Proteomics Uncovers a Role for Redox in Drought Tolerance in Wheat §

Proteomics Uncovers a Role for Redox in Drought Tolerance in

Wheat§

Mohsen Hajheidari,†,# Alireza Eivazi,†,# Bob B. Buchanan,‡ Joshua H. Wong,‡ Islam Majidi,† andGhasem Hosseini Salekdeh*,†

Department of Physiology and Proteomics, Agricultural Biotechnology Research, Institute of Iran, Karaj, Iran,and Department of Plant and Microbial Biology, University of California, 111 Koshland Hall,

Berkeley, California 94720

Received October 30, 2006

Proteomic analysis offers a new approach to identify a broad spectrum of genes that are expressed inliving systems. We applied a proteomic approach to study changes in wheat grain in response todrought, a major environmental parameter adversely affecting development and crop yield. Three wheatgenotypes differing in genetic background were cultivated in field under well-watered and droughtconditions by following a randomized complete block design with four replications. The overall effectof drought was highly significant as determined by grain yield and total dry matter. About 650 spotswere reproducibly detected and analyzed on 2-DE gels. Of these, 121 proteins showed significant changeunder drought condition in at least one of the genotypes. Mass spectrometry analysis using MALDI-TOF/TOF led to the identification of 57 proteins. Two-thirds of identified proteins were thioredoxin(Trx) targets, in accordance with the link between drought and oxidative stress. Further, because ofcontrasting changes in the tolerant and susceptible genotypes studied, several proteins emerge as keyparticipants in the drought response. In addition to providing new information on the response to waterdeprivation, the present study offers opportunities to pursue the breeding of wheat with enhanceddrought tolerance using identified candidate genetic markers. The 2-DE database of wheat seed proteinsis available for public access at http://www.proteome.ir.

Keywords: proteomics • redox proteins • drought • wheat • seed • oxidative stress • thioredoxin h • drought tolerance

Introduction

The advent of proteomics has made it possible to identify abroad spectrum of proteins in living systems. This capabilityis especially useful for cereals, as it may give clues not onlyabout nutritional value, but also about yield and how thesefactors are affected by adverse conditions. Among the cereals,wheat has received attention because of its economic andnutritional importance. Accordingly, proteins of whole grain1,2

and its different compartments have been analyzed, that is,amyloplasts, the organelles that synthesize and store copiousamounts of starch3,4 and the parent starchy endosperm thatis the source of flour.5 The results have given insight intothe broad metabolic capability of these systems and, in thecase of endosperm, how metabolism changes during develop-ment.

As a result of its relation to grain quality and yield, elevatedtemperature has also been investigated with the wheat pro-teome.1,2 These studies have shown that a number of grainproteins, 37 in one case1 and 38 in the other,2 are changed andthus could potentially serve as markers of temperature stress.Although studied in leaves of Arabidopsis6, sugar beet,7 rice,8,9

and oak,10 Elymus elongatum11, the effect of another majorenvironmental parameter adversely affecting development andcrop yieldsnotably, droughtshas, however, not been studiedin grain. To fill this gap, we have analyzed the grain proteomesof three wheat genotypes under normal and drought condi-tions. The results, summarized below, show that many proteinsare differentially expressed in response to water stress in thesegenetically distinct lines. Standing out are redox-linked pro-teins, including h-type thioredoxins (Trxs), that, as they seemto correlate with drought, could serve as genetic markers forthis type of stress.

Experimental ProceduresPlant Materials. Three spring wheat genotypes (Arvand,

Khazar-1, and Kelk Afghani) differing in origin were grown inexperimental fields of West Azarbayjan Agricultural ResearchCenter (36° 58′ N, 46° 6′ E) in the 2002-2003 growing season.

* Corresponding author. Dr. Ghasem Hosseini Salekdeh, AgriculturalBiotechnology Research Institute of Iran, P.O. Box 31535-1897, Karaj, Iran.E-mail, [email protected]; fax, +98-261-2704539.

§ Dedicated to the memory of Karoly K. Kobrehel who was a true visionaryin linking redox to grain proteins.

† Agricultural Biotechnology Research Institute of Iran.‡ University of California.# These two authors contributed equally to this paper.

10.1021/pr060570j CCC: $37.00 2007 American Chemical Society Journal of Proteome Research 2007, 6, 1451-1460 1451Published on Web 03/08/2007

Plots of 2.4 m2 were sowed at a density of 250 seeds per plot.The experimental design was a randomized complete blockdesign with four replications. Plots were uniformly irrigatedfrom emergence until the booting stage12 using a system basedon evaporation from Class A Pan. Control and water deficit-treatments were irrigated at 75 ( 5 and 150 ( 5 mm evapora-tion, respectively. At the end of the growing season, calculationsof total dry weight and grain yield were performed based onfour central rows (1 m2/plot). The mature seeds of the mainears from each plot were collected for proteome analysis. Thewhole grain, 20 g, was ground in a UDY mill (Model MS UDYCyclone Sample Mill, UDY CO., Fort Collins, CO). The samplesobtained were analyzed for protein by near-infrared reflectancespectroscopy13 using an NIR System 8100 (NIR System, Inc.,Silver Springs, MD). Data were analyzed using of Mstat-csoftware based on Steel and Torrie14 procedure.

