INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: exhaled nitric oxide; exercise-induced bronchoconstriction; asthma

The highly reactive molecule, nitric oxide (NO), is produced by a variety of cells and acts as an intracellular messenger in many biologic processes. In the respiratory system, NO is formed in both the upper and lower respiratory tract, and NO detectable in expired air is derived from the lower respiratory tract (1). Elevated levels of exhaled NO (ENO) have been reported in bronchial asthma (2). Because ENO in asthma correlates with airway hyperresponsiveness (3, 4) and decreases after treatment with inhaled or systemic steroids (2, 5), it is regarded as a marker of airway inflammation in asthma. The precise roles of NO in the pathogenesis of asthma, however, are yet to be elucidated.

NO is produced by three isoforms of NO synthases (NOS) in the lung: two isoforms (type I and type III) of constitutive NOS (cNOS) and inducible NOS (iNOS or type II NOS) (8). Type I NOS or neuronal NOS is involved in neurotransmission, and NO from this type of NOS in the lung mediates airway smooth muscle relaxation. Type III NOS or endothelial NOS is involved in vasorelaxation. Type II NOS (or iNOS) is expressed in response to various inflammatory stimuli by airway epithelial, endothelial, and inflammatory cells and has proinflammatory, antimicrobial, and immunomodulatory functions (9). ENO is derived from all three types of NOS and can be detrimental to the host because peroxynitrite, a major metabolite of NO, causes airway epithelial damage and airway hyperresponsiveness (10), or conversely, is bronchoprotective in asthma through a direct action on airway smooth muscle (13).

Exercise-induced bronchoconstriction (EIB) is one of the most confounding problems in children with asthma. Although a test for EIB does not correlate well with other tests for bronchial hyperresponsiveness such as methacholine challenge (16), EIB represents an important clinical feature of hyperresponsiveness in asthma. The mechanisms of EIB, however, remain controversial and the role of NO in the pathogenesis of EIB is still unknown. Recent reports investigating the relationship between ENO and EIB are not in agreement. Scollo and colleagues (17) did not find any change in ENO concentrations during EIB in children with asthma. On the other hand, Kotaru and coworkers (18) measured ENO during airway obstruction induced by hyperventilation, a surrogate model of EIB, and observed elevation of ENO in asthmatics.

The purpose of this study was to investigate the possible roles of NO, detrimental or protective, in the pathogenesis of EIB. To that end, we assessed the changes in the ENO concentration before and after exercise challenge in normal and asthmatic children. The effects of a -agonist and inhaled steroid on ENO changes during EIB were also analyzed.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

We recruited 39 asthmatic children (27 boys and 12 girls; mean age: 11.3 ± 0.5 yr; range: 8 to 18 yr). Diagnosis of asthma was confirmed by a history of recurrent episodes of dyspnea with wheeze, reversible bronchoconstriction, and airway hyperresponsiveness to acetylcholine, according to international guidelines (19). All of them were sensitive to more than one inhalant allergen as evidenced by positive RAST or skin tests. All the patients were on control medication: 14 with inhaled beclomethasone dipropionate (100 to 400 µg/d), 15 with sodium cromoglycate, 20 with sustained-release theophylline (10 with theophylline alone, and 10 with theophylline and other inhaled drugs). All the subjects used inhaled salbutamol or procaterol as needed. The subjects stopped all medications 24 h before exercise testing. We also recruited six normal children (three boys and three girls; mean age: 11.4 ± 0.6 yr; range: 9 to 13 yr) as a control group. They had no history of cardiac or chronic respiratory illnesses. They showed normal pulmonary function parameters and negative RAST for common inhalant allergens. None of the subjects had experienced upper respiratory infections within 6 wk before the testing. Their parents gave written informed consent. The ethics committee at National Mie Hospital approved the study protocol.

Exercise Challenge Test

Exercise challenge test was performed on a treadmill with a fixed load of 6 km/h speed and 10% inclination for 6 min. The heart rate was monitored before and at the end of the exercise. A test was considered valid only when the postexercise heart rate increased to the level more than 70% of the predicted maximal heart rate of each individual. ENO and FEV₁ were measured before and at 0, 5, 10, and 15 min after exercise. EIB was defined as more than 15% decrease in FEV₁ from the baseline level.

Measurement of ENO, FEV₁, and Ventilation

ENO was measured with a chemiluminescent NOx analyzer (CLM-500; Shimadzu, Kyoto, Japan) with a constant sampling flow rate of 2.0 L/min by the single-breath online procedure (20). The subject breathed room air at 22 to 25° C with humidity of 30 to 50%. NO concentration in ambient air was below 5 parts per billion (ppb). To exclude nasal NO, we asked the subjects to expire air with positive pressure to close the velopharyngeal aperture. The subjects were also instructed to expire at constant flow for more than 6 s. The NO concentration at plateau was defined as fractional exhaled NO concentration (FE_NO). FEV₁ was measured with a spirometer (AS-500; Minato, Osaka, Japan). Minute ventilation (E) was measured continuously with an open circuit spirometer and analyzed using an Aeromonitor (AE2805; Minato).

