Chromium(III)ATP inactivating (Na++ K+)-ATPase supports Na+-Na+ ...

Eur. J. Biochem. 157,585 - 595 (1986) 0FEBS 1986

Chromium(II1)ATP inactivating (Na' + K+)-ATPase supports Na' -Na+ and Rb+-Rb+ exchanges in everted red blood cells but not Na+,Kf transport Hartmut PAULS, Engin Halit SERPERSU, Ulrike KIRCH and Wilhelm SCHONER Institut fur Biochemie und Endokrinologie, Justus-Liebig-Universitat GieDen (Received January 3/March 11, 1986) - EJB 86 0007

+

1. The chromium(II1) complex of ATP, an MgATP complex analogue, inactivates (Na' K+)-ATPase by forming a stable chromo-phosphointermediate. The rate constant k 2 of inactivation at 37 OC of the &y-bidentate = 1.08 mM), imidazole = 15 mM) and Mg2+ = 0.7 mM). These of CrATP is enhanced by Na' cations did not affect the dissociation constant of the enzyme-chromium-ATP complex. 2. The inactive chromophosphoenzyme is reactivated slowly by high concentrations of Na+ at 37°C. The half-maximal effect on the reactivation was reached at 40 mM NaCl, when the maximally observable reactivation was studied. However, 126 mM NaCl was necessary to see the half-maximal effect on the apparent reactivation velocity constant. K + ions hindered the reactivation with a Ki of 70 pM. 3. Formation of the chromophosphoenzyme led to a reduction of the Rb' binding sites and of the capacity to occlude Rb'. 4. The lJ,y-bidentate of chromium(II1)ATP (Kd = 8 pM) had a higher affinity than the a,lJ,y-tridentate of chromium(II1)ATP (Kd = 44 pM) or the cobalt tetramine complex of ATP (Kd = 500 pM). The P,y-bidentate of the chromium(II1) complex of adenosine 5'-[PJ-methyleneltriphosphate also inactivated (Na' + K+)ATPase. 5. Although CrATP could not support Na+,K+ exchange in everted vesicles prepared from human red blood cells, it supported the Na+-Na+ and Rb+-Rb+ exchange. 6. It is concluded that CrATP opens up Na' and K + channels by forming a relatively stable modified enzymeCrATP complex. This stable complex is also formed in the presence of the chromium complex of adenosine 5'[B,y-methyleneltriphosphate. Because the P,y-bidentate of chromium ATP is recognized better than the a,lJ,ytridentate, it is concluded that the triphosphate site recognizes MgATP with a straight polyphosphate chain and that the Mg2' resides between the P- and the y-phosphorus. The enhancement of inactivation by Mg2+ and Na' may be caused by conformational changes at the triphosphate site. Incubation of (Na' + K+)-ATPase with the MgATP complex analogue chromium(II1)ATP leads to the inactivation of the enzyme by the formation of a stable phosphointermediate [l]. This inhibition occurs in the absence of Na+ and Mg", which otherwise are absolutely necessary to build up the two differently reactive forms of phosphointermediate from ATP [2 - 51. Such two forms of phosphointermediate, which are passed through during active Na+ and K + transport [2 - 51, have also been found to exist during the active Ca2 transport catalyzed by the CaZ+-ATPaseof sarcoplasmic reticulum [6]. When the latter enzyme is incubated with CrATP, Ca2+ is occluded in a stable form and the enzyme is phosphorylated and inactivated [7]. Since occlusion of Na' and Rb' ions in (Na' + K+)-ATPase has been reported [3, 8,9], it seemed possible that Na' ions might also be occluded in the CrATP-inactivated enzyme in a more stable way. +

Correspondence to W. Schoner, Institut fur Biochemie und Endokrinologie, Justus-Liebig-Universitat GieDen, Frankfurter StraBe 100, D-6300 GieBen, Federal Republic of Germany Abbreviations. AdoPPP, ATP, adenosine 5'-triphosphate; AdoPP[CH2]P, adenosine 5'-[B,y-methyleneltriphosphate; CrATP, CrAdoPPP, chromium(II1) complex of adenosine 5'-triphosphate; Co(NH&ATP, Co(NH3)4AdoPPP, cobalt(I1I)tetraamine complex of ATP. Enzyme. (Na' K+)-ATPase (EC 3.6.1.37).

+

Using the relatively stable MgATP complex analogue CrATP, it should also be possible to resolve some of the uncertainties concerning the functions of Mg2 and MgATP in the catalytic and ion transport process: MgATP is considered to be the true substrate of the enzyme in the overall reaction [lo- 121, although ATP binding occurs in the absence of Mg2+ [13,14] and although Mn2+ binding to a single highaffinity binding site should be sufficient for full activity [15]. Yet, the additional presence of Mg2+ lowers the dissociation constant of the enzyme-ATP complex [16, 171 and the ATP/ ADP exchange reaction [18]. It is unclear so far whether this effect occurs by the formation of an MgATP coinplex or due to the alteration of the conformation of the ATP binding site. The latter assumption is supported by a considerable number of studies with protein-reactive ATP analogues [19 - 221. They show that despite the binding of the analogues to the highaffinity ATP binding site, photoaffinity labelling with azidoATP derivatives is not possible except in the presence of Mg2+ [21, 221. MgATP has also been reported to inhibit the Na+-dependent ATP/ADP exchange activity, which reflects the Na+Na' exchange, at concentrations of 0.4- 1 mM [lo]. No report has been published so far on whether the Na+-Na+ exchange is affected by MgATP. The Na'-Na' exchange either needs the hydrolysis of ATP [23,24] or the simultaneous presence of ATP + ADP [25]. Nucleoside triphosphates which +

586

cannot phosphorylate the enzyme are unable to support this exchange activity [25] but they support the K+-K+ exchange [26]. Due to the availability of a stable MgATP complex analogue, these partial reactions of the sodium pump may also be studied. For all of these reasons, we continued to explore the effects of univalent and divalent cations on the inactivation of (Na' K +)ATPaseand on the stability of the phosphoprotein formed. We furthermore studied the effects of CrATP on the occlusion of univalent cations into the enzyme and tested its action on the sodium pump and its partial transport activities. Some of these results have been included in a review [27].