Protein Extraction and Two-Dimensional Gel Electro-phoresis (2-DE). Protein was extracted from the ground mealaccording to Finnie et al.15 with minor modifications. The seedsof each of the main ears collected from different plots (replica-tions) were ground in liquid nitrogen with a mortal and pestle.Approximately, 4 g of ground grain was suspended in 20 mLof extraction buffer consisting of 5 mM Tris, pH 7.5, 1 mMCaCl2, and 20 mM DTT. The soluble proteins were extractedby shaking 30 min at 4 °C and then clarified by centrifugation(30 min at 15 000g). The supernatant fraction was applied toanalytical 2-D gels. Protein was quantified according to Brad-ford16 using reagents from Bio-Rad (Hercules, CA) and bovineserum albumin as the standard. For preparing gliadins, 10 g ofmill flour was mixed with about 8 mL of distilled water,centrifuged for 5 min at 10 000g, and washed with water. Theresidue (gliadins and glutenins) were separated by centrifuga-tion at 10 000g, and the gliadin fraction was dried at roomtemperature and weighed.

The precipitate was solubilized in buffer containing 7 M urea,2 M thiourea, 4% (w/v) CHAPS, 1% (w/v) DTT, 1% (v/v) pH3-10 ampholyte, and 35 mM Tris base. For IEF, 18 cm IPGstrips with a linear gradient (pH 4-7) were rehydrated using350 µL of rehydration buffer (8 M urea, 2% (w/v) CHAPS, 20mM DTT, 0.5% (v/v) IPG buffer 4-7, and 0.01% (w/v) bromo-phenol blue) in a reswelling tray (Amersham Pharmacia Bio-tech, Uppsala, Sweden) at room temperature for 16 h. Foranalytical and preparative gels, 120 µg and 1.5 mg of protein,respectively, were added to rehydration buffer. IEF was per-formed using Multiphor II and a DryStrip Kit (AmershamPharmacia Biotech). Gels were run at 500 V for 2 h followedby 1000 V for 2 h and finally 3500 V for 14 h. The focused IPGstrips were equilibrated twice for 15 min in 10 mL of equilibra-tion solution. The first equilibration was performed in asolution containing 50 mM Tris-HCl buffer, pH 8.8, 6 M urea,30%(v/v) glycerol, 2% (w/v) SDS, and 1% (w/v) DTT, with afew grains of bromophenol blue. The second equilibration wasas the first except that DTT was replaced with 2.5% (w/v)iodoacetamide. The second dimension was developed with a12% SDS-polyacrylamide gel using a Protean II Multi Cell (Bio-Rad). Protein spots were visualized in analytical gels by stainingwith silver nitrate.17 Preparative gels were stained with colloidalCommassie Brilliant Blue G-250 (CBB).18-19

Image and Data Analysis. Wet silver-stained gels werescanned with a GS-800 calibrated densitometer (Bio-Rad,Hercules, CA) at a resolution of 600 dots and 12-bit per inchand analyzed using Melanie-3 software (GeneBio, Geneva,Switzerland) according to the user’s manual. After scanning,

spot detection, protein quantification, and spot pairing werecarried out based on Melanie-3 default settings. Then, spotpairs were investigated visually, and the scatter plots of eachdata point between gels were displayed to estimate gel similar-ity and experimental errors. The molecular mass of proteinsin the gels was estimated by co-electrophoresis with standardprotein markers (Amersham Pharmacia Biotech); pI was de-termined by measuring spot migration on 18 cm IPG (pH 4-7,linear) strips. Since there were two treatments for eachgenotype (well-watered and droughted), treatment combina-tions were analyzed by two-way analysis of variance (ANOVA).Spots were concluded to be significantly up- or down-regulatedwhen P < 0.05.

Protein Identification and Database Search. Protein spotswere excised from CBB-stained gels and analyzed using anApplied Biosystems 4700 Proteomics Analyzer at the Proteinand Proteomics Center, University of Singapore (Mass Spec-trometry Services, Department of Biological Sciences). Proteindigestion, desalting, and concentration of samples were carriedout using Montage In-Gel Digestion Kits (Millipore and AppliedBiosystems, Foster City, CA). The samples were dissolved insolvent consisting of 0.1% trifluroacetate and 50% acetonitrile(ACN) in MilliQ Water. Then 0.5 µL of sample solution wasmixed with 0.5 µL of matrix solution (5 mg/mL R-cyano-4-hydroxycinnamic acid dissolved in the above solvent), appliedto a MALDI sample target plate, and dried in air. Before eachanalysis, the instrument was calibrated with the AppliedBiosystems 4700 Proteomics Analyzer Calibration Mixture. Datainterpretation was carried out using the GPS Explorer Software(Applied Biosystems), and an automated database search wascarried out using the MASCOT program (Matrix Science, Ltd.,London, U.K.). Combined MS-MS/MS searches were con-ducted with the selection of the following criteria: MSDB20040710 (1501893 sequences; 480537664 residues), all entries,parent ion mass tolerance at 50 ppm, MS/MS mass toleranceof 0.2 Da, carbamidomethylation of cysteine (fixed modifica-tion) and methionine oxidation (variable modification). Ac-cording to MASCOT probability analysis (P < 0.05), onlysignificant hits were accepted.

