INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

The plasma membrane H⁺-ATPase generates an H⁺ electrochemical gradient across the membrane, providing a driving force for the uptake of various nutrients such as potassium, nitrate, sulfate, Suc, and amino acids, and regulates cytoplasmic pH in many cell types and tissues of plants (Serrano, 1989; Palmgren, 1991, 1998; Sussman, 1994; Michelet and Boutry, 1995; Sze et al., 1999). H⁺-ATPase is essential for plants, and regulatory mechanisms of this enzyme have been extensively investigated. The H⁺-ATPase activity is thought to be regulated by an autoinhibitory domain in the C-terminal region of the enzyme (Palmgren et al., 1990) and it has been suggested that the C-terminal function is modulated by the binding of the 14-3-3 protein using fungal toxin fusicoccin (FC; Baunsgaard et al., 1998; Fullone et al., 1998; Olsson et al., 1998; Chung et al., 1999; Fuglsang et al., 1999; Svennelid et al., 1999; Sze et al., 1999). In stomatal guard cells a recent study has indicated that the level of phosphorylation in the C terminus parallels the activity of H⁺-ATPase, and the 14-3-3 protein binds to the phosphorylated C terminus (Kinoshita and Shimazaki, 1999).

14-3-3 proteins were initially discovered as abundant soluble proteins within bovine brain tissue (Moore and Perez, 1967) and have been identified from various eukaryotic organs including insects, yeast, and plants (Ferl, 1996). There are at least 10 14-3-3 isoforms in Arabidopsis (Wu et al., 1997). In plants 14-3-3 proteins regulate the transcription by binding to the activators in the nucleus (de Vetten et al., 1992; Lu et al., 1992; Schultz et al., 1998), and they regulate metabolic enzymes such as nitrate reductase (NR; Bachmann et al., 1996a) and Suc phosphate synthase (SPS; Toroser and Huber, 1997). 14-3-3 proteins recognize the phosphorylated NR and SPS, and inhibit their activities (Bachmann et al., 1996b; Toroser et al., 1998). These 14-3-3-binding proteins possess the conserved binding motifs RSXpSXP and RXY/FXpSXP, where X represents any amino acid and pS represents phospho-Ser (Muslin et al., 1996; Yaffe et al., 1997). However, the H⁺-ATPase does not possess this conserved binding motif, and recently, the unique binding motif QQXYpTV was found at the extreme end of the C terminus, where pT represents phospho-Thr (Fuglsang et al., 1999; Svennelid et al., 1999).

Because 14-3-3 proteins possess highly conserved primary sequences and highly similar tertiary structures (Liu et al., 1995; Xiao et al., 1995), binding of 14-3-3 protein to target protein is suggested to be non-specific. Yeast and human 14-3-3 proteins actually bind to the Arabidopsis H⁺-ATPase homolog, AHA2, and all of the tested isoforms activate the enzyme (Baunsgaard et al., 1998). However, the binding activity varies among these isoforms. The different affinity and inactivation ability of 14-3-3 isoforms for NR have been reported in vitro (Bachmann et al., 1996b; Kanamaru et al., 1999). Immunological analysis has revealed that an isoform of a 14-3-3 protein is expressed in a tissue- and organelle-specific manner in the germination processes of barley embryos, although the other two isoforms are expressed throughout the embryo (Testerink et al., 1999). Promoter analysis of GF14 in Arabidopsis has revealed cell- and tissue-specific localization of the promoter activity, which is dependent on the plant maturation stage (Daugherty et al., 1996). Furthermore, subcellular localization of the specific 14-3-3 proteins in the nuclei of Arabidopsis and maize has been demonstrated (Bihn et al., 1997). These observations suggest that the specific isoform of 14-3-3 proteins may bind to the target protein in vivo and may exert the functional specificity of the 14-3-3 isoforms.

In this study we show that the 14-3-3 protein that binds to the plasma membrane H⁺-ATPase in guard cells is the isoform of vf14-3-3a, and we suggest that this isoform plays a key role in the regulation of the H⁺-ATPase in vivo.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Blue Light (BL) and FC Induce the Binding of 14-3-3 Protein to the H⁺-ATPase in Guard Cell Protoplasts (GCPs)

The plasma membrane H⁺-ATPase was immunoprecipitated from GCPs using the specific antibodies, and the immunoprecipitate was separated by SDS-PAGE. Silver staining of the immunoprecipitate revealed that a 32.5-kD protein was coprecipitated with the H⁺-ATPase. The amount of this protein was increased when GCPs were illuminated with BL or treated with FC, an activator of the plasma membrane H⁺-ATPase (Fig. 1A). A 32.5-kD protein was recognized as the 14-3-3 protein by the antibodies raised against the Arabidopsis 14-3-3 protein, GF14 (Fig. 1B, see "Materials and Methods").

