Confocal microscopy was used to investigate the temporal and spatial properties of Ca2+ transients and Ca2+ sparks in ventricular myocytes of the rainbow trout (Oncorhynchus mykiss). Confocal imaging confirmed the absence of T tubules and the long (∼160 μm), thin (∼8 μm) morphology of trout myocytes. Line scan imaging of Ca2+ transients evoked by electrical stimulation in cells loaded with fluo 4 revealed spatial inhomogeneities in the temporal properties of Ca2+ transients across the width of the myocytes. The Ca2+ wavefront initiated faster, rose faster, and reached larger peak amplitudes in the periphery of the myocyte compared with the center. These differences were exacerbated by stimulation with the L-type Ca2+ channel agonist (−)BAY K 8644 or by sarcoplasmic reticulum (SR) inhibition with ryanodine and thapsigargin. Results reveal that the shape of the trout myocyte allows for rapid diffusion of Ca2+ from the cell periphery to the cell center, with SR Ca2+ release contributing to the cytosolic Ca2+ rise in a time-dependent manner. Spontaneous Ca2+ sparks were exceedingly rare in trout myocytes under control conditions (1 sparking cell from 238 cells examined). This is in marked contrast to the rat where a total of 56 spontaneous Ca2+ sparks were observed in 9 of 11 myocytes examined. Ca2+ sparklike events were observed in a very small number of trout myocytes (15 sparks from 9 of 378 cells examined) after stimulation with either (−)BAY K 8644 or high Ca2+ (6 mM). Reducing temperature to 15°C in intact myocytes or permeabilizing myocytes to adjust intracellular conditions to favor Ca2+ spark detection was without significant effects. Possible reasons for the rarity of Ca2+ sparks in a cardiac myocyte with an active SR are discussed.
- calcium sparks
- calcium inhomogeneities
- calcium transients
- ryanodine receptor isoform
- sarcoplasmic reticulum
in adult mammalian ventricular myocytes, the well-developed T-tubular network ensures spatially homogeneous Ca2+ entry across the width of the cell (7). The T-tubular network is also responsible for the functional coupling of the sarcoplasmic reticulum (SR) to the sarcolemmal (SL) membrane throughout the entire cell volume, bringing the SL L-type Ca2+ channels (DHPRs) in close apposition to the SR Ca2+-release channels [ryanodine receptors (RyRs)] and thus providing the structural basis for excitation-contraction (E-C) coupling (52).
Myocytes from nonmammalian vertebrate species, such as the rainbow trout, are devoid of T tubules. This difference in ultrastructure is expected to have significant consequences for E-C coupling. Indeed, adult mammalian atrial myocytes lack a T-tubular network and show marked differences from ventricular myocytes in the spatial and temporal properties of cellular Ca2+ gradients. In adult mammalian atrial myocytes, voltage-dependent Ca2+ entry triggers Ca2+ release from peripheral SR couplings that subsequently induce the release of Ca2+ from SR stores in more central regions of the cell. The Ca2+ moves throughout the volume of the myocyte via propagated Ca2+-induced Ca2+ release between neighboring RyRs (19). In these cells, as with cultured ventricular myocytes and detubulated ventricular myocytes, there is a prominent V-shaped rise in the Ca2+ wavefront across the width of the cell with a marked time delay (50–300 ms) in peak Ca2+ between the cell periphery and the cell center (7, 19, 25, 53).
Mammalian neonatal myocytes and Purkinje cells also lack a T-tubular system but, in addition, have a poorly developed SR (5, 14, 51). The spatiotemporal paradigm for Ca2+ flux in these cell types is characterized by a peripheral rise in Ca2+ followed by diffusion of Ca2+ into the cell interior with limited propagated SR Ca2+ release. This arrangement results in a shallow V-shaped Ca2+ wavefront across the width of the myocyte with a delay (25–50 ms) in transient peak between cell periphery and cell center (14).
Similar to mammalian neonatal and Purkinje cells, the SR in rainbow trout myocytes is poorly developed in comparison with adult mammalian cardiac myocytes (35). Nevertheless, the possibility of an active role for the SR in cycling Ca2+ during E-C coupling in trout heart is supported by a number of observations. 1) Ultrastructure studies of rainbow trout ventricles clearly show SR surrounding the myofibrils and forming peripheral couplings with the SL (35). These couplings are often occupied by foot processes, which are probably RyRs (49). 2) Electrophysiological studies reveal significant (∼750 μM total Ca2+) steady-state Ca2+ stores in the trout SR that are releasable by caffeine (18, 38). 3) A physiological role for these Ca2+ stores is evidenced by the change in inactivation of the L-type Ca2+ channel current with changing SR Ca2+ content (18, 38). 4) Isometric tension in isolated ventricular muscle is reduced (by 5–25%, depending on temperature and stimulation frequency) when the SR is inhibited with ryanodine (Ry) (16, 36).
