Pharmacological evidence for Orai channel activation as a source of cardiac abnormal automaticity

Abstract

Calcium transport through plasma membrane voltage-independent calcium channels is vital for signaling events in non-excitable and excitable cells. Following up on our earlier work, we tested the hypothesis that this type of calcium transport can disrupt myocardial electromechanical stability. Our Western and immunofluorescence analyses show that left atrial and ventricular myocytes express the Orai1 and the Orai3 calcium channels. Adding the Orai activator 2-aminoethoxydiphenyl borate (2-APB) to the superfusate of rat left atria causes these non- automatic muscles to contract spontaneously and persistently at rates of up to 10 Hz, and to produce normal action potentials from normal resting potentials, all in the absence of external stimulation. 2-APB likewise induces such automatic activity in superfused rat left ventricular papillary muscles, and the EC50s at which 2-APB induces this activity in both muscles are similar to the concentrations which activate Orais. Importantly, the voltage-independent calcium channel inhibitor 1-[2-(4-methoxyphenyl)-2-[3-(4-methoxyphenyl) propoxy] ethyl-1H-imidazole (SKF-96365) suppresses this automaticity with an IC50 of 11 ± 0.6 μM in left atria and 6 ± 1.6 μM in papillary muscles. 1-(5-Iodonaphthalene-1-sulfonyl)-hexahydro-1,4-diazepine (ML-7), a second voltage-independent calcium channel inhibitor, and two calmodulin inhibitors also prevent 2-APB automaticity while two calmodulin-dependent protein kinase II inhibitors do not. Thus an activator of the Orai calcium channels provokes a novel type of high frequency automaticity in non-automatic heart muscle.

1. Introduction

In the normal heart, the sinoatrial node spontaneously generates the electrical stimuli that transit through the conduction system to initiate myocyte depolarization and contraction (Adamson et al., 2005). By contrast to this typical setting, arrhythmogenic electrical activity can arise in ectopic sites like atrial and ventricular muscle, the cells of the conduction system, and the muscular sleeves of cardiac veins (Haissaguerre et al., 2002; Jais et al., 1997; Waldo and Wit, 1993). Three types of voltage-dependent phenomena are hypothe- sized to produce ectopic activity: (i) The abnormal propagation of electrical impulses through cardiac muscle which re-excites myocytes to produce impulse reentry (Panfilov and Pertsov, 2001), (ii) calcium leak from sarcoplasmic reticulum stores that activates transient inward currents which trigger abnormal impulses (Lehnart et al., 2006; Shannon et al., 2003), and (iii) myocyte partial depolarization in settings like ischemia which elicits the spontaneous action potentials that define abnormal automaticity (Vassalle, 1971). However, drugs that alter either the myocyte action potential or impulse conduction are poor anti-arrhythmics (Adamson et al., 2005; The CAST Investigators, 1989). In addition, arrhythmia occurs in normal heart under non-ischemic conditions that are unlikely to cause calcium leak or partial depolarization (Haissaguerre et al., 2002). Furthermore, myocytes isolated from failing hearts can spontaneously depolarize from normal resting potentials without major changes in calcium homeostasis (Nuss et al., 1999). Together these latter reports suggest that novel mechanisms may initiate arrhythmogenic ectopy.

Two families of plasma membrane voltage-independent calcium transporters sustain the calcium homeostasis essential for cell signaling; (i) the Orai channels that form the store-operated calcium channel (SOCC) and the arachidonate-regulated calcium channel (ARC), and (ii) the transient receptor potential protein (Trp) channels. In the case of the SOCC, depletion of endoplasmic reticulum (ER) calcium stores induces ER stromal interaction molecule-1 (Stim1) to translocate to ER-plasma membrane junctions. There Stim1 recruits plasma membrane proteins including Orai1 to form the SOCC which facilitates the calcium entry required to refill cell stores (Liou et al., 2005; Parekh and Putney, 2005). For ARC activity, receptor signaling generates free arachidonic acid which activates pentamers of Orai1 and Orai3 to provoke cell calcium entry independently of ER Stim1 (Shuttleworth et al., 2004). Finally, multiple effectors activate calcium entry into cells through the Trp channels, and the SOCC itself may contain Trps (Birnbaumer, 2009). Both families of voltage-independent calcium transporters contribute to muscle and sinoatrial node function (Hunton et al., 2004; Ju et al., 2007; Lyfenko and Dirksen, 2008; Stiber et al., 2008). However, no data suggest that calcium transported through either family of proteins influences the fundamental properties of non-automatic heart muscle including its electromechanical stability.

