Role of cyclic GMP-dependent protein kinase in the contractile response to exogenous nitric oxide in rat cardiac myocytes

Abstract

Nitric oxide (NO) can directly modulate cardiac contractility by accelerating relaxation and reducing diastolic tone. The intracellular mechanisms underlying these contractile effects are poorly understood. Here we investigate the role of cyclic GMP-dependent protein kinase (PKG) in the contractile response to exogenous NO in rat ventricular myocytes. Isolated ventricular myocytes were stimulated electrically and contractility was assessed by measuring cell shortening. Some cells were loaded with the fluorescent Ca²⁺ probe indo-1 AM for simultaneous assessment of the intracellular Ca²⁺ transient. The NO donor diethylamine NONOate (DEA/NO, 10 μm) significantly increased resting cell length, reduced twitch amplitude and accelerated time to 50 % relaxation (to 100.8 ± 0.2, 83.7 ± 3.0 and 88.9 ± 3.7 % of control values, respectively). The contractile effects of DEA/NO occurred without significant changes in the amplitude or kinetics of the intracellular Ca²⁺ transient, suggesting that the myofilament response to Ca²⁺ was reduced. These effects were abolished by inhibition of either guanylyl cyclase (with 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one; ODQ, 10 μm) or PKG (with Rp-8-Br-cGMPs, 10 μm) suggesting that, at the concentration investigated, the effects of DEA/NO were mediated exclusively by PKG, following activation of guanylyl cyclase and elevation of cGMP. Direct activation of PKG with 8-pCPT-cGMP (10 μm) mimicked the effects of DEA/NO (resting cell length and time to 50 % relaxation were 100.6 ± 0.1 and 90.5 ± 1.5 % of control values, respectively).The reduced myofilament Ca²⁺ responsiveness was not attributable to an intracellular acidosis since the small reduction in pH_i induced by DEA/NO was found to be uncoupled from its contractile effects. However, hearts treated with DEA/NO (10 μm) showed a significant increase (1.4-fold; P < 0.01) in troponin I phosphorylation compared to control, untreated hearts. These results suggest that the reduction in myofilament Ca²⁺ responsiveness produced by DEA/NO results from phosphorylation of troponin I by PKG.

It is now well established that nitric oxide (NO) released either from cardiac endothelial cells or generated within cardiac myocytes themselves can directly influence cardiac contractile function (reviewed by Kelly et al. 1996; Shah & MacCarthy, 2000). In the absence of stimulation by extrinsic agonists, both endothelium-derived NO and exogenous NO donors accelerate myocardial relaxation and/or reduce diastolic tone. These effects have been observed in a variety of preparations and species, including rat cardiac myocytes (Shah et al. 1994; Vila-Petroff et al. 1999), ferret (Smith et al. 1991), cat (Mohan et al. 1996) and human (Flesch et al. 1997) papillary muscles; isolated ejecting guinea-pig hearts (Grocott-Mason et al. 1994); and normal human subjects undergoing diagnostic cardiac catheterization (Paulus et al. 1994, 1995). NO can also modulate myocardial inotropic state, although whether it is positively or negatively inotropic may depend on several factors including the concentration of NO (Sarkar et al. 2000), the rate of NO release (Balligand et al. 1993) and/or the presence of β-adrenergic stimulation (Méry et al. 1993; Sandirasegarane & Diamond, 1999).

The intracellular signalling pathways responsible for these effects of NO remain poorly understood. A widely held hypothesis is that NO activates soluble guanylyl cyclase and elevates cGMP, which triggers contractile changes via activation of cGMP-dependent protein kinase (PKG) or by modulation of phosphodiesterase activity (see below). Rat ventricular myocytes are known to express low levels of PKG (approximately 10-fold lower than in smooth muscle; Méry et al. 1991). In spite of these low levels, evidence suggests that PKG mediates the cGMP-induced reduction in the L-type Ca²⁺ current observed following prior stimulation with cAMP (Méry et al. 1991).

Recent work has proposed the intriguing possibility that positive inotropic effects of NO donors may involve cGMP-independent pathways (Sandirasegarane & Diamond, 1999; Vila-Petroff et al. 1999; Paolocci et al. 2000; Sarkar et al. 2000); e.g. by stimulation of adenylyl cyclase (Vila-Petroff et al. 1999) or altered Ca²⁺ fluxes due to nitrosylation of sarcolemmal L-type Ca²⁺ channels (Hu et al. 1997) or ryanodine receptors (Xu et al. 1998). Furthermore, positive inotropic effects of some NO donors have been attributed to protein nitrosylation by peroxynitrite (Chesnais et al. 1999a; Paolocci et al. 2000).

Modification of contractility via cGMP-mediated inhibition or stimulation of phosphodiesterase activity appears to have significance mainly when contractility is already enhanced by activation of cAMP-dependent protein kinase (PKA). Under these conditions, low levels of cGMP inhibit (whereas high levels stimulate) phosphodiesterase activity, leading to positive (or negative) inotropic effects via stimulation (or inhibition) of the Ca²⁺ current, I_Ca (Méry et al. 1993). However, NO donors generally have little effect on I_Ca under baseline conditions (Méry et al. 1991, 1993; Wahler & Dollinger, 1995).

