Differential expression of dehydrin genes in wild barley, Hordeum spontaneum, associated with resistance to water deficit

Plant, Cell and Environment

(2004)

27

, 1297–1308

© 2004 Blackwell Publishing Ltd

1297

Blackwell Science, LtdOxford, UKPCEPlant, Cell and Environment0016-8025Blackwell Science Ltd 2004? 2004

271012971308Original Article

Differential expression of dehydrin genes in wild barleyT. Suprunova

et al.

Correspondence: Professor Eviatar Nevo. Fax: 972 4 8288172; e-mail: [email protected]

*The first two authors should be considered as first authors, as theycontributed equally to this work.

Differential expression of dehydrin genes in wild barley,

Hordeum spontaneum

, associated with resistance to water deficit

T. SUPRUNOVA*, T. KRUGMAN*, T. FAHIMA, G. CHEN, I. SHAMS, A. KOROL & E. NEVO

Institute of Evolution, University of Haifa, Haifa 31905, Israel

ABSTRACT

Dehydrin gene (

Dhn

) expression is associated with plantresponse to dehydration. The aim of the present study wasto investigate the association of differential expression of

Dhn

genes (

Dhn 1, 3, 5, 6,

and

9

) with drought tolerancefound in wild barley (

Hordeum spontaneum

). Tolerant andsensitive genotypes were identified from Israeli (Tabighamicrosite) and Jordanian (Jarash and Waddi Hassa) popu-lations (based on scoring of water loss rate of 390 geno-types). The five

Dhn

genes were up-regulated bydehydration in resistant and sensitive wild barley geno-types. Notably, differences between resistant and sensitivegenotypes were detected, mainly in the expression of

Dhn1

and

Dhn6

genes, depending on the duration of dehydrationstress.

Dhn1

tended to react earlier (after 3 h) and higher(12 h and 24 h) in resistant compared to sensitive geno-types. The level of expression of

Dhn6

was significantlyhigher in the resistant genotypes at the earlier stages afterstress. However, after 12 and 24 h

Dhn6

expression wasrelatively higher in sensitive genotypes. The present resultsmay indicate that these genes have some functional role inthe dehydration tolerance in wild barley. The authors sug-gest that the observed differences of

Dhn

expression in wildbarley, originating from different micro- and macro ecogeo-graphic locations, may be the result of adaptive edaphic andclimatic selective pressures.

Key-words

: drought; ecogeographic location; edaphic selec-tive pressure; local adaptation; quantitative polymerasechain reaction.

INTRODUCTION

Water deficit is considered to be among the most severeenvironmental stresses that can cause water stress, oftenreferred as drought, and has an immediate effect on plantgrowth and yield (Araus

et al

. 2002). Plants respond towater stress through multiple physiological mechanisms atthe cellular, tissue, and whole-plant levels. These responses

are not only dependent upon the severity and duration ofthe water deficit, but also on the developmental stage andmorphological/anatomical parameter of the plants (Bartels& Souer 2004). Many genes respond experimentally towater stress, but their precise functions in either toleranceor sensitivity often remain unclear (Ludlow & Muchow1990; Smith & Griffiths 1993).

Dehydrins (DHNs), peripheral membrane proteins thatfunction in physical protection of the cell from water deficitor temperature change, are among the most frequentlyobserved proteins in plants under water stress. The

Dhn

genefamily encodes for several proteins that are typicallyaccumulated during the maturation–drying phase of seeddevelopment, in seedlings, or more mature plants inresponse to environmental stresses (Close 1997; Shinozaki& Yamaguchi-Shinozaki 1997; Zhu

et al

. 2000). These pro-teins are lipid-associated and may undergo function-relatedconformation changes at the water/membrane interface,perhaps related to the stabilization of vesicles or endo-membrane structures under stress conditions (Ismail, Hall& Close 1999; Koag

et al

. 2003). Most of the

Dhn

genes areup-regulated by environmental stresses such as drought,salinity, low temperature, or application of abscisic acid(ABA) (Close 1997; Zhu

et al

. 2000). Association betweentolerance to stresses involving dehydration (drought, salin-ity or freezing) and accumulation of members of the

Dhn

family has been established in cultivars such as wheat (Lab-hilili, Jourdier & Gautier 1995; Lopez

et al

. 2003), barley(Zhu

et al

. 2000), poplar (Pelah

et al

. 1997), and sunflower(Cellier

et al

. 1998). Variability in dehydrin genes has beenstudied in different species, e.g.

Pisum sativum

(Grosselin-demann

et al

. 1998),

Helianthus annuus

(Natali, Giordani &Cavallini 2003) and barley (Morrell, Lundy & Clegg 2003).In the barley genome, 12

Dhn

genes were identified (Choi,Zhu & Close 1999; Choi & Close 2000). Wide allelic varia-tion was found at the

Dhn4

locus in wild barley

(Hordeumspontaneum)

germplasm from Israel (Close

et al

. 2000).Moderate polymorphism associated with geographic struc-ture was found in

Dhn9

locus and high polymorphism withno geographic structure was found in

Dhn5

in a collectionof wild barley from the entire ecogeographic distribution ofwild barley, from the Mediterranean across the ZagrosMountains and into south west Asia (Morrell

et al

. 2003).Numerous studies have shown that wild progenitors of

cultivars comprise one of the major genetic recourses of

1298

T. Suprunova

et al

.

© 2004 Blackwell Publishing Ltd,

Plant, Cell and Environment,

27,

1297–1308

plant tolerance to stressful environments (Nevo

et al

. 1979,2002; Nevo 1992; Ellis

et al

. 2000; Robinson

et al

. 2000).Wild barley,

H. spontaneum,

the progenitor of cultivatedbarley, is a selfing annual grass of predominantly Mediter-ranean and Irano-Turanian distribution that penetrates intodesert environments where it maintains stable populations(Harlan & Zohary 1966). The wide ecological range of wildbarley differs in water availability, temperature, soil type,altitude, and vegetation generating a high potential ofadaptive diversity to abiotic stresses. Genetic diversity andphysiology in

H. spontaneum

from Israel and Jordan havebeen studied previously at

micro-

and

macro

geographicscales. In the

macro

geographical scale studies, adaptivegenetic diversity was detected by protein (Nevo

et al

. 1979)and DNA markers (Pakniyat

et al

. 1997; Owuor

et al

. 1999;Turpeinen

et al

. 2001, 2003; Baek, Beharav & Nevo 2003).At the

micro

eco-geographical scales, local adaptive geneticdiversity was found in several microsites in Israel (Ivandic

et al

. 2000, 2003; Gupta

et al

. 2002; Huang

et al

. 2002;Owuor

et al

. 2003). In particular, genetic and ecologicalfactors were correlated with carbon isotope discriminationin wild barley, which is related to water use efficiency(WUE), an important determinant of plant productivityunder limited water availability (Forster

et al

. 1994; Condon& Hall 1997). This adaptive genetic diversity found acrossthe genome and evidence of variability of dehydrin genes(Close

et al

. 2000; Morrell

et al

. 2003), indicate the potentialof wild barley as a source for drought resistance alleles forbreeding purposes.

Plant genetic adaptation to environmental stresses is dis-played in physiological and biochemical responses, con-trolled by up- and down-regulated changes in geneexpression. The interface between the molecular expres-sion mechanisms of stress responsive genes and plantresponse to drought stress is critical for translating molec-ular genetics into advances in crop production under stressconditions (Bruce, Edmeades & Barker 2002). We focushere on differential expression of dehydrin genes in wildbarley from the Mediterranean and Irano-Turanian regions,which is associated with resistance to dehydration.

MATERIALS AND METHODS

Plant material

Selection of wild barley genotypes, contrasting in responseto water stress, was performed using two approaches.

