Phospholamban Knockout Breaks Arrhythmogenic Ca²⁺ Waves and Suppresses Catecholaminergic Polymorphic Ventricular Tachycardia in Mice

Abstract

Rationale

Phospholamban (PLN) is an inhibitor of cardiac sarco(endo)plasmic reticulum Ca²⁺ ATPase (SERCA2a). PLN knockout (PLN-KO) enhances sarcoplasmic reticulum (SR) Ca²⁺ load and Ca²⁺ leak. Conversely, PLN-KO accelerates Ca²⁺ sequestration and aborts arrhythmogenic spontaneous Ca²⁺ waves (SCWs). An important question is whether these seemingly paradoxical effects of PLN-KO exacerbate or protect against Ca²⁺-triggered arrhythmias.

Objective

We investigate the impact of PLN-KO on SCWs, triggered activities, and stress-induced ventricular tachyarrhythmias (VTs) in a mouse model of cardiac ryanodine receptor (RyR2)-linked catecholaminergic polymorphic ventricular tachycardia (CPVT).

Methods and Results

We generated a PLN-deficient, RyR2 mutant mouse model (PLN^−/−/RyR2-R4496C^+/−) by crossbreeding PLN-KO mice with CPVT-associated RyR2-R4496C mutant mice. Ca²⁺ imaging and patch-clamp recording revealed cell-wide propagating SCWs and triggered activities in RyR2-R4496C^+/− ventricular myocytes during SR Ca²⁺ overload. PLN-KO fragmented these cell-wide SCWs into mini-waves and Ca²⁺ sparks, and suppressed triggered activities evoked by SR Ca²⁺ overload. Importantly, these effects of PLN-KO were reverted by partially inhibiting SERCA2a with 2,5-Di-tert-butylhydroquinone (tBHQ). However, Bay K, caffeine, or Li⁺ failed to convert mini-waves to cell-wide SCWs in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes. Furthermore, ECG analysis showed that PLN-KO mice are not susceptible to stress-induced VTs. On the contrary, PLN-KO protected RyR2-R4496C mutant mice from stress-induced VTs.

Conclusions

Our results demonstrate that despite severe SR Ca²⁺ leak, PLN-KO suppresses triggered activities and stress-induced VTs in a mouse model of CPVT. These data suggest that breaking up cell-wide propagating SCWs by enhancing Ca²⁺ sequestration represents an effective approach for suppressing Ca²⁺-triggered arrhythmias.

INTRODUCTION

In the heart excitation-contraction coupling is mediated by a mechanism known as Ca²⁺-induced Ca²⁺ release (CICR)^1–3. In this process, membrane depolarization activates the voltage-dependent L-type Ca²⁺ channel (LTCC), resulting in a small influx of external Ca²⁺ into the cytosol. This Ca²⁺ then binds to the cardiac Ca²⁺ release channel/ryanodine receptor (RyR2) and opens the channel, leading to a large release of Ca²⁺ from the sarcoplasmic reticulum (SR). In addition to CICR, it has long been known that SR Ca²⁺ release can occur spontaneously under conditions of SR Ca²⁺ overload in the absence of membrane depolarizations^4–9. A number of conditions, including excessive beta-adrenergic stimulation, Na⁺ overload, elevated extracellular Ca²⁺ concentrations, and fast pacing can result in SR Ca²⁺ overload which, in turn, can trigger spontaneous SR Ca²⁺ release in the form of propagating Ca²⁺ waves^4–9. It has also long been recognized that these spontaneous Ca²⁺ waves (SCWs) can alter membrane potential via activation of the electrogenic Na⁺/Ca²⁺ exchanger (NCX), leading to delayed afterdepolarizations (DADs), triggered activities, and triggered arrhythmias^{8, 10–12}. In fact, SCW-evoked DADs are a major cause of ventricular tachyarrhythmias (VTs) in heart failure^12–14. SCW-evoked DADs also underlie the cause of catecholaminergic polymorphic ventricular tachycardia (CPVT) associated with mutations in RyR2 and cardiac calsequestrin (CASQ2)¹⁵. CPVT-causing RyR2 or CASQ2 mutations have been shown to enhance the propensity for SCWs and DADs¹⁵. Given their critical role in arrhythmogenesis, suppressing SCWs represents a promising therapeutic strategy for the treatment of Ca²⁺-triggered arrhythmias.

Since RyR2 mediates SCWs, inhibiting the RyR2 channel would be effective in suppressing SCWs. Indeed, reducing the RyR2 activity by tetracaine has been shown to inhibit spontaneous Ca²⁺ release¹⁶. Further, it has recently been shown that flecainide, a Na⁺ channel blocker, suppresses SCWs in cardiac cells and CPVT in both mice and humans by modifying the gating of the RyR2 channel^17–19. Flecainide reduces the duration and increases the frequency of openings of the RyR2 channel. Similarly, we have recently shown that carvedilol, a non-selective beta-blocker, also reduces the duration and increases the frequency of RyR2 openings, and suppresses SCWs and CPVT in mice²⁰. Interestingly, by modifying the gating of RyR2, flecainide increases the frequency and reduces the mass of Ca²⁺ sparks without affecting the SR Ca²⁺ content¹⁸. These actions of flecainide effectively break up cell-wide propagating SCWs into non-propagating spontaneous Ca²⁺ release events (mini-waves or Ca²⁺ sparks)^{18, 19}. These observations have led to the suggestion that breaking up SCWs by modifying RyR2 gating represents an effective approach to suppressing SCW-evoked DADs and triggered arrhythmia¹⁹.

The sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA2a) in the heart also plays a critical role in determining the initiation and propagation of SCWs^21–25. It has been shown that increasing the activity of SERCA2a by removing phospholamban (PLN), an inhibitor of SERCA2a, elevated SR Ca²⁺ load and markedly enhanced the frequency and amplitude of Ca²⁺ sparks^26–28. Interestingly, despite severe SR Ca²⁺ leak, no spontaneous cardiac arrhythmias in PLN knockout (PLN-KO) mice have been reported. Further, cell-wide propagating SCWs were hardly observed or frequently aborted in PLN-KO cardiomyocytes²⁹. These observations raise an important question as to whether accelerating SR Ca²⁺ uptake by removing PLN is pro-arrhythmic or anti-arrhythmic. On one hand, PLN-KO elevates SR Ca²⁺ content and increases SR Ca²⁺ leak, which would enhance the propensity for Ca²⁺ leak-induced DADs. On the other hand, PLN-KO aborts SCWs, which would suppress SCW-induced DADs and triggered activities. To address this seemingly paradoxical question, we employed the PLN-KO mice along with the CPVT RyR2-R4496C mutant mice that are prone to SCWs and DAD-evoked VTs^{20, 26}. We examined the impact of PLN-KO on the spatial and temporal properties of SCWs and the occurrence of triggered activities in ventricular myocytes expressing the RyR2-R4496C mutant. We also determined the effect of PLN-KO on the susceptibility to stress-induced VTs in the CPVT RyR2-R4496C mutant mice. We found that the removal of PLN breaks SCWs and suppresses triggered activities in the RyR2-R4496C mutant ventricular myocytes, and diminishes stress-induced VTs in the RyR2-R4496C mutant mice. These data are consistent with the notion that breaking up propagating SCWs by accelerating SR Ca²⁺ uptake is effective in suppressing Ca²⁺-triggered arrhythmias.

METHODS

To determine whether removal of PLN alters the spatial and temporal profiles of intracellular Ca²⁺ signalling in RyR2 R4496C^+/− mutant ventricular myocytes, we crossbred the RyR2-R4496C^+/− mutant mice with the PLN knockout (PLN-KO) mice (PLN^−/−) to produce a PLN deficient mouse line expressing the RyR2 R4496C^+/− mutation (PLN^−/−/RyR2-R4496C^+/−). Detailed methods are provided in the Online Supplement.

RESULTS

PLN-KO breaks cell-wide propagating spontaneous Ca²⁺ waves in isolated RyR2-R4496C^+/− mutant ventricular myocytes

It is well known that cardiomyocytes display spontaneous Ca²⁺ waves (SCWs) propagating throughout the entire cell under the conditions of SR Ca²⁺ overload^4–8. Interestingly, PLN-KO markedly alters the pattern of spontaneous Ca²⁺ release by breaking up the cell-wide propagating SCWs into multiple, localized mini-waves and sparks²⁹. To determine whether PLN-KO is also able to break up cell-wide propagating SCWs in ventricular myocytes harbouring a CPVT-causing RyR2 mutation R4496C that is prone to spontaneous Ca²⁺ release, we crossbred the PLN-KO mice (PLN^−/−) with the RyR2-R4496C mutant heterozygous mice (RyR2-R4496C^+/−) to generate double mutant mice, PLN^−/−/RyR2-R4496C^+/−. Ventricular myocytes were isolated from the RyR2-R4496C^+/− and PLN^−/−/RyR2-R4496C^+/− mice, loaded with fluo-4 AM, and perfused with elevated extracellular Ca²⁺ (6 mM) to induce SR Ca²⁺ overload and SCWs. Intracellular Ca²⁺ dynamics were monitored using line-scan confocal Ca²⁺ imaging. As shown in Fig.1A, SCWs in RyR2-R4496C^+/− ventricular myocytes originated from the middle (or either end) of the cell and propagated across the entire cell, similar to those reported previously^{20, 30–32}. On the other hand, SCWs in the PLN^−/−/RyR2-R4496C^+/− ventricular myocytes frequently and simultaneously occurred at multiple sites and aborted shortly after their initiation without propagating across the entire cell. They appeared as short-lived mini-waves or clusters of Ca²⁺ sparks (Fig. 1B). Similar spontaneous Ca²⁺ release events were also detected in ventricular myocytes from PLN^−/− mouse hearts (Fig. 1C), consistent with those shown previously²⁹. Further, this impact of PLN-KO was not limited to SCWs induced by elevated external Ca²⁺. We found that PLN-KO also breaks SCWs induced by isoproterenol (Online Fig. I). Taken together, these observations indicate that PLN-KO is able to break up cell-wide SCWs in the RyR2-R4496C^+/− mutant ventricular myocytes.

Figure 1. PLN-KO breaks cell-wide propagating spontaneous Ca²⁺ waves in isolated ventricular myocytes.

Ventricular myocytes were isolated from RyR2-R4496C^+/− mutant mice, the PLN deficient, RyR2-R4496C^+/− mice (PLN^−/−/RyR2-R4496C^+/−) or the PLN^−/− mice, and loaded with the fluorescent Ca²⁺ indicator dye fluo-4, AM. The fluo-4 loaded cells were perfused with KRH buffer containing 6 mM extracellular Ca²⁺ to induce SR Ca²⁺ overload. Store overload induced spontaneous SR Ca²⁺ release events were detected by line-scan confocal Ca²⁺ imaging.