+

tion step described above. This procedure was repeated once. Finally the sediment was homogenized in the case of the NaCl reactivation experiments in 10 mM imidazole/HCl (pH 7.25) and in the case of the "Rb+-binding experiments in 25 mM imidazole/HCl (pH 7.5), 1 mM EDTA. To avoid any difficulties in the interpretation of the 86Rb+-binding experiments by enzyme-bound ATP in the control enzyme, it was incubated prior to the binding assay for 6 h at 37°C in 100 mM NaCI, 3 mM MgClz and 25 mM imidazole/HCl (pH 7.25) to hydrolyze ATP. The enzyme was then washed as described above. [86Rb]Rubidium binding

a ) Equilibrium binding. In analogy to [34], 0.5 mg of untreated or CrATP-inactivated enzyme were incubated in Chemicals polypropylene tubes at 0 "C for 5 min with 50 mM imidazolel Chromium(II1) perchlorate was from Pierce Inorganics, HCl (pH 7.25), 1 mM EDTA with or without 0.1 mM ouabain Rotterdam. All radioactive substances were from Amersham in a total volume of 1 ml containing 2-300 pM "Rb' Buchler (Braunschweig). All other chemicals were of analyti- (0.01 -0.65 pCi). After centrifugation at I00000 x g for cal grade and obtained through E. Merck (Darmstadt), 20 min and careful removal of any remaining incubation medium from the walls of the tubes, the sediment was dissolved Boehringer Mannheim (Mannheim) or Serva (Heidelberg). with 0.5 ml 1 M NaOH, neutralized with a few drops of concentrated HC1 and counted as described earlier [l]. All Preparation of MgATP complex analogues determinations were done in triplicate; the ouabain-sensitive The a,B,y-tridentate of chromium(II1)ATP was prepared Rb+-binding was plotted according to Scatchard [35]. b ) Occlussion of 8 6 R b + 0.5. 1 mg of untreated or CrATPaccording to DePamphilis and Cleland [28] with minor inactivated enzyme were preincubated at 0 "C with 25 mM modifications [l], the P,y-bidentate of CrATP and its AdoPP[CH2]P derivative according to Dunaway-Mariano Tris/HCl (pH 7.4), 15 mM histidine and 1 mM EDTA in the and Cleland [29] and the cobalt(II1) tetramine complex of presence or absence of 2 mM ATP. After addition of 100 pM ATP according to [30]. "CrATP was prepared in the same 86RbC1(0.6 pCi) and mixing, all samples were incubated for manner with the lyophilized 'lCrC13 HCl from Amersham- 5 min at 0°C in a total volume of 0.5 ml; 400 p1 of the incubation Buchler, Braunschweig. The a,B,y-tridentate of [ Y - ~ ~ P I C ~ A T P mixture was then passed through precooled columns of 0.5 ml Dowex 50 WX8,lOO- 200 mesh, at 0°C. The columns was prepared as described previously [l]. had been prepared as described by Beaugi and Glynn [36] except that Dowex 50 WX8 resin was used instead of BioRad Enzyme and assays BioRex 70. 75% of the enzyme protein was found in the (Na' K+)-ATPase from pig or sheep kidney with a effluent, and the contact time of the enzyme with the resin specific activity of 6 - 15 U/mg was prepared according to was 2 s. The effluents were collected in vials and counted in a Jsrgensen [31]; 1 unit (Na' + K+)-ATPase is defined as the scintillation solution containing Triton X-100 [13]. Occluded amount of enzyme catalyzing the hydrolysis of 1 p o l ATP/ rubidium is defined as the differencebetween the samples with min at 37°C. The quantification of the enzyme has been and without 2 mM ATP. All determinations were done in described earlier [32]like the binding of [c~-~'P]C~ATP and the triplicate. phosphorylation of the enzyme with [y-32P]CrATP[I]. Protein Incorporation of [ 51 Crlchromium was determined according to Lowry et al. [33]. ,from "CrATP into the enzyme

MATERIALS AND METHODS

-

+

+

Inactivation of (Na' Kf )-ATPase by the a,j,y-tridentate of CrATP and other MgATP complex analogues

0.2-0.4 mg enzyme protein was incubated at 37°C in a total volume of 1 ml with 100 mM imidazole/HCl (pH 7.25) and 0.1 mM CrATP for 2-6 h. In some experiments the inactivation medium contained additional substances and a lower buffer concentration. The activity was measured in 4O-pl aliquots with the coupled optical assay [32].

5.5 mg enzyme protein was incubated with 50 mM imidazole/HCl (pH 7.25) and 0.1 mM 51CrATP in a total volume of 5 ml at 37°C. At the intervals given in Fig.4 duplicate 100yl samples were withdrawn and added to 4 ml of an ice-cold mixture of 5% trichloroacetic acid containing 5% H3P04. The samples were centrifuged at 4°C at 1OOOOOxg for 20min. The sediment was washed with the same mixture, solubilized and counted as described earlier [I]. Since inactivation of (Na' + K+)-ATPase by CrATP is hindered by an excess of ATP [l], unspecific binding of 'Cr3 was measured in control experiments in the presence of 10 mM ATP (Na+-free). The values communicated in Fig.4 represent the ATP-sensitiveincorporation of radioactive chromium. +

Preparation of CrATP-inactivated (Na' (chromophosphoenzyme)

f

Kf )-ATPase

Multiples of the above reaction mixture were diluted at the end of the inactivation time with 5 vol. ice-cold distilled water. The enzyme was spun down at 100000 x g for 20 min and the sediment was taken up in 25mM imidazole/HCl (pH 7.25) containing 2 mM EDTA. Thereafter the enzyme was incubated for 30 min at 37 "C followed by the centrifuga-

Reactivation experiments Reactivation of (Na' CrATP-inactivated (Na'