The first of these was exposure to short-term droughtstress at the seedling stage. A large-scale screening of wildbarley,

H. spontaneum,

seedlings (390 genotypes from 26populations, 3–20 genotypes per population, obtained fromthe collection of the Institute of Evolution, University ofHaifa, Israel) was carried out. These genotypes originatedfrom diverse eco-geographic regions of Israel and Jordanrepresenting a wide range of water availability habitats,with an annual rainfall of 100–1000 mm. Measurement ofthe water loss rate (WLR) of detached leaves was used asa quick screening method for selection of the most contrast-ing genotypes in their response to water stress (Appendix1). The most contrasting 40 genotypes (20 genotypes withlowest values of WLR and 20 genotypes with the highestvalues of WLR) and one barley cultivar (Noga) were cho-sen according to their WLR scores. Three plants from eachof the 41 genotypes were then tested again for WLR. As aresult of this screening, two ’putatively sensitive’ genotypes(Waddi Hassa JS1 and JS2) and two ’putatively resistant’genotypes (Jarash JR1 and JR2) were chosen for geneexpression analysis.

The second approach was exposure to long-term droughtstress of adult plants for which two additional genotypes,one resistant from Terra Rossa (TR) soil and one sensitivefrom Basalt (BA) soil, were chosen on the basis of theevaluation of 15 agronomic, morphological, developmental,and fertility-related traits tested by Ivandic

et al

. (2000)from the Tabigha microsite (Nevo

et al

. 1981).Altogether, six chosen genotypes were tested again (10

plants from each) for WLR (see below) and relative watercontent (RWC) under water stress. The ecogeographicaldata of the collection sites of the genotypes chosen forexpression analysis are presented in Table 1.

Water loss rate

The water loss rate (WLR) of 390 genotypes was measuredas described by Clarke & McCaig (1982) with slight modi-fication. Seedlings were grown on moist filter paper in Petridishes at room temperature. Fully expanded first leaveswere cut and fresh weight (FW) was immediately recorded(time 0). The leaves were left on filter paper for 24 h andthe weight (W

24

) was measured again. Total dry weight(DW) was recorded after drying for 24 h at 80

∞

C. WLR wascalculated according to the formula:

Table 1.

Ecogeographical data of the site of origin of six genotypes of

H. spontaneum

from Jordan and Israel

Site of origin Genotype ResistanceAltitude(m)

Mean annualrainfall (mm)

Mean annualtemperature (

∞

C) Soil type

Wadi Hassa, Jordan JS1, JS2 S 300 250–300 15.7 Sedimentary rocks: basalt, limestone, shaleTabigha, Israel BA S 93 480 24.5 BasaltTabigha, Israel TR R 93 480 24.5 Terra RossaJarash, Jordan JR1, JR2 R 630 290 18.9 Terra Rossa

Meteorological data of Wadi-Hasa and Jarash were obtained from Meteorological Department of the Hashemite Kingdom of Jordan; dataof Tabigha described by Nevo & Beiles (1989). S, sensitive genotypes; R, resistant genotypes.

Differential expression of dehydrin genes in wild barley

1299

© 2004 Blackwell Publishing Ltd,

Plant, Cell and Environment,

27,

1297–1308

WLR (g h

-

1

g

-

1

DW) = (FW

-

W

24

)/(DW

¥

24)

The WLR of the six selected resistant and sensitive geno-types was measured according to Ristic & Jenks (2002). Thetime [

T

x

(min); x = 0] at which the blades of fully expandedfirst leaves were first weighed was considered as 0 min(T

0

= 0 min). The leaf blades were weighed five times. Thetime of each measurement (min) was recorded as time

T

x

where x = 2, 4, 6 and 8 (every 2 h); such that 2 indicatestime of the first measurement and 8 indicates time of thelast measurement. Total dry weight (DW) was recordedafter drying for 24 h at 80

∞

C. WLR was calculated for eachtime point according to the formula:

WLR (g h

-

1

g

-

1

DW) = [(F

Tx

-

F

Tx+2

)

¥

60]/[DW

¥

(T

x+2

-

T

x

)]

Leaf relative water content

Seedlings were grown on moist filter paper in Petri dishesat room temperature. Leaf relative water content (RWC)was measured under control (well-watered plants) anddrought stress conditions. For drought stress 10-day-oldplants were placed onto dry filter paper for 24 h. Freshweight (FW) of fully expanded leaves was immediatelyrecorded after leaf excision. The leaves were soaked indistilled water for 24 h at 4

∞

C in darkness and the turgidweight (TW) was recorded. Total dry weight (DW) was thenrecorded after drying for 24 h at 80

∞

C. RWC was calculatedaccording to Barrs & Weatherley (1968):

RWC (%) = [(FW

-

DW)/(TW

-

DW)]

¥

100.

Dehydration treatment for expression analysis

Seedlings were grown in a greenhouse at 22

∞

C, with aphotoperiod of 12 h light/12 h dark, in Murashige andSkoog basal salt mixture (MS) solution (Sigma ChemicalCo., St Louis, MO, USA), circulated by air pumps. Droughtstress was applied to 10-day-old seedlings by draining thesolution from the container for defined dehydration peri-ods. Leaf tissues of two seedlings per genotype were har-vested from control plants (time 0); and after 3, 12, and 24 hof dehydration, frozen in liquid nitrogen, and stored at

-

80

∞

C for RNA extraction.

RNA isolation and reverse transcriptase - polymerase chain reaction analysis

Total RNA was extracted from leaves using EZ-RNA TotalRNA Isolation Kit (Biological Industries, Beit HaemekLTD, Israel). RNA was treated by RNase-free

DNase

I(Ambion, Inc., Austin, TX, USA) for removal of templateDNA before first-strand synthesis. First-strand cDNA wasprepared from 2

m

g of total RNA, using universalOligo(dT)

15

primer and 200 units of SuperScript II reversetranscriptase (Invitrogen, Carlsbad, CA, USA), at 42

∞

C for1 h in a 20-

m

L reaction volume. The resulting single-strandcDNA was amplified with

Taq

DNA polymerase (SigmaChemical Co) using gene-specific 5

¢

-end and 3

¢

-end primersthat were designed based on different exon sequences of

Dhn

genes to produce different-size polymerase chain reac-tion (PCR) products amplified from DNA and RNA tem-plates (Choi

et al

. 1999). A fragment of barley

a

–

tubulin

gene was amplified with specific primers (forward: 5

¢

-AGTGTCCTGTCCACCCACTC-3

¢

; reverse: 5

¢

-AGCATGAAGTGGATCCTTGG - 3

¢

) as an internal control for the rel-ative amount of RNA. PCR was performed in a GeneAmpPCR 9700 (Applied Biosystems, Foster City, CA, USA)starting with a denaturation step of 3 min, followed by 26cycles (22 cycles for

a

–

tubulin

gene) of 95

∞

C for 30 s, 65

∞

Cfor 30 s, 72

∞C for 30 s, and terminated at 72 ∞C for 10 min.PCR products were separated on 1.2% agarose gels andstained with ethidium bromide for photography (EagleyeII; Stratagene, La Jolla, CA, USA).

Quantitative real-time polymerase chain reaction analysis

Gene quantification was performed using ABI PRISM 7000Sequence Detection System (Applied Biosystems). Specificprimer pairs of studied genes were designed based onsequences presented in the Gene Bank database usingPrimer Express 2 software (Table 2). Each reaction wasperformed on 5 mL of 1 : 100 (v/v) dilution of the first-strand cDNA, synthesized as described above, in a totalreaction volume of 25 mL using SYBR Green PCR MasterMix (Applied Biosystems) and 300 nM of each forward andreverse primer. Reaction conditions for thermal cyclingwere: 50 ∞C for 2 min, 95 ∞C for 10 min, followed by 40cycles of 95 ∞C for 15 s and 60 ∞C for 1 min. Amplification

Table 2. Primers used in real-time PCR expression analysis

Target geneGeneBank Accession No. Sequence of primers (5¢ to 3¢) Amplicon size (bp)

Dhn1 AF043087 GCAACAGATCAGCACACTTCCA 95GCTGACCCTGGTACTCCATTGT

Dhn6 AF043091 TTTTACCGTGTGATAGATGTTGCA 72TGCAAACCGACCAGACAAACT

a-tubulin U40042 TCCATGATGGCCAAGTGTGA 72GACATCCCCACGGTACATGAG

1300 T. Suprunova et al.

© 2004 Blackwell Publishing Ltd, Plant, Cell and Environment, 27, 1297–1308

specificity was checked with a heat dissociation protocol(melting curves in 60–95 ∞C range), as a final step of thePCR. All primer pairs showed a single peak on the meltingcurve, and a single band of the expected size was observedusing agarose gel electrophoresis.

