The OsNAC6 gene is a member of the NAC transcription factor gene family in rice. Expression of OsNAC6 is induced by abiotic stresses, including cold, drought and high salinity. OsNAC6 gene expression is also induced by wounding and blast disease. A transactivation assay using a yeast system demonstrated that OsNAC6 functions as a transcriptional activator, and transient localization studies with OsNAC6–sGFP fusion protein revealed its nuclear localization. Transgenic rice plants over-expressing OsNAC6 constitutively exhibited growth retardation and low reproductive yields. These transgenic rice plants showed an improved tolerance to dehydration and high-salt stresses, and also exhibited increased tolerance to blast disease. By utilizing stress-inducible promoters, such as the OsNAC6 promoter, it is hoped that stress-inducible over-expression of OsNAC6 in rice can improve stress tolerance by suppressing the negative effects of OsNAC6 on growth under normal growth conditions. The results of microarray analysis revealed that many genes that are inducible by abiotic and biotic stresses were upregulated in rice plants over-expressing OsNAC6. A transient transactivation assay showed that OsNAC6 activates the expression of at least two genes, including a gene encoding peroxidase. Collectively, these results indicate that OsNAC6 functions as a transcriptional activator in response to abiotic and biotic stresses in plants. We conclude that OsNAC6 may serve as a useful biotechnological tool for the improvement of stress tolerance in various kinds of plants.
Genes containing a NAC domain are plant-specific transcription factors, and are expressed in various developmental stages and tissues. The NAC domain was originally characterized from consensus sequences from petunia NAM and Arabidopsis ATAF1, ATAF2 and CUC2 (Aida et al., 1997). Many NAC proteins, including CUC2 of Arabidopsis, have crucial functions in plant development. In potato, some NAC genes have been found to be upregulated by wounding and bacterial infection (Collinge and Boller, 2001; Hegedus et al., 2003; Mysore et al., 2002), and some NAC proteins have been shown to mediate viral resistance (Ren et al., 2000; Xie et al., 1999). We recently reported that some Arabidopsis NAC proteins are involved in abiotic stress responses and tolerance (Fujita et al., 2004; Tran et al., 2004). Arabidopsis RD26 cDNA, which was originally identified from dehydrated plants, encodes a NAC protein (Fujita et al., 2004). Expression of the RD26 gene was induced not only by drought but also by high salinity, abscisic acid (ABA) and methyl jasmonic acid (MeJA). Transgenic plants over-expressing RD26 were highly sensitive to ABA, while RD26-repressed plants were insensitive. The results of microarray analysis showed that ABA- and stress-inducible genes are upregulated in the RD26-over-expressing plants and repressed in the RD26-repressed plants. Furthermore, RD26 activated a promoter of its target gene encoding glyoxalase in Arabidopsis protoplasts. These results collectively indicate that RD26 functions as a transcriptional activator in ABA-inducible gene expression as a result of abiotic stress in plants.
On the other hand, using a yeast one-hybrid system, we isolated three cDNA clones encoding proteins that bind to the promoter region of the Arabidopsis gene ERD1 (early responsive to dehydration 1), which contains the CATGTG motif (Nakashima et al., 1997; Simpson et al., 2003; Tran et al., 2004). These cDNA clones encode proteins named ANAC019 (At1 g52890), ANAC055 (At3 g15500) and ANAC072 (At4 g27410; RD26), which belong to the NAC transcription factor family. The NAC proteins bound specifically to the CATGTG motif both in vitro and in vivo, and activated transcription of a β-glucuronidase (GUS) reporter gene driven by the region containing the CATGTG motif in Arabidopsis T87 protoplasts. The expression of ANAC019, ANAC055 and ANAC072 was induced by drought, high salinity, ABA and MeJA. Using the yeast one-hybrid system, we determined the complete NAC recognition sequence, containing CATGT and with CACG as the core DNA binding site (Tran et al., 2004). Microarray analysis of transgenic plants over-expressing either ANAC019, ANAC055 or ANAC072 revealed that several stress-inducible genes including glyoxalase were upregulated in the transgenic plants. In good accordance with the upregulation of stress-inducible genes, the plants showed a significant increase in drought tolerance. Recently, we revealed that co-expression of the stress-inducible zinc finger homeodomain ZFHD1 and NAC transcription factors enhances expression of the ERD1 gene in Arabidopsis (Tran et al., 2007).
Kikuchi et al. (2000) described the molecular characteristics of the eight OsNAC genes, OsNAC1 to OsNAC8, that encode proteins with NAC domains in rice (Oryza sativa L.). Each OsNAC gene is expressed in a specific pattern in different organs, suggesting that this family has diverse and important roles in rice development. Ooka et al. (2003) performed a comprehensive analysis of NAC family genes in rice and Arabidopsis. They found 75 predicted NAC proteins in full-length cDNA data sets of rice and 105 in putative genes from the Arabidopsis genome. Recently, we monitored expression profiles of rice genes using a 1700 cDNA microarray and RNA gel-blot analyses (Rabbani et al., 2003), and found that the expression of OsNAC6 is induced by cold, drought, high salinity and ABA application. OsNAC6 showed high similarity to the Arabidopsis stress-related NAC proteins, ANAC019, ANAC055 and ANAC072 (RD26) (Ooka et al., 2003). Recently, Ohnishi et al. (2005) reported that OsNAC6 was induced by cold, high salinity, drought, ABA and JA.
