Increased Ca²⁺ sensitivity of the ryanodine receptor mutant RyR2^R4496C underlies catecholaminergic polymorphic ventricular tachycardia

Abstract

Cardiac ryanodine receptor (RyR2) mutations are associated with autosomal dominant catecholaminergic polymorphic ventricular tachycardia (CPVT), suggesting that alterations in Ca²⁺ handling underlie this disease. Here we analyze the underlying Ca²⁺ release defect that leads to arrhythmia in cardiomyocytes isolated from heterozygous knock-in mice carrying the RyR2^R4496C mutation. RyR2^R4496C−/− littermates (wild type, WT) were used as controls. [Ca²⁺]_i transients were obtained by field stimulation in fluo-3 loaded cardiomyocytes and viewed using confocal microscopy. In our basal recording conditions (2 Hz stimulation rate), [Ca²⁺]_i transients and sarcoplasmic reticulum (SR) Ca²⁺ load were similar in WT and RyR2^R4496C cells. However, paced RyR2^R4496C ventricular myocytes presented abnormal Ca²⁺ release during the diastolic period, viewed as Ca²⁺-waves, consistent with the occurrence of delayed after-depolarizations. The occurrence of this abnormal Ca²⁺ release was enhanced at faster stimulation rates and by β-adrenergic stimulation, which also induced triggered activity. Spontaneous Ca²⁺-sparks were more frequent in RyR2^R4496C myocytes, indicating increased RyR2^R4496C activity. When permeabilized cells were exposed to different cytosolic [Ca²⁺]_i, RyR2^R4496C showed a dramatic increase in Ca²⁺ sensitivity. Isoproterenol increased [Ca²⁺]_i transient amplitude and Ca²⁺ spark frequency to the same extent in WT and RyR2^R4496C cells, indicating that the β-adrenergic sensitivity of RyR2^R4496C cells remained unaltered. This effect was independent of protein expression variations since no difference was found in the total or phosphorylated RyR2 expression levels. In conclusion, the arrhythmogenic potential of the RyR2^R4496C mutation is due to the increased Ca²⁺ sensitivity of RyR2^R4496C, which induces diastolic Ca²⁺ release and lowers the threshold for triggered activity.

Catecholaminergic polymorphic ventricular tachycardia (CPVT) is an inherited arrhythmogenic disease characterized by stress-induced, adrenergically mediated bidirectional or polymorphic ventricular tachycardia occurring in structurally normal hearts¹. During exercise or acute emotions, CPVT patients develop life-threatening ventricular arrhythmias leading to syncope or sudden death. The first RyR2 mutation identified in a CPVT family was R4497C². Today, more than 70 RyR2 mutations have been reported (http://www.fsm.it/cardmoc/), and they comprise the most common genetic subtype of CPVT^3–7, although mutations in the calsequestrin gene can also cause CPVT^8,9.

Diverging results and conclusions have been generated from expression studies of RyR2^R4496C in heterologous systems. Jiang and collaborators showed that RyR2^R4496C (the mouse equivalent of the human RyR2^R4497C mutation), when expressed in human embryonic kidney (HEK) cells, exhibits increased basal activity and increased sensitivity to luminal Ca^2+.10 However, other authors found no difference in the basal activity of RyR2^R4497C, but instead showed increased activity and gating frequency after protein kinase A (PKA) phosphorylation¹¹ or sarcoplasmic reticulum (SR) Ca²⁺ overload¹². The expression studies were carried out in a variety of models, which may explain the inhomogeneous findings. Furthermore, heterologous systems lack cardiac intracellular environment with all the RyR2 accessory proteins¹³ and most Ca²⁺-handling proteins, so analysis in native cardiac myocytes is now critical to elucidate the mechanisms by which the mutation leads to cardiac arrhythmia.

Recently, a knock-in mouse-model carrier of the RyR2^R4496C mutation was developed¹⁴. Their phenotype presents extraordinary similarity with the clinical manifestations of patients carrying the RyR2^R4497C mutation, including the development of bidirectional ventricular tachycardia. When exposed to adrenaline and caffeine, the RyR2^R4496C cardiomyocytes develop delayed after-depolarizations (DADs)¹⁵, suggesting that triggered arrhythmias are elicited by adrenergic activation¹⁶. Here we demonstrate that untreated RyR2^R4496C myocytes have increased spontaneous Ca²⁺ release in diastole during electrical pacing, due to the enhanced Ca²⁺ sensitivity of mutant RyR2; this abnormality is further augmented by exposure to isoproterenol and increasing pacing rates.

