Sodium nitroprusside, a NO donor, modifies Ca2+ transport and mechanical properties in frog skeletal muscle

Journal of Muscle Research and Cell Motility 19, 865±876 (1998)

Sodium nitroprusside, a NO donor, modi®es Ca2+ transport and mechanical properties in frog skeletal muscle S. BELIA1, T. PIETRANGELO3, S. FULLE1, G. MENCHETTI1, E. CECCHINI1, Á 3* M . F E L A C O 2 , J . V E C C H I E T 4 and G . F A N O 1

Dipartimento di Biologia Cellulare e Molecolare-Sezione di Fisiologia e Bio®sica, UniversitaÁ di Perugia, Italy Istituto di Biologia e Genetica, 3Dipartimento di Scienze Biomediche-Laboratorio Fisiologia Celluare and 4Dipartimento di Medicina e Scienze dell' Invecchiamento, UniversitaÁ ``G. D' Annunzio'', Chieti, Italy

2

Received 23 September 1997; accepted in revised form 20 August 1998

Summary Recently it has been hypothesized that, in skeletal muscle, NO produced directly by high-frequency stimulation could produce contraction through reactions with thiol groups on the sarcoplasmic reticulum (SR). However, a possible cGMP-mediated relaxing effect, similar to that seen in smooth muscle, has also been demonstrated. We used puri®ed SR preparations and single ®bres from frog fast muscles incubated with different concentrations of sodium nitroprusside (SNP) in this study. The results obtained from a long low-frequency stimulation, together with those from a study on Ca2+ transport regulation, showed that the presence of NO precursor induced: an acceleration of the onset of fatigue in single ®bres; a decreased vesicular Ca2+ content due to increased Ca2+ release; a shift to open status in SR Ca2+ channels; an increase in SR Ca2+ pump activity. The data presented in this paper seem to indicate that the increased NO in the muscle ®bres can in¯uence muscle activity in different ways, perhaps depending on the metabolic status of the muscle and target (®laments, sarcolemma, SR) with which the NO (or its derivatives) acts. Ó Kluwer Academic Publishers.

Introduction During the conversion of arginine to citrulline in endothelial cells, a volatile oxygen-reactive intermediate, nitric oxide (NO), is formed by the action of nitric oxide synthase (NOs) in the presence of calmodulin and Ca2+ (Bredt & Snyder, 1990). Since NO relaxes the smooth muscle ®bres present in the vessel walls, it was originally named endothelium-derived relaxing factor (EDRF) (Palmer et al., 1987). In recent years, there has been great interest in the function of NO as a cellular messenger in different tissues, particularly the nervous system. Using the vast amount of available data, the probable mechanism by which NO acts in the cell can now be described. Under most conditions, the enzyme that synthesizes NO (NOs) (Murphy et al., 1993; Forstermann, 1994) can be activated, in the presence *To whom correspondence should be addressed at: Dipartimento di Scienze Biomediche-Laboratorio Fisiologia Cellulare, via dei Vestini 29, 66013 Chieti, Italy. E-mail: [email protected]

0142±4319/98 Ó Kluwer Academic Publishers

of calmodulin and an adequate Ca2+ concentration, to produce NO which, in turn, interacts (at least in part) with the heme group of guanylate cyclase to form cGMP (Murad et al., 1990; Schmidt et al., 1993). Thus, this cyclic nucleotide can be considered as the true second messenger in many NO-regulated processes. A recent study on the relationship between NO production and skeletal muscle activity has produced some interesting results: the presence of constitutive NOs (cNOs) was demonstrated in human tracheal muscle by immunohistochemical techniques. This preparation was also used to clone the enzyme gene (Nakane et al., 1983) and to determine NO release in muscle preparations of rat limbs (Balon & Nadler, 1994). Moreover, it has been clearly demonstrated that type II ®bres (fast) show a cNOs immunostaining. It is possible to stimulate the NOs activity with consequent NO production in muscles with a high percentage of these ®bres (Kobzik et al., 1994; Grozdanovic et al., 1995). Recently the presence of endothelial cNOs (ecNOs) type was also reported in skeletal muscle regardless of ®bre type (Kobzik et al., 1995).

866 Balon and Nadler (1994) were the ®rst to report the possible NO effect in skeletal muscle activity. They noted that a high-frequency stimulation (100 Hz) induced the release of NO into the bath solution, while the use of L-NMMA (NG ±mono-methyl-L-arginine), a speci®c NOs-inhibitor, gave the opposite result. Recently, Kobzik and colleagues (1995) demonstrated a possible cGMP-mediated relaxing effect of NO in skeletal muscle, similar to that seen in smooth muscle. In the same article, the authors hypothesized a contraction-stimulating effect of NO (acting as a reactive oxygen intermediate) through reactions with thiol groups on the sarcoplasmic reticulum (SR). This positive action on force generation was recently shown in both slow- and fast-twitch mammalian skeletal muscles (Murrant et al., 1994; Murrant & Barclay, 1995) and was also con®rmed by the effects induced by other reactive oxygen intermediates which, in unfatigued skeletal muscle, had a positive action on excitation±contraction (E±C) coupling. However, when antioxidant agents such as catalase and/or superoxide dismutase were used, an opposite effect was observed in mammalian bundles (Reid et al., 1993) and single frog ®bres (Regnier et al., 1992). The aim of this work was to verify and extend these observations by using puri®ed SR preparations and single ®bres derived from frog fast muscles (tibialis anterior, gastrocnemius, extensor digitorum longus and sartorius). In particular, the mechanical responses to a low frequency long-time stimulation were determined and a study of the regulation of Ca2+ transport which may be involved in the functional regulation of SR activity was carried out. To mimic the effect of NO increase we used the NO donor sodium nitroprusside (SNP) which is able to produce NO with a constant rate (Maragos et al., 1993; Feelish & Noack, 1987; Balon & Nadler, 1994). Materials and methods NADPH diaphorase activity Muscle ®bres derived from gastrocnemius and sartorius muscles were ®xed for 30 min in 4% paraformaldehyde, buffered with phosphate buffered saline (PBS) 0.1 M , pH 7.4 at 4°C. The samples were cut into 10 lm longitudinal and transverse sections in a cryostat (Cryocut 1800, Reichert-Jung, Germany) thawed on alum/gelatin pretreated glass slides. To demonstrate the nicotinamide adenine dinucleotide (NADH) diaphorase activity according to Scherer-Singler and coworkers (1983), these sections were immediately covered completely at room temperature for 30±60 min with the following incubation medium containing 50 mM Tris±chloride, 500 mM Nitro blue tetrazolium (NBT), 7 mM bNADPH and 0.2% Triton X-100 at pH 8.00.

