Stretch-activated channel activation promotes early afterdepolarizations in rat ventricular myocytes under oxidative stress

Abstract

Mechanical stretch and oxidative stress have been shown to prolong action potential duration (APD) and produce early afterdepolarizations (EADs). Here, we developed a simulation model to study the role of stretch-activated channel (SAC) currents in triggering EADs in ventricular myocytes under oxidative stress. We adapted our coupling clamp circuit so that a model ionic current representing the actual SAC current was injected into ventricular myocytes and added as a real-time current. This current was calculated as I_SAC = G_SAC * (V_m − E_SAC), where G_SAC is the stretch-activated conductance, V_m is the membrane potential, and E_SAC is the reversal potential. In rat ventricular myocytes, application of G_SAC did not produce sustained automaticity or EADs, although turn-on of G_SAC did produce some transient automaticity at high levels of G_SAC. Exposure of myocytes to 100 μM H₂O₂ induced significant APD prolongation and increase in intracellular Ca²⁺ load and transient, but no EAD or sustained automaticity was generated in the absence of G_SAC. However, the combination of G_SAC and H₂O₂ consistently produced EADs at lower levels of G_SAC (2.6 ± 0.4 nS, n = 14, P < 0.05). Pacing myocytes at a faster rate further prolonged APD and promoted the development of EADs. SAC activation plays an important role in facilitating the development of EADs in ventricular myocytes under acute oxidative stress. This mechanism may contribute to the increased propensity to lethal ventricular arrhythmias seen in cardiomyopathies, where the myocardium stretch and oxidative stress generally coexist.

previous studies have shown that myocardial stretch, a process termed “mechanoelectric feedback,” can induce myocyte depolarization and abnormal impulses by the stretch-activated channel (SAC) currents (7, 21). A subset of SACs has been reported as nonselective cation channels (48), which allow Na⁺, K⁺, Cs⁺, and Ca²⁺ to permeate (36). A linear current-voltage relationship has been described for SACs with reversal potentials measured in myocytes from −6 to −15 mV (17, 33, 48). In a review by Kohl et al. (24), they note that models incorporating SACs as pure charge carriers and models that include movement of ions through the channels both reproduced experimental phenomena, such as diastolic depolarization, positive chronotropic responses of pacemakers, and action potential (AP) generation due to stretch (25).

It has been suggested that oxidative stress is a mechanism for Ca²⁺ overload and the consequent electrical changes in cardiomyocytes (5). On the other hand, Ca²⁺ overload increases oxidative stress by altering mitochondrial function, which leads to increased production of oxygen radicals (9). Oxidative stress alters the function of several elements in the excitation-contraction coupling pathways (including L-type calcium channel, the ryanodine receptor, and the Ca²⁺-ATPase in the sarcoplasmic reticulum) and the electrophysiology of the ventricle (5). Exposure of guinea pig ventricular free wall to H₂O₂ induced extra ventricular beats due to early (EADs) and delayed afterdepolarizations (8).

Because both mechanical stretch and oxidative stress cause Ca²⁺ overload, we hypothesize that abnormal stretch of myocardium may proceed in concert with acute oxidative stress in facilitating the induction of EADs. We have incorporated a model system similar to that used by Riemer et al. (31) on cell models with a major difference that we incorporate a real-time modulation of isolated rat ventricular myocytes with a time- and voltage-varying current produced by a computation of current that would have been produced by a SAC conductance (G_SAC). The advantage of our system is that we are using real myocytes and thus can expose the cells to an agent that produces acute oxidative stress (H₂O₂) and determine the cooperative effects of these two modulations.

METHODS

Cell isolation.

Single ventricular myocytes were isolated from Sprague-Dawley rats of either sex weighing 165–250 g, as our laboratory described previously (44). The myocytes were kept in storage solution and were used within 6 h after isolation.

Perforated patch-clamp recording for APs.

Myocytes were placed in a chamber on an inverted microscope and continuously perfused with Tyrode solution at 2 ml/min and 36 ± 0.5°C. Recording pipettes were pulled from borosilicate glass with resistances of 1.5–3 MΩ when filled with internal solution. High-resistance seals were formed by applying light suction, and we used perforated-patch technique with amphotericin B in the pipette to allow formation of a low series resistance without intracellular dialysis for AP recordings. No corrections were made for liquid junction potential, other than zeroing the potential before touching cell surface. APs were recorded using the whole cell patch clamp with Axoclamp 2A amplifier (Axon Instruments, Foster City, CA) in current-clamp mode. Series resistance was carefully compensated after recording of the membrane potential was established. APs were initiated by current pulses of 2-ms duration and amplitude of 10–15% suprathreshold.

Solutions.

