Signalling mechanisms in contraction-mediated stimulation of intracellular NO production in cat ventricular myocytes

Abstract

In this study we sought to determine whether contractile activity has a role as a signalling mechanism in the activation of intracellular nitric oxide (NO_i) production induced by electrical stimulation of cat ventricular myocytes. Field stimulation (FS) of single ventricular myocytes elicited frequency-dependent increases in NO_i that were blocked by the calmodulin (CaM) inhibitor 10 μm W-7 and partially inhibited by the phosphatidylinositol 3′-kinase (PI-(3)K) inhibitor 10 μm LY294002. Increasing extracellular [Ca²⁺] caused a concentration-dependent increase in FS-induced NO_i that was partially inhibited by LY294002. The negative inotropic agents BDM (5 mm) or blebbistatin (10 μm) decreased cell shortening and NO_i production without concomitant changes in L-type Ca²⁺ current (I_Ca,L) or [Ca²⁺]_i transients. The positive inotropic agents EMD 57033 or CGP 48506 (1 μm) increased cell shortening and NO_i production without concomitant changes in I_Ca,L or [Ca²⁺]_i transients. FS-induced NO_i production was decreased in myocytes infected (100 multiplicity of viral infection (MOI); 24 h) with a replication-deficient adenovirus expressing a dominant-negative mutant of protein kinase B (Akt) compared with cells infected with a control adenovirus expressing β-galactosidase. FS-induced NO_i was partially inhibited by either endothelial (eNOS) or neuronal nitric oxide synthase (nNOS) inhibitors and completely blocked by simultaneous exposure to both. FS-induced [Ca²⁺]_i transients were increased by the nNOS inhibitor nNOS-I (0.24 μm), decreased by the eNOS inhibitor L-NIO (1 μm) and unchanged by exposure to both inhibitors. We conclude that in cat ventricular myocytes, FS-induced NO_i production requires both Ca²⁺-dependent CaM signalling and Ca²⁺-independent PI-(3)K–Akt signalling activated by contractile activity. FS activates NO_i production from both eNOS and nNOS, and each source of NO_i exerts opposing effects on [Ca²⁺]_i transient amplitude. These findings are important for understanding the regulation of NO_i signalling in the normal and mechanically failing heart.

Nitric oxide (NO) is a fundamental second messenger that participates in a wide variety of cellular functions. In adult rat ventricular myocytes, electrical field stimulation (FS) increases intracellular NO (NO_i) production as detected by nitrite accumulation (Kaye et al. 1996). NO_i production is dependent on elevation of intracellular [Ca²⁺] presumably to stimulate constitutive Ca²⁺–calmodulin (CaM)-dependent NO synthase (NOS) activity. However in endothelial cells, mechanical forces such as tangential sheer stress (Dimmeler et al. 1999; Fulton et al. 1999) and circumferential stretch (Kuebler et al. 2003) stimulate endothelial (eNOS)-dependent NO_i production via activation of phosphatidylinositol 3′-kinase (PI-(3)K)–protein kinase B (Akt) signalling (Fulton et al. 1999). Mechanical forces applied to cardiac muscle also stimulate NO_i production (Pinsky et al. 1997; Prendergast et al. 1997; Vila Petroff et al. 2001). For instance, sustained stretch of stimulated cardiac myocytes requires PI-(3)K–Akt signalling to activate eNOS-dependent NO_i production (Vila Petroff et al. 2001). Results from our previous work also indicate that in cat atrial myocytes, muscarinic (Dedkova et al. 2003), β₂-adrenergic (Wang et al. 2002) and α1-adrenergic (Wang et al. 2005) receptor stimulation requires PI-(3)K–Akt signalling to stimulate NO_i production. In contrast to CaM-dependent activation of constitutive NOS, PI-(3)K–Akt signalling is Ca²⁺-independent (Conus et al. 1998; Dedkova et al. 2003; Boo & Jo, 2003). These findings therefore raise the question of whether FS of cardiac myocytes stimulates NO_i production entirely through a Ca²⁺-dependent process or whether Ca²⁺-independent signalling via PI-(3)K–Akt also contributes to FS-induced NO_i production. Therefore, the primary purpose of the present study was to determine whether a Ca²⁺-independent PI-(3)K–Akt signalling mechanism activated by contractile activity, and acting in conjunction with Ca²⁺–CaM signalling, contributes to NO_i production induced by electrical FS of ventricular myocytes. Part of this work has been published in abstract form (Dedkova et al. 2004).

Methods

Adult cats of either sex were anaesthetized with sodium pentobarbital (50 mg kg⁻¹, i.p.). Once fully anaesthetized, a bilateral thoracotomy was performed, and the heart was rapidly excised and mounted on a Langendorff perfusion apparatus. After enzyme (type II collagenase; Worthington Biochemical) digestion, ventricular myocytes were isolated as previously reported (Rubenstein & Lipsius, 1995). Animal protocols used were approved by the Institutional Animal Care and Use Committee of Loyola University of Chicago, Stritch School of Medicine, Maywood, IL, USA.

Electrophysiological recordings from myocytes were performed using a perforated-patch (nystatin) whole-cell recording method, as previously described (Rubenstein & Lipsius, 1995). CsCl (5 mm) was added to all external solutions to block K⁺ conductances. L-type Ca²⁺ current (I_Ca,L) was activated by depolarizing pulses from a holding potential of −40 to 0 mV for 200 ms every 5 s. Peak I_Ca,L amplitude was measured in relation to steady-state current at the end of the pulse.

Measurements of NO_i production were obtained using the fluorescent NO-sensitive dye 4,5-diaminofluorescein diacetate (DAF-2 DA) (Kojima et al. 1998; Nakatsubo et al. 1998) as previously described (Dedkova & Blatter, 2002; Wang et al. 2002, 2005; Dedkova et al. 2003). NO_i measurements were performed at room temperature. DAF-2 fluorescence was excited at 480 nm and emitted cellular fluorescence was recorded at 540 nm. Changes in cellular DAF-2 fluorescence intensity (F) was normalized to the level of fluorescence recorded prior to stimulation (F_o), and changes in NO_i are expressed as F/F_o. NO-induced DAF-2 fluorescence is irreversible and therefore fluorescence intensity remains constant even if NO_i levels decrease. Changes in FS-induced NO_i production were measured after 3 min of stimulation at each frequency. Control measurements of NO_i at each stimulation frequency obtained in different experiments are grouped together. FS of cells was achieved by 3 ms duration suprathreshold rectangular voltage pulses delivered through a pair of extracellular platinum electrodes.

