Human mesenchymal stromal cells could deliver erythropoietin and migrate to the basal layer of hair shaft when subcutaneously implanted in a murine model

ContentslistsavailableatSciVerseScienceDirect

Tissue and Cell

j o ur na l ho me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / t i c e

Human mesenchymal stromal cells could deliver erythropoietin and migrate to the basal layer of hair shaft when subcutaneously implanted in a murine model

P.L. Mok

^a,c,∗

, S.K. Cheong

^a,b

, C.F. Leong

^c

, K.H. Chua

^d

, O. Ainoon

^c,e

aPPUKM-MAKNACancerCentre,UniversitiKebangsaanMalaysia,KualaLumpur,Malaysia

bFacultyofMedicineandHealthSciences,UniversitiTunkuAbdulRahman,Kajang,Selangor,Malaysia

cDepartmentofPathology,FacultyofMedicine,UniversitiKebangsaanMalaysia,KualaLumpur,Malaysia

dDepartmentofPhysiology,FacultyofMedicine,UniversitiKebangsaanMalaysia,KualaLumpur,Malaysia

eFacultyofMedicineandHealthSciences,UniversitiSainsIslamMalaysia,KualaLumpur,Malaysia

a r t i c l e i n f o

Articlehistory:

Received27November2011 Receivedinrevisedform5April2012 Accepted5April2012

Available online 4 May 2012

Keywords:

Mesenchymalstromalcells Nucleofection

MIDGE Erythropoietin Hemoglobin Hairregeneration

a b s t r a c t

Mesenchymalstromalcells(MSC)areanattractivecell-targetingvehicleforgenedelivery.MIDGE(an acronymforMinimalistic,ImmunologicallyDefinedGeneExpression)constructisrelativelysaferthan theviralorplasmidexpressionsystemasthedetrimentaleukaryoticandprokaryoticgeneandsequences havebeeneliminated.TheobjectiveofthisstudywastotesttheabilityofthehumanMSC(hMSC)to delivertheerythropoietin(EPO)geneinanudemicemodelfollowingnucleofectionusingaMIDGEcon- struct.hMSCnucleofectedwithMIDGEencodingtheEPOgenewasinjectedsubcutaneouslyinMatrigel atthedorsalflankofnudemice.SubcutaneousimplantationofnucleofectedhMSCresultedinincreased hemoglobinlevelwithpresenceofhumanEPOintheperipheralbloodoftheinjectednudemiceinthe firsttwoweekspost-implantationcomparedwiththecontrolgroups.Thebasallayerofthehairshaft inthedermallayerwasfoundtobesignificantlypositiveforimmunohistochemicalstainingofahuman EPOantibody.However,onlyafewbasallayersofthehairshaftwerefoundtobepositivelystainedfor CD105.Inconclusion,hMSCharboringMIDGE-EPOcoulddeliverandtransientlyexpresstheEPOgene inthenudemicemodel.ThesecellscouldbelocalizedtothehairfollicleandsecretedEPOproteinmight havepossibleroleinhairregeneration.

© 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Erythropoietin(EPO)stimulatesredblood-cellproductionand isproducedinthefetalliverand adultkidney(Jelkmann,1992;

Fisher,2003).SincethefirstsuccessfulcloningofthehumanEPO genein1985,recombinantEPO(rhuEPO)hasbecomeatherapeutic optionforrenalanemiainchronicrenalfailure(CRF)patientsand severalformsofnon-renalanemia.InlightofpredictedriseinCRF inagingpopulationsglobally,theneedtodevelopacost-effective alternativeisapparent(Tsakiris,2000;Eckardt,2001).Transplant- ingEPO-producingcells ofhumanoriginisanidealtherapeutic alternativeforCRFpatientsbecauseitwouldbemorecosteffective andlesslikelytoresultintheseriouscomplicationsofpurered-cell aplasiaduetoauto-regulation(Bennettetal.,2004).

∗Correspondingauthorat:DepartmentofPathology,FacultyofMedicine,Uni- versitiKebangsaanMalaysia,56000Cheras,KualaLumpur,Malaysia.

E-mailaddresses:rachelmok2005@gmail.com(P.L.Mok), cheongsk@utar.edu.my(S.K.Cheong),cfleong@gmail.com(C.F.Leong), ckienhui@hotmail.com(K.H.Chua),ainoon@usim.edu.my(O.Ainoon).

ThesuccessofproducingabiologicallyfunctionalEPOdepends onglycosylationoftheEPOprotein(Krantz,1991).TheN-glycans, forexample,playamajorroleinsecretion,molecularstability,sol- ubility, receptorbinding affinityand invivoelimination ofEPO (Elliottetal.,2003).Thechoiceofhostcells,theculturingcondi- tionsandthepurificationproceduresdeterminethecomposition oftheglycanisoformsofarhuEPOpreparation(Jelkmann,2008).

Todate,humanEPOgenehasalsobeentransfectedinmyoblast (Hamamorietal.,1995)andskincells(Gothelfetal.,2010)andhas successfullycorrectedanemiainrenalfailureinamurinemodel.

Mesenchymalstromalcells(MSC)havealsobeenexploredtocarry anddeliverfunctionalEPOininvivostudies(Bartholomewetal., 2001;Eliopoulosetal.,2003,2006).

MSCisatypeofstemcellsthatareeasytobeisolated,robust inexvivogrowthandamenabletogeneticmodifications(Doering, 2008).HumanMSCexpressintermediatelevelsofhumanleuko- cyteantigen(HLA)majorhistocompatibilitycomplex(MHC)class Imoleculesandcanbeinducedbyinterferon-␥toexpressHLA class IIand Fas ligand; theydo not expressthe co-stimulatory moleculesB7-1,B7-2,CD40,orCD40ligand.Theyshouldtherefore berecognizedbyalloreactiveT-cells.However,human, baboon, and murine MSC failed to elicit a proliferative response from 0040-8166/$–seefrontmatter© 2012 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.tice.2012.04.002

allogeneiclymphocytes.Whenpre-culturedwithinterferon-␥for fullHLAclassIIexpression,MSCstillescaperecognitionbyallore- activeT-cells. Unlike other non-professionalantigen-presenting cells,thisfailureisnotreversedbyprovisionofCD28-mediatedco- stimulation.Furthermore,MSCareimmunosuppressiveandinhibit T-cellalloreactivityinducedinmixedlymphocyteculturesorby non-specificmitogens(Tseetal.,2003;SingerandCaplan,2011).

