ceptors (Cervetto, L., Lagnado, L., Perry, R. J., Robin- son, D. W., and ..... (Dixon and Webb, 1964), and one example of such analysis is illustrated in Fig. 5.
Vol. 266, No. 1, Issue of January 5, pp. 189-197, 1991 Printed in U.S.A.
THEJOURNAL OF BIOLOGICAL CHEMISTRY
0 1991 by The American Society for Biochemistry and Molecular Biology, Inc.
Effect of Potassium Ionsand Membrane Potentialon the Na-Ca-K Exchanger in Isolated Intact Bovine Rod Outer Segments* (Received for publication, May 30, 1990)
Paul P. M. Schnetkamp andRobert T. Szerencsei From the Department of Medical Biochemistry, University of Calgary, Health Sciences Centre, 3330 Hospital Drive N. W., Calgary, Alberta T2N 4N1,Canada
1988; Matthews et al., 1988; Nakatani and Yau, 1988; Nicol and Bownds, 1989). A proper maintenance of internal free Ca2+concentration by Na-Ca exchange is of crucial importance for a proper function of rods, since a slight imbalance of the precarious equilibrium between Ca2+influx and Caz+ efflux, for example by elevating extracellular C, will cause the dark current to disappear within a few seconds (Hodgkin et al., 1985). Three factors make isolated ROS an ideal system to study Na-Ca exchange. First, Na-Ca exchange appears to be the only functional ion transporter when the cGMP/light-sensitive channels are closed by light (Yau and Nakatani, 1984; Lagnado et al., 1988; Schnetkamp et al., 1989). Second, the Na-Ca exchange activity in ROS is high, one of the highest among biological systems (Kaupp and Schnetkamp,1982); at maximal capacity Na-Ca exchange can change total intracellular Ca2+by 0.5 mM/s in bovine ROS (Schnetkamp, 1986), and by 0.05 mM/s in frog ROS (Schnetkamp and Bownds, 1987). Third, isolated bovine ROS contain a large Ca2+buffer capacity due to negatively charged residues on the surface of the intracellular disc membranes; a likely candidate is the high concentration (30 mM)of phosphatidylserine in the internal disc membranes (Schnetkamp, 1979, 1985a). Ca2+enriched bovine ROS can be prepared that contain 15-25 mM total Ca2+at an internal free Ca2+ concentration in the low micromolar range (Schnetkampand Kaupp, 1985; Schnetkamp, 1985a). The large intracellular Ca2+ buffer capacity enables us to observe linear kinetics of Na-Ca exchange fluxes at high resolution and over reasonable time intervals in contrast tovesicle systems, where the internalion concentrations change much more rapidly due to the small internal volume The outer segments of vertebrate rod photoreceptor cells (Reeves, 1985). The advantage of ROS for the study of Na(ROS)’ exhibit remarkably dynamic Ca2+fluxes; in the dark Ca exchange has been illustrated by measurements of Na-Ca a large Ca2+flux (1-2 PA) enters the outer segment via the exchange currents inROS (Hodgkin et al., 1987; Hodgkinand light-sensitive conductance, which is balanced by Ca2+efflux Nunn, 1987), by measurements of Na+-induced Ca” fluxes via Na-Ca exchange (Yau and Nakatani, 1984; Yauand Nak- measured with 45Ca(Schnetkamp, 1980) or with optical dyes atani, 1985). In thelight, the influx of Ca2+ceases and Na-Ca (Schnetkamp, 1986; Schnetkamp and Bownds, 1987; Schnetexchange is responsible for a rapid decrease in the intracel- kamp et al., 1989), and by purification of the exchanger from lular Ca2+ concentration (McNaughton et al., 1986; Ratto et bovine ROS (Cook and Kaupp, 1988; Nicoll and Applebury, al., 1988) and increase in extracellular Ca2+(Gold, 1986;Miller 1989). and Korenbrot, 1987);changes in intracellular Ca2+appear at External potassium ions have complex effects (both inhibleast in part to mediate light adaptation (Koch and Stryer, itory andstimulatory) on Na-Caand Ca-Ca exchange in bovine ROS (Schnetkamp, 1980, 1986), while in amphibian * This research was supported by grants from the Alberta Heritage Foundation for Medical Research and theMedical Research Council ROS potassium ions inhibit forward Na-Ca exchange (Hodgof Canada. The costs of publication of this article were defrayed in kin et al., 1987; Schnetkamp and Bownds, 1987). Recently, part by the payment of page charges. This article must therefore be two studies reported that Na-Ca exchange in both amphibian hereby marked “aduertisement” in accordance with 18U.S.C. Section and bovine ROS requires and transports K+ and a stoichi1734 solely to indicate this fact. ometry of 4Na/(lCa+lK) was suggested (Cervetto et al., 1989; The abbreviations used are: ROS, rod outer segments; Hepes, 4- Schnetkamp et al., 1988,1989). In thisstudy we have analyzed (2-hydroxyethyl)-l-piperazineethanesulfonic acid; FCCP, carbonyl TMA, tetramethylam- in detail the effects of K+ and membrane potential on the cyanide p-trifluoromethoxyphenylhydrazone; monium; BAPTA, 1,2-bis(2-aminophenoxy)ethane-N,N,N,N-tetra-kinetics of Na-Ca and Ca-Ca exchange in bovine ROS; we acetic acid. propose a three-sitemodel for the Na-Ca exchanger in bovine
Two recent studies reportedthat Na-Ca exchange in the outer segments of tiger salamander rod photoreceptors (Cervetto,L., Lagnado, L., Perry, R. J., Robinson, D. W., and McNaughton, P. A. (1989) Nature 337, 740-743) and of bovine rod photoreceptors (Schnetkamp, P. P. M., Basu, D. K., and Szerencsei, R. T. (1989)Am. J. Physiol. 257, C153-157) requires and transports K+in a 4Na/( lCa+lK) stoichiometry. In this study, we have examined the effects of K+ ions and membrane potential on the kinetics of Na-Ca and CaCa exchange in rod outer segments isolated from bovine retinas. The objective was to establish the ion selectivity andvoltage dependence of the different cation binding sites on the Na-Ca-K exchange protein. Potassium ions activated Na-Ca exchange when present on the Ca2+side, although the extentof activation decreased with decreasing Na+ concentration. Potassium ions inhibited Na-Ca exchange when present on the Na+ side; inhibition arose from competition between Na+ and K+ for a common single cation-binding site. Activation of Na-Ca exchange by K+ displayed a different ion selectivity than that observed for inhibition of Na-Ca exchange by K+. The results are interpreted in terms of a three-site model for the rod NaCa-K exchanger. The rate of forward Na-Ca exchange decreased by 1.75-fold for a 60 mV depolarization of the plasma membrane but only at lower Na+concentrations. The rate of Ca-Ca exchange was not affected by changes in membrane potential.