Fast, one-dimensional (1-D) line scan images of intracellular Ca²⁺ release (i.e. [Ca²⁺]_i transients) were recorded at room temperature using a confocal laser scanning unit (LSM 410, Carl Zeiss, Germany). Fluo-4 (incubation with 10 μm acetoxymethyl ester (AM) form of fluo-4 (fluo-4/AM) for 15 min) was excited with the 488 nm line of an argon ion laser and emitted fluorescence was collected at > 515 nm. The scan line was positioned parallel to the longitudinal axis of the cell (avoiding the nucleus) and scanned repetitively at 6 ms intervals. Line scan profiles are presented as background-subtracted F/F_o. [Ca²⁺]_i transients were recorded from the same cell at three different frequencies in the absence and presence of inotropic agents. Because of the relatively long duration of each experiment (15–20 min), we performed a series of control experiments to determine the effects of photobleaching and/or loss of Ca²⁺ indicator on [Ca²⁺]_i transient amplitudes. [Ca²⁺]_i transients were recorded using the same stimulation parameters and time course as a typical experiment but without drug exposure. The results showed that during the second half of each experiment (during the time when cells were typically exposed to drugs), [Ca²⁺]_i transient amplitudes decreased significantly by 14 ± 5%, 16 ± 3% and 18 ± 1% for each stimulation frequency (0.5, 1.0 and 2.0 Hz), respectively (n = 4). Therefore, these mean values were used to correct [Ca²⁺]_i transient amplitudes at each stimulation frequency for each drug tested. To confirm that these time-dependent changes in [Ca²⁺]_i transients were due to photobleaching and/or loss of Ca²⁺ indicator, we performed additional selected experiments with indo-1/AM, a ratiometric dye for which [Ca²⁺]_i measurements are not affected by changes in dye concentration. The results from indo-1 experiments were not different from the corrected fluo-4 results (data not shown), confirming that photobleaching and/or loss of dye were responsible for the time-dependent decreases in [Ca²⁺]_i transients. Cell shortening of myocytes during FS was determined simultaneously from line scan images.

[Ca²⁺]_i transients also were measured using indo-1/AM, as previously described (Wang et al. 2003). Myocytes were loaded with Ca²⁺ indicator by exposure to 5 μm indo-1/AM in 1 ml Tyrode solution containing 0.001 g ml⁻¹ of Pluronic F-127 for 10 min at room temperature. Cells were washed for 10 min to allow de-esterification of the indicator. For spatially averaged single cell [Ca²⁺]_i measurements, indo-1 fluorescence was excited at 357 nm and cellular fluorescence was recorded simultaneously at 405 nm (F₄₀₅) and 485 nm (F₄₈₅). Changes of [Ca²⁺]_i are expressed as the ratio: R = F₄₀₅/F₄₈₅.

A replication-defective adenovirus expressing a dominant-negative mutant of Akt (Adv-dnAkt) (Fuijo & Walsh, 1999) was obtained from Vector Biolabs, Philadelphia, PA, USA. The mutant consisted of the murine Akt coding sequence fused in frame with the haemaglutinin (HA) epitope, and bearing two mutations (T308A and S473A) rendering the transgene inactive by phosphorylation. Prior studies have demonstrated that this mutant functions in a dominant-negative fashion to inhibit insulin-like growth factor 1 (IGF1)–PI-(3)K–Akt signalling in neonatal and adult cardiomyocytes (Fujio et al. 2000). A control adenovirus expressing nuclear-encoded β-galactosidase (Adv-βgal) was used to control for non-specific effects of adenoviral infection (Heidkamp et al. 2001). Adenoviruses were amplified and purified using HEK293 cells (Eble et al. 1998), and the MOI for each virus was determined by dilution assay in HEK293 cells grown in 96-well clusters. Myocytes were plated in Dulbecco's modified Eagle's medium: Medium 199 (4: 1) culture medium onto laminin-coated glass coverslips, Nunc chamberslides, or 35 mm plastic dishes, and infected (100 MOI, 24 h) with Adv-dnAkt or Adv-βgal. Preliminary experiments using 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-gal) staining of Adv-βgal infected cells determined that a concentration of 100 MOI infected 93 ± 3% (n = 3 experiments, 400–700 cells per experiment) of cultured myocytes.

Immunocytochemistry of infected myocytes was performed using a rhodamine-conjugated mouse anti-HA monoclonal antibody (Roche Applied Science, Indianapolis, IN, USA). Cells were fixed, stained and viewed using a Zeiss LSM 510 confocal microscope, as previously described (Bayer et al. 2001). Western blotting of cell extracts obtained from Adv-dnAkt- and Adv-βgal-infected cells was performed using 10% SDS-polyacrylamide gels. Blots were probed with either rabbit anti-Akt polyclonal antibody (Cell Signalling, Beverly, MA, USA) followed by peroxidase-conjugated goat anti-rabbit IgG or rat anti-HA (Roche Applied Science) followed by peroxidase-conjugated goat anti-rat IgG, and visualized by enhanced chemiluminescence (Amersham, Arlington Heights, IL, USA).

Drugs and chemicals in this study included: LY294002 (LY) and N-(6-aminohexyl)-5-chloro-1-naphthalene sulphonamide hydrochloride (W-7) from Sigma Chemicals; DAF-2 DA, blebbistatin, l-N5-(1-iminoethyl) ornithine (L-NIO), (4S)-N-(4-amino-5[aminoethyl]aminopentyl)-N′-nitroguanidine (nNOS inhibitor I) and N^G-allyl-l-arginine (L-ALA) from EMD Calbiochem; and S-methyl-l-thiocitrulline (MTC) (Alexis Biochemical). EMD 57033 was generously provided by Dr Norbert Beier, Merck, Darmstadt, Germany. CGP 48506 was generously provided by Dr R. J. Solaro, University of Illinois, Chicago, IL, USA.

Data are presented as the mean ± s.e.m. Test measurements were compared to controls and analysed using either paired or unpaired Student's t test for significance at P < 0.05.

