A comparison of the effects of ATP and tetracaine on spontaneous Ca²⁺ release from rat permeabilised cardiac myocytes

Abstract

1. Fluo-3 fluorescence measurements were made in isolated β-escin permeablised rat cardiac myocytes using confocal microscopy. Perfusion of a mock intracellular solution containing 0.22-0.23 μm Ca²⁺ and 5 mm ATP elicited regular waves of Ca²⁺ (approximately every 5 s) due to spontaneous release of Ca²⁺ from the sarcoplasmic reticulum (SR).

2. An approximately linear relationship was noted between Ca²⁺ wave velocity (v) and amplitude (σ). Under the control conditions the ratio of velocity to amplitude (v/σ) varied little and was 99.8 ± 2.5 m s⁻¹μm⁻¹(n = 78).

3. Reduction of [ATP] in the bathing solution to 0.5 and 0.2 mm ATP progressively decreased Ca²⁺ wave frequency and propagation velocity while increasing the amplitude. The changes in Ca²⁺ wave characteristics in 0.5 mm ATP were similar to those observed during perfusion with 50 μm tetracaine. In 0.2 mm ATP the decline of [Ca²⁺] during a Ca²⁺ wave was slowed suggesting a lowered rate of Ca²⁺ re-uptake by the SR Ca²⁺ pump.

4. Reduction of [ATP] to 0.1 mm abolished Ca²⁺ waves after 15-20 s. Returning the [ATP] to 5 mm caused a burst of high frequency and large amplitude waves. Mean velocity of the first wave on returning to 5 mm ATP was larger than normal but the v/σ value was 32 ± 6 % of control (n = 6). In the similar burst on removal of 100 μm tetracaine v/σ was higher than control (166 ± 9 %, n = 6).

5. Rapid application of caffeine (10 mm) was used to assess the SR Ca²⁺ content. This showed that SR Ca²⁺ increased as [ATP] was reduced or [tetracaine] was increased. The highest SR Ca²⁺ content was observed after perfusion with 0.1 mm ATP, which was 245 ± 15 % of control values.

6. Returning [ATP] from 0.1 mm to 5 mm caused a burst of high frequency, large amplitude Ca²⁺ waves. But recovery after incubation with 300 μm tetracaine resulted in SR Ca²⁺ release with no coherent wave pattern. The reason for this discrepancy is discussed.

Decreased myocardial intracellular ATP concentration ([ATP]) occurs during prolonged periods of ischaemia. Early during the ischaemic period, intracellular phosphocreatine concentration ([CrP]) decreases and inorganic phosphate concentration [Pi] increases as the Lohmann reaction supplements intracellular ATP production in the absence of aerobic glycolysis (Bailey et al. 1981; Allen et al. 1985). After [CrP] has fallen to undetectable levels (< 0.1 mm), anaerobic glycolysis can sustain ATP production for considerable periods of time. ³²P NMR studies on isolated hearts suggest that the normal [ATP] (5-7 mm) is sustained for approximately 30 min during global ischaemia, before gradually declining to undetectable levels (< 0.1 mm) over the subsequent 30 min (Elliott et al. 1992). Changes of intracellular [Ca²⁺] are complex, particularly in the latter stages of ischaemia. Early in ischaemia, the Ca²⁺ transient amplitude and duration increase due to the accompanying intracellular acidosis (Steenbergen et al. 1987; Allen et al. 1988). Later, excitation-contraction (E-C) coupling is inhibited and replaced by regular oscillations of intracellular [Ca²⁺] due to spontaneous SR Ca²⁺ release (Lee & Allen, 1992). Abnormal SR Ca²⁺ release generates Ca²⁺-activated arrhythmic currents and may be the cause of arrhythmic activity observed during ischaemia (Opie & Clusin, 2000).

Reperfusion of the myocardium after moderate to long periods of ischaemia restores intracellular [ATP] towards normal but this phase has been associated with a paradoxical increase of intracellular [Ca²⁺] involving spontaneous SR Ca²⁺ release and re-uptake (Kihara et al. 1989). Again, these oscillations of [Ca²⁺] are thought to be the cause of arrhythmic behaviour commonly observed immediately on reperfusion in both experimental animals and in humans (Chess-Williams et al. 1990). Investigators have identified a number of different cellular mechanisms that contribute to the reperfusion paradox; including decreased intracellular and extracellular pH, increased intracellular [Na⁺] and the generation of oxygen-derived free radicals (Grinwald, 1982; Hess & Manson, 1984). Despite the significant changes of intracellular [ATP] during ischaemia and reperfusion, it is not clear to what extent ATP can modulate E-C coupling or spontaneous Ca²⁺ release. ATP is required for the normal function of the cardiac ryanodine receptor type 2 (RyR2) and the sarcoplasmic/endoplasmic reticulum Ca²⁺-ATPase type 2 (SERCA2), but the net effect of reduced [ATP] is difficult to predict. Recent work suggests that reduced [ATP] modulates RyR activity in such a way as to decrease the frequency and increase the amplitude of spontaneous Ca²⁺ release (Yang & Steele, 2000).

