β-Adrenergic stimulation increases the intra-sarcoplasmic reticulum Ca²⁺ threshold for Ca²⁺ wave generation

Abstract

β-Adrenergic signalling induces positive inotropic effects on the heart that associate with pro-arrhythmic spontaneous Ca²⁺ waves. A threshold level of sarcoplasmic reticulum (SR) Ca²⁺ ([Ca²⁺]_SR) is necessary to trigger Ca²⁺ waves, and whether the increased incidence of Ca²⁺ waves during β-adrenergic stimulation is due to an alteration in this threshold remains controversial. Using the low-affinity Ca²⁺ indicator fluo-5N entrapped within the SR of rabbit ventricular myocytes, we addressed this controversy by directly monitoring [Ca²⁺]_SR and Ca²⁺ waves during β-adrenergic stimulation. Electrical pacing in elevated extracellular Ca²⁺ ([Ca²⁺]_o= 7 mm) was used to increase [Ca²⁺]_SR to the threshold where Ca²⁺ waves were consistently observed. The β-adrenergic agonist isoproterenol (ISO; 1 μm) increased [Ca²⁺]_SR well above the control threshold and consistently triggered Ca²⁺ waves. However, when [Ca²⁺]_SR was subsequently lowered in the presence of ISO (by lowering [Ca²⁺]_o to 1 mm and partially inhibiting sarcoplasmic/endoplasmic reticulum calcium ATPase with cyclopiazonic acid or thapsigargin), Ca²⁺ waves ceased to occur at a [Ca²⁺]_SR that was higher than the control threshold. Furthermore, for a set [Ca²⁺]_SR level the refractoriness of wave occurrence (Ca²⁺ wave latency) was prolonged during β-adrenergic stimulation, and was highly dependent on the extent that [Ca]_SR exceeded the wave threshold. These data show that acute β-adrenergic stimulation increases the [Ca²⁺]_SR threshold for Ca²⁺ waves, and therefore the primary cause of Ca²⁺ waves is the robust increase in [Ca²⁺]_SR above this higher threshold level. Elevation of the [Ca²⁺]_SR wave threshold and prolongation of wave latency represent potentially protective mechanisms against pro-arrhythmogenic Ca²⁺ release during β-adrenergic stimulation.

Key points

• In the heart, Ca²⁺ waves are arrhythmogenic spontaneous sarcoplasmic reticulum (SR) Ca²⁺ release events that arise when the Ca²⁺ content in the SR reaches a critical threshold level.

• β-Adrenergic signalling induces Ca²⁺ waves in cardiac myocytes, but it remains unclear if this is due to a decrease in the Ca²⁺ wave threshold or more simply due to an increase in SR Ca²⁺ content.

• We used direct, dynamic measurement of SR Ca²⁺ levels to show that the Ca²⁺ wave threshold is unexpectedly increased during β-adrenergic stimulation.

• Our data show that the primary cause of Ca²⁺ waves following acute β-adrenergic stimulation is the increase in SR Ca²⁺ content and not a decrease in the Ca²⁺ wave threshold.

• We propose that the elevation of the Ca²⁺ wave threshold represents a protective mechanism against arrhythmogenic events during periods of β-adrenergic stimulation.

Introduction

Cardiac contraction is triggered via Ca²⁺-induced Ca²⁺ release (CICR), where influx of Ca²⁺ into the cardiomyocyte through L-type Ca²⁺ channels (LTCCs) opens ryanodine receptor (RyR) Ca²⁺ release channels in the junctional sarcoplasmic reticulum (SR) membrane and initiates coordinated release of Ca²⁺ from the SR Ca²⁺ store. SR Ca²⁺ release can also occur spontaneously during diastole. Diastolic Ca²⁺ release is detrimental to cardiac performance as it impairs relaxation, reduces the concentration of Ca²⁺ within the SR ([Ca²⁺]_SR) and predisposes the heart to arrhythmia. One highly arrhythmogenic form of diastolic Ca²⁺ release is the Ca²⁺ wave, where Ca²⁺ that is spontaneously released from an SR Ca²⁺ release unit (CRU, formed by a cluster of RyRs) diffuses to neighbouring CRUs and triggers regenerative and propagating CICR independently of the cardiac action potential (AP) (Stern et al. 1988). Extrusion of Ca²⁺ that is released during a Ca²⁺ wave via the electrogenic Na⁺–Ca²⁺ exchange mechanism (Kass et al. 1978; Fedida et al. 1987) has a strong depolarizing effect on the diastolic membrane potential and underlies arrhythmogenic delayed afterdepolarizations (Capogrossi et al. 1987; Schlotthauer & Bers, 2000; Fujiwara et al. 2008). It is therefore of paramount importance to understand the mechanisms for Ca²⁺ wave generation in the heart.

A considerable amount of evidence has emerged that a distinct ‘overload’[Ca²⁺]_SR is necessary to trigger Ca²⁺ waves, and this level has been termed the Ca²⁺ wave threshold (Díaz et al. 1997b; Venetucci et al. 2007) or store overload-induced Ca²⁺ release threshold (Jiang et al. 2004). The Ca²⁺ wave threshold is altered by agents that modulate RyR activity, with channel activators (e.g. caffeine) and inhibitors (e.g. tetracaine) decreasing (Venetucci et al. 2007; Kong et al. 2008) and increasing (Overend et al. 1997) the threshold, respectively. In animal models of heart failure where RyR channel activity is increased, the Ca²⁺ wave threshold is decreased (Belevych et al. 2012; Maxwell et al. 2012) and this may play a role in the arrhythmogenesis associated with these models. Furthermore, investigations of RyR mutations associated with catecholaminergic polymorphic ventricular tachycardia (CPVT) have shown that gain-of-function mutations (reviewed by Priori & Chen, 2011) result in a decrease in the Ca²⁺ wave threshold and increased incidence of arrhythmogenic Ca²⁺ release following increases in [Ca²⁺]_SR[e.g. in response to cardiac glycosides (Sedej et al. 2010) or β-adrenergic stimulation (Kashimura et al. 2010)]. Interestingly, although the Ca²⁺ wave threshold is lower in cardiomyocytes from CPVT RyR R4496C mutant mice than in wild-type controls, β-adrenergic stimulation increases the Ca²⁺ wave threshold in R4496C myocytes (Kashimura et al. 2010). Thus, the likely factor triggering Ca²⁺ waves in response to catecholamines in this CPVT model is the sudden increase in [Ca²⁺]_SR and not an acute lowering of the Ca²⁺ wave threshold. A similar increase in the Ca²⁺ wave threshold in response to β-adrenergic signalling was observed in wild-type mouse cardiomyocytes (Kashimura et al. 2010; Stokke et al. 2010a), suggesting that this effect is not unique to disease models. Contrasting with these data are investigations that show that β-adrenergic stimulation increases diastolic SR Ca²⁺ release in the form of Ca²⁺ sparks (Zhou et al. 2009; Ogrodnik & Niggli, 2010), SR Ca²⁺ leak (Curran et al. 2007; Bovo et al. 2012; Ullrich et al. 2012) and Ca²⁺ waves (Curran et al. 2010) at a similar or lower level of SR Ca²⁺ content (which intuitively should translate into a decrease in the Ca²⁺ wave threshold). Additionally, selective activation of signalling molecules acting downstream of β-adrenergic stimulation (Guo et al. 2006; Terentyev et al. 2008) or phosphomimetic mutation of target phosphorylation sites on the RyR (Shan et al. 2010; van Oort et al. 2010) have all been shown to augment SR Ca²⁺ release and predispose the heart to arrhythmia. Thus, it remains unclear and highly controversial whether Ca²⁺ waves during β-adrenergic stimulation are due to direct alterations in the Ca²⁺ wave threshold or more simply due to the increase in [Ca²⁺]_SR that occurs concomitant with activation of the signalling pathway.

