Probing conserved regions of the cytoplasmic LOOP1 segment linking transmembrane segments 2 and 3 of the Saccharomyces cerevisiae plasma membrane H+-ATPase.

Genetic probing was used to examine conserved amino acid clusters in the first cytoplasmic loop domain (LOOP1) linking transmembrane segments 2 and 3 of the plasma membrane H+-ATPase from Saccharomyces cerevisiae. Deletion of the LOOP1 region in PMA1 resulted in a defective enzyme. Scanning alanine mutagenesis of conserved residues produced lethal cell phenotypes in 14 of 26 amino acids, suggesting major enzyme defects. Most viable mutants showed growth characteristics that were comparable to wild type. Two mutations, I183A and D185A, produced reduced growth rates, hygromycin B resistance, and low pH sensitivity, which are phenotypes associated with defects in the H+-ATPase. However, both mutant enzymes displayed near-normal kinetics for ATP hydrolysis in vitro. Localized random mutagenesis was also performed at sites Glu195, Val196, and Ile210, which all showed lethal phenotypes upon conversion to alanine. Amino acids with polar side groups could substitute for Glu195, while Val196 could not tolerate polar side group moieties. Nine mutations at Ile210 proved lethal, including K, R, E, P, H, N, V, G, and A, while functional enzyme was obtained with S, C, M, and L. Normal rates and extents of pH gradient formation were observed for all mutant enzymes, except I183A and D185A. Detailed analysis of the I183A enzyme indicated that it hydrolyzed ATP like wild type, but it appeared to inefficiently couple ATP hydrolysis to proton transport. In total, these results affirm that conserved amino acids in LOOP1 are important to H+-ATPase function, and purturbations in this region can alter the efficiency of energy coupling.

The plasma membrane H ϩ -ATPase from Saccharomyces cerevisiae is an electrogenic proton pump that couples ATP hydrolysis to proton transport. It plays important roles in both intracellular pH regulation and the maintenance of a large electrochemical proton gradient necessary for nutrient uptake. The H ϩ -ATPase is a member of the P-type family of ion translocating enzymes, and it belongs to the P 2 -subclass (P 2 -ATPase) of enzymes that are involved in the transport of nonheavy metals (1). The H ϩ -ATPase is comprised of a membrane transport domain, which is believed to consist of 10 transmembrane segments, and a large catalytic ATP hydrolysis and transduction domain. Typical of P-type ATPases, the H ϩ -ATPase couples energy from the cytoplasmic ATP hydrolysis domain to the membrane-embedded proton transport domain. The structural basis for this coupling is poorly defined, but it is believed that long range interactions are essential to this process (2). Clues to the nature of long range conformational interactions underlying coupling have come from a variety of experimental approaches, including drug interaction (3,4), immunological probing (5), limited proteolysis (6), and genetic probing (3,(7)(8)(9). Despite these studies, a consensus structural model for coupling has not yet been elucidated. There is emerging evidence that transmembrane segments 5 and 6, and perhaps 4 and 8, may participate in the transport of ions across the bilayer (1,3,7,10). How these segments are organized with the remaining transmembrane segments to participate in ion transport is not known. It is also not clear whether coupling involves a direct interaction of protein structure elements extending from transmembrane segments 5 and 6 into the central catalytic region or whether indirect interactions, involving other cytoplasmic and membrane-associated protein structure elements, are important.
In the yeast H ϩ -ATPase, mutations in the first two transmembrane segments (M1 and M2) and the central catalytic region appear to alter the electrogenic character of the enzyme (11)(12)(13). Genetic evidence suggesting an indirect interaction between M1 and M2 and the central catalytic domain was obtained from suppressor studies, in which the phenotype produced by a primary site mutation in one sector was overcome by a secondary site mutation in the other (8,9). These studies suggest that interactions between M1 and M2 and the catalytic region are important for normal ATP-linked electrogenic proton transport.
A likely candidate to mediate these long range interactions is the cytoplasmic loop domain (LOOP1) linking M2 and M3. The LOOP1 region was originally postulated to act as a transduction domain (14), based largely on the observation that Ca 2ϩ transport appeared uncoupled from ATP hydrolysis in the Ca 2ϩ -ATPase following tryptic hydrolysis of Arg 198 (15). In contrast, proteolytic (16 -19) and genetic studies (20,21) of residues in this region suggested that perturbed enzymes are fully coupled. While the role of this region in coupling is uncertain, it has been shown to undergo distinct conformational changes during catalysis that are linked to cation binding in the Na ϩ ,K ϩ -ATPase and Ca 2ϩ -ATPase (6,22). Finally, mutations S234A and D226N in the LOOP1 region of the yeast H ϩ -ATPase were suggested to reduce the normal stoichiometry of protons transported to ATP (23), and a recent report showed that a H285Q mutation caused partial uncoupling (19). These latter results suggest that perturbations in the LOOP1 region of the H ϩ -ATPase may affect coupling.