The standard curves were generated for each studiedgene using serial dilutions of an experimental cDNA sam-ple that showed the maximal amount of target gene inpreliminary reverse transcriptase (RT)-PCR analysis. Tar-get quantities in the tested samples were automatically cal-culated by the supported software. Calculation is based onthe intensity of the reporter dye fluorescence in the thresh-old cycle (Ct) of each sample interpolated to the standardcurve. In order to account for differences in target RNApresented in each sample, both Dhn1 and Dhn6 gene quan-tities were normalized to the barley a–tubulin as house-keeping gene, which is not affected by various stressconditions (Kawasaki et al. 2001; Nemoto & Sasakuma2002; Ozturk et al. 2002). Two independent plant samplesfor each genotype were examined in triplicate.

Data analysis

Means, standard deviation (SD), and comparisons of thephysiological measurements and results of the expressionanalysis (real time PCR results) were performed using theSTATISTICA package (Statsoft 1996). One-, two- andthree-way ANOVA were employed for testing the signifi-cance of the genotype, resistance-sensitivity, and durationof stress on the expression level.

RESULTS

Physiological analysis of wild barley genotypes

Water loss rate

In order to identify contrasting genotypes in their responseto short-term severe water stress, we measured the WLR

of 390 genotypes from 26 Israeli and Jordanian populationsof H. spontaneum. The assessment of water loss fromexcised leaves has shown promise for characterizingdrought resistance of wheat genotypes and could be easilydetermined (Clarke & McCaig 1982; Ristic & Jenks 2002).The values of WLR of the tested genotypes were highlyvariable among genotypes, within and between populationsand ranged from 0.095 to 0.215 g h-1 g-1 DW (Appendix 1).The most contrasting 40 genotypes and one barley cultivar(Noga) were tested again for WLR (Fig. 1). The WLR ofthe tested cultivar H. vulgare (Noga), regarded as the mostdrought resistant cultivar in Israel, was more than0.22 g h-1 g-1 DW, higher than the most sensitive wild barley(Fig. 1). As a result of this screening, two ’putatively sensi-tive’ genotypes (Waddi Hassa JS1 and JS2) and two ’puta-tively resistant’ genotypes (Jarash JR1 and JR2) werechosen for gene expression analysis. Detailed measure-ments of WLR at the early stage of dehydration (2, 4, 6,and 8 h) showed that water loss rate is stabilizing after 6 h,and there is no significant change between these values andWLR after 24 h (this kind of analysis could not be donewith 390 genotypes because of technical reasons) (Fig. 2).The WLR of the sensitive genotypes (JS1 and JS2) origi-nating from Wadi Hassa in Jordan, were significantly higherthan the resistant genotypes (JR1 and JR2) originatingfrom Jarash in Jordan in all time points and over the entireexperiment (F1,21 = 40.59, P < 5 ¥ 10-6). WLR of the twogenotypes from the Israeli Tabigha microsite (TR and BA)were significantly different after 2 h (P = 5 ¥ 10-5), how-ever, the difference declined with time and became non-significant (P = 0.647) after 4 h of dehydration (Fig. 2).

Relative water content is considered as an alternativemeasure of plant water status reflecting the metabolic activ-ity in leaf tissues (Flower & Ludlow 1986). Drought resis-tance of a plant is related to its ability to maintain higherrelative water content in the leaves under water stress. Nosignificant difference (P = 0.328) was detected among thegenotypes in RWC in well-irrigated plants (RWC ª 96% for

Figure 1. Water loss rate (WLR) under drought stress of 40 wild barley geno-types representing Jordanian and Israeli populations.The barley seedlings (after 10 d of growth) were used for measurement of WLR according to Clarke & McCaig (1982). Each value is the mean ± SE (n = 3).

Differential expression of dehydrin genes in wild barley 1301

© 2004 Blackwell Publishing Ltd, Plant, Cell and Environment, 27, 1297–1308

each genotype) (Fig. 3). RWC of all genotypes declined withwater stress, and significantly lower (P = 0.003) RWC valueswere found in sensitive genotypes (JS1, JS2, and BA) thanin resistant genotypes (JR1, JR2, and TR) (Fig. 3).

RT-PCR analysis of expression of Dhn genes

The expression characteristics of five Dhn genes were stud-ied in the six wild barley genotypes that were defined as’sensitive’ and ’resistant’ by leaf WLR and RWC measure-ments, as described above. To determine the expression ofindividual members of the dehydrin gene family, RT-PCRanalysis with gene-specific 5¢-end and 3¢-end primers foreach Dhn gene was performed using the same RNA iso-lated from well-watered (control) and drought-stressedplants (after 3, 12, and 24 h of dehydration). Differences inexpression patterns were found in each of the Dhn genesdepending on duration of dehydration stress. Dhn1, 3, 6,and 9 were not expressed in well-watered plants, however,low levels of Dhn5 were detected in TR- and BA-

genotypes. All five tested dehydrin genes were up-regulatedunder water stress (Fig. 4). Expression of Dhn3 and Dhn9was detected in the resistant and sensitive genotypes after3 h of dehydration, with the exception of genotype JS1. Noclear differences in expression of these genes were foundbetween the tested resistant and sensitive genotypes after12 and 24 h under drought stress. Dhn 1, 5, and 6 displayeddifferences in expression levels between resistant and sen-sitive plants: earlier induction of Dhn1 (after 3 h dehydra-tion) and higher levels (after 12 and 24 h dehydration) ofexpression were observed in the resistant plants (JR1 andJR2) as compared to the sensitive plants. The expression ofDhn1 in TR-genotype was not observed after 3 h dehydra-tion, however, it was detectable by real-time PCR (seebelow). Dhn5 was expressed after 3 h of dehydration stressin all genotypes and increased after 12 h of dehydration; aslightly higher expression level was detected in the resistantgenotypes as compared to sensitive ones. Higher expressionof Dhn6 was observed in the resistant genotypes after 3 hdehydration, however, after 12 and 24 h of dehydration the

Figure 2. Dynamics of leaf water loss rate (WLR) of six wild barley gen-otypes under drought stress. Rate of epidermal water loss in intact leaves of wild barley seedlings (after 10 d of growth) was measured according to Ristic & Jenks (2002). Each point is the mean ± SD (n = 10).Time (h)

Figure 3. Relative water content (RWC) of wild barley genotypes under drought stress. The barley seedlings (after 10 d of growth) were used for mea-surement of RWC according to Barrs & Weatherley (1968). Data are shown as the means ± SD (n = 6). The letters on the top of each column indicate statistical significant difference (P < 0.05, LSD test).

1302 T. Suprunova et al.

© 2004 Blackwell Publishing Ltd, Plant, Cell and Environment, 27, 1297–1308

sensitive plants showed higher expression level as com-pared to the resistant plants (Fig. 4).

Real-time PCR analysis of dehydrin genes expression

In order to confirm the results obtained by RT-PCR, weused the quantitative real time PCR that allows quantifyingthe absolute level of the targets by fluorescence detectionof PCR product following each cycle of the reaction. More-over, the absolute level of transcription of genes withextremely low expression level could be estimated by real-time PCR (Charrier et al. 2002). From the five Dhn genestested by RT-PCR, we chose Dhn1 and Dhn6, whichshowed different trends of expressions. Namely, Dhn1 tran-scripts appeared earlier (after 3 h) and at a higher intensity(after 12 and 24 h) in the resistant plants than in the sensi-tive plants. Dhn6 was expressed higher (after 3 h) and prob-ably earlier (earlier time points were not studied) in theresistant genotypes, but showed slightly higher expressionafter 12 and 24 h of drought in the sensitive plants. Thefollowing pattern was found.

