Stalk Segment 4 of the Yeast Plasma Membrane H+-ATPase

In the P2-type ATPases, there is growing evidence that four α-helical stalk segments connect the cytoplasmic part of the molecule, responsible for ATP binding and hydrolysis, to the membrane-embedded part that mediates cation transport. The present study has focused on stalk segment 4, which displays a significant degree of sequence conservation among P2-ATPases. When site-directed mutants in this region of the yeast plasma membrane H+-ATPase were constructed and expressed in secretory vesicles, more than half of the amino acid substitutions led to a severalfold decrease in the rate of ATP hydrolysis, although they had little or no effect on the coupling between hydrolysis and transport. Strikingly, mutant ATPases bearing single substitutions of 13 consecutive residues from Ile-359 through Gly-371 were highly resistant to inorganic orthovanadate, with IC50 values at least 10-fold above those seen in the wild-type enzyme. Most of the same mutants also displayed a significant reduction in the K m for MgATP and an increase in the pH optimum for ATP hydrolysis. Taken together, these changes in kinetic behavior point to a shift in equilibrium from the E2 conformation of the ATPase toward the E1conformation. The residues from Ile-359 through Gly-371 would occupy three full turns of an α-helix, suggesting that this portion of stalk segment 4 may provide a conformationally active link between catalytic sites in the cytoplasm and cation-binding sites in the membrane.

P-type ATPases, which are found throughout prokaryotic and eukaryotic cells, use the energy from ATP hydrolysis to pump inorganic cations across cell membranes. In the most abundant subfamily, designated P 2 , the 100-kDa ATPase polypeptide is embedded in the lipid bilayer by four hydrophobic segments at the N-terminal end of the molecule and six at the C-terminal end (1). The central hydrophilic region protrudes into the cytoplasm and contains catalytic sites for ATP binding and formation of the characteristic ␤-aspartyl phosphate intermediate.
It was proposed more than a decade ago that conformational changes in the catalytic portion of the P-type ATPases are transmitted into the membrane via a stalk-like region made up of the cytoplasmic extensions from some, but not necessarily all, of the transmembrane segments (2). Recently, the stalk has been directly visualized in cryoelectron microscopic studies of the sarcoplasmic reticulum Ca 2ϩ -ATPase (3) and the plasma membrane H ϩ -ATPase of Neurospora crassa (4), both at 8-Å resolution. In the case of the Ca 2ϩ -ATPase, the stalk appeared as a narrow structure connecting the compact, wedge-shaped cytoplasmic domain with the membrane-spanning segments; it contained four rod-like densities (very likely ␣-helices), which were tentatively identified as stalk segments 2-5 (3). The Neurospora H ϩ -ATPase was strikingly similar to the Ca 2ϩ -ATPase in its membrane region, but the cytoplasmic portion was noticeably less compact, and instead of a single stalk, there were several apparent connections between the cytoplasmic portion and the membrane (4). As the authors pointed out (3)(4)(5), the two structures may represent different conformational states, since the Ca 2ϩ -ATPase was crystallized in the presence of decavanadate and the H ϩ -ATPase, in the absence of added ligands. Indeed, there is compelling evidence that the reaction cycle of the P-ATPases alternates between two major conformations (E 1 and E 2 ), that are different enough to be distinguished from one another by a variety of biochemical and biophysical methods (6) including proteolytic digestion patterns (7)(8)(9).
Recently, Soteropoulos and Perlin (10) carried out an informative mutagenesis study of stalk segments 2 and 3 in the yeast plasma membrane H ϩ -ATPase, a close relative of the Neurospora enzyme. Their approach was to replace selected amino acids in S2 (Ile-183 and Gly-186) and S3 (Gly-270 and Thr-287) with helix-breaking residues, Gly and Pro, in order to test the helical nature of S2 and S3. At both positions in S2, the mutations proved to be lethal; at the positions in S3, the cells were viable, but the Pro replacements led to a significant reduction in ATPase activity. The authors concluded that the ␣-helical nature of S2, and to a lesser extent that of S3, may help to stabilize the stalk and/or promote the proper conformational interaction between the cytoplasmic and membrane-embedded portions of the ATPase.
