JTV519 (K201) reduces sarcoplasmic reticulum Ca²⁺ leak and improves diastolic function in vitro in murine and human non-failing myocardium

Abstract

BACKGROUND AND PURPOSE

Ca²⁺ leak from the sarcoplasmic reticulum (SR) via ryanodine receptors (RyR2s) contributes to cardiomyocyte dysfunction. RyR2 Ca²⁺ leak has been related to RyR2 phosphorylation. In these conditions, JTV519 (K201), a 1,4-benzothiazepine derivative and multi-channel blocker, stabilizes RyR2s and decrease SR Ca²⁺ leak. We investigated whether JTV519 stabilizes RyR2s without increasing RyR2 phosphorylation in mice and in non-failing human myocardium and explored underlying mechanisms.

EXPERIMENTAL APPROACH

SR Ca²⁺ leak was induced by ouabain in murine cardiomyocytes. [Ca²⁺]-transients, SR Ca²⁺ load and RyR2-mediated Ca²⁺ leak (sparks/waves) were quantified, with or without JTV519 (1 µmol·L⁻¹). Contribution of Ca²⁺-/calmodulin-dependent kinase II (CaMKII) was assessed by KN-93 and Western blot (RyR2-Ser²⁸¹⁴ phosphorylation). Effects of JTV519 on contractile force were investigated in non-failing human ventricular trabeculae.

KEY RESULTS

Ouabain increased systolic and diastolic cytosolic [Ca²⁺]_i, SR [Ca²⁺], and SR Ca²⁺ leak (Ca²⁺ spark (SparkF) and Ca²⁺ wave frequency), independently of CaMKII and RyR-Ser²⁸¹⁴ phosphorylation. JTV519 decreased SparkF but also SR Ca²⁺ load. At matched SR [Ca²⁺], Ca²⁺ leak was significantly reduced by JTV519, but it had no effect on fractional Ca²⁺ release or Ca²⁺ wave propagation velocity. In human muscle, JTV519 was negatively inotropic at baseline but significantly enhanced ouabain-induced force and reduced its deleterious effects on diastolic function.

CONCLUSIONS AND IMPLICATIONS

JTV519 was effective in reducing SR Ca²⁺ leak by specifically regulating RyR2 opening at diastolic [Ca²⁺]_i in the absence of increased RyR2 phosphorylation at Ser²⁸¹⁴, extending the potential use of JTV519 to conditions of acute cellular Ca²⁺ overload.

Introduction

In cardiac muscle, excitation-contraction (EC) coupling relies on the precise regulation of intracellular Ca²⁺. Ca²⁺ also regulates a variety of intracellular processes, such as cellular metabolism, gene transcription and cell death. Dysregulation of cardiomyocyte Ca²⁺ homeostasis has been implicated in a variety of cardiac disease states (Bers, 2008). During EC coupling, Ca²⁺ influx through voltage-dependent L-type Ca²⁺ channels triggers Ca²⁺ release from the intracellular Ca²⁺ store, the sarcoplasmic reticulum (SR), via opening of ryanodine receptors (RyR2; receptor nomenclature follows Alexander et al., 2011) in the SR membrane (Ca²⁺-induced Ca²⁺ release). During diastole, RyR2s are closed and low cytosolic [Ca²⁺]_i is restored mainly by the SR Ca²⁺ ATPase (SERCA) and the sarcolemmal Na⁺/Ca²⁺ exchanger (NCX). The contribution of NCX to cytosolic Ca²⁺ removal varies among species between 7% (rats), 28% (rabbits) and up to about 50% (failing human hearts) (Dipla et al., 1999; Bers, 2001).

Leak of Ca²⁺ from the SR through RyR2s during diastole has been implicated in arrhythmias and contractile dysfunction in heart failure (Marx et al., 2000; Lindner et al., 2002), myocardial ischaemia (Hirose et al., 2008), atrial fibrillation (Vest et al., 2005) and congenital arrhythmias (Mohamed et al., 2007). Several mechanisms have been proposed to explain RyR2 dysfunction, including altered RyR2 gating due to point mutations in the RyR2 molecule (Cerrone et al., 2005), alterations in RyR2 phosphorylation (Witcher et al., 1991; Wehrens et al., 2004b; Xiao et al., 2005) and the composition of the RyR2 macromolecular complex (Marx et al., 2000), and increased SR Ca²⁺ content (store-overload induced Ca²⁺ release) (Eisner et al., 2009). Increased RyR2 Ca²⁺ leak has been associated with PKA-dependent phosphorylation of the RyR2 at Ser²⁸⁰⁹ (Marx et al., 2000) or Ser²⁰³⁰ (Xiao et al., 2007), or Ca²⁺-calmodulin-dependent kinase II (CaMKII)-mediated phosphorylation at Ser²⁸⁰⁹ (Witcher et al., 1991) or Ser²⁸¹⁵ (Wehrens et al., 2004b) (corresponding to Ser²⁸⁰⁸ and Ser²⁸¹⁴ in mice) respectively. However, the exact role of site-specific phosphorylation of the RyR2 in promoting SR Ca²⁺ leak is still under debate (Blayney and Lai, 2009; Eisner et al., 2009). We have previously shown that cellular Na⁺- and Ca²⁺ overload induced by ouabain treatment leads to Ca²⁺ leak from the SR through the RyR2 in the absence of change in RyR2 phosphorylation at Ser²⁸⁰⁸ (Sedej et al., 2010). One aim of the present study was to investigate whether SR Ca²⁺ leak induced by intracellular Ca²⁺ overload can occur in the absence of CaMKII-mediated RyR2 phosphorylation at Ser²⁸¹⁴.

JTV519 (K201) is a 1,4-benzothiazepine derivative and anti-arrhythmic drug currently under clinical investigation. While JTV519 also inhibits sarcolemmal currents, such as sodium (I_Na), L-type Ca²⁺ and potassium currents (I_Kr, I_K1) (Kaneko et al., 2009), this multi-channel blocker has been investigated based on its potential to prevent arrhythmias related to increased Ca²⁺ leak from the SR. Heart failure and atrial fibrillation are associated with increased phosphorylation of the RyR2 (Marx et al., 2000; Greiser et al., 2009). In these conditions, JTV519 (K201) reduced RyR2-mediated SR Ca²⁺ leak and related cardiomyocyte dysfunction (Kohno et al., 2003; Wehrens et al., 2005; Toischer et al., 2010). However, whether JTV519 is effective in reducing arrhythmogenic RyR2 Ca²⁺ leak in the absence of CaMKII-mediated phosphorylation on Ser²⁸¹⁴ has not yet been explored.

