Purification of the (Ca2’-Mg2’)-ATPase from Human Erythrocyte Membranes Using a Calmodulin Affinity

The (Ca’+-M&+)-ATPase from human erythrocyte membranes has been solubilized in Triton X-100 and purified on a calmodulin affinity chromatography column in the presence of phosphatidylserine, to limit the inactivation of the enzyme. The enzyme was purified at least 150 times when compared with the original ghosts and showed a specific activity of 3.8 ~mol=mg-‘*min-L. In sodium dodecyl sulfate-polyacrylamide gels, a single major band was visible at a position corresponding to a molecular weight of about 125,000; a minor band (11% of the total protein) was present at a position corresponding to M, = 205,000. Upon incubation of the pu- rified preparation with [32P]ATP, both bands were phosphorylated in proportion to their mass, suggesting that both were active forms of purified ATPase.

The (Ca'+-M&+)-ATPase from human erythrocyte membranes has been solubilized in Triton X-100 and purified on a calmodulin affinity chromatography column in the presence of phosphatidylserine, to limit the inactivation of the enzyme. The enzyme was purified at least 150 times when compared with the original ghosts and showed a specific activity of 3.8 ~mol=mg-'*min-L. In sodium dodecyl sulfate-polyacrylamide gels, a single major band was visible at a position corresponding to a molecular weight of about 125,000; a minor band (11% of the total protein) was present at a position corresponding to M, = 205,000. Upon incubation of the purified preparation with [32P]ATP, both bands were phosphorylated in proportion to their mass, suggesting that both were active forms of purified ATPase.
The (Ca'+-Mg'+)-ATPase of the erythrocyte membrane is generally accepted as the enzyme responsible for maintaining the concentration of Ca2+ inside the cell at levels much lower than in the environment (1). Molecular studies of the enzyme have been hindered by the fact that it represents only a minor fraction of the total protein of the membrane and by the fact that, in purification attempts, it co-eluted with Band 3, the most abundant protein component of the erythrocyte membrane. Isolation and purification attempts have nevertheless been made, with the aid of either Triton X-100 or deoxycholate (2-5), and have yielded partially purified soluble fractions the specific activity of which was 20 to 155 times higher than in the starting membrane material. The specific activities reported for purified preparations ranged from 0.15 to 3.1 prnol. mg-' . min-', as compared with the usual value of 0.005 to 0.05 for whole membranes. The most highly purified form previously reported contained three proteins in approximately equal amounts; only one of these proteins was phosphorylated (4). Studies on the partially purified soluble enzyme have indicated that it has a molecular weight of between 125,000 and 150,000 (2,4) and have established a specific requirement for acidic phospholipids (6). Starting from partially purified preparations, it has also been possible to reconstitute the * The work described has been made possible by support from Grant N4. 3.597-1.75 from the Swiss National Funds, Grants AM 19785 and AM 21820 from the National Institutes of Health, and the Mayo Foundation. 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 U.S.C. Section 1734 solely to indicate this fact.
(Ca" -Mg"')-ATPase activity (2) and the transport of Ca"' (5) in artificial phospholipid vesicles. Calmodulin, the ubiquitous modulator of Ca"'-dependent functions, has been shown (7-9) to activate both the (Ca"-Mg'+)-ATPase and ATP-dependent Ca2+ transport in whole erythrocyte membranes. The ability to interact with calmodulin is retained by the solubilized (Ca"-Mg"')-ATPase (10). This ability has been exploited in the present work. An affinity chromatography column containing calmodulin has been prepared, and the Ca"+-dependent formation of a calmodulin. ATPase complex in the column has been used to separate the (Ca'+ -MgZ+)-ATPase from the other membrane proteins present in the ghost solubilizate.
The specific activity of the enzyme purified in this way is at least 150 times higher than that of the original ghost membranes, and it consists primarily of a single phosphorylatable protein.

