Characterization of the plasma membrane Mg2+-ATPase from the yeast, Saccharomyces cerevisiae.

The plasma membrane of Saccharomyces cerevisiae has a Mg2+-dependent ATPase which is distinct from the mitochondrial Mg2+-ATPase and at the pH optimum of 5.5 has a Km for ATP of 1.7 mM and a Vmax of 0.42 mumol of ATP hydrolyzed/mg/min. At least three protein components of the crude membrane (Mr = 210,000, 160,000 and 115,000) are labeled with [gamma"32P]ATP at pH 5.5. These phosphoproteins form rapidly in the presence of Mg2+, rapidly turn over the bound phosphate when unlabeled ATP is added, and dephosphorylate after incubation in the presence of hydroxylamine. Vanadate, an inhibitor of the Mg2+-ATPase activity, blocks the phosphorylation of the 210,000- and 115,000-dalton proteins. At pH 7.0, only the 210,000- and 160,000-dalton proteins are phosphorylated. While these three phosphorylated intermediates have not been unambiguously identified as components of the Mg2+-ATPase, the finding of such phosphorylated components in association with that activity implies that this enzyme differs in mechanism from the mitochondrial proton pump and that it is similar in mechanism to the metal ion pumps ((Na+-K+)-ATPase and Ca2+-ATPase) of the mammalian plasma membrane.

The plasma membrane of Saccharomyces cerevisiae has a M<'-dependent ATPase which is distinct from the mitochondrial M$'-ATPase and at the pH optimum of 5.5 has a K, for ATP of 1.7 mM and a V,,,, of 0.42 pmol of ATP hydrolyzed/mg/min. At least three protein components of the crude membrane (Bfr = 210,000, 160,000 and 115,000) are labeled with [y-32P]ATP at pH 5.5. These phosphoproteins form rapidly in the presence of MgZ', rapidly turn over the bound phosphate when unlabeled ATP is added, and dephosphorylate after incubation in the presence of hydroxylamine. Vanadate, an inhibitor of the M&'-ATPase activity, blocks the phosphorylation of the 210,000-and 115,000dalton proteins. At pH 7.0, only the 210,000-and 160,000-dalton proteins are phosphorylated. While these three phosphorylated intermediates have not been unambiguously identified as components of the M$'-ATPase, the finding of such phosphorylated components in association with that activity implies that this enzyme differs in mechanism from the mitochondrial proton pump and that it is similar in mechanism to the metal ion pumps ((Na'-K')-ATPase and Ca"-ATPase) of the mammalian plasma membrane.
Two classes of membrane-bound ATPases have been described in the literature.
Class I ATPases depend on Mg"+ for activity, are involved in oxidative phosphorylation, pump protons, and are composed of integral and peripheral proteins. These enzymes have at least six subunits and can hydrolyze ATP using two subunits of M, = 56,000 and 59,000. There is no phosphorylated intermediate involved in ATP hydrolysis. Examples of class I enzymes are the Mg2'-ATPases of bacteria (l), cbloroplasts (2), and mitochondria (3). Class II ATPases are dependent upon Mg2+ for activity, and are also stimulated by other cations. They are involved in ion translocation and the maintainance of ion gradients, pump K', Na+, or Ca2', and are composed of only integral membrane proteins. These enzymes have only one or two subunits and appear to hydrolyze ATP using one subunit of M, = 100,000. In addition, an unstable phosphorylated intermediate is formed during ATP hydrolysis. Examples of class II enzymes are the (Na+-K')-ATPase isolated from brain, kidney and eel (4), and the Ca"-ATPase isolated from sarcoplasmic reticulum (5).
A Mg"-dependent ATPase has been found in both the cell  (8) has been analyzed as a function of growth and metabolic state and a detailed comparison between the mitochondrial ATPase and the plasma membrane ATPase in the yeast Schizosaccharomyces pombe has been reported (9). The Mg"'-ATPase of the plasma membrane of the related fungus Neurospora has also been characterized and compared to the mitochondrial Mg"-ATPase in this organism.
