The Influence of Some Cations on an Adenosine Triphosphatase from Peripheral Nerves

MILESTONES IN NEPHROLOGY J Am Soc Nephrol 9: 217()-2l77. 1998

JENS CHR. SKOU

Institute ofPhysiology, Universit’ ofAarhus (Denmark)

with COrnflleflts by

JENS CHR. SKOU andJOSEPH F. HOFFMAN

Reprinted from Biochirn. Biophys. Acta. 23: 394-401, 1957

Stimulation of a nerve leads to an influx of sodium ions into the fibre and hence to an increase in the intra-axonab sodium AUTHOR COMMENTARY concentratio&. Normal conditions are restored by an outward Jens Cht Skou transport of the sodium ions, and this process requires energy because the efflux takes place against an electrochemical gra- Institute ofBiophysics, University dient. The mechanism of this transport is not known. In experiments with giant axons from Sepia officinalic and ofAarhus, Denmark from lligoforbesi, HoIu.iN 1Nl) KE\NES2 found that dinitrophe- nob, azide and cyanide inhibit the active transport of sodium ions out of the nerve; this inhibition is revetsible. In the concentrations used all three substances also inhibit the oxydative phosphorybation which takes place in mitochondria; dinitro- 1953 I wrote to professor David Nachmansohn at phenol and azide do so through an uncoupling of the phos- I Columbia University in New York and asked if I could photylation4, and cyanide through an inhibition of the oxy- come in August and September and prepare some acetyl- dation5. CALDWELI observed correspondingly that addition of cholinesterase from electric eel. I wanted to test the effect of these substances, in the concentrations used by HODGKIN AND surface pressure on the activity of an enzyme incorporated KE\NES, led to a reduction of the content of energy-rich phos- into a monomolecular layer of nerve membrane lipids on a phate esters in the axoplasm ofgiant axons. This seems to mdi- cate that energy-rich phosphate esters are somehow involved in water surface. I was a medical doctor who was interested in surgery, had taken a break in my clinical training, and got a the active transport of sodium ions out of the nerve fibres. In this connexion it is of interest that LIBET7 and ABOOD position at the Institute of Physiology at Aarhus University ANt) GnRn8 weie able to demonstrate an adenosine triphos- in order to write a thesis on the mechanism of action of local phatase (ATPase) in the sheath of giant axons. A further study anesthetics. I used a monomolecular layer of nerve mem- on the ATPase in nerves and its possible robe in the active out- brane lipids on a Krebs-Ringer water surface as a model for ward transport of sodium ions seems warranted. the nerve membrane. Local anesthetics placed in the water According to LIBET, the ATPase in the sheath ofgiant axons is phase penetrated the monolayer, and there was a correlation calcium-activated, while the experiments by MooD AND GERARD between the anesthetic potency and the increase in pressure suggest that it is activated by magnesium and located in submi- in the monolayer. This suggested to me that local anesthet- croscopic particles. In peripheral nerves from the rat the latter ics, by an effect on the lipid part of the nerve membrane, authors found both a calcium- and a magnesium-activated blocked for a conformational change in the proteins in the ATPase. The calcium-activated enzyme was predominantly bocat- membrane, leading to the increase in permeability for sodi- ed in the mitochondria, while the magnesium-activated, as in urn, and thereby blocking the nerve impulse. To test this, I giant axons, was mainly located in the submicroscopic particles. needed a lipid monolayer with a protein incorporated, and to Giant axons were not available to us. In preliminary exper- show that penetration of local anesthetics into the lipid part irnents we found that a homogenate of leg nerves from the of the monolayer influenced the conformation of the protein. shore crab (Carcinus inaenas) contained both a calcium- and The protein responsible for the permeability increase for a magnesium-activated ATPase, and that their localization sodium was unknown; instead, I decided to form a mono- was sirnilai to that of the ATPase found by AB0OD AND GERARD layer of lipids with an enzyme incorporated, and use a in rat-nerve homogenates. For our study we have chosen the change in enzymatic activity as an indication of a confor- magnesium-activated enzyme, because it resembles the mag- mational change of the protein. For the experiments, I need- nesium-activated ATPase from the sheath of giant axons in ed a membrane-bound enzyme with a high activity. Acetyl- that it is located in submicroscopic particles. cholinesterase was a candidate. The present investigation is concerned with the effect on In his reply, Professor Nachmansohn said that he intend- the enzyme activity exerted by the cations normally present in ed to spend August at the Marine Biological Station in the tissue-sodium , potassium , magnesium and calcium. Milestones in Nephrology 2171