Representative line-scan images of spontaneous Ca²⁺ release in isolated RyR2-R4496C^+/− (n=39) (A), PLN^−/−/RyR2-R4496C^+/− (n=43) (B), and PLN^−/− (n=9) (C) ventricular myocytes are shown.

PLN-KO fragments cell-wide propagating SCWs in ventricular myocytes in intact RyR2-R4496C^+/− hearts

The markedly altered spatial and temporal profiles of intracellular Ca²⁺ dynamics in PLN^−/−/RyR2-R4496C^+/− or PLN^−/− ventricular myocytes may have resulted from cellular damage during cell isolation. To avoid this potential problem, we carried out line-scan confocal Ca²⁺ imaging of epicardial ventricular myocytes in intact hearts³³. Rhod-2 AM loaded hearts from the RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/−, and PLN^−/− mice were Langendorff-perfused with elevated extracellular Ca²⁺ (6 mM) and paced at 6 Hz to induce SR Ca²⁺ overload and subsequent SCWs. As seen in Fig. 2A (top panel), after interruption of electrical pacing, SCWs occurred at 1 or 2 sites and propagated throughout the entire cell in ventricular myocytes in intact RyR2-R4496C^+/− hearts. Analysis of the spatially averaged fluorescence revealed well-separated spontaneous Ca²⁺ release events with amplitudes similar to that of stimulated Ca²⁺ transients (Fig. 2A, bottom panel). On the other hand, spontaneous Ca²⁺ release in ventricular myocytes in intact PLN^−/−/RyR2-R4496C^+/− (Fig. 2B, top panel) or PLN^−/− (Online Fig. II, top panel) hearts frequently occurred at multiple sites as mini-waves or clusters of Ca²⁺ sparks. Analysis of spatially averaged fluorescence showed numerous spontaneous Ca²⁺ release events with amplitudes much smaller than that of the stimulated Ca²⁺ transients (Fig. 2B, Online Fig. II, bottom panels). This pattern of spontaneous Ca²⁺ release observed in ventricular myocytes in the intact PLN^−/−/RyR2-R4496C^+/− or PLN^−/− heart is very similar to that seen in isolated cells (Fig. 1). Thus, the distinct features of spontaneous Ca²⁺ release in isolated PLN^−/−/RyR2-R4496C^+/− or PLN^−/− myocytes reflect the intrinsic properties of intracellular Ca²⁺ handling of these cells, rather than reflecting the consequences of cellular damage during cell isolation.

Figure 2. PLN-KO fragments cell-wide propagating spontaneous Ca²⁺ waves in intact hearts.

Intact hearts isolated from RyR2-R4496C^+/− or PLN^−/−/RyR2-R4496C^+/− mice were loaded with Rhod-2-AM and Langendorff-perfused with 6 mM extracellular Ca²⁺ and paced at 6 Hz to induce SR Ca²⁺ overload. Spontaneous SR Ca²⁺ release in epicardial ventricular myocytes in intact hearts was monitored by line-scan confocal Ca²⁺ imaging. Representative line-scan images (top) and the corresponding digitized images (middle) of cell-wide propagating spontaneous Ca²⁺ waves in intact RyR2-R4496C^+/− hearts (A) and of mini-waves and Ca²⁺ sparks in intact PLN^−/−/RyR2-R4496C^+/− hearts (B) are shown. Panels A and B (bottom) show the spatial average of fluorescence signal along the scan-line. (C) Definition of wave fluorescence, full duration at half maximum (FDHM), time to peak, amplitude, and rate of rise.

To further assess the spatial and temporal properties of spontaneous Ca²⁺ release in ventricular myocytes in intact RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/− and PLN^−/− hearts, we analyzed all spontaneous Ca²⁺ release events (Figs. 2A, 2B, Online Fig. II, middle panels, and Online Fig. III) and classified them into three categories: waves, mini-waves, and sparks, based on their total fluorescence/event. As seen in Fig. 3, RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/−, and PLN^−/− ventricular myocytes displayed very different distributions of spontaneous Ca²⁺ release events. In RyR2-R4496C^+/− ventricular myocytes, 93% of the total spontaneously released Ca²⁺ was released in the form of Ca²⁺ waves, while mini-waves and Ca²⁺ sparks together consisted of only 7% of the total spontaneously released Ca²⁺ (Fig. 3A,D). In contrast, a majority of the spontaneously released Ca²⁺ in PLN^−/−/RyR2-R4496C^+/− or PLN^−/− cells was released as mini waves (77–74%), while Ca²⁺ waves and sparks consisted of 20–25% and 3-2% of the total released Ca²⁺, respectively (Fig. 3B,C,D). Furthermore, the occurrence of Ca²⁺ waves was significantly greater in RyR2-R4496C^+/− cells than in PLN^−/−/RyR2-R4496C^+/− or PLN^−/− cells (Fig. 3D). On the other hand, the occurrence of mini-waves and Ca²⁺ sparks was significantly greater in PLN^−/−/RyR2-R4496C^+/− or PLN^−/− cells than in RyR2-R4496C^+/− cells (Fig. 3E,F,G). In other words, RyR2-R4496C^+/− ventricular myocytes displayed primarily Ca²⁺ waves, whereas PLN^−/−/RyR2-R4496C^+/− or PLN^−/− ventricular myocytes exhibited predominantly mini-waves and Ca²⁺ sparks with few Ca²⁺ waves (Fig. 3A,B,C).