In this study, we investigated the temporal and spatial properties of Ca2+ transients and the occurrence of spontaneous Ca2+ sparks using laser scanning confocal microscopy in ventricular myocytes from the rainbow trout. Because this is the first published study conducted on fish myocytes, all experiments were run in tandem with rat ventricular myocytes as a positive control. Our aim was to assess the impact of myocyte morphology on cellular Ca2+ gradients in fish myocytes and to better understand the putative role of the SR in trout E-C coupling. We show that Ca2+ influx from the cell periphery releases peripheral SR Ca2+ stores; Ca2+ then diffuses to the cell interior, where it releases Ca2+ from more centrally located SR Ca2+ stores. Under the conditions of our study, spontaneous Ca2+ sparks in trout ventricular myocytes were extremely rare.
Female rainbow trout (Oncorhynchus mykiss, mean body mass of 137 ± 5 g, n = 83) were purchased from Glasshouses Trout Farm (North Yorkshire, UK). They were held indoors in large fiberglass tanks containing recirculated aerated fresh water for a minimum of 2 wk at 15°C before experimentation. Male Wistar rats were maintained in cages of four animals. Both species were exposed to a 12:12-h light-dark photoperiod and fed ad libitum with appropriate commercial pellets. All procedures were approved by the Animal Care and Welfare Committee of the University of Leeds and are in accordance with UK regulations.
Isolated myocyte preparation.
Detailed descriptions of myocyte preparation have been published previously for both rainbow trout (37, 48) and rat (8). In some experiments, myocytes were detubulated using formamide (6) or permeabilized by exposure to saponin (at 20–30 μl of 10 mg/ml stock in 1 ml of cell suspension) in a mock intracellular solution for 5 min before centrifugation and resuspension. Mock intracellular solution contained (in mM) 100 KCl, 15 HEPES, 10 phosphocreatine, 5 ATP, 5.75 MgCl2 (to give a free concentration of ∼1 mM), and 0.36 EGTA. CaCl2 concentration was varied to give free Ca2+ values between 100 and 1,000 nM. Calculations were based on the EGTA and ATP/phosphocreatine in the solution at a pH of 7.2 at 20°C.
Permeabilized cells were placed in an experimental chamber with a very small volume (<10 μl) (54). The chamber was placed on the stage of an inverted microscope (Zeiss LSM 510, Oberkochen, Germany), and Ca2+ was imaged in the same manner as described below for intact cells.
Visualization of cell membrane.
To visualize the SL membrane, myocytes from trout and rat were incubated for 2 min with the lipophilic fluorescent indicator di-8-ANEPPS (5 μM; Molecular Probes, Eugene, OR). Cells were then resuspended in extracellular solution [containing (in mM) 150 NaCl, 5.4 KCl, 1.5 MgSO4, 0.4 NaH2PO4, 2 CaCl2, 5 glucose, 5 pyruvate, and 10 HEPES and adjusted to pH 7.7 with KOH]. Myocytes were imaged using confocal laser scanning microscopy (Zeiss LSM 510) with 488-nm excitation light and detection at >505 nm. Consecutive plane scans (x-y) through the z-plane were used to construct a three-dimensional image (z-stack). We analyzed this image using Zeiss LSM 5 Image Examiner software to calculate cell length, width, and depth (see Table 1).
Myocytes from both rat and trout were loaded with 4 μM fluo 4-AM for 20 min at room temperature, followed by 30 min for deesterification, and then transferred to a bath on the stage of an inverted microscope (Zeiss LSM 510) and stimulated to contract via platinum plate electrodes (50–100 V, 5- to 15-ms pulses). Ca2+ was imaged by exciting fluo 4 at 488 nm and detecting emitted fluorescence at >505 nm. When cells were double labeled with fluo 4 and di-8-ANEPPS, emitted fluorescence was split to bandpass filters of 505–530 nm and 560–615 nm, respectively. A Zeiss Plan-Neofluor 1.2 water-immersion objective was used in all measurements. The pin-hole aperture was set to the size of the Airy disk to optimize z-axis resolution. Under these conditions, the half-width of the point spread function was measured as <0.3 μm in the x-y plane using fluorescent beads (diameter of 0.175 μm, Molecular Probes). We collected images by using repetitive line scans (1,000 lines of 512 pixels) every 2–4 ms across the width of the cell. All line scan images are presented as the original raw signal. Corresponding time traces show fluorescence as the ratio of fluorescence to background fluorescence (F/F0). F0 was measured in each cell in a region that did not have localized or transient fluorescent elevation. Peripheral Ca2+ flux was assessed by taking the average of the Ca2+ transients in the periphery of both sides of the myocyte (1–2 μm). Cell center Ca2+ measurements were assessed by measuring the Ca2+ transient in the center of the cell over a width of 2–4 μm. Every attempt was made to focus the imaging plane at the midpoint of cell depth (z-plane) to exclude contamination of our “cell center” measurements with Ca2+ influx from surface membranes. However, we cannot completely exclude the possibility of surface membrane influx, especially in cells with very narrow depths (3–4 μm). In some trout myocytes, there were intense, localized, and time invariant fluorescent signals detected. Because these probably represent dye accumulating in intracellular compartments (see Ref. 22), these areas were avoided when focusing the line scans. Ca2+ transient amplitude, transient rise time, and rate of rise were measured between diastolic and peak F/F0 signal. Signal exponential functions were fit to the decay of the transient to assess the time constant of decay.
