Purification and Molecular Properties of the (Sodium Potassium)-Adenosinetriphosphatase and Reconstitution of Coupled Sodium and Potassium Transport in Phospholipid Vesicles Containing Purified Enzyme
+
LOWELL E. HOKIN Department of P h a r m a c o l o w , UniveTsitV of W i s c o n s i n Medical School, Madison, W i s c o n s i n 53706 - -
ABSTRACT Recent work in our laboratory on the purification and characterization of the (sodium potassium)-activated adenosinetriphosphatase (NaK ATPase) has been reviewed. Two enzymes have been purified, that from the rectal salt gland of the spiny dogfish, Squalus acanthias and that from the electric organ of the electric eel, Electrophorus electricus. The enzyme appears to consist of two catalytic subunits of molecular weight of about 95,000 and one glycoprotein with a molecular weight of about 50,000. The amino acid composition, N-terminal amino acids, and the carbohydrate composition of these subunits have been determined. The phospholipid composition of the holoenzyme has also been determined. The protein component shows very little variation with evolution, but the carbohydrate and phospholipid components show considerable variation. It has been possible to form vesicles from the purified enzyme from Squalus acanthias and to demonstrate the ATP-dependent, ouabain inhibitable, coupled uphill transports of Na+ and K + . The properties of these transports are very similar to those observed previously in intact erythrocytes or resealed erythrocyte ghosts with respect to asymmetries of binding sites, stoichiometries of Na+ and K + transported, Na+-Na+ exchange, and K+-K+ exchange. It is concluded that the NaK ATPase is the molecular machine for effecting Na+ and K + transport in the intact cell membrane.
+
With the exception of a few specialized tissues (see B. Schmidt-Nielsen this Symposium) all animal tissues maintain low internal Na+ concentrations and high internal K + concentrations in the face of the inverse situation i n the extracellular fluid. This distribution of N a + and K + across cell membranes fulfills many important physiological functions. Probably the most important and most primitive function is to maintain osmotic equilibrium. Sodium and C1- are constantly leaking into the cell, accompanied by water. The animal cell, with its fragile cell membrane, would burst if the NaCl were not extruded. This problem is dealt with by the sodium pump which actively extrudes the N a + . The uptake of K + accompanies the extrusion of Na'. Part of the K + uptake is tightly coupled to the extrusion by a mechanism J. EXP. ZOOL., 194: 197-206.
which is called the Na+-K' pump and part of the K + uptake and probably most of the C1- extrusion are due to passive distribution of these ions according to the Gibbs-Donnan equilibrium, which can be quantitatively related by the Nernst equation to the potential difference across the membrane. Evolution h a s utilized the uneven distribution of Na+ a n d K'. to carry out several specialized functions. The nerve impulse is a result of a wave of increased Na' conductance down the nerve membrane allowing Na' to move down its electrochemical gradient into the nerve, and causing depolarization. This is immediately followed by a wave of increased K+ conductance which allows an 1 This work was aided by grants from the National Heart and Lung Institute (HLNS 16318) and the National Science Foundation (BMS73-01506).
197
198
LOWELL E. HOKIN
equivalent amount of K' to move out of the nerve, repolarizing the nerve. A similar situation occurs in muscle membranes during excitation. Glands and other epithelial structures have utilized Na+-K+ pump to effect the net transport or secretion of Na+ and sometimes K + across the epithelium. Because of its involvement in so many basic aspects of cellular physiology, it has been estimated that up to 30% of the energy from resting respiration is devoted to the Na+-K+pump. Another interesting aspect of the Na+-K+ pump is that it is inhibited by very low concentrations of cardiac glycosides or their aglycones. The physiological aspects of the Na+-K+pump were reviewed by Baker two years ago ('72). For many years the Na+-K+ pump was viewed as a black box, i.e., nothing was known about the molecular machine within the membrane which effected the pump. In 1957, J. Skou made a n important breakthrough. He found that in a microsome fraction (consisting mainly of small fragments of axolemma) of the leg nerve of the shore crab there was a n adenosinetriphosphatase which required Mg", N a + , and K + for optimal activity (Skou, '57). He suggested that this adenosinetriphosphatase, which henceforth shall be referred to as the NaK ATPase, might be involved in the coupled transports of N a + and K + . Over the intervening years this suggestion has been supported by a vast amount of circumstantial evidence which has established "beyond a reasonable doubt" that the NaK ATPase is intimately concerned with the coupled transports of N a + and K + and may in fact be the molecular machine which actually effects these transports. Most of the studies, in which correlations between the Na+-K+ pump and the NaK ATPase were made, were carried out in the red cell, red cell ghosts, and reconstituted red cell ghosts. The latter are essentially resealed red cell ghosts in which internal and external concentrations of ions, nucleotides, etc., can be manipulated by the investigator, These studies with red cell preparations have been carried out primarily in the laboratories of Robert Post and Ian Glynn. Valuable studies of a more biochemical nature have also been carried out with a macro-
some fraction of the electric organ of the electric eel, Electrophorus electricus in the laboratory of Wayne. Albers. Time does not permit a review of the vast amount of experimental work which has gone into this problem. Progress in this field was reviewed in the 1974 issue of Annual Reviews of Biochemistry (Dahl and Hokin, '74). An international symposium on the NaK ATPase was held under the auspices of the New York Academy of Sciences in November, 1973, and the proceedings have been published (Askari, '74). There have been many models proposed as to how the NaK ATPase may effect Na' and K+ transport (Albers et al., '68; Charnock and Opit, '68; Hokin, '69; Jardetsky, '66; Post et al., '73b; Sen et al., '69; Skou, '71; Stein et al., '73; Yang, '70). These models involve conformational changes in selected regions of the polypeptide chains of the enzyme, changes in spatial relationships between the subunits, and specific channels for Na' and K + . We do not know a great deal about how the transports of N a + and K' are brought about. Our laboratory has followed the philosophy that the only sure road to learning about this difficult problem is to purify the enzyme, study its chemical properties, attempt to reconstitute transport by incorporating the enzyme in a phospholipid bilayer, and to study conformational changes with physical techniques. This is indeed a formidable and ambitious program and after ten years of effort we are now at the threshold of the most difficult aspect of the whole problem, namely, to study the conformation of the enzyme itself and the conformational changes which presumably take place during transport. It should be pointed out that in contrast to the transport mechanism, a great deal has been learned about the reaction mechanism, particularly in the laboratories of Post and Albers [see recent review (Dahl and Hokin, '74)l. I could mention many others who have contributed to this problem. The techniques used here have been primarily kinetic studies of individual steps of the overall reaction. For the remainder of this presentation, I shall devote my attention primarily to efforts in our laboratory to purify the enzyme, to study its chemical properties, and
NA+ AND K TRANSPORT IN NAK ATPASE VESICLES
to reconstitute N a + and K + transport in phospholipid ve sicles. Purification of the NaK ATPase f r o m the rectal gland of Squalus acanthias and from the electric organ of Electrophorus electricus. The purification of enzymes which are deeply embedded in the membrane is a very difficult problem, and classical methods of protein purification which have been so successful with soluble proteins have been of little value. Two approaches to the purification of the NaK ATPase have been used. One approach is to use a detergent, usually deoxycholate, which extracts all of the extraneous proteins from the membrane, leaving the NaK ATPase membrane bound (Jorgensen et al. ’71; Kyte, ’71b). The other approach, and the one which we have followed, is to solubilize the enzyme with a non-denaturing nonionic detergent, such as Lubrol WX, and to purify the solubilized enzyme by subsequent steps. We devoted a great deal of time to purification of the enzyme from beef brain, which permits purification on a very large scale, but the best we could do was to purify the enzyme to about 50% homogeneity (Uesugi et al., ’71). These studies, which spanned about 5 years, did, however, lay out for us a purification procedure which we have found to be generally applicable to other tissues. It seemed to us that application of the “August Krogh Principle,” which Professor Krebs (’75) has discussed in this Symposium, might permit that last “leap forw a r d from 50% homogeneity to essentially a pure enzyme. I n earlier studies in the 6 0 s , we had a great deal of experience with salt glands of birds, which, as you know, are specialized to secrete NaCl which is ingested from sea water. The limitations in quantity of this material and difficulties with extensive connective tissue in this organ and some preliminary studies on preparation of Lubrol extracts of the enzyme from cell fractions of the salt gland suggested that it would not prove satisfactory. In the spring of 1971, Doctor Franklin Epstein visited our University and suggested that I try the rectal salt gland of the spiny dogfish shark Squalus acanthias. His preliminary studies had already indicated that this gland was rich in NaK ATPase. The gland in a n average
