A Conserved Asparagine in a P-type Proton Pump Is Required for Efficient Gating of Protons

Background: P-type proton pumps work against large electrochemical gradients, but how gating is controlled is poorly understood.

Results: Substitution of Asn-106 in transmembrane segment M2 significantly reduces the ability of the pump to perform uphill proton transport.

Conclusion: Asn-106 fulfills several criteria for being a gating residue in P-type proton pumps.

Significance: The presence of a gating residue provides a mechanism for preventing backflow of protons during pumping.

Abstract

The minimal proton pumping machinery of the Arabidopsis thaliana P-type plasma membrane H⁺-ATPase isoform 2 (AHA2) consists of an aspartate residue serving as key proton donor/acceptor (Asp-684) and an arginine residue controlling the pK_a of the aspartate. However, other important aspects of the proton transport mechanism such as gating, and the ability to occlude protons, are still unclear. An asparagine residue (Asn-106) in transmembrane segment 2 of AHA2 is conserved in all P-type plasma membrane H⁺-ATPases. In the crystal structure of the plant plasma membrane H⁺-ATPase, this residue is located in the putative ligand entrance pathway, in close proximity to the central proton donor/acceptor Asp-684. Substitution of Asn-106 resulted in mutant enzymes with significantly reduced ability to transport protons against a membrane potential. Sensitivity toward orthovanadate was increased when Asn-106 was substituted with an aspartate residue, but decreased in mutants with alanine, lysine, glutamine, or threonine replacement of Asn-106. The apparent proton affinity was decreased for all mutants, most likely due to a perturbation of the local environment of Asp-684. Altogether, our results demonstrate that Asn-106 is important for closure of the proton entrance pathway prior to proton translocation across the membrane.

Introduction

Proton pumps establish proton gradients across biological membranes that represent a form of energy that can be tapped to perform work such as secondary active transport across membranes, ATP synthesis in mitochondria and chloroplasts, and flagellar motions (1, 2). Despite progress in understanding the mechanism of transport of protons by biological pumps, it still remains unclear how protons are restricted from sliding backwards when transported against electrochemical gradients (3–5). Protons move extremely fast in water as protons tunnel from one water molecule to the next via hydrogen bonding between water molecules and, likewise, protons might cross biological membranes through networks of hydrogen bonds (6). The unique properties of protons caused Nagle and Morowitz (7) to predict that the path for protons in a hydrogen-bonded chain could involve water molecules within the protein to complete a proton relay system (8, 9). Irrespective of the nature of the proton transport pathway, the hydrogen bonded chain has to be efficiently interrupted at least once along the transport pathway in order to prevent backslipping of protons, which is a requirement for uphill transport of protons.

P-type ATPases are a family of cation transporters found in all eukaryotes that share the same overall structural fold (10–13). Prominent members of the P-type ATPase family are the well studied Na⁺,K⁺-ATPase and the sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA).³ P-type plasma membrane (PM) H⁺-ATPases are responsible for generating the essential electrochemical gradient across the plasma membrane of fungi and plants, and in these organisms they are the physiological equivalent of the Na⁺,K⁺-ATPase in animals (14). The PM H⁺-ATPase transports one proton per ATP hydrolyzed (15, 16) and pumps protons from the cytosol to the extracellular space apparently without any ion being counter-transported. Consequently, P-type PM H⁺-ATPases maintain high electrochemical gradients across the fungal and the plant plasma membranes (up to −300 mV and −200 mV, respectively (17, 18), much higher than the approximately −70 mV reported for the Na⁺,K⁺-ATPase).

Diverse proton transporters such as bacteriorhodopsin, the F_OF₁ ATP synthase, and the PM H⁺-ATPase appear to employ a common proton transport mechanism that involves a central aspartate residue serving as key proton donor/acceptor and an arginine residue controlling the pK_a of that aspartate residue as the minimal proton pumping apparatus (19). For the plant PM H⁺-ATPase, the central proton donor/acceptor has been shown to be the essential residue Asp-684, and, in addition, the nearby Arg-655 has been shown to be important, although not essential, for proton transport in this protein (20, 21). In addition to the involvement of these two residues and the recent identification of a proton release group (22), not much is known about how proton transport is achieved by PM H⁺-ATPases.

