The biological functions of WRKY transcription factors in plants have been widely studied, but their roles in abiotic stress are still not well understood. We isolated an ABA overly sensitive mutant, abo3, which is disrupted by a T-DNA insertion in At1g66600 encoding a WRKY transcription factor AtWRKY63. The mutant was hypersensitive to ABA in both seedling establishment and seedling growth. However, stomatal closure was less sensitive to ABA, and the abo3 mutant was less drought tolerant than the wild type. Northern blot analysis indicated that the expression of the ABA-responsive transcription factor ABF2/AREB1 was markedly lower in the abo3 mutant than in the wild type. The abo3 mutation also reduced the expression of stress-inducible genes RD29A and COR47, especially early during ABA treatment. ABO3 is able to bind the W-box in the promoter of ABF2in vitro. These results uncover an important role for a WRKY transcription factor in plant responses to ABA and drought stress.
Drought stress is one of the most severe environmental culprits that greatly restrict plant distribution and crop production (Zhu, 2002). Drought stress induces the accumulation of the plant hormone abscisic acid (ABA), which leads to stomatal closure for maintaining water status in plant cells under water-deficit conditions. The increased ABA interacts with the ABA receptors PYR/PYLs of START proteins, which interact with PP2C proteins and release the inhibition of PP2Cs on SnRK2 protein kinases. The activated SnRK2s phosphorylate downstream transcriptional factors such as ABSCISIC ACID RESPONSIVE ELEMENTS-BINDING FACTOR 2 (ABF2) and ABI5 (ABA insensitive 5, which is a bZIP protein) to regulate the expression of ABA response genes (Fujii and Zhu, 2009; Ma et al., 2009; Nakashima et al., 2009; Park et al., 2009). Various genes are up- or downregulated at the transcriptional level by both drought stress and ABA treatment. Analyzing the promoters of ABA-inducible genes has identified some conserved cis-elements, one of which is ABRE (ABA-responsive element, PyACGTGGC) (Guiltinan et al., 1990; Yamaguchi-Shinozaki et al., 1990; Iwasaki et al., 1995; Shen et al., 1996). Transcription factors such as AREB (ABRE binding protein)/ABFs, and ABI5 could bind ABRE and regulate the expression of ABA-responsive genes (Uno et al., 2000; Carles et al., 2002; Casaretto and Ho, 2003). The transcripts of ABFs, including ABF1, ABF2/AREB1, ABF3 and ABF4/AREB2, are also highly induced by the application of exogenous ABA (Choi et al., 2000; Uno et al., 2000). Plants overexpressing ABF3 and ABF4 showed increased ABA sensitivity in seed germination and seedling growth, reduced transpiration and more drought tolerance than wild-type plants (Kang et al., 2002). However, constitutive overexpression of ABF2/AREB1 did not increase the expression of downstream ABA-responsive genes, because the activation of ABF2/AREB1 needs ABA-triggered protein phosphorylation (Fujita et al., 2005). An ABA-activated 42-kDa kinase can phosphorylate and activate ABF2/AREB1 (Fujita et al., 2005; Furihata et al., 2006). When the phosphorylatable Ser/Thr residues were substituted with Asp to mimic phosphorylation, the resultant active form of AREB1 induced the expression of many downstream ABA-responsive genes, even without ABA treatment (Furihata et al., 2006).
In this study, we isolated an ABA-hypersensitive mutant, abo3 (ABA overly sensitive 3), which is caused by a T-DNA insertion in an ABA-upregulated WRKY gene. Although the knock-out of ABO3 renders plants hypersensitive to ABA in seed germination and seedling growth, abo3 mutant plants showed reduced ABA sensitivity in guard cells, lost water faster and were more sensitive to drought stress than wild-type plants. We found that abo3 mutation impaired the expression of ABF2 and downstream genes such as RD29A and COR47 during early ABA treatment. Gel shift analysis revealed that ABO3 protein binds to the W-box elements localized in the ABF2 promoter. However, transgenic plants overexpressing ABO3 showed no ABA or ABF2 expression phenotypes, which suggests that ABO3 might need co-factor(s) or modifications for transactivating downstream target genes.