Effect of -agonist Inhalation on ENO during EIB

The subjects who had EIB underwent a second exercise test according to the same protocol 2 d after the first test. They inhaled 50 or 100 µg procaterol with a metered-dose inhaler at 5 min after exercise when the maximal decrease of FEV₁ was observed. Then ENO and FEV₁ measurements were repeated at 10 and 15 min after exercise.

Statistical Analysis

FE_NO is reported in parts per billion. The results are expressed as the mean ± SEM. Data from more than three groups were compared by analysis of variance (ANOVA) with Scheffé's post hoc test. Differences between two groups were analyzed with Wilcoxon's signed rank test. Correlations between the baseline ENO and maximal percent decrease of FEV₁ were evaluated with Spearman's rank correlation test. Differences were considered significant at a value of p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Eighteen of the 39 asthmatic subjects had EIB (EIB group); the remaining 21 did not (non-EIB group). The six normal children (control group) did not have EIB. Table 1 summarizes the demographic data of the three groups. There were no differences in mean age, the baseline FEV_1.0, and preexercise and postexercise heart rate between the groups.

The baseline FE_NO value was higher in the asthmatic patients, significantly in the EIB group, than in the control group. In the subjects with asthma, we also found a correlation between the baseline FE_NO and the maximal postexercise decrease in FEV₁ (Figure 1). There was no significant difference among the groups in the baseline FEV₁ expressed as a percentage of the predicted value or the heart rate (preexercise and postexercise).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Relationship between baseline FENO and maximal postexercise decrease in FEV1 in the 39 subjects with asthma.

In the EIB group, FEV₁ was significantly decreased at 0, 5, 10, and 15 min after the exercise, and the mean maximal decrease in FEV₁, observed at 5 min, was 43.5 ± 4.4%. In the non-EIB group, a small but significant decrease in FEV₁ was observed after the exercise, with a mean maximal decrease of 8.8 ± 1.4%. In the control group, the changes in FEV₁ were not significant (Figure 2A). FE_NO decreased in all groups at the end of the exercise, with mean decreases of 24.0 ± 6.3% (p < 0.05), 8.0 ± 7.6%, and 16.7 ± 8.0% (p < 0.05) in the EIB, non-EIB, and control group, respectively. The time course of FE_NO differed in each group at 5 min after the exercise and thereafter. In the non-EIB and control groups, FE_NO increased to the level higher than the baseline. On the contrary, in the EIB group FE_NO remained at a decreased level at 5 and 10 min after the exercise (Figure 2B).

View larger version (18K): [in this window] [in a new window]   Figure 2.   Changes of FEV1 (A) and FENO (B) after exercise in control group (open circles), the non-EIB group (open squares), and the EIB group (closed squares). FEV1 and FENO are expressed as the percentage of the baseline value. *p < 0.005, significant difference compared with the baseline value (Scheffé).

It is well known that an increase in ventilation causes a decrease in FE_NO. To exclude the possibility that persistent decrease in FE_NO in the EIB group may be a simple consequence of E, 11 patients (three from the EIB group and eight from the non-EIB group) with asthma were further examined for E during and after exercise (Figure 3). Changes in E in the two groups were almost identical. E increased during exercise and reached to the highest level at the end of the exercise (0 min). However, at 5 min after exercise, E returned to the baseline level when FE_NO remained at a decreased level in the EIB group. The mean changes in FE_NO at 5 min in the particular patients were 12.5 ± 2.1% and + 4.3 ± 7.5% in the EIB and the non-EIB group, respectively (p < 0.05).

View larger version (17K): [in this window] [in a new window]   Figure 3.   Changes in E after exercise in the EIB group (closed squares; n = 3) and the non-EIB group (open circles; n = 8). There was no difference in the changes in E between the two groups.

Because it has been reported that acute bronchoconstriction is associated with a reduction in ENO concentrations in asthma (21) and inhalation of a -agonist increases the ENO level in asthma (22), we hypothesized that the prolonged decrease in FE_NO in the EIB group may be due to bronchoconstriction and that -agonist inhalation can reverse the decrease in these subjects. Unexpectedly, although inhalation of a -agonist at the peak of EIB significantly accelerated recovery of FEV₁ (Figure 4A), it did not bring about an increase in FE_NO (Figure 4B). These results indicate that the prolonged decrease in ENO in the EIB group is not simply a result of bronchoconstriction.

