INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

EXCITATION-CONTRACTION (E-C) coupling in cardiac cells occurs by the release of stored Ca²⁺ from the sarcoplasmic reticulum (SR) through ryanodine receptor channels (RyRs). RyRs are themselves activated by Ca²⁺ influx via voltage-activated L-type Ca²⁺ channels, the mechanism known as Ca²⁺-induced Ca²⁺ release (CICR). Ca²⁺ release is controlled by the L-type Ca²⁺ current, which is activated by membrane depolarization. Several years ago, localized discrete Ca²⁺ release events, called Ca²⁺ sparks, were discovered in cardiac cells with the use of the fluorescent Ca²⁺ probe fluo 3 and confocal laser scanning microscopy (4-6, 30). A spark event occurs when Ca²⁺ stored in the SR is released by one or more RyRs. The whole cell intracellular Ca²⁺ concentration ([Ca²⁺]_i) transient elicited by a strong depolarization is thought to represent the recruitment and summation of many Ca²⁺ sparks after an increase in opened L-type Ca²⁺ channels (8, 21).

Most studies performed on Ca²⁺ sparks have used adult cardiac cells that had been derived from enzymatically dissociated cardiac tissue. However, there is only limited knowledge on the developmental aspects of ion channel expression and Ca²⁺ release stores, which are essential for the functioning of E-C coupling. In the present study, cardiomyocytes differentiated from embryonic stem (ES) cells were used. These cardiomyocytes have been applied to characterize the time course of expression of voltage-dependent Ca²⁺ channels as well as the underlying signaling pathways and have been proven as a versatile tool to study the developmental aspects of cardiomyogenesis (17, 19, 22, 23).

We report for the first time on the occurrence of Ca²⁺ sparks during the process of differentiation of cardiomyocyes from nonbeating cardiac precursor cells to terminally differentiated cardiomyocytes. We also demonstrate that the characteristics of Ca²⁺ sparks in developing cardiomyocytes is related to the filling state of the SR with Ca²⁺ as well as to the volume of the SR and the amount of RyRs expressed during different stages of cardiomyocyte differentiation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Spinner flask culture technique for cultivation of embryoid bodies. The ES cell line CCE (27) was grown on mitotically inactivated feeder layers of primary murine embryonic fibroblasts for a maximum of eight passages in Iscove's medium (GIBCO Life Technologies) supplemented with 20% heat-inactivated (56°C, 30 min) fetal calf serum (GIBCO), 2 mM Glutamax (GIBCO), 100 µM 2-mercaptoethanol (Sigma; Deisenhofen, Germany), 1% minimal essential medium nonessential amino acid stock solution (GIBCO), 100 IU/ml penicillin, and 100 µg/ml streptomycin (GIBCO) in a humidified environment containing 5% CO₂ at 37°C and passaged every 2 to 3 days. At day 0 of differentiation, adherent cells were enzymatically dissociated with the use of 0.2% trypsin and 0.05% EDTA in phosphate-buffered saline (PBS) (GIBCO) and seeded at a density of 1 · 10⁷ cells/ml in 250 ml of siliconized spinner flasks (Integra Biosciences; Gernwald, Germany) containing 125 ml of Iscove's medium supplemented with the same additives as described above. After 24 h, 125 ml of medium were added for a final volume of 250 ml. The spinner flask medium was stirred at 20 rpm with the use of a stirrer system (Cell Spin, Integra Biosciences) and was partly (125 ml) changed every day.

