Microbiota in Human Periodontal Abscess Revealed by 16S ...

Periodontal abscess is an acute exacerbation of chronic periodontitis, exhibiting clinical symptoms of swelling and severe pain in the gingival ... ThisarticleispartoftheResearchTopic BioinformaticsinMicrobiota Viewall 41 Articles Articles QiZhao UniversityofScienceandTechnologyLiaoning,China ThuyDo UniversityofLeeds,UnitedKingdom TakuichiSato NiigataUniversity,Japan Theeditorandreviewers'affiliationsarethelatestprovidedontheirLoopresearchprofilesandmaynotreflecttheirsituationatthetimeofreview. Abstract Introduction MaterialsandMethods Results Discussion DataAvailability EthicsStatement AuthorContributions Funding ConflictofInterestStatement SupplementaryMaterial Footnotes References SuggestaResearchTopic> DownloadArticle DownloadPDF ReadCube EPUB XML(NLM) Supplementary Material Exportcitation EndNote ReferenceManager SimpleTEXTfile BibTex totalviews ViewArticleImpact SuggestaResearchTopic> SHAREON OpenSupplementalData ORIGINALRESEARCHarticle Front.Microbiol.,30July2019 |https://doi.org/10.3389/fmicb.2019.01723 MicrobiotainHumanPeriodontalAbscessRevealedby16SrDNASequencing JiazhenChen1†,XingwenWu2†,DantingZhu3,MengXu3,YouchengYu2,LiyingYu3*andWenhongZhang1* 1DepartmentofInfectiousDiseases,HuashanHospital,FudanUniversity,Shanghai,China 2DepartmentofDentistry,ZhongshanHospital,FudanUniversity,Shanghai,China 3DepartmentofDentistry,HuashanHospital,FudanUniversity,Shanghai,China Periodontalabscessisanoralinfectivediseasecausedbyvariouskindsofbacteria.Weaimedtocharacterizethemicrobiotacompositionofperiodontalabscessesbymetagenomicmethodsandcompareittothatofthecorrespondingpocketandhealthygingivalcrevicetoinvestigatethespecificbacteriaassociatedwiththisdisease.Samplesfromabscesspus(AB),periodontalpocketcoronallyabovetheabscess(PO),andthegingivalcreviceoftheperiodontalhealthytoothwereobtainedfrom20periodontalabscesspatients.Furthermore,healthygingivalcrevicesampleswereobtainedfrom25healthyindividuals.BacterialDNAwasextractedand16SrRNAgenefragmentsweresequencedtocharacterizethemicrobiotaanddeterminetaxonomicclassification.Thebeta-diversityanalysisresultsshowedthattheABandPOgroupshadsimilarcompositions.Porphyromonasgingivalis,Prevotellaintermedia,andotherPrevotellaspp.werethepredominantbacteriaofhumanperiodontalabscesses.TheabundancesofFilifactoralocisandAtopobiumrimaeweresignificantlyhigherinperiodontalabscessesthanintheperiodontalpocket,suggestingtheirassociationwithperiodontalabscessformation.Inconclusion,wecharacterizedthemicrobiotainperiodontalabscessandidentifiedsomespeciesthatarepositivelyassociatedwiththisdisease.Thisprovidesabetterunderstandingofthecomponentsofperiodontalabscesses,whichwillhelpfacilitatethedevelopmentofantibiotictherapystrategies. Introduction Periodontalabscessisanacuteexacerbationofchronicperiodontitis,exhibitingclinicalsymptomsofswellingandseverepaininthegingivalmargin.Itisdefinedasalocalizedsuppurativelesionthatisrelatedtoperiodontalalveolarbonelossandtheaccumulationofpusinthegingivalwalloftheperiodontalpocket(Herreraetal.,2000).Previously,weculturedobligateanaerobicbacteriafromperiodontalabscessandcharacterizedtheirantimicrobialresistanceprofiles(Xieetal.,2014),inwhichthepredominantobligateanaerobeswereblack-pigmentedPrevotella.Althoughtheresultswerepartlyinagreementwiththefindingsofpreviousstudies(Jaramilloetal.,2005;Herreraetal.,2014),somebacteriasuchasthoseofthegenusTreponemawereunculturableandsomepredominantanaerobessuchasPorphyromonasgingivalis,Tannerellaforsythia,andFusobacteriumspp.werelessfrequentlycultured,duetothecultureconditionorsuitabilityofmedium(Xieetal.,2014).Inaddition,ithasbeenreportedthatthetherapeuticeffectofantibioticregimensonperiodontalabscessislimited(SmithandDavies,1986;Herreraetal.,2000,2014),suggestingthecomplexityofassociatedpathogens.Recently,weusedhigh-throughputbarcoded16SrDNAsequencingtocharacterizethemicrobiotaintheperiodontalpocketofpatientswithperiodontitisandcomparedthesetothoseofpatientswithchronicobstructivepulmonarydisease(COPD).Asthenumberofdifferentkindsofbacteriawasdeterminedinthesubgingivalplaqueofeverypatient,wehypothesizedthatperiodontalabscessiscausedbyacombinationofmicrobiota,andspecificpathogensmightbemoredominantintheabscessthaninthepocketandinhealthycontrolssuggestingapositive-associationwithabscessformation.Therefore,aclearerunderstandingofpathogensandthemicrobiotathatcauseperiodontalabscessisnecessary. Inthisstudy,weusedhigh-throughputbarcoded16SrDNAsequencingtechniquetocharacterizethemicrobiotaofperiodontalabscess,thecorrespondingpocket,andhealthygingivalcrevicetoinvestigatethespecificbacteriaassociatedwithperiodontalabscessinhumanperiodontitis. MaterialsandMethods PatientRecruitment