Disruption of ABO3 gene in a T-DNA insertion mutant increases ABA sensitivity
Different concentrations of ABA can inhibit both seed germination and seedling growth of Arabidopsis. To study ABA and drought-tolerance mechanisms, we used a root-bending assay to search for mutants in which root growth is more sensitive or insensitive to ABA (Yin et al., 2009). As the root growth of the Columbia accession is more insensitive to ABA than some other accessions, such as Landsberg and C24, we have usually used 20–30 μm ABA or more to test the root growth sensitivity (Yin et al., 2009). Five-day-old seedlings grown on MS medium were transferred with roots upside down onto vertical agar plates containing MS medium supplemented with 20–30 μm ABA (Yin et al., 2009). One of the ABA-hypersensitive mutants, abo3, was isolated from the T-DNA collection of SALK lines (Salk_007496). The T-DNA insertion site is located in the third exon of At1g66600 encoding a WRKY transcription factor, AtWRKY63 (Figure 1a). The insertion would be expected to completely disrupt the expression of At1g66600. RT-PCR analysis confirmed that there were At1g66600 transcripts in the wild type but not in the abo3 mutant (Figure 1b). Northern blot analysis did not detect the transcripts of AtWRKY63/ABO3 in our conditions, probably because of its low expression level. There is no information on microarray data for this gene. Real-time RT-PCR analysis revealed that ABO3 expression is upregulated by ABA treatment in either Columbia or Landsberg (Figure 1c). We further checked the ABO3 expression in different abi mutants. The transcriptional induction of ABO3 by ABA was impaired in abi1-1 (Leube et al., 1998), abi2-1 (Rodriguez et al., 1998) (two dominant-negative mutations in Landsberg), abi3-1 (Parcy et al., 1997) (in Landsberg) or abi5 (a point mutation in the fifth nucleotide before the first putative ATG, detail information about this mutant can be found in TAIR; in Columbia) mutants, suggesting that ABI1 and ABI2 negatively, and ABI3 and ABI5 positively, regulate ABO3 expression. However, ABO3 transcripts were still induced in the abi4-1 mutant (Finkelstein et al., 1998), indicating that ABI4 did not regulate ABO3 expression.
We investigated ABA sensitivity during seedling establishment. As Arabidopsis seedling establishment is more sensitive to ABA than root growth, we used low concentrations of ABA in seedling establishment and high ABA concentrations in root growth. One week after seed germination on MS media containing 0.1–0.5 μm ABA, wild-type and abo3 mutant seedlings with green cotyledons were compared. As shown in Figure 1d,e, the abo3 mutant was more sensitive to ABA than the wild type during seedling establishment.
Root growth of the abo3 mutant was slower than that of the wild type on MS medium (Figure 1f,g), which suggests that the abo3 mutation influences root development. Adding different concentrations of exogenous ABA to the MS medium inhibited both root and shoot growth of abo3 and the wild type, but with more inhibition in abo3 than in the wild type (Figure 1f,g).
To determine whether the abo3 phenotypes are caused by the disruption of AtWRKY63, we amplified AtWRKY63 cDNA by RT-PCR and overexpressed it under the control of the constitutive CaMV 35S promoter in abo3 mutant plants (Figure 1d). We obtained eight independent transgenic lines, and randomly picked up three lines. All of the three lines complemented the growth and ABA-sensitive phenotype of the abo3 mutant in T3 homozygous plants. Here, we took line 4 as an example for detailed analysis. RT-PCR revealed that the AtWRKY63 transcript was overexpressed in line 4 (Figure 1b). We analyzed line 4 on MS plates supplemented with different concentrations of ABA or no ABA for both seed germination and root growth. Transgenic line 4 displayed a wild-type phenotype on MS medium containing different concentrations of ABA, or no ABA, in both seed germination (Figure 1d,e) and root growth assays (Figure 1f,g). We further performed drought-tolerance and stomatal movement tests (Figure 2a–e) and northern blot analysis for the expression of ABF2 (a putative target of AtWRKY63; Figure 5b, see detailed explanation later), which further confirmed that the abo3 mutant phenotypes were completely rescued by 35S- AtWRKY63. Therefore, the disruption of AtWRKY63 is responsible for the phenotypes observed in the abo3 mutant.
