Intracellular Ca²⁺ release contributes to automaticity in cat atrial pacemaker cells

Abstract

1. The cellular mechanisms governing cardiac atrial pacemaker activity are not clear. In the present study we used perforated patch voltage clamp and confocal fluorescence microscopy to study the contribution of intracellular Ca²⁺ release to automaticity of pacemaker cells isolated from cat right atrium.

2. In spontaneously beating pacemaker cells, an increase in subsarcolemmal intracellular Ca²⁺ concentration occurred concomitantly with the last third of diastolic depolarization due to local release of Ca²⁺ from the sarcoplasmic reticulum (SR), i.e. Ca²⁺ sparks. Nickel (Ni²⁺; 25–50 μM), a blocker of low voltage-activated T-type Ca²⁺ current (I_Ca,T), decreased diastolic depolarization, prolonged pacemaker cycle length and suppressed diastolic Ca²⁺ release.

3. Voltage clamp analysis indicated that the diastolic Ca²⁺ release was voltage dependent and triggered at about -60 mV. Ni²⁺ suppressed low voltage-activated Ca²⁺ release. Moreover, low voltage-activated Ca²⁺ release was paralleled by a slow inward current presumably due to stimulation of Na⁺–Ca²⁺ exchange (I_Na-Ca). Low voltage-activated Ca²⁺ release was found in both sino-atrial node and latent atrial pacemaker cells but not in working atrial myocytes.

4. These findings suggest that low voltage-activated I_Ca,T triggers subsarcolemmal Ca²⁺ sparks, which in turn stimulate I_Na-Ca to depolarize the pacemaker potential to threshold. This novel mechanism indicates a pivotal role for I_Ca,T and subsarcolemmal intracellular Ca²⁺ release in normal atrial pacemaker activity and may contribute to the development of ectopic atrial arrhythmias.

Electrical excitation of the mammalian heart originates from specialized pacemaker cells located in specific regions of the right atrium. The right atrium contains both primary (sino-atrial (SA) node) and secondary (latent) pacemaker cells. Latent atrial pacemakers are specialized cells localized in specific regions of the inferior right atrium (Rubenstein et al. 1987). SA node pacemaker cells maintain the normal heartbeat whereas latent atrial pacemakers contribute to various types of atrial arrhythmias. In general, the electrical activity of cardiac pacemakers is thought to depend exclusively on ion channel currents within the plasma membrane, including: (i) the hyperpolarization-activated inward current, (ii) time-dependent decay of K⁺ conductance, (iii) inward T-type and L-type Ca²⁺ currents, (iv) inward leak currents, and (v) the lack of background K⁺ conductance (Noma et al. 1984; Hagiwara et al. 1988; DiFrancesco, 1991; Wu et al. 1991; Campbell et al. 1992; Zhou & Lipsius, 1992, 1993, 1994). In particular, low voltage-activated T-type Ca²⁺ current (I_Ca,T) is thought to contribute to the last third or late phase of diastolic depolarization in both SA node and latent atrial pacemakers (Hagiwara et al. 1988; Zhou & Lipsius, 1994). However, inhibition of intracellular Ca²⁺ release from the sarcoplasmic reticulum (SR) by ryanodine depresses diastolic depolarization in both SA node (Lipsius, 1989; Rigg & Terrar, 1996; Li et al. 1997) and latent atrial (Rubenstein & Lipsius, 1989; Zhou & Lipsius, 1993) pacemakers, indicating that intracellular Ca²⁺ release also plays a prominent role in bringing the late pacemaker potential to threshold. Diastolic Ca²⁺ release is thought to depolarize the pacemaker potential by stimulating inward Na⁺-Ca²⁺ exchange current (I_Na-Ca) (Zhou & Lipsius, 1993). However, a clear demonstration of diastolic Ca²⁺ release in pacemaker cells has not yet been provided and the mechanisms responsible for initiating diastolic Ca²⁺ release in these cells are not understood. In the present study, we report that atrial pacemaker cells exhibit a local release of subsarcolemmal intracellular Ca²⁺, i.e. Ca²⁺ sparks (Cheng et al. 1993), during late diastole, which is triggered by low voltage-activated I_Ca,T. This novel pacemaker mechanism contributes to normal pacemaker function and may underlie abnormal atrial pacemaker activities promoted by alterations in intracellular Ca²⁺ release.