Dhn1

The results of analysis by real-time PCR of Dhn1 expres-sion after 3, 12, and 24 h of dehydration are presented inFig. 5. Expression of Dhn1 was not detectable in well-watered plants of all wild barley genotypes; hence, thosedata are not presented in Fig. 5. The three resistant geno-types displayed earlier induction of Dhn1 after 3 h of

drought stress. The expression level of JR1 was by farhigher (360-fold) than in the sensitive plants (P < 0.02)(Fig. 5). Variation in the expression of Dhn1 within theconsidered resistant group was observed also after 24 h ofdehydration. In the sensitive group, at the early stage (3 h)very low expression level (0.005) was observed only in JS1;variation between the genotypes was observed after 12 hdehydration (especially in JS1). All genotypes reached theirmaximal level of Dhn1 expression after 24 h of droughtstress with the unexpected exception of the sensitive geno-type JS1 that after 12 h stress displayed the maximalexpression followed by reduced activity of Dhn1. The high-est level of expression was observed after 24 h in resistantgenotypes JR1 and JR2 (between 22 to 68% higher) ascompared with JS1 and JS2, but the difference was notsignificant (P > 0.2). In general, earlier induction of Dhn1was observed in the resistant plants, and a slightly higherexpression level was displayed in the resistant plants after12 and 24 h of dehydration stress. In the TR and BA geno-types from Tabigha that were selected by their whole plantperformance under stress, similar trends of expression pat-terns were found: Dhn1 expression was higher in resistantplants originating from the TR soil as compared with thegenotype from the BA soil, but only after 3 h of droughtstress the difference was significant (P = 0.011) (Fig. 5).

Dhn6

Expression of Dhn6 was not detectable in the control plantsof all genotypes; hence, those data are not presented inFig. 6. A clear difference in expression level of Dhn6 was

Figure 4. Differential expression pat-terns of Dhn1, 3, 5, 6 and 9 detected by RT-PCR. The RT-PCR was carried out with gene specific primers, using cDNA obtained from six wild barley genotypes (JS1, JS2, BA, JR1, JR2, TR) after 0 (con-trol, C), 3, 12 and 24 h of dehydration. As a control for relative amount of RNA, RT-PCR with gene specific primers for a–tubulin (Tub) was performed.

C 3 h 12 h 24 h C 3 h 12 h 24 h C 3 h 12 h 24 h

C 3 h 12 h 24 h C 3 h 12 h 24 h C 3 h 12 h 24 h

Differential expression of dehydrin genes in wild barley 1303

© 2004 Blackwell Publishing Ltd, Plant, Cell and Environment, 27, 1297–1308

detected between sensitive and resistant plants (Fig. 6).After 3 h of dehydration stress, expression level was higher(two- to four-fold) in all resistant genotypes (JR1, JR2, andTR) than in all sensitive genotypes (JS1, JS2, and BA)(F1.4 = 37.19; P = 0.009). The opposite trend of transcriptaccumulation was observed after 12 h: expression in thesensitive plants JS1 and JS2 was significantly (P = 0.032)higher (35 to 83%) than in the resistant plants JR1 and JR2,whereas BA and TR were not different. After 24 h of dehy-dration, expression of Dhn6 was not changed in the sensi-tive JS1 and increased slightly in JS2, while reduction wasobserved in the resistant plants JR1 and JR2; the TR gen-otype showed 33% higher expression than the BA geno-type. In general, the most significant difference wasobserved after 3 h of drought stress, when all resistantplants displayed a higher level of Dhn6 expression.

DISCUSSION

Differential expression of Dhn genes

The last step in the dehydration-signalling cascade is thealternation of genes responsible for the synthesis of com-pounds that serve to protect cellular structures against thedeleterious effects of dehydration, such as proteins withprotective functions encoded for by the late embryogenesisabundant or lea genes (Bartels & Souer 2004). The dehy-drin gene family belongs to this group and it is one of themost studied drought-inducible gene families. Most of thesegenes are differentially regulated at the transcriptionallevel under environmental stresses such as water deficit,salinity, and low temperature or in response to ABA (Close1997; Zhu et al. 2000). The aim of this research was to studythe association of the differential expression of Dhn genes(Dhn 1, 3, 5, 6, and 9) with drought tolerance in wild barley,

Figure 5. Expression of Dhn1 detected by quantitative real-time PCR. Real-time PCR was carried out with cDNA obtained from six wild barley genotypes (JS1, JS2, BA, TR, JR1, and JR2) after 3, 12, and 24 h of dehydration. Quantification is based on Ct values that were normalized using the Ct value corresponding to a barley (housekeeping) a–tubulin gene. Two independent plant samples for each genotype were examined in triplicate. Each value is the mean ± SE (n = 2).

Figure 6. Expression of Dhn6 detected by quantitative real-time PCR. Real-time PCR was carried out with cDNA obtained from six barley genotypes (JS1, JS2, BA, TR, JR1, and JR2) after 3, 12, and 24 h of dehydration. Quantification is based on Ct values that were normalized using the Ct value corresponding to a barley (housekeeping) a–tubulin gene. Two independent plant samples for each genotype were examined in triplicate. Each value is the mean ± SE (n = 2).

1304 T. Suprunova et al.

© 2004 Blackwell Publishing Ltd, Plant, Cell and Environment, 27, 1297–1308

H. spontaneum, from its centres of diversity in Israel andJordan. An association between tolerance to stresses witha dehydrative component (drought, freezing, and salinity)and gene expression or proteins has been observed in somecrop species (Labhilili et al. 1995; Pelah et al. 1997; Cellieret al. 1998; Zhu et al. 2000; Lopez et al. 2003). In most cases,the drought resistance and sensitivity of the testedgenotypes were not characterized by physiologicalmeasurements.

Here, we describe intraspecific variation of Dhn genes inresponse to short-term severe dehydration stress in geno-types of wild barley contrasting in drought tolerance. Inorder to be able to relate the differential gene expressionnot only to plant reaction to water stress, but also to dehy-dration tolerance, ’drought resistant’ and ’drought sensi-tive’ genotypes were identified from a wide collection ofwild barley. The resistant genotypes were from Jarash andsensitive genotypes from Waddi Hassa, from north andsouth Jordan. These selected genotypes were different intheir response to drought at the seedling stage, accordingto measurements of leaf WLR and RWC. Drought resis-tance can be achieved by various types of mechanisms, notnecessarily physiological ones. As water deficit has a nega-tive impact on whole plant productivity, two other geno-types that were contrasted in their plant performance underwater stress in a previous study (Ivandic et al. 2000, 2003)were added to this study for comparison. These genotypeswere originally collected from alternative soil types (Basaltand Terra Rossa) along a 100-m transect (Nevo et al. 1981).These two genotypes from the Tabigha microsite, proved todiffer only after short dehydration (as displayed by theWLR after 2 h), but the difference declined after 4 h(Fig. 2). More stable differences between the selectedgroups of resistant and sensitive genotypes under 24 h ofdrought stress were found for the RWC test. An associationwas found between whole plant performance under pro-longed drought stress in field conditions and dehydrationstress at the seedling stage. This finding is especially impor-tant for the eastern Mediterranean environment, with dryconditions common during the sensitive stage of emergenceand early growth, which along with high temperatures andincreasing water demands at the end of the spring result ina low yield.

In our tests, Dhn genes (Dhn 1, 3, 5, 6, and 9) were up-regulated by water deficit imposed by dehydration, whichare similar to the results previously obtained in cultivatedbarley (Choi et al. 1999). Differences in expression patternswere observed in the Dhn genes depending on duration ofdehydration stress (Fig. 4). These results indicate that eachmember of this gene family may have different functions inthe process of plant response to drought. Clear differencesbetween sensitive and resistant genotypes were detected byRT-PCR of Dhn1 and Dhn6 and confirmed by quantitativereal-time PCR (Figs 5 & 6). Real-time PCR was alreadyapplied successfully for quantifying the level of expressionof plant genes both under normal growth conditions andunder stressful abiotic treatment (Svensson et al. 2002).Using this method, we were able to quantify and compare

the level of expression of the Dhn genes in different geno-types under various durations of stress. The results obtainedby quantitative real-time PCR show that Dhn1 tended toreact earlier and stronger in the resistant as compared tosensitive genotypes (Fig. 5). At earlier stages after stress,Dhn6 was expressed in significantly higher amounts in theresistant than in sensitive genotypes (Fig. 6). Most probablyDhn6 expression was also earlier in resistant plants; how-ever, earlier time points were not studied. While accumu-lation of Dhn1 was relatively higher in the resistantgenotypes after 12 and 24 h dehydration, accumulation ofDhn6 showed a change in the trend (except for the BAgenotype): relatively higher expression in sensitive geno-types (Fig. 6). Variation in the expression within the consid-ered resistant and sensitive groups was also observed,similar to the physiological response to dehydration.