Figure 3. Distribution of spontaneous Ca²⁺ release events in intact RyR2-R4496C mutant hearts with or without PLN.

Line-scan confocal images were digitized and spontaneous Ca²⁺ release events were detected and classified using a custom-made program as described in the supplementary methods. (A, B, C) Distribution of spontaneous Ca²⁺ release events in RyR2-R4496C^+/− mutant (A), PLN^−/−/RyR2-R4496C^+/− (B), and PLN^−/− (C) hearts according to their total fluorescence. Three types of spontaneous Ca²⁺ release events (Ca²⁺ sparks, mini-waves, and waves) were classified based on the size of the total fluorescence (see Supplementary Methods). The red line represents a Gaussian fit of the distribution of Ca²⁺ waves in RyR2-R4496C^+/− hearts. (D) The overall contribution (%) of sparks, mini-waves, and waves in RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/−, and PLN^−/− hearts. (E, F, G) The occurrence (events/scan) of Ca²⁺ waves (E), mini-waves (F), and Ca²⁺ sparks (G) in RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/−, and PLN^−/− hearts. Data shown are mean ± SEM from 19 (PLN^−/−), 39 (RyR2-R4496C^+/−), and 43 (PLN^−/−/RyR2-R4496C^+/−) line-scan images (*P < 0.001).

We next determined and compared the properties of Ca²⁺ waves, mini waves, and Ca²⁺ sparks in ventricular myocytes in intact RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/− and PLN^−/−hearts. We found that the amplitude, full duration at half maximum (FDHM), and rate of rise of Ca²⁺ waves or mini waves are significantly greater in RyR2-R4496C^+/− cells than in PLN^−/−/RyR2-R4496C^+/− or PLN^−/− cells (Fig. 4A,B). On the other hand, the amplitude and duration of Ca²⁺ sparks are significantly smaller in RyR2-R4496C^+/− cells than in PLN^−/−/RyR2-R4496C^+/− or PLN^−/− cells. Consistent with previously reported data²⁷, PLN-KO increased the amplitude and decreased the FDHM of stimulated Ca²⁺ transients (Fig. 2, Online Fig. IV). Taken together, our single cell and intact heart Ca²⁺ imaging studies demonstrate that PLN-KO suppresses SCWs in RyR2-R4496C^+/− mutant ventricular myocytes by breaking up cell-wide propagating SCWs into mini-waves and Ca²⁺ sparks and reducing the amplitude, duration, and rate of rise of SCWs.

Figure 4. Effect of PLN-KO on spontaneous Ca²⁺ release in intact RyR2-R4496C mutant hearts.

Spontaneous Ca²⁺ release events in intact RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/−, and PLN^−/− hearts were divided into Ca²⁺ waves (A), mini-waves (B), and Ca²⁺ sparks (C) as described in the legend of Fig. 3, and their amplitude (top), full duration at half maximum (FDHM) (middle), and rate of rise (bottom) were compared. Data shown are mean ± SEM from 19–43 line-scan images (^#P < 0.01, *P < 0.001, vs RC^+/−).

PLN-KO suppresses triggered activities in RyR2-R4496C^+/− ventricular myocytes

Spontaneous SR Ca²⁺ release can lead to DADs, and DADs can trigger action potentials (APs) when the amplitude of a DAD reaches the threshold for Na⁺ channel activation. Whether spontaneous Ca²⁺ release can generate DADs with amplitudes that are sufficient to trigger APs depends on the amplitude and rate of rise of the spontaneous Ca²⁺ release^{10, 34}. The substantially different spatial and temporal properties of spontaneous Ca²⁺ release in RyR2-R4496C^+/− and PLN^−/−/RyR2-R4496C^+/− cells raise the important question of whether PLN-KO can also affect the occurrence of triggered activities. To address this question, we perfused ventricular myocytes isolated from the RyR2-R4496C^+/− and PLN^−/−/RyR2-R4496C^+/− mice with 6 mM extracellular Ca²⁺ to induce SR Ca²⁺ overload and spontaneous Ca²⁺ release. We then recorded the membrane potential in these cells using the perforated patch current clamp technique. As shown in Fig. 5, RyR2-R4496C^+/− ventricular myocytes displayed frequent DADs and spontaneously triggered APs (Figs. 5Aa, C and D), which is consistent with those reported previously³¹. Interestingly, under the same conditions, PLN^−/−/RyR2-R4496C^+/− ventricular myocytes exhibited a large number of small DADs, but little or no triggered APs (Figs. 5Ba, C and D). Thus, these observations indicate that PLN-KO suppresses the occurrence of triggered APs in RyR2-R4496C^+/− ventricular myocytes.

Figure 5. Effect of PLN-KO on delayed afterdepolarizations and triggered activities.

Ventricular myocytes were isolated from RyR2-R4496C^+/− and PLN^−/−/RyR2-R4496C^+/− hearts and perfused with 6 mM extracellular Ca²⁺ to induce spontaneous Ca²⁺ release. Membrane potentials in RyR2-R4496C^+/− (A) or PLN^−/−/RyR2-R4496C^+/− (B) myocytes before (a) and after (b) the treatment with tBHQ, a SERCA2a inhibitor, were recorded using the perforated patch current clamp technique. (C, D) The frequency of spontaneously triggered APs (C) and delayed afterdepolarizations (DADs) (D) in RyR2-R4496C^+/− and PLN^−/−/RyR2-R4496C^+/− ventricular myocytes before (a) and after (b) tBHQ treatment. Data shown are mean ± SEM from 8–12 cells (^#P < 0.05, vs RC mice; *P < 0.05, vs -tBHQ). (E) Representative line-scan images of spontaneous Ca²⁺ release in isolated PLN^−/−/RyR2-R4496C^+/− ventricular myocytes before (a) and after (b) the treatment with 5 µM tBHQ (n = 21–22 cells) are shown.