Results

Figure 1A shows original recordings of NO_i production obtained from two different ventricular myocytes during electrical FS at three different frequencies (0.5, 1.0 and 2.0 Hz). In control (cntrl), each increase in stimulation frequency elicited an increase in NO_i production. In another ventricular myocyte, prior exposure to 10 μm W-7, a potent inhibitor of CaM, completely blocked NO_i production at each stimulation frequency. Figure 1B summarizes the control frequency-dependent increases in NO_i production and the effects of W-7 to block NO_i production at each stimulation frequency. Because the main hypothesis of the present study is that contraction modulates frequency-dependent NO_i production, we determined the effects of W-7 on contraction strength at each stimulation frequency. As shown in Fig. 1C, W-7 elicited a significant decrease in contraction at each stimulation frequency. This finding is consistent with the effects of CaM-dependent kinases to modulate key steps in cardiac excitation–contraction (E–C) coupling (Rapundalo, 1998). These results could be interpreted to indicate that the effect of W-7 to inhibit NO_i production was at least partially mediated by inhibition of contraction. However, we will show from the following experiments that this level of negative inotropy cannot explain the effect of W-7 to completely abolish FS-induced NO_i production. It therefore seems likely that the effect of W-7 to abolish NO_i production is mediated primarily by inhibition of CaM-dependent NOS activity.

Figure 1. NO_i production recorded following field stimulation (FS; 0.5, 1 and 2 Hz) ventricular myocytes.

A, NO_i was measured under control (cntrl) conditions, and in the presence of 10 μm W-7 in two different ventricular myocytes. W-7 abolished FS-induced NO_i production at each stimulation frequency. B, summary shows that compared to controls (cntrl), W-7 significantly inhibited NO_i production at each stimulation frequency. C, summary shows that compared to controls (cntrl) W-7 also significantly inhibited contraction. *P < 0.05.

Figure 2A shows original recordings of the concentration-dependent effects of extracellular [Ca²⁺] ([Ca²⁺]_o) on NO_i production in two different ventricular myocytes during FS at 1 Hz. In control (cntrl), step increases in [Ca²⁺]_o (0.5, 1.0, 2.0 and 4.0 mm) elicited step increases in NO_i production. Figure 2B summarizes the increases in control NO_i production at each [Ca²⁺]_o (open bars). This finding and those shown in Fig. 1 are consistent with the Ca²⁺-dependence of NOS and NO_i production. However, as shown in Fig. 2C (cntrl) and summarized in Fig. 2D (cntrl; open bars), each increase in [Ca²⁺]_o also increased contraction strength, as expected. This raises the question of whether mechanical activity contributed a signalling mechanism that, in conjunction with Ca²⁺ signalling, activates NO_i production. Because mechanical activity can mediate PI-(3)K-dependent NO_i production (Fulton et al. 1999; Vila Petroff et al. 2001), we repeated the experiments in the presence of 10 μm LY, a specific inhibitor of PI-(3)K (Vlahos et al. 1994). Recordings from another cell (Fig. 2A) show that LY markedly inhibited the [Ca²⁺]_o-mediated increase in NO_i production compared to controls. Figure 2B summarizes the inhibitory effects of LY on NO_i production (filled bars) compared with controls (open bars). These findings indicate that full activation of [Ca²⁺]_o-mediated NO_i production requires PI-(3)K signalling. Moreover, they suggest that the LY-insensitive NO_i production is mediated via Ca²⁺-dependent stimulation of NOS activity. However, as shown in Fig. 2C, LY also markedly increased contraction strength at each [Ca²⁺]_o. Figure 2D summarizes the significant positive inotropic effects of LY at each [Ca²⁺]_o compared with controls. This positive inotropic effect of LY raises the question of whether the residual LY-insensitive NO_i production may be mediated by the enhanced contraction acting via PI-(3)K signalling that LY is unable to completely block.

Figure 2. Effects of [Ca²⁺]_o on field stimulation (FS)-induced NO_i production in the absence (cntrl) and presence of LY294002.

A, original recordings of NO_i production from ventricular myocytes following FS at 1 Hz. Step increases in [Ca²⁺]_o (0.5, 1, 2 and 4 mm) increased NO_i production (cntrl). In another ventricular myocyte, 10 μm LY294002 partially inhibited NO_i production. B, summary shows that step increases in [Ca²⁺]_o elicited increases in NO_i production (open bars) and that LY294002 significantly inhibited NO_i production at each [Ca²⁺]_o (filled bars). C, original recordings of cell shortening at each [Ca²⁺]_o in the absence (cntrl) and presence of LY294002. LY294002 increases contraction strength. D, summary shows that compared with controls (open bars), LY294002 significantly increased contraction amplitudes at each [Ca²⁺]_o (filled bars). *P < 0.05.

The LY-induced positive inotropic effect is consistent with reports that PI-(3)K inhibits cAMP-mediated modulation of E–C coupling (Kerfant et al. 2004; Leblais et al. 2004; Alloatti et al. 2005). Therefore, as shown in Fig. 3, we determined the effects of LY on [Ca²⁺]_o-mediated NO_i production and contraction in the presence of 5 μm H-89, an inhibitor of cAMP-dependent protein kinase A (PKA) activity (Chijiwa et al. 1990). In other words, this experiment was designed to determine whether the effects of LY on FS-induced NO_i production were influenced by the positive inotropic effects of LY. In Fig. 3A, control responses showed a typical [Ca²⁺]_o-mediated increase in NO_i production. In another myocyte, LY plus H-89 markedly inhibited the increase in NO_i production elicited by step increases in [Ca²⁺]_o. This attenuated response to [Ca²⁺]_o was similar to that elicited by LY alone (see Fig. 2A). Moreover, as shown in Fig. 3C, H-89 blocked the positive inotropic effects of LY. The bar graphs in Fig. 3B and D summarize the effects of LY plus H-89 on [Ca²⁺]_o-mediated NO_i production and contraction, respectively. They show that inhibition of PI-(3)K and cAMP–PKA signalling significantly attenuates Ca²⁺-dependent NO_i production, similar to that elicited by PI-(3)K inhibition alone (compare Figs 3B and 2B). In addition, H-89 prevented the positive inotropic effects of LY (compare Figs 3D and 2D). The fact that LY elicited a similar inhibition of [Ca²⁺]_o-mediated NO_i production in the absence or presence of H-89, indicates that the positive inotropic effects of LY did not influence the results. Therefore it seems likely that LY inhibited NO_i production via inhibition of PI-(3)K signalling and that the residual NO_i production was due to Ca²⁺–CaM-dependent activation of NOS activity. These findings therefore indicate that PI-(3)K signalling is required for full activation of Ca²⁺–CaM-dependent NO_i production. Because PI-(3)K signalling is activated by mechanical activity and is Ca²⁺-independent, these results support the idea that beat-to-beat contractile activity stimulates a Ca²⁺-independent signalling mechanism that is required, in conjunction with Ca²⁺–CaM-dependent signalling, to activate FS-induced NO_i production.

