RESEARCHLETTER
ingoldchloridefor2minandfixedwith5%aqueoussodiumthiosulphate.A15sincubationinFastgreenwasusedforcounterstaining.Afterabriefrinseindistilledwater,dehydrationwasquicklyperformedin70%,95%andabsoluteethylalcohol;sectionswereclearedwithxyleneandmountedwithDPX.
ForVanGieson’sstainingofcollagen,nucleiwerestainedwithcelestinebluefor2min,brieflyrinsedindistilledwater,incubatedwithHarrishaematoxylinsolutionfor2minandwashedunderrunningtapwaterfor5min.Curtisstain(saturatedaqueouspicricacid,1%PonceauSandglacialaceticacidmix)wasperformedfor5min,untilcollagenwaspink.Afterabriefrinseindistilledwater,dehydrationwasquicklyperformedin70%,95%andabsoluteethylalcohol;sectionswereclearedwithxyleneandmountedwithDPX.
Immunofluorescencestainingofcryostatsectionswasperformedaspreviouslydescribed8.Forwholemountstainingofthecalvaria,allincubationtimeswereextended.Theantibodiesusedwereanti-TH(RabbitpAb,Millipore)andanti-GFAP(RabbitpAb,Dako).Confocalimageswereacquiredwithalaserscanningconfocal(ZeissLSM700).Atleast3differentsectionswereusedforquantificationusingImageJsoftware.
Fornestin/CD34immunohistochemistryofhumanBMsamples,12controlBMbiopsies(2fromhealthydonors,2frompatientswithreactiveperipheralleuko-cytosis,and8performedforlymphomastaging,butunaffectedbylymphoma)and28MPN(13ofthemwereJAK2(V617F)1)werestainedfornestinapplyingthemonoclonalantibody10C2fromAbDSerotec(OBT1610)atadilutionof1:50usinganautomatedimmunostainer(Benchmark,Ventana/Roche).Antigenretrievalwasachievedbycellconditioning(CC1fromVentana/Roche)treatmentfor60min.Incubationfor60min,signalamplificationandvisualization(amplifierandchromo-genultraviewuniversaldiaminobenzidinefromVentana/Roche)followed.Fornestin/CD34doublestainings,themonoclonalantibodyQBEnd/10(Ventana/Roche790-2927)wasusedafternestindetectionandvisualizedusinganalternativechromo-genicdetectionkit(basicaminoethylcarbazolefromVentana/Roche).Thenumberofnestin1perivascularniches(eithersinglecellsorclustersofupto3cells)wasblindlyscoredintheBMsamples.Anaverageareaof7.2mm2wasevaluatedforeachcase.ForTHimmunofluorescenceofhumanBMsamples,2controland16MPN(5essentialthrombocythaemia,4chronicmyeloidleukaemia,and7primarymyelofi-brosis)BMbiopsieswereused.SectionsweredeparaffinedandantigenretrievalwasperformedwithEDTApH9.After30minofpermeabilisationinmethanol,immu-nofluorescencestainingandquantificationwasperformedasdescribedabove.ELISA.Bio-PlexProMouseTh17cytokinePanelA6-plex(M60-00007NY,Bio-Rad)wasperformedfollowingthemanufacturer’sprotocol.Cxcl12proteinlevelsweremeasuredbyconventionalELISA.Briefly,96-wellplateswerecoatedovernightat4uCwith2mgml21ofmonoclonalanti-humanandmouseCXCL12/SDF-1antibody(MAB350,R&DSystems).Afterblocking,bonemarrowextracellularfluidswereincubatedfor2hatroomtemperature,followedbyadditionofbiotinylatedanti-humanandmouseCXCL12/SDF-1antibody(BAF310,RD).Streptavidin-horseradishperoxidaseconjugate(RPN1231V,Dako)wasusedforreportingsig-nal,andreactionwasstoppedwithhorseradishperoxidasesubstrate(TMB,ES001-500ML,Chemicon,Millipore).Stardardcurvewasperformedwithrecombinanthuman,feline,rhesusmacaqueSDF-1alpha(350-NS,R&D).
