Effects of Low Extracellular Sodium on Cytosolic Ionized Calcium Na+-Ca2+ EXCHANGE AS A MAJOR CALCIUM INFLUX PATHWAY IN KIDNEY CELLS*

The effects of extracellular Na’ (Naz) on cytosolic ionized calcium (Ca?+) and on calcium and sodium fluxes were measured in monkey kidney cells (LLC- MK2). Ca? was measured with aequorin and the ion fluxes with 45Ca and 22Na. Na+-free media rapidly increased Cat’ from 60 to a maximum of about 700 n~ in 2-3 min. After the peak, Ca?’ declined and reached a plateau of about twice the resting Ca?+. The peak Ca;+ was inversely proportional to Na: and directly proportional to the extracellular calcium concentration (Ca:+). On the other hand, a pH of 6.8 reduced and Ca:+ substitution with Sr2+ completely blocked the Caf+ response to low Naz. A Na+-free medium stimu- lated calcium efflux from the cells 4-5-fold, a response which was abolished in the absence of extracellular Ca2+. Na+-free media also stimulated calcium influx and sodium efflux. The cell calcium content, however, was not increased. These results indicate that removal of extracellular Na+ increases Ca? by stimulating calcium ,influx and not by inhibiting calcium efflux; the increased calcium influx takes place on the Na+-Ca2+ antiporter operating in the reverse mode in exchange for sodium efflux. The increased calcium efflux occurs as a consequence of the rise in Caf+ and presumably takes place on the (Ca2+-Mgz’) ATPase-dependent calcium pump. Measurement of Cytosolic Ionized Calcium-The cells were loaded with the calcium-sensitive photoprotein aequorin by a modification of the scrape-loading method of McNeil and Taylor (21, 22). They were held and perfused in a matrix of fiberglass in a cuvette placed in an aequorin photometer (22). The perfusate was a Krebs-Henseleit bicarbonate buffer (KHB) comprising (in mM) NaCl 120, KC1 5, CaClz 1.3, MgS04 1.0, KH2P04 1.0, NaHC03 24, and glucose 5, equilibrated with a gas phase of 95% 02, 5% COz. The light emitted by the Caz+-aequorin reaction was collected by a photomultiplier tube (EM1 9635A), and the current was recorded on a strip chart recorder. All experiments were performed at 25 "C. The conversion of the aequorin light signal measured in nA into ionized calcium concentration was done according to Allen and Blinks (23). At the end of each experiment, the amount of aequorin incorporated by the cells was assessed by exposing the cells to 2% Triton X-100, in the presence of 140 mM KC1,l mM MgC12, 13 mM Hepes (pH 7.4), and 10 mM CaClz to ignite all the intracellular aequorin. The light signal emitted in these conditions (Lmax) was 5-6 orders of magnitude greater than the current measured during the experiment (2). The fractional lumines- cence, i.e. the current signal (L) measured at every time point,

evoked by a decreased AMNa+ has been generally interpreted as resulting from a decreased calcium efflux on the Na+-Ca2+ exchanger (4,12-15) and not from an increased calcium influx on the antiporter operating in the reverse mode.
The distinction between the two possible explanations is not trivial. Although an inhibition of calcium efflux would have the same effect on Cap+ as a stimulation of calcium influx, the implication of these two opposite interpretations would be fundamentally different. If one assumes that calcium efflux on the Na+-Ca2+ antiporter is depressed when ApNa+ is decreased, the resulting rise in Cap+ implies that Na+-Ca2+ exchange is the most important and determinant controller of Ca?
and that the plasmalemma (Ca2+-Mg2f)-ATPasedependent calcium pump and other intracellular calcium transport mechanisms cannot compensate for the inhibition of the antiporter (12-15). On the other hand, if the decrease in ApNa+ was to increase Ca?' by stimulating calcium influx on the antiporter operating in the reverse mode, one could speculate that the Na+-Ca2+ antiporter functions as an important calcium influx pathway competing with calcium channels (11, 19) and that the (Ca2+-Mg2+)-ATPase-dependent calcium pump is the main calcium efflux pathway (5,7,(16)(17)(18). The latter hypothesis may account in part for the fact that in some tissues the absence of extracellular Na+ induces no measurable increase in cell calcium (18, 20); indeed, the increased calcium influx on the antiporter would be immediately followed by an increased calcium efflux on the (Ca2+-Mg+)-ATPase-dependent calcium pump stimulated by the elevated CaH+.
