Introduction

Top Summary Introduction Materials and Methods Results and Discussion References

Gastric H⁺,K⁺-ATPase, a P-type ATPase, is responsible for gastric acid secretion. The enzyme is located in the tubulovesicular system of the parietal cell and is translocated to the apical plasma membrane after hormonal stimulation. It catalyzes an electroneutral transport of K⁺ versus H⁺ energized by ATP hydrolysis. Gastric H⁺,K⁺-ATPase can be reversibly inhibited by imidazo pyridines like 2-methyl-8-[phenylmethoxy] imidazo-(1,2a) pyridine 3-acetonitrile (SCH 28080), apparently by binding to a high-affinity site for K⁺ (Wallmark et al., 1987, Keeling et al., 1989; Mendlein and Sachs, 1990). Munson et al. (1991) determined the binding site of a photoaffinity analog of SCH 28080 to be the domain including the first two transmembrane segments of the -subunit as well as the extracellular loop between these segments. Lyu and Farley (1997) recently reported that Na⁺,K⁺-ATPase, in which twelve amino acids of the first transmembrane segment of the -subunit had been replaced by the homologous amino acids from H⁺,K⁺-ATPase, showed a rather high sensitivity toward this drug. This suggests that the N terminal part of the -subunit of H⁺,K⁺-ATPase is the main participant in SCH 28080 binding, as it is in the binding of ouabain to Na⁺,K⁺-ATPase (Lingrel and Kuntzweiler, 1994). On the other hand, Asano et al. (1997) suggested recently that glutamic acid 820 might also be important for the binding of SCH 28080.

Recently, we prepared in Sf9 insect cells with aid of the baculovirus expression system a series of mutants of rat gastric H⁺,K⁺-ATPase in which glutamic acid 820, located in the 6th transmembrane domain of the -subunit, had been replaced by various amino acids (Hermsen et al., 1998). All mutants (E820D, E820A, E820Q, and E820N) could, like the wild-type enzyme, be phosphorylated by ATP. The hydrolysis of the phosphorylated intermediate (E-P) of mutant E820D could, like that of the wild-type enzyme, be enhanced by 1 mM K⁺. This mutant showed a normal K⁺-stimulated ATPase activity with a maximal activity at 1 mM K⁺. The dephosphorylation of the E-P of the E820Q and E820N mutants could not be stimulated by K⁺ and that of the E820A mutant only at 100 mM K⁺. Similarly, the ATPase activity of the E820Q and E820N mutants was not stimulated by K⁺, whereas, that of the E820A mutant was only slightly increased at 100 mM K⁺. For clarity, these three mutants are coined in the present paper as "K⁺-insensitive mutants".

Preincubation of the wild-type enzyme at 0°C with SCH 28080, a specific inhibitor of gastric H⁺,K⁺-ATPase, resulted in a decrease of the ATP phosphorylation level with an IC₅₀ value of 20 nM (see Table 1). Similar IC₅₀ values were obtained for the mutants E820D and E820N. The E820A mutant, however, was less sensitive to SCH 28080 and phosphorylation of the E820Q mutant could hardly be reduced by this compound. Thus, the three K⁺-insensitive mutants have a different apparent affinity for SCH 28080 when ATP phosphorylation is used as the assay method. Moreover, these findings are difficult to match with the suggestion of Asano et al. (1997) that Glu⁸²⁰ is directly involved in SCH 28080 binding, and they make it likely that a more complex mechanism must exist for SCH 28080 inhibition.

In addition, we recently demonstrated (Swarts et al., 1998) that all three K⁺-insensitive mutants (E820Q, E820N, and E820A) had an ATPase activity in the absence of K⁺. This constitutive ATPase activity of these mutants could be inhibited by SCH 28080 with similar IC₅₀ values (0.2-0.7 µM) as that of the wild-type enzyme (0.4 µM) and the E820D mutant (0.3 µM; see Table 1). In that study, we also measured phosphorylation and dephosphorylation at 21°C and found that the K⁺-insensitive mutants had a high spontaneous dephosphorylation rate that could not be further stimulated by K⁺.

