The cardiac ryanodine receptor, but not sarcoplasmic reticulum Ca2+-ATPase, is a major determinant of Ca2+ alternans in intact mouse hearts

Sarcoplasmic reticulum (SR) Ca2+ cycling is governed by the cardiac ryanodine receptor (RyR2) and SR Ca2+-ATPase (SERCA2a). Abnormal SR Ca2+ cycling is thought to be the primary cause of Ca2+ alternans that can elicit ventricular arrhythmias and sudden cardiac arrest. Although alterations in either RyR2 or SERCA2a function are expected to affect SR Ca2+ cycling, whether and to what extent altered RyR2 or SERCA2a function affects Ca2+ alternans is unclear. Here, we employed a gain-of-function RyR2 variant (R4496C) and the phospholamban-knockout (PLB-KO) mouse model to assess the effect of genetically enhanced RyR2 or SERCA2a function on Ca2+ alternans. Confocal Ca2+ imaging revealed that RyR2-R4496C shortened SR Ca2+ release refractoriness and markedly suppressed rapid pacing–induced Ca2+ alternans. Interestingly, despite enhancing RyR2 function, intact RyR2-R4496C hearts exhibited no detectable spontaneous SR Ca2+ release events during pacing. Unlike for RyR2, enhancing SERCA2a function by ablating PLB exerted a relatively minor effect on Ca2+ alternans in intact hearts expressing RyR2 WT or a loss-of-function RyR2 variant, E4872Q, that promotes Ca2+ alternans. Furthermore, partial SERCA2a inhibition with 3 μm 2,5-di-tert-butylhydroquinone (tBHQ) also had little impact on Ca2+ alternans, whereas strong SERCA2a inhibition with 10 μm tBHQ markedly reduced the amplitude of Ca2+ transients and suppressed Ca2+ alternans in intact hearts. Our results demonstrate that enhanced RyR2 function suppresses Ca2+ alternans in the absence of spontaneous Ca2+ release and that RyR2, but not SERCA2a, is a key determinant of Ca2+ alternans in intact working hearts, making RyR2 an important therapeutic target for cardiac alternans.

