Role of transmembrane segment M8 in the biogenesis and function of yeast plasma-membrane H⁺-ATPase

Abstract

Of the four transmembrane helices (M4, M5, M6, and M8) that pack together to form the ion-binding sites of P₂-type ATPases, M8 has until now received the least attention. The present study has used alanine-scanning mutagenesis to map structure-function relationships throughout M8 of the yeast plasma-membrane H⁺-ATPase. Mutant forms of the ATPase were expressed in secretory vesicles and at the plasma membrane for measurements of ATP hydrolysis and ATP-dependent H⁺ pumping. In secretory vesicles, Ala substitutions at a cluster of four positions near the extracytoplasmic end of M8 led to partial uncoupling of H⁺ transport from ATP hydrolysis, while substitution of Ser-800 (close to the middle of M8) by Ala increased the apparent stoichiometry of H⁺ transport. A similar increase has previously been reported following the substitution of Glu-803 by Gln (Petrov, V. et al., J. Biol. Chem. 275:15709–15718, 2000) at a position known to contribute directly to Ca²⁺ binding in the Ca²⁺-ATPase of sarcoplasmic reticulum (Toyoshima, C., et al., Nature 405: 647–655, 2000). Four other mutations in M8 interfered with H⁺-ATPase folding and trafficking to the plasma membrane; based on homology modeling, they occupy positions that appear important for the proper bundling of M8 with M5, M6, M7, and M10. Taken together, these results point to a key role for M8 in the biogenesis, stability, and physiological functioning of the H⁺-ATPase.

1. Introduction

The plasma-membrane H⁺-ATPase of yeast is encoded by the PMA1 gene [1] and belongs to the physiologically important family of P₂-type ATPases [2], which couple ATP hydrolysis to the transport of monovalent and divalent cations across cell membranes. Other members of the family include eukaryotic Na⁺/K⁺-, H⁺/K⁺-, Ca²⁺-, Mn²⁺- and prokaryotic Mg²⁺-ATPases. Structurally, the P₂-type cation pumps share a common topology in which four N-terminal transmembrane segments (M1–M4) and six C-terminal transmembrane segments (M5–M10) serve to anchor the 100 kDa ATPase polypeptide in the membrane. A large hydrophilic loop between M4 and M5 protrudes into the cytoplasm and contains well-conserved residues involved in ATP binding and hydrolysis.

The year 2000 saw a major step towards understanding the detailed molecular architecture of P₂-ATPases with the publication of a 2.6Å crystal structure for the Ca2+-ATPase of sarcoplasmic reticulum [3]. Building on cryoelectron microscopic studies that had defined the number of transmembrane α-helices [4] and on site-directed mutagenesis that had clearly implicated M4, M5, M6, and M8 in high-affinity Ca²⁺ binding [5, 6], the new structure established the precise way in which those four helices cooperate to form a pair of membrane-embedded Ca²⁺ binding sites. Since then, a companion crystal structure has revealed that the transmembrane domain rearranges significantly in the E₂ conformation, causing the binding sites to be lost and Ca²⁺ to be released at the lumenal surface of the membrane [7]. Most recently, additional conformations that form part of the catalytic cycle have been documented by crystal structures of the Ca²⁺-ATPase complexed with AMPPCP, ADP, AlF₄, and thapsigargin [8–13].

It is very likely that the same membrane-spanning helices shape the transport pathway of all P₂-type ATPases, given the pronounced structural similarity between the membrane domains of Neurospora plasma-membrane H⁺-ATPase (mapped at 8Å resolution) and sarcoplasmic reticulum Ca²⁺-ATPase [14]. Additional evidence comes from the fact that amino acid substitutions along M4, M5, and M6 have been shown to affect the transport properties of P₂-type Na+/K+-[15–21], H+/K+-[22–26], H+-[27–31], and Ca²⁺/Mn²⁺- [32] ATPases. M8 has been less well characterized, however, and the generality of its role is not yet clear. For this reason, we were interested to find recently that a mutation in this segment of yeast plasma-membrane H⁺-ATPase (E803N) decreases the apparent stoichiometry of proton transport, while another mutation at the same position (E803Q) actually increases the apparent stoichiometry [29]. The present study has followed up on this result by using scanning mutagenesis to examine structure-function relationships along the entire length of M8. The results pinpoint additional residues that help to dictate transport stoichiometry, and also give evidence for an unexpectedly important role of M8 in the folding and biogenesis of the H⁺-ATPase.