Western Blot Analysis. Flour protein, 100 mg, was extractedwith 1.0 mL of 0.5 N NaCl by shaking for 1 h at ambienttemperature. The supernatant fraction was recovered by cen-trifugation (10 min at 15 800g) at 4 °C. Western blot analysiswas performed on clarified salt-soluble protein extracts pre-pared from the indicated cultivars and subjected to SDS-PAGEon a pre-cast 10-20% linear gradient Criterion Tris-HCl gel(Bio-Rad, Hercules, CA) at pH 8.5 at a constant voltage of 150V for 75 min. Several concentrations of protein (5-20 µg) weredissolved in 1× Laemmli sample buffer, boiled for 5 min,clarified, and loaded onto the above gel. Proteins were trans-ferred to nitrocellulose at a constant voltage of 50 V for 1 h at4 °C using a Criterion Blotter Cell Assembly (Bio-Rad, Hercules,CA). Nitrocellulose membrane was blocked with 5% powderedmilk in buffer (20 mM Tris-HCl, pH 7.5, supplemented with0.15 M NaCl) for 2 × 30 min at 25 °C and incubated in primaryantibody overnight at 4 °C and in secondary antibody for 1 hat 25 °C. Primary antibody was wheat anti-Trx h II20 diluted1:1000; secondary antibody was goat anti-rabbit IgG-HRPConjugate (Bio-Rad, Hercules, CA) diluted 1:3000. Blots weredeveloped in TMB Substrate Kit for Peroxidase (Vector Labo-ratories, Burlingame, CA) according to the manufacturer’sinstructions.

research articles Hajheidari et al.

1452 Journal of Proteome Research • Vol. 6, No. 4, 2007

Protein Determination, Blot Scanning, and Analysis. Theblots were scanned using a UMAX PowerLook II00 scannerfitted with Photoshop 6. Images were transferred to QuantityOne Quantitation Software (Bio-Rad, Hercules, CA), version 4.0,and the volume (intensity × area) of the immunoreactive bandswas measured.

Results and Discussion

On the basis of the analysis of variance, the overall effect ofdrought was highly significant (P < 0.01) as determined by grainyield (Figure 1, left panel) and total dry matter (Figure 1, rightpanel). Ears per m2 and grains per spike were also affected withrespective drought-induced decreases of 35 and 33%. In allcases, the tolerant genotype (Khazar-1) was significantly lessaffected than its susceptible counterparts (Afghani and Arvand).

There were also protein differences. One particularly strikingchange was observed with the gliadin storage proteins whichshowed an increase of 9-fold in drought-tolerant Khazar-1versus a 3-fold increase in susceptible Arvand and Afghani.Interestingly, while total protein increased in all cases duringdrought stress, the extent was relatively modest: Khazar-1 (3%)< Afghani (7%) < Arvand (9%). These changes may reflect theeffect of stress as drought tends to increase protein contentduring grain filling because the accumulation of starch is moreseverely affected than that of nitrogen.21 Overall, it appears that,while effecting only a small increase in total grain protein,drought elicits a major redirection of resources to the synthesisof gliadins. Further, this redirection appears to be morepronounced in drought-tolerant genotypes.

Identification of Drought-Responsive Proteins. We applied2-DE to analyze the seed proteomes of the three wheatgenotypes in response to prolonged drought stress (SupportingInformation Figure 1). Drought-responsive proteins were ex-cised from preparative gels and examined by MALDI-TOF/TOF.A spot that consistently showed a particular position in thedifferent gels was considered to be the same protein.

About 650 spots were reproducibly detected in four replica-tions of each genotype, that is, Khazar-1 (Figure 2), Arvand,and Afghani. Of these, 121 proteins showed significant changeunder drought condition in at least one of the genotypes (Figure2 and Supporting Information Table 1). Mass spectrometryanalysis using MALDI-TOF/TOF led to the identification of 57proteins (Supporting Information Table 2) that were classifiedaccording to function (Table 1). Some of the proteins in Table1, for example, Trx h, serpin, R-amylase inhibitor, and mito-chondrial aldehyde dehydrogenase, have functions other thanthe one indicated.

Both the total number of drought-responsive proteins andthe number of up-regulated proteins (P < 0.05) were higher inthe tolerant genotype, Khazar-1, than in the susceptiblegenotypes, Arvand and Afghani (Figure 3). Additionally, theresponse patterns of the proteins generally differed among thethree genotypes. Most of the proteins up-regulated duringdrought in Khazar-1 were either up-regulated to a lesser extent,down-regulated, or showed no change in Arvand and Afghani.Conversely, a small number of proteins up-regulated in thesusceptible genotypes either showed no change or were down-regulated in the resistant counterpart.

Function of Drought-Responsive Proteins. The proteinsfunction in fundamental processes, including stress/defense,protein synthesis/assembly, metabolism, and storage. Four ofthe proteins were unknown, and two had no known function.Of the 57 proteins of known function, 38 (designated with anasterisk, *) have been identified as either a confirmed orcandidate Trx target, thus, highlighting a link of droughttolerance to redox.22-24 The individual proteins are discussedbelow in relation to their function and link to redox. The changein each protein in response to drought is given in Table 1.Corresponding changes in the different processes are sum-marized in Figure 4. The majority of the proteins identifiedfunctioned in stress/defense, protein synthesis/assembly, andmetabolism. A small number were storage proteins and pro-teins with unknown function.

Proteins Responding in Three Genotypes. Because of theproduction of reactive oxygen species (ROS) in multiple typesof stress, systems have been developed to minimize theirdeleterious effects during seed development and germina-tion.25,26 In particular, desiccation and resumption of respirationfollowing the hydration of dry seeds give rise to ROS that shouldbe removed or neutralized. The appearance of antioxidantenzymes at the onset of the maturation-drying phase is inaccord with this function and acquisition of desiccation toler-ance.27 In view of these events, it is not surprising that thelargest number of drought-related proteins in cereal grain areredox-related and are known to undergo a pronounced changein sulfhydryl status during development and germination.23

Four enzymes functional in stress/defense, 1-Cys peroxiredoxin,glutathione S-transferase, and two forms of Trx h (spots 37 and43), showed a decisive response in the drought-tolerant geno-type (Kharzar-1) and both susceptible counterparts.