The purpose of the present study was to examine the characteristics of spontaneous Ca²⁺ release using the high time and spatial resolution of confocal fluorescence microscopy. Cellular [ATP] was reduced (in the absence of CrP) for varying periods and then restored to mimic both ischaemia and reperfusion.

METHODS

Cell isolation

Wistar rats (200-250 g) were killed by stunning and cervical dislocation. Rat myocytes were isolated as previously described (Eisner et al. 1989). Briefly, isolated hearts were perfused retrogradely via the aorta (5 ml min⁻¹, 37 °C) with a nominally Ca²⁺-free Krebs-Henseleit solution for 10 min. This was followed by perfusion for 10 min with re-circulated Krebs-Henseleit solution supplemented with 0.6 mg ml⁻¹ collagenase (type 1, Worthington Chemical Co), 0.1 mg ml⁻¹ protease (type XIV, Sigma) and 80 μm CaCl₂. The enzyme-containing solution was subsequently washed-out by perfusion with a nominally Ca²⁺-free Krebs-Henseleit solution containing 1 % bovine serum albumin (BSA, fraction V, Sigma) for approximately 10 min before both left and right ventricles were removed. This tissue was finely chopped and gently triturated to release cells. Cells were maintained in this solution at a concentration of approximately 1 × 10⁴ cells ml⁻¹ until use.

Cell permeabilisation and fluorescence measurements

β-Escin (Sigma) was added from a freshly prepared stock solution to a 1 ml aliquot of cell suspension to give a final concentration of 0.1 mg ml⁻¹. This solution was gently stirred for 10-15 s and the β-escin was removed by gentle centrifugation and resuspending the cells in mock intracellular solution (see below). The cells were then placed in a small bath and allowed to settle onto the coverslip that formed the base.

Solution composition

Permeabilised cells were perfused with a mock intracellular solution with the following composition (mm): 100 KCl, 5 Na₂ATP, 5.4 MgCl₂, 25 Hepes, 0.1 K₂EGTA, pH 7.0. The mitochondrial inhibitors carbonyl cyanide (20 μm) and oligomycin (20 μm) (Calbiochem) were included to ensure that mitochondrial ATP production was negligible. Confocal measurements of [Ca²⁺] were obtained by the inclusion of 10 μm fluo-3 (Molecular Probes) in the solution. In these experiments, the [ATP] was varied between 0.1 and 5 mm by creating a second solution with matched composition to the control solution but without ATP and with a lower total but equivalent free Mg²⁺ concentration (1 mm MgCl₂). Mixing the control solution (5 mm ATP) and the ‘zero ATP’ solution to varying proportions created the range of ATP concentrations used in this study. The free [Ca²⁺] in the experimental solutions was adjusted to 220-230 nm; this concentration was found to produce spontaneous Ca²⁺ release from permeabilised myocytes with a frequency of approximately 0.2 Hz in isolated rat myocytes.

Data recording and analysis

Confocal linescan images were recorded using either a BioRad MRC 1024 or a BioRad Radiance 2000 confocal system. Fluorescence was excited at 488 nm and measured above 515 nm using epifluorescence optics of a Nikon Eclipse inverted microscope with Nikon Fluor × 60 oil objective. The data were acquired in linescan mode at 2 ms line⁻¹ and nominal pixel dimension ranged from 0.41 μm (zoom = 1) to 0.20 μm (zoom = 2). An example of a typical linescan image is shown in Fig. 1. Spontaneous release of Ca²⁺ from the SR occurred approximately in the middle of the cell and propagated to either end generating a commonly observed V-shaped linescan profile. The upper part of the V-shaped record is linear indicating propagation along the cell length at a constant velocity. However, the lower part of the record is not linear, containing a number of discontinuities, suggesting this section of the record is interrupted by cell movement and therefore was not suitable for analysis. An experimental record typically contained 150-200 linescan images; these were reviewed off-line and one common region (typically 40 pixels in length) was selected on the basis that uniform propagated wave-fronts could be identified in the majority (> 90 %) of the images. Using software written for the purpose, the mean fluorescence from 4 × 10 pixel segments of the 40-pixel section was calculated in each linescan image and accumulated into a single data file. A typical record obtained using this method is shown in Fig. 1Ab. The rapid upstroke of fluorescence represents the rapid release of Ca²⁺ from the SR; the decay to baseline levels is due to Ca²⁺ re-uptake by the SR and Ca²⁺ diffusion from the cell. The delays between traces represent the propagation times between sites, from these and the positions of the sites the velocity can be calculated. The four transients generate three velocity measurements; the mean value of these three measurements is shown in Fig. 1Ba for a continuous record of 18 transients. This procedure has two main advantages: (i) propagation velocity values from a series of Ca²⁺ waves are obtained from the same part of the cell thus minimising variation due to sub-cellular structures and (ii) the procedure can be automated and applied to ≈200 sequential linescan images in a single analysis run. The automated analysis included a measure of wave amplitude, thus each wave-front provided three parameters: velocity, amplitude and period (time between waves).

Figure 1. Relationship between Ca²⁺ wave amplitude and velocity.

