The aim of the present study was to determine the effect of nitric oxide and prostanoids on microcirculation and oxygen uptake, specifically in the active skeletal muscle by use of positron emission tomography (PET). Healthy males performed three 5-min bouts of light knee-extensor exercise. Skeletal muscle blood flow and oxygen uptake were measured at rest and during the exercise using PET with H2O15 and 15O2 during: 1) control conditions; 2) nitric oxide synthase (NOS) inhibition by arterial infusion of NG-monomethyl-l-arginine (l-NMMA), and 3) combined NOS and cyclooxygenase (COX) inhibition by arterial infusion of l-NMMA and indomethacin. At rest, inhibition of NOS alone and in combination with indomethacin reduced (P < 0.05) muscle blood flow. NOS inhibition increased (P < 0.05) limb oxygen extraction fraction (OEF) more than the reduction in muscle blood flow, resulting in an ∼20% increase (P < 0.05) in resting muscle oxygen consumption. During exercise, muscle blood flow and oxygen uptake were not altered with NOS inhibition, whereas muscle OEF was increased (P < 0.05). NOS and COX inhibition reduced (P < 0.05) blood flow in working quadriceps femoris muscle by 13%, whereas muscle OEF and oxygen uptake were enhanced by 51 and 30%, respectively. In conclusion, by specifically measuring blood flow and oxygen uptake by the use of PET instead of whole limb measurements, the present study shows for the first time in humans that inhibition of NO formation enhances resting muscle oxygen uptake and that combined inhibition of NOS and COX during exercise increases muscle oxygen uptake.
- skeletal muscle
- blood flow
- oxygen consumption
- nitric oxide
nitric oxide (NO) is involved in a host of signaling and regulatory pathways of mammals. Its role as an important tonic regulator of baseline vessel tone (37) and blood pressure (28) is well established. Additionally, exercise hyperemia also depends on NO since it, in synergy with other compounds, regulates blood flow of the working limb (2, 15, 19, 23, 30, 31). NO also plays a role in the regulation of muscle metabolism. By use of in vitro preparations, it has been demonstrated that the primary effect of exogenous NO on mitochondrial activity is a reversible and competitive inhibition of cytochrome oxidase activity (5, 6, 32). Some animal studies have subsequently found evidence that NO tonically inhibits mitochondrial respiration in vivo (33–35), but there has not been evidence for this in humans (10, 27, 31). It is well shown with various methods that prostanoids can act synergistically with NO to regulate vascular function in health and disease (2, 19, 23, 30, 31), but COX inhibition may also have direct effects on cellular aerobic respiration by affecting uncoupling (17, 20, 23).
Positron emission tomography (PET) is a noninvasive imaging method based on short-lived radioisotopes that can be applied to measure blood flow and its distribution in muscle. Muscle blood flow by PET can be quantified with [15O]H2O (radiowater), which is an intravenously infused inert and freely diffusible tracer. PET radiowater measures only blood flow in vessels where there is exchange of water molecules, i.e., where exchange of nutrients and oxygen occurs. In addition to the unique possibility to have direct three-dimensional insight to capillary level blood flow in resting and working human skeletal muscle with [15O]H2O tracer, it is also possible to directly measure exercising muscle oxygen consumption (V̇o2) and extraction with bolus-inhaled 15O2 tracer (26), enabling the specific determination of local muscle oxygen uptake. Moreover, PET allows for determination of blood flow and oxygen uptake specifically in the muscle, without contribution from other inactive tissues such as skin, fat, and resting muscle.
The hypothesis of this study was that use of the above-described PET technology would provide a more sensitive assessment of the effect of NO synthase (NOS) and cyclooxygenase (COX) inhibition compared with previous studies measuring whole leg blood flow and oxygen uptake (10, 27, 31), and this study shows for the first time in humans that NO blockade enhances resting oxygen uptake and that combined inhibition of NOS and COX increases muscle oxygen uptake during exercise.