Given the close link between SCWs and triggered activities^{10, 34}, the lack of triggered APs in PLN^−/−/RyR2-R4496C^+/− cells is likely attributable to the absence of SCWs in these cells. To test this possibility, we mimicked the action of PLN by partially inhibiting SERCA2a with 2,5-Di-tert-butylhydroquinone (tBHQ, 5 µM), a SERCA2a inhibitor. As shown in Fig. 5E, partial inhibition of SERCA2a by tBHQ in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes converted multiple and frequent mini-waves into cell-wide propagating SCWs similar to those observed in RyR2-R4496C^+/− ventricular myocytes. Importantly, the tBHQ treatment increased the occurrence of triggered APs (Figs. 5Bb, C,D) in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes. On the other hand, the tBHQ treatment did not markedly affect the occurrence of DADs or triggered APs in RyR2-R4496C^+/− cells (Figs. 5Ab,C,D). Therefore, these data suggest that PLN-KO suppresses triggered activities by breaking up cell-wide SCWs.

Role of RyR2, LTCC, NCX, and SR Ca²⁺ load in breaking cell-wide SCWs in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes

The conversion of mini-waves to cell-wide SCWs by tBHQ in PLN^−/−/RyR2-R4496C^+/− cells also suggests that enhanced SERCA2a activity as a consequence of PLN-KO is an important determinant of the occurrence of mini-waves. However, it is possible that PLN-KO may also lead to compensatory changes in the expression of Ca²⁺ handling proteins, which may in turn contribute to the genesis of mini-waves in PLN^−/−/RyR2-R4496C^+/− cells. To test this possibility, we assessed the expression level of RyR2, LTCC, SERCA2a, and NCX proteins in the RyR2-R4496C^+/− and PLN^−/−/RyR2-R4496C^+/− hearts using immunoblotting analysis. As shown in Fig. 6A, there were no significant differences in their expression levels except for RyR2 that exhibited a slightly higher (~10%, P<0.05) expression in PLN^−/−/RyR2-R4496C^+/− hearts than in RyR2-R4496C^+/− hearts.

Figure 6. Role of RyR2, LTCC, NCX, and SR Ca²⁺ load in the generation of mini-waves in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes.

(A) Whole heart homogenates were prepared from RyR2-R4496C^+/− and PLN^−/−/RyR2-R4496C^+/− mutant mice and used for immunoblotting analysis using antibodies against RyR2, LTCC, SERCA2a, NCX or β-actin. Data shown are mean ± SEM (n=3, *P < 0.05, vs RyR2-R4496C^+/−). (B, C, D) Ventricular myocytes were isolated from PLN^−/−/RyR2-R4496C^+/− hearts and loaded with the fluorescent Ca²⁺ indicator dye fluo-4-AM. Representative line-scan images of spontaneous Ca²⁺ release induced by elevated extracellular Ca²⁺ (6 mM) in PLN^−/−/RyR2-R4496C^+/− cells before (a) and after (b) the treatment with Bay K (B), caffeine (C), and LiCl (replacing NaCl in KRH) (D) (n = 14–18 cells) are shown. Note that none of these treatments converted mini-waves to cell-wide SCWs. (E) SR Ca²⁺ contents in fluo-4-AM loaded RyR2-R4496C^+/− (a), PLN^−/−/RyR2-R4496C^+/− (b), and PLN^−/− (c) ventricular myocytes were estimated by measuring the amplitude of caffeine (20mM) induced Ca²⁺ transients (d). Data shown are mean ± SEM (n=12–18) (*P < 0.05, vs RyR2-R4496C^+/−).

It is also possible that PLN-KO may break SCWs by altering the activity of LTCC, RyR2, or NCX in addition to SERCA2a. For instance, mini-waves could result from reduced activity of LTCC or RyR2, which would reduce Ca²⁺ influx and SR Ca²⁺ release, and thus the propagation of Ca²⁺ waves. Further, mini-waves could also result from increased activity of NCX, which would enhance Ca²⁺ removal, and thus reduce SR Ca²⁺ content and SR Ca²⁺ release. To test these possibilities, we assessed the impact of Bay K 8644 (a LTCC agonist), caffeine (a RyR2 agonist), and Li⁺ (an inhibitor of NCX) on spontaneous SR Ca²⁺ release in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes. In sharp contrast to tBHQ, Bay K, caffeine, or Li⁺ failed to convert mini-waves into cell-wide SCWs in PLN^−/−/RyR2-R4496C^+/− cells (Fig. 6B,C,D).

The SR Ca²⁺ content is also a critical determinant of spontaneous Ca²⁺ waves^{35, 36}. Accordingly, we determined the SR Ca²⁺ content in RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/−, and PLN^−/− cells. We found that PLN^−/−/RyR2-R4496C^+/− and PLN^−/− cells displayed significantly higher SR Ca²⁺ content than RyR2-R4496C^+/− cells (Fig. 6E). Thus, enhanced SERCA2a activity, rather than reduced SR Ca²⁺ content, decreased LTCC or RyR2 activity, or increased NCX activity, is a major contributor to the break-up of cell-wide SCWs.

