Salivary IgA antibody responses to Streptococcus mitis and Streptococcus mutans in preterm and fullterm newborn children

Salivary

IgA

antibody

responses

to

Streptococcus

mitis

and

Streptococcus

mutans

in

preterm

and

fullterm

newborn

children

Ruchele

Dias

Nogueira

*

,

Maria

Lucia

Talarico

Sesso

,

Mariana

Castro

Loureiro

Borges

,

Renata

O.

Mattos-Graner

,

Daniel

James

Smith

,

Virginia

Paes

Leme

Ferriani

a_{UniversityofUberaba,MinasGerais,Brazil}

b_{DepartmentofPediatrics,FacultyofMedicineofRibeiraoPreto,UniversityofSaoPaulo,SaoPaulo,Brazil}

c_{DepartmentofMicrobiologyandImmunology,PiracicabaSchoolofDentistry,UniversityofCampinas,SaoPaulo,Brazil}

d_{DepartmentofImmunology,TheForsythInstitute,MA,UnitedStates}

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Accepted7November2011

Keywords:

Streptococcusmutans Streptococcusmitis

IgA Saliva Prematurity

a

b

s

t

r

a

c

t

Objectives: TheintensitiesandspecificitiesofsalivaryIgAantibodyresponsestoantigensof

Streptococcusmutans,themainpathogenofdentalcaries,mayinfluencecolonizationby theseorganismsduringthefirst1.5yearoflife.Thus,theontogenyofsalivaryIgAresponses tooralcolonizerscontinuestowarrantinvestigation,especiallywithregardtotheinfluence ofbirthconditions,e.g.prematurity,ontheabilityofchildrentoefficientlyrespondtooral microorganisms.Inthisstudy,wecharacterisedthesalivaryantibodyresponsestotwo bacterialspecieswhichareprototypesofpioneerandpathogenicmicroorganismsofthe oralcavity(StreptococcusmitisandStreptococcusmutans,respectively)infullterm(FT)and preterm(PT)newbornchildren.

Methods:Salivasfrom123infants(70FTand53PT)werecollectedduringthefirst10hafter birthandlevelsofIgAandIgMantibodiesandthepresenceofS.mutansandS.mitiswere analysedrespectivelybyELISAandbychequerboardDNA–DNAhybridization.Two sub-groupsof24FTand24PTchildrenwerecomparedwithrespecttopatternsofantibody specificities against S. mutansand S. mitisantigens, using Westernblot assays. Cross-adsorptionof10infant’s salivawastestedtoS.mitis,S.mutansandEnterococcusfaecalis

antigens.

Results: SalivarylevelsofIgAatbirthwere2.5-foldhigherinFTthaninPTchildren(Mann– Whitney;P<0.05).SalivaryIgAantibodiesreactivewithseveralantigensofS.mitisandS. mutansweredetectedatbirthinchildrenwithundetectablelevelsofthosebacteria. Adsorp-tionofinfantsalivawithcellsofS.mutansproducedareductionofantibodiesrecognizingS. mitisantigensinhalfoftheneonates.ThediversityandintensityofIgAresponseswerelower inPTcomparedtoFTchildren,althoughthosedifferenceswerenotsignificant.

Conclusion: ThesedataprovideevidencethatchildrenhavesalivaryIgAantibodiesshortly afterbirth,whichmightinfluencetheestablishmentoftheoralmicrobiota,andthatthe levelsofsalivaryantibodymightberelatedtoprematurity.

#2011ElsevierLtd.

*Correspondingauthorat:FaculdadedeMedicinadeRibeiraoPreto,DepartamentodePuericulturaePediatria,Av.dosBandeirantes,3900, CEP14049-900RibeiraoPreto,SP,Brazil.Tel.:+551636022902;fax:+551636022297.

E-mailaddress:ruchele_nogueira@yahoo.com.br(R.D.Nogueira).

Available

online

at

www.sciencedirect.com

journalhomepage:http://www.elsevier.com/locate/aob

0003–9969#2011ElsevierLtd.

doi:10.1016/j.archoralbio.2011.11.011

Open access under the Elsevier OA license.

1.

