Ca²⁺ signaling in human induced pluripotent stem cell-derived cardiomyocytes (iPS-CM) from normal and catecholaminergic polymorphic ventricular tachycardia (CPVT)-afflicted subjects

Abstract

Derivation of cardiomyocytes from induced pluripotent stem cells (iPS-CMs) allowed us to probe the Ca²⁺-signaling parameters of human iPS-CMs from healthy- and catecholaminergic polymorphic ventricular tachycardia (CPVT1)-afflicted individuals carrying a novel point mutation p.F2483I in ryanodine receptors (RyR2). iPS-CMs were dissociated on day 30–40 of differentiation and patch-clamped within 3–6 days. Calcium currents (I_Ca) averaged ~8 pA/pF in control and mutant iPS-CMs. I_Ca-induced Ca²⁺-transients in control and mutant cells had bell-shaped voltage-dependence similar to that of I_Ca, consistent with Ca²⁺-induced Ca²⁺-release (CICR) mechanism. The ratio of I_Ca-activated to caffeine-triggered Ca²⁺-transients was ~0.3 in both cell types. Caffeine-induced Ca²⁺-transients generated significantly smaller Na⁺–Ca²⁺ exchanger current (I_NCX) in mutant cells, reflecting their smaller Ca²⁺-stores. The gain of CICR was voltage-dependent as in adult cardiomyocytes. Adrenergic agonists enhanced I_Ca, but differentially altered the CICR gain, diastolic Ca²⁺, and Ca²⁺-sparks in mutant cells. The mutant cells, when Ca²⁺-overloaded, showed longer and wandering Ca²⁺-sparks that activated adjoining release sites, had larger CICR gain at −30 mV yet smaller Ca²⁺-stores. We conclude that control and mutant iPS-CMs express the adult cardiomyocyte Ca²⁺-signaling phenotype. RyR2 F2483I mutant myocytes have aberrant unitary Ca²⁺-signaling, smaller Ca²⁺-stores, higher CICR gains, and sensitized adrenergic regulation, consistent with functionally altered Ca²⁺-release profile of CPVT syndrome.

1. Introduction

Recent breakthroughs in stem cell biology have made it possible to develop pluripotent stem cells from adult fibroblasts by transfecting them with a set of 4 “stemness” genes (inducible pluripotent stem cells, iPSC, [1]). This reprogramming allows experimental approaches that drive such cells to acquire cardiac molecular and electrophysiological phenotypes [2–4], thus creating opportunities for therapy of a host of cardiac pathologies using patient-derived cells. This approach has made it also possible to examine patient-specific mutations in ion channels and Ca²⁺ signaling proteins that might lead to arrhythmia and heart failure in iPS-CM in a laboratory setting, thus devising pharmacological patient-specific paradigms for therapy [5–11]. In light of such potentials, it is imperative that the electrophysiological and Ca²⁺ signaling properties of human iPS-CM as well as their pharmacology are fully identified and quantified.

Ca²⁺-signaling in mammalian hearts is characterized by: (1) I_Ca-gated Ca²⁺-release (CICR), providing for the characteristic bell-shaped voltage-dependence of Ca²⁺ transients that closely reflect the voltage-dependence of I_Ca; (2) the gain of CICR is voltage-dependent, not predicted from a strictly Ca²⁺-dependent process [12–14]; (3) β-adrenergic agonists enhance I_Ca, Ca²⁺ content of the sarcoplasmic reticulum (SR), Ca_i-transients and accelerate their decay kinetics, consistent with PKA-mediated phosphorylation of DHPRs, phospholamban/SERCA2a complex, and the RyR2; (4) Caffeine-triggered Ca²⁺-release activates an inward current (I_NCX) with time-course and kinetics similar to rise and fall of cytosolic Ca²⁺, reflecting the efflux of Ca²⁺ on the electrogenic Na⁺–Ca²⁺ exchanger (NCX). Although there are already a number of reports on the electrophysiology of iPS-CM [15–17] there are few detailed reports on their Ca²⁺ signaling pathways and their regulation beyond measurements of Ca²⁺ transients in intact non-voltage clamped cells [9,18] and in embryonic stem cell-derived cardiomyocytes [19].

In this report, we describe the Ca²⁺ signaling properties of human iPS-CM by quantifying the activities of Ca²⁺-signaling proteins that include the density, kinetics, and regulation of Ca²⁺ channels and NCX transporter, the size of SR Ca²⁺-stores, its regulation by β-adrenergic agonists, the voltage-dependence of I_Ca and Ca²⁺-transients, the gain of CICR, the efficiency of Ca²⁺-release mechanism, and the properties of the individual dyadic calcium release (sparks). In addition, we have attempted to quantify possible abnormalities in these parameters in cells derived from a patient afflicted with catecholaminergic polymorphic ventricular tachycardia (CPVT), carrying a recently identified ryanodine receptor mutation (p.F2483I) [9]. Our data suggests that Ca²⁺-signaling properties of adult cardiac myocytes are closely replicated in human iPS-CM. That is, I_Ca-gated SR Ca²⁺-release is the primary mechanism for the release of Ca²⁺ on depolarization of the cell by the action potential. Relaxation, in a manner similar to mammalian myocardium, is mediated by reuptake of Ca²⁺ into the SR and extrusion of Ca²⁺ by the Na⁺–Ca²⁺ exchanger, producing currents often in excess of 2–3 pA/pF. While adrenergic agonists strongly enhanced I_Ca, and accelerated the rate of decay of the Ca²⁺-transients, they had insignificant effects on NCX currents, consistent with findings in adult mammalian hearts [20,21]. These findings led us to conclude that human iPS-CM represent a reliable Ca²⁺-signaling model of mammalian cardiomyocytes.

Numbers of recent reports have implicated RyR2-mutations and the resultant abnormal Ca²⁺ signaling in development of arrhythmia and sudden death associated with intense adrenergic stimulation in patients with CPVT. It has been proposed that such mutation renders the RyRs “leaky” on exposure to β-adrenergic agonists (hyperphosphorylation & dissociation of calstabin from RyR2, [22,23] producing localized increases in Ca²⁺ that is extruded on NCX generating local depolarization (EADs & DADs), triggering at times fatal arrhythmias. Alternatively, overloading of SR Ca²⁺-stores by adrenergic agonists has been proposed [24,25] to lead to increased probability of RyR2 channel openings, resulting in abnormal release of Ca²⁺ and the resultant membrane-depolarization and arrhythmias. These ideas have been tested in a number of knock-in mice and in vitro models [26–29], but their validity in the human disease remain still somewhat clouded by both the variability of the RyR2 point mutations producing CPVT, the locus of phosphorylation on RyR2, lack of universal confirmatory results, and absence of clear-cut pharmacology [25].

