Control of intracellular calcium in the presence of nitric oxide donors in isolated skeletal muscle fibres from mouse

Abstract

In skeletal muscle, nitric oxide (NO) is commonly referred to as a modulator of the activity of the ryanodine receptor (RyR) calcium release channel. However the reported effects of NO on isolated sarcoplasmic reticulum (SR) preparations and single ryanodine receptor (RyR) activity are diverse, and how NO affects SR calcium release and intracellular calcium homeostasis under physiological conditions remains poorly documented and hardly predictable. Here, we studied the effects of NO donors on membrane current and intracellular [Ca²⁺] in single skeletal muscle fibres from mouse, under voltage-clamp conditions. When fibres were chronically exposed to millimolar levels of sodium nitroprusside (SNP) and challenged by short membrane depolarizations, there was a progressive increase in the resting [Ca²⁺] level. This effect was use-dependent with the slope of rise in resting [Ca²⁺] being increased two-fold when the depolarizing pulse level was raised from −20 to +10 mV. Analysis of the decay of the [Ca²⁺] transients suggested that cytoplasmic Ca²⁺ removal processes were largely unaffected by the presence of SNP. Also the functional properties of the dihydropyridine receptor were very similar under control conditions and in the presence of SNP. The resting [Ca²⁺] elevation due to SNP was accompanied by a depression of the peak calcium release elicited by pulses to +10 mV. The effects of SNP could be reproduced by the chemically distinct NO donor NOC-12. They could be reversed upon exposure of the fibres to the thiol reducing agent dithiothreitol. Results suggest that large levels of NO produce a redox-sensitive continuous leak of Ca²⁺ from the SR, through a limited number of release channels that do not close once they are activated by membrane depolarization. This SR Ca²⁺ leak and the resulting increase in resting [Ca²⁺] may be important in mediating the effects of excess NO on voltage-activated calcium release.

Skeletal muscle produces the biologically active gaseous messenger NO. Under normal conditions, this is predominantly achieved by the calcium/calmodulin-dependent activity of a constitutive isoform of NO synthase (Nakane et al. 1993; Silvagno et al. 1996). In situ generated NO is believed to play an important role in various physiological aspects of muscle function, and disruption of this signalling pathway is suspected to be involved in certain types of muscle disorders (for review see Stamler & Meissner, 2001). One of the main presumed targets of NO is the excitation–contraction coupling process. The contractile activity of skeletal muscle relies on acute control of the myoplasmic free calcium concentration, achieved down to the millisecond time scale by an optimized set of calcium intrusion, buffering and extrusion mechanisms. The essential source of intracellular calcium elevation is the sarcoplasmic reticulum (SR), with the gating of the ryanodine receptor (RyR) calcium release channel being under the tight control of the trans-tubular membrane potential, through a link with the voltage-sensing dihydropyridine receptor (for review see Melzer et al. 1995). Several endogenous molecules and ions have been shown to modulate the activity of the isolated RyR channel, highlighting possible auxiliary control levels of excitation–contraction coupling. Along this line, results from several studies using SR Ca²⁺ fluxes and [³H]ryanodine binding on SR vesicles, as well as single channel measurements suggested that NO can modulate SR calcium channel activity through mechanisms of S-nitrosylation and oxidation of free thiols (see Eu et al. 1999). However, overall the reported effects of NO are far from straightforward. Meszaros et al. (1996) found that NO donors reduce the rate of Ca²⁺ release from isolated SR as well as the open probability of RyRs in planar bilayers. Conversely, application of NO by various means was found to promote Ca²⁺ release from SR vesicles and to increase the open probability of isolated RyRs (Stoyanovsky et al. 1997). Studies from Suko et al. (1999) and Hart & Dulhunty (2000) showed that low concentrations of NO donors increase the open probability of the release channels while larger concentrations reduce it. Part of this complexity was suggested to arise from differences in experimental conditions with special importance given to the redox environment (see Stamler & Meissner, 2001). For instance, low concentrations of NO could prevent oxidation-induced activation of the calcium release channel, while higher concentrations would activate the channel (Aghdasi et al. 1997). Also, the effectiveness of NO in activating the RyR channels would be critically dependent on the ambient P_O2 (Eu et al. 2000), and muscle contractility would be regulated by a concerted action of these two parameters (Eu et al. 2003).

Despite this now considerable amount of information, how NO affects intracellular calcium handling in an intact muscle fibre remains surprisingly obscure. On the one hand, studies on isolated RyRs and SR give only limited insight into how NO affects the RyR activity under physiological conditions of gating of these channels. On the other hand, results from studies of muscle contractile activity are complicated by effects of NO on target proteins other than the RyR, including proteins directly involved in force production (Galler et al. 1997; Perkins et al. 1997; Andrade et al. 1998), and also possibly in membrane excitability. In this respect there is an obvious need for data concerning modulation of calcium release by NO under controlled membrane potential conditions. Here, we demonstrate that at the whole single-cell level, exogenous NO produces a use-dependent increase in resting [Ca²⁺] and a concomitant inhibition of SR calcium release at high levels of membrane depolarization. Preliminary aspects of this work have been presented to the Biophysical Society (Pouvreau et al. 2004).

Methods

Preparation of the muscle fibres

Experiments were performed on single skeletal fibres isolated from the flexor digitorum brevis muscles from 4- to 8-week-old OF1 mice. All experiments were performed in accordance with the guidelines of the French Ministry of Agriculture (87/848) and of the European Community (86/609/EEC). Details of procedures for enzymatic isolation of single fibres, partial insulation of the fibres with silicone grease and intracellular dye loading, were as previously described (Jacquemond, 1997; Collet et al. 1999; Collet & Jacquemond, 2002). In brief, mice were killed by cervical dislocation before removal of the muscles. Muscles were treated with collagenase (Sigma type 1) for 60–75 min at 37°C in the presence of Tyrode solution as the external solution. Single fibres were then obtained by triturating the muscles within the experimental chamber. The major part of a single fibre was electrically insulated with silicone grease so that whole-cell voltage clamp could be achieved on a short portion of the fibre extremity. Prior to patch clamp, indo-1 was introduced locally into the fibre by pressure microinjection through a micropipette containing 0.5 mm indo-1 dissolved in an intracellular-like solution (for details concerning microinjection, see Csernoch et al. 1998). Fibres were then left for ∼1 h to allow for dye intracellular equilibration. All experiments were performed at room temperature (20–22°C).

Electrophysiology

An RK-400 patch-clamp amplifier (Bio-Logic, Claix, France) was used in whole-cell configuration. Command voltage pulse generation and data acquisition were performed using commercial software (Biopatch Acquire, Bio-Logic) driving an A/D converter (Laboratory Master DMA board, Scientific Solutions Inc., Mentor, OH, USA). Analog compensation was systematically used to decrease the effective series resistance. Voltage clamp was performed with a microelectrode filled with the intracellular-like solution. The tip of the microelectrode was inserted through the silicone, within the insulated part of the fibre. Unless otherwise specified, membrane depolarizations were applied every 30 s from a holding command potential of −80 mV.

