Rapid inhibition of the Na⁺-K⁺ pump affects Na⁺-Ca²⁺ exchanger-mediated relaxation in rabbit ventricular myocytes

Abstract

1. The direct influence of Na⁺-K⁺ pump activity on the ability of the Na⁺-Ca²⁺ exchanger to remove Ca²⁺ was investigated in isolated adult rabbit ventricular myocytes.

2. Cell shortening was measured using an edge-detection system. Cytoplasmic [Ca²⁺] was monitored using the fluorescent indicator indo-1. Electrophysiological parameters were recorded using high-resistance microelectrodes. The Na⁺-K⁺ pump was rapidly inhibited by removal of extracellular K⁺ and measurements were taken almost immediately to minimise effects on other cellular compartments. Activity of the Na⁺-Ca²⁺ exchanger was monitored during release of Ca²⁺ from the sarcoplasmic reticulum (SR) elicited by rapid application of 15 mm caffeine.

3. When Na⁺-K⁺ pump activity was affected by K⁺ removal, cell relaxation and indo-1 fluorescence decline were slowed by approximately 40 %. The charge calculated by integrating the caffeine-induced transient inward current was unchanged, suggesting that there was no difference in the SR Ca²⁺ content in the two conditions. However Ca²⁺ flux via the Na⁺-Ca²⁺ exchanger was slower when the Na⁺-K⁺ pump was inhibited.

4. Similar experiments were performed by inhibiting the Na⁺-K⁺ pump using 0.5 mm strophanthidin. In this condition similar results to the ones observed by K⁺ removal were obtained, suggesting a specific role of the Na⁺-K⁺ pump in the phenomenon observed.

5. This study suggests that the activity of the Na⁺-K⁺ pump influences Na⁺-Ca²⁺ exchanger function in the absence of changes in SR Ca²⁺ content. This can be explained by a slower removal of Na⁺ from the subsarcolemmal space. The source of the increase in subsarcolemmal [Na⁺] requires further investigation. However, calculations derived from modelling suggest that the Na⁺-Ca²⁺ exchanger itself could be involved.

Intracellular sodium concentration ([Na⁺]) is an important factor in the regulation of contraction in cardiac muscle. One of the major Ca²⁺ regulatory mechanisms, the Na⁺-Ca²⁺ exchanger, is largely affected by changes in intracellular [Na⁺] hence influencing cell contractility (Bridge et al. 1988; Leblanc & Hume, 1990; Eisner, 1990; Blaustein & Lederer, 1999). Several cellular mechanisms are involved in [Na⁺] regulation. In particular, the Na⁺ channels allow Na⁺ entry during depolarisation. This Na⁺ is then removed by the Na⁺-K⁺-ATPase (Na⁺-K⁺ pump). The Na⁺-Ca²⁺ exchanger also brings Na⁺ into the cytoplasm in exchange for Ca²⁺ during Ca²⁺ extrusion. However, the significance of such a contribution to [Na⁺] regulation is debatable. Membrane currents ascribed to the exchanger and recorded during Ca²⁺ extrusion are too small to support a substantial change in bulk [Na⁺] during Ca²⁺ extrusion by this mechanism. Even when the Na⁺-Ca²⁺ exchanger removes Ca²⁺ during caffeine application to produce a large inward current, bulk cytoplasmic [Na⁺] can be calculated to increase by only a few micromoles, which will only minimally affect the function of the Na⁺-Ca²⁺ exchanger. However, several investigators have now proposed the existence of a subsarcolemmal space with restricted diffusion properties for Na⁺ and Ca²⁺ (e.g. Lederer et al. 1990; Langer & Rich, 1992; Trafford et al. 1995; Su et al. 1998). This means that even if Na⁺ influx via the Na⁺-Ca²⁺ exchanger during Ca²⁺ extrusion is small, it could produce a substantial increase in subsarcolemmal [Na⁺] and therefore affect the Na⁺-Ca²⁺ exchanger itself either by changing the electrochemical gradient or via a Na⁺-dependent inactivation mechanism (Hilgemann et al. 1992).