In the present study, we have performed alanine-scanning mutagenesis along the entire length of S4, again in the yeast H ϩ -ATPase. S4 almost certainly contributes to the stalk structure seen by cryoelectron microscopy, and it links the Asp residue that is phosphorylated by ATP to membrane segment 4 (M4). M4 in turn plays a central role in cation binding and translocation, based on mutagenesis studies of the sarcoplasmic reticulum Ca 2ϩ -ATPase (11)(12)(13)(14)(15)(16), plasma membrane Ca 2ϩ -ATPase (17), Na ϩ ,K ϩ -ATPase (16, 18 -20), and gastric H ϩ ,K ϩ -ATPase (21). Consistent with this picture, the mutagenesis results to be described below provide evidence that stalk segment 4 helps to mediate the E 1 /E 2 conformational change in the yeast H ϩ -ATPase.
Mutagenesis-To introduce mutations into the S4 region of the H ϩ -ATPase by polymerase chain reaction (24), two restriction fragments of the PMA1 gene were employed, both subcloned into a modified Bluescript plasmid (Stratagene, La Jolla, CA). Mutations from Lys-355 to Ala-370 were introduced into a 615-base pair BstEII/EcoRI restriction fragment, whereas mutations from Gly-371 to Leu-375 were introduced into a 495-base pair StyI/BamHI fragment. After DNA sequencing to verify the mutation, the restriction fragment was moved into plasmid PMA1.2 (22). The 3.8-kilobase pair HindIII/SacI fragment containing the entire pma1-coding region was then cloned into the yeast expression vector YCp2HSE (22), placing the mutant allele under control of two tandemly arranged heat-shock elements. Finally, the plasmids were transformed into yeast strain SY4 (see above) according to the method of Ito et al. (25).
Isolation of Secretory Vesicles and Quantitation of Expressed ATPase-Transformed SY4 cells were grown to mid-exponential phase (A 600 ϳ1) at 23°C in supplemented minimal medium containing 2% galactose, shifted to medium containing 2% glucose for 3 h, and then heat-shocked at 39°C for an additional 2 h. The cells were harvested and washed, and secretory vesicles were isolated and suspended in 0.8 M sorbitol, 1 mM EDTA, 10 mM triethanolamine/acetic acid, pH 7.2, as described previously (26). To determine the level of expressed PMA1 protein, secretory vesicles (5-20 g) were subjected to SDS-polyacrylamide gel electrophoresis and immunoblotted. Quantitative Phospho-rImager (Bio-Rad) analysis was carried out at two protein concentrations within the linear range, and the expression level was calculated from the average of two or more determinations.
ATP Hydrolysis-Unless otherwise noted, ATP hydrolysis was assayed at 30°C in 0.5 ml of 50 mM MES 1 /Tris, pH 5.7, 5 mM KN 3 , 5 mM Na 2 ATP, 10 mM MgCl 2 , and an ATP-regenerating system (5 mM phosphoenolpyruvate and 50 g/ml pyruvate kinase). The reaction was terminated after 20 -40 min, and the release of inorganic phosphate from ATP was determined by the method of Fiske and Subbarow (27). Specific activity was calculated as the difference between ATP hydrolysis in the absence and in the presence of 100 M sodium orthovanadate, a potent inhibitor of P-type ATPases. IC 50 values for vanadate inhibition were determined by measuring ATP hydrolysis in the presence of increasing concentrations of vanadate. For determination of K m values, Na 2 ATP was varied between 0.15 and 7.5 mM with MgCl 2 always in excess of ATP by 5 mM; actual concentrations of MgATP were calculated by the method of Fabiato and Fabiato (28). To determine the pH optimum for ATP hydrolysis, the pH of the assay media was adjusted to values between 5.2 and 7.5 with Tris base.
ATP hydrolysis was also assayed under conditions similar to that used for quantitation of proton transport (see below), as described previously (29). Briefly, secretory vesicles (5-10 g of protein) were diluted into 200 l of 0.6 M sorbitol, 0.1 M KCl, 20 mM HEPES/KOH, pH 6.7, Na 2 ATP (0.3 to 3.0 mM), and MgCl 2 (5 mM excess over the ATP concentration) at 30°C. The reaction was stopped after 20 -40 min by addition of trichloroacetic acid to a final concentration of 5%, and inorganic phosphate was determined.
Proton Transport and Fluorescence Quenching-ATP-dependent proton transport was assayed by measuring the initial rate of acridine orange fluorescence quenching as described previously (29). The specific initial rate of quenching was adjusted for the level of ATPase expression and reported as a percentage of the wild-type rate. In order to examine the coupling of proton transport to ATP hydrolysis, the initial rate of fluorescence quenching was determined over a range of ATP concentrations and plotted as a function of the rate of ATP hydrolysis, assayed under similar conditions (see above).