EXPERIMENTAL Woods Hole, and suggested that I join him there. In The ATPase %%?55 prepared bs homogenization and subsequent differ- ential centrifugation5 of leg nerves froni the shore crab ( Garcinus September we could then return to New York and prepare mam as). the enzyme. I did not know what I would do in Woods Hole The isolated nerves were washed and honiogenized in 10 volumes of but accepted. This was a lucky choice. It was a great experi- 0.25 M ice-cold sucrose httffered with histidine, 30 niM/l, pH 7.6. The admixture of alkali metal ions was avoided by using the histidine base, ence to spend a month in surroundings where scientists and the pH was adjusted by addition of 1 N HC1. The homogenate was came from all over the world to work during the summer, centrifuged in a Servall angle centrifuge at OC. Fragments and stroma because there was access to squids and thereby to giant were removed by centrifugation at 2,000 X g for I 5 minutes, IllitoChon- dna and s()Ilse submicroscopic particles by centrifiigation at 10,000 X g axons. I did not take part in the experimental work, but for I 5 minutes. The supernatant was ceiitrifuged at 20,000 X g (maxi- looked, listened, and learned. In between I read scientific mum available) for 3 hours. During this centnfugation the temperature rose from 0’ to 8-10#{176}C. papers, and in one written by Nachmansohn it was men- The sediment after this last centrifugation was suspended in a vol- tioned that in 1948 Libet had shown that there is an ATP unie of 0.25 /i,f ice-cold buffered sucrose corresponclmg to one half of hydrolyzing enzyme in the sheath part of the giant axon (1). the volume of the original homogenate. This suspensioll was centrifuged at 10,000 x g for 10 minutes in order to remove an remaining mito- I knew that ATP was the energy source in cells, and won- cliondria. The final supernatant was used as enzyme solution in the dered why there was an AlP hydrobyzing enzyme in the experIments: it contained from one half to two thirds of the activity orig- nerve membrane. Furthermore, being a membrane-bound inably present iii the homogenate. The enzvtne was relatively unstable; when it was stored in a refrigera- enzyme it could be a candidate for my monolayer work. tor at 4-5#{176}C, its activity fell to oe half in 3-4 days. I decided to take a book at the enzyme when I returned The disodium or diharium salts of ATP and the barium salt of ADP to Aarhus, a decision that turned out to have far-reaching (Sigma products) were converted into free acids, the Na salt by passage through an Ansherlite 120 (H forni) column and the barium salts by consequences. precipitation with equimolar amounts of sulphuric acid and subsequent In September I prepared acetylchobinesterase at Cobum- passage through au Ansl)erlite 1 20 colunn to renave residual hariuni iOns. The free acids were neutralized to pH 7 with a 1 A! solution of 2- bia University, and then returned to Aarhus where I contin- arino-2-I11ethvl-l .3-propanediol. In the concentrations used, this sub- ued my monolayer experiments. During the autumn of 1954 stance does not affect the enzyme activity and is without influence on the I started to look for the nerve membrane enzyme. I had phosphate determination. ATP and ADP were (letermined spectrophotonietricallv” and h mea- no access to giant axons, but decided to use crab nerves surement of 7’P; Na and K were determined ssith a flame spectropho- instead, and arranged with a fisherman to deliver some tometer, Beckman DU, flame attachment No. 920(). The pH was mea- crabs. The membrane fragments of sciatic crab nerves sured witls a glass electrode, Radiometer PHM 3. Inorganic phosphate was determined by the method of FIsKE ANt) StIsBRow1#{176} with aniidol as showed a Mg2tactivated ATPase activity, which was slight- the reducing agent. ly increased by Na, while K had no effect. Then I ran into The activity of the enzyme was determined in a volume of I .0 ml con- problems. I was unable to reproduce my results, as the mea- taming 0. 1 I1l of the above-i’nentioned enzynse solution. The reaction mixture was buffered with histidine, 30 mM/b, pH 7.2. Unless otherwise sured enzyme activity varied. I continued, but was interrupt- stated, the mixture contained 3 mM ATP/l. The cations were added as ed by the Christmas holidays, the spring semester teaching, solution of their chloride. All experiments were performed at 36#{176}C. and it was After a lO-mintite temperature equilibration the experiment was started and summer holidays, not until the autumn of by the addition of enzyme, and the rnixttire was inctibated for 15 or 30 1955 that I got the answer, namely that the enzyme for activ- minutes, according to the reaction velocity. The hydrolysis of ATP was ity required a combined effect of Na and K. This had not linear within the experimental period. The reaction was stopped b the addition of 0.1 1111 of 50% trichloroacetic acid, and after centrifugation occurred to me, and the explanation for the variable results aliquots of 0.4 nil were taken out for determination of inorganic phos- was that, apparently, I sometimes had converted the barium phate. salt of AlP to the Na salt, sometimes to the K salt, and In the absence of added cations, the reaction mixture contained small amounts (about 0. 1 mAt/I) of sodium and potassium, which origi- that the nerves sometimes were homogenized in KCI, some- nated from the enzyme solution and from the ATP. times in sucrose, and therefore the combination of Na and In all the diagrams except Fig. 1 the enzyme activity is expressed as K in the medium had varied. Knowing that there was no i.g of P split off from ATP in 30 mintltes. effect of K on the activity, and only a slight effect of Na,

