Citation

Material Information

Title:
Expansion of human Tregs from cryopreserved umbilical cord blood for GMP-compliant autologous adoptive cell transfer therapy
Series Title:
Molecular Therapy - Methods & Clinical Development
Creator:
Seay, Howard
Publisher:
CellPress
Publication Date:
Language:
English
Physical Description:
Journal Article

Notes

Abstract:
Umbilical cord blood is a traditional and convenient source of cells for hematopoietic stem cell transplantation. Thymic regulatory T cells (Tregs) are also present in cord blood, and there is growing interest in the use of autologous Tregs to provide a low-risk, fully human leukocyte antigen (HLA)-matched cell product for treating autoimmune diseases, such as type 1 diabetes. Here, we describe a good manufacturing practice (GMP)-compatible Treg expansion protocol using fluorescence-activated cell sorting, resulting in a mean 2,092-fold expansion of Tregs over a 16-day culture for a median yield of 1.26 × 109 Tregs from single-donor cryopreserved units. The resulting Tregs passed prior clinical trial release criteria for Treg purity and sterility, including additional rigorous assessments of FOXP3 and Helios expression and epigenetic analysis of the FOXP3 Treg-specific demethylated region (TSDR). Compared with expanded adult peripheral blood Tregs, expanded cord blood Tregs remained more naive, as assessed by continued expression of CD45RA, produced reduced IFN-γ following activation, and effectively inhibited responder T cell proliferation. Immunosequencing of the T cell receptor revealed a remarkably diverse receptor repertoire within cord blood Tregs that was maintained following in vitro expansion. These data support the feasibility of generating GMP-compliant Tregs from cord blood for adoptive cell transfer therapies and highlight potential advantages in terms of safety, phenotypic stability, autoantigen specificity, and tissue distribution.
Acquisition:
Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Howard Seay.

Record Information

Source Institution:
University of Florida Institutional Repository
Holding Location:
University of Florida
Rights Management:
All rights reserved by the submitter.

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

OriginalArticleExpansionofHumanTregsfromCryopreserved UmbilicalCordBloodforGMP-Compliant AutologousAdoptiveCellTransferTherapyHowardR.Seay,1AmyL.Putnam,4JuditCserny,1AmandaL.Posgai,1EmmaH.Rosenau,2JohnR.Wingard,2KateF.Girard,5MoreyKraus,5AngelaP.Lares,4HeatherL.Brown,6KatherineS.Brown,6KristiT.Balavage,1LeeanaD.Peters,1AshleyN.Bushdorf,1MarkA.Atkinson,1 3JeffreyA.Bluestone,4MichaelJ.Haller,3andToddM.Brusko11DepartmentofPathology,ImmunologyandLaboratoryMedicine,UniversityofFloridaCollegeofMedicine,Gainesville,FL32610,USA;2DivisionofHematologyand Oncology,DepartmentofMedicine,UniversityofFloridaCollegeofMedicine,Gainesville,FL32610,USA;3DepartmentofPediatrics,UniversityofFloridaCollegeof Medicine,Gainesville,FL32610,USA;4DiabetesCenterandDepartmentofMedicine,UniversityofCalifornia,SanFrancisco,SanFrancisco,CA94143,USA;5ViaCord,LLC,Waltham,MA02451,USA;6CbrSystems,Inc.,SanBruno,CA94066,USAUmbilicalcordbloodisatraditionalandconvenientsourceof cellsforhematopoieticstemcelltransplantation.ThymicregulatoryTcells(Tregs)arealsopresentincordblood,andthereis growinginterestintheuseofautologousTregstoprovidealowrisk,fullyhumanleukocyteantigen(HLA)-matchedcellproductfortreatingautoimmunediseases,suchastype1diabetes. Here,wedescribeagoodmanufacturingpractice(GMP)compatibleTregexpansionprotocolusing uorescenceactivatedcellsorting,resultinginamean2,092-foldexpansion ofTregsovera16-daycultureforamedianyieldof1.26 109Tregsfromsingle-donorcryopreservedunits.Theresulting TregspassedpriorclinicaltrialreleasecriteriaforTregpurity andsterility,includingadditionalrigorousassessmentsof FOXP3andHeliosexpressionandepigeneticanalysisofthe FOXP3 Treg-speci cdemethylatedregion(TSDR).Compared withexpandedadultperipheralbloodTregs,expandedcord bloodTregsremainedmorenaive,asassessedbycontinued expressionofCD45RA,producedreducedIFNg following activation,andeffectivelyinhibitedresponderTcellproliferation.ImmunosequencingoftheTcellreceptorrevealeda remarkablydiversereceptorrepertoirewithincordbloodTregs thatwasmaintainedfollowinginvitroexpansion.Thesedata supportthefeasibilityofgeneratingGMP-compliantTregs fromcordbloodforadoptivecelltransfertherapiesandhighlightpotentialadvantagesintermsofsafety,phenotypicstability,autoantigenspeci city,andtissuedistribution.INTRODUCTIONTheinabilityofahosttocontrolimmunein ammationisakeypathologicalfeatureforthedevelopmentofautoimmunediseases, includingtype1diabetes(T1D).1,2Thereisgrowinginterestinthe useofadoptivecelltransfer(ACT)therapiesinvolvingregulatory Tcells(Tregs)asameanstocontrolhostresponses,giventheiressentialroleinlimitingbothinnateandadaptiveimmuneactivation.3AkeycomponentofthetherapeuticactivityofTregsresidesintheir capacitytoexerttheirfunction(s)as “ livingdrugs. ”4,5Speci cally, Tregsfunctiontomaintaindominantperipheraltoleranceby consuminggrowthfactorsofpathogenicTcellsandexertingtheirtissue-directedsuppressiveactivitiesthroughanarrayofanti-in ammatorybiochemicalpathways,expressionofco-inhibitoryreceptors,and productionofimmunoregulatorycytokines.6TheseseeminglyredundantmechanismsallowTregstotracktovarioustissuestosuppress hostin ammationinatissue-andcontext-dependentmanner. MorerecentstudieshavealsosuggestedthatTregsmayplayan importantroleintissuerepairfollowingdamage,7 – 9extendingtheir potentialbene tstothe eldofregenerativemedicine. T1Disatissue-speci cautoimmunediseasecharacterizedbythetargetingofinsulin-producing b cellsinthepancreaticislets,primarily throughtheactivityofautoreactiveTeffectorcells(Teffs).Anumber ofimmunotherapeuticinterventionshavebeenconductedinpatients withT1D,withtheultimategoalofalteringtheeffectivebalanceof cellstorestoreregulationbyTregs.Atleasttwomajorapproaches havebeentakeninthisregard.First,agentshavebeenusedtotarget andselectivelydepleteautoreactiveTeffs,withnotabletherapeutic responsesobservedfollowingtreatmentwithalefacept(afusionproteindisruptingtheCD2:LFA3interaction)10andtepilizumab(antiCD3)11asmonotherapiesandanti-thymocyteglobulin(ATG)for leukocytedepletionincombinationwiththestemcellmobilizing agentgranulocytecolony-stimulatingfactor(G-CSF).12Oneof themostpotentinterventionstodateemployedthecombination therapyofautologousnonmyeloablativestemcelltransplantplus Received11October2016;accepted16December2016; http://dx.doi.org/10.1016/j.omtm.2016.12.003 Correspondence: ToddM.Brusko,DepartmentofPathology,Immunologyand LaboratoryMedicine,CollegeofMedicine,UniversityofFlorida,1275Center Drive,BiomedicalSciencesBuildingJ-589,Box100275,Gainesville,FL32610,USA. E-mail: tbrusko@u .edu 178MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017 2016TheAuthor(s). ThisisanopenaccessarticleundertheCCBY-NC-NDlicense( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).

