Barium Influx Mediated by the Cardiac Sodium-Calcium Exchanger in Transfected Chinese Hamster Ovary Cells

Abstract

We examined Ba²⁺ influx using isotopic and fura-2 techniques in transfected Chinese hamster ovary cells expressing the bovine cardiac Na⁺/Ca²⁺ exchanger (CK1.4 cells). Ba²⁺ competitively inhibited exchange-me diated ⁴⁵Ca²⁺ uptake with a K _i ∼ 3 mM. Ba²⁺ uptake was stimulated by pretreating the cells with ouabain and by removing extracellular Na⁺, as expected for Na⁺/Ba²⁺ exchange activity. The maximal velocity of Ba²⁺ accumulation was estimated to be 50% of that for Ca²⁺. When the monovalent cation ionophore gramicidin was used to equilibrate internal and external concentrations of Na⁺, Ba²⁺ influx was negligible in the absence of Na⁺ and increased to a maximum at 20–40 mM Na⁺. At higher Na⁺ concentrations, Ba²⁺ influx declined, presumably due to the competition between Na⁺ and Ba²⁺ for transport sites on the exchanger. Unlike Ca²⁺, Ba²⁺ did not appear to be taken up by intracellular organelles: Thus, ¹³³Ba²⁺ uptake in ouabain-treated cells was not reduced by mitochondrial inhibitors such as Cl-CCP or oligomycin-rotenone. Moreover, intracellular Ca²⁺ stores that had been depleted of Ca²⁺ by pretreatment of the cells with ionomycin (a Ca²⁺ ionophore) remained empty during a subsequent period of Ba²⁺ influx. Ca²⁺ uptake or release by intracellular organelles secondarily regulated exchange activity through alterations in [Ca²⁺]_i. Exchange-mediated Ba²⁺ influx was inhibited when cytosolic [Ca²⁺] was reduced to 20 nM or less and was accelerated at cytosolic Ca²⁺ concentrations of 25–50 nM. We conclude that (a) Ba²⁺ substitutes for Ca²⁺ as a transport substrate for the exchanger, (b) cytosolic Ba²⁺ does not appear to be sequestered by intracellular organelles, and (c) exchange-mediated Ba²⁺ influx is accelerated by low concentrations of cytosolic Ca²⁺.

introduction

The cardiac Na⁺/Ca²⁺ exchanger couples the transmembrane movement of 3 Na⁺ ions to that of a single Ca²⁺ ion in the opposite direction. It is the principal mechanism for mediating Ca²⁺ efflux in cardiac myocytes. It plays a critical role in regulating cardiac contractility by competing with the sarco(endo)plasmic reticulum Ca²⁺ ATPase (SERCA)¹ for cytosolic Ca²⁺, thereby modulating the amount of releasable Ca²⁺ stored in the sarcoplasmic reticulum (reviewed in Reeves, 1995). Kinetic results are compatible with a consecutive exchange mechanism, in which Ca²⁺ and Na⁺ are translocated in separate steps (Khananshvili, 1990; Hilgemann et al. 1991).

Exchange activity is regulated by Ca²⁺- and/or ATP-dependent processes that affect the exchanger's distribution between an active state and either of two inactive states (Hilgemann et al., 1992a , b ). Entry of the exchanger into the first inactive state (I₁ inactivation) is thought to occur when the exchanger is fully loaded with Na⁺ at the cytoplasmic membrane surface; this mode of inactivation is observed experimentally as a time-dependent decrease in “reverse” Na⁺/Ca²⁺ exchange (Na⁺ _i-dependent Ca²⁺ influx) following a step increase in [Na⁺]_i (Hilgemann, 1990). The second (I₂) mode of inactivation is promoted by the absence of cytosolic Ca²⁺ and is detected experimentally as an activation of reverse exchange activity by submicromolar concentrations of cytosolic Ca²⁺ (secondary Ca²⁺ activation) (DiPolo, 1979; DiPolo and Beaugé, 1987). Cytosolic ATP counteracts both modes of inactivation, although the precise mechanism(s) involved have not been delineated. Despite the advances in our understanding of the regulatory behavior of the exchanger in subcellular or internally dialyzed cellular preparations, much less information is available on how exchange activity is regulated in intact cells, or how it interacts with other Ca²⁺ homeostatic mechanisms. Part of the difficulty in studying these issues stems from the technical limitations of Ca²⁺ influx measurements. Both ⁴⁵Ca²⁺ fluxes and fura-2 measurements are greatly affected by the sequestration and release of Ca²⁺ from the endoplasmic reticulum and the mitochondria. A previous report (Chernaya et al., 1996) used Ba²⁺ as a substitute for Ca²⁺ to assess effects of thapsigargin (Tg) on Na⁺/Ca²⁺ exchange activity in transfected Chinese hamster ovary cells expressing the bovine cardiac Na⁺/Ca²⁺ exchanger. Here we describe detailed studies of Ba²⁺ transport by the Na⁺/Ca²⁺ exchanger in these cells. Consistent with results obtained with other cells, cytosolic Ba²⁺ is not significantly accumulated by either the endoplasmic reticulum or the mitochondria. Measurements of Na⁺ _i-dependent Ba²⁺ influx therefore provide a more direct measure of exchange activity than the corresponding Ca²⁺ flux measurements. Moreover, regulatory activation of Na_i-dependent Ba²⁺influx by [Ca²⁺]_i can be readily observed under appropriate conditions. Our results suggest that Ca²⁺ _i-dependent activation of exchange activity involves a complex interplay between [Ca²⁺]_i and various intracellular Ca²⁺ compartments.

