ATL9, a RING Zinc Finger Protein with E3 Ubiquitin Ligase Activity Implicated in Chitin

16 Pages·2010·2.62 MB·English

ATL9, a RING Zinc Finger Protein with E3 Ubiquitin
Ligase Activity Implicated in Chitin- and NADPH Oxidase-
Mediated Defense Responses
Marta Berrocal-Lobo1¤b, Sophia Stone3, Xin Yang2, Jay Antico2, Judy Callis3, Katrina M. Ramonell2*,
Shauna Somerville1¤a
1DepartmentofPlantBiology,CarnegieInstitution,Stanford,California,UnitedStatesofAmerica,2DepartmentofBiologicalSciences,UniversityofAlabama,Tuscaloosa,
Alabama,UnitedStatesofAmerica,3SectionofMolecularandCellularBiology,UniversityofCaliforniaDavis,Davis,California,UnitedStatesofAmerica
Abstract
Pathogenassociatedmolecularpatterns(PAMPs)aresignalsdetectedbyplantsthatactivatebasaldefenses.Oneofthese
PAMPsischitin,acarbohydratepresentinthecellwallsoffungiandininsectexoskeletons.Previousworkhasshownthat
chitintreatmentofArabidopsisthalianainduceddefense-relatedgenesintheabsenceofapathogenandthattheresponse
wasindependentofthesalicylicacid(SA),jasmonicacid(JA)andethylene(ET)signalingpathways.Oneofthesegenesis
ATL9 (=ATL2G), which encodes a RING zinc-finger like protein. In the current work we demonstrate that ATL9 has E3
ubiquitinligaseactivityandislocalizedtotheendoplasmicreticulum.TheexpressionpatternofATL9ispositivelycorrelated
with basal defense responses against Golovinomyces cichoracearum, a biotrophic fungal pathogen. The basal levels of
expressionandtheinductionofATL9bychitin,inwildtypeplants,dependsontheactivityofNADPHoxidasessuggesting
thatchitin-mediateddefenseresponseisNADPHoxidasedependent.AlthoughATL9expressionisnotinducedbytreatment
withknowndefensehormones(SA,JAorET),fullexpressioninresponsetochitiniscompromisedslightlyinmutantswhere
ET-orSA-dependentsignalingissuppressed.Microarrayanalysisoftheatl9mutantrevealedcandidategenesthatappearto
actdownstreamofATL9inchitin-mediateddefenses.Theseresultshintatthecomplexityofchitin-mediatedsignalingand
thepotentialinterplaybetweenelicitor-mediatedsignaling,signalingviaknowndefensepathwaysandtheoxidativeburst.
Citation:Berrocal-LoboM,StoneS,YangX,AnticoJ,CallisJ,etal.(2010)ATL9,aRINGZincFingerProteinwithE3UbiquitinLigaseActivityImplicatedinChitin-
andNADPHOxidase-MediatedDefenseResponses.PLoSONE5(12):e14426.doi:10.1371/journal.pone.0014426
Editor:DanielJ.Kliebenstein,UniversityofCalifornia,UnitedStatesofAmerica
ReceivedJune24,2010;AcceptedNovember23,2010;PublishedDecember23,2010
Copyright:(cid:2)2010Berrocal-Loboetal.Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermits
unrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalauthorandsourcearecredited.
Funding: Marta Berrocal-Lobo was supported by the Carnegie Institution of Washington and a postdoctoralfellowship from theSecretaria de Estado de
EducacionyUniversidadesandFondoSocialEuropeo,MECD/Fullbright(Spain).KatrinaM.RamonellwassupportedbytheUniversityofAlabamaandbygrants
fromtheNationalInstitutesofHealth(grantno.1-R15-GM073630-01)andtheU.S.DepartmentofAgriculture(grantno.2007-01652toK.M.R.).Thisworkwas
supportedinpartbytheNationalScienceFoundationandtheCarnegieInstitution.Thefundershadnoroleinstudydesign,datacollectionandanalysis,decision
topublish,orpreparationofthemanuscript.
CompetingInterests:Theauthorshavedeclaredthatnocompetinginterestsexist.
*E-mail:[email protected]
¤aCurrentaddress:EnergyBiosciencesInstitute,UniversityofCalifornia,Berkeley,California,UnitedStatesofAmerica
¤bCurrentaddress:CentrodeBiotecnolog´ıayGeno´micadePlantas(CBGP),(UPM-INIA),CampusdeMontegancedo,UniversidadPolite´cnicadeMadrid,Pozuelo
deAlarco´n,Madrid,Spain
Introduction strong defense responses [4,5,6,7,8]. Several receptors associated
with these PAMPs have been characterized including: the FLS2
Plantsdefendagainstpathogensusinganinnatesystemofdefense receptor that recognizes flagellin [9]; the EFR receptor, which
that has both constitutive andinducible components.Constitutive perceivesthefirst18aminoacidsofbacterialelongationfactorTu
defense responses are independent of the physical presence of a (EF-Tu)[7],chitinoligosaccharideelicitorbindingprotein(CeBiP)
pathogen and are typically chemical and physical barriers that [10],atransmembraneproteinwithtwoextracellularLysinemotifs
protect the plant from pathogen invasion [1]. Inducible plant (LysM) that is involved in chitin recognition and the LysM-RLK
defenses depend on pathogen recognition and fall into two major CERK1 that is required for chitin-initiated responses and
classes; specific gene-for-gene interactions and more general downstreamsignalling[10,11].
