The transcription factor SlAREB1 confers drought, salt stress tolerance and regulates biotic and abiotic stress-related genes in tomato


J. A. Casaretto. Fax: +56-71 200276; e-mail:


Members of the abscisic acid-responsive element binding protein (AREB)/abscisic acid-responsive element binding factor (ABF) subfamily of basic leucine zipper (bZIP) transcription factors have been implicated in abscisic acid (ABA) and abiotic stress responses in plants. Here we describe two members identified in cultivated tomato (Solanum lycopersicum), named SlAREB1 and SlAREB2. Expression of SlAREB1 and SlAREB2 is induced by drought and salinity in both leaves and root tissues, although that of SlAREB1 was more affected. In stress assays, SlAREB1-overexpressing transgenic tomato plants showed increased tolerance to salt and water stress compared to wild-type and SlAREB1-down-regulating transgenic plants, as assessed by physiological parameters such as relative water content (RWC), chlorophyll fluorescence and damage by lipoperoxidation. In order to identify SlAREB1 target genes responsible for the enhanced tolerance, microarray and cDNA-amplified fragment length polymorphism (AFLP) analyses were performed. Genes encoding oxidative stress-related proteins, lipid transfer proteins (LTPs), transcription regulators and late embryogenesis abundant proteins were found among the up-regulated genes in SlAREB1-overexpressing lines, especially in aerial tissue. Notably, several genes encoding defence proteins associated with responses to biotic stress (e.g. pathogenesis-related proteins, protease inhibitors, and catabolic enzymes) were also up-regulated by SlAREB1 overexpression, suggesting that this bZIP transcription factor is involved in ABA signals that participate in abiotic stress and possibly in response to pathogens.


abscisic acid


abscisic acid-responsive element binding protein


basic leucine zipper


Drought and salinity are two major environmental conditions that retard plant growth and severely decrease agricultural productivity. Plants can minimize drought and salt injury by adaptation mechanisms that have evolved to provide resistance to different harmful conditions (Umezawa et al. 2006). Enhanced drought and salt tolerance in crops have included traditional and marker-assisted breeding, but the complexity of the tolerance mechanisms at the physiological and genetic levels has influenced the use of genetic engineering approaches (Takeda & Matsuoka 2008; Mittler & Blumwald 2010). Thus, biotechnological efforts, for example, have made use of genes whose products encode for proteins that regulate ion homeostasis, enzymes responsible for osmotic adjustment, molecular chaperones and enzymes that contribute to cellular detoxification (Umezawa et al. 2006).

As part of the activated signalling pathways in response to environmental stimuli, the phytohormone abscisic acid (ABA) plays an important role in regulating gene expression. ABA levels increase dramatically in vegetative tissues with decreasing levels of water potential (Seo & Koshiba 2003) and regulate the expression of numerous genes under stress conditions such as drought, salt stress, adaptation to low temperatures and during acquisition of desiccation tolerance in maturing seeds (Christmann et al. 2006). These signal transduction pathways include regulatory points controlled by transcription factors that activate or repress sets of genes to confer different degrees of stress resistance (Yamaguchi-Shinozaki & Shinozaki 2005). Among the numerous transcription factors identified in plants, few are known to participate in regulating gene expression in ABA-dependent processes (Shinozaki & Yamaguchi-Shinozaki 2007). One major ABA-dependent gene activation pathway is mediated by a specific subfamily of basic leucine zipper (bZIP) transcription factors that recognize abscisic acid-responsive elements (ABREs) known as abscisic acid-responsive element binding proteins (AREBs) (Uno et al. 2000) or abscisic acid-responsive element binding factors (ABFs) (Choi et al. 2000). Orthologous genes have been reported in species such as Arabidopsis (Choi et al. 2000; Finkelstein & Lynch 2000; Uno et al. 2000) and cereals (Casaretto & Ho 2003; Lu et al. 2009). Functional characterization of these transcription factors has revealed that they can confer resistance to drought and salt stress when overexpressed in Arabidopsis (Kim et al. 2004; Fujita et al. 2005).

The Solanaceae family, which includes potato, tobacco, tomato and pepper, represents the most valuable family of vegetable crops. Although these species present adequate adaptation plasticity, only some cultivars within each species display moderate tolerance to abiotic stress. For instance, in most tomato cultivars, salinity produces detrimental effects such as reduction of seed germination, inhibition of growth and decreased fruit productivity (Cuartero et al. 2006). A small number of transcriptonal regulators have demonstrated to be involved in abiotic stress responses in the Solanaceae, such as AIM1 (Abuqamar et al. 2009), LebZIP2 (Seong et al. 2008), JERF1 (Wu et al. 2008), StEREBP1 (Lee et al. 2007), CabZIP1 (Lee et al. 2006) and TERF1 (Zhang et al. 2005). In a previous study, we described a bZIP transcription factor, named SlAREB1, identified first in two salt- and drought-resistant tomato wild relatives, Solanum lycopersicum and Solanum chilense, and in cultivated tomato (Yañez et al. 2009). In these three species, the Solanum AREB1 ortholog responds to desiccation and high salt. In addition, SlAREB1 was able to transiently activate the expression of few marker stress genes in tobacco, suggesting that it may participate in abiotic stress responses in the Solanum genus (Yañez et al. 2009).

In this article, we present the characterization of a second tomato AREB/ABF family member gene, SlAREB2. We also investigated the functional role of SlAREB1 by generating sense and antisense transgenic tomato plants that turned out to be more tolerant and more susceptible than wild-type (WT) plants, respectively. Large-scale gene expression profiling approaches (i.e. microarrays and cDNA-AFLP) revealed that SlAREB1 affects transcription of genes associated with salinity, drought and oxidative stresses. Interestingly, among the induced genes, several genes encoding proteins that normally participate in responses to pathogen attack were also identified. Consequently, these results suggest that SlAREB1 plays a significant role in responses to environmental stresses and also may participate as a link of ABA signalling to responses to biotic stimuli.


Plant material and experimental treatments

S. lycopersicum (L.) Miller cv. Moneymaker seeds were washed and sterilized before germinating them in a soil:perlite mixture (1:1). Seedlings were then transferred to 2 L pots with a vermiculite:perlite mixture (1:1) as substrate. Plants were irrigated with a 1 g L−1 Murashige and Skoog mineral nutrient solution and were grown in growth chambers in a 16 h daylight and 8 h dark period at 20–23 °C.

