Introduction

Unlike animals, plants cannot move away from disadvantageous environmental conditions such as cold, drought, and high salinity. Cold exposure (low temperature or chilling) represents one of the major environmental stresses that adversely affect rice productivity through poor germination, yellowing leaves, growth retardation, and improper microspore development (Andaya and Mackill 2003; Baruah et al. 2009; Shimono et al. 2011a; de Los Reyes et al. 2013). Therefore, screening for cold tolerance genes is particularly important in light of the current increase in occasional extreme weather events.

To cope with diverse environmental stresses plants regulate the expression of a large number of stress-associated genes. The gene products include water channel proteins, the enzymes required for the biosynthesis of osmoprotectants, lipid desaturases for membrane modification, protective proteins, such as antifreeze proteins and chaperones, detoxification proteins, and transcription factors (Shinozaki and Yamaguchi-Shinozaki 1996; Singh et al. 2002). In particular, transcription factors play an essential role in the regulation of the plant response to abiotic stresses through the binding of transcription factors to cognate cis-acting elements present in the promoter region of their target genes (Casaretto and Ho 2003; Fujita et al. 2005; Yamaguchi-Shinozaki and Shinozaki 2006; Hu et al. 2008; Zou et al. 2010; Yang et al. 2012; Peng et al. 2013). One class of proteins that is unique to plants and plays a vital role in abiotic stress response is the WRKY family of proteins that bind to the consensus cis-element W box (TTGACT/C) (Eulgem et al. 2000; Cai et al. 2008; Ciolkowski et al. 2008; Rushton et al. 2010). Previous studies have identified a large number of WRKYs, including 74 WRKY-encoding genes in Arabidopsis (Ulker and Somssich 2004) and 125 in rice (Rice WRKY working group 2012).

Until now, the functions of WRKY proteins have been mainly analyzed with a focus on defense responses upon pathogen infection. For instance, the Arabidopsis WRKYs AtWRKY18, AtWRKY40, and AtWRKY60 are negative regulators of resistance to the hemibiotrophic bacterial pathogen Pseudomonas syringae (Xu et al. 2006) and the barley WRKYs HvWRKY1 and HvWRKY2 are negative regulators of the basal defense response against the fungal pathogen Blumeria graminis (Shen et al. 2007). The rice WRKY OsWRKY45 enhances resistance to Magnaporthe oryzae and Xanthomonas oryzae pathovar oryzae (Xoo) (Shimono et al. 2007; Jiang et al. 2010; Shimono et al. 2011b; Matsushita et al. 2012). Overexpression of OsWRKY62 and OsWRKY76 compromises XA21-mediated resistance to Xoo (Peng et al. 2008; Seo et al. 2011) and overexpression of OsWRKY28 enhances rice susceptibility to M. oryzae (Chujo et al. 2013). Consistent with these findings, reduced expression of these rice WRKYs enhances resistance to Xoo and M. oryzae (Peng et al. 2010; Delteil et al. 2012).

Notably, a number of WRKY genes are involved in the coordination of multiple biological processes. For instance, AtWRKY33 regulates disease resistance, NaCl tolerance and thermotolerance (Birkenbihl et al. 2012; Jiang and Deyholos 2009; Li et al. 2011), and a pepper WRKY, CaWRKY40, modulates resistance to Ralstonia solanacearum and tolerance to heat stress (Dang et al. 2012). These findings suggest that WRKYs can serve as nodes in the crosstalk between different physiological processes; however, the functions of the majority of WRKY family members and their possible roles in signaling crosstalk remain poorly understood.

Rice OsWRKY71 is upregulated by several defense-signaling molecules, such as methyl jasmonate, salicylic acid and 1-aminocyclo-propane-1-carboxylic acid, as well as by biotic elicitors, wounding and pathogen infection. Accordingly, OsWRKY71 overexpression enhances resistance to Xoo (Liu et al. 2007; Chujo et al. 2008). On the other hand, OsWRKY71 acts as a transcriptional repressor of gibberellin-responsive genes (Zhang et al. 2004). In the present study, we show that OsWRKY71 expression is induced by cold stress. In an attempt to investigate the in vivo function of OsWRKY71, we analyzed the responses of OsWRKY71 overexpression lines upon cold treatment. Our data indicate that OsWRKY71 may have an important role in cold tolerance in rice.

