Pacing-induced Heterogeneities in Intracellular Ca²⁺ Signaling, Cardiac Alternans, and Ventricular Arrhythmias in Intact Rat Heart

Development of Synchronous Alternans

Figure 1A shows the region of the LV from which recordings were typically made, a 2D fluorescence intensity image of the left ventricular surface (Figure 1B) and the corresponding line-scan image (Figure 1C) from which high-resolution recordings of intracellular Ca²⁺ concentration were made. The square in Figure 1A indicates a typical recording site on the LV. Figure 1B shows an X-Y confocal frame-scan image of the cardiac surface with myocytes oriented in long bundles. The regions of bright fluorescence indicate the time and position of SR Ca²⁺ release transients during stimulation during the nearly 3 seconds required to construct this image.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Procedure for recording frame-scan (X-Y) and corresponding line-scan (X-t) images on the subepicardial region of intact rat LV. A, Photograph of rat LV indicating a typical recording site. B, X-Y frame-scan of epicardial myocytes 1 to 2 cell layers beneath surface of myocardium showing fluo-4 fluorescence of Ca²⁺ transients; bright corresponds to systole, dark corresponds to diastole. C, X-t line-scan image acquired with the line placed across the 14 cells as indicated by the white arrow on B. Each myocyte in B is indicated in C with a green horizontal line. The blue horizontal line indicates a particularly bright region of a capillary, which provides a landmark for relating line-scan to frame-scan images. The 2 Ca²⁺ transients recorded in C show simultaneous activation of all 14 myocytes in the image (labeled cells 1 to 14) and a typical intensity profile for average fluorescence (above).

To measure Ca²⁺ transients simultaneously in a number of myocytes with high temporal and spatial resolution, the scan line was placed across 14 cardiac myocytes on the left ventricular epicardial surface (white arrow in Figure 1B). The X-t line-scan image to the right of the frame-scan (Figure 1C) shows how cytoplasmic Ca²⁺ transients are measured in individual myocytes. The location of each myocyte is indicated by the green horizontal lines (Figure 1B) during repetitive scanning along the line shown in Figure 1C. The average fluorescence for each scan from all cells is indicated in the intensity profile above and shows typical fluorescence intensity data, with the brightest intensity occurring at peak systole, followed by a fall in intensity during Ca²⁺ removal for the cytoplasm.

The normal behavior of cardiac myocytes at basal pacing rates (at BCL [basic cycle length]=500 ms) is shown in Figure 2A. This line-scan image shows 2 responses at the basal rate followed by a 10-second test train at BCL=260 ms, followed by the first response back at basal pacing. The last basal beat is followed by smaller transients in all myocytes and the ensuing test train demonstrates nominally uniform Ca²⁺ alternans in a large–small–large–small sequence until rapid pacing ends. Nearly all cells show some degree of alternans; however, there is a great degree of heterogeneity in the responses of individual myocytes. Figure 2B shows an expanded view of the last 4 rapid pacing beats and the first after return to basal rate. The variability in steady-state alternans magnitude from cell to cell is apparent in the profiles of 4 cells selected, ranging from nearly complete alternans (cells 2 and 12) to none at all (cell 5). All cells in alternans are in the same phase, demonstrating behavior termed “synchronous alternans.” This form of alternans occurs over a range of BCLs such that the alternans ratio (AR=1−S/L, based on Ca²⁺ transient peak amplitude; where S indicates short and small transient and L, large and long transient) increases with decreasing BCL for each cell (Figure 2C). However, the BCL at which each cell develops alternans is variable, suggesting that the susceptibility to alternans onset differs between myocytes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Cell-to-cell heterogeneity in the development of synchronous whole cell Ca²⁺ alternans in epicardial myocytes in situ. A (top), Line scan image from Figure 1 corresponding to overall mean fluorescence intensity profile showing 2 Ca²⁺ transients at the basal BCL=500 ms, followed by a rapid train at BCL=260 ms (10 seconds) and finally 1 transient on return to BCL=500 ms. B, Temporal and spatial expansion of image A (horizontal teal bar) showing the last 4 beats in the fast pacing epoch and one at basal rate. Mean fluorescence intensity profiles for cells 2, 5, 8, and 12 are shown at right. C, Graph of AR vs pacing BCL for each cell in this site. Data points for each cell are indicated by their assigned numeration and thin lines are their respective fitted sigmoid curves. The bold points, curve, and error bars are the mean±SEM for the group of cells in this site.

