Characterization of the Subunit Structure of the Maize Tonoplast ATPase IMMUNOLOGICAL AND INHIBITOR

Gradient purified preparations of the maize 400-kDa tonoplast ATPase are enriched in two major polypeptides, 72 and 62 kDa. Polyclonal antibodies were pre- pared against these two putative subunits after elution from sodium dodecyl sulfate-polyacrylamide gel elec- trophoresis gel slices and against the solubilized native enzyme. Antibodies to both the 72- and 62-kDa poly- peptides cross-reacted with similar bands on immuno-blots of a tonoplast-enriched fraction from barley, while only the 72-kDa antibodies cross-reacted with tonoplast and tonoplast ATPase preparations from Neurospora. Antibodies to the 72-kDa polypeptide and the native enzyme both strongly inhibited enzyme activity, but the 62-kDa antibody was without effect. The identity and function of the subunits was further probed using radiolabeled covalent inhibitors of the tonoplast ATPase, 7-chloro-4-nitro[’4C]benzo-2-oxa-1,3-diazole ([‘4C]NBD-Cl) and N,N’-[’4C]dicyclohex- ylcarbodiimide ([14C]DCCD). [14C]NBD-C1 preferentially labeled the 72-kDa polypeptide, and labeling was prevented by ATP. [14C]DCCD, an inhibitor of the proton channel portion of the mitochondrial ATPase, bound to a 16-kDa polypeptide. Venturicidin blocked binding to the mitochondrial 8-kDa polypeptide coleoptiles were collected from 5-day-old maize seedlings (Zea mays L. cv Golden Cross Bantam), and tonoplast-enriched membranes were isolated from 1/8% dextran step gradients and stored in resus- pension buffer (0.25 M sorbitol, 1 mM EDTA, 1 mM dithiothreitol, 10 mM Tris-Mes, pH 7.5) at -70 "C. We have previously shown that the H+-translocating ATPase enriched in this preparation is indistin-guishable from the H+-translocating ATPase associated with isolated vacuoles purified from corn coleoptile protoplasts (21). Treatment of the membranes with 0.15% deoxycholate removed loosely bound protein and increased the specific activity of the tonoplast ATPase 5-fold to approximately 1 pmol/min/mg of protein. The ATPase was solubilized with 40 mM octyl-P-D-glucopyranoside and partially purified by centrifugation into a 4-ml10-25% linear sucrose gradient in 1 mM EDTA, 1 mM dithiothreitol, 20 mM octyl-P-D-glucopyranoside, and 10 mM Tris-Mes, pH 7.5. The gradient was centrifuged in a Beckman SW 65 rotor at 200,000 X g for 5 h at 4 "C, and 0.5-ml fractions were collected. The peak fractions of ATPase were enriched in the M, 72,000 and M, 62,000 bands on SDS-polyacrylamide gel electrophoresis gels, as previously reported (4). ATPase Assay-Five to 20 pl of enzyme were assayed in a 0.5-ml volume containing 2.5 mM ATP, 2.5 mM MgS04, 50 mM KCl, 25 mM Tris-Mes, pH 6.5. After a 20-30-min incubation at 37 "C, the reaction was terminated, and released Pi was determined by the method of Fiske and Subbarow (22).

Characterization of the Subunit Structure of the Maize Tonoplast ATPase IMMUNOLOGICAL AND INHIBITOR BINDING STUDIES* (Received for publication, June 9, 1986) Suzanne Mandala$ and Lincoln Taizg From the Biology Department,Thimann Laboratories,University of California at Santa Cruz,Santa Cruz,California 95064 Gradient purified preparations of the maize 400-kDa tonoplast ATPase are enriched in two major polypeptides, 72 and 62 kDa. Polyclonal antibodies were prepared against these two putative subunits after elution from sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel slices and against the solubilized native enzyme. Antibodies to both the 72-and 62-kDa polypeptides cross-reacted with similar bands on immunoblots of a tonoplast-enriched fraction from barley, while only the 72-kDa antibodies cross-reacted with tonoplast and tonoplast ATPase preparations from Neurospora. Antibodies to the 72-kDa polypeptide and the native enzyme both strongly inhibited enzyme activity, but the 62-kDa antibody was without effect.

diazole (['4C]NBD-Cl) and N,N'-['4C]dicyclohexylcarbodiimide ([14C]DCCD).
