DSpace at VNU: Killed Bacillus subtilis spores as a mucosal adjuvant for an H5N1 vaccine

DSpace at VNU: Killed Bacillus subtilis spores as a mucosal adjuvant for an H5N1 vaccine tài liệu, giáo án, bài giảng ,...

Vaccine

jo u r n al h om ep a ge : w w w e l s e v i e r c o m / l o c a t e / v a c c i n e

Killed Bacillus subtilis spores as a mucosal adjuvant for an H5N1 vaccine

Thi Van Anh Nguyend, Sung-Moo Parka, Byoung-Shik Shima, Ho Hyun Songa, In Su Cheona,

Ji Eun Janga, Jung-ah Choia, Young Ki Choie, Konrad Stadlera, Simon M Cuttingb,∗

a Laboratory Science Division, International Vaccine Institute, Seoul 151-818, Republic of Korea

b School of Biological Sciences, Royal Holloway, University of London, Egham, Surrey TW20 0EX, UK

c Key Laboratory of Gene Technology, Institute of Biotechnology, VAST, 18 Hoang Quoc Viet, Hanoi, Viet Nam

d Key laboratory of Enzyme and Protein Technology, VNU University of Science, 334 Nguyen Trai, Hanoi, Viet Nam

e Department of Microbiology, College of Medicine and Medical Research Institute, Chungbuk National University, Cheongju 361-763, Republic of Korea

a r t i c l e i n f o

Article history:

Received 28 October 2011

Received in revised form 21 February 2012

Accepted 9 March 2012

Available online 22 March 2012

Keywords:

Bacillus subtilis

Spore adjuvant

Influenza

H5N1 vaccine

a b s t r a c t

HeatkilledsporesoftheGram-positivebacteriumBacillussubtilishavebeenevaluatedasavaccine deliv-erysystemwithmucosaladjuvantpropertiesforinfluenza.KilledsporeswereabletobindH5N1virions (NIBRG-14;clade1)and,whenintra-nasallyadministeredtomice,resultingimmuneresponses,both humoralandcellmediated,wereenhancedcomparedtoimmunizationwiththevirionalone.Levelsof bothsystemicIgGandmucosalsIgAspecifictothevirionwereelevated.LevelsofIgG2a(aTh1antibody type)werestronglyenhancedwhenthevirionwasco-administeredwithkilledspores.Cytokine induc-tioninstimulatedsplenocyteswasalsoapparentindicatingbalancedTh1andTh2responses.Evidence

ofcross-neutralizationofclade2.2viruseswasshown.Inachallengeexperimentmicedosedtwotimes withsporesadsorbedwithjust20ngHA(hemagglutinin)ofinactivatedNIBRG-14werefullyprotected againstchallengewith20LD50ofH5N2virus.Interestingly,partialprotection(60%)wasobservedin ani-malsdosedonlywithkilledspores.Micedosedonlywithkilledsporeswereshowntobefullyprotected againstH5N2(5LD50)infectionindicatingthatinnateimmunityanditsstimulationbysporesmayplay

animportantroleinprotection.Supportingthiskilledsporeswere(i)showntostimulateTLR-mediated expressionofNF-␬B,and(ii)abletorecruitNKcellsintolungsandinducematurationofDCs.Thiswork demonstratesthepotentialandunderlyingmechanismfortheuseofbacterialsporesasanadjuvantfor H5N1vaccination

© 2012 Elsevier Ltd All rights reserved

1 Introduction

HighlypathogenicH5N1influenzainfectionsarestillrelatively

rareinhumansyetthecontinuedevolutionofthisvirusmakesit

apotentiallyseriousandloomingpublichealththreat[1].A

vac-cinesuitableforprotectionagainstafuturepandemiccausedby

anH5N1 virusshould providecross-clade protective immunity

sincetheevolutionofH5N1virusesinbirdsinsouth-eastAsiais

unprecedentedwhilethestrainthatmaycauseapandemic

can-notbepredictedandcurrentparenteralseasonalinfluenzavaccines

arealllargelystrain-specific[1].Inaddition,suchavaccineshould

havethreefurtherattributes:besafe,requirealowantigen

con-tent(antigen-sparing)tocombatthelowimmunogenicityofH5N1

virusesandcaterforproductionproblemsandfinally,inducealong

lastingimmunity[2–4].Regardingantigen-sparing,adjuvantssuch

∗ Corresponding author Tel.: +44 1784 443760; fax: +44 1784 414224.

E-mail address: s.cutting@rhul.ac.uk (S.M Cutting).

1 These authors contributed equally to this paper.

asAS03orMF59,oil-in-wateremulsions(basedonsqualene-like molecules)havebeenusedparenterallyinhumanssuccessfully[5]

Anumber ofstudieshave shownthatsecretoryIgA(sIgA)from therespiratorytractismoreeffectiveincross-protectionagainst heterologous influenza compared to IgG induced by parenteral vaccination[6,7].Accordingly,mucosaladjuvantssuchas modi-fiedcholeratoxinand therelated E.coli heatlabileenterotoxin (LT)havebeenshowneffectiveataugmentingresponsesto inacti-vatedinfluenzavirus[8,9]butunfortunately,havebeenlinkedto adversereactionsincludingfacialparalysis[10].Thenasalrouteof vaccinationhasbeenshowntoinducemoreappropriateimmune responsesatthemucosalsurface,theactualsiteofinfection,and hasbeenmoreeffectivetoepidemicscaused byaheterologous virus[11,12]

Sporesof theGram-positivebacterium, Bacillussubtilis,have beenusedsuccessfullyforantigendelivery.Usinggenetically engi-neeredsporesthatexpressheterologousantigensprotectionhas beenachievedagainstanumberofpathogensusinganimal mod-els,notablyClostridiumperfringens[13],Clostridiumdifficile[14], Clostridiumtetani[15],andRotavirus[16].In alltheseexamples

0264-410X/$ – see front matter © 2012 Elsevier Ltd All rights reserved.

bindingheterologousantigens[20].Usingacombinationof

electro-staticandhydrophobicinteractionsproteinshavebeenshownto

bindtothesporesurface.Thisapproachenablesthesporetoserve

asanantigencarrierwithouttheneedfor geneticmodification

Remarkably,usingsuchanapproachantigenpresentationappears

equallyasefficientwhetherthesporesareliveorinactivated(heat

killed).Thisapproachthenresemblesothermicroparticulate

adju-vants,which,byvirtueoftheirsizeandmultimericpresentation,

mimicthepathogenstheimmunesystemhasevolvedtocontrol,

andfacilitatesuptake byantigen presentingcells [21,22]

Stud-iesonthenatureofantibodyresponsesinmicehaveshownthat

sporesco-administeredwithproteinantigensatmucosalsurfaces

augmentantigen-specificsecretoryIgA(sIgA)aswellasinducing

balancedTh1/Th2responses[20,23]