Trypsinolysis-Limited trypsinolysis was performed on isolated secretory vesicles as described previously (30). Vesicles were suspended at 0.5 mg/ml in 20 mM Tris-HCl, pH 7.0, and 5 mM MgCl 2 . Following preincubation at 30°C for 5 min in the presence of 0, 1, 10, or 100 M orthovanadate, tosylphenylalanyl chloromethyl ketone-treated trypsin was added (trypsin/protein ratio of 1:4), and the incubation was contin-ued for 20 min. The reaction was terminated by the addition of 1 mM diisopropyl fluorophosphate, and the products were analyzed by immunoblotting with polyclonal antiserum against the ATPase.
Protein Determination-Protein concentrations were determined by the method of Lowry et al. (31) as modified by Ambesi et al. (26), with bovine serum albumin as standard.

Selection of Residues for Mutagenesis-
The goal of the present study was to explore the functional role of amino acid residues throughout stalk segment 4 (S4), which links membrane segment 4 (M4) with the phosphorylation site (Asp-378) of the yeast plasma membrane H ϩ -ATPase (Fig. 1). Residues from Lys-355 to Leu-375 were subjected to alanine-scanning mutagenesis, filling in the 21-amino acid stretch between previously published studies of M4 (Tyr-325 through Ala-354; Ref. 29) and the phosphorylation domain (Cys-376 through Thr-384; Refs. 32 and 33). With the exception of Ala-358, Ala-365, and Ala-370, which were replaced with Ser, each residue was changed to Ala. Each mutant allele was cloned into the expression vector YCp-2HSE, transformed into yeast strain SY4, and expressed under the control of a heat-shock promoter after turning off the wild-type PMA1 allele (22). Secretory vesicles containing newly synthesized mutant ATPase were then isolated and characterized (26).
Expression and ATP Hydrolysis-As summarized in Table I  (top part), quantitative immunoblotting revealed that Ala/Ser substitutions in S4 had only a modest effect on biogenesis, with mutant H ϩ -ATPases reaching the secretory vesicles at 35-101% of the amount seen in the wild-type control. Likewise, 19 of 21 mutant enzymes clearly retained the ability to hydrolyze ATP, with specific activities ranging from 27 to 118% after correction for the level of expression in the vesicles. Only two of the mutants showed more serious defects in ATP hydrolysis as follows: L369A (12%) and I374A (16%), and even in these cases, the uncorrected activities were 4 -5-fold greater than the background values measured in the empty plasmid control. Thus, none of the residues in S4 appeared to be completely essential for ATP hydrolysis, although most of the mutations caused at least a 50% decrease in the rate of hydrolysis.
Proton Transport-Given that S4 physically links the phosphorylated Asp residue to the membrane, it was important to examine the effects of the Ala/Ser substitutions on the ability of the ATPase to transport protons. This was measured by fluorescence quenching of the pH-sensitive dye, acridine orange (Table I, top part). In four of the mutants (K356A, L369A, I374A, and L375A), ATP-dependent quenching was clearly 1 The abbreviation used is: MES, 4-morpholineethanesulfonic acid. FIG. 1. Sequence alignment of stalk segment 4. A, the 30-amino acid stretch from the beginning of stalk segment 4 to the end of the phosphorylation region has been aligned for 9 representative P 2 -AT-Pases. Swiss-Prot sequence accession numbers are (top to bottom) as follows: P05030, P07038, P09627, P54210, P20649, Q00804, P13585, P04074, and P09626. For clarity, residues identical to the yeast PMA1 sequence are replaced with a period. The catalytic Asp residue is indicated in bold type. B, the Garnier-Robson secondary structure prediction for the yeast PMA1 H ϩ -ATPase was determined using the Peptide Structure program of the University of Wisconsin Genetics Computer Group. Residues likely to assume an ␣-helical structure are indicated by H, ␤-turns by T, and indeterminate structure by periods.
above background, but the activities were so low that quantitative determinations were not possible. In one mutant (E367A), the initial rate of proton transport (83% of wild type) appeared to exceed the rate of ATP hydrolysis (46% of wild type). Closer examination of this mutant seemed warranted, given the fact that Glu-367 is strongly conserved among P 2 -ATPases (see Fig. 1). When proton transport was measured over a range of ATP concentrations, however, and the initial rates were plotted as a function of ATP hydrolysis assayed under the same conditions, the data for E367A (like the data for three other mutants) proved to be very similar to the wild-type control (Fig. 2). It later became obvious that the discrepancy in Table I could be accounted for by the abnormal pH optimum of the E367A ATPase (see legend to Fig. 2 and below). For all of the remaining mutants, the initial rate of ATP-dependent quenching closely paralleled that of ATP hydrolysis ( Table I), suggesting that the substitutions had little or no effect on the coupling between transport and hydrolysis.