[14C]NBD-C1 preferentially labeled the 72-kDa polypeptide, and labeling was prevented by ATP. [14C]DCCD, an inhibitor of the proton channel portion of the mitochondrial ATPase, bound to a 16-kDa polypeptide. Venturicidin blocked binding to the mitochondrial 8-kDa polypeptide but did not affect binding to the tonoplast 16-kDa polypeptide. Taken together, the results implicate the 72-kDa polypeptide as the catalytic subunit of the tonoplast ATPase. The DCCD-binding 16-kDa polypeptide may comprise the proton channel. The presence of nucleotide-binding sites on the 62-kDa polypeptide suggests that it may function as a regulatory subunit.
Plant vacuoles are acidic organelles in which ions, sugars, organic acids, and hydrolytic enzymes are stored (1). Studies with isolated vacuoles and tonoplast (plant vacuolar membrane) vesicles have indicated that a proton-translocating ATPase present on the tonoplast generates an electrochemical gradient, which may be responsible for the observed accumulation of ions and solutes (reviewed in Ref. 2 ) . A similar ATP-dependent proton pump is also present on the Golgi of maize coleoptiles (3). Several recent reports have described the partial purification of a novel ATPase from plant and ~~ * This research was supported by Grant PCM-8301995 from the National Science Foundation and Grant DE-FG03-84ER13245 from the Department of Energy. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ Present address: Dept. of Human Genetics and Physiology, Yale University School of Medicine, New Haven, CT 06510.
To whom correspondence should be addressed.
Previously we have shown that an antibody produced to the M , 72,000 protein from maize cross-reacts to a similar polypeptide in Neurospora vacuolar membranes (5). In the present study, we have examined the effects of antibodies made to the M, 72,000 and 62,000 proteins on ATPase activity and have analyzed their cross-reactivity to barley tonoplast membranes. In addition, radioactively labeled inhibitors of the tonoplast ATPase were used as probes to characterize putative subunits on SDS gels.
ATPase Assay-Five to 20 pl of enzyme were assayed in a 0.5-ml volume containing 2.5 mM ATP, 2.5 mM MgS04, 50 mM KCl, 25 mM Tris-Mes, pH 6.5. After a 20-30-min incubation a t 37 "C, the reaction was terminated, and released Pi was determined by the method of Fiske and Subbarow (22).
Inhibitor Binding-To label tonoplast proteins with [l4C]NBD-C1, 50 pl of deoxycholate-washed membranes (100 pg of protein) were treated with 0.5 pCi of [l4C]NBD-C1 (109 mCi/mmol) to give a final concentration of 75 pM. Following a 1-h incubation on ice in the presence or absence of 5 mM MgATP, the membranes were diluted with 4 ml of 50 mM Tris/HCl, pH 8.0, and centrifuged at 120,000 X g for 1 h in a Beckman SW 60 rotor. The pellet was resuspended in 40 pl of dilution buffer and prepared for SDS-gel electrophoresis. Labeling of partially purified ATPase from the sucrose gradient was performed similarly except that 60 pl of ATPase (7 pg of protein, specific activity, 2.5 pmol/min/mg) were incubated with 0.3 pCi of ["CINBD-Cl to give a final concentration of 35 p~. After the 1 h of incubation, samples were prepared for gel electrophoresis without washing.
The method of Sussman and Slayman (23) was used to label maize proteins with [14C]DCCD. Mitochondrial membranes (150 pg of protein collected from the 35/50% interface of a sucrose step gradient) and deoxycholate-washed tonoplast membranes (80 pg) were pretreated in the presence or absence of 50 pg/ml venturicidin for 10 min on ice. A 90-min incubation with 0.375 pCi of[14C]DCCD at a concentration of 12 p~ was terminated by dilution into 4 ml of resuspension buffer and centrifugation a t 120,000 X g for 45 min. Pellets were resuspended in 40 p1 of buffer and prepared for gel electrophoresis. Gradient-purified ATPase was labeled with 0.25 pCi of [14C]DCCD a t a final concentration of 67 p~ and was prepared directly for electrophoresis without washing.
Antibody Production-The M, 62,000 and 72,000 antigens were prepared by cutting out bands from 8% acrylamide gels that had been loaded with 900 pg of deoxycholate-washed tonoplast membranes. Coomassie-stained gel slices (40 pg of estimated protein) were mixed with 1 ml of 0.15 M NaCl and thoroughly crushed with a glass homogenizer. Antigen was vortexed with 1 ml of complete Freund's adjuvant and 5 mg of tubercle bacilli. Twenty intradermal injections were given by the method of Vaitukaitis et al. (24) on the shaved backs of New Zealand White rabbits. Subsequent injections were given subcutaneously a t 1-3-month intervals for a total of 6 injections. Antigen for boosting was prepared by homogenizing gel slices with incomplete Freund's adjuvant.