Forty-fiveparticipantswererecruitedfromMarch2015toSeptember2015,including20periodontitispatientswithperiodontalabscessand25periodontalhealthyindividuals.Subjectswiththefollowingconditionswereexcludedfromthestudy:pregnancy,useofantibioticsoranti-inflammatorydrugsduringthepast3months,andadministrationofperiodontaltherapyduringthelast6months.ThestudywasapprovedbytheethicscommitteeofHuashanHospital,FudanUniversity(No.KY2014-023).Allparticipantsprovidedsigned,informedconsent.ThestudydesignisshowninSupplementaryFigureS1.Probingdepth(PD),clinicalattachmentloss(CAL),andsimplifiedoralhygieneindex(OHI-S)wereassessedaccordingtoWorldHealthOrganizationrecommendations(WHO,1997).Periodontitiswasdiagnosedaspreviouslydescribed(Wuetal.,2017)withthepresenceofmorethanonetoothwithatleastonesite(mesiobuccal,buccal,distobuccal,mesiolingual,lingual,anddistolingualsites)withPD≥4mm,CAL≥2mm,andbleedingonprobing.Periodontalabscesswasdiagnosedbyaperiodontalspecialistbasedonthepatients’symptomatology,clinicalandradiologicalexaminationfindingssuchasswellingandenlargementofthegingiva,historyofperiodontaldisease,andradiographofthealveolarbonedestructionaroundthecementoenameljunction.Patientswithperiodontalabscessbutwithoutperiodontitiswereexcluded.PeriodontaltoothhealthwasdefinedasPD≤2mmandwithnobleedingonprobingatallsixsites. SpecimenCollectionandIsolationofBacterialDNA Threesampleswerecollectedfromallthepatientswithabscess,includingthesampleofabscesspus(abscessgroup,AB),periodontalpocketcoronallyabovetheabscess(pocketgroup,PO),andgingivalcreviceofperiodontalhealthytooth(patientcontrolgroup,PC).Healthyteethwerealsosampledfromperiodontallyhealthyindividualsasthehealthycontrol(controlgroup,HC).Theabscesssamplesweredrainedafterdecontaminationofthemucosa.ANo.25#sterilizedpaperpoint(Gapadent,China)wasimmersedintothedeepareaofpusfor10safterdrainagewithasterilizedprobe.TheperiodontalpocketandhealthygingivalcrevicesamplesweredippedwithNo.25#sterilizedpaperpointsaspreviouslydescribed(Wuetal.,2017).Theperiodontalpocketsamplewasnotcollectedfromonepatientduetocontaminationwithpus.Allsampleswerestoredintris-EDTAbuffersolutionofpH7.4(Sigma,UnitedStates)inafreezer(−80°C).BacterialDNAwasextractedusingtheQiAampDNAMiniKit(Qiagen,Germany),accordingtothemanufacturer’sinstructions. Amplificationofthe16SrDNAandSequencing Accordingtopreviousstudies(Claessonetal.,2010;Mizrahi-Manetal.,2013;Jorthetal.,2014),hypervariableV3–V4orV4–V5regionsarerecommendedtostudythemicrobiomewhenusingthesecond-generationsequencingmethod.ItwasalsodeterminedthattheV4–V5regionshowedthebestperformanceamongallregions.Therefore,theamplificationoftheV4–V5regionsof16SrDNA,libraryconstruction,indexPCR,andPCRclean-upwereperformedaspreviouslydescribed(Wuetal.,2017).Equalamountsoftagged16SrRNAgeneampliconsofeachsampleweremixedanddenaturedwith0.1MNaOH.Themixedlibrarywasdilutedtoafinalconcentrationof10–20pMusing10mMtrisatpH8.5.Multiplexedpaired-endsequencing(2×300bpreads)ofthe16SrRNAampliconswasperformedusingaMiseqsystem(Illumina,SanDiego,CA,UnitedStates).ImageanalysisandbasecallingwereperformedontheMiseqsystemusingtheMiSeqReportersoftware(MSR).Afterde-multiplexingthedataandremovingthereadsthatfailedthepurityfilter(PF=0),thereadswereconvertedtoFASTQformat. DataAnalyses ThegeneratedFASTQfiles(.fastq)andqualityfileswereacquiredasrawandmappedsequencedatausingdefaultsettingsoftheQIIME2software(version:2018.8)(Caporasoetal.,2010).Eachoperationaltaxonomicunit(OTU)wasgeneratedwith97%similaritycutoffusingUPARSEv7.1andchimericsequenceswereidentifiedandremovedusingUCHIME.Thephylogeneticaffiliationofeach16SrRNAgenesequencewasanalyzedusingRDPClassifier1basedontheSilva(SSU132)16SrRNAdatabaseusingaconfidencethresholdof70%(Amatoetal.,2013;DuanX.B.etal.,2017;DuanX.etal.,2017;Xuetal.,2018).Theoutputwasbasedontheclassificationofreadsatseveraltaxonomiclevels.Thealpha-andbeta-diversityanalyseswerecomputedfromthepreviouslyconstructedOTUtableusingMothursoftware(v.1.30.1)(Schlossetal.,2009)andweightedUniFrac(Lozuponeetal.,2011)analysis.Inthegrouplevel,abundanceanalysiswasdeterminedfromrarefactionfilesbytheMann–Whitneytestbetweenpatientandhealthcontrolgroupsandapairedt-testintheself-controlcomparison(SPSSStatisticsv20.0andGraphPadPrismsoftwarev6.01).Atthepatientlevel,whenanalyzingthesignificantlydominantbacteriainabscesspatients,anyOTUswithabundancedifferencesgreaterthan10%wereconsideredsignificantlydominant.Theminimumabundancecutoffwassetat0.1%abundance,andabundancevalues<0.1%wereneglected. Results ParticipantCharacteristics Characteristicsofthe45enrolledsubjectsarelistedinTable1.Thesexproportionandsmokingstatusbetweenpatientsandcontrolswerenotstatisticallydifferent. TABLE1 Table1.Characteristicsdataoftheenrolledstudyparticipants. TaxonomicClassificationofthe16SrDNASequences Intotal,4.1GBrawdatacontaining1.70millionhigh-qualityandclassifiablereadswereobtainedfrom84samples.Thesequencingdepthwassimilaramongthefourgroupsasfollows:20.85±3.16kreadsintheABgroup,15.50±1.93kreadsinthePOgroup,15.40±1.40kreadsinthePCgroup,and27.31±7.21kreadsintheHCgroup.Amongthesehigh-qualityreads,99.41%wereclassifiedinto322genera,belongingto22phyla,38classes,84orders,and153families.TherewasnosignificantdifferenceintheproportionofunclassifiablesequencesamongthefourgroupsbasedonaKruskal–Wallistest(P=0.186). Alpha-DiversityAnalysis