Involvement of the nonhomologous region of subunit A of the yeast V-ATPase in coupling and in vivo dissociation.

The catalytic nucleotide binding subunit (subunit A) of the vacuolar proton-translocating ATPase (or V-ATPase) is homologous to the beta-subunit of the F-ATPase but contains a 90-amino acid insert not present in the beta-subunit, termed the nonhomologous region. We previously demonstrated that mutations in this region lead to changes in coupling of proton transport and ATPase activity and to inhibition of in vivo dissociation of the V-ATPase complex, an important regulatory mechanism (Shao, E., Nishi T., Kawasaki-Nishi, S., and Forgac, M. (2003) J. Biol. Chem. 278, 12985-12991). Measurement of the ATP dependence of coupling for the wild type and mutant proteins demonstrates that the coupling differences are observed at ATP concentrations up to 1 mm. A decrease in coupling efficiency is observed at higher ATP concentrations for the wild type and mutant V-ATPases. Immunoprecipitation of an epitope-tagged nonhomologous region from cell lysates indicates that this region is able to bind to the integral V0 domain in the absence of the remainder of the A subunit, an interaction confirmed by immunoprecipitation of V0. Interaction between the nonhomologous region and V0 is reduced upon incubation of cells in the absence of glucose, suggesting that the nonhomologous region may act as a trigger to activate in vivo dissociation. Immunoprecipitation suggests that the epitope tag on the nonhomologous region becomes less accessible upon glucose withdrawal, possibly due to binding to another cellular target. In vivo dissociation of the V-ATPase in response to glucose removal is also blocked by chloroquine, a weak base that neutralizes the acidic pH of the vacuole. The results suggest that the dependence of in vivo dissociation of the V-ATPase on catalytic activity may be due to neutralization of the yeast vacuole, which in turn blocks glucose-dependent dissociation.

The vacuolar proton-translocating ATPases (V-ATPases) 1 are a family of ATP-dependent proton pumps that couple the hydrolysis of ATP to proton movement across the membrane (1)(2)(3)(4)(5)(6)(7)(8). This proton movement results in acidification of intracellular compartments, which in turn is critical for cellular processes such as receptor-mediated endocytosis, the processing and degradation of macromolecules, intracellular trafficking of lysosomal enzymes, coupled transport of small molecules, and entry of certain envelope viruses (1). For certain specialized cells such as renal intercalated cells, osteoclasts, macrophages, and insect goblet cells, V-ATPases are present on the plasma membrane and function in processes such as renal acidification, bone resorption, pH homeostasis, and coupled potassium transport (9 -12).
The V-ATPases are multisubunit complexes composed of two functional domains (1)(2)(3)(4)(5)(6)(7)(8). The soluble V 1 domain is responsible for ATP hydrolysis and contains eight different subunits (subunits A-H) with molecular masses 70 -13 kDa. The integral V 0 domain is responsible for proton translocation and contains six different subunits (subunits a, d, e, c, cЈ, and cЉ) with molecular masses of 100 -10 kDa. The 10-kDa subunit e, previously identified in bovine and insect (13,14), has recently been shown to be essential for function of the yeast V-ATPase (15). Both the 70-kDa A subunit and the 60-kDa B subunit of V 1 possess nucleotide binding sites (16,17), with the catalytic nucleotide binding sites located on subunit A (18). The V-ATPases structurally resemble the F-ATPases (proton-driven ATP synthases) of mitochondria, chloroplasts, and bacteria (19 -23). The A and B subunits of the V-ATPases share ϳ25% amino acid sequence identity with the ␤and ␣-subunits of the F-ATPases, respectively (24 -28). Sequence alignment of the A subunit and the ␤-subunit reveals a 90-amino acid region (termed the nonhomologous domain), which is present in the A subunit but which is absent from the ␤-subunit (24 -27). Although not conserved between the V and F-ATPases, this region is highly conserved among V-ATPase A subunit sequences (24 -27).
We have previously demonstrated by site-directed mutagenesis of the VMA1 gene that encodes subunit A in yeast that changes in the nonhomologous domain can alter coupling of proton transport and ATPase activity (29). We have also observed that mutations in this domain are able to block in vivo dissociation of the V-ATPase in response to glucose depletion (29). Reversible dissociation of the V-ATPase has been shown to represent an important mechanism of regulating V-ATPase activity in vivo (30,31). In the current study, we have further characterized the changes in coupling behavior observed upon mutagenesis of the nonhomologous region and have investigated the ability of this region to interact with the integral V 0 domain. The results suggest that the nonhomologous region may serve as a trigger to activate reversible dissociation of the V-ATPase.

