Mutagenic Analysis of Functional Residues in Putative Substrate-binding Site and Acidic Domains of Vacuolar H+-Pyrophosphatase*

Vacuolar H+-translocating inorganic pyrophosphatase (V-PPase) uses PPi as an energy donor and requires free Mg2+ for enzyme activity and stability. To determine the catalytic domain, we analyzed charged residues (Asp253, Lys261, Glu263, Asp279, Asp283, Asp287, Asp723, Asp727, and Asp731) in the putative PPi-binding site and two conserved acidic regions of mung bean V-PPase by site-directed mutagenesis and heterologous expression in yeast. Amino acid substitution of the residues with alanine and conservative residues resulted in a marked decrease in PPi hydrolysis activity and a complete loss of H+ transport activity. The conformational change of V-PPase induced by the binding of the substrate was reflected in the susceptibility to trypsin. Wild-type V-PPase was completely digested by trypsin but not in the presence of Mg-PPi, while two V-PPase mutants, K261A and E263A, became sensitive to trypsin even in the presence of the substrate. These results suggest that the second acidic region is also implicated in the substrate hydrolysis and that at least two residues, Lys261and Glu263, are essential for the substrate-binding function. From the observation that the conservative mutants K261R and E263D showed partial activity of PPi hydrolysis but no proton pump activity, we estimated that two residues, Lys261 and Glu263, might be related to the energy conversion from PPi hydrolysis to H+transport. The importance of two residues, Asp253 and Glu263, in the Mg2+-binding function was also suggested from the trypsin susceptibility in the presence of Mg2+. Furthermore, it was found that the two acidic regions include essential common motifs shared among the P-type ATPases.

Vacuolar H ϩ -pyrophosphatase (V-PPase) 1 belongs to the fourth class of electrogenic proton pump in addition to the P-, F-, and V-type ATPases. The proton pumping reaction couples with the hydrolysis of PP i . V-PPase acidifies vacuoles together with vacuolar H ϩ -ATPase in the plant cell and actively exports protons from the cytosol in the bacterial plasma membrane (1)(2)(3). V-PPase has the simplest structure among the proton pumps except for bacteriorhodopsin, a light-driven proton pump. The molecular mass calculated from the cDNA sequences range from 80 to 81 kDa for V-PPases of land plants and algae (for a review, see Refs. 3 and 4), while V-PPases in photosynthetic bacteria Rhodospirillum rubrum (5) and archaebacteria Pyrobaculum aerophilum (6) are relatively small. The simplicity of the enzyme structure and its substrate is an advantage to analyze the structure-function relationship. The enzyme activity is stimulated by K ϩ at relatively high concentrations. Also, Mg 2ϩ is essential to form a Mg-PP i complex and to keep the active conformation of V-PPase (1,7,8). Ca 2ϩ prevents formation of a Mg-PP i complex (8) and directly inhibits V-PPase (9). Thus, V-PPase should have a specific binding site for its substrate (Mg-PP i ), Mg 2ϩ , K ϩ , and Ca 2ϩ in addition to a H ϩ transport channel.
Multiple amino acid sequence alignment of V-PPases of various organisms revealed highly conserved regions (2,3,10). There is a putative substrate-binding motif of DXXXXXXXKXE in the cytoplasmic loop (1,11). This was supported by immunochemical study with an antibody specific to this sequence (DVGADLVGKVE) (12). This sequence is common among V-PPases not only from land plants but also from Chara corallina (10), Acetabularia acetabulum (13), R. rubrum (5), Thermotoga maritima (GenBank™ accession number AE001702), and P. aerophilum (6). Studies using substrate analogs, such as aminomethylenebisphosphonate, have also provided information on the catalytic domain (14 -16). Furthermore, the N-ethylmaleimide-binding cysteine residue (Cys 634 ) (17) and the N,NЈdicyclohexylcarbodiimde-binding residues (Glu 305 and Asp 504 ) (18) have been identified by a combination of site-directed mutagenesis of Arabidopsis V-PPase and heterologous expression in yeast.
