Biochemical Characterization of In Vitro Phosphorylation and Dephosphorylation of the Plasma Membrane H⁺-ATPase

Abstract

Stomatal opening, which is mediated by blue light receptor phototropins, is driven by activation of the plasma membrane H⁺-ATPase via phosphorylation of the penultimate threonine in the C-terminus and subsequent binding of a 14-3-3 protein. However, the biochemical properties of the protein kinase and protein phosphatase for H⁺-ATPase are largely unknown. We therefore investigated in vitro phosphorylation and dephosphorylation of H⁺-ATPase. H⁺-ATPase was phosphorylated in vitro on the penultimate threonine in the C-terminus in isolated microsomes from guard cell protoplasts of Vicia faba. Phosphorylated H⁺-ATPase was dephosphorylated in vitro, and the dephosphorylation was inhibited by EDTA, a divalent cation chelator, but not by calyculin A, an inhibitor of type 1 and 2A protein phosphatases. Essentially the same results were obtained in purified plasma membranes from etiolated Arabidopsis seedlings, indicating that a similar protein kinase and phosphatase are involved in plant cells. Further analyses revealed that phosphorylation of the H⁺-ATPase is insensitive to K-252a, a potent inhibitor of protein kinase, and is hypersensitive to Triton X-100, a non-ionic detergent. Moreover, dephosphorylation required Mg²⁺ but not Ca²⁺, and protein phosphatase was localized in the 1% Triton X-100-insoluble fraction. These results demonstrate that a protein kinase–phosphatase pair, K-252a-insensitive protein kinase and Mg²⁺-dependent type 2C protein phosphatase, co-localizes at least in part with the H⁺-ATPase in the plasma membrane and regulates the phosphorylation status of the penultimate threonine of the H⁺-ATPase.

Introduction

Stomatal pores surrounded by a pair of guard cells in the epidermis regulate gas exchange between leaves and the atmosphere and allow CO₂ entry for both photosynthesis and the transpiration stream in higher plants (Shimazaki et al. 2007). Stomata open through activation of a H⁺ pump in guard cells in response to blue light (Assmann et al. 1985, Shimazaki et al. 1986). The blue light-activated pump creates an inside-negative electrical potential across the plasma membrane and drives K⁺ uptake through voltage-gated K⁺ channels (Assmann and Shimazaki 1999, Schroeder et al. 2001). The H⁺ pump has been demonstrated to be the plasma membrane H⁺-ATPase (Kinoshita and Shimazaki 1999). Blue light activates the H⁺-ATPase through phosphorylation of the penultimate threonine in the C-terminus and binding of 14-3-3 protein to the phosphorylated C-terminus in stomatal guard cells (Kinoshita and Shimazaki 1999, Kinoshita and Shimazaki 2002). Genetic analyses of Arabidopsis mutants and biochemical analysis using guard cell protoplasts (GCPs) from Vicia faba demonstrated that phototropins (phot1 and phot2) function as blue light receptors in a redundant manner and mediate blue light- induced stomatal opening through activation of the H⁺-ATPase in guard cells (Kinoshita et al. 2001, Briggs and Christie 2002, Kinoshita et al. 2003, Ueno et al. 2005). However, the signaling mechanism from phototropins to activation of the H⁺-ATPase is largely unknown, although RPT2 and type 1 protein phosphatases act as positive regulators in this pathway (Inada et al. 2004, Takemiya et al. 2006).

The plasma membrane H⁺-ATPase generates an electrochemical gradient across the plasma membrane that provides a driving force for uptake of numerous nutrients, including K⁺, NO³⁻, SO₄²⁻, PO₄³⁻, amino acids, peptides and sucrose, by coupling with organ-specific transporters, and for cell turgor maintenance (Sze et al. 1999, Palmgren 2001, Sondergaard et al. 2004). Because plasma membrane H⁺-ATPases are ubiquitous in all plant cell types investigated, the regulatory mechanism of this enzyme has been extensively studied (Michelet and Boutry 1995). It has been demonstrated that H⁺-ATPase activity is regulated by an autoinhibitory domain in its C-terminal region, as removal of a 7–10 kDa fragment from its C-terminal end generates a high-activity state of H⁺-ATPase with a higher V_max, a lower K_m for ATP, changed pH dependence with higher activity at physiological pH, and increased coupling of H⁺ transport with ATP hydrolysis (Palmgren et al. 1990, Morsomme et al. 1996).

Further analyses revealed that the 14-3-3 protein and phosphorylation of the H⁺-ATPase are required for its activation. The 14-3-3 proteins have been shown to bind to the phosphorylated C-terminus of the H⁺-ATPase, and the phosphorylation site was determined to be the penultimate Thr948 [numbering according to the Arabidopsis H⁺-ATPase isoform 1 (AHA1)] in the C-terminus of H⁺-ATPase from spinach leaves (Olsson et al. 1998). Moreover, this phospho-Thr948 residue and neighboring amino acids have been demonstrated to create a novel binding motif of YpT948V for the 14-3-3 protein and to regulate the physiological binding of 14-3-3 protein to H⁺-ATPase in several cell types (Fuglsang et al. 1999, Svennelid et al. 1999, Camoni et al. 2000, Maudoux et al. 2000, Kinoshita and Shimazaki 2002). There are 11 functional H⁺-ATPase genes in the Arabidopsis genome, and the penultimate threonine is conserved in all H⁺-ATPases (Palmgren 2001). Therefore, activation of the H⁺-ATPase by phosphorylation of the penultimate threonine in the C-terminus, concomitant with binding of the 14-3-3 protein to the phosphorylated C-terminus, is thought to be a common physiological mechanism for regulation of the H⁺-ATPase in plant cells (Palmgren 2001, Sondergaard et al. 2004).

