Differential expression of proteins in response to the interaction ...

2 downloads 6248 Views 292KB Size Report
proteins by LC-MS/MS technology can provide insight into the molecular .... The Institute for Geno- mics Research (TIGR) protein databases using the MASCOT.
Proteomics 2008, 8, 545–554

545

DOI 10.1002/pmic.200700115

RESEARCH ARTICLE

Differential expression of proteins in response to the interaction between the pathogen Fusarium graminearum and its host, Hordeum vulgare Jennifer Geddes1, Franc¸ois Eudes1, André Laroche1 and L. Brent Selinger2 1 2

Lethbridge Research Centre, Agriculture and Agri-Food Canada, Lethbridge, AB, Canada Department of Biological Sciences, University of Lethbridge, Lethbridge, AB, Canada

Using proteomic techniques, a study aimed at isolating and identifying proteins associated with resistance to fusarium head blight (FHB) was conducted on six barley genotypes of varying resistance. At anthesis, barley spikelets were point inoculated with Fusarium graminearum macroconidial suspensions or mock inoculum. In total, 43 acidic protein spots out of 600 were detected 3 days postinoculation to be differentially expressed due to FHB and were identified. Identification of proteins responsive to FHB included those associated with oxidative burst and oxidative stress response, such as malate dehydrogenase and peroxidases, and pathogenesis-related (PR). An increase in abundance of PR-3 or PR-5 could be associated with the resistant genotypes CI4196, Svansota, and Harbin, as well as the intermediate resistant genotype CDC Bold. On the contrary, the susceptible genotype Stander showed a decrease in abundance of these acidic PR-proteins. In the susceptible and intermediate resistant genotypes Stander and CDC Bold, as well as CI4196, the increased abundance of proteins associated with an oxidative response might have prepared the terrain for saprophytic fungal invasion. On the contrary, in the resistant sources Harbin and Svansota we did not observed change in abundance of these proteins. Not a single significant change in acidic protein abundance could be detected in Chevron. Three distinct response patterns are reported from these six barley genotypes.

Received: February 2, 2007 Revised: October 1, 2007 Accepted: November 2, 2007

Keywords: 2-Dimensional gel electrophoresis / Barley / Data mining / Incompatible interaction / Oxidative stress

1

Introduction

Fusarium head blight (FHB) or scab, caused mainly by Fusarium graminearum Schwabe (teleomorph = Gibberella zeae (Schwein) Petch), is a severe disease of barley and wheat Correspondence: Dr. Franc¸ois Eudes, Lethbridge Research Centre, Agriculture and Agri-Food Canada, 5403 1st Ave. South, P. O. Box 3000, Lethbridge, AB, T1J 4B1, Canada E-mail: [email protected] Fax: 11-403-382-3156 Abbreviations: CMC, carboxymethylcellulose; dpi, days postinoculation; FHB, fusarium head blight; JA, jasmonic acid; PR, pathogenesis-related; RUBISCO, ribulose-1,5-bisphosphate carboxylase/oxygenase; SA, salicylic acid; SAR, systemic acquired resistance; TLP, thaumatin-like protein

© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

grown in humid and semihumid climates [1, 2]. Disease symptoms develop in the spike tissue and are marked by premature necrosis and a brown/gray discoloration. FHB causes significantly lower grain yield, lower test weight, reduced grain quality, and reduced milling yield [3]. The fungus also produces trichothecene mycotoxins such as deoxynivalenol (DON) that are detrimental to both humans and livestock [4]. These mycotoxins have been implicated in pathogenesis, phytotoxicity, and the induction of apoptosis in eukaryotic cell cultures [5–7]. Shriveled grains contaminated with mycotoxins are commonly observed in susceptible cultivars infected by Fusarium spp. [8]. Infection results in a significant loss in value for both the producer and the barley milling industry [1]. Partial control of FHB in barley is through a combination of management practices and partially resistant varieties. www.proteomics-journal.com

546

J. Geddes et al.

Introduction of FHB-resistant barley cultivars would contribute to improved food safety and reduce losses suffered by barley growers and milling industries. In barley, resistance to FHB appears to be less variable than in wheat due to fewer visual symptoms following infection; however, resistance in wheat has been studied more extensively and has led to the development of highly resistant cultivars such as Sumai 3. As a result, the development of FHB-resistant cultivars is a high-priority breeding objective for many barley breeding programs worldwide. However, breeding for FHB-resistance has proven to be a challenge due to the limited understanding of the biochemical and molecular mechanisms involved in plant resistance against infection and spread of F. graminearum. Plants delay pathogen growth or resist pathogen attack by mobilizing a variety of biochemical and molecular defenses [9]. An incompatible interaction between the host and the pathogen results in the triggering of defense responses through signaling pathways; these include the production of ROS (e.g. superoxide radical (O2 ), hydrogen peroxide (H2O2), hydroxyl radical (OH)), nitric oxide, salicylic acid (SA), jasmonic acid (JA), and ethylene [10]. Signaling pathways activate a broad series of defense responses that control or eliminate the pathogen. These responses include hypersensitive response, deposition of cell wall reinforcing materials, and the synthesis of a wide-range of antimicrobial compounds including pathogenesis-related (PR) proteins and phytoalexins [11]. Host response to F. graminearum infection has been studied mainly in wheat [12–14]. Molecular characterization of cDNA clones and ESTs from Fusarium spp.-infected wheat spikes revealed an increase in transcript levels of many PR-genes [12–14]. Different classes of PR-proteins including PR-1, PR-2 (b-1,3-glucanases), PR-3 and PR-4 (chitinases), PR-5 (thaumatin-like protein (TLP)), and PR-9 (peroxidases) were induced within 6–12 h following infection [12–14]. In monocots, exogenous applications of JA to rice have resulted in the accumulation of transcripts for PR-1, -2, -3, -5, and -9 which are associated with hypersensitive cell death [15]. Moreover, SA treatments in monocots have resulted in the up-regulation of PR-2, -3, and -5 [10, 16, 17]. These findings suggest the occurrence of crosstalk between the JA and SA pathways during plant response to pathogen invasion and that defense-related proteins in monocots are activated after fungal infection and may play a role in the general defense against Fusarium spp. infection. Similar to findings in wheat, a recent transcriptome study on the interaction between barley and F. graminearum has reported the induction of transcripts encoding defense-related proteins, oxidative burst-associated enzymes, and phenylpropanoid pathway enzymes [18]. A recent metabolome profiling study led to the identification of groups of compounds that were able to discriminate resistance, and suggested the plausible functions of metabolites in wheat plant defense against F. graminearum [19]. © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Proteomics 2008, 8, 545–554