These two hypotheses, inhibition of the forward mode or stimulation of the reverse mode of Na*-Ca2+ exchange in low N&, can be easily tested. In the first case, calcium efflux should be depressed and the tissue should gain calcium. In the second case, calcium influx and efflux should be both increased and Na efflux should be enhanced but only in the presence of extracellular calcium. This series of experiments was performed in isolated kidney cells to test these different predictions. sera were purchased from Microbiological Associates (Walkersville, MD). MgATP, EGTA, EDTA, Hepes, and Triton X-100 were from Sigma; 45CaC12, 22NaC1, and scintillator Formula 963 were from New England Nuclear. Ultrapure grades of MgClz and CaClZ were purchased from Johnson Matthey Ltd., United Kingdom and BDH Chemicals, Ltd., United Kingdom, respectively.
Measurement of Cytosolic Ionized Calcium-The cells were loaded with the calcium-sensitive photoprotein aequorin by a modification of the scrape-loading method of McNeil and Taylor (21,22). They were held and perfused in a matrix of fiberglass in a cuvette placed in an aequorin photometer (22). The perfusate was a Krebs-Henseleit bicarbonate buffer (KHB) comprising (in mM) NaCl 120, KC1 5, CaClz 1.3, MgS04 1.0, KH2P04 1.0, NaHC03 24, and glucose 5, equilibrated with a gas phase of 95% 02, 5% COz. The light emitted by the Caz+-aequorin reaction was collected by a photomultiplier tube (EM1 9635A), and the current was recorded on a strip chart recorder. All experiments were performed at 25 "C. The conversion of the aequorin light signal measured in nA into ionized calcium concentration was done according to Allen and Blinks (23). At the end of each experiment, the amount of aequorin incorporated by the cells was assessed by exposing the cells to 2% Triton X-100, in the presence of 140 mM KC1,l mM MgC12, 13 mM Hepes (pH 7.4), and 10 mM CaClz to ignite all the intracellular aequorin. The light signal emitted in these conditions (Lmax) was 5-6 orders of magnitude greater than the current measured during the experiment (2). The fractional luminescence, i.e. the current signal ( L ) measured at every time point, divided by the current signal obtained after Triton-X ( L a x ) was interpolated on a standard curve relating the fractional luminescence of a constant amount of the same lot of aequorin reacting with standard calcium concentrations at the same temperature (25 "C) and in a solution whose ionic composition was assumed to be that of the intracellular milieu (140 mM KCl, 1 mM MgC12, pH 7.4).
Measurement of Cell Calcium-To determine the effects of extracellular Na+ on cell calcium, the cells were first incubated in KHB as a suspension gently stirred with a magnetic stirrer at 25 "C for 60 min. The cells were then centrifuged at 200 X g and resuspended in a Na+-free KHB with TMA as the substituting ion. After 60 min in Na+-free media, the cell samples were added to 40 ml of GKN (a phosphate-buffered saline solution containing in mM: KzHPO4 0.15, KHzP04 0.51, NaCl 135, KC1 4, glucose ll), centrifuged, and resuspended in 3 ml of deionized water. The cells were homogenized with an ultrasonic probe. Calcium was assayed by the method of Borle and Briggs (24) and cell protein by the method of Lowry et ul. (25).
Measurement of Fructionul Calcium Efflux-Kidney cells were first labeled for 60 min with 45Ca, and the isotope desaturation was performed by constant perfusion at 25 "C according to the method published by Borle et al. (26). The cells were placed into a loosely packed matrix of fiberglass over a 10-pm mesh nylon net placed at the outlet of the chamber and were perfused at a rate of 0.6 ml/min. The dead space of the system from medium reservoir to collection vials was 1.5 ml. The effluent was collected in scintillation minivials at intervals ranging from 30 s to 5 min, and the radioactivity was assayed by scintillation spectrophotometry. At the end of the experiment, the cells were removed from the chamber, solubilized in 0.5 N NaOH, and aliquots were analyzed for radioactivity and protein. Data are expressed as the ffiCa fractional efflux ratio. The fractional efflux is the radioactivity released per unit of time expressed as the percentage of the mean cell radioactivity during the collection period. The fractional efflux ratio is the fractional efflux of the experimental group divided by the fractional efflux of controls perfused concurrently.