Materials and methods

Plant materials and exposure of rice seedlings to chilling stress

Rice (Oryza sativa japonica cultivar [cv.] Dongjin) plants were grown in a greenhouse under a 14/10-h light/dark period, at a temperature of 24–28 °C and 70–80 % humidity. To investigate the expression of OsWRKY71 in rice, seedlings and mature leaves and panicles of flowering plants were subjected to cold (4 °C), abscisic acid (ABA, 100 µM), salt (300 mM NaCl), and drought (20 % polyethylene glycol and air-drying) treatments. To analyze the response to cold stress, homozygous transgenic plants of the T2 generation were used. Five-week-old seedlings were treated at 4 °C in a low-temperature environment room, and thereafter moved to the greenhouse for 6 days. Surviving seedlings were photographed and characterized to investigate the response of OsWRKY71 overexpressing transgenic plants (OsWRKY71-OX) to chilling stress. To analyze the expression of cold-responsive genes in seedlings in response to cold stress, 5-week-old seedlings of transgenic plants of T2 generation were exposed to 4 °C in the low-temperature environment room. At 0, 12, and 24 h of chilling treatment, leaf blades of cold-stressed seedlings were sampled for analysis.

RNA gel blot and RT-PCR analysis

Total RNA was prepared from various rice tissues after different abiotic stress treatments using Trizol reagent (Invitrogen, Carlsbad, CA, USA). For RNA gel blot analysis, 20 μg of isolated total RNA was fractionated on a 1.3 % agarose gel and transferred onto Hybond-N+ nylon membranes (Amersham Biosciences, Pittsburgh, PA, USA). Hybridization was then carried out with [α-32P] dCTP-labeled gene-specific probe according to standard procedures under high-stringency hybridization conditions (Cho et al. 2006). Gene-specific probes for OsWRKY71 (LOC_Os02g08440) and SalT (LOC_Os01g24710) were amplified using the primers listed in Supplementary Table S1. PCR amplification was performed in a final volume of 40 µL (100 pmol of each primer, 20 µM each of dNTPs, 10 mM Tris–HCl pH 9.0, 2 mM MgCl2, 50 mM KCl, 0.1 % Triton X-100, and 0.5 U of Taq polymerase) using 50 ng of genomic DNA as template.

For RT-PCR analysis, total RNA isolated from wild type (WT) and transgenic rice plants was reverse transcribed with oligo-dT primer and a first-strand cDNA synthesis kit (Roche, Mannheim, Germany). The synthesized first-strand cDNA was used in subsequent PCR reactions as described above with gene-specific primers and control primers for OsUBQ5 (Jain et al. 2006; Supplementary Table S1). The cold-responsive genes analyzed included OsDREB1A (LOC_Os09g35030), Glutamate decarboxylase (LOC_Os03g13300), TPP1 (Trehalose-6-phosphate phosphatase1; LOC_Os02g44230), OsTGFR (LOC_Os01g68510), WSI76 (LOC_Os07g48830), OsCORTM1 (Cold-regulated 413 thylakoid membrane1; LOC_Os05g49170) and OsMAT1 (Methionine adenosyltransferase1; LOC_Os05g04510).

Analysis of OsWRKY71 promoter:GFP expression

The 1943-bp promoter region was isolated from rice genomic DNA by PCR using pOsWRKY71-F/R primers (Supplementary Table S1). After adapter-conjugating PCR with attB1 and attB2 adaptors, the isolated promoter fragment was linked to the Green fluorescent protein (GFP) gene by the Gateway LR clonase (Invitrogen). The transformation vector contained the bar gene between the CaMV35S promoter and the terminator of nopaline synthase gene (Park et al. 2010). Through Agrobacterium-mediated rice transformation and propagation of the transgenic lines, several independent lines of the T2 generation were obtained. Among them, one representative transgenic line was selected for further study.

