Functional analysis of an Arabidopsis transcription factor, DREB2A, involved in drought-responsive gene expression

Abstract

Transcription factors DREB1A/CBF3 and DREB2A specifically interact with cis-acting dehydration-responsive element/C-repeat (DRE/CRT) involved in cold and drought stress-responsive gene expression in Arabidopsis thaliana. Intact DREB2A expression does not activate downstream genes under normal growth conditions, suggesting that DREB2A requires posttranslational modification for activation, but the activation mechanism has not been clarified. DREB2A domain analysis using Arabidopsis protoplasts identified a transcriptional activation domain between residues 254 and 335, and deletion of a region between residues 136 and 165 transforms DREB2A to a constitutive active form. Overexpression of constitutive active DREB2A resulted in significant drought stress tolerance but only slight freezing tolerance in transgenic Arabidopsis plants. Microarray and RNA gel blot analyses revealed that DREB2A regulates expression of many water stress-inducible genes. However, some genes downstream of DREB2A are not downstream of DREB1A, which also recognizes DRE/CRT but functions in cold stress-responsive gene expression. Synthetic green fluorescent protein gave a strong signal in the nucleus under unstressed control conditions when fused to constitutive active DREB2A but only a weak signal when fused to full-length DREB2A. The region between DREB2A residues 136 and 165 plays a role in the stability of this protein in the nucleus, which is important for protein activation.

Figure 1.

Domain Analysis of the C-Terminal Region of the DREB2A Protein in Protoplasts Prepared from Arabidopsis T87 Cells. (A) Schematic diagram of the reporter and effector constructs used in cotransfection experiments. The reporter construct contained 75-bp fragments of the RD29A promoter tandemly repeated three times (DRE×3), −61 RD29A minimal TATA promoter, and GUS reporter gene. The effector constructs contain the CaMV 35S promoter and the tobacco mosaic virus Ω sequence (Gallie et al., 1987) fused to the DREB2A cDNA with or without C-terminal deletion. Nos-T indicates the polyadenylation signal of the gene for nopaline synthetase. (B) Transactivation of the RD29A promoter–GUS fusion gene by DREB1A, DREB2A FL, or C-terminal-region deletion mutants of DREB2A. Deleted regions of the DREB2A mutants are indicated as numbers of amino acid residues. Striped boxes indicate the AP2/ERF DNA binding domain. To normalize for transfection efficiency, the CaMV 35S promoter–luciferase plasmid was cotransfected in each experiment. Bars indicate the standard error of six replicates. Ratios indicate the degree of expression compared with the value obtained with the empty effector plasmid.

Figure 2.

Transcriptional Activation with the C-Terminal Region of DREB2A Fused to the GAL4 Binding Domain. (A) Schematic diagrams of the reporter and effector constructs. The GUS reporter plasmid contains nine copies of the GAL4 binding site fused to the CaMV 35S minimal TATA box. The effector plasmids encode the GAL4 DNA binding domain (BD; amino acids 1 to 147) alone or the GAL4 binding domain fused to either the GAL4 activation region (amino acids 768 to 881) or C-terminal regions of DREB2A (white bar). (B) Transactivation of the GAL4 binding sites–GUS fusion gene by the fusion proteins of the GAL4 DNA binding domain with the GAL4 activation domain (AD) or C-terminal regions of the DREB2A protein. Striped boxes indicate the GAL4 DNA binding domain. To normalize for transfection efficiency, the CaMV 35S promoter–luciferase plasmid was cotransfected in each experiment. Bars indicate the standard error of six replicates. Ratios indicate the degree of expression compared with the value obtained from the plasmid that encodes only the GAL4 DNA binding domain. (C) Schematic diagrams of the DREB2A proteins. Numerals indicate the position of each domain in amino acid numbers from the N terminus. NLS, nuclear localization signal; ERF/AP2, ERF/AP2 DNA binding domain; NRD, negative regulatory domain; AD, activation domain.

Figure 3.

Effects of Overexpressing DREB2A CA in Transgenic Plants under Unstressed Conditions. (A) The 25-d-old seedlings carrying DREB2A CA driven by the 35S promoter (35S:DREB2A CA-a, -b, and -c), the DREB2A FL construct driven by the 35S promoter (35S:DREB2A FL), or pBI121 (wt). (B) Inflorescence heights of 30-d-old transgenic plants. Average inflorescence heights and standard deviations were calculated using 12 plants. Asterisks indicate that these plants have significantly shorter inflorescences than wild-type plants (Student's t test, P < 0.001). (C) Photographs of 35-d-old transgenic plants (D) Close-up of the wild-type and 35S:DREB2A CA plants shown in (C). (E) RNA gel blot analysis of DREB2A and RD29A in wild-type and transgenic plants. Each lane was loaded with 5 μg of total RNA prepared from 3-week-old transgenic Arabidopsis plants. Ethidium bromide–stained rRNA image is shown as a loading control.

Figure 4.

Expression of the DREB1A- and DREB2A-Upregulated Genes in the Plants Carrying pBI121, 35S:DREB2A CA, or 35S:DREB1A Constructs. RNAs were prepared from transgenic Arabidopsis plants that had been dehydrated for 5 h (dry), treated at 4°C for 5 h (cold), or untreated (control). The full-length cDNA was used as probe for each gene except for DREB2A, RD29A, and RD29B; the 3′-terminal–specific DNA fragments were used as probes for these three genes. (A) Expression of the transgenes. (B) Genes that showed higher expression in 35S:DREB2A plants. (C) Genes that showed equivalent expression in 35S:DREB1A plants and 35S:DREB2A CA plants. (D) Genes that showed higher expression in 35S:DREB1A plants. (E) Ethidium bromide–stained rRNA image is shown as a loading control.