BELIA et al. The medium was then poured off and the sections rinsed in tap water, then in distilled water and ®nally mounted with glycerol gel. Controls were performed with media free of substrate.

Mechanical procedures Single ®bres were dissected from the lateral head of the frog tibialis anterior muscle. Experiments were carried out at 16°C and 2.2 lm sarcomere spacing. Stimuli of alternating polarity were applied transversely to the muscle ®bre by a pair of platinum wires (0.5 mm diameter). Pulses of 0.5 ms duration and twice the threshold strength were used. Tetanic stimulation was applied in a brief volley of 0.35 s, at 4 min intervals (Lombardi & Menchetti, 1984). The optimal stimulation frequency was determined for each ®bre and ranged from 86±100 Hz. Tension was measured by means of a capacitance gauge transducer (resonance frequency of 30 kHz and 2 mV mg)1 sensitivity). Fibres were continuously perfused in dark conditions with either normal Ringer (NR) solution containing 115 mM NaCl, 2.5 mM KCl, 1.8 mM CaCl2, 3.0 mM phosphate buffer (pH 7.0), or with an experimental solution (NR + 1.0 mM SNP). Fibre fatigue was induced by low frequency (2 Hz) stimulation. Data from the tension transducer were collected by an A/D converter (R.C. Electronic) and stored on a PC. Measurements were done by speci®c software (R.C. Electronic). At the end of the experiment the ®bre was perfused overnight with NR solution and then ®bre viability was checked by evoking a series of twitches and tetani. In each experiment we stimulated ®bres in NR to obtain a constant twitch/tetanus amplitude ratio. From this point, in the control experiments, we began the procedure to induce fatigue. In the experiments in NR with 1 mM SNP, after the rapid substitution of the solution, we continued the procedure to obtain constant levels of twitch and tetanic tension and then induced fatigue. As far as the rate at which SNP produces NO is concerned, we have not measured it directly, but this value can be derived by using the method described by Feelish and Noack (1987); under our experimental conditions (16°C, pH 7.0), 1 mM SNP produces about 0.1 lmol min)1 of NO.

Vesicle preparations Puri®ed vesicles, derived from the fragmentation of terminal cisternae of skeletal muscles dissected from posterior frog legs, were prepared using a previously reported method (FanoÁ et al., 1989). Skeletal muscles were triturated in liquid N2 and homogenized in 250 ml homogenization medium (HM) at 4°C. The nuclei, contractile proteins and mitochondria were then removed by centrifuging several times. The remaining microsomes were resuspended in 5 ml HM with a Dounce homogenizer and layered on to a discontinuous sucrose gradient (1.6, 1.3, 1.1 and 0.8 M ). The gradient was centrifuged at 70 000 ´ g for 16 hours and the fraction at the 1.6±1.3 M interface step was collected, washed twice in the presence of 0.2 mM EGTA, and, after protein determinations, immediately used for the tests.

NO donor and skeletal muscle activity Ca

2+

transport

Active (ATP-dependent) uptake and passive release of Ca2+ were determined by a previously reported method (Marsili et al., 1992) which uses 45Ca2+ as a radiolabelled tracer, and a rapid ®ltration apparatus to determine the Ca2+ ¯uxes across the vesicular membranes. 45 Ca2+ uptake was measured according to Fitts and coworkers (1980) using 1 ml of Briggs solution (Briggs et al., 1977) containing 20 mM imidazole, 0.2 mM EGTA, 20 mM K-oxalate, 5 mM Na-ATP, 10 mM MgCl2, 0.18 mM CaCl2 and 100 mM KCl (pH 7.4) with the addition of 5 lM 45Ca2+ (speci®c activity 5±50 mCi mg)1) with or without test substances (1 mM SNP or 0.1 mM db-cGMP). The uptake started with the addition of 250 lg proteic substrate. At preselected times (0.5, 2, 6 min), calcium ¯ux was stopped by rapid ®ltration on Millipore ®lters (0.45 lm) placed in a Bio-Rad ®ltration apparatus. Then 0.5 ml of the ®ltered solution was collected and the radioactivity read on a bcounter (Beckman LS 800). Ca2+ uptake was calculated by taking the difference between the radioactivity measured initially and the value recorded at the time considered. The experiments for the determination of Ca2+ release were carried out on vesicles preloaded with 5 lM 45Ca2+ as previously described, but in absence of SNP or db-cGMP. After 6 min of this treatment, the Briggs solution was replaced with 1 ml of the same solution containing 5 mM adenine instead of 5 mM Na-ATP, with or without test substances (1 mM SNP or 0.1 mM db-cGMP). The calcium release was measured, on a ®ltered solution, by taking the difference between the value measured after 6 min incubation and those recorded after 2 (db-cGMP) or 4 (SNP) minutes from the addition of the relaxing solution. Ca2+/ATPase activity was measured on 20±50 lg of proteins by a previously reported method (FanoÁ et al., 1989) in a ®nal volume of 1 ml of incubation medium containing 5 mM ATP, 10 lM CaCl2 and 1 mM SNP.

