WO2004087952A2 - Genes regulated by hydrogen peroxide stress - Google Patents

Abstract

The present invention relates to a method to isolate genes involved in the reaction on and adaptation to hydrogen peroxide stress. The invention relates further to genes, isolated by said method.

Description

GENES REGULATED BY HYDROGEN PEROXIDE STRESS

The present invention relates to a method to isolate genes involved in the reaction on and adaptation to hydrogen peroxide stress. The invention relates further to genes, isolated by said method.

Until recently, the knowledge on the signaling capacity of hydrogen peroxide remained scarce. In bacteria and yeast the puzzle is starting to get elucidated. Christman et al., 1985 reported the discovery of the OxyR regulatory protein, a response regulator activating most of the H₂0₂- inducible genes in Salmonella. In higher eukaryotes and in fission yeast, ROS induce a stress- activated protein kinase pathway required for AP-1 activity (for a review: Toone et al., 2001). The AP-1 family of proteins has roles in controlling stress response genes. Technology like microarrays and cDNA-AFLP allows the study of the expression patterns of thousands of genes at the same time. Several studies report on the global gene response to H₂0₂ in yeast (Godon et al., 1998; Gasch et al., 2000), E. coli (Zheng et al., 2001), a mammalian cell line (Yoneda et al., 2001) and in plants (Desikan et al., 2001 ). Nevertheless, the mechanisms by which H₂O2 is sensed in plants and the signaling cascades it triggers remain largely unknown. Although reactive oxygen species (ROS) and more particularly hydrogen peroxide are formed in normal cell metabolism in plants, elevated ROS levels causing oxidative stress are mostly associated with adverse environmental conditions (Dat et al., 2000; Mittler and Berkowitz, 2001). During the hypersensitive response, the formation of H₂O2 is one of the earliest events (Lamb and Dixon, 1997; Greenberg, 1997). Evidence was shown for an increase of H₂0₂ upon salt stress (Burdon et al., 1996), temperature stress (Dat et al., 1998; Lopez-Delgado et al., 1998; Prasad et al., 1994; Kingston-Smith et al., 1999), UV-B irradiation (A-H-Mackerness et al., 1999), ozone (Langebartels et al., 2000), excess excitation energy (Karpinski et al., 1999) and drought stress (Guan et al., 2000; Pei et al., 2000). H₂0₂ is phytotoxic at high concentrations but acts as a messenger molecule in acclimatory signaling at lower, non-toxic concentrations. For example, a transient increase in H₂0₂ can lead to tolerance against high or low temperatures (Dat et al., 1998; Lopez-Delgado et al., 1998; Anderson et al., 1995; Prasad et al., 1994). At higher concentrations, H₂0₂ acts as a signal molecule in the orchestration of programmed cell death (Lamb and Dixon, 1997; Levine et al., 1994; Alvarez et al., 1998; Desikan et al.,1998). However, the knowledge on the molecular mechanisms and pathways involved in H₂0₂ signal transduction in plants remains scarce (Kovtun et al., 2000, Desikan et al., 2001, Vranova et al., 2002; Mittler et al., 2002; Neill et al., 2002). Previously we have used catalase deficient tobacco plants (Cat1 AS) as a model system to study the signaling aspects of H₂0₂ (Chamnongpol, 1996; Chamnongpol, 1998; Dat et al., 2003). In CAT1AS transgenics we can modulate H₂0₂ levels by exposing the plants to high light (HL) intensities (>800-1000 μmol/m²s). Under these conditions photorespiration is induced and since the CAT1AS plants only retain 10% of their residual catalase activity, they can't scavenge efficiently the photorespiratory H₂0₂ formed in the peroxisomes (Corpas et al., 2001). CatlAS plants are hence an ideal system to study the signal transduction of H_z0₂ in planta, because perturbation of H₂0₂ homeostasis can be sustained over time, no invasive techniques are needed to modulate H₂0₂ levels and, more importantly, a discussion about the physiological relevance of the H 0₂ concentrations is avoided. Earlier experiments with CAT1AS plants showed both the potential of H₂0₂ to trigger local and systemic acquired resistance and to trigger the induction of an active cell death program (Chamnonpol et al., 1998; Dat et al., 2003). Exposure of these plants for at least 4 hours to HL induced a synchronous cell death program requiring cfe novo protein synthesis and involves NADPH-oxidase activity, calcium-fluxes, G-proteins and kinase and phosphatase activities and that shares several features of programmed cell death: cell shrinkage, vacuolar collapse and chromatin condensation and fragmentation (Dat et al., 2003). When the plants are only exposed to HL for maximal 1h30' no cell death is induced, but oppositely a long-lasting resistance towards abiotic stress is established. In order to inventarize the transcriptional changes and hence obtain an insight in the signal transduction mechanisms during the onset of both cell death and acquired resistance we initiated a transcriptome analysis via cDNA-Amplified Length Polymorphism (AFLP) that allowed us to identify temporal induction patterns during a sustained H₂0₂ stress in CAT1AS plants. We used the cDNA-AFLP transcript profiling technology that allows not only to screen for differentially expressed genes in a certain system (Bachem et al., 1996; Durrant et al., 2000; Ditt et al., 2001; Breyne et al, 2002) but also allows to follow the kinetics of an expressed gene (Breyne and Zabeau, 2001). In contrary to the 'closed' nature of the microarray technology, cDNA-AFLP transcript profiling is applicable to organisms of which no extensive sequence information is available (Breyne et al., 2002). In addition, this method makes it possible to evaluate more time points or conditions at relatively low extra costs. Moreover, low abundant transcripts are detectable using the cDNA-AFLP. Finally, differentially expressed genes with a high level of sequence homology (e.g. of the same gene family) have greater chance to be detected separately than with microarrays. We used a modified cDNA-AFLP protocol in which one transcript results in only one cDNA-AFLP fragment and thus one band on the gel. By extensive cDNA-AFLP transcript profiling experiment combined with an advanced data processing, several H₂0₂ induced genes were identified. By profiling multiple time points, we were able to distinguish temporal expression patterns and hence identify groups of possibly coregulated genes.

It is a first aspect of the invention to provide a method to isolate genes involved in the hydrogen peroxide induced signaling pathway, comprising (1) growing at least one catalase deficient plant and one control plant (2) performing a high light treatment of said plants (3) harvesting a sample of said plants at the same time point (4) performing a differential display technique on the samples; and (5) analyzing the data. Differential display techniques are known to the person skilled in the art. Preferably, said differential display technique is cDNA- AFLP analysis or microarray analysis. Said catalase deficient plant may be any plant that has a lower catalase activity compared with the wild type. It may be obtained by random mutagenesis as well as by recombinant techniques. In one preferred embodiment, said plant is tobacco, and said catalase deficient plant is CAT1AS. In another preferred embodiment, said plant is Arabidopsis thaliana and said catalase deficient plant is CAT2AS or CAT2HP1. Preferably, said analysis comprises a quantification of the test and the control samples, such as a quantification of the bands obtained by the cDNA-AFLP analysis, or a quantification of the microarray data. Preferably, said quantification of the samples is followed by sequence analysis of the differentially displayed transcripts. In case of cDNA-AFLP, the data are preferably quantified using AFLP-Quantar™Pro, and the raw data are further processed using ArrayAN. A further aspect of the invention is a gene, identified with the method according to the invention. Preferably, said gene is selected from the group consisting of the sequences comprising SEQ ID N° 1 to SEQ ID N° 712. Even more preferably, said gene is selected from the group consisting of the sequences comprising SEQ ID N° 409 to SEQ ID N° 712. Most preferably, said gene is selected from the group consisting of the sequences comprising SEQ ID N° 35, SEQ ID N° 131, SEQ ID N" 142, SEQ ID N° 208, SEQ ID N° 321 and SEQ ID N° 326.

