Assessment of 16S rRNA gene primers for studying bacterial ...

To evaluate bacterial 16S rRNA gene primer pairs in regard to bacterial diversity coverage and/or accurate taxonomic assignment, four primers ... Skiptomaincontent Advertisement SearchallSpringerOpenarticles Search Assessmentof16SrRNAgeneprimersforstudyingbacterialcommunitystructureandfunctionofagingflue-curedtobaccos DownloadPDF DownloadPDF Originalarticle OpenAccess Published:10November2018 Assessmentof16SrRNAgeneprimersforstudyingbacterialcommunitystructureandfunctionofagingflue-curedtobaccos FanWang1,2,XiaoMen1,2,GeZhang1,2,KaichaoLiang3,YuhuaXin3,JuanWang3,AijunLi3,HaiboZhang1,2,HaobaoLiu3,4&LijunWu5  AMBExpress volume 8,Article number: 182(2018) Citethisarticle 6994Accesses 17Citations 8Altmetric Metricsdetails AbstractSelectionofoptimalprimerpairsin16SrRNAgenesequencingisapivotalissueinmicroorganismdiversityanalysis.However,limitedefforthasbeenputintoinvestigationofspecificprimersetsforanalysisofthebacterialdiversityofagingflue-curedtobaccos(AFTs),aswellaspredictionofthefunctionofthebacterialcommunity.Inthisstudy,theperformanceoffourprimerpairsindeterminingbacterialcommunitystructurebasedon16SrRNAgenesequencesinAFTswasassessed,andthefunctionsofgeneswerepredictedusingPhylogeneticInvestigationofCommunitiesbyReconstructionofUnobservedStates(PICRUSt).Resultsrevealedthattheprimerset799F–1193RcoveringtheamplificationregionV5V6V7gaveamoreaccuratepictureofthebacterialcommunitystructureofAFTs,withlowerco-amplificationlevelsofchloroplastandmitochondrialgenes,andmoregeneracoveredthanwhenusingtheotherprimers.Inaddition,functionalgenepredictionsuggestedthatthemicrobiomeofAFTswasinvolvedinkindsofinterestedpathways.AhighabundanceoffunctionalgenesinvolvedinnitrogenmetabolismwasdetectedinAFTs,reflectingahighlevelofbacteriainvolvedindegradingharmfulnitrogencompoundsandgeneratingnitrogenousnutrientsforothers.Additionally,thefunctionalgenesinvolvedinbiosynthesisofvaluablemetabolitesanddegradationoftoxiccompoundsprovidedinformationthattheAFTspossessahugelibraryofmicroorganismsandgenesthatcouldbeappliedtofurtherstudies.Allofthesefindingsprovideasignificancereferenceforresearchersworkingonthebacterialdiversityassessmentoftobacco-relatedsamples. IntroductionMicroorganismsaredominantdriversofbiogeochemicalprocesses(Shinichietal.2015),yetalargenumberofmicroorganisms(≥ 99%)fromtheenvironmentremainuncultivatedinthelaboratory,andtheseregulateecosystemprocessesandevenaffectourlives(Kaeberleinetal.2002;Suetal.2012).Therefore,characterizingmicrobialcommunitydiversitycanhavetheoreticalandpracticalsignificanceinunderstandingtherelationshipsbetweenmicroorganismandtheirhosts,treatmentofenvironmentalpollutants,utilizationofmicrobialresources,andhumanmedicalhealth.16SrRNAgenesequencesareconsideredthepredominantandmostreliableapproachforstudyingmicrobialstructuresinhumans,thegutsofanimals,naturalhabitats,andtherhizospheresofplants(Beckersetal.2016).16SrRNAgenevariableregionsandtheprimersetsusedforamplificationof16SrRNAgenesequencescanresultinsignificantlydifferentbacterialcommunityprofiles(Jasonetal.2013;Prosdocimietal.2013).Determiningoptimalprimerpairsfor16SrRNAgenesequencing,therefore,playsanimportantroleinmicroorganismdiversityanalysisofspecialhosts.SeveralresearchershavesuggestedthattheprimersV3FandV4R,coveringtheV3V4amplificationregion,shouldbethepreferredprimersetforfuturestudiestoachieveaccuratebacterialcommunitydiversities,andalargenumberofstudieshavefocusedonthisregion(Castelinoetal.2017;Caietal.2013;Parulekaretal.2017;Takahashietal.2014).Analogously,forplantsamples,notincludingsubterraneanplantpartsbutonlyabove-groundpartssuchasshoot-tip,leaves,etc.,theV3V4regionshasbeenusedpredominantlyforidentifying16SrRNAgenesequences(ThomasandSekhar2017;Tyxetal.2016;Wangetal.2018).Nevertheless,duetothehomologybetweenbacterial16SrRNAgenesandplantchloroplastandmitochondrialDNA,plantDNAmaybecoextractedduringbacterialDNAextractionfromvariousplantsamples,whichcancausecontaminatingsequences(Dyalletal.2004;Raven1970).Usually,ahighabundanceofCyanobacteriaatthephylumlevel,unidentifiedatthegenuslevel,presentinrawsequencingresultsindicatestheoccurrenceofcontaminatingsequencesfromanorganellaroriginsuchmitochondrialand/orchloroplastDNA(Beckersetal.2016).Toexcludetheinterferenceofcontaminatingsequences,acommonmethodistofilteroutthechloroplastDNA,mitochondrialDNA,andotherunknownsequencesduringdataprocessing.However,thismightnotrevealtherealmicrobiomestructuresofsamples.Todate,althoughseveralstudieshavefocusedoncharacterizationofbacterialcommunitiesindifferenttobaccosusing16SrRNAgenesequenceswiththeV3V4andV5V6V7primers(Huangetal.2010;Tyxetal.2016),fewreporthaveshowedoptimizedprimersforbacterialmicrobiomeinvestigationoftobaccos,whilethereareasmanyas10,000chloroplastDNAcopiesintobaccoleafcells(Shaveretal.2006).Microorganismsfromagingflue-curedtobaccos(AFTs)wouldbeusefulinproducingvaluableproducts,treatingwastewater,andcreatingbiofuelsandawiderangeofchemicalsandenzymes.Thereis,therefore,anurgentneedtoinvestigatetheperfectprimersfortobaccobacterialmicrobiomeresearch.PICRUStisanoptimaltooltopredictthefunctionsofamicrobialcommunity’smetagenomefromits16Sprofile.ItpredictsgenespresentedinorganismsandusesexistingannotationsofgenecontentemployingtheKEGG(KyotoEncyclopediaofGenesandGenomes)OrthologyandClustersofOrthologsGroups(Langilleetal.2013).PICRUSthasbeenusedtodeterminethepotentialfunctionsofthemicrobialcommunityinanumberofsamples(Kooetal.2017a,b;Wuetal.2016),whichmightprovideinformationfordevelopmentandutilizationofmicroorganismresources.Somestudiesshowedthatbacteriaintobaccocouldgeneratetoxinsandpro-inflammatorybiomoleculesandgeneratenitrite(Tyxetal.2016).Thus,characterizingbacterialdiversityintobaccosamplesandpredictingthepotentialfunctionsofthebacterialcommunitymightbeagoodwaytoreducethelevelsofcertainharmfulcompoundsintobaccoandimprovethequalityoftobaccoproducts,andotherproducts,suchascoffeeberry,tea,andsauce,etc.Inthisstudy,weusedfour16SrRNAgeneprimersetstoinvestigatethebacteriaintobaccosamplestoselecttheoptimalprimersetsforcharacterizingbacterialdiversityintobaccosamples.Moreover,thebacterialcommunityfunctionwasfurtherpredictedusingPICRUStbasedon16SrRNAgenesequencingforexploringmicroorganismsthatproducednaturalproductsanddegradedtoxiccompounds.Theaimofthisstudywastoprovideavaluablereferencetopeerresearchersworkingonbacterialdiversitydeterminationoftobacco-relatedsamples,andotherplantsamples,using16SrRNAgenesequencingtechnology,andprovideusefulinformationaboutmicroorganismsinAFTsforimprovingthequalityandreducingthelevelsofcertainharmfulcompoundsintobaccosandotherfermentationproductsinfurtherstudies.MaterialsandmethodsExperimentaldesignHigh-throughputsequencingof16SrRNAgeneswasemployedtoinvestigatetheoptimizedprimersetwhichcouldbeusedtoimprovethedeterminationofthemicrobialdiversityofAFTs.Toevaluatebacterial16SrRNAgeneprimerpairsinregardtobacterialdiversitycoverageand/oraccuratetaxonomicassignment,fourprimerspairs(V1F–V3R,V3F–V4R,V4F–V5R,andV5F–V7R)targetingdifferent16SrRNAgeneregionswereemployed.AfterindividualPCR(PolymeraseChainReaction)amplification,atotaloffourfragmentswereobtained,asshowninTable 