Intact β-Adrenergic Response and Unmodified Progression Toward Heart Failure in Mice With Genetic Ablation of a Major Protein Kinase A Phosphorylation Site in the Cardiac Ryanodine Receptor
Increased phosphorylation of the cardiac ryanodine receptor (RyR)2 by protein kinase A (PKA) at the phosphoepitope encompassing Ser2808 has been advanced as a central mechanism in the pathogenesis of cardiac arrhythmias and heart failure. In this scheme, persistent activation of the sympathetic system during chronic stress leads to PKA “hyperphosphorylation” of RyR2-S2808, which increases Ca2+ release by augmenting the sensitivity of the RyR2 channel to diastolic Ca2+. This gain-of-function is postulated to occur with the unique participation of RyR2-S2808, and other potential PKA phosphorylation sites have been discarded. Although it is clear that RyR2 is among the first proteins in the heart to be phosphorylated by β-adrenergic stimulation, the functional impact of phosphorylation in excitation–contraction coupling and cardiac performance remains unclear. We used gene targeting to produce a mouse model with complete ablation of the RyR2-S2808 phosphorylation site (RyR2-S2808A). Whole-heart and isolated cardiomyocyte experiments were performed to test the role of β-adrenergic stimulation and PKA phosphorylation of Ser2808 in heart failure progression and cellular Ca2+ handling. We found that the RyR2-S2808A mutation does not alter the β-adrenergic response, leaves cellular function almost unchanged, and offers no significant protection in the maladaptive cardiac remodeling induced by chronic stress. Moreover, the RyR2-S2808A mutation appears to modify single-channel activity, although modestly and only at activating [Ca2+]. Taken together, these results reveal some of the most important effects of PKA phosphorylation of RyR2 but do not support a major role for RyR2-S2808 phosphorylation in the pathogenesis of cardiac dysfunction and failure.
In cardiac muscle, Ca2+ release channels/ryanodine receptors (RyR2) of sarcoplasmic reticulum (SR) are key players in excitation–contraction coupling, a finely orchestrated series of events that link electrical stimuli with mechanical contractions. Depolarization of the external membrane opens L-type Ca2+ channels, allowing a small inward Ca2+ current (ICa) that triggers massive Ca2+ release from the SR. This Ca2+ amplification process, Ca2+-induced Ca2+ release, is mediated by RyR2s and provides the majority of Ca2+ that myofilaments need for contraction. Relaxation occurs when Ca2+ detaches from the myofilaments and is cycled back into the SR by the work of the Ca2+-ATPase, a process that is regulated by phospholamban (PLB).1
Increased metabolic demand during periods of stress or exercise triggers release of catecholamines, which stimulate β-adrenergic receptors in the heart. Activation of the β-adrenergic receptor pathway triggers a cascade of events that increase cAMP, which in turn activates protein kinase (PK)A. PKA then phosphorylates several target proteins in the sarcolemma, the SR, and myofilaments, notably, L-type Ca2+ channels, PLB, RyR2, and troponin I and C, leading to an increased Ca2+ entry, enhanced Ca2+-induced Ca2+ release, and faster Ca2+ uptake and relaxation rates, all of which contribute to the inotropic and lusitropic effects of β-adrenergic stimulation on the heartbeat.1–3
Despite the fact that RyR2 are among the first proteins to undergo phosphorylation by β-adrenergic receptor stimulation of cardiac muscle,4,5 the functional impact of this reaction remains unclear. On the one hand, Li et al6 have found that PKA phosphorylation of RyR2 has little functional relevance for diastolic Ca2+ release if SR Ca2+ levels remain constant. On the other hand, Marks et al have postulated that PKA phosphorylation of RyR2 is so consequential for intracellular Ca2+ homeostasis that derangement of this process may be the basis for some forms of cardiac arrhythmias7 and heart failure (HF).8–10 In between these 2 extremes, other results, mainly from in vitro experiments, imply that PKA phosphorylation increases,4,11 decreases,12–13 or has no effect14 on RyR2 activity.