199
dogfish weighed about 2 gms and except for a capsule the tissue had about the same consistency as liver. Doctor Epstein spends his summers at the Mount Desert Island Biological Laboratory in Maine and at that time of year these migratory fish are present in virtually unlimited numbers. I went to Maine that summer and found that i t was fairly easy to isolate membrane fractions from this gland which were ten times richer in NaK ATPase than microsomes from beef brain. With our studies with beef brain the following steps had evolved: isolation of microsomes by differential centrifugation, extraction of extraneous proteins with 2 M NaI by the method of Nakao et al. (‘65), solubilization of the NaK ATPase with Lubrol WX, zonal centrifugation of the Lubrol extract, and a novel ammonium sulfate fractionation which consists of incubation of the concentrated zonal fraction at 30” with 1 M ammonium sulfate, which precipitates all of the protein, followed by dilution to 0.4 M ammonium sulfate, which solubilizes impurities leaving the NaK ATPase in a membranous form (Uesugi, ’71). The ammonium sulfate fractionation only works if free Lubrol is removed by zonal centrifugation. Since there was no zonal centrifuge at the Mount Desert Island Biological Laboratory, I was unable to carry out the entire procedure there with rectal gland materials. However, I found that the enzyme was readily solubilized with Lubrol (omitting the NaI extraction step). Furthermore, it was subsequently found that the enzyme in the gland was stable for at least a year if the gland was rapidly frozen over dry ice and stored at - 70”. I further found that membrane fractions from frozen glands could be isolated; these membrane fractions were not as pure as those isolated from fresh glands but Lubrol extracts of frozen glands had the same specific activity as those from fresh glands, suggesting that Lubrol selectively solubilized the NaK ATPase leaving behind membranous impurities. After returning to the Laboratory in the autumn, we were able to further purify membrane fractions from frozen glands and to apply the basic steps worked out for the brain (omitting the NaI extraction step) and obtain enzyme which appeared to be 90-95% pure (Hokin et al.,
200
LOWELL E. HOKIN
'73). We then turned our attention to the commercially available electric eel, EEectrophorus electricus, whose electric organ is also rich in NaK ATPase. Again applying the same basic procedure which had been worked out for beef brain (omitting the NaI extraction step) we were able to obtain enzyme which was as pure as that obtained from Squalus acanthias (Dixon and Hokin, '74) and with a somewhat higher specific activity due presumably to somewhat less inactivation by Lubrol at the initial stage of extraction of the microsoma1 fraction with this detergent (Perrone et al., '75). The yields of purified enzyme from either Squalus acanthias or Electrophorus electricus are now of the order of 50 mg per preparation and two to three preparations can be done per week. This affords ample enzyme for most studies. Chemical properties of t h e purified NaK ATPase. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the purified enzymes shows that 90-95% of the protein runs as two bafids, one with a n apparent molecular weight of 95,000 and other with a n apparent molecular weight of 55,000 in the shark and 47,000 in the electric eel. About 5% of the protein runs near the tracking dye. It had been known that in the presence of N a + , Mg2+ and [ Y - ~ ~ATP P ] the enzyme became phosphorylated, and if K + was added, the enzyme rapidly became dephosphorylated (Albers et al., '63; Blostein, '66; Charnock and Post, '63a; Charnock et al., '63b; Gibbs et al., '65; Hokin et al., '65; Kanazawa et al., '67; Nagano et al., '65; Post et al., '65; Rose, '63). The phosphorylation is on a p-carboxyl of an aspartyl residue i n the enzyme (Degani et al., '74; Hokin, '74; Nishigaki et al., '74; Post and Orcutt, '73a). Since this is a n acyl phosphate bond, it is highly unstable, particularly at higher pHs, but it was. stable enough to identify which subunit carried the aspartyl-p-phosphate residue, and this turned out to be the 95,000 molecular weight subunit, which we shall refer to a s the catalytic subunit (Alexander and Rodnight, '70; Avruch and Fairbanks, '72; Collins and Albers, '72; Hart and Titus, '73; Kyte, '71a; Uesugi et al., '71). Periodic acid Schiff staining of the gels or carbohydrate analy-
sis of the isolated subunits revealed that the lower molecular weight "subunit" was a glycoprotein (Dixon and Hokin, '74; Hokin et al., '73; Kyte, '72; Lane et al., '73; Perrone et al., '75). Kyte ('72) showed that this protein from the enzyme from the outer medulla of the dog kidney contained sialic acid, glucosamine and neutral sugars, We identified the neutral sugars by gas liquid chromatography and they turned out to be fucose, galactose, mannose and possibly glucose (Perrone et al., '75) (the presence of trace amounts of sucrose from the preparation which would give glucose but not fructose on gas liquid chromatography cannot be completely ruled out). There is considerable variation in carbohydrate composition of the glycoprotein with evolution (Hokin, '74; Perrone et al., '75). Studies on the essentiality of the carbohydrates for catalytic activity are only beginning, but with neuraminidase, we have already been able to show that 68% and 100% of the sialic acid can be removed from the enzymes from Squalus acanthias and Electrophorus electricus, respectively, without any loss in enzyme activity (Perrone et al., '75). In contrast to the carbohydrate composition, the amino acid compositions of the catalytic subunit and the glycoprotein have remained remarkably constant throughout evolution (Hokin, '74). The amino acid composition of the catalytic subunit is definitely different from that of the glycoprotein, although not drastically different. However, some amino acids, such as tyrosine, differ by a factor of two-fold. The N-terminal amino acids of the catalytic subunit also differ with different species, being glycine, alanine and serine for the dog, Squalus acanthias and Electrophorus electricus. On the other hand, the N-terminal amino acid for the glycoprotein is alanine for all three species (Hokin, '74). These data, taken together suggest that there is no drastic change in the protein component of the enzyme throughout vertebrate evolution, but that there must be small changes in amino acid sequences as indicated by different N-terminal amino acids for the catalytic subunit. With respect to the important question as to whether the glycoprotein is a "subunit" of the enzyme or is a membrane
N A + AND K TRANSPORT IN NAK ATPASE VESICLES
constituent in close proximity which copurifies with the NaK ATPase the evideilce until recently had been of a n indirect nature. In addition to the two indirect lines of evidence mentioned above, namely, the constancy of the amino acid composition and the N-terminal amino acid composition throughout evolution, it can be mentioned that cross linking with dimethylsuberimidate forms a dimer between the catalytic subunit and the glycoprotein (Kyte, '72) and that polyacrylamide gels of the enzyme at various stages of purification show a parallel enrichment of the catalytic subunit and the glycoprotein (Dixon and Hokin, '74; Hokin et al., '73). Recently, evidence of a more direct nature has been obtained (Rhee and Hokin, '75). Antibodies to the purified glycoprotein from Squalus acanthias have been prepared and have been shown to inhibit the NaK ATPase activity (Rhee and Hokin, '75). There has been some controversy about the ratios of catalytic subunit to glycoprotein in different purified preparations (Hokin and Hexum, '72; Hokin, '74; Jorgensen, '74: Kyte, '72; Lane et al., '73). However, with the purified enzymes from Squalus acanthias and Electrophorus elecIricus, we obtained ratios of 2.16 and 1.95, respectively. Two catalytic subunits and one glycoprotein give a minimum molecular weight very close to 250,000. which is that determined by the X-ray inactivation studies of Kepner and Macey ('68). Assuming "half of sites reactivity," for which there is some evidence (Repke et al., '74; Stein et al., '73) this molecular weight agrees with maximal ouabain binding and phosphorylation, which is very close to 4,000 picamoles/mg of purified enzyme from either Squalus acanthias or Electrop h w u s electricus (Perrone et al., '75). Thus, if one compares the purified enzymes from Squalus acanthias and Electrophorus electricus, the stoichiometry of the catalytic subunits and the values for oubain binding and phosphorylation are all internally consistent. Reconstitution of Nu+ and K' transport in NaK ATPase-lecithin vesicles. The kinetic parameters of the enzyme in its membranous environment before detergent treatment and the kinetic parameters