Vanadate Resistance-Each of the mutant ATPases was next examined for changes in sensitivity to inorganic orthovanadate, an inhibitor that binds tightly to the E 2 conformation of P-type ATPases. Here, a striking result was obtained; of the 21 mutants studied, 9 showed a very significant degree of vana-date resistance, with IC 50 values above 10 M (Table II, top part). The resistant mutations began at position 360 and continued (with interruptions) through position 374. Upon closer inspection of this region, it was apparent that 4 of the 6 mutants still sensitive to vanadate carried substitutions of Ser for Ala or Ala for Ser. Because a less conservative substitution at one of these positions (S368F) had previously been shown to produce a highly vanadate-resistant enzyme (34), it seemed worthwhile to make further replacements of Ser-364, Ala-365, and Ala-370. To define the ends of the region, Ile-359, Val-372, and Glu-373 were also included.
Among the additional mutants that were tested, all but one (V372F) were expressed in secretory vesicles at 43% or more of the wild-type level, and all but two (A370L and V372F) had measurable rates of ATP hydrolysis (Table I, bottom part). Furthermore, for every mutant with sufficient activity to be assayed, there was a close correspondence between the initial rate of ATP hydrolysis and the initial rate of ATP-dependent proton transport (Table I), indicating that these substitutions (like the initial set described above) had little or no effect on the coupling between hydrolysis and transport.
b Vanadate-sensitive ATP hydrolysis was measured as described under "Experimental Procedures." One unit is defined as 1 mol of P i /min. Values are reported with and without correction for the amount of mutant H ϩ -ATPase expressed in the secretory vesicles.
c The initial rate of fluorescence quenching (proton transport) was determined as previously described (29). One unit is defined as 1% of total fluorescence quenching/min. d ND, not determined. Corrections for expression have not been made for mutants with measured ATPase activities below 4% of the wild-type control.
M (Table II, bottom part), confirming that substantial vanadate resistance could be observed at these positions as well. The distribution of resistance along S4 is illustrated in Fig. 3A, which demonstrates that the 13-amino acid stretch from Ile-359 through Gly-371 can undergo mutations that elevate the IC 50 more than 5-fold above the wild-type value.
Other Kinetic Properties-In previous mutagenesis studies of the yeast H ϩ -ATPase, vanadate resistance has frequently been accompanied by a reduction in the apparent K m value for MgATP and a rise in pH optimum (29,30). This "coordinated" phenotype has been interpreted as a shift in equilibrium from the vanadate-sensitive E 2 conformation toward the E 1 conformation, where MgATP and the transported proton are expected to bind with higher affinity than in E 2 . In the present study, 6 of the 14 vanadate-resistant mutants in the stalk 4 region (I359F, S364D, I366A, E367A, L369A, and A370F) displayed K m values of 0.5 mM or lower, representing at least a 3-fold change from the value seen in the wild-type control; six additional mutants (Q361A, K362A, L363A, A365F, A365L, and I374A) gave smaller but reproducible reductions to 0.6 -0.9 mM (Table II). Thus, a strong but not complete correlation could be seen between vanadate resistance and a decreased K m for MgATP (Fig. 3B).
