Diastolic transient inward current in Long QT Syndrome type 3 is caused by Ca²⁺ overload and inhibited by Ranolazine

Abstract

Long QT Syndrome variant 3 (LQT-3) is a channelopathy in which mutations in SCN5A, the gene coding for the primary heart Na⁺ channel alpha subunit, disrupt inactivation to elevate the risk of mutation carriers for arrhythmias that are thought to be calcium (Ca²⁺)-dependent. Spontaneous arrhythmogenic diastolic activity has been reported in myocytes isolated from mice harboring the well-characterized ΔKPQ LQT-3 mutation but the link to altered Ca²⁺ cycling related to mutant Na⁺ channel activity has not previously been demonstrated. Here we have investigated the relationship between elevated sarcoplasmic reticulum (SR) Ca²⁺ load and induction of spontaneous diastolic inward current (I_TI) in myocytes expressing ΔKPQ Na⁺ channels, and tested the sensitivity of both to the antianginal compound ranolazine. We combined whole cell patch clamp measurements, imaging of intracellular Ca²⁺, and measurement of SR Ca²⁺ content using a caffeine dump methodology. We compared the Ca²⁺ content of ΔKPQ^+/- myocytes displaying I_TI to those without spontaneous diastolic activity and found that I_TI induction correlates with higher sarcoplasmic reticulum (SR) Ca²⁺. Both spontaneous diastolic I_TI and underlying Ca²⁺ waves are inhibited by ranolazine at concentrations that preferentially target I_NaL during prolonged depolarization. Furthermore, ranolazine I_TI inhibition is accompanied by a small but significant decrease in SR Ca²⁺ content. Our results provide the first direct evidence that induction of diastolic transient inward current (I_TI) in ΔKPQ^+/- myocytes occurs under conditions of elevated SR Ca²⁺ load.

Introduction

Long QT syndrome variant 3 (LQT-3) is a rare, but lethal, inherited arrhythmia syndrome caused by mutations in SCN5A, the gene coding for the alpha subunit of the primary heart voltage gated Na⁺ channel [1, 2]. The first reported and best-characterized LQT-3 mutation causes the deletion of three amino acids (ΔKPQ) in the channel inactivation gate [3] resulting in a defect in channel inactivation and persistent late Na⁺ current (I_NaL) during maintained ventricular depolarization that underlies the electrocardiogram QT interval [4]. Initial analysis of the functional consequences of the ΔKPQ mutation [5, 6] focused on the link between the resulting action potential (AP) prolongation and the generation of “early afterdepolarizations” (EADs) which are driven by reactivation of L-type calcium channels and occur during the initial phases of ventricular repolarization [7]. More recently, attention has focused on the link between mutation enhanced Na⁺ entry during depolarization to transient inward current (I_TI) recorded at diastolic potentials after repolarization is complete [8]. I_TI drives arrhythmogenic “delayed after depolarizations” (DADs) which occur over the diastolic voltage range under conditions favoring sarcoplasmic reticulum (SR) calcium ([Ca²⁺]_SR) overload [9-11]. Thus the link between the ΔKPQ mutation and I_TI induction suggested the possibility that arrhythmia susceptibility in LQT-3 may be due to two distinct Ca²⁺ dependent mechanisms: one that may occur during repolarization (EAD) and one that may occur during diastole (DAD), but direct measurement of altered SR load in latter case previously has not been reported.

Here, in myocytes isolated from ΔKPQ^+/- mice, we have investigated the relationship between I_TI generation and [Ca²⁺]_SR and studied the effects of the antianginal agent ranolazine on both parameters. We chose to investigate ranolazine because it has previously been shown not only to preferentially inhibit I_NaL [12] but also because it is associated with improved diastolic relaxation in LQT-3 patients [13]. We find, as in the case of other perturbations that promote spontaneous diastolic arrhythmogenic activity, I_TI is associated with constitutively elevated [Ca²⁺]_SR. Ranolazine inhibits both diastolic Ca²⁺ transients and I_TI over the same concentration range as ranolazine inhibition of I_NaL. Furthermore, Ranolazine inhibition of I_TI in ΔKPQ^+/-myocytes is accompanied by a reduction in [Ca²⁺]_SR. Our results thus provide the first direct link between altered Na⁺ channel activity in ΔKPQ^+/- myocytes during membrane depolarization to arrhythmogenic activity driven by elevated [Ca²⁺]_SR over diastolic voltages: evidence for a second Ca²⁺ dependent arrhythmic pathway in LQT-3.