20 1
of the purified enzyme are the same [for both Squalus acanthias (Ratanabanangkoon et al., '73) and Electrophorus electricus (Dixon and Hokin, '74)]. This indicates that the purified enzyme should be functionally capable of transporting Na' and K+ if it were properly oriented i n a bilayer. In this connection, it should be pointed out that electron micrographs and X-ray diffraction studies (Simpkins and Hokin, '73) indicate that the purified enzyme is embedded in a phospholipid bilayer. The phospholipid content of the purified enzyme from Squalus acanthias and Electrophorus electricus have been reported (Perrone e t al., '75). Although showing some quantitative differences the main phospholipids are phosphatidylcholine, phosphatidylethanolamine and phosphatidylserine. We spent a great deal of time attempting to incorporate this membranous enzyme into black lipid membranes without success. We finally turned to the technique of Racker and his associates ('72), which consists of solubilization of the enzyme in a lecithin-cholate micellar solution followed by slow dialysis to remove the cholate. During dialysis vesicles are formed from phospholipid and enzyme. The intactness of the NaK ATPase vesicles (or at least a certain percentage of them) was shown by adding I4C inulin or I4C glucose before dialysis (Hilden et al., '74). After dialysis the vesicles were passed over Sephadex G-50 or Sephadex G-150. As to be expected all of the protein eluted in the void volume and associated with it were peaks of either ['"CI inulin or ["C] glucose. The free inulin or glucose eluted later. If the vesicles were incubated at 25" for 30 to 60 min in 20 mM NaC1, 100 mM KCl, 5 mM MgClz, and 40 mM imidazole HC1 (pH 7.4) on both sides and "Na+ and ATP on the outside, there was a net influx of N a + into the vesicles in the presence of ATP but not in its absence (Hilden et al., '74). There was much less influx if ATP was omitted. CTP, which is a fairly good substitute for ATP as a substrate for the NaK ATPase was about half as effective as ATP in supporting *'Na+ entry into the vesicles. UTP, which is virtually inactive as a substitute for ATP, was inactive in supporting "Na+ entry. I n the squid axon
202
LOWELL E. HOKIN
(Baker and Manil, '68) and i n a reconstituted erythrocyte ghosts (Gardner and Conlon, '72) ouabain inhibits the NaK ATPase only on the side opposite the ATP site, which would be outside in the squid axon and erythrocyte ghosts but inside in the NaK ATPase vesicles. If vesicles were prepared in the presence of ouabain so that ouabain would be trapped inside them, ATP-dependent zzNa+ entry was blocked. However, if no ouabain was included during formation of the vesicles and ouabain was added to the external surface of the vesicles, no inhibition of ATP-dependent "Na+ entry was observed. Thus, the same asymmetry for ouabain inhibition was observed in the vesicles a s is seen in the squid axon and i n resealed erythrocyte ghosts although i n the case of the NaK ATPase vesicles, transport is in the opposite direction. If Mg'+ was omitted from the system ATP-dependent N a + influx was markedly reduced. It could be argued that the ATP-dependent uptake of "Na+ was due to exchange diffusion and not net Na' transport, since exchange diffusion is also dependent on ATP. However, it can be demonstrated only in the absence of K'. To establish unequivocally that the ATP-dependent uptake of Z'Na+ was not due to exchange diffusion, vesicles were prepared by dialyzing solution containing "Na+ so that the same concentration and specific activity of "Na+ would be present o n both sides of the vesicle membrane. Under these conditions there was still a net ATP-dependent uptake of 22Na+,which could be due only to net transport of Na' into the vesicles. Since the specific activity and concentration of Na+ was known, the concentration of Na' after incubation with ATP could be determined. With initial concentration of 20 mM Na on both sides of the vesicle membrane, the final concentration inside ranged from 38 to 128 mM, depending on the vesicle preparation. If sodium was substituted for potassium during preparation of vesicles and no potassium was added during incubation, Na+-Na' exchange could be demonstrated (Hilden et al, '74). As pointed out i n the Introduction, the transports of Na+ and K + in nerve, the erythrocyte, and presumably many other
places are coupled. To study K+ transport, we prepared vesicles and allowed 42K or "Rb to leak in at 4" for 1-3 days in order to trap these cations inside the vesicles (Hilden and Hokin, '75). Since Rb+ behaves quite similarly to K T in activating the NaK ATPase and since "Rb' is a much longer lived isotope than 42K,most of the experiments were carried out with '"Rb'. When MgATP was added to the outside of the vesicles, there was a marked extrusion of either 42K+ or "Rb' so that the remaining content of isotope was only about one-sixth of that in vesicles incubated without MgATP. The extrusion of 42Kand 8fiRbwere identical. Figure 1 shows the simultaneous uptake of "Na and the extrusion of 4EKin the same batch of vesicles. I n experiments of this type, stoichiometries could be calculated. The following relationships were found for Na' transport K + transport and ATP hydrolysis: Na/ATP = 1.42, K/ATP = 1.04 and Na/K = 1.43. The ratio of 2.8 Na' transported into 2 K + transported out is very close to the value reported for the red cell membrane. Ouabain inhibited the efflux of K + but only if it was present on the inside of the vesicles, as was the case for inhibition of Na' transport. In other words, the site for inhibition by ouabain was on the same side of the membrane as the site for K + binding, a s has been shown to be the case in the squid axon and the red cell. K' efflux required the presence of external Mg2+, a s was the case for Na+ influx. CTP was about 90% as effective as ATP in supporting K + efflux, but UTP was only 20% as effective. If vesicles were prepared in the absence of NaC1, there was no ATPdependent K + efflux if Na+ was not added during incubation. However, if NaCl was added to the incubation medium, the usual K + efflux was observed. This is fairly direct evidence for the coupling of K + efflux to N a + influx. Li' was completely ineffective in substituting for Na'. K + was used in the incubation medium. The average internal concentration of K' after incubation was 8 mM, based on 42K measurements, and 6.8 mM, based on 86Rb+ measurements. Thus, a concentration gradient of about seven-fold was established between the inside and outside of the vesi-
NA+ AND K TRANSPORT IN NAK ATPASE VESICLES I
I
1
22N,+
203
0
3s
600
g 2.c
400
h
L \g 0 I0
2 u
a
I
I
1 3
0
I
1.c
P
I
42K+
-
-ATP
a
v)
9
.-AT
I
IS
100
0.5 '00
20
40
60
INCUBATION TIME (MINI Fig. 1 Time course for Z2Na+ entry and 42K+exit. Vesicles were preloaded with isotope for 2 days. During incubation at 25", 0.5 ml of vesicles were removed at the indicated times ,and the vesicular isotope contents were measured. Reproduced by permission of the Journal of Bioliogical Chemistry.
cles, directly demonstrating active transport. A K + - K + exchange can be demonstrated if ghosts are prepared in media without Na+ [see review by Glynn ('68)l. This K+-K+ exchange is dependent on the presence of ATP and inorganic phosphate. If vesicles were prepared in the absence of Na+, a K+-K+ exchange could be demonstrated if ATP and inorganic phosphate were added to the medium in which the vesicles were suspended. This K+-K+ exchange was also ouabain inhibitable, as has been observed in the red cell.
LITERATURE CITED Albers, R. W., S. Fahn and G. J. Koval 1963 The role of sodium ions i n the activation of Electrophorus electricus organ adenosine triphosphatase. Proc. Nat. Acad. Sci. U.S.A., 50: 474-481. Albers, R. W., G. J. Koval and G. J. Siege1 1968 Studies on the interaction of ouabain and other cardioactive steroids with sodium-potassiumactivated adenosine triphosphatase. Mol. Pharmacol., 4: 324-336. Alexander, D. R., and R. Rodnight 1970 Separation of neutral pH of a 32P-labelled membrane protein associated with the sodium-plusmagnesium ion-activated adenosine triphosphatase from ox brain. Biochem. J., 119: 44P-45P.
