Concerted regulation of skeletal muscle contractility by oxygen tension and endogenous nitric oxide

Experimental Procedures

Muscle Bioassay. EDL muscles were obtained from male mice at 6–8 weeks of age (nNOS^-/-and C57BL/6, The Jackson Laboratory). Each experiment used the left and right EDL from a single mouse, which were removed and suspended from proximal and distal tendons, in series with force transducers, in organ baths at 37°C containing Krebs' solution supplemented with d-tubocurarine (25 μM; to eliminate any influence of neuromuscular transmission). Muscles were allowed to equilibrate for ≈15 min at 5% CO₂ and either 1–2% O₂, 20% O_2' or 95% O₂. Bath oxygen tension was monitored with an oxygen electrode (WPI Instruments, Waltham, MA). When used, l-nitroarginine methyl ester (l-NAME) (1 mM) was added to the bath at the beginning of the equilibration period. All measurements of muscle pairs were carried out in parallel. Contractions were elicited by square-wave constant-voltage stimulation (50–80 V) of 2 msec duration through 7.0-mm-wide platinum electrodes placed parallel to the long axis of the muscle. Before testing, each muscle was adjusted to the length at which isometric twitch-force generation was maximal (optimum muscle length, L_o). At least 2 min elapsed between single-twitch stimuli, and between trains of tetanic stimuli, and tetanic testing always followed the completion of single-twitch measurements. Tetanic stimulation consisted of pulse trains of 75 msec duration, and the amplitude of individual stimuli was set at a level equal to either half-maximal or maximal single-twitch stimulation. For each muscle, L_o was measured after the final tetanic stimulation, the muscle was weighed wet, and the effective cross-sectional area was calculated by approximating the muscle as a cylinder of length, L_o, and a density of 1.06 gm/cm^-3 (12). Because L_o did not differ significantly between WT and nNOS^-/- EDL (1.35 ± 0.06 cm vs. 1.36 ± 0.06 cm, P = 0.77), differences in effective cross-sectional area were largely a result of differences in muscle mass. Between-group analyses of single or multiple parameters were carried out, respectively, with Student's t test and ANOVA.

Measurement of Muscle pO₂. EDL muscles were fixed horizontally in Krebs' solution bubbled continuously with 20% O₂/5% CO₂. A polaragraphic oxygen electrode (glass micropipette; tip diameter ≈10 μm), calibrated after each use, was used to measure pO₂ after penetration of the muscle core (13). The composition of the sparging gas was changed sequentially from 20% O₂ to 95% O₂ and 1% O₂ after equilibration at each pO₂ (≈10 min).

Myocyte Ca²⁺ Transients and Sarcomere Shortening. FDB muscles obtained from C57BL/6 and nNOS^-/- mice were incubated with 2% collagenase type 2 (Worthington Lab) for 18 h at 37°C under 95%O₂/5% CO₂ before myocytes were dispersed and stored until use in Tyrodes solution containing 1.8 mM Ca²⁺ at room temperature. For experimental observations, myocytes were transferred to a sealed microscope stage compartment and superfused continuously at room temperature with Tyrodes solution containing 1.8 mM Ca²⁺ and 0.5 mM probenecid. Observations were obtained first at 20% O₂ before adjustment to 1% O₂. Depolarizing stimulation (pacing) was delivered continuously at 0.5 Hz. Measurement of intracellular Ca²⁺ and of sarcomere length were as described (14). Ca²⁺ measurement used Fura-2/AM (5 μM) and a dual-wavelength spectrophotometer exciting alternately at 365 and 380 nm. Sarcomere length was monitored continuously with a charge-coupled device camera and changes in sarcomere length were determined by fast Fourier transform of the Z-line density trace.