Materials and Methods

ΔKPQ Mice and Isolation of Cardiac Ventricular Myocytes

Mice heterozygous for a knock-in KPQ deletion in SCN5A [5] were genotyped by PCR analysis to confirm the expression of the SCN5A ΔKPQ Na⁺ channels. Ventricular myocytes from adult mice (3 to 8 months of age) were dissociated as previously described [8, 14] and used within 6 hours. The institutional Animal Care and Use Committee at Columbia University approved the protocols for all animal studies.

Solutions

Membrane currents were measured using whole-cell procedures with an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA). Capacity current and series resistance compensation were carried out according to the amplifier manufacture. Micropipette electrodes resistance was 1.5 to 2.5 MΩ. Membrane currents were low pass filtered at 2 kHz and sampled at 2 kHz. All measurements were obtained at room temperature (22°C). Na⁺ current was recorded using the following external solution (in mM): 130 NaCl, 2 CaCl₂, 5 CsCl, 1.2 MgCl₂, 10 HEPES, 0.5 BaCl₂ and 5 glucose (pH 7.4 with CsOH). Pipette filling solution was composed of (in mM): 50 aspartic acid, 60 CsCl, 5 Na₂ATP, 0.05 EGTA, 10 HEPES, 0.0215 CaCl₂, and 1 MgCl₂ (pH 7.2, with CsOH). For imaging experiments, fluo 4 K⁺ salt (Invitrogen, Eugene, OR) 0.05 mM replaced EGTA. Free [Ca²⁺] in both pipette filling solutions was buffered to 100 nM. Tetrodotoxin (TTX; Ascent Scientific, Princeton, NJ) and Ranolazine (Sigma Chemical Co, St. Louis, MO) solutions were prepared from 5 mM and 10 mM stock solutions respectively in H₂O. Because of changes of efficacy, Ranolazine stock was prepared fresh every day.

Protocols

High concentrations of TTX (50 μM) were used to block peak Na⁺ channel current (I_NaP) and late Na⁺ current (I_NaL) and reveal background currents that were digitally subtracted offline. Both I_NaP and I_NaL were defined by the mean of 5 sweeps recorded at 1 Hz and measured as TTX-sensitive currents in response to depolarizing pulses from -75 mV to -10 mV. I_NaL was measured 200 ms after the voltage step. Ranolazine block of I_Na was achieved within 2 to 3 minutes. Data were normalized to cell capacitance and to control currents for establishing the concentration-response relationships. A 1 Hz conditioning train of voltage pulses was used to trigger I_TI in ΔKPQ^+/- myocytes [8]. Briefly, it consisted of 10 depolarizing steps from a holding potential of -75 mV (-10 mV, 400 ms) followed by a 1 s pause and a last depolarization (Figure 1A). I_TI was recorded at -75 mV for a period of 10 s following the last depolarizing step of the conditioning train of voltage pulses.

Figure 1. I_TI induction in ΔKPQ^+/- myocytes correlates with elevated [Ca²⁺]_SR.

A. Schematic of conditioning voltage pulse train used to induce I_TI (Methods). I_TI and [Ca²⁺]_SR were measured at the holding potential (-75 mV) as indicated by the arrow in the diagram. B,C. Shown are high gain current recordings during the 3 last pulses of the conditioning train followed by 10 s at the holding potential in a cell which lacked I_TI (B, left) and a cell in which I_TI was induced (C, left). In the same cells, [Ca²⁺]_SR was estimated by applying caffeine (10 mM, grey bars) and integrating the caffeine-induced I_NCX (B,C right, arrows) triggered after the conditioning train of voltage pulses. D. Bar graph summarizes mean [Ca²⁺]_SR in cells without (open, n = 9) and with (closed, n = 11) I_TI (**: p < 0.05) (D). [Ca²⁺]_SR of single experiments is plotted to reveal threshold for I_TI (E). Vertical dashed line: mean [Ca²⁺]_SR estimated by others in cells that generate I_TI [18].