The negative inotropic and relaxant effects of NO and cGMP have largely been attributed to a cGMP-mediated reduction in myofilament Ca²⁺ responsiveness, possibly via activation of PKG (e.g. Shah, 1996; Vila-Petroff et al. 1999). However, the mechanisms responsible for reduced myofilament Ca²⁺ responsiveness remain to be determined. One possibility is that phosphorylation of troponin I by PKG may have comparable effects to PKA-induced phosphorylation (Robertson et al. 1982), i.e. reduction in myofilament Ca²⁺ sensitivity by increasing the off-rate of Ca²⁺ from troponin C. There is some evidence that PKG can phosphorylate troponin I in vitro (e.g. Blumenthal et al. 1978) and that the contractile effects of NO may be related to troponin I phosphorylation (Kaye et al. 1999). However, the effects of PKG on Ca²⁺ sensitivity of skinned cardiac muscle have been contradictory, with suggestions that PKG either reduces (Pfitzer et al. 1982) or increases (Mope et al. 1980) Ca²⁺ sensitivity. Alternatively, it has been postulated that reduced myofilament Ca²⁺ responsiveness results from cytosolic acidification induced by disruption of Na⁺-H⁺ exchanger function by the NO-cGMP pathway (Ito et al. 1997). There is also some evidence that PKG activation (with 8-Br-cGMP) may suppress basal I_Ca (Matsumoto et al. 2000), although there is a general consensus that such suppression is only relevant when I_Ca is pre-stimulated via the PKA pathway (Méry et al. 1991; Wahler & Dollinger, 1995; Matsumoto et al. 2000).

The aim of the present study was to determine more precisely the role of PKG in the contractile response to the NO donor diethylamine-NONOate (DEA/NO) in isolated rat ventricular myocytes, and in particular to distinguish between different subcellular mechanisms that may induce a reduction in myofilament Ca²⁺ sensitivity.

METHODS

Cardiac myocyte studies

Male Wistar rats (250–350 g) were killed by an overdose of sodium pentobarbitone (Euthatal, 60 mg kg⁻¹, i.p.) in accordance with UK Home Office regulations. Hearts were excised rapidly and ventricular myocytes isolated by collagenase digestion as described previously (Shah et al. 1997). Cells were not exposed to physiological Ca²⁺ concentrations (1.25 mm) until at least 1 h after isolation and all experiments were performed within 6 h of cell isolation. To monitor intracellular Ca²⁺ transients, some cells were loaded with indo-1 acetoxymethylester (indo-1 AM, final concentration 2 μm), essentially as described previously (Shah et al. 1997). Some cells were loaded with carboxy-seminaphthorhodafluor-1 acetoxymethylester (C-SNARF-1 AM, final concentration 4 μm) for assessment of intracellular pH (pH_i).

Single myocyte contraction and intracellular Ca²⁺ (or pH) were studied on the stage of an inverted fluorescence microscope (Nikon Eclipse, TE300) coupled to a dual emission spectrophotometer (Photon Technology International (PTI), Princeton, NJ, USA; for further details see Shah et al. 1997). The myocyte chamber was superfused at 1–2 ml min⁻¹ with Hepes buffer containing (mm): NaCl 117, KCl 5.7, NaH₂PO₄ 1.2, MgSO₄ 0.66, glucose 10, Hepes 20, CaCl₂ 1.25, pH 7.4. Experiments were performed at 24 ± 0.5 °C to minimize cellular loss of fluorescent indicators (Spurgeon et al. 1990). Cells were field stimulated at 0.5 Hz (isolated stimulator unit model SD9, Grass Instrument Co., Quincy, MA, USA) via two platinum electrodes positioned on either side of the chamber. Myocytes were examined using a × 40 oil immersion Fluor objective (Nikon, NA 1.3) and myocyte length was monitored by video-edge detection (Crystal Biotech, Hopkinton, MA, USA) at a temporal resolution of 5 ms. Intracellular Ca²⁺ was assessed from the 410/480 nm fluorescence emission ratio in indo-1-loaded cells excited at 360 nm, following subtraction of background fluorescence. No attempt was made to calibrate the indo-1 410/480 nm signal because of potential errors resulting from dye compartmentalization (Spurgeon et al. 1990).

In SNARF-1-loaded cells, pH_i was measured from the 580/640 nm fluorescence emission ratio following excitation at 540 nm. The 580/640 nm fluorescence ratio was converted off-line to pH_i with reference to an in situ calibration curve generated using nigericin-containing calibration solutions of varying pH, according to the method described by Yasutake et al. (1996). All data were recorded on a PC running Felix software (version 1.21, PTI).