Figure 3. Effects of [Ca²⁺]_o on field stimulation (FS)-induced NO_i production in the absence (cntrl) and presence of LY294002 + H-89.

A, original recordings of NO_i production from ventricular myocytes following FS at 1 Hz. Step increases in [Ca²⁺]_o (0.5, 1, 2 and 4 mm) increased NO_i production (cntrl). In another ventricular myocyte, 10 μm LY294002 + 5 μm H-89 markedly inhibited NO_i production, in a similar manner to LY294002 alone (see Figure 2). B, summary shows that step increases in [Ca²⁺]_o elicited increases in NO_i production (open bars), and LY294002 + H-89 significantly inhibited NO_i production at each [Ca²⁺]_o (filled bars). C, original recordings of cell shortening at each [Ca²⁺]_o in the absence (cntrl) and presence of LY294002 + H-89. H-89 prevented the positive inotropic effects of LY294002. D, summary shows that contraction amplitudes in the absence (open bars) and presence of LY294002 + H-89 (filled bars) were not significantly different. *P < 0.05.

In the next series of experiments, we therefore sought to determine whether contractile activity, acting via Ca²⁺-independent PI-(3)K signalling, contributes to FS-induced NO_i production. Our first approach was to use agents that modify contractile activity without significantly affecting Ca²⁺ homeostasis. Figure 4A shows original recordings of NO_i obtained from three different ventricular myocytes. Control (cntrl) recordings show a typical frequency-dependent increase in NO_i production. BDM is a negative inotropic agent that is thought to uncouple contractile activation from excitation (Gwathmey et al. 1991; Backx et al. 1994). At relatively low concentrations (≤ 5 mm), BDM decreases contractile activity with little effect on intracellular Ca²⁺ homeostasis (Byron et al. 1996). Prior exposure to 5 mm BDM decreased FS-induced NO_i production at each stimulation frequency. EMD 57033 is a positive inotropic agent that increases the Ca²⁺ sensitivity of contractile proteins without affecting intracellular Ca²⁺ handling (Gambassi et al. 1993; Solaro et al. 1993). As shown in Fig. 4A, exposure of another ventricular myocyte to 1 μm EMD 57033 increased NO_i production at each stimulation frequency. As summarized in Fig. 4B, BDM and EMD 57033 significantly decreased and increased, respectively, FS-induced NO_i production at each stimulation frequency, compared with control.

Figure 4. Effects of negative (BDM) and positive (EMD 57033) inotropic agents on field stimulation (FS)-induced NO_i production.

A, original recording of NO_i production from three different ventricular myocytes following FS at 0.5, 1.0 and 2.0 Hz. Control (cntrl) recordings show typical frequency-dependent increases in NO_i production. Compared to control, BDM (5 mm) decreased and EMD 57033 (1 μm) increased NO_i production at each stimulation frequency. B, summary shows significant decrease (BDM) and increase (EMD 57033) in NO_i production at each stimulation frequency compared with control. *P < 0.05.

The present results suggest that contractile activity contributes to the signalling mechanisms that regulate FS-induced NO_i production. However, because contractile activity and NO_i production are both Ca²⁺-dependent processes it becomes important to determine whether each inotropic agent altered contractile activity independently of concomitant changes in Ca²⁺ handling. Therefore, in the following experiments we determined the effects of each inotropic agent on I_Ca,L, [Ca²⁺]_i transient and contraction amplitudes. Figure 5 shows the effects of 5 mm BDM on I_Ca,L (Fig. 5A and B), [Ca²⁺]_i transients (Fig. 5C and D) and cell shortening (Fig. 5E). As shown in Fig. 5A and summarized in Fig. 5B, BDM had no effect on peak I_Ca,L amplitude. In Fig. 5C, compared to controls, BDM elicited little effect on [Ca²⁺]_i transient amplitudes. As summarized in Fig. 5D and E, BDM exerted no significant effects on [Ca²⁺]_i transient amplitudes (Fig. 5D), while significantly decreasing contraction amplitude (Fig. 5E) at each stimulation frequency.

Figure 5. Effects of BDM (5 mm) on I_Ca,L (A and B), [Ca²⁺]_i transients (C and D) and contraction amplitudes (E).

A, original recordings of I_Ca,L in the absence (cntrl) and presence of BDM. BDM had no effect on I_Ca,L. B, summary shows that BDM had no significant effects on peak I_Ca,L amplitude (n = 4). C, original recordings of line scan images of [Ca²⁺]_i (upper panels) and [Ca²⁺]_i transients (lower traces) at each stimulation frequency in the absence (control) and presence of BDM. D, summary shows that BDM had no significant effect on [Ca²⁺]_i transients at each stimulation frequency (n = 5). E; summary shows that BDM significantly decreased contraction amplitudes at each stimulation frequency (n = 5). *P < 0.05.

Figure 6 shows the results of similar experiments designed to determine the effects of the positive inotropic agent EMD 57033. EMD 57033 had no effect on either I_Ca,L (Fig. 6A and B) or [Ca²⁺]_i transient amplitudes at each stimulation frequency (Fig. 6C and D). However, EMD 57033 significantly increased contraction amplitude at each stimulation frequency (Fig. 6E).

Figure 6. Effects of EMD 57033 (1 μm) on I_Ca,L (A and B), [Ca²⁺]_i transients (C and D) and contraction amplitudes (E).

A, original recordings of I_Ca,L in the absence (cntrl) and presence of EMD 57033. EMD 57033 had no effect on I_Ca,L. B, summary shows that EMD 57033 had no significant effects on peak I_Ca,L amplitude (n = 4). C, original recordings of line scan images of [Ca²⁺]_i (upper panels) and [Ca²⁺]_i transients (lower traces) at each stimulation frequency in the absence (control) and presence of EMD 57033. D, summary shows that EMD 57033 had no significant effects on [Ca²⁺]_i transients at each stimulation frequency. E, summary shows that EMD 57033 significantly increased contraction amplitudes at each stimulation frequency. In D and E: 0.5 Hz, n = 5; 1 Hz, n = 5; 2 Hz, n = 3. *P < 0.05.

In additional experiments we tested the effects of blebbistatin, another negative inotropic agent that inhibits myosin ATPase (Kovacs et al. 2004). The data summarized in Fig. 7 indicate that 10 μm blebbistatin (Blebb) significantly decreased FS-induced NO_i production (Fig. 7A) and contraction amplitude (Fig. 7D) at each stimulation frequency without exerting any significant effects on I_Ca,L (Fig. 7B) or [Ca²⁺]_i transient amplitudes (Fig. 7C).