In this study, we showed that the expression of OsNAC6 is induced not only by abiotic stresses but also by biotic stresses. In an attempt to investigate the in vivo functions of OsNAC6, we generated transgenic rice plants over-expressing this gene. We showed that many abiotic and biotic stress-responsive genes were upregulated in the transgenic plants, and that OsNAC6-over-expressing lines were tolerant to dehydration and high-salt stresses. Furthermore, the transgenic rice plants over-expressing OsNAC6 showed slightly improved tolerance to blast disease. As a means of avoiding the negative effect of OsNAC6 on the growth of rice, we propose that stress-inducible promoters can be utilized to over-express OsNAC6 under stress conditions. Here, we suggest that OsNAC6 is involved in abiotic and biotic stress-responsive gene expression in rice. We also propose that the combined usage of OsNAC6 and stress-inducible promoters might be useful for molecular breeding for stress tolerance.
Expression of OsNAC6 is induced by both abiotic and biotic stresses
The OsNAC6 gene (AB028185; AK068392; Os01 g0884300; ONAC048) encodes a protein of 303 amino acids containing the NAC domain in its N-terminal region (Figure 1a). The NAC domain contains predicted nuclear localization signals (NLS) at amino acids 71–83 and 107–123. RNA gel-blot and quantitative polymerase chain reaction (PCR) analyses showed that OsNAC6 was induced by dehydration, high salt (250 mm NaCl), cold (4°C), 100 μm ABA, 100 μm methyl jasmonate (MeJA) (Figure 1b and Supplementary Figure S1) and wounding (Figure 1c). Induction of OsNAC6 was observed in leaves infected with the blast fungus Magnaporthe grisea Kyu89-246 (Figure 1d). We also examined the effects of stress-related chemicals on the expression of OsNAC6 in rice culture cells. Quantitative PCR analysis showed that OsNAC6 was moderately induced by hydrogen peroxide (H2O2) and weakly by the elictor N-acetylchitooligosaccharide (Figure 1e).
In order to assess the effect of the promoter region on the expression of OsNAC6 under abiotic and biotic stresses in leaves and roots, we generated the transgenic rice plants containing 1.5 kb OsNAC6 promoter–GUS chimeric genes. Quantitative analysis of the OsNAC6 promoter–GUS transgenic rice plants showed that OsNAC6 was induced by dehydration, high salinity, cold, ABA, MeJA, hydrogen peroxide, wounding and blast disease (Figure 1f,g). Sequences of various cis-acting elements involved in the response to abiotic stresses were identified in the 1.5 kb promoter region of OsNAC6 (Figure 1h). We found three ABA-responsive elements (ABREs; ACGTGG/TC) (Hattori et al., 2002), three recognition sites for MYB (MYBRSs; C/TAACNA/G) (Abe et al., 2003) and six recognition sites for MYC (MYCRSs; CANNTG) (Abe et al., 2003). The OsNAC6 promoter also includes some cis-acting elements involved in the reponse to biotic stresses, such as four W-boxes (TTGAC) (Eulgem et al., 2000) and four GCC boxes (GCCGCC) (Brown et al., 2003), which are known as recognition sites for WRKY and ERF transcription factors, respectively. Additionally, the OsNAC6 promoter has three as-1 motifs (TGACG; Lam et al., 1989), which are known to be oxidative stress-responsive elements.
OsNAC6 functions as a transcriptional activator and is localized in the nucleus
An OsNAC6–sGFP fusion protein driven by the CaMV 35S promoter was transiently expressed in onion epidermal cells and analyzed by fluorescent microscopy. A SV40 NLS–sGFP fusion protein (SV40NLS–sGFP) and sGFP alone (35S–sGFP), driven by the 35S promoter, were used as a positive control (nuclear localization) and negative control, respectively. Nuclear localization was confirmed for OsNAC6 as both OsNAC6–sGFP and the positive control (SV40NLS–sGFP) were localized in the nucleus, whereas 35S–sGFP was localized in both cytoplasm and nucleus (Figure 2a). We also examined the transcriptional activity of OsNAC6 using a yeast system. A GAL4 DNA binding domain–OsNAC6 fusion protein was expressed in yeast cells, which were assayed for their ability to activate transcription from the GAL4 binding sequence. OsNAC6 promoted yeast growth in the absence of histidine and showed β-galactosidase activity, while the vector control pGBKT7 did not (Figure 2b). These data confirm that OsNAC6 functions as a transcriptional activator.
