ebook img

Induction of Antiviral Immune Response through Recognition of the Repeating Subunit Pattern of PDF

17 Pages·2017·1.42 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Induction of Antiviral Immune Response through Recognition of the Repeating Subunit Pattern of

RESEARCH ARTICLE crossm Induction of Antiviral Immune Response through Recognition of the Repeating Subunit Pattern of Viral Capsids Is Toll- Like Receptor 2 Dependent D KellyM.Shepardson,aBenjaminSchwarz,b*KyleLarson,aRachelleV.Morton,a o JohnAvera,bKimberlyMcCoy,bAlaynaCaffrey,aAnnHarmsen,aTrevorDouglas,b w n AgnieszkaRynda-Applea lo DepartmentofMicrobiologyandImmunology,MontanaStateUniversity,Bozeman,Montana,USAa; a d DepartmentofChemistry,IndianaUniversity,Bloomington,Indiana,USAb e d ABSTRACT Although viruses and viral capsids induce rapid immune responses, little fr o is known about viral pathogen-associated molecular patterns (PAMPs) that are ex- Received1August2017 Accepted10 m hibited on their surface. Here, we demonstrate that the repeating protein subunit October2017 Published14November2017 h CitationShepardsonKM,SchwarzB,LarsonK, t pattern common to most virus capsids is a molecular pattern that induces a Toll- MortonRV,AveraJ,McCoyK,CaffreyA, tp like-receptor-2 (TLR2)-dependent antiviral immune response. This early antiviral im- HarmsenA,DouglasT,Rynda-AppleA.2017. :// m mune response regulates the clearance of subsequent bacterial superinfections, Inductionofantiviralimmuneresponse throughrecognitionoftherepeatingsubunit b which are a primary cause of morbidities associated with influenza virus infections. patternofviralcapsidsisToll-likereceptor2 io Utilizing this altered susceptibility to subsequent bacterial challenge as an outcome, dependent.mBio8:e01356-17.https://doi.org/ .a s we determined that multiple unrelated, empty, and replication-deficient capsids initi- 10.1128/mBio.01356-17. m ated early TLR2-dependent immune responses, similar to intact influenza virus or InvitedEditorDennisW.Metzger,Albany .o MedicalCollege r murine pneumovirus. These TLR2-mediated responses driven by the capsid were not g EditorMaryK.Estes,BaylorCollegeof / dependent upon the capsid’s shape, size, origin, or amino acid sequence. However, Medicine o n they were dependent upon the multisubunit arrangement of the capsid proteins, Copyright©2017Shepardsonetal.Thisisan A because unlike intact capsids, individual capsid subunits did not enhance bacterial open-accessarticledistributedundertheterms p clearance. Further, we demonstrated that even a linear microfilament protein built oInfttehrneaCtrioenatailvleicCeonmsem. onsAttribution4.0 ril from repeating protein subunits (F-actin), but not its monomer (G-actin), induced 4 AddresscorrespondencetoAgnieszkaRynda- , similar kinetics of subsequent bacterial clearance as did virus capsid. However, al- Apple,[email protected]. 2 0 though capsids and F-actin induced similar bacterial clearance, in macrophages they *Presentaddress:BenjaminSchwarz,Rocky 1 MountainLaboratories,Hamilton,Montana, 9 required distinct TLR2 heterodimers for this response (TLR2/6 or TLR2/1, respectively) USA. b and different phagocyte populations were involved in the execution of these re- y K.M.S.andB.S.contributedequallytothiswork. sponses in vivo. Our results demonstrate that TLR2 responds to invading viral parti- g u cles that are composed of repeating protein subunits, indicating that this common e s architectureofviruscapsidsisapreviouslyunrecognizedmolecularpattern. t IMPORTANCE Rapid and precise pathogen identification is critical for the initiation of pathogen-specific immune responses and pathogen clearance. Bacteria and fungi express common molecular patterns on their exteriors that are recognized by cell surface-expressed host pattern recognition receptors (PRRs) prior to infection. In contrast, viral molecular patterns are primarily nucleic acids, which are recognized after virus internalization. We found that an initial antiviral immune response is in- duced by the repeating subunit pattern of virus exteriors (capsids), and thus, induc- tion of this response is independent of viral infection. This early response to viral capsids required the cell surface-expressed PRR TLR2 and allowed for improved clearance of subsequent bacterial infection that commonly complicates respiratory viral infections. Since the repeating protein subunit pattern is conserved across viral capsids, this suggests that it is not easy for a virus to change without altering fit- ness. Targeting this vulnerability could lead to development of a universal antiviral