RESULTS I had not bothered to investigate whether there was Na or K in the medium. When the substrate is ADP, no inorganic phosphate is biber- But what was the function of a nerve membrane ATPase ated (data not presented) . In the presence of ATP the activated by a combined effect of Na and K? I was inter- hydrolysis stops when inorganic phosphate corresponding ested in the effect of local anesthetics on nerve conduction, to one phosphate group has been split off (Fig. 1). and my first reaction was that it was the sodium channel in Fig. 2 shows that under the experimental conditions used the nerve membrane. However, it seemed unlikely from the pH optimum is 7.2. what was known about nerve conduction, that the opening of The enzyme activity is zero when no cations are added or the membrane for sodium was ATP-dependent. It seemed when K or Ca alone are added to the reaction mixture. more likely that the enzyme was involved in the active trans- The addition of Na gives rise to a just measurable activity, port of Na, and that AlP was the substrate for the transport. which is independent of the sodium concentration. When I knew very little about active transport, and looked for arti- Mg is added, the enzyme shows a slight activity with a des on the substrate for the active transport. I found that maximum at 3 mM/b (Fig. 3). Hodgkin and Keynes (2) had shown that poisoning of giant If sodium ions are added to a reaction mixture which axons with dinitrophenol, azide, or cyanide inhibited the already contains magnesium ions, the enzyme activity active transport, suggesting that the substrate is energy-rich increases (Fig. 4). When the concentrations ofMg and of phosphate esters. And, as AlP is an energy-rich phosphate ATP are 3 mM/I, a maximum activity is observed when the 2172 Journal of the American Society of Nephrology