Our data here suggests that despite significant quantitative intercellular differences in Ca²⁺-signaling parameters of control and RyR2 mutant cells, they both had equivalent and elevated densities of Ca²⁺ currents and NCX activity, similar bell-shaped voltage-dependence of I_Ca-gated Ca²⁺-release, and voltage-dependent CICR gain. Mutant cells, however, were consistently found to have smaller caffeine-triggered Ca²⁺-stores higher CICR gain, especially at −30 mV, consistent with longer, recurrent and often wandering Ca²⁺-sparks, compared to sporadic and brief sparks of control iPS-CM. Though adrenergic agonists produced equivalent and large enhancements of I_Ca in both mutant and control cells, they differentially altered the CICR gain, diastolic Ca²⁺, and Ca²⁺-sparks in mutant cells consistent with aberrant Ca²⁺-release profiles of Ca²⁺-overloaded CPVT-mutant myocytes, and the higher proclivity for generation of DADs and EADs in mutant hearts.

2. Methods

2.1. Cultivation of human iPS cells and preparation of cardiomyocytes

2.1.1. Culture of undifferentiated human iPS cells

The human iPS cell lines were derived from dermal fibroblasts of a CPVT-afflicted patient carrying a de novo heterozygous autosomal dominant p.F2483I mutation in RYR2 and a healthy subject. The generation, cardiac differentiation and characterization of these cell lines were reported recently [9]. The iPS cells were maintained on mitomycin C treated murine embryonic fibroblasts (MEF) prepared in our laboratory in DMEM/F12 medium supplemented with Glutamax, 20% knockout serum replacer, 1% nonessential amino acids (NAA), 0.1 mM β-mercaptoethanol (βME, Invitrogen, Darmstadt, Germany), 50 ng/ml FGF-2 (PeproTech, Hamburg, Germany). Cells were passaged by manual dissection of cell clusters every 5–6 days.

2.1.2. Cardiac differentiation

Cardiac differentiation of human iPS cells was carried out on the murine visceral endoderm-like cell line (END₂), which was provided by C. Mummery (Leiden University Medical Center, The Netherlands). END₂ cells were mitotically inactivated for 3 h with 10 μg/ml mitomycin C (Sigma–Aldrich Chemie GmbH, Munich, Germany) and 1.2 × 10⁶ cells were plated on 6 cm dishes coated with 0.1% gelatin one day before initiation of iPS cell differentiation. To initiate co-cultures, iPS cell colonies were dissociated into clumps by using collagenase IV (Sigma-Aldrich, 1 mg/ml in DMEM/F-12 at 37 °C for 5–10 min). The differentiation was carried out in 1% knockout-DMEM containing 1 mM L-glutamine, 1% NAA, 0.1 mM βME and 1% penicillin/streptomycin (100 U/ml and 100 μg/ml, respectively). The co-culture was left undisturbed at 37 °C/5% CO₂ for 5 days. First medium change was performed on day 5 and later on days 9, 12 and 15 of differentiation. Spontaneously contracting clusters were dissociated into single cardiomyocytes for experiments.

2.1.3. Preparation of iPS-CM for patch-clamp experiments

Beating areas were micro-dissected mechanically at day 30–40 of differentiation, dissociated with collagenase B, and single iPS-CM then plated on fibronectin (2.5 μg/ml)-coated glass coverslips in 6 well plates. Cells were incubated for 36–72 h before their use in electrophysiological experiments.

2.2. Measurements of cellular currents and global Ca²⁺

iPS-CM were voltage-clamped in the whole-cell configuration. L-type Ca²⁺ current (I_Ca) and I_NCX were activated by depolarizing pulses or exposure to caffeine. The voltage-clamped cells were dia-lyzed with a Cs⁺-based, moderately Ca²⁺-buffered pipette solution containing (in mM): 110 Cs⁺-Aspartate, 15 or 5 NaCl, 20 TEACl, 5 Mg-ATP, 0.2 EGTA and 0.1 Fluo-4 pentapotassium, 0.1 CaCl₂ ([Ca²⁺]_i ~ 100 mM), 10 glucose and 10 HEPES (titrated to pH 7.2 with CsOH; measured osmolarity: 295 mOsm) allowing simultaneous measurements of intracellular Ca²⁺ transients. L-type Ca²⁺ current (I_Ca) was measured by depolarization to 0 mV from a holding potential of −50 or −40 mV using a Dagan amplifier and pClamp (Clampex 10.2) software. Borosilicate patch pipettes were prepared using a horizontal pipette puller (Model P-87, Sutter Instruments, CA). The pipettes had a resistance of 3–5 MΩ. The extracellular solution in the experimental chamber contained (in mM): 137 NaCl, 5.4 KCl, 2 CaCl₂, 1 MgCl₂, 10 glucose and 10 HEPES (titrated to pH 7.4 with NaOH). To facilitate recordings of I_Ca and I_NCX, we blocked K⁺ currents, not only by including TEA in the dialyzing solution, but also by using an electromagnetically controlled puffing system that applied K⁺-free solutions in the immediate vicinity of the voltage-clamped cell. I_Ca was modulated by adding isoproterenol or Bay-K 8644 to the K⁺-free puffing solutions while rapid application of 3 mM caffeine was used to probe the magnitude of SR Ca²⁺ stores and activate I_NCX. I_Ca was normalized relative to the membrane capacitance and plotted in units of pA/pF. All experiments were carried out at room temperature (22–24 °C).

Cellular Ca²⁺ transients in voltage-clamped cells were measured fluorometrically by including 0.1 mM Fluo-4 pentapotassium salt in the dialyzing pipette solution. The dye was excited by 460 nm light from a LED-based illuminator (Prismatix, Modiin Ilite, Israel) and Ca²⁺-dependent fluorescent light (>500 nm) was detected with a photomultiplier tube that placed behind a moveable, adjustable diaphragm, which served to limit the area of detection to the voltage-clamped cell. The cellular fluorescence signals (ΔF/F₀) were normalized by dividing the changes in whole-cell fluorescence (ΔF) with the baseline fluorescence (F₀).

2.3. Ca²⁺ imaging measurements

2.3.1. TIRF imaging and analysis of ratiometric images

Single isolated beating cardiomyocytes derived from iPS cells were plated on non-coated WillCodish, glass bottom Petri dishes (Ted Pella Inc., 35 × 10 mm, 22 mm glass) and allowed to attach for three days before imaging. Intracellular Ca²⁺ signals were measured with the fluorescent Ca²⁺-indicator dye Fluo-4AM (5 μM, Invitrogen) after 30 min incubation at 37 °C and 5% CO₂. The cells were imaged using a Leica multicolor total internal reflection fluorescence (TIRF) imaging system (Leica Microsystems, Buffalo Grove, IL) fitted with a 63× oil-immersion objective lens and an Andor iXon3 camera with 512 × 512 pixels. An argon ion laser was used for excitation at 488 nm and fluorescence emission was measured at wavelengths >515 nm. Cells were imaged at 70–100 Hz with a depth of penetration less than 200 nm into the cell and focus on sub-sarcolemmal Ca²⁺ release in regions where the cell membrane was attached to the underlying glass cover slips. The pixel size in the object plane was 0.25 μm square, or 0.157 μm with 1.6 zoom [30,31]. Images were binned using 2×2 pixel averaging. The external bath solution contained (in mM): 137 NaCl, 4 KCl, 10 HEPES, 10 Glucose, 1 MgCl₂, 2 CaCl₂ with pH 7.4 using NaOH. Sparks were measured in control vs. CPVT-iPS-CM in control bath solution and in response to 3 min application of 100 μM 8-Br-cAMP.