vaccine. ® November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 1 ® Shepardsonetal. KEYWORDS patternrecognition,bacterialsuperinfection,innateimmunity,Toll-like receptors,virology,virus-hostinteractions Rapidrecognitionofpathogensbytheimmunesystemisessentialfortheinitiation ofanefficientimmuneresponse.Naturehasevolvedadelicatesystemofimmunity that allows for the recognition of harmful and/or damaging entities based on con- served elements that mark them as dangerous (1, 2). This early immune response depends on a pathogen recognition event mediated by the host pattern recognition receptors(PRRs)thatcaneitherbesecreted,expressedonthecellsurface,orexpressed intracellularly(1).Regardlessoftheircellularlocation,PRRsidentifyspecificmolecules commonamongforeignorganisms,knownaspathogen-associatedmolecularpatterns (PAMPs)(3).Formanybacteriaandfungi,therearewell-knowncellsurface-expressed PRRs,suchasToll-likereceptors(TLRs),thatareabletoalerttheimmunesystemtoa D o threat before infection can be established. In contrast, viruses are typically identified w after internalization, based on their nucleic acid recognition by intracellular PRRs (4). n lo While early reports implicated cell surface-expressed TLRs in immune responses to a viruses, including respiratory syncytial virus (RSV) or Epstein-Barr virus (EBV) (5, 6), d e classicalexternalPAMPsarenotknowntobepresentoneithertheviralcapsidorthe d f viral envelope, the most external structures of nonenveloped and enveloped viruses, r o respectively(4).Consequently,littleisknownaboutcommonexternalviralPAMPsand m howtheycouldbeinvolvedinearlyantiviralimmuneresponses. h t To address this gap in our understanding of virus recognition, we utilized empty tp : capsids, the protein shells of viruses, as a probe to determine whether and how the // m immunesystemrespondstotheinitialrecognitionoftheviruspriortoestablishment b of infection. Capsids are composed of repeating copies of either a single protein io . subunit or a small number of protein subunits (7). Importantly for this study, capsids a s retainthebasicexternalarchitectureofavirus,butaredevoidofcommonviralPAMPs m . suchasnucleicacids(4).Forthesereasons,capsidsprovideausefultoolforexamining o r howthehostmayinitiallysenseexternalviralstructure. g / Previously,weobservedsimilaritiesbetweentheimmuneresponsesinducedbyan o n infectious enveloped virus and by empty protein capsids in the lungs of mice (8–10). A We showed that antigen-nonspecific priming of the murine lung by a capsid derived p r fromtheSalmonellabacteriophageP22resultedininnateimmuneimprintingsimilarto il 4 influenza A virus (IAV) infection, and that this imprinting accelerated the immune , 2 responsetolethalIAVchallenge(8,11,12).Wealsoshowedthatevenanon-pathogen- 0 associatedcapsid,smallheatshockprotein(sHSP)fromMethanocaldococcusjannaschii, 1 9 induced immune responses similar to IAV and protected mice from a range of respi- b ratorypathogens,includingIAV,severeacuterespiratorysyndrome(SARS)coronavirus, y g Coxiellaburnetii,andStaphylococcusaureus(9–14).Immuneimprintingbyvirusescan u e eitherreduceorenhancehostsusceptibilitytosubsequentbacterialinfections(15,16). s t Indeed,ourprecedinginvestigationrevealedthatprotectionfromS.aureuschallenge is time-dependent after IAV, with enhanced bacterial clearance 2 to 3 days post-IAV infectionandreducedclearance7dayspost-IAVinfection(11,14).ConsideringthatIAV andthecapsidsinducedsimilarinitialimmuneresponses(9),wehypothesizedthatthe time-dependentchangeinhostsusceptibilitytosubsequentS.aureuschallenge,which wereportedtooccurafterIAVinfection,isnotlimitedtoIAVinfection.Therefore,we speculatedthatthisresponseisinducedbyrecognitionofcertaincommonpatternson thevirusexteriorbyhostPRRsindependentlyofinfection. Here,wereportinductionofasimilartime-dependentchangeinhostsusceptibility topost-IAVS.aureusinfection(14)afterinoculationofmicewithvariousemptycapsids unrelatedtoIAV.Thisprovidedustheopportunitytousethischangeinhostsuscep- tibilitytoasubsequentS.aureusinfectionasareadoutforthecapsid-inducedimmune responses (11, 14). We found that recognition of the multisubunit protein assembly architecture of virus capsids by the cell surface-expressed PRR TLR2 was required for inductionoftheseresponses.Ourinvestigationalsorevealedthatthistime-dependent November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 2 ® EarlyAntiviralImmuneResponseRequiresTLR2 changeinhostsusceptibilitytoasubsequentS.aureusinfectionwasinducedregardless ofthesize,shape,origin,oraminoacidsequenceoftheviruscapsid,butthatinduction ofthisresponserequiredtherepeatingproteinsubunitarrangementcommontothe