ester, this supported the idea that the ATPase was involved in the active transport. P IA I wrote the article on the crab nerve experiments in 1956, in which I suggested that the enzyme was involved in the active transport of sodium. It was published in 1957. The word pump was considered as part of the title, but I found it too provocative. The title became “The influence of some cations on the activity of an adenosine triphosphatase from U peripheral nerves?’ No wonder, that very few took notice of it, or connected it to active transport. As can be seen from Mg6K120Na80mP/l b the list of references, I referred to very few articles on active Fu;. I. Liberation of inorganic 60 7.0 8.0 9.0 Considering phosphate with ATP as substrate. p14 transport. that the activity of the enzyme re- The dotted line itIicates first P of quired a combined effect of Na and K, today I am sur- ATP. Abscissa, time in minutes: ordi- Fi;. 2. Relation between enzyme prised that I did not cite papers in which it was shown that nate. pg P removed ffl)I ATP. The activity and pH. Abscissa, pH; ordi- reaction mixttlre contained 5 m%I nate, ig P removed from ATP in 30 extracellular K4 activates the Na efflux. It may reflect the Mg, So mM Na and 120 mM K per minutes. The reaction mixture con- limited knowledge I had of the field, but also that my main litre. The initial concentration of tamed 6 mM Mg. 80 mM Na, 120 interest was the monolayer experiments. The crab nerve ATP was 1.16 mM/I. mM K and 3 hiM ATP per litre. experiments were done in between the monolayer experiments, which were my main interest, and I was not aware of sodium concentration is 6 mM/I. At higher sodium concen- the importance of the observations. trations, the activating effect of Na decreases. There was a crucial experiment I had not done. In 1958 I Addition of potassium ions to a mixture containing mag- presented a paper on the monolayer experiments at the nesium ions does not affect the enzyme activity; addition of Fourth International Congress on Biochemistry in Vienna, calcium iOflS results in an inhibition. Austria. Here I met Robert Post whom I knew from Woods If both magnesium and sodium ions are present in the mix- Hole. He told me that he had shown on red blood cells that ture, addition of potassium ions results in a further increase in 3 Na are actively transported out in exchange for 2 K. the enzyme activity (Fig. 5). When the potassium concentration After I told him about the experiments on the crab nerve is raised, the enzyme activity reaches a maximum and then de- enzyme, he asked me if the enzyme was inhibited by creases, and at high potassium concentrations it is smaller than ouabain. I did not know that H.J. Schatzmann in Switzerland it was when no potassium ions were added. The batter phenom- in 1953 had shown that cardiac glycosides, of which oua- enon is seen only from the curves representing 3 and 10 mM/b. The curve for 3 mM/I shows that high concentrations bain is the most water-soluble, specifically inhibits the active transport of sodium and potassium in red blood cells (3). of potassium inhibit only that part of the activity which is due to Nat, hut not the part which is due to Mg. When Robert Post came to Aarhus after the conference, I It is further seen from Fig. 5 that the maximum activity could show him that ouabain inhibited the enzyme, which obtained in the presence of potassium increases with the convinced him that the crab nerve enzyme was involved in sodium concentration. The potassium concentration required the active transport of sodium and potassium. for maximum enzyme activity also depends on the sodium I had started to book for the enzyme in red blood cells, concentration; it incieases when the sodium concentration is which I had learned is a classic test object for investigations increased. Maximum enzyme activity is obtained when the on active transport. Robert Post asked if he could do the potassium concentration is roughly equal to that of sodium. experiments on red blood cells when he returned to the Finally, it should be observed that the inhibition due to high United States. As I had no experience with experiments on potassium concentrations decreases when the concentration red blood cells and he had, I agreed and turned to other tis- of sodium is increased. The effect of potassium ions depends sues. In 1960 he published his paper, “Membrane adenosine triphosphatase as a participant in the active transport of sodi- FL1‘P urn and potassium in human erythrocytes” (4), in which it 4 a very convincingly was shown that there is a correlation 2 between the effect of the cations on the activity of the enzyme and on active transport. Robert Post was known in the transport field and his paper had a better title than mine, - a a so it attracted more attention. Experiments on many different tissues were published in the following years, which showed that the enzyme was involved in the active transport of Na and K, and from 0 50 100 this it was possible for me in a review paper in 1965 (5) to mM/I conclude that the enzyme system fulfilled all of the require- Fi;. 3. EIzYfl1e activity in relation to ments of a sodium-potassium pump. The enzyme was the concentration of Mg#{176},Ca named the Na,K-ATPase. I gave up monolayer experiments, Na ‘ ,or K . Abscissa, ion concen- trallon III mM/I: ordinate. pg P re- concentrated on the enzyme, and never became a surgeon. IflO’e(l Irons ATP in 30 mintltes. Milestones in Nephrology 2173 accordingly not only on the presence of magnesium and sodi- References urn ions, but also on the concentration of sodium ions. 1. tibet B: Adenosine triphosphatase (ATPase) in nerve [Abstract]. Fed Pmc7: 72, The addition of K has also an effect on the relation be- 1948 tween enzyme activity and sodium concentration (Fig. 6). In 2. HOdgkin AL, Keynes an: Active transport of cations in giant axons from Sepia and Loligo.JPhysiol 128: 28-60, 1955 the absence of K, the activity reaches a maximum at 6 mM S. Schatzmann HJ: Herzglykoside ala Hemmstoffe der aktiven Kalium Na/b; when more sodium is added, the activity decreases. Natrium Transport durch die Eythrocytenmembran. Hdv PhysiOl Pharinacol Aaa 11: 346-354, 1953 When as little as 3 mM K/b is added to the system, the activi- 4. Post RI, Merritt GR, Kinsolving CR, Aibright CD: Membrane adenosine ty is not only enhanced but shows a steady rise with the sodi- uiphosphatase as a participant in the active transport of sodium and potassi- urn concentration until it finally levels off. The sodium con- urn in the human erythrocyte.JBioI Chem235: 1796-1802, 1960 5. Skou JC: Enzymatic basis for active transport of Na and K#{176}across the cell centration at which this level is reached increases with the membrane. Physiol Rev 45: 596-617, 1965 amount of potassium added. A certain, bow activity in the absence of sodium is observed in Fig. 6. It is, as previously pointed out (cf Fig. 4), due to the presence of magnesium, and it is independent of the potassi- GUEST COMMENTARY urn concentration. In the presence of high concentrations of potassium, 200 and 350 mM/I, the addition of small amounts Joseph E Hoffman of sodium, 3 and 6 mM/b, leads to an inhibition of this low, magnesium-dependent activity. Higher concentrations of Eugene Higgins Professor of sodium have, as usual, an activating effect. Cellular and Molecular Physiology, It appeared from Fig. 4 that calcium ions inhibit the activi- Yale University School of Medicine, ty which is due to the presence of Mg4. An inhibition is also New Haven, Connecticut seen when calcium ions are added to a system containing magnesium + sodium ions or magnesium + sodium + potassium ions (Fig. 7). When Mg or Mg + Na are the only cations present, ntil the call to write this commentary, I had no reason the optimum magnesium concentration is 3 mM/b. In the U to peruse again Skou’s paper (1) since it was first pub- presence of potassium the optimum concentration of mag- bished. While rereading it some 40 years later, I was even nesium is 6 mM/l, and this value is independent of the potassium concentration. If calcium ions are also added, the opti- more impressed with the scope and clarity of the experi- mum magnesium concentration becomes still higher, and it ments, but I also reflected once more on why there was such increases with the calcium concentration (Fig. 8). a delay by the witnesses at the time (myself included) in rec- In Figs. 9 and 10 is shown the relation between enzyme ognizing that the ATPase that Skou had discovered was in activity and sodium concentration in the absence and pres- fact the Na/K pump. General acceptance was slow in com- ence of calcium ions at various concentrations of potassium ing and I believe the lag was based in part on not knowing and magnesium. It appears that the activating effect ofNa is for sure what the proximate energy source for the pump was proportional to the Mg:Ca ratio. but more so on the need to develop critical evidence show- It was previously noted (Fig. 6) that low concentrations of ing that the Na, K-ATPase, per Se, functioned as the Na/K sodium in the presence of high potassium concentrations pump. Even so, one should be aware that by 1957, active inhibited the enzyme activity due to magnesium. Figs. 9 and transport of Na and K was a rapidly maturing field. It had 10 show that this is also the case in the presence of calcium. not only been defined (2,3), but active K uptake and Na extrusion had been experimentally demonstrated (4), shown to be tightly coupled and metabolically dependent (4,5). In A9 P addition, the Na pump had its specific inhibitor identified (6) and had early on been given its name based on a solid rationale (7). Considering the energy source for the Na pump, red blood cells and their ghosts have clearly played a critical role in helping to establish the immediate substrate of the pump. The first experiments involved entrapment of ATP within ghosts at hemolysis and the finding that active uptake 1 of K occurred (8). This was so even in the presence of arse- 1T nate, which inhibits the metabolic production of ATR 0 Ca Unfortunately, the ghost system that was used continued to 6 o o o glycolyse, and the ghosts accumulated K in the presence of 0 2 4 6 8 arsenate, although less so than with added AlP. In addition,