Gene as used here refers both to the genomic sequence (including possible introns) as well as to the cDNA derived from the spliced messenger. It may refer to the promoter sequence too. However, it is clear for the person skilled in the art that for some applications, the coding sequence, such as it may be derived from the cDNA, may be operably linked to a suitable promoter. Operably linked refers to a juxtaposition wherein the components so described are in a relationship permitting them to function in their intended manner. A promoter sequence "operably linked" to a coding sequence is ligated in such a way that expression of the coding sequence is achieved under conditions compatible with the promoter sequence. Still another aspect of the invention is the use of a gene, isolated with the method according to the invention, to obtain stress tolerance. Preferably said stress tolerance is oxidative stress tolerance. Preferably said gene is selected from the group consisting of the sequences comprising SEQ ID N°1 to SEQ ID N° 712, or a homologue thereof. Even more preferably, said gene is selected from the group consisting of the sequences comprising SEQ ID N°409 to SEQ ID N° 712, or a homologue thereof. Still even more preferably, said gene is selected from the group consisting of the sequences comprising SEQ ID N°35, SEQ ID N° 131, SEQ ID N° 142, SEQ ID N° 208, SEQ ID N° 321 and SEQ ID N° 326, or a homologue thereof. A homologue as used here means for a gene that the nucleic acid of the gene has a nucleic acid sequence that is at least 70% identical, and even more preferably at least 80% identical, and even more preferably at least 85% identical, and even more preferably at least 90% identical and even more preferably at least 95% identical, and even more preferably at least 96% identical, and even more preferably at least 97% identical, and even more preferably at least 98% identical, and even more preferably at least 99% identical, as measured by a BLASTN search (Altschul et al., 1997).

Another aspect of the invention is the use of a gene, isolated with the method according to the invention, whereby said gene is encoding a protein selected from the group consisting of the sequences comprising SEQ ID N° 714 to 730, or a homologue thereof.

Still another aspect of the invention is the use of a gene encoding a protein comprising SEQ ID N° 713, or a homologue thereof, to obtain stress tolerance. Preferably, staid stress tolerance is oxidative stress tolerance. A homologue as used here means for a protein that the protein encoded by the gene has an amino acid sequence that is at least 75% identical, and even more preferably at least 80% identical, and even more preferably at least 85% identical, and even more preferably at least 90% identical and even more preferably at least 95% identical, and even more preferably at least 96% identical, and even more preferably at least 97% identical, and even more preferably at least 98% identical, and even more preferably at least 99% identical, as measured by a BLASTP search (Altschul et al. , 1997)

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 : At5g22860 is upregulated in HL-treated CAT2AS plants. Fold changes of At5g22860 transcript in CAT2AS plants compared to wild-type plants exposed to HL conditions.

Figure 2: Left: Representative phenotypes of knock-out Arabidopsis plants in a serine carboxypeptidase S28 family protein. Right: wild-type phenotype.

Figure 3: SVISS-induced tobacco knockout plants 4 weeks post co-inoculation with chimeric satellite virus iπ-vitro transcripts and TMV helper virus leaf extract. Identities of chimeric constructs are indicated on the photos with their respective transcript tag-ID.

EXAMPLES

Materials and methods to the examples

Tobacco plant growth conditions and stress treatments Seeds of transgenic CAT1AS and wild-type SR1 Nicotiana tabacum (L.) were germinated and grown as described before (Chamnongpol et al., 1998). Twelve CAT1AS and wild-type plants were simultaneously grown in a controlled growth chamber (Herasus, Balingen, Germany: 24°C, 70% relative humidity, 16h day/8h night, 150 μmol.nAs^"1) High light treatments (800- 1000 μmol.πAs^"1) were performed in a Phytotron chamber PLC 970 (Sanyo Gallenkamp, Leicester, UK). CAT1AS and SR1 plants were equally distributed in the chamber to limit the boundary effect or potential effect of microenvironments. The plants were positioned under HL for 0', 10', 20', 30', 40', 50', 1h, 2h, 4h, 6h, 8h and 11h in such a way that there was no overshadowing between leaves of the various plants. The harvesting was done after each time point by taking simultaneously 1 CAT1AS and 1 wild-type plant out of the chamber; the 6* leaf from the bottom was harvested and immediately frozen in liquid nitrogen. Special care was taken to avoid touching and all sampling was done with gloves and scissors washed with 100% EtOH between each sampling. Timing between taking the plant out of the chamber and freezing of the sample did not exceed 1 minute.

cDNA-AFLP analysis

PolyA⁺ RNA was prepared from 500 μg of total RNA (Trizol^R Reagent) using oligotex™ columns according to the manufacturers instructions (Qiagen). First strand cDNA was synthesized by reverse transcription starting from 1 μg of polyA⁺ using a biotynilated oligo-dT₂₅ primer (Genset) and Superscript II enzymes (Gibco-BRL). Second strand synthesis was done by strand replacement with E. coli ligase (Gibco-BRL), DNA polymerase I (USB) and RNAse-H (USB). Cleaning of the cDNA occurred by elution through PCR purification columns (Qiagen). Five hundred ng of double stranded cDNA was used for AFLP analysis as described by Vos et al (1995) and Bachem et al (1996) with several modifications: after the first digestion with BstYI enzyme, the 3' end fragments were collected on Dynabeads (Dynal) by means of their biotinylated tail. After digestion with Msel enzyme, the released restriction fragments were collected and used as templates in the subsequent AFLP steps. The adapters and primers used are: BstYI-adapter: 5'- CTCGTAGACTGCGTAGT

CATCTGACGCATCACTAG -5' Msel adapter: 5'- GACGATGAGTCCTGAG

TACTCAGGACTCAT -5' BstYI-primers: 5'- GACTGCGTAGTGATC(T/C)N,.₂ ^■ Msel-primers: 5'- GATGAGTCCTGAGTAAN,.;, -3'

For preamplifications, an Msel-primer without selective nucleotides was combined with a BstYI-primer containing either a T or C as 3' most nucleotide. PCR conditions were as described (Vos et al, 1995). The obtained amplification mixtures were diluted 600 fold and 5 μl was used for selective amplifications using a P³³-labeled BstYI-primer (plus 1 or 2 selective nucleotides), the Msel-primer (plus 1 or 2 selective nucleotides) and the Amplitaq-Gold polymerase (Roche) following the described procedure (Vos et al., 1995). In a pilot experiment, all timepoints (except wild type: 0', 1h and 6h HL) and 20 randomly chosen primer combinations (BstT4-Mse31 , BstT4-Mse32, BstT4-Mse33, BstT11-Mse4, BstT21-Mse1, BstT31-Mse2, BstT32-Mse3, BstT34-Mse1, BstT34-Mse2, BstT34-Mse3, BstT42-Mse1, BstT44-Mse2, BstC4-Mse32, BstC4-Mse33, BstC11-Mse4, BstC31-Mse2, BstC32-Mse2, BstC33-Mse2, BstC34-Mse2, BstC43-Mse1) were used. In the gene discovery experiment, 10 timepoints (CAT1AS: 0', 30', 50', 2h, 4h, 11h; wild type: 30', 50', 2h, 11h) all possible BstYI+1/Msel+2 primer combinations (128 in total) were used. Of every primer combination used, the according 4 BstYI+2/Mse+2 primer combinations were used on pooled samples of all time points. Amplification products were separated on 5% polyacrylamide gels using the Sequigel system (Biorad). Sodium acetate (22mg/l) was added to the buffer in the bottom tank to avoid run-off of the smallest fragments. Gels dried on 3MM Wattman paper were exposed to Kodak Biomax films and scanned in a Phosphorlmager 445 SI (Molecular Dynamics).