Ouabain, a cardiac glycoside, increases contractile force by increasing intracellular [Ca²⁺], and at higher doses induces arrhythmias, without changing phosphorylation at the PKA-dependent phosphorylation site Ser²⁸⁰⁸ (Sedej et al., 2010). In the present study, we investigated the mechanisms by which JTV519 reduced RyR2 Ca²⁺ leak resulting from ouabain-induced increase in intracellular [Ca²⁺]. We found that SR Ca²⁺ leak induced by ouabain occurred in the absence of increased RyR2 phosphorylation at the CaMKII-dependent site Ser²⁸¹⁴. JTV519 reduced SR Ca²⁺ leak by stabilizing the RyR2 in its closed state during diastole without decreasing the gain of Ca²⁺-induced Ca²⁺ release.

Methods

The use of donor hearts and research protocols for human samples were approved by the local ethics committee (ref. nr. 20–277 ex 08/09). Investigations on human tissue conform with the principles outlined in the Declaration of Helsinki. All studies involving animals are reported in accordance with the ARRIVE guidelines for reporting experiments involving animals (McGrath et al., 2010). Animals were treated according to the Guidelines for the Care and Use of Laboratory Animals (National Institute of Health, USA).

Cell isolation

Adult FVB/N mice (11–16 weeks old; total = 43) provided by the Abt. f. Labortierkunde, Medical University of Vienna, Austria, were used for all experiments. Ventricular cardiomyocytes were isolated as previously described (Heinzel et al., 2002; Sedej et al., 2010). Mice were anesthetized with isoflurane (Abbott, Wiesbaden, Germany) and killed by cervical dislocation. The heart was quickly removed and perfused on a Langendorff setup for 4 min using Ca²⁺-free Tyrode's solution (see below) with taurine (15 mmol·L⁻¹), to which collagenase Type II (285 U·mg⁻¹, Worthington, Lakewood, NJ, USA) was then added, and perfusion continued for 7 min. Following mechanical dispersion of the cells, extracellular [Ca²⁺] was increased stepwise to reach 1 mmol·L⁻¹. Only quiescent, rod-shaped ventricular cardiomyocytes were used for experiments.

Confocal [Ca²⁺]_i measurements

Cells were studied in a perfusion chamber mounted on the stage of an inverted microscope with a Plan Neofluar 40×/1.3 oil-immersion objective and equipped with a Zeiss LSM 510 Meta confocal laser point scanning system (Zeiss GmbH, Jena, Darmstadt, Germany). Fluo-4 AM (Invitrogen, Germany) was used as Ca²⁺ sensitive dye (excitation at 488 nm, emission >515 nm). Pinhole was set to 1 Airy unit, resulting in an optical slice thickness of 0.9 µm. Pixel width was between 0.17 and 0.33 µm. To record [Ca²⁺]_i transients, a 512 pixel scan line was drawn along the longitudinal axis of the cell, avoiding scanning through the nuclei, and scanned every 1.54 ms. Consecutive scan lines were stacked over time and visualized as 2D image (line scan image).

Experimental protocol

Cells were superfused with Tyrode's solution, containing (in mmol·L⁻¹) NaCl 136, KCl 5, CaCl₂ 3, MgCl₂ 1, HEPES 10, glucose 10; pH adjusted to 7.4 with NaOH at 37°C. Cells were electrically stimulated at 1 Hz for at least 10 min. Line scan images with [Ca²⁺]_i transients were recorded at steady state of the [Ca²⁺]_i transient amplitude (baseline). In order to quantify the effects of ouabain on [Ca²⁺]_i transients and on SR Ca²⁺ leak (Ca²⁺ sparks frequency) in beating cardiomyocytes during steady-state electrical stimulation, ouabain (100 µmol·L⁻¹, Sigma Aldrich, Vienna, Austria) was washed in for 7 min to induce a moderate increase in cytosolic [Na⁺]_i (Sedej et al., 2010).

In a second set of experiments, SR Ca²⁺ leak was measured as a function of SR [Ca²⁺] at steady state of intracellular Ca²⁺ fluxes. For this, at steady-state baseline, stimulation was continued without (control) or with ouabain for 7 min. Then electrical stimulation was stopped, and external solution was rapidly switched to a Na⁺- and Ca²⁺-free solution to inhibit Na⁺- and Ca²⁺ transport via the sarcolemmal NCX [(Shannon et al., 2003); solution contained, in mmol·L⁻¹, LiCl 91, LiOH 21, KCl 4, MgCl2 1, HEPES 20, EGTA 10, glucose 10; pH 7.4 with LiOH ± ouabain as above. 20 s were allowed for a steady state of intracellular Ca²⁺ fluxes, followed by 10 s to record Ca²⁺ spark frequency. Tyrode's solution with Na⁺ and Ca²⁺ (as above) was then re-introduced within <2 s, and SR Ca²⁺ content was assessed by rapid application of caffeine (30 mmol·L⁻¹). In a subset of cells, the CaMKII-inhibitor KN-93 (1 µmol·L⁻¹; Merck, Darmstadt, Germany) was added together with ouabain. Another group of cells was pre-incubated for at least 1 h with JTV519 (1 µmol·L⁻¹) before running the protocols above. In these cells, JTV519 was also present throughout the entire protocol.

To allow quantitative spark analysis in all groups, we used supraphysiological [Ca²⁺] (3 mmol·L⁻¹) in the external Tyrode's solution (see above), because in preceding experiments only a minority (32%) of control cardiomyocytes showed Ca²⁺ sparks during the scanning period at 1 mmol·L⁻¹ external [Ca²⁺] (vs. 82% of all control cells in the present study at 3 mmol·L⁻¹ external [Ca²⁺]).