Ft;. 4. Enzyme activity in relation to the the suspension contained a large proportion of intact cells, concentration of Na ,K#{176}, Ca in the pres- which together with the lactate production, also clouded the ence of Mg, 3 mM/I. Abscissa, ion concentration in mM/I: ordinate, p.g P removed interpretation of the results (9). Thus, although it could be from ATP in 30 Illillutes. 2 174 Journal of the American Society of Nephrology

said that high energy phosphate from ATP was used to run Mg 6 mfrt/l the pump, close scrutiny would indicate that there was no basis, other than by inference, to conclude that ATP itself, rather than some other metabolic intermediate, was used 3c directly by the Na pump. This same limitation applies to analogous experiments with squid axons in which AlP and other metabolic substrates and inhibitors were injected 2C directly into the fibers (10). In these experiments, the active extrusion of Na could be related to phosphate bond energy. ic It should be noted that these fibers still contained unknown amounts of cytoplasm and mitochondria. On the other hand, in the mid-b950s a cleaner and better-defined system of ghosts had been developed and used to identify AlP unequi- ‘.# 0 40 60 80 100 120 KCI mM/I vocally as the proximate substrate of the pump (1 1,12). These studies complemented work on intact red cells, where FI( .. . Eli/vIne d(tIS’itV in relation to tls(. concentratu)n of K in the pIcsenct of Mg , 6 Ill i%1/I,IIl(1 different concentrations of Na . Abscissa, potassiunl metabolism had been ingeniously manipulated, and in concentration ill IuA%’I/l; ordinate. tg P renioved froI ATP in 30 IiI5utes. which there was a strong correlation between Al? use and

The effect of Li vas studied in a few experiments. Lithium pump activity (13,14). iOIIS do i’iOt affect the activity due to magnesium, but if the In light of the now established relationship between ATP system contains both magnesium and sodium, the addition of and the pump, it would seem natural to conclude that Skou’s

Li results iii an increase ofthe activity. The action of lithiuni Na, K-ATPase was the biochemical equivalent of the pump is consequently similar to that of potassium, but it should be itself. Part of the problem was that Skou’s crab nerve prepa- noted that at low concentrations lithium has a weaker effect ration was vesicular and its relation to the nerve membrane than potassium, while its effect is stronger at high concentra- was unspecified. Another aspect was that, given the example tiotis (Fig. 1 1). of the ATPase associated with myosin, the ATPase could act I)ISCUSSION to supply the energy without being the contractile mechanism or, in Skou’s case, being the pump. In fact, in 1960, I In a brain-tissue homogenate, UrrER’2 demonstrated a magne- concluded an abstract stating, “These results ... support the sium-activated apyra.se, the activity ofwhich varied with the sodi- concept that the pump complex is or has as an intimate corn- tim concentration. Homogenates of rat nerves8 and of crab ponent an ATPase” (1 1). Thus, given direct correspondence nerves contain not only a magnesium-activated ATPase, but also of kinetic parameters, such as activation, inhibition, Na and an adenylic kinase. It seems possible that the apyrase activity K competition, between the Na pump on the one hand and demonstrated by U’ITER was due to the combined effects of magnesium-sodium-activated ATPa.se and an adenylic kina.se. the Na, K-ATPase on the other, although instructive, does The ATPase from crab nerve studied here is magnesium- not establish their identity (15). activated and located in submicroscopic particles. In these The evidence that convincingly established that the Na, respects it resembles the ATPase isolated from rat nerve and K-ATPase was indeed the Na pump itself came from many from the sheath of giant axons by AB000 AND GERARD8 and sources and spanned many years. In my view, the critical also the ATPase isolated from muscle by KIELI,Y AND steps that led to this identification involved studies on phosphorylation of the Na, K-ATPase (16-22), its purification

pyP (23-26), and its incorporation into phospholipid vesicles K 20 (27). The phosphorybation studies were important for de- fining various intermediates that led to the correlation of the events of transphosphorylation of the Na, K-ATPase with the events of transbocation of the transported ions. Purification of the Na, K-ATPase was primary in establish- ing the enzyme’s subunit composition and a myriad of other Mg 6 mN/I properties, but none more crucial than by providing a preparation that could be reconstituted into vesicles that could then be shown to transport Na in exchange for K in concert with its known enzymatic activities. Thus, we come from Skou’s original discovery to the identity of the enzyme as the molecular device that mediates the active transport mM/I of Na and K. It is also noteworthy that the Nobel Committee ultimately came to recognize, in 1997, Skou’s pivotal FI;. b. I-n1%I1R actis’itv in relation to the concentration of Na in the presence of Mg4 , 6 Ili11/l, aIKl diflirent concentrations of K . Abscissa, soclinni con- achievements. centration iI n1f/l; ordinate. ig P IeI1oved holil AlP in 30 ninLItes. Milestones in Nephrology 2175