Binding experiments Sarcoplasmic reticulum vesicles (R4) (McGrew et al., 1989) were washed in a binding buffer (1 M KCl, 10 mM Hepes, 25 lM CaCl2, pH 7.4) centrifuged at 110 000 ´ g for 90 min and resuspended at a ®nal concentration of 1 mg ml)1. Aliquots of 0.2 ml, pretreated for 5 min in the presence or absence of 1 mM SNP, were incubated at 25°C with 10 nM 3 H-ryanodine for different times (5, 10, 20, 30, 60, 120 min) for total binding. At each incubation time, the samples were ®ltered rapidly (Bio-Rad ®ltration apparatus), washed twice with 1 ml cold buffer, and the bound radioactivity in the ®lter assessed by liquid scintillation counting. In other experiments, sample aliquots of 0.2 ml were incubated with different 3H-ryanodine concentrations (3±100 nM ) for 30 min with or without 1.0 mM SNP (preincubated for 5 min). The non-speci®c binding was measured in the presence of a 1000-fold excess of cold ryanodine. The binding buffer solution to test the Ca2+ effect on ryanodine binding contained 1 M KCl, 10 mM Hepes and 1 mM EGTA and different Ca2+ concentrations (0.03, 0.1, 0.25, 0.5, 1.1 mM ) in the presence or absence of 1 mM SNP (preincubated for 5 min). Under this condition the effective free Ca2+ ranged from 0.02 to 0.5 lM in the considered

867 experimental interval (Schatzmann, 1973). Non-speci®c binding was performed with the addition of a 1000-fold excess of cold ryanodine (FanoÁ et al., 1992).

Guanylate cyclase activity The assay was performed using the Garbers and Murad method (1979) on 500 lg protein derived from puri®ed vesicles in a solution containing 50 mM Tris±HCl, 5 mM theophylline, 5 mM dithiothreitol (DTT), 0.01 mM CaCl2, 1 mM guanosine 5¢-triphosphate (GTP) (pH 7.4) in the presence or absence of 1 mM SNP. After incubation (10 min at 37°C) the reaction was stopped by 25% trichloracetic acid (TCA)(v/v), the samples were centrifuged (12 000 ´ g, 10 min) and the supernatant was extracted four times with diethylether to remove the TCA. After overnight evaporation of residual ether, 100 ll of the supernatant were used to measure the cGMP formed by a commercial radioimmunoassay kit (Amersham).

Video-imaging The myogenic cell line C2C12 (Lorenzon et al., 1997) was used to measure the [Ca2+]i variations on several cells (n ˆ 12). At the beginning of each experiment, the cells were washed with the normal external solution (NES), a buffered solution containing 140 mM NaCl, 1.8 mM CaCl2, 2.8 mM KCl, 2 mM MgCl2, 10 mM glucose and 10 mM HEPES/NaOH (pH 7.4). After washing, the cells were loaded by incubation at 37°C for 30 min with the ¯uorescent Ca2+ dye Fura-2 pentacetoxymethylester (Fura-2 AM, 5 lM ; Sigma) dissolved in NES supplemented with 10 mg ml)1 bovine serum albumin (BSA). After incubation and rinsing with NES, the cells were maintained for an additional 15 min at 37°C to allow the de-esteri®cation of the dye. The glass coverslip with the cells was placed in the microincubator chamber of a digital ¯uorescence-imaging microscopy system equipped with a 63´ oil objective lens (Leika DM-IRB, Germany). 1 mM SNP was added to the cells bathed in NES. Excitation light at 340 or 380 nm was provided by a high-speed ®lter exchanger (Till Photonics, USA) equipped with a 75 W xenon lamp. A cooled CCD camera (Hamamatsu Photonics, Hamamatsu, Japan) captured the ¯uorescence. The ratio images were calculated offline and carried out pixel by pixel in pairs of corresponding 340 and 380 nm ®les. Each temporal analysis was obtained from an area of interest in the cell and represents the mean value variation of the ¯uorescence intensity of ratio images.

Statistical analysis All the data were analysed, if not otherwise indicated, by a computerized program (Prism 2, GraphPad, USA) utilizing the t test and analysis of variance to verify the statistical differences between the experimental series.

Results NOs localization Detailed staining analysis for NADPH diaphorase clearly showed that high formazan concentrations were present in both longitudinal and transverse

868

BELIA et al.

