Partial Inhibition of Ca2+ Current by Methoxyverapamil (D600) Reveals Spatial Nonuniformities in [Ca2+]i During Excitation-Contraction Coupling in Cardiac Myocytes
Abstract The laser scanning confocal microscope was used in conjunction with the Ca2+ indicator fluo 3 to examine the spatiotemporal properties of free Ca2+ ([Ca2+]i) transients in isolated rat cardiac myocytes. We show that localized increases in [Ca2+]i (Ca2+ sparks) can be triggered by membrane depolarization in cardiac myocytes when the sarcolemmal Ca2+ current amplitude is reduced by methoxyverapamil (D600). These depolarization-evoked Ca2+ sparks are similar in amplitude and spatiotemporal properties to spontaneous Ca2+ sparks previously observed at rest. These observations support the idea that Ca2+ sparks are the result of the activation of functional elementary units of sarcoplasmic reticulum (SR) Ca2+ release. The synchronous activation of a large number of Ca2+ sparks can explain the increased amplitude and slower time course of the electrically evoked [Ca2+]i transient as well as the presence of spatial nonuniformities in [Ca2+]i during its rise. The data shown here suggest a model for excitation-contraction coupling in which the amplitude of the [Ca2+]i transient is regulated by variations in the probability of recruitment of elementary SR Ca2+ release units as well as the amount of Ca2+ released by each unit. Since the activation of each release unit will depend on the local amplitude of the Ca2+ current, this model can explain the regulation of the amplitude of the [Ca2+]i transient by the Ca2+ current. In addition, these data indicate that caution should be applied to the interpretation of signals obtained with nonlinear Ca2+ indicators during the rising phase of the [Ca2+]i transient, when the nonuniformities in [Ca2+]i are largest.
Recently, Cheng et al1 reported a new phenomenon in [Ca2+]i metabolism in rat heart muscle: the occurrence of spontaneous local increases in [Ca2+]i of ≈150 nmol/L, which last ≈40 ms in quiescent healthy ventricular cells. These “Ca2+ sparks” arise from the spontaneous gating of sarcoplasmic reticulum (SR) Ca2+ release channels (ryanodine receptors [RyRs]). The amplitude and the unitary behavior of these Ca2+ sparks indicate that they could be due to the activation of minimal functional SR Ca2+ release units. Depolarization-evoked [Ca2+]i transients arise from excitation-contraction (EC) coupling,2 which is triggered by an influx of Ca2+ across the sarcolemma (principally via L-type Ca2+ channels,3 4 also called dihydropyridine receptors [DHPRs]). This influx of Ca2+ is amplified by SR Ca2+ release, which occurs as a result of the Ca2+-induced Ca2+ release (CICR) mechanism,5 to provide sufficient Ca2+ to activate contraction. Thus, despite differences in appearance, both [Ca2+]i transients and Ca2+ sparks arise from the activation of RyRs in the SR. Thus the [Ca2+]i transient could be due to the synchronous activation of a large number of Ca2+ sparks.
However, direct visualization of such Ca2+ sparks during EC coupling is complicated by the loss of contrast in confocal [Ca2+]i images that occurs as a result of the cell-wide increase in [Ca2+]i. A normal EC coupling [Ca2+]i transient is associated with a Ca2+ flux ≈104 greater than that of a typical Ca2+ spark; thus, the detection of a single Ca2+ spark on such a high background is difficult. In the present study, the new strategy adopted was to reduce the amplitude of the triggering Ca2+ current (ICa) (with methoxyverapamil [D600]) to reduce the amplitude of the evoked [Ca2+]i transient. In these conditions, we studied the spatial and temporal properties of the evoked [Ca2+]i transient and show that the local changes in [Ca2+]i that occur are very similar to spontaneous Ca2+ sparks. We conclude that Ca2+ sparks can be evoked by depolarization and suggest that the normal [Ca2+]i transient is the result of the summation of a large number of Ca2+ sparks resulting from the activation of functional SR Ca2+ release units by the local increase in [Ca2+]i arising from Ca2+ channel activation.