majorityof,ifnotall,viruscapsids.Importantly,becausetherepeatingarrangementof protein subunits extends beyond just capsid structure, we also found that linear multisubunit protein assemblies not derived from viruses, such as actin filaments (F-actin),alsoregulatedhostsusceptibilitytoasubsequentS.aureuschallenge.While wedeterminedthatTLR2wasrequiredforinitiationoftheseresponsestobothcapsular andlinearrepeatingproteinassemblies,differentTLR2heterodimers(TLR2/6orTLR2/1) and different effector cells (neutrophils and macrophages) were involved in the exe- cutionoftheseresponseswheninducedbyeitherviruscapsidsorF-actin. Taken together, these data support a novel role for TLR2 as a receptor that recognizesrepeatingproteinsubunitpatternsregardlessoftheproteinsinvolved.The D significance of TLR2-mediated recognition of multivalent protein subunit architecture o expandsourcurrentunderstandingofimmunitytovirusesanditsimpactonimmunity w n tobacteria,andprovidesanewlayerofmotif-directedrecognition.Importantly,these lo datademonstratethattherecognitionofviralcapsidsisindependentofinfectionand a d establishesTLR2’sroleintheinitialrecognitionofviruses. e d f r RESULTS o m Exposure to empty virus capsid from P22 bacteriophage induces time- h dependent kinetics of S. aureus bacterial clearance, similar to IAV. Building on t t previousresultsdemonstratingeithersHSPcapsid-orIAV-inducedclearanceofsubse- p : / quent S. aureus challenge, we examined another capsid, derived from the bacterio- /m phageP22,andanaturalmousevirustoensurethatthisphenomenonwasnotcapsid b orvirusspecific.AswithsHSPcapsid,IAVandP22capsidssharenoproteinsimilarity, io . a althoughtheydoshareasimilarsphericalshapewithapolyvalentproteinsurface(17) s m (Fig. 1A). The ability of the host to clear secondary post-IAV S. aureus pneumonia . changes over time (11, 14), and we hypothesized that this time dependence and the o r g associatedimmuneresponsesmaybeutilizedtocharacterizesimilaritiesintherecog- / nition of the virus and the capsid. To test this initial hypothesis, wild-type (WT) mice o n wereinoculatedwitheitherIAVorP22capsidandthenchallengedwithS.aureus3,7, A or14dayslater(Fig.1B).Ascapsidsarereplicationdeficient,weinoculatedmicewith p r 3consecutivedailydoses,mimickingthereplicationofaninfectiousvirus. il 4 Compared to mock-inoculated mice, mice inoculated with P22 capsid displayed a , 2 similar increase in clearance of S. aureus as did IAV-inoculated mice at day 3 post- 0 1 inoculation (p.i.) (Fig. 1C). By day 7 p.i., both P22 capsid- and IAV-inoculated mice 9 showed at least a 10-fold decrease in bacterial clearance (Fig. 1C). At day 14 p.i., b y IAV-inoculatedmice,whichbythispointwouldstillhaveresultingimmuneeffectsfrom g a productive viral infection (18, 19), exhibited improved bacterial clearance, whereas u e bacterial clearance in P22 capsid-inoculated mice was similar to that of mock- s t inoculated mice (Fig. 1C). Inoculation of mice with pneumonia virus of mice (PVM), a causeofnaturalmouseviralinfectionthatresultsin/inducespathologysimilartothat causedbyhumanrespiratorysyncytialvirus(20),resultedinasimilartime-dependent clearance of S. aureus challenge as found with P22 capsids or IAV (see Fig. S1 in the supplementalmaterial). P22capsidinoculationinducescytokineandcellularimmuneresponses,similarto IAV, both before and after S. aureus challenge. To determine whether the similar patterns of bacterial clearance induced by IAV and by P22 capsid inoculation are also associatedwithacommoncytokineresponse,wecomparedlevelsofinterleukin-13(IL-13), whichcontributestoclearanceofS.aureusinfectioninthelung(11,14),andgamma interferon(IFN-(cid:2)),whichbothweandothershavefoundtobediseaseenhancing(11, 14,21).Comparedtomock-inoculatedWTmice,bothP22capsid-andIAV-inoculated miceexhibitedincreasedIL-13productionwhenchallengedwithS.aureusatday3p.i. (Fig.1D).However,whenchallengedatday7p.i.withS.aureus,bothP22capsid-and IAV-inoculatedmiceshoweddiminishedIL-13production.ForIFN-(cid:2)production,com- November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 3 ® Shepardsonetal. D o w n lo a d e d f r o m h t t p : / / m b io . a s m . o r g / o n A p r il 4 , 2 0 1 FIG 1 Inoculation of mice with P22 capsid or with IAV induces similar immune responses to subsequent S. aureus 9 challenge and similar patterns of time-dependent alteration of bacterial clearance. (A) Structural representation of b y influenza A virus (IAV) (courtesy of CDC) and P22 capsid (PDB accession no. 2XYY). (B) Schematic of inoculation and g S.aureuschallengetimeline.