Inhibition of nitric oxide synthase (NOS) significantly attenuates the increase in skeletal muscle glucose uptake during contraction/exercise, and a greater attenuation is observed in individuals with Type 2 diabetes compared with healthy individuals. Therefore, NO appears to play an important role in mediating muscle glucose uptake during contraction. In this study, we investigated the involvement of neuronal NOSμ (nNOSμ), the main NOS isoform activated during contraction, on skeletal muscle glucose uptake during ex vivo contraction. Extensor digitorum longus muscles were isolated from nNOSμ−/− and nNOSμ+/+ mice. Muscles were contracted ex vivo in a temperature-controlled (30°C) organ bath with or without the presence of the NOS inhibitor NG-monomethyl-l-arginine (L-NMMA) and the NOS substrate L-arginine. Glucose uptake was determined by radioactive tracers. Skeletal muscle glucose uptake increased approximately fourfold during contraction in muscles from both nNOSμ−/− and nNOSμ+/+ mice. L-NMMA significantly attenuated the increase in muscle glucose uptake during contraction in both genotypes. This attenuation was reversed by L-arginine, suggesting that L-NMMA attenuated the increase in muscle glucose uptake during contraction by inhibiting NOS and not via a nonspecific effect of the inhibitor. Low levels of NOS activity (∼4%) were detected in muscles from nNOSμ−/− mice, and there was no evidence of compensation from other NOS isoform or AMP-activated protein kinase which is also involved in mediating muscle glucose uptake during contraction. These results indicate that NO regulates skeletal muscle glucose uptake during ex vivo contraction independently of nNOSμ.
exercise (muscle contraction) can cause significant glucose uptake into skeletal muscle fibers to the extent that elevated plasma glucose levels of individuals with Type 2 diabetes (T2D) is normalized (42). Muscle contraction stimulates muscle glucose uptake via an increase in GLUT4 translocation (13, 65). Although not fully elucidated, the cellular mechanisms involved in glucose uptake during muscle contraction appear to include AMP-activated protein kinase (AMPK), calcium/calmodulin-dependent protein kinase II (Ca2+/CaMKII), protein kinase C (PKC), reactive oxygen species (ROS), or nitric oxide (NO) (36, 47). Over the last few years additional novel proteins, including Myo1c (61), PIKfyve (34), Rac1, and PAK1, (59) have been shown to at least partially regulate skeletal muscle glucose uptake during contraction. There are clearly multiple sites and levels of regulation and multiple levels of redundancy embedded within the pathway for contraction-mediated glucose uptake.
Skeletal muscle contraction activates NO synthase (NOS) leading to an increase in NO production (2, 28). Balon and colleagues (1, 48) demonstrated that NO was involved in mediating muscle glucose uptake during contraction. Our group has accumulated evidence from a range of experimental models, both in rodents and humans, that NOS inhibition attenuates the increase in skeletal muscle glucose uptake during contraction/exercise. These include mouse muscles stimulated to contract ex vivo (37, 38), rat muscles contracting in situ (51), and humans in vivo during cycling exercise in both healthy (6, 25) and individuals with T2D (25). Interestingly, the role of NO in contraction-stimulated glucose uptake appears to be greater in individuals with T2D, given that local NOS inhibition caused a significantly greater attenuation of the increase in leg glucose uptake in individuals with T2D performing cycling exercise compared with their healthy counterparts (25).
The specific isoform of NOS that might be responsible for mediating skeletal muscle glucose uptake during contraction/exercise is unknown. The nonspecific NOS inhibitors [NG-nitro-l-arginine methyl ester (L-NAME) and NG-monomethyl-l-arginine (L-NMMA)] used in these studies inhibit all NOS isoforms present in the muscle fibers. A study using eNOS−/− mice demonstrated that whole body absence of eNOS is associated with higher skeletal muscle glucose uptake during treadmill exercise (29). These mice are, however, exercise intolerant and have significantly lower exercise-induced hyperemia. Therefore, the elevated glucose uptake was postulated to be attributed to exercise-induced hypoxia (29). Neuronal NOSμ is believed to be the main isoform activated during skeletal muscle contraction as guanosine 3′,5′-cyclic monophosphate (cGMP) content, a downstream signaling of NO, in isolated extensor digitorum longus muscles (EDL) from C57Bl/6 and eNOS−/− mice and was increased during contraction, but not in nNOSμ−/− mice (28). Skeletal muscle glucose uptake during contraction has not been assessed in nNOSμ−/− mice.