Figure 7. Effects of blebbistatin (Blebb; 10 μm) on NO_i production (A), I_Ca,L (B), [Ca²⁺]_i transients (C) and contraction amplitudes (D).

Blebbistatin significantly decreased NO_i production (A) and contraction at each stimulation frequency (D) but had no significant effects on I_Ca,L (B) or [Ca²⁺]_i transients (C) at each stimulation frequency. *P < 0.05.

Figure 8 shows additional results obtained with CGP 48506, another positive inotropic agent that acts as a Ca²⁺ sensitizer (Wolska et al. 1996). Compared with controls, 1 μm CGP 48506 significantly increased FS-induced NO_i production (Fig. 8A) and contraction amplitudes (Fig. 8D) at each stimulation frequency without exerting any significant effects on I_Ca,L (Fig. 8B) or [Ca²⁺]_i transient amplitudes (Fig. 8C). Together, the present findings indicate that contractile activity modulates endo-genous NO_i production independently of concomitant changes in Ca²⁺ influx via I_Ca,L or sarcoplasmic reticulum (SR) Ca²⁺ release. In other words, contraction-mediated regulation of NO_i production acts via a Ca²⁺-independent signalling mechanism.

Figure 8. Effects of CGP 48507 (1 μm) on NO_i production (A), I_Ca,L (B), [Ca²⁺]_i transients (C) and contraction amplitudes (D).

CGP 48507 significantly increased NO_i production (A) and contraction (D) at each stimulation frequency but had no significant effects on I_Ca,L (B) or [Ca²⁺]_i transients (C) at each stimulation frequency. *P < 0.05.

As stated earlier, PI-(3)K–Akt signalling is a Ca²⁺-independent mechanism that is required for mechanically induced NO_i production in both endo-thelial and cardiac cells. To determine whether FS-induced NO_i production requires PI-(3)K signalling, we measured NO_i production in the absence and presence of LY. In Fig. 9A control (cntrl) recordings show typical frequency-dependent increases in NO_i production. In another myocyte, prior exposure to 10 μm LY markedly inhibited NO_i production at each stimulation frequency. The summary graph in Fig. 9B indicates that LY significantly inhibited NO_i production at each stimulation frequency. These findings are similar to those in which LY significantly inhibited [Ca²⁺]_o-dependent NO_i production (Fig. 2). Figure 9C shows that LY had no discernable effects on [Ca²⁺]_i transient amplitudes although it significantly increased diastolic [Ca²⁺]_i at each stimulation frequency (0.5 Hz, + 27 ± 4%; 1 Hz, 18 ± 5%; 2 Hz, 19 ± 4%; n = 4; P < 0.05 at each frequency). The summary graph in Fig. 9D indicates that LY had no significant effects on [Ca²⁺]_i transient amplitudes. However, as discussed earlier (Fig. 2), LY significantly increased contraction strength at each stimulation frequency (Fig. 9E).

Figure 9. Field stimulation (FS)-induced NO_i production in the absence (cntrl) and presence of LY294002.

A, original recordings of NO_i production obtained from two separate ventricular myocytes. Compared to control (cntrl), LY294002 partially inhibited NO_i production at each stimulation frequency. B, summary shows that LY294002 significantly inhibited NO_i production at each stimulation frequency. C, original recordings of line scan images of [Ca²⁺]_i (upper panels) and [Ca²⁺]_i transients (lower traces) at each stimulation frequency in the absence (control) and presence of LY294002. LY294002 had no effect on [Ca²⁺]_i transient amplitude but increased baseline diastolic [Ca²⁺]_i (dashed lines). D, summary shows that LY294002 had no significant effects on [Ca²⁺]_i transients at each stimulation frequency. E, summary shows that LY294002 significantly increased contraction amplitudes at each stimulation frequency.*P < 0.05.

Once again, the positive inotropic effects of LY raise the question of whether the LY-insensitive NO_i production is mediated by the enhanced LY-induced contraction. Therefore, as shown in Fig. 10, we determined the effects of LY in the presence of H-89 to inhibit cAMP-dependent PKA-mediated stimulation of contraction. As shown in Fig. 10A and summarized in Fig. 10B, LY plus H-89 significantly attenuated FS-induced NO_i production at each stimulation frequency, similar to that elicited by LY alone (see Fig. 9A and B). Figure 10C shows that LY plus H-89 has no effects of [Ca²⁺]_i transient amplitudes although the kinetics of the transients appear slowed. In addition, H-89 blocked the increase in diastolic [Ca²⁺]_i elicited by LY alone (see Fig. 9C). The summary graph in Fig. 10D shows that LY plus H-89 had no significant effects on [Ca²⁺]_i transient amplitudes. Furthermore, as shown in Fig. 10E, H-89 prevented the positive inotropic effects of LY at each stimulation frequency. Together, these findings indicate that LY significantly attenuates FS-induced NO_i production via a Ca²⁺-independent mechanism (i.e. PI-(3)K signalling). Moreover, the LY-insensitive NO_i production was not due to the positive inotropic effects of LY but rather to Ca²⁺–CaM-dependent activation of NOS activity. These findings also confirm that PI-(3)K signalling inhibits basal cAMP-dependent PKA activity, as previously reported (Kerfant et al. 2004; Leblais et al. 2004; Alloatti et al. 2005).

Figure 10. Field stimulation (FS)-induced NO_i production in the absence (cntrl) and presence of LY294002 + H-89.

A, original recordings of NO_i production from two separate ventricular myocytes. Compared to control (cntrl), exposure to LY294002 + H-89 partially inhibited NO_i production at each stimulation frequency. B, summary shows that compared to controls (open bars) LY294002 + H-89 (filled bars) significantly inhibited FS-induced NO_i production, in a similar manner to LY294002 alone (see Figure 9). C, original recordings of line scan images of [Ca²⁺]_i (upper panels) and [Ca²⁺]_i transients (lower traces) at each stimulation frequency in the absence (control) and presence of LY294002 + H-89. LY294002 + H-89 had no effect on [Ca²⁺]_i transient amplitudes but abolished the increase in baseline diastolic [Ca²⁺]_i. D, summary shows that compared to controls (open bars) LY294002 + H-89 (filled bars) had no effect on [Ca²⁺]_i transients. E, summary shows that contraction amplitudes in the absence (open bars) and presence (filled bars) of LY294002 + H-89 were not significantly different. *P < 0.05.