Normal Tyrode is as follows (mM): 148.8 NaCl, 4 KCl, 1.8 CaCl₂, 0.53 MgCl₂, 0.33 NaH₂PO₄, 5 HEPES, and 5 glucose, pH 7.4. The CaCl₂ is omitted for Ca²⁺-free Tyrode. Storage solution is as follows (mM): 120 potassium-glutamate, 20 taurine, 5 MgCl₂, 1 EGTA, 10 dextrose, and 10 HEPES, with pH adjusted to 7.4 using KOH. The pipette solution is as follows (mM): 135 KCl, 5 Na₂CrPh, 5 MgATP, and 10 HEPES, pH 7.2. Amphotericin B 150 μg/ml was added to the internal solution just before use, and the solution was renewed every 2 h.

Incorporating a SAC model into a real isolated myocyte.

We adapted our coupling clamp circuit (40), so that a model ionic current that represents SACs is injected into a isolated myocyte and added as a real-time current. This SAC model was described in detail in our recent work (39). Several studies have shown that physical stretch of isolated ventricular myocytes produced stretch-activated currents with a linear current/voltage relationship and reversal potentials between −6 and −11 mV (6, 48, 49). Thus, for the majority of experiments in this study, we set the reversal potential of SAC (E_SAC) = −10 mV and G_SAC to a constant value for each recording, but the value of G_SAC was varied to investigate the critical G_SAC required to induce EADs. G_SAC was applied as an “on or off” function. To test the effects of E_SAC on SAC-induced EADs, E_SAC of −20 and 0 mV were also employed in some experiments, as indicated.

Intracellular Ca²⁺ concentration transient measurements.

Myocytes were first loaded with 2 μM fura 2-AM for 10 min at room temperature, and fluorescence measurements are recorded with a dual-excitation fluorescence photomultiplier system (IonOptix). After loading, the cells were washed and resuspended in normal Tyrode solution. These myocytes were then placed in the cell chamber, stimulated at 1 and 3 Hz with 4-ms duration, and imaged through a Fluor ×40 oil objective. Cells were exposed to light emitted by a 75-W Xenon lamp and passed though either a 340- or 380-nm wavelength filter. The emitted fluorescence is detected at 510 nm. To take into account any inference effects of background fluorescence, the background fluorescence for each myocyte was determined by moving the myocyte out of the view and recording the fluorescence from the bath solution alone.

Statistical analysis.

Data are reported as means ± SE. Comparison of AP properties before and after application of H₂O₂ and at different pacing rates was done with a one-way analysis of variance for repeated measures, followed by Student-Newman-Keuls post hoc test to determine significant differences (P < 0.05). If the criteria for parametric analysis were not met, a nonparametric analysis using the Friedman repeated-measures analysis of variance on ranks was used, followed by Dunn's post hoc test to determine significant differences (P < 0.05). Comparison of the critical value of G_SAC required to induce abnormal activity before and after application of H₂O₂ was performed with a paired t-test.

Use of vertebrate animals.

All experiments on live vertebrates were performed in accordance with protocols approved by the institution's Animal Care and Use Committee.

Statement of responsibility.

The authors had full access to the data and take full responsibility for its integrity. All authors have read and agree to the paper as written.

RESULTS

Response of rat ventricular cells to SAC.

Using the perforated patch, we applied varying levels of G_SAC to rat ventricular myocytes, either in control solution or in the presence of 100 μM H₂O₂. The representative results were shown in Fig. 1. We paced the myocyte at 1 Hz and applied the G_SAC during the time period from 6.5 to 11.5 s. The responses of myocyte to G_SAC levels (E_SAC was set at −10 mV) of 4, 5, 6, or 7 nS (top to bottom) are shown. The depolarization of the membrane potential increases as G_SAC increases. For G_SAC = 7 nS, there was some induced automaticities. However, these automaticities were not sustained, owing to the inexcitable property of the myocyte at large depolarization (sodium channel is inactivated), and the cell was also not excitable to the direct stimulations (Fig. 1, bottom). None of the cells tested showed induced sustained automaticity or EADs in response to any level of G_SAC in the control solution.

Fig. 1.

Recordings from an isolated rat ventricular myocyte at 1-Hz stimulation with the stretch-activated channel conductance (G_SAC) applied from 6.5 through 11.5 s at various levels. Early afterdepolarization (EAD) was not induced at any levels of G_SAC. However, a transient activity was observed at a high G_SAC (7 nS), but failed to sustain, and the cell became refractory to direct stimulation.

Effect of oxidative stress on cellular response to SACs.