The variable and apparently contradicting results of PKA phosphorylation may be attributable to the presence of multiple phosphorylation sites on the RyR2 protein4,15,16 that may influence channel activity differently. However, Wehrens et al10 have reported that PKA phosphorylates mouse RyR2 at a single and unique residue, Ser2808 (corresponding to Ser2809 in human RyR2) and that genetic ablation of this phosphorylation site prevents PKA activation of RyR2 and protects hearts from the maladaptive cardiac remodeling that follows myocardial infarction. In overt HF, Marks and colleagues have detected “hyperphosphorylation” of Ser2809 in canine and human hearts,8–10 postulating that this reaction is the basis for the increased Ca2+ leak and decreased SR Ca2+ content that are characteristic of HF, but these findings were not confirmed by others.6,14,16,17 Thus, the role of PKA phosphorylation, and especially of Ser2808 phosphorylation, remains unclear.
We generated mice with genetic ablation of the Ser2808 phosphoepitope (RyR2-S2808A) and used whole-heart, isolated cardiomyocyte and single-channel experiments to test the role of β-adrenergic stimulation and PKA phosphorylation of Ser2808 in HF progression, cellular Ca2+ handling, and RyR2 activity. We found that the RyR2-S2808A mutation modifies single-channel activity, albeit modestly and at activating [Ca2+] only, allows for normal β-adrenergic stimulation of [Ca2+]i transients, leaves cellular function almost unaltered, and offers no significant protection in the maladaptive cardiac remodeling induced by chronic stress.
Materials and Methods
RyR2-S2808A mice and age-matched wild-type (WT) littermates were maintained and studied according to the protocol approved by the Institutional Animal Care and Use Committees of the University of Wisconsin Madison and by the Association for Assessment and Accreditation of Laboratory Care International.
Generation of the RyR2-S2808A Mice
Mice with the RyR2-S2808A mutation were generated by homologous recombination.
An expanded Materials and Methods section is available in the online data supplement at http://circres.ahajournals.org.
Generation of RyR2-S2808A Mice and Characterization With Phospho-Specific Antibody
We used homologous recombination to generate mice with total ablation of the RyR2-S2808 phosphoepitope by replacing Ser2808 with the nonphosphorylatable residue Ala (Figure 1A). The resultant RyR2-S2808A mice developed normally, displayed no overt phenotypic differences compared with their WT littermates, and propagated at the expected Mendelian frequencies (61 WT [24.1%], 127 heterozygotes [50.2%], and 65 homozygotes [25.7%]). Southern blot (Figure 1B), PCR (Figure 1C), and immunocytochemistry (Figure 1D) experiments confirmed that the RyR2-S2808A substitution was present in hearts of the mutant mice. Western blots using the phospho-specific antibody RyR2-pS2808 (ref. 15) displayed a strong band in WT but no signal in RyR2-S2808A SR-enriched microsomes (Figure 1E).
Echocardiographic Analysis of WT and RyR2-S2808A Mice
Hearts of RyR2-S2808A mice were structurally normal in histologic (Figure 2A) and echocardiographic (Figure 2B) analyses. However, following 4 weeks of aortic constriction, both RyR2-S2808A mice and their WT littermates exhibited evident cardiac remodeling and progressive systolic dysfunction. Cardiac hypertrophy was apparent in explanted banded hearts (Figure 2A), and the ratio of left ventricular mass/body weight increased to a greater extent in WT (82±10%) than in RyR2-S2808A (61±13%) mice (n=6 to 7), but the difference did not reach statistical significance (P=0.25). A similar trend toward cardioprotection was noted in the progression of RyR2-S2808A mice toward systolic dysfunction, both in the functional index (0.77±0.1 in WT versus 0.85±0.05 in RyR2-S2808A) and the stroke volume (25.8±9 μL in WT versus 34.1±9.7 μL in RyR2-S2808A). However, like the structural parameters, functional differences were modest and lacked statistical significance (see other echocardiographic parameters in Table I in the online data supplement). Thus, ablation of the RyR2-S2808 phosphorylation site does not appear to impact significantly the maladaptive cardiac remodeling that follows chronic stress.
