HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 16, 13831-13839, April 19, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/16/13831    most recent M200682200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Smardon, A. M. Articles by Kane, P. M. Search for Related Content PubMed PubMed Citation Articles by Smardon, A. M. Articles by Kane, P. M. Social Bookmarking                      What's this?

The RAVE Complex Is Essential for Stable Assembly of the Yeast V-ATPase^*

Anne M. Smardon, Maureen Tarsio, and Patricia M. Kane

From the Department of Biochemistry and Molecular Biology, State University of New York, Upstate Medical University, Syracuse, New York 13210

Received for publication, January 22, 2002, and in revised form, February 6, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Vacuolar proton-translocating ATPases are composed of a peripheral complex, V₁, attached to an integral membrane complex, V_o. Association of the two complexes is essential for ATP-driven proton transport and is regulated post-translationally in response to glucose concentration. A new complex, RAVE, was recently isolated and implicated in glucose-dependent reassembly of V-ATPase complexes that had disassembled in response to glucose deprivation (Seol, J. H., Shevchenko, A., and Deshaies, R. J. (2001) Nat. Cell Biol. 3, 384-391). Here, we provide evidence supporting a role for RAVE in reassembly of the V-ATPase but also demonstrate an essential role in V-ATPase assembly under other conditions. The RAVE complex associates reversibly with V₁ complexes released from the membrane by glucose deprivation but binds constitutively to cytosolic V₁ sectors in a mutant lacking V_o sectors. V-ATPase complexes from cells lacking RAVE subunits show serious structural and functional defects even in glucose-grown cells or in combination with a mutation that blocks disassembly of the V-ATPase. RAVE·V₁ interactions are specifically disrupted in cells lacking V₁ subunits E or G, suggesting a direct involvement for these subunits in interaction of the two complexes. Skp1p, a RAVE subunit involved in many different signal transduction pathways, binds stably to other RAVE subunits under conditions that alter RAVE·V₁ binding; thus, Skp1p recruitment to the RAVE complex does not appear to provide a signal for V-ATPase assembly.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Vacuolar proton-translocating ATPases (V-ATPases)¹ are highly conserved proton pumps responsible for acidification of organelles such as the lysosome/vacuole, Golgi apparatus, and endosomes in all eukaryotic cells (1-3). In some cells, VATPases are also present at high levels at the plasma membrane, where they pump protons from the cytosol out of the cell (2, 4). In all of these locations and in organisms ranging from yeast to humans, V-ATPases have a very similar structure and subunit composition. They are comprised of 13 or 14 subunits arranged as a complex of cytosolic peripheral membrane subunits containing the sites of ATP hydrolysis, the V₁ sector, attached to a membrane complex containing the proton pore, the V_o sector. ATP-driven proton transport occurs only when the two sectors are structurally and functionally coupled. Free V₁ sectors do not catalyze hydrolysis of MgATP, the physiological substrate, and free V_o sectors do not appear to form open proton pores (5-7).

The assembly pathways for V-ATPases are complicated and incompletely understood. Both mammalian cells and yeast contain free V₁ and V_o sectors in vivo (8-10), and yeast mutants lacking one subunit of either sector are able to assemble the other sector (10). Yet there is evidence that the major pathway for biosynthetic assembly of V-ATPases does not involve independent assembly of free V₁ and V_o sectors followed by attachment of the two sectors. Instead, pulse-chase studies indicate very early association of V₁ and V_o sector subunits followed by addition of subunits from both sectors (11). Definition of assembly pathways was further complicated by the observation that fully assembled V-ATPases could rapidly and reversibly disassemble into free V₁ and V_o sectors (12, 13). Disassembly of V-ATPases in yeast and insects, the two systems in which the process is best characterized, occurs in response to low extracellular glucose concentrations. Reversible disassembly is believed to be an important regulatory mechanism; disassembly of V-ATPases conserves ATP under conditions of nutrient limitation by silencing the ATPase activity of the enzyme and reassembly rapidly reactivates the pump with no need for new protein synthesis (14, 15).

The signaling pathways connecting V-ATPase assembly with extracellular glucose concentration have proven to be somewhat elusive. Many of the pathways that signal glucose availability in yeast cells do not appear to be involved in modulating the assembly state of the V-ATPase (16). Recently, a new player in V-ATPase assembly was identified. Seol et al. (17) identified a Skp1p-containing complex they called RAVE, regulator of the ATPase of vacuolar and endosomal membranes, through affinity chromatography to detect Skp1p binding partners, and subsequently showed that RAVE also bound at least four of the V₁ subunits of the yeast V-ATPase. The RAVE complex contains three members, Rav1p, Rav2p, and Skp1p. RAV1 and RAV2 were previously uncharacterized yeast open reading frames, but RAV1 has homologues in all eukaryotes. Deletion of the RAV1 and RAV2 genes resulted in a temperature-dependent Vma phenotype, characterized by sensitivity to elevated pH and poor growth on glycerol-containing medium. The V-ATPase also showed assembly defects in the rav mutants, and the V₁ subunits that were assembled on the vacuolar membrane in a rav1 mutant disassembled rapidly in response to glucose deprivation but reassembled much more slowly upon glucose re-addition than in wild-type cells. These results implicated the RAVE complex in glucose-triggered reassembly of the V-ATPase, but it was not clear whether this aspect of RAVE function accounted for the V-ATPase assembly defects of rav1 mutants observed even in the presence of glucose.