The ability of the Na⁺-Ca²⁺ exchanger to alter [Na⁺] in the subsarcolemmal space is important for two reasons: firstly, during Ca²⁺ extrusion, Na⁺ entry via the Na⁺-Ca²⁺ exchanger represents a negative feedback mechanism for the exchanger itself, therefore influencing Ca²⁺ regulation in the cytoplasm; secondly, a rise in subsarcolemmal [Na⁺] will depend on diffusion but also on Na⁺-K⁺ pump function (the major Na⁺ regulatory mechanism in cardiac cells). This could link the Na⁺-Ca²⁺ exchanger and the Na⁺-K⁺ pump functionally (e.g. Fujioka et al. 1998; Su et al. 1998; James et al. 1999; Blaustein & Lederer, 1999). The Na⁺-K⁺ pump could therefore affect Na⁺-Ca²⁺ exchanger function on a beat-to-beat basis.

The experiments described in this paper address these points by abruptly inhibiting the Na⁺-K⁺ pump and immediately recording Na⁺-Ca²⁺ exchanger function as a measure of its ability to remove Ca²⁺ from the cytoplasm. In the absence of changes in SR Ca²⁺ content and in Na⁺ entry via the Na⁺ channels, this can reveal a functional role for Na⁺ entry via the Na⁺-Ca²⁺ exchanger and its dependency on Na⁺-K⁺ pump function - a possible direct functional link between the two sarcolemmal mechanisms.

METHODS

Isolation of rabbit myocytes

Adult male New Zealand White rabbits were killed with an overdose of pentobarbital administered intravenously, the heart was removed and cells were isolated using a method described in detail elsewhere (Terracciano & MacLeod, 1994). Briefly, ventricular myocytes were enzymatically dissociated using protease (4 U ml⁻¹, Worthington), then collagenase (0.3 mg ml⁻¹, Worthington) and hyaluronidase (0.6 mg ml⁻¹, Sigma). The cells were filtered, centrifuged and then resuspended and stored in Dulbecco's modified Eagle's medium (DMEM) buffered with 25 mm Hepes (Gibco) at room temperature.

Cell shortening and cytoplasmic [Ca²⁺] measurements

Cells were placed on a superfusing chamber (volume, ≈60 μl) positioned on the stage of an inverted microscope (Nikon TE200). A thin coating of laminin (Sigma) was applied to the bottom of the chamber to improve cell adhesion.

Cell shortening was measured using a video edge-detection system described by Steadman et al. (1988). Measurements of the total cell length were not made, and the shortening of the cell was expressed as an absolute value and not as a percentage of cell length.

Cytoplasmic [Ca²⁺] was determined using the Ca²⁺-sensitive, single-excitation, dual-emission fluorescent dye, indo-1. Cells were loaded with 5 μm of the acetoxymethyl (AM) ester form of the indicator (indo-1 AM) (Molecular Probes, Inc.) for 20 min at room temperature. The supernatant was then discarded and replaced by fresh DMEM. Cells were stored in the dark at room temperature and used within 6-7 h. Calibration of indo-1 fluorescence was performed as previously described (Terracciano & MacLeod, 1997). Since all the experiments were performed at 37 °C, a K_d of 395 nm was used (Puglisi et al. 1996).

Current measurements

The electrophysiological experiments were performed using an Axoclamp-2B system (Axon Instruments). To avoid dialysis of the cells and to minimise the effects of changing the intracellular environment, high-resistance (15-30 MΩ) microelectrodes were used. The microelectrode filling solution contained: KCl, 2 m; EGTA, 0.1 mm; Hepes, 5 mm; pH 7.2. Voltage-clamp experiments were performed in switch-clamp mode. The switching rate was 3-5 kHz. Gain was increased up to 0.8 nA mV⁻¹ to obtain a maximally square voltage trace with no oscillations. The phase lag control was not used in order to avoid ‘false clamp’ (Terracciano & MacLeod, 1997; Terracciano et al. 1998). Protocols were controlled with pCLAMP 7 software (Axon Instruments). Calculations of SR Ca²⁺ content from the integral of the caffeine-induced transient inward current were performed as previously described (Terracciano et al. 1995).