Pathogen or Microbe-Associated Molecular Pattern (PAMP or While our knowledge of how plants perceive pathogens and
MAMP)-associatedresponses.Ingene-for-geneinteractions,aplant activate associated defense signaling pathways is increasing
resistance(R)generecognizesandinteractswithaspecificrace(s)of rapidly, less is known about how these processes are regulated
pathogenthatexpressesacorrespondingavirulence(Avr)gene[2]. duringtheinfection.Apredominantthemethatisemergingisthat
In the absence of gene-for-gene interactions, general elicitors or of ubiquitination as a means of targeting components of defense
PAMPs,[3]arerecognizedbythehostandactivatebroad-spectrum signaling pathways for degradation to curtail the plant immune
defense responses. Common PAMPs such as the oligosaccharide response [12,13]. Studies characterizing the roles of several
chitin(b-1,4linkedN-acetylglucosamine)andthebacterialproteins ubiquitinE3ligasesindefensehavebeguntoprovidecluesabout
flagellin and elongation factor Tu (EF-Tu) are known to activate theregulationofpathogen-inducedsignaling[12,14].Forinstance
PLoSONE | www.plosone.org 1 December2010 | Volume 5 | Issue 12 | e14426
ATL99sRoleinPlantDefense
the rice resistance (R) protein Xa21 has been shown to interact sevenareknowntobespecificallyinducedbyelicitorsandfourof
with an E3 ubiquitin ligase XB3 [15]. Interaction between XB3 the proteins (ATL9, ATL2/ACRE132, ATL6 and ATL17) are
and Xa21 is required for the accumulation of the XA21 protein induced specifically by flagellin or chitin (Figure 1B). All eight
and is necessary for Xa21-mediated resistance to Xanthomonas proteins contain a RING zinc-finger motif with six highly
oryzae pv. oryza in rice [15,16]. A RING-finger type protein from conserved cysteine residues that match the consensus sequence
pepper CaRFP1 was shown to physically interact with PR-1 for the C3HC4 type (RING zinc-finger) domain group (http://
(pathogenesis related-1) protein in leaves of plants after infection www.sanger.ac.uk/Software/Pfam/). Of the genes found only
withbothbacterialandfungalpathogens[17].Over-expressionof three,ATL9inArabidopsis,NtACRE132intobaccoandOsBIRF1
CaRFP1 in transgenic Arabidopsis conferred disease susceptibility in rice have been tested for their putative role in response to
to Pseudomonas syringae pv. tomato and reduced PR-2 and PR-5 pathogens (Figure1B).
expressionsuggestingthatCaRFP1isanE3ligasethattargetsPR
proteins [17]. The induction of ATL9 is independent of the classical
E3 ligases also appear to play a prominent role in elicitor- defense pathways
mediated defense responses. In particular, members of the ATL
Studies have shown that chitin-induced defense responses act
(Arabidopsisto´xicosenlevadura)[18,19]genefamilyhavebeenshown
throughanindependentsignalingpathwayandarenotdependent
tobeactivatedbyelicitorsandtoplayimportantrolesindefense
on SA-, JA- or ET-mediated responses [29]. In some cases,
pathways.TheArabidopsisATLgenefamilycontains80members
however, expression levels of chitin-induced genes were found to
andisaconservedgroupofRINGzinc-fingerproteinsthatencode
be slightly compromised in mutants defective in the SA- and JA-
putativeE3ubiquitinligases[20].ATL2andATL6inArabidopsis
dependent signaling pathways suggesting some level of cross-talk
and EL5 in rice, all encoding RING-finger type E3 ligases, have
[29]. In order to determine whether the induction of ATL9 was
beenshowntoberapidlyinducedinresponsetotheelicitorchitin
mediated solely by chitin or might also be regulated by the SA-,
[18,21,22,23]. Recent work by Hondo et al. [24] demonstrated
JA- or ET-dependent signaling pathways, Col-0 plants and
thatthetomatoorthologofArabidopsisATL2,LeATL6,responded
mutants impaired in each signaling pathway (SA: sid2-1, npr1-1;
to cell wall protein fraction elicitor from the biocontrol agent
JA: jar1; ET: ein2-5) were treated with chitin for 30 minutes and
Pythium oligandrum and appeared to regulate the jasmonic acid-
ATL9 expression was monitored. All plants tested showed
dependent defense gene expression. In a screen for chitin-
induction of ATL9 when treated with chitin compared to
responsive genes in Arabidopsis, we identified an ATL family
untreated controls, although in all cases levels of induction in
member, ATL9 (At2g35000; ATL2G), that responded strongly to
themutantlineswerelowerthaninwildtype.InductionofATL9
chitin treatment [6]. Loss-of-function mutations in this gene
by chitin treatment was higher in Col-0 plants than in any other
resulted in increased susceptibility to the powdery mildew
line tested (Figure 2A). Among the mutants, the highest levels of
pathogen, Golovinomyces cichoracearum (=Erysiphe cichoracearum) [6].
ATL9 induction were detected in the ein2-5 and npr1-1 mutants,
OurresultshereconfirmthatATL9isanE3ubiquitinligaseand
whileATL9levelsinsid2-1andjar-1werethelowest(Figure2A).
show that it is localized to the endoplasmic reticulum. ATL9
ThebasallevelsofATL9ineachofthemutantsareslightlylower
expressionisinducedbyinfectionwithG.cichoracearumandATL9
or higher than in Col-0 (inset Figure 2A). From these data it
function is required for basal defense against this biotrophic
appears that the induction of ATL9 expression by chitin is not
pathogen.Interestingly,ATL9expressionappearstobedependent
dependent ontheSA andJA signaling pathways.
on NADPH oxidases and mutations in ATL9 lead to an
To resolve the contributions of each signaling pathway to
impairment in the ability of plants to produce reactive oxygen
overalllevelsofATL9expression,wetreatedwildtypeplantswith
species(ROS)afterinfection.Expressionprofilingofatl9revealed
chitin,SA,1-aminocyclopropane-1-carboxylicacid(ACC)andJA
a complex interplay between chitin-mediated signaling and other
[30,31]. Expression of ATL9 and marker genes associated with
defense pathways.
each pathway (SA: PR1; JA- and Ethylene: PDF1.2) were
quantified 30 minutes after treatment. Only treatment with
Results
crab-shell chitin (CSC) induced expression of ATL9 (Figure 2B).
