recipientswww.k8k8.com

ARTICLEReceived6May2014|Accepted7Jul2014|Published5Aug2014DOCK8regulatesprotectiveimmunitybycontrollingthefunctionandsurvivalofRORgtILCsAkhileshK.
Singh1,AhmetEken1,MalloryFry1,EstelleBettelli2&MohamedOukka1,3Retinoicacidreceptor-relatedorphanreceptor-gt-positive(RORgt)innatelymphoidcells(ILCs)produceinterleukin(IL)-22andIL-17,whicharecriticalforprotectiveimmunityagainstentericpathogens.
ThemolecularmechanismunderlyingthedevelopmentandsurvivalofRORgtILCsisnotthoroughlyunderstood.
Here,weshowthatDedicatorofcytokinesis8(DOCK8),ascaffoldingproteininvolvedincytoskeletalrearrangementandcellmigration,isessentialfortheprotectiveimmunityagainstCitrobacterrodentium.
AcomparativeRNAsequencing-basedanalysisrevealsanimpairedinductionofantimicrobialpeptidesinthecolonofDOCK8-decientmice,whichcorrelateswithhighsusceptibilitytoinfectionandaverylownumberofIL-22-producingRORgtILCsintheirGItract.
Furthermore,DOCK8-decientRORgtILCsarelessresponsivetoIL-7mediatedsignalling,morepronetoapoptosisandproducelessIL-22duetoadefectinIL-23-mediatedSTAT3phosphorylation.
OurstudiesrevealanunsuspectedroleofDOCK8forthefunction,generationandsurvivalofRORgtILCs.
DOI:10.
1038/ncomms56031CenterforImmunityandImmunotherapies,SeattleChildren'sResearchInstitute,Seattle,Washington98101,USA.
2BenaroyaResearchInstitute,ImmunologyProgram,Seattle,Washington98101,USA.
3DepartmentofImmunology,UniversityofWashington,Seattle,Washington98105,USA.
CorrespondenceandrequestsformaterialsshouldbeaddressedtoM.
O.
(email:moukka@u.
washington.
edu).
NATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications1&2014MacmillanPublishersLimited.
Allrightsreserved.
Themammaliangastrointestinal(GI)tractharboursalargenumberofmicrobiota1.
ThemucosaoftheGItractisacriticalinterfaceforthehostimmunesystemanditsassociatedmicrobialecosystem2.
Atthebarriersurfaces,innatelymphoidcells(ILCs)constituteamajorfractionoftheimmunecellrepertoireandserveaprotectiveroleininnateimmuneresponsetoinfectiousmicroorganismsandintissueremodellingafterdamageinictedbyinjuryorinfection.
ILCsarelineage-negative,Id2-dependentcellswithvaryingfunctions3–5.
SimilartoT-helpercelllineage,ILCsareclassiedintothreegroups6,7basedontheirrequirementoftranscriptionfactorsandtheeffectorcytokineproduction.
Type1ILCsrequiretranscriptionfactorT-betandproduceinterferon-g,whiletype2ILCs(ILC2)dependontranscriptionfactorGATA3andproduceTh2-associatedcytokinesinterleukin-5(IL)-5andIL-13(refs5,6).
Type3ILCs(ILC3)expressRORgttranscriptionfactorandproduceIL-22and/orIL-17(refs3,5,7).
ILC3includelymphoidtissueinducercells(LTi)thatmediatelymphoidtissuedevelopmentinthefetusanditsregenerationintheadult.
Accordingly,RORgt-decientmicelackLTiandlymphnode(LN)formation8.
Althoughthetranscriptionfactorsimportantforthegenerationofthesecellshavebeenidentied,littleisknownaboutthefactorsimportantfortheirmaintenanceandregulationofcytokineproduction.
RORgtILCsproducelargequantitiesofIL-22(refs9–14).
IL-22hasbeenshowntopromotewoundhealing,proliferationandantiapoptoticpathwaysintheintestinalepithelialcells15;italsoupregulatesantimicrobialpeptidesandmucusproduction9,16.
IL-22-mediatedearlyinductionofantimicrobialpeptides,RegIIIgandRegIIIbincolonicepithelialcellsisrequiredforprotectionfromCitrobacterrodentium(C.
rodentium)infection12,16.
Accordingly,IL-22-decientmiceshowincreasedmorbidityandmortalityafterC.
rodentiuminfection16.
TheproductionofIL-22byILCsinresponsetoC.
rodentiuminfectionispromotedbyIL-23(refs12,16,17).
Dedicatorofcytokinesis8(DOCK8)interactswiththeRhofamilysmallGproteinCdc42(refs18,19).
AveryrecentstudyhasshownthatinBcells,afterligationofTLR9,DOCK8servesasanadaptorproteinandlinksTLR9toaSTAT3cascade,whichisessentialforTLR9-drivenB-cellproliferationanddifferentiation20.
DOCK8mutation(null)isassociatedwithcombinedimmunodeciencyinhumans21,22.
DOCK8-decientpatientsdevelopatopicdermatitis,Staphylococcusaureusskinabscessesorsofttissueinfection,pneumonias,elevatedserumIgE,eosinophiliaandaremoresusceptibletocancer21,23.
ThemajorityofthesepatientshavelownumbersofCD4andCD8Tcellsinthebloodandspleen21.
However,ithasnotbeeninvestigatedwhetherDOCK8-decientpatientsalsohavedefectsinILCsoralteredmicrobialcompositionintheirGItracts,whichwouldrenderthemmoresusceptibletoinfections.
Herein,wepresentevidencethatDOCK8isessentialforprotectiveimmunityagainstanentericpathogen.
UsingDOCK8-decientmice,wehaveshownthatDOCK8playsacrucialroleinthesurvivalofRORgtILCsandtheircytokineproduction.
Indeed,wefurtherdeterminedthatRORgtILCsrequireDOCK8foroptimalSTAT3activationandIL-22productiononIL-23stimulation.
Therefore,ourstudiessuggestthatbesideshavingdefectiveB-cell-mediatedresponses,DOCK8-decientpatientsmayalsohavelimitednumbersofILC3.
Thus,ourstudiesmayhelpexplainwhytheabsenceofDOCK8contributestopatients'susceptibilitytoabroadspectrumofpathogens.
ResultsDOCK8isrequiredfortheclearanceofanentericpathogen.
TodenetheroleofscaffoldingproteinDOCK8inIL-23-mediatedinductionofIL-22,weutilizedDOCK8pri/primice24.
ThesemicehaddefectsinB-cellimmunologicalsynapses,germinalcentreformationandlong-livedantibodyproduction.
AlongwithB-celldefects,thesemicealsohadadefectinT-cellsurvivalandfunction25.
Therefore,werstexaminedthetotalcellnumberaswellasthecellcompositioninDOCK8pri/primice.
Aspreviouslyshown,thesemicehavesignicantlyfewercellsinspleenandLNsincomparisonwithageandsex-matchedwild-type(WT)mice(SupplementaryFig.
1a,top).
Incontrasttotheperipheralcellcount,DOCK8pri/primicehaveacomparablenumberoflaminaproprialymphocytes(LPL)intheGItract,butsignicantlyhighercellsinthemesentericLNincomparisonwithWTmice(SupplementaryFig.
1a,bottom).
Consistentwithpreviousndings,CD4andCD8TcellsinthespleensofDOCK8-pri/primicewerereducedby50–60%ofWTlevels25,26.
Inaddition,wedidnotndsignicantchangesinthenumbersofCD4andCD8Tcellsinthelaminapropria(LP)ofDOCK8pri/primiceincomparisonwithWT(SupplementaryFig.
1b).
Interestingly,DOCK8deciencydidnotaffectthenumberofmatureBcellsinlymphoidorgansexceptinthesmallintestine(SI),wherewefoundasignicantreductionoftheB-cellpercentage(SupplementaryFig.
1c).
ThisreductionwasalsomorepronouncedwhenmicewereinfectedwithC.
rodentium(SupplementaryFigs1dand2).
TheBcellsthatwerefoundinthecolonicLPexpressedhighlevelsofmajorhistocompatibilitycomplex(MHC)ClassII(SupplementaryFig.
1d).
DOCK8hasbeenshowntobecriticalforinterstitialdendriticcell(DC)migrationduringimmuneresponses26.
Therefore,weassessedwhetherDOCK8deciencycouldaffecttherecruitmentofmyeloidcellsinresponsetoaninfection.
However,wedidnotndanyreductionintherecruitmentofCD11borCD11ccellstotheLPofDOCK8pri/primiceafterinfectionwithC.
rodentium(SupplementaryFig.
1e).
SinceithasbeenshownthatBcellsplayanimportantroleintheclearanceofentericpathogens27–29,weinfectedWT,DOCK8pri/primice,B-cell-decient(mMT)andIL-23R/micewithC.
rodentium.
Interestingly,similartoIL-23R/mice,allDOCK8pri/primicelostweightandsuccumbedtotheinfection(Fig.
1a,b).
ThespleensandliversofDOCK8pri/primicecontainedtwotofourlogmorebacteriathanWT,indicatingthatDOCK8pri/primicewereunabletocontrolbacterialdissemination(Fig.
1c).
AlthoughweobservedadefectintherecruitmentofMHCclassIIhighBcellstotheGItractofDOCK8pri/primice(SupplementaryFigs1c,dand2),mMTmicesurvivedandwereabletocleartheinfection(Fig.
1a,b).
TheseresultssuggestthatDOCK8deciencyrendersmicesusceptibletoC.
rodentiuminfectionindependentlyofBcellfunction.