Experimental protocol

Myocytes were selected for study according to previously established criteria (Capogrossi et al. 1986), i.e. myocytes were rod shaped with clear striations, showed no membrane blebs or granulations and spontaneous contractions were infrequent (< 1 min⁻¹ in the absence of stimulation). At the start of each experiment, cells were stimulated continuously at 0.5 Hz for 10–15 min to check the stability of contraction. Twitch amplitude during the stabilization period was recorded on a thermal chart recorder (WR7400, Graphtec Corporation, Yokohama, Japan). Cells which demonstrated significant changes in twitch amplitude (> ± 5 %) or spontaneous bursts of activity during the stabilization period were rejected from study. Preliminary experiments established that cells demonstrating a stable twitch amplitude over the 10–15 min stabilization period also showed no significant changes in the time course of twitch contractions (time to peak shortening, time from peak to 50 % relaxation) during this period (data not shown). Stable cells were used to examine the effects of the following interventions on myocyte contractility (each in separate groups of cells): (i) DEA/NO alone (1 nm-100 μm); (ii) DEA/NO (10 μm) in the presence of the guanylyl cyclase inhibitor, 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ, 10 μm); (iii) DEA/NO (10 μm) applied in the presence of the specific PKG inhibitor, Rp-8-Br-cGMPs (10 μm); (iv) a specific activator of PKG, 8-pCPT-cGMP (10 μm, applied for 10–15 min); (v) DEA/NO (10 μm) applied during inhibition of the Na⁺-H⁺ exchanger with HOE-642 (3 μm).

Note that DEA/NO was applied for 10–15 min (for each concentration examined) and each of the inhibitors used (ODQ, Rp-8-Br-cGMPs, HOE-642) were applied for 10 min prior to treatment with DEA/NO. All experiments were replicated in at least six cells obtained from at least three (or more) different myocyte preparations.

Assessment of troponin I phosphorylation

Isolated hearts were perfused with Hepes buffer (composition as above) in Langendorff mode (37 °C) at constant coronary flow. Left ventricular pressure (LVP) was measured using an intraventricular balloon attached to a pressure transducer (Bell & Howard Ltd, Watford, UK). At the start, the balloon was inflated to achieve an end-diastolic LVP of 10 mmHg and flow rate adjusted to establish an aortic pressure of 75 ± 5 mmHg (separate pressure transducer, Bell & Howard Ltd). Hearts were paced at 360 beats min⁻¹ via a right atrial pacing electrode. Pressure data were sampled at 400 Hz via a MacLab module (AD Instruments, Hastings, UK) and recorded on a PC using Chart software (version 3.4, AD Instruments). After 15 min stabilization, hearts were exposed either to DEA/NO (10 μm), isoproterenol (isoprenaline; 3 μm) or no drug. After a further 5 min, the left ventricle was rapidly dissected, freeze-clamped and stored at −80 °C for later analysis. Isoproterenol-treated hearts were used as positive controls for assessment of troponin I phosphorylation. Whole hearts were chosen for analyses of troponin I phosphorylation to facilitate the immunoprecipitation analysis by increasing the amount of available tissue. Troponin I is specific to the cardiac myocytes (e.g. see Li et al. 2001) and hence non-myocyte constituents of the heart (e.g. coronary microvasculature and endothelial cells) would not interfere with the experimental interpretation.

Troponin I phosphorylation in ventricular tissue homogenates was investigated using phosphorylation-independent and phosphorylation-specific monoclonal antibodies (mAbs 19 and 14, respectively, Al-Hillawi et al. 1998) exactly as described recently (Tavernier et al. 2000). Total troponin I was immunoprecipitated with mAb 19 and membranes were then probed with both mAbs 19 and 14. Protein bands were visualized by enhanced chemiluminescence and quantified by densitometry. It should be recognised that mAb 14 detects cardiac troponin I that has been phosphorylated at one specific PKA phosphorylation site (Ser24) in the N-terminal (Al-Hillawi et al. 1998). PKA-dependent phosphorylation of troponin I can also take place at Ser23 and either one (monophosphorylation) or both (biphosphorylation) sites may be phosphorylated at any one time. Consequently, the degree of phosphorylation suggested by mAb 14 detection in the present study is not directly comparable to previous estimates of troponin I phosphorylation using the incorporation of ³²P which recognises total phoshorylation at either or both sites (e.g. see McConnell et al. 1998).

Chemicals and solutions

All chemicals used were of analytical grade and were obtained from BDH (Poole, Dorset, UK) or Sigma (Poole, Dorset, UK). Collagenase type II was purchased from Worthington Biochemical Corporation (Twyford, Reading, UK). DEA/NO, ODQ and indo-1 AM were obtained from Calbiochem. (Nottingham, UK). Rp-8-Br-cGMPs and 8-pCPT-cGMP were obtained from Biolog Life Science Institute (Bremen, Germany). C-SNARF-1 AM was obtained from Molecular Probes (Eugene, OR, USA). HOE-642 was a gift from Aventis (Frankfurt, Germany).