View larger version (17K): [in this window] [in a new window]   Figure 4.   Effect of -agonist inhalation at the peak of EIB on FEV1 and FENO. Patients who had EIB at the first exercise testing performed the second testing when they inhaled -agonist, procaterol, at the peak of EIB (at 5 min after the exercise). The postinhalation change at 10 min after the exercise in FEV1 is expressed as FEV1.0 (A) and that in FENO is expressed as FENO. *p < 0.005.

We then analyzed the effect of steroid treatment on the change in the postexercise level of ENO in the EIB group (Figure 5). Twelve of the 21 patients in the EIB group had been treated with an inhaled steroid, whereas nine had been steroid-naive. The magnitude of the decrease and the time-course of FEV₁ after the exercise did not differ between the two subgroups (Figure 5A). FE_NO decreased right after the exercise in both groups. In the steroid-naive group, FE_NO remained at a decreased level until 10 min after the exercise. On the contrary, FE_NO rose close to the baseline level in the steroid-treated group. The mean FE_NO at 10 min after exercise was significantly higher in the steroid-treated group than in the steroid-naive group (Figure 5B).

View larger version (16K): [in this window] [in a new window]   Figure 5.   Changes of FEV1 (A) and FENO (B) after exercise in the steroid-naive EIB group (open circles) and the steroid-treated EIB group (closed squares). FEV1 and FENO are expressed as percentage of the baseline value. Time course of FEV1 after exercise was similar in both groups. FENO at 10 min after exercise in the steroid-treated group was significantly higher than the steroid-naive group. **p < 0.005.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study, we found that exercise challenge induced distinct responses in the ENO concentrations: ENO significantly decreased during EIB whereas it increased in non-EIB subjects, and bronchodilatation with a -agonist did not reverse the prolonged reduction of ENO in the EIB subjects. Changes in ENO were not directly associated with changes in E. These findings suggest a possible impaired NO response to exercise in patients with EIB.

Because of a possible regulatory function of NO on vascular and smooth muscle tone, changes in ENO during physical exercise have been extensively studied in normal subjects and athletes (23). These studies have revealed that FE_NO decreases whereas NO output increases during exercise. It is thought that enhanced airflow during exercise reduces the luminal NO concentration and increases the total amount of ENO per unit time. Whether NO is actually formed in the lung as a result of exercise is still a topic of debate (23, 26). The effect of exercise on ENO in various pulmonary and cardiovascular diseases has also been studied (17, 27, 28).

In agreement with previous studies in normal subjects, we found that FE_NO decreased at the end of exercise in both normal and asthmatic subjects. During postexercise recovery, FE_NO rebounded, as expected, to the baseline or higher in normal and asthmatic children who did not have EIB. Surprisingly, in patients with EIB FE_NO remained at a reduced level during EIB. We postulated two mechanisms for this observation. First, a decreased airway caliber during EIB, owing to reduction in the airway surface area, may have caused reduction in NO diffusion from the airway wall to the lumen, and then a decrease in the NO concentration in the lumen. The findings of several studies support this hypothesis: acute bronchoconstriction induced by hypertonic saline and adenosine was associated with a reduction in ENO concentrations (21), and bronchodilatation by a -agonist caused an increase in ENO levels (22). We found, however, that bronchodilatation by a -agonist at the peak of EIB did not affect the ENO concentrations in these subjects, suggesting that prolonged depression of the ENO concentrations during EIB is not a simple consequence of acute bronchoconstriction. In supporting of this, there have been several observations that acute changes in airway caliber after administration of bronchodilators or methacholine did not affect the ENO level (29, 30). Thus, we concluded that our first hypothesis seems unlikely to be correct.

Our second hypothesis for the sustained decrease in ENO during EIB is that asthmatics with EIB may have defective physiologic NO response to physical exercise. NO synthesized in nonadrenergic, noncholinergic nerves acts as an airway smooth muscle relaxant in humans (31). In guinea-pig models, bronchoconstriction induced by inhalation of cold air was significantly enhanced by an intravenous NOS inhibitor, N^G-nitro- L-arginine methyl ester, suggesting a bronchodilating effect of endogenous NO (32). Bradykinin induces NO release from the epithelium and mediates relaxation of smooth muscle (33). If it is assumed that NO is an endogenous smooth muscle relaxant in asthma, it might act as a counter-regulator of bronchospasm. Therminarias and colleagues (34) subjected athletes to exercise challenge in cold air and found that airway obstruction caused by intense exercise at 10° C was accompanied by a decrease in ENO. They postulated that impaired production of NO as an endogenous bronchodilator may be related to the airway obstruction. Silkoff and colleagues, in their elegant mathematical model (9), demonstrated that the NO diffusing capacity (D_NO) of the airways is increased in asthma. They suggest that D_NO represents an increase in the area of cNOS activity because D_NO remained elevated after steroids whereas the concentration of NO (Cw) in the airway wall decreased. D_NOCw after steroids, presumably reflecting maximal diffusion of constitutive NO, was positively correlated with the provocative concentration of methacholine producing a 20% reduction in FEV₁ (PC₂₀) and FEV₁/FVC. They postulated that the elevated D_NO in asthmatics may reflect upregulation of nonadrenergic, noncholinergic, NO-producing nerves in airways in compensation for decreased sensitivity of airway smooth muscle to the relaxant effects of endogenous NO. These observations strongly support the role of NO as a bronchodilator. Collectively, our results suggest that EIB may be at least partly due to impaired production or release of endogenous NO.