Skp1p is a very highly conserved protein primarily known as a component of SCF (Skp1-cullin-F-box) ubiquitin ligases, which are involved in signal-induced protein degradation (18-20). SCF complexes regulate a wide range of fundamental cell processes ranging from cell cycle progression to nutrient utilization to transcription. Skp1p mediates this wide range of functions by binding a variety of proteins with little structural or sequence similarity beyond an F-box motif, a degenerate, 40-amino acid sequence motif originally identified by sequence alignment of several Skp1p-binding proteins and subsequently shown to be directly involved in binding to Skp1p (19, 21). In SCF complexes Skp1p is seen primarily as an adaptor capable of bridging F-box proteins with specificity for phosphorylated substrates to be degraded and E2/E3 ubiquitin ligase components bound to the cullin (19, 21). The RAVE complex diverges from the SCF model in several important ways. Although Seol et al. (17) determined that Rav1p binds to Skp1p, RAV1 contains no identifiable F-box sequence or resemblance to the cullin proteins. More importantly, the RAVE complex does not contain the yeast cullin, Cdc53p (17), and there is no apparent role for protein degradation in disassembly or reassembly of the V-ATPase (12). Taken together, these data indicate that RAVE is a non-SCF Skp1p-containing complex. The role of Skp1p in the RAVE complex is mysterious, but the presence of Skp1p invokes rich possibilities for cross-talk between V-ATPase assembly and the wide variety of signaling pathways in yeast that ultimately act through other Skp1-containing complexes.

In this work we have taken a closer look at the role of RAVE in both biosynthetic assembly and glucose-dependent disassembly and reassembly of the V-ATPase in yeast. Our results support those of Seol et al. (17) implicating RAVE in glucose-induced reassembly of the V-ATPase but also indicate that the RAVE complex must play a more general role in V-ATPase assembly. We provide evidence that the RAVE complex appears to bind V₁ whenever it is in the cytosol and that RAVE is required for stable assembly even in a mutant V-ATPase incapable of disassembly in response to glucose deprivation. We also find that the association of Skp1p with the RAVE complex does not change with changes in carbon source or under conditions where the RAVE complex cannot bind to V₁. These results suggest that Skp1p is a stable component of the RAVE complex and that the RAVE complex is critical for stable assembly of the V-ATPase under many different conditions.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials and Growth Media-- Oligonucleotides were purchased from MWG Biotech. LA-Taq polymerase was from Panvera and native Pfu DNA polymerase was from Stratagene. Other molecular biology reagents were purchased from New England BioLabs. Tran³⁵S-label and zymolyase 100T were purchased from ICN. Concanamycin A was from Wako Biochemicals. Monoclonal antibody 9E10 against the myc epitope was obtained from Roche Molecular Biotechnology or Zymed Laboratories Inc., and alkaline phosphatase second antibodies were from Promega. Yeast and Escherichia coli media were from Difco or Fisher. All other reagents were purchased from Sigma Chemical Co.

Yeast cells were grown in yeast extract-peptone-2% dextrose (YEPD) medium or fully supplemented minimal medium (SD) lacking individual nutrients as described previously (22). For growth of vma mutant strains, YEPD was buffered to pH 5.0 with 50 mM sodium phosphate/50 mM sodium succinate (YEPD, pH 5). For analysis of the Vma growth phenotype, YEPD was buffered to pH 7.5 with 50 mM Mes, 50 mM MOPS buffer, and 60 mM CaCl₂ was added (YEPD, pH 7.5, + Ca²⁺).

Strain Construction-- The strains used in this work are listed in Table I. The rav1::LEU2 and rav2::URA3 alleles were constructed by fusion PCR. Sequences immediately upstream and downstream of the RAV1 and RAV2 open reading frames were amplified using oligonucleotides YJR1-7 and YJR2-8 to amplify the 5'-end of RAV1, oligonucleotides YJR5-11 and YJR6-12 to amplify the 3'-end of RAV1, oligonucleotides YDR1-1 and YDR2-2 to amplify the 5'-end of RAV2, and oligonucleotides YDR5-5 and YDR6-6 to amplify the 3'-end of RAV2. Oligonucleotide sequences are shown in Table II. The URA3 and LEU2 genes were amplified using oligonucleotides YDR3-3 and YDR4-4 with pRS316 as a template and YJR3-9 and YJR4-10 with pRS315 as a template, respectively (23). The deletion alleles were constructed by combining the RAV2 5' and 3'-ends and URA3 fragment into a single fusion PCR reaction containing oligonucleotides YDR1-1 and YDR6-6. The RAV1 5'- and 3'-ends and LEU2 fragment were combined in a fusion PCR reaction containing oligonucleotides YJR1-7 and YJR6-12. The PCR reactions employed a mixture of LA-Taq and native Pfu thermostable DNA polymerases in LA-Taq buffer provided by the manufacturer. The products of the fusion PCR reactions were gel-purified and used to transform wild-type strain SF838-5A in a one-step gene replacement (24). Replacement of the wild-type with the mutant alleles was confirmed by PCR. To create the rav1rav2 double mutant, the rav2::URA2 fusion product was transformed into the rav1 strain, and transformants were selected on SD-uracil.

The strain containing the integrated vma11-E145L allele was constructed by first cloning the SmaI-EcoRV fragment containing kanMX6 gene from plasmid pFA6A-KanMX6 into the SnaBI site 445 bp upstream from the start codon of a plasmid-borne copy of the vma11-E145L allele in pRHA176 (25, 26). The 3.3-kb kanMX-containing vma11-E145L was excised from the plasmid with SacII and KpnI and used to transform the SF838-5A wild-type strain. Transformants were selected on YEPD medium containing 200 µg/ml G418, and integration of the vma11-E145L allele at the VMA11 locus was confirmed by sequencing. The rav1::LEU2 mutation was then introduced into this strain as described above.