The earlier expression of Dhn1 and Dhn6 in resistantgenotypes may indicate that the resistance is due to one ormore of the following mechanisms and/or their combina-tion: (1) earlier perception of the water stress (2), moreefficient signalling pathways and transcriptional activators,and (3) higher expression of the Dhn gene. Time courseexperiments in several plants have shown that water deficitis sensed very rapidly – long before symptoms such as wilt-ing become manifest and before the relative water contentdecreases significantly. Transcripts and proteins indicativeof a dehydration response are detectable within 60 minafter the onset of dehydration in the resurrection plantCraterostigma plantaginum and in Arabidopsis thaliana(Bartels & Souer 2004). The analysis of differential geneexpression by micro-array analysis has identified a broadspectrum of transcripts whose expression is modified inresponse to dehydration in A. thaliana (Seki et al. 2002) andin a relatively drought-tolerant barley cultivar (Ozturk et al.2002). Moreover, microarray analysis showed that in theinitial response to salt stress, individual up-regulated tran-scripts characteristic of salt-tolerant rice were absent insalinity-sensitive rice (Kawasaki et al. 2001). These authorsindicate that the resistant cultivars can overcome the stressdue to their ability to induce transcripts that, among otherfunctions, stimulate protein synthesis and components ofsignalling circuits, while the sensitive cultivars show a delayin responding by up-regulation and fewer responses in total(Kawasaki et al. 2001). Our findings in resistant and sensi-tive wild barley corroborate well with these results.

Drought tolerance is a complex trait, which cannot beanalysed genetically in the same way as monogenic resis-tance. However, processes involved in drought tolerancecan be dissected by molecular genetic approaches throughmarker-based detection and mapping of relevant quantita-tive trait loci (QTL) and/or co-localization of QTLs withcandidate genes. Our finding on the role of Dhn1 in droughttolerance of wild barley is supported by several reports onco-localization of such QTLs with Dhn genes in barley, e.g.QTLs for RWC (Teulat et al. 2003) and winter-hardiness(Pan et al. 1994; Zee et al. 1995) overlapping with a clusterof Dhn genes on chromosome 5H. Further tests are under-way to unravel the genome distribution of QTLs affecting

Differential expression of dehydrin genes in wild barley 1305

© 2004 Blackwell Publishing Ltd, Plant, Cell and Environment, 27, 1297–1308

the resistance-related multitrait complexes. The targetedmapping populations are based on crosses between resis-tant and susceptible H. spontaneum as well as resistant H.spontaneum (Chen et al. in preparation) and susceptiblecultivar of H. vulgare (Zhang et al. in preparation). Thesetargeted populations will allow us to detect genomicregions shared between drought-resistance candidate genes(including the Dhn family) and resistance QTLs, strength-ening the functional interpretation of the candidate geneexpression patterns.

Microgeographic adaptation of wild barley to water stress

The drought-resistant and -sensitive genotypes selected forthis study were from a collection of wild barley represent-ing the entire ecogeographic distribution in Israel and Jor-dan, based on their physiological response to dehydrationas an objective test, rather than by their site of origin. Wefound wide variation within and between populations withrespect to WLR. The main difference between the collec-tion sites of the selected genotypes was soil type. The resis-tant genotypes from Jordan originated from Terra Rossasoil (Jarash), similar to the origin of the resistant genotypefrom the Tabigha microsite in Israel. The sensitive geno-types originated from Basalt soil type (Wadi Hassa) pos-sessed a greater water-holding capacity compared with TR,similar to the basalt-sensitive genotypes from Tabigha(Ivandic et al. 2000). The subpopulations from the Tabighamicrosite shared the same annual rainfall (480 mm), andthe two collection sites from Jordan were similar to eachother in their annual rainfall (250–300 mm). The soils inWadi Hassa, the origin site of the sensitive genotypes,formed through weathering of limestone and basalt, arevery rich and productive when combined with enoughmoisture (http://vkrp.org/studies/environmental/geomor-phology/). Wadi Hassa is one of riverbeds, which dominatedrainage of the Karak plateau. Therefore, water availabilityat the bottom of the Wadi is relatively high and the plantsare exposed to more water during their life history, despitethe macroclimatic xeric surrounding. Water availability isknown to impose strong selective pressure, as it is funda-mental to almost all aspects of plant physiology (Stephen-son 1990; Prentice et al. 1992; Bray 1994; Bohnert, Nelson& Jensen 1995; Pérez 1996). The above features mayexplain why the sensitive genotypes were found at this col-lection site.

A long history of research on the genetics of ecologicalecotypes and natural populations within plant species pro-vides support for climate as a selective pressure to whichpopulations adapt locally (Nagy 1997; Owuor et al. 1999,2003; Ivandic et al. 2000; Gupta et al. 2002; Huang et al.2002). We suggest that the observed differences, both inphysiology and Dhn gene expression, may be a result of acombination of edaphic and climatic selective pressures.Adaptation to environmental changes in populationsrequires the existence of a large spectrum of genetic diver-sity and phenotypic plasticity for physiological traits. It

characterizes species from highly fluctuating ecosystems(Hoffmann & Parsons 1991). Indeed, high genetic diversitywas found in the Terra Rossa subpopulation of Tabigha ascompared with the Basalt subpopulation (Nevo & Beiles1989; Owuor et al. 1999) and the Jarash population as com-pared with other populations from Jordan (Baek et al.2003). Drought-resistant wild barley genotypes originatingin populations with a high level of genetic diversity andphenotypic plasticity for physiological traits, can serve as agood source for identifying drought-resistant alleles forbreeding purposes.

The remarkable finding of this study is in the following:(1) two-step screening of a few hundred wild barley geno-types with respect to WLR that is related to drought toler-ance revealed large variation, and the most contrastinggenotypes were originated from ecologically divergentmicroclimatic niches; (2) the alternative for drought toler-ance groups displayed different patterns in the dynamics ofdesiccation-induced expression of dehydrin candidategenes (Dhn1 and Dhn6). This was revealed by RT-PCR andreal-time PCR analysis. Our further efforts will be devotedto elucidate whether the observed differential expressionrelates to allelic variation in the promoter region of thesegenes or, if alternatively, trans-regulatory factors areinvolved (i.e. alleles of transcription factor genes or othercomponents of the signal transduction pathways).

ACKNOWLEDGMENTS

This work was supported by grants from the German-Israeli Cooperation Project (DIP project No. DIP-B 4.3)funded by the BMBF and supported by BMBF’s Interna-tional Bureau at the DLR; The Israeli Science Foundation(No. 9030/96, 9048/99, and 9019/01) and the financial sup-port of the Israel Discount Bank Chair of EvolutionaryBiology and the Ancell-Teicher Research Foundation forMolecular Genetics and Evolution.

REFERENCES

Araus J.L., Slafer G.A., Reynolds M.P. & Royo C. (2002) Plantbreeding and drought in C3 cereals: what should we breed for?Annual Botany (Lond) 89, 925–940.

Baek H.J., Beharav A. & Nevo E. (2003) Ecological-genomicdiversity of microsatellites in wild barley, Hordeum spontaneum,populations in Jordan. Theoretical and Applied Genetics 106 (3),397–410.

Barrs C. & Weatherley P.E. (1968) A re-examination of the rela-tive turgidity technique for estimating water deficit in leaves.Australian Journal of Biological Science 15, 413–428.

Bartels D. & Souer E. (2004) Molecular responses of higher plantsto dehydration. In Plant Responses to Abiotic Stress (eds H. Hirt& K. Shinozaki), pp. 9–38. Springer-Verlag, Berlin, Heidelberg.

Bohnert H.J., Nelson D.E. & Jensen R.G. (1995) Adaptations toenvironmental stress. Plant Cell 7, 1099–2111.

