Reconstitution of the Na'K' Pump of Ehrlich Ascites Tumor and Enhancement of Efficiency by Quercetin*

Plasma membranes from Ehrlich ascites tumor cells were solubilized by octylglucoside in the presence of phospholipids. The Na+K'-ATPase was purified from this extract by adsorption and elution from thiol-Seph-arose 4B. The enzyme (specific activity, 7 pmoles of ATP hydrolyzed min" mg of protein") was reconstituted into liposomes by the octylglucoside dilution pro- cedure. An ATP-dependent Na+ intlux with low efficiency was observed. On addition of appropriate amounts of quercetin, the Na' flux/ATP hydrolysis ratio was increased from 0.4 to 1.4. different

The high aerobic glycolysis of tumors (1) is supported by different ATPases (2). In Ehrlich ascites tumor cells, the sensitivity of glycolysis to ouabain pointed to the Na'K'-ATPase as the major supplier of ADP and P, (3). The question of whether there was increased pump activity or an inefficiently operating pump was more difficult to answer. The inhibition of glycolysis by quercetin and other flavonoids at concentrations that did not inhibit Rb' uptake (4) suggested an inefficient pump, particularly since these flavonoids increased the efficiency of the mitochondrial proton pump by substituting for the ATPase inhibitor a regulatory subunit of the mitochondrial proton pump (2, 5 ) . However, the observation that quercetin a t very low concentrations inhibits lactate export in tumor cells (6) complicated the interpretation of the mode of action of these versatile compounds.
It therefore became imperative to test the effect of quercetin on the Na'K' pump itself. Studies of purified Na'K'-ATPase from electric eel localized the action of quercetin to the step E2 -P s E + Pi (7). Both formation and hydrolysis of E , -P were highly sensitive to quercetin while the formation of E , -P with ATP was unaffected. The action of quercetin again resembled that of a regulatory subunit that protects the enzyme against illicit entry of water as in the case of the E subunit of chloroplast ATPase (8).
We now wish to report the purification of the Na'K'-ATPase from Ehrlich ascites tumor cells and its incorporation into liposomes. The reconstituted proteoliposomes catalyzed an ATP-dependent influx of Na'. The efficiency of pumping * This investigation was supported by Grants CA-08964 and CA-14454, awarded by the National Cancer Institute, Department of Health, Education, and Welfare, and Grant BC-156 from the American Cancer Society. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisernent" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. was very low and was increased severalfold by addition of quercetin.
Plasma membranes were isolated from Ehrlich ascites tumor cells according to the procedures of Brunette and Till (10) with the following modifications. The cells were swollen in 10 mM CaC1. (instead of 1 mM ZnCIl), and 100 m~ CaC12 was substituted for 10 mM ZnCIP in the preparation of the polyethylene glycol-Dextran twophase system. The membranes were stored at -90°C a t a protein concentration of 6 to 10 rng/ml.
The Na'K'-ATPase was extracted by incubating 5 mg of plasma membranes in 1 ml of medium containing 2% octylglucoside (pH 7.4), 400 pg of sonicated asolectin, and buffered sucrose (40 mM 4-(2-hydroxyethy1)-1-piperazineethanesulfonic acid (Hepes)-HOH (pH 7.4), 200 mM sucrose, and 6 mM MgClr) at 4OC for 10 min. The suspension was centrifuged a t 30,000 rpm (50,000 X g) for 20 min at 0°C in a Beckman Ti-50 rotor. The supernatant, containing most of the Na'K'-ATPase activity, was then centrifuged at 48,000 rpm (144,000 X g) for 2 h at 0°C in a Beckman type 65 rotor. Most of the Na'K'-ATPase activity was found in the pellet, which was resuspended in 0.2 ml of buffered sucrose, 0.1% octylglucoside.
The Na'K'-ATPase was further purified by chromatography on a thiol-Sepharose 4B column (0.5 X 2 cm) which was equilibrated with buffered sucrose plus 0.1% octylglucoside. The enzyme (10 mg) was applied to 2.5 ml bed volume of the gel. The column was washed with 25 ml of buffered sucrose plus 0.1% octylglucoside and the Na'K'-ATPase was eluted with a linear gradient of 0 to 25 mM dithiothreitol in buffered sucrose plus 0.1% octylglucoside.
Na'K'-ATPase activity was measured with [y."P,]ATP as described (9). Reconstitution of the Na'K'-ATPase was by octylglucoside dilution (11). Measurement of Na' uptake into the proteoliposomes (12) was performed by incubating 0.05 ml of reconstituted vesicles (generally containing about 10 pg of protein) with 0.06 ml of 75 mM &Sod, 50 mM Na2S04, 50 mM imidazole-HSO, (pH 7.4). 20 mM P-mercaptoethanol, containing 5 pCi of "Na/ml for 1 min a t 37°C. The assay was started by the addition of 0.005 ml of a 100 mM Tris/ATP, 100 mM MgSO, solution and the mixture was incubated for 2 min a t 37°C. An aliquot of 0.1 ml was removed and passed through a Tris Dowex column (6 X 0.3 cm) and counted in Liquiscint as described previously (12). When the Na'/ATP ratio was determined, 1 mM ouabain was present and the final volume was 0.129 ml. Controls were adjusted to the same volume with water. Protein was determined according to Lowry et al. (13). Sodium dodecyl sulfatepolyacrylamide gel electrophoresis was performed essentially as described by Chua and Bennoun (14) using a 7.5 to 15% linear gradient of acrylamide. The separating gel was 10 cm long with a 2-cm stacking gel, and 3 6 p-mercaptoethanol was substituted for dithiothreitol.
Electrophoresis was carried out a t 20 mA for 6 h. After fixation in