Fig. 1. NADPH diaphorase activity measured on cryosections derived from sartorius (At) or gastrocnemius muscle (Bt). The formazan precipitate (dark areas) indicates the NOS activity (magni®cation 400´). Controls (Ac, Bc) were performed in the presence of NOs-substrate free media (magni®cation 250´). For more details, see Materials and methods.

sections. The dark contours (Fig. 1) indicate that enzymatic activity was present overall in the cell membranes of both sartorius (Fig. 1A) and gastrocnemius (Fig. 1B) muscle, with only a slight quantitative difference between the two muscle types. The data derived from these observations con®rm that NOs activity was present in our preparations. Ca2+ transport The NO action on the mechanism of Ca2+ status control in fast ®bres was tested using puri®ed vesicles obtained from SR fragmentation. In these substrates some qualitative and quantitative aspects of Ca2+ transport were modi®ed by the presence of the NO donor SNP. Figure 2 reports the data derived from the experiments in which vesicles were loaded with 5 lM 45 Ca2+ in the presence of 5 mM ATP and 0.18 mM Ca2+. Under these conditions (see Methods for more details), loaded Ca2+ reached a plateau after 4±6 min incubation. As can be seen from the graph, the presence of 1 mM SNP modi®ed the time-course of calcium uptake by signi®cantly decreasing vesicular Ca2+ content. This quantitative difference began to be evident after two minutes incubation and reached the maximum (63%) after six minutes at the end of the experimental time. Higher SNP concentrations, up to 10 mM (data not shown), did not modify this situation. The decreased intravesicular Ca2+ content could be due to a reduced vesicle capacity to accumulate Ca2+. This possibility was tested by measuring the

calcium ¯uxes from vesicles actively loaded for six min as previously described. Under these conditions, the presence of 1.0 mM SNP signi®cantly increased calcium release from preloaded vesicles (from 8.8 ‹ 0.4 to 15.6 ‹ 3.2 nmol Ca2+ min)1). Since the effective vesicular calcium content was from the balance between the calcium released via Ca2+ channels and the Ca2+ uptake by the Ca2+ pump, we also tested the possibility that NO could modify the status of Ca2+ channels present in these preparations by using 3H-ryanodine, an alkaloid able to interact with the open form of these channels. As reported in Fig. 3a, in the presence of 1 M KCl (a prerequisite to obtain binding of ryanodine on Ca2+ channels present in the vesicles) the alkaloid binds its receptors in less than 40 min saturation time. Under these conditions, the non-speci®c binding is very low (8±11% of the total binding) and the presence of SNP is able to increase the speci®c binding for all the tested times (10±120 min). The effect induced by the NO donor also depends on the concentration of ryanodine used. As reported in Fig. 3b, in a concentration range of 3±20 nM of 3H-ryanodine, 1.0 mM SNP shifts the plot to the left. This effect seems at least partially related to the [Ca2+]out because, in the presence of 10 nM 3 H-ryanodine, it was more effective when the external free calcium concentration was less than 0.25 mM (Fig. 4). The non-linear regression, carried out on data described in Fig. 3b, suggests that the increase of bound ryanodine was caused by an increase in site af®nity rather than in site number. In fact, the Bmax of the two

NO donor and skeletal muscle activity

869

Fig. 2. NO effect on Ca2+ transport in vesicles derived from frog sarcoplasmic reticulum of skeletal muscle. ATPdependent Ca2+ uptake was determined in the presence or absence of 1 mM SNP using 5 lM 45Ca2+ as a marker of Ca2+ movement. Ca2+ release was determined, on vesicles previously incubated (6 min) in the absence of SNP, after 4 min of incubation in the presence of 5 mM adenine with or without 1 mM SNP (see Methods for more details). The data reported are the means ‹ S D of ®ve (release) or eight (uptake) measurements. The statistical differences were tested by unpaired t tests for each experimental point (*p < 0.05, **p < 0.01) and by two-way ANOVA to determine the statistical signi®cance of the treatment (p < 0.0001).

series were almost identical (control ˆ 0.36 ‹ 0.032 nM ; treated ˆ 0.37 ‹ 0.036 nM ) while the Kd differed signi®cantly (control ˆ 11.25 ‹ 3.18 nM ; treated ˆ 4.26 ‹ 1.76 nM ). In the presence of SNP, the ATPase activity associated with Ca2+ transport did not participate in the decrease of Ca2+ uptake. Under these conditions, the activity of the Ca2+ pump increased in a dose-dependent fashion when 0.01± 1.0 mM SNP was present in the medium (Fig. 5a). In particular, SNP increased the enzymatic response starting from a concentration of 0.01 mM (+100% of activation). A possible NO target in the cell could be the heme groups, and in particular the heme contained in the soluble guanylate cyclase, the enzyme which synthesizes cGMP. This could also occur in the vesicles. In fact, the presence of 1.0 mM SNP induced a signi®cant increase in guanylate cyclase (GC) activity (Fig. 5b), with a consequent increase of cGMP in the medium. As reported in Fig. 6, db-cGMP mimicked NOdonor effects in both the release and uptake of 45Ca2+, stimulating the ®rst (+30%) and signi®cantly inhibiting the second. To see whether SNP causes a Ca2+ increase in the myoplasm, we also performed a series of experiments

using Fura 2/AM as a ¯uorescent indicator of Ca2+ variations. However, this method is not compatible with the use of muscle ®bres as a substrate and, for this reason, we decided to use a myogenic cell line (C2C12) to test this hypothesis. As can be seen in Fig. 7, the presence in the medium of 1 mM SNP induced a rapid and reversible effect in [Ca2+]i which increased signi®cantly. Mechanical activity The time course of fatigue, induced by low-frequency stimulation (2 Hz), was measured on three ®bres superinfused with NR containing 1 mM SNP and compared with three ®bres superinfused with only NR. Each curve was obtained by measuring the amplitude of 40 twitches, and each point represents the ratio between the peak tension of the twitch and the ®rst twitch of the series (Fig. 8). It can be seen that the time course of the treated ®bres curve runs below that of the controls after 20±30 s from the start of stimulation, indicating that, in the presence of SNP, the onset of fatigue was accelerated. During this activity, the relaxation rate of the twitch, which is considered peculiar to this mechanical

870

BELIA et al.