Measurement of the Ca²⁺ wave threshold in cardiomyocytes during β-adrenergic stimulation is experimentally challenging due to the enhanced activity of multiple Ca²⁺ handling proteins, which hinders targeted and controlled examination of the Ca²⁺ wave threshold. To gain new insight into the mechanisms underlying Ca²⁺ waves during β-adrenergic stimulation we utilized direct, dynamic fluorescent measurement of [Ca²⁺]_SR and the Ca²⁺ wave threshold in acutely isolated rabbit ventricular myocytes. We found that β-adrenergic stimulation increases the Ca²⁺ wave threshold, which may serve to protect the heart from arrhythmogenic Ca²⁺ release as [Ca²⁺]_SR increases during sympathetic nervous system activation.

Part of this work has been presented previously in abstract form (Domeier & Blatter, 2010; Domeier et al. 2011).

Methods

Solutions and chemicals

All reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise noted. Normal Tyrode solution contained (in mm): 135 NaCl, 4 KCl, 2 CaCl₂, 1 MgCl₂, 10 d-glucose and 10 Hepes (pH 7.4 with NaOH). Tyrode solutions with altered extracellular free calcium concentration ([Ca²⁺]_o) were prepared using iso-osmotic substitution of CaCl₂ and NaCl. Isoproterenol (ISO) was prepared daily as a stock solution of 100 mm (in H₂O, maintained at 4°C), diluted in Tyrode solution immediately prior to experimental procedures and used within 30 min. Cyclopiazonic acid (CPA) and thapsigargin (TG) were dissolved in DMSO and diluted to working concentrations in Tyrode solution. All experiments were performed with appropriate vehicle controls. All experiments were conducted at room temperature (22–24°C).

Cardiac myocyte isolation

Left ventricular myocytes were isolated from male New Zealand White rabbits (30 animals, 2.5 kg, Myrtle's Rabbitry, Thompsons Station, TN, USA). Rabbits were anaesthetized with sodium pentobarbital (50 mg kg⁻¹) and hearts were rapidly excised and retrogradely perfused with a nominally Ca²⁺-free Tyrode solution for 10 min, followed by a minimal essential Eagle's medium (MEM) solution with 20 μm Ca²⁺ and 45 μg ml⁻¹ Liberase Blendzyme TH (Roche Applied Science, Indianapolis, IN, USA) for 20 min all at 37°C. The left ventricular free wall tissue was minced, filtered and washed in MEM solution containing 50 μm Ca²⁺ and 10 mg ml⁻¹ BSA. Isolated myocytes were maintained in MEM solution with 50 μm Ca²⁺ at room temperature until Ca²⁺ indicator dye loading procedures. All protocols were approved by the Institutional Animal Care and Use Committee at Rush University, and comply with US and UK regulations on animal experimentation (Drummond, 2009).

Intra-SR Ca²⁺ measurements

Ventricular myocytes were incubated with 10 μm of membrane-permeant fluo-5N/AM (Molecular Probes-Life Technologies, Grand Island, NY, USA) for 2.5 h at 37°C to promote accumulation of dye within the SR (see Fig. 2A). Following a 30 min wash at 37°C, myocytes were equilibrated to room temperature and plated on laminin-coated coverslips for imaging experiments. Laser scanning fluorescence confocal microscopy was performed using the resonant scan head of a Nikon A1R system (Nikon Instruments Inc., Melville, NY, USA) in frame scan mode (60 frames s⁻¹ at 256 × 512 pixels, 400 nm per pixel). Fluo-5N was excited using the 488 nm line of the argon ion laser and emission was recorded at 500–530 nm. Low laser excitation intensities and limited recording times were used to minimize photobleaching of the fluo-5N indicator. Although generally negligible, in some experiments with extended recording times (e.g. Fig. 3A) photobleaching did occur and these recordings were corrected using control bleach rates (single exponential fits) obtained from each cell under steady-state conditions. Changes in [Ca²⁺]_SR are presented as (F–F_Min)/(F_Max–F_Min) or calibrated as [Ca²⁺]_SR=[400 μm× (F–F_Min)/(F_Max–F)] (Shannon et al. 2003; Zima et al. 2011). F_Min is the quench-corrected (15%) fluorescence value following complete emptying of the SR with 10 mm caffeine, and F_Max is taken as the diastolic fluorescence at 1 Hz pacing in the presence of 1 μm ISO.

Figure 2. Direct [Ca²⁺]_SR measurements following isoproterenol application.

A, confocal fluorescence image (a) of a myocyte with the SR loaded with the low-affinity Ca²⁺ indicator fluo-5N. Enlarged region of image (b) and corresponding fluorescence profile (c) showing patterning of the junctional SR network. AU, arbitrary units of fluo-5N fluorescence. B, example [Ca²⁺]_SR measurements with fluo-5N fluorescence showing an electrically induced Ca²⁺ depletion (arrow, 0.75 Hz) followed by rest under control conditions (a and b), in the presence of ISO (c) (a–c: [Ca²⁺]_o= 2 mm), and in the presence of ISO with 7.5 μm CPA in 100 μm[Ca²⁺]_o (d and e). In traces b and e 250 μm caffeine (250 Caff) was applied to pharmacologically sensitize RyRs and induce Ca²⁺ waves. Ca²⁺ waves are visible in fluo-5N fluorescence profiles as a non-triggered decrease in fluorescence (stars). Summary data of resting [Ca²⁺]_SR (C) and Ca²⁺ wave occurrence (D) under the conditions shown in B. Ca²⁺ wave occurrence is presented as fraction (%) of cells showing waves, from a total of eight cells. *P < 0.05.

Figure 3. Ca²⁺ waves during sustained rest in the presence of isoproterenol.

