Overexpression of Arachis hypogaea AREB1 Gene Enhances Drought Tolerance by Modulating ROS Scavenging and Maintaining Endogenous ABA Content

AhAREB1 (Arachis hypogaea Abscisic-acid Response Element Binding Protein 1) is a member of the basic domain leucine zipper (bZIP)-type transcription factor in peanut. Previously, we found that expression of AhAREB1 was specifically induced by abscisic acid (ABA), dehydration and drought. To understand the drought defense mechanism regulated by AhAREB1, transgenic Arabidopsis overexpressing AhAREB1 was conducted in wild-type (WT), and a complementation experiment was employed to ABA non-sensitivity mutant abi5 (abscisic acid-insensitive 5). Constitutive expression of AhAREB1 confers water stress tolerance and is highly sensitive to exogenous ABA. Microarray and further real-time PCR analysis revealed that drought stress, reactive oxygen species (ROS) scavenging, ABA synthesis/metabolism-related genes and others were regulated in transgenic Arabidopsis overexpressing AhAREB1. Accordingly, low level of ROS, but higher ABA content was detected in the transgenic Arabidopsis plants’ overexpression of AhAREB1. Taken together, it was concluded that AhAREB1 modulates ROS accumulation and endogenous ABA level to improve drought tolerance in transgenic Arabidopsis.


Introduction
Plant productivity is greatly affected by environmental stresses, including drought, low temperature and high salinity. Particularly, water stress is an important abiotic stress that leads to reduction of crop yields and affects agricultural productivity in many parts of the world. Peanuts, or groundnuts, are important oil and economic crops around the world and necessary for the national economy of China. In China, 70% of the peanut production areas are subject to varying degrees of water deficit that severely impact peanut production [1]. Therefore, studies on improvement of drought resistance of peanut are urgent and necessary. However, not much is known about drought stress responses, due to the complex genome (tetraploid) and the lack of efficient transgenic technology in peanut.
Numerous stress responsive genes that are involved in protecting plants far from stress-induced damages have been identified [2][3][4]. Abscisic acid (ABA), a universal stress response signal molecule, alters numerous transcriptome through dependent and independent pathways to enhance stress tolerance and it is involved in ion homeostasis and enhanced antioxidant defense [5,6]. Under stress conditions, ABA signaling is activated and phosphorylates downstream substrates, AREBs/ABFs, then stimulates ABA-responsive gene expression and ABA-related responses in plant [7]. AREB/ABFs are the member of basic domain leucine zipper (bZIP) transcription factors. In Arabidopsis, they include AREB1/ABF2, AREB2/ABF4, ABF3, ABI3 and ABI5, etc., which have been identified and characterized as ABA-responsive signaling molecular or downstream transcription factors [5,[8][9][10]. Functional characterizations of AREB/ABFs have revealed that they are induced in response to drought, high salinity or even ABA treatment [11][12][13]. In Arabidopsis, AREB1/ABF2, AREB2/ABF4 and ABF3 do not completely overlap in their functions and can form homo-or hetero-dimers with each other [14]. Therefore, each AREB protein may play a specific role in response to water stress. Many downstream genes of AREB/ABFs have been illustrated, including RD29B, AIL1, RAB1 and RD20, etc. [13,14]. In addition, AREB homologous genes, ABP9 (isolated from maize) and PtrABF (isolated from Poncirus tirfoliata), have been reported to modulate cellular levels of ROS (reactive oxygen species) [15,16]. ROS has a dual role in various cellular compartments. It acts as a signaling molecule to process plant protective stress responses at low concentrations, and it causes oxidative damage at higher concentrations [15,17]. It has been well known that ABA regulates ROS producing and ROS scavenging genes to modulate the ROS levels [15]. AREB protein may play a crucial role in these proceedings.
AREB/ABF-related pathways and multiple mechanisms appear to contribute to increasing stress tolerance in Arabidopsis and other species, which is useful to understand the drought stress response in peanut. We have previously isolated a dehydration inducible AREB-like gene, AhAREB1, from peanut and found that AhAREB1 gene was specifically induced by ABA, dehydration and drought [18]. Like others, AREBs/ABFs and AhAREB1 genes encode a transcription activator [19], and the amino acid sequence of AhAREB1 shared homology with other AREBs/ABFs [18]. However, the function and exact mechanisms of AhAREB1 in plant are still not clear.
Here, we report that the constitutively expressed AhAREB1 gene is associated with high tolerance to dehydration and drought stresses in Arabidopsis, also recovering the ABA-sensitivity in abi5 (ABA insensitive mutant). The potential AhAREB1 target genes were identified by comparing large-scale gene expression profiles of transgenic Arabidopsis under normal growth conditions under drought stress. According to the microarray results, expression patterns of candidate genes were further confirmed by quantitative PCR. Moreover, accumulation of ROS and ABA content was analyzed in AhAREB1-overexpressing plants.