TABLE I
Purification of ascites tumor Na'K'-ATPase The purification and assays were performed as described under "Materials and Methods." Specific activities were calculated by subtracting ouabain-insensitive activitv from the total activitv. of the Na'K+-ATPase. Heconstitution was done with asolectin by the octylglucoside dilution method (11) at the various detergent concentrations and with varying protein and phospholipid concentrations.

RESULTS AND DISCUSSION
in 10% isopropyl alcohol, 10% acetic acid, and 0.01% Coomassie Brilliant Blue R-250 for 2 h. Gels were then destained with 5% ethanol The recoveries and specific activities of fractions during the and 10% acetic acid. As molecular weight markers, /3-galactosidase purification Of the Na'K"ATPase are shown I* On (130,000), bovine serum albumin (68,000), ovalbumin (44,000). and sodium dodecyl sulfate-polyacrylamide electroPhoresis gels soybean trypsin inhibitor (20,000) were used.
( Fig. I), there were two major bands in the best preparation  Reconstitution was performed by the octylglucoside dilution procedure (11). The effects of varying phospholipid/protein ratios and optimal detergent concentration for reconstitution are shown in Fig. 2. Simultaneous measurements of ATPase activity and Na+ flux in the presence of ouabain, which inhibited unincorporated ATPase, but not the inverted ATPase, revealed a low efficiency of pump action. Values of 0.3 to 0.5 for the Na+/ATP ratio were observed compared to ratios of 1.3 to 1.5 for a reconstitution of a Na'K'-ATPase from electric eel or from mouse brain. As shown in Table I1 and Fig. 3, the efficiency of pumping with the tumor enzyme was greatly increased by the addition of appropriate concentrations of quercetin. It should be noted that the amount of quercetin added is expressed in terms of the phospholipid plus protein concentration, since both interact with the bioflavonoids and influence the effective concentration. Quercetin had little or no effect on the Na'/ATP ratio in proteoliposomes reconstituted with active electric eel or mouse brain Na+K+-ATPase (Table 11, Experiment 2).

Na+/ATP ratios of vesicles reconstituted with Na+K+-ATPases from tumor, brain, and electric eel
Although in this paper we have reported only values for the 2-min flux of 22Na+, we have established that the efficiency of pumping in the absence and presence of quercetin remains constant up to 10 min. The possibility of a Na'/Na' exchange reaction contributing to the 'lNa' flux measurements was ruled out by substituting either choline chloride or choline sulfate for the internal NaCl. Within 15%, the 'lNa' flux data  Table I1 for Experiment 1.
were identical. It should be noted that the "ouabain-insensitive ATPase" of reconstituted vesicles is completely inhibited by ouabain which is incorporated inside the vesicles.
These experiments strongly support the hypothesis that the Na+K'-ATPase in these tumor cells is defective and operates inefficiently. There are several possibilities with respect to the type of lesion. The most likely and most easily explored explanation is a defective regulatory subunit analogous to the E subunit of the H'-ATPases. During the past years, we have searched for such a subunit without success. However, we have used the highly purified eel enzyme in the search for an ATPase inhibitor and this may have been an inadequate assay system. Now that highly purified preparations of the tumor enzyme are available, we can resume our search with better hope of success. It is encouraging that a natural inhibitor of the Na'K'-ATPase has been discovered recently (15,16). According to the published data (15), the fraction isolated from brain does not seem to conform to the expected properties of a regulatory subunit since it inhibited ATPase activity and ion transport to the same extent, an action similar to that of ouabain. However, further tests are required to establish whether a differential effect can be induced.