S2808 Is Not the Only PKA Site in RyR2
Wehrens et al10 reported that RyR2 immunoprecipitated from RyR2-S2808A mice could not be phosphorylated by PKA and postulated that S2808 is the only PKA site in the entire RyR2 protein. Here we show that immunoprecipitated RyR2-S2808A protein (Figure 3A) was phosphorylated by the catalytic subunit of PKA. WT and RyR2-S2808A proteins were either untreated or pretreated with alkaline phosphatase (AkPh) before incubation with PKA and [γ-32P]ATP to assess the level of basal phosphorylation; values were then normalized to that of maximally dephosphorylated (AkPh-treated) WT-RyR2 (100%). Samples incubated with the phosphorylation cocktail minus PKA showed no incorporation of radioactivity (Figure 3A; −PKA, 0%), indicating that no endogenous kinases participated in this reaction and that all phosphorylation was PKA specific. Untreated WT-RyR2 incorporated 71±7% of the total [γ-32P]ATP, indicating that only ≈30% of PKA phosphorylation sites were basally phosphorylated. Probing with phospho-specific antibodies (Figure 3C) reveals that S2808 is mostly responsible for this basal phosphorylation; S2030, another PKA-specific site,16 appears completely dephosphorylated under basal conditions. In comparison, untreated RyR2-S2808A could be phosphorylated to 44±5% of WT-RyR2 (Figure 3A) and this value changed only slightly by pretreatment with AkPh (36±3%), suggesting that more than half of the total PKA phosphorylation sites are available for phosphorylation and that they are mostly dephosphorylated under basal conditions. Indeed, Figure 3C shows that S2030 is a phosphoepitope still available for phosphorylation in RyR2-S2808A and is likely responsible for the incorporation of [γ-32P]ATP detected in the autoradiogram. These complementary approaches strongly suggest that PKA phosphorylates at least 2 sites in WT-RyR2, S2808 and S2030, which are highly and barely phosphorylated, respectively, under basal conditions in mouse hearts (Figure 3B).16,17,18
In Vivo Phosphorylation of RyR2
Next, we assessed the in vivo phosphorylation state of the 3 known phosphoepitopes of RyR2 and their response to β-adrenergic stimulation. Figure 4A illustrates the general protocol. Explanted hearts were retrogradely perfused in a Langendorff apparatus, in which a pressure transducer in the tubing carrying perfusate recorded the relative strength of the heartbeat as hearts were perfused with an external solution containing or lacking 1 μmol/L isoproterenol (Iso) and 2 mmol/L Ca2+. In the presence of Ca2+ (Figure 4A, left), Iso readily exerted a strong inotropic effect, as noted by the increase in the output pressure of heart contractions; in the absence of Ca2+ (right), all contractions were abated, as expected, but perfusion continued nonetheless. After 20 minutes, total perfusion time (marked by asterisk), hearts were flash frozen in liquid nitrogen, pulverized, and homogenized in the presence of phosphatase inhibitors in preparation for Western blots.
Figure 4B shows that Iso perfusion activated the β-adrenergic pathway and increased PKA activity as the phosphorylation of PLB-Ser16, a specific PKA-phosphorylated residue, yielded a strong signal in the absence and presence of 2 mmol/L Ca2+. By contrast, phosphorylation of PLB-Thr17 occurred only in the presence of Ca2+ and Iso, supporting the notion that the Ca2+- and calmodulin-dependent Ca2+/calmodulin-dependent protein kinase (CaMK)II, which specifically phosphorylates this residue, is secondarily activated by β-adrenergic stimulation.19 RyR2-S2808 was again found phosphorylated in WT hearts, as Iso perfusion increased band intensity <2-fold, both in the presence and absence of Ca2+. By contrast, in the same WT hearts, Iso increased the level of RyR2-S2030 phosphorylation ≈7- and ≈3-fold in the presence and absence of Ca2+, respectively, clearly indicating that this phosphoepitope is markedly sensitive to β-adrenergic stimulation and a Ca2+-stimulated kinase (or a Ca2+-inhibited phosphatase). The CaMKII-specific RyR2-S2815 was weakly phosphorylated under basal conditions and, oddly, Iso and Ca2+ did not increase phosphorylation substantially, as would be expected from the vigorous activation of CaMKII noted by PLB-pT17.
Figure 4C displays the average response to β-adrenergic stimulation of each RyR2 phosphoepitope in WT and RyR2-S2808A hearts, normalized to its own control in the absence of Iso and corrected by the anti-RyR2 signal obtained from the same blot. In general, β-adrenergic stimulation increased phosphorylation of all 3 sites in WT hearts, but the participation of S2030 was the greatest. In RyR2-S2808A hearts, Iso increased the S2030 and S2815 phospho-signals more intensely than in WT hearts, perhaps as a compensatory mechanism for the lack of S2808 phosphorylation.
