Specific patterns of changes in wheat gene expression after treatment with three antifungal ...

15 Pages·2005·0.31 MB·English

PlantMolecularBiology (2005)57:693–707 (cid:1)Springer2005
DOI10.1007/s11103-005-1728-y
Specific patterns of changes in wheat gene expression after treatment
with three antifungal compounds
Fre´de´rique Pasquer1, Edwige Isidore1, Ju¨ rg Zarn2 and Beat Keller1,*
1Institute of Plant Biology, University of Zu¨rich, Zollikerstrasse 107, CH8008 Zu¨rich, Switzerland (*author
for correspondence; e-mail [email protected]); 2Swiss Federal Office of Public Health, Stauffac-
herstrasse 101, CH8004 Zu¨rich, Switzerland
Received5July2004;acceptedinrevisedform3February2005
Key words: cDNA microarray, defence response, gene expression pattern, plant protection products,
Triticum aestivum
Abstract
Thetwofungicidesazoxystrobinandfenpropimorphareusedagainstpowderymildewandrustdiseasesin
wheat (Triticum aestivum L). Azoxystrobin, a strobilurin, inhibits fungal mitochondrial respiration and
fenpropimorph, a morpholin, represses biosynthesis of ergosterol, the major sterol of fungal membranes.
Although the fungitoxic activity of these compounds is well understood, their effects on plant metabolism
remain unclear. In contrast to the fungicides which directly affect pathogen metabolism, benzo(1,2,3)
thiadiazole-7-carbothioic acid S-methylester (BTH) induces resistance against wheat pathogens by the
activationofsystemicacquiredresistanceinthehostplant.Inthisstudy,wemonitoredgeneexpressionin
spring wheat after treatment with each of these agrochemicals in a greenhouse trial using a microarray
containing 600 barley cDNA clones. Defence-related genes were strongly induced after treatment with
BTH,confirmingtheactivationofasimilarsetofgenesasindicotplantsfollowingsalicylicacidtreatment.
A similar gene expression pattern was observed after treatment with fenpropimorph and some defence-
related genes were induced by azoxystrobin, demonstrating that these fungicides also activate a defence
reaction.However,lessintenseresponsesweretriggeredthanwithBTH.Thesameexperimentsperformed
under field conditions gave dramatically different results. No gene showed differential expression after
treatmentanddefencegeneswerealreadyexpressedatahighlevelbeforeapplicationoftheagrochemicals.
These differences in the expression patterns between the two environments demonstrate the importance of
plant growth conditions for testing the impact of agrochemicals on plant metabolism.
Abbreviations: BTH, benzo(1,2,3)thiadiazole-7-carbothioic acid S-methylester; INA, 2,6-dichloroisonicot-
inic acid, JA, jasmonic acid; SA, salicylic acid; SAR, systemic acquired resistance
Introduction number of systemic fungicides with different
modes of action and targets have been developed
Plants have evolved effective resistance mecha- to reduce the losses caused by these diseases.
nismsthatenablethemtodefendagainstpathogen Strobilurins form a family of broad-spectrum
attacks.Nevertheless,allcropsaresusceptibletoa fungicides that are derived from a natural com-
number of major fungal pathogens that cause up pound, strobilurin A, which is produced by
to 20% of yield losses (Gullino et al., 2000). In the wood-rotting fungus Strobilurus tenacellus
cereals, rusts, mildews and Septoria are the most (Bartlettet al.,2002).Thesynthesisofderivativesof
damaging fungal diseases. In the last decades, a thismoleculehasledtoseveralactivecompounds,
694
including azoxystrobin (Gullino et al., 2000). powdery mildew (Blumeria graminis), leaf rust
Azoxystrobinandtheotherstrobilurinsareinhibi- (Pucciniatriticina)andSeptorialeafspot(Go¨rlach
tors of fungal mitochondrial respiration by block- et al., 1996) but not to Fusarium head blight
ingtheelectrontransferattheQ siteofcytochrome (Yu et al., 2001). The treatment of wheat plants
0
bc (Affourtit et al., 2000). Strobilurins currently with SA results in a lower resistance level against
1
represent10%ofthefungicidemarketandareused powdery mildew compared to plants treated with
by farmers to control fungal pathogens such as BTH, suggesting the involvement of other signal-
powderymildewandrusts.Besidestheiranti-fungal ling pathway(s) to induce this SAR-like response
action, strobilurins are also known for their (Go¨rlachet al.,1996).