In this study, we systematically probed by mutagenesisconserved regions of LOOP1 to investigate the potential role of this segment in energy coupling by the yeast H ϩ -ATPase.
Growth Phenotypes of PMA1 Mutants-Overnight cultures (18 h) of pma1 mutants (0.4 ml) were inoculated into 30 ml of YPD medium and grown with shaking at both 24 and 37°C. Aliquots (0.2 ml) were removed hourly from each culture to determine optical density at 590 nm. A microplate-based assay was used to examine the pH sensitivity and hygromycin B resistance of mutants by evaluating cell growth in YDP media with pH ranging from 2.25 to 5.5 or in YDP media containing 0 -200 g/ml hygromycin B, as described previously (13).
Site-directed Mutagenesis-Site-directed pma1 mutants were constructed as described previously (9). All mutants were prepared in phagemid vector pGW201, which consists of vector pGEM-3zf subcloned with a 6.1-kilobase HindIII fragment containing PMA1 marked with URA3 at the 3Ј non-coding end (8). The PMA1 gene was modified by a C 3 T change in the coding sequence at nucleotide position 1344 (residue Phe 448 ), which did not alter the codon for Phe 448 but resulted in loss of an EcoRI site at this position. A second silent amino acid change was made at L294 to introduce an unique SpeI site by converting T 3 C at coding position 892 and G 3 A at position 894. The entire 6.1kilobase region from pGW201 was excised and transplaced into yeast strain SH122 by the lithium acetate transformation procedure (27). All vectors containing reconstituted pma1 genes were sequenced prior to transplacement into yeast to confirm the primary site mutation and to eliminate potential secondary mutations in the target region. Isogenic pma1 mutants were prepared as described previously (8). Two basic tetrad growth patterns were observed: 4:0 (4 spores viable) and 2:2 (2 viable, 2 dead spores). A 2:2 pattern indicated a recessive lethal mutation, while a 4:0 pattern suggested either a fully functional PMA1 or a gene conversion due to lethality of the mutation in question (28). To distinguish between these possibilities, the genomic DNA of these mutants was sequenced. In all cases, pma1 mutations were confirmed by DNA sequence analysis of polymerase chain reaction-amplified chromosomal DNA from the meiotic segregants (13).
Construction of LOOP1 Deletion Mutant-Vector pGW201 was double digested at unique sites with restriction enzymes SpeI and EcoRI, which removed a 0.4-kilobase fragment encoding amino acids Glu 162 to Leu 298 . A double-stranded oligonucleotide linker consisting of the sequence GGAATTCTTAAGCAGACTAGTC (and its compliment) was digested with SpeI and EcoRI, and then ligated to equivalent sites in the deleted region in pGW201 to form vector pMP401. The linker region carried a new AflII restriction site that aided in the cloning process. The newly constructed vector, which was deleted for the LOOP1 region, contained a short polar linker consisting of Glu 162 , Phe 163 , Leu 164 , Ser 165 , and Arg 166 . The next residue, Leu 167 , is equivalent to Leu 298 in the wild type sequence. A 5.7-kilobase HindIII fragment from pMP401 was used to replace the wild type PMA1 gene by homologous recombination (8). The cellular phenotype associated with this deletion was assessed by evaluating growth of meiotic segregants on YPD plates.
Membrane Preparation and ATPase Purification-Cells grown to mid-log phase (A 590 nm ϳ 5) were resuspended at 3 g/ml in homogenization buffer consisting of 100 mM Tris-HCl, pH 7.5, 5 mM EDTA, 5 mg/ml bovine serum albumin, and 1 mM phenylmethylsulfonyl fluoride. The cells were passed through a French pressure cell at 20,000 p.s.i. at 4°C. The lysate was rapidly adjusted to pH 7.0 with 1 M Tris base, and centrifuged at 10,000 ϫ g for 10 min. The resulting supernatant was centrifuged at 100,000 ϫ g for 30 min, and the pellet was resuspended at 2 mg/ml in extraction buffer consisting of 10 mM HEPES-KOH, pH 7.0, 100 mM KCl, 0.2 mM EDTA, 1 mM dithiothreitol, and 0.45% (w/v) glycerol. Deoxycholate (10% (w/v) stock, pH 7.5) was added dropwise to the suspension, which was stirred gently on ice, to a final concentration of 0.5% (w/v). The deoxycholate-extracted suspension was centrifuged at 150,000 ϫ g for 1 h. The pellet was washed by resuspension in extraction buffer (5 ml/mg protein) and centrifuged, as above. The final pellet was resuspended at 2.5 mg/ml in extraction buffer.