Normal Ca2+ Transients and β-Adrenergic Response in RyR2-S2808A–Isolated Cardiomyocytes
β-Adrenergic stimulation of cardiac cells results in RyR phosphorylation (Figure 4) and increased Ca2+-induced Ca2+ release,1–3 but the modulatory role of PKA phosphorylation on RyRs remains debated. We measured the amplitude and kinetics of field-stimulated [Ca2+]i transients and cell shortenings in isolated ventricular myocytes from WT and RyR2-S2808A mice, both under control conditions (basal) and after β-adrenergic stimulation (+1 μmol/L Iso) of the same cells. Figure 5 shows that, except for the slower time constant of decay (τ) of the [Ca2+]i transient at 1 and 2 Hz in RyR2-S2808A cells, all parameters were similar for the 2 populations of cells. In the presence of Iso, the amplitude and kinetics of the [Ca2+]i transient as well as the magnitude of cell contraction increased, as expected, but there were no differences between the 2 groups. We also determined the size of the caffeine-releasable Ca2+ pool at various pacing frequencies to estimate the amount of SR Ca2+ load, but, again, there were no differences between WT and RyR2-S2808A cells (see supplemental Figure I). These results suggest that ablation of the RyR2-S2808 phosphorylation site has limited effects on the global [Ca2+]i transient and subsequent contraction of ventricular myocytes. The Iso normalization of all measured parameters in RyR2-S2808A cells implies that the combined functional output of all PKA-stimulated proteins obscures any modest effect resulting from the ablation of one of several phosphorylation sites in the RyR2 protein.
Ca2+ Sparks and Wavelets in Permeabilized WT and RyR2-S2808A Cells
Ca2+ sparks represent the coordinated opening of a cluster of RyR2s gating in situ20 and may be used as direct indicators of the effect of modulators on RyR2 activity. Previous experiments have shown that perfusion of cAMP onto permeabilized mouse ventricular cells activates membrane-bound PKA, which in turn phosphorylates SR proteins and increases Ca2+ spark activity.6 Figure 6A shows Ca2+ spark images obtained from permeabilized WT and RyR2-S2808A cardiomyocytes before and after the addition of 10 μmol/L cAMP while in the presence of the CaMKII inhibitor KN-93 (1 μmol/L). To prevent dephosphorylation by endogenous phosphatases, okadaic acid (5 μmol/L) was added to the perfusate. cAMP promptly increased Ca2 + release, turning discrete Ca2+ release events into coalescing and laterally propagating wavelets that hampered quantification of isolated Ca2+ sparks. We thus obtained the integral of the amplitude (ΔF/F0) multiplied by the full width (in micrometers) and full duration (in milliseconds) of all Ca2+ release events (sparks and wavelets) and plotted results against the same parameter obtained before cAMP addition; each cell, then, was used as its own control. Figure 6B shows that cAMP vigorously increased Ca2+ release in WT and RyR2-S2808A cells >200% at all time points, yielding a significant difference against Ca2+ release events in the absence of cAMP (*P<0.05; **P<0.001), but no difference was seen between WT and RyR2-S2808A cells at any time point. Thus, these results suggest that the RyR2-S2808 phosphoepitope is not a critical determinant of Ca2+ spark activity, at least at the quasidiastolic (50 nmol/L) [Ca2+] used here.
