Differential transcriptome analyses of three wheat genotypes reveal different host response pathways associated with Fusarium head blight and trichothecene resistance

Authors


E-mail: nora.foroud@agr.gc.ca; francois.eudes@agr.gc.ca

Abstract

Fusarium head blight is a disease of cereal crops caused by a group of trichothecene-producing Fusarium species. In the current study, three wheat genotypes with cv. Superb pedigree were evaluated for their ability to activate different defence-response pathways when exposed to one of three elicitors: (i) trichothecene-producing and (ii) non-producing F. graminearum strains, and (iii) the trichothecene toxin deoxynivalenol. Spikelets distal to the inoculation point were harvested at multiple time points in order to identify systemic temporal changes in transcript accumulation associated with resistance. Distinct differences were observed between the resistant genotypes and cv. Superb, as well as between the two resistant genotypes. The current data suggest that different molecular mechanisms exist not only between susceptibility and resistance responses, but between different forms of genetic resistance. It is proposed that Type 1 resistance in one of the resistant double haploid lines evaluated here involves a combination of structural features that slow fungal penetration and activation of a systemic response in uninfected tissues adjacent to the site of infection to prevent or minimize secondary infection; in contrast, Type 2 resistance is more likely a form of local resistance.

Introduction

Fusarium head blight (FHB) is a destructive disease of grain crops, with worldwide economic and health impacts. Wheat is one of the most heavily FHB-affected crops and suffers the largest economic damage. The disease is caused by a range of trichothecene-producing Fusarium species, F. graminearum (teleomorph: Gibberella zeae) being the most economically relevant (Parry et al., 1995). The fungus infects the cereal inflorescence during anthesis and grain development. Once established within the floret, fungal hyphae can spread from spikelet to spikelet through the vascular bundles of the rachis (Parry et al., 1995). Fusarium species produce trichothecene mycotoxins that accumulate in kernels of infected spikelets, reducing both food quality and grain germination. Trichothecenes are phytotoxic and are associated with aggressiveness of F. graminearum disease spread in Triticeae (Proctor et al., 1995; Bai et al., 2002; Langevin et al., 2004; Jansen et al., 2005), as was demonstrated in a series of studies conducted with a trichothecene non-producing strain of F. graminearum (FgTri5−; wildtype FgTri5+) (Proctor et al., 1995). On the other hand, trichothecenes do not appear to be necessary for initial infection (Bai et al., 2002), or infection of the wheat fruit coat (Jansen et al., 2005).

The most effective means to prevent damage caused by this disease is to cultivate crops with high levels of genetic resistance to FHB (Foroud & Eudes, 2009). Two major forms of FHB resistance have been characterized in wheat and related cereals (Schroeder & Christensen, 1963). Type 1 resistance (resistance to initial infection) is typically measured as the percentage of diseased spikes, and/or spikelets, resulting from exposure to the pathogen. Type 2 resistance (resistance to disease spread) can be evaluated by point inoculation of a single spikelet with Fusarium spores and subsequently measuring the percentage of infected spikelets (Foroud & Eudes, 2009). Several quantitative trait loci (QTLs) have been identified in FHB resistance, including the cv. Sumai 3-derived 3BS QTL associated with Type 2 resistance, and the 5A QTL involved in Type 1 resistance in cv. Sumai 3 and other genetic backgrounds (Bürstmayr et al., 2009). An improved understanding of the molecular mechanisms associated with resistance has the potential to provide essential new tools for the development of improved FHB-resistant germplasm.

Over the past decade, differential gene expression studies in the FHB-host interactions have been conducted (Pritsch et al., 2000, 2001; Boddu et al., 2006, 2007; Bernardo et al., 2007; Golkari et al., 2007, 2009; Li & Yen, 2008; Reinprecht et al., 2008; Jia et al., 2009; Steiner et al., 2009). Some of these studies involved work with near-isogenic lines. In association with the 3BS QTL Jia et al. (2009) observed higher expression of genes involved in cell wall biogenesis. In association with both the 3BS and 5A QTLs, there was higher transcript accumulation of a UDP-glucosyltransferase, which may be involved in deoxynivalenol (DON) detoxification, and of a pathogenesis-related (PR) family protein (Steiner et al., 2009). Higher accumulation of F. graminearum-induced PR-2, -4 and -5 transcripts has also been observed in cv. Sumai 3 compared with two near-isogenic lines that are susceptible to FHB (Golkari et al., 2009). Pritsch et al. (2000) tracked the F. graminearum-induced transcript accumulation of six genes encoding PR proteins, including PR-1, PR-2 (β-1,3-glucanase), PR-3 (chitinase), PR-4, PR-5 (thaumatin-like protein) and peroxidase (POX), and observed increases in transcript accumulation for all six. Among these, PR-4 and -5 accumulated earlier and in higher quantities in cv. Sumai 3 than in the susceptible cv. Wheaton. In a second study, the same group found that PR-1, -2 and -5 accumulated in the uninvaded spikelets of wheat heads (Pritsch et al., 2001). In another resistant cultivar, Ning 7840, F. graminearum infection induced up-regulation of PR-1 and a chitinase precursor, relative to the response of the same genes in the susceptible cv. Clark (Bernardo et al., 2007). Consistent with these transcriptomic patterns, over-expression of various PR proteins has been shown to improve cereal resistance to FHB or Arabidopsis resistance to F. oxysporum (Epple et al., 1997; Anand et al., 2003).

Despite advances in knowledge of the genes and pathways associated with FHB-resistance, the molecular processes that actually confer a resistance or susceptible response remain unclear. The absence of gene-for-gene resistance increases the complexity of the disease response, and makes it difficult to identify discrete mechanisms underlying resistance or susceptibility. In order to better understand the cellular processes controlling disease outcomes, and create opportunities for developing novel forms of resistance, the molecular pathways in FHB-challenged wheat tissues that result in resistance need to be defined. Because changes in gene expression are expected to reflect the suite of host cell responses during FHB infection, a time-course study has been conducted of the differential transcriptome of three wheat genotypes in their response to isolated components of FHB. The three wheat genotypes include two FHB-resistant double haploid lines GS-1-EM0040 (CIMMYT 11/Superb*2) and GS-1-EM0168 (CM82036/Superb*2, with the 3BS QTL), and a susceptible Canadian cultivar, Superb, that shares 75% genetic identity with each of the double haploid lines. The isolated components of the disease being tested were the presence (FgTri5+) (FgTri5+ vs. water), or absence (FgTri5−) (FgTri5− vs. water) of trichothecene production in the pathogen, the presence or absence of a pure trichothecene (DON) (DON vs. water), and a direct comparison of FgTri5+ vs. FgTri5− was also made. The gene expression analysis was performed on uninoculated spikelets harvested from point-inoculated wheat spikes, which would be expected to capture the gene expression changes resulting from a systemic, rather than local, host tissue response. This experimental strategy allowed for comparison of the molecular biology of the systemic host responses in susceptible vs. resistant wheat, and of two different resistant backgrounds, while at the same time allowing an exploration of how different components of FHB disease affect the disease outcomes.

Materials and methods

Plant material

Three wheat genotypes with cv. Superb pedigree were used in the current study. Superb is a susceptible Canadian cultivar; GS-1-EM0040 (CIMMYT 11/Superb*2) and GS-1-EM0168 (CM82036/Superb*2) are double haploid lines generated by in vitro selection of microspore-derived embryos using a trichothecene toxin screen (0·23 mg L−1 deoxynivalenol, 0·23 mg L−1 15-O-acety-4-deoxynivalenol, 0·47 mg L−1 nivalenol, and 0·7 mg L−1 T2 toxin) as described by Eudes et al. (2008). All three wheat lines were screened for FHB resistance and agronomic properties, as described under the ‘Phenotypic evaluation’ section. All three parents of the double haploid lines have previously been screened for a series of FHB QTLs. Cultivar CM82036 was shown to carry 3BS (barc133 and barc147) and 5A (gwm293 and gwm156) QTLs (Badea et al., 2008). No FHB resistance QTLs have been previously developed in cvs Superb or CIMMYT 11; although cv. CIMMYT 11 was shown to yield amplicons similar in size to the 2DL (gwm539) and 4B (wmc238) QTLs from cv. Wuhan 1 (Badea et al., 2008). The double haploid line GS-1-EM0168 was screened here for the 3BS and 5A QTLs, and GS-1-EM0040 for the 2DL, 4B, and 5A QTLs, as described by Badea et al. (2008).

For the microarray experiment, seeds were sown in 1 gallon pots (one plant per pot) containing Cornell Mix (Boodley & Sheldrake, 1977) and grown in a greenhouse (21/18°C, 16 h photoperiod). Plants were watered as needed and fertilized biweekly with 20-20-20 (N-P-K). At the 5–7 leaf stage they were treated with Tilt™ (leaf surfaces were sprayed with 2·5 mL L−1 propiconazole, Syngenta Crop Protection Canada) and Intercept™ (a total of 0.24 mg of Imidacloprid in 50 mL of water was applied to each pot, Bayer Crop Science Canada) preventative measures against powdery mildew and aphids.

Phenotypic evaluation

Agronomic properties (plant height, yield, maturity and lodging) were recorded in 2006 at four sites, each with three blocks of four rows, at Agriculture and Agri-Foods Canada in Lethbridge, Alberta. FHB resistance was evaluated both in the greenhouse and in the field as described in Eudes et al. (2008). Greenhouse experiments were conducted using the F. graminearum strain N2 for point inoculation at Agriculture and Agri-Foods Canada locations in Lacombe and Lethbridge, Alberta, and Ste-Foy, Québec. At the Lethbridge location, plants were incubated in a mist-irrigated greenhouse (25°C, 95% humidity, 16 h photoperiod) for 3 days post-inoculation. For cv. Superb, one experiment was conducted at both the Ste-Foy and Lacombe locations, and eight were conducted at the Lethbridge location. For both GS-1-EM0040 and GS-1-EM0168, three experiments were conducted at the Lethbridge location. Inoculations were performed as described in Eudes et al. (2008), and disease spread was calculated as the total number of discolored spikelets, including the two inoculated spikelets.