189
190
Effect of Potassium on Na-Ca-K Exchange in Rods
ROS and discuss the ion selectivity of the three sites. In the accompanying paper, we describe the unidirectional Na+, Ca2+,and K+ fluxes through the Na-Ca exchanger from bovine
ROS. MATERIALS AND METHODS
Bovine retinas were obtained from fresh eyeballs kept in a lighttight container and purchased from a local abattoir. Ca2+-enriched and Ca2+-depletedROS were purified on sucrose-Ficoll gradients as described before (Schnetkamp et al., 1979; Schnetkamp, 1986). Ca" fluxes in ROS were measured in suspensions of ROS with the Ca2+indicating dye arsenazo I11 in the suspension medium; details of these procedures were described before (Schnetkamp, 1986). In all our calculations bovine ROS are assumed to be cylinders of 1 X 20 pm and theoverall ROS rhodopsin concentration is assumed to be 3 mM (Daemen, 1973).All procedures were done under dim red illumination. Potassium-depleted ROS were prepared by treating a suspension of ROS in the standard medium (600 mM sucrose, 20 mM Hepes adjusted to pH 7.4 with arginine) with the nonspecific ionophore ) 20 mM methylamine hydrochloride for 5 min gramicidin (2 p ~ and a t room temperature. The overall rhodopsin concentration in the . ROS were sedimented for 20 min suspension was 5 p ~ Subsequently, a t 4,000rpm (Beckman 520 rotor) and resuspended in a potassiumfree medium containing 600 mM sucrose, 20 mM Hepes (adjusted to pH 7.4 with arginine), and 5% Ficoll400 to a final rhodopsin concen. above procedure removed the 20-30 mM tration of 200-300 p ~ The total K+ normally present in Ca2+-enrichedROS (as measured with atomic absorption spectroscopy). "Ca fluxes were measured with a rapid filtration technique and liquid scintillation counting under normal room lights as described in the accompanying paper. The electrogenicity of Ca2+uptake via reverse Na-Ca exchange was determined as Ca2+-induced proton influx via the electrogenic protonophore FCCP the proton influx formed a currentloop with outward Na-Ca exchange current and was measured with the pH-indicating dye phenol red in adual-wavelength spectrophotometer as described previously (Schnetkamp et al., 1989). RESULTS
In a previous study we demonstrated that Na-Ca exchange in bovine ROS requires K+ and Ca2+ on one side of the membrane and Na+ on the other (Schnetkamp et al., 1989). This suggests that the Na-Ca exchanger has possibly three sites, one for each of the above three ions. In order to evaluate the contribution of the Na' and K' gradients to the equilibrium free Ca2+concentration, it is important to know the ion selectivity of each site, in particular to which degree these sites are selective, e.g. Na' does not bind to the K+ site and vice versa. Throughout this study the term forward Na-Ca exchange will be usedto indicate Ca2+efflux via Na-Ca exchange, while the termreverse Na-Ca exchange will beused to indicate Ca2+ influx via Na-Ca exchange. Ion Selectivity of Activation of Na-Ca Exchange by Alkali Cations-First we analyzed the ion selectivity of the K+ site on the ROS Na-Ca exchanger by measuring the ability of different monovalent cations to activate Na-Ca exchange when present on the same side as Ca2+(Fig. 1).Forward NaCa exchange was analyzed in Ca2'-rich ROS depleted from internal K' by treatment with the ionophore gramicidin (see "Materials and Methods"). The subsequent experiments were done in the presence of gramicidin to equilibrate externally added alkali cations in the intracellular space (Schnetkamp, 1985a). As described before (Schnetkamp et aL, 1989), K' depletion strongly reduced the rate of forward Na-Ca exchange, whereas readdition of K+ restored forward Na-Ca exchange. Rb+ and ammonium could substitute for K+ and caused full activation of forward Na-Ca exchange, and Cs+ caused a slight activation of forward Na-Ca exchange, whereas Li+ and the organic cations choline and tetramethylammonium had no effect (Fig. lA). The observation that Li' could
50mM Na+
//
O.Ol 10
20
30
40
50
BO
70
80
,
90
Naf
do
seconds
FIG. 1. Ion selectivity of the K+ site of the Na-Ca-K exchanger inbovine ROS. A, forward Na-Ca exchange was measured in a suspension of Ca2+-richand K+-depleted ROS. ROS were incubated in 600 mM sucrose, 100 p~ arsenazo 111, 2 p~ gramicidin, 20 mM Hepes (adjusted to pH 7.4 with arginine), and 10 mMof the indicated cation chlorides; rhodopsin concentration in the suspension was 4.4 p ~ Ca2+ . release was initiated by addition of 50 mM NaCl from a 2.5 M stock solution; Ca2+release is indicated by an increase in ODWOD7',, measured in a dual-wavelength spectrophotometer. The calibration bar of 0.05 absorbance unit represented a Ca2+release of 0.55 molof Ca2+/mol of rhodopsin or a decrease in total internal Ca2+ concentration by 1.65 mM. B, reverse Na-Ca exchange was measured in Ca2+-depletedROS incubated in 600 mM sucrose, 200 p~ arsenazo 111, 1 p~ FCCP, 100 p~ BAPTA, and 20 mM Hepes (adjusted to pH 7.4 with arginine); rhodopsin concentration in the suspension was . uptake was initiated by addition of 10 mM of the 11.9 p ~ Ca2+ indicated cation chlorides; 110 p~ CaC12 was added 15 s before addition of cations to yield a value for OD6so-OD7so of 1.1 absorbance unit. In the trace labeled gramicidin, 2 p M gramicidin was present. Caz+uptake in ROS is indicated by a decrease in ODsso-OD760and is plotted upward. Calibration bar of 0.05 absorbance unit represents a Ca2+uptake of 0.18 mol of Caz+/mol of rhodopsin or an increase in total intracellular Ca" concentration by 0.54 mM. Noise in the traces illustrated here and in all other figures did not reflect photometric sensitivity; it was caused by a spin bar used to mix the suspension. Temperature, 25 "C.