METHODS

Adult cats of either sex were anaesthetized with sodium pentobarbital (70 mg kg⁻¹i.p.). After bilateral thoracotomy, the heart was rapidly excised and placed on a Langendorff perfusion apparatus for cell isolation, as described previously (Wang & Lipsius, 1996). Pacemaker cells are distinguished from atrial muscle cells by their morphology (smaller size, lack of striations, tapered ends), pacemaker activity and the presence of hyperpolarization-activated inward current (Wu et al. 1991; Zhou & Lipsius, 1992). Latent pacemaker cells were isolated from a specific region of the inferior right atrium at its junction with the inferior vena cava, and SA node pacemaker cells were isolated from the superior right atrium at its junction with the superior vena cava (Rubenstein et al. 1987). Isolated cells were plated on glass coverslips and transferred onto the stage of an inverted microscope (Axiovert 100, Carl Zeiss) attached to a confocal scanner unit (LSM 410, Carl Zeiss). The Ca²⁺ indicator fluo-3 (Molecular Probes) was introduced into the myoplasm by exposure of the cells to the membrane-permeant acetoxymethyl ester form of the dye (fluo-3 AM, 5 μM for 30 min at room temperature). During experiments, cells were continuously superfused with Tyrode solution containing (mM): NaCl, 137; KCl, 5.4; MgCl₂, 1.0; CaCl₂, 2.0; Hepes, 5; glucose, 11; titrated with NaOH to a pH of 7.4. All experiments were performed at 33 ± 2°C. Membrane voltage and currents were recorded with an Axopatch 200A amplifier (Axon Instruments) using a nystatin (150 μg ml⁻¹)-perforated patch (Horn & Marty, 1988) whole-cell recording method (Hamill et al. 1981). The internal pipette solution contained (mM): potassium glutamate, 100; KCl, 40; MgCl₂, 1.0; Na₂-ATP, 4; EGTA, 0.5; Hepes, 5; titrated with KOH to a pH of 7.2. Fluo-3 fluorescence was excited using an argon-ion laser at 488 nm and simultaneously recorded at wavelengths ≥ 515 nm. Confocal linescan images were acquired at a rate of 250 Hz. [Ca²⁺]_i images were calculated using the formula:

(Cannell et al. 1994), where R is the normalized fluorescence (F/F_rest) and K_d= 1.1 μM. A trigger signal generated by the voltage clamp protocol was recorded simultaneously by the confocal imaging system to allow synchronization of electrophysiological and Ca²⁺ measurements. For further details of methods see Hüser et al. (1996) and Blatter et al. (1997).

The animal procedures used in this study were in accordance with the guidelines of the Animal Care and Use Committee of Loyola University Medical Center.

RESULTS

Figure 1 illustrates pacemaker action potentials (A) and spatially resolved changes in intracellular Ca²⁺ concentration (B) recorded from a spontaneously beating latent atrial pacemaker cell. The subsarcolemmal Ca²⁺ concentration ([Ca²⁺]_ss) is known to regulate plasma membrane conductances, including I_Na-Ca. Therefore, changes in [Ca²⁺]_ss were visualized by positioning the repetitively scanned line (250 Hz) used to obtain the linescan image parallel to the longitudinal axis of the cell and close to the sarcolemmal membrane (Fig. 1D, inset). Each pacemaker action potential coincided with a rapid rise in [Ca²⁺]_ss due to Ca²⁺ release from the SR triggered by Ca²⁺ entry through voltage-gated Ca²⁺ channels. Ca²⁺ release occurred almost simultaneously throughout the subsarcolemmal space. No change in [Ca²⁺]_ss was observed immediately following repolarization. However, [Ca²⁺]_ss gradually increased during the late phase of the pacemaker depolarization prior to the action potential, resulting in a small diastolic ‘Ca²⁺ pedestal’ in recordings averaged over the subsarcolemmal space (Fig. 1B). Recordings of [Ca²⁺]_ss from a restricted region (∼1 μm) close to an active release site (Fig. 1C) revealed that the Ca²⁺ pedestal was caused by the summation of individual local Ca²⁺ release events, i.e. Ca²⁺ sparks. The surface plot (Fig. 1D) demonstrates the absence of Ca²⁺ sparks during early diastolic depolarization immediately following the prior pacemaker action potential, and a gradual increase in Ca²⁺ spark frequency during the late pacemaker depolarization. In this cell, the diastolic Ca²⁺ pedestal became detectable with the occurrence of the first subsarcolemmal Ca²⁺ spark at a membrane potential of -57 mV. Similar results were obtained in seven pacemaker cells.