2. Materials and methods

2.1. Yeast strains and growth conditions56

Two strains of Saccharomyces cerevisiae were used in this study: SY4 (MATa; ura3-52; leu2-3, 112; his4-619; sec6-4; GAL; pma1::YIpGAL-PMA1; ref. 33) and NY13 (MATa ura3-52). In SY4, the chromosomal PMA1 gene has been placed under control of the GAL1 promoter (P_GAL-PMA1) by the gene disruption strategy of Cid et al. [34]. SY4 also carries a temperature-sensitive mutation in the SEC6 gene, blocking the last step in plasma membrane biogenesis and leading to the accumulation of secretory vesicles [35].

2.2. Site-Directed Mutagenesis

The polymerase chain reaction [36] was used to introduce mutations into a 519bp BglII to SalI fragment of the PMA1 gene [1] that had previously been subcloned into a modified version of Bluescript (Stratagene, La Jolla, CA). Each fragment was sequenced to verify the presence of the mutation and the absence of unwanted base changes, and then moved into the full-length PMA1 gene in plasmid pPMA1.2 [33] or plasmid pVP3, a wild-type version of pGW201 [37]. To express the ATPase in secretory vesicles, a 3.7-kb HindIII to SacI piece of pPMA1.2 containing the entire coding sequence of the gene was transferred to the centromeric plasmid YCp2HSE, placing the mutated gene under heat-shock control; the resulting plasmid was transformed into strain SY4 of Saccharomyces cerevisiae [33]. To introduce the mutation into the chromosomal copy of the PMA1 gene for expression in plasma membranes, a 6.1-kb HindIII fragment containing the mutant allele linked to URA3 was excised from plasmid pVP3 and integrated into strain NY13 using the Alkali-Cation Yeast transformation kit (Bio 101). In all cases, DNA sequencing was repeated to confirm the identity of the mutant allele.

2.3. Cell Fractionation

For studies in secretory vesicles, SY4 cells transformed with the desired plasmid were grown to mid-exponential phase (OD₆₀₀ ~ 0.7–1.2) on supplemented minimal medium containing 2% galactose at 23°C, shifted to medium containing 2% glucose for 2.5 h to turn off expression of the chromosomal copy of the PMA1 gene, and then shifted to 37°C for 2 h to turn on expression of the plasmid-borne allele and block the fusion of secretory vesicles with the plasma membrane. Vesicles containing newly synthesized plasmid-encoded ATPase were isolated by differential centrifugation, further purified by gradient centrifugation [38], and suspended at a protein concentration of 1.0–3.0 mg/ml in 0.8 M sorbitol, 10 mM triethanolamine-acetic acid, pH 7.2, 1 mM EDTA, chymostatin (2μg/ml) and leupeptin, pepstatin, and aprotinin (each 1 μg/ml).

For studies on plasma membranes, NY13-derived strains were grown to mid-exponential phase in supplemented minimal medium containing 4% glucose, and a plasma membrane-enriched fraction was prepared by the method of Perlin et al. [39], followed by washing in 1 mM EGTA-Tris, pH 7.5, and resuspension in the same buffer. All preparative procedures were carried out at 0–4°C.

The amount of expressed ATPase was measured by SDS-polyacrylamide gel electrophoresis and Western blotting, as described elsewhere [38]. Blots were treated with affinity-purified polyclonal antibody against the closely related plasma-membrane H⁺-ATPase of Neurospora crassa [40] and then with ¹²⁵I-protein A (ICN, Irvine, CA), and assayed by means of a PhosphorImager equipped with ImageQuant software version 5.0 (Molecular Dynamics). The expression level of mutant ATPase relative to a wild-type control was calculated from the average of three or more determinations.

