A rice gene encoding a calcium-dependent protein kinase (CDPK), OsCDPK7, was induced by cold and salt stresses. To elucidate the physiological function of OsCDPK7, we generated transgenic rice plants with altered levels of the protein. The extent of tolerance to cold and salt/drought stresses of these plants correlated well with the level of OsCDPK7 expression. Therefore, OsCDPK7 was shown to be a positive regulator commonly involved in the tolerance to both stresses in rice. Over-expression of OsCDPK7 enhanced induction of some stress-responsive genes in response to salinity/drought, but not to cold. Thus, it was suggested that the downstream pathways leading to the cold and salt/drought tolerance are different from each other. It seems likely that at least two distinct pathways commonly use a single CDPK, maintaining the signalling specificity through unknown post-translational regulation mechanisms. These results demonstrate that simple manipulation of CDPK activity has great potential with regard to plant improvement.
Environmental stresses, such as cold, salinity and drought, have an enormous impact on crop productivity throughout the world ( Boyer, 1982; Epstein et al. 1980 ). To survive under unfavourable conditions, plants have developed a variety of sophisticated strategies ( Bohnert et al. 1995 ; Bray, 1997; Thomashow, 1999; Zhu et al. 1997 ). The products of some stress-inducible genes (e.g. enzymes involved in the biosynthesis of various osmoprotectants and late-embryogenesis-abundant (LEA) proteins) directly counteract the detrimental conditions. Transfer of these genes into plants confirmed their protective roles in stress adaptation ( Holmberg & Bulow, 1998; Xu et al. 1996 ). Although the effect of each individual gene is rather small, simultaneous transcriptional activation of a subset of these genes can confer much greater stress tolerance on Arabidopsis (Jaglo-Ottosen et al. 1998; Liu et al. 1998 ). It is generally thought that distinct mechanisms underlie the adaptation to cold and salt/drought stresses in plants ( Shinozaki & Yamaguchi-Shinozaki, 1997; Zhu et al. 1997 ). Both appear to be regulated by complex signalling networks of abscisic acid (ABA)-dependent pathways and/or ABA-independent pathways ( Ishitani et al. 1997 ; Leung & Giraudat, 1998; Shinozaki & Yamaguchi-Shinozaki, 1997). It has been predicted that the modulation of signalling regulators will be a promising method for improving the stress tolerance of plants. However, little is known about the molecular mechanisms underlying signal transduction.
Previous studies showed that the cytoplasmic Ca2+ levels in plant cells increase rapidly in response to multiple stress stimuli, including cold, salt and drought ( Sanders et al. 1999 ; Trewavas & Malho, 1997). Following this Ca2+ influx, signals are likely to be mediated by combinations of protein phosphorylation/dephosphorylation cascades. The experimental perturbation with specific reagents indicated the pivotal roles of Ca2+ influx and protein phosphorylation in these stress responses ( Knight et al. 1996 ; Knight et al. 1997 ; Monroy & Dhindsa, 1995; Monroy et al. 1993 ; Monroy et al. 1998 ; Tahtiharju et al. 1997 ). It is presumed that the majority of Ca2+-stimulated protein phosphorylation is performed predominantly by members of the Ca2+-dependent protein kinase (CDPK) family in plants ( Sanders et al. 1999 ; Trewavas & Malho, 1997). This kinase family, currently only known in plants and protozoa, contains a calmodulin-like regulatory domain with four EF hands, Ca2+-binding site, at its C-terminal end, which enables activation directly through Ca2+ binding ( Roberts & Harmon, 1992). Two of the four constitutively active mutant enzymes of two related Arabidopsis CDPKs activated a stress/ABA-responsive promoter in a transient expression system, indicating that selected members of the CDPK family are involved in that particular stress signalling ( Sheen, 1996). In addition, stress-induced gene expression of some CDPK members has been reported in various plant species, and they are proposed to mediate stress signals ( Berberich & Kusano, 1997; Botella et al. 1996 ; Urao et al. 1994 ; Yoon et al. 1999 ). However, it is not known whether or not the increases in mRNA levels are accompanied by increases in protein levels and/or kinase activities. It is also noteworthy that a number of transport proteins (e.g. aquaporins, H+-ATPases and ion channels), which are responsible for cytosolic osmoregulation and involved in stress adaptation, are regulated by CDPKs ( Bethke & Jones, 1997; Camoni et al. 1998a ; Li et al. 1998 ; Lino et al. 1998 ; Pei et al. 1996 ; Weaver & Roberts, 1992). Despite the potential importance of CDPKs, the physiological function of a specific CDPK pathway has not been elucidated so far.
