Vanadate-induced Movements of Ca2+ and K’ in Human Red Blood Cells *

Fresh human red blood cells become highly labeled with 4aCaz+ when exposed to 0.5 m~ vanadate. The effect of vanadate requires its penetration into the cell, and is attributed to the inhibition of the outwardly directed Ca2+-pumping ATPase which would otherwise “mask” the uptake of 46Ca2+. Since the inhibition of the Ca2+ pump by vanadate is not complete, a transmembrane Ca2+-Ca2+ exchange can be detected. The influx leg of the exchange is inhibited by verapamil, quinidine, and Co2+. This, as well as additional (kinetic) evidence, indicates that the influx of Ca2+ is a carrier-mediated process. Experiments in which the transmembrane K+ gradient has been abolished or decreased with ionophores, or by increasing the K+ concentration in the medium, suggest that the K+ gradient may play a role in the influx of Ca2+. The vanadate-induced accumulation of Ca2+ by red cells promotes a massive efflux of K+, indicating the activation of a Ca2+-sensitive K+-chan-nel. The results indicate the occurrence of a slow cycling of Ca2+ across the red cell membrane. The influx leg of the cycle occurs through a verapamil-sensitive channel, and is possibly driven by the discharge of the trans- membrane K+ gradient. The efflux leg of the cycle consists of the Caz+-pumping ATPase. The extrusion by The functional of cell

temperature, i.e. under conditions which should inhibit the Ca-ATPase. This procedure indeed unmasked the influx of Ca2+.
The study of the mechanismb) for the inward transport of Ca2+ into human red blood cells is of a particular interest. These cells apparently lack functions mediated by the influx of Ca2+, yet possess an efficient Ca pump, which is normally used far below its f u l l capacity (10). Intact red blood cells containing a normal amount of ATP do not take up added in amounts exceeding 0.1 to 1.0 pmol/liter of packed cells. After ATP depletion, however, they accumulate Ca2+ up to a concentration 50 times higher than in the normal state (see Ref. 10, for a review). This suggests that the activity of the Ca-ATPase, rather than the "absolute" impermeability of the cell membrane is the cause of the negligible uptake of 45Ca2+ usually observed in the red blood cells.
In the present study, we have used an alternative approach to the study of the influx of 45Ca2+ in human red blood cells. In intact fed erythrocytes, we have induced Ca2+ influx by preincubation with vanadate, a compound that inhibits the Ca-ATPase in red cell membranes (11) and in the purified state (7), and the (Na,K)-ATPase in intact red blood cells after permeating through the membrane via the anion channel (12). The results have provided support for the concept that the influx of Ca" into red blood cells is a carrier-mediated process. The sensitivity of the Ca2+ influx to verapamil suggests that the transport system in red blood cells is similar, or identical, to the slow Ca2+ channel of excitable tissues. Moreover, the results suggest that the gradient of K+ ions may play a role in the influx of Ca". 4

MATERIALS AND METHODS
Red Blood Cells-Fresh citrate-treated human blood was provided by a local blood bank, stored under sterile conditions at 0-4 "C, and used within 3 weeks.
Red blood cells were prepared immediately before the experiment. Blood was centrifuged (5 min, 2500 X g). The plasma and the buffy layer were sucked off, and the red blood cells were suspended in a medium containing 50 m Na-Hepes,' pH 7.2, 100 mM NaC1, 5 mM KC1,0.5 m~ MgC12, and 5 m glucose. The suspension was centrifuged as above, and the washing procedure was repeated four times. The washed cells were resuspended in the same medium at a concentration of 30%. 45Ca2+ Uptake Experiments-The suspension of cells (hematocrit 30%) was preincubated 15 min with 0.5 m Na-ortho-vanadate from a 0.5 M stock solution in 0.5 M Na-Hepes, pH 7.2. 45CaC12 (specific activity about 2500 cpm/nmol) was then added to a final concentration of 2.5 m~, unless indicated otherwise. After the addition of Cazi, 0.5-ml aliquots of the suspension were withdrawn at 0, 20, 40, 60, 90 The abbreviations used are: Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; TPP+, tetraphenylphosphonium; diS-C3-(5), min (unless indicated otherwise), diluted to 1.5 ml with medium containing 0.5 m~ vanadate, and centrifuged 0.5 min in an Eppendorf 3200 microcentrifuge. The supernatants were discarded and the pellets suspended in 1 ml of the medium and recentrifuged as above. The washing procedure was repeated four times. Finally, the pellets were treated with 0.5 ml of 10% (w/v) trichloroacetic acid containing 20 n" Lac&. After removal of the precipitate by centrifugation, 0.5 ml of the supernatant was used for the measurement of radioactivity in a scintillation counter using Biofluor New England Nuclear as a scintillation fluid. The experiments were performed at 25 "C. The control samples were treated as above but without vanadate in the suspension, or in the washing medium. The order of addition of the compounds tested is indicated in the figures.