IL-22/andIL-23p19/micearehighlysusceptibletoC.
rodentiuminfection16,30,31.
Therefore,weusedIL-23R/miceasapositivecontrolforinfection.
Consistentwithabove,IL-23R/micewerehighlysusceptibletoinfection(Fig.
1a,b).
Next,wedeterminedwhetherDOCK8wasrequiredwithinhaematopoieticcellsornon-haematopoieticcells.
WegeneratedchimerasbytransferringbonemarrowcellsisolatedfromeitherDOCK8pri/priorWTlittermatemiceintoirradiatedRag2/IL-2Rg/mice.
Fiveweeksaftertransfer,thesemicewereinfectedwithC.
rodentium.
Asexpected,similartoRag2/IL-2Rg/mice,allRag2/IL-2Rg/micetransferredwithDOCK8pri/pribonemarrowsuccumbedtotheinfection,whileWTbonemarrowtransferredmicesurvivedtheinfection(Fig.
1d).
TheliversofRag2/IL-2Rg/miceandRag2/IL-2Rg/micetransferredwithDOCK8pri/pribonemarrowcontainedsignicantlyhigherbacterialloadthanthosetransferredwithWTbonemarrowcells(Fig.
1e).
TheseresultssuggestthatDOCK8isrequiredwithinthehaematopoieticcompartmentandplaysanARTICLENATURECOMMUNICATIONS|DOI:10.
1038/ncomms56032NATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications&2014MacmillanPublishersLimited.
Allrightsreserved.
120%Survival%SurvivalGeneexpression(log2ratio)IL-22mRNARegIIIγmRNAIL-23p19mRNACFU(spleen)CFUperg(liver)CFUperg(liver)IL-2Rγc–/–Reg3γReg3βCxcI9Tnfrsf8IL-22IL-23RIL-7IL-7RIL-23Rin2CxcI13CxcI2Trem1IL-6AHRDOCK8pri/priDOCK8pri/priWTWTDOCK8pri/priWTDOCK8pri/priWTBodyweightchange(%)1008060402000020406080100120400,000******300,000200,000100,000Rag2IL-2Rγc–/–WTDock8pri/pri0107–7840–4–8–5–4–4–3–1011124550.
000450.
150.
00080.
00060.
00040.
000200.
10.
0500.
00030.
000150****NSLog2ratioCxcl2Cxcl13Trem1IL-6IL-23IL-7RIL-23RIL-22Tnfrsf8CxC19RegIIIγRegIIIβRin2****1051031011071051031010357911131520258503664288111012141613152025WTIL-23R–/–Dock8pri/priμMTDayspostinfection9095100105DayspostinfectionDayspostinfectionFigure1|DOCK8isrequiredfortheclearanceofentericpathogen.
(a–c)WT,IL-23R/,DOCK8pri/priandmMTmicewereinfectedwithC.
rodentium(a)percentsurvivaland(b)averagebodyweightchangeofWT(n5),IL-23R/(n5),DOCK8pri/pri(n6)andmMT(n4)miceafterinfection.
(c)C.
rodentiumcolony-formingunits(CFU)inthespleensandliversofinfectedmiceatday8postinfection(p.
i.
).
(d,e)DOCK8isrequiredinhaematopoieticcompartmentforthesurvival.
BonemarrowcellsfrommiceofindicatedgenotypeweretransferredintoirradiatedRag2/IL-2Rgc/mice.
(d)Percentsurvivaland(e)C.
rodentiumCFUintheliversofinfectedmiceatday8p.
i.
(f)DOCK8-dependentgeneexpressioninthecolonsofC.
rodentiuminfectedmice.
(g)ColonicIL-22,RegIIIgandIL-23p19expressioninresponsetoinfectionatday8p.
i.
Threeindependentexperimentswereperformedwithaminimumoffourtosixmicepergroup.
Thedatashownarethemean±s.
d.
PvalueswereobtainedbyStudent'st-test.
**Po0.
01,***Po0.
001.
NS,notsignicant.
NATURECOMMUNICATIONS|DOI:10.
1038/ncomms5603ARTICLENATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications3&2014MacmillanPublishersLimited.
Allrightsreserved.
importantroleinthefunctionofimmunecellsinvolvedinprotectiveimmunity.
TounderstandtheeffectofDOCK8onglobaltranscriptionlevels,weusedRNAsequencingtoexaminethedifferentialgeneexpressiononC.
rodentiuminfection.
RNAwaspuriedfromtheinfectedcolonsofWTandDOCK8pri/primice.
WeevaluatedthelevelsofthegeneexpressionintheinfectedcolonsofWTandDOCK8pri/primice.
Amongthemosthighlyexpressedgenes,RegIIIg,RegIIIbandIL-22werehighlydefectiveinDOCK8pri/primice,incomparisonwithWTmice(Fig.
1f).
SinceithasbeenshownthatearlyIL-22-dependentinductionofRegIIIgiscriticalforprotectiveimmunityagainstentericpathogen12,16,wemeasuredcolonicIL-22andRegIIIgtranscriptsafterinfectionwithC.
rodentium.
ConsistentwithRNA-seqdata,thelevelsofIL-22andRegIIIgexpressionwereverylowinthecolonofDOCK8pri/primiceincomparisonwithlevelsfoundinthecolonofWTmice(Fig.
1g).
ThisndingindicatesthatDOCK8isrequiredfortheinductionofIL-22expressiononinfection.
IL-22canbeproducedbymanycelltypes,includingILCsandTh22cells3,9,11,12,31–33.
Therefore,weaddressedthecellularsourceofIL-22thatwascriticalforprotectionagainstentericpathogeninfection.
ToassesswhetherDOCK8isrequiredforIL-22productionbyILCsorTh22,wecrossedDOCK8pri/primicewithRag1/micelackingTcells.
Thesemicewerealsohighlysusceptibletoentericpathogeninfectionandfailedtoclearinfection(SupplementaryFig.
3a,b).
ThespleensandliversofRag1/DOCK8pri/primicecontainedsignicantlymorebacteriathanRag1/mice,indicatingthatDOCK8isrequiredininnateimmunecellsforthecontrolofentericpathogeninfection(SupplementaryFig.
3c).
AndsimilartoDOCK8pri/primice,Rag1/DOCK8pri/primicewerealsohighlydefectiveinIL-22-dependentRegIIIginductionafterentericpathogeninfection(SupplementaryFig.
3d).
Alltogether,thesedataindicatethatDOCK8isrequiredforprotectiveimmunityagainstC.
rodentiumbyregulatingIL-22productionfrominnateimmunecells.
ExogenousIL-22protectsDOCK8KOmiceduringC.
roden-tiuminfection.
Ourdata(Fig.
1andSupplementaryFig.
3)clearlyshowedthatDOCK8wasrequiredforcontrollingC.
rodentiuminfection.
TotestwhetherexogenousIL-22couldrescuetheDOCK8pri/primiceonC.
rodentiuminfection,weinjectedIL-22cDNAviahydrodynamicdeliveryintoDOCK8pri/primice.
WehadpreviouslyshownthatasingleinjectionofIL-22-encodingcDNAintoIL-22-defectivemicerestoredexpressionofIL-22inthebloodandcolon,lastedaslongas10days34.
WegaveoneinjectionofIL-22-encodingplasmid2daysbeforeinfectionandanotheratday5postinfection.
DOCK8pri/primicethatreceivedtheIL-22plasmidsurvivedanaverageof5dayslongerandhadsignicantlylessweightlosscomparedtoemptyplasmidrecipients(Fig.
2a,b).
IntheIL-22plasmid-treatedgroup,splenicandliverbacteriacountsweresignicantlylowercomparedtotheemptyplasmid-injectedgroup(Fig.
2c).
Thus,thesesetsofexperimentscollectivelydemonstratedthatdefectiveIL-22productionwasanimportantcauseofhighbacterialoadanddeathofDOCK8pri/primiceduringC.
rodentiuminfection.
Conversely,only15%ofDOCK8pri/primicethatreceivedtheIL-22plasmidsurvivedtheinfection,suggestingthatthesemicemighthaveotherdefectsininnateimmunecells,whichwererequiredforfurthersurvival.
DOCK8hasbeenshowntobeimportantforB-cellimmunologicalsynapses,germinalcentresandlong-livedantibodyproduction24.
ItisalsoimportantforthesurvivalandfunctionofCD8TandNKTcells25,35.
TheDOCK8deciencyalsoresultsinreducedCD4TcellsanddefectiveinterstitialDCmigrationduringimmuneresponses25,26.
ThesepreviousndingssuggestthatDOCK8isalsorequiredforfullyfunctionalacquiredimmuneresponse.
ItisimportanttomentionthatmicedefectiveintherecombinationactivatinggenebecomepredisposedtoC.
rodentiuminfectionat3to4weekspostinfection,suggestingthatbothinnateandacquiredimmuneresponseisrequiredtoclearC.
rodentiuminfection(SupplementaryFig.
3a,b)12,16.
Thus,partialrestorationofsurvivalinDOCK8pri/primiceinjectedwithIL-22plasmidcouldbeduetodefectiveacquiredimmuneresponses.
RORctILCsrequireDOCK8foroptimalIL-22production.
PriorreportsundoubtedlyshowedthatILC3playsadecisiveroleinprotectionagainstC.
rodentiuminfectionbyproducingIL-22earlyon9,10,12,13,36–38.
Inthismodel,earlyduringinfection,classicalDCsproduceIL-23thatinducesIL-22productionviaILC3(refs12,16,17).
IL-22enhancesepithelialcellrepairandpreventsdisseminationofbacteria.