When scanning for Ca2+ sparks, we stopped electrical stimulation and moved the imaging plane throughout the depth (z-plane) and often focused at the surface membrane where Ca2+ sparks resulting from peripheral RyRs are thought to occur. Criteria for acceptance of an event as a spontaneous Ca2+ spark (i.e., those that occur in the absence of membrane depolarization) were as follows: amplitude >1.3 F/F0, full width >0.5 μm, a rise time between 5 and 35 ms, and a 50% decay time between 10 and 40 ms (10, 33, 34). Sparks were assessed by visual examination of scan records from quiescent cells. As a check on our visual detection, the automatic spark detection program with IDL software (9) was used on trout images that were thought most likely to give positive detections. Using a detection amplitude of 3.8 SD (9) in 12 trout cells (both quiescent and field stimulated), we identified 18 sparks by eye and 19 by the automated detection program. We performed analyses of Ca2+ transients and Ca2+ sparks using Clampfit 8.1 software (Axon Instruments) and SigmaPlot 8.0 (Jandel Scientific).
To investigate the relative role of SL and SR Ca2+ flux, cells were bathed with solutions containing either the L-type Ca2+ channel agonist or SR inhibitors. (−)BAY K 8644 (100 nM) stimulates SL Ca2+ influx by increasing the open probability of the L-type Ca2+ channel without affecting single-channel conductance. Furthermore, it does not activate an intracellular signaling cascade that may affect SR function (26). The SR Ca2+ release channel inhibitor Ry (10 μM) and the SR Ca2+ ATPase inhibitor thapsigargin (Tg; 2 μM) were applied in combination to block SR function. For SR inhibition experiments, cells were bathed with the inhibitor solution during imaging and were also pretreated with Ry and Tg for a minimum of 20 min before experimentation. To examine SR Ca2+ content, some cells were exposed to a rapid application of caffeine (10 mM).
Western blot analysis.
Ventricles from trout, rat, and cow (obtained from a local abattoir) were ground by mortar and pestle under liquid nitrogen. Bovine ventricle was included in RyR analysis because the density of RyR2 in bovine hearts is expected to be less than that in rat and thus would provide more of a continuum on which to place the novel results from the rainbow trout. Homogenization buffer (10 mM EDTA, 300 mM sucrose, 0.35 mM SDS) containing protease inhibitors was added to the ground tissue, and then this was frozen at −80°C until use. Myocardial proteins were separated by SDS-PAGE on 6% acrylamide gels as described (27) using the Mini-Protean system (Bio-Rad Laboratories). Each lane of the gel was loaded with 30 μg of protein, which was determined using BCA protein analysis kit (Sigma, St. Louis, MO). After separation, proteins were transferred to polyvinylidene difluoride membranes by semi-dry blotting for 1.5 h at 60 mA per gel. Nonspecific sites were blocked by overnight immersion at 4°C in Tris-buffered saline (50 mM Tris, 150 mM NaCl, 0.1% Tween 20, pH 7.4) containing 5% (wt/vol) dried skimmed milk. Membranes were then probed with two different primary antibodies to the mammalian RyR. The first primary antibody (34C, mouse monoclonal, Developmental Studies Hybridoma Bank) detects most known mammalian RyR isoforms. In fish skeletal muscle, this antibody detects the RyR1 isoform (31). Immunoreactivity of this antibody in trout heart is unknown, but it is has been shown to immunoreact with the cardiac RyR in frog heart (46) but not common carp (Cyprinus carpio) (11). The second antibody (C3-33, mouse monoclonal; ABR) is more specific for the cardiac isoform in mammals, only very weakly detecting RyR1. However, it is known to detect both RyR1 and RyR3 in fish skeletal muscle (31). Immunoreactivity was visualized after probing with secondary goat anti-mouse antibodies (1:10,000; Jackson ImmunoResearch Laboratories) using a peroxidase-based chemiluminescent substrate kit (Super Signal, Pierce). Band densities were assessed using Scion Image software (NIH freeware) and normalized to the RyR2 rat signal. Normalizations were only conducted on bands from the same gel with the same amount of protein.
