Role of CaMKIIdelta phosphorylation of the cardiac ryanodine receptor in the force frequency relationship and heart failure

Abstract

The force frequency relationship (FFR), first described by Bowditch 139 years ago as the observation that myocardial contractility increases proportionally with increasing heart rate, is an important mediator of enhanced cardiac output during exercise. Individuals with heart failure have defective positive FFR that impairs their cardiac function in response to stress, and the degree of positive FFR deficiency correlates with heart failure progression. We have identified a mechanism for FFR involving heart rate dependent phosphorylation of the major cardiac sarcoplasmic reticulum calcium release channel/ryanodine receptor (RyR2), at Ser2814, by calcium/calmodulin-dependent serine/threonine kinase-delta (CaMKIIdelta). Mice engineered with an RyR2-S2814A mutation have RyR2 channels that cannot be phosphorylated by CaMKIIdelta, and exhibit a blunted positive FFR. Ex vivo hearts from RyR2-S2814A mice also have blunted positive FFR, and cardiomyocytes isolated from the RyR2-S2814A mice exhibit impaired rate-dependent enhancement of cytosolic calcium levels and fractional shortening. The cardiac RyR2 macromolecular complexes isolated from murine and human failing hearts have reduced CaMKIIdelta levels. These data indicate that CaMKIIdelta phosphorylation of RyR2 plays an important role in mediating positive FFR in the heart, and that defective regulation of RyR2 by CaMKIIdelta-mediated phosphorylation is associated with the loss of positive FFR in failing hearts.

Fig. 1.

CaMKIIδ phosphorylates RyR2 at Ser2814 in response to increased heart rate. (A) RyR2 immunoprecipitated from hearts from WT, RyR2-S2814A knock-in mice, and WT samples treated with either 1 μM KN-93 or PKI incubated with CaMKIIδ or PKA in the presence of [γ³²P]-ATP. (B) Lysates of low- and high-frequency paced WT, RyR2-S2814A, RyR2-S2808A, and KN-93–treated WT hearts. Blots show total CaMKIIδ, activated CaMKIIδ (pThr287), and total PLN levels. The phosphorylation status of PLN at Ser16 and Thr17 are also shown. (C) RyR2 immunoprecipitated from baseline and high-frequency paced hearts. Blot shows phosphorylation status of RyR2-Ser2808 and RyR2-Ser2814, as well as the levels of CaMKIIδ and activated CaMKIIδ complexed with RyR2. (D–G) Quantitative summary of data (^‡P < 0.05 vs. WT; *P < 0.05 vs. baseline). All experiments were performed in triplicate. AU, arbitrary unit.

Fig. 2.

Mice lacking the CaMKIIδ phosphorylation site on RyR2 have a blunted FFR. (A) In vivo contractility as a function of pacing frequency. Contractility represented as percentage change in dP/dt_max from baseline (WT, n = 9; S2814A, n = 8; *P < 0.05). (B) Representative left ventricular pressure traces of ex vivo contractility data from Langendorff perfused hearts paced from 400 beats/min to 550 beats/min. (C) Pooled results of ex vivo contractility study for hearts paced from baseline heart rate of 400 beats/min to 550 or 700 beats/min represented as percentage change in dP/dt_max from baseline heart rate (WT, n = 13; S2814A, n = 9; S2808A, n = 4; WT + KN-93, n = 5; asterisk over KN-93 bar indicates significant difference vs. WT; *P < 0.01; **P < 0.05).

Fig. 3.

Cardiomyocytes lacking the CaMKIIδ phosphorylation site on RyR2 have a blunted FFR. (A) Representative fractional shortening traces from isolated cardiomyocytes paced at 0.5 Hz and then stepped directly to 3.0 Hz. (B) Pooled fractional shortening data (WT, n = 37 cells; RyR2-S2814A, n = 35 cells; n = 4 mice per group) presented as change in cell length normalized to baseline cell length. At 3.0 Hz the fractional shortening of WT cardiomyocytes was significantly higher than RyR2-S2814A cardiomyocytes (*P < 0.01). Both WT and RyR2-S2814A myocytes had significant increases in fractional shortening at 3.0 Hz compared with 0.5 Hz (P < 0.001 and P < 0.05, respectively). (C) Representative Ca²⁺ transients from cardiomyocytes loaded with Fura-2. Data presented as the 340/380 ratio of the dye. (D) Pooled Ca²⁺ fluorescence ratio data (WT peak, 2.35 ± 0.06, n = 30 cells; RyR2-S2814A peak, 2.15 ± 0.05, n = 32 cells; n = 4 mice per group; *P < 0.05).

Fig. 4.

CaMKIIδ is depleted from RyR2 in heart failure. (A) RyR2 immunoprecipitated from human and mouse normal and heart failure hearts probed for phosphorylation at RyR2-Ser2814, CaMKIIδ, and CaMKIIδ-Thr287. (B) Lysates from human and mouse normal and heart failure hearts probed for total CaMKIIδ and CaMKIIδ-pT287. (C–F) Quantitative summary of immunoblot data. Graphs represent the relative amount of CaMKIIδ associated with RyR2 (C), relative amount of activated CaMKIIδ associated with RyR2 normalized to the amount of CaMKIIδ present in the channel complex (D), total CaMKIIδ in heart lysate (E), and relative amount of activated CaMKIIδ normalized to the total amount of CaMKIIδ in mouse and human cardiac lysates (F). All experiments were performed in triplicate. AU, arbitrary unit. *P < 0.05.

Fig. 5.

Preventing CaMKIIδ phosphorylation of RyR2 does not impart cardioprotection in heart failure. Ejection fraction of WT and RyR2-S2814A mice before and after MI followed by echocardiography for 4 weeks (WT, n = 10; S2814A, n = 11).