ATP Hydrolysis-ATPase assays were conducted in 96-well microplates, essentially as described by Monk et al. (29). A 125-l assay mixture contained 10 mM MES 1 -Tris, pH 6.5, 5 mM MgSO 4 , 25 mM NH 4 Cl, 5 mM ATP, and 0.5-1 g membrane protein. Samples were incubated at 30°C and inorganic phosphate released from ATP was determined by the addition of 125 l of phosphate developing reagent (29). The absorbance at 600 nm (A 600 ) was determined after a 15-min incubation at 22°C. All K m and V max values were obtained by determining ATP hydrolysis as a function of substrate concentration (0 -10 mM for both ATP and MgSO 4 ), and the data were fit to the Michaelis-Menten equation. Vanadate sensitivity was determined by measuring ATP hydrolysis in standard ATPase buffer containing sodium vanadate at 0 -100 M.
Measurement of ATP-induced pH Gradient Formation-Deoxycholate-enriched H ϩ -ATPase was reconstituted into asolectin-containing liposomes, as described previously (11). ATP-induced proton transport in reconstituted vesicles was monitored by the acridine orange fluorescence-quenching assay (30). Liposomes containing 50 g of reconstituted enzyme were suspended in 1.0 ml of assay buffer containing 10 mM MES-Tris, pH 7.0, 50 mM KCl and 5 mM ATP in a 2.7-ml stirred cuvette. The reaction was initiated by the addition of 5 mM MgSO 4 . The quenching of acridine orange fluorescence was monitored with a LS5-B spectrofluorometer (Perkin-Elmer Corp.) with excitation and emission wavelengths of 420 and 550 nm, respectively.
Isolation of Mutant H ϩ -ATPase-containing Secretory Vesicles-pma1 mutants that yield distinct hygromycin B resistant phenotypes were introduced into yeast strain NY17 sec6-4, which produces secretory vesicles at elevated temperature (25,26), by a one-step gene replacement procedure. A 6.1-kilobase HindIII fragment from pGW201 containing pma1 marked with URA3 was transplaced into yeast, as described previously (9). Transformants growing in the absence of uracil (Ura ϩ colonies) were further selected for growth resistance on YPD plates containing 50 -250 g/ml hygromycin B. Hygromycin B-resistant colonies were isolated, and the PMA1 region was amplified by polymerase chain reaction. DNA sequence analysis was used to confirm that the mutations were integrated into the genome. Secretory vesicles were prepared by a modified version of the procedure described by Rao and Slayman (31). Parental strain NY17 and isogenic pma1 mutant strains were grown in YPD at 22°C until mid-log phase (A 590 ϳ 5). The temperature was increased to 37°C for 2 h, and the cells were harvested by centrifugation at 3,000 ϫ g for 20 min. The cell pellet was resuspended in homogenization buffer consisting of 50 mM Tris-HCl, pH 7.5, 1.4 M sorbitol, 5 mM EDTA, 1 mM EGTA, 5 mg/ml bovine serum albumin, 2 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, and 2.5 g/ml chymostatin. The cells were broken in a French press at 20,000 p.s.i., and the lysate was centrifuged at 10,000 ϫ g for 10 min. Secretory vesicles in the supernatant were centrifuged at 100,000 ϫ g for 60 min. The pellet was resuspended at 5 mg/ml in 10 mM MES-Tris, pH 7.0, containing 0.8 M sorbitol (31), and used directly for evaluation ATP-induced pH gradient formation. Proton-transport in secretory vesicles was determined by the same procedure used with reconstituted vesicles. In some experiments, proton-transport and ATP hydrolysis were monitored simultaneously. To monitor ATP hydrolysis under these conditions, a 40-l aliquot was removed at 30-s intervals from the assay mixture, and mixed with 200 l of phosphate colorimetric reagent (32) to measure phosphate released. Coupling efficiency was assessed by comparing the initial rate of proton pumping (percent fluorescence quenching per min) with the rate of ATP hydrolyzed per min.