Many animal viruses carry an envelope comprised of

phos-pholipidembeddedwithlipoproteinsand/orglycoproteins.Such

viruses have been shown to efficiently adsorb to hydrophobic

surfaceswherebindingincreasesasthepHdeclines[24].This

phe-nomenonshouldenable aviriontoadsorbtothesporesurface

andwastherationaleforthepresentstudy.Inthiswork,wehave

shownthatsporeshaveuniquecharacteristicsasbothanadjuvant

andforantigendelivery.Mucosaldeliveryofalowdoseof

virion-adsorbedsporesisshowntoelicitprotectiveimmunitytoinfluenza

inamurinemodelofinfection.Wealsoshowthatsporesby

them-selvesactasapotentialimmunestimulantabletoconferprotection

toinfluenzaandweprovideabasicmechanismforhowthismight

occur

2 Materials and methods

2.1 Generalmethods

MethodsforworkwithB.subtilisaredescribedelsewhere[25]

B.subtilisstrainPY79,isastandard,prototrophic,laboratorystrain

[26].Sporulationwasconfirmedbyphase-contrastmicroscopyand

sporecropswereharvestedandpurifiedasdescribedelsewhere

[27].Sporeswerekilledbyautoclaving(121◦C,15p.s.i.,30min)

2.2 Reagents

Inactivated H5N1 virus, A/Vietnam/1194/2004 (NIBRG-14;

clade 1) and A/Turkey/1/2005 (clade 2.2) as well as reference

antiserawereobtainedfromtheNationalInstituteforBiological

StandardsandControl(NIBSC),UK

2.3 Propagation,purificationandtitrationofNIBRG-14

Alivevirus(NIBRG-14)suspension(50%Egginfectivedoses

cal-culatedforNIBRG-14was6.1log10EID50per0.2ml)wasdiluted

10−3insterilePBS(0.01M,pH7.2)supplementedwith100␮gof

dardand specificsheepantiserum(NIBSC,UK) Asuspensionof NIBRG-14at60␮g/ml(HA)waspreparedasaworkingstockanda representativeSDS-PAGEgelofproteinsextractedfromNIBRG-14

isshowninSupplementaryFig.1 2.4 Imageanalysisofinfluenzavirionadsorption Purified inactivated H5N1 virions (NIBRG-14; 0.33mg/ml, equivalent to 11␮M of HA) were directly labeled with 20␮M sucinimidyl tetramethyl rhodamine (TMR, Invitrogen) at room temperature(RT)for 45minin phosphatebufferedsaline(PBS;

145mM pH 7.4) The reaction was stopped by adding 10mM Tris–HCl, and the unreacted dyes were carefully removed by sequentialdialysis(4-times)inPBS.LabeledNIBRG-14-TMRwas thencentrifuged(10,000rpm;5min)toremoveprecipitationthat mayhaveoccurredduringlabelinganddialysis.The concentra-tionofproteinand TMRin thelabeledsamplewasdetermined (usingaNanoDrop)tobe0.27mg/ml,equivalentto9␮Mof pro-teinand1.7␮MofTMR,indicatingaproteintodyeratioofabout 5:1.InactivatedPY79spores(5×109)weredirectlylabeledwith

50␮Msuccinimidylfluoresceinisothiocyanate(FITC,Invitrogen) usingthesamelabelingconditionsasthoseusedforNIBRG-14.The labeledPY79-FITCsporeswereremovedfromunincorporateddyes

bysequentialwashingwitha100-foldvolumeof10×PBS(1-times) followedby1XPBS(2-times)bycentrifugation(12,000rpm,3min) ThelabeledPY79-FITCsporeswereobservedbylightand epifluo-rescencemicroscopytoconfirmthatall(100%)sporeswerelabeled PY79-FITCspores(1×108)wereincubated(withgentleagitation) with6mlofNIBRG-14-TMRfor30minatRT.Virion-labeledspores wereharvestedbycentrifugation(12,000rpm,3min)andwashed 3-timeswithPBS(100-fold).Labeledsporeswereobservedunder laserconfocalfluorescencemicroscopeusingaCarlZeissLSM510 2.5 Adsorptionofvirionstosporesforimmunization

Autoclavedsporesweresuspendedin10XPBS(0.1M,pH7.2)

atRTfor30min,ataconcentrationof5×109sporesml−1buffer Sporeswerethenwashed3timeswith1×PBS(0.1M,pH7.2)by centrifugationandpelletsre-suspendedinthesamebuffer Inacti-vatedNIBRG-14virionswereaddedandthemixtureshakengently

atRTfor30min.Afteradsorption,themixturewasincubatedon iceuntiluse

2.6 Animals AnimalsusedinthisworkwereBALB/corC57BL/6mice(Charles RiverorTheJacksonLaboratory)forantibodyproduction,NKcell recruitmentand for analysis ofimmune responses In allcases females,aged6–8weekswereused.AnimalproceduresintheUK wereperformedundertheHomeOfficeprojectlicensePPL70/6126

InS.Koreaanimalexperimentswereperformedinaccordancewith

InstituteInstitutionalAnimalCareandUseCommittee

2.7 Analysisofimmuneresponses

Groupsofmice(n=6;BALB/c)wereimmunizedintra-nasally

(i.n.)usingthree-dosesondays1,15and36.Immunogenswere

(i)inactivatedNIBRG-14(1.2␮gHA/dose)and(ii)PY79K spores

(1×109/dose)adsorbedwithinactivatedNIBRG-14(1.2␮gHA).In

allexperimentscontrolgroupsincludednạve(PBS)andagroup

receivingonlyPY79K spores(1×109).Bloodsampleswere

col-lected beforeeach dose, and splenocyteand final bleeds were

collectedthreeweeksafterthelastimmunization.Serumwas

inac-tivatedat56◦C(30min)andstoredat−80◦Ctilluse.Salivawas

takensevendaysafterthelastdoseusingsalivacollectionswabs

(ProbactTransportSwabs,Tech.ServiceLtd.,Heywood,UK).Swabs

wereappliedfor5mintillcompletelywetandthenincubatedin

proteaseinhibitor(1mMPMSFinPBS,0.2ml)for3honice.The

extractionswerestoredat−80◦Ctilluse.