The aim of this study is to clarify the substrate-binding site of V-PPase by the method of site-directed and random mutagenesis. We prepared a line of constructs, in which charged residues in a putative substrate-binding site were replaced, expressed in Saccharomyces cerevisiae, and then examined for enzymatic properties. V-PPase has been proposed to have three conserved regions (3,10). In addition to a putative PP i -binding site in the first conserved region, we investigated the two acidic motifs in the first and third conserved regions. Each aspartic acid residue in the two acidic regions was substituted and examined for enzymatic properties. Here, the functional roles of these residues on the substrate hydrolysis, binding of free Mg 2ϩ , and a coupling reaction between PP i hydrolysis and proton transport were examined. The similarity of the conserved functional motifs of V-PPase with the P-type ATPase is also discussed.

Heterologous Expression of Mung Bean V-PPase in Yeast Cell-A
EcoRI-SalI fragment of VVP2 cDNA encoding mung bean V-PPase (Ref. 19; DDBJ accession number AB009077) was inserted into a URA3marked, high copy (2 m) yeast expression vector pKT10 (20,21). The obtained pKVVP2 plasmid was introduced into a S. cerevisiae strain BJ5458, which was deficient in major vacuolar proteinases (22), by the lithium accetate/single-stranded DNA/polyethylene glycol transformation method (23). Positive Uraϩ colonies were selected, and the expression of V-PPase was confirmed by immunoblotting with the anti-V-PPase antibodies previously prepared (24).
Plasmid Preparation for V-PPase Mutants-Site-directed mutagenesis was performed using a QuickChange site-directed mutagenesis kit (Stratagene) by the method of Kirsch and Joly (25). Mutagenic and antisense standard primers used in this study are listed in Table I. The DNA sequences of at least two independent plasmids for each mutant were determined to confirm the mutation points.
Preparation of a Random Mutation Library and Screening of V-PPase Mutant-For convenience of genetic manipulation, the AatII site of pKVVP2 plasmid in pKT10 vector was removed, and a SacI site at position 715 in the plasmid was introduced by substitution with a synonymous codon. The resulting pKVVP2/S plasmid was used as a seed for hypermutagenic polymerase chain reaction (26). The reaction mixture contained 40 pg/ml pKVVP2/S plasmid, 250 nM primers (Table  I, F668 and R1251), 10 mM Tris-HCl, pH 8.3, 50 mM KCl, 0.001% gelatin, 2.5 mM MgCl 2 , 0.5 mM MnCl 2 , 0.1 mM dATP, 0.1 mM dCTP, 1 mM dGTP, 1 mM dTTP, and 0.08 units/ml Taq DNA polymerase. Amplification was done using 20 cycles of a set of 30 s at 95°C, 45 s at 55°C, and 5 min at 72°C. The obtained polymerase chain reaction fragments were digested and inserted into the SacI-AatII site of the pKVVP2/S plasmid. This library was amplified in E. coli and introduced into yeast BJ5458. A region that was mutagenized in each mutant was amplified by polymerase chain reaction using F668 and R1251 primers. Mutation points were defined by DNA sequencing. In most mutants, plural sites were mutated.