1. Thioredoxin h. A member of a family of small proteinsthat appear to be ubiquitous, Trx h functions in scavenging,but perhaps more importantly, the protein regulates a numberof fundamental seed processes.23,28-30 All Trxs share a conserved

Figure 1. The effect of drought stress on grain yield and dry weight of three genotypes.

Redox Role in Drought Tolerance research articles

Journal of Proteome Research • Vol. 6, No. 4, 2007 1453

active-site motif, [-Trp-Cys-Gly(Pro)-Pro-Cys-]. In theirreduced state, Trxs reduce regulatory disulphide bridges ofnumerous target proteins. Subsequently, the oxidized Trxsformed are reduced by either of two mechanisms: one linkedto ferredoxin (chloroplasts and cyanobacteria) and the otherto NADPH (heterotrophic cells and cell compartments). PlantTrxs reduced by NADPH, via NADP-thioredoxin reductases(NTR), are designated the h-type and have been described forthe cytosol, mitochondria, and ER.23

Trx h is also involved in a range of biochemical processes.These include the mobilization of protein and starch ingerminating cereal seeds,31 self-incompatibility,32,33 and cellularprotection against oxidative stress, particularly during seeddesiccation and germination.34 Trxs h reduce a variety of targetproteins that contain disulfide bonds, including storage pro-teins, such as glutenins and gliadins in wheat35,36 and hordeinand glutelin in barley,37 and proteins related to oxidative stresssuch as peroxiredoxin.23

In the current study, we identified three isoforms of Trx hthat showed contrasting response patterns in the tolerant andsusceptible genotypes. One form (spot 43) increased in thetolerant genotype and decreased in the susceptible counter-parts, and another (spot 42) decreased in all three genotypes(both of these changes were statistically significant). A thirdspot, 37, significantly increased in the tolerant genotype anddecreased in the susceptible ones, although these changes werenot statistically significant.

Because of its pronounced change in drought stress, weapplied an immunological approach to confirm the proteomicresults for Trx h. As seen in Figure 5, Western blot analysisshowed that the amount of Trx h differed in the three genotypessubjected to drought, that is, the level decreased in the sensitivegenotypes (Arvand and Afghani) and increased in the tolerantcounterpart (Khazar-1). This differential change was observedwith both the major and minor bands, two well-known iso-forms of Trx h (major and minor bands are designated with adot and a star in Figure 5).38 Further, the proteomic andWestern blot analyses were complementary in showing anincrease in Trx h in Khazar-1 and a decrease in both Arvandand Afghani under drought stress (Figure 6). Thus, the level ofTrx h in whole grain decreased in sensitive and increased intolerant genotypes based on two lines of evidence. In view ofits apparent dual improving effect, it becomes of interest torelate the abundance of Trx not only to drought tolerance, butalso to dough quality.39 It is known that drought leads to adecrease in the ability of flour produced acceptable products.40

The question, therefore, arises as to whether Trx will helpminimize this deleterious effect.

The three wheat Trx h spots37,42 are closest to ArabidopsisTrx h5 (81% similarity). They are, however, also quite close toTrx h3 (79% and 78% similarity). Trx h5 and h3 are character-ized by a relatively common but atypical active site, CPPC. Thethree isoforms of Trxs in our study could, in fact, be the wheatcounterpart of Arabiopsis Trx h3 and/or Trx h5. Both Trxs are

Figure 2. 2-D gel analysis of proteins extracted from grain of Khazar-1 genotype harvested under well-watered conditions. In the firstdimension (IEF), 120 µg of protein was loaded on an 18 cm IPG strip with a linear gradient of pH 4-7. In the second dimension, 12%SDS-PAGE gels were used, with a well for molecular weight standards. Proteins were visualized using silver staining. Arrows representdrought-responsive spots of which 57 have been identified by MS (Table 1).



Table 1. Drought Responsive Proteins of Wheat Grain Identified Using MALDI TOF-TOF MS/MS and Their CorrespondingInduction Factor (Percent Volume of Spot in Stress Condition/Percent Volume of Spot in Well-Watered Condition)a



of interest in this context because h5 has been associated withoxidative stress41 and h3 confers H2O2 tolerance.42 Our resultssupport these previously assigned functions and extend the roleof Trx h to drought stress.

2. Glutathione S-Transferase. A participant in ROS scaveng-ing, glutathione S-transferase showed decisive differentialexpression in the tolerant and susceptible genotypes. Theenzyme limits oxidative damage by removing ROS formed instress and by detoxifying xenobiotics under normal conditions.The endogenous products of oxidative damage, for example,membrane lipid peroxides and products of oxidative DNAdegradation, are highly cytotoxic. Glutathione S-transferasedetoxifies these endogenously generated electrophiles by con-jugating them with GSH.43 In accord with this conclusion, theenzyme increased 2-fold under drought stress compared to thewell-watered counterpart in Khazar-1, whereas it was down-regulated by a similar factor in Arvand and Afghani (Table 1).