Introduction

The mucosal immune system represents the first line of defence in the adaptive immune response to mucosal infection.SecretoryIgA(SIgA)presentinsaliva1_maycontrol

theoralmicrobiotabyreducingtheadherenceofbacteriato theoralmucosaandteeth.2_{ThetotallevelsofSIgAinsaliva}

have beenconsidered asanindicator ofmaturationofthe mucosalimmunesysteminchildren.3–6_{Transientreductions}

inthelevels ofIgA detectedinsalivawereassociatedwith increased susceptibilitytoinfectionsofthe gastrointestinal tract.4,6_{Severalfactorsmightinfluencethedevelopmentofan}

effective mucosal immune response, including nutritional status,breastfeeding,gestationalage,expositiontoantigens andgeneticfactors.7_{Newborninfantsareknowntohavea}

higherfrequencyofmicrobialinfectionsthanolderchildren andadultssoonafterbirth,duetoimmaturityoftheimmune system.8 _{Babies born prematurely (less than 37 weeks}

gestation) have 5 times higher susceptibility to bacterial infections.9_{StreptococcisuchasS.mitisrepresentthemajority}

of bacteria that initiallycolonize the oral cavity.10–12 _After

tooth eruption new species colonize such as S. mutans,5,13

althoughsuchspeciescanbealsodetectedinchildrenbefore tooth eruption.14 _{Prospective study of 5- to 24-month-old}

children heavily exposed to S. mutans showed a complex patternofsalivaryIgAantibodyreactivitytoantigensfromS. mutansandS.mitis,15,16_{suggestingthatresponsesto}

virulence-associatedantigensearlyinlifemayinfluencetheabilityofS. mutans to colonize the oral cavity. Several recent studies showedthatSIgAispresentinsalivaandothersecretionsat birth.6,7 _{However, the influence of these antibodies in the}

establishmentoftheoralmicrobiotaisunknown.

Inthisstudy,wecharacterisedthelevelsandspecificitiesof salivary IgAantibodies toS.mitis andS.mutansantigens in newbornchildren,andcomparedintensitiesandcomplexities ofantibodyresponsesbetweenfullterm(FT)andpreterm(PT) children.

2.

Materials

and

methods

2.1. Studydesign

A total of 123 (70 FT and 53 PT)newborn children inthe Hospital of the University of Ribeirao Preto, Brazil were enrolled in this study, under mothers consent for their participation. This study was approved by the Ethical CommitteeoftheMedicalSchoolofRibeiraoPreto,SP,Brazil, 2963/2007. To be included in the study population, only healthynewborns less than 10hold were includedinthis study. Children with congenital malformations, perinatal hypoxia, intracranial haemorrhage, with length or weight incompatible with gestational ages, or under antibiotic therapywereexcludedfromthisstudy.Gestationalagewas estimatedfromthereporteddateoflastmenstruationperiod and somatic evaluation. Information on maternal and gestational background was obtained by interviewing the mother.Atotalof123children,70FTand53PTwerebornat meangestationalagesof39.2(SD:1.23,range:37–41)and34.5

(SD:2.21,range:30–36.5)weeksrespectively.Forty-one(18PT and23FT)and82(35PTand47FT)infantsweredeliveredby caesareanandvaginallysectionrespectively.Amongstthose children(n=123)fromwhomvolumesofsalivacollectedwere suitableforlaboratoryanalysis,twosubsetsofchildrenwere selected for immunological analysis: 24 PT (<37 weeks of

gestation)and24FTchildren.Forthepurposesofcomparison, thesePTandFTchildrenwerepairedbasedontotalsalivary levelsofIgA,gender,racialbackgroundandbreastfeeding.

2.2. Collectionofsamples

Samplesofwholesalivaunstimulatedwerecollectedusing sterile polypropylene transfer pipettes. Collections were performedinallchildrenatapproximately4–10hafterbirth inordertostandardizethecollectionandtheoralmicrobial exposure, and at least 3h after breastfeeding to avoid contamination with non-salivary components, but in four children(3FTand1PT)salivasampleswerecollectedbefore the first breastfeeding. Solution of 250mM EDTA, pH 5.2 (Sigma,St.Louis,MO,USA)wasaddedtoeachsamplepriorto transportonicetothelaboratorywheretheywerestoredat 808Cuntilanalysis.Samplesofcolostrumandmaternalmilk were collected by manual expression into empty sterile containers on the 1st day of lactation from 20 puerperal mothersofthesomechildreninthestudy.Aftercollection,the maternal samples also were transported on ice to the laboratory, centrifuged to remove lipids components and storedat 808Cuntiluse.