View larger version (55K): [in this window] [in a new window]   Figure 1.   Coprecipitation of the 14-3-3 protein with H+-ATPase. GCPs were incubated under background red light (RL) for 30 min, and were then illuminated with BL for 2.5 min (BL) or incubated with 10 µM of FC for 5 min (FC). Reactions were terminated by adding the solubilizing medium after individual treatments. A, The H+-ATPase was immunoprecipitated using antibodies against the plasma membrane H+-ATPase. The immunoprecipitated products from 100 µg of GCP protein were loaded on each lane and separated by SDS-PAGE, followed by staining with silver. B, Western-blot analysis of the 14-3-3 protein using polyclonal antibodies against recombinant GF14 from Arabidopsis. Experiments repeated two times on different occasions gave the same results.

FC Induces the Binding of 14-3-3 Protein to the H⁺-ATPase in Mesophyll Cell Protoplasts (MCPs)

In GCPs, 14-3-3 protein bound to the H⁺-ATPase when it was phosphorylated, and the 14-3-3 protein was suggested to be specific to the 32.5-kD protein, as described above. However, it was not clear whether or not this binding was unique to guard cells. To answer this question, MCPs that showed similar transcription levels of each three vf14-3-3 isoforms were used (see Fig. 5). Western-blot analysis using antibody against GF14 showed the presence of 14-3-3 proteins with molecular masses of 33.3, 32.5, and 28.0 kD in MCPs (Fig. 2A), but only the 32.5-kD protein was found in GCPs. Since MCPs are likely to be insensitive to BL, FC was used to activate the plasma membrane H⁺-ATPase, and then the H⁺-ATPase in MCPs was immunoprecipitated. Western-blot analysis of the coprecipitated protein revealed only the 32.5-kD 14-3-3 protein, and showed that this protein was increased by FC (Fig. 2B), suggesting that the 32.5-kD 14-3-3 protein bound to the plasma membrane H⁺-ATPase in MCPs as well. Western blotting also revealed the presence of the immunoprecipitated 14-3-3 protein in GCPs in response to BL.

View larger version (43K): [in this window] [in a new window]   Figure 2.   Immunoprecipitation of the 14-3-3 protein with H+-ATPase in MCPs and GCPs. A, Western-blot analysis of the 14-3-3 protein in the cell extracts with the polyclonal antibodies against GF14. Twenty micrgrams of GCPs and MCPs were separated by SDS-PAGE. B, Western-blot analysis for the immunoprecipitated products using antibody against GF14. MCPs were treated with 10 µM of FC (FC) for 5 min, and then the H+-ATPase was immunoprecipitated. GCPs were illuminated with BL (BL), and then immunoprecipitated. The immunoprecipitants were separated by SDS-PAGE.

Mass Profiling of the 14-3-3 Protein Bound to the H⁺-ATPase

The 14-3-3 protein that bound to the plasma membrane H⁺-ATPase in vivo is likely to exist in GCPs and MCPs. To identify the isoform, the 14-3-3 protein isolated by SDS-PAGE from the immunoprecipitate was excised and subjected to trypsinolysis. The peptides were purified and concentrated by reverse-phase chromatography and analyzed by matrix-assisted laser-desorption ionization time of flight (MALDI-TOF) mass spectrometry. After the best possible spectrum was obtained, databases (National Center for Biotechnology Information non-redundant [NCBInr] and Genpept) were searched for the match (Fig. 3). Thirty-one masses were obtained, and nine masses matched with the predicted masses of trypsin-digested vf14-3-3a peptides cloned from broad bean (Saalbach et al., 1997; Table I; Fig. 4). Matched peptides covered 39% of the vf14-3-3a. The masses of other proteins were not sufficiently matched with the obtained masses. These results indicate that one major protein was vf14-3-3a.