‘‘greeningeffect’’onthecropwhichisdefinedasa Theputativeeffectofthesecompoundsoncrops
delayedleafsenescenceandanincreasedgrain-fill- has been tested by studying possible consequences
ing period (Bartlett et al., 2002). This side effect oftheirprimaryaction(forexamplewiththemea-
seemstoresultfromtheinhibitionofethylenebio- surement of sterol content after morpholine treat-
synthesisbyreductionofproductionofsuperoxide ment (Khalil and Mercer, 1991)) and by few
whichisthemediatoroftheconversionreactionof bioassays (Grossmann and Retzlaff, 1997). How-
1-aminocyclopropane-1-carboxylic acid (ACC) to ever,littleisknownontheireffectonthewholeplant
ethylene(WuandvonTiedemann,2001). metabolism. Genome-wide expression profiling
Morpholines are another family of systemic (Schena et al., 1995) is a technology for studying
fungicides, known since 1965 (Mercer, 1991). changes of global gene expression after a specific
Fenpropimorphwasdiscoveredin1979andisstill treatment of the plant. For the two closely related
commonly used against mildews and rusts. It species wheat (Triticum aestivum L.) and barley
inhibits two enzymes of fungal sterol biosynthesis (HordeumvulgareL.),morethan950,000ESTshave
(Engels et al., 1998). The morpholine compound beencharacterised(seetheTriticeaeESTdatabase
inhibits the enzymes sterol D14 reductase and website http://wheat.pw.usda.gov/genome/). Such
D8)D7 isomerase by binding tightly to their cata- collectionsofESTscanbeusedfortheconstruction
lyticsite(Mercer,1993;Debieuet al.,2000).Some ofcDNAmicroarrays.Theputativeroleofgenesof
phytotoxic effects like growth delay and altered unknownfunctioncanbepredictedfromsimilarity
phytosterol composition have been observed in of expression patterns, even from different species
cereals treated with this fungicide (Mercer et al., (van Noort et al., 2003). In cereals, this technique
1989; Khalil and Mercer, 1991). has been used in only few studies. Rice cDNA
Systemic acquired resistance (SAR) was dis- microarrays have been used to study gene expres-
covered several decades ago and has been studied sionrelatedtosaltstresstoleranceofrice(Kawasaki
intensively(Metraux,2001).ASARresponseleads et al., 2001) and iron deficiency in barley (Negishi
to pathogen resistance in the whole plant after et al., 2002). Maize development and the response
biologicalorchemicalstimulation.Thismechanism to UV radiation were studied with maize cDNA
allowsthe plant toprotect itself against numerous microarrays (Lee et al., 2002; Casati and Walbot,
viral, bacterial or fungal pathogens, depending on 2003). In barley, expression profiling was used for
thespecies(Oostendorpet al.,2001).Thetwomain the detection of mutated genes (Zakhrabekova
chemicalSARenhancersare2,6-dichloroisonicoti- et al.,2002)andresponsestodroughtandsaltstress
nic acid (INA) and benzo(1,2,3)thiadiazole-7-car- (Ozturk et al., 2002) or studying gene-for-gene
bothioicacidS-methylester(BTH),whichareboth interactionwithpowderymildewusingtheBarley1
structurally similar to salicylic acid SA (Go¨rlach GeneChipfromAffymetrix(Caldoet al.,2004).
et al.,1996).Indicotyledonousplants,thesemole- Extensive toxicological risk assessment on the
culesinducepathogenesis-relatedgenesandspecific active compounds of pesticides and on their
genes involved in signalling. Salicylic acid plays a catabolites in the plant is an integral part of the
key-roleinthesignaltransductionpathwayleading regulatory approval process of pesticides and a
to SAR (Lawton et al., 1995). However, in mono- huge database on these compounds is available at
cotyledonous plants, the role of SA has not been World Health Organisation (http://www.who.int/
clearlydemonstrated,althoughthesynthesisofSA pcs/jmpr/jmpr.htm). Most of the studies concern
is induced by aphid damage in barley (Chaman food safety issues and human health. Therefore, a
et al.,2003).Inwheat,BTHcaninduceresistanceto high food safety level of pesticide residues and
695
theirmetabolitesincropscanbeassumed.Thereis laboratory (SFR clones) and from Clemson Uni-
also a broad knowledge on endogenous plant versity(HVSMEg,HVSMEhandHV_Cebclones)
compounds affecting human health (Stegelmeier were chosen tocover major biochemicalpathways
et al., 1999). In contrast, studies on possible (for more details about the libraries, see http://
changes in plant metabolism upon pesticide treat- wheat.pw.usda.gov/genome for our laboratory‘s
ment are not mandatory and there is a lack of library and http://www.genome.clemson.edu/pro-
information on this aspect for most pesticides. jects/barleyfortheClemsonUniversitycollection).
In order to determine whether the three com- We used a method adapted from Reymond et al.
pounds azoxystrobin, fenpropimorph and BTH (2000) to print PCR products amplified from
alterwheatgeneexpression,weproducedacDNA these clones and negative controls (human and
microarray containing 600 barley genes covering Arabidopsis thaliana cDNA) onto coated glass
the major plant biochemical pathways. Here, we slides.Eachclonewasprintedtwice.TheClemson
report the impact of the two fungicides and the clones were amplified twice in 150 ll with Taq
SAR enhancer on gene expression in wheat plants polymerase (Sigma-Aldrich, Buchs, Switzerland)
grownundercontrolledgreenhouseconditionsand using5¢endamino-modifiedM13universalprimers
compare these results to a similar trial where in 35 cycles (94 (cid:2)C, 45 s; 52 (cid:2)C, 45 s; 72 (cid:2)C, 90 s).
plants weregrown inanagriculturalenvironment. TheSFRcloneswereamplifiedinthesamevolume
with the 5¢ end amino-modified TriplExAmp
primers(94 (cid:2)C,45 s;62 (cid:2)C,45 s;72 (cid:2)C,90 s)for10
Materials and methods cycles followed by 25 cycles (94 (cid:2)C, 45 s; 55 (cid:2)C,
45 s;72 (cid:2)C,90 s).TheDNAproductswerechecked
Plant material and treatments on agarose gels and sequenced with a 377 ABI
prismTMDNAsequencer,toconfirmtheiridentity.
