Functional expression of gastric H+,K(+)-ATPase and site-directed mutagenesis of the putative cation binding site and catalytic center.

Gastric H+,K+-ATPase was functionally expressed in the human kidney HEK293 cell line. The expressed enzyme catalyzed ouabain-resistant K+-dependent ATP hydrolysis. The K+-ATPase activity was inhibited by SCH 28080, a specific inhibitor of gastric proton pump, in a dose-dependent manner. By using this functional expression system in combination with site-directed mutagenesis, we investigated effects of mutations in the putative cation binding site and the catalytic center of the gastric H+,K+-ATPase. In Na+,K+-ATPase, the glutamic acid residue in the 4th transmembrane segment is regarded as one of the residues responsible for the K+-induced conformational change (Kuntzweiler, T. A., Wallick, E. T., Johnson, C. L., and Lingrel, J. B.(1995) J. Biol. Chem. 270, 2993-3000). When the corresponding glutamic acid (Glu-345) of H+,K+-ATPase was mutated to aspartic acid, lysine, or valine, the SCH 28080-sensitive K+-ATPase activity was abolished. However, when this residue was replaced by glutamine, about 50% of the activity was retained. This mutant showed a 10-fold lower affinity for K+ (Km = 2.6 mM) compared with the wild-type enzyme (Km = 0.24 mM). Thus, Glu-345 is important in determining the K+ affinity of H+,K+-ATPase. When the aspartic acid residue in the phosphorylation site was mutated to glutamic acid, this mutant showed no SCH 28080-sensitive K+-ATPase activity. Thus, amino acid replacement of the phosphorylation site is not tolerated and a stringent structure appears to be required for enzyme activity. When the lysine residue in the fluorescein isothiocyanate binding site (part of ATP binding site) was mutated to arginine, asparagine, or glutamic acid, the SCH 28080-sensitive K+-ATPase activity was eliminated. However, the mutant in which this residue was changed to glutamine had about 30% of the activity, suggesting that amino acid replacement of this site is tolerated to a certain extent.

H ϩ ,K ϩ -ATPase is the proton pump responsible for gastric acid secretion (1,2). It consists of ␣and ␤-subunits. The ␣-subunit is the catalytic subunit with a molecular mass of 114 kDa (3) and contains the phosphorylation site, the ATP binding site, and the binding sites for proton pump inhibitors (4 -7). The ␤-subunit is a glycoprotein with a molecular mass of 60 -80 kDa (8). One of the roles of the ␤-subunit is to stabilize the ␣-subunit in the membrane. Although the cDNAs of both subunits of many species were cloned, there have been no reports of structure-function studies using site-directed mutagenesis because there has been no effective functional expression system. Here we report the functional expression of rabbit gastric H ϩ ,K ϩ -ATPase in human HEK293 cells. When the cells were co-transfected with the cDNAs of the ␣and ␤-subunits, ouabain-resistant K ϩ -dependent ATPase activity was observed. The activity was inhibited by SCH 28080 and scopadulcic acid B, specific inhibitors of the gastric H ϩ ,K ϩ -ATPase (9,10). By using this functional expression system, we investigated the role of amino acid residues of the putative cation binding site and the catalytic center.
H ϩ ,K ϩ -ATPase is a member of the P-type ATPase family. Sarcoplasmic and endoplasmic reticulum Ca 2ϩ -ATPases and Na ϩ ,K ϩ -ATPase also belong to the same family. They actively transport the ions coupled with the hydrolysis of ATP. It has been considered that P-type ATPases have the common structures in the catalytic center including the phosphorylation site and the ATP binding site. On the other hand, their cation recognition sites and transport pathways are hypothesized to be common to some extent, but divergent depending on the species of transporting cations. From the site-directed mutation and chemical labeling experiments, Glu-327 in the ␣-subunit of Na ϩ ,K ϩ -ATPase (sheep ␣-1) has been recognized as one of the pivotal residues for cation-induced conformational changes or for K ϩ occlusion (11)(12)(13). The replacement of this residue by glutamine partly reduced the affinity of the enzyme for Na ϩ and K ϩ (14). Glu-309 in sarcoplasmic Ca 2ϩ -ATPase (the counterpart of Glu-327 of Na ϩ ,K ϩ -ATPase) has been suggested to be responsible for Ca 2ϩ high affinity binding. The replacement of this residue by glutamine completely eliminated the Ca 2ϩ transport activity and the Ca 2ϩ sensitivity in the phosphorylation reaction (15). Here we mutated the corresponding residue (Glu-345) of the H ϩ ,K ϩ -ATPase ␣-subunit and compared the property of the mutant with those of Na ϩ ,K ϩ -ATPase and Ca 2ϩ -ATPase.