Therefore,weanalysedwhetherthedefectinIL-22productionbyDOCK8-decientILCsresultedfromdirectDOCK8-mediatedsignallinginILCsorresultedfromanindirecteffectofDOCK8onothercellsimportantforitsfunction.
ThedefectiveIL-22productioninDOCK8pri/primicecouldbeduetodefectiveIL-23expressioninDCs.
Therefore,wemeasuredIL-23p19,aspecicsubunitofIL-23,inthecolonofC.
rodentiuminfectedmice.
Interestingly,IL-23expressioninthecolonofC.
rodentiuminfectedDOCK8pri/primicewascomparabletothatofWTmice(Fig.
1g),suggestingthattheIL-22defectinDOCK8pri/primiceisnotduetothe120%Survival%BodyweightCFU(spleen)LiverCFUperg100806040200035791113151719212325700357911131021031031041041051051068090100110********DayspostinfectionDayspostinfectionEmptyvectorinjectedIL-22plasmidinjectedEmptyvectorinjectedIL-22plasmidinjectedFigure2|ExogenousIL-22protectsDOCK8KOmiceduringC.
rodentiuminfection.
DOCK8pri/primicewereadministeredwithIL-22expressingplasmidoremptyvectorviahydrodynamicinjectionsat2and5daysofC.
rodentiuminfection.
(a)Percentsurvival(b)averagebodyweightchangesafterinfectionand(c)colony-formingunitsinthespleensandliversofinfectedmice.
Allthedataarerepresentativeofthreeindependentexperimentswithvemicepergroup.
Thedatashownarethemean±s.
d.
PvalueswereobtainedbyStudent'st-test.
*Po0.
05,**Po0.
01,***Po0.
001.
ARTICLENATURECOMMUNICATIONS|DOI:10.
1038/ncomms56034NATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications&2014MacmillanPublishersLimited.
Allrightsreserved.
impairedIL-23expressioninDCs.
WealsoanalysedtheexpressionofDock8ininnateIL-23RcellsandhavefoundthatDock8washighlyexpressedinthesecells(SupplementaryFig.
4).
NextweinvestigatedwhetherDOCK8wasalsorequiredforIL-22productionbyRORgtILCsonIL-23stimulation.
Therefore,weexaminedIL-22expressionintheLPLsisolatedfromDOCK8pri/primiceandtheirlittermatecontrolWTmice.
Asexpected,inWTmice,ILC3(CD3negative)producedhighamountsofIL-22inresponsetoIL-23.
ComparedwiththoseinWTmice,boththepercentageandabsolutenumberofCD3IL-22cellsintheGItractofDOCK8pri/primiceweresignicantlyreduced,indicatingthatDOCK8wasimportantfortheproduc-tionofIL-22intheGItract(Fig.
3a,b).
However,thisresultcouldbemisleadingduetothepaucityofRORgtILC3inDOCK8pri/primicethatcouldhaveledtoanoverallreductionofIL-22production.
Thus,itremainsuncertainwhetherDOCK8directlyregulatesIL-22productionfromeachindividualILC3.
Toverifythis,wepuriedCD90.
2IL-23RcellsfromtheSILPLsofRag1/IL-23Rgfp/orRag1/IL-23Rgfp/DOCK8pri/primiceandculturedthemwithorwithoutIL-23.
TheIL-22proteinintheculturesupernatantwasmeasuredbyenzyme-linkedimmunosorbentassay(ELISA).
TheproductionofIL-22byDOCK8-decientRORgtILCswassignicantlyreducedcomparedwiththoseinWTmice(Fig.
3c).
Tofurthercorroborateourndings,wemeasuredIL-22andIL-17transcriptlevelsinsortedILC3sfromRag1/IL-23Rgfp/orRag1/IL-23Rgfp/DOCK8pri/primice(SupplementaryFig.
5).
TheIL-22transcriptlevelinDOCK8-decientILC3swassignicantlyreducedcomparedtothoseinWTmice.
However,IL-17AtranscriptlevelwascomparableinWTandDOCK8-decientILC3s.
TheseresultssuggestedthatDOCK8wasrequiredforRORgtILCsoptimalIL-22productioninresponsetoIL-23stimulation.
WealsoinvestigatedthefunctionofDOCK8onIL-22productionbyCD4TcellsafterC.
rodentiuminfection.
DOCK8-decientCD4TcellsarehighlydefectiveinIL-22productionafterC.
rodentiuminfection(SupplementaryFig.
6).
However,IL-22producedbyCD4TcellsforprotectionagainstC.
rodentiumisnotascrucialasIL22producedbyILC3(ref.
39).
SinceIL-23signalsviaSTAT3-dependentmechanisms,wetestedwhetherDOCK8wasalsorequiredinILC3fortheactivationofSTAT3inresponsetoIL-23stimulation.
Thus,westimulatedfreshlyisolatedILC3withIL-23andassessedSTAT3phosphorylation(pY705-STAT3)byowcytometry(Fig.
3d).
ConsistentwiththeIL-23-mediatedSTAT3phosphorylationinTcells40,IL-23inducedSTAT3phosphorylationinWTILCs;however,IL-23-mediatedSTAT3activationwasseverelyimpairedinDOCK8-decientILCs(Fig.
3d).
ThisdataclearlyshowthatDOCK8isrequiredforSTAT3activationandIL-22productiononIL-23stimulation.
ProfoundreductionofRORctILCsintheabsenceofDOCK8.
ThemajorinnatecellpopulationthatexpressesIL-23Rintheintestineisgroup3ILC,whichiscomposedofLTicellsandILC3s41.
VariousILC3sweredescribedinhumansandmicebasedontheeffectorcytokinestheyproduce.
Theseincludebutarenotlimitedto(1)IL-22-producingNCRILC3(ref.
9),whicharealsocalledILC22,NK22,NKR-LTiorNCR22(ref.
42)(2)NCRIL-17AIFN-gdoubleproducingILC3(ref.
43)andlastly(3)NCRIL-17AILC3s44.
AllofthesesubsetsofILC3srespondtoIL-23andproducetheabove-mentionedcytokine.
Thus,thegroup3ofILCsisveryheterogeneousandiscomposedbynumeroussubsetsofcellsthatcanbedistinguishedbytheexpressionsofvarioussurfacemarkerssuchaCD4andNKp46.
TodeterminewhetherDOCK8deciencyaffectsthedevelopmentSmallintestineColonGatedonCD3–3.
090.
242.
45WTDOCK8pri/pri9.
4187.
817.
479.
20.
370.
430.
270.
562.
296.
81.
8897.
4RORγtIL-220.
461008060%Max402000102103104105pY-STAT3PBS+ILC3IL-23+ILC3IL-23+KOILC343210SIColonWTDock8pri/pri***750500IL-22concentration(pgml–1)2500IL-23stimulatedUnstimulatedIL-23stimulatedUnstimulatedRag1–/–IL-23Rgfp/+Rag1–/–IL-23Rgfp/+DOCK8pri/priSIColon*******IL-22+cells(%)4.
5***31.
50IL-22+cells(*103)Figure3|DOCK8isalsorequiredfortheIL-23-mediatedinductionofIL-22.
(a)SILPLsfromWTandDOCK8pri/primicewerestimulatedwith20ngml1IL-23for4hinthepresenceofGolgistop.
RORgtandIL-22expressioninCD3cellswasanalysedbyowcytometry.
(b)PercentageandabsolutenumbersofintestinalCD3IL-22cells.
(c)SortedILC3(CD90.
2IL-23Rgfp/)fromintestinalLPLsofRag1/IL-23Rgfp/andRag1/IL-23Rgfp/DOCK8pri/primicewereculturedwithorwithoutIL-23(20ngml1).
TheIL-22producedinculturesupernatantwasmeasuredbyELISA.
(d)Fluorescence-activatedcellsortinganalysisfordetectionofpSTAT3inILC3fromRag1/IL-23Rgfp/orRag1/IL-23Rgfp/DOCK8pri/primice.
Allthedataarerepresentativeofthreeindependentexperimentswiththreemicepergroup.
Thedatashownarethemean±s.
d.
PvalueswereobtainedbyStudent'st-test.
**Po0.
01,***Po0.
001.
NATURECOMMUNICATIONS|DOI:10.
1038/ncomms5603ARTICLENATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications5&2014MacmillanPublishersLimited.
Allrightsreserved.
ofaparticularsubsetofILC3intheGItract,weexaminedtheRORgtILCcompartmentofDOCK8pri/primice.
Thus,westainedboththeSIandcolonlaminaproprialeukocytes(LPLs)withCD3,CD90.
2,CD127,RORgt,CD4andNKp46-specicantibodies.
BoththepercentageandnumberofRORgtILCs(CD3CD90.
2CD127RORgt)intheSIandcolonweresignicantlyreducedinDOCK8pri/primice(Fig.
4a–candSupplementaryFig.
7a–c).
ThenumberofRORgtILCsintheLPLsofSIofDOCK8pri/primicewasreducedto5.
6%ofthenumberinWTmice.
ThepercentageofCD4andNKp46ILCsamongRORgtILCswascomparablebetweenDOCK8pri/priandWTmice,despitethefactthattheirabsolutenumbersweresignicantlyreducedinDOCK8pri/primice(Fig.
4a–candSupplementaryFig.
7a–c).
Next,weinvestigatedwhetherthedefectofRORgtILCsinDOCK8pri/primicewasonlylimitedtothegut,orcouldbeobservedalsoinotherorgans.
16.
2GatedonCD3-CD90.
2+CD127+cells63.
110.
2WTDOCK8pri/pri17.
671.
610.
840.
879.
965.
78.