ATL9 (Arabidopsis to´xicos en levadura 9) encodes an E3 ExposureofplantstoACC,JAandSAresultedinnoappreciable
ubiquitin ligase with homology to a family of genes expression of the gene (Figure 2B) providing evidence that ATL9
induction is uniquely associated with chitin-mediated defense
induced by wounding and abiotic stress
responses. As expected, marker genes associated with each of the
Previous studies by our group have shown that mutants in the
classical defense pathways were induced by their corresponding
gene At2g35000 were more susceptible to fungal infection than
signalingmolecule(Figure2B).TheseresultssuggestthattheSA-,
wild-typeplants[6].At2g35000belongstotheATLfamily[18,19]
ethylene- and JA-mediated defense pathways do not directly
ofRING(reallyinterestingnewgene)zinc-fingerproteinsandwas
regulate ATL9 expression but might feedback on the chitin-
designatedasATL9inapreviousreview[12].TheATL9protein
induced expression of ATL9.
consists of 378 amino acids and contains an N-terminal signal
peptide;twopredictedtransmembranedomains,aC3HC4RING
ATL9 is involved in plant defense against G.
zinc-fingerdomain,aPESTdomainandaC-terminalcoiledcoil
region (Figure 1A).Three members of theArabidopsis ATL gene cichoracearum
family, ATL2, ATL6 and ATL17, are presumed to play a role in Mutations in ATL9 rendered plants more susceptible to
defense although their precise functions are unknown at present infection by the biotrophic fungus G. cichoracearum compared to
[21,25]. Using database searches we identified a total of eight thewildtypeplants.Tomorepreciselyquantifythesusceptibility
proteins with a high percentage of homologyto ATL9, including of atl9 mutants to G. cichoracearum, inoculations were performed
severalATLsinotherplantspeciessuchasOryzasativa(EL5)[26], using three independent atl9 T-DNA insertional lines, 35S:ATL9
Nicotiana tabacum (Avr9) [27] or Poplar (PtaRHE1) [28]. An plants, Col-0, the Kas-1 Arabidopsis accession and sid2-1 plants.
alignmentoftheeightproteinsequencesalongwiththeirputative Counts of mature conidiophores 6 dpi showed that all three atl9
function in defense is shown in Figure 1B. Of the eight proteins, mutants (atl9-1, atl9-2, atl9-3) supportedmore than twotimesthe
PLoSONE | www.plosone.org 2 December2010 | Volume 5 | Issue 12 | e14426
ATL99sRoleinPlantDefense
Figure1.ATL9structureandsequencealignmentbetweenATLfamilymembers.A)SchematicdiagramofATL9proteinstructure.The
proteinhasanNterminalregion(cytosolic),containingasignalpeptide(gray);twotransmembranedomains(Tm1andTm2,inblack);andaC3HC3
RINGZinc-fingerdomain(darkgray),aPESTsequenceandacoiledcoildomainthatextendsintotheERlumen(white).B)AlignmentofATL9RING
Zinc-fingerdomainaminoacidsequencewiththeotherATLfamilymembersimplicatedindefenseresponsesinplants.Asterisksindicateconserved
cysteines;conservedaminoacidsareindicatedinblack,residuesconservedinmorethanan87.5%arehighlightedingray.Theconsensussequence
for this group of RING zinc-fingers is: C-X2-CL-X-E-X7-R-X2-P-X-C-X-H-X-FH-X2-C-X-D-X-W-X6-CP-X-C, where X is any amino acid. Induction of the
correspondinggenesisindicatedas:C(chitin),F(flagellin),P(pathogens),Av(Avr9),BTH(benzothaidiazole),ACC(1-aminocyclopropane-1-carboxylic
acid),CWP(cellwallproteinfractionelicitor),(+)indicatesmutantsinthesegeneshavebeentestedforalteredpathogenresponse.
doi:10.1371/journal.pone.0014426.g001
number of mature conidiophores per fungal colony compared to at points where the fungus was attempting to penetrate the
wild type (Figure 3A). The atl9-1, atl9-2 and atl9-3 mutations all epidermal cells (Arrows, Figure 3B). DAB staining of H O was
2 2
resulted in a more pronounced susceptibility phenotype than not visible in AtrbohD/F or the sid2-1 mutant (Figure 3B).
observed in wild type plants. All plants tested were more Hydrogen peroxide was detected both at the point of fungal
susceptiblethantheresistantKas-1ecotypeandcomplementation penetration aswellasinsidesomeepidermalcells inCol-0 plants
oftheatl9mutationrestoredthewild-typedefensephenotype(data (Figure3B).Thehighestlevelsofhydrogenperoxideweredetected
not shown). Transformed plants containing an over-expression intheresistantKas-1ecotype(Figure3B)withthehighestlevelsof
construct of ATL9 were significantly more resistant to G. H2O2 localizing to points of fungal penetration and in the
cichoracearum and showed a 2-fold decrease in conidiophores per epidermal cells directly adjacent to penetrated cells. High H2O2
colony compared towild-type plants (Figure3A). levelsinKas-1correlatedwithastronginhibitionoffungalgrowth
Since previously published microarray data (NASC array) and colony development (Figure 3B). In all cases there was good
indicated that expression of ATL9 in roots was compromised in inverse correlation between the conidiophores per colony
the NADPH oxidase AtrbohC mutant, we tested the defense (Figure 3A) and the levels of H2O2 detected in leaves via DAB
phenotypeofmutantsintheNADPHoxidasesAtrbohDandAtrbohF staining(Figure3C).TakentogetherthesedataindicatethatROS
whichareexpressedinleavesandinthedoublemutantAtrbohD/ production in epidermal cells is necessary for effective defense
FagainstG.cichoracearum(Figure3A).AtrbohD,AtrbohFandtheSA- againstG.cichoracearumandthisROSproductionisimpairedinthe
compromised sid2-1 mutant showed comparable numbers of atl9 mutant, AtrbohD and AtrbohF. This provides further evidence
conidiophores per colony and were significantly more susceptible that functional ATL9, AtrbohD and AtrbohF expression are needed
than atl9, the 35S:ATL9 line, Col-0 or the resistant Kas-1 foreffective defenses againstpowdery mildew.
accession.TheAtrbohD/Fdoublemutantwasthemostsusceptible
to powdery mildew even surpassing that of sid2-1 suggesting an BasalexpressionofATL9andinductionbychitindepends
additive effectof theindividual mutations(Figure 3A). on NADPH oxidase activity
BecausetheROSresponseisessentialfordiseaseresistancewe The production of reactive oxygen species (ROS) is a key
decidedtodetermineifROSproductionwascompromisedinatl9 characteristicoftheinitialdefenseresponseofplantstopathogen
mutantandintheAtrbohmutants(Arabidopsisthalianarespiratory attack and chitin treatment is known to elicit ROS production
burst oxidase homolog). Infected leaves of each mutant were in roots although not in leaves [32]. Several NADPH oxidases
double stained with diaminobenzidine (DAB) to visualize H O in Arabidopsis have been identified and their roles in ROS
2 2
andwithtrypanbluetoobservethespreadofthefungus7dpiwith generation and plant defense characterized [33,34,35]. A search
powdery mildew (Figure 3B). Consistent with our hypothesis, of publically available microarray data (http://affymetrix.