204
LOWELL E. HOKIN
Askari, A. 1974 Properties and functions of Na+ K + )-activated adenosinetriphosphatase. Ann. N. Y. Acad. Sci., 242. Avruch, J., and G. Fairbanks 1972 Demonstration of a phosphopeptide intermediate in the Mg+ +-dependent Na+-and K+-stimulated adenosine triphosphatase reaction of the ervthro. . cyte membrane. Proc. Nat. Acad. Sci. U.S.A., 69: 1216-1220. Baker, P. F. 1972 The sodium pump in animal tissues and its role in the control of cellular metabolism and function. Metab. Pathways, VI: 243-268. Baker, P. F., and J. Manil 1968 The rates o€ action of K+ and oubain o n the sodium pump in squid axons. Biochim. Biophys. Acta, 150: 328-330. Blostein, R. 1966 Evidence for a phosphorylated intermediate of red-cell membrane adenosine triphosphatase. Biochem. Biophys. Res. Commun., 24: 598-602. Charnock, J. S., and L. J. Opit 1968 Membrane metabolism and ion transport. In: The biological basis of medicine. E. E. Bittar and N. Bitter, eds. Academic Press, New York. Volume I: 69-103. Charnock, J. S., and R. L. Post 1963a Evidence of the mechanism of ouabain inhibition of cation activated adenosine triphosphatase. Nature, 199: 910-911. Charnock, J. S., A. S. Rosenthal and R. L. Post 1963b Studies of the mechanism of cation transport. 11. A phosphorylated intermediate in the cation stimulated enzyme hydrolysis of adenosine triphosphate. Aust. J. Exp. Biol. Med. Sci., 41: 675-686. Collins, R. C., and R. W. Albers 1972 The phosphoryl acceptor protein of Na-K-ATPase from various tissues. J. Neurochem., 19: 12091214. Dahl, J. L., and L. E. Hokin 1974 The sodiumpotassium adenosinetriphosphatase. Ann. Rev. Biochem., 43: 327-356. Degani, C., A. S. Dahms and P. D. Boyer 1974 Characterization of acyl phosphate in transport ATPases by a borohydride reduction method. Ann. N. Y. Acad. Sci., 242: 77-79. Dixon, J. F., and L. E. Hokin 1974 Studies on the characterization of the sodium-potassium transport adenosine triphosphatase. Purification and properties of the enzyme from the electric organ of Electrophorus electricus. Arch. Biochem. Biophys., 163: 749-758. Gardner, J. D., and T. P. Conlon 1972 The effects of sodium and potassium on ouaban binding by human erythrocytes. J. Gen Physiol., 60: 609-629. Gibbs, R., P. M. Roddy and E. Titus 1965 Preparation, assay, and properties of an Na+- and K+-requiring adenosine triphosphatase from beef brain. J. Biol. Chem., 240: 2181-2187. Glynn, I. M. 1968 Membrane adenosine triphosphatase and cation transport. Brit. Med. Bull., 24: 165-169. Hart, W. H., Jr., and E. 0. Titus 1973 Isolation of a protein component of sodium-potassium transport adenosine triphosphatase con-
+
taining ligand-protected sulfhydryl groups. J. Biol. Chem., 248: 1365-1371. Hilden, S., and L. E. Hokin 1975 Active potassium transport coupled to active sodium transport in vesicles reconstituted from purified sodium and potassium adenosine triphosphatase from the rectal gland of Squalus acanthias. J. Biol. Chem., 250: 6296-6303. Hilden, S . , H. M. Rhee and L. E. Hokin 1974 Sodium transport by phospholipid vesicles containing purified sodium and potassium ionactivated adenosine triphosphatase. J. Biol. Chem., 249: 7432-7440. Hokin, L. E. 1969 O n the molecular characterization of the sodium-potassium transport adenosine triphosphatase. J. Gen. Physiol., 54: 327s-342s. 1974 Purification and properties of the (sodium + potassium)-activated adenosinetriphosphatase and reconstitution of sodium transport. Ann. N. Y . Acad. Sci., 242: 12-23. Hokin, L. E., and T. D. Hexum 1972 Studies on the characterization of the sodium-potassium transport adenosine triphcsphatase. IX. On the role of phospholipids in the enzyme. Arch. Biochem. Biophys., 151: 453463. Hokin, L. E., J. L. Dahl, J. D. Deupree, J. F. Dixon, J. F. Hackney and J. F. Perdue 1973 Studies on the characterization of the sodiumpotassium transport adenosine triphosphatase. X. Purification of the enzyme from the rectal gland of Squalus acanthias. J. Biol. Chem., 248: 2593-2605. Hokin, L. E., P. S . Sastry, P. R. Galsworthy and A. Yoda 1965 Evidence that a phosphorylated intermediate in a brain transport adenosine triphosphatase is an acyl phosphate. Proc. Nat. Acad. Sci. U.S.A., 54: 177-184. Jardetsky, 0. 1966 Simple allosteric model for membrane pumps. Nature, 211: 969-970. Jorgensen, P. L. 1974 Purification and characterization of ( N a + + K+)-ATPase. 111. Purifiication from the outer medulla of a mammalian kidney after selective removal of membrane components by sodium dodecylsulphate. Biochim. Biophys. Acta, 356: 36-52. Jorgensen, P. L., J. C. Skou, and L. P. Solomonson 1971 Purification and characterization of (Na+ K+)-ATPase. 11. Preparation by zonal centrifugation of highly active (Na+ + K + ) ATPase from the outer medulla of rabbit kidneys. Biochim. Biophys. Acta, 233: 381-394. Kanazawa, T., M. Saito and Y. Tonomura 1967 Properties of a phosphorylated protein as a reaction intermediate of Na+-K+ sensitive ATPase. J. Biochem. (Tokyo), 61: 555-566. Kepner, G. R., and R. I. Macey 1968 Membrane enzyme systems. Molecular size determinations by radiation inactivation. Biochim. Biophys. Acta, 163: 180-203. Krebs, H. 1975 The August Krogh principle (“For many problems there is an animal on which it can be most conveniently studied”). J. Exp. Zool., this Symposium. Kyte, J. 1971a Phosphorylation of a purified (Na+ K + > adenosine triphosphatase. Biochem. Biophys. Res. Commun., 43: 1259-1265.
+
+
N A + AND K TRANSPORT IN NAK ATPASE VESICLES 1971b Purification of the sodium- and potassium-depe,ident adenosine triphosphatase from canine renal medulla. J. Biol. Chem., 246: 4157-4 165. 1972 Properties of the two polypeptides of sodium- and potassium-dependent adenosine triphosphatase. J. Biol. Chem., 247: 7642-7649. Lane, L. K., J. H. Copenhaver, Jr., G . E. Lindenmayer and A. Schwartz 1973 Purification and characterization of and [3H] ouabain binding to the transport adenosine triphosphatase from outer medulla of canine kidney. J. Biol. Chem., 248: 7197-7200. Nagano, K., T. Kanazawa, N. Mizuno, Y. Tashima, T. Nakao, and M. Nakao 1965 Some acyl phosphate-like properties of P32-labelled sodium-potassium-activated adenosine triphosphatase. Biochem. Biophys. Res. Commun., 19: 759-764. Nakao, T., Y. Tashima, K. Nagano, and M. Nakao 1965 Highly specific sodium-potassium-activated adenosine triphosphatase from various tissues of rabbit. Biochem. Biophys. Hes. Commun., 19: 755-'758. Nishigaki, I., F. T. Chen and L. E. Hokin 1974 Studies on the characterization of the sodiumpotassium transport adenosine triphosphatase. XV. Direct chemical characterization of the acyl phosphate in the enzyme as an aspartyl /3-phosphate residue. J. Biol. Chem., 249: 491 14916. Perrone, J. R., J. F. Hackney, J. F. Dixon and L. E. Hokin 1975 Molecular properties of purified (sodium + potassium)-activated adenosinetriphosphatases and their subunits from the rectal gland of Squalus acanthias and the electric organ of Electrophorus electricus. J. Biol. Chem., 250: 4178-4184. Post, R. L., and B. Orcutt 1973a Active site of phosphorylation of N a + , K+-ATPase. In: Organization of energy transducing membranes. M. Nakao and L. Packer eds. Tokyo University Press, Tokyo, pp. 35-46. Post, R. L., S. Kume and F. N. Rogers 1973b Alternating paths of phosphorylation of the sodium and potassium ion pump of plasma membranes. I n : Mechanisms in bioenergetics. F. Azzone and L. Ernster, eds. Academic Press, New York, pp. 203-218. Post, R. L., A. K. Sen and A. S . Rosenthal 1965 A phosphorylated intermediate in adenosine triphosphate-dependent sodium and potassium transport across kidney membranes. J. Biol. Chem., 240: 1437-1445.
205
Racker, E. 1972 Reconstitution of a calcium pump with phospholipids and a purified Ca+ + adenosine triphosphatase from sarcoplasmic reticulum. J. Biol. Chem.. 247: 8198-8200. Ratanabanangkoon, K., J . Dixon and L. E. Hokin 1973 Studies on the characterization of the sodium-potassium transport adenosinetriphosphatase. XI. Comparison of kinetic properties of the Furified with the impure membranebound enzyme from Squalus acanthias. Arch. Biochem. Biophys., 156: 342-349. Repke, K. R. H., R. Schon, W. Henke, W. Schonfeld, B. Streckenbach and F. Dittrich 1974 Experimental and theoretical examination of the flip-flop model (Na, K)-ATPase function. Ann. N. Y. Acad. Sci., 242: 203-219. Rhee, H. M., and L. E. Hokin 1975 Inhibition of the purified sodium-potassium activated adenosinetriphosphatase from the rectal gland of Squalus acanthias by antibody against the glycoprotein subunit. Biochem. Biophys. Res. Commun., 63: 1139-1145. Rose, S. P. R. 1963 Phosphoprotein as in intermediate in cerebral microsomal adenosinetriphosphatase. Nature, 199: 375-378. Sen, A. K., T. Tobin and R. L. Post 1969 A cycle for ouabain inhibition of sodium- and potassium-dependent adenosine triphosphatase. J. Biol. Chem., 244: 659643604. Simpkins, H., and L. E. Hokin 1973 Studies on the characterization of the sodium-potassium transport adenosinetriphosphatase. XIII. On the organization and role of phospholipids in the purified enzyme. Arch. Biochem. Biophys., 159: 897-902. Skou, J. C. 1957 The influence of some cations on an adenosine triphosphatase from peripheral nerves. Biochim. Biophys. Acta, 23: 394-401. 1971 Sequence of steps in the ( N a + K f )-activated enzyme system in relation to sodium and potassium transport. Cur. Top. Bioenerg., 4: 357-397. Stein, W. D., W. R. Lieb, J. D. Karlish and Y. Eilam 1973 A model for active transport of sodium and potassium ions as mediated by a tetrameric enzyme. Proc. Nat. Acad. Sci. U.S.A., 70: 275-278. Uesugi, S., N. C. Dulak, J. F. Dixon, T. D. Hexum, J. L. Dahl, J . F. Perdue and L. E. Hokin 1971 Studies on the characterization of the sodiumpotassium transport adenosine triphosphatase. VI. Large scale partial purification and properties of a lubrol-solubilized bovine brain enzyme. J. Biol. Chem., 246: 531-543. Yang, J. H. 1970 Possible role of pyrophosphate linkage in the active transport of sodium ions. Proc. Nat. Acad. Sci. U.S.A., 67: 59-61.
