Heat stress-induced H2O2 is required for effective expressionof heat shock genes in Arabidopsis

Roman A. Volkov Æ Irina I. Panchuk ÆPhillip M. Mullineaux Æ Friedrich Schoffl

Received: 19 May 2005 / Accepted: 15 March 2006

� Springer Science+Business Media B.V. 2006

Abstract The mechanisms of sensing and signalling of

heat and oxidative stresses are not well understood. The

central question of this paper is whether in plant cells

oxidative stress, in particular H2O2, is required for heat

stress- and heat shock factor (HSF)-dependent expression

of genes. Heat stress increases intracellular accumulation

of H2O2 in Arabidopsis cell culture. The accumulation was

greatly diminished using ascorbate as a scavenger or

respectively diphenyleneiodonium chloride (DPI) as an

inhibitor of reactive oxygen species production. The

mRNA of heat shock protein (HSP) genes, exemplified by

Hsp17.6, Hsp18.2, and the two cytosolic ascorbate perox-

idase genes Apx1, Apx2, reached similar levels by moderate

heat stress (37�C) or by treatment with H2O2, butylperox-

ide and diamide at room temperature. The heat-induced

expression levels were significantly reduced in the pres-

ence of ascorbate or DPI indicating that H2O2 is an

essential component in the heat stress signalling pathway.

Rapid (15 min) formation of heat shock promoter element

(HSE) protein-binding complex of high molecular weight

in extracts of heat-stressed or H2O2-treated cells and the

inability to form this complex after ascorbate treatment

suggests that oxidative stress affects gene expression via

HSF activation and conversely, that H2O2 is involved in

HSF activation during the early phase of heat stress. The

heat stress induction of a high mobility HSE-binding

complex, characteristic for later phase of heat shock re-

sponse, was blocked by ascorbate and DPI. H2O2 was

unable to induce this complex suggesting that H2O2 is in-

volved only in the early stages of HSF activation. Signifi-

cant induction of the genes tested after diamid treatment

and moderate expression of the sHSP genes in the presence

of 50 mM ascorbate at 37�C occurred without activation of

HSF, indicating that other mechanisms may be involved in

stress signalling.

Keywords Ascorbate peroxidase Æ DPI Æ Heat shock

factor Æ Heat shock protein Æ Hydrogen peroxide ÆOxidative stress

Introduction

In nature, plants are frequently subject to heat stress and

like other organisms, they have evolved strategies for

preventing and repairing cellular damage caused by heat

stress. In all species studied, heat stress results in the

production of heat shock proteins (HSP), which have been

classified into a number of families based on their molec-

ular mass (HSP100, HSP90, HSP70, HSP60 and small (s)

HSP), most of which have chaperone function (for review

see Jaenicke and Creighton 1993; Boston et al. 1996). The

expression of sHSP is a signature of the heat shock

response in plants. Plants are unique in the number and

Electronic Supplementary Material Supplementary material is

available for this article at http//dx.doi.org/10.1007/s11103-006-0045-4

Roman A. Volkov and Irina I. Panchuk contributed equally

R. A. Volkov Æ I. I. Panchuk Æ F. Schoffl (&)

Zentrum fur Molekularbiologie der Pflanzen – Allgemeine

Genetik, Universitat Tubingen, Auf der Morgenstelle 28, 72076

Tubingen, Germany

e-mail: friedrich.schoeffl@zmbp.uni-tuebingen.de

P. M. Mullineaux

Department of Biological Sciences, University of Essex,

Wivenhoe Park, Colchester, Essex CO4 3SQ, UK

R. A. Volkov Æ I. I. Panchuk

Department of Molecular Genetics and Biotechnology,

University of Chernivtsy, Kotsubinsky str. 2, 58012, Ukraine

Plant Mol Biol (2006) 61:733–746

DOI 10.1007/s11103-006-0045-4

123

complexity of sHSP that they produce upon heat stress, no

isoforms are expressed in vegetative tissue under non-stress

conditions (for review see Jakob and Buchner 1994;

Schoffl et al. 1998). Despite the ubiquitous nature of this

conserved response, little is known how plants sense heat

stress or about the signalling pathways resulting in heat

shock gene expression. In all organisms the heat shock

response is primarily regulated at the transcriptional level

by heat stress transcription factors (HSF), which are acti-

vated by stress for a specific binding to heat shock pro-

moter elements (HSE). In Arabidopsis 21 different HSF

genes have been identified (Nover et al. 2001), but only

few have been functionally characterized. AtHsf1 (HSF-

A1a) and AtHsf3 (HSF-A1b) are the regulators, which

become activated very early in the heat shock response and

are necessary for efficient expression of heat shock genes

because double knock out mutants, hsf1/3 are unable to

form high molecular weight HSE-binding complexes and

mRNA accumulation of HSF target genes is significantly

impaired upon heat stress (Lohmann et al. 2004). Recent

investigations have shown that not only conventional HSP

genes are controlled by HSF. Other genes encoding key

enzymes in biochemical pathways related to environmental

responses have been identified as targets of HSF regulation

in HSF3-transgenic plants (Panchuk et al. 2002; Paniku-

langara et al. 2004; Busch et al. 2005).

Permanent production of reactive oxygen species (ROS)

such as hydrogen peroxide (H2O2), superoxide, hydroxyl

radicals, and singlet oxygen is an unavoidable consequence

of aerobic metabolism. In plant tissues, about 1% of the

total O2 consumption goes to ROS production (Puntarulo

et al. 1988). An excessive generation of ROS leads to the

damage of proteins, lipids, and DNA and causes an oxi-

dative stress, which is a central factor in abiotic and biotic

stress phenomena (Bowler et al. 1992; Chen et al. 1993;

Asada 1999; Finkel and Holbrook 2000; Moller 2001).

There is considerable evidence that oxidative stress induces

expression of HSP and chaperones in plants, which can

provide a protective function against oxidative stress. In

tomato and rice, mitochondrial HSP22 and chloroplastic

HSP26, respectively, are induced by H2O2 (Banzet et al.

1998; Lee et al. 2000). In cyanobacteria and Arabidopsis,

high light and H2O2, respectively, induced the mRNAs of

some chaperones, HSP, and heat shock transcription factors

(Desikan et al. 2001; Hihara et al. 2001). It has also been

shown that thermotolerance can be induced by compounds

that induce oxidative bursts (Dat et al. 1998).

There is ample evidence that different environmental

stresses, including also high temperature, induce oxidative

stress in plants (Foyer et al. 1997; Dat et al. 1998). Very

short heat pulses can result in oxidative bursts of super-

oxide and/or hydrogen peroxide (Vallelian-Bindschedler

et al. 1998). This suggests that there is considerable inter-

linking between heat and oxidative stress signalling and

responses.

There are several possible sources of H2O2 in plants,

which can be activated during abiotic and biotic stress to

induce H2O2 generation and thereby oxidative stress, e.g.

electron transport chains (ETC) in chloroplasts and mito-

chondria, photorespiration in peroxisomes (Noctor and

Foyer 1998; Dat et al. 2000), or enzymatic sources

including plasma membrane-located NAD(P)H oxidases

(Desikan et al. 1998; Keller et al. 1998; Torres et al.

1998), and cell wall bound peroxidases/amine oxidase

(Bolwell and Wojtaszek 1997). Upon severe heat stress the

decrease in enzymatic activities of for example catalase

(Dat et al. 1998) and ascorbate peroxidase (Panchuk et al.

2002) may diminish removal of H2O2 and consequently

contribute to enhanced levels of ROS.

To survive under environmental stress conditions, plants

undergo a process of stress acclimation, which may require

changes in the flow of metabolites, suppression of path-

ways involved in the excessive production of ROS, and the

induction of various defense genes such as HSP and ROS

scavenging enzymes (Vierling 1991; Dat et al. 2000;

Mittler 2002; Panchuk et al. 2002).

Heat stress-induced oxidative damage becomes visible

by bleaching of green tissue, a feature that documents the

interlinkage between the two responses in plants. Arabid-

opsis plants that have acquired enhanced levels of thermo-

tolerance experienced lower levels of oxidative damage

during recovery from heat stress as compared with non-

conditioned plants and this protection from oxidative dam-

age correlates with greater survival rates of such plants

(Larkindale and Knight 2002). Transgenic overexpression

of HSF constructs in Arabidopsis resulted in a moderate

increase in basal thermotolerance that was also associated

with protection from oxidative bleaching of seedlings (Lee

et al. 1995; Prandl et al. 1998). This suggests that one aspect

of thermotolerance in Arabidopsis is an increased ability to

either prevent or repair heat-induced oxidative damage. This

conclusion is in accordance with the heat stress-induced

expression of ascorbate peroxidase (Apx) genes

(Storozhenko et al. 1998; Panchuk et al. 2002) and the

identification of novel heat-tolerant, HSF-dependently

expressed APX isoform in Arabidopsis (Panchuk et al.

2002). Although ROS were originally considered to be

detrimental to cells, it is now widely recognized that redox

regulation involving ROS is a key factor modulating cellular

activities (Allen and Tresini 2000; Dat et al. 2000). Espe-

cially, accumulation of relatively low-toxic H2O2 induces

the expression of various defense-related genes, including

glutathione S transferase, phenylalanine ammonia lyase,

and HSP (Vandenabeele et al. 2003; Levine et al. 1994;

Desikan et al. 1998; Neill et al. 1999; Grant et al. 2000).

