Small-Conductance Calcium-Activated Potassium Channel and Recurrent Ventricular Fibrillation in Failing Rabbit Ventricles

Abstract

Rationale

Fibrillation-defibrillation episodes in failing ventricles may be followed by action potential duration (APD) shortening and recurrent spontaneous ventricular fibrillation (SVF).

Objective

We hypothesized that activation of apamin-sensitive small-conductance Ca²⁺-activated K⁺ (SK) channels are responsible for the postshock APD shortening in failing ventricles.

Methods and Results

A rabbit model of tachycardia-induced heart failure was used. Simultaneous optical mapping of intracellular Ca²⁺ and membrane potential (Vm) was performed in failing and non-failing ventricles. Three failing ventricles developed SVF (SVF group), 9 did not (no-SVF group). None of the 10 non-failing ventricles developed SVF. Increased pacing rate and duration augmented the magnitude of APD shortening. Apamin (1 μmol/L) eliminated recurrent SVF, increased postshock APD₈₀ in SVF group from 126±5 ms to 153±4 ms (p<0.05), in no-SVF group from147±2 ms to 162±3 ms (p<0.05) but did not change of APD₈₀ in non-failing group. Whole cell patch-clamp studies at 36°C showed that the apamin-sensitive K⁺ current (I_KAS) density was significantly larger in the failing than in the normal ventricular epicardial myocytes, and epicardial I_KAS density is significantly higher than midmyocardial and endocardial myocytes. Steady-state Ca²⁺ response of I_KAS was leftward-shifted in the failing cells compared with the normal control cells, indicating increased Ca²⁺ sensitivity of I_KAS in failing ventricles. The K_d was 232 ± 5 nM for failing myocytes and 553 ± 78 nM for normal myocytes (p = 0.002).

Conclusions

Heart failure heterogeneously increases the sensitivity of I_KAS to intracellular Ca²⁺, leading to upregulation of I_KAS, postshock APD shortening and recurrent SVF.

Introduction

Electrical storm describes a clinical condition in which the patients suffer from recurrent spontaneous ventricular fibrillation (SVF) requiring multiple defibrillation shocks within a short period of time.¹ It occurs frequently in patients with heart failure (HF) and is an important reason for rehospitalization.^2-4 In addition, recurrent SVF occurs in roughly half of the patients undergoing cardiopulmonary resuscitation, and may be associated with significant morbidity and mortality.^{3, 4} The mechanisms of electrical storm remain unclear. We recently developed a rabbit model of electrical storm with pacing-induced HF.⁵ In that model, acute but reversible postshock action potential duration (APD) shortening induced recurrent SVF by promoting late phase 3 early afterdepolarizations (EADs) and triggered activity. However, the mechanisms underlying acute APD shortening after fibrillation-defibrillation episodes in failing ventricles remain unclear. Because VF is associated with intracellular Ca²⁺ (Ca_i) accumulation, especially in failing ventricles,^6-8 we hypothesized that apamin-sensitive small-conductance Ca²⁺-activated K⁺ (SK) channels may be in part responsible for postshock APD shortening. Apamin is a neurotoxin that selectively blocks SK channels.^{9, 10} Several studies show that apamin-sensitive K⁺ current (I_KAS) is abundantly present in cardiac atrial cells but none in normal ventricular cells.^10-16 We were not able to find any studies on I_KAS regulation in heart failure (HF) ventricles. Therefore, we studied normal and HF rabbits to test the hypotheses that I_KAS is significantly increased in HF rabbit ventricular myocytes and that I_KAS blockade prevents postshock APD shortening and recurrent SVF in Langendorff-perfused HF ventricles.

Methods

An expanded Methods section is available in the Online Data Supplement. New Zealand white rabbits (N=39) were used in the study. Among them, 5 died during rapid pacing, 22 were used for optical mapping (including 12 with pacing-induced HF, 3 sham with pacemaker implanted but not paced and 7 normal control) and 12 were used for patch clamp studies (6 HF and 6 normal control).