EXPERIMENTAL PROCEDURES
Materials and Strains-Escherichia coli and yeast culture media were purchased from Difco. Restriction endonucleases, T4 DNA ligase, and other molecular biology reagents were from Invitrogen. Zymolyase 100T was purchased from Sekagaku America, Inc. Protease inhibitors and the monoclonal antibodies against the HA epitope, 3F10, were purchased from Roche Applied Science. Protein A-Sepharose, Protein G-agarose, ATP, and most other chemicals were obtained from Sigma. Concanamycin A was purchased from Fluka Chemical Corp. 9-Amino-6-chloro-2-methoxyacridine was purchased from Molecular Probes, Inc. (Eugene, OR). SDS, nitrocellulose membrane (0.45-m pore size), Tween 20, horseradish peroxidase-conjugated goat anti-rabbit IgG, and horseradish peroxidase-conjugated goat anti-mouse IgG were from Bio-Rad. The chemiluminescence substrate for horseradish peroxidase was from Kirkegaard & Perry Laboratories. Mouse monoclonal antibodies 8B1-F3 against subunit A, 13D11-B2 against subunit B, and 10D7 against Vph1p were purchased from Molecular Probes. The rabbit polyclonal antibodies against subunits d, D, F, G, and H were gifts from Dr. Tom Stevens (University of Oregon, Eugene), the antibody against subunit E was a gift from Dr. Daniel Klionsky (University of Michigan, Ann Arbor), and the antibody against subunit C was a gift from Dr. Patricia Kane (SUNY Upstate Medical University, Syracuse).
Plasmid Construction and Transformation-Both SpeI and BamHI restriction enzyme sites were introduced into the nucleic acid sequence of the nonhomologous region of the VMA1 gene using the polymerase chain reaction. The sequence of the nonhomologous region of the VMA1 gene (NHR) was amplified from the wild type plasmid (Yip5-VMA1) using primers containing the restriction enzyme sites SpeI and BamHI and then cloned into the 2-m plasmid, p426, using the SpeI and BamHI sites (p426-NHR). The sequence of the cloned nonhomologous region of the VMA1 gene was confirmed by DNA sequencing using an automated sequencer from Applied Biosystems. The oligonucleotides used for amplification of the nonhomologous region of the VMA1 gene were as follows, with the restriction enzyme sites underlined: NHR forward (SpeI), 5Ј-GGACTAGTAAACAATGCCGGGAAAGTTTCAAGT-CGGC-3Ј; NHR reverse (BamHI), 5Ј-CGCGGATCCAGGATAGTCAGC-AGATAACTT-3Ј. The three tandem repeats of the nine-amino acid HA epitope (YPYDVPDYA) were ligated into the BamHI and KpnI sites of p426-NHR (p426-NHR::HA). This resulted in insertion of three tandem repeated HA tags after amino acid residue Pro-233 of the nonhomologous region of Vma1p.
Yeast strain SF838 -5Aalpha vma1⌬ 8 were transformed using the lithium acetate method with p426-NHR::HA and p426 vector alone (32). The transformants were selected on Ura Ϫ plates as described previously (33). Growth phenotypes of the transformants were assessed on YPD plates buffered with 50 mM KH 2 PO 4 or 50 mM succinic acid to either pH 7.5 or pH 5.5.
Isolation of Vacuolar Membrane Vesicles-Vacuolar membrane vesicles were isolated using a modified protocol described by Uchida et al. (34). The yeast integrated with wild type plasmid (YIp5-VMA1), the nonhomologous mutants, or the vector YIp5 alone were cultured overnight in 1 liter of YPD (pH 5.5) to log phase. Cells were pelleted, washed once with water, and resuspended in 100 ml of 10 mM dithiothreitol and 100 mM Tris-HCl, pH 9.4. After incubation at 30°C for 15 min, cells were pelleted again; washed once with 100 ml of YPD medium containing 0.7 M sorbitol, 2 mM dithiothreitol, 100 mM MES-Tris, pH 7.5, and 2 mg of Zymolase 100T; and incubated at 30°C with gentle shaking for 60 min. The resulting spheroplasts were osmotically lysed, and the vacuoles were isolated by flotation on two consecutive Ficoll gradients. Protein concentrations were determined by Lowry assay (35).
Immunoblot Analysis-Yeast were grown to log phase at 30°C in Ura Ϫ -selective medium and whole cell lysates were prepared using 50 mM Tris-HCl, pH 6.8, 8 M urea, 5% SDS, 1 mM EDTA, and 5% ␤-mercaptoethanol, as described previously (36). Samples were subjected to SDS-PAGE and transferred to nitrocellulose. The expression of the nonhomologous region of subunit A was detected by Western blotting using the horseradish peroxidase-conjugated monoclonal antibody 3F10 against HA, whereas subunits A and B were detected using the monoclonal antibody 8B1-F3 directed against subunit A and the monoclonal antibody 13D11 against subunit B, respectively, followed by horseradish peroxidase-conjugated secondary antibodies. Immunoblots were developed using a chemiluminescent detection method from Kirkegaard & Perry Laboratories.