H2O2 is also involved in the activation of mitogen-activated

734 Plant Mol Biol (2006) 61:733–746

123

protein kinases (MAPKs) that modulate gene expression and

transduce cellular responses to extracellular stimuli (Desi-

kan et al. 1999; Grant et al. 2000; Kovtun et al. 2000;

Samuel et al. 2000).

In our present analysis we investigate the generation of

H2O2 upon heat stress and the effects of heat stress and

oxidative/antioxidative compounds on the primary

expression level (mRNA) of selected sHSP and APX

genes. We present evidence that at normal temperature,

H2O2 is an efficient inducer of sHSP mRNA expression.

On the other hand we show that heat stress-induced

expression of HSP can be counteracted by diphenylenei-

odonium chloride (DPI), an inhibitor of flavin-dependent

oxidases involved in superoxide radical generation, or by

ascorbate, a peroxide scavenger. Our data demonstrate that

H2O2 is an important component in heat stress-activated

gene expression that appears to be involved in HSF-acti-

vation and signalling.

Material and methods

Cell culture and growth conditions

Arabidopsis thaliana (ecotype Landsberg) cell suspension

culture was grown in MS medium (Murashige and Skoog

1962) containing basal salt mixture, 3% (w/v) sucrose,

0.5 lg/ml NAA, 0.05 lg/ml kinetin, pH 5.7. Three ml of

seven days old suspension culture was added to 97 ml of

fresh medium and cultivated at constant light (3000 lu-

men m)2) at 20�C with shaking (60 strikes min)1). Expo-

nentially growing cells (3–4 days old) were used for

experiments. Before treatments, cell density was adjusted

with fresh MS medium to OD 660 = 0.12 and pre-incu-

bated for 90 min at normal growth condition. Aliquots of

5 ml were used for all further treatments.

Cell viability staining test

To evaluate viability of cells, an aliquot of the suspension

culture was supplemented with Evans Blue Dye (Fluka,

Swiss) to a final concentration of 0.04%, incubated 10–

15 min at room temperature and monitored under micro-

scope. Dead cells are stained with the dye whereas viable

cells remain unstained.

Heat stress treatment of the cell culture

Heat stress was administered by subjecting cell culture

samples to 37�C or 44�C for 2 h in a shaking water bath in

the dark. Controls were incubated at 20�C under otherwise

identical condition. Following the treatments samples were

immediately assayed for quantifications of H2O2 levels, or

frozen in liquid nitrogen and used for isolation of mRNA or

preparation of protein extracts.

Supplementation of oxidative and antioxidative

compounds

All compounds, with the exception of DPI, used for cell

culture treatments were dissolved in MS medium and pH

was subsequently adjusted to 5.7. For DPI treatment a

20 mM stock (in DMSO) was prepared. Reduced ascor-

bate, DPI, or oxidative compounds were added to the final

concentrations as indicated and cells were incubated in a

shaking water bath (60 strokes min-1) in the dark at 20, 37

or 44�C, respectively. The following concentrations were

tested: ascorbate––5, 20, 50 mM; DPI––0.5, 10, 25 and

150 lM; H2O2––0.05, 0.5, 5 and 50 mM, tert-butylperox-

ide (BP)––0.05, 0.5 and 5 mM; diamid (DA)––0.05, 0.5

and 5 mM. All supplements were purchased from Sigma.

Cultivation of Arabidopsis plants, and stress treatments

Arabidopsis (ecotype Columbia 24) plants were cultivated

as described by Panchuk et al. (2002). Leaves of seven-

week-old plants were collected and incubated in section

incubation buffer (SIB: 1 mM potassium phosphate, pH

6.0, and 1% (w/v) sucrose) in a shaking water bath

(60 strokes min)1) in the dark at 20�C or 37�C, respec-

tively. Ascorbate, DPI, or oxidative compounds were ad-

ded to the final concentrations as indicated.

Measurement of intracellular H2O2 accumulation

Intracellular accumulation of ROS (e.g. H2O2) was moni-

tored using in vivo oxidation of carboxy-HDCFDA

(5–6-carboxy-2¢,7¢-dichlorodihydrofluorescein diacetate:

Molecular Probes) fluorescence probe (Royall and Ischir-

opoulos 1993). Five millilitre aliquots of Arabidopsis cell

culture were supplemented with carboxy-HDCFDA (end

concentration 10 lM) and incubated for 5 min at 20�C for

the preloading of cells with the fluorescence probe. Cells

were incubated in the dark in a shaking water bath

(60 strokes min)1) at 20, 37 or 44 �C. After 1 h, samples

were placed on ice and EDTA was added to final con-

centration of 0.3 mM. Cells were destroyed by sonication

(1 min, 30 W) using Sonifier B-12 (Branson Sonic Power

Company, Dunbury, Connecticut) and fluorescence was

measured using fluorescence spectrophotometer F 2000

(Hitachi) at excitation and emission wavelengths of 503

and 525 nm respectively. Three replicates of each sample

were routinely measured in parallel. Changes in the accu-

mulation of H2O2 were analyzed for statistical significance

according to t-test (Engel 1997). Intracellular character of

oxidation of fluorescence probe (Royall and Ischiropoulos

Plant Mol Biol (2006) 61:733–746 735

123

1993) was confirmed in control experiments applying cat-

alase to destroy potential extracellular H2O2. No difference

was found between samples incubated with or without

catalase. Spontaneous oxidation of carboxy-HDCFDA in

the absence of cell culture was below 1% of the levels

observed in the presence of the cells.

mRNA isolation and cDNA preparation

Poly(A)+-mRNA and cDNA were prepared as described by

Panchuk et al. (2002). The amount of poly(A)+-mRNA/

cDNA double-stranded products obtained after reverse

transcription was measured using PicoGreen dsDNA

Quantitation reagent (Molecular Probes). This method of

template quantification improved the reproducibility of data

of subsequent real-time PCR. For monitoring the degree of

potential template degradation during the preparation of

poly(A)+-mRNA/cDNA, two primer pairs spanning proxi-

mal and respectively distal parts of the AtAct2 mRNA were

used. Identical threshold cycles with both pairs of primers

indicated the integrity of mRNA/cDNA.

Primer design and PCR-product identity

Gene-specific primers for real-time RT-PCR quantification

were used as described by Panchuk et al. (2002), Volkov

et al. (2003). The resulting PCR products had the same size

of approximately 300 bp. The quality of PCR products was

visually inspected by electrophoresis; the generation of

only one single band of the expected size was taken as a

criterion for specificity. The identity of PCR products was

confirmed by direct DNA-sequencing.

Quantitative real-time RT-PCR

Quantification of gene-specific cDNA was performed by

real-time PCR monitoring the intercalation of SYBR-

Green (Molecular Probes) essentially as described by

Panchuk et al. (2002). Two concentrations of cDNA (1 ng

and 0.1 ng) were routinely measured in parallel and

duplicate samples were run for each concentration. All

experiments were repeated at least twice for cDNA pre-

pared for two samples of Arabidopsis cells. Using stan-

dardized conditions, deviations of threshold values were

less than 1.0 cycle for independent cDNA preparations and

less than 0.5 cycle for replicates of the same cDNA. The

quantification of mRNA levels is based on the comparison

to the level of an Act2 mRNA standard, defined as 100

relative expression units (REU: Panchuk et al. 2002),

which was determined in separate reactions. Changes in the

relative concentrations of PCR products/steady-state

mRNA levels were checked for statistical significance

according to t-test (Engel 1997).

Preparation of protein extracts

Crude cell extracts were generated by shock freezing cell

culture samples in liquid nitrogen, 1.2 Vol of 1.83 · LEB

(Low salt Extraction Buffer: 1 · LEB = 10 mM KCl,

3.3 mM MgCl2, 0.35 M sucrose, 8% (w/v) glycerol,

15 mM HEPES-KOH, pH 7.9, 2% Ficoll 400, 1 mM

PMSF, 1 · Proteinase Inhibitor Cocktail ‘‘Complett’’

(Roche), 1 · phosphatase inhibibitor Cocktails I and II

(Sigma)) were added, ground on ice for 3 min and centri-

fuged at 7000 g, 10 min, 4�C. The supernatant was cleared

again at 21,000 g, 20 min, 4�C, and protein concentration

was determined (Bradford 1976).

Electrophoretic mobility shift assay (EMSA)

The 5¢-ends of a synthetic double-stranded oligonucleotide

HSE probe were labeled by incubation with a-32P-dATP

and Klenow fragment (MBI Fermentas). The labelled

probe was purified with QIAquick Nucleotide Removal Kit

(Qiagen) and used for EMSA. Samples containing 20 lg

protein (10–12 ll low salt extract) were mixed with 3 Vol

of DB (Dilution Buffer: 15 mM HEPES-KOH, pH 7.9,

1.5 mM EDTA, 0.1 mM EGTA) containing excess of

unspecific DNA (per 1 probe: 5 lg of poly(dI-dC) and

100 ng of PCR product of coding region of NPT gene), and

incubated for 5 min at room temperature. Then 1 ng of

labelled HSE probe was added and samples were incubated

for 25 min at room temperature. To a total volume of 40 ll

binding reaction 2 ll of loading buffer (30% v/v glycerol,

0.2% w/v bromophenol blue) were added. Samples were

loaded onto a pre-run 5% polyacrylamide gel containing

3% v/v glycerol in 0.5ÆTBE (44.5 mM tris–boric acid pH

8.0, 1 mM EDTA) and subjected to electrophoresis for

2.5 hours, 350 V in cold room. Gels were dried on DE81

paper (Whatman Biometra) and exposed to Kodak BioMax

MS film.