For FHB index calculations, field evaluations were conducted, as described by Eudes et al. (2008), in Canadian nurseries: including CEROM (St-Hyacinthe, Québec), SEMICO (Ste-Rosalie, Québec) and Glenea (Winnipeg, Manitoba). Cultivar Superb was evaluated at both CEROM and SEMICO in 2006 and 2007, and at Glenea in 2006. GS-1-EM0040 was evaluated at CEROM in 2004, 2005, 2006 and 2007, at SEMICO in 2005, 2006 and 2007, and at Glenea in 2006. GS-1-EM0168 was evaluated at both CEROM and SEMICO in 2005 and 2007. Mature grain was harvested from field experiments and DON quantification was conducted as described in Eudes et al. (2008).

Treatments for the microarray

Four treatments were used in this study: water (ddH2O), FgTri5+, FgTri5− and DON. Fusarium graminearum FgTri5− (trichothecene non-producing strain GZT40) and FgTri5+ (wildtype strain GZ3639) were generously donated by Robert Proctor (Agriculture Research Services, United States Department of Agriculture, IL, USA). For FgTri5+ or FgTri5− containing inocula, four mycelium plugs (1 cm2 each) from FgTri5+ or FgTri5− cultured potato dextrose agar plates were used to inoculate 500 mL CMC broth (1·5% carboxymethylcellulose (Sigma C1011), 0·1% NH4NO3, 0·1% KH2PO4, 0·05% MgSO4·7H2O, 0·1% yeast extract; Cappellini & Peterson, 1965). The culture was incubated at room temperature under gentle agitation (150 rpm) for 2 weeks and then filtered through cheesecloth. The filtrate was washed with sterile ddH2O by centrifugation at 3220 g in a swinging bucket. After three washes, the pellet was diluted to a working concentration of 40 000 macroconidia mL−1. For DON-containing inocula, DON (D0156, Sigma-Aldrich) was dissolved in 40% ethanol to a concentration of 1 mg mL−1, and diluted to a working concentration of two parts per million (p.p.m.) in ddH2O.

Inoculations and tissue collection for the microarray

Plants were moved to a mist-irrigated greenhouse (25°C, 95% humidity, 16 h photoperiod) 1 day prior to anthesis. Two spikelets (one on each row) near the centre of the spike were point-inoculated at anthesis, by injection of 10 μL inoculum directly into the florets on either side of the spike. In order to reduce the variation in transcript profiles due to the circadian cycle, all inoculations took place between 07·00 and 10·00 h. Up to six spikelets above and below each inoculation point (for a total of up to 12 spikelets; see Fig. 1) were collected at 3, 8 and 24 h after inoculation (hai), flash frozen and stored at −80°C for RNA extractions. Zero hai uninoculated controls were also collected: up to 12 spikelets above/below the centre were harvested between 07·00 and 10·00 h. Three experimental repetitions were performed, each with three biological replicates, where experimental repetitions were differentiated by seeding dates. Each biological replicate consisted of one spike from one plant for a given condition. The three biological replicates (spikelets from three spikes) were pooled together for RNA extractions. Thus, for each experimental repetition, a total of 39 extractions were performed.

Figure 1.

 Inoculation of wheat spikes and harvesting of spikelets for RNA extractions. (a) The two central spikelets (four florets; shaded in figure) of the wheat head were each point-inoculated with 10 μL inoculum, and incubated in a mist-irrigated greenhouse. Spikelets above and below the inoculation point were collected at 3, 8 and 24 hai, and the remainder of the head was discarded. For 0 hai controls, plants were not inoculated, but spikelets were collected above and below the two central spikelets at anthesis. Total RNA was extracted from the collected spikelets. (b) Analysis of microarray results included four treatment comparisons: I, effect of Fusarium graminearum wildtype (FgTri5+); II, effect of trichothecene non-producing F. graminearum (FgTri5−); III, effect of DON; and IV, effect of FgTri5+ compared with FgTri5-.

Microarray

For each of the three experimental repetitions, harvested spikelets were ground under liquid nitrogen, and total RNA was extracted from 0·1 g of the resultant powder using QIAGEN’s RNeasy® Plant Mini kit with an on-column DNase digestion. RNA quality was verified on a 1% agarose gel and 5 μg of total RNA was used for biotin labelling. The two-cycle cDNA synthesis, cRNA labelling and hybridization to the Affymetrix GeneChip® Wheat Genome Array were done according to manufacturers’ instructions (Affymetrix). A total of 117 chips were used, corresponding to the 39 RNA extractions in each of three experimental repetitions. The Affymetrix CEL files resulting from scanning have been deposited in the Gene Expression Omnibus (NCBI) for public accessibility (accession GSE28973). CEL files were subjected to gcrma followed by MAS5 algorithms within ArrayAssist® (Stratagene) software for data transformation and normalization. Default settings were used in statistical analysis to calculate fold difference (FD) and a Bonferroni correction was applied based on a P-value <0·05. Treatment comparisons were made within each plant line, at each harvest time (3, 8 and 24 hai), with a 2·0 FD cut-off (< 0·05). Treatment comparisons are presented in Figure 1 and include: (i) effect of wildtype F. graminearum (FgTri5+); (ii) effect of the trichothecene non-producing F. graminearum (FgTri5−); (iii) effect of DON; and (iv) effect of FgTri5+ compared directly with FgTri5−. Comparisons were also made between plant lines for the 0 hai harvest with a 5·0 FD cut-off (< 0·05). Probe sets were initially annotated using an automated procedure by blastint against SwissProt, a non-redundant protein database, and the plant and fungal sections of the non-redundant nucleotide database (NCBI). Probe sets of potential interest were then manually annotated using the same program, but match(es) with homologues of a biochemically characterized gene in wheat or another plant species were preferentially selected over less characterized genes or genes from more distant relatives.

Real-time PCR

cDNA synthesis was performed on 2 μg total RNA using Invitrogen’s SuperScriptTM III Reverse Transcriptase according to manufacturer’s instructions, and the product was diluted 20-fold in nuclease-free water. Primers were designed to amplify gene products from eight unique probe sets; primer sequences were selected to have overlapping regions corresponding to the probes used in the Affymetrix GeneChip® (Table S1). The gene product corresponding to the housekeeping probe set AFFX-Ta_Sucsyn_5_at was selected as a housekeeping gene for quantitative real-time PCR (qPCR), as this gene showed little variation across gene chips for the various conditions used for microarrays, with an average signal for log-transformed data of 7·2 + 0·8. The qPCR reaction mix was set up using QIAGEN’s QuantiTectTM SYBR® Green Master Mix with 2 μL of cDNA, and 0·5 μm of both forward and reverse primers, in a final reaction volume of 15 μL. Samples were run on an Applied Biosystems 7900HT Fast Real-Time PCR System, with a 15 min hot-start at 95°C, followed by 40 cycles of 95°C (30 s) and 60°C (60 s). Fold-differences and standard error were calculated from Ct-values, normalized against the housekeeping gene (AFFX-Ta_Sucsyn_5_at), using QIAGEN’s rest 2006 software package.

Results

Affymetrix GeneChip® microarray analysis was used to investigate induced transcriptional differences in the wheat response upon exposure to different treatments (as defined in Fig. 1). Three wheat genotypes were evaluated, and included FHB-susceptible cv. Superb, and double haploid lines GS-1-EM0040 (CIMMYT 11/Superb*2), and GS-1-EM0168 (CM82036/Superb*2). GS-1-EM0168 was shown to carry the 3BS QTL, but was not positive for the 5A QTL derived from CM82036. Amplicon sizes for all QTLs screened in GS-1-EM0040 were similar to the parent cultivar Superb, thus the putative QTLs from CIMMYT 11 were not transferred to GS-1-EM0040. Phenotypic evaluation of these lines is presented in Table 1. Both double haploid lines were FHB resistant and showed reduced DON accumulation compared with cv. Superb. GS-1-EM0040 had the lowest FHB index, and GS-1-EM0168 showed the least amount of disease spread. All three lines were comparable for agronomic traits, except that GS-1-EM0168 was roughly 10 cm shorter than the other two lines, and both double haploid lines had slightly better yields compared with cv. Superb. Genotype-specific differences in constitutive gene expression were examined in the 0 hai uninoculated control samples. Genes showing at least 2 FD in constitutive expression (< 0·05) are listed in Table S2 and a subset of those showing 5 FD are presented in Table 2. In challenge experiments, all transcriptome analyses were conducted on the uninoculated spikelets from inoculated heads collected at 3, 8 and 24 hai. Four treatment comparisons (Fig. 1) within each genotype were made: (i) FgTri5+ vs. water (effect of FgTri5+); (ii) FgTri5− vs. water (effect of FgTri5−); (iii) DON vs. water (effect of DON); and (iv) FgTri5+ vs. FgTri5−. The number of challenge-dependent changes of at least 2 FD (< 0·05) are presented in Figure 2, and a subset of those genes which are discussed in the current report are presented in Table 3. A complete list of the challenge-dependent changes (2 FD, < 0·05) is presented in the supplementary materials (Table S3). Technical validation of the 2 FD cut-off value in microarray experiments were assessed by qPCR and are presented in Table 4.

Table 1.   Phenotypic evaluation of three wheat genotypes
Plant lineParentageQTLsFHB evaluationnAgronomy trial
Disease spreadFHB indexDON (ppm)Height (cm)Yield (kg ha−1)Maturity (days)Lodge
Suberb  8·3 ± 2·231·6 ± 14·58·9 ± 5·01287·81486·0102·01·8
GS-1-EM0040CIMMYT11/Superb*2 5·0 ± 1·910·7 ± 6·34·5 ± 2·11291·01525·1104·32·1
GS-1-EM0168CM82036/Superb*23BS2·7 ± 0·824·5 ± 6·55·5 ± 2·31278·31562·7103·61·4
Table 2.   Constitutive difference between transcriptomes of FHB-resistant genotypes and susceptible cv. Superb
Probe set IDPredicted gene functionGS-1-EM0040 vs. SuperbGS-1-EM0168 vs. Superb
FDP-valFDP -val
  1. Positive and negative values (FD > 5·0, < 0·05) indicate higher transcript accumulation in the resistant line and cv. Superb, respectively.