Materials and Methods
Single rat cardiac myocytes were dissociated by enzymatic treatment as described elsewhere.1 6 Aliquots of cells were exposed to 5 μmol/L fluo 3-AM7 (diluted from a stock containing 50 μg fluo 3-AM and 25 μg Pluronic [Molecular Probes]) in 100 μL dimethyl sulfoxide) for 5 minutes, followed by a 30-minute wash in extracellular solution to allow time for deesterification.
Cells were electrically stimulated with 2-ms voltage pulses delivered through parallel platinum wires. The stimulation voltage was set to 1.5 times threshold. The stimulation pulse was triggered during the confocal scan at a selectable line and pixel by custom electronics designed by the authors.
Confocal Imaging of [Ca2+]i
To examine the spatial properties of the [Ca2+]i transient at high resolution, a cell loaded with fluo 3 was imaged along a selected line in the confocal plane at 500 Hz with a Nikon Diaphot microscope coupled to a Bio-Rad MRC600 confocal imaging head. The confocal line was oriented to pass through the thickest section of the cell, either longitudinally or transversely while avoiding the nuclei. However, the orientation of the line did not appear to affect the pattern of spatial [Ca2+]i reported here. Consecutive images of the fluorescence intensity along the selected line were placed below each other to produce the line-scan images shown in Figs 1 through 4⇓⇓⇓⇓, in which the horizontal dimension is the distance along the scanned line and the vertical dimension is time (increasing from top to bottom). Each pixel in the image represents a “voxel” inside the cell defined by the point-spread function of the imaging system.8 In our modified Bio-Rad 600 confocal microscope system, the lateral resolution is just below 0.4 μm, and the axial resolution is ≈0.8 μm (both resolutions were defined by the full width at half maximum [FWHM] fluorescence intensity of images of 0.1-μm fluorescent beads).
Fluorescence signals were normalized by dividing them by the average fluorescence intensity at rest. This procedure removes the contribution of possible nonuniformities in dye concentration within the confocal plane and gives a signal that can be directly related to the [Ca2+]i.1 This procedure also allowed direct comparison between different experiments but did not materially affect the appearance of the nonuniformities in [Ca2+]i reported here.
The standard bathing solution contained (mmol/L) NaCl 137, KCl 5.4, MgCl2 1.2, CaCl2 1, and HEPES 20 (pH 7.4 at 25°C). D600 (Sigma Chemical Co), CdCl2, 2,3-butanedione monoxime (BDM, Sigma), and ryanodine (S.B. Pennick & Co) were added to the standard bathing solution as needed. Control experiments showed that BDM had no effect on the fluorescence of fluo 3.
Fig 1A⇑ shows a line-scan image of a cell under control conditions. Two hundred fifty-six milliseconds after starting the scan, the cell was electrically stimulated. After a short delay (6 ms), fluorescence rapidly increased across the entire region examined. This increase in fluorescence reached its peak in 32 ms and then declined with a half-time of 152 ms (Fig 1B⇑, top tracing). The movement of the ends of the cell started 30 ms after the stimulation, and maximal shortening of the cell was 8.7% (Fig 1B⇑, lower tracing). Although, superficially, the fluorescence transient (and hence [Ca2+]i) appeared to be continuous and uniform, closer examination of the image reveals a nonuniform structure in the early phase of [Ca2+]i transient (between 0 and 28 ms after stimulation) (Fig 1C⇑). This impression is confirmed by the plot of spatial [Ca2+]i profile at the time of initiation of the Ca2+ transient (Fig 1D⇑).
Since there was a delay before cell shortening occurred, the normalization procedure (see “Materials and Methods”) should have removed any contribution to the signal arising from differences in the dye distribution across the cell. Furthermore, the observed delay between the increase in [Ca2+]i and cell shortening excludes a possible contribution from movement artifacts. Therefore, we conclude that [Ca2+]i is inhomogeneous during the rising phase of the depolarization-evoked [Ca2+]i transient.