Data analysis

Scanned Phosphorlmager gel images were processed with AFLP-Quantar™Pro ( Keygene, Wageningen, The Netherlands). This software application allows an accurate quantification of band intensities in DNA fingerprints. Intensities of individual bands are determined and semi- automatically corrected for variations such as running and loading differences and image artifacts. The obtained raw expression data were further processed using a Microsoft Access based, in-house developed software application (ArrayAN). ArrayAN allows an accurate and automated high-throughput handling of gene expression data. An additional normalization of band intensities was performed based on the expression levels of constitutively expressed genes. Therefore, we firstly divided each band intensity with a correction factor. This correction factor was calculated by dividing the sum of all individual band intensities within one lane with the average of all sums within the respective primer combination. For each band (corresponding to a unique gene) a coefficient of variation (CV) was calculated. This CV. is the standard deviation on the expression value of each band divided by the average of expression during the time-course. Within each primer combination, 7% of the genes with the lowest CV value were marked as constitutively expressed throughout the experiment. Per lane the sum of intensities of these bands was summed and divided by the average of sums within the respective primer combination. This generated a second correction factor that was used to normalize the raw expression data generated by AFLP-Quantar™Pro. On these normalized 5 data, again the CV was calculated for each gene. Within the pilot experiment, expression values of genes with a CV higher than 0.8 were withdrawn for further analysis. In the gene discovery experiment only genes with a CV higher than 0.9 were withdrawn. Subsequently, expression values of selected genes were log transformed and variance normalized according to Tavazoie et al. (1999) before cluster analysis. We used Cluster and Treeview software 10 (Eisen et al., 1998) for hierarchical clustering and the software described by De Smet et al. (2002) for adaptive quality-based clustering.

Fragment isolation and characterization

Differentially expressed genes were cut out from the BstYI+2 Mse+2 expression pattern of the

15 pooled time points and reamplified with BstYI+2/Mse+0 primers. If not detectable on BstYI+2 Mse+2, they were cut out from the BstYI+1/Mse+2 expression pattern. From the pilot experiment, part of the reamplified tags were cloned into the pGEMT easy vector (Promega) and transformed into E.coli. The insert of two individual clones was sequenced. Only when these two sequences were the same, they were included for further analysis. The rest of the

20 reamplified tags were directly sequenced. Sequences of minor quality were not retained. The nucleotide sequences were compared with sequences deposited in the databases (GenBank, EMBL, DDBJ, PDB) and translated DNA sequences were compared with protein sequences in databases (GenBank CDS translations, PDB, SwissProt, PIR, PRF) by using the BLAST algorithm (Altschul et al., 1990). All sequences were also compared to the TIGR Gene Indices

25 database using the FastA algorithm (Pearson and Lipman, 1988) to find homologues that are longer toward the 5' terminus. These translated TIGR gene indices sequences then became the query sequences in a subsequent Blast homology search against the protein sequences in the database. Only when the result showed a significant homology (e-value<10^"3) it was taken into account for the annotation of the tag.

30

Arabidopsis thaliana plant growth conditions and high light treatments Catalase deficient Arabidopsis thaliana lines CAT2AS, CAT2HP1 , and the control line PTHW were grown in the controlled Phytotron growth chambers (Weiss Umwelttechnic GmbH, Reiskirchen-Lindstruth, Germany) with 140μmol/m²s light, a relative humidity of 70%, a

35 12h/12h day/night (6am to 6pm) regime and 22/18°C day/night temperature. The light system of these chambers was designed to stimulate the solar spectrum and consisted of metal halide, quartz halogen, and blue fluorescent lamps mounted at the ceiling of a separate lamp cabinet on top of the experimental chamber (Thiel et al., 1996). For the HL treatment, six week-old plants were transferred to a sun-simulator under identical growth conditions. High light intensities were 1800 μmol/m²s. At timepoints 0, 3, 8, 23 h during HL and 25 h of HL followed by 4, 7 and 22 h of recovery, the middle-aged leaves of 5 to 10 separate plants were harvested and pooled.

Construction of microarrays

The Arabidopsis thaliana microarray consisted out of 6,528 cDNA fragments spotted in duplicate, distant from each other. The clone set included 6,008 Arabidopsis genes composed from the unigen clone collection from Incyte (Arabidopsis Gem I, Incyte, USA) and 520 positive and negative controls. The cDNA inserts were PCR amplified using M13 primers, purified with MultiScreeπ-PCR plate (cat: MANU03050, Millipore, Belgium) and arrayed in 50% DMSO on Type VII silane coated slides (Amersham BioSciences, Buckinghamshire, UK) using a Molecular Dynamics Generation III printer (Amersham BioSciences). Slides were blocked in 2xSSPE, 0.2% SDS for 30 minutes at 25 °C.

RNA amplification and labeling

Antisense RNA amplification was performed using a modified protocol of in vitro transcription as described earlier by us in Puskas et al. (2002a, b). For the first strand cDNA synthesis, 5 μg of total RNA was mixed with 2 μg of a HPLC-purified anchored oligo-dT + T7 promoter (5'- GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGG-T₂₄(ACG)-3^,) (Eurogentec,

Belgium) 40 units of RNAseOUT (Invitrogen, Merelbeke, Belgium) and 0.9M D(+) trehalose (Sigma Belgium) in a total volume of 11 μl, and heated to 75°C for 5 minutes. To this mixture, 4 μl 5x first strand buffer (Invitrogen, Belgium), 2 μl 0.1 M DTT, 1 μl 10 mM dNTP mix, 1 μl 1.7 M D(+)trehalose (Sigma Belgium) and 1 μl, 200 Units of Superscript II (Invitrogen, Belgium) was added in 20 μl final volume. The sample was incubated in a Biometra-Unoll thermocycler at 37°C for 5 minutes, 45oC for 10 minutes, 10 cycles at 60°C for 2 minutes and at 55°C for 2 minutes. To the first strand reaction mix, 103.8 μl water, 33.4 μl 5x second strand synthesis buffer (Invitrogen, Belgium), 3.4 μl 10 mM dNTP mix, 1 μl of 10U/μl E.coii DNA ligase (Invitrogen, Belgium), 4 μl 10 U/μl E.coii DNA Polymerase I (Invitrogen, Belgium) and 1 μl 2U/μl E.coii RNAse H (Invitrogen, Belgium) was added, and incubated at 16°C for 2 hours. The synthesized double-stranded cDNA was purified with Qiaquick (Qiagen, Hilden, Germany). Antisense RNA synthesis was done by AmpliScribe T7 high yield transcription kit (Epicentre Technologies, USA) in total volume of 20 μl according to the manufacturer's instructions. The RNA was purified with RNeasy purification kit (Qiagen, Germany). From this aRNA, 5 μg was labeled by reverse transcription using random nonamer primers (Genset, Paris, France), 0.1 mM d(G/T/A)TPs, 0.05 mM dCTP (Amersham BioSciences, UK), 0.05 mM Cy3-dCTP or Cy5-dCTP (Amersham BioSciences, UK) 1x first strand buffer, 10 M DTT and 200 Units of Superscript II (Invitrogen, Belgium) in 20 μl total volume. The RNA and primers were denatured at 75°C for 5 minutes and cooled on ice before adding the remaining reaction components. After 2 hours incubation at 42°C, mRNA was hydrolyzed in 250 mM NaOH for 15 minutes at 37°C. The sample was neutralized with 10 μl of 2 M MOPS and purified with Qiaquick (Qiagen, Germany).