Figure 5.

Promoter Analysis of the DREB1A- and DREB2A-Upregulated Genes. (A) and (B) Sequence logo for the DRE core sequence (A/GCCGAC) with 20 adjacent nucleotides found in the promoter regions of DREB1A-upregulated genes (At2g02100, COR414-TM1, COR15A, COR15B, KIN1, and KIN2) and DREB2A-upregulated genes (RD29B, At1g52690, MT2A, At1g32860, At1g69870, PDC2, At3g53990, ATGRP7, and At1g22985). The asterisks indicate nucleotide positions analyzed by gel mobility shift assay. (C) Probe and competitor sequences used in the gel mobility shift assay. Sequences of the 75-bp fragment of the wild-type RD29A promoter (ACCGACAT, underlined) and the mutants that were used as competitors. The red letters indicate nucleotides mentioned in the text. The wild-type fragment was used also as a probe. (D) and (E) Competitive DNA binding assay of recombinant DREB1A and DREB2A proteins using A/C/G/TCCGACAT or ACCGACAA/C/G/T as competitors. Glutathione S-transferase (GST) or recombinant protein solutions were preincubated with or without unlabeled competitors for 5 min at 25°C. ³²P-labeled probe was then added, and the mixture was incubated for 30 min at 25°C. As competitors, 10×, 100×, or 1000× excess amounts of the unlabeled fragments were used. The retarded bands are indicated by arrowheads.

Figure 6.

Drought and Freezing Tolerance of the 35S:DREB2A CA and 35S:DREB1A Plants. The stress treatments were conducted as described in the text. Drought: water withheld from plants for 2 weeks; freezing: 4-week-old plants exposed to −6°C for 30 h and returned to 22°C for 5 d. (A) Photographs of plants before and after stress treatments. (B) Survival rates of plants exposed to drought and freezing stress. Average survival rates and standard errors were calculated using results of three replicated experiments. Twenty plants (five plants/pot) were tested in each experiment. In all experiments, the plants with asterisks had significantly higher survival rates than wild-type plants (χ² test, *P < 0.05, **P < 0.01).

Figure 7.

Phenotypes and Drought Stress Tolerance of the RD29A:DREB2A CA Transgenic Plants. (A) Gene expression profiles of DREB2A and RD29B in the wild-type plants, 35S:DREB2A CA-b plants (35S:-b), and RD29A:DREB2A CA plants (RD29A:-a and RD29A:-b). RNAs were isolated from unstressed control plants or plants that were treated for 30 min under drought condition. Accumulation of the DREB2A and RD29B mRNAs was measured by quantitative RT-PCR. Data represent means and standard errors of three replications. (B) Photographs of vector control plants and transgenic plants overexpressing DREB2A CA. The plants were grown on germination medium agar plates for 3 weeks, then transferred to soil and grown for 2 weeks. (C) Inflorescence heights of 5-week-old transgenic plants. Average inflorescence heights and standard deviations were calculated using 12 plants. Asterisk indicates that these plants had significantly shorter inflorescences than wild-type plants (Student's t test, P < 0.001). (D) Seeds were harvested from 3-month-old vector control plants, 35S:DREB2A CA-b plants, or RD29A:DREB2A CA plants, and air-dry seed was weighed. The average yield of each line was calculated from yields of 20 plants. Error bars show standard deviation. Asterisk indicates that these plants had significantly lower seed yields than wild-type plants (Student's t test, P < 0.001). (E) Photographs of plants before and after stress treatment. Stress treatment was performed as described in Figure 6. (F) Survival rates of plants exposed to drought stress. Average survival rates and standard errors were calculated using results of three replicated experiments. Twenty plants (five plants/pot) were tested in each experiment. Asterisk indicates that in all experiments, these plants had significantly higher survival rates than wild-type plants (χ² test, P < 0.001).

Figure 8.

Subcellular Accumulation Patterns of DREB2A FL and DREB2A CA Fused to sGFP. (A) Structure of the chimeric genes with sGFP fused to DREB2A FL (35S:GFP-DREB2A FL) or DREB2A CA (35S:GFP-DREB2A CA). (B) and (C) Expression of DREB2A and RD29A in the transgenic Arabidopsis plants carrying 35S:GFP, 35S:GFP-DREB2A FL, or 35S:GFP-DREB2A CA. Data represent means and standard errors of three replications. (D) Confocal microscope images of sGFP fluorescence (1, 3, 5, 7, and 9) and Nomarski microscope images (2, 4, 6, 8, and 10) of young roots from transgenic Arabidopsis plants carrying 35S:GFP (1 and 2), 35S:GFP-DREB2A FL (3 to 6), or 35S:GFP-DREB2A CA (7 to 10). Bars = 10 μm. (E) Relative intensities of green fluorescent signals in the nuclei of transgenic Arabidopsis plants carrying 35S:GFP-DREB2A FL or 35S:GFP-DREB2A CA. Average signal intensity was calculated from signals of >40 nuclei observed in three fields of view per line. Error bars represent standard deviation. The 35S:GFP-DREB2A CA plants gave significantly stronger green fluorescent signals than 35S:DREB2A FL plants (Student's t test, P < 0.001).

Figure 9.

Model of the Induction of Genes Regulated by DREB1A and DREB2A under Drought, High-Salinity, and Cold Stress Conditions. The genes downstream of the DREB proteins are categorized into three groups. The middle group contains downstream genes shared by DREB1A and DREB2A. The other groups consist of DREB1A- and DREB2A-specific downstream genes.