Therefore, the purpose of this study was to investigate the role of nNOSμ in the regulation of skeletal muscle glucose uptake during ex vivo contraction. We tested the hypothesis that mice lacking nNOSμ would show an attenuated increase in muscle glucose uptake during ex vivo contraction.
MATERIALS AND METHODS
All procedures were approved by the University of Melbourne Animal Ethics Committee and conformed to the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (2013, 8th Edition). Male and female nNOSμ+/+ and nNOSμ−/− littermates were generated by mating C57Bl/6 nNOSμ heterozygous breeding pairs originally obtained from the Jackson Laboratory (Jax Mice, Bar Harbor, ME). Genotyping was performed by standard PCR techniques on genomic DNA extracted from tail samples taken at day 21. Mice were housed in standard cages and maintained under constant temperature of 21 ± 1°C with 12-h light/dark cycle in the Biological Research Facility at the University of Melbourne. They had access to standard rodent chow and water ad libitum. All mice were at 14 to 17 wk of age when experiments were conducted. To provide an overview of the anti-nNOSμ antibody, brain and/or muscle tissues from nNOSμ+/− mice, C57Bl/6 mice, and Sprague Dawley rat were used for immunoblotting investigations.
Muscle dissection and experimental procedure.
The surgical and experimental procedures were performed as described previously (37). EDL muscles were used as it has been shown that nNOS expression is greater in fast-twitch muscles (26), and NOS inhibition has a greater effect on glucose uptake in EDL than soleus muscles when contracted ex vivo (38). Muscles were excised from deeply anaesthetized mice (Avertin: 0.25 mg/g body wt, intraperitoneally; Sigma Aldrich Chemicals, St. Louis, MO), and one tendon was attached to a fixed hook with surgical silk while the other tendon was attached to a force transducer (PanLab, Barcelona, Spain). Muscles were suspended vertically at optimal length, determined by progressively increasing muscle length in small increments until maximum twitch contraction force was obtained in Ringer solution (in m: 118.5 NaCl, 24.7 NaHCO3, 4.74 KCl, 1.18 MgSO4, 1.18 KH2PO4, 2.5 CaCl2, pH 7.4) with the addition of 0.01% BSA, 8 mM mannitol, and 2 mM sodium pyruvate. This solution was maintained at 30°C and oxygenated continuously with 95% O2 and 5% CO2 throughout the experiment. Muscles were incubated for 40 min in the presence or absence of L-NMMA (NOS inhibitor, 100 μM, Sigma Aldrich) or L-NMMA + L-arginine (NOS substrate, 1 mM, Sigma Aldrich). Muscles either remained rested or were stimulated to contract during the final 10 min of incubation. Contractions were elicited by square wave electrical pulses (350 ms train; 60 Hz, 0.2 ms; 6 V; 12 contractions/min) generated by a Grass S48 stimulator (Grass Instruments, Warwick, RI) and delivered via two platinum plate electrodes that flanked, without touching, the muscles. Contraction force was acquired with PowerLab Chart 4.0 software (AD Instruments, Castle Hill, Australia). At the end of the basal or contraction incubation, muscles were rapidly cut from the suture attachments, washed in ice-cold buffer, blotted on filter paper, snap frozen in liquid nitrogen, and kept at −80°C for biochemical analyses later.
Muscle glucose uptake measurement.
