Probing the Catalytic Subunit of the Tonoplast H+-ATPase from Oat Roots BINDING OF 7-CHLORO-4-NITROBENZO-2-OXA-1,3,-DIAZOLE TO THE 72-KILODALTON POLYPEPTIDE*

The purified tonoplast H*-ATPase from oat roots (Avena sativa L. var. Lang) consists of at least three different polypeptides with masses 72,60, and 16 kDa. We have used covalent modifiers (inhibitors) and polyclonal antibodies to identify the catalytic subunit of the H+-pumping ATPase. The inactivation of ATPase activity by 7-chloro-4-nitrobenzo-2-oxa-1,3-dia- zole (Nbd-C1, an adenine analog) was protected by MgATP or MgADP, and showed kinetic properties consistent with active site-directed inhibition. Under similar conditions, [14C]Nbd-C1 preferentially labeled the 72-kDa polypeptide of the purified ATPase. This binding was reduced by MgATP or 2’ (3’)-)0-(2,4,6-trinitrophenyl) ATP. Nbd-C1 probably modified cysteinyl-SH or tyrosyl-OH groups, as dithiothreitol reversed both ATPase inactivation and I“C]Nbd-Cl binding to the 72-kDa subunit. The finding that N- ethylmaleimide inhibition of ATPase activity was pro-tectable by nucleotides is consistent with the idea of sulfhydryl groups in the ATP-binding site. Polyclonal antibody made to the 72-kDa polypeptide specifically reacted (Western blot) with a 72-kDa polypeptide from 72-lzDa Subunit-An-tibodies purified purified holoenzyme separated on preparative 10% SDS-polyacrylamide gels. The position of the was determined by staining alignment strips, 72-kDa polypeptide unstained portion dialyzed lyophilized.

Probing the Catalytic Subunit of the Tonoplast H+-ATPase from Oat Roots BINDING OF 7-CHLORO-4-NITROBENZO-2-OXA-1,3,-DIAZOLE TO THE 72-KILODALTON POLYPEPTIDE* (Received for publication, July 22, 1986) Stephen K. Randall  The purified tonoplast H*-ATPase from oat roots (Avena sativa L. var. Lang) consists of at least three different polypeptides with masses 72,60, and 16 kDa. We have used covalent modifiers (inhibitors) and polyclonal antibodies to identify the catalytic subunit of the H+-pumping ATPase. The inactivation of ATPase activity by 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole (Nbd-C1, an adenine analog) was protected by MgATP or MgADP, and showed kinetic properties consistent with active site-directed inhibition. Under similar conditions, [14C]Nbd-C1 preferentially labeled the 72-kDa polypeptide of the purified ATPase. This binding was reduced by MgATP or 2' (3')-)0-(2,4,6trinitrophenyl) ATP. Nbd-C1 probably modified cysteinyl-SH or tyrosyl-OH groups, as dithiothreitol reversed both ATPase inactivation and I"C]Nbd-Cl binding to the 72-kDa subunit. The finding that Nethylmaleimide inhibition of ATPase activity was protectable by nucleotides is consistent with the idea of sulfhydryl groups in the ATP-binding site. Polyclonal antibody made to the 72-kDa polypeptide specifically reacted (Western blot) with a 72-kDa polypeptide from both tonoplast-enriched membranes and the purified tonoplast ATPase, but it did not cross-react with the mitochondrial or Escherichia coli F,-ATPase. The antibody inhibited tonoplast ATPase and H*-pumping activities. We conclude from these results that the 72-kDa polypeptide of the tonoplast H+-ATPase contains an ATP-(or nucleotide-) binding site that may constitute the catalytic domain.
In higher plant cells, the vacuole maintains and regulates cell turgor, and consequently controls cell expansion. This organelle transports and stores ions and metabolites (1,2). Recent studies have shown that an anion-sensitive ATPase on the tonoplast (vacuolar membrane) pumps H+ into the vacuole (3), providing the proton-motive force for transport of various solutes, such as Ca2+ (4) and anions (5)(6)(7).