Eight healthy young men (26 ± 2 yr, 184 ± 4 cm, 82 ± 8 kg) volunteered to participate in this study. The purpose, nature, and potential risks of the study were explained to the subjects before they gave their written informed consent to participate. The subjects were not taking any regular medication. The study was performed ∼3 h after the subjects had eaten their normal breakfast except that they abstained from caffeine-containing beverages for at least 24 h before the experiments. The subjects were also requested to avoid strenuous exercise within 48 h before the study. The study was performed according to the Declaration of Helsinki and was approved by the Ethical Committee of the Hospital District of South-Western Finland and the National Agency for Medicines.
Before the PET experiments, the antecubital vein was cannulated for tracer administration. For blood sampling, a radial artery cannula was placed under local anesthesia in the contralateral arm. Additionally, cannulas were placed under local anesthesia in the femoral artery and vein for local drug infusions and blood sampling, respectively. Subjects were then moved to the PET scanner with the femoral region in the gantry, and the right leg was fastened to an in-house-designed leg work dynamometer. PET measurements were first performed at baseline and thereafter during exercise without any drug infusions. Thirty minutes later, resting and exercising measurements were performed during NOS blockade with NG-monomethyl-l-arginine (l-NMMA), and finally, after a 30-min recovery period, both resting and exercising measurements were again repeated under synergistic inhibition of l-NMMA and indomethacin, which blocks COX enzymes. Additionally, radial artery and femoral vein blood samples for energy substrate and blood gas parameters were drawn for analysis during each of the conditions described above. Systemic mean arterial pressure was measured (Omron, M5–1; Omron Healthcare, Hoofddorf, The Netherlands) on every occasion studied. To inhibit NO synthases, l-NMMA (Clinalfa, Laufelfingen, Switzerland) was infused intra-arterially with a concentration of 1.0 mg·min−1·kg leg mass−1 and to inhibit cyclooxygenases, indomethacin (Confortid, Alphapharma, Denmark) was infused with a concentration of 50 μg·min−1·kg leg mass−1 (23).
Blood flow measurements and analysis.
Radiowater positron-emitting tracer [15O]H2O was produced as previously described (36), and the ECAT EXACT HR+ scanner (Siemens/CTI, Knoxville, TN) was used in three-dimensional mode for image acquisition to measure muscle blood flow. The oxygen-15 isotope was produced with a Cyclone 3 cyclotron (IBA Molecular).
Radiowater, [15O]H2O, was subsequently produced in a continuously working water module (Radiowater Generator, Hidex Oy, Finland). [15O]O2 gas and H2 gas can be processed in an oven at 700°C to water vapor, and, with the use of a diffusion membrane technique, radioactive water vapor can be trapped in the sterile saline. Radiowater was administered automatically as a 12- to 15-s bolus (10 ml) injection in the patient.
To ensure that the target gas, N2/O2 (99%/1%), for application of the [15O]O2 gas was suitable for use, a flow-through purifier containing active carbon and soda lime was used to remove ozone and nitrogen oxides. Gas chromatographic analysis was performed to verify the product and purity of the product before each study. The radiochemical purity of [15O]O2 gas was 97%.
Photon attenuation was corrected by 5-min transmission scans performed at the beginning of the PET measurements performed at rest and during exercise. All data were corrected for dead time, decay, and measured photon attenuation, and the images were reconstructed into a 256 × 256 matrix, producing 2.57 × 2.57 mm in-plane dimensions of voxels with 2.43-mm plane thickness. For the measurement of blood flow at rest, scanning began simultaneously with the infusion and consisted of the following frames: 6 × 5 s, 12 × 10 s, and 7 × 30 s at rest and 6 × 5 s and 12 × 10 s during exercise. During exercise, scanning was started 5 min after exercise onset to obtain a metabolic steady-state situation and continued until the end of the exercise bout, e.g., 2.5 min. Arterial blood radioactivity was also sampled continuously with a detector during imaging for blood flow quantification. Exercise consisted of dynamic one-legged exercise at 40 rpm with an average work load of 4.5 kg with a knee angle range of motion of ∼75–80 degrees (13). Local muscle blood flow was measured from the whole thigh muscles of the right leg, and specifically from the quadriceps femoris (QF) and posterior muscles of the right thigh. The data analysis was performed using the standard models (16) and methods (29). Heterogeneity of blood flow (relative dispersion) was calculated as coefficient of variation (SD divided by mean blood flow) of voxel blood flow values within the region of interest as described earlier (14).