Muscle glucose uptake was measured during the final 5 min of contraction or basal incubation in which the incubation buffer was rapidly exchanged with buffer containing 1 mM [1,2-3H]2-deoxy-glucose (0.128 μCi/ml) and 8 mM D-[14C] mannitol (0.083 μCi/ml) (PerkinElmer, Boston, MA). At the end of an experiment, muscles were immediately washed in ice-cold buffer, blotted on filter paper and snap frozen in liquid nitrogen. Muscle mass was recorded, and the entire frozen muscle was digested with 1 M NaOH for 10 min at 80°C which was then neutralized with 1 M HCl, vortex-mixed, and centrifuged at 13,000 g for 2 min. The supernatant was recovered, added to a liquid scintillation cocktail (PerkinElmer, Boston, MA), and radioactivity was determined with a β-scintillation counter (Packard TriCarb 2900TR, PerkinElmer, Boston, MA). Muscle glucose uptake was calculated as described previously (37).
Immunoblotting was performed with frozen muscle/brain sections (10 sections, 20-μm thickness) which were homogenized with 100 μl of solubilizing buffer [0.125 M Tris-HCl (pH 6.8), 4% SDS, 10% glycerol, 10 mM EGTA, 0.1 M DTT, and 0.001% bromophenol blue] as modified from the method described previously (41). Protein concentration of homogenate was determined by the RED 660 Protein Assay Kit (G-Biosciences, St Louis, MO). Homogenates containing 5 or 7 μg of total protein were separated on SDS-PAGE gels (Bio-Rad) and then wet transferred onto polyvinylidine fluoride membranes. Membranes were probed with the following primary antibodies overnight at 4°C: phospho-AMPKα Thr172 (No. 2531; 1:1000), AMPKα (No. 2793; 1:1,000), α-tubulin (No. 2144; 1:1,000) (Cell Signaling Technology, Danvers, MA); nNOS (No. 610308; 1:10,000), eNOS (No. 610296; 1:10,000), iNOS (No. 610328; 1:2,000) (BD Biosciences, San Jose, CA); GLUT4 (No. PA1-1065; 1:8,000) (Thermo Scientific, Rockford, IL), and actin (No. A2172; 1:40,000) (Sigma Aldrich). Following incubation with antimouse or antirabbit HRP-secondary antibodies and a series of washes in Tris-buffered saline with Tween, chemiluminescent signal was developed with ECL substrate (SuperSignal West Femto, Pierce). Blot images were taken with a charge-coupled device camera with Quantity One software (Bio-Rad). Prestained molecular weight markers on the membrane were imaged under white light source with the membrane position unchanged. When quantifying both phosphorylated and total protein abundance, membranes were first probed with phosphorylation-specific primary antibody and then stripped (62.5 mM Tris-HCl pH 6.8, 2% SDS, 0.8% β-mercaptoethanol), reblocked, and reprobed with primary antibody against the total protein. Loading control proteins were always probed using nonstripped membranes. Actin was used as loading control for all, except for GLUT4 where α-tubulin was used since actin and GLUT4 have similar molecular weights and it was not possible to probe both of these proteins without undertaking the stripping process.
NOS activity assay.
NOS activity was determined based on the conversion of radiolabeled L-[14C]arginine to radiolabeled L-[14C]citrulline as described previously (30). NOS activity was determined from the difference between samples incubated with and without L-NAME and was expressed as picomoles of L-[14C] citrulline formed per minute per milligram of protein.
All data are expressed as means ± SE. Statistical testing was performed with SPSS statistical package with t-test (nNOSμ+/+ vs nNOSμ−/−: body weight and muscle mass, expressions of total AMPKα, eNOS, and GLUT4), two-factor ANOVA (between factor: genotype and treatment condition—glucose uptake, contraction force, NOS activity), or two-factor repeated measures ANOVA (between factor: genotype and treatment condition; within factor: time—decrease in contraction force over time). If there were significant interactions, specific differences between mean values were identified using Fisher's least significance test. The significance level was set at P < 0.05. In this study no sex-specific difference in glucose uptake was observed, and so data from both female and male mice were pooled and analyzed together.
Morphological characteristics of nNOSμ+/+ and nNOSμ−/− mice.