Fig. 3. Increase of ryanodine-speci®c binding induced by 1 mM SNP on SR vesicles. (a) Time-dependent effect (10±120 min) of 1 mM SNP on speci®c ryanodine binding (n ˆ 7). The speci®c binding was calculated by the difference between the total binding measured in the presence of 10 nM 3H-ryanodine alone, and the non-speci®c binding carried out in the presence of a 1000-fold excess of cold ryanodine. The degree of non-speci®c binding was not time-related (8.3 ‹ 4.1% of total binding in the control series; 10.4 ‹ 3.8% in treated samples). (b) Dependence of ryanodine concentration (3±100 nM ) on ryanodine-speci®c binding with or without 1 mM SNP (n ˆ 5). The non-speci®c binding, assumed to be linear with respect to ryanodine concentration, was determined with 1000-fold excess cold ryanodine and subtracted from total bound 3H-ryanodine. Kd and Bmax values were automatically derived by the Prism-2 computer program. Two-way analysis of variance showed a signi®cant difference between the treatments (p < 0.001).

activity (Westerblad & Allen, 1993) was signi®cantly decreased both in NR and NR with 1 mM SNP conditions. However, we cannot observe signi®cant differences between controls and treated ®bres. Discussion The possibility that NO or its derivatives act upon the physiological processes of muscle contraction is suggested by at least three features: (1) the presence of one or two NO synthase types in the sarcolemma of skeletal muscle ®bres; (2) the increase of NO production after electrical stimulation; (3) the modi®cation of several characteristics of muscular activity with changes in intracellular NO concentration.

It was dif®cult to establish the identity of the active NO species inducing the described effects, because this compound can exist in three redox forms able to react with different chemical groups depending on physico-chemical conditions (FanoÁ et al., 1997). However, the aim of this work was to mimic the physiological conditions by an increase in available NO, as occurs during a long stimulation. Under this condition as described by Balon and Nadler (1994), the presence of SNP induced the same NO increase as that derived from arginine stimulation of muscle NOs. From the results of these experiments, it is evident that NO either directly or indirectly induces an increase in the Ca2+ available, which increases Ca2+ release from the SR by shifting the Ca2+ channel status to open. This fact is evident from data concerning the ryanodine bound to the vesicles derived from

NO donor and skeletal muscle activity

Fig. 4. Speci®c ryanodine binding induced by 1 mM SNP in the presence of different external Ca2+ concentrations. The binding (n ˆ 8) was performed after 30 min of incubation with 10 nM 3H-ryanodine in the presence or absence of 1.0 mM SNP in a solution containing different amounts of external Ca2+ (0.03±1.1 mM ). The non-speci®c binding was measured with 1000-fold excess of cold ryanodine (see Fig. 3 legend for more details). The statistical signi®cance, tested by unpaired t test with Welch's correction, showed p ˆ 0.016.

terminal cisternae, which demonstrate that, in the presence of NO donors such as SNP, the Kd value of ryanodine binding changes from 11.25 nM to 4.26 nM without any modi®cation of Bmax. Since at nanomolar concentrations the ryanodine binding is directly related to the open Ca2+ channel status, the greater af®nity with increased NO concentration, derived from release of SNP (especially when the external Ca2+ is low), indicates that the open form of the Ca2+ channels is mainly present in these preparations (Feelish & Noack, 1987; Gopalakrishna et al., 1993). A

871 direct positive effect on ryanodine channels has been more recently demonstrated both in skeletal and cardiac mammalian muscle (Stoyanovsky et al., 1997). As a consequence, a great increase in Ca2+ release was determined, and not even the stimulated activity of the Ca2+ pump was able to reverse the Ca2+ ¯ux through the vesicle membranes, and Ca2+ remained accumulated outside. This mechanism seems probable because it has been shown that other free radicals interact with thiol groups of the SR, increasing Ca2+ release as a consequence of the increase in the number of open Ca2+ channels (Stuart et al., 1992; Xiong et al., 1992). Considering also the data derived from C2C12 experiments (even if these cells are mammalian myoblasts), the possibility that NO could act directly or indirectly by stimulating Ca2+ release becomes more probable. It is not always possible to assume that when intracellular free Ca2+ increases, the force generated during the contraction is also higher. This fact is not automatic and should be more easily detectable during a short-term cycle of muscle activity (e.g. a single twitch or brief series of twitches). On the other hand, if the muscle (or its ®bres) undergoes an exhaustive fatigue procedure, then the progressive depletion of internal Ca2+ stores may occur and have a negative impact on contraction. Both in striated and smooth muscle, the increased presence of NO (or its derivatives) induces directly, or causes via cGMP, an increase in the calcium pump activity (Mohan et al., 1996), but only in the ®rst type of muscle do ryanodine-sensitive calcium channels play a fundamental role by producing a massive Ca2+ release from the SR. In contrast, in the smooth muscle the increase in intracellular Ca2+ concentration originates essentially from the activation of voltage-activated membrane channels. In arterial smooth muscle the activation of ryanodine channels generates spontaneous bursts of Ca2+ release, which are able to depress muscle force by activating Ca2+-sensitive K+ channels, resulting in hyperpolarization and vasodilatation (Nelson et al., 1995). Our efforts concerning these aspects are only partially conclusive. In fact, data derived from these experiments show that taking ®bres through an exhaustive fatigue procedure leads to a substantial decrease in their capacity to develop tension in the presence of higher NO concentrations. However, the single ®bres that were stimulated to obtain single twitches or complete tetani showed no signi®cant differences with respect to the controls in the rate of development of tension or in the maintenance of the tension levels and the relaxation rate. These data seem to indicate that, in opposing smooth muscle (Karaki et al., 1997), no modi®cation of Ca2+ ®lament sensitivity occurs when NO increases in the experimental solution.