Figure 2 shows the results of the application of G_SAC levels of 2, 3, 4, or 4.5 nS (top to bottom) after a 10-min application of 100 μM H₂O₂ for the same cell as Fig. 1. Application of G_SAC under these conditions did not produce sustained automaticity, but EADs were consistently produced with more EADs and longer EAD duration at the higher G_SAC level. The minimal G_SAC level for inducing transient automaticity (as shown in Fig. 1) was 5.6 ± 0.3 nS in the control solution, and the G_SAC level required for producing EADs in the same cells after 10-min exposure to H₂O₂ was 2.6 ± 0.4 nS (n = 14, P < 0.05). When normalized for cell capacitance, the critical value of G_SAC decreased from 92.6 ± 12.6 pS/pF in control to 43.32 ± 10.9 pS/pF in H₂O₂ (n = 14, P < 0.05). None of EADs has been triggered in myocytes by G_SAC in control solution, indicating that application of H₂O₂ sensitized the cells to EAD development in response to G_SAC. However, EADs were observed in some myocytes exposed to high levels of H₂O₂ (>1 mM), which led to significant AP prolongation (data not shown), consistent with previous reports (42). H₂O₂ concentrations lower than 100 μM produced little or no effect on APD. Thus we take the concentration of 100 μM H₂O₂ as the desired concentration. This concentration falls into the range detected in human blood plasma (15).

Fig. 2.

Recordings from the same cell in Fig. 1 after 10-min exposure to 100 μM H₂O₂ with G_SAC levels of 2, 3, 4, and 4.5 nS. EADs occurred at G_SAC levels of 4 and 4.5 nS with increasing EAD duration during SAC activation. No EAD was produced before or after the G_SAC application.

To investigate the interaction of the G_SAC current with the repolarization process of the cells, we replotted portions of Fig. 2, labeled A and B, for Fig. 3. For each part of this figure, we have used the data from Fig. 2, bottom (G_SAC = 4.5 nS) and replotted the APs just before (part A) and after G_SAC application (part B), along with the membrane current and the G_SAC current. The membrane current (top) was computed by multiplying cell capacitance by the negative of the computed time derivative of voltage of the measured membrane potential vs. time. Note that this definition of membrane current generates values that include the G_SAC current (middle). For Fig. 3A, G_SAC was turned off; so there was no G_SAC current. The membrane current showed a decreasing outward current during the repolarization process. For Fig. 3B, we illustrated these parameters for the first AP after G_SAC was turned on. Before stimulation of the cell, the membrane current (top) was zero, even though the G_SAC current was a steady −260 pA. This is because the membrane current has “adjusted” to the presence of the G_SAC current (which is inward) by the outward current produced by the steady depolarization of the cell. After activation of the cell, a time-dependent change in the G_SAC current was generated with the time course of changes of cell membrane potential. We have included a dashed line at −10 mV in the bottom panels of Fig. 3, A and B, indicating the voltage level of the E_SAC. In Fig. 3B, as the membrane potential went positive to E_SAC, the G_SAC current became transiently positive and then, as repolarization proceeded, it became negative during the time marked by arrows a and b of the middle panel, owing to the membrane potential becoming negative to E_SAC. This inward G_SAC current makes the cell repolarize more slowly, and the membrane current actually became net negative at the time marked by arrow c in the top panel. This net negative membrane current produced interruption of repolarization process in the cell, and the cell then began to depolarize again, which led to an EAD in this cell. The G_SAC current became outward during the positive phase of the EAD, which tends to repolarize the cell.

Fig. 3.

Demonstration of the time course of membrane current (top), G_SAC current (middle), and membrane potential (V; bottom) for the action potentials (APs) occurred just before (A) and after (B) SAC (4.5 nS) activation. B: points labeled a, b, and c indicate times points when the current changes polarity.

Measurements of AP parameters in response to SAC and H₂O₂.

We analyzed the effects of the following three interventions on AP parameters: 1) application of a subthreshold level of G_SAC (2 nS); 2) 10 min of exposure of 100 μM H₂O₂; and 3) 100 μM H₂O₂ at 10 min with G_SAC = 2 nS. These data were summarized in Table 1. The resting membrane potential (RMP) was unchanged by H₂O₂, but was significantly depolarized by application of G_SAC = 2 nS in both control and H₂O₂ solution. The AP amplitude was decreased by G_SAC application (due to the depolarization of RMP and the outward SAC current for the overshoot), but increased by H₂O₂. The AP durations at 30% (APD₃₀) and 50% repolarization (APD₅₀) were unchanged by G_SAC in control solution. The early repolarization in rat ventricular myocytes is very fast, due to large transient outward current (I_to) (29). The G_SAC current during early repolarization is outward because the membrane potential is greater than the E_SAC (−10 mV) and would tend to facilitate early repolarization. But because there is already a large I_to-induced fast early repolarization, G_SAC application has little effect and does not change APD₃₀ or APD₅₀. Whereas during the late repolarization phase, the stretch current is inward (membrane potential is negative to E_SAC), which tends to prolong the APD. Thus APD at 90% repolarization (APD₉₀) was significantly increased (by ∼15%) by G_SAC.