In the current study, we investigated the function of a rice cold- and salt-inducible CDPK, OsCDPK7, by using transgenic rice plants with altered levels of the protein. Over-expression of OsCDPK7 conferred both cold and salt/drought tolerance on rice plants. In contrast, suppression of OsCDPK7 expression lowered the stress tolerance. The results indicate that OsCDPK7 plays key roles in the tolerance to both stresses in rice.
cDNA cloning of OsCDPK7
Two maize CDPKs, ZmCDPK1 and ZmCDPK7, showed 97.1% identity at the amino acid level over the entire polypeptide ( Berberich & Kusano, 1997; Saijo et al. 1997 ), and the mRNA expression of both CDPK genes were induced by cold stress ( Berberich & Kusano, 1997; data not shown). Using the ZmCDPK7 cDNA as a probe, we isolated a highly homologous rice clone, designated OsCDPK7 (GenBank accession no. AB042550), from a rice root cDNA library ( Hata et al. 1997 ). In contrast to the maize genome, Southern blot analysis of rice genomic DNA with the 3′-non-coding region (nt 1815–2061) as a probe indicated that the rice genome contains only a single copy of this gene (data not shown).
OsCDPK7 and ZmCDPK7 showed 88.6% identity at the amino acid level over the entire polypeptide. In addition, OsCDPK7 showed high amino acid identity to a mung bean CDPK, VrCDPK1 (85.4%), the transcript level of which was elevated by salt stress and mechanical strain ( Botella et al. 1996 ). They were classified into the same subclass of CDPKs in a phylogenetic tree for the amino acid sequences of the entire ORFs ( Fig. 1). A further database search revealed an Arabidopsis expressed sequence tag (EST) clone (accession no. R90026) encoding a portion of CDPK (124 amino acids). The deduced protein sequence was more closely related to OsCDPK7 than ATCDPK1( Sheen, 1996; Urao et al. 1994 ), a putative Arabidopsis stress-signalling isoform (amino acid identities of 80.6 and 52.8%, respectively). Therefore, it seems likely that OsCDPK7, ZmCDPK7, ZmCDPK1, VrCDPK1 and the Arabidopsis protein may be orthologues that play identical roles under stress conditions.
The entire coding region of OsCDPK7 was subcloned into an expression vector, pGEX4T-1 (Amersham Pharmacia), and then transformed into Eschericia coli BL21 (DE3). Protein kinase activity was assayed as described ( Saijo et al. 1997 ). The recombinant OsCDPK7 protein fused with glutathione-S-transferase (GST) efficiently phophorylated histone IIIS, casein, and myelin basic protein (Sigma) in a Ca2+-dependent manner (data not shown). This broad substrate specificity is in sharp contrast to the narrow specificity of a GST–ATCDPK1 fusion protein ( Urao et al. 1994 ), indicating that OsCDPK7 and ATCDPK1 differ in substrate preferences.
OsCDPK7 is induced by cold and salt stresses
Expression of the OsCDPK7 mRNA, around 2.3 kb in size, was increased by cold and salt stresses in both shoots ( Fig. 2a) and roots (data not shown) of 10-day-old seedlings, but not by exogenous abscisic acid (ABA) application. In contrast, a rice stress-responsive gene, rab16A ( Mundy & Chua, 1988), was induced by the salt and ABA treatments ( Fig. 2a). This indicates that OsCDPK7 belongs to a subclass of stress-inducible CDPKs, which is probably conserved from monocotyledonous to dicotyledonous plants.