Analyses using the fungal toxin fusicoccin (FC), an activator of H⁺-ATPase, support this hypothesis. FC induces phosphorylation of the penultimate threonine in the C-terminus and a stable complex with the phosphorylated C-terminus and the 14-3-3 protein in several cell types (Camoni et al. 2000, Maudoux et al. 2000). FC action is thought to stabilize/accumulate the complex of phosphorylated H⁺-ATPase and 14-3-3 protein through inhibition of its dephosphorylation (Fuglsang et al. 1999, Svennelid et al. 1999, Kinoshita and Shimazaki 2001). These results suggest that a similar/identical protein kinase and protein phosphatase are involved in regulation of the phosphorylation status of the H⁺-ATPase in plant cells.

The protein kinase activity for phosphorylation of the penultimate threonine in the C-terminus of H⁺-ATPase in vitro was detected in the plasma membrane fraction from spinach leaves (Svennelid et al. 1999). Moreover, Camoni et al. (2000) showed that type 2A protein phosphatase purified from the cytosol of maize roots dephosphorylates the phosphorylated penultimate threonine of H⁺-ATPase in vitro. However, the biochemical properties of this protein kinase and phosphatase have yet to be determined, not only in stomatal guard cells but also in other cell types. In contrast, Arabidopsis Ser/Thr protein kinase, PKS5, is suggested to be a negative regulator of the H⁺-ATPase through phosphorylation of Ser931 in the C-terminus of H⁺-ATPase and prevention of an interaction between the phosphorylated penultimate threonine of H⁺-ATPase and 14-3-3 protein (Fuglsang et al. 2007). Consistent with this, phosphorylations (Thr931 and Ser938) within the H⁺-ATPase C-terminal region negatively control binding of 14-3-3 protein to the phosphorylated penultimate Thr955 in PMA2 of Nicotiana tabacum (Duby et al. 2009).

In the present study, we performed in vitro phosphorylation and dephosphorylation assays focused on the penultimate threonine in the C-terminus of plasma membrane H⁺-ATPase using isolated microsomal membranes from GCPs of V. faba and purified plasma membranes from etiolated seedlings of Arabidopsis thaliana, and characterized the biochemical properties of the protein kinase and protein phosphatase. Our results show that the protein kinase and phosphatase co-localize at least in part with the H⁺-ATPase in the plasma membrane and possess similar biochemical properties in both guard cells and etiolated seedlings.

Results

Blue light-induced phosphorylation of the H⁺-ATPase in guard cell protoplasts

Fig. 1A shows typical blue light-induced phosphorylation of the H⁺-ATPase in GCPs from V. faba, detected by immunoblot using specific antibodies against the phosphorylated penultimate threonine of the H⁺-ATPase (p-Thr), and a protein blot using the recombinant glutathione S-transferase (GST)–14-3-3 protein that specifically binds to the phosphorylated penultimate threonine of the H⁺-ATPase as a probe (Kinoshita and Shimazaki 1999, Kinoshita and Shimazaki 2001, Kinoshita and Shimazaki 2002). The H⁺-ATPase antibodies specifically recognized a protein band of 95 kDa as H⁺-ATPase. When GCPs were illuminated with a short pulse of blue light (30 s, 100 μmol m⁻² s⁻¹) superimposed on background red light (600 μmol m⁻² s⁻¹), the H⁺-ATPase was phosphorylated on the penultimate threonine and bound to the 14-3-3 protein. Treatment of GCPs with the fungal toxin FC also induced phosphorylation of the H⁺-ATPase and its binding to 14-3-3 protein. Because the amount of the 14-3-3 protein binding to the phosphorylated H⁺-ATPase was proportional to the phosphorylation level of the penultimate threonine, we therefore detected the phosphorylation status of the penultimate threonine in the H⁺-ATPase by a protein blot in the subsequent experiments. In addition, the present results using anti-pThr antibodies provide new evidence for phosphorylation of the penultimate threonine of the H⁺-ATPase in guard cells in response to blue light and FC, because the phosphoryaltion site of the H⁺-ATPase in guard cells had been determined indirectly by competition experiments in a protein blot using synthetic phosphopeptides (Kinoshita and Shimazaki, 2001, Kinoshita and Shimazaki 2002).

Fig. 1

In vitro phosphorylation and dephosphorylation of the H⁺-ATPase in microsomes from GCPs of Vicia faba. (A) Blue light- and FC-induced phosphorylation of the H⁺-ATPase in GCPs of V. faba. GCPs were pre-incubated under red light for 30 min, then illuminated with a 30 s blue light pulse superimposed on the red light, or FC at 10 μM was added to the GCP suspension. RL, red light; BL, 3.5 min after blue light commenced; +FC, 4 min after addition of FC. The phosphorylation level of the H⁺-ATPase was detected by immunoblot using anti-phospho-Thr947 antibody (p-Thr) and protein blot (PB) using GST–14-3-3 protein as a probe. The H⁺-ATPase was detected by immunoblot (IB) using a specific antibody for the H⁺-ATPase. (B) In vitro phosphorylation of the H⁺-ATPase. Reactions were started by the addition of 2 mM ATP to the reaction buffer containing microsomes (10 μg of protein) and performed for 60 min at 30°C. K-252a was added to the reaction mixture at 0.5 μM. The phosphorylation level of the H⁺-ATPase was detected by protein blot (PB) using GST–14-3-3 protein as a probe. The H⁺-ATPase was detected by immunoblot (IB) using a specific antibody for the H⁺-ATPase. (C) In vitro dephosphorylation of the H⁺-ATPase. Microsomes (10 μg of protein) from blue light-illuminated GCPs were used for the dephosphorylation (Untreated) and the reactions were performed for 30 min at 30°C (None). Calyculin A (CA) at 0.5 μM, 0.5 mM vanadate (Vana.) and 5 mM EDTA were added to the reaction buffer. Other details were the same as in B. All experiments were repeated on three occasions and yielded similar results.