+ K+)-ATPase activity. 0.4 mg of

+ K')-ATPase

(chromophospho-

587 enzyme) was incubated at 37 "C with 50 mM imidazole/HCl (pH 7.25) and various concentrations of NaCl or MgClz in a total volume of 1 ml. At the time indicated in the Figs 5 and 6, the recovery of (Na' + K+)-ATPase activity was determined by including 5O-pl aliquots of the incubation mixture into the optical assay. Maximal reactivation (equilibrium conditions) was usually obtained after 5 -6 h. Evaluation of the dissociation constants of the substrate complexes of Na' and M g 2 + with the inactive enzyme from their effects on the reactivation. The kinetics of the reactivation of CrATP-inactivated (Na' + K+)-ATPase in the presence of Na', K t or Mgz+ were fitted to the equation A, = A , - (Ae - Ao)e-k'. In this equation is A . the activity of the enzyme at the start of the reactivation, A, the activity at t min and A , the maximal activity which can be obtained under equilibrium conditions. The calculations were carried out with the computer program 'non-linear regression BMDP 3R' (Health Sciences Computing Facility, University of California, Los Angeles). A , and the velocity constant k was used to calculate the affinity of Na'. We thank Dr Klaus Failing from the Biomathematics Group of the Institute of Veterinary Physiology for his help in setting up the program. Liberation of 32Pifrom the CrATP-inactivated enzyme. 1.5 mg of [y-32P]CrATP-inactivated enzyme protein was incubated as described above in a total volume of 5.5 ml. At the times indicated (Fig. 6A) 50-pl aliquots were withdrawn and included into the optical assay and simultaneously a 100-pl aliquot was used for the determination of the released radioactive phosphate. That 100-p1sample was added to 2 ml of an ice-cold mixture of 5% trichloroacetic acid containing 5% H3P04 and 0.5 ml 5% ammonium molybdate in 2 M HzSO4. Protein was spun down by centrifugation; the sediments were washed once with the same mixture. The pellets were solubilized and counted as described earlier [l] to determine the amount of the radioactive phosphoenzyme. The liberated 32Piwas determined in the butanol/isobutanol extract of the acid supernatant [37]. All determinations were done in duplicate. Liberation of 51Cr3+. 7.5 mg 51CrATP-inactivated enzyme protein was incubated with 50 mM imidazole/HCl (pH 7.25) and 500 mM NaCl in a total volume of 3 ml. At the times indicated in Fig. 6B, duplicate 100-pl aliquots were withdrawn and added to 1 ml ice-cold 5% trichloroacetic acid containing 5% H3P04. After centrifugation and precipitation of the enzyme protein, 0.8 ml of the supernatant was counted in a liquid scintillation counter. All values were corrected against blanks without NaCl. Experiments with everted vesicles prepared from red blood cell vesicles

Everted vesicles from human red blood cells were prepared as described by Steck et al. [38, 391 with the following modifications: Tris/HCl (pH 8.0) was substituted for the sodium phosphate buffer and 0.2 mM ouabain was included into one portion of the membranes at the vesiculation step. Vesiculation was either performed in 0.5mM Tris/HCl buffer (pH KO), followed by the equilibration of the everted isolated red blood in a medium containing the desired conditions intravesicularly, or it was carried out in the medium intended for the interior of the vesicles. After vesiculation, achieved by forcing the membranes through a thin needle, 1 vol. of the preparation was placed on top of 1 vol. Dextran T70 (1.03 g/ ml) and centrifuged at 4°C in a SW 27 rotor of the Beckman Spinco ultracentrifuge at 24000 rpm overnight. Inside-out

vesicles containing univalent cations were washed and then stored maximally for 1 week at 4°C. Empty vesicles (500 pg protein/ml) were loaded by incubating them overnight at 4 "C and thereafter additionally for 30 min at 37°C in an equal volume of an appropriate buffer including the necessary ions. The vesicles were washed thereafter twice and were used immediately for the cation flux experiments. Osmolarity was maintained by using choline chloride as a neutral ion. Cation flux experiments were carried out at 37°C. The uptake of "Na' or "Rb' was measured by diluting an aliquot of 100 pl (50 mg protein) of the incubation medium 40-fold in non-radioactive ice-cold incubation medium lacking the nucleotides. After centrifugation of the vesicles for 10 min at 2000 rpm in a rotor 50 Ti and washing them once with the same medium, the sediments were solubilized in 0.5ml 1 M NaOH and counted in a scintillation counter as described previously [l].The results are mean values of three or four experiments. N a + - K + exchange experiments. Empty vesicles were equilibrated overnight in an equal volume of 25 mM Tris/HCl (pH 7.4), 20 mM NaCI, 4 mM KC1, 1 mM MgC12 at 4°C followed by an incubation period of 30min at 37°C. After washing twice with 25 mM Tris/HCl (pH 7.4) containing 5 mM NaC1, 1 mM MgC12, 20 mM choline chloride, the everted red blood cells were immediately incubated at 37 "C in 25 mM Tris/HCl (pH 7.4), 5 mM 22NaC1 (5 x lo5 cpm . pmol-I), 1 mM MgClz, 0.5 mM ATP or 0.12 mM CrATP and the necessary amounts of choline chloride to hold the osmolarity. They were further handled as described above. Na+-Na+ exchange conditions. Vesiculation was performed in 25 mM Tris/HCl (pH 7.4), 100 mM NaC1, 1 mM MgC12, 10 mM choline chloride; controls contained additionally 0.2 mM ouabain. After washing of the vesicles twice in 25 mM Tris/HCl (pH 7.4), 5 mM NaCl and 1 mM MgC12 the vesicles were incubated in the same medium containing5 mMZ2NaC1(4xlo5 cpm.pmol-'and0.12 mM CrATP or 0.5 mM ATP 0.2 mM ADP. Rh+ -Rb+ exchange conditions. Empty vesicles were equilibrated as described above with an equal volume of 280 mM Tris/HCl (pH 7.4) and 20 mM RbCl to give a final internal concentration of 10 mM RbCl and 140 mM Tris/HCl. After two washes in 140 mM Tris-HC1 (pH 7.4), 10 mM RbCl and 5 mM MgClz the vesicles were incubated at 37°C in 140 mM Tris/HCl, 10mM 86RbC1 (5x105 cpm . pmol-'), 5 m M MgCIz, 5 mM phosphate (neutralized with Tris base), 3 mM ATP or 0.12 mM CrATP. A control contained either 0.2 mM ouabain within the vesicles or 18 pM vanadate in the incubation medium. The stopping medium consisted of ice-cold 140 mM Tris/HCl (pH 7.4), 10 mM RbCl and 5 mM MgC12.