The purpose of the present study is to provide a solution for these apparently contradictory findings. Why does SCH 28080 inhibit the ATPase activity (measured at 37°C) of all mutants with a rather similar affinity, whereas the IC₅₀ values for reduction of phosphorylation at 0°C vary so much? To give a framework for the experiments to be described, a simplified version of the Post-Albers model for P-type ATPases is given in Fig. 1. This model takes into account that both the E₂ form and the E₂-P form of gastric H⁺,K⁺-ATPase are able to react with SCH 28080 (Keeling et al., 1989; Mendlein and Sachs, 1990; Van der Hijden et al., 1991).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Post-Albers reaction scheme for H+,K+-ATPase including the SCH 28080 inhibition mechanism. The clockwise reactions are indicated in the text with a positive (+) sign and the counter-clockwise reactions with a minus () sign. Reaction + 4 is the conversion of E2-P to E2.

The present study shows that the K⁺-insensitive mutants have a high tendency for the E₁ form, in particular at 21°C, and therefore do not bind SCH 28080 very well under nonphosphorylating conditions. Importantly, under ATP-hydrolyzing conditions, they are temporarily in the E₂-P form, which does bind SCH 28080. During each catalytic cycle, part of the enzyme is thus trapped by SCH 28080 and therefore inhibited. The lack of reduction of the ATP phosphorylation level by SCH 28080 of mutant E820Q does not mean that residue Glu⁸²⁰ is directly involved in SCH 28080 binding, but only indicates the very low tendency of this mutant for the E₂ conformational state of the enzyme.

	Materials and Methods

Top Summary Introduction Materials and Methods Results and Discussion References

Preparation of Mutants. Rat gastric H⁺,K⁺-ATPase was expressed in Sf9 cells as described previously (Klaassen et al., 1993; Swarts et al., 1996). The BaculoGold transfer vector pAcUW51 (Pharmingen, San Diego, CA), containing the full length cDNA of the rat H⁺,K⁺-ATPase - and -subunits was used for site-directed mutagenesis (Deng and Nickoloff, 1992). The obtained pAcUW51-HK-wt and pAcUW51-HK-mutants, with the DNA code of the -subunit under control of the polyhedrin promoter and that of the -subunit under control of the p10 promoter were used to produce recombinant viruses. The (mutated) transfer vectors and linearized AcNPV DNA (BaculoGold DNA) were co-transfected in Sf9 cells according to the instructions of the supplier. The viruses obtained by this method were further purified via a plaque assay, and expression of the -subunit was screened by Western blotting. The presence of the desired mutation in the viral genome was checked by sequence and restriction analysis.

Production of Recombinant H⁺,K⁺-ATPase. Sf9 cells were grown at 27°C in 100-ml spinner flask cultures (Klaassen et al., 1993). For production of H⁺,K⁺-ATPase 1.0-1.5 × 10⁶ cells/ml were infected at a multiplicity of infection of 1 to 3 in the presence of 1% ethanol (Klaassen et al., 1995) and incubated for 3 days using Xpress medium (BioWhittaker, Walkersville, MD) containing additionally 0.1% pluronic F-68 (Sigma, Bornem, Belgium). By using the latter incubation conditions, both the phosphorylation capacity and the H⁺,K⁺-ATPase activity of the expressed enzyme were 2 to 3 times higher than previously found (Swarts et al., 1996).

Preparation of Sf9 Membranes. The Sf9 cells were harvested by centrifugation at 2000g for 5 min. After resuspension at 0°C in 0.25 M sucrose, 2 mM EDTA, and 25 mM HEPES/Tris (pH 7.0), the membranes were sonicated 3 × 15 s at 60 W (Branson Power Company, Denbury, CT). After centrifugation for 30 min at 10,000g the supernatant was recentrifuged for 60 min at 100,000g at 4°C. The pelleted membranes were resuspended in the above mentioned buffer and stored at 20°C.

