Solubilization by Lysolecithin and Purification of the Plasma Membrane ATPase of the Yeast Schizosaccharomyces

Purified plasma membranes of Schizosaccharomyces pombe were obtained by precipitation at pH 5.2 of a crude particulate fraction, followed by differential cen- trifugations and isopycnic centrifugation in a discon- tinuous sucrose gradient. The specific activity of the Mg+-requiring plasma membrane ATPase activity (EC 3.6.1.3) was enriched from 0.3 amol min-’ X mg-’ of protein in the homogenate to 26 in the purified mem- branes. The optimal conditions for solubilization of the ATPase activity by lysolecithin were found to be: 2 mg/ml of lysolecithin, a lysolecithin to protein ratio of 8 at pH 7.5, and 15°C in the presence of 1 mu ATP and 1 mu ethylenediaminetetraacetic acid. A 6- to ?-fold purification of the solubilized ATPase activity was ob- tained by centrifugation of the lysolecithin extract in a sucrose gradient. Part of the ATPase activity which was inactivated during the centrifugation in the sucrose gradient could be restored by addition of a micellar solution of 50 ag of lysolecithin/ml during the assay. Polyacrylamide gel electrophoresis in the presence of sodium dodecyl


Jean-Pierre
Dufour and Andre Goffeau From the Laboratoire d'Enzymologie, Universitk Catholique de Louvain, 1348 Louvain-la-Neuve, Belgium Purified plasma membranes of Schizosaccharomyces pombe were obtained by precipitation at pH 5.2 of a crude particulate fraction, followed by differential centrifugations and isopycnic centrifugation in a discontinuous sucrose gradient. The specific activity of the Mg+-requiring plasma membrane ATPase activity (EC 3.6.1.3) was enriched from 0.3 amol min-' X mg-' of protein in the homogenate to 26 in the purified membranes. The optimal conditions for solubilization of the ATPase activity by lysolecithin were found to be: 2 mg/ml of lysolecithin, a lysolecithin to protein ratio of 8 at pH 7.5, and 15°C in the presence of 1 mu ATP and 1 mu ethylenediaminetetraacetic acid. A 6-to ?-fold purification of the solubilized ATPase activity was obtained by centrifugation of the lysolecithin extract in a sucrose gradient.
Part of the ATPase activity which was inactivated during the centrifugation in the sucrose gradient could be restored by addition of a micellar solution of 50 ag of lysolecithin/ml during the assay. Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate of the purified enzyme showed only one band of M, = 106,000 stained with Coomassie blue. Another ATPase component of apparent molecular weight lower than 10,000 was stained by periodic Sehiff reagent but not colored by Coomassie blue. The purified enzyme was 85% inhibited by 50 PM N,W-dicyclohexylcarbodiimide and 94% inhibited by 53 ag of Dio-S/ml.
The plasma membrane fraction purified by differential and isopycnic gradient centrifugation of a subcellular homogenate of the yeast Schizosaccharomyces pombe exhibits Mg'+-dependent adenosine triphosphatase activity (1). The properties of this plasma membrane-bound activity have been characterized and shown to be different in several aspects from those of the mitochondria-bound ATPase (1). In particular, the optimum pH of the plasma membrane ATPase is much lower than that of the mitochondrial enzyme, so that the ratio of the activities measured at pH 6.0 and pH 9.0 grossly reflect the relative proportion of plasma to mitochondrial membranes (1). A substantial amount of experimental data indicates that the plasma membrane-bound ATPase of S. pombe is involved in the cellular transport of ions and metabolites (2)(3)(4). Rather similar plasma membrane-bound ATPase activities have been identified in other fungi: Saccharomyces cerevisiae (5-g), Can&da albicans (lo), and Neurospora crassa (11)(12)(13)(14). None of these plasma membrane ATPases have yet been purified.
* This is publication 1.521 of the EURATOM Biology Division. 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 USC. Section 1734 solely to indicate this fact.
Their subunit composition is therefore unknown and cannot be compared to that of bacterial, mitochondrial, or chloroplast ATPases nor to that of the mammalian (Na* + K')-or Ca'+stimulated ATPases.