Preparation of V-PPase Enriched Membrane Fraction-Crude membrane fractions were prepared from yeast cells by the method of Kim et al. (27) with a few modifications. Yeast cells were precultured at 30°C for 2 days in AHCW/Glc medium that contained 50 mM potassium phosphate buffer, pH 5.5, 0.002% (w/v) adenine sulfate, 0.002% tryptophan, 2% glucose, 1% casamino acids (Nihon Pharmaceutical Co.), and 0.67% yeast nitrogen base without amino acids (Difco). The cell culture was diluted 64-fold and then grown for 12 h to reach an exponential phase. After being washed with 0.1 M Tris-HCl, pH 9.4, 50 mM 2mercaptoethanol, and 0.1 M glucose at 30°C for 10 min, cells were treated with a zymolyase medium at 30°C for 1 h with gentle agitation. The medium contained 0.05% zymolyase 20T (Seikagaku Kogyo Co.), 0.9 M sorbitol, 0.1 M glucose, 50 mM Tris-Mes, pH 7.6, 5 mM DTT, 0.043% yeast nitrogen base without amino acids and ammonium sulfate, and 0.25ϫ dropout solution composed of all amino acids and adenines (28). Spheroplasts were collected from the suspension by centrifugation at 3,000 ϫ g for 10 min and washed with 1 M sorbitol.
The spheroplasts were resuspended in 50 mM Tris-ascorbate, pH 7.6, 1.1 M glycerol, 1.5% polyvinylpyrrolidone (M r 40,000), 5 mM EGTA-Tris, 1 mM DTT, 0.2% bovine serum albumin, 1 mM PMSF, and 1 mg/liter leupeptin and then homogenized with a motor-driven Teflon homogenizer. After centrifugation at 2,000 ϫ g for 10 min, the precipitate was suspended in the same buffer and centrifuged again. All of the supernatant fractions were pooled and centrifuged at 120,000 ϫ g for 30 min. The precipitate (membranes) was suspended in 15% (w/w) sucrose and layered on a 35% (w/w) sucrose solution. Both sucrose solutions con-

FIG. 1. Heterologous expression of V-PPase in yeast.
A, the mung bean V-PPase cDNA was linked with the GAP promoter (pKVVP2). A transformant containing the pKT10 vector was used as a negative control. B, both constructs of wild type (WT) and vector (V) were transformed and expressed in S. cerevisiae. The membrane fractions enriched with V-PPase were prepared and then subjected to SDS-polyacrylamide gel electrophoresis (left) and subsequent immunoblot analysis with the anti-V-PPase antibody (right). The arrowheads indicate the position of V-PPase. C, the membrane vesicles prepared from S. cerevisiae strains that expressed V-PPase (WT) and a vector (V) were assayed for the PP i hydrolysis activity. The reaction medium contained 1 mM Na 4 PP i , 1 mM MgSO 4 , 50 mM KCl, 1 mM sodium molybdate, 0.02% Triton, X-100, and 30 mM Tris-Mes, pH 7.2 in the presence (right) or absence (left) of 0.5 mM KF. In an experiment (Na), NaCl at 50 mM was used instead of KCl. Released P i was measured colorimetrically. D, the membrane vesicles (200 g) were also assayed for the PP i -dependent H ϩ transport activity with 1 M acridine orange. Reaction was started by the addition of PP i (1 mM). The initial rate of fluorescence quenching was defined as the H ϩ transport activity.

TABLE I
A list of the pairs of mutagenic primer and antisense standard primer used for amino acid substitution of mung bean V-PPase The nucleotides changed for the substitution are underlined.

Name
Primer sequence Name Protein and Enzyme Assays-Protein content was determined by the method of Bradford (29). PP i hydrolysis activity was measured in a reaction medium supplemented with 0.5 mM KF (24). PP i -dependent H ϩ transport activity was determined as described previously (24) with a few modifications. Membrane preparations were preincubated with a reaction buffer containing 0.3 M sorbitol, 5 mM Tris-Mes, pH 7.6, 0.1 M KCl, 0.5 mM EGTA-Tris, 1.5 mM MgCl 2 , 0.2% bovine serum albumin, and 1 M acridine orange, and then the reaction was initiated by the addition of 1 mM Na 4 PP i .
Immunoblotting-Proteins were separated by SDS-polyacrylamide gel electrophoresis on 10% gels and transferred to a polyvinylidene difluoride membrane by the standard procedure. Immunoblotting was carried out using polyclonal antibodies against V-PPase purified from mung bean (24) and an ECL procedure (Amersham Pharmacia Biotech).