+
+
LOWELL E. HOKIN Department of P h a r m a c o l o w , UniveTsitV of W i s c o n s i n Medical School, Madison, W i s c o n s i n 53706 - -
ABSTRACT Recent work in our laboratory on the purification and characterization of the (sodium potassium)-activated adenosinetriphosphatase (NaK ATPase) has been reviewed. Two enzymes have been purified, that from the rectal salt gland of the spiny dogfish, Squalus acanthias and that from the electric organ of the electric eel, Electrophorus electricus. The enzyme appears to consist of two catalytic subunits of molecular weight of about 95,000 and one glycoprotein with a molecular weight of about 50,000. The amino acid composition, N-terminal amino acids, and the carbohydrate composition of these subunits have been determined. The phospholipid composition of the holoenzyme has also been determined. The protein component shows very little variation with evolution, but the carbohydrate and phospholipid components show considerable variation. It has been possible to form vesicles from the purified enzyme from Squalus acanthias and to demonstrate the ATP-dependent, ouabain inhibitable, coupled uphill transports of Na+ and K + . The properties of these transports are very similar to those observed previously in intact erythrocytes or resealed erythrocyte ghosts with respect to asymmetries of binding sites, stoichiometries of Na+ and K + transported, Na+-Na+ exchange, and K+-K+ exchange. It is concluded that the NaK ATPase is the molecular machine for effecting Na+ and K + transport in the intact cell membrane.
+
With the exception of a few specialized tissues (see B. Schmidt-Nielsen this Symposium) all animal tissues maintain low internal Na+ concentrations and high internal K + concentrations in the face of the inverse situation i n the extracellular fluid. This distribution of N a + and K + across cell membranes fulfills many important physiological functions. Probably the most important and most primitive function is to maintain osmotic equilibrium. Sodium and C1- are constantly leaking into the cell, accompanied by water. The animal cell, with its fragile cell membrane, would burst if the NaCl were not extruded. This problem is dealt with by the sodium pump which actively extrudes the N a + . The uptake of K + accompanies the extrusion of Na'. Part of the K + uptake is tightly coupled to the extrusion by a mechanism J. EXP. ZOOL., 194: 197-206.
which is called the Na+-K' pump and part of the K + uptake and probably most of the C1- extrusion are due to passive distribution of these ions according to the Gibbs-Donnan equilibrium, which can be quantitatively related by the Nernst equation to the potential difference across the membrane. Evolution h a s utilized the uneven distribution of Na+ a n d K'. to carry out several specialized functions. The nerve impulse is a result of a wave of increased Na' conductance down the nerve membrane allowing Na' to move down its electrochemical gradient into the nerve, and causing depolarization. This is immediately followed by a wave of increased K+ conductance which allows an 1 This work was aided by grants from the National Heart and Lung Institute (HLNS 16318) and the National Science Foundation (BMS73-01506).
197
198
LOWELL E. HOKIN
equivalent amount of K' to move out of the nerve, repolarizing the nerve. A similar situation occurs in muscle membranes during excitation. Glands and other epithelial structures have utilized Na+-K+ pump to effect the net transport or secretion of Na+ and sometimes K + across the epithelium. Because of its involvement in so many basic aspects of cellular physiology, it has been estimated that up to 30% of the energy from resting respiration is devoted to the Na+-K+pump. Another interesting aspect of the Na+-K+ pump is that it is inhibited by very low concentrations of cardiac glycosides or their aglycones. The physiological aspects of the Na+-K+pump were reviewed by Baker two years ago ('72). For many years the Na+-K+ pump was viewed as a black box, i.e., nothing was known about the molecular machine within the membrane which effected the pump. In 1957, J. Skou made a n important breakthrough. He found that in a microsome fraction (consisting mainly of small fragments of axolemma) of the leg nerve of the shore crab there was a n adenosinetriphosphatase which required Mg", N a + , and K + for optimal activity (Skou, '57). He suggested that this adenosinetriphosphatase, which henceforth shall be referred to as the NaK ATPase, might be involved in the coupled transports of N a + and K + . Over the intervening years this suggestion has been supported by a vast amount of circumstantial evidence which has established "beyond a reasonable doubt" that the NaK ATPase is intimately concerned with the coupled transports of N a + and K + and may in fact be the molecular machine which actually effects these transports. Most of the studies, in which correlations between the Na+-K+ pump and the NaK ATPase were made, were carried out in the red cell, red cell ghosts, and reconstituted red cell ghosts. The latter are essentially resealed red cell ghosts in which internal and external concentrations of ions, nucleotides, etc., can be manipulated by the investigator, These studies with red cell preparations have been carried out primarily in the laboratories of Robert Post and Ian Glynn. Valuable studies of a more biochemical nature have also been carried out with a macro-
some fraction of the electric organ of the electric eel, Electrophorus electricus in the laboratory of Wayne. Albers. Time does not permit a review of the vast amount of experimental work which has gone into this problem. Progress in this field was reviewed in the 1974 issue of Annual Reviews of Biochemistry (Dahl and Hokin, '74). An international symposium on the NaK ATPase was held under the auspices of the New York Academy of Sciences in November, 1973, and the proceedings have been published (Askari, '74). There have been many models proposed as to how the NaK ATPase may effect Na' and K+ transport (Albers et al., '68; Charnock and Opit, '68; Hokin, '69; Jardetsky, '66; Post et al., '73b; Sen et al., '69; Skou, '71; Stein et al., '73; Yang, '70). These models involve conformational changes in selected regions of the polypeptide chains of the enzyme, changes in spatial relationships between the subunits, and specific channels for Na' and K + . We do not know a great deal about how the transports of N a + and K' are brought about. Our laboratory has followed the philosophy that the only sure road to learning about this difficult problem is to purify the enzyme, study its chemical properties, attempt to reconstitute transport by incorporating the enzyme in a phospholipid bilayer, and to study conformational changes with physical techniques. This is indeed a formidable and ambitious program and after ten years of effort we are now at the threshold of the most difficult aspect of the whole problem, namely, to study the conformation of the enzyme itself and the conformational changes which presumably take place during transport. It should be pointed out that in contrast to the transport mechanism, a great deal has been learned about the reaction mechanism, particularly in the laboratories of Post and Albers [see recent review (Dahl and Hokin, '74)l. I could mention many others who have contributed to this problem. The techniques used here have been primarily kinetic studies of individual steps of the overall reaction. For the remainder of this presentation, I shall devote my attention primarily to efforts in our laboratory to purify the enzyme, to study its chemical properties, and
NA+ AND K TRANSPORT IN NAK ATPASE VESICLES
to reconstitute N a + and K + transport in phospholipid ve sicles. Purification of the NaK ATPase f r o m the rectal gland of Squalus acanthias and from the electric organ of Electrophorus electricus. The purification of enzymes which are deeply embedded in the membrane is a very difficult problem, and classical methods of protein purification which have been so successful with soluble proteins have been of little value. Two approaches to the purification of the NaK ATPase have been used. One approach is to use a detergent, usually deoxycholate, which extracts all of the extraneous proteins from the membrane, leaving the NaK ATPase membrane bound (Jorgensen et al. ’71; Kyte, ’71b). The other approach, and the one which we have followed, is to solubilize the enzyme with a non-denaturing nonionic detergent, such as Lubrol WX, and to purify the solubilized enzyme by subsequent steps. We devoted a great deal of time to purification of the enzyme from beef brain, which permits purification on a very large scale, but the best we could do was to purify the enzyme to about 50% homogeneity (Uesugi et al., ’71). These studies, which spanned about 5 years, did, however, lay out for us a purification procedure which we have found to be generally applicable to other tissues. It seemed to us that application of the “August Krogh Principle,” which Professor Krebs (’75) has discussed in this Symposium, might permit that last “leap forw a r d from 50% homogeneity to essentially a pure enzyme. I n earlier studies in the 6 0 s , we had a great deal of experience with salt glands of birds, which, as you know, are specialized to secrete NaCl which is ingested from sea water. The limitations in quantity of this material and difficulties with extensive connective tissue in this organ and some preliminary studies on preparation of Lubrol extracts of the enzyme from cell fractions of the salt gland suggested that it would not prove satisfactory. In the spring of 1971, Doctor Franklin Epstein visited our University and suggested that I try the rectal salt gland of the spiny dogfish shark Squalus acanthias. His preliminary studies had already indicated that this gland was rich in NaK ATPase. The gland in a n average