Antisera to the native ATPase was prepared by subcutaneously injecting rabbits with 1 ml (-100 pg) of partially purified ATPase from the sucrose gradient mixed with 1 ml of complete Freund's adjuvant. Two boosts using the same amount of ATPase with incomplete Freund's adjuvant were given at 1-month intervals. Approximately 40 ml of blood were collected from the ear 10 days after the last injections. Antiserum was precipitated with 40% ammonium sulfate and resuspended in 50 mM Tris/HCl, pH 7.4, 150 mM NaC1.
Specific cross-reaction to the M, 72,000 protein could be detected by the immunoblot procedure after 2 injections, and after 3 injections for the antibodies to the M. 62,000 protein.
Fluorography-Gels of 14C-labeled material were incubated in 80 ml of ENHANCE (New England Nuclear) for 1 h and then rinsed with cold water for 1 h before drying. Intensifying screens (Du Pont) were used with 32Pand '251-treated material. Radioactive bands were detected with Kodak x-ray (XAR-5) film after 1-24-day storage in the dark at -70 "C.
Electrotransfer-Proteins from SDS gels were electrotransferred to nitrocellulose, and the blots were washed, incubated with antibody, and treated with 9-Protein A, as described by Rott and Nelson (26). For the overnight incubation with antisera, a concentration of 20 pl in 40 ml of buffer was used for the antisera to the M, 62,000 and 72,000 proteins and 40 pl of antisera to gradient-purified ATPase.
Protein Assay-Protein was determined by the method of Lowry et al. (27) after trichloroacetic acid precipitation.
Neurospora and Barley Membranes-The Neurospora vacuolar membranes and ATPase fractions, prepared as described in Ref. 5, were provided by Dr. E. J. Bowman, University of California, Santa Cruz. The barley membranes were a gift from Dr. F. DuPont (United States Department of Agriculture, Albany, CA) and were obtained from a 22/30% interface of a sucrose step gradient using a modification of the procedure described in DuPont and Hurkman (28).
Reagent~-['~ClDCCD was from Amersham Corp., and [14C]NBD-C1 was from Research Products International Corp. Adjuvant and tubercle bacilli (Mycobacterium tuberculosis H37 RA) were obtained from Difco. Venturicidin came from BDH Chemicals, Ltd., Great Britain. All other chemicals were obtained as previously described (4). purified ATPase taken from the sucrose gradient (Fig. lA,  lanes 1 and 2). This antibody also cross-reacted with a M , 62,000 protein from barley membranes enriched in tonoplasts (lane 5 ) but did not recognize any Neurospora proteins when tested against vacuolar membranes or purified ATPase (lanes 3 and 4). Antibodies to the M , 72,000 protein specifically recognized that protein in maize tonoplast membranes and ATPase fractions (Fig. 1B). In addition to cross-reacting to the barley MI 72,000 protein (lane 5 ) , the anti-72 kDa antibodies also recognized a polypeptide of similar M , in Neurospora vacuolar membranes and ATPase fractions as we have previously shown (5). Proteins in Neurospora vacuolar membranes are slightly displaced upward in this molecular weight range, apparently by the presence of lower MI glycoproteins (5). Antibodies to the partially purified ATPase were produced by immunization against peak ATPase fractions taken from sucrose gradients. By the immunoblotting procedure, several polypeptides from gradient-purified ATPase fractions were recognized by this antiserum including the M, 72,000 protein (Fig. IC).

Immunological Characterization-
The effect of the three different antibodies on tonoplast ATPase activity is shown in Fig. 2. Partially purified ATPase was preincubated with the indicated amounts of antibody, and then ATPase activity was monitored. Antibodies made against the gradient-purified ATPase and the M , 72,000 protein were both inhibitory, giving 50% inhibition of activity with approximately 15 and 45 pg of serum protein, respectively. Almost 90% inhibition was achieved at the highest concentrations used. The anti-62-kDa antibodies showed no effect on ATPase activity at the highest tested concentration. Inhibition of ATPase activity in intact tonoplast vesicles by the anti-ATPase and anti-72-kDa antibody required higher concentrations.