methods

Cells

CK1.4 cells were prepared by transfection of dhfr ⁻ CHO cells (CCL 61; American Type Culture Collection, Rockville, MD) with a mammalian expression vector (pcDNA I/Neo; Invitrogen Corp., San Diego, CA) containing a cDNA insert coding for the bovine cardiac Na⁺/Ca²⁺ exchanger (Aceto et al. 1992; Pijuan et al. 1993). Control cells were prepared by transfection of the CHO cells with pcDNA3 (Invitrogen Corp.), a closely related expression vector with no cDNA insert. The cells were grown in Iscove's modified Dulbecco's medium containing 500 μg/ml geneticin (G418; GIBCO, Gaithersburg, MD), either 10% FCS (JRH Biosciences, Lenexa, KS) or 10% supplemented calf serum (Cool Calf 2; Sigma Chemical Co., St. Louis, MO) and antibiotics as described (Pijuan et al., 1993). Unless otherwise specified, all biochemicals were obtained from Sigma Chemical Co.

Solutions

Na-PSS (physiological salts solution) contains 140 mM NaCl, 5 mM KCl, 1 mM MgCl₂, 10 mM glucose, and 20 mM MOPS, pH adjusted to 7.4 (37°C) with Tris. NMDG-, Li-, and K-PSS have the same composition as Na-PSS, except that Na⁺ is replaced with NMDG⁺, Li⁺, and K⁺, respectively. The termination medium for the ⁴⁵Ca²⁺ transport assay consists of 100 mM MgCl₂, 10 mM LaCl₃, and 5 mM MOPS, pH 7.4 (Tris).

⁴⁵Ca²⁺ and ¹³³Ba²⁺ Uptake

Cells were grown to confluence in 24-well plastic dishes and preincubated for 30 min at 37°C with 1 ml/well of nominally Ca²⁺-free Na-PSS with or without 0.4 mM ouabain as indicated. The preincubation medium was then aspirated and replaced with 200 μl of assay medium (Na- or NMDG-PSS, as indicated) containing 100 μM ⁴⁵Ca²⁺ or ¹³³Ba²⁺. Radioisotopes were obtained from DuPont NEN Research Products (Boston, MA). For cells preincubated with ouabain, the assay solutions also contained 0.4 mM ouabain. After the desired interval, the wells were washed four times with 1 ml of termination medium; the contents of the wells were then extracted with 1 ml of 0.1 N HNO₃ and counted. Protein was determined in separate sample wells by the Lowry method (Lowry, 1951). Data are presented as the mean values ± SEM (error bars shown in figures) for the indicated number (n) of experiments.

Fura-2 Assays

Cells were grown to confluence in 75-cm² plastic culture flasks and washed three times with Na-PSS. The cells were released from the flask with 5 ml of Na-PSS + 5 mM EDTA, centrifuged, and resuspended in 5 ml of Na-PSS + 1 mM CaCl₂, centrifuged again, and resuspended in 4–5 ml of Na-PSS + 1 mM CaCl₂ + 1% BSA; the BSA aids in solubilizing the added fura-2-AM (Thomas and Delaville, 1991). The cells were distributed in individual 0.3-ml aliquots to plastic tubes and allowed to incubate for 30 min at 37°C to recover from isolation. Individual tubes of cells were loaded at 10-min intervals for 30 min (37°C) with 3 μM fura-2-AM (Molecular Probes, Eugene, OR) and 0.25 mM sulfinpyrazone to retard transport of fura-2 out of the cell (DiVirgilio et al., 1988; Pijuan et al., 1993). The fura-2 and sulfinpyrazone were added as 1,000-fold concentrated stock solutions in dimethyl sulfoxide. Where indicated, ouabain (0.4 mM) was also added as a 1,000-fold concentrated solution in dimethyl sulfoxide. After the 30-min loading period, the cells were rapidly centrifuged in an Eppendorf Mini-centrifuge, washed, and preincubated in 0.1 ml of the desired medium for 1 min as specified in individual experiments. The cells were then added directly to a cuvette containing 3 ml of either Na-PSS or Li-PSS and fura-2 fluorescence was monitored at 510-nm emission with alternate excitation at 350/390 nm (Schilling et al., 1989), using a Photon Technology International (South Brunswick, NJ) RF-M 2001 fluorometer. All fluorescence values were corrected for autofluorescence using cells that had not been loaded with fura-2. Data are presented as the ratio of fluorescence values at the 350/390 excitation wavelengths and represent the mean values ± SEM (error bars shown in figures) for the indicated number (n) of experiments. Calibrations were conducted with digitonin-permeabilized cells according to the procedure of Grynkiewicz et al. (1985) and yielded values of R_max = 7.5, R_min = 1.4 and Sf/Sb = 2.75 under our experimental conditions. K _D values for Ba²⁺ and fura-2 of 0.8 μM (Schilling et al., 1989), 1.4 μM (Kwan and Putney, 1990), and 2.4 μM (McCormack and Osbaldeston, 1990) have been reported; the average K _D (1.5 μM) yields [Ba²⁺] values of 0.5, 1.5, 3.1, and 6.0 μM at 350/390 ratios of 2, 3, 4, and 5, respectively. Fura-2 also responds to Ca²⁺ at the excitation wavelengths used for the Ba²⁺ measurements. Assuming K _D = 224 nM for the Ca²⁺/fura-2 complex (Grynkiewicz et al., 1985), the following 350/390 excitation ratios correspond to the indicated values for [Ca²⁺]_i: 2.0 (67 nM), 2.5 (135 nM), 3.0 (219 nM), 4.0 (457 nM), and 5.0 (887 nM).

