienceandTechnology,Tsukuba,Japan

SearchformorepapersbythisauthorYoshioKato

BiomedicalResearchInstitute,NationalInstituteofAdvancedIndustrialScienceandTechnology,Tsukuba,Japan

SearchformorepapersbythisauthorNaosukeKamei

DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,Hiroshima,Japan

DepartmentofRegenerativeMedicine,MedicalCenterforTranslationalandClinicalResearch,HiroshimaUniversityHospital,Hiroshima,Japan

SearchformorepapersbythisauthorKenjiMiyado

DepartmentofReproductiveBiology,NationalCenterforChildHealthandDevelopment,Tokyo,Japan

SearchformorepapersbythisauthorYukihitoHigashi

DepartmentofRegenerativeMedicine,MedicalCenterforTranslationalandClinicalResearch,HiroshimaUniversityHospital,Hiroshima,Japan

SearchformorepapersbythisauthorMitsuoOchi

DepartmentofOrthopaedicsSurgery,IntegratedHealthSciences,InstituteofBiomedicalandHealthScience,HiroshimaUniversity,Hiroshima,Japan

Searchformorepapersbythisauthor

Giveaccess

Sharefulltextaccess

Sharefull-textaccess

PleasereviewourTermsandConditionsofUseandcheckboxbelowtosharefull-textversionofarticle.IhavereadandaccepttheWileyOnlineLibraryTermsandConditionsofUseShareableLink

Usethelinkbelowtoshareafull-textversionofthisarticlewithyourfriendsandcolleagues.Learnmore.

CopyURL

Sharealink

ShareonEmailFacebookTwitterLinkedInRedditWechatAbstractAcknowledgements

Paracrinesignalingbybone‐marrow‐derivedmesenchymalstemcells(MSCs)playsamajorroleintissuerepair.AlthoughtheproductionofregulatorycytokinesbyMSCtransplantationisacriticalmodulatoroftissueregeneration,wefocusedonexosomes,whichareextracellularvesiclesthatcontainproteinsandnucleicacids,asanoveladditionalmodulatorofcell‐to‐cellcommunicationandtissueregeneration.Toaddressthis,weusedradiologicimaging,histologicalexamination,andimmunohistochemicalanalysistoevaluatetheroleofexosomesisolatedfromMSC‐conditionedmedium(CM)inthehealingprocessinafemurfracturemodelofCD9−/−mice,astrainthatisknowntoproducereducedlevelsofexosomes.WefoundthattheboneunionrateinCD9−/−micewassignificantlylowerthanwild‐typemicebecauseoftheretardationofcallusformation.TheretardationoffracturehealinginCD9−/−micewasrescuedbytheinjectionofexosomes,butthiswasnotthecaseaftertheinjectionofexosomes‐freeconditionedmedium(CM‐Exo).Thelevelsofthebonerepair‐relatedcytokines,monocytechemotacticprotein‐1(MCP‐1),MCP‐3,andstromalcell‐derivedfactor‐1inexosomeswerelowcomparedwithlevelsinCMandCM‐Exo,suggestingthatbonerepairmaybeinpartmediatedbyotherexosomecomponents,suchasmicroRNAs.TheseresultssuggestthatexosomesinCMfacilitatetheaccelerationoffracturehealing,andweconcludethatexosomesareanovelfactorofMSCparacrinesignalingwithanimportantroleinthetissuerepairprocess.

Thisworkfocusesonexosomes,whichareextracellularvesicles,asanoveladditionalmodulatorofcell‐to‐cellcommunication.Thisstudyevaluatedtheroleofexosomesisolatedfrommesenchymalstemcell(MSC)‐conditionedmedium(MSC‐CM)inthefracture‐healingprocessofCD9−/−mice,astrainthatisknowntoproducereducedlevelsofexosomes.RetardationoffracturehealinginCD9−/−micewasrescuedbytheinjectionofMSCexosomes,butthiswasnotthecaseaftertheinjectionofexosome‐freeCM.ThisstudyfindsthatMSCexosomesareanovelfactorofMSCparacrinesignaling,withanimportantroleinthetissuerepairprocess.

Introduction

Poororimpairedrecoveryfromfracturesisanincreasinglyimportantprobleminagingsocieties.Althoughmostfracturesarerepairedthroughintramembranousossificationandendochondralossification,inapproximately10%ofcases,unionisdelayedorcompromisedbecauseofimpairedbloodsupplyandrecruitmentofhomingstemcells,excessivedamagetotheperiosteum,inadequateimmobilization,orinfectionattheinjurysite[1].Althoughautologousbonetransplantationandvascularizedbonegraftshavebeenperformedfordelayed‐unionornonunionfractures[2],currenttherapiesareoftenineffective.Animprovedunderstandingofthemolecularmechanismsunderlyingbonerepairisneededtoinformthedevelopmentofnewtherapiesforpatientswithdelayed‐unionornonunioncomplicationsfromfractures.

Transplantationofstemcells,suchasbone‐marrow‐derivedmesenchymalstemcells(MSCs),promotestissueregeneration,includingfracturehealing[3–5].ManystudiesindicatethattransplantedMSCscontributetotissueregenerationbymechanismsotherthandifferentiationanddirecttissueintegration[6–9].Therefore,paracrinesignalingbyMSCsisattractingattentionasapotentialmechanismtoexplaintheeffectofthesecellsontissueregeneration.Theparacrineeffectismediatedbyproteinssuchascytokinesandchemokineswithantiapoptotic,anti‐inflammatory,antioxidative,andproangiogenicproperties[10].MSC‐conditionedmediumacceleratescallusformation,andmonocytechemotacticprotein‐1(MCP‐1)andstromalcell‐derivedfactor‐1(SDF‐1)areessentialintheinitialphaseoffracturehealingbyregulatingtherecruitmentanddifferentiationofMSCsandotherprogenitorcells[11–14].

Recently,extracellularvesicles,includingexosomes,haveattractedincreasingattentionasmediatorsofcell‐to‐cellcommunication[15,16].Exosomesaresmallparticlesof30–200nmthatarederivedfromthefusionofmultivesicularbodiestoplasmamembranes,andtheyaresecretedintotheextracellularenvironmentbymostcells.Tetraspanins,suchasCD9andCD81,arecharacteristicMarkersandmembranouscomponentsofexosomes[17–21].CD9isatransmembraneproteinassociatedwithangiogenesisviacelladhesion,migration,andsignaltransduction,anditisinvolvedincellfusionprocesseslinkedtofertilization,osteoclastogenesis,andmyogenesis[22–30].EggsanddendriticcellsfromCD9‐deficientmicereleasereducedlevelsofexosomes[31,32].Recentstudieshaveshownthatextracellularvesicles,suchasexosomesderivedfromMSCs,mediateregenerativefunctionsinseveraldiseases,includingthoseaffectingthekidneyandheart[33–35].WealsoreportedthatMSC‐derivedexosomespromotemuscleregenerationbyenhancingmyogenesisandangiogenesis[36].AlthoughtheparacrineeffectofMSCshasbeensuggestedtobethemajormechanismdrivingtissueregenerationafterMSCtransplantation,theparacrinefactorsproducedbyMSCshavenotbeencompletelydefined.TheobjectiveofthisstudywastoaddressthefunctionofMSC‐derivedexosomesasnovelparacrinefactorsinfracturehealingbyusingamousemodeloffracture.