Direct Recording of WT and RyR2-S2808A Channel Activity in Lipid Bilayers
Ca2+ sparks and spontaneous Ca2+ release are meaningful assays of RyR2 activity, but their interpretation is complicated by uncontrolled luminal and cytosolic factors that also influence channel activity. We reconstituted SR vesicles into planar lipid bilayers and recorded single-channel activity under controlled conditions to directly assess the effect of phosphorylation on isolated RyR2s.12,17 SR vesicles were first treated for 10 minutes with 10 U/mL protein phosphatase (PP)1 to remove any residual phosphorylation present in RyR2 channels. Using Western blots, we provide evidence that the RyR2 channels were dephosphorylated by PP1 at S2808 and S2030 under these starting conditions (Figure 7A). Dephosphorylated WT and RyR2-S2808A channels exhibited robust channel activity at cis (corresponding to cytosolic side) [Ca2+]=5 μmol/L (“Control, pCa 5.3”), as shown in the representative traces of Figure 7B. On average, the probability of the channel being open (Po) was similar for WT and RyR2-S2808A channels (0.41±0.12 [n=15] and 0.39±0.08 [n=13], respectively), as expected if the phosphorylation state were identical in both types of channels. The addition of 2 mmol/L MgATP (0.35 mmol/L free Mg2+) decreased channel activity (0.12±0.06 and 0.14±0.04 for WT and RyR2-S2808A, respectively), also as expected from the competitive effect of Mg2+ on the Ca2+-activation site.21 Decreasing cytosolic [Ca2+] to 100 nmol/L (pCa 7) deactivated channels almost completely (Po<0.0001 in both groups), with only brief openings being detected sporadically. Addition of 1 μg/mL the catalytic subunit of PKA failed to increase Po in WT and RyR2-S2808A channels (Po<0.0001); however, after the return of cytosolic [Ca2+] to prephosphorylation levels (pCa 5.3), WT channels displayed consistently lower Po (0.08±0.03) than RyR2-S2808A channels (0.22±0.05) (last traces of Figure 7B). The latter was the only difference between WT and RyR2-S2808A channels more than the course of the experiment, represented in Figure 7C and 7D. Again, using Western blots, we show that PKA effectively phosphorylates both S2808 and S2030 sites under the conditions used in the bilayer experiments (Figure 7A, “pCa7, MgATP, PKA”). We therefore assign the difference in WT and RyR2-S2808 channel activity at pCa 5.3 to the phosphorylation of S2808 in WT channels. The inhibition at high [Ca2+] caused by PKA phosphorylation of WT channels is consistent with our previous report which suggested that PKA increased the transient peak of RyR2 activity but decreased its steady-state activity (see Discussion).12
In this study, we demonstrate that stimulation of β-adrenergic receptors by Iso in whole hearts leads to phosphorylation of several RyR2 phosphoepitopes (Figure 4). RyR2 channels, therefore, appear to be integral components of the multiprotein response that leads to increased cardiac performance during β-adrenergic stimulation, but the extent of their participation in the “fight-or-flight” response remains debated. Ca2+ release by RyR2 is steeply graded by ICa and luminal Ca2+, both of which increase during β-adrenergic stimulation because of PKA phosphorylation of L-type Ca2+ channels and PLB, respectively1; hence, it appears that RyR2 channels need only be responsive to external cues to increase Ca2+ release, but this passive compliance negates the necessity for direct RyR2 protein phosphorylation. Some outstanding questions, therefore, are the following. What are the intrinsic changes in RyR2 function brought about by PKA phosphorylation? Is this functional modulation conspicuous enough to modify global Ca2+ transients, cardiac performance, and be a substrate for development of cardiac arrhythmias and HF? In an attempt to answer these questions, we generated a mouse with ablation of a major PKA phosphorylation site (RyR2-S2808A) and found that RyR2 phosphorylation results in discernible effects on single-channel activity that are obscured by a multitude of cellular adjustments in response to sympathetic stimulation. Aortic-banded RyR2-S2808A and WT mice developed cardiac hypertrophy and systolic dysfunction equally, suggesting that PKA phosphorylation of Ser2808 does not significantly impact the maladaptive cardiac remodeling following chronic stress. Because RyR2-S2808A mice compensated for the lack of Ser2808 phosphorylation by augmenting the participation of other phosphoepitopes (Figure 4), our results do not rule out that RyR2 phosphorylation is a central mechanism in the cellular response to acute or chronic stress, but they are at odds with other reports that place this mutation alone as a key determinant of SR Ca2+ leak,8,9 cardiac protection in HF progression,10 or exercise-induced cardiac arrhythmias.7
Control Mechanisms of Diastolic Ca2+ Leak: RyR2 Phosphorylation Versus SR Load
Diastolic Ca2+ leak normally offsets Ca2+ release and uptake imbalances to stabilize SR Ca2+ load during contractions; however, an increase in the rate of diastolic Ca2+ leak may induce Ca2+-dependent arrhythmias and is probably a key determinant of the contractile dysfunction that characterizes HF.22,23 In principle, an increase in diastolic Ca2+ release may be caused by any mechanism that increases SR Ca2+ load or by those that increase RyR2 activity or both. Whereas it is progressively clear that diastolic Ca2+ leak (and systolic Ca2+ release) is greatly dependent on SR load,22,24 the role of RyR2 modulation remains less obvious.23 Recently, Eisner’s group has shown that modulators of RyR2 (caffeine, tetracaine) quickly modify the magnitude of the [Ca2+]i transient but also that the change in SR load brought about by the modified RyR2 prevents long-term effects. In this scenario, modulation of RyR2 activity may influence diastolic Ca2+ leak but the effect would be contingent on SR Ca2+ load.24
