Phosphorylation-dependent interaction between plant plasma membrane H(+)-ATPase and 14-3-3 proteins.

The H(+)-ATPase is a key enzyme for the establishment and maintenance of plasma membrane potential and energization of secondary active transport in the plant cell. The phytotoxin fusicoccin induces H(+)-ATPase activation by promoting the association of 14-3-3 proteins. It is still unclear whether 14-3-3 proteins can represent natural regulators of the proton pump, and factors regulating 14-3-3 binding to the H(+)-ATPase under physiological conditions are unknown as well. In the present study in vivo and in vitro evidence is provided that 14-3-3 proteins can associate with the H(+)-ATPase from maize roots also in a fusicoccin-independent manner and that the interaction depends on the phosphorylation status of the proton pump. Furthermore, results indicate that phosphorylation of H(+)-ATPase influences also the fusicoccin-dependent interaction of 14-3-3 proteins. Finally, a protein phosphatase 2A able to impair the interaction between H(+)-ATPase and 14-3-3 proteins was identified and partially purified from maize root.

The H ؉ -ATPase is a key enzyme for the establishment and maintenance of plasma membrane potential and energization of secondary active transport in the plant cell. The phytotoxin fusicoccin induces H ؉ -ATPase activation by promoting the association of 14-3-3 proteins. It is still unclear whether 14-3-3 proteins can represent natural regulators of the proton pump, and factors regulating 14-3-3 binding to the H ؉ -ATPase under physiological conditions are unknown as well. In the present study in vivo and in vitro evidence is provided that 14-3-3 proteins can associate with the H ؉ -ATPase from maize roots also in a fusicoccin-independent manner and that the interaction depends on the phosphorylation status of the proton pump. Furthermore, results indicate that phosphorylation of H ؉ -ATPase influences also the fusicoccin-dependent interaction of 14-3-3 proteins. Finally, a protein phosphatase 2A able to impair the interaction between H ؉ -ATPase and 14-3-3 proteins was identified and partially purified from maize root.
The plant plasma membrane H ϩ -ATPase (1, 2) generates an electrochemical gradient across plant cell plasma membrane that provides the driving force for secondary active transport and for cell turgor maintenance. A number of central physiological processes, such as stomata opening, phloem loading, or root ion uptake, depend on regulation of H ϩ -ATPase activity. Despite a large body of evidence indicating that different stimuli, like hormones or light and stresses regulate these processes by affecting H ϩ -ATPase activity, up to now very little is known about molecular mechanisms controlling H ϩ -ATPase activity.
It was demonstrated that the C-terminal region of the enzyme is an autoinhibitory domain. In fact, proteolytic removal of this region (3) or heterologous expression of a C-terminal truncated enzyme results in an increase of H ϩ -ATPase activity (4). It was suggested, on the basis of similarity of biochemical parameters of activation, that H ϩ -ATPase stimulators such as lysophosphatidylcholine or the toxin from Fusicoccum amygdali, fusicoccin (FC), 1 act through C-terminal domain displacement (5-7), which therefore could represent a common mechanism for different proton pump effectors. It was also originally proposed that FC did not bind directly to the H ϩ -ATPase but rather to receptors located at the plasma membrane. A search for FC receptors led to their identification as members of the 14-3-3 eukaryotic protein family (8 -10). This discovery shed some light on the FC mechanism of H ϩ -ATPase activation. In fact it was clearly shown by different authors that FC is able to promote the association of 14-3-3 proteins with the C-terminal domain of the H ϩ -ATPase (11)(12)(13), thereby activating the enzyme (14). These lines of evidence indicate that the FC receptor is made up of a complex between 14-3-3 and H ϩ -ATPase and suggest that association of 14-3-3 proteins with the proton pump could represent a mechanism regulating the proton pump under physiological conditions.
It is a common property of 14-3-3 proteins that they have the capability to associate with target proteins through binding to consensus motifs containing phosphorylated residues (15,16).
Recently it was shown that a similar mechanism holds true also for plant 14-3-3 association with plant enzymes such as nitrate reductase (17,18) and sucrose phosphate synthase (19,20). Because it is known that H ϩ -ATPase is a phosphorylated enzyme in vivo (21)(22)(23), it is conceivable that a phosphorylation/dephosphorylation mechanism, by endogenous protein kinases and phosphatases, may regulate the association of 14-3-3 with the H ϩ -ATPase.
In the present study in vivo and in vitro data are reported demonstrating that 14-3-3 proteins are able to associate with the H ϩ -ATPase in a FC-independent manner and that the interaction depends on the phosphorylation status of the proton pump. We also demonstrate that a protein phosphatase 2A (PP2A) from maize roots affects the association of 14-3-3 proteins with the H ϩ -ATPase.