"Ca2' Efflux Experiments-The cells were loaded with 45Ca2+ as described above (1-h incubation with 45Ca2'), and then washed four times at 0 "C with standard medium containing 0.5 n" vanadate. The washing procedure reduced the radioactivity of the medium practically to the background level. The suspension was then adjusted to the original volume, divided into aliquots, supplemented with the compounds to be tested, and incubated at 25 "C. 0.9-ml aliquots were withdrawn at the times indicated in the figures, and centrifuged in an Eppendorf 3200 microcentrifuge. After careful separation of the supernatants from the pellets, the latter were precipitated with trichloroacetic acid, and counted as above. No correction was made for the supernatant trapped in the extracellular space.
K' Efflux Experiments-The incubation of the suspension with vanadate and Ca2' was performed as described for the 45Ca2' uptake experiments except that nonradioactive Ca2' was used. The aliquots withdrawn at the time indicated in the figures were centrifuged, and the supernatants separated from the pellets. The content of K' in the supernatants was measured by atomic absorption in samples diluted 100 times.
ATP Content of Red Blood Cells-The conditions were the same as for the K' efflux experiments except that the aliquots of the suspensions were added to the test tubes containing 40 pl of 70% HClO,. ATP was measured in the supernatant after the removal of the precipitated proteins by a coupled enzyme assay method (Boehringer Mannheim AG).
Concentration of Free Ca2+-The change of absorbance of Arsenazo I11 in a dual wavelength spectrophotometer was used, with 660 and 685 nm as the wavelength pair. The concentration of Arsenazo I11 was 50 and 300 p~, in the medium used for the Ca2' uptake experiments.
Concentration of TPP'-The concentration of TPP' was measured by a TPP+-selective electrode (13), in the standard medium. The concentration of added TPP' was 10 p~, the concentration of red blood cells 0.2%.
Fluorescence Changes of diS-C3-(5)-The fluorescence measurements were performed according to the method of Sims et al. (14). The concentration of cells was 0.5%, that of the dye, 0.33 p~. The total concentration of methanol was 0.17%.
Chemicals-Na-ortho-vanadate was obtained from ICN Pharmaceuticals, Plainview, NY. The solutions were prepared based on a M , of 400. 45CaC12 was obtained from the Radiochemical Centre; nigericin and monensin, from Lilly, valinomycin from Calbiochem; DIDS, Nasalt, and quinidine sulfate from Sigma; Arsenazo I11 from Fluka, Buchs, Switzerland; verapamil from Knoll, Ludwigshafen, West Germany; trifluoperazine from Smith, Kline and French Laboratories, Philadelphia, PA. R 24571 was a kind gift of Dr. Van Belle, Janssen Pharmaceutica, Beerse, Belgium; diS-C3(5) was a kind gift of Dr. A. Waggoner, Amherst College, Amherst, MA.

