Overexpression of adenylyl cyclase type 5 (AC5) confers a proarrhythmic substrate to the heart

Abstract

Inhibition of β-adrenergic receptor (β-AR) signaling is one of the most common therapeutic approaches for patients with arrhythmias. Adenylyl cyclase (AC) is the key enzyme responsible for transducing β-AR stimulation to increases in cAMP. The two major AC isoforms in the heart are types 5 and 6. Therefore, it is surprising that prior studies on overexpression of AC5 and AC6 in transgenic (Tg) mice have not examined mediation of arrhythmogenesis. Our goal was to examine the proarrhythmic substrate in AC5Tg hearts. Intracellular calcium ion (Ca²⁺_i) was imaged in fluo-4 AM-loaded ventricular myocytes. The sarcoplasmic reticulum (SR) Ca²⁺ content, fractional Ca²⁺ release, and twitch Ca²⁺ transient were significantly higher in the AC5Tg vs. wild-type (WT) myocytes, indicating Ca²⁺ overload in AC5Tg myocytes. Action potential (AP) duration was significantly longer in AC5Tg than in WT myocytes. Additionally, AC5Tg myocytes developed spontaneous Ca²⁺ waves in a larger fraction compared with WT myocytes, particularly when cells were exposed to isoproterenol. The Ca²⁺ waves further induced afterdepolarizations and triggered APs. AC5Tg hearts had increased level of SERCA2a, oxidized Ca²⁺/calmodulin-dependent protein kinase II (CaMKII), and phosphorylation of ryanodine receptors (RyR) at the CaMKII site, especially after isoproterenol treatment. This was consistent with higher reactive oxygen species production in AC5Tg myocytes after isoproterenol treatment. Isoproterenol induced more severe arrhythmias in AC5Tg than in WT mice. We conclude that AC5 overexpression promotes arrhythmogenesis, by inducing SR Ca²⁺ overload and hyperactivation of RyR (phosphorylation by CaMKII), which in turn induces spontaneous Ca²⁺ waves and afterdepolarizations.

enhanced β-adrenergic receptor (β-AR) signaling, by increasing adenylyl cyclase (AC) activity and cAMP, is one of the most common mechanisms for arrhythmias (15, 26), and conversely, inhibition of β-AR signaling is one of the most common therapeutic approaches for patients with arrhythmias (8). There are two major AC isoforms in the heart (type 5 and type 6, AC5 and AC6) and despite the fact that AC is the key enzyme responsible for transducing β-AR stimulation to increase cAMP, little has been studied on how overexpressing either AC5 or AC6 affects cardiac arrhythmias. Interestingly the prior literature on this topic has been controversial, showing that these two isoforms do not always regulate the heart identically. For example, some studies showed that AC6 overexpression protects the heart against myocardial infarction (20), whereas other studies showed it failed to protect the heart from chronic pressure overload (11). In addition, some studies have shown that AC5 overexpression rescued the cardiomyopathy induced by G_αq expression (17), but not the one induced by β1-AR (22). Furthermore, AC5 inhibition has been shown to be cardioprotective (16, 25), while AC5 overexpression to be deleterious (14).

The goal of the present investigation was to determine the role of AC5 signaling in arrhythmogenesis by using a transgenic mouse with AC5 overexpression (AC5Tg). We showed that cardiac-specific overexpression of AC5 is detrimental to cardiac tissue by directly causing Ca²⁺_i overload and oxidative stress, thus generating a proarrhythmic substrate.

MATERIALS AND METHODS

Mouse models.

Generation of cardiomyocyte-restricted AC5Tg mice has been described in detail previously (12). Mice of either sex were used at 2–5 mo of age. Animals used in this study were maintained in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council, 2011). This study was approved by the Animal Care and Use Committee at Rutgers-New Jersey Medical School.

Cell isolation.

Ventricular myocytes were enzymatically isolated from the left ventricles of adult mouse hearts. After being anesthetized with isoflurane in a covered beaker, the hearts were removed from the mice and perfused retrogradely in Langendorff fashion at 37°C with nominally Ca²⁺-free Tyrode's solution containing ∼1 mg/ml collagenase (type II; Worthington) and 0.1 mg/ml protease (type XIV, Sigma) for 13–15 min. After the enzyme solution was washed out, the heart was removed from the perfusion apparatus and swirled in a culture dish. The Ca²⁺ concentration was slowly increased to 1.0 mM, and the cells were stored at room temperature until they were ready for use. The cells were used within 8 h of isolation. All single cell electrophysiological experiments were performed at 35–37°C.

Patch-clamp methods.