Results

We first compared the contractile properties of EDL at a physiologically extreme bath oxygen tension of 20–21% O₂ (pO₂ ≈150 mmHg, ambient pO₂), which represents an oxidative stress for peripheral tissue, and at 1–2% O₂ (pO₂ ≈10 mmHg), which is characteristic of normal mammalian striated muscle in vivo (8, 9). Direct measurement at the muscle core (see Experimental Procedures and Table 1) indicated that mean muscle pO₂ was ≈40 mmHg (median 36.7 mmHg) at 20% bath O₂ and ≈3.5 mmHg (median 1.75 mmHg) at ≈1–2% bath O_2. Thus, 1–20% bath O₂ recapitulates the physiological O₂ gradient in skeletal muscle (8, 9). In addition, to allow a direct comparison with previous results, we examined EDL performance at 95% bath O₂ (pO₂ ≈700 mmHg), in which we measured a core pO₂ ≈400 mmHg, thus eliminating the physiological O₂ gradient (Table 1). Some measurements of muscle performance were obtained sequentially at different oxygen tensions to examine the time course and reversibility of pO₂-dependent changes in contractility (see Fig. 2 A).

We assessed fundamental measures of E–C coupling and force production, isometric and tetanic twitch tensions, for muscles adjusted along their length-tension curve to optimize force generation. Measurements of force evoked by single-pulse stimulation revealed that, with stimuli producing about half-maximal twitch tension, the force generated by EDL was increased by ≈45% at 1% vs. 20% O₂, and even for stimuli producing maximal twitch tensions, twitch tension amplitude was ≈20% greater at low pO₂ (P < 0.05; Fig. 1A). These observations reveal pO₂-dependent regulation of E–C coupling, and recent findings provide a likely mechanism: NO modifies by S-nitrosylation a single cysteine residue (Cys-3635) within RyR1 and thereby augments substantially Ca²⁺ flux through RyR1, which governs active contraction (10, 15). Remarkably, NO acts on this critical thiol only at physiological O₂ concentrations (≈10 mmHg or less), where the pO₂-dependent redox state of a small subset of RyR1 thiols (representing an O₂ sensor) is reduced, thereby producing an NO-responsive conformation in the channel (10). In contrast, NO has no effect at supranormal pO₂, where these allosteric effector thiols are oxidized. Thus, mechanistic analysis of NO effects on RyR1 indicates that NO should facilitate contractility at low physiological pO₂. In contrast, the original analysis of diaphragm muscle (5, 16) and subsequent studies (reviewed in ref. 6), all carried out at supraphysiological pO₂, have reported consistently that pharmacologic inhibition of NOS enhances force production by skeletal muscle ex vivo (i.e., NO is inhibitory).

We reassessed directly the role of endogenous NO in pO₂-dependent regulation of contractility employing both nNOS^-/- mice and acute inhibition of NOS with l-NAME. The isometric twitch tensions developed by EDL from nNOS^-/- mice were indistinguishable at 1% or 20% O₂, with either submaximal or maximal stimulation (Figs. 1B and 2B). Thus, in the absence of endogenous NO production by nNOS, the influence of pO₂ on twitch tension is completely eliminated. This finding, in combination with the increase in twitch tension at low vs. high pO₂ in WT EDL (Fig. 1 A), indicates a concerted influence of pO₂ and NO that results in the enhancement by endogenous NO of E–C coupling and force production at physiological oxygen tension. In addition, we observed that nNOS^-/- EDL generated greater twitch force than did WT EDL (normalized for cross-sectional area), at both low and high pO₂ (Fig. 1). At 1% O₂, this increase was nonsignificant for half-maximal twitch tension, but maximal twitch force was increased by ≈22%, and at 20% O₂, twitch force increased by ≈54% for both submaximal and maximal twitch tensions (P < 0.05). These observations indicate the operation of distinct NO-independent compensatory force-enhancing mechanisms that are regulated by O₂ in fast-twitch muscle deprived chronically of nNOS activity.

