Optical Mapping of Sarcoplasmic Reticulum Ca2+ in the Intact HeartNovelty and Significance
Ryanodine Receptor Refractoriness During Alternans and Fibrillation
Rationale: Sarcoplasmic reticulum (SR) Ca2+ cycling is key to normal excitation–contraction coupling but may also contribute to pathological cardiac alternans and arrhythmia.
Objective: To measure intra-SR free [Ca2+] ([Ca2+]SR) changes in intact hearts during alternans and ventricular fibrillation (VF).
Methods and Results: Simultaneous optical mapping of Vm (with RH237) and [Ca2+]SR (with Fluo-5N AM) was performed in Langendorff-perfused rabbit hearts. Alternans and VF were induced by rapid pacing. SR Ca2+ and action potential duration (APD) alternans occurred in-phase, but SR Ca2+ alternans emerged first as cycle length was progressively reduced (217±10 versus 190±13 ms; P<0.05). Ryanodine receptor (RyR) refractoriness played a key role in the onset of SR Ca2+ alternans, with SR Ca2+ release alternans routinely occurring without changes in diastolic [Ca2+]SR. Sensitizing RyR with caffeine (200 μmol/L) significantly reduced the pacing threshold for both SR Ca2+ and APD alternans (188±15 and 173±12 ms; P<0.05 versus baseline). Caffeine also reduced the magnitude of spatially discordant SR Ca2+ alternans, but not APD alternans, the pacing threshold for discordance, or threshold for VF. During VF, [Ca2+]SR was high, but RyR remained nearly continuously refractory, resulting in minimal SR Ca2+ release throughout VF.
Conclusions: In intact hearts, RyR refractoriness initiates SR Ca2+ release alternans that can be amplified by diastolic [Ca2+]SR alternans and lead to APD alternans. Sensitizing RyR suppresses spatially concordant but not discordant SR Ca2+ and APD alternans. Despite increased [Ca2+]SR during VF, SR Ca2+ release was nearly continuously refractory. This novel method provides insight into SR Ca2+ handling during cardiac alternans and arrhythmia.
In cardiac muscle, Ca2+ release from and reuptake into the sarcoplasmic reticulum (SR) plays a central role in contraction, and tight regulation of Ca2+-induced Ca2+ release is required for proper excitation–contraction coupling (ECC).1 Experimental studies and computational modeling have revealed that SR Ca2+ handling can also contribute to arrhythmogenic behavior. In particular, abnormal intracellular Ca2+ handling has been demonstrated to underlie the development of cardiac alternans,2–8 which is not only associated with lethal ventricular arrhythmia in patients,9,10 but has also been shown to be mechanistically important in the development of re-entrant arrhythmias through the development of spatially discordant alternans.11,12
Editorial see p 1369
In This Issue, see p 1361
At the cellular level, beat-to-beat alternation in the amplitude of the intracellular Ca2+ transient has been shown to underlie repolarization alternans,8 which is in turn expressed as clinically observed T-wave alternans.11 Various subcellular mechanisms governing the development of intracellular Ca2+ alternans have been proposed, including alternating L-type Ca2+ current (ICa),13 alternating diastolic SR Ca2+ load,3,14 and alternating refractoriness of the SR release channel (ryanodine receptor [RyR]).4,15 Many of these detailed mechanistic investigations have been performed in isolated cardiac myocytes, where total SR Ca2+ content can be measured with caffeine pulses3 or free intra-SR [Ca2+] ([Ca2+]SR) optically monitored with a low-affinity Ca2+ indicator.4,15 However, these single cell studies provide little insight into the spatially heterogeneous nature of SR Ca2+ cycling and how this affects the emergence, severity, and concordance of cardiac alternans in myocardial tissue. Thus, it is difficult to extrapolate directly findings in isolated cells to the intact heart. Furthermore, arrhythmogenic behavior, such as spatially discordant alternans and consequent ventricular fibrillation (VF), are inherently tissue-level phenomena and thus can only be studied in the intact heart.
Experimental investigations into the mechanisms of Ca2+ alternans in the intact heart and the role of Ca2+ in arrhythmogenesis have predominantly used wide-field optical mapping that can record signals over multiple sites but, until now, have been limited to dual mapping of transmembrane potential (Vm) and intracellular Ca2+ and so have been unable to examine SR Ca2+ kinetics directly. In this study, we report for the first time simultaneous mapping of Vm and [Ca2+]SR across the surface of the intact heart and use this novel approach to investigate the role of SR Ca2+ in cardiac alternans and VF. We further investigated the role of RyR refractoriness by sensitizing the RyR with low-concentration (200 μmol/L) caffeine and the combined effects of sensitized RyR and increased SR Ca2+-ATPase (SERCA) activity through β-adrenergic receptor (β-AR) stimulation with isoproterenol (ISO, 100 nmol/L).
An expended Methods section is provided in the online-only Data Supplement. All procedures involving animals were approved by the Animal Care and Use Committee of the University of California, Davis, and adhered to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. Male New Zealand white rabbits (n=27) were anesthetized with an intravenous injection of pentobarbital sodium (50 mg/kg). Hearts were removed rapidly and Langendorff perfused as described previously.16 An ECG was recorded continuously, and pacing was from the base of the left ventricular epicardium.