Protein Determination. Protein was determined with the modified Lowry method described by Peterson (1983) using bovine serum albumin as a standard.

ATP Phosphorylation Capacity. ATP phosphorylation was determined as described before (Swarts et al., 1998). Sf9 cell membranes (amounts indicated in the legends) were incubated at 0° or 21°C in 50 mM Tris-acetic acid (pH 6.0), 1.2 mM MgCl₂, and 0.2 mM EDTA with and without 0.1 mM SCH 28080 in a volume of 50 µl. After 60-min preincubation, 10 µl of 0.6 µM [-³²P]ATP was added and incubated for the indicated time at either 0° or 21°C. The reaction was stopped by adding 5% trichloroacetic acid in 0.1 M phosphoric acid. The phosphorylated protein was collected by filtration over a 0.8 µm membrane filter (Schleicher and Schull, Dassel, Germany). After repeated washing, the filters were analyzed by liquid scintillation analysis.

Dephosphorylation Studies. For dephosphorylation studies, phosphorylation was carried out and the dephosphorylation mixture was diluted 8.3 times with nonradioactive ATP (final concentration 10 µM) to prevent rephosphorylation with radioactive ATP and was further incubated for 3 to 30 s at 21°C (Helmich-de Jong et al., 1985). The reaction was stopped at the time points indicated and the amount of E-P was determined as described above.

Analysis of Data. The IC₅₀ values for SCH 28080 were iteratively determined by fitting the concentration relationship to the logistic equation Y = A + (B  A)/1 + (10^C/10^X)^D (A = bottom plateau; B = top plateau; C = IC₅₀; D = Hill coefficient; the values of X and C were entered as the logarithm of concentration) using the nonlinear regression computer program InPlot (GraphPAD Software for Science, San Diego, CA). All data are presented as mean values with S.E.M. Differences of average were tested for significance by means of Student's t test.

Chemicals. [-³²P]ATP (3000 C_immol¹, Amersham, Buckinghamshire, UK) was diluted with nonradioactive Tris-ATP (pH 6.0) to a specific radioactivity of 20 to 100 C_immol¹. SCH 28080, kindly provided by Dr. A. Barnett (Schering-Plough, Kenilworth, NJ), was dissolved in ethanol and diluted to its final concentration of 0.1 mM in 0.2% ethanol.

	Results and Discussion

Top Summary Introduction Materials and Methods Results and Discussion References

The mechanism of the SCH 28080-induced decrease of the steady-state ATP phosphorylation level and the ATPase activity of the pig gastric H⁺,K⁺-ATPase has previously been explained with the aid of the adapted form of the Post-Albers scheme given in Fig. 1 (Keeling et al., 1989; Mendlein and Sachs, 1990; Van der Hijden et al., 1991). This model will be used as a framework to explain the behavior of both the wild-type enzyme and the Glu⁸²⁰ mutants.