MATERIALS AND METHODS
All reagents were of the highest purity available. Phosphatidylserine and lecithin were grade I, obtained from Lipid Products, South SDS-polyacrylamide gel electrophoresis of the ,r'P-labeled enzyme on 5% gels of pH 6.0 was done as described (3). The gels were then frozen on dry ice, and pairs of l-mm slices were incubated overnight with shaking in 1 ml of 0.5% SDS, 10 mM Tris-Cl, pH 9, at 4O'C. After addition of 10 ml of Insta-Gel (Packard) the vials were counted in a Nuclear Chicago scintillation counter. the elution buffer was replaced by 5 mM EDTA, a small peak of material absorbing at 280 nm was eluted, coincident with about 19% of the total (Ca"'-Mg"')-ATPase activity loaded on the column.

Isolation
Proceclure-The results of a typical isolation experiment are shown in Fig. 1 and Table  I. The supernatant of Triton X-IOO-solubilized ghosts contained 21% of the total ghost protein and 277% of the final (Ca"'-Mg"')-ATPase activity of ghosts. Such a stimulation of the ATPase was always seen when calmodulin-deficient ghosts were solubilized by Triton X-100. When ghosts from which calmodulin had not been removed were solubilized, the ATPase activity was depressed instead (3).

In erythrocyte ghost membranes, a certain amount of Mg'+dependent and Ca'+-independent
ATPase activity is always observed. In the experiments described here, this Mg"+ ATPase was about 5% of the (Ca"-Mg"')-ATPase.
It may be significant that the activity eluted from the column in the presence of Ca'+ contained a higher percentage of Mg"'-ATPase activity: 17.6%.
When the isolation was carried out as described above. but in the absence of phosphatidylserine in the buffers and in the supernatant, the yield of activity in the peak eluted by EDTA was only 5%. However, this enzyme preparation could be stimulated up to four times by the addition of 100 pg of phosphatidylserine during the assay. When lecithin instead of phosphatidylserine was present during the isolation procedure, only 4.4% of the activity was eluted by EDTA. Characterization of the Isolated Enzyme-Protein determinations of the material eluted by EDTA gave a total of about 0.09% of the ghost protein used as starting material. The specific activity of the most active fraction was determined to be 3.8 )rmol/mg/min, at 37°C and in the presence of However, the purified enzyme could not be Purification of (Ca2+-Mg+)-ATPase from Erythrocyte Membranes 9957 stimulated by added calmodulin. The reasons for the absence of stimulation are currently being investigated. One possibility that is being explored is that some calmoduhn might elute with the enzyme (14), another that the large amount of Triton, relative to the small amount of enzyme protein, could interfere with the activation process.
Because the original ghosts were capable of being stimulated by calmodulin, and because the activity of the solubilized ghosts decreased with time, it was difficult to choose a proper reference point for calculation of the degree of purification. A minimal degree of purification of 147-fold was calculated by comparing the specific activity of the purified protein with the calmodulin-stimulated activity of the ghosts. A degree of purification of 570-fold was calculated by direct comparison of the specific activity eluted from the column with that of the original ghosts without calmodulin.
The range of 150-to 570-fold purification probably underestimates the purity of the enzyme, since the presence of inactive forms in the purified enzyme is probable. The solubilized enzyme is known to lose activity when kept on ice, even in the presence of phosphatidylserine.
The percentage of total ghost protein recovered in the fraction eluted by EDTA indicates a value for the total ATPase content between the 0.02% obtained by [32P]ATP labeling of ghosts (15) and the value of 0.21% obtained by kinetic titration (16) or of 0.35% by calmodulin binding (10). The value obtained from [32P]ATP labeling may be too low, due to lability of the phosphoprotein.
The higher values may be more nearly correct, with the recovery of ATPase protein being lowered by losses during purification. Fig. 2 shows an SDS-polyacrylamide gel with 10 pg of the isolated enzyme, stained with Coomassie blue. One major band with an apparent mass of 125,000 f 3,500 daltons (determined on two different enzyme preparations isolated from two different ghost preparations, by comparison with protein standards) appears. This mass corresponds to the value of 125,000 daltons for the [v-32P]ATP-labeled phosphoprotein in ghosts determined by Ronner (3), and is not far from the values of 145,000 given by Knauf et al. (15) and by Wolf et al. (4).