not activate forward Na-Ca exchange eliminates the possibility that internal K+ was required to electrically compensate for the inward Na-Ca exchange current by carrying an outward current via the channel ionophore gramicidin. Gramicidin not only equilibrates internal and externalK', but also internal and external Na'. Therefore, it is possible that forward Na-Ca exchange observed in thepresence of Na+ asthe only alkali cation was activated by internal Na' and represented a 4Na/(lCa+Na) exchange; in this case Na' would be a relatively poor substitute for K+. Rapid equilibration of the
Effect of Potassium on Na-Ca-K Exchange in Rods Na+ gradient in the presence of gramicidin had little effect on either the initial rate of Na+-stimulated Ca2+release or on the amount of Ca2+released provided that 5-10 mM K+ was present (not illustrated). A simple calculation demonstrates that forward Na-Ca exchange can be expected to cause a rapid equilibration of the Na+ gradient even in the absence of gramicidin. In the trace illustrated in Fig. lA, addition of 50 mM external Na+ in the presence of 10 mM K' caused the release of 3.5 mol of Ca2+/mol of rhodopsin in 30 s. With a stoichiometry of four Na' ions entering for each Ca2+ ion released, the above Ca2+release was accompanied by the influx of 14 mol of Na+/mol of rhodopsin and caused an increase in total intracellular Na+ by 42 mM. Activation of reverse Na-Ca exchange in Ca2+-depletedand Na+-rich ROS required extracellular K+; the ion selectivity of activation of Ca2+ uptake via reverse Na-Ca exchange by external alkali cations was very similar as observed for activation of forward Na-Ca exchange by internal alkali cations (compare Figs. 1, A and B ) . The uptake of Ca2+observed in the absence of any added alkali cations may in part be accounted for by external K+ (about 100 p ~ carried ) over from the Ficoll gradient used to purify ROS. Two further traces illustrated in Fig. 1B corroborate the notion that Ca2+ uptake in Ca2+-depletedROS observed here occurred exclusively via reverse Na-Ca exchange. First, addition of 10 mM Na+ completely inhibited Ca2+uptake and caused the release of the small amount of Ca2+ that was taken up prior to addition ofNa'. Second, addition of gramicidin and 10 mM external K+ caused the release of internal Na+ (measured by atomic absorption spectroscopy) and completely inhibited Ca2+uptake. K+-dependent and K+-independent Nu-Ca Exchange-The above results demonstrate that Na-Ca exchange is strongly activated by K+ when present on the same side as Ca2+. However, the rate of Na+-stimulated Ca2+release observed above in the absence of any other internalalkali cation than Na+ is significantly larger than the Na+-independent Ca2+ leak (Fig. lA). The residual Kf contamination in the ROS suspension amounted to less than 5 p~ as determined by atomic absorption spectroscopy and this is insufficient to support the residual Na+-stimulated Ca2+release (the sucrose media used in our experiments were passed over a mixed-bed ion exchanger). Here, we have examined how the K+ requirement is affected by the concentration of the two other substrates of the Na-Ca exchange protein. The K+ requirement of forward Na-Ca exchange depended strongly on the external Na+ concentration; it became progressively smaller as theNa' concentration was decreased (Fig. 2). At a Na' concentration of 5 mM, addition of 10 mM K' inhibited forward Na-Ca exchange since inhibition of Na+-stimulated Ca2+release by external K+ (Fig. 6) predominated activation by internal K+ (note that these experiments were carried out in the presence of gramicidin). A slight activation of forward Na-Ca exchange at 5 mM external Na+was observed at a lower K+ concentration of 0.5 mM in some, but not all preparations tested. The external K+ requirement of reverse Na-Ca exchange was examined in Ca2+-depleted ROS at two different free external Ca2+concentrations (Fig. 3). External K+ caused a large activation of Ca2+uptake at a free external Ca2+concentration of 1.5 ptM but much less activation at a free external Ca2' concentration of 150 p~ (compare inverted triangles with triangles in both panels of Fig. 3; the filled circles represent nonspecific absorption of 45Cato the borosilicate glass fiber filters when Ca2+uptake was prevented by the presence of 1 mM EDTA). Some of the Ca2+uptake observed without added K+can be accounted for by external K+ (about 200 p ~carried )
191 20 mM Na'
10 mM K+ present
"I 'I 0.05
no Na+
0
10
0
20
30
40
50
60
70
SO
90
SECONDS
FIG. 2. K' requirement of forward Na-Ca exchange depends on the external Na+ concentration. Forward Na-Ca exchange was measured in Ca2+-rich andK+-depleted ROS. Incubation medium and experimental conditions were as described in the legend of Fig. L4. The suspension contained 5.2 p~ rhodopsin. Ca2+release was initiated by addition of NaCl to theindicated final concentration. KC1 was present as indicated. The calibration bar of 3 mM represents a change in total internal Caz+concentration. Temperature, 25 "C. . . . , . . . , . .
6.0
.,
. ,
1.5pM FREE Ca'*
a + n
12.0
150pM FREE Ca2*
5.0
10.0
4.0
8.0
3.0
6.0
0
0
7
be 2.0 1 .o
0.0
-4.0
2.0
0.0
0
40
80
TIME (sec)
120
0
80 TIME (sec) 40
120
FIG. 3. K+ requirement of reverse Na-Ca exchange depends on the external Ca2+ concentration. Caz+-depleted ROS were incubated for 1 min in 600 mM sucrose, 20 mM Hepes (adjusted to pH 7.4 with arginine), 10 mMKC1 (inuerted triangles and diamonds), 5 PM gramicidin (diamonds),200 g M HEDTA (leftpanel), or 50 PM EDTA (rightpanel).The suspension contained 20 p~ rhodopsin. 45Ca uptake is expressed as the percentage of the total amount of 45Ca present in the suspension medium. Ca2' uptake was initiated at time 0 by addition of ' T a and 200 p M CaHEDTA (left panel, 200 +M HEDTA present), 200 PM CaC12 (right panel,50 PM EDTA present) or 200 p~ CaCI2plus 1 mM EDTA (circles in both panels). Temperature. 25 "C.