All statistical data are presented as means ± SE, with n equal to the number of observations, number of cells, or number of animals as specified in the text. Statistical tests were Student’s paired t-tests or ANOVA with Student-Newman-Keuls post hoc. For nonnormally distributed data, Mann-Whitney paired tests or ANOVA on ranks with Dunn’s post hoc were used. Significance was accepted as P < 0.05. Details for statistics preformed on each data set are given in the figure and table legends.
Morphometric features of trout ventricular myocytes are provided in Fig. 1 and Table 1. Most trout myocytes are long and thin, typified by the image in Fig. 1A. However, a small portion (∼10%) of ventricular myocytes are more sheetlike, being wide but narrow in depth (Fig. 1B). We believe both populations of cells are from the spongy myocardium, which makes up the majority of the fish heart. Although we cannot exclude the possibility, it is unlikely that the two populations of myocytes reflect the spongy and compact myocardium. This is because the outer compact layer in a ∼140-g trout is very thin (<1 mm), and the enzymatic digestion of the heart is terminated before disruption of this layer. Myocytes of both morphologies are combined in the mean data given in Table 1. However, an examination of individual trout ventricular myocytes reveals that the widest myocytes (i.e., 24.4 μm) also had the narrowest depth (i.e., 3.2 μm). Correspondingly, myocytes with narrow widths (i.e., 6.6 μm) had deeper depths (i.e., 5.4 μm). Thus it appears that the cross-sectional area (∼41 μm2) is fairly well maintained over a range of myocyte dimensions in the trout ventricle.
There are no T tubules in trout myocyte, which contrasts with the clear SL invaginations present in the rat ventricular myocyte (Fig. 1C). Formamide treatment detubulates rat ventricular myocytes (Fig. 1D). The narrow width and depth of the trout myocyte combined with the lack of T tubules has implications for cellular Ca2+ gradients as discussed below.
Temporal and spatial properties of Ca2+ transients.
Given the morphological differences between rat and trout myocytes, temporal and spatial differences in the Ca2+ transient are expected between these species. Figure 2 shows line-scan images of Ca2+ transients across the width of a trout myocyte, a control rat myocyte, and a rat myocyte that has been partially detubulated. The rise of Ca2+ in the periphery of the trout myocyte is larger and faster compared with the cell center (Figs. 2A and 3). There was a 5.8 ± 2.5 ms (n = 20)-delay in the initial Ca2+ rise between the cell periphery and cell center. This is in contrast to the uniform rise in the Ca2+ wavefront across the width of the rat myocyte (Fig. 2B). The difference is certainly related to the presence of T tubules in rat myocytes, which conduct the depolarizing wave deep into the cell, allowing nearly simultaneously SL Ca2+ influx and SR Ca2+ release across the width of the cell. The role of the T tubules in this response is illustrated in Fig. 2C, where the simultaneous rise in the Ca2+ wavefront is compromised after the cell has been treated with formamide to disrupt the T tubule network.
In response to field stimulation under control conditions (2 mM Ca2+), Ca2+ transients were significantly larger in rat myocytes (3.7 ± 0.8 F/F0; n = 8) than in trout myocytes (1.81 ± 0.12 F/F0; n = 22) with similar rise times (124 ± 8 ms for rat and 123 ± 13 ms for trout). Time constants of decay tended to be longer in trout, but this trend was not statistically resolvable (391 ± 36 ms for rat and 519 ± 91 ms for trout; also, see Fig. 2).
The next series of experiments used the SL L-type Ca2+ channel agonist (−)BAY K 8644 (100 nM) and SR inhibition with Ry (10 μM) and Tg (2 μM) to investigate the relative roles of SL and SR Ca2+ flux in the spatial and temporal properties of the Ca2+ transients in trout myocytes. Representative line scan images and time course ratios are given in Fig. 3, and mean results are summarized in Fig. 4. Under all treatments, the peak amplitude of the Ca2+ transient is greater at the cell periphery than at the cell center, supporting the prominent role of SL Ca2+ influx in trout E-C coupling. Stimulating SL Ca2+ flux with (−)BAY K 8644 caused an increase in Ca2+ transient amplitude at the cell periphery but not at the cell center. This may result from both the direct effect of (−)BAY K 8644 on the open probability of the L-type Ca2+ channel and on increased SR Ca2+ release at peripheral coupling sites. Fast extrusion via the Na+/Ca2+ exchanger may account for the lack of a significant increase in Ca2+ in the center of the myocyte. Inhibiting SR function decreased the amplitude of the Ca2+ transient in both regions of the cell (Figs. 3 and 4), indicating that SR Ca2+ release contributes to the Ca2+ transient throughout the entire cell volume.