Body weight was significantly lower in the age-matched nNOSμ−/− mice compared with their nNOSμ+/+ littermates (26.3 ± 0.6 vs. 28.6 ± 0.7 g; P < 0.05; n = 22 and 36). EDL muscle mass was also significantly smaller in nNOSμ−/− than in nNOSμ+/+ mice (9.9 ± 0.3 vs. 10.9 ± 0.2 mg; P < 0.05; n = 44 and 72).
The peak contraction force normalized to muscle mass was not different between muscles from nNOSμ+/+ and nNOSμ−/− mice (Fig. 1A). Inhibition of NO production by L-NMMA did not affect peak contraction force in either genotype (Fig. 1A). Throughout the 10 min of contraction, force production decreased to a similar extent over time in both genotypes in the presence or absence of L-NMMA (Fig. 1B). Reversal of inhibition of L-NMMA by L-arginine (a NOS substrate which competes with and overcomes the inhibition of L-NMMA) had no effect on peak force production or the force production over time in muscles from nNOSμ+/+ mice (data not shown).
Skeletal muscle glucose uptake.
Basal muscle glucose uptake was not different between nNOSμ+/+ and nNOSμ−/−, and incubation with L-NMMA had no impact on basal muscle glucose uptake (Fig. 2). Contraction significantly increased glucose uptake in muscles from nNOSμ+/+ mice by approximately fourfold (Fig. 2). There was a similar increase in glucose uptake during contraction in muscles from nNOSμ−/− mice which was not statistically different from that of nNOSμ+/+ (Fig. 2). L-NMMA significantly attenuated the increase in muscle glucose uptake during contraction in both genotypes with no difference between genotypes (Fig. 2). This represented about 21 and 24% attenuation in the increase in glucose uptake during contraction in muscles from nNOSμ+/+ and nNOSμ−/− mice, respectively. L-arginine fully reversed the attenuation of glucose uptake during contraction by L-NMMA in muscles from nNOSμ+/+ mice (data not shown) suggesting that the attenuation of the increase in glucose uptake during contraction by L-NMMA was due to its inhibition of NOS, and not due to a nonspecific effect of the inhibitor.
Expression of nNOSμ and nNOS splice variants.
The BD Biosciences anti-nNOS antibody, which is specific for nNOS amino acids 1095-1289 in the C terminal region, detected a number of bands. It detects nNOS spliced variants based on their different sizes due to a truncation at the N terminal region. Given that skeletal muscles of rodents (56) and humans (27, 31) express an alternatively spliced variant of the conventional nNOS (nNOSα), known as nNOSμ, which has 34 extra amino acids, it is more appropriate to refer to the main isoform of nNOS expressed in skeletal muscle as nNOSμ. The 160 kDa nNOSμ band was detected in muscles from nNOSμ+/+ and not in nNOSμ−/− mice (Fig. 3A). There was also another band at 140 kDa, presumably the nNOSβ splice variant (4), detected in muscles from nNOSμ+/+ but not in nNOSμ−/− mice (Fig. 3A). To better compare the expressions of nNOSμ and nNOS splice variants, homogenates from muscles of nNOSμ+/− mice were run together with muscle homogenates from nNOSμ+/+ and nNOSμ−/− mice (n = 1) (Fig. 3A). Both the 160 kDa and 140 kDa bands were detected in muscles from nNOSμ+/+ and nNOSμ+/− mice but not from nNOSμ−/− mice (Fig. 3A). The bands were less dense in the nNOSμ+/− compared with nNOSμ+/+ mice, as would be expected. There were bands at 60 kDa and 52 kDa detected in all genotypes (Fig. 3A). As previously reported, the 140 kDa nNOSβ band was also detected in the muscles from C57Bl/6 mice (4) and brain tissue of nNOSμ−/− mice (7) (Fig. 3B).
Other protein expression and phosphorylation.