1.Tominimizenon-specificamplificationandinvestigatethepotentialbiasesoftheIlluminahigh-throughputsequencingtechnology,each16Sprimercarriedauniqueeight-basesequenceforeachsample.Finally,thegeneratedsequencingreadswereanalyzedbysubsequentbioinformatics.Inaddition,PICRUSt(Langilleetal.2013)wasusedtopredictthefunctionalcompositionofthemetagenomeweobtainedfromtheIlluminaMiSeqplatformanalysisofthesamples.Table 1PrimersusedincurrentstudyFullsizetable SamplingandDNAisolationSampleswerecollectedfromHonghecigarettefactory(Qujing,Yunnan)andChuxiongcigarettefactory(Yuanjiang,Yunnan),respectively.Thesesampleswerestoredandtransportedusinganicebathbeforereachingthelaboratoryandthenstoredat− 20 °C.Nospecificpermitswererequiredforthisstudy.AnUltraClean®SoilDNAIsolationKit(MobioInc.,Carlsbad,CA,USA)wasusedtoextracttotalDNAfromthecollectedsamplesindividuallyaccordingtothemanufacturer’smanual.TheDNAextractedfromthreetechnicalreplicatesofeachsamplewaspooledintooneDNAsampletominimizeanypotentialDNAextractionbias.TheconcentrationofeachDNAextractwasdeterminedwith1%agarosegelelectrophoresisandaThermoNanoDrop1000Spectrophotometer(ThermoScientific,Waltham,MA,USA).PCRamplificationandIlluminaMiSeqsequencingAsbeenshowninTable 1,theV1V2V3regionofthebacterial16SrRNAgenewasamplifiedwiththeuniversalprimers8F–533R.ThePCRprogramwasasfollows:95 °Cfor5 min,26cyclesat95 °Cfor45 s,55 °Cfor50 s,and72 °Cfor45 s,withafinalextensionof72 °Cfor10 min.TheV3V4regionwasamplifiedwiththeprimers336F–806RfollowingthePCRprogram95 °Cfor5 min,27cyclesat95 °Cfor45 s,50 °Cfor50 s,and72 °Cfor45 s,withafinalextensionof72 °Cfor10 min.Theprimers515F–909RcoveredtheV4V5amplificationregionsandemployedthefollowingPCRprogram:95 °Cfor5 min,28cyclesat95 °Cfor45 s,55 °Cfor50 s,and72 °Cfor45 s,withafinalextensionof72 °Cfor10 min.TheV5V6V7regionwasamplifiedwiththeprimers799F–1193R.ThePCRprogramwas95 °Cfor5 min,27cyclesat95 °Cfor45 s,55 °Cfor50 s,and72 °Cfor45 s,withafinalextensionof72 °Cfor10 min.Allprimerscontainedan8-nucleotidebarcodesequenceuniquetoeachsample.Allreactionswereperformedintriplicatein50 μLvolumescontaining5μLof10×PyrobestBuffer,4 μLof2.5 mMdNTPs,2 μLofeachprimer(10 μM),0.3μLofPyrobestDNAPolymerase(2.5U/μL,TaKaRa,Japan),and30 ngoftemplateDNA.Ampliconswereextractedfrom2%agarosegelsandpurifiedusinganAxyPrepDNAGelExtractionKit(AxygenBiosciences,UnionCity,CA,USA)accordingtothemanufacturer’sinstructionsandquantifiedusingQuantiFluor™-ST(PromegaCorporation,Madison,WI,USA).Purifiedampliconswerepooledinequimolaramountsandpaired-endsequenced(2 × 300)onanIlluminaMiSeqplatformaccordingtostandardprotocols.BioinformaticanalysisTheextractionofhigh-qualitysequenceswasfirstlyperformedwiththeQIIMEpackage(Gregoryetal.2010).Rawsequenceswereselectedbasedonsequencelength,quality,primer,andtag,andthelow-qualitysequenceswereremoved.TheIlluminaMiseqsequencingdatahavebeendepositedinNCBISequenceReadArchivedatabasewiththeSRAaccessionnumberSRP139912.Theuniquesequencesetwasclassifiedintooperationaltaxonomicunits(OTUs)underthethresholdof97%identityusingUCLUST(Edgar2010).ChimericsequenceswereidentifiedandremovedusingUsearch(version8.0.1623