+

RESULTS Evaluation of the affinity of different MgATP complex analogues,for (Na' + K + )-ATPase

Several MgATP complex analogues have been described so far, which differ in the nature of the metal used as magnesium substituent [28- 301 as well in the kind of the complex formed [28 - 301. Since chromium(I1I)ATP has been reported previously to bind to the high-afinity ATP binding site of (Na' K+)-ATPase and to inactivate it [l], we were interested to see whether the nature of the complex might be of importance for the inhibitory action. Table 1 shows that any kind of

+

588 metal-ATP complex may inactivate the enzyme. Interestingly, interested to see how these ions and parameters affect the the inactivation rate constants k2, which are obtained at inactivation by CrATP. Fig. 1 shows that low concentrations saturating concentrations of the MgATP complex analogues, of Na+ enhance at high ionic strength the inactivation of the are in the same range for all ATP analogues, whether or not enzyme by the cl,P,y-tridentate of CrATP, whilst high Na' the terminal phosphate can be split off. However, the nature concentrations protect against the inactivation [I]. The enof the metal substituent and the kind of the complex formed hancing effect of Na+ at high ionic strength is due to the severely affects the recognition of the MgATP complex increase of the inactivation velocity constant with the a&analogue by the ATP binding site. The p,y-bidentate of CrATP tridentate as well with the F,y-bidentate complexes of CrATP showing a straight polyphosphate chain is recognized better (Fig. 1B, Table 2). This is demonstrated in a more careful than the curved a,&y-tridentate of CrATP. Moreover, the study for the B,y-bidentate of CrATP also at low ionic strength bridge oxygen between the P- and the y-phosphate group is of (Fig. 2). 1 mM NaCl was necessary to reach the half-maximal importance [14,40]. The bulky substituent in C O ( N H ~ ) ~ A T Peffect (Fig. 3 B). Also imidazole, which mimics the effects of apparently hinders the access to the ATP binding site. Na' [16, 411, could increase the rate constant k2 of inactivation with an apparent dissociation constant of 15 mM (Fig. 3A). Mg2+ enhanced the rate constant of k2 at Effects of monovalent and divalent cations on the inactivation saturating concentrations of the p,y-bidentate of CrATP as of (Nu' K' )-ATPase by CrATP well. The half-maximal effect was seen at 0.71 mM (Fig. 3 C). An unexpected finding of our previous inactivation studies In all these studies on the effects of ions on the inactivation with CrATP was that this process did not need MgZf and of (Na' + K')-ATPase by the P,y-bidentate of CrATP no Na'; but K' ions and high concentrations of Na' ions effect of Na+, imidazole or Mg2+on the dissociation constant protected the enzyme against inactivation [l]. Because ATP of the enzyme-substrate complex with CrATP was seen binding has been shown to be influenced by the presence of (Fig.2). From Table2 it is evident that Mg2' ions affect Na' and Mg2' as well by the ionic strength [16], we were the inactivation with both MgATP complex analogues in a different way : whilst the dissociation constant of the enzyme complex with the 6,y-bidentate of CrATP remains unchanged Table 1 . Comparison of the dissociation constants and inactivation in the presence of Mg2+,it is decreased when the unfavourable a,S,y-tridentate is the substrate (Table 2). We also saw that velocity constants of various MgATP analogues for (Na' + K')ATPase low concentrations of Na+ (10 mM) at high ionic strength, All inactivations were performed at 150 mM imidazole/HCl (pH 7.25) like K +, increase the dissociation constant of the favourable at 37°C P,y-bidentate of CrATP (which effect is counteracted by Mg"): but 10 mM NaCl has no effect on the dissociation Analogue Kd k2 constant of the less favourable a,B,y-tridentate of CrATP. Na' and Mg2'increase in any case the apparent inactivation CIM minvelocity constant k2 (Table 2, Figs 1-3). In contrast to the 8 0.039 P,y-Bidentate of CrAdoPPP findings in equilibrium binding experiments with radioactive a,/?,y-Tridentate of CrAdoPPP 44 0.036 AdoPP[NH]P [16], the increase of the ionic strength in the 160 0.061 b,y-Bidentate of CrAdoPP[CH2]P range of 10-70 mM had no effect on the dissociation P,y-Bidentate of Co(NH,),AdoPPP 500 0.016 constant of the p,y-bidentate of CrATP.

+

-s I

0.1

t

" Y "

\\

I

I

5

mM CrATP

m M ATP

eol-

I

control

+ 1 0 rnM No*

+ 3 m M M+

-

i

t

+ I

I

20

I

8 rnM No*

I

40

I

rnln

I

75

50

25

25

50

75 'ICrATP

+

100 [mM-']

Fig. 1. Eflects of Na', K', and M g 2 + on the inactivation of (Nu' K')-ATPase by the a,P,y-tridentate of CrATP. (A) Time course of the inactivation in the presence of 0.1 mM CrATP. 0.2 mg enzyme protein was incubated at 37°C with 50 mM imidazole (pH 7.25) and the additions given. (B) Study of the effect of different concentrations of CrATP and additions on the inactivation rale constants

589 Table 2. Effects of N a + , K+ and M g 2 + on the dissociation constants of the enzyme CrATP complexes and the inactivation velocity constants k2 4-5 units (Na' + K+)-ATPase from pig kidney were incubated at 37°C with 100 mM imidazole buffer pH 7.2 and the rate of inactivation was followed a,P,y-Tridentate of CrATP

P,y-Bidentate of CrATP

Additions

dissociation constant Ka

inactivation velocity constant k2

~

change in

PM

min-'

-fold

8.0 20.0 240.0 25.0 7.5 7.1 25.0

0.0391 0.0714 0.0300 0.0769 0.1250 0.1724 0.1180

1.00 2.50 30.00 3.12 0.93 0.88 3.12

10.0

0.1670

1.25

kd

change in

Kd

inactivation velocity constant k2

PM

min-'

-fold

1.00 1.83 0.77 1.97 3.20 4.40 3.10

44.7 37.9 229.4 39.9 11.9 13.0 14.6

0.0365 0.0597 0.0405 0.0417 0.0587 0.1022 0.0593

1.00 0.85 5.13 0.89 0.26 0.29 0.32

1.00 1.63 1.11 1.31 1.60 2.80 1.62

4.27

27.0

0.0849

0.60

2.32

_

_

k2

dissociation _ constant

kd

kz

~

None 10 mM Na+ 1 mM K + 10 mM N a + + 1 mM K + 3 mM Mgz+ 10 mM Na+ + 3 mM Mg2' 1 mM K f + 3 mM Mg2+ 10 mM Na+ + 1 mM K + + 3 mM Mg2+

