Identification of adaptation-specific differences in mRNA expression of sessile and pedunculate oak...

Summary Quercus petraea (Matt.) Liebl. and Q. robur L.hybridize frequently and occupy similar, though distinct, eco-logical niches. So far, genetic discrimination between thesespecies at the molecular level has been based mainly on neutralmarkers. Because such markers often exhibit low species dif-ferentiation because of high genetic compatibility and ex-change between Q. robur and Q. petraea at these loci, we usedadaptation-related expressed genes as markers. Accordingly,we identified osmotic-stress-induced genes in a Q. petraea cellline grown under moderate osmotic stress conditions. Two sub-traction libraries were established from callus cells culturedunder hyperosmotic stress for 1 or 48 h. Thirty-three differen-tially expressed sequence tags (ESTs) (from 70 originally iso-lated) were classified according to their putative functions. Atleast five of these gene products may contribute to osmotic-stress tolerance in oak: betaine aldehyde dehydrogenase, twotrans-acting transcription factors (one abscsic acid (ABA)-responsive, the other ABA-independent), a glutathione-S-transferase and a heat-shock cognate protein. Seven geneswere selected based on their putative function and their expres-sion monitored in vivo. Leaf tissue from Q. petraea and Q. ro-bur plantlets grown hydroponically under hyperosmotic condi-tions was harvested after 0, 1, 6, 24 or 72 h and analyzed byreal-time polymerase chain reaction (PCR). We found indica-tions of osmotic stress adaptation in Q. petraea based onup-regulation of genes related to protective functions, whereasdown-regulation of these genes was evident in Q. robur. Thus,genetic markers related to adaptive traits may be useful for dif-ferentiating Q. petraea and Q. robur genotypes.

Keywords: EST, gene expression, hyperosmotic stress, Quer-cus robur, Quercus petraea, subtractive hybridization.

Introduction

The sympatric white oaks Quercus petraea (Matt.) Liebl. (ses-sile oak) and Q. robur L. (pedunculate oak) represent discretegenetic entities, despite their ability to hybridize (Muir et al.2000). In distinct geographical areas, they even share the ma-

ternally inherited chloroplast genome (Dumolin-Lapègue etal. 1999). Thus sessile oak and pedunculate oak are geneticallyclosely related, and most molecular markers identified so farshow little or no species differentiation (Petit et al. 1993, Sam-uel et al. 1995, Mariette et al. 2002), although a few molecularmarkers exhibit high species differentiation (Bodénès et al.1997, Gömöry 2000).

The preference of species within this genus for distinctmicroclimatic conditions implies a physiological adaptiveforce. Sessile oak grows predominantly on relatively dry soilson lower altitude slopes and ridges, whereas pedunculate oakcan populate heavy, temporarily waterlogged lowland soils(Aas 1998). Several types of physiological function involvinggenes responsible for osmotic regulation of the cell could beresponsible for the contrasting adaptations of sessile oak andpedunculate oak to dry and wet environments, respectively.Osmotic adjustment is an important acclimation response ofplants to environmental change and to water deficit in particu-lar. The inheritance of capacity for osmotic adjustment islikely to be complex and multigenic.

Numerous osmotic-stress-related genes from plants havebeen characterized by: (1) large-scale analyses of differentialexpression patterns (Bray 2002, Seki et al. 2002, Rabbani et al.2003); (2) analysis of a global expression profile monitoredunder salt-stress conditions using abundance profiles in sev-eral species (Bohnert et al. 2001); and (3) small-scale analysisof differentially expressed genes obtained by subtractionmethods (Pih et al. 1997). Functions of osmotic-stress-relatedgenes include osmoprotection, water and ion transport, free-radical scavenging, chaperone activity, signal transduction,transcription control, protein synthesis and energy metabo-lism. Gene products that are prominently expressed duringwater-deficit stress include enzymes of sugar metabolism(Garcia et al. 1997), quaternary amine biosynthesis (Ishitani etal. 1995) and amino acid metabolism (Delauney and Verma1993, Liu and Zhu 1997, Sofo et al. 2004). These proteins areinvolved in redox metabolism (Roxas et al. 1997), fulfill sev-eral protective functions and act as chaperones (Ried andWalker-Simmons 1993, Moons et al. 1995, Xu et al. 1996,

Tree Physiology 25, 1317–1329© 2005 Heron Publishing—Victoria, Canada

Identification of adaptation-specific differences in mRNA expression ofsessile and pedunculate oak based on osmotic-stress-induced genes

ILGA PORTH,1,2 MARGIT KOCH,1,2 MARIA BERENYI,1 AGNES BURG1 and KORNELBURG1,3

1 Austrian Research Centers Seibersdorf, Division of Biogenetics and Natural Resources, A-2444 Seibersdorf, Austria2 These authors contributed equally to the study3 Corresponding author ([email protected])

Received August 18, 2004; accepted January 28, 2005; published online August 1, 2005

Borovskii et al. 2002, Sun et al. 2002). They also protect thecell from oxidative and chemical damage (Churin et al. 1999),take part in vacuolar biogenesis and function (Blumwald2000), signal transduction and control of gene expression(Burssens et al. 2000, Mikolajczyk et al. 2000, Xiong et al.2002, Shen et al. 2003), and proteolysis (Koizumi et al. 1993).Additionally, several transcripts with similarity to proteins ofunknown functions (Hurkman and Tanaka 1996) or to virtualproteins (Bohnert et al. 2001, Seki et al. 2001, Bray 2002) havebeen detected.

Attempts to delineate a signal transduction pathway for saltor drought stress in plants have failed, however, because toofew components have been identified (Zhu 2002). In this re-spect, much less is known about gene expression in trees thanin Arabidopsis (Xiong et al. 1999, Seki et al. 2001, 2002,Kreps et al. 2002, Yamaguchi-Shinozaki et al. 2002) or cropcultivars (Kawasaki et al. 2001, Patnaik and Khurana 2001,Sharma and Lavanya 2002, Cooper et al. 2003, Rabbani et al.2003). Apart from a recent survey to identify differentially ex-pressed genes in water-stressed maritime pine (Pinus pinasterAit.) (Dubos et al. 2003), there is little information about thegenes involved in the tolerance of tree species to unfavorableenvironmental conditions such as drought.

To gain an understanding of the difference in the molecularlevel responses of Q. petraea and Q. robur to changes in os-motic environment, we used gene expression profiling basedon adaptive traits. We identified osmotic-stress-induced genesfrom Q. petraea callus cells cultured under moderate stressconditions by subtractive hybridization. We then analyzed theexpression profiles of a selected set of genes in response to os-motic stress in hydroponically grown plantlets of Q. robur andQ. petraea by real-time quantitative polymerase chain reaction(qPCR). To reveal potentially different responses of the twospecies to the imposed osmotic stress and thus find a basis forspecies differentiation, we compared the expression of thegenes at each sampling time between species, and also relatedthe expression values to those in control plants.

Materials and methods

Oak cell culture

Cell suspension cultures originating from Q. petraea callus(Endemann and Wilhelm 1999) were grown in Erlenmeyerflasks with continuous shaking in an environmentally con-trolled chamber set to provide a 16-h photoperiod with 65–70 µmol m– 2 s –1 of illumination and a constant temperature of24 °C. Control cell cultures (K) were grown in P24 mediumcontaining the concentration of salts described by Teasdale(1992) (hereafter P24-S medium), whereas hyperosmoticallytreated cells were grown in P24 medium containing four timesthe concentration of salts (hereafter P24-4S medium) for 1 h(1T) or for 2 days (2T).

Isolation of polyA+ RNA and cDNA library construction

Total RNA was extracted from plant tissue by the guanidiumthiocyanate/lithium chloride method (Gao et al. 1995). Poly

A+ RNA was purified with Oligo-dT-labeled magnetic Dyna-beads (Dynal No. 610.02, Dynal Biotech, Oslo, Norway). ThecDNA libraries were constructed from the 1T and the 2T cellcultures using the SMART PCR cDNA Library ConstructionKit (Clontech Laboratories, Palo Alto, CA) according to themanufacturer’s instructions. After cDNA size fractionation ona CHROMA SPIN-400 column (BD Biosciences Clontech,Mountain View, CA) to exclude molecules < 500 bp, the dou-ble-stranded cDNA was ligated into the λTriplEx2 vector. Li-gations were subsequently packaged into λ bacteriophageusing the Boehringer Mannheim DNA Packaging Kit (Boeh-ringer Mannheim, Mannheim, Germany). The titers of thenon-amplified libraries were 8.9 × 106 pfu ml – 1 (1T) and 8.4 ×106 pfu ml – 1 (2T), respectively.