The effects on twitch tension of acute NOS inhibition provided additional evidence for concerted regulation by NO and O₂ of EDL function (Fig. 1 A). At a physiological O₂ concentration of 1% O₂ (characteristic of active muscle), l-NAME treatment (1 mM) reduced half-maximal and maximal twitch force by 30% and 42%, respectively (P < 0.05), which is closely comparable to the difference between twitch tensions obtained at 1% and 20% O₂. Thus, consistent with the results from EDL of nNOS^-/- mice, the magnitude of the NO-dependent enhancement of contractility revealed by l-NAME accounts fully for the potentiating effects on twitch tension of low vs. high pO₂. In striking contrast, but consistent with previous observations (5, 6, 16), twitch force was enhanced by l-NAME at 20% O₂, resulting in increases of 43% and 24% for submaximal and maximal twitch tensions, respectively (P < 0.05; Fig. 1 A). The enhancement of force production by NOS inhibition in the ambient air condition (≈40 mmHg) indicates that at higher physiological pO₂ (associated with relative muscle inactivity), NO exerts an inhibitory influence on the contractile apparatus. The mechanism of action is unlikely to involve a direct modulation of Ca²⁺ release from the SR because RyR1 is not modified by NO at supranormal pO₂ (10, 15).

We also examined the response of EDL to trains of stimuli that produce tetanic (fused) contractions, and, which, relative to single-twitch stimulation, will evoke large and sustained increases in cytosolic Ca²⁺ levels and NO production (6). We used pulse trains comprised of individual pulses with amplitudes corresponding to either half-maximal or maximal single-twitch stimulation, and in no case did stimulus amplitude influence the reported effects of pO₂ or of NO. In addition, it should be noted that, although maximal tetanic tension developed at stimulus rates of 200–250 per sec under all conditions (Figs. 2 and 3), observations in intact, behaving rodents indicate that the physiological range of sustained discharge of motor neurons supplying fast-twitch muscles of the limb extends only through ≈100 per sec (17, 18), which in fact encompasses the range over which we observed the largest NO/O₂-dependent differences in contractility (vide infra).

Normalized force frequency curves obtained from WT EDL at 1% vs. 20% O₂ did not differ significantly over the measured tetanic frequency range of 40–250 Hz. (Fig. 3A). However, consistent with the pO₂-dependent effects of acute NOS inhibition on twitch tension (Fig. 1 A), treatment with l-NAME resulted in a leftward shift of the force frequency relationship at 20% O₂ and a rightward shift at 1% O₂ (Fig. 3A). The difference between the normalized force frequency curves at 1% O₂ and 20% O₂ after l-NAME treatment was significant over the physiologically relevant range of 40–120 Hz (P < 0.05). More strikingly, plots of absolute force vs. stimulation frequency revealed that acute NOS inhibition produced substantial increases in tetanic force at 20% O₂ (Fig. 3B), which averaged 68% and were significant over the frequency range of 40–120 Hz vs. both 20% O₂ in the absence of l-NAME (P < 0.05; Fig. 3B) and 1% O₂ in the presence or absence of l-NAME (P < 0.05; Fig. 3B). Force generation/stimulation frequency curves obtained at 20% O₂ in the absence of l-NAME were not significantly different from 1% O₂ in the presence of l-NAME and 1% O₂ in the absence of l-NAME (Fig. 3B). Tetanic force generation by WT EDL at 95% O₂ was substantially greater than at 1% O₂ or 20% O₂ (with peak tensions of ≈35–40 M/cm², which is equivalent to ≈20 g; Figs. 2 A and 3 B and C), but, whereas l-NAME treatment yielded a small but significant leftward shift in the normalized force frequency curves at lower stimulus frequencies (consistent with previous findings at 95% O₂; refs. 5, 6, and 12), no significant differences were obtained between absolute force vs. frequency curves at 95% O₂ in the presence vs. absence of l-NAME (Fig. 3 C and D). Thus, the inhibitory influence of NO on force production evident at 20% O₂ (≈40 mmHg) is no longer exerted in the extreme redox environment created <95% O₂ (≈400 mmHg). Taken together, the effects of NOS inhibition on contractile responses to repetitive stimulation indicate that the influence of NO varies with pO₂ and muscle activation, which will conjointly determine the redox milieu of critical subcellular compartments (19, 20).