Due to the proficient extractive and protective properties of lysolecithin, we have been able to solubilize and purify a yeast plasma membrane ATPase of high specific activity. The purified plasma membrane ATPase of S. pombe exhibits a much simpler subunit composition than the mitochondrial, chloroplast, or bacterial ATPases. Rather similarly to the mammalian (Na+ + K+)-stimulated plasma membrane or Ca'+-stimulated sarcoplasmic reticulum ATPases, one major polypeptide of M, = 100,000 + 5000 is observed in the purified yeast plasma membrane ATPase. The purified enzyme is activated by phospholipids and is sensitive to the inhibitors Dio-9 and ZV,N'-dicyclohexylcarbodiimide. MATERIALS AND METHODS Isolation of Plasma Membranes-Plasma membranes were prepared by a modification of the previously reported method (1). The subcellular homogenate was centrifuged twice at 1,000 x g for 5 min, the pellet being discarded each time. The supernatant was centrifuged for 5 min at 3,000 x g. Centrifugation at 15,009 x g for 40 min of the 3,000 X g supernatant yielded a pellet which was suspended to 3 mg of protein/ml in a medium containing 1 mM ATP, 1 mM EDTA, 10 mM Tris adjusted at pH 5.2 with CH&OOH and centrifuged at 7,500 x g for 45 s in a Sorvall HB4 rotor. The 7,500 x g supernatant was brought to pH 7.5 with NaOH and centrifuged at 45,000 x g for 12 min. The pellet was suspended to 2 mg of protein/ml in the 1.099 g X crnm3 sucrose layer, 10 mM Tris adjusted at pH 7.5 with CH:&OOH. Two milliliters of this suspension were laid on top of a discontinuous sucrose gradient made of 5 ml of the 1.25 g x cm-' sucrose layer followed by 3 ml of the 1.20 g x cmm3 sucrose layer and 2 ml of the 1.18 g x cm-" sucrose layer buffered with 10 mM Tris adjusted at pH 7.5 with CHXOOH.
After centrifuaation at 100.000 x P for 1.5 h in a Spinco R42. rotor, the membranes enriched in plasma membrane ATPase activity were recovered by centrifugation at 100,000 x g for 1 h of the 3 ml of the 1.20 g x crne3 sucrose layer previously diluted in 25 ml of 1 mru ATP, 1 mM EDTA, 10 mru Tris adjusted at pH 7.5 with CH:&OOH.
All operations were carried out at 4°C.

Solubilization
of PZasma Membrane ATPase and Concentration-A typical procedure for solubilization of plasma membrane ATPase was to suspend 1 mg of purified plasma membrane in 350 fi of the following solution referred to in the text as "solubilization buffer": 1 rnM EDTA, 1 mM ATP, 10 m&r Tris adjusted at pH 7.5 with CH&OOH. This suspension kept at 4'C was added to 4.65 ml of so&a&d solubilixation buffer containing 5 mg of lysolecithin and incubated for 10 min at 15°C. The suspension was centrifuged at 100,000 X g for 45 min at 15°C. The supematant containing the lysolecithin extract was concentrated to 1 ml by filtration through an Amicon CF 25 cone centrifuged at 1,500 x g for at least 30 min at 4% Adenosine Triphosphatase Assay-Unless otherwise indicated, the assays for ATPase were carried out by incubation of the sample at 30°C for 8 min in a final volume of 1.0 ml containing either 6 rnM ATP, 12 mM MgCL, 25 mru Tris adjusted at pH 6.0 with CH&OOH (for plasma membrane ATPase) or 6 mu ATP, 3 mM MgC12,25 mu Tris adjusted at pH 9.0 with NaOH (for mitochondrial ATPase). When indicated in the legends, an ATP-regenerating mixture containing 2 mu phosphoenolpyruvate and 5 ~sg of pymvate kinase (EC 2.7.1.40) (200 lU/mg) was used. In some experiments when low amounts of material were available, the incubation time was extended beyond 8 min to a maximum of 60 min. The reaction was stopped by the addition of 3 ml of 7% sodium dodecyl sulfate (w/v) as described by Dulley (15). Inorganic phosphate was measured as described by Pullman and Penefsky (16).
Protein Determination-The proteins were measured by the method of Lowry, with bovine serum albumin as standard (17). Standard curves were always carried out in the presence of the medium in which the sample was suspended.