Transgenic rice plants over-expressing OsNAC6 show improved stress tolerance
To analyze the function of the OsNAC6 protein in rice plants, we generated transgenic rice plants over-expressing OsNAC6. The coding region of OsNAC6 was fused to the maize ubiquitin promoter (Christensen et al., 1992). We examined expression levels of the transgene in rice plants containing a ubiquitin promoter–OsNAC6 fusion (OsNAC6-OX plants) by RNA gel-blot analysis using an OsNAC6-specific probe (data not shown). These results were subsequently confirmed using quantitative PCR (Figure 3a). We used plants of the T2 or T3 generation of the transgenic rice lines 23–1 and 26–1 for further analyses. Growth of the OsNAC6-OX plants was compared with that of rice plants carrying the vector only (control plants) at 14 days after germination. In our hydroponic system, the OsNAC6-OX rice plants were smaller than the control plants (Figure 3b,c). We measured major agronomic traits such as plant height, number of panicles per plant, number of spikelets per panicle, and grain yield per plant for the OsNAC6-OX plants and control plants (Supplementary Table S1). The reproductive yields of the OsNAC6-OX plants were lower than those of the control plants.
Dehydration stress tolerance of the OsNAC6-OX rice plants was compared with that of the control plants. The OsNAC6-OX rice plants and the control plants were grown hydroponically for 14 days. Plants were then maintained under dry-air conditions for 12 h, and the plants were allowed to recover in nutrient solution for an additional 9–12 days. As shown in Table 1, 69–84% of the OsNAC6-OX plant survived, whereas only 27% of the control plants survived. However, the survival rates varied between experiments. Because it is not easy to control the dehydration conditions of rice plants that are grown hydroponically, we conducted a parallel stress tolerance assay under high-salt stress conditions.
Table 1. Survival rates of transgenic rice plants over-expressing OsNAC6 when subjected to dehydration stress
aThese plants have significantly higher survival rates than the vector control plants as determined using the chi-squared test (P <0.001).
OsNAC6-OX line 23–1
OsNAC6-OX line 26–1
The stress tolerance of OsNAC6-OX plants that were exposed to high salt was compared with that of the control rice plants. Transgenic and control plants were grown hydroponically for 14 days and were then transferred into 250 mm NaCl solution for 3 days. After the stress treatment, plants were allowed to recover in nutrient solution for 9–12 days. The OsNAC6-OX plants had higher photosynthetic activity than the control plants as shown by measuring chlorophyll fluorescence corresponding to the maximum photochemical efficiency of PSII in the dark-adapted state (Fv/Fm) (Figure 3d). These results demonstrated that over-expression of OsNAC6 results in improved stress tolerance in rice plants. As shown in Table 2, more than 90% of the OsNAC6-OX rice plants survived, whereas only 26% of the control plants survived. Representative results for the control and OsNAC6-OX plant lines are shown in Figure 3e–g).
Table 2. Survival rates of transgenic rice plants over-expressing OsNAC6 when subjected to high-salt stress
aThese plants have significantly higher survival rates than the vector control plants as determined using the chi-squared test (P <0.001).
OsNAC6-OX line 23-1
OsNAC6-OX line 26-1
We then checked the disease tolerance of the OsNAC6-OX plants. OsNAC6-OX plants and control plants at the 5.7–6.2 leaf stage were inoculated with rice blast fungus (M. grisea) and grown in a greenhouse for 1 week. The length and width of the disease lesions were measured. Although the lesion length of the OsNAC6-OX plants was similar to that of the control plants, the lesion width of the OsNAC6-OX plants was slightly smaller than that of the control plants after the 1 week incubation period (Figure 3h). These results suggest that the rice blast fungus and/or disease symptoms may not penetrate into the leaf veins in the OsNAC6-OX plants.
The stress-inducible promoters are useful for over-expressing OsNAC6 in rice to improve stress tolerance by suppressing the negative effects of OsNAC6 on growth
We used stress-inducible promoters instead of the constitutive ubiquitin promoter for over-expression of OsNAC6 under stress conditions in transgenic rice to minimize the negative effects of OsNAC6 on plant growth under normal growth conditions. We selected two promoters from rice, the OsNAC6 promoter and the LIP9 promoter. LIP9 encodes a low-temperature-induced protein (Aguan et al., 1991). We previously showed that LIP9 is also induced by cold, drought and high-salinity stresses, and by ABA application (Rabbani et al., 2003). We isolated the 1.5 kb promoter region of OsNAC6 and the 1.5 kb promoter region of LIP9 from the rice genome (Nipponbare) and used them as stress-inducible promoters to over-express OsNAC6. We generated rice plants carrying the OsNAC6 gene fused to either the OsNAC6 promoter (POsNAC6–OsNAC6) or the LIP9 promoter (PLIP9–OsNAC6), and examined the expression level of OsNAC6 in rice plants containing POsNAC6–OsNAC6 or PLIP9–OsNAC6 by using quantitative PCR analysis (Figure 4a,d). The OsNAC6 mRNA levels in the POsNAC6-OsNAC6 lines 1–1 and 3–1 and the PLIP9-OsNAC6 lines 3–1 and 5–1 were higher than those in the control plants. Although the OsNAC6 mRNA levels were high, these lines did not show severe growth retardation, in contrast with the OsNAC6-OX plants (Figure 4a–f). We measured the major agronomic traits of the POsNAC6-OsNAC6 plants, the PLIP9-OsNAC6 plants and the control plants (Supplementary Table S1). Although the reproductive yields of the PLIP9-OsNAC6 plants were lower than those of the control plants, no significant difference was detected between the POsNAC6-OsNAC6 plants and the controls. These results indicate that over-expression of OsNAC6 using the OsNAC6 promoter does not affect growth and reproductive yields of rice.