65CD3CD127NKp46CD4RORγt2015105255075100RORγt+ILC3(%)CD3–CD90.
2+cells(%)CD3–CD90.
2+cells(*104)RORγt+ILC3(*104)CD4+ILC3(*104)NKp46+ILC3(*104)0012301230GatedonCD3–90.
634.
643.
443.
69693.
420.
514.
14CD127Dock8pri/priSILPLGatedonCD3–,CD90.
2+,CD127+andRORγt+GatedonCD3–,CD90.
2+,CD127+RORγtWTCD90.
2CD90.
2WTCD127CD36040SILPLcells(%)SILPLRORγt+ILCs(%)RORγt+ILCsinSILPL200CD45.
2CD45.
2CD45.
1CD45.
118160,000120,00080,00040,00001260******WT(CD45.
1)Dock8pri/pri(CD45.
2)CD45.
2CD45.
1NSDock8pri/pri0.
050.
10.
150.
200.
40.
81.
2WTSILPLDOCK8pri/priSILPL*****4****01020CD4+ILC3(%)3075NKp46+ILC3(%)502500*****NSNSCD90.
2RORγtFigure4|ProfoundreductionofRORctILCsintheabsenceofDOCK8.
(a)RORgtILCswereanalysedinSILPLsbyowcytometryaftergatingonCD3CD90.
2CD127cells.
(b,c)PercentageandabsolutenumbersofCD3CD90.
2,CD3CD90.
2CD127RORgt,CD3CD90.
2CD127RORgtCD4andCD3CD90.
2CD127RORgtNKp46cellsofWTandDOCK8pri/primice.
(d,e)Cell-intrinsicfunctionsofDOCK8regulateRORgtILCs.
Mixedbonemarrowchimerasgeneratedbyreconstituting50%CD45.
1WTand50%CD45.
2DOCK8pri/primiceandtransferredintolethallyirradiatedRag2/IL-2Rgc/mice.
(d)RORgtILCs(CD3CD90.
2CD127)wereanalysed4weeksposttransfer.
(e)ThepercentageandabsolutenumberofRORgtILCs.
Threeindependentexperimentswereperformedwithaminimumofthreemicepergroup.
Thedatashownarethemean±s.
d.
PvalueswereobtainedbyStudent'st-test.
**Po0.
01,***Po0.
001.
NS,notsignicant.
ARTICLENATURECOMMUNICATIONS|DOI:10.
1038/ncomms56036NATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications&2014MacmillanPublishersLimited.
Allrightsreserved.
BoththepercentageandabsolutenumbersofRORgtILCsweresignicantlyreducedinmultipleorgansofDOCK8-decientmice,includingspleensandLNs(SupplementaryFig.
8a–c).
Thus,DOCK8isnotonlyimportantforthesurvivalofRORgtILCsofthegut,butitisalsorequiredfortheirmaintenanceinotherorgansaswell.
WefurtherexploredwhetherDOCK8requirementbyRORgtILCswasperinatalorpostnatal.
Therefore,weassessedthenumberofRORgtILCsinthegutofWTandDOCK8pri/prinewbornmice.
IncontrasttothesignicantdecreaseofRORgtILCsinthegutLPLsofadultDOCK8pri/primice,thefrequencyoftotalRORgtandperinatalCD4LTicellsinthegutLPLsisolatedfromDOCK8pri/prinewbornmicewascomparabletothosefromWTnewborns,suggestingthatDOCK8isdispensableforthedevelopmentofperinatalLTicellsinthegut(SupplementaryFig.
9a–c).
Recently,itwasshownthatILC3scouldbeidentiedasCD45lowandCD90.
2highcells39.
Therefore,tosortperinatalLTicells,westainednewbornmiceGItractLPLswithCD45,CD3,CD90.
02andRORgtandassessedwhetherCD3CD45lowandCD90.
2highcellscouldrepresentRORgtcells,asinadultmice(SupplementaryFig.
9d,upperpanel).
Consistentwithrecentlypublishedwork39,mostoftheRORgtcellsfoundinLPofnewbornmicecouldbeidentiedasCD3CD45lowandCD90.
2highcells.
Then,wesortedCD4andCD4perinatalLTicellsandassessedtheexpressionofDock8.
OurresultshowedthatCD4LTicellsdoexpressDock8transcripts,butthatexpressionlevelwassignicantlyreducedcomparedwithCD4LTiandadultILC3s(SupplementaryFig.
9d,e).
Alltogether,theseresultssuggestthatDOCK8isdispensableforthedevelopmentofperinatalLTicellsinthegut.
Tofurtherconrmourndings,wecrossedDOCK8pri/primicewithRag1/IL-23Rgfp/andgeneratedRag1/IL-23gfp/DOCK8pri/primiceinwhichIL-23RexpressioncanbefollowedbasedonGFPexpression45.
ThesemicehavefewerCD90.
2IL-23Rcellscomparedwithlittermatecontrolmice(SupplementaryFig.
10a).
WealsomeasuredthelevelofIL-23RtranscriptsinthecolonsofRag1/IL-23gfp/DOCK8pri/primiceafterC.
rodentiuminfection.
ConsistentwithDOCK8pri/primice,thesemicealsohadsignicantlyfewerIL-23RtranscriptsincomparisonwithRag1/IL-23Rgfp/mice(SupplementaryFig.
10b,c).
Altogether,theseresultsindicatethatDOCK8isinvolvedinthedevelopmentorsurvivalofRORgtIL-23RILCs.
TodeterminewhetherDOCK8wasonlyrequiredforthesurvivalofRORgtILCs,wecomparedtheeffectofDOCK8deciencyonthegenerationofILC2intheGItractLPLs(SupplementaryFig.
11a,b).
IncontrasttoILC3inthegutofDOCK8-decientmice,ILC2absolutenumbersofDOCK8-decientmicedidnotsignicantdifferfromthenumberofILC2sfoundinDOCK8-sufcientmice.
Interestingly,thepercentageofILC2wasinverselycorrelatedwiththepercentageofILC3inDOCK8-decientmiceandwassignicantlyelevatedinDOCK8-decientmice.
ThesedatasuggestthatDOCK8deciencyspecicallyinuencesthegeneration/survivalofRORgtILCs,butnotGATA3ILC2,inthegut.
ToinvestigatewhetherthereductioninRORgtILCsisduetoacell-intrinsicdefect,wegeneratedmixedbonemarrowchimericmiceinwhich50%ofthereconstitutingbonemarrowwasderivedfromWTCD45.
1donorsand50%wasderivedfromDOCK8pri/priCD45.
2congenicdonors.
Thus,WTandDOCK8pri/pricellscouldbedistinguishedonthebasisofCD45.
1andCD45.
2expressions,respectively.
WethenevaluatedthegenerationofRORgtILCsinthesechimericmice.
DOCK8-decienthaematopoieticprogenitorcellsfailedtogiverisetoRORgtILCs,whereasWTbonemarrowprogenitorsdevelopednormallyintoRORgtILCsinirradiatedRag2/IL-2Rg/mice(Fig.
4d,eandSupplementaryFig.
7d,e).
ThesedatasuggestanintrinsicrequirementofDOCK8forthedevelopmentofRORgtILCs.
OurdataclearlyshowthatDOCK8isrequiredforthegeneration/survival/functionofILC3.
TotestwhethertheinfusionofWTILC3couldprovideinnateimmuneprotectiononC.
rodentiuminfection,5103ILC3sortedfromRag1/IL-23Rgfp/hostwasadaptivelytransferredintoRag1/DOCK8pri/primice1hafterinfection.
DOCK8pri/primicethatreceivedtheILC3survivedanaverageof3.
5dayslongerandhadsignicantlylessweightlosscomparedwithuntreatedrecipients(SupplementaryFig.
12a,b).
IntheILC3-transferedgroup,splenicandliverbacteriacountsweresignicantlylowercomparedwiththeuntreatedgroup(SupplementaryFig.
12c).
Interestingly,themeansurvivalofDOCK8-decientmiceinfusedwithWTILC3siscomparabletothemeansurvivalobservedinmicethatlackfunctionILC3sthatwereinfusedwithWTILC3s12,39.
ItremainsunclearwhytheadoptivetransferofWTILCscouldonlyprovidepartialprotection,onepossibleexplanationisthatonlylimitednumberoftransferredILC3scanmigratetothecolonorthesecellshaveashortlifespan.
Inarecentstudy,DOCK8wasshowntointeractwithMyd88inBcells20.
However,wefoundthatMyd88wasdispensableforthegenerationofRORgtILCs(SupplementaryFig.
13).
Thisresultisconsistentwithpriorndings,inwhichMyd88/micedidnotshowsignicantalterationinthedevelopmentofRORgtILCs46.
DOCK8alsointeractswithWASpandregulatesnaturalkillercellcytotoxicity47.
However,wedidnotndsignicantreductionsofRORgtILC3inWASp/mice(SupplementaryFig.
13).
Thus,DOCK8isrequiredforRORgtILCsdevelopmentinacell-intrinsicmannerandindependentofMyd88orWASp-mediatedfunctions.
GenerationandsurvivalofRORctILCsrequiresDOCK8.
ToaddresswhyDOCK8pri/primicehavedrasticreductionsinRORgtILCs,wersttestedwhetherthesemicehadanintactcommonlymphoidprogenitors(CLPs)compartment.
RORgtILCsdevelopfromCLPs(LinSca-1lowc-KitIntIL-7Ra)48,49;thus,weevaluatedthenumberofthesecellsinthebonemarrowofWTandDOCK8pri/primice(Fig.
5a).