Construction of subtraction libraries

Subtractive hybridization was carried out with the PCR-SelectcDNA Subtraction Kit (Clontech) according to the protocolprovided. Two kinds of subtractions were performed: (1) for-ward subtraction with the two tester samples (1T cDNA and2T cDNA) and the control (K cDNA) as the driver sample; and(2) reverse subtraction with the control (K cDNA) as tester andthe two induced samples (1T cDNA and 2T cDNA) as driversamples. Subtractive libraries (SLs) were constructed with1TS and 2TS amplification products using the Advantage PCRCloning Kit (Clontech). Positive clones were identified byPCR reactions using nested primers provided with the Sub-traction kit. The clones containing inserts were stored onmasterplates representing the subtractive libraries SL1T andSL2T, respectively.

Identification of differentially expressed ESTs

The differential screening procedure was carried out accord-ing to instructions provided with the PCR-Select DifferentialScreening Kit (Clontech). Complementer DNA dot blots rep-resenting the corresponding subtracted libraries were preparedby dotting 5 µl of the PCR product from each individual for-ward-subtracted clone onto Hybond-N nylon membranes.Four identical nylon filters were prepared from each sub-tracted library (see Figure 1). Products from secondary PCR ofthe forward- and reverse-subtracted and of the non-subtractedcDNAs were labeled with (32P) dCTP (Prime-It II RandomPrimer Kit, Stratagene, La Jolla, CA) and used as hybridiza-tion probes. Clones showing differential hybridization weresequenced by the cycle sequencing reaction (BigDye Termina-tor Cycle Sequencing Ready Reaction Kit, Applied Bio-systems, Foster City, CA).

Library screening for full-length cDNAs

The full-length mRNAs of the 1T9 and 1T62 clones were ob-tained by isotope screening, as described by Sambrook et al.(1989). The remaining SL clones were extended at the 5′ and3′ ends to full length by PCR amplification of total DNA ex-tracts of the cDNA libraries using outward primers nested inthe SL fragment and the TriplEx sequencing primers. Avail-able sequences of the corresponding ESTs were used to designprimers with the OLIGO primer analysis software Version 5.0

1318 PORTH, KOCH, BERENYI, BURG AND BURG

TREE PHYSIOLOGY VOLUME 25, 2005

(data not shown). We used the Advantage cDNA PolymeraseKit (Clontech) and applied the recommended cycling parame-ters, except for the specific annealing temperatures and a finalextension step of 5 min at 72 °C, for long-distance PCR. ThosePCR products that showed single bands on an agarose gel(2–2.5%) stained with ethidium bromide were subsequentlypurified with the QIAGEN gel purification kit (QIAGEN,Hilden, Germany) and sequenced directly. The PCR productsshowing multiple banding patterns on the gel were subjectedto T/A cloning with Invitrogen’s TOPO cloning kit (Invitro-gen, Groningen, Netherlands) before sequencing. Sequencingreactions as well as sequence analysis were performed as pre-viously described. The resulting 5′ and 3′ end sequence frag-ments of a specific cDNA clone were added contiguously tothe corresponding EST sequence with the DNASIS 2.5 contigmanager application (Hitachi Software Engineering, Tokyo,Japan, 1994–1997). The putative full- length sequences of thecDNA clones were annotated with BLASTX available at theNCBI server.

Growth of plants and sample collection

The experimental material consisted of seedlings grown fromacorns (eight mother trees) originating from a natural mixedQ. robur–Q. petraea stand (Sigmundsherberg, Austria) andplantlets originating from somatic embryo culture (Q. roburgenotype P28H9; Prewein and Wilhelm 2003). Germinationof acorns was forced by incubating them in tap water heated to41 °C for 3 to 4 h before sowing on washed quartz sand. Ger-mination trays were covered with plastic foil with access tofresh air and watered with tap water. Quercus robur plantletswere grown under the same greenhouse conditions as the seed-lings. Fourteen-week-old seedlings were transferred to a hy-droponic system that consisted of closed Curver/Rubbermaidboxes equipped with an automatic air supply. The box lidswere perforated to allow insertion of the plantlets (each plantwas fixed in position with a stopper made of cling film). Thetwo species were equally represented in each sample set. Theplantlets were first allowed to equilibrate for 4 weeks in P24-Smedium. This medium (osmotic potential, Ψπ = –0.077 MPa)consisted only of the macro and micro elements and iron com-pounds in Milli-Q water (Millipore, Billerica, MA), adjustedto pH 5.9, and subsequently autoclaved. Plants were grown ina controlled environment chamber set at a constant tempera-ture of 22 °C and a 14-h photoperiod with quantum flux ofabout 50 µmol m– 2 s –1. The P24-S medium was renewedweekly during the acclimation period.

After acclimation, the plants were subjected to osmoticstress for 0, 1, 6, 24 or 72 h by replacing the P24-S mediumwith P24-4S medium (Ψπ = –0.283 MPa) at 1100 h on Day 1.Sample collection was started at 1000 h on Day 1, 1 h beforethe start of the osmotic stress treatment. Additional sampleswere taken at 1200 h (treatment time 1 h), 1700 h (treatmenttime 6 h), 1100 h the next day (treatment time 24 h), and Day 4at 1100 h (sampling time 72 h). Leaf samples were collectedwith a sterile scalpel and immediately submerged in RNAlatersolution (Ambion, Austin, TX), incubated overnight at 4 °Cand subsequently frozen at –20 °C.

Isolation of RNA and relative quantification of target genes

Leaves were ground in liquid nitrogen with a mortar and pestleand total RNA was isolated from the ground material with theRNeasy Plant Mini Kit (QIAGEN). Contaminating DNA wasremoved enzymatically with DNA-free (Ambion, Austin,TX). The amount of RNA was determined spectrophotometri-cally (260/280 nm), and its purity and integrity checked on anRNase-free 1% agarose gel. A two-step qPCR was performed.In the first step, the required template (first strand cDNA) wassynthesized by reverse transcription (RT) using about 1 µg to-tal RNA, SuperScript II RTase and oligo (dT) primers (Invitro-gen). The RT reaction mix was diluted 10× to reduce the inhib-itory effects of the RT buffer system. For qPCR, we appliedSybrGreen I Dye, an intercalatory fluorescent dye (includingthe passive reference ROX), and a PE 5700 fluorescence-based detection system (Applied Biosystems). In a 25-µl reac-tion volume, we used 10.5 µl of primer-pair mix (primer con-centration: 300/300 nmol), 12.5 µl of dye and 2 µl of cDNA.Reactions were performed according to the manufacturer’sprotocol, except that sample reactions were set up in triplicate.

Seven transcripts osmotically induced in oak callus cells(Table 1) were selected to monitor their expression profiles.Four genes (hsp80, GST, BADH and DPBF) were selected ac-cording to their known functions in enhancing stress tolerance(Saneoka et al. 1995, Holmstrom et al. 2000, Roxas et al. 2000,Uno et al. 2000, Sun et al. 2002). To monitor putative changesin energy metabolism, the thiazole biosynthetic enzyme wasselected. We selected lipid transfer protein (LTP) as a repre-sentative of gene products that are frequently induced by os-motic stress (Bohnert et al. 2001, Seki et al. 2001), and wechose a putative member of the oxygenase gene family 2-oxo-glutarate (2-OG) because this gene family is thought to havea wide range of functions. The β-tubulin gene of oak(Est172ma), based on a partial coding sequence (GenBankAccession No. CF369288), served as an internal standard tonormalize differences in template amounts (van Wees et al.2000, Zhang and Gurr 2001). Target sequences between 100and 120 bp were designed using Primer Express software forprimer design following the manufacturer’s instructions (Ta-ble 2). Quantification analysis was carried out according toPfaffl (2002) and the User Bulletin No. 2 ABI Prism 7700 Se-quence Detection System (relative quantification of gene ex-pression) using the threshold cycle method (Ct) and a fluores-cence level of 10 –1.

In total, 27 plantlets (representing the genotypes of maternalhalf sibs from the eight mother trees and the genotype of one invitro line) were analyzed by qPCR to monitor gene expressionin Q. petraea and Q. robur. For the control and Day 1 samples,40% Q. petraea and 60% Q. robur phenotypes were available.For investigating plantlets after 1 h, 6 h and 3 days of osmoticstress treatment, the numerical proportion of phenotypes was1:1. Relative expression ratios for each species were deter-mined by relating Ct values derived from stressed plantlets ofthis species to the respective physiological reference of thesame species (see below). Additionally, the expression ofQ. petraea (sample) was compared with that of Q. robur (con-trol) at the five sampling times.

TREE PHYSIOLOGY ONLINE at http://heronpublishing.com

ADAPTATION OF mRNA EXPRESSION IN OAK TO OSMOTIC STRESS 1319



Tabl

e1.

