aspartate3gp.com

ARTICLERe-designingInterleukin-12toenhanceitssafetyandpotentialasananti-tumorimmunotherapeuticagentPengjuWang1,XiaozhuLi2,JiweiWang1,DonglingGao1,YuenanLi1,HaozeLi1,YongchaoChu1,ZhongxianZhang1,HongtaoLiu1,GuozhongJiang1,ZhenguoCheng1,ShengdianWang2,JianzengDong3,BaisuiFeng1,4,LouisaS.
Chard5,NicholasR.
Lemoine1,5&YaoheWang1,5Interleukin-12(IL-12)hasemergedasoneofthemostpotentagentsforanti-tumorimmu-notherapy.
However,potentiallylethaltoxicityassociatedwithsystemicadministrationofIL-12precludesitsclinicalapplication.
Hereweredesignthemoleculeinsuchawaythatitsanti-tumorefcacyisnotcompromised,buttoxiceffectsareeliminated.
DeletionoftheN-terminalsignalpeptideofIL-12caneffectsuchachangebypreventingIL-12secretionfromcells.
Weuseanewlydesignedtumor-targetedoncolyticadenovirus(Ad-TD)todelivernon-secreting(ns)IL-12totumorcellsandexaminethetherapeuticandtoxiceffectsinSyrianhamstermodelsofpancreaticcancer(PaCa).
Strikingly,intraperitonealdeliveryofAd-TD-nsIL-12signicantlyenhancedsurvivalofanimalswithorthotopicPaCaandcuredperitoneallydisseminatedPaCawithnotoxicsideeffects,incontrasttothetreatmentwithAd-TDexpressingunmodiedIL-12.
ThesendingsofferrenewedhopefordevelopmentofIL-12-basedtreatmentsforcancer.
DOI:10.
1038/s41467-017-01385-8OPEN1Sino-BritishResearchCentreforMolecularOncology,NationalCentreforInternationalResearchinCellandGeneTherapy,SchoolofBasicMedicalSciences,AcademyofMedicalSciences,ZhengzhouUniversity,450052Zhengzhou,China.
2CASKeyLaboratoryofInfectionandImmunity,InstituteofBiophysics,ChineseAcademyofSciences,100101Beijing,China.
3DepartmentofCardiology,TheFirstAfliatedHospitalofZhengzhouUniversity,450052Zhengzhou,China.
4SectionofDigestiveDiseases,TheSecondAfliatedHospitalofZhengzhouUniversity,450014Zhengzhou,China.
5CentreforMolecularOncology,BartsCancerInstitute,QueenMaryUniversityofLondon,LondonEC1M6BQ,UK.
CorrespondenceandrequestsformaterialsshouldbeaddressedtoY.
W.
(email:yaohe.
wang@qmul.
ac.
uk)NATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunications11234567890Tumor-inducedimmunesuppressionisrecognizedasanimportantmechanismbywhichtumorsevadeimmune-mediateddetectionanddestruction1.
Anumberofstrate-giestoovercomethissuppressionhavebeenevaluated,butlocalIL-12expressionconsistentlyappearstobeoneofthemosteffectivemethodstoachievethisduetoitscentralroleinT-andNK-cell-mediatedinammatoryresponses2–5.
Unfortunately,clinicalapplicationofIL-12-basedtherapiesremainsproblematicduetothepotentialforrapiddevelopmentoflethalinammatorysyndrome6–10.
ThedevelopmentofstrategiestoovercomeIL-12-mediatedtoxicityiscurrentlythesubjectofintenseresearchandanumberofmodicationstoIL-12havebeenexplored.
Mostrecently,tumor-targetedoncolyticadenoviral(AdV)deliveryofmembrane-anchoredIL-12variantswasanalyzedinthecontextofefcacyagainstmetastaticpancreaticcancer11,12.
However,deliveryoftherapeuticallyeffectivedosesofAdVresultedinmembranesaturationofIL-12,leadingtoreleaseintotheserumandsubsequenttoxicity.
Morepromisingdrug-inducibleIL-12systemsalloweasiermanagementofIL-12levelsoverlongperiods,resultinginareasonabledegreeofclinicalefcacy.
However,inefcienttransductionoftumorcellswithcarriervectorsandthelackofsimultaneousinductionofinammationcurrentlylimitstheoverallanti-tumoreffectofthisapproach11,13.
Tumor-targetedoncolyticviruses(TOVs)areattractivether-apeuticcandidatesforcancertreatmentduetotheirabilitytoreplicateinanddirectlylysetumorcells,releasetumorantigensfromdestroyedcancercellsandimportantlyinducelocalinammation,whichcontributessignicantlytoreversaloflocalimmunesuppressionanddevelopmentofanti-tumorimmuneresponses14,15.
Furthermore,TOVscanbeusedtoefcientlydelivertherapeuticgenesspecicallytothetumorsiteatanincreasinglevelfollowingviralreplicationintumorcells.
Therst-generation,tumor-targetedoncolyticadenovirus,anE1B55k-deletedoncolyticadenovirus(H101)wastherstOVtherapytobelicensedforcancertreatment.
However,althoughclinicalsafetyproleswereencouraging,fewobjectiveresponseswereseen16,17.
IthassubsequentlybeenrecognizedthatdeletionsintheE1B55KandE3generegionsinthevirushadasignicantimpactontheabilityofthesevirusestoreplicateefcientlywithincells18.
BasedonourimprovedknowledgeofAdVbiology18–20,wehaveconstructedanew-generationreplicatingAdVwithtriplegenedeletions(E1ACR2,E1B19K,andE3gp19K),Ad-TD-LUC.
ThiswasusedtodeliveramodiedIL-12(nsIL-12,withdeletionoftheIL-12signalpeptide)toSyrianhamstermodelsofpan-creaticcancer(PaCa),whichareparticularlysuitablefortheseinvestigationsastheyarepermissiveforAdVreplication21,22andasshownhereforthersttime,permissiveforhumanIL-12functions.
OncolyticvirusesencodingIL-12havedemonstratedstronganti-tumoreffectsinpreclinicalmodelsofcancers23–25;however,systemicaccumulationofIL-12afterdeliverybyoncolyticvirusesremainspotentiallylethaltopatients10,26.
HerewereportthatsystemicdeliveryofthemodiednsIL-12usingouradenovirusAd-TD-nsIL-12toperitoneallydisseminatedandorthotopicpancreatictumorsisanextremelyeffectiveanti-tumortherapy.
Importantly,notoxicsideeffectsareobserved,evenwhenvirusesareadministeredathighdosesthatareusuallyassociatedwithlethalIL-12-mediatedtoxicityinthesemodels.
ResultsAd-TDreplicatesselectivelyincancercells.
Followingabetterunderstandingofthefunctionsofdifferentadenovirusgenes,wehaveconstructedanoveltumor-targetedreplicatingAdV,Ad-TD-LUC,inwhichtheE1ACR2,E1B19KandE3gp19Kgenesweredeletedandtheluciferase(LUC)openreadingframeinsertedintotheE3gp19Kregion(Fig.
1a).
Toanalyzeviralselectivityandreplicationintumorcells,weassessedviralrepli-cationinapanelofnormalandtumorcelllines(Fig.
1b–k).
Ad-TD-LUCreplicatedefcientlyinallcancercelllinesexamined(Fig.
1d–k),yetwassignicantlyattenuatedinnormalcelllines(Fig.
1b,c)incomparisontothewild-typeAd5virus,whichreplicatedtohightitersinthesecelllines.
Furthermore,Ad-TD-LUCalsodemonstratedasuperiorinvivoselectivity.
Tumortissues,lung,andliverwereexaminedafterintraperitonealadministrationofthevirusintoSyrianhamsterbearingHap-T1orthotopicpancreaticcancer.
WhiletheviralE1AgeneproducedbyAd5couldbedetectedinbothtumorandnormaltissuesatalltime-pointsexamined,E1AdetectedafterAd-TD-LUCtreatmentwasconnedtotumortissueandabsentfromnormallungandlivertissues(Fig.
1l).
Ad-TD-nsIL12replicatesinandiscytotoxictocancercells.
OncolyticvirusesencodingIL-12havedemonstratedstronganti-tumoreffectsinpreclinicalmodelsofcancers23–25;however,evendirectintratumoral(i.
t.
)administrationoftheseagentscanresultinsystemicaccumulationofIL-12,whichhasthepotentialtotriggerlethalinammatorysyndrome.