Measurement of Fractional Sodium Efflux-The same method and calculations as those just described for calcium were used to determine "Na fractional efflux. &Ca U p t~k e -~~C a uptake studies were performed according to Uchikawa and Borle (27). Kidney cells were harvested from the culture flasks with a rubber policeman and placed in suspension in KHB at 25°C. After an equilibration period of 60 min, 45Ca was added, and 1-ml aliquots were taken from 2-60 min, centrifuged through a 40-ml column of ice-cold GKN, and the cell pellet was homogenized with an ultrasonic probe in 3 ml of deionized water. &Ca, 40Ca, and protein were determined as described above. To study the effects of Na+-free media, the cells were suspended in KHB in which Na+ was substituted by TMA. Extracellular Nu+ Substitution- Table I shows the composition of the various incubation media used in these experiments. Nit: was either reduced to 24 mM and replaced with TMA, lithium, and choline   or totally replaced by the same substituting ions. In Na+-free media, the pH of 7.4 was always maintained by 24 mM choline bicarbonate in equilibrium with a gas phase of 5% COZ, 95% 0 2 . In 45Ca and "Na fractional efflux experiments the pH of 7.4 was maintained by 20 mM Hepes and N$ was totally replaced by 144 mM TMA.

Effects of Lowering Extracellular Sodium on Cytosolic Ion-
ized Calcium-The lowering or the removal of extracellular sodium and its substitution by TMA, lithium, or choline resulted in an immediate rise in the cytosolic ionized calcium of cultured monkey kidney cells. This confirms the results obtained in perfused proximal tubules of Necturus kidney by Lorenzen et al. and Windhager (13, 15) who measured Ca?+ with Ca2+-selective microelectrodes. Fig. 1 shows a composite drawing derived from 5 separate experiments in Na+-free media in which Na; was replaced by TMA. The inset presents a typical recording of such an experiment. Ca? rose from a basal concentration of 65 5 19 to 696 -C 78 nM in 90 s. For the next 10 min, Ca?+ declined to a quasi-plateau at approximately 171 f 39 nM. When N d was restored to its normal concentration of 144 mM, Ca? returned to the basal levels within 5 min. In spite of this rise in Ca?, the total cell calcium measured throughout the period of incubation in Na+-free media up to 60 min did not increase. In fact cell calcium decreased slightly from 8.35 f 1.21 nmol/mg of protein (n = 8) to 7.15 -t 0.43 nmol/mg of protein (n = 16). The peak level of Cat+ depended both on the substituting ions and on the NG concentration. Fig. 2 shows that total NG substitution with choline and lithium also increased Ca?+ but less than with TMA.    the rate of rise. In most cases, a new steady state was reached, and Ca:+ was maintained at a plateau 2-3-fold above the basal concentration ( Table 11). The peak Ca:+ and its rate of increase were inversely proportional to the N d concentration. Fig. 3 shows the relation between Ca? and N d , when TMA was used as the substituting ion.

Effects of substitution of extracellular Nu+ by TMA, choline, and lithium on the cytosolic ionized calcium of kidney cells
Effects of Lowering N d on Calcium Efflux-The rise in cytosolic ionized calcium when N& is decreased and the inverse relation between N d and Ca? strongly suggest that Na+-Caz+ ~ exchange may be implicated. The question is whether Ca?' rises because of a decreased calcium efflux or because of an increased calcium influx, that is whether the Na+-Ca2+ exchanger operates in the forward or in the reverse mode. To distinguish between these two possibilities, calcium

500-
efflux was measured by the 45Ca fractional efflux ratio technique. Fig. 4A shows that when Nh+ was totally replaced by 144 mM TMA, calcium efflux immediately increased 4.5-fold. Calcium efflux rose for 3 min and then declined to a plateau at about twice the basal level. When N%+ was reintroduced in the perfusate, calcium efflux returned to control levels in 5-10 min. This rapid and dramatic increase in calcium efflux excludes an inhibition of calcium efflux as a cause of the rise in Cap+ as suggested by many authors. On the contrary, these results suggest that calcium efflux is increased because of the rise in Cap+. Indeed, the magnitude of the stimulation of calcium efflux (4.5-fold), the duration of its rise (3 min), its decline, and finally its plateau at twice the control levels closely match the magnitude and the time course of the rise in Caf+. A smaller Na,+ reduction to 100 mM was also followed by an increase in fractional calcium efflux from 1.1 f 0.08 to 1.37 f 0.13 ( n = 6).