2.4. Metabolic labeling, immunoprecipitation, and limited trypsinolysis

For mutant ATPases that were poorly expressed in secretory vesicles, the adequacy of protein folding was determined by limited trypsinolysis as described previously [28]. Briefly, cells were incubated with [³⁵S]methionine (Amersham Pharmacia Biotech) under conditions permitting expression of the plasmid-borne but not the chromosomal PMA1 gene. Total membranes were isolated and treated with tosyl-phenylalanyl chloromethyl ketone-trypsin (Worthington Biochemical Corp.) for various times at a trypsin:protein ratio of 1:25, and the reaction was stopped by adding diisopropyl fluorophosphate to a final concentration of 1 mM. Samples were then immunoprecipitated with polyclonal anti-ATPase antibody and analyzed by SDS-polyacrylamide gel electrophoresis and fluorography [38]. For limited trypsinolysis of mutant ATPases that were expressed at the plasma membrane, a 1:4 trypsin/protein ratio was used. After the reaction was stopped with 1 mM diisopropyl fluorophosphate, Laemmli buffer was added and aliquots (usually 0.4 μg of protein) were subjected to SDS-polyacrylamide gel electrophoresis and electroblotted onto PVDF membrane. The bands were then visualized using the Enhanced Chemiluminiscence kit from Amersham.

2.5. ATPase activity

For mutant ATPases that were expressed detectably in secretory vesicles and/or plasma membranes, ATP hydrolysis was assayed at 30°C in buffer containing 50 mM MES-Tris, pH 5.7, 5 mM Na₂ATP, 5 mM MgCl₂, 5 mM KN₃ and an ATP-regenerating system (5 mM phosphoenolpyruvate and 50 μg/ml pyruvate kinase) in the absence or presence of 100 μM sodium orthovanadate. Inorganic phosphate was measured by the method of Fiske and Subbarow [41] and protein concentrations were determined by the method of Lowry et al. [42] and/or Bensadoun and Weinstein [43]; for the protein assay, an appropriate aliquot of vesicle resuspension buffer was added to the bovine serum albumin standard to compensate for changes in absorbance due to the presence of interfering compounds. For the determination of pH optimum, the pH was adjusted to values between 5.0 and 7.0. K_m determinations were made in the same medium except that ATP was replaced with varying amounts of MgATP (1:1); in each case, the concentration of MgATP was calculated according to Fabiato and Fabiato [44].

2.6. Coupling between proton pumping and ATP hydrolysis

H⁺ pumping into secretory vesicles was monitored by fluorescence quenching of the pH-sensitive dye acridine orange as described previously [29]. Assays were carried out at 29°C with continuous stirring on a Hitachi F2000 fluorescence spectrophotometer (excitation, 430 nm; emission, 530 nm) equipped with Intracellular Cation Measurement System software. Freshly prepared vesicles (40–50 μg) were suspended in 1.5 ml of 0.6 M sorbitol, 20 mM HEPES-KOH, pH 6.7, 100 mM KCl, 20 mM KNO₃, 2 μM acridine orange, and ATP (0.2–2.0 mM), and after stabilization of baseline fluorescence (120–150 s), proton pumping was initiated by the addition of MgCl₂ (5.2–7.0 mM).

Parallel measurement of ATPase activity under the same conditions was performed in 100 μl of 0.6 M sorbitol, 20 mM HEPES-KOH, pH 6.7, 100 mM KCl, 20 mM KNO₃, containing ATP (0.2–2.0 mM) and MgCl₂ (5.2–7.0 mM) at 29°C for 20–40 min. The reaction was stopped with 1 ml of 1.25% trichloracetic acid, and inorganic phosphate was measured as above.

2.7. Homology modeling

Three-dimensional models of yeast PMA1 H⁺-ATPase were built based on crystallographic structures of SERCA1a Ca²⁺-ATPase in the E₁ and E₂ conformations [3, 7]. The first step was to align the H⁺- and Ca²⁺-ATPase sequences. Given the borderline homology between the two ATPases in certain of the transmembrane helices, this process was guided by using a set of six algorithms to determine the location and length of the 10 helices in the H⁺-ATPase. Four of the algorithms (HMMTOP [45], SOUSI [46], TMHMM [47], and TopPred [48]) predicted all 10 helices with a very high degree of overlap, while DAS [49] missed M1 and TMPRED [50] missed M8. A consensus transmembrane topology for the H⁺-ATPase was constructed with this information (Table 1 in "Supplementary Material"). The non-homologous N- and C-termini were then removed from the H⁺- and Ca²⁺-ATPases, and the two sequences were aligned by means of the Clustal X algorithm [51], followed by visual inspection and modification of the alignment (Fig. 1 in "Supplementary Material"). As pointed out previously, it was necessary to introduce significant gaps into the M2–M3 and M4–M5 cytoplasmic loops, reflecting the smaller size of the A and N domains in the H⁺-ATPase.