Trypsin Digestion Analysis-To remove Mg 2ϩ and PMSF, yeast membrane preparations were washed with 20 mM Tris-Mes, pH 7.6, 20% glycerol, 50 mM KCl, 0.05 mM MgCl 2 . The membrane fraction was mixed with an equal volume of 20 mM Tris-Mes, pH 7.6, 10 g/ml trypsin, 20% glycerol, 50 mM KCl, 1 mM DTT, and 1% MOA Excellent (a mixture of alkylglucosides; Kao Co., Japan) and then incubated at 30°C for 40 min. The reaction was stopped by the addition of an SDS-sample buffer containing 5 mM PMSF. The samples were subjected to SDSpolyacrylamide gel electrophoresis and subsequent immunoblotting with anti-V-PPase antibody.

Expression of Mung Bean V-PPase in Yeast as a Functional
Enzyme-A yeast expression vector pKT10 was used to express a cDNA (VVP2) encoding mung bean V-PPase in yeast (Fig.  1A). For DNA manipulation, an EcoRI site in VVP2 was eliminated by substitution with synonymous codons. VVP2 was efficiently transcribed under the control of the promoter of glyceraldehyde-3-phosphate dehydrogenase in a medium supplemented with glucose (30). The V-PPase-enriched membrane fraction was prepared by a stepwise sucrose gradient centrifugation and then subjected to immunoblotting with the polyclonal antibodies specific to V-PPase. Transformants produced a 73-kDa protein that was reacted with the antibodies (Fig.  1B). The recombinant V-PPase accounted for 2 to 3% of the total membrane protein. The membranes showed the PP i hydrolysis activity that was insensitive to KF, an inhibitor of acid phosphatase (Fig. 1C). The activity was 3.5 times enhanced by 50 mM KCl but not by 50 mM NaCl. The membrane vesicles prepared from VVP2 transformant gave the PP i -dependent H ϩ transport activity (Fig. 1D). The pH gradient was collapsed by the addition of membrane-permeable ammonium ion, indicating the electrogenic H ϩ transport in the vesicles. The elimination of an EcoRI site from the original VPP2 sequence did not affect the expression level and the enzymatic activity. The results indicate that V-PPase expressed in yeast functions normally in the heterologous system.
A Series of V-PPase Mutants in Respect to Putative Functional Motifs-The sequence DVGADLVGKVE is a putative substrate-binding site of V-PPase (2,11,31). This motif has been demonstrated to be exposed to the cytosol (12). To evaluate the functional significance of the charged residues in this motif, we generated a series of V-PPase mutants, in which the residues were replaced with alanine (D253A, K261A, E263A) or conservative residues (D253E, K261R, E263D).
A comparison of the primary structures of V-PPases of various organisms revealed that there are two consensus acidic regions, DNVGDNVGD (acidic region 1) and DTXGDPXKD (acidic region 2). The former is near the DVGADLVGKVE motif in loop e, and the latter is in loop m between the 13th and 14th transmembrane domains (Fig. 2). Both regions can be expressed as a common motif DXXXDXXXD. To test whether the aspartate residues in the two motifs are implicated in the interaction with the substrate and Mg 2ϩ , we generated a series of mutants substituted with alanine (D279A, D283A, D287A, D723A, D727A, and D731A) and glutamic acid (D279E, D283E, D287E, D723E, D727E, and D731E). These V-PPase mutants were assayed for the enzymatic activities.