Array hybridization and post-hybridization processes

The probes were resuspended in 210μl hybridization solution containing 50 % formamide, 1x HybridizationBuffer (Amersham BioSciences, UK), 0.1 % SDS and 60μg/ml poly-dT. Hybridization and post-hybridization washing was performed at 45oC using an Automated Slide Processor, ASP (Amersham BioSciences, UK). Post-hybridization washing was performed in 1xSSC, 0.1% SDS, followed by O.lxSSC, 0.1% SDS and O.lxSSC. The complete ASP program can be downloaded from www.microarrays.be (/technology/protocols).

Scanning and data analysis

Arrays were scanned at 532 nm and 635 nm using a Generation III scanner (Amersham BioSciences, UK). Image analysis was performed with ArrayVision (Imaging Research Inc, Ontario, Canada). Spot intensities were measured as artifact removed total intensities (ARVol) without correction for background. For 24 negative control spots containing a Bacillus subtiiis specific cDNA and 6,008 Arabidopsis spots, we first addressed within-slide normalization by plotting for each single slide a "MA-plot" (Yang et al. 2002) where M = log₂ (RIG) and A = log₂ RxG for each spot. The "Lowess" normalization (r=0.2) was applied to correct for dye intensity differences. Based on the M' (adjusted M) and A values for each gene, adjusted log₂R and log₂G signal intensities were obtained. Based on the 96 adjusted log₂R and log₂G signal intensities of the negative control spots, the median and the 95 Percentile were calculated. The 95 Percentile was defined as a signal threshold. For each gene, the adjusted log R and log₂G signal intensities were compared to the signal threshold. None of the 6008 genes were uniformly below the signal threshold. Subsequently, to normalize between slides and to identify genotype-specific differentially expressed genes over time, we applied two sequential analyses of variance (ANOVAs), proposed by Wolfinger et al. (2001) as follows: 1) let y_Mm be the base-2 logarithm of the "Lowess"-transformed spot measurement from gene / (/^' = 1-6008); we subjected all 576,768 y_(Wm to a linear normalization model of the form y«_m = μ + As + A*D;R_m + ε*m. where μ is the sample mean, A_k the effect of the Wh array (k = 1-24), A*D_/R_m the channel effect (A_/A) for the /77th replication of the total collection of cDNA fragments ( = 2; left and right), and ε;_Wm the stochastic error; 2) we subjected the residuals ε_Λ/mfrom this model to 6,008 gene-specific models of the form t,_lgUmn = μ + Sy+ T„+ V_g+ V_gT„ + G + GAk ⁺ Yijumn where S_j is the main effect for the type of sample 0^" = 2; control and test sample), T_n is the main effect of the nth time point (π = 1-4), V_g is the main effect of the gth genotype (g = 3: WT, transgenel and transgene2), V₉T„ is the interaction between the genotype and time (effect of interest), GA models the spot effect, and GA models the gene-specific dye effect, and γ_jsMro„ the stochastic error. We made standard stochastic assumptions about the preceding linear models. In particular, the effects A^, A/ARm, GA*, ε_Λ/m and γ₉w were considered as random effects, where the remaining terms Sy, T„, V_s, V₉T„ and GA were assumed to be fixed effects. As a measure of variability in expression levels between genotypes overtime, we calculated the Wald statistic for each gene. The Wald statistic is usually tested against the χ² distribution. In a balanced design, the Wald statistic divided by its degrees of freedom will be distributed as F„,_m where is the number of degrees of freedom of the fixed effect, and n is the number of residual degrees of freedom for the fixed effect. For unbalanced designs, as is the case here, it may be difficult to deduce the appropriate residual number of degrees of freedom, and the F distribution is only approximate. The p-value cutoff was set arbitrarily at the stringent 0.001 level. No further adjustments for multiple testing were performed.

All standard calculations and statistics, including the "Lowess" fit, were done using Genstat (GenStat Release 6.1 for Windows.VSN International, Hemel Hempstead UK). Performing both the normalization and gene model fits was done using the built-in "REML" procedure in Genstat (Payne et al., 2000); Saving the non-hierarchical Wald test for the fixed term V₉T„ in the REML analyses was done using "WALD" procedure in Genstat (Goedhart, P and Thissen.T. Biometris Genstat Procedure Library 6th edition

(www.biometris.nl/software/qenstat).

DNA Chip Analysis — Catalase deficient lines and control plants were grown under controlled growth conditions. Six-weeks old plants were transferred to a sun-simulator. High light intensities were 1600 μmol.m².s^"1. Middle aged leaves were harvested and pooled. RNA was isolated and used to perform chip analyses (Arabidopsis ATH1 chips; Affymetrix, Santa Clara, CA). Comparative analysis of samples was performed using Affymetrix MAS5.0 software and the Silicon Genetics GeneSpring version 5.1.

T-DNA knockout- The cDNA of the Arabidopsis metacaspase gene (AtMC6) was cloned in the binary vector (Karimi et al., 2002) and this construction was used for gen inactivation. The floral dip technique was used for plant transformation. Example 1: Profiling gene expression during a sustained _0₂ stress in catalase deficient tobacco

Twelve transgenic tobacco plants (CAT1 AS) that retain only 10% of wild-type catalase activity (Chamnongpol et al., 1996) and twelve wild-type plants were simultaneously exposed to high light (HL; 1000 μmol.nAs^"1) for 0, 10min, 20min, 30min, 40 min, 50 min, 1 h, 2h, 4h, 6h, 8h and 11h, respectively. Such a HL treatment provokes a sustained in planta H₂0₂ stress within the CAT1AS plants and leads to the induction of cell death (Dat et al., 2003). Cell death was visible after 8 hours when patches of gray, chlorotic areas of dying cells appeared on both the 5^m and 6¹" leaf of CAT1AS plants, but first appearance of death cells is clear after 4 hr HL as evidenced by trypan blue staining (Dat et al., 2003). We visualized gene expression in the 6 leaf during the whole time-course in both CAT1AS and wild-type plants using a modified cDNA-AFLP protocol in which one transcript results in only one cDNA-AFLP fragment and thus one band on the gel (Breyne and Zabeau, 2001) combined with an advanced data processing. By profiling multiple time points we intended to group related genes by their temporal expression patterns and identify functional relationships between genes within each group. In a pilot (21 timepoints) and a gene discovery (10 timepoints) experiment, respectively 20 and 128 cDNA-AFLP primer combinations were used and visualized as described in materials and methods. For each primer-combination 80-120 bands (gene fragments) were observed varying in length between 50 and 600 bp. Using AFLP-Quantar™Prø software, ail lanes were identified and oblique ones corrected, all AFLP fragments were quantified resulting in individual band intensities per lane. These raw expression data were normalized for differences in total lane intensities due to differential PCR amplification and potential loading errors. AFLP- Quantar™Pro performs a default lane normalization based on total lane intensity. We experienced that this correction was insufficient and subjected the data to a second lane normalization based on constitutive bands: assuming that most of the genes visualized in one primer combination are not altered in their expression during the HL treatment, we selected the genes with the lowest coefficient of variation as a reference to calculate a conversion factor to adjust the lanes. Therefore we used the ArrayAn software that allows an easy handling of large amounts of data and automatically executes the calculation of parameters and the extra lane correction. We used an objective, mathematical approach by calculating the coefficient of variation to decide which bands were included for further analysis (this in contrast with e.g. visual inspection). Subsequently, each individual profile was variance normalized as described by Tavazoie et al., 1999. Bands corresponding to differentially expressed genes were cut out, reamplified, sequenced and subjected to further analysis (see materials and methods). Example 2: Pilot experiment reveals cluster of early H₂O2 induced genes.