Aa, a linescan image of a Ca²⁺ wave in a permeabilised rat ventricular myocyte perfused with a mock intracellular solution containing 220 μm Ca²⁺. The calibration bars indicate time and distance. Fluorescence intensity is linearly related to grey scale (see calibration bar). The horizontal lines through a wave-front represent 4 × 10 pixel bands (i.e. 4 × 2.1 μm bands). Ab, the mean fluorescence in each band (10 pixels) is plotted against time. The left-hand scale is in fluorescence units (F_o is the fluo-3 fluorescence at 225 μm Ca²⁺). The scale on the right-hand side is the corresponding values of [Ca²⁺]. Ba, a record of Ca²⁺ transients due to regular Ca²⁺ waves, the [Ca²⁺] derived from the average fluo-3 fluorescence from a 10-pixel band. The plot below is the mean velocity of the corresponding wave-front (40 pixel (8.4 μm) length). Bb, plot of transient amplitude (σ) against velocity (v); the line through the points is the best-fit linear regression (P < 0.001).

Calibration of fluorescence signal

Data from linescan image files were initially converted to normalised fluorescence units (F/F₀), where F₀ is the fluo-3 fluorescence measured in 225 nm Ca²⁺ relative to that measured in < 1 nm Ca²⁺. These values were then converted to [Ca²⁺] using a relationship derived by Cheng et al. (1993) in which:

where R is the fluorescence ratio (F/F₀), [Ca²⁺]_ref is the [Ca²⁺] in the perfusion solution (225 nm) and K_d is the Ca²⁺ affinity of fluo-3. The K_d of the indicator fluo-3 was measured as 0.56 ± 0.06 μm(n = 4) in a solution with a composition similar to the mock intracellular solution used in this study (using 10 mm EGTA to buffer Ca²⁺).

Statistics

Changes in wave velocity, amplitude and period are expressed as means ± standard error of the mean. Student's t tests were used to compare the relative changes; P < 0.05 was considered statistically significant.

RESULTS

Figure 1A and B illustrates the method for analysis of linescan images. Perfusing the permeabilised cell with a weakly Ca²⁺ buffered solution containing 220-230 nm Ca²⁺ resulted in Ca²⁺ waves at 3-6 s intervals as a result of propagated SR Ca²⁺ release and re-uptake. In Fig. 1Ba the signal represents the mean fluorescence signal in a 10-pixel segment of a line scan from a cell approaching steady-state activity. This represents the final 2 min of a 5 min recovery period following exposure to tetracaine (see later). This was chosen to ensure a wider range of wave amplitude than was seen in control conditions. The mean velocity of the wave-front is plotted at the appropriate point below the trace. Ca²⁺ wave amplitude varied by ≈50 % over the time scale of the recording. The relationship between velocity and amplitude is illustrated in Fig. 1Bb, where the mean amplitude of spontaneous release amplitude in a 40-pixel section is plotted against the corresponding mean velocity of the wave-front. The relationship is approximately linear over the range of wave amplitudes studied.

The effects of low [ATP] on spontaneous wave propagation

One of the terminal events during metabolic blockade in cardiac muscle is a reduced cytosolic [ATP] concentration. To examine the effects of this on the characteristics of spontaneous release, [ATP] in the perfusing solution was reduced from 5 to 0.5 mm (Fig. 2A) and to 0.2 mm (Fig. 2B). In each panel, the mean fluorescence signal from a 10-pixel segment is shown. Below this trace is shown the mean velocity of the wave-front calculated over a 40-pixel section (v). The lower trace is the ratio of mean velocity (v) to the amplitude of the wave-front (σ) in the 40-pixel segment (v/σ). On reduction of [ATP] to 0.5 mm ATP, wave amplitude increased, but with no accompanying increase in propagation velocity. Over 30-60 s, the velocity of the wave-front decreased, but not significantly below the control (5 mm ATP) values. v/σ showed little change in the control period before reduction in [ATP] as expected from Fig. 1. However, in low [ATP] the value of v/σ decreased to approximately 50 % of control. On re-establishing 5 mm ATP, wave amplitude returned to control values. Lowering [ATP] to 0.2 mm (Fig. 2B) caused a marked increase in wave amplitude and decrease in velocity. In this instance, v/σ fell to ≈30 % of control. On restoring [ATP] to control values, the amplitude and frequency of the spontaneous waves increased above the control value, with a dramatic increase in the velocity of the wave. However, v/σ remained close to the values achieved in the presence of 0.2 mm ATP and slowly increased towards control values over 20-30 s.

Figure 2. The effect of low ATP on Ca²⁺ wave amplitude, frequency and velocity.

A, records of [Ca²⁺] derived from the average fluo-3 fluorescence from a 10-pixel band (2.9 μm); the mean velocity of a 40 pixel wide (11.6 μm) wave-front. Calculated ratio of velocity/wave amplitude (v/σ) for each wave is shown in the lower section. The filled bar above the trace indicates period of perfusion with 0.5 mm ATP. B, records of: [Ca²⁺], wave velocity and v/σ values for spontaneous Ca²⁺ waves during perfusion with 0.2 mm ATP and recovery.

Low [ATP] abolishes spontaneous release in permeabilised cardiac myocytes.