Lysolecithin Solutions-Lysolecithin was suspended in solubilixation buffer. The suspensions were sonicated at room temperature for 30 s using the Virsonic cell disrupter, model 16-850, at 50% of maximal intensity. The critical mice&r concentration of lysolecithin was determined by three different methods: 1) measurement of the changes of the surface tension of an air-water interface as a function of the concentration of the lysolecithin in the bulk phase (18), using a Du Noiiy tensiometer model 8501 (A. K&s, Optischmechanische Werk-&&ten-Hamburg 39); 2) measurement of the fluorescence of perylene/lysolecithin mixtures at increasing lysolecithin concentrations (19) using a Zeiss ZFM, fluorimeter. 3) measurement of the changes of the absorbance at 542 nm of rhodamine 6G as a function of the concentration of lysolecithin as described by Bonsen et al. (20). The critical mice&r concentration of 1ysoIecithin in the solubiliition buffer was observed to be 8, 9.2, and 10 &ml using the fust, the second, and the third method, respectively, giving a mean value of 9.1 &ml. This value agrees well with the values reported by Saunders (21).

Sodium Dodeeyl Sulfate-Polyacrylamide Gel Electrophoresis in
Multiphasic Buffer Systems-Sodium dodecyl sulfate-gel electrophoresis was carried out as described by Neville and Giossman (22) using the Multiphasic Buffer System 54179 from the computer output of Jovin et al. (23). The gel contained 11% acrylamide and 0.1% bisacrylamide which is 11.1 T x 0.9 C using the notation of Hjerten (24).
The samples were precipitated by 10% tricbloroacetic acid for 30 min at 0°C and centrifuged twice for 2.5 min in an Eppendorf centrifuge, model 3200. The pellets were washed several times with ether, resuspended in 100 d of 1% b-mercaptoethanol, 2% sodium dodecyl sulfate, 1% glycerol, 0.005% bromophenol blue, 27 rnM HzSOI, pH 6.14, adjusted with Tris and heated at 100°C for 6 min. After electrophoresis, the gels were stained at room temperature with Coomassie blue and destained with methanol/acetic acid/water (5:1:5). Spectrophotometric scanning was carried out with the 580 to 650 nm filter of the Kipp and Zonen DD2 densitometer.

RESULTS
Purification of Plasma Membranes of High ATPase Actiu-@-One criticai factor for successful solubilization of the plasma membrane ATPase is to use enriched plasma membrane fractions of high specific pH 6.0 ATPase activity. Such plasma membrane fractions of S. pombe with high ATPase activity were obtained by selective pH precipitation followed by centrifugation through a discontinuous sucrose gradient. Table I shows the distribution of ATPase activities measured at pH 6.0 and pH 9.0 in the supernatant and the pellet of a low speed centrifugation after bringing a crude membrane fraction to the indicated pH. This procedure was inspired by the data of Fuhrman and Kramer (25) obtained with S. cerevtiiae.
As previously shown (l), the ATPase activity measured at pH 9.0 represents only the mitochondria-bound activity, while that at pH 6.0 expresses not only the plasma membrane-bound activity but also about one-third of the mitochondrial activity maximally expressed at pH 9.0. It can be seen in Table I that treatment at pH 5.4 or lower, aggregates preferentially the mitochon~a-Lund activity. From pH 5.4 and below the activity assayed at pH 6.0, in the supernatant, expresses essentially the plasma membranebound ATPase since the correction for residual mitochondrial activity amounts to less than 10% of the measured value. However appreciable inactivation of the plasma membranebound ATPase is observed at treatments below pH 5.0, as reflected by decreasing recoveries of the pH 6.0 ATPase. Treatment at pH 5.2 was chosen as the best compromise between yield and purification.