(C)WTmicewereeitherinoculatedwithIAV(blacktriangles)onday0orinoculatedwithP22 u capsid(graysquares)orPBS(mock;whitecircles)ondays0,1,and2.Differentgroupsofmicewerechallengedwith e S.aureusonday3,7,or14andweresacrificed24hafterchallengetodetermineS.aureuslungbacterialload.(D)Fold s t change of cytokine concentration determined by ELISA on the BALFs of mice from day 3 and day 7 p.i. in panel C normalizedtocytokineconcentrationinmiceinoculatedwithPBS(mockcontrol;statisticalanalysescompareindividual treatmentstomocktreatment).(EandF)WTmicewereinoculatedasdescribedforpanelC,CD11c(cid:2)(E)orLy6G(cid:2)(F)cells werepurifiedfromthelungsatday3or7,andtheirabilitytokillS.aureusfollowinga1.5-h(E)or3-h(F)incubationex vivowasdetermined.TheexperimentinpanelChadaminimumof5animalspergroup,theexvivoexperimentinpanel Dhad5biologicaland3technicalreplicates,andtheexvivoexperimentinpanelsEandFhad3combinedbiologicalwith 5technicalreplicates.Eachwasrepeatedthreetimes.Dataarerepresentedasmean(cid:3)SD.*,P(cid:4)0.05;**,P(cid:4)0.01;***, P(cid:4)0.001. paredtomock-inoculatedWTmice,bothP22capsid-andIAV-inoculatedmiceshowed diminishedproductionofthiscytokinewhenchallengedwithS.aureusatday3p.i.,but when challenged at day 7 p.i., both P22 capsid- and IAV-inoculated mice showed increasedIFN-(cid:2)production(Fig.1D). IFN-(cid:2)can impair bacterial clearance by macrophages (21), whereas IL-13 can be involved in neutrophil recruitment (22). We previously reported that both CD11c(cid:2) (alveolarmacrophages)andLy6G(cid:2)cells(neutrophils)isolatedfromIAV-inoculatedmice November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 4 ® EarlyAntiviralImmuneResponseRequiresTLR2 onday3p.i.haveanincreasedabilitytokillS.aureusexvivocomparedtothosefrom mock-inoculatedmice,whereasneutrophilsisolatedfromthelungsofIAV-inoculated miceonday7p.i.showedanimpairedabilitytokillS.aureus(14).SimilartoIAVonday 3p.i.,bothalveolarmacrophagesandneutrophilsharvestedfromlungsofWTmiceat day3post-P22capsidinoculationkilledbetween6and12timesmoreS.aureusthan cells from mock-inoculated mice (Fig. 1E and F). Additionally, similar to neutrophils harvested from mice at day 7 post-IAV inoculation, neutrophils harvested at day 7 post-P22capsidinoculationexhibitedimpairedbactericidalactivitycomparedtothose frommock-inoculatedmice(Fig.1F).AlveolarmacrophagesisolatedfromP22capsid-, IAV-, or mock-inoculated mice at day 7 showed similar bactericidal activities toward S.aureusexvivo(Fig.1E). Host responses involved in dictating susceptibility to S. aureus challenge during earlyviralinfectionareinducedpriortobacterialchallenge.Specifically,productionof D IFN-(cid:3)over the course of early viral infection (day 2 to 3) contributes to increased o bacterial clearance of S. aureus during subsequent challenge. However, decreased w n production of IFN-(cid:3)and increased production of IFN-(cid:4)later in viral infection reduce lo bacterialclearance(11,14).Therefore,weinvestigatedwhetherthecytokineresponses a d inducedbyP22capsidinoculationpriortoS.aureuschallengewerealsosimilartothose e d inducedbyIAV.ThetwoconditionsinducedsimilarearlyproductionofIFN-(cid:3)(day3) f andlateproductionofIFN-(cid:4)(day7)inWTmicecomparedtomock-inoculatedWTmice ro m (Fig. S2). This suggests that the induction of a protective immune response was a h consequence of the initial viral or capsid recognition events and not a response to t t activechallengewithS.aureus.Activationofadaptiveimmuneresponsesatlatertimes p : / followingactiveviralinfectionincreasesboththemagnitudeandthecomplexityofthe /m immuneresponse.Thus,fortheremainderofthisstudy,improvedbacterialclearance b on day 3 p.i. was used as a metric for better understanding the mechanisms of early io . a recognitionofviruscapsid. s m TLR2-dependentsignalingisrequiredforIAV-andP22capsid-mediatedclear- . anceofS.aureusonday3p.i.Ourresultsthusfarsuggestthatactivevirusinfection o r g isnotrequiredforinductionoftheimmunologicresponsesthatregulatehostsuscep- / tibility to subsequent bacterial infection. Thus, we hypothesized that a common o n structuralpatternofviralcapsidswasrecognizedbyaPRRwithbroadspecificity,such A as the TLRs. To reduce the number of possible candidate PRRs involved, the MyD88 p r adapter molecule, which is used by nearly all members of the TLR family to induce il 4 inflammatorysignalinguponrecognitionofPAMPs,wasexamined(23).myd88(cid:5)/(cid:5)mice , 2 inoculatedwitheitherIAVorP22capsidshowedimpairedclearanceofS.aureusfrom 0 1 thelungatday3p.i.comparedtosimilarlyinoculatedWTmice(Fig.2A).Inadditionto 9 acting as a mediator for TLRs, MyD88 also complexes and transmits signals from the b y