Regulation of cardiac alternans by β-adrenergic signaling pathways

Abstract

In cat atrial myocytes, β-adrenergic receptor (β-AR) stimulation exerts profound effects on excitation-contraction coupling and cellular Ca²⁺ cycling that are mediated by β₁- and β₂-AR subtypes coupled to G proteins (G_s and G_i). In this study, we determined the effects of β-AR stimulation on pacing-induced Ca²⁺ alternans. Ca²⁺ alternans was recorded from single cat atrial myocytes with the fluorescent Ca²⁺ indicator indo-1. Stable Ca²⁺ alternans occurred at an average pacing frequency of 1.7 Hz at room temperature with a mean alternans ratio of 0.43. Nonselective β-AR stimulation as well as selective stimulation of β₁/G_s, β₂/G_s + G_i, and β₂/G_s coupled pathways all abolished pacing-induced Ca²⁺ alternans. β₁-AR stimulation abolished alternans through stimulation of PKA and Ca²⁺/calmodulin-dependent protein kinase II, whereas β₂-AR stimulation exclusively involved PKA and was mediated via G_s, whereas a known second pathway in cat atrial myocytes acting through G_i and nitric oxide production was not involved in alternans regulation. Inhibition of various mitochondrial functions (dissipation of the mitochondrial membrane potential or inhibition of mitochondrial F₁/F₀-ATP synthase, mitochondrial Ca²⁺ uptake via the mitochondrial Ca²⁺ uniporter, and Ca²⁺ extrusion via mitochondrial Na⁺/Ca²⁺ exchange) enhanced Ca²⁺ alternans; however, β-AR stimulation still abrogated alternans, provided that sufficient cellular ATP was available. Selective inhibition of mitochondrial or glycolytic ATP production did not prevent β-AR stimulation from abolishing Ca²⁺ alternans. However, when both ATP sources were depleted, β-AR stimulation failed to decrease Ca²⁺ alternans. These results indicate that in atrial myocytes, β-AR stimulation protects against pacing-induced alternans by acting through parallel and complementary signaling pathways.

cardiac alternans is defined as a beat-to-beat alternation in contraction amplitude (mechanical alternans), action potential duration (electrical or action potential duration alternans), and Ca²⁺ transient amplitude (Ca²⁺ alternans) at a constant stimulation frequency (e.g., Ref. 57). Alternans has been recognized as a risk factor for cardiac arrhythmia and sudden cardiac death (47–49), including atrial fibrillation (20, 37), and although it is considered a multifactorial process, it has become increasingly clear that alternans is ultimately linked to disturbances in myocardial Ca²⁺ homeostasis and impaired intracellular Ca²⁺ concentration ([Ca²⁺]_i) regulation (for reviews, see Refs. 5, 6, 12–14, 16, 31, 36, and 43). Based on the theory of cardiac Ca²⁺ signaling, computational studies, and experimental data (for reviews and references, see Refs. 54 and 56), two parameters have emerged as critically relevant to the generation of alternans at the cellular level: fractional Ca²⁺ release from the sarcoplasmic reticulum (SR), dependent on SR load, and the efficiency of beat-to-beat cytosolic Ca²⁺ sequestration. Ca²⁺ sequestration depends on the availability of adequate ATP supplies to fuel Ca²⁺ pumps [particularly sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA)] and to serve as a substrate for phosphorylation processes as well as on organelles capable of storing Ca²⁺. We have previously shown that conditions that are accompanied by reduced ATP generation promote the generation of alternans. Our previous studies showed that inhibition of glycolysis (5, 23, 28, 29) as well as mitochondrial function (16) enhances the propensity of electromechanical and Ca²⁺ alternans in atrial myocytes.

Sympathetic innervation of the heart elicits chronotropic, inotropic, and lusitropic effects that facilitate an increase in cardiac output and are mediated by β-adrenergic receptor (β-AR) signaling (46). At the level of cardiomyocytes, catecholamine binding to β-ARs causes an increase in cellular Ca²⁺ cycling that associates with the phosphorylation of Ca²⁺-handling proteins by cellular kinases such as PKA and Ca²⁺/calmodulin-dependent protein kinase II (CaMKII). Due to the profound influence of β-AR signaling on cardiac Ca²⁺ cycling and excitation-contraction (E-C) coupling, it is not all surprising that β-AR stimulation also affects cardiac alternans (14, 15). In clinical settings, for example, administration of antiadrenergic drugs (β-blockers) has been shown to reduce microvolt-level T-wave alternans (e.g., Refs. 26 and 27), an established marker of susceptibility to sudden cardiac death (44, 49). However, the relationship between β-AR signaling and alternans is not straightforward, and opposite effects have been observed, in which β-AR stimulation either favors (7, 39) or protects (11, 23, 24) against alternans and alternans-related arrhythmias. At the cellular level, whether β-AR stimulation facilitates or protects against alternans critically depends on whether a β-AR stimulation-mediated increase in SR load and fractional release (favoring alternans) dominates over β-AR-mediated effects on Ca²⁺ sequestration (abolishing alternans) or vice versa.