To monitor intra-SR free [Ca2+] optically, hearts were loaded with the low-affinity Ca2+ indicator Fluo-5N AM (Invitrogen, Carlsbad, CA) for 60 minutes at room temperature.17 Hearts were subsequently stained with the voltage-sensitive dye RH237 (Invitrogen, Carlsbad, CA). All experiments were performed at 37°C. In a separate set of hearts (n=3), dual optical mapping of Vm with RH237 and intracellular Ca2+ with Rhod-2 AM (Invitrogen, Carlsbad, CA) was performed as previously described.16 To induce alternans and ventricular arrhythmia, the pacing cycle length (PCL) was decremented in 10 ms steps until loss of 1:1 capture or induction of VF (Online Figure III). In a subset of animals, either low-concentration caffeine (200 μmol/L; n=6) to sensitize RyR18 or ISO (100 nmol/L; n=3) to stimulate β-AR was added to the perfusate.
Data analysis was performed using 2 commercially available analysis programs (BV_Analyze, Brainvision, Tokyo, Japan and Optiq, Cairn, United Kingdom). Vm and [Ca2+]SR data sets were spatially aligned and processed with a Gaussian spatial filter (radius 3 pixels). The spectral method, which has been used clinically for detecting microvolt T-wave alternans,10 was used to detect the presence of significant action potential duration (APD) and SR Ca2+ alternans as previously described.19 The magnitude of SR Ca2+ release alternans was calculated as 1 minus the ratio of the average small beat (S) release amplitude to the average large beat (L) release amplitude (1−S/L). Diastolic SR Ca2+ load alternans was calculated as the average difference between diastolic levels (D) of S and L beats divided by the average L amplitude (D/L) during a 1 to 2 seconds recording (Figure 3B). Data are expressed as mean±SD and were compared using Student t tests, paired where appropriate. P<0.05 was considered statistically significant.
Simultaneous Optical Mapping of Vm and [Ca2+]SR
To simultaneously map Vm and [Ca2+]SR signals at high spatial and temporal resolution, hearts were loaded with RH237 and Fluo-5N AM. Online Figure IA shows a schematic diagram of the optical setup for dual optical mapping of Vm and [Ca2+]SR. When hearts were loaded with either RH237 or Fluo-5N AM alone, no significant cross talk or bleed through of signal was observed in the other channel (Online Figure IB), confirming complete spectral separation between the Vm and [Ca2+]SR signals. To validate [Ca2+]SR signal kinetics and the dynamic spatial and temporal relationship between Vm and [Ca2+]SR, hearts were paced at a PCL of 300 ms. SR Ca2+ transients had the expected morphology of a rapid downstroke after the AP upstroke and a monotonic increase back to baseline levels because Ca2+ is pumped back into the SR (Figure 1Ai and 1Aiii). As expected from intracellular Ca2+ ([Ca2+]i) measurements,16 the Vm upstroke preceded the [Ca2+]SR downstroke by an average of 9.2±1.6 ms (n=7).
Frequency-Dependent [Ca2+]SR Changes
To test the response of [Ca2+]SR to changes in heart rate, optical recordings were made during progressive increases in pacing rate (Figure 1B). Diastolic [Ca2+]SR increased rapidly (within 1–2 beats) to a new steady state as the PCL decreased, with alternans occurring at shorter PCLs (Figure 1Biii). Once pacing was stopped, diastolic [Ca2+]SR quickly returned to the baseline level as a result of a larger release during the first sinus beat, which occurs after a pause (Figure 1Biv). These traces were normalized as indicated in Figure 1B, and normalized mean diastolic and systolic [Ca2+]SR with changes in PCL are shown in Figure 1C (n=3).
Pacing-Induced APD and SR Ca2+ Alternans
Alternans was induced by decrementing the PCL in 10 ms steps. An example of increasing alternans magnitude with decreasing PCL is shown in Figure 2. In this example, significant SR Ca2+ alternans was induced at PCL=220 ms, whereas significant APD alternans did not occur until 190 ms. The PCL threshold for SR Ca2+ alternans was significantly longer than for Vm alternans (217±10 versus 190±13 ms; P<0.05). APD and SR Ca2+ alternans normally occurred in-phase (large SR Ca2+ release corresponding to long APD and vice versa). Both APD and SR Ca2+ alternans progressively increased with decreasing PCL, as shown in the maps of spectral magnitude (Figure 2A and 2B) and in example traces (Figure 2C and 2D).