inflammation-mediating receptor IL-1R (24, 25). However, il-1r(cid:5)/(cid:5) mice, either IAV or g P22inoculated,demonstratedasimilarpatternofclearanceofS.aureusuponchallenge u e asthatseeninWTmice(Fig.S3),implicatingTLRsandnotIL-1Rasalikelycandidate s t intherecognitionofandsubsequentimmuneactivationbyIAVandP22capsid. To determine which TLRs were induced in response to IAV or P22 capsid, we quantified transcript levels for each TLR, in either whole-lung samples or purified alveolarmacrophages,6hfollowingeitherIAVorP22capsidinoculation,aknowntime frameforTLRtranscriptinduction(26,27).The6-htimepointwaschosentoidentify TLRs involved in early recognition rather than in downstream immune activation events, which could be complicated by recognition of other viral structures, such as nucleic acids. Both IAV and P22 capsid inoculation induced TLR2 and TLR6 mRNA expression. While TLR2 was upregulated in both whole-lung samples and purified alveolar macrophages, TLR6 was upregulated only by alveolar macrophages (Fig. 2B). Based on the consistent upregulation of TLR2, we assessed the requirement for this receptor in the immune response induced by the capsid. In contrast to WT mice, inoculationoftlr2(cid:5)/(cid:5)miceonday3p.i.witheitherIAV(Fig.2C)orP22capsid(Fig.2D) didnotresultinanincreaseofS.aureusclearancecomparedtomock-inoculatedmice. This indicates that TLR2 signaling was required for the initiation of protective re- November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 5 ® Shepardsonetal. D o w n lo a d e d f r o m FIG 2 TLR2-dependent signaling is required for IAV- and P22 capsid-mediated bacterial clearance of S. aureus during subsequent bacterial h infection.(A)WTandmyd88(cid:5)/(cid:5)micewereinoculatedwithIAV(day0;blackbars),P22capsid(days0,1,and2;graybars),orPBS(days0,1,and t t 2;mock;whitebars)andchallengedwithS.aureusonday3,andlungbacterialburdenwasmeasured24hlater.(B)Geneexpressionoftlr2and p : tlr6wasanalyzedbyqRT-PCRonRNAisolatedfromwholeWTlungorCD11c(cid:2)lungalveolarmacrophages6haftertreatmentwitheitherIAV, // P22capsid,orPBS(black,gray,orwhitebars,respectively).(CandD)WTandtlr2(cid:5)/(cid:5)micewereinoculatedwithIAV(C)(day0;blackbars)orP22 m capsid(D)(days0,1,and2;graybars)orPBS(day0ordays0,1,and2;mock;whitebars)andchallengedwithS.aureusonday3,andlung b bacterial burden was measured 24 h later. (E) IL-13 and IFN-(cid:3)cytokine concentrations in cell-free BALF from mice in panels C and D were io . determinedbyELISA(dashedredlineisPBS).(F)WTmicewereinoculatedwithPAM2(hatchedgraybar)orPBS(mock;whitebar)(day0)and a challengedwithS.aureusonday3,andlungbacterialburdenwasmeasured24hlater.(G)LevelsofIL-13andIFN-(cid:3)weremeasuredintheBALFs sm frommiceinpanelFbyELISA(dashedredlineisPBS,andstatisticalanalysiscomparestreatmenttoPBS).Mouseexperiments(A,C,D,andF) . hadaminimumof5animalspergroup,theexvivoexperimentinpanelBhad3biologicaland3technicalreplicates,andtheexvivoexperiment o inpanelsEandGhad5biologicaland3technicalreplicates.Eachwasrepeatedthreetimes.Dataarerepresentedasmean(cid:3)SD.*,P(cid:4)0.05; rg **,P(cid:4)0.01;***,P(cid:4)0.001;****,P(cid:4)0.0001. / o n A sponses.tlr2(cid:5)/(cid:5)micedidnotpresentthecharacteristiccytokinesequelaeoftypeIIFNs p r andIL-13atday3p.i.Specifically,tlr2(cid:5)/(cid:5)miceinoculatedwithIAVorP22capsidhad il 4 noincreaseinproductionofIL-13orIFN-(cid:3)comparedtomock-inoculatedtlr2(cid:5)/(cid:5)mice, , 2 whileWTmiceinoculatedwithIAVorP22capsidhada2-foldincreaseinproduction 0 of IL-13 and IFN-(cid:3)compared to mock-inoculated WT mice (Fig. 2E; Table S1). This 19 indicatedthatTLR2signalingcontributestotheupregulationofIFN-(cid:3)andIL-13after b y eitherIAVorP22capsidrecognitiononday4p.i.(Fig.2E).Furthermore,wefoundthat g killingofS.aureusexvivobyneutrophilsandmacrophagesisolatedfromtlr2(cid:5)/(cid:5)mice u e on day 3 after inoculation with either IAV or P22 capsid did not induce increased s t bactericidalactivitycomparedtothesecellsfromWTmice(Fig.S4). ToexaminewhetherTLR2activationalonewassufficienttoelicitprotectionagainst subsequent bacterial challenge, the known TLR2 agonist PAM2CSK5 (PAM2) was ad- ministeredinplaceofIAVorP22capsid.PAM2treatmentresultedinincreasedbacterial clearancecomparedtomock-inoculatedmiceanddisplayedasimilarpatternofboth IL-13 and IFN-(cid:3)production at day 3 p.i. (Fig. 2F and G) as found with IAV and P22 capsid. These results support a role for TLR2 signaling in immune responses to secondarypost-viralandpost-capsidS.aureuschallenge. Capsid-mediated immune events are independent of the capsid size and shape, but are dependent on TLR2 and recognition of intact