Properties of H+-translocating Adenosine Triphosphatase in Vacuolar Membranes of Saccharomyces cerevisiae*

The properties of M8+-ATPase in the vacuole of Saccharomyces cerevisiae were studied, using purified in- tact vacuoles and right-side-out vacuolar membrane vesicles prepared by the method of Y. Ohsumi and Y. Anraku ((1981) J. BioL Chem. 256, 2079). The enzyme requires M&+ ion but not Ca2+ ion. Cu2+ and Zn2+ ions inhibit the activity. The optimal pH is at pH 7.0. The enzyme hydrolyzes ATP, GTP, U T P , and CTP in this order and the K, value for ATP was determined as 0.2 m ~ . It does not hydrolyze ADP, adenosyl-5'-yl imido- diphosphate, orp-nitrophenyl phosphate. ADP does not inhibit hydrolysis of ATP by the enzyme. The activities of intact vacuoles and of vacuolar membrane vesicles were stimulated 3- and 1.5-fold, respectively, by the protonophore uncoupler 3,5-di-tert-butyl-4-hydroxy- benzilidenemalononitrile and the K+/H' antiporter ionophore nigericin. Sodium azide at a concentration ex- erting an uncoupler effect also stimulated the activity. The activity was sensitive to the ATPase inhibitor N,W-dicyclohexylcarbodiimide, but not to sodium vanadate. The ATP-dependent formation of an electrochemical potential difference of protons, measured by the flow-dialysis method, was determined as 180 mV, with con- Recently, we established a procedure for preparing right-side-out vacuolar membrane vesicles of high purity from cells of the yeast Saccharomyces cerevisiae and showed that the vesicles catalyze active arginine transport which is driven by an electrochemical potential difference of protons formed by ATP hydrolysis (1). Subsequent studies on ATP hydrolysis by intact vacuoles and vacuolar membrane vesicles indicated the presence of a new Mg2'-ATPase with catalytic site This paper describes the properties of vacuolar membrane-bound M8+-ATPase and its characteristic function as a H'-translocating ATPase. The Mg2+-dependent, H+-translocating ATPase of the vacuoles was found to be a new specific marker of these organelles and to differ from mitochondrial ATPase (2) and plasma membrane-bound ATPase (3,4). mCi/mmol) to the upper chamber, 4 mM ATP was added to start the reaction. Fractions of 2.1 ml was collected and their radioactivity was determined in a liquid scintillation counter with a toluene-Triton X- 100 scintillator. The internal water space of vacuolar membrane vesicles under standard assay conditions was determined as 5.2 d/mg of membrane protein by the inulin method (17).

+ Recipient of a grant from the Toray Science Foundation of Japan.
(40 mCi/mmol) and [I4C]KSCN (60 mCi/mmol) were purchased from Amersham. 4-Methylumbelliferyl-a-~-mannopyranoside was obtained from Koch-Light Laboratories, England. SF6847' was a gift from Dr. Y. Nishizawa of Sumitomo Chemicals, Osaka. Other chemicals used, which were commercial products of analytical grade, were described in the previous paper (1). The haploid strain of S. cereuisiae, X2180-1A, from the Yeast Genetic Stock Center, Berkeley, was used (1). A mitochondrion-defective strain X2180-1A(p-) was isolated by the method of Wilkie (5) and the defect of mitochondrial DNA of the strain was examined by the method of Williamson and Fennel1 (6). Cells were grown aerobically in YEPD medium containing 1% yeast extract (Difco), 2% polypeptone, and 2% glucose at 30 "C, and harvested at a cell density of 3 X IO7 cells/ ml of medium.
Preparations of Intact Vacuoles, Vacuolar Membrane Vesicles, and Submitochondrial Particles-Intact vacuoles and vacuolar membrane vesicles were prepared by the method of Ohsumi and Anraku (1). (For details, see Scheme 1.) Unless otherwise noted, they were prepared from strain X2180-1A. The vacuoles were suspended in buffer B containing 10 mM MES/Tris (pH 6.9), 0.5 mM MgClz, and 8% Ficoll (I), and promptly used for assays of Mg2+-ATPase and other enzyme activities. Vacuolar membrane vesicles suspended in buffer C (Scheme l), were used immediately as mentioned above, or frozen at -80 "C for a month before use without significant loss of ATPase activity. Submitochondrial particles were prepared by the method of Tzagoloff and Meagher (2).