To determine whether or not the protein level increases upon stressing, we then carried out immunoblot analyses with isoform-specific antibodies raised against the C-terminal portion of OsCDPK7. A clear band corresponding to an apparent molecular mass of 51 kDa was seen in the soluble fractions of the protein extracts from plant tissues ( Fig. 2b). Significant variation in the protein level was not detected during the cold-stress period examined in either the shoots ( Fig. 2b) or roots (data not shown), even when the mRNA accumulated to a high level ( Fig. 2a). The protein levels were almost similar between shoots and roots (data not shown). In the microsomal membrane fractions precipitated by the centrifugation, no signal for the OsCDPK7 protein was detected (data not shown). Similar results were obtained under salt stress (data not shown). Thus, the intracellular localization of the enzyme protein is as yet unclear. Moreover, the shoot proteins immunoprecipitated with anti-OsCDPK7 antibodies did not exhibit any changes in kinase activity (data not shown). However, the activity seemed to be affected by contaminated Ca2+ during the preparation of protein extracts. Therefore, in this experiment, additional factor(s) that positively or negatively influence the activity other than Ca2+ could not be detected either.
Transgenic rice plants with altered expression levels of OsCDPK7
To clarify the physiological role of OsCDPK7 in rice, the full-length cDNA (nt 1–1970) for OsCDPK7 was introduced into rice cells in the sense orientation under the control of the cauliflower mosaic virus (CaMV) 35S promoter ( Fig. 3a) by means of Agrobacterium-mediated transformation ( Hiei et al. 1994 ). Then, 14 independent lines of transgenic plants (T0 generation) were generated. DNA blot analysis of these lines confirmed that independent lines contained one to several copies of the transgene per haploid genome (data not shown). Among them, two lines over-expressing OsCDPK7, S3 and S1, and another co-suppressed line, S27, were chosen for further experiments. In the next T1 generation, the levels of the OsCDPK7 protein were constitutively higher in the two over-expressing lines, S3 and S1, than those in the segregated non-transgenic (NT) line derived from S3 in both roots and shoots ( Fig. 3b). No obvious effects on plant growth and development were observed on OsCDPK7 over-expression under the normal growth conditions. On the other hand, the protein level in S27 was much lower than that in NT ( Fig. 3b). Since the protein levels were similar between homozygous and heterozygous plants in each line (data not shown), both types of T1 transformants were combined and then used for further analyses. Significant variations in the protein level in each line were not detected even under cold- and salt-stress conditions (data not shown), when the transcript levels were increased (also see Fig. 5).
Stress tolerance of transgenic OsCDPK7 rice plants
To examine cold tolerance, 10-day-old T1 seedlings were exposed to 4°C for 24 h, and then returned to the normal growth conditions to allow their recovery. The extent of cold tolerance correlated well with the level of OsCDPK7 expression ( Fig. 4a). The elevated tolerance of the OsCDPK7-over-expressing plants was confirmed by measuring the changes in the chlorophyll fluorescence yield in the youngest extended leaf of each plant. The Fv/Fm values recovered to nearly normal levels in S3 and S1 plants 48 h after cold treatment ( Fig. 4b). In contrast, the values progressively decreased in the segregated NT plants derived from S3, their leaves showing prominent chlorosis and wilting. In another independent experiment, similar results were obtained ( Table 1).
Table 1. Cold, salt and drought stress tolerance of the wild-type (WT) and transgenic OsCDPK7 rice plants
a Numbers of WT and T1 plants, the youngest leaves of which wilted 3 days after cold stress (4°C for 24 h). b Numbers of WT and T1 plants, the youngest leaves of which wilted 3 days after salt stress (200 m m NaCl for 24 h). cNumbers of wilting T2 plants 5 days after drought stress (without water supply for 3 days). dGenotypes showing a statistically significant difference from the NT plants (χ2 test, P < 0.005).