over from the Ficoll gradient used to purify ROS. The same specific radioactivity was used throughout the experiment illustrated in Fig. 3 to allow a quantitativecomparison in Ca2+ uptake at different free external Ca2+concentrations. For both Ca2+concentrations used in Fig. 3, we examined the effect of removal of internal Na' in the presence of gramicidin. At the lower Ca2+concentration, addition of gramicidin completely inhibited Ca2+uptake as expected for Ca2+ uptake via reverse Na-Ca exchange, but at higher Ca2+concentration of 150 p~ considerable Ca2+uptake was still observed in the presence of gramicidin (compare diamonds in right panel with those in left panel). This suggests that at higher external Ca2+concentrations pathways other than NaCa exchange contributed to Ca2+influx. Removal of internal
192
Effect of Potassium on Na-Ca-K Exchange in Rods
Na' required the presence of ionophores such as gramicidin, which perhaps could leadto artifacts. Therefore, we compared two different methods for measuring Ca2+uptake in Ca2+depleted ROS with a measurement of Na-Ca exchange current (Table I); if Ca2+ uptake can take placeviatwo separate pathways, no correlation between Ca2+ uptake and Na-Ca exchange current is anticipated. The Na-Ca exchange current in bovine ROS is accompanied by a countercurrentcarried by protons in the presence of the electrogenic protonophore FCCP, and this countercurrent was measured with the pHindicating dye phenol red (Schnetkamp et al., 1989). The outward Na-Ca exchange current forms a current loop with an inward current of protons carried by FCCP and measured as a Ca2+-inducedalkalinization of the medium. Electrogenic Ca2+influx via other pathways such as Ca2+channels or Ca2+ pumps is expected to produce an inward current accompanied by a Ca2'-induced acidification when FCCP is present. No evidence was obtained to suggest that Ca2+ uptake in the absence of gramicidin was significantly larger as compared with proton uptake (Table I), i.e. total Ca2+uptake did not exceed Ca2+uptake via electrogenic Na-Ca exchange under the condition that the electrogenicity is one extra charge carried by Na+ (and measured as proton uptake via FCCP) for each Ca2+taken up. Forward Nu-Ca ExchangeIs Inhibited by Extracellular K+In this section we address the question whether K+ gradients affect Na-Ca exchange in bovine ROS. We have previously shown that Na-Ca exchange in ROS transports K', and, therefore, the K+ gradient is expected to influence internal free Ca2+concentration (Cervetto et al., 1989; Schnetkamp et al., 1989). We measured the effect of K' and other cationson the rate and amplitude of forward Na-Ca exchange; these experiments were carried out in the absence of ionophores or in the presence of the electrogenic protonophore FCCP to control the membrane potential. At a concentration 1 pM, FCCP is the dominant conductance in the ROS plasma membrane, and the membrane potential is expected to be kept constantand equal tothe Nernstpotential for protons (Schnetkamp, 1985; Schnetkamp et al., 1989). When the external K+ concentration was increased, forward Na-Ca exchange was progressively inhibited independent of the presence of FCCP (Fig. 4). The pattern of inhibition of forward Na-Ca exchange by external K+was analyzed in a Dixon plot (Dixon and Webb, 1964), and one example of such analysis is illustrated in Fig. 5. Straight lines were observed with slopes TABLE I Correlation between CAz+uptake and reverse Na-Caexchange current Values for Ca2+uptake are average values k standard deviation with the number of preparations between parenthesis. Caz+ uptake was measured with arsenazo 111 as described in the legend of Fig. 1B (free Ca2+concentration about 5-10 y ~ ) ;Ca2+uptake was measured with 'Ta flux as described in the legend of Fig. 3 (right panel, free . uptake Caz+concentrations ranged between 100 and 500 b ~ )Proton was measured in 600 mM sucrose, 20 mM tetramethylammonium chloride, 0.5 mM Hepes, pH 7.4, 35 pM BAPTA, 10 mM KCl, and 40 +M phenol red. Reverse Na-Ca exchange was initiated by addition of 150 P M CaClZand calibrated by adding known amounts of HCl. Initial rate"
Amountb
Ca2+uptake (45Caflux)
1.0 (k0.4) (7)
1.1f 0.3 (7)
Proton uptake (Phenol red)
1.7 (k0.4) (7)
1.1 k 0.4 (7)
CaZ+uptake (arsenazo 111) 1.1 (&0.2) (4) 1.1 k 0.1 (4) Indicates X IO6 Ca2+/ROS/s or protons/ROS/s. Indicates mol of Caz'/molof rhodopsin or mol of protons/mol of rhodopsin.
that decreased as the external Na' concentration was increased. The lines representing different Na' concentrations up to 20 mM had a common point of intersection in the second quadrant well above the abscissa indicative for competitive inhibition between Na+ and K+. The inhibition constant for K+ was 5 , 6, and 8 mM, respectively, in three different experiments. At higher Na+ concentrations the intersection point was shifted to the left. The above behavior suggests that at low Na' concentration Na+ and K+ compete for a common site that binds asingle alkali cation; as theNa+ concentration is increased additional Na+ions will bind to theexchanger to sites to which K+ cannotbind, and in doing so the less specific cation site becomes more selective for Na', very similar to the situation observed for the Na-Ca exchanger in cardiac sarcolemma vesicles (Slaughter et al., 1983). Another graphic representation of the inhibition of Na-Ca exchange byK' analyzes the effect of K+ on the Na+-binding constant in the Hill plot (Fig. 6). Hill plots of the Na' concentration dependence of Na-Ca exchange invariably yielded straight and parallel lines with Hill coefficients between 1.7 and 2.3; for 24 different Hill plots of the Na' concentration dependence of forward Na-Ca exchange the average Hill coefficient was 2.0 (SD = 0.2). In the experiment illustrated in Fig. 6, the Na'binding constant increased from a value of 32 mM (at 5 mM K') to 54 and 82 mM (at 25 and 75 mM K+, respectively), and the Hill coefficients were between 1.7 and 1.8. The above observations on the effect of K+ on Na-Ca exchange in bovine ROS were analyzed in terms of competitive inhibition. However, K+ is also transported via the NaCa exchanger, and this suggests another way to explain inhibition of Na-Ca exchange by K' that does not require a