Other Procedures-Protein concentrations were determined by modification of the Lowry method (33). SDS-gel eletrophoresis of plasma membrane proteins in 10% precast minigels (Bio-Rad) was performed according to the manufacturer's instructions. DNA sequencing was performed with the Sequenase DNA sequencing kit (U. S. Biochemical Corp.).

RESULTS
Deleting the LOOP1 Region-The LOOP 1 region appears similar in all the P-type ATPases (34). As a first step toward understanding the role of LOOP1 in H ϩ -ATPase function, the region extending from the putative bilayer/cytoplasm boundary of transmembrane segment 2 to the same boundary point in transmembrane segment 3, Glu 162 -Leu 298 , was deleted. However, because LOOP1 demarcates the bilayer boundaries for transmembrane segments 2 and 3, it was rationalized that linking the transmembrane segments directly would be too destabilizing to the enzyme. Therefore, the LOOP1 domain was replaced with a short polar linker region consisting of Glu 162 , Phe 163 , Leu 164 , Ser 165 , Arg 166 , and Leu 167 . (Leu 167 corresponds to Leu 298 in the wild type enzyme.) The deletion construct was introduced into yeast and found to produce a recessive lethal growth phenotype indicative of a defective H ϩ -ATPase. The fact that the deletion construct produced a recessive lethal phenotype, rather than a dominant lethal phenotype resulting in gene conversion, may suggest that the mutant enzyme structure was not grossly perturbed. Nonetheless, these results demonstrate that deletion of the LOOP1 region results in a defective enzyme.
Alanine Mutagenesis of Conserved Residues in LOOP1-Several highly conserved stretches of amino acids have been identified in the LOOP1 region among the various P-type ATPases (35). These stretches in yeast PMA1 include Val 182 -Gln 188 , Glu 195 -Ile 201 , Gly 207 -Asp 213 , Ile 225 -Leu 235 , and Met 258 -Gly 270 . Scanning alanine mutagenesis was used to probe residues, in four of the five conserved regions found in LOOP1, by converting all amino acids to alanine. The Met 258 -Gly 270 region was not examined because this region contains the lowest relative percentage of conserved residues (35), and specific residues have been examined in other studies (8,36,37). Conserved amino acid clusters or regions are usually considered to be functionally important, and by probing these regions, it is more likely that interesting mutations would be isolated. Alanine substitution was used because it generally causes fewer structural perturbations, and subtly perturbed mutant enzymes were desirable due to the lethality of more severe pma1 mutations in yeast. (In one case, the naturally occurring alanine at position 229 was converted to serine.) The effects of each mutation on enzyme function, as reflected in cell viability, are shown in Table I. It was found that 54% (14/26) of the residues, when modified to alanine, resulted in sufficient enzyme perturbation to prevent cell growth (lethality). The lethal mutations were widely distributed, although clusters did appear such as Glu 195 -Pro 198 and Thr 231 -Glu 233 . Lethal mutations were frequently observed with prominent changes in side group character, such as Glu or Arg to Ala. However, they were also observed with more subtle changes, such as Val or Gly to Ala.
Characterization of Growth and Enzymatic Properties of Viable pma1 Mutants-Viable pma1 mutants from the scanning alanine mutagenesis were characterized for cell growth at 24 and 37°C. All of the pma1 mutants at 24°C showed growth properties that were comparable to that of wild type (Table I).
In contrast, two mutants, I183A and D185A, showed significantly reduced growth rates at the elevated temperature of 37°C (Fig. 1). Viable mutants were screened for growth resistance to hygromycin B, which has been correlated with the ability of the H ϩ -ATPase to sustain a hyperpolarized membrane potential (39). As previously observed, wild type yeast are highly sensitive to hygromycin B (24), while pma1 mutants show a wide range of sensitivities. Differential sensitivity was also observed for the LOOP1 mutants, as indicated by the amount of hygromycin B required to produce 50% growth inhibition in rich medium (IC 50 ) ( Table I). A majority of the pma1 mutants were found to retain strong sensitivity to hygromycin B, while five mutants I183A, G207A, V209A, S228A, and I230A were three to five times more resistant to the antibiotic. (This shift in sensitivity correlates exactly with plate assays (24) previously used to score hygromycin B resistance. 2 ) Hygromycin B resistance was also observed at 37°C, and in the case of I183A, was somewhat amplified (Table I). The pma1 mutants were also grown as a function of acidic medium pH, which provides a relative measure of the ability of the H ϩ -ATPase to regulate intracellular pH. Fig. 2 shows that the growth of wild type cells was relatively insensitive to acidic medium above pH 2.3. In contrast to wild type, pma1 mutants, I183A and D185A, showed prominent growth sensitivities below pH 2.9 at both 24°C (not shown) and 37°C. pma1 mutant S228A also showed acid sensitivity but only at elevated temperature. These data indicate that several LOOP1 mutants show typical phenotypic growth abnormalities that are associated with defects in the H ϩ -ATPase (24).