Given its crucial role in cardiac alternans, understanding how Ca 2ϩ alternans occurs would be key to the understanding of cardiac alternans. Over the past decades, major advances in the understanding of the mechanisms of Ca 2ϩ alternans have been made. It has become clear that Ca 2ϩ alternans results from altered SR 5 Ca 2ϩ cycling, which is governed by SR Ca 2ϩ release and reuptake (9, 11, 20 -25).
Inhibiting RyR2 function either by tetracaine, intracellular acidification, or metabolic inhibition has been shown to prolong SR Ca 2ϩ release refractoriness and promote Ca 2ϩ alternans in isolated cardiomyocytes (26 -30). On the other hand, increasing RyR2 function by caffeine shortens SR Ca 2ϩ release refractoriness and suppresses Ca 2ϩ alternans (31,32). These observations suggest that the activity of RyR2 is a major determinant of SR Ca 2ϩ release refractoriness and Ca 2ϩ alternans. Consistent with this view, we showed that genetically suppressing RyR2 function prolongs SR Ca 2ϩ release refractoriness and promotes Ca 2ϩ alternans in intact hearts (32). On the other hand, shortened refractoriness of SR Ca 2ϩ release as a result of CASQ2 (cardiac calsequestrin) ablation suppresses Ca 2ϩ alter- . The authors declare that they have no conflicts of interest with the contents of this article. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This article contains Figs. S1-S6. 1 Recipient of the Alberta Innovates-Health Solutions (AIHS) Fellowship Award. 2 Recipient of the Libin Cardiovascular Institute of Alberta and Cumming School of Medicine Postdoctoral Fellowship Award. 3 Recipient of the AIHS Studentship Award. 4  cro ARTICLE nans in intact hearts (33). These findings support a general notion that suppressing the activity of RyR2 prolongs the refractoriness of SR Ca 2ϩ release and promotes Ca 2ϩ alternans, whereas enhancing RyR2 activity shortens the refractoriness of SR Ca 2ϩ release and suppresses Ca 2ϩ alternans (10,31,32,34,35). However, and contrary to this expectation, enhanced RyR2 function as a result of some genetic mutations or abnormal redox modifications has been shown to promote Ca 2ϩ alternans in isolated cardiomyocytes (36 -38). For instance, the CPVT-causing, gain-of-function (GOF) RyR2 mutation R4496C has been shown to reduce the refractoriness of SR Ca 2ϩ release in the mouse trabecular muscle (39). However, the same GOF RyR2 R4496C mutation was found to promote Ca 2ϩ alternans in isolated cardiac cells (40). Therefore, it remains unclear whether enhanced RyR2 function suppresses or promotes Ca 2ϩ alternans.
These seemingly conflicting observations also raise an important question of why enhanced RyR2 activity and shortened refractoriness of SR Ca 2ϩ release are unable to suppress Ca 2ϩ alternans in the isolated RyR2 R4496C mutant cells. One possible explanation is that enhanced RyR2 activity would increase the propensity for spontaneous Ca 2ϩ release (SCR), such as Ca 2ϩ sparks and Ca 2ϩ waves, which in turn would promote Ca 2ϩ alternans (25,28,38,(41)(42)(43)(44)(45)(46). Hence, it would be of interest and importance to determine whether enhanced RyR2 function promotes Ca 2ϩ alternans in intact working hearts that exhibit little or no SCR during stimulation (47).
SR Ca 2ϩ reuptake, another important aspect of SR Ca 2ϩ cycling, is also believed to play an important role in Ca 2ϩ alternans. Overexpression of the cardiac sarco/endoplasmic reticulum Ca 2ϩ ATPase (SERCA2a) suppresses Ca 2ϩ alternans, whereas reducing SERCA2a expression or activity promotes Ca 2ϩ alternans (5, 9, 48 -51). However, severely reducing the activity of SERCA2a may suppress, rather than promote, Ca 2ϩ alternans, probably due to reduced SR Ca 2ϩ content (25,46,52). Interestingly, atrial overexpression of SERCA2a has little effect on cardiac alternans (53). Hence, the effect of altered SERCA2a activity on Ca 2ϩ alternans is complex and variable, and the relative contribution of altered RyR2 and SERCA2a activity to the genesis of Ca 2ϩ alternans is also unclear.
In the present study, we carried out laser-scanning confocal Ca 2ϩ imaging of cardiomyocytes in intact WT and RyR2 mutant hearts that exhibited little spontaneous Ca 2ϩ release. We assessed the impact of the GOF RyR2 R4496C mutation on SR Ca 2ϩ release refractoriness and Ca 2ϩ alternans and the effects of inhibiting or enhancing SERCA2a activity on Ca 2ϩ alternans in intact hearts. We demonstrate that genetically enhancing RyR2 function shortens Ca 2ϩ release refractoriness and suppresses Ca 2ϩ alternans in intact hearts without producing spontaneous SR Ca 2ϩ release. We also demonstrate that genetically enhancing SERCA2a function by phospholamban (PLB) ablation had a relatively minor impact on Ca 2ϩ alternans in intact hearts. Furthermore, we found that modest inhibition of SERCA2a function also had little impact on Ca 2ϩ alternans in intact hearts. Collectively, our data demonstrate that the activity of RyR2, but not SERCA2a, is a major determinant of Ca 2ϩ alternans in intact working mouse hearts.