Isolation of SlAREB1 and SlAREB2

cDNA synthesis of SlAREB1 and SlAREB2 was performed by RT-PCR with total RNA extracted from leaves (100 mg) from 10-week-old plants using the SV Total RNA Isolation System (Promega, Madison, WI, USA) following the manufacturer's instructions and an RT-PCR. Cloning of SlAREB1 was described in Yañez et al. (2009). Briefly, the nucleotidic sequences described in the GenBank database (LeAREB, accession number AY530758, herein SlAREB1) and a 1095 bp sequence designated SlAREB2 were isolated by PCR amplification using the specific primers SlA1-FWD, 5′-ATGGGGAGTAATTATCATTTCAAGAAC-3′ and SlA1-REV, 5′-TTACCATGGACCAGTTTGTGTCCGTCT-3′ for SlAREB1 and SlA2-FWD, 5′-ATGGGATCTTACCTGAACTTCAAGAACTTTGC-3′ and SlA2-REV, 5′-CCAAGGTCCCGTCACTGTCCTTCG-3′ for SlAREB2. Nucleotide sequences were compared to those described for tomato in The Gene Index (TGI) database ( and deduced amino acid sequences were aligned using the BioEdit sequence alignment editor tool ( Phylogenetic analyses based on the resultant alignments were then constructed by using the neighbour-joining method with the MEGA 4 programme (Tamura et al. 2007).

Expression analyses by qRT-PCR

For gene expression analysis of SlAREB1 and SlAREB2 in WT and transgenic plants, leaves and roots were sampled at different time points of each treatment, flash-frozen in liquid nitrogen and kept at −80 °C. Total RNA was extracted (100 mg for each sample) using the SV Total RNA Isolation System (Promega). RNA samples (2 µg) were treated with RNase-free DNase I (Invitrogen, Carlsbad, CA, USA). RNA integrity was visualized in 1% agarose gels, and concentration and purity (260/280 ratio) were determined with a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). RNA samples were reverse transcribed in a 20 µL reaction using the AffinityScript QPCR cDNA Synthesis Kit (Stratagene, La Jolla, CA, USA) for first-strand cDNA synthesis according to the manufacturer's instructions. qRT-PCR was employed for measuring transcript levels and was carried out in 96-well plates with a DNA Engine Opticon 2 System (MJ Research, Waltham, MA, USA) for continuous fluorescence detection. Each 20 µL reaction contained 2 µL of cDNA sample (100 ng), 10 µL 2× SYBR Green Master Mix Reagent (Stratagene) and 0.25 µM of gene-specific primer mix. The thermal cycle used was as follows: 95 °C for 10 min, then 40 cycles of 95 °C for 15 s, 58–60 °C for 15 s and 72 °C for 20 s. The tomato GAPDH gene (accession U97257) was used as constitutive control. qRT-PCR analyses were performed at least three times using sets of cDNA samples from independent plants. Data from qRT-PCR experiments were analysed according to Livak & Schmittgen (2001). Specific primers were designed using the Beacon Designer Software (Bio-Rad, Hercules, CA, USA) and were listed in Supporting Information Table S1.

Generation of transgenic plants

The full-length cDNA of SlAREB1 was inserted in sense orientation into the XbaI-SstI sites to replace the β-glucuronidase (GUS) gene in the pBI121 binary vector (Clontech, Palo Alto, CA, USA) driven by the CaMV 35S promoter. The resulting construct in sense orientation, 35S-SlAREB1(+), was confirmed by sequencing. The SlAREB1 cDNA was also inserted in the reverse orientation into the pBI121 vector, resulting in the antisense construct 35S-SlAREB1(−). These constructs were introduced into Agrobacterium tumefaciens LBA4404 by electroporation. To generate transgenic S. lycopersicum cv. Moneymaker plants, Agrobacterium-mediated transformation of cotyledon explants was performed according to Fillati, Kiser & Rose (1987) separately with both constructs. The presence of the transgene in the regenerating plantlets was confirmed by PCR using the forward primer 5′-CCACTATCCTTCGCAAGAC-3′, which anneals to the 35S promoter region, and the reverse primer 5′-TTACCATGGACCAGTTTGTGTC-3′ or 5′-ATGGGGAGTAATTATCATTTCAA-3′ for sense and antisense lines, respectively, which anneal to the SlAREB1 coding region. Genomic DNA was isolated using a commercial kit (DNeasy Plant Mini Kit; Qiagen, Valencia, CA, USA).

Stress assays and ABA treatment

Ten-week-old plants of WT and sense and antisense transgenic plants were grown in 2 L pots with a vermiculite:perlite mixture (1:1) and were irrigated with a 1.1 g L−1 Murashige and Skoog mineral nutrient solution and kept in chambers in a 16 h light/8 h dark regime at 20–23 °C. Stress treatments were performed as described by Yañez et al. (2009). The drought treatment consisted of withholding water for up to 30 d. Well-watered plants were maintained as controls by watering plants daily. A set of WT plants was also subjected to water deficit shock by removing them from the pots and allowing them to dry on the laboratory bench for up to 24 h. For salt stress, the plants were treated with a 300 mm NaCl solution (200 mL) every 72 h and were kept in the same growing conditions described above. Relative water content (RWC) was estimated in all stress treatment as RWC = (fresh weight − dry weight)/(turgid weight − dry weight) × 100. Fresh weight was measured using a scale; turgid weight was determined by soaking samples in distilled water for 24 h at room temperature; and dry weight was determined by placing the same samples at 100 °C until constant weight was reached. Survival rate was recorded after 14 d of recovery from 25 d of drought stress or 21 d after salt treatment and was defined as the number of healthy plants divided by the total number of plants. Pictures were taken to record the phenotypes. For ABA treatment, plants were sprayed with a 100 µM ABA solution (Sigma A1012; Sigma, St Louis, MO, USA) before leaves were sampled for total RNA extraction. The assays were repeated three times and were arranged with control and stressed/treated plants with six pots per replicate.

Chlorophyll content measurement

Leaves were excised from control and stressed plants and were weighed. Each sample consisted of six randomly selected folioles from three individual plants. Samples were ground with a mortar and pestle and were extracted with 80% acetone (v/v). Absorbance of the supernatant was measured using a Hitachi U-2000 spectrophotometer and chlorophyll content was estimated as described by Lichtenthaler & Wellburn (1983).

Measurement of chlorophyll fluorescence

Chlorophyll fluorescence was measured in six plants per line per treatment using a portable Fluorescence Monitoring System FMS2 (Hansatech Instruments, Norfolk, UK) according to the protocol described by van Kooten & Snel (1990). The minimal florescence (Fo) was determined with a low modulated light without affecting variable fluorescence (Fv). The maximal fluorescence (Fm) was determined by 0.8 s saturating light of 10 000 µmol m−2 s−1 and after adaptation to white actinic light (400 µmol m−2 s−1) on nine dark-adapted (30 min) leaves in each transgenic line. The maximal photochemical efficiency (Fv/Fm) of photosystem II (PSII) was estimated as Fv/Fm = (Fm − Fo)/Fm. The maximum intensity of fluorescence in light-acclimated leaves (Fm′) and levels of fluorescence during a short light-saturating pulse at a point below Fm′ were then measured and used to estimate the effective quantum yield of PSII photochemistry ΦPSII = (Fm′ − Ft′)/Fm′ (Maxwell & Johnson 2000).