ToinvestigatethepotentialofIL-12modiedtopreventsecretionfrominfectedcells,Ad-TDvirusesexpressingwild-typeIL-12(Ad-TD-IL-12)ormodiedIL-12(Ad-TD-nsIL-12)wereconstructed(Fig.
1a).
ThepotencyofIL-12-armedviruses,controlvirusandwild-typeAd5wasassessedinapanelofhumancancercelllines(Fig.
2a)andapanelofSyrianhamstertumorcelllines(Fig.
2b).
ThethreemutantvirusesweresignicantlymorepotentthanAd5inallcelllinesexamined.
AsexpectedfrompreviousdataregardingtheabilityofoncolyticadenovirusestoreplicateefcientlyinSyrianhamstermodels22,wefoundthatAd5andourmutantvirusescouldreplicateinSyrianhamstertumorcelllineseffectivelyandatanequivalentleveloveratimecourse(Fig.
2c–g).
ModifyingIL-12doesnotaffectitsbiologicalactivity.
Havingconrmedthatthethreevirusesdevelopedinthisstudywereabletoselectivelyreplicateinandkilltumorcells,wenextinvestigatedtheintracellularaccumulationandsecretionofhumanIL-12afterinfectionofSyrianhamstertumorcelllineswithAd-TD-IL-12andAd-TD-nsIL-12.
AsshowninFig.
3,whilehighlevelsofIL-12weredetectedfromsupernatantsofAd-TD-IL-12-infectedcells,verylowlevelsofIL-12weredetectedfromsupernatantsofAd-TD-nsIL-12-infectedcells(Fig.
3a–f).
IL-12was,however,detectedinlysatesofAd-TD-nsIL-12-infectedcells(Fig.
3g,h),althoughatsignicantlylowerlevelswhencomparedtotheAd-TD-IL-12-infectedcells.
ThelowerlevelsofIL-12detectedfromAd-TD-nsIL-12-infectedlysatescorrespondedwithadecreaseinIL-12mRNAdetectedfrominfectedHap-T1cellsatboth72and96hpicomparedtoIL-12mRNAdetectedinAd-TD-IL-12-infectedcells(Fig.
3h,i).
NodifferencesinmRNAlevelsweredetectedatearlytimepoints,suggestingapossibilitythatintra-cellularaccumulationofIL-12mayacttodownregulateitsownexpressionatthemRNAlevelonceanintracellularthresholdisreached.
GiventhathumanIL-12isnon-functionalinmurinesystems,nextweinvestigatedwhetherhumanIL-12couldfunctionwithintheSyrianhamsterimmunesystem.
Hereweshow,forthersttime,thatrecombinanthumanIL-12isabletoinduceTNF-α(Fig.
3j)andIFN-γ(Fig.
3k)expressioninhamstersplenocytesexvivoandinduceproliferationofSyrianhamster,butnotmurine,peripheralbloodmononuclearcells(PBMCs)(Fig.
3l),conrmingthathumanIL-12isfunctionallyactiveinSyrianARTICLENATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-82NATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunicationshamstermodelsanddeletionofthesignalpeptidehasnodetrimentalimpactonitsactivity.
Ad-TD-nsIL-12iseffectiveinasubcutaneousPaCamodel.
Toinvestigatetheanti-tumorefcacyoftheAd-TD-LUCvirusandIL-12-armedAd-TDs,subcutaneousHPD1NRpancreatictumorswereestablishedinhamsters.
Animalsweretreatedi.
t.
sixtimeswith1*109PFU/injectionofAd-TD-IL-12,Ad-TD-nsIL-12,Ad-TD-LUCorPBS.
Multiplevirusadministrationwasusedbasedonpreviousclinicalstudiesandourrecentstudythatdemon-stratesthatrepeatedinjectionofoncolyticadenoviruscanresultinimprovedanti-tumorefcacythroughengagementofthehostL-ITRabcdefghikljE1ACR2E1B19kE3p19kR-ITRL-ITR100,000NHBE*************************Ad5Ad-TD-LUCNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSNSAd5liverAd5lungAd5tumorAd-TD-LUCliverAd-TD-LUClungAd-TD-LUCtumorHamsterhepatocytePaTu8988TSUIT2MiaPaCa2A549PC3MCF-7HCT116SKOV310,0001000PFU/cellPFU/cellPFU/cellPFU/cellPFU/cellPFU/cellE1Acopies/2mgtissueDNA(Log10)PFU/cell1001000100101PFU/cellPFU/cellPFU/cell100,00010,0001000100101,000,000100,00010,0001000100100,00010,00010001001010,00010001001011,000,000100,00010,00010001009876524h48hHourspostinfectionHourspostinfectionHourspostinfectionHourspostinfectionHourspostinfectionHourspostinfectionHourspostinfectionHourspostinfectionHourspostinfectionHourspostinfectionDaysposttreatment72h24h48h72h24h48h72h24h48h72h24h48h72h24h48h72h24h48h72h24h48h72h24h48h72hDay3Day7Day1424h48h72h100,00010,000100010010,000100010010110,0001000100101Ad5WTAd-TD-LUC/IL-12/nslL-12E1ACR2Delete922–945Delete1771–1914ReplacedbyLUCorIL-12ornslL-12E1B19kE3p19kR-ITRNATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-8ARTICLENATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunications3immunesystemagainstbothadenovirusandtumor-associatedantigens27.
Inthismodel,connementofIL-12expressiontothetumormasslimitedtheassociatedtoxicityresultingin100%tumoreradicationandsurvivalofbothAd-TD-IL-12andAd-TD-nsIL-12-treatedanimals(Fig.
4a,b).
Successfuloncolyticvirusstrategiesaimnotonlytoeradicatetheprimarytumor,butalsotoinducelong-termanti-tumorimmunity.
Toevaluatethedevelopmentofanti-tumorimmunityforprotectionagainstdiseaserecurrenceelicitedbyAd-TD-nsIL-12treatment,thoseanimalsthathadclearedtheirprimarytumoraftertreatmentwithAd-TD-nsIL-12werere-challengedfourweekslaterwithHPD1NRorcontrolrenalcancerHaKcells.
Ad-TD-nsIl-12-treatedanimalsdemonstrateddevelopmentofimmunitytoHPD1NRtumorcells,asevidencedbyrapidclearanceofthese,butnotHaKcellsnorHPD1NRco-injectedwithhamsterα-CD3depletionantibody(Fig.
4c).
Ad-TD-nsIL-12cansafelycuredisseminatedPaCa.
Toevaluatetheanti-tumoractivityofAd-TD-nsIL-12inadvancedPaCa,aperitoneallydisseminatedPaCamodelthatrapidlyrecapitulates(4–6days)late-stagehumanPaCa28wasestablishedinSyrianhamsters(SupplementaryFig.
1a).
SHPC6cellswereinjectedi.
p.
and4dayslater,PBSorvirusatadoseof1*109PFUwasadministeredi.
p.
ondays0,2,and4.
Ad-TD-nsIL-12treatmentresultedin100%survivalwhichpersisteduntilterminationoftheexperiment(Fig.
5a).
Noanimaldisplayedsignsoftreatment-relatedtoxicity.
TreatmentwithAd-TD-IL-12resultedinonlya10%survivalrate,withdeathsoccurringearlierthanthoseinthePBS-treatedgroup(Fig.
5a).
AdosereductionofAd-TD-IL-12didnotalleviateIL-12-mediatedtoxicity,resultinginearlydeathofalargeproportionoftheanimalstreatedandnoimprovementinoverallefcacycomparedtothePBStreatment(Fig.
5b).
Toevaluatetheimprovementinsafetyassociatedwithadministra-tionofAd-TD-nsIL-12comparedtoAd-TD-IL-12,liverfunctionwasexaminedafterPBS,Ad-TD-IL-12(5*108PFU)orAd-TD-nsIL-12(1*109)wereadministeredi.
p.
intohamstersbearingperitoneallydisseminatedSHPC6PaCa.
Livertoxicitywasassessedbymeasuringaspartateaminotransferase(AST),alanineaminotransferase(ALT)andalkalinephosphatase(ALP)levelsintheserumondays1,3,and5post-injection.
SignicantelevationsinallthreeenzymesweredetectedintheAd-TD-IL-12-treatedgroupateachtimepointwhereasliverenzymesmeasuredinAd-TD-nsIL-12-treatedanimals,whichreceiveddoublethedoseofvirus,remainedequivalenttothosedetectedinPBS-treatedani-mals(Fig.