Ca2+ Uptake by the Red Blood Cells
Red blood cells incubated with 45Ca2' became labeled only to an extent corresponding to their low Ca2+ content, which is in the range of 0.1 to 1.0 pmol/liter of packed cells. This is in agreement with previous observations made in several laboratories (see Ref. 9 for a review). The preincubation of the cell suspension with vanadate, however, dramatically changed this pattern. The cells took up 45Ca2' in a manner dependent on the concentration of added vanadate. In a series of several experiments, the threshold was at about 0.1 m~, and saturation was approached at about 6 m~ (Fig. IA). It was necessary to buffer heavily the vanadate solution to prevent alkaliniza- tion of the medium at high concentrations, since alkalinization per se can stimulate uptake of Ca2+ (15). The rate of Ca2' uptake was slow, and reached a plateau about 60 min after its addition in the experiment shown. The variability range in all experiments that have been carried out was 20 to 80 min. This plateau level of uptake generally remained stable during the 90 to 120 min of incubation. More prolonged incubations led to the decrease of the ,%a2' content in the cells (see below, Fig. ll), a finding which will be discussed later.
In our experiments, we have routinely used 0.5 mM vanadate. This low concentration was chosen to minimize possible pH artifacts, and possible side effects due to the interaction of vanadate with cytoplasmic components (12). At this concentration, the amount of 45Ca2+ taken up was dependent on the concentration of added Ca2+. Signifcant labeling was observed at about 0.1 mM Ca2+ (not shown), whereas saturation was approached at 2.5 m~ (Fig. 1B). The latter concentration was then routinely used. At 25 "C, 0.5 m~ vanadate, and 2.5 mM Ca2+, the steady state uptake varied between 10 and 50 pmol/ liter packed cells. The increase of Ca2' content after vanadate treatment was independently checked by atomic absorption spectrophotometry and the values found were consistent with those obtained with the isotopic method (not shown). No radioactivity was bound by membranes lysed during the washing of 45Ca2' by a hypotonic buffer solution containing 0.5 mM vanadate (not shown). This indicates that the Ca2+ ions retained by the cells after vanadate treatment was in an osmoticlabile pool.
The effect of vanadate cannot be ascribed to the alkalinization of the medium. In our experiments, we have used 50 mM Na-Hepes, pH 7.2, as a buffer, and then 30% suspension of red blood cells contributed additional buffering capacity. The fiial pH in the suspension was lower than 7.2 (between 6.9 and 7.1) and the addition of vanadate did not change it by more than 0.05. As shown in Fig. 2, this minor pH change cannot explain the progressive increase in the uptake of 45Ca2' induced by vanadate.

Mechanism of the Vanadate-induced 45Ca2+ Uptake
As shown by Cantley et al. (12), vanadate inhibits the (Na,K)-ATPase of red blood cells from the inner side of the membrane. Since, in our experiments, we have used similar conditions, it could be expected that the Ca-ATPase would also be inhibited, thus unmasking the uptake of 45Ca2+. The The sidedness of the vanadate action in triggering the uptake of 45Ca2+ was explored with the inhibitor of the anion channel DIDS (Fig. 3). Preincubation of red cells with DIDS prior to the addition of vanadate inhibited substantially Ca2+ uptake, but there was no inhibition when DIDS was added after vanadate (in some experiments, the addition of DIDS after the addition of vanadate even potentiated Ca2+ uptake).
This indicates that vanadate must cross the membrane via the anion channel to stimulate the uptake of Ca2+. Once vanadate is inside the cell, the anion channel is no longer necessary for the induction and maintenance of the Ca2+ uptake. Experiments with radioactive vanadate have showed that red blood cells indeed take up vanadate via the anion channel (12), supporting the tentative conclusion reached here that vanadate induces "uptake" of Ca2+ by inhibiting the outwardly directed Ca2+ pump. The hypothesis has been tested by measuring Ca2+ efflux from red blood cells loaded with 45Ca2' by vanadate.