Enzymatic Activities of V-PPase Mutants-The V-PPase protein in each mutant was clearly detected as a 73-kDa protein in an immunoblot (Fig. 3A). PP i hydrolysis and PP i -dependent H ϩ transport activities of the mutants were assayed at 1 mM (Fig.  3, B and D) or 0.3 mM Mg 2ϩ (Fig. 3C). The membrane preparation of the wild-type strain showed high activities compared with a low basal level in the control vector. All nine V-PPase mutants substituted with alanine (D253A, K261A, E263A, D279A, D283A, D287A, D723A, D727A, and D731A) lost the activity completely (Fig. 3, B-D). Conservative exchange with glutamate in seven mutants (D253E, D279E, D283E, D287E, D723E, D727E, and D731E) also caused a loss of function of H ϩ transport (Fig. 3, B and C). The PP i hydrolysis activity was decreased in these mutants. As a control, we obtained a V262A mutant from a random mutation library. This V262A mutant had the same V-PPase activity and protein level as the wild type. These results proved that aspartate residues in the putative PP i -binding site and the acidic motifs are involved in the binding and hydrolysis of the substrate. There was no difference in the protein level of V-PPase in the membranes among the mutants (Fig. 3A). Thus, the loss of activity was not due to the absence or low expression of V-PPase.
Interestingly, the K261R and E263D mutants retained 25 and 50% of the original PP i hydrolytic activity, respectively, at 1 mM MgCl 2 , but they had no H ϩ transport activity. Thus, Lys 261 and Glu 263 may be involved in the energy transduction from PP i hydrolysis to H ϩ translocation. Furthermore, three other mutants (E263G, V259A, and C304R) with a single substitution of an amino acid were obtained by a random mutation technique. The E263G and C304R mutants gave no activity of PP i hydrolysis or H ϩ transport (Fig. 4). The V259A mutant had 60% of the original activity of PP i hydrolysis but no H ϩ transport activity as well as K261R and E263D.
Mg 2ϩ and Substrate Binding Properties in V-PPase Mutants-V-PPase requires Mg 2ϩ for activation, structural stabilization, and protection from protease digestion (1,7,8). In this study, yeast vacuolar membranes containing V-PPase were treated with trypsin in the presence or absence of Mg 2ϩ , PP i , and Mg-PP i . Trypsin preferentially cleaves at the carboxyl sides of arginine and lysine residues. As shown in Fig. 5A, Mg 2ϩ and Mg-PP i , but not PP i , at relatively high concentrations partially prevented digestion of V-PPase by trypsin. Thus, the trypsin susceptibility is a good marker for the structural change caused by the binding of Mg 2ϩ and Mg-PP i .
All V-PPase mutants were thoroughly digested by trypsin in the absence of Mg 2ϩ (Fig. 5B, panels none and 1 mM PP i ). Therefore, the amino acid substitution did not affect the trypsin cleavage site. It should be noted that Mg 2ϩ protected the enzyme from trypsin digestion in mutants of K261A, D279A, D283A, D287A, D723A, D727A, D731A, E263D, and D727E as compared with the wild type V-PPase (Fig. 5B, 1 mM Mg). Neither Mg 2ϩ nor PP i has a direct inhibitory effect on trypsin, since V-PPase in the E263A mutant was completely digested even in the presence of 1 mM Mg 2ϩ and 1 mM PP i (Fig. 5B). This observation indicates that these amino acid substitutions in- In the presence of both Mg 2ϩ and PP i , certain mutants were resistant to trypsin as was the wild type V-PPase. However, the K261A, E263A, and K261R mutants were sensitive to trypsin, and the other alanine mutants (D279A, D723A) and the glutamate mutants (D253E, D279E, D283E, D287E, and D723E) were partially sensitive under the assay conditions. These results suggest that Lys 261 and Glu 263 are essential for the substrate-binding function. The other aspartate residues at 253, 279, 283, 287, and 723 may also be implicated in the substratebinding function or located near the substrate-binding site of V-PPase. DISCUSSION V-PPase has been estimated to possess the binding sites for its substrate (Mg-PP i ), Mg 2ϩ , K ϩ , and Ca 2ϩ , independently, and a H ϩ transport pathway. In the present study, we examined the functional role of charged residues in the putative substrate-binding site (DVGADLVGKVE) and two conservative acidic regions of V-PPase by site-directed and random mutagenesis in combination with a heterologous expression system in S. cerevisiae. The V-PPase expressed in yeast exhibited the PP i hydrolysis and proton pump activities, and the enzyme protein was detected by immunoblotting. V-PPase translated in yeast was considered to be localized mainly in vacuolar membranes judging from the following three observations: vacuolar type H ϩ -ATPase was detected in the same membrane fraction; a PP i -dependent H ϩ current was detected in intact vacuoles prepared from the VVP2-transformed yeast cells by the patch clamp technique 2 ; and the amount of the V-PPase protein in a yeast strain that lacked the major vacuolar proteases was higher than that in the normal strain.