Thealpha-diversityanalysiswasconductedwithtwoindexes,namelytheShannonindeximplicatingcommunitydiversityandtheChao1indeximplicatingcommunityrichness(SupplementaryFigureS2).Plotsweregeneratedandexportedtotherarefactioncurves(Aagaardetal.,2012;SupplementaryFigureS3).TherewasnosignificantdifferenceinChao1andShannonindexesamongthefourgroupsbasedonaone-wayANOVAanalysiswithTukey’smultiplecomparisonstest. Beta-DiversityAnalysis ThemicrobialrawOTUdataweresubjectedtotheprincipalcoordinateanalysis(PCoA)toevaluatethesimilaritiesamongthefourgroups(Figure1).TheresultsshowedthatthesamplesfromtheABandPOgroupshadsimilarmicrobiotacompositions,whichcouldbegroupedintoonecluster,whereastheHCgroupformedanothercluster.ThesamplesfromthePCgroupcouldnotbegroupedintooneclusterandwerescatteredinthe3Dplot(Figure1). FIGURE1 Figure1.BetadiversityanalysisbasedonUniFracanalysis.PlotsweregeneratedusingweightedUniFracdistances.Reddotrepresentstheabscesspus(AB)group.Greendotrepresentsthepocket(PO)group.Yellowdotrepresentsthepatientcontrol(PC)group.Bluedotrepresentsthehealthycontrol(HC)group.Circlesinredandbluerepresentdifferentperiodontalbacterialcommunityclusters,respectively. AbundanceAnalysis Consistentwiththeresultsofthebeta-diversityanalysis,8ofthetop10mostdominantbacteriawerethesameinthesetwogroups.ThegeneraPorphyromonas,Treponema2,Streptococcus,Neisseria,Fusobacterium,Prevotella,Prevotella7andTannerellaaccountedfor61%ofthebacteriaintheABandPOgroups,andexhibitednosignificantdifferencesbetweentheABandPOgroupsbasedonapairedt-test(Table3andSupplementaryFigureS4).Furthermore,basedonthehierarchicalclusteringanalysisofthefourgroups,17ofthe20(85%)ABsamplesand11ofthe19(58%)POsamplesclusteredtogether(lowerclusterinFigure2).InthePCandHCcontrolgroups,8ofthetop10mostdominantbacteriaoverlapped.ThegeneraStreptococcus,Neisseria,Bacteroides,Fusobacterium,Veillonella,Prevotella,Actinomyces,andPorphyromonasaccountedfor57and59%ofthetotalbacteriainthePCandHCgroups,respectively(SupplementaryFigureS4).Fourteenofthe20(70%)PCsamplesand22ofthe25(88%)HCsamplesclusteredtogether(upperclusterinFigure2),indicatingalmostsimilarcompositionsbetweenthesetwogroups. FIGURE2 Figure2.Microbialcommunitybarplotatthegenuslevelwithclustertree.Eachlineshowstheresultsofthebacterialcommunityofonesample.Differentbacterialtaxonomiesatthegenuslevel(cutoffwassetas1%;somebelow1%wereshownas“others”)areshownindifferentcolorsontheright.Thehierarchicalclusterdiagramsinallsamplesbasedonthecommunitycomposition(basedonthealgorithmofBray–Curtis)isshownontheleft.RedcolorindicatesthatthespecimenisintheABgroup.Greencolorindicatesthatthespecimenisinthepocket(PO)group.Yellowcolorindicatesthatthespecimenisinthehealthycontrol(HC).Bluecolorindicatesthatthespecimenisinthepatientcontrol(PC). DominantBacteriainAbscessesatthePatientLevel IntheABgroup,theabundanceofthemostabundantOTUsinallsamplesdidnotexceed50%(Figure3).Thedatashowedthat9of20sampleshadtwoOTUswithanabundanceover10%and7of20sampleshadthreeormoreOTUswithanabundanceover10%.Furthermore,2of20sampleshadtwoOTUswithanabundanceover20%,includingthecombinationofLeptotrichiaceae_Unclassified(27.8%)andP.gingivalisW83(23.1%)inabscess031711C,andLautropia_unculturedbacterium(29.5%)andStreptococcus_unculturedbacterium(21.6%)inabscess052004B(Figure3).Theseresultssuggestthatdiseaseinthesepatientswascausedbybacterialco-infections. FIGURE3 Figure3.RelativeabundanceofgeneraintheABgroup.EachcolumnrepresentsasampleintheABgroup.Thegeneraweresortedindescendingorderofaverageabundanceinabscesssamples.Differentbacterialtaxonomiesatthegenuslevelareshownindifferentcolorsonthebottom.Abundancecutoffinthisbargraphwassetat1%;somebelow1%wereshownas“others.” Torevealthedominantbacteriainvolvedinabscess,wecomparedtheOTUsoftheABgroupwiththoseofthePCgroupbyperformingapairedt-testatthegrouplevel.Theresultsshowedthattheabundanceof6OTUswassignificantlyhigherintheABthaninthePCgroup(SupplementaryTableS1).However,exceptP.gingivalisW83,otherclassicalopportunisticbacteriacausingperiodontalabscesswerenotsignificantlydifferentbasedonthisanalysis.Furthermore,theOTUs,exceptP.gingivalisW83,whichwasrelativelyhighintheABgroup,werelowinaverageabundance(SupplementaryTableS1),suggestingthatgroupcomparisonisnotidealfortheanalysisofdominantbacteria. Consideringtheheterogeneityofdominantopportunisticbacteriaindifferentpatients,weperformedadirectbacterialabundancecomparisonbetweentheABandPCgroupsatthepatientlevel,andthebacteriawithabundancedifferences>10%betweentheABandPCgroupswereidentifiedassignificantlydominantbacteria(Table2andSupplementaryTableS2).Intotal,19OTUsincludingP.gingivalisW83(8/20,40%),Prevotellaspp.