In addition, a small protein peak of 205,000 (The same enzyme preparation was used as in Fig. 2. appears and, in the region of 80,000 daltons, a faint band can be detected. This peak probably represents some residual Band 3; in the experiment shown, it represented -6% of the total protein (according to the intensity of the Coomassie stain). The 205,000-dalton peak (-11% of the total protein) could be detected in all preparations isolated so far, in varying amounts (10 to 50% of the total protein). This band could represent an impurity (e.g. spectrin) not removed during isolation, or some aggregated Band 3, but it could also represent a dimer of the 125,000-dalton protein not dissociated completely in SDS.
The purified enzyme preparation was labeled with [y-32P]ATP in the presence of (Ca*+ and Me) f EGTA, and electrophoresed on 5% polyacrylamide gels, in a phosphate buffer-SDS system. A control without radioactive ATP was stained with Coomassie blue. The results are shown in Fig. 3.
The major radioactivity peak had an RF value of 0.28, which corresponded well with the RF value of 0.25 determined for the 125,000-dalton protein in the Coomassie blue-stained control gel. A small amount of radioactivity (RF = 0.13) was found in the region of the 205,000-dalton protein (RF = 0.11 in the Coomassie blue gel). The amount of radioactivity found in the small peak, 18% of the radioactivity in the large peak, corresponded well with the amount of protein found in the 205,000dalton peak, -15% of the 125,000-dalton peak. In the presence of EGTA, negligible 32P activity was found on the gel, as shown in Fig. 3B. These data strongly suggest that the 125,000-dalton protein is identical with the ghost protein Purification of (Ca'+-Mg')-ATPase from Erythrocyte Membranes labeled by [y-"'P]ATP in the presence of Ca"' and Mg'+, and that the 205,000-dalton protein may indeed represent a dimer of this protein not dissociated in SDS. This experiment also shows that the 205,000-dalton protein is not spectrin, because spectrin would not be labeled by [y-""P]ATP under these conditions (e.g. only 2 pM ATP, and in the presence of Ca"') (17). DISCUSSION Unlike the Ca"' transporting ATPase of sarcoplasmic reticulum, the (Ca"-Mg")-ATPase of erythrocytes has some unfavorable properties which have so far limited the success of the studies aimed at resolving it molecularly in a functionally active state. It is contained in the membrane in extremely low amounts, it is co-purified with Band 3, a protein which is many times more abundant than the (Ca'+-Mg"+)-ATPase, and it is rather unstable in the solubilized state. The procedure described in the present paper eliminates these drawbacks, and permits the extraction and purification from the membrane of an active enzyme. The specific, and Ca"'-dependent, association of the detergent-solubilized ATPase with the column-bound calmodulin eliminates in a single step the contaminating proteins, while the presence of acidic phospholipids throughout the entire solubilization and purification limits the inactivation of the enzyme. The purified enzyme contains a single major phosphorylatable protein which gave an apparent molecular weight of 125,000. A minor phosphorylatable protein was also present at a position corresponding to a molecular weight of 205,000. The fist band corresponds to the size already indicated by others for the (Ca"'-Mg'+)-ATPase.
The identity of the slower band is not certain. However, since aggregation of proteins in SDS-polyacrylamide gels has been reliably reported (18), it may well represent an aggregated form of the monomeric enzyme. Although its apparent molecular weight is slightly low for a dimer, its Cast-dependent phosphorylation suggests the possibility that it is indeed an aggregate of the main band. Further investigation will be necessary to determine definitely the identity of the larger protein.
The purest previously reported (Ca'+-Mg"+)-ATPase contained three polypeptide chains in approximately equal amounts; only one of these chains was phosphorylatable (4). That studv renorted that attemnts to senarate the ohosnho-rylatable protein from the others resulted in loss of activity. The present study demonstrates that affinity chromatography can separate an active phosphorylatable protein from other membrane components.
A recent report of higher specific activity of a (Cal+-Mg"+)-ATPase (19) from porcine erythrocytes presented no evidence of purity. Since porcine erythrocyte membranes have a much higher specific activity of (Ca'+-Mg")-ATPase than does human, it is not possible to compare the purity of this preparation with others, based on specific activity alone.