Solutions

The 6 mm potassium solution (6K) was made up as follows (mm): NaCl, 140; KCl, 6; MgCl₂, 1; CaCl₂, 1; glucose, 10; Hepes, 10; pH to 7.4 with 2 m NaOH. In the 0K solution KCl was omitted. Chemicals were purchased from BDH. A 0.5 m stock solution of strophanthidin (Sigma) was made in DMSO. Caffeine was added as solid to the final solution.

The temperature of the superfusing solution was approximately 37 °C. The different superfusing solutions were fed via separate tubing and were connected to the bath using a 4-way perfusion manifold with a near-zero dead space configuration (MP-4, Warner Instrument Corp., Hamden, CT, USA). The whole tubing system was water-jacketed at 37 °C to avoid changes in temperature during rapid switching. The rate of superfusion was 2-3 ml min⁻¹ except during fast application of caffeine, when it was 12-15 ml min⁻¹. Miniature solenoid valves (The Lee Company, Essex, UK) were used to produce fast changes in the superfusate. These changes were electronically controlled using pCLAMP 7 trigger outputs. The speed of solution changes was calculated by monitoring changes in the electrode tip potential when a solution containing no KCl was changed with one containing 50 mm KCl. The time needed to obtain a complete change of solution around the electrode was 150-350 ms from the application of the trigger pulse, and was dependent on the position of the electrode in the bath.

Data acquisition and statistical analysis

The data obtained from the video edge-detection system, the epifluorescence apparatus and the Axoclamp-2B system were recorded on a computer using pCLAMP 7. The rate of sampling was between 0.5 and 3 kHz.

Time to peak (TTP) was measured between the point before the initial increase of the trace and the peak of the signal. Time to 50 % relaxation (T50) was measured between the peak of the signal and the point on the declining phase corresponding to half of the total size of the twitch. To assess statistical differences between means, Student's t test was used. Unless otherwise specified, the results are expressed as mean ± standard error of the mean (s.e.m.). n represents the number of myocytes that have undergone investigation.

RESULTS

The Na⁺-K⁺ pump-mediated Na⁺ extrusion was inhibited by rapidly removing K⁺ from the superfusing solution. Caffeine (15 mm) was then applied to obtain: (1) a rapid increase in cytoplasmic [Ca²⁺] and (2) the functional inhibition of SR Ca²⁺ uptake. In the presence of caffeine, Ca²⁺ extrusion and cell relaxation can be almost entirely ascribed to the function of the Na⁺-Ca²⁺ exchanger (Bridge et al. 1990; Bassani et al. 1994). Moreover, the electrogenic Na⁺-Ca²⁺ exchanger generates an inward current during removal of cytoplasmic Ca²⁺. This current was used to evaluate: (1) the amount of SR Ca²⁺ content (Varro et al. 1993) and (2) the rate of Ca²⁺ extrusion from the cytoplasm via the Na⁺-Ca²⁺ exchanger.

The protocol to measure caffeine-induced cell shortening and membrane current is shown in Fig. 1. Cells were impaled under the conditions described above and paced at 1 Hz using voltage-clamp steps of 400 ms in duration from -80 to +40 mV in 6K solution. Stimulation was stopped and cell membrane potential was held at -80 mV. After 1 s rest, 15 mm caffeine was applied for approximately 5 s. After a further 5 s in control 6K solution, a second application of caffeine for 5 s followed. After caffeine wash out, stimulation was restored until steady-state contraction was achieved. Stimulation was stopped again and the paired caffeine application protocol repeated but this time K⁺ was removed from the superfusing solution 400 ms before applying caffeine in each instance. The traces from the second caffeine application were subtracted from those obtained during the first application in order to abolish caffeine-dependent and Ca²⁺-independent components of the recordings. This is because the second caffeine application does not elicit Ca²⁺ release from the SR since the stores are still depleted from the previous caffeine application (Terracciano et al. 1995).

Figure 1. Protocols used to elicit paired caffeine applications in the presence and absence of K⁺.

After stopping stimulation, two rapid applications of caffeine separated by a 5 s interval were performed. The left panel shows the effects of paired caffeine application in 6 mm K⁺ (6K solution) on cell shortening and membrane current. A calcium-dependent transient inward current and contraction developed upon the first application of caffeine. These were absent during the second caffeine application presumably because the SR was still empty after the previous caffeine-induced release. A contracture still developed during the second application of caffeine and it can be explained by an increase in myofilament sensitivity to Ca²⁺. The right panel shows traces recorded from another cell during two consecutive applications of caffeine in the absence of K⁺. In this case a maintained inward current was present and could be ascribed to the effects of 0K solution on membrane conductances, presumably also to the inhibition of Na⁺-K⁺ pump current.