FLgP References

1. Skou JC: The influence of some cations on an adenosine triphos- 20 phatase from peripheral nerves. Bioclzinz Biophys Ada 23: 394-401, 3. K8Na8Mg4mt./I 1957 2. Ussing HE: Transport of ions across cellular membranes. Physiol Rev 15 2. Na8Mg4mFVI 29: 127-155, 1949 1. Mg4rn?yI 3. Rosenberg T: The concept and definition of active transport. In: 10 Active Transport and Secretion, Symposia of the Society for Experimental Biology No. VIII, New York, Academic Press, 1954, pp 27-41 5 4. Harris JE: The influence of the metabolism of human erythrocytes on their potassium content.JBiol Chem 141: 579-595, 1941 5. Harris EJ, Maizels M: The permeability ofhuman erythrocytes to Na. C JPhysiol 113: 506-524, 1951 0 2 4 6 8 10 6. Schatzmann HJ: Herzgbykoside als Hemmstoffe f#{252}rden aktiven Cad2 mI’1/L Kalium- und Natriumtransport durch die Erythrocytenmembran. Helv Physiol PharmacolActa 11: 346-354, 1953 Fi. 7. Enzvnie activit’ in relation to the con- 7. Dean RB: Theories ofebectrolyte equilibrium in muscle. Biol Symp 3: centration of Ca in the presence of Mg. 331-348, 1941 Mg + Nat, or Mg + Na + K. Abscissa, 8. Gardos G: Akkumulation der kaliumionen durch menschliche Blutk calcium concentration in niM/l; ordinate, p.g 216 #{248}rpeerchen.Ada Physiol HungAcad Sd 6: 191-199, 1954 P removed from ATP in 30 minutes. 9. Hoffman JF: The link between metabolism and active transport of sodium in human red cell ghosts.JMembrBiolS7: 143-161, 1980 10. Caldwell PC, Hodgkin AL, Keynes RD, Shawl’!: The effects of inject- MEYERHOF11. There are further points of similarity between ing “energy-rich” phosphate compounds on the active transport of the magnesium-activated ATPases from crab nerve and from ions in the giant axons of Loligo.JPhysiol 152: 561-590, 1960 muscle. Their pH optima are roughly identical, 7.2 and 6.8, 11. Hoffman JF: The link between metabolism and active transport of Na in human red cell ghosts. Fed Pmc 19: 127, 1960 respectively, and they are both strongly inhibited by calcium 12. Hoffman JF: Cation transport and structure of the red cell plasma ions. The effect of sodium ions on the activity of the magne- membrane. Circulation 16: 1201-1213, 1962 siurn-activated ATPase from muscle has not been studied. 13. Dunham ET: Linkage of active cation transport to Al? utilization. Physiologist 1: 23, 1957 The experiments reported here show that the activity of 14. Dunham ET: Parallel decay of AlP and active cation fluxes in the ATPase from crab nerve is highly dependent on the rela- starved human erythrocytes. Fed Proc 16: 33, 1957 15. Post RL, Merritt CR, Kinsolving CR, Albright CD: Membrane adeno- tive concentrations of the four cations Nat, K, Mg and sine triphosphatase as a participant in the active transport of sodium Cat The presence of magnesium ions is an obligatory and potassium in the human erythrocyte. J Biol C/tern 235: requirement for the activity of the enzyme; sodium ions 1796-1802, 1960 16. SkouJC: Further investigations on a Mg4 + Na-activated adenos- increase the activity when magnesium ions are present; potas- intriphosphatase, possibly related to the active, linked transport of sium ions increase the activity when the system contains both Na and K across the nerve membrane. Biochim Biophys Ada 42: magnesium and sodium ions. In high concentrations potassi- 6-23, 1960 17. ChamockJS, Post RI: Evidence of the mechanism of ouabain inhi- urn ions inhibit that part of the activity which is due to Nat, bition of cation activated adenosine triphosphatase. Nature 199: while the activity due to Mg is not affected. Calcium ions 910-911, 1963 inhibit the activity under all conditions. 18. Albers RW, Fahn S. Koval GJ: The role of sodium ions in the activa- don of electrophorus electric organ adenosine triphosphatase. Proc In the presence of magnesium or of magnesium + sodium NaIIACWJ Sci USA 50: 474-481, 1963 the optimum magnesium concentration was found to be 3 19. Post RL, Sen AK Rosenthal AS: A phosphorylated intermediate in mM/i when the ATP concentration was also 3 mM/b. Pre- adenosine triphosphate.dependent sodium and potassium transport across kidney membranes.JBiol C/zero 240: 1437-1445, 1965 liminary studies indicate that the optimum concentration of 20. Albers RW, Koval GJ, Siegel GJ: Studies on the interaction of magnesium is equal to the ATP concentration also at other ouabain and other cardioactive steroids with sodium-potassium-acti- ATP levels. From this, one might infer that the substrate for vated adenosine triphosphatase. Mol Pharmacol4: 324-336, 1968 21. Post RL, Kume S, Tobin T, Orcutt B, Sen AK flexibility ofan active the enzyme is magnesium-ATP. center in sodium-plus-potassium adenosine triphosphatase. J 0ev The differences between the effects of sodium and potassi- Physiol 54: S306-S326, 1969 urn suggest that the activating effects of sodium and potassi- 22. KyteJ: Phosphorylation ofa purified (Na + Ki adenosine triphosphatase. Bloc/zero Biophys Res Commun 43: 1259-1265, 1971 urn differ in their point ofattack, and that the inhibitory effect 23. J#{248}rgensen PL, Skou JC: Preparation of highly active (Nat + K)- of potassium is due to an interference with the sodium acti- ATPase from the outer medulla ofrabbit kidney. Biodsetn Biophys Res vation. An hypothesis to fit this would be that the substrate Cosnmun 37: 39-46, 1969 24. J#{248}rgensenPL, SkouJC: Purification and characterization of (Na-i- + most readily attacked by the enzyme is sodium- K)-ATPase. I. The influence ofdetergents on the activity of (Na + rnagnesiurn-ATP. Addition of K might then lead, on the one K)-ATPase in preparations from the outer medulla ofrabbit kidney. hand, to a direct stimulation of the enzyme in the presence Bwchim Biophys Ada 233: 366-380, 1971 25. J#{248}rgensenPL,SkouJC, Solomonson 12: Purification and character- of Mg and Na and, on the other, to a displacement of ization of (Nat K)-ATPase. II. Preparation by zonab centrifuga- sodium from the substrate, resulting in an inhibition. tion ofhighly active (Na + K)-ATPase from the outer medulla of This assumption would explain, firstly, why potassium is rabbit kidneys. Biochim Biophys Acta 233: 381-394, 1971 26. KyteJ: Purification ofthe sodium- and potassium-dependent adeno- activating only in the presence of magnesium + sodium. sine triphosphatase from canine renal medulla. J Biol Qsem 246: Secondly, it would explain the relation between enzyme 4157-4165, 1971 27. Hilden S, Rhee HM, Hokin LE: Sodium transport by phospholipid activity and potassium concentration (cf Fig. 5) . The rise of vesicles containing purified sodium and potassium ion-activated activity on addition of K must then be due to a direct stimu- adenosine triphosphatase.JBiol Chem 249: 7432-7440, 1974 lation of the enzyme by this ion. With an increase in the po- 2176 Journal of the American Society of Nephrology

