In plants, osmotic stress-responsive transcriptional regulation depends mainly on two major classes of cis-acting elements found in the promoter regions of stress-inducible genes: ABA-responsive elements (ABREs) and dehydration-responsive elements (DREs). ABRE has been shown to perceive ABA-mediated osmotic stress signals, whereas DRE is known to be involved in an ABA-independent pathway. Previously, we reported that the transcription factor DRE-BINDING PROTEIN 2A (DREB2A) regulates DRE-mediated transcription of target genes under osmotic stress conditions in Arabidopsis (Arabidopsis thaliana). However, the transcriptional regulation of DREB2A itself remains largely uncharacterized. To elucidate the transcriptional mechanism associated with the DREB2A gene under osmotic stress conditions, we generated a series of truncated and base-substituted variants of the DREB2A promoter and evaluated their transcriptional activities individually. We found that both ABRE and coupling element 3 (CE3)-like sequences located approximately −100 bp from the transcriptional initiation site are necessary for the dehydration-responsive expression of DREB2A. Coupling our transient expression analyses with yeast one-hybrid and chromatin immunoprecipitation (ChIP) assays indicated that the ABRE-BINDING PROTEIN 1 (AREB1), AREB2 and ABRE-BINDING FACTOR 3 (ABF3) bZIP transcription factors can bind to and activate the DREB2A promoter in an ABRE-dependent manner. Exogenous ABA application induced only a modest accumulation of the DREB2A transcript when compared with the osmotic stress treatment. However, the osmotic stress-induced DREB2A expression was found to be markedly impaired in several ABA-deficient and ABA-insensitive mutants. These results suggest that in addition to an ABA-independent pathway, the ABA-dependent pathway plays a positive role in the osmotic stress-responsive expression of DREB2A.
In contrast to animals, plants cannot avoid unfavorable circumstances by adjusting their location. Instead, they have developed sophisticated sensing and response mechanisms related to environmental stresses such as drought, high salinity and extreme temperatures. Numerous researchers have reported a number of genes that provide the ability to tolerate and overcome the adverse effects of these stresses and have revealed complicated transcriptional regulatory networks associated with these genes that involve specific cis-acting elements in their promoters and interacting transcription factors (Zhu 2002, Bartels and Sunkar 2005, Nakashima et al. 2009b, Medina et al. 2011).
Many osmotic stress-responsive genes in plants are also responsive to the phytohormone ABA, indicating that ABA functions as a signal mediator for osmotic stress in plants (Cutler et al. 2010, Raghavendra et al. 2010, Fujita et al. 2011). The ABA-responsive element (ABRE; PyACGTG/TC) is a well-studied cis-element involved in ABA-induced gene expression (Hattori et al. 2002, Fujita et al. 2011). Multiple ABREs or combinations of ABREs with coupling elements are required for ABA-responsive transcription to occur (Shen et al. 1996, Hobo et al. 1999, Narusaka et al. 2003, Zhang et al. 2005, Gómez-Porras et al. 2007). The ABRE-BINDING PROTEIN/FACTOR (AREB/ABF) family belongs to group A of the basic leucine zipper (bZIP) transcription factor family and is composed of nine members in Arabidopsis (Arabidopsis thaliana) (Choi et al. 2000, Uno et al. 2000, Corrêa et al. 2008). Gain- and loss-of-function studies on the AREB/ABF genes have revealed positive and negative effects of these genes, respectively, on the osmotic stress tolerance of plants (Kang et al. 2002, Kim et al. 2004, Yoshida et al. 2010).
Dehydration-responsive element/C-repeat (DRE; A/GCCGAC) is another example of the cis-elements that govern osmotic stress-responsive transcription (Agarwal et al. 2006, Nakashima et al. 2009b). DRE was first identified as an essential cis-element in the RESPONSIVE TO DESSICATION 29A (RD29A) promoter for ABA-independent expression in response to dehydration and cold stresses (Yamaguchi-Shinozaki and Shinozaki 1994). The ABA independence of DRE-mediated transcription has been supported by several reports. A truncated RD29A promoter containing DREs without an ABRE was found to exhibit a loss of induction activity in response to exogenous ABA application but to retain induction activity under osmotic stress treatments (Yamaguchi-Shinozaki and Shinozaki 1994, Narusaka et al. 2003). Moreover, the osmotic stress-responsive transcription of RD29A was observed to be maintained in ABA-deficient mutant plants (Yamaguchi-Shinozaki and Shinozaki 1993).