results

Isotope Flux Studies

The data in Fig. 1 show the effects of various concentrations of Ba²⁺ on the rates of ⁴⁵Ca²⁺ uptake by ouabain-treated CK1.4 cells in a Na⁺-free medium. Ba²⁺ inhibited ⁴⁵Ca²⁺ uptake at both 0.1 and 1.0 mM [Ca²⁺], but was a more potent inhibitor at the lower Ca²⁺ concentration. Curiously, low concentrations of Ba²⁺ (0.1 mM) slightly stimulated ⁴⁵Ca²⁺ uptake; the explanation for this is unclear, and this phenomenon was not investigated further. At Ba²⁺ concentrations >1 mM, the inhibition was competitive (K _i = 3.1 mM) as indicated by the Dixon plot in the right panel of Fig. 1; including the full range of Ba²⁺ concentrations in the Dixon plot shifted the apparent K _i to 4.5 mM (data not shown). The results imply that Ba²⁺ interacts with the Ca²⁺ transport site on the Na⁺/Ca²⁺ exchanger.

Figure 1.

Ba²⁺ inhibits exchange-mediated ⁴⁵Ca²⁺ uptake by CK1.4 cells. (Left) Cells were pretreated with 0.4 mM ouabain and assayed for ⁴⁵Ca²⁺ uptake (15 s) in NMDG-PSS containing the indicated concentrations of BaCl₂. The concentrations of ⁴⁵Ca²⁺ used are 1 mM (filled circles) and 0.1 mM (open circles). (Right) The data from the left panel for 1–10 min are presented as a Dixon plot. The point of intersection of the two lines yields an apparent K _i for Ba²⁺ of 3 mM (n = 5).

To determine whether Ba²⁺ is transported by the exchanger, we examined ¹³³Ba²⁺ uptake. As shown in Fig. 2 A (open symbols), ¹³³Ba²⁺ uptake in an Na-free medium was stimulated by prior treatment of the cells with ouabain to elevate intracellular Na⁺. In the presence of physiological concentrations of extracellular Na⁺, ¹³³Ba²⁺ uptake by ouabain-treated cells was strongly inhibited (Fig. 2 B, open squares). For vector transfected control cells (Fig. 2, filled symbols), which do not exhibit Na⁺/ Ca²⁺ exchange activity (Pijuan et al., 1993; Chernaya et al., 1996), ¹³³Ba²⁺ uptake was <20% of the maximal levels shown by CK1.4 cells and was unaffected by ouabain treatment or Na-free conditions. The results with ¹³³Ba²⁺ are qualitatively similar to those obtained with ⁴⁵Ca²⁺; the maximal levels of ⁴⁵Ca²⁺ uptake, however, are generally three to five times higher than for ¹³³Ba²⁺ (see, for example, Fig. 6). Based on the initial rates of exchange-mediated ¹³³Ba²⁺ uptake in Fig. 2, and assuming a K _m of 3 mM, we calculate a V_max for Ba²⁺ uptake of 3.7 nmol/mg protein/15 s, or ∼50% of that reported previously for ⁴⁵Ca²⁺ uptake (Condrescu et al., 1995). As shown in Fig. 3, extracellular Na⁺ stimulated ¹³³Ba²⁺ efflux from preloaded CK1.4 cells, suggesting that the exchanger also transports Ba²⁺ out of the cells. The efflux data with ¹³³Ba²⁺ are again qualitatively similar to those obtained using ⁴⁵Ca²⁺.

Figure 2.

¹³³Ba²⁺ uptake by CK1.4 cells and vector-transfected control cells. (A) Cells were preincubated with or without 0.4 mM ouabain, as indicated, and assayed for ¹³³Ba²⁺ uptake in NMDG-PSS. (B) Cells were preincubated with ouabain and assayed for ¹³³Ba²⁺ uptake in NMDG-PSS or Na-PSS, as indicated. Open symbols, CK1.4 cells (n = 3); closed symbols, control cells (n = 2).

Figure 6.

Effect of mitochondrial antagonists on ⁴⁵Ca²⁺ and ¹³³Ba²⁺ uptake by CK1.4 cells. Cells were treated with ouabain and assayed for in NMDG-PSS for ⁴⁵Ca²⁺ (0.1 mM) or ¹³³Ba²⁺ (0.1 mM) uptake (5 min); where indicated 10 μM Cl-CCP or 2.5 μg/ml oligomycin + 2 μM rotenone were present in the assay medium (n = 6, ¹³³Ba²⁺; n = 4, ⁴⁵Ca²⁺).

Figure 3.

¹³³Ba²⁺ efflux from CK1.4 cells. Ouabain-treated cells were allowed to accumulate ¹³³Ba²⁺ for a period of 5 min and then placed in Na-PSS (open circles) or NMDG-PSS (closed circles) for the indicated intervals (n = 4).