C57BL/6wild‐typemicewereobtainedfromCLEAJapanInc.(Tokyo,Japan,http://www.clea‐japan.com).CD9−/−miceonaC57BL/6backgroundweredescribedpreviously[28]andwerebredinourinstitution.ThisstudywasreviewedandapprovedbytheEthicsCommitteeforExperimentalAnimalsofHiroshimaUniversity,andallanimalsweretreatedaccordingtotheguidelinesoftheInstitutionalAnimalCareandUseCommittee.

Thefracturemodelinvolved77malemicebetween17and19weeksold.AtransversefemoralshaftfracturewasgeneratedbyusingaC‐shapedinstrumentapplyingthree‐pointbending.Therightkneewasexposedbyusingthelateralparapatellarapproachwiththepatellamediallydislocated.Thefemurintercondylargroovewasexposedatthekneejointbyfullflexion,andaburrholeof0.5‐mmdiameterwasmadeinthecenteroftheintercondylargroove.Toavoidsignificantdisplacementofthefracturewhilemaintainingawell‐alignedandstablefracturesite,a0.5‐mm‐diameter24‐gaugeneedlewasinsertedintotheburrholeatthecenteroftheintercondylargroove,andthetipoftheneedlewasrunthroughthetopofthefemurgreatertrochanter.Athinsawcutatadepthof3mmwasappliedmidshaftafterminimallateralexposuretoweakentheboneandtoavoidcomplexfractures.Therightfemurofeachanimalwasfracturedbyathree‐pointbendingtechnique[3,37],afterwhichthemuscularfasciaeandskinwereclosed.

X‐rayandmicrocomputedtomography(μCT)imagingwereperformedbyusingahigh‐resolutionSkyscan1176in‐vivomicro‐CT(Bruker,Madison,WI,https://www.bruker.com).Three‐dimensionalCTimageswererecreatedbyCTAn(version1.15)+CTVol(version2.3)software(Bruker).Thebonedensitiesofthefemurshaftsofwild‐typemice(n=10)andCD9−/−mice(n=10)weremeasuredbycomputedtomographyusingCTAn(version1.15)+CTVol(version2.3)software(Bruker).Radiographicimagingwasperformedat0,1,2,4,and6weeksafterfracture.Boneunionwasdefinedasthepresenceofabridgingcallusontwocortices[3].

Thefemursof1‐month‐oldwild‐typeandCD9−/−micewereharvestedat3and10daysafterfracture.Thesesampleswereembeddedinparaffinafterfixationin4%paraformaldehydeat4°Cfor24hoursanddecalcificationin10%EDTA(Wako,Osaka,Japan,http://www.wako‐chem.co.jp)atroomtemperaturefor14days.Thefemursweresectioned(4.5‐μmslices)alongthelongitudinalaxiswithamicrotomeandstainedwithToluidineblue(Sigma‐Aldrich,St.Louis,MO,http://www.sigmaaldrich.com),hematoxylinandeosin(H&E)(Sigma‐Aldrich),andtartrate‐resistantacidphosphatase(TRAP)(Wako)foranalysisofhistologicdifferences.Atleasttwodifferentsectionspersamplewereanalyzedmicroscopically.Fracturehealingwasevaluatedbyusingahistologicalscoreoffracturehealing[38].TRAP‐positivecells’mm2fieldwascounted.

Sectionsfromparaffin‐embeddedfemurwerefirstdeparaffinizedinxyleneandrehydratedinethanolandwater.AfterwashingwithPBS,sectionswereblockedwith10%serumfor20minatroomtemperature.Antibodiestoα‐smoothmuscleactin(αSM)(Abcam,Austin,TX,http://www.abcam.com;ab5694;1:200)wereaddedandincubatedovernightat4°C.Afterwashingwithphosphate‐bufferedsaline(PBS),sectionswereincubatedwithbiotinylatedsecondaryantibodyfor30minutesatroomtemperatureandthenincubatedwithVectastainABC‐APalkalinephosphatase(VectorLaboratories,Burlingame,CA,http://vectorlabs.com)for30min.Slideswerewashed,andsectionswereincubatedwithalkalinephosphatasesubstrate(VectorLaboratories).

Humanbonemarrow‐derivedMSCs(MSCs)wereobtainedfromLonza(Basel,Switzerland,http://www.lonza.com)andculturedinMSCgrowthmedium(Lonza).ThehumanosteosarcomacelllineHOSwasobtainedfromtheAmericanTypeCultureCollectionandmaintainedinDulbecco’sModifiedEagle’sMedium(DMEM;Wako)with10%fetalbovineserumand1%antibiotic‐antimycoticsolution(NacalaiTesque,Kyoto,Japan,http://www.nacalai.co.jp).TheMSCswereculturedat37°Cunder5%CO2,andMSCsatpassages4–6wereusedforallexperiments.MSCswereseededat1.0×105perwellinasix‐wellplatewithMSCgrowthmedium.Onedaylater,thecellswerewashedwithserum‐freeDMEMandculturedwith2ml/wellserum‐freeDMEMfor48hours.Toisolatetheexosomes,2mlofconditionedmedium(CM)wascollectedandcentrifugedfor15minutesat2,380gandthenfurtherultracentrifugedfor70minutesat180,000gat4°C.Thesupernatantswerecollectedasexosome‐depletedconditionedmedium(CM‐Exo).Theexosomepelletswereresuspendedin100µlofPBS.TheexosomesisolatedfromthesamevolumesofculturemediumandfromthesamenumbersofcellswereresuspendedinPBSorDMEM.CMandCM‐ExowereconcentratedbyusingAmiconUltra‐2centrifugalfilters(EMDMillipore,Billerica,MA,http://www.emdmillipore.com),accordingtotherecommendedprotocol.ForlocalinjectionintoafemurfracturemodelofCD9−/−orwild‐typemice,100µlofexosomes,CM,orCM‐Exowasinjectedintothefracturedpartat1and8daysafterfracture.