Next, the over-expressing plants also showed an increased tolerance to salt stress ( Fig. 4c). In more than half of the NT-S3 plants and untransformed wild-type plants, the youngest leaves wilted 3 days after salt stress. On the other hand, S3 and S1 plants exhibited greater tolerance (statistically significant) ( Table 1). In addition, 13-day-old T2 seedlings of S1 plants showed increased drought tolerance with statistical significance ( Fig. 4d and Table 1). All the NT plants wilted 5 days after drought stress, whereas more than half of the S1 plants did not.
To determine whether or not a decrease in OsCDPK7 expression lowers the stress tolerance, we compared the number of plants whose leaves wilted under milder conditions between the OsCDPK7-suppressed plants (S27) and the segregated NT plants derived from S27. In S27 plants, the tolerance to both of these stresses was significantly decreased as follows. After milder cold stress, wilting was observed in 18/33 (55%) S27 plants and 3/9 (33%) NT plants. After milder salt stress, the wilting ratios of S27 and NT plants were 9/35 (26%) and 0/12 (0%), respectively. Taken together, we conclude that OsCDPK7 is a positive regulator commonly involved in the adaptation to at least three distinct stress agents, cold, salt and drought.
Expression of stress-inducible genes in transgenic OsCDPK7 plants
The induction of numerous stress-responsive genes is a hallmark of stress adaptation in plants ( Shinozaki & Yamaguchi-Shinozaki, 1997; Thomashow, 1999; Zhu et al. 1997 ). To elucidate further the role of OsCDPK7 in stress tolerance, we examined the effects of OsCDPK7 over-expression on the transcript levels of several stress-inducible rice genes –rab16A ( Mundy & Chua, 1988), salT ( Claes et al. 1990 ) and wsi18 ( Takahashi et al. 1994 ) – that encode a group 2 late-embryogenesis-abundant (LEA) protein, a glycine-rich protein, and a group 3 LEA protein, respectively. Since it is thought that similar mechanisms underlie the adaptation to water deficit caused by drought and high salinity ( Bray, 1997; Zhu et al. 1997 ), we analysed the expression of these genes in roots only under cold and salt stresses.
In both roots and shoots of the OsCDPK7-over-expressing plants (S3 and S1), the OsCDPK7 transcript was accumulated at high concentrations. The level was increased further under both cold and salt stresses, whereas it remained almost constant in the presence of exogeneous ABA or under drought stress conditions ( Fig. 5). Notably, OsCDPK7 over-expression did not induce the above stress-inducible genes under the normal growth conditions ( Fig. 5), in contrast to resukts obtained for the CBF1-and DREB1A-over-expressing Arabidopsis plants ( Jaglo-Ottosen et al. 1998 ; Liu et al. 1998 ). In roots of the over-expressing plants under salt stress, the transcript levels of all the stress-inducible genes were higher than the normally induced levels ( Fig. 5a). In shoots, only rab16A was highly induced by both salt and drought stress ( Fig. 5b,c). In contrast, salt induction was reduced in the suppressed plants ( Fig. 5a,b). The salt/drought stress tolerance of the over-expressing plants may be enhanced, at least in part, by the high-level accumulation of these gene products through the OsCDPK7 pathway. On the other hand, no induction of these genes was detected under cold stress. Thus, it was suggested that mechanisms of cold tolerance and salt/drought tolerance are different from each other, sharing OsCDPK7 as a common component. Moreover, it should be noted that the induction of rab16A and salT by ABA was increased or decreased by OsCDPK7 over-expression or suppression, respectively, which indicates that OsCDPK7 is also involved in ABA-dependent pathways.