capsid. TLR2’s involvement in the recognition of both IAV and P22 capsid suggested that TLR2 was respondingtosimilarstructuralpatternsonthesurfaceofthetwoparticles.Thisraised thepossibilityofTLR2actingasagenericreceptorforcommonstructuralelementsof viral capsids. Despite having no sequence similarity and having distinct host specific- ities, IAV and P22 capsid have similar sizes (~80 nm for IAV [28] and 60 nm for P22 November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 6 ® EarlyAntiviralImmuneResponseRequiresTLR2 D o w n lo a d e d f r o m h t t p FIG 3 Repeating protein subunit structure, but not spherical shape or size, is required for time-dependent : alterationofS.aureusclearance.(A)To-scalestructuralrepresentationsofTMVcapsid,P22capsid,andferritin // m capsid(assembledusingPDBaccessionnumbers4UDV,2XYY,and3AJO,respectively).(B)Micewereinoculated b withP22capsid(graysymbols),TMVcapsid(greensymbols),murine(openbluesymbols)orP.furiosus(solidblue io symbols)ferritincapsid,orPBS(mock;whitesymbols)(days0,1,and2)andchallengedwithS.aureusoneither . day3orday7.(C)tlr2(cid:5)/(cid:5)micewereinoculatedwithTMVcapsid(greensymbols),murine(openbluesymbols)or a s P.furiosus(solidbluesymbols)ferritincapsid,orPBS(mock;whitesymbols)(days0,1,and2)andchallengedwith m S.aureusonday3.(D)StructuralrepresentationofP22capsidandP22dimer.Sizesarenottoscale(PDBaccession . o no. 2XYY). (E) WT mice were inoculated with P22 capsid (gray symbols), P22 dimer (white symbols with gray r outline),orPBS(mock;whitesymbolswithblackoutline)(days0,1,and2)andwerechallengedwithS.aureuson g / eitherday3orday7.(B,C,andE)S.aureuslungbacterialburdenwasdetermined24hafterchallenge.Experiments o hadaminimumof5animalspergroupandwererepeatedtwo(C)orthree(BandE)times.Dataarerepresented n asmean(cid:3)SD.*,P(cid:4)0.05;***,P(cid:4)0.001. A p r il 4 capsid[29])andsphericalshapes.Toconsiderthepossibilitythatsizeand/orshapemay , 2 play a role in TLR2-mediated recognition of these particles, we tested a library of 0 1 capsids with various sizes and shapes for their ability to induce a similar pattern of 9 susceptibilitytoS.aureuschallengeasfoundwithP22capsidandIAV.Toexaminethe b y impact of shape, mice were inoculated with a large (~300-nm) rod-shaped capsid g derived from a tobacco mosaic virus (TMV) coat protein (CP) mutant (which forms a u e capsid in the absence of the nucleic acid cargo [30]). To examine the impact of size, s t micewereinoculatedwithsmall(~12nmindiameter[31])ferritincapsidsofeithera self(murine)orforeign(Pyrococcusfuriosus)origin(Fig.3A).Similartomiceinoculated withP22capsid,miceinoculatedwithTMVcapsidorwitheitherpyrococcal(nonself)or murine (self) ferritin capsids showed improved bacterial clearance when challenged with S. aureus at day 3 p.i., but not at day 7 p.i. (Fig. 3B). Additionally, tlr2(cid:5)/(cid:5) mice inoculated with TMV or with either ferritin capsid did not clear subsequent S. aureus challenge better than mock-inoculated mice on day 3 p.i. (Fig. 3C). These results demonstrated that the capsid recognition resulting in time-dependent alteration of hostsusceptibilitytoS.aureuspneumoniarequiredTLR2,andwasnotdeterminedby the capsid’s size or shape, and suggested that the capsid recognition preceded the recognitionofselfversusnonself. Viral capsids have long been known to induce more-efficient humoral responses thantheirisolatedsubunitsastheirrepeatedsubunitstructurebettercross-linksB-cell receptors (32, 33). While IAV is an enveloped virus, proteins (including primarily hemagglutinin, but also neuraminidase and matrix 2 protein) on the IAV surface are November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 7 ® Shepardsonetal. tightlydistributed,withanaveragespacingbetweenhemagglutininsubunitsof~9nm, leadingtoasimilarlyspacedpolyvalentproteinsurfaceasthecoatproteinoftheP22 capsid(averagespacingof~6nmbetweensubunits)(28,34).Toconsiderwhetherthe repeatingproteinsubunitarrangementwassimilarlyrequiredfortheTLR2-dependent recognition, we inoculated WT mice with the disassembled P22 coat protein subunit, whichcanbestablymaintainedinsolutionasadimer(29,35)F(ig.3D).Incontrastto miceinoculatedwithanintactP22capsid,inoculationofmicewithP22dimerneither showed improved bacterial clearance at day 3 p.i. nor decreased the clearance of S.aureusuponchallengeonday7p.i.