Recently, the first three-dimensional structure of a P-type H⁺-ATPase was published, showing the plant PM H⁺-ATPase in a quasi-occluded E1 conformation at a resolution of 3.6 Å (12). Although this structure has been an important step forward in the understanding of the PM H⁺-ATPase, the resolution is too low to unambiguously determine the orientation of amino acid side chains or the presence of water molecules and can thus only provide indications for the proton translocation mechanism of the PM H⁺-ATPase.

In the Na⁺,K⁺-ATPase, the cytosolic cation entrance pathway has been proposed to be closed by movement of transmembrane segment 1 (M1), thereby creating a hydrophobic plug that restricts the movement of the side chain of Glu-329, a residue in transmembrane segment 4 (M4) involved in liganding and gating (23, 24). However, the residues implicated in the occlusion mechanism of the Na⁺,K⁺-ATPase are not conserved in the PM H⁺-ATPase. Furthermore, the PM H⁺-ATPase faces other challenges such as the very high proton mobility. The structural basis for the higher efficiency of the PM H⁺-ATPase compared with the Na⁺,K⁺-ATPase is not known. In this study, we have investigated the role of an asparagine residue, Asn-106, in transmembrane segment 2 (M2) of the plant PM H⁺-ATPase, which is completely conserved in all PM H⁺-ATPases. Our results suggest that this residue is important for proton transport in P-type PM H⁺-ATPases and could serve to ensure closure of the proton entrance pathway.

MATERIALS AND METHODS

Construction of Mutants

The multicopy vector Yep-351 (25) containing a modified cDNA of the Arabidopsis thaliana AHA2 PM H⁺-ATPase isoform under the control of the PMA1 promoter (26) was used for the cloning of the Asn-106 mutants. The modified cDNA encodes the AHA2 PM H⁺-ATPase with a C-terminal deletion of 73 amino acid residues and the insertion of a C-terminal Met-Arg-Gly-Ser-His₆ (MRGSH₆) tag (20). The C-terminal deletion renders the PM H⁺-ATPase constitutively active, whereas the addition of the His₆ tag allows for affinity purification of the constructed mutants. Site-directed mutagenesis was performed by standard procedures using polymerase chain reaction, and all mutated sequences were verified by DNA sequencing.

Expression in Yeast

The Saccharomyces cerevisiae strain RS-72 (27) was transformed and cultured essentially as described previously (28). In RS-72 (MATa ade1–100 his4–519 leu2–3,112), the natural constitutive promoter of the yeast endogenous PM H⁺-ATPase PMA1 has been replaced by the galactose-dependent GAL1 promoter (27). As PMA1 is essential for yeast growth, RS-72 is only able to grow on galactose-containing medium. Using this strain, plasmid-borne plant PM H⁺-ATPases under the control of the constitutive PMA1 promoter can be tested for their ability to rescue pma1 mutants on glucose-containing medium. Plasma membrane enriched fractions from yeast expressing the wild type or the various Asn-106 mutants were purified as described previously (29).

Yeast Complementation Assay

Yeast was grown for 3 days at 30 °C in liquid medium containing 2% galactose. Approximately 10³ cells in 10 μl were spotted onto solid synthetic minimal medium containing either 2% galactose at pH 5.5 or 2% glucose at pH values of 6.5, 5.5, 4.5, and 3.5, respectively. Growth was recorded after incubation for 3 days at 30 °C.

Gel Electrophoresis and Western Blotting

Membrane fractions or purified proteins were separated by SDS-PAGE on 10% acrylamide. After electrotransfer of the proteins to a nitrocellulose membrane (Millipore), protein blots were probed with an antibody directed against Dpm1p, Pma1p, or the MRSGH₆ epitope present at the C terminus (Qiagen, Chatsworth, CA).

Protein Determination

Protein concentrations were determined by the method of Bradford (30) employing bovine serum albumin as a standard.

Purification and Reconstitution of ATPase

Growth of yeast for protein purification was the same as described (20). The transformed yeast cells were harvested, and the PM H⁺-ATPases were purified by membrane solubilization using n-dodecyl-β-d-maltoside and Ni²⁺-nitrilotriacetic acid affinity chromatography according to established procedures (12). Artificial liposomes were prepared from soybean lecithin or dioleoylphosphatidylcholine (DOPC) by extrusion through a 100 nm filter. Protein was reconstituted into the liposomes at a lipid to protein ratio (w/w) of 200:1 as described previously (31). Reconstitution was performed in a buffer containing 10 mm MES-KOH, pH 6.5, 50 mm K₂SO₄, and 20% (v/v) glycerol.