Role of a DVGADLVGKVE Motif and Acidic Regions in Enzyme Activity-The substitution of acidic residues in the putative substrate-binding motif and two acidic regions had a negative effect on the V-PPase activity (Fig. 3). This is a direct effect of amino acid substitution, since all V-PPase mutants expressed in yeast were accumulated in the membrane at equal levels. The PP i hydrolysis activity was markedly decreased even in the case of conservative substitution (D253E, K261R, and E263D), although the V262A mutant retained the original activity. Thus, these residues (Asp 253 , Lys 261 , and Glu 263 ) are essential for V-PPase activity.
The present study showed that V-PPase might not interact with PP i in the absence of Mg 2ϩ , since PP i did not affect trypsin 2 Y. Nakanishi, I. Yabe, and M. Maeshima, unpublished data. susceptibility even at 1 mM, but PP i plus Mg 2ϩ makes V-PPase resistant to trypsin (Fig. 5A). Furthermore, the trypsin susceptibility assay revealed that Lys 261 and Glu 263 are essential for the binding of the substrate (Fig. 5). Fig. 6B shows a scheme of our model. This work demonstrated the functional significance of acidic residues in the acidic regions (DNVGDNVGD 287 and DTXGD-PXKD 731 ) (Fig. 2). The sequences are conserved among V-PPases of various organisms including R. rubrum (5) and T. maritima (GenBank TM accession number AE001702). Substitution of the aspartate residue at 279, 283, 287, 723, 727, and 731 with alanine or glutamate residue resulted in the complete loss of the activity (Fig. 3). The digestion assay with trypsin also supported the implication of aspartate residues at 279, 283, 287, and 723 in the binding of Mg-PP i (Fig. 5). These residues may be adjacent to the DVGADLVGKVE motif in the tertiary structure of V-PPase as discussed later.
The present study also suggested the residues involved in energy coupling of V-PPase. The K261R and E263D mutants partially retained the PP i hydrolysis activity, 25 and 50% of that of the wild-type enzyme, respectively, but the mutants did not exhibit the H ϩ transport activity (Fig. 3). The V259A mutant also retained partial activity (60%) of PP i hydrolysis but no proton pump activity (Fig. 4). These three residues are located in the putative PP i -binding site. Thus, the conserved motif including Val 259 , Lys 261 , and Glu 263 may be implicated in the initial step of the energy transfer from PP i hydrolysis to the H ϩ translocation. In Arabidopsis V-PPase, Glu 427 in the transmembrane domain has been demonstrated to be involved in the energy coupling by the site-directed mutagenesis (18). The role of several charged residues in the transmembrane domains remains to be investigated for their role in H ϩ translocation.