DRE-BINDING PROTEIN 2A (DREB2A) was identified in a yeast one-hybrid screen using the RD29A promoter as bait (Liu et al. 1998). DREB2A harbors an ethylene-responsive element-binding factor/APETALA2-type (ERF/AP2-type) DNA-binding domain and functions as a transcriptional activator by interacting with a DRE cis-element in the promoter region of downstream genes (Liu et al. 1998, Nakashima et al. 2000, Sakuma et al. 2002). Overexpression of full-length DREB2A does not result in the activation of downstream gene expression. However, overexpression of a constitutively active form of DREB2A that lacks the negative regulatory domain confers enhanced tolerance to dehydration, high salinity and high temperature in plants (Sakuma et al. 2006a, Sakuma et al. 2006b). Conversely, T-DNA insertional mutants of DREB2A exhibit a loss of tolerance to high temperature (Sakuma et al. 2006b). Subsequent genome-wide transcriptomic analyses revealed that DREB2A contributes to the acquisition of environmental stress tolerance in plants. Microarray data from a DREB2A gain-of-function mutation suggest that DREB2A regulates a number of genes that are involved in osmotic and/or heat stress tolerance and that a number of these genes contain more than one DRE sequence in their promoter region (Sakuma et al. 2006a, Sakuma et al. 2006b, Maruyama et al. 2009). Recent studies suggest that DREB2A might collaborate with other transcription factors, including members of the AREB/ABF family, in downstream gene regulation through direct interaction (Lee et al. 2010, Elfving et al. 2011).
DREB2A expression is also stress inducible. When Arabidopsis plants are exposed to abiotic stress conditions, such as heat and high osmolarity, the DREB2A transcript level increases rapidly and reaches levels up to 250-fold higher than the basal level. In contrast, transcriptional induction of DREB2A in response to exogenous ABA is weak, which supports our categorization of DREB2A transcription under osmotic stress conditions as ABA-independent. However, a weak but significant induction of DREB2A expression in response to ABA application can be detected (Liu et al. 1998, Nakashima et al. 2000, Sakuma et al. 2002).
From many studies performed in recent years, we are currently beginning to understand the roles of DREB2A and its orthologs in plants. However, we still have little information regarding how plants regulate the expression of DREB2A in response to unfavorable circumstances. In this study, we report that the DREB2A promoter contains ABRE and coupling element 3 (CE3)-like sequences that are necessary for promoter activity under osmotic stress conditions, such as dehydration and high salinity. Based on establishing the significant but partial effect of ABA signaling cascades on DREB2A expression, we suggest a model in which DREB2A expression requires both ABA-dependent and ABA-independent signaling cascades.
Truncation analysis of the DREB2A promoter
To define the DREB2A promoter region involved in the response to osmotic stress, a number of truncated promoter fragments were isolated from Arabidopsis genomic DNA and fused to the GUS (β-glucuronidase) reporter gene in the plant expression vector pGK-GUS (Qin et al. 2008). Transgenic Arabidopsis plants harboring each promoter–GUS construct were generated and were exposed to controlled dehydration conditions for 5 h. The transcript levels of endogenous DREB2A and GUS were measured by quantitative real-time PCR (qRT-PCR) (Fig. 1). The transcriptional response of GUS was abolished between the N8 and N9 constructs, or between the C5 and C6 constructs (Fig. 1). These data indicated that the region (designated Region I) between −147 and −54 bp from the transcription initiation site was responsible for the dehydration-responsive expression of DREB2A (Fig. 1).