In vitro phosphorylation of the H⁺-ATPase in microsomes from GCPs

We then performed in vitro phosphorylation of the H⁺-ATPase in isolated microsomal membranes from GCPs, and detected the phosphorylation status of the penultimate threonine in the C-terminus using a protein blot (Fig. 1B). Microsomal membranes were isolated from red light-illuminated GCPs in which the H⁺-ATPase exists at a lower phosphorylation level (Fig. 1A). When we added 2 mM ATP to the microsomal membrane fraction, the H⁺-ATPase was phosphorylated, suggesting that protein kinase, which directly phosphorylates the H⁺-ATPase, localizes in the microsomal membrane. A previous investigation indicated that K-252a, a potent inhibitor of protein kinase (Kase et al. 1987), had no effect on FC-induced phosphorylation of the H⁺-ATPase in GCPs (Kinoshita and Shimazaki 2001). We thus examined the effect of K-252a on in vitro phosphorylation of the H⁺-ATPase. K-252a at 0.5 μM had no effect on phosphorylation, indicating that a K-252a-insensitive protein kinase is involved in phosphorylation of the H⁺-ATPase. The absence of any effect of K-252a on in vitro phosphorylation of the H⁺-ATPase is not due to a low concentration of K-252a, as the same concentration of K-252a inhibited in vitro light-induced autophosphorylation of phot1 in plasma membranes from Arabidopsis etiolated seedlings (Supplementary Fig. S1).

In vitro dephosphorylation of the H⁺-ATPase in microsomes from GCPs

We next examined in vitro dephosphorylation of the H⁺-ATPase (Fig. 1C). The GCPs were illuminated for 30 s with blue light superimposed on background red light to induce phosphorylation of the H⁺-ATPase in vivo (Fig. 1A). We then isolated the microsomal membrane from GCPs at 3.5 min after the start of the blue light pulse, when the H⁺-ATPase showed maximum phosphorylation, and microsomal membranes were used for in vitro dephosphorylation. Before the dephosphorylation reaction, the H⁺-ATPase was phosphorylated (Fig. 1C, Untreated). After the dephosphorylation reaction, the H⁺-ATPase was almost completely dephosphorylated (Fig. 1C, None), suggesting that protein phosphatase also localizes in the microsomal membrane. The decrease of the phosphorylation level of the H⁺-ATPase was not due to its degradation, as the same amounts of H⁺-ATPase protein were detected immunologically using H⁺-ATPase antibodies.

To classify the types of protein phosphatase, we examined the effects of protein phosphatase inhibitors on dephosphorylation. The results showed that 0.5 μM calyculin A, an inhibitor of type 1 and type 2A protein phosphatases (Ishihara et al. 1989), and 0.5 mM vanadate, an inhibitor of tyrosine phosphatase (Huyer et al. 1997), had no effect on the dephosphorylation, but 5 mM EDTA, a chelating agent for divalent cations, severely inhibited dephosphorylation. Given that the reaction buffer for the dephosphorylation only contains 2.5 mM Mg²⁺ as a divalent cation, Mg²⁺ is likely to be required for dephosphorylation.

In vitro dephosphorylation of the H⁺-ATPase in plasma membranes from etiolated Arabidopsis seedlings

Phosphorylation of thepenultimate threonine of the H⁺-ATPase, which is a prerequisite for the binding of 14-3-3 protein, is thought to be a common mechanism for regulation of H⁺-ATPases in plant cells (Palmgren 2001, Sondergaard et al. 2004). Furthermore, isolation of the membrane fraction from etiolated Arabidopsis seedlings is easier than from GCPs. Therefore, we further investigated in vitro phosphorylation and dephosphorylation of the H⁺-ATPase in plasma membranes isolated from etiolated Arabidopsis seedlings by the two-phase partitioning method. Two H⁺-ATPase isogenes (AHA1 and AHA2), which are abundant in plants (Ueno et al. 2005), were mainly expressed in etiolated seedlings (Supplementary Fig. S2). The H⁺-ATPase from etiolated seedlings was phosphorylated under the present growth conditions (Fig. 2A, Untreated). Therefore, we first examined in vitro dephosphorylation of the H⁺-ATPase. After the dephosphorylation reaction, the H⁺-ATPase was almost completely dephosphorylated without degradation of the H⁺-ATPase (Fig. 2A, None). Dephosphorylation reached the lowest level around 20 min after the start of the reaction (Supplementary Fig. S3).

Fig. 2

In vitro dephosphorylation of the H⁺-ATPase in plasma membranes from etiolated Arabidopsis seedlings. (A) In vitro dephosphorylation of the H⁺-ATPase. Plasma membranes (10 μg of protein) were used for the dephosphorylation (Untreated) and the reactions were performed for 30 min at 30°C (None). Calyculin A (CA) at 0.5 μM, 0.5 mM vanadate (Vana.), 5 mM EDTA and 10 μM FC were added to the reaction buffer. Other details were the same as in Fig. 1B. (B) In vivo treatments of etiolated seedlings with calyculin A (CA) and FC. CA (0.5 μM) and FC (10 μM) were incubated with the etiolated seedlings for 60 min at room temperature in MS liquid medium. Reactions were terminated by homogenization of etiolated seedlings using a mortar and pestle in SDS buffer. Protein (50 μg) was subjected to SDS–PAGE. Other details were the same as in Fig. 1B. (C) Effect of divalent cations on in vitro dephosphorylation. Divalent cations (Mg²⁺, Ca²⁺, Mn²⁺ and Co²⁺) were added to the reaction buffer at 2.5 mM. Other details were the same as in Fig. 1B. (D) Effect of Triton X-100 (TX-100), a non-ionic detergent, on dephosphorylation. TX-100 was added to the reaction mixture at the indicated concentrations. Other details were the same as in Fig. 1B. All experiments were repeated on three occasions and yielded similar results. The graphs show the relative phosphorylation level of the H⁺-ATPase quantified by the ratio of the phosphorylated H⁺-ATPase to the total quantity of the H⁺-ATPase. The bars represent the means with standard deviation for three independent experiments.