872

BELIA et al.

Fig. 5. Enzymatic activity (Ca2+-ATPase and guanylate cyclase) modi®cations induced by SNP. (a) The Ca2+-ATPase assay was carried out in the presence of 150±200 lg proteins at pH 7.4. The released orthophosphate was estimated on 1 ml of clear supernatant and the speci®c activity was expressed as lg Pi/min/mg. The statistical signi®cance of the data (mean ‹ S D ; n ˆ 5), tested by unpaired t test, showed a signi®cant difference between treated and untreated samples (**p < 0.01). (b) The guanylate cyclase activity was measured as cGMP formed (pmol/min/mg) from 500 lg protein of puri®ed vesicles in the presence or absence of 1 mM SNP. The data (mean ‹ S D ; n ˆ 5) showed a signi®cant difference (**p < 0.001) between SNP samples and controls.

Recently, Kobzic and colleagues (1994) proposed that the relative decrease in force generated in the presence of NO could be obtained by stimulating muscle relaxation induced by cGMP, the synthesis of which is stimulated by NO. This cyclic nucleotide may induce muscle relaxation in a way similar to that in smooth muscles. Our data in part con¯ict with this hypothesis, and even if the values seem to indicate that cGMP increases the availability of Ca2+, no modi®cation of mechanical parameters was observed with increased NO concentration during a long-term, low-frequency (2 Hz) stimulation. During a maximal tetanus, only two of the nine ®bres tested showed a partial decrease in tension (data not shown); therefore, the inhibiting effect related to cGMP increase is not evident. However, it is also possible that, during this activity, the duration of the electrical stimulation was too short, so that the cGMP stimulation by NO was not

evident. In fact, according to Balon and Nadler (1994), it is possible to observe a 40% increase in the NO concentration in medium containing skeletal muscles only after a long period (1 h) of electrical stimulation. In a previous study, we found that electrical stimulation during an isometric tetanus in frog sartorius muscle induced an immediate increase in the myoplasmic cGMP concentration, which was directly related to the developed tension (FanoÁ et al., 1982). In the same muscles, db-cGMP induced Ca2+ release from the SR and had an opposite effect on substances such as dantrolene which are known to inhibit Ca2+ release from SR (Mancinelli et al., 1983). Therefore it seems improbable, at least in frog skeletal muscle, that cGMP could act as a muscle relaxing agent. In summary, it appears that NO (at least in its reduced state) could act via guanylate cyclase activation and [cGMP]i increase as a stimulating agent for Ca2+ release from SR, probably via an activation of the

NO donor and skeletal muscle activity

873

Fig. 6. Effect of db-cGMP on Ca2+ uptake and release in vesicles derived from frog SR of skeletal muscle. The data reported are mean ‹ S D of ®ve (release) or eight (uptake) measurements. Ca2+ uptake was determined in the presence or absence of 0.1 mM db-cGMP in Briggs solution containing 5 mM ATP. Ca2+ release was determined after 2 min incubation in the presence or absence of 0.1 mM db-cGMP on vesicles preincubated for 4 min in Briggs solution alone. The statistical differences were tested by unpaired t tests for each point (**p < 0.01) and by two-way ANOVA to determine the statistical signi®cance of the treatment (p < 0.001).

Fig. 7. Single-cell temporal analysis of [Ca2+]i increase induced by a single pulse of 1 mM SNP (solid line). After a 5 min wash with NES (W, arrow) the resting level was recovered. The two transients represent the mean of corresponding areas (the blue trace for the blue area and the red trace for the red area, in the graph and in the panel, respectively) carried out pixel by pixel on a pair of corresponding 340 and 380 nm images. The ¯uorescent ratio values, collected as 1 ratio per 400 ms, indicate the [Ca2+]i according to the chromatic pseudocolour bar which indicates increasing values from blue to red. The bar indicates 30 s for temporal analysis and 20 lm for the cell dimensions. Arrows with the letters A and B indicate the time of images A and B.

874

BELIA et al.

Fig. 8. Twitch peak amplitude (mean ‹ S D ) of single muscle ®bres stimulated at 2 Hz plotted against time from the start of the stimulation (only one in every ten data points in each curve is shown). Peak tension values of twitches during fatigue are normalized to the constant level of twitch before the long-term stimulation. After the ®rst stimulation in NR, the treated samples were incubated with an experimental solution composed of NR plus 1 mM SNP. Controls were maintained and stimulated in NR. Controls (n ˆ 3) and treated (n ˆ 3) results are obtained from different ®bres. The two data-sets analysed by unpaired t test are statistically different (p < 0.0001).