Figure 3A and B shows typical fluorescence signals recorded with an [ATP] of 0.1 mm. Immediately after reduction of [ATP], Ca²⁺ wave frequency and velocity of propagation were reduced. After two to three waves, spontaneous release ceased for periods of up to 100 s (see later). Returning to 5 mm ATP caused a marked increase in wave amplitude and frequency. During this period, the velocity of wave propagation was higher than the last waves observed in 0.1 ATP, but less than control values. The values of v/σ during this burst were similar to the minimum values obtained from the last waves measured in the presence of 0.1 mm ATP. However, as the frequency of oscillations declined and the amplitude returned to normal, v/σ increased to control values. This result was observed in 14 other cells: mean values for the changes in amplitude, velocity and v/σ values are given in Table 1A. In two cells, spontaneous Ca²⁺ release was maintained in the presence of 0.1 mm ATP, but at a much lower frequency and larger amplitude than control (similar to that observed in 0.2 mm ATP). The averaged values in Table 1 exclude measurements from these cells.

Figure 3. The effect of 0.1 mm ATP on Ca²⁺ wave amplitude and velocity.

A, records of [Ca²⁺] derived from the average fluo-3 fluorescence from a 10-pixel band (2.1 μm); the mean velocity of a 40-pixel wave-front. Calculated ratio of velocity/wave amplitude (v/σ) for each wave is shown in the lower section. The continuous bar above the trace indicates period of exposure to low ATP. B, record of Ca²⁺ transients associated with spontaneous Ca²⁺ waves and rapid application of caffeine. The filled bar above the trace indicates period of exposure to 0.1 mm ATP. Caffeine (10 s duration) was applied as indicated by the arrows below the trace.

Table 1A.

Changes in characteristics of spontaneous release of Ca2+ in the presence of low ATP or tetracaine (values expressed relative to control)

	Amplitude (σ)	Velocity (v)	v/σ	Period	Caffeine amp
In 0.5 mm ATP (n = 14)	1.72 ± 0.11	0.72 ± 0.07	0.41 ± 0.03	1.16 ± 0.11	1.97 ± 0.09 (7)
In 0.2 mm ATP (n = 14)	2.20 ± 0.13	0.49 ± 0.05	0.27 ± 0.07	2.10 ± 0.05	2.33 ± 0.12 (6)
In 0.1 mm ATP* (n = 14)	1.99 ± 0.16	0.61 ± 0.08	0.31 ± 0.11	1.55 ± 0.07	2.45 ± 0.15 (7)
In 50 μm tetracaine (n = 12)	1.22 ± 0.08	0.83 ± 0.03	0.76 ± 0.05	1.15 ± 0.03	1.31 ± 0.08 (6)
In 100 μm tetracaine* (n = 11)	0.97 ± 0.02	1.09 ± 0.05	1.03 ± 0.05	0.95 ± 0.04	1.73 ± 0.17 (5)
In 300 μm tetracaine* (n = 12)	1.04 ± 0.02	0.97 ± 0.02	1.04 ± 0.05	1.06 ± 0.03	1.95 ± 0.19 (6)

^*0.1 mm ATP, 100 and 300 μm tetracaine inhibited spontaneous release of Ca²⁺. Therefore these values represent the mean of the last wave recorded under these conditions. Period is the mean time relative to control) between the last and 2nd last spontaneous release. The values in 0.5 and 0.2 mm ATP and 50 μm tetracaine represent steady-state values (relative to control). † Caffeine responses were measured in a subset of cells; the numbers are shown in parentheses.

As shown in Fig. 3B, caffeine was applied to assess SR Ca²⁺ content under control conditions and in the presence of 0.1 mm ATP when all spontaneous activity had been abolished. The amplitude of the caffeine response in 0.1 mm ATP was clearly larger than control suggesting the Ca²⁺ content of the SR was increased. The mean changes in the amplitude of caffeine responses are also given in Table 1A.

Time course of the spontaneous Ca²⁺ release in low [ATP]

Normal spontaneous Ca²⁺ release from the SR is rapid in on-set with a slower decay phase representing Ca²⁺ re-uptake by the SR and efflux from the myocyte. It is conceivable that both the rate of Ca²⁺ release and the rate of re-uptake are influenced by reduced [ATP]. This can be compared by examining the fluorescence record from a 10-pixel segment recorded under control conditions and in the presence of reduced [ATP] (Fig. 4). It is clear from the records (Fig. 4A and B) that while the rate of rise of [Ca²⁺] appeared to be unaffected by perfusion with either 0.5 or 0.2 mm ATP, the rate of decay of the Ca²⁺ transient is substantially reduced in 0.2 mm ATP. The decay phase of the fluorescence record was fitted with a single exponential; the mean time constant for the decay under control conditions was 0.29 ± 0.2 s (n = 13). In the presence of 0.5 mm ATP, the time constant was not significantly different from control (106 ± 5 %, n = 7). However, in the presence of 0.2 mm ATP, the time constant was significantly longer than control (201 ± 16 %, n = 6) suggesting that Ca²⁺ uptake rate was reduced. The rate of rise of Ca²⁺ was quantified by measuring the time between 10 and 90 % of the amplitude of the Ca²⁺ wave. The mean value under control conditions was 28 ± 2 ms, this was not significantly altered in 0.5 mm ATP (101 ± 3.2 %, n = 7) or 0.2 mm ATP (109 ± 5 %, n = 6).