Several groups have reported that 2-aminoethoxydiphenyl borate (2-APB) activates Orai1 and Orai3 with EC50s of 20 μM and 14 μM, respectively, to induce calcium entry into calcium-replete non- excitable cells (DeHaven et al., 2008; Lis et al., 2007; Peinelt et al., 2008; Zhang et al., 2008). Previously we showed that 2-APB causes sporadic or tachycardic ectopy in superfused rat left atria but had not identified a potential mechanism for this unusual activity (Wolkowicz et al., 2007a,b). Here we test whether this ectopy has characteristics of an Orai-dependent process and report data suggesting that the activation of these voltage-independent calcium channels may ignite high frequency automaticity in non-automatic cardiac muscle.

2. Materials and methods

2.1. Materials

2-APB, 1,4-dihydro-2,6-dimethyl-5-nitro-4-[2-(trifluoromethyl) phenyl]-3-pyridinecarboxylic acid methyl ester (Bay K 8644), and 1-[2-(4-methoxyphenyl)-2-[3-(4-methoxyphenyl) propoxy]ethyl- 1H-imidazole (SKF-96365) were from Tocris-Cookson (Ellisville MO, USA). 1-(2-Hydroxyethyl)-4-(3-(trifluoromethyl-10-phenothiazinyl) propyl)-piperazine (fluphenazine-N-2-choloroethane), N-[2- [[[3-(4-chlorophenyl)-2propenyl]methylamino]methyl]phenyl]- N-(2-hydroxyethyl)-4-methoxybenzenesulphonamide (KN-93), 4-[(2S)-2-[(5-isoquinolinylsulfonyl)methylamino]-3-oxo-3-(4- phenyl-1-piperazinyl)propyl] phenyl isoquinolinesulfonic acid ester (KN-62), 1-(5-iodonaphthalene-1-sulfonyl)-hexahydro-1,4- diazepine (ML-7), and (RS)-N-benzyl-N-(2-chloroethyl)-1-phenoxy- propan-2-amine (phenoxybenzamine) were from Calbiochem (Gibbstown NJ, USA). Enhanced chemiluminescence reagent was from GE Healthsciences (Piscataway NJ, USA). All other reagents were standard laboratory grade.Bay K 8644 and 2-APB were prepared as 100 and 150 mM DMSO stock solutions, respectively; SKF-96365 was suspended as a 100 mM aqueous stock solution.

2.2. Left atrial and left ventricular papillary muscle preparations

These investigations conform to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). All rats were anesthetized with isoflurane prior to the harvesting of their hearts. Rat left atrial appendages and intact right atria were isolated, superfused in Krebs-Henseliet (KH) buffer, and their mechanical function was assessed as described (Wolkowicz et al., 2007b). All left atria were pre-treated for 10 min with 300 nM (−)Bay K 8644 except those used in calmodulin inhibitor, calmodulin-dependent protein kinase II (CaMKII) inhibitor, and ML-7 studies. To acquire resting and action potentials, superfused left atria were impaled with glass microelectrodes containing 3 M KCl (Huang et al., 2004).

Anterior left ventricular papillary muscles were dissected and mounted in muscle baths that contained oxygenated KH (Topcuoğlu et al., 1999) maintained at 30 °C. Muscle resting tension was set to 0.8 g and all muscles were paced for 30 min at 1 Hz and 130% of capture voltage. Papillary muscle length then was adjusted to produce a maximum force of contraction and resting tension was set to 80% of this value. Typical muscles (n= 7) were 3.8 ± 0.25 mm long, 1.0 ± 0.10 mm wide, and weighed 4.3 ± 0.98 mg-wet-weight. They gener- ated 14.6 ± 1.3 mN of force with a time-to-peak tension of 119 ± 2 ms, 50% relaxation times of 89 ± 1 ms, and 90% relaxation times of 157 ± 7 ms. All stabilized papillary muscles were pre-treated by adding 300 nM (−)Bay K 8644 to the superfusate 10 min before the start of any experiment. Bay K 8644 was maintained in the superfusate thereafter and it increased muscle force of contraction to 149 ± 7% of control.