View larger version (34K): [in this window] [in a new window]   Figure 3.   Mass profiling of the 14-3-3 protein bound to the H+-ATPase in the BL-illuminated GCPs. The immunoprecipitated 14-3-3 protein with H+-ATPase from BL-illuminated GCPs was isolated by SDS-PAGE, and collected. Peptides of the coprecipitated 14-3-3 protein were digested by trypsin and purified by reverse-phase chromatography on C18 microcolumns. The purified peptides were analyzed by MALDI-TOF mass spectrometry. The resulting mass spectrum was searched and analyzed on databases NCBInr and Genpept to identify the peptides. Nine peptides matched with masses from vf14-3-3a were indicated.

View larger version (55K): [in this window] [in a new window]   Figure 4.   Alignment of deduced amino acid sequences of vf14-3-3 proteins from broad bean guard cells. Dots indicate amino acids that are identical to the vf14-3-3a sequence, and dashes indicate gaps introduced to allow for optimal alignment of the sequences. Asterisks below sequences show identical amino acid residues among these sequences. Double solid lines above the vf14-3-3a sequence show the primer sites for degenerate PCR. Sequences analyzed by mass profiling are boxed in black.

Isolation of Four cDNAs Encoding 14-3-3 Proteins from Guard Cells

We isolated cDNAs encoding 14-3-3 proteins in guard cells using two degenerate oligonucleotide primers from conserved amino acid sequences (VAYKNV and IMQLLRDN). The primers were used for PCR of the first-strand cDNA template, which was made from broad bean GCPs. PCR-amplified fragments of about 600 bp were obtained, and 20 of these products were sequenced. The products consisted of three vf14-3-3 isoforms, vf14-3-3a, vf14-3-3c, and vf14-3-3d. vf14-3-3b could be obtained by use of its specific probe (Saalbach et al., 1997). To isolate the full-length cDNA, 5'-RACE and 3'-RACE were performed for vf14-3-3c and vf14-3-3d, since vf14-3-3a and vf14-3-3b had been isolated from developing cotyledon of broad bean (Saalbach et al., 1997). The full-lengths of vf14-3-3c and vf14-3-3d were sequenced (Fig. 4). vf14-3-3c was 1,067 bp in length and contained a 792-bp open reading frame that encoded a putative polypeptide of 263 amino acids with a predicted molecular mass of 29,682 D. vf14-3-3d was 1,164 bp in length and contained a 774-bp open reading frame that encoded a putative polypeptide of 257 amino acids with a predicted molecular mass of 29,160 D.

Deduced amino acid sequences encoded by vf14-3-3 isoforms were aligned and revealed 59% to 79% identity between each pair of isoforms (Fig. 4; Table II). N- and C-termini of deduced amino acid sequences revealed the unique sequence of each isoform. The vf14-3-3a, vf14-3-3b, vf14-3-3c, and vf14-3-3d proteins were compared with 10 isoforms of Arabidopsis 14-3-3 protein and showed the highest similarity to GF14, GF14µ, GF14, and GF14µ, respectively.

Northern Hybridization Analysis of vf14-3-3

The transcription levels of the four distinct vf14-3-3 isoforms, vf14-3-3a, vf14-3-3b, vf14-3-3c, and vf14-3-3d, were determined in GCPs, MCPs, leaves, and roots of broad bean. Probes for northern analysis showed negligible cross-hybridization among isoforms (data not shown). As shown in Figure 5A, each vf14-3-3 probe hybridized to a single mRNA band of about 1,200 bp in each lane. The transcription level of vf14-3-3a was the higher in GCPs than in other tissues, and the level in roots was also high. MCPs showed the lowest level of vf14-3-3a. The transcription level of vf14-3-3b was highest in roots and was also high in leaves and GCPs. The level of vf14-3-3c was slightly higher in leaves than in their tissues. The level of vf14-3-3d was slightly higher in GCPs and roots than in MCPs and leaves, but longer exposure time was required to detect the signal in these tissues. The transcription level of H⁺-ATPase was highest in GCPs and lowest in MCPs. Transcription profiles of the vf14-3-3a gene were very similar to those of the H⁺-ATPase gene.

View larger version (42K): [in this window] [in a new window]   Figure 5.   Northern-blot analysis of vf14-3-3 transcripts in GCPs, MCPs, leaves, and roots. A, Each lane contained an equal amount (20 µg) of total RNA isolated from GCPs, MCPs, leaves, and roots. These were hybridized with DIG-labeled probes corresponding to vf14-3-3a, vf14-3-3b, vf14-3-3c, vf14-3-3d, and H+-ATPase. B, Staining of gel with ethidium bromide was shown as a loading control. Experiments repeated two times on different occasions gave similar results.