Removal of K⁺ produces an inward shift of membrane current (presumably also produced by the inhibition of the outward current ascribed to the Na⁺-K⁺ pump). This is Ca²⁺ independent and can be observed during the second application of caffeine. Subtraction of the two currents removes this component and allows comparison with currents recorded in the presence of K⁺, where the maintained inward current is not observed. In addition, if cell shortening is considered, the effect of caffeine on myofilament sensitivity to Ca²⁺ can also be abolished (this is the contracture observed during the second caffeine exposure and at the end of the first application of caffeine; Fig. 1, top trace of left panel). Finally, subtraction of Ca²⁺-independent effects is also useful to eliminate the possible effects of caffeine on the fluorescent indicator indo-1 (O'Neill et al. 1990).

Figure 2 shows caffeine-induced cell-shortening traces from the same cell in the presence and absence of K⁺ after the subtraction described above. When the Na⁺-K⁺ pump was rapidly inhibited, the cell relaxed more slowly in the continuous presence of caffeine. This suggests that Na⁺-Ca²⁺ exchanger-mediated relaxation is influenced by the function of the Na⁺-K⁺ pump. Results from six cells are presented in the lower panel of Fig. 2.

Figure 2. The effects of rapid removal of K⁺ on caffeine-induced cell shortening in the same cell.

In 0K, cell relaxation in the presence of caffeine, and mostly dependent on Na⁺-Ca²⁺ exchanger activity, was slowed compared with cell relaxation when 6K is present in the superfusing solution. This suggests that inhibition of the Na⁺-K⁺ pump leads to a reduced ability of the Na⁺-Ca²⁺ exchanger to remove Ca²⁺. The bar graph shows data for the monoexponential time constant fitted on the relaxation phase of the traces recorded in 6 cells. ** P < 0.05.

Figure 3 shows indo-1 fluorescence and membrane current subtractions from two consecutive caffeine applications in the presence and absence of K⁺ as described in Fig. 1. The time constant (τ) obtained by monoexponential fitting on the declining phase of the indo-1 fluorescence transient in 0K was greater than in 6K solution, suggesting that Ca²⁺ is removed more slowly when the Na⁺-K⁺ pump is inhibited and confirms the cell shortening observations.

Figure 3. Indo-1 fluorescence and membrane current changes elicited by application of caffeine in the presence and absence of extracellular K⁺.

The protocol described in Fig. 1 was applied and subtraction performed. When the Na⁺-K⁺ pump was rapidly inhibited, the indo-1 fluorescence transient declined more slowly compared with the one elicited in the presence of extracellular K⁺. The caffeine-induced transient inward current was also affected by removal of extracellular K⁺.

The caffeine-induced transient inward current (Fig. 3, lower traces) was also affected by removal of extracellular K⁺. The amplitude of the current (expressed as current density) was reduced by removal of K⁺ whereas TTP was increased. The rate of current decay was not significantly different in the presence or absence of K⁺. Mean data are shown in Table 1.

Table 1.

Effects of removal of K⁺ on indo-1 fluorescence and membrane current

	6K solution	0K solution	P value
Indo-1 ratio τ of decline (ms)	629 ± 29 (n = 11)	896 ± 81 (n = 11)	< 0.01
Indo-1 ratio amplitude	0.451 ± 0.08 (n = 11)	0.453 ± 0.09 (n = 11)	n.s.
Current TTP (ms)	151 ± 7 (n = 18)	189 ± 13 (n = 18)	< 0.01
Current τ of decay (ms)	276 ± 30 (n = 19)	260 ± 20 (n = 19)	n.s.
Current density (pA pF−1)	1.05 ± 0.1 (n = 20)	0.89 ± 0.08 (n = 20)	< 0.05
Current integral (pC)	47.1 ± 5 (n = 18)	46.8 ± 5 (n = 18)	n.s.
SR Ca2+ content (μmol (1 non-mitochondrial cell volume)−1)	28.9 ± 4 (n = 18)	28.5 ± 4 (n = 18)	n.s.

n.s., not significant; TTP, time-to-peak.