Genetically enhancing RyR2 function reduces the refractoriness of SR Ca 2؉ release in intact hearts
We have recently shown that genetically suppressing RyR2 function lengthens the refractoriness of SR Ca 2ϩ release in intact hearts (32). It is unclear whether genetically enhancing RyR2 function would shorten the refractoriness of SR Ca 2ϩ release in intact hearts. To address this question, we employed the RyR2 R4496C mutation, a CPVT-linked GOF RyR2 mutation that has been shown to significantly enhance the sensitivity of the channel to Ca 2ϩ activation (54 -56). We determined the refractoriness of SR Ca 2ϩ release in isolated Langendorff-perfused intact RyR2 WT and heterozygous RyR2 R4496C mutant hearts using the S1S2 stimulation protocol (10). As shown in Fig. 1, the amplitude of Ca 2ϩ transients in both the WT and RyR2 R4496C hearts decreased when the S1S2 interval was progressively reduced (from 200 to 40 ms) (Fig. 1, A and B). However, the WT and RyR2 R4496C hearts showed significantly different relationships between the Ca 2ϩ transient amplitude and S1S2 interval ( Fig. 1C) (p Ͻ 0.05). The Ca 2ϩ transient amplitude of the RyR2 R4496C hearts recovered faster than that of the WT hearts at S1S2 intervals between 75 and 160 ms. Therefore, these data demonstrate that, contrary to the effect of suppressing RyR2 function (32), genetically enhancing RyR2 function shortens the refractoriness of SR Ca 2ϩ release.

Enhancing RyR2 function markedly suppresses rapid stimulation-induced Ca 2؉ alternans in intact hearts
Prolonged refractoriness of SR Ca 2ϩ release is known to promote Ca 2ϩ alternans (10,31,32,34,35). Thus, a shortened refractoriness of SR Ca 2ϩ release as a result of the RyR2 R4496C mutation would be expected to suppress Ca 2ϩ alternans. To test this possibility, we determined the propensity for Ca 2ϩ alternans in isolated Langendorff-perfused intact RyR2 WT and heterozygous RyR2 R4496C mutant hearts. As shown in Fig. 2, RyR2 WT hearts exhibited significant beat-to-beat alternations in the amplitude of Ca 2ϩ transients at the stimulation frequency of 12 Hz ( Fig. 2A). On the other hand, RyR2 R4496C mutant hearts displayed little or no beat-to-beat variations in the amplitude of Ca 2ϩ transients at the same stimulation frequency (12 Hz) (Fig. 2B).
The frequency dependence of Ca 2ϩ alternans in intact WT and RyR2 R4496C hearts is shown in Fig. 2 (C and D). Substantial Ca 2ϩ alternans could be readily detected in RyR2 WT hearts stimulated at 10 -11 Hz, whereas higher stimulation frequencies (13-14 Hz) were required to induce considerable Ca 2ϩ alternans in RyR2 R4496C mutant hearts (Fig. 2, C and D). Furthermore, RyR2 R4496C hearts showed a significantly lower alternans ratio and alternans duration at each stimulation frequency between 10 and 14 Hz (Fig. 2, C and D) (p Ͻ 0.01). Taken together, these data indicate that, contrary to the effect of suppressing RyR2 function (32), genetically enhancing RyR2 function markedly suppresses rapid stimulation-induced Ca 2ϩ alternans in intact hearts.

No spontaneous Ca 2؉ sparks or Ca 2؉ waves were detected in intact RyR2-R4496C mutant hearts during Ca 2؉ alternans
It has been shown that the RyR2 R4496C mutation increases the propensity for spontaneous Ca 2ϩ release (SCR; Ca 2ϩ sparks and Ca 2ϩ waves) (40,55,57). It is thought that SCR necessitates and facilitates the occurrence of Ca 2ϩ alternans (25,28,38,(41)(42)(43)(44)(45)(46). Hence, it is of interest to determine whether SCR is involved in the occurrence of Ca 2ϩ alternans in intact working RyR2 R4496C mutant hearts. Fig. 3 shows Ca 2ϩ transients in intact RyR2-R4496C hearts continuously stimulated with increasing frequencies from 6 to 14 Hz ( Fig. 3 and Fig. S1). Surprisingly, despite the enhanced RyR2 function, intact working RyR2 R4496C mutant hearts exhibited no detectable spontaneous Ca 2ϩ sparks or Ca 2ϩ waves before or after the occurrence of Ca 2ϩ alternans. The continued high-frequency stimulations probably override SCR in the intact working RyR2-R4496C hearts. Consistent with this view, spontaneous Ca 2ϩ waves were readily observed in intact RyR2-R4496C mutant hearts, but not in WT hearts, after the cessation of pacing in the presence of high adrenergic stress (1 M epinephrine plus 0.6 mM caffeine) (Fig. 4). It is important to note that the same condition (1 M epinephrine plus 0.6 mM caffeine) also induced VTs in intact working RyR2-R4496C hearts as reported previously (47). Hence, the Ca 2ϩ alternans observed in rapidly stimulated intact RyR2 R4496C mutant hearts are unrelated to spontaneous Ca 2ϩ sparks or Ca 2ϩ waves.