Figure 4. Effects of perfusion with low ATP on the time course of the Ca²⁺ wave.

Aa, a typical fluorescence record of a Ca²⁺ wave (mean fluorescence from 10 pixels, i.e. 2.1 μm) under control conditions (5 mm ATP) and in the presence of 0.5 mm ATP. The best-fit exponential decay time constant (t) for decay phase of the Ca²⁺ transient is shown beside each transient. Ba, corresponding Ca²⁺ transient recorded in another cell under control conditions in 0.2 mm ATP. The up-strokes of transients are shown below the corresponding records (Ab and Bb).

The effect of tetracaine on the characteristics of spontaneous Ca²⁺ release in permeabilised cardiac muscle

Tetracaine is an anaesthetic agent with known effects on the Ca²⁺ sensitivity of the RyR, but with little effect on the SR Ca²⁺ pump. Therefore it was used as a reference compound for comparison with the effects of reduced [ATP]. As shown in Fig. 5A, 50 μm tetracaine decreased wave frequency, increased amplitude and slowed propagation slightly. v/σ was reduced during exposure to tetracaine. As with low ATP, wave frequency increased when tetracaine was removed although wave amplitude was not significantly greater than control. In addition, the initial wave propagation velocity was greater than expected from the wave amplitude, thus the v/σ value transiently increased above control values on removal of the drug. Increasing the concentration of tetracaine to 100 μm completely inhibited spontaneous release for periods in excess of 100 s in 11 out of 13 cells studied. However, during the quiescent period, the noise on the fluorescence signal increased progressively, reaching maximum values after approximately 50 s. In two cells that did maintain spontaneous release in 100 μm tetracaine, wave amplitude was larger than control but with reduced frequency. These atypical cells were excluded from the analysis. Removal of 100 μm tetracaine after perfusion for 100-120 s caused a transient increase in frequency of Ca²⁺ waves but wave amplitude was less than control. As can be seen from Fig. 5B, although the velocities of the waves within the burst were less than control values, v/σ values were initially higher. In parallel experiments, the response to caffeine (10 mm) under control conditions and after perfusion with tetracaine (50, 100 and 300 μm) for 100-120 s were studied. As shown in Table 1, tetracaine increased the Ca²⁺ content of the SR as assessed by caffeine-induced Ca²⁺ release in a dose-dependent manner. The use of 300 μm tetracaine also rapidly abolished Ca²⁺ waves, but removal of tetracaine after ≈100 s exposure did not result in coherent waves.

Figure 5. The effects of tetracaine on Ca²⁺ wave amplitude and frequency velocity.

Records of [Ca²⁺] derived from the average fluo-3 fluorescence from a 10-pixel band (2.1 μm); the mean velocity of a 40-pixel wave-front (11.6 μm). Calculated ratio of velocity/wave amplitude (v/σ) for each wave is shown in the lower section. The continuous bar above the trace indicates period of exposure to tetracaine: A, 50 μm; B, 100 μm tetracaine.

The differences in the characteristics of the Ca²⁺ waves observed after exposure to 0.1 mm ATP and tetracaine are highlighted in Fig. 6. Comparing the last wave in 0.1 mm ATP with the first wave observed immediately on returning to control conditions (5 mm ATP) indicates that the two events are similar. However, when 300 μm tetracaine was removed, multiple wave-fronts of varying amplitude were observed. This less organised form of burst of Ca²⁺ waves after 300 μm tetracaine makes analysis of the Ca²⁺ waves difficult, so no data for such waves are included in Table 1B. Figure 6C shows that a single, coherent wave follows removal of 300 μm tetracaine after a shorter exposure.

Figure 6. Linescan images of Ca²⁺ waves in permeabilised cardiac myocytes.

A shows the last wave recorded in 0.1 mm ATP and the first wave after return of [ATP] to 5 mm after 80 s. B shows linescan images of the last wave recorded in 0.3 mm tetracaine and the less coherent wave patterns recorded on removal of tetracaine after 80 s. C shows linescan images of the last wave recorded in 0.3 mm tetracaine and the distinct first wave recorded on removal of tetracaine after 40 s.

Table 2A.

Changes in characteristics of 1st spontaneous release of Ca²⁺ after low ATP or tetracaine (values expressed relative to control)

	Amplitude (σ)	Velocity (v)	v/σ	Period	Caffeine amp
After 0.5 mm ATP (n = 7)	1.53 ± 0.18	1.02 ± 0.04	0.71 ± 0.05	0.82 ± 0.05	1.97 ± 0.09 (7)
After 0.2 mm ATP (n = 8)	2.54 ± 0.21	0.75 ± 0.07	0.35 ± 0.05	0.38 ± 0.08	2.33 ± 0.12 (6)
After 0.1 mm ATP* (n = 8)	2.67 ± 0.24	0.63 ± 0.14	0.32 ± 0.08	0.34 ± 0.04	2.45 ± 0.15 (7)
After 50 μm tetracaine (n = 6)	1.01 ± 0.11	1.25 ± 0.03	1.28 ± 0.11	0.74 ± 0.07	1.31 ± 0.08 (6)
After 100 μm tetracaine*(n = 6)	0.52 ± 0.06	0.78 ± 0.09	1.66 ± 0.09	0.25 ± 0.05	1.73 ± 0.17 (5)
After 300 μm tetracaine (n = 6)	—	—	—	—	1.95 ± 0.19 (6)

^*See Table 1A. Period is the mean time (relative to control) from the 1st spontaneous release to the 2nd. Caffeine responses are as described in Table 1A.