H O production was much lower in atl9, compared to that of arabidopsis.info/narrays/geneswinger.pl) revealedthat expression
2 2
Col-0orKas-1plants(Figure3B).H O localizationwasfocused ofATL9inrootswasrepressedby50%intheAtrbohCmutant[36]
2 2
PLoSONE | www.plosone.org 3 December2010 | Volume 5 | Issue 12 | e14426
ATL99sRoleinPlantDefense
Figure2.InfluenceofSA-,JA-andET-mediatedpathwaysonATL9expressionmeasuredbyqRT-PCR.A)InductionofATL930minutes
afterchitintreatmentinCol-0wild-typeplantsandintheein2-5,jar1andsid2-1mutants.B)InductionofATL9(black),PR1(lightgray)andPDF1.2
(darkgray)inCol-0plantsaftertreatmentwithchitin(CSC),50mMACC,5mMJAor0.5mMSA.Datarepresenttheratiobetweenmockandtreated
plantsofthreeindependentbiologicalsamples.TheinsetshowsrelativeexpressionofATL9transcriptcomparedtotheactincontrol(RE/Actin).
doi:10.1371/journal.pone.0014426.g002
compared to wild-type plants. AtrbohC plants are also more similarpatternofreducedinductionandreducedbasalexpression
susceptibletothepowderymildewpathogenG.cichoracearum(data wasobservedfortheMAPK3geneinallthelinestested(Figure4
notshown).Thesedataraisedthepossibilitythattheinductionof and inset). These results indicate that both basal transcription
ATL9bychitinmightbedependentonNADPHoxidaseactivity. levelsofATL9aswellastheinductionofATL9bychitindepend
TotestthishypothesiswetreatedCol-0plantsandthreemutants onNADPHoxidaseactivity.Theseresultsareinlinewithprevious
impaired in NADPH activity (AtrbohD, AtrbohF, AtrbohD/F) with NASC array data showing that ATL9 transcription levels are
chitin and analyzed the transcript levels of ATL9 by qRT-PCR impairedin theAtrbohC mutant.
(Figure 4). The MAPK3 gene was used as a control since it is
knowntobeinducedbychitin[29].ThoughATL9wasinducedby
ATL9 is induced by G. cichoracearum and has a complex
chitininCol-0plants,itsinductionwassignificantlyreducedinthe
three Atrboh mutants (Figure 4). ATL9 was still expressed in pattern of expression post-inoculation
untreatedAtrbohmutantsbutathalfofthatobservedinuntreated Since atl9 mutants are more susceptible to G. cichoracearum
wild-type plants (inset Figure 4). These data indicate that basal (Figure3A) thanwild-type plantsandATL9basal expressionand
expression levels of ATL9 are dependent on AtrbohD and F.. A induction by chitin is dependent on NADPH oxidases (Figure 4
PLoSONE | www.plosone.org 4 December2010 | Volume 5 | Issue 12 | e14426
ATL99sRoleinPlantDefense
Figure 3. Susceptibility ofatl9mutants toGolovinomycetescichoracearumand hydrogen peroxide production. A) Quantification of
GolovinomycescichoracearumgrowthonCol-0plants,threedifferentatl9T-DNAinsertionalmutants(atl9-1-3),35S:ATL9,Kas-1,AtrbohD,AtrbohF,
AtrbohD/FandtheSA-compromisedmutantsid2-1.Thenumberofconidiophorespercolony(c/c)werecounted6dpi.Inoculationswereperformed
atalowdensity(5–10conidia/mm2;n=36plants).Datavaluesrepresentoneofatleastthreeindependentexperimentswithsimilarresults.B)
ProductionofhydrogenperoxideonplantsinfectedwithG.cichoracearum.Clearedleavesofinfectedplantswerestainedwithtrypanbluewhich
stainsthefungalhyphaeandconidiophoresontheleafsurface.SecondarystainingwithDABwasperformedtoindicateareasofhydrogenperoxide
production(indicatedbyperoxidaseactivity).a,Kas-1,b,Col-0,c,atl9-1,d,AtrbohD,f,AtrbohD/F,g,sid2-1.Arrowsindicatefungalpenetrationpoints
inepidermalcellswithdetectedhydrogenperoxideproduction.Theimagesweretakenatafinalmagnificationof40x.Bars.-5mm.C)Quantification
of hydrogen peroxide production in mutants. The intensity of DAB staining was quantified at each inoculated plant and compared to the
correspondinguninoculatedcontroltoobtainarelativeDABsignal.ThequantificationwasperformedwiththeaidoftheprogramImageJ.
doi:10.1371/journal.pone.0014426.g003
andinset), wewere interestedindetermining thetimingof ATL9 were monitored using qRT-PCR at several time points after
expression after fungal infection. Expression levels of ATL9, infection with powdery mildew [37,38]. Wild-type plants were
AtrbohD, AtrbohF, and the defense marker genes PR1 and PDF1.2 heavily inoculated with G. cichoracearum and tissue samples were
PLoSONE | www.plosone.org 5 December2010 | Volume 5 | Issue 12 | e14426
ATL99sRoleinPlantDefense
Figure4.ThebasalexpressionofATL9anditsinductionbychitindependsonNADPHoxidases.Plantsweretreatedwithchitinfor30
minutesandharvestedforanalysis.ExpressionlevelsofATL9weremeasuredbyqRT-PCR.TheinsetshowstherelativeexpressionofATL9transcript
compared to actin (RE/Actin) in control plants. Data represent the ratio of expression between mock and treated plants of three independent
biologicalsamples.