We have previously shown that in cat atrial myocytes, β-AR signaling is mediated by β₁- and β₂-AR subtypes. While β₁-AR-mediated effects on E-C coupling are mediated by G_s proteins coupled to adenylate cyclase (AC), which activates cAMP-dependent PKA, β₂-ARs couple to both G_s and G_i proteins (25, 60, 61). In previous studies (10, 51) of β₂-AR signaling pathways responsible for the regulation of Ca²⁺ and K⁺ currents in cat atrial myocytes, we established that β₂/G_s signaling acts via cAMP/PKA, similar to β₁/G_s signaling. In contrast, β₂/G_i signaling couples via phosphoinositide 3-kinase (PI3K)/Akt to endothelial nitric oxide (NO) synthase (eNOS) and leads to NO production. We further established that in atrial myocytes, NO has two important downstream effects: 1) NO-dependent activation of guanylate cyclase leads to an increase of cGMP and inhibition of cGMP-dependent phosphodiesterase (PDE) III, resulting in enhanced cAMP/PKA signaling (52), and 2) NO can inhibit the β₂/G_s/cAMP/PKA pathway via protein nitrosylation (10). Thus, the goal of the present study was to elucidate the interplay between these specific β-AR signaling pathways and their downstream targets, mitochondrial function, and cellular energy (ATP) sources for the regulation of pacing-induced Ca²⁺ alternans in atrial myocytes.

METHODS

Chemicals and solutions.

Chemicals were obtained from Sigma-Aldrich (St. Louis, MO) unless otherwise noted. ICI-118,551 was from AstraZeneca (Wilmington, DE), CGP-37157 was obtained from Tocris (Ellisville, MO), and KN-92, KN-93, spermine NONO-ate (SNO), and W-7 were from Calbiochem (San Diego, CA). During experiments, cells were continuously superfused with standard Tyrode solution containing (in mM) 140 NaCl, 5 KCl, 1 MgCl₂, 10 glucose, 10 HEPES, and 2 CaCl₂ (pH 7.3 adjusted with NaOH). All inhibitors, agonists, and antagonists were added to normal Tyrode solution from stock solutions.

β-AR stimulation protocols.

Four different β-AR stimulation protocols were used based on our previous studies (10, 51) on β-AR signaling in cat atrial myocytes. Nonselective β₁/β₂-AR stimulation (termed here β-AR stimulation) was achieved with isoproterenol (ISO; 0.1 μM). For specific stimulation of β₁-AR-dependent signaling (β₁-AR stimulation), cells were treated with ISO (0.1 μM) in combination with the β₂-AR subtype blocker ICI-118,551 (0.01 μM). Simultaneous activation of β₂/G_i and β₂/G_s signaling (β₂-AR stimulation) was achieved with the combination of ISO (0.1 μM) and the specific β₁-AR inhibitor atenolol (0.01 μM), whereas for selective β₂/G_s stimulation, fenoterol (41, 42) was used (0.1 μM). Concentrations were chosen based on our previous studies (10, 51) on β-AR signaling in cat atrial myocytes.

Myocyte isolation.

Cat atrial myocytes were isolated according to previously described methods (58). All protocols and procedures were approved by the Institutional Animal Care and Use Committee. Adult cats of either sex were anesthetized with pentobarbital sodium (50 mg/kg), and hearts were excised, mounted on a Langendorff apparatus, and retrogradely perfused with oxygenated collagenase-containing solution (37°C).

[Ca²⁺]_i measurements.

For [Ca²⁺]_i measurements, atrial myocytes were loaded with the membrane-permeable form of the fluorescent Ca²⁺ indicator indo-1 AM (5 μM, Invitrogen/Molecular Probes, Carlsbad, CA) in standard Tyrode solution for 15–20 min at room temperature followed by a >20-min deesterification interval in dye-free media. Indo-1-loaded cells were excited at 360 nm, and emission signals were collected simultaneously at 410-nm fluorescence (F₄₁₀) and 485-nm fluorescence (F₄₈₅) using photomultiplier tubes. Fluorescence signals were background subtracted, and [Ca²⁺]_i changes are expressed as changes in the ratio (R) of F₄₁₀ to F₄₈₅. Ca²⁺ transients and Ca²⁺ alternans were elicited by electrical field stimulation (voltage set at ∼50% above the threshold for contraction to ensure that every electrical stimulus was captured by the myocyte) using a pair of platinum electrodes. Pacing-induced stable Ca²⁺ alternans was elicited by increasing the stimulation frequency from 0.5 to 1 Hz followed by further increases in stimulation frequency in 0.1-Hz increments until stable alternans was observed. Experimental protocols were applied when individual cells revealed stable alternans for ∼1 min at a given stimulation frequency. The degree of Ca²⁺ alternans was quantified as the alternans ratio. The alternans ratio was defined as 1 − S/L, where S/L is the ratio of the small-amplitude (S) to large-amplitude (L) Ca²⁺ transient during a pair of alternating Ca²⁺ transients (28, 59). Thus, the alternans ratio had values between 1 and 0, with 0 indicating no alternans and 1 indicating the highest possible degree of alternans, with only every other stimulation resulting in a measurable Ca²⁺ transient.

Temperature.

All experiments were performed at room temperature (22–25°C). The pacing frequency threshold for the occurecnce of alternans decreases with lower temperatures (14, 23). By conducting experiments at room temperature, alternans could be studied in isolated myocytes with pacing protocols that were significantly less stressful, allowing for prolonged stimulation protocols and extended cell viability over time.