2.8 Challengestudies

The animal and viral experiments using H5N2 virus were

conductedin animal and clinical biosafetylevel 3 containment

facilities,andallpersonnelworepersonalprotectiveequipment,

includingTyveksuitsandN95orHEPA-filteredairrespirators(3M)

Groups of mice (BALB/c)were anesthetized by intra-peritoneal

(i.p.) injection of ketamine (100mgkg−1body weight)/xylazin

hydrochloride(10mgkg−1bodyweight).Forchallengewith

inac-tivated influenza virus (Section 3.4), a two-dose regimen was

used;animalswerei.n.primedonday1withwholeinactivated

H5N1(NIBRG-14)virus(containingeither0.02␮gor0.5␮gofHA

in 20␮l of PBS)in the presence or absence of 2×109 CFU of

autoclavedspores(PY79K)andwereboostedwiththesame

anti-genat day 14 Controlgroups wereanimals dosed witheither

spores or NIBRG-14 only For virus challenge, an H5N2 virus,

A/AquaticBird/Korea(H5N2),wasused.A/AquaticBird/Koreawas

usedbecausethehighlypathogenicH5N1viruswasnotavailable

atthetimeofchallenge.TheHAproteinofA/AquaticBird/Korea

is92%and93% conservedwiththeHAproteinsfromNIBRG-14

andA/Turkey/1/2005,respectively.Anesthetizedmicewere

chal-lengedi.n.fourweeksafterthesecondimmunizationwitha20LD50

(50%mouselethaldoses)ofA/AquaticBird/Korea(H5N2)as

out-linedintheresultssection.Miceweremonitoreddailyforclinical

signsofinfluenzainfectionandbodyweightrecordeddaily.Mice

thatlostgreaterthan25%ofbodyweightwereeuthanized.Thetwo

doseregimenwaschosenbasedonextensivepreliminarychallenge

trialswherewedetermined(i)thelowesti.n.doseofNIBRG-14

(0.02␮g)that didnotconfer protectionin miceusingdifferent

dosingregimens,and(ii)whetheradsorbedsporescouldenhance

NIBRG-14-specificIgGresponses.Asimilarstrategywasusedfor

evaluationofprotectionconferred byescalatingdosesofPY79K

spores(Section3.5)butafinalchallengedoseof5LD50wasused27

daysafterthelastdoseofspores

2.9 DeterminationofantibodyresponsesbyindirectELISA

The ELISA methods for measuring IgG, IgG1, IgG2a or IgA

weremadeusingthesameformat.Plateswerecoatedwith50␮l

of inactivated NIBRG-14 (20 haemagglutination units (HAU) in

carbonate–bicarbonatebuffer,pH9)perwellandincubatedat4◦C

overnight.Afterblockingwith2%BSAinPBSatRTfor1h,serum

sampleswereappliedasa2-folddilutionseries,startingwitha

1/100dilution in assaydiluent buffer(10mM PBS[pH 7.4],1%

[wt/vol]BSA,0.05%[vol/vol]Tween20).Foreachplate,apooled

serumfromthenạvegroup(1/100) wasincludedin replicated

wells.Plateswereincubatedfor2hatRTbeforetheadditionof anti-mousehorseradishperoxidaseconjugatesforserumsamples (allobtainedfromSerotec,UK).Plateswereincubatedforone addi-tional hour at RT and then developed with thesubstrate TMB (3,3,5,5-tetramethyl-benzidine;Sigma).Reactionswerestopped with2NH2SO4.Dilutioncurvesweredrawnforeachsampleand end-point titerswere calculated asthedilutions producing the sameopticaldensityasthe1/100dilutionofapre-immuneserum 2.10 Cytokinedeterminations

Splenocytes(5×105)wereseededin96-wellcellcultureplates

incompletemedium(DMEMmediumsupplementedwith10%fetal BovineSerum, 50␮g/mlpenicillin,and 50␮g/mlstreptomycin) Cellswerestimulatedwith200HAUml−1 ofinactivated

NIBRG-14, 1␮g/mlConAasa positive controlorwithmediumonlyat

37◦Cin5%CO2.Supernatantswereharvestedfromculturesafter

48hofincubation.Cytokines(IFN-␥,IL-2andIl-6)weredetected usingReady-Set-GoELISAkits(eBioscience,UK)andaccordingto themanufacturers proceduresand standardsprovided In brief, microtitreplates(96well;Maxisorb;Nunc,UK)werecoatedwith purified capture anti-mouse antibody in 10mM PBS (pH 7.4) overnightat4◦C.Freebindingsiteswereblockedwithassay dilu-entatRTfor1h.CulturesupernatantswereincubatedatRTfor

2h.Beforedetection,biotinylatedanti-mouseantibodywasadded andincubatedatRTfor1h.Streptavidin-horseradishperoxidase conjugatewasaddedandincubatedatRTfor30min.Recombinant mousecytokineswereusedasastandardandcytokinesquantified

byinterpolationfromastandardcurve

2.11 Haemagglutinationinhibition(HAI)assay Onevolumeofimmunizedmouseserumwasincubatedwith fivevolumesofcholerafiltrate(Sigma)at37◦Cforapproximately

16h,followedby1hofincubationat56◦C.To50␮lofthe2-fold dilutionseriesofserum(inPBS),25␮lofPBScontaining4HAU

of inactivatedvirus(A/Vietnam/1194/2004 or A/Turkey/1/2005) wasaddedand incubatedfor30minat37◦C Next,25␮lof1% chickenerythrocytesuspensioninPBSwasaddedfollowedby1h incubationat4◦C.Subsequenthaemagglutinationpatternswere examinedandexpressedasthevalueofthehighestserumdilution whichcancompletelyinhibithaemagglutination

2.12 TLRexpression TheTLRreportermacrophagecellline(RAWBlue,Invivogen, France)wasusedforassessmentofTLRsignaling.RawBluecells wereculturedin10%FBSDMEM(Sigma)supplementedwith4.5g/l glucose,2mMl-glutamine,100␮g/mlNormocinTMand200␮g/ml ZeocinTMat37◦Cand5%CO2.FortheTLRstimulationassay,cells wereseededto96wellplateat5.5×105/mlwith180␮lcell sus-pension perwell(∼100,000cells/well) Wells already contained

20␮lofligandsuspension(heatkilledB.subtilisPY79spores) Con-trolsusedwere1.1␮m diameterlatexbeads(Sigma),B.subtilis PY79vegetativecells,LPS(E.coli;SigmaK325) andmediaonly Cellswereincubated for23hat37◦C afterwhich supernatants wereaddedtotheQuantiBlueTMSEAPdetectionassay(Invivogen) TLRstimulationwasmadebymeasurementofOD655readingsafter 1.5hofincubationandtheassaywasrepeatedtwice

2.13 Maturationofbonemarrowderiveddendriticcells(BM-DC)

bysporeinvitro BM-DCs were generated from murine bone marrow cells Briefly, bone marrow was flushed from the tibiae and femurs

of BALB/c mice and was depleted of red blood cells (RBC)