Dephosphorylation was inhibited by EDTA, but not by calyculin A and vanadate, indicating that the protein phosphatase for the H⁺-ATPase in etiolated seedlings from Arabidopsis has a sensitivity to EDTA similar to that in GCPs from V. faba. In addition, FC completely inhibited in vitro dephosphorylation of the H⁺-ATPase. This result supports previous results showing that FC inhibits dephosphorylation of the H⁺-ATPase through stabilizing the phosphorylated H⁺-ATPase–14-3-3 protein complex (Fuglsang et al. 1999, Svennelid et al. 1999, Kinoshita and Shimazaki 2001). Indeed, the endogenous 14-3-3 proteins (mainly 35, 33.5 and 32.5 kDa) were identified immunologically in the plasma membrane (Fig. 3A).

Fig. 3

In vitro phosphorylation of the H⁺-ATPase in plasma membranes of etiolated Arabidopsis seedlings. (A) In vitro phosphorylation of the H⁺-ATPase. Plasma membranes (10 μg of protein) were used for the phosphorylation. Reactions were started by the addition of 2 mM ATP to the reaction buffer and performed for 60 min at 30°C. FC and the recombinant 14-3-3 protein (His-GF14ø) were added to the reaction mixture at 10 and 0.5 μM, respectively. The 14-3-3 proteins were detected by immunoblot (IB) using a specific antibody for the 14-3-3 protein. Other details were the same as in Fig. 1B. (B) Effect of K-252a on phosphorylation. Reactions were started by the addition of 2 mM ATP and 10 μM FC to the reaction buffer. K-252a was added to the reaction mixture at 0.5 μM. Other details were the same as in Fig. 1B. (C) Effect of Triton X-100 (TX-100), a non-ionic detergent, on phosphorylation. Reactions were started by the addition of 2 mM ATP and 10 μM FC to the reaction buffer. TX-100 was added to the reaction mixture at the indicated concentrations. Other details were the same as in Fig. 1B. All experiments were repeated on three occasions and yielded similar results. The graph shows the relative phosphorylation level of the H⁺-ATPase. The bars represent the means with standard deviation for three independent experiments.

Camoni et al. (2000) showed that in vivo treatment of maize roots with okadaic acid, an inhibitor of type 1 and type 2A protein phosphatases, stimulates H⁺-ATPase activity, and that type 2A protein phosphatase purified from cytosol of maize roots dephosphorylates the H⁺-ATPase in vitro. To confirm this, we treated the etiolated seedlings with calyculin A as a membrane-permeable inhibitor of type 1 and type 2A protein phosphatases (Neumann et al. 1995) and examined the phosphorylation level of the H⁺-ATPase (Fig. 2B). However, in vivo treatment of etiolated seedlings with calyculin A did not increase the phosphorylation level of the H⁺-ATPase, although in vivo treatment with FC induced phosphorylation of the H⁺-ATPase.

To determine which divalent cation is required for dephosphorylation of the H⁺-ATPase, we examined the effects of Mg²⁺, Ca²⁺, Mn²⁺ and Co²⁺ on the dephosphorylation (Fig. 2C). Mg²⁺ and Mn²⁺ induced dephosphorylation of the H⁺-ATPase, but Ca²⁺ and Co²⁺ had no effect. Mg²⁺ was more effective for the dephosphorylation than was Mn²⁺. These properties are similar to those of the type 2C protein phosphatase (PP2C) in Arabidopsis (Baudouin et al. 1999), suggesting that a PP2C-type phosphatase is likely to catalyze dephosphorylation of the H⁺-ATPase.

In vitro phosphorylation of the H⁺-ATPase in plasma membranes from etiolated Arabidopsis seedlings

Because the H⁺-ATPase is phosphorylated in the plasma membrane from etiolated seedlings (Fig. 2A), we performed in vitro phosphorylation of the H⁺-ATPase after in vitro dephosphorylation. The H⁺-ATPase was dephosphorylated before the phosphorylation reaction (Fig. 3A). We then performed in vitro phosphorylation of the H⁺-ATPase. When 2 mM ATP was added to the plasma membrane, the H⁺-ATPase was phosphorylated. This result is consistent with a previous study where in vitro phosphorylation of the penultimate threonine in the C-terminus of H⁺-ATPase was observed in isolated plasma membranes from spinach leaves (Svennelid et al. 1999).

The plasma membrane possesses high dephosphorylation activity for the H⁺-ATPase, and in vitro dephosphorylation of the H⁺-ATPase is inhibited by FC (Fig. 2). We therefore added 10 μM FC to the reaction mixture at the same time as ATP (Fig. 3A). As expected, the phosphorylation level of the H⁺-ATPase in the presence of both ATP and FC was higher than that in the presence of ATP alone. Addition of the recombinant 14-3-3 protein (His-GF14ø) had no effect on the phosphorylation level of the H⁺-ATPase. These results suggest that protein phosphatase dephosphorylates the H⁺-ATPase at the same time during the phosphorylation reaction, and that the endogenous 14-3-3 protein in the plasma membrane is sufficient for formation of the phosphorylated H⁺-ATPase–14-3-3 protein complex. Therefore, we performed in vitro phosphorylation in the presence of both ATP and FC in the following experiments. Phosphorylation reached the highest level around 40 min after ATP addition, and the level was sustained for at least 60 min (Supplementary Fig. S3).

We next investigated the effect of K-252a on phosphorylation (Fig. 3B). Consistent with Fig. 1B, 0.5 μM K-252a had no effect on phosphorylation of the H⁺-ATPase. These results indicate that the protein kinase for the H⁺-ATPase in etiolated seedlings from Arabidopsis also has insensitivity to K-252a similar to that observed in GCPs from V. faba.