The myocytes were patch-clamped using the whole cell configuration of the patch-clamp technique in the current-clamp or the voltage-clamp mode. To record APs, patch pipettes (2–5 MΩ) were filled with an internal solution containing (in mM) 110 K⁺-aspartate, 30 KCl, 5 NaCl, 10 HEPES, 0.1 EGTA, 5 Mg-ATP, 5 Na₂-creatine phosphate, and 0.05 cAMP (pH 7.2, adjusted with KOH). The myocytes were superfused with normal Tyrode's solution containing (in mM) 136 NaCl, 5.4 KCl, 0.33 Na₂PO₄, 1.0 CaCl₂, 1 MgCl₂, 10 glucose, and 10 HEPES (pH 7.4, adjusted with NaOH). Action potentials (APs) were elicited with 2-ms, 2- to 4-nA square pulses at various pacing cycle lengths (PCLs).

To record L-type Ca²⁺ current (I_Ca,L), patch pipettes were filled with internal solution containing (in mM) 110 Cs-aspartate, 30 CsCl, 5 NaCl, 10 HEPES, 0.1 EGTA, 5 MgATP, 5 Na₂-phosphocreatine, pH 7.2 with CsOH. The cells were perfused with a modified Tyrode's solution containing (in mM) 120 NaCl, 20 TEACl, 0.33 Na₂PO₄, 1.0 CaCl₂, 1 MgCl₂, 10 HEPES, pH 7.4 adjusted with NaOH. The myocytes were stimulated at a PCL of 6 s with a double-pulse protocol. Following a 100-ms prepulse to −40 mV from the holding potential of −80 mV (to inactivate Na⁺ current and T-type Ca²⁺ current), I_Ca,L was elicited by a subsequent test depolarization step to test potentials (−40 to +50 mV at a 10-mV step) for 300 ms. For total outward K⁺ current recording, the pipette and superfusion solutions were the same as those for AP recording. Tetrodotoxin (10 μM) and CdCl₂ (0.5 mM) were added into the Tyrode's solution to inhibit I_Na and I_Ca,L. K⁺ currents were elicited from a holding potential of −80 mV by a series of 400-ms test pulses from −40 to +40 mV in 10-mV increments. The inward Na⁺/Ca²⁺ exchange current (I_NCX) was recorded by a rapid application of 10 mM caffeine (∼2-s duration). The application of caffeine was always preceded by a set of ten 200-ms conditioning depolarization pulses from −40 to 0 mV to ensure a consistent degree of SR Ca²⁺ loading. A holding potential of −40 mV was maintained during caffeine application. Electrical signals were measured with an Axopatch 700A patch-clamp amplifier, controlled by a personal computer using Digidata 1322A interface, driven by pCLAMP10 software.

Ca²⁺_i measurement.

Myocytes were loaded with the Ca²⁺ indicator fluo-4 by incubating them for ∼30 min in bath solution containing 4 μM fluo-4 AM (Molecular Probes), after which the cells were washed and placed in a heated chamber on a Nikon Eclipse TE200 inverted microscope. Ca²⁺_i fluorescence [excitation/emission wavelengths (EX/EM): 485/530 nm] was recorded using an AndorIxon charge-coupled device camera (Andor Technology) operating at ∼50 fps. Fluorescence intensities were measured as the ratio of fluorescence (F) over the basal diastolic fluorescence (F₀).

Measurement of mitochondrial ROS production.

Changes in mitochondrial superoxide production were monitored using MitoSOX Red (Invitrogen/Molecular Probes). Isolated cardiomyocytes were loaded with MitoSOX Red (5 μM) for 15 min at 37°C followed by washout. MitoSOX Red fluorescence (EX/EM: 485/585 nm) was monitored using a Nikon Eclipse TE200 inverted microscope and recorded using an Ixon Charge-Coupled Device camera (Andor Technology) operating at ∼50 fps. The fluorescence images were obtained at the start, and after 40 min in the absence or presence of 1 μM isoproterenol.

Western blot analysis.

Total protein extract from heart tissues (or isolated myocytes) were prepared in a buffer containing protease, kinase, and phosphatase inhibitors. Specific antibodies were used to detect expression or activation of differential proteins. The films were densitometrically evaluated utilizing “Quantity One” software (Bio-Rad Laboratories). Antiserum for oxidized CaMKII was a generous gift from Dr. Mark Anderson (Univ. of Iowa). Other antibodies were from Thermo Fisher Scientific [ryanodine receptor (RyR), dihydropyridine receptor (DHPR)], Santa Cruz Biotech (PLN, β1-AR), or Abcam (P-RyR, Pan-cadherin).

Electrocardiogram.

Standard lead II ECG signals were recorded in anesthetized mice (pentobarbital 0.04 mg/g) by using bipolar limb leads through a differential amplifier (model DP301, Warner Instruments). The signals were filtered between 0.1 and 100 Hz, displayed, and recorded to computer by using AxoScope 10 through MiniDigi 1A converter (Molecular Device).