Akt is a downstream target of PI-(3)K involved in the activation of eNOS-dependent NO_i production (Dimmeler et al. 1999; Fulton et al. 1999; Vila Petroff et al. 2001). Therefore, in another approach we infected (100 MOI; 24 h) cardiac myocytes with either Adv-βgal or Adv-dnAkt, and analysed transgene expression and NO_i production after overnight culture. Figure 11A shows the faint outline of a cell infected with Adv-βgal (dashed line in upper panel), and a cell infected with Adv-dnAkt (lower panel). Compared with the cell infected with Adv-βgal, the cell infected with Adv-dnAkt expressed high levels of the dominant-negative transgene which was localized diffusely throughout the cytoplasm. In Fig. 11B, Western blots (WB) show the presence of the 60 kDa Akt mutant that cross-reacted with both the Akt and HA antibody, and which displays a similar apparent molecular weight as the endogenous enzyme. Figure 11C shows original recordings of NO_i production elicited at 0.5 and 1 Hz in myocytes uninfected (cntrl) or infected with either Adv-βgal (βgal) or Adv-dnAkt (dnAkt). Uninfected cells showed similar responses to cells infected with Adv-βgal (traces superimposed). In myocytes overexpressing Adv-dnAkt, NO_i production was marked inhibited compared with cells infected with Adv-βgal. The summary in Fig. 11D shows that FS-induced NO_i production at each stimulation frequency was similar in cells infected with Adv-βgal compared to uninfected cells and significantly inhibited in cells infected with Adv-dnAkt compared to either control or Adv-βgal. These findings, along with those shown in Figs 9 and 10, indicate that full activation of FS-induced NO_i production requires upstream activation of PI-(3)K–Akt signalling.

Figure 11. Effects of adenoviral infection with a dominant-negative Akt mutant on field stimulation (FS)-induced NO_i production.

A, immunocytochemistry of infected myocytes showed that compared with cells infected with Adv-βgal (upper panel; dashed line marks cell border), cells infected with Adv-dnAkt (lower panel) expressed high levels of the dominant-negative transgene localized diffusely throughout the cytoplasm. B, Western blots (WB) show the presence of the 60 kDa Akt mutant, which cross-reacted with both the Akt and HA antibody. C, NO_i production elicited at 0.5 and 1 Hz in an uninfected (cntrl) myocyte, and myocytes infected with either Adv-βgal (βgal) or Adv-dnAkt (dnAkt). Uninfected cells showed similar responses to cells infected with βgal. Over-expression of dnAkt prevented FS-induced NO_i production. D, summary shows that FS-induced NO_i production at each stimulation frequency was significantly inhibited in cells infected with dnAkt compared to controls. Numbers in parentheses indicate the number of cells studied. *P < 0.05.

Cardiac muscle contains eNOS and nNOS, both of which are constitutive Ca²⁺-dependent NOSs. This raises the question of which NOS isoform is activated by FS. We approached this question by employing different NOS blocking agents that are relatively selective for either eNOS or nNOS. We tested one eNOS inhibitor (L-NIO) and three different nNOS inhibitors (nNOS-I, L-ALA and MTC). Of the three nNOS inhibitors, nNOS-I is the most selective inhibitor being 2500 times more selective for nNOS than eNOS (Hah et al. 2001). The concentration of each blocking agent used was 2-fold higher than its reported IC₅₀ value. Figure 12A shows the effects of L-NIO and nNOS-I on NO_i production elicited by 0.5, 1 and 2 Hz stimulation. Control (cntrl) recordings show typical frequency-dependent increases in NO_i production. In a second cell, the presence of 1 μm L-NIO partially inhibited NO_i production at each stimulation frequency. Likewise, in a third cell 0.24 μm nNOS-I partially inhibited NO_i production at each stimulation frequency. In a fourth cell, simultaneous exposure to L-NIO plus nNOS-I essentially blocked NO_i production at each stimulation frequency. The bar graph in Fig. 12B summarizes the results obtained with all blocking agents at each stimulation frequency. Compared to control (cntrl), 1 μm L-NIO, 0.24 μm nNOS-I, 0.5 μm L-ALA and 0.6 μm MTC all exerted partial and significant inhibition of FS-induced NO_i production at each stimulation frequency. Exposure to L-NIO plus nNOS-I exerted almost complete block of FS-induced NO_i production at each stimulation frequency. These findings indicate that FS-induced NO_i production results from activation of both eNOS and nNOS and each of the two NOS isoforms contribute approximately half of the total NO_i production.

Figure 12. Effects of eNOS and nNOS inhibitors on field stimulation (FS)-induced NO_i production.

A, original recordings of NO_i production obtained from four different myocytes, each stimulated at 0.5, 1 and 2 Hz. Control (cntrl) recordings show typical frequency-dependent increases in NO_i production. L-NIO (1 μm) and nNOS-I (0.24 μm) each partially inhibited NO_i production at each stimulation frequency. Simultaneous exposure to L-NIO plus nNOS-I blocked NO_i production at each stimulation frequency. B, bar graph summarizing the data shows that compared to control (cntrl) NO_i production elicited by FS at 0.5, 1 and 2 Hz, exposure to L-NIO, nNOS-I, L-ALA and MTC each elicited partial and significant inhibition of NO_i production at each stimulation frequency. Simultaneous exposure to L-NIO plus nNOS-I blocked NO_i production at each stimulation frequency. Numbers in parentheses indicate the number of cells studied. All values are significantly different from their respective controls.

The present results indicate that changes in contractile force elicit significant changes in NO_i production and yet [Ca²⁺]_i transient amplitudes are not significantly affected. This may suggest that endogenous NO_i release does not affect SR Ca²⁺ release. On the other hand, endogenous NO_i generated by the two different NOS isoforms may exert opposing effects on SR Ca²⁺ release, resulting in no net change. Therefore, as shown in Fig. 13, we determined the effect of 0.24 μm nNOS-I (nNOS inhibitor I) or 1 μm L-NIO (eNOS inhibitor) on FS-induced [Ca²⁺]_i transients. In Fig. 13A, compared to its own control (Fig. 13Aa), exposure to nNOS-I (Fig. 13Ab) increased [Ca²⁺]_i transient amplitude, and in Fig. 13B, compared to its own control (Fig. 13Ba), exposure to L-NIO decreased [Ca²⁺]_i transient amplitude (Fig. 13Bb). If the effects of NO_i derived from each NOS isoform oppose one another, then blocking NO_i release from both sources should result in no net effect on [Ca²⁺]_i transients. As shown in Fig. 13C, compared to its own control (Fig. 13Ca), simultaneous exposure to nNOS-I plus L-NIO (Fig. 13Cb) exerted little effect on [Ca²⁺]_i transients. The bar graph (Fig. 13D) summarizes the percentage change in [Ca²⁺]_i transient amplitudes in relation to a normalized control (100%). Inhibition of nNOS (nNOS-I) significantly increased [Ca²⁺]_i transient amplitude (+33 ± 10%) and inhibition of eNOS (L-NIO) significantly decreased [Ca²⁺]_i transient amplitude (−32 ± 6%). Blocking both NOS isoforms had no significant effect on [Ca²⁺]_i transient amplitude (−6 ± 3%). These findings therefore suggest that the opposing effects of NO_i derived from eNOS and nNOS on [Ca²⁺]_i transients cancel each other and are consistent with the finding that modulation of NO_i production by changes in contraction exerted no significant effects on SR Ca²⁺ release.