Measurement of intramembrane charge movement and of the slow calcium current

Charge movement currents were calculated according to previously described procedures (Collet et al. 2003). In brief, adequately scaled control current records elicited by 50 ms-long hyperpolarizing pulses of 20 mV were subtracted from the current elicited by test depolarizing pulses of the same duration to −50, −20 and +10 mV. The amount of charge moved during a test pulse was measured by integrating the ‘on’ portion of the corrected test current records. The calculated charge was normalized to the capacitance of the fibre.

Calcium current recordings were obtained in response to 1 s-long depolarizations to various potentials. Pulses were applied every 45 s. Only the linear leak component of the current was removed by subtracting the adequately scaled value of the steady current measured during a 20 mV hyperpolarizing step applied before the test pulse. The voltage dependence of the calcium current density curves was fitted with the following equation:

where I(V) is the peak current density at the command potential V, G_max is the maximum conductance, V_rev is the apparent reversal potential, V_0.5 is the half-activation potential and k is a steepness factor.

Fluorescence measurements

The optical set-up for indo-1 fluorescence measurements was previously described (Jacquemond, 1997). In brief, a Nikon Diaphot epifluorescence microscope was used in diafluorescence mode. For indo-1 excitation, the beam of light from a high-pressure mercury bulb set on the top of the microscope was passed through a 335 nm interference filter and focused onto the preparation using a quartz aspherical doublet. The emitted fluorescence light was collected by a 40x objective and simultaneously detected at 405 ± 5 nm (F₄₀₅) and 470 ± 5 nm (F₄₇₀) by two photomultipliers. The fluorescence measurement field was 40 μm in diameter and the silicone-free extremity of each tested fibre was placed in the middle of the field. Background fluorescence at both emission wavelengths was measured next to each tested fibre and was then subtracted from all measurements. The presence of movement artifacts was routinely detected in the raw indo-1 fluorescence transients measured in response to strong membrane depolarizations. As previously described (Jacquemond, 1997), this could be assessed from non-linearity between the normalized time course of change in fluorescence at the two wavelengths. However, in most cases the motion-related changes in fluorescence were of small amplitude as compared to the calcium-dependent changes in fluorescence. Furthermore, even when clearly present in the original fluorescence traces, motion artifacts appeared to be largely eliminated in the ratio (F₄₀₅/F₄₇₀) signal.

Calibration of the indo-1 response and calculation of [Ca²⁺]_i

The standard ratio method was used with the parameters: R = F₄₀₅/F₄₇₀, R_min and R_max the fluorescence ratio in the absence of Ca²⁺ and in the presence of saturating Ca²⁺, respectively, and the dissociation constant, K_D and the ratio of F₄₇₀ in the absence of Ca²⁺ to F₄₇₀ in the presence of saturating Ca²⁺, β. Results were either expressed in terms of indo-1 percentage saturation or in actual free calcium concentration (for details of calculation, see Jacquemond, 1997; Csernoch et al. 1998). In vivo values for R_min, R_max and β were measured using procedures previously described (Collet et al. 1999; Collet & Jacquemond, 2002). No correction was made for indo-1–Ca²⁺ binding and dissociation kinetics, so that the derived time course of change in [Ca²⁺]_i does not exactly reflect the true amplitude and time course of change in ionized calcium.

In vitro measurements of the indo-1 response were performed in the presence of SNP and NOC-12. These were performed with the experimental set-up using glass capillary tubes filled with calibrating solutions (Jacquemond, 1997; Collet & Jacquemond, 2002). These solutions were prepared from two stock solutions containing (mm): KCl 100, EGTA 10 and Pipes 10; with or without 10 mm CaCl₂. Various free calcium concentrations were achieved by mixing these two solutions in different ratios. pH was adjusted to 7.20. Indo-1 (50 μm) and either SNP (1 mm) or NOC-12 (1 mm) were added to these solutions. Fitting the results from a control and a SNP calibration by a 1 : 1 binding curve gave values for R_max, R_min and K_Dβ of 4.26, 0.61 and 0.39 μm in control and 4.34, 0.59 and 0.39 μm in the presence of SNP (under our in vitro conditions β is approximately equal to 1.2 and the actual K_D is thus close to 0.3 μm), respectively. These values are sufficiently close to justify the use of common calibration parameters. Such calibrations were impossible in the presence of 1 mm NOC-12 because under the same conditions as above there was just no detectable indo-1 fluorescence for free calcium concentrations lower than the micromolar level. The ratio value at a saturating [Ca²⁺] in the presence of NOC-12 (R = 4.2 for [Ca²⁺]= 0.27 mm) was though similar to the one measured in control and in the presence of SNP. The loss of fluorescence at non-saturating [Ca²⁺] was probably due to an interaction between free indo-1 and NOC-12 under the in vitro conditions. This problem was not encountered during the in vivo experiments: indo-1 fluorescence was clearly detected from the fibres bathed in the presence of NOC-12 and no difference in the value of initial resting fluorescence ratio and in the value of peak fluorescence ratio in response to a given pulse was detected between NOC-12-treated fibres and control or SNP-treated fibres. Furthermore we calculated the slope m from plots of F₄₇₀ versus F₄₀₅ from fluorescence transients elicited by membrane depolarizations in the NOC-12-treated fibres and in the fibres used for control experiments and issued from the same muscles. The value of m is proportional to the parameters R_max, R_min and β (for details see Jacquemond, 1997) and it should thus be affected if the indo-1 properties within the fibres treated with NOC-12 differed from the ones in control fibres. The mean value for m was −0.49 ± 0.03 (n = 7) in NOC-12-treated fibres, not significantly different (P = 0.22) from the mean value from the control fibres (−0.56 ± 0.04, n = 6). We thus believe that NOC-12 did not significantly interfere with the [Ca²⁺] measurements in the fibres, presumably because of poor membrane permeability to this compound.

Solutions

The intracellular-like solution contained (mm): potassium glutamate 120, Na₂ATP 5, sodium phosphocreatine 5, MgCl₂ 5.5, glucose 5 and Hepes 5; pH adjusted to 7.20 with KOH. The standard extracellular solution contained (mm): TEA-methanesulphonate 140, CaCl₂ 2.5, MgCl₂ 2, TEA-Hepes 10 and TTX 0.002; pH 7.20. For measurements of the slow calcium current (Fig. 6) this solution also contained 1 mm 4-aminopyridine to further block outward potassium conductances, and the fibres were pressure microinjected with a solution that contained 50 mm EGTA (Collet et al. 2003).

Figure 6. Slow calcium current under control conditions and in the presence of SNP.

A, calcium current records obtained in response to 1 s-long depolarizing pulses to values between −50 and +50 mV in a control fibre (left) and in a fibre exposed to 2 mm SNP (right). Superimposed thick curves correspond to the result from fitting a single exponential function to the decaying phase of the current. B, mean voltage dependence of the peak calcium current amplitude in control fibres (•, n = 6) and in fibres exposed to 2 mm SNP (○, n = 4). In both panels, the superimposed curve was calculated using the mean parameters obtained from fitting the voltage dependence in each fibre (see text for details). C, corresponding mean values for the time constant of calcium current decay.