Changes in signal variance during inhibition of spontaneous Ca²⁺ release

The increase in the noise of the fluorescence signal in tetracaine was in contrast to that observed during 0.1 mm ATP. This difference is highlighted in Fig. 7A and B, which compares the response to 100 μm tetracaine with that to 0.1 mm ATP. In Fig. 7C, linescan images early and late during the quiescent period in tetracaine indicate that the increased noise on the fluorescence signal is due to the increased incidence of Ca²⁺ sparks. The noise of the fluorescence signal (a 40-pixel segment) was quantified by measuring the variance of the signal. The calculated variance was normalised with respect to the value immediately following the last spontaneous Ca²⁺ release event to prevent cell-to-cell variations in absolute cell fluorescence obscuring changes in the average variance values. Figure 7C shows a significant increase in variance after approximately 60 s from the last spontaneous release in 100 μm tetracaine. In the presence of 300 μm tetracaine, the rise in variance is delayed. However, in 0.1 mm ATP there are no time-dependent changes in variance. These results suggest the reason for the dissimilarities in the ‘burst’ response to removal of tetracaine and re-addition of ATP may be linked to the absence of increased spark activity during long periods of quiescence in low [ATP].

Figure 7. Changes in signal variance during quiescent periods associated with exposure to 0.1 mm ATP, 0.1 and 0.3 mm tetracaine.

A, the fluorescence record of a 10-pixel band (2.1 μm) showing the last Ca²⁺ wave in 0.1 ATP, the quiescent period and the first wave on return of ATP to 5 mm. The trace directly above the fluorescence record is the variance of the signal relative to the period immediately after the last Ca²⁺ wave (right-hand side scale). B, the comparable fluorescence record for the last wave recorded in 100 μm tetracaine and the first wave after removal of tetracaine. Relative variance is shown above the fluorescence record. C, sample linescan images early (a) and late (b) during the quiescent period in tetracaine. The sections of the variance record corresponding to these line scans are indicated in C. D, records of mean changes in the variance of the fluorescence signal during the quiescent period following exposure to 0.1 mm ATP, 0.1 and 0.3 mm tetracaine (n = 6 for each trace). The standard deviation of the variance for the first and last value of each record are shown as error bars at the beginning and end of each trace.

DISCUSSION

Cellular basis for spontaneous Ca²⁺ release; the role of Ca²⁺ sparks

At high cellular Ca²⁺ loads, spontaneous SR Ca²⁺ release occurs as a propagating Ca²⁺ wave throughout the myocyte (Capogrossi et al. 1987; Wier et al. 1987). This form of SR Ca²⁺ release does not require a Ca²⁺ current trigger; recent work using confocal microscopy has suggested a direct link between elementary Ca²⁺ release events (Ca²⁺ sparks) and propagated Ca²⁺ waves (Cheng et al. 1996; Lukyanenko et al. 1998). In Ca²⁺ overloaded cells Ca²⁺ sparks are larger in amplitude and higher in frequency compared with normal Ca²⁺ loads (Cheng et al. 1996; Lukyanenko et al. 1996). The key change in Ca²⁺ overloaded myocytes is an altered status of cardiac SR such that a localised increase of [Ca²⁺] (i.e. a Ca²⁺ spark) can recruit adjacent Ca²⁺ release events thereby generating a propagated Ca²⁺ wave (Cheng et al. 1996; Lipp & Niggli, 1998; Lukyanenko et al. 1998).

Relationship between Ca²⁺ propagation velocity and amplitude

The results illustrated in Fig. 1Bb show an approximately linear relationship between Ca²⁺ wave amplitude and propagation velocity under control conditions. This relationship has been previously reported for propagated waves in intact cardiac myocytes (Trafford et al. 1995). Although the precise mechanism underlying the relationship has not been verified, propagation of a Ca²⁺ wave is thought to be mediated via a ‘fire-diffuse-fire’ mechanism between discrete clusters of RyR (Stern, 1992; Keizer et al. 1998). Simplistically, larger Ca²⁺ release would be expected to propagate faster because the threshold [Ca²⁺] in adjacent sites would be attained earlier. Over a limited range of wave amplitudes, Keizer et al. 1998 calculated an approximately linear relationship between Ca²⁺ wave propagation velocity (v) and the coefficient σ/dc* where d is the distance between release sites, c* is the threshold for Ca²⁺ release and σ is the amount of Ca²⁺ released from the SR (Keizer et al. 1998). Normally, the parameters d and c* are considered as fixed, but recently, modelling of Ca²⁺ waves has been extended to include: (i) the influence of luminal [Ca²⁺] on c* (Györke & Györke, 1998); (ii) the asymmetry of the Ca²⁺ diffusion coefficient and the release-site density in the transverse and longitudinal directions (Izu et al. 2001; Subramanian et al. 2001); (iii) the stochastic nature of RyR2 activity (Izu et al. 2001); and (iv) the non-linear nature of Ca²⁺ buffering around the release site (Izu et al. 2001). In the current study, small but significant variations in Ca²⁺ wave velocity were observed under control conditions; correcting the Ca²⁺ wave velocity for the amplitude (i.e. the v/σ index) minimised these variations. This suggests that these variations arise from small changes in SR Ca²⁺ content between Ca²⁺ waves (rather than variations in d or c*).