3. 1-Cys-peroxiredoxin. 1-Cys-peroxiredoxin is also directlyinvolved in scavenging ROS. In plants, the cDNA of 1-Cys-peroxiredoxin was first isolated as a dormancy-related proteinexpressed in the embryo and aleurone layer of barley.44 Theantioxidant activity of this enzyme and its contribution to

desiccation tolerance during late stages of seed developmenthave been elucidated.45 These authors suggested that 1-Cys-peroxiredoxin may help maintain dormancy. In reaching asimilar conclusion, Haselkas et al.46 suggested that the anti-oxidant function of 1-Cys Prx is to sense harsh environments,thereby preventing germination under unfavorable conditions.In our study, the change in 1-Cys-peroxiredoxin was pro-nounced but was inconsistent in the three genotypes.

Proteins Responding in Two Genotypes. Three types ofprotein in the stress/defense category showed significantopposing changes in Khazar-1 and one of the susceptiblegenotypes (R-amylase inhibitor, a cold-regulated protein anddehydroascorbate reductase).

1. Dehydroascorbate Reductase. This enzyme was up-regulated in response to drought stress in the tolerant genotype,Khazar-1, and down-regulated in the susceptible genotypeseither markedly (Arvand) or marginally (Afghani). The enzymeis noteworthy in view of its role in the ascorbate-glutathionecycle for ROS removal. Dehydroascorbate reductase also showedthe highest increase in activity during seed development amongthe enzymes studied.25 As there is a shortage of ascorbic acidin seeds,47 enhanced dehydroascorbate activity early in germi-nation would help maintain the limited pool of this anti-oxidant.26

2. r-Amylase Inhibitor. The activity and expression level ofR-amylase inhibitor is associated with starch during grain fillingand seed maturation. We identified nine protein spots asisoforms of the R-amylase inhibitor. Of these, four (spots 18,20, 31, and 543) were up-regulated to a greater extent inKhazar-1 than in the sensitive genotypes. Four (spots 32, 152,153, and 190) did not change markedly in Khazar-1, but weredown-regulated to varying degrees in Afghani, and two proteins(spots 32 and 153) were down-regulated in both Afghani andArvand. Only one isoform (spot 36) was up-regulated in Arvand.Although a single form of R-amylase did not change in anopposing manner in the tolerant and susceptible genotypes,the observed up-regulation of individual forms in tolerantKhazar-1 and their down-regulation in the susceptible geno-types may offer protection against oxidative stress and helppreserve grain starch, a component crucial for germination.

Table 1 (Continued)

a According to Mascot probability analysis (P < 0.05), only significant hits were accepted; (*) thioredoxin target (22,23); (**) change statistically significantin at least one variety in response to drought stress compared to well-watered. Spots were concluded to be significantly up- or down-regulated when P < 0.05.Dark gray highlights, change statistically significant in two varieties in response to drought stress compared to well-watered;10 light gray highlights, changestatistically significant in three varieties in response to drought stress compared to well-watered.9 (§) Change differs between tolerant and at least one susceptiblevariety in response to drought stress compared to well-watered.11

Figure 3. The number of grain proteins differing significantly inabundance in drought-stressed and re-watered plants of wheatcultivars compared with well-watered controls. Solid bars, pro-teins more abundant in stressed plants; open bars, proteins lessabundant in stressed plants.



3. Cold-Regulated Protein. We identified two cold-regulatedproteins, both of which were significantly up-regulated inKhazar-1. One (spot 30) was down-regulated in one of thesusceptible genotypes, Arvand. The UniGene cluster of thisgene (Ta.13183) represents ESTs from leaf, root, and developing

and dormant seed tissues. Drought, salt, and cold conditionsof dormant seed and drought condition represent the highestnumber of ESTs. In Arabidopsis, most cold-regulated genesrespond to dehydration, and conversely, most dehydration-induced genes also respond to cold stress.48 The current worksuggests that a common set of genes is responsible for bothdrought and cold tolerance.

4. Proteins Functional in Synthesis and Assembly. Two ofthe proteins identified in this category, elongation factor-1 Rand HSP17, were up-regulated in Khazar-1 and down-regulatedin one of the susceptible genotypes in accord with a protectivefunction in drought response (Table 1). Others have observedthat this protein family, which is expressed during seeddevelopment,49 is up-regulated in mature wheat seed underheat stress.1,2 Additionally, several studies have suggested thatHSPs may function to protect cellular components during seeddesiccation.50-53 In the current study, the types of HSPsidentified were consistently up-regulated to a greater extentin drought-tolerant (Khazar-1) than in the susceptible geno-types (Afghani and Arvand).

It is noteworthy that three HSP 70 identified (spots 170, 171,and 173) were down-regulated in the susceptible genotype,Afghani. These proteins are involved in a wide range of cellularfunctions, including protein folding, and the correct assemblyof oligomeric proteins,54 prevention of the aggregation ofdenatured proteins,55 refolding of stress-denatured proteins,56

and import of proteins across membranes.57 Their down-regulation would thus likely have detrimental effects onproteins at multiple levels. The changes in expression of theseHSP70 proteins in the other genotypes was less clear.

Three forms of protein disulfide isomerase3 precursor wereidentified. One (spot 119) showed behavior not observed withother proteins in this study: down-regulated in each of the treegenotypes, but less so in the two susceptible genotypes thanin Khazar-1. The observation suggests that this form of theenzyme may be detrimental for drought tolerance.

5. Proteins Functional in Metabolism. Of the metabolicenzymes identified, glyceraldehyde-3-phosphate dehydro-genase showed the most striking response to drought stress:up-regulation in drought-susceptible Afghani and Arvand anddown-regulation in tolerant Khazar-1. This finding suggests thatglycolysis would be decreased during drought. The pattern ofthe other metabolic enzymes was less clear. Thus, methylma-lonate semialdehyde dehydrogenase, a mitochondrial enzymefunctional in the catabolism of valine and pyrimidines, wasincreased in one susceptible and down-regulated in the otheras well as in the tolerant counterpart. Starch granule-boundstarch synthase showed similar inconsistent behavior.

