Ca2+/Calmodulin-Dependent Protein Kinase II Phosphorylation of Ryanodine Receptor Does Affect Calcium Sparks in Mouse Ventricular Myocytes
Previous studies in transgenic mice and with isolated ryanodine receptors (RyR) have indicated that Ca2+-calmodulin-dependent protein kinase II (CaMKII) can phosphorylate RyR and activate local diastolic sarcoplasmic reticulum (SR) Ca2+ release events (Ca2+ sparks) and RyR channel opening. Here we use relatively controlled physiological conditions in saponin-permeabilized wild type (WT) and phospholamban knockout (PLB-KO) mouse ventricular myocytes to test whether exogenous preactivated CaMKII or endogenous CaMKII can enhance resting Ca2+ sparks. PLB-KO mice were used to preclude ancillary effects of CaMKII mediated by phospholamban phosphorylation. In both WT and PLB-KO myocytes, Ca2+ spark frequency was increased by both preactivated exogenous CaMKII and endogenous CaMKII. This effect was abolished by CaMKII inhibitor peptides. In contrast, protein kinase A catalytic subunit also enhanced Ca2+ spark frequency in WT, but had no effect in PLB-KO. Both endogenous and exogenous CaMKII increased SR Ca2+ content in WT (presumably via PLB phosphorylation), but not in PLB-KO. Exogenous calmodulin decreased Ca2+ spark frequency in both WT and PLB-KO (K0.5 ≈100 nmol/L). Endogenous CaMKII (at 500 nmol/L [Ca2+]) phosphorylated RyR as completely in <4 minutes as the maximum achieved by preactivated exogenous CaMKII. After CaMKII activation Ca2+ sparks were longer in duration, and more frequent propagating SR Ca2+ release events were observed. We conclude that CaMKII-dependent phosphorylation of RyR by endogenous associated CaMKII (but not PKA-dependent phosphorylation) increases resting SR Ca2+ release or leak. Moreover, this may explain the enhanced SR diastolic Ca2+ leak and certain triggered arrhythmias seen in heart failure.
Cardiac ryanodine receptors (RyR) are central in excitation contraction coupling (ECC) both as the sarcoplasmic reticulum (SR) Ca2+ release channel and as a scaffolding protein that localizes numerous regulatory proteins to the junctional complex.1 Ca2+ sparks reflect the synchronous activation of a cluster of ≈6 to 20 RyR at a single junction, producing both the diastolic SR Ca2+ leak and the temporally synchronized SR Ca2+ release during ECC.1–3 Their activity can be modulated by many factors including divalent cations, adenine nucleotides, calmodulin (CaM), caffeine, oxidation, and phosphorylation.1,2
Protein kinase A (PKA)-dependent RyR phosphorylation has been reported to increase RyR open probability at the single channel level.4–6 It was further suggested that hyperphosphorylation of RyR at Ser2809 by PKA in heart failure (HF) caused dissociation of FK-506 binding protein (FKBP12.6) from the RyR, resulting in enhanced diastolic Ca2+ leak, reduced SR Ca2+ content, and contractile dysfunction.7 However, Li et al8 found that cAMP-induced RyR phosphorylation had no effect on resting RyR-mediated SR Ca2+ leak (assessed via Ca2+ sparks) in phospholamban (PLB) knockout mouse myocytes (PLB-KO). It remains controversial whether PKA-dependent RyR phosphorylation does occur in HF and whether it causes dissociation of FKBP12.6 from RyR or mediates enhanced SR Ca2+ leak.8–10
Ca2+/calmodulin-dependent protein kinase II (CaMKII) has been reported to phosphorylate or regulate L-type Ca2+ current (ICa), PLB, and RyR.11–16 At the single-channel recording level, reports suggest that CaMKII can either enhance5,16,17 or depress RyR activity.18 However, in voltage clamped intact myocytes, Li et al19 reported that endogenous CaMKII increased SR Ca2+ release for a given SR Ca2+ content and ICa trigger. In addition, protein phosphatases (PP1 and PP2A) can reduce SR Ca2+ release channel activity for a given SR load and ICa,20 but can also enhance Ca2+ spark frequency.21 Moreover, in transgenic mice overexpressing CaMKIIδC, fractional SR Ca2+ release was enhanced and resting spontaneous SR Ca2+ spark frequency was dramatically increased, despite lower SR Ca2+ load and lower diastolic [Ca2+]i.22 Moreover, CaMKII is associated with RyR22 and is upregulated in heart failure,22 where SR Ca2+ leak may contribute to systolic dysfunction and arrhythmogenesis.22
The goal here was to test the effects of CaMKII on resting SR Ca2+ release (as Ca2+ sparks) in saponin-permeabilized mouse ventricular myocytes (wild type [WT] and PLB-KO). Activation of endogenous CaMKII or application of preactivated exogenous CaMKII (autophosphorylated) resulted in RyR phosphorylation and enhanced resting Ca2+ spark frequency (CaSpF), independent of SR Ca2+ load. CaM also depressed RyR activity with a half-maximal effect at 100 nmol/L CaM.