EXPERIMENTAL PROCEDURES
Chemicals-FC was prepared according to Ballio et al. (24). Okadaic acid, microcystin-LR, and anti-PP2A antibodies were purchased from Calbiochem, (La Jolla, CA). [␥-32 P]ATP (specific activity, 110 TBq/ mmol) and thrombin were from Amersham Pharmacia Biotech. Alkaline phosphatase, from calf intestine, was from Roche Molecular Biochemicals. Protein kinase A, catalytic subunit, and myelin basic protein (from bovine brain) were from Sigma. Chemicals for gel electrophoresis were from Bio-Rad. All other reagents were of analytical grade.
Plant Material-Maize seeds (Zea mays L. cv. Paolo) from Dekalb (Mestre, Italy) were germinated and grown in the dark for 6 days, as already described (25).
In Vivo Incubation of Maize Roots with Okadaic Acid and FC-Maize roots (8 g) were cut into small pieces (approximately 5 mm) and incubated in 20 ml of 20 mM Tris-Mes (pH 7.0), containing 300 mM sucrose, for 1 h at room temperature. When indicated, 1 M okadaic acid (OA) and 10 M FC were added.
Purification of H ϩ -ATPase from Maize Roots-Two-phase partitioned plasma membranes were obtained from 200 g of maize roots as described previously (26). When indicated, plasma membranes were treated with 0.5% Triton X-100 (26). For the purification of H ϩ -ATPase, plasma membrane proteins were solubilized with dodecyl-␤-D-maltoside as described by Johansson et al. (27) and fractionated by anion exchange HPLC (26).
Purification of Protein Phosphatase from Maize Roots-The cytosolic fraction obtained from the purification of H ϩ -ATPase (supernatant from the 50,000 ϫ g centrifugation) was used as starting material. Cytosol was dialyzed overnight at 4°C against 20 mM Tris-Mes (pH 7.0) and loaded onto a DEAE-biogel (Bio-Rad) column (100 ϫ 20 mm) equilibrated with 20 mM Tris-Mes (pH 6.5) containing 1 mM DTT and 1 mM PMSF (buffer A). Elution was performed in 40 min by a linear gradient from 0 to 0.5 M NaCl in buffer A, at a flow rate of 2.0 ml/min.
Fractions of 2 ml were collected and tested for the ability to abolish the interaction between HPLC-purified H ϩ -ATPase and 14-3-3 in an in vitro overlay assay (13). Active fractions were pooled, dialyzed, and concentrated to 3 ml in a Sartorius collodion bag (cut-off, 12 kDa) and loaded onto a DEAE TSK-5-PW HPLC (Bio-Rad) column (75 ϫ 7.5 mm), equilibrated at 0.5 ml/min with 20 mM histidine-HCl buffer (pH 6.5) containing 0.5 mM DTT, 1 mM PMSF. Elution was performed in 40 min, by a linear gradient from 0 to 0.5 M NaCl, in the same buffer. Fractions of 1 ml were collected, active fractions were pooled, and aliquots of 200 l were loaded onto a Bio-sil TSK-250 (Bio-Rad) HPLC gel filtration column (300 ϫ 7.5 mm), equilibrated in 20 mM Tris-Mes buffer (pH 7.0), containing 150 mM NaCl and 0.5 mM DTT, at 1.0 ml/min flow rate. Active fractions were collected and used for the biochemical characterization of protein phosphatase activity.
Identification of H ϩ -ATPase-Dephosphorylating Protein Phosphatase-To identify protein phosphatase activities able to act on H ϩ -ATPase, a functional assay was set up by testing the ability of chromatographic fractions to inhibit H ϩ -ATPase/14-3-3 association. 2 g of HPLC-purified plasma membrane H ϩ -ATPase were incubated with 1 g of each fraction in 50 mM Tris-Mes (pH 7.0), containing 10 mM MgCl 2 , 1.2 mM CaCl 2 , 1 mM EGTA (free calcium about 100 M), 1 mM DTT, 1 mM PMSF, 10 M leupeptin, 5 M chymostatin, 5 M pepstatin, for 30 min at room temperature. Treated H ϩ -ATPase was then used in the overlay assay (13). Specificity of protein phosphatase was assessed incubating samples under the same conditions, but in the presence of 2 M OA and 2 M microcystin-LR (MC).