From the concentration dependence of the effect of vanadate ( Fig. lA), one cannot expect full inhibition of the Ca2+ pump at 0.5 m. The possibility that the exchange of Ca2+ across the membrane, although far from the steady state, leads to the accumulation of Ca2+ was tested, and demonstrated, in the experiments shown in Fig. 4. Red blood cells treated with vanadate and 45Ca2+, and washed free of external radioactivity, released the accumulated 45Ca2+ when incubated under the conditions of the uptake experiments, but without 45Ca2+. The efflux was inhibited by vanadate and abolished by 150 p~ La3+, which completely inhibits the ATP-dependent Ca2+ efflux from resealed Ca2+-loaded red cells (2). On the other hand, the addition of unlabeled Ca2+ to the medium markedly accelerated the efflux of radioactivity, a n effect which was inhibited by high concentration of vanadate, or, much more markedly, by La3+ (Fig. 4B). The stimulation of 45Ca2+ efflux by 40Ca2' approached saturation at about 2.5 mM.
The experiment shows unequivocally that exchange of Ca2+ occurs across the membrane of red blood cells treated with vanadate and Ca2+. The rate of 45Ca2+ efflux in the presence of 40Ca2+ reflects at best the magnitude of the efflux compo-into Erythrocytes nent of the Ca2+ exchange process in vanadate-treated cells.
Due to the isotopic dilution, however, this is true only for the initial rate of the efflux. A quantitative estimation gave the value of about 12 pmol of Ca2' exchanged/liter of packed cells/h (Table I), a limit reflecting the activity of the Ca-ATPase activated by calmodulin under the experimental conditions.
The efflux of 45Ca2+ from 45Ca2+-loaded red blood cells was not influenced by the presence of Naf ions in the medium (not shown). This was tested using choline chloride as a substitute for NaC1, and excludes the possibility that a portion of the 45Ca2+ efflux occurs via a Na+/Ca2' exchange.
The initial rates of Ca2+ uptake averaged 22 pmol/liter/h ( Table I) and decreased with the increasing age of the cells.

FIG. 3.
Effect of DIDS on the vanadate-induced 4sCaz' uptake and on the "Ca2' efflux from 46Ca2+-loaded cells. A, a portion of the suspension (A) was preincubated with 0.58 mM DIDS for 5 min. Then, 0.5 m~ vanadate was added. After 10 min of further incubation, the uptake was started by the addition of 45Ca2+ (2.5 mM). A second portion of the suspension (0) was pretreated for 10 min with vanadate and then for 5 min with DIDS, before the uptake was started with 45Ca2+. Another portion of the suspension (0) was preincubated as the first portion, but dimethyl sulfoxide was used instead of DIDS (final concentration 0.5%). The control suspensions were treated in the same way but without vanadate (closed symbols). B, red blood cells were loaded with %a2+ as described under "Materials and Methods" and preincubated with 0.58 m DIDS for 5 min at 0 "C. Then 2.5 nm %aZ+ was added (time 0) (0). The control (0) was treated in the same way but with 0.5% (v/v) of dimethyl sulfoxide instead added. In experiments carried out in the same day, or within 2 days, the matched values of initial rates of influx and efflux showed that the former exceeds the latter by about 100%, a difference large enough to yield reliable results even in experiments where the conditions could alter one, or both, components of the exchange. It is necessary to mention at this point that, in the case of Ca2+ influx, the initial rates are slightly underestimated, especially in the experiments in which the uptake reached the steady state rapidly (ie. within 40 min) because the first non-zero time value was at 20 to 30 min. The experiment in Fig. 4 implies the necessity of controlling both inward and outward components of the exchange process in experiments in which the 45Ca2+ uptake is altered.
The effects of vanadate and La3+ on the efflux of Ca2+ from 45Ca2+-loaded cells indicate that vanadate (and La3+) inhibit the Ca-ATPase also in its natural environment. If the inhibition of the ATPase is the only explanation for the accumulation of 45Ca2+ in red blood cells, then other inhibitors of the ATPase should induce the same effect. This assumption was supported by experiments in which the anticalmodulin drugs trifluoperazine and R24571 (16, 17) were found to induce 45Ca2+ uptake. The effect of trifluoperazine became clearly evident at about 0.1 m~, a concentration where nonspecific membranolytic effects could play a role (18). R24571, however, induced a very evident uptake of 45Ca2' at about 50 p~, and si&1cantly potentiated the effect of vanadate at lower concentrations, in agreement with the idea that both inhibitors act on the same target (Fig. 5). The potentiation was not due to nonspecific membrane damage, since the effect was abolished by nigericin (Fig. 5) which inhibits the vanadate-induced 45Ca2f uptake (see below).