Fura-2 Measurement of Ba²⁺ Influx

An alternate method of assaying Ba²⁺ movements is to use the Ca²⁺-indicating dye fura 2 (Schilling et al., 1989). As shown in Fig. 4 A, the addition of Ba²⁺ to fura 2–loaded CK1.4 cells produced an increase in the 350/ 390 excitation ratio which was enhanced by Na⁺ _o-removal (Li⁺-substitution). Pretreating the cells with ouabain to elevate [Na⁺]_i (Fig. 4 B) increased the 350/390 excitation ratio under Na⁺-free conditions, consistent with the expected increase in Na⁺ _i-dependent Ba²⁺ influx via the exchanger. Similar results were obtained when N -methyl-d-glucamine was used as the Na⁺ substitute instead of Li⁺ (data not shown). When vector-transfected control cells were used instead of CK1.4 cells, the increase in the 350/390 ratio was not affected by Na⁺-removal or by ouabain-treatment and was approximately equivalent to that seen in the presence of Na⁺ for the CK1.4 cells (data not shown).

Figure 4.

Ba²⁺ influx in CK1.4 cells detected with fura-2. Cells were loaded with fura-2 without (A) or with (B) 0.4 mM ouabain, washed, and preincubated for 1 min in Na-PSS + 1 mM CaCl₂ and then diluted 30-fold into cuvettes containing either Li-PSS or Na-PSS, as indicated; 0.3 mM EGTA was included in both media. BaCl₂ (1 mM) and EGTA (10 mM) were added as indicated by the arrows (n = 5).

Adding 10 mM EGTA after a period of Ba²⁺ accumulation in the Na-free medium resulted in little or no decline in the 350/390 ratio (Fig. 4) suggesting that cytosolic Ba²⁺ is transported poorly or not at all by the ATP-dependent Ca²⁺ pumps. In comparable experiments conducted with Ca²⁺ instead of Ba²⁺, EGTA addition results in a rapid decline in [Ca²⁺]_i (see Condrescu et al., 1995; Chernaya et al., 1996). The results with Ba²⁺ in the fura-2 experiments differ from the results of the ¹³³Ba²⁺ flux studies (Fig. 3), which indicated that ¹³³Ba²⁺ was lost from the CK1.4 cells, even in the absence of extracellular Na⁺ (cf., discussion).

Dependence of Ba²⁺ Influx on [Na⁺]_i

The effects of ouabain treatment indicate that Ba²⁺ entry is accelerated when [Na⁺]_i is increased. The influence of [Na⁺]_i is examined more directly in the experiments shown in Fig. 5. Fura-2–loaded CK1.4 cells were placed in cuvettes containing K-PSS with various concentrations of Na⁺ (mM concentrations given in Fig. 5 next to individual traces) and treated with 1 μM gramicidin to bring about rapid equilibration of monovalent cations across the plasma membrane. Thus, in the presence of gramicidin, Na⁺ concentrations should be equal on both sides of the cell membrane. In the absence of Na⁺, Ba²⁺ influx was negligible. (The initial, abrupt rise in the fura-2 ratio upon addition of Ba²⁺ is due to the presence of small amounts of extracellular fura-2.) Increasing concentrations of Na⁺ produced progressively higher rates of Ba²⁺ influx with maximal rates at 20–40 mM Na⁺. With higher Na⁺ concentrations, Ba²⁺ influx declined to a level that was only slightly higher, at 140 mM Na⁺, than that observed in the absence of Na⁺. The slopes of the fura-2 traces are presented as a function of [Na⁺] in the inset to the right panel of Fig. 5. The increasing rates of Ba²⁺ influx within the range of 0–20 mM [Na⁺] most likely reflect the stimulatory effects of cytosolic Na⁺ in activating exchange activity. The decline in Ba²⁺ influx rates at higher Na⁺ concentrations is probably due to competition between external Na⁺ and Ba²⁺ for transport sites on the exchange carrier. When control transfected cells were used in similar experiments, Ba²⁺ influx was low (comparable to that seen in the absence of Na⁺ for the CK1.4 cells), and variations in [Na⁺] had no effect. We conclude that the Na⁺/Ca²⁺ exchanger provides the major route of Ba²⁺ entry in the CK1.4 cells.

Figure 5.

Ba²⁺ influx in gramicidin-treated CK1.4 cells. Cells were loaded with fura-2, washed, and preincubated in K-PSS + 1 mM CaCl₂, centrifuged, and resuspended in mixtures of K-PSS + Na-PSS that yielded the final Na⁺ concentrations (in mM) indicated next to each trace; each solution also contained 0.3 mM EGTA. Gramicidin (1 μM) was added immediately after the cells and 1 mM BaCl₂ was added at 30 s. (n = 3–5; for 70 mM Na⁺, n = 2). Inset: Dependence of the rate of Ba²⁺ influx on [Na⁺]. The slopes of the traces at each [Na⁺], normalized to that for [Na⁺] = 0, are plotted on the ordinate scale.

Ba²⁺ Uptake and Organellar Sequestration

Previous reports have suggested that Ba²⁺ is not significantly sequestered by the endoplasmic reticulum (Kwan and Putney, 1990; Rasgado-Flores et al., 1986). Ba²⁺ uptake by mitochondria has been reported, but the results vary markedly with different cell types (cf., discussion). We conducted the following experiments to determine whether Ba²⁺ was sequestered by intracellular organelles in CK1.4 cells and to assess whether the presence of cytosolic Ba²⁺ interfered with other Ca²⁺ transport processes.