ABRE is necessary for DREB2A expression under osmotic stress
For further characterization of the cis-elements that influence DREB2A promoter activity, we screened the Region I sequence using two web-based plant cis-element databases: PlantCARE (Lescot et al. 2002) and PLACE/Signal Scan (Higo et al. 1999). Based on these screenings, ABRE and CE3 (GACGCGTGTC)-like sequences were identified as putative cis-elements in Region I (Fig. 2A). Although no clear-cut prediction was available, two additional GC-rich boxes were identified and designated Box1 and Box2 (Fig. 2A, B). In addition, we obtained genomic sequences upstream of the DREB2A orthologs in other plant species from the Phytozome database (http://www.phytozome.net/) and compared them with the DREB2A promoter sequence. We found that the CE3–ABRE cassette is conserved in the promoter region of DREB2A orthologs in many other eudicot plants, although CE3 was found to be replaced by ABRE in plants outside the Arabidopsis genus (Supplementary Fig. S1).
Because the transcriptional activity of a single copy of the Region I sequence was not strong enough to define, a new promoter–GUS construct was prepared containing four unidirectional tandem repeats of Region I (I × 4). In transgenic Arabidopsis plants, I × 4 showed dehydration stress-induced accumulation of GUS mRNA as N0 did (Fig. 2C). Next, we introduced base substitutions into each of the putative ABRE sequences in the I × 4 construct and designated this new construct Am × 4 (Fig. 2B, C). This ABRE-mutated construct resulted in the loss of induction of transcriptional activity in response to dehydration (Fig. 2C). These results raised the possibility that the ABRE sequence might be one of the core elements in the DREB2A promoter involved in the response to dehydration stress. To test this hypothesis, we examined the promoter activity of the N0 derivatives, including base substitutions in ABRE, Box1 or Box2 (N0-Am, N0-1m or N0-2m, respectively) and truncations of the ABRE or CE3-like (N0ΔAB or N0ΔCE, respectively) sequence in transgenic plants under dehydration conditions (Fig. 2B, D). All of the ABRE- or CE3-like-mutated promoters were unable to respond to dehydration treatment, whereas mutations in Box1 (N0-1m) and Box2 (N0-2m) had only minor effects on the promoter activity (Fig. 2B, D). Because ABRE was originally identified as a cis-element involved in ABA signaling, we suspected that the ABRE sequence in the DREB2A promoter might receive an ABA signal. Two-week-old plants containing one of several different promoter–GUS constructs were treated with ABA for 5 h, and their transcriptional activities were measured individually by qRT-PCR (Fig. 2E). We were able to detect the responses of the wild-type constructs (N0 and I × 4: 3.8- and 9.0-fold increases, respectively) to ABA, but the fold changes were smaller than the increases observed in response to dehydration stress (Fig. 2C: 25.9- and 25.1-fold increases, respectively). The ABRE-mutated constructs (N0-Am and Am × 4) did not respond to ABA (Fig. 2E). These data suggest that the ABRE sequence in Region I is responsible for the dehydration stress-induced activity of the DREB2A promoter and is also involved in the weak induction of DREB2A by ABA.
AREB/ABFs bind to the ABRE sequence of the DREB2A promoter
ABRE belongs to the larger G-box cis-element group, which is generally considered to be a target of the bZIP or basic helix–loop–helix (bHLH) transcription factor families (Choi et al. 2000, Uno et al. 2000, Sibéril et al. 2001). Therefore, we hypothesized that members of these transcription factor families might influence ABRE-dependent transcriptional regulation of the DREB2A promoter. Because the ABRE sequence in the DREB2A promoter is distinct from the standard target sequence of bHLH (CANNTG; Sibéril et al. 2001), we attempted to identify candidate transcription factors in the bZIP family that bind to the ABRE sequence. Using the methods described by Satoh et al. (2004), we screened a pool comprised of representative members of each bZIP family subgroup using the I × 4 and N0 constructs (Supplementary Fig. S2). An active form of AREB1 (AREB1ΔQT; Fujita et al. 2005), which represents subgroup A, was associated with the most prominent transcription activity of the DREB2A promoter. The bZIP family proteins that belong to subgroups B, C, E and I were also associated with transcription activity, although this activity was weak (Supplementary Fig. S2). The AREB/ABF family members are important for ABRE-mediated gene expression (Choi et al. 2000, Uno et al. 2000, Kang et al. 2002, Kim et al. 2004, Corrêa et al. 2008, Yoshida et al. 2010, Fujita et al. 2011). Therefore, we selected the AREB/ABF family as one of the candidate families for transcription factors that interact with ABRE in the DREB2A promoter.