At this point, then, it can be concluded that vanadate induces the uptake of Ca2+ after penetrating into the cell, and that the uptake of Ca2+ does not require the simultaneous movement of vanadate ions. The potentiation of the effect of vanadate by other inhibitors of the Ca-ATPase suggests that the induction of Ca2+ uptake is not restricted to vanadate.

Factors Influencing the Uptake of 45Ca2' Effect of Divalent Cations and of Znhibitors of Ca2+
Znflux-The inward movement of 45Ca2+ in vanadate-treated red blood cells could occw on a specialized carrier, or by defects in the membrane structure. Several observations support the existence of a carrier-mediated uptake. One is the saturability of the uptake (Fig. lB), which agrees with what observed in ATP-depleted cells (19). This fact cannot be attributed to the activation of the Ca2'-ATPase by calmodulin due to the increased intracellular Ca2+ level, because the Ca2+-Ca2+ exchange in Ca2+-loaded cells was also saturable with in the same concentration range. The concept of a Ca2+ carrier is supported also by the finding that the influx of Ca2' was inhibited by several compounds, and appears to be specific for Ca2+.
In experiments not presented here, we have found that the addition of up to 8 mM Sr2' or Ba2+ had only a negligible inhibitory effect on 45Ca2+ uptake. Co2+ ions, however, inhibited strongly (ICw at about 150 p~) (Fig, 6A). Since even at the highest concentration used (1.5 mM) Co2+ did not stimulate the efflux of 45Ca2+ from loaded red cells (Fig. 6B), the inhibition of the 45Ca2+ uptake was not due to the activation of the Ca-ATPase (e.g. by displacement of vanadate). The "Ca2+ antagonist" verapamil (20, 21) inhibited the vanadate-induced 45Ca2+ uptake (Fig. 7) in the range 10 to 100 p~ (IC50 at about 70 p~) .
Even at the highest concentration tested (200 p~) , verapamil did not stimulate the activity of the Ca2' pump (Fig. 7B). Concentrations of verapamil in excess of 0.2 mM (e.g. 0.5 m~) induced less inhibition and very high con-  centrations even increased the uptake above the control (not shown). Nonspecific damages of the membrane at these concentrations have already been described for a number of amphiphilic substances (see Ref. 15 for a review). It is of interest for the discussion below that quinidine, an inhibitor of the Ca2'-sensitive K' channel (22), was found to inhibit the uptake of 4sCa2+ at the concentration usually employed for the inhibition of the channel (not shown).
Effect ofMonovalent Ions-Since the uptake of 45Ca2' must be electrically compensated by the movement of other ions, the lack of suitable co-or counter-ions should restrict the movement of Ca2' and thus lead to inhibition of the uptake. Because of the unusual permeability properties of the red blood cell membrane for monovalent ions one could predict a role of C1-and/or H' (OH-) as charge-compensating species if Ca2' enters the cell by means of simple diffusion or of an uniport-type carrier. The substitution of C1"containing salts with isotonic sucrose in the medium had, however, no inhibitory effects. Instead, the uptake of Ca2' was strongly increased (in some experiments up to 10 times) as compared with the standard medium (Fig. 8). The uptake decreased at prolonged times of incubation and was stimulated by DIDS. In the presence of the latter, the decrease at longer incubation times was not seen. The enhancement of the uptake of Ca2+ in the C1"free medium might be due to several factors (e.g. depolarization of the membrane, diminution of the competitive action of Na+, increase of the K' gradient). The possibility of nonspecific membrane leaks was ruled out by measuring the uptake in media in which the cation composition was kept constant but that of C1-was varied, keeping the total anion content constant with Hepes (not shown). Increased uptake was observed also in the medium in which the C1 concentration approached the physiological level. It is clear then, that the possibility that C1-accompanies Ca2' into the cell has no experimental support. When the impermeable anion gluconate replaced Hepes, in experiments similar to that of Fig. 8, a strong inhibition of Ca2' uptake was observed. The dependence of the Ca2' uptake on the gluconate concentration, however, suggests that the inhibition was due to the Ca2+binding properties of gluconate (not shown).