doi:10.1371/journal.pone.0014426.g004
takenatfourearlytimepoints:1 hr.,1.5hr.,2hr.and4 hr.post- construct was made containing a C-terminal fusion of the green
infectionandattwolateinfectiontimepoints;24hr.and48hours fluorescentprotein(GFP)totheATL9geneunderthecontrolofits
post-infection (hpi). Quantitative analysis of the gene expression native promoter (ATL9p:ATL9:GFP). Stable transgenic lines were
revealed that ATL9 is induced very early in the infection process created by transforming Col-0 plants with Agrobacterium tumefaciens
(1 hr.;Figure5)andthenexpressionquicklydropsofftonegligible containing the ATL9p:ATL9:GFP construct. Upon visualization,
levels at 4 hpi in a pattern similar to that of PR1 expression fluorescentsignalwasdetectedintheendoplasmicreticulum(ER)
(Figure5).ATL9expressionagainroseinthelatertimepoints(24 of leaf epidermal cells (Figure 7A–C). Localization of ATL9 was
and 48 hpi). By 24 hpi, ATL9 expression was again induced to observedintheERofalltissuestestedandwasparticularlyhighin
levelssimilartothoseobservedinthe1 hourtimepoint(Figure5). roots(Figure7A–C;datanotshown).Formorepreciselocalization
AtrbohD,AtrbohF,andPDF1.2werenotsignificantly induced early asecondconstructcontainingafusionofGFPtotheC-terminusof
in the infection process but were present at higher levels in the ATL9 driven by the CaMV 35S promoter (35S:ATL9:GFP) was
later stages of theinfection beginning at 24hpi (Figure 5).These
datamayindicatethatanearlystepinpathogenrecognitionorthe
infection process is critical for activation of AtrbohD, AtrbohF, and
PDF1.2.
ATL9 has ubiquitin ligase activity
TheRINGzinc-finger domainisknowntobeessential forthe
functionofubiquitinE3ligases[39]andtwoothermembersofthe
ATLfamily,ATL2andATL6,havebeenidentifiedasE3ligases
in Arabidopsis [18,21]. In order to determine whether ATL9
might function as an E3 ligase, we analyzed ATL9 activity using
an in vitro ubiquitination assay (Figure 6). A GST:ATL9 fusion
proteinwasaffinitypurifiedanditsE3ligaseactivitywasassayedin
vitro.Multipleformsofubiquitinatedproteinsweredetectedinthe
completeassaybytheanti-ubiquitinantibodies(Figure6,lane1).
TheomissionofAtUBC8(Figure6,lane2),GST-ATL9(Figure6,
lane3)orubiquitin(Figure6,lane4)fromtheassayresultedina
lossofproteinpolyubiquitination(Figure6,lane1)confirmingthat
ATL9 has ubiquitinligase activity.
Figure 5. Relative expression levels of selected genes at
The ATL9 protein is localized to the Endoplasmic various time points after inoculation with Golovinomyces
cichoracearum. Col-0 plants were inoculated with G. cichoracearum
Reticulum (ER)
athighconcentration.qRT-PCRdatarepresenttheratiobetweenmock
Defense-related proteins can be found in a diverse array of andtreatedplantsofthreeindependentbiologicalsamples.
cellular compartments.TodeterminethelocalizationofATL9,a doi:10.1371/journal.pone.0014426.g005
PLoSONE | www.plosone.org 6 December2010 | Volume 5 | Issue 12 | e14426
ATL99sRoleinPlantDefense
detectedregardlessofthetimeaftertreatmentwithchitin(datanot
shown).
Microarray expression analysis of atl9
To identify genes whose expression is dependent on ATL9
function,Col-0plantsandatl9mutantsweretreatedwithchitinfor
30 minutes and their expression profiles were examined using
AffymetrixATH1arrays.Ofthe22,677genesrepresented,16,530
were defined as present and of these 4,375 genes showed altered
expression in the atl9 mutant compared to wild-type plants
(Figure 8A and 8B). Genes were divided into (1) those that
exhibitedgenotype-specificdifferencesinexpressionbetweenwild
typeandatl9and(2)thosethatrespondedtotreatmentwithchitin.
Figure 8Bshowsthat themajority ofgenes that exhibitedaltered
expression responded to chitin treatment regardless of plant
genotype(3510),whileasmallernumber(525)fluctuatedbetween
wild-typeplantsandtheatl9mutantirrespectiveoftreatment.An
intermediatenumberofgenes(340)werechangedbothbytheloss
ofATL9expressionandbychitintreatment(Figure8B).Dataon
selected genes withineachcategory are giveninTable S1.
Severalgeneswithknownrolesinplantdefensewereidentifiedthat
were specifically altered between wild-type plants and atl9 mutants
(Table S1). Two genes associated with SA-mediated defense
responses, PR-1 and SID2/ICS1 were repressed in plants lacking
ATL9(TableS1).PR-1levels(At2g14580)weredecreased3-foldinthe
atl9 mutant compared with Col-0 plants. This finding is surprising
sincewehaveshownthatPR-1geneexpressionisinducedbyinfection
withG.cichoracearumbutnotbychitin-treatmentalone(Figure5,Table
Figure6.ATL9iscapableofmediatingproteinubiquitination S1). Expression of SID2/ICS1 (isochorismate synthase; At1g74710)
in a E2-dependent manner. The complete in vitro ubiquitination wasslightlydown-regulatedintheatl9mutant(0.77-fold)comparedto
assay(lane1)containedrecombinantyeastE1enzyme,recombinant6x Col-0 though in both cases the gene was induced by chitin. PEN3
histidine-tagged Arabidopsis E2 enzyme UBC8, GST-tagged ATL9 and
expression was reduced by 50% in the atl9 mutants in contrast to
ubiquitin.OmissionofAtUBC8(lane2),GST-ATL9(lane3)orubiquitin
PEN1 and PEN2 whose expression was unchanged (Table S1).
(lane4)fromtheassayresultedinalossofproteinpolyubiquitinationas
indicated by a lack of the ubiquitinated proteins compared to the ExpressionofPDF2.5,anantimicrobialpeptide,wasslightlylowerin
complete assay (lane 1). Ubiquitinated proteins were visualized by atl9 plants compared to Col-0. However, when treated with chitin
western blot analysis using anti-ubiquitin antibodies. * Indicate non- PDF2.5 was strongly induced in wild-type plants but its expression
specific cross-reactive proteins. ¤ Represent AtUBC8-Ub(n). Molecular remained unchanged in the atl9 mutant (Table S1). Similarly the
weightmarkers(kDa)areshowntotheleftoftheblot.