To examine mitochondrial accumulation of Ba²⁺, we measured the effects of mitochondrial inhibitors on ⁴⁵Ca²⁺ and ¹³³Ba²⁺ uptake. As shown in Fig. 6, the uncoupler Cl-CCP (2 μM) and the combination of oligomycin (2.5 μg/ml) + rotenone (2 μM) inhibited exchange-mediated ⁴⁵Ca²⁺ uptake in ouabain-treated CK1.4 cells by 42 ± 2 and 32 ± 2%, respectively, but had no effect on ¹³³Ba²⁺ uptake (−11 ± 3 and 6 ± 4% inhibition, respectively). The results suggest that any mitochondrial accumulation of Ba²⁺ was small compared to that observed with Ca²⁺, consistent with the reported selectivity of the mitochondrial uniporter toward divalent cations (Ca²⁺ >> Ba²⁺; Saris and Åkerman, 1980). To address the question of whether cytosolic Ba²⁺ could be sequestered by the endoplasmic reticulum, we depleted intracellular Ca²⁺ stores using the Ca²⁺ ionophore ionomycin and then asked whether the stores would refill with Ba²⁺ during a subsequent period of Ba²⁺ influx. We assessed the degree of filling of intra-cellular stores by measuring the response of fura-2– loaded cells to the addition of extracellular ATP. ATP binds to P_2U receptors in Chinese hamster ovary cells and elicits the production of inositol (1,4,5)-trisphosphate (InsP₃), leading to release of sequestered Ca²⁺ from InsP₃-sensitive stores (Iredale and Hill, 1993; Pijuan et al., 1993). This is illustrated in the inset to Fig. 7, where ATP elicits a pronounced [Ca²⁺]_i transient when added to CK1.4 cells (control trace). In contrast to this behavior, when an aliquot of cells was pretreated for 1 min with 10 μM ionomycin before adding the cells to the cuvette, no Ca²⁺ transient was observed (Fig. 7, inset). This indicates that the ionomycin had released essentially all the sequestered Ca²⁺ from the stores during the 1-min preincubation. It is important to note that in all traces shown in Fig. 7, 0.3% BSA was present in the cuvette to scavenge residual ionomycin (cf., Chernaya et al., 1996); this was done to ensure that the presence of the ionophore would not interfere with possible Ba²⁺ accumulation in the stores. In comparable experiments carried out with Ca²⁺, the InsP₃-sensitive stores refilled rapidly in the presence of extracellular Ca²⁺ (Chernaya et al., 1996).

Figure 7.

Test for Ba²⁺ sequestration by InsP₃-sensitive stores. (control) Cells were preloaded with fura-2 in the presence of 0.4 mM ouabain, washed, and incubated for 1 min in Na-PSS + 1 mM CaCl₂ and then diluted 30-fold into cuvettes containing Li-PSS + 0.3 mM EGTA + 0.3% BSA. BaCl₂ (1 mM), EGTA (10 mM) and ATP (0.3 mM) were added as indicated. (ionomycin) Cells were loaded with fura-2 in the presence of ouabain, washed, and incubated for 1 min in Na-PSS + 0.3 mM EGTA containing 10 μM ionomycin; the cells were then centrifuged and resuspended in Li-PSS + 0.3 mM EGTA + 0.3% BSA. BaCl₂, EGTA, and ATP were added as in the control trace. (Inset) Experimental conditions are as described for the control and ionomycin-treated cells above, except that ATP (0.3 mM) was added as indicated (n = 6).

In the main panel of Fig. 7, ouabain-treated CK1.4 cells with intact Ca²⁺ stores were allowed to accumulate Ba²⁺ for 2.5 min in an Na⁺-free medium; then 10 mM EGTA was added to block any further Ba²⁺ influx, and ATP was subsequently added (after a 30-s delay) to measure the amount of Ca²⁺ present in the stores. As shown, ATP evoked a robust [Ca²⁺]_i transient under these conditions (control trace). These results indicate that the InsP₃-sensitive stores had retained their Ca²⁺ load during the period of Ba²⁺ accumulation and responded normally to ATP addition. Note that after the peak of the [Ca²⁺]_i transient, the fura-2 signal declined toward the level seen before ATP addition; continued incubation resulted in stabilization of the fura-2 signal at the level observed before ATP addition, consistent with the absence of Ba²⁺ removal from the cytosol (data not shown). Similar results were seen when 2 μM ionomycin was used to elicit Ca²⁺ release instead of ATP (data not shown). Thus, the presence of cytosolic Ba²⁺ did not block either the ATP-evoked Ca²⁺ release pathway or Ca²⁺ removal from the cytosol. The latter process may occur through either the SERCA or the plasma membrane Ca²⁺ pumps, or both. In the case where Ca²⁺ release was elicited by ionomycin, only the plasma membrane Ca²⁺ pump would be expected to contribute to Ca²⁺ removal from the cytosol since intracellular organelles would be unable to accumulate Ca²⁺ in the presence of the ionophore. These results therefore indicate that Ca²⁺ extrusion by the plasma membrane Ca²⁺ ATPase is not blocked by the presence of cytosolic Ba²⁺.

When ionomycin-pretreated cells (with depleted Ca²⁺ stores) were subjected to the protocol described above, ATP did not elicit an increase in the 350/390 ratio (Fig. 7, main panel). The absence of an increase in [Ba²⁺]_i indicates that the InsP₃-sensitive stores remained empty during the period of Ba²⁺ uptake. Note that Ba²⁺ is rapidly conducted by InsP₃-gated Ca²⁺ channels (Bezprozvanny and Ehrlich, 1994) and would therefore have been released from the stores if present. The results indicate that Ba²⁺ is not sequestered by InsP₃-sensitive stores in these cells, probably due to its inability to serve as a transport substrate for the SERCA Ca²⁺ pump.