The sequences around the phosphorylation site and the FITC 1 binding site are well conserved in some of the P-type ATPases (16). In Na ϩ ,K ϩ -ATPase and Ca 2ϩ -ATPase, amino acid replacement in the phosphorylation site is not tolerated (17,18). In Ca 2ϩ -ATPase, amino acid replacement at the FITC binding site is tolerated, and the structure is able to withstand basic amino acids, but not a negatively charged amino acid (18). In the present paper, we also replaced Asp-387 of the phosphorylation site and Lys-519 of the FITC binding site of the H ϩ ,K ϩ -ATPase ␣-subunit, measured the enzyme activity of the * This study was supported in part by a grant-in-aid for encouragement of young scientists (to S. A.), and scientific research on priority areas (to N. T.) from the Ministry of Education, Science and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ To whom correspondence should be addressed. Fax: 81-764-34-5051. mutants, and compared the effects of these mutations with those of Na ϩ ,K ϩ -ATPase and Ca 2ϩ -ATPase.

EXPERIMENTAL PROCEDURES
Materials-HEK293 cells (human embryonic kidney cell line) were a kind gift from Dr. Jonathan Lytton (Brigham & Women's Hospital, Harvard Medical School, Boston, MA). pcDNA3 vector was purchased from Invitrogen (San Diego, CA). Restriction enzymes and other DNA and RNA modifying enzymes were from Toyobo (Osaka, Japan), New England Biolabs, Life Technologies, Inc., or Pharmacia Biotech Inc. (Tokyo, Japan). RNAzol was purchased from Biotecx Laboratories (Houston, TX), Oligotex-dT30 and MutanK kit were from Takara (Ohtsu, Japan), and ZAP-cDNA synthesis kit and Gigapack II Gold lambda packaging extract were from Stratagene. Block Ace was purchased from Dainippon Pharmaceutical Co. (Osaka, Japan). All other reagents were of molecular biology grade or the highest grade of purity available.
Preparation of cDNA Library-Total RNA was prepared from gastric mucosa of a male Japanese white rabbit using RNAzol. Poly(A) RNA was isolated using Oligotex-dT30. A gastric cDNA library was prepared using the ZAP-cDNA synthesis kit and Gigapack II Gold lambda packaging extract according to their instruction manuals.
cDNA Cloning of ␣and ␤-Subunits of H ϩ ,K ϩ -ATPase-cDNA was synthesized from the total RNA from rabbit gastric mucosa using Molony murine leukemia virus reverse transcriptase, primed with oligo(dT). The ␣-subunit cDNA between nucleotides 829 and 1412 (counted from initiation ATG as position 1), and the ␤-subunit cDNA between nucleotides 320 and 649 were amplified by PCR. PCR primers for the ␣-subunit were 5Ј-ATCATCGGGCGCATCGCCTC-3Ј and 5Ј-GAGAA-CTTGAGCAGCGCCGT-3Ј. PCR primers for the ␤-subunit were 5Ј-CCGTCGACCCAGCCTCACGCACACCCTGA-3Ј and 5Ј-CCGGATCC-CGACCTGCAGCGGCGTGAGG-3Ј. The sequences of these primers were from the cDNA sequences of the ␣and ␤-subunits of H ϩ ,K ϩ -ATPase (19,20) except that the ␤-subunit primers contained SalI and BamHI restriction sites (underlined). The 583-base pair fragment of the ␣-subunit cDNA and the 329-base pair fragment of the ␤-subunit cDNA were purified on an agarose gel, labeled with [ 32 P]dCTP, and used for screening the gastric cDNA library. The screening of the cDNA library was carried out as described below. Duplicate nitrocellulose filters (Hybond-C, Amersham) were prehybridized for 2 h at 41°C in 1 M NaCl, 5 mM EDTA, 4 mM sodium phosphate, 0.1% SDS, 5 [tomes] Denhardt's, 50% formamide, 100 g/ml salmon sperm DNA, 50 mM Tris-HCl (pH 8.0). Hybridization was carried out in the same solution supplemented with the 32 P-labeled probe at 41°C overnight. After washing under high stringency conditions, the filters were dried and exposed on an x-ray film. The positive plaques were screened by autoradiography. cDNAs in pBluescript SK(Ϫ) vector were prepared by in vivo excision from Uni-ZAP XR as described in the instruction manuals.