Na+-Ca2+
Since the rise in Ca;+ induced by low N d is not caused by a decreased calcium efflux, the only possible explanation is that calcium enters the cytosol from the extracellular fluids or, less likely, from some intracellular pool of sequestered calcium. Supporting the former idea, Fig. 4B shows that in the absence of extracellular calcium, perfusion with a Na+free medium did not increase calcium efflux. This suggests that when calcium influx is abolished, the rise in Ca? and the secondary increase in calcium efflux do not take place. These results also indicate that mobilization of calcium from intracellular pools does not contribute much, if anything, to the rise in Caf+ evoked by low N d .
Effects of Substitution of Extracellular Calcium with Strontium-To further test the idea that a low N%+ triggers an influx of calcium into the cell, the influence of strontium and of various extracellular concentrations of calcium was investigated. Sr2+ activates aequorin only weakly, so that even if its enters the cell through the various calcium influx pathways, its presence in the cytosol would not enhance the aequorin luminescence. Moreover, Sr2+ is an inhibitor of Na+-Ca" exchange (28). Fig. 5 shows a typical experiment in which kidney cells were exposed sequentially to 1) a Na+-free medium containing 1.3 mM Ca2+ with TMA as substituting ion (120 mM TMA and 24 mM choline bicarbonate), 2) a Na+free medium in which all the calcium was replaced by 1.3 mM Sr", and 3) a Na+-free medium containing again 1.3 mM Ca2+. First, N d substitution with TMA produced the typical rise in Ca:+ in the presence of extracellular calcium. When extracellular calcium was replaced by strontium, the Na+-free medium produced practically no change in luminescence, indicating that the main source of calcium must be extracellular and that intracellular pools of sequestered calcium contribute essentially nothing to the rise in Ca:+ induced by TMA. When Cd+ was reintroduced for the third stimulation, the peak Cap+ and the plateau were depressed. A possible explanation is that a significant influx of strontium during the second stimulation has increased the cytosolic concentration of SI ?+ which competes with Ca2+ for binding sites on the aequorin molecules. calcium as suggested by the previous experiments, the peak Ca?' should be dependent upon the extracellular calcium concentration. Therefore, we studied the effects of Cg' from 30 W M to 1.3 mM on the peak Ca?+ induced by the total substitution of N&+ by 120 mM TMA and 24 mM choline bicarbonate. In these experiments, the cells were continuously exposed to 2 pg of FCCP/ml to eliminate or minimize the intracellular buffering of Ca?' by mitochondria (see below). Fig. 6 shows a typical recording of such a test, and Fig. 7 the relation between the peak Ca?' and the extracellular calcium concentration in 9 different experiments. The results clearly show that the rise in Ca:' evoked by a Na+-free medium is 1onA L The 4 stimulations recorded in the same cell preparation at 30-min intervals are representative of 3 separate experiments. The periods of Na+-free perfusion are indicated by the horizontal lines below the tracings along with the extracellular Ca" concentration prevailing during each stimulation. These experiments were performed in the presence of 2 pg/ml FCCP to prevent the progressive decay of the responses observed in sequential stimulations and caused by mitochondrial buffering (Fig. 12). totally dependent on extracellular calcium and support the postulate that it is caused by a stimulation of calcium influx. Effect of Lowering Nu,+ on Calcium Influx-To test whether the substitution of extracellular Na' with TMA increases calcium influx, the uptake of 45Ca by kidney cells in suspension was studied in control media (KHB) and in media where N$ was totally replaced by 120 mM TMA and 24 mM choline bicarbonate. Fig. 8 shows that 45Ca uptake was significantly increased in Na+-free media. The early time points of the 45Ca uptake curve represent both calcium binding to the cell surface and unidirectional calcium influx, so that it is impossible to calculate exactly the magnitude of the increase in calcium influx. If one assumes that calcium binding is not increased by TMA substitution, calcium influx can be estimated to be raised 4-6-fold. In addition, 2 mM Mn2+ which is an inhibitor of Na+-Ca2+ exchange, markedly reduced the increased calcium influx evoked by a Na+-free medium.