Results

Heat stress increases H2O2 level in Arabidopsis

We have monitored the effect of moderate (37�C) and

severe (44�C) heat stress on the intracellular accumulation

of ROS, e.g. H2O2, in Arabidopsis cell culture. The assay

used is based on in vivo oxidation of carboxy-HDCFDA

fluorescence probe (Royall and Ischiropoulos 1993). The

viability of cells in suspension culture, grown under opti-

mal conditions, was tested. Viability staining showed that

approximately 99% of the cells were alive (supplemental

Figure 1). To ensure a good physiological state of the cells,

suspension culture was diluted with an excess of the fresh

736 Plant Mol Biol (2006) 61:733–746

123

medium and pre-cultivated prior to application of stress. To

avoid light-dependent effects the heat stress was adminis-

tered in the dark. In addition to heat stress, we investigated

also the influences of peroxide scavenger (ascorbate) and

inhibitior (DPI) of ROS production by supplementing cell

culture samples with different concentrations of ascorbate

or DPI during temperature treatments. Ascorbate and DPI

treatments had no negative effects on the viability of

Arabidopsis cells.

Figure 1A shows that heat stress causes a strong in-

crease in H2O2 levels: approximately 2.3-fold at 37�C and

2.5-fold at 44�C within 1-hour treatment. The accumula-

tion of H2O2 is a very fast process, occurring within the

first 15 min of heat stress. Longer exposure, as measured

during the first, the second, and fourth hour of continuous

heat stress, resulted in no further increase, rather in a de-

crease of H2O2 levels. DPI, supplemented at a concentra-

tion of 25 lM had a negative effect on the accumulation of

H2O2 at 20�C and especially at 37�C, but not at 44�C

(Fig. 1B). Higher DPI concentrations of up to 150 lM

resulted in only a minor further decrease of the H2O2 level

at 20�C and 37�C. Supplementation of ascorbate caused

much stronger negative effects on H2O2 levels at all tem-

peratures tested (Fig. 1B). At a concentration of 50 mM

the reduction was approximately 4-, 7- and 5-fold at 20, 37

and 44�C, respectively. The effect was dosage-dependent:

25 mM caused a larger reduction than 5 mM ascorbate.

In contrast to DPI, ascorbate was also very efficient in

reducing the H2O2 levels at normal temperature and at

44�C. These data show that heat stress enhances intracel-

lular production of H2O2, which can be blocked completely

by ascorbate supplementation.

Oxidative compounds induce heat shock

gene expression

In order to test whether at normal temperature the appli-

cation of oxidative compounds is capable of inducing heat

shock gene expression, we determined the mRNA levels of

sHSP and APX genes in Arabidopsis cell culture by

quantitative RT-PCR. Small Hsp17.6 and Hsp18.2 exem-

plify typical heat-induced and HSF-dependent expressed

genes in Arabidopsis (Volkov et al. 2003; Lohmann et al.

2004) and cytosolic ascorbate peroxidase genes Apx1 and

Apx2 have been previously identified as novel HSF-

dependent heat shock genes, which are significantly in-

duced at the mRNA level in leaf tissues upon heat stress

(Panchuk et al. 2002).

In our analysis we determined the mRNA levels of these

genes when induced by different concentrations of H2O2,

BP, or DA at 20�C and compared them with the levels in

untreated (20�C) or heat-treated (37�C or 44�C) cells

(Fig. 2). Two of the oxidative compounds used, H2O2 and

BP, are peroxides. While H2O2 is located mainly in

aqueous phase, BP has relatively high affinity to cellular

membranes. DA is a synthetic SH-groups-oxidizing com-

pound, which changes intracellular redox status and pro-

motes protein disulfide cross-linking. These differences

may result in different cellular responses after treatment.

The viability staining test revealed that 50 mM H2O2

has a slightly negative effect on cell viability: approxi-

mately 90% of the cells survived after 2 hours treatment.

Lower concentrations (5 mM or less) of the oxidative

compounds tested appeared to be non-toxic, 95–98% of

cells survived after the treatments.

In cell culture, several genes tested were induced by

moderate heat stress (37�C) whereas no induction was

Ascorbate

28 36

100

238240

948544 51

0

100

200

300

400

mM

Flu

ore

scen

ce(a

rbit

ary

un

its)

0 5 50

DPI

141

61112130

240 238

100 95

243 259

70

251275

293280

0

100

200

300

400

µM

Flu

ore

scen

ce(a

rbit

ary

un

its)

0 5 50 0 5 50

20˚C 37˚C 44˚C

20˚C 37˚C 44˚C0

0.5 25 150

DM

SO 0

0.5 25 150

DM

SO 0

0.5 25 150

DM

SO

96139

79108

278

216238240

100

152101

71

0

100

200

300

400

Flu

ore

scen

ce(a

rbit

ary

un

its)

0-15

min

0-1h

1-2h

3-4h

0-15

min

0-1h

1-2h

3-4h

0-15

min

0-1h

1-2h

3-4h

(A)

(B)

Fig. 1 Intracellular oxidation levels in Arabidopsis suspension cell

culture upon heat-stress. (A) Time-dependent changes after prolonged

heat-treatment: the mean rate of oxidation of fluorescence probe was

measured during 15 min (0–15 min), or 1 h (0–1 h) immediately after

beginning of treatment, or from 1 to 2 h (1–2 h), or from 3 to 4 h (3–

4 h); (B) Effects of application of different concentration of DPI

(dissolved in DMSO), or ascorbate, on the intracellular oxidation levels

in Arabidopsis cell incubated for 1 h at 20, 37 or 44�C. All treatments

were performed in the dark. Bars show means SD (n = 4–5)

Plant Mol Biol (2006) 61:733–746 737

123

found after severe heat treatment (44�C) similar to the

response observed in Arabidopsis leaves (Panchuk et al.

2002; Volkov et al. 2003). It should be noted that 44�C

heat stress is lethal to unconditioned Arabidopsis plants

(Wunderlich et al. 2003; Lohmann et al. 2004). The cur-

rent data on tissue culture show that at 20�C H2O2, BP, and

DA are potent inducers of mRNA expression of the heat-

inducible genes tested (Fig. 2). The magnitude of mRNA

induction by H2O2, BP, or DA, and by moderate heat stress

(37�C) was very similar for Hsp17.6 whereas for Hsp18.2

the oxidative stress-induced levels reach approximately 1/

3rd of the level induced by heat stress at 37�C. The optimal

concentrations for the induction of the two sHSP genes are

0.5 mM H2O2, 0.5–5 mM BP, and 0.5 mM DA. The same

concentrations are also optimal for the induction of Apx2

mRNA, but interestingly, the levels induced by H2O2 or BP

were approximately 3-fold higher compared to the levels

after heat stress at 37�C; DA-induced levels reach

approximately only 50% of the heat-induced levels. The

induction factors range between approximately 100-fold

for Hsp17.6, 300–1000-fold for Hsp18.2 and 10–25-fold

for Apx2.

The expression profile of Apx1 differed from the pat-

terns of the three other genes by: (i) A relatively high basal

level of mRNA present under non-stress condition, (ii) a

moderate stimulating effect (maximum about 3-fold

induction) by application of oxidative stress compounds,

and only a weak stimulating effect (about 1.5-fold induc-

tion) by heat stress at 37�C, (iii) a wider range of effective

concentrations of all three different compounds. In com-

mon with Apx2 were the higher levels of oxidative stress-

induced mRNA of Apx1 compared to heat stress. Among

other members of APX gene family, only Apx4 was

induced by heat stress and 3 genes, Apx4, Apx6 and tApx

were induced by application of oxidative compounds

(supplemental Figure 2). The maximal induction was ob-

tained by application of 5 mM H2O2, which was one order

of magnitude higher than the optimal concentration for the

induction of Apx1, Apx2 and both sHSP genes. Interest-

ingly, no significant changes were detected for mRNA

levels of microsomal Apx3 and Apx5.

These data show that oxidative compounds, in particular

H2O2, cause an efficient increase in transcript levels of

sHSP genes. The transcript levels reached, as exemplified

by Hsp17.6 and Hsp18.2, approximately the same levels as

induced by heat stress. The mRNA expression of Apx1 and

Apx2 is more efficiently induced by oxidative stress com-

pared to heat stress.

H2O2 is required for efficient heat induction

of mRNA levels

If an increase in H2O2 level is required for heat stress-

activated gene expression, it could be expected that inhi-

bition of production or scavenging of ROS should exert

profound negative effects on heat-induced mRNA levels.