With 20 primer combinations we could follow the expression of 1628 genes in both wild type and CAT1AS plants during HL. In total 262 genes were considered as differentially expressed (CV>0.8) during the time-course and subjected to both a hierarchical clustering analysis and quality based clustering approach (Eisen et al.,1998; De Smet et al., 2002). Both clustering algorithms revealed four main clusters. The first contains 128 genes induced only in the CAT1AS plants, what indicates H₂0₂ responsiveness. In the Eisen clustering result, groups of genes that are upregulated during the first hour after high light induction are clearly distinguishable and suggest very early and direct H₂0₂ responsiveness. A second cluster of 28 genes shows a decreased expression in both lines. 19 genes are induced only in SR1 plants in the third cluster. Finally, a fourth cluster contains a significant portion of 77 genes that are lower expressed from the start in CAT1 AS plants and downregulated in the SR1 plants after 4- 6hrs of high light treatment. We hypothesize that this cluster houses genes that are downregulated by the HL treatment that causes stress in the wild type plant. This would mean that, although SR1 and CAT1AS are phenotypically undistinguishable, on the molecular level CatlAS plants are in a stressed status comparable with the SR1 plants after 4hrs of HL.

Example 3: Gene discovery experiment as a base for unraveling H₂0₂ response in plants

In a gene discovery experiment the redundant time points of the pilot experiment were omitted for further analysis: six time points from CAT1AS and four from SR1 were selected in such a way that early induced genes and groups of genes with similar expression patterns still could be identified. With all possible BstYI+1/Mse+2 primer combinations we followed the expression of 13752 genes. Out of these, 1207 genes showed a CV higher than 0.9 and were considered as differentially expressed and subjected to both a hierarchical clustering analysis and a quality based clustering approach. In analogy with the pilot experiment the same four clusters of similar gene expression could be distinguished. In total 691 genes are H₂0₂ upregulated and again groups of genes induced during the first hour are clearly distinguishable. Secondly,

101 genes are repressed in both lines. The third cluster contains 110 genes that are induced only in SR1 plants. Finally, 252 genes show lower expression from the start in CAT1AS plants and are downregulated in the SR1 plants after 4-6hrs of high light treatment.

Example 4: Sequence analysis

Tags that were differentially expressed were excised from the gel out of the BstYI+2/Mse+2 expression pattern of the pooled time points and reamplified using the corresponding BstYI+2 Mse+0 primers. All sequences were stripped of adapter sequences and compared to nucleotide and protein databases using Blast and FastA algorithms as described in materials and methods. In total from the pilot and gene discovery experiment 156 and 556 sequences could be determined unambiguously varying in length from 15 up to 660 basepairs. The other sequences were a mixture of PCR products and could not be directly sequenced. Considerably, of the 691 H₂O₂ upregulated genes in the gene discovery experiment we were able to obtain 400 sequences. 109 fragments show homology to genes with an unknown function. Surprising is the high number of tags (230) that don't show a significant hit with any of the screened databases.

Example 5: Comparison ofH202 signaling in tobacco and Arabidopsis thaliana

Using cDNA-AFLP transcript profiling technology, an estimated 50% of the transcriptome (14.000 transcript fragments) upon high light stress was investigated in catalase deficient tobacco. Approximately 1400 transcript fragments (10%) were identified as differentially expressed and the sequence identity of 713 transcript fragments revealed genes involved in defence and hypersensitive response, mitochondrial and proteolytic and posttranscriptional processes, hormone interplay, lipid metabolism and signal transduction. Interestingly, photosynthetic components were clearly repressed under normal growth conditions in catalase deficient tobacco plants, a situation only reached after 4-6 hours of HL in wild type plants, indicating a signaling function of (peroxisomal) hydrogen peroxide towards photosynthesis. In analogy with tobacco, microarray technology was used on two different lines of high light treated catalase deficient Arabidopsis plants. Of the in total 6004 ESTs (±25% of the genome) present on the array, 1552 (24%) displayed differential expression between catalase deficient lines and the control line. Cluster analysis revealed clusters with similar expression patterns as identified in tobacco. Although there exists at first sight a large difference between the percentages of differentially expressed genes between Arabidopsis and tobacco, a comparison of the molecular profiles was made to determine if similar processes were involved in the response upon elevated H202 in catalase deficient tobacco and Arabidopsis thaliana. Three different approaches were followed to compare the differentially expressed sequences of both systems. In the first approach, the tobacco transcript fragments were compared to the MIPS Arabidopsis thaliana Genome Database (MatDB, http://mips.gsf.de/proj/thal/db/index.html) of the Munich Information Center for Protein Sequences (MIPS). The resulting At-code was looked up in the At-codes of the 1552 differentially expressed genes of the Arabidopsis experiment. Using this method, 61 homologues were identified. In a second method, a similar process was followed but the corresponding protein sequences of the At-codes were compared resulting in 216 homologues. In the third approach, the FastA program was used to identify extended sequences of the tobacco transcript in The Institute for Genomic Research (TIGR) gene index database (http://www.tigr.org/tdb/tgi/). This database contains the assemblies of known ESTs of several plant species and often provides a longer tobacco 3' sequence of the primarily 5' region sequences identified in the tobacco cDNA-AFLP experiment. This TIGR gene index was the subject of a new homology search against the MatDB database. Finally, the homologous A. thaliana gene was compared on the protein level with the differentially expressed genes from the Arabidopsis thaliana experiment. Using this approach, 246 homologues were identified. For different homology searches used in the three approaches, a minimal e-value of 10e-05 was used as cut-off. The results of the sequence comparisons, their gene description and e-values are represented in table 1

Example 6: Homologous genes and gene families are involved in H202 signaling in catalase deficient tobacco and Arabidopsis thaliana.

Although in tobacco and Arabidopsis thaliana catalase deficient model systems only an estimated 25% of the total traπscriptome upon H202 stress was analyzed or identified, the search for homologous genes that show differentially expression in both systems, revealed numerous homologous genes or genes belonging to the similar functional classes. Transcripts coding for enzymes involved in production or scavenging of ROS included polyamine oxidases, peroxidases and catalase. A significant overlap was observed between genes involved in defense. Most significantly was the induction of heat shock proteins in both model systems, reinforcing an important role for these proteins during oxidative stress, most likely by protecting crucial enzymes and proteins against the harmful effects of ROS. Beside, several glutathione-S-transferases, an Avr9 elicitor response protein, a Pto-kinase interactor, and wound-and sensescence associated proteins were homologous. The 12 oxophytodienoate reductase and UDP-glucose:salicylic acid glucosyltransferase betray the involvement of jasmonic acid and salicylic acid. Both systems show increased transcripts of mitochondrial cytochrome C and of a mitichondrial substrate carrier. Protein turnover represents clearly a common response upon an increase in H202. Several ubiquitin genes together with E2 ubiquitin-conjugating enzymes are differentially expressed in both tobacco and Arabidopsis. Interesting was the large overlap in downregulated components of the photosynthetic apparatus. Several chlorophyll a/b-binding protein, photosystem I & II subunits and the small subunit of ribulose-1,5-bisphosphate carboxylase were repressed in both catalase deficient species. These genes were downregulated after 4-6 hours of high light treatment in wild type tobacco and in wild type Arabidopsis already after 3 hours. Although, the timing of repression differed between catalase deficient plants. In tobacco, the photosynthetic genes were already repressed under normal growth conditions whereas in Arabidopsis, full downregulation was reached only after three hours of high light, comparable to the timing of repression in wild type Arabidopsis plants. Nevertheless, since the sampling during the first hours of high light only included 0 and 3 hours, it cannot be excluded that these genes were earlier downregulated in catalase deficient Arabidopsis compared to wild type plants. Most exciting was the presence of similar groups of signal transduction components and transcription factors in both catalase deficient species. Receptor-like kinases and a homologue of the ARR1 protein are good candidates to act as important sensory mechanisms during oxidative stress. A group of protein kinases and phosphatases could play key-roles in further downstream signal transduction represented by the large group of commonly expressed transcription factors. Several AP2 and ethylene-responsive element binding protein homologues, WRKY and heat shock transcription factors, a bZIP-domain containing factor, MYB-related and scarecrow transcription factors show differential expression. Further detailed molecular analysis of these signal transduction components certainly will help to elucidate H202 signaling and provide a better insight in the induction of defense and cell death in plants.