5c).
Asexpected,serumlevelsofIL-12werehigheraftertreatmentwithAd-TD-IL-12comparedtoAd-TD-nsIL-12levels,whichremainedconstantthroughoutthetime-pointsanalyzed(Fig.
5d).
Afurtherexperimentwascarriedoutinwhichonei.
p.
injectionofhighdose(3*109PFU)Ad-TD-nsIL-12wasadministered.
Notreatment-relateddeathswereobservedandthetreatedanimalswerecompletelycured,survivinguntilterminationoftheexperiment(Fig.
5e).
Onceagain,inductionoflong-term,tumor-specicimmunitywasobservedafterre-challengeofhamsterspreviouslytreatedwithAd-TD-nsIL-12inthismodel(Fig.
5f).
Ad-TD-nsIL-12showsefcacyinanorthotopicPaCamodel.
Ad-TD-nsIL-12wasevaluatedfurtherusinganorthotopicPaCamodeltomimicaclinicallyunresectablediseasescenario.
ThehamsterHapT1PaCacelllinewaseffectiveforestablishingsizeable,metastasizingorthotopictumorsbyday6post-injection(SupplementaryFig.
1b,c)29.
HapT1-tumorsestablishedinthetailofthepancreasweretreatedi.
pwithPBS,Ad-TD-IL-12orAd-TD-nsIL-12(1*109PFU/injection)everyotherdayforsixdoses.
Inthismodel,treatmentwithAd-TD-nsIL-12signicantlyimprovedsurvivalcomparedwithPBSandAd-TD-IL-12-treatedanimals,thelattergroupexperiencingrapidmortalityduetoIL-12-mediatedtoxicity(Fig.
6a).
ThismodelwasusedtoassessanincreaseddoseofAd-TD-nsIL-12(2.
5*109PFU/injection,sixtimes)vs.
areduceddoseofAd-TD-IL-12(5*108PFU/injection,sixtimes).
TreatmentwithahigherdoseofAd-TD-nsIL-12dramaticallyincreasedthesurvivalratecomparedtoPBS,Ad-TD-LUCandAd-TD-IL-12-treatedanimals.
ReductionoftheAd-TD-IL-12doseadminis-teredalleviatedtoxiceffects,butefcacyremainedsignicantlylowerthanthatfollowingtreatmentwithAd-TD-nsIL-12(Fig.
6b).
Atthesedoses(Ad-TD-IL-12at5*108PFU/injectionandAd-TD-nsIL-12at2.
5*109PFU/injectiondeliveredsixtimesi.
p)wesawevidenceoftumorcellinfectionbyallvirusesbyimmunohistochemicaldetectionofviralproteinsinthetumorsectionsatday7followingthelastinjection(Fig.
6c)andbythetitrationofinfectiousvirionsrecovered(Fig.
6d),whichinallcaseswereabsentfromtumorsamplesbyday14.
AnalysisofviralE1ADNAproducedbyAd-TD-nsIL-12intumorornormaltissuesusingthismodelfurthersupportedthetumorspecicityofthensIL-12-armedadenovirusatadoseof2.
5*109PFU/injection,withE1Abeingdetectedatmuchlowerlevels(closetothelimitofassaysensitivity)innormallungorlivertissuecomparedtotumortissue(Fig.
6e).
AnalysisofthegrowthoforthotopicHap-T1tumorsfollowingthisregimedemonstratedthatbothAd-TD-IL12andAd-TD-nsIL-12causeddramaticregressionofthetumorsizeby14dayscomparedtoPBSandcontrolAd-TD-LUC(Fig.
6f),andbothviruses,attheirrespectivedoses,enhancedCD3+T-cellinltrationwithintumortissue(Fig.
6g).
AsobservedusingtheSHPC6modelofdisseminatedPaCa,systemicIL-12accumulationwasdetectedforthedurationofviralinfectiononlyaftertreatmentwithAd-TD-IL12(Fig.
6h),despitebothvirusesbeingdetectedatcomparablelevelsintumorcells(Fig.
6c,d).
Moreover,pathologicalexaminationofhepatocytesaftertreatmentwithAd-TD-LUCandAd-TD-nsIL-Fig.
1AnoveloncolyticadenovirusAd-TD-LUCistumor-selectiveinvitroandinvivo.
aThebasicstructuresoftheAd5genomeandderivedmutantsareshown.
Thearrowsindicatethedeletedregionsinthegenomeofadenovirus.
IL-12comprisesp40subunitandp35subunitlinkedbyanelasticpeptide(VPGVGVPGVG),withasignalpeptidelinkedtothep40subunit.
Non-secreting(ns)IL-12lacksthissignalpeptide.
bkTumorselectivitywasconrmedbyassessingreplicationofAd-TD-LUCandwildtypeAdVinnormalcellsandinapanelofhumancancercells.
Celllinesexaminedwerenormalhumanbronchialepithelialcells(NHBE);primaryculturehamsterhepatocytes;pancreaticcancercelllines(PaTu8988T;SUIT2;MiaPaCa2);lungcancer(A549);prostatecancer(PC3);colorectalcancer(HCT116);ovariancancer(SKOV3)andbreastcancer(MCF-7).
NHBECellswereinfectedatanMOIof100particles/cell,therestwereinfectedatanMOIof5PFU/cell.
Replicationassayswerecarriedoutoveraperiodof72h.
InfectiousvirusproductionwasassessedbytitrationonJH293cellsandthetiterasPFU/cellcalculatedandshownasmeanandstandarderrorofthemean(SEM).
Statisticalanalysiswasperformedusingatwo-wayANOVAtestwithBonferronipost-test;*p45cells/HPF.
ND:notdetected.
hIL-12expressioninserawasdetectedbyELISA.
iRepresentativehistopathologyofHEstainingofliversafterthreei.
pinjectionsofvirusintohamstersbearingorthotopicPaCatumors(n=3/group)onday0,2,and4,usingthesamedoseasinb.
LiverswerecollectedonedayfollowingthelastinjectionandanalyzedusingHEstaining(*200)ARTICLENATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-88NATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunicationsthesignalpeptiderequiredforsecretionfromcellsandusedouradenovirusAD-TDtodeliverthisdirectlyorsystemicallytoanumberofdifferenttumormodels.
WhenAd-TD-nsIL-12wasinjectedi.
pintohamstersbearingintraperitoneallydisseminatedSHPC6PaCa,onlylowlevelsofIL-12weredetectedinperipheralbloodonday1,3,5,andnoobviouschangesofALT,ASTandALPliverenzymesweredetected(Fig.
5).
DeliveryofunmodiedIL-12resultedinhighlevelsofperipheralIL-12andelevationsinliverenzymelevels,consistentwithsevereliverbloodvesselcongestion,eosinophilicdegenerationandhepatocyteapoptosisandnecrosis(Fig.
6),indicativeofinductionofsystemictoxicity.
BypreventingIL-12secretionfromtumorcells,nsIL-12,tumor-associatedantigensandnewAd-TD-nsIL-12virionsarereleasedonlywhenthetumorcellswerelysedafterviralinfection,limiting100acdeghifb806040200100806040200050100150200050100150200P=0.
017*PBSPBSPBSPBSPBSPBSPBSPBSAd-TD-LUCAd-TD-LUCAd-TD-LUCAd-TD-LUCAd-TD-LUCAd-TD-LUCAd-TD-LUCAd-TD-LUCAd-TD-LUC(2.
5*109)Ad-TD-IL-12Ad-TD-IL-12Ad-TD-IL-12Ad-TD-IL-12tumorAd-TD-nsIL-12tumorAd-TD-nsIL-12liverAd-TD-nsIL-12lungAd-TD-IL-12liverAd-TD-IL-12lungAd-TD-IL-12Ad-TD-IL-12Ad-TD-IL-12Ad-TD-IL-12Ad-TD-IL-12Ad-TD-IL-12(5*108)Ad-TD-nsIL-12Ad-TD-nsIL-12Ad-TD-nsIL-12Ad-TD-nsIL-12Ad-TD-nsIL-12Ad-TD-nsIL-12Ad-TD-nsIL-12Ad-TD-nsIL-12PBSAd-TD-LUCAd-TD-IL-12Ad-TD-nsIL-12Ad-TD-nsIL12(2.
5*109)DaysposttreatmentDaysposttreatmentDaysposttreatmentDaysposttreatmentDaysposttreatmentDaysposttreatmentDaysposttreatmentDaysposttreatmentSurvival(%)Survival(%)**P=0.