Enhancing SERCA2a function by phospholamban knockout (PLB-KO) has a relatively small impact on Ca 2؉ alternans in intact RyR2 WT or E4872Q mutant hearts
It is clear that modulating the activity of RyR2 has a major impact on Ca 2ϩ alternans (26 -30, 32). However, the relative impact of modulating the activity of SERCA2a, another key component of SR Ca 2ϩ cycling, on Ca 2ϩ alternans is unclear. To this end, we assessed the impact of PLB-KO on Ca 2ϩ alternans in intact RyR2 WT or E4872Q mutant hearts. The RyR2 E4872Q mutation has been shown to suppress Ca 2ϩ activation of RyR2 and the occurrence of spontaneous Ca 2ϩ waves (32,58,59). As expected, PLB-KO markedly increased the amplitude of Ca 2ϩ transients and reduced the transient decay time (T 50 ) as compared with WT hearts (Fig. S2), consistent with its stimulatory action on SERCA2a and SR Ca 2ϩ reuptake (60). Surprisingly, PLB-KO did not significantly alter the average alternans ratios at stimulation frequencies from 5 to 14 Hz (Fig. 5). PLB-KO significantly reduced the average alternans durations only at stimulation frequencies of 10, 11, and 12 Hz (Fig. 5D). Thus, compared with enhancing RyR2 function, enhancing SERCA2a function as a result of PLB-KO has relatively small impact on Ca 2ϩ alternans in intact WT hearts.
We also assessed whether PLB-KO could rescue the enhanced Ca 2ϩ alternans in RyR2 E4872Q mutant hearts with suppressed RyR2 function. Similar to what was observed in RyR2 WT hearts (Fig. S2), PLB-KO significantly increased the amplitude of Ca 2ϩ transients and reduced the T 50 in intact RyR2 E4872Q mutant hearts (Fig. S3). PLB-KO also significantly increased the SR Ca 2ϩ content in WT ventricular myocytes, as expected, and dramatically increased the SR Ca 2ϩ content in RyR2 E4872Q mutant cells (Fig. S4). However, PLB-KO had no significant impact on the average alternans ratio of RyR2 E4872Q hearts at stimulation frequencies from 5 to 14 Hz (Fig.  6). PLB-KO significantly reduced the average alternans duration only at the stimulation frequency of 9 Hz (Fig. 6D). Thus, enhancing SERCA2a function by PLB-KO does not suppress the enhanced Ca 2ϩ alternans in RyR2 E4872Q mutant hearts. These observations also suggest that, although the SR Ca 2ϩ content is an important regulator of SR Ca 2ϩ handling, it does not seem to play a critical role in Ca 2ϩ alternans.