Seedsofspringwheat(TriticumaestivumL.,variety They were concentrated during purification with
Greina) were grown in the greenhouse (16 h light/ multiscreen-PCR (Millipore, Volketswil, Switzer-
20 (cid:2)C, 8 h night/16 (cid:2)C, 2–4 seeds per pot). At land).Twosetsofslideswereproduced,usingtwo
growthstage32(Tottman,1987),plantsweretrea- differentDNAspotters.Inthefirstset,cDNAsam-
tedwithBTH(Bion(cid:3),Novartis(Basel,Switzerland) plesweredilutedinprintbuffer(NoAbBiodiscoveries
60 g/ha, as recommended by the manufacturer). Inc., Mississauga, Canada) at a concentration of
ThesecondapplicationofBion(cid:3)andthesprayingof 0.5–1 lg/ll. PCR products were printed using a
the two fungicides (azoxystrobin, Amistar(cid:3) from GMS417Arrayer(Affymetrix,SantaClara,USA)on
Syngenta(Basel,Switzerland)andfenpropimorph, epoxy-coatedglassslides(NoAbBiodiscoveriesInc.).
Corbel(cid:3) from BASF (Wa¨denswil, Switzerland) at Thesecondsetwasproducedaccordingtotheprotocol
the concentration of 1 l/ha) were made at growth of P. Reymond (http://www.unil.ch/ibpv/WWWPR/
stage39(Tottman,1987).Otherplantswerekeptas Docs/protocols.htm), using the printing facilities of
untreatedcontrols.Inthefield,theplantsweresown Lausanne University (printing robot (OmniGrid),
in 5-row plots (1.3 m wide, 1.2 m long, approxi- GeneMachines, Ann Arbor, USA). The purified
mately 50 seeds/row) near Zu¨rich, Switzerland, at probes were diluted with a 2 · spotting solu-
theSwissFederalResearchStationforAgroecology tion(6 · SSC,3 Mbetain)tothefinalconcentrationof
and Agriculture (FAL Reckenholz, 440 m above 0.5–1 lg/ll DNA, 3 · SSC, 1.5 M betain and then
sea level). For each treatment, four plots were printedonQMTAldehydeslides(PeqlabBiotechnol-
sprayedwithonefungicideeach,followingthesame ogieGmbh,Erlangen,Germany).
protocol as for the treatment in the greenhouse.
Four further plots were left untreated and used as RNA isolation and preparation of the fluorescent
control.Forbothtrials,flagleaveswereharvestedat targets
24 h,1and2 weeksaftertreatment.
For each treatment and time point, several flag
Preparation of the cDNA microarray leaves were pooled to reduce biological variation.
Total RNA was isolated using TRIzol reagent
A total of 600 cDNA clones of barley (Hordeum (InvitrogenLifeTechnologies,Basel,Switzerland),
vulgareL.,varietiesMorexandCI16155)fromour and 40 lg of RNA was used for the reverse-tran-
696
scription, using either Cy5-labelled dCTP for the Cy3. Genes were considered induced or repressed
treated samples or Cy3-dCTP (Amersham Bio- by Significance Analysis of Microarray (SAM,
sciences, Otelfingen, Switzerland) for the non- Excel Add-in available at http://www-stat.stan-
treated control samples, adapted from Reymond ford.edu/(cid:1)tibs/SAM/). This program allowed the
et al.(2000).Thereactionwasincubatedfor2 hat determination of both differentially expressed
42 (cid:2)C. After pooling the treated and control genes and corresponding false discovery rates
samples, RNA was degraded, the labelled cDNAs (FDR, (Tusher et al., 2001; Hu et al., 2003)). The
werepurifiedusingtheMinElutePCRpurification number of differentially expressed genes and the
kit (Qiagen, Basel, Switzerland) and diluted in FDR were determined in order to have one gene
hybridisation solution (3· SSC, 0.2% SDS, 0.2% considered as falsely detected (Samimi et al.,
yeast tRNA). 2005). The obtained FDR were generally below
10%inthegreenhousebuthigherwhentherewere
Hybridisation on microarrays, scanning very few genes differentially expressed after the
and analysis treatments.Therefore,whenonly3–6genesoutof
the 600 genes of the chip were differentially
The target solution was denatured for 1 min at expressed,theFDRwerebetween16–37.5%.Such
95 (cid:2)C and applied to the microarray slide which high ratios have already been observed in a recent
wasthencoveredwithacoverslip.Hybridisations studyonplant–insectinteractionswherefewgenes
were performed in chambers placed in a water were differentially expressed and high FDR were
bathat65 (cid:2)Cfor14–16 hr.Theslideswerewashed obtained (Reymond et al., 2004).
twicein1 ·SSC,0.03%SDSfor6 min,thentwice ClusteranalyseswerecarriedoutusingGenesis
in 0.2 · SSC for 5 min and finally twice in 0.05 · software (Sturn et al., 2002). Reproducibility
SSC for 5 min. They were subsequently dried by between the replicates was checked by measuring
centrifugation. correlation and data are presented in trees created
Scanning of the microarray slides was per- with Genesis software.