Assays of Mg+-ATPase and Other Enzymes-The standard reaction mixture for assay of Mg*+-ATPase (0.6 ml) contained 10 mM MES/Tris, pH 7.0,4 mM ATP, 4 mM MgCL, 25 mM KC1, and enzyme. For assay of Mg2+-ATPase in intact vacuoles, 8% Ficoll was added to the standard mixture. Where indicated, the inhibitor in ethanolic solution was added to the mixture at a final concentration of ethanol of less than 1% (v/v). Incubation was for IO min at 30 "C, and the reaction was stopped by adding 0.35 ml of 15% sodium dodecyl sulfate. Inorganic phosphate liberated was measured by the method of Futai et a!. (7) and 1 unit of enzyme was defined as the amount liberating 1 pmol of inorganic phosphate/min under the standard conditions described above. MgZ+-ATPase activity increased linearly with up to 200 pg and 50 pg of protein per reaction mixture with intact vacuoles and vacuolar membrane vesicles, respectively. a-Mannosidase (8), glucose-6-phosphate dehydrogenase (9), succinate dehydrogenase (IO), NADPH-cytochrome c reductase (Il), chitin synthetase (12), and alkaline phosphatase (13) were assayed by published methods. Protein was determined by the method of Lowry et al. (14).
Determination of Quinacrine Quenching-The method of Tsuchiya and Rosen (15) was used with a modification. The reaction mixture (2.5 m l ) , which contained 50 mM Tricine/NaOH, pH 7.5, 5 mM MgCIz, 2 p~ quinacrine, and 180 pg of protein of membrane vesicles, was incubated in a quartz cuvette in a Hitachi-MPF-4 spectrofluorometer at 25 "C for 2 to 5 min. To this mixture, ATP (0.5 mM) was added to start quenching. The excitation and emission wavelengths used were 365 nm and 451 nm, respectively. mCi/mmol) to the upper chamber, 4 mM ATP was added to start the reaction. Fractions of 2.1 ml was collected and their radioactivity was determined in a liquid scintillation counter with a toluene-Triton X-100 scintillator. The internal water space of vacuolar membrane vesicles under standard assay conditions was determined as 5.2 d/mg of membrane protein by the inulin method (17).

RESULTS
Purity and Yield of Vacuoles-The distributions of marker enzymes (a-mannosidase for vacuoles (18), glucose-6-phosphate dehydrogenase for cytosol (19), succinate dehydrogenase for mitochondria (20), NADPH-cytochrome c reductase for microsomes ( l l ) , and chitin synthetase for plasma membrane (21)) were examined by measuring enzyme activities in the vacuole and spheroplast lysate fractions. Table I shows that the recovery of a-mannosidase activity in the vacuole fraction was about 13% and that its specific activity in this fraction was increased 28.5-fold. The recoveries of the four other marker enzyme activities in this fraction were found to be less than 0.1% of their total activities. These results indicate that the vacuole fraction obtained was virtually free from mitochondria and other membranous organelles.
Intactness of Vacuoles-Alkaline phosphatase is known to be localized in the internal space of vacuoles (22). The vacuole       fraction showed a latent phosphatase activity which was activated maximdy 4.5-fold on addition of Triton X-100 or cholate (Table 11). This indicated that the vacuoles were intact but that the vacuolar membranes were injured by detergents, resulting in breakage of the latency for alkaline phosphatase. The membrane-bound a-mannosidase activity was activated slightly by the detergents used, indicating that the substrate 4-methylumbelliferyl-a-~-mannopyranoside used was partly permeable through the membrane.
Mg2+-ATPase activity was detected in the vacuoles with A T P -M e as an impermeable substrate and this activity was inhibited by a high concentration of detergents (Table 11). This suggests strongly that the catalytic site of the enzyme in intact vacuoles is exposed to the outer surface of vacuolar membranes.
MF-ATPase as a Marker of Vacuolar Membranes-Intact vacuoles were treated hypotonically and disrupted me-  . We found that 80% and 96% of the Mg2+-ATPase and amannosidase activities, respectively, but less than 5% of the alkaline phosphatase activity, were recovered in the vacuolar membrane vesicles (Table 111). Furthermore, the effects of Triton X-100 and cholate on the Mg2+-ATPase of the membrane vesicles were the same as in intact vacuoles. From these observations we concluded that Mg+-ATPase, a new membrane-bound marker of vacuoles, has a catalytic site exposed to the cytoplasm and that the vacuolar membrane vesicles prepared as described above are right-side-out, consistent with morphological and functional evidence reported previously (1). Properties of M2'-ATPase-The enzyme required Mg2+ ion for ATP hydrolysis (Fig. lA) . The optimal ratio of ATP to M$+ of 1 indicated that an ATP-Mg2+ complex is substrate for the enzyme. Ca2+ ion had no effect on the activity. The optimal pH of the enzyme was detemined to be pH 7.0 (Fig.   1B). The enzyme hydrolyzed ATP and three other ribonucleoside triphosphates, GTP, UTP, and CTP with this order of preference (Table IV). ADP and p-nitrophenyl phosphate, which were not hydrolyzed by the enzyme, did not inhibit the activity. The K , value for ATP was determined as 0.2 mM (Fig. 2), which is €!-fold smaller than that of the ATPase of plasma membranes of S. cereuisiae (3). Fig. 3 shows the effects of various cations on Mg2'-ATPase. The activity was not affected by Kf or Na+ (data not shown), but was inhibited strongly by Cu2+ and Zn2+ ions. These cations inhibited the activity noncompetitively; the KlI2 values of inhibition by Cu2+ and Zn2+ ions were determined as less than 10 p~ and about 20 p~, respectively. Ca2+ and NH,' ions stimulated the activity at concentrations of more than 0.1 mM. We are now investigating these stimulating effects with respect to energy coupling of active cation transport by the vesicles.