The tonoplast ATPase from oat roots has been well characterized. It is resistant to orthovanadate, stimulated by C1-, and inhibited by NO; (8)(9)(10)(11). Both H+ pumping and ATPase activities are sensitive to inhibition by NEM,' DCCD, Nbd-C1, and DIDS (10,11). The tonoplast ATPase is insensitive to azide or oligomycin (10,ll). These properties are generally shared by the vacuolar ATPase of fungi (12) and a H+-ATPase associated with clathrin-coated vesicles (13) and the chromaffin granule (14). These H+ pumps may represent a third class of ATPases which differ from the ElE2-type (plasma membrane) and the FIFo-type ATPases.
To identify the catalytic subunit of the tonoplast H+-ATPase from oat roots, we have examined NEM and Nbd-C1 for their effectiveness as active site-directed inhibitors. Nbd-C1 has been a useful probe for the identification and characterization of the nucleotide-binding sites in FIFO-and EIE,type ATPase (22,23). Here we show that the Nbd-C1 inhibition of ATPase activity follows inactivation kinetics consistent with a site-specific covalent modification by Nbd-C1. ["C] Nbd-C1 specifically labeled the 72-kDa polypeptide and this binding is protected by a competitive inhibitor, TNP-ATP. These results provide evidence that the 72-kDa subunit contains the substrate-binding site.

MATERIALS AND METHODS
Plant Material-Oat seeds (Avena sativa L. var. Lang) were germinated over an aerated solution of 0.5 mM CaS04. Roots were harvested after 3-5 days of growth.

7135
and was purified (11) with the following modifications. The solubilized preparation was separated with Sepharose CL-GB, instead of Sepharose 4B, to obtain greater flow rates in the presence of glycerol.
As ATP and DTT prevented Nbd-C1 inhibition of the ATPase, they were omitted from the solubilization and elution buffers. MgSO, (0.5 mM) was added to stabilize the enzyme activity. Although this modified procedure did not change the apparent mass of the tonoplast ATPase, it did result in the appearance of a few additional polypeptides. Fractions with the highest specific ATPase activities were utilized (11).
ATPase and Protein Assays-ATPase activity was measured either by a coupled assay (24) or by Pi release (25). In the former assay, ADP production was monitored spectrophotometrically by measuring NADH oxidation in a coupled lactate dehydrogenase-pyruvate kinase reaction and an ATP-regenerating system (24) in a 1.0-ml reaction mixture (11). Protein concentration was determined according to Lowry (26) after precipitating the proteins twice with 7.5% trichloroacetic acid to remove glycerol (27).
Preparation of Polyclonal Antibodies to the 72-lzDa Subunit-Antibodies were raised in New Zealand White rabbits. To obtain the purified 72-kDa subunit, the purified holoenzyme (11) was separated on preparative 10% SDS-polyacrylamide gels. The position of the subunit was determined by staining alignment strips, The 72-kDa polypeptide was cut out and electroeluted from the unstained portion of the gel. The antigen was dialyzed against water and lyophilized. Rabbits were immunized with 20-40 gg of protein by intramuscular injection (30). The first and second immunizations contained Freund's complete and incomplete adjuvant respectively, while subsequent injections were in PBS. Booster shots were given at 3-week intervals. After the titer built up (about 5 months), blood was collected and the serum obtained was stored at -70 "C. The antiserum was partially purified with a one-step batch DEAE-cellulose absorption (30) procedure.
Zmmunoblotting-After separating polypeptides by SDS-PAGE, the unstained gels were blotted onto diazophenylthioether paper (31,32). The paper was incubated with antiserum to 72-kDa polypeptide (1:200 dilution) for 3 h at 22 "C; the blots were thoroughly washed (32), and probed with goat anti-rabbit IgG conjugated to alkaline phosphatase. 5-Bromo-4-chloro-3-indolyl phosphate in 0.1 M Tris-C1, 1 mM MgC12, and 1.0% agarose at pH 8.3 was utilized for the detection of alkaline phosphatase activity. Reactive bands were visualized as a dark blue color due to the product, indigo.