ATPase and Phosphorylation Assays

ATPase activity was determined as described previously (32) using purified protein in a reactivation buffer (20% (v/v) glycerol, 100 mm MES, pH 6.5, 1 mm EDTA, 1 mm dithiothretiol, 10 μg/μl soybean l-α-phosphatidylcholine (lecithin), 0.7% (w/v) n-dodecyl-β-d-maltoside). Phosphorylation by [γ-³²P]ATP was performed as described previously (20).

Measurement of Proton Transport

Proton transport into DOPC or lecithin liposomes was measured as described (33) by monitoring fluorescence quenching of 9-amino-6-chloro-2-methoxyacridine. The reaction medium contained 10 mm MES-KOH, pH 6.5, 50 mm K₂SO₄, 10% glycerol, 1 mm ATP, 1 μm 9-amino-6-chloro-2-methoxyacridine, and 125 nm valinomycin. Proton pumping reactions were started by the addition of MgSO₄ to a final concentration of 2 mm, and the proton gradient was collapsed by the addition of carbonyl cyanide-m-chlorophenylhydrazone to a final concentration of 125 nm.

Measurement of Membrane Potential

The buildup of membrane potential in reconstituted DOPC proteoliposomes was measured using oxonol VI (33). The reaction medium had a final volume of 3 ml and contained 10 mm Mes-KOH, pH 6.5, 50 mm K₂SO₄, 10% glycerol, 1 mm ATP, and 50 nm oxonol VI. The measurements were performed at 10 °C to allow the formation of a stable steady-state fluorescence level. After a 5-min pre-equilibration, reactions were initiated by the addition of MgSO₄ to a final concentration of 2 mm. Fluorescence was followed until steady state, and the membrane potential was collapsed by the addition of valinomycin to a final concentration of 125 nm.

RESULTS

Asn-106 Is a Candidate Gating Residue in P-type Proton Pumps

In the only available crystal structure of a P-type PM H⁺-ATPase, the 3.6 Å structure of the plant PM H⁺-ATPase AHA2 (12), it seems that Asn-106, located in M2, is within hydrogen bonding distance of the central proton donor/acceptor Asp-684 (Fig. 1A). Cavities likely to be filled with water are seen on both sides of the Asn-106-Asp-684 pair in the quasi-occluded E1 state. Asn-106 is completely conserved in all P-type H⁺-ATPases, suggesting an important function, but no mutagenesis data have been presented to support this assumption. To probe the role of Asn-106, this residue was replaced with alanine (short, hydrophobic residue), aspartate (negatively charged residue), lysine (positively charged residue), glutamine, or threonine (longer or shorter polar residue), giving rise to constructs N106A/N106D/N106K/N106Q/N106T, respectively. Mutations at Asn-106 were introduced in a truncated version of AHA2 lacking 73 C-terminal residues comprising most of its regulatory domain and thereby rendered constitutively active (aha2Δ73; termed wild type in this study for clarity) (20, 29).

FIGURE 1.

Asn-106 in AHA2 is located close to Asp-684 in the E1 state, and the corresponding residue in the SR Ca²⁺-ATPase, Asn-101, undergoes large displacements during the transition from the E1 to the E2 conformation. A, the crystal structure of AHA2 (PDB code 3B8C, green) is superimposed on the structure of SERCA in the Ca²⁺-bound E1 state with bound AMPPCP (PDB code 1T5S, magenta). Asn-101 and Asp-800 from SERCA, and Asn-106 and Asp-684 from AHA2 are shown as sticks. B, superposition of the SERCA in the Ca²⁺-bound E1 state with bound AMPPCP (PDB code 1T5S, magenta), the E2P state with bound BeF₃⁻ (PDB code 3B9B, cyan), and the proton bound E2 state with bound AMPPCP and thapsigargin (PDB code 2C8K, yellow). Asn-101 (*) and Asp-800 (#) are shown as sticks. The viewpoint is the same as in A.