Removal of 5Ј-Noncoding Sequence of the ␣-Subunit cDNA-The ␣-subunit cDNA between nucleotides Ϫ28 and 495 was amplified by PCR. The PCR primers were 5Ј-CCGAATTCAAGGAGGGCAGCG-CAGCGAG-3Ј and 5Ј-GCCTCGAGGCTCTTGAACTCCTGATAGTAGC-3Ј. The 540-base pair fragment was purified on a gel and digested with EcoRI and BstEII. The cDNA cassette between EcoRI and BstEII of the ␣-subunit cDNA construct was replaced by the PCR-derived fragment in order to remove the cDNA sequence in its 5Ј-noncoding region (nucleotides Ϫ68 to Ϫ29). This truncated cDNA was used as the cDNA construct for the ␣-subunit.
Constructs of the ␣and ␤-Subunit cDNAs in Mammalian Expression Vector pcDNA3-The ␣and ␤-subunit cDNAs were digested with EcoRI and XhoI. The obtained fragments were each ligated into pcDNA3 vector treated with EcoRI and XhoI.
DNA Sequencing-DNA sequencing was done by the dideoxy chain termination method using an Autoread DNA sequencing kit and an ALF-II DNA Sequencer (Pharmacia Biotech Inc.).
Site-directed Mutagenesis-Site-directed mutagenesis was carried out by the method of Kunkel (21) using a MutanK kit. Synthetic oligonucleotides, 21-23 bases long and containing one or two mutated bases near the center, were hybridized with the uridine-containing single strand template of the construct of the H ϩ ,K ϩ -ATPase ␣-subunit. After sequencing, the appropriate fragment of mutant ␣-subunit cDNA was excised and ligated back into the relevant position of the wild-type construct of the ␣-subunit.
Cell Culture and Transfection-HEK293 cells were maintained in a humidified incubator at 37°C under 5% CO 2 atmosphere in Dulbecco's modified Eagle medium (high glucose) (Life Technologies, Inc.) containing 2 mM L-glutamine, 100 M minimum essential medium nonessential amino acids, 10% fetal calf serum, penicillin G (100 units/ml), and streptomycin (100 g/ml). DNA transfection was performed by the calcium phosphate method with 10 g of cesium chloride-purified DNA per 10-cm dish. Cells were harvested 2 days after the DNA transfection.
Preparation of Membrane Fractions from HEK293 Cells-Membrane fractions from the transfected cells and control (mock-transfected) cells were prepared from cells harvested from five to twenty 10-cm Petri dishes. Cells were washed with 5 ml of PBS, scraped, and suspended in PBS containing 5 mM EDTA. After washing twice with PBS, cells were incubated in 2 ml of low ionic salt buffer (0.5 mM MgCl 2 , 10 mM Tris-HCl, pH 7.4) at 0°C for 10 min. Phenylmethylsulfonyl fluoride (1 mM) and aprotinin (0.09 unit/ml) were added to the suspension. The cells were homogenized with 25 strokes in a Dounce homogenizer, and the homogenate was diluted with an equal volume of a solution containing 500 mM sucrose and 10 mM Tris-HCl, pH 7.4. The cell suspension was homogenized with 25 more strokes. The homogenized suspension was centrifuged at 800 ϫ g for 10 min, and the supernatant was centrifuged at 100,000 ϫ g for 90 min, and the pellet was suspended in a solution containing 250 mM sucrose and 5 mM Tris-HCl, pH 7.4.
Gastric Vesicles-Gastric vesicles enriched in H ϩ ,K ϩ -ATPase were prepared from mucosa in the fundic region of hog stomachs by differential and density gradient centrifugation as described elsewhere (22).