Therefore we tested the influence of DPI and ascorbate on

mRNA levels of Hsp17.6, Hsp18.2, Apx1, and Apx2 during

heat stress. Figure 3 shows that ascorbate supplementations

of 5 and 20 mM had a moderate effect on Hsp17.6,

Hsp18.2, 50 mM a profound negative effect on the heat

stress-induced mRNA levels of Hsp17.6, Hsp18.2 and

Apx2. 50 mM ascorbate reduced the mRNA levels of

Hsp17.6, Hsp18.2 and Apx2 by factors of 23 and 28,

respectively. There was no negative effect of 50 mM

ascorbate treatment on the basal mRNA levels of all tested

genes: the levels increased slightly by a factor of less than

2, when cells were treated at 20�C.

5.9 5.64.3

40

1.96.5

110

3.8 3.3

28

10571

24

0

50

100

150

200

mR

NA

leve

l (R

EU

)

24

2640

15 38

1890

29 9.2288

27003100

46

2600

110

1000

2000

3000

4000

mR

NA

leve

l (R

EU

)403

272250

340290

628800

590

230

630541491

640

0

250

500

750

1000

mR

NA

leve

l (R

EU

)

3.21.1

920

0.8 0.6

370

6.12

10

320150

0.8

340

0

300

600

900

1200

mR

NA

leve

l (R

EU

)Hsp 17.6

Apx 1

Hsp 18.2

Apx 2

T,(˚C) 20˚37˚44˚ 20˚ 20˚ 20˚

0.05 0.5 5 50

0.05 0.5 5

0.05 0.5 5Treatment

+BP,mM +DA,mM

T,(˚C) 20˚37˚44˚ 20˚ 20˚ 20˚

0.05 0.5 5 50

0.05 0.5 5

0.05 0.5 5Treatment

+ H2O2,mM+ H2O2,mM + BP,mM + DA,mM

Fig. 2 Messenger RNA levels

of Hsp17.6, Hsp18.2, Apx1 and

Apx2 in Arabidopsis cells after

different stress treatments. All

treatments were performed in

the dark. Poly(A)+-mRNA

levels were quantified by real-

time RT-PCR. Expression levels

are represented in comparison to

the expression of actin2 mRNA

standard, which was defined as

100 relative expression units

(REU). Bars show means SD

(n = 4–6). Note that different

scales are used in graphs. BP,

tertbutylperoxide; DA, diamide

738 Plant Mol Biol (2006) 61:733–746

123

DPI treatment resulted also in a suppression of heat

stress-induced mRNA levels of Hsp17.6, Hsp18.2 and

Apx2. 25 lM had only little effect, whereas 150 lM DPI

caused 5.5-, 6.6-, and 3-fold reduction of heat-induced

mRNA levels of Hsp17.6, Hsp18.2 and Apx2. There was no

negative effect of DPI on basal gene expression of cells

incubated at 20�C or DMSO-treated cells (control of the

DPI solvent) after heat stress. The heat stress-dependent

mRNA expression profile of Apx1 differs from the patterns

of the other genes by: (i) Minor but statistically insignifi-

cant negative effects after ascorbate treatment, (ii) only a

moderate reduction (about 50%) of expression levels after

DPI treatment. These data demonstrate that ROS, espe-

cially H2O2, are necessary for heat-induced gene expres-

sion of sHSP genes and Apx2.

Regarding to the possibility that the effects observed for

the suspension culture may differ from that in planta, we

have tested changes of mRNA levels in leaves of seven-

weak-old Arabidopsis plants after incubation in dark for

2 h in SIB (i) at room temperature in the presence of

5 mM, 50 mM H2O2 or 5 mM BP and (ii) at 37�C in the

presence of 50 mM ascorbate (supplemental Figure 3) in

comparison to the respective controls. In leaves, similar to

the cell culture, H2O2 and BP markedly induced expression

of Hsp17.6, Hsp18.2, Apx1 and Apx2, by factors of maxi-

mal 633, 125, 3.2 and 86, respectively. However, in con-

trast to the cell culture, Hsp17.6, Hsp18.2 and Apx2 mRNA

levels induced by heat stress at 37�C were respectively 3.0-

fold, 60-fold and 2.7-fold higher than the maximal levels

induced by H2O2 or BP. Comparing with cell culture, a

much higher concentration of H2O2 (50 mM) was neces-

sary to achieve the maximal induction, and BP was a more

effective inducer than H2O2 at 5 mM concentration. Sup-

plementation of the incubation medium with 50 mM

ascorbate during heat stress of leaves at 37�C reduced the

heat-inducible mRNA levels of Hsp17.6, Hsp18.2 and

Apx2 by factors of 1.5, 1.9 and 2.8, respectively. This

shows that the effect of ascorbate is less pronounced in

leaves than in cell culture. These quantitative differences in

responses between Arabidopsis cell culture and leaves may

reflect the better penetration of compounds in the suspen-

sion culture cells and probably tissue-specific differences

in leaves. Taken together the data show that the Arabid-

opsis suspension culture cells, as compared to leaves,

exhibit similar but more pronounced effects (changes at

mRNA of target genes) and are therefore a useful model

system for investigating stress signalling.

Oxidative stress-induced HSF-DNA binding complexes

The common induction of gene expression by heat stress

and oxidative stress raises the possibility of common sig-

nalling pathways. The analysis of HSF knock out mutants

provided evidence that at least two different HSF, HSF-

A1a and HSF-A1b (originally designated HSF1 and HSF3

by Hubel and Schoffl et al. 1994; Prandl et al. 1998), are

involved in the control of Hsp17.6 and Hsp18.2 tran-

scription during the initial phase of the heat shock response

(Wunderlich et al. 2003; Lohmann et al. 2004). This early

phase is characterized by the formation of high molecular

weight DNA-protein binding complexes, which were

identified by electrophoretic mobility shift analyses

(EMSA). Therefore, we examined the potential of oxida-

tive compounds to induce the formation of heat stress-

specific HSF complexes with a double-stranded HSE probe

(Fig. 4A) containing six copies of active pentanucleotide

HSE-modules (Scharf et al. 2001). Figure 4B shows that

in H2O2-treated cells similar high molecular weight

250

340

286 202 246 238339

280173 187

0

250

500

750

1000

mR

NA

leve

l(R

EU

)

4.3

40

7.6

35 311.4 2.9

3213

39

0

50

100

150

200

mR

NA

leve

l(R

EU

)

24

2640

45

14001500

11312

1650

478

2080

0

1000

2000

3000

4000

mR

NA

leve

l (R

EU

)

1.1

920

1.5

500600

400.7

550

140

0

300

600

900

1200

mR

NA

leve

l (R

EU

)Hsp 17.6

Apx 1

Hsp 18.2

Apx 2

T,˚C 20 37˚ 20 37˚ 37 37˚ 20 37˚ 37 37˚

50 5 20 50 25 25 150

+ Ascorbate,mM +DPI,µM

DM

SO

Treatment

T,˚C 20˚ 37˚ 20˚ 37˚ 37˚ 37˚ 20˚ 37˚ 37˚ 37˚

50 5 20 50 25 25 150

+ Ascorbate,mM +DPI, µM

DM

SO

Treatment

990

Fig. 3 Messenger RNA levels

of Hsp17.6, Hsp18.2, Apx1 and

Apx2 in Arabidopsis cells after

heat treatment in the presence of

different concentration of

ascorbate, DPI or DMSO

(solvent control for DPI). All

treatments were performed in

the dark. Poly(A)+-mRNA

levels were quantified by real-

time RT-PCR. Expression levels

are represented in comparison to

the expression of actin2 mRNA

standard, which was defined as

100 relative expression units

(REU). Bars show means SD

(n = 4–6). Note that different

scales are used in graphs

Plant Mol Biol (2006) 61:733–746 739

123

complexes were formed at 20�C as compared to heat-

treated (37�C, 15 min) cells. Ascorbate but not DPI treat-

ment blocked the formation of the heat stress-induced

complex. Interestingly, light grown cells, when assayed

immediately after harvest, show also a high molecular

weight HSE-binding complex that is typical for heat

stressed cells, but which is not detectable anymore after

15 min incubation in the dark. Similar to H2O2, application

of BP also results in formation of high molecular complex

at 20�C, whereas DA treatment was unable to induce this

complex; the HSE-binding complex was also not induced

at 44�C (data not shown).

Since it was known that longer periods of heat stress

cause a change in the pattern of HSE-protein binding

complexes in leaves (Lohmann et al. 2004), we tested also

cells treated for 15 min, 1 and 2 h in the dark. It should be

noted that the components forming the late higher mobility

HSF-DNA binding complex are unknown. Evidently HSF-

A1a and –A1b, which have been identified as early

response regulators, are not part of the late complex. After

2 h (Fig. 4B and C) the high molecular weight complexes

that are typical for the initial phase of the heat shock

response have disappeared or were reduced in all lanes

except for light grown cells. An other HSE-binding com-

plex of higher mobility, which is typical for the later stages

of the heat shock response (Lohmann et al. 2004), was

observed in cells treated for 2 h at 37�C, but it was not

formed in H2O2-treated cells at 20�C, or heat treated cells

incubated in the presence of 50 mM ascorbate or 150 lM

DPI. Application of different concentrations demonstrated

that 10 lM DPI moderately and 25 lM DPI significantly

reduced intensity of the higher mobility complex (Fig. 4C).