199
dogfish weighed about 2 gms and except for a capsule the tissue had about the same consistency as liver. Doctor Epstein spends his summers at the Mount Desert Island Biological Laboratory in Maine and at that time of year these migratory fish are present in virtually unlimited numbers. I went to Maine that summer and found that i t was fairly easy to isolate membrane fractions from this gland which were ten times richer in NaK ATPase than microsomes from beef brain. With our studies with beef brain the following steps had evolved: isolation of microsomes by differential centrifugation, extraction of extraneous proteins with 2 M NaI by the method of Nakao et al. (‘65), solubilization of the NaK ATPase with Lubrol WX, zonal centrifugation of the Lubrol extract, and a novel ammonium sulfate fractionation which consists of incubation of the concentrated zonal fraction at 30” with 1 M ammonium sulfate, which precipitates all of the protein, followed by dilution to 0.4 M ammonium sulfate, which solubilizes impurities leaving the NaK ATPase in a membranous form (Uesugi, ’71). The ammonium sulfate fractionation only works if free Lubrol is removed by zonal centrifugation. Since there was no zonal centrifuge at the Mount Desert Island Biological Laboratory, I was unable to carry out the entire procedure there with rectal gland materials. However, I found that the enzyme was readily solubilized with Lubrol (omitting the NaI extraction step). Furthermore, it was subsequently found that the enzyme in the gland was stable for at least a year if the gland was rapidly frozen over dry ice and stored at - 70”. I further found that membrane fractions from frozen glands could be isolated; these membrane fractions were not as pure as those isolated from fresh glands but Lubrol extracts of frozen glands had the same specific activity as those from fresh glands, suggesting that Lubrol selectively solubilized the NaK ATPase leaving behind membranous impurities. After returning to the Laboratory in the autumn, we were able to further purify membrane fractions from frozen glands and to apply the basic steps worked out for the brain (omitting the NaI extraction step) and obtain enzyme which appeared to be 90-95% pure (Hokin et al.,
200
LOWELL E. HOKIN
'73). We then turned our attention to the commercially available electric eel, EEectrophorus electricus, whose electric organ is also rich in NaK ATPase. Again applying the same basic procedure which had been worked out for beef brain (omitting the NaI extraction step) we were able to obtain enzyme which was as pure as that obtained from Squalus acanthias (Dixon and Hokin, '74) and with a somewhat higher specific activity due presumably to somewhat less inactivation by Lubrol at the initial stage of extraction of the microsoma1 fraction with this detergent (Perrone et al., '75). The yields of purified enzyme from either Squalus acanthias or Electrophorus electricus are now of the order of 50 mg per preparation and two to three preparations can be done per week. This affords ample enzyme for most studies. Chemical properties of t h e purified NaK ATPase. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the purified enzymes shows that 90-95% of the protein runs as two bafids, one with a n apparent molecular weight of 95,000 and other with a n apparent molecular weight of 55,000 in the shark and 47,000 in the electric eel. About 5% of the protein runs near the tracking dye. It had been known that in the presence of N a + , Mg2+ and [ Y - ~ ~ATP P ] the enzyme became phosphorylated, and if K + was added, the enzyme rapidly became dephosphorylated (Albers et al., '63; Blostein, '66; Charnock and Post, '63a; Charnock et al., '63b; Gibbs et al., '65; Hokin et al., '65; Kanazawa et al., '67; Nagano et al., '65; Post et al., '65; Rose, '63). The phosphorylation is on a p-carboxyl of an aspartyl residue i n the enzyme (Degani et al., '74; Hokin, '74; Nishigaki et al., '74; Post and Orcutt, '73a). Since this is a n acyl phosphate bond, it is highly unstable, particularly at higher pHs, but it was. stable enough to identify which subunit carried the aspartyl-p-phosphate residue, and this turned out to be the 95,000 molecular weight subunit, which we shall refer to a s the catalytic subunit (Alexander and Rodnight, '70; Avruch and Fairbanks, '72; Collins and Albers, '72; Hart and Titus, '73; Kyte, '71a; Uesugi et al., '71). Periodic acid Schiff staining of the gels or carbohydrate analy-
sis of the isolated subunits revealed that the lower molecular weight "subunit" was a glycoprotein (Dixon and Hokin, '74; Hokin et al., '73; Kyte, '72; Lane et al., '73; Perrone et al., '75). Kyte ('72) showed that this protein from the enzyme from the outer medulla of the dog kidney contained sialic acid, glucosamine and neutral sugars, We identified the neutral sugars by gas liquid chromatography and they turned out to be fucose, galactose, mannose and possibly glucose (Perrone et al., '75) (the presence of trace amounts of sucrose from the preparation which would give glucose but not fructose on gas liquid chromatography cannot be completely ruled out). There is considerable variation in carbohydrate composition of the glycoprotein with evolution (Hokin, '74; Perrone et al., '75). Studies on the essentiality of the carbohydrates for catalytic activity are only beginning, but with neuraminidase, we have already been able to show that 68% and 100% of the sialic acid can be removed from the enzymes from Squalus acanthias and Electrophorus electricus, respectively, without any loss in enzyme activity (Perrone et al., '75). In contrast to the carbohydrate composition, the amino acid compositions of the catalytic subunit and the glycoprotein have remained remarkably constant throughout evolution (Hokin, '74). The amino acid composition of the catalytic subunit is definitely different from that of the glycoprotein, although not drastically different. However, some amino acids, such as tyrosine, differ by a factor of two-fold. The N-terminal amino acids of the catalytic subunit also differ with different species, being glycine, alanine and serine for the dog, Squalus acanthias and Electrophorus electricus. On the other hand, the N-terminal amino acid for the glycoprotein is alanine for all three species (Hokin, '74). These data, taken together suggest that there is no drastic change in the protein component of the enzyme throughout vertebrate evolution, but that there must be small changes in amino acid sequences as indicated by different N-terminal amino acids for the catalytic subunit. With respect to the important question as to whether the glycoprotein is a "subunit" of the enzyme or is a membrane
N A + AND K TRANSPORT IN NAK ATPASE VESICLES
constituent in close proximity which copurifies with the NaK ATPase the evideilce until recently had been of a n indirect nature. In addition to the two indirect lines of evidence mentioned above, namely, the constancy of the amino acid composition and the N-terminal amino acid composition throughout evolution, it can be mentioned that cross linking with dimethylsuberimidate forms a dimer between the catalytic subunit and the glycoprotein (Kyte, '72) and that polyacrylamide gels of the enzyme at various stages of purification show a parallel enrichment of the catalytic subunit and the glycoprotein (Dixon and Hokin, '74; Hokin et al., '73). Recently, evidence of a more direct nature has been obtained (Rhee and Hokin, '75). Antibodies to the purified glycoprotein from Squalus acanthias have been prepared and have been shown to inhibit the NaK ATPase activity (Rhee and Hokin, '75). There has been some controversy about the ratios of catalytic subunit to glycoprotein in different purified preparations (Hokin and Hexum, '72; Hokin, '74; Jorgensen, '74: Kyte, '72; Lane et al., '73). However, with the purified enzymes from Squalus acanthias and Electrophorus elecIricus, we obtained ratios of 2.16 and 1.95, respectively. Two catalytic subunits and one glycoprotein give a minimum molecular weight very close to 250,000. which is that determined by the X-ray inactivation studies of Kepner and Macey ('68). Assuming "half of sites reactivity," for which there is some evidence (Repke et al., '74; Stein et al., '73) this molecular weight agrees with maximal ouabain binding and phosphorylation, which is very close to 4,000 picamoles/mg of purified enzyme from either Squalus acanthias or Electrop h w u s electricus (Perrone et al., '75). Thus, if one compares the purified enzymes from Squalus acanthias and Electrophorus electricus, the stoichiometry of the catalytic subunits and the values for oubain binding and phosphorylation are all internally consistent. Reconstitution of Nu+ and K' transport in NaK ATPase-lecithin vesicles. The kinetic parameters of the enzyme in its membranous environment before detergent treatment and the kinetic parameters