formedusingaScanArray5000(PerkinElmerLife
Sciences, Rodgau – Ju¨gesheim, Germany) at the Northern blot analysis
resolution of 10 lm/pixel. Photomultiplier and
laser power settings were adjusted in order to The results from microarray experiments were
obtain similar intensity levels of signal for the partially validated by RNA blot analysis. The
control spots (house-keeping genes and alien same quantity of RNA (40 lg) was electropho-
cDNA, Stratagene, Amsterdam, Netherlands) for resed and transferred to a nylon membrane as
both channels. Pictures were analysed by Imagene previously described (Feuillet et al., 1997). The
4.1 software (BioDiscovery Inc., Los Angeles, labelled probes WCI2, 5, WIR1c, actin, PR 1a/1b
USA). Spots flagged as empty by the software or (HV_CEb0006J08f)) were prepared using stan-
manually were removed from the analysis. Nor- dard procedures (Sambrook et al., 1989) with
malisationofthesignalintensitiesbetweenthetwo clones previously used as templates for the barley
channels was performed using the global method, cDNA microarray. The RNA blots were analysed
and between the slides by scale normalisation using Biomax MS-1 film (Kodak, Lausanne,
(Yang et al., 2002). Experiments with reversion of Switzerland).
the dyes did not show significant differences in
hybridisation level, so only one way of labelling
(control with Cy3 and treated sample with Cy5 Results
dye) was further used. Hybridisations were per-
formed in triplicates, using three different samples Barley cDNA microarray design
treated by the same compound for the three time
points. To better assess the gene expression of To study gene expression in wheat, we made a
untreated plants from both growth conditions, cDNA microarray containing 600 cDNAs. Barley
microarrayswerealsoperformedforthethreetime andwheataretwospeciesshowinghighsimilarityof
points. The samples from the greenhouse were theirgeneorder(FeuilletandKeller,2002)andhigh
labelledwiththeCy5andthosefromthefieldwith conservation of gene sequences between the two
697
species(Bennetzenet al.,1998).Severalstudieshave
shown that coding sequence identity can reach
100%andisusuallyover90%(Ramakrishnaet al.,
2002;SanMiguelet al.,2002;Caldwellet al.,2004).
Therefore, there is good cross-hybridisation
between nucleic acids of barley and wheat.
Furthermore,astheprobesarelongerthan400 bp,
the cDNA microarray allows small inter-species
differences with only minimal consequences on
hybridisation (Adjaye et al., 2004; Close et al.,
2004).Thegenesusedforourchipwerechosenfrom
barley(Hordeumvulgarecv.Morex)cDNAlibraries
from pre-anthesis spikes, spikes 5–45 days after
pollination, plants challenged with the powdery
mildew pathogen (HVSMEg clones, HVSMEh
clones and HV_Ceb clones, Clemson University,
respectively) and from our laboratory collection
(SFR clones). In blast analyses, we found that the
600 sequences all had corresponding ESTs in the
wheat libraries (E value between 0 and e-20 and
mean identity level of 90%), reflecting the high
homology level between the barley and wheat Figure1. Analysis of gene expression after chemical treat-
sequences(seesupplementaryTable 6).Expression ments. Hierarchical clustering of the differentially expressed
genes, as displayed by the software Genesis, is shown after
of genes encoding enzymes from the major bio-
treatmentwithazoxystrobin(A),BTH(B)andfenpropimorph
chemical pathways of primary and secondary
(F) in the greenhouse at the time points 24h, 1 and 2weeks
metabolism was analysed. In preliminary tests, aftertreatment.Thecolourscalebarrepresentstheratiovalues.
wheatderivedlabelledsamplesresultedinsuccess- Genes with higher expression level after treatment appear in
ful cross-hybridisation with the barley probes on red; those with lower hybridisation intensity appear in blue.
After 24h, the expression patterns of BTH-treated and
our microarray slides (data not shown). This is in
fenpropimorph-treated plants clustered together (blue tree on
agreement with the generally observed high
topoftheclusters).Forthesubsequenttimepoint,expression
conservation of coding sequences in wheat and profiles of azoxystrobin-treated and BTH-treated plants were
barley(Goffet al.,2002;Closeet al.,2004). moresimilartoeachother,andafter2weeks,treatmentswith
fenpropimorph and azoxystrobin resulted in similar differen-
tially expressed genes. The lower panel shows the microarray
Effect of the BTH treatment
resultsofthesubgroupofninegenesalsoanalysedbyNorthern
blot (Figure2). From top to bottom: actin gene, the lipoxy-
The analysis of BTH treated plants grown in the genasegene(WCI2),thethiolproteasegene(WCI4),thewheat
greenhouse revealed that 17 genes were differen- chemically induced 1 and 5 genes, a wheat induced resis-
tiallyregulatedafter24 h(falsediscoveryrateFDR tance gene (WIR1c), the pathogenesis-related genes PR 1
(HV_CEb0010L20f)andPR1a/1b(HV_CEb0006J08f)andthe
5.8%), 24 after 1 week (FDR 4.1%) and 9 after
proteindisulfide-isomerasegene(HVSMEg0005I10f).