Under these conditions, the ratio of pH 6.0 to pH 9.0 activity was raised to 8.9 and 59% of the total pH 6.0 activity was recovered in the supematant. Fig. 1 shows that when the particulate fraction obtained by centrifugation of the supematant of a pH 5.2 treatment at Effects of pH on aggregation of plasma and mitochondrial membrane-bound ATPases One milliliter containing 3.76 mg of protein of the 15,000 x g pellet measured in the unfractionated sample. The recoveries were obtained obtained as described under "Materials and Methods" was added to by addition of the ATPase units in the pellet and supernatant com-15 ml of 10 nnu Tris, 1 rnM ATP, 1 mM EDTA, the pH of which was pared (in per cent) to the original ATPase units measured in the previously adjusted to the indicated value with acetic acid. After &&action&d sample. ATPase activity measured at pH 9.0 reflects centrifugation at 7,500 x g for 45 s in the Sorvall HB4 rotor, the only mitochondrial activity, while that measured at pH 6.0 reflects supematant was brought to pH 7.5 and the pellet was suspended in the nlasma-membrane activity plus 30% of the maximal mitochondrial 1 &l of the solubilixation buffer. The temperature was mai&.ained at activity, expressed at pH 9.0 &I the same sample (1). The specific 4'C during thii procedure. The ATPase assays were performed with activities of the unfractionated sample were 2.4 and 1.3 ~01 X mm-' the ATP-regenerating mixture. The ATPase units recovered in the x mg -' for pH 6.0 ATPase activity and mitochondrial ATPase supematant-were expressed in per cent of the original ATPase units activity respectively. Effect in crude membranes with followins DH values: Activity Specific ATPase activity (E"ol x min-' X -. 7 Eighty-five milligrams of protein of the 15,000 x g pellet treated at pH 5.2 as indicated under "Materials and Methods" and suspended in 2 ml of 10 KIM Tris, 1 mM ATP, 1 mM EDTA, 1.099 g x cm-" of sucrose adjusted at pH 7.5 with CHCOOH was laid on a discontinuous sucrose gradient as described under "Materials and Methods." After the centrifugation, the gradient was fractionated in 12 x 1 ml and the density of each even-numbered fraction was measured with a calibrated refractometer (Officine Galileo). ATPase assays at pH 6.0 (plasma membrane) and pH 9.0 (mitochondria) were carried out in presence of ATP-regenerating mixture.
The recoveries were 97 and 91% for the plasma membrane and mitochondrial ATPase activities, respectively. The protein recovery was 91%.
45,000 X g for 12 min is submitted to centrifugation at 100,000 X g through a discontinuous sucrose gradient, the plasma membrane ATPase activity equilibrates at the interface between the 1.20 and 1.25 g x cme3 sucrose layers.
The combination of the pH precipitation and centrifugation steps results in a lo-fold purification of the plasma membrane ATPase activity and a similar decrease of the mitochondrial one as shown in Table II. The plasma membrane ATPase specific activity attained 25.7 with a pH 6.0 to pH 9.0 ratio of 26.8. An appreciable increase of pH 6.0 ATPase specific activity is observed in the pellet obtained after the centrifugation at 45,000 X g of the 7,500 X g supernatant after pH 5.2 treatment. This indicates that the pH treatment releases a considerable amount of soluble proteins which were enclosed in the crude membrane fraction. These proteins which are present in the 7,500 X g supernatant but eliminated in the 45,000 x g supernatant do not contain plasma membrane ATPase activity.
The purification of the plasma membrane was monitored by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. It is apparent from the gel patterns illustrated in Fig. 2, A  component. A molecular weight of 100,000 f 5,000 was est,imated for this peptide by comparison of its mobility to that of proteins of known molecular weight (Fig. 2C).
Solubilization of Plasma Membrane ATPase-After unsuccessful attempts with several nonionic detergents (Triton X-100, Brij 35, Lubrol WX) or bile salts (sodium cholate and sodium deoxycholate) which were found to inactivate the ATPase activity, lysolecithin was finally used to solubilize the plasma membrane. The conditions of solubilization have, however, to be carefully controlled. Fig. 3A shows that, at a  . 3. Parameters of the solubilization by lysolecithin of the plasma membrane ATPase activity. A, effect of concentration of lysolecithin. The specific activity of plasma membrane-bound ATPase was 12.4. Aliquota of 36 4 of plasma membrane fraction (213 pg of protein) were added to solubilization buffer containing the indicated concentration of lysolecithm at the constant lysolecithin/protein ratio of 7. These suspensions were incubated for 10 min at 15°C and centrifuged at 100,000 x g for 45 min at 15*C in the Spinco R-40 rotor. The supernatants and the pellets, each resuspended in 500 $ of solubilisation buffer, were stored at 15 and O"C, respectively. The yield of solubiliition of ATPase activity (0) and of protein (U) are the ratio of ATPase units (or protein units) in the supematant to those iu the original purified membrane when the recoveries were less than 90% or higher than 110%. Otherwise, the ratio of ATPase units bilk&ion is thus function of the inactivation orthe stimulation of the ATPase activity as well as the effkiencs of the extraction. The recoveries were-determined as described under Table I. The same conventions were used in B and C. The recoveries of ATPase and constant lysolecithin/prot&n ratio of 7, solubilixation is obtained between 0.25 and 2 mg of lysolecithm/mI. Under these experimental conditions lysolecithin is in mice&r form since ita critical miceIIar concentration is 9.1 &ml as determined by the three different methods mentioned under "Materials and Methods." Such treatment solubilixes about 65 to 70% of the ATPase activity and 75 to 80% of the proteins. In addition, lysolecithin protects the plasma membrane ATPase activity during solubilixation since recoveries of 91 to 97% of ATPase activities were obtained. Fig. 38 demonstrates that the optimal ratio of 8 mg of lysolecithin/mg of protein solubilizes the highest amount of ATPase and protein &its. The extraction of ATPase activity is clearly bimodial with a partial extraction at a ratio of 4 and a more completed extraction at a ratio of 8 mg of lysolecithin/mg of protein.
