Nitric oxide synthase inhibition reduces O2 cost of force development and spares high-energy phosphates following contractions in pump-perfused rat hindlimb muscles

Authors


Corresponding author R. T. Hepple: Faculty of Kinesiology, University of Calgary, 2500 University Drive NW, Calgary, Alberta, T2N 1N4, Canada. Email: hepple@ucalgary.ca

Abstract

The purpose of the present experiments was to test the hypotheses that: (i) nitric oxide synthase (NOS) inhibition reduces the O2 cost of force development across a range of contractile demands; and (ii) this reduced O2 cost of force development would be reflected in a sparing of intramuscular higher energy phosphates. Rat distal hindlimb muscles were pump perfused in situ and electrically stimulated (200 ms trains with pulses at 100 Hz, each pulse 0.05 ms duration) for 1 min each at 15, 30 and 60 tetani min−1 and for 2 min at 90 tetani min−1 in three groups: 0.01 mm adenosine; 1 mm d-NAME and 0.01 mm adenosine (d-NAME); and 1 mm l-NAME and 0.01 mm adenosine (l-NAME). The gastrocnemius–plantaris–soleus muscle group was freeze clamped post-contractions for metabolite analyses. Force was 19% higher and oxygen uptake inline image was 20% lower with l-NAME versus adenosine, and there was a 35% reduction in inline image/time-integrated tension versus adenosine and 24% versusd-NAME that was independent of contraction frequency. l-NAME treatment produced a 33% sparing of muscle phosphocreatine (PCr), and intramuscular lactate was no different between groups. In contrast, d-NAME reduced force by 30%, inline image by 29% and the O2 cost of force development by 15% compared with adenosine, but had no effect on the degree of intramuscular ATP and PCr depletion. These results show that NOS inhibition improved the metabolic efficiency of force development, either by improving the ATP yield for a given O2 consumption or by reducing the ATP cost of force development. In addition, these effects were independent of contraction frequency.

Nitric oxide (NO) is a potent signalling molecule that has numerous physiological effects that could influence skeletal muscle contraction (Kobzik et al. 1994; Reid, 2001; Stamler & Meissner, 2001). Among these are effects on vascular smooth muscle (e.g. vasodilation: Brock et al. 1998; Hickner et al. 1997) excitation–contraction coupling (e.g. increased Ca2+ sensitivity and release: Andrade et al. 1998; Heunks et al. 2001a,b) and mitochondria (e.g. reversible inhibition of cytochrome oxidase: Brown, 1995; Giulivi, 1998; Giulivi et al. 1998; Brown & Borutaite, 1999). Consistent with these diverse sites of action for NO, the effects of NO synthase (NOS) inhibition on skeletal muscle aerobic metabolism can be quite diverse. Nitric oxide synthase inhibition has been shown both to accelerate oxygen uptake inline image kinetics over a range of exercise intensities in the horse (Kindig et al. 2000, 2002) and in humans (Jones et al. 2003, 2004; Wilkerson et al. 2004) and to increase O2 extraction in the horse (Kindig et al. 2000) and dog (Shen et al. 2000). In the horse, however, increased oxygen extraction withl-NAME could not prevent a decline in inline image at maximal workloads, probably because of the reduced O2 delivery consequent to NOS inhibition (Kindig et al. 2000). Several studies have shown that NOS inhibition with Nω-nitro-l-arginine methyl ester (l-NAME) attenuates muscle O2 consumption during contractions in humans (Hillig et al. 2003) and in perfused canine skeletal muscles (King-VanVlack et al. 2002), although a reduction in blood flow coincident with NOS inhibition in these studies complicates a straightforward interpretation (i.e. it is not possible to discern whether inline image is reduced consequent to an l-NAME-induced restriction in muscle O2 delivery). In contrast, two recent studies showed that the attenuating effect of l-NAME on inline image during constant-frequency high-intensity contractions exists even when rates of muscle convective O2 delivery were matched between groups (Grassi et al. 2005; Krause et al. 2005), showing that this effect is not secondary to impaired O2 availability. Consistent with this interpretation, our previous study also showed that although the O2 cost of force production was reduced by l-NAME, muscle lactate efflux was lowered by l-NAME treatment, suggesting that the reduced O2 cost of force development did not require a greater reliance on glycolytic metabolism (Krause et al. 2005).