Cell signalling
 Ta.20570·1.A1_atSimilar to 1-aminocyclopropane-1-carboxylate oxidase [Sorghum bicolor]−16·70·002  
 TaAffx.27775·1.S1_atSimilar to receptor-like protein kinase 4 [Oryza sativa (japonica cultivar-group)]5·10·017  
Defence-related protein
 Ta.1282·4.S1_atLipid transfer protein 3 (LTP3) [Triticum aestivum]38·90·000  
 Ta.21646·1.S1_atSimilar to lipid transfer protein [O. sativa (japonica cultivar-group)]  −10·80·025
 Ta.24761·1.S1_atSimilar to NP_1878641 HSP70; ATP binding [Arabidopsis thaliana]−16·80·000  
 Ta.28695·6.S1_atMetallothionein (LOC542898)  −28·40·001
 Ta.352·1.S1_atSimilar to dehydration-responsive protein RD22 [O. sativa (japonica cultivar-group)]73·70·001  
 TaAffx.36658·1.S1_atSimilar to disease resistance protein Hcr2-5D [O. sativa (japonica cultivar-group)]5·90·006  
Hydrolysis
 Ta.10441·1.S1_atSimilar to phospholipase D delta (AtPLDdelta) (PLD delta)  5·50·012
 TaAffx.97452·1.A1_atSimilar to lipase [O. sativa (japonica cultivar-group)]−11·00·003  
 TaAffx.97535·1.S1_atSimilar to lipase [O. sativa (japonica cultivar-group)]−8·40·008  
Metabolism
 Ta.15908·1.S1_atSimilar to argininosuccinate lyase  −6·70·001
 Ta.27757·1.S1_atSimilar to phosphoglycerate mutase family, putative [O. sativa (japonica cultivar-group)]  −8·40·012
 Ta.332·1.S1_atSimilar to aldose reductase [O. sativa (japonica cultivar-group)]  8·30·014
Nucleic acid binding protein
 Ta.14115·2.S1_atSimilar to NP_1809941 UREG; metal ion binding/nickel ion binding/nucleotide binding [A. thaliana]  5·40·015
 Ta.30583·1.S1_atSimilar to XP_4666301 zinc finger protein [O. sativa (japonica cultivar-group)]−9·90·035  
 Ta.7269·1.S1_atSimilar to putative high mobility group protein (hmg1 gene), cultivar Wyuna, from endosperm tissue  6·70·005
 Ta.9409·1.S1_atSimilar to transcriptional coactivator p15 (PC4) family protein-like [O. sativa (japonica cultivar-group)]  11·60·007
Oxidase
 Ta.25832·1.A1_atSimilar to cytochrome P450 [T. aestivum]−14·60·000  
 Ta.5022·1.A1_atSimilar to NADH dehydrogenase (ubiquinone)5·70·020  
Phenylpropanoid metabolism
 Ta.25383·1.A1_atSimilar to hydroxycinnamoyl transferase [O. sativa (japonica cultivar-group)]  −6·20·004
Protein degradation
 Ta.22794·1.S1_atSKP1 protein [T. aestivum] (LOC543011)9·00·005  
 Ta.22794·1.S1_x_atSKP1 protein [T. aestivum] (LOC543011)8·50·011  
 Ta.7124·1.A1_atSimilar to serine protease-like protein [O. sativa (japonica cultivar-group)]  −8·80·001
Ribosomal protein
 Ta.28366·1.S1_a_atSimilar to NP_7043381 ribosomal protein, putative [Plasmodium falciparum 3D7]−5·50·003  
Structural protein
 Ta.7378·6.S1_atAlpha-tubulin (LOC543387)8·10·0012·60·033
 TaAffx.14498·1.S1_atAlpha/beta-gliadin precursor (LOC543192)56·50·000  
Transport protein
 Ta.9751·1.A1_atSimilar to ABC transporter [O. sativa]−13·20·000  
 TaAffx.104805·1.S1_atSimilar to coatomer protein complex, beta prime subunit [O. sativa (japonica cultivar-group)]  12·40·007
 TaAffx.113846·1.S1_s_atSimilar to aquaporin PIP1 (Pip1)259·10·000  
Protein with unknown function
 Ta.1038·1.S1_a_atUnknown18·40·000  
 Ta.11279·1.A1_atSimilar to NP_9220971 hypothetical protein [O. sativa (japonica cultivar-group)]  −11·30·012
 Ta.12402·1.S1_atSimilar to hypothetical protein [O. sativa (japonica cultivar-group)] gi|48717095|dbj|BAD22868·1|−6·20·000  
 Ta.13689·1.S1_atUnknown6·50·000  
 Ta.15830·1.S1_atUnknown  7·20·002
 Ta.24733·1.S1_atSimilar to ESTs AU077636 (E30684), AU064227 (E20182), AU082144 (E20182) correspond to a region of the predicted gene [O. sativa (japonica cultivar-group)]−6·80·001  
 Ta.24736·1.S1_atUnknown  8·30·039
 Ta.25945·1.S1_atUnknown11·50·002  
 Ta.28079·1.S1_x_atUnknown−5·40·023  
 Ta.28185·1.S1_x_atSimilar to NP_5652511 unknown protein [A. thaliana]  −5·50·008
 Ta.28236·1.S1_x_atSimilar to XP_4726411 OSJNBa0027P08·6 [O. sativa (japonica cultivar-group)]−6·30·017  
 Ta.28862·1.S1_atUnknown  5·20·005
 Ta.2963·1.S1_atUnknown11·30·000  
 Ta.3017·1.S1_atUnknown  5·40·013
 Ta.3017·1.S1_x_atUnknown  5·30·030
 Ta.30560·1.S1_atUnknown  27·90·001
 Ta.30560·2.S1_a_atUnknown  8·10·005
 Ta.3211·1.S1_atSimilar to O. sativa (japonica cultivar-group) cDNA clone:001-020-F07, full insert sequence  10·10·010
 Ta.3764·1.S1_atUnknown−6·50·001  
 Ta.3780·2.S1_a_atSimilar to NP_9155411 P0529E05·20 [O. sativa (japonica cultivar-group)]−8·80·001  
 Ta.481·1.A1_atSimilar to T. aestivum clone wl1.pk0003.h12:fis, full insert mRNA sequence  −5·30·004
 Ta.4978·2.A1_a_atUnknown  5·50·031
 Ta.6066·2.S1_a_atUnknown  7·30·003
 Ta.7304·1.A1_atSimilar to XP_4682341 hypothetical protein [O. sativa (japonica cultivar-group)]6·40·001  
 Ta.7919·1.S1_atUnknown  7·00·010
 Ta.8888·1.A1_atUnknown6·70·002  
 Ta.913·1.S1_atSimilar to O. sativa (japonica cultivar-group) cDNA clone:001-103-C02, full insert sequence−12·40·001  
 Ta.913·1.S1_x_atSimilar to O. sativa (japonica cultivar-group) cDNA clone:001-103-C02, full insert sequence−12·60·002  
 Ta.9590·1.S1_atUnknown  8·40·023
 TaAffx.100436·1.S1_atUnknown−8·80·004  
 TaAffx.104444·1.S1_atUnknown  14·60·003
 TaAffx.12620·1.A1_atUnknown  5·80·003
 TaAffx.134071·1.S1_atSimilar to hypothetical protein [O. sativa (japonica cultivar-group)]  −6·80·004
 TaAffx.144000·1.S1_s_atUnknown−5·20·031  
 TaAffx.16415·2.S1_atUnknown−6·00·001  
 TaAffx.25439·1.S1_s_atUnknown  5·10·019
 TaAffx.25558·1.S1_atUnknown  7·30·001
 TaAffx.31525·1.S1_atUnknown−5·60·003  
 TaAffx.43135·1.A1_x_atUnknown5·70·014  
 TaAffx.50853·1.S1_atUnknown  6·80·002
 TaAffx.54367·1.S1_atUnknown  11·10·000
 TaAffx.58902·1.S1_x_atSimilar to NP_5683981 unknown protein [A. thaliana]  −16·50·002
 TaAffx.59322·1.S1_atUnknown  5·40·017
 TaAffx.7810·2.S1_atUnknown5·80·0015·90·006
 TaAffx.78888·1.S1_atUnknown−5·70·011  
 TaAffx.78888·1.S1_x_atUnknown−5·30·011  
 TaAffx.8202·1.S1_atUnknown−9·60·004  
 TaAffx.92842·1.S1_atUnknown13·30·007  
Figure 2.

 Number of treatment-dependent differentially regulated transcripts in each plant line at 3, 8 and 24 hai. For each Venn diagram, the top left circle indicates the number of up (+) or down (−) regulated genes in GS-1-EM0040 for a given treatment comparison at a given harvest time; similarly, the top right circle represents GS-1-EM0168 and the bottom one represents cv. Superb. The overlapping regions between circles indicate the number of differentially regulated transcripts in common between plant lines.

Table 3.   Challenge-dependent changes in defence-related genes (FD > 2, < 0·05) in FHB-susceptible cv. Superb and resistant genotypes GS-1-EM0040 and GS-1-EM0168
Probe set IDPredicted gene functionTreatment comparisonhaiPlant lineFDP-val
  1. Positive and negative FD values indicate higher and lower transcript accumulation, respectively, in the FgTri5+ (FgTri5+ vs. water), FgTri5− (FgTri5− vs. water), DON (DON vs. water), and in FgTri5+ vs. FgTri5−.