Fig. 5 demonstrates that high concentrations of N a + reactivate the CrATP-inactivated (Na' K )-ATPase slowly and that K + ions counteract this effect. The reactivation is caused by the hydrolysis of the chromophosphointermediate, as is evident from Fig. 6 showing the release of 51Cr3+together with inorganic phosphate (not shown) from the [51Cr]~hromo-phosphointermediate and the decrease of the chromo[32P]phosphointermediate during the Na+-dependent reactivation. When the effects of the Na' concentration on the reactivation velocity and the maximal reactivation were investigated, it became evident that, depending on the parameter studied, different concentrations of Na' were necessary to achieve half-maximal effects (Fig.7): 126 mM Na' was necessary to reach the half-maximal effect on the reactivation velocity constant (Fig. 7A), whilst 40 mM Na' already led to the half-maximal reactivation under equilibrium conditions (Fig.7B). Since K + ions inhibit the reactivation process, we were interested to determine the affinity of K + in this process. A Dixon plot of the K + effect at different Na' concentrations shows that K + is bound with an inhibition constant of 70 pM (Fig. 8). Mgz+ ions could replace N a + in the reactivation of the inactive enzyme (Fig. 9). The half-maximal effect of Mg2 + in the reactivation is seen at 67 mM MgC12. The puzzling situation that K f ions, contrary to the phosphointermediate formed from ATP, did not enhance the hydrolysis of the chromophosphointermediate (but stabilized it, Fig. 5) cannot be explained by the fact that the enzyme does not recognize K + . It is possible, however, that the stimulation of hydrolysis needs the occlusion of K f and therefore the conformational change from an Na+-EI form to a K f occluding E2 form as part of the transport cycle. If this were the case, the occlusion of 86Rb+,a congener of K + , should be blocked. As is shown in Fig. 10, this is the case, indeed. Moreover, the number of Rb+ binding sites decreases in the CrATP-inactivated enzyme (Fig. 11). We were unable to detect the occlusion of Na+ into the CrATP-inactivated enzyme with the experimental set up given by Glynn et al. [8].

+

no NaCI o/

VB,y bidentate of CrATP [mM-1]

Fig. 2. Study on the influence of the Na' concentration on the apparent inactivation rate constant of (Nu+ K+)-ATPasefrom sheep kidney as a function of the concentration of the P,y-bidentate of CrATP. Incubation of the enzyme at 37°C was carried out at 10 mM imidazole/ HC1 (pH 7.25)

+

The transfer of chromium f r o m CrATP to the enzyme stabilizes the phosphointermediate It was suggested previously that the transfer of chromium might stabilize the phosphointermediate and lead thereby to the enzyme's inactivation. This is demonstrated now in Fig. 4. Using the a,j,y-tridentate of 51CrATP,we saw an incorporation of maximally 143 pmol radioactive chromium/unit enzyme together with the inactivation. This value corresponds favourably with the capacity of the phosphorylation site of 121 18 pmol/unit enzyme [l]. Studies on the cation sensitivity of the chromium phosphoenzyme

K + ions increase the hydrolysis of the phosphointermediate formed from ATP but do not affect the hydrolysis of the chromophosphoenzyme [l]. Hydrolysis of the phosphointermediate may also take place in the presence of N a + [42]. We were therefore interested to see whether Na' ions might affect the hydrolysis of the chromophosphointermediate.

+

Studies on the role of CrATP as a substrate o f t h e sodium p u m p and its partial reactions

The nucleotide requirements of the sodium pump and of its partial reactions of the Na+-Na+exchange and of the K + -

590

25

50

10

2

100

20

1

3

2

m M imidazole-HCI (pH 725) mM NaCI mM MgCll Fig.3. Study of the effects of N a + , Mg2+ and imidazole on the inactivation rate constant k2 upon incubation with the P,y-bidentate of CrATP.

(A) Effect of the concentration of imidazole in the presence and the absence of 10 mM NaCl and 3 mM MgClz on the inactivation rate constant k2. (B) Effect of NaCl on the inactivation rate constant k2 at 10 mM imidazole/HCl (pH 7.25). (C) Effect of MgCI2 on the inactivation rate constant k2 at 10 mM imidazole/HCl (pH 7.25)

A 80

-

-

/-a--

800

-

P

E

2 0 Inv)

-'

0.00 mM

20

51CrATP

60

60

120

0.2

min

nmolrs %3+

0.4

0.6

0.8

1.0

incorpcrotcd Img protein

Fig.4. Comparison of the inactivation of (Na+ + K+)-ATPase activity and the incorporation ofs1@+ into . the enzyme protein upon incubation with 0.08 mM 5'CrATP. (A) Time course of the inactivation and the incorporation of radioactive chromium. (B) Titration of the number of sites incorporating radioactive chromium with the inactivation

100 mM NaCI

I 0 0 rnY NoCl 0.6 mM KCl

control

I

2

3

4

5

6

hours

Fig. 5 . Time course of' the reactivation of CrATP-inactivated (Na' K + )-ATPase at various NaCl concentrations

+

K + exchange differ [4,25,26]. Since CrATP is relatively stable and inactivates (Na' + K+)-ATPase slowly, it should not supported the 22Na+ uptake into K+-containing everted vesicles from human red blood cells prepared according to Steck et al. [38,39]. This is what we find (Fig. 12A). However, CrATP supported an ouabain-sensitive 22Na+ uptake into N a t -containing vesicles with a mean rate of 2.65 nmol . mg-' . h-' with a standard error of the mean value of 1.16 nmol mg-' . h-' in five independent experiments. This exchange rate is equal in magnitude to the Nat-Na+ exchange rate in the presence of ADP + ATP of 2.52 k 1.16 nmol . mg-' . h-' (Fig. 12B). Consistent with the former report of Simons [26] that ATP analogues with a blocked phosphate group support as a cofactor the K + - K + exchange reaction, CrATP also supported the ouabain-sensitive "Rb+ uptake into Rb+containing everted vesicles from red blood cells (Fig. 12C).