A, simultaneous [Ca²⁺]_i (rhod-2, upper traces in black) and [Ca²⁺]_SR (fluo-5N, lower traces in grey) measurements following rest from electrical stimulation under control conditions (7 mm[Ca²⁺]_o, a= 1.1 Hz, b= 1.3 Hz) and in the presence of ISO (c: 2 mm[Ca²⁺]_o, 1.3 Hz). Electrical stimulation frequency was incrementally increased to elevate [Ca²⁺]_SR and determine the threshold level for Ca²⁺ waves. In the example shown, Ca²⁺ waves were only observed when [Ca²⁺]_SR was higher than the control threshold level, unless RyRs were pharmacologically sensitized using 250 μm caffeine. Dashed lines: 1, no-wave threshold control; 2, wave threshold control; 3, wave threshold ISO. B, positive correlation between electrical stimulation frequency (range 0.5–1.3 Hz) and resting [Ca²⁺]_SR measured using fluo-5N fluorescence. Frequency and [Ca²⁺]_SR values are normalized to the respective values found at the lowest pacing frequency in each experimental trial.

Cytosolic Ca²⁺ ([Ca²⁺]_i) measurements, indo-1

For ratiometric epifluorescence [Ca²⁺]_i measurements, myocytes were loaded with 5 μm indo-1/AM (Molecular Probes-Life Technologies) for 10 min followed by a 30 min wash, all at room temperature. Single myocytes were exposed to 360 nm light to excite indo-1, and fluorescence emission was simultaneously monitored at 410 nm (F₄₁₀) and 485 nm (F₄₈₅). Fluorescence emission at each wavelength was background subtracted and [Ca²⁺]_i is presented as the indo-1 ratio (R=F₄₁₀/F₄₈₅) or change in indo-1 ratio (ΔR=R_peak–R_baseline).

Cytosolic Ca²⁺ measurements, rhod-2

Myocytes preloaded with fluo-5N were subsequently loaded with the spectrally distinct Ca²⁺ indicator rhod-2/AM (Molecular Probes-Life Technologies, 5 μm, 10 min loading time, 10 min wash, all at room temperature) to monitor cytosolic Ca²⁺ waves simultaneously with [Ca²⁺]_SR. Laser scanning fluorescence confocal microscopy was performed (Nikon A1R, 60 frames s⁻¹) with rhod-2 excitation at 543 nm and fluorescence emission recorded at >600 nm. Rhod-2 fluorescence signals are presented as F/F₀, where F₀ is the diastolic fluorescence at the start of each experimental recording.

Experimental protocols to determine the Ca²⁺ wave threshold

Control conditions

Two series of experiments were conducted to determine the Ca²⁺ wave threshold. In one series, the Ca²⁺ wave threshold was determined based on direct measurement of [Ca²⁺]_SR using fluo-5N, whereas the second series relied on cytosolic [Ca²⁺]_i measurements (with indo-1) in response to caffeine challenge to estimate SR Ca²⁺ content. Because rabbit ventricular myocytes did not typically exhibit Ca²⁺ waves under control conditions ([Ca²⁺]_o= 2 mm), a high extracellular Ca²⁺ solution was applied ([Ca²⁺]_o= 7 mm). In both series of experiments rabbit ventricular myocytes underwent several consecutive experimental trials defined here as an episode consisting of electrical field stimulation of cells for a period of 30–60 s, followed by 8 s of rest. The stimulation–rest protocol was repeated at incrementally increasing pacing frequencies (0.05–0.2 Hz increments) to increase [Ca²⁺]_SR (see Fig. 3B) until Ca²⁺ waves were observed during the rest period. In fluo-5N experiments the [Ca²⁺]_SR where waves were observed was defined as the ‘wave threshold’ and the [Ca²⁺]_SR at the preceding lower frequency where waves were not observed was defined as the ‘no-wave threshold’. Control measurements confirmed that the control wave and no-wave thresholds were stable and reproducible over time. Additionally, in experiments where waves were observed, SR Ca²⁺ refilling time and wave latency were calculated. SR Ca²⁺ refilling time was defined as the interval between the AP-induced depletion nadir (minimum [Ca²⁺]_SR value) and the time point when [Ca²⁺]_SR recovered to the [Ca²⁺]_SR observed prior to the depletion (dark grey in Fig. 6A). Wave latency was calculated as the interval between the time point when [Ca²⁺]_SR recovered from an AP-induced depletion and the onset of the wave (light grey in Fig. 6A). In indo-1 experiments the SR Ca²⁺ content associated with a particular stimulation frequency was determined in a separate pacing train using the amplitude of the Ca²⁺ transient in response to 10 mm caffeine. This value was determined immediately following cessation of pacing.

Figure 6. Wave latency is highly dependent on [Ca²⁺]_SR.

A, representative fluo-5N fluorescence traces from a cell with [Ca²⁺]_SR at the wave threshold (a, threshold marked with dashed line 1) and above the wave threshold (b, steady-state [Ca²⁺]_SR before wave marked with dashed line 2). In each trace, arrow denotes AP stimulation triggering [Ca²⁺]_SR depletion, followed by SR Ca²⁺ refilling (dark grey) and a wave latency time (light grey) prior to the Ca²⁺ wave (marked by star). B, individual paired experiments obtained from nine cells, each yielding a wave latency measurement at two different [Ca²⁺]_SR levels under control conditions (Ctl low, Ctl high). C, individual paired experiments obtained from six cells, each yielding a wave latency measurement in the presence of ISO (grey square) alone and in the combined presence of ISO plus CPA or TG (ISO+CPA/TG black square). D, summary data (averages from data shown in B and C) of the relationship between [Ca²⁺]_SR and Ca²⁺ wave latency under control conditions at two different [Ca²⁺]_SR levels (circles), in ISO (grey square) and in ISO+CPA/TG (black square). Statistical significance analysis of these data is presented in the text.

β-Adrenergic stimulation

In all experiments β-adrenergic stimulation (ISO, 5 min) increased [Ca²⁺]_SR compared with control conditions and triggered Ca²⁺ waves. To determine the critical [Ca²⁺]_SR threshold level where Ca²⁺ waves occurred (or ceased to occur when [Ca²⁺]_SR was lowered) a protocol was applied that resulted in a gradual decrease of [Ca²⁺]_SR levels after the initial increase, all in the maintained presence of ISO. For this, [Ca²⁺]_o was lowered (as indicated) and sarcoplasmic/endoplasmic reticulum calcium ATPase (SERCA) was partially blocked by CPA or TG while β-adrenergic stimulation was maintained. In the fluo-5N experiments this protocol allowed for direct assessment of the [Ca²⁺]_SR where waves occurred during β-adrenergic stimulation, with subsequent paired comparison with the respective control threshold levels. With the indo-1 experimental approach wave and no-wave thresholds under control conditions were determined as described above. However, in the presence of ISO+CPA only no-wave data points were used for analysis. This analysis design was implemented because the indirect assessment of SR Ca²⁺ content from the caffeine-induced cytosolic Ca²⁺ transient would have to occur after a Ca²⁺ wave. With CPA or TG present the amount of [Ca²⁺]_SR remaining in the SR following a Ca²⁺ wave is significantly lower than that which preceded the wave due to the impaired reuptake by SERCA (Díaz et al. 1997a; Domeier et al. 2010; see Figs 2B, 3A and 4A), i.e. application of caffeine after a Ca²⁺ wave would grossly underestimate the wave threshold.