Low [ATP] inhibits Ca²⁺ waves by inhibiting the RyR

The effects of reduced cytosolic [ATP] on spontaneous Ca²⁺ waves described here could arise through: (i) inhibition of Ca²⁺ uptake via the SR Ca²⁺-ATPase and/or (ii) inhibition of Ca²⁺ release via the RyR. The combination of lower frequency, increased amplitude of spontaneous Ca²⁺ release and increased SR Ca²⁺ content (i.e. caffeine-induced Ca²⁺ release) has been previously noted in studies using the drug tetracaine (Overend et al. 1997; Györke et al. 1997). This suggests that the dominant effect of lowering [ATP] is to reduce the activity of the RyR rather than inhibition of SR Ca²⁺ uptake. Recent studies on intact ventricular myocytes under metabolic inhibition have reported a reduced frequency of spontaneous Ca²⁺ release and an increased SR content (Overend et al. 2001). A decreased cytosolic [ATP] may be the intracellular mechanism although it should be remembered that there are many intracellular changes taking place during metabolic inhibition.

Low [ATP] also inhibits uptake of Ca²⁺ into the SR

In permeabilised cells the rate of fall of [Ca²⁺] at any one site within the cell is determined largely by uptake into the SR. The high spatial resolution of the linescan method used in this study allows the rate of fall of [Ca²⁺] to be assessed, unaffected by concomitant changes in Ca²⁺ wave propagation velocity. In this study, the rate of fall of [Ca²⁺] was reduced at 0.2 mm ATP but not at 0.5 mm, consistent with published values of the K_m for ATP of the SR Ca²⁺-ATPase of 0.18 mm (Shigekawa et al. 2000). Despite slowed Ca²⁺ uptake, the SR is able to accumulate more Ca²⁺ than normal and therefore release sufficient Ca²⁺ to overcome the lower RyR Ca²⁺ sensitivity at 0.2 mm ATP and allow propagation of the Ca²⁺ wave. The increased time between waves observed at 0.5 and 0.2 mm ATP reflects two factors. (i) Spontaneous Ca²⁺ release has to be larger to provide sufficient Ca²⁺ to trigger efflux at adjacent but less sensitive release sites. In support of this, decreases in frequency and increases in amplitude of Ca²⁺ waves were also observed in tetracaine. (ii) The lower rate of SR Ca²⁺ uptake observed in 0.2 mm ATP will increase the time required for sufficient Ca²⁺ to accumulate within the SR.

The effect of low ATP and tetracaine on wave propagation

Despite the increased amplitude of the Ca²⁺ waves in low [ATP] or tetracaine, the velocity of propagation was less than control values. A slower, but larger Ca²⁺ wave runs counter to the proportional relationship between amplitude and velocity seen under control conditions (i.e. constant v/σ). In the steady state, reduced ATP appears to reduce the value of v/σ. In 0.5 mm ATP, this value was ≈40 % of control values, while in 0.2 mm ATP, the value reached 27 %. In individual cells, values below 25 % of control were not detected. Based on the modelling of Keizer et al. 1998 and Izu et al. 2001, a decreased v/σ value would result from an increase in c*, i.e. an increase in the threshold [Ca²⁺] for Ca²⁺ release. Calculations by Izu et al. (2001) showed that slowing of wave velocity can also be achieved by altered kinetics of RyR2. Studies of the effects of ATP on isolated RyR indicate that reduced [ATP] decreased the probability of opening (P_o) and the mean open time but not the single channel conductance (Xu & Meissner, 1996; Kermode et al. 1998). The extent of the effects are disputed; in one study, P_o decreased from 0.8 in 5 mm ATP to 0.6 in zero ATP (Xu & Meissner, 1996), while in another the P_o decreased to 0.02 in zero ATP (Kermode et al. 1998). Furthermore, the Ca²⁺ sensitivity of RyR is decreased in low [ATP] (Xu & Meissner, 1996). Therefore lowered [ATP] would be expected to reduce RyR activity with the net effect of an increased threshold [Ca²⁺] for Ca²⁺ release, but it is not clear whether reduced P_o or reduced Ca²⁺ sensitivity is the main contributor to this effect. Interestingly, tetracaine (50 μm) decreased the v/σ to a similar extent to low [ATP], yet studies on isolated RyR suggest that at these concentrations tetracaine simply reduces P_o of RyR2 (Xu et al. 1993; Györke et al. 1997).