common binding site for Na' and K'. An outward K' gradient is part of the driving force for extruding Ca2+ via a 4Na/ (lCa+lK) exchange; increasing the external K' concentration reduces the outward K' gradient and causes the release of less Ca2+.When the inward Na+ gradient is larger at higher external Na+ concentration the inhibitory effect of K+ becomes much less, since Na-Ca exchange is driven by the fourth power of the Na' gradient. We tested the idea that inhibition of forward Na-Ca exchange by external K' is caused by the diminished K+ gradient in two ways. First, inhibition should be limited to cations that can replace K+ in activating Na-Ca exchange (e.g. Rb', Fig. I), but other cations such as Li' and tetramethylammonium (TMA') should beineffective. We observed that all cationstested including Li+ and TMA' inhibited forward Na-Ca exchange (Fig. 7). Second, we examined inhibition of Na-Ca exchange by K+ and ammonium under conditions that no gradient of these cations was maintained. Ammonium ions could activate Na-Ca exchange (Fig. 1) and inhibited forward Na-Ca exchange veryeffectively (Fig. 7), although no gradient for ammonium ions across the ROS plasma membrane was maintained formore than a second due to permeation of the neutral species ammonia and acetic acid (Schnetkamp, 1985b). Likewise, the inhibition of forward Na-Ca exchange by K+ was not alteredin the presence of ionophores (e.g. FCCP plus valinomycin, nigericin, or gramicidin) that resulted in the rapid dissipation of the K' gradient (notillustrated). The above results demonstrate that K+ concentration and not K+ gradient causes inhibition of NaCa exchange when K' is present on the same side as Na+. Effect of Membrane Potential on Nu-Ca and Ca-Ca Exchange-Na-Ca exchange in isolated bovine ROS is an obligatory electrogenic process when an external ionophore is added to provide a countercurrent for the Na-Ca exchange current (Schnetkamp et al., 1989). Therefore, we examined the voltage dependence of both electrogenic Na-Ca exchange
Nu-Ca-K Exchange in Rods
Effect of Potassium on
A no ionophores FIG. 4. Inhibition of forward NaCa exchange by external K+. Ca2+enriched ROS were incubated in 600 mM sucrose, 100 PM arsenazo II1,l PM FCCP banel B onlv), KC1 as indicated in mM, and 20 mM -Hepes (adjusted to pH 7.4 with arginine); the suspension contained 4 p M rhodopsin. Ca2+release was initiated at time 0 by addition of 10 mM NaC1. The calibration bar of 0.05 absorbance unit represented a Ca2+release of 0.69molof Caz+/mol of rhodopsin ora decrease in total internal Ca2+concentration by 2.07 mM. Temperature, 25 "C.
193
0 lpM FCCP
0
5 mM K' 10
10
m M Na'
:e, I
25
25
50
50
o,05
0
I
0.05
no Na+
0
20
40
80
60
100
120
0
20
40
seconds
60
80
100
120
seconds
1OmM NaCI A 15mM NaCl m20rnM NaCI *30mM NaCI
4.0
10 mM Na' w 3.0 I-
a E
-
\
2.0
1 .o
I
0.0
25 50 75 100 KC1 (mM) FIG. 5. Dixon plot of inhibition of forward Na-Ca exchange by external K+. Forward Na-Ca exchange was measured at different Na+ concentrations as a function of extracellular K+ concentration as described in legend of Fig. 4 (FCCP was always present). For each Na+ concentration, the initial rate of Na+-induced Caz+release was plotted as a function of external K+ concentration in Dixon plots. Temperature, 25 "C. 0
1.0 -
h
> I
0.5
-
0.0
-
E
>
5 u
-0.5
;
: -
-1.0
-1.5 -2.0' 4.0
'
'
"
I
"
10.0
100.0
NaCl (mM) FIG. 6. Hill plots of forward Na-Ca exchange at three different external K+ concentrations. Forward Na-Ca exchange was measured a t different Na+ concentrations for three different values of the extracellular K+ concentration; protocol was as described in the legend of Fig. 4 (FCCP was always present). The initial rate of Na+-induced Ca2+release was measured and converted in a Hill plot. Temperature, 25 "C.
7
leak SECONDS
FIG. 7. Inhibition of forward Na-Ca exchange by external cations. Forward Na-Ca exchange in the presence of 1 PM FCCP was measured as described in the legend of Fig. 4. The indicated cations were present at a concentrationof 50 mM. Rhodopsin concentration in the suspension was 5.3 PM. Ca2+release was initiated at time 0 by addition of 10 mM NaC1. The calibration bar of0.05 absorbance units represented a release of0.55molof Ca2+/mol of rhodopsin or a decrease of the total internal Ca2+concentration by 1.65 mM. Temperature, 25 "C.
and electroneutral Ca-Ca self-exchange. The membrane potential of ROS in suspension can be controlled with the aid of ionophores such as FCCP and the K+-selective ionophore valinomycin. Different membrane potentials were established with different extracellular K+ concentrations in the presence of valinomycin. In order to differentiate between direct inhibition of forward Na-Ca exchange by external K+ and inhibition of forward Na-Ca exchange by K+-induced depolarization of the plasma membrane in thepresence of valinomycin, we chose as a control the effects of K+ on forward Na-Ca exchange in the presence of the electrogenic protonophore FCCP. In thepresence of valinomycin the membrane potential is controlled by the K+gradient, whereas in thepresence of FCCP the membrane potential is held constant and equal to the Nernst potential for protons. This procedure is illustrated in Fig. 8. When valinomycin was added at an external K+ concentration of 5 mM, the plasma membrane probably hyperpolarized due to the outward K+ gradient, and the rate of forward Na-Ca exchange increased (to 9.4 X lo6 Ca2+/outer segment/s) as compared with that observed in the presence
Effect of Potassium on Na-Ca-K Exchange in Rods
194
5 mM K +
Na'
50 mM
100 mM K+
FIG. 8. Voltage dependence of forward Na-Ca exchange. Ca2+-enriched ROS were incubated in 100 PM arsenazo 111, 1 PM FCCP, or 1PM valinomycin (as indicated), 145 mM TMACl + 5 mMKC1 (labeled 5 mM KC]) or 50 mM TMACl + 100 mM KC1 (labeled 100 r n KCl), ~ 20 mM Hepes (adjusted to pH 7.4 with arginine); the suspension contained 3.9 PM rhodopsin. Ca2+ release was initiated at time 0 by addition of 50 mM NaCl. The calibration bar of 0.05 absorbance unit represented a Ca" release of 0.91 mol of Ca2+/mo1of rhodop- 0.05 sin or a decrease of the total internal Caz+ concentration by 2.73 mM. Temperature, 25 "C.
0
I
I
I
I
I
I
1
I
1
0
10
20
30
40 seconds
50
60
70
FIG. 9. Voltage
.. .-
:..: ..