Purified plasma membranes from the viable pma1 mutants were used to assess the biochemical properties of mutant enzymes. The relative amount of H ϩ -ATPase in the membrane was assessed by examining the amount of the characteristic molecular mass ϳ100-kDa polypeptide band by SDS-gel electrophoresis. It was determined from this analysis that the level of assembled H ϩ -ATPase for all the scanning alanine mutants was within 20% of that of wild type, when evaluated as a percent of the total plasma membrane protein. Most of the mutant enzymes showed K m values within 25% of wild type, while three mutant enzymes, V196M, V209A, and S228A, showed values about 40% of wild type (Table II). Fifty percent of the mutant enzymes showed V max levels that were 50 -60% of wild type, including enzymes with modifications in the final cluster (I225A, S228A, A229S, and I230A). These results were qualitatively similar at 37°C, although the values for V max were two to three times greater. There was a weak correlation 2 D. Seto-Young, unpublished results.  V  158  120  56  42  I230A  V  174  124  102  64  T231A  DN  G232A RL E233A RL a Viability was determined from the growth of dissected spores on YPD medium. Two viable spores and two dead spores indicated recessive lethality (RL), while four viable spores indicated either viability of a pma1 mutant (V) or a dominant negative mutation (DN) resulting from gene conversion (8,38), which was confirmed following DNA sequencing of the chromosomal pma1 gene. b Average doubling time (min) determined from duplicate growth curves. Standard error of the mean was less than 15% for each growth set.
c Hygromycin B concentration (g/ml) required to inhibit growth 50% (IC 50 ) in YPD growth medium.
between the level of ATP hydrolysis under V max conditions and growth phenotypes, as has been observed previously (11,13). In particular, the I183A and D185A mutants, which show prominent growth phenotypes displayed normal hydrolytic properties. All of the mutant enzymes showed K i values for vanadate that ranged from 1.0 to 2.5 M, which are comparable to wild type (not shown). There were no enzymes showing significant vanadate insensitivity (Ͼ100-fold), which has been observed with other pma1 mutant enzymes (40). These results suggest that the mutant enzymes were only mildly perturbed in ATP hydrolysis by the introduction of alanine at the various positions.
Localized Random Mutagenesis of Glu 195 , Glu 196 , and Ile 210 -Localized random mutagenesis was used to introduce diverse mutations at defined sites. Ile 210 is located in a sequence region that is largely tolerant of mutation (Table I). However, a I210A mutation is lethal, and it was of interest to determine whether a more subtle perturbation of this residue might be tolerated. Random mutagenesis resulted in 12 different amino acid substitutions at position 210. Nine mutations were lethal, including K, R, E, P, H, N, V, G, and A. Functional enzyme was obtained with S, C, M, and L substitutions. Cys 210 and Leu 210 mutant enzymes supported relatively normal rates of growth (Fig. 1) and sensitivities to hygromycin B (Fig. 2). In contrast, the mutants carrying I210M and I210S grew slower (Fig. 1), were resistant to hygromycin B (Table III), and were sensitive to acid loading (Fig. 2). The growth properties of the I210M-and I210S-containing mutants were consistent with a decreased V max (Ͻ40% of wild type) for these enzymes in vitro, relative to wild type and other mutant enzymes at position 210 (Table II). These results suggest that nonpolar residues of similar hydrated volume, Ile, Leu, and Met or small, relatively polar residues, Ser and Cys, may fill a structural role at this position. Clearly, the conserved nature of Ile 210 does not equate with it being essential.