PAGE 2

preconditioningwithcyclophosphamideandG-CSFalongwithhigh dosesofATG13 – 16orthechemotherapeuticagent udarabine.17T1D patientswhoreceivedthesetherapiesexperiencedasigni cantremission,somefor4yearsorlonger,andwereabletostopexogenousinsulin.However,signi cantsideeffectslimittheuseofsuchaggressive protocols,particularlyinthesettingofpediatricpatients.Thesecond majorapproachinvolvestheadministrationoftherapiestobolsterthe regulatoryarmoftheTcellcompartment.Inthisregard,effortshave beenmadetotreatsubjectswithT1Dwithexogenouslow-doseinterleukin-2(IL-2),aselectiveTreggrowthfactor.18Alternatively,autologouspolyclonalTregscanbeisolatedfromperipheralbloodandcan alsobebolsteredbyexvivoexpansionpriortoACTtherapy.Initial phaseIdoseescalationtrialsconductedinsubjectswithgraftversus hostdisease(GvHD)19 – 21andsubjectswithrecent-onsetT1D22,23havedemonstratedbothinitialpersistenceandsafety.Althoughnot poweredtoassessef cacy,theseinitialT1Dtrialsdidobservesubjects whomaintainedrelativelystableC-peptideresponseswhenassessed bymixed-mealtolerancetests.Moreover,preliminarydatafromthese effortssuggestedthatTregpersistencecorrelatedwithpreservationof C-peptideinanumberofsubjects. Umbilicalcordblood(CB)hasbeenusedasasourceofhematopoietic stemcells(HSCs)fortransplantationsince1988.24,25HSCtherapies canrestorefunctioninpatientssufferingfrommalignancies,bone marrowfailuredisorders,inbornerrorsofmetabolism,andimmunologicaldisorders.26Harvestedcordbloodunits(CBUs)canalsobe usedasasourceofnaiveTregsthathavebeenestablishedasapotential treatmenttocontrolxeno-GvHD,27andmultiplephaseIclinicaltrials evaluatingthesafetyofTregACTtherapyinGvHDhavereportedthat exvivo-expanded,partiallyhumanleukocyteantigen(HLA)-matched Tregsfromnon-autologousCBUsarewelltolerated.28,29Wehavepreviouslydemonstratedthattheinfusionofautologous,cryopreserved CB(cryoCB)alone30orincombinationwithoraldocosahexaenoic acidandvitaminDissafeinsubjectswithT1D.31However,thisunmanipulatedCBproductcontainsonlyaminorfractionofTregs, whichmayexplaintheinabilityofthesetherapiestopreserveC-peptideproduction.32Wehypothesizedthatapurepopulationofnaive TregswithhighproliferativecapacitycouldbeisolatedfromcryoCB units(cryoCBUs)andexpandedtotherapeuticallyef caciouscell numbers.Thisapproachmayhaveseveralpotentialadvantagesover peripheralbloodTregs.CBisincreasinglystoredatbirth,andacquisitionofthismaterialdoesnotrequirealarge-volumeblooddrawor leukapheresisprocedurethatmaybeonerousinpediatricpatientpopulations.3Inaddition,CByieldsahighlynaiverepertoireofTregs,presumablypriortotheexpansionofanypathogenicautoreactiveTeff populationsthatcouldpotentiallycontaminateaTregpreparation whenisolatedfromT1Dpatientsatthetimeofdiseaseonset.Moreover,evidencefrombothanimalmodelsandhumaninvitrostudies suggeststhatnaiveTregsexhibitincreasedlong-termphenotypicstabilitycomparedwithmemoryTregs,andthisnotionisalsore ectedin theepigeneticpro leofthesecells.33,34Inthisstudy,wedevelopedaprotocoltoisolateCB-derivedCD4+CD25+CD127lo/ Tregs(CBTregs)andexpandthemforpotential autologousTregACTtherapeuticapplications.Byutilizing uorescence-activatedcellsorting(FACS)forisolation,anextremelypure populationofTregscanbeisolatedandinducedtoproliferateto therapeuticdosesconsistentwithpriortrialefforts.23Ourresulting protocolwasthenvalidatedundergoodmanufacturingpractices (GMP)conditionsanddemonstratedtomeetbothclinicalandsterilityreleasecriteriaconsistentwithapriorTregtrial.23OurGMPprotocolresultedinahighlypureandpotentpopulationofTregsthat maintainedtheircanonicalphenotypicandfunctionalqualities. OurdatademonstratethatTregsderivedfromcryoCB(cryoCB Tregs)canbeeffectivelyisolatedandexpandedwithapurityandpotencycomparablewithTregsisolatedfromCBoradultperipheral blood(APB).Overall,theseresultssupportadditionaltestingof cryoCBTregsasapotentialtherapyforthetreatmentofT1Dand otherin ammatoryandautoimmunediseases.RESULTSIdenticationandIsolationofCD4+CD25+CD127lo/Tregsfrom CryoCBUsMethodsforthecryopreservationandhandlingofCBUshavebeen optimizedfortherecoveryofCD34+HSCs.Wesoughttodetermine whethercryoCBTregscouldsurviveandsustainsuf cientsurface markerexpressiontofacilitateisolationbyFACSandtoenable invitroexpansion.35PrioreffortsutilizingCBforisolationofTregs havereliedprimarilyonmicrobead-andcolumn-basedpositiveselectionofCD4+CD25+Tcells.Ina2008study,microbeadisolationof CBTregsproducedanaverageproductofabout48%Tregs(CD25+FOXP3+orCD127FOXP3+)priortoexpansion,27and,ina2011 study,only64%ofcells(range,31% – 96%)wereconsideredTregs (CD127FOXP3+)attheendofexpansion.28OurFACSplots( Figure1 A)indicatethat,althoughtheTregpopulationwasclearly moredistinctinCB(whetherfreshlyisolatedorpost-cryopreservation)overAPB,itstillcontainsaminorfractionofCD4+CD127+CD25intermediateTcellsthatwouldnothavebeeneliminatedby microbeadisolation,whichlikelycontributedtothenon-FOXP3expressingcellsfoundcontaminatingtheculturesinthosepreviousefforts.27Hence,wefoundFACSisolationofCD4+CD25+CD127lo/ Tregstobesuf cientandnecessary,eveninthesetting ofCB,forproducingahighlypureendFOXP3+Helios+Tregwithout theneedforadditionalimmunosuppressiveagents(e.g.,rapamycin) thatnegativelyaffectexpansionkinetics.36Weobservedvariancein theTregstainingpro lesinAPBcomparedwithfreshandcryoCB; however,theCD4+CD25+CD127lo/ Tregpopulationwasreadily distinguishablefromconventionalCD4+Tcells(Tconvs)forall samples.Tregrecoverywascomparableacrossallthreesources( Figure1 B).Wedidnotethatthepost-sortingpurityofCD4+CD25+CD127lo/ Tregs(byextracellularmarkers,attainedbytakingpart ofthesortedproductandrerunningitthroughthesorter)was modestlyreducedimmediatelypost-sortingincryopreservedversus freshCBTregs(APBTregs=99.0% 0.82%,CBTregs=99.47% 0.41%,cryoCBTregs=96.91% 1.49%,****p<0.0001; Figure1 C). However,wenotethatthisreductioninpost-sortingmarkerstability didnotnegativelyaffectthe nalpurityoftheTregpopulationpostexpansion,asdemonstratedbelow.www.moleculartherapy.org MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017179

PAGE 3

Thenumberofcellsthatcanbeinitiallyisolatedhasthepotentialto dictatetherapeuticdosingcapacityandevensubsequentclinicalef cacy.WeobservedthattheTregyieldwascomparableinCBUsversus cryoCBUs( FigureS1 A).WhentheprecisevolumeofCBcollectedafterchildbirthcouldbedeterminedfromtheCBUdocumentation,we observedtheaveragenumberofcryoCBTregsrecoveredpermilliliter ofCBtobe3.22 103(SD=1.65 103,n=9)comparedwitha highermeanvaluefromCBof7.19 103(SD=2.96 103,n= 11,**p<0.01; FigureS1 B).BecausestorageandprocessingproceduresvaryforanyCBbank,wecomparedtheyieldofTregsfrom bothacademicresearchandprivateindustrysources.Overall,theprivatebanksdidyieldahigheraveragenumberofTregs( FigureS1 C); however,wenotedthatthesampleswereselectedfromalarge numberofavailableresearchunits,andthispre-selectionbiaswill notbefeasibleforautologouspatientunits.Wethereforesoughtto identifyasurrogateindicatorofpotentialTregcellyieldfrominformationuniformlyrecorded.Whenweanalyzedthreesourcesof cryoCB(anacademicbankattheUniversityofFlorida[UF,abiorepositoryofcryopreservedresearchcordsnotdeemedeligiblefor banking,obtainedfromtheLifeCordPublicCordBloodBank]and twoprivatebanks,ViaCordandCordBloodRegistry)forfactors potentiallypredictingTregyield,wefoundthattheCD34+cellcount priortocryopreservationwaspositivelycorrelatedwithpost-thaw TregrecoveryinCBobtainedfromtheUFbiorepository(r2= 0.823;**p<0.01),butsuchcorrelationwasnotobservedinCB obtainedfromtheprivatebanks(r2=0.028,p=0.75;r2=0.006, A B C Figure1.IsolationofTregsfromAPB,CB,andCryoCBbyFACSProducesaHighlyPurePopulation (A)CD4+gatedscatterplotsshowinggatingforCD25+CD127lo/ TregsfromAPB,CB,andcryoCB.AlthoughallTregrecoveryandpurityvaluesreachedtargetranges,a statisticallysignicantsacriceismadeforcryoCBTregs.(B)CD25+CD127lo/ TregsasapercentageofCD4+TcellsweresignicantlylessduringFACSpuricationof cryoCBcomparedwithAPB(**p<0.001,Tukey'smultiplecomparisonstest).(C)Post-sortingpurityofCD25+CD127lo/ Treg(attainedbyre-examiningasmallportionofthe collectedsample)wascomparableforcellsisolatedfromAPBandCBbutlowerforcryoCBTregs(Tukey'smultiplecomparisonstest,****p<0.0001).MolecularTherapy:Methods&ClinicalDevelopment 180MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017

PAGE 4

p=0.84; FigureS1 C).Wedidnotseeanysigni cantcorrelationbetweenpre-cryopreservationtotalnucleatedcell(TNC)countand Tregyield( FigureS1 D),nordidweobserveanyeffectofCBstorage durationonTregrecovery( FigureS1 E). WealsotestedthefeasibilityofisolatingcryoCBTregsfromthe partitioned20%fractionofcryoCBUs(adividedchamberofthe cryoCBUconsistingofabout20%ofthetotalvolume)toleavetheremaining80%-fractionoftheunitforfuturetreatments.Followingthe samemethodsstatedpreviouslyforthawingandcellisolation,we foundthat,onaverage,30,044 6,035cryoCBTregscouldbe FACS-puri edfromthe20%-fraction(n=5; Table1 ).Following stimulationandexpansionusingprotocol2,wefoundthat1.98 108 7.02 107cryoCBTregscouldbeobtainedandusedforpotentialautologousACTtherapy( Table1 ).Basedontheseresults,it wouldappearthatisolationandexpansionofevenaminorfraction ofacryoCBUisfeasible,butthedownstreamutilityofthiswillrequire dosendingstudiesfordeterminingthenecessarycellnumbersfor ACT.TregsDerivedfromCryoCBUsRetainTheirLineage CharacteristicsandSuppressivePhenotypeWedevelopedandoptimizedmethodsfortheexpansionofcryoCB TregsutilizingmethodsadaptedfromPutnametal.36TregswereisolatedfrombothfreshandcryoCBUs,andthesewerecomparedwith TregsderivedfromAPB.Tregswereexpandedinvitroaccordingto threeuniqueprotocols(fromthispointforwardreferredtoasprotocols1 – 3),asdescribedunder MaterialsandMethods anddepictedin FigureS2 .Brie y,allthreeprotocolsincludedTcellstimulationon days0and9alongwiththepresenceofexogenousIL-2intheculture medium.Protocol1wasa14-dayexpansionschedulethatutilizedan anti-CD3(cloneOKT3)loadedhumanK562erythroleukemia-type arti cialantigen-presentingcell(aAPC)line(KT/64/86;aAPCsexpressingCD64andCD86)onday0,followedbyanti-CD3/antiCD28beadrestimulationonday9.Becausebeadsoffertheadvantage ofoff-the-shelfGMPcompliance,protocols2and3didnotinclude aAPCsbut,rather,utilizedanti-CD3/anti-CD28beadsforeachround ofTcellstimulation.Forprotocol2,analyseswereperformedonday 27,whereasprotocol3initiallyinvolvedanexpected23-dayexpansionschedulebutwassubsequentlyshortenedwhenweobserveda suf cientTregyieldonday16foratargetedtherapeuticdosefor ACT(asshownbelow). We rstcomparedexpansionfromCBTregsversuscryoCBTregs andfoundthatneithercellyieldnorfoldexpansionwerediminished bypriorcryopreservationwithexpansionbyprotocol1( Figures2 A – 2C)orprotocol2( Figures2 D – 2F).CB-derivedTregs,whethercryopreservedorfreshlyisolated,hadincreasedproliferativecapacity comparedwithAPBTregs(**p<0.01and*p<0.05,respectively; Figure2 C).Whenprotocol1andprotocol2weredirectlycompared,we foundthatCBandcryoCBTregcountsweresimilaronday14(the naldayofprotocol1;p=0.65andp=0.42,respectively).However, protocol2offeredahighermeanTregyield,althoughweobserved greatervariationacrossruns,likelybecauseofthethirdroundofstimulationwithanti-CD3/anti-CD28GMPbeadsandextendedduration ofculture.Usingprotocol3,wefoundthatTregsexpandedef ciently ( Figures2 A,2D,and2G),yieldingnumberssuf cientforatherapeuticdose(estimatinga1.2 109cellrequirementtoyield30 106cells/kgforanaverage40-kgpediatricpatientbyday16( Figure3 H). Wethusdeterminedthatprotocol3offerstheoptimalGMPcompliantapproachforCBTregexpansionbasedonoff-the-shelf reagentavailabilityandexpansionparameters. WenextexaminedAPB,CB,andcryoCBforpost-expansionTregpurityandCD8+TcellcontaminationaccordingtotheUniversityof California,SanFrancisco(UCSF)phase1trialclinicalreleasecriteria forpolyclonalTregsexpandedfromAPB.23Additionally,weassessed post-expansionsuppressivefunctionfromcryopreservedexpanded Tregpreparations.Clinicalmicrobiologyreleasecriteriawereevaluatedfromculturealiquotscollectedatmultipletimepoints throughouttheexpansionperiod,andalltestsforendotoxinaswell asbacterialandfungalcontaminationwerefoundtobenegative. Asexpected,theCD4+Tcellpopulationwasalmostentirely composedofFOXP3+Helios+TregsinbothCBandcryoCBTreg expansions,whereasTregpuritywassigni cantlylowerincells expandedfromAPB(protocols1and2,****p<0.0001; Figure3 A). Indeed,allexpandedCBandcryoCBTregpreparationsgreatlysurpassedthereleaserequirementfor R 60%FOXP3+.23EpigeneticanalysisoftheTreg-speci cdemethylatedregion(TSDR)withinthe FOXP3 conservednon-codingsequence2(CNS2)locuscon rmed thatthymicTregpuritywasgreatestamongTregsisolatedand expandedfromfreshorcryopreservedCB(protocol1:CB= 97.8% 1.0%,cryoCB=96.9% 3.5%;protocol2:CB=92.1% 4.6%,cryoCB=93.9% 8.2%;protocol3:cryoCB=89.0% 9.8%). APBTregsdemonstratedsigni cantlylessdemethylationatthe TSDRcomparedwithcryoCBTregs(protocol1:mean=78.5% 10.8%,**p<0.01;protocol2:mean=80.9% 11.2%,**p<0.01; Figure3 B).Asexpected,CBTconvcontrolcellsexhibitednearlycompletemethylationoftheTSDR(3.8% 2.6%demethylated,n=5; Table1.The20%FractionofacryoCBCanProduceaSubstantialNumber ofExpandedcryoCBTregsforACTTherapy UnitNo. Day0:cryoCB TregsIsolated Day27ofExpansion: CellCount Fold Expansion 126,568176,000,0006,625 242,158298,000,0007,069 339,145221,000,0005,646 429,456105,000,0003,565 527,894189,000,0006,776 Average33,044198,000,0005,992 Abene tofusingthe20%fraction(adividedchamberofthecryoCBUconsistingof about20%ofthetotalvolume)isthattheremaining80%canbeusedforfutureautologoustherapieswhenneeded.Todeterminethefeasibilityofusinga20%fractionto derivecryoCBTregsfromacryoCBU,weisolatedandexpandedcryoCBTregsusing protocol2( MaterialsandMethods ).UsingFACS,3.3044 104 6.035 103cryoCB Tregswereisolated(n=5)andexpandedto1.98 108 6.28 107cells.www.moleculartherapy.org MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017181