(3/20,15%),Prevotellaintermedia(2/20,10%),P.gingivalisTDC60(1/20,5%),andPrevotellaheparinolytica(1/20,5%),werefoundtobesignificantlydominantintheABgroupcomparedwithabundancesinthePCgroupinthecorrespondingnumberofpatients(Table2).Additionally,20OTUs,includingStreptococcusspp.(6/20,30%),Actinomycesspp.(3/20,15%),Lautropiaspp.(3/20,15%),Neisseriaspp.(3/20,15%),Veillonellaspp.(3/20,15%),Fusobacteriumspp.(2/20,10%),P.intermedia(2/20,10%),andBacteroidesneonati(2/20,10%),werefoundtobesignificantlymoredominantinthePCgroupthaninABgroup(Table2).Interestingly,P.intermediawasidentifiedassignificantlydominantintheABgroupofsomepatientsandthePCgroupofotherpatients,indicatingthatitmighthaveaheterogeneousfunctioninabscessformationindifferentpopulations. TABLE2 Table2.TheDominantOTUsinABgroupcomparedtoPCgroupatpatientlevel. SpecificBacteriainAbscessComparedWithThoseinPocketsattheGroupLevel Atthegrouplevel,acomparisonbetweentheABandPOgroupsrevealedspecificbacteriaassociatedwithacutedisease.Theabundanceofbacteriainthesetwogroupswashighlysimilar.Atthegenuslevel,onlyFilifactorandAtopobiumexhibitedsignificantlyhigherabundanceintheABgroup,whereasninegenerapresentedsignificantlylowerinabundanceintheABgroupthaninthePOgroup(Table3).Similarly,3OTUsincludingFilifactoralocisandAtopobiumrimaepresentedsignificantlyhigherabundanceintheABgroup,suggestingthattheymightfunctioninperiodontalabscessformation.Moreover,4OTUsexhibitedsignificantlylowerabundanceintheABgroupthaninthePOgroup(Table3).However,all4OTUswerenotaccuratelyclassifiedtothespecieslevel. TABLE3 Table3.SpecificbacteriaofsignificantdifferenceinABcomparedwithPOgroupbasedonapairedt-testatgrouplevel. BacteriaAssociatedWithAbscessattheGroupLevel Atthegrouplevel,acomparisonbetweentheABandHCgroupswasmadeinthepresentstudy.Atthegenuslevel,24generaincludingPorphyromonas,Treponema2,Tannerella,Filifactor,Parvimonas,andPrevotella1weresignificantlymoreabundantintheABgroupthanintheHCgroup(SupplementaryTableS3).Moreover,25generaincludingStreptococcus,Neisseria,Veillonella,Capnocytophaga,Actinomyces,Selenomonas3,andPrevotella2weresignificantlylessabundantintheABgroupthanintheHCgroup(SupplementaryTableS3).AttheOTUlevel,28OTUsincludingP.gingivalis,Treponema2spp.,P.intermedia,F.alocis,andT.forsythiaexhibitedsignificantlyhigherabundanceintheABgroupthanintheHCgroup.Incontrast,22OTUsincludingStreptococcusspp.,Veillonellaspp.,Actinomycesspp.,andNeisseriaspp.showedlessabundanceintheABgroupthanintheHCgroup(Table4). TABLE4 Table4.MeanrelativeabundanceofOTUswithsignificantstatisticaldifferencebetweenABandHCgroupsatgrouplevel. PorphyromonasgingivalisandP.intermediawerefoundtobedominantintheABgroupatthepatientlevel,andtheirabundancewassignificantlyhigherintheABgroupthanintheHCgroupatthegrouplevel.Additionally,Prevotellaspp.wasidentifiedtobethedominantspeciesintheabscessesofafewpatients,butitsabundancewasnotsignificantlydifferentbetweentheABandHCgroupsanditcouldnotbeclassifiedatthespecieslevel.Incontrast,Streptococcusspp.,Actinomycesspp.,Lautropiaspp.,Neisseriaspp.,andVeillonellaspp.werethedominantspeciesinthePCgroupcomparedwiththoseintheABgroupatthepatientlevel,andtheirabundancewassignificantlylowerintheABgroupthanintheHCgroupatthegrouplevel. SpecificBacteriaAssociatedWithPeriodontitisattheGroupLevel WealsocomparedthebacterialabundancebetweenthePOandHCgroupsatthegroupleveltodetermineifourdatawereconsistentwithwell-knownperiodontitis-associatedbacteria.Atthegenuslevel,23generaincludinggenusPorphyromonas,Treponema2,Tannerella,Fretibacterium,Prevotella1,Filifactor,Dialister,andDesulfobulbusweresignificantlyhigherinthePOgroupthanintheHCgroup,whereas17generaincludingStreptococcus,Bacteroides,Veillonella,Bergeyella,andKingellawerelowerinthePOgroupthanintheHCgroup(SupplementaryTableS4).AttheOTUlevel,theabundanceof29OTUsincludingP.gingivalis,P.intermedia,Treponemaspp.,T.forsythia,F.alocis,andP.heparinolyticawerehigherinthePOgroupthanintheHCgroup.Theresultswerepartlyconcordantwithapreviousstudyabouttheperiodontalredandorangecomplex(Newmanetal.,2015).Incontrast,theabundanceof23OTUsincludingStreptococcusspp.andB.neonatiwerelowerinthePOgroupthanintheHCgroup(SupplementaryTableS5). Discussion