Coprecipitation of vf14-3-3a with H⁺-ATPase

The 14-3-3 protein bound to the plasma membrane H⁺-ATPase, and the bound isoform was suggested to be a vf14-3-3a protein. To confirm this, recombinant vf14-3-3a and vf14-3-3b, major isoforms in guard cells, were expressed in Escherichia coli and purified. The bound 14-3-3 protein and recombinant vf14-3-3 proteins were subjected to SDS-PAGE (Fig. 6A). The 14-3-3 proteins bound to the H⁺-ATPase in the BL- and FC-treated GCPs migrated at the same position as recombinant vf14-3-3a, but at a position different from that for vf14-3-3b migration. Furthermore, the bound 14-3-3 proteins and recombinant vf14-3-3 proteins were digested with cyanogen bromide (CNBr), which cleaves the C terminus of Met residue. As shown in Figure 6B, the CNBr-digested 14-3-3 proteins bound to the H⁺-ATPase produced two major bands of 13.0 and 9.6 kD, and two minor bands of 7.0 and 2.2 kD, in the BL- and FC-treated GCPs, suggesting that the 14-3-3 proteins bound to the H⁺-ATPase stimulated by BL and FC are identical. The CNBr-digested recombinant vf14-3-3a produced exactly the same peptide profile as shown above, but without the 7.0-kD peptide. This 7.0-kD band is probably due to a contaminant produced by the immunoprecipitation procedure, since the procedure without GCPs yielded the same product (data not shown). Digestion of recombinant vf14-3-3b with CNBr gave a profile completely different from that of the bound 14-3-3 protein. These results suggest that the 14-3-3 proteins bound to the plasma membrane H⁺-ATPase are vf14-3-3a.

View larger version (69K): [in this window] [in a new window]   Figure 6.   Comparison of immunoprecipitated 14-3-3 protein with recombinant vf14-3-3a and b. A, Silver staining of the immunoprecipitated 14-3-3 protein and recombinant vf14-3-3a and b. GCPs were incubated under RL for 30 min, and were then illuminated with BL for 2.5 min or treated with 10 µM of FC for 5 min. GCPs were immunoprecipitated using antibodies against H+-ATPase. One hundred micrograms of GCP protein was used for the immunoprecipitation in each lane (BL and FC). Fifty nanograms of recombinant vf14-3-3a and b were loaded on each lane. B, Silver staining of the CNBr-digested 14-3-3 proteins coprecipitated with H+-ATPase and recombinant vf14-3-3s. One and one-half milligrams of immunoprecipitated products and 600 ng of recombinant vf14-3-3s were separated by SDS-PAGE, and were then transferred to nitrocellulose membrane. The 14-3-3 proteins were excised from the membrane and treated with CNBr. The peptides were separated by Tricine SDS-PAGE and stained with silver.

Binding Affinity of Recombinant 14-3-3 Protein to the H⁺-ATPase

Because the 14-3-3 protein homologs have been indicated to have different affinity for NR (Bachmann et al., 1996; Kanamaru et al., 1999), the binding of vf14-3-3a and vf14-3-3b to the plasma membrane H⁺-ATPase was investigated. GCPs were illuminated with BL to phosphorylate H⁺-ATPase, and the H⁺-ATPases were separated by SDS-PAGE and blotted to the membrane for Far western analysis. When 14-3-3 proteins at 0.1 µM were applied, the amount of bound 14-3-3 protein increased with the increase in the amount of H⁺-ATPase (Fig. 7). However, the amount of bound 14-3-3 protein was larger in vf14-3-3a than in vf14-3-3b. At 2.5 µg of GCPs, the binding amount was more than 2-fold larger in vf14-3-3a than in vf14-3-3b. These results suggest that vf14-3-3a possessed a higher binding affinity for the H⁺-ATPase than did vf14-3-3b.