Cardiac Myocyte Isolation
Mouse ventricular myocytes were enzymatically isolated as previously described (see the online data supplement, available at http://circres.ahajournals.org),8 and PLB-KO mice were provided by Dr E.G. Kranias (University of Cincinnati, OH). Procedures were performed according to Guiding Principles in the Care and Use of the Animals approved by the Council of the American Physiological Society.
Ca2+ Sparks in Permeabilized Myocytes
Myocytes were permeabilized with saponin (50 μg/mL) for 30 seconds (see the online data supplement for details) and placed in internal solution (in mmol/L): EGTA 1, HEPES 10, K-aspartate 120, ATP 5, free MgCl2 1, reduced glutathione 10, free [Ca2+] 50 nmol/L for WT and 10 or 25 nmol/L for PLB-KO (see the online supplement), creatine phosphokinase 5 U/mL, phosphocreatine 10, dextran (relative molecular mass: 40 000) 4%, K4Fluo-3 0.05, pH 7.2. PKA inhibitory peptide PKI (15 μmol/L; Calbiochem no. 116805) was in all bath solutions (except when exogenous PKA was added). After baseline Ca2+ sparks were recorded, myocytes were exposed to solution designed to phosphorylate RyR (see below). After this period, solution was replaced by the original internal solution containing 10 μmol/L okadaic acid (OA) to prevent dephosphorylation in subsequent Ca2+ spark measurements. Control experiments showed that OA by itself (5 minutes) caused no change in CaSpF in 19 of 20 myocytes. Ca2+ sparks were recorded as previously described.8 SR Ca2+ load was evaluated by Ca2+ transient amplitude on caffeine application. CaSpF is expressed as sparks per picoliter per second (p1−1s−1).
Activation of Endogenous CaMKII
To activate endogenous CaMKII in permeabilized ventricular myocytes [Ca2+]i was elevated for 1 minute to 500 nmol/L with 1.2 μmol/L exogenous CaM. Phosphatase inhibitor OA (2 μmol/L) was included to prevent dephosphorylation of CaMKII and its targets, and 15 μmol/L PKI was included.
Preactivation of Exogenous CaMKII
We also used preactivated exogenous CaMKII. CaMKIIα (12 mg/mL; gift from Dr J.H. Brown, University of California, San Diego) was preactivated (autophosphorylated) by incubation with (in μmol/L): ATP 100, CaM 2.4, and CaCl2 200 for 10 minutes at 37°C. In some experiments, CaMKII inhibitor AIP (1 μmol/L) was included as a negative control. Preactivated CaMKII was then rapidly diluted into final superfusate with free [Ca2+] restored to 50 nmol/L (with 6 μg/mL CaMKII, 1.2 μmol/L CaM, 10 μmol/L OA, and 15 μmol/L PKI) and exposed to permeabilized myocytes.
Immunoprecipitation and CaMKII Phosphorylation of RyR
RyR was immunoprecipitated from homogenate or cell lysate with antibody (MA3–916, Affinity BioReagents) in 0.5 mL of 62.5 mmol/L Tris-HCl buffer (pH 7.5), 0.9% NaCl, 0.5 mmol/L NaF, 1% Triton-X, and protease inhibitors for 2 hours at 4°C (see the online supplement for details). CaMKII phosphorylation of immunoprecipitated RyR was initiated by adding autophosphorylated CaMKII and γP32-ATP (final specific activity 300 μCi/μmol). Reactions were terminated with stop solution (10% sodium dodecyl sulfate, 300 mmol/L EGTA, and 0.25 mmol/L DTT) and size-fractionated on 5% SDS-PAGE, and RyR radioactivity was quantified using Unscan-it software. RyR protein was determined by immunoblotting. FKBP12.6 was measured6 using anti-FKBP12 (supplied by Dr A.R. Marks, Columbia University, New York; see the online supplement).
