<%BANNER%>
PRIVATE ITEM
Digitization of this item is currently in progress.
Fortunella margarita Transcriptional Reprogramming Triggered by Xanthomonas citri subsp. citri
CITATION DOWNLOADS PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/AA00008948/00001
 Material Information
Title: Fortunella margarita Transcriptional Reprogramming Triggered by Xanthomonas citri subsp. citri
Series Title: BMC Plant Biology
Physical Description: Archival
Language: English
Creator: Khalaf, Abeer A.
Gmitter, Frederick G. Jr.
Conesa, Ana
Dopazo, Joaquin
Moore, Gloria A.
Publisher: BioMed Central
Publication Date: 2011
 Notes
Abstract: Background: Citrus canker disease caused by the bacterial pathogen Xanthomonas citri subsp. citri (Xcc) has become endemic in areas where high temperature, rain, humidity, and windy conditions provide a favourable environment for the dissemination of the bacterium. Xcc is pathogenic on many commercial citrus varieties but appears to elicit an incompatible reaction on the citrus relative Fortunella margarita Swing (kumquat), in the form of a very distinct delayed necrotic response. We have developed subtractive libraries enriched in sequences expressed in kumquat leaves during both early and late stages of the disease. The isolated differentially expressed transcripts were subsequently sequenced. Our results demonstrate how the use of microarray expression profiling can help assign roles to previously uncharacterized genes and elucidate plant pathogenesis-response related mechanisms. This can be considered to be a case study in a citrus relative where high throughput technologies were utilized to understand defence mechanisms in Fortunella and citrus at the molecular level. Results: cDNAs from sequenced kumquat libraries (ESTs) made from subtracted RNA populations, healthy vs. infected, were used to make this microarray. Of 2054 selected genes on a customized array, 317 were differentially expressed (P < 0.05) in Xcc challenged kumquat plants compared to mock-inoculated ones. This study identified components of the incompatible interaction such as reactive oxygen species (ROS) and programmed cell death (PCD). Common defence mechanisms and a number of resistance genes were also identified. In addition, there were a considerable number of differentially regulated genes that had no homologues in the databases. This could be an indication of either a specialized set of genes employed by kumquat in response to canker disease or new defence mechanisms in citrus. Conclusion: Functional categorization of kumquat Xcc-responsive genes revealed an enhanced defence-related metabolism as well as a number of resistant response-specific genes in the kumquat transcriptome in response to Xcc inoculation. Gene expression profile(s) were analyzed to assemble a comprehensive and inclusive image of the molecular interaction in the kumquat/Xcc system. This was done in order to elucidate molecular mechanisms associated with the development of the hypersensitive response phenotype in kumquat leaves. These data will be used to perform comparisons among citrus species to evaluate means to enhance the host immune responses against bacterial diseases.
General Note: Publication of this article was funded in part by the University of Florida Open-Access publishing Fund. In addition, requestors receiving funding through the UFOAP project are expected to submit a post-review, final draft of the article to UF's institutional repository, IR@UF, (www.uflib.ufl.edu/ufir) at the time of funding. The Institutional Repository at the University of Florida (IR@UF) is the digital archive for the intellectual output of the University of Florida community, with research, news, outreach and educational materials
 Record Information
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution.
Resource Identifier: doi - 10.1186-1471-2229-11-159
System ID: AA00008948:00001

Downloads

This item is only available as the following downloads:

( XLSX )

( XLSX )

( PDF )

( XLSX )

( XLSX )

( XLSX )


Full Text

PAGE 1

RESEARCHARTICLE OpenAccessFortunellamargarita Transcriptional ReprogrammingTriggeredby Xanthomonas citri subsp. citriAbeerAKhalaf1,2*,FrederickGGmitterJr2,AnaConesa3,JoaquinDopazo3andGloriaAMoore1AbstractBackground: Citruscankerdiseasecausedbythebacterialpathogen Xanthomonascitri subsp. citri( Xcc) has becomeendemicinareaswherehightemperature,rain,humidity,andwindyconditionsprovideafavourable environmentforthedisseminationofthebacterium.Xccispathogeniconmanycommercialcitrusvarietiesbut appearstoelicitanincompatiblereactiononthecitrusrelative Fortunellamargarita Swing(kumquat),intheform ofaverydistinctdelayednecroticresponse.Wehavedevelopedsubtractivelibrariesenrichedinsequences expressedinkumquatleavesduringbothearlyandlatestagesofthedisease.Theisolateddifferentiallyexpressed transcriptsweresubsequentlysequenced.Ourresultsdemonstratehowtheuseofmicroarrayexpressionprofiling canhelpassignrolestopreviouslyuncharacterizedgenesandelucidateplantpathogenesis-responserelated mechanisms.Thiscanbeconsideredtobeacasestudyinacitrusrelativewherehighthroughputtechnologies wereutilizedtounderstanddefencemechanismsin Fortunella andcitrusatthemolecularlevel. Results:cDNAsfromsequencedkumquatlibraries(ESTs)madefromsubtractedRNApopulations,healthy vs.infected,wereusedtomakethismicroarray .Of2054selectedgenesonacustomizedarray,317were differentiallyexpressed(P<0.05)inXccchallengedkumquatplantscomparedtomock-inoculatedones.Thisstudy identifiedcomponentsoftheincompatibleinteractionsuchasreactiveoxygenspecies(ROS)andprogrammedcell death(PCD).Commondefencemechanismsandanumberofresistancegeneswerealsoidentified.Inaddition, therewereaconsiderablenumberofdifferentiallyregulatedgenesthathadnohomologuesinthedatabases.This couldbeanindicationofeitheraspecializedsetofgenesemployedbykumquatinresponsetocankerdiseaseor newdefencemechanismsincitrus. Conclusion: FunctionalcategorizationofkumquatXcc-respon sivegenesrevealedanenh anceddefence-related metabolismaswellasanumberofresista ntresponse-specif icgenesinthekumquattranscriptomeinresponsetoXcc inoculation.Geneexpressionprofile(s)wereanalyzedtoa ssembleacomprehensiveandinclusiveimageofthemolecular interactioninthekumquat/Xccsystem.Thiswasdoneinordertoelucidatemolecularmechanismsassociatedwiththe developmentofthehypersensitivere sponsephenotypeinkumquatleaves.Thesedatawillbeusedtoperform comparisonsamongcitrusspeciestoevaluatemeanstoenhancethehostimmuneresponsesagainstbacterialdiseases.BackgroundCitrustreesaresusceptibletoanumberofdiseaseswith differentdegreesofeconomicimpact.Oneofthemost severeintermsofeconomiclossesiscitruscankerdisease(sometimesreferredtoasAsiaticcitruscanker) causedby Xanthomonascitri subsp. citri ,(synonym, Xanthomonasaxonopodis pv. citri strainA;Xac-A).Xcc isabiotrophicbacterialphytopathogenthatbelongsto thegenus Xanthomonas ofthe a -subdivisionvwithin Proteobacteria.Susceptibilitytocitruscankerdisease variesamongcitrustypesandrelatives,butmostofthe commerciallygrowncitrustypesaresusceptiblehoststo Xcc[1].Diseasesymptomsincludecankerlesionson thegreenaerialpartsoftheplantaswellasfruit;infectionscanresultinbothfoliarandfruitabscission, therebydecreasingtheproductivityofaffectedtrees.In *Correspondence:abeera@ufl.edu1PlantMolecularandCellularBiologyProgram(PMCB),HorticulturalSciences Department,UniversityofFlorida,Gainesville,Fl.,32611,USA FulllistofauthorinformationisavailableattheendofthearticleKhalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 2011Khalafetal;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommons AttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,andreproductionin anymedium,providedtheoriginalworkisproperlycited.

PAGE 2

additiontherecanbereducedprofitabilityasaresultof blemishedfruitthatcanbeharvestedbutnotsoldin thefreshmarket. Plantshaveevolvedmultip ledefencemechanismsto survivepathogenattacks[2].Thefirstbranchofthe indispensableplantinnateimmunitysystemistriggered bypathogen-associatedmolecularpatterns(PAMPs) suchasthelipopolysacchar ides(LPSs),peptidoglycan andbacterialflagellin,aswellasthechitinandglucan fromfungi.Thesecondbranchutilizesthenucleotidebindingsite-leucine-richrepeat(NBS-LRR)encodedby R(resistance)genesnamedtheeffector-triggeredimmunity(ETI)[3].The Xanthomonas spp.phytobacterial pathogenshaveevolveduniquepathogenesismechanismstoavoidhostrecognitionandsuppresshost defences[4,5].Bacterialeffectorproteinsaredelivered viathebacterialtypeIIIsecretionsystem(TTSS)into theplantcelltoevaderecognitionbythedifferentplant surveillancesystems[6].Theseeffectorsingeneralcontributetohostresistanceo rsusceptibilityaswellasto modifyinghostresponses.Afundamentalelementofthe ETIinresistantplantsisalocalizedcellcollapseora hypersensitiveresponse(HR)atinfectionsitesinan attempttorestrictthegrowthofthepathogen[7,8]. Thisisacommonfeatureofdiseaseresistantresponses inincompatibleplant-pathogen,andoccasionallysome non-host,interactions[9,10].Someofthe Xanthomonas spp.effectorproteins,forinstancePthA/AvrBs3,are essentialtoelicitcitruscankersymptomsandif expressedbyitselfinsidehostcells, pthA issufficientto causesymptomsofcitruscankerdisease[11-15].Inthe meantimehowever,otherrecentstudiesshowthatother typesofproteinsareinjectedthroughtheXccTTSS anddonotnecessarilyalterthephysiologicalandtranscriptionalresponsestothepathogenincitrus [8,10,16,17]. Whilecertaingenesinvolvedinsystemicacquired resistance(SAR)havebeencharacterizedandusedas markersforstudyingplantdefencemechanisms[18], crosstalkbetweensignalsandhormonepathwayshas alsobeenproposed[19-21].Consequently,plantresistanceiscorrelatedwiththeactivationofacomplexnetworkofdefencepathwaysandtheresponseofthehost planttoamicrobialassaultisthereforeexpectedto resultindrasticchangesinthepatternsofgeneexpressionthroughouttheplant[22,23]. Kumquats(Fortunella spp.),closerelativestocitrus species,arereportedtohavehighlevelsoffieldresistancetocitruscanker[1].Previously,wehaveshown asharplycontrastingphenotypeingrapefruitand kumquatwhenbothplantswerechallengedwitha highconcentrationofXccr(OD600nm=0.3)[24]. Grapefruit(Citrusparadisi Macf.cv.Duncan), consideredtobehighlysusceptibletothebacterium, showedthecharacteristicsequenceofcankerlesion development.Initiallylesionsappearedaswatersoaking,followedbythedevelopmentofaraisedcorky form;eachsuchlesionisareservoirofnewbacterial inoculum.Bacterialexudateswerevisiblebetween10 and21dayspost-inoculation.Incontrast,PCDwas observedinkumquatleavesintheformofaHR3-5 daysafterinoculationwiththecanker-causingbacterium.Onlynecroticlesionswereobservedandthebacterialpopulationovertimewasshowntohavean avirulent incompatiblegrowthpatternwherebacterial multiplicationceaseduponthedevelopmentofnecrosis[8,25]. Newtoolshavebeendevelopedinrecentyears throughadvancesingenomics,proteomics,andbioinformaticsthathaveparticularutilityforexamining pathogen:hostinteractioncomplexities[22,26-28].The purposeofthisstudywastoexaminesimultaneous changesinexpressionprofilesforgenesdifferentially expressedintheearlystages(6-72hpi)ofcitruscanker infectioninkumquat,particularlythosepreviously implicatedinPCD-relatedresponsessuchasHR.ResultsandDiscussionInthisstudy,identificationofdifferentiallyexpressed kumquatgenesduringitsinteractionwithXccwaspursuedinanattempttounravelthenatureoftheresistancemechanism(s)employedbytheplant.Previously, kumquatsuppressionsubtractivehybridization(SSH) cDNAlibrarieswereconstru ctedfromXcc-inoculated vs.mockinoculatedleaves[24].SinceSSHallowsdifferentialamplificationofraretargetsequencesduetothe eliminationofmoreabundanthouse-keepingcDNA transcriptsfoundincommonfrombothsamples,the techniquehasthepotentialofuncoveringpertinent cDNAsequences.Subtractionwasdoneinbothdirections,forward(inoculated-mock)andreverse(mockinoculated)andtheresultingcDNAsweresubsequently sequenced.Preliminaryscreeningmacroarrayswere usedtoconfirmenrichmentofthesubtractedlibraries withdifferentiallyexpressedgenes(datanotshown).MicroarrayexperimentaldesignKumquatmicroarraychiphybridizationdatawere assessedforoverallsignali ntensityandconsistencyof theexpressionratiooveralltimepoints,whichresulted intheexclusionofchipswithinconsistentresults. Figure1isascatterplotshowingM-valuesfromtwo differentbiologicalreplicate-hybridizationswith Xcc-inducibletargets(Cy5-labeled)andmockinoculated non-infectedtargets(Cy3-labeled)confirminghighdata consistencylevels(R2=0.921).Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page2of17