As the POsNAC6–OsNAC6 and PLIP9–OsNAC6 constructs were fused to the sGFP protein, we determined the stress-inducible cellular localization of the OsNAC6–sGFP fusion proteins after the high-salt stress treatment using fluorescent microscopy. The OsNAC6–sGFP fusion proteins were observed in the nuclei of lateral root cells after 24 h of 250 mm NaCl treatment in POsNAC6-OsNAC6 line 3–1 (Figure 4g) and PLIP9-OsNAC6 line 5–1 (Figure 4i). It is important to note that clear cellular localization of the OsNAC6–sGFP fusion proteins was not observed in POsNAC6-OsNAC6 line 3–1 (Figure 4h) or PLIP9-OsNAC6 line 5–1 (Figure 4j) without high-salt stress treatment.
As described above, the tolerance to high salinity of the POsNAC6-OsNAC6 plants and the PLIP9-OsNAC6 plants was compared with that of the control rice plants. As shown in Table 3 15% of the control rice plants died, whereas 64–95% of the POsNAC6-OsNAC6 plants and the PLIP9-OsNAC6 plants survived. Representative results are shown in Figure 4k–n.
Table 3. Survival rates of transgenic rice plants over-expressing stress-inducible OsNAC6. when subjected to high-salt stress
aThese plants have significantly higher survival rates than the vector control plants as determined using the chi-squared test (P <0.001).
POsNAC6-OsNAC6 line 1-1
POsNAC6-OsNAC6 line 3-1
PLIP9-OsNAC6 line 3-1
PLIP9-OsNAC6 line 5-1
We also checked the disease tolerance of the POsNAC6-OsNAC6 plants and the PLIP9-OsNAC6 plants. The transgenic plants were inoculated with M. grisea as described above. The lesion sizes on the POsNAC6-OsNAC6 and PLIP9-OsNAC6 plants were similar to that on the control plants (data not shown).
Stress-related genes are upregulated in rice plants over-expressing OsNAC6
We were interested to elucidate the molecular mechanism of stress tolerance mediated by OsNAC6 in rice. Therefore, we identified genes that were upregulated in the OsNAC6-OX plants by using the 22 K rice oligo microarray (Agilent Technologies, Inc.). We used two independent OsNAC6-OX lines (26–2 and 27–1) for the microarray analysis. Both of these OsNAC6-OX lines also showed growth retardation and improved tolerance to high-salt stress (data not shown). We identified 163 upregulated genes in OsNAC6-OX plants (Supplementary Table S2). Based on microarray analysis of stress-inducible genes, 58 genes (37% of 155 genes for which microarray data are available) were found to be induced by dehydration, high-salt stress or cold stress (Supplementary Table S2 and unpublished results). Thirty-four genes (22%), 33 genes (21%) and 20 genes (13%) were induced by high salinity, dehydration and cold, respectively. Sequence analysis of the upregulated genes in the OsNAC6-OX plants identified genes with predicted functions that are involved in the regulation of stress responses. Specifically, five protein kinases and 11 transcription factors containing domains such as MYB, AP2/ERF and zinc finger were identified. In addition, genes encoding proteins with predicted functions in stress tolerance such as detoxification (glutathione S-transferase etc.), redox homeostasis (aldo/keto reductase family protein, NAD(P)H-dependent oxidoreductase etc.) and proteolytic degradation (ubiquitin, peptidase, etc.) were identified. Interestingly, we also identified genes encoding proteins with predicted functions in biotic stress responses and tolerance, such as peroxidase (14 genes), β-1,3-glucanase-like protein (four genes) and chitinase (three genes).