NBD-C1 Binding and Inhibition-To further characterize the subunit composition of the tonoplast ATPase, radiolabeled inhibitors of ATPase activity were utilized. NBD-C1 had been used to covalently bind the / 3 subunit of the mitochondrial ATPase (29). NBD-C1 inhibited maize tonoplast ATPase activity as shown in Fig. 3. When the incubation with NBD-Cl was performed in the presence of 5 mM ATP,  inactivation of the ATPase was partially prevented. In the concentration range of 35-70 p~ NBD-C1, used to label tonoplast proteins (Fig. 4), approximately 75-90% inhibition was obtained in the absence of ATP and 48-60% inhibition in the presence of ATP.
Tonoplast proteins were labeled with ['4C]NBD-Cl in the presence or absence of 5 mM MgATP. The Coomassie-stained gel shown in Fig. 44 indicates that many polypeptides were present in deoxycholate-washed tonoplasts while gradientpurified ATPase fractions primarily consisted of the M , 72,000 and 62,000 proteins. The band most heavily labeled by [14C] NBD-C1 in the fluorograph (Fig. 4B) was the M , 72,000 polypeptide. The M , 72,000 protein was also the only one to show marked protection from labeling by MgATP.
DCCD Binding and Inhibition-The presence of a separate polypeptide ( M , -8,000) that is involved in proton conductance distinguishes FoF1-from E,E,-type ATPases.
[14C]DCCD has been used to identify the proton channel of both types of ATPases (23, 30), since DCCD is thought to inhibit H'-ATPases by blocking proton conductance (30). Tonoplast ATPase activity was also inhibited by DCCD. Fifty percent inactivation of ATPase was measured at a concentration less than 10 ~L M DCCD (Fig. 3).
[14C]DCCD was used to compare labeling of maize mitochondrial and tonoplast proteins. Fig.  5 shows that [14C]DCCD-labeled a polypeptide in mitochondrial membranes at a M, of approximately 8,000, expected for the Fo proteolipid. Binding to this subunit was prevented by venturicidin, as has been previously demonstrated (11). Another protein of M , -25,000 was also labeled in mitochondrial membranes. This may be explained by the existence of several other mitochondrial proteins, such as subunits of cytochrome oxidase and cytochrome bcl, that are sensitive to DCCD (30). In tonoplast membranes,[14C]DCCD also labeled a low molecular weight polypeptide, which differed from the mitochondrial proteolipid by having twice the relative mobility on the gel (Mr 16,000) and by not showing sensitivity to venturicidin. The same polypeptide was also identified as the major ["C] DCCD binding protein in purified tonoplast ATPase fractions from the sucrose gradient. The low level of binding to the M , 72,000 and 62,000 proteins in the gradient fraction may be nonspecific binding due to the method of directly loading the sample on the gel or may represent specific binding as has been reported for the @ subunit of the mitochondrial ATPase (31). When mitochondrial and tonoplast membranes were treated the same way (i.e. without dilution and pelleting) these subunits also showed preferential labeling.
The presence of multiple subunits, including a low M , polypeptide that binds [14C]DCCD, is reminiscent of FoFItype ATPases. The soluble F, portion is composed of 5 or more subunits with molecular weights ranging between 7,500 and 62,000 Da (34). The catalytic polypeptide is thought to be the p subunit of M , -55,000. The membrane-bound Fo portion contains the M , 8,000 proton channel that binds ["C] DCCD and 2 other subunits. In contrast, the proton-translocating EIEz ATPase found on plant and fungal plasma membranes consists of a single type of polypeptide of MI 100,000. This subunit functions catalytically and also binds ["C] DCCD (23). With respect to sensitivity to inhibitors, the tonoplast ATPase appears to more closely resemble the FoF1type than the EIEz-type enzyme (11,32). The tonoplast and FoF1-type ATPases are not inhibited by vanadate, suggesting that they do not form phosphorylated intermediates characteristic of the reaction cycle of E1E2-type enzymes (4,7,11,32). In comparisons of the effectiveness of a large variety of inhibitors on all three types of ATPases in Neurospora (11) and between the mitochondrial and tonoplast ATPases in oat roots (32), it was demonstrated that many inhibitors had strikingly similar effects on the mitochondrial and tonoplast enzymes. However, the tonoplast ATPase is resistant to the mitochondrial inhibitor, azide (4,7,9,32).