Considering a dose-dependent inhibition of intracellular

production of H2O2 by DPI (see Fig. 1) the effects on HSF-

binding complex formation correlate well with the intra-

cellular H2O2 levels.

Thus, our experiments indicate that peroxides (H2O2 and

BP) stimulate the formation of high molecular weight

HSE-binding complexes in a similar way as heat stress.

Interestingly, formation of HSE-binding complexes was

not found in DA-treated cells, although this compound

effectively induces expression of heat stress genes. In

contrast to heat stress, the formation of the ‘‘late’’ high

mobility HSE-binding complex (after 2 h) is not initiated

in H2O2-treated cells and prevented in heat-stressed cells

by simultaneous treatment with ascorbate or DPI. This

indicates that H2O2 is involved only in the early stages of

HSF activation. Upon moderate heat-stress H2O2 appears

to be essential for the induction of heat stress response but

by exogenous application it seems to either prolong the

initial phase or an additional signal (e.g. denaturation of

proteins upon heat stress) may be necessary for the late

phase of induction.

Discussion

Arabidopsis cell culture as a model system

Our analysis provides evidence that oxidative stress has a

profound effect on the heat-stress-dependent induction of

HSF target genes in plant cells. The experiments were

conducted with Arabidopsis thaliana suspension culture

cells, which are, with respect to the application of inducers

and scavengers of ROS, particularly hydrogen peroxide,

convenient compared to whole plant or organs. Although

cell culture was used by several authors (Desikan et al.

(A)

(B)

(C)

Fig. 4 Stress-dependent increase in the DNA-binding activity of

HSF. (A) Oligonucleotide probe used for gel mobility shift assay,

HSE modules are underlined. (B) Gel mobility shift assay with

protein extracts prepared from Arabidopsis cells incubated at 20 or

37�C in the presence or absence of 0.5 mM hydrogen peroxide,

50 mM ascorbate or 150 lM DPI. (C) Effect of different concentra-

tions of DPI on the induction of DNA-HSF complexes. The stress

inducible DNA-HSF complexes specific for early and late phases of

stress response are indicated by black and open arrows, respectively;

constitutive DNA-HSF complexes are indicated by asterisks. All

treatments were performed in the dark except the light control

treatment at 20�C

740 Plant Mol Biol (2006) 61:733–746

123

1996; 1998; 1999; Clarke et al. 2000; Vacca et al. 2004)

for investigation of plant cell stress response, the question

arose wether the effects observed really reflect the situation

in planta. In our experiments we used exponentially

growing Arabidopsis cell culture, which represents a sus-

pension of microcolonies of chlorophyll-containing callus

cells originally obtained from leaves. Our data show that

genes investigated in cell culture respond to a heat stress in

a similar way as in leaves by (i) induction of mRNA of

known heat-inducible genes upon moderate heat-treatment

at 37�C, (ii) no induction at 44�C, (iii) rapid formation of a

high molecular weight HSE-binding complex during early

phase of heat shock response and the dynamic changes in

the pattern of binding complexes during the later phase

(after 2 h) of the response. Hence, regulation of the heat

stress response appears to be similar in cell culture and

leaves. Accordingly, the cell culture represents a conve-

nient model system for studies of stress signalling. On the

other hand, quantitative differences between leaves and

cell culture in the expression of stress genes and the

presence of HSE-binding complexes in light growing cells

at 20�C indicate tissue specificity of stress response.

In order to reduce the complexity of signalling inter-

ference between light and heat stress we have routinely

used heat stress treatment in the dark for studying the

expression of heat shock and HSF-dependent target genes

in Arabidopsis (Lohmann et al. 2004, Busch et al. 2005).

With this experimental design we intended to exclude the

effects of photo-oxidative stress generated by heat stress in

chloroplasts.

Possible source of heat-induced H2O2

We have shown that heat stress generates enhanced levels

of H2O2 in tissue culture cells. The data indicate that the

level rises very rapidly within the first 15 min with a

subsequent decline during longer exposure to heat stress.

This response is reminiscent of the oxidative burst occur-

ring after pathogen attack (Desikan et al. 1996; Clarke

et al. 2000).

Our experiments demonstrate that in Arabidopsis cells

the level of H2O2 significantly increases, 2.3–2.5 fold, both

after moderate (37�C) and respectively severe (44�C) heat

stress. Similar, in mustard seedlings subjected to severe

heat stress (55�C, 1.5 h) in the dark the level of endoge-

nous H2O2 increased by 65% in comparison with plants

grown at 24�C (Dat et al. 1998). Severe heat stress (at 60–

65�C) that induces programmed cell death in tobacco tissue

culture cells leads to much higher and sustained levels of

H2O2 (Vacca et al. 2004). There are different possibilities

for the generation of H2O2 within plant cells. In the

absence of light, mitochondria may be the source of

intracellular generation of ROS, in particular the proton-

pumping complexes I and III (CI, NADH dehydrogenase,

and CIII, ubiquinol-cytochrome bc1 reductase) located in

the inner membrane (Moller 2001), and it was proposed

that ROS could be also generated by a plant-specific non-

pumping internal NADPH dehydrogenase, NDin(NADPH)

(Moller 2001). Heat treatment at 41�C increases oxygen

respiration rate and results in excessive production of ROS

in yeast mitochondria (Sugiyama et al. 2000). The main

sites of ROS production appear to be external NADH de-

hydrogenases, NDE1 and NDE2 (Davidson and Schiestl

2001). These two proteins functionally substitute the pro-

ton-pumping complex CI present in mitochondria of the

majority of eukaryots (Moller 2001).

In order to identify possible ROS generating mecha-

nisms in Arabidopsis cells upon heat stress in the dark, we

tested the effect of DPI, an inhibitor of flavoenzymes

(O’Donnell et al. 1994). It has been demonstrated that DPI

inhibits NDin(NADPH), CI and NDin(NADH) activities in

plant mitochondria with a Ki of 0.17, 3.7, and 63 lM,

respectively (Agius et al. 1998). In the plasma membrane

of soybean DPI inhibits NADPH oxidase with a Ki of 0.1

lM, whereas activity of NADH oxidase was only slightly

affected even by application of 100 lM DPI (Morre 2002).

Our data show that the application of 0.5 lM DPI, which

should extensively inhibit NADPH-oxidases, had no effect

on the intracellular H2O2 level. However, the negative ef-

fect on H2O2 levels, generated by 25 lM DPI, indicates

that CI could be the main site of ROS production at 37�C.

However, it should be noted, that intracellular concentra-

tion may be higher than that of exogenous DPI in the

medium (Moller 2001). Therefore, it cannot entirely be

excluded that NADH oxidase may be partially inhibited by

application of 25 lM DPI. Further increase in DPI con-

centrations up to 150 lM had only little effect on further

decrease of H2O2 levels, which may indicate that NADH-

oxidases play a minor role in ROS generation. However,

other enzymes (e.g. mitochondrial CIII) seem to participate

in the generation of H2O2, which is indicated by the fact

that a large fraction (58%) of H2O2 generated at 37�C is

DPI-insensitive and the heat-induced H2O2 levels gener-

ated at 44�C heat stress are completely unaffected by DPI.

Taking into account that the intracellular level of H2O2

represents a balance between production and elimination,

the heat inactivation of scavenging enzymes may be in-

volved in changing the steady state levels of H2O2 during

heat stress. It was shown that in wild type Arabidopsis

leaves the activity of APX was not changed following

treatment at 37�C, but was severely compromised at 44�C

(Panchuk et al. 2002). Superoxide dismutase and glutathi-

one reductase activities remained unchanged under these

conditions.

Plant Mol Biol (2006) 61:733–746 741

123

H2O2-dependent expression of heat-inducible genes

Our experiments show that externally added H2O2 and

other oxidative compounds are effective inducers of heat-

inducible genes, Hsp17.6, Hsp18.2, Apx1, and Apx2. The

effective concentration of H2O2 was 0.5 mM. This con-

centration is similar to the one (0.2 mM) being effective in

driving the induction of an AtHsp18.2 promoter–LUC re-

porter construct in Arabidopsis protoplasts (Kovtun et al.

2000). However, the effective concentration of H2O2 at its

cellular site is unknown, but probably much lower than the

externally applied concentration. It has been shown that the

half-life of exogenous 20 mM H2O2 is 2 min, because

Arabidopsis cultures have a high scavenging capacity

(Desikan et al. 1998, 2001). The effect of scavenging is

strongly dependent on cell growth and experimental con-

ditions. Cell cultures used in our experiments had the

capacity to reduce 0.5 mM exogenous H2O2 within 1 hour

to approximately 10%, 5 mM to about 30% (see supple-

mental Figure 4).

The induction of heat stress genes by oxidative stress

has been previously reported, however, the mRNA levels

have not been compared with those after induction by heat

stress. We have used real-time PCR quantification of

mRNA levels of Hsp17.6, Hsp18.2, Apx1 and Apx2. The

sHSP genes and Apx2, which are practically not expressed

in unstressed cells, are strongly induced to comparable

levels by moderate heat stress (37�C) or by application of

H2O2 (0.5 mM). It was previously shown that in Arabid-

opsis leaves the mRNA levels of heat-inducible sHSP and

Apx2 reach the maximum after 1–2 h treatment at 37�C

with a decline thereafter (Panchuk et al. 2002; Volkov

et al. 2003). This fast accumulation and transient expres-

sion is a signature of HSF-regulated heat shock genes

(Panchuk et al. 2002; Lohmann et al. 2004). The Apx1

expression profile is different. It shows a significant

expression at normal temperature, its mRNA level after

heat stress is only about 2-fold increased after heat stress

(Panchuk et al. 2002), and Apx1 shows a strong induction

by H2O2, which is also demonstrated for Apx2 (this paper).