Effects of Different Extracellular Calcium Concentrations-
Effect of Lowering Nu: on Sodium Efflux-If the observed rise in Caf' occurs by Na+-Ca2+ exchange one would predict that the increased calcium influx on the Na+-Ca2+ antiporter should be accompanied by an simultaneous efflux of sodium that would be dependent upon the presence of extracellular calcium. To test this prediction, the effects of sodium substitution with TMA on 22Na efflux were studied in the presence and in the absence of extracellular calcium. Fig. 9A shows that the "Na fractional efflux was increased 40% in Na+-free media. 22Na efflux reached a peak in 3.5 min and then declined to a plateau about 20% above basal levels and returned to the control rate within 2 min when sodium was reintroduced in the perfusate bathing the cells. The time course of the increased sodium efflux and of its various phases closely matched that of the rise in Ca?' and that of calcium efflux. Fig. 9B shows that in the absence of extracellular calcium, sodium efflux was not stimulated by lowering Na;. These results strongly support the hypothesis that Na+-free media stimulates calcium influx and sodium efflux on the Na'-Ca2+ antiporter operating in the reverse mode. The increased so-B , Na,=o,cao=o

A .
active transport would be expected to fall (13).
Effect of Ouabain on Caf' in Nu+-free Media-One can postulate that the fall in Nat caused by Na+ efflux could be calcium influx on the antiporter may only occur in exchange for sodium efflux, the resulting rise in Caf+ would be limited by the availability of intracellular sodium. Once Nat is depleted, calcium influx would decrease, allowing the plasma-lemma1 calcium pump and intracellular buffering of Ca:+ by mitochondria to lower Ca?' to a new steady state plateau. If Ca?' and its duration will depend upon Nat when the cells are exposed to Na+-free media. To test this hypothesis the In the first control test with TMA and choline, Caf+ rose to 421 nM and followed the typical time course of peak, decline, plateau, and return to control levels. In the presence of ouabain, the response became progressively more marked and prolonged. The second and third stimulation of Ca?' in Na+free medium was obtained 10 and 60 min after the cells were exposed to the glycoside. The initial peak Ca?' was increased by ouabain from 421 nM in the first trace to 471 and 504 nM, 10 and 60 min after ouabain. The rate of decline was also markedly slowed down from 58 nM/min in the control test to 17.1 nM/min and 11.2 nM/min in the second and third test. Finally the Ca?+ plateau was also increased from 143 to 163 and 363 nM in the 3 responses, respectively. Such results can be interpreted as an enhancement of calcium influx on the Na+-Caz+ exchanger by progressively higher concentrations of intracellular Na+ caused by the inhibition of the Na+ pump by ouabain. They support the idea that the decline in Ca?' after the initial peak may be caused by a depletion of intracellular sodium as sodium efflux on the Na+-Ca2+ antiport is stimulated. Effects of pH on Ca?+ in Na+-free Media-Na+-Ca2+ ex-  9. Effect of the total replacement of extracellular Na+ by TMA on sodium efflux from kidney cells. Na+ fractional efflux measured with "Na represents the experimental to control ratio. A , effect of perfusion with Na+-free media in the presence of 1.3 mM CaZ+. B, effect of perfusion with Na+ free media in the total absence of extracellular Ca2+. C, effect of perfusion with Na+-free media in the total absence of extracellular K+. The periods of Na+free perfusion are indicated by the horizontal lines below the graphs.

Minutes
Each data point represents the mean +-S.E. of 3 experiments.

Na+-Ca2+ Exchange and Ca?' in Kidney Cells
change is very pH sensitive in cardiac sarcolemmal vesicles and was shown to be a sigmoidal function of pH; it is inhibited at pH 6 and stimulated at pH 9 (28). If the rise in Cap+ induced by Na+-free media was caused by calcium influx on the Na+-Ca2+ antiporter, it should also be influenced by pH. To test this hypothesis, the rise in Caf+ induced by N d substitution with TMA was studied at pH 6.8, 7.4, and 7.8 by increasing or decreasing the pCOz of the gas phase equilibrating the KHB buffer. Fig. 11 shows a typical recording representative of 3 separate experiments, During the first stimulation produced by TMA at pH 7.4, Caf' rose from 26 to 380 nM in 1.3 min and then fell to a plateau of 150 nM after 10 min. When the pH was reduced to 6.8, TMA substitution increased Caf+ from 36-241 nM in 1.2 min, and the plateau of 37 nM was reached in 12 min. When the pH was raised to 7.8, the rise in Caf' induced by TMA reached 592 nM in 2.2 min, and a plateau of 145 nM was reached after 36 min. It appears, therefore, that the calcium influx pathway responsible for the rise in Caf' is markedly pH sensitive. It is inhibited at low pH and enhanced at high pH.