We found that patients who were receiving inhaled steroid treatment showed faster recovery of ENO during EIB than those who were steroid-naive. The NO concentration in any biologic system depends on the rate of enzymatic formation by NOS and the rate of consumption by various biomolecules (35). Continuous transcriptional activation of iNOS gene, leading to abundant expression of its messenger RNA (mRNA) in the airway epithelium of patients with asthma, has been described (36). Decreased expression of iNOS in those receiving an inhaled corticosteroid has also been observed (36). Although these data correspond to our findings that the baseline ENO was significantly higher in asthma than in the normal control group, the faster recovery of ENO from the decreased level during EIB in steroid-treated patients cannot be explained by that report. This contradiction may be explained by consideration of NO metabolism. There are two major metabolic pathways of NO: S-nitrosothiols formed upon reaction of NO with redox-activated thiols; and oxidative metabolites, nitrite and nitrate, formed by reactive oxygen species (35). S-nitrosothiol functions as an active storage pool for NO, and decreased concentrations of the metabolite in tracheal aspirates from children with severe asthma attacks have been reported (37). Assuming that ENO arise partly from degradation of S-nitrosothiol, patients with EIB may have been defective in this metabolite and consequently showed prolonged depression of ENO during EIB, and steroid treatment may have partially restored the S-nitrosothiol level, resulting in a "near-normal" NO response to exercise. However, one may argue against protective function of NO from EIB in the steroid-treated children because they still had EIB despite a "near-normal" NO response. Although precise explanation for this observation is not known, ENO concentrations at 10 min after the exercise in the steroid-treated EIB group were still lower than those in the non-EIB group (see Figures 2 and 5) and may have been insufficient to reverse bronchoconstriction.

In contrast to our observation, Scollo and colleagues (17) reported that the ENO concentration did not change during EIB in asthmatic children. Possible explanations of this discrepancy may be the method of NO measurement, online versus off-line, different patient backgrounds, or different methods for exercise challenge. However, the most significant difference between the two studies is the use of inhaled steroid. In the Scollo study, nine of 10 patients with EIB had been treated with inhaled steroid whereas only 12 of 21 EIB patients had been on inhaled steroid in our study. We observed distinct decrease in ENO during EIB only in steroid-naive patients, not in steroid-treated patients. This suggests that the former study may have missed an "abnormal NO response" in steroid-naive patients. There have been no other studies that investigated NO and EIB in children, and further study is necessary to clarify the role of ENO in the pathogenesis of EIB.

Finally, we also found that the baseline ENO concentrations were significantly elevated in asthma and correlated with the magnitude of postexercise decrease in FEV₁. These findings are in agreement with previous reports that propose ENO as a marker of airway inflammation and bronchial hyperresponsiveness (2, 17). Although the elevation in ENO at the baseline and the decrease in ENO during EIB seem paradoxical, these data may point to two attributes of NO in the airways of bronchial asthma, detrimental and regulatory. Recently, Paul-Clark and colleagues reported that NOS inhibitors administered locally exacerbated inflammation in the carrageenin-induced pleurisy in rats whereas those administered systemically ameliorated inflammation (38). They showed that inhibition of NOS at the site of inflammation caused increase in mediators such as histamine, leukotriene B4, and reactive oxygen species, suggesting that the local production of NO is protective by its ability to modulate concentrations of proinflammatory mediators. Differential effects of NOS inhibitors in their model indicate opposite roles for NO in induction of acute inflammation.

In conclusion, we demonstrated that exercise induced a decrease in the ENO concentration and that the decrease persisted during EIB in children with asthma, while prompt recovery of ENO concentrations was observed in non-EIB asthmatic and normal children; these findings suggest an impaired NO response in patients with EIB. In addition to an inflammatory aspect of NO as previously reported, there may be a possible bronchoprotective role for NO in the pathogenesis of EIB.