In addition to the above-mentioned genes, the salt/ABA induction but not cold induction of rab16B ( Yamaguchi-Shinozaki et al. 1989 ) and oslea3 ( Moons et al. 1997 ), both of which encode LEA proteins, was found to be increased in roots of the over-expressing plants (data not shown). The expression of the following rice cold-responsive genes was also examined: alcohol dehydrogenase-1 ( Xie & Wu, 1989), lip5, lip9', and lip19 ( Aguan et al. 1991 ), Δ1-pyrroline-5-carboxylate synthetase ( Igarashi et al. 1997 ) and glutathione reductase-2 ( Kaminaka et al. 1998 ). However, they exhibited no correlated induction with the OsCDPK7 expression levels (data not shown). Thus, we have not yet identified genes induced only by cold stress. Nevertheless, it still remains possible that currently unknown genes are regulated by the OsCDPK7 pathways under cold stress.
Considering its importance as the most major crop in the world, a better understanding of stress signalling in rice would undoubtedly have an enormous impact. Here we characterized a rice CDPK, OsCDPK7, the mRNA level of which is increased under cold- and salt-stress conditions in both shoots and roots. Database searches revealed the presence of genes encoding this type of stress-inducible CDPK in both monocots and dicots. OsCDPK7 belongs to a subclass of stress-inducible CDPKs, which is conserved throughout higher plants but distinct from that of ATCDPK1, a putative stress-signalling isoform ( Sheen, 1996; Urao et al. 1994 ). The difference in substrate specificity also indicated that OsCDPK7 is involved in other signalling pathway(s) than that of ATCDPK1 under stress conditions.
Immunoblot analysis of the protein level of OsCDPK7 in rice plants with the isoform-specific antibodies detected no significant increase under the stress conditions, even when the mRNA level was elevated. There are other examples of putative stress-signalling mitogen-activated protein kinases (MAPKs) in plants in which transcriptional up-regulation is not correlated with an increase in the amount of protein ( Bogre et al. 1997 ; Seo et al. 1999 ). It may be speculated that the activated MAPK is degraded immediately after it has transduced signal(s), and the transcriptional up-regulation of the MAPK gene is to compensate for the loss of the MAPK protein ( Hirt, 1999). To know whether or not this is the case for OsCDPK7, it is important to examine turnover of the protein under the stress conditions. Since the OsCDPK7 protein is expressed at an almost constant level in the presence or absence of stress stimuli, there must be a post-translational mechanism(s) regulating the kinase activity in plant cells. At present, Ca2+ is the only known regulator of the activity of CDPKs. However, it seems likely that additional unknown mechanism(s) may be involved in control of the OsCDPK7 activity under stress conditions, taking into account the broad substrate specificity of this kinase. In this regard, recent studies suggested that some of the isoforms might be regulated by interaction with other proteins, e.g. 14-3-3 proteins ( Camoni et al. 1998b ; Moorhead et al. 1999 ). Even though neither activation of the OsCDPK7 enzyme upon stress nor protein–protein interaction was detected, the possibility that such regulation mechanisms are involved in vivo cannot be ruled out at present. Moreover, it is still possible that changes in protein localization could be important in the regulation of OsCDPK7.
We then carried out a functional analysis of the role of OsCDPK7 in stress tolerance of rice plants, using transgenic plants with altered levels of the OsCDPK7 protein. The mRNA derived from the endogenous promoter and that derived from the CaMV 35S promoter might differ in their stability, because the transgene is truncated in the 3′-non-coding region. However, it seems likely that the same intact protein is translated from both types of the mRNA, since Western blot analyses detected a single band in each lane of the over-expressing plants ( Fig. 3b). Over-expression of OsCDPK7 conferred both cold and salt/drought tolerance on rice plants. In contrast, suppression of OsCDPK7 expression lowered the stress tolerance. Therefore, OsCDPK7 plays key roles in the tolerance to the two types of stress in rice. To our knowledge, this is the first demonstration of the physiological role of a CDPK isoform at the whole-plant level.