Although these nonuniformities during EC coupling may be explained by the activation of a large number of discrete Ca2+ sparks (as hypothesized by Cheng et al1 ), to resolve their genesis more clearly the spatial contrast of the signal requires enhancement. Depolarization-evoked Ca2+ release from the SR is completely graded by the amplitude of the L-type ICa9 10 11 ; thus, a reduction in the ICa amplitude (by reducing the number of available Ca2+ channels) could reduce the amplitude of the [Ca2+]i transient by reducing the probability of Ca2+ spark activation. D600, a use-dependent Ca2+ channel antagonist,12 was chosen for this purpose, since it would progressively reduce ICa over several beats, allowing examination of the block development. Fig 2A⇑ shows four images taken during alternate stimuli at 0.5 Hz in the presence of 1 μmol/L D600. Beneath each image, the corresponding spatially averaged [Ca2+]i transient is plotted. The size of the [Ca2+]i transient gradually decreased with successive stimuli (from rest), and within five beats, the peak normalized fluorescence reached a steady state level that was about one sixth of the control level shown in Fig 1⇑. It was notable that in the presence of D600, there was a dramatic decrease in the spatial uniformity of the [Ca2+]i transient and that this spatial nonuniformity in [Ca2+]i increased during the negative staircase arising from the use dependence of the action of D600. Thus, by the time that steady state was reached (stimulus 5 onward), only a few discrete sites responded to the stimuli. Although the first stimulus resulted in a small contraction, the cell remained mechanically quiescent during subsequent stimuli; thus, the large nonuniformity in [Ca2+]i observed in these conditions should not contain motion artifacts. The latter point was confirmed by the addition of 10 mmol/L BDM to the bathing solution, which completely abolished all contractions in the presence of D600 but did not alter the appearance of the spatial nonuniformity in [Ca2+]i (data not shown).
Visually, the isolated Ca2+ release sites activated by cellular depolarization are very similar to the spontaneous Ca2+ sparks reported by Cheng et al.1 However, it should be noted that although the spontaneous Ca2+ sparks can occur without ICa,1 the depolarization-activated Ca2+ sparks depended on the availability of L-type ICa (see below). A more quantitative analysis of the depolarization-evoked release at these sites is shown in Fig 3⇑. The normalized fluorescence in regions where Ca2+ release occurred (indicated by thin bars in Fig 3A⇑) reached a peak of ≈2.0. In these regions, the decline of fluorescence exhibited two phases (Fig 3B⇑, tracing a): a fast phase with a half-time (t) of 20 ms followed by a slower phase with t of 188 ms. In other regions (marked by the thicker bars in Fig 3A⇑), the time to peak of the [Ca2+]i transient was delayed slightly (Fig 3B⇑, tracing b), presumably as a result of the time taken for Ca2+ to diffuse from other (unresolved) release sites outside the confocal plane. In these regions, only one phase in the decay of fluorescence, which had a t of 198 ms, was evident. At later times, the tracings from both regions overlapped, showing that the [Ca2+]i transient was more uniform during the reuptake phase of the [Ca2+]i transient. The signal from sites that released Ca2+ will contain a component that is due to the diffusion of Ca2+ from adjacent release sites (both inside and outside the confocal plane). To correct for this component, the background fluorescence change (δb) between sites of Ca2+ release was subtracted from the fluorescence change at sites of release. After this first-order correction, the local [Ca2+]i transient had a time to peak of 14 ms and a t of decline of 20 ms (Fig 3B⇑, tracing a-δb). The amplitude of the peak fluorescence change, the spatial properties, and the kinetics of these evoked sparks are very similar to those of spontaneous Ca2+ sparks (spontaneous Ca2+ sparks had a peak normalized fluorescence of 1.9, a time to peak of 10 ms, t of decay of 23 ms, and FWHM of 2.0 μm).
When 1 to 10 μmol/L ryanodine was added in these conditions, no evoked Ca2+ sparks were observed (data not shown), indicating that the evoked Ca2+ sparks were due to Ca2+ release from the SR. In addition, evoked Ca2+ sparks were blocked by 0.1 mmol/L Cd2+, showing that the evoked Ca2+ sparks were the result of SR Ca2+ release triggered by ICa (data not shown), probably as a result of the normal CICR mechanism.