2 weeks (FDR 8.9%), which represents about 5%
ofthegenespresentonthechip.Allofthesediffer-
entially expressed genes were over-expressed and of 2- to nearly 60-fold 24 h after treatment when
none repressed (Figure 1, Tables 1–3). The differ- compared to untreated plants (Table 1). Most of
entially regulated transcripts mainly belong to thesegenesremainedover-expressed1and2 weeks
defence-related genes. Genes encoding glucanase, aftertreatmentbutwithdecreasingover-expression
lipoxygenase and pathogenesis-related genes ratioscomparedtothoseobtainedat24 h(Tables2
(PR1a/1b, 2, B1), the wheat induced resistance and3).Theseresultswereconfirmedforsomegenes
genes(WIR1B, 1C and 232)and the wheat chemi- (PR1, PR1a/1b,PDI, WCI1, WCI2, WCI4,WCI5
cally induced genes 2 (WCI2, lipoxygenase), and WIR1c) by Northern blot analysis (Figure 2)
1 (WCI1, jasmonate-induced), 4 (WCI4, protease) and are in agreement with other studies (Go¨rlach
and5(WCI5)showedanincreasedgeneexpression et al.,1996;StadnikandBuchenauer,1999).Genes
698
e
A24h 1.30.90.6 1.23.06.5 2.81.63.21.2 2.71.02.5 1.52.70.8 1.61.82.2 2.91.52.70.9 1.41.20.9 2.12.7)9.0 egen
at
c
h 872 172 2902 767 539 320 6270 407 923 di
F24 17.4.2. 6.10.8. 8.1.3.2. 2.3.2. 3.2.2. 2.2.3. 3.2.1.2. 3.2.2. 1.1.)1. sin
e
u
B24h 56.744.335.0 12.77.36.1 5.55.33.93.5 3.13.13.0 2.82.82.7 2.62.62.5 2.12.11.91.9 1.61.61.4 1.41.4)1.2 veval
ati
12) neg
A). E-1) 3.11. E-1) and
( 1 n
oxystrobin rmprotein v:3)(EC1. rmprotein neinductio
eenhousetrial24haftertreatmentwitheitherBTH(B),fenpropimorph(F)oraz Putativefunction UP|Q41520(Q41520)Lipoxygenase(Fragment)(EC1.13.11.12)PIR|T06273Benzothiadiazole-inducedproteincloneWCI-1-wheathomologuetoUP|Q41522(Q41522)Thiolprotease UP|LOX1_HORVU(P29114)Lipoxygenase1(EC1.13.11.12)UP|PR1A_HORVU(P32937)Pathogenesis-relatedprotein1A/1BprecursorUP|PR12_HORVU(P35792)Pathogenesis-relatedproteinPRB1-2precursor UP|Q94F70(Q94F70)Putativethaumatin-likeproteinsimilartoUP|COPD_ORYSA(P49661)CoatomerdeltasubunitUP|Q41581(Q41581)WIR1proteinsimilartoUP|Q8H841(Q8H841)Putativereceptor-likeproteinkinase UP|E13B_HORVU(P15737)Glucanendo-13-beta-glucosidaseGIIprecursorUP|P93180(P93180)Pathogenesis-relatedprotein4precursorPIR|T06278Benzothiadiazole-inducedproteincloneWCI-5-wheat UP|PDI_HORVU(P80284)Proteindisulfide-isomeraseprecursor(PDI)(EndospeUP|WIRB_WHEAT(Q01481)WIR1BproteinUP|LX23_HORVU(Q8GSM2)Lipoxygenase2.3chloroplastprecursor(LOX2:H UP|PDI_HORVU(P80284)Proteindisulfide-isomeraseprecursor(PDI)(EndospeUP|ENPL_HORVU(P36183)EndoplasminhomologprecursorUP|Q43765(Q43765)Chitinase(EC3.2.1.14) UP|PR1_HORVU(Q05968)Pathogenesis-relatedprotein1precursorsimilartoUP|Q7XH17(Q7XH17)Putativereceptor-likeproteinkinase4ZeamayssimilartoGP|13897320Somaticembryogenesisreceptor-likekinase2{}similartoUP|Q8H8H7(Q8H8H7)Putativeflavanone3-hydroxylase UP|Q9M4C7(Q9M4C7)Alleneoxidesynthase(EC4.2.1.92)UP|Q6RYF4(Q6RYF4)CoatomeralphasubunitUP|O65189(O65189)Glucanendo-13-beta-glucosidase UP|Q43764(Q43764)Chitinase(EC3.2.1.14)homologuetoUP|FKB7_WHEAT(Q43207)70kDaPeptidylprolylisomeraseUP|CHS1_HORVU(P26018)Chalconesynthase1 gresultasinthedatabaseofClemsonUniversity.ntiallyexpressedbySAManalysisareinboldtype.Thepositivevaluesindicatege%forB,FandA,respectively.