The pH affected the solubilixation markedly. As shown in Fig. SC, alkalinixation of the solubilixation medium increases steadily the release of proteins. However the solubiition of the plasma membrane ATPase is optimal at pH 7.5 and ATPase activity is lost irreversibly below pH 6.0 and above the optimal pH of 7.5. Increasing the ionic strength, by increasing concentrations of buffer did not modify signiticantly the solubilixation of the ATPase activity. Similarly, prolongation of the incubation time up to 30 min had little effect (not shown).
The presence of 1 mM EDTA or 1 mu ATP in the solubil-i&ion medium increases significantly the specific activity of the solubilized ATPase (Table III). These effects are not additional but combined additions of EDTA and ATP are slightly more favorable than ATP alone. The presence of 1 mu ATP does also exhibit protective effects against inactivation during the subsequent purification of the ATPase (not Bhown).
The temperature does also influence markedly the solubil-i&ion which is incomplete at 4'C and is optimal and more specific at WC (not shown).
Purification of Plasma Membrane ATPase-Aa described protein units were of 91 to 110% and 89 to 108%, respectively. B, effects of the lysolecithin/protein ratio. The specific activity of plasma membrane-bound ATPase was 12.4. Aliquots of 85 pi of plasma membrane fraction (504 gg of protein for the first seven points) or 42 4 (249 I.L~ of protein for the last seven points) were added to solubil-i&ion buffer at each of the indicated lysolecithin/protein ratio. The final concentration of lysolecithin was 0.5 mg/ml of solubilization buffer. The suspension was shaken 10 min at 15'C and centrifuged at 100,096 x g for 1 h at WC in the Spinco R-40 rotor. The supernatants and the pellets resuspended in 560 ~1 of solubilisation buffer were stored at 15 and O"C, respectively. The recoveries of ATPase and protein units were of 89 to 110% and 89 to 99%, respectively. C, effects of pH. The specific activity of plasma membrane-bound ATPase was 12.4. One milliliter of solubilization medium contained 213 gg of protein and 1491 pg of lysolecithin. The pH was adjusted with CH&OOH or NaOH. Other conditions were as described under A. For protein units, the recoveries were of 91 to 101%. For the ATPase units, the recoveries were of 80,91,102,101,96,62,38, and 26% at pH 5.5,6.0,6X+ 7.0, 7.5,8.0,8.5, and 9.0 respectively.

III
Effects of ATF and EDTA on sohbilization by lysolecithin of the plasma membrane ATPase activity The specific activity of plasma membrane-bound ATPass was 9.5. The final concentration of lysolecithin was 1.5 mg/ml. The lysolecithin/protein ratio was 7. The pH was adjusted at pH 7.5 with CHCOOH. Each sample containing 213 gg of protein was incubated 10 mm in 1 ml of the indicated solubltion medium and centrifuged at 100,006 x g for 45 min in the Spine0 R-40 rotor. Other conditions were as described under Fig. 3A 6.5 +ATplm~ above, the optimal solubilixation of plasma membrane ATPase requires a high lysolecithin/protein ratio and therefore extensive dilution to maintain a relatively low concentration of lysolecithin. Concentration of the diluted lysolecithin extract was carried out by filtration through an Amicon CF 25 cone which retains molecules of molecular weight higher than 25,000. For volumes smaller than 3 ml, about 90% of the ATPase activity can be recovered by ultrafiltration at 15°C or below, but this yield is decreased to 70% when the filtration is carried out at 2O'C. It is also decreased when larger volumes are concentrated. The concentrated extract loses about 50% of its activity in 2 days at 4°C and is completely inactivated when exposed to higher temperatures. Fortunately, the concentrated extract is stable during storage at -180°C for 2 days or longer and can be thawed and frozen four to six times with loss of only 30% of its initial activity.