2.3. The concentration-dependence of 2-APB-induced ectopy

Two sets of experiments assessed how 2-APB affects calcium- loaded cardiac muscle. First, left atria (n= 7) were superfused in KH at 30 °C, paced at 0.1 Hz, and eight increasing concentrations of 2-APB from 0 to 30 μM were added to the superfusate. After three to 10 min incubation at each concentration, pacing was stopped and we measured the rate at which these left atria spontaneously contracted. In separate experiments, 7.5 or 20 μM 2-APB was added to the superfusate of two groups of 0.1 Hz-paced left atria (n= 3 per) and their mechanical function was recorded for 15 min with or without pacing. Finally, we recorded the electrical activity of untreated, 1 Hz- paced (n= 3) and 2-APB-treated, spontaneously contracting left atria (n= 3).

The second set of experiments undertook similar analyses in papillary muscles. Seven concentrations of 2-APB from 0 to 28 μM were added to the superfusate of 1 Hz-paced papillary muscles (n= 7). After three to 10 min incubation at each concentration, pacing was stopped and we measured the rate at which papillary muscles spontaneously contracted. Two additional groups of muscles (n= 5 per) were superfused in KH at 30 °C, paced at 1 Hz, and 0 or 20 μM 2-APB was added to the superfusate. Five minutes later we stopped pacing and measured the rate at which these muscles spontaneously contracted.

2.4. The temperature dependence of 2-APB-induced left atrial ectopy

To estimate a physiological rate for 2-APB-induced ectopy, we superfused 0.1 Hz-paced left atria (n= 9) in KH at 30 °C, and added 20 μM 2-APB to the superfusate. Once these muscles spontaneously contracted, pacing was stopped and we measured the rate of this spontaneous activity. Additional 0.1 Hz-paced left atria (n= 9) then were superfused at 23 °C, 20 μM 2-APB was added to the superfusate, and the steady-state rate of spontaneous contraction was measured in the absence of pacing. Muscle bath temperature was rapidly increased to 37 °C and left atrial spontaneous contraction rate was measured again in the absence of pacing. For comparative purposes, the contraction rate of a group of untreated, unpaced intact rat right atria (n = 7) was measured at 30 °C as was the spontaneous contraction rate of a second group of untreated, unpaced right atria (n= 7) measured at 23 and 37 °C.

2.5. Assessment of 2-APB-induced ectopy as a triggered or an automatic event

Three groups of left atria (n= 5 per) and three groups of papillary muscles (n= 5 per) were superfused in KH at 30 °C, and paced at 1 Hz. Pacing was stopped and one group of each type of muscle remained untreated. 25 μM 2-APB was added to the superfusate of the second group of muscles 30 s after the last stimulus. The third group of left atria or papillary muscles were treated with 25 μM 2-APB 30 s after the last stimulus, and then with 40 μM SKF-96365 30 s later. We assessed whether any of these unpaced muscles spontaneously contracted during the next 10 min of superfusion.
Fig. 1. 2-APB induces spontaneous cardiac mechanical activity in a hyperbolic manner: (A) Rates of spontaneous mechanical activity recorded from left atria (■; n= 7) and left ventricular papillary muscles (○; n= 7) treated with increasing 2-APB. (B) Mechanical function of a paced left atrium (0.1 Hz), pre-treated with 300 nM Bay K 8644 (BayK) and exposed to 7.5 μM 2-APB (2-APB). Inset: These atria do not contract without stimulation. (C) Mechanical function of a left atrium pre-treated with Bay K 8644 (BayK) and exposed to 20 μM 2-APB (2-APB). Inset: These unpaced atria contract spontaneously, persistently, and rapidly.

2.6. Rationale for inhibitors and dosages tested

Experiments in Section 2.7 and 2.8 use pairs of inhibitors of (i) calcium entry via voltage-independent calcium channels, (ii) calmodulin or (iii) CaMKII to test whether 2-APB-induced ectopy requires these channels and/or calmodulin-linked signaling.(i) As one test for the participation of voltage-independent calcium channels in 2-APB-induced ectopy, we determined whether concentra- tions of SKF-96365 that inhibit voltage-independent calcium entry suppress ectopy (Ju et al., 2007; Prakash et al., 2004). Next we tested whether 20 μM ML-7 suppresses 2-APB-induced ectopic activity. This concentration of this naphthalenesulphonamide effectively blocks voltage-independent calcium entry (Wang et al., submitted for publication; Watanabe et al., 1998). (ii) We used 60 μM fluphenazine- N2-chloroethane and phenoxybenzamine to assess whether calmodulin plays a role in 2-APB-induced ectopy. This concentration of these two compounds effectively and irreversibly inhibits calmodulin in cell extracts and whole tissue (Early et al., 1984; Hait et al., 1987). Finally, we probed whether CaMKII participates in this ectopy using 20 μM of the CaMKII inhibitors KN-93 (Anderson et al., 1998) and KN-62 (Davies et al., 2000), a concentration within the range that inhibits CaMKII in complex biological systems.