ProteinsfromCM,CM‐Exo,andexosomeswereseparatedonMini‐ProteanTGXPrecastGels(Bio‐RadLaboratories,Hercules,CA,http://www.bio‐rad.com)andtransferredtoapoly(vinylidene)difluoridemembrane(Bio‐RadLaboratories).Antiflotillin‐1antibody(610820;diluted1:500;BDBiosciences,SanJose,CA,http://www.bdbiosciences.com;purified),mouseanti‐humanCD9antibody(BDBioscience;diluted1:200),andanti‐CD81antibody(SantaCruzBiotechnology,Dallas,TX,http://www.scbt.com;sc‐7637;diluted1:200)wereusedasprimaryantibodies.Horseradishperoxidase(HRP)‐conjugatedgoatanti‐mouseIgGantibody(SantaCruzBiotechnology;dilutedsc‐2005)orHRP‐conjugatedgoatanti‐rabbitIgGantibody(SantaCruzBiotechnology;sc‐2030)wereusedasthesecondaryantibodies.Thesignalwasdetectedbychemiluminescencewithimmuno‐enhancer(Wako)byusingtheImageQuantLAS4000system(GEHealthcare,Uppsala,Sweden,http://www3.gehealthcare.co.jp).

Werecentlydevelopedahigh‐resolutionfrequencytransmissionelectric‐field(FTE)imagingsystembasedonfieldemissionscanningelectronmicroscopy(SEM)ofJSM‐7000F(JEOL,Tokyo,Japan,http://www.jeol.co.jp),whichenablesobservationofthebiologicalspecimensinwaterwithoutmetalstaining[39].ExosomeswereimagedbyusingtheFTEimagingsystem.Theobservationconditionswerecapturedunderthefollowingparameters:×20,000magnifications,1,280×1,024pixels,80‐secondscanningtime,7‐mmworkingdistance,3‐to4‐kVacceleratingEB,and10‐pAcurrent.OriginalFTEimageswerefilteredbyusingatwo‐dimensionalGaussianfilter(GF)withthekernelsizeof7×7pixelsandtheradiusof1.2σ.BackgroundsubtractionwasachievedbysubtractingFTEimagesfromthefilteredimagesusingabroadGF(400×400pixels,160σ).

Profilingofangiogenesis‐relatedproteins(55proteins)andcytokines(102proteins)inCM,CM‐Exo,andexosomeswasundertakenbyusingaHumanAngiogenesisArraykitandaHumanXLCytokineAntibodyArraykit(ProteomeProfiler,R&DSystems,Minneapolis,MN,https://www.rndsystems.com),accordingtotherecommendedprotocols.Theconcentrationsofmonocytechemotacticprotein‐1(MCP‐1),MCP‐3,andstromalcell‐derivedfactor‐1(SDF‐1)intheCM,CM‐Exo,andexosomesfromMSCsandHOSwerequantifiedbyusingtheBio‐PlexMultiplexsuspensionassaysystem(Bio‐RadLaboratories),accordingtotherecommendedprotocol.

SmallRNAswerepurifiedfromMSCandHOSexosomesbyusingthemirVanamiRNAIsolationKit(ThermoFisherScientificLifeSciences,OakwoodVillage,OH,https://www.thermofisher.com).PurifiedsmallRNAswereconcentratedbyusinganevaporator.TheconcentrationsandqualityofthesmallRNAsfromthesamevolumesofculturemediumforthesamenumbersofcellsweredeterminedwiththeBioAnalyzer2100(AgilentTechnologies,SantaClara,CA,http://www.agilent.com),andsmallRNAwasusedastheinputforthenCounterHumanmiRNAExpressionAssayKit(NanoStringTechnologies,Seattle,WA,http://www.nanostring.com)asdescribedpreviously[36].

Allquantitativedataareexpressedasmean±SD.StatisticalanalyseswereperformedbyPearson’schi‐squaretest,theMann‐WhitneyUtest,ortheSteel‐Dwasstest.Avalueofp.05wasconsideredstatisticallysignificant.

SkeletaldevelopmentinCD9−/−micewasnormal.Bodyweight,bonedensity,andtheentirewidthofthetibialgrowthplateweremeasuredinwild‐typeandCD9−/−mice.Bodyweightandbonedensitywerenotsignificantlydifferentbetweenwild‐typeandCD9−/−miceat17weeksold,justbeforefracture(Fig.1A,1B).Furthermore,toconfirmwhethertherewasaninherentdifferenceinbonegrowth,wemeasuredthewidthofthetibialgrowthplatesinwild‐typeandCD9−/−miceat1monthold.ThewidthsofthetibialgrowthplatesweresignificantlyreducedinCD9−/−micecomparedwithwild‐typemice(Fig.1C).Next,weexaminedtheprocessoffracturehealinginCD9−/−mice.Wild‐typemiceexhibitedendochondralossificationthroughcallusformation2weeksafterfracture,andboneunionwasapparentat3weeks(Fig.2A).However,CD9−/−miceshowedasignificantdelayinfracturehealingcomparedwithwild‐typemice.Boneunionratewas25%inCD9−/−miceat3weeksafterfracture,whichwassignificantlylowerthanwild‐typemice(Fig.2B).TheaverageperiodforboneunioninCD9−/−micewasalsosignificantlylongerthanwild‐typemice(datanotshown).HistologicalanalysesofthefracturerevealedthatendochondralossificationwasparticularlydelayedinCD9−/−micecomparedwithwild‐typemice(Fig.2C)at2and3weeksafterthefracture.Overall,theseresultsshowedthathealingcapacitywasdecreasedinCD9−/−micecomparedwithwild‐typemice.

Bodyweight,bonedensity,andtheentirewidthofthetibialgrowthplateinCD9−/−mice.(A,B):ThebodyweightandbonedensityweremeasuredinWT(n=10)andCD9−/−(n=10)miceat17weeksold,justbeforefracture.(C):ThetibialgrowthplatelengthwasmeasuredinWTandCD9−/−miceat1monthold.GrowthplatelengthsinCD9−/−miceweresignificantlyshorterthanthoseinWT.Scalebars=200μm.Valueswereexpressedasmeans±SE.StatisticalanalysiswasperformedbyMann‐WhitneyUtest.∗,p.05.Abbreviation:WT,wild‐typemice.

ImpairedfracturehealinginCD9−/−mice.(A,B):AtransversefemoralshaftfracturewasproducedbyusingaC‐shapedinstrumentapplyingthree‐pointbendinginWT(n=15)andCD9−/−(n=15)mice.Radiographicimagingbyx‐rayandboneunionratewereperformedat0,1,2,4,and6weeksafterthefracture.StatisticalanalysiswasperformedbyPearson"schi‐squaretest.∗,p.05.(C):ThefemursofWTandCD9−/−micewereharvestedat2and3weeksafterfracture.Afterradiographicimagingbyx‐rayandµCTwasperformed,thefemurswerestainedwithToluidineblue,andH&E.Scalebars=100μm.Abbreviations:μCT,microcomputedtomography;PO,postoperative;0W,0weeks;WT,wild‐typemice.