Interestingly,wedidnotndanymajordifferenceinthepercentageofCLPsamongWTorDOCK8pri/primice.
ThesedataareconsistentwiththefactthatDOCK8-decientmicehavenormalLNsandDOCK8mayberequiredforRORgtILCssurvivalonlyafterbirth.
Next,weassessedwhetherDOCK8-decientRORgtILCshadadevelopmentaldefectduetodefectsintranscriptionfactorsimplicatedintheirgeneration.
ThesecellsrequiretranscriptionfactorsId2(refs4,50),RORgt51andTox52fortheirdevelopment.
PreviousworkinourlaboratoryhasclearlyshownallCD90.
2IL-23RcellsareRORgtILCs34.
Therefore,wepuriedCD90.
2IL-23RandCD90.
2IL-23RcellsfromRag1/IL-23Rgfp/DOCK8pri/priandRag1/IL-23Rgfp/miceandmeasuredtranscriptsforId2,ToxandRORgtbyquantitativePCR.
However,weobservedcomparableexpressionofthesetranscriptionfactorsinCD90.
2IL-23RcellspuriedfromtheGItractofRag1/IL-23Rgfp/DOCK8pri/primiceincomparisontocontrols(SupplementaryFig.
14).
ThesedatafurthersupporttheideathatDOCK8isnotrequiredforearlydevelopmentofRORgtILCs,butmayberequiredfortheirmaintenanceorsurvival.
Recently,thearylhydrocarbonreceptor(Ahr),aligand-dependenttranscriptionfactor,wasshowntobeimportantforthegenerationandmaintenanceofadultRORgtILCs10,37,38.
Therefore,wealsoexaminedAhrtranscriptsbyquantitativePCR,butitslevelofexpressioninCD90.
2IL-23RcellswascomparabletothelevelsfoundinWTcells(SupplementaryFig.
14).
ThesedataindicatethattheeffectofDOCK8onRORgtILCsoccursindependentlyofAhrNATURECOMMUNICATIONS|DOI:10.
1038/ncomms5603ARTICLENATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications7&2014MacmillanPublishersLimited.
Allrightsreserved.
expression.
Recently,ithasbeenshownthatNotchsignallingisnecessaryforadult,butnotfetal,developmentofRORgtILCs48.
Thus,weexaminedtheexpressionofNotch-1andNotch-2transcriptinCD90.
2IL-23RcellsfromRag1/IL-23Rgfp/andRag1/IL-23Rgfp/DOCK8pri/primice.
CD90.
2IL-23RcellsfromRag1/IL-23Rgfp/GatedonLin–cellsWTDOCK8pri/priGatedonCD90.
2+RORγt+2.
43842.
63CD12720.
198.
11.
8896.
41.
290.
47Rag1–/–Rag1–/–DOCK8pri/pri1.
880088.
20011.
82.
9410.
63.
3Sca-182.
5c-Kit85.
31.
18AnnexinVAnnexinV7AADRORγtRORγt100lsotypeWTDock8pri/pri9006003000**Rag1–/–IL-23Rgfp/+Rag1–/–IL-23Rgfp/+Dock8pri/pri6,0004,0002,0000ControlControlIL-7IL-7pSTAT5(MFI)IL-7Rα(MFI)WTDOCK8pri/priMedium+WTILC3IL-7+WTILC3Medium+KOILC3IL-7+KOILC3806040%Max200%Max1008060402000102103pY-STAT51041050IL-7Rα102103104105GatedonCD45+CD3–CD90.
2+RORγt+CD90.
2Figure5|DOCK8isrequiredforthegenerationandsurvivalofRORctILCs.
(a)DotplotsshowingthebonemarrowCLPsfromWTandDOCK8pri/primice.
(b)DOCK8-decientRORgtILC3shaveenhancedapoptosis.
Rag1/andRag1/DOCK8pri/priSILPLswereexaminedbyowcytometryforRORgtandapoptosismarkers(AnnexinVand7-AAD).
Allthedataarerepresentativeofthreeindependentexperimentswiththreemicepergroup.
(c)HistogramshowingtheIL-7RaexpressiononILC3(leftpanel)andbargraphshowingtheMFIofIL-7Raexpression(rightpanel)(d)HistogramshowingthephosphorylationofSTAT5inILC3onIL-7stimulation(leftpanel)andbargraphshowingtheMFIofpSTAT5inILC3isolatedfromtheSILPLsofRag1/IL-23Rgfp/orRag1/IL-23Rgfp/DOCK8pri/primice(rightpanel).
Allthedataarerepresentativeoffourindependentexperimentswithtwotofourmicepergroup.
Thedatashownarethemean±s.
d.
PvalueswereobtainedbyStudent'st-test.
**Po0.
01.
ARTICLENATURECOMMUNICATIONS|DOI:10.
1038/ncomms56038NATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications&2014MacmillanPublishersLimited.
Allrightsreserved.
DOCK8pri/primicehadcomparableexpressionofNotch-2,butslightlyhigherexpressionofNotch-1(SupplementaryFig.
15).
WealsoassessedthetranscriptofHes1,aNotchtranscriptionaltarget,inCD90.
2IL-23Rcellsbutitsexpressionwascomparable.
However,CD90.
2IL-23RcellsfromRag1/IL-23Rgfp/DOCK8pri/primicehadsignicantlyhigherexpressionofHes1transcript(SupplementaryFig.
15,toppanel).
Moreimportantly,wetransducedbonemorrowhaematopoieticstemcells(HSCs)ofDOCK8-decientmicewithretrovirusthatcontainedconstitutivelyactiveintracellularNotchandinfusedthetransducedHSCsintoRag1IL2Rg-decientmicerecipient.
ThreeweeksafterHSCreconstitution,weanalysedthenumberofILC3inthegutofthesetransducedHSC-reconstitutedmice.
Inthesemice,thenumberofILC3wasveylimited.
Thus,theoverexpressionofintracellularNotchcouldnotrescuetheILC3compartmentintheabsenceofDOCK8(SupplementaryFig.
15,bottompanel).
Therefore,itisveryunlikelythatILC3sdefectsobservedinDOCK8-decientmiceareduetoadefectinNotchmediatedfunctions.
DOCK8hasbeenimplicatedinthesurvivalandfunctionofCD8TcellsandNKTcells25,35.
Therefore,wenextinvestigatedifDOCK8alsoinuencesthesurvivalofRORgtILCsintheGItract.
WestainedSILPLswith7-AADandAnnexinVtoexaminetheapoptoticcelldeath.
IncontrasttoCD90.
2RORgtcellsintheGItractofRag1/mice,CD90.
2RORgtcellsisolatedfromRag1/DOCK8pri/primiceconsistentlyshowedhigherAnnexinVand7-AADstaining(Fig.
5b),indicatingthatintestinalRORgtILCswithoutDOCK8aremorepronetoapoptosis.
However,theexpressionofantiapoptoticgeneBcl-2wasnotaffectedinDOCK8pri/primiceILCs,asshownbyuorescence-activatedcellsortinghistogram(SupplementaryFig.
16a).
Next,weanalysedtheexpressionofantiapoptoticgenes,Bcl-2,Bcl-xLandMcl-1inCD90.
2IL-23RcellsfromRag1/IL-23Rgfp/andRag1/IL-23Rgfp/DOCK8pri/primice.
CD90.
2IL-23RcellsfromRag1/IL-23Rgfp/DOCK8pri/primicehadlowerexpressionofBcl-xLbutnosignicantdifferenceinexpressionsofBcl-2orMcl-1(SupplementaryFig.
16b).
Thus,DOCK8maypromotethesurvivalofRORgtILCsbyinducingBcl-xLexpression.
SincethesurvivalandthegenerationofRORgtILCsarelargelydependentonIL-7(refs53,54)mediatedsignalling,weanalysedtheIL-7RaexpressiononRORgtILCs(Fig.
5c).
BothRORgtILCsisolatedfromDOCK8pri/priorWTcontrolmicehadcomparablelevelofexpressionofIL-7Ra.
Thus,RORgtILCdefectsinDOCK8-decientmicearenotduetodefectiveIL-7expression(Fig.
1f,rightpanel)inthegutordefectsinIL-7Raexpressiononthesecells.
Next,weexploredwhetherDOCK8-decientILC3hadnormalresponsetoIL-7stimula-tion.
ThusweisolatedcellsfromSILPLsofRag1/IL-23Rgfp/andRag1/IL-23Rgfp/DOCK8pri/primiceandmeasuredphosphorylationofSTAT5(pSTAT5)inresponsetoIL-7.
Tooursurprise,DOCK8-decientILC3scouldnotphosphorylateSTAT5asefcientlyasWTILC3s(Fig.
5d).
DOCK8-decientILC3shaveanimpairedIL-7mediatedsignalling,whichmaycompromisetheirsurvivalinthegut.
Alltogether,thesedatasuggestthatDOCK8iscriticalforprotectiveimmunityagainstentericpathogenbyregulatingthegenerationandsurvivalofRORgtILCsintheGItract.
DiscussionInthepresentstudy,wehavedescribedhowadeciencyinDOCK8leadstocompromisedprotectiveimmuneresponseagainstanentericpathogen.
WehaveshownthatDOCK8expressioninthehaematopoieticcompartmentisrequiredforprotectiveimmunityanditsdeciencyresultsindrasticreductionofRORgtILCsintheGItract.
WehavealsopresentedevidencethatDOCK8positivelyregulatesthesurvivalofRORgtILCsinacell-intrinsicmanner.
ThedevelopmentofRORgtILCsisdependentonseveraltranscriptionfactorsincludingId2(refs4,50),RORgt51,Tox52andIkaros55.