Mutations at Asn-106 Abolish the Ability of the Plant PM H⁺-ATPase to Complement pma1

When situated in the plasma membrane of a living cell, the PM H⁺-ATPase faces the challenge of pumping protons against a steep electrochemical gradient involving both a pH gradient and a membrane potential (positive on the outside). A heterologous yeast complementation assay at different medium pH values was performed to test the activity of the Asn-106 mutants in vivo. We employed the yeast strain RS-72, in which the gene encoding the endogenous yeast PM H⁺-ATPase, PMA1, has been placed under the control of a galactose promoter and, as a consequence, the strain is only capable of growing on galactose-containing medium unless the pma1 deficiency is complemented (27). The ability of the plant PM H⁺-ATPase and mutants hereof to complement the function of Pma1p was tested on glucose-containing medium. Under these conditions, the wild type plant PM H⁺-ATPase readily complemented pma1. However, none of the Asn-106 mutants were able to support yeast growth in the pH range from pH 3.5 to 6.5 (Fig. 2). Both wild type and the Asn-106 mutants were found to co-localize with Pma1p in the plasma membrane enriched fraction, and similar expression levels were observed for the wild type and mutant proteins (Fig. 3A). This demonstrates that Asn-106 mutants cannot operate under in vivo conditions characterized by the presence of an electrochemical gradient.

FIGURE 2.

Mutagenesis of Asn-106 abolishes the ability of the PM H⁺-ATPase to complement pma1. Yeast complementation assay was performed for no PM H⁺-ATPase (empty vector), wild type PM H⁺-ATPase, and the various Asn-106 mutant PM H⁺-ATPases. In the yeast strain RS-72, the endogenous yeast PM H⁺-ATPase, Pma1p, has been placed under the control of a galactose promoter, whereas the introduced, plasmid-borne plant PM H⁺-ATPases are under the control of the constitutive PMA1 promoter (27). Yeast growth on glucose is therefore dependent on a functional plasmid-borne PM H⁺-ATPase. Transformed yeast cells, RS-72, were spotted on either galactose-containing medium (Gal) at pH 5.5 or glucose-containing media (Glu) at different pH values. Growth was recorded after 3 days.

FIGURE 3.

Heterologously expressed wild type and Asn-106 mutants of AHA2 are located in the yeast plasma membrane. A, total membranes (Total) from yeast expressing wild type or Asn-106 mutated PM H⁺-ATPases were subjected to sucrose step gradient fractionation. The proteins present in each fraction were determined by protein blot analysis. Membrane marker proteins are as follows: Dpm1p, endoplasmic reticulum; Pma1p, PM. B, Coomassie staining of affinity purified PM H⁺-ATPase proteins.

Expression of the Wild Type and the Asn-106 Mutant PM H⁺-ATPases

The inability of the mutants to support yeast growth prompted us to further characterize the kinetic properties of the pumps. Each mutant enzyme expressed in yeast strain RS-72 was purified to homogeneity (Fig. 3B). The ATPase kinetic properties of the wild type and the Asn-106 mutant enzymes were evaluated in a detergent solubilized and partially lipidated state to describe any qualitative changes in the Asn-106 mutant enzymes (Table 1). All of the enzymes showed ATP-dependent phosphate release with apparent ATP affinities in the range of 0.07 to 0.25 mm. Thus, the Asn-106 mutant proteins have retained the ability to bind and hydrolyze ATP.

TABLE 1.

Kinetic data of purified wild type AHA2 PM H⁺-ATPase and Asn-106 mutant enzymes

Values are presented as the mean of at least three independent experiments ± S.D. Asterisks denote significant difference from the wild type (two-tailed p < 0.05).

PM H+-ATPase	Specific activity	pH optimum	Apparent ATP affinity	Vanadate sensitivity (IC50)
	μmol Pi min−1 mg−1		μm	μm
Wild type	25.7 ± 1.8	7.1 ± 0.1	45 ± 5	8.7 ± 1.1
N106A	3.6 ± 1.1*	6.6 ± 0.1*	69 ± 7*	305 ± 69*
N106D	18 ± 9.6	6.5 ± 0.1*	255 ± 22*	2.6 ± 1.0*
N106K	1.7 ± 1.0*	6.4 ± 0.1*	155 ± 39*	2037 ± 201*
N106Q	3.9 ± 0.5*	6.6 ± 0.1*	65 ± 11*	162 ± 32*
N106T	3.6 ± 0.4*	6.4 ± 0.2*	102 ± 26*	332 ± 43*