PAGE 3

Functionalannotationandanoverviewofglobalgene expressionTheB2GOprogram[29]wasusedtoassignGO(Gene Ontology)termsforhitsobtainedthrougheBLAST homologysearchesinNCBI.AgeneralviewofthesimilarityofthequerysetwiththeNCBIdatabase,thedistributionofthecutoffforthee-valueaswellasthe distributionofspecieswithsimilarsequencesareshown inAdditionalfiles1,2,and3.TheGOannotationscore isconsideredtobemoreintuitivethanregularblastevaluessinceGOannotationiscarriedoutbyapplying anannotationrule(AR)ontheontologyterms.Additionally,querysequencede scriptionsareobtainedby applyingalanguageprocessingalgorithmthatextracts informativenamesandavoidslow-contenttermssuch as hypotheticalprotein or expressedprotein .Using Blast2Gosuitedefaultparameters,1042probeswere providedwithGOannotations(Additionalfile4). Approximately25%ofthetranscriptsonthearraydo notshowsimilaritytoproteinspresentinpublicdatabases.Someofthesecouldrepresentexclusivegenesof thecitrusorkumquatlineages,butafractionofthese uncharacterizedsequencesmaypossiblyrepresent lowqualityor3 UTRsequences.Similarpercentagesof unknownsequenceshave beenreportedinother small-scaleESTprojects[30-32]andthereforethispatterncanbeconsideredcharacteristicofthisapproach. Sinceacitrusgenomesequenceisnowavailable,future studieswillhaveawealthofcitrusgenomicsequence informationthatcanbeutilizedtoidentifykumquatspecificaswellasnovelcitrusgenesinvolvedindiverse defencemechanisms[28]. GeneontologyanalysisprovidedanextremelyinformativesnapshotoftheXcc/kumquatinteraction.The hierarchicalstructurefor thegeneontologyofagroup ofsequencescanbevisualizedasatreebymeansof directedacyclicgraphs(DAG)[33].Forinstance,the molecularfunctionsofthenetworkimplicatedinthe kumquatresponsetoXccinfectionisillustratedinthe DAGpresentedinFigure2.Thegraphdemonstratesa treecontrolledbytheSeqfilterthatorganizesthe numberofnodestobedisplayed.Seqisthenumberof differentsequencesannotatedatthechildGOterm.On thewhole,thebiologicalmeaningfordifferent sequencesinthedatasetwasbestillustratedintermsof threeGOgenecategories;thebiologicalprocesses (Figure3)underlyingmolec ularfunctions(Additional file5),andthecellularcompartmentswhereproteins werelocalized(Additionalfile6).KumquattranscriptionalchangesinresponsetoXcc infectionAnimportantaspectofthedatawasthat,formany genes,transcriptabundancevariedovertimepoints,and anumberofgeneswereonlyup-ordown-regulatedat oneortwotimepoints(Figure4).Twoapproacheswere usedtoidentifypatternsofgeneexpression.First;the ASCA-geneanalysismethodo logyrevealedthatmostof thetotalvariabilityinthedatawasrelatedtotime-associatedchanges[34].AccordingtoASCA,289probes wereselectedasdifferentiallyexpressed,172ofwhich wereatstatisticallysignificantlevels(Additionalfile7). Moreover,thetime-associatedvariationcouldbedivided intotwomainvariabilitypatterns.Onepattern (accountingfor20%ofthevariation)representedgenes whoseexpressionlevelschangedsignificantlyat24hpi fromtheirlevelsat6hpiandthenrecoveredtovalues similartothestartingvalues(ortoevengreatervalues intheoppositedirection)at72hpi.However,themajor pattern(80%ofthetime-asso ciatedvariation)indicated astronggeneexpressionchangebetween6and24hpi followedbypreservationofexpressionlevelsat72hpi. Thisindicatesthatthestrongestresponsetoinfection occurredat6to24hpi,andthemajorityofgenesmaintainedtheirchangeforupto3dayswithasmallerpercentagerevertingtoinitialvalues. Thesecondapproach,maSigProanalysis,indicated that317probesweredifferentiallyexpressedthroughout time,(adjustedpvalue<0.05andR2ofthemodelfit >=0.6;Additionalfile8).Theresultsofbothapproaches werecombinedinto433probesthatwerethenfiltered usingmorestringentconditionstoprovideaunique Figure1 ScatterplotanalysisoftheM-valuesfromtwo microarrayhybridizationsusingRNAsamplesfromtwo independentkumquatplantsinoculatedwith5108cfu/ml Xcc.Eachspotrepresentsthenormalizedhybridizationsignal intensityforeachtranscriptonthemicroarray.RNAsamplesfrom non-inoculatedandinoculatedleaftissuewerelabeledwithCy3 andCy5,respectively. (h6_1) :6hourspostinoculationhybridization resultsfromslide#1hybridizedtoplantA-RNAsamplesvs. (h6_2) : 6hourspostinoculationhybridizationresultsfromslide#2 hybridizedtoplantB-RNAsamples. Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page3of17

PAGE 4

Figure3 Amultilevelpiechartshowingthedistributionofprobesonthechip .Biologicalprocesseswithinallofthelowestnodeswith thegivennumberofsequencesorscorevalueplotjointlywithane-valuecutoff(e-06). Figure2 ADirectedAcyclicGraph(DAG)visualizingthehierarchical structureoftheGeneOntology(GO)ininoculatedkumquat leaves .Childrenthatrepresentamorespecificinstanceofaparenttermhave isa relationshiptotheparent.Thedarkerthecolorofthenode themorenumberofBlasthitsandthehigherannotationscoreithas.Allnodescontainthehitannotationscoresinnumbers. Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page4of17

PAGE 5

result.Theunionratherthantheintersectionofthetwo approacheswastakenbecausethetwomethodsreveal differentaspectsofthedataandarethuscomplementary.TheASCA-genemethodologyfocusesonshared geneexpressionchangestofindimportantgenes,while maSigProtreatsgenesindependentlyandevaluatessignificanttimedependent-changes.AlthoughASCA-gene methodologymaymisssomegeneswhoseexpression patternisrarebutsignificant,thesewillbecapturedby thegene-wisemaSigProapproach.Alternatively,maSigProcanmissgeneswithlesspronouncedchanges, whichcanberecoveredbyASCA-geneiftheirprofileis abundantwithinthedataset.Theuseofboth approachestogetherresultedintheidentificationof437 differentiallyexpressedgenes312ofwhichwithacceptablep-valuesthatcouldbe dividedinto4clusters accordingtotheirexpressionpatterns(Figure5,Additionalfile9).Thecriterionforthisdivisionisasfollows. FromtheASCAanalysisweobtainedthemainpatterns ofvariation:ClusterpatternAindicatesastrongchange inexpressionbetween6hpiand24hpi,whichisthen maintainedattime72hpi.ClusterBpatterniscomprisedofgenesdifferentiallyexpressedat6hpiascomparedtoeither24hpior72hpi.Foreachpattern,the correlationofthemeanvalueofeachgeneateachtime pointwiththeprofilesindicatedbyASCA-genewascalculated;subsequentlygenesweredividedinto4clusters dependingonwhetherexpressionlevelschangedin positiveornegativedirections.Inthisanalysis,genes cannotbeclassifiedsimplyasinducedorrepressed, becausethisdependsonthetimepointsconsidered;for example,genesinclusterpatternCarerepressedat24 hpiandtheninducedat72hpi.Functionalcategorizationoftranscriptsunderlineskey elementsinkumquatresponsetoXccinfectionBasedontheassumptionthatalteredgeneprofilesduringplant-microbeinteract ionscanbecorrelatedwith symptoms,geneontologyandannotation,webelieve thatXccrepresentsatypicalexampleofhowthebacterialpathogencanmanipulatethehostsystemsinits favouraselucidatedpreviouslyindifferentstudies [3,4,11,35,36].Informationonallofthespecifictranscriptsdiscussedinthesubsequentparagraphsisgiven inTable1.Cernadas etal .inoculated Pera sweet orangewitheitherXcc,whichcausestypicalcanker symptomsonthiscitrustype,or Xanthomonasaxonopodis pv. aurantifolii pathotypeC(Xaa),whichonlyproducessymptomsonMexicanlime,followedbya detailedtranscriptionala nalysisforthesweetorange plants[36].Althoughtheanalysesdoneinthatstudy cannotbedirectlycomparedwithourstudybecauseof differencesinmethodology,somegeneralizationsare notedbelow. Thedistributionoffunctio nswithinthesignificantly expressedgenesinXccinfectedkumquatsindicatesthat thehighestnumberoftranscripts(~30%)wasassociated withresponsetostress,electrontransport,and/oroxidativestress(asshowninFigure3),anindicationofan earlyregulatorychangesintheplantimmunesystemby Xcc.Earlierstudies,suchasthatofCernadas etal. ,have cometothesameperception[ 36].EachidentifiedclusterwassubjectedtofunctionalanalysisbyeitherstudyingthedistributionofGOtermsorperforming enrichmentanalysistoseeiftherewerefunctionalcategoriesthatweresignificantlyrepresented.Atotalof137 genes,whichmakesupmorethan30%ofthegenesthat weresignificantlyexpressed,weredown-regulatedinthe intervalbetween6hpiand24hpi,.Mostofthemwere groupedinClustersAandC(Figure6).Theexpression levelsofthegenesinbothoftheseclustersreacheda minimalexpressionlevelat24hpifollowedbyeithera minor(ClusterA)ormajor(ClusterC)recoveryby72 hpi(Table1).Forinstance,theexpressionofthethioredoxinfgenehomologue(KLLFI3-F09)thatbelongsto clusterAreacheditsmaximumlevelofexpression(+1.8 fold)by72hpiafterslightdecreaseat24hpi.Thelipoxygenasegenehomologue(KSLFII1-F07)thatbelongsto clusterCwas1.5folddown-regulatedat6hpifollowed bya3foldincreaseinexpressionwhencomparedto theitsexpressionlevelat6hpisample.GenesinCluster AandCwerefrequentlyrelatedtooxidativestress Figure4 Venndiagramdemonstratingthenumberofupregulatedgenes(numbersinred)vsthedown-regulated (numbersingreen)subsequenttoXccinoculation .Resultswere basedonthemeaninductionsofsixexperimentalreplicates.Genes withM-values>0.5(1.5fold)wereconsideredup-regulatedwhileMvalues<0wereconsidereddown-regulated. Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page5of17