Table 1.

AP characteristics and response to G_SAC and to 100 μM H₂O₂

	Control Solution	100 μM H2O2
Critical
GSAC, nS	5.6±0.3	2.6±0.4*
Normalized
GSAC, pS/pF	92.6±12.6	43.32±10.9*

	GSAC = 0 nS	GSAC = 2 nS	GSAC = 0 nS	GSAC = 2 nS
RMP, mV	−78.8±0.5	−69.9±2.2*	−79.3±0.7†	−76.3±2.7*‡
APD30, ms	4.1±0.7	4.0±0.9	5.4±0.8*†	5.8±1.4‡
APD50, ms	8.7±1.1	8.5±1.3	11.7±2.1*†	11.8±0.8*†
APD90, ms	28.9±2.3	32.9±2.9*	37.5±3.4*†	44.8±4.2*†‡
APA, mV	107.5.4±2.5	96.0±2.5*	102.7±3.5*†	99.1±2.6‡

Values are means ± SE; n = 14. G_SAC, stretch-activated channel conductance; RMP, resting membrane potential; APD₃₀, APD₅₀, and APD₉₀: action potential duration at 30, 50, and 90% repolarization, respectively; APA, action potential amplitude.

^*P < 0.05 compared with control, G_SAC = 0 nS.

^†P < 0.05 compared with control, G_SAC = 2 nS.

^‡P < 0.05 compared with H₂O₂, G_SAC = 0 nS.

The effect of exposure to 100 μM H₂O₂ on APD was dramatic with significant prolongation of APD₃₀, APD₅₀, and APD₉₀. This increase in APDs may reflect an increase in the L-type calcium current (I_Ca) and the late sodium current due to exposure to H₂O₂ (5, 42, 46). When G_SAC = 2 nS is applied to cells in conjunction with H₂O₂, there was a variable effect on the APD compared with control (see Table 1). Since H₂O₂ increased APD₃₀ compared with control, when the stretch current (which is outward during early membrane potential level) was applied, the outward SAC current accelerated membrane repolarization and thus shortened APD₃₀. The addition of G_SAC in the presence of H₂O₂ did not alter APD₅₀ compared with H₂O₂ alone. This is likely because the membrane potential level at APD₅₀ is close to the E_SAC; the current from SAC is very small, thus making little contribution to the APD₅₀. The application of G_SAC in the presence of H₂O₂ further increases APD₉₀ over the control value (a total 55% increase, compared with 30% increase due to H₂O₂ alone and a 25% increase due to G_SAC alone).

Changing the E_SAC model alters the SAC-induced depolarization and the EAD inducibility.

Because the E_SAC varies from different species and cardiac tissue types, cell response to the same G_SAC may be different. In addition, it has been reported that H₂O₂ alters intracellular Ca²⁺ and Na⁺ concentrations by facilitating I_Ca and slowing the late sodium current (42, 46). These H₂O₂-induced changes of intracellular ions may subsequently alter the E_SAC. Furthermore, the G_SAC for multiple ions may also alter the E_SAC with time during stretch. We have tested the SAC effects on EAD induction by varying E_SAC from −20 to 0 mV. In the presence of 100 μM H₂O₂, EADs and automaticity were facilitated by setting the E_SAC to 0 mV, but attenuated by setting the E_SAC to −20 mV (n = 6). The representative traces were shown in Fig. 4A. For this myocyte, a relatively low value of G_SAC (2.4 nS) was applied for 10 s after the end of 1-Hz pacing. This G_SAC caused the largest depolarization at 0-mV E_SAC and the least depolarization at −20-mV E_SAC. This low value of G_SAC did not induce automaticity at −20- or −10-mV E_SAC. However, sustained automaticity was successfully induced at 0-mV E_SAC. The effects of E_SAC on EAD induction are likely a result of altered cell depolarization level. Indeed, in a total of six rat ventricular myocytes, application of 4.0-nS G_SAC induced 17.6 ± 0.7 mV depolarization for the −10 mV-E_SAC, whereas this G_SAC induced a 13.3 ± 0.9 mV depolarization for the −20-mV E_SAC and a 23.3 ± 1.0 mV depolarization for the 0-mV E_SAC (Fig. 4B, P < 0.05). In addition to the depolarization level, setting E_SAC to a higher level also shifts the membrane potential toward the I_Ca reactivation window, which further facilitates EAD induction.