Image analysis

For the steady-state [Ca²⁺]_i transient analysis, line scan images were segmented by the onset of the whole-line averaged [Ca²⁺]_i transients and four to five consecutive transients were averaged using custom made algorithms coded in IDL (IDL 7.0, ITT Visual Information Solutions, Paris, France) as previously described (Heinzel et al., 2008; Lenaerts et al., 2009). The amplitude of the cytosolic [Ca²⁺]_i transient was calculated by normalizing the peak (F_peak) of the [Ca²⁺]_i-dependent fluorescence (F) to F averaged during 30 ms before the onset of the Ca²⁺ transient (F₀). F₅₀ was defined as the half-maximum of the [Ca²⁺]_i transient amplitude. SR [Ca²⁺] was estimated by the amplitude of the caffeine-induced [Ca²⁺]_i transient. The contribution of SERCA to the electrically stimulated cytosolic [Ca²⁺]_i transient (Twitch_SERCA%) was estimated from the decay rates (mono-exponential fit) of the caffeine-induced [Ca²⁺]_i transient (λ_caff) and the electrically stimulated [Ca²⁺]_i transient (λ_stim) in the same cell by the formula: Twitch_SERCA%= (λ_stim – λ_caff) / λ_stim× 100 (Fowler et al., 2005). Relative changes in diastolic cytosolic [Ca²⁺]_i were quantified in the same cell by dividing F₀ at steady state in the treatment phase (see above) with F₀ at steady-state baseline (F_0,bsl). This F/F₀ ratio could be slightly below 1.0 in control conditions, reflecting loss of fluorescence signal due to repetitive laser scanning. A pseudo-calibration of the Ca²⁺-dependent fluorescence signals was performed to estimate SR [Ca²⁺] in the presence of ouabain as outlined in the Appendix S1.

Ca²⁺ sparks were quantified during the decline of the [Ca²⁺]_i transient at 1 Hz stimulation at baseline and during the treatment phase, as well as during the last 10 s of long diastole following stimulation (with the drugs present). Line scan images were analyzed offline, and Ca²⁺ sparks were quantified as singular Ca²⁺ release events with a maximum width smaller than 4 µm. Time to peak [Ca²⁺]_i, peak amplitude, full width at half maximum (FWHM) and full duration at half maximum (FDHM) of Ca²⁺ sparks were calculated automatically from the local [Ca²⁺] transients. Ca²⁺ wave propagation speed was measured following linear fitting of the Ca²⁺ wave front.

Western blots

Suspensions of isolated cardiomyocytes (viability > 50%) were stimulated in an electrical field and exposed to extracellular solutions as in the treatment phase described above. A subset of cell suspensions was treated with isoprenaline (1 µmol·L⁻¹) in combination with the serine/threonine phosphatase inhibitor calyculin A (1 µmol·L⁻¹) to induce RyR2 phosphorylation at Ser²⁸¹⁴ (positive control) (Huke and Bers, 2008). After treatment, the cells were pelleted and snap-frozen in liquid nitrogen. Samples were homogenized in lysis buffer composed of (in mmol·L⁻¹) Tris–HCl 20, pH 7.4; NaCl 137, NaF 20, sodium orthovanadate 1, sodium pyrophosphate 1, β-glycerophosphate 50, EDTA 10, EGTA 1, PMSF 1, glycerol 10%, NP-40 1%, aprotinin 4 µg·mL⁻¹, pepstatin A 4 µg·mL⁻¹ and leupeptin 4 µg·mL⁻¹. 40 µg of cell homogenates were run on 4–12% gradient SDS-polyacrylamide gels and transferred to nitrocellulose membranes overnight. Non-specific binding was blocked for 1 h at room temperature using 5% dried milk in Tris-buffered saline (pH 7.4) containing 0.1% Tween-20. Membranes were probed with anti-phospho-Ser²⁸¹⁴-RyR2 antibody (Badrilla, Leeds, UK) overnight at 4°C. Anti-rabbit IgG linked with HRP (GE Healthcare, Berkshire, UK) was used as a secondary antibody. Signal was detected using the Pierce Supersignal West Pico Reagents (Thermo Fisher Scientific, Leicestershire, UK), and the optical density of the bands was determined by Quantity One software (Bio-Rad Laboratories, Hertfordshire, UK). After stripping, membranes were re-probed with the RyR-specific antibody MA3-916 (Thermo Scientific).

Human muscle strip experiments

Functional experiments were performed on 21 isolated endocardial muscle strips (trabeculae) from five non-failing hearts (ejection fraction 57 ± 4%) not suitable for transplantation (sepsis of the donor, n= 1; excessive myocardial hypertrophy, n= 1; advanced coronary artery disease, n= 1; old age, n= 2). Human endocardial trabeculae were prepared as previously described (von Lewinski et al., 2005). Donor hearts were kept in cardioplegic Tyrode's solution at 4°C containing (in mmol·L⁻¹) Na⁺ 152, K⁺ 3.6, Cl^- 135, HCO³⁻ 25, Mg²⁺ 0.6, H₂PO⁴⁻ 1.3, SO₄²⁻ 0.6, Ca²⁺ 0.2, glucose 11.2, insulin 10 I.U.·L⁻¹ and 2,3-butanedione-monoxime (BDM) 30, equilibrated with carbogen (95% O₂, 5% CO₂) to a pH of 7.4, and transported to the laboratory. BDM was included in the cardioplegic solution to protect the myocardium during transportation from the operating room and from cutting injury. The cardioplegic effects of the solution are fully reversible upon wash-out (von Lewinski et al., 2008). Trabeculae (cross-sectional area < 0.6 mm²) were dissected under a stereo-microscope. Then the trabeculae were mounted between miniature hooks, connected to an isometric force transducer (Scientific Instruments GmbH, Gilching, Germany) and superfused with modified Tyrode's solution (no BDM, stepwise increase of [Ca²⁺]_e to 2.5 mmol·L⁻¹) at 37°C. Isometric twitches were evoked by electrical stimulation (1 Hz, pulse duration 5 ms) at the preload at which maximum steady-state twitch force was achieved (L_max). JTV519 (1 µmol·L⁻¹) was added to a subset of trabeculae. The trabeculae were stimulated for 1 h before the step-wise wash-in of ouabain at the increasing concentrations indicated.