Fig 1. Adsorption of H5N1 onto spores Laser scanning confocal micrographs showing individual fluorescent-labeled PY79-FITC spores (panel A, green signal) bound with fluorescent-labeled virus particles NIBRG-14-TMR (panel B, red signal) on dual excitations of 488 nm and 543 nm, respectively Panel C is a merged image of panels A and B Panels D and E are images of a control sample in which PY79-FITC spores were incubated with only TMR dye at a 4-fold higher concentration than that of NIBRG-14-TMR Panel D the green signal of individual fluorescent-labeled PY79-FITC spores, while Panel E is a dark image without any observable red signal of NIBRG-14-TMR Panel F shows magnified representative images (taken from panel C and other observations) of clusters of NIBRG-14-TMR bound on PY79-FITC spores (For interpretation of the references

to color in this figure legend, the reader is referred to the web version of the article.)

with ACK buffer Cells were plated in six-well culture plates

(1×106cells/well) inRPMI1640medium (Invitrogen,USA)

sup-plementedwith2mMglutamine,1mMsodiumpyruvate,50␮M

␤-mercaptoethanol,100IU/mlpenicillin,100␮g/mlstreptomycin,

10%FBS,and20ng/mlrecombinantmouseGM-CSFand10ng/ml

recombinantmouseIL-4andthenincubatedat37◦Cand5%CO2for

5days.BM-DCswerestimulatedwith2␮g/mlofCholeratoxin(CT)

and1,10,and100MOIofsporesfor24h.Cellswerestainedwith

CD11cplusCD80,CD86orMHCclassIIfor20minat4◦C.Stained

cellswereanalyzedbyflowcytometry,FACSCalibur.All

cytomet-ricdatawereexpressedasmeanfluorescenceintensity(MFI)and

analyzedusingFlowJosoftware.Theassaywasrepeatedtwice

2.14 IL-12p40enzyme-linkedimmunosorbentassay

Toidentifythematurationofdendriticcell,IL-12p40was

mea-suredusingReady-set-goELISAkitandperformedaccordingtothe

manufacturers’instruction(eBioscience,Inc.)

2.15 Preparationoflungsingle-cellsuspensions

Toobtainsingle-cellsuspensions,lungsfrombothBALB/cand

C57BL/6 were minced and incubated in RPMI-1640 containing

10%fetalbovineserum(FBSand with2mg/mlofcollagenaseD

(Roche) and 0.1mg/ml of DNaseI (Sigma) at 37◦C for 45min

Asingle-cellsuspensionwaspreparedafterRBClysis usingACK

buffer,filteredthrougha100-nmcellstrainer(BD),andpelletedby

centrifugation.SamplesweretreatedwithFcblock(BD

Pharmin-gen)beforestaining.Cellswerestainedwithanti-CD3-FITCplus

anti-NK1.1-PerCPCy5.5,andanti-CD3-Allophycocyanin(APC)plus

anti-DX5-PE.Stainedcellswereanalyzedbyflowcytometry,FACS

Calibur.AllcytometricdatawasanalyzedusingFlowJosoftware Theassaywasrepeatedtwice

2.16 Statistics AlldatawereanalyzedwithGraphPadPrismversion5 (Graph-Padsoftware,SanDiego,CA).A Pvalueof>0.05 wasconsidered non-significant

3 Results

3.1 Adsorptionofvirustospores PurifiedH5N1(NIBRG-14)virionswerelabeledwithTMRand mixedwithasuspensionofpuresporesofB.subtilisstrainPY79 thathadbeenlabeledwithFITC.Sporeswereharvested,washed repeatedlyandexaminedbyconfocalimaging(Fig.1).FITC-labeled sporeswerefoundtobecoatedwithTMR-labeledvirions(Fig.1B andC).Generally,NIBRG-14wasuniformlylabeledonthespore surfacealthoughsomeclusteringwasapparent(visibleasintense redpatches,seeFig.1B,CandF)whichispresumablyduetovirion aggregates.Toconfirmthatthesporewasbindingtothevirionand notTMRweincubatedPY79-FITC-labelledsporeswithTMRonlyat

a4-foldhigherconcentrationthanthatusedwithNIBRG-14-TMR Followingthis,onlygreenlabelingofsporeswasobserved(Fig.1D) andnoredfluorescenceduetoTMRbindingtothesporesurface wasdetectable(Fig.1E).ToverifythatsporesboundtoNIBRG-14 virionsacompetitivebindingexperimentwasperformedby incu-bationofPY79-FITCwithamixtureofnon-labeledH5N1/labeled H5N1-TMRataratioof10:1.Undertheseconditionswecouldnot observearedsignalduetoH5N1-TMRbindingtoPY79-FITCspores

Fig 2. NIBRG-14-specific humoral responses Panel A, mice were immunized i.n with three doses (days 1, 21 and 42) of PY79 K spores pre-adsorbed with inactivated NIBRG-14 1.2 ␮g HA () and serum IgG1 and IgG2a titers determined Controls included nạve mice (×), mice immunized with PY79 K spores alone () or with inactivated NIBRG-14 (1.2 ␮g HA) alone () Bleeds were taken on days 0, 20, 41 and 63 Data are presented as arithmetic means and error bars are standard deviation Panel B, salivary IgA titers from the same study as panel A Samples were taken on days 0, 20, 41 and 63 Data shows individual mouse (䊉) and arithmetic means (–).

(datanotshown).TheseresultsdemonstratethatanH5N1virion

canbindtothesurfaceofB.subtilisspores.Itisnotpossibleatthis

stagetoaccuratelydeterminethenumberofH5N1virionsthatbind

toasinglespore.However,weestimatethenumbertobeatleast

eightvirionspersporebasedonthefollowingcalculation:(i)the

totalsurfaceareaofanellipsoidalsporehavingalengthof1.38␮m

anddiameterof0.63␮mwascalculatedas∼2.60␮m2;(ii)asingle

fluorophoreemitsaspotwithadiameterequaltothewavelength

ofemissionlight,inthecaseofTMRthisis572nm.Asthesizeof

asinglevirionisabout50nm,whichismuchsmallerthanthesize

ofasinglefluorescentspot,asinglevirionconjugatedwitheither

singleormultipleTMRfluorophoreswouldthereforealsoemita

spotwithadiameterofabout572nm.Thus,theemissionareaof

asinglevirionwascalculatedtobe0.32␮m2.Assumingauniform

intensityofTMRonindividualspores,therewouldbeatleast8

virions(2.60␮m2/0.32␮m2=8.1)boundperspore

3.2 Heat-killedsporesenhancehumoralandcellularimmune

responsestoNIBRG-14

Virusneutralisingantibodiesincludingbothlocalandsystemic

playanimportantroleinimmunitytoinfluenzainfection[29–31]