This study was undertaken in an effort to further characterize the subunits and to identify the catalytic site of the maize tonoplast ATPase using immunological and inhibitor binding methods. Previously we have shown that the tonoplast ATPase represents a distinct immunological class based on cross-reactivity studies (5). Antibodies to the Neurospora plasma membrane ATPase and the p subunit of the E. coli FoFl ATPase do not cross-react with the vacuolar membrane ATPase of Neurospora, and the antibody to the maize M, 72,000 tonoplast polypeptide does not recognize mitochondrial or plasma membrane polypeptides (5). Uchida et al. (7) have shown that antiserum to the F1 ATPase of S. cerevisiae inhibits mitochondrial but not the vacuolar membrane ATPase. In the present study, we have shown that antibodies to the maize MI 72,000 and MI 62,000 polypeptides both crossreacted with barley, while only the antibody to the maize M, 72,000 subunit cross-reacted with Neurospora (Fig. 1). This indicates that the larger subunit is more highly conserved. In a study on the immunological properties of FoFl ATPases, Rott and Nelson (26)   consistently cross-reacted to antibodies made to the 0 subunit from a different species. Antibodies made to a or y subunits cross-reacted weakly or not at all to subunits from other species.
The importance of the M , 72,000 protein for the ATPase was also evident from measuring antibody effects on activity (Fig. 2). ATPase activity was inhibited only by the two antibodies that recognized the M , 72,000 protein in immunoblots: the antiserum made against gradient-purified ATPase fractions and the 72-kDa polypeptide. Thus, the immunological data support the M , 72,000 polypeptide as the catalytic subunit. Studies with radioactively labeled inhibitors are consistent with this conclusion. NBD-Cl was chosen as a potentially useful label of the catalytic site since Ferguson et al. (29) have found that NBD-Cl reacts exclusively with the p subunit of the mitochondrial ATPase. NBD-C1 inhibited the tonoplast ATPase in an ATP-protectable manner (Fig. 3). Fig. 4 shows that [14C]NBD-C1 primarily labeled the M , 72,000 polypeptide in maize tonoplast membranes. The M , 72,000 polypeptide was also the only band to show significant protection from labeling by ATP. Similar results have been obtained for the vacuolar ATPase of Neurospora ( 5 ) . Manolson et al. (6) reported that the radioactively labeled substrate analog ["PI-3-(0)-benzoy1)benzoyl ATP preferentially labeled the 57-kDa polypeptide of beet tonoplast ATPase, rather than the 67-kDa polypeptide. However, Manolson et al. (6) concluded that the binding site on the 57-kDa subunit is probably a regulatory site since BzATP was not a simple competitive inhibitor of the ATPase. In contrast, we have observed that [3'P]8-azido-ATP binds to both the 72-and 62-kDa polypeptides of maize tonoplast ATPase,' suggesting that both catalytic and regulatory sites are being labeled. By analogy to the F, ATPase, which has as many as three nucleotide-binding sites on the a and @ subunits (34, 35), nucleotide-binding sites may be present on both of the major subunits of the tonoplast ATPase. Since other evidence strongly supports the 72-kDa polypeptide as the catalytic subunit, we concur with Manolson et al. (6) that the 62-kDa polypeptide probably represents a regulatory subunit.
In conclusion, further characterization of the subunit composition of the maize tonoplast ATPase has provided evidence for the oligomeric nature of the enzyme. Results from the immunoIogica1 and inhibitor binding studies have suggested that the M , 72,000 polypeptide is the catalytic subunit of the ATPase. The M , 62,000 protein may be a regulatory subunit S. Mandala and L. Taiz, unpublished data.
with nucleotide-binding sites, while the M , 16,000 polypeptide that binds DCCD is most likely the proton channel of the ATPase. Similarities between FoF1-type ATPases and the vacuolar ATPases from plants and fungi are evident from this and other studies (11, 32). As previously noted ( 5 ) , the FoF, ATPase that most closely resembles the tonoplast ATPase is one from the anaerobic bacterium, Clostridium pasteurianum (36). This enzyme has a simpler subunit composition ( M , = 66,000, 58,000, 43,000, and 15,000) than that found in mitochondria, chloroplasts, and aerobic bacteria (36). The possibility that this enzyme is a primitive form from which more complex F,F,-type ATPases has evolved has been proposed (33).

Immunological and Binding Studies
on Maize Tonoplast ATPase 12855