Ca2+-induced Ca2+-release mechanism. Partial support for this hypothesis is also derived from the observations that agents that can increase NO concentrations in mammalian muscles increase the force developed during a short-term contraction (Murrant & Barclay, 1995). It is also possible that, after a long stimulation, the increases of NO and/or cGMP cause a partial or total depletion of Ca2+ stores, and consequently a negative effect on E±C coupling. Finally, it can also be hypothesized that the NO variations in the muscle ®bre can, during long-term activation, in¯uence muscle activity in relation to its redox status and ®bre metabolism. This does not mean that NO (or its derivatives) is a necessary step during the E±C cycle, but that in speci®c situations it

can regulate muscle response by modulating the open status of the Ca2+ channels. Therefore its role may be evident only under particular conditions, such as when the systems for buffering these free radicals are saturated. The data presented here form a working hypothesis that may stimulate greater interest in the role of NO in skeletal muscle activity.

Acknowledgements The authors thank Dr Douglas Root and Dr Scott Norton from the University of North Texas, Dr C. Franzini-Armstrong from the University of Pennsylvania, and Mrs C. Bennet from the UniversitaÁ di

NO donor and skeletal muscle activity Perugia (Istituto Patologia Generale, Italy) for their revision of this paper. This work was supported, in part, by a research grant from MURST to G. FanoÁ (40%±Contrazione muscolare). References BALON, T. W. & NADLER, J. L.

(1994) Nitric oxide release is present from incubated skeletal muscle preparation. Appl. Physiol. 77, 2519±21. B R E D T , D . S . & S N Y D E R , S . W . H . (1990) Isolation of nitric oxide synthase, a calmodulin-requiring enzyme. Proc. Natl Acad. Sci. USA 87, 682±5. B R I G G S , F . N ., P O L A N D , J . L . & S O L A R O , R . J . (1977) Relative capabilities of sarcoplasmic reticulum in fast and slow mammalian skeletal muscle. J. Physiol. 266, 587±94. Á , G ., B E L I A , S ., F U L L E , S ., A N G E L E L L A , P ., P A N A R A , FANO F ., M A R S I L I , V . & P A S C O L I N I , R . (1989) Functional aspects of calcium transport in sarcoplasmic reticulum vesicles derived from frog skeletal muscle treated with saponin. J. Muscle Res. Cell Motil. 10, 326±30. Á , G ., M A R S I L I , V ., P R O T A S I , F ., M A R I G G I O Á , M . A ., FANO F U L L E , S ., C E C C H I N I , E . & M E N C H E T T I , G . (1992) Relationship between an endogenous calcium binding protein (S-100ab) and Ca2+ transport in frog skeletal muscle. Basic and Applied Myology 2, 309±16. Á , G ., M E N C H E T T I , G . , D E L L A T O R R E , G ., V O L P I , L . , FANO S E C C A , T . & O R L A C C H I O , A . (1982) Cyclic adenosine monophosphate and cyclic guanosine monophosphate variations during isometric tetanus in frog sartorius muscle. Can. J. Physiol. Pharmacol. 60, 79±83. Á , G., TIJSKENS, P., COSCIA, F., CUCCURULLO, F. & FANO M E N C H E T T I , G . (1997) The intricate story of nitric oxide as regulator of skeletal muscle activity: a brief report. Current Topics in Pharmacology 3, 247±53. F E E L I S H , M . & N O A C K , E . A . (1987) Correlation between nitric oxide formation during degradation of organic nitrates and activation of guanylate cyclase. Eur. J. Pharmacol. 139, 19±30. FITTS, R. H., WINDER, W. W., BROOKE, M. H., KAISER, K . K . & H O L L O S Z Y , J . O . (1980) Contractile, biochemical and histochemical properties of thyrotoxic rat soleus muscle. Am. J. Physiol. 238, C15±C20. F O R S T E R M A N N , U . (1994) Biochemistry and molecular biology of nitric oxide synthases. Arzneim.-Forsch./Drug Res. 44, 402±7. G A R B E R S , D . L . & M U R A D , F . (1979) Guanylate cyclase assay methods. Adv. Cyclic Nucleotide Res. 10, 58±67. GOPALAKRISHNA, R., CHEN, Z. H. & GUNDIMEDA, U. (1993) Nitric oxide and nitric oxide-generating agents induce a reversible inactivation of protein kinase C activity and phorbol ester binding. J. Biol. Chem. 268, 27180±85. G R O Z D A N O V I C , Z . , N A K O S , G ., D A H R M A N N , G ., M A Y E R , B . & G O S S R A U , R . (1995) Species-independent expression of nitric oxide synthase in the sarcolemma region of visceral and somatic striated muscle ®bers. Cell Tissue Res. 281, 493±9. K A R A K I , H . , O Z A K I , H . , H O R I , M . , M I T S U I -S A I T O , M . , HARADA, K., MIYAMOTO, S., NAKAZAWA, H., WON, K . J . & S A T O , K . (1997) Calcium movements, distribu-