6. Storage Proteins. One globulin was up-regulated understress selectively only in drought-sensitive genotypes anddown-regulated in Khazar-1. The results suggest that it isadvantageous to down-regulate the synthesis of this storageprotein and direct available resources to minimize the effectsof stress during drought.

7. Unknown. Two proteins of unknown function (spots 28and 29) were up-regulated in the drought-tolerant genotypeand down-regulated in one or both of the susceptible coun-terparts. Another protein, spot 99, that was down-regulated upto 2-fold in the susceptible genotypes and not significantlychanged in the tolerant counterpart, was identified as ahypothetical protein from Sporobolus stapfianus (resurrectiongrass) with 89% identity to rice glyoxalase I (BAA36759).Glyoxalase I plays a central role in the prevention of glycation

Figure 4. Functional annotation of the identified drought-responsive proteins in three genotypes classified by biologicalfunction described in Table 1. The numbers represent the percentand number of proteins in each class.



reactions by detoxifying R-oxoaldehydes such as methylglyoxaland glyoxal through conjugation with glutathione. Thesederivatives are further metabolized by glyoxalase II, therebyminimizing their mutagenic and cytotoxic activities that, amongother effects, lead to arrested growth.58 A decrease in glyoxalaseI activity in situ during aging or oxidative stress results inincreased glycation and tissue damage.59 Transgenic tobaccounderexpressing glyoxalase I showed an enhanced accumula-tion of methylglyoxal which resulted in the inhibition of seedgermination.60 By contrast, overexpression of glyoxalase Iresulted in improved tolerance against methylglyoxal andhigher levels of resistance to salinity stress compared to

nontransgenic controls.61 Glyoxalase I has also been found tobe one of several genes induced in drought and cold stressesin Arabidopsis.62 Further study is needed to determine howglyoxalase is linked to the drought-tolerance mechanism inplants.

Concluding Remarks

In highlighting the role of redox, the present findings addinsight to our understanding of the response of wheat todrought stress. The results provide evidence that droughtcauses a redirection in protein synthesis to increase theformation of gliadins, interestingly, to a greater extent indrought-tolerant genotypes than in susceptible genotypes. Ofthe 57 drought-responsive proteins identified, two-thirds wereTrx targets, highlighting the link between drought and oxidativestress that was earlier observed with leaves of Arabidopsis.6

Further, because of contrasting changes in the tolerant andsusceptible genotypes studied, several proteins emerge as keyparticipants in the drought response. Included are two proteinsthat are clearly selectively up-regulated in the drought-tolerantgenotype, Khazar-1, notably, Trx h and glutathione S-trans-ferase. Other possible candidates include cold-regulated pro-tein, dehydroascorbate reductase, elongation factor-1 R, HSP17,and unknown protein (P0022BO5.25). There are also twoproteins that behaved in an opposing manner, that is, down-regulated in the tolerant Khazar-1 and up-regulated in thesusceptible Afghani and Arvand: glyceraldehyde-3-phosphatedehydrogenase and at least one globulin storage protein. Oneenzyme, protein disulfide isomerase3 precursor, was consist-ently down-regulated, but more extensively in the tolerant thanin the susceptible genotypes.

In addition to providing new information on the responseto water deprivation, the present study offers opportunities topursue the breeding of wheat with enhanced drought tolerance.Specific areas worthy of further study include:

Figure 5. Western blot analysis of Trx h in grain extracts from drought-susceptible and tolerant wheat genotypes grown in the field.The gel was developed with 5 µg of flour protein. Similar results were obtained with 10 and 20 µg samples. Seeds were taken fromplants grown in one of the four indicated replicate field plots (R1-R4).

Figure 6. Comparison of Western blot and proteomic analysisfor drought-induced changes in Trx h isoforms in grain ofdrought-susceptible and tolerant wheat varieties grown underfield conditions. The amount of protein in the Western blot wasassessed by estimating the volume of each band by densitomet-ric scanning. The values shown represent averages taken fromthe relevant lanes designated in Figure 5. The average ofsignificant induction factors of all Trx h isofrorms in the pro-teomic (2-DE) analysis are presented as indicated.



• Identification of the remaining 63 proteins found to changeduring drought stress in the genotypes in the current investiga-tion. This information may give further insight into drought-responsive pathways and the genes that control them. Effortsshould be made to take advantage of ongoing improvementsin proteomics technology to increase sensitivity in an attemptto identify less abundant proteins such as transcription factors.

• Development of additional wheat genotypes tolerant andsusceptible to drought. Added analyses will provide an assess-ment of the extent to which present findings can be generalizedand whether the redox and related proteins identified in thisstudy undergo selective change in other genotypes. This workis also central to efforts to identify candidate markers fordrought tolerance.

• The effect of drought on the redox state of grain proteinsin susceptible versus tolerant genotypes. A change in redoxstate can markedly alter the activity of certain enzymes, forexample, by oxidative deactivation or glutathionylation, inaddition to alerting the solubility and digestibility propertiesof storage proteins.23

• Analysis of the proteome in mapping grain populations.Proteome maps will help determine whether the clustering ofdrought-responsive proteins has a genetic basis and whetherthey can be applied as markers for breeding.