Thi

rty-

thre

eos

mor

egul

ated

tran

scri

pts

from

Que

rcus

petr

aea

callu

sce

llscu

lture

dfo

r1h

(1T

)or2

days

(2T

)in

hype

rosm

otic

P24-

4Sm

ediu

man

dis

olat

edby

subt

ract

ive

hybr

idiz

atio

nan

ddi

f-fe

rent

ials

cree

ning

.Abb

revi

atio

ns:E

ST=

expr

esse

dse

quen

ceta

g;B

S=

bios

ynth

esis

;CH

=ch

aper

one;

CK

=cy

toki

nesi

s;C

O=

cellu

laro

rgan

izat

ion;

EM

=en

ergy

met

abol

ism

;OS

=ox

idat

ive

stre

ss;

PD=

prot

ein

dest

inat

ion;

PS=

prot

ein

synt

hesi

s;R

F=

regu

lato

ryfu

nctio

n;T

F=

tran

spor

tfac

ilita

tion;

TS

=tr

ansc

ript

ion;

UC

=un

clas

sifi

ed;a

ndU

K=

unkn

own.

Osm

otic

EST

Clo

neFu

nctio

nal

Not

eG

enB

ank

BL

AST

XSp

ecie

sG

enB

ank

Scor

eE

-tr

eatm

ent

cate

gory

3ca

tego

ryen

try

hom

olog

yre

fere

nce

valu

e(n

ucel

otid

e)

1h

1T7

1C

O1

CF3

6927

6Fi

ddle

head

-lik

epr

otei

nG

ossy

pium

hirs

utum

AY

0728

2343

40.

01T

81

PDC

N72

5666

Ubi

quiti

n/r

ibos

omal

prot

ein

s27a

Sola

num

tube

rosu

mB

F153

684

111

7e-2

41T

92

CO

1,2

CF3

6929

0L

ipid

tran

sfer

prot

ein

prec

urso

rG

.hir

sutu

mA

F195

864

163

3e-3

91T

113

UK

1C

F369

263

Unk

now

npr

otei

nA

rabi

dops

isth

alia

naN

M_1

2356

331

4e-

119

1T12

1B

S1

CF3

6926

4O

xido

redu

ctas

e,2O

G-F

e(II

)A

.tha

lian

aN

M_1

1278

641

3e-

114

oxyg

enas

efa

mily

1T15

1R

F1

CF3

6926

5T

GF

β-In

duci

ble

nucl

ear

prot

ein

Ory

zasa

tiva

AP0

0430

037

6e-

103

1T21

1PS

1C

F369

266

60S

Rib

osom

alpr

otei

nL

10Vi

tis

ripa

ria

AF1

8075

835

6e-

110

1T24

2U

K1

CF3

6926

7U

nkno

wn

prot

ein

A.t

hali

ana

AB

0122

4364

5e-0

91T

251

UK

1C

F369

268

Unk

now

npr

otei

nA

.tha

lian

aA

B02

6651

178

2e-4

41T

262

PS1

CF3

6927

040

SR

ibos

omal

prot

ein

S20

A.t

hali

ana

NM

_125

624

200

2e-5

01T

311

TS

CN

7256

67D

ehyd

ratio

n-re

spon

sive

elem

ent

Lyco

pers

icon

escu

lent

umA

F506

825

506e

-05

bind

ing

prot

ein

31T

343

EM

CN

7256

73T

rios

epho

spha

teis

omer

ase

A.t

hali

ana

NM

_127

687

375

e-10

31T

411

EM

1C

F369

272

Pyru

vate

kina

seA

.tha

lian

aN

M_1

1515

945

70.

01T

571

TS

1,2

CF3

6927

3D

c3Pr

omot

er-b

indi

ngfa

ctor

-2H

elia

nthu

san

nuus

AF0

0145

492

2e-2

51T

621

BS

1,2

CF3

6927

4B

etai

neal

dehy

dede

hydr

ogen

ase

Am

aran

thus

AF0

1715

066

90.

0hy

poch

ondr

iacu

s1T

631

RF

CN

7256

68E

ukar

yotic

tran

slat

ion

initi

atio

nA

.tha

lian

aN

M_1

0334

023

01e

-59

fact

or5,

puta

tive/

ef-5

1T69

1U

K1

CF3

6927

5U

nkno

wn

prot

ein

A.t

hali

ana

NM

_123

621

184

2e-4

51T

721

PS1

CF3

6927

740

SR

ibos

omal

prot

ein

S4O

.sat

iva

Y15

009

244

1e-8

6

2da

ys2T

21

EM

CN

7256

69K

etol

-aci

dre

duct

oiso

mer

ase

A.t

hali

ana

NM

_115

723

247

3e-6

42T

31

UC

1C

F369

283

Puta

tive

mtn

19O

.sat

iva

AP0

0446

431

4e-

113

2T5

1PS

1C

F369

286

Plas

tidri

boso

mal

prot

ein

L19

O.s

ativ

aA

P005

008

847e

-28

gene

prec

urso

r2T

111

OS

1,2

CF3

6927

8G

luta

thio

ne-S

-tra

nsfe

rase

A.t

hali

ana

NM

_106

118

197

4e-4

92T

121

TF

1C

F369

279

Mito

chon

dria

lim

port

inne

rO

.sat

iva

AF1

5011

310

83e

-29

mem

bran

etr

ansl

ocas

esu

buni

tTim

92T

132

BS

1,2

CF3

6928

0O

xido

redu

ctas

e,2O

G-F

e(II

)A

.tha

lian

aN

M_1

1278

610

88e

-23

oxyg

enas

efa

mily

2T14

1E

M1,

2C

F369

287

Thi

azol

ebi

osyn

thet

icC

itru

ssi

nens

isZ

8298

329

8e-

127

enzy

me

prec

urso

r

1Fu

llle

ngth

sequ

ence

sre

cove

red

bysc

reen

ing

1Tan

d2T

stre

ssed

cDN

Alib

rary

,res

pect

ivel

y.2

Tra

nscr

ipts

inve

stig

ated

inre

al-t

ime

PCR

.3

Clo

necl

assi

fica

tion

asde

scri

bed

inR

esul

ts.

Con

tinue

dov

erle

af

Statistical analysis

Relative quantification analysis was performed with the rela-tive expression software tool REST (Pfaffl et al. 2002), allow-ing group-wise comparison and statistical evaluation of databy means of a pairwise reallocation test. The mathematicalmodel in REST used to calculate the relative expression ratiois:

( )E

E

C

C

targettarget(MEANcontrol – sample)

refr

t

t( )

∆

∆ ef(MEANcontrol – sample)(1)

It is based on the mean threshold cycle deviation (∆Ct) of acontrol and a sample, normalized by a reference transcript andincludes a real-time PCR efficiency correction (E) E = 10 –1/k,where k represents the slope of the standard curve. The expres-sion ratios for the seven transcripts obtained at the differentsampling times were tested for significance by performing 104

randomizations with the pairwise reallocation test of REST.The absolute expression (Ct values) of β-tubulin was some-

what variable. The variability was independent of the speciesand experimental conditions but dependent on certain ex-tracts/individuals. The expression ratio for Q. robur/Q. pet-raea was below two (1.9) and in no case significant (P =0.431). The PCR efficiency for β-tubulin amplification was92%. The Ct values were 32.6 for Q. petraea and 30.2 forQ. robur because the calculated regression line included all ex-periments and individual samples. Therefore, although therewas some variation in the absolute values, this may not reflectsignificant regulation of the expression of the β-tubulin gene,to which we related the expression of our target genes. How-ever, for qPCR, the correction for any variation in the amountof starting material between samples is relevant when sampleswere obtained from different individuals.

The REST software tool allows for group-wise comparisonfor more samples in an experimental trial. Moreover, the ran-domization test makes no assumptions about distributions ofmeasured gene expression (except for random allocation tocontrol and treatment groups) and thus accounts for the ob-served variances in gene expression measurements, when nor-mal distributions are not expected (Pfaffl et al. 2002). Hence,significant expression ratios for the investigated genes wouldbe observed between 7.2 for down-regulation and 6.3 forup-regulation when comparing the species directly at each sin-gle sampling time, and between 18.5 for down-regulation and4.5 for up-regulation when comparing the species over thetime course (see Figures 2 and 3). These values greatly exceedthe variation in β-tubulin gene expression.

Results

Osmotic-stress-induced genes

The subtractive hybridization technique identified differencesin gene expression between Q. petraea cell suspensions cul-tured under osmotic stress and control conditions. Originally,the isolated gene fragments (ESTs) yielded average sequences



Tabl

e1

Con

t’d.