A recent study by Grassi and colleagues showed that inhibition of NOS by l-NAME significantly reduced muscle fatigue development during 4 min of contractions at 60% peak inline image in the pump-perfused canine gastrocnemius muscle (Grassi et al. 2005). Nitric oxide synthase inhibition during this study did not produce a reduction in total inline image or speed inline image kinetics as predicted, despite an attempt to overcome the vasoconstrictive effects of l-NAME by pretreatment with acetylcholine and pump-controlled perfusion (Grassi et al. 2005). In contrast, this study (Grassi et al. 2005) reported a lower inline image (the difference between resting and steady-state inline image), similar to our previous report in pump-perfused rat distal hindlimb muscles (Krause et al. 2005), and a trend towards sparing of muscle phosphocreatine (PCr) and reduced substrate level phosphorylation during the maintained muscle force development, suggesting a better maintained metabolic adaptation during contractions in the presence of l-NAME. In light of these data, we aimed to examine whether the effects of NOS inhibition on muscle inline image and metabolic adaptation during pump-controlled blood flow conditions could be explained by potential non-NOS-related effects of l-NAME, using the additional control of the enantiomer Nω-nitro-d-arginine methyl ester (d-NAME). Furthermore, we aimed to establish whether the effects of l-NAME on muscle inline image were evident across a range of metabolic demands using an incremental intensity contraction protocol. We hypothesized that l-NAME would lower the O2 cost of force development over a range of contractile demands, whereas d-NAME would not, suggesting that this effect of l-NAME was consequent to NOS inhibition. Furthermore, we hypothesized that a lower O2 cost of force development with NOS inhibition would also reduce the requirements of high-energy phosphate (HEP) utilization to maintain the ATP turnover rate during contraction, and so we would observe a sparing of ATP and PCr with l-NAME. To meet these objectives, we used a pump-perfused rat hindlimb model to match muscle convective O2 delivery between three groups of animals: (i) 1 mml-NAME and 0.01 mm adenosine added to the perfusate; (ii) 1 mmd-NAME and 0.01 mm adenosine added to the perfusate; and (iii) 0.01 mm adenosine added to the perfusate (to act as a control group alongside d-NAME in these studies).

Methods

Animals

Adult male Sprague–Dawley rats (n= 6, l-NAME; n= 7, d-NAME; and n= 6, Adenosine) were obtained from Charles River and housed in pairs in the Heritage Medical Research Building's animal resource centre vivarium, under a 12 h–12 h light–dark cycle at 22°C and fed Purina rat chow and water ad libitum.

Surgical procedures

An in situ single perfused rat hindlimb preparation (Gorski et al. 1986; Hepple et al. 2002) was used to allow control of muscle blood flow during experiments. After anaesthetizing the animal with 70–75 mg kg−1 sodium pentobarbitone (i.p.), the left hindlimb was prepared by removing the skin and isolating the sciatic nerve. The sciatic nerve was then ligated and cut proximally for placement of a platinum nerve hook electrode connected to an electrical stimulator (Grass S48). As previously, the gluteal nerve was severed to avoid stimulation of the upper hindlimb musculature (Hepple et al. 2002). Following the sciatic nerve preparation, the Achilles tendon was severed with a portion of the calcaneus intact, and the gastrocnemius–plantaris–soleus muscle group was secured with 1.0 non-compliant silk thread in preparation for attachment to a force transducer (FT-10, Grass Instruments). The contralateral iliac artery and vein were ligated, and the right gastrocnemius–plantaris–soleus muscle group was excised, trimmed of fat, and weighed. Following surgery, the left hindlimb was secured to an aluminium baseplate by a bone clamp placed around the proximal femur and a cable tie placed around the ankle, to minimize movement artifact during muscle contractions. Catheters (22 gauge in the artery and 20 gauge in the vein) were then inserted in the left iliac artery and vein and advanced into the femoral artery and vein, respectively, to initiate flow to and from the hindlimb. Once perfusion of the hindlimb was initiated (prior to contractions), the animal was killed with an intracardiac injection of 25 mg sodium pentobarbitone. All exposed tissues of the experimental hindlimb were wrapped in warm saline-soaked gauze, Saran wrap and aluminium foil to avoid moisture and heat loss throughout the experiment. Muscle and perfusate temperatures were maintained at 37°C.

