Function of the COOH-terminal Domain of Vph1p in Activity and Assembly of the Yeast V-ATPase*

We have previously shown that mutations in buried charged residues in the last two transmembrane helices of Vph1p (the 100-kDa subunit of the yeast V-ATPase) inhibit proton transport and ATPase activity (Leng, X. H., Manolson, M., Liu, Q., and Forgac, M. (1996) J. Biol. Chem . 271, 22487–22493). In this report we have further explored the function of this region of Vph1p (residues 721–840) using a combination of site-directed and random mutagenesis. Effects of mutations on stability of Vph1p, assembly of the V-ATPase complex, 9-amino-6-chloro-2-methoxyacridine quenching (as a measure of proton transport), and ATPase activity were assessed. Additional mutations were analyzed to test the impor-tance of Glu-789 in TM7 and His-743 in TM6. Although substitution of Asp for Glu at position 789 led to a 50% decrease in 9-amino-6-chloro-2-methoxyacridine quenching, substitution of Ala at this position gave a mutant with 40% quenching relative to wild type, suggesting that a negative charge at this position is not absolutely essential for proton transport. Similarly, a positive charge is not essential at position His-743, since the H743Y and H743A mutants retain 20 and 60% of wild-type quenching, respectively. Interestingly, H743A approaches wild-type ATPase activity at elevated pH while the E789D 2 ) as described previously (34) in the presence or absence of 10 n M bafilo-mycin A 1 , a specific inhibitor of the V-ATPase (35). SDS-polyacrylamide gel electrophoresis was carried out as described by Laemmli (36). For immunoblot analysis, whole cell lysates were prepared from overnight cultures grown in Ura 2 SD medium using glass bead homog-enization as described previously (37). Whole cell lysates or vacuolar membrane vesicle were subjected to SDS-PAGE and Western blots were probed with the mouse monoclonal antibodies 8B1-F3 against the 69- kDa subunit or 10D7 against the 100-kDa subunit (Molecular Probes, Inc.), followed by horseradish peroxidase-conjugated secondary anti- body (Bio-Rad). Blots were developed using a chemiluminescent detec-tion method obtained from Kirkegaard & Perry Laboratories. Quanti- tations were done using an IS-1000 Digital imaging system (Alpha Innotech Corp.). Metabolic labeling of cells and immunoprecipitations of the V-ATPase were carried out as described previously (23), except that the immunoprecipitated samples were analyzed by SDS-PAGE on 15% acrylamide gels instead of 12% acrylamide gels. The pH profile of V-ATPase activity for individual mutants was measured using a buffer system containing 20 m M each of the following: MES, MOPS, Tris, and glycine.

The vacuolar (H ϩ )-ATPases (or V-ATPases) 1 are a widely distributed family of ATP-driven proton pumps responsible for acidifying the interior of various intracellular organelles and providing the energy for numerous coupled transport processes (for reviews, see Refs. [1][2][3][4][5][6][7][8][9]. The V-ATPases also play an important role in such processes as receptor-mediated endocytosis, intracellular membrane traffic, and protein processing and degradation. V-ATPases in the plasma membrane of specialized cells function in renal acidification, pH homeostasis, bone resorption, and tumor metastasis (7, 10 -12).
The V-ATPases are multisubunit complexes composed of two structural domains (1)(2)(3)(4)(5)(6)(7)(8)(9). The peripheral V 1 domain is a 500-kDa complex with the structure A 3 B 3 C 1 D 1 E 1 F 1 G 1 H 1 which is responsible for ATP hydrolysis, while the integral V 0 domain is a 250-kDa complex with the structure a 1 d 1 cЉ 1 (c, cЈ) 6 which is responsible for proton translocation across the membrane. In Saccharomyces cerevisiae, the V-ATPase subunits are encoded by at least 14 genes (1). Acidification of the central vacuole by the V-ATPase is important both to maintain the activity of degradative enzymes and to drive uptake of solutes such as Ca 2ϩ and amino acids (5). Disruption of any of the V-ATPase subunit genes leads to a conditional lethal phenotype (Vma Ϫ ) in which cells are unable to grow at neutral pH and in the presence of elevated Ca 2ϩ but are able to grow at acidic pH (5,13,14).