Fig. 4.

The effect of SAC reversal potential (E_SAC) on G_SAC-induced cell depolarization and automaticity in rat ventricular myocytes. A: at a relatively low value of G_SAC (2.4 nS), increasing E_SAC levels from −20 to 0 mV significantly increased the G_SAC-induced cell depolarization. The sustained automaticity was successfully induced by the G_SAC at E_SAC = 0 mV, but not at the E_SAC value of −10 or −20 mV. B: bar graph showing the cell depolarization induced by 2.4-nS G_SAC at the E_SAC level of −20, −10, and 0 mV (n = 6).

Increasing pacing frequency promotes EAD development.

We have tested the frequency-dependent effect on APD by using 3-Hz stimulation, alternating with the 1-Hz stimulation, in both control and H₂O₂ solutions. Representative results are shown in Fig. 5A, in which we have superimposed responses to G_SAC = 2 nS at 1-Hz (solid line) or 3-Hz (dotted line) stimulation in the control solution (top) and after 10-min exposure to 100 μM H₂O₂ (bottom). Application of G_SAC in control solution produced a depolarization on which the repetitive APs were superimposed. APs labeled a occurred just before G_SAC was turned on; those labeled d occurred after the G_SAC has been turned off; and those labeled b and c occurred progressively during the G_SAC application. After 10 min in H₂O₂ solution, the cell produced EADs during G_SAC application only when stimulated at 3-Hz but not at 1-Hz rate. Figure 5B shows the same data for the period just before and after turning on the G_SAC to show the progressive changes in AP shapes.

Fig. 5.

A: superimposed recordings from an isolated rat ventricular cell stimulated at 1 Hz (solid lines) and 3 Hz (dotted lines) before (top) and after 10 min exposure to 100 μM H₂O₂ (bottom), with G_SAC (2 nS) activation period indicated by double-headed arrows. EADs occurred in H₂O₂ solution only at 3-Hz stimulation during this low G_SAC application. B: a faster time scale illustrating the changes in APs (replotted from A).

In control solution, the 1- and 3-Hz APs are nearly superimposed when G_SAC was turned off (Fig. 5B, top, labeled a at 6 s), but, while the G_SAC was turned on (from 6.5 to 11.5 s), the APs at 3 Hz had longer duration than those at 1 Hz. For the same cell in the 100 μM H₂O₂ solution, APs at 1 and 3 Hz had longer APD than in control solution, but are not different from each other (1 vs. 3 Hz) when the G_SAC was off (Fig. 5B, bottom, labeled a). However, when the G_SAC was turned on, APs at 3 Hz had significantly longer APD than those at 1 Hz and increased their APD progressively to form EADs (bottom, labeled b and c), while the APs at 1 Hz did not form EADs. In summary of these findings, fast pacing did not significantly alter APD in rat ventricular myocytes in control and in 100 μM H₂O₂ solution. However, when SAC was activated, fast pacing significantly prolonged APD in myocytes for both solutions and facilitated EAD development in myocytes exposed to 100 μM H₂O₂.

These changes in AP waveforms are shown more clearly in Fig. 6, in which we have superimposed the APs at 3-Hz (dotted lines) with those at 1-Hz rate (solid lines) with the four vertical panels illustrating the APs at the four time points indicated by a, b, c, and d of Fig. 5A. For Fig. 6A (control solution), the fast stimulation slightly prolonged APD at G_SAC off, but dramatically prolonged APD at G_SAC on (b and c), and this effect was completely reversed after the G_SAC was turned off (d).

Fig. 6.

APs recorded in control solution (A) and after 10 min exposure to H₂O₂ (B). Panels from top to bottom illustrate APs labeled a, b, c, and d in Fig. 5A at 1-Hz (solid lines) and at 3-Hz stimulation (dotted lines).

For the same cell after 10-min exposure to H₂O₂ (Fig. 6B), the APD was prolonged before the G_SAC is turned on (compared with the control solution), but the further effect on APD of the fast pacing was absent. While the G_SAC was turned on (panels b and c), there was further lengthening of APD at 1-Hz stimulation and a very pronounced further lengthening of APD at 3 Hz, leading to an EAD in panel c. Similar to the control solution, the effect of G_SAC is completely reversed after the G_SAC has been turned off (panel d).

These data are summarized in Table 2. G_SAC significantly depolarized the RMP at both pacing frequencies. In both control and H₂O₂ solution, increasing pacing frequency tended to increase APD₉₀, but the increase was significant only when G_SAC was applied.

Table 2.