Three-dimensional models for the H⁺-ATPase were then constructed with the program MODELLER 8v0 [52], using the recently published 2. 4Å E1 and 3.1Å E₂ structures of SERCA1a ATPase (PDB codes 1SU4 and 1IWO, respectively) as templates. Model building by this program is based on the satisfaction of spatial restraints and the optimization of a molecular probability density function employing methods of conjugate gradients for energy minimization and molecular dynamics with simulated annealing [52].

Twenty models were made for each conformation, and five of each were submitted to http://www.jcsg.org/scripts/prod/validation/sv2.cgi for structural evaluation. Algorithms used to identify the best E₁ and E₂ models included PROCHECK [53] and WHAT_CHECK [54] to assess the stereochemistry of main-chain and side-chain residues and the packing quality and planarity of rings; ERRAT [55] to examine non-bonded interactions between different atom types; and PROVE [56] to check atomic volumes. The results indicated high quality for both the E₁ and E₂ models: Ramachandran plots with 90.6% and 89.8% of residues in the most favored regions (E₁ and E₂, respectively); overall quality factors of 73.7 and 78.1 in the ERRAT analysis; and average z-scores of 0.61 and 0.59 in PROVE.

3. Results

3.1. Selection of residues for mutagenesis

The present study has used alanine-scanning mutagenesis to examine the functional role of amino acid residues throughout M8 of the yeast PMA1 H⁺-ATPase. Based on an earlier hydropathy analysis, residues from Met-791 through Arg-811 were included in the study. Kuhlbrandt and co-workers [57] have recently presented a structural model for the closely related Pma1 H⁺-ATPase of Neurospora crassa, obtained by mapping an 8Å cryoelectron microscopic structure of that enzyme onto the 2.6Å E₁ structure of SERCA1 Ca²⁺-ATPase; the resulting model shows M8 extending from Gly-791 through Thr-812 in N. crassa and thus from Gly-789 through Thr-810 in S. cerevisiae. The E₁ and E₂ models developed for the S. cerevisiae ATPase in the present study gave a very similar picture, with M8 running from Phe-788 through Thr-810.

Not surprisingly, M8 has remained nearly identical throughout the Pma1 H⁺-ATPases of ascomycetous fungi (Fig. 1, S. cerevisiae through Aspergillus nidulans). There is also significant similarity to published sequences from three other fungi (Ustilago maydis, Filobasidiella neoformans, and Uromyces fabae) and two algae (Dunaliella bioculata and Cyanidium caldarum), although all but one of these enzymes more closely resemble the plasma-membrane H⁺-ATPases of higher plants (for example, Arabidopsis thaliana AHA1). Mammalian P₂-ATPase sequences have diverged so extensively towards the C-terminal end of the molecule that it is difficult to align them with certainty, but the structural comparison of Kuhlbrandt and co-workers [57] suggests that the alignment shown in Fig. 1 is at least approximately correct. The apparent correspondence between E803 of the yeast H⁺-ATPase, E908 of SERCA1a Ca²⁺-ATPase, E938 or Q941 of HKA H⁺,K⁺-ATPase, Q933 of NKA Na⁺,K⁺-ATPase, and Q971 of PMCA1 Ca²⁺-ATPase will become important in interpreting the results to be presented below.

Fig. 1.

Alignment of M8 sequences from P₂-type ATPases. The alignment was carried out as described in "Materials and Methods." Accession numbers (top to bottom) are as follows: P05030, P49380, AF109913, P28877, P24545, P09627, P07038, Q07421, AAB069598, AAC27991, AJ315590, AL603947, AF077766, AY149918, P54211, E13998, P20649, P04191, P11505, P54707, and P04074.

3.2. Expression and activity of the M8 mutants in secretory vesicles

Initially, the yeast H⁺-ATPase gene containing each of the 21 Ala mutations was cloned into the expression vector YCp-2HSE, transformed into yeast strain SY4, and expressed under control of a heat-shock promoter after turning off the wild-type PMA1 allele [33]. In strain SY4, newly synthesized wild-type ATPase has been shown to travel rapidly from the endoplasmic reticulum to post-Golgi secretory vesicles, where it becomes arrested due to a temperature-sensitive block in vesicle fusion with the plasma membrane. Thus, the trafficking ability of mutant ATPases can be assessed by isolating the secretory vesicles and immunoblotting with anti-PMA1 antiserum.