Figure 13. Effects of nNOS and eNOS inhibition on [Ca²⁺]_i transients.

A, compared to its own control (a), 0.24 μm nNOS-I increased [Ca²⁺]_i transient amplitude (b). B, compared to its own control (a), 1 μm L-NIO decreased [Ca²⁺]_i transient amplitude (b). C, compared to its own control (a), simultaneous exposure to nNOS-I plus L-NIO had no effect on [Ca²⁺]_i transient amplitude (b). D, bar graph summarizing the data shows that compared to control, nNOS-I significantly increased, L-NIO significantly decreased and simultaneous exposure to both inhibitors had no significant effects on [Ca²⁺]_i transient amplitude. Numbers in parentheses indicate the number of cells studied. *P < 0.05.

Discussion

In the present study, FS of cat ventricular myocytes stimulated NO_i production. This finding is similar to that reported in adult rat ventricular myocytes (Kaye et al. 1996), although in the present study NO_i was measured directly using an intracellular NO-sensitive dye. The new findings of the present study are: (1) that NO_i production induced by electrical FS requires both Ca²⁺-dependent CaM signalling and Ca²⁺-independent PI-(3)K–Akt signalling activated by contractile activity; and (2) that FS-induced NO_i production results from activation of both eNOS and nNOS, and each source of NO_i exerts opposing effects on SR Ca²⁺ release.

In the study by Kaye et al. (1996), nitrite release was correlated with pacing-induced increases in [Ca²⁺]_i and was significantly attenuated by chelation of intracellular Ca²⁺. These authors concluded that pacing-induced release of NO_i is regulated by Ca²⁺-dependent activation of constitutive eNOS activity, consistent with the Ca²⁺–CaM dependence of constitutive NOS activities. The present experiments also show that FS-induced NO_i production is blocked by inhibition of Ca²⁺-dependent CaM (W-7) and directly increased by increasing stimulation frequency or by increasing [Ca²⁺]_o (both of which increase Ca²⁺ influx). Therefore, in cat ventricular myocytes FS-induced NO_i production requires Ca²⁺–CaM signalling. Our previous work in cat atrial myocytes indicates that Ca²⁺–CaM signalling is also required for receptor-mediated stimulation of eNOS and NO_i production (Dedkova et al. 2003; Wang et al. 2005).

However, Ca²⁺ influx and SR Ca²⁺ release also activate contraction, and mechanical forces are known to stimulate NO_i production in both endothelial (Dimmeler et al. 1999; Fulton et al. 1999; Kuebler et al. 2003) and cardiac (Prendergast et al. 1997; Vila Petroff et al. 2001) cells. In the present study, interventions that either decreased (BDM or blebbistatin) or increased (EMD 57033 or CGP 48506) contractile strength elicited concomitant changes in FS-induced NO_i production without concomitant changes in Ca²⁺ influx via I_Ca,L or changes in the [Ca²⁺]_i transient. Although W-7 and the two negative inotropic drugs (BDM and blebbistatin) all elicited similar negative inotropic effects, only W-7 completed blocked FS-induced NO_i production. Moreover, our previous studies have shown that W-7 also abolishes the effects of acetylcholine (Dedkova et al. 2003) and phenylephrine (Wang et al. 2005) to stimulate NO_i production in atrial myocytes. It therefore seems likely that the primary mechanism by which W-7 blocked NO_i production was through inhibition of CaM-dependent NOS activation rather than via inhibition of contraction-mediated signalling. The present study also showed that inhibition of PI-(3)K signalling by LY294002 or inhibition of Akt by over-expression of a non-phosphorylatable, dominant-negative mutant of Akt each significantly inhibited FS-induced NO_i production. These results are consistent with those reported in endothelial and cardiac cells showing that mechanical forces act via Ca²⁺-independent PI-(3)K–Akt signalling to stimulate NO_i release. Shear stress on human endothelial cells (Dimmeler et al. 1999) and sustained stretch of cardiac muscle (Vila Petroff et al. 2001) each acts via PI-(3)K–Akt to phosphorylate Ser₁₁₇₇ on eNOS. It is important to note, however, that in the present study inhibition of Ca²⁺-dependent CaM by W-7 abolished FS-induced NO_i production (Fig. 1) whereas inhibition of PI-(3)K signalling by LY294002 significantly inhibited but did not abolish FS-induced NO_i production (Figs 9 and 10) or Ca²⁺-mediated NO_i production (Fig. 2). In other words, Ca²⁺–CaM signalling is absolutely required for activation of FS-induced NO_i production whereas PI-(3)K signalling is required for full activation of CaM-dependent NO_i production. The idea that PI-(3)K–Akt signalling modulates Ca²⁺-dependent NO_i production is consistent with reports that in endothelial cells activation of PI-(3)K–Akt signalling enhances the sensitivity of eNOS to Ca²⁺ (Dimmeler et al. 1999) as well as increases eNOS catalytic activity, and reduces the dissociation of CaM from eNOS (McCabe et al. 2000). This dual mechanism of PI-(3)K–Akt and Ca²⁺–CaM signalling can readily explain the frequency-dependence of FS-induced NO_i production, as shown in the schematic diagram in Fig. 14. FS elevates [Ca²⁺]_i through Ca²⁺ influx via I_Ca,L and SR Ca²⁺ release. Elevation of [Ca²⁺]_i stimulates both Ca²⁺-dependent CaM signalling as well as contractile strength. Contraction-mediated stimulation of PI-(3)K–Akt signalling functions to set the gain of Ca²⁺–CaM-dependent NOS activity. We therefore conclude that beat-to-beat contraction of ventricular myocytes activates a Ca²⁺-independent signalling component that in conjunction with Ca²⁺-dependent CaM signalling is required for full activation of FS-induced NO_i production.