MSCalsopossessthe capabilitytomigratetosites oftissue damageorinflammation.Thisabilitytomigratetositesofacute tissueinjuryhasbeendemonstratedinthesettingofbonefracture, cerebralischemiaandininfarctedheart(CaplanandBruder,2001;

Tomitaetal.,2002;ParekkadanandMilwid,2010).Localdelivery ofEPObyMSCisclinicallyusefultorescuedamagedtissuesfrom compromisedlocalcirculationsuchascriticalhind-limbischemia, myocardialinfarctionorspinalcordinjury.Thereareampleani- malstudiestoshowthatEPOcouldinduceformationofnewblood vesselsindamagedtissues(Meeretal.,2005;Zhangetal.,2007).

Ourpreviousexperimentshaveshownthathumanmesenchy- malstromalcells(hMSC)isolatedfromadultbonemarrowcould be used as a carrier for the EPO gene. The transfected hMSC couldstablyexpresstheEPOproteinasdeterminedbyenzyme- linkedimmunosorbentassay(ELISA)(Moketal.,2012), leading toinductionofdifferentiationofhumanhematopoieticstemcells intoerythroidcolonyinvitro(Moketal.,2008).Inthesestudies, transfectionwasachievedthrough nucleofectionofa newcon- struct,namedMIDGE(Minimalistic,ImmunologicallyDefinedGene Expression).MIDGEisa linear,double-strandedDNA consisting solelyoftheexpressioncassette,cappedwithhairpinstructures attheendsforprotectionagainstexonucleasedegradation.The MIDGE construct offers the following advantages: (i) the con- structcontainsonlythenecessaryelementsrequiredtoexpress thedesiredgenewithnoresistancemarkersorotherunwanted genes,(ii)immune-stimulatorysequences,suchasCpGs,aremini- mized,(iii)theconstructeliminatestheriskofrecombinationand mutagenesisofwild-typeviruses,and(iv)theconstructcouldbe transferredeasilyasitissmallerinsizecomparedwiththeplasmid system(Schakowskietal.,2001).

In the current study, we aimed to test the hMSC in carry- inganddeliveringtheEPOgeneinaMIDGEconstructfollowing nucleofectioninaninvivosystem.Thenucleofectedcellsweresub- cutaneouslyimplantedinthedorsalflankofnudemiceandthe effectsoftheexpressedEPOproteinonthehemoglobinlevelsin theperipheralbloodwerestudiedfortwomonths.Theimplanted cellsexpressingEPOproteinwerefoundtobelocalizedinthesub- cutaneouslayerofmice.

2. Materialsandmethods

2.1. NucleofectionofMIDGE-EPOintohMSC

NucleofectionofMIDGE-EPOintohMSCwasperformedusing U-23pulsingprogramaspreviouslydescribed(Moketal.,2008).

Generally,2␮gofMIDGE-EPOwasusedtonucleofect5×10⁵hMSC and seededon 35mm cellculture petri dishes containing Dul- becco’s Modified Eagle Medium supplemented with 10% fetal bovine serum and 1% penicillin and streptomycin antibiotics (Gibco-Invitrogen, Grand Island, NY). After 24h, the adherent nucleofectedcellswerethentrypsinizedforimplantationintonude mice.

2.2. Westernblotanalysis

TodetectthepresenceofEPOinthesupernatantofnucleofected cells,Westernblotanalysiswasperformedusingarabbitpolygo- nalanti-humanEPOpolyclonalantibody(Cat.No.H-162)(Santa

CruzBiotechnology;SantaCruz,CA).TheEPOimmunoreactivity wasvisualized using theWesternBreeze Chromogenic Western Blot Immunodetection Kit (Gibco-Invitrogen). For positive con- trol,rhuEPO(Cat.No.02625)(StemCellTechnologies,Vancouver, Canada)wasused.

2.3. ImplantationofhMSCintomice

In a preliminarystudy onthe sustainability of hMSCin the Matrigeloneweekpost-implantation,aBalb/cmousewasused.

Approximately5×10⁵hMSCwerestainedusingthePKH-26red fluorescentcelllinkerminikit(Sigma–Aldrich,St.Louis,MO,USA) according to the protocol recommended by the manufacturer.

Thesecellswerethenwashedwithphosphatebufferedsaline(PBS) bycentrifugingthecellsat200×gfor10min,suspendedin50␮l ofPBSandmixedwith0.5mlthawedBDMatrigelMatrixHighCon- centration(Cat.No.356231)(BDBiosciences,Bedford,MA)at4^◦C.

TheMatrigelmixturewastheninjectedsubcutaneouslyintothe leftflankofthemouse.Atbodytemperature,theMatrigelwould rapidlyacquireasemisolidform.

To study the delivery of EPO protein by the nucleo- fected hMSC, experiments were performed on 12-week old Nu/Nu male mice (BioLASCO, Taipei, Taiwan) with proto- cols approved by the Universiti Kebangsaan Malaysia Animal Ethics Committee (PP/PAT/2008/AINOON/12-AUGUST/227-JAN- 2009-DEC-2011).Sixmicewereusedforeachgroupofexperiment:

miceimplanted subcutaneously withMatrigelmixed withPBS, mice implanted with non-nucleofected hMSC encapsulated in Matrigel and mice implanted with nucleofected hMSC (hMSC- EPO)encapsulatedinMatrigel.Approximately5×10⁵cellswere injectedintothesubcutaneoustissuesofthenudemiceaccording totheprotocolstatedabove.Non-nucleofectedcellswereprepared bynucleofectingcellswithoutMIDGE-EPO.

2.4. DeterminationofhumanEPOandhemoglobinlevelsinthe blood:

Undergeneralanesthesia,100␮lofbloodwaswithdrawnfrom the retro-orbital venous plexus of the mice using heparinized capillarytube(HirschmannLaborgerate,Eberstadt,Germany)and transferredintoa0.5mlethylenediaminetetraaceticacid(EDTA) tube(Greiner Bio-One GmBH,Kremsm ˝unter, Austria). Thetube containingthebloodsamplewasthencentrifugedat400×gfor 20mintoobtaintheplasmaforhumanEPOmeasurementusing theHumanErythropoietinELISAImmunoassayKit(StemCellTech- nologies).Hemoglobinmeasurementwasperformedaccordingto theproceduresdescribedinChoudhrietal.(1997).