Statistical analysis.

Results are presented as means ± SE for the indicated number of cells (n). Statistical differences were determined using Student's t-tests for paired or unpaired data. Differences were considered statistically significant at P < 0.05.

RESULTS

Effect of β-adrenergic stimulation on pacing-induced Ca²⁺ alternans.

Stable Ca²⁺ alternans was induced by incrementally increasing the pacing frequency. Ca²⁺ alternans was observed at stimulation frequencies of >1 Hz. The average pacing frequency where stable maintained alternans was observed in cat atrial myocytes at room temperature was 1.7 Hz. The average alternans ratio was 0.43 ± 0.02 under control conditions.

In the heart, β-AR stimulation exerts positive inotropic effects and enhances the efficacy of E-C coupling through the regulation of intracellular Ca²⁺-handling proteins and Ca²⁺ signaling pathways. Stimulation of atrial myocytes revealing stable pacing-induce Ca²⁺ alternans with the nonselective β-AR agonist ISO (0.1 μM) abolished alternans (alternans ratio decreased from 0.43 ± 0.02 to 0.03 ± 0.01), as shown in Fig. 1A. Furthermore, ISO increased the Ca²⁺ transient amplitude of both the small and large alternans Ca²⁺ transient by 232% and 85%, respectively, indicative of the positive inotropic effect of β-AR stimulation.

Fig. 1.

Effects of β-adrenergic receptor (β-AR) stimulation on pacing-induced Ca²⁺ alternans. Pacing-induced Ca²⁺ alternans recordings are from individual atrial myocytes before and during β-AR stimulation. A: nonselective β-AR stimulation with isoproterenol (ISO; 0.1 μM). B: β₁-AR-specific stimulation using ISO (0.1 μM) combined with the β₂-AR antagonist ICI-118,551 (0.01 μM). C: β₂-AR-specific stimulation with ISO (0.1 μM) combined with the β₁-AR antagonist atenolol (0.01 μM). D: specific stimulation of the β₂/G_s pathway with fenoterol (0.1 μM). Bottom recordings show Ca²⁺ transients at expanded times (a, b, and c) from the top recordings. E: summary of selective and nonselective β-AR stimulation on alternans ratio. Numbers in parentheses indicate numbers of individual atrial myocytes tested. *P < 0.001 vs. control (CTL).

While cat atrial myocytes have both β₁- and β₂-subtypes of the AR (53), our previous study (51) revealed that the β₃-subtype is not involved in E-C coupling regulation in cat atrial tissue. Therefore, in the following experiments, we focused on β₁- and β₂-AR-mediated signaling pathways. As shown in Fig. 1B, selective stimulation of the β₁-AR pathway by the combined application of ISO (0.1 μM) and the β₂-AR blocker ICI-118,551 (0.01 μM) abolished Ca²⁺ alternans to the same degree as ISO alone. Similarly, selective stimulation of the β₂-AR pathway [with ISO in the presence of the β₁-AR blocker atenolol (0.01 μM)] or exclusively of the β₂/G_s pathway [with the selective β₂/G_s agonist fenoterol (0.1 μM)] abolished Ca²⁺ alternans, as shown in Fig. 1, C and D, respectively. Similar to nonselective β-AR stimulation, the selective activation of β₁- and β₂-ARs had a positive inotropic effect on the Ca²⁺ transient amplitude. Thus, the combined or selective activation of β-AR subtypes leads to an abolishment of pacing-induced Ca²⁺ alternans. With all four stimulation protocols, the ARs dropped to similarly low values between 0.03 and 0.04 (Fig. 1E), indicative of the complete absence of Ca²⁺ alternans.

β-AR effects on Ca²⁺ alternans involves PKA.

It is well established that β₁- and β₂-AR-mediated effects on E-C coupling are mediated by G_s proteins coupled to AC, which generates cAMP, which, in turn, activates PKA. PKA is involved in the phosphorylation processes of Ca²⁺-handling proteins [L-type voltage-gated Ca²⁺ channels (LTCC), SERCA, and the ryanodine receptor (RyR) Ca²⁺-release channel], which profoundly affects Ca²⁺ signaling during E-C coupling. We therefore tested the effect of PKA inhibition alone and in combination with β-AR stimulation on Ca²⁺ alternans. As shown in Fig. 2A, the PKA inhibitor H-89 (1 μM) enhanced Ca²⁺ alternans and increased the alternans ratio by 76%, to 0.72 ± 0.06. These data suggest that basal PKA activity exerts partial protection against alternans. Subsequent β-AR stimulation in the presence of the PKA blocker decreased Ca²⁺ alternans, but with different efficiency depending on which β-AR subtype was activated. β₁-AR stimulation completely abolished Ca²⁺ alternans (Fig. 2A, top), whereas selective activation of the β₂/G_s pathway, while showing a partial decrease of alternans, failed to lower the alternans ratio below control levels (Fig. 2A, middle). The summary data of changes of the alternans ratio are shown in Fig. 2A, bottom.

Fig. 2.