Optical mapping studies

We simultaneously mapped intracellular Ca²⁺ (Ca_i) and membrane potential (V_m) using Rhod-2 AM and RH237, respectively, in ventricles of Langendorff-perfused hearts. Rapid ventricular pacing and 3 to 5 ventricular fibrillation (VF)-defibrillation episodes were mapped. Apamin (1 μmol/L) was added to the perfusate for 30 min before the same protocols were repeated. We then waited for 30 min to determine if the effects of apamin can be washed out.

Patch Clamp Studies

Whole-cell patch-clamp technique was used to record apamin-sensitive K⁺ currents (I_KAS) in isolated ventricular cardiomyocytes. All experiments were performed at 36°C. To study Ca²⁺-dependence of I_KAS, various combinations of EGTA and CaCl₂ were used in the pipette solutions.

Real-time PCR

Ventricular tissues were sampled from 5 normal and 5 HF rabbits prior to cell isolation for patch-clamp studies. Real-time PCR was performed using iQ SYBR Green Supermix with iCycler (BioRad, Hercules, CA).

Statistical analyses

Statistical analyses were performed using SPSS PASW Statistics 17 software (IBM, Chicago, IL). We used Student's t-test for a comparison between two groups. One-way ANOVA with post hoc Bonferroni tests were performed for comparisons among three or more groups. A p≤0.05 was considered as statistically significant.

Results

Evidence of HF

All rabbits that survived the rapid pacing protocol showed clinical signs of HF, including appetite loss, tachypnea, lethargy, pleural effusion, ascites, and visible congestion of lung, liver, and gastrointestinal tract. Echocardiograms of sham-operated rabbits (N=3) at baseline and at second surgery showed no changes of left ventricular (LV) end-diastolic dimension (12.5±0.9 mm vs 13.0±0.6 mm), end-systolic dimension (6.5±1.5 mm vs 6.9±1.7 mm), or fractional shortening (0.49±0.08 vs 0.48±0.11) (p=NS for all comparisons). In contrast, HF rabbits (N=13) showed significant increases in end-diastolic dimension (14.5±0.2 mm vs 19.9±0.5 mm) and in end-systolic dimension (8.4±0.3 mm vs 17.9±0.3 mm) and reduced fractional shortening (0.42±0.01 vs 0.10±0.01) (P<0.05 for all comparisons). The percent LV fibrosis was 5 %±1 % for normal (N=3) or sham-operated hearts and 21 %±6 % for failing hearts (N=5) (P<0.05). Action potential duration measured at 80% repolarization (APD₈₀) with pacing cycle length (PCL) of 300 ms were not different between sham-operated and normal rabbits (166±5 ms vs 168±2 ms, p>0.05). Therefore, these data were combined into a single non-failing group. There were no significant differences in RR intervals during the sinus rhythm between non-failing and failing hearts (464±38 ms vs 509±42 ms, p=NS). Figure 1A shows marked cardiac enlargement (right panel) typical for all failing hearts. Action potential duration (APD) prolongation was present in all failing ventricles. Figure 1B shows APD₈₀ at 300 ms PCL recorded from the site marked by an asterisk in APD₈₀ maps in Figure 1C. Overall, the failing ventricles (N=7) had a longer APD₈₀ than non-failing ventricles (N=7) (181±5 ms vs 167±2 ms, p<0.05, Figure 1D). The Ca_i transient duration at 80% repolarization (Ca_iTD₈₀) was 184±5 ms for failing ventricles and 176±3 ms for non-failing ventricles (p=NS).

Figure 1. Action potential duration (APD) and intracellular Ca transient duration (Ca_iTD) in non-failing and failing ventricles.