Arrhythmia scoring system.

The definitions and point values of various arrhythmic events were as follows (7, 10, 27): 0 point: no arrhythmia; 1 point: premature ventricular contraction (PVC); 2 point: paired PVC; 3 point: bigeminal or trigeminal PVC or nonsustained ventricular tachycardia (VT) (≥ 3, but <10 consecutive PVC); 4 point: sustained VT (≥ 10 consecutive PVC); 5 point: ventricular fibrillation (VF).

Statistics.

Results are expressed as means ± SE. Individual groups were compared by two-tailed Student's t-test or Fisher's exact tests as indicated in the text. Results were considered statistically significant if the P value was <0.05.

RESULTS

Characterization of AP and Ca²⁺ handling properties in AC5Tg mice at baseline.

To determine whether the overexpression of AC5 affects the cellular electrical and Ca²⁺ handling properties, we compared APs, Ca²⁺ transients (CaT), and SR Ca²⁺ load in isolated ventricular myocytes. As shown in Fig. 1, an apparent late plateau was observed after the rapid repolarization phase in AC5Tg myocytes, causing prolongation of the AP duration (APD) at 90% repolarization (APD₉₀) (WT 46.7 ± 6.9 vs. AC5Tg 65.2 ± 7.6 ms, P < 0.05), while there was no change in the APD at 50% repolarization (APD₅₀) (Fig. 1, A and B). In some cells, we also observed early afterdepolarizations (EADs) in AC5Tg myocytes (Fig. 1C, left). Blocking Na⁺/Ca²⁺ exchange current (I_NCX) by replacing Na⁺ with Li⁺ in the Tyrode's perfusate eliminated EADs and shortened the APD₉₀ (Fig. 1C), suggesting the involvement of a high Ca²⁺ load and forward-mode I_NCX in the APD prolongation in AC5Tg myocytes. Consistent with this assumption, the twitch CaT (F/F₀, WT 2.4 ± 0.1 vs. AC5Tg 3.4 ± 0.1), the SR Ca²⁺ content (F/F₀, WT 3.4 ± 0.2 vs. AC5Tg 4.2 ± 0.2), and the fractional Ca²⁺ release (WT 68.3 ± 3.4% vs. AC5Tg 84.0 ± 2.3 %) were significantly increased, and the 50% duration of twitch CaT (T₅₀) (WT 227.3 ± 14.2 vs. AC5Tg 127.3 ± 11.1 ms) was shortened in AC5Tg myocytes (Fig. 2).

Fig. 1.

Prolonged action potential (AP) durations (APDs) in adenylyl cyclase type 5 transgenic (AC5Tg) myocytes. A: representative AP recordings from ventricular myocytes of wild-type (WT) and AC5Tg mice. B: summary of APD at 50% repolarization (APD₅₀) and APD at 90% repolarization (APD₉₀) values. *P < 0.05 compared with WT control myocytes. Cell number is indicated above each bar (from 3–5 mice each). C: effect of Na⁺-free, Li⁺-containing Tyrode's solution on the AP properties in AC5Tg, suggesting the contribution of Ca²⁺ overload and Na⁺/Ca²⁺ exchange current (I_NCX) on the prolonged APD and early afterdepolarizations (EADs).

Fig. 2.

Increased sarcoplasmic reticulum (SR) Ca²⁺ contents and intracellular Ca²⁺ (Ca²⁺_i) transient amplitudes in AC5Tg myocytes. A: twitch Ca²⁺ transient (CaT) and SR Ca²⁺ content measured as the height of caffeine (10 mM)-induced Ca²⁺ transient in a WT ventricular myocyte. B: same as A, except in an AC5Tg ventricular myocyte. C–F: summarized data for SR Ca²⁺ contents, twitch CaT, fractional SR Ca²⁺ release, and 50% duration of twitch CaT in WT and AC5Tg ventricular myocytes. Cell number is indicated above each bar (from 3–5 mice each). *P < 0.05, **P < 0.01 compared with WT control myocytes.

Alterations in membrane currents in AC5Tg myocytes.

To obtain the direct evidence for potential upregulation of I_NCX, we measured SR Ca²⁺ release and I_NCX in response to caffeine (10 mM) application in fluo-4-loaded myocytes. We found significant increases in the amplitudes of both caffeine-induced Ca²⁺ transients (SR Ca²⁺ content) and peak I_NCX current density in response to caffeine in AC5Tg myocytes compared with WT controls (Fig. 3, A–D). These results were consistent with AP morphology changes as shown in Fig. 1A. Since I_NCX current is dependent on SR Ca²⁺ release, we further measured the ratio of peak I_NCX density against corresponding peak F/F₀ [I_NCX/(F/F₀)] as described previously (18). The normalized I_NCX exhibited the same magnitude in WT (0.39 ± 0.05) and in AC5Tg (0.40 ± 0.06). These results suggest that the enhanced I_NCX in AC5Tg is due to the elevation of Ca²⁺_i loading rather than increased NCX expression.