Data analysis

Steady-state twitches and Ca²⁺ transients were averaged over 30 s periods. Cell twitch amplitude was expressed as percentage of resting cell length. Twitch kinetics were quantified by measuring the time to peak shortening (T_pk) and the time from peak shortening to 50 % relaxation (RT₅₀). Similar measurements were derived to quantify Ca²⁺ transient kinetics. Peak shortening and re-lengthening velocity were measured from the derivative of the cell length trace and were expressed relative to the control value. Effects of different interventions on pH_i were expressed in pH units, having first subtracted the effect attributed to time alone (see below). All data are presented as mean ±s.e.m. Student's paired or unpaired t tests were employed as appropriate, and P < 0.05 was considered statistically significant.

RESULTS

Effects of DEA/NO on myocyte contraction

DEA/NO (100 nm-100 μm) had a dose-dependent negative inotropic effect and produced a small increase in resting cell length (Fig. 1). It also reduced twitch duration, manifest both as an earlier onset of relaxation (shorter T_pk) and reduced relaxation time (RT₅₀). Since there is evidence that low doses of NO donors may have positive inotropic effects (e.g. Vila-Petroff et al. 1999), additional experiments were performed to examine the contractile effects of lower doses of DEA/NO. These experiments revealed that 1 and 10 nm DEA/NO had no significant effects on twitch amplitude, resting cell length, T_pk or RT₅₀ (paired t tests compared to control values, P > 0.05).

Figure 1. Dose-dependent relaxant and negative inotropic effects of DEA/NO in isolated cardiac myocytes.

A, twitch contractions from a typical cell. Numbers next to each trace indicate the concentration of DEA/NO applied (μm). B-E, mean data (±s.e.m.) for changes in resting cell length, twitch amplitude, time to peak shortening (T_pk), and time from peak shortening to 50 % relaxation (RT₅₀). Control data (0 DEA/NO) is from a total of 18 cells investigated. Note that the effects of lower concentrations (10⁻⁹ and 10⁻⁸m) and higher concentrations (10⁻⁷ to 10⁻⁴m) of DEA/NO were investigated in separate group of cells (n≥ 9 cells for each concentration) but there were no significant differences in the control data between these groups (unpaired t test, P > 0.05). *Statistical significance compared to corresponding control data (paired t tests, P < 0.05).

DEA/NO (0.1–10 μm) had no significant effects on the size or kinetics of the Ca²⁺ transient in indo-1-loaded cells. However, at the highest concentration examined (100 μm), it produced a small decrease in peak indo-1 fluorescence ratio (to 96.6 ± 1.1 % of control values, P < 0.05, n = 5) although the diastolic ratio and Ca²⁺ transient kinetics were unaffected. Figure 2A (left panel) shows the contractile effects of DEA/NO (10 μm), occurring independently of changes in intracellular Ca²⁺ transient, in a typical cell. A phase-plane plot (loop) of cell length vs. simultaneously measured indo-1 fluorescence ratio (Spurgeon et al. 1992; Shah et al. 1994) for these twitches is also shown (Fig. 2A, right panel). It is evident that DEA/NO shifts the relaxation phase of the loop downwards and to the right, which is indicative of a reduction in myofilament Ca²⁺ responsiveness (Spurgeon et al. 1992; Shah et al. 1994). Furthermore, the onset of relaxation (re-lengthening) occurs at higher intracellular [Ca²⁺] with DEA/NO than control. Comparable phase-plane loops were observed for six other cells. Figure 2B summarizes the effects of 10 μm DEA/NO on each of the parameters measured (expressed as percentage change relative to control value). Actual measured values for each of the parameters displayed in Fig. 2B are given in Table 1. DEA/NO (10 μm) produced a small but significant increase in resting cell length and a decrease in twitch amplitude, T_pk and RT₅₀, but had no significant effects on diastolic (R_d) or peak systolic (R_pk) indo-1 fluorescence ratio. Neither peak shortening velocity (V_s) nor re-lengthening velocity (V_l) were significantly altered. To further investigate the mechanisms responsible for the apparent reduction in myofilament Ca²⁺ responsiveness, all subsequent experiments were conducted using 10 μm DEA/NO.

Figure 2. Reduction in myofilament Ca²⁺ responsiveness induced by DEA/NO (10 μm).

A, left panel, representative example of effect on myocyte contraction and simultaneous intracellular Ca²⁺ transients (indo-1 410/480 nm ratio); right panel, phase-plane loops of cell length against intracellular Ca²⁺ for the twitches shown in left panel. Thin lines denote control conditions, and thick lines denote data recorded after 10 min exposure to DEA/NO. B, averaged percentage changes relative to the control value (mean ±s.e.m.) in contractile parameters and indo-1 410/480 nm ratio following application of DEA/NO. *P < 0.05. Numbers in parentheses indicate the number of cells averaged. Actual values (mean ±s.e.m.) are given in Table 1. RL, resting cell length (n = 15); TA, twitch amplitude (n = 15); T_pk, time to peak shortening (n = 15); RT₅₀, time from minimum length to half relaxation (n = 15); V_s, maximum shortening velocity (n = 15); V_l, maximum re-lengthening velocity (n = 15); R_d, diastolic indo-1 410/480 nm ratio (n = 7); R_pk, peak indo-1 410/480 nm ratio (n = 7).

Table 1.