Isolation procedure of cardiomyocytes. Whole 5- to 7-day-old embryoid bodies were enzymatically dissociated to obtain very early developmental stage (VEDS) cardiac precursor cells. To obtain early developmental stage (EDS) and late developmental stage (LDS) cardiomyocytes, embryoid bodies were removed from the spinner flasks on day 6 of cell culture and plated to 10-cm tissue culture petri dishes (Falcon, Becton-Dickinson; Franklin Lakes, NJ). During the subsequent days, an increasing number embryoid bodies displayed spontaneous contractions, indicating cardiomyocyte differentiation. Single cardiomyocytes were isolated from clusters of spontaneously beating areas within embryoid bodies by a modified procedure of Isenberg and Klöckner (14). Beating areas were excised with a sterile microscalpel and collected in low-Ca²⁺ solution containing (in mM) 120 NaCl, 5.4 KCl, 5 MgSO₄, 5 sodium pyruvate, 20 glucose, 20 taurine, and 10 HEPES (pH 6.9 with NaOH). The tissue was then incubated in enzyme medium (30 µM CaCl₂; 1 mg/ml collagenase B; Boehringer-Mannheim; Mannheim, Germany) for 20 min at 37°C. Tissue fragments were transferred into a medium containing (in mM) 85 KCl, 30 K₂HPO₄, 5 MgSO₄, 1 EGTA, 2 Na₂ATP, 5 Na pyruvate, 5 creatine, 20 taurine, and 20 glucose, pH 7.2, where they were kept at room temperature for 1 h and then resuspended in Iscove's medium. Isolated cells were plated on sterile coverslips (Cellocate, Eppendorf; Hamburg, Germany) and kept in the incubator for 24-48 h. Spontaneously contracting myocytes were observed within 12 h after cell preparation. Because VEDS cardiac precursor cells could not be identified by spontaneous contractions, the cells were immunostained after the Ca²⁺ measurements with an antibody directed against sarcomeric -actinin.

Ca²+ imaging and confocal laser scanning microscopy. Isolated cardiomyocytes on coverslips were loaded for 15 min at 37°C with 10 µM fluo 3-acetoxymethyl ester dissolved in dimethyl sulfoxide (final concentration 0.1%) and pluronic F-127 (Molecular Probes) (final concentration <0.025%). After the cells were loaded, they were washed in E1 buffer containing (in mM) 135 NaCl, 5.4 KCl, 1.8 CaCl₂, 1 MgCl₂, 10 glucose, and 10 HEPES (pH 7.4 at 37°C). Fluo 3 fluorescence imaging was performed on a laser scanning confocal microscope (model LSM 410, Zeiss; Jena, Germany) equipped with an argon ion laser and coupled to an inverted microscope (Axiovert 135, Zeiss). The objective lens was a Zeiss ×25 oil immersion Plan-Neofluar with a numerical aperture of 0.8. Fluo 3 fluorescence was excited with the use of the 488-nm line of the argon laser. The laser excitation beam was directed to the specimen through a 510-nm dichroic beam splitter, and the emitted fluorescence was collected through a 515-nm long-pass emission filter in front of a photomultiplier tube. For Ca²⁺ spark recording and quantitative analysis, the line-scan mode of the confocal laser-scanning microscope was used. Single myocytes were scanned repetitively (250 Hz) along a line positioned along the longitudinal axis of the cell avoiding the nuclei. The axial cell depth amounted to ~2 µm. A line-scan image was constructed by stacking all 512 lines vertically. The magnification was set by the objective and the hardware zoom factor of the laser scanning microscope to give a pixel size of ~0.1 µm². The volume of each pixel (voxel) in the line was ~0.1 µm³.

Fluorescence staining of RyRs. RyRs in cardiac cells were stained by use of the fluorescent ryanodine derivative BODIPY FL-X (Molecular Probes; Eugene, OR). Briefly, cardiac cells isolated from embryoid bodies were fixed for 30 min in methanol-acetone (7:3), permeabilized in PBS supplemented with 0.1% Triton X-100 (Sigma), and incubated for 20 min with 0.2 µM BODIPY FL-X ryanodine dissolved in dimethyl sulfoxide. The cells were subsequently washed three times in PBS, and the BODIPY fluorescence was analyzed by confocal laser scanning microscopy and the 488-nm line of the argon-ion laser. Emission was recorded using a long-pass 515-nm filter set.