Effect of SERCA2a inhibition on Ca 2؉ alternans in intact hearts
We next assessed the effect of 2,5-di-tert-butylhydroquinone (tBHQ), an inhibitor of SERCA2a, on Ca 2ϩ alternans in intact RyR2 WT and E4872Q mutant hearts. As expected, tBHQ at 3 and 10 M significantly reduced the amplitude of Ca 2ϩ transients. It also prolonged the T 50 in RyR2 WT hearts by 13% (at 3 M) and 16% (at 10 M) (Fig. S5). This is consistent with the inhibitory action of tBHQ on SERCA2a and SR Ca 2ϩ reuptake. However, despite its significant impact on SERCA2a, tBHQ at 3 M did not significantly affect the average alternans ratio or duration in intact RyR2 WT hearts stimulated at a wide range of frequencies (from 5 to 14 Hz) (Fig. 7). Surprisingly, tBHQ at 10 M markedly reduced both the average alternans ratio and duration in intact RyR2 WT hearts at stimulation frequencies of 11-14 Hz (Fig. 7). These data indicate that, depending on the extent of SERCA2a inhibition, reducing hearts were loaded with Rhod-2 AM. Hearts were first stimulated at 5 Hz for 30 beats (S1), followed by a single S2 stimulation. A series of S1S2 stimulations were repeatedly applied with progressively reduced S1S2 intervals from 200 to 40 ms. Ca 2ϩ transients were recorded using line-scanning confocal imaging. C, the relationship between A2/A1 ratio of the Ca 2ϩ transient amplitude and S1S2 interval is shown. Data shown are mean Ϯ S.D. (error bars) (n ϭ 5 hearts for WT, n ϭ 7 hearts for R4496C) (*, p Ͻ 0.05).

Role of RyR2 and SERCA2a in Ca 2؉ alternans
SERCA2a function either has a relatively minor impact on Ca 2ϩ alternans or can lead to marked suppression of Ca 2ϩ alternans.
We also assessed the impact of tBHQ on Ca 2ϩ alternans in intact RyR2 E4872Q mutant hearts (32,58). Similar to those observed in RyR2 WT hearts, tBHQ at 3 and 10 M significantly reduced the amplitude of Ca 2ϩ transients. It also prolonged the T 50 of Ca 2ϩ transients by 8% (at 3 M) and 16% (at 10 M) (Fig.  S6). As with intact WT hearts, tBHQ at 10 M also significantly reduced the average alternans ratio and duration in intact RyR2 E4872Q mutant hearts (Fig. 8). On the other hand, tBHQ at 3 M had a minor effect on the average alternans ratio or duration in intact E4872Q hearts (Fig. 8). Collectively, these data indicate that compared with RyR2, SERCA2a plays a relatively minor role in Ca 2ϩ alternans in intact hearts.

Discussion
Beat-to-beat alternations in the amplitude of the cytosolic Ca 2ϩ transient (Ca 2ϩ alternans) are thought to be the primary cause of cardiac alternans (1-11), which is a major risk factor for ventricular arrhythmias and sudden cardiac arrest (12)(13)(14)(15)(16)(17)(18)(19). Despite its important role in arrhythmogenesis, the molecular mechanism underlying Ca 2ϩ alternans remains undefined. An increased body of evidence suggests that Ca 2ϩ alternans results from abnormal SR Ca 2ϩ cycling (9, 11, 21-25). Because SR Ca 2ϩ cycling is governed by SR Ca 2ϩ release via RyR2 and SR Ca 2ϩ reuptake by SERCA2a (20), altered RyR2 or SERCA2a function would be expected to affect SR Ca 2ϩ cycling, thus leading to Ca 2ϩ alternans. However, how changes in the activity of RyR2 or SERCA2a affect Ca 2ϩ alternans is unclear. To address this question, here we determined the impact of genetically or pharmacologically enhancing or suppressing RyR2 or SERCA2a function on Ca 2ϩ alternans in intact working hearts. We found that altering RyR2 function, but not SERCA2a function, has a major impact on Ca 2ϩ alternans. These findings shed new insights into the molecular mechanism of Ca 2ϩ alternans and have important therapeutic implications for cardiac alternans.
Recent studies have consistently shown that suppressing the function of RyR2 prolongs the refractoriness of SR Ca 2ϩ release and promotes Ca 2ϩ alternans (26 -30, 32). However, the impact of enhanced RyR2 function on Ca 2ϩ alternans is unclear. On the one hand, enhancing RyR2 function would increase spontaneous Ca 2ϩ release (Ca 2ϩ sparks/Ca 2ϩ waves), which would promote Ca 2ϩ alternans (25,28,38,(41)(42)(43)(44)(45)(46). On the other hand, enhancing RyR2 function would shorten SR Ca 2ϩ release refractoriness, which would suppress Ca 2ϩ alternans (10,31,32,34,35,39). To ascertain these seemingly paradoxical effects of enhanced RyR2 function on Ca 2ϩ alternans, we determined the impact of a disease-causing RyR2 mutation (R4496C) with enhanced channel activity on Ca 2ϩ alternans in the setting of intact working hearts. We found that, despite the enhanced RyR2 activity, intact working RyR2 R4496C mutant hearts displayed little or no spontaneous Ca 2ϩ sparks or waves during electrical stimulation, similar to what was reported previously (47). This is also consistent with the observation that increased heart rate alone (as in programmed electrical stimulation) can rarely trigger VTs in patients with CPVT (61). In contrast, accelerating heart rate (in the absence of excessive adrenergic stress) suppresses spontaneous Ca 2ϩ release and prevents VTs in both CPVT animal models and patients (62). Furthermore, we found that, in the absence of Ca 2ϩ sparks/waves, enhancing RyR2 function shortens SR Ca 2ϩ release refractoriness and suppresses Ca 2ϩ alternans. Hence, contrary to depressed RyR2 function, which promotes Ca 2ϩ release refractoriness and Ca 2ϩ alternans and suppresses stress-provoked CPVT, enhanced RyR2 function protects against Ca 2ϩ alternans but promotes CPVT. These observations suggest that the mechanisms underlying stress-provoked CPVT and Ca 2ϩ alternans are different. It is of interest to note that, contrary to our findings, enhanced