Comparison of contractile parameters and indo-1 fluorescence of rat ventricular myocytes before (control) and after treatment with DEA/NO (10 μm)

	Control	DEA/NO	n	Pvalue
Resting cell length (μm)	104.5 ± 4.0	105.2 ± 4.0	15	P < 0.01
Twitch amplitude (% resting cell length)	10.7 ± 0.9	9.3 ± 0.9	15	P < 0.01
Tpk (ms)	227 ± 12	210 ± 13	15	P < 0.01
RT50 (ms)	113 ± 7	100 ± 6	15	P < 0.01
Peak shortening velocity (μm s−1)	149 ± 19	137 ± 24	15	P > 0.05
Peak re-lengthening velocity (μm s−1)	123 ± 17	122 ± 22	15	P > 0.05
Diastolic indo-1 410/480 nm ratio	0.609 ± 0.122	0.624 ± 0.353	7	P > 0.05
Peak indo-1 410/480 nm ratio	1.030 ± 0.192	1.012 ± 0.188	7	P > 0.05

Data are shown as mean ±s.e.m. for n number of cells. Student's paired t test was used to test for statistical significance.

Effect of guanylyl cyclase or PKG inhibition on contractile response to DEA/NO

The role of guanylyl clase was investigated using the guanylyl cyclase inhibitor 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ, 10 μm). This dose of ODQ has been shown to completely abolish the NO-stimulated increase in cGMP in rat ventricular myocytes (Vila-Petroff et al. 1999). Ca²⁺ transient data were not analysed since ODQ itself was found to be fluorescent. Averaged data from 20 cells (Fig. 3A–D) indicated that ODQ alone had no significant effects on any of the contractile parameters measured, but that in its maintained presence the contractile effects of DEA/NO were completely abolished.

Figure 3. Effect of the guanylyl cyclase inhibitor ODQ on responses to DEA/NO (10 μm).

A-D, mean data (±s.e.m.) for changes in resting cell length, twitch amplitude, time to peak shortening (T_pk), and time from peak shortening to 50 % relaxation (RT₅₀). Open columns, values prior to addition of any drugs; hatched columns, ODQ alone; filled columns, DEA/NO in the presence of ODQ. n = 20.

In separate experiments, pretreatment of myocytes with the PKG inhibitor Rp-8-Br-cGMPs (10 μm) completely blocked the contractile effects of DEA/NO (n = 10 cells, Fig. 4B–E). Representative twitches, Ca²⁺ transients and corresponding phase-plane loops from one such experiment (Fig. 4A, left and right panels respectively) show full inhibition of the effects of DEA/NO. Similar phase-plane loops were observed for five other cells. Rp-8-Br-cGMPs alone had no significant effects on any of the parameters measured (Fig. 4B–E).

Figure 4. Effect of the PKG inhibitor Rp-8-Br-cGMPs (10 μm) on responses to DEA/NO (10 μm).

A, left panel, twitch contractions and simultaneous intracellular Ca²⁺ transients from a typical cell in control conditions (thin lines), after addition of Rp-8-Br-cGMPs (dotted lines), and following application of DEA/NO in the continued presence of Rp-8-Br-cGMPs (thick lines); right panel, phase-plane loops of cell length against intracellular Ca²⁺ for the twitches shown in left panel. B-E, mean data (±s.e.m.) for changes in resting cell length, twitch amplitude, time to peak shortening (T_pk), and time from peak shortening to 50 % relaxation (RT₅₀). Open columns, values prior to addition of any drugs; hatched columns, Rp-8-Br-cGMPs alone; filled columns, DEA/NO in the presence of Rp-8-Br-cGMPs. n = 10.

These results suggest that either inhibition of guanylyl cyclase or of PKG prevents the reduction in myofilament Ca²⁺ responsiveness produced by DEA/NO.

Effects of direct PKG activation

The specific PKG activator, 8-pCPT-cGMP (10 μm), induced similar effects on myocyte twitch shortening as those observed with DEA/NO. Results from a typical cell are illustrated in Fig. 5A and average data from 14 cells in Fig. 5B–E. A significantly increased resting cell length and significantly shortened T_pk and RT₅₀ were also seen with 8-pCPT-cGMP. It also reduced twitch amplitude in most cells although this did not achieve statistical significance. As with DEA/NO, these effects occurred with no significant change in amplitude or kinetics of the intracellular Ca²⁺ transient (Fig. 5A, left panel) and were associated with a rightward and downward shift of the corresponding phase-plane loop (Fig. 5A, right panel). Similar phase-plane loops were observed in eight other cells. Hence, direct activation of PKG appeared to induce contractile changes by decreasing myofilament Ca²⁺ responsiveness.

Figure 5. Effect of 8-pCPT-cGMP (10 μm) in isolated cardiac myocytes.

A, left panel, representative example of effect on myocyte contraction and intracellular Ca²⁺ transients; right panel, phase-plane loops of cell length against intracellular Ca²⁺ for the twitches shown in left panel. B-E, mean data (±s.e.m.) for changes in resting cell length, twitch amplitude, time to peak shortening (T_pk), and time from peak shortening to 50 % relaxation (RT₅₀). Open columns, values prior to addition of any drugs; filled columns, after 8-pCPT-cGMP. n = 14; *P < 0.05.