To combine a myc-tagged Rav1p with the various vma deletion alleles, two different strategies were used. The vma mutant strains containing myc₉-tagged RAV1 (Table I) were obtained by amplifying vma::URA3 alleles from genomic DNA of existing vma1 (27), vma2 (28), vma3 (29), vma4 (30), and vma7 and vma10 (11) strains by PCR. In each case, oligonucleotides recognizing sequences 200-500 bp 5' and 3' of the URA3 insertion into each gene were used, and the PCR fragment containing the disrupted allele was gel-purified. Wild-type strain RDY1512 (17) was then transformed with the mutant alleles, and transformants were selected on SD-uracil medium. Correct integration was confirmed by either PCR or immunoblot. The vma mutant strains containing myc₁₃-tagged RAV1 (Table I) were obtained by introducing a RAV1-myc13-kanMX allele, which placed the myc₁₃ epitope immediately before the stop codon of RAV1, into existing vma deletion strains. The myc₁₃-kanMX cassette was amplified from plasmid pFA6a-myc₁₃-kanMX (25), and RAV1 fragments upstream and downstream of the stop codon were amplified with oligonucleotides YJR1-7 and RAV1M2 and RAV1M3 and YJR6-12, respectively. The final myc₁₃-tagged RAV1 was generated by fusion PCR as described previously (31). Previously characterized vma5 (32), vma13 (33), and vma8 (11) strains were transformed with the fusion PCR fragment. Transformants were selected on YEPD, pH 5, medium containing 200 µg/ml G418 and confirmed by PCR and/or immunoblotting.

Immunoprecipitations-- Immunoprecipitation of V-ATPase complexes from biosynthetically labeled cells was performed as described previously (12). For immunoprecipitation of cytosolic RAVE·V₁ complexes, cytosolic fractions were obtained from the indicated strains (150 A₆₀₀ of each) as described previously (6). Protein from 10% of the cytosolic fraction was directly precipitated with 10% trichloroacetic acid, and the remaining 90% was combined with 100 µl of anti-V₁ monoclonal antibody (8B1 or 13D11 (29)) or 5 µg of anti-myc antibody (9E10, Roche Molecular Biochemicals) followed by 60 µl of a 50% (v/v) suspension of Protein A-Sepharose CL4B. The immunoprecipitated proteins were solubilized at 75 °C in cracking buffer (50 mM Tris-HCl, pH 6.8, 8 M urea, 5% SDS, 5% -mercaptoethanol) for analysis by SDS-PAGE and immunoblotting.

For comparison of immunoprecipitations from the various vma strains (Fig. 6), cells (100 A₆₀₀ of each strain) were lysed by agitation with glass beads as described (34) and cytosol was obtained by centrifugation for 30 min at 100,000 × g in a Beckman TLA-100 Ultracentrifuge. Protein concentration in the cytosolic fractions from the various strains was measured by Lowry assay (35); 0.4 mg of protein was directly trichloroacetic acid-precipitated, and 4 mg was immunoprecipitated and analyzed by immunoblotting as described below.

Other Biochemical Techniques-- Vacuolar vesicles were isolated as described (36). ATP hydrolysis activity was determined at 37 °C using a coupled enzyme assay (37), in the presence and absence of 100 nM concanamycin A. For immunoblotting, vacuolar vesicles were pelleted by centrifugation and solubilized in cracking buffer. For all immunoblots, samples were separated by SDS-PAGE then transferred to nitrocellulose. Blots were probed with mouse monoclonal antibodies 8B1, 13D11, 7A2, and 10D7 against V₁ subunits A, B, and C, and V_o subunit a, respectively. Myc-tagged RAVE subunits were detected with mouse monoclonal antibody 9E10. Rabbit polyclonal antisera against yeast Skp1 was a generous gift from Ray Deshaies, California Institute of Technology. Primary antibodies were bound by alkaline phosphatase-conjugated goat anti-mouse or goat anti-rabbit secondary antibody and detected by colorimetric assay as described (38).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Is Association of the V₁ Sector and the RAVE Complex Glucose-dependent?-- Seol et al. (17) reported that the reduced levels of V₁ sectors that were present at the vacuolar membrane in a rav1 mutant in vivo dissociated rapidly from the membrane upon glucose deprivation but were slow to re-associate upon glucose re-addition. This result implicated the RAVE complex in the glucose-dependent reassembly of the V-ATPase. To probe this aspect of RAVE activity further, we examined the effects of extracellular glucose on the RAVE·V₁ interaction by using a monoclonal antibody against the V₁ B subunit to immunoprecipitate free V₁ complexes from cytosol along with bound RAVE subunits. Immunoblots of the cytosolic fractions probed with antibody against another V₁ subunit, subunit A, or against Rav1p and Rav2p proteins tagged with a myc₉ epitope are shown in Fig. 1A. These immunoblots demonstrate that there are detectable levels of V₁ subunits in the cytosol even in the presence of glucose (+), but the level of V₁ subunits in the cytosol increases markedly when the cells are deprived of glucose for 15 min () and then falls to the original level when glucose is restored to the cells (/+). (Although this immunoblot was probed only with the anti-A subunit antibody, previous results indicate that all of the V₁ subunits except subunit C are present in the cytosolic V₁ complexes (6, 12).) In contrast, the levels of Rav1p and Rav2p in the cytosol remain the same regardless of the level of extracellular glucose. In the right panels of Fig. 1A, complexes immunoprecipitated under non-denaturing conditions with an anti-B subunit antibody were probed for the presence of V₁ subunit A, Rav1p, and Rav2p on immunoblots. The levels of co-precipitated A subunit reflect the levels present in the cytosolic fractions, as expected; they increase in the absence of glucose and decrease in the presence of glucose. It is striking, however, that the levels of coprecipitated Rav1p and Rav2p also rise and fall with the changes in glucose concentration. This indicates that a part of the population of cytosolic RAVE complexes is recruited to interact with V₁ sectors released from the membrane by glucose deprivation and that these RAVE complexes are able to rapidly release the bound V₁ when extracellular glucose is restored.

