Raffinose family oligosaccharides (RFO) accumulating during seed development are thought to play a role in the desiccation tolerance of seeds. However, the functions of RFO in desiccation tolerance have not been elucidated. Here we examine the functions of RFO in Arabidopsis thaliana plants under drought- and cold-stress conditions, based on the analyses of function and expression of genes involved in RFO biosynthesis. Sugar analysis showed that drought-, high salinity- and cold-treated Arabidopsis plants accumulate a large amount of raffinose and galactinol, but not stachyose. Raffinose and galactinol were not detected in unstressed plants. This suggests that raffinose and galactinol are involved in tolerance to drought, high salinity and cold stresses. Galactinol synthase (GolS) catalyses the first step in the biosynthesis of RFO from UDP-galactose. We identified three stress-responsive GolS genes (AtGolS1, 2 and 3) among seven Arabidopsis GolS genes. AtGolS1 and 2 were induced by drought and high-salinity stresses, but not by cold stress. By contrast, AtGolS3 was induced by cold stress but not by drought or salt stress. All the GST fusion proteins of GST-AtGolS1, 2 and 3 expressed in Escherichia coli had galactinol synthase activities. Overexpression of AtGolS2 in transgenic Arabidopsis caused an increase in endogenous galactinol and raffinose, and showed reduced transpiration from leaves to improve drought tolerance. These results show that stress-inducible galactinol synthase plays a key role in the accumulation of galactinol and raffinose under abiotic stress conditions, and that galactinol and raffinose may function as osmoprotectants in drought-stress tolerance of plants.
Plants respond to water deficit and adapt to drought conditions through various physiological and biochemical changes, including transition in gene expression during water deficit. The mechanisms of drought-stress response have been investigated most extensively in a model plant, Arabidopsis thaliana, and in other plants including crops and resurrection plants (Bray, 1997; Ingram and Bartels, 1996; Shinozaki and Yamaguchi-Shinozaki, 1997; Shinozaki and Yamaguchi-Shinozaki, 1999). The drought-stress signal is mediated through both abscisic acid (ABA)-dependent and ABA-independent pathways to regulate expression of various genes that are involved in drought-stress tolerance and response. These gene products are thought to function in the accumulation of osmoprotectants; detoxification protection of cells; protein turnover; stress-signalling pathways; transcriptional regulation, and so on (Bohnert et al., 1995; Bray, 1997; Ingram and Bartels, 1996; Shinozaki and Yamaguchi-Shinozaki, 1997; Shinozaki and Yamaguchi-Shinozaki, 1999). In particular, osmoprotectants such as proline, glycinebetaine and mannitol have been shown to be important for the improvement of stress tolerance in plants by manipulation of genes encoding key enzymes of the osmolyte synthesis or degradation pathway (Kavi-Kishor et al., 1995; Nanjo et al., 1999a; Nanjo et al., 1999b; Tarczynski et al., 1993; Thomas et al., 1995). Raffinose family oligosaccharides (RFOs) such as raffinose and stachyose accumulate during seed development and are thought to play some role in desiccation tolerance of seeds. During seed development, RFOs accumulate at the late stage of soyabean seed maturation and desiccation (Castillo et al., 1990; Saravitz et al., 1987). In maize, raffinose accumulates during the seed desiccation process and is thought to function in stress tolerance, whereas sucrose accumulates independently of desiccation tolerance. Desiccation tolerance of seeds is not achieved in the absence of raffinose accumulation (Brenac et al., 1997). These results suggest that the ratio of sucrose to RFO is critical for desiccation tolerance of seeds, rather than the total amount of sugars. Although young, excised soyabean seeds are not tolerant to desiccation, slow dehydration induces stress tolerance, which is strongly correlated with a significant increase in stachyose content (Blackman et al., 1992).