Effects of DEA/NO on intracellular pH

It has been suggested that decreased myofilament Ca²⁺ responsiveness associated with NO may be secondary to intracellular acidification (Ito et al. 1997). To investigate this possibility, we studied the effects of DEA/NO (10 μm) on pH_i in C-SNARF-1-loaded myocytes. These experiments were complicated by the finding that in myocytes that were electrically stimulated to contract at 0.5 Hz, there was a small but significant reduction in pH_i with time alone (0.037 ± 0.008 pH units in 10 min; P < 0.05, comparing time 0 and time 10 min by paired t test; n = 35 cells from six rats; Fig. 6A). Therefore, in experiments examining the effects of DEA/NO (or other interventions) on pH_i, it was necessary to account for the effect of time alone. This was done in two ways. In each cell, pH_i was first monitored during a 10 min stimulation period without added drugs (i.e. 0–10 min) and then the net effects of DEA/NO (or other drugs) applied subsequently (i.e. 10–20 min) were calculated by subtracting the effects occurring due to time alone (see Fig. 6A). Using this method, DEA/NO reduced pH_i by 0.060 ± 0.010 pH units over and above the effects of time alone (n = 36 cells; P < 0.05 paired t test; Fig. 6B). Alternatively, the change in pH_i in the presence of DEA/NO including the effect of time (i.e. 0.095 ± 0.002 pH units, n = 36 cells, time interval 10–20 min) was compared with the change in pH_i recorded from separate cells not exposed to DEA/NO over the same time period (i.e. 0.049 ± 0.003 pH units, n = 35 cells). The difference between these values was also statistically significant (unpaired t test; P < 0.01).

Figure 6. Effect of DEA/NO (10 μm) on pH_i in C-SNARF-1-loaded ventricular myocytes.

A, effect of time alone (open squares, n = 35 cells) and DEA/NO added at 10 min (filled circles, n = 36 cells). Arrow indicates DEA/NO application. *P < 0.05 compared to the effects of time alone (unpaired t test). B, mean (±s.e.m.) data for changes in pH_i with DEA/NO alone (filled columns), DEA/NO in the presence of Rp-8-Br-cGMPs (hatched columns), or DEA/NO in the presence of HOE-642 (open columns). Numbers in brackets indicate the number of cells averaged. *P < 0.05.

Using the former approach (each cell as its own time control), the change in pH_i produced by DEA/NO during inhibition of PKG with Rp-8-Br-cGMPs (10 μm) was found to be 0.053 ± 0.011 pH units (n = 28 cells), which was not significantly different from the effects of DEA/NO alone (Fig. 6B). Therefore, inhibition of PKG failed to inhibit the change in pH_i induced by DEA/NO, despite its effectiveness at abolishing the contractile effects (Fig. 4). Since reduction in pH_i may result from inhibition of forward mode H⁺ extrusion by the Na⁺-H⁺ exchanger (Ito et al. 1997), the effects of DEA/NO were next investigated in the presence of the Na⁺-H⁺ exchanger inhibitor HOE-642 (cariporide, 3 μm). As shown in Fig. 6B, HOE-642 significantly inhibited the reduction in pH_i induced by DEA/NO (P < 0.01; n = 29 cells).

In separate experiments, we then examined whether HOE-642 had any effect on the contractile effects of DEA/NO. Figure 7 illustrates that Na⁺-H⁺ exchange inhibition with HOE-642 failed to prevent the contractile effects of DEA/NO. It should be noted that HOE-642 alone caused a small but significant reduction in T_pk and twitch amplitude (Fig. 8C and D), but DEA/NO induced a further reduction in twitch amplitude and twitch duration when compared to the effects of HOE-642 alone.

Figure 7. Effect of the Na⁺-H⁺ exchange inhibitor HOE-642 (3 μm) on responses to DEA/NO (10 μm).

A, twitch contractions from a typical cell in control conditions (thin lines), after addition of HOE-642 (dotted lines), and following application of DEA/NO in the continued presence of HOE-642 (thick lines). B-E, mean data (±s.e.m.) for changes in resting cell length, twitch amplitude, time to peak shortening (T_pk), and time from peak shortening to 50 % relaxation (RT₅₀). Open columns, values prior to addition of any drugs; hatched columns, HOE-642 alone; filled columns, DEA/NO in the presence of HOE-642. *P < 0.05 compared to control or HOE-642 alone.

Figure 8. Immunoblots for total troponin I (A) and phosphorylated troponin I (B) in LV myocardium.

Left panel, hearts treated with either DEA/NO (10 μm) or vehicle alone (control). Right panel, hearts treated with either isoproterenol (Iso, 3 μm) or vehicle alone (control).