PCC1gene(pathogenandcircadiancontrolled1)wasdown-regulated
doi:10.1371/journal.pone.0014426.g006
inresponsetothelossofATL9.AsinthecaseofPDF2.5,PCC1was
stronglyup-regulatedinCol-0plantsuponchitintreatmentandwas
co-bombarded into onion epidermal cells with a known ER down-regulatedintheatl9mutant(TableS1).
luminal marker ER-rk (signal peptide of AtWAK2:mCherry:H- Most of the defense genes surveyed were not significantly
DEL) [40]. When visualized with fluorescent microscopy, affected by the loss of atl9 expression and exhibited similar
35S:ATL9:GFP co-localized with the ER-rk marker in the onion responses to chitin-treated control plants (Table S1). However,
epidermal cells confirming its localization to the ER (Figure 7D– several defense genes had enhanced expression levels after chitin
F). In parallel we also transiently expressed an N-terminal treatment in the atl9 mutant including PEN1, PAD4, two disease
construct of ATL9, 35S:GFP:ATL9, in tobacco with a 35S:GFP resistance-likeproteins,ACRE1b,ATL17andproteinphosphatase
construct as an internal negative control for GFP localization to 2C (Table S1). Interestingly, AtrbohD expression was induced to
the nucleus. Again, ATL9 was expressed transiently in the ER of higher levels in chitin-treated atl9 compared with wild-type
tobaccoepidermalcellsconfirmingourresultsinArabidopsisand controls. Similar increases were not observed in the expression
onion (Figure S1A). Fluorescence was detectable within the of AtrbohC orAtrbohF.
nucleus in the tobacco cells expressing the 35S:GFP construct
and the GFP signal could be seen in the ER surrounding the Defense genes associated with elicitor-mediated
nuclear membrane (Figure S1B). A western blot with the protein responses
extract from the transiently transformed tobacco tissue was Toidentifydefense-relatedgenesthatwerespecificallylinkedto
performed and hybridized with anti-GFP antibody to confirm thechitinsignalingpathway,wecomparedtheexpressionofgenes
thesize of theATL9and GFPfusion proteins(Figure S1C). knowntobeinducedbybothabioticandbioticstressesinchitin-
SincesomePAMPresponserelatedproteinsaretransportedto treated Col-0 wild type and atl9 plants (Table S1). We first
the vesicle after stress responses [41], we wanted to determine if analyzedtheexpressionlevelsoftheATL9familymembersATL2-
ATL9 responded in a similar manner. To determine if ATL9 ACRE132, ATL6 and ATL17 (Table S1). In all cases, the ATL
localization changed after elicitor treatment, transgenic plants familymemberswereinducedatsignificantlyhigherlevelsinwild-
expressingATL9p:ATL9:GFPweretreatedwithchitin(100 mg/L) typeplants,comparedtoatl9mutants.Thisstronglysuggeststhat
by direct infiltration of the leaves and observed at five minute ATL9isimportantininitiatingtheexpressionofgenesinvolvedin
intervals for one hour. No change in ATL9 localization was chitin-mediatedresponses.Wewerealsointerestedindetermining
PLoSONE | www.plosone.org 7 December2010 | Volume 5 | Issue 12 | e14426
ATL99sRoleinPlantDefense
Figure 7. The ATL9 protein localizes to the ER. A–C) Confocal images of leaf epidermal cells in transgenic Arabidopsis plants expressing
ATL9p:ATL9:GFPshowingproteinlocalizationtotheendoplasmicreticulum.B)ConfocalimageofaleaftrichomeshowingATL9p:ATL9:GFPlocalizing
totheER.D–F)Co-localizationofthe35S:ATL9:GFPfusionandER-rkmarker.Onionepidermalcellswereco-bombardedwith35S:ATL9:GFPandER-rk
andvisualizedusingfluorescencemicroscopy.D)35S:ATL9:GFP.E)ER-rk(mCherry).F)GFP/mCherryoverlay.Bars:3mm(A,B),1mm(C),10mm(D-F).
doi:10.1371/journal.pone.0014426.g007
the responses of the known elicitor receptors, CERK1 (LysM lipoxygenase-3 (LOX3) gene, which is involved in octadecanoid
RLK1),FLS2,BAK1andEFRintheatl9mutant.Theinduction biosynthesis leading totheproductionof JA inplants,was highly
of the gene encoding the chitin receptor LysM RLK1 was induced inboth atl9andCol-0 plants by chitin.
essentially unchanged in the atl9 mutant and wild-type plants.