For measurements of intramembrane charge movement the extracellular solution contained (mm): TEA-methanesulphonate 140, CaCl₂ 0.25, MgCl₂ 4, CdCl₂ 0.5, LaCl₃ 0.3, 4-aminopyridine 1, TEA-Hepes 10, TTX 0.002 and N-benzyl-p-toluene sulphonamide (BTS) 0.1; pH 7.20.

The NO donors sodium nitroprusside (SNP) and N-ethyl-2-1-ethyl-2-hydroxy-2-nitrosohydrazino-ethanamine (NOC-12) were purchased from Sigma. They were diluted in the extracellular solution just before use.

Statistics

Least-squares fits were performed using a Marquardt-Levenberg algorithm routine included in Microcal Origin (Originlab, Northampton, MA, USA). Data values are presented as means ± s.e.m. for n fibres, where n is specified in the Results. Statistical significance was determined using a Student's unpaired t test assuming significance for P < 0.05.

Results

Use-dependent increase in resting [Ca²⁺] in the presence of the NO donor SNP

The most straightforward effect of SNP on intracellular calcium in a voltage-clamped muscle fibre is illustrated in Fig. 1. Figure 1A shows a series of 15 successive indo-1 saturation transients elicited by a 50 ms-long depolarizing pulse to +10 mV applied every 30 s, in a control fibre (top) and in a fibre exposed to 2 mm SNP (bottom). For each series of traces, the panel on the right presents an enlarged view of the first (continuous line) and last response (dotted line) superimposed. Whereas the response was very stable in the control fibre, the resting indo-1 saturation level progressively increased in the fibre exposed to SNP and there was a concomitant depression in the peak level of the indo-1 transient. The left panel in Fig. 1B shows the time course of the mean value for resting [Ca²⁺] measured in fibres that were challenged by the same pulse protocol as in Fig. 1A, under control conditions (filled circles) and in the presence of either 1 mm or 2 mm SNP in the extracellular solution (open symbols). In all of these experiments, fibres were bathed in the SNP-containing solution for at least 15 min before voltage clamp was achieved. The mean initial resting [Ca²⁺] did not significantly differ between control fibres and fibres exposed to SNP. In the presence of 1 mm SNP, over a period of 8 min following application of the first depolarizing pulse, resting [Ca²⁺] increased from 98 ± 15 nm to 205 ± 29 nm (n = 4, P = 0.017). In the presence of 2 mm SNP, resting [Ca²⁺] increased from 119 ± 10 nm to 204 ± 23 nm over a period of 6 min (n = 13, P = 0.002). In the control fibres, the resting [Ca²⁺] after 8 min was 106 ± 11 nm as compared to 95 ± 12 nm at the beginning of the experiment (n = 10, P = 0.5). Under the same conditions, SNP at a concentration of 0.1 mm was found to be ineffective (not shown) with a resting [Ca²⁺] of 117 ± 28 nm after 8 min as compared to 108 ± 19 at the beginning of the experiment (n = 3, P = 0.8). As described next in relation to Fig. 5, a use-dependent increase in resting [Ca²⁺] was also observed with the chemically distinct NO donor NOC-12, making it unlikely to be due to a NO-independent pharmacological effect.

Figure 1. Indo-1 calcium transients in the presence of the NO donor SNP.

A, indo-1 saturation signals elicited by 50 ms-long voltage-clamp depolarizations to +10 mV applied every 30 s in a control fibre (top) and in a fibre exposed to 2 mm SNP (bottom). For each fibre, the first (line) and last signal (dots) are shown superimposed on an expanded time scale on the right part of the panel. B, time course of the mean resting [Ca²⁺] (left) and of the mean peak change in [Ca²⁺] (right) in control fibres (n = 10) and in SNP-exposed fibres (1 mm SNP, n = 4; 2 mm SNP, n = 13 up to t = 6 min and n = 7 thereafter) under the same conditions as in A. In B, the peak change in [Ca²⁺] was normalized to its initial value in each fibre.

Figure 5. The NO donor-induced increase in resting [Ca²⁺] is accompanied by a differential effect on the peak change in [Ca²⁺] elicited at −20 and +10 mV.

The pulse protocol shown in A was applied every 30 s in control fibres (left), in fibres exposed to SNP (middle) and in fibres exposed to NOC-12 (right). B, superimposed examples of indo-1 signals elicited at the beginning of the experiment (line) and after a certain time (dots) under each condition (9 min in control and SNP, 14 min in NOC-12). C, time course of the mean peak change in [Ca²⁺] elicited by the pulse to −20 mV (squares) and to +10 mV (circles), in control conditions (n = 7), in SNP-treated fibres (n = 10) and in NOC-12-treated fibres (n = 4). D, corresponding time course of the mean resting [Ca²⁺] measured at the beginning of each trace. E, dependence of the peak Δ[Ca²⁺] elicited by the pulses to −20 mV (○) and to +10 mV (•) upon the preceding resting [Ca²⁺] in fibres exposed to SNP and NOC-12 and stimulated by the 3-step protocol. See text for details. Data are from the same fibres as in Fig. 5B–D, but also include measurements taken at later times when available.

The mean values of absolute peak change in [Ca²⁺] (Δ[Ca²⁺]) from the first transient elicited in control and in SNP-treated fibres under the conditions of Fig. 1 were 1.0 ± 0.2 μm (control, n = 10), 0.83 ± 0.2 μm (SNP 1 mm, n = 4) and 0.72 ± 0.12 (SNP 2 mm, n = 13), which were not significantly different. The right panel of Fig. 1B shows the time course of the relative mean value for the peak Δ[Ca²⁺] in these fibres during the experiments. Values were obtained by normalizing the peak Δ[Ca²⁺] values in each fibre to the corresponding peak Δ[Ca²⁺] value measured in response to the first depolarization. On average, the peak Δ[Ca²⁺] value was depressed by ∼50% within 8 min in the presence of SNP.

Figure 2A shows a series of indo-1 transients measured from a fibre in the presence of 2 mm SNP and stimulated by 50 ms-long pulses to +10 mV, in the absence of extracellular calcium (CaCl₂ was replaced by MgCl₂ in our standard solution). It shows essentially the same pattern as in the presence of calcium: there was a progressive rise in the resting level of the transient while its amplitude dropped. The right panel of Fig. 2A shows the time course of the mean values of resting [Ca²⁺] under these conditions (open squares, n = 6) superimposed onto the mean values measured with 2 mm SNP in the presence of calcium (open triangles). There was no difference between the two, so calcium entry from the extracellular medium is very unlikely to play a role in the NO-induced rise in resting [Ca²⁺]. Mean values for the slopes of the increase in resting [Ca²⁺] from all fibres that received 50 ms-long pulses to +10 mV under the different conditions used in this study are presented in Fig. 2B. Values for the slopes were obtained by fitting a linear function to the individual series of data points against time in each fibre. NOC-12 appeared less effective than SNP which may be related to the fact that it releases NO more slowly than SNP (see for example Zou & Cowley, 1997; Sun et al. 2003).