Fig. 3.

Alterations in membrane currents in AC5Tg myocytes. A and B: caffeine (Caff; 10 mM)-induced Ca²⁺ transients (i.e., SR Ca²⁺ contents) and I_NCX in AC5Tg (n = 6) vs. WT myocytes (n = 6). C and D: simultaneously recorded I_NCX in the same cells as in A and B. E and F: representative current traces of L-type Ca²⁺ current (I_Ca,L) and current-voltage (I-V) relations recorded from AC5Tg vs. WT myocytes. G and H: representative current traces of outward potassium current (I_K) and I-V relations for both peak (I_peak) and steady-state currents (I_ss) recorded from AC5Tg (n = 6) vs. WT myocytes (n = 6). *P < 0.05 compared with WT control myocytes. Myocytes used in the observations were from 3–5 mice in each group.

The I_Ca,L and potassium current (I_K) are the other two major current components determining the AP duration. We therefore evaluated the current density level of I_Ca,L and total I_K. As shown in Fig. 3, E and F, the I_Ca,L current density was increased in the AC5Tg myocytes at various holding potentials. However, the outward I_K,total densities were not altered in the AC5Tg myocytes compared with WT (Fig. 3, G and H).

Alterations in other Ca²⁺ handling proteins in AC5Tg myocytes.

To determine if the alterations in Ca²⁺ handling properties are mediated by changes in expression level or function of any Ca²⁺ handling proteins, we next carried out Western blotting analyses. As shown in Fig. 4A, the expression levels of both SR Ca²⁺ ATPase (SERCA2a) and phospholamban (PLN) were upregulated (SERCA2a, from 1.00 ± 0.08 to 2.94 ± 0.19; PLN, from 1.00 ± 0.13 to 3.11 ± 0.25, n = 3, P < 0.05, respectively) in AC5Tg hearts, while no alteration was found in total calsequestrin (CSQ) expression (Fig. 4A). On the other hand, as shown in Fig. 4B, we did not observe any significant alteration in the expression level of β1-AR, L-type calcium channel (DHPR), and Na⁺/Ca²⁺ exchanger type 1 (NCX1) in the membrane preparation of AC5Tg vs. WT hearts.

Fig. 4.

Western blot analysis of Ca²⁺ handling proteins in AC5Tg hearts. A: expression levels of SR Ca²⁺ ATPase (SERCA2a), phospholamban (PLN), and calsequestrin (CSQ) in AC5Tg vs. WT. Representative blot bands (Aa) and summarized bar graph (Ab) are shown. CSQ shows no difference and serves as an internal control. Data were normalized to WT on the same gel (n = 3). B: expression levels of β1-AR, DHPR, and Na⁺/Ca²⁺ exchanger type 1 (NCX1) from membrane preparation of AC5Tg vs. WT hearts. Representative blot bands (Ba) and summarized bar graph (Bb) are shown. Pan-cadherin serves as an internal control. Data were normalized to WT on the same gel (n = 4). Equal sample loading was also confirmed by Ponceau S staining of the Western blot membrane. C: total ryanodine receptor (RyR) expression and phosphorylation of RyR (pRyR) at position Ser2808 (PKA site) and Ser2814 (CaMKII site) at baseline (vehicle injection, n = 5 in WT and n = 6 in AC5Tg) and after isoproterenol treatment (1 μM) for 40 min (n = 4) (D). Representative blot bands (Ca and Da) and summarized bar graph (Cb and Db) are shown in each panel. Data were normalized to WT on the same gel. ns, nonsignificant; *P < 0.05, **P < 0.01 compared with WT.

In addition, no differences were observed in total RyR2 expression levels between AC5Tg and WT hearts at either baseline or 40 min after isoproterenol injection (Fig. 4, C and D). Since the phosphorylation status of RyR2 is closely related to hyperactivity of RyR2 and intracellular (arrhythmogenic) Ca²⁺ waves, we also compared the phosphorylation levels of RyR2 at protein kinase A (PKA) (S2808) and CaMKII (S2814) sites by using phosphospecific antibodies. As shown in Fig. 4, C and D, the level of phosphorylation at the CaMKII site was significantly higher in AC5Tg hearts compared with WT at both baseline and 40 min after isoproterenol injection, while there was no difference in the levels of phosphorylation at the PKA site in AC5Tg vs. WT. This may be caused by a higher CaMKII activity in AC5Tg hearts than in WT hearts.