Kinases
 TaAffx.110222·1.S1_x_atSimilar to leucine-rich repeat-containing extracellular glycoprotein [Sorghum bicolor] / somatic embryogenesis receptor kinase SERK [Medicago truncatula]DON vs. water24Superb2·20·0043
 Ta.4972·1.A1_atSimilar to aspartate kinase-homoserine dehydrogenase [Oryza sativa (japonica cultivar-group)]DON vs. water24Superb−2·10·0462
 Ta.3748·1.A1_atSimilar to hexokinase 1DON vs. water24Superb−2·20·0122
 Ta.10354·2.S1_x_atSimilar to pyrophosphate-fructose 6-phosphate 1-phosphotransferase alpha subunit (PFP) (6-phosphofructokinase, pyrophosphate dependent)DON vs. water24Superb−2·10·0263
 TaAffx.113624·2.S1_atSimilar to serine/threonine-protein kinase BRI1-like 3 precursor (Brassinosteroid insensitive 1-like protein 3)Tri5− vs. water8Superb2·30·0065
 Ta.3748·1.A1_atSimilar to hexokinase 1Tri5− vs. water24Superb−2·20·0149
 Ta.18587·1.S1_x_atSimilar to systemin receptor SR160 precursor (Brassinosteroid LRR receptor kinase) [O. sativa (japonica cultivar-group)]Tri5− vs. water24Superb−2·30·0002
 Ta.20980·2.S1_atSimilar to putative serine/threonine kinase receptor precursor (S-receptor kinase) (SRK)Tri5+ vs. Tri5−24Superb2·20·0118
 TaAffx.31923·1.S1_atSimilar to serine/threonine protein kinase [O. sativa (japonica cultivar-group)]Tri5+ vs. water3Superb2·00·0281
 TaAffx.113624·2.S1_atSimilar to serine/threonine-protein kinase BRI1-like 3 precursor (Brassinosteroid insensitive 1-like protein 3)Tri5+ vs. water8Superb2·70·0054
 TaAffx.12878·1.A1_atSimilar to wall-associated kinase 3 [Triticum aestivum]Tri5+ vs. water8Superb2·10·0065
 TaAffx.5982·1.S1_atSimilar to putative MAPKK kinase [Hordeum vulgare subsp. vulgare]DON vs. water3GS-1-EM0168−2·10·0445
 Ta.1684·3.S1_atSimilar to nucleoside diphosphate kinase III, chloroplast precursor (NDK III) (NDP kinase III) (NDPK III)Tri5− vs. water3GS-1-EM0168−2·10·0420
 Ta.19909·1.A1_atSimilar to putative MAP kinase activating protein C22orf5Tri5− vs. water8GS-1-EM01682·10·0094
 TaAffx.114172·1.S1_s_atSimilar to serine/threonine-specific protein kinase [O. sativa (japonica cultivar-group)]Tri5− vs. water24GS-1-EM0168−2·20·0096
 Ta.1684·3.S1_atSimilar to nucleoside diphosphate kinase III, chloroplast precursor (NDK III) (NDP kinase III) (NDPK III)Tri5+ vs. Tri5−3GS-1-EM01682·70·0094
 Ta.16965·2.A1_atSimilar to UNR-interacting protein (serine-threonine kinase receptor-associated protein)Tri5+ vs. Tri5−8GS-1-EM0168−2·10·0444
 Ta.4696·1.S1_atSimilar to receptor protein kinase-like protein [O. sativa (japonica cultivar-group)]Tri5+ vs. water8GS-1-EM01682·60·0286
 TaAffx.30003·1.S1_x_atSimilar to Aegilops tauschii protein kinase 1 mRNA, complete cdsDON vs. water8GS-1-EM00402·10·0033
 Ta.5481·1.S1_atSimilar to glucosidase II beta subunit precursor (protein kinase C substrate, 60·1 kDa protein, heavy chain) (PKCSH) (80K-H protein) (Vacuolar system)DON vs. water8GS-1-EM00402·40·0269
 Ta.29379·1.A1_atSimilar to kinase interacting protein 1 -like [O. sativa (japonica cultivar-group)]DON vs. water8GS-1-EM00402·00·0010
 TaAffx.4882·1.S1_atSimilar to receptor-type protein kinase LRK1 [O. sativa (japonica cultivar-group)]DON vs. water8GS-1-EM00402·00·0146
 TaAffx.50893·1.S1_s_atSimilar to serine/threonine protein kinase [T. aestivum]DON vs. water8GS-1-EM00402·30·0080
 TaAffx.104820·1.S1_atSimilar to receptor kinase [O. sativa (japonica cultivar-group)]DON vs. water24GS-1-EM0040−2·00·0441
 TaAffx.129414·2.S1_atSimilar to receptor-like protein kinase precursorDON vs. water24GS-1-EM0040−2·40·0317
 TaAffx.10874·1.S1_atSimilar to receptor-protein kinase [O. sativa (japonica cultivar-group)]DON vs. water24GS-1-EM0040−2·00·0101
 Ta.27812·1.A1_atSimilar to receptor-protein kinase [O. sativa (japonica cultivar-group)]DON vs. water24GS-1-EM0040−2·10·0076
 Ta.22638·1.A1_atYRK1 (LOC543113)DON vs. water24GS-1-EM0040−2·30·0297
 TaAffx.54339·1.S1_atSimilar to mitogen activated protein kinase homologue MMK2Tri5− vs. water3GS-1-EM00402·10·0080
 Ta.3322·3.S1_x_atSimilar to ankyrin protein kinase-like [Poa pratensis]Tri5− vs. water3GS-1-EM0040−2·20·0230
 Ta.6954·3.S1_x_atSimilar to choline kinase [O. sativa (japonica cultivar-group)]Tri5− vs. water3GS-1-EM0040−2·40·0374
 TaAffx.38017·1.A1_atSimilar to receptor-type protein kinase LRK1 [O. sativa (japonica cultivar-group)]Tri5− vs. water3GS-1-EM0040−2·00·0113
 TaAffx.52905·1.S1_x_atPhosphoribulokinase (LOC543283)Tri5− vs. water8GS-1-EM00402·30·0016
 TaAffx.101059·2.S1_atSimilar to 4-diphosphocytidyl-2-C-methyl-d-erythritol kinase, chloroplast precursor (CMK)Tri5− vs. water8GS-1-EM00402·20·0095
 TaAffx.37109·1.S1_atSimilar to receptor protein kinase [Arabidopsis thaliana]Tri5− vs. water8GS-1-EM00402·00·0073
 TaAffx.12878·1.A1_atSimilar to wall-associated kinase 3 [T. aestivum]Tri5− vs. water8GS-1-EM00402·40·0004
 Ta.8249·3.S1_atSimilar to calmodulin-domain protein kinase [O. sativa (japonica cultivar-group)]Tri5− vs. water24GS-1-EM00402·30·0247
 TaAffx.56854·1.S1_atSimilar to serine/threonine kinase receptor precursor-like protein [O. sativa (japonica cultivar-group)]Tri5− vs. water24GS-1-EM00402·00·0050
 TaAffx.5699·1.S1_atSimilar to putative receptor protein kinase CRINKLY4 precursorTri5+ vs. Tri5−3GS-1-EM00402·00·0302
 TaAffx.54339·1.S1_atSimilar to mitogen activated protein kinase homologue MMK2Tri5+ vs. Tri5−3GS-1-EM0040−2·30·0140
 TaAffx.101059·2.S1_atSimilar to 4-diphosphocytidyl-2-C-methyl-d-erythritol kinase, chloroplast precursor (CMK)Tri5+ vs. Tri5−8GS-1-EM0040−2·10·0233
 TaAffx.59615·1.S1_atSimilar to receptor-like protein kinase [O. sativa (japonica cultivar-group)]Tri5+ vs. Tri5−24GS-1-EM0040−2·10·0208
 TaAffx.4882·1.S1_atSimilar to receptor-type protein kinase LRK1 [O. sativa (japonica cultivar-group)]Tri5+ vs. Tri5−24GS-1-EM0040−2·10·0256
 TaAffx.84282·1.S1_atSimilar to CDPK-related protein kinase (PK421)Tri5+ vs. water3GS-1-EM0040−2·40·0001
 Ta.19909·1.A1_atSimilar to putative MAP kinase activating protein C22orf5Tri5+ vs. water3GS-1-EM0040−2·50·0264
 TaAffx.64399·1.S1_atSimilar to receptor-like kinase RHG1 [Glycine max]Tri5+ vs. water3GS-1-EM0040−2·80·0030
 Ta.5481·1.S1_atSimilar to glucosidase II beta subunit precursor (protein kinase C substrate, 60·1 kDa protein, heavy chain) (PKCSH) (80K-H protein) (Vacuolar system)Tri5+ vs. water8GS-1-EM00402·50·0043
 Ta.6683·1.A1_x_atSimilar to mitogen activated protein kinase kinase kinase 12 (Leucine-zipper protein kinase) (ZPK) (Dual leucine zipper bearing kinase) (DLK)Tri5+ vs. water8GS-1-EM00402·70·0402
 Ta.5204·1.S1_atSimilar to NP_2006381 ATP binding / kinase/ protein serine/threonine kinase [A. thaliana]Tri5+ vs. water8GS-1-EM00402·20·0221
 Ta.1357·2.A1_atSimilar to protein kinase [O. sativa (japonica cultivar-group)]Tri5+ vs. water8GS-1-EM00402·00·0309
 Ta.3631·2.S1_atSimilar to putative 3,4-dihydroxy-2-butanone kinaseTri5+ vs. water8GS-1-EM00402·20·0346
 TaAffx.18131·1.S1_atSimilar to putative phytosulfokine receptor precursor (Phytosulfokine LRR receptor kinase)Tri5+ vs. water8GS-1-EM00402·10·0481
 Ta.18665·1.S1_atSimilar to putative protein kinase G11A [O. sativa (japonica cultivar-group)] gi|55296796|dbj|BAD68122·1| putative protein kinase G11A [O. sativa (japonica)Tri5+ vs. water8GS-1-EM00402·00·0096
 TaAffx.37109·1.S1_atSimilar to receptor protein kinase [A. thaliana]Tri5+ vs. water8GS-1-EM00402·40·0029
 TaAffx.12878·1.A1_atSimilar to wall-associated kinase 3 [T. aestivum]Tri5+ vs. water24GS-1-EM0040−2·10·0147
Ribosomal protein
 TaAffx.107003·1.S1_atSimilar to 50S ribosomal protein L4, chloroplast (CL4) [A. thaliana]DON vs. water3Superb−2·00·0156
 TaAffx.128418·38.S1_atSimilar to 18S Soybean (G. max) 18S ribosomal RNADON vs. water3Superb−2·00·0002
 TaAffx.128418·24.S1_atSimilar to 28S ribosomal RNA [T. aestivum]DON vs. water8Superb−2·60·0298
 TaAffx.35350·1.S1_atSimilar to ribosomal protein S4 [Panax ginseng]Tri5− vs. water3Superb2·20·0008
 Ta.12751·3.A1_atRibosomal protein L13a (LOC606336)Tri5+ vs. Tri5−8Superb2·30·0122
 TaAffx.6593·2.S1_atSimilar to chloroplast 50S ribosomal protein L33Tri5+ vs. water24Superb−2·10·0095
 TaAffx.108878·1.S1_x_atSimilar to 60S ribosomal protein L5 [Neurospora crassa]Tri5+ vs. water24Superb−2·10·0186
 TaAffx.111195·1.S1_atSimilar to 5S ribosomal RNA [T. aestivum]DON vs. water3GS-1-EM0168−3·20·0217
 Ta.30687·2.S1_x_atRibosomal protein S29 (LOC606342)DON vs. water8GS-1-EM01682·20·0368
 Ta.9507·2.S1_atRibosomal protein L6 (LOC606333)DON vs. water8GS-1-EM01682·10·0271
 Ta.9507·2.S1_x_atRibosomal protein L6 (LOC606333)DON vs. water8GS-1-EM01682·00·0258
 Ta.30687·2.S1_x_atRibosomal protein S29 (LOC606342)DON vs. water24GS-1-EM0168−4·30·0330
 TaAffx.56089·1.S1_atUbiquitin/ribosomal fusion protein (LOC542909)Tri5− vs. water3GS-1-EM01682·00·0222
 TaAffx.111195·1.S1_atSimilar to 5S ribosomal RNA [T. aestivum]Tri5− vs. water3GS-1-EM0168−2·20·0486
 Ta.30687·2.S1_x_atRibosomal protein S29 (LOC606342)Tri5− vs. water8GS-1-EM01683·00·0214
 TaAffx.129824·5.S1_x_atSimilar to ribosomal protein S11 [T. aestivum]Tri5− vs. water24GS-1-EM0168−3·50·0487
 TaAffx.128896·21.S1_x_atSimilar to chloroplast 30S ribosomal protein S8Tri5− vs. water24GS-1-EM0168−2·40·0196
 TaAffx.129824·1.S1_s_atSimilar to chloroplast 30S ribosomal protein S11Tri5− vs. water24GS-1-EM0168−2·10·0439
 TaAffx.129824·9.S1_x_atSimilar to 30S ribosomal protein S11Tri5− vs. water24GS-1-EM0168−3·50·0455
 Ta.30687·2.S1_x_atRibosomal protein S29 (LOC606342)Tri5− vs. water24GS-1-EM0168−2·60·0082
 Ta.30687·2.S1_x_atRibosomal protein S29 (LOC606342)Tri5+ vs. Tri5−3GS-1-EM0168−3·00·0308
 Ta.30687·2.S1_x_atRibosomal protein S29 (LOC606342)Tri5+ vs. Tri5−24GS-1-EM01682·40·0109
 Ta.30687·2.S1_x_atRibosomal protein S29 (LOC606342)Tri5+ vs. water3GS-1-EM0168−4·00·0131
 Ta.13787·1.S1_x_atSimilar to plastid-specific 30S ribosomal protein 1, chloroplast precursor (CS-S5) (CS5) (S22) (ribosomal protein 1) (PSRP-1)Tri5+ vs. water24GS-1-EM0168−2·30·0321
 TaAffx.111195·1.S1_atSimilar to 5S ribosomal RNA [T. aestivum]DON vs. water8GS-1-EM00403·00·0124
 Ta.9507·2.S1_atRibosomal protein L6 (LOC606333)DON vs. water8GS-1-EM00402·70·0370
 Ta.9507·2.S1_x_atRibosomal protein L6 (LOC606333)DON vs. water8GS-1-EM00402·70·0264
 TaAffx.128418·24.S1_atSimilar to 28S ribosomal RNA [T. aestivum]Tri5− vs. water8GS-1-EM00402·40·0324
 TaAffx.129824·5.S1_x_atSimilar to ribosomal protein S11 [T. aestivum]Tri5+ vs. water8GS-1-EM00402·40·0410
 TaAffx.107897·1.S1_atSimilar to chloroplast 30S ribosomal protein S18Tri5+ vs. water8GS-1-EM00402·20·0385
 Ta.10990·1.A1_atSimilar to 40S ribosomal protein S19Tri5+ vs. water8GS-1-EM00402·00·0192
 TaAffx.113782·1.S1_atSimilar to 60S ribosomal protein L11-1 (L16A)Tri5+ vs. water24GS-1-EM0040−2·00·0218
Phenylpropanoid pathway
 Ta.7022·1.S1_atSimilar to phenylalanine ammonia-lyaseDON vs. water3Superb−2·10·0260
 Ta.7022·1.S1_x_atSimilar to phenylalanine ammonia-lyaseDON vs. water3Superb−2·30·0164
 TaAffx.111664·1.S1_atSimilar to cinnamyl alcohol dehydrogenase [O. sativa (japonica cultivar-group)]Tri5− vs. water3GS-1-EM01682·00·0064
 Ta.9220·3.S1_atSimilar to phenylalanine ammonia-lyaseDON vs. water8GS-1-EM00402·00·0130
 Ta.9220·3.S1_x_atSimilar to phenylalanine ammonia-lyaseDON vs. water8GS-1-EM00402·60·0157
 TaAffx.84154·1.S1_atSimilar to phenylalanine ammonia-lyaseDON vs. water8GS-1-EM00402·30·0197
 Ta.3609·1.S1_a_atSimilar to NP_1764821 4-coumarate-CoA ligase [A. thaliana]DON vs. water8GS-1-EM00402·10·0012
 Ta.10418·2.S1_x_atSimilar to chalcone synthase [O. sativa (japonica cultivar-group)]DON vs. water8GS-1-EM00402·20·0333
 Ta.9220·3.S1_x_atSimilar to phenylalanine ammonia-lyaseTri5− vs. water8GS-1-EM00402·70·0027
 Ta.8086·1.A1_atSimilar to hydroxycinnamoyl transferase [O. sativa (japonica cultivar-group)]Tri5− vs. water8GS-1-EM00402·30·0358
 Ta.9172·3.S1_atSimilar to naringenin-chalcone synthaseTri5− vs. water24GS-1-EM00402·50·0495
 Ta.9220·3.S1_x_atSimilar to phenylalanine ammonia-lyaseTri5+ vs. water8GS-1-EM00402·10·0025
Pathogenesis-related
 Ta.29496·2.S1_atSimilar to peroxidase 12 precursor (Atperox P12) (PRXR6) (ATP4a)DON vs. water8Superb−2·20·0371
 Ta.14461·3.S1_x_atSimilar to nectarin 1 precursor (Superoxide dismutase [Mn])DON vs. water8Superb−2·60·0240
 Ta.27762·1.S1_x_atThaumatin-like protein (Ta-TLP)DON vs. water24Superb2·80·0285
 TaAffx.116570·1.S1_atPathogenesis-related protein 4 (PR4)DON vs. water24Superb2·20·0497
 Ta.24715·1.S1_atPeroxidase (pox3)DON vs. water24Superb2·00·0373
 Ta.24501·1.S1_atThaumatin-like protein (LOC543292)DON vs. water24Superb2·50·0298
 Ta.82·1.S1_atPeroxidase (LOC543287)DON vs. water24Superb4·10·0072
 Ta.28354·3.S1_x_atGlutathione transferase (gstu3)DON vs. water24Superb2·30·0055
 Ta.21307·1.S1_x_atSimilar to peroxidase [O. sativa (japonica cultivar-group)]DON vs. water24Superb2·30·0156
 Ta.8304·1.S1_x_atSimilar to pathogenesis-related PR1a [Triticum monococcum]DON vs. water24Superb2·10·0099
 Ta.14946·1.S1_atSimilar to NP_5664261 ATHCHIB (basic chitinase) [A. thaliana]DON vs. water24Superb5·80·0101
 Ta.21120·1.S1_atSimilar to glucan endo-1,3-beta-d-glucosidaseDON vs. water24Superb3·20·0014
 Ta.22562·1.S1_atSimilar to glucan endo-1,3-beta-d-glucosidaseDON vs. water24Superb2·50·0023
 TaAffx.32266·1.A1_atSimilar to probable phospholipid hydroperoxide glutathione peroxidase (PHGPx) (Salt-associated protein)DON vs. water24Superb−2·70·0474
 Ta.21386·1.S1_atSimilar to probable nonspecific lipid-transfer protein 2 (LTP 2)Tri5+ vs. Tri5−8Superb2·10·0484
 TaAffx.78864·1.S1_atSimilar to glutathione S-transferase [O. sativa (japonica cultivar-group)]Tri5+ vs. Tri5−8Superb−2·30·0109
 Ta.5235·1.S1_x_atPeroxidase precursor (prx)Tri5+ vs. Tri5−24Superb2·10·0033
 Ta.82·1.S1_atPeroxidase (LOC543287)Tri5+ vs. water24Superb2·70·0273
 Ta.21120·1.S1_atSimilar to glucan endo-1,3-beta-d-glucosidaseTri5+ vs. water24Superb3·10·0166
 Ta.22562·1.S1_atSimilar to glucan endo-1,3-beta-d-glucosidaseTri5+ vs. water24Superb2·10·0447
 Ta.30944·1.S1_s_atGlutathione transferase F2 (gstf2)DON vs. water8GS-1-EM01682·70·0346
 TaAffx.108531·2.S1_atSimilar to peroxidaseDON vs. water8GS-1-EM01682·30·0014
 Ta.18560·1.S1_s_atSimilar to class III peroxidase [O. sativa (japonica cultivar-group)]DON vs. water8GS-1-EM01682·10·0115
 Ta.18647·1.S1_s_atSimilar to probable nonspecific lipid-transfer protein AKCS9 precursor (LTP)DON vs. water24GS-1-EM0168−2·10·0248
 Ta.8828·3.S1_a_atSimilar to peroxidase [Ananas comosus]Tri5− vs. water3GS-1-EM01682·40·0381
 Ta.4876·1.A1_x_atPeroxidase (Pox4)Tri5− vs. water24GS-1-EM0168−2·20·0110
 TaAffx.113452·1.S1_atSimilar to chitinase [T. aestivum]Tri5− vs. water24GS-1-EM0168−2·20·0251
 Ta.14850·1.S1_atSimilar to glutathione S-transferase [O. sativa (japonica cultivar-group)]Tri5+ vs. Tri5−3GS-1-EM0168−2·10·0423
 Ta.25024·1.S1_x_atSimilar to w.s.to NP_1773131 peroxidase [A. thaliana]Tri5+ vs. Tri5−8GS-1-EM0168−2·00·0293
 Ta.5385·1.S1_atPeroxidase (POX2)Tri5+ vs. Tri5−24GS-1-EM01682·70·0215
 Ta.27389·1.S1_atSimilar to gamma-purothionin- poulard wheatTri5+ vs. Tri5−24GS-1-EM01682·10·0029
 Ta.30501·1.S1_atSimilar to chitinaseTri5+ vs. Tri5−24GS-1-EM01682·40·0307
 Ta.14281·1.S1_atDefensin (Tad1)DON vs. water8GS-1-EM00402·10·0474
 TaAffx.32251·1.S1_atSimilar to nonspecific lipid-transfer protein 4·3 precursor (LTP 4·3)DON vs. water8GS-1-EM00402·10·0218
 Ta.28·1.S1_atGlucan endo-1,3-beta-d-glucosidase (LOC543330)DON vs. water24GS-1-EM0040−2·50·0117
 Ta.28354·3.S1_x_atGlutathione transferase (gstu3)DON vs. water24GS-1-EM0040−2·30·0480
 Ta.13013·2.S1_x_atSimilar to pathogenesis-related protein 10b [S. bicolor]DON vs. water24GS-1-EM0040−2·40·0041
 Ta.303·3.S1_x_atSimilar to glutathione transferaseDON vs. water24GS-1-EM0040−2·50·0428
 Ta.26048·1.S1_x_atSimilar to glucan endo-1,3-beta-d-glucosidaseDON vs. water24GS-1-EM0040−2·50·0132
 Ta.23376·2.S1_s_atSimilar to peroxidase 47 precursor (Atperox P47) (ATP32)Tri5− vs. water3GS-1-EM0040−2·10·0177
 Ta.11124·1.A1_atSimilar to glucan endo-1,3-beta-d-glucosidase [H. vulgare subsp. vulgare]Tri5− vs. water3GS-1-EM0040−2·30·0340
 Ta.4876·1.A1_x_atPeroxidase (pox4)Tri5− vs. water8GS-1-EM00402·30·0453
 Ta.4328·1.S1_atSimilar to pathogenesis-related protein 10b [S. bicolor]Tri5− vs. water8GS-1-EM00402·30·0150
 Ta.761·1.S1_atSimilar to glutathione S-transferase [O. sativa (japonica cultivar-group)]Tri5− vs. water8GS-1-EM00402·40·0381
 Ta.26983·1.A1_atSimilar to chitinaseTri5− vs. water8GS-1-EM00402·40·0375
 Ta.30739·1.A1_atSimilar to pathogenesis-related 1a [T. monococcum]Tri5+ vs. Tri5−8GS-1-EM0040−2·00·0384
 Ta.4601·2.S1_atBeta-glucosidase (TaGlu1a)Tri5+ vs. water8GS-1-EM00402·60·0329
 Ta.14281·1.S1_atDefensin (Tad1)Tri5+ vs. water8GS-1-EM00402·40·0136
 Ta.14580·1.S1_atSimilar to peroxidase precursorTri5+ vs. water8GS-1-EM00402·20·0120
 Ta.30755·1.S1_atSimilar to nonspecific lipid-transfer protein 2G (LTP2G) (lipid transfer protein 2 isoform 1) (LTP2-1) (7 kDa lipid transfer protein 1)Tri5+ vs. water8GS-1-EM00402·40·0435
 Ta.13754·1.S1_s_atSimilar to lipid transfer protein-related [A. thaliana]Tri5+ vs. water8GS-1-EM00402·60·0434
 TaAffx.86266·1.S1_atSimilar to beta-glucosidase homologue precursorTri5+ vs. water8GS-1-EM00402·10·0381
 Ta.8258·1.S1_x_atSimilar to type 2 non-specific lipid transfer protein precursor [T. aestivum]Tri5+ vs. water24GS-1-EM0040−2·50·0397
 Ta.8258·2.S1_atSimilar to type 2 non-specific lipid transfer protein [T. aestivum]Tri5+ vs. water24GS-1-EM0040−2·50·0417
Jasmonic acid signalling
 Ta.526·1.S1_x_atSimilar to lipoxygenaseDON vs. water3Superb−2·80·0095
 TaAffx.58772·1.S1_atSimilar to 12-oxophytodienoate reductase 3 (12-oxophytodienoate-10,11-reductase 3) (OPDA-reductase 3) (LeOPR3)Tri5− vs. water8Superb2·20·0006
 Ta.1967·1.S1_x_atSimilar to lipoxygenaseTri5− vs. water3GS-1-EM01682·90·0492
 Ta.12757·1.A1_atSimilar to lipoxygenase-like protein (lox gene) [H. vulgare subsp. vulgare]Tri5− vs. water24GS-1-EM0168−2·20·0171
 Ta.1207·1.S1_atSimilar to oxo-phytodienoic acid reductase [O. sativa (japonica cultivar-group)]Tri5+ vs. Tri5−3GS-1-EM01682·20·0434
 Ta.526·1.S1_x_atSimilar to lipoxygenaseTri5+ vs. Tri5−3GS-1-EM0168−2·00·0380
 TaAffx.134501·1.A1_atSimilar to lipoxygenaseTri5+ vs. Tri5−24GS-1-EM01682·20·0479
 Ta.1207·1.S1_atSimilar to oxo-phytodienoic acid reductase [O. sativa (japonica cultivar-group)]Tri5+ vs. water3GS-1-EM01683·20·0037
 Ta.1207·1.S1_x_atSimilar to oxo-phytodienoic acid reductase [O. sativa (japonica cultivar-group)]Tri5+ vs. water3GS-1-EM01682·20·0322
 Ta.1207·1.S1_atSimilar to oxo-phytodienoic acid reductase [O. sativa (japonica cultivar-group)]DON vs. water24GS-1-EM0040−2·10·0150
 Ta.12757·1.A1_atSimilar to lipoxygenase-like protein (lox gene) [H. vulgare subsp. vulgare]DON vs. water24GS-1-EM0040−2·40·0153
 TaAffx.128684·1.S1_x_atSimilar to 12-oxo-phytodienoic acid reductase [Zea mays]DON vs. water24GS-1-EM0040−2·10·0222
 Ta.30827·1.A1_x_atSimilar to 32·6 kDa jasmonate-induced protein [H. vulgare]Tri5− vs. water24GS-1-EM00402·20·0438
Ethylene signalling
 TaAffx.93223·1.A1_atSimilar to 1-aminocyclopropane-1-carboxylate oxidase (ACC oxidase) (Ethylene-forming enzyme) (EFE)DON vs. water8Superb−2·10·0253
 Ta.6397·1.A1_atSimilar to Adenosylmethionine decarboxylaseDON vs. water24Superb−2·20·0018
 Ta.4470·1.S1_atSimilar to ethylene-binding protein-like [O. sativa (japonica cultivar-group)]/AP2 domain-containing transcription factor-like [O. sativa (japonica cultivar-group)]Tri5− vs. water3Superb2·30·0490
 TaAffx.57475·1.S1_x_atSimilar to methionine adenosyltransferaseTri5− vs. water3GS-1-EM01682·10·0473
 TaAffx.128576·1.S1_atSimilar to ethylene-forming enzyme [O. sativa (japonica cultivar-group)]Tri5+ vs. water8GS-1-EM00402·00·0191
Table 4.   qPCR validation of twofold difference (FD) in microarray
Probe set IDGene descriptionPlant lineComparisonhaiMicroarrayqPCR
FDP-valFDError
  1. Positive and negative FD values indicate up- and down-regulation, respectively, in the given treatment comparison.