Example 7: A knock-out Arabidopsis in At5g22860 (a serine carboxypeptidase-Hke protein) shows increased stress sensitivity.

A genome-wide transcriptome analysis was performed on a control line and a catalase deficient transgenic lines (CAT2AS). RNA from plants exposed to High Light for 0, 3, 8 and 23 hours was hybridised to the Arabidopsis ATH1 Affymetrix chip. This analysis revealed that a transcript coding for a serine carboxypeptidase S28 family protein (At5g22860) was gradually responsive to increasing H 0₂ levels as it was specifically upregulated in in the catalase deficient plants during the HL treatment (see Figure 1).

We identified a knock-out transgenic Arabidopsis line for At5g22860 in a T-DNA transgenic line (0X6-19). Segregation analysis is indicative for a single T-DNA insertion. Among the T2 plants (T1 being the primary transgenic plant), 25% of the plants showed a stress-sensitive phenotype. In vitro grown plants under control growth conditions were clearly smaller, and their leaves colored brown-purple (indicative of increased anthocyanin accumulation) (Figure 2). From this T2 generation, stress sensitive plants and plants with no phenotype (but resistant to the selectable marker for plant transformation) were transferred to soil for seedsetting. T3 plants from stress sensitive plants all showed the observed phenotype and of the T3 plants originating from T2 plants with no phenotype, again 25% showed the stress sensitive phenotype. These segregation ratios are indicative for a T-DNA knock-out phenotype, independent from a potential effect provoked by the original MC6 transgene. The flanking sequences of the T-DNA insertion were amplified by (TAIL-) PCR according to Liu et al. (1995) with modifications. A 400 bp genomic DNA fragment flanking the left border of the T-DNA insert was isolated. Direct sequence analysis of this fragment demonstrates that the T- DNA was inserted into the serine carboxypeptidase S28 family protein (At5g22860).

Example 8: Transgenic Arabidopsis (overproduction and RNAi) of candidate regulators in H₂0₂ signal transduction

To identify within the established Arabidopsis H₂0₂ transcriptome dataset key genes for a detailed functional analysis we applied two criteria: (i) selected genes or homologues are present in both the tobacco and

Arabidopsis dataset; (ii) selected genes are rapidly upregulated by increased hydrogen peroxide levels.

Finally 17 genes were retained for further analysis:

At4g31920

At1g33600

At1g56450

At3g20410 At5g24400

At1g20580

At4g32730

At1g68050

At2g37590 At5g05410

At3g54620

At2g38470

At2g23320

At5g04410 At5g09330

At3g29035

At1g01720

Full-length cDNAs were inserted into binary vectors (pK7WG2D for overexpression and pK7GWIWG2 for silencing; Karimi et al., 2002) via recombinational cloning (GateWay). Transgenic plants were generated via the floral dip method. All constructs were introduced into a wild-type Arabidopsis background and in catalase deficient plants.

Primary transgenics were obtained by selection on Kanamycine and assessed for endogenous and transgene dependent gene expression. Transgenic lines with increased or decreased levels of the target genes were retained for further analysis. Selected homozygous lines for all constructs in both wild-type and catalase deficient background will be assessed for:

(i) increased biotic and abiotic stress resistance (by exposing them to a variety of stresses) (ii) the perturbation of downstream transcriptional regulons governed by the target genes (by genome-wide transcriptome analysis; molecular phenotyping). Example ^.-Functional screen with viral induced gene silencing technology identified candidate genes in stress resistance

To identify within the established tobacco H₂0₂ transcriptome dataset key genes for a detailed functional analysis we initiate a functional screening by a transient gene-silencing strategy in tobacco.

Therefore we used the satellite-virus-induced silencing system (SVISS) (Gossele et al., 2002). The PCR reamplified cDNA-AFLP transcript fragments were cloned directly into the satellite virus vector and verified by sequence analysis after cloning. Inoculations were performed as described in Gossele et al. (2002). Viral (sTMV) RNA concentration and systemic spreading was assessed by Northern blot analysis.

Various target genes selected form the H₂0₂ tobacco dataset were effectively silenced and produced knockout phenotypes indicative for an involvement in defense responses. Wild-type TMV/STMV inoculated plants shows almost no viral symptoms. Plants co-inoculated with wild- type TMV leaf extract and target gene-specific chimeric satellite virus in-vitro transcripts show various phenotypes: changes in pigmentation; hypersensitive response-like lesions; cell death; deformed leaf (see also Figure 3).

a) BC3-M42-008 [SCARECROW gene regulator-like] SEQ ID N° 142 b) BT34-M2-030 [R2R3 Myb protein] SEQ ID N° 326 c) BC4-M44-043 [homologous to F-box protein FKF1/AD03] SEQ ID N° 208 d) BC3-M33-088 [AP-2 related transcription factor] SEQ ID N° 131 e) BT31-M2-065 [Dof zinc finger protein] SEQ ID N° 321 f) BC1-M43-022 [HBF-1 ; bZIP transcription factor] SEQ ID N°35

Table 1: Comparison of differentially expressed genes in high light treated catalase deficient tobacco and Arabidopsis thaliana.

Three different approaches were followed to compare all differentially expressed sequences in catalase deficient tobacco and Arabidopsis. In a first method, the MIPS Arabidopsis homologues of the tobacco transcript fragments were searched and their At-codes compared to these of the Arabidopsis experiment. In a second approach, the comparison of MIPS sequences was done on the protein sequence level. Thirdly, the tobacco transcript fragment was prolonged via the TIGR gene indices database, compared to the MIPS database followed by comparison on the protein level. Transcr.Fragm., tobacco transcript fragment identified in cDNA-AFLP experiment; TIGR, homology identified via TIGR, At-Seq, homology identified via At-protein sequence comparison; At, , homology identified via At-code comparison; Tob_cluster, cluster to which tobacco transcript fragment belongs; Length, length in basepair of the tobacco transcript fragment;, Description, description of the transcript fragment as identified in the tobacco experiment; TIGR_GI, accession number of TIGR gene-index corresponding to tobacco transcript fragment; evaluel, e-value of the TIGR homology; At_code_A, MIPS At-code of the Arabidopsis thaliana homologue, Bits, bits value of the homology; e-value2, e-value of the MIPS homology; Atcode_B, MIPS-At-code of the closest homologue from the Arabidopsis experiment, cluster, cluster to which the Arabidopsis transcript belongs; MIPS_description, description for the Arabidopsis gene as described in MIPS database; Seqjength, length of the shortest protein sequence; %ld, percentage of identity between the two Arabidopsis proteins; Cov_Length, length of the resulting sequence homology, % Cov, percentage of coverage relative to sequence length; Score; bit value of the homology; e-value3, , e-value of the homology between MIPS protein sequences.