Figure 2. Increase in resting [Ca²⁺] in fibres exposed to NO donors in the presence and absence of extracellular calcium.

A, indo-1 saturation signals elicited by 50 ms-long depolarizations to +10 mV applied every 30 s in a fibre bathed in 2 mm SNP in the nominal absence of extracellular calcium. The panel on the right shows the time course of the mean resting [Ca²⁺] measured in fibres bathed in 2 mm SNP in the presence of extracellular calcium (▵, same fibres as in Fig. 1B) and in the absence of extracellular calcium (□, n = 6), under the same conditions of stimulation. B, mean values for the slope of increase in resting [Ca²⁺] in control and NO donor-exposed fibres. Values were obtained both from fibres stimulated with simple pulses (as in Fig. 1) and from fibres stimulated with the 3-step protocol (as in Fig. 5). Values in the presence of millimolar levels of NO donors are significantly elevated as compared to control values (control, n = 13; SNP 0.1 mm, n = 3; SNP 1 mm, n = 13; SNP 2 mm, n = 12; SNP 2 mm 0 Ca, n = 6; NOC-12, n = 4). Mean values in 1 and 2 mm SNP are not significantly different.

The holding current remained stable in the presence of SNP: over the course of the experiments described in Fig. 1 the mean holding current value after 6 min was 101 ± 5% and 89 ± 3% of its initial value in control and in SNP-treated (1 and 2 mm) fibres, respectively. This indicates that the increase in resting [Ca²⁺] was neither accompanied by, nor due to, a drastic alteration of the membrane conductance.

The increase in resting [Ca²⁺] observed in the presence of SNP was a clearly use-dependent phenomenon. This could be assessed from the fact that the initial resting [Ca²⁺] was completely independent of the amount of time the fibres were left in the presence of SNP before they were voltage clamped and the first depolarizing pulse applied (within a period of 15 min to 2 h). Also, in two fibres held at −80 mV in the presence of 1 mm SNP, we took indo-1 fluorescence measurements every 30 s for 8 min before applying depolarizing pulses to +10 mV. Only when the pulses were applied did resting [Ca²⁺] start to increase (not illustrated).

Myoplasmic Ca²⁺ removal in the presence of SNP

An increase in resting [Ca²⁺] can result from either an increased calcium influx into, or a reduced calcium efflux from, the myoplasmic compartment. Because in skeletal muscle, myoplasmic calcium extrusion is essentially under the control of the SR Ca²⁺-ATPase, the effect reported here could be thought to result from an inhibition of its activity. Results presented in Fig. 3 suggest that this is not the case. Under voltage-clamp conditions, Ca²⁺ release turns off rapidly following the end of a step depolarization, so that the following decay of the Ca²⁺ transient can be reasonably assumed to reflect the Ca²⁺ removal capabilities of the fibre. Figure 3A shows two superimposed [Ca²⁺] transients normalized to the same maximum amplitude, from a control fibre (left panel) and from a fibre bathed in 2 mm SNP (right panel). These transients were calculated from the data shown in Fig. 1A; they correspond to the first and to the 15th (7 min later) response elicited in the control fibre and in the SNP-exposed fibre. The thick curve superimposed onto the decay of each trace corresponds to the result from fitting a single exponential plus constant function. In the control fibre, the time course of the two transients was identical. In the SNP treated fibre, despite the increase in resting [Ca²⁺] (from 0.11 to 0.22 μm) and decrease in peak Δ[Ca²⁺] (from 0.55 to 0.2 μm) the time course of decay of the two transients was also similar (time constant of 116 ms for the first record and of 99 ms for the 15th record). This was generally the case although in a few fibres a substantial slowing of the decay occurred at late times during the experiment, when large levels of resting [Ca²⁺] were reached (not illustrated). We performed exponential fits to the decay of complete series of [Ca²⁺] transients measured from the fibres used in Fig. 1. Results are shown in Fig. 3B. Data from fibres exposed to 1 mm and 2 mm SNP were pooled together (n = 17). Figure 3B shows the time course of the mean resting [Ca²⁺] (left) and of the mean time constant of [Ca²⁺] decay (middle) in control (filled circles) and in SNP-treated fibres (open circles). It shows that, over a 6-min period, the mean time constant of decay never significantly differed between control and SNP-treated fibres; within the same period the mean resting [Ca²⁺] from SNP-treated fibres increased from 117 ± 9 nm to 200 ± 18 nm (P = 0.0002). We also tested whether fibre-to-fibre variability could mask a differential evolution of the Ca²⁺ decay in control and in the presence of the NO donor. For this, in each fibre the time constant of decay of the successive Ca²⁺ transients was normalized to the value of the time constant measured from the first Ca²⁺ transient elicited. The right panel of Fig. 3B shows the corresponding mean normalized values against time. It shows that after several minutes there was a tendency for a slow increase in the time constant of decay against time in the presence of SNP. For example, after 6 min the time constant was increased by 40.6 ± 14% in SNP as compared to 12.4 ± 6% under control conditions. However fitting the values of the normalized time constants against time by a linear function in each fibre, gave a mean increase of 3.5 ± 1% per 30 s in SNP-treated fibres (n = 17), as compared to 1.1 ± 0.5% per 30 s in control fibres (n = 10); these values are not significantly different (P = 0.09). So there was no systematic slowing of the [Ca²⁺] decay in the presence of SNP. Although this is insufficient to definitely exclude the SR Ca²⁺ pump as a target for NO under our conditions, it clearly indicates that SNP-treated fibres were able to maintain their cytoplasmic Ca²⁺ removal capabilities.

Figure 3. Resting [Ca²⁺] and rate of decay of the [Ca²⁺] transients in fibres exposed to SNP.

A, [Ca²⁺] transients normalized to the same maximal amplitude in response to the first and to the 15th 50 ms-long depolarization to +10 mV applied to a control fibre (left) and to a fibre exposed to 2 mm SNP (right). Fibres are the same as in Fig. 1A. The result from fitting a single exponential plus constant function is shown superimposed onto the decay phase of each transient. In the control fibre the two traces and corresponding fits are undistinguishable. The time constant of decay was 83 ms initially and was 88 ms after 6 min. In the SNP-treated fibre, the time constant of decay was 116 ms initially and was 99 ms after 7 min. B, time course of the mean resting [Ca²⁺] (left), mean time constant of [Ca²⁺] decay (middle) and mean time constant of [Ca²⁺] decay normalized to its initial value (right) in control fibres (•, n = 10) and in fibres exposed to 1 mm and 2 mm SNP (○, n = 17). All fibres were stimulated by 50 ms-long pulses to +10 mV applied every 30 s.