Although the V-ATPases are structurally and evolutionarily related to the F-ATPases (15)(16)(17)(18)(19)(20), there is no obvious structural homolog for the 100-kDa subunit of the V-ATPases. In yeast, the 100-kDa subunit is encoded by two homologous genes, VPH1 and STV1 (21,22). Both genes encode proteins containing an amino-terminal hydrophilic domain of about 45 kDa and a carboxyl-terminal hydrophobic domain of about 55 kDa which contains six to seven putative transmembrane helices. Disruption of both VPH1 and STV1 gives a typical vma Ϫ phenotype, whereas disruption of VPH1 alone leads to only partial loss of vacuolar acidification and reduced growth at neutral pH. The incomplete vma Ϫ phenotype of the VPH1 knockout is due to the presence of Stv1p, which can partially compensate for the absence of Vph1p. Immunocytochemical studies indicate that Vph1p is targeted to the central vacuole whereas Stv1p is normally targeted to some other intracellular compartment, possibly endosomes (22). In the absence of Vph1p, however, overexpression of STV1 results in mistargeting of sufficient Stv1p to the vacuole to give a nearly wild-type phenotype (22). Disruption of STV1 alone has no obvious phenotypic consequences.
Our previous mutagenesis studies of Vph1p suggested a possible role of the 100-kDa subunit in proton translocation (23). Mutagenesis of several charged residues in the last four transmembrane helices of Vph1p led to significant loss of ATP-de-pendent proton transport without discernable effects on either stability or assembly of the V-ATPase complex. In particular, mutation of Glu-789 in the last transmembrane helix to Gln led to almost complete loss of both proton transport and ATPase activity. These results are similar to those obtained for the Escherichia coli F 0 a subunit (24 -27), suggesting that the 100-kDa subunit may function as the a subunit homolog in the V-ATPases.
To further explore the function of this COOH-terminal region of Vph1p, particularly the last two transmembrane helices, we have carried out additional site-directed as well as random mutagenesis. Our results suggest that this segment of Vph1p plays a crucial role in organization as well as ATP-dependent proton transport by the V-ATPase complex.

EXPERIMENTAL PROCEDURES
Materials and Strains-Zymolyase 100T was obtained from Seikagaku America, Inc. Tran 35 S-Label was purchased from ICN. Bafilomycin A1 was a kind gift from Dr. Karlheinz Altendorf (University of Osnabruck). Leupeptin was from Boehringer Mannheim. 9-Amino-6chloro-2-methoxyacridine (ACMA) was from Molecular Probes, Inc. E. coli and yeast culture media were purchased from Difco Laboratories. PCR reagents and other molecular biology reagents were from Promega and New England Biolabs. ATP, phenylmethylsulfonyl fluoride, and most other chemicals were purchased from Sigma.
Mutations were confirmed by DNA sequencing using the dideoxy chain termination method (29) (Sequenase TM Version 2.0 DNA Sequencing Kit, U. S. Biochemical Corp.). Fragments containing the indicated mutations were then substituted back into the vector containing the wild-type VPH1 cDNA. No other mutations were detected in the final product.
Random Mutagenesis-Random mutagenesis was carried out using a polymerase chain reaction (PCR) based strategy. Oligonucleotide primers corresponding to nucleotides close to both ends of the COOH-terminal segment (from amino acid residue 721 to 840) were used to amplify this region of the VPH1. The first cycle of PCR was carried out in the presence of 200 M of each of three dNTPs and 0.5 M of the fourth dNTP, with 200 M of the fourth dNTP added in the subsequent 24 cycles. Four separate PCR reaction mixtures (with dA, dC, dG, and dT, each at low concentration in the first cycle) were combined, purified, and digested with EcoRI and BamHI. The purified EcoRI-BamHI PCR fragments were then cloned back into the wild-type VPH1 cDNA in the yeast expression shuttle vector pRS316. The ligation mixture was then used to transform E. coli XL-blue cells and the plasmids purified from the pooled transformants and used for transformation of yeast.