Figure 1(a) shows the metabolic pathway of RFOs in plants. Galactinol synthase (GolS) catalyses the first committed step in the biosynthesis of RFOs and plays a key regulatory role in carbon partitioning between sucrose and RFOs (Saravitz et al., 1987). Therefore GolS potentially catalyses a metabolic key step, and its gene provides an experimental tool to manipulate the level of RFOs in seeds or vegetative tissues to analyse the function of RFOs as osmoprotectants. The GolS activity in kidney bean seeds increases on exposure of plants to cold, and the expression of GolS genes is induced by cold stress in Arabidopsis and Ajuga reptans plants (Liu, J. J. et al., 1998; Sprenger and Keller, 2000). During cold acclimation, RFOs accumulate in leaves of A. reptans (Bachmann et al., 1994). However, functions of RFOs in stress tolerance, and roles of GolS in RFO biosynthesis, have not been elucidated.
At least seven GolS-related genes are found in the Arabidopsis genome, but little is known about their roles in the accumulation of galactinol and raffinose in plants under water-deficit conditions. In the present study we report that three Arabidopsis GolS genes, AtGolS1, 2 and 3, are differentially induced by drought, low temperature, high salinity and ABA. We showed that their gene products have galactinol synthase activity. Furthermore, transgenic plants that overexpressed the AtGolS cDNA showed improved drought tolerance. This study provides a direct evidence that the stress-inducible GolS gene controls the level of RFOs, and that galactinol and raffinose play important roles in drought-stress tolerance.
Accumulation of galactinol and RFOs during water-deficit stress in Arabidopsis
We measured the accumulation of endogenous levels of galactinol and RFOs in Arabidopsis reproductive organs or soil-grown Arabidopsis plants during dehydration, high salinity and cold stresses. As shown in Figure 1(b), galactinol, raffinose and stachyose accumulated only in seeds in reproductive organs. Sucrose is the first major sugar in Arabidopsis mature seeds, and stachyose and raffinose are the second major sugars. This suggests that galactinol, raffinose and stachyose act as osmoprotectants in Arabidopsis seeds as in other plants such as soyabean (Glycine max) (Blackman et al., 1992). So the sugar content was also measured in Arabidopsis plants that had been treated with dehydration, high salinity (150 mm NaCl) or cold stress, and in unstressed plants (water) as a control (Figure 1c). While the unstressed plants had no detectable amount of RFOs, galactinol and raffinose accumulated in all the stress-treated plants. O-methyl-inositol, which functions in Coleus as an osmoprotectant, was not detected under these stresses in Arabidopsis (Pattanagul and Madore, 1999). These results imply that galactinol and raffinose function as an osmoprotectant in plants under osmotic stresses. However, stachyose, the first major RFO in seeds, did not accumulate in these stressed plants.
Isolation and characterization of GolS genes from Arabidopsis
Seven GolS-related genes were identified from the A. thaliana (Columbia) genome databases (AtDB, Stanford University; http://genome-http://www.stanford.edu/Arabidopsis/) by a homology search against the Oryza sativa GolS gene homologue, wsi76 (Takahashi et al., 1994), using the blast program. Their GenBank accession numbers are AC002337, AC009323, AC003970, AC002292, AB005244, AL049171/AL161564and AC004473. We named these genes AtGolS1, 2, 3, 4, 5, 6 and 7, respectively. The putative GolS genes were amplified from the Arabidopsis full-length cDNA library (Seki et al., 1998) by PCR using the synthetic oligonucleotide sets (see Experimental procedures for details), and into the pBluescript vector. To investigate the stress-inducibility and tissue-specific expression of the seven AtGolS genes, we carried out the Northern blot analysis using Arabidopsis plants grown on agar plates (Figure 2a). Among seven AtGolS genes, three genes were induced by osmotic stress. These cDNAs were isolated by screening the Arabidopsis full-length cDNA library using the above-mentioned PCR-amplified cDNA fragments as probes and named AtGolS1, 2 and 3, respectively. Phylogenic analysis revealed that AtGolS1, 2 and 3 are highly homologous with each other (Figure 2c), and are all highly homologous to known GolS genes from A. reptans (Sprenger and Keller, 2000); soyabean (Kerr et al., 1993); zucchini (Kerr et al., 1993); and rice (Takahashi et al., 1994). All of the reported GolS and AtGolS1 proteins have a putative serine phosphorylation site at position 270, but AtGolS2 and 3 do not. There is a characteristic hydrophobic pentapeptide (APSAA) at the carboxy-terminal end of all known GolS.