Data Analysis

pClamp 8.0 and 10.2 (Molecular Devices, Sunnyvale, CA), Excel (Microsoft), and Origin 7.0 (OriginLab, Northampton, MA) were used for data acquisition and analysis. Curve fitting was carried out using programs provided in the OriginLab software package. Ranolazine IC₅₀ was estimated by fitting concentration-response data (I_TI, I_NaL and I_NaP) to a logistic function (Origin). I_TI density was estimated by defining a baseline diastolic current and measuring the area under the curve for the difference between this baseline and any diastolic events. Total charges for each recording was calculated as the sum of all diastolic events occurring within 10 seconds after the conditioning train protocol and is expressed in pC/cell.

Normalization of [Ca²⁺]_SR

[Ca²⁺]_SR was estimated by integrating caffeine-induced NCX transient inward current (I_NCX). The resulting values, in pC/pF were normalized to μmol/L cytosol using the constant 121.32 derived from measurements made in 3-8 month old rats [15, 16]. Because balance between different Ca²⁺ transports in rat and mice is similar, we can assume that our normalization is valid [17]

Ca²⁺ imaging

Fluo 4 K⁺ salt was excited at 488 nm and fluorescent emission collected at >505 nm using laser-scanning confocal microscopy (Zeiss Live 5, Carl Zeiss GmbH, Jena, Germany). The linescan mode was used to record fluorescent temporal profiles. Recordings lasted 30 s and were acquired at a scan rate of 500 Hz. Resting levels of fluorescence (F0), acquired immediately before the depolarizing conditioning train of voltage pulses, were used to normalize the images. Image normalization was done using a custom made macro in imageJ (Rasband, W.S., ImageJ, U. S. National Institutes of Health, Bethesda, MD, USA, http://rsb.info.nih.gov/ij/, 1997-2008). For comparison, diastolic Ca²⁺ waves and systolic Ca²⁺ transients profiles were integrated and normalized to control conditions.

Statistics

Data are presented as mean values ± SEM. Two-tailed Student's t test was used to compare two means; a value of p<0.05 was considered statistically significant.

Results

The experiments in this study were designed to investigate a possible role of intracellular Ca²⁺ in I_TI generation in myocytes expressing Na⁺ channels harboring the ΔKPQ mutation and thus mutation-enhanced I_NaL. Because of the well documented and strong correlation found between [Ca²⁺]_SR and spontaneous Ca²⁺ release from the sarcoplasmic reticulum (SR) in other models [18], we designed experiments to investigate the relationship between I_TI induction and [Ca²⁺]_SR. We used a modified previously-described conditioning train protocol [8] to induce I_TI (see Methods) and application of caffeine to induce release of and hence measure [Ca²⁺]_SR [19, 20]. The latter technique, described in Methods, is referred to as a caffeine dump protocol. First, transient inward current (I_TI) was measured at -75 mV after applying conditioning trains of voltage pulses (-10 mV, 400 ms, 1 Hz). Using this protocol, I_TI was induced in 76 % of ΔKPQ^+/- myocytes (n = 57 vs. n = 18 without I_TI), a number similar to that which was previously reported [8]. We separated the data into two groups: data from ΔKPQ^+/- myocytes displaying I_TI (n = 11) and from those myocytes in which I_TI was not induced (n = 9) (Figure 1B and C, left). Then, in the same cells we measured [Ca²⁺]_SR. The protocol was repeated, but within 1-2 s after the final conditioning pulse, the cell was exposed to 10 mM caffeine to trigger Ca²⁺ release from the SR [19, 20]. SR Ca²⁺ release is monitored via a transient inward Na⁺/Ca²⁺ exchanger current (I_NCX) (Figure 1B and C, right). Data for similar experiments carried out in myocytes isolated from WT littermate mice are summarized in supplemental data (Figure 1).

[Ca²⁺]_SR estimated by species specific normalization of integrated I_NCX to cell volume (Methods), revealed that induction of I_TI in these myocytes correlates with a significantly higher [Ca²⁺]_SR compared to quiescent cells (quiescent: 108.2 ± 7 μmol/L of cytosol, n = 9, displaying I_TI: 134.1 ± 7 μmol/L of cytosol, n = 11, p < 0.05) (Figure 1D). Because I_TI probability depends not only on luminal SR Ca²⁺ [21] but also on RyR₂ gating [22], we estimated the [Ca²⁺]_SR threshold for spontaneous I_TI by plotting together [Ca²⁺]_SR values from single experiments (Figure 1E) [23]. In previous investigations, Diaz et al have reported a mean value for [Ca²⁺]_SR of 120 μmol/L cytosol for cells exhibiting spontaneous diastolic activity [18]. I_TI induction in the present experiments in ΔKPQ myocytes occurs in the same range of [Ca²⁺]_SR as indicated by the dashed vertical line in Figure 1E). This is also the case for I_TI induction in myocytes from WT littermates (Supplemental data). Thus, similar to other perturbations, induction of I_TI in myocytes expressing ΔKPQ Na⁺ channels occurs under conditions in which Ca²⁺ store overload promotes spontaneous Ca²⁺release [18, 21].