Ca2+ sparks were analyzed as previously described (see the online supplement).8 Ca2+ spark amplitudes were normalized to fluorescence baseline (F0) as F/F0, duration was full-duration half-maximum (FDHM), and width was full-width half-maximum (FWHM). Results are expressed as mean±SEM. Significance (P<0.05) was determined using Student t test.
Exogenous CaMKII on Ca2+ Sparks and SR Ca2+Content in WT mice
Figure 1A shows representative line scan images from a WT myocyte before, during application of preactivated CaMKIIα with CaM, and following washout of CaM/CaMKIIα (with dephosphorylation inhibited by OA). Figure 1B shows mean data, where CaSpF did not increase during CaM/CaMKII exposure for 5 minutes (205±20 versus 199±39 sparks pl−1s−1), but 5 minutes after washout of CaM/CaMKII CaSpF greatly increased (by 108%, n=10) versus control.
The dramatic increase in CaSpF during CaM/CaMKII washout is consistent with CaMKII-dependent activation of RyR. However, the failure to see increased CaSpF during CaM/CaMKII exposure could be attributable to an acute inhibitory effect of CaM (see CaM Alone Inhibits Ca2+ Spark Frequency).
CaMKII can phosphorylate PLB to cause increased SR Ca2+ uptake and content, and CaSpF depends strongly on SR Ca2+ content.1,8 Figure 1C shows that after exposure to CaMKII there was a small (14%), but significant increase in SR Ca2+ content (assessed by caffeine-induced Ca2+ transients). Given the steep leak versus SR Ca2+ content relationship,1,8 we cannot rule out the possibility that part of the 108% increase in CaSpF after CaMKII exposure was secondary to the 14% increase in SR Ca2+ content.
Exogenous CaMKII on Ca2+ Sparks and SR Ca2+Content in PLB-KO Mice
To eliminate the potential complicating effect of CaMKII on PLB we repeated the same protocol in PLB-KO mice (Figure 2). In PLB-KO myocytes, CaM/CaMKIIα addition raised the mean CaSpF by 28%, but not significantly (P=0.06). After washout, CaSpF increased significantly by 60% (P<0.0001, n=9; Figure 2B). In PLB-KO myocytes, SR Ca2+ content was not increased by exposure to activated CaMKII (Figure 2C). Thus, the increased CaSpF cannot be secondary to increased SR Ca2+ content, but could be caused by CaMKII-dependent RyR modulation. As a control, these experiments were also done with CaMKII inhibitory peptide AIP in the CaMKII preactivation cocktail. Figure 2D shows that AIP prevented the increase in CaSpF. Thus CaMKII activity is required for the observed increase in CaSpF.
There were small, but significant changes in Ca2+ spark characteristics (Table). For example, there was larger spark duration and spatial spread after exposure to activated CaMKII (in both WT and PLB-KO). These observations would be consistent with CaMKII causing minor increases in the duration of SR Ca2+ release during Ca2+ sparks. The small increase in Ca2+ spark amplitude in WT cells during CaM/CaMKII exposure might be secondary to the increase in SR Ca2+ content caused by the combination of SR Ca-ATPase stimulation (via PLB phosphorylation) and inhibition of SR Ca2+ leak (via CaM, see CaM Alone Inhibits Ca2+ Spark Frequency).
CaM Alone Inhibits Ca2+ Spark Frequency
In both WT and PLB-KO mice, we often saw an initial acute inhibition of CaSpF, which we hypothesized was caused by CaM-dependent RyR inhibition as observed in bilayer and SR vesicle studies.23,24 To test whether CaM had direct effects on RyR (versus via CaMKII), we applied different CaM concentrations to myocytes under conditions where CaMKII should not be activated: (1) in WT myocytes at 50 nmol/L [Ca2+]i without CaMKII inhibitor AIP (this low [Ca2+] should not activate CaMKII; Figure 3A) and (2) in PLB-KO myocytes at 25 nmol/L [Ca2+]i with AIP (to inhibit any endogenous CaMKII; Figure 3B). The right panels of Figure 3 show that CaM decreased CaSpF in a dose-dependent manner with half-inhibition at ≈100 nmol/L CaM (K0.5) in both WT and PLB-KO myocytes. Furthermore, the inhibitory effect was partially reversed by CaM washout (Figure 3A and 3B). CaM did not alter Ca2+ spark properties (not shown).