The slope of the SNP-induced rise in resting [Ca²⁺] depends on the amplitude of the depolarizing pulses

The above results are consistent with NO being responsible for a depolarization-induced continuous release of SR Ca²⁺. One possibility is that some release channels that open during depolarization remain open after membrane repolarization (see for example Collet & Jacquemond, 2002). If true, one may expect the amplitude of the rise in resting [Ca²⁺] to depend upon the extent of the preceding activation of voltage-induced SR Ca²⁺ release. Figure 4 shows results that support this possibility. It illustrates the effect of SNP (2 mm) in fibres that were successively stimulated by membrane depolarizations of two different amplitudes. Indo-1 transients were first elicited by 50 ms-long pulses to −20 mV during 5 min and the pulse amplitude was then raised to +10 mV. An example of corresponding indo-1 signals is shown in Fig. 4B. The left panel in Fig. 4C shows the time course of the mean value for resting [Ca²⁺] from four fibres under the same conditions. There was a clear increase in the slope of rise in resting [Ca²⁺] when the pulse amplitude was raised: on average, the mean slope increased from 6.3 ± 1 nm min⁻¹ with the pulses to −20 mV to 15.9 ± 3 nm min⁻¹ with the pulses to +10 mV (n = 4, P = 0.015).

Figure 4. The slope of the SNP-induced rise in resting [Ca²⁺] increases with the amplitude of the depolarizing pulse.

A, shows the pulse protocol corresponding to −20 mV during the first 5 min and then to +10 mV. B, indo-1 saturation signals: a 50 ms long voltage-clamp depolarization was applied every 30 s in a fibre exposed to 2 mm SNP. C, time course of the mean resting [Ca²⁺] (left) and of the mean peak change in [Ca²⁺] (right), under the same conditions as in B (n = 4). The mean slope of increase in resting [Ca²⁺] was 6.3 ± 1 nm min⁻¹ with the pulse to −20 mV and 15.9 ± 3 nm min⁻¹ with the pulse to +10 mV (P = 0.015).

Resting [Ca²⁺] and peak Δ[Ca²⁺] in the presence of NO donors

The right panel in Fig. 4C shows the mean values for the peak Δ[Ca²⁺] along the course of that series of experiments. Surprisingly, in response to the pulses to −20 mV, the peak Δ[Ca²⁺] was stable or even tended to slowly increase with time. Conversely and in accordance with results shown in Fig. 1, the peak Δ[Ca²⁺] progressively decreased when pulses to +10 mV were applied. The reasons for this apparent discrepancy were unclear. As the resting [Ca²⁺] progressively increased throughout these experiments, and this in a use-dependent manner, the Δ[Ca²⁺] elicited by pulses to −20 mV systematically started from lower levels of resting [Ca²⁺] than the Δ[Ca²⁺] elicited by pulses to +10 mV. One could then argue that the difference in the direction of the change in Δ[Ca²⁺] (stable or progressive increase at −20 mV versus progressive decrease at +10 mV) was not related to the pulse amplitude but to the different levels of resting [Ca²⁺]. The purpose of the following section was to discriminate between these two possibilities, that is to determine whether this differential effect would still be observed under conditions where the Δ[Ca²⁺] elicited by both pulses began from a similar level of resting [Ca²⁺]. For this we measured calcium transients in response to pulses to both −20 and +10 mV at different times following exposure to NO donors. A protocol consisting of three 50 ms-long depolarizing pulses to −50, −20 and +10 mV (Fig. 5A, referred to as the 3-step protocol in the following sections) was applied every 30 s in control fibres and in fibres exposed to either SNP or NOC-12.

Figure 5B shows superposition of the first indo-1 response (line) and of the response recorded after a certain time in a control fibre (left, dotted trace recorded after 9 min), in a fibre exposed to 1 mm SNP (middle, dotted trace recorded after 9 min), and in a fibre exposed to 1 mm NOC-12 (right, dotted trace recorded after 14 min). Figure 5C shows the time course of the mean values for the peak Δ[Ca²⁺] at −20 mV (squares) and +10 mV (circles) along the course of these experiments, while Fig. 5D shows the corresponding mean values of resting [Ca²⁺]. Qualitatively, NOC-12 produced the same effect as SNP on resting [Ca²⁺]. Also, in both SNP- and NOC-12-treated fibres, there was a clear tendency for the rise in resting [Ca²⁺] to be accompanied by a depression of the peak Δ[Ca²⁺] at +10 mV. Within the same period, the average peak Δ[Ca²⁺] during the pulse to −20 mV remained fairly stable. It should to be noted that in some of the SNP-exposed fibres, the 3-step protocol was applied over much longer periods of time, in which case the peak Δ[Ca²⁺] at −20 mV also ended up being depressed (not illustrated). Within the series of experiments described in Fig. 5, the membrane current density at the different voltages tested did not significantly differ between control and NO donor-treated fibres, either at the beginning of the experiments or 9 min later, indicating that there was no change in membrane conductance associated with the effects of NO donors (not shown).

The progressive reduction of the peak calcium transient elicited by a pulse to +10 mV in the presence of SNP or NOC-12 seemed to parallel the increase in resting [Ca²⁺]. This may indicate that under these conditions, Ca²⁺ release was depressed as a consequence of the rise in the resting [Ca²⁺], for instance through a calcium-dependent negative feedback mechanism on the release channel activity. However if true, it is hard to understand how this effect was not observed for pulses to −20 mV over the same range of resting [Ca²⁺]. This is further examined in Fig. 5E which shows the dependence of the peak Δ[Ca²⁺] elicited by the pulses to −20 mV (open circles) and +10 mV (filled circles) upon the immediately preceding resting [Ca²⁺] level in the fibres exposed to SNP and NOC-12 and stimulated by the 3-step protocol (data are from the same fibres as in Fig. 5B and C but also include measurements taken at later times when available). For this, individual pairs of values of resting [Ca²⁺] and peak Δ[Ca²⁺] were pooled and sorted according to the resting [Ca²⁺] level. The peak Δ[Ca²⁺] values for which the corresponding resting [Ca²⁺] fell within the same range (with an increment arbitrarily set to 0.04 μm) were then averaged and plotted against the mean resting [Ca²⁺]. For a pulse to +10 mV the peak Δ[Ca²⁺] dropped continuously as resting [Ca²⁺] increased, whereas for the pulse to −20 mV, a significant drop in peak Δ[Ca²⁺] only occurred for resting [Ca²⁺] near and above 0.2 μm. In order to assess this difference quantitatively we fitted the two sets of data points with a Hill equation. This gave values for the concentration of half inhibition and Hill coefficient of 0.23 μm and 6.5 for the data points from pulses to −20 mV and 0.17 μm and 3.0 for data points from pulses to +10 mV, respectively.

Functional properties of the dihydropyridine receptor in the presence of SNP

We tested the possibility that NO may act on the voltage-sensing steps of excitation–contraction coupling by investigating the properties of the dihydropyridine receptor in terms of slow calcium channel and of voltage sensor. This was achieved by measuring, in separate sets of experiments, the slow calcium current and the intramembrane charge movement in control fibres and in fibres exposed to SNP.