Bray E.A. (1994) Alterations in gene expression in response towater deficit. In Stress-Induced Gene Expression in Plants (ed.A.S. Basra), pp. 1–23. Harwood Academic Publishers,Ludhiana, India.

Bruce W.B., Edmeades G.O. & Barker T.C. (2002) Molecular and

1306 T. Suprunova et al.

© 2004 Blackwell Publishing Ltd, Plant, Cell and Environment, 27, 1297–1308

physiological approaches to maize improvement for droughttolerance. Journal of Experimental Botany 53, 13–25.

Cellier F., Conejero G., Breitler J.C. & Casse F. (1998) Molecularand physiological responses to water deficit in drought-tolerantand drought-sensitive sunflower lines (Helianthus annuus L):accumulation of dehydrin transcripts correlates with tolerance.Plant Physiology 116, 319–328.

Charrier B., Champion A., Henry Y. & Kreis M. (2002) Expres-sion profiling of the whole Arabidopsis shaggy-like kinase mul-tigene family by real-time reverse transcriptase-polymerasechain reaction. Plant Physiology 130, 577–590.

Choi D.W. & Close T.J. (2000) A newly identified barley gene,Dhn12, encoding YSK2 DHN, is located on chromosome 6H andhas embryo-specific expression. Theoretical and Applied Genet-ics 100, 1274–1278.

Choi D.W., Zhu B. & Close T.J. (1999) The barley (Hordeumvulgare L.) dehydrin multigene family: sequences, allelic types,chromosome assignments, and expression characteristics of 11Dhn genes of cv. Dicktoo. Theoretical and Applied Genetics 98,1234–1247.

Clarke J.M. & McCaig T.N. (1982) Evaluation of techniques fordrought resistance in wheat. Crop Science 22, 503–506.

Close T.J. (1997) Dehydrins: a commonalty in the response ofplants to dehydration and low temperature. Physiologia Plan-tarum 100, 291–296.

Close T.J., Choi D.W., Venegas M., Salvi S., Tuberosa R.,Ryabushkina N., Turuspekov Y. & Nevo E. (2000) Allelic vari-ation in wild and cultivated barley at the Dhn4 locus, whichencodes a major drought-induced and seed protein, DHN 4.8thInternational Barley Genetics Symposium 22–27 October 2000,(ed. S. Logue) Adelaide, South Australia.

Condon A.G. & Hall A.E. (1997) Adaptation to diverse environ-ments: genotypic variation in water-use efficiency within cropspecies. In Agricultural Ecology (ed. L.E. Jackson), pp. 79–116.Academic Press, San Diego, CA, USA.

Ellis R.P., Forster B.P., Robinson D., Handley L.L., Gordon D.C.,Russell J.R. & Powell W. (2000) Wild barley: a source of genesfor crop improvement in the 21st century? Journal of Experi-mental Botany 51, 9–17.

Flower D.J. & Ludlow M.M. (1986) Contribution of osmoticadjustment to the dehydration tolerance of water-stressedpigeonpea (Cajanus cajan (L.) millsp.) leaves. Plant, Cell andEnvironment 9, 33–40.

Forster B.P., Handley L.L., Pakniyat H., Scrimgeour C.B., NevoE. & Raven J. (1994) Genetic and ecological factors controllingcarbon isotope discrimination in barley. Aspects of Applied Biol-ogy 38, 139–143.

Grosselindemann E., Robertson M., Wilmer J.A. & Chandler P.M.(1998) Genetic variation in pea (Pisum) dehydrins: sequenceelements responsible for length differences between dehydrinalleles and between dehydrin loci in Pisum sativum L. Theoret-ical and Applied Genetics 96, 1186–1192.

Gupta P.K., Sharma P.K., Balyan H.S., Roy J.K., Sharma S.,Beharav A. & Nevo E. (2002) Polymorphism at rDNA loci inbarley and its relation with climatic variables. Theoretical andApplied Genetics 104, 473–481.

Harlan R.J. & Zohary D. (1966) Distribution of wild wheat andbarley. Science 153, 1074–1080.

Hoffmann A.A. & Parsons P.A. (1991) Evolutionary Genetics andEnvironmental Stress. Oxford University Press, Oxford, UK.

Huang Q., Beharav A., Li Y., Kirzhner V. & Nevo E. (2002)Mosaic microecological differential stress causes adaptive mic-rosatellite divergence in wild barley, Hordeum spontaneum, atNeve Yaar, Israel. Genome 45, 1216–1229.

Ismail A., Hall A.E. & Close T.J. (1999) Purification and partialcharacterization of a dehydrin involved in chilling tolerance dur-

ing seedling emergence of cowpea. Plant Physiology 120, 237–244.

Ivandic V., Hackett C.A., Zhang Z.J., Staub J.E., Nevo E., ThomasW.T.B. & Forster B.P. (2000) Phenotypic responses of wild bar-ley to experimentally imposed water stress. Journal of Experi-mental Botany 51, 2021–2029.

Ivandic V., Thomas W.T.B., Nevo E., Zhang Z. & Forster B.P.(2003) Association of simple sequence repeats with quantitativetrait variation including biotic and abiotic stress tolerance inHordeum spontaneum. Plant Breeding 122, 300–304.

Kawasaki S., Borchert C., Deyholos M., Wang H., Brazille S.,Kawai K., Galbraith D. & Bohnert H.J. (2001) Gene expressionprofiles during the initial phase of salt stress in rice. Plant Cell13, 889–905.

Koag M.C., Fenton R.D., Wilkens S. & Close T.J. (2003) Thebinding of maize DHN1 to lipid vesicles. Gain structure and lipidspecificity. Plant Physiology 131, 309–316.

Labhilili M., Jourdier P. & Gautier M.F. (1995) Characterizationof cDNAs encoding Triticum durum dehydrins and their expres-sion patterns in cultivars that differ in drought tolerance. PlantScience 112, 219–230.

Lopez C.G., Banowetz G.M., Peterson C.J. & Kronstad W.E.(2003) Dehydrin expression and drought tolerance in sevenwheat cultivars. Crop Science 43, 577–582.

Ludlow M.M. & Muchow R.C. (1990) A critical evaluation of traitsfor improving crop yields in water-limited environments.Advances in Agronomy 43, 107–153.

Morrell P.L., Lundy K.E. & Clegg M.T. (2003) Distinct geographicpatterns of genetic diversity is maintained in wild barley (Hor-deum vulgare ssp. spontaneum) despite migration. Proceedingsof the National of Academy of Sciences of the USA 100, 10812–10817.

Nagy E. (1997) Selection for native characters in hybrids betweentwo locally adapted plant subspecies. Evolution 51, 1469–1480.

Natali L., Giordani T. & Cavallini A. (2003) Sequence variabilityof a dehydrin gene within Helianthus annuus. Theoretical andApplied Genetics 106, 811–818.

Nemoto Y. & Sasakuma T. (2002) Differential stress responses ofearly salt-stress responding genes in common wheat. Phy-tochemistry 61, 129–133.

Nevo E. (1992) Origin, evolution, population genetics andresources for breeding of wild barley, Hordeum spontaneum, inthe Fertile Crescent. In Barley: Genetics, Biochemistry, Molecu-lar Biology and Biotechnology (ed. P.R. Shewry), pp. 19–43.CAB International, Wallingford, Oxon, UK.

Nevo E. & Beiles A. (1989) Genetic diversity of wild emmer wheatin Israel and Turkey. Theoretical and Applied Genetics 77, 421–455.

Nevo E., Brown A.H.D., Zohary D., Storch N. & Beiles A. (1981)Microgeographic edaphic differentiation of allozyme polymor-phism of wild barley (Hordeum spontaneum, Poaceae). PlantSystematic and Evolution 138, 287–292.

Nevo E., Korol A.B., Beiles A. & Fahima T. (2002) Evolution ofWild Emmer and Wheat Improvement. Springer Verlag,Heidelberg.

Nevo E., Zohary D., Brown A.H.D. & Haber M. (1979) Geneticdiversity and environmental associations of wild barley, Hor-deum spontaneum, in Israel. Evolution 33, 815–833.