Transformation and Selection-Yeast cells (MM112) lacking functional endogenous 100-kDa subunit were transformed using the lithium acetate method (30) with the wild-type VPH1 in pRS316 as a positive control and pRS316 vector alone as a negative control. The transformants were selected on uracil minus (Ura Ϫ ) plates as described previously (31). Growth phenotypes of the mutants were assessed on YPD plates buffered with 50 mM KH 2 PO 4 and 50 mM succinic acid to either pH 5.5 or 7.5. For testing calcium sensitivity, YPD plates buffered with 50 mM MES and 50 mM MOPS, pH 7.5, were supplemented with 50 mM calcium chloride.
To screen the randomly mutated VPH1, the pooled plasmids were introduced into the yeast MM112 using the lithium acetate method and the yeast colonies grown on Ura Ϫ plates. Colonies were screened by replica plating on plates buffered to pH 5.5 and 7.5 since mutants lacking V-ATPase activity fail to grow on medium buffered to pH 7.5 (with or without Ca 2ϩ ), but can grow on medium buffered to pH 5.5 (5,13,14). Mutants unable to grow at pH 7.5 were verified by retransforming the recovered plasmid into yeast and examining the growth phenotype. DNA sequencing (29) of the recovered plasmid was used to identify the site of mutation and to confirm that no other mutations were present.
Other Procedures-Vacuolar membrane vesicles were isolated as described previously (23). Protein concentrations were determined by the method of Lowry et al. (32). ATPase activities were measured using a coupled spectrophotometric assay (33) with the modification of using 0.35 mM NADH instead of 0.5 mM NADH. ATP-dependent proton transport was measured by fluorescence quenching using the fluorescence probe ACMA in transport buffer (25 mM MES/Tris, pH 7.2, 5 mM MgCl 2 ) as described previously (34) in the presence or absence of 10 nM bafilomycin A 1 , a specific inhibitor of the V-ATPase (35). SDS-polyacrylamide gel electrophoresis was carried out as described by Laemmli (36).
For immunoblot analysis, whole cell lysates were prepared from overnight cultures grown in Ura Ϫ SD medium using glass bead homogenization as described previously (37). Whole cell lysates or vacuolar membrane vesicle were subjected to SDS-PAGE and Western blots were probed with the mouse monoclonal antibodies 8B1-F3 against the 69-kDa subunit or 10D7 against the 100-kDa subunit (Molecular Probes, Inc.), followed by horseradish peroxidase-conjugated secondary antibody (Bio-Rad). Blots were developed using a chemiluminescent detection method obtained from Kirkegaard & Perry Laboratories. Quantitations were done using an IS-1000 Digital imaging system (Alpha Innotech Corp.).
Metabolic labeling of cells and immunoprecipitations of the V-ATPase were carried out as described previously (23), except that the immunoprecipitated samples were analyzed by SDS-PAGE on 15% acrylamide gels instead of 12% acrylamide gels. The pH profile of V-ATPase activity for individual mutants was measured using a buffer system containing 20 mM each of the following: MES, MOPS, Tris, and glycine.

RESULTS
We have carried out mutagenesis of Vph1p in a yeast strain in which both the VPH1 and STV1 genes encoding the 100-kDa subunit of the V-ATPase have been disrupted, leading to typical Vma Ϫ growth phenotype. These cells are unable to grow at neutral pH (pH 7.5) but are able to grow at acidic pH (pH 5.5). Mutations were introduced into the VPH1 gene using either site-directed or random mutagenesis and the mutant forms of VPH1 expressed in the double deletion strain using the shuttle vector pRS316.
We have previously shown that mutation of Glu-789 in transmembrane helix seven of Vph1p to Gln results in significant loss of proton transport and ATPase activity while having no measurable effect on stability or assembly of the V-ATPase complex (23). To further investigate the role of this buried charged residue, additional site-directed mutations were constructed. Both ATP-dependent ACMA quencing (as a measure of proton transport) and bafilomycin-sensitive ATPase activity were measured in vacuoles isolated from wild-type and mutant strains. As can be seen in Fig. 1, ACMA quenching of the Glu-789 mutants varied from 20 to 50% relative to wild-type, with the highest activity observed for the E789D mutant. To assess the effects of these mutations on assembly of the V-ATPase complex, cells were metabolically labeled with 35 S and the V-ATPase solubilized with detergent and immunoprecipitated using an antibody directed against the A subunit. As can be seen in Fig. 2, none of the mutations tested had a detectable effect on assembly of the V-ATPase complex.