In other PLB-KO myocytes where AIP was not included, very high CaM (≥2 μmol/L) increased CaSpF in a minority of cells (2 out 6), an effect never observed at low CaM concentration or when AIP was included (Figure 3B). This may be because of some degree of CaMKII activation at very high CaM (despite low [Ca2+]i). This unexpected increase of CaSpF could be abolished by AIP, supporting our conclusion that RyRs are activated by CaMKII but inhibited by CaM.
In Vitro Time-Dependent RyR Phosphorylation by CaMKII
To test whether our preactivation approach can phosphorylate RyR, we measured 32P incorporation in immunoprecipitated RyRs (Figure 4A, n=4). In the absence of OA, RyR phosphorylation by CaMKIIα reached a peak in 5 to 8 minutes (normalized to RyR loading for each sample), but declined at 10 minutes. However, with OA, 32P incorporation reached a higher maintained peak at 8 minutes. This raises 2 points. First, our preactivated CaMKIIα readily phosphorylates RyR, and second, phosphatases that coimmunoprecipitate with RyR may limit phosphate incorporation. The decline of 32P at 8 to 10 minutes without OA might be partly caused by the exhaustion of 32P-ATP available (and dephosphorylation), because [ATP] was lower here than in the Ca2+ spark measurements (100 μmol/L versus 5 mmol/L) to optimize 32P specific activity.
Exogenous PKA and CaSpF
The approach used here previously showed that cAMP-dependent RyR phosphorylation could increase CaSpF in WT, but not in PLB-KO mice (or where only nonphosphorylatable PLB was expressed).8 Here, we further test whether exogenous PKA catalytic subunit alters CaSpF, as we see for preactivated CaMKII. Figure 4B shows that exogenous PKA increased CaSpF in WT, but not in PLB-KO. These results with active PKA agree with the earlier results with cAMP (which depended on endogenous PKA).8 However, the PKA results contrast dramatically with the present results with activated CaMKII.
Because the PKA effect on RyR may depend on dissociation of FKBP12.6, we also measured FKBP12/12.6 expression and RyR association. There is no change in total FKBP12 expression in PLB-KO versus WT hearts (Figure 4C). Nor was there any difference in FKBP12.6, which coimmunoprecipitates with RyR in control conditions or after treatment with PKA or PKI (Figure 4D). We also did not find any appreciable loss of RyR-associated FKBP12.6 on myocyte permeabilization as used here (not shown). Thus, we did not detect FKBP12.6 dissociation under conditions where in WT myocytes there was enhanced CaSpF (presumably because of PLB phosphorylation, enhanced SR Ca-ATPase activity, and SR Ca content).8
Endogenous CaMKII on Ca2+ Sparks in Permeabilized Myocytes
Since endogenous CaMKIIδ is known to associate with the RyR,17,25–27 we sought to test whether endogenous CaMKII can activate RyR in myocytes as seen above for exogenous CaMKII. Figure 5A shows line scans during a 5-minute control period, as [Ca2+]i was elevated to 500 nmol/L (with 1.2 μmol/L CaM, 15 μmol/L PKI, and 2 μmol/L OA added to activate endogenous CaMKII) and on washout (where CaM was removed and [Ca2+]i restored to the initial level).
Figure 5B and 5C shows that endogenous CaMKII activation increased CaSpF in both WT and PLB-KO mice (mean for 3 to 12 minutes in sparks pl−1s−1 increased from 139±31 to 350±37 in WT, P=0.009, and from 267±29 to 529±75 in PLB-KO, P=0.004). Maximum CaSpF occurred within 4 minutes of CaMKII activation and was maintained for the entire 10 minutes studied (OA was present). The same protocol was performed with the specific CaMKII inhibitor peptide (AIP 1 μmol/L) included during and after [Ca2+]i and CaM elevation. AIP prevented the increase in CaSpF in both WT and PLB-KO (Figure 5B and 5C). SR Ca2+ content was significantly increased in WT myocytes, but did not change significantly in PLB-KO. This demonstrates that the increased CaSpF was not caused by increased SR load in the PLB-KO. These results are consistent with exogenous CaMKII effects on both PLB and RyR.
In WT myocytes PLB-dependent stimulation of SR Ca-ATPase must be slightly stronger than the enhanced SR Ca2+ leak, resulting in higher SR Ca2+ content. However, in PLB-KO, the enhanced SR Ca2+ leak should lower SR Ca2+ content, but did not. This may be because the SR Ca2+-ATPase is so active in PLB-KO myocytes that the leak does not greatly depress load (see Figure II in the online supplement).