SCH 28080 Reduction of the ATP Phosphorylation Capacity. The effect of SCH 28080 on the phosphorylation level had previously been measured by preincubation with the drug for 1 h at 0°C followed by a 10-s phosphorylation at 0°C using 0.1 µM ATP (pH 6.0) (Hermsen et al., 1998). Because we recently observed with some mutants (E820D, E820Q) that a higher phosphorylation level could be reached at a (pre)incubation temperature of 21°C (Swarts et al., 1998), dose-inhibition curves for SCH 28080 were carried out at both temperatures. Fig. 2A shows that at 0°C the wild-type enzyme showed a sigmoid dose-inhibition curve with an IC₅₀ value of 30 nM. At 21°C the dose-inhibition curve for the wild-type enzyme shifted to the right (IC₅₀ = 0.2 µM), indicating a temperature-sensitive decrease in apparent affinity for SCH 28080. These findings can be explained by assuming that the wild-type enzyme is at least partly present in the E₂ conformation, which forms an E₂-SCH complex during the preincubation period (reaction 7). As a result, a new equilibrium between E₁E₂E₂-SCH is achieved, which depends on the SCH 28080 concentration used. Upon addition of ATP, the residual E₁ reacts and E₁-P is formed (reaction +2); this is rapidly converted into the E₂-P conformation (reaction +3). The decrease in apparent affinity for SCH 28080 upon increasing the temperature from 0° to 21°C might be due to a shift in the E₁E₂ equilibrium to the left, resulting in formation of less E₂-SCH complex at suboptimal SCH 28080 concentrations. Alternatively, the increase in temperature might also shift the E₂E₂-SCH equilibrium to the left, so indirectly resulting in more E₁ and, thus, a higher phosphorylation level.

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Effect of SCH 28080 on ATP phosphorylation capacity of the H+,K+-ATPase mutants at 0° and 21°C. Membranes (4-8 µg) obtained from Sf9 cells infected with either wild-type virus (A), mutant E820A (B), or mutant E820Q (C) were preincubated for 60 min at either 0° or 21°C in the presence of 1.2 mM MgCl2, 0.2 mM EDTA, 50 mM Tris-acetic acid (pH 6.0) and the indicated SCH 28080 concentrations. After phosphorylation for 10 s at 0° or 21°C with 0.1 µM [-32P]ATP the phosphorylation level (E-P; pmol · mg1 protein) was determined and corrected for that of mock-infected cells. Mean values ± S.E.M. of three enzyme preparations.

Time Course of SCH 28080 Effect. We investigated the difference between the effect of high SCH 28080 concentrations on the phosphorylation level at 21°C for the wild-type enzyme (decrease) and the E820Q mutant (increase) in more detail by adding 100 µM SCH 28080 15 s after the start of the phosphorylation. Figure 3 shows that addition of SCH 28080 resulted in an immediate decrease in the phosphorylation level of the wild-type enzyme and in an immediate increase in that of the E820Q mutant. In both cases a new steady-state level was reached within 20 to 30 s; this level was similar to that obtained when the enzyme was preincubated for 60 min with SCH 28080 prior to the addition of ATP (dotted line for E820Q). The rate of equilibrium change is in the same order of magnitude as described for the binding of SCH 28080 at this temperature (Keeling et al., 1989), suggesting that the binding rate primarily determined the rate of this process.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Time course of the effect of SCH 28080 on the ATP-phosphorylation capacity of the wild-type enzyme and the E820Q mutant. Membranes (4 µg) obtained from Sf9 cells infected with either wild-type virus (A) or mutant E820Q (B) were preincubated for 60 min at 21°C in the presence of 1.2 mM MgCl2, 0.2 mM EDTA, and 50 mM Tris-acetic acid (pH 6.0). At t = 0 s, 0.1 µM [-32P]ATP (final concentration) was added and the level of the phosphorylated intermediate was followed in time. At t = 15 s, 100 µM SCH 28080 (final concentration) was added. Data were corrected for those of mock-infected cells. Representative for 2 to 3 experiments.