Figure 4. Isoproterenol increases the intra-SR Ca²⁺ wave threshold, determined by direct [Ca²⁺]_SR measurements.

A, simultaneous [Ca²⁺]_i (rhod-2, upper traces in black) and [Ca²⁺]_SR (fluo-5N, lower traces in grey) measurements following rest from electrical stimulation (arrows) under control conditions (a: 7 mm[Ca²⁺]_o, 0.7 Hz; b: 7 mm[Ca²⁺]_o, 0.75 Hz), in the presence of ISO (c: 2 mm[Ca²⁺]_o, 0.75 Hz) and in the presence of ISO with 3 μm CPA (d and e: 1 mm[Ca²⁺]_o, 0.75 Hz). Dashed lines: 1, no-wave, threshold control; 2, wave threshold, control; 3, no-wave threshold, ISO. Subsequent application of 250 μm caffeine induced a Ca²⁺ wave (e). Waves are marked by stars. B, normalized [Ca²⁺]_SR values from 18 individual experimental trials (from 12 cells) in ISO+CPA/TG separated by the criteria whether waves were observed (filled circle) or not (open circles), together with the mean ± SEM of [Ca²⁺]_SR for the two groups. For each individual trial [Ca²⁺]_SR was normalized to the maximal fluorescence observed in the presence of 1 μm ISO ([Ca²⁺]_{SR,normalized}= 1) and the no-wave threshold under control conditions ([Ca²⁺]_{SR,normalized}= 0). C, quantification of the [Ca²⁺]_SR associated with the no-wave threshold under control conditions (open bar) and in the presence of ISO+CPA/TG (shaded bar). *P < 0.002 vs. control, n= 7 cells.

Statistics and data analysis

Only rod-shaped myocytes with clear striations were used for experimental protocols. Approximately 20% of myocytes exhibited visible damage during the repetitive stimulation/rest protocols (e.g. membrane blebbing or irreversible cellular contracture). These myocytes were not utilized for data analysis. Data are presented as individual observations from single myocytes or as mean ± standard error of the mean (SEM). Statistical comparisons were performed using Student's t test for paired or unpaired data, with significance set at P < 0.05. The number of individual cells or experimental trials is given as n as indicated.

Results

In this investigation we tested the hypothesis that β-adrenergic receptor stimulation alters the intra-SR Ca²⁺ threshold for spontaneous Ca²⁺ waves. For this, we monitored spontaneous Ca²⁺ waves that occurred during rest from steady-state electrical stimulation. A major difficulty associated with studies of SR Ca²⁺ release following activation of β-adrenergic stimulation is that intracellular Ca²⁺ handling is profoundly altered by this signalling pathway, most notably an increase in Ca²⁺ entry into the cell via the LTCC (Tsien et al. 1986) and an increase in SR Ca²⁺ content via effects on phospholamban and the SERCA pump (Kranias & Solaro, 1982; Lindemann et al. 1983). Because of these effects, it becomes experimentally difficult to control both [Ca²⁺]_i and [Ca²⁺]_SR, which is essential when monitoring SR Ca²⁺ release as RyR gating is influenced by both cytosolic and intra-luminal [Ca²⁺] (Fill & Copello, 2002). We therefore defined experimental conditions where Ca²⁺ cycling (in particular [Ca²⁺]_i) was the same between control conditions and in the presence of β-adrenergic stimulation (1 μm ISO, 5 min), and reasoned that if during rest from electrical stimulation steady-state [Ca²⁺]_i is equivalent between the two respective conditions then the [Ca²⁺]_SR where Ca²⁺ waves occur will directly reflect the intra-SR Ca²⁺ wave threshold.

Acute application of ISO led to an increase in Ca²⁺ transient amplitude (Fig. 1A and C, arrows indicate electrically evoked Ca²⁺ transients), resting [Ca²⁺]_i (Fig. 1A and D) and SR Ca²⁺ content as assessed from the amplitude of the [Ca²⁺]_i increase in response to 10 mm caffeine (Fig. 1B and E). Under these conditions, spontaneous Ca²⁺ waves were frequently observed during the rest period following steady-state stimulation (Fig. 1A, star). Subsequent addition of a low extracellular Ca²⁺ solution ([Ca²⁺]_o= 100 μm) containing ISO and the SERCA inhibitor CPA (7.5 μm) returned critical parameters of cellular Ca²⁺ handling to levels observed under control conditions, with Ca²⁺ transient amplitude (Fig. 1C), resting [Ca²⁺]_i (Fig. 1D) and SR Ca²⁺ content (Fig. 1E) becoming similar to the original control conditions.

Figure 1. Effects of isoproterenol on intracellular Ca²⁺ handling.

A, example cytosolic Ca²⁺ transients measured with indo-1 (0.75 Hz, electrical stimuli marked by arrows) followed by rest under control conditions ([Ca²⁺]_o= 2 mm) and in the presence of 1 μm ISO. Control resting [Ca²⁺]_i level is marked by dashed line. Note spontaneous Ca²⁺ wave that occurred during the rest period in the presence of ISO (star). B, example [Ca²⁺]_i increases in response to 10 mm caffeine (black bars) used to assess SR Ca²⁺ content. Caffeine was applied immediately after identical 0.75 Hz pacing trains as shown in A, prior to the occurrence of spontaneous Ca²⁺ waves. Summary data of Ca²⁺ transient amplitude (0.75 Hz; C), resting [Ca²⁺]_i (D) and SR Ca²⁺ load (E) under control conditions, in the presence of ISO, and in the presence of ISO with 7.5 μm cyclopiazonic acid (CPA) in 100 μm[Ca²⁺]_o. *P < 0.05 versus control.

We next loaded the SR of cardiomyocytes with the low affinity Ca²⁺ indicator fluo-5N to directly and dynamically monitor the [Ca²⁺]_SR when waves occur. Figure 2A shows the typical fluo-5N SR staining pattern. Based on observations of increased SR Ca²⁺ release in the form of Ca²⁺ sparks (Zhou et al. 2009; Ogrodnik & Niggli, 2010), Ca²⁺ leak (Curran et al. 2010; Bovo et al. 2012; Ullrich et al. 2012) and Ca²⁺ waves (Curran et al. 2007) in the presence of β-adrenergic stimulation, our initial hypothesis was that ISO would decrease the [Ca²⁺]_SR threshold where waves occurred (i.e. Ca²⁺ waves would be observed in the presence of ISO at a [Ca²⁺]_SR at or below the level where waves were not observed under control conditions). In rabbit ventricular myocytes Ca²⁺ waves were not typically observed under control conditions with 2 mm[Ca²⁺]_o (Fig. 2Ba and D). However Ca²⁺ waves could be triggered using a low concentration (250 μm) of caffeine (Fig. 2Bb and D), a RyR agonist that sensitizes the release channel to Ca²⁺ and lowers the intra-SR Ca²⁺ threshold for Ca²⁺ waves (Venetucci et al. 2007; Kong et al. 2008). Application of ISO caused a large increase in [Ca²⁺]_SR (Fig. 2Bc and C) and Ca²⁺ waves were observed (Fig. 2Bc and D). We then applied ISO+CPA ([Ca²⁺]_o= 100 μm) to lower [Ca²⁺]_SR. When [Ca²⁺]_SR decreased to a similar level as under control conditions no Ca²⁺ waves were observed (Fig. 2Bd and D). Subsequent addition of 250 μm caffeine to myocytes treated with ISO+CPA triggered Ca²⁺ waves similarly to control conditions (Fig. 2Be and D). In summary, the data shown in Figs 1 and 2 indicate that under conditions of identical [Ca²⁺]_i and [Ca²⁺]_SR ISO stimulation failed to increase the propensity of Ca²⁺ waves and therefore does not appear to lower the Ca²⁺ wave threshold.