In Vivo Interaction between the Nonhomologous Region and the V 0 Domain-Dissociation and reassembly of the V-ATPase in response to glucose depletion and glucose readdition were measured as described previously (37) with some modification. The yeast cells containing the integrated wild type plasmid (Yip5-VMA1) or the vma1⌬ strain expressing the HA-tagged nonhomologous region of Vma1p or the vector alone were grown to log phase in Ura Ϫ -selective medium. The cells were converted to spheroplasts by treatment with Zymolase 100T and incubated in YEP medium with or without 2% glucose for 15 min at 30°C. An aliquot of the spheroplasts incubated in the absence of glucose was subsequently incubated in the presence of 2% glucose for an additional 15 min. Spheroplasts were pelleted and lysed in phosphate-buffered saline containing 1% C 12 E 9 , protease inhibitors (2 g/ml aprotinin, 0.7 g/ml pepstatin, 5 g/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride), and 1 mM dithiobis(succinimidyl propionate). The V 0 domain was immunoprecipitated using the monoclonal antibody 10D7-A7 against Vph1p and protein A-Sepharose, whereas the HA-tagged nonhomologous region was immunoprecipitated using the monoclonal antibody 3F10 against HA and protein G-agarose. Where indicated, the V 1 domain and intact V 1 V 0 were immunoprecipitated using the monoclonal antibody 8B1-F3 against subunit A and protein A-Sepharose. Samples were then separated on 4 -20% gradient gels and transferred to nitrocellulose. Western blotting was performed using the monoclonal antibody 3F10 against HA to detect the nonhomologous region as well as the polyclonal antibody against subunit d and the monoclonal antibody 10D7-A7 against Vph1p to detect the V 0 domain, followed by a horseradish peroxidase-conjugated secondary antibody. To detect V 1 subunits, monoclonal antibody 8B1-F3 against subunit A or 13D11 against subunit B or rabbit polyclonal antibody against subunits C, D, E, F, G, or H was employed as primary antibody in Western blot analysis. Western blot analysis of expression levels of the nonhomologous region of Vma1p and other V-ATPase subunits in whole cell lysates was performed as previously described (37). Protein samples were precipitated with an equal volume of 20% trichloroacetic acid, washed with cold acetone, and redissolved in 50 mM Tris-HCl, pH 6.8, 8 M urea, 5% SDS, 1 mM EDTA, and 5% ␤-mercaptoethanol. Samples were then subjected to SDS-PAGE and transferred to nitrocellulose. Blots were incubated with antibodies against either the HA epitope tag or subunits of the yeast V-ATPase followed by horseradish peroxidase-conjugated secondary antibodies as previously described (37). Immunoblots were developed using a chemiluminescent detection method from Kirkegaard & Perry Laboratories.
Biochemical Characterization-ATPase activity was measured using a coupled spectrophotomeric method as described previously (38) with some modification. Isolated vacuoles were incubated in ATPase assay buffer (50 mM NaCl, 30 mM KCl, 20 mM HEPES-NaOH, pH 7.0, 0.2 mM EGTA, 10% glycerol, 1 mM MgCl 2 , 1.5 mM phosphoenolpyruvate, 0.35 mM NADH, 20 units/ml pyruvate kinase, and 10 units/ml lactate dehydrogenase) with 0.1% Me 2 SO or 1 M concanamycin A at room temperature for 10 min. The assay was initiated by the addition of ATP at the indicated concentrations, and the absorbance at 341 nm was measured continuously using a Kontron UV-visible spectrophotometer. ATP-dependent proton transport was measured by the initial rate of ATP-dependent fluorescence quenching with the fluorescence probe 9-amino-6-chloro-2-methoxyacridine in the presence or absence of 1 M concanamycin A using a PerkinElmer Life Sciences LS50B spectrofluorometer (39).
Quinacrine Staining-Vacuolar accumulation of quinacrine was examined as described previously with slight modifications (40,41). Approximately 4 -5 ϫ 10 6 log phase yeast cells were harvested, resuspended in 500 l of YPD buffered with 50 mM Na 2 HPO 4 (pH 7.6) containing 200 M quinacrine or in 500 l of YPD buffered with 50 mM Na 2 HPO 4 (pH 7.6) containing 200 M chloroquine, and incubated with shaking at room temperature for 5 min. Cells treated with chloroquine were then sedimented and resuspended in 500 l of YPD buffered with 50 mM Na 2 HPO 4 (pH 7.6) containing 200 M quinacrine and incubated with shaking at room temperature for 5 min. Cells were sedimented at 10,000 ϫ g for 5 s and resuspended in 50 l of 2% glucose buffered with 50 mM Na 2 HPO 4 (pH 7.6). 8-l samples were applied to a microscope slide and visualized immediately with a Zeiss Axiovert fluorescence microscope. Cells were viewed under Nomarski optics to observe cell morphology and in fluorescence mode using a fluorescein filter with a 40ϫ objective to observe quinacrine staining.