When such an experiment was carried out with the M8 mutants, four of the 21 Ala-substituted ATPases were found in the vesicles at less than 10% of the wild-type level (Table 1), suggesting that they were poorly folded and blocked in biogenesis at an earlier step of the pathway. To test this idea, each of the four mutants was metabolically labeled with ³⁵S-Cys and Met; total membranes were isolated and exposed to a limiting concentration of trypsin; and H⁺-ATPase fragments were immunoprecipitated with polyclonal anti-PMA1 antiserum and analyzed by SDS-polyacrylamide gel electrophoresis and autoradiography. Indeed, as illustrated in Fig. 2, the mutant ATPases were significantly more sensitive to trypsin than the wild-type control. Thus, M8 appears to play an important role in the proper folding of the yeast Pma1 H⁺-ATPase; this result will be discussed in a later section.

TABLE 1.

Effect of M8 mutations on expression and ATP hydrolysis in secretory vesicles

Mutation	Expressiona	ATP hydrolysisb			Km MgATP	IC50 vanadate	pH optimum
		uncorrected		corrected
	%	U/mg	%	%	mM	μM
Wild typec	100	5.58	100	100	1.4	1.5	5.7
Noned	2	0.08	1	*	*	*	*
M791A	84	1.94	39	49	1.3	1.5	5.7
N792A	71	1.61	34	53	1.1	2.2	5.4
G793A	69	1.07	27	38	1.1	2.9	5.4
I794A	8	0.16	4	*	*	*	*
M795A	35	0.72	15	45	0.4	1.9	5.7
F796A	9	0.19	4	*	*	*	*
L797A	19	0.83	20	105	0.7	1.8	5.6
Q798A	1	0.05	1	*	*	*	*
I799A	5	0.10	2	*	*	*	*
S800A	73	2.96	56	89	1.4	2.1	5.7
L801A	19	0.42	7	*	*	*	*
T802A	94	2.77	47	52	1.1	1.8	5.7
E803A	20	0.91	15	74	0.7	4.2	5.7
N804A	85	2.26	41	46	1.6	2.7	5.7
W805A	63	2.27	40	65	0.9	2.2	5.7
L806A	47	2.68	58	120	1.8	2.0	5.7
I807A	18	0.42	11	*	*	*	*
F808A	69	2.66	53	83	0.8	2.6	5.7
I809A	82	3.00	72	88	1.5	1.8	5.7
T810A	80	3.03	69	84	1.1	2.1	5.7
R811A	38	1.99	29	86	0.9	2.2	5.7
N792D	65	4.41	63	98	1.3	1.3	5.7
N792Q	62	5.28	73	118	1.2	1.5	5.7
N792H	70	3.13	44	64	1.1	1.5	5.7
G793E	61	1.08	19	36	0.8	2.0	5.4
Q798E	111	5.92	87	78	1.1	1.5	5.7

^aThe specific expression of 100-kDa ATPase protein was calculated by quantitative immunoblotting as described in "Materials and Methods.”

^bATP hydrolysis was measured as outlined in "Materials and Methods.” This series of experiments included 31 wild-type preparations with an average ATPase activity of 5.58+/−0.21 μmol P_i/min.mg. Data for mutant ATPases are the average of 4–12 determinations except in the case of N792D, N792Q, N792H, G793E, I794A, Q798A, I799A, L801A, and I807A, for which there were two determinations; each mutant value was corrected for expression relative to a wild-type control run in parallel on the same day.

^cSecretory vesicles were isolated from cells containing the expression plasmid YCp2HSE with the wild-type PMA1 gene (positive control).

^dSecretory vesicles came from cells carrying the plasmid with no PMA1 gene (negative control). Asterisks indicate that no corrections were made for preparations with low expression and/or ATPase activity.

Fig. 2.

Trypsinolysis of newly synthesized I794A, F796A, Q798A, and I799A ATPases. ³⁵S-labeled total membranes were incubated at a trypsin/protein ratio of 1:25 for 0, 2, or 10 min at 30°C. The reaction was stopped by the addition of 1 mM diisopropyl fluorophosphate, and Pma1 protein was immunoprecipitated and subjected to SDS-polyacrylamide gel electrophoresis and fluorography as described in "Materials and Methods." WT, wild type.