TaAffx.106139·1.S1_atUnknownSuperbFgTri5− vs. water24−2·10·004−2·4±0·16
Ta.526·1.S1_x_atLipoxygenaseGS-1-EM0168FgTri5+ vs. FgTri5−3−2·00·038−3·1±0·92
TaAffx.108735·1.S1_atUnknownGS-1-EM0040FgTri5+ vs. FgTri5−8+2·10·003+2·2±0·27
Ta.1207·1.S1_atSimilar to oxo-phytodienoic acid reductase [O. sativa (japonica cultivar-group)]GS-1-EM0168FgTri5+ vs. FgTri5−3+2·20·043+4·2±0·06
Ta.16723·2.S1_x_atSimilar to unknown protein [O. sativa (japonica cultivar-group)] gi|48716457|GS-1-EM0040DON vs. water8+2·20·020+2·4±1·54
Ta.24501·1.S1_atThaumatin-like proteinSuperbDON vs. water24+2·50·030+2·2±0·68
Ta.1967·1.S1_x_atLipoxygenaseGS-1-EM0168FgTri5− vs. water3+2·90·049+2·3±0·78
TaAffx.111195·1.S1_at5S ribosomal RNA [T. aestivum]GS-1-EM0040DON vs. water8+3·00·012+2·0±0·49
Ta.1207·1.S1_atSimilar to oxo-phytodienoic acid reductase [O. sativa (japonica cultivar-group)]GS-1-EM0168FgTri5+ vs. water3+3·20·004+3·3±0·98