Since C1-and H' (OH-) ions are in Donnan equilibrium in red blood cells (23-25), the removal of C1-should influence the distribution of H' across the red blood cell membrane as well. If in media containing physiological concentration of C1-, the H' gradient is oriented outwardly (24,25), the substitution of C1-for impermeable anions should lead to the diminution of the H+ gradient, and, eventually, to its reversal. Since, under these conditions, a stimulation of the uptake was seen, a role of H' as possible counter-ions for the inward Ca2+ movement appears unlikely. Surprisingly, however, the protonophoric agent FCCP inhibited the uptake of 45Ca2' induced by vanadate (Fig. 9)  vanadate, and 0.4 m~ DIDS dissolved in 0.5% dimethyl sulfoxide; V, 2.5 mM Ca2', 0.5 m~ vanadate, and 0.5% dimethyl sulfoxide. V and V were taken from an independent experiment. iments, FCCP did not change the pH of the suspension even when the buffer was omitted from the medium (not shown).
In other experiments, no changes in the transmembrane movement of the lipid-soluble cations TPP+, and dis-c3(5) were seen after the addition of FCCP (not shown). Furthermore, the inhibition by FCCP was still observed in media containing 100 m~ Na-gluconate instead of NaC1, i.e. under conditions where the direction of H+ is expected to be reversed (24, 25).
The inhibition by FCCP could also be due to the stimulation of the K' efflux since the gradient of K' may be important for the maintenance of Ca' ' influx into vanadate-treated red blood cells (see below). FCCP-induced electrogenic H+ movements may represent a charge-compensating device for the outwardly directed movement of K' , in addition to the efflux of C1-. FCCP indeed induced a small increase of the K+ efflux from vanadate-treated cells, but the effect was minor, and could only account for a marginal portion of the inhibition or the 45Ca2' by FCCP.' At the moment, therefore, no conclusive and complete explanation for the effect of FCCP is possible. Possibly, FCCP acts directly on the Ca2'-transporting system, e.g. by protonation of Ca2' binding sites.

The Transmembrane K' Gradient in the Uptake of 45Ca2+
The possibility that the transmembrane K' gradient is involved in the process of Ca2+ uptake has been investigated. Preincubation of vanadate-treated cell with nigericin or monensin (4.7 and 5.3 p~, respectively) which exchange Na' and K+ for H+ and thus collapse Na+ and K+ gradients (27) strongly inhibited the influx of Ca'+ (Fig. 1OA). The ionophores had no effect on control suspensions not treated with vanadate nor did they change the pH of the suspension. Nigericin did not stimulate the efflux of 45Ca2+ from Ca'+loaded cells (Fig. lOB), but promoted a rapid release of KC ( Fig. 1 0 0 . The inhibition of the Ca'+ influx by nigericin (and monensin) is likely to be due to the collapse of the K' gradient, since the collapse of the Na+ gradient should activate, rather The inhibition of the uptake was estimated assuming proportionality between K' release and inhibition of Ca2+ uptake. The magnitude of the inhibition was estimated by intrapolation between the release of K' in control (0% inhibition) and nigericin-treated cells differed by~20%. than inhibit, the Ca2+ influx, if the latter were linked to Na+ by known transport mechanisms (Na/Ca exchange). The role of H' is made unlikely by the data mentioned above. The role of the K+ gradient is supported by the decreased uptake of Ca' ' in media of increasing K+ concentration under conditions where Na+ was decreased correspondingly, and the concentration of C1-(and thus the osmolarity) was kept constant (Fig.  1lA). The time course of Ca2+ uptake in suspensions exposed to various K' concentrations varied. At low K+, the amount of Ca2+ taken up declined after prolonged incubation, possibly due to the opening of K+ channels by the increased concentration of cell Ca".