The expressions of actin and α-tubulin proteins were not different between genotypes and were suitable for use as loading controls. Total AMPKα expression was similar between muscles from nNOSμ+/+ and nNOSμ−/− mice, indicating no compensatory induction of AMPKα in muscles from nNOSμ−/− mice (Fig. 4A). In the resting state, muscles from nNOSμ+/+ and nNOSμ−/− mice had similar AMPKα Thr172 phosphorylation relative to AMPKα abundance (Fig. 4B). As expected, contraction significantly increased AMPKα Thr172 phosphorylation with an ∼10-fold increase observed in both genotypes but no difference between them (Fig. 4B). As shown previously in rats (51) and C57Bl/6 mice (37), NOS inhibition during contraction had no impact on AMPKα Thr172 phosphorylation in muscles from nNOSμ+/+ and nNOSμ−/− mice (Fig. 4B). In line with our previous report (63), we were unable to detect any iNOS in muscles from nNOSμ+/+ and nNOSμ−/− mice (data not shown). Endothelial NOS expression was not different between muscles from nNOSμ+/+ and nNOSμ−/− mice (Fig. 5A), indicating that the absence of nNOSμ in skeletal muscles did not cause an upregulation of eNOS protein expression to compensate for a reduction of NO production at resting state. GLUT4 protein expression was similar in muscles from nNOSμ+/+ and nNOSμ−/− mice (Fig. 5B).
In muscles from nNOSμ+/+ mice contraction significantly increased NOS activity, and L-NMMA inhibited it to a level significantly below the basal state (Fig. 6). This is consistent with our previously reported data from C57Bl/6 mice (37). There was some, albeit very small, NOS activity detected in muscles from nNOSμ−/− mice which represented ∼4% of nNOSμ+/+ value in the basal state (Fig. 6). A similar degree of residual NOS activity was reported previously in brain tissues from nNOSμ−/− mice (21).
In this study, we report for the first time that nNOSμ is not essential for the regulation of skeletal muscle glucose uptake during ex vivo contraction in mice. Interestingly, NOS inhibition significantly attenuated the increase in skeletal muscle glucose uptake during contraction to a similar extent in muscles from both nNOSμ+/+ and nNOSμ−/− mice. The NOS substrate L-arginine overcame the attenuation of contraction-stimulated glucose uptake by L-NMMA. These results suggest that NO mediates glucose uptake during ex vivo contraction independently of nNOSμ.
The regulation of skeletal muscle glucose uptake during contraction is a complex and integrated process that involves multiple signaling cascades and can be affected by several external factors such as stimulation protocol and contractile force production (22, 23). In the present study, and as previously shown (10, 45), nNOSμ−/− mice had lower body and muscle masses compared with the wild-type controls. It is not entirely clear why nNOSμ−/− mice have lower body and muscle masses, but it may be because although they have normal food intake (39) they have higher locomotor activity (54) compared with controls, suggesting a higher energy expenditure in these mice. It is unlikely that this affected glucose uptake during contraction in muscles from nNOSμ−/− mice because, similar to previous reports (10, 45), their smaller muscles produced normal contraction force when normalized to muscle mass (Fig. 1A). The same was found for fatigability where force decreased to a similar extent over time in both genotypes in the presence or absence of L-NMMA (Fig. 1B). As such, it is expected that muscles from both nNOSμ+/+ and nNOSμ−/− mice were activated equally with a similar metabolic challenge and, therefore, had the same stimulus for glucose uptake. This was supported by the similar increases in AMPK phosphorylation (Fig. 4B) which is a sensitive marker for metabolic status/perturbation (9). In addition, fasting glucose and insulin levels (55, 62) as well as serum lipid profiles (62, 64), all of which may affect muscle glucose uptake, have been found to be similar in nNOSμ−/− and control mice.
The normal glucose uptake during contraction in muscles from nNOSμ−/− mice raises the critical question of whether NO is important in mediating this process. The finding that NOS inhibition significantly attenuated skeletal muscle glucose uptake during contraction in muscles from both nNOSμ+/+ and nNOSμ−/− mice (Fig. 2) was also surprising and interesting. These suggest that 1) the attenuation may be due to non-NOS specific effects of the NOS inhibitor (L-NMMA), 2) the critical role of nNOSμ in producing NO in muscle fibers during contraction may have been compensated, or 3) nNOSμ is not essential for mediating skeletal muscle glucose uptake during contraction.