SDS-Polyacrylamide Gel Electrophoresis-SDS-polyacrylamide gel electrophoresis was carried out as described elsewhere (23). Membrane preparations (30 g of protein) or gastric vesicles (0.5 or 1 g of protein) were incubated in a sample buffer containing 2% SDS, 2% ␤-mercaptoethanol, 10% glycerol, and 10 mM Tris-HCl, pH 6.8, at room temperature for 2 min and applied to the SDS-polyacrylamide gel.
Immunoblot-Proteins were blotted onto nitrocellulose filters in a transfer solution containing 20% methanol, 192 mM glycine, and 25 mM Tris, pH 8.3, as described elsewhere (24). The filters were soaked in 4% Block Ace solution for 20 min at room temperature. They were incubated for 2 h with the primary antibody diluted with 0.8% Block Ace solution. The blots were washed with Tris-buffered saline containing 0.1% Tween 20 and incubated with peroxidase-conjugated anti-rabbit IgG for 1 h. Finally, the blots were stained with Peroxidase Immunostain kit from Wako Pure Chemicals (Osaka, Japan).
Assay of H ϩ ,K ϩ -ATPase Activity-The H ϩ ,K ϩ -ATPase activity was measured in 1 ml of a solution containing 50 g of membrane protein, 40 mM Tris-HCl, pH 7.4, 3 mM MgCl 2 , 3 mM ATP, 5 mM NaN 3 , 1 mM ouabain, and 15 mM KCl. After the reaction at 37°C for 30 min, the inorganic phosphate released was measured as described elsewhere (26). The K ϩ -ATPase activity was calculated as the difference between the activities in the presence and absence of KCl. Furthermore, the SCH 28080-sensitive K ϩ -ATPase activity was calculated as the difference between the K ϩ -ATPase activities in the presence and absence of 50 M SCH 28080.
Protein was measured using the BCA Protein Assay Kit from Pierce with bovine serum albumin as a standard.

RESULTS
Expression of the ␣and ␤-Subunits of H ϩ ,K ϩ -ATPase-We prepared the cDNAs encoding the ␣and ␤-subunits of rabbit gastric H ϩ ,K ϩ -ATPase according to the cDNA sequences presented elsewhere (19,20) with the exception that some part of the 5Ј-untranslated region (nucleotides Ϫ68 to Ϫ29) of the ␣-cDNA was removed as described under "Experimental Procedures." These cDNAs were introduced in a mammalian expression vector, pcDNA3, which contained a strong cytomegalovirus promoter. Fig. 1 shows the immunoblot of the crude membrane fractions from HEK293 cells; mock-transfected (lane 1) and transfected with the cDNA constructs of ␣-(lane 2), ␤-subunit (lane 3) or ␣plus ␤-subunits (lane 4). The blots were detected by using an anti-␣ subunit antibody. A hog gastric vesicle preparation was used as a positive control (lanes 5 and 6). In the mock-transfected cells and cells transfected with the ␤-subunit cDNA, no band was detected. When the cells were transfected with the ␣-subunit cDNA, a single weak band was detected around 95 kDa. When the cells were co-transfected with the ␣and ␤-subunit cDNAs, a significantly denser band was observed. The ␣-subunit expression on the immunoblot from 30 g of the crude membrane fraction of the expressed cells appears to be comparable to that of 0.5 g of the gastric vesicle preparation.