View larger version (46K): [in this window] [in a new window]   Figure 7.   Binding of the recombinant vf14-3-3s to the phosphorylated H+-ATPase. A, Expression and purification of GST-vf14-3-3 fusion proteins in E. coli. The GST-vf14-3-3s were separated by SDS-PAGE and stained with silver. B, Far western-blot analysis for the H+-ATPase. GCPs were incubated under background RL for 30 min, and were then illuminated with BL for 2.5 min. GCPs at the indicated amounts were subjected to SDS-PAGE, and Far western blot-analysis was conducted using 0.1 µM of GST-vf14-3-3a and b as probes. C, Western-blot analysis for the H+-ATPase. GCPs were loaded as shown in B. D, Relative amount of vf14-3-3s bound to the H+-ATPase. Binding levels were determined densitometrically.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

In this study we aimed to determine the 14-3-3 isoform that binds to the plasma membrane H⁺-ATPase in guard cells. It is advantageous to identify this isoform in the materials studied here because the affinity between the H⁺-ATPase and 14-3-3 protein is much higher than the affinity between proteins of other systems (Fuglsang et al., 1999). We can therefore easily obtain the 14-3-3 protein in its bound form to the target protein of the H⁺-ATPase (Baunsgaard et al., 1998; Kinoshita and Shimazaki, 1999). The 14-3-3 protein that binds to the H⁺-ATPase seems to be the only 32.5-kD 14-3-3 isoform in guard cells. In mesophyll cells, analysis of the coprecipitated proteins with the H⁺-ATPase revealed that a 32.5-kD 14-3-3 protein bound to the plasma membrane H⁺-ATPase upon its activation by FC, although the cells expressed at least three 14-3-3 isoproteins (Fig. 3). From these results we suspect that the specific 14-3-3 isoprotein binds to the plasma membrane H⁺-ATPase during its activation in vivo.

To test this hypothesis, we cloned several isoforms expressed in broad bean guard cells, and identified four cDNAs encoding 14-3-3 proteins. These were designated vf14-3-3a, vf14-3-3b, vf14-3-3c, and vf14-3-3d, according to the previous report (Saalbach et al., 1997). Among them, vf14-3-3a and b were most highly expressed in guard cells. The expression profile of vf14-3-3a in GCPs, MCPs, leaves, and roots was very similar to that of the plasma membrane H⁺-ATPase, suggesting a close relationship between vf14-3-3a and the H⁺-ATPase. We therefore expressed the recombinant proteins of vf14-3-3a and vf14-3-3b in E. coli, and subjected the proteins to SDS-PAGE to compare the molecular masses. The 14-3-3 protein that bound to the H⁺-ATPase migrated at the same position as that of recombinant vf14-3-3a, but not that of vf14-3-3b. Furthermore, digestion of the bound 14-3-3 protein with CNBr produced a peptide profile identical to that of vf14-3-3a. These results suggest that the bound 14-3-3 protein is vf14-3-3a. In accord with this notion, mass profiling of the proteins after trypsinolysis revealed that the masses of nine of the 31 peptides from the bound 14-3-3 protein matched with the predicted masses from vf14-3-3a peptides.

There are several to 10 isoforms of 14-3-3 proteins in plant cells. The 14-3-3 proteins are known to inhibit NR and SPS by binding to the consensus motif in the enzyme. However, there are several isoforms that bind to the targets and regulate the enzymes (Bachmann et al., 1996; Baunsgaard et al., 1998; Kanamaru et al., 1999). For example, three isoforms of Arabidopsis 14-3-3 proteins inactivate NR and bind to NR, but some isoforms show neither of these activities (Bachmann et al., 1996). These experiments were done in vitro and there is no evidence that the isoform specifically binds to the target enzymes in vivo. To our knowledge this is the first evidence that shows that specific isoforms bind to their target protein of plant cells in vivo.

The 14-3-3 proteins were initially found as FC-binding proteins in plants (Korthout and de Boer, 1994; Marra et al., 1994; Oecking et al., 1994). It was subsequently demonstrated that FC binds to the complex between the 14-3-3 protein and the C terminus of the plasma membrane H⁺-ATPase (Baunsgaard et al., 1998; Fullone et al., 1998; Sze et al., 1999). In the above works the partial amino acid sequence of 14-3-3 proteins in the plasma membrane bound to the H⁺-ATPase were determined in Commelina communis, oat, and maize. However, because there is no insufficient sequence data in these plant species, the isoform specificity could not be determined. Fullone et al. (1998) recently determined the amino acid sequence of the C terminus of the 14-3-3 protein that bound to the plasma membrane using FC receptor preparations. They indicated that two 14-3-3 protein species were present in the preparations and that one of them, the species having a molecular mass of 33 kD, was the isoform GF14-6 in maize. The other 14-3-3 isoform present in the preparations, which had a molecular mass of 31 kD, was not identified. The 33-kD isoform is likely to have been the 14-3-3 protein that binds to the plasma membrane H⁺-ATPase in the maize. However, it is still not clear whether the isoform binds to the H⁺-ATPase in vivo.