OurstudiesshowthatDOCK8isnotrequiredforthedevelopmentoffetalLTicellsbecauseDOCK8-decientmice,incontrasttoRORgt-decientmice,developnormalLNsandhavenormalLTicellsintheirGItract,soitisthereforelikelythatDOCK8playsanimportantroleinthemaintenanceofRORgtILCsafterbirthorinadultmice.
OurndingshaveuncoveredapreviouslyunsuspectedroleforDOCK8inprotectiveimmunityatGItractmucosa.
Recently,AhrwasshowntobeimportantforthegenerationandmaintenanceofadultRORgtILCs10,37,38.
Morerecently,thetranscriptionfactorT-bet(encodedbyTbx21)wasalsoshowntobecriticalforthedevelopmentofNKp46ILC3(refs56–58).
However,neitherAhrnorT-betisrequiredforthegenerationoffetallymphoidtissueinducer(LTi)cellsandasaresult,thesemicehavenormalLNformation.
Unliketheabove-mentionedtranscriptionfactors,DOCK8isuniqueinthatitisnotaffectingaparticularsubsetofILC3,butisrequiredforthesurvivalofallRORgtILCs,includingbothCD4andNKp46subsets.
TheabsenceofDOCK8mayinuencethemicrobialcompositionintheGItractandthatmayinuencetheRORgtILCsgenerationorsurvival.
However,itisstillunderdebatewhetherthedevelopmentofIL-22-producingILCsrequiresintestinalsymbioticmicrobes11,36,42,46,54,58.
OurmixedbonemarrowchimericexperimenthasfurtherdemonstratedthatRORgtILCsfromDOCK8-decientmiceareimpairedinacell-intrinsicmanner.
AlthoughourndingsindicatethatDOCK8isrequiredforthegenerationandsurvivalofRORgtILCs,DOCK8isalsoimportantforthesurvivalandfunctionofCD8Tcells25andNKTcells35.
Therefore,DOCK8ingeneralmayberequiredforthesurvivalofmanyimmunecells.
However,morestudiesarerequiredtodeterminebywhichmechanismsDOCK8regulatescellsurvival.
DOCK8-decientpatientsdevelopatopicdermatitis,Staphylo-coccusaureusskinabscessesorsofttissueinfections,pneumonias,elevatedserumIgE,eosinophiliaandGItractinfections21,23,59.
ThemajorityofthesepatientshavelownumbersofCD4andCD8Tcellsinthebloodandspleen21.
However,ithasnotbeeninvestigatedwhetherDOCK8-decientpatientsalsohavedefectsinILCsoralteredmicrobialcompositionintheirGItracts.
Itiscriticalbecausedysbiosisinthemicrobialcommunitycanresultininammatorydiseasesnotonlyoftheintestine,butalsooforgansatdistalsites60.
ItiswidelyacceptedthatRORgtILCsarecriticalforprotectiveimmunityatbarriersurfaces11,36.
Therefore,ourndingsthatDOCK8isrequiredforthegenerationofRORgtILCsprovidesadditionalinsightontheclinicalfeaturesandthesevereimmunodeciencyobservedinDOCK8-decientpatients.
Recentdatahaveprovidedevidencethatcommensalbacterial-derivedsignalslimitsteady-stateserumIgEconcentrationandbasophil-associatedallergicinammation61.
SinceDOCK8-decientpatientshaveanautosomalrecessiveformofhyperimmunoglobulinemiaEsyndromecharacterizedbyrecurrentinfections,increasedsusceptibilitytoatopiceczemaandaverageserumIgEconcentrations10timeshigherthanthosefoundincontrolsubjects21,22,itistemptingtospeculatethattheincreasedsusceptibilitytobacterialinfectionandtheelevatedlevelsofIgEseeninDOCK8-decientpatientscouldbetheresultofdysbiosisinthemicrobialcommunityorinadequatecontrolofpathogensduetoadefectinthesurveillanceoftheGItractbyILCs.
Inadditiontoinfectionsusceptibility,DOCK8-decientpatientsalsoNATURECOMMUNICATIONS|DOI:10.
1038/ncomms5603ARTICLENATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications9&2014MacmillanPublishersLimited.
Allrightsreserved.
havehighersusceptibilitytocancer21,23,59.
Usingamousemodel,ithasbeenshownthatNKp46LTicellsareessentialformelanomarejectionmediatedbyIL-12(ref.
62).
Intheabovestudy,authorsshowedthatIL-12-mediatedrepressionofsubcutaneousmelanomaisindependentofadaptiveimmunityorcNKcells,butrequiresinnatelinease-negativeNKp46LTicells.
TumourrejectionbyNKp46LTiwasindependentofIFN-g,IL-22,lymphotoxin,perforinorIL-17,butrequiredtheexpressionofadhesionmoleculesintumourvessels,whichcouldfacilitatetheabilityofleukocytestoevadethetumour62.
Thus,ourndingsthatDOCK8-decientmicehavedrasticreductionsinRORgtILCscouldexplainwhyDOCK8-decientpatientsaremorepronetocancer.
Insummary,ourdatashowthatDOCK8iscriticalforprotectiveimmunityagainstentericpathogeninfection,inde-pendentofitsroleinB-cell-mediatedfunctions.
OurndingshaverevealedanunanticipatedDOCK8requirementforthegenerationandsurvivalofRORgtILCs.
WealsoprovideevidencethatdrasticreductionsinthenumbersofRORgtILCs,aswellasdefectiveIL-22productionbythesecells,areresponsibleforearlymortalityonC.
rodentiuminfectioninDOCK8-decientmice.
Toourknowledge,thisistherststudythathasidentiedthescaffoldingproteinDOCK8asanimportantmoleculeforthegeneration,survivalandoptimalfunctionofILCsandprotectiveimmunityagainstentericpathogen.
GiventhatDOCK8deciencyinhumanpatientsresultsinrecurrentinfectionsandincreasedvulnerabilitytocancer,theseconsequencesmaypartlybeduetodefectiveRORgtILCsintheGItract.
MethodsMice.
Micewerebredandmaintainedunderspecicpathogen-freeconditionsinaccordancewithanimalfacilitiesattheSeattleChildren'sResearchInstitute(SCRI).
ExperimentswereconductedinaccordancewiththeguidelinesoftheSCRIAnimalCareandUseCommittee.
IL-23R/andDOCK8-decient(DOCK8pri/pri)miceweregeneratedaspreviouslydescribed24,45.
WT(C57BL/6)andRag1/micewerepurchasedfromJacksonLaboratories(BarHarbour,ME),whileRag2/IL-2Rg/micewerepurchasedfromTaconic.
Mostoftheanimalsusedinthestudywere6–8-weeksoldmaleandfemalemice.
C.
rodentiuminfectionandassessmentofCFU.
C.
rodentiumwasculturedovernightat37°CinLuriabrothwithgentleagitation.
Afterspinningdownat4,500r.
p.
m.
,theyweresuspendedinPBSandtheconcentrationwasmeasuredusinga600nmcelldensitymetre(WPABiowaveCO8000).
Micewerefastedfor8hbeforeoralgavageof2109colony-formingunitsin200mlPBS.
Spleensandliversofinfectedmicewereusedtoanalysethecolony-formingunits.
SpleensandliversweremechanicallycrushedinPBSandafterserialdilution,platedonLBagarplatesandincubatedovernightat37°C.
Bonemarrowtransfer.
Donorbonemarrowcells(5106)wereintravenouslyinjectedintorecipientmiceirradiatedat550rads.
Irradiatedmiceweretreatedwithantibiotics(Baytril)for2weeksafterinjectionandmicewereinfected6weeksposttransfer.
Mixedbonemarrowtransfer.
RecipientRag2/IL-2Rg/micewereirra-diated4hbeforetransferat550radandtreatedwithantibiotics(Baytril)for2weeks.
Irradiatedmicereceived50/50mixtureofCD45.
1WTandCD45.
2DOCK8pri/pribonemarrowcells(5106)intravenouslyandmicewereanalysedforRORgtILCs5weeksposttransfer.
IsolationofLPcells.
ForGItractLPLisolation,colonsandSIwereisolated,ushedwithice-coldPBS,theattachedfatremovedandthencutopenlong-itudinally.
Thesampleswerecleanedbyshakinginice-coldPBSfourtimesbeforetissuewascutinto1cmpieces.
Theepithelialcellswereremovedbyincubatingthetissueinpredigestionbuffer(HBSS,2mMEDTAandantibiotics)for30minat37°Cwithshaking.
TheLPLcellswereisolatedbyincubatingthetissuesindigestionbuffer(HBSS,10%fetalbovineserum,1mgml1Collagenase(Wor-thington)andDNase(Sigma-Aldrich)for20min.
Thedigestedtissueswerethenlteredthrougha40-mmlter.
Thisstepwasrepeatedtogetthecompletedigestionofthetissues.
Cellswerewashedonceintheice-coldmedium(RPMI,10%fetalbovineserumandantibiotics),resuspendedin10mlofthe40%fractionofa40:80Percollgradientandoverlaidon5mlofthe80%fractionina15mlFalcontube.
Percollgradientseparationwasperformedbycentrifugationfor20minat1,800r.
p.
m.
atroomtemperature.
LPLcellswerecollectedattheinterphaseofthePercollgradient,washedandresuspendedinmedium.
Flowcytometricanalysis.
Forowcytometricanalysis,cellswerestainedwithantibodiestothefollowingmarkers:CD3((FITCorPE,eBioscience,alexauor647,BioLegend)(1:250dilution)),CD90.