Total RNA was extracted from 1-month-old leaves that were subjected to air-drying (drought), low temperature (4 °C) or high salinity (300 mM NaCl) for 0 to 6 h, and reverse transcribed with an oligo-dT primer and a first-strand cDNA synthesis kit (Roche). Real-time qRT-PCR was carried out with 50 ng of cDNA template using the GFP-specific primers GFP-F/R. Fluorescence from the amplicon was detected by a Stratagene Mx3000p Real-Time PCR machine and Mx3005P software v2.02 (Stratagene, USA). All results were measured from triplicate reactions for each sample. To normalize the real-time qRT-PCR data, the OsUBI1 (LOC_Os06g46770) gene primers OsUbi1-F/R were used as a reference.

Subcellular localization of OsWRKY71-GFP

Full-length cDNA of OsWRKY71 was amplified by PCR using full-length F/R primers (Supplementary Table S1) and the PCR product was cloned into the pENTR™/D-TOPO vector (Invitrogen). The insert was confirmed by sequencing and then subcloned into the destination vector p2FGW7 for N-terminal GFP fusion (Karimi et al. 2002) using the Gateway LR clonase (Invitrogen). The resulting construct was introduced into maize mesophyll protoplasts using PEG-calcium mediated transformation method (Cho et al. 2009). After incubation for 24 h, expression of the GFP fusion protein was monitored using a confocal microscope (LSM510 META, Carl Zeiss, Germany). GFP fluorescence was excited at 488 nm and detected in the emission range of 500–525 nm. OsABF1-RFP was used as a nuclear marker (Amir Hossain et al. 2010).

Assay for transcriptional activity of OsWRKY71

For the transient expression assay, an effector vector was constructed by fusing the full-length cDNA of OsWRKY71, amplified by PCR using primers containing NdeI and EcoRI sites (Supplementary Table S1), to the GAL4 binding domain (BD) sequence under the control of the CaMV35S promoter and the Ω sequence. The previously described GAL4-responsive reporter construct (Ohta et al. 2000) was used as a reporter vector, and contains 5X GAL4, a minimal TATA, the Ω sequence and the Luciferase (LUC) gene. The maize ubiquitin promoter:β-Glucuronidase (GUS) construct (ZmUBQ:GUS) was used as an internal control (Han et al. 2013). Maize mesophyll protoplasts isolated from the second leaves of dark grown plants were transfected with the effector, reporter and internal control vectors as described previously (Han et al. 2013). Transfected protoplasts were incubated in W5 solution for 12 h (Zhang et al. 2011). LUC and GUS activity assays were performed as previously described (Han et al. 2013). All experiments were repeated three times with similar results.

Production of OsWRKY71 overexpressing transgenic plants

Full-length cDNA of OsWRKY71 was amplified by PCR using full-length F/R primers (Supplementary Table S1), cloned into pENTR™/D-TOPO vectors (Invitrogen), and then subcloned into the binary plant vector pH2GW7 (Karimi et al. 2002) using the gateway LR clonase (Invitrogen). To produce transgenic rice plants expressing CaMV35S:OsWRKY71, the Agrobacterium tumefaciens LBA4404 strain harboring this vector was grown on AB media supplemented with 10 mg L−1 streptomycin and 50 mg L−1 hygromycin for 3 days at 28 °C, and transgenic calli were obtained via the Agrobacterium-mediated co-cultivation method as described previously (Jeon et al. 2000). OsWRKY71-OX transgenic plants were selected on medium containing 50 mg L−1 hygromycin and 250 mg L−1 cefotaxime. T2 homozygous lines were selected by testing for hygromycin resistance of progeny plants and analyzed for further characterization.

Phenotype analysis of OsWRKY71-OX lines in response to cold stress

To examine the cold tolerance of OsWRKY71-OX lines, we analyzed both the chlorophyll fluorescence and survival ratio of transgenic plants (Lee et al. 2004). In brief, the youngest leaves of 10-day-old seedlings were cut and floated on distilled water at 4 °C in a cold chamber under continuous light of 110 μmol m−2 s−1 for up to 40 h and the chlorophyll fluorescence was measured using the Plant Efficiency Analyzer (Hansatech, Kings Lynn, UK). The ratio Fv/Fm was calculated to examine damage in the photosystem (Genty et al. 1989). For survival analysis, 10-day-old seedlings were treated in the cold chamber for up to 50 h. After recovery for 6 days under normal growth conditions, the third-leaf survived plants were counted and their fresh weights were measured. Dry weights were measured after completely drying at 80 °C for 2 days.