PAGE 6

response.Mostoftheactivityforgenesintheseclusters islocatedinthemitochondria,thecellmembraneand thechloroplast(Figure6). ClusterBwasthelargest clusterandincluded235geneswithup-regulated expressionlevelsbetween6hpito24hpifollowedby sustainedexpressionuntil72hpi(Figure5,Figure7A). ClusterDcontained61membersthathadalowsteady expressionupto24hpi,andweresubsequentlyupregulated(Figure5,Figure7B).Th isclusterincludesgenes, suchastheglycosyltransferase-likegene(KSLFI7-F12), thatmediatethetransferofglycosylresiduesfromactivatednucleotidesugarstoacceptormolecules(aglycones),akeymechanismindeterminingthediversity, activityandchemicalcomplexityofplantnatural products.Inplants,UGTs(uridinediphosphatesugar glycosyltransferases)generallyuseUDP-glucoseand occasionallyUDP-xyloseforglucosylationofphenylpropanoidaglycones.Albrech tandBowman[37]proposed usingUGTsandotherglycosyltransferasesasprospectivegeneticengineeringca ndidatesduetotheirimportantroleinresistanceandtolerancetocitrustristeza virus(CTV)aswellascitrushuanglongbing(HLB)in trifoliateoranges(Poncirustrifoliata L.Raf.).Phenolics aremainlysynthesizedinplantsviathephenylpropanoid pathwayandareincorporatedintomanyimportant compoundsincludingplanthormones,secondarymetabolitesinvolvedinstress,defenceresponses,andxenobioticssuchasherbicides[38].Inaddition, Figure5 Clusterpatterns .Theoverallaveragegeneexpressionprofilesforgenesfromdifferentfunctionalclustersateachtime-point. Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page6of17

PAGE 7

phenylpropanoidpathwayintermediates,forexamplepcoumaricacid,caffeicacid,ferulicacidandsinapicacid, andpathwayderivatives,includingflavonoidaglycones andglycosides,exhibitantimicrobialactivity[39,40].KumquattranscriptionalchangesinresponsetoXcc infection ROSvsROSscavengingInordertomaintainhomeostasisandovercomethe damagingeffectsofROS(reactiveoxygenspecies),abalancebetweenSODs(superoxidedismutase)andthedifferentH2O2-scavengingenzymesisconsideredtobe criticalindeterminingthelevelsofO2 -andH2O2in plantcells[41,42].Accordingly,thereisaconstant interplaybetweentheantioxidantstateandprocesses generatingROS.ROSareproducedinchloroplasts,peroxisomes,andmitochondriainresponsetobioticas wellasabioticstresses[43,44].Accordingly,theexpressionofdifferentenzymesthatproduceROSwereevidentlystringentlycontrolledandcoordinatedduringthe kumquat/Xccinteraction.Forinstance,whileformate dehydrogenase(FDH;KLLRI2-G05),amitochondrial NADdependentenzyme,was1.5foldupregulatedby6 hpi,amineoxidase(KSLFI3-G05)thatcontributestothe synthesisofH2O2andsecondarymetaboliteswasdownregulatedby1.5and1.6foldat6and24hpirespectively inresponsetoXccchallenge(Table1).Concurrently, Xcc-inoculatedkumquatplantsoverexpressedgenes relatedtoROSscavengingtorestrictdamagetothe inoculatedpartsoftheplant,inthiscasetheleaves.For instance,CuZnSOD(KLLF13-A03)expressioninkumquatswasincreased~1.5foldat6hpiandwasstabilized at24hpiand72dpi(Table1).Thesamephenomenon wasobservedpreviouslyintomatoinfectedwith Botrytis cinerea ,asignofincreasedROSproductionbythehost aspartofthedefenceresponsetoinfection[45]. Furthermore,whiletheexpressionofsomeofthegenes linkedtoproteaseinhibitorsandendopeptidaseactivities suchasproteaseinhibitorhomologue(KLLFI2-D02)was suppressedbythebacteria,otherserine-typeendopeptidaseinhibitorssuchasanATP-dependentionprotease (KSLFIV1-H05)was>2foldup-regulatedasearlyas6 hpisubsequenttoXccinoculation.Inthesamecontext, theredoxcouplingascorbat e-glutathionecycle,known toberesponsibleforperoxidedetoxification[46],was repressedby6hpiinthekumquatdataset;examples includedehydroascorbatereductase(KSLRI1-F02)and glutathioneperoxidase(KL LFII3-G07).Ascorbateand glutathionearenon-enzymaticantioxidantmolecules thathavearoleinothercycles,includingthosethat synthesizeandinsomecasesmodulateflavonoids,alkaloids,phenoliccompounds, a -tocopherolandcarotenoids,allofwhichcontributeinscavengingROS[47]. Dehydroascorbatereductas eactivityisindispensible whentheascorbateperoxidase(APX)levelsarehigher thannormalundercertainco nditionstoensurepreservationofthereducedformofascorbate.Bothproteins inadditiontocertaintypesoftrypsininhibitorsmight alsocatalyzeaplantresponse[48].Asimilarstudyto investigatetheXanthomonas-grapefruitcompatible interactionmightpresentaplatformtocomparegene expressionprofilesofsomegenesofinterestinboth plants. Accumulatingevidenceindicatesthatproteinubiquitinationanddegradation,laststepsinproteinturnover, areinvolvedinplantdefenceresponses.Anumberof recentstudieshaveinvestigatedapossibleroleofUboxE3ubiquitinligasesinPTI(PAMPS-triggered immunity),ETI(effector-triggeredimmunity),aswell asplantcelldeathanddefence[49,50].Inthepresent study,6ubiquitinationpathway-relatedgenes,for exampleubiquitin-conjugatingenzymeucb7(CSL1A02),wereisolatedinthekumquatforwardsubtracted libraries;moreinvestigationoftheirexpressionlevels afterinfectionwillfollow.Otherinducedgenesthat areinvolvedintheproteolysisprocessarepresentin clustersA,BandC.GenesinvolvedinphotosynthesisAdistinctdown-regulationintheexpressionofribulose1,5-bisphosphatecarboxylase/oxygenaseat6hpi,followedbyanincreaseinexpressionthatreaches maximumexpressionat24hpi,wasobservedinthe microarraydataset(Rubiscosmallandlargesubunits; forexampleKLLFIII3-G09andKSLRII2-F01)(Table1). Rubisco,themostabundantproteininleaves,isthe mainsourceofenergyproductioninplantcells.A decreaseinphotosynthesiswaspreviouslyshownin Arabidopsis leavesasearlyas3hafterchallengewith the P.syringae avirulentstrain,whileafter48htherate ofphotosynthesiswaslowerwiththevirulentstrain [51].Mostofthephotosyn theticmachineryinchallengedkumquatleaveswasrepressedat6hpi,including chlorophyllA/Bbindingprotein(KLLFIII3-A06and KLLFIII3-E08).Threephotosynthesis-relatedgeneswere differentiallydown-regulatedduringthefirst24hours. InPto-mediatedresistance,30photosynthesis-related genesand12genesencodingchloroplast-associatedproteinsweresuppressed[52].Theseresultsshowthat plantsreducephotosyntheticpotentialtoinduceHRfollowingpathogenattack.Further,Quirinoetal.[53]suggestedthatHRandsenescencearetwoprogramsthat involvebiochemicalsimilaritiesaswellasanoverlap. Theresearchreinforcedtheideaofaconnection betweendefenceresponseandsenescence.Evidently,the down-regulationofgenesinvolvedinphotosynthesis duringtheXcc/kumquatinteractionrepresentsacost fortheplantfitnesswhereenergyresourceswereredirectedtodefenceresponse.Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page7of17