PAGE 5

Figure3 B).CD8+Tcellcontaminationwasminimal,particularlyin cellsexpandedfromCB(protocol1:APBTregs=0.8% 0.4%,CB Tregs=0.4% 0.3%,cryoCBTregs=0.5% 0.3%; Figure3 C),presumablyfromthelowerfrequencyofCD8+TcellinCB.37Again, thesevalueswerewellbelowtheclinicalreleasecriteriaof % 5% CD4CD8+contamination.Correspondingly,foreachprotocol, >99%ofexpandedcryoCBTregswereCD4+,inaccordancewith thepolyclonalAPBTregreleasecriteria.23Notably,interferon g (IFNg )productionwassigni cantlyhigheramongTregsisolated andexpandedfromAPB(protocol1,7.5% 3.2%;protocol2, 9.7 4.4%)comparedwithbothfreshandcryopreservedCBpreparations(protocol1:CB=1.8% 0.9%,**p<0.01;cryoCB=1.7% 0.9%,**p<0.01;protocol2:CB=2.2% 1.2%,**p<0.01;cryoCB= 2.2% 1.2%,**p<0.01; Figure3 D).CD4+TcellsfromCB,as expected,havenearlyuniformexpressionoftheCD45RAisoform characteristicofnaiveTcells( Figure3 E).Importantly,weobserved thatTregsexpandedfromCBretainedhighlevelsofCD45RAexpression,evenfollowinginvitroexpansion( Figure3 F),incontrastto expandedAPBTregsthatconverttotheCD45ROisoform.38Finally, Tregswereevaluatedforfunctionalsuppressivecapacityafterexpansion.Importantly,TregsexpandedfromcryoCB,CB,andAPBall demonstratedtheabilitytosuppressbothpolyclonalCD4+and CD8+Tcellresponses( Figure4 ).CBTregsExhibitaHighlyDiverseReceptorRepertoirethatIs MaintainedfollowingExpansionTregTcellreceptor(TCR)diversityhasbeendemonstratedto bebene cialinmaintainingself-tolerance.39Moreover,areport A C B D G EF HID a y0 D a y 9 D ay 14 0 cryoCBTreg CBTreg T o t a lCel l s (x1 09)1 0.5 APBTreg C B cryoC B AP B 0 n.s.Fold Ex p a nsi on(x1 03)2 4 6 *** *D a y 0 Day 9 Day 14 D a y18 D ay 2 7 0 T o ta lCells (x 109)CBTreg cyroCBTreg 3 2 1 APBTreg C B cr y o CB APB n.s.F oldExpansion(x104)2 4 6 8 ** *D a y0 Day9 Da y 16 0 cryoCBTreg Tr egCount( x109)2 1 3c ry o CB 0 F old E xp ansion(x103)2 4 6 8CB c ry oC B A PB 10610710810910101011 n.s.Tr e gCo u n t n.s. **CB c r yo C B A PB 10610710810910101011 n.s.TregCount n.s. n.s.cr yoCB Tr eg 10610710810910101011 TregC o unt Figure2.TregsIsolatedfromCBorCryoCBWereExpandedInVitroAccordingtoThreeProtocols (AI)Tregexpansionefciencyisshownfor(AC)protocol1,(DF)protocol2,and(GI)protocol3.(A,D,andG)ThetotalTregcountovertimewascomp arableforCBversus cryoCB(two-wayANOVA,NS=p R 0.05).(B,E,andH)TheTregfoldexpansionfrombaselineonthenaldayofculturewasnotsignicantlydifferentforCBversuscryoCB(Tukey's multiplecomparisonstest,NS=p R 0.05).(C,F,andI)TheTregcountonthenaldayofculturewasnotsignicantlydifferentforCBversuscryoCB(Tukey'smultiplecomparisons test,NS=p R 0.05).Protocolmethodologiesareoutlinedin FigureS1 .p<0.05wasconsideredstatisticallysignicant.*p<0.05,**p<0.01,***p<0.001,****p<0.0001.MolecularTherapy:Methods&ClinicalDevelopment 182MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017

PAGE 6

byYangetal.40demonstratedadistinctivemurineTCRrepertoire amongTregsgeneratedearlyinde velopmentduringtheperinatal period,whichexhibitlessclonalexpansionandareuniquely capableofdefendingtissuesagainstautoimmunedestruction comparedwithTregsderivedfromadultmice.Therefore,we soughttodeterminetherelative diversityofthepolyclonalTreg populationsderivedfromCBrelativetothoseobservedinAPB Tregs.Forthisanalysis,wecon ductedimmunosequencingofthe complementarity-determiningregion3(CDR3) b chainloopof Figure3.TregsIsolatedandExpandedfromCB, CryoCB,andAPBWereEvaluatedforPostexpansionPurity,Stability,andNaivety (A)FOXP3+Helios+TregfrequenciesamongCD4+cells weresignicantlyhigherforTregsexpandedfrom CBorcryoCBcomparedwithAPB.Tregswere expandedviaprotocol1(ANOVA,**p<0.001,NS= p R 0.05).(B)ThepercentageofcellswithdemethylatedFOXP3-TSDRwassignicantlygreateramong TregsexpandedfromCBorcryoCBbyprotocols1,2, and3comparedwithTregsexpandedfromAPBby protocol1(**p<0.01,***p<0.001).Asexpected,low FOXP3-TSDRpercentdemethylationwasobserved forCD4+Tconvcontrolsisolatedandexpandedfrom CB.(CandD)Thepercentagesof(C)CD8+cells and(D)IFNg+cellsweresignicantlylowerforTregs expandedfromCBorcryoCBversusAPB(ANOVA, *p<0.05,**p<0.01,***p<0.001,****p<0.0001, NS=p R 0.05).TregsorTconvswereexpandedvia protocol1.(C)CD8+frequencieswerecomparable amongCD4+Tconvsisolatedandexpandedfrom APB,CB,orcryoCB.(D)Asexpected,IFNg+cell frequencieswerehigheramongTconvsisolatedand expandedfromAPB,CB,orcryoCB.(EandF) RepresentativehistogramsshowingCD45RAexpressionbygeometricmeanuorescenceintensity(gMFI)in CBTcellsandinsomeexpandedpopulations(E)and meangMFIforCD45RAexpressiononexpandedTreg andTconvpopulations(F),separatedbyexpansion protocol.theTCR(TCR b ),ahighlyvariableregion formedasaresultofTCRV(D)Jgene segmentrecombinationthatservestoengage antigenpeptidespresentedbyHLAmolecules.41WecomparedTregTCR b V-gene ( TRBV )andJ-gene( TRBJ )frequenciespreandpost-expansionfromeitherCBorAPB andfoundthatV-gene( Figure5 A)and J-gene( Figure5 B)distributionswerenot noticeablychangedasaresultofexpansion foreitherTregsource.Similarly,uponevaluatingthefrequencyofCDR3sequence lengths,forwhichanon-Gaussiandistributionwouldindicateamonoclonalexpansion, weobservedanormaldistributionwitha meanlengthof42aminoacids(aa)inbothpre-andpost-expansionTregs( Figure5 C). InassessingTCRdiversity,wequanti edCBTregproductive clonality,ascoreofthesampleproductivediversitynormalized tosampleproductiveentropy.Aproductiveclonalityscorenear1 wouldindicateveryfewproductiverearrangements(monoclonal) andlittleTCRdiversity,whereasascorenear0wouldindicate amorepolyclonalsample.ForCBTregs,productiveclonalitydidwww.moleculartherapy.org MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017183