Fi;. . Eli/vine activity in relation to the concentration of Mg in the presence of K . 1 20 Ii,I/l. Na ,Mo mAt/I, and diflerent concentrations of (a . Abscissa, nlagnesiunl concentration in n3f/l: ordinate, pg p removed froni ATP in 30 minutes.

FI;. 10. Enzvne activity in relation 10 the concentration of Na in tlit )r(s-

ence of K , 350 hiM/I, and dillerent concentrations of Mg and (a . tassiuni concen tration this stimulation must be counteracted Abscissa, sodisins concentration in IisU/l; ordinate pg P renu)ved fron ATP in by an inhibition due to displacement of sodium from the 3() niinutes. substrate. The activity must consequently pass through a maximum and eventually approach the level observed when tions are not accounted for in this way, e.g. the decrease in no sodium was added to the system. the activating effect of Na above a certain concentration in Thirdly, it would explain the dependence of the potassium a system which does not contain K, and the inhibitory effect effect on the sodium concentration. The higher the sodium of small amounts of sodium in the presence of a high potas- concentration, the higher must be the potassium concentra- sium concentration. Further studies may clarify these points. tion required to give a certain displacement of sodium from The effect of calcium on the enzyme is purely inhibitory. the substrate. Since a higher potassium concentration means a The inhibition is counteracted by the addition of extra inag- stronger activation of the enzyme, it follows that the maximum nesium and may therefore be due to a competition between enzyme activity obtained by the addition of potassium must calcium and magnesium. increase with the sodium concentration. The potassium con- It may now be asked whether the present studies have pro- centration required to give maximum activity must also duced any evidence ofa connexion between this enzyme and increase with the sodium concentration. Fig. 5 shows that max- the active extrusion of sodium ions from the axon. imum activity was obtained when the concentration of potassi- The process responsible for this transport must presum- tim was approximately equal to that of sodium. The concen- ably be located in or in close proximity to the nerve mem- tration of potassium required to reduce the enzyme activity to brane. In homogenates of crab nerve the enzyme studied the level observed when no sodium was added must accord- here is located in submicroscopic particles, and we do not ingly also increase with the sodium concentration. know its localization in the intact nerve, but it is suggestive Although the above-mentioned assumption could explain that ABOOD AND GERRi were abbe to isolate from the sheath a majority of the data, it should be noted that a few observa- of giant axons an ATPase which was also magnesium-activated and located in submicroscopic particles. As previously pointed out, studies by HODGKIN ANt) KEYNES and by CAI.DWELL indicate a connexion between the sodium extrusion from nerve and the metabolism of energy-rich phosphate esters. The substrate of ATPase is an energy-rich phosphate ester. HoD;iiN AND KE\Ni.stt have shown that the sodium efflux from giant axons depends on and is directly proportional to the intra-axonal sodium concentration; they injected sodium into the axons, and the intra-axonal sodium concentrations in their experiments varied from the 40 mM/b normally present to 130 mM/I. In the crab nerve the intra-axonal potassium concentra- 0 50 00 io ou tion is 342 mM/kg axoplasm’1; this is calculated from deter- MaCI mi’/I minations of the potassium and sodium concentrations in