Effects of DEA/NO on phosphorylation of troponin I

Langendorff-perfused hearts treated with DEA/NO (n = 4) showed a significant reduction in peak systolic LVP (from 132 ± 10 to 100 ± 8 mmHg, P < 0.05), mean aortic pressure (from 76 ± 1 to 51 ± 2 mmHg, P < 0.05) and left ventricular dP/dt max (maximum rate of systolic pressure development, from 4371 ± 633 to 3619 ± 509 mmHg s⁻¹, P < 0.05). In contrast, control hearts (n = 4) showed no significant changes in these parameters over the same time period (data not shown).

Troponin I phosphorylation was studied using phosphorylation-independent and phosphorylation-specific monoclonal antibodies (mAbs 19 and 14, respectively) as described in the Methods. As illustrated in Fig. 8A (left panel), the total amount of troponin I detected with mAb 19 was similar for DEA/NO-treated and control groups. However, probing with mAb 14 revealed a significant increase in phosphorylated troponin I in DEA/NO treated hearts compared with controls (Fig. 8B, left panel). As expected, there was also a significant increase in troponin I phosphorylation in isoproterenol-treated hearts compared with their corresponding controls (Fig. 8A–B, right panel). Densitometric analysis showed that DEA/NO increased the level of troponin I phosphorylation by 42 ± 6 % (P < 0.01, n = 4 hearts).

DISCUSSION

The intracellular signalling mechanisms responsible for the contractile effects of NO in cardiac myocytes have not been fully characterized. Previous studies (Shah et al. 1994; Vila-Petroff et al. 1999) have suggested that PKG is a likely candidate in mediating the decreased myofilament Ca²⁺ responsiveness associated with NO donors but the underlying mechanism has remained elusive. Intracellular acidification caused by cGMP-mediated inhibition of forward mode Na⁺-H⁺ exchange has been proposed to potentially explain the reduced myofilament Ca²⁺ sensitivity (Ito et al. 1997) but has not been unequivocally linked with observed contractile effects. Similarly, there is contradictory evidence regarding possible linkage between PKG and troponin I phosphorylation (Mope et al. 1980, Pfitzer et al. 1982). The main findings of the present study are that the contractile effects of DEA/NO (1–10 μm) in rat cardiac myocytes result from a reduction in myofilament Ca²⁺ responsiveness, mediated exclusively via the cGMP-PKG pathway and most probably culminating in the phosphorylation of troponin I by PKG. The small reduction in intracellular pH induced by DEA/NO appears to be largely unrelated to its contractile effects.

Contractile effects of DEA/NO

DEA/NO increased resting cell length, reduced twitch amplitude and caused faster twitch relaxation in myocytes studied in the absence of pre-stimulation by other agonists, in accordance with contractile effects previously reported for other NO donors (e.g. sodium nitroprusside (SNP), Ito et al. 1997; Chesnais et al. 1999b; S-nitroso-N-acetyl-penicillamine (SNAP), Vila-Petroff et al. 1999; Chesnais et al. 1999b; and the cGMP analogue 8-bromo-cGMP, Shah et al. 1994). Some studies have reported positive inotropic effects of NO donors at low concentrations (Vila-Petroff et al. 1999) or even high concentrations (Müller-Strahl et al. 2000; Sarkar et al. 2000). In contrast, in the present study DEA/NO was found to have no positive inotropic effects, even at the lowest concentrations examined (1 and 10 nm). The reasons for these differing results are unclear but could involve differences in the species used, the experimental protocol or the NO donor investigated.

At concentrations of 0.1–10 μm, DEA/NO had no significant effect on intracellular Ca²⁺ transients, consistent with its contractile effects being due solely to an alteration in myofilament Ca²⁺ responsiveness. However, at the highest concentration investigated (100 μm) there was a small reduction in the peak Ca²⁺ transient although the kinetics of Ca²⁺ transient decline remained unaltered. It has previously been suggested that cGMP and PKG are associated with phosphorylation of phospholamban on the sarcoplasmic reticulum (SR) which could modify cardiac contractility by accelerating SR Ca²⁺ uptake (Sabine et al. 1995). However, the absence of an effect of DEA/NO and the PKG activator 8-pCPT-cGMP on the rate of Ca²⁺ transient decline, suggests that acceleration of SR Ca²⁺ uptake, secondary to phospholamban phosphorylation, is not a major mechanism in the present study.

The contractile effects of DEA/NO (10 μm) were completely abolished by inhibition of guanylyl cyclase with ODQ in the present study; i.e. the negative inotropic and relaxant effects of DEA/NO were attributable exclusively to a cGMP-mediated reduction in myofilament Ca²⁺ responsiveness. Similarly, Vila-Petroff et al. (1999) found that the negative inotropic effect of high concentrations of SNAP was abolished by ODQ. In contrast, positive inotropic effects of NO donors have consistently been found to be insensitive to inhibition of guanylyl cyclase, (Vila-Petroff et al. 1999; Sarkar et al. 2000) suggesting that these effects occur through separate pathways.