Measurements were made with pacing cycle length (PCL) of 300 ms in 7 non-failing (including 4 normal and 3 sham-operated ventricles) and 7 failing ventricles. A, Non-failing and failing hearts. B, The black and red lines indicate optical tracings of membrane potential (V_m) and intracellular calcium (Ca_i), respectively. The optical tracings were recorded from the site labeled by an asterisk in APD₈₀ map and Ca_iTD₈₀ map in C. C, APD₈₀ map (upper panels) and Ca_iTD₈₀ map (lower panels) obtained from normal (left panels) and failing ventricles (right panels). Averages of APD₈₀ and Ca_iTD₈₀ in the study were measured from all pixels in the map, excluding the atrial signals and the pixels at the edge of ventricles. D, Average of APD₈₀ and Ca_iTD₈₀ at fixed PCL of 300 ms in 7 non-failing and 7 failing ventricles. Bar graphs represent mean±SEM; APD₈₀ and Ca_iTD₈₀ indicate APD and Ca_i transient duration, respectively, measured at 80% repolarization; HF, heart failure; no-HF, non-failing heart; LAD, left anterior descending artery; LV, left ventricle; RV, right ventricle.

Effects of Apamin on Acute APD Shortening and Postshock SVF

After fibrillation-defibrillation episodes, acute but transient shortening of APD occurred in all 12 failing ventricles. Three of them developed SVF in the postshock period (the “SVF-Group”). Figure 2A, left panel, shows V_m (black line) and Ca_i (red line) optical signals recorded during one of the SVF episodes. There were 2 postshock beats (1 and 2) before VF termination. Optical maps confirmed complete cessation of wavefronts after beat 2, consistent with type B successful defibrillation.¹⁷ However, there was a large Ca_i transient and short APD (beat 3). The first beat of SVF (beat 4) began during late phase 3 of the preceding sinus beat. In Figure 2A(b), an isochronal map illustrates that beat 4 originated from the basal portion of the RV. The right two columns are ratio maps of V_m and Ca_i, respectively, of the same beat. The Ca_i remained elevated throughout the mapped field while Vm had already repolarized. The focal beat originated from the left upper quadrant during persistent Ca_i elevation, consistent with the late phase 3 EAD mechanism.^{5, 18} Five SVF episodes were recorded after initial successful defibrillation in this failing heart. Apamin prevented both the postshock APD shortening and SVF (Figure 2B). We repeated fibrillation-defibrillation episodes four times, but no SVF was observed in the postshock period after apamin. Among the 12 failing ventricles studied, 3 developed a total of 11 episodes of SVF 104±23 sec after initial successful defibrillation. Apamin administration completely prevented SVF after 3-5 repeated fibrillation-defibrillation episodes. Figure 3A shows additional optical traces and APD₈₀ map during baseline sinus rhythm (SRm) and during a fibrillation-defibrillation episode in a ventricle with postshock SVF. Upper panels illustrate optical traces of sinus rhythm and in the postshock period in the absence of apamin. Acute APD shortening occurred after defibrillation while Ca_i remained elevated after repolarization (beats 1 and 2). The difference between Ca_iTD₈₀ and APD₈₀ is shown in the Ca_iTD₈₀-APD₈₀ map. A large difference was present between Ca_iTD₈₀ and APD₈₀ in beats 1 and 2. After pretreatment with apamin, beats 1 and 2 after defibrillation had less APD shortening as compared with baseline, associated with smaller differences between Ca_iTD₈₀ and APD₈₀ (Figure 3B). The pseudoECG tracings in panels A(b) and B(b) show recurrent SVF in the absence of apamin (A) but not in the presence of apamin (B). We were not able to washout apamin effects in any of the 12 failing ventricles.

Figure 2. Pretreatment with apamin prevents refibrillation.