RESULTS~
NEM and Nbd-Cl Inhibition of ATPase Activity Is Protectable by ATP and ADP-We found previously that both NEM and Nbd-C1 inhibition of the tonoplast H+-ATPase activity were protected by the substrate, MgATP (10,11). These preliminary results suggested that both inhibitors were potentially useful probes for identifying the catalytic subunit. The inhibition by NEM and Nbd-C1 were examined in greater detail. Tonoplast membranes were pretreated with NEM or * Figs Nbd-C1 at different concentrations, temperature, pH, or time. The pretreatments were terminated by diluting 100-fold and ATPase activity was measured at pH 7.0 and at 23 "C.
Inhibition by NEM was pH dependent with maximum sensitivity at pH 8.0 (Fig. 1). This suggests that NEM may be reacting with the unprotonated sulfhydryl group of a cysteine residue as suggested for NEM inhibition of the plasma membrane ATPase from N. crassa (34). NEM (at pH 8) inhibited ATPase activity with a IW of 2-3 p~ (Fig. 2 A ) . At this pH, the protection by MgATP was relatively small compared to pH 7 where the IW was 20 p~ (Fig. 2A).
The reactivity of Nbd-C1 with the tonoplast ATPase was also strongly dependent on both pH and temperature. Fig. 3 shows the loss of ATPase activity as a function of Nbd-C1 concentration at pH 7.0 and 4 "C. Under these conditions, the ATPase was about 50-fold less sensitive to inhibition than at pH 8 (and 23 "C) where the I,, was 0.8 pM (Fig. 2B). It is probable that the increased inhibition is due to the heightened reactivity of Nbd-C1 towards amino groups at alkaline pH (35, 36). Although the inhibition by Nbd-C1 is protected by ATP, inhibition may not be site-directed under these conditions (see below). Therefore all subsequent incubations of Nbd-C1 with the ATPase were conducted at pH 7.0 and 4 "C.
Inhibition by NEM (Fig. 4A) and Nbd-C1 ( Tonoplast membranes were preincubated in 2.5 mM Hepes-BTP, 10 mM BTP-Cl for 30 min at pH 7.0 and 4 "C. Pretreatments were terminated by dilution into the coupled assay medium (as in Fig. 1). This implied that ATP binding to the catalytic site protects against inhibition by Nbd-C1. MgATP at 0.4 mM protected approximately 50% of the inhibition due to 100 p~ NEM (Fig. 4A). In contrast to NEM, Nbd-C1 inhibition was not as well protected by MgADP (Fig. 4B). Adenosine conferred little protection against inhibition by either NEM or Nbd-C1 (Fig. 4).
Kinetics of Nbd-Cl and NEM Inhibition of ATPase Actiuity-To determine if Nbd-C1 and NEM were valid probes for identifying a catalytic subunit of the ATPase, we conducted a preliminary study to test for site-specific inhibition. In sitespecific inhibition, the initial binding of the inhibitor to the enzyme is due to its structural analogy to the physiological substrate (ATP). If the inhibitor is bound at the active site prior to covalent modification, the loss of activity as a function of time should show pseudo first order kinetics, and the rate of inactivation should reach a maximum with increasing concentrations of inhibitor (38,39). If the presence of substrate retards the rate of inactivation, then a simple interpretation is that the substrate and inactivator compete for binding at the enzyme's active site.
Tonoplast vesicles were incubated with Nbd-C1 at 4 "C for various times, pH 7.0. ATPase activity was inhibited 33% by 15 p~ Nbd-C1 in the absence of ATP. This inhibition was almost completely protected by ATP. The log activity plotted as a function of time, decreased linearly for 40 min (Fig. 5). At 15 p~ Nbd-C1, this loss of activity showed a rate constant of 9.96 X min-l in the absence of MgATP, and 9.3 X min" in the presence of MgATP. The inactivation rate constant (kinas) showed apparent saturation at high concentrations of Nbd-C1 (Fig. 6). A double-reciprocal plot (l/kinaCt uersus l/Nbd-Cl) yielded a kd of 36 p~ (representing the affinity of Nbd-C1 for the ATPase) and a limiting (maximum) inhibition rate constant of 52 X min" (Fig. 6, inset). These results suggest that Nbd-C1 inactivates the ATPase as an active site-directed inhibitor. In contrast, NEM inhibition displayed more complex kinetics. The rates of inactivation became nonlinear by 20 min at low concentrations of NEM (3 p~) (Fig. 7), indicating NEM may modify several sites on the ATPase.