2.7. The effect of SKF-96365 on 2-APB-induced automaticity

Four experiments tested whether SKF-96365 affects either 2-APB- induced ectopy or normal muscle function. (i) 0.1 Hz-paced left atria (n= 7) and 1 Hz-paced papillary muscles (n= 7) were superfused in KH at 30 °C, and 20 μM 2-APB was added to the superfusate. Once these muscles began to spontaneously contract, 50 μM SKF-96365 was added to the superfusate. We measured muscle spontaneous contraction rate immediately before and then three minutes after the addition of SKF-96365. We also measured muscle mechanical function before the addition of 2-APB to the superfusate and compared it to muscle function measured after the addition of 2- APB and SKF-96365. (ii) 0.1 Hz-paced left atria (n= 7) and 1 Hz- paced papillary muscles (n= 7) were superfused in KH at 30 °C. 20 μM 2-APB was added to the superfusate and once spontaneous contraction occurred, increasing concentrations of SKF-96365 from A 0 to 30 μM were added to the superfusate. The rate of spontaneous contraction was measured three minutes after the addition of any concentration of SKF-96365. The IC50s at which SKF-96365 sup- pressed 2-APB-ectopy were calculated from this data. (iii) To assess whether SKF-96365 affects left atrial electrical properties, 30 μM SKF- 96365 was added to the superfusate of 2-APB-treated, spontaneously contracting left atria (n= 3).

2.8. The sensitivity of 2-APB-induced automaticity to ML-7, calmodulin, and CaMKII inhibitors

Six pharmacological analyses buttressed a role for Orai-linked calcium signaling in 2-APB-induced automaticity. All experiments were performed at 30 °C. (i) A group of left atria (n=6) were superfused, paced at 0.1 Hz, and left untreated. Groups of 0.1 Hz-paced left atria (n=6 per) were pre- incubated for 15 min with (ii) 20 μM ML-7, (iii) 60 μM fluphenazine-N-2- chloroethane, (iv) 60 μM phenoxybenzamine, (v) 20 μM KN-62, and (vi) 20 μM KN-93. 20 μM 2-APB was added to the superfusate of all muscles, and the rate at which they spontaneously contracted was measured during the next 15 min.
To assess whether these drugs affected left atrial mechanical function, we measured the mechanical function of five groups of superfused 0.1 Hz paced, untreated left atria (n= 5 per) before and 15 min after exposure to identical concentrations of these five drugs.

2.9. Analyses of myocyte Orai expression

Western analyses: Triton X-100-soluble left atrial and left ventricular protein (80 μg; n= 3) were Western analyzed for Orai1 and Orai3 using 1 μg/ml of the ProSci 4281 and 4117 antibodies, respectively (ProSci, Poway CA, USA) (Wolkowicz et al., 2002). 5% non-fat dry milk suspended in phosphate-buffered saline (PBS) was used to block nitrocellulose transfer membranes and to dilute antibodies. Westerns using Orai antibodies pre-incubated with blocking peptides established the specificity of these blots. Immunofluorescence analyses: Left atrial and left ventricular frozen sections were fixed for 20 min in 3.7% formaldehyde-PBS, washed in PBS, and then incubated with 0.2% sodium dodecyl sulfate-20 mM dithiothreitol-PBS for seven minutes at 55 °C. Permeabilized sections were washed with PBS, blocked for 1 h with 1% albumin-10% goat serum-PBS, washed with PBS, and exposed overnight at 4 °C to either (i) PBS, (ii) 3 μg/ml of the Orai1 antibody, (iii) 3 μg/ml of the Orai3 antibody or (iv) an anti-bacterial chloramphenicol acetyltransferase antibody (# 5307-310127; 5-Prime-3Prime, Boulder CO, USA); each of these was suspended in 1% albumin-PBS. Sections then were washed with PBS and incubated for one hour at room temperature with 2 μg/ml of the Invitrogen A31618 goat anti-rabbit Alexafluor 488 secondary antibody diluted in 1% albumin-2% goat serum-PBS (Invitrogen, Carlsbad CA, USA). After washing with PBS, sections were exposed for three hours at room temperature to 2 μg/ml of the Sigma A7732 muscle-specific alpha-actinin antibody (Sigma, St. Louis MO, USA). Sections were washed with PBS and exposed for one hour at room temperature to 3 μg/ml of the ThermoScientific 3551 goat anti-mouse DyLight 594 secondary antibody diluted in 1% albumin-2% goat serum-PBS (ThermoScientific, Rockford IL, USA). A final PBS wash was performed. Sections then were sealed and analyzed using a Nikon Eclipse TE-2000-U fluorescence microscope and the Metamorph 2.0 software package. Control sections treated with PBS or with the chloramphenicol acetyltransferase antibody produced weak fluores- cence (Not shown).