Constitutive differences in gene expression

Expression levels of transcripts corresponding to 174 and 156 unique probe sets were constitutively different (FD > 2·0, < 0·05) in spikelets of cv. Superb vs. GS-1-EM0040 and cv. Superb vs. GS-1-EM0168, respectively (Table S2; FD > 5·0 presented in Table 2). Among the constitutive differences between GS-1-EM0168 and cv. Superb was the observed lower expression in GS-1-EM0168 of two defence-related transcripts, including one with similarity to rice lipid transfer protein (LTP; 10·8 FD) and one annotated as a metallothionein (28·4 FD). The most dramatic constitutive difference observed was of a wheat aquaporin PIP1, which was highly over-expressed (259·1 FD) in GS-1-EM0040 compared with cv. Superb. Other transcripts that were over-expressed in GS-1-EM0040, compared with cv. Superb, with potential connections to plant defence, include a gene with similarity to a dehydration-responsive protein from rice (RD22; 73·7 FD) and a wheat LTP (TaLTP3; 38·9 FD).

Challenge-induced differences in global gene expression

Differences in transcript accumulation at >2·0 FD were observed in all treatment comparisons and at all harvest times (Fig. 2) with an average of 36 differentially regulated transcripts for any combination of genotypes and treatments. The largest number of expression differences was observed in GS-1-EM0040, whereas the number of differences observed in GS-1-EM0168 and susceptible cv. Superb genotypes were lower and comparable in the two genotypes. Additionally, by 24 hai no up-regulation of transcripts was observed in GS-1-EM0168 in response to the three elicitors evaluated (FgTri5+, FgTri5− or DON). DON treatment led mainly to down-regulation of transcripts in all wheat lines at 3 hai. In cv. Superb the strong DON-induced down-regulation was also observed at 8 hai, and by 24 hai, two-thirds of the differentially regulated genes were down-regulated. In the resistant genotypes, DON-induced down-regulation was also observed at 24 hai; but at 8 hai, a total of 122 genes were up-regulated and none were down-regulated in GS-1-EM0040, while 40 genes were up-regulated and six down-regulated in GS-1-EM0168.