Direct muscle V̇o2 and extraction fraction analysis. Muscle V̇o2 was measured directly from the exercising muscle by PET and 15O-labeled oxygen according to previous studies (18, 26). Briefly, [15O]O2 gas was administered into the patient as a bolus. Radioactive [15O]O2 gas was collected in a rubber bellows with a total volume of 900 ml. The rubber bellows is equipped with a radioactivity detector and remote-controlled membrane valves. Total radioactivity of the gas bolus can be measured, and the patient will inhale the remotely controlled radioactive gas bolus diluted with room air. After 5 min of steady-state exercise, 15O2 was inhaled, and tissue and arterial time-activity curves were collected. Muscle oxygen extraction and consumption were then quantitated by nonlinear fitting from the 15O2 data and by including a separate compartment for free and myoglobin-bound oxygen fitting (26).
With the 15O2 inhalation, also muscle oxygen extraction fraction (OEF) can be quantified irrespectively of blood flow. Direct muscle V̇o2 measurements were also performed at rest, but these analyses were not possible to calculate because of significant oxygen binding to myoglobin and significantly reduced oxygen supply during NOS and COX enzyme inhibitions. Therefore, V̇o2 was calculated by multiplication of limb oxygen extraction with muscle blood flow, which provides a good estimation for muscle V̇o2 at rest. This Fick's principle was also used for method comparison to direct 15O2 determination-analyzed V̇o2. To calculate exercising muscle V̇o2 by Fick's principle, blood flow obtained from QF muscle was used. This was reasoned to be the best estimate for V̇o2, since it is assumed that the knee extensions are performed almost solely by QF and that oxygen extraction is higher in exercising QF than is the value obtained over the leg, still recognizing that the latter value represents the oxygen extraction of the whole leg.
Magnetic resonance imaging. Structural magnetic resonance imaging (MRI) was performed ∼1 wk before the PET study as described earlier (12), when subjects were also accustomed to the one-leg knee extension exercise model in a PET scanner. MRI scanning was performed to get total leg volume of the working leg, since NOS- and COX-inhibiting drug infusions were based on effective concentrations per liter leg volume (23). The mean total leg volume of the subjects was 12.2 ± 1.5 liters.
Blood parameter analysis.
Blood samples for energy substrate lactate and blood oxygen analysis were drawn in each study condition in the middle of PET measurements from the femoral vein and radial artery and analyzed with standardized hospital practices. Briefly, lactate was analyzed with enzymatic methods (Roche Modular P analyzer; Roche Diagnostics, Mannheim, Germany), and blood oxygen was analyzed with a Radiometer ABL 835 blood gas analyzer.
Statistical analyses were performed with SAS 8.2 and SAS Enterprise 4.2 programs (SAS Institute, Cary, NC). Statistical analyses were performed using two-way ANOVA for repeated measures. If a significant main effect(s) was found, pairwise differences were identified using the Tukey-Kramer post hoc procedure. Results are expressed as means ± SD, and P ≤ 0.05 was considered statistically significant.
Blood flow and oxygen uptake at rest.