Malondialdehyde (MDA) content assay

MDA, a product of fatty acid peroxidation, was determined by the thiobarbituric acid (TBA) reaction method according to Heath & Packer (1968). Leaf tissue (0.5 g) was ground with a cold mortar and pestle in 5 mL 20% trichloroacetic acid (TCA) reagent. The homogenate was centrifuged at 5000 g for 10 min and the supernatant was used for determination. Each reaction sample containing 1 mL extract and 1 mL 0.5% TBA reagent [0.5% (m/v) TBA dissolved in 20% (m/v) TCA] was heated at 95 °C for 25 min, quickly cooled on ice and centrifuged at 5000 g for 10 min. Absorbance was read at 532 and 600 nm to subtract non-specific absorption, and MDA content was calculated by using an extinction coefficient of 155 mm−1 cm−1.

Microarray analysis

Total RNA was obtained from leaves and roots from WT and SlAREB1-overexpressing line S3 (three biological replicates each) using 100 mg of flash-frozen, pulverized tissue-cultured plants and the RNeasy Plant Mini Kit (Qiagen) following the manufacturer's instructions. RNA recovery and purity were determined spectrophotometrically. Labelling and gene chip (Tomato Affymetrix GeneChip Genome Array, representing more than 9200 transcripts) hybridization and signal generation and detection were carried out by the Genome Explorations Inc. Affymetrix core facility (Memphis, TN) according to the manufacturer's protocols. A total of 12 chips were run for the six leaf and six root RNA samples (i.e. no technical replicates). The final log2 signal values and ‘present’ calls were generated using the Affymetrix MAS software version 5.0. The signal intensity for each gene was calculated as the average intensity difference, represented by [z(PM-MM)/(number of probe pairs)], where PM and MM denote perfect-match and mismatch probes. A significance analysis test (unpaired t-test with Benjamini–Hochberg false discovery rate correction) was performed to test the equality of the mean signal values between the WT and the transgenic line. Fold change was calculated as the ratio of overall signal values from the two groups. Differentially expressed genes were identified as having P-values ≤0.05 for all samples in one group, fold change ≥2.0 (increase or decrease) and t-test P-values ≤0.05. Differentially expressed genes were then annotated using the NetAffx database provided by Affymetrix ( In some cases, the annotations of unknown genes were further refined by performing additional Basic Local Alignment Search Tool (BLAST) searches.

cDNA-AFLP analysis

cDNA-AFLP analysis was performed basically according to Vos et al. (1995) with minor modifications. Total RNA from leaves was extracted from tissue-cultured WT and transgenic tomato plants (0.5 g) using the SV Total RNA Isolation System (Promega). Double-stranded cDNA was synthesized from 5 µg DNaseI-treated RNA using the SuperScript Double-Stranded cDNA Synthesis Kit (Invitrogen) according to the manufacturer's instructions. Five hundred nanograms of double-stranded cDNA was used for AFLP analysis and was digested with EcoRI and MseI. Digested fragments were ligated to EcoRI and MseI adapters to generate the cDNA template for amplification. PCR was performed in two consecutive reactions: pre-amplification using primers complementary to the adapters followed by a selective amplification of a 1:10 dilution of the previous PCR products using two AFLP primers containing three selective nucleotides. AFLP reactions were performed according to the AFLP Analysis System Manual (Invitrogen). Amplification products were separated on 6% polyacrylamide gels where the size resolution ranged between 50 and 500 bp. DNA fragments were visualized by silver staining using the Silver Sequence DNA Sequencing System (Promega). Transcript-derived fragments (TDFs) that showed clear differential amplification in transgenic versus WT plants were excised from the gels and were extracted using the QIAquick Gel Extraction Kit (Qiagen). TDFs were recovered by PCR under the same conditions as used in the selective amplification. PCR products were purified, cloned into the pGEM-T vector (Promega) and subsequently sequenced (Macrogen, Seoul, Korea). All sequences were analysed for homology using the BLASTN and BLASTX algorithms by comparison with the GenBank database and the TGI database ( Confirmation of expression of selected genes from the microarray and cDNA-AFLP data was performed by qRT-PCR as previously described using total RNA samples from the transgenic lines S2, S3, A4 and A6 and with specific primers for each tomato gene.

Statistical analysis

The statistical software package STATISTICA 6.0 ( was used for data analysis. An analysis of variance followed by multiple comparison with Tukey's honestly significant difference (HSD) mean separation test was performed to determine the statistical significance of differences of the mean values at P ≤ 0.05.


Identification of AREB-like sequences in tomato

As a result of studying stress-related genes in wild and cultivated tomato species, a gene encoding a bZIP protein, SlAREB1, homologous to a cDNA clone (accession AY530758), was previously identified (Yañez et al. 2009). Further homology searches in the TGI database predicted the existence of at least one additional full-length cDNA (TC175608) and another partial sequence for AREB-like genes. This second full-length cDNA sequence, herein designated as SlAREB2, has 1095 bp and encodes a protein of 365 amino acids. Both predicted proteins, SlAREB1 and SlAREB2, are related to the AREB/ABF/ABI5 subfamily of bZIP transcription factors described in Arabidopsis and cereals (Lu et al. 2009). The nucleotide sequences of SlAREB1 and SlAREB2 share a 53% identity. SlARERB1 presents high homology with a cDNA from Vitis vinifera (73% identity) and is 66% identical to the ABF2/AtAREB1 bZIP factor from Arabidopsis (Yañez et al. 2009). However, SlAREB2 presents significant homology (81% identity) with the Nicotiana tabacum PHI-2 gene corresponding to a phosphate-induced transcription factor (Sano & Nagata 2002; Fig. 1) and is 51% identical to AtABF3. Comparison of their deduced amino acid sequences are indicative of bZIP transcription factors with a basic region and four heptad repeats of leucines and methionines forming the leucine zipper domain located in the C-terminus. These SlAREBs also present four highly conserved regions outside the basic domain, three clustered at the N-terminal half and another located in the C-terminus (Fig. 1a). Phylogenetic analysis assembled with the predicted amino acid sequences clustered SlAREB1 with homologous bZIPs from different angiosperm species (Beta vulgaris, V. vinifera, Populus trichocarpa and Arabidopsis AREB1), whereas SlAREB2 was grouped with NtPHI-2 and other AREB/ABF factors from Arabidopsis. Both tomato proteins appear separated from four related monocot proteins (Fig. 1b).

Figure 1.

Amino acid sequence comparison of AREB-like bZIP factors identified in S. lycopersicum cv. Moneymaker with homologous sequences. Amino acid sequences of SlAREB1 and SlAREB2 were deduced from nucleotide sequences of cDNA isolated by RT-PCR. (a) Sequences were aligned with a similar sequence from V. vinifera (VvbZIP) and the closest homolog from Arabidopsis (AtAREB1) and from a solanaceous species, NtPHI-2. Conserved regions of AREB-like bZIPs are underlined, and the bZIP signature (basic region and Leu zipper) is indicated over a double line. Identical and similar residues between the sequences are shaded in black and grey, respectively. (b) Phylogenetic relationship among SlAREBs and AREB-like proteins. Bootstrap values from 1000 replicates were used to asses the robustness of the tree. Accession numbers of sequences used in a multiple alignment used to generate the tree are Arabidopsis thaliana AtAREB1 (BAB12404), AtAREB2 (BAB12405), AtAREB3 (NP_191244), AtABF1 (NP_564551), AtABF3 (NP_849490), AtbZIP15 (CAD11866), AtbZIP67 (Q8RYD6); Beta vulgaris BvAREB1 (CAP66259); Hordeum vulgare HvABI5 (AAO06115); N. tabacum NtPHI-2 (BAB61098); Oryza sativa OsbZIP23 (BAG92808) and OsTRAB1 (BAA83740); Phaseolus vulgaris PvbZIP6 (AAK39132); P. trichocarpa PtbZIP (XP_002302435); VvbZIP (CAB85632); and Zea mays (NP_001150949).