2((PE-Cy5,PE-Cy7Biolegend;1:2,000dilution)),CD4((PE,PE-Cy5,APC-euor780oreuor450,eBioscience;1:250dilution)),IL-22((APC,eBioscience;1:125dilution)),RORgt((PE,eBioscience;1:200dilution)),CD127((PEorAPC-euor780,eBioscience),NKp46(eour660,eBiosciences),c-Kit(PE-Cy7,BDBiosciences;1:125dilution)),SCA-1((PerCP-Cy5.
5,eBiosciences;1:500dilution)),CD8((PE-Cy7,eBioscience;1:2,500dilu-tion)),CD19((FITC,eBioscience),B220(Alexauor647,Biolegend;1:500dilu-tion)),MHCclassII((PE,eBioscience),CD11c(APC,eBioscience)andCD11b(APC-euor780,eBioscience;1:500dilution)).
Forcytokineproduction,cellswerestimulateddirectlyexvivobyincubatingfor4hwith20ngml1IL-23(eBioscience)inthepresenceofGolgiplug(BD).
Cellswerexedandpermeabi-lizedbyusingFoxp3x/permbuffersetCat#421403(Biolegend),asdescribedbymanufacturersandstainedwithIL-22andRORgt.
ForpSTAT3,cellsweresti-mulatedbyincubatingfor20minwith50ngml1rIL-23(eBioscience)followedbyxationwith2%paraformaldehyde.
Paraformaldehyde-xedcellswereper-meabilizedwithphosowpermbufferIII(BDbiosciences)andstainedwithanti-STAT3(pY705)antibodyconjugatedtoAlexauor647(BDbiosciences).
Geneexpression.
TotalRNAwasextractedwithTrizol(Invitrogen,Carlsbad,CA)andreversetranscribedwiththeiScriptcDNAsynthesiskit(BioRad,Hercules,CA).
TherealtimePCRwasconductedusingStepOnePlusRealTimePCRsystem(AppliedBiosystems)withprimerandprobelistedinSupplementaryTable1.
ThereactionwassetusingTaqManFastUniversalPCRMasterMixCat#43552042(AppliedBiosystems).
ThesampleswerenormalizedtothecontrolhousekeepinggeneRPL-19andreportedaccordingtothedCTmethod.
MeasurementofIL-22byELISA.
SortedSILPLs(3,000)wereculturedintheabsenceorpresenceof20ngml1rIL-23(eBioscienceCA,USA)for24h.
ThesupernatantwascentrifugedandmeasuredforIL-22proteinconcentrationbysandwichELISAkit#88-7522-86(eBioscienceCA,USA).
HydrodynamicgenedeliveryofIL-22.
Forhydrodynamicinjections,15mgplasmidDNAwasresuspendedin2mlofPBSandintravenouslyinjectedin5–10susinghydrodynamictailveininjection.
IL-22proteinproductionwasvalidatedfrombloodofmicethathadreceivedthehydrodynamicplasmidinjections.
RNAsequencing.
TotalRNAwaspuriedfromC.
rodentiuminfectedcolonwithRNeasyminikit(Qiagen).
RNAsequencingwasperformedatGenomicCoreFacilitySouthwesternMedicalCenter,UniversityofTexas.
Statisticalanalysis.
Forstatisticalanalysis,Pvalueswerecalculatedwithatwo-tailedunpairedStudent'st-test.
Pvaluesoflessthan0.
05wereconsideredsignicant.
Errorbarsdenote±s.
d.
References1.
Backhed,F.
,Ley,R.
E.
,Sonnenburg,J.
L.
,Peterson,D.
A.
&Gordon,J.
I.
Host-bacterialmutualisminthehumanintestine.
Science307,1915–1920(2005).
2.
Hooper,L.
V.
&Macpherson,A.
J.
Immuneadaptationsthatmaintainhomeostasiswiththeintestinalmicrobiota.
Nat.
Rev.
Immunol.
10,159–169(2010).
3.
Spits,H.
&Cupedo,T.
Innatelymphoidcells:emerginginsightsindevelopment,lineagerelationships,andfunction.
Annu.
Rev.
Immunol.
30,647–675(2012).
4.
Yokota,Y.
etal.
Developmentofperipherallymphoidorgansandnaturalkillercellsdependsonthehelix-loop-helixinhibitorId2.
Nature397,702–706(1999).
5.
Spits,H.
&DiSanto,J.
P.
Theexpandingfamilyofinnatelymphoidcells:regulatorsandeffectorsofimmunityandtissueremodeling.
Nat.
Immunol.
12,21–27(2011).
6.
Walker,J.
A.
,Barlow,J.
L.
&McKenzie,A.
N.
Innatelymphoidcells–howdidwemissthemNat.
Rev.
Immunol.
13,75–87(2013).
7.
TaitWojno,E.
D.
&Artis,D.
Innatelymphoidcells:balancingimmunity,inammation,andtissuerepairintheintestine.
CellHostMicrobe12,445–457(2012).
8.
Eberl,G.
Induciblelymphoidtissuesintheadultgut:recapitulationofafetaldevelopmentalpathwayNat.
Rev.
Immunol.
5,413–420(2005).
9.
Cella,M.
etal.
AhumannaturalkillercellsubsetprovidesaninnatesourceofIL-22formucosalimmunity.
Nature457,722–725(2009).
ARTICLENATURECOMMUNICATIONS|DOI:10.
1038/ncomms560310NATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications&2014MacmillanPublishersLimited.
Allrightsreserved.
10.
Kiss,E.
A.
etal.
Naturalarylhydrocarbonreceptorligandscontrolorganogenesisofintestinallymphoidfollicles.
Science334,1561–1565(2011).
11.
Sawa,S.
etal.
RORgammatinnatelymphoidcellsregulateintestinalhomeostasisbyintegratingnegativesignalsfromthesymbioticmicrobiota.
Nat.
Immunol.
12,320–326(2011).
12.
Sonnenberg,G.
F.
,Monticelli,L.
A.
,Elloso,M.
M.
,Fouser,L.
A.
&Artis,D.
CD4()lymphoidtissue-inducercellspromoteinnateimmunityinthegut.
Immunity34,122–134(2011).
13.
Tumanov,A.
V.
etal.
LymphotoxincontrolstheIL-22protectionpathwayingutinnatelymphoidcellsduringmucosalpathogenchallenge.
CellHostMicrobe10,44–53(2011).
14.
Zenewicz,L.
A.
etal.
Innateandadaptiveinterleukin-22protectsmicefrominammatoryboweldisease.
Immunity29,947–957(2008).
15.
Pickert,G.
etal.
STAT3linksIL-22signalinginintestinalepithelialcellstomucosalwoundhealing.
J.
Exp.
Med.
206,1465–1472(2009).
16.
Zheng,Y.
etal.
Interleukin-22mediatesearlyhostdefenseagainstattachingandeffacingbacterialpathogens.
Nat.
Med.
14,282–289(2008).
17.
Satpathy,A.
T.
etal.
Notch2-dependentclassicaldendriticcellsorchestrateintestinalimmunitytoattaching-and-effacingbacterialpathogens.
Nat.
Immunol.
14,937–948(2013).
18.
Cote,J.
F.
&Vuori,K.
IdenticationofanevolutionarilyconservedsuperfamilyofDOCK180-relatedproteinswithguaninenucleotideexchangeactivity.
J.
Cell.
Sci.
115,4901–4913(2002).
19.
Ruusala,A.
&Aspenstrom,P.
IsolationandcharacterisationofDOCK8,amemberoftheDOCK180-relatedregulatorsofcellmorphology.
FEBSLett.
572,159–166(2004).
20.
Jabara,H.
H.
etal.
DOCK8functionsasanadaptorthatlinksTLR-MyD88signalingtoBcellactivation.
Nat.
Immunol.
13,612–620(2012).
21.
Zhang,Q.
etal.
CombinedimmunodeciencyassociatedwithDOCK8mutations.
NewEngl.
J.
Med.
361,2046–2055(2009).
22.
Engelhardt,K.
R.
etal.
Largedeletionsandpointmutationsinvolvingthededicatorofcytokinesis8(DOCK8)intheautosomal-recessiveformofhyper-IgEsyndrome.
J.
AllergyClin.
Immunol.
124,1289–1302e1284(2009).
23.
Su,H.
C.
Dedicatorofcytokinesis8(DOCK8)deciency.
Curr.
Opin.
AllergyClin.
Immunol.
10,515–520(2010).
24.
Randall,K.
L.
etal.
Dock8mutationscrippleBcellimmunologicalsynapses,germinalcentersandlong-livedantibodyproduction.
Nat.
Immunol.
10,1283–1291(2009).
25.
Randall,K.
L.
etal.
DOCK8deciencyimpairsCD8Tcellsurvivalandfunctioninhumansandmice.
J.
Exp.
Med.
208,2305–2320(2011).
26.
Harada,Y.
etal.
DOCK8isaCdc42activatorcriticalforinterstitialdendriticcellmigrationduringimmuneresponses.
Blood119,4451–4461(2012).
27.
Simmons,C.
P.
etal.
CentralroleforBlymphocytesandCD4TcellsinimmunitytoinfectionbytheattachingandeffacingpathogenCitrobacterrodentium.
Infect.
Immun.
71,5077–5086(2003).
28.
Maaser,C.
etal.
ClearanceofCitrobacterrodentiumrequiresBcellsbutnotsecretoryimmunoglobulinA(IgA)orIgMantibodies.
Infect.
Immun.
72,3315–3324(2004).
29.
Bry,L.
&Brenner,M.
B.