We next evaluated the expression and activation level of CaMKII. As shown in Fig. 5, the total expression of CaMKII remained the same. While the phosphorylated form of CaMKII (ratio) was also at the same level in WT and AC5Tg at either baseline or after isoproterenol treatment, the level of oxidized forms of CaMKII was higher in AC5Tg hearts compared with WT at both baseline and after isoproterenol treatment. In addition, the oxidation level of CAMKII in AC5Tg was further upregulated by isoproterenol treatment (P < 0.05).

Fig. 5.

Western blot analysis of CaMKII activation in AC5Tg hearts. A: Western blot bands of total, phosphorylated (P-CaMKII), and oxidized CaMKII (O-CaMKII) in AC5Tg vs. WT hearts at baseline (BL) and after isoproterenol treatment (ISO; 1 μM). Ba: summarized bar graph showing that phosphorylated (at Thr286) CaMKII levels remained the same in all the groups. Bb: summarized bar graph showing that oxidized CaMKII levels were significant higher in AC5Tg hearts, especially after isoproterenol treatment. Heart number is indicated above each bar. *P < 0.05 compared with WT; #P < 0.05 compared with AC5TG baseline.

Enhanced ROS production under β-AR stimulation in AC5Tg myocytes.

Previous studies have shown that β-AR stimulation increases oxidative stress level (ROS production) in the heart (1, 30), and ROS play an important role in cardiac arrhythmogenesis (23, 28). Our recent study has also revealed that manganese superoxide dismutase (MnSOD) is downregulated in AC5Tg mice at baseline and in response to chronic β-AR stimulation (14). Next, we compared the tolerance to oxidative stress in AC5Tg vs. WT mouse myocytes exposed to acute isoproterenol treatment. As shown in Fig. 6, MitoSox-Red fluorescence (an indicator of mitochondrial superoxide level) was measured and compared between WT and AC5Tg myocytes. While treatment with isoproterenol (1 μM, for 40 min) increased MitoSox-Red fluorescence in both WT and AC5Tg myocytes, the magnitude of enhancement was more striking in AC5Tg than in WT. This result suggests that the AC5Tg cardiomyocytes may be less tolerant to oxidative stress induced by acute β-AR stimulation. We assume that isoproterenol-promoted ROS production is likely to account for the augmentation of SR Ca²⁺ leak and generation of Ca²⁺ waves during β-AR stimulation via combined phosphorylation and oxidation of RyRs.

Fig. 6.

Reactive oxygen species (ROS) production in AC5Tg myocytes. A: representative images showing MitoSOX Red-loaded WT and AC5Tg cardiomyocytes under control conditions and after exposure to 1 μM isoproterenol for 40 min. B: summary of MitoSOX Red fluorescence in WT and AC5Tg cardiomyocytes in the absence (vehicle) or presence of 1 μM isoproterenol (Iso). Cell number is indicated above each bar (from 3–5 mice each). *P < 0.05, **P < 0.01.

Ca²⁺ overload and increased spontaneous Ca²⁺ release in AC5Tg myocytes.

The aforementioned alterations in electrophysiology and Ca²⁺ handling in AC5Tg myocytes present a proarrhythmic substrate. We next tried to obtain direct evidence by evaluating the incidence of spontaneous Ca²⁺ wave and afterdepolarizations. As shown in Fig. 7, another important finding was that AC5Tg myocytes frequently developed Ca²⁺ aftertransients (oscillations) following each paced beat (Fig. 7, A and B) after stimulation with isoproterenol. These spontaneous premature Ca²⁺ release events occurred in a significantly larger fraction of AC5Tg cells compared with WT myocytes even under basal conditions, but the incidence was even more striking when cells were exposed to isoproterenol (42.9% vs. 61.9%) (Fig. 7C). The spontaneous Ca²⁺ waves produced early (EADs) and/or delayed afterdepolarizations (DADs) and triggered APs (Fig. 7E). These data strongly suggest that SR Ca²⁺ overload and spontaneous Ca²⁺ release-induced EADs and DADs may facilitate the incidence of arrhythmias in AC5Tg mice in vivo.

Fig. 7.