β-AR effects on Ca²⁺ alternans mediated by PKA. A, top and middle: recordings of Ca²⁺ alternans before and during inhibition of PKA with H-89 (1 μM) followed by β₁-AR (top) or β₂/G_s stimulation (middle). Bottom, average alternans ratios in the presence of H-89 and during subsequent β₁-AR and β₂/G_s stimulation. B, top: recordings of Ca²⁺ alternans where cells were first exposed to β-AR stimulation followed by treatment with H-89 (10 μM). Bottom, average alternans ratios. During β₂-AR stimulation, H-89 caused the maximal degree of alternans, with the alternans ratio approaching 1. Black circles mark the timing of electrical stimulation when the small-amplitude Ca²⁺ transient was strongly reduced in amplitude. Numbers in parentheses indicate numbers of individual atrial myocytes tested. [Ca²⁺]_i, intracellular Ca²⁺ concentration; R, ratio of fluorescence at 410 nm to fluorescence at 485 nm.

The effect of PKA inhibition was tested with a different protocol, as shown in Fig. 2B. Here, β-AR stimulation preceded the application of the PKA inhibitor. As expected, β₁-AR, β₂-AR, and selective β₂/G_s stimulation completely abolished Ca²⁺ alternans. Subsequent application of H-89 (10 μM) caused the maximal degree of alternans during β₂-AR stimulation, with an alternans ratio approaching 1. In contrast, during β₁-AR stimulation, PKA inhibition also enhanced Ca²⁺ alternans beyond control levels (alternans ratio: 0.72 ± 0.15); however, the effect was less pronounced than during β₂-AR stimulation. Taken together, these data indicate that β₂-AR stimulation abolished alternans through a PKA-dependent pathway, whereas the decrease of alternans mediated by β₁-AR signaling appeared only partially be dependent on PKA. CaMKII is a second crucial intracellular kinase that regulates Ca²⁺-handling proteins through phosphorylation, and it has been shown that in the heart, β₁-AR stimulation activates a dual signaling pathway mediated by cAMP/PKA and CaMKII (50). Therefore, in the following set of experiments, we tested the involvement of CaMKII in β-AR-mediated regulation of Ca²⁺ alternans.

Role of calmodulin and CaMKII in the regulation of Ca²⁺ alternans.

Inhibition of the CaMKII signaling pathway on pacing-induced Ca alternans is shown in Fig. 3. Exposure to the calmodulin antagonist W-7 (25 μM) led to a profound enhancement of Ca²⁺ alternans, again suggesting that basal kinase activity protects against the development of a high degree of alternans. In the example shown in Fig. 3A, top, W-7 almost completely abolished the small-amplitude alternans Ca²⁺ transient, and, on average, the alternans ratio increased to 0.90 ± 0.05. Similarly, the CaMKII antagonist KN-93 (1 μM) enhanced Ca²⁺ alternans and increased the alternans ratio to 0.77 ± 0.06 (Fig. 3A, bottom), whereas KN-92, the inactive analog of KN-93, serving as a negative control, had only a marginal effect on the alternans ratio. Subsequent β₁-AR stimulation decreased Ca²⁺ alternans, but in the presence of W-7 and KN-93, the alternans ratio remained above control levels (0.69 ± 0.18 and 0.67 ± 0.10, respectively). In contrast, in the presence of the inactive analog KN-92, β₁-AR stimulation was capable of completely abolishing Ca²⁺ alternans.

Fig. 3.

β-AR effects on Ca²⁺ alternans mediated by Ca²⁺/calmodulin-dependent protein kinase II (CaMKII). A, top: recordings of Ca²⁺ alternans before and during inhibition of calmodulin with W-7 (25 μM) followed by β₁-AR stimulation. Bottom, average alternans ratios in the presence of W-7, the CaMKII inhibitor KN-93 (1 μM), or the inactive analog KN-92 (1 μM) followed by β₁-AR stimulation. B, top: decrease of Ca²⁺ alternans by β₁-AR stimulation was reversed and alternans was enhanced by subsequent treatment with KN-93. Bottom, average alternans ratios during β₁-AR stimulation followed by treatment with W-7, KN-93, or KN-92. KN-92, the inactive analog of KN-93, served as a negative control. Black circles mark the timing of electrical stimulation when the small-amplitude Ca²⁺ transient was strongly reduced in amplitude. Numbers in parentheses indicate numbers of individual atrial myocytes tested.

Figure 3B shows the reversed protocol. β₁-AR stimulation abolished pacing-induced Ca²⁺ alternans, as expected; however, the subsequent application of W-7 and KN-93 increased the alternans ratio above control levels, to 0.78 ± 0.11 and 0.76 ± 0.13, respectively. The inactive analog KN-92 had no influence on the abolishing effect of β₁-AR stimulation (Fig. 3B, bottom). These data indicate that the effect of β₁-AR stimulation to abolish alternans largely depends on CaMKII activity and, to a lesser extent, on the cAMP/PKA pathway, whereas the β₂-AR-dependent effects on Ca²⁺ alternans primarily involve PKA. This was further confirmed by the observation that in the presence of KN-93, β₂-AR stimulation still lowered the alternans ratio below control levels (data not shown).

Combined inhibition of PKA and CaMKII is required to abolish β₁-AR effects on Ca²⁺ alternans.