Proteomic techniques provide tools for studying plant stress responses and possible mechanisms of plant resistance. Using a 2-DE-based protein separation method, a global protein expression profile can be generated and compared. One of the major advantages of the 2-DE technique is that differentially expressed proteins can clearly and reproducibly be detected when infected and uninfected plants are compared. Proteins showing differential expression between treatments may have an important role in the response of the plant to stress. Identification of these differentially expressed proteins by LC-MS/MS technology can provide insight into the molecular mechanisms of resistance and underlying functions of these proteins in determining resistance in barley plants. In this study, a systematic comparison of acidic protein profiles among barley spikelets from six genotypes inoculated with F. graminearum or a mock control was made at 1 and 3 days post-inoculation (dpi). The objective of this study was to identify differentially expressed proteins in FHB-resistant and FHB-susceptible barley genotypes under infected and uninfected conditions, as well as describe the possible mechanisms of resistance in a range of barley genotypes.

2

Materials and methods

2.1 Plant growth Six barley genotypes showing a wide range of phenotypic responses to point inoculation with F. gramineraum [20] were used in this study; three barley genotypes representing FHBresistant sources (Harbin, CI4196, and Svansota), two cultivars of intermediate-resistance to FHB (Chevron and CDC Bold), and one very susceptible cultivar (Stander). Seeds were planted in 15-cm pots and placed in a greenhouse at 21/187C with a 16 h photoperiod until anthesis [21]. Plants were watered daily and treated once with Tilt™ (2.5 mL/L propiconazole, Syngenta Crop Protection Canada, Guelph, ON) during the tillering stage and Intercept™ (0.004 g/L of soil, imidacloprid, Bayer Crop Science Canada, Toronto, ON) once sufficient root development was established to prevent powdery mildew and aphids, respectively. 2.2 Preparation of macroconidia inoculum A single isolate of F. graminearum strain N2, from an infected wheat head, (J. Gilbert, Winnipeg, MB, Canada) was cultured on potato dextrose agar (PDA) for 5 days at room temperature. A F. graminearum macroconidial suspension was produced by transferring four PDA plugs (1 cm61 cm) of the established fungal culture to 500 mL of carboxymethylcellulose (CMC) broth (CMC 15 g, NH4NO3 1 g, KH2PO4 1 g, MgSO4 ?7H2O 0.5 g, yeast 1 g, and H2O 1 L). The culture was incubated on a rotary shaker (150 rpm) at 227C for 2 wk [21]. A hemacytometer was used to count macroconidia. The F. graminearum culture was diluted with www.proteomics-journal.com

Plant Proteomics

Proteomics 2008, 8, 545–554

547

water to produce a suspension of 40 000 F. graminearum macroconidia per milliliter. A mock inoculum was prepared by diluting sterile CMC broth to the same extent as the F. graminearum culture.

natant was stored at 2807C until protein electrophoresis. The Bradford method [25] was used to quantify protein concentration. Three biological replicates were completed for each genotype and treatment.

2.3 Barley spike inoculations

2.5 IEF and SDS-PAGE

To check the developmental stage of the barley plant, the leaf sheath was pulled back from the spike, without damaging the spike and leaf. The macroconidial suspension of F. graminearum was applied to the spikelet using point inoculation at the anthesis stage. Spikelets were inoculated by carefully spreading the palea and lemma and injecting 10 mL of 40 000 F. graminearum macroconidia/mL suspension inside the spikelets using a micropipette [21, 22]. Diluted CMC broth was similarly applied to spikelets on a separate plant to serve as a control. Every second group of spikelets per head was inoculated in the fall 2005. Following inoculation, pots were placed inside a mist-irrigated greenhouse (95% relative humidity) for 72 h at 25/217C with a 16 h photoperiod. Infected heads were harvested at 1 and 3 dpi. Harvested spikes were immediately placed in liquid nitrogen and transferred to a 2807C freezer for storage until protein extraction. Inoculations of two to three spikes per plant were repeated on three different plants per genotypes.

A solubilized protein sample (50 mg for analytical and preparative gels) was mixed with lysis buffer to a total volume of 300 mL and loaded on a 17 cm pH 4–7 BioRad Ready Gel Strip (163-2008, BioRad Laboratories) with the in-gel rehydration method according to the manufacturer’s instructions. For the second dimension separation, the strips were equilibrated for 10 min on a rotary shaker (60 rpm) with 2% DTT and 2.5% iodacetamide (163-2109, BioRad Laboratories). The strips were positioned on top of a 12.5% polyacrylamide gel in the presence of SDS and sealed with 1% agarose. The gels were run for 30 min at 30 mA followed by 5 h at 60 mA using a BioRad Protean II Cell.