2.10. Statistics

Data are the mean±S.E.M. Fisher’s least protected significance difference test compared two means. Two-way repeated measure analysis of variance compared means between different groups. Significance was assigned at Pb 0.05.

3. Results

3.1. 2-APB induces high frequency ectopic activity in superfused left atrial and ventricular muscle

2-APB activates Orai-linked calcium transport in non-excitable cells with EC50s of 20 and 14 μM (Peinelt et al., 2008), and it induces ectopy in superfused rat left atria (Wolkowicz et al., 2007a,b). Since altered calcium homeostasis underlies some forms of ectopy (Anderson et al., 1998), we tested how concentrations of 2-APB that activate the Orais affect intact, isolated heart muscle.

We find that superfused, calcium-loaded left atria and papillary muscles exposed to ≤8 μM 2-APB require pacing to contract (Fig. 1A and B (Inset)). However, exposing muscles to higher concentrations of 2- APB causes them to contract in the absence of a pacing stimulus, and at 30 °C this ectopic activity reaches a maximum of 230 ± 8 spontaneous contractions(min)−1 for left atria and 282 ±13 spontaneous contractions(min)−1 for papillary muscles (Fig. 1A and C: Inset). The EC50s for 2-APB-induced ectopy are 14 ±2.2 μM for left atria and 12 ±1.3 μM for
papillary muscles, and both types of muscle respond to 2-APB in a hyperbolic manner with left atria, for example, exhibiting a Hill coefficient of 8.5 ±0.52. This ectopy reaches a maximum at approxi- mately 20 μM 2-APB, maintains constant rates at higher concentrations of 2-APB (Fig. 1A), and persists indefinitely without pacing (Fig. 1C: Rest).

We recorded the electrical activity of 2-APB-treated, spontaneously contracting left atria to assess whether they produce normal action potentials and to establish whether partial or phase 4 depolarization occurs in these muscles. We find that left atrial action potential upstroke velocity, amplitude, and duration are identical between 2-APB-treated, spontaneously contracting and untreated, paced left atria (Fig. 2A and B, and Table 1). Further, spontaneously contracting left atria maintain resting potentials identical to untreated muscles (Table 1). Ectopic depolarizations capture rat left atria in a 1:1 manner. Thus, the electrical properties of spontaneously contracting left atria match those of normal, untreated, paced muscles. We find no evidence for phase 4 destabilization in these spontaneously contracting muscles (Fig. 2A: Diastolic interval).

Next we ascertained the temperature dependence of left atrial 2- APB-induced ectopy to estimate its possible physiological rate. We measure a Q10 of 3.0 ± 0.11 for this spontaneous activity, a value significantly greater than the Q10 for normal right atrial automaticity, 2.5 ± 0.17 (pb 0.01). At 37 °C, 20 μM 2-APB provokes 543 ± 13 spontaneous left atrial contractions(min)−1 while right atrial normal automaticity occurs at the significantly slower rate of 359 ± 9 spontaneous contractions(min)−1 (pb 0.01).

3.2. 2-APB induces automatic activity in non-automatic cardiac muscle

We next assessed whether 2-APB induces triggered events that closely associate with a preceding external stimulus or whether it provokes automatic activity independent of an external stimulus (Vassalle, 1971; Waldo and Wit, 1993). Untreated left atria (Fig. 3A: Rest) and papillary muscles (Not shown) are quiescent without pacing. By contrast, 2-APB provokes ectopy in quiescent left atria 242 ± 12 s after its addition to the showing that no difference occurs between the resting potentials of 2-APB-treated spontaneously contracting left atria and untreated, paced muscles (Table 1), 10 to 20 μM 2-APB appears to induce a novel type of automaticity in non-automatic cardiac muscle.