The black and red lines indicate optical tracings for V_m and Ca_i, respectively, obtained from the endocardial surface in a failing ventricle. A(a), Before apamin, a defibrillation shock (arrow) was followed by two postshock beats (1 and 2). Optical maps (not shown) confirmed complete cessation of wavefronts after beat 2, consistent with type-B successful defibrillation. After a short pause, there was a large Ca_i transient (red dot), short APD (beat 3) and first beat (beat 4) of SVF arising from late phase 3. A(b), Snapshots of Ca_i and V_m ratio maps at times from 10 ms before to 20 ms after the onset of beat 4. Note Ca_i remained elevated throughout the mapped field while V_m has already repolarized. The beat 4 arose during persistently high Ca_i and initiated the SVF. B, After apamin (1 μM) infusion, the postshock beats (5 and 6) had longer APD than the beats 1-4 in A. The APD and Ca_iTD were approximately the same, and SVF episodes were completely prevented. PM, papillary muscle; S, interventricular septum; RV, right ventricle.

Figure 3. Effects of apamin (1 μM) on ΔAPD₈₀ and the differences between Ca_iTD₈₀ and APD₈₀ in SVF Group.

Optical traces (upper panels) of V_m (black line) and Ca_i (red line) were recorded from site marked by an asterisk in APD₈₀ map (lower panels) in a second heart of the SVF Group. A(a), Epicardial optical traces of V_m and Ca_i and APD₈₀ map during sinus rhythm before pacing-induced VF (left panels). Right upper panel shows beats 1 and 2 had acute shortening of APD in the immediate postshock period, resulting in the Ca_i elevation during late phase 3 and phase 4. Lower panel shows the corresponding APD₈₀ maps, and the maps of the difference between Ca_iTD₈₀ and APD₈₀ in beats 1 and 2. A(b), pseudoECG of the same heart, showing a SVF episode 74 sec after initial successful defibrillation. B(a), After apamin, the postshock beats 1 and 2 had longer APD₈₀ than those before apamin (A(a)). The Ca_iTD₈₀ was similar to that in A, and the differences between Ca_iTD₈₀ and APD₈₀ in beats 1 and 2 were reduced. B(b), pseudoECG show the absence of SVF after initial successful defibrillation. Arrow indicates defibrillation; PI VF, pacing-induced VF; SRm, Sinus rhythm.

In four of the 12 failing hearts, we also studied different doses of apamin on postshock APD₈₀. The results show that apamin concentration of 0.1, 0.3 and 1.0 μmol/L prolonged postshock APD₈₀ to similar duration while 0.01 μmol/L apamin had little effects on postshock APD₈₀ (online Figure I).

Effects of Apamin on Baseline APD

Examples of APD₈₀ maps (bottom panel) and corresponding optical traces of action potentials from a single representative pixel (top panel) are presented in Figure 4A. Apamin induced more APD₈₀ prolongation in failing than in non-failing ventricles at PCL of 300 ms. The APD prolongation predominantly occurred in the terminal portion of repolarization. Figure 4B shows the magnitude of APD₈₀ prolongation map obtained from the same non-failing and failing ventricles presented in figure 4A. At PCL of 300 ms, the percentage of APD prolongation with apamin was 5.4±0.7 % for failing ventricles compared with 1.4±0.6 % in non-failing ventricles (p<0.05) (Figure 4C).

Figure 4. Effect of apamin in non-failing and failing ventricles.

Measurements of APD₈₀ were made with PCL of 300 ms. A, The black and blue lines (upper panels) indicate the optical V_m tracing before and after application of apamin, respectively. The tracings were obtained from the site labeled by an asterisk in the APD₈₀ map (lower panels). B, Examples of ΔAPD₈₀ map after apamin in non-failing and failing ventricles. C, Magnitude of APD₈₀ prolongation before and after apamin in 7 non-failing and 7 failing ventricles. Bar graphs represent means±SEM.