2.4 Protein extraction and quantification Protein samples were extracted using the acetone and TCA method described by Wang et al. [23] as reported by Zhou et al. [24] with some modifications. Barley spikelets that received either F. graminearum or mock treatment were ground in liquid nitrogen in a prechilled mortar. Finely ground powder was collected in a 50 mL Falcon tube and weighed. Five milliliters of 10% w/v TCA (T0699, Sigma– Aldrich) and 0.07% v/v 2-mercaptoethanol (M-3148, Sigma– Aldrich) was made up in cold (2207C) acetone and was added to 0.5 g of ground tissue. The samples were incubated for 1 h at 2207C to precipitate proteins and then centrifuged for 20 min at 12 0006g. The pellet of precipitated proteins and debris was washed several times with 5 mL cold 90% acetone containing 0.07% v/v 2-mercaptoethanol until the pellet was colorless. A 20 min centrifugation at 12 0006g was used to pellet the proteins after each wash. Pellets were air dried for 20 min, and the proteins were resuspended in 1 mL of lysis buffer for 20 min. Lysis buffer contained 8 M urea (161-0731, BioRad Laboratories, Mississauga, ON, Canada), 2% CHAPS (BP571-5, BioRad Laboratories), 50 mM DTT (161-0611, BioRad Laboratories), and 0.2% Biolyte carrier ampholytes pH 3–10 (ZM0021, Invitrogen Canada, Burlington, ON, Canada). After centrifugation at 12 0006g for 20 min to remove debris, the supernatant was collected and immediately cleaned using the BioRad ReadyPrep 2-D cleanup kit (163-2130, BioRad Laboratories) according to the manufacturer’s instructions. A 5 mL sample was removed for a protein assay and the remaining super© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

2.6 Staining of SDS-PAGE gels Protein spots were stained with Sypro Ruby (S-12000, Invitrogen Canada) and quantified according to the stain manufacturer’s instructions. Images from all Sypro Ruby stained gels were captured using a Typhoon 9400 scanner (GE Healthcare, Baie D’Urfe, QC, Canada) with the following scanning settings: scan resolution, 300 dots/cm; photomultiplier (PMT), 600 V; normal sensitivity; filters, 610 BP30/Green (532 nm). Protein spots with significantly altered expression following F. graminearum infection were manually excised for LC-MS/MS analyses. 2.7 Image analysis Computer software Phoretix 2D Expression (v2005, from Nonlinear Dynamics, Durham, NC 27703, USA) was used to analyze images of Sypro Ruby-stained gels. Three images representing three biological replicates for 1 and 3 dpi with F. graminearum or mock inoculum, and for each of the six barley genotypes were grouped to calculate the averaged volume of all the individual protein spots. To reduce experimental variation arising during processing of 2-DE, a normalized volume for each individual protein spot was calculated using 100 times the volume of the protein divided by the total volume of all proteins detected on the same image. Warping, matching, and volume comparisons of proteins among the treatments were generated by the software and confirmed manually. Both 1 and 3 dpi samples were compared for each of the six genotypes X treatment separately; at each time point averaged gels of the mock treatment were subtracted from averaged gels of the F. graminearum treatment, and only gels from 3 dpi were further analyzed. A minimum threshold of two-fold change in average expression volumes over triplicates images was set to identify more abundant or less abundant proteins. For each of the 116 www.proteomics-journal.com

548

J. Geddes et al.

protein spots identified using the Phoretix 2D Expression program a t-test was performed and a p-value (p,0.05) is reported in Table 1 and Supporting Information Table S1. We defined de novo proteins as those proteins only present in the F. graminearum treatments, although this kind of change theoretically describes a maximal increase in abundance. Proteins showing altered expression were compared among the resistant and susceptible barley genotypes. Several protein homologs were identified, indicating limited variance of protein positioning across the 2-DE gels. 2.8 LC-MS/MS Excised Sypro Ruby-stained protein spots were sent to the University of British Columbia Laboratory of Molecular Biophysics (URL: http://www.lmb.ubc.ca/analytical.html) or to the National Research Council of Canada (NRC, Ottawa, ON, Canada) for LC-MS/MS analysis. According to the NRC protocol, the proteins were destained and reduced with DTT, alkylated with acrylamide, and digested with trypsin (Promega, Madison, WI, USA). The resulting peptide solution was analyzed on a Micromass capillary LC (CapLC) and QTof API US (Manchester, UK) LC-MS system. A peptide capillary trapping (CapTrap; Michrom Bioresources, Auburn, CA, USA) was used for online desalting, followed by back flushing onto a 0.0756100 mm peptide map (PepMap) C18 column (LC Packings, Amsterdam, Netherlands). Peptides were eluted from the column with a 30 min linear gradient of 3–45% solvent B (solvent B: 97.9% ACN, 2% H2O, 0.1% formic acid) at a flow rate of ,300 nL/min. The standard micromass nanospray source with blunt-tip 90 mm od and 20 mm id fused silica emitter was held at 807C, capillary voltage 13.4 kV, cone voltage 32 V. Data acquisition was performed in data dependent mode, with up to three precursors for MS/MS selected from each MS survey scan. The.pkl files generated by Micromass ProteinLynx software were searched against the National Center for Biotechnology Information Non-Redundant (NCBI NR) and the The Institute for Genomics Research (TIGR) protein databases using the MASCOT MS/MS Ion Search (www.matrixscience.com). Search parameters allowed for one missed cleavage and set the peptide tolerance to 61.2 Da and MS/MS tolerance to 60.6 Da with a carbamidomethyl fixed modification and an oxidation variable modification. A match was considered successful when the protein identification score was located out of the random region of MASCOT.

3

Results

3.1 Protein identification Approximately 600 protein spots were resolved in the pH 4–7 range on all of the 2-DE gels. Figure 1A shows the total protein expression profile from FHB-resistant barley genotype, Harbin, following F. graminearum point inoculation at 3 dpi. © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Proteomics 2008, 8, 545–554