Previously, photo-oxidative stress-induced Apx2 expres-

sion restricted to bundle sheath cells was found in Ara-

bidopsis leaves (Fryer et al. 2003).

H2O2 exerts a pivotal role in plant life, it is widely

recognized as a key signalling compound that can mediate

cross tolerance in plants towards other stresses (Bolwell

1999; Bowler and Fluhr 2000), but it is also a toxic com-

pound that causes detrimental effects and cell death in

plants. It is known that heat stress stimulates the accumu-

lation of H2O2 in plant cells (Foyer et al. 1997; Dat et al.

1998; Vacca et al. 2004). Conversely, transcriptome anal-

ysis of Arabidopsis tissue culture cells subjected to H2O2

revealed the induction of a number of genes with roles in

biotic and abiotic stress responses, including also several

genes encoding HSP and HSF (Desikan et al. 2001). The

expression of sHSP genes following application of H2O2

was also reported for tomato (Banzet et al. 1998) and rice

(Lee et al. 2000). Similar, upon high light stress, which

results in excessive production of H2O2, representatives of

all HSP gene families, e.g. HSP101, HSP90, HSP81,

HSP70 and sHSP, were induced in wild type Arabidopsis

(Rossel et al. 2002) and further upregulated upon high light

in Apx1 (Pnueli et al. 2003) and catalase (Vandenabele

et al. 2003) deficient mutants. It was proposed that upon

oxidative stress chaperone function of HSP may be nec-

essary to limit oxidation-mediated disulfide bridge-induced

protein aggregation (Rossel et al. 2002).

Our data show that heat treatment at 37�C has an effect

on gene expression, similar to exogenous 0.5 mM H2O2. It

was shown that this concentration did not affect viability of

Arabidopsis suspension culture cells. These data are con-

sistent with the observation that the induction of Arabid-

opsis cell death requires H2O2 concentration of more than

5 mM (Neil et al. 1999). Thus the intracellular H2O2 levels

induced by moderate heat treatment seems to be not toxic

but may play a regulatory role in the cellular network that

leads to altered gene expression the adaptation of cells to

different biotic and abiotic stresses.

Heat and oxidative stress signalling

In order to test whether oxidative stress and heat stress

induction share common components in signalling stress

responses, we have studied the influence of inhibitors and

scavengers of H2O2 production on heat-induced levels of

mRNAs. Both DPI and ascorbate exert profound negative

effects on the mRNA induction of Hsp17.6, Hsp18.2, and

Apx2. By contrast the mRNA levels of Apx1 are compro-

mised to a much lower extent. This result is a clear indi-

cation for the involvement of H2O2 in the heat stress and

HSF-dependent expression of typical heat shock genes.

Ascorbate scavenging appears to be much more potent in

blocking heat-induced expression of genes compared to

DPI (Fig. 3). This is in accordance with the higher levels of

H2O2 present after DPI treatment, which is in contrast to

ascorbate, unable to completely block heat-induced gen-

eration of H2O2 (Fig. 1B). It was also demonstrated that

programmed cell death (PCD), triggered by heat stress in

tobacco tissue culture cells, is prevented by antioxidants,

also by ascorbate (Vacca et al. 2004). This phenomenon is

linked to rapid ROS production in PCD cells, which show

an early inhibition of glucose oxidation that was accom-

panied by a strong impairment of mitochondrial function.

In our experiments application of 25–150 lM DPI or

5 mM ascorbate at 37�C reduces intracellular H2O2 to the

level in control untreated cells at 20�C. This correlates with

742 Plant Mol Biol (2006) 61:733–746

123

a decrease of Hsp17.6, Hsp18.2, and Apx2 mRNA level but

it remains considerably higher than in control cells. Only

50 mM ascorbate (which results in H2O2 level of 36%

comparing to untreated control at 20�C––see Fig. 1) is able

to reduce Apx2 expression to the control level, but mRNA

levels of both sHSP genes remain still increased. To ex-

plain the data two possibilities should be considered: (i) An

additional H2O2-independent stress signalling pathway is

active in heat-treated cells, resulting in an induction of

sHSP genes in the presence of 50 mM ascorbate; (ii) heat

shock causes a ‘‘sensitization’’ of H2O2-dependent sig-

nalling pathway allowing expression of heat shock genes in

cells with reduced H2O2 levels.

It is well known that the heat-inducible binding of

transcription factor HSF to the HSE promoter sequences

control heat stress-dependent expression of HSP genes.

According to the chaperone titration model (Morimoto

1998; Schoffl et al. 1998) in the majority of eukaryotes,

HSF are located in the cytoplasm of unstressed cells in an

inactive monomeric form as a complex with HSP70/HSP90

and probably some other proteins. Upon stress, dissociation

of these complexes and activation/trimerisation of HSF

occurs, followed by relocation in the nucleus. However,

little is known about exact molecular mechanisms of HSF

activation in plants. The question arose whether oxidative

stress (e.g. H2O2) or changes in the redox status affect HSF

activation, which is required for initiating the transcription

of target genes. Using EMSA for the identification of HSF-

HSE binding complexes we have shown that both, heat

stress and oxidative stress resulted in the formation of high

molecular weight complexes (Fig. 4), a signature of early

HSFA1a/A1b-dependent gene expression in heat-stressed

leave tissue of Arabidopsis (Lohmann et al. 2004). Fur-

thermore, HSF binding to HSE was prevented if the heat

treatment was performed in the presence of ascorbate but,

interestingly, not when supplemented with DPI (Fig. 4).

The DPI insensitivity of HSE-complex formation probably

reflects the fact that heat stress-induced H2O2 production is

not completely suppressed by DPI. Upon heat treatment at

37�C in the dark the intensity of high molecular weight

complex was practically the same both in the absence of

DPI (increased H2O2 level) and in the presence of 150 lM

DPI (H2O2 level as in control cells at 20�C), whereas no

high molecular weight complex was found at 20�C. This

indicates that upon moderate heat shock the activation of

HSF may occur in the absence of increased of H2O2 levels

(as shown after DPI treatment), although, activation of

HSF-HSE binding appears H2O2-dependent (as shown after

ascorbate treatment). Accordingly, HSF seems to be more

susceptible to H2O2-dependent activation upon heat treat-

ment as at normal temperature. It seems possible that

heat shock per se and H2O2 cooperate in HSF activation

by dissociating cytoplasmic HSF-chaperone complexes,

leading to HSF trimerization, DNA-binding, and tran-

scriptional activation. Such a cooperation would explain a

‘‘sensitization’’ of H2O2-dependent signalling that leads to

heat shock gene expression after DPI treatment at 37�C.

Interestingly, the high molecular weight HSE binding

complexes were also found in untreated light grown tissue

culture cells but rapidly disappeared when cells were

incubated in the dark. This suggests that light-dependent

ROS may activate HSF-binding, that can be rapidly re-

versed in the dark. Whether the light-dependent complex is

functional in transcriptional activation of target genes is

unknown, it may perhaps play a role in the low level basal

expression of genes.

Discrepancies between the HSF-DNA binding complex

formation and only low mRNA levels of target genes, e.g.

observed after heat stress and DPI treatment (Fig. 4), may

indicate that DNA binding and transcriptional activation

are two separate processes. This two step process is a well

known phenomenon in mammalian cells (Hensold et al.

1990; Jurivich et al. 1992) and probably also occur in plant

cells.

What is the mechanism of H2O2 in heat stress signalling

and heat shock gene expression? Our data suggest that

oxidative stress is required for effective transcription of

stress genes, which correlates with the induction of HSE-

binding activity during the early phase of the heat shock

response. It has been shown that in Drosophila and human

cells H2O2 is a potent activator of HSF trimerization and

consequently DNA-binding (Zhong et al. 1998; Ahn and

Thiele 2003). Recombinant human HSF1, but not HSF2,

has the capacity to directly sense heat and oxidative stress

in vitro (Ahn and Thiele 2003). Our data indicate that

Arabidopsis HSF may be also a subject of oxidative stress

activation. The most suitable candidates for H2O2 activa-

tion would be AtHSFA1a or AtHSFA1b, which are in-

volved in the formation of the high molecular weight HSE-

binding complexes and which seem to be able to func-

tionally replace each other in Arabidopsis (Lohmann et al.

2004). The involvement of H2O2 and HSF is further

implicated by the negative effect of a transdominant neg-

ative HSF mutant on the expression of Apx1 in Arabidospis

(Davletova et al. 2005). Alternatively, other components

(e.g., chaperones, composition yet unknown in plants) of

the inactive HSF complex, present under non-stress con-

ditions, may be targets for inactivation by oxidative stress.