We next tested the alternative hypothesis that β-adrenergic stimulation increases the [Ca²⁺]_SR threshold where waves occur. For this we designed experiments to observe Ca²⁺ waves under control conditions and to determine if, in the presence of ISO, Ca²⁺ waves were absent at the same [Ca²⁺]_SR level or would require even higher [Ca²⁺]_SR to occur. As spontaneous Ca²⁺ waves were not observed in normal [Ca²⁺]_o of 2 mm (Fig. 2Ba and D), [Ca²⁺]_o was raised to 7 mm and electrical stimulation frequency was incrementally increased to elevate [Ca²⁺]_SR (Fig. 3B) until Ca²⁺ waves were observed during rest from stimulation (see Methods for details on the stimulation–rest protocol). The highest near steady-state [Ca²⁺]_SR where Ca²⁺ waves were absent was defined as the control no-wave threshold (level 1 in Fig. 3Aa) while the lowest [Ca²⁺]_SR that associated with Ca²⁺ waves was defined as the control wave threshold (level 2 in Fig. 3Ab). Across all experiments (n= 21 cells), the stimulation frequencies associated with the no-wave and wave thresholds were 0.83 ± 0.05 and 0.93 ± 0.05 Hz, respectively. We then applied ISO to the same cell which increased [Ca²⁺]_SR due to its stimulatory effect on SERCA. As [Ca²⁺]_SR subsequently declined during an extended period of rest we determined the lowest [Ca²⁺]_SR that preceded a Ca²⁺ wave (ISO wave threshold; level 3 in Fig. 3Ac). As shown in Fig. 3A (trace c), Ca²⁺ waves were frequent with ISO but were never observed at a [Ca²⁺]_SR below the control no-wave threshold (level 1), unless RyRs were sensitized with a low concentration of caffeine (250 μm).

We next supported these findings with a complementary experimental approach that allowed Ca²⁺ waves to be observed in a defined time interval (8 s) following rest from steady-state stimulation. For this the rest period where threshold levels were determined was preceded by a 30–60 s interval during which cells were electrically stimulated to establish identical steady-state conditions for the SR Ca²⁺ release mechanism. After control ([Ca²⁺]_o= 7 mm) no-wave (level 1 in Fig. 4Aa) and wave (level 2 in Fig. 4Ab) thresholds were determined, ISO was applied for 5 min leading to an increase in [Ca²⁺]_SR and frequent Ca²⁺ waves (Fig. 4Ac). [Ca²⁺]_SR was then gradually reduced using application of ISO in the presence of low [Ca²⁺]_o (1 mm) and the SERCA inhibitor CPA (3 μm). In the example shown waves were no longer observed following rest from stimulation (Fig. 4Ad), unless RyRs were pharmacologically sensitized with 250 μm caffeine (Fig. 4Ae). Level 3 in Fig. 4Ad represents the no-wave threshold in the presence of ISO. Additional experiments were performed using the alternative SERCA inhibitor TG (1 μm) with similar results, and these results were pooled to provide the summary data shown in Fig. 4B and C. From 12 individual myocytes measurements of the no-wave and wave thresholds, as well as [Ca²⁺]_SR in the presence of ISO, followed by exposure to a SERCA blocker were obtained. In this set of experiments the no-wave threshold was 0.88 ± 0.01 [(F–F_Min)/(F_Max–F_Min)] and the wave threshold was 0.90 ± 0.01 (P < 0.05, n= 12 cells, paired comparison). In the presence of ISO alone, all 12 cells exhibited Ca²⁺ waves, presumably due to the substantial increase of [Ca²⁺]_SR above the control threshold with (F–F_Min)/(F_Max–F_Min) values approaching 1. Myocytes were subsequently exposed to ISO+CPA or ISO+TG for examination of the [Ca²⁺]_SR associated with waves (filled circles in Fig. 4B) or lack of waves (open circles). From each individual cell one or two measurements could be obtained for a total number of 18 experimental trials. In contrast to ISO alone, in the presence of ISO+CPA/TG waves were only observed in 8 of 18 experimental trials where [Ca²⁺]_SR was above the control no-wave threshold. The observation that in a majority of experimental trials waves were absent in the presence of ISO at [Ca²⁺]_SR levels where in control conditions waves were observed was indicative of an increase in the Ca²⁺ wave threshold. Quantitative analysis of the [Ca²⁺]_SR associated with the no-wave threshold under the respective conditions showed that this level was significantly higher in the presence of ISO+CPA/TG (Fig. 4C). The no-wave threshold under control conditions was at [Ca²⁺]_SR= 2.58 ± 0.24 mm (n= 7 cells) and increased to 3.39 ± 0.23 mm in the presence of ISO+CPA/TG (n= 7 cells; P= 0.002 vs. control). At [Ca²⁺]_SR below the control no-wave threshold, Ca²⁺ waves were not observed with ISO+CPA/TG (data not shown), unless RyRs were sensitized pharmacologically with 250 μm caffeine. In the presence of 250 μm caffeine Ca²⁺ waves were observed in all experimental trials (eight Ca²⁺ waves in eight cells) at a [Ca²⁺]_SR below the control no-wave threshold, indicative of a decrease in the Ca²⁺ wave threshold. Quantification of the no-wave threshold in the presence of 250 μm caffeine and ISO+CPA/TG showed that wave activity ceased when [Ca²⁺]_SR dropped to 1.00 ± 0.11 mm (n= 5 cells), a level that was significantly (P < 0.001) below the control and ISO+CPA/TG no-wave thresholds.

The direct and dynamic [Ca²⁺]_SR measurements clearly revealed an increase in the Ca²⁺ wave threshold during β-adrenergic stimulation. To further support these direct measurements we obtained additional measurements of the no-wave threshold in the presence of ISO+CPA using the amplitude of the [Ca²⁺]_i increase in response to 10 mm caffeine to estimate SR Ca²⁺ content.

