Nitric oxide-dependent activation of CaMKII increases diastolic sarcoplasmic reticulum calcium release in cardiac myocytes in response to adrenergic stimulation

Abstract

Spontaneous calcium waves in cardiac myocytes are caused by diastolic sarcoplasmic reticulum release (SR Ca(2+) leak) through ryanodine receptors. Beta-adrenergic (β-AR) tone is known to increase this leak through the activation of Ca-calmodulin-dependent protein kinase (CaMKII) and the subsequent phosphorylation of the ryanodine receptor. When β-AR drive is chronic, as observed in heart failure, this CaMKII-dependent effect is exaggerated and becomes potentially arrhythmogenic. Recent evidence has indicated that CaMKII activation can be regulated by cellular oxidizing agents, such as reactive oxygen species. Here, we investigate how the cellular second messenger, nitric oxide, mediates CaMKII activity downstream of the adrenergic signaling cascade and promotes the generation of arrhythmogenic spontaneous Ca(2+) waves in intact cardiomyocytes. Both SCaWs and SR Ca(2+) leak were measured in intact rabbit and mouse ventricular myocytes loaded with the Ca-dependent fluorescent dye, fluo-4. CaMKII activity in vitro and immunoblotting for phosphorylated residues on CaMKII, nitric oxide synthase, and Akt were measured to confirm activity of these enzymes as part of the adrenergic cascade. We demonstrate that stimulation of the β-AR pathway by isoproterenol increased the CaMKII-dependent SR Ca(2+) leak. This increased leak was prevented by inhibition of nitric oxide synthase 1 but not nitric oxide synthase 3. In ventricular myocytes isolated from wild-type mice, isoproterenol stimulation also increased the CaMKII-dependent leak. Critically, in myocytes isolated from nitric oxide synthase 1 knock-out mice this effect is ablated. We show that isoproterenol stimulation leads to an increase in nitric oxide production, and nitric oxide alone is sufficient to activate CaMKII and increase SR Ca(2+) leak. Mechanistically, our data links Akt to nitric oxide synthase 1 activation downstream of β-AR stimulation. Collectively, this evidence supports the hypothesis that CaMKII is regulated by nitric oxide as part of the adrenergic cascade leading to arrhythmogenesis.

Figure 1. Inhibition of NOS attenuates SCaW formation in ISO treated myocytes.

A) Average [Ca]_SRT (n = 34–40) for each treatment (raw data at the top). B) Percentage of myocytes showing at least one SCaW. C) Data in B, normalized to myocyte [Ca]_SRT. D&E) [Ca]_SRT matched data (D) and the average number of SCaWs exhibited (E, n = 13–15). t-test, *p<0.05).

Figure 2. ISO-dependent leak is attenuated by NOS inhibitor, L-NAME.

A) The leak-dependent shift of Ca²⁺ from the cytosol to the SR. Each point represents a loading protocol (from low to high [Ca]_SRT; resting, 1 field stimulation, 0.25 Hz, 0.5 Hz and 1 Hz stimulation, respectively). B) The SR Ca²⁺ leak (right) in [Ca]_SRT matched data (left, n = 10–14). C) The [Ca]_SRT (right) needed to induce the same SR Ca²⁺ leak (left) in leak matched data (left, n = 11–17). *Statistically different from control, #different from ISO (t-test, p<0.05).

Figure 3. Inhibition of NOS1 but not NOS3 reverses the ISO-dependent increase in SR Ca²⁺ leak.

A) Leak/load relationship. B) Matched data such that the average [Ca]_SRT was the same for all treatments (left) and resultant leaks (right, n = 13–17). C) Data matched such that the average SR Ca²⁺ leak was the same for all treatments (left) and the [Ca]_SRT needed to induce that leak (right, n = 11–19). *different from control, # different from ISO (t-test, p<0.05).

Figure 4. NOS1^−/− mice show attenuated CaMKII-dependent leak.

A) Matched data such that [Ca]_SRT was the same for all treatments (left) an resultant SR Ca leaks (right, n = 10–22). B) Matched data such that [Ca]_SRT was the same in NOS1^−/− and NOS1^−/−+SNAP (left) and the resultant SR Ca leaks (right), demonstrates that SNAP restores the leak/load relationship in NOS1^−/− myocytes. C) Summary data (top, n = 4 hearts each) and representative immunoblot (bottom) of phosphorylated RyR at CaMKII-specific residue, Ser2814, and total RyR expression in WT and NOS1^−/− heart lysates. D) Western blots showing total CaMKII normalized to GAPDH (left) and CaMKII phosphorylated at T286 (right, n = 5) in WT and NOS1^−/− hearts. Representative blot showing at bottom. E) Summary data (top) and representative immunoblot (bottom) of oxidize CaMKII after ISO stimulation in WT and NOS1^−/− heart lysates. *Statistically different from control, # different from WT+ISO (t-test, p<0.05).

Figure 5. NO increases CaMKII-dependent SR Ca²⁺ leak.

A) NO-dependent DAF-2 fluorescence (n = 6). Spearman correlation = 1.0 for SNAP, 0.9 for ISO, and −0.05 for control. B) SNAP-dependent SR Ca²⁺ leak. The SR Ca²⁺ leak (right) in [Ca]_SRT matched data (left, n = 9–13). C) Data was matched such that leak was the same (left) with the [Ca]_SRT needed to induced that leak shown on the right (n = 12–17). D) Purified CaMKII pre-activated with 200 µM Ca²⁺ and CaM. H₂O₂ (Lane 2) or 500 µM SNAP (Lane 3) was added followed by EGTA. ATP³² was added along with purified β2a L-type Ca²⁺ channel subunit on nickel beads. Incorporation of P³² was measured as an indicator of Ca-independent sustained kinase activity. Lane 1 is CaMKII without Ca²⁺, CaM, or ATP; Lane 4 is CaMKII without Ca²⁺, CaM, or ATP plus the addition of SNAP (500 µM) alone. Lane 5 is P³² incorporation in the continued presence of Ca²⁺ and CaM. E) Cardiac myocytes were field stimulated at 0.5 Hz under the indicated conditions. CaMKII was then immunoprecipitated from cellular homogenates which were then blotted with antibody to S-NO. *different from ISO, **different from both ISO and control (t-test, p<0.05).

Figure 6. Akt Activates NOS.

A) Western blot indicating that ISO increases Akt phosphorylation at S473 in a dose-dependent manner in isolated rabbit cells. B) ISO-dependent increase in p-Akt is blunted by Akt-Inhibitor X (top, right, * different from control, ** different from ISO, paired t-test, p<0.05). Cells were treated with ISO or ISO+Akt Inhibitor X. Akt Inhibitor X decreased the SR Ca leak and the activation of Akt (top left and bottom). C) Down-regulated Akt expression (plus the constitutively expressed Akt) increased the total Akt over the constitutive expression alone (top). Akt-dn decreased ISO-dependent SR Ca leak when data was selected to give the same average [Ca]_SRT (bottom). D) NOS1 phosphorylation at Akt phosphorylation site S1416, representative immunoblot (left) and summary data (left). (* different from control, ** different from ISO, paired t-test, p<0.05).