Besides, other cellular proteins may be damaged by H2O2,

which will require chaperones for repair. According to the

chaperone titration model a withdrawal of chaperones from

HSF complexes, induced by a higher load of denatured

proteins, would result in trimerization and activation of

HSF (Zou et al. 1998). Hence, H2O2 generated upon heat

stress may both directly and indirectly contribute to the

activation of HSF.

Plant Mol Biol (2006) 61:733–746 743

123

Our data show that H2O2 plays an important role during

the early phase of heat shock response (up to 2 h) being

involved in activation of HSF–HSE binding and tran-

scription of heat shock genes. In contrast, the high mobility

HSE binding complex, characteristic for the later phase of

heat shock response, was not induced by application of

H2O2, BP, or DA at room temperature. Also, mRNA of

heat shock genes induced by application of H2O2 was not

effectively translated as shown by Western analysis using

antibody directed against sHSP of Arabidopsis (R.A. Vol-

kov, I.I. Panchuk, F. Schoffl, unpublished results). Never-

theless, H2O2 is required for the activation of the later

phase of heat shock response because application of

ascorbate or DPI at 37�C in concentration dependent

manner suppressed induction of the late high mobility HSE

binding complex. Hence, a combination of H2O2-depen-

dent and independent mechanisms controls the later phase.

H2O2- and HSF-independent components

of the heat shock response

Heat-inducible expression of sHSP genes, occurring in the

presence of 50 mM ascorbate, seems to be not only H2O2-

independent, but also HSF-independent because high

molecular weight HSE binding complex was not induced by

this treatment. Similar, the complex was not detected after

the treatment with DA, although, the expression of heat

shock genes was induced. It was demonstrated that DA

treatment promotes the formation of the monomer oxidised

form of human HSF1 (Manalo and Liu 2001). This intra-

molecular disulfide cross-linked conformer was resistant to

the in vitro heat-induced trimerisation and activation. The

presumptive involvement of other transcription factors in

the transcription of HSP genes is also implicated by the data

that in the hsfA1a/b double knock out mutants of Arabid-

opsis, which are unable to form high molecular weight HSE-

binding complexes, mRNA accumulation of HSF target

genes is significantly but not completely impaired upon heat

stress (Lohmann et al. 2004).

Still, little is known about transcription factors modu-

lating oxidative stress response in plants. Besides HSF,

WRKY and bZIP proteins were proposed as possible can-

didates (Vranova et al. 2002). In Arabidopsis, the zinc

finger protein Zat12 appears to be involved in Apx1

expression (Rizhsky et al. 2004). The finding that light

stress-induced Apx1 and Zat12 transcript accumulation is

inhibited in plants expressing a dominant-negative HSF21

construct suggest that HSF function is required upstream of

Zat12 and hence, that HSF function is required at a rela-

tively early stage of the oxidative stress signalling and

acclimation response (Davletova et al. 2005). Further

analysis will be required to determine the potential of

AtHSF-A1a and AtHSF-A1b, which at present are the most

likely candidates that may sense and integrate heat and/or

oxidative stress and trigger gene expression in the envi-

ronmental stress responses in Arabidopsis. The roles of

other transcription factors and alternative mechanisms

acting at transcriptional/posttranscriptional levels will be

the target of future investigations

Acknowledgements Part of this work was supported by SFB 446

funded by the Deutsche Forschungsgemeinschaft.

References

Agius SC, Bykova NV, Igamberdiev AU, Moller IM (1998) The

internal rotenone-insensitive NADPH dehydrogenase contributes

to malate oxidation by potato tuber and pea leaf mitochondria.

Physiol Plant 104:329–336

Ahn SG, Thiele DJ (2003) Redox regulation of mammalian heat

shock factor 1 is essential for Hsp gene activation and protection

from stress. Genes Development 17:516–528

Allen RG, Tresini M (2000) Oxidative stress and gene regulation.

Free Radic Biol Med 28:463–499

Asada K (1999) The water–water cycle in chloroplasts: scavenging of

active oxygens and dissipation of excess photons. Annu Rev

Plant Physiol Plant Mol Biol 50:601–639

Banzet N, Richaud C, Deveaux Y, Kazmaier M, Gagnon J, Trianta-

phylides C (1998) Accumulation of small heat shock proteins,

including mitochondrial HSP22, induced by oxidative stress and

adaptive response in tomato cells. Plant J 13:519–527

Bolwell GP (1999) Role of active oxygen species and NO in plant

defense responses. Curr Opin Plant Biol 2:287–294

Bolwell GP, Wojtaszek P (1997) Mechanisms for the generation of

reactive oxygen species in plant defense: a broad perspective.

Physiol Mol Plant Pathol 51:347–366

Boston RS, Viitanen PV, Vierling E (1996) Molecular chaperons and

protein folding in plants. Plant Mol Biol 32:191–222

Bowler C, Van Montagu M, Inze D (1992) Superoxide dismutases

and stress tolerance. Annu Rev Plant Physiol Plant Mol Biol

43:83–116

Bowler C, Fluhr R (2000) The role of calcium and activated oxygens

as signals for controlling cross-tolerance. Trends Plant Sci

5:241–245

Bradford MM (1976) A rapid and sensitive method for the quantifi-

cation of microgram quantities of protein utilizing the principle

of protein-dye binding. Anal Biochem 72:248–254

Busch W, Wunderlich M, Schoffl F (2005) Identification of novel

Heat Shock Factor dependent genes and biochemical pathways

in Arabidopsis thaliana. Plant J 41:1–14

Chen Z, Silva H, Klessig RF (1993) Active oxygen species in the

induction of plant systemic acquired resistance by SA. Science

262:1883–1886

Clarke A, Desikan R, Hurst RD, Hancock JT, Neill SJ (2000) NO way

back: nitric oxide and programmed cell death in Arabidopsis

thaliana suspension cultures. Plant J 24:667–677

Dat JF, Foyer CH, Scott IM (1998) Changes in salicylic acid and

antioxidants during induction of thermotolerance in mustard

seedlings. Plant Physiol 118:1455–1461

Dat J, Vandenbeele S, Vranova E, Van Montagu M, Inze D, Van

Breusegm F (2000) Dual action of the active oxygen species

during plant stress responses. Cell Mol Life Sci 57:779–795

Davidson JF, Schiestl RH (2001) Mitochondrial respiratory electron

carriers are involved in oxidative stress during heat stress in

Saccharomyces cerevisiae. Mol Cell Biol 21:8483–8489

744 Plant Mol Biol (2006) 61:733–746

123

Davletova S, Rizhsky L, Liang H, Shengqiang Z, Oliver DJ, Coutu J,

Shulaev V, Schlauch K, Mittler R (2005) Cytosolic ascorbate

peroxidase 1 is a central component of the reactive oxygen gene

network of Arabidopsis. Plant Cell 17:268–281

Desikan R, Clarke A, Hancock JT, Neill SJ (1999) H2O2 activates a

MAP kinase-like enzyme in Arabidopsis thaliana suspension

cultures. J Exp Bot 50:1863–1866

Desikan R, Burnett EC, Hancock JT, Neill SJ (1998) Harpin and

hydrogen peroxide induce the expression of a homologue of

gp91-phox in Arabidopsis thaliana suspension cultures. J Exp

Bot 49:1767–1771

Desikan R, Hancock JT, Coffey MJ, Neill SJ (1996) Generation of

active oxygen in elicited cells of Arabidopsis thaliana is medi-

ated by a NADPH oxidase-like enzyme. FEBS Lett 382:213–217

Desikan R, Mackerness SAH, Hancock JT, Neill SJ (2001) Regula-

tion of the Arabidopsis transcriptome by oxidative stress. Plant

Physiol 127:159–172

Engel J (1997) Signifikante Schule der schlichten Statistik, Filander

Verlag, Furth

Finkel T, Holbrook NJ (2000) Oxidants, oxidative stress and the

biology of ageing. Nature 408:239–247

Foyer CH, Lopez-Delgado H, Dat JF, Scott IM (1997) Hydrogen

peroxide- and glutathione-associated mechanisms of acclimatory

stress tolerance and signaling. Plant Physiol 100:241–254

Fryer MJ, Ball L, Oxborough K, Karpinski S, Mullineaux PM, Baker

NR (2003) Control of Ascorbate Peroxidase 2 expression by

hydrogen peroxide and leaf water status during excess light

stress reveals a functional organisation of Arabidopsis leaves.

Plant J 33:691–705

Grant JJ, Yun B.-W, Loake GJ (2000) Oxidative burst and cognate

redox signaling reported by luciferase imaging: identification of

a signal network that functions independently of ethylene, SA

and Me-JA but is dependent on MAPKK activity. Plant J

24:569–582

Hensold JO, Hunt CR, Calderwood SK, Housman DE, Kingston RE

(1990) DNA binding of heat shock factor to the heat shock

element is insufficient for transcriptional activation in murine

erythroleukemia cells. Mol Cell Biol 10:1600–1608

Hihara Y, Kamei A, Kanehisa M, Kaplan A, Ikeuchi M (2001) DNA

microarray analysis of cyanobacterial gene expression during

acclimation to high light. Plant Cell 13:793–806

Hubel A, Schoffl F (1994) Arabidopsis heat shock factor: Isolation

and characterization of the gene and the recombinant protein.