These fungal Mg"'-ATPases have characteristics of both class I and class II enzymes. They are located in the plasma membrane and the Neurospora enzyme is involved in proton pumping (10)(11)(12). It is not known, however, whether phosphoprotein intermediates are formed during ATP hydrolysis. The studies described in this paper were designed to characterize the activity of the Mg'+-dependent ATPase from the plasma membrane of S. cerevisiae with respect to its sensitivity to inhibitors, response to exogenously added cations, pH dependence, and ability to be stably phosphorylated by [y"'P]ATP.
These experiments indicate that the enzyme is distinct from the mitochondrial Mg"'-ATPase and the known mammalian metal ion pumps. In addition, the ATP hydrolytic activity of the enzyme has been associated with the formation of vanadate-sensitive phosphorylated intermediates, which implies that the S. cerevisiae plasma membrane Mg"+-ATPase is similar in mechanism to the class II type plasma membrane ATPases. Inhibitors and 5 mM MgC12 were added as described in the figure legends and the reaction was always started by the addition of enzyme, terminated by the addition of 0.5 ml of cold 5% trichloroacetic acid, and subsequently kept at 0°C. After the addition of trichloroacetic acid, an aliquot was taken to determine the extent of hydrolysis, the mixture was made of 10 mM ATP and 10 mM P,, and 1 ml of cold wash fluid (5% trichloroacetic acid, 10 mM P,, 2.5 mM ATP) was added. After 10 min, the precipitated protein was sedimented at 12,000 x g for 10 min and washed twice with 5 ml of cold wash fluid. weighed. The amount of radioactive phosphate in each peak was then calculated using the percentage of counts in each peak, the number of counts originally placed on the gel, and the specific activity of the [+'P]ATP. Molecular Weight Determination-Molecular weights were determined using the gel electrophoresis system described above with the following gel standards: myoglobin (M, = 17,200), bovine serum albumin (M, = 65,000), phosphorylase b (M, = 94,000) and myosin (M, = 220,000). As the graph of the log of the molecular weight uersus the relative mobility of the protein standards gave a straight line, the molecular weight of the phosphorylated intermediates were determined from this standard curve using the relative mobilities obtained from the autoradiogram for these proteins.

Characterization-The
ATPase activity of a microsomal membrane fraction of the yeast S. cerevisiae is Mg"-dependent. The optimum Mg" concentration of 5 mM corresponds to a Mg"+-ATP ratio of approximately 1 and a large excess of Mg"+ (25 mM) causes only a 10 to 20% inhibition of the enzyme (Fig. 1). The dependence of ATPase activity on the concentration of ATP is described by an approximately hyperbolic curve with a K, of 1.7 mu + 0.03 and a V,,,, of 0.42 & 0.02 pmol of ATP hydrolyzed/mg/min. Some fractionation of the Mg"-ATPase activity could be achieved by sedimentation of the microsomal membrane in a sucrose density gradient as shown in Fig. 2. The Mg'+-ATPase activity was associated with a membrane protein peak of bouyant density 1.22 g/ml which is similar to the density observed for the plasma membrane of S. cerevisiae by others (26,27). Comparison with Other Known ATPases- Fig.  3 shows the pH dependence of the microsomal membrane MgZ+-ATPase activity in the presence and absence of oligomycin, an inhibitor of the mitochondrial ATPase. At the pH optimum of 5.5, over 60% of the ATPase activity is oligomycin-resistant and can be attributed to the plasma membrane Mg*'-ATPase. The yeast mitochondrial Mg"-ATPase has a pH optimum near 9 (9) and is sensitive to oligomycin. Therefore, the oligomycin-sensitive ATPase activity at pH 9.0 to 9.5 is probably caused by contamination of the microsomal preparation by the mitochondrial enzyme. Also, the pH 9.0 to 9.5 activity is completely sensitive to the mitochondrial ATPase inhibitors efrapeptin and Dio 9 (28), while the pH 5.5 activity is only partially sensitive to these inhibitors (data not shown). Since the metal ion pumps of the mammalian plasma membrane enzyme are stimulated by certain cations, the yeast in the presence of exogenously added cations. The yeast enzyme is not affected by Ca2+ (5 to 500 PM) and it is slightly inhibited by Na+ (54% inhibited by 375 mM) and K' (28% inhibited by 50 mM). Experiments using similar concentrations of Tris showed that this inhibition appears to be due to increasing ionic strength. Strophanthidin, a specific inhibitor of the (Na+-K')-ATPase did not inhibit the yeast enzyme. Therefore, the yeast enzyme does not have the properties of the mammalian plasma membrane metal ion pumps.