Effects of Apamin on APD After Rapid Pacing

We used rapid pacing to increase Ca_i. After abrupt cessation of rapid ventricular pacing, APD shortening (ΔAPD₈₀) was greater in failing than in non-failing ventricles. The dependence of ΔAPD₈₀ on the PCL (Pacing Protocol I) and pacing duration (Pacing Protocol II) are presented in Figure 5 and online Figure II, respectively. The shorter PCL and longer pacing duration elicited more APD shortening, and apamin reduced ΔAPD₈₀ after these rapid pacing protocols in failing ventricles. Figure 5A shows the effects of Pacing protocol I on ΔAPD₈₀ in a non-failing ventricle at baseline (control) and after apamin administration. Figure 5B shows that apamin did not significantly change ΔAPD₈₀ in non-failing ventricles. Figure 5C shows an example obtained from a failing ventricle with Pacing Protocol I. In this example, apamin reduced ΔAPD₈₀ from 23 ms to 8 ms. Figure 3D summarizes the effects of apamin in failing ventricles. Apamin significantly reduced ΔAPD₈₀ in failing but not in non-failing ventricles. Similar results were noted for the dependence of ΔAPD₈₀ on pacing duration (pacing protocol II, online Figure II).

Figure 5. APD after cessation of rapid pacing.

In A and C, the first and 3^rd row show action potential at fixed-rate pacing at 300 ms PCL. The 2^nd and 4^th rows show action potentials during rapid pacing followed by a post-pacing beat. The preceding cycle length of the first post-pacing beat was controlled by an S₂ given 300 ms after the last S₁. The APD of the first post-pacing beat was measured to 80% repolarization (APD₈₀). The same pacing protocol was performed at baseline and after apamin (1 μM). B, The effect of apamin on ΔAPD₈₀ in all non-failing ventricles studied. D, The effects of apamin in all failing ventricles studied. Bar graphs represent means±SEM, “1:1” indicates the shortest PCL associated with 1:1 capture. * p<0.05 vs control.

Effects of Apamin on Postshock APD in no-SVF Group

In 9 of 12 failing ventricles, SVF did not occur after fibrillation-defibrillation episodes (the “no-SVF” Group).⁵ Similar to the SVF Group, acute APD shortening was also observed in this group after a fibrillation-defibrillation episode. Figure 6A shows the optical tracing and corresponding APD₈₀ maps in sinus rhythm (left panels) and after defibrillation (right panels) in a failing ventricle without SVF. Similarly, apamin prolonged the APD immediately after defibrillation shock in failing ventricles that did not develop SVF (Figure 6A(b)). Note the difference between Ca_iTD₈₀ and APD₈₀ was reduced by apamin. In contrast, no acute shortening of APD was observed immediately after a successful defibrillation shock in non-failing ventricles. Pretreatment with apamin did not result in significant APD changes in non-failing ventricles (Figure 6B).

Figure 6. Effects of apamin on postshock APD in no-SVF Group and non-failing hearts.

A (a), Left panels shows baseline Vm (black line) and Ca_i (red line) recorded from site marked by an asterisk in APD₈₀ map during baseline sinus rhythm. Right upper panel shows postshock beats 1 and 2. Right lower panel shows the APD₈₀ maps, and the differences between Ca_iTD₈₀ and APD₈₀ maps in beats 1 and 2. B, An example of non-failing heart. As compared with Control (B(a)), there were little changes of APD and Ca_iTD after defibrillation when the tissues were pretreated with apamin (B(b)). SRm, Sinus rhythm.

Postshock APD Changes of All Ventricles Studied

We plotted APD₈₀ of the first two postshock beats against the preceding diastolic interval (DI) for all episodes. Apamin did not change postshock APDs in non-failing ventricles (figure 7A), but prolonged postshock APDs significantly in failing ventricles of both the no-SVF Group (Figure 7B) and in the SVF Group (Figure 7C). We performed statistical analyses with repeated-measure ANOVA to compare the ΔAPD₈₀ before and after apamin. The ΔAPD₈₀ was 5.8 ms (-2.2 ms to 13.8 ms; p=NS) for non-failing ventricles, indicating that apamin did not significantly change the postshock APD. For no-SVF and SVF groups of failing ventricles, the ΔAPD₈₀ was 15.7 ms (9.4 ms to 21.9 ms, p<0.001) and 27.4 ms (14.3 ms to 40.4 ms, p<0.001), respectively, showing that postshock APD was lengthened significantly by apamin.

Figure 7. Effects of apamin in failing ventricles.