abo3 mutation impairs ABA-induced stomatal closure, and the abo3 mutant is more sensitive to drought stress than the wild type
Plant drought tolerance is influenced by a combination of multiple molecular and cellular pathways. Some ABA-hypersensitive mutants, such era1 (Pei et al., 1998) and abo1 (Chen et al., 2006), are more tolerant to drought stress, whereas the sad1 mutant is more sensitive to both ABA and drought stress (Xiong et al., 2001). Plants overexpressing AtHD2C are insensitive to ABA but show increased drought tolerance (Sridha and Wu, 2006). Because abo3 mutants are more sensitive to ABA in both seedling growth and seed germination, we tested the drought-tolerance phenotype of the abo3 mutant. Plants were grown for 3 weeks and then exposed to dehydration by withholding water for 2 weeks, abo3 mutant plants showed a less resistant phenotype to drought than wild-type plants or complementation line 4 (Figure 2a). After drought treatment for two more weeks, plants were re-watered and continually cultured for 1 week to count the plant survival rate. The abo3 mutant displayed a higher death rate when compared with the wild type or with line 4 (Figure 2c). We further evaluated water loss rates with detached leaves. Detached leaves of abo3 lost water more quickly than wild-type or line-4 leaves (Figure 2b), consistent with the increased drought sensitivity of the mutant.
Because water loss mainly depends on stomatal regulation, we then compared stomatal apertures under treatment with different concentrations of ABA. The epidermal peels of abo3 and wild-type or complementation line-4 plants were incubated in a buffer solution under strong light conditions for 12 h to fully open the stomata. Then, the peels were treated with different concentrations of ABA for 2 h. The ratio of stomatal length to width indicates the degree of stomatal closure. abo3 and wild-type or line-4 plants showed the same ratio of length to width of fully opened stomata without ABA treatment, but wild-type or line-4 plants showed higher ratios of length to width of stomata than abo3 mutant plants after treatment with ABA (Figure 2d,e). These results suggest that stomata closure in abo3 is less sensitive to ABA than that in the wild type or line 4. The drought sensitivity of the abo3 mutant as well as its impaired stomatal closure under exogenous ABA treatment suggests that ABO3 functions in ABA-mediated drought stress response pathways.
To test the potential impact of ABO3 overexpression on plant responses to ABA and drought, we introduced the 35S-AtWRKY63 construct to wild-type plants, and obtained 11 independent transgenic lines. Through northern blot, we selected two homozygous T3 lines (OE8 and OE11) with different ABO3 expression levels for further phenotype analysis (Figure 3c). However, we did not observe any ABA-related phenotype either in seed germination or in root bending of the overexpression lines (Figure 3b–d). Drought treatment and stomata closure experiments also lacked any visible phenotype (data not shown). These results suggest that ABO3 protein may need other components or ABA or drought-induced post-translational modifications to activate downstream genes in plant stress responses.
ABO3 is a nuclear protein with transcriptional activation activity
To explore the spatial expression pattern of ABO3, we obtained 14 independent T3 transgenic Arabidopsis lines that harbored an ABO3 promoter::GUS reporter construct, and six lines were analyzed by GUS staining with similar expression patterns. We found the expression of the ABO3 promoter::GUS fusion gene very low and detectable only after staining for a long time (2 days) (Figure 4a; i, whole seedling; ii, root; iii, stem; iv, leaf; v, flower). The GUS expression appeared higher in roots and flowers than in stems and leaves. A very low level of GUS staining was detected in the early development of seedlings (24–72 h after seed germination; Figure 4bi–iii), or only after ABA treatment in guard cells (Figure 4bv). Consistent with this GUS staining pattern, RT-PCR results also indicated relatively higher expression in roots and flowers (Figure 4avi). Interestingly, the GUS expression in the leaves was concentrated in trichomes (Figure 4iv, Leaf, biv). The low expression of ABO3 is consistent with public microarray data, in which no expression data are available for AT1G66600/ABO3.
To examine the subcellular distribution of the ABO3 protein, we fused ABO3 with GFP (35S::ABO3-GFP). Confocal imaging showed the ABO3-GFP fusion protein localized exclusively in the nuclei of onion (Allium cepa) epidermal cells in a transient expression assay (Figure 4c). As a control, the GFP protein was found in both the nucleus and cytoplasm. The nuclear localization of ABO3 is consistent with its predicted function as a transcription factor.