3.3. Left atrial and ventricular myocytes express Orai1 and Orai3

The EC50 values and the hyperbolic response of heart muscle to 2- APB suggest the involvement of the Orais in producing this automaticity. To provide initial support for this hypothesis, we evaluated whether left atrial and papillary muscle myocytes express Orai1 or Orai3. Western analyses show that rat left atria (Fig. 4A) and papillary muscle (Not shown) contain both voltage-independent calcium channels. Immunofluorescence analyses of left atrial (Fig. 4B and C) and papillary muscle (Not shown) frozen sections demonstrate that cardiac cells which express Orai1 or Orai3 also contain muscle alpha- actinin. Both Orais distribute in a punctate pattern in alpha-actinin- containing cardiac myocytes, typical of the active form of these proteins (Fig. 4B and C: Arrows; Luik et al., 2006). Negative controls show only weak, diffuse fluorescence (Not shown).

3.4. SKF-96365, ML-7, and calmodulin inhibitors suppress 2-APB-induced automaticity

We performed pharmacological analyses to buttress a hypothesis that Orai calcium signaling underlies the high frequency automaticity we observe in cardiac muscles exposed to 2-APB.(Table 1). SKF-96365 alone does not affect untreated left atrial or papillary muscle mechanical function (Not shown). Importantly, ML-7, a second inhibitor of voltage-independent calcium entry (Wang et al., submitted for publication; Watanabe et al., 1998) and a homolog of the store-operated calcium channel inhibitor 1-(5- chloronaphthalene-1-sulfonyl)-hexahydro-1,4-diazepine (ML-9) (Norwood et al., 2000; Smyth et al., 2008), suppresses 2-APB automaticity in superfused left atria (Fig. 6A) but does not affect the mechanical function of paced, untreated atria (Not shown).

Finally, the calmodulin inhibitors phenoxybenzamine and fluphenazine-N-2-chloroethane suppress 2-APB-induced left atrial automaticity (Fig. 6A and B: Lower panel) but do not affect normal muscle function (Not shown). The CaMKII inhibitors KN-93 and KN- 62 do not affect 2-APB-induced automaticity (Fig. 6A).

4. Discussion

We reported that 2-APB produces mechanical instability in isolated superfused rat left atria and that the rate of this ectopy depends on the apparent muscle calcium load, increasing from a sporadic to a tachycardic rate in muscles treated with various calcium-loading agents (Wolkowicz et al., 2007a,b). The studies reported here address four questions that arise from these initial reports. First, does this ectopic activity occur in ventricular muscle as well as atria? Second, is this activity a reentrant, a triggered or an automatic event? Third, can its rate match the rates reported for arrhythmic drivers? Finally, what cellular mechanism might initiate this ectopic activity?

Superfused rat left ventricular papillary muscles respond to 2-APB in a manner identical to rat left atria (Fig. 1A and B). Specifically, 2- APB produces mechanical instability in superfused papillary muscles with a similar EC50 and hyperbolic response as measured in left atria. This ectopy occurs at similar rates and with similar inhibitor sensitivities in both types of muscle. These data suggest that 2-APB targets a common site or sites in left atria and ventricular papillary muscle to provoke a persistent high frequency ectopy.

Prolonged periods of left atrial and ventricular muscle quiescence precede 2-APB- induced high frequency ectopy (Fig. 3B). These results argue against reentry in the instigation of this ectopy since abnormal impulse conduction usually requires an initiating event or an arrhythmogenic substrate (Mandapati et al., 2000; Panfilov and Pertsov, 2001). Neither is present in these unpaced, normal muscles. The fact that 2-APB induces rapid ectopy at near constant rates in ≤ 4 mg wet-weight muscle also argues against a reentrant mechanism