Ca2+ spark amplitude increased in WT myocytes on endogenous CaMKII activation (Table), possibly because of enhanced SR content and spark duration (FDHM). In PLB-KO, spark amplitude also increased, possibly because of prolonged release duration. Figure 6A shows histograms of Ca2+ spark durations (± activation of endogenous CaMKII) in PLB-KO myocytes. For control, 90% of Ca2+ sparks were 10 to 40 ms in FDHM, whereas after CaMKII this value was only 77%. There were essentially no Ca2+ sparks with FDHM >70 ms in control (0.3% of events), but CaMKII increased this to 4.7% of the already more frequent events. Integrating the individual Ca2+ spark events (from longest to shortest duration, Figure 6B) shows that half the Ca2+ sparks are longer than 32 ms in CaMKII versus 26 ms in control. Figure 6C shows that CaMKII activation resulted in occasional macrosparks (that exceed the spatial spread of normal Ca2+ sparks; 0.11±0.03/s) and miniwaves (propagating Ca-induced Ca-release; 0.021±0.011/s), but these events were virtually absent in control (0.008±0.0003/s macrosparks and 0 miniwaves). The much higher frequency of long Ca2+ sparks and propagating SR Ca2+ release events may reflect a higher propensity for initiation of delayed afterdepolarizations (DADs) and consequent triggered arrhythmias on CaMKII-dependent RyR phosphorylation.
Phosphorylation of RyR and PLB in WT Mouse Ventricular Myocytes
Finally, we assessed CaMKII-dependent RyR phosphorylation after treating ventricular myocytes as in the Ca2+ spark measurements. Here we used RyR back-phosphorylation after myocyte incubation by subsequent treatment with CaMKIIα (and γ32P-ATP, high [Ca2+], and [CaM]). In this case the amount of 32P incorporated is highest when the least RyR phosphorylation already occurred during the primary incubation (ie, basal and AIP+alkaline phosphatase–treated samples in Figure 7A, lane 2 and 1). Thus, the reciprocal of back-phosphorylation is an index of prior phosphorylation (Figure 7B). Ca/CaM-treated myocytes had nearly maximal phosphorylation levels, similar to when cells were exposed to exogenous activated CaMKIIα. The time course of in vivo RyR phosphorylation is consistent with that of CaSpF (Figure 5B). Myocytes treated with AIP and alkaline phosphatase (Figure 7A, lane 1) had ≈98% of the control basal phosphorylation (lane 2), suggesting the basal RyR phosphorylation by CaMKII is very low.
CaMKII also phosphorylates PLB at Thr-17, and this is a control for CaMKII target phosphorylation. Figure 7A and 7C shows that basal PLB phosphorylation was too low to detect. After treatment with Ca/CaM (even after restoration of [Ca2+]i to 50 nmol/L with OA), PLB phosphorylation quickly reached its maximum, with no apparent increase with exogenous CaMKIIα (n=4). Thus, both RyR and PLB were highly phosphorylated by treatment of high Ca/CaM, under conditions where CaSpF were evaluated in Figures 5 and 6⇑.
We demonstrated in a controlled mouse PLB-KO myocyte setting that: (1) CaMKII-dependent RyR phosphorylation significantly increases CaSpF and duration, independent of SR Ca2+ content (using both exogenous and endogenous CaMKII), (2) exogenous PKA catalytic subunit does not activate SR Ca2+ leak via Ca2+ sparks, and (3) CaM decreases CaSpF with K0.5≈100 nmol/L [Ca2+]i.
Most previous work has removed the RyR from its native cellular environment (SR vesicles or single-channel bilayer recordings) or used intact myocytes where [Ca2+]i, SR Ca2+ content, and CaMKII activation are difficult to control. Our permeabilized myocytes bridge these, in that [Ca2+]i, SR Ca2+content, and CaMKII activity are relatively controlled and the local RyR environment is minimally perturbed. In addition, PLB-KO myocytes prevent the major complicating effect of both CaMKII and PKA on the SR Ca-ATPase. With PLB present, activation of PKA and CaMKII can increase SR Ca2+ content, which itself alters CaSpF, SR Ca2+ leak, global Ca2+ transients, and Ca2+ spark amplitude.1–3,28
CaM and Resting SR Ca2+ Release
Our control CaM studies indicate that CaM depresses RyR activity in the cellular setting at diastolic [Ca2+]i, with a K0.5 for CaM ≈100 nmol/L. This is consistent with extensive earlier work on CaM binding to and inhibition of RyR gating in isolated RyR in bilayers or vesicles,23,24 with respect to both K0.5 and binding kinetics (ie, CaSpF changed over tens of seconds). Notably, this K0.5 is comparable to the free [CaM] in ventricular myocytes at diastolic [Ca2+]i,29 such that gradual changes in RyR regulation could occur as free [CaM] changes (although significant changes of RyR-CaM binding during a single heartbeat are unlikely).23,24
At high free [CaM] (≥2 μmol/L), inhibition of CaSpF began to reverse in an AIP-sensitive manner. This indicates that some local CaMKII activation may occur even at diastolic [Ca2+]i, when [CaM] is high. This emphasizes the importance of quantitative studies of CaM effects in a cellular environment.