In single-channel experiments, we previously showed that PKA phosphorylation of canine RyR2 results in complex kinetic responses, mainly increasing its transient peak Po in response to a rapid rise in activating [Ca2+], but also modestly decreasing its activity at steady [Ca2+] (pCa 5).12 PKA phosphorylation, therefore, produces a RyR2 with faster activation and relaxation kinetics to the trigger Ca2+, which in cellular settings would translate into increased rates of Ca2+ release and faster turn off of release for a given ICa, both of which are detected experimentally.25 At no point did we observe an increase in RyR2 activity at diastolic [Ca2+]. Presently, we show that mouse RyR2, like canine12 and human17 RyR2, fails to modify its activity in response to PKA phosphorylation (Figure 7) at diastolic [Ca2+] (pCa 7). Ablation of the S2808 phosphoepitope, presumably the center of PKA effects,7–10 did not confer gain-of-function. Nevertheless, after return to activating cytosolic [Ca2+] (pCa 5), PKA-phosphorylated RyR2-S2808A channels consistently displayed increased activity compared with their WT counterparts. Western blots conducted under conditions closely resembling those in the bilayer experiments clearly show dephosphorylation of S2808 by PP1 and phosphorylation of S2808 and S2030 by PKA, supporting the notion that the purported effects (or lack thereof) of phosphorylation seen in bilayers were indeed manifestations of the phosphorylation state of RyR2.
Overall, the above results and those obtained in cells suggest important conclusions. (1) PKA phosphorylation of RyR2 alone (either at S2808 or S2030 or both) does not appear to constitute a major mechanism to increase diastolic Ca2+ leak; additional factors appear to be needed, likely, increased SR Ca2+ load.6,22,24 In support of this, cAMP increased Ca2+ spark activity in permeabilized cells of WT and RyR2-S2808A hearts (Figure 6), most likely by phosphorylation of PLB and the subsequent increase in SR load, as shown by Li et al.6 (2) PKA phosphorylation of S2808 does modify RyR2 activity, ie, it attenuates RyR2 activity when [Ca2+] is high (shown by differences in Po of WT and RyR2-S2808A channels at pCa 5.3; Figure 7C and 7D, dotted line), probably contributing to the relative Ca2+ insensitivity of RyR2 in vivo26 and the faster turn off of release seen in bilayers12 and cellular25 settings, but the effect appears modest and is likely obscured by other modulators of RyR2s that are relevant when Ca2+ is high, namely, Ca2+ itself (ICa), and the SR Ca2+ load that increases during diastole. This is supported by the lack of prominent effects of S2808 ablation on Ca2+ transients and the full restoration of cellular function on β-adrenergic stimulation (Figure 5). The role of other RyR2 phosphorylation sites that are Ca2+ activated (such as the CaMKII-phosphorylated S2815 site) remains to be defined, but this site played no role in our cellular experiments because we omitted calmodulin and added KN-93, a specific CaMKII blocker. (3) PKA readily phosphorylated S2030 in vitro (Figures 3 and 7⇑) and likely in vivo as well (Figure 4), but phosphorylation of this site, like S2808, was insufficient to activate RyR2 at diastolic [Ca2+]. Although conclusions on the role of S2030 in cellular Ca2+ handling cannot be firmly drawn because RyR2-S2808A hearts tended to overphosphorylate this site (Figure 4), in vitro experiments were done at quasiequivalent levels of S2030 phosphorylation (Figure 7) and yielded modest effects as well.
Comparison With Other Results
The effects of PKA phosphorylation of RyR2 have been investigated in several animal species, including human, and at different levels of integration, from single-channel effects to whole-heart physiology. The results do not always converge, and a unifying theme has not been found, for reasons that are only now starting to emerge. First, whereas all 3 clearly identified phosphorylation sites are present in the RyR2 protein of all mammals investigated, there appears to be variability in their basal level of phosphorylation, which could account for different outcomes. For example, Carter et al18 found high basal levels of phosphorylation of S2809 (≈70% of all sites) in sheep heart homogenates, like we did here in mice, and associated steep changes in RyR2 activity with small changes in the phosphorylation of this site. However, not all species appear as eager to phosphorylate S2808 basally, ie, canine and rat RyR2 can be substantially phosphorylated at this site by PKA.17 Second, while investigating the effects of PKA, very few studies have carefully controlled SR load, which, as discussed above, profoundly influences RyR2 activity and may conceivably override phosphorylation effects.22–24 Furthermore, SR load may dramatically influence the propensity of RyR2 to undergo phosphorylation, as the latter process appears to be modulated by the open state of the channel.27 All of this may account, in whole or in part, for the activation,7–11 inhibition12,13 or lack of effect14 of PKA phosphorylation on RyR2.