Saturation mutagenesis of amino acid residues Glu 195 and Val 196 was performed because these residues define the begin-ning of a stretch where four consecutive alanine replacements, Glu 195 -Pro 198 , were lethal, indicating a region sensitive to perturbation. Random mutations were introduced with a oligonucleotide primer, in which the 6 nucleotide bases in codons 195 and 196 were represented by an equal distribution of the four dNTPs. Phagemid from 92 transformed colonies were sequenced, and 20 different substitutions were recovered, either as single (positions 195 or 196) or double amino acid substitutions (positions 195 and 196). The mutant pma1 genes were introduced into yeast resulting in the growth phenotypes listed in Table III. The charge and size character of the Glu 195 side group appears important because hydrophobic substitutions were lethal, while the strongly polar but smaller Asp could substitute, along with the smaller but weakly polar Cys. Gly could also substitute, which may suggest some steric constraints in this region. On the other hand, Val 196 could not tolerate a polar change or the small side chain of Gly, while Met could substitute. Ala could substitute at position 196 provided that Cys was present at Glu 195 . However, the double mutant, E195C,V196A, grew poorly and displayed hygromycin B resistance (Table III). These enzymes displayed about 60% of the wild type level of activity and showed normal K m values (Table  II); the double mutant showed less than 50% of the normal enzyme amount in the membrane (not shown). These results suggest that neither Glu 195 nor Val 196 are essential residues, but their side group character may play an important structural role.
Assessing Proton Transport by Mutant Enzymes-The proton transport properties of pma1 mutant enzymes were assessed by reconstituting the enzymes in asolectin liposomes and following ATP-induced pH gradient formation by the acridine orange fluorescence quenching assay. All reconstituted enzyme preparations were compared by examining the rate and extent of ATP-induced pH gradient formation at a fixed level of total activity (ATP hydrolysis units) in the assay. A survey of the mutant enzymes indicated that all but two, I183A and D185A, showed proton transport characteristics that were nearly identical to wild type (Fig. 3A). The mutant enzymes D185A and I183A consistently showed less pH gradient formation relative to wild type. The ATPase activity of reconstituted vesicles was confirmed before and after the proton pumping assay, and SDS-gel electrophoresis was used to confirm that the enzyme remained intact following reconstitution (Fig. 3B).
To rule out the possibility that the diminished proton transport observed for the I183A mutant enzyme was due to alterations occurring during detergent extraction and reconstitution, the I183A mutant was evaluated in naturally occurring secretory (Sec) vesicles. This was accomplished by transplacing the I183A pma1 mutant gene into strain NY17, which contains the temperature-sensitive Sec6-4 mutation that results in accumulation of Sec vesicles at 37°C (25). The new strain contained only the pma1-I183A gene and induction of Sec vesicles at 37°C resulted in the incorporation of the mutant enzyme into the vesicles. It was not possible to introduce the D185A mutation because it did not confer significant hygromycin B resistance (Table I) suitable for selection in NY17 cells. ATPmediated proton transport in the Sec vesicles was assessed in the same way as for the reconstituted liposomes. Fig. 4A shows the pH gradient profiles for wild type and the I183A mutant enzymes in the presence of increasing amounts of vanadate to partially inhibit enzyme activity. Wild type and mutant enzyme activities were adjusted to produce nearly equivalent rates of ATP hydrolysis in each assay, as illustrated in Fig. 4B. Proton transport was observed to differ between the mutant and wild type enzymes at equivalent rates of ATP hydrolysis, which was consistent with the reconstitution data. The most significant differences were observed at the highest rates of ATP hydrolysis (Fig. 4B, curves c, d, g, and h). It can be seen that the extent of ATP-induced proton transport by the I183A mutant enzyme in the Sec vesicles was lower than that of wild type. A semiquantitative assessment of proton transport by the two enzymes determined by dividing the initial rate of proton transport (fluorescence quenching) by the rate of ATP hydrolysis under conditions when the enzymes were turning over at approximately 70% of capacity indicated that the mutant enzyme was approximately 45 Ϯ 5% less efficient than the wild type enzyme in ATP-mediated proton pumping. These results, along with the whole-cell growth data, are consistent with the notion that the I183A mutant enzyme couples protons less efficiently than the wild type enzyme.