Oralperiodontalabscessisanoralinfective,painfuldiseasethatcanspread(Yonedaetal.,2011;Herreraetal.,2014;Satoetal.,2016),anditisavaluablepotentialsignofundiagnosedtype2diabetes(Alagl,2017).Severalstudieshaveidentifiedthedominantmicrobiotabyculture-baseddiagnosticmethods(NewmanandSims,1979;Jaramilloetal.,2005;Xieetal.,2014).Inthepresentstudy,consideringtheheterogeneityofdominantopportunisticbacteriaindifferentpatients,apatientlevelanalysisbetweenabscessandhealthyperiodontiumwasmade,whichshowedthatP.gingivalis,andPrevotellaspp.includingP.intermediawerefoundtobedominantintheabscessofsomepatientscomparedtothoseofhealthyperiodontiumbasedon16SrDNAmetagenomicsequencing.Comparedtothefindingsofourpreviousculture-basedstudy,ourstudyconfirmedthatPrevotellaspp.,andespeciallyP.intermedia,isthedominantspeciesinhumanperiodontalabscess(Xieetal.,2014). However,thisstudydiffersfromthetraditionalculture-basedmethodinthefollowingaspects.First,theculturemethodusuallydetectsthemostabundantbacterium,butthesecondorthethirdmostabundantbacteriacanbeneglected.Forexample,inabscesssamplesfromtwopatients,P.gingivalisW83wasthemostdominantbacterium,andtheabundancewas39.4and28.5%,respectively.Furthermore,thesecondhighestabundantbacteriuminbothsampleswasFusobacteriumspp.,forwhichabundancewasonly12.3and17.9%,respectively,whichcouldbeneglectedintheculturemethod.Second,themetagenomicsequencingmethodcandetectunculturablebacteriaandisnotrestrictedtomediumselectivityoraddictiveantibiotics.Forexample,inoneabscesssample,thedominantbacteriumwasTreponema2spp.(Table2),whichisunculturable.Inaddition,themajordominantbacteriumidentifiedinthepresentstudywasP.gingivalis,whichwasnotdetectedinsomebytheculturemethod(NewmanandSims,1979;Jaramilloetal.,2005;Xieetal.,2014).Thismightbeattributedtomediumselectivityoradditionofselectiveantibioticsthatinhibitthisbacterium. Porphyromonasgingivalisisamemberofperiodontalredcomplex(Socranskyetal.,1998;Newmanetal.,2015),whichisthemostpredominantbacterialclusterdetectedinsubgingivalplaque,andcaninducetheproductionofinterleukin-1inmacrophages(Saitoetal.,1997)andtriggerpolyclonalB-cellactivation(Champaiboonetal.,2000)associatedwithbleedingonprobingandalveolarboneloss(Socranskyetal.,1998).Moreover,itmightbeassociatedwithseveralgeneraldysfunctionsincludingcardiovasculardisease(Kozarovetal.,2005),rheumatoidarthritis(BerthelotandLeGoff,2010),Alzheimer’sdisease(Dominyetal.,2019),andconceptioninwomen(Pajuetal.,2017).Prevotellaspp.includingP.intermediaisamemberoftheperiodontalorangecomplex(Newmanetal.,2015),thesecondmostpredominantbacterialclusterdetectedinsubgingivalplaque,inadditiontobeingrecognizedpathogensofperiodontalinfection.InastudybyJaramilloetal.(2005),themostfrequentsubgingivalbacteriumwasFusobacteriumspp.(75%),followedbyP.intermediaandP.nigrescens(60%),aswellasP.gingivalis(51%).InpartialagreementoftheresultsofourpreviousstudythatP.intermediaisthemostprevalentbacteriuminperiodontalabscess(Xieetal.,2014),inthepresentstudy,thesecondandthirdmostdominantbacteriawerePrevotellaspp.andP.intermedia(25%). Likebrain,lung,andpyogenicliverabscesses,whicharecausedbymultiplekindsofbacteria(Brook,2009;Webbetal.,2014;Yazbecketal.,2014),periodontalabscessismorecomplexthanpreviouslythought.Inthepresentstudy,sevenof20sampleshadthreeormoreOTUswithanabundancegreaterthan10%,andmostOTUswereopportunisticbacteria,suggestingpathogenheterogeneityandbacterialco-infectioninperiodontalabscessdiseases.Abscessoccursinasitethatinhabitsmultiplenormalandopportunisticbacteria,whicharesymbioticandpromoteabscessformation(Newmanetal.,2015).Thesignificantlydominantbacteriaintheabscesswerealsodiverseindifferentpatientsandthedifferencebetweentheabscessandthepocketremainsunknown. Tothebestofourknowledge,thebacteriaintheperiodontalabscessandperiodontalpocketwerecomparedforthefirsttime.Periodontalabscessmightrepresentacuteexacerbationofperiodontitisthatisfavoredbychangesinthesubgingivalmicrobiota,withanincreaseinbacterialvirulenceoradecreaseinhostdefense(Herreraetal.,2014),resultinginthedisruptionofchronicphase(pocket)homeostasisandconversiontotheacutephase(abscess).ItisnoteworthythatonlytwoOTUs,F.alocisandA.rimae,weresignificantlyhigherinabundanceintheABgroupthaninthePOgroupatthegrouplevel,indicatingthattheycouldbeassociatedwiththeexacerbationofchronicperiodontitistoacuteperiodontalabscess,althoughbacterialabundanceanddiversitywerehighlysimilarbetweentheABandPOgroups.F.alocisisaGram-positiveanaerobicrod,whichisnowsuggestedtobeanewperiodontalpathogen(Schlaferetal.,2010;Arunietal.,2015;Camelo-Castilloetal.,2015)withuniquepropertiessuchasresistancetooxidativestress(Arunietal.,2011),theabilitytocausechronicinflammation(Fineetal.,2013),andthecapacitytotriggerapoptosisofgingivalepithelialcells(Moffattetal.,2011).Inthepresentstudy,consistentwiththefindingsofpreviousstudiesthatF.alocisispositivelyassociatedwithperiodontitis,F.alociswasfoundtobemoreabundantinperiodontalpocketsthaninthehealthyperiodontium.Furthermore,itwasmoreabundantinperiodontalabscessthaninthepocket,suggestingthatitisapotential,acuteabscess-related,periodontalpathogen.A.rimaeisananaerobic,Gram-positive,rod-shapedbacterium,whichhasbeensuggestedtobeanendodonticabscess-relatedmicroorganism(Tennertetal.,2014;Georgeetal.,2016).Inthepresentstudy,A.rimaewasfoundtobemoresignificantlyabundantintheabscessthaninthepocketandhealthyperiodontiumofthesamepatient.However,ithasbeenreportedthatA.rimaeismoreprevalentinhealthysubjects(Kumaretal.,2003),suggestingitsroleinperiodontitisformation,whichiscomplexandrequiresfurtherstudy. Thebacteriaassociatedwithperiodontitishavebeenwellinvestigatedpreviously(Liuetal.,2012;Wangetal.,2013).Inthepresentstudy,thefindingthattheabundanceof29OTUsincludingP.gingivalis,P.intermedia,T.forsythia,andF.alociswashigherinthePOgroupthanintheHCgroupwaslargelyinagreementwiththefindingsofpreviousstudies(Liuetal.,2012;Wangetal.,2013).Thesedatafurtherstrengthenthereliabilityofthisstudytoinvestigatetheopportunisticpathogensanddominantmicrobiotaassociatedwithperiodontalabscess. Meanwhile,thereweresomelimitationstoourstudy.First,onlytheV4–V5region,andnotthefull-lengthgene,wassequenced,whichmightresultinsomeunclassifiedOTUslikeStreptococcus_Unclassifiedatthespecieslevel.Second,differentbacterialdatabasescouldleadtodifferencesindetectedspecies,whichrequiresfurthercomparisonswithpreviouslypublishedstudiestoconfirmthesuitabilityofthepresentresearch.Third,thepresentstudydidnotquantifythebacterialloadsinsamples,andquantitativeresearchwillbemorehelpfulinunravelingtherelationshipbetweentheseverityofperiodontalabscessandcertainbacteria. Inconclusion,weused16SrRNA-basedmetagenomicstocharacterizethebacterialprofileofperiodontalabscessinhumansandcompareditwiththecorrespondingperiodontalpocketandhealthyperiodontium.Theresultsshowedthatthebacterialcompositionofperiodontalabscessismorecomplexandmainlyinvolvesbacterialco-infections.Further,P.gingivalis,P.intermedia,andPrevotellaspp.werethepredominantbacteriainhumanperiodontalabscesses.Twospecies,F.alocisandA.rimae,werefoundtobepositivelyassociatedwithabscessformation,althoughtheirbacterialabundanceanddiversityinperiodontalabscessandperiodontalpocketswerehighlysimilar.Recognitionofthebacterialprofileofperiodontalabscessmightrevealnewstrategiesforthediagnosis,surveillance,andtreatmentofperiodontalabscess,includingtheaccurateuseofantibioticsandprobiotics. DataAvailability AllsequencingdatawereuploadedanddepositedtotheSRAdatabasewithprojectnumberPRJNA547446. EthicsStatement ThisstudywasapprovedbytheEthicsCommitteeoftheHuashanHospital,FudanUniversity(No.KY2014-023). AuthorContributions JCandXWconceivedanddesignedthestudy,acquired,analyzed,andinterpretedthedata,anddraftedandcriticallyrevisedthemanuscript.DZ,MX,andYYacquiredthedataandcriticallyrevisedthemanuscript.LYandWZconceivedanddesignedthestudy,analyzedandinterpretedthedata,anddraftedandcriticallyrevisedthemanuscript.Allauthorsapprovedthefinalmanuscriptandagreedtobeaccountableforallaspectsofthework. Funding ThisworkwassupportedbytheNationalNaturalScienceFoundationofChina(81471987). ConflictofInterestStatement Theauthorsdeclarethattheresearchwasconductedintheabsenceofanycommercialorfinancialrelationshipsthatcouldbeconstruedasapotentialconflictofinterest. SupplementaryMaterial TheSupplementaryMaterialforthisarticlecanbefoundonlineat:https://www.frontiersin.org/articles/10.3389/fmicb.2019.01723/full#supplementary-material Footnotes ^http://rdp.cme.msu.edu/ References Aagaard,K.,Riehle,K.,Ma,J.,Segata,N.,Mistretta,T.A.,Coarfa,C.,etal.(2012).Ametagenomicapproachtocharacterizationofthevaginalmicrobiomesignatureinpregnancy.PLoSOne7:e36466.doi:10.1371/journal.pone.0036466 PubMedAbstract|CrossRefFullText|GoogleScholar Alagl,A.S.(2017).Periodontalabscessasapossibleoralclinicalsigninthediagnosisofundiagnoseddiabetesmellitusofelderlyinadentalclinicsetup-a7-yearcross-sectionalstudy.J.Investig.Clin.Dent.8:e12217.doi:10.1111/jicd.12217 PubMedAbstract|CrossRefFullText|GoogleScholar Amato,K.R.,Yeoman,C.J.,Kent,A.,Righini,N.,Carbonero,F.,Estrada,A.,etal.(2013).Habitatdegradationimpactsblackhowlermonkey(Alouattapigra)gastrointestinalmicrobiomes.ISMEJ.7,1344–1353.doi:10.1038/ismej.2013.16 PubMedAbstract|CrossRefFullText|GoogleScholar Aruni,A.W.,Mishra,A.,Dou,Y.,Chioma,O.,Hamilton,B.N.,andFletcher,H.M.(2015).Filifactoralocis–anewemergingperiodontalpathogen.Microb.Infect.17,517–530.doi:10.1016/j.micinf.2015.03.011 PubMedAbstract|CrossRefFullText|GoogleScholar Aruni,A.W.,Roy,F.,andFletcher,H.(2011).FilifactoralocishasvirulenceattributesthatcanenhanceitspersistenceunderoxidativestressconditionsandmediateinvasionofepithelialcellsbyPorphyromonasgingivalis.Infect.Immun.79,3872–3886.doi:10.1128/IAI.05631-11 PubMedAbstract|CrossRefFullText|GoogleScholar Berthelot,J.M.,andLeGoff,B.(2010).Rheumatoidarthritisandperiodontaldisease.JointBoneSpine77,537–541.doi:10.1016/j.jbspin.2010.04.015 PubMedAbstract|CrossRefFullText|GoogleScholar Brook,I.(2009).Microbiologyandantimicrobialtreatmentoforbitalandintracranialcomplicationsofsinusitisinchildrenandtheirmanagement.Int.J.Pediatr.Otorhinolaryngol.73,1183–1186.doi:10.1016/j.ijporl.2009.01.020 PubMedAbstract|CrossRefFullText|GoogleScholar Camelo-Castillo,A.,Novoa,L.,Balsa-Castro,C.,Blanco,J.,Mira,A.,andTomas,I.(2015).Relationshipbetweenperiodontitis-associatedsubgingivalmicrobiotaandclinicalinflammationby16Spyrosequencing.J.Clin.Periodontol.42,1074–1082.doi:10.1111/jcpe.12470 PubMedAbstract|CrossRefFullText|GoogleScholar Caporaso,J.G.,Kuczynski,J.,Stombaugh,J.,Bittinger,K.,Bushman,F.D.,Costello,E.K.,etal.