The remaining Ala mutants reached the secretory vesicles at 18% to 94% of the level shown by the wild-type ATPase, and after correction for the expression level, were able to hydrolyze ATP at 37% to 120% of the control rate (Table 1). Apart from two mutants (L801A and I807A) with activity too low to be characterized further, none of these mutations caused a significant change in the kinetic properties of the H⁺-ATPase, including the K_m for MgATP, the IC₅₀ for inhibition by orthovanadate, or the pH optimum for ATP hydrolysis (Table 1).

3.3. Proton pumping by the M8 mutants

A different picture emerged when secretory vesicles containing the mutant ATPases were assayed for H⁺ pumping over a range of MgATP concentrations, using acridine orange as a fluorescent probe. As illustrated in Fig. 3, wild-type ATPase typically displays a linear relationship between the initial rate of fluorescence quenching and the rate of ATP hydrolysis, with a slope that corresponds to the H⁺/ATP stoichiometry [27, 29]. Strikingly, Ala substitutions at four positions close to the extracytoplasmic end of M8 (M791A, N792A, G793A, and M795A) and one position immediately beyond the cytoplasmic end (R811A) led to a reduction in the slope of the quenching-vs-hydrolysis plot (Fig. 3), consistent with the idea that these mutant ATPases are partially or completely uncoupled. Reductions were also obtained with other amino acid replacements at two of the same positions: N792D, Q, and H and G793E, which mimicked naturally occurring substitutions found in other fungal species (Fig. 1). By contrast, little or no change in coupling was seen following Ala substitutions at six positions in the middle of M8 (I797A, T802A, N804A, W805A, L806A, T810A) and a Gln-to-Glu substitution (Q798E) at one additional position in the same region (not shown).

Fig. 3.

H⁺ transport as a function of ATP hydrolysis in mutant ATPases that display partial uncoupling. Secretory vesicles expressing mutant ATPases were isolated as described in "Materials and Methods." To assay H⁺ transport, the initial rate of acridine orange fluorescence quenching was determined over a range of MgATP concentrations (0 to 2.0 mM) and plotted as a function of the rate of ATP hydrolysis measured under the same conditions. WT, wild type.

The most interesting results came from substitutions at Ser-800 and Glu-803, located three positions (or a turn of the helix) apart in the middle of M8. In the case of S800A, as previously reported for E803Q [29], an actual increase was seen in the slope of the quenching-vs-hydrolysis plot, corresponding to an apparent increase in H⁺/ATP stoichiometry (Fig. 4). Two other substitutions at Glu-803 (E803N and E803A) led to nearly complete uncoupling. As will be discussed in a later section, these findings gain significance given the alignment between Glu-803 and one of the known Ca²⁺ binding residues (Glu-908) of SERCA Ca²⁺-ATPase [3].

Fig. 4.

H⁺ transport as a function of ATP hydrolysis in mutant ATPases with an apparent increase in H⁺:ATP stoichiometry. See legend to Fig. 3 for details.

3.4. Expression and activity of the M8 mutant ATPases at the plasma membrane

To look further at the role of M8 in biogenesis and function, the four Ala mutations already shown to block delivery of the ATPase to secretory vesicles (see Table 1) were integrated into the chromosomal copy of the PMA1 gene. The goal of this experiment was to ask whether any of the mutant ATPases might fold properly and travel to the plasma membrane in the absence of the heat-shock step (23° to 37°) that is an integral part of the secretory vesicle expression system.

Two of the four alleles (Q798A, I799A) were unable to support growth at temperatures as low as 23°, suggesting that proper folding could not be achieved even under these conditions. The remaining two mutants (I794A and F796A) grew well at steady-state temperatures of 23° and even 30°, although the growth rate fell significantly at 37°, especially in the case of I794A (not shown). Plasma membranes isolated from cells grown at 30° had noticeably reduced amounts of ATPase compared with the wild-type control (34% and 72% in I794A and F796A, respectively); after correction for the expression level, ATPase activities were 41% of control in I794A and 37% in F796A (Table 2). As assayed by limited trypsinolysis, I794A ATPase appeared to be well folded, while F796A ATPase was slightly more sensitive to trypsin than the wild-type control (Fig. 5). Taken together, the data indicate that Q798A and I799A ATPases have profound folding defects, while I794A and F796A ATPases have borderline defects that become especially acute under the heat-shock conditions employed in the secretory vesicle expression system.

TABLE 2.