Role of Diastolic [Ca2+]SR During Alternans
Diastolic [Ca2+]SR has the ability to change on a beat-to-beat basis in response to increasing heart rate, as shown in Figure 1B. Changes in diastolic [Ca2+]SR can also contribute to Ca2+ alternans, with a larger diastolic SR Ca2+ load facilitating a larger SR Ca2+ release on the subsequent beat and a lower load leading to a smaller release.3 To investigate the role of diastolic [Ca2+]SR in contributing to alternans in the intact heart, the emergence of alternation in diastolic [Ca2+]SR was compared with alternation of SR Ca2+ release. SR Ca2+ release alternans occurring without changes in diastolic [Ca2+]SR were routinely observed (Figure 3Bi). SR Ca2+ release alternans typically occurred before diastolic [Ca2+]SR alternans (ie, at longer PCLs), with significant SR Ca2+ load alternans only occurring at shorter PCLs (Figure 3D and 3E). Because of heterogeneity of SR Ca2+ handling throughout the heart, SR Ca2+ release alternans both with and without diastolic SR Ca2+ load alternans were routinely observed occurring simultaneously in the same heart at the same PCL (200 ms in the example of Figure 3A and 3B). Thus in the intact heart, diastolic [Ca2+]SR alternans are not required for SR Ca2+ release alternans to occur. We suggest that RyR refractoriness may initiate SR Ca2+ release alternans at longer PCLs and might cause alternans of both diastolic [Ca2+]SR and APD as PCL shortens further.
Role of RyR Refractoriness on SR Ca2+ and APD Alternans
RyR release kinetics play a key role in ECC, and under normal conditions, the number of RyRs opening during Ca2+-induced Ca2+ release varies little from beat-to-beat, resulting in consistent SR Ca2+ release.1 However, when a shorter diastolic interval occurs, such as during a premature stimulus, a smaller SR Ca2+ release is observed because of incomplete recovery of RyRs from refractoriness. Caffeine sensitizes RyRs to Ca2+, and at high concentration, it can cause SR Ca2+ release even when the SR is refractory to activation by ICa.20 Therefore, we tested whether low-concentration caffeine (200 μmol/L) could modify RyR refractoriness and shift the alternans threshold to a shorter PCL. As expected, caffeine significantly reduced diastolic [Ca2+]SR after 20 minutes of treatment (Figure 4C; P<0.05). To quantify RyR recovery from refractoriness, we evaluated the ratio of the S2-induced SR Ca2+ release amplitude to the S1 release amplitude (S2/S1 ratio) at S2-coupling intervals ranging from 500 to 180 ms before and after sensitization of RyR with caffeine. Caffeine resulted in a significant increase in the S2/S1 ratio at S2-coupling intervals of 200 and 180 ms (P<0.05; Figure 4A and 4B), suggesting a decrease in RyR refractoriness with caffeine treatment. As expected, caffeine did not alter the time course of SR Ca2+ recovery (τ, Figure 4D).
In line with these data, RyR refractoriness was found to play a key role in the onset of APD and SR Ca2+ alternans. Figure 5 demonstrates the effect of caffeine (200 μmol/L) on the emergence and severity of APD and SR Ca2+ alternans. The spectral magnitude of both APD and SR Ca2+ alternans at a PCL of 180 ms was reduced after caffeine treatment (Figure 5A and 5B versus Figure 5C and 5D). Figure 5E to 5G shows that caffeine significantly reduced the magnitude of SR Ca2+ release alternans at PCLs of 200 to 160 ms (Figure 5E), the magnitude of diastolic SR Ca2+ load alternans at PCLs of 180 to 160 ms (Figure 5F), and the pacing threshold for induction of both SR Ca2+ and APD alternans (baseline: 217±10 and 190±13 ms versus caffeine: 188±15 and 173±12 ms; P<0.05 for SR Ca2+ and APD, respectively [Figure 5G]).
To assess further the role of RyR refractoriness on alternans and the interplay between RyR refractoriness and SERCA function, β-AR stimulation was performed with ISO (100 nmol/L). Like caffeine, β-AR stimulation sensitizes RyR, but unlike caffeine, it also significantly increases SERCA function. As expected, ISO reduced SR Ca2+ release time (time to nadir) from 56.3±3.5 to 42.2±1.8 ms, and SR Ca2+ recovery (τ) decreased from 72.8±6.3 to 58.6±4.8 ms (P<0.05 for both; Online Figure VIB and VIC). ISO decreased the magnitude and pacing thresholds of SR Ca2+ and APD alternans from 247±6 to 173±12 ms (P<0.05) and 200±10 to 153±12 ms (P=0.06), respectively (Online Figures VII and VIII). Importantly, increased SERCA function with ISO prevented diastolic SR Ca2+ load alternans even at rapid pacing rates, yet alternans of SR Ca2+ release still occurred because of encroachment on RyR refractoriness (Online Figure VIIIB–VIIID).