After pacing-induced Ca²⁺ alternans was enhanced by the inhibition of PKA (alternans ratio in the presence of H-89: 0.72 ± 0.06; Fig. 4), subsequent β₁-AR stimulation reversed the effect of H-89 and decreased the alternans ratio below control levels (alternans ratio: 0.13 ± 0.09). Subsequent exposure to W-7 reversed the effect of β₁-AR stimulation. In the example shown in Fig. 4, top, combined inhibition of PKA and CaMKII almost completely abolished the small-amplitude Ca²⁺ transient despite β₁-AR stimulation. On average, the alternans ratio rose to 0.72 ± 0.12. In conclusion, the results indicate that β₁-AR stimulation protects against alternans by invoking two complementary signaling pathways, one involving cAMP/PKA and the other depending on CaMKII activity.

Fig. 4.

β₁-AR effects on Ca²⁺ alternans are mediated by both PKA and CaMKII. A: Ca²⁺ transient recordings revealing Ca²⁺ alternans from an atrial myocyte subsequently treated with the PKA inhibitor H-89 followed by β₁-AR stimulation and the subsequent addition of W-7. B: average alternans ratios. The data indicate that during combined inhibition of PKA and CaMKII, β₁-AR failed to decrease Ca²⁺ alternans. Black circles mark the electrical stimulation when the small-amplitude Ca²⁺ transient was strongly reduced in amplitude. Numbers in parentheses indicate numbers of individual atrial myocytes tested.

Role of downstream targets of β₂-AR signaling in the protection against alternans.

β₂-ARs couple to both G_s and G_i proteins (25, 60, 61). As detailed in the Introduction, in cat atrial myocytes, the β₂/G_i signaling pathway couples via PI3K/Akt to eNOS and leads to NO production. The downstream effects of NO are twofold and result in enhanced cAMP/PKA signaling via inhibition of cGMP-dependent PDE III but also inhibition of the β₂/G_s/AC/cAMP/PKA pathway (1, 10). Since NO plays a key role in the β₂/G_i signaling pathways, we applied a blocker of eNOS [N^ω-nitro-l-arginine methyl ester (l-NAME)] or supplied NO in excess with an exogenous NO donor (SNO). As shown in Fig. 5A, neither l-NAME nor SNO had a profound effect on the alternans ratio of pacing-induced Ca²⁺ alternans. Selective stimulation of the β₂/G_s pathway in the presence of l-NAME or SNO resulted in a complete abolishment of Ca²⁺ alternans, suggesting that the β₂/G_s pathway is sufficient for complete protection against alternans during β₂-AR stimulation. Concomitant stimulation of the β₂/G_i pathway had no additional effect, indicating that the β₂/G_i pathway (and thus NO) does not play a significant role in the regulation of alternans. This was further confirmed by the observation that in the presence of pertussis toxin to disrupt G_i regulation (61), β₂/G_s stimulation fully abolished alternans (data not shown) and that even in the presence of excessive amounts of NO, β₂-AR stimulation led to an abrogation of alternans (Fig. 5A). In addition, inhibition of PDE III by milrinone (Fig. 5B) completely abolished Ca²⁺ alternans and had a strong positive inotropic effect on the Ca²⁺ transient amplitude, presumably by increasing cytosolic cAMP levels. β₂-AR stimulation in the presence of milrinone had no additional effect on the alternans ratio since the alternans ratio was already nearly 0 (data not shown). In conclusion, these data indicate that β₂-AR-mediated abolishment of Ca²⁺ alternans is mediated by the β₂/G_s/AC/cAMP/PKA pathway.

Fig. 5.

Effects of nitric oxide (NO) and phosphodiesterase (PDE) on Ca²⁺ alternans. A: average alternans ratios under control conditions and in the presence of the endothelial NO synthase inhibitor N^ω-nitro-l-arginine methyl ester (l-NAME; 50 μM), the exogenous NO donor spermine NONOate (SNO; 100 μM), and the PDE III inhibitor milrinone (5 μM) (solid bars). Shaded and open bars show the effects of l-NAME and SNO on alternans ratios during specific β₂/G_s and nonselective β₂-AR stimulation, respectively. B: recordings of pacing-induced Ca²⁺ transients followed by treatment with milrinone. Numbers in parentheses indicate numbers of individual atrial myocytes tested.

Under β-AR stimulation, normal contraction of myocytes and intracellular Ca²⁺ cycling are critically dependent on the balance between myocardial energy demand and energy supply from mitochondrial and glycolytic sources. Thus, in the next experiments, we explored the role of β-AR signaling in Ca²⁺ alternans during metabolic inhibition.

Alternans depends on the interaction between mitochondrial function and β-AR signaling.

We have previously shown that inhibition of various mitochondrial functions (ranging from inhibition of the electron transport chain, F₁/F₀-ATP synthase, and mitochondrial dehydrogenases to mitochondrial Ca²⁺ transport) enhances the propensity and degree of Ca²⁺ alternans, suggesting that mitochondrial function, specifically ATP production, plays a critical role in the generation of alternans (16).