Figure 6A shows calcium current traces measured from a control fibre (left) and from a fibre in the presence of 2 mm SNP (right). Traces were obtained in response to 1 s-long membrane depolarizations ranging between −50 and +50 mV with a 10 mV increment. For membrane potential values more depolarized than −10 mV, the decaying phase of the current was fitted with a single exponential function, the result of which is shown superimposed to the corresponding traces. The voltage dependence of the mean peak current density and of the mean time constant of current decay are shown in Fig. 6B and C, respectively, for control fibres (filled circles, n = 6) and for fibres in the presence of 2 mm SNP (open circles, n = 4). There were only minor differences in the properties of the current between control and SNP-treated fibres. The voltage dependence of the peak current was fitted in each fibre as described in the Methods. Values for G_max, V_rev, V_0.5 and k were 173 ± 25 S F⁻¹, 65 ± 3 mV, −6.6 ± 0.4 mV and 6.1 ± 0.4 mV in control fibres and 163 ± 14 S F⁻¹, 61.4 ± 4 mV, −10.2 ± 1.2 mV and 4.7 ± 0.4 mV in the presence of SNP, respectively. The 3–4 mV leftward shift of the voltage dependence in the presence of SNP was significant (P = 0.01). Other parameters including values for the time constant of current decay did not significantly differ between the two conditions.

Concerning the properties of intramembrane charge movement under exposure to NO, we were specifically interested in achieving these measurements under the protocol conditions of Fig. 5, in order to see whether the differences observed between Ca²⁺ release at −20 and +10 mV in the presence of NO would correlate with a similar effect on the voltage sensor. Figure 7A shows charge movement records measured in response to the pulses of the 3-step protocol in a fibre exposed to 2 mm SNP. Top records were obtained in response to the first application of the protocol while bottom records were measured 9 min later. The amount of ‘on’ charge moved at −50, −20 and +10 mV was 2.6, 9.7 and 27.9 nC μF⁻¹, respectively, to start with and 3.1, 12.8 and 25.6 nC μF⁻¹, respectively, 9 min later. Figure 7B shows the mean amount of ‘on’ charge moved against time in control fibres (filled circles, n = 8) and in fibres exposed to 2 mm SNP (open circles, n = 7). In order to test whether there was any systematic change during the course of the experiments, we fitted a linear function to the amount of charge measured at the three membrane potentials in each individual fibre, as a function of time. Statistical analysis using one sample t tests showed that whatever the pulse potential, the mean slope never significantly differed from zero in both control and SNP-treated fibres (P > 0.3). Averaging the data measured at the different times in each fibre gave a mean amount of charge moved at −50, −20 and +10 mV of 1.81 ± 0.14, 8.71 ± 0.85 and 21.90 ± 1.85 nC μF⁻¹ in control fibres (n = 8), and of 2.80 ± 0.38, 11.70 ± 1.23 and 23.89 ± 1.99 nC μF⁻¹ in SNP treated fibres (n = 7), respectively. Normalizing the values at −50 and −20 mV to that at +10 mV in each fibre gave a ratio of 0.084 ± 0.003 and 0.118 ± 0.015 in control and SNP-treated fibres at −50 mV, respectively, and of 0.40 ± 0.02 and 0.49 ± 0.02 in control and SNP treated fibres at −20 mV, respectively. The significantly larger ratios observed at −50 and −20 mV are consistent with a slight negative shift of the voltage dependence of charge movement in the presence of SNP.

Figure 7. Intramembrane charge movement in the presence of SNP.

A, intramembrane charge movement records obtained from a fibre bathed in 2 mm SNP at the indicated values of membrane depolarization. The top series of traces was obtained at the beginning of the experiment; the bottom series was obtained 9 min later. The fibre was stimulated by the 3-step protocol applied every 30 s (as in Fig. 5). Mean values for the amount of ‘on’ charge moved by pulses to −50, −20 and +10 mV in control fibres (•, n = 8) and in fibres bathed in 2 mm SNP (○, n = 7), against time.

Dithiothreitol reverses the effects of SNP and NOC-12

The reducing agent dithiothreitol (DTT) was previously shown to reverse the effects of NO on isolated RyRs (Hart & Dulhunty, 2000). It was also demonstrated that DTT can permeate across cell membranes and can act on proteins within the reticulum of living cells (Braakman et al. 1992). In Fig. 8A–E each row shows indo-1 saturation traces recorded from the same fibre in response to 50 ms-long pulses to +10 mV. The left panel shows the first response under the indicated condition (either control or in presence of SNP or NOC-12 at 2 mm). The middle panel shows the response some time after repeating the same pulse every 30 s under that same condition, while the panel on the right shows the response measured after replacing the extracellular solution with either the control solution (wash) or with the control solution plus 10 mm DTT. Figure 8A and B shows that the increase in resting [Ca²⁺] and decrease in amplitude of the indo-1 transient produced in the presence of either SNP or NOC-12 were not reversed upon removing the NO donor from the external solution. Conversely, when DTT was present in the washing solution, resting [Ca²⁺] clearly returned towards its original level while the amplitude of the transient tended to increase again (Fig. 8D and E). Figure 8C shows that DTT had very little effect when tested under control conditions: in three control fibres the mean resting [Ca²⁺] and mean peak Δ[Ca²⁺] after applying DTT corresponded to 95 ± 5% and 114 ± 3% of the control values, respectively. In a total of eight fibres exposed to NO donors (six fibres exposed to SNP and two fibres exposed to NOC-12 for a period of time that varied from 6 to 23 min), washing with the DTT-containing solution produced a systematic decrease in resting [Ca²⁺] and increase in peak Δ[Ca²⁺], although to a variable extent. Mean values for the relative change in resting [Ca²⁺] and peak Δ[Ca²⁺] (normalized to the initial value measured in each fibre) at the end of the application of NO donor, and after washing with DTT are reported in Fig. 8F and G.

Figure 8. Effects of NO donors are reversed by dithiothreitol.

A–E, all traces correspond to indo-1 saturation records elicited by 50 ms-long depolarizations to +10 mV under the indicated conditions. Each row corresponds to a different fibre. Responses that are shown were measured at the beginning of the experiment (left), 10–15 min later (middle) and after replacing the extracellular solution either with the control solution (wash) or with the control solution also containing 10 mm dithiothreitol (DTT) (right). Depolarizations were applied every 30 s. F, mean relative change in resting [Ca²⁺] when exchanging the NO-donor-containing solution for the DTT-containing solution (n = 8). G, corresponding mean relative change in peak Δ[Ca²⁺]. In F and G, values from each fibre were normalized to the value measured at the beginning of the experiment. Mean data are from eight fibres repeatedly depolarized in the presence of SNP (n = 6) or NOC-12 (n = 2) for a period ranging between 6 and 23 min before washout with DTT.

Discussion

The present work provides new insights into the role of NO in the regulation of intracellular calcium in intact skeletal muscle fibres. Main conclusions are threefold: (i) excess NO does not have a strong direct and acute effect on voltage-activated Ca²⁺ release; (ii) excess NO opens up release channels in a use-dependent and redox-sensitive manner; and (iii) the resulting chronic resting Ca²⁺ leak is likely to play a major role in the ultimate alteration of excitation–contraction coupling produced by excess NO. This overall response of skeletal muscle has to be of strong relevance under conditions of nitrosative stress.