Northern analysis revealed that the expression of vf14-3-3a was higher in leaves than in MCPs, although the leaves consisted primarily of mesophyll cells. This difference might be attributed to the fact that vascular bundles were present in leaves and absent in MCPs. The companion cells in the phloem of the vascular system contains a high density of the plasma membrane H⁺-ATPase, which drives phloem loading in cooperation with Suc/H⁺ cotransporter (Dewitt and Sussmann, 1995; Moriau et al., 1999). Thus, the finding that vf14-3-3a was more highly expressed in leaves than in MCPs might indicate that vf14-3-3a coexisted with the plasma membrane H⁺-ATPase in companion cells. It was confirmed that the transcript of the H⁺-ATPase genes was more highly expressed in leaves than in MCPs (Fig. 5). The plasma membrane H⁺-ATPase is encoded in the multigene family, and the kinetic properties of the isoforms are distinct (Palmgren, 1998; Sze et al., 1999). Isoforms of the plasma membrane H⁺-ATPases of VHA1 and VHA2 are expressed in guard cells (Hentzen et al., 1996; Kinoshita and Shimazaki, 1999) and thus, vf14-3-3a is thought to bind to these isoforms. The mRNAs of VHA1 and VHA2 have been shown to be expressed in leaves, stems, and roots (Hentzen et al., 1996).

In the present study the recombinant vf14-3-3a and b could bind to the plasma membrane H⁺-ATPase in vitro (Fig. 7), although the affinity for the H⁺-ATPase was higher in vf14-3-3a than in vf14-3-3b. In contrast, vf14-3-3a was the only protein bound to the H⁺-ATPase in guard cells in vivo in response to BL and FC. One possible reason for this specific binding may be due to the predominant expression of the vf14-3-3a in guard cells (Figs. 2 and 5). Another interpretation is that vf14-3-3a plays a specific role in the activation of the plasma membrane H⁺-ATPase, and vf14-3-3a may be localized in the vicinity of this membrane. The latter hypothesis is supported by the result that only vf14-3-3a bound to the H⁺-ATPase in MCPs, although MCP expressed two other isoproteins of 14-3-3 (Fig. 2). In accord with this interpretation subcellular localization of the specific isoforms has been reported in the nuclei of Arabidopsis and maize (Bihn et al., 1997). Why, then, do guard cells prefer vf14-3-3a for the regulation of H⁺-ATPase? It is possible that vf14-3-3a is more efficient than vf14-3-3b at activating the plasma membrane H⁺-ATPase, since the former is more efficient at binding to H⁺-ATPase (Fig. 7).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Materials

Plants of broad bean (Vicia faba cv Ryosai Issun) were cultured hydroponically for 4 to 7 weeks in a greenhouse under a 14-h fluorescent light/10-h dark cycle at 22°C and a relative humidity of 50% to 60% (Shimazaki et al., 1992). Fully expanded second and third leaves were harvested.

Protoplast Preparation

GCPs were isolated enzymatically as described previously (Kinoshita and Shimazaki, 1999). Mesophyll protoplasts were also prepared according to the previous method (Shimazaki et al., 1982). Isolated GCPs were stored in 0.4 M mannitol and 1 mM CaCl₂, and the mesophyll cell protoplasts were stored in 0.6 M mannitol and 1 mM CaCl₂ on ice until use in the dark. Protein concentrations were determined by the method of Bradford (1976).

Immunoprecipitation

Immunoprecipitation using the antibodies against the H⁺-ATPase was done as described previously (Kinoshita and Shimazaki, 1999). GCP suspension of 0.5 to 1.0 mL (1.0 mg of protein mL¹) was incubated in 5 mM MES [2-(N-morpholino)-ethanesulfonic acid]-NaOH (pH 6.0), 0.4 M mannitol, 1 mM CaCl₂, and 10 mM KCl under background RL for 40 min at 24°C. Then the GCPs were illuminated with BL for 2.5 min or treated with 10 µM FC for 5 min. The reaction was terminated by disruption of GCPs with the addition of an equal volume of medium to the GCPs suspension. The medium contained 100 mM MOPS [3-(N-morpholino)-propanesulfonic acid]-KOH (pH 7.5), 5 mM EDTA, 200 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 20 µM leupeptin, 4 mM dithiothreitol (DTT), 20 mM NaF, 2 mM ammonium molybdate, 200 nM calyculin A, and 2% (w/v) Triton X-100. The GCP was centrifuged at 10,000g for 3 min. The resulting supernatant was mixed with antibodies against the plasma membrane H⁺-ATPase at 0.5% (v/v). After incubation for 12 h at 4°C, protein A-agarose (Santa Cruz Biotech, Santa Cruz, CA) was added to the supernatant at 2.5% (v/v), and was kept at 4°C for 12 h. The sample was then centrifuged at 10,000g for 3 min and the pellet was washed three times with 1 mL of ice-cold Tris-buffered saline (TBS). The obtained pellet was resuspended in the solubilizing medium (Laemmli, 1970) and was centrifuged at 10,000g for 1 min to remove agarose; the supernatant was then subjected to SDS-PAGE.