Figure 5A shows original traces to determine control no-wave (level 1) and wave (level 2) thresholds, followed by the no-wave threshold in the presence of ISO+CPA (level 3). As shown in example traces of Fig. 5A and summary data of Fig. 5B, the SR Ca²⁺ content associated with the disappearance of Ca²⁺ waves with ISO+CPA (ISO no-wave threshold) was significantly higher than the wave threshold observed under control conditions. Importantly, resting [Ca²⁺]_i was similar under these experimental conditions (Fig. 5C; dotted line in Fig. 5A), which excluded the possibility that ISO-dependent changes in global cytosolic [Ca²⁺]_i were responsible for the observed effects on the wave threshold. Thus, the observed differences in SR Ca²⁺ content (Figs 4A and 5B) where waves occurred reflect a true increase in the intra-SR Ca²⁺ wave threshold.

Figure 5. Isoproterenol increases the intra-SR Ca²⁺ wave threshold determined by cytosolic [Ca²⁺]_i measurements.

A, example indo-1 [Ca²⁺]_i measurements following rest from electrical stimulation (arrows) under control conditions (7 mm[Ca²⁺]_o, a= 0.6 Hz, b= 0.7 Hz) and in the presence of ISO with 3 μm CPA (c: 1 mm[Ca²⁺]_o, 0.7 Hz). Caffeine (10 mm) was applied after rest from a separate pacing train to assess SR Ca²⁺ content at each pacing frequency under control conditions and to determine no-wave (dashed line 1) and wave (dashed line 2) threshold SR Ca²⁺ content. In the presence of ISO+CPA, when waves were not observed following rest from pacing, 10 mm caffeine was applied to assess the no-wave threshold in ISO (dashed line 3). Star denotes Ca²⁺ wave. Dotted line: resting [Ca²⁺]_i. B, summary data of SR Ca²⁺ content under the conditions shown in A (n= 8 cells). C, average resting [Ca²⁺]_i was similar under these experimental conditions (n= 7 cells). *P < 0.05.

Recent work has proposed that, in addition to the Ca²⁺ wave threshold, altered RyR ‘refractoriness’ may be a critical factor regulating SR Ca²⁺ release and the generation of Ca²⁺ waves (Ramay et al. 2011; Belevych et al. 2012). This concept is supported by direct measurements of [Ca²⁺]_SR which show that after [Ca²⁺]_SR recovers to a steady-state level following electrical stimulation a finite time interval (latency period) remains prior to Ca²⁺ wave initiation (Figs 2B, 3A and 4A; see also Domeier et al. 2010; Belevych et al. 2012; Maxwell et al. 2012). We therefore examined SR Ca²⁺ refilling, [Ca²⁺]_SR and wave latency in experiments where waves were observed. In these experiments two measurements of [Ca²⁺]_SR, SR Ca²⁺ refilling times and Ca²⁺ wave latency were obtained from an individual cell (nine cells total), one at the wave threshold (Fig. 6Aa) and one after [Ca²⁺]_SR was increased above the control wave threshold (Fig. 6Ab). [Ca²⁺]_SR was increased by increasing the pacing frequency prior to the rest period during which Ca²⁺ waves were observed. Under control conditions SR Ca²⁺ refilling time (marked by dark grey area in Fig. 6A) was independent of initial [Ca²⁺]_SR (416 ± 25 ms at [Ca²⁺]_SR= 0.95 ± 0.01 (F–F_Min)/(F_Max–F_Min), versus 416 ± 28 ms at [Ca²⁺]_SR= 0.91 ± 0.01; n= 9 cells). In the presence of ISO, SR Ca²⁺ refilling time was significantly shorter than under control conditions (227 ± 26 ms, n= 6 cells, P < 0.001 compared with control), and was slowed by subsequent SERCA inhibition with CPA or TG (367 ± 96 ms). The acceleration of SR refilling in the presence of ISO is explained by the stimulatory β-adrenergic effect on SERCA.

Analysis of wave latency revealed several interesting features. In control conditions (Fig. 6B; circles in Fig. 6D) as well as in the presence of ISO (Fig. 6C; squares in Fig. 6D), wave latency was highly dependent on [Ca²⁺]_SR, with higher [Ca²⁺]_SR leading to a significantly shorter wave latency. Furthermore, in the presence of ISO the dependence of wave latency on [Ca²⁺]_SR shifted towards longer latencies, i.e. for a given SR Ca²⁺ content, the wave latency was prolonged in the presence of ISO compared with control conditions. For example, despite the fact during ISO treatment [Ca²⁺]_SR was significantly higher than under control conditions [ISO (grey square) 1.0 ± 0 vs. control (grey circle) 0.95 ± 0.01; P < 0.001], latency times were similar (ISO 1.02 ± 0.22 s vs. control 1.44 ± 0.27 s; not significantly different at P= 0.38). Furthermore, comparison of control conditions (Fig. 6D, grey circle) with ISO+CPA/TG treatment (Fig. 6D, black square) showed that at a similar [Ca²⁺]_SR (control 0.95 ± 0.01 vs. ISO+CPA 0.93 ± 0.02; not significantly different at P= 0.28) latency times were significantly shorter under control conditions (control 1.44 ± 0.27 s vs. ISO+CPA/TG 4.42 ± 0.85 s; P < 0.01).

In summary, our study shows that during β-adrenergic stimulation the propensity for Ca²⁺ waves is increased because of significantly increased [Ca²⁺]_SR and SR Ca²⁺ overload. Counteracting the increase in [Ca²⁺]_SR are changes in two [Ca²⁺]_SR-dependent parameters, with an increase in Ca²⁺ wave threshold and a prolonged wave latency.

Discussion

During sympathetic nervous system activation the heart experiences chronotropic, inotropic and lusitropic effects that act in concert to facilitate an increase in cardiac output. At the level of the individual cardiomyocyte catecholamine binding to β-adrenoceptors causes an increase in cellular Ca²⁺ cycling that associates with Ca²⁺ waves, suggesting an acute lowering of the intra-SR Ca²⁺ wave threshold. However, our data using direct, dynamic measurements of this threshold clearly show that to the contrary, the threshold for Ca²⁺ waves is increased by acute β-adrenergic stimulation. The direct [Ca²⁺]_SR measurements presented here support recent quantification of the Ca²⁺ wave threshold in mouse ventricular myocytes which show a similar effect of the Ca²⁺ wave threshold increasing following acute β-adrenergic stimulation (Kashimura et al. 2010; Stokke et al. 2010a).