20 1
of the purified enzyme are the same [for both Squalus acanthias (Ratanabanangkoon et al., '73) and Electrophorus electricus (Dixon and Hokin, '74)]. This indicates that the purified enzyme should be functionally capable of transporting Na' and K+ if it were properly oriented i n a bilayer. In this connection, it should be pointed out that electron micrographs and X-ray diffraction studies (Simpkins and Hokin, '73) indicate that the purified enzyme is embedded in a phospholipid bilayer. The phospholipid content of the purified enzyme from Squalus acanthias and Electrophorus electricus have been reported (Perrone e t al., '75). Although showing some quantitative differences the main phospholipids are phosphatidylcholine, phosphatidylethanolamine and phosphatidylserine. We spent a great deal of time attempting to incorporate this membranous enzyme into black lipid membranes without success. We finally turned to the technique of Racker and his associates ('72), which consists of solubilization of the enzyme in a lecithin-cholate micellar solution followed by slow dialysis to remove the cholate. During dialysis vesicles are formed from phospholipid and enzyme. The intactness of the NaK ATPase vesicles (or at least a certain percentage of them) was shown by adding I4C inulin or I4C glucose before dialysis (Hilden et al., '74). After dialysis the vesicles were passed over Sephadex G-50 or Sephadex G-150. As to be expected all of the protein eluted in the void volume and associated with it were peaks of either ['"CI inulin or ["C] glucose. The free inulin or glucose eluted later. If the vesicles were incubated at 25" for 30 to 60 min in 20 mM NaC1, 100 mM KCl, 5 mM MgClz, and 40 mM imidazole HC1 (pH 7.4) on both sides and "Na+ and ATP on the outside, there was a net influx of N a + into the vesicles in the presence of ATP but not in its absence (Hilden et al., '74). There was much less influx if ATP was omitted. CTP, which is a fairly good substitute for ATP as a substrate for the NaK ATPase was about half as effective as ATP in supporting *'Na+ entry into the vesicles. UTP, which is virtually inactive as a substitute for ATP, was inactive in supporting "Na+ entry. I n the squid axon
202
LOWELL E. HOKIN
(Baker and Manil, '68) and i n a reconstituted erythrocyte ghosts (Gardner and Conlon, '72) ouabain inhibits the NaK ATPase only on the side opposite the ATP site, which would be outside in the squid axon and erythrocyte ghosts but inside in the NaK ATPase vesicles. If vesicles were prepared in the presence of ouabain so that ouabain would be trapped inside them, ATP-dependent zzNa+ entry was blocked. However, if no ouabain was included during formation of the vesicles and ouabain was added to the external surface of the vesicles, no inhibition of ATP-dependent "Na+ entry was observed. Thus, the same asymmetry for ouabain inhibition was observed in the vesicles a s is seen in the squid axon and i n resealed erythrocyte ghosts although i n the case of the NaK ATPase vesicles, transport is in the opposite direction. If Mg'+ was omitted from the system ATP-dependent N a + influx was markedly reduced. It could be argued that the ATP-dependent uptake of "Na+ was due to exchange diffusion and not net Na' transport, since exchange diffusion is also dependent on ATP. However, it can be demonstrated only in the absence of K'. To establish unequivocally that the ATP-dependent uptake of Z'Na+ was not due to exchange diffusion, vesicles were prepared by dialyzing solution containing "Na+ so that the same concentration and specific activity of "Na+ would be present o n both sides of the vesicle membrane. Under these conditions there was still a net ATP-dependent uptake of 22Na+,which could be due only to net transport of Na' into the vesicles. Since the specific activity and concentration of Na+ was known, the concentration of Na' after incubation with ATP could be determined. With initial concentration of 20 mM Na on both sides of the vesicle membrane, the final concentration inside ranged from 38 to 128 mM, depending on the vesicle preparation. If sodium was substituted for potassium during preparation of vesicles and no potassium was added during incubation, Na+-Na' exchange could be demonstrated (Hilden et al, '74). As pointed out i n the Introduction, the transports of Na+ and K + in nerve, the erythrocyte, and presumably many other
places are coupled. To study K+ transport, we prepared vesicles and allowed 42K or "Rb to leak in at 4" for 1-3 days in order to trap these cations inside the vesicles (Hilden and Hokin, '75). Since Rb+ behaves quite similarly to K T in activating the NaK ATPase and since "Rb' is a much longer lived isotope than 42K,most of the experiments were carried out with '"Rb'. When MgATP was added to the outside of the vesicles, there was a marked extrusion of either 42K+ or "Rb' so that the remaining content of isotope was only about one-sixth of that in vesicles incubated without MgATP. The extrusion of 42Kand 8fiRbwere identical. Figure 1 shows the simultaneous uptake of "Na and the extrusion of 4EKin the same batch of vesicles. I n experiments of this type, stoichiometries could be calculated. The following relationships were found for Na' transport K + transport and ATP hydrolysis: Na/ATP = 1.42, K/ATP = 1.04 and Na/K = 1.43. The ratio of 2.8 Na' transported into 2 K + transported out is very close to the value reported for the red cell membrane. Ouabain inhibited the efflux of K + but only if it was present on the inside of the vesicles, as was the case for inhibition of Na' transport. In other words, the site for inhibition by ouabain was on the same side of the membrane as the site for K + binding, a s has been shown to be the case in the squid axon and the red cell. K' efflux required the presence of external Mg2+, a s was the case for Na+ influx. CTP was about 90% as effective as ATP in supporting K + efflux, but UTP was only 20% as effective. If vesicles were prepared in the absence of NaC1, there was no ATPdependent K + efflux if Na+ was not added during incubation. However, if NaCl was added to the incubation medium, the usual K + efflux was observed. This is fairly direct evidence for the coupling of K + efflux to N a + influx. Li' was completely ineffective in substituting for Na'. K + was used in the incubation medium. The average internal concentration of K' after incubation was 8 mM, based on 42K measurements, and 6.8 mM, based on 86Rb+ measurements. Thus, a concentration gradient of about seven-fold was established between the inside and outside of the vesi-
NA+ AND K TRANSPORT IN NAK ATPASE VESICLES I
I
1
22N,+
203
0
3s
600
g 2.c
400
h
L \g 0 I0
2 u
a
I
I
1 3
0
I
1.c
P
I
42K+
-
-ATP
a
v)
9
.-AT
I
IS
100
0.5 '00
20
40
60
INCUBATION TIME (MINI Fig. 1 Time course for Z2Na+ entry and 42K+exit. Vesicles were preloaded with isotope for 2 days. During incubation at 25", 0.5 ml of vesicles were removed at the indicated times ,and the vesicular isotope contents were measured. Reproduced by permission of the Journal of Bioliogical Chemistry.
cles, directly demonstrating active transport. A K + - K + exchange can be demonstrated if ghosts are prepared in media without Na+ [see review by Glynn ('68)l. This K+-K+ exchange is dependent on the presence of ATP and inorganic phosphate. If vesicles were prepared in the absence of Na+, a K+-K+ exchange could be demonstrated if ATP and inorganic phosphate were added to the medium in which the vesicles were suspended. This K+-K+ exchange was also ouabain inhibitable, as has been observed in the red cell.
LITERATURE CITED Albers, R. W., S. Fahn and G. J. Koval 1963 The role of sodium ions i n the activation of Electrophorus electricus organ adenosine triphosphatase. Proc. Nat. Acad. Sci. U.S.A., 50: 474-481. Albers, R. W., G. J. Koval and G. J. Siege1 1968 Studies on the interaction of ouabain and other cardioactive steroids with sodium-potassiumactivated adenosine triphosphatase. Mol. Pharmacol., 4: 324-336. Alexander, D. R., and R. Rodnight 1970 Separation of neutral pH of a 32P-labelled membrane protein associated with the sodium-plusmagnesium ion-activated adenosine triphosphatase from ox brain. Biochem. J., 119: 44P-45P.