Several studies have reported that arginine analog NOS inhibitors have non-NOS specific effects and these effects appeared to be analog specific. For example, L-NAME but not L-NMMA was shown to bind to muscarinic receptors (8), and L-NNA inhibited rat liver arginase while L-NMMA and L-NAME did not (49). The nonspecific effects reported for L-NMMA, the NOS inhibitor used in this study, are unclear. It was reported that the vasoconstrictor effect of L-NMMA involved superoxide production (50, 60) via activation of cyclooxygenase with the resultant increased production of superoxide inactivating endothelium-derived relaxing factor (NO) (50). On the other hand, L-NMMA inhibited endothelium-dependent release of superoxide induced by increased blood flow (44). Any nonspecific effect of L-NMMA remains to be further established. In fact, the protective effect of L-NMMA on myofiber membrane injury is enantiomer specific, as the inactive D-NMMA conferred no protection (32) suggesting that L-NMMA's effect in skeletal muscle is specific via inhibition of NOS. In this study, it appears likely that L-NMMA attenuated the increase in muscle glucose uptake during contraction via inhibition of NOS. L-NMMA significantly attenuated the increase in NOS activity in parallel with the attenuation in the increase in glucose uptake during contraction (Fig. 2 and 6). Furthermore, the attenuation of glucose uptake by L-NMMA was completely reversed by L-arginine which is a NOS substrate that competes with L-NMMA for NOS binding site. Although L-arginine could compete with L-NMMA for cellular uptake (5), it has been shown previously that L-arginine infusion overcame the attenuation of leg glucose uptake during exercise in humans by L-NMMA while L-arginine infusion per se has no effect on leg glucose uptake (6). These findings suggest that the attenuation was due to inhibition of NOS and not a nonspecific effect.
The normal increase in glucose uptake during contraction in muscles from nNOSμ−/− mice could also be due to other NOS isoforms or splice variants that generated NO. Endothelial NOS was the other NOS isoform detected in these muscles; however, eNOS was previously shown to be unlikely involved in skeletal muscle glucose uptake during contraction/exercise. The absence of eNOS in diaphragm and soleus muscles does not alter the NO released during ex vivo contraction (19). Similarly, it was demonstrated that cGMP formation, a downstream mediator of NO signaling, is increased in EDL muscles of control and eNOS−/− but not nNOS−/− mice following electrical stimulation (28). These findings indicated that eNOS in muscle fibers is likely not involved in the production of NO during ex vivo muscle contraction. Furthermore, it has been demonstrated that mice with whole body eNOS deficiency do not have an attenuation of muscle glucose uptake during treadmill exercise (29). Taken together, these studies suggest that eNOS is not essential for skeletal muscle glucose uptake during contraction/exercise and unlikely to compensate for the loss of nNOSμ in skeletal muscles.
Nevertheless, the data from the present study suggests that the attenuation of glucose uptake during contraction is via the inhibition of a NOS isoform. It has been previously suggested that nNOSβ, an alternative splice variant of nNOS, is expressed in nNOSμ−/− mice (16) because these mice are developed by targeted deletion of exon 2 (21). Neuronal NOSβ and nNOSγ are still transcribed and intact in various tissues in these mice because they do not contain exon 2 (12, 16). Neuronal NOSβ is catalytically active with about 80% activity of nNOSμ, while nNOSγ is considered functionally inactive (7). To date, the presence of nNOSβ in skeletal muscles of nNOSμ−/− mice remains controversial. Some studies demonstrated indirect evidence from immunohistochemistry suggesting the presence of nNOS splice variants, presumably nNOSβ (see below), in skeletal muscles from nNOSμ−/− mice (45, 52). However, we and others (3) did not detect the presence of nNOSβ in skeletal muscles of nNOSμ−/− mice (Fig. 3A). We immunoblotted a band at 140 kDa, likely to be nNOSβ (as there is currently no commercially available nNOSβ-specific antibody), only in muscles from nNOSμ+/+ and nNOSμ+/− but not nNOSμ−/− mice (Fig. 3A). This 140 kDa band was also detected in skeletal muscles of C57Bl/6 mice and in brain tissue of nNOSμ−/− mice (Fig. 3B) (4, 7), supporting the identification of this band as nNOSβ. The reason(s) for the discrepancy between our findings and those of Baum et al. (3) and others (45, 52) regarding the presence of nNOSβ in skeletal muscles from nNOSμ−/− mice is difficult to discern. It is worthwhile noting that the studies suggesting the presence of nNOSβ in skeletal muscles from nNOSμ−/− mice were deduced from immunohistochemical assessments, not from immunoblotting (45, 52). Therefore, the identity and molecular weight of the stained protein(s) were unknown.