(Fig.3E).Infact,miceinoculatedwithP22dimer showedasimilarpatternofsusceptibilitytoS.aureuschallengeasdidmock-inoculated mice (Fig. 3E). This finding further supports the hypothesis that antiviral immune responsesmediatetime-dependentchangesinhostsusceptibilitytosubsequentS.au- reus challenge via sequence-independent recognition of the structural pattern of the D viral capsid. These results strongly suggest/imply a pivotal role for repeating protein o subunit structures in the induction of antiviral immune responses by a mechanism w n involvingacommonrecognitionpathwayforthispattern. lo TLR2 recognizes repeating protein structures and does not require the sub- a d units to be spherically organized. The induction of common antiviral immune e d responses and the requirement for TLR2 by all of the intact capsids examined sug- f r gested that TLR2 might be involved in recognizing the repeating subunit pattern of o m viral capsids. In nature, repeating protein subunit patterns are not limited to capsids, h i.e., multisubunit protein assemblies surrounding a hollow cargo volume. Thus, we t t p rationalized that if TLR2 recognition is initiated solely by multisubunit protein assem- : / blies, then it might extend to noncapsid structures. Therefore, we examined filamen- /m tous (F) actin as a model noncapsid repeating protein subunit assembly. Unlike the b io capsidstestedthusfar,F-actinisanextendedhelicalmicrofiber,butitisstillcomposed . a ofarepeatingproteinsubunit(Fig.4A). s m InoculationofmicewithF-actinallowedfora10-foldincreaseinbacterialclearance . o atday3p.i.,butrenderedmicelessabletoclearS.aureusfollowingchallengeatday r g 7p.i.thanmock-inoculatedmice(Fig.4B).Thus,F-actininoculationresultedinasimilar / o time-dependentpatternofhostsusceptibilitytobacterialchallengeasdidinoculation n witheitherthecapsidsorviruses(Fig.1C,3B,andS1).Additionally,aswasfoundwith A P22 dimer (Fig. 4C), the globular F-actin monomers (G-actin) did not induce the p r time-dependentalterationofsusceptibilitytoS.aureuschallengecomparedtoF-actin- il 4 inoculated mice, further supporting our hypothesis that repeating protein assemblies , 2 composedofrepeatingproteinsubunitsareinducingthisresponse.F-actininoculation 0 1 also did not increase clearance of S. aureus in tlr2(cid:5)/(cid:5) mice at day 3 p.i. compared to 9 F-actin-inoculated WT mice, indicating a requirement for TLR2 in bacterial clearance b y after F-actin inoculation (Fig. 4D), similar to IAV and P22 capsid (Fig. 2C and D). g However, in contrast to IAV and P22 capsid (Fig. 2C), there was not a significant u e difference in the production of IFN-(cid:3)or IL-13 in the F-actin-inoculated tlr2(cid:5)/(cid:5) mice s t comparedtothatfoundinF-actin-inoculatedWTmice(Fig.4E),suggestingthatF-actin inducedadifferentTLR2-mediatedmechanismthanP22capsid.AsF-actinisknownto be recognized by the c-type lectin receptor Clec9a (DNGR1) (36, 37), we determined whetherClec9awasinvolvedintheantiviralimmuneresponseregulatinghostsuscep- tibilitytoasubsequentbacterialchallenge.Inoculationofclec9a(cid:5)/(cid:5)micewithF-actin resultedinincreasedclearanceofS.aureuscomparedtomock-treatedclec9a(cid:5)/(cid:5)mice (Fig. S5), indicating that Clec9a is not required for the F-actin-mediated immune responses that regulate host susceptibility to a subsequent S. aureus pneumonia. Together, these results support a general role for TLR2 recognition of proteins with repeating subunit structure regardless of their origin, but also suggest that TLR2 recognition can be promiscuous and, as such, can result in different downstream cytokineresponses. The repeating protein subunit assemblies of P22 capsid and F-actin mediate inductionofprotectiveimmuneresponsesthroughTLR6andTLR1,respectively. TheinductionofdifferentcytokineresponseswithasharedTLR2dependencebyP22 November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 8 ® EarlyAntiviralImmuneResponseRequiresTLR2 D o w n lo a d e d f r o m h t t p : / / m b io . a s m . o r g / o n FIG4 F-actininoculationinducessimilartime-dependentalterationofbacterialclearanceafterS.aureus A challenge,butdifferentcytokinescomparedtoIAVorP22capsidinoculation.(A)Astructuralrepresen- p tmaticioenwoefrFe-iancoticnualantdedGw-aicthtinF-gaecntienra(oteradnfgroemsyPmDbBoalsc)coerssPiBoSn(nmuomcbke;wrsh3iJte8Ksyamndbo1lJs6)Z(d,raeyssp0e,c1ti,vaenlyd.(2B))aWndT ril 4 werechallengedwithS.aureusoneitherday3orday7,andlungbacterialburdenwasevaluated24h , afterchallenge.(C)WTmicewereinoculatedwith50(cid:5)gofeitherF-actin(orangesymbols),G-actin(dark 2 0 redsymbols),orPBS(mock;whitesymbols)(days0,1,and2)andwerechallengedwithS.aureusonday 1 3,andlungbacterialburdenwasevaluated24hafterchallenge.