The time to the peak of the transient was slowed across the whole cell by (−)BAY K 8644 application, (Figs. 3B and 4B), reflecting the increased amplitude due to increased or prolonged Ca2+ entry. In the cell center, SR inhibition slowed the rate of rise of the Ca2+ transient but affected neither the absolute time to reach peak amplitude (Fig. 4, B and C) nor the delay in the initiation of the rise in the Ca2+ transient in the cell center compared with the cell periphery (not shown). This suggests that there is limited propagation between RyRs and that SR Ca2+ release is more dependent on the diffusion of Ca2+ from cell periphery to cell center. There were no spatial or temporal differences in the decay of the Ca2+ transient across the width of trout ventricular myocytes under control conditions (Fig. 4D). However, SR inhibition significantly slowed transient decay in the cell center, indicating a role for the SR in Ca2+ resequestration during relaxation. Thus the SR appears to play a clear role in the time course and amplitude of the Ca2+ transient in the center of the trout myocyte. The role of SR Ca2+ cycling in the time course of peripheral Ca2+ gradients is less marked, probably due to the large SL Ca2+ flux, resulting from the large surface area-to-volume ratio in these cells.
To further our investigation into the role of the SR during trout E-C coupling, we scanned isolated intact myocytes for spontaneous Ca2+ spark activity. Despite the large number of cells examined (n = 238 cells from n = 83 fish), spontaneous Ca2+ sparklike events were only observed in one trout ventricular myocyte under control conditions (2 mM Ca2+). Acute decreases in the temperature of the perfusion solution from 21 to 15°C had no effect on the sparking activity of trout ventricular myocytes (n = 28 cells; not shown). Low levels of Ry (<1 μM) also had no effect on spark activity (not shown). Increasing Ca2+ concentration of the superfusate from 2 to 6 mM resulted in the detection of five “sparklike events” (from 3 out of a total of 180 cells observed). Similarly, stimulating the cells with (−)BAY K 8644 to increase spark frequency resulted in detection of 10 events (from 6 out of a total of 198 cells observed). An example of these sparklike events is shown in Fig. 5C. In all cases, these sparklike events occurred at the periphery of the cell. In contrast, Ca2+ sparks were readily apparent across the width of rat ventricular myocytes under both control conditions (56 sparks, from 9 out of a total of 11 cells examined; Fig. 5A) and in the presence of agonists (58 sparks from 10 out of a total of 10 cells examined; not shown). Mean data ± SE and statistical analyses for Ca2+ spark characteristics from rat and trout ventricular myocytes are summarized in Table 2. In general, trout ventricular sparklike events were smaller, narrower, and had faster rates of rise and 50% decay than those from rat (Table 2).
There are many possibilities as to why spontaneous Ca2+ sparks were almost completely absent under control conditions in trout myocytes. Critical factors that modulate spark activity include SR luminal Ca2+ content, cytosolic (subspace) Ca2+ concentration, RyR isoform, and the spatial clustering of RyRs (13). The simplest explanation for the rarity of Ca2+ sparks in trout myocytes is that there was no Ca2+ in the SR. However, this seems unlikely as all cells were stimulated at 0.5 Hz for a minimum of 3 min before quiescent imaging, which should result in a stable SR Ca2+ content (38). Furthermore, the changes in the Ca2+ transients after SR inhibition (see Figs. 3C and 4) indicate that not only is there Ca2+ in the SR but that it is released and resequestered during a twitch. However, to confirm that the lack of sparks under control conditions was not the result of a low SR Ca2+ content, Ca2+ transients were recorded in trout ventricular myocytes during caffeine (10 mM) application (Fig. 6). Caffeine application induces large contractions and Ca2+ transients that were 4.5-fold greater than those stimulated by field electrodes. The response to caffeine was abolished in cells that were treated with Ry-Tg (not shown), clearly indicating significant SR Ca2+ stores in the trout myocytes under the conditions of the present study.
The absence of Ca2+ sparks in the presence of significant SR Ca2+ stores may suggest that the conditions in our intact myocytes were not suitable for spark detection. Indeed, we did observe postrest potentiation of Ca2+ transients (not shown), which suggests an active SR that does not release Ca2+ spontaneously at rest. Furthermore, if we increased extracellular Ca2+ to levels of more than 6 mM, spontaneous Ca2+ waves would occur in some cells. Interestingly, the waves were usually localized to discreet regions and rarely propagated throughout the entire length of the cell. This observation suggests lack of communication between RyR clusters on the trout SR and supports the finding of lack of propagative SR Ca2+ release from the cell periphery to center during the transient.