Effect of Triton X-100 on in vitro phosphorylation and dephosphorylation of the H⁺-ATPase

Analyses of in vitro phosphorylation and dephosphorylation of the H⁺-ATPase indicated that both protein kinase and phosphatase co-localize with the H⁺-ATPase in the plasma membrane. We therefore examined the effect of Triton X-100 (TX-100), a non-ionic detergent, on in vitro phosphorylation and dephosphorylation of the H⁺-ATPase. We found that in vitro phosphorylation of the H⁺-ATPase was hypersensitive to TX-100 (Fig. 3C). The phosphorylation was drastically inhibited when the reaction buffer contained ≥0.05% TX-100. In contrast, dephosphorylation was not inhibited even in the presence of 0.1% TX-100, and was partially inhibited in the presence of ≥0.5% TX-100 (Fig. 2D).

Detection of dephosphorylation of the H⁺-ATPase in the Triton X-100-insoluble fraction

During these experiments, we found that the H⁺-ATPase localizes in the 1% TX-100-insoluble fraction of the plasma membrane (Fig. 4A). We then examined in vitro phosphorylation and dephosphorylation of the H⁺-ATPase using the insoluble fraction. Interestingly, EDTA-sensitive dephosphorylation was detected in the TX-100-insoluble fraction; however, phosphorylation was not (data not shown). These results suggest that protein phosphatase is likely to form a complex with the H⁺-ATPase. Therefore, we next performed immunoprecipitation of the H⁺-ATPase from the plasma membrane in the presence of 1% TX-100 using specific antibodies raised against the conserved catalytic domain of the H⁺-ATPase, and examined in vitro dephosphorylation of the H⁺-ATPase in the immunoprecipitate. The H⁺-ATPase antibodies specifically recognized a protein band of 95 kDa as H⁺-ATPase in the immunoprecipitate, and the immunoprecipitated H⁺-ATPase was phosphorylated (Fig. 4B). However, dephosphorylation of the H⁺-ATPase was not detected in the immunoprecipitate (data not shown). Silver staining of the proteins in the immunoprecipitate separated by SDS–PAGE revealed that at least nine proteins co-precipitate with the H⁺-ATPase (Supplementary Fig. S4).

Fig. 4

Analyses of the H⁺-ATPase complex from etiolated Arabidopsis seedlings. (A) In vitro dephosphorylation of the H⁺-ATPase. Triton X-100-insoluble fractions (10 μg of protein) were used for the dephosphorylation (Untreated) and the reactions were performed for 30 min at 30°C (None). Calyculin A (CA) at 0.5 μM, 0.5 mM vanadate (Vana.) and 5 mM EDTA were added to the reaction buffer. Other details were the same as in Fig. 1B. (B) Detection of the H⁺-ATPase in the immunoprecipitate. Immunoprecipitations were performed with (+) or without (−) the plasma membrane (Mem.) (0.1 mg of protein) from etiolated seedlings using a specific antibody for the H⁺-ATPase. The phosphorylation level of the H⁺-ATPase was detected by protein blot (PB) using GST–14-3-3 protein as a probe. The H⁺-ATPase was detected by immunoblot (IB) using a specific antibody for the H⁺-ATPase. All experiments were repeated twice on different occasions and gave similar results.

Discussion

Stomatal guard cells are an ideal system to investigate regulation of the H⁺-ATPase because this enzyme is clearly activated by blue light as a physiological signal via phosphorylation of the penultimate threonine in the C-terminus (Kinoshita and Shimazaki 1999, Kinoshita and Shimazaki 2002, Ueno et al. 2005). However, the biochemical properties with respect to the protein kinase and the protein phosphatase for the penultimate threonine in the C-terminus of the H⁺-ATPase in guard cells are unknown. We therefore first established in vitro phosphorylation and dephosphorylation assays in microsomal membranes from Vicia GCPs. We found that phosphorylation of the H⁺-ATPase in vitro is detected in microsomal membranes from Vicia GCPs and is insensitive to K-252a, a potent protein kinase inhibitor (Fig. 1B). Consistent with this, FC-induced phosphorylation of the H⁺-ATPase was almost insensitive to K-252a in GCPs (Kinoshita and Shimazaki 2001). The results indicate that the protein kinase for the H⁺-ATPase is insensitive to K-252a. Dephosphorylation of the H⁺-ATPase in vitro was also detected in microsomal membranes from Vicia GCPs and was inhibited by EDTA, but not by calyculin A, an inhibitor of type 1 and type 2A protein phosphatases (Fig. 1C), indicating that a divalent cation-dependent protein phosphatase catalyzes dephosphorylation of the H⁺-ATPase. Illumination of the microsomal membranes with blue light during in vitro phosphorylation and dephosphorylation reactions had no effect on the phosphorylation level of the H⁺-ATPase (data not shown), although both blue light receptor phototropins and the H⁺-ATPase co-localize in the microsomal membrane in GCPs (Kinoshita et al. 2003). These results suggest that the components of the blue light signaling pathway may be lost in isolated microsomal membranes, and that the H⁺-ATPase is unlikely to be a direct substrate of the receptor kinase phototropins.