WeexaminedwhetherdelayedfracturehealinginCD9−/−micewasrescuedbyinjectionofMSC‐derivedexosomes.ExosomeswereisolatedfromMSC‐conditionedmediumbyultracentrifugation(Fig.3A).ExosomemarkersCD9,CD81,andflotillin‐1wereexaminedintheconditionedmedium(CM),supernatant(CM‐Exo),andexosomepelletafterultracentrifugationbyimmunoblotting.ExosomemarkersweredetectedintheCMandexosomepreparations,butnotintheCM‐Exo,afterultracentrifugation(Fig.3B).TheFTEimagingsystemwasdirectlyabletoimagetheuntreatedexosomesinsolution.Theseexosomeimagesclearlyshowedasphericalshapeofapproximately80nm(Fig.3C).TheCM‐Exowaslessthanexosomepreparations(Fig.3C).

IsolationofexosomesfromMSC‐conditionedmedium.(A):MSCswereseededat1.0×105cellsperwellinasix‐wellplatewithMSCgrowthmedium.Onedaylater,thecellswerewashedwithserum‐freeDMEMandculturedwith2mlperwellserum‐freeDMEMfor48hours.Toisolatetheexosomes,2mlofconditionedmedium(CM)wascollectedandcentrifugedfor15minutesat2,380gandthenfurtherultracentrifugedfor70minutesat180,000g.Thesupernatantswerecollectedasexosome‐depletedconditionedmedium.ThepelletswereresuspendedinPBSforuseasexosomes.(B):Immunoblottingforexosomemarkers,CD9,CD81,andflotillin‐1,inExo,CMandCM‐Exo.(C):ImagesoftheuntreatedExoandCM‐ExoinsolutionbyFTEsystem.Originalmagnification×20,000.Scalebars=1μm.Abbreviations:CM,conditionedmedium;CM‐Exo,exosome‐depletedconditionedmedium;DMEM,Dulbecco’smodifiedEagle’smedium;Exo,exosomes;MSC,mesenchymalstemcell;PBS,phosphate‐bufferedsaline.

ToexaminetheeffectofexosomesonfracturehealinginCD9−/−mice,100µlofCM,CM‐Exo,orexosomeswereinjectedintothefracturesite.DelayedfracturehealinginCD9−/−micewasrescuedbytheinjectionofCMandexosomes(Fig.4A).TheinjectionofCMandexosomeswasassociatedwithabundantcallusformation2weeksafterfracture;boneunionat3weeksafterfracturewassimilartowild‐typemice.However,delayedfracturehealinginCD9−/−micewasnotrescuedbyCM‐Exo(Fig.4A).TheaverageperiodforboneunionwassignificantlydifferentbetweenMSCexosome‐treatedandCMExo‐treatedorcontrolgroups(Fig.4B).TheseresultsindicatedthatexosomesrescueddelayedfracturehealinginCD9−/−mice.HistologicalanalysiswithtoluidineblueandH&Estainingrevealedthattherewasacceleratedformationofhypertrophicchondrocytesandwovenboneintheexosome‐injectedCD9−/−mice,andtherewassignificantimprovementinfracturehealinginexosome‐injectedCD9−/−micecomparedwithPBS‐injectedcontrolCD9−/−mice(Fig.4C,4D).Indeed,manyTRAP‐positivecellswerepresentinthecallusesofexosome‐injectedCD9−/−mice10daysaftersurgery,whereasveryfewTRAP‐positivecellsweredetectedinthecallusesofcontrolmice(Fig.4D).CellspositiveforthevascularmarkerαSMwerealsoincreasedincallusesofexosome‐injectedCD9−/−mice,incontrasttocallusesofPBS‐injectedCD9−/−mice(Fig.4D).Overall,thesedataareconsistentwithsupplementingexosomesfacilitatinganimprovedfracturehealingprocessinCD9−/−mice.Furthermore,toexaminewhetherthepromotionoffracturehealingbyexosomesdependsonMSC‐derivedexosomesspecifically,exosomescollectedfromthehumanosteosarcomacelllineHOSwereinjectedintofracturedfemursofCD9−/−miceandcomparedwithmicetreatedwithMSCexosomes.AlthoughMSCexosomespromotedcallusformationat10daysafterthefracture,HOSexosomesfailedtopromotecallusformation(Fig.4E).

MSCexosomesrescueddelayedfracturehealinginCD9−/−mice.(A):Representativeradiologicimagesofcontrol(n=15),CM(n=9),CM‐Exo(n=9),andExo(n=9)treatedCD9−/−miceat0,1,2,4,and6weeksafterfracture.(B):Theaverageperiodforboneunion.StatisticalanalysiswasperformedbySteel‐Dwasstest.#,p.05versuscontrol;∗,p.05versusCM‐Exo.(C):RepresentativehistologicalevaluationofcallusformationafterExoinjectioninCD9−/−mice.ThefemursofPBScontrolandExo‐treatedCD9−/−micewereassayed10daysafterfracture.ThefemurswerestainedwithToluidineblue(PBS,n=6;Exo,n=5),TRAP(PBS,n=4;Exo,n=4),andαSMantibody(PBS,n=4;Exo,n=4)forhistologicanalysisofcallus.Scalebars=200μm(rightcolumn)and100μm(leftcolumn).(D):Quantificationoffracturehealing,osteoclasts(TRAP)andangiogenesis(αSM)incalluses.Valueswereexpressedasmean±SE.StatisticalanalysiswasperformedbyMann‐WhitneyUtest,∗,p.05.(E):RepresentativeradiologicimagesandµCTimagesoffemurs.ThefemursofPBS‐injectedcontrolCD9−/−mice,MSC‐Exo‐treatedCD9−/−mice,andHOS‐Exo‐treatedCD9−/−micewereharvestedat10daysafterfracture.ThefemurswerestainedwithToluidineblue(PBS,n=6;MSC‐Exo,n=5;HOS‐Exo,n=5)forhistologicanalysis.Scalebars=100μm.Fracturehealingscoresweregeneratedafterhistologicalevaluation.StatisticalanalysiswasperformedbySteel‐Dwasstest.∗,p.05.Control,noinjectiongroup.Abbreviations:CM,conditionedmedium;CM‐Exo,exosome‐depletedconditionedmedium;µCT,microcomputedtomography;Exo,exosomes;HOS,humanosteosarcomacells;MSC,mesenchymalstemcell;PBS,phosphate‐bufferedsaline;αSM,α‐smoothmuscleactin;TB,Toluidineblue;TRAP,tartrate‐resistantacidphosphatase;0W,0weeks;WT,wildtype.