Protein profile changes observed following F. graminearum inoculation were more significant and only reported for 3 dpi. No results are presented from F. graminearum versus mock inoculation at 1 dpi. Comparison of 2-DE images from three biological replicates indicated there were both qualitative and quantitative differences between protein profiles from F. graminearum and mock-inoculated spikelets in resistant barley genotypes CI4196 (R), Harbin (R), Svansota (R), in intermediate resistant CDC Bold (I), and susceptible Stander (S). No protein difference was significant in intermediate resistant Checron (I). In total, 43 protein spots were hand-selected based on significant average differences (t-test) between mock-inoculated samples and F. graminearuminoculated samples, differences between F. graminearuminoculated samples at 1 and 3 dpi. These protein spots were observed in at least two of the three gels and identified with more than two peptides. Table 1 shows the 42 protein spots, representing 33 different proteins grouped according to an oxidative burst and defense response, PR proteins, and proteins associated with metabolism and regulation. 3.2 Proteins associated with an oxidative burst and defense response Seventeen protein spots were associated with an oxidative burst response, leading to the identification of 12 different proteins in four barley genotypes (Table 1). Putative malate dehydrogenase (gi)34911788) was observed in all four genotypes. It had lower abundance in resistant Svansota (22.2fold), and in susceptible Stander (22.3-fold), and higher abundance in both intermediate resistant genotype CDC Bold (3.8-fold) and in resistant genotype CI4196 (13.4-fold). Cytoplasmic malate dehydrogenase (gi)18202485) showed 2.5-fold reduced abundance in susceptible genotype Stander, while intermediate-resistant genotype CDC Bold showed 19.7-fold increased abundance. The susceptible genotype Stander showed a 14.8-fold increase in abundance of cytosolic malate dehydrogenase (gi)37928995), and 26.4-fold lower abundance of glutathione transferase F5 (gi)23504745). Intermediate-resistant genotype CDC Bold showed a 14.2fold increase in abundance of peroxiredoxin Q (Q5S1S6). Four different peroxidases were identified in three barley genotypes: peroxidase gi)22587 was 3.1-fold more abundant in CDC Bold. Ascorbate peroxidase gi)3688398 was more abundant in CDC Bold (2.9-fold) and CI4196 (6.7-fold). Another ascorbate peroxidase gi)15808779 was 3.0-fold more abundant in Stander. CDC Bold showed 8.7-fold increased abundance of a stromal ascorbate peroxidase (gi)32879781). Glutathione-dependent dehydroascorbate reductase 1 (gi)6939839), enzymes associated with peroxidases, was observed only in CDC Bold (4.7-fold increase). A jasmonateinduced protein (gi)400094) showed higher abundance in CI4196 (2.4-fold), and a universal stress-like protein (gi)53791695) was identified in CDC Bold with 2.7-fold higher abundance. Resistant CI4196, intermediate resistant CDC Bold, and susceptible Stander presented a higher www.proteomics-journal.com

Plant Proteomics

Proteomics 2008, 8, 545–554

549

Table 1. Identification of differentially expressed proteins showing a minimum two-fold change in abundance in response to F. graminearum infection in six barley lines

Accession number Protein identification

Oxidative burst and defense response gi)3688398 Ascorbate peroxidase gi)15808779 Ascorbate peroxidase gi)32879781 Stromal ascorbate peroxidase gi)22587 Peroxidase gi)230504745 Glutathione transferase F5 gi)18202485 Cytoplasmic malate dehydrogenase gi)34911788 Putative malate dehydrogenase gi)37928995 Cytosolic malate dehydrogenase gi)400094 Jasmonate-induced protein 1 gi)53791695 Universal stress protein-like gi)6939839 Glutathione-dependent dehydroascorbate reductase 1 Q5S1S6 Peroxiredoxin Q PR gi)563489 Chitinase gi)56090131 TLP4 gi)75103125 TLP3 Q94649 TLP7 Metabolism and regulation gi)28416583 Ubiquitin-like protein gi)46805452 Chloroplast inorganic pyrophosphatase gi)29124123 Actin-depolymerizing factor 3 gi)847873 Mg-chelatase subunit gi)167096 RUBISCO activase isoform 1 gi)1167948 RUBISCO activase isoform 1 gi)100796 Phosphoribulokinase gi)482311 Photosystem II oxygen-evolving complex 1

CI4196 Fold changea) p-valueb)

6.7 0.04 – – – – – – – – – – 13.4 0.04 – – 2.4 0.03 – – 4.7 0.02 – –

Harbin Fold change p-value

– – – – – – – – – – – – – – – – – – – – – –

Svansota Fold change p-value

– – – – – – – – – – – – 22.2 0.00 – – – – – – – –

Chevron Fold change p-value

CDC Bold Fold change p-value

– – – – – – – – – – – – – – – – – – – – – –

2.9 0.04 – – 8.7 0.00 3.1 0.00 – – 19.7 0.00 3.7 0.03 – – – – 2.7 0.00 – –

Stander Fold change p-value

– – 3.0 0.00 – – – – 26.4 0.01 22.5 0.01 22.3 0.03 14.8 0.00 – – – – – –

– –

– –

– –

14.2 0.03

– –

33.3 0.00 – – – – – –

– – – – ?

– – – – – – 2.5 0.01

– – – – – – – –

– – 7.3 0.01 – – – –

– – 2.3 0.00 – – 216.2 0.00

– – – – 2.9 0.03 – – 7.6 0.01 – – – – – –

– – 22.3 0.05 – – – – – – – – 22.7 0.04 – –

– – – – – – – – – – – – – – – –

– – – – – – – – – – – – – – – –

2.8 0.00 – – – – 3.3 0.05 2.2 0.00 – – – – – –

– – – – – – – – – – 8.7 0.00 – – 2.0 0.00

© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

– –

www.proteomics-journal.com

550

J. Geddes et al.

Proteomics 2008, 8, 545–554

Table 1. Continued

Accession number Protein identification

CI4196 Fold changea) p-valueb)

Harbin Fold change p-value

Svansota Fold change p-value

Chevron Fold change p-value

CDC Bold Fold change p-value

Stander Fold change p-value

gi)131344 Photosystem II oxygen-evolving complex protein 2 precursor gi)20790 Type 1 chlorophyll a/b-binding protein gi)2507469 Triose phosphate isomerase, cytosolic gi)47607439 Mitochondrial ATP synthase gi)55233175 b-cyanoalanine synthase gi)62732953 Chloroplast fructose-bisphosphate aldolase precursor gi)6683813 UMP/CMP kinase b gi)7488889 Type III a membrane protein gi)75133690 Putative reversibly glycosylated polypeptide gi)7528175 Histone H4 fragment

– –

– –

– –

– –

?