Role of RyR2 and SERCA2a in Ca 2؉ alternans
RyR2 function has been shown to promote Ca 2ϩ alternans in isolated cardiomyocytes where spontaneous Ca 2ϩ release is present (36, 38,40). These observations suggest that the presence or absence of spontaneous SR Ca 2ϩ release may influence whether enhanced RyR2 function will promote or suppress Ca 2ϩ alternans and that the nature and mechanisms of Ca 2ϩ alternans with or without spontaneous SR Ca 2ϩ release may be different. Further studies are needed to fully understand the role of spontaneous Ca 2ϩ release in the genesis of Ca 2ϩ alternans.

Role of RyR2 and SERCA2a in Ca 2؉ alternans
The role of SERCA2a in Ca 2ϩ alternans is complex. On the one hand, reduced SERCA2a function would decrease SR Ca 2ϩ content and thus SR Ca 2ϩ release, which would suppress Ca 2ϩ alternans. On the other hand, reduced SERCA2a function would prolong Ca 2ϩ transient decay and elevate cytosolic Ca 2ϩ concentration, which would promote Ca 2ϩ alternans (25,31,46,52). Similarly, enhanced SERCA2a function would increase SR Ca 2ϩ content and thus SR Ca 2ϩ release, which would promote Ca 2ϩ alternans. In contrast, enhanced SERCA2a function would hasten Ca 2ϩ transient decay and reduce cytosolic Ca 2ϩ The average alternans ratio (C) and average alternans duration (D) intact RyR2-E4872Q and PLB-KO/EQ at different stimulation frequencies are shown. Data shown are mean Ϯ S.D. (error bars) (n ϭ 5 hearts for E4872Q, n ϭ 8 hearts for PLB-KO/EQ); two-way ANOVA with a Bonferroni's post hoc test (*, p Ͻ 0.05). For the analysis of alternans ratios, the F statistics are as follows: for the row factor (pacing frequency), F ϭ 54.76, p Ͻ 0.01; for the column factor (genotypes), F ϭ 13.26, p Ͻ 0.01; interaction between the column and row factors, F ϭ 1.276; p ϭ 0.257. For the analysis of alternans durations, the F statistics are as follows: for the Row factor (pacing frequency), F ϭ 38.66, p Ͻ 0.01; for the column factor (genotypes), F ϭ 13.874, p Ͻ 0.01; interaction between the column and row factors, F ϭ 0.726; p ϭ 0.6834.