CM,CM‐Exo,andMSCexosomeswereinjectedintothefracturedfemursofwild‐typemicetoexaminewhetherexosomespromotefracturehealinginthewild‐typecontext.AlthoughtheinjectionofCMandexosomespromotedboneunionat2weeksafterfracture,boneunionwasnotenhancedbyCM‐Exoinwild‐typemice(Fig.5A).ThetimingofboneunionwassignificantlyshorterinmicetreatedwithCMandexosomescomparedwiththecontrolandCM‐Exogroups(Fig.5B).Theseresultsindicatedthatexosomespromotefracturehealinginwild‐typemice.

MSCexosomespromotefracturehealinginwild‐typemice.(A):Representativeradiologicimagesofthefemursofcontrol(n=15),CM(n=9),CM‐Exo(n=9),andExo(n=9)treatedWTmice.(B):Theaverageperiodforboneunion.StatisticalanalysiswasperformedbySteel‐Dwasstest.∗,p.05versuscontrol.Control,noinjectiongroup.(C):Theconcentrationsofthebone‐repair‐relatedcytokines,MCP‐1,‐3,andSDF‐1weremeasuredinCM,exosomes,andCM‐ExofromMSCsandHOScellsbyusingBio‐Plexassays.SomevaluesinCMandCM‐ExofromMSCswereusedwith29,575.6pg/mlasthemaximumconcentrationwithinrange,becausethevaluesofMCP‐1inCMandCM‐ExofromMSCswereoutofrangeofthestandardcurve.Valueswereexpressedasmean±SE.StatisticalanalysiswasperformedbySteel‐Dwasstest.∗,p.05versusMSC‐Exo;∗∗,p.01versusMSC‐Exo.hMSCwasfromninedonors(n=17).Abbreviations:CM,conditionedmedium;CM‐Exo,exosome‐depletedconditionedmedium;Exo,exosomes;HOS,humanosteosarcomacells;MCP‐1,monocytechemotacticprotein‐1;MSC,mesenchymalstemcell;SDF‐1,stromalcell‐derivedfactor‐1;0w,0weeks.

Cytokineantibodyarrays(102proteins)andangiogenesisantibodyarrays(55proteins)wereusedtoidentifyproteinsthatwerepresentinCM,CM‐Exo,andexosomespreparations.Interestingly,theseassaysrevealedthatthelevelsofcytokinesandangiogenicfactorssuchasvascularendothelialgrowthfactor(VEGF)inMSCexosomeswerelowerthanthoseinCMandCM‐Exo(supplementalonlineFig.1A,1B).Tomoreaccuratelyassesstheconcentrationsofthecytokines,MCP‐1,MCP‐3,andSDF‐1,allknowntobeessentialforfracturehealing[11–14],thelevelsofthesefactorsinCM,CM‐Exo,MSCexosomes,andHOSexosomeswerequantifiedbyusingaBio‐Plexsystem(Fig.5C).SDF‐1,MCP‐1,andMCP‐3levelsinMSCexosomesweresignificantlylowerthanthelevelsinCMandCM‐Exo(Fig.5C).

ToassaymicroRNAs(miRNAs)inMSCexosomes,wecomparedMSCexosomeswithHOSexosomesusingtheNanoStringsystem.Wepreviouslydescribedthetop20mosthighlyexpressedmiRNAsinMSCexosomesandCM‐Exo[36].Thetop30mosthighlyexpressedmiRNAsinMSCandHOSexosomeswereverysimilar(Table1).DifferentiallyexpressedmiRNAsareshownastheratioofmiRNAexpressionlevelinMSCtoHOSexosomes(Table2).Ofthetop10listedinTable2,miR‐4532,miR‐125b‐5p,andmiR‐4516alsoappearedinthetop30mosthighlyexpressedmiRNAsinMSCexosomes(Table1).miR‐338‐3pandmiR‐548aawereexpressedatmorethanthreefoldhigherlevelsinMSCexosomesrelativetoCM‐Exo.

Table1.Top30mosthighlyexpressedmiRNAsinMSCexosomesandHOSexosomesaccordingtoNanoStringmiRNAexpressionprofiling.

Impairedtissuerepairhasbeenlinkedtocompromisedchemotaxis[24],vascularization[22,23],andcellfusion[26]inCD9−/−mice[40–42].Inadditiontothesephenotypes,exosomereleaseisreducedinthesemice[31,32].Thus,wehypothesizedthatimpairedfracturehealinginCD9−/−micemaybelinkedtocompromisedexosomereleasecomparedwithwild‐typemice,andaugmentedfracturehealingmightbeachievedinthismousemodelbysupplementingexosomes.Inthisstudy,therewerenosignificantdifferencesinweightandbonedensitybetweenwild‐typeandCD9−/−miceat17weeksold.AlthoughinvitrostudiesshowthatCD9positivelyregulatesosteoclastogenesisviacellularfusionandMAPKsignaling[29,30],osteoclastactivityunderthetibialgrowthplateissimilarinwild‐typeandCD9−/−miceat8weeksold[43].Similarly,therewerealsonoobviousdifferencesinosteogenesisbetweenwild‐typemiceandCD9/CD81double‐knockoutmice[43].Thus,thesedatasuggestthatosteoclastogenesisandosteogenesisarenormalinCD9−/−mice.However,ourstudyrevealedthatthewidthofthetibialgrowthplatewassignificantlyreducedinCD9−/−micecomparedwithwild‐typemiceat1monthold.Theseresultsmayindicatethattheinitialphaseofendochondralossification,includingchondrocyteproliferationandhypertrophicdifferentiation,wasimpairedinthismodelsystem.Indeed,CD9−/−miceexhibitedretardationofcallusformationcomparedwithwild‐typemice.Furthermore,cellrecruitmentisknowntobeimportantintheinitialphaseoffracturehealing.TheexpressionofCD9onstemcells,includingCD34+hematopoieticstemcells,regulatesmigration,adhesion,andhoming[44,45].Thus,theretardationoffracturehealinginCD9−/−micemightberelatedtoimpairedhomingofstemandprogenitorcellstothefracturesite.However,inourstudy,theexpressionofCD9didnotseemtobeimportantinstemcellhomingfortissueregenerationbecausetheinjectionofMSCexosomesintoCD9−/−micerescuedtheretardationoffracturehealing.Theseresultssuggestthatthereductionand/ordysfunctionofexosomesareatleastoneofthecausesofdelayed‐fracturehealinginCD9−/−mice.