To increase the probability of detecting Ca2+ sparks, we permeabilized the myocytes to directly control the intracellular environment (54). Our aim was to increase intracellular Ca2+ concentration with a high level of intracellular Ca2+ buffering to reveal individual Ca2+ sparks. However, even at a free Ca2+ concentration of >750 nM, no Ca2+ sparks were detected (in 26 cells examined from n = 5 fish). Therefore, we conclude that, although there is Ca2+ in the SR of trout ventricular myocytes, it is not spontaneously released or is released at amplitudes undetectable under the conditions of our study.
The presence of Ca2+ sparks has been linked to the isomer of RyR present in vertebrate striated muscle (32). Therefore, another explanation for the lack of sparks in trout myocytes is that the isoform of the trout RyR does not spontaneously spark. At present, there are no single-channel recordings or sequencing data of trout heart RyR that would enable determination of receptor isoform or subtype. In this study, we used two commercially available monoclonal antibodies to the mammalian RyR (34C and C3-33) and employed SDS-PAGE and Western blot analyses to examine the immunoreactivity of the RyR in trout heart. The results from this investigation are shown in Fig. 7.
A single band was detected for trout ventricle with a similar mobility to that of the RyR in rat and cow preparations using the antibody 34C. In contrast, trout ventricles demonstrated very low immunoreactivity with the antibody C3-33 to the mammalian cardiac isoform (RyR2) (Fig. 7). It is tempting to interpret these data as suggesting that the trout cardiac RyR is immunologically distinct from the cardiac RyR2 isoform, which is known to show spontaneous Ca2+ sparks in mammals. However, signal density was considerably smaller in trout compared with the mammalian species for both antibodies. This may be because the antibodies were designed to target mammalian RyR sequences. Additionally, low reactivity may be because the density of the RyR on the trout SR membrane is approximately one-third that of the rat (44, 45). However, increasing the amount of trout protein loaded onto the gel from 30 to 90 μg only marginally increased signal density (not shown).
We report the first measurements of temporal and spatial properties of cellular Ca2+ gradients in fish cardiac myocytes. There were two principal findings. First, the lack of T tubules in trout ventricular myocytes resulted in spatial inhomogeneities in the temporal characteristics of the Ca2+ transient. Ca2+ rose faster and to greater levels in the periphery of the cell and then diffused rapidly to the cell center. Ca2+ release from the SR contributes to the Ca2+ transient in both the cell periphery and the cell center. Second, despite an active SR, spontaneous Ca2+ sparks were extremely rare in trout ventricular myocytes.
Inhomogeneities in the Ca2+ transient.
Transverse line scan images from trout ventricular myocytes revealed a shallow U-shaped Ca2+ wavefront during contraction (Fig. 2A). The faster rise of Ca2+ in the periphery of the trout myocyte is due to the absence of a T-tubular system. Indeed, the majority of myocytes that lack a T-tubular system (e.g., atrial myocytes, cultured myocytes, Purkinje cells) exhibit temporal and spatial inhomogeneities in the rise of the Ca2+ transient, which is seen as a pronounced V-shaped Ca2+ wavefront with a 20- to 100-ms time delay between the periphery and cell center (see Ref. 7 for a recent review). The small time delay in trout myocytes (∼6 ms) may relate to the fact that not only are the cells narrow in width but they are also narrow in depth. Indeed, a small cross-sectional area (∼41 μm2) was maintained across a range of myocyte shapes such that the maximum distance from cell periphery to cell center was rarely more than 3 μm. We estimate the rate of Ca2+ movement over this short distance to be ∼46 μm/s, which falls into the lower range of Ca2+ movement calculated for detubulated myocytes (53) and Ca2+ waves (20) in mammalian cardiac myocytes. It is interesting to note that, in pacemaker cells from the toad, which have similar dimensions to those of the trout ventricle, no temporal inhomogeneities were observed during the field-stimulated Ca2+ transient (22). However, SR in toad pacemaker cells appears to be more active than SR in trout, and thus rapid propagation of Ca2+ across the narrow distances may appear instantaneous in this cell type, even in the absence of T tubules.
Contribution of SL and SR flux to cellular Ca2+ gradients.