TNP-ATP Is a Competitive Inhibitor of the Tonoplast ATPase-In an effort to find another inhibitor that would also bind at the substrate-binding site, we examined the effect of TNP-ATP on ATPase activity. TNP-ATP is an ATP analog that is very slowly hydrolyzed and is an extremely potent competitive inhibitor of the mitochondrial F,-ATPase (40). This ATP analog was useful in characterizing nucleotidebinding sites on other ATPases (41). We found that TNP-ATP is a more potent competitive inhibitor of the tonoplast polypeptide was consistently reduced when the ATPase was preincubated in the presence of MgATP. In the experiment shown, Nbd-C1 binding was reduced 33% by 3 mM ATP and 86% by 10 PM TNP-ATP (Fig. 9A). The protection by MgATP (33%) was less than might be expected from the near complete protection of activity (Figs. 4B and 5). A large component (67%) of the Nbd-C1 (not protected by ATP) may be bound to reactive site(s) on the 72-kDa polypeptide that are not directly related to catalysis. The near complete protection by TNP-ATP would suggest that some of these reactive sites may bind nucleotides. The results, nevertheless, show that a portion (33%) of the [14C]Nbd-C1 was bound to sites protectable by the substrate.
Although initial experiments suggested that NEM could be useful in probing the functional subunits of the ATPase, ["C] NEM labeled several polypeptides (including a 72-kDa polypeptide) in the purified ATPase preparation, and little specific MgATP protection could be discerned from these experiments (not shown). This may be due to NEM modification of several sites on the subunits of the ATPase.
Reversal of Nbd-C1 Inhibition and Binding by DTT-The carbonium ion, formed as chloride leaves Nbd-C1, is highly reactive with nucleophiles (such as-SH,-OH, and-NH2) and can form 4-substituted 7-nitrobenzofurazan derivatives (42). Nbd-C1 could potentially react with cysteine, tyrosine, or lysine residues (42). To distinguish between these alternatives, we tested the effect of DTT on Nbd-C1 inhibition and binding. Sulfhydryl reducing agents can reverse S-Nbd or tyrosyl 0-Nbd adducts but not N-Nbd adducts (22,43). Dithiothreitol partially reversed Nbd-C1 inhibition (75%) ( Table  I) suggesting that the site of inhibition on the tonoplast ATPase contains sulfhydryl (cysteinyl) or hydroxyl (tyrosyl) groups. Consistent with these results, dithiothreitol reduced [14C]Nbd-C1 binding to the 72-kDa polypeptide of the purified ATPase. It is clear that a substantial amount (at least 65%) of the Nbd-C1 binding was reversed (compare lane 1 in Fig.  9A to lane 2 in Fig. 9B). These results suggest that the nucleotide-binding site on the 72-kDa subunit contains either a cysteine or a tyrosine, or both residues. It is possible the remaining (25-35%) [14C]Nbd-C1 is bound as N-Nbd (44).