Chloroquine Treatment and in Vivo Dissociation of the V-ATPase in Response to Glucose Depletion-Dissociation and reassembly of the V-ATPase in response to glucose depletion and glucose readdition were measured as described previously with some modifications (37). Yeast containing integrated wild type Vma1p were grown in YPD, pH 5.5, overnight to an absorbance at 600 nm of Ͻ1.0. The cells were converted to spheroplasts by treatment with Zymolase 100T and incubated in YEP medium with or without 2% glucose in the presence or in the absence of 200 M chloroquine for 15 min at 30°C. Spheroplasts were pelleted and lysed in phosphate-buffered saline containing 1% C 12 E 9 (polyoxyethylene 9-lauryl ether), protease inhibitors (2 g/ml aprotinin, 0.7 g/ml pepstatin, 5 g/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride), and 1 mM dithiobis(succinimidyl propionate). The V-ATPase complexes were immunoprecipitated using antibody 13D11 against the B subunit and protein A-Sepharose followed by separation on 8% acrylamide gels and transfer to nitrocellulose. Western blotting was then performed using the antibody 8B1-F3 against the A subunit and antibody 13D11 against the B subunit to detect the V 1 domain and antibody 10D7-A7 against Vph1p to detect the V 0 domain followed by a horse-radish peroxidase-conjugated secondary antibody. Dissociation of the V-ATPase complex is reflected as a reduction in the a subunit immunoprecipitated using the antibody directed against subunit B of the V 1 domain.

ATP Dependence of the Coupling Efficiency of Proton Transport and ATP Hydrolysis for Wild Type and Mutant Forms of Vma1p Containing Mutations in the Nonhomologous
Region-We had previously demonstrated that mutations in the nonhomologous region could result in changes of the coupling of proton transport and ATP hydrolysis by the V-ATPases (29). In FIG. 1. ATP dependence of ATPase activity, proton transport, and the coupling ratio for vacuolar membranes isolated from yeast expressing the wild type A subunit or mutant forms containing mutations in the nonhomologous region of subunit A. Vacuolar membranes (10 g of protein) were isolated from cells expressing either wild type or mutant forms of subunit A containing the indicated mutations in the nonhomologous region and assayed for either concanamycin A-sensitive ATPase activity (a) or concanamycin A-sensitive, ATP-dependent proton transport (b), as determined from the initial rate of fluorescence quenching using the fluorescence probe 9-amino-6-chloro-2-methoxyacridine (see "Experimental Procedures"). Activities were measured at the indicated ATP concentrations (maintaining an Mg 2ϩ /ATP ratio of 2) and in the presence or absence of 1 M concanamycin A. Each value is the average of measurements on two independent vacuole preparations (error bars correspond to the S.E.). c, the coupling ratio calculated by dividing the initial rate of fluorescence quenching by the ATPase activity, shown as a function of ATP concentration. d, proton transport measurements done in the presence of 1.5 mM phophoenolpyruvate and 6 units/ml pyruvate kinase added to reconvert any ADP present to ATP. Measurements were done in duplicate on a single preparation of vacuolar membranes. f, the coupling ratio calculated by dividing the proton transport data shown in d by the ATPase activities shown in a. particular, vacuolar membranes isolated from two mutants (P223V and P233V) showed reduced proton transport but nearly the same level of ATPase activity as membranes isolated from yeast expressing wild type Vma1p, suggesting that these mutations resulted in a partial uncoupling of proton transport and ATPase activity (29). By contrast, the vacuolar membranes from cells expressing the P217V mutation showed only 30 -40% of wild type levels of ATPase activity but significantly higher proton transport than wild type. This result suggested a tighter coupling of proton transport and ATPase activity in this mutant and led to the conclusion that the wild type V-ATPase is not optimally coupled. The latter result is consistent with the hypothesis that vacuolar acidification may be modulated by changes in the tightness of coupling of proton transport and ATP hydrolysis by the V-ATPase (42,43).