ThetherapeuticpotentialofMSCscontinuestoreceivewidespreadattention[10].WethereforefocusedonthetherapeuticimplicationsofMSCexosomesasanovelparacrinefactorsecretedbyMSCs.ThepresentstudyshowedthatCMandMSCexosomesacceleratefracturehealing.However,exosomesderivedfromHOS,anosteosarcomacellline,didnothavethiscapability.Previously,itwasreportedthatserum‐freeconditionedmediumderivedfromhumanMSCsacceleratescallusformationinamousedistractionosteogenesismodel.MCP‐1/‐3andinterleukin‐3(IL‐3)/IL‐6inCMrecruitbonemarrowstromalcells,matureendothelialcells,andprogenitorcells,andinduceangiogenesisandosteogenicdifferentiation[14].AnothergroupalsoreportedthatMCP‐1andSDF‐1promoteendochondralbonerepairbyrecruitingMSCstothesiteofinjury[12,13].ThesestudiesdemonstratedthefunctionalcapabilitiesofMCP‐1andSDF‐1usingneutralizingantibodiesinMCP‐1−/−andSDF‐1+/−mousemodels.However,ourdatarevealedthattheMCP‐1,MCP‐3,andSDF‐1concentrationsinMSCexosomesweresignificantlylowerthaninCMand/orCM‐Exo.Interestingly,althoughCM‐Exowasnotabletopromotefracturehealing,thelevelsofseveralangiogenicfactors,SDF‐1,MCP‐1,andMCP‐3,inCM‐ExowerehigherthaninMSCexosomes.Notably,MCP‐1levelsweremarkedlylowinHOSexosomescomparedwithMSCexosomes.Angiogenesisalsoplaysanimportantroleinfracturehealing[46].CMcontainshighlevelsofangiogenicfactorssuchasVEGFandIL‐6andenhancesboneingrowthandfracturehealingbystimulatingendothelialcells[47].However,aswepreviouslyreported,althoughVEGFandIL‐6levelsinMSCexosomesarelowcomparedwithCMandCM‐Exo,MSCexosomesinduceangiogenesistoagreaterextent[36].TheseresultssuggestthattheaccelerationoffracturehealingbyMSCexosomesisnotsolelyattributabletoMCP‐1,‐3,SDF‐1,andangiogenicfactors.ExosomescontainmRNAsandmiRNAs,aswellascytokines,andfunctionalRNAscanbetransferredfromonecelltoanotherviatheexosome[48–50].WepreviouslyidentifiedmiRNAsinMSCexosomes,CM,andCM‐Exo.Amongthem,miR‐21,anantiapoptoticmiRNA,wasmostabundantinMSCexosomesandMSC‐CM[36].MiR‐21promotesosteogenicdifferentiationofMSCs[51,52],andlocalinjectionofmiR‐21overexpressingMSCspromotesfracturehealinginaratmodel[53].OuranalysisrevealedthatthelevelsofmiR‐21werehighinbothHOSandMSCexosomes.ManyhighlyexpressedmiRNAsinMSCexosomesweresimilartothoseinHOSexosomes.HighlyexpressedmiRNAsarecommoninexosomesfrommanydifferentcells[36,54,55].However,amongthetop30mosthighlyexpressedmiRNAsinMSCexosomes,several,suchasmiR‐4532,miR‐125b‐5p,andmiR‐338‐3p,weremoreabundantorweresignificantlydifferentiallyexpressedinMSCexosomesrelativetoHOSexosomesorCM‐Exo.ThesemiRNAsarestillunknowntobeinvolvedintissueregeneration,includingfracturehealing.Thus,afuturestudyprobablyshouldbemorefocusedonstemcell‐specificmiRNA.AlthoughitisdifficulttodeterminewhichmiRNAcontributedmosttotheaccelerationoffracturehealingbyMSCexosomes,miRNAinMSCexosomesmayplayacriticalroleintissueregenerationasauniquesetofmiRNA.TheaccelerationoffracturehealingbyMSCexosomesmaynotonlyberelatedtotherecruitmentofstemcellsorprogenitorcellsbycytokinessuchasMCP‐1andSDF‐1,butalsotheinductionofosteogenesisandangiogenesis,whichisatleastinpartmediatedbymiRNAsinexosomes.MiRNAscontroltissuedevelopmentandhomeostasisthroughfine‐tuninggeneexpression[56].MSC‐exosomemiRNAsmightcontributetothedynamicsoftissueregenerationbyinfluencingthemicroenvironmentattherepairsite.Furthermore,itwasrecentlyreportedthatexosomesareeffectivecarriersofsiRNAandmiRNA[57–59].Thus,engineeredexosomesmaybeapotentiallyefficientRNAdeliverysystemforthetreatmentofvariousdiseasesandtissueregeneration,andMSCexosomesmayprovideausefultooltouncovernewmechanismsinvolvedintissuerepair.

ThisstudydemonstratedthatMSC‐derivedexosomesplayanimportantroleinfracturehealingthroughfacilitationofendochondralossification.TheinjectionofexosomesrescuedtheretardationoffracturehealinginCD9−/−mice,whereasexosome‐freeconditionedculturemedium(CM‐Exo)failedtoacceleratefracturerepair.Thus,weconcludethatexosomesareanovelcomponentoftheparacrineeffectofMSCs,whichmodulatefracturehealing.ExosomesmaycontributesignificantlytothemechanismoftissueregenerationbyMSCtransplantation.

WethankT.MiyataforexcellenttechnicalsupportandM.Miyakiforhelpinthestatisticalanalysis.ThisresearchwassupportedbyMEXT/JPSKAKENHIGrant‐in‐AidforScientificResearch(A)21249079(M.O.);EXT/JPSKAKENHIGrant‐in‐AidforScientificResearch(B)15H04959(S.M.);andYoungScientists(A)Grant24689057(S.M.).

T.F.:conceptionanddesign,collectionand/orassemblyofdata,dataanalysisandinterpretation,manuscriptwriting,finalapprovalofmanuscript;S.M.:conceptionanddesign,dataanalysisandinterpretation,manuscriptwriting,finalapprovalofmanuscript;H.I.:collectionand/orassemblyofdata,finalapprovalofmanuscript;T.O.andY.K.:collectionand/orassemblyofdata,manuscriptwriting,finalapprovalofmanuscript;N.K.:collectionand/orassemblyofdata,dataanalysisandinterpretation,finalapprovalofmanuscript;K.M.:provisionofstudymaterialorpatients,dataanalysisandinterpretation,finalapprovalofmanuscript;Y.H.:dataanalysisandinterpretation,manuscriptwriting,finalapprovalofmanuscript;M.O.:collectionand/orassemblyofdata,dataanalysisandinterpretation,manuscriptwriting,finalapprovalofmanuscript.