(2010).QIIMEallowsanalysisofhigh-throughputcommunitysequencingdata.Nat.Methods7,335–336.doi:10.1038/nmeth.f.303 PubMedAbstract|CrossRefFullText|GoogleScholar Champaiboon,C.,Yongvanitchit,K.,Pichyangkul,S.,andMahanonda,R.(2000).TheimmunemodulationofB-cellresponsesbyPorphyromonasginginvalisandinterleukin-10.J.Periodontol.71,468–475.doi:10.1902/jop.2000.71.3.468 PubMedAbstract|CrossRefFullText|GoogleScholar Claesson,M.J.,Wang,Q.,O’Sullivan,O.,Greene-Diniz,R.,Cole,J.R.,Ross,R.P.,etal.(2010).Comparisonoftwonext-generationsequencingtechnologiesforresolvinghighlycomplexmicrobiotacompositionusingtandemvariable16SrRNAgeneregions.NucleicAcidsRes.38:e200.doi:10.1093/nar/gkq873 PubMedAbstract|CrossRefFullText|GoogleScholar Dominy,S.S.,Lynch,C.,Ermini,F.,Benedyk,M.,Marczyk,A.,Konradi,A.,etal.(2019).PorphyromonasgingivalisinAlzheimer’sdiseasebrains:evidencefordiseasecausationandtreatmentwithsmall-moleculeinhibitors.Sci.Adv.5:eaau3333.doi:10.1126/sciadv.aau3333 PubMedAbstract|CrossRefFullText|GoogleScholar Duan,X.-B.,Wu,T.-X.,Guo,Y.-C.,Zhou,X.-D.,Lei,Y.-L.,Xu,X.,etal.(2017).Marginalbonelossaroundnon-submergedimplantsisassociatedwithsalivarymicrobiomeduringbonehealing.Int.J.OralSci.9,95–103.doi:10.1038/ijos.2017.18 PubMedAbstract|CrossRefFullText|GoogleScholar Duan,X.,Wu,T.,Xu,X.,Chen,D.,Mo,A.,Lei,Y.,etal.(2017).Smokingmayleadtomarginalbonelossaroundnon-submergedimplantsduringbonehealingbyalteringsalivarymicrobiome:aprospectivestudy.J.Periodontol.88,1297–1308.doi:10.1902/jop.2017.160808 PubMedAbstract|CrossRefFullText|GoogleScholar Fine,D.H.,Markowitz,K.,Fairlie,K.,Tischio-Bereski,D.,Ferrendiz,J.,Furgang,D.,etal.(2013).AconsortiumofAggregatibacteractinomycetemcomitans(Aa),StreptococcusparasanguinisandFilifactoralocisarepresentinsitespriortobonelossinalongitudinalstudyoflocalizedaggressiveperiodontitis.J.Clin.Microbiol.51,2850–2861.doi:10.1128/jcm.00729-13 PubMedAbstract|CrossRefFullText|GoogleScholar George,N.,Flamiatos,E.,Kawasaki,K.,Kim,N.,Carriere,C.,Phan,B.,etal.(2016).Oralmicrobiotaspeciesinacuteapicalendodonticabscesses.J.OralMicrobiol.8:30989.doi:10.3402/jom.v8.30989 PubMedAbstract|CrossRefFullText|GoogleScholar Herrera,D.,Alonso,B.,deArriba,L.,SantaCruz,I.,Serrano,C.,andSanz,M.(2014).Acuteperiodontallesions.Periodontol65,149–177.doi:10.1111/prd.12022 PubMedAbstract|CrossRefFullText|GoogleScholar Herrera,D.,Roldan,S.,andSanz,M.(2000).Theperiodontalabscess:areview.J.Clin.Periodontol.27,377–386.doi:10.1034/j.1600-051x.2000.027006377.x PubMedAbstract|CrossRefFullText|GoogleScholar Jaramillo,A.,Arce,R.M.,Herrera,D.,Betancourth,M.,Botero,J.E.,andContreras,A.(2005).Clinicalandmicrobiologicalcharacterizationofperiodontalabscesses.J.Clin.Periodontol.32,1213–1218.doi:10.1111/j.1600-051X.2005.00839.x PubMedAbstract|CrossRefFullText|GoogleScholar Jorth,P.,Turner,K.H.,Gumus,P.,Nizam,N.,Buduneli,N.,andWhiteley,M.(2014).Metatranscriptomicsofthehumanoralmicrobiomeduringhealthanddisease.mBio5,e01012–e01014.doi:10.1128/mBio.01012-14 PubMedAbstract|CrossRefFullText|GoogleScholar Kozarov,E.V.,Dorn,B.R.,Shelburne,C.E.,Dunn,W.A.Jr.,andProgulske-Fox,A.(2005).HumanatheroscleroticplaquecontainsviableinvasiveActinobacillusactinomycetemcomitansandPorphyromonasgingivalis.Arterioscler.Thromb.Vasc.Biol.25,e17–e18.doi:10.1161/01.ATV.0000155018.67835.1a PubMedAbstract|CrossRefFullText|GoogleScholar Kumar,P.,Griffen,A.,Barton,J.,Paster,B.,Moeschberger,M.,andLeys,E.(2003).Newbacterialspeciesassociatedwithchronicperiodontitis.J.Dent.Res.82,338–344.doi:10.1177/154405910308200503 PubMedAbstract|CrossRefFullText|GoogleScholar Liu,B.,Faller,L.L.,Klitgord,N.,Mazumdar,V.,Ghodsi,M.,Sommer,D.D.,etal.(2012).Deepsequencingoftheoralmicrobiomerevealssignaturesofperiodontaldisease.PLoSOne7:e37919.doi:10.1371/journal.pone.0037919 PubMedAbstract|CrossRefFullText|GoogleScholar Lozupone,C.,Lladser,M.E.,Knights,D.,Stombaugh,J.,andKnight,R.(2011).UniFrac:aneffectivedistancemetricformicrobialcommunitycomparison.ISMEJ.5,169–172.doi:10.1038/ismej.2010.133 PubMedAbstract|CrossRefFullText|GoogleScholar Mizrahi-Man,O.,Davenport,E.R.,andGilad,Y.(2013).Taxonomicclassificationofbacterial16SrRNAgenesusingshortsequencingreads:evaluationofeffectivestudydesigns.PLoSOne8:e53608.doi:10.1371/journal.pone.0053608 PubMedAbstract|CrossRefFullText|GoogleScholar Moffatt,C.E.,Whitmore,S.E.,Griffen,A.L.,Leys,E.J.,andLamont,R.J.(2011).Filifactoralocisinteractionswithgingivalepithelialcells.Mol.OralMicrobiol.26,365–373.doi:10.1111/j.2041-1014.2011.00624.x PubMedAbstract|CrossRefFullText|GoogleScholar Newman,M.G.,andSims,T.N.(1979).Thepredominantcultivablemicrobiotaoftheperiodontalabscess.J.Periodontol.50,350–354.doi:10.1902/jop.1979.50.7.350 PubMedAbstract|CrossRefFullText|GoogleScholar Newman,M.G.,Takei,H.H.,andKlokkevold,P.R.