DISCUSSION
Mutagenesis of 26 conserved residues in the LOOP1 domain revealed that conversion to alanine in approximately one-half the residues (14) induced significant enzyme perturbation, resulting in a lethal cell phenotype (Table I). Fig. 5 summarizes the scanning alanine mutagenesis growth data reported in this study, and illustrates that only eight residues could be changed to alanine without a major affect on the H ϩ -ATPase. The remaining four viable mutants displayed typical defects associated with a functionally perturbed H ϩ -ATPase, including hygromycin B resistance and low pH sensitivity. Many of the lethal mutations occurred in a region predicted to form ␤-strand structure (41) (Fig. 5), although a cluster of lethal and viable hygromycin B-resistant mutants occurred in a transition region between the predicted ␣-helical stalk segment and the ␤-strand sector. It is possible that the lethal mutations generated in this study were each sufficient to cause a structural perturbation that led to either an assembly-defective enzyme or an assembled but catalytically deficient enzyme. In this regard, it has been widely suggested that the LOOP1 region interacts with the large central cytoplasmic domain, LOOP2, and this interaction is important for catalysis. Mutations in the LOOP1 domain are known to influence vanadate sensitivity, phosphate binding, and the rate of phosphorylated intermediate turnover (8, 37, 42, 43). Protein-protein interactions often occur at the interfaces of conserved residues, and the clusters of  mutations in conserved regions producing lethal phenotypes in this study may be consistent with such interacting regions. LOOP1 would be expected to interact with LOOP2 through local interactions of individual amino acid residues between structural domains. Such interactions could help explain why many of the individual amino acid substitutions in portions of the conserved region of LOOP1 failed to significantly inhibit ATPase activity. In addition, these interactions help account for the saturation mutagenesis data obtained at positions Glu 195 , Val 196 , and Ile 210 , in which both viable and lethal amino acid substitutions were identified (Table III). If a specific side group were essential, then all substitutions would be lethal. However, if the targeted residue was part of a local interacting domain, then more diverse side groups within a given molecular size or charge family could be accommodated, as was observed. Overall, the subtle effects of viable LOOP1 mutations in highly conserved regions suggest that these regions play a minimal role in catalysis and are more likely to be of structural importance. Subtle mutations should have a minimal effect on the structural properties of LOOP1, while more perturbing mutations should induce both local and global shape changes, which could lead to a severe catalytic deficiency or lack of enzyme assembly. If LOOP1 contributes to the active center, then local structural changes could influence catalysis. It is well established that mutations in LOOP1 strongly influence the sensitivity of the yeast H ϩ -ATPase to vanadate (8,37,44), as well as phosphate interactions (37). These mutations could directly alter the phosphorylation domain or they could alter the distribution of conformational intermediates leading to vanadate insensitivity, as previously proposed (8,16,36,44). The effects of these mutations on vanadate sensitivity, the reduced catalytic activity observed for a factor Xa-engineered mutant (16), and some of the I210 mutations in this study (Table II), are consistent with the effects of LOOP1 mutations on the Ca 2ϩ -ATPase (20,21). In the Ca 2ϩ -ATPase, mutations in the LOOP1 ␤-strand domain are believed to alter the progression of catalysis by blocking the interconversion of catalytic intermediates during the E 1 P-E 2 P transition (20,21). Such perturbations in LOOP1 are believed to indirectly affect catalytic turnover. Thus, deleterious mutations in LOOP1 could alter the progression of catalysis by indirectly blocking the interconversion of catalytic intermediates. Definitive structural changes in LOOP1 appear to be an important part of the overall catalytic cycle for the Na ϩ ,K ϩ -ATPase (6). Finally, it is known that formation of E 1 P exposes a tryptic cleavage site on LOOP1 in the Ca 2ϩ -ATPase, whereas the site is largely protected in E 2 P (18).