The previous studies were conducted at a single ATP concentration (0.5 mM) that is below the typical cytoplasmic ATP concentration measured in yeast cells (44). In addition, it had previously been reported for the V-ATPase from bovine brain clathrin-coated vesicles that the tightness of coupling of proton transport and ATPase activity by the V-ATPase varied as a function of ATP concentration, with higher concentrations of ATP leading to a progressive uncoupling of these activities (42). We therefore wished to determine whether the changes in coupling behavior observed for strains expressing mutant forms of Vma1p were preserved at higher and lower concentrations of ATP. Fig. 1a shows the ATP dependence of concanamycin-sensitive ATPase activity over the range of 0.1-2.0 mM ATP for vacuolar membranes isolated from cells expressing wild type Vma1p as well as the P217V, P223V, and P233V mutants. Fig. 1b shows the ATP dependence of concanamycinsensitive proton transport (as assessed from the initial rate of ATP-dependent quenching of 9-amino-6-chloro-2-methoxyacridine fluorescence). Fig. 1c shows the ATP dependence of the ratio of proton transport to ATPase activity over the same range of ATP concentrations.
Consistent with previous results (42), proton transport peaks and then decreases at higher ATP concentrations, whereas ATPase activity increases and then plateaus over the same concentration range. The ATP concentration at which optimal proton transport is observed, however, varies with the A subunit expressed. Thus, for the P233V mutant, optimal proton transport is observed at 0.3 mM ATP, whereas for the P217V mutant the optimal concentration is 0.5 mM, and for the wild type and P223V mutants, this concentration is 0.7 mM (Fig. 1b). Also consistent with previous results (29), at 0.5 mM ATP the P217V mutant is more tightly coupled and the P223V and P233V mutants are less tightly coupled relative to the wild type enzyme (Fig. 1c). This relative coupling efficiency is generally preserved over the ATP concentration range of 0.1-1.0 mM, although at 0.3 mM ATP, the coupling of the P233V mutant is actually higher than wild type because of the peak in proton transport activity observed for this mutant at this ATP concentration (Fig. 1c). The lower coupling of the P223V and P233V mutants relative to the wild type is preserved up to 2.0 mM ATP. By contrast, for the P217V mutant, the coupling ratio decreases to and then falls below that of the wild type at 1.5 and 2.0 mM ATP (Fig. 1c), suggesting that the hypercoupling of this mutant may not be preserved under in vivo conditions. It has previously been shown that ADP inhibits V-ATPase activity, acting as a competitive inhibitor with a K i of 11 M for the bovine coated vesicle V-ATPase (45) and 310 M for the yeast V-ATPase (34). The ATPase assay employed is a coupled spectrophotometric assay using an ATP-regenerating system, whereas the proton transport assay employed does not include an ATP-regenerating system. It is therefore possible that the apparent decrease in coupling efficiency at high ATP concentrations is due to inhibition of proton transport by ADP that either accumulates during the course of the assay or is added as a contaminant with ATP. To address this question, pyruvate kinase and phosphoenolpyruvate were added to the proton transport assay to ensure that any ADP that was generated was quickly reconverted to ATP. The levels of pyruvate kinase and phosphoenolpyruvate were shown to be adequate to completely convert any ADP present to ATP using the coupled spectrophotometric assay used to measure ATPase activity (data not shown). As seen in Fig. 1d, the ATP dependence of proton transport of the wild type and mutant forms of the V-ATPase is similar to that observed in the absence of the regenerating system (Fig. 1b), indicating that the decreased coupling efficiency observed at higher ATP concentrations cannot be accounted for on the basis of ADP inhibition of proton transport.
Association between the Nonhomologous Region of the Subunit A and the V 0 Domain of the V-ATPase-In order to identify V-ATPase subunits with which the nonhomologous region of the A subunit was able to interact, an HA-tagged form of this region was expressed in cells disrupted in the VMA1 gene. As shown in Fig. 2, Western blot of whole cell lysates demonstrated the stable expression of this domain. Disruption of genes encoding subunits of the V-ATPase leads to a characteristic vma Ϫ phenotype in which cells are unable to grow at neutral pH (46,47). As expected, expression of the nonhomologous region in the absence of the remainder of the A subunit was not able to rescue the vma Ϫ phenotype. Interestingly, co-expression of the HA-tagged nonhomologous region and an A subunit construct lacking this region led to a wild type phenotype (data not shown). This result suggests that the nonhomologous region is able to fold into a stable domain that is able to bind to the remainder of the A subunit and promote assembly of at least a partially functional V-ATPase complex.