Role of PKG

Elevation of intracellular cGMP in cardiac myocytes can potentially influence several different pathways, including the activation of PKG, and inhibition or stimulation of phosphodiesterase activity and consequent changes in cAMP levels. It is also potentially feasible that high levels of cGMP could induce changes in contractility via cross-activation of PKA, as has been reported in smooth muscle cells (e.g. see Ruiz-Velasco et al. 1998). In the present study, the contractile effects of DEA/NO were found to be completely blocked by specific inhibition of PKG with Rp-8-Br-cGMPs, suggesting that this was the sole mechanism of action. In turn, PKG activation appeared to result mainly in a reduction in myofilament Ca²⁺ responsiveness. In support of this, specific activation of PKG using 8-pCPT-cGMP induced contractile changes similar to those produced by DEA/NO and in the absence of alterations in the Ca²⁺ transient.

The observation that inhibition of PKG completely abolished the contractile effects of DEA/NO also implies that activation of the cAMP-PKA pathway, either by inhibition of phosphodiesterase activity or cross-activation of PKA, makes little or no contribution to the contractile effects of DEA/NO. Furthermore, activation of the PKA pathway would be expected to increase the contraction amplitude and the Ca²⁺ transient and accelerate the rate of Ca²⁺ transient decline. In contrast our experiments showed a reduction in twitch amplitude and no change in the intracellular Ca²⁺ transient.

Mechanism of reduction in myofilament Ca²⁺ responsiveness

It has been suggested that reduced myofilament Ca²⁺ sensitivity produced by NO donors is attributable to intracellular acidification resulting from cGMP-dependent inhibition of the Na⁺-H⁺ exchanger (Ito et al. 1997). These authors reported that the reduction in twitch amplitude (to 72 % of baseline) caused by SNP in rat cardiac myocytes was associated with a reduction in pH_i by approximately 0.1 pH units. In hypertrophied rat myocytes, they found that SNP had no significant effect on either twitch contraction or pH_i. In addition, the SNP-induced change in pH_i could be abolished by inhibition of Na⁺-H⁺ exchange with 5-(N-ethyl-N-isopropyl)-amiloride (EIPA). However, this study did not investigate whether EIPA also blocked the contractile effects of SNP.

Results from the present study showed that DEA/NO also produced a small reduction in pH_i, but that this persisted during inhibition of PKG with Rp-8-Br-cGMPs. Since PKG inhibition abolished the contractile effects of DEA/NO but not the associated acidification, it is unlikely that the decreased pH_i was a causative mechanism for the decreased myofilament Ca²⁺ responsiveness. Furthermore, whereas the change in pH_i induced by DEA/NO could be blocked by the Na⁺-H⁺ exchange inhibitor HOE-642, the latter agent did not inhibit the changes in contraction induced by DEA/NO. HOE-642 is known to have a high degree of selectivity for the cardiac Na⁺-H⁺ exchanger, and is highly effective in rat ventricular myocytes at the doses used in the present study (Shipolini et al. 1997). The present study therefore indicates that the contractile effects of DEA/NO are uncoupled from its effects on pH_i. It is interesting to note that such uncoupling has recently also been reported for the positive inotropic effect and alkalosis produced by angiotensin II in cardiac myocytes (Vila-Petroff et al. 2000).

Previous in vitro studies have suggested that PKG phosphorylates cardiac troponin I at the same sites (Ser23/24) as those phosphorylated by PKA (Blumenthal et al. 1978; Lincoln & Corbin, 1978). It is reasonable to propose therefore that PKG-induced phosphorylation of troponin I could produce similar effects to those reported for the PKA-induced troponin I phosphorylation, i.e. a decrease in myofilament Ca²⁺ sensitivity (Robertson et al. 1982) and/or increased rates of cross-bridge cycling (Hoh et al. 1988; Strang et al. 1994). However, experiments examining the effects of PKG on troponin I phosphorylation and myofilament Ca²⁺ sensitivity in skinned cardiac muscle have produced conflicting results (Mope et al. 1980; Pfitzer et al. 1982). In the present study, we found a clear increase in the level of troponin I phosphorylation following treatment with DEA/NO. The phosphorylation-specific antibody used in the current study specifically recognizes troponin I that is phosphorylated at the PKA-sensitive Ser24 in the N-terminal. It seems likely therefore that the reduction in myofilament Ca²⁺ responsiveness produced by DEA/NO is mediated by PKG-dependent phosphorylation of troponin I at the same site(s) as those phosphorylated by PKA. PKA-induced phosphorylation of troponin I is also thought to be responsible for the increase in cross-bridge cycling rate (as indicated by increased maximum shortening velocity or frequency for minimum dynamic stiffness) associated with β-adrenergic stimulation (Hoh et al. 1988; Strang et al. 1994; Fentzke et al. 1999). Whether PKG-induced troponin I phosphorylation has comparable effects on cross-bridge dynamics remains to be determined.

Conclusion

In rat ventricular myocytes, the negative inotropic and relaxant effects of DEA/NO can be attributed to a reduction in myofilament Ca²⁺ sensitivity mediated exclusively by the cGMP-PKG pathway. The reduction in myofilament Ca²⁺ responsiveness appears to result from phosphorylation of troponin I by PKG, but is uncoupled from the mild intracellular acidosis associated with DEA/NO treatment.