H ϩ ,K ϩ -ATPase Activity of the Membrane Fractions-The K ϩstimulated ATPase activity was measured in the membrane fractions from the cells transfected with the ␣-subunit, ␤-subunit, or ␣and ␤-subunit cDNAs. 1 mM ouabain and 5 mM NaN 3 were added to the reaction mixture to inhibit endogenous Na ϩ ,K ϩ -ATPase and mitochondrial ATPase, respectively. These inhibitors did not affect the gastric H ϩ ,K ϩ -ATPase activity (data not shown). A significant SCH 28080-sensitive K ϩ -ATPase activity was detected only in the cells transfected with both the ␣and ␤-subunit cDNAs (Table I). This activity did not increase in the presence of the monovalent cation ionophore, gramicidin, indicating that the membranes were leaky. The expressed K ϩ -ATPase was sensitive to SCH 28080, an inhibitor specific for the gastric H ϩ ,K ϩ -ATPase. Fig. 2 shows the effect of SCH 28080 on the K ϩ -ATPase activity of the membrane fraction from the ␣␤-transfected cells. SCH 28080 inhibited the K ϩ -ATPase activity in a dose-dependent manner, with an IC 50 value of 2.1 M. The inhibitory effect of SCH 28080 on the expressed H ϩ ,K ϩ -ATPase was weaker than its effect on the hog gastric vesicles, for which the IC 50 value of 0.87 M was reported (27). SCH 28080 is an inhibitor competitive with luminal K ϩ . It is protonated in an acidic compartment (28). The difference in potency of SCH 28080 between our system and the gastric vesicles may be due to the facts that the K ϩ and H ϩ permeabilities of the HEK membrane were higher than those of gastric vesicles and that SCH 28080 was not well protonated. In fact, the IC 50 value of SCH 28080 in leaky gastric vesicles at pH 7.5 was reported to be 1.5 M (28). The K ϩ -ATPase activity was also sensitive to scopadulcic acid B, a proton pump inhibitor found in a Paraguayan medicinal herb (10). Scopadulcic acid B at 100 M inhibited 40% of the K ϩ -ATPase activity. Fig. 3 shows the effect of K ϩ concentration on the expressed SCH 28080-sensitive K ϩ -ATPase activity. Low concentrations of K ϩ (less than 3 mM) stimulated the ATPase activity, while high concentrations of K ϩ (more than 10 mM) were inhibitory. The K m value for K ϩ obtained from the least-squares curvefitting in the range of the low K ϩ concentrations was 0.24 mM, which is in agreement with the value obtained with gastric vesicles (29,30).
Proton transport activity in the membrane fraction was measured using acridine orange fluorescence; however, no significant quenching of the fluorescence was observed.
Hereafter, we studied the significance of the putative functional sites on the ␣-subunit of H ϩ ,K ϩ -ATPase: 1) putative cation binding site, 2) phosphorylation site, and 3) FITC binding site (putative ATP binding site) by using the present functional expression system.
Site-directed Mutations of the Putative Cation Binding Site-In Na ϩ ,K ϩ -ATPase, the glutamic acid residue in the 4th transmembrane segment of the ␣-subunit (Glu-329 in rat ␣1, Glu-327 in rat ␣2 and sheep ␣1) was reported to be involved in determining the affinity for Na ϩ and K ϩ and the cation-induced conformational changes (12,14,31). The mutant in which this glutamic acid was mutated to glutamine or leucine showed lower affinity for Na ϩ and K ϩ than wild-type Na ϩ ,K ϩ -ATPase (14,31). In sarcoplasmic Ca 2ϩ -ATPase, the corresponding glutamic acid residue (Glu-309) was reported to be involved in high affinity binding of Ca 2ϩ . This residue is critically important for the function and unalterable (15). To study whether Glu-345 of the ␣-subunit of the H ϩ ,K ϩ -ATPase is essential for enzyme activity, we mutated this residue to aspartic acid, lysine, glutamine, or valine and measured the SCH  1-4) and 0.5 and 1 g of gastric vesicles (lanes 5 and 6) were applied to the gel. Lanes: 1, mock-transfected cells; 2, cells transfected with the ␣-subunit cDNA; 3, the ␤-subunit cDNA; 4, ␣and ␤-subunit cDNAs; 5, gastric vesicles (0.5 g); 6, gastric vesicles (1 g).

TABLE I SCH 28080-sensitive K ϩ -ATPase activities of the HEK cells
transfected with ␣-, ␤-, or (␣ ϩ ␤)-subunit cDNAs SCH 28080-sensitive K ϩ -ATPase activity was measured in the presence of 1 mM ouabain and 5 mM NaN 3  28080-sensitive K ϩ -ATPase activity. These mutants were expressed almost in the same amount as the wild-type ␣-subunit, judging from the immunoblotting pattern with the anti-␣ antibody (Fig. 4A). Mutations of Glu-345 to aspartic acid (E345D), lysine (E345K), or valine (E345V) eliminated the SCH 28080sensitive K ϩ -ATPase activity. No activity was detected even in the presence of high concentrations of KCl (data not shown). However, the mutant having glutamine for glutamic acid (E345Q) retained 50% of the wild-type K ϩ -ATPase activity (Table II). The same situation was reported for Na ϩ ,K ϩ -ATPase (31). Fig. 3 shows the SCH 28080-sensitive K ϩ -ATPase activity of the mutant E345Q as a function of the K ϩ concentration. The K m value of E345Q mutant for K ϩ stimulation was 2.6 mM, a value about 10-fold higher than that of wild-type enzyme, and the V max value was about 50% of the wild type enzyme (Table III).