Perfusion medium

Bovine whole blood was collected weekly from a local abattoir. The erythrocytes were ‘washed’ in three changes of Krebs–Henseleit buffer by centrifugation at 5000g, with aspiration of the supernatant and buffy coat between washes. The washed erythrocytes were then stored at 4°C in Krebs–Henseleit buffer containing 5 mm glucose and used within 3 days of collection. The standard perfusion medium for all three groups consisted of a Krebs–Henseleit bicarbonate buffer (pH 7.4) containing 4.5% bovine serum albumin (Sigma Chemicals; dialysed 48 h), bovine erythrocytes (haematocrit 43%, verified by direct observation in centrifuged capillary tubes), 5 mm glucose, 100 mU ml−1 insulin, 1000 mU ml−1 heparin and 0.15 mm pyruvate.

Experimental protocol

Animals were divided into three groups. Note that we have previously observed that it is necessary to add adenosine to the perfusate to obtain adequate vasodilatation in the pump-perfused rat hindlimb when using l-NAME (Krause et al. 2005); thus, adenosine was a component of the perfusion medium in each of these groups. In one group, Nω-nitro-l-arginine methyl ester (l-NAME, 1 mm; Sigma Chemicals) and adenosine (0.01 mm; Sigma Chemicals) were added to the perfusate (l-NAME group, n= 6). In a second group, the ‘inactive’ enantiomer of l-NAME (Nω-nitro-d-arginine methyl ester, d-NAME, 1 mm; Sigma Chemicals) and adenosine (0.01 mm; Sigma Chemicals) were added to the perfusate (d-NAME control group, n= 7). The third group received adenosine only (0.01 mm; Sigma Chemcials; adenosine group, n= 6), and also acted as a control group for l-NAME and d-NAME. This concentration of l-NAME was chosen because a similar dose used previously demonstrated successful inhibition of NOS (King-VanVlack et al. 2002). Similarly, previous studies in the horse (Kindig et al. 2000, 2002) using an oral dosing regime yielded approximately the same blood l-NAME concentration when blood volume was estimated from total body mass in these experiments. The concentration of adenosine was chosen as that which produced a similar degree of vasodilation in the l-NAME-treated group to that seen under control conditions previously (Krause et al. 2005).

As described in our previous study (Krause et al. 2005), the perfusate was heated to 37°C and equilibrated with 95% O2–5% CO2 through 7 m of gas-permeable tubing contained in a flask, prior to entering the hindlimb. This yielded an average arterial O2 content of 20.1 ± 0.2% by volume. Flow was controlled using a peristaltic pump (Gilson Miniplus 3), where the rate of perfusion was confirmed following each experiment by timed collection in a graduated cylinder. As in our previous experiments (Krause et al. 2005), an incubation period of 30 min was standardized between groups prior to initiating contractions. During this period, the rate of perfusion was gradually increased to the level desired during contractions and was selected to permit matching of muscle mass-specific blood flow between experimental groups.

Muscle length and stimulation voltage (∼7 V) were adjusted to yield maximal tension. The gastrocnemius–plantaris–soleus muscle complex was stimulated to contract (200 ms trains, with pulses at 100 Hz, each pulse being of 0.05 ms duration) for 1 min each at 15, 30 and 60 tetani min−1 and for 2 min at 90 tetani min−1. Blood samples were drawn anaerobically from the arterial catheter prior to contractions and from venous effluent every 30 s during contractions, and analysed for [haemoglobin], O2 saturation inline image, partial pressures of O2 inline image and CO2 inline image, and [lactate] using a blood gas analyser (Nova Biomedical Stat Profile M). Arterial and venous O2 content was determined using the formula:

display math

Oxygen uptake inline image across the hindlimb was calculated by multiplying the arteriovenous oxygen content difference by the blood flow. Similarly, lactate efflux across the hindlimb was calculated as the product of blood flow and the arteriovenous difference in [lactate]. The time delay in erythrocyte movement from the hindlimb vasculature to the venous sampling port was measured as done previously (Hepple et al. 2003), and this time difference was used to assess force production at the time when blood O2 content was measured (thus allowing inline image per Newton force to be obtained). Time-integrated tension was calculated for each blood sampling point as done previously (Krause et al. 2005).