REFERENCES

- A-H- ackerness, S., Surplus, S.L., Blake, P., John, C.F., Buchaπan-Wollaston, V., Jordan, B.R. and Thomas, B. (1999). Ultraviolet-B-induced stress and changes in gene expression in Arabidopsis thaliana: role of signaling pathways controlled by jasmonic acid, ethylene and reactive oxygen species. Plant Cell and Environment, 22, 1413-

1423.

- Alvarez, M.E., Pennell, R.I., Meijer, P.J., Ishikawa, A., Dixon, R.A. and Lamb, C. (1998). Reactive oxygen intermediates mediate a systemic signal network in the establishment of plan immunity. Cell, 92, 773-784. - Anderson, M.D., Prasad, T.K. and Stewart, C.R. (1995). Changes in isoizyme profiles of catalase, peroxidase and glutathione reductase during accumulation to chilling in mesocotyls of maize seedlings. Plant physiol, 109, 1247-1257.

- Bachem, C.W.,van der Hoeven, R.S., de Bruijn, S.M., Vreugdenhil, D„ Zabeau, . And Visser, R.G. (1996). Visualisation of differential gene expression using a novel method of RNA fingerprinting based on AFLP: analysis of gene expression during potato tuber development. Plant J, 9, 745-753.

- Breyne, P., Dreesen, R., Vandepoele, K., De Veylder L., Van Breusegem, F., Callewaert, L., Rombauts, S., Raes, J., Cannoot, B., Engler, G. Inze, D. And Zabeau, M. (2002) Transcriptosome analysis during cell division in plants. Proc Natl Acad Sci USA 99, 14825-14830.

- Breyne, P and Zabeau, M. (2001) Genome-wide expression analysis of plant cell cycle modulate genes. Curr Opin Plant Biol, 4, 136-142.

- Burdon, R.H., O'Kane, D., Fadzillah, N., Gill, V., Boyd, P.A., Finch, R.R.(1996) Oxidative stress and responses in Arabidopsis thaliana and Oryza sativa subjected to chilling and salinity stress. Biochem Soc Trans 24, 469-472.

- Chamnongpol, S., Willekens, H., Langebartels, C, Van Montagu, M., Inze, D. and Van Camp, W. (1996). Transgenic tobacco with a reduced catalase activity develops necrotic lesions and induces pathogenesis-related expression under high light. Plant J 10, 491-503. - Chamnongpol, S., Willekens, H., Moeder, W., Langebartels, C, Sandermann, H.Jr.,

Van Montagu, M., Inze, D. And Van Camp, W. (1998) Proc Natl Acad Sci USA, 95, 5818-5823.

- Christmann, M.F., Morgan, R.W., Jacobson, F.S. and Ames, B.N. (1985) Positive control of a regulon for defenses against oxidative stress and some heat-shock proteins in Salmonella typhimurium. Cell, 3, 753-762. - Corpas, F.J., Barroso, J.B. and del Rio, L.A. (2001). Peroxisomes as a source of reactive oxygen species and nitric oxide signal molecules in plant cells. Trends Plant Sci 6, 145-150.

Dat, J.F., Inze, D and Van Breusegem, F. (2001) Catalase deficient tobacco plants: tools for in planta studies of the role of hydrogen peroxide. Redox Rep 6, 37-42.

Dat, J.F., Lopez-Delgado, H., Foyer, CH. and Scott, I.M. (1998) Parallel changes in H₂0₂ and catalase during thermotolerance induced by salicylic acid or heat acclimation in mustard seedlings. Plant Physiol 116, 1351-1357.

- Dat, J.F., Pellinen, R., Van de Cotte, B., Langebartels, C, Kangasjarvi, J., Inzέ, D., Van Breusegem, F. (2003). Changes in hydrogen peroxide homeostasis trigger an active cell death process in tobacco. Plant J. Feb, 33(4), 621-632.

- Desikan, R., Reynolds, A., Hancock, J.T. and Neill, S.J. (1998). Harpin and hydrogen peroxide both initiate programmed cell death but have differential affects on defence gene expression in Arabidopsis suspension cultures. Biochem J 15, 115-120. - Desikan, R., A-H-Mackerness, S., Hancock, J.T. and Neill, S.J. (2001). Regulation of the Arabidopsis transcriptosome by oxidative stress. Plant Physiol 127, 159-172.

- De Smet, F., Mathys, J., Marchal, K., Thijs, G., De Moor, B. And Moreau, Y. (2002). Adaptive quality-based clustering of gene expression profiles. Bioinformatics 18, 735- 746. - Ditt, R.F., Nester, E.W. and Comai, L. (2001). Plant gene expression response to

Agrobacterium tumefaciens. Proc Natl Acad Sci USA 98, 10954-10959.

- Durrant, W.E., Rowland, O., Piedras, P., Hammond-Kosack, K.E. and Jones, J.I. (2000). cDNA-AFLP reveals a striking overlap in race-specific resistance and wound response gene expression profiles. Plant Cell 12, 963-977. - Eisen, M.B., Spellman, P.T., Brown, P.O. and Botstein, D. (1998). Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95, 14863- 14868.

- Gasch, A.P., Spellman, P.T., Kao, CM., Carmel-Harel, O., Eisen, M.B., Storz, G., Botstein, D. and Brown, P.O. (2000). Genomic expression programs in the response of yeast cells to environmental changes. Mol Biol Cell 11 , 4241 -4257.

- Godon, C, Lagniel, G., Lee, J., Buhler, J.M., Kieffer, S., Perrot, M., Boucher, H., Toledano, M.B. and Labarre, J. (1998). The H₂0₂ stimulon in Saccharomyces cerevisiae. J Biol Chem 273, 22480-22489.

- Gossele, W., Rache, I., Meulewaeter, F., Comelissens, M and Metzalff, M. (2002) SVISS - a novel transient gene silencing system for gene function discovery and validation in tobacco plants. Plant J. 32, 859-866. - Greenberg, J.T. (1997). Programmed cell death in plant-pathogen interactions. Annual Review of Plant Physiology and Plant Molecular Biology, 48, 525-545.

- Guan, L.M., Zhao, J. and Scandalios, J.G. (2000). Cis-elements and trans-factors that regulate expression of the maize Cat1 antioxidant gene in response to ABA and osmotic stress: H₂0₂ is the likely intermediary signaling molecule for the response.

Plant J 22, 87-95.

- Karimi, M., Inze, D., and Depicker, A. (2002) GATEWAY vectors for Agrobacterium- mediated plant transformation. Trends Plant Sci. 7,193-195.

- Karpinski, S., Reynolds, H., Karpinska, B., Wingsle, G., Creissen, G. and Mullineaux P. (1999). Systemic signaling and acclimation in response to excess excitation energy in

Arabidopsis. Science, 284, 654-657.

- Kingston-Smith, A.H., Harbinson, J. and Foyer, CH. (1999). Acclimation of photosynthesis, H₂0₂ content and antioxidants in maize (Zea mays) grown at sub- optimal temperatures. Plant Cell and Environment 22, 1071-1083. - Kovtun, Y., Chiu, W.L., Tena, G and Sheen, J. (2000). Functional analysis of oxidative stress-activated mitogen-activated protein kinase cascade in plants. Proc Natl Acad Sci USA 97, 2940-2945.