Western-Blot Analysis

14-3-3 proteins were detected immunologically with antibodies raised against recombinant GF14 from Arabidopsis according to the method of Gallagher et al. (1992) with slight modifications. Proteins of GCPs and MCPs were subjected to SDS-PAGE and transferred onto nitrocellulose membranes (Hybond-C, Amersham, Buckinghamshire, UK) using Trans-blot (Bio-Rad, Tokyo). The membranes were incubated in blocking medium containing 20 mM Tris-HCl (pH 7.4), 140 mM NaCl, 0.05% (w/v) Tween 20, and 5% (w/v) fatty-acid free milk at room temperature for 30 min, and were then reacted with polyclonal antibodies at 5,000-fold dilution at 4°C overnight. After the membrane was washed three times with T-TBS containing 20 mM Tris-HCl (pH 7.4), 140 mM NaCl, and 0.05% (w/v) Tween 20, it was reacted with alkaline phosphatase-conjugated goat anti-rabbit IgG antibodies (Bio-Rad) at 3,000-fold dilution for 1 h at room temperature. The alkaline phosphatase reaction was developed by 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium.

Far Western-Blot Analysis

Far western blotting was performed as described previously (Kinoshita and Shimazaki, 1999). Proteins were subjected to SDS-PAGE and were transferred onto nitrocellulose membranes. The membranes were then incubated at room temperature in a solution of 25 mM HEPES [4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid]-NaOH (pH 7.7), 25 mM NaCl, 5 mM MgCl₂, 1 mM DTT, and 6 M guanidine-HCl for 15 min, and then the concentration of guanidine-HCl was gradually decreased to 0.3 M. The membranes were incubated at room temperature in the solution of 25 mM HEPES-KOH (pH 7.7), 75 mM KCl, 0.1 mM EDTA, 1 mM DTT, 0.04% (w/v) Tween 20, and 2% (w/v) fatty-acid free milk for 1 h, and were then reacted with 0.1 µM glutathione S-transferase- (GST) 14-3-3 fusion protein for overnight at 4°C in the same medium. The membranes were washed three times with T-TBS, then were reacted with anti-GST antibodies (Pharmacia Biotech, Tokyo) at 5,000-fold dilution in the blocking medium. After three final washings in T-TBS, the membranes were reacted with anti-goat IgG antibodies (Sigma) at 5,000-fold dilution in the blocking medium.

Mass Profiling

Immunoprecipitates with antibody raised against H⁺-ATPase from 8 mg of protein of GCPs were separated by SDS-PAGE and 32.5-kD bands were excised. Proteins (32.5-kD) were extracted from gel pieces and were re-subjected to SDS-PAGE to concentrate. The 32.5-kD band was excised from the gel and subjected to mass profiling. In brief, the excised 32.5-kD band was extensively washed and dehydrated with ammonium bicarbonate/acetonitrile buffer. The band was rehydrated in 20 µg mL¹ trypsin solution and was incubated at 37°C for 12 h. Tryptic peptides were recovered by acetonitrile/trifluoroacetic acid extraction and were purified and concentrated by reverse-phase chromatography on C18 microcolumns. The peptides were mixed with matrix solution (-cyano-hydroxycinnamic acid) and MALDI plate. Mass spectrometry was done using Voyager DE PRO in reflective mode. After the best possible spectrum was obtained, the databases (NCBInr and Genpept) were searched for the matches.