We next tested the effects of ranolazine on I_TI and spontaneous diastolic Ca²⁺ transients in the same cells, because ranolazine previously has been shown to have high selectivity for late (I_NaL) vs. peak (I_NaP) Na⁺ current (IC₅₀ = 15 vs 135 μM) [12], and has been reported to improve diastolic function in LQT-3 patients [13]. Figure 2 illustrates these experiments. Shown in the figure are currents (upper rows), line scans (middle rows), and averaged Ca²⁺ signals (lower rows) recorded during the last pulse of the conditioning train along with prolonged recordings at -75 mV following the conditioning train. The protocol induced I_TI (Figure 2A, arrow, upper panel), that was associated with spontaneous elevation of cytosolic [Ca²⁺] ([Ca²⁺]_i) (i.e.: Ca²⁺ wave) (Figure 2A, middle panel, arrows). Exposure of the cell to ranolazine (20 μM) inhibited I_TI (Figure 2B, upper panel) and the accompanying diastolic Ca²⁺ wave (Figure 2B, lower panel). Examination of the currents recorded during the last conditioning pulse reveals ranolazine block of I_NaL in the experiments.

Figure 2. Ranolazine (20 μM) inhibits I_TI and related diastolic Ca²⁺ wave.

Representative currents (upper rows), line scans (middle rows), and averaged Ca²⁺ signals (lower rows) recorded during the last pulse of the conditioning train protocol (Methods) and followed by 10 s at the return holding potential (-75 mV). A. Control conditions reveal I_TI induction (A, upper panel, arrow) that was associated with a Ca²⁺ wave (A, middle and lower panels, arrows). B. Exposure to ranolazine (20 μM) inhibited I_TI (B, upper panel) and the accompanying Ca²⁺ wave (B, middle and lower panels).

Summary data for this series of experiments are presented in Figure 3. In ΔKPQ myocytes, ranolazine (20 μM) significantly decreased mean I_TI density per cell by 50 % ± 8 %, from -57.9 ± 8 pC/cell in control condition to -29.5 ± 6 pC/cell after ranolazine (n = 17, p < 0.05) (Figure 3A) and the underlying fluorescent Ca²⁺ wave density (reduced by 39 % ± 10 % (n = 17, p < 0.05) (Figure 3B). Data were normalized to counterbalance cell-to-cell variability associated with Ca²⁺ dye loading. In the same cells ranolazine had no significant effect on systolic Ca²⁺ transients, measured as the integral of the fluorescent signal of the last Ca²⁺ transient during the conditioning train in the absence and presence of ranolazine (Figure 3C; CTL: 100 %, RAN: 99.6 ± 6 %, n = 17, p > 0.05). In cells isolated from WT littermates ranolazine had no significant effect either on systolic Ca²⁺ or on I_TI and underlying diastolic Ca²⁺ transients (Supplemental data, figures 2 and 3).

Figure 3. Summary data: ranolazine inhibits I_TI, diastolic Ca²⁺ wave, but not systolic Ca²⁺ transient.

Both systolic and diastolic Ca²⁺ events were determined by integrating normalized fluo-4 transient signals as described in Methods. The bar graphs summarize (A) mean I_TI density per cell (n = 18); (B) mean Ca²⁺ wave density (n = 18); and (C) mean systolic Ca²⁺ density as measured during the final pulse of the conditioning train (n = 18) in control (open) and after exposure to ranolazine (20 μM) (closed). **: p < 0.05, paired.