- Lamb, C. and Dixon, R.A. (1997). The oxidative burst in plant disease resistance. Annual Review of Plant Physiology and Plant Molecular Biology, 48, 251-275. - Langebartels, C, Ernst, D., Kangasjarvi, J. And Sandermann, H.Jr. (2000) Ozone effects on plant defense. Methods Enzymol 319, 520-535.

- Levine, A., Tenhaken, R., Dixon, R and Lamb, C. (1994). H₂0₂ from oxidative burst orchestrates the plant hypersensitive disease resistance response. Cell, 79, 583-593.

- Liu, Y.G., Mitsukawa, N., Oosumi, T. and Whittier, R.F. (1995). Efficient isolation and mapping of Arabidopsis thaliana T-DNA insert junctions by thermal asymmetric interlaced PCR. Plant J. 8, 457-463.

- Lopez-Delgado, H., Dat, J.F., Foyer, CH. and Scott, l.M. (1998). Induction of thermotolerance in potato microplants by acetylsalicylic acid and H₂0₂. J Exp Botany 49, 713-720. - Mittler, R., Shulaev, V., Seskar, M and Lam, E. (1996). Inhibition of programmed cell death in tobacco plants during a pathogen-induced hypersensitive response at low oxygen pressure. Plant cell 8, 1991-2001.

- Mittler, R. and Berkowitz, G. (2001). Hydrogen peroxide, a messenger with too many roles? Redox Rep 6, 69-72. - Neill, S., Desikan, R. and Hancock, J. (2002). Hydrogen peroxide signaling. Curr Opin

Plant Biol 5, 388-395. - Payne, R. Arnold, G. (2000). Genstat Release 6.1 Reference Manual, Part 3 Procedure Library PL14. Oxford: VSN International

- Pearson, W.R. and Lipman, D.J. (1988). Improved tools for biological sequence comparison. Proc Natl Acad Sci USA, 85, 2444-2448. - Pei, Z.M., Murata, Y., Benning, G., Thomine, S., Klϋsener, B., Allen, G.J., Grill, E.,

Schroeder, J.I. (2000). Calcium channels activated by hydrogen peroxide mediate abscisic acid signalling in guard cells. Nature, 406, 731-734.

- Prasad, T.K., Anderson, M.D., Martin, B.A. and Stewart, C.R. (1994). Evidence for chilling-induced oxidative stress in maize seedlings and a regulatory role for hydrogen peroxide. Plant cell, 6, 65-74.

- Puskas, L. G., Zvara, A., Hackler, L., Micsik, T., and van Hummelen, P. (2002a). Production of bulk amounts of universal RNA for DNA microarrays. Biotechniques 33, 898-900, 902, 904.

- Puskas, L. G., Zvara, A., Hackler, L, and Van Hummelen, P. (2002b). RNA amplification results In reproducible microarray data with slight ratio bias. Biotechniques

32, 1330-4, 1336, 1338, 1340.

- Tavazoie, S., Hughes, J.D., Campbell, M.J., Cho, R.J. and Church, G.M. (1999). Systematic determination of genetic network architecture. Nat Genet 22, 213-215.

- Toone, W.M., Morgan, B.A. and Jones, N. (2001). Redox control of AP-1-like factors in yeast and beyond. Oncogene, 20, 2336-2346.

- Vos, P., Hogers, R., Bleeker, M., Reijans, M., van de Lee, T., Homes, M., Frijters, A., Pot, J., Peleman, J., Kuiper, M. (1995). AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res, 23, 4407-4414.

- Vranova; E., Inze, D. And Van Breusegem, F. (2002) Signal transduction during oxidative stress. J Exp Bot 53, 1227-1236.

- Wolfinger, R. D., Gibson, G., Wolfinger, E. D., Bennett, L, Hamadeh, H., Bushel, P., Afshari, C, and Paules, R. S. (2001). Assessing gene significance from cDNA microarray expression data via mixed models. Journal of Computational Biology : a Journal of Computational Molecular Cell Biology 8, 625-637. - Yoneda, K., Peck, K., Chang, M.M., Chmiel, K., Sher, Y.P., Chen, J., Yang, P.C., Chen,

Y. and Wu, R. (2001) Development of high-density DNA microarray membrane for profiling smoke- and hydrogen peroxide-induced genes in a human bronchial epithelial cell line. Am J Respir Crit Care 164, S85-S89.

- Zheng, M., Wang, X., Templeton, L.J., Smuslki, D.R., LaRossa, R.A. and Storz, G. (2001). DNA microarray-mediated transcriptional profiling of the Escherichia coli response to hydrogen peroxide. J Bacteriol, 183, 4562-4570.

Claims

1. A method to isolate a gene involved in the hydrogen peroxide induced signaling pathway, comprising (1) growing at least one catalase deficient plant and one control plant (2) performing a high light treatment of said plants (3) harvesting a sample of said plants at the same time point (4) performing a differential display technique on the samples; and (5) analyzing the data.

2. The method according to claim 1 whereby said differential display technique is cDNA- AFLP analysis

3. The method according to claim 1 whereby said differential display technique is microarray analysis.

4. The method according to any of the claims 1-3 whereby said plant is tobacco.

5. The method according to any of the claims 1 -3 whereby said plant is Arabidopsis thaliana

6. The method according to any of the claims 1-4, whereby said catalase deficient plant is CAT1AS.

7. The method according to any of the claims 1, 2, 3 or 5, whereby said catalase deficient plant is CAT2AS or CAT2HP1.

8. The method according to any of the previous claims, whereby said analysis comprises a quantification of test and control samples, followed by sequence analysis of the differentially displayed transcripts.

9. A gene, identified with the method according to any of the previous claims

10. A gene according to claim 9, selected from the group consisting of the sequences comprising SEQ ID N° 1 to SEQ ID N° 712.

11. A gene according to claim 9 or 10, selected from the group consisting of the sequences comprising SEQ ID N° 409 to SEQ ID N" 712.

12. A gene according to claim 9 or 10, selected from the group consisting of the sequences comprising SEQ ID N° 35, SEQ ID N° 131, SEQ ID N° 142, SEQ ID N° 208, SEQ ID N° 321 and SEQ ID N° 326.

13. A gene according to claim 9, encoding a protein selected from the group consisting of the sequences comprising SEQ ID N° 713 to SEQ ID N° 730.

14. The use of a gene, isolated with the method according to claim 1-8, to obtain stress tolerance.

15. The use of a gene, according to claim 14, whereby said gene is selected from the group consisting of the sequences comprising SEQ ID N° 1 to SEQ ID N° 712, or a homologue thereof.

16. The use of a gene, according to claim 15, whereby said gene is selected from the group consisting of the sequences comprising SEQ ID N° 409 to SEQ ID N° 712, or a homologue thereof.

17. The use of a gene, according to claim 14, whereby said gene is selected from the group consisting of the sequences comprising SEQ ID N° 35, SEQ ID N° 131, SEQ ID

N° 142, SEQ ID N° 208, SEQ ID N° 321 and SEQ ID N° 326, or a homologue thereof.

18. The use of a gene, according to claim 14, whereby said gene is encoding a protein selected from the group consisting of the sequences comprising SEQ ID N° 714 to SEQ ID N° 730, or a homologue thereof.

19. The use of a gene, encoding a protein comprising SEQ ID N°713, or a homologue thereof, to obtain stress tolerance.

20. The use according to any of the claims 14-19, whereby said stress is oxidative stress.