2.0
10.0
of FCCP (6.7 X lo6 Ca2+/outersegment/s). At a high external K+ concentration of 100 mM, addition of valinomycin probably depolarized the ROS plasma membrane and caused a decrease in the rate of forward Na-Ca exchange (to 0.6 X lo6 Ca2+/outersegment/s) as compared with that observed in the presence of FCCP (3.9 X lo6 Ca2+/outersegment/s). The rate of forward Na-Ca exchange in the presence of valinomycin was divided by the rate in the presence of FCCP, and this ratio was plotted in Fig. 9 as a function of changes in the membrane potential (represented by the logarithm of the external K+ concentration); alinear relationship was observed of forward Na-Ca exchange with a 1.9-foldincrease in the rate for a -60 mV change in the membrane potential, or an e-fold increase in rate for -85 mV change in membrane potential. For five different preparations an average 1.75- (S.D. = 0.17) fold increase was observed for a -60 mV change in membrane potential. In contrast, changes in membrane potential did not alter the rate of forward Na-Ca exchange when the external Na+ concentration was raised to 150 mM. Our results are very
dependence
of
concentration, both in the presence of FCCP as well as in the presence of valinomycin (similar to the data illustrated in Fig. 7). For each K+ concentration, the rate observed with valinomycin pres-
100.0
similar to those determined by direct electrical measurements in tiger salamander rods (Lagnado et al., 1988). Voltage dependence of Na-Ca exchange could be due to a number of factors including the electrogenicity of Na-Ca exchange, i.e. the translocation of a net charge, or the placement of the Na+-binding site within the electric field of the membrane as suggested by Lagnado et al. (1988). The Ca-Ca self-exchange process (e.g. Schnetkamp, 1980) or (Ca+K)/ (Ca+K) exchange (accompanying paper) does not involve transport of a netcharge and enables us to evaluate if aspects of Na-Ca exchange other than its electrogenicity contribute to itsvoltage dependence. For example, several kinetic models of the Na-Ca exchanger predict a voltage dependence for CaCa exchange (e.g. Hilgemann, 1988).We measured the uptake of 45Caas anindicator of Ca-Ca exchange (Schnetkamp, 1979, 1980) and examined its dependence on membrane voltage with the same protocol as used above for the voltage dependence of Na-Ca exchange (Fig. 10). Ca-Ca self-exchange (free Ca2+concentration -K,) differed from forward Na-Ca ex-
Effect of Potassium on Nu-Ca-K Exchange in Rods
FIG. 10. Voltagedependence
195
24.0
24.0
18.0
18.0
of
Ca-Ca exchange. Electroneutral '%a-
40Caexchange was measured inCa2+-rich ROS by means of a rapid filtration technique. ROS were incubated in 20 mM Hepes (adjustedto pH 7.4 with arginine), 100 p~ CaC12, 150 p~ HEDTA, 2 pM FCCP, or 2 p~ valinomycin (as indicated) and a mixture of TMACl and KC1 (total concentration 150 mM) to give the desired KC1 concentration. The rhodopsin concentration was 15 p M . 45Cauptake is expressed as the percentage of the total amount of 'Ta present in the suspension medium. Uptake was initiated by addition of "Ca at time 0. Temperature, 25 'C.
X
3
1 L
Z
-0 12.0
12.0
0 In
6\"
6.0
6.0
0.0
0.0 0 80
40
TIME (sec.)
120
0 80
40
120
TIME (sec.)
effects of ions and membrane potential on Ca2+ fluxes reported in this study were caused by direct effects of ions and membrane potential on the exchanger. A more complex situation is encountered in Ca+-depleted ROS, net Ca2+uptake could be observed when all Na' was removed from ROS by treatment with gramicidin (Fig. 3, right panel). Perhaps this is an artifact caused by addition of gramicidin, since in the absence of gramicidin a good quantitative correlation was observed between Ca2+uptake and outward Na-Ca exchange current (Table I). Effects of Alkali Cations on Na-Ca Exchange-The objective of our experiments was to analyze the different cation-binding sites on the Na-Ca exchanger and to examine the effect of DISCUSSION occupation of these sites by different cations on Ca2+transThe effect of different cations on Na-Ca exchange was port. We wished to establish if Na+, K', and Ca2+ share studied in intact ROS isolated from bovine retinas. In order common sites, since such knowledge is required to evaluate to obtain maximal Na-Ca exchange fluxes, ROS were exposed the effect of ion gradients on regulation of intracellular free to rather unphysiological isolation protocols designed to yield Ca2+concentration via Na-Ca exchange, the ultimate physeither Ca2+-rich ROS or Ca'+-depleted/Na+-rich ROS; sub- iological function of Na-Ca exchange. If K+ and Na+ share sequent removal of the large load of Ca2+or Na+was examined common transport siteson the Na-Ca exchanger, the internal via the forward and reverse modes of Na-Ca exchange, re- free Ca2+ concentration cannotbe obtained in a simple manfunction of external Ca2+ concentrationand the spectively. It has been shown that Na-Ca exchange is the only nerasa electrogenic transporter operating in the plasma membrane transmembrane Na+ and K+ gradients. In other preparations of isolated ROS when the cGMP-dependent channelis closed including squid giant axon, heart sarcolemma vesicles and (Lagnado et al., 1988; Schnetkamp et al., 1989); it has also synaptosomes, a two-site model for the Na-Ca exchanger has been shown that the light-sensitive channel and the Na-Ca been proposed (reviewed by Reeves, 1985);an A-site binds a exchanger are the only two proteinscontributing to Ca2+ single Ca2+ion or two Na+ ions, a B-site binds alkali cations fluxes across the ROS plasma membrane (Yau and Nakatani, in a nonselective manner when Ca2+occupies the A-site, but 1984; Lagnado et aL, 1988; Schnetkamp, 1979, 1986). These occupation of the A-site by Na+ makes the B-site selective for conclusions were derived from electrical measurements on Na+ as well. This model was prompted by the observation amphibian ROS and from an analysis of Ca2+fluxes in bovine that Ca-Ca exchange can be stimulated by alkali cations in a ROS. Analysis of Na+, Ca2+,and K+ fluxes in isolated Ca2+- relatively nonselective manner (e.g. Li+, Na+, and K+). However, the occupant of the B-site is transported only whenboth rich bovine ROS in the accompanying paper demonstrates that these fluxes can all be accounted for by the different A- and B-sites are occupied by Na+ and not when the A-site partial reactions of the Na-Ca-K exchanger present in the is occupied by Ca2+and transportsCa2+.The bovine ROS NaROS plasma membrane; no cation fluxes due to other electro- Ca exchanger appears to have a site(s)very similar to the Agenic or electroneutral transporters were detected that ex- site of other exchangers, two Na+ ions appear to compete for ceeded a flux of about lo5 cations/ROS/s or a current of 0.01 a common site with one Ca2+(Schnetkamp, 1986).The bovine pA (the rate of Na-Ca exchange fluxes observed in bovine rod Na-Ca exchanger has an alkali cation-binding site with ROS ranged between 5-10 X lo6 cations/ROS/s). Hence, the some features in common with the B-site of other exchangers, change in two respects, no clear voltage dependence was observed and Ca-Ca self-exchange was not inhibited by external K+. The slight inhibition of 45Ca-Caexchange noticed in this experiment at 25,50, and 150 mM KC1 in thepresence of valinomycin is probably due to experimental variability and was not observed in three otherexperiments. External K+did have other effects on Ca-Ca exchange; as described before (Schnetkamp, 1980), external K+ increased the rate of Ca-Ca exchange a t Ca2+concentrations well below the Km of the exchanger for Ca2+(e.g. 0.1 p ~ )and , decreased the rateof CaCa exchange at Ca2+ concentrations well above the K,,, (e.g. 200 pM) (not illustrated).