Expression of SlAREB1 and SlAREB2 increases under drought, salt stress and ABA

The expression profiles of SlAREB1 and SlAREB2 were examined in tomato plants exposed to a desiccation shock and to severe salt stress (300 mm NaCl). Transcript levels of SlAREB1 were rapidly induced in leaves by water deficit shock treatment after 2 h, reaching maximum levels of 30-fold at 6 h and remaining almost constant until the last time point measured (Fig. 2a). Up-regulation of SlAREB1 was also observed in roots, although this time, the induction gradually increased after 2 h post-treatment, reaching a maximum level of ca. 10-fold at 12 h (Fig. 2b). Interestingly, the magnitude of induction of SlAREB2 in both tissues was about twofold, much lower than that of SlAREB1 during the time course of the assays (Fig. 2). SlAREB1 was also induced transiently by high salinity in leaves and in roots (13- and 4-fold at 6 h, respectively). However, transcript levels of SlAREB2 remained almost unchanged (Fig. 3). These results suggest that probably, SlAREB1, but not SlAREB2, has a predominant participation in the regulation of gene expression associated with stress responses in tomato.

Figure 2.

Expression profiles of SlAREB1 and SlAREB2 in wild-type tomato plants under water deficit shock. Gene expression analyses were examined by qRT-PCR using total RNA from leaves (a) and roots (b) of 10-week-old tomato plants subjected to water deficit shock for 0, 2, 6, 12 and 24 h. Bars indicate mean relative expression values ± standard error (n = 3), normalized with GADPH as a constitutive expressed gene. The bars with different letters are significantly different from each other (P < 0.05). Capital and lowercase letters were used for SlAREB1 and SlAREB2, respectively.

Figure 3.

Expression profiles of SlAREB1 and SlAREB2 in wild-type tomato plants under salt stress. Gene expression analyses were examined by qRT-PCR using total RNA from leaves (a) and roots (b) of 10-week-old tomato plants after treatment with 300 mm NaCl for 0, 6, 12, 24, 48 and 72 h. Bars indicate mean relative expression values ± standard error (n = 4), normalized with GADPH as a constitutive expressed gene. The bars with different letters are significantly different from each other (P < 0.05). Capital and lowercase letters were used for SlAREB1 and SlAREB2, respectively.

WT tomato plants were treated with ABA to evaluate the regulation of these genes by this stress-related hormone. As shown in Fig. 4, the expression levels of SlAREB1 and SlAREB2 were up-regulated in leaves and in roots after 2 h of ABA application, with ca. threefold induction at 12 and 2 h in leaves and in roots, respectively, suggesting that both transcription factors have a potential role in regulating ABA-induced gene expression.

Figure 4.

Expression of SlAREB1 and SlAREB2 in leaves of tomato plants treated with ABA. Gene expression analyses were examined by qRT-PCR using total RNA from leaves (a) and roots (b) of tomato plants treated with 100 µM ABA for 0, 1, 2, 8, 12 and 24 h. Bars indicate mean relative expression values ± standard error, normalized with GADPH. The bars with different letters are significantly different from each other (P < 0.05). Capital and lowercase letters were used for SlAREB1 and SlAREB2, respectively.

Constitutive expression of SlAREB1 in transgenic tomato enhances tolerance to salt stress and drought

Since SlAREB1 was strongly up-regulated by ABA, water deficit and salt stress, stable expression experiments were used to investigate whether this transcription factor is involved in stress-dependent gene regulation in transgenic tomato plants. Over 20 sense and antisense independent transformed lines were generated. The presence of T-DNA was examined by PCR, and expression of SlAREB1 was evaluated through qRT-PCR. Four sense (S1, S2, S3, and S6) and four antisense (A1, A3, A4, and A6) lines were selected and propagated in order to consider T2 transgenic lines with different expression levels of SlAREB1 for stress tolerance assays (Fig. 5). The relative expression level of SlAREB1 in these sense lines varied from 12-fold (S6) to 23-fold (S2) with respect to WT plants. In antisense lines, expression of the endogenous SlAREB1 was down-regulated to about 15–35% of the expression level in WT plants (Fig. 5). Under normal growth conditions, the transgenic lines showed no abnormal morphological phenotype, except for the S3 sense plants that were slightly shorter (ca. 10% reduction in height) compared with WT and antisense plants.

Figure 5.

SlAREB1 transcript levels in transgenic S. lycopersicum plants. Expression analysis of SlAREB1 in leaves of transgenic lines that overexpress (sense) or down-regulate the expression of SlAREB1 (antisense) was performed by qRT-PCR. Bars indicate mean relative expression values ± standard error (n = 3), normalized with GAPDH. Asterisks indicate significant differences between WT and transgenic lines (P < 0.01). Integration of the T-DNA into the plant genome was confirmed by PCR and is indicated above each graph. WT, wild type; S1–S6, overexpressing lines; A1–A6, antisense lines.