Expression profiles of these AtGolS genes, using the same plant materials as used for the sugar analysis, are shown in Figure 2(d). AtGolS1 and 2 were induced by drought and high-salinity stresses but not by cold stress. By contrast, AtGolS3 was induced by cold stress but not by drought or high-salinity stress. Although AtGolS1 and 2 showed similar patterns of gene expression, the level of AtGolS2 mRNA was higher than that of AtGolS1 mRNA. In mature seeds, a high level of expression of AtGolS1 and 2 was observed, whereas the expression of AtGolS3 was hardly observed.
Enzymatic properties of the bacterially expressed AtGolS proteins
To examine whether the AtGolS1, 2 and 3 genes encode a galactinol synthase, we analysed the biochemical properties of the recombinant AtGolS1, 2 and 3 proteins expressed in Escherichia coli. The DNA fragments for the AtGolS1, 2 and 3 coding regions were amplified by PCR and fused to the GST (glutathione S-transferase) gene in-frame using the pGEX4T-1 (Amersham Pharmacia Biotech, Uppsala, Sweden) to construct a chimeric plasmid pGST-AtGolS1, 2 and 3, respectively. The GST-AtGolS1, 2 and 3 proteins were overexpressed in E. coli and purified from the crude cell extract using a glutathione-Sepharose 4B.
We then examined whether the purified GST-AtGolS1, 2 and 3 recombinant proteins synthesize galactinol in the presence of myoinositol, a substrate of GolS. As shown in Figure 3, galactinol as well as myoinositol was detected after the incubation with GST-AtGolS1, 2 and 3 (Figure 3b–d), whereas only myoinositol was detected in the control (Figure 3a). These results show that AtGolS1, 2 and 3 encode galactinol synthase.
Expression analysis of the AtGolS genes in Arabidopsis plants under stress conditions
We analysed the expression of three AtGolS genes under various abiotic stress conditions using Arabidopsis plants grown on agar plates (Figure 4). AtGolS1 and 2 were induced by drought and high-salinity stresses, but not by cold stress, whereas AtGolS3 was strongly induced by cold stress, but not by drought or high-salinity stress. Although rice wsi76, a GolS homologue, is not induced by ABA (Takahashi et al., 1994), AtGolS1 and 2 were weakly induced by ABA but AtGolS3 was not.
AtGolS3 is controlled by DREB1A
Overexpression of transcription factors DREB1A/CBF3 (DREB: dehydration responsive-element binding factor; CBF: C-repeat binding factor) and CBF1 improves the tolerance to drought, high salinity and cold stresses (Jaglo-Ottosen et al., 1998; Kasuga et al., 1999; Liu, Q. et al., 1998), and causes accumulation of galactinol and raffinose in the transgenic plants (Gilmour et al., 2000; our unpublished data). The DREB1/CBF genes are rapidly induced in response to low temperature, and encode transcriptional activators that control the expression of stress-inducible genes containing DRE in their promoters. Overexpression of DREB1A/CBF3 in transgenic Arabidopsis plants has been shown to induce overexpression of many stress-inducible target genes under unstressed conditions, such as RD (responsive to dehydration), ERD (early responsive to dehydration), COR (cold-regulated), and KINgenes, and to enhance stress tolerance of the transgenic plants to drought, freezing and high-salinity stresses (Gilmour et al., 2000; Kasuga et al., 1999; Seki et al., 2001). Then the expression of AtGolS genes in DREB1A-overexpressing transgenic plants (35S::DREB1A plants) was examined. As shown in Figure 5(a), the AtGolS3 mRNA was significantly observed in the 35S::DREB1A plants under unstressed conditions, and accumulated greatly under cold-stress conditions. By contrast, AtGolS1 and 2 mRNAs were unchanged in the 35S::DREB1A plants and did not accumulate under stress conditions. These results suggest that AtGolS3 is a target gene of DREB1A, but the others are not. In the AtGolS3 promoter, we found two DRE (TACCGACAT) and two DRE-like A/GCCGAC core motifs (Yamaguchi-Shinozaki and Shinozaki, 1994) at −814 and −786 bp, and at −390 and −379 bp, respectively, upstream from the transcription start site (Figure 5b). These results suggest that the increase in raffinose levels of DREB1A-overexpressing plants (Gilmour et al., 2000) is due to the increased expression of AtGolS3, which may improve freezing tolerance of the transgenic plants. On the other hand, AtGolS1 and 2 are not controlled by DREB1A.