The obtained image sequences were collected and surveyed with Leica software (LAS AF), but displayed and analyzed in detail using a custom-designed program (Con2i). The method of analysis is illustrated in Fig. 1. Initially each frame was subjected to 2 × 2 or 3 × 3 pixel averaging. This was followed by scaling that equalized the fluorescence intensity throughout the diastolic intervals (Panel A). An average diastolic image (AVE, F₀(x,y), Panel B) was then calculated based on a large number of selected diastolic images with little or no indication of Ca²⁺ release activity. Such averaged images showed the footprints of the examined cells that typically had well defined outlines around regions of stable attachment where the fluorescence intensity varied significantly. This variability may reflect the stationary distribution of intracellular organelles and some undulation of the cell membrane and extracellular matrix. To compensate for the stationary variability each frame was divided by the average (#380/AVE, F(x,y,t)/F₀(x,y), Panel B). The resulting ratiometric images showed a fairly uniform intensity as a mottle of dark blue hues which extended throughout the cell except for hotspots of focal Ca² release activity that were displayed in warmer colors of green, yellow, orange and red (arrows). Panel C shows a sequence of frames (#28–43) where a Ca²⁺ spark first appeared abruptly as a small bright spot (~1 μm across, frame #29) that grew in size (#30–33) and flared up at a slightly different location (#34) before it finally faded and disappeared (#43). The time course of each Ca²⁺ sparks could be follow by defining a region of interest based on its first appearance (see Figs. 7, 9 and supplement 1). This approach was used to evaluate the duration of Ca²⁺ sparks as the intervals where the intensity exceeded 50% of the peak value (Fig. 7C). To chart the evolution of Ca²⁺ sparks and detect shifts from one release site to another, we also used an algorithm that approximated each image of local Ca²⁺ release with a Gaussian distribution (Panel D) and followed changes in location, amplitude, variance and helicity from frame to frame. This approach allowed the location of Ca²⁺ release sites to be determined with an accuracy of 0.1–0.3 μm depending on the signal-to-noise ratio.

Fig. 1.

Automated analysis of Ca²⁺ sparks based on TIRF-imaging. (A) Equalization of the diastolic fluorescence intensity. The image (ROI) shows a single frame with superimposed regions of interest that were marked in a semi-transparent manner in red (cell) and blue (surroundings). The fluorescence intensities in these regions were plotted versus time (red and blue curves), and were approximated throughout the diastolic intervals by a black curve and a line that together determined a scale factor that varied with time and was used to compensate for the slow decline in fluorescence that often occurred in the intervals between beats. (B) Ratiometric images. The distribution of fluorescence intensity in each frame (e.g. #38) was divided by an average fluorescence intensity calculated based on multiple selected fluorescence images without noticeable Ca²⁺-release activity. (C) A sequence of partial ratiometric images where the development and decay of local Ca²⁺ release is shown in bright colors on a dark blue mottled background representing the resting Ca²⁺ activity. (D) Determination of unitary properties of Ca²⁺ release based on a Gaussian approximation. The two sample images (#29 and #34) are shown with superimposed white circles that mark the location and standard deviations of Gaussian approximations. The curves above and to the left of the images show how well horizontal and vertical mid sections through the images were approximated by Gaussian curves. The graph at the bottom of Panel D quantifies the horizontal shift in center of the Gaussian approximations as seen also in the curves above the sample images. The shown partial frames correspond to 30 × 30 pixels or 9.4 μm × 9.4 μm.

2.3.2. Confocal Ca²⁺ imaging

Voltage-clamped iPS-CM dialyzed with fluorescent and non-fluorescent Ca²⁺ buffers (0.2 mM Fluo-4, 0.5 mM EGTA, with 0.5 mM Ca²⁺) were imaged at a frame rate of 120–240 Hz on a Noran Odyssey confocal microscope as previously described [31]. The focal plane was adjusted to the midsection of the cell intersecting its nuclei.

2.4. Confocal immuno-fluorescence imaging of RyR2 distributions

Immuno-labeled RyR2 in fixed iPS-CMs were imaged on a Leica confocal microscope following published procedures [32], but using a primary antibody to RyR2 (1:100 anti RyR Mab by Thermo Scientific). Briefly, immunocytochemical approaches were used to detect RyR2 in cultures of control, IMRC8 iPSC-CM. Media was removed and cells were rinsed with phosphate-buffered saline (PBS), and then fixed with ice cold methanol for 10 min. After washing the cells with PBS, the cells were blocked in 1% bovine serum albumin and 0.1% Triton in PBS. Then the blocking solution was removed, and primary antibody (1:100 anti RyR Mab by Thermo Scientific) was applied overnight at 4 °C in block. After washing out the cells 3 times with PBS, anti-mouse FITC (1:100) in block was applied for 2 h at room temperature. The secondary antibody dilution was removed, washed, and the cover slip was mounted on a slide using a drop of ProLong Gold with DAPI and allowed to cure for 24 h at room temperature and stored at −20 °C. Imaging was performed on a Leica TCS SPS AOBS confocal microscope system using a 488 nm argon ion laser to examine FITC expression of RyR2 and a 405 nm diode laser to examine nuclear DAPI expression using a 63× oil objective with zoom function.

2.5. Statistical analysis

Average values are presented in histograms and in the text as the mean ± the standard error of the mean for “n” cells. The distribution of data in individual cells is shown in separate panels. T-test was used to determine statistical significance. Significant findings are labeled with one (p < 0.05, *) or two stars (p < 0.01, **).

3. Results

3.1. Calcium current in iPS-CM from healthy and CPVT subjects

Cardiomyocytes from two control human iPS cell lines (clones 5 (C5) and 8 (C8)) and one CPTV iPS cell line (clone 1, NP0014-C1)) were used in the comparative electrophysiological and Ca²⁺-signaling experiments. To approximate as closely as possible, the internal media of intact contracting cells, I_Ca was measured in cells dialyzed with low Ca²⁺-buffered solutions (in mM: 0.2 EGTA, 0.1 Fluo4, 0.1 Ca²⁺, see Section 2), containing either 5 or 15 mM Na⁺. Fig. 2D, shows that I_Ca averaged about 8 pA/pF in control and mutant cells, in bathing solutions containing 2 mM Ca²⁺, with significant distribution in I_Ca density between individual cells of three cell lines (Fig. 2A). The variability in the current density could not be attributed solely to the age of cells in culture nor to cell size (cell capacitance, indicated at the bottom of each column) or intracellular Na⁺ concentrations (Panels B and D).

Fig. 2.