Salt stress assays were carried out with these plants. All antisense lines showed signs of damage that became more notorious during the time course of the assays. Stress-induced symptoms were visible 15 d after salt treatment with chlorosis and wilting of lower leaves. At 21 d, these plants exhibited reduced growth and increased chlorosis and necrosis. After 30 d, the plants were totally chlorotic and showed a collapse of the shoot tissue (Fig. 6a). The same was true for WT plants, though signs of stress appeared few days later than in the antisense lines. However, SlAREB1-sense lines exhibited no significant decline in shoot biomass, displaying less chlorosis, wilting and necrosis during the assay (Fig. 6a). After 21 d of treatment, over 80% of the SlAREB1-sense lines survived with less injury, whereas WT plants and antisense lines were notably affected. The lowest survival rates (less than 20%) were observed in antisense lines A1, A3 and A6 (Fig. 6b). Damage was evaluated by measuring diverse physiological parameters. RWC in a WT plants decreased to about 65 and 55% at 15 and 30 d post-treatment, respectively. In antisense lines, this parameter went down to about 55–60% at 15 d and to 30–40% at 30 d post-treatment (Fig. 6c). Reduction of RWC in SlAREB1 overexpression followed a similar pattern but to only ca. 65% at 30 d, demonstrating that sense lines had higher water content related to the higher turgor observed compared with WT and antisense plants. In addition, total chlorophyll content correlated with the chlorosis observed in the stressed plants. Whereas WT plants lost about 20% of chlorophylls at day 30, in antisense and sense lines, this reduction was, in average, about 40 and 8%, respectively (Fig. 6d). Chlorophyll fluorescence was used to screen the photosynthetic performance of the PSII stability that reflects the level of cellular damage. The maximum quantum efficiency of PSII (Fv/Fm) and the quantum yield of PSII (ΦPSII) were measured at different times during treatment. At the beginning of the assays, WT and transgenic plants exhibited normal high values of Fv/Fm; however, salt treatment caused a constant and marked inhibition of PSII for both antisense and WT plants, as revealed by a decrease in the Fv/Fm values from 0.8 to 0.25 in antisense lines and down to 0.4 in WT plants. In contrast, no significant differences in Fv/Fm values were observed in the SlAREB1-sense lines during the first 21 d of treatment (Fig. 6e). These results followed the same trend when ΦPSII values were measured (Fig. 6f). SlAREB1-antisense and WT plants showed an important decrease of ΦPSII during week 1 of treatment to values of ca. 0.1 and 0.2 for antisense lines and WT plants, respectively. In two SlAREB1-sense lines (S2 and S3), these values were maintained between 0.5 and 0.45, whereas in line S1, ΦPSII values decreased to 0.4 during week 1 and then to 0.37 at day 30, values always above those of WT plants. In addition, since stress conditions such as salinity and drought generate reactive oxygen species (ROS), thus inducing oxidative damage to macromolecules as membrane lipids, we measured MDA as marker for lipid peroxidation. MDA levels increased rapidly in antisense and WT plants to more than threefold over those in sense lines at 30 d (Fig. 6g), suggesting that overexpression of SlAREB1 induces potential antioxidative processes preventing significant membrane damage.

Figure 6.

Figure 6.

Expression of SlAREB1 in tomato affects tolerance to salt stress. Ten-week-old tomato plants were treated with 300 mm NaCl (200 mL every 72 h). (a) Pictures show representative plants at 0, 15, 21 and 30 d post-treatment. (b) Survival rate of the plants shown after 21 d of salt treatment and 14 d of recovery from salt stress. Asterisks above each column indicate a significant difference (P < 0.05) between WT and transgenic lines. Each bar represents an average of nine plants ± standard error. (c) Leaf relative water content (RWC) of plants at 0, 15 and 30 d (mean ± standard error; n = 6). (d) Chlorophyll pigment content of plants at 0, 15 and 30 d (mean ± standard error; n = 6). (e) Chlorophyll fluorescence analysis indicating maximum photosynthetic efficiency of photosystem II (PSII) (Fv/Fm) of the plants during the assay. (f) Operating quantum yield of PSII (ΦPSII). Photochemistry parameters indicate mean values ± standard error (nine measurements in three independent experiments). (g) Malondialdehyde (MDA) content in the plants during the assay (mean ± standard error of three independent experiments). WT, wild type; A, antisense lines; S, sense lines; FW, fresh weight.

Figure 6.

Figure 6.

Expression of SlAREB1 in tomato affects tolerance to salt stress. Ten-week-old tomato plants were treated with 300 mm NaCl (200 mL every 72 h). (a) Pictures show representative plants at 0, 15, 21 and 30 d post-treatment. (b) Survival rate of the plants shown after 21 d of salt treatment and 14 d of recovery from salt stress. Asterisks above each column indicate a significant difference (P < 0.05) between WT and transgenic lines. Each bar represents an average of nine plants ± standard error. (c) Leaf relative water content (RWC) of plants at 0, 15 and 30 d (mean ± standard error; n = 6). (d) Chlorophyll pigment content of plants at 0, 15 and 30 d (mean ± standard error; n = 6). (e) Chlorophyll fluorescence analysis indicating maximum photosynthetic efficiency of photosystem II (PSII) (Fv/Fm) of the plants during the assay. (f) Operating quantum yield of PSII (ΦPSII). Photochemistry parameters indicate mean values ± standard error (nine measurements in three independent experiments). (g) Malondialdehyde (MDA) content in the plants during the assay (mean ± standard error of three independent experiments). WT, wild type; A, antisense lines; S, sense lines; FW, fresh weight.

To determine whether the transgenic lines were tolerant to drought conditions, WT and transgenic T2 plants were grown in the same pots and growing conditions. Water was withheld for up to 35 d and then physiological parameters were measured. Whereas WT plants started to show desiccation symptoms between 15 and 20 d post-treatment, the four antisense lines showed the same symptoms 4–6 d earlier. On the contrary, SlAREB1-overexpressing lines started to lose turgor after day 25 (Fig. 7a). After 25 d of drought stress and 14 d of watering, between 75 and 100% of the SlAREB1-sense lines survived, while 45% of WT and between 0 and 20% of antisense lines recovered healthy leaves (Fig. 7b). Notoriously, line S3 could stay turgid even up to 40 d with no watering, when all the rest dried out (not shown). All plants lost water content continuously until the last day measured (25 d), when most antisense and WT plants were wilted (RWC of ca. 40%, Fig. 7c). At that time point, RWC in sense lines ranged between 60 and 75%. Chorophyll content, photosynthetic efficiency and quantum yield also reflected the level of damage during the drought assays (Fig. 7d–f). Inhibition of PSII for both SlAREB1-antisense lines was constant from day 15, when loss of turgor was obvious. In WT plants, however, this decline was delayed but severe after 25 d. The same was true for three sense lines except for line S3, which stayed green the whole period. Nevertheless, at the end of the experiment, all sense lines were less chlorotic and displayed higher Fv/Fm values than WT plants (Fig. 7d,e). These results reveal that the expression level of SlAREB1 correlates with the degree of tolerance to salinity and water deficit in transgenic tomato.

Figure 7.

Figure 7.

Drought tolerance of transgenic tomato plants. A water withholding assay was performed with 10-week-old plants for up to 35 d. (a) Images of representative plants of each group at 1, 15 and 25 d after initiation of the assay. (b) Survival rate of the plants shown 14 d after drought stress for 25 d. Asterisks above each column indicate a significant difference (P < 0.05) between WT and transgenic lines. Each bar represents an average of nine plants ± standard error. (c) Relative water content (RWC) (mean ± standard error; n = 6). (d) Chlorophyll pigment content of plants at 0, 15 and 35 d (mean ± standard error; n = 6). (e) Chlorophyll fluorescence (Fv/Fm) and (f) photosystem II (PSII) quantum yield (ΦPSII) of the plants during the assay (mean values ± standard error of nine measurements in three independent experiments). WT, wild type; A, antisense lines; S, sense lines.

Figure 7.

Figure 7.