Creation of AtGolS-overexpressing transgenic plants and evaluation of their drought-stress tolerance
To further analyse the role of AtGolS in the accumulation of galactinol and raffinose under drought-stress conditions, we generated transgenic Arabidopsis plants that overexpressed sense AtGolS2 transcripts (Figure 6a). We selected two transgenic lines from overexpressors of AtGolS2 (AtGolS2-S8, S29) which showed constitutive strong expression of each AtGolS transgene (Figure 6b).
Then we measured galactinol and raffinose contents in these transgenic plants under unstressed conditions. Under normal growth conditions, galactinol and raffinose were not detected in wild-type plants by HPLC analysis. In contrast, each sense transgenic plant showed significant accumulation of galactinol and raffinose under normal growth conditions (Figure 6c).
To examine whether altered expression of AtGolS genes affected the tolerance to drought stress in the transgenic plants, the AtGolS2 sense transgenic plants (S8, S29) were grown for 3 weeks under normal conditions, and then exposed to drought stress by stopping water supply. At 14 days after drought-stress treatment, each AtGolS2 sense transgenic plant clearly showed stronger stress tolerance to drought than the wild-type plants. When plants were rehydrated, all the sense transgenic plants recovered, but none of the control plants survived (Figure 6d). These results indicate important roles of galactinol and raffinose in drought-stress tolerance, and a key role of AtGolS in the production of galactinol and raffinose.
Water content of the soil and transpiration rate of the transgenic plants
We also measured the water content of the soil in pot-grown plants during dehydration. As shown in Figure 7(a), the water content of the soil in the pots used for growth of these AtGolS sense transgenic plants was higher than that of wild-type plants. This suggests that the drought tolerance of AtGolS sense transgenic plants is due to less transpiration of their leaves.
Then we measured the transpiration from leaves of the AtGolS sense transgenic plants and wild-type plants that had been grown under normal conditions (Figure 7b). It became clear that the transpiration rate of the AtGolS sense transgenic plants was lower than that of wild-type plants.
Accumulation of galactinol and raffinose and expression of galactinol synthase genes during abiotic stress treatment
Accumulation of galactinol and RFO (raffinose and stachyose) was observed during seed maturation in Arabidopsis (Figure 1b), as in soyabean and maize (Blackman et al., 1992; Brenac et al., 1997). This suggests important roles of these sugars in the desiccation tolerance of mature seeds. We analysed these sugars in Arabidopsis rosette plants treated with drought, high-salinity and cold stresses, and showed that galactinol and raffinose, but not stachyose, accumulate in all the stress-treated plants. This suggests that galactinol and raffinose are involved in the stress tolerance in plants exposed to abiotic stress, as well as in mature seeds.
To examine the function of galactinol and raffinose in stress tolerance, seven genes for galactinol synthase (GolS), a key enzyme in RFO synthesis, were isolated based on the Arabidopsis genome sequence. Among these seven genes, three GolS genes were induced by drought, high-salinity or cold stress, and were named AtGolS1, 2 and 3 (Figure 2a). RNA gel-blot analysis revealed that AtGolS1 and 2 are induced by drought and high-salinity stresses, but not by cold stress; but AtGolS3 is induced by cold stress, but not by drought or high-salinity stresses (Figure 2d). This suggests that AtGolS1 and 2 mainly function in drought and high-salinity stress tolerance, whereas AtGolS3 functions in cold stress tolerance. These expression patterns resemble the relation between DREB1 and DREB2 which encode key transcription factors involved in cold and dehydration-responsive gene expression, respectively (Liu, Q. et al., 1998). The expression pattern of DREB1 is similar to that of AtGolS3, which is induced only by cold stress. In contrast, the expression pattern of DREB2 is similar to that of AtGolS1 and 2 that is induced by drought and high-salinity stresses, but not by cold stress (Liu, Q. et al., 1998). In fact, we showed that AtGolS3 is controlled by DREB1A (Figure 5). However, AtGolS1 and 2 are not controlled by DREB1A. In this study we identified AtGolS3 as a new target gene of DREB1A. Interestingly, AtGolS3 was induced by cold stress but not by drought stress (Figures 4 and 5). Most of the DREB target genes are induced by both drought and cold stresses. AtGolS3 is the first example of a DREB target gene that is induced only by cold stress. The AtGolS3 promoter contains four DRE related motifs and one ABA-responsive element (ABRE). Combinations of these cis-acting elements may be involved in cold-specific expression of AtGolS3. Another possibility is that some negative factor induced during dehydration stress may repress the drought-responsive expression of AtGolS3.