2.3. DetectionofStreptococcalspeciesintheoralcavity

ThepresenceofS.mutansandS.mitisinthesamplesofsalivaof newbornchildrenwasinvestigatedbychequerboardDNA–DNA hybridization with species-specific probes as described by Socranskyet al.17_{Briefly, 0.5M NaOH,pH 13.4(Sigma) was}

addedtosalivasamples.Afterboiling,sampleswereapplied andfixed byexposuretoultra-violetlight(Hybrilinker, UVP, Upland, CA,USA)inindividual lanesonanylon membrane (AmershamBiosciences,LittleChalfont,Buckinghamshire,UK) usingthechequerboardslotblot device(Minislot30, Immu-netics, Cambridge, MA, USA). Digoxigenin-labelled whole genomic DNA probes were prepared for each one of the reference strains(S.mutans[UA159] and S.mitis [ATCC506]) usingarandomprimertechnique.ThesetwoDNAprobeswere hybridizedperpendicularlytothelanesofthebacterialsamples usingtheMiniblotter45(ImmuneticsCambridge,MA,USA)at 708C.Boundprobesweredetectedusing phosphatase-conju-gatedantibodytodigoxigenin (RocheAppliedScience, Man-nheim, Germany) and revealedwith CDP-Star1 _(Amersham

Biosciences, Little Chalfont, Buckinghamshire, United King-dom).Twolanesineachruncontainedstandardsat concen-trationsof105_and106_{cellsofeachspecies.Thesensitivityofthe}

assaywasadjustedtopermitdetectionof104_{cellsofagiven}

withthestandardsonthesamemembrane.Failuretodetecta signalwasrecordedaszero.

2.4. Concentrationofprotein,IgAandIgM

Totalconcentrationofproteininsalivawasdeterminedbythe methodofBradford(Sigma)tocheckforvariationsinsalivary flow.TotallevelsofIgAandIgMweredeterminedincapture ELISAassaysaspreviouslydescribed.15

2.5. SalivaryandmaternalmilkIgAreactivityagainst

bacterialantigens

PatternsofreactivityofsalivaryIgAandbreastmilkantibody againstS.mitis(ATCC906)and S.mutans(UA 159)Agswere determined in Western blot assays. Sixteen micrograms of antigen extracts prepared as previously described15 _were

loadedper lane, separated by sodium dodecyl sulphate–6% polyacrylamidegelelectrophoresis,andtransferredto nitrocel-lulose membranes. After being stained with Red Ponceau (Sigma),membraneswerewashedandblockedovernightat48C (in Tris-buffered saline–Tween, pH 7.5, 5% nonfat milk). Incubations with samples diluted 1:100 were performed at roomtemperaturefor2h.Asnegativecontrols,membranes were incubated only with blocking buffer, and as positive controls,membraneswereincubatedwithastandardsaliva sampleobtainedfromanadultwhosepatternofreactionwithS. mutans and S. mitis antigen extracts had been previously measured.ThesecondaryantibodywasgoatIgGanti-human IgAconjugatedwithhorseradishperoxidase(1:4000dilution). Antibody reactions were developed using an ECL system (AmershamBiosciences).Forthispurpose,immunoblotswere incubatedwithECLdetectionsolutionandthenexposedtothe

same X-rayfilm for 5min. Thedeveloped X-rayfilms were scannedinascanningdensitometer(Bio-RadGS-700Imaging Densitometer) to analyse patterns of antigen recognition, includingthenumberandintensityofreactivebands.Afilm blankvaluewassubtractedfromthevalueofthereactiveband. In order to determine whether any of the antibodies detectedwereuniquelyspecificforasinglespecies,tensaliva samples (3 PT and 7 FT) were adsorbed sequentially with antigensofcellsofS.mitis,S.mutansandEnterococcusfaecalisas describedbyColeetal.18_{Antibodyactivitiesremainingafter}

adsorption(percent)weredeterminedbydividingtheoptical density at 450nm of each adsorbed saliva by that of the corresponding unabsorbedsaliva at the same dilution and multiplyingby100.