Site-directed Mutations of the Phosphorylation Site-H ϩ ,K ϩ -ATPase is phosphorylated during its reaction cycle. The phosphorylation site is Asp-387 of the ␣-subunit of rabbit H ϩ ,K ϩ -ATPase (19). We prepared several mutants, in which Asp-387 was replaced with glutamic acid (D387E), histidine (D387H), or asparagine (D387N). The extent of expression of these mutants was almost the same as that of the wild-type ␣-subunit, as judged by immunoblotting with the anti-␣ subunit antibody (Fig. 4B). In each mutant, however, no significant SCH 28080sensitive K ϩ -ATPase activity was observed (Table II).
Site-directed Mutations of the FITC Binding Site-FITC specifically binds to a lysine residue of H ϩ ,K ϩ -ATPase (Lys-519 in rabbit gastric H ϩ ,K ϩ -ATPase) and inhibits the enzyme activity. The presence of ATP protects the ATPase from labeling and inactivation by FITC (32). Therefore, this residue is thought to be part of or near to the nucleotide binding site. The amino acid sequence around the FITC binding site is well conserved in P-type ATPases, i.e. H ϩ ,K ϩ -ATPase, Na ϩ ,K ϩ -ATPase, and sarcoplasmic and plasma membrane Ca 2ϩ -ATPases (3,16,(33)(34)(35)(36).
Here we mutated Lys-519 to arginine (K519R), asparagine (K519N), glutamine (K519Q), or glutamic acid (K519E) and measured the K ϩ -ATPase activity in the membrane fractions. These mutants were expressed to approximately the same ex-tent as the wild-type ␣-subunit (Fig. 4C). The mutations to arginine (K519R), glutamic acid (K519E), and asparagine (K519N) eliminated the ATPase activity. The mutant K519Q retained 30% of the SCH 28080-sensitive K ϩ -ATPase activity of the wild-type enzyme (Table II). DISCUSSION Gastric H ϩ ,K ϩ -ATPase belongs to the family of P-type ATPases, which form phosphorylated intermediates in their catalytic cycles and are inhibited by vanadate (37). Sarcoplasmic and endoplasmic reticulum Ca 2ϩ -ATPases and Na ϩ ,K ϩ -ATPase also belong to this group, and these ATPases were cloned and expressed, and their structure-function relationships have been studied extensively (14, 17, 18, 31, 33-35, 38 -41). There are many reports describing the cDNA cloning of the ␣and ␤-subunits of the gastric H ϩ ,K ϩ -ATPase from rat, pig, human, rabbit, and dog (3,19,20,(42)(43)(44)(45)(46)(47). There have been few reports, however, of the functional expression of the gastric H ϩ ,K ϩ -ATPase. Recently, gastric H ϩ ,K ϩ -ATPase subunits were expressed in renal proximal tubular epithelial cells (LLC-PK 1 ), but the enzyme functions were not measured (48). Very recently, Mathews et al. (49) reported the functional expression of the ATPase in Xenopus oocytes. The lack of an effective expression system slowed the study of the structure-function relationships of the gastric H ϩ ,K ϩ -ATPase. Here we report the functional expression of gastric H ϩ ,K ϩ -ATPase in HEK293 cells. The cDNAs for the ␣and ␤-subunits of H ϩ ,K ϩ -ATPase were introduced separately to pcDNA3 vectors, and the cDNAs were transfected separately or simultaneously into HEK293 cells with the calcium phosphate method. When the ␤-subunit cDNA alone was transfected, the ␣-subunit was not detected by immunoblotting, suggesting that there is no endogenous H ϩ ,K ϩ -ATPase in HEK293 cells. No significant SCH 28080sensitive K ϩ -ATPase was detected in the membrane fraction of these cells. When the ␣-subunit cDNA was transfected without the ␤-subunit cDNA, a slight band of the ␣-subunit was observed, but SCH 28080-sensitive K ϩ -ATPase activity was not detected. When the ␣and ␤-subunit cDNAs were co-transfected, the ␣-subunit was clearly seen on immunoblot, and a significant and reproducible SCH 28080-sensitive K ϩ -ATPase  Ϫ0.01 Ϯ 