AtGolS1 and 2 were slightly induced by exogenous ABA treatment. The AtGolS1 and 2 promoters have two ABREs which may function in ABA-responsive expression, whereas the AtGolS3 promoter has one ABRE (Figure 5b). AtGolS3 is not induced by ABA, which indicates that one ABRE is not functional in the promoter.
Drought tolerance of AtGolS-overexpressing transgenic plants and a key role of AtGolS in the production of galactinol and raffinose
To examine the function of galactinol and raffinose as osmoprotectants in plants under water-deficit stresses, we created AtGolS2-overexpressing transgenic Arabidopsis plants. In the transgenic plants, not only was the expression of AtGolS genes increased, but also galactinol and raffinose were accumulated (Figures 6b,c). The leaves of the transgenic plants under normal growth condition contained both galactinol and raffinose at a concentration of 200–500 µg g−1 FW. This content was equivalent to that in the transgenic plants in which proline or glycinebetaine was superfluously accumulated (Hayashi et al., 1997; Nanjo et al., 1999b). These transgenic plants also showed drought-stress tolerance (Figure 6d). Thus we showed that galactinol synthase has a key role in controlling galactinol and raffinose in plants, and that galactinol and raffinose function in drought tolerance. Transgenic plants that overexpressed DREB1A/CBF3 are tolerant to dehydration and cold stress (Kasuga et al., 1999; Liu, Q. et al., 1998), and accumulate more galactinol and raffinose as well as other sugars in comparison with the wild-type plant (Gilmour et al., 2000; our unpublished data). The transgenic plant accumulates proline as well as galactinol and raffinose (Gilmour et al., 2000; our unpublished data). These results suggest that galactinol and raffinose function as osmoprotectants like proline.
We measured the water content in the soil in pot-grown plants under drought stress, and showed that the water content of AtGolS sense transgenic plants is higher than that of wild-type plants. The transpiration rate of the AtGolS sense transgenic leaves was significantly suppressed in comparison with that of vector control plants. It is well known that ABA suppresses transpiration of leaves by closing stomata; the synthesis of ABA may be activated by galactinol or raffinose, which causes stomata closure. Transgenic plants overexpressing AtNCED3 cDNA accumulate ABA, which causes stomatal closure making plants drought tolerant (Iuchi et al., 2001). The AtNCED3 gene is an Arabidopsis drought-inducible gene encoding 9-cis epoxycarotenoid dioxygenase (NCED), a key enzyme in ABA biosynthesis. The phenotype of the AtNCED3-overexpressing plants was similar to that of the AtGolS-overexpressing plants (data not shown).
Using isolated chloroplast membranes of spinach (Spinacia oleracea), raffinose was suggested to reduce the inactivation of electron (DCIP reduction) and cyclic photophosphorylation in photosynthesis under freezing, drought and high-temperature stresses (Santarius, 1973). Furthermore, Santarius and Milde (1977) reported accumulation of sucrose and raffinose in chloroplasts of frost-hardy leaves. Thylakoid membranes may be partially protected by raffinose in chloroplasts. This membrane stabilization depends on the concentration of sugars and their molecular size. The trisaccharide raffinose may be more effective in membrane stabilization than either disaccharide sucrose or monosaccharide glucose.
Plant material and growth conditions
Arabidopsis thaliana ecotype Columbia was used in this study. Wild-type seeds were sown in 9 cm plastic pots filled with a 1 : 1 mixture of perlite/vermiculite and watered with 1000-fold diluted Hyponex (Hyponex, Osaka, Japan). Transgenic plants were grown on germination medium (GM) agar plates (Valvekens et al., 1988) containing 30 mg l−1 hygromycin (GMH) or 30 mg l−1 kanamycin (GMK). Two-week-old plate-grown plants were transferred to soil pots. The plants were then grown under continuous illumination of approximately 2500 lux at 22°C.