Previous Results Concerning CaMKII and Resting Ca2+ Release
Most earlier studies of cardiac RyR gating in lipid bilayers have shown that CaMKII enhances RyR open probability,5,16,17 but not all results agree.18 A recent surprising result in bilayers was that phosphatase treatment also increased RyR activity in ≈50% of channels studied.21 Discrepancies among bilayer studies are unresolved.
Similarly, most work in intact ventricular myocytes suggests that CaMKII enhances cardiac RyR activation. Transgenic mice overexpressing CaMKIIδC show increased CaMKII associated with RyR, enhanced RyR phosphorylation, and increased fractional SR Ca2+ release and resting CaSpF (despite lower SR Ca2+ content and diastolic [Ca2+]i).22 Acute CaMKIIδC overexpression in rabbit ventricular myocytes (via adenovirus) increased fractional SR Ca2+ release and CaSpF (normalized for SR Ca2+ content).30 When SR Ca2+ content and ICa trigger were controlled and matched, activation of endogenous CaMKII greatly enhanced fractional SR Ca2+ release,19 and phosphatase manipulation gave functionally similar results.20,31 One study32 found opposite results, that constitutively active CaMKII inhibited SR Ca2+ release, whereas CaMKII inhibition (by AC3-I) enhanced SR Ca2+ release (although SR Ca2+ content was not measured in the same protocols). Currie et al26 used permeabilized rabbit ventricular myocytes and found that the CaMKII inhibitor AIP inhibited CaSpF and ryanodine binding (an indicator of RyR activation). Thus, our results and most others indicate that endogenous CaMKII activity can phosphorylate RyR and enhance RyR opening, both at rest (as SR Ca2+ leak or Ca2+ sparks), and during E-C coupling.
WT mouse myocytes had similar behavior to PLB-KO here (see the online supplement). However, conclusions for WT are complicated by CaMKII phosphorylation of PLB at Thr-17. CaMKII stimulation of both SR Ca-ATPase and RyR-mediated SR Ca2+ leak would affect SR Ca2+ content oppositely. In WT, the SR Ca-ATPase stimulation predominates with respect to SR Ca2+ content (which was slightly increased). However, both effects would tend to enhance SR Ca2+ leak and CaSpF. This emphasizes the value of PLB-KO mice in isolating RyR effects mechanistically. It also suggests that these combined actions in normal (or failing) myocytes could be more arrhythmogenic than either alone.
Does RyR Become Phosphorylated by CaMKII Under Physiological Conditions?
Wehrens et al17 showed that RyR phosphorylation at Ser2815 by CaMKII increases at higher heart rate, and the same was found in isolated myocytes.33 Increasing heart rate also accelerates SR Ca2+ uptake and normally increases SR Ca2+ content. Thus, under physiological conditions, the sort of CaMKII-dependent RyR regulation described here is likely to be dynamically functional.
PKA- Versus CaMKII-Dependent Changes in Ca2+ Sparks
The role of PKA-dependent RyR phosphorylation and enhancement of diastolic leak is controversial, and the field cannot be reviewed here. Briefly, 1 group has a cogent body of evidence from single RyR bilayer gating and extensive biochemical data that phosphorylation of cardiac RyR at S2809 by PKA causes FKBP12.6 dissociation from RyR, which would increase SR Ca2+ leak and decrease SR Ca2+ content.6,7 Other labs have been unable to confirm certain aspects of this hypothesis.8–10 Our group could not detect any effect on CaSpF on maximal RyR phosphorylation by endogenous PKA driven by exogenous cAMP in PLB-KO myocyte (in experiments like those described here).8 Here, we further tested the effect of PKA catalytic subunit on RyR activity (in direct parallel to our CaMKII studies). Again, we found that PKA enhanced CaSpF in WT mouse myocytes (consistent with enhanced SR Ca2+ uptake and content), but had no effect at all on CaSpF in PLB-KO mouse myocytes. This contrasts sharply with profound CaSpF enhancement seen here with both exogenous and endogenous CaMKII in PLB-KO myocytes. Thus, CaMKII has a powerful effect on resting Ca2+ release (versus PKA).