Results

OsWRKY71 expression in response to cold treatment

Analysis of publicly available microarray data revealed that expression of OsWRKY71 is greatly increased when rice plants are challenged by cold stress (Supplementary Fig. S1). This prompted us to further investigate the expression of OsWRKY71 in rice in response to cold and other environmental stresses by RNA blot analysis (Fig. 1). OsWRKY71 expression was examined by subjecting rice seedlings and panicles of flowering plants to cold (4 °C), ABA (100 µM), salt (300 mM NaCl), and drought (20 % PEG and air-drying) treatments. OsWRKY71 expression was rapidly increased in cold-treated seedlings as early as 2 h, but was barely detectable in untreated samples (Fig. 1a). Notably, none of the ABA, salt, and drought treatments increased OsWRKY71 expression relative to that of untreated leaves, even 8 h after each treatment (Fig. 1b), indicating its unique responsiveness to cold treatment. The cold-responsive expression of OsWRKY71 was consistently observed in the mature leaf and panicle after 3 days (I) and 7 days (II) exposure to cold stress (Fig. 1c). In contrast, the gene expression of SalT was upregulated in response to salt and drought, as previously reported (Claes et al. 1990).

Fig. 1
figure 1

Analysis of cold-responsive expression of OsWRKY71. a RNA blot analysis of the time course of OsWRKY71 expression under control or cold-treated conditions. Two-week-old seedlings were transferred to the 4 °C growth chamber for 0, 2, 4, 8, 12, and 24 h of cold treatment and then sampled for RNA blot analysis. Untreated seedlings were sampled at the same time as controls. rRNAs are shown as a loading control. b RNA blot analysis of OsWRKY71 under control, ABA (100 µM), salt (300 mM NaCl), cold (4 °C), and drought (air dry* and 20 % PEG**) conditions. All stresses were applied to 2-week-old seedlings for 8 h. c RNA blot analysis of OsWRKY71 in mature leaves and panicles under control, drought, salt, and cold conditions. Mature leaves and panicles a week before anthesis were sampled after 3 (I) or 7 days (II) of various stress treatments. SalT expression was included as a drought- and salt-responsive marker gene. d Promoter activity analysis of pOsWRKY71:GFP in transgenic rice. A strong GFP signal was detected only after cold (4 °C) treatment, in contrast to drought (air dry) and salt (300 mM NaCl) treatments. The fold values were represented as a graph relative to the level of GFP mRNA in the lowest-expressing transgenic plants, as indicated by an asterisk

Full size image

To further confirm that OsWRKY71 is regulated by cold at the transcriptional level, we constructed a fusion vector containing the approximately 2-kb promoter sequence of OsWRKY71 and the reporter gene GFP (pOsWRKY71:GFP) and introduced the vector into rice. pOsWRKY71:GFP transgenic rice plants were treated with cold, drought and salt. GFP expression analysis indicated that OsWRKY71 promoter activity was dramatically increased after only 2 h of cold treatment (Fig. 1d).

Determination of subcellular localization of OsWRKY71

To determine the subcellular localization of OsWRKY71, an OsWRKY71-GFP fusion construct under the control of the CaMV35S promoter was generated. In a transient expression assay with maize mesophyll protoplasts, OsWRKY71-GFP was detected mostly in the nucleus and to a lesser extent in the cytosol, as confirmed by the nuclear marker construct, OsABF1-RFP (Fig. 2a). Therefore, we concluded that the OsWRKY71 protein might be targeted to the nucleus.