Spatial Heterogeneity of Alternans and Spatial Discordance
A key advantage of mapping-based techniques is the ability to record spatially heterogeneous phenomena, which may play an important role in the development of arrhythmia. At the whole heart level, Vm alternans may become spatially discordant, where one area of tissue exhibits a long–short APD sequence and another area has the opposite, short–long sequence. Discordant alternans are associated with large spatial gradients of repolarization and can initiate re-entrant arrhythmias.11,21 In this study, spatial heterogeneity in the magnitude of alternans (of both APD and SR Ca2+) was routinely observed with some areas seemingly more prone to the development of alternans than others (alternans emerge in these areas at longer PCLs and have a larger magnitude as the PCL is decreased). As demonstrated in Figure 2A and 2B, local onset of SR Ca2+ alternans was first observed in the central region of the mapping field of view at PCL=220 ms. APD alternans followed at a PCL of 190 ms and emerged at the same location. Interestingly, when spatially discordant alternans developed at PCL=150 ms, the area of largest alternans magnitude remained similar to the original onset location. Spatial heterogeneity of alternans magnitude was also clearly observed in Figure 5A and remained even after caffeine treatment (Figure 5C). Additional examples of spatial heterogeneity of APD and SR Ca2+ alternans magnitude are shown in Online Figure V. Spatially discordant alternans were observed in 12 of 14 hearts with discordance emerging at the same PCL for both SR Ca2+ and APD alternans (140±12 ms). Not surprisingly, VF could only be induced in those hearts that first exhibited spatial discordance, supporting the role of discordant alternans in the initiation of VF.
Spatially Discordant Alternans and the Role of RyR Refractoriness
To investigate the role of RyR refractoriness in contributing to spatially discordant alternans, the threshold and magnitude of discordant alternans were compared before and after 200 μmol/L caffeine treatment. Figure 6 shows an example of spatially discordant alternans from a heart paced at 140 ms under control conditions (Figure 6A and 6B) and after caffeine treatment (Figure 6C and 6D). Spatially discordant APD and SR Ca2+ alternans were induced with distinct phases separated by black nodal lines (Figure 6A and 6C) both before and after caffeine treatment. Example traces of APD and SR Ca2+ from 2 locations show opposite phases on either side of the nodal line (Figure 6B and 6D, black versus white box), although APD and SR Ca2+ alternans remained in-phase with one another (ie, large SR Ca2+ transient corresponds to long APD and vice versa). Although 200 μmol/L caffeine reduced the spectral magnitude of SR Ca2+ alternans at PCL=140 ms (69±41 versus 37±32 AU; P<0.05), it did not reduce the spectral magnitude of APD alternans (20±21 versus 19±12 AU; P=0.9; Figure 6E) nor did it reduce the pacing threshold for induction of either spatially discordant APD or SR Ca2+ alternans (Figure 6F). Caffeine also did not affect the pacing threshold for induction of VF (118±8 versus 113±6 ms; Figure 6G). These data suggest that mechanisms other than RyR kinetics, such as restitution of APD and conduction velocity (CV), may become more critical for the transition from spatially concordant to spatially discordant alternans and the transition of alternans to VF. An analysis of CV restitution kinetics during alternans indicates that the transition from spatial concordance to spatial discordance occurs only after CV restitution is evoked (Online Figure IV).
Comparison of [Ca2+]SR and [Ca2+]i During VF
To investigate further the role of [Ca2+]SR and [Ca2+]i during VF, hearts were loaded with either RH237 and Fluo-5N AM for dual imaging of Vm and [Ca2+]SR or RH237 and Rhod-2 AM for Vm and [Ca2+]i. In 8 of 14 hearts, rapid pacing induced nonsustained VF. Figure 7A shows an example of a heart with dual imaging of Vm and [Ca2+]SR during ventricular pacing and VF. When normalized to the amplitude of the SR Ca2+ transient during ventricular pacing, minimal SR Ca2+ release occurred during VF (<10% of normal SR Ca2+ release amplitude, Figure 7Aiii), suggesting near-complete RyR refractoriness during VF. Additional signal locations during this same VF episode are shown in Online Figure IX. However, dual imaging of Vm and [Ca2+]i (Figure 7B) shows that during VF, [Ca2+]i oscillated with a peak-to-peak amplitude that is ≈30% of the normal intracellular Ca2+ transient amplitude during pacing (Figure 7Biii). Although the absolute [Ca2+]SR cannot be quantified with Fluo-5N in the intact heart, a comparison of the fluorescence during VF and after spontaneous cardioversion to sinus rhythm showed that, during VF, [Ca2+]SR is ≈30% higher than normal diastolic levels (Figure 8iii).
In this study, we developed the first methodology for simultaneous optical mapping of Vm and SR Ca2+ at high spatial–temporal resolution in the intact heart. The voltage-sensitive dye RH237 and low-affinity Ca2+ indicator Fluo-5N AM were used together to map Vm and [Ca2+]SR, respectively, with high fidelity and no fluorescent cross talk. We used this novel approach to investigate the role of SR Ca2+ in cardiac alternans and VF. The results demonstrate that (1) APD and SR Ca2+ alternans predominantly occur in-phase with one another, but SR Ca2+ alternans emerge first, at longer PCLs. As expected, both APD and SR Ca2+ alternans worsen with decreasing PCL. (2) RyR refractoriness plays a key role in the onset of SR Ca2+ alternans. This was demonstrated by the observation that SR Ca2+ release alternans often occur in the absence of diastolic [Ca2+]SR alternans and that release alternans tend to emerge at longer PCLs, with diastolic [Ca2+]SR alternans occurring as PCL decreases. Accordingly, sensitizing RyR with low-concentration caffeine (200 μmol/L) reduced the pacing threshold for onset, as well as the magnitude of SR Ca2+ and APD alternans. Even when SERCA function is sufficiently increased (with 100 nmol/L ISO) to prevent alternans of diastolic [Ca2+]SR, release alternans were not prevented because of encroachment on RyR refractoriness. (3) Spatially discordant alternans of both APD and SR Ca2+ emerge at even shorter PCLs. Although caffeine reduces the magnitude of spatially discordant SR Ca2+ alternans, it does not prevent the onset of discordance nor does it reduce the magnitude of spatially discordant APD alternans. (4) During VF, [Ca2+]SR is substantially higher than during normal rhythms, but the RyRs remain almost continuously refractory with minimal Ca2+ release from the SR. This key finding may have important clinical implications for the treatment of VF and for preventing spontaneous reinitiation of arrhythmic events.