Muscle metabolite analyses

Immediately following the incremental frequency contraction bout, the gastrocnemius–soleus–plantaris muscle complex was freeze clamped in situ and stored at −70°C until further analysis. Frozen muscle samples were subsequently pulverized under liquid nitrogen using a pestle and mortar to mix all the fibres within the gastrocnemius–soleus–plantaris complex. This procedure did not allow the determination of differences in fibre type-specific changes in HEP content with the different treatments. The data therefore represent a ‘snap-shot’ of the metabolic environment across these muscles following the fatiguing contractions and different treatment interventions. A portion of this crushed and frozen muscle was freeze dried, dissected free of blood and connective tissue and powdered in a percussion pestle and mortar. Powdered aliquots of muscle tissue were then extracted according to the method of Harris et al. (1974). Briefly, 6–10 mg of powdered muscle was weighed out into a 1.5 ml microcentrifuge tube. Muscle metabolites were then extracted from the powdered tissue using 0.5 m ice-cold perchloric acid containing 1 mm EDTA and frequent vortexing for 10 min (in between vortexing the samples were kept on ice). Proteins and unwanted cell debris were removed by centrifugation at 22 000g for 3 min at 4°C, and the resulting supernatant was neutralized with 2.2 m KHCO3 for at least 15 min. The solution was again centrifuged at 22 000g for 3 min at 4°C, and the supernatant (extract) removed and frozen at −70°C until further analysis of muscle adenosine triphosphate (ATP), phosphocreatine (PCr), creatine (Cr) and lactate concentrations. These metabolites were assayed by measuring the reduction or oxidation of NAD and NADH, respectively, via enzyme- and substrate-loaded reactions by spectrophotometry (Harris et al. 1974).

Muscle glycogen was determined from an aliquot of freeze-dried and powdered muscle. Appoximately 2–3 mg of muscle powder was solubilized in 120 μl of 0.1 m NaOH at 80°C for 10 min, and then neutralized with 480 μl of buffer (0.1 m HCl and 0.2 m citric acid with 0.2 m Na2HPO4, pH 5.0, mixed in a ratio of 1:3). Glycogen debranching enzyme, amyloglucosidase (AGG; Roche diagnostics, UK), was then added (15 μl) to the neutralized solution and incubated at room temperature for 1 h, yielding glucosyl units from glycogen hydrolysis. Samples were then centrifuged at 22 000g for 3 min, and the supernatant (extract) removed and stored in screw-cap microcentrifuge tubes at −70°C for subsequent determination of glucosyl unit concentration. The glycogen extract was assayed for glucosyl unit content by measuring the reduction of NAD via an enzyme- and substrate-loaded reaction by spectrophotometry (Harris et al. 1974).

Blood flow distribution

Blood flow distribution to the gastrocnemius–plantaris–soleus muscle group was estimated in the l-NAME and adenosine groups, based upon the previously published results obtained from our laboratory (Krause et al. 2005). Briefly, this work showed that the gastrocnemius–plantaris–soleus muscle group received 15.6% of the total hindlimb blood flow, irrespective of treatment withl-NAME and/or adenosine. Thus, mass-specific blood flow to the gastrocnemius–plantaris–soleus muscle group was estimated as the product of total hindlimb blood flow and 15.6%, normalized to the mass of this muscle group. To confirm our prediction and take account of any potential alterations in blood flow distribution induced by d-NAME, we studied a separate group of animals (n= 2) in which 1 mm of d-NAME and 0.01 mm adenosine was added to the perfusate. Blood flow was gradually increased until the desired rate of flow was reached, and subsequently ∼290 000 coloured microspheres were infused, according to the methods previously described (Hepple et al. 2002). Since these experiments showed that d-NAME did not alter blood flow distribution (see Results) from that observed with the other treatments in our previous study (Krause et al. 2005), mass-specific blood flow to the gastrocnemius–plantaris–soleus muscle group was estimated in the d-NAME group in the same manner described above for the other groups (i.e. the proportion of blood flow to the gastrocnemius–plantaris–soleus muscle group is not significantly altered by any of these treatment interventions).