Isolation and Sequencing of cDNA Clones

Two degenerate oligonucleotides (GTNGCNTAYAARAAYGT and CNCKNARNARYTGCATDAT) were deduced from the highly conserved amino acid sequences (VAYKNV and IMQLLRD, respectively) These oligonucleotides were used in degenerate PCR with the first-strand cDNAs as a template. The first-strand cDNAs were synthesized from total RNA of GCPs by avian myeloblastosis virus reverse transcriptase (Takara, Tokyo) using Oligo(dT) as a primer. Primers (ATGGCCACCGCACCAA and TAACTAATTAGCAGTCACACATTTT) for vf14-3-3a and (ATGGCTTCCACCAAGGAT and CACCAAACACCAGCCTC) for vf14-3-3b were generated on the basis of a full-length open reading frame (Saalbach et al., 1997) and were used for cloning of vf14-3-3a and vf14-3-3b, respectively. Primers (GCTGCACAGGATATTGCTGCTG and CAGTCTGACAAAGCTTGTGCCA) for vf14-3-3c and (GCTTCCACTGCTGCAGAG and TCCTGAAAGGGCCTGTCACC) for vf14-3-3d were used for the 3'-RACE. The 3'-RACE was performed according to the standard procedure of the 3'-RACE system (Gibco-BRL, Tokyo). Five primers (TGTCCAACAGCTCCAAAATACTAGC, CATGGTCATCATTCTTTCTACCTTCCTCC, and CGCCGCACGTAGCGAC) for vf14-3-3c and (CACATTCACCTCATTCCCTTTCGTC and CACGAGGCCCTACGTCC) for vf14-3-3d were used for the 5'-RACE. The 5'-RACE was performed according to the standard procedure of the 5'-RACE system, version 2.0 (Gibco-BRL). PCR products were cloned into a pCRII vector (Invitrogen, San Diego). Sequences were determined from both strands of the cDNA using an ALFred DNA sequencer (Pharmacia Biotech, Piscataway, NJ). Nucleotide and amino acid sequences were analyzed using the GENETYX software system (Software Development Co., Tokyo).

RNA Isolation and Northern Hybridization

RNA was isolated from GCPs, MCPs, leaves, and roots of broad bean with ISOGEN (Nippon Gene, Tokyo). Digoxigenin-labeled probes of vf14-3-3a and vf14-3-3b (1-786 bp), of vf14-3-3c (52 to 1,014 bp), and of vf14-3-3d (27 to 1,136 bp) were obtained by PCR using a PCR DIG labeling mix (Roche, Tokyo). Northern hybridization was using a Digoxigenin Luminescent Detection Kit (Roche) according to the manufacturer's instruction. Signals were detected using CDP-Star system (Roche).

Expression and Purification of vf14-3-3a and vf14-3-3b Proteins

The recombinant vf14-3-3a and vf14-3-3b proteins were expressed and purified from Escherichia coli (JM109). The coding sequences of vf14-3-3a and vf14-3-3b were amplified by PCR using synthetic oligonucleotides (CCGGATCCATGGCCACCGCAC and CCGGATCCTTACTGTGGTTCATCATTGC) and (CCGGATCCATGGCTTCCACCAAGG and CCGGATCCTCACTCTGCATCATCAC) that contained a BamHI site on one end, respectively, and cloned in frame with GST into the pGEX-2T plasmid vector (Pharmacia Biotech). Expression and purification were performed using Bulk and RediPack GST purification modules (Pharmacia Biotech) according to the manufacturers' instructions.

CNBr Digestion

The 14-3-3 protein bound to the H⁺-ATPase and the recombinant 14-3-3 proteins were digested by CNBr according to the previously described (Kinoshita and Shimazaki, 1999). The 14-3-3 proteins were separated by SDS-PAGE and transferred onto nitrocellulose membranes. After staining the proteins with 0.2% (w/v) Ponceau S, the band containing the 14-3-3 protein was excised, and was then digested by incubation with 100 mg mL¹ CNBr in 70% (v/v) formic acid at room temperature overnight with gentle rotation. The supernatant containing digested proteins was dried, and the dried peptides were dissolved in 200 µL of water, then dried again. The resulting peptides were subjected to Tricine SDS-PAGE according to the previously described (Shagger and Jagow, 1987) with slight modifications. The peptides were stained with silver.

Light Source

RL for background illumination was obtained from a tungsten lamp (EXR 150 W, Sylvania, Danvers, MA) by passing the light through a red glass filter (2-61, Corning, Corning, NY) and BL was obtained from a tungsten lamp (EXR 300W, Sylvania) through a blue glass filter (5-60, Corning). Photon flux densities of RL and BL were 600 and 150 µmol m² s¹, respectively, as determined with a quantum meter (model 185A, LI-COR, Lincoln, NE).