Dual Optical Mapping of Vm and [Ca2+]SR
For the past few decades, dual optical mapping of Vm and intracellular Ca2+ in intact hearts has provided a wealth of information toward a more complete understanding of normal cardiac ECC and the mechanisms of ventricular arrhythmias.16,22,23 In mammalian hearts, the majority of the intracellular Ca2+ transient is comprised of Ca2+ release from and reuptake into the SR (≈70% in the rabbit heart).24 Thus, the ability to discern precisely whether changes in intracellular Ca2+ are because of transmembrane Ca2+ flux or SR Ca2+ release/reuptake provides important insight into the mechanisms of ECC and Ca2+-mediated arrhythmias. Low-affinity fluorescent Ca2+ indicators, such as Fluo-5N and Mag-Fluo4, can be used for this purpose. These indicators have dissociation constants (Kd) in the range of 10 to 400 μmol/L and exhibit minimal fluorescence in the cytosol compared with the SR lumen and the fluorescence decreases on SR Ca2+ release.25 Several groups have recently reported using these low-affinity Ca2+ indicators to monitor [Ca2+]SR optically in isolated myocytes to investigate fractional SR Ca2+ release25 and the mechanisms of Ca2+ alternans.4,15 Furthermore, Mag-Fluo4, combined with a pulsed local field fluorescent microscope, has been used to record [Ca2+]SR from a single location on the epicardial surface of the intact mouse heart.17,26 Although these studies have provided new and important information on SR Ca2+ release and reuptake during normal and pathological circumstances, until now, methods to image SR Ca2+ activity across the entire surface of the intact heart had not been developed. This spatial information is vitally important for further understanding the heterogeneous nature of SR Ca2+ cycling in the heart and the role of SR Ca2+ in spatially distinct arrhythmic phenomena such as focal arrhythmia sources, spatially discordant alternans, and VF.
This study used dual optical mapping techniques to monitor, for the first time, Vm and [Ca2+]SR simultaneously at high spatial and temporal resolution. We developed an optical setup using the voltage-sensitive indicator RH237 and the low-affinity Ca2+ indicator Fluo-5N AM that results in minimal fluorescent cross talk between Vm and [Ca2+]SR signals (Online Figure IB). More importantly, detailed analysis of signal kinetics and the dynamic temporal relationship between Vm and [Ca2+]SR shows that SR Ca2+ transients have the expected morphology of a rapid downstroke, indicating SR Ca2+ release after the AP upstroke and a monotonic increase back to baseline levels as Ca2+ is pumped back into the SR (Figure 1Ai and 1Aiii). As expected, diastolic [Ca2+]SR rapidly increases or decreases in response to changes in heart rate (Figure 1B–1C), consistent with studies in isolated cardiac myocytes.25 The average Vm-[Ca2+]SR delay of 9.2±1.6 ms measured with this approach is comparable to Vm-[Ca2+]i delays reported in previous studies.16,22 These data indicate that dual optical mapping of Vm and [Ca2+]SR in the intact heart is indeed a reliable tool for study of ECC and the mechanisms of cardiac arrhythmias.
Mechanisms of SR Ca2+ Alternans
Cardiac alternans play an important role in ventricular arrhythmias and have been studied extensively both clinically and experimentally.11,27–29 Clinically, pacing- or exercise-induced T-wave alternans have been reported as a highly sensitive marker for sudden cardiac death.11,27 The search for the mechanisms underlying T-wave alternans has revealed that intracellular Ca2+ alternans are a key determinant.8,11,29 Consistent with previous studies,4,15 our results show that SR Ca2+ transients alternate as heart rate increases (Figure 2). Although APD alternans are typically in-phase with SR Ca2+ alternans, APD alternans emerge at significantly faster heart rates (Figures 2E and 5G). Importantly, APD alternans emerge from the same location as do the earliest and strongest SR Ca2+ alternans (Figure 2A and 2B), indicating that SR Ca2+ alternans likely contribute to the development of APD alternans.