It was notable that these evoked Ca2+ sparks changed position from stimulus to stimulus. To clarify this point, eight consecutive line-scan images at steady state are shown in Fig 4⇑ where the rising phase of the [Ca2+]i transients are aligned and displayed vertically. It is clear that a sparking region in one beat could become a nonsparking region in another beat and vice versa. This stochastic variation in the location of SR Ca2+ release could be due to the stochastic variation in the probability of activation of either sarcolemmal Ca2+ channels or SR Ca2+ release channels or both. In addition, the fact that release in one site does not propagate supports the idea that the functional units of SR Ca2+ release cannot normally activate CICR in adjacent regions.1 13
Depolarization-Evoked Ca2+ Sparks in Cardiac Myocytes
After adding the use-dependent Ca2+ channel blocker D600, the depolarization-evoked (spatially averaged) [Ca2+]i transient decreased in amplitude, and this decrease in amplitude was accompanied by a decrease in the spatial uniformity in [Ca2+]i. Eventually, EC coupling takes the form of multiple local increases in [Ca2+]i (evoked Ca2+ sparks). These Ca2+ sparks cannot be explained by cell movement and are too large to be explained by random noise in the detection system. Furthermore, the spatial position of some of the evoked Ca2+ sparks varies from beat to beat. Later, during relaxation, nonuniformities in the [Ca2+]i transient largely disappear. Therefore, it is unlikely that these evoked Ca2+ sparks can be ascribed to artifacts associated with the use of a nonratiometric Ca2+ indicator such as fluo 3.
The evoked Ca2+ sparks are the result of EC coupling for the following reasons: (1) They are synchronously triggered by the electrical stimulation. (2) They require ICa through sarcolemmal Ca2+ channels because they are abolished by 0.1 mmol/L Cd2+. (3) They are not due solely to ICa since they are abolished by ryanodine. The evoked Ca2+ sparks are remarkably similar to the spontaneous Ca2+ sparks recently reported by Cheng et al.1 Both types of Ca2+ sparks are associated with a peak normalized fluorescence of ≈2.0 with a time to peak of ≈10 ms and a t of decline of ≈20 ms. In addition, both spontaneous and evoked Ca2+ sparks are limited to small cell volumes, delimited by a region ≈2 μm in diameter when the peak fluorescence level is reached (10 ms). We conclude that the Ca2+ sparks evoked by depolarization are identical to those seen at rest, despite the fact that spontaneous Ca2+ sparks do not require sarcolemmal Ca2+ influx for their generation.1
Since EC coupling occurs at discrete sites as described above, it is possible that the evoked Ca2+ sparks are the result of the activation of elementary units in EC coupling. Estimates of the Ca2+ flux associated with single Ca2+ sparks as well as pharmacological data suggested that a spontaneous spark may represent the gating of a single RyR or a few RyRs acting in concert.1 To explain the known properties of cardiac EC coupling, it has been suggested that a small group of RyRs may form functional units for EC coupling,14 in which case the Ca2+ sparks observed here could be the result of the activation of such functional units. Within the framework of this model, the resolution of the activation of functional SR release units is explained by a decrease in the probability of SR release as a result of the reduction in sarcolemmal Ca2+ influx by D600. Since the evoked Ca2+ sparks in these conditions are well separated in position, they cannot fuse; thus, most of the evoked Ca2+ sparks represent single, unitary events. The fact that an evoked Ca2+ spark has a size (and therefore the Ca2+ flux) similar to spontaneous Ca2+ sparks further supports the idea that Ca2+ sparks represent fundamental (minimal) units of SR Ca2+ release, both at rest and during EC coupling. Therefore, we have resolved EC coupling to the level of individual discrete sites, and it seems likely that EC coupling under normal conditions is nothing more than the spatiotemporal summation of a large number of these elementary events.