When the integrals of these currents were calculated (Fig. 4), no difference in the charge carried was detected. This suggests that, when the Na⁺-K⁺ pump is rapidly inhibited as described, no difference in SR Ca²⁺ content has occurred yet. The rate of charge movement was slower when K⁺ was absent in the superfusing solution, suggesting once again that Ca²⁺ is removed more slowly via the Na⁺-Ca²⁺ exchanger. The bar graph in the lower part of Fig. 4 shows the results obtained from six cells. Charge data are shown in Table 1.

Figure 4. Integral of caffeine-induced inward currents elicited by application of caffeine in 0K and 6K.

The total amount of charge was unchanged in the two groups but increased more slowly in the absence of K⁺ in the superfusing solution (data for T50 of decay is shown in the bar graph; n = 6; **P < 0.05). This result suggests that, although the same amount of Ca²⁺ is released from the SR upon application of caffeine, Ca²⁺ is removed more slowly when the Na⁺-K⁺ pump is inhibited. This finding supports the idea that under normal conditions, during Ca²⁺ extrusion, the Na⁺-Ca²⁺ exchanger produces an increase in subsarcolemmal [Na⁺]. Accumulation of Na⁺ modifies the ability of the Na⁺-Ca²⁺ exchanger to remove further Ca²⁺. When the Na⁺-K⁺ pump is inhibited this accumulation is affected and produces more obvious changes in Na⁺-Ca²⁺ exchanger-mediated Ca²⁺ extrusion.

Removal of extracellular K⁺ allows rapid inhibition of the Na⁺-K⁺ pump. However this technique is not specific and other ionic regulatory mechanisms, such as K⁺ channels, can be affected. To test that the effects observed are the consequence of specific inhibition of the Na⁺-K⁺ pump, strophanthidin was used. The inhibition of the pump achieved with this compound takes several seconds and depends on its concentration. Strophanthidin (500 μm) was used to achieve a maximal inhibition of high and low affinity components of the Na⁺-K⁺ pump (e.g. Gao et al. 1996). The following experiments assessed how long this concentration of strophanthidin takes to inhibit Na⁺-K⁺ pump current (I_p).

Cells were superfused with 6K solution and voltage clamped at -80 mV. To minimise membrane conductances and abolish Na⁺-Ca²⁺ exchanger current, 5 mm Ni²⁺ and 2 mm Ba²⁺ were added to the superfusing solution. After 2 min in this solution, 500 μm strophanthidin was applied using the same perfusion system described to remove K⁺ in the previous experiments. An inward shift in membrane current of 0.79 ± 0.09 pA pF⁻¹ (n = 12) ascribed to inhibition of I_p was observed. Ninety percent of this shift developed in 18.6 ± 2.8 s (n = 12) from the application of strophanthidin suggesting that, even at high concentration, strophanthidin takes about 20 s to inhibit approximately 90 % of Na⁺-K⁺ pump current. A similar protocol to that described in Fig. 1 was then performed but 500 μm strophanthidin was added to the superfusing solution instead of removing K⁺. The first application of caffeine was performed 20 s later to allow an almost complete inhibition of the Na⁺-K⁺ pump. The second application of caffeine was still performed after 5 s. Subtractions of the traces recorded during the second application of caffeine from the ones recorded during the first application were performed. Figure 5 shows how strophanthidin delays Ca²⁺ removal via the Na⁺-Ca²⁺ exchanger. The integral of caffeine-induced currents normalised to cell capacitance was not changed (6K solution: 0.44 ± 0.1 pC pF⁻¹; n = 5; 6K + strophanthidin: 0.48 ± 0.1 pC pF⁻¹; n = 5) and neither was the amplitude of the caffeine-induced indo-1 transient (6K solution: 0.51 ± 0.05 pC pF⁻¹; n = 11; 6K + strophanthidin: 0.50 ± 0.05 pC pF⁻¹; n = 11), suggesting that the same amount of Ca²⁺ released from the SR was present in control and after a 20 s strophanthidin application. These results, although obtained several seconds after stopping stimulation, confirm the observations obtained with K⁺ removal and support the role of the Na⁺-K⁺ pump in delaying Ca²⁺ extrusion via the Na⁺-Ca²⁺ exchanger with mechanisms independent of SR Ca²⁺ release.