Challenge-induced differential gene expression of defence response pathways

Genes displaying differential expression are associated with several distinct pathways. Some of these genes are known to be involved directly in the plant defence response, or potentially involved in regulating defence responses (Table 3). Differential expression of kinases was observed in all three wheat lines. A total of 38 differences were observed across the various treatment conditions in GS-1-EM0040. These differences occurred in response to all treatment comparisons (less so in the FgTri5+ vs. FgTri5− comparison). In contrast, only eight and 11 differences were observed across the various conditions in GS-1-EM0168 and cv. Superb, respectively.

Differential expression of genes encoding ribosomal components was also observed. At 8 hai in GS-1-EM0040, all three elicitors (FgTri5+, FgTri5− and DON) induced up-regulation of ribosomal genes. In GS-1-EM0168, DON (8 hai) and FgTri5− (3 and 8 hai) also induced up-regulation of similar genes. Down-regulation of these genes was observed in susceptible cv. Superb in response both to DON (3 and 8 hai) and FgTri5+ (24 hai).

Differential expression of PR genes was observed in all three wheat genotypes, but occurred mainly in GS-1-EM0040 and cv. Superb (Table 3). At 8 hai, challenge with the fungus induced PR gene up-regulation in GS-1-EM0040, including one chitinase (Chi; FgTri5−), three glucanases (Glu; FgTri5+), one defensin (FgTri5+) and one peroxidase (POX; FgTri5+ vs. FgTri5−). FgTri5+ induced up-regulation of three PR genes (two Glu and one POX) in cv. Superb by 24 hai. DON also induced up-regulation of PR genes, including eight genes in cv. Superb (24 hai), three in GS-1-EM0168 and another three in GS-1-EM0040 (8 hai).

Differential expression of phenylpropanoid pathway genes was primarily limited to GS-1-EM0040, where up-regulation of two genes encoding l-phenylalanine ammonia-lyase (PAL) was observed at 8 hai: one of these genes (probe sets Ta.845145·1.S1_at) occurred in response to DON, and the other (probe sets Ta.9220.S1_at and Ta.9220.S1_x_at) was induced by DON, FgTri5+ and FgTri5− (Table 3). In susceptible cv. Superb, DON induced down-regulation of a different PAL (probe sets Ta.7022·1.S1_ at and Ta.7022·1.S1_x_at) at 3 hai. PAL is the entry point to phenylpropanoid metabolism and is involved in the formation of a variety of phenolic metabolites, including lignins, flavonoids and SA (Dixon et al., 2002). In GS-1-EM0040, up-regulation of genes involved in both lignin and flavonoid biosynthesis was observed in response to both FgTri5− and DON. Up-regulation of genes whose products have often been associated with SA signalling in Arabidopis, but not necessarily in wheat (Huckelhoven et al., 1999; Molina et al., 1999; Yu & Muehlbauer, 2001; Lu et al., 2006), were an ankyrin kinase-like protein that was down-regulated in GS-1-EM0040 (FgTri5− at 3 hai) and the SA-regulated PR-1, which was up-regulated in cv. Superb 24 hai after point inoculation with DON.

Jasmonic acid (JA) synthesis involves a well-characterized multistep pathway that converts linolenic acid to the final hormone. Among the corresponding genes, up-regulation of oxo-phytodienoic acid reductase was observed in GS-1-EM0168 in response to FgTri5+ (3 hai) when compared to either FgTri5− or the water controls. Two genes encoding lipoxygenases, a class of enzymes thought to initiate JA biosynthesis, were also up-regulated in response to the FgTri5+ treatment when compared with water (3 hai) or FgTri5− (24 hai) (Table 3). Up-regulation of JA biosynthesis or responsive genes was also observed in cv. Superb and GS-1-EM0040 genotypes, but only in response to FgTri5− treatment. Point inoculation with DON, on the other hand, induced down-regulation of JA signalling genes in both GS-1-EM0040 (24 hai) and cv. Superb (3 hai), but not in GS-1-EM0168.

Quantitative real-time PCR validation of microarrays

Eight genes from nine comparisons were randomly selected from the microarray data set with >2·0 FD for confirmation of the expression patterns by an alternative method, quantitative real-time PCR (qPCR). While the absolute expression FD values obtained by qPCR differed from those reported in the microarray data, all nine comparisons showed the same direction of change in the qPCR and microarray data sets, and displayed FD values >2·0 (Table 4).

Discussion

The differential transcriptomes of three wheat genotypes were evaluated both before and after challenge with different strains of FHB and DON (Fig. 1), using Affymetrix GeneChip® microarray technology. These wheat lines differ in their susceptibility to FHB, but are genetically related through the use of the FHB-susceptible cv. Superb as a common parent. The two double haploid lines GS-1-EM0040 (CIMMYT 11/Superb*2) and GS-1-EM0168 (CM82036/Superb*2), showed improved resistance compared with cv. Superb, which was shown to be moderately susceptible (Table 1). GS-1-EM0168 has the 3BS QTL and showed very little, if any, disease spread, indicating classical Type 2 resistance. More disease spread was observed in GS-1-EM0040 compared with GS-1-EM0168, although a reasonable amount of resistance was still detected, suggesting moderate levels of Type 2 resistance in this line. Both double haploid lines had lower FHB index values compared with cv. Superb, but the value in GS-1-EM0040 (10·7 + 6·3) was even lower than that of GS-1-EM0168 (24·5 + 6·5). FHB index is used as an indicator of the combined effect of both Type 1 and Type 2 resistances (Foroud & Eudes, 2009), and as the Type 2 resistance is higher in GS-1-EM0168, the lower FHB index in GS-1-EM0040 suggests a higher level of Type 1 resistance in GS-1-EM0040. Both double haploid lines also showed reduced DON accumulation compared with cv. Superb.

Among the observed genotype-dependent differences was the higher constitutive expression of a Type 1 lipid transfer protein (LTP), TaLTP3, and a plasma membrane aquaporin, TaPIP1, in GS-1-EM0040 compared with cv. Superb (Table 2). While LTPs have been demonstrated to inhibit growth of some microorganisms, including Fusarium species (Sun et al., 2008), preliminary experiments suggest that TaLTP3 does not inhibit growth of the F. graminearum strain GZ3639 (N.A. Foroud, F. Eudes and B.E. Ellis, unpublished data). Some LTPs have been implicated in cuticular wax deposition at the plant cell wall (Samuels et al., 2008), and PIP over-expression has been reported to lead to increased cuticle thickness in rice (Hanba et al., 2004). The primary mode of invasion by FHB-causing species is by direct penetration of the host epidermis accompanied by cuticular degradation (Kang & Buchenauer, 2000b), presumably mediated by the activity of the cutinases and lipases previously implicated in FHB disease progress (Voigt et al., 2005). Furthermore, F. graminearum germinates and grows more easily on leaves of A. thaliana CER1, a cuticular wax-deficient mutant, than on wildtype plants (T. Ouellet and V. Henderson, AAFC, Canada, unpublished data). Interestingly, recent studies suggest that TaLTP3 gene expression is highly correlated with the 5A QTL associated with Type 1 resistance to FHB in wheat cultivars Wuhan and NuyBai, and in a series of derived lines (T. Ouellet and X.-Y. Long, AAFC, Canada, unpublished data). However, no QTLs for resistance have been identified specifically in GS-1-EM0040, which was also screened here for the 5A QTL. Nonetheless, it is possible that TaLTP3 is involved in the mechanism of resistance in wheat genotypes with the 5A QTL, and also in GS-1-EM0040, if the gene(s) of interest associated with the 5A QTL interact with TaLTP3 in mediating resistance.

The largest number of expression differences occurred in GS-1-EM0040, in response to all three elicitors (FgTri5+, FgTri5− or DON) (Fig. 2). The high degree of differential gene expression observed in this genotype included kinases, which are often involved in early signalling events in response to various stimuli (Table 3). Thus, the systemic tissues in GS-1-EM0040 appear to be the most strongly affected by FHB elicitors compared with the other two lines. In contrast to the impressive response observed in GS-1-EM0040, GS-1-EM0168 showed fewer differences in global gene expression (Fig. 2), and by 24 hai none of the three elicitors induced up-regulation of transcripts in this genotype. These data suggest that by 24 hai, GS-1-EM0168 has either already completed the changes involved in setting in motion a resistance response, or that the plant is reallocating resources for a defence response at the site of infection.

Interestingly, while the DON treatment led mainly to down-regulation of transcripts in all wheat lines at the 3 and 24 hai time points, in cv. Superb, the strong down-regulation pattern in response to DON was also observed at 8 hai. In contrast to this, a total of 122 genes were up-regulated at 8 hai in GS-1-EM0040 in response to DON and none were down-regulated, while in GS-1-EM0168, 40 genes were up-regulated and six down-regulated. The pattern of down-regulation suggests that DON may be delaying the plant’s defence response pathway in the susceptible genotype, but is less effective in doing so in the resistant genotypes.

These results demonstrate that when wheat florets respond to challenge with F. graminearum or with DON, the challenge is perceived in distal tissues and induces systemic changes in gene expression. The resulting gene expression patterns may be important in determining whether or not effective resistance develops in the challenged plant.

The trichothecenes produced during infection by F. graminearum are known to induce ribotoxic stress and programmed cell death in eukaryotic cells, and have been shown to inhibit ribosomal peptidyl transferase activity (Mclaughlin et al., 1977). Interestingly, in the current study, direct application of DON induced up-regulation of a range of genes encoding ribosomal components in both resistant wheat lines, but not in the susceptible genotype (Table 3). In both GS-1-EM0040 and GS-1-EM0168, this up-regulation was observed at 8 hai. Some ribosomal gene up-regulation was also observed in response to fungal treatments in the resistant lines. In the susceptible cv. Superb genotype, down-regulation of ribosomal genes was observed in response to both DON and FgTri5+. Thus, when challenged with either DON or DON-producing fungi, the resistant wheat genotypes displayed increased expression of genes encoding components of the ribosomal machinery, whereas expression of these genes declined in susceptible wheat. In a parallel study where the same treatments in the same three wheat lines were compared for differences in protein accumulation, several proteins accumulated to higher levels in either resistant line compared with cv. Superb following challenge with DON (N.A. Foroud, F. Eudes and B.E. Ellis, unpublished data). This result could potentially be related to the stronger pattern of DON-induced down-regulation of gene expression in cv. Superb. The pattern of DON-induced down-regulated molecular events suggests that trichothecenes may delay the plant’s defence response pathway in susceptible genotypes, but is less effective in doing so in resistant lines.