These data suggest that induction of I_TI is correlated with spontaneous Ca²⁺ waves that occur during diastole and that ranolazine significantly decreases both spontaneous diastolic events. We observed effects of ranolazine on both to I_TI current density measured and the frequency of I_TI occurrence. Figures 4 and 5 summarize these effects of ranolazine on I_TI over a broad concentration range. Figure 4 shows examples of I_TI recorded after conditioning trains in the absence and presence of 10 μM, 50 μM, and 1000 μM ranolazine. Figure 5 reports the group data, and focuses on the effects of ranolazine on total I_TI density. Diastolic I_TI density was summed for each cell, and compared in both conditions: control and ranolazine. At a concentration of 10 μM ranolazine failed to prevent the generation of spontaneous diastolic I_TI (CTL: -62.6 ±10 pC/cell, RAN: -48.9 ± 14 pC/cell, n = 5, p > 0.05) (Figure 5A, left). Increasing ranolazine concentration to 50 μM significantly decreased I_TI from -35 ± 4 pC/cell in control conditions to -9.6 ± 4 pC/cell (n = 8, p < 0.05) (Figure 5A, center). The maximal inhibition we observed was reached with 1000 μM (CTL: -46.8 ± 11 pC/cell, RAN: -5.3 ± 5.3 pC/cell, n = 4, p < 0.05) (Figure 5A, right). Data were fit with a logistic function shown as the smooth curve in the figure (see Methods), resulting in a predicted IC₅₀ of 19.5 ± 3 μM. Ranolazine also affects the frequency of I_TI occurrence (data not shown). There was no significant effect of 10 μM ranolazine mean I_TI frequency (0.88 ± 0.23 I_TI per cell control vs. 1.13 +/- 0.35 I_TI per cell in 10 μM ranolazine, n = 8) but 50 μM, 100 μM and 1000 μM significantly did so (1.35 ± 0.21 I_TI per cell control vs. 0.6 ± 0.27 I_TI per cell in 50 μM ranolazine, n = 10, p < 0.05; 0.89 ± 0.26 I_TI per cell control vs. 0.13 ± 0.12 I_TI per cell in 100 μM ranolazine, n = 9, p < 0.05 and 1.25 ± 0.25 I_TI per cell control vs. 0.13 ± 0.11 I_TI per cell in 1000 μM ranolazine, n = 4, p < 0.05).

Figure 4. Ranolazine inhibits I_TI in a concentration-dependent manner.

Representative high gain recordings from three different myocytes show currents in response to the 3 last pulses of conditioning trains followed by 10 s return to the holding potential before (left) and after (right) exposure to the indicated ranolazine concentrations: 10 μM (upper row), 50 μM (middle row) and 100 μM (lower row). Arrows indicate I_TI in relevant panels. Zero current is indicated (0) in each panel.

Figure 5. Ranolazine inhibition of I_TI and I_NaL share similar concentration-dependence.

A. Bar graphs show mean I_TI density per cell in control conditions (open) and after exposure to 10 μM (left; n = 5), 50 μM (middle; n = 8) and 1000 μM (right; n = 4) ranolazine (closed). **: p<0.05, paired. B. Summary plot of the concentration-dependence ranolazine inhibition of I_Na peak (●), I_NaL (○) and I_TIs (). Smooth curves are best fits of concentration-response relationships to I_NaL, I_NaP and I_TI data (see Methods).

For comparison, we also measured the effects of ranolazine on I_NaP and I_NaL in the same cells (Figure 5B). In agreement with previously published results [12], ranolazine preferentially inhibited I_NaL when compared to its effects on I_NaP. I_NaP was inhibited by 13.4 % ± 3 % (n = 8), 23.4 % ± 3 % (n = 12), 33.6 % ± 3 % (n = 9) and 87 % ± 3 % (n = 4) after perfusion with 10 μM, 50 μM, 100 μM, and 1000 μM ranolazine respectively (Figure 5B, closed black circles). I_NaL was inhibited by 29.8 % ± 4 % (n = 8, ranolazine 10 μM); 62 % ± 2 % (n = 12, ranolazine 50 μM); 79 % ± 4 % (n = 9, ranolazine 100 μM); and by 92 % ± 4 % (n = 4, ranolazine 1000 μM) (Figure 5B, open black circles). I_NaL (open) and I_NaP (filled) data were fit with logistic functions (Methods). Resulting best fits are plotted as smooth curves (Figure 5B). Estimated IC₅₀s for both sets of data, I_NaL and I_NaP (26.8 μM and 206.1 μM respectively) are in the range of previously published results [12]. The results reveal strikingly similar concentration-dependent ranolazine inhibition of I_NaL and I_TI (19.5 μM), and clearly a distinct pharmacological profile from ranolazine inhibition of I_NaP.