Interestingly, substitution of Ala at position Glu-789 resulted in a mutant which retains 40% wild-type ACMA quenching, suggesting that a charged residue at this position is not absolutely essential for proton transport. Because we had also previously observed an approximately 40% decrease in pumping activity on substitution of Ala for His-743 in TM6 (23), we constructed a double mutant in which Glu-789 and His-743 were interchanged. As can be seen in Fig. 1, the ACMA quenching activity of the E789H/H743E double mutant remained low relative to wild type. Because ACMA quenching is not linearly related to proton transport, the apparent differences between ACMA quenching and ATPase activity observed for some of the mutants may not be meaningful.
To further explore the function of the COOH-terminal domain (residues 721 to 840), PCR-based random mutagenesis was performed on this region. Following mutagenesis and transformation, transformants were selected for their inability to grow at pH 7.5. Mutants that failed to grow at neutral pH were then screened for stability of the 100-kDa subunit by Western blot analysis on whole cell lysates using monoclonal antibody 10D7 which is specific for Vph1p. Mutants which showed reduced levels of Vph1p were not subjected to further study. Plasmids from mutants showing normal levels of Vph1p in the whole cell lysate were rescued, purified, and reintroduced into the double deletion strain. Plasmids which continued to confer a mutant phenotype were sequenced and Western blots performed on cells transformed with the purified plasmids. As can be seen in Fig. 3, panel A, all of the mutants isolated showed nearly wild-type levels of Vph1p in whole cell lysates. It should be noted that some of the mutants which were discarded in the initial screening may have failed to show normal Western blotting by 10D7 because the mutations resulted in disruption of the epitope recognized by this antibody rather than because they resulted in destabilization of the protein.
To assess whether the mutant 100-kDa proteins were targeted correctly to the vacuole, Western blot analysis was performed using antibody 10D7 against Vph1p on isolated vacuolar membranes. As can be seen in Fig. 3, panel B, two of the mutants (F809L and G814D) showed significantly lower levels of Vph1p in the vacuolar membrane, while several other mutants showed smaller reductions. Because disruption of VPH1 leads to the loss of assembly of V 1 onto the vacuolar membrane (21), we also carried out Western blot analysis on vacuolar membranes using monoclonal antibody 8B1-F3 against the A subunit as a measure of assembly of the V-ATPase complex. As can be seen in Fig. 3, panel C, several of the mutants show significant reductions in the level of A subunit associated with the vacuolar membrane, including L780S, L800S, W802R, F809L, and G814D. These mutants would therefore appear to be defective in assembly of the V-ATPase complex. Alternatively, mutations in the 100-kDa subunit may lead to changes in accessibility or stability of the epitope recognized by the monoclonal antibody 8B1-F3 unrelated to assembly of the V-ATPase complex.
Both ATP-dependent ACMA quenching and bafilomycin-sensitive ATPase activity were measured in vacuoles isolated from wild-type and mutant cells (Fig. 4). Also shown in Fig. 4 is ATPase activity and ATPdependent proton transport (as assessed by ACMA quenching) were measured on aliquots of purified vacuolar membrane vesicles containing 2.5 g of protein as described under "Experimental Procedures." Activities are expressed relative to the ⌬vph1⌬stv1 strain expressing the pRS316 plasmid containing the wild-type VPH1 gene (defined as 100%). All of the ATP-dependent ACMA quenching in vacuolar membrane vesicles isolated from cells expressing the wild-type Vph1p was inhibitable by 10 nM bafilomycin. No ATP-dependent ACMA quenching or bafilomycin-sensitive ATPase activity was observed in vacuolar membrane vesicles isolated from cells transformed with the vector alone. The specific activity of the bafilomycin-sensitive ATPase observed for wild-type vacuoles was 3.3 mol of ATP/min/mg of protein at saturating ATP and 37°C. Each bar represents the average of two or three determinations made on two or three independent vacuolar membrane preparations, with the error corresponding to the standard deviation.
quantitation of the levels of the 100-kDa subunit in cell lysates and vacuoles and levels of the A subunit in vacuoles derived from Western blot analysis (Fig. 3) using an Alpha Innotech digital imaging system. As can be seen, ACMA quenching of the mutants ranged from 0 to 30% of wild type. The reduced proton transport activity of the mutants is consistent with their growth phenotype. Those mutants showing significantly lower ACMA quenching than predicted on the basis of the observed levels of A subunit present on the vacuolar membrane include H729R, L739S, H743R, H743Y, L746S, W802R, V803D, and V803F. The G814D mutant showed higher levels of ACMA quenching and ATPase activity than predicted on the basis of the level of vacuole-associated A subunit. It is possible that in this case, mutation of the 100-kDa subunit has resulted in increased protease sensitivity of the A subunit at a site that does not effect activity but does alter its recognition by the monoclonal antibody 8B1-F3.