Despite the fact that ribosomes are known targets of DON-toxicity, this is the first report, to the authors’ knowledge, on differential expression of ribosomal proteins in the plant response to trichothecenes; although, Jia et al. (2009) observed F. graminearum induced down- and up-regulation of 41 and four ribosomal genes, respectively, in two near-isogenic lines with or without the 3BS QTL for Type 2 resistance. In this case, the elicited response was observed in both lines. In the current study, the elicited response was observed only in the more resistant lines, where a pattern of up-regulation of ribosomal genes was DON-induced. It may be relevant that both of the resistant lines evaluated here were originally recovered from an in vitro selection screen using a trichothecene toxin (Eudes et al., 2008). GS-1-EM0040 and GS-1-EM0168 have therefore already demonstrated both resistance to FHB and tolerance for the trichothecene virulence factor, which may result from over-production of ribosomes or DON-sensitive ribosome components.

Microbial elicited changes in ribosomal gene expression in other plant–pathogen interactions have also been reported (Hall et al., 2007), and it has been proposed that modification of ribosomal composition can have a direct impact on regulating the translation of specific gene products (Mauro & Edelman, 2002). Thus, in addition to the potential alleviation of trichothecene-induced ribotoxicity, trichothecene-induced changes in ribosomal gene expression may represent a plausible mechanism for regulating translation of specific defence responses.

Toxic stress in plants during pathogenic invasion arises not only from the direct impact of toxins produced by the pathogen, but also from the plant’s production of reactive oxygen species (ROS) as part of its suite of defence responses. In order to protect itself against ROS-induced cellular damage, plants also produce antioxidants, including the PR proteins glutathione-S-transferase (GSTs) and peroxidases (POXs). The elicitor-induced up-regulation of PR gene expression observed in GS-1-EM0168 consisted mainly of antioxidant gene expression, and typically occurred earlier in this genotype than in the other two wheat lines evaluated (Table 3).

Most of the PR protein expression consisted of genes expressing antifungal proteins, and were observed primarily in cv. Superb and GS-1-EM0040 while occurring earlier in the latter. At 8 hai, challenge with the fungus (FgTri5+ and FgTri5−) induced up-regulation of several antifungal genes in GS-1-EM0040 and delayed fungal induced up-regulation of fewer genes was observed in cv. Superb (24 hai). DON also induced up-regulation of antifungal genes, but this was observed mainly in cv. Superb (24 hai) (Table 3). Thus, up-regulation of PR genes in GS-1-EM0040 not only occurred earlier, but was more pronounced in response to the fungus than to DON; whereas in cv. Superb, up-regulation occurred later and primarily in response to DON. Because DON is presumably a single effector, whereas challenge with a live fungus brings many potential effectors into play, this response difference is not surprising.

Fusarium-induced PR protein gene expression has previously been reported in cereals (Pritsch et al., 2000, 2001; Boddu et al., 2006, 2007; Bernardo et al., 2007; Golkari et al., 2007, 2009; Geddes et al., 2008; Jia et al., 2009; Steiner et al., 2009). In some reports, early accumulation and/or higher accumulation of PR proteins differentiated resistant genotypes from susceptible ones (Pritsch et al., 2000; Geddes et al., 2008; Golkari et al., 2009). Additionally, PR gene expression has also previously been observed in response to FHB in the systemic tissues of resistant and susceptible wheat (Pritsch et al., 2001). In the current report, early expression of PR genes was observed in spikelets distal to the inoculation point in both resistant genotypes compared with cv. Superb, consistent with the model that an earlier plant defence response will often distinguish a susceptible from a resistant response, where the earlier response is observed in the resistant varieties. Thus the observations corroborate previous reports that early PR protein expression may contribute to FHB resistance. Accumulation of these gene products in the systemic tissue could presumably prevent secondary infection or reduce disease spread into these tissues.

Regulation of PR protein expression is often mediated by host signalling metabolites, including salicylic acid (SA), jasmonic acid (JA), and ethylene (ET). In a recent report, exogenous applications of methyl-jasmonate increased susceptibility to F. graminearum infection in Arabidopsis during the early stages of infection, but improved resistance could be observed if the treatment was performed 12 hai (Makandar et al., 2010). The importance of JA signalling during later stages of infection is suggestive of a role in resistance to disease spread. Furthermore, gene expression studies have implicated both JA and ET signalling pathways in FHB-resistance in Type 2 resistant cv. Sumai 3 (Li & Yen, 2008). In the current study, while no significant differential expression patterns were observed for those genes believed to be involved in SA and ET signalling, up-regulation of JA biosynthesis and signalling genes was observed in treatment/genotype combinations where a strong resistance to disease spread was anticipated, i.e. in response to FgTri5+ challenge in GS-1-EM0168 (cv. Sumai 3-derived resistance with the 3BS QTL), and in response to FgTri5− challenge in cv. Superb and GS-1-EM0040 (where point inoculation with the trichothecene non-producing strain phenocopies classical Type 2 FHB-resistance where no disease spread is observed). Interestingly, treatment with pure DON, an important aggressiveness factor in FHB disease spread within the host, resulted in down-regulation of JA signalling genes in both cv. Superb and GS-1-EM0040. Perhaps the JA defence response pathway can be interdicted by DON-producing fungi in a susceptible interaction. It is well established that trichothecene production is an important factor for overcoming the host barrier at the rachis node (Jansen et al., 2005; Ilgen et al., 2009), and that trichothecenes are required for disease spread (Proctor et al., 1995; Bai et al., 2002; Langevin et al., 2004; Maier et al., 2006). JA signalling can be a positive regulator of genes involved in lignin biosynthesis (Xue et al., 2008) and signalling through this pathway could potentially help modify cell wall lignification at the rachis node of an infected spikelet.

With the exception of a single gene in GS-1-EM0168, up-regulation of genes in the phenylpropanoid pathway was observed only in GS-1-EM0040. Phenylpropanoid compounds are secondary metabolites with diverse roles in the plant defence response. The first step in this pathway is the conversion of phenylalanine to cinnamate via the activity of PAL (Dixon et al., 2002). Fusarium graminearum-induced PAL up-regulation has previously been reported in Type 2 resistant cv. Sumai 3 (Golkari et al., 2009) and derived line CM82036 (Steiner et al., 2009) in association with the 3BS QTL. In the current study, where uninoculated spikelets were investigated, the CM82039-derived line with the 3BS QTL, GS-1-EM0168, did not show elicitor-induced up-regulation of PAL. Up-regulation of PAL transcripts was observed only in GS-1-EM0040 with no known QTLs for resistance, at 8 hai in response to all three elicitors (Table 3). In susceptible cv. Superb, DON induced down-regulation of PAL at 3 hai. Cinnamate can be converted to p-coumarate, which is the substrate for flavonoid and lignin biosynthesis (Dixon et al., 2002). Up-regulation of genes from both pathways was observed in GS-1-EM0040 at 8 hai.

Cell wall (CW) lignification or thickening of the lemma and of the rachis node has been implicated in FHB-resistance (Kang & Buchenauer, 2000a; Jansen et al., 2005). POX activity is involved in the process of CW lignification (Schopfer, 1996). In GS-1-EM0040, up-regulation of POX, which was limited to the fungal treatments (FgTri5+ and FgTri5−), occurred at 8 hai (Table 3). Up-regulation of genes involved in lignin biosynthesis was prominent in GS-1-EM0040, and also occurred at 8 hai. Aquaporins, in addition to their ability to transport water across cellular membranes, can move small molecules, including H2O2, which is the substrate for POX (Bienert et al., 2007). The higher constitutive expression of aquaporin TaPIP1 in GS-1-EM0040 (Table 2) may facilitate accumulation of H2O2 at the cell wall for lignification. If this lignification were to occur in the glumes or lemma of uninfected spikelets, prevention of Fusarium penetration through the spike surface may be observed, thus increasing Type 1 resistance. On the other hand, cell wall lignification at the rachis node, a known barrier for disease spread in FHB (Kang & Buchenauer, 2000a; Jansen et al., 2005; Ilgen et al., 2009), of an already infected spikelet would improve Type 2 resistance. Furthermore, as up-regulation of PAL has been associated with cv. Sumai 3-derived Type 2 resistance (Golkari et al., 2009; Steiner et al., 2009), it was surprising that up-regulation of the phenylpropanoid pathway was limited in cv. Sumai 3-derived GS-1-EM0168, which has such a strong level of Type 2 resistance. On the other hand, because the analysis was performed in the uninoculated tissues, it remains likely that up-regulation of lignin biosynthesis genes would occur at or near the site of infection in this Type 2 resistant genotype. As the global gene expression patterns revealed so few systemic changes in elicited GS-1-EM0168 spikes, and because disease containment is the mechanism of resistance to FHB spread, it is more likely that Type 2 resistance is a direct result of changes at the site of infection, possibly mediated by JA signalling, rather than a systemic response. On the other hand, in GS-1-EM0040, where many changes in plant defence gene expression were elicited in the uninoculated spikelets, Type 1 resistance in this genotype may involve structural features that minimize fungal penetration in combination with the activation of a systemic response providing further protection against secondary infection.

Acknowledgements

We are grateful to the Alberta Agriculture Research Initiative for funding this research. Special thanks to colleagues at Agriculture and Agri-Food Canada: Ana Badea for conducting QTL analysis on the double haploid lines; Denise Nilsson and Bernie Genswein for technical support and feedback in data processing; Jiro Hattori for his contribution to probe set annotation; Dave Pearson and greenhouse staff for their support and hard work in maintaining our plants. We would also like to thank the field crew at Lethbridge and our colleagues at the nurseries for phenotypic evaluation of the double haploid lines.

Ancillary