At rest, thigh muscle blood flow was lower (P < 0.05) during both NOS inhibition and combined NO and COX inhibition than during the control condition (Figs. 1 and 2A). Limb OEF was higher (P < 0.05) during both of these conditions than in control (Fig. 2B). During NOS inhibition, but not during combined NOS and COX inhibition, muscle V̇o2 was 20% higher (P < 0.05) than during the control condition (Fig. 2C). NOS and COX inhibition (P < 0.05) increased the heterogeneity of resting muscle blood flow (Figs. 1 and 3). Mean arterial blood pressure levels were higher (P < 0.05) during NOS blockade alone and during combined NOS and COX inhibition (Table 1).
Blood flow and oxygen uptake during exercise.
During exercise, NOS blockade did not affect working QF muscle blood flow (Figs. 1 and 4A), but limb OEF was higher (P < 0.05; Fig. 5A). Combined inhibition of NOS and COX resulted in a lower (P < 0.05) QF blood flow (Fig. 1) and higher (P < 0.05) OEF (Fig. 5A), but muscle V̇o2 determined by Fick's principle was not statistically different from the control condition during NOS or NOS + COX inhibition (Fig. 5B). Direct oxygen-15-determined muscle OEF and V̇o2 analysis, however, revealed that OEF of the working QF muscle was higher (P < 0.05) during NOS inhibition alone than during control and higher (P < 0.05) during combined NOS and COX blockade than during NOS inhibition alone (Fig. 5A). As a consequence, working QF muscle V̇o2 tended to be increased during NOS inhibition (P = 0.07) and was significantly enhanced when both NOS and COX were inhibited (Fig. 5B). The robustness of these findings is also illustrated in Fig. 6, where the utilization of oxygen in these three different study conditions can be judged from authentic oxygen-15 inhalation tissue time-activity curves. Compared with overall limb OEF values, direct muscle OEF was not different between control or NOS inhibition but was significantly higher during double inhibition (P = 0.001) (Fig. 5A). There were no differences between the V̇o2 determination methods regarding muscle V̇o2 (P = 0.2), and muscle V̇o2 values determined by Fick's principle correlated well with the direct V̇o2 measurements from the muscle (Fig. 5, A and B).
There were no significant changes in blood flow heterogeneity in exercising QF between measurements during exercise (Fig. 4C), but posterior muscle blood flow almost doubled (2.0 ± 0.7 vs. 3.9 ± ml·100 g−1·min−1) and flow heterogeneity in the same muscles increased (42 ± 3 vs. 65 ± 5%) from rest to control exercise (P < 0.001 in both). Yet, NOS inhibition or combined NOS and COX inhibition did not affect posterior muscle blood flow (P = 0.59; Fig. 4B) or flow heterogeneity (P = 0.39; Fig. 4D). There were no changes in blood pressure with the NOS and COX inhibition during exercise (Table 2).
The power output of the applied knee-extension exercise was calculated to be <10 watts, with an average workload of 4 kg. Thus power output was low in nature, as also low lactate levels indicate (Table 2).
Blood flow in subcutaneous adipose tissue.
Double inhibition reduced subcutaneous adipose tissue blood flow of the leg at rest, but, during exercise, adipose tissue blood flow was not changed significantly during inhibitions (Fig. 7).
The results of this study show that inhibition of NO formation enhances oxygen uptake in resting skeletal muscle despite a marked reduction in blood flow. Moreover, during exercise, NOS and COX inhibition reduced blood flow in QF and enhanced muscle V̇o2 due to an increased oxygen extraction.
The effect of NO inhibition on muscle blood flow and V̇o2.