The two-dimensional gel electrophoresis (2-DE) data-bases of wheat seed proteins contain clickable 2-DE gel imagesand descriptive textual information such as protein name,Mr/pI values, MS score, and sequence coverage. These andother information are available for public access at http://www.proteome.ir.

Abbreviations: CBB, Coomassie Brilliant Blue; BSA, bovineserum albumin; Trx, thioredoxin; ROS, reactive oxygen species.

Acknowledgment. This project was partially funded bygrants from the Agricultural Biotechnology Research Institute,Iran, to G.H.S. and from the California Agricultural ExperimentStation to B.B.B. We are grateful to the Iranian National ScienceFoundation for the financial assistance to establish 2-DEdatabase and to Mohammad Ghareyazie and Taha Abachi fortheir technical assistance in creating the database.

Supporting Information Available: List of droughtresponsive proteins and identified proteins in wheat. Thismaterial is available free of charge via the Internet at http://pubs.acs.org.

References

(1) Majoul, T.; Bancel, E.; Triboı, E.; Ben Hamida, J.; Branlard, G.Proteomics 2003, 3, 175-183.

(2) Skylas, D. J.; Cordwell, S. J.; Hains, P. G.; Larsen, M. R.; Basseal,D. J.; Walsh, B. J.; Blumenthal, C.; Rathmell, W.; Copeland, L.;Wrigley, C. W. J. Cereal Sci. 2002, 35, 175-188.

(3) Andon, N. L.; Hollingworth, S.; Koller, A.; Greenland, A. J.; Yates,J. R.; Haynes, P. A. Proteomics 2002, 2, 1156-1168.

(4) Balmer, Y.; Vensel, W. H.; Dupont, F. M.; Buchanan, B. B.;Hurkman, W. J. J. Exp. Bot. 2006, 57, 1591-1602.

(5) Vensel, W. H.; Tanaka, C. K.; Nai, N.; Wong, J. H.; Buchanan, B.B.; Hurkman, W. J. Proteomics 2005, 5, 1594-1611.

(6) Rey, P.; Pruvot, G.; Becuwe, N.; Eymery, F.; Rumeau, D.; Peltier,G. Plant J. 1998, 13, 97-107.

(7) Hajheidari, M.; Abdollahian-Noghabi, M.; Askari, H.; Heidari, M.;Sadeghian, S. Y.; Ober, E. S.; Salekdeh, G. H. Proteomics 2005, 5,950-60.

(8) Salekdeh, Gh. H.; Siopongco, J.; Wade, L. J.; Ghareyazie, B.;Bennett, J.; Proteomics 2002, 2, 1131-1145.

(9) Ali, G. M.; Komatsu, S. J. Proteome Res. 2006, 5, 396-403.(10) Jorge, I.; Navarro, R. M.; Lenz, C.; Ariza, D.; Jorrin, J. Proteomics

2006, 6 (Suppl. 1) S207-214.

(11) Gazanchian, A.; Hajheidari, M.; Sima, N. K.; Salekdeh, G. H. J.Exp. Bot. 2007, 58, 291-300.

(12) Zadoks, J. C.; Chang, T. T.; and Konzak, C. F. Weed Res. 1974, 14,415-421.

(13) Norris, K. H.; Hruschka, W. R.; Bean, M. M.; Slaughter, D. C. CerealFoods World 1989, 34, 696-705.

(14) Steel, R. G. D.; Torrie, J. H. Principles and Procedures of Statistics,2nd ed.; McGraw Hill: New York, 1980.

(15) Finnie, C.; Melchior, S.; Roepstorff, P.; Svensson, B. Plant Physiol.2002, 129, 1308-1319.

(16) Bradford, M. Anal. Biochem. 1976, 72, 396-403.(17) Blum, H.; Beier, H.; Gross, H. J. Electrophoresis 1987, 8, 93-99.(18) Neuhoff, V.; Arold, N.; Taube, D.; Ehrhardt, W. Electrophoresis

1988, 9, 255-262.(19) Smith, D. M.; Tran, H. M.; Epstein, L. B. In Cytokines: A Practical

Approach, 2nd ed.; Balkwill, F. R., Ed.; IRL Press: Oxford, U.K.,1995; pp 111-128.

(20) Johnson, T. C.; Wada, K.; Buchanan, B. B.; Holmgren, A. PlantPhysiol. 1987, 85, 446-451.

(21) Triboı, E.; Martre, P.; Triboı-Blondel, A. M. J. Exp. Bot. 2003, 54,1731-1742.

(22) Balmer, Y.; Koller, A.; del Val, G.; Manieri, W.; Schurmann, P.;Buchanan, B.B. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 370-375.

(23) Buchanan, B. B.; Balmer, Y. Annu. Rev. Plant Biol. 2005, 56, 187-220.

(24) Wong, J. H.; Cai, N.; Balmer, Y.; Tanaka, C. K., Vensel, W. H.;Hurkman, W. J.; Buchanan, B. B. Phytochemistry 2004, 65, 1629-1640.

(25) De Gara, L.; de Pinto, M. C.; Moliterni, V. M. C.; D’Egidio, M. G.J. Exp. Bot. 2003, 54, 249-258.

(26) Pallanca, J. E.; Smirnoff, N. Plant Physiol. 1999, 120, 453-462.(27) Baily, C.; Audigier, C.; Ladonne, F.; Wagner, M. H.; Coste, F.;

Corbineau, F.; Come, D. J. Exp. Bot. 2001, 52, 701-708.(28) Schurmann, P.; Jacquot, J.-P. Annu. Rev. Plant Physiol. Plant Mol.