2.5. ImmunohistochemicalstainingonMatrigelimplantsand otherorgans

At 2 and 8 weeks post-implantation, the mice were sacri- ficedbycervicaldislocationtoharvesttheimplant,togetherwith theskinandthemusclelayer.Theharvestedimplantswerefix- ated and sectionsof 3␮mwere prepared for hematoxylin and eosin(H&E)(Sigma–Aldrich),andimmunohistochemicalstaining usingtheLSAB+System-HorseradishPeroxidaseKit(DakoCytoma- tion,Glostrup, Denmark). For thedetection of humanEPO and CD105,anti-humanrabbitEPO(Cat.No.sc-7956;dilution1:200) (SantaCruzBiotechnology)andanti-humanmouseCD105(Cat.No.

555690;dilution1:200)(BDBiosciencesPharmingen,SanDiego, CA)primaryantibodieswereused.TolocalizehMSCinothermajor organs,thelung,heart,spleen,kidney,liverandbrainwerealso harvestedonthesecondandeightweeks,andstainedaccordingto theaboveprocedure.

2.6. Statisticalanalysis

Resultsareexpressedasmean±S.E.M.Theresultswereana- lyzedwithMann–WhitneyUtest.Two-sidedpvaluelessthan0.05 wereconsideredstatisticallysignificant.

3. Results

3.1. NucleofectedhMSCexpressedandsecretedhumanEPO proteinintothesupernatantofculturemedium

IsolatedhMSC wereexpanded as previouslydescribed (Mok etal.,2008).ThehMSCwereshowntoexhibitCD73,CD90and CD105,butnotCD34andCD45phenotypes(datanotshown;see Moketal.,2003;Choongetal.,2007;Wongetal.,2008).Cultured cellsfrompassage3to5werethennucleofectedwith2␮gMIDGE- EPOusingtheU-23pulsingprogram.Adayfollowingnucleofection andpriortoimplantationofthenucleofectedcellsintothenude mice,aWesternblotanalysiswasperformedtodemonstratesuc- cessfulnucleofectionofMIDGE-EPOinthehMSCsamples,andthat thesecellsexpressedandsecretedtheEPOprotein(Fig.1).Super- natantsobtainedfromthreeindependentlynucleofectedsamples showedbandswithamolecularmass(36kDa)correspondingtothe rhuEPOusedasthepositivecontrol.QuantificationofEPOprotein inthesupernatantusingHumanErythropoietinELISAImmunoas- say Kit (Stem Cell Technologies, Vancouver, Canada) showed highest level of expressed protein on day 5 (1.70±1.30U/ml) and sustained at (0.41±0.32)U/ml to (0.80±1.37)U/ml from day 15 to day 50 post-nucleofection (data not shown; see Moketal.,2012).

3.2. NucleofectedhMSCcouldtransientlydeliverhumanEPOinto nudemiceandraisedthehemoglobinlevelsintheblood

ThepracticaluseofhMSCforthesecretionoftherapeuticgene productsincell-basedtherapyapplicationreliesontheircapacityto engraftandsustaininvivofollowingtransplantation(Ankrumand Karp,2010;KarpandTeo,2009).Hence,wechosetouseMatrigelas abiomaterialsubstratetoembedthecellstoavoidlossofcellsinto theemptyspacesofsubcutaneouslayerandothertissues,andto enhancesurvivabilityinthelessoxygenatedsubcutaneoustissue, whichmighteventuallyleadtocelldeath,andthereforeaffectour results(Tomaetal.,2002;McGinleyetal.,2011).Themechanical propertiesofthesubstratetowhichcellsadherehavebeenfound tomediatemanyaspectof cellularfunctionincludingprolifera- tion,migration,differentiation,andalsotoimprovegenedelivery activities(Kongetal.,2005).

Prior to implantation of the nucleofected cells, a prelimi- nary study was performed on a Balb/c mouse to determine the sustainability of hMSC in the Matrigelimplant. hMSC was labeledwithPKH-26redfluorescentstainbeforemixingwithcold Matrigeland themixturewassubcutaneously injectedinto the leftflankof aBalb/c mouse.Oneweekfollowing theimplanta- tion,theMatrigelwasharvested(Fig.1B),fixatedandsectioned to monitor migration of the implanted cells. Under the fluo- rescence microscope, mostred fluorescent cells were foundat theedgeoftheMatrigelimplant(Fig.1C)indicating thathMSC couldsurviveand sustain inthe Matrigelsubstratein thesub- cutaneous tissuesof immunocompetentmouse one week after injection.

Thenucleofectedcellswerenextinjectedintothesubcutaneous tissues of nude mice.Following implantation, blood was with- drawnfromtheretro-orbitalvenousplexuseveryweekforeight

Fig.1.ValidationofEPOexpressionandsecretionbythenucleofectedhMSCandsustainabilityofinjectedhMSCintheMatrigelsubstrate.(A)WesternblotdetectionofEPO (indicatedwitharrow)inthesupernatantofthreenucleofectedhMSCsamples(labeledashMSC1-EPO,hMSC2-EPOandhMSC3-EPO)24hafternucleofection.Arecombinant humanEPO(rhuEPO)wasusedasthepositivecontrol.Supernatantsharvestedfromano-DNAcontrolfornucleofectionandculturemediumwereusedasnegativecontrols.

(B)Priortoinjection,hMSCwaslabeledwithPKH-26redfluorescentstain.ThefigureshowsMatrigelexcisedfromthemouseoneweekpost-implantation.(C)TheMatrigel implantwasfixatedinparaffinwax,sectionedandobservedunderafluorescentmicroscope(magnification40×).

weeks.Thebloodwasthenanalyzedforthepresenceofhuman EPOintheplasmaandforthehemoglobinlevels.Inthefirstweek followingimplantation, theEPOconcentrationwassignificantly higherin thetest groupofhMSC-EPO (3.94±0.95mU/ml)than thecontrolgroupsofhMSC(0.60±0.03mU/ml)orMatrigelonly (0.64±0.04mU/ml)(Fig.2A).Thelevel ofEPOinthehMSC-EPO group,however,droppedto1.01±0.24mU/mlonweek2,which wasslightlyhigherthanthecontrols,andtoalevelcorresponding tothecontrolgroupsonweek3onwards(Fig.2A).