Specifically,HA-specificantibodiesneutralizethevirusand

pre-ventinfection[32].Wethereforeevaluatedtheantibodyresponses

specifictoH5N1inmicedosedthree-timesintra-nasally(i.n.)with sporesadsorbedwithinactivatedNIBRG-14.Groupsofmice(n=6) wereimmunizedi.n.withPY79sporesthathadbeenkilledby auto-claving(PY79K)andadsorbedwithNIBRG-14(1.2␮gHA).Control groupsincludedanimalsdosedwithonlykilledsporesor

NIBRG-14.NIBRG-14-specific IgGresponsesweremeasuredbyindirect ELISAinserumsamples(Fig.2).AnalysisoftheIgG1andIgG2a subclasses(Fig.2A)revealedthatheat-killedsporesinduced virion-specificIgG1andIgG2ainthegroupwheresporeswereadsorbed withNIBRG-14and thatIgG2aresponseswereboth earlierand higher(10-fold)thanthose toinactivatedvirusdeliveredalone Thevalue ofthe IgG1:Ig2aratios(SupplementaryFig.2)of the adsorbedspore-NIBRG-14groupwerelower(∼1.8)thanthevalue

ofNIBRG-14 alone(∼5),indicatinga Th1biasinducedbykilled spores.Secretory IgA(sIgA)wasalsodeterminedinsaliva sam-ples(Fig.2B).NIBRG-14dosedaloneinducedvirus-specificsIgA, whilelevelsdetectedinanimalsdosedwithsporesadsorbedwith NIBRG-14werehigher,albeitnotsignificantly(P=0.2240) Althoughcellmediatedimmunitydoesnotpreventinfectionit doesplayanimportantroleintherecoveryofinfluenza-associated complications.PotentT-helperresponsesplayanimportantrolein stimulatingantibodyproductionandtheproliferationofcytotoxic

Tlymphocytes(CTL)andbytheproductionofcytokines[33]

IL-2andIFN-␥producedbyTh1cellsandIL-6,apro-inflammatory

Fig 3.NIBRG-14 cytokine responses Cytokine levels were determined from antigen (NIBRG-14) stimulated splenocytes by ELISA Spleens were taken from mice immunized using a three-dose i.n regime Groups were nạve, autoclaved spores (PY79 K ), NIBRG-14 (1.2 ␮g HA) and PY79 K spores adsorbed with NIBRG-14 (1.2 ␮g HA).

Fig 4. Route of administration Salivary sIgA (panel A) and IgG1 and IgG2a titers (panel B) from mice immunized i.n with three doses (days 1, 15 and 36) of PY79 K spores (1 × 10 9 ) pre-adsorbed with inactivated NIBRG-14 (1.2 ␮g HA, ), or, 1 × 10 9 PY79 K spores administered by the sub-cutaneous (s.c.) route and NIBRG-14 (1.2 ␮g HA) by the i.n route () Controls included nạve mice (×) and mice immunized i.n with inactivated NIBRG-14 alone (1.2 ␮g HA, ) Bleeds were taken on days 0, 14, 35 and 57 Data are presented as arithmetic means and error bars are standard deviation.

cytokine produced by Th2 cells, were measured in NIBRG-14

stimulatedsplenocytes (Fig.3).In eachcase significantlyhigher

responseswerefoundinanimalsdosedwithheat-killed spores

mixedwithNIBRG-14thantotheinactivatedNIBRG-14alone

(IL-2,P=0.0005;IFN-␥,P<0.0001;IL-6,P=0.0001).Theseresultsshow

thatheat-killedvirionadsorbedsporescouldelicitbothTh1-and

Th2-specificcellmediatedandantibodyresponses

Serum anti-HA antibody responses were also further

con-firmedbythehemagglutinationinhibition(HAI)assay,using1%

chicken erythrocytes and inactivated NIBRG-14 Table1 shows

thatHAItitersofmicedosedwithPY79K-NIBRG-14were

signifi-cantlyhigher(P=0.0185)thantothosedosedwithNIBRG-14alone

(187±65vs.88±44)andwithallmicehavinganendpointtiter

higherthan160comparedtojust onemouserespondingwhen

dosedwithNIBRG-14alone(P=0.0189).Importantly,with

inac-tivatedA/Turkey/1/2005,aclade2.2H5N1virus,5outof6mice

respondedwithHAItitersgreaterthan160(200±98)comparedto

nomicedosedwithNIBRG-14alone

We next evaluated thedose of spores required to generate

animmuneresponseusingthethree-dosei.n.regimen.NIBRG-14

(1.2␮gofHA/dose)wasadsorbedwithescalatingconcentrationsof

heat-killedspores(PY79Kat1×107,1×108and1×109).We

cal-culatedthatthehighestdoseofsporeswouldhaveadsorbed1␮gof

NIBRG-14,1×108100ngand1×10710ng.NIBRG-14-specificIgG

(serum)andsIgA(saliva)responsesweredeterminedbyindirect

ELISA(SupplementaryFig3).LevelsofIgGingroupsdosedwith

sporesadsorbedwithNIBRG-14(titre=>2000)were,inallcases,

significantlyhigher(P<0.05)thanincontrolgroups(titre=<1000;

naive,micedosedwith1× 109 PY79K or withNIBRG-14 alone)

(SupplementaryFig.3A).AlthoughtheIgGtiterswerehighestusing

1×109sporesadsorbedwithNIBRG-14nosignificantdifferences (P>0.2)werefoundwithresponsesobtainedusingthelowerspore concentrations.AsimilarprofileofsIgAresponseswasobserved (SupplementaryFig.3B)withhigherantibodytiters(correlating withhighersporedoses)butthesewerenotsignificantlydifferent (P>0.08).Interestingly,usingthethree-doseregimenitwasclear thatadministrationofNIBRG-14aloneproducednodetectablesIgA responseswhilesporesadsorbedwithNIBRG-14produced sub-stantiallevels(n.b.,thiscontrastswiththepreviousexperiment wherelowlevelsofNIBRG-14-specificsIgAwereobservedinmice dosedwithNIBRG-14alone)

3.3 Killedsporescanaugmentimmuneresponseswithoutvirion adsorption

Inordertoinvestigatetheadjuvantpropertyofsporeswe mea-suredtheimmuneresponsestosporesadsorbedwithNIBRG-14 thathadbeendeliveredbyoneroute(i.n.)vs.administrationof sporesandvirionbyseparateroutes(sub-cutaneousandi.n respec-tively).Groupsofmice(n=6)weredosedondays1,15and36using two,3-dose,regimens.Either,1×109 PY79Kspores mixedwith inactiveNIBRG-14(1.2␮gHA)usingthei.n.route,or,1×109PY79K

sporesbythesub-cutaneous(s.c.)routeand1.2␮gofNIBRG-14by thei.n.route(administeredonthesameday).NIBRG-14-specific sIgA(saliva)andserumIgG1andIgG2alevelsweredetermined (Fig.4).Virion-specificsIgAresponses(Fig.4A)showedasignificant increase(P=0.0003)whensporesandvirionwereadministeredby thesameroute.Interestinglythough,dosingofsporesandvirions

byseparateroutesdidshowaclearinduction(P=0.0485)of virion-specificsIgAwhereasnasaladministrationofNIBRG-14generated

Table 1

Hemagglutination inhibition.