PAGE 8

Table1FunctionalcategorizationofcDNAsidentifiedfrommicroarrayanalysis.PutativeFunction ClusterP-ValueID M-Value 6hpi 24hpi72hpi OXIDATIVEBURST/STRESS,APOPTOSIS *Thioredoxinf A0.093KLLFI3-F09 +0.093+0.037+0.913 *Peroxidase C 0.046KLLRI2-F09 -2.370-0.280-0.380 *ClassIIIperoxidase B 0.018KSLFI1-H03 +1.643+0.323+1.053 Glycosyltransferase-like protein D 0.004KSLFI7-F12 -2.237+0.103-0.243 *Glutathioneperoxidase D 0.024KSLFI1-B02 -2.143-0.027-0.287 *Formatedehydrogenase B 0.041KLLRI2-G05 +0.930-0.447+0.173 *CuZn-superoxidedismutase B 0.035KLLFI3-A03 +0.523-0.267-0.086 *Proteaseinhibitor B 0.015KLLFI2-D02 -1.987+0.653+0.510 *lonproteasehomologue D 0.074KSLFIV1-H05 +1.220-0.607+0.240 *Dehydroascorbate reductase C 0.012KSLR1-F02 -2.033-0.040-0.223 *Glutathioneperoxidase C 0.021KLLFII3-G07 0.267 -0.237-0.247 *Ubiquitin-conjugatingenzymeubc7 B 0.015CSL1-A02 -0.163-0.390+0.320 Catalase(EC1.11.1.6)CAT-2 C 0.041KLLFI1-F11 -1.343-0.207-0.253 *Amineoxidase A 0.035KSLFI3-G05 -0.553-0.720+0.493 Hydroperoxidelyase B 0.012CSL2F2-A01 -2.963-0.427-0.167 Benzoicacidsalicylicacidmethyltransferase C 0.034KLLRI2-C03 +0.103-0.233-0.263 *1-aminocyclopropane-1carboxylateoxidase B 0.008KSLFI7-H12 -1.993+0.183-0.310 PHOTOSYNTHESIS Chlorophyllabbindingprotein B 0.006KLLFIII3-A06 -2.057+0.217+0.563 ChloroplastphotosystemII22kda B 0.15KLLFIII3-E08 -0.717+0.433-0.050 DEFENCE *Pathogenesis-relatedprotein1a A 0.054KSLFI3-H10 -1.803-0.077-0.407 *SABP2 B 0.008KLLRI2-G01 +0.787-0.093-0.317 *Beta-1,3-glucanase B 0.024KSLFII1-C07 -1.91 -0.760-0.167 Phenylalanine-ammonialyase D 0.016KSLFI4-F04 -2.58 -0.013-0.147 Pathogenesis-relatedprotein4-1 A 0.032KLLFII2-G01 +0.07-0.210-0.353 *ClassIVchitinase C 0.626KLLRI2-D05 +0.01 -0.3-0.28 *NDR1homologue C 0.132KLLFII2-E03 -0.91 -0.74-0.04 *Trypsininhibitor A 0.011KSLFIII1-H12 -0.40 +1.05+0.303 *Trypsininhibitor B 0.008KLLFIII3-F03 -3.20 +0.197-0.490 *HSR203J-likeprotein C 0.003KSLFI3-C10 +0.073-0.09-0.303 *DND1[Arabidopsis thaliana] D 0.074KLLRI2-B05 -0.203 -0.580 +0.310 *Baxinhibitor-1 C 0.01KLLFIII2-E02 +0.233+0.083+0.06 *Latex-abundant(caspase -like) B 0.090KLLRI2-A12 +0.230+0.173+0.09 *Zincfingerprotein B 0.017KSLFI6-C10 +2.387+0.000+0.663M-valueisthebasetwologarithmoftheratiobetweenthebackground-subtractedforegroundintensitymeasuredintheredandthegreenchannels. TheseESTswereidentifiedashavingacy5cy3ratio>1.5forfouroutofsixspotsonthemicroarrays. -PutativefunctiondeterminedwiththeGeneontologysequencedescription -Cluster:TheclustertowhichtheputativegenebelongsaccordingtoBlast2GOfunctionalanalysis. -P-valueassociatedtothestatisticalanalysisfordifferentialexpressionadjustedformultiplecomparisons. -ID:Assignedatselection. -M-Value:Ametricforcomparingagene smRNA-expressionlevelbetweentwodistinctexperimentalconditions;inthiscasemockinoculatedvsXccinoculated. -(-)meansdown-regulatedwhereM-value<0while(+)isup-regulatedwhereM-value>0 Table1.includes(*)genesthatarediscussedinthetext. Whileothersequencesthatmightnotbementionedinthetextbutshowinterestinggeneexpressionprofiles.Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page8of17

PAGE 9

CellwallremodellingXccinoculationofkumquatwasfollowedbythedownregulationofvariousgenesrelatedtocellwallremodelingandrapidexpansionsuchasendoglucanases.The expressionlevelofakumquathomologueofthiswall looseningprotein(KLLFI2C10)wasinsignificant.On theotherhand,genesrelatedtocellwallreorganization, forexamplexyloglucanendotr ansglycosylase/hydrolase (XET, anenzymeinvolvedincellwallelongationand restructuring),weresignificantlyup-regulatedby24hpi (KSLFIII1-H08).In Pera sweetorange,amajor differenceintheresponsetoinoculationofthetwobacterialstrainswasthatXccstronglyupregulatedseveral cellwallremodellingenzymes,whileXaaupregulated genesrelatedtoendoglucanaseinhitorsandligninbiosynthesis.Aphenomenonthatweobservedinkumquat plantsisthedevelopmentofafewminutenecrotic flecksontheleaveswheninoculatedwithlowconcentrationsofthebacterium(Xcc).Neitherleafabscission norwatersoakedlesionswereobservedontheleaves laterunderourconditions.Itisalsoworthmentioning thatalthoughCernadasetal.usedarelativelyhigh Figure6 Blast2GOdirectedacyclicgraphshowing molecularfunction afterXccinoculationamongtr anscriptsrepresentingthe enrichedfunctionalcategories(P<0.25) .(A)ClusterA.(B)ClusterC. Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page9of17

PAGE 10

concentrationofXaa(OD600nm=0.6,~doublewhat weusedforXccwithkumquat)onlypustleswere recordedinsweetorangeinoculatedwithXaathatwere notfollowedbynecrosis[24].Usinglightmicroscopy, wehavepreviouslyshownmesophylcollapseinkumquatleaveswhichwasfollowedbyleafabscission72hrs postinoculationwithXcc.Alternatively,grapefruit mesophylcellsfrominoculatedleavesshowedenlargement(hypertrophy)anddivision(hyperplasia)followed byraisedcircularlesionsthatbecameraisedand developedintowhiteoryellowspongypustules.These pustulesthendarkenedandthickenedintobrowncorky cankerlesions[24].Pustuleformationandhypertrophy werelinkedpreviouslytothePthAeffectorin Nicotiana benthamiana [35].Alternatively,accumulationofthe tomato XTH (xyloglucanendotransglucosylase/hydrolase LeXTH1 )protein6hoursafterattachmentoftheparasitehasprovidedevidenceforaroleof XTH indefence reactionsassociatedwiththeincompatibletomatoCuscuta interactionaswaspresentedinAlbertetal.[54]. Figure7 Blast2GOdirectedacyclicgraphshowing molecularfunction (P<0.25)amongtranscriptsinducedat24hpi .(A)ClusterB.(B) ClusterD. Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page10of17

PAGE 11

ResistancegenesandrelatedproteinsMostofthedifferentiallyexpressedgenesinthekumquat/Xccinteractionhavealsobeenidentifiedinplantinsectinteractions[55],non-hostresistance[56]andRgenemediatedresistance[52],suggestingahighlevelof convergencebetweendifferenttypesofresistance mechanisms.Theexpressionlevelsofsomehomologues tothePR3(endochitinase)(KFII1-B11)and-galactosidase(BG1)(KRI2-D05)geneswerefoundtobemodestlyactivated;2and1foldup-regulatedrespectively6 hpi,althoughbothgenesweredown-regulatedby24hpi asshownbythemicroarrayandtheqRT-PCR(quantitativereal-timePCR)data(Table2,AdditionalFile10). Chitinaseexpressionisaplantdefencestrategytypically usedagainstwallcomponentsoffungiandinsects[57]. AccordingtotheqRT-PCR,thekumquatchitinasegene was>2-foldup-regulatedby6hpi,afterwhichitwas suppressedat24hpi,andthenitsexpressionincreased. Otherstudieshavealsoreportedtheinductionofchitinasesinresponsetobacterialpathogensbuttheirfunctionisnotwellknown[58]. Interestingly,theexpressionlevelsofthekumquat PR1(Pathogenesis-relatedgene1)genehomologue,normallyamarkerofsalicylicacid-inducedsystemic acquiredresistance(SAR)thatisusuallyup-regulated afterpathogeninfections,waslowerintheinfectedsamplesasearlyas6hpicomparedtocontrolmockinoculatedsamples(KLLFII2-A04)(accordingtothe microarrayresultsandnonpublisheddata).Theregion upstreamofthePR1promoter,W-boxsequences,was shownpreviouslytoactasanegativecis-actingelement intheexpressionofdefencerelatedgenes[59].This impliesadifferentbasaldefenceresponseandSARregulationmechanismfromthatofArabidopsisandother dicotyledonousplantsinkumquatafterXccinfection. PRgenesthatwereidentifiedincitrusvaryintheir responsestodifferentpathogensasshowninthisstudy andotherswherevariationinexpressionlevelsofdifferentmembersofthePRgenefamilywasdependenton thenatureofdifferentelicitors[60-62].Thebiological activityofalargemajorityofPRgenesinplantsduring bioticstressisyettoberevealed.Moreinterestingly, accordingtothemicroarrayresults,expressionlevelsof anumberofhomologuestootherdefencerelatedgenes suchastheNDR1gene(KLLFII2-E03)werealso repressed24hpiafterXccinfection. LRRproteinsareknowntobeapartoftheearlysignaltransductioncascadeinvolvedintherecognitionof pathogenAvrproducts[63].Sequencesforanumberof homologuesknowntobepartofdifferenthormonal defencepathways(forinstancetranscriptionfactors, receptorlikeandreceptor-likekinases)werefoundto bedifferentiallyexpressedinkumquatafterXcc inoculation.KeymolecularfeaturesofkumquatPCDAnumberofgeneshomologoustoknownresistant response-specificgeneswereexpressedinthekumquat transcriptomeconcurrentlyfollowingXccinoculation, listedanddiscussedbelow: (i)KSLFI3-C10ishomologousto hsr203J ,acarboxylesterase(CXE)geneimplicatedpreviouslyinthe incompatibleinteractionsbetweentobaccoandthebacterialpathogen Ralstoniasolanacearum .Itspromoteris highly,rapidly,andspecific allyactivatedinresponseto HRinducingbacterialinoculation,doesnotrespondto variousstressconditions,andisstronglydependenton hrp (hypersensitiveresponseandpathogenicity)genesof thepathogenicbacterium[64].Ithasbeenproposed thatitsexpressionshouldbeausefulmarkerforprogrammedcelldeathoccurringinresponsetodiverse pathogens. (ii)KLLRI2-B05shareshomologywithDND1 (DEFENCENODEATH1),whichencodesacyclic nucleotide-gatedionchannelthatallowspassageofCa2 +,K+andothercations.The Arabidopsisthalianadnd1 mutantfailedtoproduceHRcelldeathinresponseto anavirulentpathogeninfection[65]. (iii)Duringprogrammedcelldeathorapoptosiscytochromecisreleasedtothecytoplasmfromtheintermembranespaceofthemitochondrion[66,67].Oncein Table2qRT-PCRanalysisofgenesexpressedinresponsetoXccinoculationa(5108cfu/ml)concentrationofthe MiamistrainX04-59.GeneIDFunction Microarray (FoldChange) Real-timePCR (FoldChange) 6h24h72h0t6h24h72h120h KLLFII2-C05Basicleucinezippertranscriptionfactor -1.04+1.06-1.211-0.53+1.13+1.57+0.26 KLLRI2-D05ClassIchitinase(CHI1) +1.01-1.23-1.211+2.9-0.169.21-4.53 CSL1-D05Adiseaseresistanceleucine-richrepeatprotein+1.06+2.04+1.081+3.09+7.73+1.66+0.19 KLLRI2-H10Receptor-likeserinethreoninekinase +1.62-1.02-1.191+2.65+0.51+0.53+0.47 KLLFII1-B11Aputativebeta-galactosidaseBG1 -1.13-1.24+1.011-1.24-0.10+0.91+0.19 CSL2-A02Amitogen-activated proteinkinase3 +1.02-1.11+1.161+5.66-14.1+1.22+0.58 Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page11of17