Biol. 2000, 51, 371-400.(29) Baumann, U.; Juttner, J. Cell. Mol. Life Sci. 2002, 59, 1042-1057.(30) Yano, H.; Kuroda, S.; Buchanan, B. B. Proteomics 2002, 9, 1090-

1096.(31) Wong, J. H.; Kim, Y. B.; Ren, P. H.; Cai, N.; Cho, M. J.; Hedden,

P.; Lemaux, P. G.; Buchanan, B. B. Proc. Natl. Acad. Sci. U.S.A.2002, 99, 16325-16330.

(32) Bower, M. S.; Matias, D. D.; Fernades-Carvalho, E.; Gu, M.;Rothstein, S. J.; Goring, D. R. Plant Cell 1996, 8, 1641-1650.

(33) Cabrillac, D.; Cock, J. M.; Dumas, C.; and Gaude, T. Nature 2001,410, 220-223.

(34) Serrato, A. J.; Cejudo, F. J. Planta 2003, 217, 392-399.(35) Kobrehel, K.; Wong, J. H.; Balogh, A.; Kiss, F.; Yee, B. C.; Buchanan,

B. B. Plant Physiol. 1992, 99, 919-924.(36) Lozano, R. M.; Wong, J. H.; Yee, B. C.; Peters, A.; Kobrehel, K.;

Buchanan, B. B. Planta 1996, 200, 100-106.(37) Marx, C.; Wong, J. H.; Buchanan, B. B. Planta 2003, 216, 454-

460.(38) Johnson, T. E.; Wada, K.; Buchanan, B. B.; Holmgren, A. Plant

Physiol. 1987, 85, 446-451.(39) Wong, J. H.; Kobrehel, K.; Nimbona, C.; Yee, B. C.; Balogh, A.;

Kiss, F.; Buchanan, B. B. Cereal Chem. 1993, 70, 113-114.(40) Guttieri, M. J.; Stark, J. C.; O’Brien, K.; Souza, E. Crop Sci. 2001,

41, 327-335.(41) Laloi, C.; Mestres-Ortega, D.; Marco, Y.; Meyer, Y. Reichheld, J.

P. Plant Physiol. 2004, 134, 1006-1016.(42) Mouaheb, N.; Thomas, D.; Verdoucq, L.; Monfort, P.; Meyer, Y.

Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 3312-3317.(43) Marrs, K. A. Annu. Rev. Plant Physiol. Plant Mol. Biol. 1996, 47,

127-158.(44) Aalen, R. B.; Opsahl-Ferstad, H.-G.; Linnestad, C.; Olsen, O. A.

Plant J. 1994, 5, 385-396.(45) Stacy, R. A. P.; Munthe, E.; Steinum, T.; Sharma, B.; Aalen, R. B.

Plant Mol. Biol. 1996, 31, 1205-1216.(46) Haslekås, C.; Viken, M. K.; Grini, P. E.; Nygaard, V.; Nordgard, S.

H.; Meza, T. J.; Aalen, R. B. Plant Physiol. 2003, 133, 1148-1157.(47) Arrigoni, O.; De Gara, L.; Tommasi, F.; Liso, R. Plant Physiol. 1992,

99, 235-238.(48) Shinozaki, K.; Yamaguchi-Shinozaki, K. Curr. Opin. Plant Biol.

2000, 3, 217-223.(49) Sun, W.; Bernard, C.; Van de Cotte, B.; Van Montagu, M.;

Verbruggen, A. Plant J. 2001, 27, 407-415.(50) Coca, M. A.; Almoguera, C.; Jordano, J. Plant Mol. Biol. 1994, 25,

479-492.(51) DeRocher, A. E.; Vierling, E. Plant J. 1994, 5, 93-102.(52) Alamillo, J.; Almoguera, C.; Bartels, D.; Jordano, J. Plant Mol. Biol.

1995, 29, 1093-1099.



(53) Wehmeyer, N.; Vierling, E. Plant Physiol. 2000, 122, 1099-1108.(54) Marocco, A.; Santucci, A.; Cerioli, S.; Motto, M.; Di Fonzo, N.;

Thompson, R. D.; Salamini, F. Plant Cell 1991, 3, 507-515.(55) Sheffield, W. P.; Shore, G. C.; Randall, S. K. J. Biol. Chem. 1990,

265, 11069-11076.(56) Gaitanaris, G. A.; Papavassiliou, A. G.; Rubock, P.; Silverstein, S.

J.; Gottesman, M. E. Cell 1990, 61, 1013-1020.(57) James, P.; Pfund, C.; Craig, E. A. Science 1997, 275, 387-389.(58) Thornalley, P. J. Biochem. J. 1990, 269, 1-11.

(59) Thornalley, P. J. Biochem. Soc. Trans. 2003, 31, 1343-1348.(60) Yadav, S. K.; Singla-Pareek, S. L.; Ray, M.; Reddy, M. K.; Sopory,

S. K. Biochem. Biophys. Res. Commun. 2005, 337, 61-67.(61) Veena, R. V. S.; Sopory, S. K. Plant J. 1999, 17, 385-295.(62) Seki, M.; Narusaka, M.; Abe, H.; Kasuga, M.; Yamaguchi-

Shinozaki, K.; Carninci, P.; Hayashizaki, Y.; Shinozaki, K. PlantCell 2001, 13, 61-72.

PR060570J



Documents

Proteomics Uncovers a Role for Redox in Drought Tolerance in Wheat §