Among the nine members of the AREB/ABF family, AREB1, AREB2 and ABF3 were selected because they have been reported to regulate environmental stress responses and ABA signaling coordinately in the vegetative stage (Fujita et al. 2005, Yoshida et al. 2010). To test the abilities of the three AREB/ABF members to bind to the DREB2A promoter, we used a yeast one-hybrid (Y1H) system (Fig. 3A, B). Either a 42 bp fragment of the DREB2A promoter proximal to the ABRE sequence (WT) or a mutated ABRE sequence (MT) was introduced into the yeast genome (Fig. 3A). The three AREB/ABF members were expressed individually in the yeast cells as fusion proteins with the GAL4 activation domain, and their binding affinities for the promoter were evaluated. On synthetic dextrose media (SD media) containing 3-amino-1,2,4-triazole (3-AT), the yeast lines expressing each tested AREB/ABF member showed more rapid growth than did the void vector control (VEC) in an ABRE-dependent manner (Fig. 3B). These results suggested that the AREB/ABF members can bind to the ABRE sequence in yeast. AREB1 is activated by phosphorylation, and two constitutively active forms (AREB1ΔQT, internal deletion; AREB1M8, phosphorylation mimic) and one constitutively inactive form (AREB1M5, phosphorylation negative) of this protein have been reported (Fujita et al. 2005, Furihata et al. 2006). We adapted these derivatives of AREB1 for use in our Y1H system and found no significant differences among the binding signals of these mutant forms (Fig. 3B).
To confirm the interaction of AREB/ABFs with the ABRE sequence in the DREB2A promoter in planta, chromatin immunoprecipitation (ChIP) was performed using Arabidopsis plants expressing hemagglutinin (HA)-tagged AREB1 driven by the 35S promoter (Fig. 3C–E). Because the relationship between stress signal acceptance and the DNA-binding affinity of the AREB/ABF proteins is unknown, we treated the plants with ABA or dehydration before ChIP. In both mock- and stress-treated plant lines, positive binding signals were detected by PCR (Fig. 3D) and quantitative PCR (Fig. 3E). Taken together with the results from the Y1H system, these findings suggest that the three AREB/ABF members could interact directly with the ABRE sequence of the DREB2A promoter. In addition, the activation of AREB1 by an ABA or stress signal seems to be distinct from its binding affinity for the target sequence.
AREB/ABF transcription factors can activate the DREB2A promoter
We evaluated the transactivity of the three AREB/ABF members with respect to the DREB2A promoter through a transient expression assay using Arabidopsis mesophyll protoplasts (Fig. 4). In addition to the three AREB/ABF members, we used AREB1ΔQT, which exhibits constitutive transactivity in the absence of ABA (Fujita et al. 2005). An RD29Bpro construct, including five tandem repeats of doubled ABREs containing a 77 bp fragment from the RD29B promoter region, was used as a positive control reporter. This construct has been reported to show a steady response to ABA-dependent AREB/ABF activation in a transient expression assay system (Fig. 4A) (Fujita et al. 2005, Furihata et al. 2006, Yoshida et al. 2010). In comparison with the RD29Bpro reporter construct, all of the tested AREB/ABF members showed reduced but still significant transactivity related to the N0 construct in an ABA-dependent manner (Fig. 4B). Similar to the loss-of-function analysis in the Y1H system, the ABRE-mutated MT construct showed no activity associated with any of the AREB/ABF members (Figs. 3B, 4B). These results suggest that the investigated AREB/ABFs might be among the transcription factors that interact with the ABRE sequence in the DREB2A promoter and activate its expression.