Data analysis

Values are presented as means ±s.e.m. A one-way ANOVA with a Student–Newman–Keuls post hoc test was employed to analyse differences in body mass, muscle mass, perfusion conditions, peak tension and the intramuscular metabolite concentrations between treatment groups. Two-way ANOVAs (treatment × time) with a Student–Newman–Keuls post hoc test were employed to detect significant differences between adenosine versus treatment groups for measurements made during the contraction bout.

Results

Descriptive data and perfusion conditions

No differences were observed in body mass or muscle mass between groups (Table 1). In a separate set of animals (n= 2), d-NAME did not alter muscle blood flow distribution (gastrocnemius, 49 ± 1 ml min−1 (100 g)−1; plantaris, 91 ± 8 ml min−1 (100 g)−1; and soleus, 142 ±3 ml min−1 (100 g)−1) from that observed with the other treatments seen previously (adenosine: gastrocnemius, 54 ± 1 ml min−1 (100 g)−1; plantaris, 103 ± 12 ml min−1(100 g)−1; soleus, 168 ± 17 ml min−1 (100 g)−1; l-NAME: gastrocnemius, 49 ± 2 ml min−1 (100 g)−1; plantaris, 87 ± 2 ml min−1 (100 g)−1; and soleus, 168 ± 14 ml min−1 100 g−1; Krause et al. 2005). On this basis, there were no differences in the estimated muscle mass-specific blood flow or convective O2 inline image delivery to the gastrocnemius–plantaris–soleus muscle group between any of the treatments (Table 2).

Table 1.  Descriptive data
  l-NAME (n= 6) d-NAME (n= 7)Adenosine (n= 6)
  1. Values are means ±s.e.m. GPS, gastrocnemius–plantaris–soleus muscle group.

Body mass (g)411.83 ± 19.4403.85 ± 11.0426.85 ± 21.3
GPS mass (g)  2.66 ± 98.1  2.58 ± 95.2  2.75 ± 64.5
Total contracting mass (g) 4.58 ± 0.1 4.57 ± 0.2 4.89 ± 0.1
Total perfused mass (g)15.79 ± 0.415.76 ± 0.716.87 ± 0.5
Table 2.  Perfusion conditions
  l-NAME (n= 6) d-NAME (n= 7)Adenosine (n= 6)
  1. Values are means ±s.e.m. GPS, gastrocnemius–plantaris–soleus muscle group; inline image, oxygen delivery. *P < 0.05 versusd-NAME.

Total pressure (mmHg)148 ± 6144 ± 7131 ± 6
Net pressure (mmHg)87 ± 491 ± 2 75 ± 4*
Blood flow (ml min−1)10.3 ± 0.210.2 ± 0.110.6 ± 0.2
GPS blood flow (ml min−1 (100 g)−1)61 ± 262 ± 261 ± 2
GPS inline image (μmol min−1 (100 g)−1)465 ± 11499 ± 17492 ± 19
Resting inline image(μmol min−1 (100 g)−1)23.8 ± 2.825.0 ± 3.026.3 ± 1.3

Contractile and metabolic responses

Initial (peak) tension (Fig. 1A) was not different between l-NAME (13.4 ± 0.8 N g−1) and adenosine only groups (12.2 ± 0.3 N g−1; P= 0.175). However, there was a significantly lower peak tension in d-NAME (10.9 ± 0.6 N g−1) versusl-NAME (P < 0.01). Time-integrated tension was significantly lower in the d-NAME group (average over the 5 min contraction bout, 0.549 ± 0.017 N s−1 g−1) versus the other treatments (P < 0.05). Furthermore, time-integrated tension with l-NAME (0.788 ± 0.019 N s−1 g−1) was significantly higher than both adenosine only (0.664 ± 0.019 N s−1 g−1) and d-NAME groups (P < 0.05), and these differences were independent of contraction intensity (treatment–time interaction, P > 0.05; Fig. 1B).

Figure 1.

Tension (A) and time-integrated tension (B) as a function of time and contraction frequency

In this and all subsequent figures, values are means ±s.e.m. There was a significant main effect for a lower tension development in d-NAME (n= 7) versus adenosine (n= 6) and a lower time-integrated tension in both adenosine and d-NAME versus the l-NAME group (n= 6; P < 0.05).