The WRKY proteins comprise a superfamily of transcription factors that can be divided into three subgroups according to the conserved WRKY DNA-binding domain sequences (Wu et al., 2005). ABO3/AtWRKY63 belongs to the third subgroup of WRKY proteins (Wu et al., 2005). To investigate whether ABO3 possesses transcriptional activation activity, we assayed the GAL4 DNA binding domain-ABO3 fusion protein in yeast for its ability to activate transcription of the GAL4 upstream sequence-driven His and LacZ reporter gene expression. The yeast growth on the His-deficient medium, LacZ staining and relative quantitative assay of β-galactosidase activity in Figure 4D indicate that the transactivation sites of ABO3 are located in the W domain and C-terminal region of ABO3.
ABO3 positively regulates the expression of the ABF2 transcription factor
The drought and ABA response phenotypes of the abo3 mutant suggest that the expression of some ABA-responsive genes might be altered in the mutant. To verify this possibility, we selected various stress-inducible marker genes, including ABF2, ABF3, RD29A, COR47, DREB2A, RD22 and KIN1, and analyzed their expression under treatments with 20 and 100 μm ABA by northern blots. RD29A, COR47 and KIN1 are stress-responsive genes with a presumed protective function against stress damage, and part of their stress induction is dependent on ABA (Kurkela and Franck, 1990; Lin et al., 1990; Guo et al., 1992; Yamaguchi-Shinozaki and Shinozaki, 1993a, 1994; Wang and Cutler, 1995); expression of another stress-responsive gene, RD22, is induced by salt, drought and ABA through a pathway dependent on ABA-induced MYC and MYB genes (Yamaguchi-Shinozaki and Shinozaki, 1993b). ABF2, ABF3 and DREB2A are ABA-responsive transcription factors (Liu et al., 1998; Uno et al., 2000; Kang et al., 2002; Kim et al., 2004). As shown in Figure 5a, all of the tested genes were induced by different concentrations of ABA treatments after 2 and 5 h. The levels of ABF3, DREB2A, RD22 and KIN1 did not differ between abo3 and wild-type plants. However, the expression of ABF2 was greatly reduced in abo3 compared with wild-type plants, especially at 2 h. The transcripts of RD29A and COR47 were reduced in the abo3 mutant compared with the wild type upon treatment after 2 h, but not after 5 h. These results indicate that ABO3 is critical for the expression of some ABA-inducible genes at early time points during treatment.
We further monitored the expression patterns of ABF2, RD29A and COR47 during a 2-h, 20-μm ABA treatment (Figure 5b). The transcript of ABF2 was detected at 30 min in the wild type but not in the abo3 mutant, and was less abundant in the abo3 mutant than in the wild type at 60 and 120 min. The transcript levels of RD29A and COR47 were lower in the abo3 mutant than in the wild type at all three time points. Here, we also included abo3 plants transformed with 35S-AtWRKY63 to show that the construct restored the expression of ABF2, RD29A and COR47 to the wild-type levels at 120 min.
ABO3 is able to bind the W-box in the ABF2 promoter in vitro
Plant WRKY proteins recognize various W-box elements with a TGAC core sequence in the promoters of many defense-related genes (Yu et al., 2001). The promoter region of ABF2 contains two potential W-box elements within a 1-kb region upstream of the transcription initiation site: one reverse sequence, APN1, is located between −868 and −838 bp, and the other forward sequence, APN2, is located between −193 and −163 bp. A double-stranded APN1 or APN2 probe containing one typical W-box was used in an electrophoresis mobility shift assay (EMSA) to determine whether the ABO3 protein can recognize the TGAC core sequence. Recombinant ABO3-GST protein (Figure 6a) from Escherichia coli was incubated with [γ-32P]ATP-labeled APN1/2 probes. The retarded complexes were detected with either APN1 or APN2, which indicates that ABO3 can bind to these two sequences, with the binding to APN2 being stronger (Figure 6b). To determine the binding specificity, we tested mutant probes (mAPN1/2) in which the TTGAC sequence of each W-box was changed to TTAAA. The mutant probes failed to produce retarded bands on incubation with the recombinant ABO3 protein (Figure 6b). Upon incubation with 100 times more unlabeled PNA1/2 probes, the retarded bands became weaker. This competition assay further indicates that ABO3 specifically recognizes the W-box in vitro. These results suggest that the interaction between DNA and ABO3 requires the WRKY-recognition sequences. Thus, ABO3 is capable of recognizing two W-boxes in the ABF2 promoter, which is consistent with its important role in vivo in controlling ABF2 expression under ABA treatment.