Figure 1. Ca²⁺ sparks precede the action potential in spontaneously beating latent atrial pacemaker cells.

Simultaneous recordings of membrane voltage (V_m; A) and subsarcolemmal Ca²⁺ concentration ([Ca²⁺]_ss; B). The lower traces in each panel show amplified recordings of diastolic voltage (A) and [Ca²⁺]_ss (B). The amplified traces in B show an increase in [Ca²⁺]_ss during the late pacemaker depolarization preceding the action potential (‘Ca²⁺ pedestal’). [Ca²⁺]_ss started to increase at threshold potentials of about -58 mV. C, local [Ca²⁺] transients recorded by averaging over a short distance (1 μm) of the scanned line intersecting with a site displaying localized Ca²⁺ release (arrow in D). The arrows mark individual Ca²⁺ sparks. D, surface plot representation of the linescan image illustrating changes in [Ca²⁺]_ss during the diastolic depolarization preceding the first action potential (indicated by the horizontal bar in C). The inset in D indicates the positioning of the line.

To determine whether the mechanism responsible for triggering diastolic Ca²⁺ release depended on membrane depolarization or whether it was a voltage-independent mechanism inherent to the SR, we studied Ca²⁺ release in pacemaker cells under voltage clamp conditions. The pacemaker action potential was approximated by a ramp-step voltage protocol (Fig. 2A–C, left panels) which followed nine pre-conditioning pulses to ensure steady-state loading of SR Ca²⁺. The [Ca²⁺]_ss transient elicited by L-type Ca²⁺ current (I_Ca,L) was preceded by a smaller Ca²⁺ release triggered during the ramp depolarization at a potential of -57 mV. This low voltage-activated Ca²⁺ release (LVCR) seen under voltage clamp conditions mimicked the diastolic Ca²⁺ pedestal recorded from spontaneously beating pacemaker cells, and successfully reproduced the pattern of changes in [Ca²⁺]_ss. Like the diastolic Ca²⁺ pedestal, LVCR resulted from the summation of individual subsarcolemmal Ca²⁺ sparks. As shown in Fig. 2A–C (right panels), when the ramp depolarization was omitted and membrane voltage was held constant (-70 mV) prior to the voltage step, diastolic Ca²⁺ release was not observed. These findings clearly rule out the possibility that a spontaneous time-dependent mechanism is responsible for the increase in frequency of Ca²⁺ sparks during late diastole. Rather, Ca²⁺ release during the ramp and therefore during the late phase of diastolic depolarization is dependent on membrane depolarization. The voltage threshold of -57 mV excludes activation of L-type Ca²⁺ channels and subsequent Ca²⁺-induced Ca²⁺ release (CICR) as the underlying mechanism. It is compatible, however, with CICR induced by Ca²⁺ entry through low voltage-activated T-type Ca²⁺ channels (Zhou & January, 1998; Sipido et al. 1998). Similar results were obtained in four pacemaker cells.

Figure 2. Pacemaker Ca²⁺ sparks are triggered by a voltage-dependent mechanism.