Group a A/Vietnam/1194/04 (NIBRG-14, clade 1, H5N1) A/Turkey/1/05 (clade 2.2, H5N1)

a Serum from immunized or unimmunized (nạve) mice.

b Mice that responded with final HI titer >160.

c Geometric mean titer.

d Autoclaved spores.

nolocalantibodyresponses(note thatinanearlier experiment

showninFig.2wedidobservelowlevelsofvirion-specificsIgA

when animalsweredosed withvirionalone) Thesame profile

ofserumantibodyresponses(Fig.4B)wereobservedwithspores

showingthegreatestaugmentationofantibodyresponseswhen

administeredusingthesameroute.Althoughourdatadoesreflect

anypossiblechangesinproductionofsalivaresultingfrom

intra-nasaldeliverytheseresultsshowfirstly,thatthegreatestadjuvant

effectisseenwhensporesareadsorbedwithvirion.Second,that

sporescanaugmentimmuneresponsestoNIBRG-14when

admin-istered by a separateroute.We interpret this asevidence that

spores,bythemselveshavenaturaladjuvantproperties

3.4 Antigenco-deliveredwithsporesprotectsmicefromalethal

doseofinfluenza

Groupsofmice(n=5)wereimmunizedi.n.withPY79Kspores

adsorbedwithNIBRG-14(either0.02␮gor0.5␮g HAperdose)

usingtwodosesandthenchallengedwith20timesthe50%lethal

doseofA/AquaticBird/Korea(H5N2)virus(Fig.5).Therationale

forthechoiceofatwo-doseregimenisexplainedinSection2.All

micedosedwithvirion-adsorbedspores(0.02␮gor0.5␮gofHA)

werefullyprotectedtolethalchallenge.Animalsdosedwithspores

adsorbedwith0.02␮gNIBRG-14didthoughshowatransientdrop

inbodyweightbutwentontorecover.Allmiceinthenạve

con-trolgroupaswellasthosedosedwith0.02␮gofNIBRG-14lost

weightrapidlyandhadtobeeuthanized.Animalsimmunizedi.n

withNIBRG-14(at0.5␮gHA)alonewereaffordedsomeprotection

(4/5mice)butmostinterestingly,heat-killedsporesadministered

withoutadsorbedantigenwerealsoabletoprovideprotectionin3

outof5animals.Inarepeatexperimentusingasimilardosing

regi-menthesameresultswerefound(SupplementaryFig.4).Therefore

theseresultsshowthatheat-killedsporesmixedwithaslittleas

0.02␮gofantigenserveasanadjuvantbeingabletoprovidefull

protectiontoalethaldoseofH5N2

3.5 Administrationofheat-killedsporesaloneisprotective

Beforethechallengereportedabove,NIBRG-14antibodytiters

(IgG)fromimmunizedmiceweremeasured(Fig.6).Thisrevealed

thatanti-NIBRG-14IgGtitersfromthegroupdosedwith0.5␮g

(HA)ofNIBRG-14alonewerehigherthanthegroupdosedwith

sporesadsorbedwith0.02␮g(HA)NIBRG-14,yettheresulting

pro-tectionwaslower(100%vs.80%).Thisindicatesthatantibodiesto

A/AquaticBird/Korea(H5N2)alonearenotsufficientfor

protec-tionandmayimplyamoredirectroleofthesporeinaugmenting

immuneresponsesandprotection.Insupportofthisitisapparent

thatsporesthemselves,withoutdeliveryofvirion,areabletoconfer

some level of protection (60%; Fig 5).Further, when

adminis-teredbyseparateroutessporeswerestillabletoenhanceantibody

responsestotheviruswhendeliveredbyadifferentroute.One

potentialexplanationisthatsporesthemselvesinduceaninnate

immuneresponse,which,althoughshort-livedcouldprovide pro-tection.Toaddressthisweimmunizedgroupsofmice(n=4)with escalating doses of heat-killed spores (ranging from 1× 107 to

2× 109)andthesametwo-doseregimenasthatusedinthe vacci-nationexperimentsdescribedabove.Twenty-sevendaysafterthe 2nddosemicewerechallengedwith5LD50ofH5N2(Fig.7).Mice dosedwith5×108or2×109sporeswerefullyprotectedagainst H5N2challengeforaminimumdurationof14days

3.6 Toll-likereceptor(TLR)signalling InductionofTLRsisoneprincipalmechanismbywhichinnate immunitycanbeactivatedandwhereinteractionofaligandwith

aTLRleads,viatranscriptionofNF-␬Btargetgenes,toregulation

ofthefunctionofantigen-presentingcells(APCs)and upregula-tionofco-stimulatorymoleculessuchasCD80andCD86[34].For

Survival rates 120

100 X X X X X X X X X X X X X X X

80

60

40

20

0

Body weight 120

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0

0

Days post challenge

60 80 100

20

40

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0

Days post challenge Fig 5.Protection against H5N2 challenge BALB/c mice (n = 5 per group) were immunized as described in materials and methods on days 1 and 14 Four weeks after the second immunization (day 42), mice were i.n challenged with 20LD50

of a mouse-adapted strain of the H5N2 virus Mice were then monitored daily to determine survival and body weight changes Groups were mice dosed with PY79 K spores (2 × 10 9 ) adsorbed with 0.02 ␮g of inactive NIBRG-14 (), 0.02 ␮g of inactive NIBRG-14 (), PY79 K (2 × 10 9 ) adsorbed with 0.5 ␮g of inactive NIBRG-14 (X), 0.5 ␮g inactive NIBRG-14 () Control groups were nạve (䊉) and animals dosed with PY79 K (2 × 10 9 ) only () ***P < 0.001 vs control group.