PAGE 12

thecytoplasm,itactivatescaspases(cysteineaspartatespecificproteases),killerproteinsthatdismantlethecell [68].Twokeyproteinsknowntobecorecomponentsof theapopticmachineryinanimals,caspaseand Bax-Inhibitor1 genehomologues,wereidentifiedinourdataset. Accordingtoourdata,ahomologuethathascaspase activityKLLRI2_A12wasslightlyup-regulatedby6hpi inkumquatchallengedleaves.Baxisamemberofthe Bcl2familythatplaysaregulatoryrolepreventingapoptosisbyinhibitingadaptersneededfortheactivationof caspases[54].Akumquathomologueofthe Bax-Inhibitor1 gene(KLLFIII2-E06)wasshowntobeslightlyup regulated6hpiinresponsetoXccchallengeaswaspreviouslyshownwith Arabidopsisthaliana BaxInhibitor-1 ( AtBI-1 ),isolatedduringadifferentialscreenofplants challengedwiththephytopathogen Pseudomonassyringae [69].Inthesamecontext, Baxinhibitorhasbeen showntotriggercytochromecreleasefrommitochondriaboth invitro and invivo inanimals. (iv)Endopeptidaseinhibitorsareoftenpartofan inducible,jasmonicacidassociateddefencepathwaythat accumulatesuponwounding,pathogen,orherbivore damageinleaves[26].Theantagonisticinteraction betweenproteasesandendopep tidaseinhibitorsisconsideredtobeacelldeathcontrolmechanism[70].Li et al.,2008 demonstratedthataserineprotease(Kunitz trypsin)inhibitor(KTI1)ofArabidopsisisinvolvedin modulatingPCDinplant-pathogeninteractions[71]. RNAisilencingoftheAtKTI1generesultedinenhanced lesiondevelopmentafterinf iltrationofleaftissuewith thePCD-elicitingfungaltoxinfumonisinB1(FB1)or theavirulentbacterialpathogen Pseudomonassyringae pvtomatoDC3000carryingavrB(PstavrB).Trypsin inhibitor(KSLFIII1-H12andKLLFIII3-F03)andamiraculinserinetypeendopepti daseinhibitor(FI2-A05) sequenceswerefoundinoriginalearlysubtraction librariesrepresentingtranscriptsexpressedduringearly infection(30min.pi-24hpi).WhileKLLFIII3-F03(trypsinhomologue)geneexpressionwassignificantlysuppressed6hpi,theexpressionofKSLIII1-H12was slightlysuppressedandthen2foldupregulated24hpi. Furtheranalysisshouldbedonetostudythedifference betweenthemechanism(s)ofactionofthesetwogenes.SuppressionofdefenceresponsesAveryevidentdown-regulationofaconsiderablenumberofgeneswasrecordedby6hpiwhichmaybe causedbydefencesuppressionimposedbyXcceffectors (ClustersAandC;Figure6Aand6B).Ithasbeen shownpreviouslythatXccexploitstheTypeIIIsecretionsystem(T3SS)toinjectdifferenteffectorproteins intocitrusplantsinordertoavoidhostrecognitionand subsequentlyMAMPS/PAMP -triggeredimmunity.The bacterialeffectorproteinssuppressplantdefences includingbasaldefence,gene-for-generesistance,and nonhostresistance.Therewasnoaccumulationofany SARgenetranscriptsincludingPR1,amarkerfor enhanceddefence;inadditionsomeotherkeyelements intheSAdefencepathwayweresuppressed.Onthe otherhand,theS-adenosyll-methionine:benzoicacid salicylicacidcarboxylmethyltransferasegene(KLLRI2C03)wasatleast1-foldupre gulatedinkumquatleaves by6hpiinresponsetoXccinoculation;thegeneis knowntoplayaroleinplantdefenceresponses[72].In addition,theSA-bindingprotein2(SABP2KLLRI2G01),alipaseproteinthatbelongstothehydrolase superfamily,wasfoundtobeup-regulatedat6hpiby atleast2fold;thegenewaspreviouslyfoundtobe requiredfortheplantimmuneresponseintobacco[73].RealtimeQuantitativePolymeraseChainReaction ValidationValidationofthepresentedmicroarraydatasetwascarriedoutusingTaqMangene expressionassayfora numberofhomologuesonthearray.Genesthatwere implicatedinplantdefenceincludingbasicleucinezippertranscriptionfactor(KLLFII2-C05),aputativechitinaseprotein(CHI1)(KLLRI2D05),aputativedisease resistanceleucinerichprotein(CSL1-D05),areceptor likeproteinkinase(KLLRI2-H10),abetagalactosidase likeprotein(KLLFII1-B11)andaputativemitogen-activatedprotein(CSL2-A02)we reselectedforvalidation usingqRT-PCR.AssummarizedinTable2andAdditionalfile10,theqRT-PCRdatacorrelatedwiththe microarrayresultsconfirmingtheup-ordown-regulationofallanalyzedgenesalthoughasexpectedthe qRT-PCRwasmoresensitive.ConclusionsInthisstudy,a F.margarita custommicroarrayrepresenting1024unigeneswasusedtostudytheresponse toinoculationwith X.axonopodis pv. citri .AverydistinctthoughdelayedHRwasobservedinXcc-inoculated kumquatplantswhereinitiallythebacteriumgrewexponentially,followedbyasuddenleaftissuecollapse (necrosiswithnocankerlesions)2-5daysafterinoculation[24].AcomparabledelayedHRwasobservedin tomatoresistanceresponsetoraceT3mediatedby AvrXv3effectorandRxvT3Rprotein[74].Thecurrent kumquatanalysisallowedsimultaneousinvestigationof theexpressionofmorethanonegroupofgenesknown tobelinkedtomorethanonebiologicalprocessand cellularcompartmentinre lationtotheHRcausedby Xccinfection.Alargenumberofgeneswerefoundto bedifferentiallyexpressedafterinfection.Mostofthe genesinvolvedindefencemechanismsinkumquat appeartobeassociatedwiththephenomenathatprecedetheHRincludingoxidativeburst,proteindegradation,andregulationofphotosynthesisaswellasthe productionofROSthatisassociatedwiththeoxidativeKhalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page12of17

PAGE 13

burst.Oneverydistinctobservationwasthatsomeof thedefencegenessuchasPR1andNDR1weredownregulatedinkumquatsinresponsetoXccinoculationas earlyas6hpi,aphenomenoncurrentlyunderfurther examination.Whatclearlyappearstobearesistant responseandadrasticdecreaseinthebacterialpopulation,inadditiontotheactivationofgenesinvolvedin ROSproductionaswellasandprogrammedcelldeath, seemstobeacommonmechanismthatispursuedby morethanonecitrusbacterialpathogenwithnoassociated-resistancegenesyetidentified[24,36].Future workwillcomparedifferencesingeneresponseinboth resistantandsusceptiblecitrustypes.MethodsPlantmaterialandinoculationwithbacteriaFortunellamargarita (Lour.)Swingle(Nagamikumquat) plantswereusedinalloftheexperimentsdescribedin thisstudy.Plantswereapproximately2yearsoldatthe timeoftheexperimentandweremaintainedinthe quarantinegreenhousefacilityattheDivisionofPlant Industry,FloridaDepartmentofAgriculture(Gainesville, FL,USA)undercontrolledconditions.Leavesfromaset ofsixkumquatplantswereinfiltratedwithbacterialculturesaccordingtoLundetal.[75].Thebacterialstrain usedwas Xanthomonascitri subsp. citri A;MiamiX0459(Xcc).Theinoculumwasadjustedto5108cfu/ml. Asimilarsetofplantswasmock-inoculatedusingsterile tapwaterascontrols.Leavesfromthetwosetsofplants wereusedinsubsequentexperiments.MicroarrayplatformThekumquatmicroarraychipwasdevelopedand printedattheUniversityofFlorida(Gainesville,FL, USA).ThearrayincludedESTschosenfrom4previouslyconstructedNagamikumquatforwardand reverseleafsubtractioncDNAlibraries.ThecDNA librarieswereconstructedusingRNAextractedfrom leaftissuecollectedatdifferentintervalspostinoculationwithXcc(seebelow)andpooledintoearlyandlate librarysetstocaptureawidespectrumofdifferentially expressedtranscripts[24].RandomDNAsequencing wasperformedfrom5 and3 endsofrandomlyselected clonesusinguniversalprimers,generatingsequence informationfrom2788and1655clonesfromtheearly andlateleafsubtractionlibrariesrespectively. Theinitialdatasetwasreducedtoatotalof2304transcriptsthatwereselectedaccordingtosequencealignmentsimilaritieswithproteinsintheGenbankdatabase. Sequenceswereselectedbasedonqualityandlength. Thedatasetincluded2254kumquatESTscomprising 738contigsand1516singletons,inadditionto50 cDNAcontrolelements.Eachprobewasprintedin3 locationsonthearrayusingtheOmnigridMicroarrayer (GeneMachines,SanCarlos,CA,USA)sothatall cloneshad3technicalreplicatesoneachslide,generatingatotalof6912spots.Postprintingslideprocessing wasperformedasdescribedinHelleretal.(1997)[76] withsomemodifications.I nbrief,acombinationof sequentialbakingandUVcrosslinkingwasimplemented whereslideswerebakedfor80minat80Cinadrying ovenwithoutvacuum.Theslideswerethenwashed twicein0.1%SDSfor5minuteseachtoremoveany unboundDNA.ExperimentalplanToidentifygenesthatareconsideredtobedifferentially expressedinkumquat,atime-courseexperimentwas designedutilizingthekumquat/Xccpathosystem.Six independentXccormock-inoculatedkumquatplants wereused;eachplantwasconsideredanindependent biologicalreplicate.Sincecitruscankerisanon-systemic disease,6-10leavespertreatedkumquatplantwere independentlyinfiltratedusing5108cfu/mlXcc.All RNAsamplesisolatedfromhealthymock-orXcc inoculated(infected)leaves wereprocessedindependently.Itisunlikelythatdifferentialgeneexpression observedwascausedbythepressureinfiltrationinoculationmethodused,sincethisfactorwasnormalizedby treatingthemockinoculatedplantsintheexactsame wayasinfectedplants. Individualleaveswereharvestedfromtheinoculated andmock-inoculatedplantsatspecifictime-pointspostinoculation(pi)accordingtodesignatedconditionsfor eachexperiment;therewere3timepointsforthe microarrayexperimentand5forsubsequentreal-time PCRassays.Forthemicroarrayexperiment,thethree timepoints(6hpi,24hpi,and72hpi)werechosen basedupontheinternalbacterialpopulationspreviously detectedatthesetimesfollo winginoculationandthe knowledgethatkumquatleavesabscised3-5daysafter inoculation.Inaddition,previousexperimentsrevealed thatthereweresometranscriptsdifferentiallyexpressed asearlyas30minpost-inoculationwithXcc.Finally, theRNAyieldandtheabundanceofcellulartranscripts decreasedastheleavesapproachedtotalPCD,ashas beenshownpreviouslybyothers[53].Timepointsof0 and120hpiwereaddedfortherealtimePCRassays. Thehealthymock-inoculatedandXcc-inoculatedleaf sampleswereimmediatelyfrozeninliquidnitrogen.For eachrespectivetimepoint,totalRNAwasextracted usingRNeasycolumns(QIAGEN,Valencia,CA,U.S.A.) accordingtothemanufacturer sprotocol.RNApurity, concentrationandqualitywereassessedusingaspectrophotometerandaBioAnalyzer2100(AgilentTechnologies,PaloAlto,CA).Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page13of17