Next, we performed in vitro phosphorylation and dephosphorylation of the H⁺-ATPase in the plasma membrane from etiolated Arabidopsis seedlings. We found that the protein kinase for the H⁺-ATPase co-localizes at least in part with the H⁺-ATPase and shows insensitivity to K-252a similar to that in GCPs (Fig. 3B). These results suggest that the biochemical properties of the protein kinase are similar in both GCPs and etiolated seedlings. Interestingly, further biochemical analysis using the plasma membrane from etiolated seedlings revealed that phosphorylation of the H⁺-ATPase is hypersensitive to the non-ionic detergent TX-100 (Fig. 3C). Phosphorylation of the H⁺-ATPase was detected only in the presence of ≤0.025% (w/v) TX-100. The critical micellar concentration (CMC), i.e. the minimal detergent concentration at which micelles are observed, of TX-100 was demonstrated to be around 0.016% (w/v) (Hait and Moulik 2001). The detergent lyses the lipid barrier of the membrane and forms a detergent– lipid–protein complex at a concentration above its CMC (Lin and Guidotti 2009), indicating that the intact structure of the plasma membrane is a prerequisite for protein kinase activity. In support of this, a non-ionic detergent Tween-20, which is structurally different from TX-100 (Linke 2009), also inhibited phosphorylation of the H⁺-ATPase at a concentration above its CMC (Supplementary Fig. S5). It is noteworthy that blue light-induced ATP hydrolysis of the H⁺-ATPase in GCP extracts was also detected in the presence of ≤0.025% TX-100 (Kinoshita and Shimazaki 1999). These results suggest that the protein kinase activity for the H⁺-ATPase is closely correlated with H⁺-ATPase activity.

Dephosphorylation of the H⁺-ATPase in vitro was also detected in the plasma membrane from etiolated Arabidopsis seedlings, and was inhibited by EDTA, but not by calyculin A (Fig. 2A), suggesting that the biochemical properties of protein phosphatase are similar in both GCPs and etiolated seedlings. To characterize the protein phosphatase for the H⁺-ATPase, we further examined the divalent cation dependence of dephosphorylation of the H⁺-ATPase in the plasma membrane from etiolated Arabidopsis seedlings. The results showed that a Mg²⁺-dependent, Ca²⁺-independent PP2C is likely to be involved in dephosphorylation of the H⁺-ATPase (Fig. 2C). PP2C-type phosphatases are monomeric enzymes present in both prokaryotes and eukaryotes. Seventy-six Arabidopsis genes were identified as PP2C-type phosphatase candidates, and these fall into 10 groups, except for six genes that could not be clustered (Schweighofer et al. 2004). Members of this family are involved in the regulation of signaling pathways, e.g. ABI1, ABI2 and HAB1 for the ABA signaling pathway (Merlot et al. 2001, Robert et al. 2006, Moes et al. 2008, Rubio et al. 2009), AIP1 for the calcium-dependent signaling pathway involving calcineulin B-like calcium sensors (CBLs) and CBL-interacting protein kinase 23 (Lee et al. 2007), and KAPP for receptor kinase CLAVATA1 signaling (Trotochaud et al. 1999). However, the physiological functions of PP2Cs in Arabidopsis are largely unknown, except for the above PP2Cs. Further investigations are needed to determine which PP2C-type phosphatase is involved in the dephosphorylation of the H⁺-ATPase.

In contrast to the present results, Camoni et al. (2000) indicated that the type 2A protein phosphatase catalyzes dephosphorylation of the H⁺-ATPase in maize roots. However, calyculin A, an inhibitor of type 2A protein phosphatase, had no effect on the in vitro dephosphorylation of the H⁺-ATPase in either microsomes from Vicia GCPs or plasma membrane from etiolated Arabidopsis seedlings (Figs. 1C, 2A). Moreover, treatment of etiolated seedlings of Arabidopsis with calyculin A did not increase the phosphorylation level of the H⁺-ATPase (Fig. 2B). From these results, we conclude that type 2A protein phosphatase is not likely to dephosphorylate the H⁺-ATPase under our experimental conditions.

Analysis of in vitro dephosphorylation of H⁺-ATPase using 1% TX-100-insoluble fractions of the plasma membrane indicated that the protein phosphatase is likely to form a complex with the H⁺-ATPase (Fig. 4A). However, in vitro dephosphorylation of the H⁺-ATPase was not detected in the immunoprecipitate when we performed immunoprecipitation in the presence of 1% TX-100 using specific antibodies recognizing the conserved catalytic domain of the H⁺-ATPase, suggesting that direct binding of antibodies to the catalytic domain of the H⁺-ATPase may inhibit its dephosphorylation or dissociate the protein phosphatase from the H⁺-ATPase. Interestingly, we found that the H⁺-ATPase immunoprecipitate contains at least nine co-precipitated proteins (Supplementary Fig. S4). Kanczewska et al. (2005) demonstrated that activation of the H⁺-ATPase by phosphorylation of the penultimate threonine in the C-terminus and binding of the 14-3-3 protein converts a dimer of the H⁺-ATPase into a hexamer. However, the composition of the H⁺-ATPase complex is largely unknown except for the 14-3-3 protein (Fuglsang et al. 1999, Kinoshita and Shimazaki 1999, Svennelid et al. 1999, Camoni et al. 2000, Maudoux et al. 2000). Further investigations to identify the co-precipitated proteins in the H⁺-ATPase immunoprecipitate using mass spectrometry will elucidate the composition of the H⁺-ATPase complex.

In conclusion, we have shown that the protein kinase for phosphorylation of the penultimate threonine in the C-terminus of the H⁺-ATPase is insensitive to a potent kinase inhibitor K-252a, and that its activity is closely related to the structure of the plasma membrane. Moreover, the protein phosphatase is likely to be Mg²⁺-dependent PP2C-type phosphatase. The protein kinase and phosphatase possess similar biochemical properties in both guard cells and etiolated seedlings. It should be noted that blue light induced phosphorylation of the H⁺-ATPase in guard cells (Fig. 1A), but not in etiolated seedlings (Supplementary Fig. S6), although the same H⁺-ATPase isogenes, AHA1 and AHA2, were mainly expressed in both cell types (Supplementary Fig. S2, Ueno et al. 2005). These results suggest that regulation of protein kinase and phosphatase in etiolated seedlings is different from that in guard cells. Our results also indicate that a kinase and phosphatase pair co-localizes at least in part with the H⁺-ATPase in the plasma membrane and regulates the activation status of the H⁺-ATPase. This is the first evidence demonstrating the biochemical properties of the protein kinase and phosphatase in regulation of the phosphorylation status of the penultimate threonine in the H⁺-ATPase in plant cells.