A through C, Postshock APD₈₀ versus diastolic interval in non-failing ventricles (A), failing ventricles without SVF (B) and with SVF (C) of all hearts studied. The postshock APD₈₀ without apamin (unfilled circles) are shorter than with apamin (filled circles) in failing ventricles (B and C) but not in non-failing ventricles (A). D and E, Bar graphs show postshock APD₈₀ and the difference between postshock APD₈₀ and Ca_iTD₈₀ with and without apamin in non-failing, no-SVF and SVF groups. Bar graphs represent means±SEM. * p<0.05 vs control; § p<0.05 vs no-HF group; ‡ p<0.05 vs no-HF and no-SVF group.

We investigated the postshock APD₈₀ and the differences between Ca_iTD₈₀ and APD₈₀ in all ventricles studied. The results are presented in Figure 7D and 7E, respectively. There were no significant differences in VF duration before versus after apamin in non-failing ventricles (324±14 sec vs 338±16 sec, p=NS), no-SVF (353±33 sec vs 352±44 sec, p=NS), and SVF (336±43 sec vs 362±33 sec, p=NS) groups. Before apamin, the postshock APD₈₀ shortened to 147±2 ms, 126±5 ms and 156±2 ms (p<0.05) in no-SVF group, SVF group and non-failing ventricles, respectively. Apamin minimized acute transient postshock APD₈₀ shortening in failing ventricles but not in non-failing ventricles. Specifically, apamin increased postshock APD₈₀ in no-SVF group from147±2 ms to 162±3 ms (p<0.05) and in SVF group from 126±5 ms to 153±4 ms (p<0.05). There was no change of APD₈₀ in non-failing group before (156±2 ms) and after apamin (162±4 ms, p=NS). The differences between Ca_iTD₈₀ and APD₈₀ were also decreased with apamin in no-SVF (36±2 ms vs 22±3 ms, p<0.05) and SVF (77±8 ms vs 41±4 ms, p<0.05) groups but not in non-failing group (27±2 ms vs 25±3 ms).

Effects of Glibenclamide on ΔAPD₈₀ in Failing Ventricles

To determine if ATP-sensitive potassium current (I_KATP) activation was responsible for the APD shortening in failing ventricles, we examined the effect glibenclamide (10 μmol/L), which mainly inhibit I_KATP channel in this range of concentrations,¹⁹ in 3 failing ventricles. As shown in online Figure III, glibenclamide did not significantly affect the magnitude of post-rapid pacing and postshock APD shortening compared with control.

I_KAS in Failing Ventricles

The density and properties of I_KAS were examined in cardiomyocytes isolated from normal and failing rabbit left ventricles using the voltage-clamp technique in whole-cell mode. Figure 8A shows representative current traces obtained with a step-pulse protocol (300 ms pulse duration; holding potential, -50 mV; see inset) in the absence and presence of 100 nM apamin in the bath solution. Mean I_KAS density (determined as the apamin-sensitive difference current) was significantly larger in failing than in normal ventricular epicardial myocytes (I_KAS density at 0 mV with an intrapipette free-Ca²⁺ of 863 nM: 8.39 ± 1.00 pA/pF, n = 6 cells from 5 failing rabbits, vs. 2.83 ± 0.87 pA/pF, n = 6 cells from 4 normal rabbits, p < 0.01). In contrast, when the intrapipette free-Ca²⁺ was buffered to 100 nM, the apamin-sensitive current was much smaller, with no significant difference between normal and failing ventricular myocytes (Fig. 8D). Figure 8B illustrates the I_KAS-voltage (I-V) relationships. Transmural distribution of I_KAS in failing ventricles was studied using cardiomyocytes isolated from three layers (6 epicardial cells, 5 midmyocardial cells, and 7 endocardial cells from 5 animals). Mean I_KAS density in epicardial myocytes (8.52±2.47 pA/pF, range 5.21-11.46 pA/pF) was significantly larger than midmyocardial cells (2.18±2.14 pA/pF, range 0.00-5.70 pA/pF) and endocardial cells (1.87±2.36 pA/pF, range 0.00-6.04 pA/pF) (Figure 8C). To further elucidate the mechanisms underlying I_KAS upregulation, Ca²⁺-dependence of I_KAS was studied in epicardial cells using pipette solutions containing increasing intracellular free Ca²⁺ concentrations. Figure 8D demonstrates that the steady-state Ca²⁺ response of I_KAS was leftward-shifted in the failing cells compared with the normal cells. The data were fitted with the Hill equation, yielding K_d of 232 ± 5 nM for failing and 553 ± 78 nM (p=0.002) for normal cells, and Hill coefficients of 2.38 ± 0.13 for failing and 1.50 ± 0.30 for normal cells (p = 0.01). These results are compatible with the notion that the K⁺ channels carrying I_KAS have increased sensitivity to cytosolic Ca²⁺ in failing ventricles.