PAGE 7

notsigni cantlychangeduringexpansion(n=5;NS,p=0.15,twotailedpairedStudent ’ sttest; Figure5 D).Apre-expansionproductive clonalityof1.64 10 3 8.82 10 4increasedto2.60 10 3 1.70 10 3post-expansion,indicatingthatCBTregsexpressa verydiversereceptorrepertoirebothbeforeandafterexpansion. Forperspective,previousstudieshaveobservedthemeanproductive clonalityofunexpandedperipheralbloodTregsobtainedfromthree children(9.2 – 16.1yearsold)tobe5.05 10 2 5.09 10 3(n= 3),42and,similarly,wediscoveredunexpandednaiveTcellsobtained fromAPBtobe6.03 10 2 2.78 10 2( Figure5 D).Additionally, APBTregswereFACS-puri edfromasingleT1Dpatientsample fromclinicaltrialNCT01210664,23andtheproductiveclonalityvalue wasfoundtobe6.10 10 2(n=1;J.A.B.,unpublisheddata).Taken Figure4.SuppressiveFunctionofCB,CryoCB,and APBTregs (AF)Tcells(responders)wereisolatedfromPBMCs, labeledwithCTV,andplatedinincreasingproportions withexpandedTregs(suppressors)fromprotocol1 (AandB),protocol2(CandD),andprotocol3(EandF) thatwerelabeledwiththecellproliferationdyeeFluor670, inratiosasindicatedalongthexaxis.Thecellswere activatedinvitrowithsolubleanti-CD3andanti-CD28, andtheproliferationofCD4+orCD8+respondercells (Tresp)wasmeasuredviaowcytometryafterlabelingwith uorescentlyconjugatedanti-CD4andanti-CD8antibodiestodistinguishthepopulations.Percentsuppressionwascalculatedas:100 [100 (percentageof proliferatingcellswithTregpresent)/(percentageofproliferatingcellswithoutTresppresent)].Suppressivecapacity wasnotsignicantlydifferentforTregsexpandedfrom cryoCB,CB,orAPB(two-wayANOVA,p=n.s.,all).together,thesedataimplyasubstantialreductioninTregTCRrepertoirediversityfrominfancytoadulthood(*p<0.0001).DISCUSSIONTcellsforACTtherapieshavebeenappliedfor treatmentofcertaincancersandinfections(reviewedbyBuschetal.43),andclinicaltrialsof autologousTregs,harvestedbyleukapheresis andexpandedfromperipheralblood,have demonstratedsafetyinadultswithimmunemediateddisorders,includingGvHDand T1D.22,23Althoughperipheralbloodoffersa largepoolofavailablecells,anincreasingnumberofperipheralbloodTcellsadoptaterminallydifferentiatedmemoryphenotypewith subjectage,reducingexpansioncapacityand thelikelihoodofsuccessfulengraftmentand survival.43Conversely,CB-derivedTregsare farlesslikelytoadoptamemoryphenotype giventhelackofcentralandeffectormemory subsets.Assuch,CBTregsmayofferanoptimalpopulationof CD28+CD27+naiveTregswithimprovedpurityfollowingFACS isolationfromCBversusperipheralbloodandcapacityforgreater stabilityduringextendedexpansions,facilitatinglong-termengraftment.44CryoCBTregs,therefore,areapotentiallyoptimalsource forautologousTregsforACTtherapies. InfusionofautologouswholeCBUswasshowntobesafebutdidnot preserveC-peptideinpediatricT1Dpatients,30suggestingapotential needforhigherdosesofTregs.Thispromptedthedevelopmentofa GMP-compliantprotocoltoexpandTregsfromcryoCBUstoproduceahighlypureandfunctionallysuppressivecellpopulationfor autologousACTtherapy.HerewecomparedthreeexpansionMolecularTherapy:Methods&ClinicalDevelopment 184MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017

PAGE 8

A B C D Figure5.ImmunosequencingoftheTCR b ChainDemonstratesIncreasedReceptorDiversityofCBTregs ImmunosequencingoftheTCR b chainwasperformedforTregsisolatedfromfreshCBpriortoandafterinvitroexpansionaccordingtoprotocol1.(AandB)TRBVgene usage(A)andTRBJgeneusage(B)werenotdifferentamongpre-expansion(black)andpost-expansion(gray)CBTregsamples.(C)ThedistributionofCD R3nucleotide lengthswascomparableinpre-expansion(closedcircles)andpost-expansion(opencircles)CBTregsamples.(D)Analysisofproductiveclonality( aninverseindicatorofTCR clonaldiversity)didnotshowasignicantdifferencebetweenpost-expansionCBTregs(graysquares)andpre-expansionsamples(blacksquares).C omparedwithpre-or post-expansionCBTregs,productiveclonalitywassignicantlyhigheramongunexpandedTregsisolatedfromperipheralblood(PB)samplesobtaine dfromyoungdonors (age9.216.1years),adultPBnaiveCD4+Tcells,oradultPBmemoryCD4+Tcells(Tukey'smultiplecomparisonstest,****p<0.0001).www.moleculartherapy.org MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017185

PAGE 9

protocolsusingTregsisolatedfromfreshandcryoCBaswellasAPB. Althoughprotocol1iswellvalidated,itutilizesaAPCsand,therefore, presentsreagentandGMPproductionchallenges.Protocol2isGMPcompliantandutilizesmultiplemicrobeadrestimulationstoyieldthe greatestnumberofexpandedTregs.Unfortunately,themicrobeads, whichbindmoreeffectivelytocells,aredif culttoremovefrom cultures,and,atthetimeoftheseexperiments,hadnotyetbeen validatedforclinicalreleasecriteria(althoughthemicrobeadshave sincebeenusedintheThRILtrial45).Protocol3utilizedGMPapprovedmaterialsforexpandedTcellproductssuitablefortesting inhumansthathavebeensuccessfullyutilizedinaphaseItrialof APBTregs.23Thepost-sortingpuritywaslowerfromcryopreservedversusfresh CB,butexpansioncurveswerenearlyidenticalfromthetwosources ofCB.Meanwhile,bothCBandcryoCBTregsexpandedmoreef cientlythanTregsfromAPB.Acriticalrequirementforanyeffective TregACTtherapyisthatitmustproducesuppressive,lineage-stable Tregsthatdonotproducein ammatorycytokines.CryoCBTreg expansionproductsfromthethreeuniqueprotocolsperformed comparablyinananalysisofTreg-suppressivecapacityandstability (de nedasFOXP3andHeliosco-expressionanddemethylationof the FOXP3 TSDR).Importantly,expandedcryoCBTregsmetpreviouslydeterminedclinicalreleasecriteriapertainingtothepercentage ofcellsthatmaintainFOXP3positivity,lowCD8+Tcellcontamination,andsterility.23Thetargetdoseisnotyetdetermined,butadose escalationtrialusingTregsexpandedfromABPhasdemonstrated safetywithdosesashighas2.9 109infusedTregs.23Wewere abletoexpandcryoCBTregstonumbersnearandevenabovethis value.Additionalclinicalstudiesareneededtode nitivelyidentify thetargetdoseforpatientswithT1D.AsaninitialphaseItrial,we proposetomoveforward,speci cally,withthegoaloftreatingpediatricsubjectswithT1Dinadoseescalationsafetytrial(at10,20,and 30 106Treg/kgbodyweight).ThesedosesareinlinewithpriorclinicaltrialsusingperipheralbloodTregs.22,46Currentexpansionswere performedusingTregsisolatedfromthecompleteCBU,andongoing effortsareneededtogenerateTregexpansionsfromafractionofthe CBUtoallowforpotentialrepeatdosing. TissuedistributionofTcellsiscontrolledbythespeci cityofthe TCR.47Moreover,activationthroughtheTCRisrequiredforTregsuppressiveactivity.Hence,weaskedhowthediversityofCBTreg TCRscomparedpre-andpost-expansionbyconductingnext-generationsequencingoftheTCR b -chain(TRB)locus.Thedatasupport thenotionthatourinvitroexpansionprotocolinducesbroadexpansionofallcloneswithoutsigni cantdeviationstowardoligoclonality. The b -chainV-genefrequencyandthesampleclonalityindicatethat, althoughtheTCRdiversitydecreasesslightlyduringinvitroexpansion,post-expansionCBTregsremainfarmorepolyclonalthan APBTregs.42Thishasimportantimplicationswhenconsidering theuseofAPBorCBasasourceofTregmaterial,eitherfreshor followingTCRorchimericantigenreceptor(CAR)genetransferapproaches,whichareincreasinglyindevelopment,particularlygiven thesuccessesinanimalmodelsusingTCR-transgenicTregs.48This notionofanoptimalTregcellpopulationwillprovideforoptimal suppressionoftissue-reactiveTeffpopulationswhilealsominimizing thepotentialforuntowardimmunosuppressionofpathogensandpotentialtumorantigens.Wearguethatduetotheiruniquelythymic origin,CBTregscontainreceptorscapableofseedingawidevariety ofperipheraltissuesearlyinlife,wouldlackanyenrichedclonesof pathogensthatmaybecomeenrichedthroughchronicinfectionin life(e.g.,tocytomegalovirus[CMV]orEpstein-Barrvirus[EBV]), andwouldlackTregsarisingtoneo-antigensthatsuppressprotective tumorimmunity.49Additionally,CBTregswillremainsubjecttothe normalhomeostaticcontrolsofahost,includingtheabilitytoattenuatesuppressionintheinstanceofacuteviralinfection.50Asawhole, thesedatasuggestthatournewGMP-compliantprotocolyieldscells suitableinnumber,phenotype,andfunctionfortestingasapotential treatmentinhumanswithautoimmunedisease,includingT1D. TheseexpandedcryoCBTregswouldprovideapracticalalternativeto theneedforalarge-volumeblooddraworleukapheresisprocedure currentlyneededforautologousACTtherapy,abene tofparticular importanceforpediatricT1Dpatients.Thatsaid,unlikeperipheral bloodcollection,thesingularopportunitytostoreautologousCBcreatesasituationinwhicheffortstoutilizefractionsofstoredCBUs,facilitatedbymulti-compartmentstoragebags,areneededtoallowforlongitudinaltreatmentregimens.Giventhehighsafetypro leobserved withwholeCBinfusion30aswellasperipheralbloodTregACTtherapy,5,23weexpectautologousCBTregACTtherapytobewelltolerated.Moreover,expandedCBTregsmayofferanallogeneiccellsource whenautologouscelltherapyisdemonstratedtobesafe.Indeed,allogeneiccells(whensuf cientlymajorhistocompatibilitycomplex [MHC]-matched)fromCBdonorswithoutT1Dorassociatedrisk allelesmightofferimprovedtherapeuticbene tsgiventhepotential advantagesofnon-autoimmunebackgroundgeneticsandintrinsic functionaswellasgreateravailabilityfromabroaderpatientbase. Insummary,TregscanbeconsistentlyisolatedfromcryoCBUsusing standardmethodsandreagentsandstimulatedtoexpandintoanumberofcellsthatcanbeconsideredwithintherapeuticdoserangesof priortrials.Therefore,the ndingsreportedheresupportcontinued testingofTregsexpandedfromcryoCBforsafetyandef cacyasapotentialtreatmenttopreventorcureT1D.MATERIALSANDMETHODSSampleCollectionandMononuclearCellIsolation“ Fresh ” CB(freshisde nedasprocessedwithin48hrofcordblood draw)wascollectedbyNewYorkBloodCenter ’ sNationalCord BloodProgramintoCBUscontaining35mLofcitratephosphate dextrose(CPD)anticoagulant.51CBUs(n=11)werethenshipped totheUniversityofFloridaDiabetesInstitutelaboratoriesandimmediatelyprocessedforisolationofcordbloodmononuclearcells (CBMCs)bydensitygradientcentrifugation(Ficoll-PaquePLUS, GEHealthcare). CryoCBUs(n=19)wereobtainedfromprivatecordbloodbanks (CbrSystemsandViaCord,asubsidiaryofPerkinElmer)orthroughMolecularTherapy:Methods&ClinicalDevelopment 186MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017