Figure 5. Indo-1 fluorescence changes elicited by application of caffeine in the presence and absence of strophanthidin.

Twenty seconds after stopping stimulation, caffeine was applied twice, as described in Fig. 1, in the presence and absence of 0.5 mm strophanthidin. Subtraction was performed and the traces shown in the top panel were obtained. Monoexponential curves were fitted on the decay of these indo-1 fluorescence transients. The bar graph shows that the time constant calculated for these curves was significantly slower in the presence of strophanthidin than in control conditions (n = 11; **P < 0.01).

DISCUSSION

Rapid inhibition of the Na⁺-K⁺ pump: experimental approach and assumptions

The results reported suggest that rapid inhibition of the Na⁺-K⁺ pump in the absence of significant changes in SR Ca²⁺ content affects Ca²⁺ extrusion via the Na⁺-Ca²⁺ exchanger. A possible explanation for such findings is that subsarcolemmal [Na⁺] changes very rapidly after inhibition of the Na⁺-K⁺ pump, hence influencing the kinetics of the Na⁺-Ca²⁺ exchanger.

The absence of effects of ryanodine or caffeine (Bers, 1987) on the inotropic action of acetylstrophanthidin suggested that sarcolemmal mechanisms alone can play a part in Na⁺-K⁺ pump inhibition and inotropism, and the Na⁺-Ca²⁺ exchanger is the central mechanism in this effect. However, these changes had always been observed after minutes of application of cardiac glycosides, presumably because Na⁺-K⁺ pump inhibition was achieved quite slowly.

A very important feature of this study is that the Na⁺-K⁺ pump was rapidly inhibited and measurements were taken almost immediately. This is the major difference between this and other investigations (e.g. Su et al. 1998) where K⁺ was removed for several minutes before measurements were taken, in which time SR Ca²⁺ content changed and could have been responsible for the effects observed on Na⁺-Ca²⁺ exchanger function. In the present study changes in SR Ca²⁺ content did not have time to develop and the observed slower Na⁺-Ca²⁺ exchanger-mediated Ca²⁺ removal can be ascribed to a sarcolemmal or subsarcolemmal functional interaction between the Na⁺-K⁺ pump and the Na⁺-Ca²⁺ exchanger.

To achieve a rapid inhibition (in the time scale of a Ca²⁺ transient), extracellular K⁺ was removed to abolish the electrochemical gradient for the Na⁺-K⁺ pump. This, coupled with a rapid switching of superfusing solutions, allowed us to observe effects only after ≈400 ms. In this time course, previous studies have shown that bulk cytoplasmic [Na⁺] does not change, and that removal of K⁺ produces a significant increase in activity only after several minutes (Eisner et al. 1981). To explain this discrepancy one hypothesis is that K⁺ removal has other effects on Na⁺-Ca²⁺ exchanger function together with inhibiting the Na⁺-K⁺ pump, and these could be involved. In other words, this technique could be non-specific and some other unknown mechanisms could be involved. To test this possibility, a more specific inhibition of the pump was sought and the experiments reported in Fig. 5 were performed. In identical conditions to those used during the removal of K⁺, 0.5 mm strophanthidin was applied. However, since at this concentration strophanthidin requires approximately 20 s to produce an average 90 % inhibition of the pump, measurements were performed after a 20 s interval. With this protocol a similar slowing of the caffeine-induced Ca²⁺ transient was found, suggesting that in the experiments described above, the effects of K⁺ removal on Ca²⁺ extrusion can be ascribed to the inhibition of the Na⁺-K⁺ pump. The experiments with strophanthidin cannot be used as substitutes for the experiments performed after K⁺ removal because of the different time course of the Na⁺-K⁺ pump inhibition. However, the evidence that similar results are found with both techniques confirms that the effects observed after K⁺ removal can be ascribed to Na⁺-K⁺ pump inhibition.