Genetic analysis of abo3 with other ABA-responsive mutants
To further investigate the function of ABO3 in the ABA signaling pathway, we crossed abo3 with abi mutants abi1-1, abi2-1, abi3-1, abi4-1 and abi5, and obtained double mutants of abi1-1/abo3, abi2-1/abo3, abi3-1/abo3, abi4-1/abo3 and abi5/abo3. As shown in Figure 7a,b, the double mutants of abi1-1/abo3, abi2-1/abo3, abi3-1/abo3 and abi5/abo3 displayed a similar ABA-resistant phenotype as their corresponding single abi mutants, although abi1, abi2 and abi3 are in the Landsberg background, which is more sensitive to ABA than Columbia in seed germination. However, the seed germination of the abi4-1/abo3 double mutant was more sensitive than abi4-1 was to ABA, which is consistent with the real-time RT-PCR analysis that upregulation of the ABO3 transcript by ABA is not influenced by abi4 mutation. We also compared the expression of these ABI genes under ABA treatment by real-time RT-PCR, and found that there was no apparent expression difference between the wild type and abo3 (Figure 7b). These results indicate that ABO3 may act at, or upstream of, ABI1, ABI2, ABI3 and ABI5 in the ABA pathway, but might act in a different pathway with ABI4. It seems that ABO3 does not directly regulate the expression of these ABI genes, although they may be genetically related to each other in the ABA signaling pathway.
Our data demonstrate a critical role for a WRKY transcription factor encoded by ABO3 in ABA response and drought tolerance. The expression of ABF2/AREB1, a key transcriptional activator for ABA-responsive genes is largely dependent on ABO3. ABO3 is capable of binding the core sequences in the promoter region of ABF2/AREB1, so ABF2 is likely to be one target of ABO3. Interestingly, the expression of ABO3 itself is upregulated by ABA, but overexpression of the ABO3 protein doesn’t induce any ABA- or drought-related phenotype, except for mutant complementation. These observations suggest that there might be a complex network of ABA-responsive transcription factors involved in stress and ABA responses.
Previous studies found that ABF2/AREB1 is an important mediator in glucose signaling and drought stress tolerance in vegetative tissues (Kim et al., 2004; Fujita et al., 2005). Overexpression of an activated form of AREB1 (AREB1ΔQT) increased the expression of some stress-inducible genes and greatly enhanced drought tolerance in Arabidopsis (Fujita et al., 2005). However, AREB1ΔQT-overexpressing plants did not change ABA-mediated stomatal closure, which suggests that the increased drought tolerance was attributed mainly to the elevated expression of downstream genes, such as LEA-type genes (Fujita et al., 2005). Nevertheless, we found that mutation in ABO3 impaired the sensitivity of ABA-mediated stomatal closure. Our data suggest that the drought-sensitive phenotypes exhibited by the abo3 mutant may be attributable to both the impaired stomatal closure and lower expression of some downstream ABA-responsive genes.
Contrary to the reduced ABA sensitivity of guard-cell movement, mutation in ABO3 resulted in an increased sensitivity of both seed germination and root growth to ABA. Furthermore, abo3 mutant plants showed some root developmental defects. These results suggest that ABO3 functions in regulating both plant growth and stress tolerance related to ABA response pathways. In this case, plant hypersensitivity to ABA in seed germination and seedling growth is not accompanied by higher drought tolerance. Similar results were also reported for AtHD2C, CaXTH3, SAD1 and AtTPS1, and overexpression of rice OsMYB3R-2 in Arabidopsis (Xiong et al., 2001; Avonce et al., 2004; Cho et al., 2006; Sridha and Wu, 2006; Dai et al., 2007). ABO3 appears to play a negative role in the inhibition of seed germination and root growth by ABA, but plays a positive role in ABA-mediated stomatal closure. In addition to the reduced expression of ABF2 in the abo3 mutant, the expression of RD29A and COR47 was also reduced in the abo3 mutant compared with the wild type, especially during early ABA treatment. ABO3 might directly regulate the expression of RD29A and COR47, or indirectly control their expression, through transcription factors such as ABF2.