MATERIALS AND METHODS

Ventricular cardiomyocytes from male and female RyR2^R4496C+/− mice (RyR2^R4496C) and their wild type RyR2^R4496C−/− (WT) littermates were isolated using a standard enzymatic digestion¹⁷. [Ca²⁺]_i transients and Ca²⁺-sparks were viewed in isolated myocytes by confocal microscopy and analyzed using homemade routines.

A complete Methods section can be found in the Online Supplemental Material.

RESULTS

Abnormal Ca²⁺ release in RyR2^R4496C myocytes

Electrophysiological experiments in RyR2^R4496C myocytes have evidenced DADs and triggered action potentials in the presence of adrenergic stimulation after stopping electrical stimulation¹⁵. Figure 1 shows Ca²⁺ images in ventricular myocytes paced at 4 Hz in the presence of 1 μmol/L isoproterenol. The WT cardiomyocyte showed no spontaneous Ca²⁺ release after electrical pacing interruption (Fig. 1A). On the contrary, the RyR2^R4496C cell showed Ca²⁺-waves evoking two [Ca²⁺]_i transients, just after stimulation stopped, which was followed by several Ca²⁺-waves, consistent with triggered activity and DADs (Fig. 1B). These data suggest that abnormal Ca²⁺ release underlies the electrical abnormalities in RyR2^R4496C cardiomyocytes. Since DADs are initiated by Ca²⁺-waves that are in turn initiated by Ca²⁺-sparks¹⁸, we analyzed Ca²⁺-sparks in RyR2^R4496C cardiomyocytes. [Ca²⁺]_i transients and waves characteristics are presented below.

Figure 1. Triggered activity observed in RyR2^R4496C myocytes.

Line-scan images of ventricular myocytes isolated from (A) a WT mouse and (B) RyR2^R4496C mouse during electrical stimulation (4Hz). The corresponding fluorescence traces are shown below. Red lines indicate electrical stimulation. After pacing stopped, the WT cell remains silent while the RyR2^R4496C cell shows Ca²⁺-waves that induce two full contractions followed by isolated Ca²⁺-waves consistent with DADs.

Ca²⁺-sparks in RyR2^R4496C myocytes

Figure 2A shows representative images of Ca²⁺-sparks¹⁹. Ca²⁺ spark frequency was double in RyR2^R4496C cells compared with WT cells (p<0.001, Figure 2B). This could be due to an increase in (i) the SR Ca²⁺ load, (ii) the level of RyR2 expression and/or phosphorylation, (iii) the diastolic [Ca²⁺]_i or (iv) changes in the intrinsic channel properties.

Figure 2. Ca²⁺-spark in intact RyR2^R4496C myocytes.

A. Line-scan images obtained in a WT and a RyR2^R4496C cell (R4496C) in the absence (top) or presence (bottom) of 1 μmol/L isoproterenol. B. Average of Ca²⁺-spark occurrence in WT cells without (white bar, n=37) and with 1 μmol/L isoproterenol (hatched bar, n=10), and in RyR2^R4496C cells without (blue bar, n=43) and with 1 μmol/L isoproterenol (hatched blue bar, n=17). C. Percentage of increase induced by 1 μmol/L isoproterenol on Ca²⁺-spark frequency in 10 WT and 17 RyR2^R4496C cells. D. Average of sites where Ca²⁺-sparks are recorded within the same cell during the recording period (18 s). E. Probability in each cell to present sites that fire repetitively. F. Maximum number of Ca²⁺-sparks recorded in the same site. For D, E & F: n= 37 WT cells, n=10 WT cells in the presence of 1 μmol/L isoproterenol; n=43 RyR2^R4496C myocytes and n=17 RyR2^R4496C myocytes under 1 μmol/L isoproterenol perfusion. *p<0.05, **p<0.01, ***p<0.001 vs. the same group in the absence of isoproterenol. †p<0.05, ††p<0.01, †††p<0.001 vs WT.

We estimated SR Ca²⁺ load in quiescent ventricular myocytes. RyR2^R4496C showed reduced SR Ca²⁺ content (F/F₀: 7.0±0.5, n=11 in RyR2^R4496C vs. 8.5±0.5 in WT, n=10, p<0.05), ruling out SR Ca²⁺ overload. No major alteration in Ca²⁺-spark characteristics was observed (Online Table I).