Residues Involved in Binding of Free Mg 2ϩ -The presence of Mg 2ϩ decreased the susceptibility of V-PPase to trypsin (Fig.  5A), which is consistent with a previous report (8). Interestingly, V-PPase became resistant to trypsin in several V-PPase mutants (K261A, D279A, D283A, D287A, D723A, D727A, D731A, E263D, and D727E) compared with the wild type V-PPase, while the D253A and E263A mutants were completely digested by trypsin (Fig. 5). This suggests that the two residues Asp 253 and Glu 263 are essential for Mg 2ϩ binding (Fig. 6). On the other hand, Lys 261 and other aspartate residues in the acidic regions seem to interfere with V-PPase in the binding of Mg 2ϩ (Fig. 6B, broken lines). Probably, these aspartate residues weaken the interaction between Mg 2ϩ and a binding site that includes Asp 253 and Glu 263 (Fig. 6B). Since Asp 253 and Glu 263 may be involved in the binding of a Mg-PP i complex as mentioned above, the two acidic residues may be located in the substrate-binding pocket and also have an ability to interact with free Mg 2ϩ . Probably, there is another, high affinity binding site for free Mg 2ϩ independent of the substrate-binding site. However, those residues could not be detected in the present assay system of trypsin susceptibility in the presence of 1 mM Mg 2ϩ .
Common Motifs in Acidic Regions among V-PPase and P-type ATPase-With respect to the conserved acidic regions, we found that V-PPases share common motifs with the P-type ATPases such as Ca 2ϩ -ATPase. In the P-type ATPase, the ATP-binding and ATP-hydrolyzing domain has been proposed to consist of several conserved motifs of DKTGT, DPPR (or DKVR), (T/S)GD(N/K), and GDGXNDA (32,33). Among these motifs, the TGDN and GDGXNDA motifs are conserved in V-PPases as a GDN motif including Asp 283 in the first acidic region and a GDTIGD motif including Asp 723 and Asp 727 in the second acidic region (Fig. 7). It has been proposed that the TGDN motif is involved in ATP hydrolysis together with a DPPR motif in the P-ATPases and that the adenosine moiety of ATP interacts with the other motifs such as a KGAP motif of the P-ATPase (32,33). It has been recently reported that the amino acid residues that are involved in the binding of adenosine moiety and the hydrolysis of phosphoanhydride bond of ATP can be topologically distinguished in the crystal structure of sarcoplasmic reticulum Ca 2ϩ -ATPase (34). The DPPR (or DKVR) motif, which is located in the hydrolysis pocket, is present as DDRR 272 in loop e and DNAK 695 in loop m of V-PPases except for A. acetabulum V-PPase (3,13), but an adenosine-binding KGAP motif is not conserved in the V-PPase. Aravind et al. (35) have proposed that two aspartate residues in the GDGXNDX motif of the P-ATPases have a role in hydrolysis of a phosphoanhydride bond of ATP. The present study revealed not only the importance of the two acidic regions in addition to the PP i -binding motif proposed previously but also the functional and sequence similarity of V-PPases to the Ptype ATPases. However, it cannot be concluded that V-PPase is a member of the P-type ATPase family, since V-PPase lacks a phosphorylation domain (DKTGTLT) that is common to the P-ATPases including plasma membrane H ϩ -ATPase, Ca 2ϩ - ATPase and other heavy metal-transporting ATPase (32). At present, it is unclear whether V-PPase forms an intermediate phosphorylation form during PP i hydrolysis.
In summary, Lys 261 and Glu 263 of mung bean V-PPase are essential for the substrate-binding function, and Asp 253 and Glu 263 are essential for the Mg 2ϩ -binding function (Fig. 6). All aspartate residues (Asp 253 , Asp 279 , Asp 283 , Asp 287 , Asp 723 , Asp 727 , and Asp 731 ) in the PP i -binding motif and two acidic regions may be involved in interaction with the substrate. It has also been reported that Glu 305 and Asp 504 of Arabidopsis V-PPase, Glu 301 and Asp 500 for mung bean V-PPase, respectively, are essential for PP i hydrolysis activity (18). Thus, we propose that these two acidic regions (279 -287 and 723-731) and a common DXXADLVGKXE (253-263) motif form a core catalytic domain of the V-PPase together with a few other motifs. To verify our model, we need to perform a high resolution crystallographic study to determine the structure.