Subsarcolemmal [Na⁺] regulation affects Ca²⁺ removal and cell relaxation

Since bulk [Na⁺] is unlikely to increase in the short time interval used during K⁺ removal, changes in the function of the Na⁺-Ca²⁺ exchanger in removing Ca²⁺ could be explained by the existence of a restricted area of [Na⁺] in the subsarcolemma, which cannot be detected with microelectrode or fluorescence techniques. It is subsarcolemmal [Na⁺] in this area that is sensed by the Na⁺-Ca²⁺ exchanger, and the one that regulates its function.

Subsarcolemmal [Na⁺] can rapidly increase if the Na⁺ channels open (during the action potential) but in the experiments reported here the holding potential was -80 mV and it is unlikely that the Na⁺ channels were opened at the time of caffeine application. Nevertheless, it is possible that when caffeine is applied a subsarcolemmal ‘Na⁺ transient’ could still be declining from the last stimulus. Na⁺-K⁺ pump activity may affect this decline, leading to a higher level of Na⁺.

Another possibility is that Na⁺ enters via the Na⁺-Ca²⁺ exchanger itself when this mechanism brings Na⁺ into the cell to exchange it with Ca²⁺ during Ca²⁺ extrusion. If one considers the caffeine-induced transient inward current and a stoichiometry of 3Na⁺:1Ca²⁺ for the Na⁺-Ca²⁺ exchanger, this amount can be calculated to be a few hundred micromoles and therefore not able to affect bulk [Na⁺]. However, there is evidence that an area of restricted diffusion (‘fuzzy space’) for Na⁺ exists. Lederer et al. (1990) were the first to suggest the presence of an area of restricted diffusion where a large increase in [Na⁺] can be achieved by a relatively small Na⁺ entry. If the volume of the restricted area is 0.3 % of the total volume (as proposed by Lederer et al. 1990), Ca²⁺ extrusion via the Na⁺-Ca²⁺ exchanger during caffeine application can produce millimolar changes in subsarcolemmal [Na⁺]. The function of the Na⁺-Ca²⁺ exchanger itself is affected by this increase in subsarcolemmal [Na⁺] resulting in a slower Ca²⁺ extrusion. Such an increase in subsarcolemmal [Na⁺] is presumably prevented or reduced by a functional Na⁺-K⁺ pump. If the pump is inhibited the increase in [Na⁺] is more evident. The results shown here seem to support this hypothesis.

In this and previous studies, a temporal lag between cytoplasmic [Ca²⁺] changes and current has been observed. This has been ascribed to an area of restricted diffusion for Ca²⁺ (Trafford et al. 1995). In the present study a slower Ca²⁺ extrusion was associated with an unchanged current decay when the Na⁺-K⁺ pump was inhibited. This suggests that changes in [Na⁺] regulation can affect the relationship between Ca²⁺ removal and membrane current, supporting once again the existence of an area of restricted diffusion for Na⁺. Whether the two areas of restricted diffusion for Ca²⁺ and Na⁺ are coincident is not known. Blaustein & Lederer have recently discussed the existence of the ‘PlasmERosome’, an area comprising plasmalemmal mechanisms including the Na⁺-Ca²⁺ exchanger and the Na⁺-K⁺ pump, the junctional SR and the ‘intervening restricted cytoplasmic space’ (Blaustein & Lederer, 1999). The ‘PlasmERosome’ is the ideal candidate for explaining the diffusion properties of Na⁺ and Ca²⁺ observed in this and other studies.

Calculations to explain the results based on models of Ca²⁺ extrusion via the Na⁺-Ca²⁺ exchanger

For a better understanding of the phenomenon involved in the functional interactions between Na⁺-K⁺ pump and Na⁺-Ca²⁺ exchanger, quantification of changes in subsarcolemmal [Na⁺] was sought. However, this could not be done using electrophysiological techniques. Calculation of the reversal potential of the Na⁺-Ca²⁺ exchanger (E_Na-Ca), which is dependent on subsarcolemmal [Na⁺], has been tried using a technique previously described (Kimura et al. 1986, 1987; Terracciano et al. 1998). However, in these experimental conditions, E_Na-Ca follows the holding potential resulting in unreliable [Na⁺] quantification (Convery & Hancox, 1999). Measuring E_Na by calculating the reversal potential for the tetrodotoxin (TTX)-sensitive Na⁺ current has also been attempted. The inaccuracy of the clamp technique during the first few milliseconds of the voltage step, when most of the Na⁺ current flows, made these experiments inconclusive. Fluorescence techniques can be excluded because they only allow calculation of bulk [Na⁺]. In order to discuss the behaviour of subsarcolemmal [Na⁺] this parameter was therefore calculated using the experimental data described in this paper and well-known equations that model Na⁺-Ca²⁺ exchanger current.