We performed Western-blots of total and phosphorylated RyR2 in hearts in basal conditions and following isoproterenol perfusion. Neither the total RyR2 expression nor the level of phosphorylated RyR2 (P-Ser 2809) was different between WT and RyR2^R4496C (Online Fig. I). We also performed functional experiments challenging the cells with 1 μmol/L isoproterenol. This procedure increased Ca²⁺-spark occurrence in both WT and RyR2^R4496C myocytes (Fig. 2B) by the same percentage (Fig. 2C). Ca²⁺-spark characteristics in the presence of isoproterenol are provided in Online Table I. Similar results were found using a lower isoproterenol concentration (100 nmol/L) (Online Fig. IIA). Furthermore, treatment of RyR2^R4496C myocytes with either a PKA blocker (KT5720) or a CaMKII blocker (KN93) failed to decrease Ca²⁺-spark frequency (Online Fig. III). These data rule out an increase in the total RyR2 expression or a higher level of basal phosphorylation as an explanation for the higher Ca²⁺-spark occurrence in RyR2^R4496C myocytes.

Resting cytoplasmic [Ca²⁺]_i, measured using Fura-2, was similar between WT and RyR2^R4496C cells (ratios: 0.56±0.02 in 16 WT myocytes, 0.57±0.01 in 45 RyR2^R4496C cells, p>0.05). Therefore the increased Ca²⁺-spark occurrence in quiescent RyR2^R4496C cells was not caused by differences in the resting intracellular Ca²⁺.

Ca²⁺-sparks are produced by the opening of RyR2 clusters. The increase in total Ca²⁺ spark frequency in RyR2^R4496C could be due to a greater number of clusters firing Ca²⁺-sparks or to the increased propensity of some clusters to fire repetitively, becoming “eager” clusters. We analyzed our data discriminating specific sites presenting multiple Ca²⁺-sparks during the recording time (about 20 s). Firing sites were counted as the sites where we recorded at least one Ca²⁺-spark. Fig. 2D shows that the RyR2^R4496C myocytes presented more firing sites, and that isoproterenol increased the number of sites in both WT and RyR2^R4496C cells. This indicates that the RyR2^R4496C cells presented more Ca²⁺-sparks due to the existence of more active RyR2 clusters (Fig. 2E). We also measured the maximum number of Ca²⁺-sparks recorded at the same site in each group and found that this was also significantly increased in RyR2^R4496C myocytes and further enhanced by β-adrenergic stimulation (Fig. 2F). Taken together, these data suggest that RyR2^R4496C cells present more Ca²⁺-sparks because of more active RyR2s clusters and a greater probability of repetitive openings of these clusters in the RyR2^R4496C myocytes.

We next explored whether RyR2^R4496C presents abnormal Ca²⁺ sensitivity. We analyzed Ca²⁺-sparks in permeabilized cells exposed to various cytoplasmic [Ca²⁺]_i. Figure 3A illustrates enhanced Ca²⁺-sparks occurrence in a RyR2^R4496C cell at 30 nmol/L [Ca²⁺]_i. At all tested [Ca²⁺]_i, Ca²⁺-sparks were much more frequent in RyR2^R4496C than in WT myocytes, consistent with increased cytosolic Ca²⁺ sensitivity (Fig. 3B). Analysis of the Ca²⁺-spark characteristics in permeabilized cells essentially confirmed the results obtained in intact cells (Online Table II). We estimated SR Ca²⁺ content in permeabilized cells and found that at all tested [Ca²⁺]_i the caffeine-evoked [Ca²⁺]_i transient was significantly decreased in the RyR2^R4496C cells (Fig. 3C). Thus the higher Ca²⁺ spark occurrence in RyR2^R4496C myocytes was not due to either a higher level of Ca²⁺ stored in the SR or an alteration of calsequestrin level evaluated by Western-blots (data not shown). Fig. 3D shows the luminal Ca²⁺ dependence of Ca²⁺-sparks, apparently consistent with increased luminal Ca²⁺ sensitivity. However cytosolic and luminal Ca²⁺ vary concurrently. The high Ca²⁺-spark occurrence recorded in RyR2^R4496C at very low intracellular Ca²⁺ might suggest that, rather than increasing RyR2^R4496C Ca²⁺ sensitivity, this mutation renders the RyR2 intrinsically active. We repeated the experiments at zero Ca²⁺ (0.5 mmol/L EGTA). In this condition, the occurrence of Ca²⁺-sparks was indistinguishable between WT and RyR2^R4496C cells (Fig. 3.E&F, left), indicating that RyR2^R4496C hyperactivity requires cytosolic Ca²⁺. To ensure that the SR was not depleted in our experimental conditions, we applied caffeine. A robust caffeine-induced Ca²⁺ transient could be evoked (Fig. 3E&F, right), proving that there was significant luminal Ca²⁺ to promote Ca²⁺-sparks. Altogether, our results show that the RyR2^R4496C mutation increases the Ca²⁺ sensitivity of the channel. The RyR2 has 2 affinity Ca²⁺ binding sites on the cytosolic portion: one of high affinity that activates the channel and one of low affinity that inactivates it. Since ryanodine binds to open RyRs, we examined the Ca²⁺-dependence of [³H]ryanodine binding in heart crude membrane preparations. Bell-shape curves were obtained for both WT and RyR2^R4496C but Ca²⁺-induced maximal activation of RyR2^R4496C was reached at one order of magnitude lower in RyR2^R4496C (Fig. 3G), indicating that the cytosolic Ca²⁺ sensitivity of RyR2 is greatly increased. We normalized the [³H]ryanodine binding and fitted data to the Hill equation²⁰ to get the values for Ca²⁺ affinity to the RyR activation (Ka, 21.9±8.3 μmol/L vs. 4.9±1.0μmol/L, n=4, p<0.003 for WT and RyR2^R4496C membranes, respectively) and inactivation (6.5±0.6 mmol/L vs. 4.8±1.1 mmol/L; p>0.05 for WT and RyR2^R4496C membranes) sites. Theses results show a 4.5 fold increase of cytosolic Ca²⁺ sensitivity for the RyR2^R4496C.