The enhanced salt/drought induction of the genes for LEA proteins by OsCDPK7 over-expression appeared to contribute, at least in part, to the improved salt/drought tolerance in rice plants. Notably, the over-expression of a barley (Hordeum vulgare L.) group 3 LEA protein, HVA1, conferred both salt and drought stress tolerance on transgenic rice plants ( Xu et al. 1996 ). Although OsCDPK7 over-expression also elevated the ABA-induced levels of the above genes, whether or not the salt induction of these genes depends on ABA remains to be examined.
In contrast to salt/drought stress, no induction of these genes was observed under cold stress or in the absence of a stress stimulus. Although little is known about the OsCDPK7-mediated cold signalling pathway, we suggest that OsCDPK7 promotes cold and salt/drought tolerance through distinct pathways. Moreover, it seems likely that OsCDPK7 is kept normally inactive, since there is no constitutive induction of the above stress-inducible genes upon OsCDPK7 over-expression. Thus, OsCDPK7 over-production only is insufficient to trigger the downstream signalling, and stress stimuli must be required to activate this CDPK. Consistent with this speculation, no significant effects were observed with regard to development, growth and fertility in over-expressing plants grown in a greenhouse (unpublished results). This is very favourable for crop improvement. For instance, DREB1A-over-expressing plants showed severe growth retardation under normal growth conditions, presumably because of the constitutive high-level expression of stress-inducible genes ( Liu et al. 1998 ). In contrast to this transcription factor, it appears that the activity of OsCDPK7 is under a stringent post-translational control in rice cells.
Finally, we propose a model for the OsCDPK7 signalling pathway under the above stress conditions, in which the amount of activated OsCDPK7 determines the transduction current ( Fig. 6). The results also suggest that OsCDPK7 acts at one of the branch points of stress signal transduction in rice. Nevertheless, there seems little or no cross-talk downstream of these CDPK pathways even when each signal is amplified by OsCDPK7 over-expression. There must be unknown mechanisms that maintain the signalling specificity. Future analyses, especially on protein–protein interactions and protein localization, are awaited to verify this model. The present work provides new approaches for engineering of crop plants with improved stress tolerance, as well as for understanding the principles governing CDPK-mediated stress signal transduction.
Using a cDNA for ZmCDPK7 as a probe, cDNA cloning of OsCDPK7 was carried out as described by Saijo et al. (1997) from a rice (Oryza sativa Nipponbare) root cDNA library ( Hata et al. 1997 ). Two positive clones were isolated, and both of them were confirmed to encode OsCDPK7 by sequencing. Clones 1 and 2 correspond to nt 1–1970 and nt 453–2126, respectively, of the composite cDNA sequence for OsCDPK7.
To prepare anti-OsCDPK7 antibodies, a cDNA fragment encoding the C-terminal portion (amino acid residues 449–551) of OsCDPK7 was subcloned into the pET32a vector (Novagen), and then transformed into E. coli BL21 (DE3). The expressed recombinant protein was purified on a histidine tag affinity column (Novagen), and then the peptide tags on the N-terminal side were removed by enterokinase digestion. Polyclonal rabbit antibodies raised against the antigen were purified using a column of Sepharose 4B (Amersham-Pharmacia) coupled with the GST–OsCDPK7 fusion protein. The purified antibodies did not cross-react with another CDPK isoform, ZmCDPK9, extracted from E. coli (BL21) cells harbouring pETPK9 ( Saijo et al. 1997 ).