PLN-KO protects the RyR2-R4496C^+/− mice from stress-induced VTs

It has been shown that the RyR2-R4496C mutant mice are highly susceptible to CPVT, which is caused by DAD-induced triggered activities^{20, 30–32}. The lack of triggered activities in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes upon SR Ca²⁺ overload raises the possibility that PLN-KO may also suppress CPVT. To directly test this possibility, we recorded ECG in WT littermates, RyR2-R4496C^+/−, RyR2-R4496C^+/+, PLN^−/−/RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/+, and PLN^−/− mice before and after the injection of a mixture of caffeine and epinephrine. Similar to those reported previously²⁰, caffeine and epinephrine induced long-lasting ventricular tachyarrhythmias (VTs) in RyR2-R4496C^+/− mice, but not in their WT littermates (Fig. 7). The RyR2-R4496C^+/+ homozygous mice are especially vulnerable to stress-induced VTs, displaying sustained VTs for the entire 30 min-period of recording after the injection of the triggers²⁰. Remarkably, caffeine and epinephrine induced little or no VTs in the PLN^−/− mice or PLN^−/−/RyR2-R4496C^+/− mice, and only short-lasting VTs in the PLN^−/−/RyR2-R4496C^+/+ mice (Fig.8). These data indicate that PLN-KO mice are not susceptible to CPVT, and that PLN-KO protects the RyR2-R4496C mutant mice from stress-induced VTs.

Figure 7. RyR2-R4496C^+/− mice are susceptible to CPVT.

Representative ECG recordings of WT littermates (A) and RyR2-R4496C^+/− (B) mice before (a) and after (b) the injection of epinephrine (1.6 mg/kg) and caffeine (120 mg/kg). VT duration (%) in WT littermates and RyR2-R4496C^+/− mutant mice within each 3-min (C) or 30-min (D) period of ECG recordings. Data shown are mean ± SEM from 9–11 mice (*P < 0.05, vs WT).

Figure 8. PLN-KO protects against stress-induced VTs in mice.

Representative ECG recordings of PLN^−/− (A), PLN^−/−/RyR2-R4496C^+/− (B), and PLN^−/−/RyR2-R4496C^+/+ (C) mice before (a) and after (b) the injection of epinephrine (1.6 mg/kg) and caffeine (120 mg/kg). VT duration (%) in PLN^−/−, PLN^−/−/RyR2-R4496C^+/−, and PLN^−/−/RyR2-R4496C^+/+ mice within each 3-min (D) or 30-min (E) period of ECG recordings. Data shown are mean ± SEM from 9–10 mice (*P < 0.05, vs PLN^−/−).

PLN^−/−/RyR2-R4496C^+/− mice display no severe defects in cardiac structure

Enhanced SR Ca²⁺ leak as a result of overexpression of the Ca²⁺/calmodulin dependent protein kinase II (CaMKII) in the heart has been shown to cause severe heart failure and dilated cardiomyopathy^{37, 38}. It would be of interest to determine whether enhanced SR Ca²⁺ leak as a result of PLN-KO could induce severe structural changes in the heart. To this end, we performed echocardiography on conscious WT, RyR2-R4496C^+/−, PLN^−/−/RyR2-R4496C^+/−, and PLN^−/− mice. We found that the RyR2-R4496C^+/− mutation itself did not induce gross changes in cardiac structure and function (Online Table I), which is in agreement with those reported previously^{30, 31}. We also found no severe structural defects in the PLN^−/−/RyR2-R4496C^+/− or PLN^−/− hearts, despite the chronic SR Ca²⁺ overload and enhanced spontaneous Ca²⁺ leak (mini waves and Ca²⁺ sparks) in the PLN^−/−/RyR2-R4496C^+/− or PLN^−/− cardiomyocytes. This is consistent with previous observation that PLN^−/− mice show enhanced myocardial contractility but no gross defects in cardiac structure^{26, 39, 40}. There are, however, some small differences between PLN^−/−/RyR2-R4496C^+/− and WT mice and between PLN^−/− and WT mice (Online Table I). Thus, as with PLN^−/− hearts, PLN^−/−/RyR2-R4496C^+/− hearts show no severe defects in cardiac structure.

DISCUSSION

A novel and surprising finding of the present study is that, despite severe SR Ca²⁺ leak, PLN-KO mice are not susceptible to stress-induced VTs. In fact, on the contrary, PLN-KO protects a mouse model harbouring the CPVT-causing RyR2-R4496C mutation from stress-induced VTs. Single cell and intact heart Ca²⁺ imaging reveal that PLN-KO effectively breaks the cell-wide propagating SCWs into mini-waves and Ca²⁺ sparks. Furthermore, PLN-KO markedly suppresses SCW-evoked triggered activities in RyR2-R4496C mutant ventricular myocytes. These observations indicate that spontaneous SR Ca²⁺ leak in the forms of mini-waves and Ca²⁺ sparks (leaky SR) without generating cell-wide propagating SCWs is not necessarily linked to triggered activities and triggered arrhythmias. Our data suggest that breaking up cell-wide propagating SCWs into mini-waves and Ca²⁺ sparks is protective against Ca²⁺ triggered arrhythmias.