An unexpected feature of these results was that the ionomycin-treated cells showed a sharply reduced rate of Ba²⁺ influx compared to control cells that had not been treated with ionomycin. As described in more detail below, this appears to be a secondary consequence of the reduced [Ca²⁺]_i in the cells with depleted Ca²⁺ stores: following removal of the ionomycin with BSA, the stores would be expected to re-sequester residual cytosolic Ca²⁺ and lower [Ca²⁺]_i. The average value of [Ca²⁺]_i during the 15 s before the addition of Ba²⁺ in this experiment was 23 ± 7 nM for the ionomycin/BSA-treated cells vs. 41 ± 5 nM for the control cells (P < 0.001, paired t test; n = 6).

Store Depletion and Ba²⁺ Influx

To examine the issues raised above in greater detail, we studied the effects of ionomycin on exchange-mediated Ba²⁺ influx in cells treated with thapsigargin (Tg), a selective blocker of SERCA Ca²⁺ pumps (Lytton et al., 1991). CK1.4 cells were pretreated with ionomycin, Tg, or both agents and then washed with 1% BSA to scavenge residual ionomycin. The cells were then placed in Li-PSS + 0.3 mM EGTA and fura-2 fluorescence was monitored after the addition of 1 mM BaCl₂. In each case, the cells had also been loaded with cytosolic Na⁺ by including 0.4 mM ouabain in the fura-2 loading medium (see legend to Fig. 8 for further details).

Figure 8.

Ba²⁺ influx in CK1.4 cells pretreated with Tg, ionomycin or both. Aliquots (0.3 ml) of CK1.4 cells were loaded with fura-2 in the presence of 0.4 mM ouabain, washed, and preincubated for 1 min in 0.1 ml of Na-PSS + 0.3 mM EGTA containing either 200 nM Tg, 10 μM ionomycin, both Tg and ionomycin, or neither agent (control). The cells were then centrifuged, resuspended in 0.3 ml of Na-PSS + 0.3 mM EGTA + 1% BSA and incubated for an additional 1 min. After an additional centrifugation, the cells were resuspended in Li-PSS + 0.3 mM EGTA in the cuvette; BaCl₂ (1 mM) was added as indicated (n = 3).

As shown in Fig. 8, Ba²⁺ influx was greatest in the Tg-treated cells and least in the cells pretreated with ionomycin; in the latter case, Ba²⁺ influx was essentially identical to that seen in Na-PSS (data not shown), implying that the exchanger was inactive under these conditions. Remarkably, Ba²⁺ influx in cells that had been treated with both ionomycin and Tg was greatly reduced compared to cells treated with Tg alone. Control cells that had been subjected to the same preincubation protocol, but in the absence of either ionomycin or Tg, also showed reduced Ba²⁺ influx; this is due to the mechanical manipulations involved in the protocols, which lead to release of ATP, depletion of the Ca²⁺ stores, and a decline in [Ca²⁺]_i (unpublished observations; cf. below). The rate of Ba²⁺ influx shows a reasonably good correlation with the values of [Ca²⁺]_i observed during the 15 s interval before the addition of Ba²⁺. The [Ca²⁺]_i values corresponding to Tg, control, Tg + ionomycin or ionomycin treatments were 34 ± 3, 26 ± 1, 20 ± 3, and 17 ± 1 nM respectively (n = 3 in each case).

Our interpretation of these results (cf. discussion) is that exchange activity is regulated secondarily through the influence of intracellular organelles on the cytosolic Ca²⁺ concentration. Release of Ca²⁺ from internal stores elevates [Ca²⁺]_i and accelerates exchange activity (cf. Chernaya et al., 1996), whereas the exchanger is deactivated when these organelles resequester residual cytosolic Ca²⁺ and lower [Ca²⁺]_i. The effects of organellar Ca²⁺ release and sequestration are further documented in the accompanying manuscript (Vázquez et al., 1997), which describes time-dependent changes in exchange activity associated with ATP-induced Ca²⁺ release from InsP₃-sensitive Ca²⁺ stores. Two surprising aspects of results in Figs. 7 and 8 deserve special emphasis: (a) exchange activity becomes activated at quite low values of [Ca²⁺]_i (25–50 nM) and (b) exchange activity can be deactivated even when the SERCA pumps are blocked by Tg, implying that Tg-resistant Ca²⁺ pumps also participate in regulation of cytosolic Ca²⁺.

discussion

The results presented here show that Ba²⁺ provides an advantageous alternative to Ca²⁺ for measurements of exchange activity in intact cells. Ba²⁺ competitively inhibited ⁴⁵Ca²⁺ uptake by Na⁺/Ca²⁺ exchange and therefore presumably binds to the Ca²⁺ transport sites on the exchanger (Fig. 1). ¹³³Ba²⁺ uptake was stimulated by ouabain-treatment and was completely suppressed by 140 mM extracellular Na⁺ (Fig. 2); similar behavior was observed using fura-2–loaded cells to assess Ba²⁺ influx (Fig. 4). The effects of Na⁺ in the gramicidin-treated cells (Fig. 5) indicated that Ba²⁺ influx was blocked in the absence of cytosolic Na⁺ and was accelerated by increasing concentrations of cytosolic Na⁺. We conclude that the exchanger provides the major pathway for Ba²⁺ entry in these cells. The data with gramicidin-treated cells also suggests that exchange activity is quite sensitive to Na⁺ _i; a small, but significant increase in Ba²⁺ entry was observed at concentrations as low as 2.3 mM Na⁺ (Fig. 5).