The AhAREB1 Overexpression Greatly Improves Drought Tolerance and ABA Sensitivity in Transgenic Arabidopsis Plants
To examine the role of AhAREB1 in plants, we performed Arabidopsis transformation using plasmid p35S::AhAREB1 in which the gene was under the control of the cauliflower mosaic virus (CaMV), 35S promoter. Twenty four independent transgenic lines were identified in which the T2 plants segregated 3:1 for hygromycin resistance, and the hygromycin resistant T3 homozygous plants were then screened and confirmed by one-step RT-PCR analysis (data not shown). Three independent homozygous lines (A22, A38 and A39) were chosen for investigation. Among these transgenic lines, A22 displayed low-level expression of AhAREB1, while A38 and A39 showed high-level expression of AhAREB1 ( Figure 1G). To evaluate the ABA sensitivity, the non-sensitivity mutant, abi5 (abscisic acid insensitive 5-1, abi5-1), was used as control, and the transgenic lines T-abi5 (transformed AhAREB1 into abi5 line, named T-abi5) were used to determine recovery of ABA sensitivity. The growth of transgenic plants, expression of AhAREB1, germinating seeds, green cotyledon and axial root length were detected, respectively. There was no significant difference in growth phenotype between transgenic plants and wild-type (WT) at the seedlings stage. However, we observed a slight growth inhibition in A38 and A39 plants on the soil-grown after six weeks (Supplementary File 1, Figure S1 and Table S1). During the germination process, no significant difference was observed between A22, T-abi5, abi5 and WT under normal growth conditions. However, A38 and A39 plants showed a low germination rate. Germination of A22 and A38 plants was more inhibited by ABA than WT ( Figure 1C). In addition, after transforming the AhAREB1 into abi5, all of them improved their ABA sensitivity under 1.5 μmol/L of ABA, which indicates that AhAREB1 recovered the ABA sensitivity in abi5 ( Figure 1C). Similar results were observed in a green cotyledon and primary root growth experiment. Under normal growth conditions, the green cotyledon of all transgenic plants were 100%, but decreased dramatically after 1.5 μmol/L ABA treatment in A38, A39 and T-abi5 plants, while abi5 plants showed non-sensitivity to ABA ( Figure  1D). An ABA inhibition of primary root growth was observed for WT, A22, A38, abi5 and T-abi5 plants. A38 and T-abi5 plants were significantly inhibited by 5.0 μmol/L ABA, while A22 and WT were less inhibited. On the other hand, abi5 lines showed higher resistance to ABA than other lines ( Figure 1A,B). The above findings demonstrated that overexpressed AhAREB1 improves ABA sensitivity. Furthermore, the survival rate of A22, A38 and A39 plants was markedly higher in comparison with WT plants. After rehydration, the A22, A38 and A39 plants achieved 68%, 88% and 84% survival rates, respectively, while WT had only 40% ( Figure 1E,F). The results also indicated that overexpressed AhAREB1 elevated drought tolerance in the plants.

Identification of Target Genes of AhAREB1 Using Genome-Wide Microarray Expression Analysis
To identify the target genes of AhAREB1, we compared the expression profiles of 14 day-old A38 plants with those of WT plants under both normal and stress conditions (dehydration 0.5 h) using the Arabidopsis gene expression 385k arrays (Roche NimbleGen). Genes 493 and 317 show differential expression between A38 and WT plants under control growth conditions and dehydration stress (A38-CK vs. WT-CK or A38-dry vs. WT-dry), respectively. Genes 749 and 986 show difference in expression in A38 plants and WT plants under the control condition compared with dehydration stress (A38-CK vs. A38-dry or WT-CK vs. WT-dry), respectively (Supplementary File 2).
In addition, we confirmed the expression of these genes using quantitative PCR (q-PCR). Under dehydration, the expression pattern of RD26, RD20, CAT2, NCED3, CCS and CSD3 is similar, which was strongly inducted at 0.5 h or 1.5 h both in A38 plants and WT plants. The expression of RD29A was more strongly induced in WT plants than A38 plants after dehydration for 0.5 h, which showed a dramatic decrease after dehydration for 1.5 h. On the other hand, the expression levels of RD29B, ABA1 and AtABCG40 were not affected after dehydration for 0.5 h compared with the normal condition in A38 plants, although RD29B and ABA1 were more highly expressed in A38 than WT plants. Meanwhile, under the normal condition, NCED3 and AtABCG40 transcripts were slightly inhibited in A38 plants when compared to WT plants. These findings are almost consistent with the microarray results. In addition, RD29B was markedly induced in WT plants after dehydration, irrespective of duration, 0.5 h or 1.5 h. However, no significant difference in expression of RD29B was found in A38 plants after dehydration, although its expression was higher in A38 plants than WT plants under normal conditions ( Figure 2). This result is not completely consistent with microarray results. A38 and WT seedlings were grown on MS agar medium with or without dehydration treatment at the time points indicated. 18SrRNA gene expression level was used as an internal control. Error bars represent standard deviation among the three reduplicate experiments. WT represents wild-type plants under normal conditions; WTD represented wild-type plants that were treated with dehydration for 0.5 h or 1.5 h; A38 represented A38 lines under normal conditions; A38D represented A38 lines that were treated with dehydration for 0.5 h or 1.5 h. * indicated that values of the WTD, A38 and A38D were significantly different from those of WT with p < 0.05, after dehydration for 0.5 h or 1.5 h, respectively. # indicated that values of A38D were significantly different from A38 at p < 0.05 after dehydration for 0.5 h or 1.5 h, respectively.

Effect of Constitutive Expression of AhAREB1 on ROS Levels and ABA Content in Arabidopsis thaliana
To acquire further insights into AhAREB1 regulation mechanism in plants under water stress, ROS levels and ABA levels were examined, associated with the related target genes of AhAREB1, from microarray expression analysis. Two prominent ROS species, O 2 − and H 2 O 2 , were examined by nitro blue tetrazolium (NBT) and 3,3-diaminobenzidin (DAB) staining under normal growth and dehydration stress conditions, respectively. As shown in Figure 3A,B, the steady state levels of O 2 − (visualized as dark blue products) and H 2 O 2 (visualized as deep brown products) were reduced in the leaves of A22, A38 or T-abi5 plants compared to WT or abi5 plants under normal conditions and dehydration. We also measured the activity of ROS-scavenging enzymes, catalases (CATs) and superoxide dismutases (SODs), in all plants under control conditions and dehydration. The CATs activities in A38 plants were higher than that WT plants, not only under control, but also under dehydration conditions, while the CATs activities in abi5 plants were lower than that WT plants ( Figure 3C). Meanwhile, after dehydration, water stress caused significant increase of SOD activity in A22, A38 and T-abi5 plants ( Figure 3C). These results demonstrate that AhAREB1 plays a pivotal role in plant tolerance to drought stress by the controlling of ROS accumulation. Based on microarray gene expression analysis, we found that many genes involved in ABA synthesis or metabolism were significantly differently expressed in A38 plants. Therefore, we hypothesize that there would more ABA content and drought stress resistance in the A38 plants. To examine this, ABA levels were measured in transgenic plants. The results show that A38 and T-abi5 have higher ABA content than WT or abi5 plants under both conditions, normal or dehydrated ( Figure 3D). These data suggest that the AhAREB1 gene may improve endogenous ABA synthesis in transgenic plant, and the roles of target genes are consistent with stress resistant phenotypes of the transgenic plants.

Discussion
In our previous work, we have shown that AhAREB1 gene, which shares high sequence homology with GmAREB1, SlAREB and ABF2/AREB1, responds to environmental stimuli [18]. Further study showed that AhAREB1 functions as a transcriptional activator in yeast [19]. Here, we showed that (1) AhAREB1-overexpressed plants had a higher survival rate under drought stress and were more sensitive to ABA than WT plants; (2) the ABA sensitivity in the abi5 mutant could be recovered by AhAREB1; (3) ABA-and drought-responsive genes (such as RD20, RD26, RD29A, RD29B), ABA homeostasis genes (ABA1, AtABCG40, NCED3 and CYP707A2) and ROS scavenging genes (CAT2, CCS, SOD1 and CSD3) were differentially expressed between AhAREB1-overexpressed plants and WT plants; (4) constitutive expression of AhAREB1 reduces ROS levels by increasing the CATs and SODs activity, but improving the ABA content in Arabidopsis.
Previous findings showed that AREB/ABF transcription factors are positive regulators of ABA signaling in response to drought stress [8,12]. Also, overexpression of the AREB/ABF genes or stress-related genes in plants can result in slight growth inhibition [10,[20][21][22]. Our results shown in Supplementary File 1, Figure S1A and Table S1 are consistent with these previous reports. The phytohormone, ABA, not only regulates abiotic stress response, but also imparts growth retardation [5], because it can reprogram the transcriptional events widely in plants via AREB/ABF transcription factors [8,12]. As expected, many ABA-and drought-responsive genes were upregulated in the AhAREB1-overexpressed plants. They included RD20, RD26, RD29A and RD29B genes, and so on, which are known to function in ABA-mediated stress-signal transduction [23,24]. Besides, higher survival rate and ABA sensitivity were observed in the AhAREB1-overexpressed plants, suggesting that plants have higher tolerance or resistance to stress [25]. Further, transformation of AhAREB1 into abi5 recovered the ABA sensitivity of the mutant plants. The protein, ABI5, also is a bZIP transcription factor, which acts as a central regulator of ABA signaling in Arabidopsis thaliana [26,27]. Previous research has shown that AREB/ABF and ABI5 are homologous protein families [28]. Our results suggest that AhAREB1 functions in ABA signal transduction and may overlap with ABI5 roles redundantly. In this study, we did not detect any difference in expression of ABI5 in A38 plants, using microarray under control conditions or dehydration stress (Supplementary File 2), while the expression level of ABI5 increased in WT under dehydration stress.
It is well known that ROSs act as intracellular messengers [29][30][31][32]. Conversely, ROS, when accumulated in excess, causes damage to plants [15,31]. Therefore, appropriate regulation of ROS levels is important to improve tolerance to abiotic stress. Previous studies indicated that ABA via ABRE transcription factors regulates ROS-producing and ROS-scavenging genes, such as SODs, APXs and CATs genes, to modulate the cellular ROS levels [32][33][34][35]. It has been known that constitutive expression of maize ABRE2 orthologous gene, ABP9, causes reduced cellular levels of ROS and enhances drought stress tolerance [15]. In this study, constitutive expression of AhAREB1 in Arabidopsis enhanced the ROS scavenging genes (CAT2, CCS, SOD1 and CSD3) expression, also reducing cellular ROS levels and maintaining higher CAT and SOD activity than the WT plants at normal and dehydration conditions. These results clearly indicate that the AhAREB1 gene is involved in modulation of ROS accumulation in Arabidopsis and enhances tolerance to dehydration, which is consistent with the previous reports [15].
Several genes encoding enzymes for ABA biosynthesis or catabolic activities were altered in AhAREB1-overexpressed plants. NCED3 is a key enzyme of ABA biosynthesis under drought stress [36] in Arabidopsis, whose transcription level shows rapid response to dehydration or drought stress and gradually decreases following the time of stress [37,38]. ABA1 is an enzyme important in de novo ABA biosynthesis, which exhibited drought stress resistance, while overexpressed in Arabidopsis plants [39]. In A38 transgenic plants, NCED3 was downregulated in expression under control conditions, but upregulated under dehydration, both in WT and A38 plants. However, ABA1 was upregulated under control conditions and had no significant difference in expression under dehydration, either in WT or A38 plants. It also indicated that AhAREB1 might participate in the negative regulation of NCED3 gene transcription. Several ABRE-like elements are found in the NCED3 promoter, given that AhAREB1 protein may be capable of activating NCED3 via specific interactions with ABREs. Meanwhile, CYP707A2 belongs to ABA 8'-hydroxylase and has a significant role in ABA catabolism during imbibitions and regulates seeds dormancy [40][41][42]. In Arabidopsis, four members (CYP707A1-A4) were identified as the enzymes responsible for catalyzing ABA 8'-hydroxylation [40][41][42][43]. From their tissue specificity or expression patterns, CYP707As may have distinctive roles in plants. For example, CYP707A3 was involved in dehydration and rehydration response [44]. Our microarray data also showed that CYP707A3 was greatly induced by dehydration both in WT (>3-fold) and A38 (>2-fold) plants (Supplementary File 2). Here, CYP707A2 seemingly responded only to ABA content or AhAREB1, for its transcripts always were induced in A38 plants whenever under control conditions or drought stress conditions. ABA content is associated with drought stress resistance in plants during the early stage [5,32,45,46]. These genes are thought to play a major role in maintaining a balance of endogenous ABA content. Therefore, detection of ABA content is very important and necessary in A38 plants. It is noteworthy that ABA content was higher in A38 transgenic plants than WT, both under control conditions and dehydration. Combined with the gene expression data, it suggests that endogenous ABA was mainly synthesized by ABA1 and NCED3, under control condition and drought stress, respectively, in our transgenic plants. Although CYP707A2 expression was upregulated in A38 plants, both under control and stress conditions, it may be a feedback loop between its transcription and ABA content/ABA signaling by AhAREB1 modulation. In fact, CYP707A2 plays a central role during seed imbibition and regulates dormancy [42,43], and its transcription was not induced by dehydration or rehydration [44]. It simply means that the high expression of CYP707A2 was not associated with stress resistance or only has a small role in ABA catabolism during the seedling stage. Under dehydration, the endogenous ABA content was further regulated by a balance between biosynthetic and catabolic activities through NCED3 and CYP707A3. The AtABCG40 gene encodes a plasma membrane ABA uptake transporter and is necessary for timely responses to ABA [47]. In A38 plants, the transcription of AtABCG40 was downregulated at the whole plant level. This suggested that ABA transporter rate may be lower in A38 plants than in WT plants. As a result, A38 plants keep a high ABA content and show stress resistance.

Plant Materials
Seeds of peanut (Arachis hypogaea L. cv. Shanyou 523) were sown in growth medium (vermiculite, perlite and soil; 1:1:2) and grown in plastic pots in a growth chamber with a photoperiod of 16 h light at 26 °C and 8 h of darkness at 22 °C, as described previously [48]. Seeds of Arabidopsis wild-type (WT) were surface sterilized in 70% ethanol for 2 min and in 1% sodium hypochlorite for 10 min. After five washes with sterile water, the seeds were sown on MS (Murashige and Skoog) medium supplemented with 2% sucrose and 0.8% agar. After 2 days of vernalization at 4 °C, the seeds were germinated and grown in a growth chamber under a daily cycle of 16 h light and 8 h dark at 20 ± 2 °C. Seven days after sowing, the seedlings were planted in plastic pots in a medium of vermiculite, peat moss and perlite (1:2:1).

Plasmid Construction and Arabidopsis Transformation
To generate the overexpression construct of the AhAREB1 coding region, full-length cDNA was generated by RT-PCR with the following primers: 5'-CTG AGATCT ATG AAC TTC AGG GGC TAT GGT GAT-3' and 5'-CTGGGTGACC CTA CCA GGG ACC TGT AAC TGT CCTT-3' (the underlined parts are BglII and BstEII sites, respectively). The PCR product, confirmed by DNA sequencing, was cloned into modified binary vector pCAMBIA1301 between the BglII and BstEII sites, under the control of the constitutive cauliflower mosaic virus 35S promoter to generate the 35S::AhAREB1 frame. The overexpression construct was introduced into Agrobacterium tumefaciens strain GV3101 and then transformed into Arabidopsis WT plants by the floral dip method. Transgenic T1 lines were selected on MS medium containing 40 mg/L kanamycin (Sigma-Aldrich, St. Louis, MO, USA). T2 seeds from the selected transgenic plants were germinated on a medium with 40 mg/L hygromycin, and then the homozygous lines were selected. The homozygous T3 progeny were then examined for the expression level of the target genes by One Step SYBR PremeScript RT-PCR kit (Takara Biotechnology, Dalian, China) using two specific primers, smRT-AhAREB1-F and -R (Supplementary File 1, Table S3).

Assays of Seed Germination and Green Cotyledons; Growth of Roots
Seed germination, green cotyledons and root growth were treated with exogenous ABA and assayed, as described previously [48]. Sterilized seeds of transgenic Arabidopsis and WT were sown on 1/2 MS agar plates, which were supplemented with 1.5 μmol/L ABA to observe germination. To score green cotyledons, the agar plates were supplemented with 0.5 and 1.5 μmol/L ABA, and germinated seeds with fully expanded green cotyledons were recorded on day 4 after sowing. Three independent experiments were performed (30 seeds in each line per test). Assays of root growth were carried out by transferring 4 days-old seedlings of transgenic Arabidopsis and the WT, under normal sprouting, onto 1/2 MS medium agar plates with 5 μmol/L ABA treatments. Primary root lengths of more than 20 seedlings in each line were measured on day 8 after the start of treatment. The experiment was triplicated.

Drought Stress Tolerance Assays
Survivability tests under the drought conditions were conducted on three independent transgenic lines and the wild-type. Seeds were surface-sterilized and cultivated on 1/2 MS medium for 10 days after radical emergence and then were transferred to soil. At the rosette stage (4 weeks after sowing), half of the samples were subjected to drought stress. These samples were subjected to drought stress by withholding irrigation for 10 days to evaluate drought tolerance visually (the leaves withered). These samples were then rewatered (once a day) for 4 days. The remaining half of the samples was grown under a standard irrigation regime (watering once a day) for 14 days, as a control. The survival rates of the plants were calculated as the ratio of plants with green leaves to the total number of plants used. All experiments were repeated at least 3 times, and more than 32 plants of each line were used in each replicate.

Microarray Analysis
Two-week-old A38 lines and WT grown on the plate were harvested directly or after dehydration treatment for 0.5 h by opening the lids of the plates directly, and they were subjected to microarray experiments using the Arabidopsis Gene expression 385k Arrays (Roche NimbleGen, Madison, WI, USA). Total RNA isolated using TRIzol reagent (Takara, Dalian, China) was used for the preparation of Cy5-labeled and Cy3-labeled cDNA probes. All microarray experiments, including data analysis, were carried out as described previously [49]. The reproducibility of microarray analysis was assessed by dye swap in each experiment.  Table S3). The relative expression levels were calculated using the relative 2 −ΔΔCT method and determined as described previously [48].