Role of RyR2 and SERCA2a in Ca 2؉ alternans
concentration, which would suppress Ca 2ϩ alternans (25,31,46,52). Thus, changes in the activity of SERCA2a would promote or suppress Ca 2ϩ alternans, depending on the relative changes in the cytosolic Ca 2ϩ concentration and the SR Ca 2ϩ content. Furthermore, because the activity of SERCA2a oppositely affects the cytosolic Ca 2ϩ concentration and SR Ca 2ϩ content, changes in the SERCA2a activity would be expected to have only a minor impact on Ca 2ϩ alternans due to the resultant opposite changes in the cytosolic Ca 2ϩ concentration and SR Ca 2ϩ content. Indeed, consistent with this view, we found that genetically enhancing the SERCA2a function by PLB-KO significantly increased the amplitude of SR Ca 2ϩ release, which would promote Ca 2ϩ alternans, but decreased the decay time of Ca 2ϩ transients, which would suppress Ca 2ϩ alternans. As a result of these opposing effects, PLB-KO did not markedly alter Ca 2ϩ alternans in intact WT hearts. We also found that PLB-KO did not rescue the enhanced Ca 2ϩ alternans in intact RyR2 E4872Q mutant hearts.
We also investigated the impact of reduced SERCA2a function on Ca 2ϩ alternans. Partially reducing SERCA2a activity using a low concentration of tBHQ (3 M) reduced the amplitude of SR Ca 2ϩ release, which would suppress Ca 2ϩ alternans, but increased the decay time of Ca 2ϩ transients, which would promote Ca 2ϩ alternans. As a result, these opposing actions of tBHQ in the amplitude and decay time of Ca 2ϩ transients led to no marked alteration in Ca 2ϩ alternans in intact WT hearts. We also found that tBHQ (3 M) had no major impact on the enhanced Ca 2ϩ alternans in intact RyR2 E4872Q mutant hearts. Interestingly, further inhibition of SERCA2a activity using a higher concentration of tBHQ (10 M) significantly suppressed Ca 2ϩ alternans in intact WT or E4872Q mutant hearts. This suppression on Ca 2ϩ alternans probably resulted from the (error bars) (n ϭ 11 hearts before tBHQ treatment, n ϭ 8 hearts after 3 M tBHQ, n ϭ 6 hearts after 10 M tBHQ); two-way ANOVA with a Dunnett's post hoc test (**, p Ͻ 0.01). For the analysis of alternans ratios, the F statistics are as follows: for the row factor (pacing frequency), F ϭ 153.906, p Ͻ 0.01; for the column factor (tBHQ treatment), F ϭ 108.024, p Ͻ 0.01; interaction between the column and row factors, F ϭ 17.8275; p Ͻ 0.01. For the analysis of alternans durations, the F statistics are as follows: for the row factor (pacing frequency), F ϭ 84.08, p Ͻ 0.01; for the column factor (tBHQ treatment), F ϭ 86.24, p Ͻ 0.01; interaction between the column and row factors, F ϭ 7.386; p Ͻ 0.01.

Role of RyR2 and SERCA2a in Ca 2؉ alternans
stronger effect of 10 M tBHQ on the reduction in SR Ca 2ϩ content, as a result of stronger inhibition of SERCA2a activity, than on the increase in Ca 2ϩ transient decay. Taken together, our findings indicate that moderate changes in the SERCA2a function have a relatively minor impact on Ca 2ϩ alternans in intact working hearts. However, it is important to note that altered SERCA2a function also affects the propensity for spontaneous Ca 2ϩ release. Thus, changes in SERCA2a function may play an important role in Ca 2ϩ alternans in the setting of disease hearts where spontaneous Ca 2ϩ release is enhanced.
Although genetically engineered mouse models harboring a RyR2 GOF or loss-of-function mutation or a PLB deletion allow us to determine the effect of specifically reducing or enhancing RyR2 or SERCA2a function on Ca 2ϩ alternans, whether our findings from these mouse hearts could be translated into the human hearts is unclear. It is known that intracellular Ca 2ϩ handling and electrophysiological properties of the mouse hearts are substantially different from those of the human hearts. Hence, the significance and relative contribution of RyR2 and SERCA2a function to Ca 2ϩ alternans in human hearts has yet to be determined.
In summary, the present study demonstrates for the first time that genetically enhancing RyR2 function shortens SR Ca 2ϩ release refractoriness and protects against Ca 2ϩ alternans in intact working hearts. On the other hand, enhancing or suppressing SERCA2a function has a relatively minor impact on Ca 2ϩ alternans in intact working hearts. These findings indicate that the activity of RyR2, but not SERCA2a, is a major determinant of Ca 2ϩ alternans. Thus, RyR2 represents a promising therapeutic target for cardiac alternans.