To further assess the effects of the mutations on assembly of the V-ATPase complex, cells were metabolically labeled with 35 S and the V-ATPase solubilized with detergent and immunoprecipitated using antibody 8B1-F3 against the A subunit. As shown in Fig. 5, the immunoprecipitation pattern observed for the H743Y, H743R, L746S, and V803D mutants is virtually indistinguishable from that observed for wild type. The H729R and L739S mutants both show nearly normal assembly, although a reduction in intensity of a band at 14 kDa can be observed. The remaining mutants all show varying degrees of defective assembly, as indicated by reductions in the intensity of the V 0 bands (molecular mass 100, 36, and 17 kDa) immunoprecipitated using an antibody against a V 1 subunit. Interestingly, although Western blot analysis would suggest relatively good assembly of the V803F mutant (Figs. 3 and 4), immunoprecipitation indicates a significant dissociation of the V 1 and V 0 complexes, suggesting that the V803F mutation may lead to destabilization of the V-ATPase complex such that dissociation occurs during detergent solubilization and immunoprecipitation.
To further characterize the activity properties of the mutants derived by both site-directed and random mutagenesis, the pH dependence of V-ATPase activity was measured over the range 5.5 to 10.5 for vacuoles isolated from both wild-type and mutant strains. As can be seen in Fig. 6, whereas the wild-type enzyme shows a well defined pH optimum at approximately 7.5, several of the mutants (including E789A, E789Q, and H743A) show a pH optimum shifted to significantly higher pH. In the case of the H743A mutant, the ATPase activity of the mutant enzyme actually reaches wild-type levels at pH 9. By contrast, the E789D mutant shows a pH optimum approximately 0.5 units lower than for the wild-type, consistent with the slightly lower pK a of Asp relative to Glu in aqueous solution. These results suggest that, while not absolutely essential for proton transport, both Glu-789 and His-743 are in a position to significantly influence V-ATPase activity.

FIG. 5. Effect of Vph1p random mutations on V-ATPase assembly.
Yeast cells (the ⌬vph1⌬stv1 strain) expressing the wild-type VPH1 gene, the indicated mutations, or the vector alone were grown overnight in methionine-free media, converted to spheroplasts, metabolically labeled with 35 S, and the V-ATPase solubilized and immunoprecipitated using the antibody 8B1-F3 followed by SDS-PAGE and autoradiography as described in the legend to Fig. 2. The positions of the V-ATPase subunits are indicated and were confirmed by comparison with the migration of 14 C-labeled molecular weight standards.

FIG. 4. Comparison of effects of
Vph1p random mutations on bafilomycin-sensitive ATPase activity, ATPdependent ACMA quenching, Vph1p stability, and the presence of Vph1p and subunit A in isolated vacuolar membrane vesicles. Bafilomycin-sensitive ATPase activity and ATP-dependent ACMA quenching were measured on vacuolar membrane vesicles as described in the legend to Fig. 1. Each bar represents the average of two determinations made on a single vacuolar preparation. Determination of the levels of Vph1p in whole cell lysates and vacuolar membranes and the A subunit in vacuolar membranes was carried out by Western blot analysis as described in the legend to Fig. 3 followed by quantitation using an IS-1000 Digital Imaging System from Alpha Innotech Corporation. All values are expressed relative to the ⌬vph1⌬stv1 strain expressing the wild-type VPH1 gene (WT).