Consistent with our previous study, as shown in Fig. 6, inhibition of mitochondrial ATP production by block of mitochondrial F₁/F₀-ATP synthase with oligomycin (A), dissipation of the mitochondrial membrane potential with FCCP (B), inhibition of mitochondrial Ca²⁺ uptake via the mitochondrial Ca²⁺ uniporter with ruthenium red (C), or Ca²⁺ extrusion via the mitochondrial Na⁺/ Ca²⁺ exchanger with CGP-37157 (D) all enhanced Ca²⁺ alternans and increased the alternans ratio to values between 0.53 and 0.80 (E). β-AR stimulation in the presence of oligomycin, FCCP, or ruthenium red lowered the alternans ratio and decreased and even abolished alternans. There were no significant differences between β₁- and β₂-AR-mediated signaling. In contrast, β-AR stimulation after the inhibition of Ca²⁺ extrusion with CGP-37157 failed to significantly decrease alternans. We (16) have previously shown that treatment with CGP-37157 led to cyclosporine-sensitive opening of the mitochondrial permeability transition pore (MPTP), which is linked to irreversible impairment of mitochondrial function. Presumably due to the irreversible nature of MPTP opening, β-AR stimulation failed to protect against Ca²⁺ alternans. In contrast, these results indicate that potentially reversible inhibition of mitochondrial functions can be counteracted through β-AR signaling.

Fig. 6.

Effect of β-AR stimulation during inhibition of mitochondrial function. Examples of effects on Ca²⁺ alternans of inhibition of various mitochondrial functions and processes followed by the activation of β-AR signaling pathways are shown. A: inhibition of mitochondrial F₁/F₀-ATP synthase with oligomycin (Oligo; 1 μg/ml) followed by β₁-AR stimulation. B: dissipation of mitochondrial membrane potential with the protonophore FCCP (0.1 μM) followed by β₂/G_s stimulation. C: inhibition of mitochondrial Ca²⁺ uptake via the Ca²⁺ uniporter with ruthenium red (RuR; 10 μM) followed by nonselective β₂-AR stimulation. D: inhibition of mitochondrial Ca²⁺ extrusion via the mitochondrial Na⁺/Ca²⁺ exchanger with CGP-37157 (CGP; 2.5 μM) followed by β₂/G_s stimulation. E: summary data of average alternans ratios in the presence of mitochondrial inhibitors followed by β-AR stimulation. Numbers in parentheses indicate numbers of individual atrial myocytes tested.

Depletion of cellular ATP sources diminishes the ability of β-AR stimulation to protect against alternans.

A common result of inhibition of diverse mitochondrial functions is a decrease in mitochondrial ATP production and exhaustion of cellular energy reserves. The diminished ATP supply potentially affects ATP-dependent Ca²⁺ transport processes, such as cytosolic Ca²⁺ removal. Decreased cytosolic Ca²⁺ sequestration has been identified as a key factor causing cardiac alternans. We have previously shown that selective inhibition of glycolytic ATP production (5, 23, 28, 29) or ATP synthesis by mitochondrial oxidative phosphorylation (16) (Fig. 6) enhanced alternans. In the following set of experiments, we addressed the question of whether β-AR stimulation is capable of decreasing Ca²⁺ alternans during selective inhibition of glycolytic or mitochondrial ATP production or when both sources of ATP are eliminated.

In this set of experiments, mitochondrial ATP production was blocked selectively by oligomycin (inhibition of mitochondrial F₁/F₀-ATP synthase). Glycolysis was selectively inhibited by pyruvate, as previously described (23, 28, 29). Pyruvate is effectively metabolized by mitochondria, which leads to an increase in citrate and ATP levels (38), which, in turn, inhibits the fructokinase reaction and reduces flux through glycolysis. Inhibition of both ATP sources was achieved by either combined exposure to oligomycin + pyruvate or exposure to 2-deoxyglucose (2-DG). 2-DG is phosphorylated by hexokinase to 2-DG-6-phosphate, which cannot be further metabolized by glycolytic enzymes. Furthermore, 2-DG acts as a sink for ATP because the hexokinase reaction consumes ATP. In the absence of any other metabolic fuel (as under our experimental conditions), 2-DG effectively blocks glycolytic and mitochondrial ATP production because inhibition of glycolysis eliminates the production of substrates for mitochondrial ATP generation. In the presence of oligomycin, 2-DG, pyruvate (Fig. 7A), and oligomycin + pyruvate (Fig. 7B), Ca²⁺ alternans was enhanced, raising average alternans ratios to 0.70–0.80 (Fig. 7C). β-AR stimulation (β₁ and β₂) was effective in completely abolishing Ca²⁺ alternans during inhibition of mitochondrial ATP production with oligomycin. Similarly, β-AR stimulation was effective in abolishing Ca²⁺ alternans that were previously enhanced by inhibition of glycolytic ATP formation. In contrast, when both sources of ATP were abolished, β-AR stimulation failed to decrease alternans (Fig. 7B). Furthermore, diastolic [Ca²⁺]_i substantially increased and Ca²⁺ transients were prolonged, consistent with compromised cytosolic Ca²⁺ removal. As shown in Fig. 7C, separate or combined inhibition of glycolytic and/or mitochondrial ATP production consistently enhanced Ca²⁺ alternans. When only one of the two cellular ATP sources was inhibited, β-AR stimulation could overcome the inhibitory effect and was able to decrease alternans. However, when both ATP sources were abolished, β-AR stimulation failed to protect against alternans.

Fig. 7.