Overlay Assay-The cDNA of the 14-3-3 isoform GF14-6 from maize was cloned into pGEX-2TK and expressed in Escherichia coli, as described previously (13). The expression system produces a glutathione S-transferase-fused 14-3-3 containing a cAMP-dependent protein kinase phosphorylation site and a thrombin site between the two polypeptides. The radiolabeled GF14-6 was obtained as described by Fullone et al. (13). The specific activity of 32 P-labeled 14-3-3 was 3.3 MBq/mg.
The overlay assay was carried out according to Fullone et al. (13), with minor modifications. Briefly, two-phase partitioned plasma membranes (10 g protein) or HPLC-fractionated H ϩ -ATPase (2 g) were subjected to SDS-PAGE and blotted onto nitrocellulose membrane. The membrane was blocked with 5% fatty acid-free milk in phosphatebuffered saline and then incubated in phosphate-buffered saline containing 2% fatty acid-free milk and the 32 P-labeled GF14-6 (8.3 kBq/ml) overnight at 4°C. After incubation, the membrane was washed three times with phosphate-buffered saline, dried, and subjected to autoradiography at Ϫ80°C. The relative amount of bound 14-3-3 proteins was estimated from overlay assays by densitometry using Scion Image software from Scion Corporation.
Protein Phosphatase Assay-Protein phosphatase activity was estimated using 32 P-labeled myelin basic protein (MBP) as substrate. 32  After a 20-min incubation, an equal volume of 20% trichloroacetic acid was added, and the mixture was centrifuged for 2 min at 5000 ϫ g. The radioactivity of 50-l aliquots of the supernatant was measured with a Wallac 1410 ␤-counter.
Western Blotting-SDS-PAGE was performed as described by Laëmmli (28), in a Mini Protean apparatus (Bio-Rad). Proteins were transferred onto nitrocellulose membranes using a semidry LKB apparatus FIG. 1. Interaction between 14-3-3 proteins and H ؉ -ATPase. The overlay assay was performed using the H ϩ -ATPase partially purified from control or FC-treated maize roots. The H ϩ -ATPase (2 g of total protein) was subjected to SDS-PAGE, blotted onto nitrocellulose, and incubated with 32 P-labeled 14-3-3 overnight at 4°C. After incubation, the membrane was washed three times, dried, and subjected to autoradiography at Ϫ80°C. (2 h, 0.8 mA cm Ϫ2 ). Immunodecoration of 14-3-3 and H ϩ -ATPase was performed according to Marra et al. (26), using the ECL detection system from Amersham Pharmacia Biotech, following the manufacturer's instructions. Immunodetection of protein phosphatase was carried out using antibodies recognizing the catalytic subunit of mammalian PP2A (Calbiochem) and anti-rabbit secondary antibodies conjugated to horseradish peroxidase (Bio-Rad).
Analytical Methods-Protein concentration was determined by the method of Bradford (29) using bovine serum albumin as the standard. H ϩ -ATPase activity was assayed according to Serrano (30).

RESULTS
The H ϩ -ATPase Purified from Maize Roots Interacts in Vitro with 14-3-3 Proteins in the Absence of FC-The H ϩ -ATPase was solubilized and partially purified by anion exchange HPLC from maize roots previously incubated in the presence or in the absence of 10 M FC (26). The partially purified H ϩ -ATPase (2 g) from FC-treated and control tissue were run on 10% SDS-PAGE, blotted on nitrocellulose membrane, and incubated with 32 P-labeled 14-3-3, both in the absence and in the presence of 10 M FC in the incubation mixture.
Results shown in Fig. 1A demonstrate that 14-3-3 proteins associate with H ϩ -ATPase in the absence of FC and that in vivo preincubation of the tissue with FC increases the interaction with the proton pump; the extent of the increase was 40%, as estimated by densitometric analysis. In Fig. 1B, the effect of in vitro incubation of FC on the interaction between 14-3-3 proteins and H ϩ -ATPase from FC-treated and from untreated tissue is shown. It is worth noting that FC in vitro administration strongly enhances the interaction not only between 14-3-3 and H ϩ -ATPase purified from control tissue but remarkably (42%, by densitometric analysis) also with the H ϩ -ATPase purified from FC-treated tissue.
These results demonstrate that 14-3-3 proteins are able to interact with the H ϩ -ATPase also in a FC-independent manner and indicate that in vivo FC treatment affects the capability of the proton pump to associate in vitro with 14-3-3 proteins, suggesting that the phosphorylation status of the purified H ϩ -ATPase may be involved.
In Vivo Administration of OA Stimulates H ϩ -ATPase Activity and Increases Plasma Membrane-bound Levels of 14-3-3-OA, a strong protein phosphatase inhibitor, was administered in vivo to maize roots to ascertain whether the phosphorylation status of H ϩ -ATPase influences the association with 14-3-3 proteins and consequently H ϩ -ATPase phosphohydrolytic activity.