The temporal and spatial inhomogeneities in the Ca2+ transient across the width of the trout myocyte may be attributable to 1) differential contribution from SL and SR Ca2+ sources during excitation or 2) a consequence of spatial separation resulting from ultrastructural organization. Of course, these two possibilities are linked through Ca2+-induced Ca2+ release and through the propagative SR Ca2+ release between RyRs. We used pharmacological methods to try to understand how these factors were related in the trout ventricle. Inhibiting the SR reduced the amplitude of the Ca2+ transient in all regions of the trout myocyte and significantly slowed the rate of rise in the cell center. Thus, despite the narrow dimensions, centrally located SR stores provide faster Ca2+ cycling in the center of the myocyte than that predicted from SL Ca2+ diffusion alone. Indeed, increasing SL Ca2+ influx with (−)BAY K 8644 produced larger transients in the cell periphery where the L-type Ca2+ channels are located but did not result in larger transients in the cell center. Furthermore, increases in SL Ca2+ influx did not appear to induce greater SR Ca2+ release from peripheral couplings or result in greater propagated SR release in the cell center. Similar results have been observed during application of (−)BAY K 8644 (or isoprenaline) in pig ventricular myocytes, which have homogeneously distributed RyRs but several cellular regions devoid of T tubules. In these cells, increased SL Ca2+ influx increased Ca2+ levels at DHPR-RyR couplings in the T tubules but did not affect the amplitude or time course of the Ca2+ transient in areas spatially separated (15). Differences between the amplitude of the free Ca2+ transient at the cell center and periphery may also be influenced by early peripheral Ca2+ efflux via Na+/Ca2+ exchanger and by the progressive binding of Ca2+ to the myofilaments.
In general, our investigation into the relative roles of SL and SR Ca2+ flux during the Ca2+ transient suggests that voltage-dependent Ca2+ entry at the periphery causes SR Ca2+ release from peripheral couplings, which increases transient amplitude. Peripheral Ca2+ then diffuses to the cell center where it initiates the subsequent release of more centrally located SR Ca2+ stores. Thus trout myocytes show temporal and spatial properties of Ca2+ gradients midway between those of a mammalian atrial cell (devoid of T tubules but robust SR resulting in propagative SR Ca2+ release in the cell center) and a mammalian neonatal myocyte (devoid of T tubules and a poorly developed SR, with Ca2+ diffusing from the periphery to the center with little SR Ca2+ recruitment).
Ca2+ sparks are rare in trout ventricular myocytes.
Ca2+ sparks have been reported with variable frequency from a number of species and in a number of cell types, including cardiac (10, 23, 52), skeletal (47), and smooth muscle (21). Critical factors that modulate spark activity among different animals and cell types include SR luminal Ca2+ content, cytosolic (subspace) Ca2+ concentration, RyR isoforms, and the spatial clustering of RyRs (13). We show spontaneous Ca2+ sparks are extremely rare in ventricular myocytes from the rainbow trout. We attempted to modulate spark activity in trout myocytes with a number of experimental interventions. To increase SR Ca2+ content, we field-stimulated trout myocytes in physiological saline containing 6 mM Ca2+ and then scanned for spontaneous release in quiescent cells. Under these conditions, we did observe a few sparklike events in trout ventricular myocytes (15 events in 378 cells examined). Treatment with (−)BAY K 8644 increases spark frequency in mammalian cardiac myocytes (24), and the increased frequency of Ca2+ sparks in rat myocytes after treatment with (−)BAY K 8644 in the present study is in line with previously published effects. When (−)BAY K 8644 was applied to the trout, the detection of a few sparklike events suggests evidence for sparking ability, albeit under nonphysiological conditions. The characteristics of Ca2+ sparklike events in trout myocytes and Ca2+ sparks observed in rat myocytes from our study fall into the range of those reported in the literature (see Table 2 and Refs. 10, 21, 22, 28, 50, 51). The kinetics and spatial positioning of the trout sparklike events correlate well with those observed in sinoatrial node cells from the cane toad (22).
The rarity of Ca2+ sparks in trout ventricular myocytes is surprising considering the sizeable SR Ca2+ content (Fig. 6 and see Refs. 18, 38). Indeed, SR Ca2+ stores are releasable by caffeine (Fig. 6), suggesting that they should be releasable by Ca2+. However, there is very little known about the cardiac RyR in fish heart. The limited physiological data available do reveal striking functional differences between trout and mammalian cardiac RyRs. Trout RyRs appear less sensitive to cold temperatures than those of mammals (42), allowing maintained SR function over a wide temperature range (18, 38). This may explain why lowering of temperature from 20 to 15°C had no effect on spontaneous Ca2+ spark activity in trout myocytes in the present study. Species differences in sensitivity to SR luminal Ca2+ are also evident from physiological studies. Mammalian RyRs begin to open spontaneously and generate Ca2+ sparks and Ca2+ waves under conditions of SR Ca2+ overload (i.e., >150 μM Ca2+) (3, 43). In contrast, SR Ca2+ content can be very high in trout myocytes (>2,000 μM total Ca2+) without spontaneous release (18, 38). The lack of Ca2+ sparks during exposure of permeabilized myocytes to high Ca2+ in the present study may suggest that trout RyRs are also less sensitive to cytosolic Ca2+ than mammalian RyRs. The Ca2+ dependence of radioactive Ry binding was less in trout heart than in rat, suggesting a less Ca2+-sensitive RyR in trout (M. Vornanen, unpublished observation). Similar observations have been made regarding the Ca2+ sensitivity of RyRs from neonatal mammalian myocytes (14).