Distribution of calcium currents (I_Ca) density and average time constant (tau) in control (IMR-C8) and mutant (NP0014-C1) iPS-CM. Cells were cultured for 3–6 days. I_Ca was recorded from a holding potential −40 mV with step depolarization to 0 mV. The number at the bottom of each bar indicates the capacitance of individual cells. (Panels A and B) Distribution of I_Ca density in control and mutant iPS-CM recorded with [Na⁺]_i at 5 or 15 mM Na⁺_i. C: Average fast (tau f) and slow (tau s) time constants of inactivation of I_Ca in control and mutant iPS-CM recorded with 5 mM Na⁺ in the pipette solution. (D) Average I_Ca density of control and mutant iPS-CM and representative I_Ca traces from each group. The average membrane capacitance values were: Panel A, IMR-C8, 32.7 ± 2.3 pF, NP0014, 43.6 ± 3.7 pF, and Panel B, NP0014, 45.8 ± 4.2 pF.

The time course of inactivation of I_Ca in low Ca²⁺-buffered dialyzing solutions was best fit with two exponentials (tau_fast = ~10 ms, and tau_slow = ~60 ms, Fig. 2C). Increasing the intracellular Na⁺ from 5 to 15 mM had little effect on the magnitude or kinetics of I_Ca. Mostly, spontaneously beating single cells were selected for Ca²⁺ signaling experimentations.

3.2. I_Ca-gated Ca²⁺-release

Fig. 3A and B compares the voltage-dependence of I_Ca and the accompanying Ca_i-transients in representative control and mutant cells. In a manner similar to that observed in mammalian myocardium for developed tension and Ca_i-transients [33–35], a bell-shaped voltage-dependence was found for both parameters that activated at −30 mV, peaked at ~0–10 mV and decreased back to baseline levels between +60 and +80 mV (middle panels in Fig. 3A and B). I_Ca reactivated on repolarizations from positive potentials (+60 to +80 mv) also triggered Ca²⁺-release (inset traces, Panels A and B) as was first described for the rat ventricular myocytes [34]. This strict dependence of Ca²⁺-transients on I_Ca and not on the direction of voltage change, is considered as phenotypic cardiac EC-coupling characteristic, distinguishing cardiac from skeletal muscle that is primarily regulated by voltage-induced Ca²⁺-release (VICR) mechanism [33–35]. The findings of Fig. 3A and B suggest that human iPS-CMs express only the cardiac-type EC-coupling. Although there was considerable variability as to the degree of development of the maintained components of Ca_i-transients in both mutant and control cell lines, Ca_i-transients always decreased at positive potentials and activated Ca_i-transients on repolarizations from these potentials (60–80 mV, inset traces, Fig. 3A and B). Comparison of control and mutant cells did not reveal significant differences in the rate of activation and decay or the voltage dependence of I_Ca-gated Ca_i-transients, suggesting no apparent global EC-coupling defects resulting from p.F2483I RyR2 point mutation under baseline conditions.

Fig. 3.

Voltage-dependence and subcellular distribution of I_Ca -activated Ca²⁺ -transients. Representative current–voltage (I–V) relations normalized relative to the membrane capacitance and the corresponding fluorescence (Fluo-4) Ca²⁺ signal recorded from control (C5, Panel A) and mutant (NP0014, Panel B) iPS-CM. Currents were recorded with a 250 ms step depolarizations from holding potential of −50 mV in 10 mV steps to +60 mV. The middle panel shows the I–V curves for I_Ca and the corresponding Ca²⁺ Fluo-4 signal. The internal solution contained 5 mM Na⁺, and was Ca²⁺-buffered with 0.1 mM Fluo-4, 0.2 mM EGTA, and 0.1 mM Ca²⁺. The right panel shows the I_Ca and fluorescence traces at +60 or +70 mV activating rises in Ca²⁺ on repolarization. (C) Confocal image of baseline fluorescence (Fluo-4, F₀). (D) Color-coded regions of interest (ROI) corresponding to nuclei (n), patch electrode (e) and cytoplasmic regions with increasing distances from the cell membrane (red, orange, green). (E) Differences in the time course of the normalized Ca²⁺-dependent fluorescence (ΔF/F₀). The color coded traces corresponds to the ROI in Panel D. The 2-D confocal fluorescence images were recorded at 120 Hz using a focal plane intersecting the nuclei. The initial response is shown on an expanded time scale. The voltage-clamp pulse (from −60 to −30 mV) and the resulting (Na⁺ and Ca²⁺) membrane currents are shown at the top. (F) Confocal image of immunofluorescence labeled RyR2 (green) in a small cluster of control iPS-CM with DAPI-labeled nuclei (blue).

Confocal spatial imaging of I_Ca-triggered Ca²⁺ release of Fig. 3C–E shows that Ca²⁺ release within the first 25 ms of depolarization occurred primarily near the cell surface membrane before rising in the interior cytoplasmic space. Fig. 3C shows the typical representation of the rounded appearance of a bi-nucleated (n, n) voltage-clamped Fluo-4 dialyzed iPS-CM, where step-depolarization produced regional Ca²⁺ signals that developed more quickly around the perimeter of the cell (red trace in Fig. 3C representing the red mask of Fig. 3D) than in its deeper layers (orange, green) or nuclei (blue). The release of Ca²⁺ from superficial SR Ca²⁺ stores is consistent with Fig. 3F, showing an immuno-fluorescence distribution of RyRs in a cluster of iPS-CMs suggesting punctate clustering of RyR2s near the surface membrane. These distributions are reminiscent of adult atrial or immature ventricular cardiomyocytes [36], but differ from those of adult ventricular cells where synchronous activation is supported by a t-tubular network [31] as occasionally observed in a small subset of iPS-CM (Suppl. Fig. 2).

3.3. SR Ca²⁺-stores

3.3.1. Fractional Ca²⁺-release

Fractional Ca²⁺release, in the context of this study, is defined as the ratio of I_Ca- to caffeine-triggered Ca_i-transients. This ratio is generally close to unity (0.6–0.9) in adult mammalian ventricular cells under our Ca²⁺ loading conditions, but significantly lower (<0.4) in atrial cells or with reduced Ca²⁺ loading of the SR [37,38]. Since Ca²⁺ release in cardiac myocytes depends on the magnitudes of both I_Ca and Ca²⁺ stores, the ratio of Ca²⁺ released by I_Ca and caffeine (representing near-full release) may be taken as the effectiveness of Ca²⁺channel-gated release. Fig. 4 quantifies the fractional Ca²⁺ release in control and mutant myocytes dialyzed with 5 or 15 mM Na⁺ (Fig. 4C). The higher intracellular concentrations of Na⁺ used in the low Ca²⁺-buffered pipette solutions are thought to increase the myocyte Ca²⁺-load [35]. The fractional release averaged ~0.3, in control and mutant cells (Fig. 4C), with some cells showing ratios as high as 0.6 and 0.9 (Fig. 4A and B). The fractional release was not significantly altered in 15 mM Na⁺ dialyzed cells in either control or mutant cells (Panel C). Fig. 4D shows representative traces of Ca_i-transients activated by I_Ca or triggered by caffeine recorded in control and RyR2-mutant myocytes, (upper and lower traces, respectively).

Fig. 4.