Upregulated genes in rice plants over-expressing OsNAC6 and DEX-treated rice plants over-expressing OsNAC6-GR
As OsNAC6-OX plants showed growth retardation, it is possible that the growth differences may have an effect on the transgenic plants at a molecular level. In order to precisely control the temporal over-expression of OsNAC6, we generated transgenic rice plants over-expressing OsNAC6 fused to a rat glucocorticoid receptor (GR) (Miesfeld et al., 1986). The coding region of OsNAC6 was fused to the GR gene and driven by the maize ubiquitin promoter. In this system, the OsNAC6–GR fusion protein moves from cytoplasm to the nucleus when the transgenic plants are treated with dexamethasone (DEX). Expression levels of the transgene in rice plants over-expressing the OsNAC6–GR chimeric gene fused to the ubiquitin promoter (OsNAC6-GR plants) were examined by RNA gel-blot analysis. The OsNAC6 mRNA levels in OsNAC6-GR plant lines 37 and 39 were higher than those in the control plants (Supplementary Figure S2a). We used OsNAC6-GR plants of the T2 or T3 generation for further analyses. Hydroponically grown OsNAC6-GR rice plants were similar to the control rice plants with regard to plant height (Supplementary Figure S2b,c). The tolerance of the OsNAC6-GR rice plants to high-salt stress was compared with that of the control plants as described above. Fourteen-day-old plants were transferred to a nutrient solution supplemented with or without 10 μm DEX for a period of 24 h. Plants were then grown in 250 mm NaCl solution for 3 days and subsequently grown hydroponically for 9–12 days (recovery period). Although 37% of the control rice plants died, more than 90% of the OsNAC6-GR plants survived when the plants were treated with DEX (Supplementary Figure S2d and Supplementary Table S3). No significant differences were detected among survival rates of control rice plants and OsNAC6-GR-over-expressing plants when the DEX treatment was omitted. Two independent transgenic rice plant lines over-expressing OsNAC6-GR were used for microarray analysis. Briefly, cDNA probes were prepared from mRNAs isolated from the 24 h DEX-treated rice plants over-expressing OsNAC6-GR and from 24 h DEX-treated control plants. These probes were subsequently hybridized with the microarray, and the expression profiles were analyzed using the 22 K rice oligo microarray. Twenty-seven upregulated genes were identified in the DEX-treated OsNAC6-GR plants, including genes encoding stress-related proteins such as a PR protein 1, probenazole-inducible proteins (PBZ1s), thioredoxin, peroxidese and lipoxygenase (Supplementary Table S4). Four genes were found to be upregulated in both the OsNAC6-OX plants and the DEX-treated OsNAC6-GR plants, including AK058583 for lipoxygenase, AK104277 for peroxidase, AK105331 for an unknown protein, and AK110725 for a DUF26-like protein containing a Ser/Thr protein kinase motif of unknown function. Microarray analysis of stress-inducible genes indicated that AK104277 and AK110725 were induced by high salinity, and AK105331 was induced by dehydration and cold (Supplementary Table S4 and unpublished results).
Expression analysis of the upregulated genes in rice plants over-expressing OsNAC6 and DEX-treated rice plants over-expressing OsNAC6-GR
To provide further confirmation of candidate genes that are direct targets of OsNAC6, we selected two genes (AK104277 and AK110725) that were upregulated in both the OsNAC6-OX plants and the DEX-treated OsNAC6-GR plants. We subjected them to further analysis. We confirmed the expression patterns of OsNAC6, AK104277 and AK110725 using quantitative PCR analysis (Figure 5a–c, i–ii). Expression of AK104277 and AK110725 was induced by stresses and was especially affected by dehydration and high salinity (Figure 5b,c, iv). To determine whether OsNAC6 is capable of transactivating the promoters of these genes, we performed transactivation experiments using protoplasts prepared from rice cultured cells. Protoplasts were co-transfected with GUS reporter constructs containing 1.2 kb promoter regions of AK104277 and AK110725 and effector plasmids containing the OsNAC6 cDNA fused to the ubiquitin promoter (Figure 5d). OsNAC6 activated expression of the GUS reporter gene fused to the promoter region of AK104277 and AK110725. We conducted protoplast transactivation experiments using a promoter containing repeated ABRE-responsive elements (Nakashima et al., 2006). Although we observed increased GUS activity in the presence of the rice AREB-like gene, we did not observe increased GUS activity in the presence of OsNAC6 (data not shown). These results suggest that OsNAC6 functions as a transcription activator that is involved in regulating the expression of AK104277 and AK110725.
In this paper, we have shown that OsNAC6 is induced by dehydration, high salinity, cold, ABA, MeJA, wounding, hydrogen peroxide, elicitor N-acethylchitooligosaccharide and blast disease (Figure 1 and Supplementary Figure S1). Collectively, these results indicate that OsNAC6 might be involved in both biotic and abiotic stress responses in rice. The sequences of various cis-acting elements, including ABREs, MYBRS, MYCRS, W-boxes, GCC boxes and as-1 motifs, are present in the 1.5 kb promoter region of OsNAC6 (Figure 1h). The ABREs, MYBRS and MYCRS are known as recognition sites of AREB/ABF, MYB and MYC transcription factors, respectively (reviewed by Yamaguchi-Shinozaki and Shinozaki, 2005). These cis-elements and their respective transcription factors have important roles for ABA signaling and abiotic stress responses (Yamaguchi-Shinozaki and Shinozaki, 2006). The W-boxes and GCC boxes are known as recognition sites for WRKY and ERF transcription factors, respectively. These appear to be involved in the regulation of various plant-specific physiological processes such as pathogen defense and senescence (Brown et al., 2003; Eulgem et al., 2000). The as-1 motifs are known as oxidative stress-responsive elements (Lam et al., 1989). These sequences may function as abiotic and/or biotic stress-responsive cis-acting elements in the OsNAC6 promoter.