Fig 6. Pre-challenge humoral responses Groups of mice (n = 5) were immunized i.n.

two times (days 1 and 15) with whole inactivated H5N1 virus (NIBRG-14) that

con-tained 0.02 ␮g or 0.5 ␮g of HA in the presence or absence of 2 × 10 9 autoclaved PY79

spores (S) Sera was collected 1 week after the last immunization and tested

indi-vidually for the presence of NIBRG-14-specific antibodies by indirect ELISA Nạve

animals received PBS only Results are expressed as the reciprocal log2 titers and

error bars indicate the standard deviation ***P < 0.001 vs PBS group.

analysisweusedtheRAWBluemurinemacrophagecelllinethat

carriesallTLRgenes(withtheexceptionofTLR5)andhasbeen

engi-neeredtoexpressanNF-␬B-induciblegeneexpressingaprotein,

SEAP,thatismeasurablecolorimetrically.Incubationofheat-killed

PY79sporeswithRAWBlue cellsproduced highlevelsofSEAP

activityinadose-dependentmanner(Fig.8).Aspositivecontrols

E.coliLPSwasalsoshowntoinduceSEAPexpressionaswellas

Fig 7. Protection mediated by spores alone BALB/c mice (n = 4 per group) were

immunized (i.n.) with different doses of killed (autoclaved) PY79 spores on days

1 and 14 Twenty-seven days after the second immunization (day 42), mice were

i.n challenged with 5LD50 of a mouse-adapted strain of the H5N2 virus Mice were

then monitored daily to determine survival and body weight changes Statistical

Fig 8.TLR induction RAW blue murine macrophage cells containing the NF-␬B inducible SEAP gene were incubated with autoclaved PY79 spores, latex beads, E coli LPS or PY79 live vegetative cells (VC, 1 × 10 7 CFU) for 23 h SEAP activity was detected using the QuantiBlue SEAP detection kit Media is the background level from RAW Blue culture Three sets of readings were taken and error bars (denoting 1 SD) are shown The entire experiment was repeated with similar results.

livevegetativecellsofPY79whichhavebeenshownelsewhereto stimulateTLR2expressionpresumablyduetotheirpeptidoglycan content[35].Onefurthercontrolweusedwas1.1␮mlatexbeads thatareofequivalentsizetobacterialspores.Thesefailedtoinduce SEAPexpressionanddemonstratethatTLRactivationbyheat-killed sporesmustarisefromaspecificligandpresentedonthesurface

ofPY79spores

3.7 SporespromoteDCmaturationandrecruitmentofNKcells

Tofurtheridentifytheimpactofsporesoninnatecell activa-tion,weevaluatedtheco-stimulatorymoleculesofbone-marrow deriveddendritic cells(BM-DCs) andmigration ofnaturalkiller (NK)cellsinairwaysbyflowcytometricanalysis.Itiswellknown thatDCiscrucialfortheeffectiveinductionofadaptiveimmune responses by increasing the MHC class molecules, B-7 family moleculesandreleasingcytokinessuchasIL-12p40[36,37].Thus,

weconfirmedtheIL-12p40secretioninBM-DCtreatedwith dif-ferent spore MOIs As expected, spores induced IL-12p40 in a dose-dependentmanner((1–100,sporestoDCs)(Fig.9)

Next,weinvestigatedtherecruitmentofNKcellsintothe air-ways.ToidentifymigrationofNKcells,twostrainsofmice,BALB/c andC57BL/6wereadministered1×109CFUofkilledspores(n=5)

or2␮gofCTbythei.n.routeandsacrificedafter24h.Strikingly, lungsfollowingsporeadministrationshowedmuchhigher recruit-mentof bothNK1.1+ NKcells(CD3−NK1.1+)and DX5+ NK cells (CD3−DX5+)cellsthanthoseofthePBSandCTgroups(Fig.10) TheseresultsdemonstratethatthepresenceofNKcellsisrequired forsporemediatedcellrecruitmentintothelungsandsuggeststhat theseNKcellsmayplayanimportantroleintheinductionofthe innateimmuneresponses

4 Discussion

Resistancetoinfluenzavirusinfectionanddiseasereliesupon mucosalandsystemicimmunitywithsIgAactiveprimarilyinthe upperrespiratorytractandserumIgGinthelowertract[4] Cell-mediatedimmuneresponses,whileimportant,aremorefocused

on clearing virus-infected cells than in prevention Parenteral influenzavaccinescaninduceneutralizingIgGbutfailtoproduce sIgAwiththisbeingoneofthemajordrawbackswithcurrent vac-cinationstrategies[2,4,38].Inthisworkwehaveexploredtheuse

ofbacterialsporesasbothanantigendeliveryvehicleandasan adjuvantusingH5N1asanexample.Unlikeotherstudieswhere

Fig 9.Killed spores induce DC maturation and IL-12p40 production Immature BM-DCs were treated with PBS, different MOIs of spores or CT After incubation for 24 h, BM-DCs were stained with CD11c-FITC plus CD80-PE, CD86-PE and MHC class II-PerCP-eFluor710 Stained cells were analyzed by flow cytometry, FACS Calibur All cytometric data were expressed as mean fluorescence intensity (MFI) and analyzed by using FlowJo software At the end of 24 h, supernatants from BM-DC was collected for measurement

of Il-12p40 by ELISA The results are shown as mean IL-12p40 in ng/ml/5 × 10 5 cells BM-DC ± SD averaged from triplicate cultures Significant differences by paired t-test between maturation from adjuvant and PBS treated DCs by MHC class II (*P < 0.05; **P < 0.01; ***P < 0.001), CD80 ( + P < 0.05; ++ P < 0.01; +++ P < 0.001), and CD86 ( # P < 0.05;

## P < 0.01; ### P < 0.001), respectively.

theBacillussporehasbeengeneticallymanipulated,here,wehave

simplyusedthesporeinaninactive,killed,form.Therationale

for this approach is that B.subtilis spores, approximately1␮m

inlength,havebeenshowntocarrynaturaladjuvantproperties

withthecapacitytoadsorbproteinantigens[20,23].Ourstudies

areencouraginganddemonstratethatsporescanefficientlybind

toH5N1virions, and whenadministered nasallyinduce

virion-specificserumandmucosalantibodyresponsesgreaterthanifthe

virionwereadministeredalone.Justtwodosesofsporesmixed

with0.02␮g(HA)ofH5N1weresufficienttoconfercomplete

pro-tectioninamurinemodelofinfection.Thesefindingsthencompare

favorablywithotheradjuvantandvaccinescurrentlyunder

devel-opment.Forexample,VLP(viruslikeparticles)vaccinescarrying

recombinantHA(0.6␮g)alsowerefoundtoconferprotectionto

avianinfluenzafollowingi.n.(2doses)administrationbut

requir-ingconsiderablymore(30-times)immunogenthanthatofspores

[39].Interestingly,inactivatedcells(notspores)ofBacillusfirmus

(referredtoasdilapidatedB.firmusorDBFs)havealsobeenshown

tocarryanaturaladjuvantpropertyandcaninduceboth

protec-tiveandcross-protectiveimmunityagainstinfluenzavirusAH1N1

[40,41]

TheHAIassayisconsideredthegoldstandardforpredicting

pro-tection(50%)againstaseasonalinfluenzastrain(i.e.,HAItiters>40)