59 1

-c .g

i'l

0.025

Y

e

3

: 1

,

,

,

,

,

40

80

120

160

200

~

c 0

I, =

0.020

-

0.015

-o

orno

2

e

$

: 125.9

0.005

rnM

280 min

NaCl (mMI

Fig.1. Evaluation of the apparent affinity of N a + from the reactivation of CrATP-inactivated (Nu' K')-ATPase. (A) Plot of the reactiva30 60 90 120 150 180 min tion velocity constant versus the NaCl concentration. (B) Plot of the K+)-ATPase reached at a certain Fig. 6. Fate of enzyme-boundphosphate and chromium during the NaCl- maximal reactivation of (Na+ concentration of NaCl (quilibrium) versus the molarity in the reactivainduced reactivation of CrATP-inactivated (Na' f K')-ATPme. (A) Time course of the disappearance of the stable ~ h r o m o [ ~ ~ P ] p h o stion - assay. A is the reactivation found in the presence of NaCl under phointermediate during the reactivation with 100 mM NaCl. (B) Time equilibrium conditions; A,, is the activity of the CrATP-inactivated course of the liberation of 'lCr3+ during the hydrolysis of the enzyme prior to the reactivation; A,, is the enzyme activity after the maximal reactivation with NaCI. The bars in the figure represent the [ 5 lCr]chromophosphointermediateduring the reactivation in the mean values f SD of eight experiments presence of 500 mM NaCl

Y

I

+

+

DISCUSSION

+

Inactivation of (Na' K+)-ATPase by CrATP occurs by the transfer of chromium together with the phosphorylation of the enzyme [l](Fig. 4) and the apparent stabilization of this phosphointermediate. With respect to the transfer of Cr3' as an Mg2+ analogue, CrATP behaves like MgATP [43]. Similarly to MgATP as substrate, the rate of phosphorylation (and inactivation in the case of CrATP) is stimulated by the presence of low concentrations of Na' (Fig. 3 B) and by Mg2+ (Fig. 3 C). In contrast to the Mg-phosphointermediate, however, the hydrolysis of the Cr-phosphointermediate takes place only in the presence of high concentrations of Na+ and is hindered by K' (Figs 5 and 8). CrATP therefore allows only the Na+-ATPase activity to continue, this being a partial reaction of (Na' K+)-ATPase activity [44, 451. In fact, a very slow hydrolysis of CrATP is found in the presence of 10 mM NaCl and 5 mM MgC12 (not shown). Formation of the stable chromophosphointermediate lowers the number of the available Rb+ binding sites (Fig. 11) and hinders the occlusion of Rb' into the enzyme protein (Fig.12). Since occlusion of Rb', a congener of K + , needs the formation of the E2 conformational state [ 2 - 5 , 361, the

+

NoCl

7 0.2

0.4

0.6

250 rnM NaCl

0.8

1.0

rnM KCI

Fig. 8. Dixon plot of the inhibitory effect of K+ on the reactivation of CrATP-inactivated (Na' K')-ATPase

+

formation of the chromophosphointermediate apparently arrests the enzyme in an El-like conformational state. In that state hydrolysis of the chromophosphointermediate and the release of Cr3+ together with the reactivation occurs only, when either high (extracellular?) concentrations of Na'

592

100

I

A-Ao

Arnax- A 0

10

x 100

30

500

100

m M HgCl

Fig. 9. Evaluation of the apparent affinity of Mg2 in the reactivation of CrATP-inuctivated (Na' Ki)-ATPase

-+

+

0.01

A

B

0°%,

120

150 min

0.03

O.OL

0.05

Fig. 11. Ouabuin-sensitive 86Rbi binding to a CrATP-inactivated and to an untrated control enzyme

( K o .= ~ 40 mM and 126 mM respectively, Figs 5 - 7) are present or when high concentrations of Mg2+ replace chromium (Fig. 9). The latter finding is consistent with the demonstration of an enhancement of the hydrolysis of the phosphointermediate formed from ATP by Mg2+ [43], whilst the former finding points to the Na' ATPase activity [42,44, 451. Extracellular sodium stimulates the ATP-ADP exchange by the sodium pump [46] by a partial reversal of the sodium pump cycle (Fig. 13). This phenomenon is, however, unable to explain the enhancing effect of high (extracellular?) sodium concentrations on the hydrolysis of the phosphointermediate, which is clearly not a K+-like effect (Figs 5 and 8). Possibly this modifying effect of Na+ occurs through a shift of the (Na)E1 P/Na+ . E2-P equilibrium towards the right, which is counteracted by K + = 70pM). Since Rb+ binding (Fig.11) and occlusion is not possible with the chromophosphointermediate(Fig. 3 0),this potassium binding site may reside at the intracellular side of the enzyme. Although Cr3+ is transferred to the phosphointermediate from CrATP during the inactivation and although a bound divalent cation in the high-affinity ATP binding site is considered sufficient for the catalytic activity [15], interaction of Mg2+ with a low-affinity binding site = 0.7 mM) additionally enhanced the rate of phosphorylation and inactivation (Fig.3C). It is well known that binding of Mg2+ to this low-affinity site alters the conformation of the highaffinity ATP binding site at the purine [19, 223 and ribosyl subsites [21] and hinders the Na+-dependent ATP/ADP exchange [18]. It is therefore likely that the enhancing effect of Mg2 on the rate of phosphorylation and inactivation occurs by an Mg2+-induced conformational change also at the triphosphate subsite of the ATP binding site. Such a conformational change could (without alteration of the dissociation constant of the enzyme complex with the P,y-bidentate of CrATP, Table 2) bring the aspartyl phosphate acceptor group nearer to the terminal phosphate and thereby enhance the rate of transDhosDhorvlation. An alteration of the ATP bindinn site by Mg" is -also indicated by the finding that Mgzi

-

Fig. 10. Comparison ofthe time course of inactivation of (Na' + K'JATPase by 0.1 m M CrATP with the capability to occlude Rb+. (A) Decrease of (Na+ + K')-ATPase activity (6.5 U/mg in the control enzyme); (B) occlusion of 86Rb+ (588 pmol in the control enzyme); (C) binding of radioactivity from 0.1 mM [ G I - ~ ~ P I C ~(650 A T Ppmol were maximally incorporated). Incorporation of radioactivity from [ E - ~ ~ P I C ~ Ainto T P the enzyme protein was determined by acid precipitation as described earlier [I]. .~The data are given as mean values f SD of 2 - 7 experiments

a02

moles Rb* bwndlrnoles Rb* free

+

593

-

Na*- Na* axc h ~ p g t

I C 0

1

N N

15 30 45 min 15 30 45 min 2 4 6min Fig. 12. Comparison of the effects of ATP and CrATP as energy sources for Nu', K + transport and as cofactors to induce Na'-Na+ and Rb'Rh+ exchanges in everted vesicles prepared from human red blood cells. (A) Study of the uptake of "NaC1 into everted vesicles prepared from human red blood cells containing 2 mM KCl and 10 mM NaCl from a medium containing 5 mM "NaCl and either 0.5 mM ATP or 0.12 mM of the P,y-bidentate of CrATP. Controls contained 18 pM sodium vanadate additionally. (B) Study of the uptake of "Na' into everted vesicles prepared from human red blood cells containing 100 mM NaCI. The incubation medium contained 5 mM "NaC1 and 0.5 mM ATP + 0.2 mM ADP or 0.12 mM of the fi,y-bidentate of CrATP. Controls contained 0.2 mM ouabain additionally within the vesicles. (C) 86Rb' uptake into everted vesicles prepared from human red blood cells containing 10 mM RbCI. The incubation medium contained 10 m M 86RbC1, 5 m M phosphate (Tris) and either 3 mM ATP or 0.12 mM of the j,y-bidentate of CrATP. Experiments with ATP or ATP + ADP (0, W) and with CrATP ( 0 ,0); open symbols are experiments in the presence of ouabain or vanadate

I

1

Na.E,ATP

N ~ J

\ATP

\

/

E,ATP

'\+

No. E,-P

HzO

Nay

NaeE,

E,

Nai

.-/--a

-+-

/NaT

Change of the Mg-Phosphoenzyme Komplex ?