We conducted a transactivation assay using a yeast system, and found that OsNAC6 functions as a transcriptional activator (Figure 2b). A transient assay showed that the OsNAC6–sGFP fusion protein is localized in the nucleus (Figure 2a), thereby providing further evidence that OsNAC may function as a transcription factor in vivo. The results of microarray analysis revealed that many stress-related and stress-responsive genes were upregulated in the OsNAC6-OX rice plants (Supplementary Table S2). Sequence analysis indicated that some of these genes encode proteins with predicted functions that are involved in abiotic stress responses and tolerance. Interestingly, some genes encode proteins with predicted functions that are involved in biotic stress responses, such as PR 1a, peroxidase, endochitinase and a β-1,3-glucanase-like protein. A large number of peroxidase genes were identified among the upregulated genes (14 genes; Supplementary Table S2). Peroxidases are enzymes that use hydrogen peroxide as the electron acceptor to catalyze a number of oxidative reactions. It is thought that peroxidases provide protection to cells under oxidative stress. Sasaki et al. (2004) reported the expression profiles of 22 rice class III peroxidase genes after infection with rice blast fungus. It is notable that the peroxidase gene R2184 was responsive to blast infection and JA, and was upregulated in the OsNAC6-OX rice plants (AK102307; Supplementary Table S2). Microarray analysis of stress-inducible genes showed that approximately 40% of the upregulated genes in the OsNAC6-OX rice plants were responsive to dehydration, high salinity or cold stress (Supplementary Table S2 and unpublished data). These data indicate that a considerable number of the upregulated genes might be the direct targets of OsNAC6. However, it is possible that some genes might be the secondary effect of OsNAC6 over-expression, as the OsNAC6-OX rice plants showed a growth retardation phenotype. Microarray analysis of the DEX-induced system using OsNAC6-GR plants revealed that four genes were upregulated in both the OsNAC6-OX and DEX-treated OsNAC6-GR plants (Supplementary Tables S2 and S4). A transient transactivation assay showed that OsNAC6 activates the transcription of at least two genes, including genes encoding peroxidase and a DUF26-like protein (Figure 5d). The transgenic rice plants over-expressing OsNAC6 showed improved stress tolerance (Figures 3 and 4, Tables 2 and 3, and Supplementary Figure S2). It is possible that these enzymes and other target genes might work together cooperatively to avoid cell damage by oxidative stress, and thereby function to protect the cell from such stress. We recently generated RNAi rice plants using the C-terminal region of OsNAC6. Our preliminary data suggest that 14-day-old RNAi rice plants were taller than control plants, and the salt-stress tolerance of the RNAi plants was similar to that of the control plants. In future studies, detailed analysis of the RNAi knock-down plants and/or T-DNA-tagged knock-out plants will help us to understand the function of OsNAC6.
The OsNAC6-OX plants showed growth retardation and lower productivity, although they showed improved stress tolerance (Tables 1 and 2 and Figure 3). Although many kinds of transcription factors have been recently reported to improve stress tolerance (Fujita et al., 2005; Furihata et al., 2006; Jaglo-Ottosen et al., 1998; Kang et al., 2002; Liu et al., 1998; Sakuma et al., 2006), growth retardation is commonly observed in transgenic plants over-expressing stress-related transcription factors (reviewed by Nakashima and Yamaguchi-Shinozaki, 2005; Zhang et al., 2004). For example, the DREB1/CBF genes of Arabidopsis have been shown to improve abiotic stress tolerance in a number of different species; however, their constitutive over-expression in transgenic plants showed an undesirable dwarf phenotype (Gilmour et al., 2000; Liu et al., 1998). We have reported that a combination of the Arabidopsis DREB1A gene and the stress-inducible RD29A promoter improved environmental stress tolerance in Arabidopsis and tobacco by gene transfer (Kasuga et al., 1999, 2004). We also showed that stress-inducible promoters, including the LIP9 promoter and especially the OsNAC6 promoter, effectively over-expressed OsNAC6 in rice and simultaneously improved stress tolerance without growth retardation effects (Figure 4 and Table 3). Thus, we conclude that both of these stress-inducible promoters can be effectively used in rice to over-express stress-tolerant genes under stress conditions, including OsNAC6, as a means to improve stress tolerance by suppressing the negative effects of OsNAC6 on plant growth under normal growth conditions.
Recent studies have revealed several molecules, including transcription factors and protein kinases, as promising candidates for common players involved in cross-talk between stress signaling pathways (Fujita et al., 2006). As mentioned above, the expression of Arabidopsis ANAC019, ANAC055 and ANAC072 (RD26) was induced by dehydration, high salinity, ABA and MeJA (Fujita et al., 2004; Tran et al., 2004). In this paper, we show that expression of OsNAC6 is induced by biotic stress as well as abiotic stresses. OsNAC6 is one of the closest orthologue of Arabidopsis ATAF2 (Delessert et al., 2005; Ooka et al., 2003), which is a gene that is highly induced by wounding and is responsive to MeJA and SA, but not to ABA (Delessert et al., 2005). Over-expression of ATAF2 leads to an increased biomass and yellowing of the leaves, but there was no obvious phenotype in two independent ATAF2 T-DNA insertion lines. At the transcriptome level, ATAF2 over-expression resulted in the repression of a number of PR proteins. Conversely, four of these pathogenesis-related transcripts were increased in both ATAF2 knock-out lines. Plant over-expressing ATAF2 showed a higher susceptibility to the soil-borne fungal pathogen Fusarium oxysporum. These results indicate that ATAF2 functions as a repressor of PR proteins in Arabidopsis. Although ATAF2 is responsive to drought stress, the functional role of ATAF2 in abiotic stress is not known. The transgenic rice plants over-expressing OsNAC6 showed growth retardation and exhibited slightly improved tolerance to blast disease (Figure 3h). Furthermore, many biotic-stress-related genes, including those for PR proteins, were upregulated in the OsNAC6-OX plants (Supplementary Table S2) and the DEX-treated OsNAC6-GR plants (Supplementary Table S4). Our results indicate that OsNAC6 does not function as a repressor but rather as an activator of PR proteins in rice. Although ATAF2 is one of the closest orthologue of OsNAC6 in Arabidopsis, ATAF2 may have a different function than OsNAC6.