Thi

rty-

thre

eos

mor

egul

ated

tran

scri

pts

from

Que

rcus

petr

aea

callu

sce

llscu

lture

dfo

r1h

(1T

)or2

days

(2T

)in

hype

rosm

otic

P24-

4Sm

ediu

man

dis

olat

edby

subt

ract

ive

hybr

idiz

atio

nan

ddi

ffer

entia

lscr

eeni

ng.A

bbre

viat

ions

:EST

=ex

pres

sed

sequ

ence

tag;

BS

=bi

osyn

thes

is;C

H=

chap

eron

e;C

K=

cyto

kine

sis;

CO

=ce

llula

rorg

aniz

atio

n;E

M=

ener

gym

etab

olis

m;O

S=

oxid

ativ

est

ress

;PD

=pr

otei

nde

stin

atio

n;PS

=pr

otei

nsy

nthe

sis;

RF

=re

gula

tory

func

tion;

TF

=tr

ansp

ortf

acili

tatio

n;T

S=

tran

scri

ptio

n;U

C=

uncl

assi

fied

;and

UK

=un

know

n.

Osm

otic

EST

Clo

neFu

nctio

nal

Not

eG

enB

ank

BL

AST

XSp

ecie

sG

enB

ank

Scor

eE

-tr

eatm

ent

cate

gory

3ca

tego

ryen

try

hom

olog

yre

fere

nce

valu

e(n

ucel

otid

e)

2da

ys2T

161

CK

CN

7256

70K

inet

ocho

repr

otei

nB

rass

ica

junc

eaA

B05

7431

886e

-17

2T19

1E

M1

CF3

6928

2N

AD

-Dep

ende

ntm

alat

ede

hydr

o-P

runu

spe

rsic

aA

F367

442

574

e-16

2ge

nase

2T25

1E

MC

N72

5672

NA

DH

-Ubi

quin

one

oxid

ored

ucta

seS.

tube

rosu

mX

8580

824

63e

-64

2T26

1U

CC

N72

5671

End

ochi

tinas

epr

ecur

sor

Pha

seol

usvu

lgar

isM

1396

830

19e

-81

2T32

1PS

1C

F369

284

40S

Rib

osom

alpr

otei

nS1

1A

.tha

lian

aN

M_1

1475

217

72e

-43

2T33

2C

H1,

2C

F369

271

Hea

tsho

ckco

gnat

epr

otei

n80

L.e

scul

entu

mM

9654

972

90.

02T

342

PS1

CF3

6928

560

SR

ibos

omal

prot

ein

l13a

A.t

hali

ana

NM

_124

253

335

7e-9

1

1Fu

llle

ngth

sequ

ence

sre

cove

red

bysc

reen

ing

1Tan

d2T

stre

ssed

cDN

Alib

rary

,res

pect

ivel

y.2

Tra

nscr

ipts

inve

stig

ated

inre

al-t

ime

qPC

R.

3C

lone

clas

sifi

catio

nas

desc

ribe

din

Res

ults

.

of about 520 bp, ranging from 200 to 1200 bp. According totheir hybridization patterns, clones were classified into threegroups (Figure 1): (1) clones that hybridized to the forward-subtracted probe but not to the reverse-subtracted probe; theseclones represented the 42 putative osmo-induced genes; (2) 24clones that hybridized both to the forward- and reverse-sub-tracted cDNA but showed a stronger hybridization signal withthe forward subtracted probe (these clones were analyzed fur-ther on virtual Northern blots to determine the nature of thispattern); and (3) four clones that hybridized both to the for-ward- and reverse-subtracted cDNA but exhibited a strongerhybridization signal with the reverse-subtracted probe; theseclones represented the four down-regulated genes (including1T11 and 1T34). Thus, 65% of the originally isolated 107 SLclones were differentially expressed (Figure 1). In the SL1Tand SL2T libraries, 47% (34) and 41% (14) of the clones wereinduced, respectively, whereas two clones per library wereidentified as down-regulated. Only the clones listed in Table 1were further analyzed.

Sequence comparison of the 33 newly isolated osmotic-stress-induced genes of oak using BLASTX showed homol-ogies to various proteins (Table 1). Best homologies of ourtranscripts were found exclusively for plant gene products.

Additional sequence information on both the 3′ and 5′ endswas obtained for 24 ESTs, and one clone was isolated as afull-length cDNA, thus, altogether, 25 cDNAs were putativelyfull length (Table 1). The analyzed ESTs were related to thefollowing functional categories: (1) protein synthesis machin-ery (six ribosomal proteins identified), (2) protein destination(ubiquitin) and chaperonins (heat-shock cognate protein), (3)cellular organization (LTP, fiddlehead protein), (4) oxidativestress (glutathione-S-transferase; GST), cytokinesis (kineto-chore protein), (5) transcription (Dc3 promoter-binding factor,dehydration-responsive element binding protein) and (5a)other regulatory functions (TGF β-inducible protein, eukary-otic translation initiation factor), (6) transport facilitation (in-ner membrane translocase subunit), (7) energy metabolism(pyruvate kinase, thiazole biosynthetic enzyme, NAD-dependentmalate dehydrogenase, ketol-acid reductoisomerase, NADH-ubiquinone oxidoreductase, triosephosphate isomerase), (7a)other biosynthetic functions (betaine aldehyde dehydrogenase(BADH), oxidoreductase/oxygenase gene family) and (8) un-classified (noduline gene mtN19, endochitinase).

Gene expression

To investigate the osmoregulated expression of genes identi-fied in Q. petraea cell cultures in vivo, we subjected plantletsof Q. petraea and Q. robur to hyperosmotic stress in a hydro-ponic system and monitored the expression profiles of sevengenes, which were selected based on homologies and putativefunctions in osmotic stress (see Table 1), by real-time qPCRafter 0, 1, 6, 24 and 72 h of treatment (Figure 2).

Time course of gene expression

We related gene expression after various treatment times tothat obtained before the imposition of the hyperosmotic treat-ment (zero time). As shown in Figure 2, the seven selectedgenes were significantly up-regulated (0.001 ≤ P ≤ 0.05) in re-sponse to the P24-4S treatment. Generally, significant up-reg-ulation was observed more frequently in Q. petraea (thiazolebiosynthetic enzyme (1 h), LTP (24 h), heat shock protein(24 h), BADH (1 and 24 h), oxidoreductase (24 and 72 h),glutathione-S-transferase (24 h)) than in Q. robur (thiazolebiosynthetic enzyme (1 h)) (Figure 2). In Q. robur plantlets,for example, expression of BADH was down-regulated signif-icantly at three sampling times: 1 h (two times), 6 h (ninetimes; P = 0.002) and 72 h (four times; P = 0.007), whereas inQ. petraea plantlets, BADH expression was significantly up-regulated at two sampling times: 1 h (4.5 times) and 24 h (Fig-ure 2e). No up-regulation of gene expression was observed ineither species at 6 h after the onset of the P24-4S treatment(Figure 2).

Expression of the investigated transcripts did not increasesmoothly with increasing time of hyperosmotic treatment (Fig-ure 2). For example, the gene coding for thiazole biosyntheticenzyme was significantly up-regulated after 1 h of hyper-osmotic treatment in both Q. petraea (P = 0.004) and Q. robur(P = 0.031), whereas the gene coding for BADH was signifi-cantly up-regulated after 1 h of hyperosmotic treatment inQ. petraea (P = 0.006) and significantly down-regulated inQ. robur (P = 0.023). Thus, at 1 h after stress induction, the ab-solute difference in expression between sessile and pedun-culate oak (relative to the respective controls) was twofold forthiazole biosynthetic enzyme and more than ninefold forBADH. Expression of these genes decreased during the subse-quent 5 h of treatment (Figures 2a and 2e) and then increased



Table 2. Amplification primers for qPCR using the Sybr Green detection system. The primer sequences corresponding to seven targets and an en-dogenous reference (Est172ma) are listed in the 5′ to 3′ orientation. All primers had a melting temperature of about 60 °C.