Figure 3. The increase in Ca^2+-spark frequency in RyR2^R4496C cells depends on the intracellular Ca²⁺ concentration ([Ca²⁺]_i).

A. Line-scan images obtained at 30 nmol/L [Ca²⁺]_i in WT and RyR2^R4496C (R4496C) permeabilized cells. B. Average of Ca²⁺-spark frequencies obtained in permeabilized myocytes exposed at 7.5 nmol/L (n= 9 vs. n=7), 15 nmol/L (n=10 vs. n=6), 30 nmol/L (n= 23 vs. n=19), 50 nmol/L (n= 37 vs. n=44) and 100 nmol/L (n=8 vs. n=8) [Ca²⁺]_i. WT cells vs. RyR2^R4496C cells. Lines are Boltzmann fitting of the data. C. Caffeine-evoked (20 mmol/L) [Ca²⁺]_i transient in permeabilized myocytes at 7.5 nmol/L (n= 13 vs. n=11), 15 nmol/L (n=15 vs. n=10), 30 nmol/L (n=19 vs. n=20), 50 nmol/L (n=21 vs. n=33) and 100 nmol/L (n=9 vs. n=9) [Ca²⁺]_i, WT cells vs. RyR2^R4496C cells, expressed as % of the caffeine-evoked transient at 100 nmol/L in WT cells. The x axis labels indicate cytosolic [Ca²⁺]_i in nmol/L. D. Ca²⁺-sparks frequencies plotted as a function of caffeine-evoked [Ca²⁺]_i transient. Numbers indicate cytosolic [Ca²⁺]_i. E. Images of Ca²⁺-sparks obtained at 0 nmol/L [Ca²⁺]_i in WT (top, left) and RyR2^R4496C (R4496C, bottom, left) cells and during (right) 20 mmol/L caffeine application (arrow). F. Left. Ca²⁺-spark frequency obtained in WT myocytes (n=6) and in RyR2^R4496C myocytes (n=6). Right. Caffeine-evoked [Ca²⁺]_i transient amplitude (in F/F₀) measured at zero [Ca²⁺]_i in 15 WT cells and 16 RyR2^R4496C cells. G. Specific Ca²⁺-dependent [³H]ryanodine binding curves to crude membrane fractions of WT (n=4) and R4496C+/+ (n=4) heart tissues. Values have been normalized and fitted to the equation y=Bmax*([Ca²⁺]^na/([Ca²⁺]^na+K_a^na))(1−([Ca²⁺]ⁿⁱ/([Ca²⁺]ⁿⁱ+K_iⁿⁱ)))+C. Open symbols, WT; blue symbols, RyR2^R4496C. Solid line for WT and dotted line for RyR2^R4496C cells or homogenates. *p<0.05, **p<0.01, ***p<0.001.

[Ca²⁺]_i transients in RyR2^R4496C myocytes

To determine whether the alteration of diastolic Ca²⁺ spark frequency in RyR2^R4496C cardiomyocytes has an impact during systole, we compared [Ca²⁺]_i transients and cell contraction at different pacing rates (2 Hz, 3 Hz, and 4 Hz) (Online table III). At 2 Hz, the [Ca²⁺]_i transient amplitude, its time to peak, the [Ca²⁺]_i transient decay time and cellular contraction, were similar (p>0.05) in WT and RyR2^R4496C cells, consistent with normal heart function in mice at rest. As stimulation rate increased, weaker [Ca²⁺]_i transients were evoked both in WT and RyR2^R4496C cells (Fig. 4A). However, the decrease in [Ca²⁺]_i transient amplitude was more pronounced (p<0.05) in RyR2^R4496C cells. This reduction was associated with both weaker cellular contraction (Fig. 4B) and slower decay time (Fig. 4C), with no difference in the time to peak (Fig. 4D).