Maize roots segments were incubated with 1 M OA or 10 M FC or both. After incubation, plasma membranes were isolated, the activity of the H ϩ -ATPase was tested, and the amount of plasma membrane-associated 14-3-3 was estimated by West- Association of 14-3-3 Proteins with the H ϩ -ATPase ern blotting analysis. Results are shown in Fig. 2. OA induces a slight (about 20%) but reproducible stimulation of H ϩ -ATPase phosphohydrolytic activity (Fig. 2A); under the same conditions FC brings about a much stronger stimulation (116%), whereas incubation with OAϩFC provoked the highest stimulation (245%). The stimulatory effect was much more evident after treatment of isolated plasma membranes with 0.5% Triton X-100. In particular the effect of OA increased to 146%, that of FC to 375%, and that of OAϩFC to 884%. Remarkably, Western blot analysis showed that stimulation of H ϩ -ATPase was correlated with amounts of 14-3-3 associated with plasma membranes (Fig. 2B).
As shown in Fig. 2B (panel b), Triton X-100 was able to selectively remove 14-3-3 proteins from plasma membranes. In fact 14-3-3 proteins were almost completely undetectable in control membranes, whereas their levels were progressively less reduced in OA, FC, and OAϩFC samples, indicating that a tighter association was induced by in vivo treatment with FC, as expected, and interestingly also by OA. These results indicate that phosphorylation of H ϩ -ATPase is a prerequisite for stimulatory 14-3-3 association in the absence of FC.
Dephosphorylation of H ϩ -ATPase Inhibits Association of 14-3-3 Proteins-To verify whether the phosphorylation status of the H ϩ -ATPase modulates the association with 14-3-3 proteins, the proton pump was subjected to a dephosphorylating treatment with alkaline phosphatase and tested for the ability to bind 32 P-radiolabeled 14-3-3 protein in an overlay assay (13). Results showed in Fig. 3A clearly demonstrate that dephosphorylated proton pump is almost unable to bind 14-3-3 proteins; addition of FC in vitro partially restored the interaction (Fig.  3B).
Identification of a Protein Phosphatase from Maize Roots That Negatively Regulates the Interaction between H ϩ -ATPase and 14-3-3 Proteins-Results obtained with alkaline phosphatase prompted us to investigate whether protein phosphatase activities able to modulate binding of 14-3-3 to H ϩ -ATPase were present in maize roots. To this purpose, cytosol from maize roots was fractionated by anion exchange DEAE-biogel chromatography followed by DEAE and gel filtration HPLC purification (Fig. 4A). Fractions were directly tested for their ability to inhibit the association between H ϩ -ATPase and 14-3-3 proteins in the overlay assay. Specificity of the protein phosphatase activity was evaluated by assaying fractions also in the presence of 1 M OA and 1 M MC. Relative amounts of bound 14-3-3 were estimated by densitometric analysis. In Fig.  4B the effect of the HPLC size exclusion-purified fraction on the interaction between H ϩ -ATPase and 14-3-3 proteins in the overlay assay is shown: protein phosphatase treatment of the H ϩ -ATPase nearly completely inhibited binding of 14-3-3 proteins. OA addition (1 M) in the incubation mixture restored the interaction. Interestingly, protein phosphatase treatment drastically reduced also the FC-dependent interaction between H ϩ -ATPase and 14-3-3 proteins.
Biochemical Characterization of the Partially Purified Protein Phosphatase-Biochemical and immunological analysis allowed us to ascertain that a protein phosphatase activity with properties typical of the PP2A family was present in fractions able to inhibit the H ϩ -ATPase/14-3-3 interaction. The characterization of the partially purified protein phosphatase (size exclusion fraction) was carried out using 32 P-MBP as substrate. Specific activity and yield of each chromatographic step are shown in Table I.