thegr ber uencindifferend10
n m 1 R qea
Table1.Differentiallyexpressedgenesi GeneIDAccessionnu wci2TAU32428wci1TAU32427wci4TAU32430 HVSMEh0095N14fBE455009HV_CEb0006J08fBE215358HV_CEb0024H14fBE559397 wir232TATHAUaHVSMEg0016C06fBG344787wir1cTARNAWIRHV_CEb0010H13fBE216428 HV_CEb0010G19fBE216411HV_CEb0003K12fBE214507wci5TAU32431 HVSMEg0005I10fBG343356wir1bWHTWIR1PHVSMEg0013J19fBE060855 HVSMEh0088H07fBE195158HV_CEb0002C16fBE214080HV_CEb0003J11fBE214483 HV_CEb0010L20fBE216529HV_CEb0017B21fBE558296HV_CEb0003A03fBE214285HV_CEb0021J19fBE519892 HVSMEg0011M01fBE060255HVSMEg0003M07fAW982621HV_CEb0003D01fBE214349 HV_CEb0003A01fBE214283aHVSMEh0100J22fBG418805HV_CEb0003P20fBE214619 aClonesthathavenotgiventhesameseIntensityratiosofgenesdeterminedtobrepression.TheFDRwere5.8%,4.5%
699
e
A1week 5.14.4)1.1 1.81.42.01.5 1.11.71.7 1.2)1.01.1 1.81.6)1.2 1.1)1.4)1.1)1.1 1.0)1.2)1.0 1.0)1.31.1)2.5)3.9 categen
di
k n
e i
we 6.33.81.2 1.31.12.01.2 2.11.31.8 1.41.41.1 1.91.31.0 2.31.11.11.2 1.30.90.8 2.11.22.0 1.51.6 ues
F1 ) )) ) ) ) ) ))) )) ) )) al
v
B1week 13.211.75.4 5.34.94.53.9 3.83.83.7 3.73.53.5 3.33.33.0 3.02.92.72.6 2.52.42.4 2.32.11.9 1.2)1.1 egative
n
A). d
n
ystrobin( asativa} ductiona
x z n
azo Ory nei
Table2.Differentiallyexpressedgenesinthegreenhousetrial1weekaftertreatmentwitheitherBTH(B),fenpropimorph(F)or GeneIDGeneaccessionPutativefunction wci1TAU32427PIR|T06273Benzothiadiazole-inducedproteincloneWCI-1-wheatwci4TAU32430HomologuetoUP|Q41522(Q41522)ThiolproteaseHV_CEb0010L20fBE216529UP|PR1_HORVU(Q05968)Pathogenesis-relatedprotein1precursor HV_CEb0024H14fBE559397UP|PR12_HORVU(P35792)Pathogenesis-relatedproteinPRB1-2precursorwir232TATHAUUP|Q94F70(Q94F70)Putativethaumatin-likeproteinwci2TAU32428UP|Q41520(Q41520)Lipoxygenase(Fragment)(EC1.13.11.12)HV_CEb0003K12fBE214507UP|P93180(P93180)Pathogenesis-relatedprotein4precursorTriticumaestivumHV_CEb0010N06fBE216563HomologuetoGP|17981573KinaseR-likeprotein{}HV_CEb0006J08fBE215358UP|PR1A_HORVU(P32937)Pathogenesis-relatedprotein1A/1BprecursorHVSMEh0095N14fBE455009UP|LOX1_HORVU(P29114)Lipoxygenase1(EC1.13.11.12) HVSMEg0002G09fAW982228HomologuetoUP|Q75RZ2(Q75RZ2)PutativecaffeoylCoAO-methyltransferaseHV_CEb0021J19fBE519892similartoUP|Q8H8H7(Q8H8H7)Putativeflavanone3-hydroxylasewir1cTARNAWIR1UP|Q41581(Q41581)WIR1protein wci5TAU32431PIR|T06278Benzothiadiazole-inducedproteincloneWCI-5-wheatHV_CEb0010G19fBE216411UP|E13B_HORVU(P15737)Glucanendo-13-beta-glucosidaseGIIprecursoraHV_CEb0011F02fBE216529UP|PR1_HORVU(Q05968)Pathogenesis-relatedprotein1precursor HV_CEb0015B10fBE558194weaklysimilartoGP|22535653PutativeproteinkinaseXa21receptortypeprecursor{HV_CEb0011F16fBE216724HomologuetoUP|Q8W4U9(Q8W4U9)ClathrinassemblyproteinAP17-likeproteinHV_CEb0011I04fBE216768similartoUP|Q84S61(Q84S61)Putativeserine/threoninekinaseproteinHVSMEh0088K24fBE195244UP|Q42839(Q42839)Chitinase(EC3.2.1.14) HV_CEb0010N04fBE216561HomologuetoUP|Q41328(Q41328)Ptokinaseinteractor1HV_CEb0017B21fBE558296similartoUP|Q7XH17(Q7XH17)Putativereceptor-likeproteinkinase4HVSMEg0007A12fBG343889similartoUP|Q6J2K7(Q6J2K7)ProteintyrosinephosphataseaHVSMEh0099N09fBE601871HomologuetoUP|TBP2_WHEAT(Q02879)TATA-boxbindingprotein2HV_CEb0010P06fBE216610HomologuetoUP|Q43220(Q43220)Peroxidase(EC1.11.1.7)HVSMEh0102L08fBE603237similartoUP|Q8S8Z0(Q8S8Z0)Proteinphosphatase2C HVSMEh0081G18fBE193520similartoSP|P37837Transaldolase(EC2.2.1.2)HordeumvulgarevulgareHVSMEh0081M04fBE193575HomologuetoGB|CAA75793Sucrosesynthase2{subsp.} aClonesthathavenotgiventhesamesequencingresultasinthedatabaseofClemsonUniversity.IntensityratiosofgenesdeterminedtobedifferentiallyexpressedbySAManalysisareinboldtype.Thepositivevaluesindicategerepression.TheFDRwere4.1%,16%and25%forB,FandA,respectively.
700
e belongingtootherfunctionalclasseswerealsoup-
eks gen regulated. Two protein disulfide isomerase genes
A2we 3.31.61.21.0 2.42.22.5 2.72.11.4 ndicate ashnodwoedneancoinadtoumcteiornporfotReiNnA(CsyOnPth)essiusbaufnteirt 2g4enhe,
i
es suggestinganincreasedbiosynthesisofsecretedor
u
ks al cell surface proteins (Harter, 1995; Ciaffi et al.,
e v
F2we 2.21.41.61.1 2.91.91.9 3.91.7)1.1 ative 2n0o0n1e)-.3G-heyndersoxenylcaosdeinagndthecapffuetoaytilveCporAotOein-msefltahvyal--
g
e
A). dn transferase, involved in flavonoid biosynthesis,
n( ks an showedanactivationoftranscriptiononlyafterone
xystrobi B2wee 6.44.03.43.2 3.23.12.9 2.92.72.4 duction week.