Materials and Methods

Plant materials and growth conditions

Vicia faba (cv. Ryosai Issun) was grown hydroponically in a greenhouse at 22°C (Kinoshita and Shimazaki 1999). Arabidopsis thaliana (ecotype Columbia-0) seeds were grown on Murashige–Skoog (MS) plates containing 1% (w/v) sucrose in the dark at 24°C for 5 d.

Isolation of GCPs from V. faba

GCPs were isolated enzymatically from the abaxial epidermis of 3- to 5-week-old leaves of V. faba according to a previously described method (Kinoshita and Shimazaki 1999). Isolated GCPs were stored in the dark in 0.4 M mannitol and 1.0 mM CaCl₂ on ice until use. Protein concentrations were determined with the Bio-Rad protein assay kit.

Determination of phosphorylation levels of the H⁺-ATPase in GCPs

Phosphorylation levels of the plasma membrane H⁺-ATPase in GCPs of V. faba were determined as described previously (Tominaga et al. 2001), with some modifications. The GCPs (0.25 mg of protein ml⁻¹) were incubated in suspension buffer containing 5 mM MES-KOH pH 6.0, 10 mM KCl, 1 mM CaCl₂ and 0.4 M mannitol under red background light for 30 min at 24°C and then illuminated with a 30 s pulse of blue light, or FC at 10 μM was added to the GCP suspension. An aliquot of withdrawn GCP suspension was centrifuged briefly for the indicated times. The resulting GCP pellet was suspended in extraction medium containing 10 mM MOPS-KOH pH 7.5, 2.5 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride (PMSF), 10 μM leupeptin, 25 μg ml⁻¹ DNase I and 0.4% (w/v) TX-100. After incubation for 10 min at 24°C, the GCP extract was mixed with an equal volume of SDS buffer [3% (w/v) SDS, 30 mM Tris–HCl pH 8.0, 3 mM EDTA, 30% (w/v) sucrose, 0.012% (w/v) Coomassie Brilliant Blue, 15% (v/v) 2-mercaptoethanol] and used for SDS–PAGE. Phosphorylation levels of the H⁺-ATPase were detected by immunoblot using polyclonal antibodies against the phosphorylated Thr947 in the H⁺-ATPase of Arabidopsis (AHA2) and protein blot using GST–14-3-3, as described below.

Filters

Background red light at 600 μmol m⁻² s⁻¹ was obtained from a tungsten lamp (Philips EXR 300 W) by passing the light through a red glass filter (Corning 2-61). Blue light at 100 μmol m⁻² s⁻¹ was obtained from a tungsten lamp (Sylvania EXR 150 W) by passing the light through a glass filter (Corning 5-60). The photon flux density was measured with a quantum meter (Li-Cor, model 185A, Lincoln, NE, USA).

Antibodies

Polyclonal antibodies against the penultimate phosphorylated Thr947 of the plasma membrane H⁺-ATPase of Arabidopsis (AHA2) were raised in rabbits using the phosphorylated synthetic peptide (CIETPSHYpTV, where pT represents phosphorylated threonine) as an antigen (Medical & Biological Laboratories Co., Ltd., Nagoya, Japan). Polyclonal antibodies against the conserved catalytic domain of the plasma membrane H⁺-ATPase of Arabidopsis (AHA2) were raised in rabbits. The AHA2 DNA fragment was amplified from first-strand Arabidopsis cDNA with PCR using the specific primers 5′-GCCGGATCCATGGATGTCCTGTGCAGTGAC-3′ and 5′-GCCGGATCCTCAAGCACCACGAGCAGC-3′. The resulting amplified DNA fragment of 967–1,845 bp of AHA2, which contains BamHI sites at both ends, was cloned into the BamHI site of the pET30a vector (Merck, Darmstadt, Germany). The purified proteins from E. coli (BL21) were used as an antigen. The antiserum was used for immunoblots in Arabidopsis. Polyclonal antibodies against the plasma membrane H⁺-ATPase of V. faba (VHA1), the 14-3-3 protein (GF14ø) of Arabidopsis and phot1 of Arabidopsis were described previously (Kinoshita and Shimazaki 1999, Inoue et al. 2008).

Immunoblot

Immunoblot was performed according to Kinoshita and Shimazaki (1999) with some modifications. Following separation of proteins by SDS–PAGE in a 9% gel, individual proteins were transferred to nitrocellulose membranes at 1.5 mA cm⁻² in transfer buffer (48 mM Tris, 39 mM glycine, 20% methanol) by electroblotting (Trans-blot, Bio-Rad Laboratories, Hercules, CA, USA). The membranes were pre-incubated in blocking buffer for 0.5 h (0.05% Tween-20, 5% non-fat dry milk, 20 mM Tris–HCl, pH 7.4, 140 mM NaCl) and then reacted with polyclonal antibodies against VHA1, AHA2, phosphorylated Thr947 of AHA2, and 14-3-3 protein at a dilution of 1 : 3,000 in blocking buffer for 16 h at 4°C. The membrane was then rinsed three times for 5 min each in T-TBS (0.05% Tween-20, 20 mM Tris–HCl, pH 7.4, 140 mM NaCl) and reacted with a goat anti-rabbit IgG secondary antibody conjugated to alkaline phosphatase or horseradish peroxidase (HRP) (Bio-Rad Laboratories) at a dilution of 1 : 3,000 in blocking buffer at room temperature for 2 h. Development of the alkaline phosphatase reaction was performed by addition of 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium. Chemiluminescence with the HRP reaction using a chemiluminescence substrate (Pierce) was detected by LIGHT CAPTURE AE-2150 (ATTO, Tokyo, Japan).