It is not clear whether the structural features contributed by LOOP1 play a direct role in ATP hydrolysis by the enzyme. Nonetheless, perturbations in LOOP1 can indirectly influence catalysis, and this may be relevant to coupling. A survey of proton transport by the viable LOOP1 pma1 mutant enzymes showed that, in nearly all cases, there was no significant difference in the proton transport properties between wild type and mutant enzymes. This finding suggests that perturbations FIG. 3. pH gradient formation by reconstituted mutant enzymes. ATP-induced pH gradient formation by purified and reconstituted pma1 mutant enzymes, as indicated, was assessed by the acridine orange fluorescence quenching assay. A, representative traces showing pH gradient formation by mutant enzymes. All assays were performed in duplicate with wild type enzyme serving as the control enzyme. B, representative SDS-gel electrophoresis profiles of reconstituted enzymes following proton transport assays, as shown in panel A. Approximately 5 g of reconstituted protein were analyzed per lane.  I210A  RL  I210S  V  219  235  74  77  I210C  V  131  114  24  30  I210M  V  177  266  80  99  I210L  V  150  100  46  40  I210K  RL  I210G  DN  I210R  RL  I210V  DN  I210N  DN  I210E  DN  I210H  DN  I210P  DN   a See legend to Table I. in LOOP1 do not generally alter proton transport. Interestingly, several of the LOOP1 mutations, which failed to elicit significant differences in in vitro proton transport or the kinetics for ATP hydrolysis, displayed prominent growth phenotypes such as temperature sensitive growth, resistance to hygromycin B, and sensitivity to low external pH suggesting potential defects in electrogenic proton transport by the H ϩ -ATPase under stress conditions. It is apparent from the relationship between H ϩ -ATPase function and cellular physiology that subtle perturbations in the enzyme can have pronounced affects on phenotype. This likely reflects the fact that the H ϩ -ATPase in the cell is not operating at maximal capacity due to the influence of numerous constraining factors such as membrane potential, pH gradient, turgor pressure, and other regulatory phenomena. Thus, small perturbations in enzyme function may be amplified at the cellular level. In contrast, mutant pma1 enzymes analyzed in vitro do not have the same physiological constraints placed upon them, and an analysis of kinetic capacity under maximum turnover conditions (V max ) may not reveal significant differences. However, two mutations, I183A and D185A, were notable because they induced in vitro and in vivo properties suggestive of coupling defects. Both mutations induced temperature sensitive growth properties at 37°C that resulted in diminished rates of cell growth (Table I, Fig. 1), sensitivity to low external pH (Fig. 2), and resistance to hygromycin B (Table I). When reconstituted in liposomes, both mutant enzymes appeared to pump protons less efficiently than wild type (Fig. 3A). This deficiency was not a reconstitution artifact because a more detailed analysis of proton transport by the I183A mutant enzyme in naturally occurring Sec vesicles confirmed the reconstitution study (Fig. 4). These mutations are predicted to lie near the interface between where the ␣-helical stalk region ends and the ␤-strand region begins. It remains to be determined whether this region directly or indirectly impacts the coupling process. The three-dimensional organization of transmembrane segments could bring M2 within close proximity to M4 and M5, which have been implicated in ion binding (1,7). Our findings are fully consistent with the previous assertion of Serrano and Portillo, who suggested that mutations at D226, E233, and S234 in LOOP1 diminish the efficiency of coupling in the H ϩ -ATPase (23,43). It should be noted that D226A and FIG. 4. pH gradient formation and ATP hydrolysis of the I183A mutant enzyme in Sec vesicles. Pma1 containing the I183A mutation was transplaced into strain NY17 by homologous recombination (8), and a homozygous mutant strain was selected on hygroymcin B-containing medium. Secretory vesicles containing the I183A mutant enzyme were collected following induction at 37°C. A, ATP-induced pH gradient formation by wild type (NY17) and mutant (I183A) enzymes was assessed by following the quenching of acridine orange fluorescence. The rate and extent of pH gradient formation was reduced by titrating the enzymes with increasing amounts of vanadate from 0 to 10 M. B, the rate of ATP hydrolysis by reconstituted enzymes under vanadate-limiting conditions, shown in panel A, was determined by simultaneously measuring the extent of phosphate liberated from ATP as a function of time by assaying small aliquots (40 l) of the reaction mixture. The approximate adjusted rate of ATP hydrolysis for both enzymes at each reaction condition was: 0.75 mol of P i /min (d and h), 0.55 mol of P i /min (c and g), 0.26 mol of P i /min (b and f), and 0.1 mol of P i /min (a and e). E234A were found to be lethal in this study (Table I). In addition, a recent report by Wach et al. (19) that a H285Q mutation induces partial uncoupling is also relevant. This mutation is predicted to reside on the ␣-helical stalk segment emerging from transmembrane segment 3, but it could be in close proximity to I183A, which lies in the ␣-helical stalk segment emerging from transmembrane segment 2. The likely close proximity of these two helical elements (6) and the partial uncoupling effects of local mutations suggests a role for local structure of the stalk region in coupling. Yet, it cannot be ascertained whether the partial uncoupling observed is a consequence of disrupting the physical mechanism of coupling or whether it reflects an alteration in the transition between E 1 P and E 2 P intermediate states, as has been previously proposed (19). The fact that mutations in both the Ile 170 and His 285 region induce vanadate insensitivity (8,16) may be more consistent with the latter proposal. Although, none of the mutations in this study altered vanadate sensitivity.
Collectively, the data in this study suggest that changes in LOOP1 can influence energy coupling. Unfortunately, it cannot yet be determined whether this involves a direct or indirect interaction with the central catalytic loop domain, LOOP2, which is involved in nucleotide binding and the phosphorylation-dephosphorylation of the critical aspartate residue.