The Ca²⁺ wave threshold is a useful experimental parameter that has greatly advanced our understanding of the mechanisms of cardiac arrhythmia, including those associated with diseases such as CPVT and heart failure. Experimental evidence and computational modelling implicate numerous intracellular factors in Ca²⁺ wave generation and propagation, including RyR open probability, release junction (sarcomere) spacing, SR Ca²⁺ uptake via SERCA, and cytosolic and intra-SR Ca²⁺ buffering processes (Izu et al. 2006; Ramay et al. 2010; Swietach et al. 2010). The appearance of Ca²⁺ waves as [Ca²⁺]_SR increases can be explained by well-established RyR gating mechanisms, most notably that the RyR open probability increases at high [Ca²⁺]_SR due to the channel's inherent sensitivity to [Ca²⁺]_SR (Gyorke & Gyorke, 1998; Fill & Copello, 2002). Furthermore, because Ca²⁺ release terminates at a set level of [Ca²⁺]_SR (Zima et al. 2008) Ca²⁺ sparks will be high-amplitude release events at high [Ca²⁺]_SR, which will increase the ability of Ca²⁺ released during a spark to overcome cytosolic buffering and diffuse to adjacent release junctions and trigger CICR. These properties result in the appearance of Ca²⁺ waves at a defined [Ca²⁺]_SR (the wave threshold), and make the time to wave initiation highly dependent on [Ca²⁺]_SR (Fig. 6).

Direct measurements of [Ca²⁺]_SR (Fig. 6) illustrate that Ca²⁺ waves do not occur immediately after [Ca²⁺]_SR reaches a steady-state concentration (Domeier et al. 2010; Belevych et al. 2012; Maxwell et al. 2012) and therefore a time-dependent ‘refractory’ component of wave generation exists (Belevych et al. 2012). While this appears to conceptually challenge the Ca²⁺ wave threshold as being the critical parameter that determines whether a Ca²⁺ wave will be observed, our data suggest that changes in wave refractory (latency) time may simply be a manifestation of how high [Ca²⁺]_SR is above the Ca²⁺ wave threshold. In an individual cell the Ca²⁺ wave threshold was highly reproducible among experimental trials, yet different wave latency times were recorded depending on [Ca²⁺]_SR, with shorter times observed above the wave threshold than at the wave threshold (Fig. 6B). This relationship was similarly observed in canine ventricular myocytes (Belevych et al. 2012). Based on these data our working model (Fig. 7) is that the cell has a distinct intra-luminal [Ca²⁺]_SR threshold for Ca²⁺ wave development (Díaz et al. 1997b; Jiang et al. 2004; Venetucci et al. 2007). At [Ca²⁺]_SR below this threshold the probability for wave development following rest from steady-state stimulation is low and thus the cell has a wave latency period that approaches infinity (Fig. 7a). When [Ca²⁺]_SR reaches the Ca²⁺ wave threshold the probability that a Ca²⁺ wave will occur increases, yet due to the stochastic nature of Ca²⁺ wave initiation there will be a finite latency period associated with the wave (Fig. 7b). As [Ca²⁺]_SR continues to increase above this distinct wave threshold, the Ca²⁺ wave probability remains high but now the latency period will be shortened due to the stimulatory effect of luminal [Ca²⁺]_SR on RyR open probability, increasing the likelihood of a spontaneous release event to trigger a Ca wave (Fig. 7c) (Gyorke & Gyorke, 1998). During β-adrenergic stimulation, despite the observation that the wave threshold is increased, the massive gain of Ca²⁺ in the SR results in a further shortening of the latency period (Fig. 7f). Overall, the latency period is inversely proportional (Fig. 6D) to the degree by which [Ca²⁺]_SR exceeds the wave threshold. For a given [Ca²⁺]_SR level, however, the latency period is longer in the presence of ISO (Fig. 7d) than under control conditions (Fig. 7c), and is similar when [Ca²⁺]_SR exceeds the respective wave thresholds to the same degree (Fig. 7c vs. Fig. 7e). Thus, there may be an important distinction between the parameter of the Ca²⁺ wave threshold and that of the wave latency period. A cell may have multiple latency periods depending on how high [Ca²⁺]_SR is above a distinct wave threshold (Fig. 6), and thus the wave threshold measurement is a more biophysically accurate way to assess the ability of the SR to sequester (or spontaneously release) Ca²⁺. Alternatively, however, the wave latency period may be the most physiologically relevant way to describe diastolic Ca²⁺ waves with respect to cellular SR Ca²⁺ content (i.e. inotropic state of heart) and rate of AP stimulation (i.e. chronotropic state of the heart). In vivo these complex relationships will determine if an arrhythmogenic Ca²⁺ wave will be observed during diastole in the beating heart.

Figure 7. Schematic diagram of proposed relationship between SR Ca²⁺ content, Ca²⁺ wave threshold and latency.

Schematic diagram of the SR showing the relationship between [Ca²⁺]_SR (shaded area), the Ca²⁺ wave threshold (dashed black line) and Ca²⁺ wave latency when [Ca²⁺]_SR is below (a), at (b) and above (c) the Ca²⁺ wave threshold. In the presence of ISO (d, e and f) the Ca²⁺ wave threshold (dashed black line) is increased when compared with control conditions (dashed grey line). When [Ca²⁺]_SR is similar between control condition (c) and ISO condition (d), latency is shorter in control due to [Ca²⁺]_SR being higher above the respective wave threshold. When [Ca²⁺]_SR is elevated by the same relative amount above the wave threshold under control conditions (c) and with ISO (e) wave latencies are similar even though the total [Ca²⁺]_SR is higher with ISO. Under most experimental conditions [Ca²⁺]_SR increases dramatically in the presence of ISO, thus elevating [Ca²⁺]_SR well above the respective Ca²⁺ wave threshold, resulting in a very short wave latency (f).

β-Adrenergic modulation of the RyR and Ca²⁺ waves

Alterations in RyR activity (e.g. via phosphorylation or redox modification) will lead to subsequent changes in the intra-SR Ca²⁺ wave threshold, as shown by RyR inhibitors (Overend et al. 1997) or activators (Venetucci et al. 2007; Kong et al. 2008) raising or lowering the threshold, respectively. The signalling events that regulate RyR activity downstream of β-adrenergic stimulation remain highly contentious, with conflicting reports throughout the literature (for example see commentaries by Yamaguchi & Meissner, 2007; Houser, 2010; Bers, 2012). Most investigations, however, report an increase in RyR activity which would result in a decrease in the Ca²⁺ wave threshold. The observation of an increased Ca²⁺ wave threshold is therefore counterintuitive. Discrepant results may emerge due to the multiple assays of RyR function which may not be mechanistically equivalent, particularly following activation of complex cellular signalling pathways (reviewed by George, 2008). As an example, systolic Ca²⁺ release in response to the AP, representing activation of RyRs via cytosolic Ca²⁺, may exhibit distinct regulation when compared to diastolic Ca²⁺ release triggered by SR Ca²⁺ overload and activation of RyRs by [Ca²⁺]_SR. The properties of local diastolic Ca²⁺ release events (Ca²⁺ spark frequency, Ca²⁺ spark amplitude, non-spark-mediated SR Ca²⁺ leak) may also be differentially modulated (Zhou et al. 2009) and subsequently alter the ability of Ca²⁺ waves to initiate and propagate. Recently, novel genetic approaches have been utilized to ablate (Wehrens et al. 2006; Benkusky et al. 2007; Chelu et al. 2009) or chronically activate (Shan et al. 2010; van Oort et al. 2010) RyR phosphorylation sites in vivo, and gain additional insight into RyR regulation that may occur downstream of β-adrenergic signalling. However, this strategy may result in a cellular phenotype that is distinct from that observed when all cellular effectors are activated in parallel following receptor activation. Complex signal integration at the level of the RyR is expected given the multiple regulatory sites on the channel protein and the role of the macromolecular RyR complex as a subcellular signalling platform.