Because we find that ranolazine inhibits both I_TI and underlying spontaneous diastolic SR Ca²⁺ events, we next investigated the impact of ranolazine on [Ca²⁺]_SR. We simultaneously measured the impact of ranolazine (50 μM) on I_TI induction and [Ca²⁺]_SR in the same cells as illustrated and summarized in Figure 6. A conditioning train protocol (Methods) was imposed to induce I_TI, and after I_TI detection, the protocol was repeated and followed by caffeine 10 mM application (within 1-2 s after the last conditioning pulse) to estimate [Ca²⁺]_SR. In cases when I_TI appeared before caffeine (Figure 6A, right), I_TI density was summed to the caffeine-induced I_NCX to correct for [Ca²⁺]_SR loss. This procedure was carried out in the absence (Figure 6A) and then again in the presence of ranolazine (50 μM, Figure 6B). [Ca²⁺]_SR was significantly decreased by ranolazine (by 9.1 % ± 1.6 %, n = 5, p < 0.05) (Figure 6C). That this small, but significant, decrease in [Ca²⁺]_SR correlates with decreased I_TI incidence, is consistent with the known non-linearity observed between [Na⁺]_i and [Ca²⁺]_SR and Ca²⁺ waves [18, 24, 25].

Figure 6. Ranolazine decreases [Ca²⁺]_SR.

Representative high gain recordings during the 3 last pulses of the conditioning train and 10 s at the holding potential (-75mV) after the train in the absence (A) and presence of 50 μM ranolazine (B). Induction (A, left, arrow) and ranolazine-dependent inhibition of I_TI (B, left) was verified before estimating [Ca²⁺]_SR in both conditions using caffeine-induced I_NCX (A and B, right, arrows). Zero current is indicated (0) in each panel. C. Bar graph summarizes normalized mean [Ca²⁺]_SR before and after exposure to ranolazine, n = 5, p < 0.05, paired (C).

Discussion

The results presented in this study provide the first direct evidence that an inherited LQT-3 mutation (ΔKPQ) that disrupts Na⁺ channel inactivation during prolonged depolarization is associated with elevated [Ca²⁺]_SR load and consequential spontaneous electrical activity (I_TI) that occur over a diastolic potential range, after the cell has repolarized. This activity may be particularly critical to the generation of pause-dependent arrhythmias in LQT-3 patients which have been reported in the clinical literature [26].

[Ca²⁺]_SR triggers I_TI in ΔKPQ^+/-

Our results clearly show that ΔKPQ^+/- myocytes with higher [Ca²⁺]_SR are more prone to spontaneous diastolic activity than those with lower [Ca²⁺]_SR, and our results are consistent with a causal role of mutation-enhanced late current (I_NaL) in I_TI generation. We previously reported a higher incidence of I_TI induction in myocytes from ΔKPQ mice compared with their WT littermates [8]. Now, our pharmacological results show that ranolazine inhibition of dysfunctional late Na⁺ channel current (I_NaL) and I_TI in ΔKPQ myocytes share a common concentration dependence and is accompanied by reduction in [Ca²⁺]_SR load. Interestingly, I_TI, induced in myocytes from WT littermates, which lack mutation-enhanced I_NaL, is not inhibited by ranolazine, strengthening the link between I_TI, spontaneous diastolic Ca²⁺ transients and I_NaL. The efficacy of the drug in suppressing spontaneous diastolic activity is likely to be due in large part to its effects both on mutation-enhanced I_NaL and on [Ca²⁺]_SR load as supported by the nonlinearity between [Na⁺]_i and [Ca²⁺]_SR [18, 24, 25].