Figure 8. Apamin-sensitive currents (I_KAS) in failing rabbit ventricles.

(A) Representative K⁺ current traces obtained from normal (upper panel) and failing (lower panel) ventricular myocytes. Voltage-pulse protocol is shown in the inset. “Baseline” shows current traces in the absence of apamin (I_baseline); “Apamin 100 nM” shows currents in the presence of 100 nM apamin (I_apamin); “Difference” shows I_KAS calculated as I_baseline – I_apamin. (B) I_KAS-voltage (I-V) relationship obtained from normal (N=6) and failing (N=6) ventricular epicardial cells. (C) I_KAS density at 0 mV with an intrapipette free-Ca²⁺ of 863 nM recorded from epicardial (Epi), midcardial (Mid), and endocardial (Endo) cells from 5 failing ventricles. *p<0.05. (D) Steady-state Ca²⁺ response of I_KAS in normal and failing epicardial ventricular myocytes. The data were fitted with the Hill equation: y = 1/[1 + (K_d/X)ⁿ], where y represents the I_KAS normalized to the currents with 10 μM intrapipette Ca²⁺; x is the concentration of Ca²⁺; K_d is the concentration at half-maximal activation; and n is the Hill coefficient. Error bars represent SEM. Numbers in parentheses indicate the number of cells patched.

Quantitative Polymerase Chain Reaction

The SK2 mRNA levels (normalized to GAPDH) in normal (N=5) and failing (N=5) ventricles were 100±16.4% and 118±8.5%, respectively (p=0.39).

Discussion

The primary finding of this study is that heart failure heterogeneously increases the sensitivity of I_KAS to intracellular Ca²⁺, leading to upregulation of I_KAS, postshock APD shortening and recurrent SVF.

I_KAS in the Normal and Failing Ventricles

The primary function SK channels in the nervous system is to produce afterhyperpolarization following a neural action potential and to protect the cell from the deleterious effects of continuous tetanic activity.²⁰ Seminal studies from multiple laboratories show that SK channels are also expressed in cardiac cells, more so in the atria than in the ventricles.^10-16 These findings suggest that specific ligands for SK currents may offer a unique therapeutic opportunity to directly modify atrial cells without interfering with ventricular myocytes.¹² Our study confirmed that I_KAS plays little role in APD regulation in normal ventricles. However, the same is not true for failing ventricles in which I_KAS is upregulated in epicardial cells. Previous studies showed that multiple ventricular K⁺ currents (I_to, I_K1, I_Kr and I_Ks) are downregulated in animal models of HF, contributing to the reduced repolarization reserve.²¹ In addition, ATP-sensitive K⁺ currents (I_KATP) are also disrupted in HF. The I_KATP, which shortens APD and protects the heart from Ca_i overload, is essential in maintaining the homeostasis during the metabolically demanding adaptive response to stress.²² In a model of HF induced by transgenic expression of the cytokine tumor necrosis factor alpha, structural remodeling in failing ventricles led to disruption of energetic signal-channel communication, resulting in disturbing the I_KATP-mediated protection.²³ Another study²⁴ shows that I_KATP channel activation under metabolic stress was impaired in myocytes from rat hypertrophied ventricle. Our study found that suppression of I_KATP did not reduce post-pacing or postshock APD shortening in failing ventricles, but apamin significantly reduced the magnitude of post-pacing or postshock APD shortening. Taken together, these findings suggest that because of I_KATP channel malfunction and downregulation of multiple other K⁺ currents, the failing ventricles may rely on I_KAS to shorten the APD in situations of Ca_i overload.