Effects of β-AR stimulation on Ca²⁺ alternans during inhibition of glycolytic and mitochondrial ATP production. A: effects of inhibition of glycolysis with pyruvate (Pyr; 10 mM) followed by nonselective β-AR stimulation with ISO (0.1 μM) on pacing-induced Ca²⁺ alternans. B: effects of combined inhibition of glycolytic (Pyr) and mitochondrial (Oligo; 1 μg/ml) ATP production followed by nonselective β-AR stimulation with ISO. C: summary data of average alternans ratios during glycolytic and/or mitochondrial inhibition followed by β-AR stimulation. β-AR stimulation could overcome the effect of separate inhibition of glycolytic or mitochondrial ATP generation; however, during inhibition of both ATP sources, β-AR stimulation failed to decrease Ca²⁺ alternans. Numbers in parentheses indicate numbers of individual atrial myocytes tested. 2-DG, 2-deoxyglucose.

In conclusion, our results indicate that β-AR stimulation protects against proarrhythmic Ca²⁺ alternans and involves parallel and complementary pathways regulated by both β₁- and β₂-AR subtypes. β-AR stimulation-dependent decrease of alternans involves both PKA and CaMKII and depends on a sufficient ATP supply.

DISCUSSION

In this study, we investigated the interplay between β-AR signaling, mitochondrial function, and the occurrence of Ca²⁺ alternans in atrial tissue. The key findings of our investigation were as follows.

First, pacing-induced Ca²⁺ alternans could be decreased by several distinct pathways linked via G proteins to β₁- and β₂-AR subtypes. The effect of β-AR stimulation involved activation of the kinases PKA and CaMKII. β₁-AR stimulation was linked via G_s to PKA and CaMKII activation. β₂-AR stimulation only activated PKA via G_s, similarly to β₁-AR stimulation; however, the previously identified β₂/G_i/eNOS/NO signaling patway in atrial myocytes had no effect on alternans. Furthermore, basal PKA and CaMKII activity exerted partial protection against alternans.

Second, pacing-induced Ca²⁺ alternans was enhanced by disruption of mitochondrial function; however, β-AR stimulation could still decrease Ca²⁺ alternans in the presence of compromised mitochondria, except when mitochondrial impairment included MPTP opening.

Finally, the abolishing effect of β-AR stimulation on Ca²⁺ alternans depended on sufficient cellular ATP reserves and energy supplies. Selective inhibition of glycolysis or mitochondrial ATP production via oxidative phosphorylation enhanced Ca²⁺ alternans; however, alternans could still be decreased by β-AR stimulation, suggesting that one ATP source could compensate for the other. Severe depletion of cellular ATP reserves by combined inhibition of glycolytic and mitochondrial ATP production prevented β-AR stimulation from abrogating Ca²⁺ alternans, suggesting that the protective effect of β-AR signaling ultimately depends on a sufficient ATP supply.

The three β-AR signaling cascades investigated here have been previously established in atrial myocytes (10, 51). In these studies, the importance of these signaling pathways for E-C coupling was demonstrated for the regulation of ion currents (L-type Ca²⁺ current and ATP-dependent K⁺ current). The present study extended these earlier observations by investigating how these pathways regulate or modulate Ca²⁺ alternans in atrial myocytes.

An important finding of our study was that the abolishment of Ca²⁺ alternans by β-AR stimulation involved the activation of cellular kinases. Whereas β₁-AR effects depended on both PKA and CaMK activity, β₂-AR-mediated effects appeared to involve solely PKA. Both kinases have well-established phosphorylation targets, including Ca²⁺-handling proteins that play a key role in E-C coupling and mediate the inotropic effects of β-AR stimulation. Among these proteins, the activity of LTCCs, RyRs, and SERCA are all enhanced by β-AR stimulation-dependent phosphorylation. While β-AR stimulation enhances LTCC currents, Ca²⁺ entry, and SR Ca²⁺ load and release, it does not seem to have a direct effect on LTCC as far as alternans are concerned, because the L-type Ca²⁺ current typically does not alternate during alternans in atrial myocytes (23, 45; but see Refs. 17 and 32). Thus, the abolishing effect appears to be downstream of LTCC augmentation. A prime candidate for a downstream target is beat-to-beat Ca²⁺ sequestration. It has been convincingly shown that cellular processes, transport pathways, and organelles that clear cytosolic Ca²⁺ protect against alternans and that conditions that impair cytosolic Ca²⁺ sequestration promote alternans (for reviews and references, see Refs. 54 and 56). β-AR stimulation strongly enhances SERCA activity, which is accomplished via phosphorylation of the SERCA inhibitory protein phospholamban (30, 33). Phospholamban phosphorylation, principally by PKA (although there is evidence for a role of CaMKII), reduces its inhibitory interactions with SERCA and leads to a dramatic increase in SERCA activity.