204
LOWELL E. HOKIN
Askari, A. 1974 Properties and functions of Na+ K + )-activated adenosinetriphosphatase. Ann. N. Y. Acad. Sci., 242. Avruch, J., and G. Fairbanks 1972 Demonstration of a phosphopeptide intermediate in the Mg+ +-dependent Na+-and K+-stimulated adenosine triphosphatase reaction of the ervthro. . cyte membrane. Proc. Nat. Acad. Sci. U.S.A., 69: 1216-1220. Baker, P. F. 1972 The sodium pump in animal tissues and its role in the control of cellular metabolism and function. Metab. Pathways, VI: 243-268. Baker, P. F., and J. Manil 1968 The rates o€ action of K+ and oubain o n the sodium pump in squid axons. Biochim. Biophys. Acta, 150: 328-330. Blostein, R. 1966 Evidence for a phosphorylated intermediate of red-cell membrane adenosine triphosphatase. Biochem. Biophys. Res. Commun., 24: 598-602. Charnock, J. S., and L. J. Opit 1968 Membrane metabolism and ion transport. In: The biological basis of medicine. E. E. Bittar and N. Bitter, eds. Academic Press, New York. Volume I: 69-103. Charnock, J. S., and R. L. Post 1963a Evidence of the mechanism of ouabain inhibition of cation activated adenosine triphosphatase. Nature, 199: 910-911. Charnock, J. S., A. S. Rosenthal and R. L. Post 1963b Studies of the mechanism of cation transport. 11. A phosphorylated intermediate in the cation stimulated enzyme hydrolysis of adenosine triphosphate. Aust. J. Exp. Biol. Med. Sci., 41: 675-686. Collins, R. C., and R. W. Albers 1972 The phosphoryl acceptor protein of Na-K-ATPase from various tissues. J. Neurochem., 19: 12091214. Dahl, J. L., and L. E. Hokin 1974 The sodiumpotassium adenosinetriphosphatase. Ann. Rev. Biochem., 43: 327-356. Degani, C., A. S. Dahms and P. D. Boyer 1974 Characterization of acyl phosphate in transport ATPases by a borohydride reduction method. Ann. N. Y. Acad. Sci., 242: 77-79. Dixon, J. F., and L. E. Hokin 1974 Studies on the characterization of the sodium-potassium transport adenosine triphosphatase. Purification and properties of the enzyme from the electric organ of Electrophorus electricus. Arch. Biochem. Biophys., 163: 749-758. Gardner, J. D., and T. P. Conlon 1972 The effects of sodium and potassium on ouaban binding by human erythrocytes. J. Gen Physiol., 60: 609-629. Gibbs, R., P. M. Roddy and E. Titus 1965 Preparation, assay, and properties of an Na+- and K+-requiring adenosine triphosphatase from beef brain. J. Biol. Chem., 240: 2181-2187. Glynn, I. M. 1968 Membrane adenosine triphosphatase and cation transport. Brit. Med. Bull., 24: 165-169. Hart, W. H., Jr., and E. 0. Titus 1973 Isolation of a protein component of sodium-potassium transport adenosine triphosphatase con-
+
taining ligand-protected sulfhydryl groups. J. Biol. Chem., 248: 1365-1371. Hilden, S., and L. E. Hokin 1975 Active potassium transport coupled to active sodium transport in vesicles reconstituted from purified sodium and potassium adenosine triphosphatase from the rectal gland of Squalus acanthias. J. Biol. Chem., 250: 6296-6303. Hilden, S . , H. M. Rhee and L. E. Hokin 1974 Sodium transport by phospholipid vesicles containing purified sodium and potassium ionactivated adenosine triphosphatase. J. Biol. Chem., 249: 7432-7440. Hokin, L. E. 1969 O n the molecular characterization of the sodium-potassium transport adenosine triphosphatase. J. Gen. Physiol., 54: 327s-342s. 1974 Purification and properties of the (sodium + potassium)-activated adenosinetriphosphatase and reconstitution of sodium transport. Ann. N. Y . Acad. Sci., 242: 12-23. Hokin, L. E., and T. D. Hexum 1972 Studies on the characterization of the sodium-potassium transport adenosine triphcsphatase. IX. On the role of phospholipids in the enzyme. Arch. Biochem. Biophys., 151: 453463. Hokin, L. E., J. L. Dahl, J. D. Deupree, J. F. Dixon, J. F. Hackney and J. F. Perdue 1973 Studies on the characterization of the sodiumpotassium transport adenosine triphosphatase. X. Purification of the enzyme from the rectal gland of Squalus acanthias. J. Biol. Chem., 248: 2593-2605. Hokin, L. E., P. S . Sastry, P. R. Galsworthy and A. Yoda 1965 Evidence that a phosphorylated intermediate in a brain transport adenosine triphosphatase is an acyl phosphate. Proc. Nat. Acad. Sci. U.S.A., 54: 177-184. Jardetsky, 0. 1966 Simple allosteric model for membrane pumps. Nature, 211: 969-970. Jorgensen, P. L. 1974 Purification and characterization of ( N a + + K+)-ATPase. 111. Purifiication from the outer medulla of a mammalian kidney after selective removal of membrane components by sodium dodecylsulphate. Biochim. Biophys. Acta, 356: 36-52. Jorgensen, P. L., J. C. Skou, and L. P. Solomonson 1971 Purification and characterization of (Na+ K+)-ATPase. 11. Preparation by zonal centrifugation of highly active (Na+ + K + ) ATPase from the outer medulla of rabbit kidneys. Biochim. Biophys. Acta, 233: 381-394. Kanazawa, T., M. Saito and Y. Tonomura 1967 Properties of a phosphorylated protein as a reaction intermediate of Na+-K+ sensitive ATPase. J. Biochem. (Tokyo), 61: 555-566. Kepner, G. R., and R. I. Macey 1968 Membrane enzyme systems. Molecular size determinations by radiation inactivation. Biochim. Biophys. Acta, 163: 180-203. Krebs, H. 1975 The August Krogh principle (“For many problems there is an animal on which it can be most conveniently studied”). J. Exp. Zool., this Symposium. Kyte, J. 1971a Phosphorylation of a purified (Na+ K + > adenosine triphosphatase. Biochem. Biophys. Res. Commun., 43: 1259-1265.
+
+
N A + AND K TRANSPORT IN NAK ATPASE VESICLES 1971b Purification of the sodium- and potassium-depe,ident adenosine triphosphatase from canine renal medulla. J. Biol. Chem., 246: 4157-4 165. 1972 Properties of the two polypeptides of sodium- and potassium-dependent adenosine triphosphatase. J. Biol. Chem., 247: 7642-7649. Lane, L. K., J. H. Copenhaver, Jr., G . E. Lindenmayer and A. Schwartz 1973 Purification and characterization of and [3H] ouabain binding to the transport adenosine triphosphatase from outer medulla of canine kidney. J. Biol. Chem., 248: 7197-7200. Nagano, K., T. Kanazawa, N. Mizuno, Y. Tashima, T. Nakao, and M. Nakao 1965 Some acyl phosphate-like properties of P32-labelled sodium-potassium-activated adenosine triphosphatase. Biochem. Biophys. Res. Commun., 19: 759-764. Nakao, T., Y. Tashima, K. Nagano, and M. Nakao 1965 Highly specific sodium-potassium-activated adenosine triphosphatase from various tissues of rabbit. Biochem. Biophys. Hes. Commun., 19: 755-'758. Nishigaki, I., F. T. Chen and L. E. Hokin 1974 Studies on the characterization of the sodiumpotassium transport adenosine triphosphatase. XV. Direct chemical characterization of the acyl phosphate in the enzyme as an aspartyl /3-phosphate residue. J. Biol. Chem., 249: 491 14916. Perrone, J. R., J. F. Hackney, J. F. Dixon and L. E. Hokin 1975 Molecular properties of purified (sodium + potassium)-activated adenosinetriphosphatases and their subunits from the rectal gland of Squalus acanthias and the electric organ of Electrophorus electricus. J. Biol. Chem., 250: 4178-4184. Post, R. L., and B. Orcutt 1973a Active site of phosphorylation of N a + , K+-ATPase. In: Organization of energy transducing membranes. M. Nakao and L. Packer eds. Tokyo University Press, Tokyo, pp. 35-46. Post, R. L., S. Kume and F. N. Rogers 1973b Alternating paths of phosphorylation of the sodium and potassium ion pump of plasma membranes. I n : Mechanisms in bioenergetics. F. Azzone and L. Ernster, eds. Academic Press, New York, pp. 203-218. Post, R. L., A. K. Sen and A. S . Rosenthal 1965 A phosphorylated intermediate in adenosine triphosphate-dependent sodium and potassium transport across kidney membranes. J. Biol. Chem., 240: 1437-1445.
205
Racker, E. 1972 Reconstitution of a calcium pump with phospholipids and a purified Ca+ + adenosine triphosphatase from sarcoplasmic reticulum. J. Biol. Chem.. 247: 8198-8200. Ratanabanangkoon, K., J . Dixon and L. E. Hokin 1973 Studies on the characterization of the sodium-potassium transport adenosinetriphosphatase. XI. Comparison of kinetic properties of the Furified with the impure membranebound enzyme from Squalus acanthias. Arch. Biochem. Biophys., 156: 342-349. Repke, K. R. H., R. Schon, W. Henke, W. Schonfeld, B. Streckenbach and F. Dittrich 1974 Experimental and theoretical examination of the flip-flop model (Na, K)-ATPase function. Ann. N. Y. Acad. Sci., 242: 203-219. Rhee, H. M., and L. E. Hokin 1975 Inhibition of the purified sodium-potassium activated adenosinetriphosphatase from the rectal gland of Squalus acanthias by antibody against the glycoprotein subunit. Biochem. Biophys. Res. Commun., 63: 1139-1145. Rose, S. P. R. 1963 Phosphoprotein as in intermediate in cerebral microsomal adenosinetriphosphatase. Nature, 199: 375-378. Sen, A. K., T. Tobin and R. L. Post 1969 A cycle for ouabain inhibition of sodium- and potassium-dependent adenosine triphosphatase. J. Biol. Chem., 244: 659643604. Simpkins, H., and L. E. Hokin 1973 Studies on the characterization of the sodium-potassium transport adenosinetriphosphatase. XIII. On the organization and role of phospholipids in the purified enzyme. Arch. Biochem. Biophys., 159: 897-902. Skou, J. C. 1957 The influence of some cations on an adenosine triphosphatase from peripheral nerves. Biochim. Biophys. Acta, 23: 394-401. 1971 Sequence of steps in the ( N a + K f )-activated enzyme system in relation to sodium and potassium transport. Cur. Top. Bioenerg., 4: 357-397. Stein, W. D., W. R. Lieb, J. D. Karlish and Y. Eilam 1973 A model for active transport of sodium and potassium ions as mediated by a tetrameric enzyme. Proc. Nat. Acad. Sci. U.S.A., 70: 275-278. Uesugi, S., N. C. Dulak, J. F. Dixon, T. D. Hexum, J. L. Dahl, J . F. Perdue and L. E. Hokin 1971 Studies on the characterization of the sodiumpotassium transport adenosine triphosphatase. VI. Large scale partial purification and properties of a lubrol-solubilized bovine brain enzyme. J. Biol. Chem., 246: 531-543. Yang, J. H. 1970 Possible role of pyrophosphate linkage in the active transport of sodium ions. Proc. Nat. Acad. Sci. U.S.A., 67: 59-61.
+