Fig.2. TimecourseofcirculatinghumanEPOandhemoglobinlevelsafterimplan- tationofhMSCintothenudemice.Bloodsampleswerecollectedatdifferenttime pointsfordeterminationof(A)plasmahumanEPOand(B)hemoglobinlevelsinmice implantedwithnucleofectedcells(hMSC-EPO)andincomparisonwiththecon- trols(mixtureofMatrigelandhMSCorMatrigelalone).N=6forallgroups.*p<0.05 betweengroupofmiceimplantedwithnucleofectedhMSC(Matrigel+hMSC-EPO) andnon-nucleofectedcells (Matrigel+hMSC). ^†p<0.05betweengroupof mice implantedwithhMSCandMatrigelalone.

Ourresultsintheinjectednudemiceshowedthattherewas nosignificantdifferenceinthehemoglobinlevelsbetweenthetwo controlgroupsofhMSCandMatrigelexceptinweeks1and3.The hemoglobinlevelwassignificantlyraisedinmiceimplantedwith hMSC-EPOonweek1 (364.34±9.47mg/ml)compared withthe controlgroupof non-nucleofectedhMSC(277.48±15.28mg/ml) (Fig.2B).Thesignificantriseinhemoglobincontinueduntilweek 7, except in week 3 when the test group of hMSC-EPO had lowerplasmaHb(332.16±20.39mg/ml)thanthecontrolgroup implanted withMatrigelonly (355.36±18.99mg/ml). On week 6, the mice implanted with hMSC-EPO had a higher Hb level (354.95±25.06mg/ml) than in thecontrol groups, hMSC alone (299.37±16.41mg/ml)andMatrigelalone(319.19±18.99mg/ml), butthedifferenceswerenotsignificant(Fig.2B).

3.3. LocalizationofhMSCexpressingEPOproteininthe subcutaneouslayerofnudemice

TostudytheinvivofateoftheinjectedhMSC,H&Eandimmuno- histochemicalstainingswereperformedtolocateEPO-producing hMSCintheMatrigelimplantandtheadjacenttissuesoftheskin andmuscleaftertwoandeightweeksofsubcutaneousinjection.

TheH&E stainingonthe epidermaland dermal layerofthe skinofimplantsharvestedfromeachofthehMSC-EPOandnon- nucleofectedhMSCgroupsafter2weeksofimplantationisshown inFig.3AandB.IntheMatrigelsectionfrombothsamples,thepres- enceoffoamymacrophages,eosinophilsandfibroblast-likecells wasobserved(Fig.3CandD).

WhenthesectionswerestainedforhumanEPO,itwasfound thatthebasallayerofthehairshaftwasstainedpositivelyinmouse implantedwithhMSC-EPO(Fig.4A,CandE).Miceinjectedwith non-nucleofectedhMSCwerestainednegativelyforhumanEPO (Fig.4B,DandF).InthemiceimplantedwithhMSC-EPO,onlya fewhairshafts,andnotthefibroblast-likecellsintheMatrigelsec- tion,werestainedpositivelyforhumanCD105(Fig.4G)indicating thatthebasallayerofthehairshaftwasofhMSCorigin.Todeter- minepossiblemigrationtootherorgans,wehaveharvestedthe lung,heart,spleen,kidney,liverandbrainandstainedthemwith bothantibodies.However,allorganswerestainednegativelyfor EPO-producingorCD105positivecells.Thesamenegativestain- ingresultswereobtainedfromalltheorgansincludingMatrigel implantsorhairshaftsafter8weekspost-implantation(datanot shown).

4. Discussion

WehavepreviouslyisolatedhMSCfromadultmarrowandour invitrostudiesdemonstratedthepotentialoftheMIDGEconstruct forextendedandstableEPOproteinexpressionupto55daysin hMSCcultures(Moketal.,2012).Inthiswork,wecarriedoutan invivostudyinnudemicetotesttheabilityofhMSCtodeliverEPO proteinviatheMIDGEconstruct.

Our results showed that hMSC could deliver EPO protein intotheperipheralblood ofnude micefollowingsubcutaneous implantation.PresenceofsecretedhumanEPOcouldenhancethe productionofredbloodcellsand,thus,resultedinasignificantrise inhemoglobinlevelsinthehMSC-EPOgroup(Fig.2).Despitethe transientexpressionofhumanEPOintheplasma,thesustenance ofhigherhemoglobinlevelsuntilweek7wasprobablyduetothe longlifespanofredbloodcells(Bannerman,1983)orprolonged effectofEPOproteinitself(Katoetal.,1997).Thedropofconcen- trationatweek3inthecurrentstudycorrespondedtoourinvitro datathatquantifiedontheEPOproteinlevelinthesupernatant ofnucleofectedcells.Thedeclineintranscriptionalactivity,which wasinaccordwiththosereportedforothergenefromplasmids

Fig.3. ObservationofinjectedhMSCintheMatrigelimplantandtissuechangesintheskinandmuscleofnudemicebyhematoxylinandeosin(H&E)staining.Sectionsof implantconsistinglayersofskin,Matrigelsubstrateandmuscleharvestedfrommouseinjectedwith(A)hMSC-EPO(40×)or(B)hMSCalone(40×).IntheMatrigelsection, (C)eosinophils(indicatedwitharrows)(400×)and(D)foamymacrophagesandfibroblast-likecells(bothindicatedwitharrows)(400×)werefoundinthemouseinjected withhMSC-EPOorhMSCalone.

(HerweijerandWolff,2003),couldbeduetopromotersilencing (Kimetal.,2011).Althoughourpreviousdatashowedsustained geneexpressionfromday15today50post-nucleofection(Mok etal.,2012),thesecretedEPOproteinmightnotbedetectedby ELISAmethodwhendilutedintothebloodstream.Inadditionto promotersilencing,immuneresponsestothecellsintheinvivo systemcouldbetriggeredbytheEPOgeneconstruct(Lifshitzetal., 2009,2010)aswellasthemethodofgeneintroductionintothe cells(Belletal.,2010).InMHCunmatchedallogeneicmicemodel, EPOauto-antibodycouldalsoformandneutralizetheactivityof secretedEPOprotein(Eliopoulosetal.,2005;Campeauetal.,2009).