Vanadate, a known inhibitor of plasma membrane ATPases (29)(30)(31), inhibits the pH 5.5 activity of the enzyme (Fig. 4) without affecting the pH 9.5 activity (data not shown). The KI for vanadate is approximately 11 pM, which is similar to its KI for the Na+ activity of the (Na'-Kf)-ATPase (30) and the Neurospora Mg2+-ATPase (29). In this experiment, inhibition by vanadate was never greater than 85%, probably due to contamination of the plasma membrane ATPase by the mitochondrial vanadate-resistant Mg"+-ATPase. Since vanadate has been proposed as a transition state analogue for proteinphosphate hydrolysis (33)  while that present after ATP chase might be caused by an occluded site on the protein due to the high concentration of crude membrane required for the phosphorylation. Both of the previous phenomena could also be explained by the action of another type of phosphorylation system or the presence of more than one enzyme in each peak.
Vanadate, an inhibitor of the yeast plasma membrane Mg"+-ATPase (Fig. 4), also inhibits the phosphorylation of the phosphoproteins found in peaks A and C. In addition, Fig. 6 shows that the phosphorylation of peak B is relatively resistant to vanadate. For the phosphorylation experiment, the concentration of vanadate was 1.6 mM, a concentration which inhibited 96% of the hydrolytic activity. At the high protein concentration used for the assay, micromolar concentrations of vanadate did not inhibit hydrolysis (see "Discussion").
Incubation of the labeled enzyme with hydroxylamine, which breaks down the aspartate-P, linkage of the Ca"-ATPase (34) and the (Na+-K')-ATPase (35,36), also causes the dephosphorylation of the phosphoprotein associated with the yeast plasma membrane Mg"'-ATPase. Incubation of the phosphoprotein in the acetate buffer used for hydroxylamine treatment did not cause the disappearance of any peaks (Fig.  7C), while incubation in this buffer in the presence of hydroxylamine caused the dephosphorylation of the intermediates present in all the peaks (Fig. 70). Also, the presence of hydroxylamine in the reaction mixture completely inhibited the hydrolysis activity of the enzyme (data not shown).
The phosphorylation reaction associated with the yeast enzyme, like the hydrolysis reaction, appears to be pH-dependent. Fig 8. shows that at pH 7.0 very little of peak C is formed and only the formation of peak A appears to be Mg'+dependent. The proteins in both peaks A and B, however, rapidly dephosphorylate in the presence of excess unlabeled ATP (Fig. 8C). Control experiments show that no label can be seen in the region of peaks A and B in the absence of enzymatic activity (Fig. 80) and that the large peak with a mobility of 0.7 is free ATP (Fig. 8E) as observed by other workers.