Owuor E.D., Beharav A., Fahima T., Kirzhner V.M., Korol A. &Nevo E. (2003) Microscale ecological stress causes RAPDmolecular selection in wild barley, Neve Yaar microsite, Israel.Genetic Resources and Crop Evolution 50, 213–224.

Owuor E.D., Fahima T., Beharav A., Korol A.B. & Nevo E. (1999)RAPD divergence caused by microsite edaphic selection in wildbarley. Genetica 105, 177–192.

Ozturk Z.N., Talame V., Deyholos M., Michalowski C.C., Gal-

Differential expression of dehydrin genes in wild barley 1307

© 2004 Blackwell Publishing Ltd, Plant, Cell and Environment, 27, 1297–1308

braith D.W., Gozukirmizi N., Tuberosa R. & Bohnert H. (2002)Monitoring large-scale changes in transcript abundance indrought- and salt-stressed barley. Plant Molecular Biology 48,551–573.

Pakniyat H., Powell W., Baird E., Handley L.L., Robinson D.,Sorimgeour C.M., Nevo E., Hackett C.A., Caligari P.D.S. &Forster B.P. (1997) AFLP variation in wild barley (Hordeumspontaneum C. Koch) with reference to salt tolerance and asso-ciated ecogeography. Genome 40, 332–341.

Pan A., Hayes P.M., Chen F., Chen T.H.H., Blake T., Wright S.,Karsai I. & Bedo Z. (1994) Genetic analysis of the componentsof winter hardiness in barley (Hordeum vulgare L.). Theoreticaland Applied Genetics 89, 900–910.

Pelah D., Wang W., Altman A., Shoseyov O. & Bartles D. (1997)Differential accumulation of water stress-related proteins,sucrose synthase and soluble sugars in Populus species that dif-fer in their water stress response. Physiologia Plantarum 99,153–159.

Pérez de la Vega M. (1996) Plant genetic adaptedness to climaticand edaphic environment. Euphytica 92, 27–38.

Prentice I.C., Cramer W., Harrison S.P., Leemans R., MonserudR.A. & Solomon A.M. (1992) A global biome model based onplant physiology and dominance, soil properties, and climate.Journal of BioGeography 19, 117–134.

Ristic Z. & Jenks M.A. (2002) Leaf cuticle and water loss in maizelines differing in dehydration avoidance. Journal of Plant Phys-iology 159, 645–651.

Robinson D., Handley L.L., Scrimgeour C.M., Gordon D.C., For-ster B.P. & Ellis R.P. (2000) Using stable isotope natural abun-dances (delta 15N and delta 13C) to integrate the stressresponses of wild barley (Hordeum spontaneum C. Koch.) gen-otypes. Journal of Experimental Botany 51 (342), 41–50.

Seki M., Narusaka M., Ishida J., et al. (2002) Monitoring theexpression profiles of 7000 Arabidopsis genes under drought,cold and high-salinity stresses using a full-length cDNA microar-ray. Plant Journal 31, 279–292.

Shinozaki K. & Yamaguchi-Shinozaki K. (1997) Gene expressionand signal transduction in water-stress response. Plant Physiol-ogy 115, 327–334.

Smith J.A.C. & Griffiths H. (1993) Water Deficits: Plant Responsesfrom Cell to Community. Bios Scientific Publishers, Oxford,UK, pp. 1–332.

Statsoft Inc. (1996) STATISTICA for Windows (Computer Pro-gram Manual). Statsoft. Inc. Tulsa, OK, USA.

Stephenson N.L. (1990) Climate control of vegetation distribu-tion: the role of the water balance. American Naturalist 135,649–670.

Svensson A.S., Johansson F.I., Moller I.M. & Rasmusson A.G.(2002) Cold stress decreases the capacity for respiratory NADHoxidation in potato leaves. FEBS Letters 517, 79–82.

Teulat B., Zoumarou-Wallis N., Rotter B., Ben Salem M., BahriH. & This D. (2003) QTL for relative water content in field-grown barley and their stability across Mediterranean environ-ments. Theoretical and Applied Genetics 108, 181–188.

Turpeinen T., Tenhola T., Manninen O., Nevo E. & Nissila E.(2001) Microsatellite diversity associated with ecological factorsin Hordeum spontaneum populations in Israel. Molecular Ecol-ogy 10, 577–591.

Turpeinen T., Vanhala T., Nevo E. & Nissila E. (2003) AFLPgenetic polymorphism in wild barley (Hordeum spontaneum)populations in Israel. Theoretical and Applied Genetics 106,1333–1339.

Zee V.K., Qiang Chen F., Hayes P.M., Close T.J. & Chen T.H.H.(1995) Cold-specific induction of a dehydrin gene-family mem-ber in barley. Plant Physiology 108, 1233–1239.

Zhu B., Choi D.W., Fenton R. & Close T.J. (2000) Expression ofthe barley dehydrin multigene family and the development offreezing tolerance. Molecular and General Genetics 246, 145–153.

Received 16 March 2004; received in revised form 5 May 2004;accepted for publication 14 June 2004

APPENDIX 1

Water loss rate (WLR) under drought stress of 390 wild barley genotypes representing Jordanian and Israeli populations.The barley seedlings (after 10 d of growth) were used for measurement of WLR (g h-1 g-1 DW) according to Clarke & McCaig(1982).

Population WLR Population WLR Population WLR Population WLR Population WLR

IP1-6 0.1166 IP11-34 0.1417 IP22-16 0.1455 JP2-2 0.1353 JP6-10 0.1733IP1-8 0.157 IP11-37 0.1203 IP22-30 0.1625 JP2-4 0.1507 JP6-15 0.097IP1-10 0.1754 IP11-49 0.1156 IP22-35 0.1782 JP2-5 0.1158 JP6-18 0.1745IP1-12 0.1845 IP11-52 0.155 IP22-42 0.1909 JP2-6 0.1628 JP6-20 0.1517IP1-13 0.1577 IP14-1 0.1621 IP22-43 0.2022 JP2-7 0.1284 JP6-23 0.1573IP1-15 0.1242 IP14-6 0.187 IP22-55 0.1362 JP2-8 0.172 JP6-24 0.1548IP1-24 0.119 IP14-7 0.2055 IP23-42 0.2152 JP2-9 0.1492 JP6-25 0.1606IP1-25 0.1255 IP14-10 0.1437 IP23-64 0.1669 JP2-10 0.1458 JP6-28 0.154IP1-33 0.1614 IP14-14 0.2019 IP23-69 0.1102 JP2-11 0.14 JP6-32 0.1844IP1-39 0.148 IP14-15 0.1148 IP33-26 0.1616 JP2-12 0.1414 JP6-43 0.1539IP9-1 0.1156 IP14-20 0.1982 IP33-28 0.1349 JP2-13 0.1464 JP6-46 0.1081IP9-9 0.1773 IP14-24 0.2095 IP33-29 0.1517 JP2-15 0.1316 JP6-49 0.1108IP9-27 0.1504 IP14-30 0.2099 IP33-30 0.1245 JP2-18 0.1265 JP6-50 0.1006IP9-33 0.1408 IP14-36 0.1935 IP33-31 0.2033 JP2-21 0.143 JP7-1 0.1211IP9-35 0.1365 IP18-1 0.1704 IP33-37 0.1305 JP2-22 0.146 JP7-2 0.1066IP9-37 0.1274 IP18-4 0.1578 IP33-41 0.1783 JP2-23 0.1137 JP7-6 0.1298IP9-39 0.1359 IP18-5 0.1708 IP33-54 0.1431 JP2-25 0.1135 JP7-12 0.1346IP9-42 0.1478 IP18-8 0.1142 JP1-1 0.1408 JP2-27 0.1384 JP7-13 0.125IP9-43 0.139 IP18-14 0.1704 JP1-3 0.159 JP2-29 0.106 JP7-14 0.1234IP9-52 0.1139 IP18-15 0.1493 JP1-4 0.1961 JP5-3 0.0973 JP7-15 0.1095