Spontaneous Ca²⁺ waves and triggered activities (TAs) in AC5Tg myocytes. A: representative Ca²⁺ fluorescence recordings in the absence (Ctl) and presence of isoproterenol (1 μM) in a WT myocyte. B: representative Ca²⁺ fluorescence recordings in the absence (Ctl) and presence of isoproterenol (1 μM) in an AC5Tg myocyte. Prominent Ca²⁺ aftertransients (asterisks) were observed in AC5Tg myocytes after isoproterenol challenge. C: incidence rate of Ca²⁺ oscillations and waves in WT and AC5Tg myocytes in the presence and absence of 1 μM isoproterenol. Myocytes used in observations were from 3–5 mice in each group. E: simultaneous line-scan image along the long axis of the myocyte (Ea), Ca²⁺_i transients (Eb), and APs (Ec) from an AC5Tg myocyte in the presence of isoproterenol (1 μM). Aberrant spontaneous Ca²⁺ waves (indicated by the arrows in Ea, and w's in Eb) causing delayed afterdepolarizations (DADs) (d) and triggered beats (*) in Eb and Ec. Pacing marks are indicated by the arrows. D: same as E, except from a WT myocyte. Neither apparent Ca²⁺ waves nor triggered activities were observed in this specific cell.

Higher susceptibility to isoproterenol-induced arrhythmias in AC5Tg mice.

In the following experiments, we assessed the effect of catecholamine challenge (isoproterenol injection, 6 mg/kg ip) in anesthetized AC5Tg vs. WT mice. Representative ECGs (lead II) recorded from AC5Tg and WT are shown in Fig. 8. We observed that only AC5Tg mice exhibited frequent PVCs and polymorphic VTs after isoproterenol treatment (Fig. 8, B, i and ii). The arrhythmia score was significantly higher in AC5Tg than WT (Fig. 8C). These data suggest that AC5Tg mice are more susceptible to catecholamine challenge-induced arrhythmias. It should be noted that the time courses for isoproterenol-induced heart-rate increase (within 1 min) and occurrence of VT/VF (peak at 20–40 min after injection) are significantly different, which is consistent with the notion that different signaling pathways may be involved in these two influences.

Fig. 8.

Isoproterenol-induced polymorphic ventricular tachycardias (VTs) in AC5Tg mice. Surface ECG (lead II) signals recorded from an anesthetized WT mouse (A) and an AC5Tg mouse (B). Representative traces before isoproterenol injection (Ctl) and 30 min after isoproterenol injection are shown in each case. Multiple PVCs (i) and polymorphic VT (ii) were observed after Iso injection. C: summarized arrhythmia scores in AC5Tg and WT. The significant comparison between two groups was conducted with Fisher's exact tests (P < 0.01).

DISCUSSION

AC5 and AC6 are the two major AC isoforms in the heart, responsible for the generation of cAMP in response to β-AR stimulation. Considering this and the fact that β-AR stimulation is a major mechanism mediating arrhythmogenesis, it is surprising that no prior study on AC5Tg or AC6Tg models have studied their role in arrhythmogenesis, which is the central feature of the present investigation. In the present study, we examined regulation of the electrophysiological and Ca²⁺ handling properties and arrhythmic propensity in the AC5Tg cardiomyocytes. The major findings are 1) the AC5Tg cardiomyocytes are characterized by Ca²⁺_i overload, i.e., increased SR Ca²⁺ content and intracellular CaT; 2) isoproterenol treatment causes more SR spontaneous Ca²⁺ release/Ca²⁺ waves in AC5Tg than in WT. This subsequently accounts for higher predisposition to arrhythmias in AC5Tg mice in vivo; 3) the molecular basis for these functional changes may be due to the increased levels of expression and activity of Ca²⁺ handling proteins, such as SERCA, I_Ca,L, and NCX in AC5Tg myocytes; and 4) exaggerated ROS generation, oxidation of CaMKII, and phosphorylation of RyR2 also contribute to the higher arrhythmic susceptibility in AC5Tg mice.

Prior studies in AC5Tg and AC6Tg mice, although not examining arrhythmogenesis, have examined Ca²⁺ regulation. A recent study by Timofeyev et al. (21) reported that AC5 and AC6 exhibit distinct subcellular compartmentalization and response differently on regulation of I_Ca,L in ventricular myocytes. Their results suggest that the AC5 isoform is mainly localized in the t-tubular region and shows less effect on I_Ca,L regulation by β-AR stimulation due to restriction by phosphodiesterase. However, our present data demonstrate that AC5 does play a critical role in the regulation of Ca²⁺ handling and arrhythmogenesis, when it is overexpressed. This discrepancy may be caused by the secondary remodeling of antioxidative ability (14) and Ca²⁺ handling proteins in AC5Tg mice. Another possibility may be that AC5 present in other subcellular regions, where it is coupled to other cellular signaling pathways, such as regulation of RyR activity and/or mitochondrial function (ROS production), bypasses I_Ca,L.

Electrophysiological basis for the arrhythmogenic substrate in AC5Tg cardiomyocytes.