The issues above that cloud the interpretation of PKA effects may be overcome in studies in which precise control of RyR2 phosphorylation is achieved, SR load is controlled, and animal models with selective ablation of the incumbent phosphorylation site may be tested side-by-side with WT controls. However, even then, outstanding differences may remain, as demonstrated with this study and the results of Wehrens et al.10 Using mice with an identical genetic mutation (RyR-S2808A), they place S2808 phosphorylation in a privileged position to control RyR2 activity in health and disease. First, they found that RyR2-S2808A does not dwell in subconducting states after PKA phosphorylation, as WT channels do. Others11,13,14,18 and we12,17 have not found subconducting states in RyR2 as the hallmark of PKA phosphorylation. Second, in their hands, RyR2-S2808A channels cannot be PKA phosphorylated.10 Our in vitro phosphorylation protocol (Figure 4) clearly shows incorporation of [γ-32P]-ATP in RyR2-S2808A channels, and phospho-specific antibodies reveal that S2030 is also phosphorylated by PKA, as previously shown.16 This difference is thus perplexing, but a potential explanation is that the S2030 site appears sensitive to endogenous phosphatases, as we found that SR-embedded RyR2-S2808A proteins were substantially less phosphorylated than their immunoprecipitated counterparts (not shown). Wehrens et al10 could have immunoprecipitated more phosphatases than we did in our assays. Lastly, their RyR2-S2808A mice were substantially protected after myocardial infarction, whereas we found no protection following aortic banding. Again, a potential source of disparate results is that myocardial infarction, which involves substitution of dead cells with scar tissue, may bring about radically different cellular adaptation than aortic banding, which exerts even stress on the left ventricle and is unlikely to ramp up apoptosis. Both models, however, produce chronic activation of the β-adrenergic pathway and should equally elicit long-term activation of PKA.
In summary, we found that PKA phosphorylation of RyR2 leads to modest effects discernible in single-channel assays. Phosphorylation of the S2808 residue appears to attenuate RyR2 activity at high Ca2+ only, leaving diastolic Ca2+ responses unaffected and hence unlikely to mediate a proposed PKA-driven increase in SR Ca2+ leak. This effect is obscured by the complex interactive array of proteins phosphorylated by PKA during sympathetic stimulation, as [Ca2+]i transients in WT and RyR2-S2808A cardiomyocytes appear similar during β-adrenergic stimulation. Ablation of this site offers little protection during chronic stress, implying a limited role for RyR2-S2808 phosphorylation in the pathogenesis of HF.
We thank Larry Whitesell for the animal surgery and Joe Warren for the generation of the chimeric mice.
Sources of Funding
This study was supported by NIH grants HL-55438 and HL-76826 (to H.H.V.).
Original received March 28, 2007; revision received July 23, 2007; accepted August 14, 2007.
Endoh M. Changes in intracellular Ca mobilization and Ca sensitization as mechanisms of action of physiological interventions and inotropic agents in intact myocardial cells. Jpn Heart J. 1997; 39: 1–44.
Witcher DR, Kovacs RJ, Schulman H, Cefali DC, Jones LR. Unique phosphorylation site on the cardiac ryanodine receptor regulates calcium channel activity. J Biol Chem. 1991; 266: 11144–11152.
Yoshida A, Takahashi M, Imagawa T, Shigekawa M, Takisawa H, Nakamura T. Phosphorylation of ryanodine receptors in rat myocytes during β-adrenergic stimulation. J Biochem (Tokyo). 1992; 119: 186–190.
Li Y, Kranias EG, Mignery GA, Bers DM. Protein kinase A phosphorylation of the ryanodine receptor does not affect calcium sparks in mouse ventricular myocytes. Circ Res. 2002; 90: 309–316.