o n Impact of the fenpropimorph treatment
z i
a e
eenhousetrial2weeksaftertreatmentwitheitherBTH(B),fenpropimorph(F)or Putativefunction PIR|T06273Benzothiadiazole-inducedproteincloneWCI-1-wheathomologuetoUP|Q41522(Q41522)ThiolproteasePIR|T06278Benzothiadiazole-inducedproteincloneWCI-5-wheatUP|Q41581(Q41581)WIR1protein UP|Q94F70(Q94F70)Putativethaumatin-likeproteinUP|PR1A_HORVU(P32937)Pathogenesis-relatedprotein1A/1BprecursorUP|PR1_HORVU(Q05968)Pathogenesis-relatedprotein1precursor UP|Q41520(Q41520)Lipoxygenase(Fragment)(EC1.13.11.12)UP|PR12_HORVU(P35792)Pathogenesis-relatedproteinPRB1-2precursorUP|Q9SM34(Q9SM34)Putativegermin-likeproteinprecursor ntiallyexpressedbySAManalysisareinboldtype.Thepositivevaluesindicategen%forB,FandA,respectively. AeaxfftpteerrresBsiTfoeHnnpprtaorteptaeirtmmnoewrnpaths(sFimtirgieulaartremt1oe)nttha,elthotonhueegohbotvsaeeinvraeendll
r e5
expressedgenesintheg Accessionnumber TAU32427TAU32430TAU32431TARNAWIR1 TATHAUBE215358BE216529 TAU32428BE559397AJ237942 sdeterminedtobedifferwere8.9%,20%and37. F24ighur,e12a.ndRN2Aweebklostafatenraflyusnigsicsihdoewtrineagtmdeiffnetrienntthiaelgerxepenrehsosuiosne
Differentially D Eb0006J08fEb0010L20f Eb0024H14f ratiosofgenen.TheFDR tfatWrehsinraCepqleI.ru5oa)Cpw,li:ithamyenowaocrtohnpne-hctatrh.rteoeNaimlntiendiocdeufaclltcalehyobdenerlitblnreeolsddoilus,tptsca,eArnoda:cbeelagiszpgeonweonxexesyeyrseg(t(eWruWnosabIeCRsdineI:14,ctg,h)Be,enW:teahBCce(tTWIipn1HaCg,taIhe2nnFo)de-:,
Table3. GeneI wci1wci4wci5wir1c wir232HV_CHV_C wci2HV_CpBTag Intensityrepressio g((HHenVVe_SsiCMs-ErEebgl0a00t0e00d65JIg01e80nff)e),s.aPnRd t1he(HpVro_tCeiEnb0d0is1u0lLfid20e-fi)soamnderaPsRe1gae/1nbe
701
additional genes (the allene oxide synthase, the WCI5, WIR1b and WIR1c were similar to each
putative flavanone 3-hydroxylase, four PR other with a small up-regulation detected by
and one putative kinase genes) were significantly microarray analysis only 24 h after treatment. No
over-expressed 24 h after fenpropimorph treat- statistically significant induction of these genes
ment (FDR 4.5%) compared to the BTH-treated was observed with microarray analysis forthelast
plants (Table 1). For WCI2 and 1, the treated/ two time points. However, over-expression of
control ratios were lower than the ones obtained WCI5and WIR1cwas detectedforall time points
with BTH, suggesting a weaker impact of this by Northern analysis (Table 1, Figure 2). Inter-
compound on plant metabolism compared to the estingly,thewheatchemicallyinducedgenes4and
SARenhancer(Tables1–3).Defence-relatedgenes 2 that had the strongest up-regulation after BTH
such as the glucanase and PR genes were induced and fenpropimorph treatment were not induced
after 24 h but did not show any differential after 24 h but showed up-regulation after 1 and
expression later, suggesting a rapid but transient 2 weeks, respectively. This indicates that the
response. The putative thaumatin-like gene responsesaftertheapplicationofazoxystrobinare
(WIR232) showed a different pattern, with an only partially overlapping with the pattern gener-
induction after 24 h, no differential expression ated by the two other compounds.
after 1 week but again an increase of mRNA A chalcone synthase gene showed a high
amount after 2 weeks compared to untreated repression level 24 h after treatment. This gene,
plants. This gene seems to be involved in two like the lipoxygenase genes, is regulated by ethyl-
phasesofthereactionandcouldberegulatedbya ene (Wan et al., 2002). The antioxidant and
different pathway. Expression of the PR1, PR1a/ ‘‘green’’ effects of azoxystrobin are attributed to a
1b, WCI2, WCI4, WCI5 and WIR1c and PDI reductionofethyleneproduction(Grossmannand
genes were also analysed by Northern blot analy- Retzlaff, 1997). Thus, the reduction of chalcone
ses, giving results similar to the microarray synthase(HV_CEb0003P20f)expressionaswellas
hybridisations (Figure 2). Interestingly, the allene the absence of over-expression of the three lipox-
oxide synthase gene was up-regulated after 24 h, ygenase genes (WCI2, HVSMEg0013J19f and
suggesting an increase of JA biosynthesis. This HVSMEh0095N14f) after 24 h confirm that
indicates that the JA synthesis would be induced azoxystrobin has a specific effect on wheat gene
afterthetreatmentwithfenpropimorphwhereasit expression which is different from the action of
was not after BTH treatment. As the putative fenpropimorph and BTH (Table 1). After 1 week,
flavanone-3-hydroxylase gene is also induced, the two genes involved in sugar metabolism (a gene
flavonoid synthesis pathway could be triggered similar to a transaldolase gene and a sucrose syn-
early after fenpropimorph treatment. thase 2 gene) were down-regulated.