Immunoprecipitation of the nonhomologous region from cells expressing only this domain using an anti-HA antibody followed by Western blot analysis revealed co-precipitation of both subunit a and subunit d of the V 0 domain (Fig. 3a) but no co-precipitation of V 1 subunits (Fig. 3b). This result suggests that the nonhomologous region is able to directly bind to the V 0 FIG. 2. Western blot analysis of the expression level of the HA-tagged nonhomologous region of Vma1p. Whole cell lysates were prepared from the vma1⌬ strain expressing wild type Vma1p (WT), the vector alone (vector) or the HA-tagged nonhomologous region of Vma1p (NHR). Samples were subjected to SDS-PAGE on a 4 -20% acrylamide gradient gel and transferred to nitrocellulose. Western blotting was then performed using the monoclonal antibody 8B1 against subunit A, the monoclonal antibody 13D11 against subunit B, or the monoclonal antibody 3F10 against the HA epitope, as described under "Experimental Procedures." Also shown are the growth phenotypes of the cells on YPD plates buffered to pH 7.5. domain. In support of this interpretation, immunoprecipitation of the V 0 domain using the monoclonal antibody 10D7 directed against subunit a results in co-precipitation of the HA-tagged nonhomologous region (Fig. 3a). Because subunit F could not be detected in the immunoprecipitation using either the anti-HA or anti-A subunit antibodies, we cannot rule out the participa-tion of subunit F in binding of the nonhomologous region to the V 0 domain. However, because subunit F appears to reside in the central stalk connecting the V 1 and V 0 domains (48 -50), in contrast to the likely peripheral location of the nonhomologous region (see below), we think subunit F is unlikely to serve such a function.
Whereas the location of the binding site for the nonhomologous region within the V 0 domain has not been identified, the hydrophilic N-terminal domain of subunit a is a likely candidate. This a subunit domain has been shown from topological studies to reside on the cytoplasmic side of the membrane (51) and has been shown to bind to both subunit H and the Cterminal domain of subunit A (52). Cross-linking studies have also identified the proximity of this domain to subunit E (49), which forms part of the peripheral stalk connecting the V 1 and The spheroplasts were incubated for 15 min in the presence (ϩD) or absence (ϪD) of 2% glucose and then lysed in phosphate-buffered saline containing 1% C 12 E 9 , protease inhibitors, and 1 mM DSP. The V 0 domain was immunoprecipitated (IP) using the monoclonal antibody 10D7 against subunit a (left panel), whereas the nonhomologous region was immunoprecipitated using the monoclonal antibody 3F10 against the HA epitope (right panel). The proteins were separated by SDS-PAGE and transferred to nitrocellulose. Western blot analysis was performed using the monoclonal antibody 10D7 against subunit a, a rabbit polyclonal antibody against subunit d, or the monoclonal antibody 3F10 against the HA epitope, as described under "Experimental Procedures." FIG. 5. Expression level of the nonhomologous region does not change in response to glucose depletion. The vma1⌬ strain expressing the wild type (WT) Vma1p, the vector alone (vector), or the HA-tagged nonhomologous region of the Vma1p (NHR) were grown overnight, converted to spheroplasts, and incubated for 15 min in the presence (ϩD) or absence (ϪD) of 2% glucose. Whole cell lysates were prepared, and Western blot analysis was performed using the monoclonal antibodies 10D7 against subunit a, 8B1-F3 against subunit A, 13D11 against subunit B, 3F10 against the HA epitope tag, and a polyclonal antibody against subunit d as described under "Experimental Procedures."

FIG. 3. Interaction between the nonhomologous region of
Vma1p and the V 0 domain of the V-ATPase. a, the vma1⌬ strain expressing the wild type (WT) Vma1p, the vector alone (vector), or the HA-tagged nonhomologous region of Vma1p (NHR) were grown overnight and converted to spheroplasts. The spheroplasts were lysed in phosphate-buffered saline containing 1% C 12 E 9 , protease inhibitors, and 1 mM DSP. The nonhomologous region was immunoprecipitated (IP) using the monoclonal antibody 3F10 against the HA epitope (left panel), whereas the V 0 domain was immunoprecipitated using the monoclonal antibody 10D7 against subunit a (right panel). The proteins were separated by SDS-PAGE and transferred to nitrocellulose. Western blot analysis was performed using the monoclonal antibody 10D7 against subunit a, a rabbit polyclonal antibody against subunit d, or the monoclonal antibody 3F10 against the HA epitope, as described under "Experimental Procedures." b, the experiment was performed as described in a except that the vector alone was not included, immunoprecipitation was performed using the antibody against HA (left panel) and the monoclonal antibody 8B1-F3 against subunit A (right panel), and Western blotting was performed using the monoclonal antibody 3F10 against the HA epitope or monoclonal antibodies 8B1-F3 against subunit A or 13D11 against subunit B or rabbit polyclonal antibodies against subunit C, D, E, G, or H. Although Western blotting was also attempted using a rabbit polyclonal antibody against subunit F, no immunoreactive band was visible in either the immunoprecipitation performed using the antibody against HA or the antibody against subunit A, possibly due to loss of subunit F from the immunoprecipitate below the level detectable using the anti-F subunit antibody. V 0 domains (50,53). Whereas electron microscopic images suggest that the nonhomologous domain may correspond to the knoblike structures seen on the outer surface of V 1 near the top of the A 3 B 3 hexamer (54, 55), this does not exclude interaction with the soluble domain of subunit a. This is because a marked change in the position of the N-terminal domain of subunit a is observed in the intact V-ATPase as compared with the isolated V 0 domain, with this domain adopting a more extended conformation perpendicular to the membrane in the V 1 V 0 structure (56,57).