HZO

(KIE,

(K)Ez-P

F1:

5

a

K3-P

u D

+

Fig. 13. A kinetic model of the relationships between Na'. K' transport and the conformational changes of (Na' K+)-ATPase including the possible sites ofaction of Nu' and K ' on the hydrolysis of the chromophosphoenzyme. Modification of the reaction scheme of Karlish et al. [551

lowers the dissociation constant of the less favourable a,B,ytridentate of CrATP whereas that of P,y-bidentate of CrATP is unaltered (Table 2 ) . Since binding of sodium to a highaffinity site = 1.08 mM) and of imidazole = 15 mM) also enhances the rate of phosphorylation and inactivation from the P,y-bidentate of CrATP (Fig. 3), it is likely that these effects too are due to changes of the conformation of the triphosphate subsite of the ATP binding site. The P,y-bidentate of CrATP binds with a higher affinity to the high-affinity ATP binding site of (Na' Kf)-ATPase than the cr,j,y-tridentate (Tables 1 and 2). This implies that

+

Mg2+ should be complexed between the b- and y-phosphate groups of ATP in the MgATP complex, the real substrate of (Na' + K+)-ATPase [lo- 121. Moreover, the increased dissociation constant of the enzyme with the cr,P,y-tridentate of CrATP, which has a curved triphosphate chain as compared to the P,y-bidentate of CrATP, suggests that MgATP is recognized with a straight polyphosphate chain of ATP at the highaffinity ATP binding site [l]. Moreover the subsite recognizing the phosphates of ATP seems to be of limiting space for Mg2+ and its analogues: Co(NH&ATP, which bears a bulky substituent, is bound more weakly than CrATP (Table 1).

All MgATP complex analogues inactivate (Na' + K ')ATPase after the formation of a dissociable enzyme-substrate complex (Table 1). This is astonishing in so far as the Plymethylene derivative of CrATP cannot phosphorylate the enzyme. It is likely therefore that the formation of the enzymesubstrate complex with this CrATP analogue is followed by the formation of a stable MgATP complex by a conformational change with the inactivation velocity constant k2 of 0.061 min-'. Formation of a stable CrATP complex which even survives acid precipitation (Fig. 10 C) has been reported earlier [l]. This behaviour indicates the formation of a tightly bound MgATP (E; . MgATP) due to a conformational change after the formation of a dissociable enzyme-MgATP complex (El . MgATP). Consistent with th~s,a conformation change of the active site after binding of MgATP has been proposed from rapid kinetic measurements [47 -491, proton relaxation studies [50] and from photoinactivation experiments with 8-azido-ATP and its CrATP derivative [22]. Consistent with the demonstration that CrATP forms a tightly bound enzyme-CrATP complex and therefore a stable chromophosphointermediate (Fig. 13), CrATP did not fuel the sodium pump [27, 511 (Fig.12A). But like other ATP analogues with a blocked phosphate group it supported the Rb+-Rb' exchange in everted human red blood cells in the presence of phosphate (Fig. 12C). This nucleoside-triphosphate-induced effect is usually interpreted as the liberation of E2-occluded K + by an ATP-induced conformational change of one subunit of the a,/?-dimer of (Na' K')-ATPase, of which one chain is phosphorylated from phosphate and the other contains the bound ATP [52]. In contrast to former reports communicating that exclusively ATP (in the absence [23, 241 or presence [25, 261 of ADP) supports the Na+-Na+ exchange and not the P,y-methylene derivative of ATP [4,25, 531, CrATP also substituted for ATP in this reaction (Fig. 12B). We were, however, unable to demonstrate an occlusion of "Na' into (Na' + K+)-ATPase in the presence of CrATP. This may indicate that the formation of a tightly bound MgATP complex (or CrATP complex) at the highaftinity ATP binding site opens up the N a + channel of the enzyme in the El conformational state. Since the phosphorylation of the enzyme from the tightly bound CrATP complex is a slow process (Figs 1 and lo), the Na' channel may rest in an open state under these conditions. It thus appears that phosphorylation of the enzyme, which otherwise occurs rapidly from MgATP, leads to the occlusion of Na' [8]. This process appears to be identical with the gating of sodium (Fig.13). Since the interaction of ATP with (Na' K+)ATPase opens up K + and Na+ channels and also switches the enzyme from the E2 conformational state to the El state [2 - 51, the ATP-induced conformational changes of the sodium pump seem to be of the same eminent importance as the ATP-induced conformational changes of myosin ATPase in the mechanism of muscle contraction 1541. Since CrATP arrests the enzyme in an El-like conformational state this or other MgATP analogues could be helpful for studying the interactions of a subunits within the a,/?-dimer. Studies into this direction are in progress.

+

+

The skilful technical assistance of Mrs Barbara Bredenbrocker and of Miss Sabine Kehm during these studies is gratefully acknowledged. This work has been supported by the Deutsche Forschungsgemeinschaft (Bonn-Bad Godesberg) through Sonderforschungsbereich 169 'Struktur und Funktion membranstandiger Proteine' (Frankfurt/Main) and by the Fonds der Chemischen Industrie (Frankfurt/Main).