– –

– – – – – – – – – –

– – – – – – – – – –

– – – – – – – – – –

– – – – – – – – – –

11.8 0.02 27.4 0.00 – – – – – –

– – – – – –

– – – – – –

3.1 0.00 – – – –

– – – – – –

– – ? – ? –

– – – – – –

– –

? –

– –

– –

– –

– –

– – – – 22.6 0.00 22.6 0.01 22.2 0.01

a) Fold change represents the change in abundance for proteins identified at 3 dpi between F. graminearum and mock inoculation. A minimum two-fold change was expressed by those proteins identified with LC-MS/MS. The values reported represent the fold changes of proteins following analysis with the Phoretix 2D Expression program. b) A t-test was performed to determine the p-value.

abundance of peroxidase precursors, peroxidases, and malate dehydrogenases. On the contrary, we did not observe significant change in protein profiles from intermediate resistant Chevron and resistant Harbin, while resistant Svansota showed a reduced abundance of one malate dehydrogenase (Table 1).

3.3 PR proteins Only four identified proteins were associated with pathogenesis (Table 1): three PR-5, TLPs and one PR-3, chitinase. TLP3 (gi)75103125) was de novo expressed in resistant genotype Harbin, and TLP4 (gi)5609013) was more abundant in susceptible genotype Stander (2.3-fold) and in intermediate-resistant genotype CDC Bold (7.3-fold). Figure 1B shows an increase in abundance of TLP4 from intermediate-resistant genotype CDC Bold at 3 dpi with F. graminearum. TLP7 (Q94649) showed decreased abundance in susceptible genotype Stander (216.2-fold) and increased abundance in resistant genotype Svansota (2.5-fold). Chitinase 2b (gi)563489), was 33.3-fold more abundant in resistant genotype CI4196. © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

3.4 Proteins associated with plant metabolism, regulation, and unknown functions Eighteen different protein spots were associated with metabolism, regulation and structure functions (Table 1). Change in abundance of these proteins was not observed in more than one genotype. In Stander, two proteins were found more abundant, a ribulose-1,5-bisphosphate carboxylase/oxygenase (RUBISCO) activase isoform 1 (gi)1167948) and a protein involved in the photosystem II oxygen-evolving complex 1 (gi)131344) (8.7 and 2.0-fold, respectively). Three proteins were less abundant in Stander, a mitochondrial ATP synthase (gi)47607439), a b-cyanoalanine synthase (gi)55233175), and a chloroplast fructose-biphosphate aldolase precursor (gi)62732953) (22.6, 22.6, and 22.2-fold, respectively). In CDC Bold, four proteins were more abundant, an Ubiquitin like protein (gi)28416583), a Mg-chelatase subunit (gi)847873), a Rubisco activase isoform 1 (gi)167096), and a type 1 chlorophyll a/b-binding protein (gi)20790) (2.8, 3.3, 2.2, and 11.2-fold, respectively). A cytosolic triose phosphate isomerase (gi)2507469) was less abundant in CDC Bold (27.4-fold). As well, a type IIIa membrane protein (gi)7488889) and a putative reversibly glycosylated polywww.proteomics-journal.com

Proteomics 2008, 8, 545–554

Plant Proteomics

551

or 3 dpi, suggesting that JA and SA pathways were activated within the first 24 h. Limitations of the methodology used could also explain the absence of such observations: e.g., use of a narrow pH range of 4–7; threshold limit of 2-DE gels. However, the acidic proteins identified provided insights into the diverse barley responses to F. graminearum aggression. 4.1 Proteins associated with an oxidative burst and oxidative stress response

Figure 1. (A) Total protein expression profile of barley spikelets extracted from FHB-resistant line, Harbin, 3 dpi with F. graminearum. This is a representative figure from three biological replicates. (B) Enlarged inlets and expression histograms of a TLP4 from the intermediate-resistant barley line CDC Bold showing an increase in abundance following F. graminearum infection (expression profile: 1 is mock protein 3 dpi and 2 is FHB protein 3 dpi).

peptide (gi)75133690) were de novo expressed in CDC Bold. In Svensota, a UMP/CMP kinase b (gi)6683813) was more abundant (3.1-fold). In Harbin, a chloroplast inorganic pyrophosphatase (gi)46805452) and a phosphoribulokinase (gi)100796) were less abundant (22.3 and 22.7-fold, respectively). As well, a Histone H4 fragment was de novo expressed in Harbin (gi)7528175). In CI4196, only one protein was more abundant (2.9-fold), an actin-depolymerizing factor 3 (gi)29124123). No change in abundance of acidic proteins associated with metabolism, regulation and structure functions were observed in Chevron.

4

Discussion

A total of 19 different acidic proteins associated with mechanisms of resistance to FHB were identified in comparative 2-DE gel-analysis of six barley genotypes. We did not observe enzymes involved in the synthesis of precursors of the JA, SA, or phenylalanine-ammonia lyase (PAL) pathways, nor the altered expression of b-1,3-glucanases (PR-2) at either 1 © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