It has been well demonstrated that afterdepolarizations and TAs are among the main mechanisms responsible for cardiac arrhythmias. Both EADs and DADs are arrhythmogenic when they reach the threshold and initiate ectopic cardiac beats. Our recent study has shown that both I_Ca,L and I_NCX are important inward currents that account for EAD generation under different pathological conditions; while reactivation of I_Ca,L plays a predominant role in EAD genesis under oxidative stress, spontaneous calcium waves and subsequent I_NCX are a predominant cause for EADs under the Ca²⁺ overload condition (31). Our present data suggest that the prolongations of APD and EADs recorded from AC5Tg myocytes are most likely due to the increase in both I_Ca,L and I_NCX (inward currents) while the outward I_K remains unchanged (Fig. 3). Furthermore, the spontaneous Ca²⁺ waves and subsequent I_NCX (or I_ti) contribute to the genesis of EADs and DADs. Since there was no alteration of DHPR expression (Fig. 4B), we assume that the increase of I_Ca,L in AC5Tg myocytes was mediated by PKA-dependent regulation. This assumption fits with the leftward shift in the I-V relation of I_Ca,L activation as shown in Fig. 3F.

Remodeling of Ca²⁺ handling proteins in AC5Tg mouse heart.

Alteration of Ca²⁺ handling occurs under various pathological conditions such as ischemia, hypertrophy, and heart failure (2, 5, 19, 29). Our results have shown that the Ca²⁺ handling properties are also changed in AC5Tg ventricular myocytes. The increased twitch CaT amplitudes and Ca²⁺ contents in the AC5Tg myocytes (Fig. 2) suggest that the AC5 overexpression increases the propensity for Ca²⁺ overload. It is known that PLN binds to SERCA2a and inhibits SR Ca²⁺ uptake by SERCA2a in ventricular myocytes. We found that in AC5Tg myocytes, both SERCA and PLN levels were high (Fig. 4A) and the overall effect on SR Ca²⁺ uptake is enhanced, as indicated by the shorter CaT duration (T₅₀, Fig. 2, B and F) and higher SR Ca²⁺ contents. This result suggests that SERCA upregulation plays a major role in increasing SR Ca²⁺ uptake, while increased PLN level shows less inhibitory effect.

While the total expression of RyR remained unchanged in AC5Tg hearts, the activation of RyR for both Ca²⁺-induced Ca²⁺ release (CICR) and diastolic spontaneous Ca²⁺ leak may be enhanced by CaMKII phosphorylation in AC5Tg hearts, since the activation of CaMKII was upregulated, especially after isoproterenol treatment. Thus, more spontaneous Ca²⁺ waves occurred in AC5Tg myocytes under both baseline and isoproterenol stimulation (Fig. 7). Therefore the overall alteration of Ca²⁺ handling properties in AC5Tg myocytes seem to be due to the combination of direct activation of Ca²⁺ handling proteins (e.g., L-type Ca²⁺ channels) by AC5-cAMP-PKA and compensatory changes in Ca²⁺ handling protein expression (e.g., SERCA and PLN).

Ca²⁺ overload occurs when there is a large increase in intracellular Ca²⁺ levels that cannot be handled by the cell (i.e., beyond the limits of the ability of cardiac cells to handle an increased calcium load; 24). Ca²⁺ overload may occur in the sarcoplasmic reticulum, mitochondria, or the cytoplasm. When Ca²⁺ overload happens it results in abnormal cell function, such as arrhythmias and heart failure. This seems to be the case in AC5Tg myocytes, especially during stimulation with isoproterenol, when more spontaneous Ca²⁺ waves, EADs, DADs, and arrhythmias were observed at the single cell level and in vivo as evidenced by the ECG recordings.

While we observed that I_Ca,L was increased in AC5Tg myocytes, most likely due to PKA-dependent activation, RyR2 phosphorylation level at S2808 (PKA site) did not show changes at either baseline or after isoproterenol treatment. We do not have a ready explanation for the differential effects of PKA on I_Ca,L and RyR phosphorylation. One possibility is that the stimulatory effect of the overexpressed AC5 may be compartmentalized and masked by the action of PDE, which would restrict local activation of PKA and PKA-dependent phosphorylation of RyR2.

Increased oxidative stress in AC5Tg cardiomyocytes.

Our recent studies have shown that the protein expression levels of MnSOD are significantly reduced (by 38%) in AC5Tg hearts, suggesting that AC5Tg hearts are more susceptible to oxidative stress (14). Recent studies have shown that β-AR stimulation promotes ROS production from mitochondria (1), which may account for Ca²⁺ wave generation (3). Our present results are consistent with these observations and further indicate that both Ca²⁺ overload and higher ROS levels are required for the higher propensity of arrhythmias in AC5Tg mice.