Drought tolerance of transgenic tomato plants. A water withholding assay was performed with 10-week-old plants for up to 35 d. (a) Images of representative plants of each group at 1, 15 and 25 d after initiation of the assay. (b) Survival rate of the plants shown 14 d after drought stress for 25 d. Asterisks above each column indicate a significant difference (P < 0.05) between WT and transgenic lines. Each bar represents an average of nine plants ± standard error. (c) Relative water content (RWC) (mean ± standard error; n = 6). (d) Chlorophyll pigment content of plants at 0, 15 and 35 d (mean ± standard error; n = 6). (e) Chlorophyll fluorescence (Fv/Fm) and (f) photosystem II (PSII) quantum yield (ΦPSII) of the plants during the assay (mean values ± standard error of nine measurements in three independent experiments). WT, wild type; A, antisense lines; S, sense lines.

SlAREB1 regulates abiotic and biotic stress-related genes

In order to identify target genes regulated by SlAREB1, Affymetrix tomato microarrays were used to compare transcripts from leaves and roots from WT and one SlAREB1-overexpressing line (S3) grown in normal conditions. Two hundred seventy-seven and 73 probe sets were identified as exhibiting differential expression (log2 > 0.1 and P ≤ 0.05) in leaves and in roots, respectively, between the transgenic line versus the WT. The resulting data indicate that most of the differentially expressed transcripts correspond to induced genes (230 in leaves and 51 in roots, Supporting Information Table S2). Annotation of those genes was performed using the Affymetrix NetAffx database, and, in some cases, the annotations of unknown genes were further refined by performing additional BLAST searches. In spite of this, a significant number of expressed sequences ended up with no known gene ontology annotation for biological process or molecular function because of the scarce information for tomato genes. From the differentially expressed genes with some function information, over 50 are related to general responses to abiotic and biotic stresses. Others are involved in metabolism, cell wall modification, transport, transcription and translation, and protein modification. Highly induced genes in leaf tissue are represented in Table 1. Note that most of these genes encode proteins that respond to abiotic stresses [e.g. lipid transfer proteins (LTPs) and molecular chaperones], others related to oxidative stress (a peroxidase and a putative glutathione S-transferase), others involved in responses to biotic stimuli (e.g. pathogenesis-related proteins, protease inhibitors, chitinases, and glucanases) and some transcription regulators, among others. When transcripts from roots were compared, few of these (48) were also found to be up-regulated, and the number of those genes related to stress responses was small. However, repressed genes in SlAREB1-overexpressing leaves (47) and roots (22) include genes involved in diverse biological processes, but most are of unknown function (Supporting Information Table S3).

Table 1.  Gene products of selected highly induced transcripts (P ≤ 0.05) in SlAREB1-overexpressing line S3 compared to wild-type (WT) plants
AccessionGene productGene ontologyLog2 fold
X71593PeroxidaseOxidative stress9.9
AI777697Photoassymilate-responsive protein (PAR1)Response to biotic stimulus9.1
AI776170Phytophthora-inhibited protease 1 (PIP1)Response to biotic stimulus8.8
U81996Non-specific lipid transfer proteinResponse to stress8.8
BG631079Glucan endo-1,3-beta-D-glucosidaseResponse to biotic stimulus8.3
BT013355Pathogenesis-related protein P2Response to fungus7.7
Y10149Subtilisin-like protease (subtilase)Proteolysis7.2
AI781668Proteinase inhibitor IStress response7.2
M69247Pathogenesis-related protein P4Response to biotic stimulus7.1
Z15141ChitinaseResponse to biotic stimulus6.7
CN384809Aminocyclopropane-carboxylate oxidaseEthylene biosynthesis6.4
BG627802Predicted protodermal factor protein (PDF)Regulation of transcription6.3
BG629234Proteinase inhibitor 1Stress response6.0
AY093595PR-5xResponse to biotic stimulus6.0
AW038686TSW12 protein – non-specific lipid proteinResponse to abiotic stress5.9
AI898214Putative calreticulin proteinMolecular chaperone5.8
BI923438Putative transcription factor (zinc finger)Regulation of transcription5.7
AI773603Photoassimilate-responsive protein (PAR1)Response to biotic stimulus5.6
BT014164Phosphoglycerate mutase family proteinMetabolic process5.6
BG630825Putative calreticulin proteinMolecular chaperone5.4
AF096246Ethylene-responsive transcriptional coactivatorRegulation of transcription5.3
BE354113Putative calreticulin proteinMolecular chaperone5.3
BG628643Threonine deaminaseAmino acid biosynthesis5.3
BG629612Acidic endochitinase precursorDefence against pathogens5.2
AI489249Metallocarboxypeptidase inhibitorDefence response5.2
Y08804PR proteinResponse to biotic stimulus5.0
M80604Beta-1,3-glucanaseCarbohydrate metabolism5.0
AI773917Arogenate dehydrogenaseAmino acid biosynthesis4.9
BG630869Squamosa promoter binding proteinPutative transcriptional regulator4.9
AJ010942Hexose transporter proteinTransmembrane transport4.8
BG127217Gamma-thionin protein or defensina proteinResponse to biotic stimulus4.7
AY530758AREB-like proteinRegulation of transcription4.6
BT012820Elicitor-inducible cytochrome P450Controlling growth. biotic stress4.5
AW033344DC1 domain-containing proteinDefence response4.3
AI779909Putative glutathione S-transferaseOxidative stress4.3
CN385070hsr203J proteinHydrolase. response to biotic stimulus4.2
BI206504WRKY transcription factorRegulation of transcription4.2
U50151Leucine aminopeptidaseProteolysis. defence response4.1
BG628403Non-specific lipid transfer proteinLipid transport. response to stress3.8
AY034148Alternative oxidase 1aRespiration. oxidation reduction3.5
BT012812EF 1-alpha (AA 1-448)Translation3.5
X94944Lipid desaturase-like proteinLipid metabolism3.3
AF118567PolygalacturonaseCarbohydrate metabolism. cell wall2.9
U30465ChitinaseResponse to biotic stimulus2.9
AF506005Acid invertaseCarbohydrate metabolism2.7
AF050496Ca2+-ATPaseIon transport2.4
U89255Pti4Regulation of transcription2.3
M69248PR proteinResponse to biotic stimulus2.2
AF146691Eli3 proteinMetabolic process2.1
BT013586HSP70Response to stress2.1
L21194Proteinase inhibitorResponse to stress2.0
Y15846TSI-1 proteinResponse to biotic stimulus2.0
BG629070Chlorophyll a/b-binding protein precursorPhotosynthesis. light harvesting1.9
X51904TAS14 peptideResponse to stress1.5

Since commercial tomato microchips do not cover all tomato genes (known or unknown), a cDNA-AFLP approach was used as an additional method to detect new SlAREB1-regulated genes. The differentially expressed fragments were investigated by selective amplification using 36 primer combinations. More than 900 bands were generated, but only bands longer than 100 bp in length that were more abundant in the SlAREB1-overexpressing line were isolated. Sixty-seven significantly up-regulated TDFs ranging from 100 to 600 bp were cloned and sequenced and were compared with databases using BLAST (Supporting Information Table S4). About half of them showed homology with genes associated with stress responses (Table 2) and only five [U81996 and X56040, two LTPs; Y08804, a pathogenesis related (PR) protein; AF506005, an acid invertase; and X51904, a late embryogenesis abundant (LEA) protein] were pulled out in the microarray analysis as well.