Fractional Ca²⁺ release in control and mutant iPS-CM. Fractional release was calculated by dividing the I_Ca -triggered Ca_i -transient by that generated by application of caffeine (ΔF_(ICa)/ΔF_(Caff)). Cells were voltage-clamped from −50 to 0 mV or were held at −50 mV while subjecting them to 0.5 s long 3 mM caffeine pulses. (A and B) Distribution of the value of efficiency in control (A) and mutant iPS-CM (B). The numbers at the bottom of each bar indicate the membrane capacitance of each cell. (C) Average fractional Ca²⁺ release with 5 and 15 mM Na⁺ in the pipette solution. (D) Representative Ca²⁺ signals triggered by I_Ca or caffeine in control (a) and mutant (b) iPS-CM.

3.3.2. SR calcium load

To quantify the magnitude of SR Ca²⁺ stores iPS-CMs were subjected to rapid puffs of caffeine containing solutions. In whole-cell patch-clamped iPS-CMs caffeine applications produced large transient rises in the global Ca²⁺ accompanied by activation of large inward I_NCX. Fig. 5A and B represents tracings of the time course of caffeine-triggered Ca_i-transients and I_NCX in control and CPVT-mutant cells. Although there was significant variability in the magnitude of caffeine-triggered Ca²⁺ release between individual cells within control (Panel C) and mutant (Panel D) cell lines, on average, there was markedly smaller caffeine-triggered Ca²⁺release in mutant compared to control cells (Panel E), suggesting significantly smaller SR Ca²⁺ stores in the mutant cells. Consistent with this idea, the Na⁺–Ca²⁺ exchanger current (I_NCX) representing the electrogenic extrusion of Ca²⁺, as cytosolic Ca²⁺ rises by caffeine-triggered Ca²⁺ release, was also significantly larger (~2.2 pA/pF) in control compared to the mutant cells (~1.5 pA/pF) suggesting smaller caffeine-triggered Ca²⁺-stores in mutant cells. The time course of Ca²⁺ rise and fall was also quite variable in different cells, but there was no consistent change in the duration of the release or its relaxation rate between control and mutant cells.

Fig. 5.

Caffeine-releasable Ca²⁺ stores. (A–E) Simultaneous measurements of caffeine-induced I_NCX current and Ca²⁺ -dependent fluorescence (ΔF/F₀) in control (IMR-C8) and mutant (NP0014) iPS-CM dialyzed with 5 mM Na⁺ and superfused with standard Tyrode’s solution. Cells were voltage-clamped to −50 mV and exposed rapidly to 3 mM caffeine for 500 ms. (A and B) Representative caffeine-induced NXC currents and corresponding fluorescence Ca²⁺ signal from control (a) and mutant iPS-CM (b). (C and D) Distribution of I_NCX values in control and mutant iPS-CM. (E) Average values of caffeine-activated Ca²⁺ signals (top) I_NCX currents (bottom) in each group. Stars indicate significance levels (*p < 0.05, **p < 0.01). (F and G) Effect of intracellular Na⁺ on the amount of Ca²⁺ released from SR in response to application of caffeine. Average values of Ca²⁺ release from SR calculated from the integral of the caffeine-induced I_NCX in control and mutant iPS-CM with 5 or 15 mM Na⁺ in the internal solution. (G) Average valves of gain factor and I_Ca density at −30 and 0 mV. The gain factor is plotted in units corresponding to the fraction in % of the caffeine-induced Ca²⁺ release that is release by a Ca²⁺ current with a density of 1 pA/pF.

Fig. 5F quantifies the rise in cytosolic Ca²⁺ as measured by the integral of I_NCX activated by caffeine-triggered Ca²⁺ release in control and mutant cells as the cellular Na⁺ load was increased from 5 to 15 mM. This procedure enhances the entry of Ca²⁺ on NCX leading to increased Ca²⁺-load of the SR. Note that increasing Na⁺ from 5 to 15 mM enhanced only slightly the Ca²⁺ load of the SR in control cells. In sharp contrast, however, the elevation of [Na⁺]_i in mutant myocytes increased significantly the Ca²⁺ load of the SR. This proclivity to Ca²⁺ loading in mutant myocytes may contribute to leakiness of SR and arrhythmogenesis under the in vivo conditions where mutant RyR2 would become more susceptible in causing localized releases of Ca²⁺ and activation of early and delayed after-depolarization (EADs and DADs).

3.3.3. Gain of I_Ca-gated Ca²⁺-release

The degree to which I_Ca triggers Ca²⁺ release normalized for the Ca²⁺ content of the SR store is defined as the gain of CICR [12]. The gain has been used as a parameter to evaluate the effectiveness of CICR under different physiological and pharmacological conditions in cardiac myocytes [13,14]. Surprisingly, the gain of I_Ca-induced Ca²⁺ release in mammalian myocytes is voltage-dependent such that more Ca²⁺ is released per Coulomb of Ca²⁺ influx at voltages negative to zero (high gain) than at positive voltages (0 to +40 mV, lower gain). Irrespective of the mechanisms suggested by different investigators [12], the voltage-dependence of CICR is unique to cardiac EC-coupling, as I_Ca-gated Ca²⁺-release in neurons, for instance, shows no voltage-dependence [39]. To test the voltage-dependence of the gain factor in the human iPS-CMs, we quantified Ca²⁺ release triggered by I_Ca as a function of SR Ca²⁺ store at two potentials: −30 mV, where I_Ca is marginally activated, and at 0 mV, where I_Ca is near its maximal value. Fig. 5G illustrates the quantification of the gain factor at both −30 and 0 mV, in 7 control and 15 mutant myocytes. In a manner similar to adult cardiomyocytes the gain was significantly higher at −30 than at 0 mV, in both control and mutant myocytes (Fig. 5G). The CICR gain though equivalent at 0 mV was significantly larger at −30 mV in mutant compared to control myocytes. Considering the smaller Ca²⁺ content of the SR in mutant myocytes (Fig. 5), this finding suggests a more sensitive Ca²⁺ release mechanism in mutant cells for equivalent I_Ca densities (Fig. 2), especially under conditions that would increase the SR load.

The gain factor was calculated using Eq. (1):

$$\text{Gain}=\frac{100\times \mathrm{\Delta }{F}_{I_{\text{Ca}}}}{\mathrm{\Delta }{F}_{I_{\text{Caff}}}\times I_{\text{Ca}}}$$ (1)

(where ΔF_ICa and ΔF_Icaff are the Ca²⁺-signals produced, respectively, by activation of I_Ca and exposure to caffeine) averaged ~20 in CPVT mutant compared to ~12 in control iPS-CM at −30 mV (Fig. 5G). The higher gain of mutant cells at the threshold of activation of I_Ca may contribute to the instability of release mechanism at voltages near resting potentials, especially under Ca²⁺-overload conditions.