875 tion and function in smooth muscle. Pharmacol. Rev. 49, 157±230. KOBZIK, L., REID, M. B., BREDT, D. S. & STAMLER, J. S. (1994) Nitric oxide in skeletal muscle. Nature 372, 546±8. KOBZIK, L., STRINGER, B., BALLIGAND, J. L., REID, M. B. & S T A M L E R , J . S . (1995) Endothelial type nitric oxide synthase in skeletal muscle ®bers: mitochondrial relationship. Biochem. Biophys. Res. Commun. 211, 375±81. L O M B A R D I , V . & M E N C H E T T I , G . (1984) The maximum velocity of shortening during the early phases of the contraction in frog single muscle ®bres. J. Muscle Res. Cell Motil. 5, 503±13. LORENZON, P., GIOVANNELLI, A., RAGOZZINO, D., E U S E B I , F . & R U Z Z I E R , F . (1997) Spontaneous and repetitive calcium transient in C2C12 mouse myotubes during ``in vitro'' myogenesis. Europ. J. Neurosci. 9, 800± 809. Á , G., FERRONI, L., SECCA, T. & MANCINELLI, L., FANO D O L C I N I , B . M . (1983) Evidence for an inotropic positive action of cGMP during excitation±contraction coupling in frog sartorius muscle. Can. J. Physiol. Pharmacol. 61, 590±94. MARAGOS, C. M., WANG, J. M., HRABIE, J. A., O P P E N H E I M , J . J . & K E E F E R , K . (1993) Nitric oxide/ nucleophile complexes inhibit the in vitro proliferation of A375 melanoma cells via nitric oxide release. Cancer Res. 53, 564±8. MARSILI, V., MANCINELLI, L., MENCHETTI, G., FULLE, S. Á , G . (1992) S-100ab increases calcium release in & FANO

puri®ed sarcoplasmic reticulum vesicles of frog skeletal muscle. J. Muscle Res. Cell Motil. 13, 511±15.

MCGREW, S. G., WOLLEBEN, C., SIEGEL, P., INOUI, M. &

(1989) Positive cooperativity of ryanodine binding to the calcium release channel of sarcoplasmic reticulum from the heart and skeletal muscle. Biochemistry 28, 1686±91.

FLEISCHER, S.

MOHAN, P., BRUSAERT, D. L., PAULUS, W. J. & SYS, S. U.

(1996) Myocardial contractile response to nitric oxide and cGMP. Circulation 93, 1223±9.

MURAD, F., ISHII, K., FORSTERMANN, U., GORSKY, L., K E R W I N , J . F . , P O L L O C K , J . & H E L L E R , M . (1990) EDRF is an intracellular second messenger and autacoid to regulate cyclic GMP synthesis in many cells. In The Biology and Medicine of Signal Transduction (edited by N I S H I Z U K A , Y ., et al.), pp. 143±64. New York: Raven Press.

MURPHY, S., SIMMONS, M. L., AGULLO, L., GARCIA, A., FEINSTEIN, D. L., GALEA, E., REIS, D. J., MINC-

G O L O M B , D . & S C H W A R T Z , J . P . (1993) Synthesis of nitric oxide in CNS glial cells. TINS 16, 323±8. M U R R A N T , C . L . & B A R C L A Y , J . K . A . D . (1995) Endothelial cell products alter mammalian skeletal muscle function in vitro. Can. J. Physiol. Pharmacol. 73, 736±41. MURRANT, C. L., WOODLEY, N. E. & BARCLAY, J. K. A .D. (1994) Effect of nitroprusside and endothelium-derived products on slow-twitch skeletal muscle function in vitro. Can. J. Physiol. Pharmacol. 72, 1089±93.

NAKANE, M., SCHMIDT, H. H. H. W., POLLOCK, J.,

(1983) Cloned human brain nitric oxide synthase is highly expressed in skeletal muscle. FEBS Lett. 316, 175±80.

FORSTERMANN, U. & MURAD, F.

876

BELIA et al.

(1995) Relaxation of arterial smooth muscle by calcium sparks. Nature 270, 633±7. P A L M E R , R . M . J . , F E R R I G E , A . G . & M O N C A D A , S . (1987) Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 327, 524±6. R E G N I E R , M . , L O R E N Z , R . R . & S I E C K , G . C . (1992) Effects of oxygen radical scavengers on force production in single living frog skeletal muscle ®bers (abstract). FASEB J. 6, A819. R E I D , M . B . , K H A W L I , F . A . & M O O D Y , M . R . (1993) Reactive oxygen in skeletal muscle: III. Contractility of unfatigued muscle. J. Applied Physiol. 75, 1081±7. S C H A T Z M A N N , H . J . (1973) Dependence on calcium concentration and stoichiometry of the calcium pump in human red cells. J. Physiol. 235, 551±69. NELSON, M. T., CHENG, H. & RUBART, M.

SCHERER-SINGLER, U., VINCENT, S. R., KIMURA, H. & M C G E E R , E . G . (1983) Demonstration of a unique population of neurons with NADPH diaphorase histochemistry. J. Neurosci. Meth. 9, 229±34.

SCHMIDT, H. H. H. W., LOHMANN, S. M. & WALTER, U.

(1993) The nitric oxide and cGMP signal transduction system: regulation and mechanism of action. Biochim. Biophys. Acta 1178, 153±75.

STOYANOVSKY, D., MURPHY, T., ANNO, P. R., KIM, Y.-M. & S A L A M A , G . (1997) Nitric oxide activates skeletal and cardiac ryanodine receptors. Cell Calcium 21, 19± 29.

STUART, J., PESSAH, I. N., FAVER, T. G. & ASCBRAMSON, J . J . (1992) Photooxidation of skeletal muscle sarcoplasmic reticulum induces rapid calcium release. Archiv. Biochem. Biophys. 292, 512±21.

XIONG,

H.,

BUCK,

E.,

STUART,

J.,

PESSAH,

I.

N.,

(1992) Rose bengal activates the Ca2+ release channel from skeletal muscle sarcoplasmic reticulum. Archiv. Biochem. Biophys. 292, 522±8. W E S T E R B L A D , H . & A L L E N , D . G . (1993) The contribution of (Ca2+)i to the slowing of relaxation in fatigued single ®bres from mouse skeletal muscle. J. Physiol. 468, 729± 40. SALAMA, G. & ABRAMSON, J. J.