Beat-to-Beat Variation of Evoked Ca2+ Sparks
Although under normal steady state conditions the [Ca2+]i transient due to EC coupling appears to be constant from beat to beat, the gating of channels underlying this process (ie, DHPRs and RyRs) is all or none at the microscopic level. Therefore, EC coupling should be inherently a stochastic process. Since Ca2+ release was limited to a few sites when the L-type ICa was inhibited, we conclude that E-C coupling displays all-or-none behavior at the microscopic level and that variations in spatially averaged [Ca2+]i can result from alterations in the probability of Ca2+ release. It should be noted that a reduction in the amount of Ca2+ available for release at all points in the cell is not compatible with the observed data. Thus, D600 reduces the amplitude of the spatially averaged [Ca2+]i transient by reducing the probability of activation of SR Ca2+ release units within the cell. However, whether the reduction in the probability of SR release unit activation is solely determined by the reduced probability of L-type Ca2+ channel opening (in the presence of D600) is unclear. It is also possible that the consequent reduction in SR Ca2+ content reduces the ability of a given L-type Ca2+ channel to activate a functional release unit (see below).
Since the sites where [Ca2+]i increases after stimulation remain discrete, it is clear that EC coupling does not normally spread throughout the cell.1 15 Furthermore, the underlying structural organization of the cell cannot explain such gross nonuniformity in the [Ca2+]i transient, because [Ca2+]i transients are normally far more uniform (Fig 1⇑). However, although most regions on the scanned line appeared to be capable of generating Ca2+ sparks, some regions seemed less likely, and some more likely, to initiate Ca2+ sparks on depolarization. The simplest explanation for this would be that the efficiency of EC coupling is not spatially homogeneous. Nevertheless, the variation in spark location is explained best by the reduced probability of L-type Ca2+ channel opening in the presence of D600. Whether this is due to the effect of D600 on channel gating or time-dependent variation in the location of blocked channels is unclear. However, both of these possibilities are formally equivalent, since they both result in reduction in the channel open probability. Although further experiments are required to define the contribution from each element (DHPR, RyR, and their coupling processes) to the nonuniform behavior reported here, the observed discrete nature of EC coupling imposes important constraints on models of EC coupling. The experimentally observed, graded nature of the [Ca2+]i transient9 10 11 must therefore arise from changes in the probability of Ca2+ spark activation as well as changes in the SR Ca2+ content. (The latter will alter the Ca2+ release flux associated with the activation of each function unit of EC coupling16 and may also affect the sensitivity of the RyR to [Ca2+]i1 17 ).
Relation Between Ca2+ Sparks and the [Ca2+]i Transient During EC Coupling
A possible objection to the idea that Ca2+ sparks represent fundamental units of EC coupling is the observation that the Ca2+ sparks and the [Ca2+]i transient have very different time courses. However, the difference in time course may simply be the consequence of the number of Ca2+ sparks elicited by the depolarization. For a solitary Ca2+ spark, the situation is formally equivalent to the diffusion of Ca2+ from a point source into a (virtually) infinite absorbing medium. There is a continuous loss of Ca2+ from the release site(s) because of diffusion driven by local [Ca2+]i gradients, and this diffusion from the site of release will accelerate the rate of decline of [Ca2+]i. However, during normal EC coupling, the diffusion from the site of release will be limited by the elevation in [Ca2+]i in neighboring regions (due to the release of Ca2+ in those regions also). This situation is formally equivalent to placing diffusional barriers (across which there is no net flux) between the sites of Ca2+ release. Therefore, when the Ca2+ sparks are close together (eg, during normal EC coupling), the volume into which Ca2+ release occurs is effectively reduced. This will have two effects: (1) The amplitude of the spatially averaged [Ca2+]i transient will increase. (2) The rate of decline of the transient will be reduced because the total number of uptake sites will be reduced (at a constant uptake site concentration, the uptake rate depends on the volume of the region of interest).