[42]butthiswasnotthecaseusingavianH5N1.Wefoundtiters

of>40in serumfromanimalsdosedwithvirusaloneyetthese

animalswerenotprotected.Thisobservationsupportsthatof

oth-ers[39]andindicatesthatnewcorrelatesofprotectionforH5N1

infectionareneeded.IfefficacyisproveninhumanstheuseofB

subtilissporescouldstretchthelimitedsupplyofstockpiled

vac-cines enabling a simplified adjuvant forinfluenza Evidence for

balancedcell-mediatedresponseswasalsoobservedincludingthe

productionofcytokinescorrelatingwithactivatedTh1andTh2cells

althoughevidenceforaTh1-biaswasobservedbasedontheratioof

IgGisotypes.Animportantconsiderationoffutureinfluenza

adju-vantsishomotypicimmunityandcross-cladeprotection.Clade2

virusesaregeneticallyandantigenicallydistinctfromclade1

iso-latesandhaveemergedrecentlysoitisencouragingthatserum

frommiceimmunizedwithsporesadsorbedwithaclade1virus

ispotentiallycapableofneutralizingavirusfromoneofthemost

divergedsubclades(clade2.2)[2].Presumably,thiscross-clade

pro-tectionarises fromsIgA which is knowntobecross protective

againstvariantstrainsofinfluenza[43,44]

Oneofthemostinterestingfindingsfromthisworkisthat

heat-killedsporescanconferprotectionwithoutco-deliveryofantigen

Our dose-dependent studiesshow 100% protectionfor 14 days

againstalowerlethaldose(5LD50)ofH5N2andthemost straight-forwardexplanationisthatofinnateimmunity.Suchprotection shouldbetypicallyshort-livedandindeedmightaccountforwhy

4weeksafterchallengewithspores(seeFig.5)only60%protection againstH5N2(20LD50).Interestinglythough,againstalower5LD50

challengedose100%protectionwasachieved27daysafterthefinal immunization.Short-terminnateprotectionhowever,isaconcept thatisnowbeingquestionedwithrecentstudiesshowingadefined innatememorycellpopulationthatcanprotectagainstalethal sys-temicinfectionofvacciniavirus[45].Thereisevidencethatinthe GI-tract,specializedM(microfold)cellshavebeenshownableto takeupbacterialsporestothePeyer’spatcheslocatedinthelamina propria[46].Thisregiontogetherwiththeassociated intraepithe-liallymphocytes(IELs)containsrestingmemoryTcells,Bcells,NK cells,macrophagesanddendriticcells.NKcellsinmicehavebeen showntobeactivatedfollowingoraladministrationofbothliveand deadB.subtilisspores[47]whileanumberofcytokines(including IFN-␥)havebeenshowntobeinducedbymacrophagesinfected withB.subtilissporesinvitro[48]andinvivo[23,35,47,49]

Wehaveprovidedfivestrandsofevidenceshowingthat heat-killedspores induceinnate immunity.First,immunization with spores alone can protect against H5N2 infection Second, that sporesefficientlystimulatematurationofDCs.Third,recruitment

ofNKcellsintothelungs.Fourth,sporesadministeredbya par-enteralrouteenhanceresponsestoH5N1deliverednasally.Finally, thatsporesinduceexpressionoftheNK-␬Bpathway.We specu-latethenthatinteractionofasporeligandwithoneormoreTLRs couldactivatetheinnateimmunesysteminthenasal-associated lymphoidtissue(NALT).Furtherworkwillofcourseberequired

toconfirmwhichTLRsaretargetedbysporesandspecifically,the identityoftheligand/sinvolved.However,itisnotablethatrecent studieshaveshownclearlythatpre-stimulationofTLR2andTLR4 withappropriateligandsprotectsmiceagainstlethalinfectionwith highlypathogenicinfluenzavirusesinanimalmodels[50].We pre-dictthenthatsporesmayactsimilarly,byinteractingwiththeTLRs andstimulatinghostdefenses.Inotherstudieswehaveshownthat killedsporesadministeredorallytoGoldenSyrianhamsterscan significantlydelaysymptomsofdiseaseandmortalitywhen chal-lengedwithClostridiumdifficile[14]suggestingthatthisadjuvant propertyisspecificnotonlytoH5N1.Interestingly,other micropar-ticulateadjuvantsappearabletoinduceTLR2andTLR4expression, for example,poly(anhydride) nanoparticleswhich promote Th1 biasedcellularresponses[51]andadjuvantsderivedfrom Gram-positivebacteriaincludingi.n.deliveryofpeptidoglycan-richGEM particles(Gram-positiveenhancermatrix;derivedfromfood-grade

Fig 10.Killed spores induce NK cell recruitment in the lungs Flow cytometric analysis of NK cell populations in lungs from BALB/c and C57BL/6 mice was performed 24 h after treatment with PBS, 2 × 10 9 CFU of spores or 2 ␮g of CT intranasal route The two different types of NK cells were defined as NK1.1 + (CD3−NK1.1 + ) and DX5 + (CD3− DX5 + ) The frequency of NK cells was presented as a percentage of total lymphocytes.

Lactococcuslactis)[52].StudiesusingB.firmusDFPshavealsoshown

upregulationofanumberoftype1interferons,cytokinesandTLR

genes[40]aswellasshort-term(7days)protectiontoH1N1[41,53]

andliveBacilluscereussporesadministeredinanimalfeedhave

beenshowntoenhancespecificPBMCproliferativeresponsesto

InfluenzaH1N1andH3N2antigensinpiglets[54]

Regarding how the virionbinds, in previous workwe have

shownthatproteinscanreadilybindtonegativelychargedspores

whenthepHoftheaqueousphasefallsbelowthepIofthe

respec-tiveprotein[20].Inadditiontoelectrostaticbindinghydrophobic

bondingalsocontributesandthismaybeafactorinvirion

adsorp-tionand,inadditiontothetwoviralproteins(neuraminidaseand

hemagglutinin)itispossiblethatthelipidcomponentoftheviral

envelopecontributestobinding.Asshownherethough,binding

tothesporeisnotessentialforgeneratingsystemicandmucosal

responsesbutthesearesignificantlyimprovedifcombined There-fore,sporesserve asboth adeliverysystemandadjuvant Asa deliverysystemtheyresembleothermicroparticulateadjuvants underdevelopmentthatmimicthesizerangeofnaturalpathogens (20nmto2␮m),suchasliposomes[55],ISCOMs[56]and emul-sions[5]butoffertheadditionaladvantageofsimplifiedproduction andstability Asa microparticulateadjuvant itispredictedthat sporeswouldintroduceadsorbedantigensdirectlyintotheMHC classIandclassIIpresentationpathways[23]andthiswouldoccur eitherinparallelorindependentlytotheinteractionofsporeswith theTLR

Killedspores maybeaviable vaccine-adjuvant forinfluenza withregardtosafety.Weshowherethatthenumberofspores thatarerequiredforgeneratingserumantibodyresponsescanbe reducedby2logswithoutanysignificanteffectonantibodytiter