Muscle mass-specific resting inline image was not different between treatments (Table 2). However, during contractions inline image was statistically lower in both l-NAME (average over the 5 min contraction bout, 187 ± 7 μmol min−1 (100 g)−1) and d-NAME groups (168 ± 7 μmol min−1 (100 g)−1) versus adenosine only (236 ± 7 μmol min−1 (100 g)−1; P < 0.05), and this effect was independent of contraction frequency (Fig. 2). The O2 cost of force development, estimated as the quotient of inline image and time-integrated tension at each time interval during which blood was drawn, is shown in Fig. 3. The O2 cost of force development was significantly higher in the adenosine only group (average over the 5 min contraction bout, 110 ± 3 nmol N−1) versus the other treatment groups (P < 0.05), and l-NAME (71 ± 3 nmol N−1) was significantly lower than both adenosine only and d-NAME groups (93 ± 3 nmol N−1; P < 0.05; i.e. l-NAME < d-NAME < adenosine). There was a significant main effect difference (P < 0.05) in muscle mass-specific lactate efflux, where l-NAME and d-NAME were lower than adenosine only (Fig. 4). None of these differences were dependent upon contraction intensity (treatment–time interaction, P > 0.05).

Figure 2.

inline image as a function of time and contraction frequency

There was a significant main effect for a lower inline image (no interaction with time) in l-NAME (n= 6) and d-NAME (n= 7) versus the adenosine group (n= 6).

Figure 3.

Estimated oxygen cost of force development as a function of time and contraction frequency

There was a significant main effect (P < 0.05) for a lower oxygen cost of force development (no interaction with time) between all groups, where l-NAME (n= 6) was lower than both adenosine (n= 6) and d-NAME (n= 7; i.e. l-NAME < d-NAME < adenosine).

Figure 4.

Lactate efflux as a function of time and contraction frequency

There was a significant main effect for a lower lactate efflux (no interaction with time) for l-NAME (n= 6) and d-NAME (n= 7) versus the adenosine group (n= 6; P < 0.05).

Muscle metabolite concentrations

No difference in muscle ATP concentration was observed between treatments (P= 0.106; Fig. 5); however, a significant sparing of muscle PCr was evident in the l-NAME group compared with d-NAME (P < 0.01) and was close to being significant compared to adenosine (P= 0.057). PCr concentration at the end of the entire fatiguing contraction period was 33% higher than the adenosine only group and 54% higher than the d-NAME group (Fig. 5). The sum of ATP and PCr concentrations for each treatment group, used as an index of total high-energy phosphate (HEP) pool, revealed a higher total high-energy phosphate content in l-NAME compared with d-NAME (P < 0.01) and adenosine alone (P < 0.05). Muscle total creatine was not different between treatment groups (data not shown). No difference in muscle lactate concentration was evident, despite the lower lactate efflux with l-NAME and d-NAME compared with adenosine (P < 0.05, no interaction with time) during contractions (Fig. 3).

Figure 5.

Muscle metabolite concentrations immediately following the incremental frequency contraction bout

*P < 0.05 significantly different from d-NAME (n= 7) and adenosine (n= 6); †P < 0.05 significantly different only from l-NAME (n= 6).