The high proportion of identified proteins associated with an oxidative burst and/or oxidative stress response in the six barley genotypes investigated, indicated that such a plant defense response was prevalent against F. graminearum. The plasma membrane of plant cells produces ROS, H2O2 and O2 , in response to both biotic and abiotic stimuli that play an important role in plant–pathogen interactions [26, 27]: (i) it hinders penetration of the pathogen by stimulating peroxidase activity and by crosslinking cell walls at the site of contact; (ii) it poses a stress on the pathogen as well as the host cell generating the oxidative burst; and (iii) it acts as a diffusible signal that leads to systemic acquired resistance (SAR) [28]. High intracellular levels of H2O2 cause the activation of plant cell death and defense mechanisms during pathogen invasion [29]. Malate dehydrogenases and peroxiredoxin Q contribute in eliminating toxic radicals produced during oxidative stress [30, 31]. Ascorbate peroxidase (PR-9) associated with the JA signaling pathway was part of the main ROS-removing system for cellular protection against oxidative stress [32]. Overall, we could identify two main oxidative response patterns (Table 1). In Svansota, Harbin, and Chevron, we observed no change in the abundance of all reported acidic proteins associated with an oxidative burst and oxidative stress responses; i.e., ascorbate peroxidases and other peroxidases, and malate dehydrogenases (with the exception of one putative malate dehydrogenase in Svansota). Diverse explanations could be developed in the absence of change in these proteins abundance, such as an increased turnover, a response with basic proteins, an allocation of energy for the development of other active mechanisms of defense: e.g., PRproteins. However, our experimental design cannot support any of these explanations. In contrast, CI4196 and CDC Bold showed dramatic increases in abundance (2.9 to 19.7-fold increase of specific proteins) of the oxidative stress defense response proteins: e.g., peroxidase precursors and peroxidases, and malate dehydrogenases. Stander showed similar response with a large reallocation of malate dehydrogenase to the cytosol, and increase in oxidative stress response. Such a massive oxidative burst and plant defense against free oxygen radicals may have direct oxidative action on F. graminearum, but could also lead to a hypersensitive reaction (not observed in this study). Hypersensitive reaction is a common mechanism used by plants to contain and confine pathogens [33]; however, as a saprophyte, F. graminearum infection would progress even in the presence of dead plant tissue. www.proteomics-journal.com

552

J. Geddes et al.

Genetic differences among the six barley genotypes may provide a possible explanation for the observed disparity among FHB-resistance [34]. 4.2 PR proteins Although no protein precursors of the JA, SA, and PAL pathways were identified, a classic model of cereal defense response reported [35] and the analysis of acidic PR-proteins observed in this study could provide valuable information regarding the barley response to F. graminearum attack. The SA pathway activates selective PR-proteins, and induces H2O2 accumulation as a signal for SAR [32, 36, 37]. PR-3 (chitinases) and PR-5 (TLPs) are induced in cereals following treatment with either JA or SA, and PR-1 and PR-9 are only activated by the JA pathways [10, 15–17]. Svansota, Harbin, and Chevron showed no significant change in abundance of the acidic PR-9 proteins, while CDC Bold, CI4196, and Stander showed increases in abundance for these PR-9 proteins, indicating the possible use of the JA antioxidant signaling pathway in pathogen defense. Activation of the JA pathway was not an absolute requirement for the induction of an incompatible interaction and FHBresistance. We could not eliminate the possibility that a complete analysis of acidic and basic PR-proteins may reveal activation of basic PR-1 and PR-9 proteins in the resistant genotypes. Resistant genotype CI4196 showed a significant increase in abundance of chitinase 2b (gi)563489; abundance, 0.26% of total protein). The corresponding gene cht2b was discovered in the barley cultivar Pallas, inoculated with powdery mildew (Genbank accession no. X78672). Chitinases are hydrolytic enzymes that inhibit the growth of many fungi in vitro by hydrolyzing the chitin of fungal cell walls. The oligomeric products of digested chitin can also act as signal molecules to stimulate further defense responses [14, 38]. An expression profile for ESTs from highly homologous barley chitinase II indicated high expression in leaf, spike, and stem, moderate expression in the sheath, and no expression in the root seed and flower (http://www.ncbi.nlm.nih.gov/ UniGene/clust.cgi?ORG=Hv&CID=173). The combined induction of a strong oxidative burst, response to oxidative stress, and abundance of induced chitinase in CI4196 may explain its high level of stable resistance. Three different TLPs were reported in four barley genotypes. A de novo produced TLP3 (gi)75103125) was observed in resistant genotype Harbin, Interestingly, a TLP7 (Q94649) was 2.5-fold more abundant in the resistant genotype Svansota, while considerably reduced (216.2-fold) in the susceptible genotype Stander. An increase in abundance of a TLP4 (gi)56090131) was observed in the intermediate-resistant genotype CDC Bold and in the susceptible genotype Stander. Inhibition of fungal growth by PR-5 has been reported, but varies depending on TLP isoform and the fungal species. TLP antifungal activity differences among isoforms might depend on the binding capacity to various fungal (1,3)-b-D© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Proteomics 2008, 8, 545–554

glucans in vivo, and the interaction with other PR-proteins: e.g., PR-2 and PR-3 [39]. Chevron did not show activation of any acidic PR-proteins associated with either the SA signaling antioxidant pathway or the JA signaling pathway for defense or the use of an alternative mode of defense against F. graminearum invasion. 4.3 Proteins associated with plant metabolism and regulation Proteins associated with metabolism, regulation, and protein structure also presented altered expression patterns following F. graminearum infection. A modification in expression patterns of these proteins was most likely a byproduct of invasion while the fungus attempted to acquire resources from the plant, for growth and survival, in a compatible interaction. Fructose bisphosphate aldolase identified in Stander is involved in basic cellular metabolism; b-cyanoalanine synthase, identified in Stander, is involved in the synthesis of cysteine, suggesting that the alteration of the amino acid synthesis and nitrogen metabolism was a result of F. graminearum infection. A change in abundance of RUBISCO, photosystem II oxygen-evolving complex, Mgchelatase, ATP synthase, and chlorophyll was reported as a possible result of the reduced photosynthetic potential of the plant following oxidative stress [40]. A type III membrane protein which has role in transportation, also showed altered expression patterns due to fungal infection and the resulting plant oxidative stress. 4.4 Comparison of defense response among barley genotypes Barley genotype responses to F. graminearum inoculation were diverse and significantly different; a summary of these responses is presented in Table 1. FHB resistant barley genotype CI4196 had the strongest increase in abundance of PR3 and PR-9 proteins, and strong active oxidative burst and oxidative stress response. Such a response is typical of an incompatible interaction between a pathogen and the host plant and indicative of activation of strong and diversified defense responses. Early-defense responses and the continued and prolonged production of ROS during an oxidative burst, may have contributed to activation of chitinase 2b in the spikelets. The FHB susceptible genotype Stander followed the classic model of a compatible interaction, response to oxidative stress and lack of or delays in PR-protein induction, and in this particular case a decrease in abundance of PR-5. Reported acidic protein profile changes suggested different defense mechanisms in the other four barley genotypes. The most surprising observations were made for Svansota, Harbin and Chevron, which had little or no acidic defense protein change in abundance. Only acidic PR-5 proteins had increased abundance or were de novo expressed in Svansota and Harbin, respectively. www.proteomics-journal.com