The present study demonstrates that inhibition of NOS increases muscle V̇o2 at rest by 20%. This is in accordance with in vitro studies which suggest that NO competes with oxygen and tonically inhibits mitochondrial respiration by binding to the cytochrome c oxidase complex (3, 4, 8, 9, 22). Some animal studies have also previously found increased V̇o2 when NO formation is blocked (33–35). Previous studies that have blocked NOS in humans by infusion of either l-NAME or l-NMMA (10, 27) have shown a similar lowering of resting blood flow as in the current study but no effect on oxygen uptake. The reason for the discrepancy between these studies in regard to oxygen uptake is unsettled, but it is likely to be explained by the fact that PET measurements of muscle blood flow and V̇o2 are more muscle specific than measurements of limb blood flow and oxygen extraction. It is important to emphasize that PET radiowater measures only nutritive blood flow, i.e., where exchange of nutrients and oxygen occurs. Thus it is very likely that, during the NOS inhibition at rest, NO-dependent blood flows to tissues such as skin are reduced to a larger extent than muscle blood flow. This is, however, masked in global blood flow measures, and the result is that total blood flow is seen to be reduced to a similar extent as the oxygen extraction is increased (and thus no changes in V̇o2 are detected). Interestingly, in this respect, especially skin blood flow is pronouncedly decreased during NOS inhibition (7, 11), but also resting blood flow in subcutaneous adipose tissue is largely dependent on NO (1). However, blood flow decreases in adipose tissue may not be the most important explanatory factor, since it was found in the current study that NOS inhibition at rest decreased adipose tissue blood flow to a similar degree as in muscle (Fig. 7) (34% in adipose tissue and 38% in muscle).
Heterogeneity of oxygen supply at rest tended to be increased when formation of NO was inhibited, and more so with combined blockade of NOS and COX enzyme (Figs. 1 and 3). This finding is in accordance with the proposition that NO facilitates O2 distribution within the tissue (38) and means that, in some regions of the thigh musculature, likely in metabolically the least active parts, blood flow was reduced to a larger extent than was the mean flow reduction as a whole. This also suggests that impairments in normal function of NO and COX products (prostacyclin specifically), seen for instance in ageing (25), may contribute to impairments in effective tissue perfusion distribution, which is also common to hypertension, obesity, and diabetes (21). Finally, based on somewhat increased blood pressure and decreased heart rate, it is likely that sympathetic outflow to the muscle may also have changed, which could also have affected flow heterogeneity. We believe that this is the cause of baroreceptor-mediated sympathetic withdrawal, but it is however, also clear from the calculations of vascular conductance that the conductance also decreases with blockade (Table 1); thus, vascular changes occur irrespectively of other minor systemic alterations.
Muscle V̇o2 determined during exercise and NOS inhibition showed that, in contrast to at rest, there was no significant alteration in blood flow or oxygen uptake, although limb oxygen extraction specified to active muscle was increased (Fig. 5, A and B). It therefore seems that the inhibitory effect of NO on respiration is lessened when aerobic metabolism is increased by voluntary muscle contractions.
The effect of combined NOS and COX inhibition on muscle blood flow and V̇o2.
Combined inhibition of NOS and COX caused a similar reduction in resting blood flow as NOS inhibition alone (Figs. 1 and 2A). We did not address the effect of sole COX inhibition on muscle hemodynamics mainly since previous studies have shown that acute COX inhibition alone does not affect blood flow (23). However, because combined inhibition of NOS and COX did not alter resting muscle V̇o2, but single NOS inhibition did, it is plausible that COX products may counteract the effects of NO on aerobic respiration (17, 20, 23). The fact that V̇o2 of the exercising muscle was increased during the combined inhibition of NOS and COX may be explained by mitochondrial uncoupling due to the COX inhibition. COX inhibition is known to increase mitochondrial inner membrane permeability and thus mitochondrial proton leak, which leads to impairments in ATP production (20). It is therefore likely that, during double inhibition, more oxygen had to be consumed to restore the normal aerobic ATP production needed for a given level of exercise. Some enhancement in oxygen extraction may also have been due to NOS inhibition, since NOS inhibition alone tended to (P = 0.07) increase V̇o2 also during exercise.