Discussion

The purpose of the present study was threefold. Firstly, we aimed to test the hypothesis that NOS inhibition with l-NAME would lower the O2 cost of force development over a range of contractile demands. Secondly, we hypothesized that high-energy phosphate (HEP) would be spared if the O2 cost of force development was lower. Thirdly, we investigated the potential non-NOS-related effects of l-NAME on the functional and metabolic responses in the pump-perfused rat hindlimb model by using the additional control of the ‘inactive’ enantiomer d-NAME in our study. The present results support those from our previous study (Krause et al. 2005), showing that NOS inhibition via l-NAME reduces the quotient of inline image and time-integrated tension, an effect consistent with a reduction in the O2 cost of force development. The present study extends these findings to show that this effect is evident across a range of contraction intensities. However, an important finding of the present study is that muscle ATP + PCr (high-energy phosphate; HEP) was spared with l-NAME treatment, in conjunction with a reduction in O2 cost of force development. This demonstrates that the lower O2 cost of contractions is not associated with an acceleration of HEP utilization to maintain ATP production during force development, but in fact demonstrates that HEP sparing is consistent with an improved contractile economy by l-NAME. A further interesting finding from this study is related to our results obtained with d-NAME treatment. We have shown that the ‘inactive’ enantiomer of the NOS inhibitor l-NAME induced contractile suppression (e.g. reduced peak tension and tension development throughout the contraction bout) and, coupled with this, the O2 cost of force development was lower when compared with adenosine alone (although this latter effect was less than half the magnitude of that observed with l-NAME). Thus, contrary to its purported role as an ‘inactive’ enantiomer of l-NAME, not only does d-NAME suppress muscle force development, but it also produces a reduction in inline image and the O2 cost of force development compared with adenosine alone, but this latter effect was of smaller magnitude than that observed with l-NAME. In addition, d-NAME was not associated with the sparing of HEP following contractions that was seen with l-NAME. These observations are worth some consideration when interpreting data in the literature and designing future studies using l-NAME, since these findings suggest that l-NAME could modulate some of its effects through a pathway unrelated to NOS inhibition. Nonetheless, our results show that the majority of the reduction in O2 cost of force development by l-NAME is caused by NOS inhibition, and is present across a range of contractile demands.

Implications of the effect of d-NAME on force production

There was a significant main effect for a higher mean force over the entire contraction bout in the l-NAME group than with adenosine alone. This effect may be explained by the actions of adenosine, which has been shown to induce production of NO via activation of the A2 receptor (Li et al. 1995), such that the augmentation of muscle function with NOS inhibition may reflect antagonism of contractile depression mediated through NO production (Reid et al. 1998) secondary to A2 receptor activation by adenosine. Interestingly, in accounting for potential effects of l-NAME that are unrelated to NOS inhibition, we observed a reduction in muscle contractility by d-NAME. Specifically, d-NAME reduced initial peak force by 10% (not significant, P= 0.129) and 19% (P < 0.01) versus adenosine only and l-NAME, respectively, and reduced force output across the entire contraction bout by an average of 30% versusl-NAME and 17% versus adenosine. In addition, d-NAME reduced inline image (Fig. 2) and lactate efflux (Fig. 4) compared with adenosine, but these data were not statistically different compared with l-NAME. The mechanism of the force suppression observed with d-NAME cannot be determined from the present studies, but underscores the importance of accounting for potential non-specific actions of l-NAME. Indeed, this suppression of force by d-NAME may help to explain the lack of increase in peak inline image during intense muscle contractions by l-NAME versus control conditions in both the present study and our previous study (Krause et al. 2005). Specifically, l-NAME may be producing opposing effects on muscle contractile function such that the effect of NOS inhibition (which would tend to augment contractile function) may be counteracted by a non-specific suppression of contractile function, which in turn prevents inline image from meeting or exceeding that observed under control conditions by depressing contractile ATP demand.

Effect of l-NAME on contractile economy

Similar to the results reported in our previous study (Krause et al. 2005), l-NAME enhanced force production but reduced inline image during electrically evoked muscle contractions versus adenosine (despite matching muscle convective O2 delivery between groups), such that the quotient of inline image and time-integrated tension was reduced relative to adenosine. On this basis, these results indicate a reduced O2 cost of force development with l-NAME. The present results extend our previous observations (Krause et al. 2005) to show that this effect is present over a wide range of contraction intensity. These data are contradictory to a previous study which showed that inline image was higher with the NOS inhibitor nitro-l-arginine during walking and running exercise in the dog (Shen et al. 2000). However, much of the increase in inline image during running already existed at rest and during walking in this previous study (Shen et al. 2000). In the present study, we did not see any difference in muscle inline image at rest between groups and observed a decrease in inline image during contractions, a finding also supported by previous studies using l-NAME to inhibit NOS (King-VanVlack et al. 2002; Hillig et al. 2003; Krause et al. 2005). The possibility that these disparities between studies are caused by differences in the NOS inhibitor employed bears consideration, and suggests that future studies should examine multiple NOS inhibitors to address this point.