Fig. 2
figure 2

Subcellular localization of OsWRKY71 and transcriptional repression assay. a Expression of OsWRKY71-GFP fusion protein in maize protoplast cells. The OsWRKY71-GFP signal is green; OsABF1-RFP is used as a nuclear marker. Bright light field and merged images are also shown. b Structures of the reporter and effector vectors used for the transcriptional activity assay of OsWRKY71. ZmUBQ:GUS was used as an internal control. Nos, nopaline synthase terminator. c Luciferase activity assay. The effector vector, either empty vector (EV) or BD-OsWRKY71 (OsWRKY71), was co-infected with the reporter LUC vector and the internal control vector ZmUBQ:GUS. The fold change was calculated as the LUC/GUS activity ratio. The values are the averages with standard deviation of the results from three independent experiments. Normalized LUC activity recorded after transfection with the empty vector control alone was arbitrarily set at 1. *P < 0.05

Full size image

Transcriptional repression ability of OsWRKY71

Next, we examined the transcriptional activation ability of OsWRKY71 using the maize protoplast transient expression system (Han et al. 2013). The effector constructs for this system contain the CaMV35S promoter, the TMV translation enhancer Ω sequence, and either the GAL4 DNA BD or the OsWRKY71 cDNA insert fused to the BD (Fig. 2b). As a reporter, we used a GAL4-responsive reporter construct (Ohta et al. 2000) that contains five copies of the GAL4 binding site in tandem and a minimal TATA region of the CaMV35S promoter, the Ω sequence, and the firefly LUC gene (Fig. 2b). In the maize protoplasts, the reporter LUC activity was considerably reduced when co-infected with the effector BD-OsWRKY71 fusion vector compared with the empty vector control (Fig. 2c). This result suggests that OsWRKY71 may function as a transcriptional repressor to regulate downstream target genes.

Characterization of transgenic rice plants overexpressing OsWRKY71

To characterize the function of OsWRKY71 in rice, transgenic rice plants constitutively overexpressing OsWRKY71 under the control of the CaMV35S promoter were generated using Agrobacterium-mediated transformation. The transcript level of OsWRKY71 was much higher in two independent transgenic T2 homozygous lines, OsWRKY71-OX12 (OX12) and OsWRKY71-OX21 (OX21), than in the transformation background WT (cv. Dongjin) and the segregated non-transgenic line (NT) (Fig. 3a). There was no apparent difference in agricultural phenotypic characteristics such as plant height and yield potential among these transgenic lines (data not shown).

Fig. 3
figure 3

Production and characterization of OsWRKY71-OX transgenic rice plants. a RT-PCR analysis of OsWRKY71 expression in OsWRKY71-OX12 (OX12) and OsWRKY71-OX21 (OX21) transgenic lines. OsUBQ5 was amplified as an internal RT-PCR control. b OsWRKY71 overexpression confers cold tolerance in rice. Five-week-old rice seedlings of wild type (WT), segregant non-transgenic (NT), and OsWRKY71-OX transgenic lines (OX12 and OX21) were exposed to cold stress (4 °C) for 50 h and then allowed to recover for 6 days under normal conditions

Full size image

To investigate whether OsWRKY71 overexpression correlated with cold tolerance in rice, two 5-week-old control lines, WT and NT, and two OsWRKY71-OX lines, OX12 and OX21, were exposed to cold stress (4 °C) for 50 h and then allowed to recover their growth for 6 days under normal growth conditions. We found that OX12 and OX21 transgenic plants recovered much better, maintaining a greater number of green leaves, than WT and NT control plants, which displayed severely dried leaves (Fig. 3b).

Cold stress tolerance was examined by measuring the third-leaf survival ratio of 10-day-old seedlings. Most plants of the OsWRKY71-OX lines OX12 and OX21, 94 and 97 %, respectively, survived after cold stress treatment followed by recovery. In contrast, only 54 and 75 % of the control WT and NT plants, respectively, survived (Table 1). Our results clearly indicate that OsWRKY71 overexpression in rice plants increases tolerance to low-temperature stress.