Changes in [Ca²⁺]_ss (line plot presentation, A; linescan image, C) and membrane currents (B) in a latent atrial pacemaker cell in response to the voltage clamp protocol shown below the current trace. A depolarizing voltage clamp ramp from -70 to -45 mV (400 ms) prior to the voltage step to +10 mV (150 ms) elicited LVCR which reproduced the diastolic Ca²⁺ pedestal seen in spontaneous pacemaker cells (left panels). In the same cell, when the voltage ramp was omitted (right panels), no sparks were detected prior to the [Ca²⁺]_ss transient elicited by the step depolarization. Prolongation of the interval between beats also did not result in detection of spontaneous Ca²⁺ sparks (not shown).

The results shown in Fig. 3 support the idea that I_Ca,T triggers diastolic Ca²⁺ release. Micromolar concentrations of Ni²⁺ (25-50 μM), a blocker of cardiac T-type Ca²⁺ channels (Hagiwara et al. 1988; Zhou & Lipsius, 1994; Lee et al. 1999), decreased the slope of diastolic depolarization (Fig. 3A), prolonged pacemaker cycle length by 230 % compared with control and suppressed the diastolic Ca²⁺ pedestal which preceded the [Ca²⁺]_ss transients triggered by the pacemaker action potentials (Fig. 3B). Under voltage clamp conditions, Ni²⁺ also inhibited LVCR in the voltage range between -60 and -45 mV (Fig. 3C) but was without effect on Ca²⁺ release triggered by I_Ca,L at potentials positive to -45 mV (not shown). At the same time, no significant effect of Ni²⁺ was observed on peak I_Ca,L or [Ca²⁺]_ss transients triggered by step depolarizations or action potentials. Finally, Fig. 3D illustrates that in a latent atrial pacemaker cell the diastolic increase in [Ca²⁺]_i initiated at -58 mV was paralleled by a slow inward current in response to voltage ramps. The onset of the inward current dip coincided with the rise in [Ca²⁺]_ss. The slow ramp depolarization and the relatively large amplitude of this inward current makes it unlikely that it is due to activation of I_Ca,T. A similar Ca²⁺ release-dependent current that develops during the late phase of the pacemaker potential has been previously identified in these cells as I_Na-Ca (Zhou & Lipsius, 1993). Similar results were obtained in six pacemaker cells.

Figure 3. Nickel inhibits diastolic Ca²⁺ release in spontaneously active and voltage clamped pacemaker cells.

A, Ni²⁺ (50 μM) reversibly slowed the rate of a spontaneously beating latent atrial pacemaker cell by selectively decreasing the slope of the late phase of diastolic depolarization. B, Ni²⁺ (50 μM) inhibited the diastolic Ca²⁺ pedestal (ii, and surface plots in iii and iv) without a significant reduction of the [Ca²⁺]_ss transient triggered by the action potential (i). Spontaneous [Ca²⁺]_ss transients were aligned to the point of maximum rate of rise. ctrl, control. C, in another voltage clamped pacemaker cell, Ni²⁺ (25 μM) inhibited LVCR initiated at -59 mV. Peak I_Ca,L elicited by the step depolarization was unaffected by Ni²⁺. D, [Ca²⁺]_ss and membrane current in response to a ramp-step depolarization in a third latent atrial pacemaker cell. During the ramp depolarization the current deviated from the expected linear change in background current (dashed line). The onset of inward current coincided with the rise in [Ca²⁺]_ss indicative of I_Na-Ca stimulation.