Protein blot

Protein blot was performed according to Kinoshita and Shimazaki (1999) with some modifications. Briefly, the proteins were subjected to SDS–PAGE in a 9% gel and blotted onto nitrocellulose membranes. The membrane was incubated with blocking buffer for 0.5 h at room temperature and then reacted with 0.1 μM GST–14-3-3 in the blocking buffer for 16 h at 4°C. GST alone did not bind to the H⁺-ATPase (Kinoshita and Shimazaki 1999). After three washes with T-TBS, the membrane was incubated with anti-GST antibodies (GE Healthcare UK Ltd., Buckinghamshire, UK) at a dilution of 1 : 3,000 for 2 h at room temperature in blocking buffer. The membrane was then washed three times for 5 min each in T-TBS and reacted with anti-goat IgG secondary antibodies conjugated to alkaline phosphatase or HRP at a dilution of 1 : 3,000 for 2 h at room temperature in blocking buffer. Detection of the alkaline phosphatase reaction and the HRP reaction was as described above.

Preparation of microsomes and plasma membranes

The microsomes from GCPs of V. faba and plasma membranes from etiolated seedlings of A. thaliana were prepared using aqueous two-phase partitioning according to a previously described method (Kinoshita et al. 1995).

Preparation of the Triton X-100-insoluble fraction

The plasma membrane (at 0.5 mg of protein ml⁻¹) from etiolated Arabidopsis seedlings was suspended in extraction buffer containing 50 mM MOPS-KOH pH 7.5, 5 mM EDTA, 100 mM NaCl, 0.5 mM PMSF, 10 μM leupeptin, 2 mM dithiothreitol (DTT), 10 mM NaF, 0.5 mM ammonium molybdate, 50 nM calyculin A and 1% (w/v) TX-100. After incubation for 30 min on ice, the suspension was centrifuged at 100,000 × g for 60 min. The resulting pellet was used as the TX-100-insoluble fraction.

In vitro phosphorylation and dephosphorylation of the H⁺-ATPase

For in vitro phosphorylation, the microsomal membranes from GCPs of V. faba (10 μg of protein), plasma membranes from etiolated Arabidopsis seedlings (10 μg of protein) or TX-100-insoluble fractions from plasma membrane (10 μg of protein) were suspended in basal reaction buffer (25 μl) containing 50 mM MOPS-KOH, pH 7.2, 2.5 mM MgCl₂, 1 mM DTT and 0.025% (w/v) TX-100. The phosphorylation reaction was performed by addition of 2 mM ATP to the suspension for 60 min at 30°C. For in vitro dephosphorylation, the microsomal membranes (10 μg of protein), plasma membranes (10 μg of protein) or TX-100-insoluble fractions (10 μg of protein) were suspended in reaction buffer (25 μl) containing 50 mM MOPS-KOH, pH 7.2, 2.5 mM MgCl₂, 1 mM DTT and 0.1% (w/v) TX-100. The dephosphorylation reaction was performed for 30 min at 30°C. The reactions were terminated by the addition of 25 μl of SDS buffer, and the solubilized proteins were separated on a 10% SDS–polyacrylamide gel. Signal quantification was performed using a CS Analyzer Ver. 3 (ATTO, Tokyo, Japan).

Purification of recombinant 14-3-3 protein

To obtain the recombinant 14-3-3 protein (GF14ø) from Arabidopsis (At1g35160), it was expressed in E. coli. The full-length coding sequence of GF14ø was amplified from the first-strand cDNA by PCR using two primers (5′-GCCGGA TCCATGGCGGCACCACCA-3′ and 5′-GCCGGATCCTTAGATCTCCTTCTGTTCTTCAG-3′), and the resulting fragment, which contains BamHI sites at both ends, was cloned into the BamHI site of the pET30a vector. The proteins (His-GF14ø) purified from E. coli (BL21) using Ni-NTA–agarose (QIAGEN, Valencia, CA, USA) according to the manufacturer’s protocol were used for in vitro phosphorylation.

Immunoprecipitation of the H⁺-ATPase

Immunoprecipitation was performed according to a previously described method (Kinoshita et al. 2005) with modifications. Briefly, the plasma membranes from etiolated Arabidopsis seedlings (at 0.5 mg of protein ml⁻¹) were suspended in extraction buffer containing 50 mM MOPS-KOH, pH 7.5, 5 mM EDTA, 100 mM NaCl, 0.5 mM PMSF, 10 μM leupeptin, 2 mM DTT, 10 mM NaF, 0.5 mM ammonium molybdate, 50 nM calyculin A and 1% (w/v) TX-100, and then mixed with 5% (v/v) anti-AHA2 antibody immobilized to Dynabeads Protein A (Veritas, Tokyo, Japan). The mixture was incubated for 1 h at 4°C with gentle mixing. The immunoprecipitate was washed three times with 1 ml of ice-cold Tris-buffered saline and used for in vitro phosphorylation and dephosphorylation assays or SDS–PAGE.

Funding

This work was supported by the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan [Grant-in-Aid for Scientific Research (B) 20370021 to T.K.; Grant-in-Aid for Scientific Research for Priority Areas 21027014 and 20053007 to T.K., 17084005 to K.S.; Grant-in-Aid for Scientific Research (S) 21227001 to K.S.]; the Asahi Glass Foundation [to T.K.].

Abbreviations

Abbreviations
 
• AHA2

  H⁺-ATPase of Arabidopsis

 
• CMC

  critical micellar concentration

 
• DTT

  dithiothreitol

 
• FC

  fusicoccin

 
• GCP

  guard cell protoplast

 
• GST

  glutathione S -transferase

 
• HRP

  horseradish peroxidase

 
• PMSF

  phenylmethylsulfonyl fluoride

 
• PP2C

  type 2C protein phosphatase

 
• p-Thr

  antiphospho-threonine

 
• TX-100

  Triton X-100.