ATP Affinity, pH Dependence, and Vanadate Sensitivity of the Asn-106 Mutants Indicate Shift in the Conformational Equilibrium of the Mutant Enzymes

In general, P-type ATPases are characterized by two major conformations, E1 and E2. The first state, E1, shows a high affinity for the exported ion (H⁺), and is phosphorylated by ATP to form E1P, which in turn is sensitive to dephosphorylation by ADP, and for the PM H⁺-ATPase, K⁺. The second state, E2, is characterized by a low affinity for the exported ion and high affinity for the imported ion, is insensitive to ADP, and is inhibited by the phosphate analog orthovanadate (34). The observed increase in apparent ATP affinity for the Asn-106 mutants (Table 1) could indicate a shift in the conformational equilibrium of the Asn-106 mutants. To investigate this, we analyzed the effect of proton concentration (pH) and vanadate on the catalytic activity of the mutant proteins.

Increasing the proton concentration has a stimulatory effect on PM H⁺-ATPase activity, similar to the effect observed for other P-type ATPases by increasing the concentration of the transported cation(s); therefore, the pH dependence of ATP hydrolysis by P-type H⁺-ATPases is, at least partially, assumed to reflect their substrate affinity (19). All of the Asn-106 mutants showed a shift in the pH dependence toward more acidic values, indicative of a lower proton affinity for the mutant enzymes when compared with the wild type (Fig. 4A, Table 1).

FIGURE 4.

Asn-106 mutant PM H⁺-ATPases show reduced affinity for protons and altered vandate sensitivity. A, representative example of the pH dependence of ATP hydrolysis by the wild type and the Asn-106 mutant PM H⁺-ATPases. The data shown are representative of three independent experiments. B, vanadate sensitivity of ATP hydrolysis by the wild type and the Asn-106 mutant PM H⁺-ATPases at pH 6.5. Values are presented ± S.D. Legends are the same as in A.

The sensitivity toward vanadate inhibition was reduced 20-fold or more for four of five mutant enzymes (N106A, N106K, N106Q, and N106T) compared with the wild type, with the N106K mutant being almost completely insensitive toward vanadate (Fig. 4B, Table 1). This indicates that these four mutants have difficulties in proceeding from the E1 to E2. In contrast, the N106D mutant enzyme displayed a significantly increased sensitivity toward vanadate, indicative of this mutant being arrested in the E2 conformation (Fig. 4B, Table 1). The shift of the N106D mutant enzyme toward the E2 state is also supported by its significantly reduced apparent ATP affinity and its lowered pH optimum.

The Asn-106 Mutants Are Arrested in Different Conformational States

Although the above data suggest that the N106D mutant PM H⁺-ATPase accumulates in the E2 conformation, they do not lend themselves toward any conclusion on the conformational state of the other Asn-106 mutants. To further investigate the conformational shift of the Asn-106 mutants, we studied the two phosphorylated conformational states, E1P and E2P, by measuring phosphorylation levels and sensitivity toward potassium. The amount of phosphorylated intermediate was measured under steady-state conditions for the wild type and the Asn-106 mutants (Fig. 5A). All of the mutants showed a shift in the steady-state phosphorylated intermediate levels (EP) compared with the wild type. The N106D mutant PM H⁺-ATPase accumulated significantly lower levels of phosphorylated intermediate (approximately one quarter) than the wild type, which supports the notion of this mutant being arrested in the E2 conformation. Interestingly, the other four mutants did not accumulate equal amounts of phosphorylated intermediate. Thus, the N106K and N106T mutant enzymes showed reduced EP levels compared with the wild type (approximately one-fourth) (Fig. 5A), which in combination with their decreased vanadate sensitivity could suggest that these mutants are stalled in the E1 position of the catalytic cycle. In contrast, the N106A and N106Q mutant PM H⁺-ATPases accumulated significantly higher EP levels than the wild type (∼2-fold) (Fig. 5A).

FIGURE 5.