Dephosphorylation Kinetics of the Phospho-Intermediates. If the above reasoning is correct, the type of E-P obtained with the E820Q mutant will be different in the absence of SCH 28080 (E₂-P) from that obtained in the presence of SCH 28080 (E₂-P-SCH). We therefore studied the dephosphorylation kinetics of the E-Ps obtained under various conditions in the absence of K⁺. Figure 4A shows the dephosphorylation process of the phospho-intermediate of the wild-type enzyme (reaction 4). The measured rate constant at 21°C was 0.08 ± 0.011 s¹ (n = 4). In the presence of 0.1 mM SCH 28080, no E-P was formed.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Dephosphorylation rate of the phospho-intermediates of the H+,K+-ATPase (mutants) generated in the absence and presence of SCH 28080. Membranes (13-19 µg) obtained from Sf9 cells infected with either wild-type virus (A), mutant E820A (B), or mutant E820Q (C) were preincubated with () and without 0.1 mM SCH 28080 () in the presence of 1.2 mM MgCl2, 0.2 mM EDTA and 50 mM Tris-acetic acid (pH 6.0) for 30-60 min at 21°C and phosphorylated with 0.1 µM [-32P]ATP. After 10 s (t = 0), the incubation medium was diluted from 60 to 500 µl with non-radioactive ATP (final concentration 10 µM) in the presence of 50 mM Tris-acetic acid (pH 6.0), 1 mM Mg2+, with () and without () SCH 28080 to prevent rephosphorylation with radioactive ATP and incubated for 3-30 s. The residual phosphorylation level (E-P; pmol.mg1 protein), corrected for the mock infected levels, is plotted as a function of time. Mean values ± S.E.M. of three enzyme preparations.

1

1

1

Effects of Inorganic Phosphate, Vanadate, and K⁺ on the ATP Phosphorylation Capacity. Table 2 describes the results from some additional experiments at 21°C that confirm the model of Fig. 1. In this table, results with two other mutants (E820D and E820N) are included. In the absence of SCH 28080 the 10-s phosphorylation level at 21°C varied for these mutants between 2.9 and 5.4 pmol E-P per mg protein. The observed difference might either be because under these conditions the maximal phosphorylation level had not been reached for all mutants or to differences in the amount of correct folded enzyme. It is known, at least for P-type ATPases, that in the baculovirus system, only part of the expressed protein is active (De Tomaso et al., 1993; Xie et al., 1996; Swarts et al., 1996; Liu and Guidotti, 1997).

View this table: [in this window] [in a new window]   TABLE 2 The effect of SCH 28080, inorganic phosphate, vanadate, and K+ on the steady-state ATP phosphorylation level Sf9 membranes (4-8 µg) were preincubated in the presence of 50 mM Tris-acetic acid (pH 6.0), 1.2 mM MgCl2, 0.2 mM EDTA, and in addition with either 0.1 mM SCH 28080, 1 mM inorganic phosphate pH 6.0 (Pi), 0.1 mM SCH 28080 + 1.0 mM Pi, 1 mM vanadate, or 10 mM KCl at 21°C. After 60 min 0.1 µM [-32P]ATP (final concentration) was added and after 10 s the amount of E-P was determined. The data are corrected for the levels obtained with membranes of mock infected Sf9 cells. Averages with S.E.M. for three enzyme preparations are presented. Differences of the data compared to the control values are tested for significance by means of Student's t test (*p < .05; **p < .01).

SCH 28080 Inhibition in the ATPase Activity Measurements. Under ATP hydrolyzing conditions (37°C and long incubation periods), the wild-type enzyme as well as the Glu⁸²⁰ mutants come temporarily in the E₂ or E₂-P forms both of which bind SCH 28080. During each catalytic cycle, part of the enzyme is thus trapped by SCH 28080 and therefore inhibited. This explains why the IC₅₀ values of SCH 28080 for the wild-type enzyme and the E820 mutants in the ATPase differ only slightly (Table 1).

Concluding Remarks. The present study shows that differences in sensitivity for SCH 28080 on the ATP phosphorylation level between various Glu⁸²⁰ mutants of gastric H⁺,K⁺-ATPase are not due to a direct involvement of Glu⁸²⁰ for SCH 28080 binding, but to differences in K⁺-sensitivity and the direction of the E₂E₁ equilibrium. In general, if mutation of a certain amino acid has an effect on the degree of enzyme inhibition, it does not prove that the mutated residue is directly involved in binding of the inhibitor.