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Interaction of RAVE subunits with cytosolic V1 sectors from wild-type and vma3 cells. A, wild-type cells carrying myc9-tagged Rav1 (RDY1512) or myc9-tagged Rav2 (RDY1513) were grown in YEPD (2% glucose), divided, pelleted by centrifugation, and either maintained in 2% glucose (+), resuspended in YEP medium lacking glucose for 15 min (), or resuspended in YEP medium lacking glucose for 15 min followed by re-addition of glucose to a 2% final concentration followed by another 15-min incubation (/+). Cytosol was prepared after the indicated treatments as described (6). Protein from 10% of the cytosol was directly trichloroacetic acid-precipitated (Cytosol), and the remaining cytosol was combined with anti-V1 subunit B antibody for immunoprecipitation of V1·RAVE complexes (IP/-V1) as described under "Experimental Procedures." Both the trichloroacetic acid precipitates and immunoprecipitated proteins were solubilized in cracking buffer, fractionated by SDS-PAGE, and transferred to nitrocellulose. The blots were probed with another V1 subunit (-Vma1p) or anti-myc epitope antibody against the indicated Rav protein. Cytosol from the equivalent of 0.6 A600 units of the starting culture was loaded in the cytosol samples; immunoprecipitate from 4.5 times as much cytosol was loaded for visualization of V1 subunits and from 9 times as much cytosol was loaded for visualization of myc-tagged RAVE subunits. B, the experiment shown in A was repeated, but the cells were incubated with 100 µg/ml cycloheximide during all of the treatments. C, the experiment shown in A was repeated, but in cells carrying the vma3::URA3 mutation (RDY1512 vma3 and RDY1513 vma3).

Disassembly and reassembly of the yeast V-ATPase has been shown to be entirely post-translational, because both processes can occur in the presence of 100 µg/ml cycloheximide (12). The turnover of RAVE subunits during glucose deprivation and re-addition has not been examined, however. Fig. 1B confirms that both the release of V₁ into the cytosol and its reassembly with the membrane occur in the presence of 100 µg/ml cycloheximide. More significantly, this figure demonstrates that Rav1p and Rav2p are relatively stable through a 15-min glucose deprivation followed by a 15-min glucose re-addition. The levels of RAVE subunits coprecipitated with V₁ are also very similar in the presence of cycloheximide. This indicates that the release of V₁ from RAVE upon glucose re-addition to glucose-deprived cells is not stimulated by degradation of Rav1p and/or Rav2p, but instead, binding of cytosolic V₁ to RAVE is reversible and responsive to extracellular glucose concentrations.

Reversible binding of RAVE to cytosolic V₁ in response to extracellular glucose could be achieved either by an intrinsic glucose sensitivity in the RAVE·V₁ interaction or by a number of more complex mechanisms. To test whether the RAVE·V₁ interaction was inherently glucose sensitive, we conducted the same experiment shown in Fig. 1A in a mutant strain lacking stable Vo sectors (vma3). In this strain, all of the V₁ subunits are synthesized, stable, and assembled into V₁ sectors lacking subunit C, but the V₁ sectors are entirely cytosolic (10, 38). Comparison of the cytosolic V₁ sectors released from the membrane by glucose deprivation to those present in the cytosol of a vma3 mutant revealed no differences in their subunit composition or biochemical properties (6). Fig. 1C demonstrates that the levels of the V₁ A subunit, Rav1p, and Rav2p in the cytosol remain constant through glucose deprivation and re-addition. If the RAVE·V₁ interaction were intrinsically glucose-sensitive, we might expect that higher levels of RAVE subunits would be coprecipitated with the V₁ sectors in the absence of glucose. The right panel of Fig. 1C demonstrates that this is not the case. Instead, similar levels of Rav1p and Rav2p are coprecipitated with V₁ regardless of extracellular glucose concentration. This result suggests that the RAVE·V₁ interaction is not intrinsically glucose sensitive; instead, RAVE seems to bind to V₁ whenever it is present in the cytosol.

V-ATPase Complexes Formed in rav Mutants Are Structurally Defective under All Conditions-- As described above, Seol et al. (17) observed a kinetic delay in reassembly of V₁ sectors upon glucose re-addition to a glucose-deprived rav1 mutant. They also saw lower levels of membrane-bound V₁ sectors even in the presence of glucose in a rav1 mutant, however. To assess the extent of the V-ATPase assembly defects in rav mutant strains grown in glucose, we quantitated the extent of V-ATPase assembly in the rav mutants by immunoprecipitation of fully and partially assembled complexes from whole cell lysates under non-denaturing conditions (Fig. 2). We have previously shown that a monoclonal antibody that recognizes the B subunit of the V-ATPase (-V₁ in Fig. 2) is able to coprecipitate the subunit-alone, assembled V₁ complexes or the fully assembled V-ATPase (including both V₁ and V_o subunits) (10). In Fig. 2, this antibody immunoprecipitates all of the V-ATPase subunits indicated on the left of the gel from wild-type cells but immunoprecipitates much lower levels of the a, d, c, and c' subunits, all of which are part of the V_o sector, from the rav1, rav2, and rav1rav2 mutants. This suggests that these mutants contain higher levels of V₁ sectors not bound to V_o than the wild-type cells. A second monoclonal antibody (designated -V_o in Fig. 2), recognizes the a subunit of the V_o sector, but only immunoprecipitates V_o sectors not bound to V₁ (10). Using this antibody, we can assess the level of free V_o sectors in cells. Consistent with the results with the anti-V₁ antibody, the anti-V_o subunit antibody immunoprecipitates higher levels of V_o subunits a, d, c, and c' from the rav mutant strains than from the wild-type strain. By quantitating the level of V_o subunits immunoprecipitated by each of the antibodies, we are able to determine the percentage of the total V_o sectors assembled with V₁ (12). On the average in two experiments, 60% of V_o sectors coprecipitated with V₁ from the wild-type cells, but only 9%, 14%, and 8% were assembled with V₁ in the rav1, rav2, and rav1rav2 mutants, respectively.