The finding that muscle V̇o2 was elevated during exercise with the double inhibition is in contrast to what has been reported in three human studies in humans exercising with the knee extensors (15, 23, 31). Limb V̇o2 was in these studies calculated as the product of limb blood flow and limb arteriovenous O2 difference, but now more muscle-specific determinations were used. However, additional explanations are plausible. For one thing, the workload was lower in the present compared with the previous studies, and, consequently, blood flow and V̇o2 were also low. Moreover, in a recent study, where the effect of similar blockers as were used in this study were applied to directly determine their effect on mitochondrial respiration, it was demonstrated that indomethacin impaired the mitochondrial respiration in a dose-related manner (Boushel R, Hellsten Y, and Saltin B, unpublished observations). NOS inhibition, however, had the reverse effect. Thus this opens up a new question for whether an increase or decrease in muscle oxygen uptake during blockade can be a function of the dose of indomethacin and l-NMMA used, which should be tested further in the future.
In this study, we did not compare PET blood flow measures, for instance, to ultrasound Doppler measures, which could have been useful for comparative reasons. However, as our correlations between Fick principle-determined and direct oxygen-15-determined oxygen extractions and consumptions during exercise suggest (Fig. 5, C and D), bulk blood flow-based V̇o2 indeed largely derive from metabolic and flow changes in working muscles. They may not, however, be always easily detectible but can be teased out by specific PET measures, such as following the authentic oxygen as it is first inhaled and then released from blood to muscle and consumed. It is these changes that can be seen in Fig. 6 that render our findings especially compelling. Despite that, there are also possible limitations in our study, one being the fairly small sample size, which may have precluded us to detect some true physiological differences. One of these may be the similar effects of combined NOS and COX blockade on V̇o2 compared with single NOS inhibition at rest. In addition, the nonsignificant (P = 0.07) tendency for increase in muscle V̇o2 during exercise under the single NOS inhibition could be because of a type II error. In this respect, it is also important to consider that the degree of NOS and COX inhibition may not have been the same at rest and during exercise since blood flow was increased substantially. However, previous studies have shown that similar concentrations of l-NMMA have been effective in reducing exercise hyperemia substantially during double blockade (e.g., Ref. 24), and pilot studies in our laboratory suggest that enhanced concentrations of l-NMMA do not reduce blood flow further. The effect of l-NMMA on NOS activity in human muscle has not been determined, but systemic l-NAME inhibition has been found to decrease NOS enzyme activity in skeletal muscle substantially, by ∼70% (10).
In conclusion, the present findings on muscle blood flow by PET methodology are in accordance with previous results showing reduced whole limb blood flow at rest and during exercise with NOS and COX inhibition as determined by thermodilution and Doppler ultrasound. However, in contrast to previous whole limb determinations, the present measurements of V̇o2 specifically in the active muscle show that oxygen uptake at rest and during exercise was increased by NOS inhibition at rest and by NOS and COX inhibition during exercise. We propose that the increased muscle oxygen uptake at rest is the result of a direct effect of NO on mitochondrial respiration. During exercise, the increased oxygen uptake with NOS and COX inhibition may reflect an increased proton leak so that more O2 is needed to restore the ATP production needed for a given level of exercise.
The study was conducted within the Centre of Excellence in Molecular Imaging in Cardiovascular and Metabolic Research supported by the Academy of Finland, University of Turku, Turku University Hospital, and Abo Academy. The present study was financially supported by The Ministry of Education of the State of Finland (grants 74/627/2006, 58/627/2007, and 45/627/2008), Academy of Finland (Grants 108539 and 214329, and Centre of Excellence funding), The Finnish Cultural Foundation and its South-Western Fund, The Finnish Sport Research Foundation, Turku University Hospital (EVO funding), Novo Nordisk Foundation, and The Danish Medical Research Council.
No conflicts of interest are declared by the authors.
This study could not have been performed without the contribution of the personnel of the Turku PET Centre, and we thank them for their excellent assistance during the study.
- Copyright © 2011 the American Physiological Society