Effects of Repeated Stimulations and of Mitochondrial
Inhibitors-The availability of intracellular sodium may not be the only condition that affects the peak Cap+ and the time course of its fall evoked by Na+-free media. Fig. 12 shows that the peak Cap+ response to repeated exposures to Na+-free media grew progressively smaller. In some instances the peak disappeared and only the plateau phase of Caf+ elevation was obtained. This could be caused by a depletion of intracellular sodium (13) or alternatively by an increased buffering action of intracellular organelles, particularly by mitochondria. To estimate the contribution of mitochondrial buffering to the decline in peak Cap+ levels, we tested two mitochondrial inhibitors, FCCP and NaCN, with choline as Na: substitute.  when extracellular sodium was totally replaced by choline, Caf' increased to 460 nM; a second stimulation 9 min later increased it to 350 nM; after 1-h exposure to FCCP a third stimulation increased Caf' to the same level (data not shown). Fig. 14 shows another recording representative of 2 experi-confirming previous results obtained in rat and rabbit kidney cells (18. 30). L NacN FIG. 14. Effect of NaCN on the rise in Cai evoked by the total replacement of extracellular Na+ with choline. The first response was induced by choline after a series of stimulations and was markedly blunted. When 2 mM NaCN was added to the perfusate, a second rise in Caf' occurred. The last three rises in Cai were produced by total substitution of Na with choline in the presence of NaCN. The periods of Na+-free perfusion are indicated by the horizontal lines below the graphs. ments performed with 2 mM NaCN with choline as the substituting ion. The first stimulation of Fig. 14 was obtained after a series of N$-free exposures and was markedly blunted. Ca:+ rose from 10 to 75 nM. After NaCN was added to the perfusate, Ca?+ transiently increased to 118 nM. In the presence of cyanide, N$ substitution by choline increased Ca? to 190 nM. In two subsequent stimulations after 60 and 90 min, CaT+ rose to 104 and 125 nM, respectively. This series of experiments suggests that mitochondrial buffering of cytosolic calcium contributes to the progressive decrease in the peak CaT+. When this buffering is eliminated or blunted by FCCP or NaCN the high levels of Ca? evoked by Na+-free media are restored.

DISCUSSION
The physiological and functional importance of Na+-Ca2+ exchange in transporting epithelia is still controversial. On one hand, the existence of a Na+-Ca2+ antiporter in basolateral membrane vesicles isolated from renal, intestinal, and toad bladder epithelial cells has been conclusively demonstrated (6,10,28,29). On the other hand, the functional importance of Na+-Ca2+ exchange as a major pathway for the active transport of calcium out of intact epithelial cells has been challenged on several grounds by many investigators (5, 7, 16-18): calcium transport on the ATP-dependent calcium pump exceeds that of the Na+-Ca2+ exchanger by a factor of 4-5-fold (7); the effects of Na+ on calcium transport are claimed to be competitive or allosteric and not due to Na+-Ca" countertransport (16, 17); and there is no gain in tissue calcium in the absence of extracellular Na+ (18, 30). The conflict may arise from the common assumption that Na+-Ca" exchange is operating in the forward mode as a calcium efflux pathway. If one assumes that the Na+-Ca2+ antiport is not functioning when ApNa+ and ApCa2+ are exactly balanced or that it operates in the reverse mode as a calcium influx pathway, the apparent contradiction may no longer exist.