PAGE 14

Fluorescentprobe,hybridization,andscanningPriortoslidehybridizationwithprobe,slideswereprehybridizedinasolutioncontaining5SSC,0.1%SDS and1%bovineserumalbuminat42Cfor45minto eliminatenonspecificbindingoftheprobetotheslide. SlideswerewashedusingMilliQRNasefreewater,then isopropanol,onceeach,andair-dried.Slidesweremaintainedatthehybridizationtemperatureuntilloaded withprobe.cDNAlabelingwasperformedusingthe Genisphere(Hatfield,PA,USA)3DNAArray50ExpressionArrayDetectionKitaccordingtothemanufacturer sprotocolfortotalRNA.Foreachtime-point 125 gofDNA-freetotalRNAisolatedfromanindependentplant(biologicalreplicate)perslidewasreverse transcribedforeachofthemockinoculatedandthe infectedleafsamplesseparatelyusingAmbionreverse transcriptase(Ambion;LaJolla,CA)inthepresenceof GenispheredTprimers.Two-stephybridizationwas performedasfollows.Thefirsthybridization,carriedout at48Covernight,contained10 loftheconcentrated cDNA(heatdenaturedprobe)madeusingeitherthe Cy5-RTprimercapturesequenceortheCy3-RTprimer capturesequence,inGenis phere2formamide-based hybridizationbuffer.Threesuccessiveposthybridization washeswereperformed,firstin2XSSC,0.2%SDSat 55Cfor10min,then2SSCfor10minandfinallyin 0.2SSCfor10minatroomtemperature.Thesecond hybridization(foradditionofthedye)wasperformed using2.5 lofeitherCy3orCy5dendrimer,2 lof high-enddifferentialbufferand58 lofhybridization buffer.Foreachtimepoint,threemock-inoculatedsamplesfromthreeindividualplantswerelabeledwithCy-3 andtheXccinfectedsampleswerelabeledwithCy-5 andbothwerehybridizedtothesamearray.Post-hybridizationwasheswereconductedasperformedearlier followingtheprimaryhybridization,withtheadditionof 0.1mldithiothreitol(DTT)intothefirstandsecond washsolutionstoreduceoxidationoffluorescentdyes.cDNAmicroarraysetupandqualitycontrolControlmeasuressuchasthedetectionsensitivitylevel weredeterminedusinginternalcontrolprobesandnonspecificcontrolelements.HumangenomicDNA,the greenfluorescentproteingene,andthelambdacontrol templateDNAfragmentwereincludedasnegativecontrols.Additionally,cDNAspreviouslyimplicatedin pathogendefencesuchasPR1andNPR1from Arabidopsis ,NDR1fromcitrus,wereprinted3timesonthe arraytotesttheabilityofthemicroarraymethodto detectchangesingeneexpression.Thesewereconsideredtobespecificpositivecontrols.Inaddition,the microarrayratioforeachgeneanalyzedwasnormalized againstthemicroarrayratioobtainedfor18S.TranscriptomedataanalysisAgilent sFeatureExtractionSof tware(AgilentCorp., PaloAlto,CA)wasusedtoanalyzethemicroarraydata. DatawereuploadedintothestatisticalplatformR[77] forstatisticalanalysisandtheLimmapackagewasused forpre-processing.DatawereLowestransformedfollowedbyscalingbetweenarrays[78].Twoarrayslides werechosenforeachtime-pointexperimentbasedon theconsistencyofthesignalacrossthereplicates.The folddifferenceinexpressionwascomputedas:2 averageratio(2tothepoweroftheaverageratio).cDNAs withanaverageratioof1.0orhigherwereconsidered differentiallyexpressed,whichrepresentsa1.5foldor higherdifferenceinexpression.Statisticalanalysiswas performedusingtwodifferentapproaches.Time-dependentgeneexpressionchangeswereanalyzedbythe maSigPromethodology[79].Dataweresubsequently subjectedtoASCA-geneanalysisthatcombines ANOVAandmultivariatemethodstoidentifymainand secondarypatternsofgeneexpressionassociatedwith differentexperimentalfactors[34].Statisticalanalysis identifiedanumberofselectedgenesthatwerefurther groupedintoclusters. Functionalinformationa bouttheESTsrepresented inthearraywasobtainedbyBlast2GOanalysisusing defaultparameters[29].Blast2GOusesBlastandan elaboratedannotationalgorithmtoassignGeneOntology(GO),EnzymeCodeandInterProfunctionallabels toasetofuncharacterizedsequences[27].Thefunctionalcharacterizationoftheseclusterswasdoneby applyingtheFunctionalEnrichment(FE)method includedinBlast2GOwhichimplementstheGossip algorithm.FEmethodsasses swhichfunctionalcategoriesareover-representedwithinagroupofgenesin relationtoabroaderlist,inthiscasethewholekumquatarray.Finally,themajo rinducedtranscriptional changesconsideredfunctionalclassesasawholewere studiedwiththePCA-maSigFunmethod[80].This methodcombinesPrincipalComponentAnalysisand maSigProtocharacterizethe expressionprofiles associatedwithcellularfunctions.Sequencedatafrom thisworkhavebeendepositedintheNCBIGenbank databaselibraries(http://www.ncbi.nlm.nih.gov/ GenBank/index.html),usingtheBankItdbESTdatabase,andaccessionnumberswereobtained.[Genbank: GW687757toGW690680].(SeeAdditionalFile11).QuantitativeReal-timeQuantitativePCRKumquatleaveswereinfiltratedwithXcc(5108cfu permilliliter),thentotalRNAwasisolatedfrominoculatedleaves0,6,24,72,120hpiforboththemicroarray andthequantitativereal-timePCRexperimentaspreviouslystatedinthe experimentalplan .Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page14of17