Vest JA, Wehrens XH, Reiken SR, Lehnart SE, Dobrev D, Chandra P, Danilo P, Ravens U, Rosen MR, Marks AR. Defective cardiac ryanodine receptor regulation during atrial fibrillation. Circulation. 2005; 111: 2025–2032.
Reiken S, Gaburjakova M, Guatimosim S, Gómez AM, D’Armiento J, Burkhoff D, Wang J, Vassort G, Lederer WJ, Marks AR. Protein kinase A phosphorylation of the cardiac calcium release channel (ryanodine receptor) in normal and failing hearts. Role of phosphatases and response to isoproterenol. J Biol Chem. 2003; 278: 444–453.
Wehrens XH, Lehnart SE, Reiken S, Vest JA, Wronska A, Marks AR. Ryanodine receptor/calcium release channel PKA phosphorylation: a critical mediator of heart failure progression. Proc Natl Acad Sci U S A. 2006; 103: 511–518.
Hain J, Onoue H, Mayrleitner M, Fleischer S, Schindler H. Phosphorylation modulates the function of the calcium release channel of sarcoplasmic reticulum from cardiac muscle. J Biol Chem. 1995; 270: 2074–2081.
Valdivia HH, Kaplan JH, Ellis-Davies GC, Lederer WJ. Rapid adaptation of cardiac ryanodine receptors: modulation by Mg2+ and phosphorylation. Science. 1995; 267: 1997–2000.
Stange M, Xu L, Balshaw D, Yamaguchi N, Meissner G. Characterization of recombinant skeletal muscle (Ser2843) and cardiac muscle (Ser2809) ryanodine receptor phosphorylation mutants. J Biol Chem. 2003; 278: 51693–51702.
Rodriguez P, Bhogal MS, Colyer J. Stoichiometric phosphorylation of cardiac ryanodine receptor on serine 2809 by calmodulin-dependent kinase II and protein kinase A. J Biol Chem. 2003; 278: 38593–38600.
Xiao B, Jiang MT, Zhao M, Yang D, Sutherland C, Lai FA, Walsh MP, Warltier DC, Cheng H, Chen SR. Characterization of a novel PKA phosphorylation site, serine-2030, reveals no PKA hyperphosphorylation of the cardiac ryanodine receptor in canine heart failure. Circ Res. 2005; 96: 847–855.
Jiang MT, Lokuta AJ, Farrell EF, Wolff MR, Haworth RA, Valdivia HH. Abnormal Ca2+ release, but normal ryanodine receptors, in canine and human heart failure. Circ Res. 2002; 91: 1015–1022.
Carter S, Colyer J, Sitsapesan R. Maximum phosphorylation of the cardiac ryanodine receptor at serine-2809 by protein kinase a produces unique modifications to channel gating and conductance not observed at lower levels of phosphorylation. Circ Res. 2006; 98: 1506–1513.
Mundiña-Weilenman C, Vittone L, Ortale M, Chiappe de Cingolani G, Mattiazzi A. Immunodetection of phosphorylation sites gives new insights into the mechanisms underlying phospholamban phosphorylation in the intact heart. J Biol Chem. 1996; 271: 33561–33567.
Cheng H, Lederer WJ, Cannell M. Calcium sparks: elementary events underlying excitation-contraction coupling in heart muscle. Science. 1993; 262: 740–744.
Meissner G, Rios E, Tripathy A, Pasek DA. Regulation of skeletal muscle Ca2+ release channel (ryanodine receptor) by Ca2+ and monovalent cations and anions. J Biol Chem. 1997; 272: 1628–1638.
Shannon TR, Pogwizd SM, Bers DM. Elevated sarcoplasmic reticulum Ca2+ leak in intact ventricular myocytes from rabbits in heart failure. Circ Res. 2003; 93: 592–594.
Bers DM, Eisner DA, Valdivia HH. Sarcoplasmic reticulum Ca2+ and heart failure. Roles of diastolic leak and Ca2+ transport. Circ Res. 2003; 93: 487–490.
Venetucci LA, Trafford AW, Eisner DA. Increasing ryanodine receptor open probability alone does not produce arrhythmogenic calcium waves: threshold sarcoplasmic reticulum calcium content is required. Circ Res. 2007; 100: 105–111.
Niggli E, Lederer WJ. Voltage-independent Ca2+ release in heart muscle. Science. 1990; 250: 565–568.