The activity was inhibited by DCCD and the K m value of inhibition was determined as 6 p~ (0.15 pmol of DCCD/mg of membrane protein) (Fig. 4). However, unlike the H+-translo-  cating ATPase of submitochondrial particles from S. cereuisiae (21, the Mg2'-ATPase of the vacuoles was not inhibited by sodium azide and was in fact stimulated at concentrations of more than 1 mM sodium azide, probably due to its uncoupler effect (23). Sodium vanadate, which is reported to be an inhibitor of Mg2'-ATPase in the plasma membranes of S. cereuiszae (3), did not inhibit the activity noticeably (Fig. 4).
Evidence that Mg2"ATPase of Vacuoles is W-Translocating ATPase-The Mg2+-ATPase activities of intact vacuoles and vacuolar membrane vesicles were stimulated 3-and 1.5-fold, respectively, by the protonophore uncoupler SF6847 (24,25) and the K+/H+ antiporter ionophore nigericin (Table   V), indicating that these reagents decreased the proton gradient formed by ATP hydrolysis. Valinomycin had no effect on the activities.
ATP hydrolysis-dependent formation of a proton gradient was directly demonstrated by recording the change in quenching of quinacrine fluorescence. The fluorescence signal of quinacrine was quenched by incubating the vacuolar membrane vesicles from strain X2180-1A ( p -) with ATP, reflecting uptake of protons and formation of ApH (Fig. 5). These results c o n f m those on the quenching of 9-aminoacridine fluorescence reported previously (l), and indicate that the uptake of protons is coupled with ATP hydrolysis by Mg2+-ATPase. ATP-dependent alkalinization of reaction mixture containing vesicles and ATP in 2 mM glycylglycine buffer, pH 6.3, was also observed with a pH electrode (data not shown).
The electrochemical potential difference of protons across the vacuolar membrane generated upon ATP hydrolysis was determined quantitatively by the flow-dialysis method with ['4C]methylamine for measuring the formation of ApH and with [14C]KSCN for measuring the membrane potential (Fig.   6). The AjiW thus calculated (16) was 180 mV, with contribution of 1.7 pH units, interior acid, and of a membrane potential of 75 mV, interior positive.

DISCUSSION
The fist half of this paper describes biochemical criteria for the preparations of intact vacuoles and vacuolar membrane vesicles according to the marker concept of cellular organelles of de Duve (26). Our results indicate that intact vacuoles, which had high a-mannosidase activity and were virtually free from other cellular organelles, were obtained reproducibly, judging by the recovery of marker enzymes (Tables I and 11), and that they were converted to right-side-out vesicles (1) with high recovery of a-mannosidase and Mg2+-ATPase by brief hypotonical treatment and centrifugation (Table I11 and Scheme 1).
We found a new Mg2'-ATPase in the vacuolar membranes and concluded that it is an important marker enzyme of these organelles. The characteristic properties of the enzyme are summarized in Table VI in comparison with those of mitochondrial and plasma membrane Mg"-ATPase of S. cereuisiae. It is clear that vacuolar membrane ATPase differs from the mitochondrial enzyme in its pH optimum and sensitivities to oligomycin, sodium azide, and CU'+ ion. In addition, the vacuolar membrane enzyme is not inhibited at all by ADP. It also differs from the plasma membrane enzyme (3) with respect of its pH optimum, K , value for ATP, and sensitivity to sodium vanadate.
We obtained evidence that the vacuolar membrane Mg'+-ATPase is characterized as a H+-translocating ATPase and generates an electrochemical potential difference of protons of 180 mV, interior acid, across the membrane (Table V and  Submitochondrial particles were used for assays. Not tested. extents of inhibition of Mg2'-ATPase activity by DCCD and Cu2' ion and of arginine uptake by the vesicles (1). These facts indicate that the enzyme is a fundamental energy-donating system for various specific antiporters for amino acids (1) and Ca'+ ion' in the vacuolar membrane. The proteinchemical properties of the enzyme are under investigation in our laboratory.