).Thetaxonomyofeach16SrRNAgenesequencewasanalyzedwithUCLUSTagainsttheSilva11916SrRNAgenedatabaseusingaconfidencethresholdof90%.PredictionofthefunctionalcompositionofthemetagenomeusingPICRUStThePICRUStsoftwarepackage(Langilleetal.2013)wasusedtoinferthepotentialgeneticcapabilityandspecificcontributionsofbacterialtaxatothemetagenomesofthetobaccosamples.PICRUStrequiresaphylogenetictreeofmarkergenesthatincludescompletereferencegenomesandthesequencesfromthesamplesunderstudy.Forthisresearch,weusedtheKEGGdatabase(Kanehisa2000)forannotations,andthe16SrRNAgenesequencefromeachofthesegenomeswasobtainedfromtheUnitedStatesDepartmentofEnergyJointGenomicInstitute’sIntegratedMicrobialGenomes(IMG)database(Markowitzetal.2012).ResultsIlluminaMiSeqsequencingTwosampleswerecollectedfromHonghecigarettefactory(Qujing,Yunnan)andChuxiongcigarettefactory(Yuanjiang,Yunnan),andsentforhigh-throughputsequencingofthe16SrRNAgenewithfourselectedbacterial16SrRNAgeneprimerpairs(Table 1).GeneratedreadsandOTUsfromdifferentprimersareshowninTable 2.Rarefactionanalysiswasperformedtoevaluatewhetherthesizeoftheclonelibraryrepresentedthediversityintheoriginalsamples(Zhaoetal.2007).Inourstudy,individualrarefactioncurvesdidreflectthattheprimersgiveanacceptablemeasureofspeciesdiversity(Fig. 1).Thecomparisonoftheprimerpairs(fourregions)revealedthehighestOTUrichnessandobservedspeciesfortheV4V5andV5V6V7regions.ThetwosamplesyieldedsimilarnumbersofOTUswith211OTUs(A1-V4V5),259OTUs(B1-V4V5),230OTUs(A1-V5V6V7),and233OTUs(B1-V5V6V7),andtheseweresimilartothenumberofobservedspecies,209(A1-V4V5),258(B1-V4V5),226(A1-V5V6V7),and232(B1-V5V6V7),respectively(Table 2).Table 2NumbersofsequencesanalyzedforthetwosamplesFullsizetable Fig. 1RarefactionanalysesfortheobservednumberofOTUsfromthetwosamples(fourregions)atageneticdistanceof3%.Rarefactioncurvesforeachregionaredisplayedindifferentcolors,asampleA1andbsampleB1Fullsizeimage ComparisonofdiversityandabundanceofbacterialtaxaSamplesexhibitedasimilarrangeoftaxonomicdiversityatthephylumandgenuslevels.Overall,18and16generafromfivephylaweredetectedwithanabundanceofatleast0.1%inthetwosamples,respectively(Fig. 2).Atthephylumlevel(Fig. 2a,b),detectionofbacterialtaxawasdependentonthechoiceof16Sregions:e.g.,Cyanobacteria(V1V2V3 > V3V4 > V4V5 > V5V6V7),ProteobacteriaandActinobacteria(V5V6V7 > V3V4 > V5V6 > V1V2V3),andFrimicutes(V5V6V7 > V3V4orV5V6 > V1V2V3).Atgenuslevelalso,theresultsshowedthattheabundanceofdetectedbacteriawascloselyrelatedtothecoveredregionsoftheselectedprimerpairs.Forexample,Sphingomonas,Bacillus,Methylobacterium,Lactobacillus,Nocardioides,andPseudomonasshowedhighabundancewithprimers799F–1193R,andextremelylowabundancewithprimers8F–533R.SimilarcaseswerealsoobservedforRhizobium,Aureimonas,Prevotella_9,andothers.PantoeaandRhodanobacterdisplayedhighabundancewithprimers799F–1193R,butwerenotdetectionwiththeotherprimers.Unidentifiedtaxashowedhighabundancewith8F–533R,butlowabundancewithprimers515F–909Rand799F–1193R.Ofnote,CercisgiganteaandNicotianatabacum(commontobacco)weretwoplantspeciesdetected.Cercisgiganteawasdetectedonlywithprimers515F–909Rwithhighabundance,andN.tabacum(commontobacco)wasnotdetectedwithprimers799F–1193R.Figure 3furtherillustratesthatprimers799F–1193Ramplifiedmoregenerathantheotherprimersfromthetop50genera.Fig. 2Relativesequenceabundanceofbacterialphyla(a,c)andgenera(b,d)associatedwithdifferentprimersets.Phylaandgeneradetectedasextremelylowpercentages(