0.01 (n ϭ 4) activity could be demonstrated. Therefore, the ␤-subunit increases the expression of the ␣-subunit and is essential for the functional expression of the H ϩ ,K ϩ -ATPase. As Na ϩ ,K ϩ -ATPase is a ubiquitous enzyme, it is likely that endogenous Na ϩ ,K ϩ -ATPase ␤-subunit exists in HEK293 cells. It would appear, however, that the ␣-subunit of the H ϩ ,K ϩ -ATPase does not assemble with the ␤-subunit of the Na ϩ ,K ϩ -ATPase in a functional form (50), although the possibility that a very weak ouabain-resistant K ϩ -ATPase activity was manifested by the H ϩ ,K ϩ -␣/Na ϩ ,K ϩ -␤ hybrid molecule cannot be excluded. So far, there has been no report that indicates the functional assembly between H ϩ ,K ϩ -ATPase ␣-subunit and Na ϩ ,K ϩ -ATPase ␤-subunit. The H ϩ ,K ϩ -ATPase ␣-subunit seems to discriminate the H ϩ ,K ϩ -ATPase ␤-subunit from the Na ϩ ,K ϩ -ATPase ␤-subunit.
On the other hand, there are several reports that the Na ϩ ,K ϩ -ATPase ␣-subunit can assemble with the H ϩ ,K ϩ -ATPase ␤-subunit in Xenopus oocytes (51,52) and HeLa cells (53). The hybrid molecule (Na/K-␣ and H/K-␤) showed Na/K pump current and Rb ϩ uptake, although these activities were much smaller than those in the authentic Na ϩ ,K ϩ -ATPase ␣/␤ complex (51). In this case, the H ϩ ,K ϩ -ATPase ␤-subunit manages to act as a surrogate for the Na ϩ ,K ϩ -ATPase ␤-subunit.
The expressed K ϩ -ATPase activity described here did not increase in the presence of gramicidin, which stimulates K ϩ -ATPase activity in gastric vesicles (54). This may be due to the leakiness of the HEK cell membrane to K ϩ and H ϩ . In fact, when we measured proton transport of the membrane fraction using acridine orange, the fluorescence was not significantly quenched.
We mutated amino acid residues involved in the putative cation binding (Glu-345), the formation of the phosphorylated intermediate (Asp-387), and the ATP binding (Lys-519). We expressed these mutants in our system and compared the properties of the mutants with those of wild-type enzyme. For the Glu-345 mutants, three in four mutants we prepared (E345D, E345K, E345V) did not show the SCH 28080-sensitive K ϩ -ATPase activity. The remaining mutant, E345Q, retained 50% of the K ϩ -ATPase activity of the wild-type enzyme. In sarcoplasmic reticulum Ca 2ϩ -ATPase, Glu-309 (the counterpart of Glu-345 of the H ϩ ,K ϩ -ATPase) is supposed to be one of the amino acid residues constituting the Ca 2ϩ high affinity site, because the replacement of Glu-309 residue by glutamine resulted in complete loss of Ca 2ϩ transport activity and phosphorylation from ATP, and because the phosphorylation of this mutant with inorganic phosphate was observed even in the presence of Ca 2ϩ (15). This residue is well conserved in P-type ATPases, including Na ϩ ,K ϩ -ATPase, plasma membrane Ca 2ϩ -ATPase, yeast H ϩ -ATPase, and gastric H ϩ ,K ϩ -ATPase. However, Glu-329 in rat kidney Na ϩ ,K ϩ -ATPase ␣-subunit has been shown not to be essential for active transport of Na ϩ and K ϩ , because the replacement of this residue to glutamine retains the enzyme activity (14). Furthermore, mutations of Glu-327 (the counterpart of Glu-345 in Na ϩ ,K ϩ -ATPase ␣ 2 -subunit) to glutamine and leucine allow the enzyme to retain function, whereas mutations to aspartic acid and alanine do not (31). Our results presented here suggest that Glu-345 is not absolutely essential for the ATPase function in gastric H ϩ ,K ϩ -ATPase as is the case with Na ϩ ,K ϩ -ATPase. Because Glu-345 can be replaced by glutamine, the negative charge of the glutamic acid residue in this site is not indispensable for the function of H ϩ ,K ϩ -ATPase. Rather, the bulkiness of the side chain in this