In contrast to unpaced 2-APB-treated left atrial or papillary muscles which contract spontaneously at rapid rates (Fig. 3B: 2- APB), such ectopy does not occur in unpaced muscles that are exposed first to 2-APB and then to 40 μM SKF-96365 (Fig. 3C: SKF- 96365). SKF-96365 reverses 2-APB-induced ectopy with an IC50 of 11 ± 1.6 μM in left atria and 6 ± 1.6 μM in papillary muscles (Fig. 5A and B: SKF-96365). Excitation–contraction coupling remains intact in [2-APB and SKF-96365]-treated left atria and papillary muscles, as they require pacing to contract, and these contractions occur with forces, times-to-peak tension, and relaxation times identical to those measured prior to the activation of 2-APB-automaticity (e.g., Fig. 5B: 0.1 Hz). Left atria that are treated first with 2-APB and then with SKF-96365 require pacing to produce action potentials; these action potentials arise from equivalent resting potentials and are identical to those acquired from untreated, paced left atria as these muscles could not easily support an appropriate reentrant circuit (Vaidya et al., 1999).

The fact that 2-APB provokes ectopy in the absence of external stimulation, that this ectopy occurs from normal resting potentials, and that these spontaneously active left atria produce normal action potentials (Fig. 2 and Table 1) all suggest that 2-APB induces a novel type of automaticity. We draw this conclusion because the first characteristic is typical of automaticity while the latter two suggest a novel arrhythmogenic mechanism distinct from earlier ones that depend on partial muscle depolarization (Waldo and Wit, 1993). The normal resting potentials and action potentials measured in auto- matically contracting left atria also indicate that 2-APB acts through a mechanism that does not grossly distort muscle electrical activity (El-Sherif et al., 1992). The identical upstroke velocities measured in paced and in spontaneously contracting left atria argue against the presence of an inward current that slowly depolarizes muscle to threshold potential. Rather, our data suggest that the activation of intracellular events downstream of a 2-APB target causes normal, non-automatic heart muscle to enter into an automatic state. Similar automatic activity has been observed previously (Nuss et al., 1999) albeit in untreated myocytes isolated from failing hearts. It may be that the mechanism induced by 2-APB in our normal muscles also underlay the data in this earlier report.

Focal micro-reentry or triggered reentry is thought to be the cause of the persistent or paroxysmal high frequency ectopic sources reported in many clinical arrhythmias (Mandapati et al., 2000; Nattel, 2005; Panfilov and Pertsov, 2001; Sanders et al., 2005; Vaidya et al., 1999). 2-APB induces left atrial automaticity at rates of 9Hz or greater at 37 C indicating that this type of ectopy might occur in vivo at rates similar to those reported for atrial arrhythmic drivers. Concern about ventricular rundown (Urthaler et al., 1997) precludes measuring automaticity in papillary muscles at 37 °C. However, if left atria and ventricle were to have similar Q10s, then 2-APB automaticity might occur at 12 Hz in left ventricle, and thereby be profoundly arrhythmogenic. In support of this possibility, we find that Langendorff perfusion of rat hearts with a concentration of 2-APB that strongly activates automaticity, 22 μM, induces reversible ventricular fibrillation (Wang et al., submitted for publication). Thus the mechanism through which 2-APB disrupts the electromechanical stability of non-automatic heart muscle may be an unexpected means to initiate arrhythmia in vivo.

While 2-APB targets multiple proteins in non-excitable and excitable cells (Maruyama et al., 1997; Missiaen et al., 2001), the hyperbolic concentration-dependence of automaticity with EC50s near 13 μM strongly suggests the Orais as its initiating mechanism (Peinelt et al., 2008).
Four other sets of data support the Orais as a likely target for 2- APB. First, Western and immunofluorescence analyses demonstrate that cardiac myocytes express Orai1 and Orai3 (Fig. 4A to C).

Consequently, these voltage-independent channels afford a myocyte target through which 10 to 20 μM 2-APB might influence heart function. Second, SKF-96365 prevents and reverses atrial and ventricular high-frequency automaticity with an IC50 of ~ 10 μM (Fig. 3C and 5). Similar concentrations of SKF-96365 suppress Orai- dependent calcium entry in non-excitable cells, isolated ventricular myocytes, and intact sinoatrial node preparations. Since SKF-96365 does not affect left atrial contractility (Fig. 5), resting or action potentials (Table 1), it seems unlikely that off-target effects (Hong et al., 1994) contribute to SKF-96365 inhibition of 2-APB-automa- ticity. Third, ML-7 an inhibitor of voltage-independent calcium entry also prevents 2-APB automaticity (Fig. 6A). Finally, two calmodulin inhibitors suppress 2-APB-automaticity. Neither ML-7 nor these calmodulin inhibitors affect untreated left atrial mechanical function. By contrast, the CaMKII inhibitors KN-93 and KN-62 do not affect 2- APB-automaticity (Fig. 6). These results suggest the novel type of automaticity defined here requires an Orai-linked process and active calmodulin but not CaMKII. Further evidence for subtle alterations in muscle calcium homeostasis during automaticity is seen in observa- tions that diastolic tension remains constant in most experiments (Fig. 3B) but that small changes occur under certain conditions (Fig. 5; Inset). Thus Orai calcium entry may contribute in a novel manner to maintaining resting myocyte calcium levels. Biochemical and molecular studies will elucidate how calcium signaling initiates automaticity in 2-APB-treated muscle.