2.6. Statisticalanalysis

Associations between concentrations of IgA,IgM and total protein, and patterns ofantibody reactions were testedby Spearmancorrelationanalysis.Differencesinthe densitome-tryvaluesofreactiveAgsbetweenthesubsetsofPTandFT childrenwereanalysedbyaMann–WhitneyUtest. Compar-isonsofthefrequenciesofchildrenwithdistinctIgAantibody specificities were tested by a chi-square test. A P-value of<0.05wasconsideredstatisticallysignificant.

3.

Results

3.1. LevelsofIgAandIgM

ImmunoglobulinAandIgMweredetectedinallsalivasamples tested(n=123).Therewerestatisticallysignificantdifferences

Fig.1–LevelsofsalivaryIgA(A)andIgM(B)observedamongst123children,Preterm(PT)andFullterm(FT)measuredinthe samesamples.Squaresrepresentabsolutesamountsofimmunoglobulinsinsalivas(mg/ml).Asterisksrepresentthe

inlevelsofsalivaryIgAbetweenPT(median:0.78,interquartile range[IQR]:0.43–1.49)andFT(median:1.09,IQR:0.55–2.75) (Mann–WhitneyUtest,P<0.05).Apositivecorrelationwas

observed between salivary levels of IgA and IgM in each group(Spearman’s,r>0.75,P<0.01). Fluctuation of

abso-lutelevels ofIgA(A)andIgM(B) areshowninFig.1. The medianconcentrationoftotalproteininsalivawas834.3mg/

ml (IQR: 613.9–1219.4), with similar levels in FT and PT infants (Mann–Whitney, P>0.05). The median ratios of

valuesofIgAnormalizedbyproteinconcentration(median ratio, 0.10, IQR: 0.05–0.20) determined for PT was signifi-cantlylowerthanthatobservedinFTinfants(medianratio, 0.22,IQR:0.06–0.40;Mann–Whitney;P<0.05).Nosignificant

differencesweredetectedinmedianratiosofvaluesofIgM normalized by protein concentration between groups (PT=medianratio,0.08,IQR:0.02–0.15vsFT=medianratio, 0.10, IQR:0.02–0.20,Mann–Whitney; P>0.05). Themedian

concentrationoftotal IgAinmaternalmilkwas2567.8mg/

ml (IQR: 834.0–3986.3) not differing between mothers of preterm and full-term infants (Mann–Whitney, P>0.05).

Also, the levels of immunoglobulins and proteins were similar in infants delivered by caesarean section or vaginally(Mann–Whitney;P>0.05).

3.2. ComplexityandintensityofIgAresponsetoS.

mutansandS.mitisAgs

Detectionofstreptococciinoralsamplesusingchequerboard DNA–DNAhybridizationassaysshowedthatnochildrenhave

S.mitisorS.mutansinsalivasamplesatthelevelstested.Fifty and37.5%ofPT(n=12)andFT(n=9)respectivelydidnotshow IgA-reactivebands totheantigenextracts tested.However, amongst the IgA-reactive children, several bands of IgA reactivitywithS.mutansandS.mitisantigenswereidentified, especiallyinFTchildren.Examplesofimmunoblotsincubated with salivas from three representative pairs of PT and FT childrenagainstAgsfromS.mutansandS.mitisareshownin

Fig.2A.MaternalandchildpatternsofIgA-reactivitywithS. mitisandS.mutansantigenswerecompared.Interestingly,few coincident bands were noted between mother and child. Medianpercentagevaluesofcoincidentbandstototalnumber ofbandsidentifiedwere5and8%forS.mitisandS.mutans, respectively.Threepairsofexamplesofimmunoblots com-paringthemothermilkandherbaby’ssalivaareshownin

Fig.2B.Inaddition,theimmunoblotsfromtwochildren(1PT and1FT)whowerenotyetbreastfedpresentedIgAresponse to antigens from S. mutans and S. mitis (Fig. 2A, pair 10).

AntigensinbothspeciesareshowntoreactwithsalivaryIgA inbothpairs.Takentogether,thesedatasuggestthattheIgA reactivitiesseeninthechildren’ssalivasarederived,atleastin part,fromthechild.