Due to the low background at pH 7.0, the time course of dephosphorylation was examined in detail at this pH. Fig. 8 shows that at 25°C the phosphate bound in peaks A and B could be chased by a 10-s pulse of excess unlabeled ATP. In order to examine the dephosphorylation more closely, the enzymatic reactions were slowed by incubation of the phosphorylation and dephosphorylation reactions at 0°C. As is shown in Table I, at the earliest time point, 3 s after the addition of unlabeled ATP, only 16 to 20% of the label remains bound to the protein. The results show that dephosphorylation is as fast as the turnover time for ATP hydrolysis, and they suggest that a phosphorylated intermediate is a possible intermediate in the mechanism of hydrolysis (see "Discussion").

DISCUSSION
The ATPase of the plasma membrane of the yeast, S. cereuisiae, has been characterized with regard to its sensitivity to inhibitors, pH dependence, response to added cations, and sedimentation properties. The enzyme was shown to be Mg2+dependent, oligomycin-resistant, have a pH optimum of 5.5, a maximum velocity of 0.42 pmol of ATP hydrolyzed/mg/min, and a K, of 1.7 mM. The bouyant density of the membrane fraction associated with the ATPase activity is 1.22 g/ml. The enzyme is not stimulated by exogenously added cations and is resistant to strophanthidin and sensitive to vanadate. The activity of this plasma membrane Mg"-ATPase is associated with the Mg"+-dependent formation of at least three stable phosphoproteins.
These proteins of M, = 210,000, 160,000, and 115,000 show rapid Mg"-stimulated phosphoryl-ation, dephosphorylate rapidly in the presence of excess unlabeled ATP, and exhibit kinetics of phosphorylation and dephosphorylation which are compatible with their involvement as intermediates in the hydrolysis reaction. The formation of two phosphoprotein peaks, A and C, is inhibited by vanadate, while the stability of all three phosphoprotein peaks is sensitive to hydroxylamine.
These preliminary studies on the activity of the microsomal Mg"-ATPase have shown the enzyme to be distinct from both the mitochondrial Mg"'-ATPase, and the well characterized metal ion pumps of the mammalian plasma membrane. The enzyme shows a pH optimum of 5.5 characteristic of the yeast plasma membrane ATPase and well removed from the pH optimum of 9 exhibited by the mitochondrial Fl ATPase in yeast (9). Also, the bouyant density of the ATPase associated membrane fraction is 1.22 g/ml, well within the range reported for the yeast plasma membrane (26,27), and sufficiently different from the bouyant density of 1.16 to 1.18 reported for the yeast mitochondrial membrane (26). The yeast plasma membrane enzyme also appears to be resistant to oligomycin and only partially sensitive to the mitochondrial inhibitors, efrapeptin and Dio 9 (28). In addition, the yeast enzyme is not stimulated by any other ions that stimulate the mammalian metal ion pumps, nor is it inhibited by strophanthidin, a specific inhibitor of the (Na'-K')-ATPase. Vanadate has recently been shown to inhibit plasma membrane ion pumps; in fact, the (Na'-K')-ATPase of both the red blood cell (31) and the kidney (30), the Mg"+-ATPase of Neurospora (29) and the Ca"'-ATPase of the red blood cell (32) are all vanadate-sensitive.
Interestingly, vanadate seems to inhibit only those ATPases in which a phosphorylated intermediate is formed during turnover (29)(30)(31)(32)37) and it appears to bind as a transition state analogue at the site where phosphate is released (30,38). The results presented here show that vanadate inhibits the yeast enzyme and would suggest that the Mg"'-ATPase described here also functions by formation of a phosphorylated intermediate. At least three rapidly forming, Mg"'-sensitive phosphoproteins, which phosphorylate and dephosphorylate rapidly enough to be intermediates, were found in the yeast plasma membrane.