Previously published studies addressed the question of whether Ba²⁺ is transported by the Na⁺/Ca²⁺ exchanger. Trosper and Philipson (1983) reported that ⁴⁵Ca²⁺ efflux from preloaded cardiac sarcolemmal vesicles was accelerated in the presence of Ba²⁺ and suggested that this was due to Ca²⁺/Ba²⁺ exchange. Tibbits and Philipson (1985) measured exchange-mediated ¹³³Ba²⁺ uptake by cardiac sarcolemmal vesicles; maximal rates of uptake were 20-fold lower than for Sr²⁺ or Ca²⁺. Measurements of outward exchange currents in cardiac myocytes indicated that Ba²⁺ was transported only poorly, if at all, by the exchanger (Kimura et al., 1987; Shimoni and Giles, 1987). However, more recent experiments with excised membrane patches from cardiac myocytes and/or frog oocytes expressing the cardiac exchanger indicate that significant outward exchange currents can be observed with Ba²⁺ as the transported substrate (personal communications from Dr. Donald Hilgemann, University of Texas Southwestern Medical Center, Dallas, TX and Dr. Larry Hryshko, St. Boniface General Hospital Research Center, Winnipeg, Canada). In contrast to reports that Ba²⁺ is transported only poorly by the exchanger, our estimated maximal influx rates with ¹³³Ba²⁺ are ∼50% of those observed with ⁴⁵Ca²⁺. It is not clear why there is such a disparity among the different reports on the relative rates of Ca²⁺ and Ba²⁺ as exchange substrates. We have recently observed that exchange-mediated Ba²⁺ influx in CHO cells is much more temperature sensitive than Ca²⁺ influx (unpublished observations); while this in itself does not explain the differing results among the various experimental reports, it does suggest that Ba²⁺ translocation by the exchanger might involve mechanistic constraints that do not apply to Ca²⁺. Thus, the varying results obtained among different investigators might reflect subtle differences in experimental conditions, species differences, or variations in membrane composition that exert disproportionate effects on exchange-mediated Ba²⁺ movements.

Approximately 1–2 nmol/mg protein of ¹³³Ba²⁺ were accumulated by ouabain-treated cells under Na⁺-free conditions (Figs. 2 and 6). With a cellular water content of 6 μl/mg protein, this is equivalent to 200–300 μM total intracellular Ba²⁺. In fura-2 experiments, the 350/390 ratios during Ba²⁺ influx approached 4.0, which is equivalent to a cytosolic concentration of approximately 3 μM. Thus, much of the intracellular Ba²⁺ is buffered, although it is unclear which cellular constituents are involved in this process. With ⁴⁵Ca²⁺, the cells accumulated 7 nmol/mg protein under the conditions of Fig. 6; with different conditions (1 mM ⁴⁵Ca²⁺, 40 mM Na⁺ + 100 mM K⁺ in the assay media), ⁴⁵Ca²⁺ accumulations of up to 40 nmol/mg protein have been observed (data not shown). [Ca²⁺]_i rarely exceeded 1 μM during exchange-mediated Ca²⁺ influx, even under conditions favoring extensive Ca²⁺ accumulation. Thus, the ratio of total cellular cation to the free cytosolic concentration was much higher for Ca²⁺ than for Ba²⁺, a result consistent with a greater degree of organellar sequestration of Ca²⁺ (cf. below).

An unexpected disparity was observed between the efflux data obtained with ¹³³Ba²⁺ and fura-2. In the ¹³³Ba²⁺ studies, preaccumulated Ba²⁺ was lost from the cells with a half-time of 3–4 min under Na⁺-free conditions (Fig. 3). With fura-2–loaded cells, however, little or no decline in cytosolic [Ba²⁺] could be detected when 10 mM EGTA was added (under Na⁺-free conditions) following a period of Ba²⁺ influx (Fig. 4). This ratio of EGTA to Ba²⁺, under the conditions of our experiments, yields a free [Ba²⁺] of 1.6 μM (MAXC program; Bers et al., 1994) and completely blocks Ba²⁺ influx. Moreover, despite the data in Fig. 3 indicating that extracellular Na⁺ stimulates ¹³³Ba²⁺ efflux, we were unable to demonstrate an Na⁺ _o-dependent decline in c ytosolic [Ba²⁺] in fura-2–loaded cells using several different protocols to preload cells with Ba²⁺ (data not shown). This behavior remains unexplained at the present time.

A major advantage of Ba²⁺ over Ca²⁺ in measuring exchange activity is that Ba²⁺ was not sequestered by the endoplasmic reticulum in these cells (Fig. 7). This conclusion confirms results reported previously using other cell types (Rasgado-Flores et al., 1987; Kwan and Putney, 1990). Moreover, Ba²⁺ did not appear to be significantly accumulated by mitochondria, as judged by the absence of an effect of mitochondrial inhibitors on ¹³³Ba²⁺ uptake (Fig. 6). Previous studies of Ba²⁺ uptake by mitochondria yielded conflicting results. Uptake of divalent cations by isolated mitochondria showed a selectivity sequence Ca²⁺ > Sr²⁺ >> Mn²⁺ > Ba²⁺ (Saris and Åkerman, 1980; Vanio et al., 1970). In rat liver mitochondria, Ba²⁺ uptake was further inhibited by K⁺ and Mg²⁺ (Vanio et al., 1970), suggesting that the cytosolic concentrations of these ions would greatly reduce the capacity of mitochondria to accumulate Ba²⁺. Studies with permeabilized synaptosomes also showed little if any Ba²⁺ accumulation by mitochondria (Rasgado-Flores et al., 1987). On the other hand, Ba²⁺ accumulation was readily detectable in fura-2–loaded rat heart mitochondria (McCormack and Osbaldeston, 1990). Mitochondria in pancreatic B cells (Howell and Tyhurst, 1976) and rabbit vascular smooth muscle (Somlyo et al., 1974) showed dense Ba²⁺ deposits when the cells were exposed to high external concentrations of Ba²⁺. While our results do not rule out mitochondrial Ba²⁺ accumulation in the CHO cells, they suggest that this process is likely to be negligible compared to Ca²⁺ sequestration. Thus, measurements of Ba²⁺ uptake by either isotopic flux measurements or fura-2 should be a more direct indicator of exchange activity than the corresponding measures of Ca²⁺ uptake.