Fu;. . Enivnic activity ill relation to the concentration ofNa in the presence whole crab nerves on the assumption that the intra-axonal of Mg . () I1U’I/l, and difkrent concentrations of K and Ca . Abscissa, sodium concentration is the same as in giant axons, viz. 40 SO(litlIii concentrations in niAf/I; ordinate. ig P renioved from ATP in 30 IliiIiiltCS. mM/kg axoplasm. Milestones in Nephrology 2177

P intensification of the enzyme activity. This observation, as well as some p1.g other characteristics of the system, suggest that the adenosine triphos-

phatase studied here may be involved in the active extrusion of sodium 12 from the nerve fibre.

Mg6 NaC mM/I REFERENCES

R. D. 1:ss. J.!‘hy.siol.(IOfl(IOfl).I 14 ( 1951 ) I 19. 8 : A. L. H()1)(kIN ANE) R. I). Kiros.J. J’hssiol. (London), 128 (1935) 28.

‘ \v. F. Losiis AXE) F. L1I’SIANN. Federation Proc., 12 (1953) 218. 4j. D.Jto..ui, Biodzern.J., 49 (1951) 271. 1). KnIJN, Proc. Ro’. Soc. (London), B 121 (1936) 165.

I) P. C. (u1avF:11,J. Physiol. (London), I 32 ( I 956) 35 P. 7 B. LIRFT, Federation Proc.. 7 ( 1948) 72. S L. G. .kisxn R. W. (;F.iu),J. G’l!u!nr (ornp. Phss,ol., 43 ( 1954) 379. 9 H. M. KsLct’R.J. Biol. (:h,., 167 (1947) 445. I0 C. H. Fisiu SN!) Y 5u88usow,J. Biol. (.5cm., 66 ( 1925) 375.

‘J -------- 11 #{149} %#{149}KIF:I.1YANI) 0. Mn}:RIIOF,/. Biol. (:hen., 176 (1948) 591. 0 40 80 ru lOU mM/I 12 M. F. UTIFR,J. Biol. (e,,i., 185 (1950) 499. ‘3 A. L. l-Icoa.iaN SNI) R. D. Ki:NFs,J. I’/sio1. (london), 131 ( 1956) 592.

‘4 R. D. Ki:vrs AND R. R. 1.:uis,J. Phsio1. (linidon;, 1 13 ( 1951 ) 73.

Fi;. I I . Enzyme activity in relation to the concentration of K or Li in the presence of Mg , 6 mM/i, and Na , 6 mM/I. Abscissa, ion concentration in inAI/1; ordinate, p.g P removed froisi ATP in 30 niinutes.

The present experiments showed that in the presence of 350 mM K/b the activity of the magnesium-activated ATPase is highly dependent on the sodium concentration, and there is an approximate direct proportionality between the enzyme activity and sodium concentration within the concentration range used in the experiments of HODGKIN AND KEYNES (cf Fig. 10). The extent of the changes in enzyme activity is influenced by the Mg:Ca ratio in the system, but the linearity is observed at all magnesium and calcium concentrations used. If, in analogy to what happens in the nerve after stimulation, a rise in the sodium concentration is accompanied by a decrease in the potassium concentration, the ATPase activity is further enhanced (cf Fig. 9). The above considerations show that the crab-nerve ATPase studied here seems to fulfill a number of the conditions that must be imposed on an enzyme which is thought to be involved in the active extrusion of sodium ions from the nerve fibre. Further studies on the enzyme and its relation to the cations may serve to throw light on the nature of this process.

SUMMARY

Leg nerves from the shore crab ( (ardnus inaeiias) contain an adenosine triphosphatase which is located in the submicroscopic particles. The info- ence of sodium, potassium, magnesium and calcium ions on this enzyme has been investigated. The presence of magnesium ions is an obligatory requirement for the activity of the enzyme. Sodium ions increase the activity when magnesium ions are present. Potassium ions increase the activity when the system contains both magnesium and sodium ions. Potassium ions in high concen-

tration inhibit that part of the activity which is due to Na , while the activity due to Mg is not affected. Calcium ions inhibit the enzyme under all conditions. When Mg or Mg + Na are present in the system, the optimum magnesium concentration is equal to the concentration of ATP. If potassium ions are added, the optimum magnesium concentration is doubled. If calcium ions are also added, the optimum magnesium concentration becomes still higher, and it increases with the calcium concentration. A majority of these observations may be explained by assuming (a) that the substrate most readily attacked by the enzyme is sodium- magnesium-ATP, (b) that potassium ions stimulate the enzyme directly, and (c) that an increase in the concentration of potassium ions leads to a displacement of sodium ions from the substrate and accordingly to an inhibition of the reaction. If the system contains the four cations in concentrations roughly equal to those in the crab-nerve axoplasin, an increase in the sodium concentration as well as a decrease in the potassium concentration will lead to an