Pleasenote:Thepublisherisnotresponsibleforthecontentorfunctionalityofanysupportinginformationsuppliedbytheauthors.Anyqueries(otherthanmissingcontent)shouldbedirectedtothecorrespondingauthorforthearticle.

1RhinelanderFW,BaragryR.Microangiographyinbonehealing.I.Undisplacedclosedfractures.JBoneJointSurg.1962;44-A:1273–1298.
2EnnekingWF,EadyJL,BurchardtH.Autogenouscorticalbonegraftsinthereconstructionofsegmentalskeletaldefects.JBoneJointSurgAm.1980;62:1039–1058.
3KodamaA,KameiN,KameiGetal.Invivobioluminescenceimagingoftransplantedbonemarrowmesenchymalstromalcellsusingamagneticdeliverysysteminaratfracturemodel.JBoneJointSurgBr.2012;94:998–1006.
4Granero-MoltóF,MyersTJ,WeisJAetal.Mesenchymalstemcellsexpressinginsulin-likegrowthfactor-I(MSCIGF)promotefracturehealingandrestorenewboneformationinIrs1knockoutmice:AnalysesofMSCIGFautocrineandparacrineregenerativeeffects.StemCells.2011;29:1537–1548.WileyOnlineLibrary
5PittengerMF,MartinBJ.Mesenchymalstemcellsandtheirpotentialascardiactherapeutics.CircRes.2004;95:9–20.
6WatsonL,EllimanSJ,ColemanCM.Fromisolationtoimplantation:Aconcisereviewofmesenchymalstemcelltherapyinbonefracturerepair.StemCellResTher.2014;5:51
7CaplanAI,DennisJE.Mesenchymalstemcellsastrophicmediators.JCellBiochem.2006;98:1076–1084.WileyOnlineLibrary
8Granero-MoltóF,WeisJA,MigaMIetal.Regenerativeeffectsoftransplantedmesenchymalstemcellsinfracturehealing.StemCells.2009;27:1887–1898.WileyOnlineLibrary
9GerstenfeldLC,CullinaneDM,BarnesGLetal.Fracturehealingasapost-nataldevelopmentalprocess:Molecular,spatial,andtemporalaspectsofitsregulation.JCellBiochem.2003;88:873–884.WileyOnlineLibrary
10LiangX,DingY,ZhangYetal.Paracrinemechanismsofmesenchymalstemcell-basedtherapy:Currentstatusandperspectives.CellTransplant.2014;23:1045–1059.
11ToupadakisCA,WongA,GenetosDCetal.Long-termadministrationofAMD3100,anantagonistofSDF-1/CXCR4signaling,altersfracturerepair.JOrthopaedicRes.2012;30111853–1859.WileyOnlineLibrary
12IshikawaM,ItoH,KitaoriTetal.MCP/CCR2signalingisessentialforrecruitmentofmesenchymalprogenitorcellsduringtheearlyphaseoffracturehealing.PLoSOne.2014;9:e104954
13KitaoriT,ItoH,SchwarzEMetal.Stromalcell-derivedfactor1/CXCR4signalingiscriticalfortherecruitmentofmesenchymalstemcellstothefracturesiteduringskeletalrepairinamousemodel.ArthritisRheum.2009;60:813–823.WileyOnlineLibrary
14AndoY,MatsubaraK,IshikawaJetal.Stemcell-conditionedmediumacceleratesdistractionosteogenesisthroughmultipleregenerativemechanisms.Bone.2014;61:82–90.
15MuJ,ZhuangX,WangQetal.Interspeciescommunicationbetweenplantandmouseguthostcellsthroughedibleplantderivedexosome-likenanoparticles.MolNutrFoodRes.2014;58:1561–1573.WileyOnlineLibrary
16LudwigAK,GiebelB.Exosomes:Smallvesiclesparticipatinginintercellularcommunication.IntJBiochemCellBiol.2012;44:11–15.
17PanBT,JohnstoneRM.Fateofthetransferrinreceptorduringmaturationofsheepreticulocytesinvitro:Selectiveexternalizationofthereceptor.Cell.1983;33:967–978.
18JohnstoneRM.Exosomesbiologicalsignificance:Aconcisereview.BloodCellsMolDis.2006;36:315–321.
19ZakharovaL,SvetlovaM,FominaAF.Tcellexosomesinducecholesterolaccumulationinhumanmonocytesviaphosphatidylserinereceptor.JCellPhysiol.2007;212:174–181.WileyOnlineLibrary
20ThéryC,OstrowskiM,SeguraE.Membranevesiclesasconveyorsofimmuneresponses.NatRevImmunol.2009;9:581–593.
21MathivananS,JiH,SimpsonRJ.Exosomes:Extracellularorganellesimportantinintercellularcommunication.JProteomics.2010;73:1907–1920.
22IwasakiT,TakedaY,MaruyamaKetal.DeletionoftetraspaninCD9diminisheslymphangiogenesisinvivoandinvitro.JBiolChem.2013;288:2118–2131.
23KamisasanukiT,TokushigeS,TerasakiHetal.TargetingCD9producesstimulus-independentantiangiogeniceffectspredominantlyinactivatedendothelialcellsduringangiogenesis:Anovelantiangiogenictherapy.BiochemBiophysResCommun.2011;413:128–135.
24QiJC,WangJ,MandadiSetal.HumanandmousemastcellsusethetetraspaninCD9asanalternateinterleukin-16receptor.Blood.2006;107:135–142.
25RecordM.Intercellularcommunicationbyexosomesinplacenta:Apossibleroleincellfusion?.Placenta.2014;35:297–302.
26DouC,LiJ,KangFetal.DualeffectofcyanidinonRANKL-induceddifferentiationandfusionofosteoclasts.JCellPhysiol.2016;231:558–567.WileyOnlineLibrary
27TachibanaI,HemlerME.Roleoftransmembrane4superfamily(TM4SF)proteinsCD9andCD81inmusclecellfusionandmyotubemaintenance.JCellBiol.1999;146:893–904.
28MiyadoK,YamadaG,YamadaSetal.RequirementofCD9ontheeggplasmamembraneforfertilization.Science.2000;287:321–324.
29IshiiM,IwaiK,KoikeMetal.RANKL-inducedexpressionoftetraspaninCD9inlipidraftmembranemicrodomainisessentialforcellfusionduringosteoclastogenesis.JBoneMinerRes.2006;216965–976.WileyOnlineLibrary