Genetic analysis combining with various abi mutants suggests that ABO3 is one of the components in the ABA signal pathway. Double mutants of abo3 with abi1, abi2, abi3 and abi5 all showed the ABA insensitivity, which is similar to the ABA-insensitive phenotype exhibited by these ABA-insensitive mutants in seed germination. ABO3 transcripts were not induced by ABA more in these abi mutants. Recent work revealed that ABI1 and ABI2 function redundantly in early ABA signaling after the perception of ABA by ABA receptors (Ma et al., 2009; Park et al., 2009). ABI3 acts as a regulator of the downstream target ABI5, which can be phosphorylated by OST1 in vitro (Lopez-Molina et al., 2002; Umezawa et al., 2009). Interestingly, only the abi4-1/abo3 double mutant did not display an ABA-insensitive phenotype like the abi4-1 mutant. Consistent with this observation, the induced transcripts of ABO3 by ABA were not disturbed in the abi4 mutant. These data suggest that ABO3 and ABI4 might operate in different ABA signaling pathways.
We did not observe any phenotypic changes in transgenic plants overexpressing ABO3. The full function of ABO3 might need some ABA-triggered post-translational modification, or ABO3 might need to cooperate with other ABA-related proteins for its capacity. Recent work indicated that ABF2 can be phosphorylated by OST1 (Fujii and Zhu, 2009). We also tried to discover whether ABO3 could be phosphorylated by OST1 in vitro, but got negative results (data not show). Previous work also suggests that the senescence-related WRKY53 protein may need the mitogen-activated protein kinase kinase kinase MEKK1 for its W-box binding activity (Miao et al., 2007). It is speculated that ABO3 may be modified by some protein kinases other than OST1. On the other hand, our evidence indicated that WRKY18, WRKY40 and WRKY60 interact both physically and functionally in a complex for plant responses to different pathogens (Xu et al., 2006). In a future study, the possibility of whether ABO3 works in a complex with other proteins needs to be explored.
Plant growth conditions and abo3 T-DNA insertion
Plants were grown in 340-ml pots filled with a mixture of peat/forest soil and vermiculite (3:1) in a glasshouse at 21°C with a light intensity of 50 μmol m−2 sec−1 and 70% relative humidity (RH) under long-day conditions (16-h light/8-h dark). Seedlings were germinated and grown on MS medium (Chen et al., 2005) supplemented with 3% (w/v) sucrose and 0.8% agar under the same growth conditions. The plate-grown seedlings were transferred to soil after 1 week.
For abo3 (Salk_007496) T-DNA insertion identification, total genomic DNA was extracted from two rosette leaves. PCR was performed using primer pairs: SALK_007496LP 5′-CTTCTTCTCGAGACATGGCAG-3′ and SALK_007496RP 5′-TTGTTCCATGTTGTGAGGTTG-3′; TF 5′-GCGTGGACCGCTTGCTGCAACT-3′.
Plasmid constructs and plant transformation
RT-PCR was performed on total RNA extracted from 2-week-old Arabidopsis thaliana Columbia seedlings with the use of TRIzol (Invitrogen, Cat. No. 15596-026). The first-strand cDNA was synthesized with an 18-mer oligo(dT) primer. Subsequent PCR amplification of ABO3 cDNA involved the primer pair ABO3F, 5′-ATGTTTTCAAACATCGATCACAAGGCTGTGG-3′, and ABO3R, 5′-CAACATCAGGTCTTCCGATGAAAATAGAGGAAATTCATTCC-3′. The PCR products were cloned into the CaMV 35S promoter driving the binary vector pMDC32 (for complementation and overexpression analysis) or pMDC85 (for protein targeting; Curtis and Grossniklaus, 2003) with Gateway Technology (Invitrogen). A promoter fragment, 1544 bp of the ABO3 gene, pABO3, was amplified from Columbia genomic DNA by PCR with the primer pair 5′-TGAGCCCTCTGATCTCCTTACGACTTTACGTGTCTTTGTC-3′ and 5′-CATCGATCACAAGGCTGTGGCAGC-3′. The amplified fragment was cloned into the pCAMBIA 1391 for a transcriptional fusion of the ABO3 promoter with the GUS coding region. The constructs were introduced into Agrobacterium tumefaciens strain GV3101 and transformed by floral infiltration into wild-type Arabidopsis (Columbia accession, for gene overexpression and GUS staining assays) and abo3 mutants (for gene complementation).
Stomatal aperture bioassays and water loss measurements
Stomatal closing assays were conducted as described by Chen et al. (2006). Rosette leaves were floated in solutions containing 50 μm CaCI, 10 mm KCl, 10 mm Mes-Tris, pH 6.15, and exposed to light (150 μmol m−2 sec−l) for 2 h. Subsequently, ABA was added to the solution at 0.5 and 1 μm to assay for stomatal closing. After ABA treatment for 2 h, stomatal apertures were measured as described by Chen et al. (2006). Values are means ± SE, n = 60. Significance (P < 0.05) was assessed by the Student’s t-test.