Conceivably, PLB-KO mice might have less FKBP12.6, which could mask a PKA effect in those myocytes. However, we found that there was neither a change in the extent of FKBP12.6 expression nor association with the RyR in these myocytes; nor was FKBP12.6 lost from the RyR on permeabilization and treatment with PKA or PKI. Thus, we remain unable to detect significant stimulation of CaSpF by PKA (unless PLB is present).
Both PKA and CaMKII affect RyR gating during ECC, when the ICa trigger and SR Ca2+ content are matched.19,34 Notably, PKA has no effect on the amount of SR Ca2+ release, fractional release, or ECC gain, but does enhance the initial rate of SR Ca2+ release (and its turn-off). These effects closely resemble single RyR channel behavior35 where PKA increased peak RyR opening during a rapid rise in local [Ca2+], but accelerated RyR closure as well. This again contrasts with CaMKII, where endogenous CaMKII activation enhances fractional SR Ca2+ release (for a given SR Ca content and ICa trigger) without greatly altering Ca2+ transient kinetics.23,24 Recent findings using overexpression of CaMKIIδC via transgenesis or via acute adenoviral transfer agree with this.22,30
Different roles of PKA and CaMKII in diastolic Ca2+ release and ECC are consistent with a difference in their molecular basis. Wehrens et al17 reported that phosphorylation of RyR occurs at Ser-2815 by CaMKII and Ser-2809 by PKA, and the 2 kinases produced different RyR gating phenotypes. This is also controversial, as PKA may also phosphorylate RyR at Ser2030 and CaMKII may phosphorylate Ser2809 and other unidentified RyR sites.10,15,16
Ca2+ Leak in Heart Failure: Pathological Implications
CaMKII is upregulated in HF in humans36 and animals.27,37 In an arrhythmogenic rabbit HF model, more CaMKII (and less phosphatase 1 and 2A) is associated with RyR, CaMKII is more activated (autophosphorylated), and RyR is more highly phosphorylated.27 In addition, there is increased diastolic SR Ca2+ leak that can be inhibited by CaMKII blockers,27,38 as was seen on CaMKII overexpression.22,30 There was also a decrease in RyR-bound CaM in this HF model,27 which could further enhance SR Ca2+ leak.
Thus, in HF there is likely to be CaMKII-dependent enhancement of diastolic SR Ca2+ leak via RyR, Ca2+ sparks, macrosparks, and Ca2+ waves. These diastolic SR Ca2+ release events can contribute to reduced SR Ca2+ content.38 However, they are also believed to underlie transient inward currents and delayed afterdepolarizations, which can initiate ventricular tachycardia, and the incidence of these events is increased in HF.
In conclusion, we demonstrated that activation of CaMKII (endogenous or exogenous) and reduced [CaM] greatly enhance resting SR Ca2+ release events (Ca2+ sparks and waves) in permeabilized ventricular myocytes under relatively controlled physiological conditions where neither SR Ca2+ content nor [Ca2+]i were altered. These CaMKII dependent effects may normally serve as a positive influence on Ca2+ transients at higher heart rate (eg, by enhancing fractional release during E-C coupling in association with enhanced SR Ca2+). However, when phosphorylation is perturbed in pathophysiological states such as HF, inappropriately high RyR phosphorylation by CaMKII may contribute to reduced SR Ca2+content and arrhythmogenesis.
We thank Brian French, Jaime O’Brien, Tina Valdez, and Christopher K. Means for assistance and Dr Sabine Huke for advice.
Sources of Funding
This work was supported by grants from the National Institutes of Health HL-30077, HL-80101 (D.M.B.), a predoctoral fellowship, and a Scientist Development Grant from the American Heart Association (T.G. and T.Z.).
Original received February 14, 2006; revision received June 1, 2006; accepted June 28, 2006.
Bers DM. Excitation-Contraction Coupling and Cardiac Contractile Force. 2nd ed. Dordrecht, Netherlands: Kluwer Academic Press; 2001.