073P=0.
003100μm100μm100μm50403020100Hexon+cellsperHPFCD3+cellsperHPFDay3Day3Day3Day3Day3Day3Day7Day7Day7Day7Day7Day7Day14Day14Day14Day14Day14Day14NSNSNSNSNSNSNSNSNSNDNDNDNDNDNDVirusPFU/gtumor(Log10)E1Acopies/2mgtissueDNA(Log10)9876598765*******************************Tumorvolume(mm3)2500200015001000100050005432106004002000[IL-12](pg/ml)NATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-8ARTICLENATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunications9thedisseminationofIL-12tothelocaltumorenvironmentwhereeffectiveanti-tumorimmunitywaselicited(Fig.
6).
IthasbeenreportedthathumanIL-12doesnotstimulatemousePBMCproliferation42,butnostudyhasbeenpublishedassessingtheeffectofhumanIL-12onhamsterPBMCs.
PBMCproliferationassayspresentedheredemonstratethathumanIL-12stimulatesbothhumanandhamsterPBMCsproliferation,butnotmousePBMCs.
Inaddition,rhIL-12wasabletoeffectivelystimulateIFNγandTNFαexpressionbyactivatedsplenocytesexvivo(Fig.
3).
TheseresultssuggestSyrianhamstersasan4000abcdgef3000**********PBSAd-TD-nslL-12+mlgGAd-TD-nslL-12+α-CD3α-CD3PBSAd-TD-nslL-12+mlgGAd-TD-nslL-12+α-CD4α-CD42000Tumorvolume(mm3)10000400030002000Tumorvolume(mm3)10000020PBSAd-TD-LUCCD450μmQuantificationoflymphocyteinfiltrationCD3Ad-TD-nslL-12Ad-TD-lL-1240Daysposttreatment6002040Daysposttreatment605PBSAd-TD-LUCAd-TD-lL-12Ad-TD-nslL-12PBSAd-TD-LUCAd-TD-lL-12Ad-TD-nslL-12*****************************************4504030%totalLNcells%totalsplenocytes%totalsplenocytes%totalsplenocytes20100152520201510501510501050504030%totalLNcells20100504030%totalLNcells201003210Quantificationoflymphocyteinfiltration5******************43210Quantificationoflymphocyteinfiltration5************43210CD3Day7CD4CD3Day21CD4CD3CD4Day3draininglymphnodeDay3spleenDay7spleenDay21spleenDay7draininglymphnodeDay21draininglymphnodeCD4CD3+CD4–CD4CD3+CD4–CD4CD3+CD4–CD4CD3+CD4–CD4CD3+CD4–CD4CD3+CD4–Day3ARTICLENATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-810NATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunicationseffectivemodelforassessingtheanti-tumoreffectsofbothoncolyticadenovirusesandhumaninammatorycytokines.
GivenpreviousreportsoftheabilityofIL-12topromoteT-cellproliferationandprogrameffectorfunctionswithinthestablepopulationofhumanCD8+effectormemoryTcellsinvivo43,44,weexaminedtherelianceofourtreatmentontheseimmunesubsets.
Ad-TD-nsIL-12therapygeneratedrobustanti-tumormemoryTcellsasevidencedbyrejectionofsecondarytumors.
ThelackofsuitableresearchtoolsforanalysisofSyrianhamsterimmunesubsetshaslimitedthescopeofresearchpossibleinthismodel,butwehavedevelopedandvalidatedanumberofSyrianhamster-specicantibodiesandqPCRprimersthatallowinitialinvestigationoftheimmuneresponseselicitedbyourviruses.
DepletionofCD4+Tcellshadnoeffectontreatmentefcacy,whereasdepletionofCD3+cellpopulationshadanegativeimpactontreatmentefcacyinvivo(Fig.
7).
TakentogethertheseresultssuggestanimportanceofCD3+C4(CD8+)cellsfortreatmentefcacy.
Furtherinvestigationofimmunefunctions,usingFACSana-lysisandqPCRdetectionmethodsdevelopedinourlaboratoryrevealedthatAd-TD-nsIL-12therapyhadnoeffectonsplenicTcellsandonlytransientlyincreasedthepercentagesofCD4+andCD8+(CD3+CD4)populationsinlocaldraininglymphnodescomparedwiththecontrolvirus,suggestingthatanti-tumorefcacyofnsIL-12treatmentdependsmoreonmodica-tionofT-cellfunctionthanalterationsoftotalnumbersofcir-culatinglymphocytes(Fig.
7),althoughcurrentlythelackoftoolsavailableforanalysisofT-cellactivationstatusinSyrianhamstermodelsprecludesconrmationofthis.
IL-12effectsaremediatedlargelythroughIFN-γandindeed,tumorIFN-γlevelswereelevatedbybothAd-TD-IL-12andAd-TD-nsIL-12comparedtotreatmentwithAd-TD-LUCorPBS(Fig.
8).
However,Ad-TD-nsIL-12elevatedlocallevelsofIFN-γandotherinammatorycytokineslaterandtoalesserextentthanAd-TD-IL-12.
SplenicIFN-γandIP-10andlymphnodeIFN-γwerealsoproducedatlowerlevelsaftertreatmentwithAd-TD-nsIL-12comparedtoAd-TD-IL-12.
TheseobservationssuggestthattheimprovedsafetyassociatedwithnsIL-12deliverymaybeduetodampening,butnotabrogationoftheinammatoryresponsesgeneratedbyIL-12andthishypothesiswarrantsfurtherinvestigation.
TumorIL-10levelswerealsoelevatedmoresig-nicantlyaftertreatmentwithAd-TD-IL12comparedtoAd-TD-nsIL12orAd-TD-LUC.
AwealthofevidencenowexiststosuggestthatIL-10functionsasbothanimmune-suppressiveandimmune-stimulatingoranti-tumorcytokine,dependingonthecontextofitsexpression45–47,thustheroleofIL-12-inducedIL-10inthissettingrequiresfurtherstudy.
Inthisstudy,wehavedevelopedarobustandsafeIL-12basedimmune-viro-therapeuticagentthatcanbedeliveredlocallyorsystemicallytotumors.
Syrianhamstermodelsofpancreaticcancer,whichcanaccuratelyreectstagesoftumorigenesisinhumans,weretreatedeffectivelyandsafelyusingAd-TD-nsIL-12usingtreatmentregimescomparabletothosecurrentlyappliedclinically(Figs.
46).
However,furtherassessmentofthephar-macologyandtoxicologyofthisapproachisstillwarrantedpriortotranslationofthistherapyintoearlystageclinicaltrialsforpancreaticcancer.
Ofnote,thisagentmightbealsoeffectiveforotheradvancedsolidtumorssuchasgastric,colorectal,andovariancancersforwhichtheviruscanbeintraperitoneallyinjected.
Insummary,theapproachdevelopedinthisstudyoffersrenewedhopefordevelopmentofIL-12-basedtreatmentsforcancer.
MethodsStudydesign.
TheobjectiveofthisstudywastoinvestigatethefunctionalefcacyandsafetyofmodiedIL-12whendeliveredtoinvivotumormodelsusinganewoncolyticadenovirus.
AllhamsterprocedureswereapprovedbytheAnimalWelfareandResearchEthicsCommitteeofZhengzhouUniversity(Zhengzhou,China).
Forinvivoexperiments,powercalculationswerecarriedouttodeterminerequiredsamplesizesusingG*Power348,F-testANOVArepeatedmeasuressettingparametersofα=0.
1,power=90%,effectsize=0.
5,group=3.
Insubcutaneoustumormodels,animalswererandomlyassignedbyanindependentanimalcare-takertotreatmentgroupsbymatchingtumorsizespriortotreatment.
Tumorgrowthwasmeasuredusingelectroniccallipersuntilatumormeasured2.
0cmindiameterorulcerated,atwhichpointtheanimalwaskilled.
Tumorgrowthcurveswereterminateduponthedeathoftherstanimalineachgroup,butgroupsurvivalwasmonitoreduntiltheexperimentalendpoint(allanimalsingroupdeceased,or60daysposttreatment).
Fororthotopicanddisseminatedpancreaticcancermodels,animalswereassignedrandomlytotreatmentgroupsandanimalsurvivalwasmonitoredbyassessmentofanimalwell-beingeveryotherday.
Experimentalendpointsinthesemodelswere200daysposttreatment.