DISCUSSION
The V-ATPases are evolutionarily related to the F-ATPases of mitochondria, chloroplasts, and bacteria (15)(16)(17)(18)(19)(20), both in overall structure (1,38,39) and sequence homology of several of the subunits (40 -44). For the F-ATPase of E. coli, proton translocation through the F 0 domain requires the participation of all three subunits: subunit a with molecular mass 30 kDa, subunit b with molecular mass 17 kDa, and subunit c with molecular mass 8 kDa (16, 24 -27, 45). The c subunit is composed of a hairpin loop of two transmembrane helices, with a buried glutamic acid (or aspartic acid) residue located in the middle of the second transmembrane helix (46). Both chemical modification with dicylohexylcarbodiimide (47,48) and mutagenesis studies (49,50) have shown that this buried carboxyl is essential for proton transport through F 0 . The b subunit, which contains a single membrane span and a cytoplasmic loop, is believed to be involved in attachment and coupling of the F 1 and F 0 domains (51,52). The a subunit, which contains 5-7 transmembrane helices (45,53), has also been shown by reconstitution studies (54) and site-directed mutagenesis (24 -27) to be essential for proton translocation through F 0 .
While the V-ATPases contain a 17-kDa c subunit that is homologous to the c subunit of F 0 (44), sequence analysis has revealed no obvious structural homolog to the a subunit in the V-ATPase complex. Based upon site-directed mutagenesis studies of the 100-kDa subunit in yeast, we proposed that the 100-kDa subunit may function as the a subunit homolog in the V-ATPase complex (23). In particular, mutation of charged residues in the last two transmembrane helices of the 100-kDa subunit (Glu-789 and His-743) led to significant decreases in proton transport and ATPase activity without appreciable effects on stability or assembly of the V-ATPase. To further elucidate the role of this COOH-terminal domain in activity and assembly of the V-ATPase complex, additional site-directed and random mutagenesis studies have been performed.
Surprisingly, we have found that a charged residue at position 789 in TM7 of the 100-kDa subunit is not absolutely required for ATP-dependent proton transport by the V-ATPase. Thus, the E789A mutant has 40% of wild-type levels of ACMA quenching, even higher than that observed for the E789Q mutant. Similarly, a charged residue at position 743 in TM6 is also not essential for proton transport. Nevertheless, both Glu-789 and His-743 appear to be located in a position to influence ATP-dependent proton transport activity. Thus, replacement of Glu-789 with Ala or Gln leads to a significant increase in the pH optimum of ATPase activity whereas the E789D mutant shows a lower pH optimum, consistent with the lower pK a of Asp relative to Glu. The H743A mutant also shows a higher pH optimum. While the pH optimum of an enzyme is a complex function of several different parameters, the observed changes suggest that these residues can influence V-ATPase activity.
One possible way in which these residues may influence activity is by participating in the access channels used to conduct protons to or away from the buried carboxyl groups of the c subunits. Previous studies on the a subunit of the E. coli F 0 have suggested a critical role for an acidic residue (Glu-219) and a positively charged residue (Arg-210) in the penultimate transmembrane helix and a positively charged residue (His-245) in the last transmembrane helix of the a subunit (24 -27). Based on these results, a role for these residues in proton conduction has been proposed, either as participants in a proton wire mechanism (55,24,25) or through direct interaction with the buried carboxyl group of subunit c (16,27,53). In the latter case, these residues have been suggested to alter the pK a of the c subunit carboxyl during the catalytic cycle of the enzyme, thereby facilitating its reversible protonation and deprotonation (16,53). Our results suggest that while Glu-789 and His-743 in the 100-kDa subunit may participate in proton translocation through V 0 , they cannot be absolutely essential to the proton conduction pathway, since otherwise substitution of a nonpolar amino acid such as Ala should completely block proton transport. Nevertheless, the observed changes in pH optimum of the mutants may reflect a change in the ability to protonate or deprotonate the c subunit carboxyl group at neutral pH.
It is important to realize that, unlike the case with the studies on F 0 , the present analysis only measures ATPase activity and ATP-dependent proton transport by the V-ATPase. It is thus possible that the mutations tested may effect some aspect of activity other than proton translocation, such as ATP hydrolysis or coupling of proton transport and ATPase activity. Attempts to measure passive proton translocation through the yeast V 0 domain using the previously reported acid treatment (56) have been unsuccessful, and the amounts of purified protein available from the yeast system do not permit the sort of dissociation and reassembly of V 0 employed for the bovine V-ATPase (57). In addition, mutations that led to uncoupling of proton transport from ATP hydrolysis would be predicted to inhibit transport much more than hydrolysis. No such uncoupled mutants have thus far been identified. Nevertheless, the location of Glu-789 and His-743 within putative membrane spanning segments and their effects on ATP-dependent proton transport are consistent with a role for these residues in proton translocation.