Thenumbers of children that presented detectable IgA antibodiestoantigensofeachStreptococcalspeciesandmean numbers ofreactivebands detected areshown in Table 1. Although IgA antibody responses were detected more fre-quently to S. mitis Ags (n=23, [11 PT and 12 FT]) when comparedtoS.mutansantigens(n=18,[7PTand11FT])those differences were not significant (Mann–Whitney, P>0.05).

Additionally, the number of IgA-reactive bands to S. mitis

antigenswas significantlyhigher inFTthan inPTchildren (Mann–WhitneyUtest,P0.05).SixpercentoftheSDS–PAGE gelsanalysedallowedthevisualizationofimportantantigens fromS.mutans:Ag I/II(185kDa),GTFC(160kDa)andGbpB (56kDa)andofS.mitis: IgA-protease (202kDa). Twenty-one percentofchildren(n=10,[3PTand7FT])hadIgAreactiveto Ag202kDa–S.mitisand16.5(n=8,[2PTand6FT])and17(n=8 [4FTand4PT])%ofchildrenpresentedIgAreactiveto185and 160kDa–S.mutansAgsrespectively(Table1).Wedidnotfind childrenwithIgAreactivewithbandsinthe56kDaregionofS. mutans blots. There were no significant differences in the numberofPTand FTchildrenwithIgA responsestothese antigens(QuiSquaretest,q<2.01;P>0.27).

TherewerevariationsintheintensitiesandnumbersofIgA antibody reactions with the recognized bands amongst childreninbothgroups.Table1showsthesumsofintensities ofIgAreactionswithallbandsdetectedforeachspecies(total intensities)observedinchildrenoftheFTandPTgroups.In general,FTchildrenpresentedthehighestintensityofIgAto allantigenstestedbutthosedifferenceswerenotstatistically significant(Mann–WhitneyUtest, P>0.2),likelyduetothe

highvariabilityinintensitiesofresponseamongstchildrenof eachgroup.

TheresultsshowedthatSIgAantibodyfrom10samples(3 PTand7FT)testeddidnotreactwithE.faecalisantigens,as SIgAresponsestoS.mutansandS.mitiswerenotreducedbyE. faecaliscross-adsorption.Ontheotherhand,whensamples (n=10)wereadsorbedwithcellsofS.mitis,thereweremean reductionsof22%ofSIgAtoS.mutansin5children(4FTand1

PT). In the same children (n=5), there was also a mean reduction of 45% of SIgA to S. mitis when samples were adsorbedpreviouslywithcellsofS.mutans.

4.

Discussion

SalivaryIgAantibodiesplayseveralrolesinthemodulationof theestablishmentofthemicrobiotacompatiblewithhealth homeostasis19_{andformafirstlineofdefenceagainstspecific}

pathogens.19 _{Salivary IgA antibodies neutralize antigenic}

componentsinvolvedinmicrobialvirulenceandmightblock surfaceadhesinsimportantforcolonizationofthemucosa.20

Inthesaliva,secretoryIgApredominates,butearlyinlife,IgM isalsonormallydetected.6_{Previously,itwasdescribedthat}

IgA canbedetectedinsalivaatbirth.6,7_{Hereweshowthat}

detectable levels of salivary IgA antibodies reactive with commonbacterialspeciesoftheoralcavitycanbedetectedin thefirsthoursoflife.

With respecttogestational age, notonlywere thetotal levelsofsalivaryIgAhigherinFT(upto2.5-fold)butalsothe complexityofIgAagainstbacterialspecies(Fig.2A,Table1), suggesting that prematurity can lead to a delay in IgA responsesatinitialstagesofantigenicchallenge.Longitudinal comparisons oflevels ofIgAinPTand FTinfantscouldbe helpful to clarify the extent to which this difference is maintainedovertime.