View larger version (71K): [in this window] [in a new window]   Fig. 2.   V-ATPase assembly is defective in rav mutant strains. Wild-type (SF838-5A) and congenic rav1, rav2, and rav1rav2 strains were converted to spheroplasts and biosynthetically labeled with Tran35S-label for 60 min. Proteins were then solubilized under non-denaturing conditions and immunoprecipitated with monoclonal antibody 13D11 against the B subunit of the V1 sector (-V1) or 10D7 (-Vo) against the a subunit of the Vo sector. The immunoprecipitated proteins were separated by SDS-PAGE and visualized on a PhosphorImager (Molecular Dynamics). The positions of known V-ATPase subunits are indicated on the left.

It has been proposed that free V₁ and V_o sectors exist in dynamic equilibrium with fully assembled V-ATPase complexes even in the presence of glucose (14, 39), so the assembly defect in the rav mutants shown in Fig. 2 could arise from the same kinetic delay in reassembly that was observed by glucose deprivation and re-addition. It is also possible, however, that RAVE plays a distinct role in biosynthetic assembly of the V-ATPase. To separate effects of the RAVE complex upon reassembly of disassembled V-ATPase complexes from effects on biosynthetic assembly, we looked at effects of a rav1 mutation on a vma mutant that is incompetent for disassembly of the V-ATPase. The vma11-E145L mutant contains a mutation at the conserved transmembrane glutamate residue of one of the three proteolipid subunits. In this mutant, V-ATPase complexes assemble and are transported to the vacuole, but these complexes have no ATPase activity (26). It has been reported that these complexes are also unable to disassemble in response to glucose deprivation (16), and this result is shown in Fig. 3A. Wild-type and congenic vma11-E145L mutant cells were biosynthetically labeled and then chased in the presence of glucose (+), in the absence of glucose (), or in the absence of glucose followed by glucose re-addition (/+). As shown in Fig. 1, there is extensive disassembly of the wild-type V-ATPase into free V₁ and V_o sectors upon glucose deprivation, as indicated by a loss of V_o subunits from immunoprecipitates with the anti-V₁ antibody and appearance of these subunits in the anti-V_o immunoprecipitate. The dissociated sectors reassemble upon glucose re-addition, as indicated by a reduction of V_o subunits immunoprecipitated by the anti-V_o antibody and an increase in the V_o complexes immunoprecipitated by the anti-V₁ antibody. In contrast, the same level of V_o subunits is coprecipitated by the anti-V₁ antibody from the vma11-E145L cells regardless of the chase conditions, indicating that V-ATPase complexes in this mutant cannot disassemble. We can therefore use this mutant to dissect the effects of the RAVE complex on reassembly of V₁ sectors from potential effects on biosynthetic assembly of the V-ATPase. If RAVE is required only for reassembly of disassembled complexes, we would predict that rav mutants would have no effect on the level of V₁·V_o association in the vma11-E145L mutant, because the V-ATPase does not disassemble in this mutant. This is not what we see in Fig. 3B, however. In the presence of both the rav1 and vma11-E145L mutations, lower levels of V_o subunits are coprecipitated by the anti-V₁ sector antibody and higher levels are coprecipitated by the anti-V_o antibody, indicating that there are higher levels of free V₁ and V_o sectors than in the strain containing only the vma11-E145L mutation. This result clearly indicates that RAVE has a more global, constitutive role in V-ATPase assembly and that it is essential for stable assembly even under conditions where there is no need for reassembly of cytosolic V₁ sectors.

View larger version (50K): [in this window] [in a new window]   Fig. 3.   V-ATPase assembly is defective even in a strain where the V-ATPase cannot disassemble. A, wild-type cells (SF838-5A) and congenic cells with the vma11-E145L allele integrated at the VMA11 locus were converted to spheroplasts and labeled with Tran35S-label for 60 min. Excess methionine and cysteine was then added to the spheroplasts to stop the labeling, and the spheroplasts were pelleted by centrifugation then resuspended in medium containing 2% glucose (+) or no added glucose () for 15 min before solubilization and immunoprecipitation with the anti-V1 and anti-Vo monoclonal antibodies. A third sample (/+) was incubated for 15 min without added glucose, then glucose was added to a final concentration of 2% for an additional 15 min, and then sample was processed for immunoprecipitation. All of the samples were subjected to SDS-PAGE and visualized on a PhosphorImager (Molecular Dynamics). The positions of known V-ATPase subunits are shown on the left. B, the two strains described in A, along with congenic strains of each carrying the rav1 mutation, were labeled for 60 min as described in A and then solubilized under non-denaturing conditions. Fully and partially assembled subcomplexes were immunoprecipitated with anti-V1 and anti-Vo antibodies and analyzed as described in A.

To better understand the role of the RAVE complex in V-ATPase assembly, we examined the genetic and biochemical characteristics of the rav mutants in more detail. Seol et al. (17) reported that cells with mutations in the RAVE complex exhibit growth defects characteristic of loss of V-ATPase function (Vma phenotype), but that these defects were temperature-dependent and milder than in a V-ATPase subunit deletion mutant. They assessed the Vma growth phenotype by growth on medium buffered to pH 7.5 or medium containing glycerol as the sole carbon source. We extended these studies to monitoring growth of rav1, rav2, and rav1rav2 mutants on medium buffered to pH 7.5 containing 60 mM CaCl₂, which provides a more stringent selection for the Vma phenotype. As shown in Fig. 4, the vma4 mutant, which lacks a structural subunit of the V-ATPase, is completely unable to grow on the pH 7.5 + CaCl₂ medium at 30° or 37 °C but does grow on pH 5.0 medium at both temperatures. The rav2 mutant grows as well as wild-type cells under both conditions at 30 °C but is unable to grow at all on the pH 7.5 + CaCl₂ plates at 37 °C. The rav1 and rav1rav2 double mutants grow poorly in elevated pH and calcium even at 30 °C and are unable to grow at 37 °C. The results shown here parallel those of Seol et al. but emphasize that the V-ATPase in the rav1 and the double mutant is clearly defective at both 30 °C and 37 °C.