Asn-106 mutant PM H⁺-ATPases are halted in different conformational states. A, measurement of steady-state phosphorylation levels at pH 6.5 of wild type and Asn-106 mutant PM H⁺-ATPases. Levels of steady-state phosphorylated intermediate (EP) are given relative to wild type (WT). B, ATP hydrolytic activity of the wild type and the Asn-106 mutants was measured without monovalent ions (black) or with the addition of 50 mm Li⁺ (dark gray), Na⁺ (light gray), or K⁺ (white). The activity without monovalent ions for each protein was set to 100%. The data presented are the mean of at least three independent measurements ± S.D.

Potassium ions allosterically stimulate the hydrolytic activity of the PM H⁺-ATPase by promoting dephosphorylation of the E1P conformation of the pump (35). To assess whether the phosphorylated intermediate observed for the N106A and N106Q mutants is primarily E1P or E2P, potassium was tested for its effect on the mutant enzymes. Hydrolytic activity was determined for the wild type and the Asn-106 mutant enzymes with and without the addition of 50 mm Li⁺, Na⁺, or K⁺, respectively (Fig. 5B). Under the experimental conditions, the strongest effect was seen upon addition of K⁺ to either wild type, N106A, or N106Q; neither Li⁺ nor Na⁺ showed any significant effect on the activity. The stimulation of activity observed for the N106A and N106Q mutant enzymes indicates that these two mutants accumulate in the E1P state.

A slight but significant decrease in the hydrolytic activity was observed for both the N106K and N106T mutant enzymes in the presence of Na⁺ and for the N106K mutant in the presence of K⁺ as well. This decrease is most likely a consequence of the E1 conformational state of these mutant enzymes in which it is possible for the monovalent cations to disrupt the proton transfer reactions in the cytosolic entrance pathway (36).

Asn-106 Mutants Transport Protons at Reduced Rates

To test whether the Asn-106 mutants were capable of transporting protons, the wild type and the mutant proteins were reconstituted into artificial liposomes and proton transport into the proteoliposomes was measured using the ΔpH probe 9-amino-6-chloro-2-methoxyacridine. All of the Asn-106 mutants were capable of performing proton transport into soybean lecithin or DOPC liposomes, albeit at reduced rates compared with the wild type (Table 2). The observed proton transport rates in lecithin vesicles were found to correlate with the ATPase activity for all Asn-106 mutants (Table 1). When proton pumping was tested in the absence of the ionophore valinomycin, thus allowing for the buildup of a membrane potential, all of the PM H⁺-ATPase proteins showed reduced activity (Fig. 6A, Table 2). This indicates that the establishment of a membrane potential slows the PM H⁺-ATPase in general and the Asn-106 mutants in particular. This phenomenon was especially noticeable in DOPC liposomes, where proton pumping above the background could only be detected for the wild type.

TABLE 2.

Proton pumping efficiency of purified wild type AHA2 PM H⁺-ATPase and Asn-106 mutant enzymes reconstitututed into liposomes composed of either soybean lecithin or DOPC

Values are presented as the mean of at least three independent experiments ± S.D. Numbers in parentheses represent the observed rates expressed as decrease in fluorescence per second. Asterisks denote significant difference from the wild type (two-tailed p < 0.05).

PM-H+-ATPase	Proton pumping (% of wild type)
	Lecithin		DOPC
	+Valinomycin	−Valinomycin	+Valinomycin	−Valinomycin
Wild type	100 ± 12 (1.8% s−1)	100 ± 8 (0.9% s−1)	100 ± 43 (0.54% s−1)	100 ± 41 (0.04% s−1)
N106A	12 ± 7*	5 ± 3*	2 ± 1*	NDa
N106D	131 ± 21	74 ± 21	27 ± 1*	NDa
N106K	3 ± 1*	2 ± 1*	NDa	NDa
N106Q	12 ± 6*	8 ± 5*	13 ± 3*	NDa
N106T	22 ± 14*	13 ± 5*	33 ± 8*	NDa

^a ND indicates not detected. No proton pumping significantly above the level of empty liposomes could be measured.

FIGURE 6.