LocaldeliveryofEPObyMSChasmanifestedamorepotentther- apeuticeffectfortreatmentofcerebralischemia(Esneaultetal., 2008)andmyocardiuminfarction(Coplandetal.,2008)thanwith deliveryofMSCorEPOaloneininvivostudies.Inneurodegener- ation,hMSCcouldserveasavehicletodeliverEPOproteininto injuredtissuesasrhuEPOcouldnotpassthroughtheblood–brain barrier.WhilehighconcentrationsofrhuEPOcouldbeadminis- teredintravenouslytoenablesomeproteintoreachtheinjured sites,therearealsorisksofthrombosisandhypertensiontobecon- sidered(BaskinandLasker,1990;LooandBeguin,1999;Ghezziand Brines,2004;Lieutaudetal.,2008).Therefore,inatissuerepar- ativetreatmentstrategythatonlyrequiresshort-termsupplyof EPOprotein,deliveryofhMSCharboringtheMIDGE-EPOshouldbe sufficientlybeneficial.

The transient EPO expression demonstrated by our current study,however, shouldnot hamperfutureeffortto isolateand poolcellsthatcarrytheMIDGE-EPOgenebeforeimplantationin ordertoimprovetheoutcomeoftheEPOproteindelivery.There havebeenanumberofreportsthathaveusedviralvectorscarry- inggenefortransductionintohMSCandshowedsustainablyhigh EPOexpressionintheanimalmodels(Bartholomewetal.,2001;

Dagaetal.,2002;Eliopoulosetal.,2003,2006;Liuetal.,2000;

Wangetal.,2009).Althoughtheresultsarepromising,weshould notruleouttheriskofoncogeneactivationandtheimmunotox- icitythattheviralvectorsmightpose.StableexpressionofhMSC nucleofectedwithMIDGE-EPOhassignificantbenefitforcorrection ofanemiainpatientssufferingfromchronickidneydisease(Brines andCerami,2008).Currenttreatmentforpatientswithchronickid- neydiseasesufferingfromanemiainvolvesalong-termregiment ofrepetitiverhuEPOinjections.Itisnotonlycostly,butalsocar- riesriskofinfection,possibledevelopmentofpurered-cellaplasia andpoorcompliancetotreatment.Ofnote,testsonthesafetyof MIDGEconstruct,includingstudyonthesiteofintegration,should bewarrantedbeforeitcouldbetranslatedintoclinicaluse.

To monitorthe implantedcells, the Matrigelimplants were harvestedfromtheinjectedmice.Twoweekspost-implantation, thepresenceoffoamymacrophages,eosinophilsand fibroblast- likecells wasobserved(Fig.3).Despite evidencesthat MSCare transplantableacrossallogeneicbarriers,immunereactioncanstill occurinaxenogenicmodel.Inastudy,injectionofhMSCatthesite ofinfarctedmyocardiumofSpraque-Dawley(SD)ratshastriggered infiltrationofimmunecellsasearlyasseconddaypost-injection.

The infiltrationbecame massiveand no presenceof hMSC was foundafteroneweek.Theinfiltrationofcellswere,however,less prominentinimmunosuppressedSDandathymicrats(Grinnemo etal.,2004).Meanwhile,thefibroblast-likecellscouldhavemostly originatedfrom thehostas ourresultsshowedthat theywere stainednegativelytoanti-humanCD105.ItislikelythatthehMSC embeddedintheMatrigelwaseitherremovedbymacrophagesas aresultoftheinflammatoryreactionorhadmigratedoutofthe Matrigel. Thiswassupportedby ourunreportedinvitro obser- vation, which showed that the cells encapsulated in thesame concentrationofMatrigel,whenincubatedwithcompletemedium, couldactuallymigrateoutfromitandattachontothesurfaceof culturedish.

Fig.4. LocalizationofEPO-producinghMSCinthebasallayerofhairshaft.Immunohistochemicalstainingusingananti-humanEPOantibodyshowed(A,CandE)positive stainingonthebasallayerofthehairshaftinthedermalofnudemouseimplantedwithhMSC-EPOand(B,DandF)negativestainingonthebasallayerofthehairshaftinthe nudemouseimplantedwithnon-nucleofectedhMSCalone.Therewasafewhairshaftsthatdemonstratedpositivestainingtoanti-humanCD105antibody(G)(indicated byarrow)inthemouseinjectedwithhMSC-EPO.

TheEPO-producingcellsintheMatrigelimplantsweresignif- icantlystainedpositiveforEPOinthebasallayerofhairshaftin thedermallayeroftheinjected mouse(Fig.4A, Cand E),sug- gesting thathMSC had migratedto theskinlayerand possibly contributedtotheformationofhairfollicles.ThehMSCcouldhave formedthebasallayerandlosttheCD105expressionmarkeras therewasonlya minutenumberofhairfolliclesdemonstrating positivestainingforCD105(Fig.4G).Thereareevidencesshowing thatadherentfibroblast-likecellsderivedfromdermalpapillaof hairbulbsdemonstratedpropertiessimilartoMSCandthattrans- planteddermalpapilladerivedfromMSCcouldinducenewhair folliclesinathymicmice(Hooduijnetal.,2006;Stennetal.,2007;

Yooetal.,2010).Ofparticularinterest,itremaineduncleartous whethercontributiontoformation of hairfollicles wasspecific toEPO-producingcellsaswewereunabletolocatepositivecells withCD105expressiononthebasallayerofhairshaftofmouse transplantedwithhMSCalone.Histologicalmicroscopicapproach issubjectedtovariabilityandsamplingerrorsasonlyafewsections (andusuallyonlyalimitednumberofopticalfieldspersection) aresampledfromtheharvestedexplantandtissues.Inadditionto that,animalshavetobebiopsiedoreuthanizedforthecollectionof

specimensandtherefore,longitudinaltrackingofcellfateisdif- ficultorimpossibleinthesameanimal.Thus,wesuggestfuture investigationonthespecificityofEPO-producingcellstomigrate andengraftintheenvironmentsurroundingthehairfolliclesby usingamoresensitivecelltrackingmethod,forexample,labelingof cellswithironparticlesandvisualizationwithmagneticresonance imaging(Terrovitisetal.,2010).