Plant Mol Biol 26:353–362

Jaenicke R, Creighton TE (1993) Junior chaperones. Curr Biol 3:234–

235

Jakob U, Buchner J (1994) Assisting spontaneity: the role of HSP90

and smHSPs as molecular chaperones. Trends Biochem Sci

19:205–211

Jurivich DA, Sistonen L, Kroes RA, Morimoto RI, (1992) Effect of

sodium salicylate on the human heat shock response. Science

255:1243–1245

Keller T, Damude HG, Werner D, Doerner P, Dixon RA, Lamb C

(1998) A plant homolog of the neutrophil NADPH oxidase

gp91phox subunit gene encodes a plasma membrane protein with

Ca2+ binding motifs. Plant Cell 10:255–266

Kovtun Y, Chiu W-L, Tena G, Sheen J (2000) Functional analysis of

oxidative stress-activated mitogen-activated protein kinase cas-

cade in plants. Proc Natl Acad Sci USA 97:2940–2945

Larkindale J, Knight MR (2002) Protection against heat stress-induced

oxidative damage in Arabidopsis involves calcium, abscisic acid,

ethylene, and salicylic acid. Plant Physiol 128:682–695

Lee JH, Hubel A, Schoffl F (1995) Derepression of the activity of

genetically engineered heat shock factor causes constitutive

synthesis of heat shock proteins and increased thermal tolerance

in transgenic Arabidopsis. Plant J 8:603–612

Lee BH, Won SH, Lee HS, Miyao M, Chung WI, Kim IJ, Jo J (2000)

Expression of the chloroplast-localized small heat shock protein

by oxidative stress in rice. Gene 245:283–290

Levine A, Tenhaken R, Dixon R, Lamb C (1994) H2O2 from the

oxidative burst orchestrates the plant hypersensitive disease

resistance response. Cell 79:583–593

Lohmann C, Eggers-Schumacher G, Wunderlich M, Schoffl F (2004)

Two different heat shock transcription factors regulate immedi-

ate early expression of stress genes in Arabidopsis. Mol Gen

Genom 271:11–21

Manalo DJ, Liu AY-C (2001) Resolution, detection, and character-

ization of redox conformers of human HSF1. J Biol Chem

276:23554–23561

Mittler R (2002) Oxidative stress, antioxidants and stress tolerance.

Trends Plant Sci 7:405–410

Moller IM (2001) Plant mitochondria and oxidative stress: Electron

transport, NADPH turnover, and metabolism of reactive oxygen

species. Annu Rev Plant Physiol Plant Mol Biol 52:561–591

Morimoto RI (1998) Regulation of the heat shock response: cross talk

between a family of heat shock factors, molecular chaperones,

and negative regulators. Genes Development 12:3788–3769

Morre DJ (2002) Preferential inhibition of the plasma membrane

NADH oxidase (NOX) activity by diphenyleneiodonium chlo-

ride with NADPH as donor. Antioxid Redox Signal 4:207–212

Murashige T, Skoog F (1962) A revised medium for rapid growth and

bioassays with tobacco tissue cultures. Physiol Plant 15:437–497

Neill S, Desikan R, Clarke A, Hancock J (1999) H202 signaling in

plant cells. In: Smallwood MF, Calvert CM, Bowels DJ (Eds)

Plant responses to environmental stress. BIOS Sci Publ Ltd,

Oxford, pp 59–64

Noctor G, Foyer CH (1998) Ascorbate and glutathione: keeping ac-

tive oxygen under control. Annu Rev Plant Physiol Plant Mol

Biol 49:249–279

Nover L, Bharti K, Doring P, Mishra S, Ganguli A, Scharf K-D (2001)

Arabidopsis and the Hsf world: How many heat stress tran-

scription factors do we need? Cell Stress Chaperones 6:177–189

O’Donnell VB, Smith GC, Jones OT (1994) Involvement of phenyl

radicals in iodonium inhibition of flavoenzymes. Mol Pharmacol

46:778–785

Panchuk II, Volkov RA, Schoffl F (2002) Heat stress and heat shock

transcription factor dependent expression and activity of ascor-

bate peroxidase in Arabidopsis. Plant Physiol 129:838–853

Panikulangara TJ, Eggers-Schumacher G, Wunderlich M, Stransky H,

Schoffl F (2004) Galactinol synthase 1, a novel heat-inducible

and HSF-target gene responsible for heat-induced synthesis of

raffinose family oligosaccharides in Arabidopsis. Plant Physiol

136:3148–3158

Pnuelli L, Liang H, Rozenberg M, Mittler R (2003) Growth sup-

pression, altered stomatal responses, and augmented induction of

heat shock proteins in cytosolic ascorbate peroxidase (Apx1)-

deficient Arabidopsis plants. Plant J 34:185–201

Prandl R, Hinderhofer K, Eggers-Schumacher G, Schoffl F (1998)

HSF3, a new heat shock factor from Arabidopsis thaliana,

derepresses the heat shock response and confers thermotolerance

when overexpressed in transgenic plants. Mol Gen Genet

258:269–278

Puntarulo S, Sanchez RA, Boveris A (1988) Hydrogen peroxide

metabolism in soybean embryonic axes at the onset of germi-

nation. Plant Physiol 86:626–630

Rizhsky L, Davletova S, Liang H, Mittler R (2004) The zinc finger

protein Zat12 is required for cytosolic ascorbate peroxidase 1

expression during oxidative stress in Arabidopsis. J Biol Chem

279:11736–11743

Rossel JB, Wilson IW, Pogson BJ (2002) Global changes in gene

expression in response to high light in Arabidopsis. Plant Physiol

130:1109–1120

Plant Mol Biol (2006) 61:733–746 745

123

Royall JA, Ischiropoulos H (1993) Evaluation of 2¢,7¢-dichlorofluor-

escin and dihydrorhodamine 123 as fluorescent probes for

intracellular H2O2 in cultured endothelial cells. Arch Biochem.

Biophys 302:348–355

Samuel MA, Miles GP, Ellis BE (2000) Ozone treatment rapidly

activates MAP kinase signalling in plants. Plant J 22:367–376

Scharf KD, Siddique M, Vierling E (2001) The expanding family of

Arabidopsis thaliana small heat stress proteins and a new family

of proteins containing a-crystallin domains (Acd proteins). Cell

Stress Chaperones 6:225–237

Schoffl F, Prandl R, Reindl A (1998) Regulation of the heat-shock

response. Plant Physiol 117:1135–1141

Storozhenko S, De Pauw P, Van Montagu M, Inze D, Kushnir S

(1998) The heat-shock element is a functional component of the

Arabidopsis APX1 gene promoter. Plant Physiol 118:1005–1014

Sugiyama K, Izawa S, Yoshiharu I (2000) The Yap1p-dependent

induction of glutathione synthesis in heat shock response of

Saccaromyces cerevisiae. J Biol Chem 275:15535–15540

Torres MA, Onuchi H, Hamada S, Machida C, Hammond-Kosack

KE, Jones JDG (1998) Six Arabidopsis thaliana homologues of

the human respiratory burst oxidase (gp91phox). Plant J 14:365–

370

Vacca RA, De Pinto MC, Valenti D, Passarella S, Marra E, De Gara L

(2004) Production of Reactive Oxygen Species, alteration of

cytosolic ascorbate peroxidase, and impairment of mitochondrial

metabolism are early events in heat shock-induced programmed

cell death in tobacco bright-yellow 2 cells. Plant Physiol

134:1100–1112

Vandenabeele S, Van Der Kelen K, Dat J, Gadjev I, Boonefaes T,

Morsa S, Rottiers P, Slooten L, Van Montagu M, Zabeau M, Inze

D, Van Breusegem F (2003) A comprehensive analysis of

hydrogen peroxide-induced gene expression in tobacco. Proc

Natl Acad Sci USA 100:16113–16118

Vallelian-Bindschedler L, Schweizer P, Mosinger E, Metraux JP

(1998) Heat-induced resistance in barley to powdery mildew

(Blumeria graminis f. sp. hordei) is associated with a burst of

active oxygen species. Mol Plant Pathol 52:185–199

Vierling E (1991) The roles of heat shock proteins in plants. Annu

Rev Plant Physiol Mol Biol 42:579–620

Volkov RA, Panchuk II, Schoffl F (2003) Heat-stress dependent and

developmental modulation of gene expression: the potential of

house-keeping genes as internal standards in mRNA expression

profiling using real-time RT-PCR. J Exp Bot 54:2343–2349

Vranova E, Inze D, Van Breusegem F (2002) Signal transduction

during oxidative stress. J Exp Bot 53:1227–1236

Wunderlich M, Werr W, Schoffl F (2003) Generation of dominant-

negative effects on the heat shock response in Arabidopsis tha-

liana by transgenic expression of a chimaeric HSF1 protein fu-

sion construct. Plant J 35:442–451

Zhong M, Orosz A, Wu C (1998) Direct sensing of heat shock and

oxidation by Drosophila heat shock transcription factor. Mol

Cell 2:101–108

Zou J, Guo Y, Guettouche T, Smith DF, Voellmy R (1998) Repres-

sion of heat shock transcription factor HSF1 activation by

HSP90 (HSP90 complex) that forms a stress-sensitive complex

with HSF1. Cell 94:471–480

746 Plant Mol Biol (2006) 61:733–746

123