Effect of the treatment with azoxystrobin Impact on gene expression of the plant protection
compounds in the field
The strobilurin fungicide is known to produce a
‘‘green effect’’ on wheat plants as non-fungicidal Fungicide and BTH treatments were made in the
secondary effect, with darker green leaves, field to compare gene expression with plants trea-
enhanced concentration of chlorophyll and ted in the greenhouse. Surprisingly, no gene
increased biomass production (Grossmann and showed differential expression after treatment of
Retzlaff, 1997). It also induces some antioxidant the plants in the field, whatever the compound
activity (Wu and von Tiedemann, 2002). After used.TheWCI2geneshowedhigherexpressionin
azoxystrobintreatment,onlyfewgenesshowedan the Northern blot 24 h after BTH treatment
alteration of their expression, with a high FDR of (Figure 3) but this was not statistically significant
10%,16%and37.5%astherewereonlyten,four in the microarray experiments. The absence of
and three genes differentially expressed after 24 h, differentially expressed genes in the field trial was
1 and 2 weeks, respectively (Tables 1–3). The supported by a high reproducibility between
PR1, 1A/1B and B1-2 genes were significantly replicates (see supplementary Figure 4). The
over-expressedafter24 honly,andtheWCI1gene non-treatedplantsshowedahighexpressionofthe
after 1 and 2 weeks. The expression levels of defence-related genes that were over-expressed
702
like the RNase S-like protein gene or the inositol-
3-phosphatesynthasegeneweremoreexpressedin
the greenhouse than in the field (supplementary
Table 4). However, the apparent over-expression
of the RNAse S-like gene could be due to the
time-shiftwhencollectingthesamples(morningin
the greenhouse and afternoon for the field exper-
iments) asit has been shown that this geneislight
responsive (Gausing, 2000).
Discussion
A SAR enhancer and two commonly used fungi-
cideswithdifferentandspecificchemicalmodesof
actionwereusedinthisstudytobetterunderstand
their impact on wheat gene expression and plant
metabolismusingcDNAmicroarrays.Arelatively
Figure3. RNA blot analysis showing differential expression
24h,1and2weeksafterfungicidetreatmentinthefieldtrial. small proportion of the genes present on the chip
C: non-treated control, A: azoxystrobin, B: BTH, F: fen- (around 5%) showed differential expression after
propimorph. Seven labelled probes were used: actin gene as the treatments. Although few transcripts showed
qualitycontrolofthe blots, alipoxygenasegene (WCI2), and
differential expression profiles, common and
thewheatchemicallyinducedgenes(WCI4,WCI1andWCI5),
compound-specificpatternswereobservedafterthe
the pathogenesis related genes PR 1 (HV_CEb0010L20f) and
PR 1a/1b (HV_CEb0006J08f). Twenty four hours after the three different treatments. Most of the differen-
BTHtreatmentand2weeksaftertheazoxystrobintreatment, tially expressed transcripts detected belonged to
the WCI2 gene showed high expression but in microarray defence-related gene families, such as the PR and
experimentsthiswasnotstatisticallysignificantandonlyfound
WIR genes, revealingthat fungicides can have not
inoneofthethreereplicateslides.
onlyanimpactontheirfungaltargetsbuthavealso
animpactontheplantitselfmoreorlesssimilarto
after the treatments in the greenhouse trial BTH.Interestingly,nogenecodingforkey-enzymes
(Figure 3andsupplementalTable 4).Forallthree of the primary metabolism showed differential
timepoints,thePR1,PR1a/1b,WIR232,WIR1c,a expression pattern after BTH and fenpropimorph
b1-3 glucanase, two peroxidase and four chitinase treatments,showing that these compoundsmainly
genes were significantly higher expressed com- affect specific sets of genes and defence pathways.
paredtotheplantsgrowninthegreenhouse.Other Onlytwogenesbelongingtothesugarmetabolism
defence-related genes were also more expressed in pathway and coding for sucrose synthase 2 and
the field for one or two time points (PR4, WIR1a, transaldolase, respectively, were down-regulated
b1-3 glucanase, glutathione peroxidase, two chal- 1 weekafterazoxystrobintreatment.
cone synthase, two kinase genes). These results BTH, fenpropimorph and azoxystrobin
suggest that many of the genes induced by these induced genes known to be involved in plant
fungicides in the greenhouse are constitutively defence against pathogens. The PR1 and wheat
expressed during growth in the agricultural envi- induced resistance genes (WIR1b, WIR1c, 232)
ronment. However, not all the defence-related were activated early after the treatment with these
genes did behave similarly. The WCI 1, 2 and 4 compounds, following a similar pattern as in Ara-
genes did not show any significant changes in bidopsis and tobacco after BTH treatment (Fried-
expression between the two growth conditions richet al.,1996;Lawtonet al.,1996).However,the
(supplementalTable 4).Thealterationofresponse induction of the PR1 genes after BTH treatment
totheplantprotectioncompoundscouldbedueto contradicts previous results (Molina et al., 1999;
a combination of stresses occurring in the field Yuet al.,2001)wherenoactivationofthesegenes
(Rizhsky et al., 2002). In contrast to the genes was observed, but is in agreement with the results
over-expressed under field conditions, some genes ofGo¨rlachet al.(1996).Thus,theinductionofthe