Concern may be raised that alternate 2-APB targets contribute to the electromechanical destabilization defined here. 2-APB does block

Fig. 6. Calmodulin inhibitors and ML-7 but not CaMKII inhibitors prevent 2-APB-induced automaticity: (A) (Open bar) Untreated 0.1 Hz-paced left atria (n= 6) were exposed to 20 μM 2-APB and spontaneous contraction was measured 10–15 min later. 0.1 Hz-paced left atria were exposed to 60 μM fluphenazine-N-2-chloroethane (Speckled bar; n= 6) or phenoxybenzamine (Hatched bar;n= 6) for 15 min, then to 20 μM 2-APB. Spontaneous contraction rates were measured 10–15 min later. Both agents decrease these rates (‡;pb 0.01). (Vertical stripes) Left atria (n= 6) were pre-treated with 20 μM ML-7 and then exposed to 2-APB. Their rate of spontaneous activity was measured 10–15 min later and ML-7 decreased this rate (‡;pb 0.01). Left atria (n= 6 per) were exposed to 20 μM KN-62 (Solid bar) or KN-93 (Diamond bar) for 15 min, then to 20 μM 2-APB. Neither drug affected their spontaneous contraction rates measured 10–15 min later. (B) (Upper) Typical mechanical function of a calcium replete (No Bay K 8644 pre-treatment) left atrium exposed to 20 μM 2-APB (2-APB) measured in the absence of pacing. (Lower) Typical mechanical function of such a left atrium treated first with 60 μM fluphenazine-N2-chloroethane (FCE) and then with 2-APB measured in the absence of pacing.connexin43 but with an IC50 of 51 μM, and it also affects other cell proteins at similar or higher concentrations (Bai et al., 2006; Missiaen et al., 2001). However, the low sensitivity of such sites compared to the EC50 of ~ 13 μM that we report argues against the participation of proteins like connexin43 in automaticity. 2-APB does inhibit Trp channels at the concentrations used here (Birnbaumer, 2009) and the inositol-tris phosphate receptors at higher concentrations (Maruyama et al., 1997). It is not apparent how 2-APB inhibition of these channels could destabilize heart muscle, as both Trps and the inositol-tris phosphate receptors are most likely inactive in quiescent heart. In addition, the main known action for SKF-96365 is to inhibit voltage- independent calcium channels. Thus, in a scenario where 2-APB induces automaticity by inhibiting Trps, SKF-96365 would have to suppress automaticity by relieving Trp channel inhibition thereby restoring them to a normal state.

By contrast, our hypothesis predicts that 2-APB activates Orai- linked calcium entry which provokes novel myocyte signaling events. These events and associated post-translational modifications might occur during the period between 2-APB addition to the superfusate and the appearance of high frequency ectopy (Fig. 3B; Rest), and provoke this novel type of cardiac automaticity. SKF-96365 would prevent Orai calcium entry and suppress ectopy. Nonetheless, our data do not rigorously exclude the possibility that Trps or inositol-tris phosphate receptors participate in 2-APB automaticity; this prospect requires further testing.

In conclusion, cardiac myocytes express both Orai1 and Orai3, and it appears that the activation of one or both of these voltage- independent calcium channels initiates a type of automaticity in non- automatic heart muscle. If this novel mechanism were to occur in vivo, it could produce persistent or intermittent high frequency ectopy depending on the duration of a physiological or pathophysiological activator signal for the Orai channels (Parekh and Putney, 2005; Shuttleworth et al., 2004).CM 4620 Such Orai-directed signals, signaling pathways or ligands may be novel therapeutic targets against arrhythmia.