Because of the relative instability of the enzyme, it was necessary to use purification methods which can be carried in less than 24 h. Purification of the plasma membrane ATPase by chromatography through DE52 (Whatman), DEAE-Sepharose (Pharmacia), or DEAE-Bio-Gel (LKB) yielded variable results and at this stage, these procedures cannot be recommended. On the other hand, good results were obtained by centrifugation through a sucrose density gradient. As shown in Fig. 4, centrifugation through a linear 10 to 30% (w/w) sucrose gradient separates the ATPase activity from the bulk of contaminating proteins which are of lower sedimentation rate. Most of the ATPase activity put on the gradient was recovered between 1.070 and 1.078 g x cm-' of sucrose. No mitochondrial activity was detected. The recovery of 28% of the ATPase activity put on the sucrose gradient (assayed in the absence of lysolecithin) indicated that the plasma membrane ATPase activity was partially inactivated during the centrifugation. However the addition of lysolecithin during the assay restored the activity as shown in Fig. 4. The stimulation factor by addition of lysolecithin during the assay is more pronounced after purification in the sucrose gradient than before. In the experiment of Fig. 4, the stimulation was 1.2-fold for the concentrated extract and lo-fold for the purified enzyme. The amount of lysolecithin required to give halfmaximal activation was 6.5 pm01 of lysolecithin x mg-' of protein for the purified enzyme (Fig. 5). It must also be pointed out that lysolecithin does not restore the activity lost during concentration and aging of the lysolecithin extract which amounts to 60 to 70% of the ATPase activity initially solubilized. The plasma membrane ATPase peak fraction of the sucrose gradient was analyzed by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. Staining by Coomassie blue reveals the presence of only a single subunit as demonstrated in Fig. 6, showing a spectrophotometric scan and a photograph of a gel containing the purified ATPase. The molecular weight of the major subunit estimated by simultaneous electrophoresis of samples containing purified ATPase and marker proteins is 105,000. This molecular weight is identical to that which was predicted from the observation of the gel pattern during purification of the plasma membrane. In addition to this major band, a white opaque zone is observed slightly ahead of the tracking dye.
No periodic acid-Schiff staining was observed for the protein band; however, the opaque zone of high mobility was intensely colored red. After staining by Coomassie blue, the relative area of the M, = 100,000 peak was estimated in the plasma membrane and purified enzyme (Table IV). It can be concluded from this gross estimation that the M, = 100,000 peak is enriched 6-to 7-fold during purification through the sucrose gradient. Concomitant loss of activity already mentioned is responsible of the much lower increase of specific activity which reaches 35 pm01 X min-' X mg-' of 'protein in the purified enzyme (Table IV). This loss of activity and the unexpected simplicity of the subunit composition of the purified plasma membrane ATPase raises the suspicion that unidentified ATPase components could have been lost during the purification. It is very difficult to completely rule out this possibility; however, from the data of Table V   for N,N'-dicyclohexylcarbodiimid~ inhibition, nor that for inhibition by Dio-9 are lost in the purified enzyme. Actually, both inhibitions are more pronounced for the purified enzyme than for the plasma membrane-bound activity.

DISCUSSION
Purification and Solubilization of the Plasma Membrane-once solubilized, the plasma membrane ATPase loses activity during the purification procedure. It is therefore essential to start from plasma membrane fractions as purified as possible. It is of special importance to reduce the mitochondrial contamination. This is achieved quickly and conveniently by the procedure which is reported above. This plasma membrane preparation, which has proven adequate for extraction of the ATPase, contains more than 20 components which can be distinguished by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (see Fig.  2A). Such a preparation could therefore be of further use for identification and purification of other plasma membranebound components. It also provides plasma membrane vesicles which could be used to study transport functions and, for instance, to verify whether indeed the plasma membrane ATPase is an electrogenic proton pump maintaining a cellular membrane potential as previously proposed (4).