Rosette leaves of mutant and wild-type plants (about 0.5 g leaves from the same stage of different plants) growing under normal conditions for 3 weeks were detached and weighed immediately on a piece of weighting paper, and then placed on a laboratory bench (40% RH) and weighed at designated times, with three replicates. The percentage loss of fresh weight was calculated on the basis of the initial weight of the plants.
Analysis of abo3 and abi1-1, abi2-1, abi3-1, abi4-1 and abi5 double mutants
Total genomic DNA was extracted from four rosette leaves. The oligonucleotide primers used for amplification of the ABIs gene fragments were as follows: ABI1dcapsF, 5′-GATATCTCCGCCGGAGAT-3′; ABI1dcapsR, 5′-CCATTCCACTGMTCACTTT-3′; ABI2dcapsF, 5′-CATCATCTGCTATGGCAGG-3′; ABI2dcapsR, 5′-CCGGAGCATGAGCCACAG-3′; ABI3dcapsF, 5′-CGGTTTCTCTTGCAGAAAGTCTTGAAGCAAGTC-3′; ABI3dcapsR, 5′-TTGCCTCTAGCTCCGGCAAGT-3′; ABI4dcapsF, 5′-ATGGACCCTTTAGCTTCCCAACATC-3′; ABI4dcapsR, 5′-AGTTACCGGAACATCAGTGAGCTCG-3′; ABI5dcapsF, 5′-AGCTGAACAGGGACAAGTAACTGAAGTTTG-3′; ABI5dcapsR, 5′-CTCTGACGTCAACTTCGTTTCTCTAGTTACCATTTAT-3′. PCR reactions were performed in a 20-μl reaction system, containing 10 ng of genomic DNA, 0.5 μm of each primer, and 1 U of Taq polymerase. Amplification conditions were 94°C for 4 min, followed by 35 cycles consisting of 94°C for 15 sec, 55°C for 30 sec and 72°C for 60 sec. PCR products were digested by different restrictive enzymes and compared between wild-type and different abi mutants. PCR products for abi1-1 and abi2-1 loci were digested by NcoI, as described by Yin et al. (2009). The G→A point mutation in abi3-1 occurs 2143 bp upstream of the start codon (Parcy et al., 1997), and PCR products were digested by SalI. In abi4-1, a G-deletion mutation was found at 619 bp that causes the early termination of translation (Finkelstein et al., 1998), and PCR products were digested by NIaIV. A G→A point mutation occurred in the fifth nucleotide before the first putative ATG in the abi5 mutant, and PCR products were digested by MseI. The primer sequences of ABI genes for real-time RT-PCR are available upon request.
Protein targeting and histochemical analysis
The recombinant ABO3-GFP fused plasmid was introduced into onion epidermal cells by particle bombardment. GFP analysis was performed as described by Gong et al. (2002). Hygromycin-resistant, ABO3 promoter–GUS transgenic Arabidopsis seedlings and plant parts (T1 generation) were stained in GUS assay buffer (3 mm 5-bromo-4-chloro-3-indolyl-β-d-glucuronic acid, 0.1 m Na-phosphate, pH 7, 0.1% Triton X-100, 8 mmβ-mercaptoethanol) for 12 h at 37°C, followed by incubation in 70% ethanol to remove chlorophyll.
Semiquantitative and quantitative RT-PCR and RNA gel blot analysis
RNA was isolated from 200-mg tissue samples with Trizol solution (Invitrogen). First-strand cDNA was synthesized by using 5 μg of total RNA and Moloney murine leukemia virus reverse transcriptase (Promega, M5101). cDNA was used to amplify ABO3 by PCR; Actin was used as the internal standard. Cycling conditions were 5 min at 94°C and 20 cycles of 30 sec at 94°C, 30 sec at 58°C and 40 sec at 72°C. The volume of each cDNA pool was adjusted to give the same exponential-phase PCR signal strength for Actin after 20 cycles. The RT-PCR product was analyzed by electrophoresis on a 1.5% agarose gel. All PCRs were performed in triplicate.