Normalized force-frequency curves from EDL of nNOS^-/- mice were essentially indistinguishable at 1%, 20%, and 95% O₂ (Fig. 3 E and H). Thus, in the chronic absence of nNOS-derived NO, pO₂ exerts no influence on the form of the force-frequency relationship obtained by repetitive stimulation. In addition, l-NAME had no significant effect on either normalized or absolute force frequency curves from nNOS^-/- mice (Fig. 3 G and H). However, consistent with the increase in twitch force generated by EDL of nNOS^-/- vs. WT (Fig. 1B), tetanic force production by EDL of nNOS^-/- mice at 20% O₂ (Figs. 2B and 3F) was greater than at 1% O₂ and exceeded force generation by WT EDL at 1% or 20% O₂ (P < 0.05; tetanic force production did not differ significantly between nNOS^-/- EDL at 1% O₂, WT EDL at 1% O₂, and WT EDL at 20% O₂). Thus, pO₂-dependent differences in force generation were revealed in the absence of nNOS as well as by acute NOS inhibition (Fig. 3 A and B). Furthermore, the pO₂-dependent difference in force generation by EDL from nNOS^-/- mice increased with increasing tetanic stimulus frequency (Figs. 2B and 3F). As for WT EDL, force generation by nNOS^-/- EDL was greater at 95% O₂ than at 1% O₂ and at 20% O₂ (Fig. 3 F and G). It should be noted that at 95% O₂, force production by nNOS^-/- EDL and WT EDL was similar (Fig. 3 C and G), i.e., the enhanced contractility of nNOS^-/- EDL at low and intermediate pO₂ is much less evident at grossly supranormal oxygen tension. These observations establish that tetanic force generation is increased in nNOS^-/- muscle at higher physiological pO₂ and suggest that mechanisms operating to increase force production may include increased generation of or sensitivity to force-enhancing factors, such as reactive oxygen species (ROS, which may be produced in greater quantities with increasing pO₂ and muscle activation), and that pathological amounts of ROS may obscure redox signaling (19, 20). We have observed only minimal effects of Cu/Zn superoxide dismutase, applied to intact muscle (19), on force production by EDL (unpublished results), but a role for ROS generated and acting intracellularly remains to be examined.

A direct demonstration of concerted regulation by pO₂ and NO of both Ca²⁺ flux and contractile function was provided by analysis of depolarization-induced changes in cytoplasmic Ca²⁺ levels and sarcomere length in myocytes isolated from FDB (of WT and nNOS^-/- mice and examined at 1% O₂ and 20% O₂ (14) (Fig. 4). Both calcium transients and sarcomere shortening evoked by brief electrical stimulation were significantly greater at 1% O₂ than at 20% O₂ in myocytes derived from WT FDB (P < 0.05), whereas pO₂ had no effect on either measure in nNOS^-/- myocytes. In addition, both the amplitude of Ca²⁺ transients and the magnitude of sarcomere shortening were significantly greater at 20% O₂, but not at 1% O₂, in nNOS^-/- vs. WT myocytes (P < 0.05). Thus, these observations confirm at the level of single muscle cells both the facilitatory action of NO on contractility at low physiological pO₂ and the pO₂-regulated compensatory enhancement of contractile function in nNOS^-/- muscle. At the molecular level, the pO₂-dependent influence of NO on Ca²⁺ transients, which are known to reflect primarily Ca²⁺ release through RyR1 (21), supports a role for RyR1 as a principal locus of action in concerted redox regulation by pO₂ and NO (10, 15).