Inhibition of Tonoplast ATPase Activity and H+ Pumping with Antisera to the 72-kDa Polypeptide-Polyclonal antibodies were made to the 72-kDa polypeptide. The antibody made to the 72-kDa polypeptide was tested for specificity. In Western (immuno) blots, the antibody reacted with a single polypeptide (mass = 72 kDa) in tonoplast-enriched membranes, or the purified tonoplast ATPase (Fig. 10). It did not crossreact with either the mitochondrial or Escherichia coli F,-

TABLE I DTT reversal of Nbd-Cl inhibition
Tonoplast vesicles were treated for 30 min (4 "C) with 10 ~L M Nbd-C1 in the presence of 10 mM BTP-Cl, pH 7.0. The reactions were terminated by a 100-fold dilution into the coupled assay mixture (-ATP, -NADH). DTT was added to a final concentration of 2 mM and the mixture was incubated for the indicated times at 35 "C. ATPase activity was measured after the addition of ATP and NADH at 23 "C. Data is the average of two exDeriments. Lanes 1, 2, and 3 are tonoplast membranes, purified tonoplast ATPase, and purified mitochondrial F1-ATPase (46), respectively, from oat roots, and lune 4 is F1-ATPase from E. coli.
ATPase. No cross-reactivity was seen with a 100-kDa polypeptide (mass of the plasma membrane H+-ATPase) (45) in a plasma membrane-enriched fraction (not shown). This antibody did react with a single polypeptide of approximately 72 kDa in membrane preparations of several other plant species, including corn, tobacco, and carrot (not shown). A minor band (in native membranes) at 65 kDa ( Fig. 10) was probably a proteolytic breakdown product of the 72-kDa polypeptide.
To confirm the role of the 72-kDa subunit in catalytic activity we tested the effect of anti-72 on ATPase activity. Anti-72 specifically inhibited the tonoplast H+-ATPase (Table 11). The antisera did not affect hydrolysis of ATP by either the mitochondrial Fl-ATPase or the plasma membrane ATPase. ATP-dependent H+ pumping in tonoplast-enriched vesicles was also inhibited by anti-72 (Fig. 11). These preliminary results show that the anti-72 of the tonoplast ATPase was highly specific, and that it inhibited ATPase activity probably by perturbing the catalytic domain of the 72 kDa.

DISCUSSION
In this paper we have used Nbd-C1 as an active site-directed probe to identify a putative catalytic subunit of the tonoplast H+. ATPase complex. We first showed that Nbd-C1 inhibition of ATPase activity is consistent with a modification of the catalytic site based on the following results. (i) Loss of activity followed pseudo first-order kinetics and the inactivation rate constant saturated with respect to Nbd-C1 concentration (Figs. 5 and 6); (ii) the physiological substrate, MgATP, protected against inactivation by Nbd-C1 with an average dissociation constant of 0.24 mM, similar to the K,,, for MgATP hydrolysis of 0.25 mM (10,ll); and (iii) MgADP, the product and a competitive inhibitor, protected against inactivation by Nbd-C1 (Fig. 4B). These results support the idea that Nbd-C1, an adenine analog, binds and modifies the cat- ATPases were incubated with 0.5 mg of antiserum in physiologically buffered saline, pH 7.4, or PBS alone for 1 h at 0-4 "C (except MFl, 23 "C). Activity was measured after a 10-fold dilution into the coupled assay medium, pH 7.0, containing 150 m M NaC1. Mitochondrial F1-ATPase from oat roots was CHC13-extracted as described previously (46). The plasma membrane-enriched fraction was obtained by centrifugation of a microsomal pellet onto a 6-12% dextran interface. All ATPase activities are expressed as the inhibitor-sensitive component (the difference of activity in the presence and absence of 20 mM BTP-NO3, 100 p~ sodium vanadate, or 0.5 mM sodium azide). Preimmune antiserum (1 mg/ml) had no effect on tonoplast ATPase activity. ATPase activities of tonoplast membranes, plasma membranes, and mitochondrial Fl-ATPase in the presence of PBS alone were 9.7.4.0, and 30.8 pmol of ADP h" mg", respectively.  Table 11, and then diluted 10-fold into a reaction mixture (1 ml) containing 15 mM Hepes-BTP at pH 7.0,0.5 pM acridine orange, 1.5 mM ATP, and 150 mM NaCl. H+ pumping was initiated by addition of MgSO, to give a final concentration of 5 mM, and activity was monitored by following the quenching of acridine orange fluorescence (excitation and emission wavelengths, 495 and 525 nm) at 23 "C (4). The reaction mixtures included a, PBS alone, or different amounts of antiserum made to 72-kDa polypeptide (in pg of protein); b 150; c, 500; or d, 1500. In e, lo00 pg of preimmune serum were added. Initial rates expressed as percent quench min" for 11 pg of vesicle protein were: a, 6.6; b, 6.5; c, 3.2; d, 0.8; and e, 6.5. Final concentration of gramicidin = 2.5 pg/ml. alytic nucleotide-binding site. Nbd-C1 is therefore a useful probe to identify the catalytic site and subunit.