Voltage clamp experiments

A subset of cells was transferred to a recording bath chamber mounted on a confocal microscope (Zeiss LSM 700, Germany) equipped with a patch-clamp setup (Axopatch 200B, Molecular Devices, Sunnyvale, CA, USA). Cardiomyocytes were superfused at 37°C with Tyrode's solution and voltage clamped in whole-cell configuration. Pipette contained (in mmol·L⁻¹) CsAsp 120, TEACl 20, MgATP 5, MgCl₂ 0.5, HEPES 10, pH 7.2 with CsOH. To measure L-type Ca²⁺ current, K⁺ was replaced by Cs⁺ in the external solution, and cells were clamped at −40 mV to inactivate I_Na. Ca²⁺ currents were then activated by 250 ms depolarizing steps to +10 mV, during 1 Hz stimulation. Simultaneously, confocal Ca²⁺ transients were recorded as described above. Current recordings were analysed using Clampfit 10.0 (Molecular Devices). L-type Ca²⁺ current was quantified as the difference between the negative peak and the current at the end of pulse. Ca²⁺ current density was determined by normalizing the current amplitude to cell capacitance.

Data analysis

Data are shown as mean ± SEM. As experiments were performed on mice of similar age and identical genetic and domestic background, data from individual experiments and observations (cardiomyocytes in Figures 1, 2C, 3 and 5; tissue homogenates from individual hearts in Figure 2A,B; solitary Ca²⁺ sparks in Figure 4) were pooled for statistical analysis. Data from experiments on individual human trabeculae were pooled as non-failing myocardium based on in vivo global left ventricular function before explantation, irrespective of co-morbidities (see section Human Muscle Strips Experiments). The number of experiments as a basis for statistical analysis is underlined in the figure legends. Data were compared using Student's t-test (Figures 1, 4, 5B), one-way (Figures 2, 3 and 5C) or two-way (Figure 6) anova followed by the Holm–Sidak post hoc tests for multiple comparisons when an overall significance was established. Chi-squared test was used to compare proportions (Figure 3C). P < 0.05 was considered significant.

Figure 1.

Increased cellular Ca²⁺ and SR Ca²⁺ leak with ouabain in beating ventricular myocytes. (A) Line scan image of a cardiomyocyte at baseline (above, 1 Hz stimulation) and following 7 min ouabain (100 µmol·L⁻¹; below). During ouabain, the amplitude of the [Ca²⁺]_i transients is larger, and Ca²⁺ sparks occur. (B) Peak [Ca²⁺]_i, Ca²⁺ spark frequency and [Ca²⁺]_i at diastole (30 ms before upstroke of [Ca²⁺]_i, given as % of baseline) are significantly increased by application of ouabain (n= 18 cells from n= 7 mice, *P < 0.05 vs. baseline).

Figure 2.

Ouabain-induced SR Ca²⁺ leak is independent of CaMKII-activation. (A) Western blot examples of RyR2 phosphorylated at Ser²⁸¹⁴ and RyR2 total in mouse left ventricular cardiomyocyte suspensions. (B) Mean values of quantitative densitometry. While phosphorylation at the CaMKII-dependent RyR2 phosphorylation site Ser²⁸¹⁴ was not altered with 7 min exposure to ouabain (100 µmol·L⁻¹; n= 6 mice), isoprenaline (ISO) and calyculin A (CalA) induced a strong increase in phosphorylation (positive control; n= 3 mice). (C) In intact cardiomyocytes, spark frequency in the presence of ouabain was not altered by the CaMKII-inhibitor KN-93 (1 µmol·L⁻¹; n= 8 cells from three mice). *P < 0.05 significantly different from control (CTRL) and ouabain (n= 18 cells from seven mice).

Figure 3.

JTV519 decreases SR Ca²⁺ leak and SR Ca²⁺ content. (A) Confocal line scan image of [Ca²⁺]_i transients during electrical stimulation, Ca²⁺ sparks during prolonged diastole and caffeine-induced release of SR Ca²⁺ content (right side) in a mouse cardiomyocyte (see Methods for details). White line represents average [Ca²⁺]_i across the scan line. With intracellular Ca²⁺ fluxes at equilibrium during a prolonged diastole, Ca²⁺ spark frequency (B), the incidence of Ca²⁺ waves (C), diastolic cytosolic [Ca²⁺]_i (D) and [Ca²⁺]_I transient amplitude (E) were increased with ouabain (control (CTRL) + Ouabain), related to an increased SR [Ca²⁺] (F). In the presence of JTV519 (1 µM), Ca²⁺ spark frequency and diastolic cytosolic [Ca²⁺]_i with ouabain were significantly lower than in CTRL + Ouabain. JTV519 decreased the amplitude of the [Ca²⁺]_i transient and SR [Ca²⁺]. n mice or cells per group: CTRL: 7/11; ouabain: 7/18; JTV: 8/16; JTV + ouabain: 6/33. *P < 0.05 versus CTRL; ^#P < 0.05 significantly different from CTRL + Ouabain; ^§P < 0.05 significantly different from JTV519 P < 0.01 for group differences in Figure 3C.

Figure 5.

JTV519 does not affect wave propagation or fractional SR Ca²⁺ release despite inhibition of L-type Ca²⁺ influx. (A) Example of a line scan image with the slope of the Ca²⁺ wave front marked red. (B) Ca²⁺ wave propagation velocity in ouabain-treated cells was unchanged in the presence of JTV519 (ouabain: n= 12 cells from three mice; JTV519 + ouabain: n= 7 cells from three mice; one wave/cell). (C) Fractional Ca²⁺ release ([Ca²⁺]_i transient amplitude/SR [Ca²⁺]) at 1 Hz stimulation was not influenced by JTV519 (n cells as in Figure 3). Representative recording of a L-type Ca²⁺ current (D) and simultaneously recorded line scan image of the cytosolic Ca²⁺ transient (E) in a control (CTRL) and JTV519-treated cardiomyocyte (1 Hz). (F) Mean values for L-type Ca²⁺ current density and Ca²⁺ transient amplitude (CTRL: n= 7 cells from three mice; JTV519: n= 6 cells from three mice). *P < 0.05 significantly different from control.