site appears to be important, because the glutamic acid cannot be replaced by aspartic acid. These features are also comparable to those of Na ϩ ,K ϩ -ATPase (31). In the present experiment, replacement of Glu-345 by glutamine reduced the affinity for K ϩ 10-fold. Therefore, this residue appears to be involved in determining the K ϩ affinity. The role of this glutamic acid residue in H ϩ ,K ϩ -ATPase is also comparable to that in Na ϩ ,K ϩ -ATPase, whose affinity for Na ϩ and K ϩ was reduced by the replacement of glutamic acid by glutamine (14,31). The difference in the manner by which this glutamic acid residue contributes to the functioning of H ϩ ,K ϩ -ATPase, Na ϩ ,K ϩ -ATPase, and Ca 2ϩ -ATPase might reflect the difference in the structures of the ion sensors or the difference in the manner by which K ϩ participates in the reaction cycles. Both H ϩ ,K ϩ -ATPase and Na ϩ ,K ϩ -ATPase actively translocate K ϩ , while K ϩ functions as an accelerator from one side of the membrane in Ca 2ϩ -ATPase (55). Recently, Kuntzweiler et al. (12) studied the effects of cations on [ 3 H]ouabain binding and have shown that Glu-327 in sheep Na ϩ ,K ϩ -ATPase ␣ 1 -subunit stabilizes the K ϩ -induced conformation in the reaction cycle. Until now there has been no direct evidence as to whether the K ϩ recognition system is common between H ϩ ,K ϩ -ATPase and Na ϩ ,K ϩ -ATPase. The present results suggest that the structure and mechanism for K ϩ recognition is similar (or partly identical) between Na ϩ ,K ϩ -ATPase and H ϩ ,K ϩ -ATPase. However, there are striking functional differences between Na ϩ ,K ϩ -ATPase and H ϩ ,K ϩ -ATPase; the former electrogenically transports Na ϩ and K ϩ , and the latter non-electrogenically transports H ϩ and K ϩ .
As to the Asp-387 mutants, all the mutants prepared (D387E, D387H, and D387N) were inactive, although they were expressed in sufficient quantities. The experimental results on the mutations to asparagine and histidine have shown that the existence of phosphate acceptor moiety at this site is indispensable for the enzyme function. Because the aspartic acid cannot be replaced by glutamic acid, the bulkiness of the side chain in the phosphorylation site appears to be very strict. These results are in agreement with the results obtained with sarcoplasmic Ca 2ϩ -ATPase and Na ϩ ,K ϩ -ATPase (17,18). The primary structure around the phosphorylation site is well conserved in the P-type ATPases. The result presented here suggests the common three-dimensional structure around the phosphorylation site among these three P-type ion-transporting ATPases, and the requirement of a stringent structure for their functions.
As to the Lys-519 mutants, three in four mutants we prepared (K519E, K519N, and K519R) were inactive. The K519Q mutant retained 30% of the K ϩ -ATPase activity of the wildtype enzyme. In sarcoplasmic Ca 2ϩ -ATPase, mutation of the corresponding lysine residue to arginine, glutamine, and glutamic acid led to activities of 60%, 25%, and 5% of the activity of the wild-type enzyme, respectively (18), indicating that this site is tolerant of amino acid replacement, although it cannot withstand a negative charge. This is also the case with H ϩ ,K ϩ -ATPase. However, it was surprising that the effect of mutation to a basic amino acid, arginine, was quite different between Ca 2ϩ -ATPase and H ϩ ,K ϩ -ATPase; the replacement did not bring severe damage to Ca 2ϩ -ATPase, whereas H ϩ ,K ϩ -ATPase could not withstand the replacement. Although the amino acid sequence around the FITC binding site is well conserved between H ϩ ,K ϩ -ATPase and Ca 2ϩ -ATPase (16), some steric difference must exist between the two FITC binding sites.