Along the same line of arguments, intact mitochondrial function plays an important role in the protection against proarrhythmic alternans. Mitochondria participate in Ca²⁺ signaling during E-C coupling in two main ways: as a potential Ca²⁺ sink that can take up and store significant amounts of Ca²⁺ and as the quantitatively most important source of ATP production that fuels phosphorylation processes or serves as a regulatory cofactor of ion channels and transporters. While it is still a matter of debate (for a discussion of this controversy, see Refs. 9 and 40) as to whether mitochondrial Ca²⁺ uptake and extrusion make a significant contribution to the beat-to-beat regulation of [Ca²⁺]_i (and therefore may also make a direct contribution to Ca²⁺ alternans), there is no doubt that impairment of mitochondrial function at various levels generates conditions that favor alternans. Dissipation of mitochondrial membrane potential and inhibition of mitochondrial function at the level of the electron transport chain, F₁/F₀-ATP synthase, or Ca²⁺ uptake and extrusion all enhance Ca²⁺ alternans (for a review, see Ref. 16; see also Figs. 6 and 7 in the present study). In addition, altering mitochondrial function changes the cellular redox environment (8), and redox modification of the RyR can cause alternans (3). These results, together with the observation that inhibition of glycolysis also induced electromechanical and Ca²⁺ alternans (see Refs. 23, 28, and 29; see also Figs. 6 and 7 in the present study), suggest that a sufficient ATP supply is crucial for the prevention of alternans and emphasizes a key role of kinases and phosphorylation processes in this context. Our results indicate that selective elimination of one of the two ATP sources per se enhanced Ca²⁺ alternans but was not sufficient to prevent the protective effect of β-AR signaling against Ca²⁺ alternans. This observation suggests that if at least one ATP source is functional, the amounts of ATP produced are sufficient to sustain the β-AR effects on Ca²⁺ signaling. Even though the bulk cellular ATP pool is depleted dramatically during inhibition of mitochondrial ATP production (since mitochondria generate up to 90% of cellular ATP in cardiac myocytes, which is primarily consumed to sustain contractile activity), glycolysis-derived ATP appears to suffice to support β-AR effects on Ca²⁺ alternans. This may in part be explained by the observation that ATP derived from glycolysis is used preferentially by membrane ion transport mechanisms. The close physical association of glycolytic enzymes with ATP-dependent ion transporters and channels form functional microcompartments where source and utilization sites for ATP are intimately linked. Such arrangements have been demonstrated for Na⁺-K⁺-ATPase (18), ATP-dependent K⁺ channels (55), and the SERCA pump (62) and have been proposed for the RyR Ca²⁺-release channel as a potential contributing factor to alternans (5, 23, 28, 29).

Modulation of RyR function through β-AR stimulation represents an additional mechanism through which β-AR stimulation protects against alternans. RyRs are subject to phosphorylation by PKA and CaMKII (2, 19, 22, 34, 35). The signaling events that occur downstream of β-AR stimulation and their effects on RyR activity remain highly contentious (for reviews, see Refs. 21 and 63), and the ramifications of RyR phosphorylation for cardiac function in health and disease have remained a highly debated and controversial issue (for a recent update on the debate, see Ref. 4). Nonetheless, we (45) have previously shown that in atrial myocytes, refractoriness of SR Ca²⁺ release via RyRs and beat-to-beat alternations in the kinetics of RyR and SR Ca²⁺-release restitution constitute a key defining factor of cardiac alternans. Factors that sensitize the RyR to Ca²⁺-induced Ca²⁺ release can accelerate restitution and therefore can overcome Ca²⁺ alternans. This can be demonstrated with the RyR-sensitizing agent caffeine at low concentrations, and the effects were strikingly similar to the effects of ISO stimulation (45). Thus, enhanced Ca²⁺ sequestration together with β-AR-mediated effects on the RyR appear to act in tandem to protect against pacing-induced Ca²⁺ alternans.

The abolishment of pacing-induced Ca²⁺ alternans by β-AR stimulation is controversial and not unique to atrial tissue. As discussed in the Introduction, β-AR stimulation has also been shown to facilitate alternans. At the cellular level, it is believed that this can occur when a strong effect of β-AR stimulation on SR Ca²⁺ load and fractional SR Ca²⁺ release outweighs the protective effect of cytosolic Ca²⁺ sequestration (54). Furthermore, we (23) have previously shown that in feline ventricular myocytes, ISO stimulation reversibly abolished electrical (action potential duration) and mechanical (cell shortening) alternans, where the alternations in action potential duration subsided in concert with the changes in contration, indicating that electrical and mechanical (or Ca²⁺) activity are tightly coupled during alternans.

In summary, we demonstrated here that β-AR stimulation abolishes pacing-induced Ca²⁺ alternans even under conditions where inhibition of mitochondrial function facilitates Ca²⁺ alternans. The protective effect of β-AR stimulation is mediated by parallel and complementary signaling cascades. The redundancy in the β-AR signaling-mediated protection against proarrhythmic alternans represents an inherent safety mechanism to prevent arrhythmogenic Ca²⁺ release in the heart during enhanced sympathetic tone.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants HL-62231, HL-80101, and HL-101235 and the Leducq Foundation (to L. A. Blatter) as well as the American Heart Association-Midwest Affiliate (to S. M. Florea).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: S.M.F. and L.A.B. conception and design of research; S.M.F. performed experiments; S.M.F. and L.A.B. analyzed data; S.M.F. and L.A.B. interpreted results of experiments; S.M.F. and L.A.B. prepared figures; S.M.F. and L.A.B. drafted manuscript; S.M.F. and L.A.B. edited and revised manuscript; S.M.F. and L.A.B. approved final version of manuscript.