In addition to Glu-789 and His-743, random mutagenesis has identified several other residues whose mutation leads to significant loss of proton transport. These include His-729 in the polar loop preceeding TM6, Leu-739 and Leu-746 in TM6, and V803D in the polar COOH-terminal tail following TM7. In the case of H729R and L739S, the observed decrease in proton pumping may be attributable to the loss of the 14-kDa VMA7 gene product (subunit F), which has been shown to be essential for activity of the yeast V-ATPase complex (58). On the other hand, complete deletion of Vma7p leads to the absence of a properly assembled V 0 domain as well as the failure of the V 1 domain to assemble onto vacuolar membranes (58), which is not what is observed for the H729R and L739S mutants. Instead, it is possible that these mutations lead to a destabilization of the interaction of subunit F with the V-ATPase complex such that this protein more readily dissociates following assembly. If correct, these results suggest a possible direct interaction between subunit F and the 100-kDa subunit. For the remaining mutations affecting proton transport, the mechanism by which they inhibit activity is uncertain. It is possible that for the L746S mutation, the polar substitution for a hydrophobic residue has resulted in a partial rotation of TM6 such that residues participating in proton translocation are no longer optimally oriented. It is interesting in this regard that a number of mutations in nonpolar residues adjacent to Glu-219 in the E. coli a subunit result in optimization of the activity of a A24D/D61G double mutant in subunit c, leading to the suggestion that there is a direct hydrophobic interaction between the two helices of subunit c and the penultimate helix of subunit a (53).
In addition to mutations affecting proton transport or ATPase activity, random mutagenesis has identified a number of residues whose mutation leads to a significant decrease in assembly of the V-ATPase complex (Fig. 7). These include L780S in TM7 and a cluster of five mutations in the COOHterminal polar tail following TM7, including L800S, W802R, V803F, F809L, and G814D. This cluster of five mutations between residues 800 and 814 suggests that this region plays some critical role in assembly of the V-ATPase complex, possibly providing a scaffold for assembly of V 1 onto the V 0 domain. Interestingly, the last two mutations in this group (F809L and G814D) may have resulted in some mistargeting of the 100-kDa subunit, since significantly reduced levels of the 100-kDa subunit are observed in the vacuolar membrane despite the presence of normal levels of Vph1p in whole cell lysates (Fig. 3). Alternatively, these mutations may lead to an altered stability of the 100-kDa subunit such that it appears at normal levels in the whole cell lysate but becomes proteolyzed during vacuole isolation. It has previously been shown that the yeast 100-kDa subunit contains targeting information since Vph1p is targeted to the central vacuole while Stv1p is targeted to some other Mutations that led to aberrant assembly of the V-ATPase (shaded circles) include L780S, L800S, W802R, V803F, F809L, and G814D. The V803F mutation is included in this group because it leads to destabilization of the V-ATPase complex (Fig. 5). Two of these mutations (darkly shaded circles), F809L and G814D, may also cause mistargeting of the 100-kDa subunit. intracellular compartment, possibly endosomes (22). It is thus possible that mutations in the region 809 -814 are altering the correct targeting of the 100-kDa subunit. This region cannot itself correspond to the targeting signal since these residues are identical between Vph1p and Stv1p (21,22). Nevertheless, the mutations introduced may be disrupting the signal such that it cannot be correctly read. Additional experiments (particularly immunolocalization studies) will be required, however, to determine whether these mutants are actually mistargeted in yeast.
In summary, our results suggest that the COOH-terminal domain of the 100-kDa subunit plays an important role in both proton transport and assembly of the V-ATPase complex. Mutations affecting ATP-dependent proton transport have principally been localized to the transmembrane segments while mutations affecting assembly are present mainly in the COOHterminal tail following the last transmembrane helix. Further work will be required to elucidate how these residues interact with other regions of the 100-kDa subunit and how this domain interacts with other subunits in the V-ATPase complex.