Animalcaretakerswereblindedtotreatmentgroupsinallcases.
Celllines.
TheSyrianhamsterpancreaticductaladenocarcinoma(PDAC)celllinesHPD1NR(maintainedinRoswellParkMemorialInstitute(RPMI)1640with10%FCS),Hap-T1,SHPC6,IPANandtherenaltumorcelllineHaKweremaintainedinDulbecco'smodiedEagle'smedium(DMEM)with10%FCS.
SHPC6andIPANwerekindlyprovidedbyW.
Wold(St.
LouisUniversity,St.
Louis,MO,USA).
HPD1NRandHap-T1celllineswerepurchasedfromtheGermanCollectionofMicroorganismsandCellCultures.
TheHaKcelllinewaspurchasedfromtheAmericanTypeCultureCollection(ATCC).
JH293,thehumankidneyepithelialcelllinetransformedwithAd5DNA,wasobtainedfromtheCancerResearchUKCentralCellServices(London,UnitedKingdom)andmaintainedinDMEMwith10%FCS.
Normalhumanbronchialepithelialcells(NHBE)wereobtainedfromCambrex(Cambridge,UK)andmaintainedinBronchialEpithelialGrowthMed-ium(BEGM)(Cambrex).
Primaryculturehamsterhepatocytes(maintainedinDMEMwith10%FCS)wereisolatedandculturedinourlab.
Allcellsweremaintainedat37°Cinahumidiedatmospherecontaining5%CO2andconrmedtobemycoplasma-freebeforebeingusedexperimentally.
Virusesandantibodies.
Adenovirustype5(Ad5)mutantsweregeneratedbyhomologousrecombinationasdescribedpreviously49.
ThecompleteAd5genomewasusedasthebackboneinallnewmutantsandwasderivedfromthepTG3602plasmid.
Thefollowingvirusesweregenerated:Ad-TD(tripledeletion:E1ACR2,E1B19K,andE3gp19K-deleted),Ad-TD-LUC(luciferase,frompGL3-BasicVectorFig.
7Ad-TD-nsIL-12efcacyisdependentonhamsterCD3+/CD4immunecellsubsets.
a,bSyrianhamsterswereinoculatedsubcutaneouslywith2*106HPD1NRcells.
Theestablishedtumorswereinjectedi.
t.
with1*109PFUAd-TD-nsIL-12orPBS(n=7/group)onday0,2,4,6,8,and10.
ControlIgGandeithermouseanti-hamsterCD3mAb(4F11)aorCD4mAbbwereinjectedi.
p.
atdosesof500μg/injectioneveryfourthdayfromthedaybeforetheviraltherapytotheendoftheexperimentandFACSanalysisusedtoconrmthedepletion.
MeantumorvolumesandSEMareshownforeachgroup.
Statisticalanalysiswascarriedoutusingaone-wayANOVAwithposthocTukey'sMultipleComparisonTest.
*p75cells/HPF.
Statisticalanalysiswascarriedoutusingaone-wayANOVAwithposthocTukey'sMultipleComparisonTest.
*p<0.
05,**p<0.
01,***p<0.
001.
gSpleensandlymphnodeswerecollectedandanalyzedbyFACSforCD3andCD4expressionatthetimepointsshown.
MeanexpressionandSEMisplotted(n=3/group).
Statisticalanalysiswascarriedoutusingaone-wayANOVAwithposthocTukey'sMultipleComparisonTest.
*p<0.
05,**p<0.
01,***p<0.
001NATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-8ARTICLENATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunications11(Promega),Ad-TD-IL-12(p40withsignalpeptideandp35withoutsignalpeptidefragmentswereclonedfromthecDNAderivedfromRPMI-8866cells,linkedwithelastincDNAbyPCR)andnsIL-12(non-secretedIL-12,p40,andp35withoutsignalpeptidefragmentswereclonedfromtheformerIL-12byPCR).
LUC,IL-12,andnsIL-12genesweredrivenbytheendogenousE3gp19kpromoter50.
Themousemonoclonalantibody(mAb)againstSyrianhamsterCD3e(clone4F11,IgG1isotype),mAbagainstmouseCD4(GK1.
5,cross-reactingwithSyrianham-ster)andanti-KLHmAbwerepreparedaspreviouslydescribed30,50.
Cytotoxicityassay.
Cellswereseededat2*103to4*103cellsperwellin96-wellplatesandinfectedwithviruses16–18hlateratstartingMOI=1000PFU/cell.
Cellsurvivalonday6afterviralinfectionwasdeterminedbyMTSassayasdetailedbythemanufacturer(Promega)andEC50values(viraldosekilling50%oftumorcells)werecalculatedaspreviouslydescribed18.
Allassayswereperformedatleastthreetimes.
Viralreplicationassay.
Cellswereseededat2*105cellsperwell,inthreewellsof6-wellplatesinmediumwith10%FCS,andinfectedat100particlesor5plaque-formingunits(PFU)/cellofAd5orAd-TD-LUC/IL-12/nsIL-121618hlater.
Primaryculturehamsterhepatocyteswereinfectedimmediatelyafterisolation.
Sampleswerecollectedintriplicateat24-hourintervalsupto72or96hafterinfection,freeze-thawedthreetimes,andtitratedonJH293cellstodeterminethe50%tissuecultureinfectivedose(TCID50)aspreviouslydescribed18.
TNF-αandIFN-γexpressioninducedbyIL-12.
Hamstersplenocyteswereobtainedfromhealthydonors.
Atotalof4*107freshsplenocyteswereseededina12-wellplateataconcentrationof2*106cells/mlandstimulatedwithPHAfor3days.
Theactivatedsplenocyteswereseededina12-wellplateatadensityof4*106cells/wellandincubatedwith100U/mlofhumanIL-2for24h.
TheactivatedsplenocyteswerewashedwithPBS,thenseededina12-wellplateatadensityof4*106cells/wellandincubatedwithhumanrIL-12for24or48hat37°C.
ThemRNAlevelsofTNF-αandIFN-γweredetectedbyqPCR.
PBMCproliferationassay.
Human,hamster,andmouseperipheralbloodwasobtainedfromhealthydonors.
PBMCswereseparatedonaFicoll-Hypaquedensitygradient.
Atotalof8*106freshPBMCswereseededina6-wellplateatacon-centrationof2*106cells/mlandstimulatedwithPHAfor3days.
TheactivatedPBMCswereseededina96-wellplateatadensityof1*105cells/wellinanalvolumeof50landincubatedwith50lofthelysateofHap-T1cellsinfectedwithAd-TD-IL-12orAd-TD-nsIL-12for48hat37°C.
PBMCproliferationwasdeterminedbyMTSassay.
rIL-12(2ng/ml)wasusedasapositivecontrol.
Sampleswereset-upintriplicate.
ELISA.
Celllysate,supernatantandserumsamplesweregeneratedandprocessedasdescribedinothersectionsofthemethods.
IL-12levelswerequantiedusinganIL-12-specicELISA(eBioscience)accordingtothemanufacturer'sinstructionsand1:10–200dilutionswereused,withsamplestestedintriplicate.
Real-timePCR.
FreshtissueswerehomogenizedandtotalRNAextractedusingTrizol(Invitrogen).
cDNAwassynthesizedbyreversetranscription(Promega).
qPCRwascarriedoutusingtheBio-RadCFXReal-TimePCRSystemandtheSYBRGreenassay(Bio-Rad)toquantifycytokineexpression.
PrimersofSyrianhamsterTNF-α,IFN-γ,IP10,CD83,IL-1β,IL-2,IL-6,IL-10,IL-12,IL-21,andβ-actingenes(Table1)weredesignedusingPrimerPremier5.
0software(Premier)andconstructedbySigma-Aldrich.
DatawereanalyzedusingBio-RadCFXManager(Bio-Rad).
TissueDNAwasextractedusingtheDNeasybloodandtissuekit(Qiagen).
qPCRwascarriedoutusingtheABISTEPONEPLUSsystemandtheSYBRGreenassay(TAKARA)todetectviralgenomeintissues.
TheprimersofadenovirusE1AgeneareE1A-Forward:5′-TGATCGATCCACCCAGTGAC-3′andE1A-Reverse:5′-ATGACAAGACCTGCAACCGT-3′.
DatawereanalyzedusingABISTEPONEPLUSsystemsoftware.
ResultswerenormalizedtoPBS-treatedgroupvalues.
Histopathologicalexaminationandimmunohistochemistry.