The relationship between BCL and AR for each cell was fitted to a sigmoid and the threshold cycle length (effective cycle length at AR=0.20 [ECL₂₀]) and the midpoint (ECL₅₀) were calculated as demonstrated in Figure 3A. To determine which characteristics of Ca²⁺ transients might predict sensitivity of each cell to alternans development, the values for ECL₂₀ and ECL₅₀ were plotted as a function of a number of properties of the Ca²⁺ transients at the basal rate of 500 ms (Figure 3B through 3E). The only reliable predictors of vulnerability to alternans were related to transient duration (Figure 3B) and Ca²⁺ reuptake rate, as indicated by the transient decay time (Figure 3C). In contrast, there was no apparent relationship between vulnerability to alternans onset and rise time (Figure 3D), total Ca²⁺ released (Figure 3E), magnitude of release, or time-to-peak (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Predictors of cellular vulnerability to Ca²⁺ alternans development. A, Graph of AR vs BCL for a cell fitted to a sigmoid curve showing calculation of ECL₂₀ (effective cycle length threshold for alternans onset) and ECL₅₀ (at half maximal alternans development). B, Graphs of ECL₂₀ and ECL₅₀ vs transient duration at 50% recovery (TD₅₀). N=172 myocytes from 11 hearts. C through E, Graphs of ECL₂₀ vs decay time, rise time, and total Ca²⁺ release of the basal Ca²⁺ transient, respectively.

Development of Dyssynchronous and Subcellular Alternans

A different pattern of Ca²⁺ signaling among neighboring myocytes occurs at higher pacing rates. Immediately after acceleration of BCL from 500 to 230 ms (Figure 4A), a number of cells developed alternans that is out of phase with the majority of cells in the field. The left expanded segment below the complete image in Figure 4A (white bar, Figure 4B) shows that the initial responses exhibit disorganized behavior, particularly for cells 3 and 4 as indicated by their respective intensity profiles. The right image (Figure 4C) is a temporal expansion of the last 4 rapid pacing beats and the first after return to the basal rate (teal bar in Figure 4A). Several myocytes are out of phase with others in the immediate vicinity, including cells 2, 5, 8, 10, and 12, as indicated by their respective intensity profiles. This “dyssynchronous alternans” may remain stable for the duration of the test train or may last only for several beats and become synchronous with surrounding myocytes. However, note that myocytes both distant from one another and immediately adjoining may demonstrate dyssynchrony.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4. Development of dyssynchronous Ca²⁺ alternans and subcellular Ca²⁺ alternans during rapid pacing. A, Line-scan image of same experiment showing three Ca²⁺ transients at basal rate (BCL=500 ms), followed by a train at BCL=210 ms, then a return to basal. B, Region of early disorganized behavior extracted (horizontal white bar) from the top image during rapid pacing epoch illustrates only cells 3 and 4 in the image and their respective intensity profiles to the immediate right. C, Temporal expansion (horizontal teal bar) of the last 4 beats of the fast pacing epoch and 1 after return to basal pacing from B. Intensity profiles from cells 2, 5, 8, 10, and 12 are shown at right. D, Expansion of cells 3 and 4 during the early behavior shown in bottom left of B (horizontal teal bar).

Dyssynchrony among myocytes on the LV occurred with an overall incidence of 21% (119/575 myocytes in 31 sites from 18 hearts). These results demonstrate the very high degree of microscopic heterogeneity in Ca²⁺ cycling that occurs in a rate-dependent manner.

The recording in Figure 4B is a temporal and spatial expansion of cells 3 and 4 in the left lower image of Figure 4A (teal bar) and demonstrates a third distinct form of Ca²⁺ signaling that is evident in several cells in this image. Under these conditions, Ca²⁺ release also varies in different regions within both cells 3 and 4 during several cycles of rapid pacing. For example, during the forth cycle of the image shown in Figure 4D, the top parts of cells 3 and 4 show little Ca²⁺ release, whereas the bottom part shows prominent release. The opposite pattern of release then occurs in the fifth cycle. However, the spatial and temporal pattern of heterogeneous intracellular Ca²⁺ release changes throughout rapid pacing (Figure 4A). This intracellular alternans pattern illustrates a phenomenon known as “subcellular alternans.”^11–13 This figure demonstrates that the same conditions that promote dyssynchronous Ca²⁺ release between different myocytes simultaneously promotes heterogeneities in Ca²⁺ release within individual myocytes.