PAGE 15

TotalRNAwasisolatedseparatelyforeachrespective timepointusingtheTRIzolreagent(Invitrogen,Carlsbad,CA,U.S.A.)accordingtomanufacturer sinstructions.RNAsampleswerefurtherpurifiedusingthe RNeasyPlantMiniKit(Qiagen,Valencia,CA)including DNaseITURBODNase(Ambion,Austin,TX,U.S.A.) treatmentaccordingtothemanufacturer sinstructions. RNAqualityandquantit ywasthenassessedusing microspectrophot ometry(NanodropTechnologies,Inc., Wilmington,DE).cDNAwassynthesizedfrom1 gof totalRNAusingAppliedBiosystems(HighCapacity cDNAReverseTranscriptionKitPN4368813,4374966). ATaqMangeneexpressionassaywasthenusedtovalidatethetranscriptaccumulationlevelsofaspecificsubsetofgenesfromthekumquatmicroarrays.Reactions wereperformedintheABIPrism7900HTsequence detector(AppliedBiosystems,FosterCity,CA,U.S.A.). PrimersforqRT-PCRweredesignedusingthePrimer Express3.0software(Appli edBiosystems),anddata werenormalizedusingaTaqmanribosomalRNAcontrol,inadditiontothekumquat18Sribosomalgene thatservedasaninternalcontrolwhereeachreal-time PCRreactionwasdoneinparallelwiththe18Sprimers. Forinternalcontrols,anumberofgenes,forexample actinshowedinconsistencies.The18ssurprisingly showedcoherencythroughouttheinteraction,andthis wasnoticedduringthemicroarraysandwassubsequentlyconfirmedwiththeRealtimePCR(RT-qPCR) 18Sexpressioncurve.qRT-PCRwascarriedoutat50C for2minutes,95Cfor10minutes,followedby40 cyclesat95Cfor15secondsand60Cfor1minute.AdditionalmaterialAdditionalfile1:FigureS1 S imilarityofthequerysetwiththeNCBI database. Additionalfile2:FigureS2 .Distributionofthecutoffforthee-value afterblastxtoNCBInr. Additionalfile3:FigureS3 .Speciesdistributionchartofkumquat transcriptsafterblastxtoNCBInr. Additionalfile4:TableS1 .KumquatGOdistributionanalysiswith sequenceIDs,descriptionsandannotationtoeachgeneontology category. Additionalfile5:FigureS4 .AmaSigProenrichedMultilevelPiechart illustratingthedistributionofMolecularFunctionswithinthestatistically significant(P<0.025insinglet-test)kumquatexpressedgenesonthe chip. Additionalfile6:FigureS5 .AchartoftheCellularComponents distributionwithinthemaSigProenrichedstatisticallysignificant(P< 0.025insinglet-test)kumquatexpressedgenesinthechip. Additionalfile7:TableS2 .DifferentiallyexpressedASCA-selectedlistof genes,atstatisticallysignificantlevels,includingpvalues,Mvalues. Additionalfile8:TableS3 .DifferentiallyexpressedmaSigPro-selected listofgenes,includingpvalues,Mvalues. Additionalfile9:TableS4 .Combinedanalysisofstatisticallysignificant maSigPro+ASCAselectedgenes(Pvalue 0.05). Additionalfile10:FigureS6 .QuantitativerealtimePCR(qRT-PCR) analysesofsixselectedkumquatESTs(AAM60932ABM67698,AA089566, AAk81874,AAC35981,AAV91900)inkumquatinoculatedwith Xanthomonasaxonopodis pv. citri strainusinga(5108cfu/ml) concentrationoftheMiamiAstrainX04-59.Leaftissuewassampledfor bothinoculatedandmock-inoculatedplantsat0,6,24,48,72and120 hpi.(Anaverageofthreeindependentbiologicalreplications). Additionalfile11:TableS5 .Sequencedatafromthisworkhavebeen depositedintheNCBIGenbankdatabaselibraries[Genbank:GW687757to GW690680]. AcknowledgementsandFunding ThisresearchwassupportedinpartbytheCitrusResearchandEducation Center(CREC),Dr.FrederickGmitter,Jr.andbyagrantfromtheFlorida CitrusProductionResearchAdvisoryCouncil,FloridaDepartmentofCitrus (FDACSContractNumber67), PeterMcClure,Chairman Wegratefullyacknowledgethegenerousfinancialandintellectualsupport fromFloridacitrusgrowers,Mrs.HariotandMr.BarnetteE. Barney Greene Jr.thathelpedcompletethiswork. WewouldalsoliketothankDr.HenryBaker,ProfessorandInterimChairand Dr.HassanBadrane(Medicine:MolecularGeneticsandMicrobiologyDeptat UF)forprovidinghelpwithmicroarraysequipments.Wewouldalsoliketo thankDr.JeffreyJonesandDr.VicenteFebresfortheirobjectivecritical reviewofthemanuscript. Authordetails1PlantMolecularandCellularBiologyProgram(PMCB),HorticulturalSciences Department,UniversityofFlorida,Gainesville,Fl.,32611,USA.2PMCB,Citrus ResearchandEducationCenter,UniversityofFlorida,LakeAlfred,Fl.,USA.3CentrodeInvestigacinPrncipeFelipe,Valencia,SPAIN. Authors contributions AKcarriedoutthemoleculargeneticstudies,participatedindataanalysis, anddraftedthemanuscript.ACcarriedouttheB2GOstatisticalanalysis supervisedbyJD.FGparticipatedinthedesignofthestudyinadditionto coordinationbetweentheauthorsandreviewofthemanuscript.GM participatedinitsexperimentaldesignandreviewedthemanuscript.All authorsreadandapprovedthefinalmanuscript. Received:22June2011Accepted:11November2011 Published:11November2011 References1.GottwaldTR,GrahamJH,SchubertTS: Citruscanker:Thepathogenandits impact. OnlinePlantHealthProgress 2002. 2.G owackiS,MacioszekV,KononowiczA: Rproteinsasfundamentalsof plantinnateimmunity. Cellular&MolecularBiologyLetters 2010,1-24. 3.JonesJD,DanglJL: Theplantimmunesystem. Nature 2006, 444(7117) :323-329. 4.FujikawaT,IshiharaH,LeachJE,TuyumuS: Suppressionofdefense responseinplantsbyavrBs3/pthAgenefamilyofXanthomonasspp. MolPlant-MicrobeInteract 2006, 19 :342-349. 5.YangB,SugioA,WhiteFF: Avoidanceofhostrecognitionbyalterations intherepetitiveandC-terminalregionsofAvrXa7,atypeIIIeffectorof Xanthomonasoryzaepv.oryzae. MolPlantMicrobeInteract 2005, 18(2) :142-149. 6.AlfanoJR,CollmerA: ThetypeIII(Hrp)secretionpathwayofplant pathogenicbacteria:traffickingharpins,Avrproteins,anddeath. J Bacteriol 1997, 179(18) :5655-5662. 7.SanabriaN,GoringD,NurnbergerT,DuberyI: Self/nonselfperceptionand recognitionmechanismsinplants:acomparisonofself-incompatibility andinnateimmunity. NewPhytol 2008, 178(3) :503-514. 8.GottigN,GaravagliaBS,DaurelioLD,ValentineA,GehringC,OrellanoEG, OttadoJ: Xanthomonasaxonopodispv.citriusesaplantnatriuretic peptide-likeproteintomodifyhosthomeostasis. ProcNatlAcadSciUSA 2008, 105(47) :18631-18636. 9.BestwickCS,BrownIR,BennettMH,MansfieldJW: Localizationof hydrogenperoxideaccumulationduringthehypersensitivereactionofKhalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page15of17

PAGE 16

lettucecellsto Pseudomonassyringae pv phaseolicola PlantCell 1997, 9(2) :209-221. 10.AliR,MaW,Lemtiri-ChliehF,TsaltasD,LengQ,vonBodmanS, BerkowitzGA: Deathdon thavenomercyandneitherdoescalcium: ArabidopsisCYCLICNUCLEOTIDEGATEDCHANNEL2andinnate immunity. PlantCell 2007, 19(3) :1081-1095. 11.ShiotaniH,FujikawaT,IshiharaH,TsuyumuS,OzakiK: ApthAhomolog from Xanthomonasaxonopodis pv. citri responsibleforhost-specific suppressionofvirulence. JBacteriol 2007, 189(8) :3271-3279. 12.BruningsAM,GabrielDW: Xanthomonascitri: breakingthesurface. MolecularPlantPathology 2003, 4(3) :141-157. 13.RomerP,HahnS,JordanT,StraussT,BonasU,LahayeT: Plantpathogen recognitionmediatedbypromoteractivationofthepepperBs3 resistancegene. Science 2007, 318(5850) :645-648. 14.YangB,WhiteFF: DiversemembersoftheAvrBs3/PthAfamilyoftypeIII effectorsaremajorvirulencedeterminantsinbacterialblightdiseaseof rice. MolPlantMicrobeInteract 2004, 17(11) :1192-1200. 15.DuanYP,AC,GZ,GE,GabrielDW: Expressionofasingle,host-specific, bacterialpathogenicitygeneinplantcellselicitsdivision,enlargement, andcelldeath. MolecularPlant-MicrobeInteractions 1999, 12(6) :556-560. 16.TasicL,BorinPF,KhaterLC,RamosCH: Cloningandcharacterizationof threehypotheticalsecretionchaperoneproteinsfromXanthomonas axonopodispv.citri. ProteinExprPurif 2007, 53(2) :363-369. 17.WhiteFF,PotnisN,JonesJB,KoebnikR: ThetypeIIIeffectorsof Xanthomonas. MolPlantPathol 2009, 10(6) :749-766. 18.SchenkPM,KazanK,MannersJM,AndersonJP,SimpsonRS,WilsonIW, SomervilleSC,MacleanDJ: SystemicgeneexpressioninArabidopsis duringanincompatibleinteractionwithAlternariabrassicicola. Plant Physiol 2003, 132(2) :999-1010. 19.SpoelSH,DongX: Makingsenseofhormonecrosstalkduringplant immuneresponses. CellHostMicrobe 2008, 3(6) :348-351. 20.LorenzoO,PiquerasR,Sanchez-SerranoJJ,SolanoR: ETHYLENERESPONSE FACTOR1integratessignalsfromethyleneandjasmonatepathwaysin plantdefense. PlantCell 2003, 15(1) :165-178. 21.NemhauserJL,HongF,ChoryJ: Differentplanthormonesregulatesimilar processesthroughlargelynonoverlappingtranscriptionalresponses. Cell 2006, 126(3) :467-475. 22. Glazebrook J: Useofmicroarrayanalysistodissecttheplantdefense response. MethodsMolBiol 2007, 354 :121-130. 23.Leon-ReyesA,SpoelSH,DeLangeES,AbeH,KobayashiM,TsudaS, MillenaarFF,WelschenRA,RitsemaT,PieterseCM: EthyleneModulatesthe RoleofNPR1inCross-TalkBetweenSalicylateandJasmonateSignaling. PlantPhysiol 2009. 24.KhalafAA,MooreGA,JonesJB,GmitterFGJr: Newinsightsintothe resistanceofNagamikumquattocankerdisease. Physiologicaland MolecularPlantPathology 2007, 71 :240-250. 25.AlfanoJR,CollmerA: Bacterialpathogensinplants:Lifeupagainstthe wall. PlantCell 1996, 8(10) :1683-1698. 26.AlbaR,FeiZ,PaytonP,LiuY,MooreSL,DebbieP,CohnJ,D AscenzoM, GordonJS,RoseJK, etal : ESTs,cDNAmicroarrays,andgeneexpression profiling:toolsfordissectingplantphysiologyanddevelopment. PlantJ 2004, 39(5) :697-714. 27.GotzS,Garcia-GomezJM,TerolJ,WilliamsTD,NagarajSH,NuedaMJ, RoblesM,TalonM,DopazoJ,ConesaA: High-throughputfunctional annotationanddataminingwiththeBlast2GOsuite. NucleicAcidsRes 2008, 36(10) :3420-3435. 28.TalonM,GmitterFGJr: Citrusgenomics. IntJPlantGenomics 2008, 2008 :528361. 29.ConesaA,GotzS,Garcia-GomezJM,TerolJ,TalonM,RoblesM: Blast2GO:a universaltoolforannotation,visualizationandanalysisinfunctional genomicsresearch. Bioinformatics 2005, 21(18) :3674-3676. 30.CohnJR,MartinGB: Pseudomonassyringae pv. tomato typeIIIeffectors AvrPtoandAvrPtoBpromoteethylene-dependentcelldeathintomato. PlantJ 2005, 44(1) :139-154. 31.DesikanR,SAH-M,HancockJT,NeillSJ: RegulationoftheArabidopsis transcriptomebyoxidativestress. PlantPhysiol 2001, 127(1) :159-172. 32.GangDR,WangJ,DudarevaN,NamKH,SimonJE,LewinsohnE, PicherskyE: Aninvestigationofthestorageandbiosynthesisof phenylpropenesinsweetbasil. PlantPhysiol 2001, 125(2) :539-555. 33.ConesaA,GotzS: Blast2GO:Acomprehensivesuiteforfunctional analysisinplantgenomics. IntJPlantGenomics 2008, 2008 :619832. 34.NuedaMJ,ConesaA,WesterhuisJA,HoefslootHC,SmildeAK,TalonM, FerrerA: Discoveringgeneexpressionpatternsintimecoursemicroarray experimentsbyANOVA-SCA. Bioinformatics 2007, 23(14) :1792-1800. 35. Kay S,HahnS,MaroisE,HauseG,BonasU: ABacterialEffectorActsasa PlantTranscriptionFactorandInducesaCellSizeRegulator. Science 2007, 318(5850) :648-651. 36.CernadasRA,CamilloLR,BenedettiCE: Transcriptionalanalysisofthe sweetorangeinteractionwiththecitruscankerpathogens Xanthomonasaxonopodispv.citriandXanthomonasaxonopodispv. aurantifolii. MolPlantPathol 2008, 9(5) :609-631. 37.AlbrechtU,BowmanKD: Transcriptionalresponseofsusceptibleand tolerantcitrustoinfectionwithCandidatusLiberibacterasiaticus. Plant Science 2011. 38.RossJ,LiY,LimE,BowlesDJ: Higherplantglycosyltransferases. Genome Biol 2001, 2(2) ,Reviews3004. 39.DixonRA: Naturalproductsandplantdiseaseresistance. Nature 2001, 411(6839) :843-847. 40.MugfordST,QiX,BakhtS,HillL,WegelE,HughesRK,PapadopoulouK, MeltonR,PhiloM,SainsburyF, etal : Aserinecarboxypeptidase-like acyltransferaseisrequiredforsynthesisofantimicrobialcompoundsand diseaseresistanceinoats. PlantCell 2009, 21(8) :2473-2484. 41.NandaAK,AndrioE,MarinoD,PaulyN,DunandC: ReactiveOxygen Speciesduringplant-microorganismearlyinteractions. Journalof IntegrativePlantBiology 2010, 52(2) :195-204. 42.TorresMA: ROSinbioticinteractions. PhysiologiaPlantarum 2010, 138(4) :414-429. 43.dePintoMC,ParadisoA,LeonettiP,DeGaraL: Hydrogenperoxide,nitric oxideandcytosolicascorbateperoxidaseatthecrossroadbetween defenceandcelldeath. PlantJ 2006, 48(5) :784-795. 44.VranovaE,InzeD,VanBreusegemF: Signaltransductionduringoxidative stress. JExpBot 2002, 53(372) :1227-1236. 45.KuzniakE,SklodowskaM: Theeffectof Botrytiscinerea infectiononthe antioxidantprofileofmitochondriafromtomatoleaves. JExpBot 2004, 55(397) :605-612. 46.HalliwellB,FoyerCH: Ascorbicacid,metalionsandthesuperoxide radical. BiochemJ 1976, 155(3) :697-700. 47.FoyerCH,NoctorG: Redoxhomeostasisandantioxidantsignaling:a metabolicinterfacebetweenstressperceptionandphysiological responses. PlantCell 2005, 17(7) :1866-1875. 48.TrumperS,FollmannH,HaberleinI: Anovel-dehydroascorbatereductase fromspinachchloroplastshomologoustoplanttrypsininhibitor. FEBS Lett 1994, 352(2) :159-162. 49. Yang C-W,Gonzalez-LamotheR,EwanRA,RowlandO,YoshiokaH, ShentonM,YeH,O DonnellE,JonesJDG,SadanandomA: TheE3 ubiquitinligaseactivityofArabidopsisPLANTU-BOX17andIts functionaltobaccoHomologACRE276arerequiredforcelldeathand defense. PlantCell 2006, 18(4) :1084-1098. 50.TrujilloM,IchimuraK,CasaisC,ShirasuK: NegativeregulationofPAMPtriggeredimmunitybyanE3ubiquitinligasetripletinArabidopsis. Curr Biol 2008, 18(18) :1396-1401. 51.BonfigKB,SchreiberU,GablerA,RoitschT,BergerS: Infectionwithvirulent andavirulentP.syringaestrainsdifferentiallyaffectsphotosynthesisand sinkmetabolisminArabidopsisleaves. Planta 2006, 225(1) :1-12. 52.MysoreKS,D AscenzoMD,HeX,MartinGB: Overexpressionofthedisease resistancegenePtointomatoinducesgeneexpressionchangessimilar toimmuneresponsesinhumanandfruitfly. PlantPhysiol 2003, 132(4) :1901-1912. 53.QuirinoBF,NohYS,HimelblauE,AmasinoRM: Molecularaspectsofleaf senescence. TrendsPlantSci 2000, 5(7) :278-282. 54.AlbertM,WernerM,ProkschP,FrySC,KaldenhoffR: Thecellwallmodifyingxyloglucanendotransglycosylase/hydrolaseLeXTH1is expressedduringthedefencereactionoftomatoagainsttheplant parasiteCuscutareflexa. PlantBiol(Stuttg) 2004, 6(4) :402-407. 55.Zhu-SalzmanK,SalzmanRA,AhnJE,KoiwaH: Transcriptionalregulationof sorghumdefensedeterminantsagainstaphloem-feedingaphid. Plant Physiol 2004, 134(1) :420-431. 56.LeeS,KimSY,ChungE,JoungYH,PaiHS,HurCG,ChoiD: ESTand microarrayanalysesofpathogen-responsivegenesinhotpepper (CapsicumannuumL.)non-hostresistanceagainstsoybeanpustule pathogen(Xanthomonasaxonopodispv.glycines). FunctIntegrGenomics 2004, 4(3) :196-205.Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page16of17