Further experiments indicated that LVCR is specific for pacemaker cells. As shown in Fig. 4A, non-pacemaker atrial myocytes completely lacked LVCR when stimulated with a ramp-step protocol. When atrial cells were stimulated with a continuous voltage ramp from -70 to 0 mV (Fig. 4B), Ca²⁺ release was invariably initiated at potentials positive to -40 mV (n = 20), consistent with Ca²⁺ release triggered by I_Ca,L. With the continuous ramp protocol (Fig. 4B), Ca²⁺ release triggered by I_Ca,L was associated with a slow transient inward current characteristic of I_Na-Ca. In addition, the inhibitory effect of ryanodine on the electrical activity of SA node pacemaker cells (Lipsius, 1989; Rigg & Terrar, 1996; Li et al. 1997) suggests that LVCR also operates in SA node pacemakers. In Fig. 4C and D, single SA node pacemaker cells isolated from the same cat hearts were voltage clamped using the ramp-step (C) or continuous ramp (D) protocols. LVCR was detected in both cases at threshold voltages of -55 mV (C) and -52 mV (D). Using the continuous ramp protocol (D), the Ca²⁺ release appeared biphasic, compatible with Ca²⁺ release triggered by low voltage-activated I_Ca,T (-52 mV) and high voltage-activated I_Ca,L (-36 mV). The larger Ca²⁺ release triggered at more positive voltages by I_Ca,L was associated with a transient stimulation of inward I_Na-Ca (Fig. 4D). It therefore appears that LVCR is specific for pacemaker cells and is not found in regular atrial muscle cells. Similar results were obtained in three SA node pacemaker cells.

Figure 4. LVCR is present in primary pacemakers from the SA node, but absent in non-pacemaker atrial muscle cells.

Response of membrane current (i) and [Ca²⁺]_ss (ii) to a ramp-step depolarization (left panels) or a continuous ramp (right panels) in a non-pacemaker atrial muscle cell (A and B) and a SA node pacemaker cell (C and D). In the regular atrial muscle cell, Ca²⁺ release was triggered at -36 mV. In the SA node pacemaker cell, two components of Ca²⁺ release were elicited at -52 and -36 mV.

DISCUSSION

In both SA node (Hagiwara et al. 1988) and latent atrial pacemakers (Zhou & Lipsius, 1994) activation of I_Ca,T contributes to the last third or late phase of diastolic depolarization. Likewise, the contribution of SR Ca²⁺ release to both SA node (Lipsius, 1989; Rigg & Terrar, 1996; Li et al. 1997) and latent atrial (Rubenstein & Lipsius, 1989; Zhou & Lipsius, 1993) pacemaker function has been implied from the observation that ryanodine, an alkaloid that blocks SR Ca²⁺ release (Rousseau et al. 1987), inhibits the late phase of diastolic depolarization. The present report directly demonstrates, in both SA node and latent atrial pacemaker cells, that highly localized subsarcolemmal Ca²⁺ release from the SR, i.e. Ca²⁺ sparks, occurs concurrently with the late phase of diastolic depolarization. Moreover, several findings indicate that diastolic Ca²⁺ spark activity was triggered by I_Ca,T: (1) diastolic Ca²⁺ sparks were elicited at low voltages compatible with I_Ca,T and clearly distinguished from I_Ca,L-triggered Ca²⁺ release (positive to -40 mV); (2) diastolic Ca²⁺ release, as well as diastolic depolarization, were inhibited by micromolar concentrations of Ni²⁺, a blocker of cardiac T-type Ca²⁺ channels (Hagiwara et al. 1988; Zhou & Lipsius, 1994; Lee et al. 1999). Moreover, the fact that atrial myocytes lack LVCR and only exhibit Ca²⁺ release triggered by I_Ca,L emphasizes the unique role of LVCR in pacemaker cells. The absence of this mechanism in working atrial myocytes can be explained by the fact that atrial pacemaker cells possess a 5-fold higher density of I_Ca,T than non-pacemaker atrial cells (Zhou & Lipsius, 1994). This higher I_Ca,T density in pacemaker cells increases the probability that T-type Ca²⁺ channels will trigger Ca²⁺ release from subsarcolemmal Ca²⁺ stores. Although the mechanism is present in both SA node and latent atrial pacemaker cells, its relative contribution to pacemaker function may be more prominent in latent than SA node pacemaker function. This is based on our previous findings which indicate that inhibition of SR Ca²⁺ release by ryanodine slows pacemaker rate significantly more in latent atrial than SA node pacemaker cells (Lipsius, 1989), and that latent atrial pacemaker cells exhibit a more prominent presence and size of subsarcolemmal SR cisternae compared with SA node pacemaker cells (Rubenstein et al. 1987). The present results also suggest that LVCR stimulates inward I_Na-Ca, which would depolarize the pacemaker potential to threshold (Fig. 3D). I_Na-Ca has been reported to contribute significantly to automaticity in latent atrial pacemakers (Zhou & Lipsius, 1993) as well as SA node pacemakers (Campbell et al. 1992). The depolarizing effect of I_Na-Ca is expected to elicit further voltage-dependent activation of I_Ca,T-Ca²⁺ spark activity, resulting in a positive feedback mechanism contributing to the late phase of diastolic depolarization.