Recently, Hu et al. (2006) reported that over-expression of the stress-responsive gene SNAC1 (STRESS-RESPONSIVE NAC1; ONAC033; AK104551) enhances drought and salt tolerance in transgenic rice without growth retardation. Although SNAC1 has homology to OsNAC6, OsNAC6 belongs to the ATAF subgroup and SNAC1 belongs to the OsNAC3 subgroup (Ooka et al., 2003). The expression of SNAC1 under biotic stress and the role of SNAC1 in biotic stress are unknown. Completion of the rice genome project revealed that the rice genome contains six homologous genes including OsNAC6 and SNAC1 (Ooka et al., 2003). Our microarray experiments showed that five of these homologues were induced by high salinity, dehydration and cold stresses (unpublished results). At the present time, the gene expression patterns, functional relatedness to growth and tolerance to abiotic and biotic stresses, and the target genes of these homologous factors remain to be solved.
In conclusion, this study shows that OsNAC6 is induced by biotic as well as abiotic stresses. The enhanced tolerance to high salinity and dehydration of the rice plants over-expressing OsNAC6 suggest that this gene may be useful for improvement of stress tolerance in various kinds of transgenic plants.
Plant materials and stress treatments
Seedlings of rice (O. sativa cv. Nipponbare) were grown in basal nutrient solution as described by Makino et al. (1988). Two-week-old rice plants were dehydrated in an plastic box for dehydration treatment. The plants were transferred from the basal nutrient solution to nutrient solution containing 250 mm NaCl, 100 μm ABA, 100 μm methyl jasmonate (MeJA), 100 μm salicylic acid (SA), 100 μm ethephon or 20 mm H2O2 for stress treatments. For cold treatment, plants were transferred to and kept at 4°C for the indicated times. For wounding, leaves of 2-week-old plants were scratched using a needle. Wounded leaves were subsequently maintained on water-saturated filter paper for the indicated times. For biotic stress, the leaves of 4-week-old transgenic plants were sprayed with a spore suspension of the rice blast fungus (M. grisea) Kyu89–246 (MAFF number 101506, National Institute of Agrobiological Sciences, Tsukuba, Japan).
Cultured cell materials and stress treatments
A suspension culture of a rice cell line (Oc) was used (Hattori et al., 1995) for various stress treatments. Cultured cells were dehydrated on Whatman 3MM filter paper, cultured in the liquid medium R2P containing 20 mm H2O2, 1 μg/ml N-acetylchitooligosaccharide elicitor (Yamaguchi et al., 2005), or transferred to and cultured at 4°C for indicated times. Control cells were harvested at the same time as the stressed cells.
RNA gel-blot analysis and quantitative PCR analysis
Isolation of total RNAs, RNA gel-blot hybridization and quantitative PCR analysis were performed as described previously (Nakashima et al., 2006). The 3′ end-specific DNA fragment of OsNAC6 between 483 and 909 nt from the start codon (ATG) and 18S rRNA were used as probes for hybridization. OsNAC6 primers 5′-CCAGCCCAAGATCAGCGAGT-3′ (744–763 nt from ATG) and 5′-TCATGTACTGGGGCA AGCCA-3′ (887–906 nt from ATG) were used for quantitative PCR analysis. The 18S rRNA primers 5′-ATGGTGGTGACGGGTGAC-3′ and 5′-CAGACACTAAAGCGCCCGGTA-3′ were used for the normalization of the quantitative PCR analysis.
Production and analysis of transgenic rice plants containing OsNAC6 promoter–GUS fusions
The promoter region of OsNAC6 (1516 bp upstream of ATG) was amplified from the rice genome by PCR using KOD DNA polymerase (Toyobo, http://www.toyobo.co.jp/bio). The OsNAC6 promoter fragment was inserted into the promoterless GUS expression vector pBIH containing the hygromycin resistance gene in pBI101 (Y. Yoshiba, unpubl. data). Confirmation of the fusion constructs, transformation of rice, and growth of the transgenic plants were performed as described previously (Ito et al., 2006). Quantitative analysis of GUS activity was performed as described previously (Nakashima and Yamaguchi-Shinozaki, 2002).
Transient expression of the sGFP protein in onion epidermal cells
OsNAC6 was examined for the presence of an activation domain using a yeast assay system as described previously (Fujita et al., 2004).