View larger version (53K): [in this window] [in a new window]   Fig. 4.   Comparison of growth phenotypes of wild-type, vma, and rav mutants strains. Logarithmically growing cells (SF838-5A and the congenic vma4, rav1, rav2, and rav1rav2 mutants) were suspended to a density of 1 OD/ml, then serially diluted in a microtiter plate. Constant volumes of each dilution were transferred to plates of YEPD, pH 5.0 (pH 5.0), or YEPD, pH 7.5, containing 60 mM CaCl2 (pH 7.5 + Ca2+) for growth at 30 °C or 37 °C.

The immunoprecipitation experiment shown in Fig. 2 was conducted at 30 °C, but we also compared the extent of V₁·V_o assembly in cells incubated at 25 °C and 37 °C. The assembly defects of the mutants were similar at all three temperatures (not shown). These results indicate that assembly of V₁ and V_o sectors occurs in the rav mutants, as demonstrated by Seol et al. (17), but that attachment of V₁ to V_o is severely defective even at temperatures that are semi-permissive for growth at elevated pH and calcium. This suggests that RAVE is required for stable V₁·V_o assembly under all conditions, but that the structural defects that occur in the absence of RAVE function may be better tolerated in vivo at 25-30 °C.

To further characterize V-ATPase complexes in the rav mutants, we isolated vacuolar vesicles from the single and double mutants and assessed their concanamycin-sensitive ATPase activity and the levels of V₁ and V_o subunits present. As shown in Fig. 5, the vacuolar vesicles from the three mutant strains contained only 1-2% as much concanamycin A-sensitive ATPase activity as the wild-type cells. (This is approximately the same level of activity we observe in a vma strain that lacks one of the structural subunits of the enzyme (40).) Immunoblots of the isolated vesicles shown in Fig. 5 indicate that the defect in activity reflects, at least in part, an assembly defect. Although the mutant vesicles contain as much of the V_o subunit a as the wild-type vesicles, they have much lower levels of V₁ subunits A, B, C, and E. By comparing immunoblots containing different dilutions of the wild-type vesicles to the mutant vesicles (not shown), we can estimate that the mutants contain 20% of the A, B, and E subunits in the wild-type membranes but only 5% of the C subunit. (Although the C subunit is not visible in the blot shown in Fig. 5, it could be detected at higher loads of vesicles.) These results support those shown in Fig. 2 and emphasize that even the population of V-ATPase complexes providing function to the mutants at 30 °C is structurally defective.

View larger version (31K): [in this window] [in a new window]   Fig. 5.   Immunoblot of V-ATPase subunits and measurement of V-ATPase activity in vacuolar vesicles. Vacuolar vesicles were isolated from wild-type cells (SF838-5A) or the congenic rav1, rav2, and rav1rav2 mutants. Vesicles were solubilized in cracking buffer, separated by SDS-PAGE and probed with monoclonal antibodies against the Vo a subunit and V1 subunits A, B, and C or polyclonal antisera against the E subunit as described under "Experimental Procedures." 15 µg of vacuolar protein was loaded for visualization of the a and C subunits; 5 µg was loaded for visualization of the A, B, and E subunits. ATPase activity sensitive to 100 nM concanamycin A was measured for each vacuole preparation and compared with the activity of the wild-type strain. Wild-type V-ATPase specific activity was 4.6 µmol/min/mg for the vacuolar preparation shown here. The percentage of the wild-type V-ATPase activity is shown at the bottom.

RAVE Binds to the Stalk Region of Cytosolic V₁ Sectors-- One key step in understanding how RAVE exerts its effects on V₁·V_o association is to define the site of interaction between the RAVE and V₁ complexes. Seol et al. (17) provided evidence that the V₁·RAVE interaction was severely compromised in a rav1 mutant, suggesting that Rav1p contains an important part of the V₁ binding site on RAVE. The subcomplexes formed when any one of the V₁ subunits is missing have been examined (10, 11, 41), and we used this information to narrow down the region of the V₁ complex involved in binding to RAVE. Cytosolic fractions were prepared from all of the individual V₁ subunit deletion strains, subcomplexes were immunoprecipitated with monoclonal antibody against the B subunit, and the immunoprecipitated complexes were separated by SDS-PAGE and immunoblotted to detect the level of myc-tagged Rav1p bound to the subcomplexes. The immunoblots of the cytosolic fractions before immunoprecipitation shown in Fig. 6 indicate that Rav1p is present at comparable levels in all of the vma mutant strains. (There was a consistent shift in the size of the myc₁₃-tagged Rav1p in the vma5 mutant. We cannot account for this difference yet, but we have shown that the Vma phenotype of this mutant is fully complemented by VMA5 on a plasmid, suggesting that the RAV1 gene is not mutated.) Vma5p (subunit C) is the only V₁ subunit not present in cytosolic V₁ complexes (6); nearly wild-type levels of Rav1p are coprecipitated with V₁ from the vma5 strain (Fig. 6, lower panel). Deletion of Vma13p (subunit H) results in cytosolic complexes containing all V₁ subunits except subunits C and H (6); wild-type levels of Rav1p are also coprecipitated with the anti-V₁ antibody from the vma13 strain, indicating that the H subunit is not essential for Rav1p binding. Vma7p, Vma8p, Vma4p, and Vma10p all encode putative stalk subunits of V₁, and low resolution structures of Radermacher et al. (42, 43) indicate that the two stalks envisioned in the intact V-ATPase (44) may collapse into a single stalk in free V₁ complexes. Nevertheless, we previously observed that different combinations of stalk subunits coprecipitated with the B subunit from deletion strains lacking individual stalk subunits (10, 11). In Fig. 6, Rav1p is coprecipitated at nearly wild-type levels from the vma7 and vma8 mutants, is not coprecipitated from the vma4, and is coprecipitated only at very low levels from vma10 mutants. (Note that Vma4p is unstable in the vma10 mutant, so that the vma10 strain contains very low levels of Vma4p (45).) Rav1p is also poorly coprecipitated with the A subunit from a vma2 (B subunit deletion) or with the B subunit from a vma1 (A subunit deletion) strain. In these strains predominantly the A and B subunits are immunoprecipitated by their respective antibodies, and almost all of the stalk subunits are missing (10, 11). We believe that the absence of the stalk subunits accounts for the defect in Rav1 binding in these strains, because if the A and/or B subunits were the site of Rav1 binding, we would expect to retain binding in the vma4 and vma10 mutants, where the A and B subunits still bind to each other at apparently normal stoichiometry (10, 11). Taken together, these results suggest that the binding site for RAVE may be in the stalk region of V₁, involving the E (Vma4p) and possibly the G (Vma10p) subunits, although we cannot eliminate the possibility that deletion of these subunits changes the conformation of the remaining subunits in a manner that prevents RAVE binding (see "Discussion").