Plant Proteomics

Proteomics 2008, 8, 545–554

Intermediate resistant CDC Bold presented similarities with resistant genotype CI4196, but showed an increase in abundance of TLP4 (PR-5) instead of chitinase (PR-3), or TLP7 as in Svansota, and did not show an increase in abundance of an acidic jasmonate-induced protein 1. Our findings suggest that TLP7 might be more efficient than TLP4 against F. graminearum mycelium. A study of basic proteins or of the timing of defense response would highlight more phenotypic differences associated with resistance between these two genotypes. Chevron had no significant changes in defense-related acidic protein abundance. A study of constitutively expressed defense proteins might shade light on FHB resistance in barley, and in particular in Chevron’s response to F. graminearum.

5

Conclusion

To the best of our knowledge, this is the first report of the application of proteomic techniques in studying the interaction between a series of barley genotypes representing various levels of resistance to FHB and the F. graminearum pathogen. The proteomic investigation of resistant and susceptible genetically unrelated barley genotypes to FHB revealed different induced response patterns to fungal invasion. The observed and deduced responses encompassed an oxidative burst, JA and SA antioxidant signaling pathways, induction of PR-proteins, and alteration of protein synthesis, photosynthesis, regulation, and other metabolic pathways. Our results indicated that the induced plant defense responses following fungal infection were diverse among resistant, intermediate and susceptible barley genotypes. We were able to detect several components of SAR in the susceptible genotype Stander; such as the production of antioxidant proteins, and a decrease in abundance of PRproteins. Resistant genotypes CI4196, Harbin, and Svansota differed in oxidative stress response, but showed a common induction response of acidic PR-3 and PR-5 proteins. An increase in abundance of oxidative responses and cell death in susceptible and intermediate genotypes, induced by the trichothecene producing fungus, might prepare the terrain for invasion of saprophytic F. graminearum. Transcriptome analysis and proteome studies of acidic and basic proteins, in response to the fungus, the trichothecenes, and their interactive effect in FHB pathogenesis would help to complete the picture of resistance put forth from these 2-DE proteomic studies.

We thank Dr. Jeannie Gilbert of AAFC-Winnipeg, MB, Canada, for the F. graminearum fungal strain N2, Suzanne Perry at the UBC Laboratory of Molecular Biophysics, and Luc Tessier at NRC Ottawa for the protein identification service and technical assistance. We also thank Dr. Wenchun Zhou for his technical advice and assistance. The funding for this project was provided by Alberta Agricultural Research Institute and MII with WGRF. The authors have declared no conflict of interest. © 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

6

553

References

[1] McMullen, M., Jones, R., Gallenburg, D., Scab of wheat and barley: A re-emerging disease of devastating impact. Plant Dis. 1997, 81, 1340–1348. [2] Parry, D. W., Jenkinson, P., McLeod, L., Fusarium ear blight (scab) in small grain cereals - a review. Plant Pathol. 1995, 44, 207–238. [3] Stack, R., Return of an old problem: Fusarium head blight of small grains, APSNet monthly Feature May 1999, DOI: 10.1094/PHP-2000-0622-01-RV. [4] Bai, G. H., Shaner, G., Management and resistance in wheat and barley to Fusarium head blight. Annu. Rev. Phytopathol. 2004, 42, 135–161. [5] Kang, Z., Buchenauer, H., Immunocytochemical localization of Fusarium toxins in infected wheat spikes by Fusarium culmorum. Physiol. Mol. Plant Pathol. 1999, 55, 275–288. [6] Shifrin, V. I., Anderson, P., Trichothecene mycotoxins trigger a ribotoxic stress response that activates c-Jun N-terminal kinase and p38 mitogen-activated protein kinase and induces apoptosis. J. Biol. Chem. 1999, 274, 13985–13992. [7] Desjardins, A. E., Hohn, T. M., Mycotoxins in plant pathogenesis. Mol. Plant Microbe Interact. 1997, 10, 147–152. [8] Bai, G.-H., Shaner, G., Wheat scab: Perspective and control. Plant Dis. 1994, 78, 760–766. [9] Bowles, D. J., Defense-related proteins in higher plants. Annu. Rev. Biochem. 1990, 59, 873–907. [10] Lu, Z.-X., Gaudet, D., Puchalski, B., Despins, T. et al., Inducers of resistance reduce common bunt infection in wheat seedlings while differentially regulating defense-gene expression. Physiol. Mol. Plant Pathol. 2006, 67, 138–148. [11] Veronese, P., Ruiz, M. T., Coca, M. A., Hernandez-Lopez, A. et al., In defense against pathogens. Both plant sentinels and foot soldiers need to know the enemy. Plant Physiol. 2003, 131, 1580–1590. [12] Li, W. L., Faris, J. D., Muthukrishnan, S., Liu, D. J. et al., Isolation and characterization of novel cDNA clones of acidic chitinases and b-1,3-glucanases from wheat spikes infected by Fusarium graminearum. Theor. Appl. Genet. 2001, 102, 353–362. [13] Pritsch, C., Muehlbauer, G. J., Bushnell, W. R., Somers, D. A., Fungal development and induction of defense response genes during early infection of wheat spikes by Fusarium graminearum. Mol. Plant Microbe Interact. 2000, 2, 159–169. [14] Pritsch, C., Vance, C. P., Bushnell, W. R., Somers, D. A. et al., Systemic expression of defense response genes in wheat spikes as a response to Fusarium graminearum infection. Physiol. Mol. Plant Pathol. 2001, 58, 1–12. [15] Schweizer, P., Buchala, A., Silverman, P., Seskar, M. et al., Plant Physiol. Jasmonate-inducible genes are activated in rice by pathogen attack without a concomitant increase in endogenous jasmonate acid levels. Plant Physiol.1997, 114, 79–88. [16] Molina, A., Garcia-Olmedo, F., Developmental and pathogen-induced expression of three barley genes encoding lipid transfer proteins. Plant J. 1993, 4, 983–991. [17] Wasternack, C., Hause, B., Jasmonates and octadecanoids: Signals in plant stress responses and development. Prog. Nucleic Acids Res. Mol. Biol. 2002, 72, 165–221.