Previously,wesuggestedthatpatternsofspecificityofIgA antibody responses to S. mutans antigens might be more importantthantotallevelsofreactiveIgAantibodies.15_Inthis

study, weobserved thatpatterns ofproteinbands reactive withsalivaryIgAwerevariableamongstnewborn(Fig.2A).We reasonedthatmucosalresponses,mostfrequentlydetectedin newbornstoantigensofS.mitis,apioneercolonizeroforal mucosa, might develop earlier than to S. mutans, which colonizechildrenatalaterage.5,13,14_{Byseparatingproteinsin}

6%SDS–PAGEgelsitispossibletovisualizethethreemain cell-associatedantigensofS.mutans,AgI/II,21_GTFC22_andGbpB5

withmolecularmassesof185,160and56kDa respectively. These antigensareinvolvedinthecapacityofS.mutansto adhereandaccumulateinthedentalbiofilm.Apreviousstudy

Table1–ComparisonsofthefrequenciesofchildrenwithpositivesalivaryIgAresponsesandsumofmeanintensitiesto totalandspecificAgsfromStreptococcusmitis(SMI)andStreptococcusmutans(SM)betweenpreterm(PT)andfullterm(FT) groups.

Antigens Number(%)ofchildrenwith

salivarylgAreactive

SumofmeanIntensityof salivarylgAreactiveSD

PT

n=24

FT

n=24

PT

n=24

FT

n=24

Streptococcusmitis

TotalSMIAgs 11(46) 12(50) 296575 1020129482

202kDaAgs(lgA1-protease) 3(13) 7(29) 85244 107226 MeannumberofreactiveAgsSD 1.01.5a _2.74.1a

Streptococcusmutans

TotalSMAgs 7(29) 11(46) 215567 466896

160kDaAgs(GTFC) 2(8) 6(25) 1653 47125

185kDaAgs(AgI/II) 4(17) 4(17) 2188 46104

MeannumberofreactiveAgsSD 0.91.6 1.52.0

showed that some five-month-old children presented with salivaryIgAreactivetoallthisantigens,especiallytoGbpBand mayhavearoleinmodulatingthelevelofcolonizationbyS. mutans.15 In the present study, approximately 30% of the childrenevaluated(n=16/48)presentedIgAagainstAgI/IIand GTFC, but not against GbpB (Table 1). Also, 20% of saliva samplesfromnewbornchildrenwerereactivewithaS.mitis

202kDacomponent(Table1),suggestingthepresenceofIgA reactive toIgA1-protease, an antigen importantfor S.mitis

establishmentintheoralcavity.23,24

Inthepresentstudyweanalysedthespecificityofsalivary SIgAantibodiesreactivewithS.mutans,S.mitisandE.faecalis,to test whether SIgA antibodiesreactive with commensal oral bacteriawereinducedbythesebacteriaandwere,therefore, specific to them or whether they were induced by cross-reactionswithotherbacteria.Theresultsofcross-adsorption showedthatinhalfofthesalivatested(n=5of10),therewasa reductionofthesalivaryIgAtoS.mutanswhentheplatewas previouslyabsorbedwithS.mitisantigens.Asimilarresultof levelsofsalivaryIgAtoS.mitisoccurredwhentheplatewas covered with S. mutans. The elimination of salivary IgA antibodiesreactivewiththetestspeciesfollowingsequential adsorptionofsalivasampleswitheachstreptococcalspecies supportspartiallytheconclusionthattheantibodieswere cross-reactiveratherthan species specific,as described previously.18,25

AlthoughinourpreviousstudywedetectedSIgAantibodies reactivewithS.mutansinsalivaofpredentateinfantswhodid notharbourS.mutans,15_{here,thepresenceofspecificantibody}

atbirthisunlikelytohavebeeninducedwithin10h,sinceit takesatleastaweekfortheuptake,processingofantigen,Bcell selection and migration to local sites, differentiation into plasma cells leading to antibody secretion and endosomal transferintotheglandlumen.Thus,somehypothesescanbe raisedtoaddressthisearlyresponseofSIgAtoS.mutansandS. mitis antigens.Firstly, thepresence ofresidual ofIgA from maternalmilkintheoralcavityofchildrencannotbeexcluded, eventhoughthesampleshavebeencollectedatleast3hafter breastfeeding.Forthisreason,wecomparedimmunoblottingof infantsalivary samples withtheirrespectivematernalmilk samples(Fig. 2B). The majority ofantigens thatweremore frequently reactivein theinfant salivary samples were not recognizedbymaternalmilk(Fig.2B).Additionally, immuno-blotsfromchildrenwhodidnotreceivematernalmilk(Fig.1A, pair10)presentedwithIgAantibodyreactivitywithS.mutans

andS.mitisantigens.Thepersistenceofsecretoryantibodiesin theoralcavity(e.g.,followingbreastfeeding)stronglydepend ontheiradhesiontosalivarypellicleontoothsurfaces.26Since newborns are edentulous, this condition for persistence of maternalIgAisabsent.