The main problem encountered during this study has been to extract the plasma membrane ATPase with minimal inactivation of the enzyme activity. This was achieved with lysolecithin. The ATPase extraction, however, is not selective since the specific activity of the lysolecithin extract is slightly lower than that of the native membranes, The low concentrations of lysolecithin and the high lysolecithin to protein ratio required for optimal solubilization cause an extensive dilution of the plasma membrane fraction. Under these rather unusual conditions, lysolecithin solubilizes the ATPase instead of stripping the membrane as in other preparations of mitochondrial or sarcoplasmic ATPases yielding particulate ATPase . lysolecithin complexes (26,2?). From the results of Fig. 3B, it can be proposed that the plasma membrane ATPase is a transmembranous protein since it is less efficiently solubilized than other proteins which might be more superficially located on the plasma membrane.
Properties of the Purified Membrane ATPase-This paper is the Fist report of purification of a plasma membrane ATPase from a fungi. In a two-step purification, we have succeeded in purifying an active soluble plasma membrane ATPase of specific activity above 30 pmol x min-' x mg-'.
The purified enzyme is highly stable during storage at -180°C but loses its original activity when stored unfrozen. In addition, considerable loss of ATPase activity is observed during the sucrose gradient centrifugation and might be explained as follows. The buoyant density of the lysolecithin micelles (or lysolecithin.lipid mixed micelles) is lower than that of the protein (or the protein.lysolecithin complexes). The two types of complexes are therefore separated during centrifugation (28) and a subsequent reduction of the protection of ATPase by lysolecithin follows. This loss of activity is not totally irreversible since addition of a micellar solution of lysolecithin restores at least partly the ATPase activity of the purified enzyme. The value of 6.5 pm01 of lysolecithin/mg of protein for half-maximal stimulation of the purified ATPase might be compared with those reported for the ATPase of beef heart submitochondrial particles which required 1.1 pm01 of lysolecithin/mg of protein for half-maximal activation (29), wheras 4 pmol of lysolecithin/mg was required for the plasma membrane ATPase complex of Escherichia coli. The variation of the extent of stimulation that we observe from one experiment to another might be attributed to variable quantities of residual lipids or to variable state of inactivation of the enzyme.
As revealed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the single-peptide structure of the yeast plasma membrane ATPase permits clear assessment of the homogeneity of the purified enzyme. The structure of the yeast plasma membrane ATPase is so different from that of the mitochondrial, bacterial, or chloroplast ATPases that the possibility for existence of common subunits are very reduced, even though the mitochondria and plasma membrane-bound enzyme share some common properties such as sensitivity to . The fact that sensitivity to these two inhibitors is conserved and even increased in the purified plasma membrane ATPase suggests that the enzyme structure has not been seriously altered during purification. It, however, must be mentioned that the pnrified plasma-membrane ATPase contains a component of high electrophoretic mobility which is stained by periodic acid-Schiff but not by Coomassie blue. This might indicate glycolipidic, glycoproteic, or lipidic material. Since the mito-chondriaI ATPase proteolipid responsible for binding of N,iV'dicyclohexylcarbodiimide exhibits similar properties (30, 31)) it cannot be excluded that the plasma-membrane ATPase unidentified component is similar to the mitochondriaI proteolipid.
It is of interest to compare the yeast plasma membrane ATPase with transport ATPases from other sources. The mammahan cell ATPase has a major component of M, = 90,090 to 1OO,ooO, it has also a smaher peptide with nnknown functions and ita sensitivity to ouabain inhibition and stimulation by (Na+ + K') seems not to be shared by the yeast enzyme (1). The ATPase from the sarcoplasmic reticulum also is composed of only one M, = 100,000 peptide which is solubihzed by neutral detergents, but it clearly differs from the yeast enzyme in its stimulation by Ca" (1). The availability of a pnrified and active plasma membrane ATPase of S. pombe obtained by the simple and reproducible purification procedure presented here, should now permit further molecular characterization of the purified enzyme as weII as reconstitution studies into artificial membranes. It might be expected that the possibilities of physiological and genetic manipulations offered by yeast cells are going to be of great advantage in the study of the structure and function of this plasma membrane enzyme.