The expression of the ABO3 gene in ABA induction was analyzed by two-step real-time quantitative RT-PCR with use of the fluorescent intercalating dye SYBR-Green in a LightCycler detection system (Bio-Rad, Chromo 4). β-actin was used as a standard control. First, total RNA samples (2 μg per reaction) from 10-day-old seedlings were reversably transcribed into cDNAs by AMV reverse transcriptase, following the manufacturer’s instructions (TAKARA, RR019 v.0609). Then, the cDNAs were used as templates in real-time PCR reactions with the gene-specific primers ABO3RTF, 5′-CTGTGGCAGCACTCCTTCATGG-3′, and ABO3RTR, 5′-CACAAGACCTGCCATGTCTCGAG-3′. The amplification of the target gene was monitored every cycle by SYBR-Green fluorescence. The cycle threshold Ct, defined as the PCR cycle at which a statistically significant increase of reporter fluorescence is first detected, was used as a measure for the starting copy numbers of the target gene. Relative quantitation of the target ABO3 expression level involved the comparative Ct method (LightCycler; Bio-Rad). The relative expression level of the ABO3 gene was calculated by the equation Y = 10Ct/3 × 100% (Ct is the difference in Ct between the control β-actin products and the target ABO3 products, i.e. Ct = CtABO3−Ctβ-actin). PCR products were confirmed on an agarose gel, and the RT-PCR data were normalized to the relative efficiency of each primer pair.
Seedlings grown on MS medium for 10 days were transferred to a solution containing 20 or 100 μm ABA, or no ABA (for control), for the described times. Total RNA (20 μg) was isolated and analyzed as previously described (Gong et al., 2002). The fragments of RD29A, COR47, RD22, ABI1, KIN1, DREB2A, ABF2 or ABF3 were 32P-labeled and used as probes for a northern blot (Chen et al., 2006). rRNA or tubulin was used as a loading control.
Transcriptional activation analysis
For transcriptional activation analysis, the full-length or partial sequence of ABO3 cDNA was amplified from the ABO3-GFP plasmid fused in-frame with the GAL4 DNA binding domain in the pGBKT7 vector (Clontech), using the following primers: ABO3YF, 5′-CGGGGATCCCGATGTTTTCAAACATCGATCACAAGGCTG-3′; ABO3YNR, 5′-CGGCTGCAGGAAGCCGTCATCAAGGCGG-3′; ABO3YWR, 5′-CGGCTGCAGCATGCACCCCAAAGGCTTTACATG-3′; ABO3YCF, 5′-GCGAATTCGCATGTAAAGCCTTTGGGGTGCATG-3′; and ABO3YR, 5′-CGGCTGCAGTCAAAACAA CATCAGGTCTTCCGATG-3′. The constructs were transformed into yeast strain AH109 containing the His3 and LacZ reporter genes. The transformed yeast cells were grown on synthetic defined (SD) plates, with or without His, and were subjected to β-galactosidase assay. The Beyotime kit (RG0036) was used for the β-galactosidase quantitative assay.
Purification of recombinant protein and electrophoretic mobility shift assay (EMSA)
Recombinant and fusion GST-ABO3 plasmids were transferred to the E. coli BL21 cell line. Under 16°C and 0.2 mm isopropyl β-d-1-thiogalactopyranoside (IPTG) for overnight incubation, the recombinant proteins were induced and purified by GST-agarose affinity.
The EMSA was performed essentially as described by Promega (http://www.promega.com/multimedia/tnt02.htm). The DNA protein binding reaction involved the incubation of 0.02 pmol of [γ-32P]ATP-labeled synthesized DNA fragments APN1 and APN2 with 1 μg of purified recombinant GST-ABO3 in a total volume of 20 μl. For mutated DNA fragments, the 5′-TTGACC/T-3′ motifs of W-boxes were substituted by 5′-TTAAA-3′. The reaction products were analyzed on 5% non-denaturing polyacrylamide gels. The specificity of the DNA binding protein for the putative binding sites was established by using either the described fold excess of unlabeled or mutated [γ-32P]ATP-labeled DNA fragments.
We thank the Arabidopsis Biological Resource Center (http://abrc.osu.edu) for providing T-DNA lines. This work was supported by the National Transgenic Research Project (2008ZX08009-002) to ZG, the National Nature Science Foundation of China (90717004, 30721062) to ZG, and (30670182) to ZC, and the Program of Introducing Talents of Discipline to Universities (B06003) to ZG.