A second advantage of using Ba²⁺ for studies of exchanger regulation is that activation of exchange activity by [Ca²⁺]_i can be readily observed (cf. below). This is experimentally difficult to demonstrate with Ca²⁺ influx measurements because Ca²⁺ entering the cells would itself accelerate exchange activity through positive feedback involving secondary Ca²⁺ activation. This implies that cytosolic Ba²⁺ is considerably less effective than Ca²⁺ for activating exchange activity at the Ca²⁺ regulatory sites. Recent measurements of exchange currents in excised patches from frog oocytes expressing the exchanger (NCX1) have verified this conclusion (personal communication, Dr. L. Hryshko).

The effects of Tg and ionomycin shown in Figs. 7 and 8 indicate that exchange activity depends secondarily on the Ca²⁺ sequestering activities of intracellular organelles through their influence on [Ca²⁺]_i. Two aspects of these results deserve special mention. First, the exchanger becomes activated at surprisingly low cyto-solic Ca²⁺ concentrations. The experiments with fura-2 (Figs. 4 B, 5, 7, and 8) indicate that high levels of exchange-mediated Ba²⁺ influx are observed when cyto-solic Ca²⁺ levels (during the period just before the addition of Ba²⁺) were between 35 and 70 nM; on the other hand, exchange activity was greatly reduced when [Ca²⁺]_i declined below 20 nM (Figs. 7 and 8). Note that these experiments were carried out in the absence of extracellular Ca²⁺, and so the lower values of [Ca²⁺]_i were substantially below “resting” values under physiological conditions (50–75 nM; Pijuan et al., 1993; Vázquez et al., 1997). Because of the relatively simple calibration procedure used, the values cited for [Ca²⁺]_i may be slightly inaccurate, but any errors are unlikely to be large enough to affect the general conclusions drawn below.

The [Ca²⁺]_i values that activate exchange activity in our experiments were clearly much lower than values obtained with excised sarcolemmal patches, where concentrations of 300–600 nM are required for half-maximal activation of outward exchange currents (Hilgemann et al., 1992b ). On the other hand, our results agree closely with those of Miura and Kimura (1989) and of Noda et al.(1988), who observed half-maximal activation of outward exchange currents in guinea pig myocytes at [Ca²⁺]_i = 22 and 47 nM, respectively. The reasons for the differences between the behavior of excised patches and intact cells are a matter for speculation: They might involve the loss of regulatory cellular components in the patches, or local interactions with intracellular Ca²⁺ storage organelles that elevate [Ca²⁺]_i in the vicinity of the exchanger above the value in the bulk cytosol. In any event, our results imply that in cells under resting physiological conditions, where [Ca²⁺]_i is typically 50–75 nM, the exchanger is at least partially activated.

The second aspect of our results that merits detailed consideration is the complex interaction between exchange activity and intracellular organelles. The data in Fig. 7 and 8 indicate that when intracellular Ca²⁺ stores were depleted with ionomycin, subsequent removal of the ionophore allowed resequestration of cytosolic Ca²⁺ by the stores, resulting in a reduction of [Ca²⁺]_i below levels needed to activate the exchanger. Deactivation of exchange activity by organellar Ca²⁺ sequestration was not observed in preliminary experiments with cells expressing an exchanger deletion mutant that is not regulated by [Ca²⁺]_i. The unexpected finding that exchange activity was reduced in ionomycin-treated cells where the SERCA Ca²⁺ pump had been blocked with Tg (Fig. 8) suggests that Tg-resistant Ca²⁺ pumps contributed to reducing [Ca²⁺]_i; at present, the identity of these pumps is unknown. It should be noted that these experiments were conducted in the absence of extracellular Ca²⁺; if Ca²⁺ had been present externally, Ca²⁺ entry through store-dependent influx pathways (Putney, 1990) would have increased [Ca²⁺]_i and activated exchange activity. These considerations raise the possibility that the exchanger participates in a capacitative feedback mechanism for regulating the filling state of intracellular stores. Thus, when intracellular stores are filled to capacity with Ca²⁺, their ability to sequester additional Ca²⁺ is reduced, leading to a rise in [Ca²⁺]_i and activation of exchange activity. Activation of the exchanger under physiological conditions would stimulate Ca²⁺ efflux, thereby reducing net Ca²⁺ entry into the cell and limiting any further increase in store size. Conversely, when the stores contain a reduced Ca²⁺ load, sequestration of Ca²⁺ from the cytosol would reduce [Ca²⁺]_i and attenuate exchange activity, thereby allowing additional filling of the stores. Capacitative feedback between intracellular stores and exchange activity, in conjunction with capacitative Ca²⁺ entry mechanisms (Putney, 1990), could be an important mechanism for controlling the Ca²⁺ content of InsP₃-sensitive stores in neutrophils, pancreatic β cells and vascular smooth muscle cells.