It is likely that I_TI sensitivity of ΔKPQ myocytes is linked to changes in intracellular sodium concentration ([Na⁺]_i) that may occur as a result of mutation-increased Na⁺ entry. Using Anemonia sulcata (ATX II) toxin-induced changes in Na⁺ entry as a pharmacological model of LQT-3, others have reported both induction of spontaneous diastolic activity and its inhibition by ranolazine [27-29] but in these models ATX II-induced Na⁺ entry is nearly 10 fold greater than that caused by the naturally occurring ΔKPQ mutation [30]. Thus the impact of mutation-enhanced late current on [Na]_i in the present experiments is expected to be subtle. But even very small changes are expected to impact [Ca²⁺]_i dynamics. Eisner, et al. showed that [Ca²⁺]_i varies as a power function of [Na⁺]_i, as assessed by tonic and twitch tension [24, 25]. Consequently even a slight increase in [Na⁺]_i can cause a steep increase in [Ca²⁺]_i. This would predict a small net increase in Ca²⁺ influx per depolarization that is augmented by mutation-altered I_NaL, and hence a net increased [Ca²⁺]_SR as determined by the balance between different cellular Ca²⁺ transport rates [31]. Correlation between elevated [Ca²⁺]_SR and spontaneous SR Ca²⁺ release has been known for many years. Diaz et al. showed that the relation between luminal SR Ca²⁺ content and Ca²⁺ stores instability is a very steep function [18]. Store instability results mostly from the sensitization of the RyR₂s by luminal SR Ca²⁺ [21, 32, 33]. Our results indicate an approximate threshold for spontaneous diastolic Ca²⁺ release in ΔKPQ^+/- myocytes that is the range reported by others for spontaneous diastolic activity generated by distinctly different perturbations (Figure 1, dashed vertical line) [18]. Taken together our data suggest that SR in ΔKPQ^+/- myocytes have normal Ca²⁺ release properties, as determined by the RyR₂ ability to open, but higher content than normal and that mutation-enhanced Na⁺ entry contributes to elevated [Ca²⁺]_SR.

Mechanism of I_TI inhibition by ranolazine

It is clear from the experiments reported in this study that ranolazine effectively and preferentially inhibits diastolic Ca²⁺ spontaneously released from the SR at concentrations that have no significant effect on SR Ca²⁺ released in response to depolarization (Figure 3). The effectiveness of ranolazine at inhibiting diastolic events is particularly interesting in view of the fact that the drug has been reported to improve diastolic relaxation in LQT-3 patients [13]. As noted above the effects of ranolazine on diastolic events share the same concentration-dependence as ranolazine inhibition of I_NaL, suggesting that pharmacologic inhibition of I_NaL should be accompanied by concomitant reduction in SR Ca²⁺ content. We in fact found a significant reduction [Ca²⁺]_SR, but the magnitude of this effect was small, approximately a 9 % reduction in [Ca²⁺]_SR (Figure 6). Nevertheless, Diaz et al reported that the relation between luminal SR Ca²⁺ content and Ca²⁺ stores instability is a very steep function, thus though small, the inhibitory effect of ranolazine on SR load is likely to be a major factor in the effects of the drug on diastolic Ca²⁺ transients and I_TI.

Ranolazine has other effects which may also contribute to its inhibition of I_TI. For example ranolazine has also been studied in a pharmacological LQT3 model [34]. Sossala et al. reported effects of ranolazine on [Ca²⁺]_SR in isolated rabbit myocytes treated with ATX-II to increase systolic Na⁺ entry via severe modulation of Na⁺ channel inactivation. In this study ranolazine significantly inhibited ATX II-induced I_NaL and reduced frequency-dependent increases in diastolic tension without affecting SR Ca²⁺ load. Though similar to our subtle effects of ranolazine on SR load, Sossala et al suggested that a ranolazine-induced increase of NCX transport might also contribute to this action. Ranolazine is also known to affect other ion channels in the heart, including Ca_v1.2 channels [35], and, in addition has been shown to have important metabolic effects in heart [36, 37], although these effects are expected to be minimal at 20 μM ranolazine.

In summary we find that expression of Na⁺ channels harboring the ΔKPQ mutation in myocytes is associated with elevated [Ca²⁺]_SR load, spontaneous diastolic Ca²⁺ release, and consequential spontaneous diastolic electrical activity (I_TI). Ranolazine effectively inhibits mutation-induced I_NaL and I_TI, and reduces [Ca²⁺]_SR in ΔKPQ myocytes. These properties make ranolazine a strong candidate for therapeutic management of LQT-3 and other disorders such as heart failure in which enhanced late Na⁺ channel activity and spontaneous diastolic Ca²⁺ release may occur [29, 38-41].