Subcellular Alternans in Intact Epicardium

To investigate subcellular alternans within individual myocytes of intact LV, we recorded line-scan images along the longitudinal axis of individual myocytes in the left ventricular epicardium (Figure 5). This approach allowed high spatial and temporal resolution Ca²⁺ imaging along myocyte length during stimulation. In addition, these hearts were stained with di-8-ANEPPS to outline the cell membranes, so that cell ends could be located during positioning of the line and during analysis. Figure 5A shows a 2D image of a single cell outlined by di-8-ANEPPS staining. The fluorescence intensity was high in this image because systole occurred during recording. The white box shows the cell outline, with the brightest portions appearing as lines along the cell ends (yellow). The scan line was placed from end to end (blue line, Figure 5B) as indicated by the dashed arrows showing the recording site. The top line-scan image in Figure 5C illustrates that this cell responded to a decrease from BCL=500 to 220 ms without development of Ca²⁺ alternans (mean whole cell intensity profile above the image). The middle (Figure 5C) illustrates that decreasing the BCL to 190 ms produced uniform large–small oscillations in successive Ca²⁺ transients, with the result being a whole cell alternans. However, faster pacing (BCL=180 ms, Figure 5D) induced subcellular alternans. In contrast to the lower rate, which showed only transient subcellular alternans (BCL=190 ms, Figure 5C), Ca²⁺ release never become uniform and subcellular alternans occurred throughout rapid pacing. The intensity profile demonstrates little alternans, owing to the fact that Ca²⁺ release involves nearly half of the cell and is out of phase with each successive beat in this example. Subcellular alternans was observed in 24/35 single myocytes (69%) in 4 hearts and was induced over the range of BCL from 220 to 130 ms in different cells.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5. Characteristics of subcellular alternans in individual myocytes of di-8-ANEPPS–stained left ventricular epicardium. A, X-Y frame-scan image of the cell in situ from which recording was made in panel B. Intensity is high (yellow) at cell ends, allowing visualization of cell length for subsequent line scans. Cell dimensions are outlined with a white box. B. Blue line indicates position of scan-line; white arrows, cell boundaries and location of X-Y frame-scan transposition for the line-scan. C (top), Long-axial line-scan image of a single myocyte in the intact LV and its intensity profile showing one Ca²⁺ transient at basal BCL=500 ms, followed by a train of transients at a fast pacing BCL=220 ms, then a return to basal rate. D, Pacing at BCL=190 ms. E, Pacing at BCL=180 ms.

Note that subcellular alternans occurs in the same cellular regions over multiple cycles (Figure 5 and supplementary Video 1). Interestingly, both dyssynchronous whole cell and subcellular Ca²⁺ alternans can occur simultaneously in microscopic regions of LV (supplemental Video 2).

Alternans Subtypes Are Temperature Independent

It is well known that low temperature increases the likelihood of alternans development.^3,14–16 Consequently, it is important to confirm that these patterns of inter- and intracellular Ca²⁺ alternans are also present at physiological temperature.

Figure 6 shows data from experiments in which temperature was maintained at 35°C following dye loading. Basal pacing rate was maintained at BCL=350 ms because of the higher spontaneous rate at higher temperature. Figure 6A shows that rapid pacing (BCL=200 ms) induced synchronous alternans among the different cells in the optical field. The 3 cells indicated at the right (cells 1, 4, and 6) demonstrate ARs varying from nearly 0 to approximately 0.90. Figure 6D summarizes data from both temperatures (data for 23°C are replotted from Figure 3B) and demonstrates the close relationship between vulnerability to alternans onset and transient duration at both temperatures. This relationship is shifted to shorter transient durations, reflecting the fact that durations are indeed shorter at higher temperature. However, the increase in the steepness of the relationship suggests an increased sensitivity of alternans development to increasing rate at physiological temperature.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6. Alternans subtypes at physiological temperature (35°C). A, BCL was reduced from basal (350 ms) to 200 ms during the test train. Intensity profiles during transverse scanning at right show fluorescence changes during last 4 beats at BCL=500 ms and the first spontaneous beat following the train in 3 representative myocytes. B, Pacing at BCL=140 ms in same site as A. C, Longitudinal recording of a single myocyte in the LV shows basal pacing at BCL=350 ms and pacing at BCL=130 ms for 10sec. D, Summary graph of ECL₂₀ vs transient duration at 50% recovery (TD₅₀) from experiments performed at room temperature (same data as in Figure 3; open circles) and at 35°C (filled circles; n=90 myocytes from 7 sites in 3 hearts). Asterisks denote the first 12 (A and B) and 6 (C) stimuli of test train.