An alternative hypothesis could be associated with the plurispecific protection at mucosal surfaces, proposed by Quanandcoworkers,27whofoundthatSIgAantibodiesfrom humansalivareactedwithactin,myosinandtubulinbutalso withantigens fromStreptococcus pyogenes. Also, those anti-bodies could result from stimulation without antigenic exposure,astheresultofanti-idiotypeinduction28_or

intra-uterinestimulation.Thus,severalbacteriahavebeenisolated fromumbilicalcord blood,amniotic fluidandfoetal mem-braneswithoutclinicalorhistologicalevidenceofinfectionor inflammationinpairsofmothersandchildren.29

In summary,the results show that detectable levels of salivaryIgAantibodiesreactivetooralbacterialspeciescanbe detectedwithinthefirsthoursafterbirth.Furthermore,the salivary IgA concentrations and IgA antibody specificities appear to influenced by the gestational age, which might reflect thelevel ofimmunological maturity ofthe mucosal immunesystem.Thesefindingssupportfurtherstudyabout the investigation of antibody and microbial sources from motherinordertoclarifytheroleanddevelopmentofmucosal immuneresponseinneonates.

Acknowledgments

Conflictofinterest:Theauthorsdeclarenoconflictofinterest. Funding: This study was supported by Fundac¸a˜o de Amparo a` PesquisadoEstado deSa˜o Paulo (FAPESP),proc. 07/57346-5, and proc.07/50807-7 andConselhoNacionalde Pesquisa(CNPq),proc.472928/2007-4.

Ethicalapproval:ThisstudywasapprovedbytheEthical CommitteeoftheMedicalSchoolofRibeiraoPreto,SP,Brazil, referencenumber:2963/2007.

r

e

f

e

r

e

n

c

e

s

1. RussellMW,HajishengallisG,ChildersNK,MichalekSM. Secretoryimmunityindefenseagainstcariogenicmutans streptococci.CariesRes1999;33_(1):4–15.

2. MarcotteH,LavoieMC.Oralmicrobialecologyandtherole ofsalivaryimmunoglobulinA.MicrobiolMolBiolRev

1998;62_{(1):71–109.}

3. BrandtzaegP,FarstadIN,HaraldsenG,JahnsenFL.Cellular andmolecularmechanismsforinductionofmucosal immunity.DevBiolStand1998;92_:93–108.

4. GleesonM,CrippsAW,ClancyRL.Modifiersofthehuman mucosalimmunesystem.ImmunCellBiol1995;73_{(5):397–404.}

5. SmithDJ,KingWF,AkitaH,TaubmanMA.Associationof salivaryimmunoglobulinAantibodyandinitialmutans streptococcalinfection.OralMicrobiolImmunol

1998;13_{(5):278–85.}

6. WanAK,SeowWK,PurdieDM,BirdPS,WalshLJ,Tudehope DI.Immunoglobulinsinsalivaofpretermandfull-term infants.Alongitudinalstudyfrom0-18monthsofage.Oral MicrobiolImmunol2003;18_(2):72–8.

7. MaruyamaK,HidaM,KohgoT,FukunagaY.Changesin salivaryandfecalsecretoryIgAininfantsunderdifferent feedingregimens.PediatrInt2009;51_(3):342–5.

8. ClappDW.Developmentalregulationoftheimmune system.SeminPerinatol2006;30_(2):69–72.

9. KiliegmanRM,RottmanCJ,BehrmanRE.Strategiesforthe preventionoflowbirthweight.AmJObstetGynecol

1990;162_{(4):1073–83.}

10.PearceC,BowdenGH,EvansM,FitzsimmonsSP,JohnsonJ, SheridanMJ,etal.Identificationofpioneerviridans streptococciintheoralcavityofhumanneonates.JMed Microbiol1995;42_(1):67–72.