Production and analysis of transgenic plants over- expressing OsNAC6
To generate transgenic rice plants over-expressing OsNAC6 (OsNAC6-OX), we used the pBIG-ubi vector (Becker, 1990; Ito et al., 2006). To generate transgenic rice plants containing stress-inducible promoter–OsNAC6 constructs, we used the pGreen II 0129 vector (Hellens et al., 2000) containing OsNAC6 fused to the sGFP gene (Chiu et al., 1996; Nakashima et al., 1997). The promoter region of OsNAC6 (1516 bp upstream of ATG) and the promoter region of the LIP9 gene (1061 bp upstream of ATG) were amplified from the rice genome by PCR using KOD DNA polymerase (Toyobo). The OsNAC6 and LIP9 promoter fragments generated were inserted upstream of the OsNAC6–sGFP chimeric gene. We introduced the constructs into wild-type rice cv. Nipponbare by Agrobacterium-mediated transformation (Goto et al., 1999), and T2 and T3 seeds were used for experiments. We grew the transgenic rice plants as described previously (Ito et al., 2006). Cellular localization of the OsNAC6–sGFP proteins in the lateral root cells of the transgenic rice plants was observed with confocal laser scanning microscope LSM5 PASAL (Carl Zeiss). We used the pBIG-ubi vector to generate transgenic rice plants over-expressing the OsNAC6-GR chimeric gene (OsNAC6-GR) (Becker, 1990; Ito et al., 2006). The rat GR gene (Miesfeld et al., 1986) was fused to the C-terminus of OsNAC6.
Stress tolerance in the transgenic rice plants
For high-salt stress treatment, plants were pre-grown hydroponically under normal conditions for 14 days and transferred to 250 mm NaCl solution for 3 days. Subsequent to the stress treatment, plants were transferred to normal hydroponic conditions and were grown for an additional 9–12 days. For dehydration treatment, plants that had been pre-grown hydroponically for 14 days were removed from the solution and kept in plastic boxes for 12 h. These plants were rehydrated and grown for an additional 9–12 days. The numbers of plants that survived and continued to grow were counted. The statistical significance of the values was determined using the chi-squared test. Photosynthetic activity (Fv/Fm) was measured according to the method described by Kasuga et al. (2004).
Disease tolerance test
OsNAC6 over-expression rice plants and control plants at the 5.7–6.2 leaf stage were inoculated with M. grisea Kyu89-246 using a puncher (1.5 mm in diameter), and the infected plants were grown in greenhouse for 1 week. The length and width of the disease lesions was measured (n =12–18).
Seedlings of transgenic rice that had been grown hydroponically for 14 days were harvested, and total RNAs were isolated by the TRIzol method (Invitrogen; http://www.invitrogen.com/). Total RNA was used for the preparation of Cy5- and Cy3-labeled cDNA probes, and these were subjected to microarray experiments using the 22 K rice oligo microarray (Agilent Technologies Inc., http://www.home.agilent.com). All microarray experiments including data analysis were carried out as described previously (Ito et al., 2006). Briefly, two independent transgenic lines were used for both OsNAC6-OX and OsNAC6-GR plants in addition to the vector control plants. The reproducibility of the microarray analysis was assessed by dye swapping in each experiment. Upregulated genes were selected according to the following three criteria as described previously (Tran et al., 2007): (i) the fold increase value was greater than three for all four comparisons of OsNAC6-OX or OsNAC6-GR, (ii) the hybridization intensity for OsNAC6-OX or OsNAC6-GR was greater than 1000 for all four comparisons, and (iii) the P-value was <0.001 for all four comparisons.
Transactivation experiments with rice protoplasts
The ORF for the OsNAC6 cDNA was inserted into a plant expression vector ubi vector (unpubl. data) containing a ubiquitin promoter (Christensen et al., 1992; Dubouzet et al., 2003). A ubiquitin–luciferase plasmid was used as an internal control to normalize GUS values. We inserted the 1.2 kb promoter regions of AK058583 or AK110725 to generate reporter constructs. Suspension culture of a rice cell line (Oc), protoplast isolation and electroporation were performed according to the techniques described by Hattori et al. (1995). Luciferase and GUS assays were performed as described previously (Nakashima et al., 2006). The transactivation data are based upon three independent transformations.
We are grateful for the excellent technical support provided by Ekuko Ohgawara, Emiko Kishi, Kyoko Murai, Kyoko Yoshiwara and Satoshi Kidokoro of Japan International Research Center for Agricultural Sciences. We thank the Rice Genome Resource Center at National Institute of Agrobiological Sciences for use of the microarray analysis system and the release of rice full-length cDNA clones. We thank Dr Eiichi Minami of the National Institute of Agrobiological Sciences and Dr Naoto Shibuya of Meiji University for their technical advice and supply of the elicitor. We thank Dr Yoshu Yoshiba of Hitachi for kindly supplying the pBIH vector. This work was supported in part by the Program for Promotion of Basic Research Activities for Innovative Biosciences (BRAIN). It was also supported in part by a project grant to K.Y.-S. from the Ministry of Agriculture, Forestry and Fisheries, Japan.