Figure 6B shows an example from the same heart in which cells in the optical field followed the stimulation at 140 ms for approximately 10 beats, after which a 2:1 block occurred that was characterized by dyssynchronous alternans. This is shown clearly in the intensity profiles for cells 16 and 17 (right). Approximately 26% of cells showed dyssynchronous alternans (23/90 myocytes in 6/7 sites in 3 hearts), an incidence rate that was close to that found at room temperature.

Figure 6C shows a longitudinal recording in a myocyte from a separate heart in which normal uniform activation of Ca²⁺ transients was evoked at 350 ms basal BCL (left line scan image). However, when BCL was reduced to 130 ms, subcellular alternans occurred in a manner similar to that found at room temperature. Overall, 9/14 (≈64%) myocytes demonstrated subcellular alternans at 35°C, indicating a similar incidence rate to that observed during recording at the lower temperature.

Overall, there is a shift in their rate dependence of all 3 forms of alternans to shorter cycle lengths because Ca²⁺ cycling is highly sensitive to temperature, reflecting the shorter transient durations at 35°C. One result is an increase in rate sensitivity of synchronous alternans development with increasing rate at 35°C, so that more cells developed alternans with smaller changes in rate than at room temperature. Thus, hypothermia may have important influences on the exact stimulation rates at which the different alternans subtypes develop but is not in itself responsible for either incidence rates or the types of alternans that develop during rapid pacing. This observation poses a possible explanation for why alternans develops at slower pacing rates at lower temperatures and why most investigators choose to study alternans at lower temperatures.

Heterogeneity in Intracellular Ca²⁺ Signaling and Arrhythmias

Finally, we recorded Ca²⁺ transients in microscopic regions of LV in 12 rat hearts in which ventricular tachycardia (VT) could be induced by rapid pacing. The results of one experiment are shown in Figure 7A. The ECG recording at the top of the figure shows the last 8 beats of the induced VT caused by pacing at a BCL=120 ms, which was followed by a series of 7 beats of spontaneous VT. A fairly stable pattern of both dyssynchronous and subcellular alternans was present throughout the end of the test train and during the induced episode of VT. Subcellular alternans is apparent in the image of a single myocyte expanded below and is particularly noticeable in the intensity profiles of 2 selected regions of that single cell expanded at the bottom of Figure 7A.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7. Subcellular Ca²⁺ alternans and dyssynchronous Ca²⁺ alternans during VT. A, VT was induced by fast pacing (BCL=120 ms; white arrows above ECG) and persisted after pacing was stopped. Recordings show the ECG, short-axis line-scan image, spatial expansion of 1 myocyte from the line-scan, and intensity profiles of 2 subcellular regions from that cell. B, Long-axis line-scan images of 2 myocytes (top) and that of a single myocyte (bottom) with corresponding intensity profiles (white-on-black, below each image) and ECG recordings (black-on-white, above each image) during spontaneous VT.

The characteristics of subcellular alternans during VT were investigated further in Figure 7B, which shows examples of subcellular alternans measured in stable spontaneous VT. The top image illustrates the variability in the spatial and temporal patterns of subcellular alternans in 2 complete cells and part of a third cell, whereas the bottom image illustrates the behavior within a single cell. Of 12 hearts in which VT was either induced or occurred spontaneously, 11 hearts demonstrated dyssynchrony within the site from which recordings were made. Longitudinal single-cell recordings were made in 9 of these hearts, and all demonstrated subcellular alternans in at least 1 cell in the visual field during VT. Typical examples of VT in multicellular sites of both rat and guinea pig heart (supplemental Videos 3 and 4, respectively) demonstrate simultaneous dyssynchrony and subcellular alternans during spontaneous VT. These movies demonstrate the highly complex Ca²⁺ dynamics that occur during arrhythmogenesis both among and within myocytes in microscopic regions of LV. At the present time, there are insufficient data to prove that the development of VT is mechanistically related to alternans under these experimental conditions. However, the occurrence of both dyssynchronous and subcellular alternans in nearly every heart that developed VT raises the possibility that these forms of heterogeneity may contribute to pacing-induced VT in this experimental model.