Protein phosphatase activity was completely independent of Ca 2ϩ and Mg 2ϩ ions in concentrations ranging from 1 M to 1 mM (data not shown), whereas it was drastically affected (Fig.  5A) by OA (IC 50 ϭ 0.9 nM) and MC (IC 50 ϭ 0.6 nM), in accordance with data reported for PP2A (31). A further confirmation was provided by Western blotting analysis using antibodies raised against a conserved sequence of the catalytic domain of mammalian PP2A. As shown in Fig. 5B, the antibodies recognized a single polypeptide of approximately 37 kDa, the same molecular mass reported for the catalytic subunit of plant PP2A (31).  Association of 14-3-3 Proteins with the H ϩ -ATPase

DISCUSSION
Results from recent research have contributed much progress to the understanding of the mechanism of H ϩ -ATPase stimulation by the phytotoxin FC. In fact it was clearly demonstrated by different authors that FC promotes the irreversible association of stimulatory 14-3-3 proteins with the H ϩ -ATPase. Because an ever-increasing body of evidence points to 14-3-3 proteins as general regulators of key enzymes in the plant cell (32), results from FC research also suggest that 14-3-3 proteins may be involved in the regulation of H ϩ -ATPase under physiological conditions. In fact, different pieces of evidence (11,13) suggest that 14-3-3 proteins can also interact with the proton pump in a FC-independent manner. Although the molecular basis of the FC-independent interaction between H ϩ -ATPase and 14-3-3 proteins is still unknown, phosphorylation/dephosphorylation events on the H ϩ -ATPase are likely, because a common feature of 14-3-3 association with target proteins involves the interaction with consensus sequences containing a phosphorylated amino acid (15,16).
This work was aimed at ascertaining whether a similar mechanism can account for the interaction between 14-3-3 proteins and H ϩ -ATPase. To this purpose, the ability of H ϩ -ATPase to associate in vitro with 14-3-3 proteins after in vivo treatments able to alter its phosphorylation status was investigated. Data reported here show that in vivo incubation of maize roots with OA provoked a significant increase of the plasma membrane-bound 14-3-3 proteins, as well as a strong stimulation of H ϩ -ATPase activity. These results strongly suggest that phosphorylation of H ϩ -ATPase is essential for its binding to 14-3-3 proteins and concomitant stimulation. We also demonstrated that FC in vivo incubation of maize roots, a treatment resulting in an increase of membrane-bound 14-3-3 and stimulation of H ϩ -ATPase activity, was able to increase the in vitro interaction of H ϩ -ATPase with 14-3-3 proteins. Also this finding is suggestive of the occurrence of a post-translational modification on the proton pump, influencing its capability to interact with 14-3-3 proteins. Interestingly, Olsson et al. (23) demonstrated that in vivo FC treatment protects from dephosphorylation a phosphothreonine residue (Thr-948 in the Arabidopsis isoform AHA1) located in the very end of the Cterminal domain of the H ϩ -ATPase. It was hypothesized that the effect is due to direct binding of 14-3-3 to the phosphorylated threonine or alternatively to a 14-3-3-induced conformational change protecting the amino acid from dephosphorylation. Our results indicate that the phosphothreonine residue may be physically involved in the binding of 14-3-3 proteins, being part, as proposed by Olsson at al. (23), of a novel consensus sequence regulating the physiological association of 14-3-3 proteins with the H ϩ -ATPase.
Furthermore, in vivo data indicate that the phosphorylation status of the H ϩ -ATPase is relevant also for the FC-dependent binding of 14-3-3 protein; in fact, the H ϩ -ATPase purified from in vivo FC or OA-treated roots bound 14-3-3 protein more efficiently also in the presence of FC. This latter result is also worth noting, because results obtained up to now point instead to different, independent mechanisms for phosphorylation and FC-mediated binding of 14-3-3 proteins to the H ϩ -ATPase (13,14).
A further demonstration of the relevance of phosphorylation as a regulatory mechanism was provided by in vitro dephosphorylation treatments of the proton pump; in fact, incubation of the H ϩ -ATPase with commercially available alkaline phos-phatase and more remarkably with a PP2A partially purified from maize roots, completely abolished the interaction with 14-3-3 proteins. Notably, this treatment was also able to impair the FC-dependent association of 14-3-3 proteins. In fact, FC in vitro addition only partially restored the association of 14-3-3 proteins with the proton pump.
In conclusion, our data, taken together, unequivocally demonstrate that the interaction between 14-3-3 proteins and H ϩ -ATPase is dependent on the phosphorylation status of the enzyme; moreover, they also indicate that phosphorylation/ dephosphorylation events, mediated by a still unidentified protein kinase and possibly a PP2A, very likely represent the basis for the physiological 14-3-3-mediated regulation of H ϩ -ATPase. In addition, our data raise the question as to how FC promotes the association of 14-3-3 proteins with the H ϩ -ATPase. In fact, whereas it was proposed that FC can act by a mechanism able to replace the phosphorylation requirements of the H ϩ -ATPase (13,14), our data indicate that the phosphorylation status of the H ϩ -ATPase can influence also the FC-dependent 14-3-3 association with the proton pump and consequently that FC does not merely mimic the phosphorylation effect but rather acts through a more complex mechanism.