Diastolic [Ca2+]SR varies little from beat-to-beat under normal conditions, indicating a balance between SERCA uptake and RyR release during each ECC cycle. An imbalance between SERCA uptake and RyR release may result in beat-to-beat alternation of diastolic [Ca2+]SR and SR Ca2+ release. Studies performed in normal isolated myocytes show that either alternating diastolic [Ca2+]SR3,14 or alternating refractoriness of RyR4,15 may contribute to SR Ca2+ alternans. In the intact heart, however, our results indicate that SR Ca2+ release alternans routinely occur without changes in diastolic [Ca2+]SR (Figure 3Bi). Although diastolic SR Ca2+ load alternans can also occur, they tend to emerge at faster heart rates after the onset of release alternans (Figure 3E). It is logical that SR Ca2+ release alternans (when large enough) would cause alternation of diastolic [Ca2+]SR.4 A larger SR Ca2+ release can reduce net Ca2+ entry through ICa because of Ca2+-dependent inactivation and increase Ca2+ efflux via the Na+-Ca2+ exchanger, thus causing net cellular Ca2+ loss and secondary diastolic [Ca2+]SR alternans. In this way, diastolic [Ca2+]SR alternans may amplify Ca2+ release and APD alternans. Furthermore, at fast heart rates, SERCA may no longer be able to fully sequester all the Ca2+ released before the onset of the next AP, thus further contributing to alternans of diastolic [Ca2+]SR. However, our data demonstrate that in the intact heart, diastolic [Ca2+]SR alternans are not a requirement for SR Ca2+ release alternans to occur, suggesting that other mechanisms such as RyR refractoriness may contribute to the onset of SR Ca2+ alternans.
Role of RyR Refractoriness in SR Ca2+ Alternans
An important property of RyRs is that they have a period of refractoriness after each opening. Under steady state conditions, the number of individual RyRs opening during Ca2+-induced Ca2+ release varies little from beat-to-beat, resulting in consistent SR Ca2+ release (Figure 1). However, because heart rate increases and the diastolic interval between beats becomes shorter, the recovery of RyRs from refractoriness limits SR Ca2+ release. If only a subset of RyRs have fully recovered from refractoriness, less Ca2+ will be released. On the next beat, the portion of RyRs that were not available for release on the previous beat has now fully recovered from refractoriness, resulting in a larger SR Ca2+ release. This alternation of SR Ca2+ release can lead to intracellular Ca2+ alternans.
In this study, RyR refractoriness was demonstrated and quantified using an S1–S2 pacing protocol, where a premature stimulus (S2) induces a smaller SR Ca2+ release compared with the previous S1-induced release (Figure 4A). As expected, sensitization of RyR with low-concentration caffeine15,30 results in a significant increase in the S2-induced SR Ca2+ release (Figure 4A and 4B), indicating a decrease in RyR refractoriness with caffeine. Interestingly, sensitization of RyR decreases the induction threshold and magnitude of not only SR Ca2+ alternans but also APD alternans (Figure 5), indicating that RyR refractoriness plays a key role in the onset of both SR Ca2+ and APD alternans. β-AR stimulation also sensitizes RyR, but unlike caffeine, β-AR stimulation significantly increases SERCA activity (Online Figure VIC) via phosphorylation of phospholamban. Accordingly, β-AR stimulation with ISO reduced the magnitude and induction threshold for both SR Ca2+ and APD alternans (Online Figures VII and VIII), likely via the combined effects of reduced RyR refractoriness and increased SERCA activity. This increase in SERCA activity prevented alternans of diastolic [Ca2+]SR even at the fastest pacing rates (Online Figure VIIIB and VIIID), yet SR Ca2+ release alternans still occurred because of encroachment on RyR refractoriness (Online Figure VIIIB and VIIIC), but as with caffeine, release alternans occurred at a faster pacing rate compared with baseline because of sensitization of RyR.
Spatially Discordant Alternans
Experimental studies have revealed complex spatiotemporal patterns of cardiac alternans, including spatially discordant alternans which are particularly arrhythmogenic because of the extreme gradients of repolarization generated.11 Several mechanisms contributing to spatially discordant alternans have been identified, including CV restitution, spatial heterogeneities of Ca2+ cycling, and intercellular uncoupling.31 Because of the significant role RyR refractoriness plays in contributing to the onset of SR Ca2+ and APD alternans, we sought to determine whether RyR refractoriness also contributes to the emergence or severity of spatially discordant alternans.
Spatially discordant APD and SR Ca2+ alternans were routinely induced both before and after caffeine treatment (Figure 6A and 6C). Although caffeine reduced the severity of discordant SR Ca2+ alternans (as measured by the average spectral magnitude at PCL=140 ms; Figure 6E), it did not reduce the severity of discordant APD alternans, the induction threshold for discordance (Figure 6F), or the threshold for induction of VF (Figure 6G). These data suggest that other dynamical mechanisms such as restitution of CV may govern the onset and severity of spatially discordant alternans rather than SR Ca2+ handling.12,21,32 An analysis of CV restitution kinetics during alternans revealed that the transition to spatial discordance occurs only after CV restitution is evoked (Online Figure IV). Thus, although RyR refractoriness plays a key role in the onset of concordant alternans, the transition to and maintenance of discordant alternans may rely more heavily on dynamical instabilities.