Figure 4. RyR2^R4496C myocytes show a rate-dependent decrease in [Ca²⁺]_i transients and SR Ca²⁺ load.

A. Average of [Ca²⁺]_i transients amplitude (expressed as F/F₀, where F is the peak fluorescence signal and F₀ the diastolic fluorescence) of WT) vs. RyR2^R4496C myocytes obtained by field stimulation at 2 Hz (n= 50 vs. n=52), 3 Hz (n=16 vs. n=21) and 4 Hz (n=14 vs. n=21). B. C& D. Bar graphs comparing the average cell shortening (B), [Ca²⁺]_i transient decay time (C), or time to peak (D) at 4 Hz in WT (n=13) vs. RyR2^R4496C myocytes (n=19). E. Line-scan images of caffeine-evoked [Ca²⁺]_i transients obtained in WT and RyR2^R4496C (R4496C) cells obtained after field stimulation at 4 Hz. F. Average caffeine-evoked [Ca²⁺]_i transients (expressed as peak F/F₀, as in panel A) in WT vs. RyR2^R4496C) myocytes obtained following field stimulation at 2 Hz (n=33 vs. n=36), 3 Hz (n=7 vs. n=22) and 4 Hz (n= 8 vs. n=25). G. [Ca²⁺]_i transient-SR Ca²⁺ load relationship in WT and RyR2^R4496C myocytes at various pacing rates. Open circles and bars, WT; blue circles and bars, RyR2^R4496C. N numbers from panels A&F.*p<0.05.

Because [Ca²⁺]_i transient amplitude depends on SR Ca²⁺ load, we investigated the SR Ca²⁺ content. Images of caffeine-evoked [Ca²⁺]_i transients evoked after electrical stimulation at 4 Hz are shown in Figure 4E. As shown in Figure 4F, caffeine-evoked [Ca²⁺]_i transients were significantly smaller (by 24.7%) after pacing the cell at 4 Hz in RyR2^R4496C compared to WT myocytes, while no significant difference was observed at lower frequencies. Plotting the peak [Ca²⁺]_i transient vs. the SR Ca²⁺ load for the three different pacing rates (Fig. 4G) provided similar correlations in WT and RyR2^R4496C myocytes. These results suggest that a decrease in SR Ca²⁺ load accounts for the reduction in systolic [Ca²⁺]_i transients and the associated lower contraction observed at the highest pacing rates in both cell groups. Interestingly, the decrease in SR load with increasing pacing rate was accentuated in RyR2^R4496C myocytes. To get an idea of how much Ca²⁺ is released at each twitch with respect to the total amount of Ca²⁺ stored, we evaluated the fractional release by normalizing the electrically-evoked [Ca²⁺]_i transient to the caffeine-evoked [Ca²⁺]_i transient in each cell tested. We found no difference between WT and RyR2^R4496c cells. For example at 4 Hz, fractional release was 0.80±0.04 in WT cells (n=8) and 0.82±0.06 in RyR2^R4496C myocytes (n=21). Altogether, these experiments unmask rate-dependent systolic Ca²⁺-release defects in RyR2^R4496C cells in relation with impaired recovery of SR Ca²⁺ load, which might be secondary to diastolic Ca²⁺ leak.

However, the relevance of this systolic defect may be limited in vivo, since increased heart rate is usually associated with high sympathetic drive, thus increased contractility. We next analyzed [Ca²⁺]_i transients under β-adrenergic stimulation. Images of [Ca²⁺]_i transients evoked by field stimulation at 4 Hz are shown in Figure 5A, top. The increase in [Ca²⁺]_i transient amplitude induced by isoproterenol was similar in WT and RyR2^R4496C cells, (Fig. 5A, bottom, Online Fig. IIB, Online Table III). We also found a similar effect of isoproterenol on SR Ca²⁺ load in both experimental groups (Fig. 5B). Altogether, these data suggest that the systolic dysfunction at high pacing rates persists during sympathetic stimulation (p<0.01), and that the incremental effect of β-adrenergic stimulation is identical in WT and RyR2^R4496C.

Figure 5. Conserved β-adrenergic responsiveness in RyR2^R4496C myocytes.