An important question is how PLN-KO rescues the CPVT phenotype of the RyR2-R4496C mutant mice in the face of severe diastolic SR Ca²⁺ leak? Increased SR Ca²⁺ leak is often observed in cardiomyocytes from heart failure and is thought to be a major cause of Ca²⁺ triggered arrhythmias^12–14. This is because diastolic SR Ca²⁺ leak can alter the membrane potential via the activation of the electrogenic Na⁺/Ca²⁺ exchanger (NCX), resulting in DADs. These DADs can potentially trigger ectopic APs that in turn can lead to triggered arrhythmia^{8, 10–12}. However, whether a DAD is able to trigger an AP depends on its amplitude. An AP is triggered when the amplitude of a DAD reaches the activation threshold for Na⁺ channels. Furthermore, the amplitude of DADs is dependent on the amplitude and rate of rise of spontaneous SR Ca²⁺ release^{10, 34}. It has been estimated that a total SR Ca²⁺ release of 50–70% of the SR Ca²⁺ load is required to generate DADs with amplitudes sufficient to produce an AP¹⁰. Therefore, the small diastolic SR Ca²⁺ leak in the form of brief, localized Ca²⁺ sparks or even mini-waves themselves are unlikely to produce DADs with amplitudes that are high enough to cause triggered activities. It is the SR Ca²⁺ overload induced cell-wide propagating SCWs that are capable of producing triggered activities. In accordance with this view, we detected a large number of small DADs but only a few triggered APs in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes that displayed severe SR Ca²⁺ leak in the form of Ca²⁺ sparks and mini-waves. On the other hand, we observed a number of triggered APs in RyR2-R4496C^+/− ventricular myocytes that exhibited cell-wide propagating SCWs. Interestingly, triggered APs were readily detected in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes after transforming mini-waves to cell-wide propagating SCWs by partially inhibiting SERCA2a with tBHQ. On the other hand, increasing the activity of LTCC with Bay K or the activity of RyR2 with caffeine or decreasing the activity of NCX with Li⁺ failed to convert mini-waves to cell-wide SCWs in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes. Further, we found that the SR Ca²⁺ content was elevated in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes compared to that in RyR2-R4496C^+/− cells. Thus, enhanced SERCA2a activity as a result of PLN-KO likely contributes to the break-up of cell-wide SCWs in PLN^−/−/RyR2-R4496C^+/− ventricular myocytes, rather than reduced SR Ca²⁺ load or altered RyR2, LTCC, or NCX activity due to potential PLN-KO induced compensatory changes. The enhanced SERCA2a activity as a result of PLN ablation would result in a rapid re-sequestration of the released Ca²⁺ into the SR. This would effectively buffer or reduce the cytosolic Ca²⁺ level that is important for the propagation of Ca²⁺ waves via Ca²⁺ induced Ca²⁺ release, thus limiting the spatial spread of Ca²⁺ waves²⁹. This effect on SCWs would reduce the amplitude of DADs and thus decrease the propensity for triggered APs and triggered arrhythmias. It is of interest to note that Davia et al.⁴¹ have shown that adenovirus-mediated overexpression of SERCA2a in adult rabbit ventricular myocytes reduced the occurrence of aftercontractions. Our present findings are consistent with those of Davia et al. and further demonstrate that enhanced SERCA2a activity suppresses triggered activities by breaking up cell-wide SCWs.

Although PLN-KO is effective in suppressing stress-induced VTs in the CPVT RyR2-R4496C mutant mice, whether PLN-KO would be beneficial in suppressing stress-induced VTs in other animal models or in humans with CPVT remains to be determined. Albeit not specifically on stress-induced arrhythmias, a number of studies have investigated the impact of PLN-KO on heart failure and cardiomyopathies^42–44. For example, it has been shown that PLN-KO rescues the heart failure and dilated cardiomyopathy phenotypes in a mouse model in which the cytoskeletal, muscle specific LIM protein (MLP) is ablated⁴². PLN-KO has also been shown to reverse the cardiac hypertrophy phenotype in a mouse model with calsequestrin overexpression⁴³. However, PLN-KO does not rescue cardiac dysfunction in all mouse models of heart failure and cardiomyopathies tested^45–47. For instance, it has recently been shown that despite the rescue of SR Ca²⁺ handling, PLN-KO exaggerates heart failure and mortality in CaMKIIδc overexpressing mice⁴⁶. It was suggested that PLN deficiency in the CaMKIIδc overexpressing mice resulted in markedly increased SR Ca²⁺ load in the face of enhanced diastolic SR Ca²⁺ leak due to CaMKIIδc-dependent hyperphosphorylation of RyR2. The combination of increased SR Ca²⁺ load and enhanced SR Ca²⁺ leak predisposes cardiomyocytes to cell death and other Ca²⁺-mediated abnormalities. Similarly, the combination of enhanced SR Ca²⁺ load as a result of overexpression of the skeletal muscle SR Ca²⁺ ATPase (SERCA1a) or PLN-KO and increased SR Ca²⁺ leak as a consequence of CASQ2-KO led to myocyte apoptosis, dilated cardiomyopathy, and early mortality⁴⁸. On the other hand, we found that the PLN-KO RyR2-R4496C mutant mice show no severe structural and functional defects. Thus, unlike that seen in the CaMKIIδc overexpressing mice or CASQ2-KO mice, PLN-KO does not lead to cardiac dysfunction in the PLN^−/−/RyR2-R4496C^+/− mice even in the face of enhanced spontaneous SR Ca²⁺ release. The exact reasons for this discrepancy are not clear. Spontaneous SR Ca²⁺ release in the CaMKIIδc-overexpressing or CASQ2-KO mice may be much more severe than that in the RyR2-R4496C^+/− mice. Consistent with this view, both CaMKIIδc-overexpressing and CASQ2-KO mice, but not RyR2-R4496C^+/− mice, exhibit dilated cardiomyopathy, heart failure or hypertrophy^{38, 49}. Thus, it is possible that the enhanced SERCA2a activity as a result of PLN-KO may not be able to fully compensate for the much more severe SR Ca²⁺ leak caused by CaMKIIδc overexpression or CASQ2-KO, leading to chronic diastolic SR Ca²⁺ leak, cardiomyopathies and heart failure. Therefore, whether PLN-KO produces beneficial effects would be dependent on the cause and severity of the defects of the disease model. It is also important to note that, opposite to those observed in PLN-KO mice, PLN deficiency in humans as a result of nonsense mutations is associated with severe dilated cardiomyopathy and heart failure⁵⁰. Hence, the beneficial effects of PLN-KO may also be species dependent.