Immunohistochemistry. Cardiac cells were identified by immunolabeling sarcomeric -actinin. Cardiac cells on coverslips were fixed in ice-cold methanol-acetone (7:3) for 30 min and were subsequently permeabilized with PBS supplemented with 0.1% Triton X-100. Blocking against unspecific staining was performed by incubation for 1 h in PBS containing 1% milk powder. For -actinin staining, a mouse monoclonal antibody (Sigma) was used in a concentration of 20 µg/ml. The secondary antibody was a Cy3-labeled goat anti-mouse antibody (Dianova; Hamburg, Germany) in a concentration of 6.5 µg/ml. Excitation was performed by use of a 543-nm He-Ne laser of the confocal setup. Emission was recorded with the use of a 575- to 640-nm band-pass filter set. The RyR was labeled by using a goat polyclonal antibody directed against the amino terminus of RyR (Santa Cruz Biotechnology; Santa Cruz, CA) and applied in a concentration of 4 µg/ml. As a secondary antibody, a Cy5-labeled donkey-anti-goat IgG (Dianova) was used in a concentration of 5 µg/ml and was excited by a 633-nm He-Ne laser of the confocal setup. Emission was recorded with the use of a long-pass 655-nm filter set.

Statistical analysis. Data are given as means ± SE, with n denoting the number of experiments with either cardiac cells or embryoid bodies derived from at least three independent preparations. Student's t-test for unpaired data was applied as appropriate. A value of P < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Time course of cardiac differentiation of ES cells. ES cells of the cell line CCE were cultivated in spinner flask culture to the three-dimensional multicellular tissue of embryoid bodies. Within the embryoid bodies, cardiac precursor cells (5- to 7-day-old), EDS (8- to 11-day-old), and LDS cardiomyocytes (15- to 25-day-old) were differentiated. Spontaneous contractions were not visible in VEDS embryoid bodies but started 1 day after plating (day 7). The number of contracting embryoid bodies increased during subsequent days and reached a maximum of 83 ± 4% on day 5 after plating (n = 7) (Fig. 1A). In parallel to the increase in spontaneously contracting embryoid bodies, a marked increase in the frequency of contractions from 60 contractions per minute on day 3 after plating to 80 ± 2 contractions per minute on days 9-10 after plating was observed during the time course of cardiomyocyte differentiation (n = 7) (Fig. 1B).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Time course of cardiomyogenesis in embryoid bodies cultivated with the spinner flask technique. A: percentage of spontaneously beating embryoid bodies after plating. Embryoid bodies were cultivated for 6 days in spinner flasks and subsequently plated to cover slips. Number of beating embryoid bodies was evaluated by transmission light microscopy. B: beating frequency in relation to the age of the embryoid bodies.

View larger version (80K): [in this window] [in a new window]   Fig. 2.   Structural organization of clusters of developing cardiomyocytes in plated whole mount embryoid bodies. Embryoid bodies were immunostained with a monoclonal antibody directed against sarcomeric -actinin. Shown are representative 6-day-old (A), 9-day-old (B), and 19-day-old (C) embryoid bodies. No sarcomeric organization was visible in very early developmental stages (VEDS) cardiac precursor cells. Bar in A represents 45 µm. Bars in B and C represent 15 µm.

[Ca²+]_i transients in cardiomyocytes differentiated from ES cells. It has been shown (32) that EDS cardiomyocytes derived from the D3 ES cell line (days 8-9) were characterized by small fluctuations of [Ca²⁺]_i from intracellular stores, which were independent from the membrane potential as well as the activity of voltage-dependent Ca²⁺ channels. In the present study, [Ca²⁺]_i oscillations of cardiomyocytes were recorded at different stages of cardiomyocyte differentiation, i.e., in 5- to 7-day-old VEDS cardiac precursor cells, in which spontaneous contractions were absent, as well as in spontaneously contracting EDS (days 8-11) and LDS (days 15-25) cardiomyocytes. As shown in Fig. 3, A-D, slow [Ca²⁺]_i fluctuations with largely variable frequency and sometimes irregular shape occurred in cardiac precursor cells (n = 10). Incubation with 10 mM of caffeine resulted in a transient rise in [Ca²⁺]_i, indicating the presence of functional RyRs. Furthermore, superfusion of VEDS cardiac precursor cells with 140 mM K⁺ resulted in an increase of [Ca²⁺]_i, indicating the presence of voltage-activated Ca²⁺ channels (see Fig. 3E). Interestingly, no propagation of the Ca²⁺ responses between different cardiac precursor cells organized in cardiac cell clusters was observed, indicating an absence of intercellular coupling. In EDS and LDS cardiomyocytes, the frequency of [Ca²⁺]_i oscillations increased to 40-70 per minute (see Fig. 3F), which correlated to the frequency of contractions. In cell clusters of these cardiac cells propagating [Ca²⁺]_i waves were observed, indicating intercellular coupling presumably via gap junctions (data not shown).

View larger version (43K): [in this window] [in a new window]   Fig. 3.   Spontaneous intracellular calcium concentration ([Ca2+]i) transients in cardiomyocytes differentiated from embryonic stem (ES) cells. A-D: [Ca2+]i transients of VEDS cardiac cells. Tracings were recorded from different cells within the same cell cluster. Ca2+ responses displayed large differences in amplitude and frequency, indicating the absence of intercellular coupling. E: presence of caffeine-sensitive Ca2+ stores and voltage-dependent Ca2+ influx pathways in VEDS cardiomyocytes. VEDS cardiomyocytes were superfused with solution containing 10 mM of caffeine, which led to transient increase in [Ca2+]i. Cells were then superfused with high (140 mM) K+ solution, which resulted in an elevation of [Ca2+]i, indicating the presence of voltage-dependent Ca2+ influx pathways. Tracing was recorded from a cell cluster different from the cluster displayed in A-D. F: [Ca2+]i transients in a 10-day-old early developmental stage (EDS) cardiac cell, which displayed spontaneous contractions at a frequency of 66 beats/min.

Ca²+ sparks during the time course of cardiomyocyte differentiation. Investigations on the occurrence of Ca²⁺ sparks in embryonic cardiomyocytes have not yet been undertaken. In adult cardiomyocytes, E-C coupling is mediated by the release of Ca²⁺ from the SR via RyRs. It is generally accepted that local [Ca²⁺]_i transients, i.e., sparks are triggered by depolarization-induced Ca²⁺ entry through L-type Ca²⁺ channels. The RyRs are thought to be situated very close to the L-type Ca²⁺ channel, where they sense an increase in local [Ca²⁺]_i when a nearby Ca²⁺ channel opens (33).

1

View larger version (157K): [in this window] [in a new window]   Fig. 4.   Representative false color line-scan images of nonstimulated VEDS cardiac precursor cells (A), EDS (B), and late developmental state (LDS; C) cardiomyocytes differentiated from ES cells. Line-scan images were recorded after spontaneous global [Ca2+]i transients had declined to baseline values. Position along the cell is presented horizontally, and successive scan lines in time are displayed vertically. Each image includes 512 lines scanned at 250 Hz. Time is running from top to bottom. Localized Ca2+ sparks were absent in noncontracting VEDS cardiac precursor cells, whereas typical, spatially restricted sparks were observed in contracting EDS and LDS cardiomyocytes. Arrowheads indicate from where in the images the tracings of Fig. 5 were recorded. Vertical bar, 0.2 s. Horizontal bar, 5 µm.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Representative time-course plots of [Ca2+]i changes in a series of line-scan images obtained from VEDS cardiac precursor cells (A), EDS (B), and LDS (C) cardiomyocytes. Line-scan images were recorded after spontaneous global [Ca2+]i transients had declined to baseline levels. Arrowheads indicate the section of the respective trace that is represented by the confocal images in Fig. 4. Traces represent a spatial width of 2 µm.

View larger version (22K): [in this window] [in a new window]   Fig. 6.   Distribution of spark frequencies (A), spark amplitudes (F/F0) (B), and the spatial diameter of sparks (full width at half maximum) (C) in EDS (solid bars) and LDS (open bars) cardiomyocytes.

Expression of RyRs during cardiomyocyte differentiation. The observed changes in the characteristics of Ca²⁺ sparks in EDS and LDS cardiomyocytes may be owing to differences in the expression of L-type Ca²⁺ channels and RyRs, as well as the Ca²⁺ load of the SR. A threefold increase in the voltage-dependent Ca²⁺ current density during the time course of cardiomyocyte differentiation of ES cells has been reported (19). To evaluate the expression of RyRs, beating areas of cardiomyocytes in whole mount embryoid bodies were labeled with BODIPY FL-X ryanodine, which has been recently used to localize RyRs in the rat parotid gland (35) and in pancreatic -cells (11). Furthermore, RyR expression was evaluated in single cell preparations by a polyclonal anti-RyR antibody and immunohistochemical methods (Fig. 7). BODIPY FL-X ryanodine staining revealed that RyRs were expressed in EDS as well as in LDS cardiomyocytes. However, the level of expression of RyRs as evaluated by quantification of BODIPY FL-X ryanodine fluorescence was significantly increased by 53 ± 3% in LDS cardiomyocytes compared with EDS cardiomyocytes (n = 3). Immunohistochemistry of the RyR in single VEDS, EDS, and LDS cardiomyocytes revealed a continuous increase in RyR immunofluorescence intensity in LDS cardiomyocytes compared with VEDS and EDS cardiomyocytes, indicating an increasing density of RyRs during cardiomyocyte differentiation (n = 3). Furthermore, a significant increase in the volume of the SR positive for RyRs was observed during cardiac cell maturation (see Fig. 7). In VEDS and EDS cardiomyocytes, the most pronounced staining was observed in the near-nuclear region. In LDS cardiomyocytes, the cell compartment displaying SR positive for RyR-immunofluorescence was significantly increased compared with VEDS and EDS cardiomyocytes and occupied nearly the whole cytoplasmic compartment.

View larger version (88K): [in this window] [in a new window]   Fig. 7.   Expression of ryanodine receptor channels (RyRs) in VEDS (A), EDS (B), and LDS (C) cardiac cells differentiated from ES cells as evaluated by RyR immunofluorescence using a polyclonal antibody directed against the RyR. Top: RyR immunofluorescence; the color code bar represents RyR immunofluorescence intensity. Middle: -actinin staining by the use of a monoclonal anti--actinin antibody. Bottom: overlay images of RyR immunofluorescence (yellow-red) and -actinin immunofluorescence (green). RyR immunofluorescence, as well as the extension of the SR positive for RyRs, is increasing during the course of cardiomyocyte differentiation. Bar represents 10 µm.

Ca²+ load of the SR in EDS and LDS cardiomyocytes. The observed differences in the frequency and amplitude of Ca²⁺ sparks in EDS and LDS cardiomyocytes may be related to differences in the Ca²⁺ load of the SR and/or differences in cytoplasmic [Ca²⁺]_i. To evaluate these issues, calibration measurements for [Ca²⁺]_i were performed, and Ca²⁺ was released from intracellular stores by superfusion with 10 mM of caffeine (Fig. 8, A and B). It was observed that basal [Ca²⁺]_i was not significantly different in EDS and LDS cardiomyocytes with 92.8 ± 9.5 nM (n = 22) and 99.8 ± 4.3 nM (n = 18), respectively. Superfusion with solutions containing 10 mM of caffeine resulted in a transient increase in [Ca²⁺]_i. The amplitude of the [Ca²⁺]_i transient was significantly increased in LDS cardiomyocytes and amounted to 2,120 ± 180 nM (n = 49), whereas in EDS cardiomyocytes a peak [Ca²⁺]_i of 780 ± 30 nM (n = 63) was observed, which clearly indicates an elevated Ca²⁺ content of the SR in LDS compared with EDS cardiomyocytes. However, the reduced amplitude of the caffeine-induced Ca²⁺ response in EDS cardiomyocytes may result from the lower number and density of RyRs in EDS compared with LDS cardiomyocytes rather than a decreased Ca²⁺ load of the SR. To exclude this possibility, EDS cardiomyocytes were preincubated before caffeine application in solution containing 10 mM of Ca²⁺. This experimental protocol resulted in a significant increase in the peak [Ca²⁺]_i to 1,990 ± 315 nM, which was not significantly different from the values obtained in LDS cardiomyocytes (n = 54) (Fig. 8B, inset). Interestingly, [Ca²⁺]_i spiking was restored in ~80% of EDS cardiomyocytes after the decline of the caffeine-induced [Ca²⁺]_i transient to basal [Ca²⁺]_i, despite the continuous presence of caffeine, whereas in the majority of LDS cardiomyocytes [Ca²⁺]_i spiking was abolished.

View larger version (25K): [in this window] [in a new window]   Fig. 8.   Ca2+ load of the SR in EDS (A) and LDS (B) cardiomyocytes. Fluo 3-loaded cardiomyocytes were superfused with 10 mM caffeine (arrows), which released Ca2+ from the SR, resulting in a transient increase in [Ca2+]i. B, inset: absolute peak [Ca2+]i values after caffeine treatment of EDS and LDS cardiomyocytes under control extracellular Ca2+ concentration ([Ca2+]o) conditions as well as of EDS cardiomyocytes preincubated for 5 min in solution containing 10 mM Ca2+ before the addition of caffeine. Note that [Ca2+]i spiking continued in EDS cardiomyocytes in the continuous presence of caffeine, whereas spiking was absent in LDS cardiomyocytes. *P < 0.05, significantly different from untreated EDS cardiomyocytes.

1

1

View larger version (49K): [in this window] [in a new window]   Fig. 9.   Effects of superfusion with 10 mM Ca2+ on the the global [Ca2+]i changes (A) and spatially localized Ca2+ sparks (B and C) in EDS cardiomyocytes. A: time point of [Ca2+]i elevation is indicated by an arrow. B: representative false color line-scan image showing Ca2+ sparks under high- [Ca2+]i conditions. Image includes 512 lines scanned at 250 Hz. Arrowhead indicates from where in the line-scan image the tracing in C was recorded. Horizontal bar represents 15 µm, whereas the vertical bar represents 0.2 s. The time is running from top to bottom. C: representative time-course plots Ca2+ sparks in a series of line-scan images obtained from EDS cardiomyocytes superfused with 10 mM Ca2+. Arrowheads indicate which section of the respective trace is represented by the confocal image. Trace represents a spatial width of 2 µm.

View larger version (24K): [in this window] [in a new window]   Fig. 10.   Distribution of spark spark frequencies (A) and amplitudes (F/F0) (B) of EDS cardiomyocytes superfused with 10 mM Ca2+ solution. A and B, insets: mean spark frequencies and spark amplitudes in EDS cardiomyocytes under control (1.8 mM) Ca2+ and under high (10 mM) Ca2+, respectively. *P < 0.05, significantly different from untreated EDS cardiomyocytes.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study reports for the first time on events of elementary Ca²⁺ release in cardiomyocytes differentiated from pluripotent ES cells. This in vitro system allows the investigation of early steps of cardiomyogenesis. Our data demonstrate that Ca²⁺ sparks displaying comparable characteristics with sparks investigated in cardiac cells of different mammalian species (4, 20, 21, 28, 34) occurred in EDS and LDS cardiomyocytes differentiated from murine ES cells. During the time course of cardiomyogenesis, a prominent increase in the number of cells displaying typical spatially restricted Ca²⁺ sparks was observed.

Absence of Ca²+ sparks in VEDS cardiac precursor cells. Sparks were totally absent in VEDS cardiac precursor cells despite the obvious presence of voltage-dependent Ca²⁺ channels and the presence of caffeine-sensitive Ca²⁺ stores. These cardiac precursor cells did not show any signs of spontaneous contractions under light microscopy inspection. In clusters of cardiac precursor cells, global [Ca²⁺]_i fluctuations with large variations in amplitude and frequency occurred which may indicate that cells of different developmental stages are present in close association. The [Ca²⁺]_i fluctuations in cardiac precursor cells may be driven by the D-myo-inositol 1,4,5-trisphosphate [Ins(1,4,5)P₃] pathway and/or mitchondria, because they have been previously shown to persist in high-K⁺ solution and in the presence of blockers of voltage-dependent Ca²⁺ channels (32). However, we never observed Ca²⁺ puffs and blibs, which have been previously associated to Ca²⁺ release via Ins(1,4,5)P₃-sensitive Ca²⁺ stores (3). The presence of Ins(1,4,5)P₃-sensitive Ca²⁺ stores in ES cells has been recently demonstrated (19) and is likewise evident in adult cardiomyocytes, where Ins(1,4,5)P₃ plays a role in the regulation of cardiac autonomic [Ca²⁺]_i spiking (15).

Occurrence of Ca²+ sparks at low frequency and amplitude in EDS cardiomyocytes. Typical Ca²⁺ sparks that occurred at low frequency were obvious in EDS cardiomyocytes, which contracted visibly as evaluated by transmission light microscopy. In beating areas of whole mount EDS embryoid bodies displaying spontaneous contractions, immunohistochemical staining of -actinin showed a distinct sarcomeric organization, although a random organization of the myofibrils prevailed. With prolonged culture time, the frequency of the contractions as well as the frequency and the amplitude of the Ca²⁺ sparks increased, whereas no significant difference in the spatial diameter and the half-time of spark rise and decay was observed. The changes in the characteristics of Ca²⁺ sparks may be owing to several causes. These include the maturation of the contractile apparatus and the SR during cardiac cell differentiation, as well as the number of voltage-dependent Ca²⁺ channels and the density of RyRs. Comparable results have been recently achieved with neonatal rabbit ventricular cells, which displayed a significantly reduced frequency of Ca²⁺ sparks compared with adult ventricular cells despite the presence of caffeine releasable Ca²⁺ stores (9). In the latter study, it was discussed that a paucity or immaturity of T-tubular diadic junctions between L-type Ca²⁺ channels and SR Ca²⁺ release channels in immature cells could functionally isolate sarcolemmal Ca²⁺ entry from triggering SR Ca²⁺ release (9). Ca²⁺ sparks have been demonstrated (25, 30) to occur at sites associated with T tubules, which are spaced at regular intervals along the length of the cells. It has been shown that most neonatal mammalian cardiomyocytes do not develop T tubules until 8-10 days of age, which may at least partially account for the absence of localized Ca²⁺ sparks in VEDS cardiac precursor cells observed in the present study. Additionally, the spatial association of dihydropyridine receptors and RyRs may be a key factor for CICR and has been recently shown to augment during the differentiation of rabbit neonatal heart cells (29). However, it should be mentioned that in adult atrial myocytes, which lack T tubules, Ca²⁺ sparks occur in cellular regions that are occupied by nonjunctional SR, i.e., from SR, which contains RyRs but is not closely associated with the L-type Ca²⁺ channels of the plasma membrane (2, 13). This points towards the notion that the absence of detectable Ca²⁺ sparks in VEDS cardiomyocytes may be the result of the lower densities of RyRs and the possibility that RyRs are not clustered in the same fashion as in LDS cardiomyocytes and/or that spontaneous Ca²⁺ release events may indeed occur but escape confocal imaging because of their small magnitude.

Expression of RyRs in ES cell-derived cardiomyocytes. The data of the present study demonstrate an increase in BODIPY FL-X ryanodine staining in beating areas of whole mount LDS cardiomyocytes, suggesting an increased expression of RyRs during the time course of cardiac cell differentiation. Furthermore, RyR immunohistochemistry revealed a pronounced increase in the volume of the SR positively labeled for RyRs during cardiomyogenesis of ES cells. This rise in the number of RyRs parallels the increase in the L-type Ca²⁺ current density, which has been previously reported (23). In contrast to the increasing expression of RyRs, the expression of Ins(1,4,5)P₃ receptors was significantly downregulated (data not shown), which may indicate a switch from the prevalence of Ins(1,4,5)P₃-mediated Ca²⁺ signaling pathways in early stages to CICR pathways in late stages of cardiomyogenesis.

Evidence for a lower Ca²+ load in the SR of EDS compared with LDS cardiomyocytes. The Ca²⁺ load of the SR is an additional feature that may influence the characteristics of Ca²⁺ sparks. In the present study, the Ca²⁺ load was evaluated by investigation of the caffeine-releasable Ca²⁺ pool. Interestingly, it was observed that Ca²⁺ spiking returned after Ca²⁺ release after treatment of EDS cardiomyocytes with caffeine, whereas spiking was abolished in LDS cardiomyocytes. This may either indicate that additional Ca²⁺ stores, which are not caffeine sensitive, are present in EDS cardiomyocytes, or further Ca²⁺ transport pathways, e.g., the Na⁺/Ca²⁺ exchange (9), are involved in Ca²⁺ spiking. The data of the present study clearly demonstrate by Ca²⁺ release experiments using the SR store agonist caffeine that the filling state of the SR in LDS cardiomyocytes is significantly increased compared with EDS cardiomyocytes. An influence of the SR Ca²⁺ load on the frequency as well as the amplitude of Ca²⁺ sparks has been reported (6, 28, 31) and has been interpreted as an increased open probability of SR Ca²⁺ release channels owing to the elevated SR Ca²⁺ content and/or an increased sensitivity of the channels to activation by cytosolic Ca²⁺ (28). Furthermore, it has been suggested that increased SR Ca²⁺ load augments the effectiveness of a given Ca²⁺ current to activate SR Ca²⁺ release (1, 10, 16). The data of the present study demonstrate a lower expression level of RyRs and a smaller extension of the SR in EDS compared with LDS cardiomyocytes, which correlates with a reduced L-type Ca²⁺ current density in EDS cardiomyocytes that has been previously described (23). Hence, our data corroborate observations in fetal rat ventricular cardiomyocytes, which displayed fewer RyRs compared with adult cardiac cells (26). The elevation of the SR Ca²⁺ load by superfusion with 10 mM external Ca²⁺ increased spark frequency and amplitude to the level of LDS cardiomyocytes. This may suggest that the density of RyRs and voltage-dependent Ca²⁺ channels or possible differences in the microachitecture between the RyR and the L-type Ca²⁺ channel in EDS compared with LDS cardiomyocytes play a secondary role for the observed differences in the spark characteristics. However, elevation of [Ca²⁺]_o results in a larger unitary current through L-type Ca²⁺ channels due to the higher driving force for Ca²⁺, which may cause a larger rise in [Ca²⁺]_i in the diadic cleft. Elevated [Ca²⁺]_i in the diadic cleft could compensate for potential differences in the microarchitecture of the RyR-dihydropyridine receptor arrangement and thereby affect properties of the Ca²⁺ sparks. Furthermore, the elevated cytosolic [Ca²⁺]_i that occurs after incubation of EDS cardiomyocytes in high-[Ca²⁺]_o solution may directly trigger the opening of RyR through CICR. This then results (see Ref. 24) in the occurrence of sites that generate sparks at high frequency. Frequent spark sites were observed in the present study when EDS cardiomyocytes were incubated in 10 mM external Ca²⁺.

Mechanisms of E-C coupling. According to the local control theory of E-C coupling in adult cardiac cells, CICR is locally initiated by Ca²⁺ influx through a single L-type Ca²⁺ channel that activates a group of SR Ca²⁺ release channels at T tubule SR junctions, which, during the action potential, equal a global [Ca²⁺]_i transient. In newborn rabbit ventricular myocytes, the absence of the T tubule network has been evidenced, which led to the assumption that the immature SR may play a minor role in E-C coupling. On the basis of the observations in newborn cardiomyocytes that the myofibrils are located subsarcolemmally, the cytosolic Ca²⁺ buffering is lower, and Na⁺/Ca²⁺ expression is increased, it was suggested that the Na⁺/Ca²⁺ exchange pathway is the predominant pathway for contraction and relaxation in newborn rabbit ventricular myocytes. This assumption can be verified by using a mathematical model (9). Compelling evidence has been recently provided that mechanisms different from the classical model of E-C coupling occur likewise in early cardiac cells differentiated from ES cells (32). In these EDS cardiac cells, spontaneous [Ca²⁺]_i oscillations drive the cell contractions, which led to the assumption that in these cells the voltage-dependent Ca²⁺ current may primarily serve for store refilling. The absence of Ca²⁺ sparks in noncontracting VEDS cardiac precursor cells and the age-dependent increase in spontaneously contracting EDS cells displaying typical sparks support these previous investigations and envision [Ca²⁺]_i release mechanisms distinct from the known CICR mechanism as primary rhythm generator in VEDS and EDS cardiac cells. This primary rhythm generator is then gradually replaced during the later stages of cardiac cell differentiation by the CICR mechanism, which is characterized by distinct spatially localized Ca²⁺ spark events.