11.HohwyJ,ReinholdtJ,KilianM.Populationdynamicsof

Streptococcusmitisinitsnaturalhabitat.InfectImmun

2001;69_{(10):6055–63.}

12.SeowWK,LamJH,TsangAK,HolcombeT,BirdPS. OralStreptococcusspeciesinpre-termandfull-term children–alongitudinalstudy.IntJPaediatrDent

13.CaufieldPW,CutterGR,DasanayakeAP.Initialacquisition ofmutansstreptococcibyinfants:evidenceforadiscrete windowofinfectivity.JDentRes1993;72_(1):37–45.

14.AlvesAC,NogueiraRD,StippRN,PampoliniF,MoraesAB, Gonc¸alvesRB,etal.Prospectivestudyofpotentialsourcesof

Streptococcusmutanstransmissioninnurseryschool children.JMedMicrobiol2009;58_(4):476–8.

15.NogueiraRD,AlvesAC,NapimogaMH,SmithDJ, Mattos-GranerRO.CharacterizationofsalivaryimmunoglobulinA responsesinchildrenheavilyexposedtotheoral

bacteriumStreptococcusmutans:influenceofspecific antigenrecognitionininfection.InfectImmun

2005;73_{(9):5675–84.}

16.NogueiraRD,AlvesAC,KingWF,GoncalvesRB,Ho¨flingJF, SmithDJ,etal.Age-specificsalivaryimmunoglobulinA responsetoStreptococcusmutansGbpB.ClinVaccineImmunol

2007;14_(6):804–7.

17.SocranskySS,SmithC,MartinL,PasterBJ,DewhirstFE, LevinAE.‘Checkerboard’DNA–DNAhybridization.

Biotechniques1994;17_{(4):788–92.}

18.Cole.etal.Humoralimmunitytocommensaloralbacteria inhumaninfants:salivarysecretoryimmunoglobulinA antibodiesreactivewithStreptococcusmitisbiovar1,

Streptococcusoralis,Streptococcusmutans,andEnterococcus faecalisduringthefirsttwoyearsoflife.InfectImmun

1999;67_:1878–86.

19.PetersonDA,McNultyNP,GurugeJL,GordonJI.IgAresponse tosymbioticbacteriaasamediatorofguthomeostasis.Cell HostMicrobe2007;15_{(2):328–39.}

20.BrandtzaegP.Mucosalimmunity:induction,dissemination, andeffectorfunctions.ScandJImmunol2009;70_{(6):505–15.}

21.DemuthDR,DavisCA,CornerAM,LamontRJ,LeboyPS, MalamudD.CloningandexpressionofaStreptococcus

sanguissurfaceantigenthatinteractswithahumansalivary agglutinin.InfectImmun1988;56_{(9):2484–90.}

22.HanadaN,KuramitsuHK.Isolationandcharacterizationof theStreptococcusmutansgtfCgene,codingforsynthesisof bothsolubleandinsolubleglucans.InfectImmun

1988;56_{(8):1999–2005.}

23.PoulsenK,ReinholdtJ,JespersgaardC,BoyeK,BrownTA, HaugeM,etal.Comprehensivegeneticstudyof

streptococcalimmunoglobulinA1proteases:evidencefor recombinationwithinandbetweenspecies.InfectImmun

1998;66_{(1):181–90.}

24.Ko¨no¨nenE,Jousimies-SomerH,BrykA,KilpT,KilianM. Establishmentofstreptococciintheupperrespiratorytract: longitudinalchangesinthemouthandnasopharynxupto2 yearsofage.JMedMicrobiol2002;51_{(9):723–30.}

25.MaJK,HikmatBY,WycoffK,VineND,ChargelegueD,YuL, etal.Characterizationofarecombinantplantmonoclonal secretoryantibodyandpreventiveimmunotherapyin humans.NatMed1998;4_(5):601–6.

26.BammanLL,GibbonsRJ,ImmunoglobulinA.antibodies reactivewithStreptococcusmutansinthesalivaofadults, children,andpredentateinfants.JClinMicrobiol

1979;10_:538–43.

27.QuanCP,BernmanA,PiresR,AvrameasS,BouvetJP. NaturalpolyreactivesecretoryimmunoglobulinA

autoantibodiesasapossiblebarriertoinfectioninhumans.

InfectImmun1997;65_{:3997–4004.}

28.SmithDJ,TaubmanMA.Ontogenyofimmunitytooral microbiotainhumans.CritRevOralBiolMed1992;3_:109–33.