Mechanism of Synchronous Alternans

There has been a great deal of study recently about how oscillations in SR Ca²⁺ release can produce Ca²⁺ alternans in isolated myocytes.^23–25 One modeling study in particular⁷ provided strong theoretical support for a major role in alternans for the relationship between the diastolic SR Ca²⁺ load and its influence on Ca²⁺ release. A separate contribution to alternans development and magnitude might also arise from the degree of Ca²⁺-induced inactivation of the L-type Ca²⁺ current, which alternates with the level of cytoplasmic Ca²⁺ during Ca²⁺ alternans. However, these studies have largely ignored the role of Ca²⁺ transient duration in regulating the availability of Ca²⁺ for subsequent release with each cycle. Thus, a large and long (L) transient will be interrupted before SR Ca²⁺ content can recover at fast pacing rates, whereas small and short (S) transients will allow more complete recovery for the next cycle, which will then be large. One implication of this observation is that because Ca²⁺ transients in human heart failure^26,27 and in animal models^28–30 are prolonged, this effect could potentially increase the vulnerability to alternans onset.

Our results provide the first direct evidence that different cells in intact heart have different vulnerabilities to alternans development based on intrinsic properties of Ca²⁺ signaling within individual cells. These results are similar to recent observations in both dog³¹ and guinea pig³² and are consistent with the idea that the rate of Ca²⁺ cycling and, specifically, the duration of the Ca²⁺ transient might determine both the magnitude of and rate sensitivity to alternans. Moreover, this was the first demonstration of the variability in Ca²⁺ transients that exists at the microscopic level; previously, it has always been largely assumed that Ca²⁺ transients are uniform within each left ventricular region (eg, base versus apex) despite the fact that important differences exist between regions.^1,3,9,10 Our observations demonstrate that this is not the case, and, furthermore, there is a high degree of heterogeneity in Ca²⁺ dynamics within microscopic regions of the LV at both low and physiological temperatures.

It is not yet clear how dyssynchrony develops, but the result is that neighboring groups of cells and, in many cases, immediately adjoining cells are capable of developing alternans that is out of phase. The result is a much greater cell-to-cell variability in Ca²⁺ signaling at the microscopic level than was previously recognized. It is not yet known how or even if these microscopic heterogeneities in Ca²⁺ signaling might contribute to regionally discordant electrical alternans. However, dyssynchrony could contribute to the substrate for reentrant arrhythmias, if its influence in the LV is sufficient in magnitude and distribution to contribute to a dispersion in refractoriness across the LV.

Subcellular Alternans, Cellular Ca²⁺ Heterogeneities, and Arrhythmias

It has only recently been shown that intracellular SR Ca²⁺ release can also go into oscillation in localized regions within individual myocytes.^11,33 Part of the cell releases a large amount of Ca²⁺, whereas another cellular region released far less; the next activation produced the opposite pattern, causing a regional alternans within atrial cells but rarely in ventricular myocytes.¹² A nearly identical phenomenon occurred during partial inhibition of ryanodine receptors (RyRs) in rat ventricular myocytes using low concentrations of tetracaine or acidosis.¹³

The mechanism for subcellular alternans is most probably similar to that underlying Ca²⁺ alternans in general. One cellular region with slow Ca²⁺ reuptake is unable to respond to 2 successive beats during rapid pacing and so requires an additional cycle to replenish stores for large releases, whereas another region, the transient of which can recover more quickly, is able to follow each stimulus. We have also recorded instances in which different cellular regions simply oscillate completely out of phase with one another. The result is regional L/S alternations that are out of phase at different sites within the cell on a beat-to-beat basis.

When we studied the incidence of rate-dependent subcellular alternans development, we found that this is not an unusual phenomenon in the intact heart. In fact, nearly 2/3 of myocytes in intact rat LV demonstrated subcellular alternans at rates that produced pacing-induced VT and again during episodes of spontaneous VT. These observations raise, for the first time, the possibility that nonuniform intracellular Ca²⁺ release occurs during rapid pacing and arrhythmias.