Role of SR Ca2+ and RyR Refractoriness During VF
The role of intracellular Ca2+ during VF has been studied previously.33–35 In those studies, diastolic intracellular Ca2+ overload was routinely observed, which may set the stage for spontaneous SR Ca2+ release on termination of VF. Previous reports on the mechanisms of postshock arrhythmias after successful defibrillation of VF suggest that spontaneous SR Ca2+ release, leading to late phase 3 early afterdepolarizations or delayed afterdepolarizations, may be responsible for VF reinitiation.36,37 Our results demonstrating near-complete RyR refractoriness and an increase in [Ca2+]SR during VF support these findings. On spontaneous termination of VF, [Ca2+]SR is ≈30% higher than during normal sinus rhythm (Figure 8iii). Because the likelihood of spontaneous SR Ca2+ release increases with increasing [Ca2+]SR,24 the time period immediately post-VF may represent a particularly vulnerable time for spontaneous SR Ca2+ release. This observation may have important clinical implications for successful defibrillation and the treatment of electrical storm.38
Our findings may also offer new insights into the interpretation of intracellular Ca2+ signals during VF. Consistent with previous studies, our dual optical mapping data of Vm and [Ca2+]i show peak-to-peak oscillation during VF that is ≈30% of the amplitude of the normal intracellular Ca2+ transient (Figure 7Biii). However, dual imaging of Vm and [Ca2+]SR demonstrates that minimal SR Ca2+ release occurs during VF (Figures 7Aiii and 8iii; Online Figure IX). Because SR Ca2+ release and reuptake accounts for ≈70% of the total intracellular Ca2+ transient in the normal rabbit heart24 (with the remaining 30% representing transsarcolemmal Ca2+ flux through L-type Ca2+ channels and Na+-Ca2+ exchanger), typical intracellular Ca2+ recordings during VF are likely reflective of transsarcolemmal Ca2+ fluxes only. It is possible that ICa may also be reduced during VF because of the depolarized diastolic Vm (Figure 7Bii) and decreased diastolic interval, which would lead to a secondary reduction in SR Ca2+ release. However, a peak-to-peak amplitude of ≈30% of the normal intracellular Ca2+ transient amplitude during VF (Figure 7Biii) is more consistent with fully available ICa and RyR refractoriness as a mechanism of reduced SR Ca2+ cycling during VF.
RyR Refractoriness as a Therapeutic Target
Given the role of RyR refractoriness in contributing to the onset and progression of alternans and in contributing to SR Ca2+ overload during VF, sensitizing RyR to reduce alternans or improve SR Ca2+ release during VF may represent an enticing therapeutic target. However, RyR sensitization and reduced refractoriness (either via catecholaminergic polymorphic ventricular tachycardia-linked mutations or enhanced phosphorylation) can significantly enhance arrhythmogenic spontaneous SR Ca2+ release events (Ca2+ sparks and waves) and lead to delayed afterdepolarizations and focal arrhythmias.39 Therefore, the pro- or antiarrhythmic consequences of RyR sensitization ultimately depend on the underlying cardiac pathology and arrhythmia mechanism in question.
Vm and [Ca2+]SR were imaged using single-wavelength emission. Unlike myocyte experiments,25 it was not feasible to calibrate [Ca2+]SR signals in the intact heart (as is true for Vm and [Ca2+]i). However, our [Ca2+]SR measurements agree qualitatively with rabbit myocyte measurements, and thus relative [Ca2+]SR changes are likely valid. The detailed mechanisms of alternans and VF induction in pathological conditions might differ from those of healthy hearts used here. However, future studies using simultaneous optical mapping of Vm and [Ca2+]SR in the intact heart may provide a novel tool for examining SR Ca2+ handling and arrhythmia in pathological conditions, and the findings of the present study will provide an important mechanistic foundation.
These findings shed new light on the role of SR Ca2+ in the progression from normal rhythms to cardiac alternans and subsequent arrhythmia. Based on the results of this study, we propose a continuum of mechanisms (RyR→SERCA→APD/CV restitution) responsible for the onset and progression of alternans: RyR refractoriness is first encroached on, which leads to SR Ca2+ release alternans. When heart rate increases, SR Ca2+ release alternans increase, diastolic SR Ca2+ load begins to alternate (impacted by SERCA function), and APD alternans emerge. At even faster rates, dynamical instabilities, such as APD and CV restitution, may lead to spatially discordant alternans, which can facilitate re-entrant arrhythmia and VF. Importantly, our results demonstrate that SR Ca2+ release is nearly continuously refractory during VF despite an increase in diastolic [Ca2+]SR and [Ca2+]i, suggesting that the intracellular Ca2+ transients observed during VF are mainly comprised of transsarcolemmal Ca2+ currents.
Sources of Funding
This study was supported by the National Institutes of Health R01 HL111600 (C.M. Ripplinger), P30 HL101280 (C.M. Ripplinger and D.M. Bers), P01 HL80101 (D.M. Bers), and T32 GM099608 (N.M. De Jesus and D.M. Bers), and the American Heart Association 12SDG9010015 (C.M. Ripplinger).
In January 2014, the average time from submission to first decision for all original research papers submitted to Circulation Research was 14.35 days.
The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.114.302505/-/DC1.