3.3.4. Modulation of SR-stores by isoproterenol

In this set of experiments, we quantified the effect of isoproterenol on the magnitude of caffeine-triggered Ca²⁺ stores and found in control cells a significant enhancement in the magnitude of the store in the presence of isoproterenol (Fig. 6A–E). Interestingly, although the caffeine-triggered Ca²⁺ releases (ΔF/F₀) were significantly enhanced by Isoproterenol (Panel E, upper bar-graphs), the accompanying I_NCX were significantly reduced (lower bar-graphs) as may be expected by isoproterenol-induced enhancement of recirculation fraction (R_f) of released Ca²⁺ back into the SR [33]. An attempt to test similar effects of isoproterenol on the size of the caffeine-induced Ca²⁺ transients in mutant cells produced recordings that were difficult to quantify and too variable to yield conclusive results. The variability in the caffeine-releasable fraction of the store in mutant cells treated with isoproterenol appeared to be caused by increased frequency of spontaneously occurring releases in the presence of the hormone that altered significantly the size of the available release-stores, prior to application of caffeine. Fig. 6F shows recordings from mutant cells where such spontaneous Ca²⁺ releases (arrows) are seen interlaced with those activated at 5 s intervals by I_Ca. As shown in Fig. 6H, the spontaneous Ca²⁺ releases were abundant only when the Na⁺ concentration was increased from 5 to 15 mM, in which case their frequency was larger in mutant (NP0014) than in control cells (WT) and was further increased by isoproterenol. Panel G shows accompanying minor changes in baseline fluorescence (Diastolic Ca²⁺). These findings are consistent with idea that Ca²⁺ overload of the myocyte is a critical in manifestations of CPVT syndrome.

Fig. 6.

Effects of isoproterenol. (A–E) Caffeine-induced I_NCX current in control iPS-CM (IMR-C5) recorded with 5 mM Na⁺_i before and after exposure to isoproterenol. (A and B) Representative I_NCX currents and corresponding fluorescence Ca²⁺ signal before (A) and after (B) exposure to 75 nM isoproterenol. (C and D) Distribution of Ca_i -transients (ΔF/F₀) and I_NCX values before (C) and after (D) exposure to isoproterenol. (E) Comparison of average values of ΔF/F₀ and I_NCX in each group (*p < 0.05, **p < 0.01). (F and G) Effects of isoproterenol on spontaneous Ca²⁺ release activity in regularly depolarized (0.2 Hz) dialyzed with pipette solutions containing 5 or 15 mM Na⁺. (F) Changes in Ca²⁺ transients (Fluo-4), I_Ca and I_NCX at the onset of exposure to 100 nM isoproterenol in mutant iPS-CM (NP0014-C1) dialyzed with 5 (top) or 15 mM Na⁺ (bottom). (G and H) Effects of isoproterenol on baseline Ca²⁺ (F₀, Panel G) and the fraction (in %) of cells with spontaneous Ca²⁺ releases (intervening between the I_Ca -triggered transients; Panel H) of control (WT) and mutant (NP0014) biPS-CM) dialyzed with 5 or 15 mM Na⁺ before and after exposure to isoproterenol. The numbers of examined cells are shown in parentheses.

3.4. Focal Ca²⁺ releases and Ca²⁺ sparks

In intact spontaneously beating cells from control and CPVT-mutant incubated in Fluo-4AM containing solutions, we measured the properties of focal Ca²⁺-release using two-dimensional confocal imaging with moderate success. Much clearer and more distinct Ca²⁺-sparks and focal releases were, however, observed when we used a TIRF imaging system (Leica Inc.) with higher spatial resolution (0.1 μm, Fig. 1). Measurements of sparks in control and mutant cells, Fig. 7C, shows a histogram of spark-duration suggesting that control sparks rarely lasted longer than 20 ms, while mutant sparks frequently reached durations of 100 ms or more, consistent with longer mean open time of RyR2s.

Fig. 7.

Ca²⁺ sparks in control (A and D) and mutant (B and E) iPS-CM. (A and B) From left to right the panels show: (1) Images of average diastolic fluorescence distributions, (2) color coded regions of interests corresponding to locations of Ca²⁺ sparks, and (3) the time course of the normalized fluorescence intensity at these locations. The histogram in Panel C shows the distributions of the duration of Ca²⁺ sparks measured at half peak amplitude for control cells (black, average = 40.4 ± 3.5 ms, n = 76) and mutant cells (red, average = 89 ± 7.5 ms, n = 50). (D and E) Image sequences showing the evolution of Ca²⁺ sparks. The colored labels correspond to the traces in Panels A and B.

Our method of analysis made it possible to determine the location Ca²⁺ sparks in the focal plane with an accuracy of ~0.1 μm (see Section 2 and Fig. 1). Considering that the imaging resolution in the vertical direction, determined by the evanescent field of illumination was also in the order of 0.1 μm, we achieved an overall resolution that revealed distinct sites with properties that were often preserved from one Ca²⁺ release event to the next. Fig. 7D and E suggests that the Ca²⁺ release sites may be classified in different categories. One type of Ca²⁺ release site produced rare and brief Ca²⁺ releases that with imaging speed of 70–100 Hz produced strong, highly localized fluorescence hot-spots in only a single frame (Panel D, Control, #116). These sparks faded and spread in 1–2 frames, but remained centered at the same location. At a second type of Ca²⁺ release site, the events were equally brief, but recurred every few hundred milliseconds (D: Control #56; E: mutant #31). While the latter sites were found in both control and mutant cells, sites with much longer lasting release events were found predominantly in the mutant cells. In some cases the epicenter of the release site remained stationary in several frames (E: Mutant #21 and 103), yet the Ca²⁺ release often appeared to terminate abruptly rather than gradually (Suppl. Fig. 1D). In other cases it leaped from site to site (E: Mutant # 109, Suppl. Fig. 1E). The long-lasting wandering Ca²⁺ release events in the mutant cells may generate inward I_NCX that in turn may predispose for EADs and DADs. We consistently found that different Ca²⁺ release sites had distinct properties that were repeatable at different times, but varied from site to site. It may be significant to note that Ca²⁺ release sites in mutant cells had a broad spectrum of Ca²⁺ release times that included the normal brief sparks possibly supporting the idea that the mutation does not by itself cause prolonged wandering Ca²⁺releases but does so only when other factors come into play.

3.5. Pharmacology of iPS-CM

Pharmacological agents known to activate adrenergic cascade or serve as agonists of Ca²⁺ channel enhance I_Ca and potentiate Ca_i-transients in cardiac muscle. Fig. 8 shows the effectiveness of isoproterenol in enhancing I_Ca and I_Ca-triggered Ca_i-transients in control and CPVT-mutant myocytes. Generally we found that isoproterenol while doubling I_Ca, only slightly (~20%) enhanced the Ca²⁺ transients (Fig. 8C and D). Similarly Bay K 8644 also potentiated I_Ca strongly, but only moderately enhanced the Ca_i-transients (Suppl. Fig. 3). There appeared to be no significant differential potentiating effects of these agents in control vs. the mutant cells.

Fig. 8.