The absence of spontaneous sparks and the difficulty in evoking Ca2+ sparks in trout ventricular myocytes may arise from the spatial coupling between RyRs on the SR membrane. In mammalian myocardium, Ca2+ sparks arise from the opening of a number of receptors (∼20) clustered together that act in concert to produce a spark (4). At present, there is no information on the size of RyR clusters in trout myocardium. Radioligand binding studies indicate the ratio of RyRs to DHPRs in trout ventricle is about one compared with approximately four in the rat ventricle (45), which suggests smaller clusters. Furthermore, immunologic studies with carp heart show small RyR clusters that are distributed in a punctate manner throughout the myocytes, which is in contrast to the larger and more regularly organized RyR clusters observed in rat myocytes in the same study (11). Ca2+ waves are thought to propagate due to recruitment of neighboring RyR clusters (10). In the present study, Ca2+ waves could sometimes be observed in intact trout myocytes during superfusion with high (>6 mM) Ca2+ solutions, but they rarely propagated throughout the length of the myocyte. These observations suggest a lack of continuity in the spatial distribution of RyR clusters in the SR membrane of trout. Clearly, further studies are necessary to clarify the impact of RyR cluster size and spatial separation on cellular Ca2+ gradients in trout heart.
Immunoreactivity of the trout RyR.
The occurrence of Ca2+ sparks has been correlated with RyR isoform subtype in a variety of muscle types. Ca2+ sparks are readily apparent in adult mammalian cardiac muscle (10), frog skeletal muscle (39), and developing mammalian skeletal muscle (40). Ca2+ sparks are absent (41) or extremely scarce (12) in adult mammalian skeletal muscle. The mammalian cardiac isoform (RyR2) is known to spark and be responsible for Ca2+-induced Ca2+ release, whereas the mammalian skeletal isoform (RyR1) does not spark and Ca2+ release is not reliant on Ca2+ influx. We show here that the trout cardiac RyR has poor immunoreactivity with the mammalian cardiac muscle isoform (C3-33; RyR2). This may suggest that the trout RyR isoform is distinct from the sparking mammalian cardiac RyR isoform and may thus explain why we were unable to detect Ca2+ sparks in the isolated myocytes. However, the low immunoreactivity may be related to low affinity of the antibody for the trout cardiac RyR. Stronger immunoreactivity was observed between the trout ventricle and the 34C antibody that detects RyR in a number of nonmammalian myoycardium, including frog atrium (46) and avian myocardium (2) but not common carp (Cyprinus carpio) (11). We know of no studies investigating the occurrence of Ca2+ sparks in the myocardium of these vertebrates. Future studies should be directed at sequencing or single-channel studies to uncover RyR isoform subtype in fish hearts. Such information may clarify the lack of Ca2+ sparks in these cells and may also provide insight into the mechanisms of E-C coupling in trout.
The present study represents the first investigation into spatial and temporal properties of Ca2+ transients and Ca2+ sparks in the fish heart. The rainbow trout heart is an interesting model to study cellular Ca2+ flux because myocyte morphology suggests that, similar to frog heart, a large surface area-to-volume ratio allows SL Ca2+ flux alone to provide enough Ca2+ to satisfy the requirements of the myofilaments. However, physiological studies have shown that, under certain conditions, contractility of the trout heart can rely heavily on SR Ca2+ release (1, 17, 36). The results of the present study support the idea of an active SR that contributes both to the amplitude and time course of the Ca2+ transient across the entire width of the cell. Indeed, we show that SR inhibition reduced the peak amplitude of the Ca2+ transient by ∼40% in both the periphery and the cell center. The lack of detectable spontaneous Ca2+ sparks in trout myocytes may result from the RyR isoform subtype, a reduced Ca2+ sensitivity, small clusters of receptors, spatial separation of RyR clusters, or poor coupling between RyRs and DHPRs. Further studies are required to understand the evolution of E-C coupling design in trout hearts and the role of sparking RyRs therein.
We are grateful to the National Science and Engineering Research Council of Canada and the University of Leeds for financial support.
We thank Derek Steele and Zoukang Yang for imaging advice and Glasshouses Trout Farm for help with the fish. We also thank the anonymous reviewers for insightful comments and suggestions.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society