A. Top. Line-scan images of [Ca²⁺]_i transients obtained in WT and RyR2^R4496C (R4496C) cells evoked by field stimulation at 4 Hz in the absence and the presence (ISO) of 1 μmol/L isoproterenol. Bottom. Percentage of [Ca²⁺]_i transient amplitude increase induced by 1 μmol/L isoproterenol at 2 (n=26 vs. n=25), 3 (n=4 vs. n=12) and 4 (n=3 vs. n=12) Hz, WT vs. RyR2^R4496C cells. B. Percentage of increase induced by 1 μmol/L isoproterenol on caffeine evoked [Ca²⁺]_i transients after stimulating the cell at 2 Hz (n= 26 vs. n=25), 3 Hz (n=4 vs. n=12) and 4 Hz (n=3 vs. n=12). WT (white hatched bars), RyR2^R4496C (blue hatched bars)

Nature of the arrhythmogenic activity of RyR2^R4496C myocytes

The arrhythmic activity of the isolated cell presented in Fig. 1 does not reflect the reality of ventricular myocytes, which are under constant electrical stimulation. Moreover, CPVT is induced by stress, meaning β-adrenergic stimulation and elevated heart rhythm. During electrical stimulation and in the presence of isoproterenol, the RyR2^R4496C myocytes developed spontaneous [Ca²⁺]_i transients and after-contractions, consistent with triggered activity, as opposed to the WT cells (Fig. 6A&B). During diastole, the RyR2^R4496C myocytes showed spontaneous Ca²⁺ release as Ca²⁺-sparks or small Ca²⁺-waves. Occasionally (arrow in Fig. 6B), spontaneous Ca²⁺ releases reached the threshold to produce non-sustained triggered activity and after-contractions. Ca²⁺-spark evoked Ca²⁺-waves were observed during the diastolic period in 17% of the RyR2^R4496C cells paced at 2 Hz (9 out of 54 RyR2^R4496C myocytes). The percentage of cells exhibiting these events dramatically increased to 67% (12 out of 18 cells) when RyR2^R4496C cells were paced at 4 Hz in the presence of isoproterenol. This behavior was almost absent in WT cells, both in basal conditions (2Hz: 1 out of 55 WT cells vs. 9 out of 54 RyR2^R4496c cells p<0.01) and at 4 Hz and isoproterenol (2 out of 14 WT cells vs. 12 out of 18 RyR2^R4496C cells, p<0.01). As shown in Fig. 1B, one single Ca²⁺-wave was able to evoke triggered activity after electrical stimulation stopped. However, this was never the case during constant stimulation. In order to trigger a full [Ca²⁺]_i transient, we measured that 7.1±0.6 Ca²⁺-waves*100μm⁻¹ had to overlap during diastole, which only occurred under stress conditions. Ca²⁺-waves spread at similar velocity in the absence (117.3±14.7 μm/s, n=13) or in the presence of 1μmol/L isoproterenol (113.8±7.2 μm/s, n=22).

Figure 6. Arrhythmic activity depends on β-adrenergic stimulation and pacing rate.

A. Line-scan images obtained in a WT cell paced at 2 Hz and superfused with 1 μmol/L isoproterenol. The corresponding fluorescence [Ca²⁺]_i transients and cell shortening profiles appear below. B. The same for a RyR2^R4496C cell. The image shows multiple Ca²⁺-sparks and/or Ca²⁺-waves during diastole. The red arrow indicates triggered activity, maintained in panel c and spontaneously terminated in panel d. Red lines indicate electrical stimuli. C. Occurrence of abnormal diastolic Ca²⁺ release at 2, 3 and 4 Hz, in myocytes in absence (solid bars) or in presence (hatched bars) of 1 μmol/L of isoproterenol; WT, white bars; RyR2^R4496C, blue bars.*p<0.05, **p<0.01 with respect to WT. †p<0.05 with respect to RyR2^R4496C. ‡p<0.05, ‡‡p<0.01, ‡‡‡p<0.001 with respect to WT in the presence of isoproterenol. #p<0.05, ###p<0.001 with respect to RyR2^R4496C at 2 Hz.

We found that mouse RyR2^R4496C myocytes presented a higher incidence of Ca²⁺-waves not only in the presence of isoproterenol but also at higher pacing rates (Fig. 6C). We then measured Ca²⁺-sparks in the diastolic period at different stimulation frequencies, when it was possible to discriminate them. During diastole, the maximum number of Ca²⁺-sparks*s⁻¹ in RyR2^R4496C myocytes under β-adrenergic stimulation increased with pacing rates (4.7±1.7 when paced at 2 Hz, n=9; 15.0±4.4 when paced at 3 Hz, n=4, p<0.05 compared with 2 Hz; and 24.8±5.4 when paced at 4 Hz, n=5, p<0.001 compared with 2 Hz). This could be due to the higher diastolic [Ca²⁺]_i induced by increasing pacing rates. The diastolic Ca²⁺ fluorescence measured in the same cells increased progressively with pacing rate (39.5±2.7 at 2 Hz, 46.1±1.6 at 3 Hz and 50.9±3.0 at 4 Hz, p<0.05 compared with 2 Hz).