The stress-responsive activity of the DREB2A promoter is impaired in ABA-insensitive and ABA-deficient mutants
Studies on two triple null mutants [null for AREB1, AREB2 and ABF3 (areb1 areb2 abf3; areb triple) or SnRK2D, SnRK2E and SnRK2I (srk2d srk2e srk2i; srk2 triple)] have shown AREB/ABFs and SnRK2s to be key factors influencing gene expression as part of a signaling cascade in response to ABA and osmotic stress signals (Fujii and Zhu 2009, Fujita et al. 2009, Nakashima et al. 2009a, Yoshida et al. 2010). Microarray and qRT-PCR data from these reports revealed transcriptional alleviation of DREB2A under high salinity and dehydration stress conditions in these two triple mutants (Supplementary Table S1) (Fujita et al. 2009, Yoshida et al. 2010). Based on these data, we performed dehydration treatments using representative ABA mutant lines (Fig. 5A). ABA INSENSITIVE 1 and 2 (ABI1 and ABI2), which encode type 2C protein phosphatases, have negative regulatory roles in ABA signaling, and abi1-1C and abi2-1C plants mutant for these genes exhibit ABA insensitivity (Leung et al. 1997, Raghavendra et al. 2010, Umezawa et al. 2010). The ABA-DEFICIENT 2 (ABA2) and 9-cis-EPOXYCAROTENOID DIOXYGENASE 3 (NCED3) genes encode key enzymes in the ABA biosynthetic pathway, and aba2-2 and nced3-2 mutant plants exhibit ABA deficiency (Cheng et al. 2002, Nambara and Marion-Poll 2005). Three-week-old Arabidopsis plants of these mutant lines and the wild type were subjected to dehydration, and their DREB2A transcript levels were measured by qRT-PCR before treatment and after 5 h of treatment (Fig. 5A). The transcript levels of RESPONSIVE TO DESSICATION 20 (RD20) and RESPONSIVE TO ABA 18 (RAB18) were also monitored as representative ABA-inducible genes and not under regulation by DREB2A (Sakuma et al. 2006a, Fujii and Zhu 2009, Fujita et al. 2009, Yoshida et al. 2010). We observed significantly impaired stress-inducible expression of DREB2A in both areb triple and srk2 triple mutant plants, in agreement with previous microarray data (Fig. 5A; Supplementary Table S1) (Fujita et al. 2009, Yoshida et al. 2010). The other mutant lines showed similar impaired induction of DREB2A transcript levels (Fig. 5A).
We observed a reduced transcriptional response of DREB2A to dehydration stress in these mutant lines. Next, we analyzed whether these mutations in the SnRK2–AREB signaling cascade affected the DREB2A promoter activity. The N0 construct was introduced into the areb triple mutant plants, and the transcript level of the GUS gene was measured before and after dehydration treatment (Fig. 5B). We detected significantly impaired dehydration induction of GUS expression in the mutant plants compared with wild-type plants harboring the N0 construct (Fig. 5B). These results suggest that DREB2A expression in response to dehydration stress is influenced by ABA signaling cascades. However, the remaining dehydration stress-responsive expression was detected at significant levels in all of the mutants analyzed, which suggests that ABA-independent signaling influences the DREB2A expression under dehydration stress conditions.
In this study, we performed a truncation assay for the DREB2A promoter to restrict the promoter region that affects transcriptional regulation under dehydration stress (Fig. 1). We discovered that ABRE and CE3-like sequences, which are known to be cis-elements involved in ABA-induced gene expression, are important for the dehydration and high-salinity stress-responsive promoter activity of DREB2A (Fig. 2; Supplementary Fig. S3). This discovery was unexpected because DREB2A gene expression has been reported to be weakly induced by exogenously applied ABA and strongly induced by dehydration and high salinity stresses (Liu et al. 1998, Nakashima et al. 2000, Sakuma et al. 2002). The comparison of the promoter sequence of Arabidopsis DREB2A with its orthologs in other sequenced eudicot species revealed a conservation of the CE3/ABRE–ABRE cassette with a similar positional relationship (Supplementary Fig. S1). This raises the possibility of a crucial role for the CE3/ABRE–ABRE cassette in the DREB2A promoter activity. Multiple ABRE motifs were found in the promoter of a DREB2A ortholog in the lycophyte, Selaginella moellendorffii, which might suggest an ancient origin for this ABRE regulation. In addition, an HSE sequence for the heat shock-responsive expression of DREB2A (Yoshida et al. 2011) is also conserved in these eudicot species in the vicinity of the CE3/ABRE–ABRE cassette (Supplementary Fig. S1). This promoter element positioning may facilitate the proper expression of DREB2A and its orthologs under various stress conditions.
We examined the relationship between DREB2A expression and an ABA signaling cascade. Subsequent analyses revealed that the DREB2A promoter can be a direct target of AREB1, AREB2 and ABF3. Our in vitro and in vivo data indicate that the three AREB/ABF members could recognize the ABRE sequence in the DREB2A promoter and activate it in response to osmotic stress conditions under which activation levels were reduced (Figs. 3, 4). In addition, loss-of-function experiments demonstrated positive roles for ABA and ABRE in the transcriptional regulation of DREB2A (Fig. 5). All of the analyzed ABA-deficient and ABA-insensitive mutant lines presented significantly impaired transcriptional responses of DREB2A to osmotic stress treatments, and the DREB2A promoter–GUS construct was associated with a similar phenomenon (Fig. 5; Supplementary Fig. S3). Collectively, our results established the contribution of ABA signaling cascades to DREB2A transcriptional regulation through the DREB2A promoter region under osmotic stress conditions.
Although an influence of ABA signaling cascades on DREB2A expression was revealed in this study, we hypothesize that the DREB2A expression that occurs in response to osmotic stress is not equivalent to that in response to ABA. The transcriptional response of DREB2A to exogenously applied ABA is weaker than the response to osmotic stress (Fig. 2; Liu et al. 1998, Nakashima et al. 2000, Sakuma et al. 2002), and >20% of DREB2A expression was retained in ABA-deficient mutants, whereas the expressional responses of RD20 and RAB18 were almost abolished in these mutant lines (Fig. 5A; Supplementary Fig. S3). Therefore, we hypothesized that the DREB2A promoter requires other stress signals to stimulate expression, which is expected to be independent of ABA signaling.
Several clues to explain this ABA-independent pathway can be obtained from different resources. In our previous study, only one-fifth of genes downstream of the three SnRK2s were found to be regulated by the three AREB/ABFs addressed here (Fujita et al. 2009). In the case of DREB2A, the srk2 triple mutation is >2-fold more effective in response to the osmotic stress-responsive transcription than the areb triple mutations (Fig. 5A; Supplementary Table S1; Fujita et al. 2009, Yoshida et al. 2010). These data support the conclusion that factors other than AREB/ABFs can influence DREB2A expression under the control of the SnRK2s. Considering the reported possibility that SRK2E, which is one of the three SnRK2 members, is capable of receiving ABA-independent osmotic stress signals (Yoshida et al. 2006), it might be expected that DREB2A expression is regulated by an ABA-independent pathway under the control of the SnRK2 members. Although downstream factors for such a pathway have not yet been identified, this possibility may explain the greater effect of the srk2 triple mutation on DREB2A expression compared with the ABA-deficient mutations (Fig. 5A). Several reports have suggested a considerable number of candidates as unknown signal mediators of DREB2A expression in response to osmotic stresses, such as WRKY, MYB or other bZIP family members (Liao et al. 2008, Zhou et al. 2008, Shang et al. 2010, Xie et al. 2010).
The DREB2A promoter sequences also provide us with several lines of evidence. The ABRE sequence contains an ACGT core, which is considered to be a target for plant bZIP proteins (Sibéril et al. 2001, Fujita et al. 2011). Additionally, an ACGT core can be found in other cis-elements of different transcription factor family members in plants, such as NAC and MYC (Boter et al. 2004, Takasaki et al. 2011). The CE3-like sequence, as a putative coupling element for the ABRE in the DREB2A promoter, should be studied further. CE3 was originally identified as a coupling element involved in achieving ABA induction in the barley HVA1 promoter (Shen et al. 1996). A genome-wide in silico analysis found that only a few Arabidopsis ABRE cis-elements are coupled with CE3 compared with the rice sequences (Gómez-Porras et al. 2007). These data support the hypothesis that the osmotic stress-responsive expression of DREB2A is regulated in a different manner to that of other ABA-responsive genes in Arabidopsis.