PAGE 10

anonymousdonationtoaUFbiorepositoryofcryopreservedresearch CBUsobtainedfromLifeCord,apubliccordbloodbank.CryoCBwas collectedandprocessedaccordingtothestandardoperatingproceduresoftheindividualbanksfromunitsmeetingNetCord-FACT InternationalStandardsforCordBloodCollection,Banking,and ReleaseforAdministration,mostnotablyataminimumgestational ageof34weeks,butwithadditionalexclusioncriteriaincluding infectiousdiseasesinthemother,severepregnancycomplications, orprematuredeliverywithabirthweightoflessthan1,500gor whenperinatalasphyxiaispresentinthefetus,andcellcounts.52AllsampleswereobtainedunderInstitutionalReviewBoard(IRB)exemptapprovedprotocolsatUF(non-humanexemptprotocol IRB201300072).CryoCBUswerethawedasdescribedpreviously,53resuspendedin60mLofdextran-HSAwashsolution(dividedinto two50-mLconicaltubes),allowedtowarmtoroomtemperature, andunderlaidwith15mLFicoll-PaquePLUSperconicaltubefor densitygradientcentrifugationandCBMCisolation. Peripheralbloodwascollectedafterinformedconsent,inaccordance withtheIRBatUF(protocolIRB201400703),fromhealthycontrol subjects(meanage,28.8 5.4years;range,22.5 – 38.7years;n=19) insodium-heparinizedVacutainertubes(BectonDickinson)and processedwithin24hrforisolationofperipheralbloodmononuclear cells(PBMCs)bydensitygradientcentrifugation(Ficoll-PaquePLUS, GEHealthcare).SampleProcessingandIsolationofTCellsbyFACSUnlessotherwisenoted,cellisolationandcultureguidelines, includingcellnumbers,volumes,andrespectiveculture asks,have beendescribedpreviously.36Forresearch-gradeTregexpansions, TcellssubsetswereisolatedonaFACSAriaIIIhigh-speedcellsorter (BDBiosciences)withthefollowingantibodies:CD4-Paci cBlue (cloneRPA-T4),CD127-PE(clonehIL-7R-M21),andCD25-APC (clone2A3).CD4+CD25+CD127lo/ TregsandCD4+CD127+Tconvs weresortedinto400 m Loffetalbovineserum(FBS,UnitedStates DepartmentofAgriculture[USDA]-approvedorigin,AtlantaBiologicals)usinganaseptictechnique.Forprotocol3,Tregisolationwas describedpreviously.23InVitroExpansionProceduresGiventhenaiveandhighlyrefractorynatureofCBTregs,wetesteda numberofstimulationconditionsandexpansiondurationstooptimize nalcellyieldandpurity.Thisentailedtheuseofactivating aAPCsand/oractivationwithcommerciallyavailableantibody-conjugatedmicrobeads.Cellculturevolumeand askrecommendations arelistedin TableS1 .Theconditionsforthesethreeprotocolscanbe visualizedin FigureS2 andareoutlinedbelow.Protocol1:14-DayaAPCStimulationandAnti-CD3-andAntiCD28-CoatedMicrobeadRestimulationFACS-isolatedcellswereplatedaccordingtoPutnametal.36aAPCs werepreparedfromKT64/86,aK562-derivedcelllineconstitutively expressinghigh-af nityFcreceptor,CD64,andCD86forco-stimulation(akindgiftfromDrs.JamesRileyandBruceLevine,University ofPennsylvania).aAPCsweregenerated,cultured,andpreparedfor co-cultureasdescribedpreviously.54,55Brie y,Fc-bindingreceptors onKT64/86werepre-clearedofserumimmunoglobulinsbyculture inserum-freemedium(SFM)overnightandthenirradiatedat 10,000rad.Anti-CD3(cloneOKT-3,MiltenyiBiotec)monoclonal antibody(mAb)wasloadedonKT64/86at1 m g/106cellsat4Cfor 30min,washedtwicewithSFM,andcryopreservedinCryoStor CS10(BioLifeSolutions).AfterTregFACSpuri cation(described above),KT64/86aAPCswereaddedtocultureata1:1aAPC-toTregcellratio.CBTregsandcryoCBTregswereexpandedin completeRPMI1640(cRPMI) — consistingofRPMI1640(Life Technologies)supplementedwith10%FBS(AtlantaBiologicals), 1MHEPES,1mMsodiumpyruvate,100 minimumessentialmedium(MEM)non-essentialaminoacidsolution,50mM2-mercaptoethanol(2-ME),and100Upenicillin/streptomycin(Gibco) — plus 600IU/mLProleukin(humanrecombinantIL-2[hrIL-2],PrometheusLaboratories).APBTregswereexpandedincRPMIplus300 IU/mLProleukin(hrIL-2,PrometheusLaboratories).Onday2,the culturevolumewasdoubled,andIL-2wasadded(attheaforementionedconcentrations,assumingconsumption).Cellswereresuspended,andfreshmediumandIL-2wereaddedondays4,6,8,11, and13.Onday9,cellswererestimulatedwithDynabeads(human T-activatorCD3/CD28forTcellexpansionandactivation,DynalInvitrogen)ata1:1ratio.Protocol2:27-DayGMP-CompliantAnti-CD3-/Anti-CD28CoatedMicrobeadStimulationandMultipleRestimulationsFACS-isolatedcellswereplatedevenlyat2.5 104to5.0 104Tregs/ wellina96-well at-bottomplate(Costar)andactivatedwithantiCD3/anti-CD28-coatedmicrobeads(MACSGMPExpActTregkit forresearchuse,MiltenyiBiotec)ata4:1bead-to-cellratio.CBTregs andcryoCBTregswereexpandedincRPMIplus600IU/mLProleukin(hrIL-2,PrometheusLaboratories).APBTregswereexpandedin cRPMIplus300IU/mLProleukin(hrIL-2,PrometheusLaboratories). Onday2,theculturevolumewasdoubled,andfreshIL-2wasadded (attheaforementionedconcentrations,assumingconsumption). Cellswereresuspended,andfreshmediumandIL-2wereaddedon days4,6,8,11,13,15,17,20,22,24,and26,assumingconsumption. Ondays9and18,cellswererestimulatedwithfreshanti-CD3/antiCD28-coatedbeadsata1:1ratio.Protocol3:16-DayGMPAnti-CD3-/Anti-CD28-Coated MicrobeadStimulationandRestimulationExceptwherenoted,protocol3followedthepreviouslypublished protocol.38Brie y,FACS-isolatedcellswereplatedandactivated withDynabeadsata4:1bead-to-cellratio.Cellswerecultured eitherinX-VIVO15orinX-VIVO15customizedbyLonza bysubstituting100%oftheglucoseinthebasemediumwith D-glucose(6,6-2H2,99%)(CambridgeIsotopeLaboratories,catalogno.DLM-349-MPT)supplementedwith10%humanheatinactivatedpooledAPBserum.Onday2,theculturevolume wasdoubled,andIL-2wasadded(300IU/mL,Proleukin).Cells wereresuspended,andfreshmediumandIL-2wereadded (600IU/mL,Proleukin)ondays5,7,12,and14,assumingwww.moleculartherapy.org MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017187

PAGE 11

consumption.Onday9,cellswererestimulatedwithadditional Dynabeadsata1:1ratio. Additionally,the nalreleasecriteriaforthisprotocolrequirethat the nalcellproductbeevaluatedforpurity( % 5%CD8+cells, <100beads/3 106cells,andendotoxin % 3.5endotoxinunits [EU]/mL),phenotype( R 95%CD4+cellsand R 60%FOXP3+),sterility(negativeformycoplasma,anaerobicandaerobicbacteria, gramstain,fungalculture,potassiumhydroxide[KOH]exam),and viability( R 85%).23AnalysisofPopulationsPre-andPost-expansionPhenotypicanalysesandsuppressionanalysiswereperformedin batchesonrestedcells1dayafterthawingpost-cryogenicstorage. Followinginvitroexpansion,Tregswereevaluatedforcontinued expressionofCD4,CD8,CD25,CD127,FOXP3,Helios,and IFNg .Onday14(protocol1),day27(protocol2),orday16(protocol3, FigureS2 ),2 106cellswereremovedfromcultureandactivatedfor4hrwithphorbolmyristateacetate(PMA;10 m g/mL)and ionomycin(500nM)inthepresenceofGolgiStop(4 m L/6mLofculture,BDBiosciences).Cellswerethenstainedforsurfacemarkers (CD4-PE-Cy7,cloneRPA-T4;CD8-APC-H7,cloneSK1;CD25BB515,clone2A3;CD127-PE,clonehIL-7R-M21;3 m L/test). IntracellularstainingwasachievedwiththeFOXP3stainingkit (BioLegend)accordingtothemanufacturer Â’ sinstructionsandmodiedasfollows.Cellswerewashedand xedfor30minat23Cusing xation/permeabilizationbuffer.Cellswerewashed,resuspendedin permeabilizationbuffer,andincubatedovernightat4C.Thenext day,thesampleswerewashedinpermeabilizationbuffer.Cells weresubsequentlyblockedwithhumanimmunoglobulinG(IgG) (5 m g/test)for5minandstainedwithanti-humanFOXP3-PE(clone 259D,5 m l/test)andHelios-Paci cBlue(clone22F6,5 m L/test).Flow cytometricdatawerecollectedonanLSRFortessacytometer(BD Biosciences)andanalyzedwithFlowJosoftware(version10.0.0.7r2, TreeStar).BisulteSequencingandTSDRReal-TimePCRThymicTreglineagecommitmentandstabilityaresubjecttoepigeneticcontrolattheconservednon-codingsequence2(CNS2)of the FOXP3 gene,56commonlyreferredtoastheTSDR.57Weassessed thedegreeofCpGdemethylationattheTSDRofculturesfollowing expansionbyreal-timePCRTSDRasdescribedpreviously.58GenomicDNAwasisolatedfrompuri edTcellsusingtheDNeasy tissuekit(QIAGEN)followingthesupplier Â’ srecommendations. Bisul tetreatmentofgenomicDNAwasperformedwiththeEZ DNAmethylationkit(ZymoResearch).Brie y,500nggenomic DNAwassodiumbisul te-treatedovernightat50Cinthedark,followedby10-minincubationonice.Afterwashinganddesulphonation,bisul te-convertedDNAwaselutedwith16 m Lelutionbuffer. Bisul te-convertedDNAsampleswereusedforPCRampli cationof FOXP3 TSDRfragmentsinareactionvolumeof50 m Lcontaining 25 m LofZymoTaqPreMixbuffer(ZymoResearch)and0.5 m M eachoftheprimersFOXP3_TSDRfw(ATATTTTTAGATAGGGA TATGGAGATGATTTGTTTGG)andFOXP3_TSDRrev(AATAA ACATCACCTACCACATCCACCAACAC).Afteractivationat 95Cfor10min,ampli cationwasperformedasfollows:50cycles at95Cfor30s,at55Cfor30s,andat72Cfor1min.Ampli ed PCRproductswerepuri edwiththeQIAquickPCRpuri cation kit(QIAGEN)andsequencedinbothdirections,applyingthePCR primersandABIBigDyeTerminatorv1.1cyclesequencingchemistry (AppliedBiosystems),followedbycapillaryelectrophoresisonanABI 3100geneticanalyzer.Trace leswereinterpretedusingESME,a Linuxsystem-basedsoftwarethatnormalizessequencetraces,correctsforincompletebisul teconversion,andallowsforquanti cation ofmethylationsignals.Foreachsample,bothPCRampli cationand sequencingwererepeatedonce.TheprimersusedforsequencingreactionsarethesameasPCRampli cationofbisul te-converted genomicDNA. TSDRreal-timePCRwasperformedin96-wellwhitetrayswitha LightCycler480system(RocheDiagnostics)witha nalreactionvolumeof20 m L,containing10 m LLightCycler480ProbesMasterMix (RocheDiagnostics),3 m Lbisul te-convertedDNAsampleorstandards,and1 m Mofeachprimer(TSDR_ForwardGGTTT GTATTTGGGTTTTGTTGTTATAGT,TSDR_ReverseCTATAAA ATAAAATATCTACCCTCTTCTCTTCCT).Theprobes(6-carboxy uorescein[FAM]-labeledmethylated FOXP3 ,CGGTCGG ATGCGTC;VIC-labeledunmethylated FOXP3 ,TGGTGGTTG GATGTGTTG)wereaddedtoa nalconcentrationof150nM.All sampleswereanalyzedinduplicate.Afterinitialdenaturationat 95Cfor10min,thesamplesweresubjectedto50cyclesat95C for15sandat61Cfor1min. Variousratiosofmethylatedandunmethylated FOXP3 TSDRtemplateDNAweremixedandusedtogenerateastandardcurve. GenomicDNAextractedfromsortedTlymphocytes,whichareeither fullymethylatedorunmethylated,werebisul te-convertedandused forPCRampli cationofthe FOXP3 TSDRfragmentwiththesame pairsofprimersandprotocolasbisul te-convertedgenomicDNA (describedabove). Ampli edPCRproductswerepuri edwiththeQIAquickgel extractionkit(QIAGEN).Theconcentrationofpuri edDNAwas determinedwithaGENanoVuespectrophotometer(GEHealthcare). 17differingratiosofmethylatedandunmethylatedbisul te-convertedDNAmixtureswerepreparedandusedasreal-timestandards togenerateastandardcurve.InVitroSuppressionAssaysTheinvitro-suppressivecapacityofexpandedTregpopulationswas determinedbasedontheirabilitytosuppresstheproliferationofallogeneicCD4+andCD8+responderTcellsderivedfromasinglePBMC setfromahealthyadultdonor(standardizedresponder).PBMCs wereharvestedfromwholebloodbydensitygradientcentrifugation (Ficoll-PaquePLUS,GEHealthcare)andcryopreservedinmultiple aliquots( 10 106cells/vial).Foreachassay,anindividualvialof PBMCswasthawedandlabeledwithCellTraceViolet(CTV,MolecularTherapy:Methods&ClinicalDevelopment 188MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017

PAGE 12

2.5 m M,LifeTechnologies),whereascryoCBandCBTregswere freshlycollectedoneitherday14orday21ofexpansionandlabeled withthecellproliferationdyeeFluor670(1.25 m M,eBioscience)and platedatratiosof1:1, 1 = 2 :1, 1 = 4 :1, 1 = 8 :1, 1 = 16 :1,and1:0 TregstoresponderPBMCs(5 104Tregsat1:0condition) in96-wellround-bottomtissuecultureplates(CellTreat)in 200 m L/well.AllinvitrosuppressionassayswereculturedincRPMI. Thecellswerestimulatedwithsolubleanti-CD3(2 m g/mL,clone OKT3,BDBiosciences)andanti-CD28(1 m g/mL,cloneCD28.2, BDBiosciences)andincubatedfor4daysat37Cina5%CO2incubator.Onday4,thecellswereharvestedandstainedwithLive/Dead Yellowaccordingtothemanufacturer ’ sprotocols(LifeTechnologies),CD4-PE-Cy7(cloneRPA-T4,3 m L/test),CD8-APC-H7(clone SK1,3 m L/test),CD25-BB515(clone2A3,3 m L/test),andCD127-PE (clonehIL-7R-M21,3 m L/test).CD4+andCD8+responderproliferationwasassessedvia owcytometry.Theresponderpopulationswere gatedbyforwardscatterarea(FSC-A)andsidescatterarea(SSC-A), andlivecellsweregatedforCD4+andCD8+.Percentsuppressionwas calculatedas:100 [100 (percentageofproliferatingcellswithTregpresent)/(percentageofproliferatingcellswithoutTresppresent)]. AnalysiswascompletedusingFlowJosoftware(TreeStar,v10.2), andproliferationwasdeterminedbasedonthedivisionindexof live-gatedresponderpopulations.TCellReceptor b ChainSequencingSamplesweresequencedusingtheimmunoSEQassay(Adaptive Biotechnologies)utilizingthedeep-levelresolutiontoidentifyand quantitatetheTCR b chain( TRB ).Inbrief,thesomaticallyrearranged TRB wasampli edfrom159.36 – 1,200nggenomicDNAusing atwo-step,ampli cationbias-controlledmultiplexPCRapproach, andlibrariesweresequencedwithrawIlluminasequencereadsdemultiplexedandprocessedasdescribedpreviously.59,60Clonality wascalculatedaccordingtotheequationprovidedbyAdaptiveTechnologiesandtheirimmunoSEQsoftware:1 (entropy)/log2(#of productiveuniquereads). AdditionaldatafromperipheralbloodsamplesusedforTCRproductiveclonalitycomparisonsweretakenfromtheimmuneACCESS openaccessdatabase(AdaptiveBiotechnologies).Henderson etal.42isolated,byFACS,CD4+CD25+CD127lowTregsfromthe peripheralblood(PB)ofthreeyoungdonors,whoseagesranged from9.2 – 16.1years.Inthesecondstudyincluded,Emerson etal.61FACS-isolatedCD4+CD45RA+CD62L+(naive)andCD4+CD45RACD45RO+(memory)Tcellsfrom17APBsamples.Each studyalsousedtheimmunoSEQassay(AdaptiveBiotechnologies) for TRB sequencing.DataAnalysisStatisticalanalyseswereperformedusingPrism(v7.01,GraphPad). Unlessotherwisenoted,statisticalanalysiswasperformedusing two-tailedunpairedStudent ’ sttestwithWelsh ’ scorrection,onewayANOVAwithGeisser-GreenhousecorrectionandTukey ’ smultiplecomparisonstest,ortwo-wayANOVAwithTukey ’ smultiple comparisonstest,asnotedinthe gureslegends(wherep<0.05 wasconsideredstatisticallysigni cant;*p<0.05,**p<0.01,***p< 0.001,****p<0.0001).SUPPLEMENTALINFORMATIONSupplementalInformationincludestwo guresandonetableandcan befoundwiththisarticleonlineat http://dx.doi.org/10.1016/j.omtm. 2016.12.003 .AUTHORCONTRIBUTIONSH.R.S.researchedthedataandwrotethemanuscript.A.L.Putnam andJ.C.researchedthedataandreviewed/editedthemanuscript. A.L.Posgaicontributedtodiscussionsandwrotethemanuscript. E.H.R.,J.R.W.,K.F.G.,M.K.,A.L.,H.L.B.,K.S.B.,K.T.B.,L.P.,A.B., M.A.A.,andJ.A.B.contributedtodiscussionsandreviewed/edited themanuscript.M.J.H.conceivedthestudyandreviewed/editedthe manuscript.T.M.B.conceivedthestudyandwrotethemanuscript.CONFLICTSOFINTERESTK.F.G.andM.K.areemployeesofViaCord,LLC,asubsidiaryof PerkinElmer,and,assuch,arePerkinElmerstockshareholders. H.L.B.andK.S.B.areemployeesofCbrSystems,Inc.(CordBlood Registry),asubsidiaryofAMAGPharmaceuticals,and,assuch,are AMAGPharmaceuticalsstockshareholders.J.A.B.andA.L.Putnam areco-inventorsonpatents(US20080131445A1andUS7722862 B2) ledinconnectionwiththemanufacturingoftheTregproduct. J.A.B.hasreceivedfundingfromCaladriusBiosciencesandotherinkindcontributionsfromBDBiosciences.Theremainingauthors declarethattheyhavenocompetinginterests.ACKNOWLEDGMENTSWewouldliketothankDrs.JamesRileyandBruceLevine(University ofPennsylvania)forkindlyprovidingtheKT64/86(aAPC)cellline. WewouldalsoliketothankMichaelR.Lee(UCSF)forhistechnical contributionstowardthiseffort.Theseeffortsweresupportedbynonrestrictedresearchgrantsandkindgifts(ofCBunits)fromViaCord, LLC(asubsidiaryofPerkinElmer)andCBRSystems,Inc.(asubsidiaryofAMAGPharmaceuticals)(toM.J.H.andT.M.B).Additional projectsupportwasprovidedbyresearchgrantsfromJDRFinthe formofCordBloodCenterGrant4-2007-1065(toM.A.A.)and CareerDevelopmentAward2-2012-280(toT.M.B.),agrantfrom theMcJunkinFamilyFoundation(toM.J.H.),andNIHGrantsP01 AI42288(toM.A.A.andT.M.B.)andR01DK106191(toT.M.B).REFERENCES1. Bach,J.F.(2003).Autoimmunediseasesasthelossofactive “ self-control ” .Ann.NY Acad.Sci. 998 ,161 – 177 2. Thompson,J.A.,Perry,D.,andBrusko,T.M.(2012).AutologousregulatoryTcellsfor thetreatmentoftype1diabetes.Curr.Diab.Rep. 12 ,623 – 632 3. Trzonkowski,P.,Bacchetta,R.,Battaglia,M.,Berglund,D.,Bohnenkamp,H.R.,ten Brinke,A.,Bushell,A.,Cools,N.,Geissler,E.K.,Gregori,S.,etal.(2015).Hurdles intherapywithregulatoryTcells.Sci.Transl.Med. 7 ,304ps18 4. Riley,J.L.,June,C.H.,andBlazar,B.R.(2009).HumanTregulatorycelltherapy:takea billionorsoandcallmeinthemorning.Immunity 30 ,656 – 665 5. Gitelman,S.E.,andBluestone,J.A.(2016).RegulatoryTcelltherapyfortype1diabetes:Maytheforcebewithyou.J.Autoimmun. 71 ,78 – 87 .www.moleculartherapy.org MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017189

PAGE 13

6. Tang,Q.,andBluestone,J.A.(2008).TheFoxp3+regulatoryTcell:ajackofalltrades, masterofregulation.Nat.Immunol. 9 ,239 – 244 7. Nosbaum,A.,Prevel,N.,Truong,H.A.,Mehta,P.,Ettinger,M.,Scharschmidt,T.C., Ali,N.H.,Pauli,M.L.,Abbas,A.K.,andRosenblum,M.D.(2016).CuttingEdge: RegulatoryTCellsFacilitateCutaneousWoundHealing.J.Immunol. 196 ,2010 – 2014 8. Lu,B.,Yang,M.,andWang,Q.(2016).Interleukin-33intumorigenesis,tumorimmuneevasion,andcancerimmunotherapy.J.Mol.Med.(Berl.) 94 ,535 – 543 9. Villalta,S.A.,Rosenthal,W.,Martinez,L.,Kaur,A.,Sparwasser,T.,Tidball,J.G., Margeta,M.,Spencer,M.J.,andBluestone,J.A.(2014).RegulatoryTcellssuppress musclein ammationandinjuryinmusculardystrophy.Sci.Transl.Med. 6 258ra142 10. Rigby,M.R.,Harris,K.M.,Pinckney,A.,DiMeglio,L.A.,Rendell,M.S.,Felner,E.I., Dostou,J.M.,Gitelman,S.E.,Grif n,K.J.,Tsalikian,E.,etal.(2015).Alefaceptprovidessustainedclinicalandimmunologicaleffectsinnew-onsettype1diabetespatients.J.Clin.Invest. 125 ,3285 – 3296 11. Woodle,E.S.,Bluestone,J.A.,Zivin,R.A.,Jolliffe,L.K.,Auger,J.,Xu,D.,and Thistlethwaite,J.R.(1998).Humanized,nonmitogenicOKT3antibody,huOKT3 gamma(Ala-Ala):initialclinicalexperience.Transplant.Proc. 30 ,1369 – 1370 12. Haller,M.J.,Gitelman,S.E.,Gottlieb,P.A.,Michels,A.W.,Rosenthal,S.M.,Shuster, J.J.,Zou,B.,Brusko,T.M.,Hulme,M.A.,Wasserfall,C.H.,etal.(2015).Anti-thymocyteglobulin/G-CSFtreatmentpreserves b cellfunctioninpatientswithestablished type1diabetes.J.Clin.Invest. 125 ,448 – 455 13. Voltarelli,J.C.,Couri,C.E.,Stracieri,A.B.,Oliveira,M.C.,Moraes,D.A.,Pieroni,F., Coutinho,M.,Malmegrim,K.C.,Foss-Freitas,M.C.,Simes,B.P.,etal.(2007). Autologousnonmyeloablativehematopoieticstemcelltransplantationinnewlydiagnosedtype1diabetesmellitus.JAMA 297 ,1568 – 1576 14. Snarski,E.,Milczarczyk,A.,Torosian,T.,Paluszewska,M.,Urbanowska,E.,Krl,M., Boguradzki,P.,Jedynasty,K.,Franek,E.,andWiktor-Jedrzejczak,W.(2011). IndependenceofexogenousinsulinfollowingimmunoablationandstemcellreconstitutioninnewlydiagnoseddiabetestypeI.BoneMarrowTransplant. 46 ,562 – 566 15. Li,L.,Shen,S.,Ouyang,J.,Hu,Y.,Hu,L.,Cui,W.,Zhang,N.,Zhuge,Y.Z.,Chen,B., Xu,J.,andZhu,D.(2012).Autologoushematopoieticstemcelltransplantationmodulatesimmunocompetentcellsandimproves b -cellfunctioninChinesepatientswith newonsetoftype1diabetes.J.Clin.Endocrinol.Metab. 97 ,1729 – 1736 16. D ’ Addio,F.,ValderramaVasquez,A.,BenNasr,M.,Franek,E.,Zhu,D.,Li,L.,Ning, G.,Snarski,E.,andFiorina,P.(2014).Autologousnonmyeloablativehematopoietic stemcelltransplantationinnew-onsettype1diabetes:amulticenteranalysis. Diabetes 63 ,3041 –3046 17. Cant-Rodrguez,O.G.,Lavalle-Gonzlez,F.,Herrera-Rojas,M.A.,Jaime-Prez,J.C., Hawing-Zrate,J.A.,Gutirrez-Aguirre,C.H.,Mancias-Guerra,C.,Gonzlez-Llano, O.,Zapata-Garrido,A.,Villarreal-Prez,J.Z.,andGmez-Almaguer,D.(2016). Long-TermInsulinIndependenceinType1DiabetesMellitusUsingaSimpli ed AutologousStemCellTransplant.J.Clin.Endocrinol.Metab. 101 ,2141 – 2148 18. Klatzmann,D.,andAbbas,A.K.(2015).Thepromiseoflow-doseinterleukin-2therapyforautoimmuneandin ammatorydiseases.Nat.Rev.Immunol. 15 ,283 – 294 19. Hippen,K.L.,Merkel,S.C.,Schirm,D.K.,Nelson,C.,Tennis,N.C.,Riley,J.L.,June, C.H.,Miller,J.S.,Wagner,J.E.,andBlazar,B.R.(2011).Generationandlarge-scale expansionofhumaninducibleregulatoryTcellsthatsuppressgraft-versus-hostdisease.Am.J.Transplant. 11 ,1148 – 1157 20. Trzonkowski,P.,Bieniaszewska,M.,Ju sci nska,J.,Dobyszuk,A.,Krzystyniak,A., Marek,N.,My sliwska,J.,andHellmann,A.(2009).First-in-manclinicalresultsof thetreatmentofpatientswithgraftversushostdiseasewithhumanexvivoexpanded CD4+CD25+CD127-Tregulatorycells.Clin.Immunol. 133 ,22 – 26 21. DiIanni,M.,Falzetti,F.,Carotti,A.,Terenzi,A.,Castellino,F.,Bonifacio,E.,Del Papa,B.,Zei,T.,Ostini,R.I.,Cecchini,D.,etal.(2011).TregspreventGVHDandpromoteimmunereconstitutioninHLA-haploidenticaltransplantation.Blood 117 3921 – 3928 22. Marek-Trzonkowska,N.,My sliwiec,M.,Dobyszuk,A.,Grabowska,M.,Derkowska, I.,Ju sci nska,J.,Owczuk,R.,Szadkowska,A.,Witkowski,P.,M ynarski,W.,etal. (2014).Therapyoftype1diabeteswithCD4(+)CD25(high)CD127-regulatory Tcellsprolongssurvivalofpancreaticislets-resultsofoneyearfollow-up.Clin. Immunol. 153 ,23 – 30 23. Bluestone,J.A.,Buckner,J.H.,Fitch,M.,Gitelman,S.E.,Gupta,S.,Hellerstein,M.K., Herold,K.C.,Lares,A.,Lee,M.R.,Li,K.,etal.(2015).Type1diabetesimmunotherapy usingpolyclonalregulatoryTcells.Sci.Transl.Med. 7 ,315ra189 24. Gluckman,E.,Broxmeyer,H.A.,Auerbach,A.D.,Friedman,H.S.,Douglas,G.W., Devergie,A.,Esperou,H.,Thierry,D.,Socie,G.,Lehn,P.,etal.(1989). HematopoieticreconstitutioninapatientwithFanconi ’ sanemiabymeansofumbilical-cordbloodfromanHLA-identicalsibling.N.Engl.J.Med. 321 ,1174 – 1178 25. Wagner,J.E.,Kernan,N.A.,Steinbuch,M.,Broxmeyer,H.E.,andGluckman,E. (1995).Allogeneicsiblingumbilical-cord-bloodtransplantationinchildrenwithmalignantandnon-malignantdisease.Lancet 346 ,214 – 219 26. Warwick,R.,andArmitage,S.(2004).Cordbloodbanking.BestPract.Res.Clin.Obstet.Gynaecol. 18 ,995 – 1011 27. Hippen,K.L.,Harker-Murray,P.,Porter,S.B.,Merkel,S.C.,Londer,A.,Taylor,D.K., Bina,M.,Panoskaltsis-Mortari,A.,Rubinstein,P.,VanRooijen,N.,etal.(2008). UmbilicalcordbloodregulatoryT-cellexpansionandfunctionaleffectsoftumornecrosisfactorreceptorfamilymembersOX40and4-1BBexpressedonarti cialantigen-presentingcells.Blood 112 ,2847 – 2857 28. Brunstein,C.G.,Miller,J.S.,Cao,Q.,McKenna,D.H.,Hippen,K.L.,Curtsinger,J., Defor,T.,Levine,B.L.,June,C.H.,Rubinstein,P.,etal.(2011).Infusionofexvivo expandedTregulatorycellsinadultstransplantedwithumbilicalcordblood:safety pro leanddetectionkinetics.Blood 117 ,1061 – 1070 29. Brunstein,C.G.,Miller,J.S.,McKenna,D.H.,Hippen,K.L.,DeFor,T.E.,Sumstad,D., Curtsinger,J.,Verneris,M.R.,MacMillan,M.L.,Levine,B.L.,etal.(2016).Umbilical cordblood-derivedTregulatorycellstopreventGVHD:kinetics,toxicitypro le,and clinicaleffect.Blood 127 ,1044 – 1051 30. Haller,M.J.,Wasserfall,C.H.,Hulme,M.A.,Cintron,M.,Brusko,T.M.,McGrail, K.M.,Sumrall,T.M.,Wingard,J.R.,Theriaque,D.W.,Shuster,J.J.,etal.(2011). Autologousumbilicalcordbloodtransfusioninyoungchildrenwithtype1diabetes failstopreserveC-peptide.DiabetesCare 34 ,2567 – 2569 31. Haller,M.J.,Wasserfall,C.H.,Hulme,M.A.,Cintron,M.,Brusko,T.M.,McGrail, K.M.,Wingard,J.R.,Theriaque,D.W.,Shuster,J.J.,Ferguson,R.J.,etal.(2013). Autologousumbilicalcordbloodinfusionfollowedbyoraldocosahexaenoicacid andvitaminDsupplementationforC-peptidepreservationinchildrenwithType 1diabetes.Biol.BloodMarrowTransplant. 19 ,1126 – 1129 32. Haller,M.J.,Viener,H.-L.,Wasserfall,C.,Brusko,T.,Atkinson,M.A.,andSchatz, D.A.(2008).Autologousumbilicalcordbloodinfusionfortype1diabetes.Exp. Hematol. 36 ,710 – 715 33. Zhou,X.,Bailey-Bucktrout,S.,Jeker,L.T.,andBluestone,J.A.(2009).Plasticityof CD4(+)FoxP3(+)Tcells.Curr.Opin.Immunol. 21 ,281 – 285 34. Hoffmann,P.,Eder,R.,Boeld,T.J.,Doser,K.,Piseshka,B.,Andreesen,R.,and Edinger,M.(2006).OnlytheCD45RA+subpopulationofCD4+CD25highTcells givesrisetohomogeneousregulatoryT-celllinesuponinvitroexpansion.Blood 108 ,4260 – 4267 35. Woods,E.J.,Liu,J.,Pollok,K.,Hartwell,J.,Smith,F.O.,Williams,D.A.,Yoder,M.C., andCritser,J.K.(2003).Atheoreticallyoptimizedmethodforcordbloodstemcell cryopreservation.J.Hematother.StemCellRes. 12 ,341 – 350 36. Putnam,A.L.,Brusko,T.M.,Lee,M.R.,Liu,W.,Szot,G.L.,Ghosh,T.,Atkinson,M.A., andBluestone,J.A.(2009).ExpansionofhumanregulatoryT-cellsfrompatientswith type1diabetes.Diabetes 58 ,652 – 662 37. Wong,M.T.,Ong,D.E.H.,Lim,F.S.H.,Teng,K.W.W.,McGovern,N.,Narayanan,S., Ho,W.Q.,Cerny,D.,Tan,H.K.,Anicete,R.,etal.(2016).AHigh-DimensionalAtlas ofHumanTCellDiversityRevealsTissue-Speci cTraf ckingandCytokine Signatures.Immunity 45 ,442 – 456 38. Bluestone,J.A.,Buckner,J.H.,Fitch,M.,Gitelman,S.E.,Gupta,S.,Hellerstein,M.K., Herold,K.C.,Lares,A.,Lee,M.R.,Li,K.,etal.(2015).Type1diabetesimmunotherapy usingpolyclonalregulatoryTcells.Sci.Transl.Med. 7 ,315ra189 – 315ra189 39. Wing,J.B.,andSakaguchi,S.(2011).TCRdiversityandTregcells,sometimesmoreis more.Eur.J.Immunol. 41 ,3097 – 3100 40. Yang,S.,Fujikado,N.,Kolodin,D.,Benoist,C.,andMathis,D.(2015).Immunetolerance.RegulatoryTcellsgeneratedearlyinlifeplayadistinctroleinmaintainingselftolerance.Science 348 ,589 – 594 .MolecularTherapy:Methods&ClinicalDevelopment 190MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017

PAGE 14

41. Hughes,M.M.,Yassai,M.,Sedy,J.R.,Wehrly,T.D.,Huang,C.Y.,Kanagawa,O., Gorski,J.,andSleckman,B.P.(2003).TcellreceptorCDR3looplengthrepertoire isdeterminedprimarilybyfeaturesoftheV(D)Jrecombinationreaction.Eur.J. Immunol. 33 ,1568 – 1575 42. Henderson,L.A.,Volpi,S.,Frugoni,F.,Janssen,E.,Kim,S.,Sundel,R.P.,Dedeoglu,F., Lo,M.S.,Hazen,M.M.,BethSon,M.,etal.(2016).Next-GenerationSequencing RevealsRestrictionandClonotypicExpansionofTregCellsinJuvenileIdiopathic Arthritis.ArthritisRheumatol. 68 ,1758 – 1768 43. Busch,D.H.,Frle,S.P.,Sommermeyer,D.,Buchholz,V.R.,andRiddell,S.R.(2016). RoleofmemoryTcellsubsetsforadoptiveimmunotherapy.Semin.Immunol. 28 28 – 34 44. Horwitz,M.E.,Chao,N.J.,Rizzieri,D.A.,Long,G.D.,Sullivan,K.M.,Gasparetto,C., Chute,J.P.,Morris,A.,McDonald,C.,Waters-Pick,B.,etal.(2014).Umbilicalcord bloodexpansionwithnicotinamideprovideslong-termmultilineageengraftment. J.Clin.Invest. 124 ,3121 – 3128 45. Sa nia,N.,Vaikunthanathan,T.,Fraser,H.,Thirkell,S.,Lowe,K.,Blackmore,L., Whitehouse,G.,Martinez-Llordella,M.,Jassem,W.,Sanchez-Fueyo,A.,etal. (2016).SuccessfulexpansionoffunctionalandstableregulatoryTcellsforimmunotherapyinlivertransplantation.Oncotarget 7 ,7563 – 7577 46. Marek-Trzonkowska,N.,Mysliwiec,M.,Dobyszuk,A.,Grabowska,M.,Techmanska, I.,Juscinska,J.,Wujtewicz,M.A.,Witkowski,P.,Mlynarski,W.,Balcerska,A.,etal. (2012).AdministrationofCD4+CD25highCD127-regulatoryTcellspreserves b -cellfunctionintype1diabetesinchildren.DiabetesCare 35 ,1817 – 1820 47. Li,M.O.,andRudensky,A.Y.(2016).TcellreceptorsignallinginthecontrolofregulatoryTcelldifferentiationandfunction.Nat.Rev.Immunol. 16 ,220 – 233 48. Tang,Q.,Henriksen,K.J.,Bi,M.,Finger,E.B.,Szot,G.,Ye,J.,Masteller,E.L., McDevitt,H.,Bonyhadi,M.,andBluestone,J.A.(2004).Invitro-expandedantigen-speci cregulatoryTcellssuppressautoimmunediabetes.J.Exp.Med. 199 1455 – 1465 49. Cohen,C.J.,Gartner,J.J.,Horovitz-Fried,M.,Shamalov,K.,Trebska-McGowan,K., Bliskovsky,V.V.,Parkhurst,M.R.,Ankri,C.,Prickett,T.D.,Crystal,J.S.,etal. (2015).Isolationofneoantigen-speci cTcellsfromtumorandperipherallymphocytes.J.Clin.Invest. 125 ,3981 – 3991 50. BacaJones,C.,Pagni,P.P.,Fousteri,G.,Sachithanantham,S.,Dave,A.,RodriguezCalvo,T.,Miller,J.,andvonHerrath,M.(2014).RegulatoryTcellscontroldiabetes withoutcompromisingacuteanti-viraldefense.Clin.Immunol. 153 ,298 – 307 51. Butler,M.G.,andMenitove,J.E.(2011).Umbilicalcordbloodbanking:anupdate. J.Assist.Reprod.Genet. 28 ,669 – 676 .52. Lauber,S.,Latta,M.,Klter,H.,andMller-Steinhardt,M.(2010).TheMannheim cordbloodbank:experiencesandperspectivesforthefuture.Transfus.Med. Hemother. 37 ,90 – 97 53. Rosenau,E.H.,Sugrue,M.W.,Haller,M.,Fisk,D.,Kelly,S.S.,Chang,M.,Hou,W., Eldjerou,L.,Slayton,W.,Cogle,C.R.,andWingard,J.R.(2012).Characteristicsof thawedautologousumbilicalcordblood.Transfusion 52 ,2234 – 2242 54. Maus,M.V.,Thomas,A.K.,Leonard,D.G.,Allman,D.,Addya,K.,Schlienger,K., Riley,J.L.,andJune,C.H.(2002).Exvivoexpansionofpolyclonalandantigen-speci ccytotoxicTlymphocytesbyarti cialAPCsexpressingligandsfortheT-cellreceptor,CD28and4-1BB.Nat.Biotechnol. 20 ,143 – 148 55. Suhoski,M.M.,Golovina,T.N.,Aqui,N.A.,Tai,V.C.,Varela-Rohena,A.,Milone, M.C.,Carroll,R.G.,Riley,J.L.,andJune,C.H.(2007).Engineeringarti cialantigen-presentingcellstoexpressadiversearrayofco-stimulatorymolecules.Mol. Ther. 15 ,981 – 988 56. Klein,L.,andJovanovic,K.(2011).RegulatoryTcelllineagecommitmentinthe thymus. Seminarsinimmunology Volume23 (Elsevier),pp.401 – 409 57. Polansky,J.K.,Kretschmer,K.,Freyer,J.,Floess,S.,Garbe,A.,Baron,U.,Olek,S., Hamann,A.,vonBoehmer,H.,andHuehn,J.(2008).DNAmethylationcontrols Foxp3geneexpression.Eur.J.Immunol. 38 ,1654 – 1663 58. Fuhrman,C.A.,Yeh,W.I.,Seay,H.R.,SaikumarLakshmi,P.,Chopra,G.,Zhang,L., Perry,D.J.,McClymont,S.A.,Yadav,M.,Lopez,M.C.,etal.(2015).Divergent PhenotypesofHumanRegulatoryTCellsExpressingtheReceptorsTIGITand CD226.J.Immunol. 195 ,145 – 155 59. Carlson,C.S.,Emerson,R.O.,Sherwood,A.M.,Desmarais,C.,Chung,M.-W., Parsons,J.M.,Steen,M.S.,LaMadrid-Herrmannsfeldt,M.A.,Williamson,D.W., Livingston,R.J.,etal.(2013).UsingsynthetictemplatestodesignanunbiasedmultiplexPCRassay.Nat.Commun. 4 ,2680 60. Robins,H.S.,Campregher,P.V.,Srivastava,S.K.,Wacher,A.,Turtle,C.J.,Kahsai,O., Riddell,S.R.,Warren,E.H.,andCarlson,C.S.(2009).Comprehensiveassessmentof T-cellreceptor b -chaindiversityinalphabetaTcells.Blood 114 ,4099 – 4107 61. Emerson,R.,Sherwood,A.,Desmarais,C.,Malhotra,S.,Phippard,D.,andRobins,H. (2013).EstimatingtheratioofCD4+toCD8+Tcellsusinghigh-throughput sequencedata.J.Immunol.Methods 391 ,14 – 21 .www.moleculartherapy.org MolecularTherapy:Methods&ClinicalDevelopmentVol.4March2017191

PAGE 15

OMTM,Volume 4SupplementalInformation ExpansionofHumanTregsfromCryopreserved UmbilicalCordBloodforGMP-Compliant AutologousAdoptiveCellTransferTherapyHowardR.Seay,AmyL.Putnam,JuditCserny,AmandaL.Posgai,EmmaH. Rosenau,JohnR.Wingard,KateF.Girard,MoreyKraus,AngelaP.Lares,HeatherL. Brown,KatherineS.Brown,KristiT.Balavage,LeeanaD.Peters,AshleyN. Bushdorf,MarkA.Atkinson,JeffreyA.Bluestone,MichaelJ.Haller,andToddM.Brusko

PAGE 16

Supplement

PAGE 17

Supplementary Figure 1. Regulatory T cell (Tregs) recovery from fresh umbilical cord blood (CB) and cryopreserved umbilical cord blood (cryoCB). The number of Tregs isolated (A) per umbilical cord blood unit s ( CB U) was not significantly test, NS = 0.05), but (B) Treg per mL of cord was significantly less is cryoCB test; ** P < 0.01). (C) CD34 + cell c ount prior to cryopreservation was significantly correlated with Treg recovery from cryoCB units obtained from an academic CB bank [University of Florida Stem Cell Lab (UFSCL), black triangles] (* P < 0.05, r 2 = 0.823) but there was no correlation for cryoC B units obtained from either private bank institution (Private Bank 1 and 2, gray and open triangles, respectively) (NS = 0.05, r 2 = 0.028). (D) Total nucleated cell (TNC) count prior to cryopreservation and (E) duration of cryopreservation (d=days) we re not significantly correlated with Treg recovery from cryoCB (NS = 0.05; r 2 = 0.001 and r 2 = 0.009, respectively). (C E) Statistical tests applied were linear regression with Pearson correlation.

PAGE 19

Figure S1. Timeline for in vitro expansion of Tregs isolated from fresh or cryopreserved umbilical cord blood according to three unique protocols. Days (D 0 D 27 ) are indicated along the x axis, and the manipulations to the culture are indicated using symbols defined in the figure. Protoco l 1 follows a standard laboratory procedure for expanding Tregs for experimental purposes using art ificial antigen presenting cells ( aAPC ) loaded with anti CD3 (OKT3). Protocols 2 and 3 involve Treg stimulation with anti CD3 and anti CD28 conjugated beads.

PAGE 20

Supplementary Table 1. Culture Fl ask/Plate Minimum Suggested Maximum* 96 well, round bottom plate 25L 50 200L 360L 96 well, flat bottom plate 50L 100 200L 360L 48 well plate 100L 190 285L 1300L 24 well plate 200L 380 570L 2.6mL 12 well plate 400L 0.76 1.14mL 6mL 6 well plate 800L 1.9 2.9mL 15mL T 12.5 1mL 2.5 3.75mL 8mL T 12.5 (upright) 250L 500L 20mL T 25 2mL 5 7.5mL 15mL T 25 (upright) 500L 1mL 40mL T 75 8mL 15 22.5mL 75mL T 75 (upright) 2mL 4mL 175mL T 175 20mL 35 52.5mL 80mL T 175 (upright) 5mL 10mL 400mL T 225 28mL 45 67.5mL 125mL T 225 (upright) 7mL 14mL 1000mL *Some wide mouth flasks with have a lower maximum volume

PAGE 21

Table S1. Cryopreserved cord blood Treg (cryoCB Treg) culture plating volume recommendations. A key technical aspect of ensuring the expansion of CB Treg is plating the cells for stimulation and restimulation at 5.0 x 10 5 cells per mL of culture media in a vessel that will allows for doubling the volume at day 2 of culture without disturbing the cells. To that end, we made a reference table to assist those trying to reproduce the protocol.