The present study also extends our present knowledge of the basis for the lower inline image during contractions when using l-NAME to inhibit NOS, in that we now know that HEP sparing occurs in parallel with the reduced O2 cost of force development with l-NAME. This conclusion is predicated upon the previous finding that l-NAME does not alter resting muscle ATP and PCr concentrations (Grassi et al. 2005), and so the higher HEP levels in l-NAME at the end of the contraction bout can be interpreted as a lower net HEP utilization in this group. The parallel reduction of HEP in conjunction with the reduction of inline image seen with l-NAME in the present study may be explained by the inversely proportional relationship between the amount of PCr utilization during contractions and inline image (Mahler, 1985; Connett, 1988; Meyer, 1988). Thus, sparing of HEP would occur under conditions where a smaller stimulation of aerobic metabolism is required.

In light of these findings with regard to HEP sparing with l-NAME, we grapple with two distinct explanations. Firstly, one may consider HEP sparing to be representative of improved kinetics of oxidative metabolism by NO removal (NO binds to complex IV and reversibly inhibits mitochondrial respiration (Brown, 1995; Giulivi et al. 1998)), thus rendering a better coupling of the electron transport chain to allow greater resynthesis of ATP at a given inline image. As such, this would reduce the requirement of PCr utilization to maintain ATP turnover and so could explain our observation of HEP sparing with l-NAME. However, a recent study demonstrated a slightly lower PCr utilization and substrate level phosphorylation during contraction in canine gastrocnemius muscle treated with l-NAME, despite a reduced fatigue development and no effect on inline image and inline image kinetics during contraction compared with control conditions (Grassi et al. 2005). This suggests that improved kinetics of oxidative metabolism as an explanation of HEP sparing with NOS inhibition may not be justified; however, it warrants further verification. In addition, the aforementioned canine study measured HEP during contractions and the ‘fundamental’ kinetic phase (Grassi et al. 2005), whereas in the present study we report HEP sparing from samples taken at the end of the contraction bout and when the muscle was highly fatigued. Thus, interpolating the HEP sparing with inline image kinetics is not possible in this study, and so our interpretations of the HEP sparing can only be considered as an outcome which could be reflective of the contraction protocol employed here (i.e. HEP sparing may not have occurred across the whole range of contractile demands). We therefore suggest that further analysis of the HEP sparing effect at each contraction intensity with l-NAME during the reduction in inline image as observed in our model is warranted to dissect the importance of these events on muscle function more accurately.

A second explanation of the HEP sparing in the present study is that the ATP cost per contraction cycle, possibly at the level of the actin–myosin cross-bridges, is altered by NO removal. This idea is consistent with previous data from studies in rat diaphragm muscle fibres subjected to both NO donors and NOS inhibitors (Reid et al. 1998). An interesting observation from this in vitro contraction study was that time-to-peak tension was prolonged with the NOS inhibitor 7-nitroindazole (1 mm), and a trend for a prolonged half-relaxation time was also apparent (Reid et al. 1998). On this basis we would expect a lower ATP utilization for a given level of force production consequent to a lower rate of cross-bridge attachment–detachment cycling. This scheme implies that the ATP cost of contraction may be altered with NOS inhibition; a point that should be considered in future investigations.

Conclusion

In summary, our results show that: (i) the reduced O2 cost of force development seen with l-NAME is caused primarily by its effects on NOS inhibition; (ii) this effect is produced over a range of contractile demands; (iii) this reduction in O2 cost of force development was accompanied by a similar sparing of HEP; and (iv) this effect could be either caused by a higher ATP yield per unit of O2 consumed and/or that the ATP cost of excitation–contraction coupling is reduced. Unexpectedly, we also found that d-NAME caused a significant suppression of skeletal muscle contractile function and a small reduction in O2 cost of force development (less than half of that seen with l-NAME). This suggests that the real magnitude of the l-NAME potential for improving contractile function may be partly masked by non-specific, d-NAME-like actions of this NOS inhibitor.

Appendix

Acknowledgements

This study was supported by operating grants from the Canadian Institutes of Health Research (MOP 48185) and NSERC (RPG238805). Dr R. T. Hepple is a Canadian Institutes of Health Research Institute of Ageing New Investigator. R. A. Howlett was a Parker B. Francis fellow, and also funded by NIH AR40155 grant.

Ancillary