Thetissuescol-lectedatdifferenttimepointswereprocessedandstainedbyH&Estainingorimmunohistochemistry(IHC)forCD3andCD4aspreviouslydescribed18.
FACSanalysis.
Spleensanddraininglymphnodeswereextractedfromhamsters,combinedwithcompleteT-cellmedium(RPMImedium,10%FCS,1%penicillin/streptomycin,1%sodiumpyruvate)thenpushedthrougha70mcellstrainertocreateasinglecellsuspension.
Cellswerecentrifugedandthepelletincubatedin5mlredbloodcelllysisbuffer(Sigma-Aldrich).
Splenocytesandlymphnodecells(1*106)werepreparedandstainedwithmAbagainstSyrianhamsterCD3andmAbagainstmouseCD4-conjugatedwithFITC(GK1.
5,cross-reactivewithSyrianhamster,BD),followedbysecondaryAPC-conjugatedrabbitanti-mouseantibodyforCD3staining(BD).
CellswereacquiredonaFACSscanner,anddatawereanalyzedusingFlowJosoftware(TreeStarInc).
Assessmentofinfectiousvirus.
FreshtumortissueswerehomogenizedandtitratedonJH293cellstodeterminethe50%tissuecultureinfectivedose(TCID50)aspreviouslydescribed18.
ViralDNAlevelswereevaluatedintumorsandliversusingreal-timePCRasdescribedaboveandpreviously30.
Invivoanimalstudies.
Subcutaneoustumormodel:Atotalof2*106HPD-1NRcellswereimplantedsubcutaneouslyintotherightankoffemale,4-weekto5-week-oldSyrianhamsters.
Whentumorsreached68mmindiameter,hamsterswerestratiedintogroupsofsevenanimalstoreceive50or100lintratumoral(i.
t)injectionsofAd-TD-LUC,Ad-TD-IL-12,Ad-TD-nsIL-12,orPBSasspeciedinthegurelegends.
Theinjectionswereintroducedthroughasinglecentraltumorpuncturesite,andthreetofourneedletracksweremaderadiallyfromthecenterwhileviruswasinjectedastheneedlewaswithdrawn.
Tumorvolumeswereesti-mated(volume=(length*width2*π)/6)twiceweeklyuntiltumorsreached2.
0cmindiameterortumorulcerationoccurred.
Forbiologicaltime-pointexperimentstoinvestigatefunctionalmechanisms,hamsterswerestratiedintogroupswhentumorsreached67mmindiameterandtreatedonceonday0.
Onday3,7,and21,tumors,lymphnodes,andspleenswerecollectedfromthreeanimalsineachgrouptoinvestigatelymphocytepopulations,immunohistochemicalstainingforCD3andCD4expressionintumortissuesandtumor-specicimmunity.
Orthotopicpancreaticcancermodel.
4-weekto5-week-oldSyrianhamsterswereanaesthetizedwith10%chloralhydratebyanintraperitoneal(i.
p)injection.
Hamsterswereplacedinthedorsaldecubitusposition,andaleftsubcostalincisionwasmade.
Thepancreaswascarefullyexposed,and50loftumorcellsuspension(6*107cells/ml)wasinjectedintothetailofthepancreaswitha25-gaugeneedle.
Atechnicallysuccessfulinjectionwascharacterizedbytheformationofavisiblebubblewithinthepancreaticparenchyma.
Theneedlewasslowlywithdrawntoavoidmacroscopiccellspreadfromtheinjectionsite.
Thepancreaswasthenreturnedtotheperitonealcavity,andtheabdominalwallwasclosedintwolayerswithnylonsutures.
Sixdayslater,thetumorvolumesreachedto4–5mmindiameter,andsevenhamsterspergroupwereeachinjectedi.
pwith500lPBS,1*109PFUAd-TD-LUCor5*108PFUAd-TD-IL-12or1*109PFUAd-TD-nsIL-12ondays0,2,4,6,8,and10.
Thesurvivalofhamsterswasmonitored.
Intraperitoneallydisseminatedpancreaticcancermodel.
Atotalof1*107SHPC6cellswereseededintothelowerrightperitonealcavityofSyrianhamsters.
Table1qPCRprimersdesignedtoamplifymRNAofSyrianhamsterimmunemodulatorsGeneForwardReverseIFN-γ5′-TGTTGCTCTGCCTCACTCAG-3′5′-CACCAGCCTTTTGCCAGTTC-3′IP105′-GACCGACCGGTAAAACCGAG-3′5′-CACGTGGGCAGGATTGACTT-3′CD835′-CCCAGAGCAGGCAAAACAAC-3′5′-TTCCTGAAAGGTGACTCGGC-3′IL1b5′-CGTGGACCTTCCAGGATGAG-3′5′-AGCTGTCGAATGGGAGCATC-3′IL-25′-CTCGCATCCTGTCTTGCACT-3′5′-AGCATCATGGGGAGTTTCGG-3′IL-65′-ATAGTCACGCCTAGCCCAAC-3′5′-TCTTGGTTCTTGGCCACTCC-3′IL-105′-AGTAACTGCACCCACTTCCC-3′5′-TGGCAACCCAAGTAACCCTT-3′IL-125′-AGGCTCTGAATCTCAACGGC-3′5′-GATTGTCACAGCACGGATGC-3′IL-155′-CGGGTCATTTTGCACGAGTA-3′5′-CCTTGCAGCCATGTTCTGTT-3′IL-215′-ACGCTCAGCTTTTGCCTGTT-3′5′-GCTCTTCTTCTGCCTTCTCGT-3′TNF-α5′-TTCTCCTTCCTGCTTGTGGC-3′5′-CAGGCTTGTCGTTCGAATTTTG-3′β-actin5′-AGATGACCCAGATCATGTTTGAGA-3′5′-CAGGATGGCATGAGGGAGAG-3′ARTICLENATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-812NATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunicationsFourdayslater,10hamsterspergroupwereinjectedi.
pwith500lPBS,1*109PFUAd-TD-LUCor5*108PFUAd-TD-IL-12or1*109PFUAd-TD-nsIL-12ondays0,2,4.
Alternativedosesweregivenasindicatedinthetextorgurelegend.
Thesurvivalofhamsterswasmonitored.
Hepatotoxicityevaluation.
Threehamsterspergrouppertimepointwereinjectedi.
pwith1*107SHPC6cells.
Fourdayslater,ninehamsterspergroupwereeachinjectedi.
pwith500lPBS,1*109PFUAd-TD-LUCor5*108PFUAd-TD-IL-12or1*109PFUAd-TD-nsIL-12ondays0.
Bloodwascollectedfromhamstersfromtheretro-orbitalsinus,andtheserawereanalyzedfortransaminaselevelsandIL-12expressiononday1,3,5.
Re-challengeoftumorfreeanimals.
ThehamstersthatunderwentcompletesubcutaneoustumorregressionfollowingAd-TD-IL-12orAd-TD-nsIL-12treat-mentwerere-challengedwith4*106HPD-1NRcells(twicethenumberofcellscomparedtotheprimarytumorcellinoculation)or5*106HaKcellsafterprimarytumorshadnotbeendetectedfor1month.
Onegroupofthecuredanimalswereinjectedi.
pwithanti-CD3mAbonthedaybeforethere-challengewith4*106HPD-1NRcells.
Tumorvolumesweremeasuredtwiceweekly.
SHCP6cellre-challengeisdescribedinthegurelegend.
CD3andCD4depletioninvivo.
Atotalof2*106HPD-1NRcellswereimplantedsubcutaneouslyinto4-weekto5-week-oldSyrianhamsters.
Whentumorsreached300mm3,hamstersweredistributedbetweenthetreatmentandthecontrolgroupsbymatchedtumorsizetoreceivei.
tinjectionsof1*109PFUAd-TD-nsIL-12orPBSonday0,2,4,6,8,and10.
DepletionmAbagainstSyrianhamsterCD3(clone4F11),mAbagainstmouseCD4orcontrolIg(mouseanti-KLHmAb)wereadministeredi.
pwithdosesof500μg/injectioneveryfourthdayfromthedaybeforethevirotherapytotheendoftheexperimentandFACSanalysisusedtoconrmdepletion.
Tumorvolumesweremeasuredasintheefcacyexperiments.
Statisticalanalysis.
StatisticalanalysiswascarriedoutusingGraphPadPrism5andSPSS19.
0software.