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Comparison of Rav1p-V1 interactions in wild-type cells and vma deletion strains. Cytosol was prepared from mutant strains containing myc-tagged Rav1 and lacking each of the V1 subunits of the yeast V-ATPase in parallel with the congenic wild-type strain (shown as the first sample of each set). Protein was directly trichloroacetic acid-precipitated from 0.4 mg of cytosol from each strain (Cytosol). 4.0 mg of cytosol from each strain was subjected to immunoprecipitation with anti-V1 B subunit antibody as described in Fig. 1. All of the samples were solubilized in cracking buffer, subjected to SDS-PAGE, and transferred to nitrocellulose. The level of Rav1p in each sample was analyzed by probing immunoblots with anti-myc antibody capable of detecting the myc-tagged Rav1 protein.

Skp1p Is a Constitutive Subunit of the RAVE Complex-- Many of the Skp1p-containing SCF complexes are relatively unstable; the subunit composition of the complexes changes in response to cellular conditions, either because of preferential binding of post-translationally modified subunits to the complex or because of degradation of the F-box subunit (46, 47). The RAVE complex was originally identified through the binding of Rav1p and Rav2p to Skp1p, but this interaction was observed under a uniform set of growth conditions. We investigated the possibility that Skp1p is transiently recruited to the RAVE complex. Seol et al. (17) provided evidence that Skp1p does not bind directly to V₁ but instead requires the presence of Rav1p as a bridge to V₁. We first determined whether the association of Skp1p with Rav1p changed in response to glucose deprivation and re-addition. As shown in Fig. 1A, RAVE transiently interacts with V₁ released from the membrane by glucose deprivation; we were interested in determining whether the level of Skp1p interaction might change with the level of V₁ interaction. Fig. 7A indicates that this does not occur. The total level of Skp1p in the cytosol is constant with glucose deprivation and re-addition (left panel), and the amount of Skp1p co-precipitated with the myc-tagged Rav1p is also constant under all of these conditions. We next asked whether Skp1p remained bound to Rav1p in a vma4 strain, where very little RAVE is bound to V₁ (Fig. 6). Fig. 7B shows that Skp1p is also coprecipitated with myc-tagged Rav1p to similar levels from the wild-type and vma4 strains. These data indicate that the association of Skp1p with Rav1p is stable and eliminate models in which Skp1p is recruited to the RAVE complex in response to, or as a stimulus of, V₁ binding.

View larger version (52K): [in this window] [in a new window]   Fig. 7.   Interaction of Skp1p with RAVE under varied conditions. A, cytosol was prepared from RDY1512 cells treated as described in Fig. 1, but the myc antibody was used to immunoprecipitate myc-tagged Rav1 and associated proteins (IP/-myc). Immunoblots were probed with anti-myc, to detect myc-tagged Rav1p or polyclonal antibody against Skp1p (Skp1). B, cytosol was prepared from RDY1512 and the congenic vma4::URA3 strain after growth of cells to log phase in glucose, then trichloroacetic acid-precipitated or subjected to immunoprecipitation as described in A. The levels of myc-tagged Rav1p and Skp1p were detected as described in A.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

RAVE Has a Constitutive Role in V-ATPase Assembly-- The results shown here support the suggestion of Seol et al. (17) that RAVE plays a role in reassembly of V-ATPase complexes disassembled in response to glucose deprivation. We have diagrammed this function in Fig. 8. We have shown that RAVE is recruited into complexes with cytosolic V₁ under conditions where V₁ is released from the vacuolar membrane, but that this interaction is reversible under conditions where V₁ reattaches to V_o. These results suggest that the RAVE·V₁ interaction is sensitive to extracellular glucose concentration, but the results with the vma3 mutant indicate that this sensitivity involves more than a change in intrinsic affinity that might be caused by modification of one of the binding partners, for example. One possible mechanism is that V_o successfully competes with RAVE for V₁ binding when glucose is present but is not an effective competitor in the absence of glucose. Although this model would account for the differences between RAVE·V₁ binding in the wild-type and vma3 cells, it would also require rapid communication between membrane-bound V_o complexes and cytosolic V₁·RAVE complexes. Future experiments will address the function of RAVE in glucose-induced reassembly in more detail.

View larger version (23K): [in this window] [in a new window]   Fig. 8.   Models for RAVE activity in biosynthesis and reassembly of the V-ATPase.