In Situ NBT Staining and Measurement of SOD Activity
In situ accumulation of superoxide (O 2 − ) was examined based on histochemical staining by nitro blue tetrazolium (NBT). For O 2 − detection, the treated samples were immersed in 1 mg/mL fresh NBT solution (prepared in 10 mmol/L phosphate buffer, pH 7.8) and incubated in light at 25 °C until dark spots were observed. The stained samples were then bleached in concentrated ethanol and kept in 70% ethanol. For extraction of superoxide dismutase (SOD, EC 1.15.1.1), about 0.5 g of leaf sample was ground in liquid nitrogen with a pre-cooled pestle and mortar and homogenized in 5 mL of extraction buffer containing 50 mmol/L phosphate buffer (pH 7.8) and 1% polyvinylpyrrolidone (PVP). The homogenate was centrifuged at 10,000× g for 20 min at 4 °C, and the resulting supernatant was collected for enzyme activity analysis. Activities of SOD were spectrophotometrically measured using an SOD Detection Kit (A001, Jiancheng, Nanjing, China), according to the manufacturer's instruction. Protein content was estimated using BCA Protein Assay Kit (P0012BCA, Beyotime, Shanghai, China), according to the manufacturer's instruction. One unit of SOD activity is defined as the amount of protein. The results were expressed as enzyme activity per milligram protein (U/mg).

In Situ DAB Staining and Measurement of CAT Activity
In situ accumulation of hydrogen peroxide (H 2 O 2 ) was examined based on histochemical staining by 3,3-diaminobenzidin (DAB). For H 2 O 2 detection, the treated samples were immersed in 1 mg/mL fresh DAB solution (pH 5.8) (prepared in 10 mmol/L phosphate buffer, pH 3.8) and incubated in dark at 28 °C for 8 h. The stained samples were then bleached in 80% ethanol and boiled for several minutes (2-5 min), then dropped for staining by concentrated ethanol and kept in 4 °C, as described by Cheng and Song [50].
For extraction of catalase (H 2 O 2 oxidoreductase; EC 1.11.1.6), about 0.5 g of leaf sample was ground in liquid nitrogen with 2-3 mL pre-cooled phosphate buffer (pH 7.0), then poured into a 25 mL volumetric flask with constant volume to scale. The mix slurry was kept at 5 °C for 10 min. The homogenate was centrifuged at 10,000× g for 15 min at 4 °C, and the resulting supernatant was collected for enzyme activity analysis. Activities of CAT were measured spectrophotometrically by UV absorption at 240 nm. Protein content was estimated using a BCA Protein Assay Kit (P0012BCA, Beyotime, Shanghai, China), according to the manufacturer's instruction. One unit of CAT activity is defined as the amount of protein. The results were expressed as enzyme activity per milligram protein (U/mg).

Quantification of ABA Levels
To determine the ABA levels in WT and transgenic plants, 28 day-old seedlings were dehydrated for 10 days. ABA extraction and detection were performed, as previously described [51]. In brief, 300-500 mg of plant material was frozen in liquid nitrogen and ground using a Mixer-Mill. Plant material was treated with 1-2 mL of ABA extraction buffer (10 mmol/L HCl and 1% (w/v) polyvinyl polypyrrolidone in methanol). The mixture was kept at 4 °C overnight. After centrifugation, the supernatant was neutralized with 1 mol/L NaOH and ABA levels were quantified using the liquid chromatogram. Raw values for ABA levels were standardized by plant mass and extraction volume.

Subcellular Localization
For the localization of AhAREB1, the coding frame of AhAREB1 prepared by PCR was cloned into the frame in front of the GFP coding region of modified pBI121 (Clontech, Mountain View, CA, USA) using the Xba1 and Sma1 sites. The construct was used to transform Arabidopsis (Col-0), and T1 plants were used to determine GFP localization. Roots of 10 day-old seedlings were examined for the green (GFP expression) fluorescence, employing a confocal microscope (LSM 5 Pascal, Zeiss, Jenaer, Germany).

Conclusions
In summary, we demonstrated that AhAREB1 acts as a transcriptional activator of stress-relative, ROS-modulated genes and ABA-induced genes under drought or dehydration stress, and it may play an important role in drought stress tolerance via ABA homeostasis and control of ROS accumulation.
(20094407110004), Guangzhou Science and Technology Project Funding (2012J4100115). We thank the Arabidopsis Information Resource Center for providing abi5 mutants, the Genevestigator Database for providing a seven day trial and Tin Yang (School of Chemistry and Environment) for technical advice on HPLC analysis.