Effect of mutations in M8 on expression and activity of ATPase in plasma membranes

Strain	Expressiona	ATPase activityb
		Uncorrected		Corrected
		U/mg	%	%
WTc	100	7.22	100	100
I794A	35	0.94	14	39
F796A	72	1.66	24	34

^aThe specific expression of 100-kDa ATPase protein was calculated by quantitative immunoblotting as described in “Materials and Methods.” Data represent an average of 2 experiments except for the wild type, for which there were four determinations. Each mutant value was corrected for expression relative to a wild type control run on the same day.

^bATP hydrolysis was measured at pH 6.25 and expressed in units of μmol P_i per min per mg of protein as outlined in “Materials and Methods.” Data represent an average of 2 experiments except for the wild type, for which there were four determinations.

^cPlasma membranes were isolated from cells containing the wild-type PMA1 gene linked to URA3.

Fig. 5.

Trypsinolysis of I794A and F796A ATPases expressed at the plasma membrane. Plasma membranes were isolated at described in "Materials and Methods" and incubated at a trypsin/protein ratio of 1:25 for 0, 2, or 10 min at 30°C. After the reaction was stopped by the addition of 1 mM diisopropyl fluorophosphate, the membranes were subjected to SDS-polyacrylamide gel electrophoresis, and PMA1 protein was assayed by immunoblotting.

3.5. Effect of M8 mutations on growth at low pH

Since Pma1 H⁺-ATPase functions physiologically to pump protons out of the cell, it was of interest to ask whether the apparent changes in H⁺/ATP coupling observed in certain of the M8 mutants might alter the ability to grow at acidic pH. For this purpose, yeast strains expressing either an “undercoupled” or “overcoupled” form of the ATPase at the plasma membrane were tested for growth in liquid medium over a range of pH values. As shown in Figure 6, growth of the wild-type strain was progressively inhibited as the pH was lowered from 5.5 to 2.5. Expression of an overcoupled Pma1 ATPase (E803Q) was unable to counteract the inhibition, indicating that the improvement in pump efficiency seen under standard acridine orange quenching conditions (pH 6.7) did not have a measurable effect under the low pH conditions employed in the growth experiment. By contrast, the inhibition was significantly exacerbated in mutant strains expressing an undercoupled ATPase (e.g., I794A), consistent with the idea that a reduced efficiency of protein pumping becomes increasingly deleterious at low external pH values.

Fig. 6.

pH dependence of growth by M8 mutants. Wild-type, I794A, and E803Q were inoculated at 0.1 OD/ml into minimal medium containing 2% glucose and 50 mM acetic acid, adjusted to the appropriate pH with either NaOH or HCl. The cultures were grown at 30°C for 16 hours (late log phase), and cell densities were measured at OD₆₀₀. Control OD values at pH 5.5 were 6.17 for the wild type, 4.34 for I794A, and 5.44 for E803Q.

4. Discussion

Based on high-resolution crystal structures of SERCA Ca²⁺-ATPase [3, 7], it is now clear that the membrane domain of P₂-type ATPases moves through a carefully orchestrated series of conformational steps as cations are bound at one side of the membrane, transported, and released at the other side. Most of the movement takes place in six of the transmembrane helices (M1–M6), which bend (M1, M3, M5), unwind partially (M2), and even shift normal to the membrane (M1–M4). By comparison, transmembrane helices M7, M8, M9, and M10 are much less mobile and have been suggested to anchor the ATPase in the membrane [58]. Nevertheless, M8 contributes directly to one of the two Ca²⁺-binding sites in the SERCA ATPase [3, 5, 6], and its role must be taken into account in efforts to understand the structural basis for cation specificity and stoichiometry. With this aim in mind, and motivated specifically by earlier results on the yeast H⁺-ATPase [29], we have carried out a detailed structure-function study of M8 by means of alanine-scanning mutagenesis.