Surprisingly,expressionlevelsoftheflagellinreceptorgene,FLS2 Discussion
were strongly induced by chitin in Col-0 plants but not in the
mutant (Table S1), suggesting that induction of FLS2 may be ATL9 is an E3 ubiquitin ligase that is integral to defense
dependentonATL9expression.ExpressionofBAK1whichactsas against fungal pathogens
aco-receptorofflagellinwithFLS2[9,42]andthebacterialEF-Tu Recent work has highlighted the ubiquitin-proteasome system
receptor EFR [7] were both induced by chitin treatment but at (UPS) and its associated E3 ubiquitin ligases as regulators of the
similar levels in both the atl9 mutant and in wild-type plants. In plant defense response and it is clear that these proteins play an
addition,ACRE1b,a genehighlyinduced bytreatment withflg22 important part in disease resistance [12,44]. In the current study
[5] was also strongly induced by chitin treatment in wild-type we have shown that ATL9 is a RING-type E3 ubiquitin ligase
plants. Taken together these results indicate that activation of stronglyinducedbychitin.TheATLgenefamilyencodesagroup
elicitor-mediated signaling by chitin and flagellin are closely of proteins that share three specific characteristics: 1) rapid
intertwined inplants. induction(,1 hour)afterelicitortreatment,2)ahighlyconserved
To better understand theinteractions between chitin-mediated RING-H2zinc-fingerdomainwithatleastsixcysteinesandthree
defensesignalingandsignalingmediatedthroughSA,JAandET histidines conserved and 3) at least one amino-terminal trans-
theexpressionlevelsofkeygenesineachofthesignaltransduction membrane domain. Most members of the ATL gene family are
pathways were monitored (Table S1). The induction of MAPK3 predicted to function as single subunit E3 ubiquitin ligases with
and MAPK5, both known to be important in signaling pathways seventeen members of the ATL family known to be expressed in
initiatedbyflagellin[43]wereverysimilarinatl9andCol-0plants Arabidopsis [25]. Although determination of a common function
indicatingthattheseMAPkinasesareupstreamofATL9activity for them is still in progress, mutations in members of the ATL
orare workinginan independentsignal transduction network. A family have been shown to have an altered defense response to
similar induction by chitin in both the atl9 mutant and wild-type pathogens [25]. The ATL2 gene was shown to be specifically
wasalsoobservedforMAPKK4.OthergenesinvolvedinSA-,JA- induced by chitin but not by other elicitors of classic defense
and ET-mediated signaling (RAR1, SGT1, HSP90, COI1, ETR1, pathways [18]. Constitutive over-expression mutants of ATL2
EIN3,PAD2,JAR1orSAG101)showednodifferencesinexpression inducedhighlevelsofpathogen-relatedgenessuchasNPR1-1and
levels in either chitin-treated control or atl9 plants or in mock- the phenylpropanoid biosynthetic enzymes phenylalanine ammo-
treated plants pointing to the fact that these genes are not nia lyase and chalcone synthase [21,45]. The EL5 gene in rice is
influenced by ATL9 expression. However, the expression of the also a member of the ATL family and is rapidly and transiently
PLoSONE | www.plosone.org 8 December2010 | Volume 5 | Issue 12 | e14426
ATL99sRoleinPlantDefense
Figure 8. Microarray analysis ofatl9mutant after chitin treatment. A) Hierarchical cluster of ratio values (relative to the water control
treatment)of16,530genesanalyzedinCol-0wild-typeandatl9plantstreatedwithcrabshellchitin(CSC)for30minutes.Eachgeneisrepresentedby
asinglerowandeachcolumnrepresentsanindividualtreatment.Redrepresentsup-regulatedgenes;blue,down-regulatedgenes;andyellow,genes
withnochangeinexpression.Groupsofgenesexpresseddifferentiallyaredelineatedwithrectangles.Threegenes(blackarrows)werechangingdue
tointeractionsbetweentreatmentandgenotypeandonegene(redarrow)wasgenotypeinteractionspecific.B)Venndiagramrepresentationof
resultsfromhierarchicalclustering.Atotalof4,375genesweredifferentiallyexpressedbetweengenotypesandtreatments.Thestatisticusedfor
clusteringwastwo-wayANOVA.Genotypeandtreatmentgroupswereanalyzedusingap-valueof0.5withp-value.0.5= notsignificant;p-value
,0.5= significant.
doi:10.1371/journal.pone.0014426.g008
induced by chitin [23,26]. EL5, like ATL9, is an E3 ubiquitin OurdatashowthatATL9isnecessaryforresistanceagainstG.
ligase and is hypothesized to play a role in defense responses cichoracearum (Figure 3A). ATL9 transcription levels appear to be
through protein turnover via the UPS [22]. Additionally, the tightlyregulated;afterchitintreatmentthegeneisinducedwithin
ACRE-132gene,anATLproteinfromtobaccohasbeenshownto 30 minutes [6] and in plants infected with G. cichoracearum ATL9
be induced during the defense response mediated by the expression can be detected in less than one hour by qRT-PCR
interaction between Avr9 and Cf9 in the response to the fungal (Figure 5). We hypothesize that ATL9 expression is also down-
pathogen Cladosporium fulvum[46]. regulated quickly (,1 hour) with ATL9 binding its target for
ATL9 is unique among ATL family members that have been eliminationviatheUPSfollowedbyrapidATL9degradationvia
characterizedsinceitcontainsaPEST(Pro-Glu-Ser-Thr)domain. itsPESTdomain.Over-expression ofATL9producedplantsthat
PESTdomainsarecommoninproteinsthatarerapidlydegraded weremoreresistantthanwild-type(Figure3A)suggestingthatthe
in the cell [47] and are suggested to serve as proteolytic signals gene is degraded rapidly under normal conditions in wild type
[48] involved in ubiquitination and internalization of proteins. plantspost-infection.Furtherexperimentswithdeletionconstructs
While few studies in plants have examined the role of PEST in the PEST domain of ATL9 will be useful in characterizing
domains and the UPS in defense, there are numerous studies in ATL99sregulation andinconfirming ourhypothesis regarding its
mammals describing PEST domains and their role in defense degradation.
against cellular pathogens [49].Thepresence ofa PEST domain
in the ATL9 protein may imply that there are multiple ways to ATL9 appears to be involved in ER stress responses and
control its expression and that ATL9 may play a critical role in ERAD during innate immunity
plantdefensesimilartothatoftheMclfamilyofgenesinhumans In2003,Takemotoetal.showedcytoplasmicaggregationand
whicharenecessaryforcellsurvivalintheimmunesystem[50].In accumulation of ER and Golgi bodies around the sites of fungal
plants, the tomato Ve gene contains a PEST domain that is penetration [52]. These cellular rearrangements suggest that the
necessary fordefense againstVerticillim species [51]. production and secretion ofplantmaterials areactivatedaround
PLoSONE | www.plosone.org 9 December2010 | Volume 5 | Issue 12 | e14426
ATL99sRoleinPlantDefense
sites of infection or penetration. Transgenic plants expressing an haveshownthatAtrbohDisresponsibleforreactiveoxygenspecies
ATL9p:ATL9:GFP fusion construct showed that ATL9 was produced after infection with avirulent bacteria or oomycete
localized to the ER membrane (Figure 7A–C). Localization of pathogens whereas AtrbohF is integral to the regulation of the
the ATL9 protein at the ER membrane not only allows the hypersensitiveresponse[61].Inthecurrentstudy,wedemonstrate
protein to remain close to the site of infection/penetration but thattheinductionofATL9bychitinisdependentonbothAtrbohD
also places the protein where it can be quickly degraded after and AtrbohF and that MAPK3 activation is also dependent on
carrying out its E3 activity. We hypothesize that ATL9 may Atrboh expression. These results suggest that the early steps in
attach ubiquitin to either plant- or fungal-derived proteins that chitin recognition are reliant on Atrboh functionality (Figure 3).