Theresultswererepresentedasmean±standardordeviation(SD)or±standarderrorofthemean(SEM).
Differencesbetween2015105543210RelativeexpressionlevelsRelativeexpressionlevelsRelativeexpressionlevelsRelativeexpressionlevelsD3D7D21TumorIFN-γPBSAd-TD-LUCAd-TD-nslL-12Ad-TD-lL-12TumorIP-102015105543210D3D7D21***********************15105543210TumorCD83D3D7D21**************************************15105543210151055432101510554321015105543210D3D7D21TumorIL-1β1206510543210TumorIL-2TumorIL-10TumorIL-12SpleenIP-10TumorIL-21TumorIL-6D3D7D21252015105543210D3D7D21D3D7D219050101086420D3D7D21150100501086420D3D7D21SpleenIFN-γ20D3D7D21DLNIFN-γD3D3D7D7D21D219630Fig.
8Ad-TD-nsIL-12efcacyismediatedbyproductionofinammatorymediators.
2*106HPD1NRcellswereseededintotherightankofSyrianhamsters.
Whentumorvolumesreached300mm3,ninehamsterspergroupwereeachinjectedi.
twith100lPBS,1*109PFUAd-TD-LUCorAd-TD-IL-12/nsIL-12onday0.
Spleens,draininglymphnodes(DLN)andtumorswerecollectedandanalyzedbyqPCRtoanalyzeexpressionlevelsofCD83,IFN-γ,IP10,IL-1β,IL-2,IL-6,IL-10,IL-12,andIL-21attheindicatedtimepoints.
AllexperimentswereperformedintriplicateandthemeanandSEMareshownforeachgroupcomparedusinganindependentt-test.
*p<0.
05,**p<0.
01,***p<0.
001NATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-8ARTICLENATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunications13groupswereanalyzedusingtheStudent'st-test,one-wayANOVAtestorKaplan–Meiersurvivalanalysis.
Differenceswereconsideredstatisticallysignicantwhenthep-valuewaslessthan0.
05.
Dataavailability.
Alldatapresentedareavailablefromtheauthorsonrequest.
Received:11January2017Accepted:14September2017References1.
Pardoll,D.
M.
Theblockadeofimmunecheckpointsincancerimmunotherapy.
Nat.
Rev.
Cancer12,252–264(2012).
2.
Manetti,R.
etal.
Naturalkillercellstimulatoryfactor(interleukin12[IL-12])inducesThelpertype1(Th1)-specicimmuneresponsesandinhibitsthedevelopmentofIL-4-producingThcells.
J.
Exp.
Med.
177,1199–1204(1993).
3.
Heinzerling,L.
etal.
TumorregressionofhumanandmurinemelanomaafterintratumoralinjectionofIL-12-encodingplasmidDNAinmice.
Exp.
Dermatol.
11,232–240(2002).
4.
Pan,W.
Y.
etal.
Cancerimmunotherapyusingamembrane-boundinterleukin-12withB7-1transmembraneandcytoplasmicdomains.
Mol.
Ther.
20,927–937(2012).
5.
DelVecchio,M.
etal.
Interleukin-12:biologicalpropertiesandclinicalapplication.
Clin.
CancerRes.
13,4677–4685(2007).
6.
Motzer,R.
J.
etal.
PhaseItrialofsubcutaneousrecombinanthumaninterleukin-12inpatientswithadvancedrenalcellcarcinoma.
Clin.
CancerRes.
4,1183–1191(1998).
7.
Sangro,B.
etal.
PhaseItrialofintratumoralinjectionofanadenovirusencodinginterleukin-12foradvanceddigestivetumors.
J.
Clin.
Oncol.
22,1389–1397(2004).
8.
Cohen,J.
IL-12deaths:explanationandapuzzle.
Science270,908(1995).
9.
Bortolanza,S.
etal.
Treatmentofpancreaticcancerwithanoncolyticadenovirusexpressinginterleukin-12inSyrianhamsters.
Mol.
Ther.
17,614–622(2009).
10.
Leonard,J.
P.
etal.
Effectsofsingle-doseinterleukin-12exposureoninterleukin-12-associatedtoxicityandinterferon-gammaproduction.
Blood90,2541–2548(1997).
11.
Poutou,J.
etal.
Safetyandantitumoreffectofoncolyticandhelper-dependentadenovirusesexpressinginterleukin-12variantsinahamsterpancreaticcancermodel.
GeneTher.
22,696–706(2015).
12.
Nagarajan,S.
&Selvaraj,P.
Glycolipid-anchoredIL-12expressedontumorcellsurfaceinducesantitumorimmuneresponse.
CancerRes.
62,2869–2874(2002).
13.
Gonzalez-Aparicio,M.
etal.
Oxaliplatinincombinationwithliver-specicexpressionofinterleukin12reducestheimmunosuppressivemicroenvironmentoftumoursanderadicatesmetastaticcolorectalcancerinmice.
Gut.
60,341–349(2011).
14.
Tong,A.
W.
etal.
Oncolyticvirusesforinductionofanti-tumorimmunity.
Curr.
Pharm.
Biotechnol.
13,1750–1760(2012).
15.
Kaufman,H.
L.
,Kohlhapp,F.
J.
&Zloza,A.
Oncolyticviruses:anewclassofimmunotherapydrugs.
Nat.
Rev.
DrugDiscov.
14,642–662(2015).
16.
Mulvihill,S.
etal.
SafetyandfeasibilityofinjectionwithanE1B-55kDagene-deleted,replication-selectiveadenovirus(ONYX-015)intoprimarycarcinomasofthepancreas:aphaseItrial.
GeneTher.
8,308–315(2001).
17.
Hecht,J.
R.
etal.
AphaseI/IItrialofintratumoralendoscopicultrasoundinjectionofONYX-015withintravenousgemcitabineinunresectablepancreaticcarcinoma.
Clin.
CancerRes.
9,555–561(2003).
18.
Wang,Y.
etal.
E3genemanipulationsaffectoncolyticadenovirusactivityinimmunocompetenttumormodels.
Nat.
Biotechnol.
21,1328–1335(2003).
19.
Spurrell,E.
etal.
STAT1interactionwithE3-14.
7Kinmonocytesaffectstheefcacyofoncolyticadenovirus.
J.
Virol.
88,2291–2300(2014).
20.
Liu,T.
C.
etal.
AnE1B-19kDagenedeletionmutantadenovirusdemonstratestumornecrosisfactor-enhancedcancerselectivityandenhancedoncolyticpotency.
Mol.
Ther.
9,786–803(2004).
21.
Thomas,M.
A.
etal.
Syrianhamsterasapermissiveimmunocompetentanimalmodelforthestudyofoncolyticadenovirusvectors.
CancerRes.
66,1270–1276(2006).
22.
Bortolanza,S.
etal.
HumanadenovirusreplicatesinimmunocompetentmodelsofpancreaticcancerinSyrianhamsters.
Hum.
GeneTher.
18,681–690(2007).
23.
Tsai,Y.
S.
etal.
EnhancementofantitumoractivityofgammaretroviruscarryingIL-12genethroughgeneticmodicationofenvelopetargetingHER2receptor:apromisingstrategyforbladdercancertherapy.
CancerGeneTher.
17,37–48(2010).
24.
Roth,J.
C.
etal.
EvaluationofthesafetyandbiodistributionofM032,anattenuatedherpessimplexvirustype1expressinghIL-12,afterintracerebraladministrationtoaotusnonhumanprimates.
Hum.
GeneTher.
Clin.
Dev.
25,16–27(2014).
25.
Pan,W.
Y.
etal.
Cancerimmunotherapyusingamembrane-boundInterleukin-12withB7-1transmembraneandcytoplasmicdomains.
Mol.
Ther.
:J.
Am.
Soc.
GeneTher.
20,927-937(2012).
26.
Bortolanza,S.
etal.
Treatmentofpancreaticcancerwithanoncolyticadenovirusexpressinginterleukin-12inSyrianhamsters.
Mol.
Ther.
17,614–622(2009).
27.
Li,X.
etal.
TheefcacyofoncolyticadenovirusismediatedbyT-cellresponsesagainstvirusandtumorinSyrianhamstermodel.
Clin.
CancerRes.
23,239–249(2017).
28.
Spencer,J.
F.
,Sagartz,J.
E.
,Wold,W.
S.
&Toth,K.
NewpancreaticcarcinomamodelforstudyingoncolyticadenovirusesinthepermissiveSyrianhamster.