Single T-type channel openings may be sufficient to trigger Ca²⁺ sparks (Stern, 1992; López-López et al. 1995). Local elevation of [Ca²⁺]_ss will stimulate Na⁺-Ca²⁺ exchange carrier molecules in the vicinity of the SR Ca²⁺ release sites, providing additional inward current for as long as [Ca²⁺]_ss stays elevated. A Ca²⁺ spark (∼20-50 ms at 33°C) outlasts a single T-type Ca²⁺ channel opening (mean open time, 2.3 ms; Balke et al. 1992). Therefore, stimulation of I_Na-Ca both amplifies the unitary T-type current event and prolongs in time inward current flow, thereby allowing efficient temporal summation of I_Ca,T-Ca²⁺ spark events. The fact that inhibition of SR Ca²⁺ release by ryanodine markedly prolongs pacemaker cycle length in latent atrial (Zhou & Lipsius, 1993) and SA node (Lipsius, 1989; Rigg & Terrar, 1996; Li et al. 1997) pacemaker cells in the presence of intact, functional T-type Ca²⁺ channels, emphasizes the importance of this mechanism in normal pacemaker function. In ryanodine-treated cells, I_Ca,T alone cannot sustain the steep slope of the pacemaker depolarization. The currents associated with a single ‘pacemaker spark’ are below the resolution of whole-cell recordings. We can, however, speculate that with a membrane resistance of ∼2 GΩ (Wu et al. 1991) and estimates of Ca²⁺ fluxes underlying Ca²⁺ sparks of ∼3 pA (Blatter et al. 1997) only a few local events, well within the range of spark frequencies observed in spontaneously beating cells, are necessary to depolarize the late diastolic slope to threshold (ΔV, ∼5-10 mV).

The present findings provide another example of subcellular Ca²⁺ release events that control membrane excitability. For example, in smooth muscle cells Ca²⁺ sparks can activate Ca²⁺-dependent K⁺ channels to cause hyperpolarization (Nelson et al. 1995) or Ca²⁺-activated Cl⁻ channels to cause depolarization (ZhuGe et al. 1998). Moreover, interstitial cells of Cajal, the gastrointestinal pacemakers, exhibit rhythmic Ca²⁺ release from IP₃-dependent Ca²⁺ stores activating a Ca²⁺-dependent cation current that drives the pacemaker depolarization (Thomsen et al. 1998). Ca²⁺ release in these cells, however, appears to be driven by an oscillator within the endoplasmic reticulum/Ca²⁺ store. In contrast, Ca²⁺ release and I_Na-Ca in cardiac pacemaker cells serve as a feedback amplifier tightly coupled to the plasma membrane pacemaker mechanism via activation of I_Ca,T. Because intracellular Ca²⁺ release is sensitive to a variety of second messenger signalling mechanisms as well as variations in SR Ca²⁺ load, this mechanism is a potential site for regulation of heart rate by autonomic neurotransmitters, hormones or drugs. Moreover, because latent atrial pacemakers appear to rely more on SR Ca²⁺ release for their pacemaker mechanism than SA node pacemakers, drugs or disease processes that enhance intracellular Ca²⁺ content and/or release may be expected to enhance latent atrial pacemaker automaticity. These findings, therefore, may provide insight into the etiologies and potential treatments of atrial arrhythmias resulting from enhanced forms of automaticity.