Cheng H, Lederer WJ, Cannell MB. Calcium sparks: elementary events underlying excitation-contraction coupling in heart muscle. Science. 1993; 262: 740–744.
Takasago T, Imagawa T, Shigekawa M. Phosphorylation of the cardiac ryanodine receptor by cAMP-dependent protein kinase. J Biochem (Tokyo). 1989; 106: 872–877.
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.
Wehrens XH, Lehnart SE, Huang F, Vest JA, Reiken SR, Mohler PJ, Sun J, Guatimosim S, Song LS, Rosemblit N, D’Armiento JM, Napolitano C, Memmi M, Priori SG, Lederer WJ, Marks AR. FKBP12.6 deficiency and defective calcium release channel (ryanodine receptor) function linked to exercise-induced sudden cardiac death. Cell. 2003; 113: 829–840.
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.
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.
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.
Xu A, Hawkins C, Narayanan N. Phosphorylation and activation of the Ca2+-pumping ATPase of cardiac sarcoplasmic reticulum by Ca2+/calmodulin-dependent protein kinase. J Biol Chem. 1993; 268: 8394–8397.
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.
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.
Wehrens XH, Lehnart SE, Reiken SR, Marks AR. Ca2+/calmodulin-dependent protein kinase II phosphorylation regulates the cardiac ryanodine receptor. Circ Res. 2004; 94: e61–70.
Maier LS, Zhang T, Chen L, DeSantiago J, Brown JH, Bers DM. Transgenic CaMKIIdeltaC overexpression uniquely alters cardiac myocyte Ca2+ handling: reduced SR Ca2+ load and activated SR Ca2+ release. Circ Res. 2003; 92: 904–911.
Fruen BR, Bardy JM, Byrem TM, Strasburg GM, Louis CF. Differential Ca2+ sensitivity of skeletal and cardiac muscle ryanodine receptors in the presence of calmodulin. Am J Physiol Cell Physiol. 2000; 279: C724–733.
Balshaw DM, Xu L, Yamaguchi N, Pasek DA, Meissner G. Calmodulin binding and inhibition of cardiac muscle calcium release channel (ryanodine receptor). J Biol Chem. 2001; 276: 20144–20153.
Zhang T, Maier LS, Dalton ND, Miyamoto S, Ross J Jr, Bers DM, Brown JH. The deltaC isoform of CaMKII is activated in cardiac hypertrophy and induces dilated cardiomyopathy and heart failure. Circ Res. 2003; 92: 912–919.
Ai X, Curran JW, Shannon TR, Bers DM, Pogwizd SM. Ca2+/calmodulin-dependent protein kinase modulates cardiac ryanodine receptor phosphorylation and sarcoplasmic reticulum Ca2+ leak in heart failure. Circ Res. 2005; 97: 1314–1322.
Shannon TR, Ginsburg KS, Bers DM. Quantitative assessment of the SR Ca2+ leak-load relationship. Circ Res. 2002; 91: 594–600.
Wu X, Bers DM [Ca]i regulates calmodulin (CaM) location in adult rabbit cardiac myocytes. Biophys J. 2005: 83a. Abstract.
Kohlhaas M, Zhang T, Seidler T, Zibrova D, Dybkova N, Steen A, Wagner S, Chen L, Brown JH, Bers DM, Maier LS. Increased sarcoplasmic reticulum calcium leak but unaltered contractility by acute CaMKII overexpression in isolated rabbit cardiac myocytes. Circ Res. 2006; 98: 235–244.
duBell WH, Gigena MS, Guatimosim S, Long X, Lederer WJ, Rogers TB. Effects of PP1/PP2A inhibitor calyculin A on the E-C coupling cascade in murine ventricular myocytes. Am J Physiol Heart Circ Physiol. 2002; 282: H38–H48.
Wu Y, Colbran RJ, Anderson ME. Calmodulin kinase is a molecular switch for cardiac excitation-contraction coupling. Proc Natl Acad Sci U S A. 2001; 98: 2877–2881.
Huke S, Bers DM Frequency-dependent acceleration of relaxation (FDAR) does not require phospholamban phosphorylation by CaM-kinase II in rat cardiomyocytes. Biophys J. 2005: 311a. Abstract.
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.
Hoch B, Meyer R, Hetzer R, Krause EG, Karczewski P. Identification and expression of delta-isoforms of the multifunctional Ca2+/calmodulin-dependent protein kinase in failing and nonfailing human myocardium. Circ Res. 1999; 84: 713–721.
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.