EST name Designation Sequence (forward primer) Designation Sequence (reverse primer)

1T9 1T9-341F CCGGTAGCATCTCTGGACTTA 1T9-458R ACCTCTCACTTCACGTTTTTGCA1T57 1T57-434F GGTGGATGTAAGCAAAGCTGTCT 1T57-534R AACATCACATTGTAGGCTGAAGCA1T62 1T62-975F AAGGCTTTTCAGGCAGGAATAGTTT 1T62-1097R GTTTTCAATTCCCCATTCCCCTAATT2T11 2T11-513F CTATTGGCTCACATGGATGCAA 2T11-615R AGAGCATTGACCCATGAGAGAAA2T13 2T13L-305F GCACTGCCAATCTCAGAGATGA 2T13L-405R AAGCCCATCGAGGTAACTGATG2T14 2T14-897F GGATGCCATTGTGAGGCTTACTA 2T14-1001R CCAAATGTGGGTCCCATTCTTG2T33 2T33-1189F GATTCTGAGGACCTTCCTCTCA 2T33-1290R GAGCTCAATGCACTTCTTAACCAAATTEst172ma Est172ma-420F GCCATTACACTGAAGGAGCTGA Est172ma-499R GTGACACACTTGAAACCCTTGC

again between 6 and 24 h after imposition of osmotic stress. Atthe 24-h sampling, expression of the BADH transcript wasup-regulated 1.6 times in Q. petraea (P = 0.001) (Figure 2e).Transcripts that showed a 24-h delay in response to the stresstreatment included LTP (Q. petraea), GST (Q. petraea), oxido-

reductase (Q. petraea) and heat-shock protein (Q. petraea)(Figures 2b, 2f, 2c and 2d, respectively). One day after imposi-tion of osmotic stress in Q. petraea plantlets, up-regulation ofLTP was 1.7 times (P = 0.002), GST 1.9 times (P = 0.001),oxidoreductase 2.9 times (P = 0.001) and heat-shock protein



Figure 1. Differential screening of the 73 positive clones originating from the 1T experiment (SL1T) and 34 positive clones from the 2T experi-ment (SL2T). Hybridization with forward- (1TS, 2TS) and reverse-subtracted (K1TS, K2TS) probes and non-subtractive controls (KAB, 1TAB,2TAB) is displayed, yielding 70 clones as defined by their position on the filter as well as by specific annotation (shown in brackets). ForSL1T/1TS hybridization (50 clones): 2A(1T9), 4A(1T25), 6A(1T41), 8A(1T57), 9A, 10A, 3B, 4B(1T26), 5B(1T34), 6B, 7B, 1C, 2C(1T11), 3C,4C, 5C, 6C, 7C, 8C, 9C, 1D, 2D(1T12), 4D, 6D, 9D, 2E, 3E(1T21), 5E, 7E, 9E(1T69), 3F, 4F, 5F, 8E, 8F(1T62), 1G(1T7), 2G(1T15), 3G,4G(1T31), 5G, 6G, 8G(1T63), 9G, 1H(1T8), 3H(1T24), 4H, 5H, 7H, 8H, 9H(1T72); and for SL2T/2TS hybridization (20 clones): 1A, 4A(2T25),5A(2T33), 1B(2T2), 3B(2T18), 4B(2T26), 5B(2T34), 1C(2T3), 2C(2T11), 3C(2T19), 4C, 2D(2T12), 3D, 1E(2T5), 2E(2T13), 2F(2T14), 3F, 1G,2H(2T16), 4H(2T32).

80 was 2.0 times (P = 0.001) that in the control group. Theoxidoreductase gene still showed elevated expression inQ. petraea after 72 h of treatment (P = 0.041), whereas expres-sion was down-regulated in Q. robur (P = 0.001) and 17 timesless than in Q. petraea (Figure 2c). After 6 h of osmotic treat-ment, GST and abscisic acid (ABA)-responsive element-bind-ing factor transcripts were down-regulated in both Q. petraea(GST, P = 0.001 and ABA-responsive element-binding factor,P = 0.001) and Q. robur (GST; P = 0.054; and ABA-responsiveelement-binding factor, P = 0.053), although expression of theGST transcript was 3.6 times less, and that of the ABA-respon-sive element-binding factor was 3.7 times less in Q. petraeathan in Q. robur (Figures 2f and 2g). Except for the gene cod-ing for the ABA-responsive element-binding factor, the othergenes showed either an early or a delayed response to osmoticstress.

Comparison of gene expression in Q. robur and Q. petraea

Figure 3 shows the expression of selected ESTs as a ratio of theexpression determined in Q. petraea to that in Q. robur. Theoxidoreductase gene (Figure 3c) had low expression in controlQ. petraea plants compared with control Q. robur plants (P =0.001) in which oxidoreductase gene expression increasedcontinuously during the 72-h treatment period, reaching a

much higher level of expression in Q. petraea (P = 0.002). Thegenes for thiazole biosynthetic enzyme (P = 0.01; Figure 3a),LTP (P = 0.001; Figure 3b) and BADH (P = 0.01; Figure 3e)showed a marked response immediately after the onset of theosmotic treatment (1 h) in Q. petraea, but not in Q. robur,whereas in the latter stages of the treatment, the difference be-tween species in expression of the thiazole biosynthetic en-zyme and BADH transcripts was small. In contrast, a secondpeak of LTP expression (P = 0.02) appeared after 24 h inQ. petraea but not in Q. robur. The heat shock protein (Fig-ure 3d) and the ABA-responsive element-binding factor (Fig-ure 3g) showed oscillating patterns, but there were no signifi-cant differences between species. Expression of the gene cod-ing for GST also oscillated between species, but expressionwas significantly (P = 0.002) higher after 24 h in Q. petraeathan in Q. robur. Although expression of the genes coding foroxidoreductase, BADH and GST was much lower in controlQ. petraea plants than in control Q. robur plants, expression ofthese genes increased more in response to osmotic stress inQ. petraea than in Q. robur. Furthermore, when gene expres-sion was high in control Q. petraea plantlets (e.g., thiazolebiosynthetic enzyme and LTP), even greater expression levels



Figure 2. Expression profiles ofselected genes in Quercuspetraea and Q. robur leaf tissueassayed at 0, 1, 6, 24 and 72 hafter the start of a hyperosmoticP24-4S treatment. (a) Thiazolebiosynthetic enzyme precursor,(b) lipid transfer protein precur-sor, (c) oxygenase/oxido-reductase, (d) heat-shockcognate protein 80, (e) betainealdehyde dehydrogenase, (f )glutathione-S- transferase and(g) Dc-3 promoter-binding fac-tor-2. Significance is indicatedas: ***, 0.001 ≤ P < 0.005; **,0.005 ≤ P ≤ 0.01; and *, 0.01 <P ≤ 0.05. All expression valuesare normalized to the expres-sion of β-tubulin. Relative ex-pression is calibrated based onthe corresponding gene expres-sion in Q. petraea and Q. roburplantlets grown in P24-S.

were observed in response to the osmotic stress treatment (Fig-ures 3a and 3b).

Discussion

Functions of genes isolated from tissue culture

We characterized 33 osmotic-stress-induced transcripts fromtissue cultures of Q. petraea. Of the isolated ESTs, 85% wererelated to specific protein functions, whereas 15% (five tran-scripts) corresponded to expressed sequences of A. thalianawith as yet unknown functions. Forty percent (12 transcripts)of the gene products were related to general metabolic turn-over, representing components of the protein-synthesizing andenergy metabolism machinery. Genes encoding proteins oftranscriptional control (2), proteins with general regulatoryfunctions (2), as well as gene products related to cell organiza-tion (2), were exclusively found after 1 h of osmotic stress (Ta-

ble 1). Endochitinase and mtN19 from the 2T library were notrelated to functional categories. Five of the remaining eightgenes showed homologies to biosynthetic enzymes (3), onemolecular chaperone and one enzymatic oxygen scavenger.The two putative oxidoreductases/oxygenases (1T12, 2T13)identified by suppression-subtracted hybridization exhibitedno best local alignment (Lalign), whereas ClustalW revealedsimilarity of 2T13 to the 3′ region of 1T12 (data not shown).Therefore, 1T12 and 2T13 was considered to be differentmembers of one gene family. One representative each for pro-tein destination (ubiquitin-related), cytokinesis and transportfacilitation was also annotated.

Transcripts (Table 1) that have been identified as targetgenes for stress tolerance in other plant species include theglycine betaine biosynthetic enzyme BADH (1T62), factorsinvolved in transcription control (dehydration-responsive ele-ment-binding protein 3 (1T31) (Kasuga et al. 1999) and Dc3promoter-binding factor (DPBF; 1T57)), oxygen scavengers



Figure 3. Expression ratios ofselected genes in Quercuspetraea versus Q. robur at 0,1, 6, 24 and 72 h after the startof a hyperosmotic P24-4Streatment. (a) Thiazolebiosynthetic enzyme precursor,(b) lipid transfer protein pre-cursor, (c) oxygenase/oxido-reductase, (d) heat-shockcognate protein 80, (e) betainealdehyde dehydrogenase, (f)glutathione-S-transferase and(g) Dc-3 promoter-binding fac-tor-2. Significance is indicatedas: ***, 0.001 ≤ P < 0.005; **,0.005 ≤ P ≤ 0.01; *, 0.01 < P≤ 0.05; and +, 0.05 < P ≤ 0.1.All expression values are nor-malized to the expression ofβ-tubulin.

like GST (2T11) and heat-shock proteins (HSPs) (hsp80;2T33). The BADH is one of the two biosynthetic enzymes ofglycine betaine, which is frequently classified as a “compati-ble osmolyte” involved in osmotic adjustment because it accu-mulates in the cytosol of water-stressed plant cells. Bray et al.(2000) suggested that glycine betaine functions directly to pre-vent salt-induced inactivation of ribulose-bisphosphate-car-boxylase and destabilization of the oxygen-evolving complexof photosystem II. Metabolic engineering of betaines to en-hance plant tolerance to abiotic stresses led to improved toler-ance to water stress in some transgenics although the resultswere species dependent and were diminished by limitation ofcholine supply to the chloroplast (Apse and Blumwald 2002).