Nevertheless,itseemedtoustheenvironmentsurroundingthe hairfolliclescouldhaveformedasuitablenichefortheengraftment ofinjectedhMSC,andsecretedEPOproteinfromthenucleofected cells might have enhanced effects in hairgrowthin ourstudy.

Recently,therewasfunctionalevidenceoftheEPOproteininhair regenerationinbothinvitroandinvivomodel.Kangetal.(2010) hasreportedthatrhuEPOcouldsignificantlyelongatehairshafts withincreasedproliferationof matrix keratinocytesin cultured humanhairfollicles.Additionally,theydemonstratedpresenceof EPO receptor in humandermal papilla cells (DPC)and invitro incubationwithrhuEPOhassignificantlystimulatedrateofDPC proliferationinadose-dependentmanner.EPOreceptor(EPO-R) signalingpathwaymediatorssuchasEPO-RandAktwerephos- phorylatedbyEPOproteininhumanDPCs.Subcutaneousinjection

ofbeadscontainingrhuEPOintotheC57BL/6micehadnotonlypro- motedanageninductionfromtelogenbutalsoprolongedanagen phase.Inanotherstudy,invitroexperimentsonauditoryhaircells showeda protectiveeffectofEPOin ischemia-and gentamicin- inducedhair-celldamages(Naldietal.,2009).

5. Conclusions

Inconclusion,hMSCharboringMIDGE-EPOcoulddeliverand transientlyexpresstheEPOgeneinthenudemicemodel.These cellscouldmigrateandformthebasallayerofthehairshaft,and secretedEPOcouldpossiblybeusefulinhairregenerationstrategy.

Authors’contributions

P.L.Mok:Conceptionand design, collectionandassembly of data, data analysis and interpretation, manuscript writing. S.K.

Cheong:Conceptionanddesign,financialsupport,dataanalysisand interpretation,approvalofmanuscript.C.F.Leong:Administrative support,provisionofstudymaterialsofpatients,dataanalysisand interpretation.K.H.Chua:Conceptionanddesign,dataanalysisand interpretation.O.Ainoon:Conceptionanddesign,dataanalysisand interpretation,approvalofmanuscript.

Competinginterests

Theauthorsdeclarethattheyhavenocompetinginterests.

Acknowledgements

We acknowledgethegeneroussupport fromMalaysiaToray Science Foundation (MTSF), Scientex Foundation and National Cancer Council (MAKNA) of Malaysia. We thank the staff and fellow researchers from the Cell Banking Unit, Genetic Cancer Laboratory,HemostasisLaboratory,TissueEngineeringLaboratory, Hematology Laboratory and Histopathology Laboratory of Uni- versitiKebangsaanMalaysiaMedicalCenter,andAnimalHolding FacilityinUniversitiKebangsaanMalaysiaBangifortheirtechni- calguidanceandassistance.WearegratefultoDr.Naseemfrom HistopathologyDepartmentofUKMMedicalCenterforreadingthe slides,andProfessorChooKong BungofUniversitiTunkuAbdul Rahmanforeditorialcommentsonthemanuscript.

References

Ankrum,J.,Karp,J.M.,2010.Mesenchymalstemcelltherapy:twostepsforward,one stepback.TrendsMol.Med.16(5),203–209.

Bannerman,R.M.,1983.Themouseinbiomedicalresearch.In:Hematology,vol.II.

AcademicPress,NewYork,pp.294–308.

Bartholomew,A.,Patil,S.,Mackay,A.,Nelson,M.,Buyaner,D.,Hardy,W.,Mosca,J., Sturgeon,C.,Siatskas,M.,Mahmud,N.,Ferrer,K.,Deans,R.,Moseley,A.,Hoffman, R.,Devine,S.M.,2001.Baboonmesenchymalstemcellscanbegeneticallymod- ifiedtosecretehumanerythropoietininvivo.Hum.GeneTher.12,1527–1541.

Baskin,S.,Lasker,N.,1990.Erythropoietin-associatedhypertension.N.Engl.J.Med.

323(14),999.

Bell,J.B.,Aronovich,E.L.,Schreifels,J.M.,Beadnell,T.C.,Hackett,P.B.,2010.Dura- tionofexpressionandactivityofSleepingBeautyTransposaseinmouseliver followinghydrodynamicDNAdelivery.Mol.Ther.18(10),1796–1802.

Bennett,C.L.,Luminari,S.,Nissenson,A.R.,Tallman,M.S.,Klinge,S.A.,McWilliams, N.,McKoy,J.M.,Kim,B.,Lyons,E.A.,Trifilio,S.M.,Rasich,D.W.,Evens,A.M.,Kuzel, T.M.,Schumock,G.T.,Belknap,S.M.,Locatelli,F.,Rossert,J.,Casadevall,N.,2004.

Purered-cellaplasiaandepoetintherapy.N.Engl.J.Med.351,1403–1408.

Brines,M.,Cerami,A.,2008.Erythropoietin-mediatedtissueprotection:reduc- ingcollateraldamagefromtheprimaryinjuryresponse.J.Intern.Med.264, 405–432.

Campeau,P.M.,Rafei,M.,Frac¸ois,M.,Birman,E.,Forner,K.,Galipeau,J.,2009.

Mesenchymalstromalcellsengineeredtoexpresserythropoietininduceanti- erythropoietinantibodiesand anemiain allorecipients. Mol. Ther. 17(2), 369–372.

Caplan,A.I.,Bruder,S.P.,2001.Mesenchymalstemcells:buildingblocksformolec- ularmedicineinthe21stcentury.TrendsMol.Med.7(6),259–263.

Choong,P.F.,Mok,P.L.,Cheong,S.K.,Then,K.Y.,2007. Generatingneuron-like cellsfromBM-derivedmesenchymalstromalcellsinvitro.Cytotherapy9(2), 170–183.

Choudhri,T.F.,Hoh,B.L.,Solomon,R.A.,Connolly,E.S.,Pinsky,D.J.,1997.Useof aspectrophotometrichemoglobinassaytoobjectivelyquantifyintracerebral hemorrhageinmice.Stroke28,2296–2302.

Copland, I.B.,Jolicoeur, E.M.,Gilis, M.,Cuerquis, J.,Eliopoulos, N.,Annabi, B., Calderone,A.,Tanguay,J.,Ducharme,A.,Galipeau,J.,2008.Couplingerythro- poietinsecretiontomesenchymalstromalcellsenhancestheirregenerative properties.Cardiovasc.Res.79,405–415.