The phosphorylation of two phosphoprotein peaks, A and C, was inhibited by vanadate, suggesting that one or both of these may be intermediates in the ATPase activity'. All three of the phosphoprotein peaks formed at pH 5.5 are sensitive to incubation in the presence of hydroxylamine (which dephosphorylates the phosphorylated intermediates of the Ca"-ATPase (34) and the (Na+-K')-ATPase (35, 36)), and hydroxylamine also inhibits the hydrolysis of the yeast plasma membrane enzyme. Taken together, these results suggest, but do not prove, that peaks A or C, or both, may be phosphorylated intermediates in the activity of the Mg'+-ATPase of the yeast plasma membrane. The view that the phosphoproteins found in the yeast plasma membrane are intermediates in the Mg"'-ATPase reaction must be taken with caution because the amount of phosphoprotein found is low, the phosphorylated intermediates cannot be directly linked to the ATPase reaction, and all the phosphoproteins do not fit the proposed role as intermediates.
The quantity of phosphorylated intermediate is low compared to that found for the other plasma membrane ion pumps. If the amount of bound phosphate is stoichiometric with the number of copies of the enzyme, the estimated turnover number is 1500 to 2000 s-' (using a specific activity of 0.2 pmol of ATP hydrolyzed/mg/min, and between 0.5 and 1.5 pmol of phosphate bound/mg of protein). This is considerably higher than the turnover of 100 to 500 s-' (39) for the metal ion pumps of the eukaryotic plasma membrane. This may be due to the fact that the dephosphorylation is not the rate-limiting step, the intermediates are not stable during isolation, or the formation of the phosphoprotein is reduced due to&he high amount of protein required for the assay.
A direct relationship between the phosphoproteins and ATP hydrolysis has not been demonstrated.
The use of a more purified enzyme preparation would allow one to resolve the ambiguities resulting from the use of a crude system (i.e. the incomplete dephosphorylation of the phosphoproteins seen in Fig. 5, the Mg"+ independence of the phosphorylation reaction for peak B protein seen in Fig. 8, and the high concentration of vanadate required to inhibit phosphorylation), and to definitely associate the phosphoprotein(s) with the ATPase hydrolytic activity. For example, the Mg'+ independence of the protein in peak B and its resistance to vanadate make it unlikely that peak B is a phosphorylated intermediate associated with the yeast plasma membrane enzyme.
The high concentration of vanadate required for the inhibition of hydrolysis in the phosphorylation assay may be indicative of the tendency of vanadate to bind to itself and to proteins. Vanadate has been shown to form oligomers with itself at low pH, to bind to proteins such as hemoglobin, and to be reduced to an inactive form under certain physiological conditions (31)." Considering the large amounts of crude membrane fraction in the phosphorylation assay, the high amount of vanadate may be necessary to maintain sufficient quantities of free vanadate in the reaction mixture.
The functional significance of the Mg"'-dependent plasma membrane ATPase has not been established, but it may be involved in cellular transport. The transport of amino acids, sugars, and other small metabolites in various fungi has been shown to involve proton gradients. In Neurospora, an ATPdependent electrogenic proton pump has been characterized and shown to be involved in the plasma membrane transport system (10)(11)(12). In the yeast S. pombe, intracellular ATP is required for active cellular uptake of small metabolites, and a plasma membrane Mg'+-ATPase has been characterized (9,40 (42). This study and other papers have described a Mg"+-dependent ATPase associated with the plasma membrane in S.
cereuisiae. This result and the involvement of proton gradients in cellular transport processes suggests the hypothesis that the yeast plasma membrane Mg"'-ATPase is an ion pump involved in the maintenance of the proton gradient necessary for active transport to occur in this organism. Although this hypothesis is attractive, a direct connection between ATPase activity and proton translocation has not been demonstrated in yeast. It is useful to compare the properties of the yeast Mg"-ATPase with other ATP cation pumps. These pumps fall into two broad classes, those that pump metal cations and those that pump protons. The metal cation pump ATPases have been shown to have phosphorylated intermediates involved in their activity, while no phosphorylated intermediate has been detected for the proton pumps. Recently, several ATPases have been observed and partially characterized which do not completely fit into either of these classes. The (H+-K')-ATP-