Abolition of Ca²⁺ waves

The absence of Ca²⁺ waves in 0.1 mm ATP or tetracaine was not due to loss of Ca²⁺ from the SR since the caffeine- induced Ca²⁺ release was in some cases twice the magnitude of the control values. This suggests either that Ca²⁺ waves were not initiated and/or that they were not propagated. Frequently the Ca²⁺ spark triggering the wave is out of the plane of focus of the wave (Cheng et al. 1996), so routine monitoring of the triggering event of the Ca²⁺ wave under different conditions was not possible. In some cells, however, Ca²⁺ sparks or limited SR Ca²⁺ release were observed in 0.1 mm ATP (see Fig. 3A and B) despite the absence of propagated Ca²⁺ waves. This suggests triggering events in the form of Ca²⁺ sparks still occur in 0.1 mm ATP although their frequency may be less than normal. Thus the absence of Ca²⁺ waves is due to the inability of the Ca²⁺ spark to propagate. Similarly, in tetracaine, spark activity failed to initiate a propagated Ca²⁺ wave, but unlike low [ATP], Ca²⁺ spark frequency increased during long periods of quiescence in a fashion similar to that reported by Györke et al. (1997). In models of Ca²⁺ wave propagation in cardiac cells there are critical conditions for Ca²⁺ wave propagation (Stern, 1992; Keizer et al. 1998). In particular Keizer et al. (1998) showed that there is a maximum value of c* d/σ above which waves cannot occur (c* is the threshold for Ca²⁺ release from the SR, d is the distance between release sites and σ is the amount of Ca²⁺ released from the SR). Therefore, one explanation for the inhibition of wave propagation is an increase in the threshold [Ca²⁺] for SR Ca²⁺ release (c*) to a greater extent than the concomitant increase in the amount of Ca²⁺ released (σ). In this study, using v/σ as an index of 1/c*, the results suggest that in 0.1 mm ATP the value for c* increases by ≈5, while the Ca²⁺ available for release has only doubled. However, care should be taken in translating peak [Ca²⁺] of a wave directly to SR Ca²⁺ release due to saturation of mobile and fixed Ca²⁺ buffers (Izu et al. 2001).

In a previous study (Yang & Steele, 2000), steady-state spontaneous Ca²⁺ release was recorded at [ATP] as low as 0.1 and 0.01 mm in the presence of 10 mm CrP. Only steady states were studied and not the transition phases to low [ATP] and on recovery. No CrP was used in the current study in order to simulate the conditions during anoxia/ischaemia in the heart. The concentration of ATP does not decrease within intact hearts until [CrP] has fallen to undetectable levels (Allen et al. 1985). Inclusion of CrP in the perfusion solution will have two effects: to buffer [ATP] and to minimise the rise in [ADP] within regions where the rate of ATP breakdown is high. Diffusion calculations predict that significant intracellular gradients of [ATP] would not develop at the concentrations used in this study (Fabiato & Fabiato, 1975). However, intracellular domains or sub-compartments may exist where ATP turnover is high leading to local gradients of [ATP] and [ADP] (Jones, 1986). Therefore, inclusion of CrP may influence regional levels of ATP and ADP and may explain the different results observed in the two studies.

Removal of RyR inhibition and wave activity

On removal of inhibition of Ca²⁺ release, whether due to low [ATP] or tetracaine, there follows a burst of high frequency waves. This has been reported previously in intact cells and probably represents the loss of the extra Ca²⁺ gained while release was inhibited; once inhibition of release is removed the SR is unable to maintain such a high Ca²⁺ content. However, differences in the details of this response exist between low [ATP] and tetracaine. These differences are most marked after long periods of inhibition (Fig. 6). After a prolonged period in tetracaine, the burst of activity is so intense that individual waves are not easily resolved, as Ca²⁺ release takes place almost simultaneously at several locations throughout the cell. After a similar period in low ATP the waves return as discrete events (Fig. 6). This may simply reflect the relatively slow entry of ATP back into the preparation due to its lower diffusion constant, such that waves return while inhibition of release is still relatively high, thus damping the burst of activity. Another point worthy of consideration is the increased spark activity (Fig. 7) after prolonged exposure to tetracaine. This may reflect a greater lumenal influence on the open probability of the RyR, which is absent in the case of low ATP. Therefore, on removal of tetracaine the additional lumenal influence over RyR activity could explain the highly disordered Ca²⁺ release since very high rates of spontaneous sparks would prevent well-organised waves emerging (Izu et al. 2001). The lack of lumenal influence of Ca²⁺ on RyR activity in the absence of ATP has been previously reported in isolated RyR in lipid bilayer (Lukyanenko et al. 1996).

In summary, reduction of ATP to 0.1 mm inhibited spontaneous Ca²⁺ waves in permeabilised cardiac myocytes due to decreased sensitivity of RyR to cytosolic Ca²⁺. When a similar situation was induced by tetracaine, limited, non-propagated Ca²⁺ release occurred after approximately 60 s of quiescence. This effect, which is thought to be due to increased lumenal [Ca²⁺] causing a secondary increase in RyR activity, is absent in 0.1 mm ATP. The large and frequent Ca²⁺ waves observed on restoration of [ATP] were not observed in the period immediately after prolonged tetracaine exposure; this difference is linked to an increased RyR activity prior to removal of tetracaine.