We next wished to determine whether the interaction between the nonhomologous region and the V 0 domain is glucosedependent. This was tested because glucose depletion causes reversible dissociation of the V 1 and V 0 domains in vivo (30). As can be seen in Fig. 4, immunoprecipitation of the V 0 domain with the anti-a subunit antibody following incubation of spheroplasts in the absence or presence of glucose revealed a decrease in the amount of the HA-tagged nonhomologous region bound to V 0 upon removal of glucose. The readdition of glucose after its removal caused reassociation of the nonhomologous region with the V 0 domain (data not shown). Thus, binding of the nonhomologous region of subunit A to the V 0 domain is glucose-dependent, with disruption of this interaction occurring upon the removal of glucose from the medium. Future experiments will be carried out to determine whether mutations in the nonhomologous region that block in vivo dissociation (29) are also able to prevent glucose-dependent dissociation of the nonhomologous region from V 0 .
Immunoprecipitation was next performed using the anti-HA monoclonal antibody. Surprisingly, although decreased co-precipitation of the V 0 subunits was observed upon glucose depletion, a decrease in the amount of the nonhomologous region immunoprecipitated was also observed (Fig. 4, right panel). This decrease in immunoprecipitation of the HA-tagged nonhomologous region in the absence of glucose could be due either to decreased levels of this domain in the cell (due, for example, to increased degradation) or to the occlusion of the epitope tag (by binding of the nonhomologous region to another protein). To distinguish these possibilities, the levels of the nonhomologous region present in whole cell lysates following glucose depletion were determined by Western blot. As can be seen in Fig. 5, the level of the nonhomologous region present in the whole cell lysate was unchanged following glucose depletion. This result suggests that the decreased immunoprecipitation of the nonhomologous region by the anti-HA antibody upon glucose removal ( Fig. 4) is due to occlusion of the antibody binding site rather than to increased degradation of this domain in the absence of glucose. As with interaction of the nonhomologous region with the V 0 domain, the decrease in the HA-tagged nonhomologous region immunoprecipitated with the anti-HA antibody observed upon glucose removal is reversed upon glucose readdition (data not shown). The identity of the protein to which the nonhomologous region binds upon its release from the V 0 domain remains to be determined.
Effect of Vacuolar Neutralization on in Vivo Glucose-dependent Dissociation of the V-ATPase-In yeast, dissociation of the V 1 and V 0 domains in response to glucose depletion has been shown to require catalytic activity of the V-ATPase (16,58). Consistent with this, the R219K mutation in the nonhomologous region that inhibits Ͼ90% of both proton transport and ATPase activity in isolated vacuolar membranes results in a block in glucose-dependent dissociation (29). This dependence of in vivo dissociation on V-ATPase activity could be explained in at least two possible ways. First, the V-ATPase may need to adopt a particular conformational state in order for dissociation FIG. 6. Chloroquine causes neutralization of acidic compartments in yeast as determined by staining with quinacrine. The vma1⌬ strain expressing the wild type Vma1p or the vector alone (vma1⌬) were grown to midlog phase and then incubated for 5 min at room temperature in YPD buffered with 50 mM Na 2 HPO 4 (pH 7.6) containing 200 M quinacrine (upper panels) or 200 M chloroquine followed by sedimentation and resuspension in the same buffer containing 200 M quinacrine followed by shaking for 5 min at room temperature (lower panels). Cells were then pelleted, resuspended in 50 l of 2% glucose buffered with 50 mM Na 2 HPO 4 (pH 7.6), and viewed with phase-contrast optics (left panels) or by fluorescence microscopy using a fluorescein filter for observation of quinacrine staining (right panels). Each panel is a composite of four fields.