K + Movements Induced by the Influx of Ca"
That Caz+ can trigger the loss of K' from erythrocytes has been shown by Gardos in 1958 (28). The incubation of red blood cells with vanadate alone did not induce a significant loss of K+ (Fig. lOB), although the (Na, K)-ATPase is markedly inhibited under these conditions (12). This is in agreement with the extremely low monovalent cation permeability of the red cell membrane (29). Incubation with Ca' ' alone had no effects on the K+ movements, but the presence of both Ca2' and vanadate greatly enhanced the efflux of K' . Halfmaximal release was reached in 15 to 20 min, and the plateau after about 60 min. Since the content of Ca' ' in cells, albeit increasing under these conditions, usually remained in the micromolar range, the affinity of the K' channel for Ca2+ evidently is in this range. This is in agreement with the findings on sealed ghosts (30), but is at variance with data obtained with ionophore-loaded red blood cells, where substantially higher values were found (31).
The efflux of K' was efficiently blocked by DIDS (Fig.  10B). This suggests net efflux of KC1, with chlorides leaving the cell via the anion channel. Since DIDS did not inhibit the influx of 45Caz' (Fig. 3), the Ca2'-activated K+ channel is not directly involved in the inward translocation of Ca". Thus, the decreased Ca2' uptake in high K' media was not due to the inhibition of the Ca"-activated K+ channel (32, 33). DISCUSSION The experiments presented here indicate quite compellingly that the loading of the red cells with 45Ca2' in the presence of vanadate is due to the inhibition of the outwardly directed Ca2' pump. The inhibition results from the interference of vanadate with the Ca2+ pump, and not, as could in principle also be suggested, from the exhaustion of the intracellular ATP store. Indeed, the level of intracellular ATP did not deviate significantly from the starting concentration of about 0.6 mM throughout the entire duration of the Ca2+ uptake experiments.
The uptake of Ca2+ after inhibition of the Ca2+ pumping of the plasma membrane seems to be a widely distributed phenomenon. It can be revealed by the exposure of Ehrlich ascites tumor cells (8) or liver slices (9) to low temperature, of liver slices to anaerobiosis (9), or by ATP depletion of human red blood cells (28). In dog red blood cells, uptake of 45Ca2' can be induced by the removal of external Na' ions ( 3 4 , which indicates the existence of a Na'/Ca2+ exchange system. These observations suggest that, in resting cells, a slow cycling of Ca2+ across the plasma membrane occurs. Since verapamil inhibits the uptake of Ca2' in red blood cells, it seems probable that Ca2+ uptake in erythrocytes and during activation of excitable cells is mediated by the same structural components of the plasma membrane. In the present work, the rate of Ca2+ cycling could be derived from the initial rates of Ca2+ influx and efflux, and from the degree of the inhibition of the Ca2'-ATPase by vanadate. Assuming that saturating concentrations of vanadate inhibit the ATPase 100%, the dependence of Ca2+ uptake on the vanadate concentration suggests that at 0.5 n" vanadate the inhibition approaches 40%. The average data for influx and efflux of Table I suggest a turnover of Ca2+ at 25 "C of between 40 and 200 pmol/liter of packed cells/h. This exceeds the value estimated by Lew and Ferreira (10) from the difference in ATP consumption in iodoacetamide-poisoned red blood cells in the presence and in the absence of Ca". The difference in temperature between our conditions and those of Lew and Ferreira (10) cannot completely explain the difference in the estimated turnover of Ca2+, which differ by a factor of between 3 and 15, assuming a Ca/ATP stoichiometry of 1. Our data suggest a higher utilization of the capacity of the Ca2+ pump and estimate it at about 1%. This low utilization value could explain the requirement for very high concentrations of vanadate to unmask the uptake of 45Ca2+, since in intact cells the Ca2' pump must be inhibited more than by 99%.