(D)WTortlr2(cid:5)/(cid:5)micewereinoculated 9 with F-actin (orange symbols) or PBS (mock; white symbols) (days 0, 1, and 2) and challenged with b S. aureus on day 3, and lung bacterial burden was evaluated 24 h after challenge. (E) Cytokine y concentrationdeterminedbyELISAontheBALFsofmicefrompanelC.ExperimentsshowninpanelsB, g u C,andDhadaminimumof5animalspergroup,andtheexvivoexperimentinpanelEhad5biological e and3technicalreplicates.Eachwasrepeatedthreetimes.Dataarerepresentedasmean(cid:3)SD.*,P(cid:4) s 0.05;**,P(cid:4)0.01;***,P(cid:4)0.001;N.S.,notsignificant. t capsid and F-actin suggested that there might be differences in cellular recruitment and/oractivation.Cellularanalysisofthebronchoalveolarlavagefluid(BALF)revealed thatP22capsidinoculationinducedrecruitmentofneutrophils(CD11b(cid:2)Ly6G(cid:2);similar to previous results from IAV [14]) while F-actin did not (Fig. 5A and S6A). In fact, the cellular composition of BALF from mice inoculated with F-actin resembled that of mock-inoculated mice and consisted of mostly resident alveolar macrophages (~93% CD11c(cid:2)SiglecF(cid:2)[Fig.5AandS6A]).Todeterminewhetherthesedifferencesincellular recruitmentintothelungtranslatedintodifferencesinthefunctionofthesecells,we tested their bactericidal capacity against S. aureus in the ex vivo assay. P22 capsid increased the bactericidal capacity of both neutrophils (Ly6G(cid:2)) and macrophages (CD11c(cid:2))isolatedfromWTmiceonday3p.i.,whileF-actinincreasedthebactericidal capacity of only macrophages (Fig. 5B). These results indicated that while both P22 November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 9 ® Shepardsonetal. D o w n lo a d e d f r o m h t t p : / / m b io . a s m . o r g / o n A FIG5 P22capsidandF-actinrequiredifferentTLR2heterodimersfortheirabilitytoinduceS.aureuskillingbymacrophages.(A)WTmicewereinoculatedwith p r P10202(cid:5)camp)s.id(B,)FW-aTctmin,icoerwPBerSe(dinaoycsu0la,t1e,danwdit2h),eaitnhderdPif2fe2rceanptisaildc(oguranytssoymfwbhoiltse),bFl-oaoctdince(ollrsafnrgoemsBymALbFowls)e,roerdPoBnSe(monocdka;yw3hpit.ei.s(yrempbreosles)n(tdaatiyvse0i,m1a,gaensds2h)o;wCDn1;b1ca(cid:2)r, il 4 orLy6G(cid:2)cellswerepurifiedfromthelungsatday3;andtheirabilitytokillS.aureusfollowinga1.5-h(CD11c(cid:2))or3-h(Ly6G(cid:2))incubationexvivowas , 2 determined.(C)PercentinitialbodyweightsofmiceinoculatedwithP22capsid(graysymbols),F-actin(orangesymbols),orPBS(mock;whitesymbols)(days 0 0,1,and2)andchallengedwithS.aureusonday3.(D)PercentinitialbodyweightsofmiceinoculatedwithPAM2,PAM3,orPBS(mock)(day0)andchallenged 1 withS.aureusonday3.(EandF)DifferentialcountsofwhitebloodcellsfromBALFofmicetreatedwithPAM2,P22capsid,orPBS(mock)(E)orPAM3,F-actin, 9 orPBS(mock)(F)andchallengedwithS.aureusonday3.(GandH)ImmortalizedBMmacrophagesfromWT,tlr2(cid:5)/(cid:5),tlr1(cid:5)/(cid:5),andtlr6(cid:5)/(cid:5)micewerestimulated b y withP22capsid(graysymbols)(G)orF-actin(orangesymbols)(H)orPBS(mock;whitesymbols)for12h.Cellswerewashedandincubatedata10:1ratioof g S.aureustocellsfor5h.CFUweredeterminedafterovernightincubationofsamplesonTSAplatesfollowingcelllysis.ExperimentsshowninpanelsAand u CtoFhadaminimumof5animalspergroup,theexvivoexperimentinpanelBhad3combinedbiologicalwith5technicalreplicates,andtheexperiment e showninpanelsGandHhad5technicalreplicatespercellline.Eachwasrepeatedthreetimes.StatisticsforpanelsCandDareforthechangeinweightloss s t fromday3today4(mock,F-actin,andPAM3)orfromday0today3(P22capsidandPAM2).Dataarerepresentedasmean(cid:3)SD.**,P(cid:4)0.01;***,P(cid:4)0.001. capsidandF-actinutilizeTLR2tomodulatehostsusceptibilitytosubsequentS.aureus challenge, the P22 capsid-induced responses are more inflammatory than the re- sponsesinducedbyF-actin.Differencesincellularrecruitmentandinflammationoften leadtodifferencesinbiologicaleffects;thus,wetestedwhetherP22capsidandF-actin haddistinctbiologicaleffects.P22capsidinoculationresultedinsignificantweightloss fromday0today3inWTmice(Fig.5C),butthemicewerethenprotectedfromweight lossuponS.aureuschallenge(day3to4,P(cid:6)0.315)(Fig.5C).F-actininoculationhad noimpactonweightlossinitially(day0to3,P(cid:6)0.501)(Fig.5C),butfollowingS.aureus challenge, the mice lost a significant amount of weight, similar to mock-inoculated mice(day3to4,Fig.5C). The observed differences between mice inoculated with P22 capsid and those inoculatedwithF-actinintermsofcellularrecruitmentandactivation,aswellasbody November/December2017 Volume8 Issue6 e01356-17 mbio.asm.org 10

Description:
Utilizing this altered susceptibility to subsequent bacterial challenge as an outcome, we determined that duced by the repeating subunit pattern of virus exteriors (capsids), and thus, induc- tion of this KEYWORDS pattern recognition, bacterial superinfection, innate immunity, Toll-like receptors
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.