Daga,A.,Muraglia,A.,Quarto,R.,Cancedda,R.,Corte,G.,2002.Enhancedengraft- mentofEPO-transducedhumanbonemarrowstromalcellstransplantedina 3Dmatrixinnon-conditionedNOD/SCIDmice.GeneTher.9,915–921.

Doering,C.B.,2008.Retroviralmodificationofmesenchymalstemcellsforgene therapyofHemophiliaA.MethodsMol.Biol.433(1),203–212.

Eckardt,K.U.,2001.After15yearsofsuccess– perspectivesoferythropoietinther- apy.Nephrol.Dial.Transplant.16,1745–1749.

Eliopoulos,N.,Al-Khaldi,A.,Crosato,M.,Lachapelle,K.,Gallipeau,J.,2003.Aneovas- cularizedorganoidderivedfromretrovirallyengineeredbonemarrowstroma leadstoprolongedinvivosystemicdeliveryoferythropoietininnonmyeloab- lated,immunocompetentmice.Nature10(6),478–489.

Eliopoulos,N.,Gagnon,R.F.,Francois,M.,Gallipeau,J.,2006.Erythropoietindelivery bygeneticallyengineeredbonemarrowstromalcellsforcorrectionofanemia inmicewithchronicrenalfailure.J.Am.Soc.Nephrol.17,1576–1584.

Eliopoulos,N.,Stagg,J.,Lejeune,L.,Pommey,S.,Galipeau,J.,2005.Allogeneicmar- rowstromalcellsareimmunerejectedbyMHCclassI-andclassII-mismatched recipientmice.GeneTher.106,4057–4065.

Elliott,S.,Lorenzini,T.,Asher,S.,Aoki,K.,Brankow,D.,Buck,L.,Busse,L.,Chang, D.,Fuller,J.,Grant,J.,Hernday,N.,Hokum,M.,Hu,S.,Knudten,A.,Levin,N., Komorowski,R.,Martin,F.,Navarro,R.,Osslund,T.,Rogers,G.,Rogers,N.,Trail, G.,Egrie,J.,2003.Enhancementoftherapeuticproteininvivoactivitiesthrough glycoengineering.Nat.Biotechnol.21,414–421.

Esneault,E.,Pacary,E.,Eddi,D.,Freret,T.,Tixier,E.,Toutain,J.,Touzani,O.,Schumann- Bard,P.,Petit,E.,Roussel,S.,Bernaudin,M.,2008.Combinedtherapeuticstrategy usingerythropoietinandmesenchymalstemcellspotentiatesneurogenesis aftertransientfocalcerebralischemiainrats.J.Cereb.BloodFlowMetab.28, 1552–1563.

Fisher,J.W.,2003.Erythropoietin:physiologyandpharmacologyupdate.Exp.Biol.

Med.228(1),1–14.

Ghezzi,P.,Brines,M.,2004.Erythropoietinasanantiapoptotic,tissue-protective cytokine.CellDeathDiffer.11,S37–S44.

Gothelf, A., Hojman, P., Gehl, J., 2010. Therapeutic levels of erythropoietin (EPO)achievedafter geneelectrotransfer toskin inmice.Gene Ther. 17, 1077–1084.

Grinnemo,K.H.,Månsson,A.,Dellgren,G.,Klingberg,D.,Wardell,E.,Drvota,V.,Tam- mik,C.,Holgersson,J.,Ringdén,O.,Sylvén,C.,LeBlanc,K.,2004.Xenoreactivity andengraftmentofhumanmesenchymalstemcellstransplantedintoinfarcted ratmyocardium.J.Thorac.Cardiovasc.Surg.127,1293–1300.

Hamamori,Y.,Samal,B.,Tian,J.,Kedes,L.,1995.Myoblasttransferofhumanery- thropoietingeneinamousemodelofrenalfailure.J.Clin.Invest.95,1808–1813.

Herweijer,H.,Wolff,J.A.,2003.Progressandprospects:nakedDNAgenetransfer andtherapy.GeneTher.10,453–458.

Hooduijn,M.J.,Gorjup,E.,Genever,P.G.,2006.Comparativecharacterizationofhair follicledermalstemcellsandbonemarrowmesenchymalstemcells.StemCells Dev.15(1),49–60.

Jelkmann,W.,2008.Developmentsinthetherapeuticuseoferythropoiesisstimu- latingagents.Br.J.Haematol.141,287–297.

Jelkmann,W.,1992.Erythropoietin:structure,controlofproduction,andfunction.

Physiol.Rev.71(s),449–489.

Kang,S.G.,Shinojima,N.,Hossain,A.,Gumin,J.,Yong,R.L.,Colman,H.,Marini,F., Andreef,M.,Lang,F.F.,2010.Erythropoietinpromoteshairshaftgrowthincul- turedhumanhairfolliclesandmodulateshairgrowthinmice.J.Dermatol.Sci.

59(2),86–90.

Karp,J.M.,Teo,G.S.L.,2009.Mesenchymalstemcellhoming:thedevilisinthedetails.

CellStemCell4(3),206–216.

Kato,M.,Kamiyama,H.,Okazaki,A.,Kumaki,K.,Sugiyama,Y.,1997.Mechanismfor thenonlinearpharmacokineticsoferythropoietininrats.JPET283,520–527.

Kim,M.,O’Callaghan,P.M.,Droms,K.A.,James,D.C.,2011.Amechanisticunder- standingofproductioninstabilityinCHOcelllinesexpressingrecombinant monoclonalantibodies.Biotechnol.Bioeng.108(10),2434–2446.

Kong,H.,Liu,J.,Riddle,K.,Matsumoto,T.,Leach,K.,Mooney,D.J.,2005.Non-viral genedeliveryregulatedbystiffnessofcelladhesionsubstrates.Nat.Mater.4, 640–644.

Krantz,S.B.,1991.Erythropoietin.Blood77(3),419–434.

Lieutaud,T.,Andrews,P.J.,Rhodes,J.K.,Williamson,R.,2008.Characterizationof thepharmacokineticsofhumanrecombinanterythropoietininbloodandbrain whenadministeredimmediatelyafterlateralfluidpercussionbraininjuryand itspharmacodynamiceffectsonIL-1betaandMIP-2inrats.J.Neurotrauma25 (10),1179–1185.