NO as a use-dependent Ca²⁺ release channel opener

Despite numerous studies using SR preparations and isolated ryanodine receptors (for review see Eu et al. 1999; Stamler & Meissner, 2001) no clear picture has emerged on how the function of skeletal muscle calcium release channels is modulated by NO. Of course, this modulation remains even more obscure when it comes to the channel within its native environment, under its physiological mode of gating. One aspect of the complexity certainly resides within the fact that NO is likely to have different or even opposite effects depending on various parameters, one of which being its concentration. However, even within studies that compared effects of low (typically submicromolar) and high (> 1 μm) levels of NO, the high levels were either reported to close the channel (Hart & Dulhunty, 2000), or to activate the channel (Aghdasi et al. 1997), or to close and activate the channel at low and high P_O2, respectively (Eu et al. 2000).

Our results show that in an intact muscle fibre, excess NO produces a use-dependent rise in resting [Ca²⁺]. This effect was still observed in the absence of extracellular calcium, ruling out the possibility of an increased calcium entry through the plasma membrane. Analysis of the decay of the [Ca²⁺] transients revealed that cytoplasmic Ca²⁺ removal was still effective in the presence of NO. The fact that some fibres exhibited a slowing of the decay after multiple transients in the presence of NO donors may actually have occurred as a consequence of the raised resting Ca²⁺ level. Indeed, the time course of Ca²⁺ decay following membrane repolarization depends upon SR uptake but is also affected by the saturation level of some intrinsic Ca²⁺ buffers, in particular the Ca-Mg binding sites on parvalbumin (Melzer et al. 1986; Garcia & Schneider, 1993; Jiang et al. 1996; Baylor & Hollingworth, 2003). As resting [Ca²⁺] progressively increased in the presence of NO donors, this may per se have sufficiently reduced the calcium buffering capacity of parvalbumin to significantly affect the rate of Ca²⁺ removal during a pulse, independently of any putative effect on the SR pump (see also Caputo et al. 1999).

We cannot definitely eliminate the possibility that the increase in resting [Ca²⁺] in the presence of NO donors occurs through NO-modified calcium pumps. However, we believe that it is more likely to primarily result from an increased population of open release channels. Also in favour of this interpretation is the striking similarity between the effects of NO donors and the effects of the specific release channel opener ryanodine (Collet & Jacquemond, 2002) under the same conditions.

The dihydropyridine receptor could also be thought of as a potential target of NO and this actually may be responsible for the apparent slight negative shift in the voltage dependence of activation of its voltage sensing and calcium channel functions. On the other hand, the NO-induced progressive decline in voltage-activated Ca²⁺ release is clearly not related to a depression of intramembrane charge movement, ruling out a role of the voltage sensor in this effect.

One could also speculate that the SR Ca²⁺ leak results from an effect of NO on a small proportion of voltage sensors that would then maintain some calcium release channels open, but without significant modification of the total amount of charge moved by a large depolarization. However, we believe that it is the calcium release channel that is more likely to be directly affected by NO because it would require a very remarkable alteration of the gating properties of the voltage sensors for them to remain under a chronic active configuration at −80 mV. Also in line with this interpretation is the large number of studies showing alteration of the isolated skeletal muscle calcium release channel function by NO, whereas concurrent data concerning the dihydropyridine receptor is clearly lacking.

The rise in resting [Ca²⁺] observed here was not reported by Andrade et al. (1998) who worked on the same preparation under conditions of field stimulation. Specific differences in the experimental procedures are probably responsible for this discrepancy. We believe that the time of application of the NO donors, as well as the pattern of stimulation, may well be the main responsible factors. Andrade et al. (1998) applied tetanic stimulations at 1-min intervals and reported effects within 1 min of application of NO donors. None of the effects we describe in our study would fit within such a short period of time. Conversely, Andrade et al. (1998) also reported effects that were stable within 3–5 min exposure to the drugs and that were then fully reversible. This is clearly different from our results where, after a few minutes, resting [Ca²⁺] and [Ca²⁺] transients were definitely not stable and effects were irreversible upon wash-out of the NO donors. It thus seems that the two studies point to different effects of NO. The one by Andrade et al. (1998) focuses on a rapid and fully reversible increase in the [Ca²⁺] transient upon acute NO application. Note that this increase is relatively modest (∼10%) and would be very hard to unravel under our conditions of chronic exposure to NO, within the variability of the peak [Ca²⁺] transient value from fibre to fibre. The fact that they did not observe the rise in resting [Ca²⁺] may be related to differences in the pattern of stimulation under the two conditions. Andrade et al. (1998) used trains of 0.5 ms-long stimulations. According to their figures the duration of the trains was about 400 ms and they report a stimulation frequency during the trains between 20 and 60 Hz. Assuming the duration of activation of the voltage sensors during a single action potential to be 2 ms, a 400 ms long train at 40 Hz could be approximated to correspond to an ∼30 ms-long depolarization. This, together with the 1-min interval between two trains may have been insufficient to induce a clear rise in resting [Ca²⁺] over a period of a few minutes. Alternatively, we cannot exclude the possibility that NO only acts on sites or steps that occur under a sustained depolarization.

One question stands, that is why are millimolar concentrations of NO donors required to observe an effect under our conditions, whereas much lower concentrations are also effective in ryanodine receptor single channel studies (see for example Sun et al. 2003; for NOC-12)? This could be related either to the endogenous NO-scavenging capacity of the intact fibre preparation or conversely to an already substantial level of intracellular NO under our conditions. One may also think of a different NO sensitivity of the release channel between its isolated form and under native conditions. Along this line and according to the work from Eu et al. (2000), the discrepancy could be argued to result from a reduced sensitivity of the release channel to NO under conditions of ambient O₂ tension; in other words, under our control conditions, release channels may operate in an oxidized state with an altered sensitivity to NO. However, the fact that DTT had no effect on the resting [Ca²⁺] and on the calcium transient under our control conditions argues against this possibility. Alternatively, it remains possible that lower concentrations of NO donors would also be effective under our conditions, but would require much longer periods of time than explored here and/or more intense conditions of stimulation.

The use-dependence of the rise in resting [Ca²⁺] is of mechanistic interest. The simplest explanation is that only within the population of release channels that open during membrane depolarization, some will react with NO, lose their control by the voltage sensors and thus remain open after membrane repolarization. One may then speculate that at rest, the release channel in its closed state keeps its NO target(s) out of either access or reactivity and only the conformational change associated with its opening allows the interaction. Alternatively, the use-dependence may actually reflect a [Ca²⁺] dependence of the process, meaning that the effectiveness of NO would be determined by the level of free Ca²⁺ reached in response to the membrane depolarization. Such a mechanism could be consistent with the reported calmodulin dependence of NO modulation of the isolated ryanodine receptor channel activity (Eu et al. 2000; Sun et al. 2001). The use-dependence of the NO effect certainly has physiological relevance. Indeed it is expected that the eventual effect of an elevated intracellular NO level will depend on the pattern of muscle activity.