Our results show that decreasing ApNa+ by lowering extracellular Na+ to 100, 24, or 0 mM immediately increased Ca%+. The peak Caf+ was inversely proportional to the concentration of extracellular Na+ and directly proportional to the concentration of extracellular Ca2+. Our measurements of calcium and sodium fluxes further showed that decreasing ApNa+ by totally replacing extracellular Na+ with TMA stimulated calcium efflux from the cells, a response abolished in Ca2+-free media; it significantly increased calcium influx. Finally the cell calcium content was not increased in Na+-free media, \ ,~ I From these observations, we can eliminate several explanations for the rise in CaT+ commonly offered by many investigators: the rise in Ca:+ induced by a low extracellular Na+ is not caused by an inhibition of calcium efflux since calcium efflux increased 4-fold (4, [10][11][12][13][14][15]29). The rise in Ca?' is not caused by a mobilization of calcium from intracellular pools of sequestered calcium as proposed by Mandel and Murphy (30); indeed, when extracellular Ca2+ is substituted by S?' , there is no longer any increase in Ca?+, and after the cells are pretreated with mitochondrial inhibitors that release mitochondrial calcium, the effects of low Na,+ are enhanced and not reduced (Figs. 13 and 14). Finally, we can eliminate the possibility that the increased sodium efflux is taking place on the Na+-K+ pump; it is not abolished by the elimination of extracellular K+ and, on theoretical ground, one can assume that intracellular Na+ would be reduced in Na+-free media (13) and that the Na+-K+ pump would be inhibited by the high cytosolic Ca".
All the evidence tends to support the view that the rise in Ca:+ induced by low N d is caused by an increased calcium influx and that this increased calcium influx is taking place on the Na+-Ca2+ antiporter: 1) the increased calcium efflux is accompanied by an increased Na efflux; 2) the increased calcium influx is inhibited by Mn2+ which is a recognized blocker of Na+-Ca2+ exchange (28); 3) the resulting increase in Caf' is blocked by Sr2+; and 4) like the Na+-Ca2+ exchange mechanism, it is inhibited by low pH and enhanced by high p H finally, 5 ) the increased Na efflux is abolished in Ca2+free media.
The following sequence of events could explain the effects of extracellular Na+ substitution. 1) Lowering extracellular Na+ suddenly reduces ApNa+ to the point where the driving force for calcium influx ApCa2+ greatly exceeds that for sodium influx. 2) A sudden and massive influx of calcium takes place on the Na+-Ca2+ antiporter in exchange for sodium efflux.
3) The increased calcium influx floods the cytosol and increases Ca?' . 4) The rise in Ca?' immediately stimulates the (Ca'+-M%+)-ATPase-dependent calcium pump and calcium efflux is markedly stimulated; the close interrelation among calcium influx, sodium efflux, the rise in Ca?+, and calcium efflux is supported by their similar time course characterized by the 3 phases of peak, decline, and plateau. 5 ) After reaching a peak, CaT+ declines to a new plateau, at about twice the control levels of Ca?'; the fall in Ca?+ may be caused by at least 3 processes: (a) a decline in calcium influx limited by the decrease in Nar that follows the sudden rise in Na+ efflux and the absence of Na+ influx; (b) a rise in calcium efflux on the calcium pump; and (c) intracellular buffering particularly by mitochondria. Indeed, pretreatment with ouabain which presumably increases Nai or with mitochondrial inhibitors enhances the peak Cai and delays the decline. 6) The plateau phase of elevated Cai+ is determined by the combined actions of an increased calcium influx, efflux, and mitochondrial buffering, and it is presumably affected by the intracellular concentration of Na+, of ATP, and by other unknown factors. Initially, however, mitochondrial calcium buffering may not be an important determinant of Cat' in Na+-free media, and calcium efflux on the calcium pump constitutes the main compensation to the elevatd Ca?+; indeed the cell calcium content does not increase, and the effects of mitochondrial inhibitors are seen only after several stimulations when the response to low N$ is declining.