Limitations of the study

Here we report that I_TI is associated with constitutively elevated [Ca²⁺]_SR in ΔKPQ^+/- myocytes and that Ranolazine inhibits both diastolic Ca²⁺ transients and I_TI over the same concentration range as ranolazine inhibition of I_NaL. This implies a critical role of mutation enhanced I_NaL in the generation of I_TI. and spontaneous diastolic Ca²⁺ release as suggested in our previous report [8]. But, because Ca²⁺ waves are generally caused by elevated [Ca²⁺]_SR, any mechanism affecting SR load will similarly affect the I_TI probability and thus affect the results. The use of genetically engineered mice provide insight into the pathophysiology of myocytes isolated from them, but, particularly in experiments such as those described in this paper in which altered intracellular Ca²⁺ dynamics are studied it must be remembered that, while contraction in mice and rats primarily relies on [Ca²⁺]_SR, in human as well as rabbits and ferrets, contraction relies more on Ca²⁺ entry through Ca_V1.2 calcium channels. Consequently the exact role of I_NaL on [Ca²⁺]_SR and resulting I_TI generation may vary with these species-dependent differences [15, 31].

Supplementary Material

01. Supplemental Figure 1. I_TI induction in ΔKPQ^-/- myocytes correlates with elevated [Ca²⁺]_SR.

A,B. Shown are high gain current recordings during the 3 last pulses of the conditioning train followed by 10 s at the holding potential (Methods) in a ΔKPQ^-/- cell (A). In the same cell, [Ca²⁺]_SR was estimated by applying caffeine (10 mM, grey bars) and integrating the caffeine-induced I_NCX (B, arrow) triggered after the conditioning train of voltage pulses. C. Bar graph summarizes mean [Ca²⁺]_SR in cells without (open, -94.2 ± 7.6 μmol/L cytosol, n = 3) and with I_TI (closed, -112.6 ± 5.5 μmol/L cytosol, n = 7, p > 0.05) (D). [Ca²⁺]_SR of single experiments is plotted to reveal threshold for I_TI (D). Mean [Ca²⁺]_SR of spontaneously active cells as reported by the literature (120 μmol/L cytosol) is represented by the vertical dotted line for comparison with data from ΔKPQ^+/- myocytes.

02. Supplemental Figure 2. Ranolazine (20 μM) does not inhibit I_TI and related diastolic Ca²⁺wave in ΔKPQ^-/- myocytes.

Representative currents (upper rows), line scans (middle rows), and averaged Ca²⁺ signals (lower rows) recorded during the last three pulses of the conditioning train protocol (Methods) and followed by 10 s at the return holding potential (-75 mV). A. Control conditions reveal I_TI induction (A, upper panel, arrow) that was associated with a Ca²⁺ wave (A, middle and lower panels, arrows). B. Exposure to ranolazine (20 μM) failed to inhibit I_TI (B, upper panel) and the accompanying Ca²⁺ wave (B, middle and lower panels).

03. Supplemental Figure 3. Summary data: ranolazine has no significant effect on I_TI, diastolic Ca²⁺wave and systolic Ca²⁺ transient.

Both systolic and diastolic Ca²⁺ events were determined by integrating normalized fluo-4 transient signals as described in Methods. A. The bar graph summarizes the mean I_TI density per cell in control (open, -19.8 ± 5.3 pC/cell) and after exposure to ranolazine (20 μM) (closed, -15.3 ± 11.5 pC/cell, n = 3, p > 0.05, paired). I_TI was induced in 3 out of 8 cells. B. Mean normalized Ca²⁺ wave density in absence and presence of the drug (RAN: 89 % ± 15.8 %, n = 3, p > 0.05, paired). C. Mean systolic Ca²⁺ density as measured during the final pulse of the conditioning train (RAN: 141.7 % ± 28.1 %, n = 8, p > 0.05, paired).

Footnote

LQTS

long QT syndrome

LQT-3

The long QT syndrome variant 3

I_TI

transient inward current

SR

sarcoplasmic reticulum

[Ca²⁺]_SR

SR Ca²⁺ content

[Ca²⁺]_i

cytosolic [Ca²⁺]

I_NaL

persistent Na⁺current

I_NaP

peak Na⁺ current

EADs / DADs

early and delayed afterdepolarizations

ΔKPQ

KPQ deletion

TTX

Tetrodotoxin

RyR₂

ryanodine receptor

I_NCX

Na⁺/Ca²⁺exchanger current

ATX-II

sea anemone toxin