The attenuation in the increase in glucose uptake during contraction in skeletal muscles of nNOSμ−/− mice by L-NMMA suggests that NO was generated in these muscles, and, indeed, low levels of NOS activity were detected in these muscles. However, NOS activity measured in vitro may not accurately reflect the level of NO production in vivo. Unfortunately, direct measurement of NO, a gaseous signaling molecule with a very short biological half-life (17, 33), is technically very challenging, especially during muscle contraction ex vivo. Measurement of an end product of NO such as nitrite and nitrate as an indicator of NO production (40) will also not accurately reflect the production of NO in skeletal muscle during contraction. This is in part because the production of nitrite and nitrate may not be proportionate with NO production because of the presence of different levels of NO chelating molecules and the degree of S-nitrosylation or S-glutathionylation occurring inside and outside the contracting muscles. In addition, nitrite and nitrate can be converted back to NO, especially under hypoxic conditions such as during muscle contraction (15). Total levels of S-nitrosylation or S-glutathionylation in the muscles was not determined because it is rather nonspecific and may complicate the data interpretation.
It is difficult to imagine how this low level of NOS activity (∼4% of nNOSμ+/+) could be sufficient to maintain a normal physiological process mediated by NO. Nevertheless, this low level of NOS activity may be significant in these nNOSμ−/− mice given that nNOSμ−/− mice with only about 7% NOS activity detected in the central nervous system appear grossly normal and lack any histopathological abnormalities (21). This suggests that even a low level NOS activity might be sufficient to protect these mice from severe pathology (12). In addition, in isolated skeletal muscles from nNOSμ−/− mice, muscle integrity and performance also appear to be normal (45). This is in contrast with skeletal muscles with total loss of all nNOS splice variants (KN2 mice) which fatigued faster and had a severe disruption of the microtubule cytoskeleton (45). These results suggest that the low level of NOS activity is critical in preserving most, if not all of, the normal neurological and muscular functions in nNOSμ−/− mice.
It is difficult to reconcile how this small amount of NOS activity could be responsible for normal glucose uptake during contraction. Perhaps, the spike/fluctuation in NOS activity (NO production) during contraction rather than the absolute level might be more important for transduction of NO signaling and therefore stimulation of glucose uptake during contraction. During NOS inhibition at the L-NMMA dose we used (100 μM), skeletal muscle NOS activity has never been shown to be inhibited to zero, instead ranging between 10% to just slightly below basal levels as in this study (Fig. 6) and others (18, 37, 38). Relative to muscles from nNOSμ−/− mice, the NOS activity post-NOS-inhibition in muscles from C57Bl/6 and nNOSμ+/+ mice appears to be sufficient, yet there is still a significant attenuation of glucose uptake during contraction observed in the present study (Fig. 2) and other studies (37, 38). A caveat worth noting is that contraction protocols are quite intense in these studies as muscle NOS activity is not significantly increased at low exercise/stimulation intensities (30, 57). Furthermore, genetically modified animals with low residual contraction-induced AMPK phosphorylation or activity have normal muscle glucose uptake during contraction/exercise (14, 24, 35, 58), whereas total elimination of AMPK phosphorylation or activity caused up to 70% attenuation in the increase in muscle glucose uptake during contraction (43). These findings suggest that the low levels of NOS activity may be sufficient to maintain a normal skeletal muscle glucose uptake during contraction. In addition, as little as 5% of Akt phosphorylation is sufficient to elicit maximum effects of insulin on GLUT4 translocation in L6 myotubes (20). These studies clearly indicate that there is considerable reserve in the capacity of signaling intermediates for signaling transduction in mediating glucose uptake, and therefore the importance of low levels/residual signaling in mediating muscle glucose uptake should not be overlooked.