Our data strongly support the idea that, in Na+-free media, the Na+-Ca2+ antiporter operates in the reverse mode as a Relation between the net driving force for calcium influx on the Na+-Ca2+ antiporter ApCa2+-3ApNa+ and Caf' obtained at different N d concentrations (Fig. 3). The relevant ApCa2+-3ApNa+ at a particular N& was calculated as shown in Fig.  15. calcium influx pathway. The question remains whether in physiological conditions, the antiporter operates in the reverse or forward mode, in other words whether it normally behaves as a calcium influx pathway or as a calcium efflux pathway. That will depend of course on the stoichiometry of the exchange mechanism, on the intracellular Na+ and Ca2+ concentration, and on the cell membrane potential difference A+, which influence ApCa2+ and ApNa+. Most investigators found that the stoichiometry of the Na+-Ca2+ antiporter measured in isolated plasma membrane vesicles or in intact cells is 3 Na+ for 1 Ca2+ in both forward and reverse modes (9-11, 28, 31-35). The intracellular Na+ concentration of kidney cells reported in the literature ranges between 12 and 35 mM (13,15,(36)(37)(38) with an average of about 15 mM, and the membrane potential difference of renal proximal tubule cells ranges between -48 and -60 mV (13-15, 36, 38, 39). Accepting a cytosolic ionized calcium of M, an extracellular concentration of Na+ and of Ca2+ of 140 and 1.0 mM, respectively, a Na? of 14 mM and A+ of -60 mV one can calculate ApCa2+ and ApNa+ in normal physiological conditions.' ApCa2+ is found to be 34.56 kJ/mol and ApNa+ 11.52 kJ/mol. Assuming a stoichiometry of 3 Na+/l Ca2+, the two driving forces are exactly equal: ApCa'+ = 3ApNa+ = 34.56 kJ/mol. We can conclude that, unless Na: is significantly less than 14 mM or A+ significantly greater than -60 mV the driving force from the sodium electrochemical potential is not great enough to energize calcium efflux in kidney cells. If in normal physiological conditions the two driving forces were exactly balanced, Na+-Ca2+ exchange would not be functioning. On the other hand, one can speculate that if the intracellular Na+ concentration was greater than 14 mM and A+ less than -60 mV, which is more likely according to all the values published in the literature, the driving force for calcium influx would exceed that for sodium influx (ApCa2+ > 3ApNa+) and the Na+-Ca2+ antiporter would operate as a calcium influx pathway. The reverse mode of Na+-Ca'+ exchange could be the usual and physiological mode. The net driving force for calcium influx on the Na+-Ca2+ antiporter would be ApCa2+ -3ApNa+ which would vary with the extracellular Na+ concentration as shown in Fig. 15. If in physiological conditions the Na+-Ca'+ antiporter was working in the forward mode as a calcium efflux pathway, one would predict that a small decrease in ApNa+ would reduce calcium efflux while a large decrease in ApNa+ would increase calcium influx as the antiporter shifts in the reverse mode. Since a small reduction in N d to 100 mM (which slightly lowers ApNa" by -0.84 kJ/ mol) results in a 25% stimulation in calcium efflux, it is likely that under physiological conditions the antiporter is not functioning (ApCa2+ = 3ApNa+) or is operating as a calcium influx pathway (ApCa" < 3ApNa+). If so, the peak Ca?+ obtained when N d is lowered should be directly proportional to the influx of calcium on the Na+-Ca2+ antiporter and, therefore, directly proportional to its driving force. Fig. 16 shows that there is an excellent correlation between the peak CaP+ measured at various Na,+ in our experiments and the net driving force ApCa2+ -3ApNa+ calculated for the appropriate NG.
We conclude that in kidney cells, the Na+-Ca2+ antiporter can operate as a calcium influx pathway. The magnitude of the calcium influx depends on the relative electrochemical potential gradients for Na+ and Ca'+. When ApNa+ is decreased by lowering N d or by increasing Nat, calcium influx on the antiporter is markedly stimulated, and it increases CaP+ to an extent that is directly related to the driving force ApCa2+ -3ApNa+. Calcium efflux on the (Ca2+-Mg2+)-ATPase-dependent calcium pump is secondarily enhanced by the high CaT+. From these considerations, we can speculate that the only transport mechanism available for calcium efflux in kidney cells is the (Ca2+-M?)-ATPase-dependent calcium pump. Only in exceptional and unlikely conditions, such as hyperpolarization (A+ > -60 mV) or very low intracellular Na+ (Na: < 10 mM), could the Na+-Ca2+ exchange operate as a calcium efflux pathway. These results and their interpretation may help solve some of the conflicts existing between the proponents of Na+-Ca2+ exchange as an important transport The electrochemical potential gradient for Caz+ is ApCa2+ = zF (E, -Ec,z+) where z = ionic valence, F = Faraday constant 96,500 coulombs/eq, and E , = membrane potential difference A$. Ea2+ is the Nernst equilibrium potential (RT/zF )In (Ca&aJ where R = gas constant 8.3 J/mol and T = the absolute temperature, 300 "C. The electrochemical potential gradient for Na+ is ApNa+ = zF (E, -EN^+). system and the investigators who challenge its importance as a calcium efflux mechanism.