Determination of refractoriness of SR Ca 2؉ release
The refractoriness of voltage-induced release of Ca 2ϩ from the SR was determined by using the S1S2 stimulation protocol as described previously with some modifications (10,32). Briefly, Ca 2ϩ transients in Rhod-2 AM-loaded hearts were first induced at 5 Hz for 5 s (S1), followed by a single S2 stimulation at a specific interval. The hearts were repeatedly stimulated by a series of S1S2 protocols with progressively decreased S1S2 intervals (from 200 to 40 ms). Ca 2ϩ transients before and after S2 stimulation were continuously recorded by using the Nikon-A1R confocal microscope in the line-scan mode.

Laser-scanning confocal Ca 2؉ imaging of intact hearts
WT and mutant mice were sacrificed by cervical dislocation. Their hearts were quickly removed and loaded with 4.4 M Rhod-2 AM (Biotium, Inc., Hayward, CA) in oxygenated Tyrode's buffer (118 mM NaCl, 5.4 mM KCl, 25 mM NaHCO 3 , 1 mM MgCl 2 , 0.42 mM NaH 2 PO 4 , 11.1 mM glucose, 10 mM taurine, 5 mM creatine, and 1.8 mM CaCl 2 , pH 7.4) via the retrograde Langendorff perfusion system at 25°C for 45 min (47,63). The Langendorff-perfused hearts were placed in a recording chamber mounted onto the Nikon A1R microscope for in situ confocal imaging (line-scan) of Ca 2ϩ signals from epicardial ventricular myocytes. The temperature of the heart was kept at 35°C throughout the experiment with 5 M blebbistatin (Toronto Research Chemicals, Toronto, Canada) to prevent motion artifact. The pixel size of the resulting line-scan images ranged between 1.8 and 2 ms in the temporal dimension and between 0.1 and 0.4 m in the spatial dimension. Ca 2ϩ alternans in the WT and mutant hearts in the absence or presence of tBHQ (3 or 10 M) was induced by rapid electrical stimulation at increasing frequencies (5-14 Hz, 6 V).

Image and signal processing
The signal and image processing methods were implemented using MATLAB (The Mathworks Inc., Boston, MA) as described previously (32). Briefly, line-scan fluorescence images were filtered according to the noise level estimated by the median absolute deviation of the pixel intensities. Individual cells in the images were manually marked, and the average fluorescence in each cell was obtained for further analysis. A wavelet peak detection algorithm was used to detect individual calcium release events in the average fluorescence signals. For each event detected in each cell, we determined the peak amplitude (local minimum-maximum difference) and the alternans ratio (relative amplitude difference between consecutive peaks). The presence of alternans periods was established when six consecutive peaks presented an alternans ratio above 0.05. Alternans duration was defined as the percentage of alternans periods over the total line-scan duration. Average magnitudes were obtained by taking the mean over each line scan.

Statistical analysis
GraphPad Prism version 6.0 was used for statistical analyses. All values shown are mean Ϯ S.D. unless indicated otherwise. To test for differences between groups, we used Student's t test (two-tailed) or one-or two-way ANOVA with Dunnett's or Bonferroni's post hoc test when appropriate. A p value Ͻ0.05 was considered to be statistically significant.