196
Effect of Potassium on Na-Ca-K Exchange in
and such a site may be invoked to explain the inhibition of forward Na-Ca exchange by external K' (Figs. 4-6). In the absence of Ca2+, Na' and K+ compete for the B-site, and occupation of the B-site by K+ prevents Na-Ca exchange; as the Na+ concentration increases the A-site is occupied by two Na' ions, which makes the B-site selective for Na+ as well and inhibition by K+ is relieved. However,the effects of alkali cations on Na-Ca exchange in bovine ROS are very different from those observed in other preparations in most aspects. Therefore, we will interpret our results in terms of a threesite model of the ROS Na-Ca exchanger, which will retain as many features from the two-site model as possible. The model should account for the two principal transport modes of the exchanger (4Na/(lCa+lK)and(lCa+lK)/(lCa+lK))and should account for the observations that Na' competes with K' for a common site, that Na+ competes with Ca2+for a common site, but that Ca2+ and K+ do not compete for a common site. A Three-site Model for Na-Ca Exchange in Bovine ROSTwo general kinetic schemes can be considered for an exchange protein, a consecutive scheme and a simultaneous scheme (for a discussion of these kinetic schemes as applied to Na-Ca exchange consult Laiiger,1987 and Hilgemann, 1988). For a consecutive scheme the exchanger has one set of binding sites that are alternately exposed to thetwo aqueous compartments; this one set of sites can accommodate either four Na+ ions or one Ca2+ion plus one K+ ion. A consecutive scheme predicts the presence of Na-Na and (Ca+K)/(Ca+K) self-exchange. For a simultaneous scheme the exchanger has two sets of sites, one that accommodates four Na' ions and another thataccommodates one Ca2+plus one K+; during the translocation step the two sets of sites trade places. In a simultaneous scheme the two sets of sites could be very different, and self-exchange processes are not an intrinsic property of a simultaneous scheme. The observation of the self-exchange processes of (Ca+K)/(Ca+K) and Na-Na exchange in bovine ROS (accompanying paper) dictates that in a simultaneous scheme the two sets of sites need to be very similar. A more extensive discussion on the above topic is presented elsewhere (Schnetkamp, 1990). For both kinetic schemes discussed above, athree-site model forthe Na-Ca exchanger in bovine ROS should be able to accommodate either four Na+ ions or one Ca2+ion plus one K+ ion. Wewill name the three sites of the ROS Na-Ca exchanger the A-, B-, and K-site, respectively. It should be kept in mind that our model offers only a simple framework to interpret observations and is not intended to provide an accurate description of the molecular events underlying transport of cations via the Na-Ca exchange protein. The A-site can bind either one Ca2+or two Na' ions; other alkali cations appear to have no affinity for this site since K+ (and Rb', Cs') does not compete with Ca2+for common sites (Fig. 10; Schnetkamp, 1980). Divalent and trivalent cations such as M$+, Mn2+, Sr", and La3+ allbind to this site resulting in inhibition of both Na-Ca exchange and Ca-Ca exchange, but only Sr2+ andCa2+(and under some conditions Ba2+) are transported when occupyingthis site (Schnetkamp, 1979, 1980). Occupation of the A-site by Ca2+can result in transport of Ca2+via Ca-Ca exchange independent of occupation of the other two sites, since Ca-Ca exchange does not require K+ (accompanying paper, Schnetkamp, 1980, 1990). It is possible that forward Na-Ca exchange in the absence of internal K' (Fig. 2) reflects Caz+release with the K-site empty (this issuewillbe discussed further in the accompanying paper). Occupation of the A-site by two Na' ions is not sufficient for transport (see below).