Effects of adrenergic stimulation on I_Ca, Ca_i -transients, and gain factor in control and mutant iPS-CM. (A and B) I_Ca traces and the corresponding Ca²⁺ fluorescence in control and mutant iPS-CM before (lack) and after (red) treatment with Isoproterenol. Cells were depolarized from −40 mV to 0 mV. (C and D) Average values of peak I_Ca and ΔF/F₀ in each group before and after exposure to isoproterenol. (D and E) Effects of isoproterenol on the gain factor at −30 and 0 mV. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

We also analyzed the extent to which isoproterenol altered the gain of CICR in control and mutant myocytes. Fig. 8 E quantifies the gain factor in control and mutant myocytes at −30 mV where the gain is high and at 0 mV where the gain is lower, as observed in adult ventricular myocytes and confirmed in Fig. 5G in iPS-CMs. Interestingly, isoproterenol appeared to lower the gain at both voltages in control, but not in CPVT-mutant myocytes. Such differential effect of isoproterenol in control and CPVT-mutant myocytes may contribute to higher sensitivity of mutant RyRs to be activated in presence of adrenergic agonists as compared to control cells.

3.6. Effect of dBcAMP on Ca²⁺ signals in intact mutant iPS-CMs

Using TIRF microscopy we found that adrenergic stimulation of mutant iPS-CMs often caused Ca²⁺-overload. In other cells, that tolerated exposure to 8-Br-cAMP, or di-butryl cAMP for 1–3 min, we generally observed increase in the frequency of spontaneous beating (Fig. 9A vs. E), and occasionally periods of increased diastolic Ca²⁺ release activity in the form of bursts of Ca²⁺ sparks and low level Ca²⁺ waves (B vs. F). The location of Ca²⁺ sparks were obtained by computerized analysis and are shown as maps superimposed on sample images showing Ca²⁺ sparks (C vs. G) or the onset of synchronous activation of Ca²⁺ release (D vs. H). These maps suggest that isoproterenol may increase the tendency of focal Ca²⁺ releases to wander or jump (white lines) from one frame to the next. The sites of the mapped Ca²⁺ sparks (white triangles) showed little correlation with the highly reproducible regions of synchronous Ca²⁺ release (D and H) suggesting that the spontaneous diastolic Ca²⁺ sparks may occur at sites that are not fully integrated in I_Ca-induced Ca²⁺ release mechanism.

Fig. 9.

TIRF imaging of Ca²⁺ sparks in mutant iPS-CM before (A–D) and after (E–H) 3 min exposure to 100 μM dBcAMP. From top to bottom the matched panels show: The time course of cellular Ca²⁺ transients (A and E), Ca²⁺ sparks at selected color coded sites (during the diastolic interval showed in gray above, B and F), and maps of the locations of Ca²⁺ sparks superimposed on sample ratiometric images of Ca²⁺ sparks (C and G) or the onset of Ca²⁺ release (D and H at the times indicated by *s in A and E). Connecting lines show movements of the center of release from one frame to the next.

4. Discussion

The major finding of this study is that spontaneously beating cells derived from human skin fibroblast show similar Ca²⁺ signaling properties as those of adult mammalian hearts, i.e. they express a robust I_Ca-gated CICR signaling pathway modulated by adrenergic and Ca²⁺ channel agonists, Table 1. Nevertheless, there were significant quantitative but not qualitative differences in the density of I_Ca, the SR Ca²⁺ load, and the response to adrenergic and Ca²⁺ channel agonists among the control as well as mutant cells, perhaps as a consequence of developmental stage of the myocytes. The most prominent change in the Ca²⁺signaling profile of p.F2843I RyR2-mutant cells was their consistently smaller caffeine-triggered Ca²⁺ stores and higher CICR gain, resulting most likely from higher frequency of recurrent longer and spatially wandering sparks. Although isoproterenol was equally effective in enhancing I_Ca, Ca_i-transients, and caffeine-triggered Ca²⁺-stores in control and mutant cells (Figs. 6 and 8), it differentially increased the diastolic Ca²⁺ levels, and the frequency of spontaneous Ca²⁺ releases in mutant cells (Fig. 6F). Given the altered Ca²⁺ signaling profile of CPVT-mutant cells, it is likely that such cells when exposed to conditions that increase the SR Ca²⁺ load, such as adrenergic stimulation, higher beating frequency, and increased cytosolic Na⁺ load, would render the mutant myocytes in intact heart more susceptible to triggering of DADs or EADs, and arrhythmogenesis. In this study we have attempted to perform a detailed investigation of Ca²⁺ signaling parameters at room temperature (22–24 °C) in iPS-CMs derived from a control subject and from a single patient with the novel F2843I mutation of the cardiac ryanodine receptor. Thus it remains to be examined how the cells may perform at physiological temperatures and although CPVT is a Mendelian disease, there may be modifying factors that impose on the role of RYR2 in EC-coupling and therefore cause variability from one patient to another carrying the same mutation. It should be also noted that working with single cells did not allow direct examination of arrhythmogenic characteristics that depend on electrical communication between cells.

Table 1.

Comparison of baseline Ca²⁺ signaling parameters and the effects of adrenergic stimulation in control and mutant iPS-CM. The upper part of the table shows a comparison of Ca-signaling parameters in control and mutant iPS-CM while the lower part shows the effects of adrenergic stimulation in each of these two groups.

	Control iPS-CM	F2483I mutant iPS-CM
ICa density (pA/pF)	8.5 ± 1.3	8.6 ± 1.0	← →
Fractional Ca2+ release	0.26 ± 0.05	0.4 ± 0.09	← →
SR Ca2+ stores
ΔF/F0 (Caff)	6.3 ± 1.0	2.7 ± 0.5	↓
INCX (pA/pF)	2.13 ± 0.27	1.32 ± 0.13	↓
CICR gain
−30 mV	12.7 ± 2.5	21.5 ± 2.5	↑
0 mV	3.8 ± 0.7	4.4 ± 0.5
Ca2+ sparks	Short and sporadic	Longer and wandering
Adrenergic stimulation
ΔICa	% 90.6 ± 19.6* ↑	% 138.8 ± 40.3* ↑
CICR gain
−30 mV	30–18* ↓	27–19*
0 mV	9.0–5.0* ↓	5.4–5.1
Diastolic Ca2+ (F0)
5 Na+i	0.13–0.14* ↑	0.1–0.13* ↑
15 Na+i	0.18–0.19* ↑	0.17–0.2* ↑
% increase in # cells generating spontaneous Ca2+ release
5 Na+i	3% ↓	9% ↑
15 Na+i	18% ↑	24% ↑

^*indicates significance at the p<0.05 level.

4.1. Ca²⁺ stores and their regulation: control vs. mutant

In cardiac-type EC-coupling there are three critical elements that contribute to the effectiveness of CICR: the density of I_Ca, the activation-state of RyRs, and magnitude of the Ca²⁺ stores. This has been amply demonstrated by many investigators and is confirmed here by the bell-shaped voltage-dependence of Ca_i-transients reflecting the voltage-dependence of I_Ca (Fig. 3). This characteristic appears to be fully intact in CPVT-mutant myocytes. Interestingly, the fractional Ca²⁺ release seems to be higher in mutant cells, but the effect was not statistically significant (Fig. 4C).