Although the results could also be interpreted as Nbd-C1 modification at a site remote from, but conformationally coupled to, the nucleotide-binding site, the chemical modification studies taken together with the immunological results strongly support the idea that the 72-kDa polypeptide contains the catalytic site: (i) [14C]Nbd-C1 preferentially bound to a 72-kDa subunit of the purified tonoplast ATPase (Fig.  9); (ii) [14C]Nbd-C1 binding to the 72-kDa subunit was protected by the substrate, MgATP, and a potent competitive inhibitor, TNP-ATP; (iii) DTT reversed the Nbd-C1 inactivation of ATPase and prevented the binding of [14C]Nbd-C1 to the 72-kDa polypeptide (Fig. 9B Fig. 11).
The catalytic site may contain a cysteine, a tyrosine, or both residues. The ATPase is inhibited by NEM (Fig. 2 A ; Refs. 10 and 11) suggesting an essential role for "SH groups.
The reversal of Nbd-C1 inhibition of activity by DTT (Table   I), suggests that a cysteinyl "SH or tyrosyl " O H group had been modified (22,43). We hope to identify the chemical nature of the Nbd derivatives from their UV spectra, fluorescent properties, and by amino acid analyses. Such studies will demonstrate whether a tyrosine residue is in the catalytic site of the tonoplast ATPase as was found for the , ! ? subunit of the F1-ATPase (22, 43) and the Na/K-ATPase of eel electroplax (23).
Our findings with the tonoplast ATPase from a higher plant tissue confirm and extend a recent report of the vacuolar ATPase from the fungus, N . crassa. Although Bowman et al. (19) did not demonstrate whether Nbd-C1 modified the active site specifically, they showed that Nbd-C1 binding to a 70-kDa polypeptide of the vacuolar membrane ATPase was protected by ATP. In a recent study with tonoplast ATPase from corn coleoptiles, it was also concluded that the 72-kDa polypeptide is the catalytic subunit (47).
The role of the 60-kDa polypeptide is less clear. We have observed ATP-protectable, [14C]NEM binding to both the 60-kDa as well as the 72-kDa ~u b u n i t .~ Manolson et al. (15) showed that Bz-ATP (an ATP analog) preferentially bound to the 57-kDa polypeptide of the tonoplast ATPase from red beet, but it did not inhibit with simple competitive kinetics. Taken together, the results show that the 57-60 kDa subunit also has a nucleotide-binding site, possibly a regulatory site as in the mitochondrial Fl-ATPase (48). Alternatively, the catalytic site may be composed of both the 72-kDa and the 60-kDa polypeptides, and different probes may react with different portions of the same nucleotide-binding pocket, as postulated for the a and the fi subunits of the mitochondrial F1-ATPase (49).
Although the H+-ATPases from clathrin-coated vesicles (13) and chromaffin granules (14) are inhibited by Nbd-C1, the catalytic subunits of these enzymes have not as yet been identified. The polypeptide profiles of the partially purified ATPases from clathrin-coated vesicles (13) or chromaffin granules (14) show sufficient similarities (major polypeptides at about 70, 60, and 16 kDa) with the plant tonoplast H+-ATPase to suggest these enzymes may belong to one class of H+-ATPases. Furthermore, the remarkable similarity of these ATPases to that of the primitive anaerobic bacteria, Clostridium pasteurianum (50), suggests that these ATPases may have had an evolutionary antecedent in the prokaryotic kingdom.