1308 T. Suprunova et al.

© 2004 Blackwell Publishing Ltd, Plant, Cell and Environment, 27, 1297–1308

IP10-1 0.1294 IP18-17 0.116 JP1-5 0.1864 JP5-5 0.1623 JP7-20 0.111IP10-4 0.1094 IP18-29 0.1637 JP1-7 0.1789 JP5-6 0.1346 JP7-21 0.1067IP10-6 0.1384 IP18-32 0.2179 JP1-8 0.144 JP5-9 0.1013 JP7-23 0.1109IP10-15 0.1306 IP18-36 0.204 JP1-9 0.1277 JP5-10 0.1038 JP7-24 0.2008IP10-18 0.1496 IP20-2 0.2071 JP1-10 0.1435 JP5-17 0.1141 JP7-25 0.097IP10-24 0.1465 IP20-4 0.2028 JP1-12 0.133 JP5-20 0.1125 JP7-26 0.1176IP10-25 0.1322 IP20-28 0.1834 JP1-13 0.1658 JP5-25 0.1047 JP7-27 0.0984IP10-27 0.1636 IP20-30 0.2056 JP1-14 0.1338 JP5-30 0.1275 JP7-28 0.1366IP10-34 0.1682 IP20-33 0.2037 JP1-15 0.1283 JP5-31 0.1221 JP7-32 0.1058IP10-50 0.1098 IP20-36 0.203 JP1-16 0.141 JP5-32 0.1123 JP7-49 0.1259IP11-9 0.1151 IP20-37 0.1218 JP1-17 0.1578 JP5-33 0.1183 JP8-1 0.1428IP11-10 0.1154 IP20-40 0.1634 JP1-18 0.1512 JP5-50 0.1188 JP8-2 0.1143IP11-12 0.1104 IP20-45 0.1976 JP1-19 0.1485 JP6-1 0.1362 JP8-3 0.1376IP11-13 0.1178 IP20-48 0.1418 JP1-20 0.1455 JP6-2 0.17 JP8-4 0.1532IP11-16 0.134 IP22-14 0.2083 JP1-28 0.1182 JP6-4 0.1454 JP8-6 0.134IP11-20 0.1398 IP22-15 0.151 JP2-1 0.1355 JP6-7 0.1551 JP8-7 0.1128JP8-8 0.1104 JP11-18 0.1886 JP15-9 0.1069 JP17-5 0.1667 JP20-4 0.1971JP8-9 0.2032 JP11-19 0.1602 JP15-12 0.1995 JP17-8 0.1521 JP20-5 0.1692JP8-10 0.2151 JP11-21 0.1402 JP15-13 0.1337 JP17-10 0.1944 JP20-6 0.1648JP8-11 0.1871 JP11-22 0.1775 JP15-17 0.1655 JP17-11 0.1688 JP20-7 0.1527JP8-12 0.1166 JP11-23 0.1535 JP15-18 0.2094 JP17-12 0.1524 JP20-8 0.1992JP8-13 0.1454 JP11-26 0.1518 JP15-20 0.1792 JP17-13 0.1643 JP20-9 0.1518JP8-19 0.1075 JP11-27 0.1554 JP15-22 0.2063 JP17-15 0.2079 JP20-11 0.1989JP8-21 0.1082 JP11-28 0.1774 JP15-23 0.1579 JP17-17 0.179 JP20-12 0.2003JP8-22 0.1217 JP11-30 0.1363 JP15-24 0.1987 JP17-18 0.1831 JP20-13 0.1783JP8-23 0.1033 JP11-31 0.1854 JP15-28 0.139 JP17-19 0.2014 JP20-16 0.1556JP8-25 0.1539 JP11-32 0.1901 JP15-31 0.1563 JP17-20 0.1969 JP20-17 0.1492JP8-28 0.1421 JP11-39 0.1458 JP15-33 0.1624 JP17-21 0.1204 JP20-18 0.1612JP8-31 0.1013 JP11-44 0.1197 JP15-44 0.1932 JP17-24 0.175 JP20-26 0.1247JP9-5 0.0995 JP11-45 0.1358 JP15-48 0.1977 JP17-25 0.1542 JP20-29 0.1616JP9-6 0.199 JP13-1 0.1501 JP15-49 0.2071 JP17-26 0.1233 JP20-38 0.1747JP9-7 0.1441 JP13-6 0.1295 JP16-1 0.146 JP17-33 0.1663 JP20-41 0.2001JP9-10 0.1489 JP13-11 0.1503 JP16-2 0.1435 JP18-1 0.1431 JP20-45 0.1356JP9-11 0.1394 JP13-12 0.1669 JP16-3 0.1674 JP18-3 0.1322 JP20-47 0.1931JP9-14 0.1636 JP13-13 0.152 JP16-5 0.134 JP18-5 0.1405 JP23-1 0.1612JP9-15 0.1286 JP13-15 0.1669 JP16-6 0.1089 JP18-6 0.1413 JP23-2 0.2076JP9-16 0.1163 JP13-16 0.1256 JP16-7 0.1401 JP18-8 0.1499 JP23-6 0.1755JP9-18 0.2046 JP13-18 0.1548 JP16-13 0.1749 JP18-9 0.128 JP23-8 0.2072JP9-20 0.1684 JP13-20 0.1428 JP16-16 0.154 JP18-10 0.11 JP23-10 0.1703JP9-21 0.1409 JP13-21 0.1436 JP16-20 0.1343 JP18-11 0.1983 JP23-11 0.1872JP9-23 0.1651 JP13-22 0.1785 JP16-21 0.1907 JP18-12 0.1119 JP23-12 0.1833JP9-25 0.1289 JP13-23 0.1214 JP16-24 0.1507 JP18-19 0.1804 JP23-13 0.1271JP9-26 0.1239 JP13-24 0.1306 JP16-25 0.1837 JP18-23 0.1353 JP23-14 0.1277JP9-28 0.1175 JP13-25 0.1369 JP16-26 0.136 JP18-24 0.143 JP23-15 0.1699JP9-29 0.1159 JP13-27 0.1403 JP16-27 0.199 JP18-25 0.1479 JP23-16 0.1674JP9-30 0.1526 JP13-28 0.1767 JP16-29 0.1495 JP18-26 0.1263 JP23-17 0.1567JP9-31 0.1263 JP13-29 0.1308 JP16-30 0.1499 JP18-27 0.1219 JP23-18 0.0969JP9-33 0.1263 JP13-31 0.1753 JP16-31 0.1464 JP18-29 0.1821 JP23-19 0.096JP11-2 0.1222 JP13-32 0.1385 JP16-32 0.1898 JP18-30 0.1976 JP23-21 0.1878JP11-4 0.1675 JP13-33 0.1393 JP16-33 0.1445 JP18-32 0.1738 JP23-23 0.1449JP11-11 0.1833 JP15-1 0.1302 JP16-47 0.1871 JP18-33 0.2045 JP23-27 0.1611JP11-13 0.1663 JP15-5 0.1202 JP17-1 0.205 JP18-50 0.1935 JP23-28 0.1959JP11-16 0.1769 JP15-6 0.1631 JP17-3 0.1342 JP20-1 0.1522 JP25-1 0.1892JP11-17 0.1644 JP15-7 0.1469 JP17-4 0.1594 JP20-2 0.1507 JP25-2 0.1326JP25-4 0.1777 JP25-12 0.1133 JP25-16 0.1321 JP25-22 0.1522 JP25-29 0.1222JP25-5 0.2026 JP25-13 0.1209 JP25-17 0.1476 JP25-24 0.1106 cv Noga 0.1789JP25-6 0.1554 JP25-14 0.1303 JP25-18 0.1894 JP25-26 0.1289 cv Morex 0.1112JP25-8 0.127 JP25-15 0.1896 JP25-21 0.1575 JP25-27 0.1197 cv Mona 0.1118

Population WLR Population WLR Population WLR Population WLR Population WLR

APPENDIX 1 Cont.

Population collection sites: Israeli Populations (IP-1, 9, 10, 11, 14, 18, 20, 22, 23, 33) are described by Nevo et al. (1979); Jordanian Populations (JP - 1, 2, 5, 6, 7, 8, 9, 11, 13, 15, 16, 17, 18, 20, 23, 25) are described by Baek et al. (2003).