CancerGeneTher.
16,912–922(2009).
29.
Abraham,A.
T.
,Shah,S.
R.
&Davidson,B.
R.
TheHaP-T1Syriangoldenhamsterpancreaticcancermodel:cellimplantationisbetterthantissueimplantation.
Pancreas29,320–323(2004).
30.
Tysome,J.
R.
etal.
Anoveltherapeuticregimentoeradicateestablishedsolidtumorswithaneffectiveinductionoftumor-specicimmunity.
Clin.
CancerRes.
18,6679–6689(2012).
31.
Coughlin,C.
M.
etal.
TumorcellresponsestoIFNgammaaffecttumorigenicityandresponsetoIL-12therapyandantiangiogenesis.
Immunity9,25–34(1998).
32.
Bellone,G.
etal.
Cytokineexpressionproleinhumanpancreaticcarcinomacellsandinsurgicalspecimens:implicationsforsurvival.
CancerImmunol.
,Immunother.
55,684–698(2006).
33.
Miao,J.
W.
,Liu,L.
J.
&Huang,J.
Interleukin-6-inducedepithelial-mesenchymaltransitionthroughsignaltransducerandactivatoroftranscription3inhumancervicalcarcinoma.
Int.
J.
Oncol.
45,165–176(2014).
34.
Heise,C.
&Kirn,D.
H.
Replication-selectiveadenovirusesasoncolyticagents.
J.
Clin.
Invest.
105,847–851(2000).
35.
Kirn,D.
Clinicalresearchresultswithdl1520(Onyx-015),areplication-selectiveadenovirusforthetreatmentofcancer:whathavewelearnedGene.
Ther.
8,89–98(2001).
36.
O'Shea,C.
C.
,Choi,S.
,McCormick,F.
&Stokoe,D.
Adenovirusoverridescellularcheckpointsforproteintranslation.
CellCycle4,883–888(2005).
37.
Fulci,G.
etal.
Depletionofperipheralmacrophagesandbrainmicrogliaincreasesbraintumortitersofoncolyticviruses.
CancerRes.
67,9398–9406(2007).
38.
Oberg,D.
etal.
ImprovedpotencyandselectivityofanoncolyticE1ACR2andE1B19Kdeletedadenoviralmutantinprostateandpancreaticcancers.
Clin.
CancerRes.
16,541553(2010).
39.
Cao,H.
etal.
[AntitumourefcacyofanoveloncolyticadenovirusAd-TD-RFPforhumannasopharyngealcarcinoma].
ZhonghuaErBiYanHouTouJingWaiKeZaZhi47,672–676(2012).
40.
Brunda,M.
J.
etal.
Antitumorandantimetastaticactivityofinterleukin12againstmurinetumors.
J.
Exp.
Med.
178,1223–1230(1993).
41.
Portielje,J.
E.
etal.
Repeatedadministrationsofinterleukin(IL)-12areassociatedwithpersistentlyelevatedplasmalevelsofIL-10anddecliningIFN-gamma,tumornecrosisfactor-alpha,IL-6,andIL-8responses.
Clin.
CancerRes.
9,76–83(2003).
42.
Schoenhaut,D.
S.
etal.
CloningandexpressionofmurineIL-12.
J.
Immunol.
148,3433–3440(1992).
43.
Hsieh,C.
S.
etal.
DevelopmentofTH1CD4+TcellsthroughIL-12producedbyListeria-inducedmacrophages.
Science260,547–549(1993).
44.
Chowdhury,F.
Z.
,Ramos,H.
J.
,Davis,L.
S.
,Forman,J.
&Farrar,J.
D.
IL-12selectivelyprogramseffectorpathwaysthatarestablyexpressedinhumanCD8+effectormemoryTcellsinvivo.
Blood118,3890–3900(2011).
45.
D'Andrea,A.
etal.
Interleukin10(IL-10)inhibitshumanlymphocyteinterferongamma-productionbysuppressingnaturalkillercellstimulatoryfactor/IL-12synthesisinaccessorycells.
J.
Exp.
Med.
178,1041–1048(1993).
46.
Chard,L.
S.
etal.
Avacciniavirusarmedwithinterleukin-10isapromisingtherapeuticagentfortreatmentofmurinepancreaticcancer.
Clin.
CancerRes.
21,405–416(2015).
47.
Mumm,J.
B.
etal.
IL-10elicitsIFNgamma-dependenttumorimmunesurveillance.
CancerCell20,781–796(2011).
48.
Faul,F.
,Erdfelder,E.
,Lang,A.
G.
&Buchner,A.
G*Power3:aexiblestatisticalpoweranalysisprogramforthesocial,behavioral,andbiomedicalsciences.
Behav.
Res.
Methods39,175–191(2007).
49.
Chartier,C.
etal.
EfcientgenerationofrecombinantadenovirusvectorsbyhomologousrecombinationinEscherichiacoli.
J.
Virol.
70,4805–4810(1996).
50.
Dondji,B.
etal.
CD4Tcellsmediatemucosalandsystemicimmuneresponsestoexperimentalhookworminfection.
ParasiteImmunol.
32,406–413(2010).
AcknowledgementsThisprojectwassupportedbytheNationalNaturalScienceFoundationofChina(81272525and81201792),MinistryofScienceandTechnologyofChina(2013DFG32080)andTheMRC(MR/M015696/1).
Allcontributingauthorshaveagreedtothesubmissionofthismanuscriptforpublication.
Originatingfromajointcolla-borationbetweentheSino-BritishResearchCentreandBeijingBio-TargetingARTICLENATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-814NATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunicationsTherapeuticsInc.
whoidentiedanunexpectedmutantAd-TD-IL12inwhichthecytokineisexpressedinanon-secretingform.
AuthorcontributionsY.
W.
supervisedthewholeproject.
Y.
W.
andP.
W.
designedthestudyandconrmedthemutantIL12isnon-secreted.
P.
W.
performedthemajorityoftheexperiments;X.
L.
didbiologicaltimepointsexperimentsandanalyzedthedata.
D.
G.
andZ.
C.
performedthehistopathologicalstudy;J.
W.
,Y.
L.
,H.
L.
andZ.
Z.
contributedtoanimalexperiments;Y.
C.
puriedandtitratedadenovirus;H.
L.
andG.
Z.
madepShuttleplasmidsfortheviralconstruction;S.
W.
participatedinthedesignandinterpretationofsomeexperiments;S.
W.
,J.
D.
,B.
F.
andN.
R.
L.
criticallyreviewedthepaper.
P.
W.
,L.
S.
C.
andY.
W.
inter-pretedallresultsandwrotethemanuscript.
AdditionalinformationSupplementaryInformationaccompaniesthispaperatdoi:10.
1038/s41467-017-01385-8.
Competinginterests:Y.
W.
,P.
W.
,D.
G.
,andN.
R.
L.
areinventorsofaledpatentapplicationrelevanttothisstudy.
Theremainingauthorsdeclarenocompetingnancialinterests.
Reprintsandpermissioninformationisavailableonlineathttp://npg.
nature.
com/reprintsandpermissions/Changehistory:AcorrectiontothisarticlehasbeenpublishedandislinkedfromtheHTMLversionofthispaper.
Publisher'snote:SpringerNatureremainsneutralwithregardtojurisdictionalclaimsinpublishedmapsandinstitutionalafliations.
OpenAccessThisarticleislicensedunderaCreativeCommonsAttribution4.
0InternationalLicense,whichpermitsuse,sharing,adaptation,distributionandreproductioninanymediumorformat,aslongasyougiveappropriatecredittotheoriginalauthor(s)andthesource,providealinktotheCreativeCommonslicense,andindicateifchangesweremade.
Theimagesorotherthirdpartymaterialinthisarticleareincludedinthearticle'sCreativeCommonslicense,unlessindicatedotherwiseinacreditlinetothematerial.
Ifmaterialisnotincludedinthearticle'sCreativeCommonslicenseandyourintendeduseisnotpermittedbystatutoryregulationorexceedsthepermitteduse,youwillneedtoobtainpermissiondirectlyfromthecopyrightholder.
Toviewacopyofthislicense,visithttp://creativecommons.
org/licenses/by/4.
0/.
TheAuthor(s)2017NATURECOMMUNICATIONS|DOI:10.
1038/s41467-017-01385-8ARTICLENATURECOMMUNICATIONS|8:1395|DOI:10.
1038/s41467-017-01385-8|www.
nature.
com/naturecommunications15