FIG. 7. Chloroquine inhibits glucose-dependent dissociation of the yeast V-ATPase. The vma1⌬ strain expressing the wild type Vma1p was grown overnight and converted to spheroplasts. The spheroplasts were incubated for 15 min in YEP medium with or without 2% glucose in the presence or absence of 200 M chloroquine at 30°C and then lysed in phosphate-buffered saline containing 1% C 12 E 9 , protease inhibitors, and 1 mM DSP. The V-ATPase complex was immunoprecipitated using the monoclonal antibody 13D11 against subunit B. The proteins were separated by SDS-PAGE on 8% acrylamide gels followed by transfer to nitrocellulose. Western blot analysis was performed using the monoclonal antibody 10D7 against subunit a, 8B1 against subunit A or 13D11 against subunit B as described under "Experimental Procedures." Glucose-dependent dissociation appears as a reduction in the amount of the V 0 subunit a immunoprecipitated using the antibody against the V 1 subunit B. to occur, and by blocking activity the V-ATPase is prevented from adopting this critical conformation. Alternatively, there may exist a pH sensor that senses the pH of the compartment lumen and prevents dissociation of the V-ATPase if the luminal pH is too alkaline. This sensor might either be part of the V-ATPase itself or a separate transmembrane protein. Thus, if the V-ATPase is inactive, the luminal pH becomes high, and dissociation of the V-ATPase is blocked.
To distinguish between these two possibilities, the pH of the vacuolar compartment needs to be increased without directly inhibiting V-ATPase activity. To accomplish this, cells were treated with the weak base chloroquine, which has previously been shown to lead to neutralization of intracellular compartments (59). As can be seen in Fig. 6, treatment of wild type yeast cells with chloroquine leads to the loss of staining of cells with the fluorescence dye quinacrine, which accumulates in acidic compartments in cells (40). This loss of quinacrine staining is also observed upon disruption of the VMA1 gene in yeast (Fig. 6), consistent with the expected neutralization of the vacuolar compartment in this strain. To assess the effect of chloroquine treatment on in vivo dissociation of the V-ATPase, spheroplasts were incubated in the absence or presence of chloroquine and the absence or presence of glucose, followed by detergent solubilization, immunoprecipitation of the V 1 domain using a monoclonal antibody against subunit B, and Western blotting using antibodies against subunits A and B of the V 1 domain and subunit a of the V 0 domain. Dissociation appears as a decrease in the amount of subunit a immunoprecipitated using an antibody against subunit B (37). As can be seen in Fig.  7, chloroquine treatment blocks in vivo dissociation of the V-ATPase in response to glucose depletion. To test whether chloroquine inhibition of in vivo dissociation was due to a direct inhibitory effect on V-ATPase activity, the effect of chloroquine on V-ATPase activity in isolated vacuolar membranes was measured. An assay of wild type vacuolar membranes in the presence of 200 M chloroquine reduced concanamycin-sensitive ATPase activity to 81 Ϯ 2% of control levels. Initial results suggested a larger inhibitory effect, but this was shown to be due to absorbance of chloroquine at 341 nm, the wavelength at which the coupled spectrophotometric assay is performed. This absorbance effect was eliminated by measuring the decrease in NADH absorbance at 360 nm instead of 341 nm, at which wavelength interference due to absorbance of light by chloroquine is negligible. Because 80% of control activity is far in excess of what is required to allow in vivo dissociation of the V-ATPase to occur (16,58), chloroquine does not appear to block glucose-dependent dissociation of V 1 V 0 by directly inhibiting V-ATPase activity.
These results suggest that in vivo dissociation of the V-ATPase complex is sensitive to the luminal pH, such that if the luminal pH is too alkaline, dissociation of the V-ATPase is blocked. Such a mechanism makes sense as a way to prevent intracellular compartments from becoming too alkaline, where severe secondary effects, such as accumulation of undegraded proteins, may occur. Because the pH of the Golgi is more alkaline than that of compartments such as the lysosome (60), this pH dependence of in vivo dissociation may also account for why V-ATPase complexes containing Stv1p (which normally resides in the Golgi) show no dissociation when localized to the Golgi but dissociate normally when targeted to the vacuole (37). At odds with this idea is the observation that Stv1pcontaining complexes have much lower proton transport activity than Vph1p-containing complexes and hence might not be expected to sufficiently acidify the vacuolar compartment to allow dissociation to occur (37). However, Vma8p mutants possessing as little as 6% of wild type proton transport activity (in vitro) are still competent for glucose-dependent dissociation (61), so the exact level of activity required for dissociation remains uncertain. CONCLUSIONS In summary, we have shown that the coupling differences resulting from changes in the nonhomologous region are largely preserved over a range of ATP concentrations, although partial uncoupling is observed at higher ATP. The nonhomologous region is shown to bind to the V 0 domain, and this interaction is observed to be glucose-dependent, suggesting that changes in the interaction between the nonhomologous region and V 0 may trigger in vivo dissociation of the complex. Finally, in vivo dissociation of the V-ATPase is shown to be sensitive to the luminal pH of the compartment, although the identity of the pH sensor controlling dissociation remains to be determined.