Proton transport by the Asn-106 mutant PM H⁺-ATPases is inhibited by the formation of a membrane potential. A, proton transport was initiated by the addition of MgSO₄ to reconstituted lecithin proteoliposomes containing either wild type PM H⁺-ATPase (blue) or the mutant enzymes, N106D (green). Proton pumping was assayed both in the presence (dark colors) and absence (light colors) of 125 nm valinomycin. The decrease in 9-amino-6-chloro-2-methoxyacridine (ACMA) fluorescence reflecting proton accumulation in the vesicle lumen was abolished by the addition of the protonophore carbonyl cyanide-m-chlorophenylhydrazone (CCCP) to 125 nm. The experiment shown is representative of three to four independent experiments. B, proton transport measured as above for wild type (blue), N106A (pink), or N106D (green) PM H⁺-ATPases reconstituted into DOPC vesicles in the presence of 125 nm valinomycin (val). The experiment shown is representative for three to four independent experiments.

In lecithin liposomes, the N106D mutant had a higher initial proton transport activity than the wild type in the presence of valinomycin, consistent with its pH optimum and wild type-like maximal activity. However, the rate of proton pumping for the N106D mutant declined more rapidly than the rate observed for the wild type (Fig. 6A), suggesting that this mutant is strongly inhibited by the electrochemical gradient. When reconstituted in DOPC vesicles, the N106D mutant displayed a lower rate of proton pumping compared with the wild type (Fig. 6B), and this rate of proton pumping was coupled to a lower ATPase activity in the DOPC liposomes (data not shown). This could suggest that DOPC liposomes are tighter toward protons and therefore build up higher electrochemical gradients than lecithin liposomes.

Asn-106 Is Important for Membrane Potential Formation

The wild type pump was assayed for its ability to generate a membrane potential in vitro using the probe oxonol VI. Membrane potential formation could successfully be measured in DOPC proteoliposomes (Fig. 7A) but not in lecithin proteoliposomes. Membrane potential formation was fast, and a maximum value was approached in less than a minute. Asn-106 mutant enzymes could all establish membrane potentials (Fig. 7B) but were much slower at establishing the steady-state level (Fig. 7, A and C). The rates of membrane potential formation were comparable with the proton transport rates measured in DOPC vesicles in the presence of valinomycin (Table 2).

FIGURE 7.

Asn-106 mutant PM H⁺-ATPases show delayed formation of steady-state membrane potential. A, typical membrane potential measurements using oxonol VI for the wild type or Asn-106 mutants reconstituted into DOPC vesicles. Proton pumping was initiated by addition of MgSO₄. After a steady state was reached, valinomycin was added to collapse the membrane potential. The experiment shown is representative for several independent experiments. a.u., arbitrary units. B, peak fluorescence obtained by the Asn-106 mutant enzymes in A, relative to the level in the wild type. Results shown are the mean of three to four independent experiments ± S.D. C, the time dependence, given as t½, of the wild type and the Asn-106 mutants. t½ represents the time required to achieve 50% of the fluorescence steady-state level. Results shown are the mean of three to four independent experiments ± S.D.

DISCUSSION

Asn-106 Is Important for Uphill Proton Translocation

Asn-106 is completely conserved in all P-type PM H⁺-ATPases but no mutational study involving this residue has previously been reported. The data presented here suggest an important role for Asn-106 in uphill proton transport.

Substituting Asn-106 resulted in a marked reduction in the ability of the modified pumps to transport protons except for the N106D mutant (Table 2). We noted that not only was the initial proton transport rate affected, but the transport rate seemed to decline more rapidly for the Asn-106 mutants than observed for the wild type (Fig. 6A). Removal of valinomycin from the assay, thus allowing for the build-up of a membrane potential, had a more pronounced effect on the Asn-106 mutants than the wild type, and together, these data indicate that the formation of an electrochemical gradient across the membrane affects the Asn-106 mutants to a higher extent than the wild type. The inability of the mutants to pump protons against a potential gradient is consistent with their inability to complement the function of the yeast PM H⁺-ATPase in vivo.

Asn-106 Is a Gatekeeping Residue in the Cytosolic Proton Entrance Pathway

All Asn-106 mutant enzymes showed a shift in their pH optimum, a property that in proton pumps is correlated with substrate affinity. Thus, the reduced ability of the Asn-106 mutant PM H⁺-ATPases to pump protons could be the result of an interruption of the proton binding event, either by decreasing the affinity of the site for the proton itself or by allowing a greater degree of diffusion out of the site. If the binding event itself was disrupted, we would expect the pump to be stalled in the same conformational state for all mutants, namely in the E1-E1P transition, which is coupled to the binding event. However, our results show that the Asn-106 mutant PM H⁺-ATPases are blocked in different catalytic steps of the reaction cycle: in the reversion of the E2 state back to E1 (N106D), in the phosphorylation step associated with binding and occlusion (N106K&N106T), or in the transition to the E2P state (N106A/N106Q). This strongly supports the notion that Asn-106 is implicated in events that are associated with major conformational rearrangements such as substrate binding and occlusion.