Table 2.  Nucleotide homology of transcript-derived fragments (TDFs) obtained from the cDNA-amplified fragment length polymorphism (AFLP) analysis with stress-associated gene sequences from the GenBank database
TDFLength (bp)AccessionHomologySpecies
  1. GTP, guanosine triphosphate; CBL, calcineurin B-like protein.

 1155EU159402Inositol-1,4,5-triphosphate-5-phosphatase (5PT1)A. thaliana
 3160U81996Non-specific lipid transfer protein (Le16)S. lycopersicum
 7185BG631079Glucan endo-1,3-beta-D-glucosidaseS. lycopersicum
10508DQ000202VIP2 transcription regulatorNicotiana benthamiana
11288Y08804PR proteinS. lycopersicum
13298AY789637Ethylene-responsive element binding proteinCapsicum annuum
14233ABA40447GTP-binding-like proteinSolanum tuberosum
19235NP564391Radical-induced cell death1 (RCD1)A. thaliana
20340NM102560Lipid transfer protein relatedA. thaliana
24287NM126777DEAD box RNA helicaseA. thaliana
30190X56040TSW12 proteinS. lycopersicum
34175AF202179Disease resistance protein BS2Capsicum chacoense
37259AK246591Microsomal glutathione s-transferaseS. lycopersicum
42257DQ996037Asc/Asc glutathione S-transferase-likeS. lycopersicum
43235NM122420CBL-interacting protein kinase (CIPK25)A. thaliana
50420AY192370Ethylene response factor 4 (ERF4)S. lycopersicum
53196AY684102Methyl jasmonate esterase (MJE)S. tuberosum
47209AY570725Trehalose-6-phosphate phosphataseN. tabacum
59226AF506005Acid invertaseS. lycopersicum
64158X51904TAS14 peptideS. lycopersicum

Confirmation of the differential expression of few up-regulated genes was performed by qRT-PCR in two sense and two antisense lines. The analyses included genes induced by abiotic stress: a non-specific LTP (LE16, U81996), TSW12 (AW038686), a calreticulin (CRT) (CRT1, AI898214), SlRCD1 (NP187268), an acid invertase (INV/LIN6, AF566005), and the LEA gene TAS14 (X51904). In addition, we included the analysis of another LEA gene, LE25, known to be stress induced in tomato, and the other tomato AREB homolog, SlAREB2. With the exception of SlRCD1 and INV/LIN6, expression of all these genes was significantly up-regulated in at least one overexpressing line (Fig. 8). Similar induction in those transgenic lines was found for four genes known to respond to biotic stimuli, PIN1 (BG629234), PR5 (AY093595), a chitinase (CHI3, Z15141) and TSI-1 (CN384694), supporting the outcome of the microarray and cDNA-AFLP analyses and the thesis that SlAREB1 is involved in regulating stress-induced genes in tomato.

Figure 8.

Expression of candidate target genes for SlAREB1 in transgenic tomato plants. Transcript levels of selected genes that were described in the microarray and cDNA-amplified fragment length polymorphism (AFLP) studies as up-regulated by overexpression of SlAREB1: LE16/LTP1 (U81996), TSW12 (AW038686), CRT1 (AI898214), INV/LIN6 (AF566005), PIN1 (BG629234), PR5 (AY093595), CHI3 (Z15141), TSI-1 (CN384694), SlRCD1 (NP187268), TAS14 (X51904), LE25 (M76552) and SlAREB2 (AK325753). Analyses were performed by qRT-PCR using total RNA isolated from 2-week-old plants grown on Murashige and Skoog (MS) plates. Bars indicate mean relative expression values ± standard error. GADPH was used as a housekeeping gene. Asterisks indicate significant differences between transgenic and WT plants (P < 0.05). WT, wild type; S2 and S3, sense lines; A4 and A6, antisense lines.


In the present study, characterization of two AREB-type transcription factors, SlAREB1 and SlAREB2, is described in tomato. In a previous work, we showed that transcription of SlAREB1 can respond to environmental stimuli in three Solanum species (Yañez et al. 2009). Here, we provide evidence that expression levels of SlAREB1 correlate with the degree of drought and salt tolerance presented by transgenic tomato plants, but most important, we identified a number of genes regulated by SlAREB1 that are associated with both abiotic and biotic stress responses.

SlAREB1 and SlAREB2 are two ABA-, stress-responsive bZIPs in tomato

AREB/ABF bZIP transcription factors have been described to mediate stress-associated gene regulation in aerial tissues and roots in Arabidopsis (Kang et al. 2002; Kim et al. 2004; Fujita et al. 2005) and rice (Lu et al. 2009; Amir Hossain et al. 2010). Other homologs appear more involved in desiccation tolerance and germination arrest in seeds and in young vegetative tissues in Arabidopsis and cereals (Finkelstein & Lynch 2000; Casaretto & Ho 2003; Lu et al. 2009). About nine members identified in both Arabidopsis and rice (Lu et al. 2009) present the four distinctive conserved regions containing specific target residues for phosphorylation (Furihata et al. 2006). In tomato, however, database scrutiny has delivered only three homologous sequences, here named SlAREB1, SlAREB2 and a third that only covers the N-terminal half with the typical conserved regions (not shown). Also, another characteristic of these factors is that they are induced by ABA (Choi et al. 2000; Uno et al. 2000). Transcription of SlAREB1 and SlAREB2 was also shown to be up-regulated by ABA. While in review of the present manuscript, the corresponding SlAREB1 factor from tomato cv. CL5915-93D was described to bind ABREs in the RD29A and SlLAP promoters in vitro and to transactivate two dehydrin genes (Hsieh et al. 2010).

Expression of SlAREB1 and SlAREB2 was induced by salinity and water deficit in leaves and roots, but that of SlAREB1 was more striking. Previously, we showed that SlAREB1 expression was minimal in mature seeds but high in roots, leaves and flowers (Yañez et al. 2009). The expression profile of SlAREB2 in different tomato tissues is similar to that of SLAREB1 (data not shown), but whether SlAREB2 has a more important role in other tissues (e.g. during seed maturation) remains to be investigated. Because SlAREB1, but not SlAREB2, was highly responsive to salt and dehydration treatments (Figs 2 & 3), it is also plausible that, in addition to an ABA, other hormonal signals participate in enhancing SlAREB1 transcription when plant tissues are under stress. Sequences of related protein sequences in dicot species other than Arabidopsis (Fig. 1b) have now been deposited in databases; however, except for rice, the role of these AREB transcription factors has not been investigated in other important crop species. Overexpression of SlAREB1 resulted in a noticeable improved tolerance to salt and water deficit compared to WT plants. Moreover, stress symptoms, such as lower RWC, diminished photosynthetic efficiency and damage by lipoperoxidation, were accelerated in SlAREB1-down-regulating plants (Figs 6 & 7).