areactivenearsitesofinfection.Severalimportantgenesinplant OurdataalsodemonstratethatbothAtrbohDandAtrbohFmutants
defense have been shown to localize to the ER. The Cf-9 are more susceptible to G. cichoracearum than wild type plants
resistance gene, which confers resistance to Cladosporium fulvum (Figure 3A–B), indicating that Atrboh activity is necessary for
racesexpressingtheAvr-9avirulencegene,containsaC-terminal defense against powdery mildew.
di-lysine motif (KKXX) targeting Cf-9 to the ER [53]. ACRE- Recent work [62] has demonstrated that Atrboh genes can
132, also a member of the ATL family, has been shown to be modulateandantagonizeSA-dependentcelldeathsignals.ROS
induced during Avr-9/Cf-9 interaction and may be an active producedbyAtrbohDlimitedthespreadofthecelldeathinduced
player in that response [46]. The barley HSP90 protein is bythebacterialpathogenP.syringaeDC3000andthenecrotroph
localized to the ER plasma membrane like ATL9 [54] and has B. cinerea. Our current study clearly shows that both ArbohD,
been shown to interact with RAR1 and SGT1, two genes AtrbohF mutants were more susceptible than wild type to G.
involved in R-gene mediated resistance to fungi, in yeast two- cichoracearumandthosequantitativeresultsaredirectlycorrelated
hybrid experiments.HSP90, RAR1 and SGT1 are hypothesized with the low levels of hydrogen peroxide accumulation at the
to form a chaperone complex that mediates the folding of R- point(s) of penetration detected in these mutants (Figure 3B–C).
proteinsandtheirincorporationintofunctionalcomplexes.Data HydrogenperoxideproductioninAtrbohD,AtrbohF,andAtrbohD/
supporting the role of an HSP90-RAR1-SGT1 complex as a Faswellasintheatl9mutantswasessentiallyhalfofthatinwild-
chaperone active in R-gene mediated defenses have been shown type plants. Interestingly the Kas-1 accession [63] that is
in both barley and Arabidopsis [44,55,56]. Though our resistant to G. cichoracearum had increased levels of peroxide
understanding of the role of ATL family members in defense is (Figure 3C) at sites of attempted penetration (Figure 3B)
incomplete,itisintriguingtospeculatethatATL9orotherfamily although no cell death was observed in these cells and plants
members may form complexes similar to HSP90-RAR1-SGT1 were phenotypically resistant. These data imply that careful
that are necessary for defense against G. cichoracearum. regulation of hydrogen peroxide production in Kas-1 plants is
E3 ligases in the ER are often associated with endoplasmic sufficient to inhibit fungal penetration without cell death [62]
reticulum associated degradation (ERAD) a process involved in suggesting that salicylic acid may activate defense signaling in
targeting misfolded proteins for ubiquitination and subsequent cells that are spatially removed from infection sites without
degradation by the proteasome [57,58]. In the ER, the ATL9 activating cell death. Torres et al. [62] further hypothesize that
proteinwouldhaveitsRINGdomainandPESTdomainexposed Atrbohs are necessary for the suppression of unwanted cell death
totheERlumen.WehypothesizethatATL9actstoubiquitinatea in cells where increased levels of SA are required to activate
negativeregulatorofplant defense responsesintheERlumenor defenseresponses.G.cichoracearumresistanceinKas-1plantsmay
someunknownluminalproteinthatisnotproperlyfoldedduring bedependent onAtrboh’s abilitytoinducehigherlevelsofH O
2 2
stress-induced ERAD. The ATL9 protein shares homology with production at sites where the fungus is attempting to penetrate
the well-characterized ERAD E39s Hrd1 and gp78 [59]. Both epidermal cells while simultaneously repressing cell death in
Hrd1andgp78areknowntobeinducedduringcellularstressand those same cells. This hypothesis is supported by the fact that
participate in the unfolded protein response. The yeast Hrd1 Atrbohmutantshadvirtuallynoperoxidaseactivityatpenetration
proteinisatthecenterofalargeproteincomplexinvolvedinthe sites and were much more susceptible to G. cichoracearum
ubiquitylation of ER luminal and membrane proteins. Since (Figure3A–B).Althoughaconnectionbetweenthe Atrbohgenes,
ATL9isanERmembraneresidentproteinwhoseRINGdomain chitin induction of ATL9 and resistance to G. cichoracearum
projectsintotheERlumen,itisinterestingtospeculatethatATL9 remains to be definitively established, our results clearly show
may also function in a group of proteins similar to the Hrd1 that NADPH oxidase activity is required for ATL9 induction
complex that is involved in ERAD induced by biotic stresses. after chitin recognition. This is in line with our results showing
Further experiments directed towards identifying both ATL9 thatthe expressionlevelsofATL9areimpaired inbothmutants
interacting partners and its’ protein target will be helpful in compared with the wild type plants (inset Figure 4B),as in root
understanding itspreciserole. tissues of AtrbohC. The activity of AtrbohD, AtrbohF and ATL9
are needed for effective defense against powdery mildew. This
Expression and chitin ATL9 gene response depends on study provides evidence for the first time that AtrbohD and
NADPH oxidases AtrbohFactivity,inplantleaves,arerequiredforG.cichoracearum
resistance and for some chitin-elicited defense responses.
The oxidative burst andtheproduction of ROSisan essential
component of plant defense. This pathway is mediated by the
activation of the NADPH complex which is necessary for the PossiblemodeofactionofATL9inchitinrecognitionand
production of H O . Following H O production, several chitin-mediated innate immunity
2 2 2 2
downstream signaling events are initiated including calcium We hypothesize that upon chitin recognition/fungal infection
mobilization to the cytoplasm, protein phosphorylation, MAP intracellular calcium levels increase leading to the activation of
kinase activation and defense gene expression [53,60]. In Arabidopsis NADPH oxidases (Figure 9). The subsequent
Arabidopsis, the NADPH family of oxidases consists of ten production of ROS and activation of MAPK signaling cascades
members (AtrbohA –AtrbohJ) which are homologs of gp9phox, a induce ATL9 expression and its transport to the ER membrane.
subunit of the mammalian NADPH oxidase complex [33]. Here ATL9, either alone or working in a complex would
Infection studies of Arabidopsis mutants in AtrbohD and AtrbohF ubiquitinate its target protein(s) for degradation and appropriate
PLoSONE | www.plosone.org 10 December2010 | Volume 5 | Issue 12 | e14426