Dc3 is a carrot late embryogenesis abundant gene that is ex-pressed abundantly during somatic and zygotic embryogen-esis. Although its expression is usually embryo-specific, it canalso be induced by ABA. The basic leucine zipper transcrip-tion factors—DPBF is one representative—are members ofthe ABA-dependent signal transduction pathway that operatesunder high-salt and drought conditions (Uno et al. 2000).

Both GST and HSPs have been targets for genetic engineer-ing of stress tolerance in plants. The GSTs, which are multi-functional and may possess glutathione peroxidase activity,serve as effective scavengers of reactive oxygen species (ROS)by catalyzing the conjugation of reduced glutathione to elec-trophilic sites of H2O2 as well as of organic hydroperoxides.Transgenic tobacco seedlings with elevated activities of GSTand peroxidase were found to acclimate better to salt stress(Roxas et al. 1997).

The HSPs, which respond to a variety of stresses and havebeen used as markers of the stress response (Feder and Hof-mann 1999), function as molecular chaperones that are in-volved in protein refolding and prevention of proteinunfolding by, for example, ROS. Osmotic stress is always ac-companied by oxidative stress, caused by ROS that damagemacromolecules and thereby affect protein stability. The HSPsare required in the cytosol, as well as in chloroplast and mito-chondrial compartments, to protect important functional units(e.g., oxygen-evolving complex of photosystem II, electrontransport of mitochondrial complex I) against ROS attack. Insome cases, engineered over-production of small HSPs hasbeen associated with improved stress tolerance (Sun et al.2002).

Lipid transfer proteins can be induced by osmotic stress(Trevino and O’Connell 1998, Bohnert et al. 2001, Seki et al.2001, Yubero-Serrano et al. 2003) and are involved in acclima-tion to water stress (Kader 1996, Wu et al. 2004). Although thein vivo function of LTPs has not been unequivocally demon-strated (Lindorff-Larsen et al. 2001, Beisson et al. 2003), it hasbeen suggested that LTPs serve as plant defense proteins byactivating defense responses, possibly by facilitating the for-mation of hydrophobic protective layers (Blein et al. 2002). Invitro experiments have demonstrated that LTPs can transferseveral classes of glyco- and phospholipids without specific-ity, thus they are also referred to as non-specific LTPs.

Members of the oxygenase gene family include genes cod-ing for biosynthetic enzymes for stress-related organic com-

pounds like ABA (Iuchi et al. 2000, Thompson et al. 2000) andglycine betaine (Russell et al. 1998). Two osmotic-stress-induced ESTs of oak were homologous (e–114; e–23) to the genefamily of 2-oxoglutarate (2OG) and Fe(II)-dependent oxygen-ases, a class of enzymes which is widespread in eukaryotesand catalyzes the oxidation of organic substrates using a di-oxygen molecule. In plants, members of this family catalyzethe formation of plant hormones (e.g., ethylene, gibberellins)and metabolites (e.g., flavones or anthocyanidins), the latterhave been characterized as non-enzymatic oxygen scavengers(Cushman and Bohnert 2000). Homologs within this genefamily may be involved in regulation of RNA stability or DNArepair through controlled methylation and demethylation(Aravind and Koonin 2001).

Differential gene expression between Q. petraea and Q. robur

A lower level of expression was found at 6 h than at 0 h for allthe genes studied, which may be associated with the effect ofcircadian rhythm on gene expression. The 6-h samples werethe only samples taken at 1700 h, and all the other treated sam-ples were taken at 1200 h. However, relating gene expressionin Q. petraea to that in Q. robur (Figure 3) showed that, underosmotic stress, down-regulation at 6 h was more pronouncedin the Q. petraea genotypes than in the Q. robur genotypes (ex-cept for oxygenase and BADH; Figures 3c and 3e) than at anyother sampling time.

Both species showed increased expression of the gene cod-ing for thiazole biosynthetic enzyme after 1 h of osmotic treat-ment (Figures 2a and 3a), but the response was significantlymore pronounced for Q. petraea than for Q. robur (Figure 3a).After this initial increase and a temporary decrease at 6 h, geneexpression remained similar to that in the control plants, con-firming that energy metabolism is not up-regulated by pro-longed water deficit, as suggested by Bray (2002). Similarly,the gene coding for DPBF, which is reported to participate inthe ABA signal transduction pathway (Uno et al. 2000), wasnot significantly up-regulated by osmotic stress (Figure 2g). Inthe gene coding for DPBF, four mutations are known that areof great value for species differentiation (Scotti-Saintagne etal. 2004; I. Porth et al., unpublished observations).

Osmotic stress simultaneously causes oxidative stress, re-sulting in adverse effects on important cellular functions. Ac-cumulation of ROS can disrupt several steps of energy metab-olism; it is important, therefore, for an organism to protectpossible points of attack. Up-regulation of three genes (codingfor GST, oxidoreductase and BADH) involved in scavengingand preventing ROS attack was observed only in sessile oak(Figures 2f, 2c, 2e and 3f, 3c, 3e, respectively). Expression ofthe BADH transcript increased immediately in response to os-motic stress, suggesting immediate protection, whereas thegenes coding for GST and oxidoreductase may have a greaterrole in the process of environmental adaptation.

The progression of gene expression for BADH (Figure 2e)and oxygenase 2OG (Figure 2c) differed strikingly betweenQ. robur and Q. petraea. The transcript for oxygenase enzymewas significantly down-regulated in Q. robur, whereas it was



up-regulated in Q. petraea (Figure 2c). Figure 3c indicatesthat, in the absence of osmotic stress, expression of this genewas much higher in Q. robur than in Q. petraea, whereas dur-ing osmotic stress, it increased steadily in Q. petraea. Basedon the wide-range effects of the products of this gene familyon various cell functions, we speculate that it plays an impor-tant role in the response of Q. petraea to osmotic and waterstress.

Expression of the gene coding for the BADH enzyme,which catalyzed the synthesis of the osmo-protective moleculeglycine betaine, was down-regulated in Q. robur, whereas itwas up-regulated in Q. petraea (Figure 2e). Although expres-sion of this gene was lower in control Q. petraea plants than incontrol Q. robur plants (Figure 3e), it showed an immediate(within 1 h) and marked response to the osmotic treatment inQ. petraea that was also evident after 24 h of treatment,whereas Q. robur made no such response, perhaps accountingfor the different tolerances of these species to water and os-motic stress. The BADH expression profile (Figure 2e) sug-gests that synthesis of glycine betaine is important both at theinitiation of osmotic stress and during subsequent adaptation.Both putative functions of glycine betaine could contribute tothe high osmotic stress tolerance of Q. petraea.

Expression of the gene coding for HSP was constant andsimilar in both species (Figure 3d), except for a slight increasein Q. petraea after 24 h of osmotic stress treatment(Figure 2d). Thus, HSP expression appears to follow the over-all stress response.

Expression of the LTP gene, responsible for formation andreinforcement of plant surface layers, was rather constant inQ. robur compared with Q. petraea (Figure 2b). The initial ex-pression level in Q. petraea was high and after an immediateincrease (1 h), it declined steeply between 1 and 5 h after theonset of the osmotic stress treatment and then increasedstrongly by 24 h. Expression of LTP is commonly up-regu-lated in response to water stress and may represent an impor-tant protective mechanism. The high expression of LTP inQ. petraea, which is adapted to dry sites, likely reflects speciesspecificity in expression level.

We conclude that Q. petraea is better adapted to water stressthan Q. robur because functions reportedly necessary for wa-ter stress tolerance were mostly up-regulated in Q. petraea,whereas the same functions were less responsive to osmoticstress in Q. robur. Our gene expression profiling data supportthe results of recent physiological studies (Ponton et al. 2002)and indicate: (1) the preference of sessile oak for well-drainedsites, whereas pedunculate oak tends to occur on moistground; and (2) the different sensitivities of these white oaks todrought, with Q. petraea being more drought tolerant thanQ. robur.

References

Aas, G. 1998. Morphologische und Ökologische Variation Mittel-europäischer Quercus-Arten: ein Beitrag zum Verständnis der Bio-diversität. Libri Botanici, Band 19, IHW Verlag, 213 p.