CriticalroleofTcell-dependentserumantibody,butnotthegut-associatedlymphoidtissue,forsurvivingacutemucosalinfectionwithCitrobacterrodentium,anattachingandeffacingpathogen.
J.
Immunol.
172,433–441(2004).
30.
Mangan,P.
R.
etal.
Transforminggrowthfactor-betainducesdevelopmentoftheT(H)17lineage.
Nature441,231–234(2006).
31.
Basu,R.
etal.
Th22cellsareanimportantsourceofIL-22forhostprotectionagainstenteropathogenicbacteria.
Immunity37,1061–1075(2012).
32.
Sonnenberg,G.
F.
,Fouser,L.
A.
&Artis,D.
Borderpatrol:regulationofimmunity,inammationandtissuehomeostasisatbarriersurfacesbyIL-22.
Nat.
Immunol.
12,383–390(2011).
33.
Colonna,M.
Interleukin-22-producingnaturalkillercellsandlymphoidtissueinducer-likecellsinmucosalimmunity.
Immunity31,15–23(2009).
34.
Eken,A.
,Singh,A.
K.
,Treuting,P.
M.
&Oukka,M.
IL-23Rinnatelymphoidcellsinducecolitisviainterleukin-22-dependentmechanism.
MucosalImmunol.
7,143–154(2013).
35.
Crawford,G.
etal.
DOCK8iscriticalforthesurvivalandfunctionofNKTcells.
Blood122,2052–2061(2013).
36.
Satoh-Takayama,N.
etal.
Microbialoradrivesinterleukin22productioninintestinalNKp46cellsthatprovideinnatemucosalimmunedefense.
Immunity29,958–970(2008).
37.
Qiu,J.
etal.
Thearylhydrocarbonreceptorregulatesgutimmunitythroughmodulationofinnatelymphoidcells.
Immunity36,92–104(2012).
38.
Lee,J.
S.
etal.
AHRdrivesthedevelopmentofgutILC22cellsandpostnatallymphoidtissuesviapathwaysdependentonandindependentofNotch.
Nat.
Immunol.
13,144–151(2012).
39.
Guo,X.
etal.
Inductionofinnatelymphoidcell-derivedinterleukin-22bythetranscriptionfactorSTAT3mediatesprotectionagainstintestinalinfection.
Immunity40,25–39(2014).
40.
Parham,C.
etal.
AreceptorfortheheterodimericcytokineIL-23iscomposedofIL-12Rbeta1andanovelcytokinereceptorsubunit,IL-23R.
J.
Immunol.
168,5699–5708(2002).
41.
Spits,H.
etal.
Innatelymphoidcells—aproposalforuniformnomenclature.
Nat.
Rev.
Immunol.
13,145–149(2013).
42.
Sanos,S.
L.
etal.
RORgammatandcommensalmicrooraarerequiredforthedifferentiationofmucosalinterleukin22-producingNKp46cells.
Nat.
Immunol.
10,83–91(2009).
43.
Buonocore,S.
etal.
Innatelymphoidcellsdriveinterleukin-23-dependentinnateintestinalpathology.
Nature464,1371–1375(2010).
44.
Powell,N.
etal.
ThetranscriptionfactorT-betregulatesintestinalinammationmediatedbyinterleukin-7receptorinnatelymphoidcells.
Immunity37,674–684(2012).
45.
Awasthi,A.
etal.
Cuttingedge:IL-23receptorgfpreportermicerevealdistinctpopulationsofIL-17-producingcells.
J.
Immunol.
182,5904–5908(2009).
46.
Reynders,A.
etal.
Identity,regulationandinvivofunctionofgutNKp46RORgammatandNKp46RORgammat-lymphoidcells.
EMBOJ.
30,2934–2947(2011).
47.
Ham,H.
etal.
Dedicatorofcytokinesis8interactswithtalinandWiskott-AldrichsyndromeproteintoregulateNKcellcytotoxicity.
J.
Immunol.
190,3661–3669(2013).
48.
Possot,C.
etal.
Notchsignalingisnecessaryforadult,butnotfetal,developmentofRORgammat()innatelymphoidcells.
Nat.
Immunol.
12,949–958(2011).
49.
Mebius,R.
E.
etal.
Thefetallivercounterpartofadultcommonlymphoidprogenitorsgivesrisetoalllymphoidlineages,CD45CD4CD3cells,aswellasmacrophages.
J.
Immunol.
166,6593–6601(2001).
50.
Boos,M.
D.
,Yokota,Y.
,Eberl,G.
&Kee,B.
L.
Maturenaturalkillercellandlymphoidtissue-inducingcelldevelopmentrequiresId2-mediatedsuppressionofEproteinactivity.
J.
Exp.
Med.
204,1119–1130(2007).
51.
Eberl,G.
etal.
AnessentialfunctionforthenuclearreceptorRORgamma(t)inthegenerationoffetallymphoidtissueinducercells.
Nat.
Immunol.
5,64–73(2004).
52.
Aliahmad,P.
,delaTorre,B.
&Kaye,J.
ShareddependenceontheDNA-bindingfactorTOXforthedevelopmentoflymphoidtissue-inducercellandNKcelllineages.
Nat.
Immunol.
11,945–952(2010).
53.
Satoh-Takayama,N.
etal.
IL-7andIL-15independentlyprogramthedifferentiationofintestinalCD3-NKp46cellsubsetsfromId2-dependentprecursors.
J.
Exp.
Med.
207,273–280(2010).
54.
Vonarbourg,C.
etal.
RegulatedexpressionofnuclearreceptorRORgammatconfersdistinctfunctionalfatestoNKcellreceptor-expressingRORgammat()innatelymphocytes.
Immunity33,736–751(2010).
55.
Wang,J.
H.
etal.
SelectivedefectsinthedevelopmentofthefetalandadultlymphoidsysteminmicewithanIkarosnullmutation.
Immunity5,537–549(1996).
56.
Sciume,G.
etal.
DistinctrequirementsforT-betingutinnatelymphoidcells.
J.
Exp.
Med.
209,2331–2338(2012).
57.
Rankin,L.
C.
etal.
ThetranscriptionfactorT-betisessentialforthedevelopmentofNKp46innatelymphocytesviatheNotchpathway.
Nat.
Immunol.
14,389–395(2013).
58.
Klose,C.
S.
etal.
AT-betgradientcontrolsthefateandfunctionofCCR6-RORgammatinnatelymphoidcells.
Nature494,261–265(2013).
59.
Sanal,O.
etal.
AdditionaldiversendingsexpandtheclinicalpresentationofDOCK8deciency.
J.
Clin.
Immunol.
32,698–708(2012).
60.
Littman,D.
R.
&Pamer,E.
G.
Roleofthecommensalmicrobiotainnormalandpathogenichostimmuneresponses.
CellHostMicrobe10,311–323(2011).
61.
Hill,D.
A.
etal.
Commensalbacteria-derivedsignalsregulatebasophilhematopoiesisandallergicinammation.
Nat.
Med.
18,538–546(2012).
62.
Eisenring,M.
,vomBerg,J.
,Kristiansen,G.
,Saller,E.
&Becher,B.
IL-12initiatestumorrejectionvialymphoidtissue-inducercellsbearingthenaturalcytotoxicityreceptorNKp46.
Nat.
Immunol.
11,1030–1038(2010).
AcknowledgementsThisworkwasfundedbytheUSNationalInstitutesofHealthgrantR21(1R21AI101936),toM.
O.
WethankChristopherC.
Goodnow(JohnCurtinSchoolofMedicalResearchandAustralianPhenomicsFacility,AustralianNationalUniversity,Australia)forprovidingDOCK8pri/primice,DavidRawlings(SeattleChildren'sResearchInstitute,Seattle,Washington,USA)forprovidingmMT,WASp/andMyd88/mice,IaanisAifantis(DepartmentofPathology,NewYorkUniversity,NY,USA)forprovidingintracellularNotchretrovirusconstructandSimonGlatigny(BenaroyaResearchInstitute,ImmunologyProgram,Seattle,WA,USA)forhelpingwithRNAsequencing.
WewouldalsoliketothankDawnMarieParesforheroutstandingadministrativeassistance.
NATURECOMMUNICATIONS|DOI:10.
1038/ncomms5603ARTICLENATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications11&2014MacmillanPublishersLimited.
Allrightsreserved.
AuthorcontributionsA.
K.
S.
designed,performedandanalysedmostexperimentsandwrotethemanuscript.
A.
E.
andM.
F.
performedandanalysedtheIL-22rescueexperiments.
E.
B.
designedandanalysedtheexperiments.
M.
O.
supervisedthestudy,analysedthedataandeditedthemanuscript.
AdditionalinformationAccessioncodes:RNA-sequencingdatahavebeendepositedinthegeneexpressionomnibus(GEO)databaseunderaccessioncodeGSE58765.
SupplementaryInformationaccompaniesthispaperathttp://www.
nature.
com/naturecommunicationsCompetingnancialinterests:Theauthorsdeclarenocompetingnancialinterests.
Reprintsandpermissioninformationisavailableonlineathttp://npg.
nature.
com/reprintsandpermissions/Howtocitethisarticle:Singh,A.
K.
etal.
DOCK8regulatesprotectiveimmunitybycontrollingthefunctionandsurvivalofRORgtILCs.
Nat.
Commun.
5:4603doi:10.
1038/ncomms5603(2014).
ARTICLENATURECOMMUNICATIONS|DOI:10.
1038/ncomms560312NATURECOMMUNICATIONS|5:4603|DOI:10.
1038/ncomms5603|www.
nature.
com/naturecommunications&2014MacmillanPublishersLimited.
Allrightsreserved.