DISCUSSION

We show for the first time that beating cardiomyocytes bearing the RyR2^R4496C mutation, equivalent to that found in several CPVT families, exhibited arrhythmogenic behavior related to a dramatically enhanced occurrence of Ca²⁺-sparks and Ca²⁺-waves during diastole. This elevated spontaneous Ca²⁺ release was further enhanced by β-adrenergic stimulation and increasing pacing rates, mimicking human exercise-induced ventricular tachycardia. The high activity of RyR2^R4496C was due to a dramatic increase in its Ca²⁺ sensitivity, which lowered the release threshold to produce spontaneous activity during the diastolic period.

Mice bearing the RyR2^R4496C mutation, which is the equivalent to the human RyR2^R4497C mutation first identified in a CPVT family¹⁴, present ventricular tachycardia in response to adrenergic stimulation and caffeine in vivo. Isolated cells were patch-clamped and action potentials recorded. Under these conditions, DADs and triggered activity could be recorded when electrical pacing was interrupted¹⁵. Here we found parallel evidence of spontaneous intracellular Ca²⁺ release and Ca²⁺-waves in similar experimental conditions (Fig. 1). However, in life, ventricular myocytes are continuously paced unless there is a problem with automatic or conducting cells. Moreover, CPVT arises under stress conditions with adrenergic stimulation which, among other effects, increases heart rate.

This report is the first to show that isolated RyR2^R4496C ventricular myocytes displayed arrhythmogenic activity related to spontaneous Ca²⁺ release while they are electrically stimulated, thus mimicking human CPVT and demonstrating that RyR2^R4496C was at the origin of the arrhythmia. In isolated cardiomyocytes paced at 2Hz, we observed multiple Ca²⁺-sparks capable of triggering localized Ca²⁺-waves in more than 16% of RyR2^R4496C myocytes (Fig. 6C). With pacing rate increased to 4 Hz and under β-adrenergic stimulation, the RyR2^R4496C myocytes were remarkably more prone to evoke Ca²⁺-waves (in up to 66.7% of cells). We thus found that RyR2^R4496C cells showed higher spontaneous Ca²⁺ release even in basal conditions, and this feature was further enhanced by β-adrenergic stimulation and pacing rate, reaching threshold for triggered activity.

The higher diastolic Ca²⁺ release in RyR2^R4496C cells is correlated by higher frequency of spontaneous Ca²⁺-sparks (Fig. 2). This increased activity could depend on the expression or phosphorylation level of RyR2, the amount of Ca²⁺ stored in the SR, and/or the sensitivity of RyR2^R4496c to luminal²¹ or cytosolic Ca²⁺. We found no difference between WT and RyR2^R4496C hearts in total RyR2 expression, FKBP12.6 association¹⁴ or RyR2 phosphorylation level, even after β-adrenergic stimulation (Online Fig. I). Moreover, although Ca²⁺-waves and high Ca²⁺-spark frequency usually reflect Ca²⁺overload¹⁸, RyR2^R4496C myocytes presented this behavior even at lower SR Ca²⁺ load.

Our data in permeabilized cardiomyocytes show that at all cytoplasmic [Ca²⁺]_i tested, the Ca²⁺-spark frequency was higher in RyR2^R4496C than in WT cells, showing that the RyR2^R4496C is hyperactive at any given [Ca²⁺]_i, and indicating Ca²⁺ hypersensitivity. However, Ca²⁺-spark frequencies in WT and RyR2^R4496C cells were similar in absence of cytosolic Ca²⁺. Under these conditions, SR Ca²⁺ load was also similar in both experimental groups, suggesting that RyR2^R4496C sensitivity to luminal Ca²⁺ is maintained under these unphysiologic circumstances. Nevertheless, in the presence of cytosolic Ca²⁺, RyR2^R4496C behaves as hypersensitive to both luminal¹⁰ and cytosolic Ca²⁺(Fig. 3). It is not easy to unequivocally assign distinct roles for cytoplasmic vs. luminal Ca²⁺ in situ due to the inherent interdependence of these Ca²⁺ compartments in living cells.