β-Adrenergic modulation of SERCA and Ca²⁺ waves

Although the Ca²⁺ wave threshold is primarily set by the regulation of RyR release channels, it is likely further tuned by altering the function of additional cellular proteins involved in wave generation. One of the main effects of β-adrenergic stimulation is to increase SERCA activity, which is accomplished via phosphorylation of the SERCA inhibitory protein phospholamban. Phospholamban phosphorylation, principally by protein kinase A although there is some evidence for a role of Ca²⁺/calmodulin-dependent protein kinase II, reduces its inhibitory interactions with SERCA and leads to a dramatic increase in SERCA activity and [Ca²⁺]_SR. Indeed, SERCA-mediated Ca²⁺ uptake increasing [Ca²⁺]_SR above the intra-SR Ca²⁺ wave threshold is one of the principal mechanisms by which β-adrenergic stimulation increases Ca²⁺ wave probability, and is supported by data showing that genetic ablation of phospholamban dramatically increases the propensity of Ca²⁺ waves (Huser et al. 1998). SERCA activity may also directly contribute to Ca²⁺ wave propagation and thus, in part, determine the intra-SR Ca²⁺ wave threshold. It has been shown experimentally that SERCA inhibition increases Ca²⁺ wave velocity (Lukyanenko et al. 1999), and computer simulations predict that increasing SERCA activity prevents Ca²⁺ wave propagation by decreasing [Ca²⁺]_i and the ability of Ca²⁺ to diffuse between Ca²⁺ release junctions (Ramay et al. 2010). However, this cytosolic Ca²⁺ buffering effect may be countered by the ability of SERCA to locally increase [Ca²⁺]_SR and create intra-SR ‘Ca²⁺ sensitization’ wavefronts which precede the cytosolic Ca²⁺ wave (Keller et al. 2007; Maxwell & Blatter, 2012). This provocative model is based on experimental evidence contradictory to that of Lukyanenko et al. (1999), which shows that SERCA inhibition decreases Ca²⁺ wave velocity (Keller et al. 2007). While this model of wave propagation can be reproduced by computer simulations (Ramay et al. 2010), it is critically dependent on the speed by which Ca²⁺ diffuses within the SR network, which is difficult to determine empirically and remains controversial (Swietach et al. 2008; Picht et al. 2011). Electrophysiological measurement of the Ca²⁺ wave threshold in rat myocytes showed that SERCA inhibition decreased the intra-SR Ca²⁺ wave threshold (O’Neill et al. 2004), and these data are supported by studies of transgenic mice with reduced SERCA abundance that also have a decreased Ca²⁺ wave threshold, although interpretation of data from this model may be complicated by concomitant compensatory changes in RyR function (Stokke et al. 2010a,b). However, in mouse ventricular myocytes where the Ca²⁺ wave threshold was increased following β-adrenergic stimulation, the Ca²⁺ wave threshold did not appear to correlate with SERCA activity (Kashimura et al. 2010), and it remains unclear if the increase in SERCA activity that accompanies β-adrenergic stimulation plays a role in increasing the intra-SR Ca²⁺ wave threshold.

β-Adrenergic modulation of cellular buffering and Ca²⁺ waves

As predicted by mathematical models, changes in the buffering properties of the cytosol and/or SR would also alter Ca²⁺ wave propagation and the intra-SR Ca²⁺ wave threshold. β-Adrenoceptor activation leads to a decrease in myofilament Ca²⁺ sensitivity which is believed in part to underlie the lusitropic effect of catecholamines on the heart. This change will likely lead to altered Ca²⁺ buffering properties and Ca²⁺ transport kinetics, although it remains unclear if these changes alter Ca²⁺ wave propagation. Recent evidence suggests that increased Ca²⁺ sensitivity promotes arrhythmogenesis, leading to the hypothesis that increased myofilament Ca²⁺ affinity may augment Ca²⁺ wave propagation by altering diastolic contracture and shortening the distance between release clusters (Izu et al. 2006; Chen-Izu et al. 2007). Changes in intra-SR Ca²⁺ buffering properties have also been shown to affect Ca²⁺ wave propagation (Kubalova et al. 2004). While there is experimental evidence that the intra-SR Ca²⁺ binding protein calsequestrin may be regulated by phosphorylation (Cala & Jones, 1991), and that phosphorylation increases Ca²⁺ binding capacity of calsequestrin (Beard et al. 2008), it is unclear if this phosphorylation occurs in response to acute activation of intracellular signalling pathways. Furthermore, the addition of exogenous intra-SR Ca²⁺ buffers alters Ca²⁺ wave frequency and amplitude without changing the intra-SR Ca²⁺ wave threshold (Kubalova et al. 2004), and thus it is unlikely that the acute effect of β-adrenergic stimulation on the intra-SR Ca²⁺ wave threshold is due to changes in intra-SR Ca²⁺ binding protein function.

Summary

β-Adrenergic stimulation enhances intracellular Ca²⁺ cycling and increases SR Ca²⁺ load to the extent that spontaneous and potentially detrimental Ca²⁺ waves occur. As shown in this study the β-adrenoceptor-mediated increase in SR Ca²⁺ load is also accompanied by an increase of the intra-SR Ca²⁺ wave threshold and a prolongation of the wave latency period. These two factors may be interpreted as a protective mechanism against diastolic arrhythmogenic Ca²⁺ release in the heart.

Glossary

AP

action potential

[Ca²⁺]_i

cytosolic free calcium concentration

[Ca²⁺]_o

extracellular free calcium concentration

[Ca²⁺]_SR

intra-sarcoplasmic reticulum free calcium concentration

CICR

calcium-induced calcium release

CPA

cyclopiazonic acid

CPVT

catecholaminergic polymorphic ventricular tachycardia

CRU

calcium release unit

DMSO

dimethyl sulfoxide

ISO

isoproterenol

LTCC

L-type calcium channel

RyR

ryanodine receptor

SERCA

sarcoplasmic/endoplasmic reticulum calcium ATPase

SR

sarcoplasmic reticulum

TG

thapsigargin

Author contributions

T.L.D., J.T.M. and L.A.B. contributed to the conception and design of the study, analysis and interpretation of data, and writing of the article, and have approved the final version of the manuscript. T.L.D. and J.T.M. performed the experimental work.

Author's present address

Timothy L. Domeier: Department of Medical Pharmacology and Physiology, University of Missouri School of Medicine, Columbia, MO 65212, USA.