Interestingly, the time course for heart rate increase (fight or flight response) was very different from that for the arrhythmia generation (pathological response) after isoproterenol stimulation, suggesting different mechanisms may exist for these two responses. While the classical stimulatory G protein-AC-cAMP-PKA pathway likely contributes to the heart rate increase and acute Ca²⁺ overload (I_Ca,L activation and CaT enhancement), increased arrhythmogenicity may be mediated by the synergistic effect of Ca²⁺ overload and higher ROS production-CaMKII activation-RyR release. This notion makes sense since it has been suggested that both Ca²⁺ overload and reduction of SR threshold for spontaneous Ca²⁺ release are required for a higher incidence of arrhythmias (3, 9). Further mechanistic studies are needed to assess the synergistic roles of differential signaling pathways in causing arrhythmias in AC5Tg hearts. An alternative explanation may involve potential vagal reflex following isoproterenol injection. It has been well known that EAD generation is slow rate dependent. Meanwhile, DADs may be overdriven by fast rate. Therefore, when the heart rate is dramatically increased shortly after isoproterenol injection, the arrhythmogenic effect may be masked. However, when the strong vagal reflex brings heart rate down, the arrhythmogenic aspects of isoproterenol administration are likely to be revealed.

Our previous studies have shown that the cAMP synthesis ability is significantly enhanced in AC5Tg heart (∼20-fold higher) compared with WT (14). Isoproterenol treatment (chronic infusion for 7 days at a dose of 60 mg·kg⁻¹·day⁻¹) promotes the cAMP synthesis rate by ∼22-fold in WT, and interestingly it can further enhance it by ∼2-fold in AC5Tg hearts (13). Combined with our present study, these data imply that isoproterenol-induced arrhythmias in AC5Tg hearts require both PKA (I_Ca,L activation) and ROS-CaMKII (RyR2 activation) pathways.

Sympathetic stress and arrhythmias.

The sympathetic nervous system is activated both in response to physiological stress, e.g., exercise, and pathological stress, e.g., heart failure. Severe arrhythmias (VTs/VFs) and sudden cardiac deaths have been linked to sympathetic stimulation in several familial/genetic arrhythmias, such as catecholaminergic polymorphic ventricular tachycardia (CPVT) and long QT syndrome (LQTS) type II. In this study, we found that arrhythmias only occur in AC5Tg mice after isoproterenol injection, and the mechanism resembles that for CPVT. Although no structural changes (e.g., point mutations) seem to be involved, AC5Tg cardiomyocytes exhibited spontaneous RyR Ca²⁺ releases (Ca²⁺ waves) and EAD/DAD generation at the single cell level after isoproterenol treatment. PVCs and polymorphic VTs were generated as a result of isoproterenol injection in AC5Tg mice. Therefore, it seems likely that a CPVT phenotype can be also induced by a functional upregulation of multiple Ca²⁺ handling proteins under sympathetic stress conditions. In addition, elevated psychosocial/mental stress (with higher sympathetic tone) favors the occurrence of cardiac arrhythmia and sudden death (4, 6). Our present study may also provide new insights into its underlying molecular and cellular mechanism(s).

Conclusions.

Our results showed that AC5 overexpression exerts a proarrhythmic effect, by inducing SR Ca²⁺ overload, increased oxidative stress, spontaneous Ca²⁺ release/Ca²⁺_i waves, and triggered activities, suggesting a novel genetic model for isoproterenol-induced arrhythmias. Since cardiac arrhythmias are a major cause of morbidity and mortality in patients with heart disease, modification of AC signaling in general, and reduction in AC5, in particular, may provide a novel therapeutic approach for arrhythmias induced by increased β-AR signaling.

GRANTS

This work was supported by National Institutes of Health Grants R01-HL-97979, 5P01-AG-027211, 5R21-HL-097264, 1R01-HL-102472, 5R01-HL-033107, 5T32-HL-069752, 5R01-HL-095888, 5P01-HL-069020, 5R01-HL-091781, R01-HL-106511, R01-HL-093481, and 1R01-HL-119464 and by the National Natural Science Foundation of China Grants 81170597 and 81470510.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

Author contributions: Z.Z., G.J.B., H.W., N.F., R.G., X.S., L.Y., and L.-H.X. performed experiments; Z.Z., G.J.B., H.W., N.F., R.G., X.S., L.Y., and L.-H.X. analyzed data; Z.Z., G.J.B., H.W., L.Y., S.F.V., and L.-H.X. interpreted results of experiments; Z.Z., G.J.B., H.W., N.F., R.G., X.S., L.Y., and L.-H.X. prepared figures; Z.Z. and L.-H.X. drafted manuscript; Z.Z., G.J.B., H.W., N.F., R.G., X.S., L.Y., D.E.V., S.F.V., and L.-H.X. approved final version of manuscript; D.E.V., S.F.V., and L.-H.X. edited and revised manuscript; L.-H.X. conception and design of research.