Although RAVE may be a "regulator" of V-ATPase assembly important in reassembly of disassembled complexes, this name may underestimate its importance in generating stable V-ATPase complexes. The results presented here suggest that V-ATPase complexes formed in the absence of a functional RAVE complex are structurally defective, even under conditions where cycling between assembled V-ATPase complexes and free V₁ and V_o complexes should be blocked (the vma11-E145L mutant). This suggests a critical role for RAVE in biosynthetic assembly, and Fig. 8 suggests two points at which RAVE might act in biosynthesis of V-ATPases. It is probably easiest to envision RAVE as stimulating association of assembled V₁ and assembled V_o sectors in biosynthesis, much as it does in reassembly. This model sees RAVE as intervening in the last step of an independent assembly pathway (Fig. 8) and appears to account for the presence of free V₁ and V_o sectors in rav mutants (Fig. 2). The problem with this model is that we have previously shown that concerted assembly of V₁ and V_o subunits is the predominant assembly pathway when all of the subunits are present (11), and pulse-chase studies in both yeast and mammalian cells failed to show the attachment of newly synthesized V₁ and V_o sectors predicted by the independent assembly pathway (8, 11). These data lead us to hypothesize that RAVE intervenes at some point in the concerted assembly pathway in Fig. 8. Although this role seems more dissimilar to the proposed role of RAVE in reassembly, it still may be that RAVE catalyzes incorporation of a critical subunit or exerts another common assembly function in both settings. It is notable that, when one subunit of the V₁ or V_o sector of the yeast V-ATPase is missing, then the independent assembly pathway appears to take over, resulting in assembly of either free V_o complexes or free V₁ complexes (11). Therefore, formation of free, assembled V₁ and V_o sectors might be predicted in a rav mutant whether the concerted or independent assembly pathway is disrupted. Regardless of the exact stage at which the RAVE complex exerts its effects on biosynthetic assembly, the results clearly indicate that there are common features between biosynthetic assembly of V-ATPases and reassembly of disassembled ATPase complexes that were not obvious in the past, including a requirement for RAVE intervention in both processes.

How Does RAVE Exert Its Effects on V-ATPase Assembly?-- We do not yet have an answer to this central question, but the data presented here provide some important clues. The results shown in Fig. 6 strongly suggest that subunit E (Vma4p) and possibly subunit G (Vma10p) are critical for binding of RAVE to cytosolic V₁ complexes. Although it is certainly possible that the vma4 and vma10 mutations alter the conformation of V₁ in a manner that prevents RAVE binding to another V₁ subunit, it is significant that deletion of three other stalk subunits, Vma13p, Vma8p, and Vma7p, has much less effect on Rav1 binding, clearly indicating that RAVE does not require an intact V₁ complex for binding. The E and G subunits show genetic and biochemical interactions with each other (45, 48, 49), and current evidence would place them in the peripheral, or stator, stalk of the V-ATPase (50). The cytoplasmic domain of the V_o a subunit is believed to be part of this stalk (51), so this result might suggest that RAVE binds at a site on V₁ that becomes available only when it is released from V_o and thus plays a role in proper assembly of the peripheral stalk. This model for RAVE action implies that the RAVE complex does not bind fully assembled, membrane-bound V-ATPase complexes. We see small amounts of the RAVE subunits in isolated vacuolar membranes from glucose-grown cells, but the subunits are also present in vacuoles from a vma3 strain, which has no V-ATPase subunits at the vacuole (29), indicating that they do not interact specifically with membrane-bound V-ATPases (data not shown). Subunit C (Vma5p) has also been assigned to the peripheral stalk (48) and it is intriguing that a number of subunit C mutations generate phenotypes with some similarity to the rav mutations; specifically, these vma5 mutants do not exhibit a strong Vma phenotype in vivo, but the mutant V-ATPase complexes show moderate to severe assembly defects in vitro (40). It is possible that RAVE is directly involved in attachment of subunit C to the peripheral stalk; this could account for the phenotypic similarity between vma5 and rav mutants and would also be consistent with the observation that RAVE binds to cytosolic V₁ complexes under conditions where subunit C is released, such as glucose deprivation. Alternatively, RAVE and subunit C could play parallel but distinct roles in generating a stably assembled V-ATPase.

RAVE is one of a number of recently identified Skp1p-containing complexes that does not adhere to the SCF model or have any apparent proteolytic function. Although the function of Skp1p in these complexes is not understood, it does appear to play an essential role in many of the complexes (52, 53). Consistent with this, we have also found that mutant rav1 proteins that bind V₁ but not Skp1p generate a Vma mutant phenotype similar to that of a rav1 mutant.² It is also significant that Skp1p is a stable component of the RAVE complex (Fig. 7). Under a number of conditions, the level of Skp1p appears to be limiting for cell growth, and the instability of F-box proteins Grr1p and Cdc4p has been invoked as a mechanism for releasing limited levels of Skp1p for alternative cellular functions (46, 47). In this context, it is somewhat surprising that Skp1p remains bound to RAVE regardless of extracellular carbon source or the potential for V₁ binding. We have not determined the percentage of the total Skp1p dedicated to the RAVE complex, so it is possible that RAVE only occupies a small portion of the Skp1p pool relative to the Grr1p- or Cdc4p-containing complexes. It is also possible that the RAVE complex has essential functions beyond its role in V-ATPase assembly. Discovering these functions may further elucidate the biochemical function of RAVE and possibly, the functions of other non-proteolytic Skp1p containing complexes.

	ACKNOWLEDGEMENTS

We thank Ray Deshaies for generously providing reagents and sharing results prior to publication and Mark Longtine for providing the pFA6 series of plasmids.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant GM50322 (to P. M. K.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, State University of New York, Upstate Medical University, 750 East Adams St., Syracuse, NY 13210. Tel.: 315-464-8742; Fax: 315-464-8736; E-mail: kanepm@upstate.edu.

Published, JBC Papers in Press, February 13, 2002, DOI 10.1074/jbc.M200682200

² A. M. Smardon and P. M. Kane, unpublished data.

	ABBREVIATIONS

The abbreviations used are: V-ATPase, vacuolar proton-translocating ATPase; RAVE, regulator of the ATPase of vacuolar and endosomal membranes; YEPD, yeast extract-peptone-2% dextrose medium; SD, fully supplemented minimal medium; SCF, Skp1-cullin-F-box; E2, ubiquitin conjugating enzyme; E3, ubiquitin-protein isopeptide ligase; Mes, 4-morpholineethanesulfonic acid; MOPS, 4-morpholinepropanesulfonic acid.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES