Developmental aspects of cardiac Ca²⁺ signaling: interplay between RyR- and IP₃R-gated Ca²⁺ stores

Abstract

The dominant mode of intracellular Ca²⁺ release in adult mammalian heart is gated by ryanodine receptors (RyRs), but it is less clear whether inositol 1,4,5-trisphosphate (IP₃)-gated Ca²⁺ release channels (IP₃Rs), which are important during embryogenesis, play a significant role during early postnatal development. To address this question, we measured confocal two-dimensional Ca²⁺ dependent fluorescence images in acutely isolated neonatal (days 1 to 2) and juvenile (days 8–10) rat cardiomyocytes, either voltage-clamped or permeabilized, where rapid exchange of solution could be used to selectively activate the two types of Ca²⁺ release channel. Targeting RyRs with caffeine produced large and rapid Ca²⁺ signals throughout the cells. Application of ATP and endothelin-1 to voltage-clamped, or IP₃ to permeabilized, cells produced smaller and slower Ca²⁺ signals that were most prominent in subsarcolemmal regions and were suppressed by either the IP₃R-blocker 2-aminoethoxydiphenylborate or replacement of the biologically active form of IP₃ with its L-stereoisomer. Such IP₃R-gated Ca²⁺ releases were amplified by Ca²⁺-induced Ca²⁺ release (CICR) via RyRs since they were also reduced by compounds that block the RyRs (tetracaine) or deplete the Ca²⁺ pools they gate (caffeine, ryanodine). Spatial analysis revealed both subsarcolemmal and perinuclear origins for the IP₃-mediated Ca²⁺ release events RyR- and IP₃R-gated Ca²⁺ signals had larger magnitudes in juvenile than in neonatal cardiomyocytes. Ca²⁺ signaling was generally quite similar in atrial and ventricular cardiomyocytes but showed divergent development of IP₃-mediated regulation in juveniles. Our data suggest that an intermediate stage of Ca²⁺ signaling may be present in developing cardiomyocytes, where, in addition to RyR-gated Ca²⁺ pools, IP₃-gated Ca²⁺ release is sufficiently large in magnitude and duration to trigger or contribute to activation of CICR and cardiac contraction.

cardiac excitation-contraction coupling in adult mammalian cardiomyocytes involves Ca²⁺ entry via L-type Ca²⁺ channels, Ca²⁺-induced Ca²⁺ release (CICR) from ryanodine receptors (RyR) on sarcoplasmic reticulum (SR), and Ca²⁺-mediated cellular contraction. In ventricular cells, Ca²⁺ signaling responses travel rapidly throughout the cell, as the action potential invades the network of t-tubules (3). However, it is not known in detail how the adult mode of Ca²⁺ signaling is established during embryonic and postnatal development, where the mammalian heart undergoes rapid changes in morphology and gene expression, including major modifications of the organelles and proteins that are associated with Ca²⁺ signaling. For instance, embryonic and neonatal cardiomyocytes lack the t-tubular network, whereas juvenile cardiomyocytes have an immature t-tubular system (15, 45). Similarly, SR in the fetal heart is sparse and has decreased capability to load Ca²⁺ compared with adult SR (33). Inositol 1,4,5-trisphosphate (IP₃)- and RyRs are also differentially expressed with development and appear at the same time period when Ca²⁺ oscillations are first observed (12, 42). Interestingly, in embryonic cardiomyocytes, Ca²⁺ oscillations precede the expression of the ionic channels responsible for spontaneous depolarization (35, 42), and transgenic mice deficient in pacemaking channels demonstrate early embryonic spontaneous beating (41, 44, 50). Consistent with this idea, ryanodine (Ry) has been shown to diminish, but not eliminate, embryonic and neonatal cell contraction (36, 45, 51). Although functional RyRs are clearly present in postnatal cardiomyocytes (18), observations that inhibitors of SR function exert a more pronounced suppressive effect on cardiac action potential-mediated Ca²⁺ transients in juvenile (88%) vs. neonatal (15%) cardiomyocytes indicate an increased reliance on and upregulation of RyR as a function of age (10, 11). Antisense knockdown of IP₃R₁ in embryonic stem cells markedly reduces cardiac cellular oscillations (35). These data suggest that multiple mechanisms may be involved in cardiac pacing, including both IP₃R- and RyR-gated Ca²⁺ signaling.

There is a clear contrast between cardiac Ca²⁺ signaling of embryonic and adult cardiomyocytes, but it remains unclear how the embryonic Ca²⁺ signaling phenotype transitions into the adult form. Here we studied the characteristics and roles of RyR- and IP₃R-gated Ca²⁺ signaling in developing cardiomyocytes. Since culturing cardiomyocytes changes the physiology and gene expression of the cells, we used only acutely isolated developing cardiomyocytes. To focus on the intracellular Ca²⁺ release mechanisms, we eliminated the effects of membrane depolarization either by voltage-clamping the cells or permeabilizing the cell membrane. In the first case, we dialyzed the cells with high concentrations of fast fluorescent and slow nonfluorescent Ca²⁺ chelators [1 mM Fluo-4 + 2 mM EGTA; intracellular Ca²⁺ concentration ([Ca²⁺]_i) ≅ 100 nM] to accentuate the rate of Ca²⁺ release while suppressing large local rises in [Ca²⁺]_i (1, 49) that may support propagated waves of CICR. To allow reliable Ca²⁺ measurements in permeabilized cells, they were superfused with a simulated internal solution, which was also designed to accentuate the Ca²⁺ release, but contained lower concentrations of the Ca²⁺ buffers (0.04 mM Fluo-4 + 0.4 mM EGTA; [Ca²⁺]_i ≅ 100 nM).

We tested the hypothesis that maturation of cardiomyocytes is accompanied by changes in the relative contribution of functional RyR- and IP₃R-gated Ca²⁺ stores and that developing cardiomyocytes have specific Ca²⁺ signaling mechanisms, such as IP₃-mediated Ca²⁺ release, that are capable of triggering Ca²⁺ transients, in part, by local CICR. We found that developing cardiomyocytes have functional RyR- and IP₃R-gated Ca²⁺ releases with unique pharmacology and spatial profiles. Maturation from neonatal to juvenile cardiac phenotype involved an upregulation of both types of Ca²⁺ release channel. The IP₃-mediated Ca²⁺ signaling was capable of generating both subsarcolemmal and perinuclear Ca²⁺ releases in developing cardiomyocytes, the former of which can directly contribute to CICR transients in the subsarcolemmal regions.

MATERIALS AND METHODS

Experiments were carried out in accordance with institutional and federal guidelines. Protocols were approved and supervised by Georgetown University Animal Care and Use Committee.

Cell isolation.

For preparation of cardiomyocytes from neonatal and juvenile rats, with the juveniles first being deeply anesthetized, the animals were decapitated, the chest cavities were opened, and the hearts were removed. The excised hearts were retrogradely perfused at 37°C through the aorta, first for 2 min with Ca²⁺ free Tyrode's solution composed of (in mM) 137 NaCl, 5.4 K-glutamate, 1 MgCl₂, 10 glucose, and 10 HEPES titrated to pH 7.4 with NaOH and then with Ca²⁺ free Tyrode's solution supplemented with collagenase (0.8 mg/ml) and protease (0.1 mg/ml) for 4 min. The atrial and ventricular tissues were dissected and gently agitated to dissociate single cells. The freshly dissociated cells were plated onto glass coverslips, stored at room temperature in Tyrode's solution containing 0.2 mM CaCl₂, and used within 8 h in Ca²⁺ imaging experiments where the membrane potential was either voltage-clamped or abolished by membrane permealibization. Cells that maintained their native elongated shape following the isolation procedure were selected for detailed examination over those that that assumed a more globular appearance.

Drugs were dissolved in the external experimental solutions and applied rapidly using a concentration-clamp device (6). Rapid application of 100 μM ATP (Sigma, St. Louis, MO), 10 mM caffeine (Sigma), 100 nM endothelin-1 (Sigma), and 1 or 20 μM IP₃ (D-myo-IP₃; Calbiochem, Gibbstown, NJ) was used to probe Ca²⁺ stores. All experiments were carried out at room temperature (22–24°C).

Whole-cell voltage-clamp procedures.

Some cells were voltage-clamped in the whole cell configuration (16) to 1) hold the membrane potential steady at −70 mV during recording of drug-induced Ca²⁺ transients, 2) measure membrane currents [e.g., Na⁺-Ca²⁺ exchanger current (I_NaCa) and Cl⁻ current] that were activated during these interventions, and 3) subject the cells to repeated depolarizing pulses from −90 to 0 mV to maintain and equilibrate intracellular Ca²⁺ stores. The voltage-clamped cells were dialyzed with a Ca²⁺-buffered pipette solution containing (in mM) 110 CsOH, 110 aspartic acid, 5 NaCl, 20 TeaCl, 5 Mg-ATP, 0.2 cAMP, 2 EGTA, 1 K₅Fluo-4, 1.2 CaCl₂, and 20 HEPES (titrated to pH 7.2 with CsOH) (56). With the consideration of the binding constants of Fluo-4 (K_d ≅ 400 nM) and EGTA (K_d ≅ 100 nM), the addition of 1.2 mM CaCl₂ was calculated to buffer [Ca²⁺]_i of the dialyzing solution at ∼100 nM. Membrane currents were measured with a Dagan voltage-clamp amplifier using pClamp software. The conditioning pulses were initiated 3 min after rupture of the membrane, and fluorescence measurements were started 3 to 4 min later. The extracellular solution used during experiments contained (in mM) 137 NaCl, 5.4 K-glutamate, 2 CaCl₂, 1 MgCl₂, 10 glucose, and 10 HEPES (titrated to pH 7.4 with NaOH).

Permeabilization.

After being prestained with 20 μM Fluo-4 acetoxymethyl ester (Fluo-4 AM), the cells were permeabilized with saponin (9, 63). First, the cells were suspended in a solution containing (in mM) 100 K-aspartate, 20 KCl, 0.5 EGTA, 0.75 MgCl₂, and 10 HEPES 10 (titrated to pH 7.2 with KOH). The cell surface membrane was permeabilized by adding 0.005% (wt/vol) saponin for 30 s. After 30 s, the bath solution was exchanged for a saponin-free simulated internal solution composed of 100 mM K-aspartate, 15 mM KCl, 5 mM KH₂PO₄, 5 mM MgATP, 0.4 mM EGTA, 0.04 mM K₅Fluo-4, 0.2 mM CaCl₂ ([Ca²⁺]_i ≅ 100 nM), 0.75 mM MgCl₂, 10 mM phosphocreatine, 5 U/ml creatine phosphokinase, 8% dextran, and 10 mM HEPES (titrated to pH 7.2 with KOH).

Two-dimensional confocal imaging.

Intracellular Ca²⁺ signals were measured with the fluorescent Ca²⁺-indicator dye Fluo-4, which was used in conjunction with the nonfluorescent Ca²⁺-chelator EGTA, as mentioned above. The cells were imaged using a Noran Odyssey XL rapid two-dimensional laser scanning confocal microscopy system (Noran Instruments, Madison, WI) attached to a Zeiss Axiovert TV135 inverted microscope fitted with an × 63 water-immersion objective lens. The excitation wavelength of the argon ion laser was set to 488 nm, and fluorescence emission was measured at wavelengths >515 nm. Cells were imaged confocally at 4–240 frames/s depending on the experiment.

The average resting fluorescence intensity (F₀) was calculated from several frames measured immediately before drug application. Images were filtered by 3 pixel × 3 pixel averaging. The amplitudes of the Ca²⁺-dependent cellular fluorescence signals were quantified as ΔF/F₀ = F/F₀ − 1, where F is the peak value and, like F_o, was measured by integration over the entire area of each cell. Tau₅₀ quantifies the rate of rise of the Ca²⁺ signals and was calculated as 1.25× the time required for the transient to develop from 10% to 50% at the peak value (Fig. 1). Generally, the fluorescence images used in illustrations were normalized by dividing the images during activation [F(x,y)] by the average image before activation [F_o(x,y); Figs. 1–5, 7, 8]. The division was performed only provided F_o(x,y) exceeded a threshold that was chosen to generate a sharp transition from the color distributions of the cells to their surroundings, which are shown in black. The green-blue speckle seen in single frames before activation indicates the level of noise in the recordings, and thereby the type of low level signals that should be discarded as random fluctuations. In contrast, the signals of physiological interest were characterized by their larger amplitudes (yellow, orange, and red distributions) and systematic progression from frame to frame.

Fig. 1.

Caffeine-induced changes in Ca²⁺-dependent fluorescence in voltage-clamped neonatal (Neo), juvenile (Juv), and adult cardiomyocytes. A: normalized fluorescence images in neonatal (a and b), juvenile (c and d), and adult (e and f) cardiomyocytes show the noise in a single frame as blue mottle before activation (a, c, and e; F/F_o ≅ 1, see color scale) and local increases in intracellular Ca²⁺ as brighter color distributions (green-yellow-orange) following application of 10 mM caffeine (b, d, and f). The timing of the images (a–f) is indicated in B, which shows the time course of the caffeine-induced cellular Ca²⁺ transients (ΔF/F₀) for representative neonatal, juvenile, and adult cardiomyocytes. The histograms show pooled data for cellular Ca²⁺ transients in terms of their magnitude (C) and the time to development of 50% of the full response (D). The level of significance (***P < 0.001, unpaired t-test) and the number of cells examined (n) are indicated in the histograms. The confocal fluorescence images were recorded at a rate of 120 frames/s, and the cells were voltage-clamped at −70 mV and dialyzed with Ca²⁺-buffered pipette solution.

Fig. 2.

ATP-induced increases in intracellular Ca²⁺ concentration ([Ca²⁺]_i) in Ca²⁺-buffered developing cardiomyocytes. A: normalized fluorescence images from a neonatal cardiomyocyte show low noise before activation (control; F/F_o ≅ 1) and subsarcolemmal increases in [Ca²⁺]_i following application of 100 μM ATP. B: average cellular Ca²⁺ transients (ΔF/F_o) in neonatal (Neo; solid bars; 1 to 2 days old) and juvenile (Juv; hatched bars; 8–10 days old) cardiomyocytes in response to 100 μM ATP (left) or 100 nM endothelin-1 (ET-1; right). Voltage-clamped cells are same color scale as Fig. 1.

Fig. 3.

Caffeine- and ATP-activated Ca²⁺ transients and inward currents in juvenile cardiomyocytes. A: superimposed cellular Ca²⁺ transients (ΔF/F₀) upon extracellular application of 10 mM caffeine (Caff; red) and 100 μM ATP (green) for the voltage-clamped cell shown in the inset and in D. B: membrane current recorded at −70 mV during 2-s exposures to caffeine (red) and ATP (green). The duration of the solution exchange is indicated by the 2-s time bars in A and B. C: average values for Na⁺-Ca²⁺ current (I_NaCa)-flux during exposure to caffeine and ATP are based on the integral of the measured membrane current during its negative excursion (B). D: normalized images of Ca²⁺-dependent fluorescence before activation (a and e) and during exposure to caffeine (b–d) or ATP (f–h) at the times indicated in B. Voltage-clamped juvenile cells are same color scale as Fig. 1.

Fig. 4.

Inositol (1,4,5)-trisphosphate (IP₃)-triggered Ca²⁺ release in permeabilized neonatal (A) and juvenile (B) cardiomyocytes. A and B: cellular Ca²⁺ transients (FΔ/F₀) upon extracellular application of 20 μM IP₃ and normalized fluorescence images before (a and c) and after (b and d) application of 20 μM IP₃ at the times indicated next to the traces. The histogram in C compares the average responses in neonatal and juvenile cardiomyocytes (confocal imaging at 120 frames/s; the nonvoltage clamped cells were stained in Fluo-4AM and then permeabilized and finally maintained in simulated internal solution; nonpaired t-test; same color scale as Fig. 1). *P ≤ 0.05.

Fig. 5.

IP₃-gated Ca²⁺ signals in juvenile, permeabilized cardiomyocytes are reversibly suppressed by the IP₃-gated Ca²⁺ release channels (IP₃R) antagonist 2-aminoethoxydiphenylborate (2-APB; A and B) and are triggered only weakly by L-IP₃ (AB). A and C: normalized fluorescence images that were recorded in representative cardiomyocytes at the times indicated next to the tracings of cellular Ca²⁺ transients, i.e., before interventions (a, f, and h), during activation with 20 μM IP₃ alone (b and g), and when IP₃ was applied after 3 min incubation with 2 μM 2-APB (c), after 3 min of washout (e), or as the biologically inactive stereoisomer L-IP₃ (i). B: average amplitude of fluorescence signals (ΔF/F₀) evoked by IP₃ (green) in cells that were then incubated with 2-APB (violet) and again challenged with IP₃ (turquoise). D compares the average amplitudes of the cellular Ca²⁺ transients in another set of cells that were exposed sequentially to IP₃ (green) and L-IP₃ (yellow; permeabilized cells, paired t-test). *P ≤ 0.05; **P ≤ 0.01.

Fig. 7.

Spatial analysis of IP₃-mediated Ca²⁺ signaling events in permeabilized neonatal (A) and juvenile (B) cardiomyocytes. Each panel shows a map that identifies 8 regions of interest, 8 traces (1–8) showing the time course of the fluorescence changes in these regions, and 8 normalized fluorescence images (a–h) that were measured at the times indicated by letters and arrow heads next to the traces. The vertical scale bars indicate ΔF/F₀ values of 1 (A) or 4 (B; 20 μM IP₃; permeabilized cells; same color scale as Fig. 1).

Fig. 8.

Time course of local IP₃-activated Ca² signal before and after depletion of the RyR-gated Ca²⁺ stores. A: cellular Ca²⁺ changes (ΔF/F₀) upon extracellular application (arrow) of 20 μM IP₃ before (green) and after (orange) 3-min incubation with 10 mM caffeine. The boxed regions of the traces and of the sample frame (B) were expanded to show details of the onset of the IP₃-induced responses on an expanded time scale (C) and as rapidly changing normalized fluorescence images recorded at 240 frames/s (D and E) at the times indicated in C (permeabilized juvenile cardiomyocyte; expanded color scale).

To reduce photobleaching, the laser beam that was used for excitation of Fluo-4 was electronically shuttered and triggered to open by the command of the pClamp program only during acquisition of data.

Statistical analyses.

All data are presented as means ± SE, n, where n is the number of cells examined. The unpaired Student's t-test was used to evaluate differences between measurements from different populations of cells (e.g., neonatal vs. juvenile cardiomyocytes), whereas the paired t-test was used to compare different pharmacological interventions in the same population of cells (Figs. 5, B and D, and 6, C and D). Differences were considered statistically significant when P < 0.05. Higher degrees of significance are indicated with two (P < 0.01) or three (P < 0.001) asterisks.

Fig. 6.

Contribution of ryanodine receptors (RyR)-gated Ca²⁺ stores to the IP₃-triggered Ca²⁺ releases. A: cellular Ca²⁺ transients (ΔF/F₀) in a permeabilized juvenile cardiomyocyte upon extracellular application of 20 μM IP₃ (left), 10 mM caffeine (inset), and IP₃ after 3 min incubation in caffeine (right). Selected frames (a-c) show normalized fluorescence images at the times indicated. B: comparison of the average Ca²⁺ transients (ΔF/F₀) in atrial (a) and ventricular (v) cardiomyocytes from neonatal (Neo; open bars) and juvenile (Juv; hatched bars) rats. The Ca²⁺ transients were evoked by 10 mM caffeine (red) or 20 μM IP₃, either by itself (green) or after 3 to 4 min incubation with 10 mM caffeine (orange), 40 μM ryanodine (blue), or 1 mM tetracaine (purple). C: IP₃-activated Ca²⁺ transients normalized relative to the caffeine-induced transients measured in the same cells. D: suppression of IP₃-activated Ca²⁺ signals by pretreatment with caffeine, ryanodine, and tetracaine. Permeabilized cells are same color scale as Fig. 1; n values are indicated at each bar. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.

RESULTS

RyR Ca²⁺ signals in dialyzed cardiomyocytes.

Excitation-contraction coupling in neonatal cardiomyocytes appears to be regulated by Ca²⁺ influx through the sarcolemma, with the L-type Ca²⁺ channel being the main source for trans-sarcolemmal Ca²⁺ flux (2, 7, 14, 34, 55). However, although RyRs are functional in neonatal cardiomyocytes, Ry only reduces Ca²⁺ transients by ∼15% in the day 1 neonatal rat heart versus 88% in juvenile cardiomyocytes, suggesting an increased reliance on RyR as a function of age (10, 11). Thus, to test the hypothesis that RyRs are upregulated as a function of maturation, we compared RyR-gated Ca²⁺ release in acutely isolated neonatal (days 1 to 2) and juvenile (days 8 to 10) cardiomyocytes, using the results from the widely studied adult cardiomyocytes as a benchmark for full maturation of RyR-mediated Ca²⁺ signaling.

Freshly isolated cardiomyocytes were voltage-clamped and dialyzed with a Ca²⁺ buffered internal solution, and conditioning pulses from −90 to −10 mV were applied at 5-s intervals to maintain the Ca²⁺ load of the SR throughout the experiments. In between conditioning pulses and with the cardiomyocytes being voltage-clamped at −70 mV to prevent activation of voltage-gated ion channels, RyR-mediated Ca²⁺ release was activated using rapid application of 10 mM caffeine, a RyR agonist, and changes in intracellular Ca²⁺ were detected using confocal Ca²⁺ imaging.

Figure 1A shows a comparison of the caffeine-activated Ca²⁺ transients in neonatal (b), juvenile (d), and adult (f) cardiomyocytes, with the normalized images recorded before activation being shown as well (a, c, and e, respectively), to give an indication of the spatial distribution of the caffeine-induced Ca²⁺ signal. The fluorescent signal Ca²⁺ transients of the neonatal (Neo), juvenile (Juv), and adult cardiomyocytes imaged at 120 Hz are shown in Fig. 1B. Comparisons of both the images (Fig. 1A) and the Ca²⁺ transients (Fig. 1B) indicate that maturation from neonate to adult is accompanied by larger and more uniformly distributed caffeine-activated Ca²⁺ transients. Indeed, analysis of the amplitude of the caffeine-induced Ca²⁺ transients (Fig. 1C) reveals that the magnitude of the caffeine Ca²⁺ transient increased with the age of the cardiomyocytes, with the neonates having the smallest amplitude (ΔF/F₀ = 0.160 ± 0.022; n = 7) and significantly larger amplitudes being observed in both juveniles (Juv: ΔF/F₀ = 0.483 ± 0.065; n = 3; P = 0.0003) and adults (ΔF/F₀ = 0.705 ± 0.18; n = 2; P = 0.0005). Figure 1D shows that the time required for development of 50% of the full amplitude of the caffeine-induced Ca²⁺ transient was generally shorter in the juvenile (Tau₅₀ = 52 ± 14 ms; n = 3) and adult (Tau₅₀ = 61 ± 1 ms; n = 2) compared with the neonatal (Tau₅₀ = 220 ± 50 ms; n = 7) cardiomyocytes. The data demonstrate an increase in the velocity and magnitude of RyR-gated Ca²⁺ release as a function of development and support our hypothesis that RyRs are upregulated during postnatal development, with neonatal and juveniles representing two stages of progressive maturation.

ATP-activated IP₃R Ca²⁺ signals in dialyzed developing cardiomyocytes.

Although IP₃-mediated Ca²⁺ signaling plays a large role in excitation-contraction coupling (ECC) of embryonic cardiomyocytes (43, 53), the role of IP₃-activated Ca²⁺ signaling in acutely isolated developing cardiomyocytes has never been studied before. The documented differences in the maturity of RyR-gated Ca²⁺ stores in neonatal and juvenile cardiomyocytes make these stages of development ideal to test whether parallel changes take place in IP₃-activated Ca²⁺ signaling, with IP₃-activated Ca²⁺ signaling perhaps having a reduced role as the role of RyR-gated Ca²⁺ signaling increases. To test this hypothesis, we compared the Ca²⁺ signals of acutely isolated neonatal and juvenile cardiomyocytes in response to activation of IP₃ by rapid application of 100 μM ATP, which binds to extracellular purinergic receptors to activate the breakdown of PIP₂ into IP₃ (52). Similarly, 100 nM endothelin-1 (ET-1), which binds to specific cell surface receptors to generate IP₃ (54) was tested as an alternative means to activate the IP₃-signaling pathway.

Ca²⁺-buffered, voltage-clamped rat cardiomyocytes produced variable responses to ATP and ET-1, ranging from rapid whole cell Ca²⁺ transients to an increase in the diastolic [Ca²⁺]_i. Figure 2A shows representative normalized images of a voltage-clamped neonatal cardiomyocyte before (control) and after exposure to ATP, revealing that ATP activated an increase in Ca²⁺ fluorescence mostly confined to the subsarcolemmal area of the cell. The spatial profiles for ET-1 Ca²⁺ responses were virtually identical (not shown). As summarized in Fig. 2B, application of ATP generated an increase in Ca²⁺-dependent fluorescence in both age groups with the neonatal cardiomyocytes producing a ΔF/F₀ = 0.040 ± 0.004 (n = 9) and the juveniles a ΔF/F₀ = 0.10 ± 0.04 (n = 6). Exposure of the same cells to ET-1 produced similar results, revealing an increase in the basal Ca²⁺ level of ΔF/F₀ = 0.030 ± 0.004 (n = 6) in the neonates and ΔF/F₀ = 0.040 ± 0.009 (n = 6) in the juveniles. Interestingly, maturation from the neonatal to the juvenile state appeared to be accompanied not only by an increase in RyR-gated Ca²⁺ stores but also by maintenance, or slight upregulation, of the IP₃R-gated Ca²⁺ stores. Although the differences in IP₃-mediated Ca²⁺ signaling only represent a trend, it was a surprising observation that we subjected to more extensive testing in permeabilized cells.

Figure 3 shows a representative example selected from a number of voltage-clamped Ca²⁺-buffered juvenile cardiomyocytes where we compared the magnitude and time course of the ATP- and caffeine-activated Ca²⁺ transients. As shown in Fig. 3A, superimposing the whole cell Ca²⁺ fluorescence traces for the two interventions on the same graph reveals different kinetics and magnitudes to the caffeine and ATP responses. Tau₅₀ of the caffeine-induced Ca²⁺ transient was shorter than that of ATP (26 ms vs. 240 ms). In addition, the amplitude of the caffeine-induced Ca²⁺ transient was larger than that of ATP (ΔF/F₀ = 0.70 vs. 0.25). The cellular images shown in Fig. 3D, taken at the time points indicated in the corresponding traces in Fig. 3A, show that caffeine-induced rise in Ca²⁺ occurred simultaneously in both the peripheral and central regions of the cell (Fig. 3D, b-d). The appearance of the caffeine-activated Ca²⁺ signal at the top the cell (Fig. 3D, b and c) corresponds to the area of the cell exposed to caffeine first. These data indicate that RyRs are distributed throughout the developing cardiomyocytes and agree with the findings of other researchers (15).

The ATP Ca²⁺ signal appears to develop first in the membrane-associated peripheral region (Fig. 3D, f and g), with delayed and incomplete invasion of the center of the cell (Fig. 3D, h). The subsarcolemmal origin of the response correlates with the data from neonatal cardiomyocyte shown in Fig. 2. The following, much larger magnitude of the ATP-activated Ca²⁺ transient, as well as the spread of the signal to some parts of the cell, suggests activation of the RyRs secondary to the primary IP₃-activated Ca²-release.

In cells where the caffeine- and ATP-induced Ca²⁺ transients were relatively large (Fig. 3A), we found that they were accompanied by significant transient inward currents (Fig. 3B; I_NaCa) of the type that in previous studies has been associated with activation of the Na⁺-Ca²⁺ exchanger by elevation of [Ca²⁺]_i (4). Note that the initial Ca²⁺ fluorescence time course (Fig. 3A) is similar to that of the transient inward current (Fig. 3B), with both sets of traces revealing a bigger delay for the ATP-induced response. Despite the great disparity in the amplitudes of the cellular Ca²⁺ transients produced by the two interventions, the amplitudes of the associated inward currents were comparable (n = 3), suggesting that the Na⁺-Ca²⁺ exchanger is quite sensitive to the localized subsarcolemmal Ca²⁺ transients evoked by ATP. The ATP-activated inward current was often followed by an outward maintained current that reversed at E_Cl and could be blocked by DIDS (unpublished data, Tufan and Morad). As summarized for three cells in Fig. 3C, the integral I_NaCa flux was larger for the caffeine- than for the ATP-activated Ca²⁺ release (1.47 ± 0.34 pC/pF vs. 0.27 ± 0.03 pC/pF), mirroring the larger magnitude of caffeine-induced Ca²⁺ fluorescence shown in Fig. 3A (Figs. 1 and 2). These results indicate that IP₃-induced Ca²⁺ release even contributes to cellular Ca²⁺ transients in highly Ca²⁺-buffered cells exclusive of Na⁺ or L-type Ca²⁺ channel activation.

IP₃R-gated Ca²⁺ signals in permeabilized cardiomyocytes.

Although the voltage-clamped, Ca²⁺-buffered cardiomyocytes provided insight into the magnitude of RyR- and IP₃R-gated Ca²⁺ signals, one of the limitations in the previous set of experiments is the possibility that variability in the expression of extracellular receptors may limit the size and frequency of ATP- and ET-1-activated Ca²⁺ signals. Therefore, direct application of IP₃ in permeabilized cells, although not representing a physiological response, was used to evoke Ca²⁺ signals and to compare the size of IP₃R-gated Ca²⁺ stores in neonatal and juvenile cardiomyocytes, as well as the ability of IP₃-activated Ca²⁺ release to directly activate CICR.

As shown in Fig. 4, application of 20 μM IP₃ activated Ca²⁺ transients in both acutely isolated neonatal (Fig. 4A) and juvenile (Fig. 4B) cardiomyocytes. The insets show normalized fluorescence images before (Fig. 4B, a and c) and after (Fig. 4B, b and d) application of IP₃. As summarized in Fig. 4C, IP₃ evoked significantly larger Ca²⁺ transients in juvenile than in neonatal cardiomyocytes (ΔF/F₀ = 0.65 ± 0.10, n = 18 vs. ΔF/F₀ = 1.45 ± 0.36, n = 25, P = 0.07). The use of 1 μM instead of 20 μM IP₃ produced fluorescence signals with similar amplitudes, but with delayed kinetics (data not shown).

To ensure that the observed effects of IP₃ were due to IP₃-dependent Ca²⁺ release, myocytes were exposed to IP₃ before and after incubation in solution containing the IP₃R antagonist 2-aminoethoxydiphenylborate (2-APB). The traces and the corresponding images in Fig. 5A show that IP₃ given before (Fig. 5A, a and b) and after washout of 2 μM 2-APB (Fig. 5A, e) produced an increase in the cytosolic [Ca²⁺] that was absent when the cells were treated with 2-APB alone (Fig. 5A, c) and with IP₃ in the presence of 2-APB (Fig. 5A, d). Average values from six cells (Fig. 5B) demonstrate that the Ca²⁺ signal produced by IP₃ alone (ΔF/F₀ = 0.20 ± 0.04) was much larger than the Ca²⁺ signals that were measured after incubation with 2-APB when the cells were either exposed to IP₃ (ΔF/F₀ = 0.07 ± 0.01; P = 0.02) or a recording was performed without intervention to ascertain the decay of the preceding Ca²⁺ signal and evaluate noise, bleaching, and the stability of the cells (ΔF/F₀ = 0.07 ± 0.01; P = 0.01).

Similarly, Fig. 5C shows sample records from a series of experiments where individual cells were found to respond more strongly to the naturally occurring D-IP₃ (Fig. 5C, f and g) than to its stereoisomer L-IP₃ (Fig. 5C, h and i). As illustrated in the Fig. 5D, the average Ca²⁺ signals produced by 20 μM D-IP₃ were greatly reduced when the cells were exposed to the same concentration of the L-isomer (ΔF/F₀: 0.20 ± 0.03 vs. 0.06 ± 0.01; n = 7; P = 0.0011). These control experiments show that the observed IP₃-activated Ca²⁺ transients are specific for IP₃Rs and IP₃.

Contribution of RyR-gated Ca²⁺ release to IP₃-mediated Ca²⁺ signal.

To evaluate the contribution of RyR-gated Ca²⁺ release to the IP₃-activated Ca²⁺ signals, we performed experiments where Ca²⁺ release via RyRs was blocked by different interventions. Figure 6A shows changes in the IP₃-activated Ca²⁺ signal in a representative juvenile cardiomyocyte before (Fig. 6A, a and b, left) and after (Fig. 6A, c, right) incubation of the cardiomyocyte with 10 mM caffeine to deplete RyR-gated Ca²⁺ secondary to the initial large Ca²⁺ release as seen in the inset. Fig. 6B compares the accumulated results from atrial (Fig. 6B, a) and ventricular (v) cardiomyocytes from neonates (Neo) and juveniles (Juv). The Ca²⁺ signals in response to caffeine (red) were almost fourfold (3.9 ± 0.9) larger in juvenile than in neonatal cells, and similar increases were observed in response to IP₃ in the absence (green) or presence (orange; Fig. 6B) of caffeine. These observations are based on pooled data from atrial and ventricular cardiomyocytes (Fig. 4C) since the permeabilization studies generally showed no significant differences between these cell types when examined at the same age. The only departure from this pattern was found in the response to IP₃ alone, where the upregulation in juvenile cardiomyocytes cells surprisingly appeared to be limited to the ventricular cells. When normalized relative to the caffeine-induced Ca²⁺ signals, those evoked by IP₃ amounted to about half (53 ± 9%; n = 43; dashed line) and, overall, showed no significant increase in juveniles although in this age group the difference between atrial and ventricular cardiomyocytes was noticeable (Fig. 6C). Pretreatment with caffeine reduced the IP₃-induced Ca²⁺ signals to 39 ± 5% (n = 27), and similar reductions were found when the RyR-gated Ca²⁺ stores were depleted with 40 μM ryanodine or 1 mM tetracaine was added to block the gating of the RyRs (Fig. 6D).

These data, which are consistent with the findings from Ca²⁺-buffered voltage-clamped cardiomyocytes (Fig. 2), suggest that, although the maturation of juvenile cardiomyocytes is accompanied by an increase in both caffeine-sensitive RyR-gated Ca²⁺ stores and IP₃R-gated Ca²⁺ stores, a change in the relative contribution of the two signaling pathways is seen only as differential development of atrial versus ventricular cells.

Spatial analysis of IP₃-mediated Ca²⁺ signaling events.

The above results suggest that the IP₃-activated Ca²⁺ release directly activates the RyRs by CICR in both neonatal and juvenile cardiomyocytes, but it remains to be shown where in the cell this communication may occur and how it may correlate with the locations of the IP₃- and RyRs. Figure 7 illustrates the spatial development of IP₃-induced Ca²⁺ transients in representative neonatal (Fig. 7A) and juvenile (Fig. 7B) cardiomyocytes. Regions of interest in the cells were identified with numbers on the sketched cells, and the Ca²⁺ signals in these numbered regions were followed in the corresponding numbered traces (1–8), thereby mapping the time course of the IP₃-generated local Ca²⁺ responses. Labeled time marks on the traces correspond to the normalized cellular images (Fig. 7B, a–h). In both age groups, the IP₃-activated responses were first observed at subsarcolemmal locations (Figs. 7A, c, and B, b) and traveled along the periphery of the cell before variably moving into the center. This wave-like spread of the IP₃-generated Ca²⁺ signals near the cell surface suggests subsarcolemmal colocalization of IP₃Rs and RyRs and was characteristic of permeabilized cells, where failure of whole cell signal integration may be directly linked to the loss of membrane potential. The minimal involvement of the central portions of the cell (e.g., Figs. 7A region/trace 3, and B, region/trace 4) may be indicative of a spatial disconnect between IP₃Rs and RyRs in the center.

Spatial analysis of IP₃-activated Ca²⁺ release after depletion of RyR-gated Ca²⁺ stores.

Although the strong IP₃-induced Ca²⁺ signals, which spread near the surface of neonatal and juvenile cells, are likely to reflect subsarcolemmal colocalization of IP₃Rs and RyRs, as found in adult atrial cells (31), it is unclear whether these signals may swamp the less conspicuous Ca²⁺ release activity of functional IP₃Rs that may be present at central locations but may operate without the support of CICR from nearby RyR-gated Ca²⁺ stores. Therefore, we measured the subcellular distribution of IP₃-induced Ca²⁺ signals under conditions where the RyR-mediated component of the Ca²⁺ release was inhibited.

Figure 8 shows the spatial development of an IP₃-induced Ca²⁺ transient in a representative juvenile cardiomyocyte before and after incubation in caffeine to deplete and prevent refilling of the RyR-gated Ca²⁺ stores. As shown in Fig. 8A, the whole cell Ca²⁺ fluorescence changes were followed before (green) and after (orange) depletion of the RyR-gated Ca²⁺ pools via long-term incubation (3 min) in 10 mM caffeine. Application of IP₃ (arrow) under control conditions produced a strong response (green trace) that had a rapid upstroke (Tau₅₀ = 12 ms) indicative of CICR and was regenerative with a period of ∼800 ms. The cellular response to IP₃ was greatly suppressed in the presence of caffeine (orange trace). To focus on the genesis of these Ca²⁺ signals, we expanded the time scale of the first 40 ms of the recordings (Fig. 8C) and inspected associated consecutive frames recorded at 240 Hz (Fig. 8, D and E). Before the caffeine treatment, the IP₃-activated Ca²⁺ transient was generally seen to originate at a single location near the cell surface (Fig. 8D, image d) and to spread steadily, primarily in the longitudinal direction of the cell (starting at Fig. 8D, frame f), eventually developing large cellular Ca²⁺ transients with distributions (not shown) similar those in Fig. 7. Although the large cellular Ca²⁺ transients disappeared after long-term incubation with caffeine (Fig. 8A), IP₃ was now seen to generate brief localized Ca²⁺ signals throughout the cytoplasm (Fig. 8E). In some cases, these Ca²⁺ signals could be seen clearly in consecutive frame (Fig. 8E, d and e, lower arrow), but more often a clear Ca²⁺ release event in a single frame was accompanied only by faint signals at the same location in the previous or following frame. Some sites appeared to produce repeated Ca²⁺ releases at intervals of ∼20 ms (e.g., Fig. 8E, e → j, top arrow, and d and e → h, bottom arrow). These results suggest that functional IP₃R-gated Ca²⁺ stores are present throughout the cardiomyocyte. Similar results were seen in the two neonatal and three juvenile cells analyzed.

To test for the presence of IP₃R Ca²⁺ stores in the perinuclear area, as has been described in cultured neonatal cardiomyocytes and adult cardiac cells (25), we performed experiments where we carefully adjusted the confocal plane to focus on the nucleus. Figure 9 shows whole cell changes in Ca²⁺ fluorescence in a representative juvenile cardiomyocyte in response to IP₃ under control conditions (Fig. 9b) and after incubation in 40 μM ryanodine to deplete RyR-gated Ca²⁺ stores (Fig. 9c). The IP₃ response under control conditions revealed a whole cell Ca²⁺ transient with a short rise time, whereas the IP₃ response after incubation with ryanodine was of a smaller magnitude (ΔF/F₀ = 3.05 for IP₃ vs. 0.76 for IP₃ in Ry) and developed much more slowly, which is consistent with the loss of RyR-mediated CICR. Fluorescence images at the peak of the cellular Ca²⁺ transients show that the Ca²⁺ signal invaded the entire cell before treatment with ryanodine (Fig. 9, image b), but afterward was largely limited to the nuclear region (Fig. 9, image c). Subsequent exposure to caffeine produced no response (data not shown). Similar results were seen in five cardiomyocytes. These findings indicate that developing cardiomyocytes have substantial RyR-insensitive IP₃R-gated Ca²⁺ stores in the perinuclear area.

Fig. 9.

IP₃-mediated perinuclear Ca²⁺ release in a permeabilized juvenile cardiomyocyte. Representative cellular Ca²⁺ transients (ΔF/F₀) and fluorescence images (a–c) evoked by extracellular application of 20 μM IP₃ before (left) and after (right) 5-min incubation in 40 μM ryanodine. The normalized fluorescence images were recorded at 240 frames/s before activation (a) and at the peak of the IP₃-activated responses, which normally invaded the entire cell (b) but were confined to the nuclear region after ryanodine treatment (c).

DISCUSSION

Here we have demonstrated that freshly dissociated cardiomyocytes from neonatal and juvenile rats have functional IP₃R-gated Ca²⁺ stores with kinetics and magnitudes that differ significantly from those of the RyR-gated Ca²⁺ releases (Figs. 1–4). The IP₃-mediated Ca²⁺ release is suppressed, not only by blocking the IP₃R with 2-APB or substituting L- for D-IP₃ (Fig. 5) but also by depleting the RyR-gated Ca²⁺ stores with caffeine or Ry or blocking the RyRs with tetracaine (Fig. 6). As discussed below, the kinetics, magnitudes, and subcellular distributions of the Ca²⁺ signals suggest that the primary subsarcolemmal IP₃-mediated Ca²⁺ release in both age groups is capable of activating local RyRs to produce a secondary CICR, even in the presence of high concentrations of fluorescent (Fluo-4) and nonfluorescent (EGTA) Ca²⁺ buffers (Figs. 3, 4, and 7). Comparisons of the RyR- and IP₃R-gated Ca²⁺ releases during development show that both release systems are upregulated from the neonatal (days 1 to 2) to the juvenile (days 8-10; Fig. 6B) stage but that the IP₃-mediated response with age overall changes little when normalized relative to CICR (Fig. 7C). Spatial analysis revealed Ry-insensitive IP₃-mediated Ca²⁺ signals that are present throughout the cells at central (Fig. 8) and perinuclear (Fig. 9) locations but are most prominent in subsarcolemmal regions (Figs. 1–5 and 7) where they evoke CICR (Figs. 3, 4, and 6) and significant inward I_NaCa (Fig. 3). Overall, our data indicate that developing cardiomyocytes have an intermediate Ca²⁺ signaling phenotype between the early postnatal and adult cardiomyocytes showing vestigial IP₃R- gated Ca²⁺ signaling in both atrial and ventricular myocytes capable of directly triggering CICR.

RyR-gated Ca²⁺ release in postnatal developing cardiomyocytes.

Analysis of the caffeine-activated Ca²⁺ release pools in voltage-clamped, Ca²⁺ buffered cells suggests enhancement of the size and kinetics of RyR-gated Ca²⁺ stores during transition from neonatal to juvenile cardiomyocytes (Fig. 1). In permeabilized myocytes this enhancement was nearly fourfold (Fig. 6). The upregulation of the size of the RyR-gated Ca²⁺ stores is consistent with previous studies of developing cardiomyocytes (15, 45). The faster kinetics in juvenile versus neonatal cells reported here (Fig. 1D) may reflect an altered sensitivity or distribution of the RyRs (10).

Characterization of IP₃R-gated Ca²⁺ release in postnatal developing cardiomyocytes.

Although other cellular studies of cardiac development have used cultured cardiomyocytes, to our knowledge this is the first study where IP₃R-gated Ca²⁺ signals were studied in acutely isolated cells. We pursued this strategy because cultured cardiomyocytes have been known to form spontaneously beating interconnected monolayers and drift to heterogeneous populations with variable functional and electrophysiological characteristics (39), dedifferentiation of cardiac structure (27, 38), and altered expression of IP₃Rs (58) and other Ca²⁺ signaling proteins (48), all of which tend to complicate the interpretation of results and compromise physiological relevance. Nevertheless, it should also be recognized that, even without subsequent culturing, the enzymatic dispersion of cells appears to cause marked changes in the phosphoinositol-signaling pathways (57, 58), and it may be asked whether permeabilized cells may be better suited than dialyzed voltage-clamped cells to evaluate IP₃-signaling by excluding confounding factors resulting from heterogeneity or enzymatic degradation of sarcolemmal ATP- and ET-1 receptors.

Application of ATP and ET-1 in voltage-clamped, Ca²⁺-buffered cells (Figs. 2 and 3) and IP₃ in permeabilized cells (Figs. 4–9) produced results, ranging from a rise in the basal [Ca²⁺]_i level to Ca²⁺ waves and transients throughout the cells at subsarcolemmal, central (Fig. 8), and perinuclear (Fig. 9) locations. Both voltage-clamp (Fig. 3) and permeabilization (Fig. 7) experiments revealed IP₃-induced Ca²⁺ transients that followed a subsarcolemmal activation pathway. Although the IP₃R-gated Ca²⁺ signals were upregulated with age (Figs. 2 and 6), the magnitude of these responses was found to depend on prior exposure to compounds that targeted the RyR-gated Ca²⁺ stores (Figs. 6 and 8), indicating that the two types of Ca²⁺ signal are interdependent.

Interplay between RyR- and IP₃R-gated Ca²⁺ stores.

The reduction in the IP₃-mediated Ca²⁺ signals following prolonged exposure to caffeine (Fig. 6) suggests that Ca²⁺ release via IP₃R may stimulate CICR from nearby, functionally distinct, RyR-gated Ca²⁺ stores. Alternatively, some subcellular Ca²⁺ storage compartments may be gated by both Ry- and IP₃Rs. The first possibility is consistent with our observations that the IP₃-mediated Ca²⁺ signals 1) generally displayed an initial slow rise that is followed by a fast caffeine-sensitive spike (Figs. 3A, 4, 6A, and 8); 2) propagated slowly under the membrane (Figs. 7 and 8); and 3) were suppressed equally by caffeine and Ry, both of which deplete the RyR-gated Ca²⁺ pools, and by tetracaine (Fig. 6), which is known to enhance the SR Ca²⁺ content by stabilizing RyRs in the closed state (13, 19, 37, 46). In addition, CICR is a promiscuous process that can be initiated by diverse Ca²⁺ triggers including Ca²⁺ channels within the same t-SR junctions, reverse mode Ca²⁺ transport by NCX (26), T-type Ca²⁺ current (47), or photolysis of caged Ca²⁺ (29), and, therefore, probably also to the IP₃-induced Ca²⁺ signals. Our use of moderate and high concentrations of Ca²⁺ buffers was, in part, intended to reduce the free diffusion distance of Ca²⁺ (1) and thereby intercept the IP₃R-gated Ca²⁺ signals before they could trigger CICR. In this regard, we may have been successful in limiting the spread of CICR to the interior of the cells, but it would appear that higher concentrations of fast Ca²⁺ buffers, in combination with ultrastructural studies, would be required to quantify distances between IP₃- and RyRs.

On the other hand, the alternative possibility that some IP₃-gated Ca²⁺ pools may be partially emptied by prior exposure to caffeine does not necessarily require a high degree of colocalization of the two receptor types, as seen in adult atrial cells (31), only equilibration of Ca²⁺ between the subcompartments they gate. Such transfer of Ca²⁺ has been observed between SR and perinuclear stores, but it notably only occurs on the time scale of minutes (59). Finally, the open probability of IP₃Rs is steeply dependent on Ca²⁺ in the range from 10 to 100 nM (40), but our buffering of [Ca²⁺]_i at 100 nM probably minimized such effects. Overall, present and published results favor the interpretation that the initial IP₃-mediated Ca²⁺ release is boosted by local CICR via RyRs.

Spatial profile of IP₃R-gated Ca²⁺ signaling.

In highly Ca²⁺-buffered cells, application of ATP or ET-1 (Figs. 2 and 3) generated increases in the cytosolic Ca²⁺ at the cell periphery that, in juvenile cells, infrequently triggered global Ca²⁺ transients. In permeabilized cells, IP₃ produced RyR-sensitive global Ca²⁺ transients that originated in subsarcolemmal regions and propagated to different degrees into the cell interior (Figs. 7 and 8). These IP₃-activated Ca²⁺ transients in neonatal and juvenile cardiomyocytes have spatial Ca²⁺ profiles similar to those of adult atrial cells (21, 28, 31, 32) where they may be explained by the existence of two regularly arranged populations of atrial RyRs, of which those near the cell surface colocalize with IP₃R in the junctional SR (28, 31). Our caffeine-activated Ca²⁺ profiles (Figs. 1, 3, 6, and 9) suggest that functional RyRs are uniformly distributed throughout the cell in the developing cardiomyocytes. However, mapping of IP₃-activated CICR pathways in either voltage-clamped (Fig. 3) or permeabilized (Fig. 7) cells suggests that a similar spatial disconnect exists between the sarcolemma and the interior SR, as that found in atrial cardiomyocytes (32). The presence of functional IP₃Rs throughout the cell (Fig. 8E), in combination with the predominant subsarcolemmal localization of IP₃R-gated Ca²⁺ signals (Figs. 2–5 and 7), suggests that those in the interior may be relatively few or spatially dissociated from the centrally located RyRs.

Thus, in developing cardiomyocytes, IP₃Rs may have a closer spatial proximity to RyRs in the peripheral versus central and nuclear regions. The general absence of differences in RyR- and IP₃R-gated Ca²⁺ signaling between atrial and ventricular cardiomyocytes in the two age groups (Fig. 6) may indicate that developing cardiomyocytes are physiologically similar to atrial cardiomyocytes, with the distinctive ventricular physiology becoming more apparent after postnatal day 10. The finding that IP₃-triggered Ca²⁺ signals of juveniles are larger in ventricular than in atrial myocytes (Fig. 6B) is surprising considering the importance of IP₃-mediated signaling in adult atrial cells (32).

Role of IP₃R-gated Ca²⁺ signaling in postnatal cardiomyocytes.

There is evidence that IP₃-dependent Ca²⁺ release can exert a positive inotropic effect in adult atrial and ventricular cells (8, 62), but it is generally found to be minor in nondiseased tissue, especially the ventricle. In our voltage-clamped and permeabilized cells, ATP- or IP₃-activated Ca²⁺ transients suggest that IP₃ is capable of triggering Ca²⁺ transients independent of depolarization of the membrane. In addition, the finding that IP₃-triggered Ca²⁺ transients that were suppressed in the presence of caffeine, ryanodine, and tetracaine implicates a role for IP₃ in CICR in developing postnatal cardiomyocytes, where the resulting inward NCX current (Fig. 3) may contribute to membrane depolarization, as seen in pacing embryonic heart cells (23, 43) and in the genesis of arrhythmias (25).

The observation of a Ry-insensitive perinuclear IP₃-induced Ca²⁺ signal (Fig. 9) is consistent with previous studies (20, 22, 30, 60) and may implicate Ca²⁺-dependent nuclear transcription factors (25). Similarly, it has been suggested that the Ca²⁺ that is released in response to IP₃ may be taken up preferentially by mitochondria that thereby influence the spread of the Ca²⁺ signals (32) and are stimulated to synthesize ATP (17, 22, 24). These observations suggest that neurohumoral regulation of IP₃ production could affect the inotropic, chronotropic, and metabolic states of developing cardiomyocytes.

The prominent role of the IP₃-signaling pathways in cardiac diseases is generally characterized by activation of the renin-angiotensin and endothelin, enhanced synthesis of IP₃, and upregulation of IP₃Rs (25). In the present context, it may be relevant to emphasize the unexpected finding that IP₃ in juvenile rat ventricular cardiomyocytes generates strong subsarcolemmal Ca²⁺ signals (Figs. 2, 3, 5, and 7), with the distribution typically seen in atrial cells (21, 28, 31, 32) where upregulation of IP₃R is associated with the relatively common human condition of atrial fibrillation (5, 61).

GRANTS

This work was supported by National Heart, Lung, and Blood Institute RO1-HL-16152 (to M. Morad).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).