A novel and consistent finding of our studies, also reported for P2328S iPSC-CMs [40] and in R4496C knock-in mice model [27], was that the caffeine-triggered Ca²⁺-stores were significantly smaller in mutant cells (Fig. 5E). The lower Ca²⁺ store content is likely to be caused by enhanced mean open time of RyRs (leakiness) reported in almost every animal model of CPVT, irrespective of the alternative mechanisms proposed. It is clear that a lower Ca²⁺-content of SR in hearts with hyperactive RyR2s would serve as a protective compensatory mechanism to reduce irregularities in rate and rhythm in whole animal models. Our finding of similar decreases in the SR Ca²⁺-content in human cellular model of CPVT, suggests that such compensatory decreases are likely to result from the molecular cross-talk between the cellular Ca²⁺-signaling proteins.

The Ca²⁺-content of the SR was measured either by using Fluo-4 Ca²⁺-sensing dye or by the integral of NCX current generated in response to rise of intracellular Ca²⁺ by caffeine. The total charge carried by I_NCX (Fig. 5F) provides a more accurate estimate of Ca²⁺ extruded from the cytosol following application of caffeine, as this measurement is free of complications resulting from dye-saturation or quantification of Ca²⁺ using a non-ratio-metric fluorescent dyes. Fig. 5F and G confirms the findings of Fig. 5E that the mutant Ca²⁺ stores are significantly smaller than those of control myocytes, but in addition shows that cellular conditions that increase the cytosolic load of Ca²⁺, for instance, higher intracel-lular Na⁺, can strongly enhance the Ca²⁺ content of the stores in mutant cells. A similar loading of Ca²⁺ store seems also to occur on exposure of cells to isoproterenol (Fig. 6E). In the latter case, even though the caffeine-triggered Ca²⁺-release (Fluo-4 signal) is strongly enhanced by isoproterenol, the accompanying I_NCX is suppressed significantly as maybe predicted from the PKA-induced enhancement of SERCA2a/PLB activity and the increased re-uptake of Ca²⁺ into the SR, i.e. the enhancement of recirculating fraction (R_f) of Ca²⁺ [33].

4.2. Gain of CICR

In a purely CICR-gated signaling system, only the numbers of coulombs of Ca²⁺ that enter the cell and interact with RyRs determine the magnitude of released Ca²⁺. Such a strictly Ca²⁺-, but not voltage-dependent property of CICR has been observed in neurons [39]. In cardiac myocytes, however, CICR gain shows significant voltage-dependence, such that the gain increases exponentially at voltages negative to 0 mV [41]. This voltage-dependence remains controversial [12] as it may depend not only on the unitary currents and open probability of the Ca²⁺ channels [42], but also on their mean open time and the structure of the dyadic junctions which in turn may be altered under pathological conditions with respect to geometry and complements of Ca²⁺ channels and RyRs. Interestingly, in atrial myocytes only the surface, but not the centrally located RyRs appear to show the cardiac phenotypic voltage-dependent gain. Woo et al. [43] by introducing various fragments of carboxyl-tail of Ca²⁺-channel into atrial myocytes probed for possible interactions between specific domains of carboxyl tail and RyR2 in mediating the voltage-dependent gain of CICR. In support of direct interaction of DHPR with RyRs it was found that only the CaMKII/CaM binding fragment (LA) of the carboxylic tail of the DHPR [44] rendered voltage-dependence to the centrally located “naked” RyRs when the fragment was introduced into the atrial myocyte [43]. Irrespective of the underlying mechanism for the voltage-dependence of CICR, it is clear that hiPSC-CMs show not only voltage-dependent gain of CICR, but also comparatively enhanced gain at negative potentials in mutant cells (Fig. 5G), making the CPVT-mutant myocyte Ca²⁺-signaling mechanisms poised for release.

4.3. CPVT Ca²⁺-signaling models: of mice and men

The CPVT cellular model presented here was made possible by the finding that the p.F2483I mutation could be expressed in the iPSC-derived cardiomyocytes from the fibroblasts of a patient afflicted with CPVT [9], making it possible to study the pathophysiology of human disease in the laboratory setting. Considering the large number (>80) of point mutations in RyRs that bring about CPVT, it is difficult to generalize that this point mutation would have the same Ca²⁺-signaling profile as those reported for other point mutations either in transgenic mice (R2474S [45]; S2246L [28]; P2328 in [26]; and R4496C [27,46]) or in iPS-CM (M4109R [47]; S406L [48]; and P2328S [40] 2012). It is more likely that there will be subtle differences in Ca²⁺-signaling profiles of the various models, because the proposed mechanisms responsible for enhancement of diastolic Ca²⁺ release in CPVT1 vary from: (1) altered sensitivity of mutant RyR2, (2) activation by luminal Ca²⁺, (3) decreased FKBP12.6 binding to mutant RYR2 and (4) to abnormal local intermolecular RYR2 domain interactions. Our findings, summarized in Table 1, suggest that p.F2483I mutation produces unitary Ca²⁺ release events that are longer in duration, more wandering and recurrent in nature, and therefore are likely to lead to activation of Ca²⁺-waves and generation of EADs and DADs. Although in some mutant cells we observed higher frequencies of spark occurrence, we believe that such measurements in intact cells are less reliable as the frequency of sparks, even in control cells, may vary greatly depending on physiological state of the myocyte. Cells that showed higher spark frequencies are likely to be more Ca²⁺-overloaded. Interestingly, we found consistently that exposure of myocytes to isoproterenol increased the proclivity of sparks to migrate/wander and generate diastolic rises of Ca²⁺ as the frequency of spontaneous beating increased (Figs. 6 and 9). In 15 mM Na⁺ dialyzed mutant cells, held at −40 mVs, we often found higher frequencies of spontaneous releases of Ca²⁺ (accompanied by activation of equivalent I_NCX) than in control cells, or mutant cells dialyzed with 5 mM Na⁺, consistent with the idea that Ca²⁺-overloading conditions increase the proclivity to aberrant Ca²⁺ releases in mutant Cells.

The consistent decrease in the Ca²⁺-content of the SR (Fig. 5) in the mutant myocytes observed in our study and in recently reported CPVT-iPS-CM carrying P2328S mutation in RYR2 [40] might be expected considering the generally agreed upon finding of increased mean open time of mutant RyRs, irrespective of underlying mechanisms. Nevertheless, in mice knock-in CPVT model with S2246L mutation in RYR2 [28] Ca²⁺-content of the SR appears not to have changed significantly even though spark frequency was markedly increased. On the other hand, the Gomez group [27] found that R4496C mutation decreased the Ca²⁺ content of the SR, while increasing the spark frequency, consistent with our findings of smaller Ca²⁺-content of the SR. It is of course possible that the transgenic mice models with heart rates exceeding 500 beats/min do not adequately represent the pathology of human CPVT where the heart is working at ~60 beats/min. We posit that the lower Ca²⁺-content of the SR in human CPVT may serve as a compensatory protective mechanism to dampen the higher gain of CICR, and the longer and more recurrent spark activity of the CPVT myocytes. The findings that both higher cellular Na⁺ load and exposure to isoproterenol strongly increase the SR Ca²⁺ load in our model leads us to suggest that such conditions would overcome the protective compensatory decreases in the SR Ca²⁺content, that leads to aberrant Ca²⁺-release and arrhythmias.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ceca.2013.04.004.