Percent change in AP parameters with increased stimulation frequency

	1 Hz, GSAC = 0 nS	%Change From GSAC = 0 nS
1 Hz, GSAC = 2 nS		3 Hz, GSAC = 0 nS	3 Hz, GSAC = 2 nS
Control solution
RMP, mV	−78.6±0.6	−2.7±0.4*	−4.3±0.8	−23.0±4.1*
APD30, ms	3.9±0.7	−4.5±5.1	4.0±3.3	−1.1±1.5
APD50, ms	8.5±2.1	−3.9±2.5	2.7±2.8	0.9±1.3
APD90, ms	28.6±2.6	6.2±5.6*	2.2±2.3	33.6±14.1*
APA, mV	98±25	−12.6±17	−0.4±1.2	−0.6±2.0
10-min Exposure to 100 μM H2O2
RMP, mV	−79.5±0.5	−4.5±0.6†	−14.7±7.8	−19.6±5.4†
APD30, ms	3.8±0.9	−1.8±1.4	−1.5±2.0	−21±19.8
APD50, ms	8.3±0.3	−4.9±6.2	−2.5±1.9	6.6±4.8
APD90, ms	33.3±6.2	2.3±2.4	4.9±4.6	24±4.2†
APA, mV	103.0±3.8	−8.7±6.1	11.5±12.3	−1.1±2.5

Values are means ± SE; n = 12.

^*P < 0.05 compared with 1 Hz for control solution with G_SAC = 0 nS.

^†P < 0.05 compared with 1 Hz for 100 μM H₂O₂ with G_SAC = 0 nS.

H₂O₂-induced increase in Ca²⁺ transient at different stimulation frequencies.

To explore the mechanisms for the fast pacing-induced facilitation of EADs in the presence of H₂O₂, we measured intracellular Ca²⁺ levels in rat ventricular myocytes at stimulation rate of 1 and 3 Hz. In a total of 10 myocytes, application of 100 μM H₂O₂ significantly increased Ca²⁺ transient at both 1- and 3-Hz stimulation frequencies, with a significantly larger increase at 3 Hz compared with 1 Hz. H₂O₂ did not change the diastolic Ca²⁺ levels at 1 Hz. However, it significantly increased the diastolic Ca²⁺ at 3-Hz stimulation. In other words, in the presence of 100 μM H₂O₂, fast pacing causes a significant Ca²⁺ load, in addition to a larger increase in Ca²⁺ transient (Fig. 7). The alteration of Ca²⁺ cycling at high stimulation frequency in the presence of 100 μM H₂O₂ likely contributes to the facilitation of SAC-induced EADs at fast-pacing rate.

Fig. 7.

H₂O₂-induced increase in diastolic Ca²⁺ and Ca²⁺ transient in rat ventricular myocytes. A: the representative Ca²⁺ transient traces in response to the application of H₂O₂ (100 μM) and the changes in stimulation frequencies (1 and 3 Hz). B: bar graph showing changes in the diastolic Ca²⁺ and Ca²⁺ transient in response to the application of H₂O₂ and changes in stimulation frequencies (n = 10). F/F₀, ratio of fluorescence intensity to the resting fluorescence intensity.

DISCUSSION

Effects of model stretch on AP parameters.

Stretch has been reported to depolarize RMP and to increase, decrease, or have a crossover effect on APD (37). The inconsistent effect on APD may reflect differences of stretching single myocytes, tissue, and intact hearts. Additionally, the different effects on APD may reflect the variety of experimental techniques used to stretch the myocardium and to record the electrical activity (3). Here, we show that incorporation of a SAC model into isolated rat ventricular myocytes resulted in depolarization of the membrane potential and increase in APD₉₀. These results are consistent with a study by Zeng et al. (48), in which they demonstrated that stretch of isolated rat ventricular myocytes resulted in depolarization of RMP and an increase in both APD₅₀ and APD₉₀. Note that these experiments were done at room temperature, and the APD was quite long. Similarly, Kiseleva et al. (23) showed that stretch of a rat left ventricular slice resulted in an increase in APD₉₀ and no change in APD₅₀ or APD₂₅, consistent with our findings (Table 1). The effects of myocardial stretch on APD may arise from the complex interactions between the SAC current and the intrinsic membrane currents (and channels) of the myocytes. Even if the G_SAC were to be constant during stretch with a constant value of E_SAC, the time dependence and polarity of the SAC current depend on the membrane potential of the myocyte. With a constant value of −10 mV for E_SAC, the SAC current would be inward during diastole, but would change in sign as the membrane potential becoming positive to −10 mV, and would then reverse again in sign as the repolarization progresses to negative values, as illustrated in Fig. 3. The large variability in the repolarization process in ventricular myocytes of different species (and in different regions of the ventricle) may significantly alter the effects of a given SAC current. In addition, the E_SAC also varies with species. Thus the role of SAC current in regulating myocyte depolarization and repolarization will be different for different species. Our SAC model provides opportunity of altering the G_SAC, thereby changing the magnitude of the SAC current. This allows us to specifically study the effects of SAC current and eliminate the interplay of other substrates resulting from the real mechanical stretch. In addition, this SAC model can also be applied to mimic SAC currents in different species by accordingly setting the reversal potential to a relative level.