Discussion

Our findings demonstrate that skeletal muscle contractility varies with oxygen tension and that this pO₂-dependent modulation is effected in substantial part through the action of NO generated by nNOS in situ. Concerted regulation by NO and O₂ of E–C coupling and force production operates at both lowand high-oxygen tension. The influence of NO at low physiological pO₂ is facilitatory, whereas NO inhibits contractility at higher physiological pO₂. Only the inhibitory influence has been described in previous studies (6) of mammalian striated muscle, which have been conducted exclusively at entirely nonphysiological oxygen tension (95% O₂). Furthermore, because those results (obtained with pharmacological inhibitors) were not consistent with the reported absence of altered muscle function in nNOS^-/- or nNOS-deficient mdx mice (2, 3, 22), the importance of NO in E–C coupling and its potential pathophysiological significance in dystrophic disease have remained in question.

The facilitatory action of NO at normal muscle pO₂ supports the physiological importance of NO-dependent enhancement of calcium release from skeletal muscle SR through S-nitrosylation of RyR1 at physiological but not supranormal pO₂, observed in vitro (10). In addition, it has recently been shown that endogenous NO enhances RyR-mediated release of Ca²⁺ from SR of cardiac myocytes (23), an action mediated specifically by nNOS (14). The inhibitory effects of NO on contractility at higher pO₂ may reflect the operation of distinct molecular mechanisms, which are revealed most prominently under conditions of oxidative stress. Redox-sensitive muscle proteins that are reported to be modified by NO or S-nitrosoglutathione (GSNO) include the Ca²⁺-dependent ATPase of the SR (24, 25), which regulates SR-cytosol Ca²⁺ homeostasis, and constituents of the myofilaments (26, 27), which directly subserve contractile function. In addition, activation of guanylate cyclase has been implicated in depression of muscle force by NO/GSNO in previous studies (5, 28) conducted at supraphysiological pO₂.

Regardless of the identity of the full spectrum of proteins acted on by NO/nitrosylation in muscle cells, our experimental results support a model under which the targets, mechanisms of action, and functional effects of endogenous NO will vary continuously with muscle redox status, linked to oxygen tension (and presumably ROS production). The contingent effects of NO would thus couple modulation of contractility to ongoing muscle activity and metabolic demand. Low pO₂ in working muscle would increase contractility and rises in pO₂ in resting muscle would attenuate contractility, which is analogous to the pO₂-dependent regulation of red blood cell-derived NO vasoactivity that couples peripheral blood flow to local metabolic demand (i.e., hypoxic vasodilation and hyperoxic vasoconstriction; refs. 29 and 30). Accordingly, O₂-regulated, NO/S-nitrosothiol-based signals would coordinately regulate both blood flow and contractility in striated muscle beds. This model may apply more generally across cell types and subcellular compartments in which NO-derived bioactivity is exerted, and may thus provide a framework for elucidating the role of NO in redox-based regulation of protein function and cellular physiology (1), which has been largely masked in studies performed to date at supraphysiological pO₂.

Increased force production as a result of oxidative insult (see 95% O₂ in Figs. 2 and 3) or NOS deficiency (Figs. 1, 2, and 3 F and G) may provide a compensatory mechanism to maintain performance by injured or diseased (dystrophic) muscle. Indeed, it may be the case that chronic deprivation of NO results in oxidative stress. However, this greater force production comes at a price: NO and O₂ no longer exert coordinate control (RyR activity is not regulated by S-nitrosylation at supraphysiological pO₂). This broader perspective may provide a novel molecular basis for diverse myopathies that are associated with altered expression, distribution or activity of NOS, and/or impairment of oxygenation or blood flow, previously subsumed under the rubric of oxidative/nitrosative stress, but here viewed as symptomatic of an imbalance in NO/O₂-based signaling that is essential for normal calcium-regulated muscle function. That is, rather than viewing oxidative/nitrosative stress simply in terms of oxidative injury to muscle, our results suggest that impairments in muscle performance may result from physiological alterations in NO/O₂ balance that disrupt nitrosylation/redox-based signaling.