Lifshitz,L.,Prutchi-Sagiv,S.,Avneon,M.,Gassmann,M.,Mittelman,M.,Neumann,D., 2009.Non-erythroidactivitiesoferythropoietin:functionaleffectsonmurine dendriticcells.Mol.Immunol.46,713–721.

Lifshitz,L.,Tabak,G.,Gassmann,M.,Mittelman,M.,Neumann,D.,2010.Macrophages asnoveltargetcellsforerythropoietin.Haematologica95(1),1823–1831.

Liu, L.N.,Hendricks,K., Buyaner,D., Boehnlein,E., Junker, U.,Deans,R., Van- guri,P.,Mosca,J.D.,2000.Systemicerythropoietinexpressionfromhuman

mesenchymalstemcellsdependonbothcelladherenceanddeliveryroute.In:

Abstractinthe6thAnnualISHAGE,SanDiego,USA,15–18June.

Loo,M.,Beguin,Y.,1999.Theeffectofrecombinanthumanerythropoietinonplatelet countsisstronglymodulatedbytheadequacyofironsupply.Blood93(10), 3286–3293.

McGinley,L.,McMahon,J.,Strappe,P.,Barry,F.,Murphy,M.,O’Toole,D.,O’Brien,T., 2011.Lentiviralvectormediatedmodificationofmesenchymalstemcellsand enhancedsurvivalinaninvitromodelofischaemia.StemCellRes.Ther.2,12.

Meer,P.V.D.,Lipsic,E.,Henning,R.H.,Boddeus,K.,vanderVelden,J.,Voors,A.A.,van Veldhuisen,D.J.,vanGilst,W.H.,Schoemaker,R.G.,2005.Erythropoietininduces neovascularizationandimprovescardiacfunctioninratswithheartfailureafter myocardialinfarction.J.Am.Coll.Cardiol.46(1),125–133.

Mok,P.L.,Cheong,S.K.,Leong,C.F.,Chua,K.H.,Ainoon,O.,2012.Extendedandstable geneexpressionvianucleofectionofMIDGEconstructintoadulthumanmarrow mesenchymalstromalcells.Cytotechnology64,203–216.

Mok,P.L.,Cheong,S.K.,Leong,C.F.,Othman,A.,2008.Invitroexpressionoferythro- poietinbytransfectedhumanmesenchymalstromalcells.Cytotherapy10(2), 116–124.

Mok,P.L.,Leong,C.F.,Cheong,S.K.,2003.Isolationandidentificationofmesenchymal stemcellsfromhumanbonemarrow.Malays.J.Pathol.25(2),121–127.

Naldi,A.M.,Gassmann,M.,Bodmer,D.,2009.ErythropoietinbutnotVEGFhasa protectiveeffectonauditoryhaircellsintheinnerear.Cell.Mol.LifeSci.66, 3595–3599.

Parekkadan,B.,Milwid,J.M.,2010.Mesenchymalstemcellsastherapeutics.Annu.

Rev.Biomed.Eng.12,87–117.

Schakowski,F.,Gorschluter,M.,Junghans,C.,Schroff,M.,Buttgereit,P.,Ziske,C., Schöttker,B.,König-Merediz,S.A.,Sauerbruch,T.,Wittig,B.,Schmidt-Wolf,I.G., 2001.Anovelminimal-sizevector(MIDGE)improvestransgeneexpressionin coloncarcinomacellsandavoidstransfectionsofundesiredDNA.Mol.Ther.3, 793–800.

Singer,N.G.,Caplan,A.I.,2011.Mesenchymalstemcells:mechanismsofinflamma- tion.Annu.Rev.Pathol.Mech.Dis.6,457–478.

Stenn,K.,Parimoo,S.,Zheng,Y.,Barrows,T.,Washenik,K.,2007.Bioengineeringthe hairfollicle.Organogenesis3(1),6–13.

Terrovitis,J.V.,Smith,R.R.,Marbán,E.,2010.Assessmentandoptimizationofcell engraftmentaftertransplantationintotheheart.Circ.Res.106,479–494.

Toma,C.,Pittenger,M.F.,Cahill,K.S.,Byrne,B.J.,Kessler,P.D.,2002.Humanmes- enchymalstemcellsdifferentiatetoacardiomyocytephenotypeintheadult murineheart.Circulation105(1),93–98.

Tomita,M.,Adachi,Y.,Yamada,H.,Takahashi,K.,Kiuchi,K.,Oyaizy,H.,Ikekuburo,K., Kaneda,H.,Matsumura,M.,Ikehara,S.,2002.Bonemarrow-derivedstemcells candifferentiateintoretinalcellsininjuredratretina.StemCells20,279–283.

Tsakiris,D.,2000.Morbidityandmortalityreductionassociatedwiththeuseof erythropoietin.Nephron85(Suppl.1),2–8.

Tse,W.T.,Pendleton,J.D.,Beyer,W.M.,Egalka,M.C.,Guinan,E.C.,2003.Suppression ofallogeneicT-cellproliferationbyhumanmarrowstromalcells:implications intransplantation.Transplantation75(3),389–397.

Wang,G.,Liu,L.N.,Lee,K.,Buyaner,D.,Hendricks,K.,Kalfoglou,C.S.,Bohnlein, E.,Junker,U.,Mosca,J.D.,2009.Comparisonofdrugandcell-baseddelivery- engineeringadultmesenchymalstemcellstodeliverhumanerythropoietin.

GeneTher.Mol.Biol.13,321–330.

Wong,C.Y.,Cheong,S.K.,Mok,P.L.,Leong,C.F.,2008.Differentiationofhuman mesenchymalstemcellsintomesangialcellsinpost-glomerularinjurymurine model.Pathology40(1),52–57.

Yoo,B.Y.,Shin,Y.H.,Yoon,H.H.,Seo,Y.K.,Park,J.K.,2010.Applicationofmesenchy- malstemcellsderivedfrombonemarrowandumbilicalcordinhumanhair multiplication.J.Dermatol.Sci.60,74–83.

Zhang,D.,Zhang,F.,Zhang,Y.,Gao,X.,Li,C.,Yang,N.,Cao,K.,2007.Combining erythropoietininfusionwithintramyocardialdeliveryofbonemarrowcellsis moreeffectiveforcardiacrepair.Transpl.Int.20(2),174–183.