Figure 4.

JTV519 has SR [Ca²⁺]_i– independent effects on SR Ca²⁺ leak. (A) Ca²⁺ spark frequency with ouabain as a function of SR [Ca²⁺] in the absence and presence of JTV519 (each symbol represents one cell; n cells or mice as in Figure 3). (B) In cardiomyocytes matched for similar SR [Ca²⁺] (box in A), Ca²⁺ spark frequency was reduced in the presence of JTV519 (n= 13 cells from seven mice for ouabain and n= 12 cells from five mice for JTV + ouabain). *P < 0.05 significantly different from ouabain without JTV519. Quantitative analysis of sparks from cardiomyocytes with matched SR [Ca²⁺] revealed similar Ca²⁺ spark amplitude (C) and time to peak (D), but faster Ca²⁺ decay (time to 50% decay, RT₅₀) (E) and thus shorter spark duration (full duration at half maximum, FDHM) (F) in the presence of JTV519 (n= 98 sparks with ouabain, n= 66 sparks with JTV + ouabain). *P < 0.05 significantly different from ouabain without JTV519.

Figure 6.

JTV519 effects in human non-failing muscle strips. (A) Original recordings of force development over time in control (CTRL: upper panel) and JTV519-treated muscle strips stimulated at 1 Hz in response to ouabain. Force development at baseline tended to be lower with JTV519 (see text). (B) Ouabain-induced increase in systolic force development was significantly higher in the presence of JTV519. (C) Time to half-maximal relaxation was measured at an ouabain-induced 25% increase of systolic force. Inotropy at higher ouabain concentration (1 µmol·L⁻¹) was associated with impaired relaxation (longer relaxation time) in CTRL but not in JTV519-treated muscle strips. (D) Increase in diastolic force indicating diastolic deterioration with prolonged ouabain exposure tended to occur later in JTV519-treated cells. n muscle strips/hearts: 10/5 for ouabain, 11/5 for JTV + ouabain. *P < 0.05 significantly different from control, ^#P < 0.05 significant differences between 115%, 130% and 150% diastolic force in both groups.

Materials

All chemicals were purchased from Sigma Aldrich, if not otherwise stated.

Results

Ouabain increases intracellular [Ca²⁺]_i and diastolic SR Ca²⁺ release

Treatment with ouabain (100 µmol·L⁻¹) for 7 min during 1 Hz stimulation significantly increased systolic [Ca²⁺]_i (Figure 1). Time to peak [Ca²⁺]_i was increased with ouabain (46 ± 4 ms vs. 33 ± 3 ms at baseline, P < 0.01, n= 18), whereas half-time of decay of the [Ca²⁺]_i transient was unchanged (185 ± 10 vs. 193 ± 10 ms). The frequency of Ca²⁺ sparks during the decay of the [Ca²⁺]_i transient was significantly increased in the presence of ouabain, indicating increased RyR2-mediated Ca²⁺ leak (Figure 1). Additionally, ouabain treatment significantly increased diastolic cytosolic [Ca²⁺]_i (Figure 1).

Acute treatment with ouabain is not associated with CaMKII-mediated RyR2 phosphorylation

We have previously shown that ouabain treatment does not alter phosphorylation of RyR2 at Ser²⁸⁰⁸ (Sedej et al., 2010). We now quantified RyR2 phosphorylation at the CaMKII-dependent site Ser²⁸¹⁴. The ouabain-induced increase in cellular Ca²⁺ load was not associated with altered phosphorylation of RyR2 at Ser²⁸¹⁴ (Figure 2A/B, n= 4). Isoprenaline (100 nmol·L⁻¹) and calyculin A (1 µmol·L⁻¹), however, induced a significant increase in Ser²⁸¹⁴ phosphorylation (positive control). To confirm that CaMKII activation was not involved in increased Ca²⁺ spark frequency observed after acute ouabain treatment, Ca²⁺ spark frequency was measured in isolated cardiomyocytes during prolonged diastole in the presence of the CaMKII-inhibitor KN-93 (1 µmol·L⁻¹, n= 8) and was found to be unchanged, compared with ouabain alone (Figure 2C).

Effects of JTV519 on SR Ca²⁺ leak and [Ca²⁺]_i transient

The effects of JTV519 (1 µmol·L⁻¹) on ouabain-induced SR Ca²⁺ leak were investigated during prolonged diastole to allow for a steady state of intracellular Ca²⁺ fluxes (Figure 3A-F: control: n= 11; ouabain: n= 18; JTV: n= 16; JTV + ouabain: n= 33). With sarcoplasmic Ca²⁺ reuptake and RyR2-mediated Ca²⁺ flux at balance, net SR Ca²⁺ leak remained significantly elevated with ouabain as reflected by a significant increase in Ca²⁺ spark frequency (Figure 3B) and Ca²⁺ waves (Figure 3C). SR Ca²⁺ content as assessed by caffeine-induced [Ca²⁺]_i transient was increased with ouabain (Figure 3D). In JTV519-treated cardiomyocytes (1 µmol·L⁻¹, pre-incubation for 1 h), Ca²⁺ spark frequency in response to ouabain was significantly reduced as compared with ouabain alone (Figure 3B), indicating a reduced SR Ca²⁺ leak. Similarly, in the presence of JTV519 the number of cardiomyocytes developing Ca²⁺ waves was lower (Figure 3C, 12/19 = 67% for ouabain, 6/33 = 18% for JTV519 + ouabain, P < 0.05).Ouabain induced a significant increase in diastolic cytosolic [Ca²⁺] also in the presence of JTV519. However, in the presence of JTV519, diastolic cytosolic [Ca²⁺] with ouabain was significantly lower than in the absence of JTV519 (Figure 3D).