Apse, M.P. and E. Blumwald. 2002. Engineering salt tolerance inplants. Curr. Opin. Biotechnol. 13:146–150.

Aravind, L. and E.V. Koonin. 2001. The DNA-repair protein AlkB,EGL-9, and leprecan define new families of 2-oxoglutarate- andiron-dependent dioxygenases. Genome Biology 2(3). Available athttp://genomebiology.com/2001/2/3/research/0007.

Beisson, F., A.J.K. Koo, S. Ruuska et al. 2003. Arabidopsis genes in-volved in acyl lipid metabolism. A 2003 census of the candidates, astudy of the distribution of expressed sequence tags in organs, and aweb-based database. Plant Physiol. 132:681–697.

Blein, J.P., P. Coutos-Thévenot, D. Marion and M. Ponchet. 2002.From elicitins to lipid-transfer proteins: a new insight in cell signal-ling involved in plant defence mechanisms. Trends Plant Sci. 7:293–296.

Blumwald, E. 2000. Sodium transport and salt tolerance in plants.Curr. Opin. Cell Biol. 12:431– 434.

Bodénès, C., S. Joandet, F. Laigret and A. Kremer. 1997. Detection ofgenomic regions differentiating two closely related oak species,Quercus petraea (Matt.) Liebl. and Quercus robur L. Heredity 78:433– 444.

Bohnert, H.J., P. Ayoubi, C. Borchert et al. 2001. A genomics ap-proach towards salt stress tolerance. Plant Physiol. Biochem. 39:295–311.

Borovskii, G.B., I.V. Stupnikova, A.I. Antipina, S.V. Vladimirova andV.K. Voinikov. 2002. Accumulation of dehydrin-like protein in themitochondria of cereals in response to cold, freezing, drought andABA treatment. BioMed Central Plant Biol. Vol. 2, Article 5.

Bray, E.A. 2002. Classification of genes differentially expressed dur-ing water-deficit stress in Arabidopsis thaliana: an analysis usingmicroarray and differential expression data. Ann. Bot. 89:803–811.

Bray, E.A., J. Biley-Serres and E. Weretilnyk. 2000. Responses toabiotic stresses. In Biochemistry and Molecular Biology of Plants.Eds. B. Buchanan, W. Gruissem and R.L. Jones. John Wiley andSons, New York, pp 1158 –1203.

Burssens, S., K. Himanen, B. Van de Cotte, T. Beeckman, M. VanMontagu, D. Inzé and N. Verbruggen. 2000. Expression of cell cy-cle regulatory genes and morphological alterations in response tosalt stress in Arabidopsis thaliana. Planta 211:632–640.

Churin, Y., S. Schilling and T. Börner. 1999. A gene family encodingglutathione peroxidase homologues in Hordeum vulgare (barley).FEBS Lett. 459:33–38.

Cooper, B., J.D. Clarke, P. Budworth et al. 2003. A network of ricegenes associated with stress response and seed development. Proc.Natl. Acad. Sci. USA 100:4945– 4950.

Cushman, J.C. and H.J. Bohnert. 2000. Genomic approaches to plantstress tolerance. Curr. Opin. Plant Biol. 3:117–124.

Delauney, A.J. and D.P.S. Verma. 1993. Proline biosynthesis andosmoregulation in plants. Plant J. 4:215–223.

Dubos, C., G. Le Provost, D. Pot, F. Salin, C. Lalane, D. Madur, J.M.Frigerio and C. Plomion. 2003. Identification and characterizationof water-stress-responsive genes in hydroponically grown mari-time pine (Pinus pinaster) seedlings. Tree Physiol. 23:169–179.

Dumolin-Lapègue, S., A. Kremer and R.J. Petit. 1999a. Are chlor-oplast and mitochondrial DNA variations species-independent inoaks? Evolution 53:1406 –1413.

Endemann, M. and E. Wilhelm. 1999. Factors influencing the induc-tion and viability of somatic embryos of Quercus robur L. Biol.Plant. 42:499–504.

Feder, M.E. and G.E. Hofmann. 1999. Heat-shock proteins, molecu-lar chaperones and the stress response: evolutionary and ecologicalphysiology. Annu. Rev. Physiol. 61:243 –282.



Gao, G., J. Widmer, D. Stapleton, T. Teh, T. Cox, B.E. Kemp andL.A. Witters. 1995. Catalytic subunits of the porcine and rat 5-AMP-activated protein kinase are members of the SNF1 proteinkinase. Biochim. Biophys. Acta 1266:73–82.

Garcia, A.B., J.D. Engler, S. Iyer, T. Gerats, M. Van Montagu andA.B. Caplan. 1997. Effects of osmoprotectants upon NaCl stress inrice. Plant Physiol. 115:159 –169.

Gömöry, D. 2000. A gene coding for a non-specific NAD-dependentdehydrogenase shows a strong differentiation between Quercusrobur and Quercus petraea. For. Genet. 7:167–170

Holmstrom, K.O., S. Somersalo, A. Mandal, T.E. Palva and B. Welin.2000. Improved tolerance to salinity and low temperature in trans-genic tobacco producing glycine betaine. J. Exp. Bot. 51:177–185.

Hurkman, W.J. and C.K. Tanaka. 1996. Effect of salt stress on germingene expression in barley roots. Plant Physiol. 110:971–977.

Ishitani, M., T. Nakamura, S.Y. Han and T. Takabe. 1995. Expressionof the betaine aldehyde dehydrogenase gene in barley in responseto osmotic stress and abscisic acid. Plant Mol. Biol. 27:307–315.

Iuchi, S., M. Kobayashi, K. Yamaguchi-Shinozaki and K. Shinozaki.2000. A stress-inducible gene for 9-cis-epoxycarotenoid dioxy-genase involved in abscisic acid biosynthesis under water stress indrought tolerant cowpea. Plant Physiol. 123:553–562.

Kader, J.C. 1996. Lipid-transfer proteins in plants. Annu. Rev. PlantPhysiol. Plant Mol. Biol. 47:627–654.

Kasuga, M., Q. Liu, S. Miura, K. Yamaguchi-Shinozaki and K. Shi-nozaki. 1999. Improving plant drought, salt and freezing toleranceby gene transfer of a single stress-inducible transcription factor.Nat. Biotechnol. 17:287–291.

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

Koizumi, M., K. Yamaguchi-Shinozaki, H. Tsuji and K. Shinozaki.1993. Structure and expression of 2 genes that encode distinctdrought-inducible cysteine proteinases in Arabidopsis thaliana.Gene 129:175–182.

Kreps, J.A., Y. Wu, H.S. Chang, T. Zhu, X. Wang and J.F. Harper.2002. Transcriptome changes for Arabidopsis in response to salt,osmotic and cold stress. Plant Physiol. 130:2129 –2141.

Lindorff-Larsen, K., M.H. Lerche, F.M. Poulsen, P. Roepstorff andJ.R. Winther. 2001. Barley lipid transfer protein, LTP1, contains anew type of lipid-like post-translational modification. J. Biol.Chem. 276:33,547–33,553.

Liu, J. and J.K. Zhu. 1997. Proline accumulation and salt-stress-in-duced gene expression in a salt-hypersensitive mutant of Arabi-dopsis. Plant Physiol. 114:591–596.

Mariette, S., J. Cottrell, U.M. Csaikl et al. 2002. Comparison of levelsof genetic diversity detected with AFLP and microsatellite markerswithin and among mixed Q. petraea (Matt.) Liebl. and Q. robur L.stands. Silvae Genet. 51:72–79.

Mikolajczik, M., O.S. Awotunde, G. Muszynska, D.F. Klessig andG. Dobrowolska. 2000. Osmotic stress induces rapid activation of asalicylic acid-induced protein kinase and a homolog of proteinkinase ASK1 in tobacco cells. Plant Cell 12:165–178.

Moons, A., G. Bauw, E. Prinsen, M. Van Montagu and D. Van derStraeten. 1995. Molecular and physiological responses to abscisicacid and salts in roots of salt-sensitive and salt-tolerant Indica ricevarieties. Plant Physiol. 107:177–186.

Muir, G., C.C. Fleming and C. Schlötterer. 2000. Brief communica-tions (taxonomy): species status of hybridizing oaks. Nature 405:1016.

Patnaik, D. and P. Khurana. 2001. Wheat biotechnology: a mini-re-view. Eur. J. Biochem. 4:1–29.

Petit, R.J., D.B. Wagner and A. Kremer. 1993. Ribosomal DNA geneand chloroplast DNA polymorphisms in a mixed stand of Quercusrobur and Q. petraea. Ann. For. Sci. 50:41s– 47s.