The model is shown in Fig. 6. In the upper panel the recorded membrane current (a) and a calculated current from the caffeine-induced Ca²⁺ transient were compared (b). The latter was calculated using the equation described by DiFrancesco & Noble (1985) and Beuckelmann & Wier (1989) with the assumption that [Na⁺]_i is 6 mm and it does not change during caffeine application. A correction was then made for the restricted diffusion for Ca²⁺ described by Trafford et al. (1995) and subsarcolemmal [Ca²⁺] ([Ca²⁺]_m) was calculated from bulk [Ca²⁺] ([Ca²⁺]_b) using eqn (4) from this study:

The constant γ was assumed to be 133 ms as calculated by Trafford et al. (1995). However, even after this correction, a current different from the one recorded was still obtained (c).

Figure 6. Calculations of subsarcolemmal [Na⁺].

The upper panel shows the recorded caffeine-induced transient inward current (a), a calculated current from the caffeine-induced Ca²⁺ transient (b) and the latter modified for the presence of a diffusion barrier for Ca²⁺ (c). To calculate currents b and c [Na⁺]_i = 6 mm. The lower panel shows the calculated changes in [Na⁺]_i that would explain the difference between currents a and c. These changes are concomitant with Ca²⁺ extrusion via the Na⁺-Ca²⁺ exchanger and can be explained by the existence of restricted diffusion for Na⁺.

If one assumes that [Na⁺] varies and it is responsible for this lag, [Na⁺] changes are calculated and described in Fig. 6, lower panel. These [Na⁺] changes are theoretically able to explain the lag between [Ca²⁺] and membrane current. Since Na⁺ entry via the Na⁺-Ca²⁺ exchanger during Ca²⁺ extrusion can only increase bulk [Na⁺] by a few micromoles, it is possible that it accumulates and diffuses slowly under the sarcolemma, representing a negative feedback mechanism for the Na⁺-Ca²⁺ exchanger. This observation suggests that, together with an area of restricted diffusion for Ca²⁺, an area of restricted diffusion for Na⁺ is possible, so that a Na⁺ transient can be obtained during removal of Ca²⁺.

The effects on this Na⁺ transient when the Na⁺-K⁺ pump is inhibited are shown in Fig. 7. The decay of this [Na⁺] transient is slower, suggesting that part of the decay is due to Na⁺ extrusion via the Na⁺-K⁺ pump. This affects the Na⁺-Ca²⁺ exchanger ability to remove Ca²⁺, resulting in slower relaxation.

Figure 7. The effects of inhibition of the Na⁺-K⁺ pump on Na⁺ diffusion from the subsarcolemmal space.

The upper panel shows changes in subsarcolemmal [Na⁺] during application of caffeine, calculated in the same cell in the presence and absence of K⁺. Monoexponential curves were fitted to these traces to calculate the rate of decay of [Na⁺] changes. The time constant was significantly larger in the absence than in the presence of extracellular K⁺ (lower panel). This suggests that inhibition of the Na⁺-K⁺ pump delays the diffusion of Na⁺ away from the Na⁺-Ca²⁺ exchanger and could affect Na⁺-Ca²⁺ exchanger function (n = 5; **P < 0.05).

In conclusion, this study shows that inhibiting the Na⁺-K⁺ pump in the absence of changes in SR Ca²⁺ content affects the ability of the Na⁺-Ca²⁺ exchanger to remove Ca²⁺. This is achieved in the time course of a Ca²⁺ transient. This phenomenon can be explained, amongst other hypotheses, by a slower removal of Na⁺ from the subsarcolemmal space, thereby affecting the Na⁺-Ca²⁺ exchanger function. The source of this temporary increase in Na⁺ could be the Na⁺-Ca²⁺ exchanger itself and can be explained by the existence of a subsarcolemmal area of restricted diffusion for Na⁺.