Although disparate, a large majority of the previously reported effects of NO on the ryanodine receptor channel activity have been either suggested or proved to occur through S-nitrosylation or oxidation of sulfhydryl groups within the channel protein. Both of these types of effects can be reversed by the reducing agent DTT (see for example Suko et al. 1999; Hart & Dulhunty, 2000) and we also found that the rise in resting [Ca²⁺] observed in the presence of NO donors could be reversed by this agent favouring a similar mechanism under the present conditions. This further stresses the potential importance of the intracellular redox environment in determining the physiological effects of NO on intracellular [Ca²⁺] homeostasis (Eu et al. 2000, 2003).

Rise in resting [Ca²⁺] and SR Ca²⁺ release

The use-dependent resting [Ca²⁺] elevation was the most immediate manifest expression of an altered intracellular Ca²⁺ homeostasis in the presence of excess NO. We believe that the accompanying depression of the peak Δ[Ca²⁺] occurs as a consequence of the underlying elevated SR Ca²⁺ leak, most likely through a direct Ca²⁺-dependent feedback mechanism. It indeed seems hard to conceive that direct effects of NO on the release channel protein could on one hand increase the channel activity at rest, while on the other hand depress the activity triggered by membrane depolarization. This would require a complicated model with for instance different populations of channels with opposite sensitivity to a given level of NO, which would be far too speculative. Rather our results are consistent with a Ca²⁺-dependent inactivation process. Several lines of studies provided evidence for the operation of Ca²⁺-dependent inactivation of Ca²⁺ release by showing that one component of the voltage-activated SR Ca²⁺ flux could be suppressed by a preceding rise in Ca²⁺, in frog (Simon et al. 1991; Jong et al. 1995) and in mammalian muscle fibres (Delbono, 1995; Garcia & Schneider, 1995). It is interesting that in mammalian fibres half-maximum inactivation was reached for an average myoplasmic [Ca²⁺] of 0.22–0.25 μm close to the range within which the peak Δ[Ca²⁺] was depressed here. Of course the bulk resting [Ca²⁺] that is measured largely underestimates the actual local [Ca²⁺] within the triadic junction that is expected to affect the opening probability of channels during a pulse.

The fact that, under our conditions, the apparent [Ca²⁺] dependence of the depression differed for a pulse to −20 and to +10 mV remains unexplained. Even more surprising is that Szentesi et al. (2000) showed that in rat muscle fibres, an identical voltage pre-pulse produced a larger attenuation of the Ca²⁺ release elicited by a small test pulse than by a large one. This is qualitatively opposite to what was found here where, for the same resting [Ca²⁺] elevation, the peak Δ[Ca²⁺] for a pulse to −20 mV was apparently less sensitive to depression than the one for a pulse to +10 mV. Reasons for this discrepancy probably lie within the different experimental conditions in the two studies, including the transient versus chronic type of resting [Ca²⁺] elevation. The difference that we observe between −20 and +10 mV under excess NO is also not related to an alteration of the properties of the voltage-sensing steps of excitation–contraction coupling, because there was no differential evolution of intramembrane charge movement at these two potentials in the presence of SNP. Of interest, we previously observed a qualitatively similar phenomenon under other conditions of raised resting [Ca²⁺] due to increased SR Ca²⁺ leak: we previously showed that ryanodine also produces a use-dependent rise in resting [Ca²⁺] (Collet & Jacquemond, 2002). Figure 10 of Collet & Jacquemond (2002) shows that, in the presence of ryanodine, the amplitude of the [Ca²⁺] transients elicited by pulses to −30 mV tended to increase with time whereas when pulses to −10 mV were given, the amplitude of the transients were progressively depressed. So this is another condition that favours resting [Ca²⁺] elevation through SR Ca²⁺ leak and that is associated with (depending on the conditions) peak [Ca²⁺] transient being depressed, or remaining constant/becoming elevated. Although still unclear, the underlying mechanism certainly has to do with the resting SR Ca²⁺ leak and (within the present context) not with an effect of NO on release channels.

Aside from Ca²⁺-dependent inactivation, it should be stressed that a use-dependent SR Ca²⁺ leak occurring under prolonged periods of activity should also produce SR Ca²⁺ depletion. Ultimately, this should be expected to contribute to further depress voltage-activated Ca²⁺ release. Along this line one may even suggest that in the presence of NO, due to the SR Ca²⁺ leak, Ca²⁺ would be diffusing away along the unclamped region of the fibre and that this would lead to SR Ca²⁺ depletion in the studied portion of fibre. However we believe that it is not the case, or at least that it is not the most significant mechanism. This is because the reversibility of the NO effects by DTT is extremely rapid, including the recovery of the peak [Ca²⁺] transient. If the decrease of the peak [Ca²⁺] was predominantly due to diffusion of Ca²⁺ away from the site of measurement, one would expect some time needed to replenish the SR once the NO-induced extra-SR Ca²⁺ leak is blocked. Instead, from the last pulse in the presence of NO to the first pulse given in the presence of DTT (Fig. 8D and E) there was an instantaneous drop in resting [Ca²⁺] together with some recovery of the peak [Ca²⁺] transient. So clearly some of the inhibition produced in the presence of NO reversed as resting [Ca²⁺] decreased. On average it took approximately 2 min to exchange the NO donor-containing solution with the DTT-containing one, which should be insufficient for Ca²⁺ trapped in the SR of the unclamped region of the fibre to be back into the SR within the studied portion. This strongly argues against SR Ca²⁺ depletion mediating the overall depression of release observed in the presence of NO.

From a more general point of view, our results indicate that it can be inappropriate to attempt to correlate an increased or decreased activity of ryanodine receptors in fragmented SR preparations under a given condition, to a respective enhancement or depression of voltage-activated SR Ca²⁺ release as a physiological effect in an intact preparation under that same condition. In the present case, whereas increased opening of the release channels appears as the predominant effect of excess NO, the end result is a depression of voltage-activated SR Ca²⁺ release.

The modifications of intracellular Ca²⁺ handling by NO donors described here may be representative of what is happening under conditions of endogenous overproduction of NO. In vitro, millimolar concentrations of SNP (as used here) were reported to produce NO levels within hundreds of nanomolar to the micromolar range (Zou & Cowley, 1997; Alonso-Galicia et al. 1998). This is in the upper range of what is presumed to be the operating intracellular concentrations (nm to μm) of NO (Stamler & Meissner, 2001). In normal muscle, endogenous NO production increases with muscle activity (see Stamler & Meissner, 2001). Raised intracellular levels of NO are thus expected under intense conditions of activity and also in several types of pathological conditions, including various inflammatory processes where the inducible NO synthase isoform (iNOs) expression is stimulated (see Kaminski & Andrade, 2001) and the autosomal dominant limb-girdle muscular dystrophy where the neuronal NO synthase (nNOs) activity is enhanced (Sunada et al. 2001). An activity-dependent increase in resting [Ca²⁺] may lead to the alteration of muscle function under such conditions.