APD prolongation and EADs.

The mechanisms of EADs are complex with possible contributions of decreased outward current or increased inward current during the late plateau phase and with a significant contribution of I_Ca “window current” as part of the mechanism for the upstroke of the EAD (20). The sodium-calcium exchange current may also contribute to EAD development (38). The generation of EADs has long been recognized as a potential mechanism of the generation of extrasystoles, which may lead to reentrant activity (32).

Prolongation of APD has being regarded as a major mechanism for the increased abnormal activity, such as EAD and delayed afterdepolarization, and ventricular arrhythmias in heart failure (HF) (19). However, in vivo studies showed that inhibition of Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) suppressed EADs in HF without shortening APD, indicating that APD prolongation is not a required mechanism for the induction of EADs (45). Instead, a CaMKII-related increase in I_Ca (41) is likely a mediator. Our results here showed that prolongation of APD by SAC activation itself did not produce EADs in rat ventricular myocytes in the control condition, suggesting that the APD prolongation itself may not be enough for EAD generation.

Previous studies have shown increases in I_Ca induced by acute exposure to H₂O₂ (5). At higher concentration levels (100–400 μM), H₂O₂ exposure induced EADs, in addition to a significant APD prolongation (42). It is important to note that these results were only observed with the perforated patch technique, but not with the “break-through” whole cell clamp, indicating that intracellular Ca²⁺ unbuffering is required. We have shown that the combination of stretch (created by a model SAC) and oxidative stress (approximated with low doses of H₂O₂) induced EADs in rat ventricular myocytes. Stretch alone was not sufficient to induce EADs at any value of G_SAC. The addition of H₂O₂, which lengthened the APD at all repolarization levels, sensitized the cells to EADs in the presence of stretch conductance. This may be due to the ability of H₂O₂ to increase I_Ca, Ca²⁺ load, and Ca²⁺ transient.

The G_SAC required for inducing EADs in H₂O₂ averaged 2.6 ± 0.4 nS (Table 1). For average RMP of −80 mV and E_SAC of −10 mV, this conductance generates a current of −182 pA. Kamkin et al. (21) showed that stretching an isolated rat ventricular myocyte for 8 μm induced a current of −269 pA at holding potential of −45 mV. Thus the amount of stretch-activated current required to induce EADs in rat ventricular myocytes under 100 μM H₂O₂ from our SAC model can be reached by the physical stretch of rat ventricular myocytes.

Effects of pacing frequency.

Increasing pacing frequency from 1 to 3 Hz did not significantly increase APD in either control or H₂O₂ conditions. However, addition of G_SAC, along with higher pacing rate, dramatically increased APD₉₀ in both control and H₂O₂ conditions (Table 2). This fast-pacing-induced APD prolongation under the condition of SAC activation plays a role in facilitating EAD induction in acute oxidative stress. This pointed out a mechanism for the premature excitation-triggered EADs in stretched myocardium, such as in HF and other dilated cardiomyopathies. The mechanisms underlying the frequency-dependent APD prolongation in the presence of SAC activation is unclear. It is unlikely related to the frequency-dependent inhibition of I_to, because the prolongation is presented only for the late, but not the early, repolarization. Instead, an increase in the intracellular Ca²⁺ load at fast pacing during SAC activation (10) may contribute. Nevertheless, our data suggest that the SAC activation-associated, frequency-dependent APD prolongation plays an important role in facilitating EAD generation when the myocytes are exposed to H₂O₂, where the Ca²⁺ load and the transient are significantly elevated. This implicates SAC activation in premature beat triggered EADs in the condition of oxidative stress.

Limitations.