Late Phase 3 EAD and Ventricular Arrhythmogenesis

When heart rate increases, APD shortens physiologically to preserve a sufficient diastolic interval for ventricular filling and coronary flow. However, excessive APD shortening might be arrhythmogenic by promoting late phase 3 EAD¹⁸ or by inducing transmural heterogeneity of repolarization.²⁵ Because of SERCA downregulation²⁶ and decreased driving force of Na⁺/Ca²⁺ exchange current,²⁷ myocytes in failing hearts have higher diastolic Ca_i concentration, slower Ca_i transient decay, and more Ca_i accumulation during burst pacing than those in normal hearts. Therefore, during rapid pacing or fibrillation in failing ventricles, the rapid depolarization may increase Ca_i which in turn activates I_KAS and shortens APD in a spatially heterogeneous fashion. Marked APD shortening after fibrillation-defibrillation episodes with persistent Ca_i elevation during the late phase 3 of action potential results in the development of late phase 3 early afterdepolarization and recurrent SVF in failing hearts.⁵

Heterogeneous Remodeling of I_KAS

The upregulation of I_KAS during HF is highly heterogeneous. Although epicardial cells in average gained much more I_KAS density than midmyocardial and endocardial cells during HF, large variations of I_KAS density is present within the same layer of cells. Previous studies showed that strong intercellular coupling can reduce the differences of APD between different layers of cells, making M cells “invisible” in intact human ventricles.²⁸ Similarly, the highly heterogeneous distribution of I_KAS density and the tight intercellular coupling of intact ventricles allowed cells with high I_KAS density and shortened APD to influence the APD of the neighboring cells with low I_KAS density. Therefore, optical mapping studies showed that postshock APD shortening occurs both in the epicardium and in the endocardium, and that apamin prevented postshock APD shortening in both layers of the ventricles.

Clinical Implications

Recurrent SVF (electrical storm) is a frequent complication in patients with advanced HF.² A recent genome wide association study²⁹ showed that variants at the 1q21 locus cluster at KCNN3 (SK3) is associated with atrial fibrillation. The knowledge that I_KAS is upregulated in HF should direct broader studies into the role of such currents in determining heart disease risk and as possible targets for preventive therapy in both atrial and ventricular arrhythmias.

Study Limitations

Apamin is a neural toxin and may induce significant neurological side effects when used in live animals and in humans. Therefore, it is necessary to develop non-toxic blockers of I_KAS before we can test the effects of I_KAS blockade on VF storm in human patients. Another limitation is that apamin is a potent inhibitor of the L-type Ca current (I_Ca,L).³⁰ While I_Ca,L blockade cannot be used to explain APD lengthening after fibrillation-defibrillation episodes, this pharmacological effect could help prevent SVF. Therefore, whether or not apamin prevented SVF by I_KAS blockade or by a combined effects of I_KAS and I_Ca,L blockade remain unclear.

Non-standard Abbreviations and Acronyms

APD

action potential duration

Ca_i

intracellular calcium

Ca_iTD

intracellular calcium transient duration

EADs

early afterdepolarizations

HF

heart failure

I_KAS

apamin-sensitive potassium current

I_KATP

ATP-sensitive potassium current

LV

left ventricle

PCL

pacing cycle length

SK

small-conductance Ca²⁺-activated K⁺ channels

SRm

sinus rhythm

SVF

spontaneous ventricular fibrillation

Vm

membrane potential

VF

ventricular fibrillation