Pfaffl, M.W., G.W. Horgan and L. Dempfle. 2002. Relative expres-sion software tool (REST) for group-wise comparison and statisti-cal analysis of relative expression results in real-time PCR. NucleicAcids Res. Available at http://nar.oupjournals.org/cgi/content/full/30/9/e36.

Pih, K.T., H.J. Jang, S.G. Kang, H.L. Piao and I. Hwang. 1997. Isola-tion of molecular markers for salt stress responses in Arabidopsisthaliana. Mol. Cell 7:567–571.

Ponton, S., J.L. Dupouey, N. Bréda and E. Dreyer. 2002. Comparisonof water-use efficiency of seedlings from two sympatric oak spe-cies: genotype × environment interactions. Tree Physiol. 22:413– 422.

Prewein, C. and E. Wilhelm. 2003. Plant regeneration from encapsu-lated somatic embryos of pedunculate oak (Quercus robur L.) InVitro Cell. Dev.–Plant 38:613–617.

Rabbani, M.A., K. Maruyama, H. Abe et al. 2003. Monitoring expres-sion profiles of rice genes under cold, drought and high-salinitystresses and abscisic acid application using cDNA microarray andRNA gel-blot analyses. Plant Physiol. 133:1755–1767.

Ried, J.L. and M.K. Walker-Simmons. 1993. Group 3 late embryo-genesis abundant proteins in desiccation-tolerant seedlings ofwheat (Triticum aestivum L.). Plant Physiol. 102:125–131.

Roxas, V.P., R.K. Smith, Jr., E.R. Allen and R.D. Allen. 1997. Over-expression of glutathione-S-transferase/glutathione peroxidase en-hances the growth of transgenic tobacco seedlings during stress.Nat. Biotechnol. 15:988–991.

Roxas, V.P., S.A. Lodhi, D.K. Garrett, J.R. Mahan and R.D. Allen.2000. Stress tolerance in transgenic tobacco seedlings that over-express glutathione-S-transferase/glutathione peroxidase. PlantCell Physiol. 41:1229–1234.

Russell, B.L., B. Rathinasabapathi and A.D. Hanson. 1998. Osmoticstress induces expression of choline monooxygenase in sugar beetand amaranth. Plant Physiol. 116:859–865.

Sambrook, J., E.F. Fritsch and T. Maniatis. 1989. Molecular cloning.A laboratory manual. 2nd Edn. Cold Spring Harbor LaboratoryPress, Cold Spring Harbor, NY, 676 p.

Samuel, R., W. Pinsker and F. Ehrendorfer. 1995. Electrophoreticanalysis of genetic variation within and between populations ofQuercus cerris, Q. pubescens, Q. petraea and Q. robur (Fagaceae).Bot. Acta 108:290 –299.

Saneoka, H., C. Nagasaka, D.T. Hahn, W.J. Yang, G.S. Prema-chandra, R.J. Joly and D. Rhodes. 1995. Salt tolerance of glycine-betaine-deficient and -containing maize lines. Plant Physiol. 107:631–638.

Scotti-Saintagne, C., S. Mariette, I. Porth, P.G. Goicoechea, T. Barre-neche, C. Bodénès, K. Burg and A. Kremer. 2004. Genome scan-ning for interspecific differentiation between two closely relatedoak species (Quercus robur L. and Q. petraea (Matt.) Liebl.). Ge-netics 168:1615–1626.

Seki, M., M. Narusaka, H. Abe, M. Kasuga, K. Yamaguchi-Shino-zaki, P. Carninci, Y. Hayashizaki and K. Shinozaki. 2001. Monitor-ing the expression pattern of 1300 Arabidopsis genes underdrought and cold stresses by using a full-length cDNA microarray.Plant Cell 13:61–72.

Seki, M., M. Narusaka, J. Ishida et al. 2002. Monitoring the expres-sion profiles of 7000 Arabidopsis genes under drought, cold andhigh-salinity stresses using a full-length cDNA microarray. Plant J.31:279–292.



Sharma, K.K. and M. Lavanya. 2002. Recent developments in trans-genics for abiotic stress in legumes of the semi-arid tropics.JIRCAS Working Report. Japan International Research Center forAgricultural Sciences, Tsukuba, Japan, pp 61–73.

Shen, Y.G., W.K. Zhang, S.J. He, J.S. Zhang, Q. Liu and S.Y. Chen.2003. An EREBP/AP-2-type protein in Triticum aestivum was aDRE-binding transcription factor induced by cold, dehydration andABA stress. Theor. Appl. Genet. 106:923–930.

Sofo, A., B. Dichio, C. Xiloyannis and A. Masia. 2004. Lipoxygenaseactivity and proline accumulation in leaves and roots of olive treesin response to drought stress. Physiol. Plant. 121:58–65.

Sun, W., M. Van Montagu and N. Verbruggen. 2002. Small heat shockproteins and stress tolerance in plants. Biochem. Biophys. Acta1577:1–9.

Teasdale, R. 1992. Formulation of plant culture media and applica-tions therefore. International Publication No. WO9207460, PatentNo. Europe: 92902531.0, Forbio PTY, Queensland, Australia.

Thompson, A.J., A.C. Jackson, R.A. Parker, D.R. Morpeth, A. Bur-bidge and I.B. Taylor. 2000. Abscisic acid biosynthesis in tomato:regulation of zeaxanthin epoxidase and 9-cis-epoxycarotenoid di-oxygenase mRNAs by light/dark cycles, water stress and abscisicacid. Plant Mol. Biol. 42:833–845.

Trevino, M.B. and M.A. O’Connell. 1998. Three drought-responsivemembers of the nonspecific lipid–transfer protein gene family inLycopersicon pennellii show different developmental patterns ofexpression. Plant Physiol. 116:1461–1468.

Uno, Y., T. Furihata, H. Abe, R. Yoshida, K. Shinozaki and K. Yama-guchi-Shinozaki. 2000. Arabidopsis basic leucine zipper transcrip-tion factors involved in an abscisic acid-dependent signal trans-duction pathway under drought and high-salinity conditions. Proc.Natl. Acad. Sci. USA 97:11,632–11,637.

Van Wees, S.C.M., E.A.M. De Swart, J.A. Van Pelt, L.C. Van Loonand C.M.J. Pieterse. 2000. Enhancement of induced disease resis-tance by simultaneous activation of salicylate- and jasmonate-de-pendent defense pathways in Arabidopsis thaliana. Proc. Natl.Acad. Sci. USA 97:8711–8716.

Wu, G., A.J. Robertson, X. Liu, P. Zheng, R.W. Wilen, N.T. Nesbittand L.V. Gusta. 2004. A lipid transfer protein gene BG-14 is differ-entially regulated by abiotic stress, ABA, anisomycin, and sphin-gosine in bromegrass (Bromus inermis). J. Plant Physiol. 161:449 – 458.

Xiong, L., M. Ishitani and J.K. Zhu. 1999. Interaction of osmoticstress, temperature and abscisic acid in the regulation of gene ex-pression in Arabidopsis. Plant Physiol. 119:205–211.

Xiong, L., H. Lee, M. Ishitani and J.K. Zhu. 2002. Regulation of os-motic stress-responsive gene expression by LOS6/ABA1 locus inArabidopsis. J. Biol. Chem. 277:8588–8596.

Xu, D., X. Duan, B. Wang, B. Hong, T.H.D. Ho and R. Wu. 1996. Ex-pression of a late embryogenesis abundant protein gene, HVA1,from barley confers tolerance to water deficit and salt stress intransgenic rice. Plant Physiol. 110:249–257.

Yamaguchi-Shinozaki, K., M. Kasuga, Q. Liu, K. Nakashima, Y. Sa-kuma, H. Abe, Z.K. Shinwar, M. Seki and K. Shinozaki. 2002. Bio-logical mechanisms of drought stress response. JIRCAS WorkingReport. Japan International Research Center for Agricultural Sci-ences, Tsukuba, Japan, pp 1–8.

Yubero-Serrano, E.M., E. Moyano, N. Medina-Escobar, J. Muñoz-Blanco and J.L. Caballero. 2003. Identification of a strawberrygene encoding a non-specific lipid transfer protein that responds toABA, wounding and cold stress. J. Exp. Bot. 54:1865–1877.

Zhang, Z. and S.J. Gurr. 2001. Expression and sequence analysis ofthe Blumeria graminis mitogen-activated protein kinase genes,mpk1 and mpk2. Gene 266:57–65.

Zhu, J.K. 2002. Salt and drought stress signal transduction in plants.Ann. Rev. Plant Biol. 53:247–273.



Identification of adaptation-specific differences in mRNA expression of sessile and pedunculate oak...

Documents

Transcript of Identification of adaptation-specific differences in mRNA expression of sessile and pedunculate oak...