www.proteomics-journal.com

554

J. Geddes et al.

[18] Boddu, J., Cho, S., Kruger, W. M., Muehlbauer, G. J., Transcriptome analysis of the barley-Fusarium graminearum interaction. Mol. Plant Microbe Interact. 2006, 19, 407–417. [19] Hamzehzarghani, H., Kushalappa, A. C., Dion, Y., Rioux, S. et al., Metabolic profiling and factor analysis to discriminate qualitative resistance in wheat cultivars against Fusarium head blight. Physiol. Mol. Plant Pathol. 2005, 66, 119–133. [20] Geddes, J., Eudes, E., Tucker, J. R., Legge, W. G., Selinger, L. B., FHB phenotypic evaluation and deoxynivalenol quantification of Fusarium graminearum infected Hordeum vulgare using three artificial inoculation methods. Can. J. Plant Pathol. 2008, in press. [21] McCallum, B. B., Tekauz, A., Influence of inoculation method and growth stage on Fusarium head blight in barley. Can. J. Plant Pathol. 2002, 24, 77–80. [22] Evans, C. K., Xie, W., Dill-Macky, R., Mirocha, C. J., Biosynthesis of deoxynivalenol in spikelets of barley inoculated with macroconidia of Fusarium graminearum. Plant Dis. 2000, 84, 654–660. [23] Wang, W., Scali, M., Vignani, R., Spadafora, A. et al., Protein extraction for two-dimensional electrophoresis from olive leaf, a plant tissue containing high levels of interfering compounds. Electrophoresis 2003, 24, 2369–2375. [24] Zhou, W. C., Eudes, F., Laroche, A., Identification of differentially regulated proteins in response to a compatible interaction between the pathogen Fusarium graminearum and its host, Triticum aestivum. Proteomics 2006, 6, 4599– 4609. [25] Bradford, M. M., A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye loading. Anal. Biochem. 1976, 72, 248– 254. [26] Foyer, C. H., Mullineaux, P. M. (Eds.), Causes of Photooxidative Stress and Amelioration of Defense Systems in Plants, CRC Press, Florida 1994, pp. 177–198. [27] Levine, A., Tenhaken, R., Dixon, R., Lamb, C., H2O2 from the oxidative burst orchestrates the plant hypersensitive disease resistance response. Cell 1994, 79, 583–593. [28] Noctor, G., Foyer, C. H., Ascorbate and glutathione: Keeping active oxygen under control. Annu. Rev. Plant Physiol. Plant Mol. Biol. 1998, 49, 249–279. [29] Takahashi, H., Chen, Z., Du, H., Liu, Y., et al., Development of necrosis and activation of disease resistance in transgenic

© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Proteomics 2008, 8, 545–554 tobacco plants with severely reduced catalase levels. Plant J. 1997, 11, 993–1005. [30] Foyer, C. H., Mullineaux, P. M. (Eds.), Causes of Photooxidative Stress and Amelioration of Defense Systems in Plants, CRC Press, Florida 1994, pp. 77–104. [31] Ostergaard, O., Finnie, C., Laugesen, S., Roepstorff, P., Proteome analysis of barley seeds: Identification of major proteins from two-dimensional gels (pI 4-7). Proteomics 2004, 4, 2437–2447. [32] Moller, I. M., Plant mitochondria and oxidative stress: Electron transport, NADPH turnover, and metabolism of reactive oxygen species. Ann. Rev. Plant Physiol. Plant Mol. Biol. 2001, 52, 561–591. [33] Dixon, R. A., Harrison, M. J., Lamb, C. J., Early events in the acitivation of plant defense responses. Annu. Rev. Phytopathol. 1994, 32, 479–501. [34] Takeda, K., Current topics on the scab disease resistance in barley and wheat. Proc. Jpn. Assoc. Mycotoxicol. 1992, 36, 13. [35] Li, W. L., Faris, J. D., Chittoor, J. M., Leach, J. E., et al., Genomic mapping of defense response genes in wheat. Theor. Appl. Genet. 1999, 98, 226–233. [36] Spoel, S. H., Koornneef, A., Claessen, S. M. C., Korzelius, J. P., et al., NPR1 modulates cross-talk between salicylate- and jasmonate-dependent defense pathways through a novel function in the cytosol. Plant Cell 2003, 15, 760–770. [37] Lamb, C., Dixon, R. A., The oxidative burst in plant disease resistance. Annu. Rev. Plant Physiol. Plant Mol. Biol. 1997, 48, 251–275. [38] Mauch, F., Mauch-Mani, B., Boller, T., Antifungal hydrolases in pea tissue: II. Inhibition of fungal growth by combinations of chitinase and b-1,3-glucanase. Plant Physiol. 1988, 88, 936–942. [39] Osmond, R. I., Hrmova, M., Fontaine, F., Imberty, A., et al., Binding interactions between barley thaumatin-like proteins and (1,3)-b-D-glucans: kinetics, specificity, structural analysis, and biological implications. Eur. J. Biochem. 2001, 268, 4190–4199. [40] Palatnik, J. F., Valle, E. M., Carrillo, N., Oxidative stress causes Ferredoxin-NADP1 reductase solubilization from thylakoid membranes in methyl viologen-treated plants. Plant Physiol. 1997, 115, 1721–1727.

www.proteomics-journal.com