It soon became apparent that four of the 21 Ala mutations blocked the ability of newly synthesized Pma1 ATPase to travel from the endoplasmic reticulum to the secretory vesicles and thus that M8 must play a significant role in ATPase biogenesis. As shown in Fig. 1, the residues involved (Ile-794, Phe-796, Gln-798, and Ile-799) lie immediately downstream of the variable N-extracytoplasmic end of M8 (positions 791 through 793) and mark the beginning of a 17-amino acid stretch (positions 794 through 811) that has been well conserved during fungal evolution. In limited trypsinolysis experiments, Ala substitutions for these four residues could be seen to disrupt protein folding under conditions that normally permit trafficking through the secretory pathway (namely, expression of the mutant allele at 37°C under control of a heat-shock promoter; ref. 33). When the same mutations were integrated into the chromosomal copy of the PMA1 gene for expression at the plasma membrane, two of the four ATPases (Q798A and I799A) were unable to support growth, while the other two (I794A and F796A) were partially temperature-sensitive. Fig. 7 is a three-dimensional model of the membrane domain of Pma1 H⁺-ATPase, built using the E1Ca structure of SERCA1a Ca²⁺-ATPase as a template. In this particular view, which looks into the membrane domain from the extracytoplasmic surface, M8 is surrounded by five other transmembrane helices, with the M8 residues important for folding and biogenesis reaching out towards M5 (Ile-799), M6 (Phe-796), M7 (Ile-794 and Gln-798), and M10 (Gln-798). Thus, these four residues seem well positioned to play a structural role by contributing to the proper assembly of helices within the M domain. A prominent role for M8 in H⁺-ATPase biogenesis has previously been suggested by Lin and Addison [59], who constructed a set of fusion proteins containing carboxy-terminal fragments of the closely related Neurospora Pma1 H⁺-ATPase and then measured the ability of the fusion proteins to insert into microsomal membranes in vitro. In those experiments, M7 displayed a strong ability to initiate translocation into the membrane bilayer, and M8 displayed a similarly strong ability to stop translocation [59].

Fig. 7.

Homology model of the yeast H⁺-ATPase illustrating residues important for biogenesis and stability. The model was built using the E1Ca structure of SERCA1a Ca²⁺-ATPase as a template (see "Materials and Methods"). In this figure, which was prepared with PyMol version 0.97 [71], the membrane domain of the ATPase is viewed from the extracellular surface of the membrane. Numbers indicate transmembrane segments 1 to 10.

The remaining mutants that were examined in the present study were all capable of expression in the secretory vesicle system, and gave evidence for an important contribution of M8 to the proton transport pathway. Although Ala substitutions at most positions had little or no effect on the coupling between ATP hydrolysis and transport, substitutions at a cluster of four positions near the extracytoplasmic end of M8 (M791A, N792A, G793A, and M795A) and one position at the cytoplasmic end (R811A) led to partial uncoupling, and substitutions at two additional positions near the middle of M8 (S800A and E803Q) produced an apparent increase in the stoichiometry of H⁺ ions transported per ATP split (see Figs. 8 and 9 for structural models). The "uncoupling" mutations presumably distort the M domain in a way that blocks H⁺ transport either partially or completely, while leaving ATP hydrolysis free to continue at a reduced rate. By contrast, the "overcoupling" mutations seem to shift the ATPase from its customary state, in which one H⁺ ion is transported per ATP split, towards an altered state with a stoichiometry of 2 H⁺ ions per ATP. A similar change in the apparent stoichiometry of the Neurospora Pma1 H⁺-ATPase has been reported by Warncke and Slayman [60] based on current-voltage analysis in intact cells. In this case, the reversal potential of the ATPase moved from −400 mV in actively metabolizing cells to −200 mV in energy-restricted cells, consistent with a shift in stoichiometry from 1 H⁺/ATP to 2 H⁺/ATP. Work is currently under way to express the S800A and E803Q mutations under conditions suitable for electrophysiological studies (A. Rivetta, unpublished studies). In the meantime, it is worth noting that evidence has also appeared for mutational increases in pump efficiency in PMA2 ATPase of Nicotiana plumbaginifolia [61], decreases in pump efficiency in the F₁F₀ ATPase of Escherichia coli [62–66] and both increases and decreases in the vacuolar V-ATPase of yeast [67–70]. Further study of such mutants, although methodologically challenging, seems certain to yield useful insights into the fundamental mechanisms of proton transport.

Fig. 8.

Homology model of the yeast H⁺-ATPase showing residues at which mutations lead to partial uncoupling. See "Materials and Methods" and the legend to Fig. 7 for details on the model. In this figure, the membrane domain of the ATPase is viewed from the extracellular side.

Fig. 9.

Homology model of the yeast H⁺-ATPase showing residues at which mutations lead to an apparent increase in H⁺/ATP stoichiometry. See "Materials and Methods" and the legend to Fig. 7 for details on the model. In this figure, the membrane domain of the ATPase is viewed from the extracellular surface of the membrane.