Preparation of protein extracts and immunoblot analysis
Plant tissues were ground in liquid nitrogen and then homogenized in an extraction buffer (50 m m Tris–HCl (pH 7.6), 2 m m EDTA, 1 m m MgCl2, 2 m m DTT, 1 m m NaF, 10 m mβ-glycerophosphate, 0.1 m m Na3VO4, 1 m m phenylmethanesulphonyl fluoride, one tablet per 50 ml protease inhibitor cocktail (Boehringer), 10% (v/v) polyvinylpolypyrrolidone). The homogenate was centrifuged first at 10 000 g for 10 min, and then at 100 000 g for 45 min. The supernatant was transferred into clean tubes, immediately frozen in liquid nitrogen, and then stored at −80°C. For immunoblot analysis, the supernatant was desalted by passing it through a Sephadex G-25 column (Amersham-Pharmacia) equilibrated with 50 m m Tris–HCl (pH 7.6), containing 1 m m DTT. Western blot analysis was performed essentially as described by Ueno et al. (2000) with anti-OsCDPK7 antibodies. The signals were then detected by an enhanced chemiluminescence system (Boehringer).
A full-length cDNA clone (nt. 1–1970) for OsCDPK7 was introduced into a Ti-based vector, pMSH1 ( Kawasaki et al. 1999 ), in the sense orientation downstream of the CaMV 35S promoter. The construct was introduced into rice calli (Oryza sativa cultivar Notohikari) by means of Agrobacterium-mediated transformation, according to the published protocol ( Hiei et al. 1994 ). Transformed calli were selected for hygromycin resistance, and then transgenic plants were regenerated.
Estimation of stress tolerance of transgenic rice plants
In a growth chamber (16 h light/8 h darkness, 28°C), sterilized seeds were germinated on Murashige and Skoog agar medium ( Murashige & Skoog, 1962) for 5 days, and then transplanted in soil. For cold-stress treatment, 10-day-old seedlings were exposed to 4°C for 24 h under continuous light conditions, and then returned to the normal growth conditions. Stress treatments under milder conditions were conducted as follows. Ten-day-old seedlings were first incubated at 15°C for 24 h, then at 4°C for 24 h, and finally returned to the normal conditions. For salt-stress treatment, 10-day-old seedlings were transferred to a nutrient solution, 0.1% (v/v) Hyponex (Hyponex Japan), containing 200 m m NaCl or 150 m m NaCl for the milder treatments, and then incubated for 24 h under the normal light/dark cycle at 28°C. Immediately after the salt stress, the roots of the plants were rinsed with water, and then hydroponically grown in a fresh nutrient solution without NaCl. Finally, in order to determine whether or not each T1 plant has the transgene, we amplified a portion of OsCDPK7 (nt 661–945) from the genomic DNA of each plant by means of PCR, using 5′-ACATCGTCATGGA-GCTCTGCGCC-3′ and 5′GAGCTACGTAATATGGGCTTCCG-3′ as primers. A 285 bp fragment without intron sequences was amplified from each transgenic plant. The plants of each T1 line examined included both the homozygous and heterozygous plants. For drought-stress treatment, 13-day-old T2 seedlings of the plants homozygous for the transgene (S1) and those of non-transgenic plants (NT) had water withheld for 3 days, and were then irrigated normally for 5 days.
Determination of chlorophyll fluorescence
Measurement of chlorophyll fluorescence was performed with a pulse-amplitude modulation fluorometer (PAM-2000; Walz, Effeltrich, Germany). Fluorescence signals from the youngest extended leaf of each rice plant, which had been dark-adapted for 15 min, were measured at the indicated times. The ratio of Fv to Fm (Fv/Fm) representing the activity of photosystem II was used to assess the functional damage to the plants ( Genty et al. 1989 ).
We wish to thank Dr T. Kusano, Dr T. Izawa and Dr T. Kawasaki (Nara Institute of Science and Technology) for their helpful discussions and encouragement, and Dr T. Endo (Kyoto University) for help in the measurement of chlorophyll fluorescence. This work was supported by Grants-in-Aid for Scientific Research on Priority Areas (Nos. 10170217 and 11151217) from the Ministry of Education, Science, Sports and Culture of Japan, and the program ‘Research for the Future’ of the Japan Society for the Promotion of Science (JSPS) (JSPS-RFTF96L00604). Y.S. was supported by a Research Fellowship from the JSPS for Young Scientists.