Table 1 Cold tolerance of OsWRKY71-OX transgenic rice judged by the third-leaf survival ratio
Full size table

The elevated tolerance of OsWRKY71-OX transgenic rice plants to cold stress was further evaluated by measuring chlorophyll fluorescence in two control and two transgenic lines that were subjected to low temperature (4 °C) for 20 or 40 h. The ratio F v/F m between the variable fluorescence (F v) and the maximum fluorescence (F m) is used to estimate the quantum yield of PSII photochemistry (Krause and Weis 1991). Under normal growth conditions (0 h), all tested plants showed an F v/F m of approximately 0.85, which is close to the value of 0.8 observed in all plant species studied (Fig. 4a). When exposed to cold stress, the control plants showed a significant reduction in the F v/F m values: 0.57 and 0.58 after 20 h and 0.2 and 0.18 after 40 h in WT and NT, respectively. In contrast, OsWRKY71-OX lines maintained relatively high F v/F m values of 0.7 and 0.71 after 20 h and 0.3 and 0.37 after 40 h for OX12 and OX21, respectively (Fig. 4a). Concordant with these findings, OX12 and OX21 plants retained higher levels of both fresh weight and dry weight than control plants (Fig. 4b, c). These results further suggest that overexpression of OsWRKY71 enhances cold tolerance in transgenic rice.

Fig. 4
figure 4

Phenotype analysis of OsWRKY71-OX transgenic rice seedlings in response to cold treatment. a Changes in chlorophyll fluorescence of wild type (WT), segregant non-transgenic (NT), and OsWRKY71-OX transgenic lines (OX12 and OX21) at 20 h and 40 h of cold stress treatment. Fresh weight (b) and dry weight (c) after 6 days of recovery. *P < 0.05

Full size image

Expression of cold-responsive genes in OsWRKY71-OX lines

To obtain clues regarding the molecular mechanism of OsWRKY71-mediated cold tolerance, we monitored the expression of previously known cold-responsive genes in OsWRKY71-OX lines by RT-PCR analysis. Transcription levels of the selected cold marker genes were examined in both control and transgenic lines under cold (4 °C) treatment for 12 or 24 h. We found remarkable upregulation of the rice DREB gene OsDREB1A (Dubouzet et al. 2003; Ito et al. 2006) in the control, WT, NT, and transgenic OX12 and OX21 plants. These plants showed no difference in the pattern of OsDREB1A expression (Fig. 5), which peaked 12 h after cold treatment. Likewise, expression of two marker genes, Glutamate decarboxylase (Su et al. 2010) and TPP1 (Pramanik and Imai 2005), reached a similar peak 12 h after cold treatment in control and transgenic lines and thereafter decreased. OsCORTM1 (OsCOR413-TM1; Ma et al. 2009) and OsMAT1 (Ma et al. 2009), as well as the PCR control gene OsUBQ5, did not respond to cold treatment in control and transgenic lines.

Fig. 5
figure 5

Expression analysis of cold-responsive marker genes in OsWRKY71-OX lines in response to cold stress. Leaves of 5-week-old seedlings exposed to 4 °C were harvested at 0, 12, and 24 h. OsUBQ5 was used as an internal control. WT wild type, NT segregant non-transgenic line, OX12 and OX21 OsWRKY71-OX transgenic rice lines

Full size image

Interestingly, two cold-responsive genes, OsTGFR and WSI76, which are the rice homologs of target genes of Arabidopsis DREBs (Dubouzet et al. 2003; Chen et al. 2008; Saito and Yoshida 2011), showed differential induction in OsWRKY71-OX plants under normal and cold stress conditions (Fig. 5). OsTGFR expression was detected in OsWRKY71-OX plants before and after cold treatment. WSI76 expression was more highly induced after cold treatment in OX12 and OX21 than in control lines. This result suggests that OsWRKY71 may have a regulatory function in the expression of OsTGFR and WSI76 in response to cold treatment.

Discussion

Among several families of transcription factors that are involved in the transcriptional regulation of plant stress-responsive genes, increasing attention has recently been paid to the role of WRKY transcription factors in regulating the response to abiotic and biotic stresses. However, relatively little progress has been made in understanding the function of WRKY proteins under abiotic stresses (Chen et al. 2012).