Rods
The B-site is a non-selective alkali cation site; occupation of this site by Li+ or K' has no clearly noticeable effect when Ca2+occupies the A-site but causes inhibition of Na-Ca exchange when Na+ occupies the A-site (Fig. 7). We suggest that competition between Na' and K+for the B-site accounts for the competitive inhibition of forward Na-Ca exchange by external K+ as judged from Dixon plots (Fig. 5). Transport from the B-site occurs only when the A-site is occupied by Na' similar as in the two-site model discussed byReeves (1985). The thirdsite, the K-site is only found on the ROS Na-Ca exchanger. When the A-site is occupied by Ca2+,the K-site is selective for K+, Rb', and ammonium (Fig. 1)and itsoccupant is transported (Schnetkamp et al., 1988, 1989, and accompanying paper). Occupation of the A-site by Na' renders both the B- and K-sites selective for Na' providing binding sites for the four Na' ions that are transportedin exchange for one Ca2+and one K'. However, when Ca2+occupies the A-site, the K-site has very little affinity for Na+ anddoes not transport Na+ (accompanying paper). When Ca2+occupies the Asite, simultaneous occupation of the K-site by K+ confers a number of properties on the ROS Na-Ca exchanger not found for Na-Ca exchangers in other tissues: 1) Occupation of the K-site byK' greatly increases the maxima1velocity of Na-Ca exchange (Schnetkamp et al., 1989; Figs. 1 and 2) but has little effect on the maximal velocity of Ca-Ca exchange (Schnetkamp, 1980,1990). 2) Occupation of the K-site by K+ results in transport of K+ in both theNa-Ca exchange mode and in the Ca-Ca selfexchange (accompanying paper). Thus, Ca-Ca exchange in bovine ROS is accompanied by K-K exchange as judged from Ca2'-activated =Rb fluxes, whereas activation of Ca-Ca exchange by K+ in cardiac sarcolemma vesicles is not accompanied by "Rb fluxes (Slaughter et aL, 1983). 3) The K-site enables the exchanger to couple transport of four Naf ions instead of three Na+ions against one Ca2+,and enables the exchanger to lower free intracellular Ca2+concentration in the submicromolar range, even when the Na+ gradient is somewhat compromised because of the steady Na+ influx through the cGMP-gated channels. 4) Occupation of the K-site by K+ alters the properties of the A-site. Thus, K' decreases the K,,, of the A-site for Caz+, relieves competition of Ca2+transport by Mg2+ (Schnetkamp, 1980), and increases the K,,, of the A-site for Na+ (fig. 6). In other words, highK' confers a conformation on the exchanger that favors binding of Ca2+rather than binding of Na+ and also favors binding of Ca2+rather thanbinding of M e , useful attributes for a protein designed to remove Ca2+ from the cytoplasm. Effect of Membrane Potential on Na-Ca and Ca-Ca Exchange-The voltage dependence of Na-Ca exchange and CaCa exchange can provide information on the individual steps in the overall transport cycle (Eisner and Lederer, 1985; Lauger, 1987; Hilgemann, 1988). The rateof Ca-Ca exchange showed very little dependence on membrane voltage when tested at an external free Ca2+ concentration nearthe K,,, of the exchanger for Ca2+; thissuggests that the external Ca2'binding site is not within the electric field of the membrane. It also suggests that the Ca2+-loadedconformation of the exchanger does not carry a net charge. The rate of forward Na-Ca exchange was decreased by depolarization of the ROS plasma membrane, but only at lower Na' concentrations (Fig. 8 and itsdiscussion) suggesting that theexternal Na+binding site(s) is located within the electric field of the membrane (Lagnado et al., 1988). Voltage-independent binding of Ca2+ suggests that theA-site discussed above is not located within
Effect of Potassium Na-Ca-K on Exchange the electric field of the membrane, whereas voltage-dependent binding of Na' suggests that the A-site is located within the electric field of the membrane (it is generally thought that binding of Na+ to the A-site controls the rate of Na-Ca exchange and the K,,, of the exchanger for Na+ appears to reflect binding of Na' to theA-site). The apparentdifference in location of the A-site may reflect the conformational change of the cation-binding sites that accompanies binding of the first Na' ion to the A-site (Reeves, 1985). REFERENCES Cervetto, L., Lagnado, L., Perry, R. J., Robinson, D.W., and McNaughton, P. A. (1989) Nature 337, 740-743 Cook, N. J., and Kaupp, U. B. (1988) J. Bwl. Chem. 2 6 3 , 1138211388 Daemen, F. J. M. (1973) Biochim. Biophys. Acta 300, 255-288 Dixon, M., and Webb, E. C. (1964) The Enzymes,Longmans, London Eisner, D. A., and Lederer, W. J. (1985) Am. J . Physiol. 248, C189c202 Gold, G. H. (1986) Proc. Natl. Acad. Sci. U. S. A. 8 3 , 1150-1154 Hilgemann, D. W. (1988) Prog. Biophys. Mol. Bwl. 5 1 , 1-45 Hodgkin, A. L., and Nunn, B. J. (1987) J. Physiol. 3 9 1 , 371-398 Hodgkin, A. L., McNaughton, P. A., and Nunn,B. J. (1985) J. Physiol. 368,447-468 Hodgkin, A. L., McNaughton, P. A., and Nunn,B. J. (1987) J. Physiol. 39 1,347-370 Kaupp, U.B., and Schnetkamp, P. P. M. (1982) Cell Calcium 3, 83112 Koch, K.-W., and Stryer, L. (1988) Nature 3 3 4 , 64-66 Lagnado, L., Cervetto, L., and McNaughton, P. A. (1988) Proc. Nat. Acad. Sci. U. S. A. 85,4548-4552
in Rods
197
Lauger, P. (1987) J. Membr. Biol. 9 9 , 1-11 Matthews, H. R., Murphy, R. L. W., Fain, G. L., and Lamb, T. D. (1988) Nature 3 3 4 , 67-69 McNaughton, P. A., Cervetto, L., and Nunn, B. J. (1986) Nature 3 2 2 , 261-263 Miller, D. L., and Korenbrot, J. I. (1987) J . Gen. Physiol. 90, 397425 Nakatani, K., and Yau, K.-W. (1988) Nature 334,69-71 Nicol, G. D., and Bownds, M. D. (1989) J. Gen. Physwl. 94,233-259 Nicoll, D.A., and Applebury, M. L. (1989) J. Biol. Chem. 264,1620716213 Ratto, G. M., Payne, R.,Owen, W. G., and Tsien, R. Y. (1988) J. Neurosci. 8 , 3240-3246 Reeves, J. P. (1985) Curr. Top. Membr. Tramp. 25, 77-127 Schnetkamp, P. P. M. (1979) Biochim. Biophys. Acta 554, 441-459 Schnetkamp, P. P. M. (1980) Biochim. Biophys. Acta 598,66-90 Schnetkamp, P. P. M. (1985a) J. Membr. Biol. 8 8 , 249-262 Schnetkamp, P. P. M. (1985b) J. Membr. Biol. 88,263-275 Schnetkamp, P. P. M. (1986) J. Physwl. 373, 25-45 Schnetkamp, P. P. M. (1990) Progr. Biophys. Mol. Biol., in press Schnetkamp, P. P. M., and Bownds, M. D. (1987) J. Gen. Physiol. 89,481-500 Schnetkamp, P. P. M., and Kaupp, U. B. (1985) Biochemistry 24, 723-727 Schnetkamp, P. P. M., Klompmakers, A.A., and Daemen, F. J. M. (1979) Biochim. Biophys. Acta 552,379-389 Schnetkamp, P. P. M., Szerencsei, R. T., and Basu, D. K. (1988) Bioph-ys. J. 53, 389a Schnetkamp, P. P. M., Basu, D. K., and Szerencsei, R. T. (1989)Am. J. Physiol. 257. C153-Cl57 Slaughter, R. S., Sutko, J. L., and Reeves, J. P. (1983) J. Biol. Chem. 258,3183-3190 Yau, K.-W., and Nakatani, K. (1984) Nature 3 1 1 , 661-663 Yau, K.-W., and Nakatani, K. (1985) Nature 313, 579-582