Finally, the presence of both dyssynchronous and subcellular alternans in nearly every heart that developed VT raises the possibility that alternans is involved in pacing-induced VT in this model. Conversely, if we had found that these forms of heterogeneity in Ca²⁺ signaling were absent during VT, it would be highly unlikely that they play any role in the arrhythmia. To our knowledge, these are the first direct measurements of Ca²⁺ signaling in single cells in intact heart during arrhythmias, and the results demonstrate an extraordinary heterogeneity in behavior at the level of individual myocytes during arrhythmias. Given that the electrical space constant far exceeds the behavior of individual myocytes, pacing-induced heterogeneities in ion channel activation and action potential in individual cells are unlikely to affect macroscopic electrical activity of the ventricular syncytium. The smoothing effect of electrical current spread across many cells would, if anything, filter the most drastic cell-to-cell differences in Ca²⁺ signaling and resulting discrepancies in ion channel and action potential activation, thus minimizing their influence in arrhythmia development. However, a reduced space constant such as occurs in disease states, where intercellular communication is reduced,^34,35 would minimize protection against nonuniformities in refractoriness, possibly promoting the development of re-entrant arrhythmias.

In the context of the relationship between transmembrane potential (V_m) and Ca²⁺ signaling heterogeneities at the cellular level, it is important to note that there are several reports of subcellular changes in V_m that might be relevant to the intracellular heterogeneities in Ca²⁺ signaling observed here. Several studies have found that cardiac cells do not depolarize uniformly during field stimulation, suggesting that a cellular “action potential” is subject to local changes in ion channel activation kinetics.^36–38 The fact that we found Ca²⁺ signaling heterogeneities within individual myocytes of intact heart raises the possibility that different cellular regions might demonstrate local action potential variability depending on activation and recovery kinetics of Ca²⁺ sensitive conductances (I_NCX, I_ClCa, among others). A role for this type of nonuniform electrical activation has not been identified in arrhythmogenesis as yet. Furthermore, it is not known if the heterogeneities in cellular V_m are also present in intact heart because these studies were performed in isolated myocytes. However, it is known that current flow from cell to cell depends on the uniformity of impedances during downstream conduction. Depolarizing ionic current as well as changes in resistance and capacitance between—and possibly within—myocytes can affect the spread and rise of depolarization.³⁹ Thus it is possible that heterogeneities in cellular Ca²⁺ could also affect the rise time and shape of the action potential and the uniformity of propagation both between neighboring myocytes as well as within individual cells.

Lastly, it is worth mentioning that the observations of highly heterogeneous behavior in intact heart leading potentially to subcellular alternans have been described in theoretical studies published in recent years.⁴⁰ This intracellular behavior is mathematically quite similar to that described for spatially discordant APD alternans in theoretical models⁴¹ across larger tissue regions. Alternatively, the behavior underlying subcellular alternans and possibly dyssynchronous alternans might best be described in terms of Turing-type instability within individual cells and between myocytes in small cardiac regions, respectively.⁴² It will be important in the future to analyze the cellular behavior observed here in terms of these theoretical considerations which could provide important insights into the basis for arrhythmia initiation and stabilization in the whole organ.

Limitations of This Study

One of the limitations of this study is that the arrhythmias were generated in hearts that were hypothermic. Low temperature affects ion channel behavior, intracellular Ca²⁺ cycling, and other factors that might contribute to arrhythmogenesis at nonphysiological temperatures. Ion channel activation is affected not only by a direct effect of cooling on kinetics but also reduced membrane fluidity, which alters channel environment in lipid membranes as well as membrane transport systems and therefore local transmembrane ionic balances. Conduction velocity is slowed because of reduced rapid Na⁺ current and gap junctional conductance. These factors will all contribute to arrhythmias at room temperature, making a definitive relationship between altered Ca²⁺ signaling and arrhythmias speculative at best.

One of the potential complicating factors in our interpretation of results is that cytochalasin D is required to prevent contraction. This agent is known to have direct effects on transients in addition to blocking contraction, raising the possibility that some of our observations about Ca²⁺ dynamics made in the presence of cytochalasin D might be influenced by its presence rather than occurring as a result of intrinsic properties of Ca²⁺ cycling in the heart. Cytochalasin D slows rise time and prolongs Ca²⁺ transient duration in rat ventricular myocytes⁴³ but has little effect on dog heart (up to 80 μmol/L),⁴⁴.⁴⁵ Finally, the ability of cytochalasin D to slow Ca²⁺ transient kinetics was associated with a tendency to reduce the incidence and duration of reentrant ventricular arrhythmias in mouse heart.⁴⁶ Overall, these data suggest that cytochalasin D is not affecting excitation/contraction coupling to a sufficient extent to explain the heterogeneities in Ca²⁺ signaling observed here.