Here, we provide evidence that OsWRKY71 positively regulates cold tolerance in rice plants. First, we showed that OsWRKY71 was induced specifically by cold (Fig. 1). Second, we found that it functioned as a nuclear transcriptional repressor (Fig. 2). Third, OsWRKY71 overexpression enhanced cold tolerance in transgenic rice, and this elevated tolerance was further confirmed by measurement of chlorophyll fluorescence, F v/F m (Figs. 3, 4; Table 1). Finally, we identified two potential downstream targets of OsWRKY71, OsTGFR and WSI76, in response to cold treatment (Fig. 5). Together, these data suggest that OsWRKY71 may have an important transcriptional regulatory role in response to cold stress, and that the OsWRKY71-mediated pathway is likely essential for cold tolerance in rice. It is noteworthy that OsWRKY71 was not upregulated by ABA and other abiotic stresses such as salt and drought in rice seedlings (Fig. 1), implying that OsWRKY71 may function in an ABA-independent cold response pathway.

Numerous transcriptional expression data have revealed cold-responsive genes in common with those under observation in the present study that are associated with cold stress in different plant species (Fowler and Thomashow 2002; Maruyama et al. 2004; Seki et al. 2002; Dubouzet et al. 2003; Wang et al. 2004; Fowler et al. 2005; Skinner et al. 2005; Ito et al. 2006; Badawi et al. 2007; Cheng et al. 2007; Zhou et al. 2008; Mittal et al. 2012; Zhang et al. 2012). For instance, DREB1s act as upstream regulators during cold stress and are rapidly induced several fold after cold treatment. Our experimental observation of expression patterns of cold-responsive genes other than OsCORTM1 and OsMAT1, such as OsDREB1A, Glc decarboxylase, and TPP1, indicated that many of the rice homologs of cold-responsive genes are also rapidly regulated in cold-exposed rice (Fig. 5), supporting conserved regulatory mechanism(s) in a range of plant species.

In Arabidopsis there are seven members belonging to the GolS gene family, of which GolS3 is controlled by DREB1A and induced by cold stress (Taji et al. 2002). Expression of the rice homolog of GolS3, WSI76, was found to be increased under chilling stress (Saito and Yoshida 2011). In this context, it is noteworthy that we found a remarkable increase in the transcripts of WSI76 in OsWRKY71-OX transgenic rice compared with the controls WT and NT after cold treatment (Fig. 5). In Boea hygrometrica, a WRKY is involved in the regulation of GolS expression (Wang et al. 2009). OsTGFR is a homolog of target genes of Arabidopsis DREBs (Dubouzet et al. 2003; Chen et al. 2008; Ma et al. 2009; Saito and Yoshida 2011). We found an alteration in OsTGFR expression in OsWRKY71-OX transgenic rice in response to cold compared with expression in the controls. The enhancement of stress tolerance in OsWRKY71 transgenic plants likely results from changes in expression of these cold-responsive genes. In previous studies, OsWRKY28 (Chujo et al. 2013) and OsWRKY76 (Yokotani et al. 2013), closely related paralogues of OsWRKY71, exhibited transcriptional repressor activity. Consistent with these findings, we found that OsWRKY71 functioned as a transcriptional repressor (Fig. 2c). This result suggests that OsWRKY71 exerts its function through transcriptional suppression of target genes that may negatively regulate the downstream genes OsTGFR and WSI76. How the transcriptional repressor OsWRKY71 regulates downstream target genes in response to cold stress remains to be determined.

Strong constitutive overexpression of transcription factors often results in negative growth phenotypes in transgenic plants. For example, growth retardation has been observed in transgenic rice plants overexpressing stress-responsive transcription factors such as OsDREB1A, OsDREB1B, OsWRKY89 and OsNAC6 (Ito et al. 2006; Nakashima et al. 2007; Wang et al. 2007). Although constitutive expression of these transgenes produced the best result in respect to cold tolerance, the overall advantage was reduced by significant differences in agronomical traits such as flowering time and yield potentials of transgenic lines. In the present study, OsWRKY71 overexpression resulted in no obvious phenotypic alteration, and plant architecture was indistinguishable from that of WT under normal growth conditions. Given that OsWRKY71 overexpression improved cold tolerance with an unaltered growth phenotype under normal conditions, our data imply the potential application of OsWRKY71 for genetic improvement of cold tolerance in rice. To further understand the molecular mechanism of enhanced cold tolerance responses mediated by OsWRKY71, it would be useful to analyze global gene expression changes in OsWRKY71-OX in response to cold stress and to identify and characterize downstream targets in future studies.