JTV519 did not affect the [Ca²⁺]_i transient amplitude in control conditions (Figure 3E) but reduced the effect of ouabain on the systolic peak [Ca²⁺]_i transient (Figure 3E) and on SR Ca²⁺ load (Figure 3F). JTV519 did not affect the decay of the [Ca²⁺]_i transient at 1 Hz (τ: 153 ± 15 ms in control vs. 140 ± 10 ms in JTV519). The calculated contribution of SERCA to the decay tended to be reduced with JTV519 (87.9 ± 1.1% vs. 90.6 ± 1.0% in control; P= 0.089; n= 11 cells from five mice for control and n= 13 cells from four mice for JTV519).

JTV519 has SR Ca²⁺ load-independent effects on SR Ca²⁺ leak

As SR Ca²⁺ leak from the RyR2s is a function of the SR Ca²⁺ content (Shannon et al., 2002), reduced Ca²⁺ spark frequency in the presence of JTV519 and ouabain may in part be explained by a lower SR Ca²⁺ load. Figure 4A shows Ca²⁺ spark frequency as a function of SR Ca²⁺ load in ouabain-treated cardiomyocytes without and with JTV519. While JTV519-treated cardiomyocytes (open circles) were found to be more on the left side of the x axis, reflecting reduced SR Ca²⁺ load, there was considerable overlap allowing matching cardiomyocytes of both groups with comparable SR Ca²⁺ load (marked by box in Figure 4A). At similar SR Ca²⁺ load, JTV519-treated cells showed a significantly lower Ca²⁺ spark frequency as compared with cardiomyocytes treated with ouabain alone (P < 0.01; Figure 4B).

Differential effects of JTV519 on RyR2-mediated SR Ca²⁺ release

In order to evaluate the effects of JTV519 on RyR2-mediated Ca²⁺ leak in more detail and independent on its effect on SR [Ca²⁺], we analysed Ca²⁺ spark morphology in cardiomyocytes matched for SR Ca²⁺ content (Figure 4C-F; n= 78 sparks in 13 cells with ouabain and 30 sparks in 12 cells for ouabain + JTV519). Using pseudo-calibration as detailed in the Appendix S1, we calculated an amplitude of the caffeine-induced [Ca²⁺] transient of 694 ± 81 nM in control, 1550 ± 199 nM with ouabain and 681 ± 62 nM in ouabain + JTV519 based on the data in Fig 3. When comparing cardiomyocytes with similar average SR Ca²⁺ load (1241 ± 39 nM in ouabain vs. 1140 ± 131 nM in ouabain + JTV519), spontaneous Ca²⁺ sparks in the presence of JTV519 had a similar amplitude (Figure 4C), time to peak [Ca²⁺]_i (Figure 4D) and width (full width at half maximum, 1.84 ± 0.08 vs. 1.77 ± 0.06 µm), but significantly faster decay (Figure 4E) and thus reduced duration (Figure 4F), suggesting a shorter open duration of RyR2 in the presence of JTV519. Ca²⁺ waves represent SR Ca²⁺ release propagated (triggered) by cytosolic Ca²⁺ released from neighbouring RyR2 clusters. We quantified the velocity of Ca²⁺ waves in cardiomyocytes as a measure of RyR2 opening properties during Ca²⁺ release triggered by cytosolic [Ca²⁺]_i increase. Ca²⁺ wave velocity was not different in cardiomyocytes treated with JTV519+ouabain as compared to ouabain alone (Figure 5A,B), despite a reduced SR Ca²⁺ load with JTV519 (F/F₀: 5.7 ± 0.4 in control + ouabain vs. 7.4 ± 0.6 in JTV519 + ouabain, n= 6 and five cells with waves respectively; P < 0.05). This suggests that JTV519 did not lower the opening probability of RyR2s in response to an increase in cytosolic [Ca²⁺]_i. During steady-state electrical stimulation, the ratio between the amplitude of the cytosolic [Ca²⁺]_i transient and SR Ca²⁺ content (i.e. fractional Ca²⁺ release) was not altered by JTV519 (Figure 5C), again supporting the notion that JTV519 reduced the rate of spontaneous (diastolic) RyR2 opening, but not its sensitivity to cytosolic [Ca²⁺]_i.

In order to investigate the effects of JTV519 on the gain of EC coupling, we recorded whole-cell L-type Ca²⁺ influx and the triggered cytosolic [Ca²⁺] transients in control and JTV519-treated cardiomyocytes (Figure 5D-F). In the absence of ouabain, JTV519 did not affect SR Ca²⁺ content (Figure 3D). However, L-type Ca²⁺ current amplitude was significantly reduced with JTV519 (Figure 5F). On the other hand, cytosolic [Ca²⁺] transients were unchanged, confirming our observations in intact cells (Figure 3E). Together, these results suggest that the gain of EC coupling is increased with JTV519.

Effects of JTV519 on ouabain-induced contractile dysfunction in human myocardium

We evaluated the effects of JTV519 (1 µmol·L⁻¹) in conditions of ouabain-induced Ca²⁺ overload in human, non-failing, ventricular myocardium. In the contracting muscle strips, ouabain time-dependently increased diastolic force (Figure 6A), indicating an increase in diastolic [Ca²⁺]. Increasing ouabain from 0.1 to 1.0 µmol·L⁻¹ did not further increase developed (systolic) force, but instead led to impaired relaxation (Figure 6A, upper panel and Figure 6C), reflecting cytosolic Ca²⁺ overload. In JTV519-treated muscle strips, average systolic force development at baseline tended to be lower as compared with control (8.0 ± 2.0 vs. 16.3 ± 3.3 mN·mm⁻², P= 0.06; n= 11 and 10 respectively). However, the positive inotropic effect of ouabain on developed force at 0.1 and at 1.0 µmol·L⁻¹ was significantly greater in the presence of JTV519 (Figure 6B). The ouabain-induced impairment of relaxation at comparable systolic force response was prevented in the presence of JTV519 (Figure 6A lower panel, Figure 6C). Similarly, the time to deterioration of diastolic function upon prolonged exposure to ouabain tended to be longer in the presence of JTV519 (Figure 6D). Comparable to our findings in murine cardiomyocytes, these results indicated that, JTV519 could reduce Ca²⁺ overload-induced diastolic dysfunction in non-failing human myocardium.