- Nonstandard Abbreviations and Acronyms
- β-adrenergic receptor
- action potential
- action potential duration
- intracellular Ca2+
- intra-sarcoplasmic reticulum free Ca2+
- conduction velocity
- excitation–contraction coupling
- L-type Ca2+ current
- pacing cycle length
- ryanodine receptor
- sarcoplasmic reticulum Ca2+-ATPase
- sarcoplasmic reticulum
- ventricular fibrillation
- transmembrane potential
- Received August 31, 2013.
- Revision received February 18, 2014.
- Accepted February 25, 2014.
- © 2014 American Heart Association, Inc.
- Díaz ME,
- O’Neill SC,
- Eisner DA
- Picht E,
- DeSantiago J,
- Blatter LA,
- Bers DM
- Goldhaber JI,
- Xie LH,
- Duong T,
- Motter C,
- Khuu K,
- Weiss JN
- Verrier RL,
- Klingenheben T,
- Malik M,
- El-Sherif N,
- Exner DV,
- Hohnloser SH,
- Ikeda T,
- Martínez JP,
- Narayan SM,
- Nieminen T,
- Rosenbaum DS
- Pastore JM,
- Girouard SD,
- Laurita KR,
- Akar FG,
- Rosenbaum DS
- Qu Z,
- Garfinkel A,
- Chen PS,
- Weiss JN
- Shkryl VM,
- Maxwell JT,
- Domeier TL,
- Blatter LA
- Myles RC,
- Wang L,
- Kang C,
- Bers DM,
- Ripplinger CM
- Kornyeyev D,
- Reyes M,
- Escobar AL
- Bassani JW,
- Bassani RA,
- Bers DM
- Choi BR,
- Salama G
- Efimov IR,
- Nikolski VP,
- Salama G
- Bers DM
- Shannon TR,
- Guo T,
- Bers DM
- Valverde CA,
- Kornyeyev D,
- Ferreiro M,
- Petrosky AD,
- Mattiazzi A,
- Escobar AL
- Hirayama Y,
- Saitoh H,
- Atarashi H,
- Hayakawa H
- Wilson LD,
- Rosenbaum DS
- Pastore JM,
- Rosenbaum DS
- Hwang GS,
- Hayashi H,
- Tang L,
- Ogawa M,
- Hernandez H,
- Tan AY,
- Li H,
- Karagueuzian HS,
- Weiss JN,
- Lin SF,
- Chen PS
- Maruyama M,
- Joung B,
- Tang L,
- Shinohara T,
- On YK,
- Han S,
- Choi EK,
- Kim DH,
- Shen MJ,
- Weiss JN,
- Lin SF,
- Chen PS
- Nademanee K,
- Taylor R,
- Bailey WE,
- Rieders DE,
- Kosar EM
Novelty and Significance
What Is Known?
T-wave alternans (a noninvasive marker of increased risk for arrhythmia and sudden cardiac death) can be caused by beat-to-beat alternation in amplitude of the intracellular Ca2+ transient.
Several mechanisms may contribute to Ca2+ transient alternans, including alternating sarcoplasmic reticulum (SR) Ca2+ load and alternating refractoriness of ryanodine receptors (RyR: the SR Ca2+ release channel).
Previous studies on detailed mechanisms of Ca2+ alternans have mostly been performed in isolated cardiac myocytes, where direct measurement of intra-SR free [Ca2+] is possible.
What New Information Does This Article Contribute?
Using a novel imaging approach to monitor intra-SR free [Ca2+] and transmembrane potential in the intact heart, we found that as heart rate increases, RyR refractoriness is the first mechanism encroached on and leads to initial development of SR Ca2+ release alternans.
At faster heart rates, SR Ca2+ load also begins to alternate, which further increases the magnitude of SR Ca2+ release alternans and consequent repolarization alternans.
At extremely fast activation rates during ventricular fibrillation (VF), SR Ca2+ load is high, but RyRs remain nearly continuously refractory, resulting in minimal SR Ca2+ release.
For the first time, we performed dual optical mapping of intra-SR free [Ca2+] and transmembrane potential to determine the role of SR Ca2+ handling during cardiac alternans and VF. Previous studies on the detailed mechanisms of SR Ca2+ handling during alternans have mostly been performed in isolated cardiac myocytes. However, cardiac arrhythmias (including alternans and VF) are spatiotemporally complex, emergent phenomena that must be studied in intact tissue. This study revealed that, in the intact heart, RyR refractoriness is the mechanism that is first encroached on as heart rate increases and contributes to the initial development of SR Ca2+ alternans. When heart rate increases further, the SR Ca2+-ATPase can no longer sequester all the Ca2+ released on the prior beat before the onset of the next action potential. Therefore, SR Ca2+ load also begins to alternate, which further increases the magnitude of SR Ca2+ release alternans and consequent repolarization alternans. During VF, ventricular activation rates are extremely fast, leading to near-continuous RyR refractoriness and thus minimal SR Ca2+ release, despite an increase in SR Ca2+ load. These data are the first to define the role of SR Ca2+ handling in the intact heart during cardiac alternans and VF.