(2015).Carranza’sClinicalPeriodontology,12thEdn.St.Louis,MI:Elsevier. GoogleScholar Paju,S.,Oittinen,J.,Haapala,H.,Asikainen,S.,Paavonen,J.,andPussinen,P.J.(2017).Porphyromonasgingivalismayinterferewithconceptioninwomen.J.OralMicrobiol.9:1330644.doi:10.1080/20002297.2017.1330644 PubMedAbstract|CrossRefFullText|GoogleScholar Saito,A.,Sojar,H.T.,andGenco,R.J.(1997).Interleukin-1geneexpressioninmacrophagesinducedbysurfaceproteincomponentsofPorphyromonasgingivalis:roleoftyrosinekinasesinsignaltransduction.OralMicrobiol.Immunol.12,135–140.doi:10.1111/j.1399-302x.1997.tb00369.x PubMedAbstract|CrossRefFullText|GoogleScholar Sato,J.,Kuroshima,T.,Wada,M.,Satoh,A.,Watanabe,S.,Okamoto,S.,etal.(2016).UseofFDG-PETtodetectachronicodontogenicinfectionasapossiblesourceofthebrainabscess.Odontology104,239–243.doi:10.1007/s10266-015-0218-211 PubMedAbstract|CrossRefFullText|GoogleScholar Schlafer,S.,Riep,B.,Griffen,A.L.,Petrich,A.,Hubner,J.,Berning,M.,etal.(2010).Filifactoralocis–involvementinperiodontalbiofilms.BMCMicrobiol.10:66.doi:10.1186/1471-2180-10-66 PubMedAbstract|CrossRefFullText|GoogleScholar Schloss,P.D.,Westcott,S.L.,Ryabin,T.,Hall,J.R.,Hartmann,M.,Hollister,E.B.,etal.(2009).Introducingmothur:open-source,platform-independent,community-supportedsoftwarefordescribingandcomparingmicrobialcommunities.Appl.Environ.Microbiol.75,7537–7541.doi:10.1128/AEM.01541-09 PubMedAbstract|CrossRefFullText|GoogleScholar Smith,R.,andDavies,R.(1986).Acutelateralperiodontalabscesses.Br.Dent.J.161,176–178.doi:10.1038/sj.bdj.4805920 PubMedAbstract|CrossRefFullText|GoogleScholar Socransky,S.S.,Haffajee,A.D.,Cugini,M.A.,Smith,C.,andKent,R.L.(1998).Microbialcomplexesinsubgingivalplaque.J.Clin.Periodontol.25,134–144.doi:10.1111/j.1600-051x.1998.tb02419.x PubMedAbstract|CrossRefFullText|GoogleScholar Tennert,C.,Fuhrmann,M.,Wittmer,A.,Karygianni,L.,Altenburger,M.J.,Pelz,K.,etal.(2014).Newbacterialcompositioninprimaryandpersistent/secondaryendodonticinfectionswithrespecttoclinicalandradiographicfindings.J.Endod.40,670–677.doi:10.1016/j.joen.2013.10.005 PubMedAbstract|CrossRefFullText|GoogleScholar Wang,J.,Qi,J.,Zhao,H.,He,S.,Zhang,Y.,Wei,S.,etal.(2013).Metagenomicsequencingrevealsmicrobiotaanditsfunctionalpotentialassociatedwithperiodontaldisease.Sci.Rep.3:1843.doi:10.1038/srep01843 PubMedAbstract|CrossRefFullText|GoogleScholar Webb,G.J.,Chapman,T.P.,Cadman,P.J.,andGorard,D.A.(2014).Pyogenicliverabscess.FrontlineGastroenterol.5,60–67.doi:10.1136/flgastro-2013-100371 PubMedAbstract|CrossRefFullText|GoogleScholar WHO(1997).OralHealthSurveys:BasicMethods,4thEdn.Geneva:WorldHealthOrganization. GoogleScholar Wu,X.,Chen,J.,Xu,M.,Zhu,D.,Wang,X.,Chen,Y.,etal.(2017).16SrDNAanalysisofperiodontalplaqueinchronicobstructivepulmonarydiseaseandperiodontitispatients.J.OralMicrobiol.9:1324725.doi:10.1080/20002297.2017.1324725 PubMedAbstract|CrossRefFullText|GoogleScholar Xie,Y.,Chen,J.,He,J.,Miao,X.,Xu,M.,Wu,X.,etal.(2014).Antimicrobialresistanceandprevalenceofresistancegenesofobligateanaerobesisolatedfromperiodontalabscesses.J.Periodontol.85,327–334.doi:10.1902/jop.2013.130081 PubMedAbstract|CrossRefFullText|GoogleScholar Xu,L.,Chen,X.,Wang,Y.,Jiang,W.,Wang,S.,Ling,Z.,etal.(2018).Dynamicalterationsinsalivarymicrobiotarelatedtodentalcariesandageinpreschoolchildrenwithdeciduousdentition:a2-Yearfollow-upstudy.Front.Physiol.9:342.doi:10.3389/fphys.2018.00342 PubMedAbstract|CrossRefFullText|GoogleScholar Yazbeck,M.F.,Dahdel,M.,Kalra,A.,Browne,A.S.,andPratter,M.R.(2014).Lungabscess:updateonmicrobiologyandmanagement.Am.J.Ther.21,217–221.doi:10.1097/MJT.0b013e3182383c9b PubMedAbstract|CrossRefFullText|GoogleScholar Yoneda,M.,Kato,S.,Mawatari,H.,Kirikoshi,H.,Imajo,K.,Fujita,K.,etal.(2011).LiverabscesscausedbyperiodontalbacterialinfectionwithFusobacteriumnecrophorum.Hepatol.Res.41,194–196.doi:10.1111/j.1872-034X.2010.00748.x PubMedAbstract|CrossRefFullText|GoogleScholar Keywords:high-throughputsequencing,oralmicrobiota,periodontalabscess,16SrDNAmetagenomic,periodontalpocket Citation:ChenJ,WuX,ZhuD,XuM,YuY,YuLandZhangW(2019)MicrobiotainHumanPeriodontalAbscessRevealedby16SrDNASequencing.Front.Microbiol.10:1723.doi:10.3389/fmicb.2019.01723 Received:09March2019;Accepted:12July2019;Published:30July2019. Editedby: QiZhao,ShenyangAerospaceUniversity,China Reviewedby: TakuichiSato,NiigataUniversity,Japan ThuyDo,UniversityofLeeds,UnitedKingdom Copyright©2019Chen,Wu,Zhu,Xu,Yu,YuandZhang.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(CCBY).Theuse,distributionorreproductioninotherforumsispermitted,providedtheoriginalauthor(s)andthecopyrightowner(s)arecreditedandthattheoriginalpublicationinthisjournaliscited,inaccordancewithacceptedacademicpractice.Nouse,distributionorreproductionispermittedwhichdoesnotcomplywiththeseterms. *Correspondence:LiyingYu,[email protected];WenhongZhang,[email protected] †Theseauthorshavecontributedequallytothiswork COMMENTARY ORIGINALARTICLE Peoplealsolookedat SuggestaResearchTopic>