PAGE 17

57.BrunnerF,StintziA,FritigB,LegrandM: Substratespecificitiesoftobacco chitinases. PlantJ 1998, 14(2) :225-234. 58.KasprzewskaA: Plantchitinases regulationandfunction. CellMolBiolLett 2003, 8(3) :809-824. 59.LebelE,HeifetzP,ThorneL,UknesS,RyalsJ,WardE: Functionalanalysisof regulatorysequencescontrollingPR-1geneexpressioninArabidopsis. PlantJ 1998, 16(2) :223-233. 60.MitsuharaI,IwaiT,SeoS,YanagawaY,KawahigasiH,HiroseS,OhkawaY, OhashiY: CharacteristicexpressionoftwelvericePR1familygenesin responsetopathogeninfection,wounding,anddefense-relatedsignal compounds(121/180). Moleculargeneticsandgenomics:MGG 2008, 279(4) :415-427. 61.CamposMA,RosaDD,TeixeiraJC,TargonMLPN,SouzaAA,PaivaLV, Stach-MachadoDR,MachadoMA: PRgenefamiliesofcitrus:theirorgan specific-bioticandabioticinducibleexpressionprofilesbasedonESTs approach. GeneticsandMolecularBiology 2007, 30 :917-930. 62.LiC-W,SuR-C,ChengC-P,YouS-J,HsiehT-H,ChaoT-C,ChanM-T: Tomato RAVTranscriptionFactorIsaPivotalModulatorInvolvedintheAP2/ EREBP-MediatedDefensePathway. PlantPhysiology 2011, 156(1) :213-227. 63.DoddsPN,LawrenceGJ,EllisJG: Sixaminoacidchangesconfinedtothe leucine-richrepeatbeta-strand/beta-turnmotifdeterminethedifference betweenthePandP2rustresistancespecificitiesinflax. PlantCell 2001, 13(1) :163-178. 64.PontierD,GodiardL,MarcoY,RobyD: hsr203J,atobaccogenewhose activationisrapid,highlylocalizedandspecificforincompatibleplant/ pathogeninteractions. PlantJ 1994, 5(4) :507-521. 65.CloughSJ,FenglerKA,YuIC,LippokB,SmithRKJr,BentAF: The Arabidopsisdnd1 defense,nodeath geneencodesamutatedcyclic nucleotide-gatedionchannel. ProcNatlAcadSciUSA 2000, 97(16) :9323-9328. 66.GreenDR,ReedJC: Mitochondriaandapoptosis. Science 1998, 281(5381) :1309-1312. 67.LoganDC: Thedynamicplantchondriome. SeminarsinCell& DevelopmentalBiology 2010, 21(6) :550-557. 68.MartinouJ-C,DesagherS,AntonssonB: Cytochromecreleasefrom mitochondria:allornothing. NatCellBiol 2000, 2(3) :E41-E43. 69.SanchezP,deTorresZabalaM,GrantM: AtBI-1,aplanthomologueofBax inhibitor-1,suppressesBax-inducedcelldeathinyeastandisrapidly upregulatedduringwoundingandpathogenchallenge. PlantJ 2000, 21(4) :393-399. 70.HoeberichtsFA,WolteringEJ: Multiplemediatorsofplantprogrammed celldeath:Interplayofconservedcelldeathmechanismsandplantspecificregulators. BioEssays 2003, 25(1) :47-57. 71.LiJ,BraderG,PalvaET: Kunitztrypsininhibitor:anantagonistofcell deathtriggeredbyphytopathogensandfumonisinb1inArabidopsis. MolPlant 2008, 1(3) :482-495. 72.XuR,SongF,ZhengZ: OsBISAMT1,ageneencodingS-adenosyl-Lmethionine:salicylicacidcarboxylmethyltransferase,isdifferentially expressedinricedefenseresponses. MolBiolRep 2006, 33(3) :223-231. 73.KumarD,KlessigDF: High-affinitysalicylicacid-bindingprotein2is requiredforplantinnateimmunityandhassalicylicacid-stimulated lipaseactivity. ProcNatlAcadSciUSA 2003, 100(26) :16101-16106. 74.GiblyA,BonshtienA,BalajiV,DebbieP,MartinGB,SessaG: Identification andexpressionprofilingoftomatogenesdifferentiallyregulatedduring aresistanceresponseto Xanthomonascampestris pv. vesicatoria Mol PlantMicrobeInteract 2004, 17(11) :1212-1222. 75.LundST,StallRE,KleeHJ: Ethyleneregulatesthesusceptibleresponseto pathogeninfectionintomato. PlantCell 1998, 10(3) :371-382. 76.HellerRA,SchenaM,ChaiA,ShalonD,BedilionT,GilmoreJ,WoolleyDE, DavisRW: Discoveryandanalysisofinflammatorydisease-relatedgenes usingcDNAmicroarrays. ProcNatlAcadSciUSA 1997, 94(6) :2150-2155. 77.SmythGK,(ed): Limma:Linearmodelsformicroarraydata. NewYork: Springer;2005. 78.YangYH,DudoitS,LuuP,LinDM,PengV,NgaiJ,SpeedTP: Normalization forcDNAmicroarraydata:arobustcompositemethodaddressingsingle andmultipleslidesystematicvariation. NucleicAcidsRes 2002, 30(4) :e15. 79.ConesaA,NuedaMJ,FerrerA,TalonM: maSigPro:amethodtoidentify significantlydifferentialexpressionprofilesintime-coursemicroarray experiments. Bioinformatics 2006, 22(9) :1096-1102. 80.NuedaMJ,SebastianP,TarazonaS,Garcia-GarciaF,DopazoJ,FerrerA, ConesaA: Functionalassessmentoftimecoursemicroarraydata. BMC Bioinformatics 2009, 10(Suppl6):S9.doi:10.1186/1471-2229-11-159 Citethisarticleas: Khalaf etal .: Fortunellamargarita Transcriptional ReprogrammingTriggeredby Xanthomonascitri subsp. citri BMCPlant Biology 2011 11 :159. Submit your next manuscript to BioMed Central and take full advantage of: Convenient online submission Thorough peer review No space constraints or color gure charges Immediate publication on acceptance Inclusion in PubMed, CAS, Scopus and Google Scholar Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Khalaf etal BMCPlantBiology 2011, 11 :159 http://www.biomedcentral.com/1471-2229/11/159 Page17of17