Figure 14. Schematic diagram of proposed signalling mechanisms responsible for field stimulation (FS)-induced NO_i production in ventricular myocytes.

FS increases [Ca²⁺]_i by Ca²⁺ influx via L-type Ca²⁺ current (I_Ca,L) and Ca²⁺ release from the sarcoplasmic reticulum (SR). Elevation of [Ca²⁺]_i stimulates both Ca²⁺-dependent calmodulin (CaM) signalling and contraction. CaM stimulates constitutive endothelial (eNOS) and neuronal nitric oxide synthase (nNOS) isoforms. Contraction activates phosphatidylinositol 3′-kinase (PI-(3)K)–protein kinase B (Akt) possibly via stimulation of β₁-integrin–focal adhesion kinase (FAK) signalling. PI-(3)K–Akt signalling sets the sensitivity of Ca²⁺-dependent CaM-mediated stimulation of eNOS and nNOS activities, and hence the level of FS-induced NO_i production. Each box indicates the pharmacological and molecular tools used in the present study.

The mechanisms by which contractile activity stimulates PI-(3)K signalling remain unclear. Of the four different PI-(3)K subtypes described, three are activated by tyrosine kinases and one is activated by G-protein coupled receptors (see Schaller, 2001; Crackower et al. 2002). As depicted in Fig. 14, one possible mechanism may involve mechanical signal transduction via integrins and activation of focal adhesion kinase (FAK), a non-receptor tyrosine kinase (see Samarel, 2005). FAK (and its related kinase proline-rich tyrosine kinase 2 (PYK2)) can bind the p85 subunit of PI-(3)K and thereby activate downstream Akt signalling (Schaller, 2001). In neonatal rat ventricular myocytes, cyclic stretch increases FAK phosphorylation which is blocked by expression of a dominant-negative FAK mutant (Torsoni et al. 2003). Moreover, in cat atrial myocytes, stimulation of β₁-integrin receptors stimulates FAK-dependent PI-(3)K–Akt activity (Lipsius et al. 2006). The results of a recent study of cultured neonatal rat ventricular myocytes indicate that stimulation of integrin signalling by RGD pentapeptide acts via FAK signalling to stimulate NO_i production (van der Wees et al. 2006), although in this cell type PI-(3)K signalling did not appear to be involved.

The present study also shows that inhibition of PI-(3)K signalling increased contraction strength and this effect was blocked by inhibition of cAMP-dependent PKA activity. It is interesting that inhibition of PI-(3)K had no effect on [Ca²⁺]_i transient amplitude (or I_Ca,L amplitude; authors' unpublished observation). These findings are consistent with reports that PI-(3)K inhibits a compartmentalized pool of cAMP (Kerfant et al. 2004). It seems likely that the positive inotropic effects of PI-(3)K inhibition are related to the concomitant increase in diastolic [Ca²⁺]_i because both effects were blocked by inhibition of cAMP-dependent PKA signalling.

The results of the present experiments indicate that FS-induced NO_i production results from activation of both eNOS and nNOS. As stated earlier, Akt mediates phosphorylation of eNOS in both endothelial cells (Dimmeler et al. 1999; Fulton et al. 1999) and cardiac myocytes (Vila Petroff et al. 2001). The nNOS isoform also contains a corresponding Akt phosphorylation motif. However, cotransfection of Akt with nNOS failed to increase NO_i release in COS-7 cells (Fulton et al. 1999). Then again, the addition of an N-myristoylation site on nNOS to enhance its interactions with biological membranes resulted in Akt-induced stimulation of nNOS in a manner similar to that seen with eNOS (Fulton et al. 1999). These authors concluded that both eNOS and nNOS may be activated by Akt when anchored to membranes. The present results therefore suggest that PI-(3)K–Akt signalling acts in conjunction with Ca²⁺–CaM signalling to stimulate NO_i production from both eNOS and nNOS in ventricular myocytes.

In the present study contraction-mediated modulation of NO_i production failed to significantly affect [Ca²⁺]_i transient amplitudes. However, selective pharmacological inhibition of either eNOS or nNOS resulted in opposite changes in [Ca²⁺]_i transient amplitudes. These results therefore suggest that endogenous NO_i derived from eNOS stimulates and from nNOS inhibits the [Ca²⁺]_i transient. The opposing effects of NO_i derived from each NOS isoform appear to cancel each other, resulting in no net effect on [Ca²⁺]_i transient amplitude. This idea is supported by the finding that [Ca²⁺]_i transient amplitude is unaffected when both sources of NO_i are inhibited simultaneously. These results also indicate that contraction-mediated NO_i production does not feed back on contraction to affect subsequent contraction-mediated NO_i production. The mechanisms by which endogenous NO_i derived from the two different NOS isoforms affect [Ca²⁺]_i transient amplitude are complex and clearly beyond the scope of the present work. However, there is evidence that nNOS is localized to the SR (Xu et al. 1999; Barouch et al. 2002) and eNOS is localized within caveolae (Feron et al. 1996; Okamoto et al. 1998). Therefore, each spatially confined source of NO_i has the potential to act directly and/or indirectly to locally regulate multiple steps in cardiac (E–C) coupling (see Casadei & Sears, 2003). As shown in the present study, NO derived from these two sources can exert opposing effects. For example, NO derived from nNOS enhances and eNOS inhibits β-adrenergic receptor stimulation of contractility via actions at different effector sites (Barouch et al. 2002). In transgenic mice, those lacking nNOS exhibited depressed force–frequency responses and those lacking eNOS had normal responses (Khan et al. 2003). These studies, as well as the present work, support the idea that NOS isoforms exert different and possibly opposite effects on cardiac E–C coupling. Of course, E–C coupling is not the only target of endogenous NO_i signalling. NO exerts a wide variety of effects on cardiac cell function via second messenger cGMP signalling as well as direct S-nitrosylation reactions. For example, NO_i modulates autonomic nerve regulation of cardiac function (Massion et al. 2005) and exerts significant cardioprotection from ischaemic–reperfusion injury, possibly via effects on mitochondrial metabolism (Massion et al. 2005; Jones & Bolli, 2006). Conversely, the present results suggest that autonomic nerve regulation of cardiac contraction also modulates NO_i production. Moreover, contractile dysfunction of the failing heart may directly contribute to a decrease in NO_i production (Recchia et al. 1998), causing the failing heart to lose NO-mediated cardioprotection and thereby become more susceptible to the progression of disease.