Finally, the observation of large nonuniformities in [Ca2+]i when the amplitude of ICa is reduced argues that caution should be applied to the interpretation of spatially averaged [Ca2+]i measurements obtained with nonlinear Ca2+ indicators. For example, in these conditions, fura 2 or indo 1 will underestimate spatially averaged [Ca2+]i.18 This should be less of a problem for a lower affinity indicator such as fluo 3, especially when images are taken with a confocal microscope, which allows direct observation of nonuniformities in [Ca2+]i. Nevertheless, it should be appreciated that the peak of the Ca2+ spark must underestimate the peak [Ca2+]i in a single sarcomere, although it may be possible to construct a computer model to correct for this problem in the future.
In conclusion, we have demonstrated that Ca2+ sparks can be evoked by depolarization and that these evoked Ca2+ sparks have a time course and amplitude similar to spontaneous Ca2+ sparks observed at rest. We were able to clearly resolve these evoked Ca2+ sparks during EC coupling because the addition of a low dose of D600 reduced the probability of their activation. Thus, the well-known modulation of the [Ca2+]i transient by the amplitude of ICa9 10 11 must include alterations in the probability of recruitment of elementary Ca2+ release sites. This local regulation of EC coupling occurs as a result of the microarchitecture of the cell, where the close apposition of the junctional SR membrane to the T-tubular membrane ensures that the activation of a functional SR release unit is primarily determined by the local amplitude of the Ca2+ flux through Ca2+ channels. In addition, alterations in SR Ca2+ content will contribute to the physiological regulation of the [Ca2+]i transient by altering the amount of Ca2+ released by each functional release unit as well as by possibly affecting the probability of release unit activation.
This study was supported by grants from the National Institutes of Health (HL-36974 and HL-26975), DRIF awards from the University of Maryland at Baltimore, the Medical Biotechnology Center, and the British Heart Foundation.
Reprint requests to W.J. Lederer, Department of Physiology and the Medical Biotechnology Center, University of Maryland at Baltimore School of Medicine, 660 W Redwood St, Baltimore, MD 21201.
- Received April 11, 1994.
- Accepted September 29, 1994.
- © 1995 American Heart Association, Inc.
Cheng H, Lederer WJ, Cannell MB. Calcium sparks: elementary events underlying excitation-contraction coupling in heart muscle. Science. 1993;262:740-744.
Callewaert G. Excitation-contraction coupling in mammalian heart cells. Cardiovasc Res. 1992;26:923-932.
Fabiato A. Simulated calcium current can both cause calcium loading in and trigger the calcium release from the sarcoplasmic reticulum of a skinned cardiac cell. J Gen Physiol. 1985;85:291-320.
Fabiato A. Time and calcium dependence of activation and inactivation of calcium-induced calcium release of calcium from the sarcoplasmic reticulum of a skinned cardiac Purkinje cell. J Gen Physiol. 1985;85:247-290.
Mitra R, Morad M. A uniform enzymic method for dissociation of myocytes from hearts and stomachs of vertebrates. Am J Physiol. 1985;246:H1056-H1060.
Minta A, Kao JP, Tsien RY. Fluorescent indicators for cytosolic calcium based on rhodamine and fluorescence chromophores. J Biol Chem. 1989;264:8171-8178.
Wilson T. Confocal Microscopy. London, England: Academic Press Inc; 1990.
Cannell MB, Berlin JR, Lederer WJ. Effect of membrane potential changes on the calcium transient in single rat cardiac muscle cells. Science. 1987;238:1419-1423.
duBell WH, Houser SR. Voltage and beat dependence of Ca2+ transient in feline ventricular myocytes. Am J Physiol. 1989;257:H746-H759.
Barcenas-Ruiz L, Wier WG. Voltage dependence of intracellular Ca2+ transients in guinea pig ventricular myocytes. Circ Res. 1987;61:148-154.
O’Niell SC, Mill JG, Eisner DA. Local activation of contraction in isolated rat ventricular myocytes. Am J Physiol. 1990;258:C1165-C1168.
Thedford SE, Lederer WJ, Valdivia HH. Activation of sarcoplasmic reticulum calcium release channels by intraluminal Ca++. Biophys J. 1994;66:A20. Abstract.