Discussion

In the present study, we showed that ouabain-induced acute SR Ca²⁺ leak was not attenuated by CaMKII inhibition and occurred without change in RyR2 phosphorylation at Ser²⁸¹⁴. In light of our previous findings (Sedej et al., 2010), we conclude that JTV519 has an SR Ca²⁺ load-independent, RyR2-stabilizing effect that does not depend on increased RyR2 phosphorylation at PKA- and CaMKII-dependent sites. Furthermore, our data suggest that, while reducing the RyR2 open probability with regard to spontaneous openings during diastole, JTV519 does not decrease the efficacy of Ca²⁺-induced Ca²⁺ release that triggers contraction. In human ventricular muscle strips, JTV519 acts negative inotropic but improves diastolic function and preserves the inotropic response in conditions of ouabain-induced Ca²⁺ overload.

Several studies have associated an increased phosphorylation of the RyR2 at its PKA-(Marx et al., 2000; Xiao et al., 2005) or CaMKII-dependent (Ai et al., 2005; Neef et al., 2010) sites with an increased open probability of the channel resulting in SR Ca²⁺ leak. CaMKII is activated by increased cytosolic Ca²⁺ turnover and could promote SR Ca²⁺ leak also by its phospholamban-dependent effect to increase SERCA activity and thus SR [Ca²⁺] (Picht et al., 2007). In contrast, in rat cardiomyocytes, overexpression of constitutively active CaMKII even reduced SR Ca²⁺ leak (Yang et al., 2007) despite increased RyR2 phosphorylation. As a result, it has been proposed that changes in RyR2 phosphorylation may not be an accurate reflection of altered RyR2 activity (Yamaguchi and Meissner, 2007). We have shown previously that in conditions of ouabain-induced SR Ca²⁺ overload, SR Ca²⁺ leak occurs in the absence of RyR2 phosphorylation at the PKA-dependent site (Sedej et al., 2010). Our present results extend these findings, indicating that RyR2 phosphorylation at Ser²⁸¹⁴ was also not a prerequisite for increased SR Ca²⁺ leak, implying that therapeutic approaches targeted at upstream kinases modulating RyR2 phosphorylation such as CaMKII may not be protective in similar conditions of acute Ca²⁺ overload. Increased SR Ca²⁺ leak with ouabain was observed in the decay phase of the [Ca²⁺] transient in steadily beating cardiomyocytes (Figure 1B), as well as in prolonged diastole following stimulation when intracellular Ca²⁺ fluxes are at steady state (Figure 3D).

JTV519 has been shown to reduce SR Ca²⁺ leak and related arrhythmias in some (Kohno et al., 2003; Wehrens et al., 2005; Toischer et al., 2010), but not all (Liu et al., 2006) cardiac disease models. Despite recent advances in the understanding of the molecular interactions of JTV519 and the RyR2 (Tateishi et al., 2009), the molecular mechanisms by which JTV519 modifies RyR2 gating to attenuate acquired gain-of-function defects of the RyR2 are not completely understood. Dissociation of the RyR2-associated protein FKBP12.6 was thought to induce instability of the RyR2 in the closed state resulting in increased SR Ca²⁺ leak (Marx et al., 2000). JTV519 increases the association of FKBP12.6 with the RyR2 (Wehrens et al., 2005) and was reportedly ineffective in FKBP12.6 knock-out mice (Wehrens et al., 2004a). However, Hunt et al. (2007) showed that K201 (JTV519) suppressed spontaneous SR Ca²⁺ release in heterologous expression systems, irrespective of FKBP12.6 association, suggesting that JTV519 may alter RyR2 gating by inducing conformational changes of the channel protein itself. In line with these findings, Tateishi et al. (2009) found that JTV519 was able to correct domain unzipping between the central and N-terminal domain of the RyR2 to stabilize the channel.

A gain-of-function defect of the RyR2 can also result from point mutations in the channel protein, as found in patients presenting with catecholaminergic polymorphic ventricular tachycardia (CPVT). In knock-in mice carrying the RyR2^R4496C mutation, JTV519 (1 µmol·L⁻¹) was not able to prevent arrhythmias induced by the β-adrenoceptor agonist isoprenaline, either in vitro or in vivo (Liu et al., 2006). We have recently shown in the same model of CPVT, that delayed afterdepolarizations and spontaneous action potentials can also be induced by ouabain. Interestingly, ouabain-induced cellular arrhythmias could be prevented by JTV519 treatment (Sedej et al., 2010), suggesting that SR Ca²⁺ leak triggered by isoprenaline and store-overload induced SR Ca²⁺ leak induced by ouabain result from different pathophysiological states of the RyR2 complex. Based on these findings, we chose to further explore the effects of JTV519 underlying ouabain-induced arrhythmias in the mouse model. JTV519 significantly attenuated the increase in Ca²⁺ spark frequency. In JTV519-treated cells ouabain still induced an increase in diastolic cytosolic [Ca²⁺] which may be attributed to NCX-mediated Ca²⁺ influx upon ouabain-induced Na⁺/K⁺-ATPase inhibition. However, diastolic cytosolic [Ca²⁺] with ouabain was significantly lower in JTV519 treated cells as compared with control cells, reflecting less contribution of SR Ca²⁺ leak to diastolic cytosolic [Ca²⁺] in the presence of JTV519. On the other hand, JTV519 also reduced SR Ca²⁺ content. In rat cardiomyocytes, JTV519 dose-dependently reduced the rate of Ca²⁺ uptake by SERCA (Loughrey et al., 2007); however, at 1 µmol·L⁻¹ JTV519, SR Ca²⁺ content remained unchanged in rats as well as in dogs (Kohno et al., 2003; Loughrey et al., 2007). Similarly, the authors reported unchanged amplitude of the L-type Ca²⁺ current at 1 µmol·L⁻¹. In contrast, in our study, L-type Ca²⁺ currents were significantly reduced (Figure 5F), and we found a trend towards a small decrease in SERCA activity already at 1 µmol·L⁻¹ JTV519. Our results suggest that in murine cardiomyocytes, L-type Ca²⁺ channels and maybe also SERCA are more susceptible to inhibition by JTV519.