Stress treatment for agar plate-grown plants
Three-week-old plate-grown plants were harvested from GM agar plates, then dehydrated on Whatman 3MM paper at room temperature and approximately 60% humidity under dim light. Plants subjected to treatment with ABA and salt stress were grown hydroponically in solution containing 100 mm ABA and 250 mm NaCl, respectively, under dim light. Cold treatment was conducted under dim light by exposure of plants grown at 22°C to a temperature of 4°C. In each case, plants were subjected to the stress treatments for various periods and were frozen in liquid nitrogen.
Stress treatment for soil-grown plants
Dehydration, salinity and cold stresses were imposed by transferring 3-week-old soil-grown plants to vats without sufficient water, to 1000-fold diluted Hyponex containing 150 mm NaCl, and to 4°C with 1000-fold diluted Hyponex under continuous illumination, respectively, for 0, 3, 7 and 10 days.
Isolation of galactinol synthase genes from Arabidopsis
Seven open reading frames that have sequence similarity with GolS genes were identified from the Arabidopsis thaliana ecotype Columbia databases (AtDB, Stanford University; http://genome-http://www.stanford.edu/Arabidopsis/) by a homology search against Oryza sativa GolS gene homologue wsi76 (Takahashi et al., 1994) using the blast program. Genomic DNA was isolated from Arabidopsis plants. Synthetic oligonucleotides for PCR based on the sequences in the DNA database were used to isolate GolS genes. The synthetic oligonucleotides were as follows: 5′-CAAGGATCCGCAGATCACGTGCTAATCAC-3′ and 5′-CAAGGATCCCCTGGCAATCAAGCAGCGGA-3′ (set A; accession number AC002337); 5′-CGCCACAGTACAAGATCGGTTA-3′ and 5′-CATGAAGAGGCGTATGCAGC-3′ (set B; accession number AC009323); 5′-CTTTCTCGGACAAGATGGCA-3′ and 5′-GTGTTGACAAGAACCTCGCT-3′ (set C; accession number AC003970); 5′-GGCCCCTGAGATTTCCGTAA-3′ and 5′-GCAGGAAAGTAGGTCATTTCTG-3′ (set D; accession number AC002292); 5′-GATTGTCGAGAAGAGGATCG-3′ and 5′-GGAAGGGGCAAGGCTTGTTA-3′ (set E; accession number AB005244); 5′-GGCTCAAATGTCGATGACCGTCGAGAAGAG-3′ and 5′-GCAGCGGAAGGGGCAAGACTAATAAGGAC-3′ (set F; accession numbers AL049171and AL161564); and 5′-CAAGGATCCAAGGTGCAGCAGGAGAGTAGG-3′ and 5′-CAAGGATCCTGCATCCGAGAAGGCTCCTAA-3′ (set G; accession number AC004473). DNA fragments of putative GolS genes were amplified from the Arabidopsis full-length cDNA library (Seki et al., 1998; Seki et al., 2001) by PCR using the synthetic oligonucleotide sets A–G, respectively. Their DNA fragments were cloned into the pBluescript II SK+ cloning vector (Stratagene, La Jolla, CA, USA). Three cDNAs for stress-inducible GolS genes were isolated by screening the Arabidopsis full-length cDNA library using the cDNA fragments that were PCR-amplified with the synthetic oligonucleotide sets A–C as probes, and named AtGolS1, 2 and 3, respectively.
Total RNA was isolated according to the method described by Nagy et al. (1988). Total RNA was fractionated in a 1% agarose gel containing formaldehyde and was blotted onto a nylon filter (Sambrook et al., 1989). The filters were hybridized in DIG Easy Hyb (Roche Diagnostics, Indianapolis, IN, USA) at 68°C with DIG-labelled fragments using DIG RNA Labeling Kit (SP6/T7) (Roche Diagnostics); the probes were PCR-amplified from the first-strand cDNAs by gene-specific primers. Following hybridization, the filters were washed twice in 2 × SSC, 0.1% SDS for 5 min at room temperature and twice in 0.5 × SSC, 0.1% SDS at 68°C for 15 min. The washed filters were developed with an immunostaining system using a DIG DNA detection kit (Roche Diagnostics).