30YiT,KimHJ,ChoJYetal.TetraspaninCD9regulatesosteoclastogenesisviaregulationofp44/42MAPKactivity.BiochemBiophysResCommun.2006;347:178–184.
31MiyadoK,YoshidaK,YamagataKetal.ThefusingabilityofspermisbestowedbyCD9-containingvesiclesreleasedfromeggsinmice.ProcNatlAcadSciUSA.2008;105:12921–12926.
32ChairoungduaA,SmithDL,PochardPetal.Exosomereleaseofβ-catenin:AnovelmechanismthatantagonizesWntsignaling.JCellBiol.2010;190:1079–1091.
33BrunoS,GrangeC,DeregibusMCetal.Mesenchymalstemcell-derivedmicrovesiclesprotectagainstacutetubularinjury.JAmSocNephrol.2009;20:1053–1067.
34HerreraMB,FonsatoV,GattiSetal.Humanliverstemcell-derivedmicrovesiclesacceleratehepaticregenerationinhepatectomizedrats.JCellMolMed.2010;146b1605–1618.WileyOnlineLibrary
35LaiRC,ArslanF,LeeMMetal.ExosomesecretedbyMSCreducesmyocardialischemia/reperfusioninjury.StemCellRes(Amst).2010;4:214–222.
36NakamuraY,MiyakiS,IshitobiHetal.Mesenchymal-stem-cell-derivedexosomesaccelerateskeletalmuscleregeneration.FEBSLett.2015;589:1257–1265.WileyOnlineLibrary
37MaesC,CoenegrachtsL,StockmansIetal.Placentalgrowthfactormediatesmesenchymalcelldevelopment,cartilageturnover,andboneremodelingduringfracturerepair.JClinInvest.2006;116:1230–1242.
38OetgenME,MerrellGA,TroianoNWetal.Developmentofafemoralnon-unionmodelinthemouse.Injury.2008;39:1119–1126.
39OguraT.Nanoscaleanalysisofunstainedbiologicalspecimensinwaterwithoutradiationdamageusinghigh-resolutionfrequencytransmissionelectric-fieldsystembasedonFE-SEM.BiochemBiophysResCommun.2015;459:521–528.
40ZhangJ,DongJ,GuHetal.CD9iscriticalforcutaneouswoundhealingthroughJNKsignaling.JInvestDermatol.2012;132:226–236.
41CharrinS,LatilM,SoaveSetal.NormalmuscleregenerationrequirestightcontrolofmusclecellfusionbytetraspaninsCD9andCD81.NatCommun.2013;4:1674
42KawanoN,MiyadoK,YoshiiNetal.AbsenceofCD9reducesendometrialVEGFsecretionandimpairsuterinerepairafterparturition.SciRep.2014;4:4701
43TakedaY,TachibanaI,MiyadoKetal.TetraspaninsCD9andCD81functiontopreventthefusionofmononuclearphagocytes.JCellBiol.2003;161:945–956.
44LeungKT,ChanKY,NgPCetal.ThetetraspaninCD9regulatesmigration,adhesion,andhomingofhumancordbloodCD34+hematopoieticstemandprogenitorcells.Blood.2011;117:1840–1850.
45BrzoskaE,KowalskiK,Markowska-ZagrajekAetal.Sdf-1(CXCL12)inducesCD9expressioninstemcellsengagedinmuscleregeneration.StemCellResTher.2015;6:46
46SaranU,GeminiPiperniS,ChatterjeeS.Roleofangiogenesisinbonerepair.ArchBiochemBiophys.2014;561:109–117.
47WangCY,YangHB,HsuHSetal.Mesenchymalstemcell-conditionedmediumfacilitatesangiogenesisandfracturehealingindiabeticrats.JTissueEngRegenMed.2012;6:559–569.WileyOnlineLibrary
48ValadiH,EkströmK,BossiosAetal.Exosome-mediatedtransferofmRNAsandmicroRNAsisanovelmechanismofgeneticexchangebetweencells.NatCellBiol.2007;9:654–659.
49ZhangY,LiuD,ChenXetal.SecretedmonocyticmiR-150enhancestargetedendothelialcellmigration.MolCell.2010;39:133–144.
50KosakaN,IguchiH,YoshiokaYetal.SecretorymechanismsandintercellulartransferofmicroRNAsinlivingcells.JBiolChem.2010;285:17442–17452.
51LiH,YangF,WangZetal.MicroRNA-21promotesosteogenicdifferentiationbytargetingsmallmothersagainstdecapentaplegic7.MolMedRep.2015;12:1561–1567.
52MengYB,LiX,LiZYetal.microRNA-21promotesosteogenicdifferentiationofmesenchymalstemcellsbythePI3K/beta-cateninpathway.JOrthopRes.2015;337957–964.WileyOnlineLibrary
53SunY,XuL,HuangSetal.mir-21overexpressingmesenchymalstemcellsacceleratefracturehealinginaratclosedfemurfracturemodel.BioMedResInt.2015;2015:412327
54BaglioSR,RooijersK,Koppers-LalicDetal.Humanbonemarrow-andadipose-mesenchymalstemcellssecreteexosomesenrichedindistinctivemiRNAandtRNAspecies.StemCellResTher.2015;6:127
55KatoT,MiyakiS,IshitobiHetal.ExosomesfromIL-1βstimulatedsynovialfibroblastsinduceosteoarthriticchangesinarticularchondrocytes.ArthritisResTher.2014;16:R163
56MiyakiS,AsaharaH.MacroviewofmicroRNAfunctioninosteoarthritis.NatRevRheumatol.2012;8:543–552.
57Alvarez-ErvitiL,SeowY,YinHetal.DeliveryofsiRNAtothemousebrainbysystemicinjectionoftargetedexosomes.NatBiotechnol.2011;29:341–345.
58OhnoS,TakanashiM,SudoKetal.SystemicallyinjectedexosomestargetedtoEGFRdeliverantitumormicroRNAtobreastcancercells.MolTher.2013;211185–191.
59ShimboK,MiyakiS,IshitobiHetal.Exosome-formedsyntheticmicroRNA-143istransferredtoosteosarcomacellsandinhibitstheirmigration.BiochemBiophysResCommun.2014;445:381–387.
AlphaMedPress|318BlackwellStreet|Durham|NC|ContactUs

Pleasecheckyouremailforinstructionsonresettingyourpassword.Ifyoudonotreceiveanemailwithin10minutes,youremailaddressmaynotberegistered,andyoumayneedtocreateanewWileyOnlineLibraryaccount.

Can"tsignin?Forgotyourusername?

Enteryouremailaddressbelowandwewillsendyouyourusername

Iftheaddressmatchesanexistingaccountyouwillreceiveanemailwithinstructionstoretrieveyourusername