REFERENCES 1 . Pauls, H., Bredenbrocker, B. & Schoner, W. (1980) Eur. J. Biochem. 109,523 - 533. 2. Schuurmans Stekhoven, F. & Bonting, S. L. (1981) Physiol. Rev. 61, 1-76. 3. Jerrgensen, P. L. (1982) Biochim. Biophys. Acta 694,27-68. 4. Glynn, I. M. & Karlish, S. J. D. (1975) Annu. Rev. Physiol. 37, 13-55. 5. Dahl, J. L. & Hokin, L. E. (1974) Annu. Rev. Biochem. 43, 327356. 6. De Meis, L. & Vianna, A. L. (1979) Annu. Rev. Biochem. 48, 275 - 292. 7. Serpersu, E. H., Kirch, U. & Schoner, W. (1982) Eur. J . Biochem. 122,347 - 354. 8. Glynn, I. M., Richards, D. E. & Hara, Y. (1985) in The sodium pump (Glynn, I. & Ellory, C., eds) pp. 589 - 598, The Company of Biologists Ltd, Cambridge. 9. Esmann, M. & Skou, J. C. (1985) Biochem. Biophys. Res. Commun. 127,857-863. 10. Robinson, J. D. (1976) Biochim. Biophys. Acta 440, 711 -722. 11. Hexum, T., Samson, F. E., Jr & Himes, R. H. (1970) Biochim. Biophys. Acta 212, 322-331. 12. Plesner, L. & Plesner, I. W. (1981) Biochim. Biophys. Acra 643, 449 -462. 13. Norby, J. G. & Jensen, J. (1971) Biochim. Biophys. Acta 233, 104-116. 14. Hegyvary, C. & Post, R. L. (1971) J . Biol. Chem. 246, 52345240. 15. O'Connor, S. E. & Grisham, C. M. (1979) Biochemistry 18,2315 2323. 16. Rempeters, G. & Schoner, W. (1983) Biochim. Biophys. Acta 727, 13-21. 17. Robinson, J. D. (1980)J. Bioenerg. Biomernbr. 12, 165-173. 18. Robinson, J. D. & Flashner, M. S. (1979) in Na,K-ATPase: Structure andkinetics (Skou, J. C. & Norby, J. G., eds) pp. 275 285, Academic Press, London. 19. Patzelt-Wenczler, R. & Mertens, W. (1981) Eur. J . Biochem. 121, 197-202. 20. Hasselberg, M. & Schoner, W. (1984) Hoppe-Seyler'.y Z . Physiol. Chem. 356,999- 1000. 21. Rempeters, G. & Schoner, W. (1981) Eur. J. Biochem. 121, 131 137. 22. Scheiner-Bobis, G. & Schoner, W. (1985) Eur. J . Biochem. 152, 739 - 746. 23. Forgac, M. & Chin, G. (1981) J . B i d . Chem. 256,3645-3646. 24. Cornelius, F . & Skou, J. C. (1985) Biochim. Biophys. Acta 818, 211 -221. 25. Glynn, I. M. & Hoffman, J . F. (1971) J . Physiol. (Lond.) 218, 239 - 256. 26. Simons, T. J. B. (1975) J . Physiol. (Lond.) 244, 731 -739. 27. Schoner, W., Rempeters, G., Bobis, G., Pauls, H., PatzeltWenczler, R., Serpersu, E. H. & Anner, B. (1984) Symp. Biol. Hung. 26,215 - 230. 28. De Pamphilis, M. L. & Cleland, W. W. (1980) Biochemistry 12, 3714- 3724. 29. Dunaway-Mariano, D. & Cleland, W. W. (1980) Biochemistry 19, 1496-1505. 30. Cornelius, R. D., Hart, P. H. & Cleland, W. W. (1977) Inorg. Chem. 16, 2799-2805. 31. Jerrgensen, P. L. (1974) Biochim. Biophys. Acta 356, 36-52. 32. Schoner, W., v. Ilberg, Ch., Kramer, R. & Seubert, W. (1967) Eur. J. Biochem. I , 334-343. 33. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chern. 193,265-275. 34. Jsrgensen, P. L. & Petersen, J. (1982) Biochim. Biophys. Acta 705, 38-47. 35. Scatchard, G. (1949) Annu. N . Y. Acad. Sci. 51,660-672. 36. Beauge, L. A. & Glynn, I. M. (1979) Nature (Lond.) 280, 510512. 37. Schoner, W., Beusch, R. & Kramer, R. (1968) Eur. J . Biochem. 7, 102-110.

595 38. Steck, T. L., Weinstein, R. S. & Wallach, D. F. H. (1970) Science ( Wush. DC) 168,255 - 257. 39. Steck, T. L. & Kalant, J. A. (1974) Methods Enzymol. 31, 172180. 40. Patzelt-Wenczler, R. & Schoner, W. (1981) Eur. J . Biochem. 114, 79 - 87. 41. Schuurmans Stekhoven, F. M. A. H., Swarts, H. G. P., de Pont, J. J. H. H. M. & Bonting, S. L. (1985) Biochim. Biophys. Actu 815, 16-24. 42. Post, R. L., Hegyvary, C. & Kume, S. (1972) J . Biof. Chem. 247, 6530- 6540. 43. Fukushima, Y. & Post, R. L. (1978) J . Biol. Chem. 253, 68536862. 44. Czerwinski, A., Gitelman, H. J. & Welt, L. G. (1967) Am. J. Physiol. 213, 786- 792. 45. Neufeld, A. H. & Levy, H. M. (1969) J . B i d . Chem. 244,64936491. 46. Cavieres, J. D. (1980) J . Physiol. (Lond.) 308, 57 P.

47. Kanazawa, T., Saito, M. & Tonomura, Y. (1970) J. Biochem. (Tokyo) 67,693-711. 48. Tonomura, Y. & Fukushima, Y. (1974) Ann. N. Y . Acud. Sci. 242,92- 105. 49. Fukushima, Y. & Tonomura, Y. (1975) J. Biochem. (Tokyo) 77, 521-531. 50. Grisham, C. M. & Hutton, W. C. (1978) Biochem. Biophys. Res. Commun. 81, 1406-1411. 51. Schoner, W., Pads, H., Sepersu, H. E., Rempeters, G., PatzeltWenczler, R. & Hesselberg, M. (1983) Curr. Top. Membr. Trump. 19, 361 - 366. 52. Sachs, J. R. (1981) J. Physiol. (Lond.) 316, 263-277. 53. Glynn, I. M. & Hoffman, J . F. (1969) J. Physiol. (Lond.) 207, 34 P-35 P. 54. Tada, M., Yamamoto, T. & Tonomura, Y. (1978) Physiol. Rev. 58, 1-79. 5 5 . Karlish, S. J. D., Yates, D. W. & Glynn, I. M. (1978) Biochim. Biophys. Actu 525, 252-264.