SlAREB1 up-regulates several abiotic stress-related genes

In most cases, ABA-induced transcription through AREBs relies on post-translational modifications of these factors that are activated via phosphorylation of specific conserved amino acid residues (Furihata et al. 2006). Both SlAREB sequences described herein present the same regions and conserve the same putative phosphorylation sites (Fig. 1). In Arabidopsis, up-regulation of downstream genes that account for the enhanced tolerance of transgenic plants cannot be achieved by expression of the AREB1 gene alone (Furihata et al. 2006). However, unlike Arabidopsis, overexpression of SlAREB1 in tomato resulted in the activation of numerous genes, with a significant number of them being associated with abiotic stress responses.

The improved tolerance to the stress treatments observed in our study must be related to SlAREB1-activated downstream genes. From the microarray data, a selection of annotated and highly induced genes in leaf tissue (Table 1) includes genes associated with abiotic (including oxidative) stresses. Among them, there are three LTPs, U81996, AW038686 and BG628403. These are members of a multi-gene family associated with the transfer of phospholipids between membranes and are thought to participate in the regulation of the intracellular fatty acid pools, in defence reactions against pathogens and in the adaptation of plants to various environmental conditions (Carvalho & Gomes 2007). LTP1/LE16 (Plant et al. 1991) and TSW12 (Torres-Schumann, Godoy & Pintor-Toro 1992) encode proteins with high similarity to LTPs, and their expression has been detected mainly in stems of tomato plants treated with NaCl, mannitol or ABA. CRT is a key endoplasmic reticulum (ER)-localized Ca2+-binding protein that plays important roles in a variety of cellular processes. Some studies have also highlighted the significance of CRTs in plant growth and development as well as in biotic and abiotic stresses (Jia et al. 2009). INV/LIN6 encodes an extracellular invertase that is up-regulated by methyl jasmonate, ABA, salt stress and wounding (Proels & Roitsch 2009). Higher transcript levels of the TAS14 gene (TDF64) in the sense line was predictable since it encodes for a dehydrin (a type 2 class of LEA protein) whose expression in tomato is known to be induced by ABA, mannitol and NaCl (Parra et al. 1996). Unlike the study of Furihata et al. (2006) in Arabidopsis, besides TAS14, no other LEA genes were found in our microarray analysis.

Because this tomato chip also only covers less than 10,000 transcripts, we included a cDNA-AFLP assay to detect additional target genes. In this analysis, we were able to obtain sequences of 5 genes picked up with the microarrays and 17 other genes. About half of the TDFs can be associated with stress responses. Of those, one interesting gene is represented by TDF19, which showed homology to RCD1, involved in sensitivity to oxidative stress. Interestingly, RCD1 has been described to interact with SOS1, a membrane Na+/H+ antiporter that confers salt tolerance in Arabidopsis, indicating a crosstalk between ion homeostasis and oxidative stress detoxification involved in plant salt tolerance (Katiyar-Agarwal et al. 2006).

Repressed genes in leaf tissues were not many and, since a biological function was assigned to only a few of them, we cannot conclude about any significant process affected by the overexpression of SlAREB1. Likewise, out of the small number of differentially expressed transcripts in roots, few could be annotated (Supporting Information Tables S2 & S3). Most genes selected for validation (Fig. 8) showed expected expression levels in sense and antisense lines; they were significantly up-regulated in at least one SlAREB1-overexpressing line. Probably, the presence of other transcription factor (AREB-like or a different class) would explain why the expression of a gene was, in some cases, similar in WT and in antisense lines. To the group of validated genes, we included two genes absent in the microarray, LE25, a known tomato LEA gene that has been described to respond to environmental stress (Kahn et al. 1993), and SlAREB2. Expression of both genes was significantly affected in one sense line and an expected trend was observed in the remaining lines (Fig. 8). Whether there is self- or cross-regulation among AREB factors in tomato as it was described in Arabidopsis (Finkelstein & Lynch 2000) is yet to be determined.

SlAREB1 up-regulates biotic stress-related genes

The elevated number of SlAREB1-up-regulated genes that came up in the microarray analysis and designated with an ontology term associated with biotic stress was unanticipated. There are over 20 in the list of top selected genes (Table 1), and these are represented by PR proteins, protease inhibitors and degrading enzymes, among others. Expression was confirmed for only a few (Fig. 8). The PR protein encoding by PR5 has been detected in the sap of fungus-infected tomato plants in both compatible and incompatible interactions (Rep et al. 2002) and in salt-stressed soybean plants (Onishi et al. 2006). PIN1 as well as PIN2 are tomato genes encoding serine proteinase inhibitors that are induced by wounding and jasmonic acid (JA) (Schilmiller & Howe 2005). The chitinases-encoding genes CHI3 and CHI2.1 (U30465), along with a β-1,3 endoglucanase (M80604), were initially identified among other chitinases and β-1,3 glucanases induced in fungus-infected tomato leaves (Danhash et al. 1993). TSI-1 encodes a salicylic acid-induced intracellular PR protein from tomato and has also been found to accumulate in potato under salt stress (Aghaei, Ehsanpour & Komatsu 2008).

Other genes responsive to pathogen interactions were also picked up in the cDNA-AFLP experiment. BS2 (TDF34) is a resistance gene that confers resistance to the bacterial pathogen Xanthomonas campestris, which causes bacterial spot disease in tomato (Tai et al. 1999). Others are represented by TDF50 and TDF53, which encode for an ethylene response factor (ERF4) and a methyl jasmonate esterase (MJE). Genes that participate in ethylene signal transduction and synthesis and other regulators of transcription were also recognized in the microarray analysis. The plant hormones ethylene, salicylic acid, JA and ABA are known to act synergistically or antagonistically in the regulation of plant responses to pathogens and abiotic stress factors (Fujita et al. 2006; Asselbergh, De Vleesschauwer & Höfte 2008). Previous studies have pointed out the induction of wound- and pathogen attack-associated genes by salt stress in tomato. For instance, the observation that the tomato def-1 mutant (impaired in jasmonate signalling) displays a severe reduction in the accumulation of PIN genes under salt stress suggests the participation of JA in salt stress-induced gene regulation (Dombrowski 2003). Several reports on the involvement of ABA in the modulation of disease resistance have been reviewed, including the participation of some transcription factors (Asselbergh et al. 2008). Our data suggest that SlAREB1 may participate in biotic responses in tomato by regulating a variety of genes, perhaps serving as a link for ABA signals to responses to other plant hormones.

Recent studies also suggest a crosstalk between plant responses to pathogens and abiotic stresses, including the expression of overlapping sets of genes in response to both stresses (Cheong et al. 2002; Abuqamar et al. 2009; Mauch-Mani & Flors 2009). In this regard, it is worth mentioning that the information obtained in this study is novel for a major ABA signal transcription factor and comprises a list of genes with possible relevant functions in response mechanisms that operate in stress tolerance in a crop species.


This work was supported by grants from Fondo Nacional de Desarrollo Científico y Tecnológico (FONDECYT, grant number 1060843) and International Foundation for Science (IFS grant number C-4075) to J.A.C.