In summary, we show that PLN-KO effectively breaks SCWs into mini-waves and Ca²⁺ sparks in mouse ventricular myocytes expressing the SCW-prone, CPVT-causing RyR2-R4496C mutant. We further show that PLN-KO markedly suppresses SCW-evoked triggered activity and completely protects the RyR2-R4496C^+/− mutant mice against CPVT. Thus, as with inhibiting RyR2 activity, breaking up SCWs by enhancing SERCA2a activity represents an effective means in suppressing Ca²⁺ triggered arrhythmias.

Limitations

In this study, we used confocal linescan imaging to estimate and compare the SR Ca²⁺ contents in cardiomyocytes with different genotypes by measuring the amplitude of caffeine evoked Ca²⁺ transients. Although this approach yielded useful information on the relative SR Ca²⁺ contents of different groups of cells, it did not provide a quantitative assessment of the SR Ca²⁺ content. Further, the amplitude of caffeine evoked Ca²⁺ transients could be influenced by various factors such as cytosolic Ca²⁺ buffering. Since increased SERCA2a activity as a result of PLN ablation would enhance the removal of cytosolic Ca²⁺ (equivalent to increased cytosolic Ca²⁺ buffering), the increase in the relative SR Ca²⁺ content detected in PLN^−/−/R4496C^+/− cells would have been underestimated due to this increased Ca²⁺ removal/buffering. However, since PLN ablation increases the SR Ca²⁺ content in PLN^−/−/R4496C^+/− cardiomyocytes, the lack of cell-wide propagating SCWs in these cells is unlikely to be due to a reduced SR Ca²⁺ content.

Novelty and Significance.

What Is Known?

• Spontaneous Ca²⁺ waves (SCWs) are a major cause of Ca²⁺-mediated arrhythmias.

• Mutations in the cardiac ryanodine receptor (RyR2) associated with catecholaminergic polymorphic ventricular tachycardia (CPVT)- enhance the propensity for SCWs.

• Ablation of phospholamban (PLN), an inhibitor of sarco(endo)plasmic reticulum Ca²⁺ ATPase (SERCA2a), increases sarcoplasmic reticulum (SR) Ca²⁺ leak, but aborts SCWs.

What New Information Does This Article Contribute?

• PLN ablation breaks cell-wide propagating SCWs into mini-waves and Ca²⁺ sparks in intact CPVT RyR2 mutant hearts.

• Despite markedly increased SR Ca²⁺ leak, PLN ablation protects against CPVT in RyR2 mutant mice.

• Breaking up SCWs by enhancing Ca²⁺ uptake represents an effective means for suppressing Ca²⁺-mediated arrhythmias.

Enhanced SR Ca²⁺ sequestration as a result of PLN ablation has been shown, on one hand, to increase potentially arrhythmogenic SR Ca²⁺ leak, but, on the other hand, to abort SCWs that could lead to triggered activity. These seemingly paradoxical actions raise an important question as to whether enhanced SR Ca²⁺ sequestration is pro-arrhythmic or anti-arrhythmic. To this end, we assessed the effect of PLN ablation on SR Ca²⁺ leak, SCWs, triggered activity, and stress-induced ventricular tachycardia’s (VTs) in a mouse model of CPVT. We found that enhanced SR Ca²⁺ uptake as a consequence of PLN ablation fragmented cell-wide propagating SCWs, but increased SR Ca²⁺ leak in the forms of mini-waves and Ca²⁺ sparks. Despite the presence of severe SR Ca²⁺ leak, PLN ablation suppressed triggered activity evoked by SR Ca²⁺ overload and protected CPVT-associated RyR2 mutant mice from stress-induced VTs. Our data show that enhancing SR Ca²⁺ sequestration suppresses CPVT in mice. These observations suggest that breaking up cell-wide propagating SCWs either by reducing the duration of RyR2 openings or by enhancing SR Ca²⁺ sequestration represents a promising strategy for suppressing Ca²⁺ mediated cardiac arrhythmias.

Nonstandard Abbreviations and Acronyms

RyR2

cardiac ryanodine receptor

CPVT

catecholaminergic polymorphic ventricular tachycardia

DAD

delayed afterdepolarization

PLN

phospholamban

SR

sarcoplasmic recticulum

SCWs

spontaneous Ca²⁺ waves

SERCA2a

cardiac sarco(endo)plasmic recticulum Ca²⁺-ATPase

AP

action potential

tBHQ

2,5-Di-tert-butylhydroquinone

CASQ2

cardiac calsequestrin

LTCC

L-type Ca²⁺ Channel

Na⁺/Ca²⁺

sodium/calcium exchange