Another possibility for the normal contraction-stimulated glucose uptake in muscles from nNOSμ−/− mice is the compensatory increase in ROS signaling due to lack of NO to scavenge superoxide (O2·−) (11). ROS are produced during contraction (46, 53) and the excess O2·− could lead to an increased production of hydrogen peroxide (H2O2) which then activates glucose uptake. Our group (38) and others (53) have shown that ROS is involved in the regulation of muscle glucose uptake during ex vivo contraction. Therefore, H2O2 may have compensated for the deficient NO production to stimulate normal muscle glucose uptake during ex vivo contraction in muscles from nNOSμ−/− mice. Nevertheless, it has been shown that EDL muscles from nNOSμ−/− mice have a similar H2O2 concentration, determined by in vitro 2,7-dichlorofluorescein fluorescence assay, compared with C57Bl/6 control mice (11) opposing the compensatory effect of H2O2 in glucose uptake in muscles from nNOSμ−/− mice. There was no compensatory increase in AMPK phosphorylation following contraction in skeletal muscles from nNOSμ−/− mice. Likewise eNOS and GLUT4 protein expression was comparable between genotypes, and no iNOS was detected in the muscles that could possibly explain the normal glucose uptake during ex vivo contraction. Therefore, the normal muscle glucose uptake during contraction in muscles from nNOSμ−/− mice was unlikely to be attributed to a compensatory effect of other potential signaling involved in muscle glucose uptake during contraction. We do not have data on p38 MAPK as the involvement of this kinase in skeletal muscle glucose uptake during contraction remains to be firmly established (47).
In summary, nNOSμ is not essential for glucose uptake during ex vivo contraction of mouse skeletal muscle. Given that NOS inhibition attenuated muscle glucose uptake during contraction, and evidence that eNOS may not be important in this regard, we suggest that other nNOS splice variants could be the putative NO-producing enzymes responsible for the normal increase in muscle glucose uptake during ex vivo contraction. It will be important that future studies examine the effects of nNOS-specific inhibitors and mice with all nNOS isoforms deleted (KN2 mice) on muscle glucose uptake during contraction. It is also worth evaluating the possible redundant effects of eNOS and nNOS in muscle glucose uptake during contraction using double knockout mice, though we showed no evidence of eNOS compensating for nNOS in the present study.
The authors acknowledge funding from the National Health and Medical Research Council (NHMRC) of Australia and the Diabetes Australia Research Trust for the support for this study.
No conflicts of interest, financial or otherwise, are declared by the author(s).
Y.H.H., S.R., and G.K.M. conception and design of research; Y.H.H., T.F., and X.Z. performed experiments; Y.H.H. analyzed data; Y.H.H., R.M.M., G.S.L., A.C.B., S.R., and G.K.M. interpreted results of experiments; Y.H.H. prepared figures; Y.H.H. drafted manuscript; Y.H.H., T.F., X.Z., R.M.M., G.S.L., A.C.B., S.R., and G.K.M. edited and revised manuscript; Y.H.H., T.F., X.Z., R.M.M., G.S.L., A.C.B., S.R., and G.K.M. approved final version of manuscript.
The authors thank Timur Naim from the University of Melbourne and Robert Lee-Young from Baker IDI Heart and Diabetes Institute for their technical assistance in the experiments or biochemistry analysis.
- Copyright © 2015 the American Physiological Society