RNAisolationandqPCR.RNAisolationwasperformedusingtheDynabeadsmRNADIRECTMicroKit(Invitrogen).ReversetranscriptionwasperformedusingtheReverseTranscriptionSystem(Promega),followingthemanufacturer’srecommendations.Theexpressionlevelofeachgenewasdeterminedbyusingtherelativestandardcurvemethod.Briefly,astandardcurvewasperformedbydoingserialdilutionsofamouseorhumanreferencetotalRNA(Clontech).Theexpress-ionlevelofeachgenewascalculatedbyinterpolationfromthestandardcurve.AllvalueswerenormalizedwithGapdhasendogenouscontrol.Thefollowingprimerswereused:Human:
GFAP-Fw:CCGACAGCAGGTCCATGTGGFAP-Rv:GTTGCTGGACGCCTTGC
NESTIN-Fw:CAACAGCGACGGAGGTCTCNESTIN-Rv:GCCTCTACGCTCTCTTCTTTGAGAPDH-Fw:GCATGGCCTTCCGTGTTC
GAPDH-Rv:CCTGCTTCACCACCTTCTTGATMouse:
Adrb2-Fw:AGCAATAGCAACGGCAGAACAdrb2-Rv:TTCACAAAGCCTTCCATGCCAdrb3-Fw:AAACTGGTTGCGAACTGTGGAdrb3-Rv:TAACGCAAAGGGTTGGTGAC
Angpt1-Fw:CTCGTCAGACATTCATCATCCAGAngpt1-Rv:CACCTTCTTTAGTGCAAAGGCTCasp1-Fw:TTGGAGCTCAAGTTGACCTCAGCasp1-Rv:TGTCAGAAGTCTTGTGCTCTGG
Cxcl12-Fw:TGCATCAGTGACGGTAAACCACxcl12-Rv:TTCTTCAGCCGTGCAACAATCGapdh-Fw:GCATGGCCTTCCGTGTTC
Gapdh-Rv:CCTGCTTCACCACCTTCTTGATGfap-Fw:CGGAGACGCATCACCTCTGGfap-Rv:AGGGAGTGGAGGAGTCATTCGIl1b-Fw:GAAATGCCACCTTTTGACAGTGIl1b-Rv:TGGATGCTCTCATCAGGACAGIl1r-Fw:GTGCTACTGGGGCTCATTTGTIl1r-Rv:GGAGTAAGAGGACACTTGCGAATIl1rn-Fw:GAGAAACAACCAGCTCATTGCIl1rn-Rv:GGATGCCCAAGAACACACTATGKitl-Fw:CCCTGAAGACTCGGGCCTA
Kitl-Rv:CAATTACAAGCGAAATGAGAGCCLifr-Fw:TACGTCGGCAGACTCGATATTLifr-Rv:TGGGCGTATCTCTCTCTCCTTMbp-Fw:AATCGGCTCACAAGGGATTCAMbp-Rv:TCCTCCCAGCTTAAAGATTTTGGMobp-Fw:CCAGGCTCTCCAAGAACCAGMobp-Rv:GGTCCACGATCTCACGCTTNestin-Fw:CCCTGAAGTCGAGGAGCTGNestin-Rv:CTGCTGCACCTCTAAGCGAPkp4-Fw:GAACCTGTCATACCGGCTGGPkp4-Rv:TTCCGAGTCTTTGCTGGGAGAPlekhb1-Fw:CTGGAAGCGGAATTGGTTCGPlekhb1-Rv:TGCCGTCTCGTCATGGTAGTAPlp1-Fw:TGAGCGCAACGGTAACAGGPlp1-Rv:TTCCCAAACAATGACACACCCS100b-Fw:TGGTTGCCCTCATTGATGTCTS100b-Rv:CCCATCCCCATCTTCGTCC
Slit2-Fw:CCATGTAAAAATGATGGCACCTGSlit2-Rv:ATCACAGTCCTGACCCTTGAA
RNaseqandmicroarray.Fornext-generationsequencing,totalRNAwasisolatedusingtheArcturusPicopureRNAisolationkit(LifeTechnologies)fromsmallnumbersofFACSsortedCD452CD312Ter1192GFP1cells,obtainedfromtheBMofNes-gfp;Mx1-cre;JAK2(V617F)miceandcontrollittermates6weeksafterpIpCtreatment.Eachsamplewasapoolfrom3differentanimals.RNAwasamplifiedandpreparedforRNA-SequsingtheOvationRNA-SeqSystemv2(NuGEN)followingthemanufacturer’srecommendations.TheRNAsequencinglibrarywaspreparedwiththeTruSeqRNASamplePreparationv2Kit(Illumina,SanDiego,CA)toconstructindex-taggedcDNA.Thequality,quantityandthesizedistributionoftheIlluminalibrariesweredeterminedusingtheDNA-1000Kit(AgilentBioanalyzer).LibrariesweresequencedontheGenomeAnalyzerIIx(Illumina)followingthestandardRNAsequencingprotocolwiththeTruSeqSBSKitv5.Fastqfilescontainingreadsforeachlibrarywereextractedanddemulti-plexedusingCasavav1.8.2pipeline.Sequencingadaptorcontaminationswereremovedfromreadsusingcutadaptsoftwaretool(MIT)andtheresultingreadsweremappedandquantifiedonthetranscriptome(NCBIM37Ensemblgene-build65)usingRSEMv1.1734.Expressiondatawascomparedbetweenbothsam-plesbytheanalysisofindividualselectedgenesfordifferentialexpression,andthroughgene-setenrichmentanalyses(GSEA)todetectcoordinatedchangesinsetsofgenesrepresentingpathways,functionalsignaturesortranscriptionfactortargets.GSEAwereperformedasdescribed(http://www.broadinstitute.org/gsea/index.jsp),usingaweightedstatistic,fold-changeranking,1,000gene-setpermu-tationsandseveralgenesetdatabasesfoundintheMolecularSignaturesDatabase.Formicroarrayanalyses,totalRNAwasisolatedaspreviouslydescribedfromBMCD452CD312Ter1192Nes-GFP1cellsobtainedfromNes-gfpmice10weeksaftertransplantationwithMx1-cre;JAK2(V617F)(n53)orcontrolcells(n51).RNAwasamplifiedusingtheNuGenOvationsystemandhybridizedtotheAffymetrixMoGene1.0STarray.DatanormalizationwasperformedusingtheRobustMulti-arrayAverage(RMA)algorithm.Toperformprincipalcomponentanalysis(PCA)comparisonwithpreviouslypublisheddata,GEOdatasetsweredownloadedandpre-processedusingtheGEOqueryBioconductorpackage35.Normalizeddatasetswereadjustedtothesameintensityrange,andbatcheffectcorrectionwasper-formedusingComBat36.
Statisticalanalyses.StatisticalanalysesandgraphicswerecarriedoutwithGraph-PadPrism5softwareandMicrosoftExcel.Unlessspecified,datasetswerecom-paredbyunpairedtwo-tailedt-tests;Pvalueslessthan0.05wereconsideredstatisticallysignificant.
31.Frenette,P.S.,Subbarao,S.,Mazo,I.B.,vonAndrian,U.H.&Wagner,D.D.Endothelialselectinsandvascularcelladhesionmolecule-1promotehematopoieticprogenitorhomingtobonemarrow.Proc.NatlAcad.Sci.USA95,14423–14428(1998).?2014Macmillan Publishers Limited. All rights reservedLETTERRESEARCH
32.33.34.Stemple,D.L.&Anderson,D.J.Isolationofastemcellforneuronsandgliafromthemammalianneuralcrest.Cell71,973–985(1992).Pajtler,K.etal.Productionofchickembryoextractforthecultivationofmurineneuralcreststemcells.J.Vis.Exp.45,e2380(2010).Biernaskie,J.A.,McKenzie,I.A.,Toma,J.G.&Miller,F.D.Isolationofskin-derivedprecursors(SKPs)anddifferentiationandenrichmentoftheirSchwanncellprogeny.NatureProtocols1,2803–2812(2007).Davis,S.&Meltzer,P.S.GEOquery:abridgebetweentheGeneExpressionOmnibus(GEO)andBioConductor.Bioinformatics23,1846–1847(2007).Johnson,W.E.,Li,C.&Rabinovic,A.AdjustingbatcheffectsinmicroarrayexpressiondatausingempiricalBayesmethods.Biostatistics8,118–127(2007).35.36.?2014Macmillan Publishers Limited. All rights reservedRESEARCHLETTER
ExtendedDataFigure1|BMMSCreductionincompoundmutantmiceandrecipientsofmutanthaematopoieticcellsisnotduetocell
differentiation.a,b,BloodcountsofNes-gfp;Mx1-cre;JAK2(V617F)andcontrolmice4-8weeksafterpIpCtreatment(a)andNes-gfpmice10-12weeksaftertransplantationwithBMcellsfromMPNandcontrolmice(b).