The results of Lew and Ferreira (10) on ATP-depleted red cells, and the present results on vanadate-treated cells, have shown that the efflux of 45Ca2+ from previously loaded cells is stimulated by the addition of 40Ca2+. It is clear from the data presented that, in vanadate-treated cells, the entry and exit legs of this Ca2+ exchange are mediated via independent routes, the former involving the verapamil-sensitive component, the latter the Ca2'-pumping ATPase. Possibly, Ca2+ is lost by the cells during the washing procedure, decreasing its concentration below the level at which the pump is saturated. The cold calcium added will penetrate into the cells, restore the original Ca2+ level, and activate the pump. The observation that several inhibitors of the influx of Ca2+ have no effect on the efflux of 45Ca2+ from the loaded cells may be rationalized by recalling that the influx component of the Ca2+-Caz+ exchange process is in excess of the requirements for optimal rates of the Ca2+-Caz+ exchange (see Table I). Thus, its inhibition, even if substantial, may well have no effect on the exchange.
Concerning the route of Ca2+ influx, one possibility is a leak through which Ca2+ is driven by its large transmembrane gradient. However, the previously mentioned observation on the saturability of 45Ca2' and on the inhibition of the uptake by verapamil and Co2+, as well as the experiments of Lew and Ferreira (10,19), favor the possibility of a carrier-mediated transport. Although not all criteria for it were verified, the experimental data obtained would be difficult to interprete in terms of membrane leaks.
The present study has shown that the transmembrane gradient of K+ is important in the maintenance of the uptake of Ca". This is somewhat surprising, since the Ca2+ concentration at the two sides of the membrane differ by a factor of 1000, and that of K+ by a factor of only 20. The concentration (activity) of Ca2' at the membrane surface could, however, be lower than in the bulk aqueous phase, emphasizing the possible role of the cell surface in the translocation of Ca2+ as suggested by Langer (35) for the activation of heart muscle cells.
At the present state of knowledge, it cannot be decided whether the electrical or the concentration components of the K+ gradient play a role in the inward movement of Ca". The membrane of red blood cells in the "resting" state does not discriminate between Na+ and K+ (28) but becomes selectively permeable to KC after the activation of the K+ channel. This could lead to membrane hyperpolarization, and it could drive electrophoretically Ca2+ into the cells. This interpretation would explain the experiments with nigericin-treated red blood cells, and those on the uptake of 45Ca2+ in media with varying K' concentrations, and is supported by the observation made in our laboratory that valinomycin stimulates slightly the uptake of 45Ca2+. However, following this interpretation, the depolarization of the membrane by substitution of chlorides in the medium with impermeable anions, or sucrose, should also lead to the inhibition of the 45Ca2+ uptake, a prediction which was not supported by the experimental results (see Fig. 8). The postulate of an electrogenic Ca2+/K+ antiporter, analogous to the electrogenic Ca2+/H' antiporter suggested by Hinnen et al. (8) could, however, accommodate all experimental results obtained.
Irrespective of the mechanism of the Ca2+/K' interaction in the process of the 45Ca2+ uptake, there is a link between the difference of the K' levels at the two sides of the membrane and the extent and the velocity of the Ca2+ uptake (Fig. 11). The activation of the Ca2+-sensitive KC channel will lead to the dissipation, or to the decrease, of the K' gradient, decreasing the uptake of 45Ca2+ and contributing to the establishment of its low steady state intracellular level. The decreased levels of cell Ca2+ during the later stages of the uptake in low Kf media, as well as the observation that this decrease is not seen in media of higher K+ content, support the involvement of the Ca2'-sensitive K+ channel in the maintenance of the steady state level of intracellular Ca2+, and, more in general, the concept that the transmembrane K+ gradient is an important ingredient in determining the balance of Ca2+ between cells and medium. The experiment with DIDS-treated red blood cells, in which the Ca2+-sensitive K' channel is blocked (Fig.  lOC), offer additional support for this suggestion.
The role of the K+-dependent Ca2+ influx system, and of the Ca2+-activated K+ channel as regulators of the CaZ+ homeostasis is probably not dramatically important in red blood cells where the cycling of Ca2+ occurs at a slow rate. However, cells with high Ca2+-cycling rates could effectively modulate large fluctuations in intracellular Ca2+ without the significant losses of K' observed in red blood cells. The main role of the Ca2+-sensitive K+ channel might just be to defend the intracellular milieu from uncontrolled Ca2+ increases.