While unzipping of amino and central RyR2 domains has been reported to be involved in some forms of enhanced RyR2 activity^22,23, the R4496C mutation is far from those domains, making that mechanism unlikely. Differential FKBP12.6 association also cannot explain the increased RyR2 sensitivity reported here, because there is unaltered RyR2-FKBP12.6 association in this animal model¹⁴. The increase in Ca²⁺-spark frequency of RyR2^R4496C is likely to reflect an enhancement of its open probability (P_o), consistent with data obtained by single channel analyses¹⁰. Our data demonstrate that, in its normal environment (i.e., in native cardiomyocytes) RyR2^R4496C has augmented Ca²⁺ sensitivity rather than increased P_o per se. Indeed, Ca²⁺-spark occurrence, measured in permeabilized cells exposed to different [Ca²⁺]_i concentrations, was significantly increased in RyR2^R4496C at all [Ca²⁺]_i tested except at zero Ca²⁺, indicating that the channel needs Ca²⁺ to become hyperactive. The RyR2^R4496C mutation is located in the C terminal portion of the channel, (cytosolic side^24,25), close to the proposed molecular region involved in Ca²⁺-dependent activation (residues 4485–4494)²³. The RyR has highly reactive cysteines capable of forming disulfide bonds²⁶. It is thus plausible that the highly reactive cysteine introduced by the mutation, interacts with other cysteines of the channel, inducing a conformation change that renders the RyR hypersensitive to Ca²⁺. The conformational change might render more accessible to Ca²⁺ the E3987 residue, identified as important in Ca²⁺ sensitivity²⁷. However, the low affinity Ca²⁺ sensing of the RyR2^R4496C seems to be normal since the Ca²⁺ inhibition found in the ³[H⁺]ryanodine binding experiments is similar to WT RyR2 (Fig. 3G). Experiments in the RyR2^R4496C tertiary structure are needed to investigate whether this point mutation induces conformational changes favoring Ca²⁺ binding to the activating sites in the RyR2.

Even though RyR2^R4496C basal activity was dramatically higher than that of WT (Fig. 2), β-adrenergic stimulation increased their activity to the same extent suggesting that: (i) the two mechanisms (increased Ca²⁺ sensitivity of the RyR2^R4496C mutant and the effect of β-adrenergic stimulation) are distinct and cumulative and (ii) the β-adrenergic regulation of RyR2^R4496C is not modified. Nevertheless, β-adrenergic stimulation further increased the already elevated diastolic Ca²⁺ leak in RyR2^R4496C cells probably by increasing the SR Ca²⁺ load²⁸, further enhancing the RyR2^R4496C cell propensity to trigger DADs and allowing the occurrence of spontaneous activity (Figs. 1&6)^29,30.

At basal conditions (2Hz in our experimental setting), the [Ca²⁺]_i transients in WT and RyR2^R4496C myocytes were similar. The [Ca²⁺]_i transient decay times were also similar, suggesting a normal function of sarco-endoplasmic reticulum Ca²⁺ ATPase (SERCA). However, the rate-dependent decrease in [Ca²⁺]_i transients and contraction, which is normal in mice cardiomyocytes, was more pronounced in RyR2^R4496C (Fig. 4). This reduction can be accounted for by a decrease in SR Ca²⁺ load (Fig. 4). Therefore, the increase in diastolic Ca²⁺ leak becomes critical for the systolic function only at the highest pacing rates. Such a negative staircase, observed in normal mice, may depend on the interval between two consecutive twitches, while SERCA replenishes the SR with Ca²⁺. Enhancement of this phenomenon in RyR2^R4496C cells seems to indicate that, because RyR2^R4496C myocytes show more Ca²⁺-waves during diastole at high pacing rate, an imbalance between Ca²⁺ leak and Ca²⁺ re-uptake results in SR Ca²⁺ depletion, although a possible alteration in RyR2^R4496C refractoriness could account for this phenomenon. However, in humans the staircase is positive, which further supports the lack of contractile impairment in CPVT patients.

The decrease in SR Ca²⁺ with pacing rate can also reflect the higher Ca²⁺ leak at higher stimulation frequencies (Fig. 6C) and could partly depend on a phenomenon known as Ca²⁺ current facilitation. By this phenomenon, the total amount of Ca²⁺ entry is enhanced when stimulation frequency is increased, mainly due to slowing of the Ca²⁺ current inactivation³¹. This longer Ca²⁺ entry during the diastolic period can further activate RyR2^R4496C, thereby evoking more Ca²⁺-waves and rhythmic disorders at higher pacing rates.

In conclusion, our study shows that beating RyR2^R4496C cardiomyocytes present high spontaneous Ca²⁺ release during diastole, due to a dramatic increase in Ca²⁺ sensitivity of the RyR2^R4496C. This diastolic Ca²⁺ leak is responsible for both DADs and decreased SR Ca²⁺ load at high pacing rates. Our findings in cardiomyocytes provide a link between the data observed with the heterologous expression of RyR2^R4496C mutation and the macroscopic phenotype observed at the whole heart (ECG). By characterizing the function of mutant RyR2, we provide a detailed definition of how CPVT mutations cause DAD and triggered arrhythmias. Furthermore, the identification of abnormal Ca²⁺ sensitivity in RyR2 as the key factor for arrhythmogenesis supports the interest of the RyR2 in the development of novel therapeutic targets.