In SERCA, the residue corresponding to Asn-106 is Asn-101, and during the catalytic cycle, the position of this asparagine residue undergoes large displacements (Fig. 1B). P-type ATPases show a high degree of overall structural similarity, and therefore, it seems reasonable to assume that Asn-106 in the PM H⁺-ATPase will be subjected to similar movements as Asn-101 in SERCA during the catalytic cycle. According to this assumption, the conformational changes accompanying the E2-E1 and the E1-E1P transitions would shift Asn-106 into the proximity of Asp-684, and it can be speculated that hydrogen bonding between Asn-106 and the protonated Asp-684 would lead to occlusion of the binding site and lock the bound proton to Asp-684. Mutagenesis of Asn-106 would thus perturb the local environment of Asp-684, thereby explaining the observed decrease in apparent proton affinity, and consequently the apparent ATP affinity as well, for all the Asn-106 mutants. Interestingly, mutagenesis of Asn-106 only affects the conformational steps that bring Asn-106 into the vicinity of Asp-684, supporting the hypothesis that Asn-106 acts as a gatekeeper in cooperation with Asp-684 to ensure efficient vectorial transport of protons in the PM H⁺-ATPase.

A Common Theme for Gating in Proton Pumps?

Proton transport is inherently quite different from that of other cations, which could suggest that specific mechanistic features must be present in proton transporters. In other proton transporters such as the CLCec Cl⁻/H⁺ exchanger and cytochrome c oxidase, gating has been suggested to be dependent on only a single amino acid residue. Thus, in the CLC-ec1 Cl⁻/H⁺ exchanger, gating of the proton pathway is achieved by the rotation of the side chain of a single glutamate residue, located at the beginning of the pathway (37, 38). The uptake of protons in cytochrome c oxidase occurs through an entrance pathway named the D pathway, a 25 Å long cavity, which consists of a network of polar residues and at least 11 crystallographically identified water molecules. However, in all the available crystallographic data, the pathway is interrupted shortly after the entrance by an asparagine residue, and molecular modeling has suggested that this asparagine residue serves as a gate that controls the polarity of the water chain within the D pathway through rotation of its side chain (39). Mutagenesis of this asparagine residue to an aspartate, a cysteine, or a threonine residue leads to a complete abolishment of proton pumping while retaining the oxidase activity and proton uptake for the chemical reaction, whereas mutagenesis of the asparagine residue to an alanine or a valine abolishes proton pumping and reduces the oxidase activity (40–43).

In the proton translocating F₁F_O ATP synthase, a completely conserved asparagine residue is located in M4 of the a subunit, and mutagenesis of this residue results in only small reductions in the hydrolytic activity, whereas proton transport is reduced to 5–50% (44). Interestingly, the modeling of a stretch of M4 from the a subunit with the c ring structure was recently performed for the Na⁺-transporting F₁F_O ATP synthase from Ilyobacter tartaricus, which showed that the conserved asparagine residue is located at the periphery of the binding site in correlation with cross-linking studies in both I. tartaricus and the proton translocating F₁F_O ATP synthase from Escherichia coli (45, 46). Furthermore, cross-linking studies showed that the protonation status of the stator arginine significantly influences the cross-linking pattern of this asparagine residue (45). Recently, chemical reactivity studies have indicated that this asparagine residue is accessible from the periplasmic half-channel, and it was suggested that the asparagine could be involved in gating H⁺ access from the periplasmic half-channel to the binding site (47).

In conclusion, we here provide evidence that Asn-106 functions as a gatekeeper residue in the PM H⁺-ATPase. This gating mechanism is distinct from that found for the SERCA Ca²⁺-ATPase. However, because of a similar placement of a conserved asparagine in a number of structurally divergent proton pumps, we speculate that gating by an asparagine residue is a common theme among proton pumps in nature.