Not surprisingly, given the many ways in which the pH dependence of the H ϩ -ATPase might be altered, the situation was less clear-cut with regard to pH optimum. Four of the vanadate-resistant mutants (L363A, I366A, E367A, and A370F) displayed a conspicuous alkaline shift of 0.3 to 0.6 pH FIG. 2. Coupling between ATP hydrolysis and proton transport in S4 mutants. The initial rate of fluorescence quenching (proton transport) was determined over a range of ATP concentrations (0.3-3.0 mM) and plotted as a function of the rate of ATP hydrolysis measured under similar conditions as described previously (26). The data represent the average of 2-4 independent preparations, and the lines were drawn by least squares analysis.  units (Table II). The rest retained an essentially normal pH optimum, except for a few with a minor shift in the acid direction (e.g. K362A). E 1 to E 2 Conformational Change Assessed by Limited Trypsinolysis-If the cluster of kinetic changes seen in many of the stalk 4 mutants reflects a shift in equilibrium from the vanadate-sensitive E2 conformation toward the vanadate-insensitive E1 conformation, it should be possible to detect the shift by limited trypsinolysis (29,30). In the experiment of Fig. 4, secretory vesicles were incubated with trypsin for 20 min in the presence of 0, 1, 10, or 100 M vanadate. The reaction was stopped by the addition of diisopropyl fluorophosphate (1 mM), and the digestion products were separated by SDS-polyacrylamide gel electrophoresis and immunoblotted with anti-PMA1 antibody. Under the conditions of the experiment, the wild-type ATPase displayed significant protection by vanadate, with the accumulation of 97-, 70-, and 60-kDa fragments. By contrast, the fragments were barely visible in E367A and not seen at all in I366A, consistent with the idea that the mutant enzymes had difficulty reaching the vanadate-protectable E2 conformation. DISCUSSION This study reports an impressive stretch of 13 consecutive vanadate-resistant mutations in stalk segment 4 of the yeast PMA1 H ϩ -ATPase. While similar mutations have been described previously for the yeast ATPase (29,30,32,(35)(36)(37), they have been scattered throughout the 100-kDa polypeptide, with no discernible structural or functional pattern (Fig. 5). The cluster in S4 therefore deserves special interest.
At the structural level, S4 is customarily depicted as an ␣-helix, acting with three other stalk helices to connect the cytoplasmic and membrane-embedded portions of the ATPase. Evidence for the ␣-helical nature of S4 came originally from the use of standard algorithms (as reviewed in Ref. 38) and has since been reinforced by cryoelectron microscopy of the sarcoplasmic reticulum Ca 2ϩ -ATPase, where an ␣-helical backbone was found to fit comfortably into each of four rod-like densities within the stalk region (5). Significantly, however, the residues represented by the vanadate-resistant mutants from Ile-359 through Gly-371 would not be restricted to one face of such a helix; instead, they would occupy three full turns near the middle of S4.
Functionally, it seems likely that the mutations cause a shift in E 1 -E 2 equilibrium toward the vanadate-insensitive E 1 conformation, given the fact that most of them display coordinated changes in the K m value for MgATP and the pH optimum of the H ϩ -ATPase. In an earlier paper (29), we described three similar mutants, spaced at intervals along transmembrane segment M4 (I332A, V336A, and V341A). Likewise, Blostein and co-workers (39,40) have pointed out that a shift in E 1 -E 2 equilibrium could explain the kinetic behavior of two Na ϩ ,K ϩ -ATPase mutants: E233K, located in the M2-M3 cytoplasmic loop (39,40), and ␣1M32, a deletion mutant lacking 32 amino acids at the N terminus. In both mutants, vanadate resistance was accompanied by a decrease in the K m value for MgATP and by marked K ϩ activation of Na-ATPase activity at micromolar ATP concentrations, a condition under which the E 2 (K) to E 1 step is normally rate-limiting.
The concentration of E 1 -E 2 mutants in stalk segment 4 of the yeast H ϩ -ATPase suggests that S4 may provide a critical, conformationally mobile link between the cytoplasmic phosphorylation site and cation-binding sites in the membrane. Independent evidence for this idea has come from mutagenesis studies by Inesi and co-workers (41) on stalk segment 4 of the sarcoplasmic reticulum Ca 2ϩ -ATPase. Here, single substitutions of conserved S4 residues significantly slowed the rate of ATP hydrolysis and Ca 2ϩ uptake, even though measurable levels of phosphorylated intermediate were formed (41). Single substitutions of non-conserved residues were less damaging, but multiple substitutions of such residues interfered with the Ca 2 ⅐E 1 P to Ca 2 ⅐E 2 transition (as reflected by the rate of phosphoenzyme turnover) and the E 1 to E 2 transition (as reflected by the time course of EGTA inactivation, Ref. 42). Thus, once again it seems likely that stalk 4 has a profound influence on ratelimiting conformational transitions.
To visualize the way in which S4 performs this function, high resolution structures for both the E 1 and E 2 conformations will be required. Reporter groups engineered into specific locations along S4 will also help to track the conformational changes; studies along these lines are presently under way in several laboratories including our own.