As a stress-responsive transcription factor, DREB2A governs many stress-responsive genes that may lead plants to tolerate unfavorable circumstances as well as potentially having adverse effects on plant growth and reproduction (Sakuma et al. 2006a, Sakuma et al. 2006b, Maruyama et al. 2009). Therefore, a complex series of steps is expected to be involved in the regulation of DREB2A expression in plants. In this study, we demonstrated that ABA signaling cascades play important roles in DREB2A expression in response to osmotic stress. The ABRE and CE3-like sequences in the DREB2A promoter might function as cis-elements related to DREB2A expression, and the ABRE sequence could physically interact with AREB1, AREB2 and ABF3 (Fig. 6). Furthermore, a lack of ABA signaling resulted in a significant decrease of the DREB2A promoter activity under osmotic stress conditions, and the presence of ABA-independent signaling cascades was revealed through the limited effect of an ABA signal on DREB2A expression. In addition to the recently revealed post-translational regulation of DREB2A activity in response to osmotic stress (Qin et al. 2008), multiple signaling cascades contribute to the osmotic stress-responsive expression of DREB2A via both ABA-dependent and ABA-independent pathways (Fig. 6). The existence of such multiple and complex pathways for regulating DREB2A expression could allow plants to respond rapidly and precisely to osmotic stress.
Materials and Methods
Plant material and growth conditions
The plants used in this study were A. thaliana (L.) Heynh. ecotype Columbia. Seeds were sterilized and stratified at 4°C for 2 d in the dark, then sown and grown on growth medium (GM) agar plates in a growth chamber under 16 h of fluorescent light (40 ± 10 μmol m−2 s−1) at 22°C and 70% relative humidity (Fujita et al. 2005).
Arabidopsis transformation was achieved using the floral-dip method (Clough 2005). Harvested T1 candidate seeds were selected on GM agar plates containing 50 mg l−1 kanamycin sulfate or 10 mg l−1 hygromycin with 25 mg l−1 cefotaxime. To obtain homozygous lines, T2 and T3 generations were subjected to the same antibiotic resistance selection protocol. The verified T3 lines were used for further analyses.
Isolation of DREB2A promoter fragments and promoter–GUS construction
All members of a truncation series of the DREB2A promoter were amplified from Arabidopsis genomic DNA. Nucleotide substitutions in the promoter construct were generated by mismatched amplification with each specific primer set (Supplementary Table S2). Each amplified promoter fragment was inserted into the BamHI site of pGK-GUS (Qin et al. 2008) or pGH-GUS (see Supplementary data). The sequence of each construct was verified by comparison with The Arabidopsis Information Resource (TAIR; http://www.arabidopsis.org/) database. Verified constructs were individually introduced into Agrobacterium tumefaciens strain GV3101 (pMP90) by a freeze–thaw method and were subsequently used for Arabidopsis transformation.
Stress and ABA treatment
Three-week-old Arabidopsis plants (principal growth stage 1.08–1.10; Boyes et al. 2001) grown on GM agar plates were subjected to dehydration treatments. For each treatment, 3–5 plants were placed on a piece of Parafilm™ at the same distance (1.0–1.5 cm) and held under 60–65% relative humidity. Following treatment, samples were collected and stored at −80°C. As negative controls, non-treated samples were stored directly at −80°C.
Two-week-old Arabidopsis plants (principal growth stages 1.02–1.03) grown on GM agar plates were used for high salinity or ABA treatments. For each treatment, 5–6 plants were soaked in water containing 50 μM ABA or 250 mM NaCl. Following the treatments, samples were placed on a piece of paper to remove surface water before storage at −80°C. In mock treatments, the same volume of solvent in which the ABA stock solution had been prepared was used. Stored samples were used for RNA extraction and qRT-PCR.
RNA extraction and qRT-PCR
Total RNA was isolated from each plant sample using RNAiso plus™ (TAKARA BIO INC.; http://www.takara-bio.co.jp/) according to the manufacturer's instructions. Isolated RNA was employed for cDNA synthesis using Superscript III™ reverse transcriptase (Invitrogen™; http://www.invitrogen.com/). qRT-PCR was performed using a 7500 real-time PCR system and 7500 system SDS software, version 188.8.131.52 (Applied Biosystems™; http://www.appliedbiosystems.com/). SYBR® Premix Ex Taq™ (TAKARA BIO INC.) was used for amplification. Arabidopsis 18S rRNA was used as a quantitative control for the amount of template. The sequences of the primer sets are shown in Supplementary Table S2.
The PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/; Lescot et al. 2002) was used for scanning the cis-elements in the DREB2A promoter sequence. The significance of candidate cis-elements was evaluated through scanning further with the other plant cis-element database PLACE (http://www.dna.affrc.go.jp/PLACE/; Higo et al. 1999).
Yeast one-hybrid (Y1H) assay
We used a modified MATCHMAKER™ One-Hybrid System (Clontech Laboratories; http://www.clontech.com/) to perform the Y1H assay. Specifically, pHisi-1 and pLacZi vectors, both harboring four tandem repeats of the DREB2A promoter fragment (WT, MT; Fig. 3A), were integrated into the genome of yeast strain YM4271 followed by selection on SD media agar plates without uracil or histidine (SD/–UH).
To verify promoter–protein interactions in yeast, putative target proteins and their derivatives were individually introduced into the pGADT7 vector (Clontech Laboratories) using BamHI and XhoI restriction enzyme sites. Following sequence verification, each construct was transformed into yeast lines harboring each repeated DREB2A promoter region using a Fast™ Yeast Transformation kit (Geno Technology Inc.; http://www.gbiosciences.com/). Transformed colonies were spotted on SD agar media that did not contain uracil, histidine or leucine (SD/–UHL) but did contain 0, 15 or 40 mM 3-AT in a concentration gradient (OD600). After several days of incubation at 30°C, the interaction affinity was evaluated based on the growth status of each spot.
Chromatin immunoprecipitation (ChIP)
Samples of 1–1.3 g were collected from whole 3-week-old Arabidopsis plants (principal growth stage 1.08–1.10) harboring the doubled HA (2 × HA)-tagged AREB1 protein-expressing construct (see Supplementary data) and subjected to ABA or dehydration treatments, as described above. ChIP was performed using an EpiQuik™ Plant ChIP Kit (Epigentek; http://www.epigentek.com/) according to the provided user guide. To immunoprecipitate the HA-tagged proteins, an anti-HA antibody (catalog No. H3663, Sigma-Aldrich; http://www.sigmaaldrich.com/) was used. The primer sets employed to amplify the DREB2A promoter regions and 18S rDNA are shown in Supplementary Table S2.
Transient expression assay with Arabidopsis mesophyll protoplasts
Transient transformation using Arabidopsis mesophyll protoplasts was achieved as described by Yoshida et al. (2010). The effector constructs were adapted from our previously reported research (Fujita et al. 2005, Furihata et al. 2006, Yoshida et al. 2010).
Sequence data from this article can be found in the TAIR database under the following accession numbers: ABA2 (AT1G52340), ABF3 (AT4G34000), ABI1 (AT4G26080), ABI2 (AT5G57050), AREB1/ABF2 (AT1G45249), AREB2/ABF4 (AT3G19290), DREB2A (AT5G05410), NCED3 (AT3G14440), RAB18 (AT5G66400), RD20 (AT2G33380), RD29A (AT5G52310), RD29B (AT5G52300), and SRK2D (AT3G50500), SRK2E (AT4G33950), SRK2I (AT5G66880).
Supplementary data are available at PCP online.
This work was supported by the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of Japan [a Grant-in-Aid for Scientific Research on Innovative Areas (22119004), Targeted Proteins Research Program (07088014) and a Grant-in-Aid for Young Scientists (B) (21780314) (to J.M.)]; the Science and Technology Research Partnership for Sustainable Development (SATREPS) of the Japan Science and Technology Agency/Japan International Cooperation Agency; Genomics for Agricultural Innovation, Development of Abiotic Stress-Tolerant Crops by DREB Genes of the Ministry of Agriculture, Forestry and Fisheries (MAFF) of Japan; the Program for the Promotion of Basic Research Activities for Innovative Biosciences (BRAIN) of Japan.
We thank K. Amano and E. Kishi in particular for their excellent technical assistance, K. Morimoto for work with protein, and M. Toyoshima for skillful editorial assistance.
ABA-responsive element-binding factor
ABA-responsive element-binding protein
basic leucine zipper
coupling element 3
dehydration-responsive element-binding protein
quantitative real-time PCR