In Northern blot analysis, filters were hybridized with [32P]-labelled fragments at 42°C, washed twice with 0.1 × SSC, 0.1% SDS at 65°C for 15 min, and autoradiographed.
Expression of GST-AtGolS gene fusion proteins
Full-length cDNA fragments of AtGolS1, 2 and 3 were amplified by PCR using synthetic oligo-DNAs 5′-CGCGGATCCATGGCTCCGGGGCTTACTCAAAC-3′ and 5′-CGCGGATCCCCACCGACAATTTTAACTCCTGG-3′ (AtGolS1); 5′-CGCGGATCCATGGCACCTGAGATCAATACC-3′ and 5′-CGCGGATCCGAGGCGTATGCAGCAACGAGC-3′ (AtGolS2); and 5′-CGCGGATCCATGGCACCTGAGATGAACAACAAGTTG-3′ and 5′-CGCGGATCCCTGGTGTTGACAAGAACCTCGCTC-3′ (AtGolS3). The PCR fragments were cloned into the EcoRV site of pBluescript II SK+. The fragments were isolated from the resultant plasmids by EcoRI digestion, and inserted into the EcoRI site of pGEX4T (Amersham Pharmacia Biotech AB, Uppsala, Sweden) to yield pGST–AtGolS1, 2 and 3. These chimeric genes were expressed in E. coli. We have previously described the procedures for the purification of the fusion protein in detail (Iuchi et al., 2000).
Assay of galactinol synthase activity
The enzyme activity of galactinol synthase was assayed as described (Liu et al., 1995). The reaction mixture (1 ml) contained buffer (50 mm Hepes–Na, 2 mm DTT pH 7.0), 4 mm MnCl2, 4 mm UDP-Gal, 20 mm myoinositol, 160 µg BSA and 0.017 mg enzyme purified as above. The enzyme was pre-incubated with the reaction mixture at 30°C for 15 min and then incubated with the substrates at 30°C for 30 min. The reaction was stopped by adding 2 ml ice-cold 100% ethanol and centrifuging in a microcentrifuge. The extract was dried up, and products were analysed by high-performance liquid chromatography (HPLC) using the sugar analysis system DX500 (CarboPac MA1, Sunnyvale, CA, USA) and a pulsed amperometry detector (Dionex Corporation, Sunnyvale, CA, USA).
Sugar analysis of plants
The plant cell extract was prepared from various plant tissues. Plant tissues were frozen with liquid nitrogen, crushed, added to 10 ml 80% ethanol preheated to 80°C, then boiled for 10 min at 90°C. Then the series of steps was repeated twice (a total of three rounds of these steps was performed). Next, sugars of the cell extracts were analysed by HPLC as described above.
Construction of transgenic plants
To generate transgenic plants with sense AtGolS cDNA, we constructed chimeric genes in which the coding sequences of the AtGolS2 cDNA were fused in a sense orientation between the cauliflower mosaic virus 35S promoter and the nos terminator sequence of the pBIG2113Not vector in which the kanamycin-resistance gene (NPT-II) was replaced with the hygromycin-resistance gene (HPT) of the pBE2113Not expression vector (Liu, Q. et al., 1998; Mitsuhara et al., 1996), and named the plasmids 35S-AtGolS2 (sense). We introduced these plasmids into wild-type Arabidopsis seedlings by Agrobacterium-mediated transformation. T2 seeds were used for subsequent experiments.
Measurement of leaf transpiration rate
The transpiration rate was measured in fully expanded leaves with a portable photosynthesis system (model LI-6400, Li-Cor, Lincoln, NE, USA) under the following conditions: 100 µmol m−2 sec−1, 350 p.p.m. CO2, 22°C and 70% relative humidity.
We thank Ms Keiko Maeda and Ms Setsuko Kawamura for their excellent technical assistance. This work was supported in part by the Special Coordination Fund of the Science and Technology Agency of the Japanese Government and by a Grant-in-Aid from the Ministry of Education, Science and Culture of Japan to K.S., and by the Program for Promotion of Basic Research Activities for Innovative Biosciences.