Notethesimilarerythrocytosis,neutrophiliaandthrombocytosisincompoundmutantmiceandNes-gfpmicetransplantedwithmutantcells.Eachdot
representsamouse.c,GFPfluorescenceinskullBMofNes-gfpmice6–8weeksaftertransplantationwithMPNandcontrolBMcells(n58;scalebar,100mm).d,FrequencyofCD452CD312Ter1192Nes-GFP1BMcellsinMPN(n59)andcontrol(n57)mice6–8weeksafterpIpCinduction.e,Giemsa-stainedfibroblasticcolony-formingunits(CFU-F)fromimmunomagnetically-enrichedBMCD452Ter1192cells30weeksaftertransplantationwithMPN
andcontrolBMcells.f,FACSanalysesofBMCD452CD312Ter1192cellsfromMPNandcontrolmice8weeksafterpIpCtreatment(n53).ThespecifiedMSCmarkerswereused.Mean6s.e.m.g,BMreticulinstaining;MPNmiceshowedincipientfibrosis(arrow)6weeksafterpIpC
treatment(magnification,3200).h,i,Lineage-tracingstudiesofBMnestin1cellsinMPN.h,Experimentaldesign.Nes-creERT2;RCE:loxPmiceweretransplantedwithMPNandcontrolBMcellsandfedwithtamoxifendiet.Diseasedevelopmentwasmonitoredover28weeks.i,BloodcountsshowingprogressiveneutrophiliaandthrombocytosisinrecipientsofMPNBMcells(n53).j,FemoralhaematoxylinandeosinstainingsshowingabnormalboneformationinrecipientsofMPNBMcells(magnification,3100).*P,0.05,**P,0.01,***P,0.001(unpairedtwo-tailedttest).
?2014Macmillan Publishers Limited. All rights reservedLETTERRESEARCH
ExtendedDataFigure2|TreatmentwiththeJAK1/2inhibitorruxolitinibreduceshaematopoieticcellexpansioninMPNbutitdoesnotrescueBMMSCs.Nes-gfpmiceweretransplantedwithBMcellsfrompIpC-inducedMx1-cre;JAK2(V617F)miceandtreatedwithruxolitinib(n54)orvehicle(n56)over8weeks,starting2weeksaftertransplantation.a,Peripheralbloodcounts.b,Frequencyofimmunophenotypically-definedBM
CD452CD312Ter1192cells.Mean6s.e.m.*P,0.05(unpairedtwo-tailedttest).
?2014Macmillan Publishers Limited. All rights reservedRESEARCHLETTER
ExtendedDataFigure3|Depletionofnestin1cellsortheirCxcl12
productionacceleratesMPN.a,Experimentaldesignforinvivonestin1celldepletion.Nes-creERT2;iDTAandcontroliDTAmiceweretransplantedwithBMcellsfrompIpC-treatedMx1-cre;JAK2(V617F)andcontrolmiceandtreatedwithtamoxifen(n53).b–e,UnlikeSchwanncells,MSCsarereducedinNes-creERT2;iDTABM.b,c,FrequencyofBMCD452CD312Ter1192CD901cells(b)andfibroblasticcolony-formingunits(CFU-F,c)inBM
CD452Ter1192cellsfromNes-creERT2;iDTAandcontrolmiceafter4-weektamoxifentreatment.d,Expressionofglialfibrillaryacidicprotein(Gfap)mRNA(normalizedtoGapdh)intheBMofmiceina.e,QuantificationofGfapimmunostainingoffemoralBMfrommiceinb,c.f–k,EarlymutantHSCexpansionandmobilisationcorrelateswithreducedCxcl12expressioninBMnestin1cells.f,g,BMlin2sca-11c-kit1(f)CD342Flt32long-term(LT)andCD341Flt32short-term(ST)HSCs,andCD342Flt32multipotentprogenitors
(MPP)and(g)frequencyofhaematopoieticcolony-formingunitsin
culture(CFU-C)fromBMnucleatedcells,spleenorperipheralbloodfromNes-Gfp;Mx1-cre;JAK2(V617F)(n56)andcontrol(n511)mice6weeksafterpIpCtreatment.h,Appearance(inset;scalebar,1cm)andhaematoxylinandeosinstainingsofspleens(magnification,3100).i–k,Cxcl12protein(i)andmRNAlevels(j,k)inBMextracellularfluid(i),nucleatedcells(j)andstromalNes-GFP1cells(k)ofMx1-cre;JAK2(V617F)andcontrolmice6weeksafterpIpCtreatment(i,j)orNes-gfpmice10weeksaftertransplantationwithmutantBM(n54)orcontrolcells(n57)(k).l,CirculatingplateletsinNes-creERT2;Cxcl12fl/fl(n511)andcontrolCxcl12fl/fllittermates(n514)18weeksaftertransplantationwithBMcellsfromMPNmiceand12weeksaftertamoxifentreatment.Mean6s.e.m.*P,0.05,**P,0.01(unpairedtwo-tailedttest).
?2014Macmillan Publishers Limited. All rights reserved