Our use of a computed current injected into an isolated myocyte to model the effects of SAC channels has advantages and disadvantages compared with physically stretching cells. Stretch-induced arrhythmias could depend on Ca²⁺ entry through SACs or a release of Ca²⁺ from internal stores in response to the SAC-induced depolarization. Mechanical deformation of ventricular myocytes induced an increase in intracellular Ca²⁺ that was sensitive to gadolinium, which blocks SACs (35). Additionally, intracellular buffering of Ca²⁺ abolished the effects of stretch in guinea pig ventricular myocytes (3). These findings suggest that Ca²⁺ may enter through the SACs. In contrast, a recent study showed that SAC currents in human atrial myocytes did not change with Ca²⁺-free external solution, suggesting that the current through SACs was not carried by Ca²⁺, but more likely by Na⁺ (22). In fact, the SAC has been found permeable to several ions, including K⁺, Na⁺, Ba²⁺, Ca²⁺, and Cl⁻ (47). A possible interpretation is that the conductance for each of these ions may vary from species and tissue types. Although our model of SACs does not include actual exchange of ions across the cell membrane, it does reproduce changes in AP characteristics, consistent with data from physically stretched myocytes and tissue (21, 23, 48). Additionally, a similar SAC formulation was incorporated into guinea pig ventricular cell model and induced EADs in this model by SAC activation (31). Different from this study, here we used a SAC integrated into real rat ventricular myocytes rather than the model simulation, thus allowing pharmacological manipulations.

An increase in diastolic Ca²⁺ and Ca²⁺ transient amplitude has been observed in ventricular myocytes by fast pacing (2) or by exposure to H₂O₂ in a concentration-dependent manner (43). Our results show that fast pacing induced a significantly larger increase in diastolic Ca²⁺ and Ca²⁺ transient in rat ventricular myocytes in the presence of H₂O₂ than in control solution. The H₂O₂-mediated enhancement of diastolic Ca²⁺ and Ca²⁺ transient at fast rate is likely a contributor to SAC-induced abnormal impulses at high-pacing frequency in the presence of 100 μM H₂O₂. Although alterations in intracellular calcium homeostasis have been implicated in H₂O₂-induced injury, the mechanisms by which calcium homeostasis is altered remain unclear. It is possible that oxidant increases intracellular Ca²⁺ concentration by altering the activity of ion channels and/or transport proteins, either directly or through effects on other systems that modulate their activity, in addition to the mechanism of oxidant-induced cellular damage (34). Previous studies reported that H₂O₂ induces alterations of several ion channels and transporters, including Na current (42), L-type Ca current (14, 18), K currents (4), ryanodine receptors (1), sarcoplasmic reticulum Ca pump (27), and the Na/Ca exchanger (13, 16). Activation of several signaling pathways is involved in these alterations, most importantly the PKC (43) and the CaMKII (11, 46) pathways. These ionic alterations, in combination with the cellular injury caused by high-concentration H₂O₂, contribute significantly to the enhanced Ca²⁺ load, facilitating abnormal impulse induction (46). In this study, we used a low concentration of H₂O₂ (100 μM), a level that has been reported in human blood plasma (15). The changes in ionic currents, transporters, and the signaling pathways that alter intracellular calcium homeostasis by this low-level H₂O₂ have not been well documented. We did not observe EADs or automaticity triggered by this concentration of H₂O₂. EADs and sustained automaticity were induced only in the combination of H₂O₂ and SAC activation, implicating SAC activation in arrhythmogenesis under the condition of oxidative stress. However, as a limitation of this SAC model, we are unable to include stretch-induced changes in ion channels and transporters and the related signaling pathways. Furthermore, changes in cell length due to contraction were also not included. Cell shortening may decrease the open probability of SACs (26); thus our model system may overestimate the magnitude of the stretch current during systole.

Perspective.

Many forms of cardiomyopathies, including HF and cardiac infarction, where the myocardium is persistently stretched, predispose ventricular myocytes to a heightened susceptibility to EADs, an important trigger for arrhythmogenesis (28). A number of studies have highlighted the importance of SAC currents in the generation of EADs and arrhythmias (30). In this study, we showed that incorporation of a real-time simulation of SACs into an isolated rat ventricular myocyte induced EADs in the presence of low doses of H₂O₂, and that this effect is enhanced by increasing the stimulation rate. We conclude that SAC activation plays an important role in triggering EADs under oxidative stress. Furthermore, our data suggested that SAC activation is an important mediator for the premature beats-triggered (fast-pacing) abnormal impulses. These findings implicate SAC activation in promoting ventricular arrhythmias in cardiomyopathies with increased tension in ventricular wall and oxidative stress. This SAC-related mechanism may explain, at least in part, the clinical findings that reducing myocardial wall tension and oxygen demand by left ventricular offloading using intra-aortic balloon counterpulsation significantly reduced ventricular arrhythmias in patients with medically refractory ventricular arrhythmias (12).

GRANTS

This work was supported by National Heart, Lung, and Blood Institute grants (HL-088168, Y. Wang; HL-088488, M. B. Wagner; HL-077485, R. W. Joyner), American Health Assistance Foundation grant (H2007-019, Y. Wang), and the support from Emory Children's Healthcare of Atlanta (to Y. Wang) and Todd Franklin Cardiac Laboratory (to R. W. Joyner).