Given that SR Ca²⁺ leak is a function of SR [Ca²⁺] (Shannon et al., 2002), reduced SR Ca²⁺ load with JTV519 may contribute to its anti-arrhythmic effect in conditions of Ca²⁺ overload. However, in cardiomyocytes matched for SR Ca²⁺ load in the presence of ouabain, Ca²⁺ spark frequency was still significantly lower in JTV519-treated cardiomyocytes, confirming that JTV519 reduced RyR2-mediated Ca²⁺ leak, independently of SR Ca²⁺ content. In line with a specific effect on RyR2 open time, at similar SR Ca²⁺ load, the duration of individual Ca²⁺ sparks was significantly shorter with JTV519 (Figure 4E/F).

Altered gating properties of the RyR2 may be related to an altered sensitivity to SR luminal [Ca²⁺] (Kubalova et al., 2005) or altered sensitivity to triggering [Ca²⁺]_i at the cytosolic site of the receptor (Marx et al., 2000). While Ca²⁺ spark frequency at similar SR Ca²⁺ was reduced with JTV519, we found that the propagation velocity of cytosolic Ca²⁺ waves was unchanged despite lower SR Ca²⁺ load in the presence of JTV519. Thus, at an average amplitude of the caffeine-induced [Ca²⁺]_i transients of ∼1200 nM, reflecting elevated SR [Ca²⁺] (Howlett et al., 2006; Ozdemir et al., 2008; Dybkova et al., 2011), cytosolic Ca²⁺ release events of similar amplitude (Figure 4C) were at least as effective in triggering Ca²⁺ release from neighbouring RyR2 in the absence and presence of JTV519. Our results suggest that luminal rather than cytosolic Ca²⁺ sensitivity of the RyR2 was reduced with JTV519. In line with this assumption, fractional Ca²⁺ release was unchanged by JTV519 (Figure 5C). Simultaneous measurement of L-type Ca²⁺ currents and cytosolic [Ca²⁺] transients revealed that [Ca²⁺]; transients of similar amplitude were triggered by significantly lower Ca²⁺ influx in the presence of JTV519 (Figure 5), suggesting that the gain of Ca²⁺-induced Ca²⁺ release was increased during treatment with JTV519. Inhibition of L-type Ca²⁺ channels by JTV519 has been reported earlier (Inagaki et al., 2000; Loughrey et al., 2007). In contrast, however, Loughrey et al. (2007) reported that, in rat cardiomyocytes, acute (90 s) application of JTV519 (1 µmol·L⁻¹) induced a reduction in Ca²⁺ wave propagation velocity (permeabilized cells) and fractional Ca²⁺ release (whole-cell patch clamp). Inhibition of L-type Ca²⁺ currents was only observed at higher concentrations of JTV519 (3 µmol·L⁻¹). The reason for these differences are unclear but may be related to species differences, time dependent effects (1 h pre-incubation with JTV519 in the present study) and recording conditions (intact vs. permeabilized/dialysed cells, 3 mmol·L⁻¹ external [Ca²⁺] in the present study).

In a recent study, Toischer et al. (2010) investigated the acute effects of JTV519 on diastolic function in terminally failing human myocardium. In failing trabeculae, 1 µmol·L⁻¹ JTV519 decreased systolic (developed) tension, whereas at a lower dose (0.3 µmol·L⁻¹) developed tension was increased. JTV519 decreased diastolic tension with higher external [Ca²⁺] or higher stimulation frequency, indicating that diastolic dysfunction, was attenuated by JTV519. In the present study, we characterized for the first time the effects of JTV519 in non-failing human ventricular muscle strips. As in failing human myocardium (Toischer et al., 2010) and in accordance with what we observed in murine cardiomyocytes, 1 µmol·L⁻¹ JTV519 was negatively inotropic. The positive inotropic effect of ouabain, however, was significantly enhanced by JTV519, suggesting that in the presence of JTV519, more of the ouabain-induced additional intracellular Ca²⁺ was retained in the SR for release with each beat. Accordingly, JTV519-treated cardiomyocytes were relatively protected from diastolic dysfunction (prolonged relaxation time, Figure 6), reflecting cytosolic Ca²⁺ accumulation at higher doses of ouabain. Thus, in non-failing human ventricular myocardium, JTV519 protects from diastolic dysfunction in vitro induced by cardiomyocyte Ca²⁺ overload.

In summary, increased PKA- or CaMKII-dependent phosphorylation at Ser²⁸⁰⁸/Ser²⁸¹⁴ of mouse RyR2 was not required for increased SR Ca²⁺ leak during acute ouabain exposure. JTV519 decreased SR Ca²⁺ leak induced by ouabain through its specific effects on RyR2 opening properties, and protected the ventricular myocardium from Ca²⁺-overload-induced diastolic dysfunction in vitro and from arrhythmogenic events. Our results extend the potential use of JTV519 in conditions of acute cellular Ca²⁺ overload.

Glossary

λ_caff

decay rate of the caffeine-induced cytosolic [Ca²⁺]_i transient

λ_stim

decay rate of the electrically stimulated cytosolic [Ca²⁺]_i transient

CalA

calyculin A

CaMKII

Ca²⁺-calmodulin-dependent kinase II

CPVT

catecholaminergic polymorphic ventricular tachycardia

EC

excitation-contraction

F₀

minimal fluorescence signal during end-diastole (treatment phase)

F₀

_bsl, minimal fluorescence signal during end-diastole (baseline)

FDHM

full duration at half maximum

F_diast

normalized diastolic Ca²⁺-dependent fluorescence signal

F_peak

systolic peak amplitude of the Ca²⁺-dependent fluorescence signal

FWHM

full width at half maximum

L_max

length at maximum steady-state force

pRyR

phosphorylated RyR

RyR2

ryanodine receptor 2

SERCA

sarco-endoplasmic reticulum Ca²⁺ ATPase

SR

sarcoplasmic reticulum

Twitch_SERCA%

percentage contribution of SERCA to cytosolic [Ca²⁺]_i transient decay

Conflict of interest

None declared.

Supporting information

Additional Supporting Information may be found in the online version of this article:

Appendix S1 Extended methods section.

Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.