INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Inhalation of ozone may cause a deterioration of lung function as well as an increase in nonspecific airway responsiveness (1) and bronchial allergen responses (2). After repeated exposures to ozone, lung function is deteriorated in the first days of exposure, however, subsequent exposure leads to attenuation of the response (3). This phenomenon has been termed adaptation, and the underlying physiologic mechanisms have not been clarified until now. In contrast to lung function, airway hyperresponsiveness, which is thought to be more closely linked to airway inflammation than to lung function, shows no or only minor adaptation (3). Two studies have analyzed airway inflammation after repeated ozone exposure, by bronchoalveolar lavage, and compared the data with either single ozone exposures (4) or repeated air exposures (5). It was concluded from both studies that airway inflammation is attenuated after repeated exposure, even though BALF protein concentrations, which were still elevated after repeated exposures, indicated ongoing cellular injury. This was also demonstrated in animal studies, which suggested a persisting or increased inflammation after repeated exposures, particularly within the airway mucosa (6, 7). Therefore, the topic of persisting airway inflammation induced by ozone has yet to be clarified.

To elucidate this unresolved issue, we compared the cellular and biochemical composition of BALF obtained after repeated exposures to 200 ppb ozone with data obtained after single exposures to 200 ppb ozone or filtered air. In addition to previous studies, we also determined the cellular composition of bronchial mucosal biopsies and assessed a visual score during bronchoscopy to estimate the degree of macroscopically detectable airway inflammation. We aimed to further quantify the extent of airway injury by measuring soluble markers of oxidative damage, antioxidants, and lipids in BALF.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Twenty-three nonsmoking healthy subjects (FEV₁, 105 ± 13% of predicted) (8) participated in this study (Table 1). Twelve subjects had a positive skin prick test to at least one common allergen (Allergopharma, Reinbek, Germany) and seven subjects reported symptoms of rhinitis but never asthma after allergen exposure. All subjects were free of symptoms at the time of the study and those with seasonal allergy were studied outside their season. Within 4 wk preceding each study day, none of the subjects had suffered an upper respiratory tract infection. Subjects were instructed about the aim of the study and gave their written informed consent. The study was approved by the Ethical Committee of the Medical Board of the State of Schleswig-Holstein.

Study Protocol

On Day 1 of the study each subject's history was taken, and a physical examination as well as a baseline lung function measurement, methacholine inhalation challenge, skin prick test to at least 26 common allergens, and electrocardiography were performed. On Day 2, a 4-h ozone exposure (200 ppb) by mouthpiece was performed to adjust exercise conditions and to measure the ventilation rate during exposure. At least 4 wk later, subjects underwent the first of three different exposure protocols, which were performed in random order at least 4 wk apart. The protocol comprised either a single exposure to filtered air, a single exposure to 200 ppb ozone, or a repeated exposure to 200 ppb ozone on 4 consecutive days. Twenty hours after each of the single or the fourth of the repeated exposures a bronchoscopy was performed by a doctor who was blinded with regard to the subject's exposure.

Ozone Exposure System

The exposure system has been described in detail previously (2). Ozone generated from 100% oxygen (Linde, Hamburg, Germany) was mixed into a stream of clean air. Ozone was monitored continuously and the target ozone concentration of 200 ppb was maintained within peak deviations of less than 5%. Temperature and relative humidity of the inspired air did not differ between exposures. Unlike our previous studies, subjects did not breathe through a mouthpiece, but wore a Plexiglas helmet (size 30 × 30 × 30 cm). The subject's head was placed inside the helmet, which was closed around the neck and supported by a flexible arm. Ozone or filtered air was delivered to the helmet directly in front of mouth and nose and left the helmet via an opening behind the head. This setup permitted subjects to breath freely through either the nose or the mouth. Oxygen saturation and heart rate were monitored at the ear lobe (Oximeter, Copenhagen, Denmark). During exercise, subjects were seated on a bicycle ergometer.

Assessment of Symptoms

The severity of symptoms associated with exposures was recorded in a written questionnaire on a five-point scale. The panel of symptoms comprised nose and throat irritation, cough, chest tightness, shortness of breath, headache, nausea, thirst, and dizziness. For evaluation, symptoms of upper respiratory tract, lower respiratory tract, and general symptoms were added to give three summary scores. Changes in symptoms were expressed as differences between scores assessed before and immediately after exposures since ozone effects were most pronounced at this time.

Lung Function Measurement

Using a pneumotachograph-based system (Jaeger, Würzburg, Germany), inspiratory vital capacity (VC) and FEV₁ were determined. Measurements were performed repeatedly until values within 5% variation were obtained.

Ozone Exposure

Within the 4-h exposure periods, subjects underwent eight half-hour periods of rest and exercise. In each period they breathed for 10 min at rest and for 15 min during exercise. After the exercise, lung function as well as symptom scores were assessed; this took about 5 min. After the last period, lung function was measured repeatedly for 1 h. Symptom scores were assessed immediately and at 1, 6, and 20 h after exposure. During the screening visit (Day 2) the work load had been adjusted on an individual basis to achieve a minute ventilation that could be maintained for 4 h. The same work load was chosen on all exposure days, the mean (SD) value being 14.8 ± 2.1 L/(min · m²).

Bronchoscopy and Lavage Procedures

Bronchoscopy was performed according to international guidelines (9) as previously described in detail (10). Briefly, all subjects were premedicated with 0.5 mg atropine, inhaled 0.1 mg fenoterol, and received supplemental oxygen during the procedure and topical anesthesia of the nose and pharynx by 0.5% lidocaine. The bronchoscope (BF20-D; Olympus, Tokyo, Japan) was introduced transnasally or orally. Endobronchial examination was performed, and the results were summarized in macroscopic scores ranging from zero to 3 and comprising the variables: bronchitis, erythema, hypersecretion, edema, and hypervulnerability (11). The lavage was performed with 150 ml sterile prewarmed (37° C) saline in the lingula or middle lobe, alternating between exposures. BALF was aspirated into siliconized glassware and immediately transported to the laboratory. After centrifugation (1,000 rpm for 15 min at 4° C) cells were transferred to the cytocentrifuge and the supernatant was frozen (80° C). Biopsies were taken after the lavage from the carina of the upper lobe.

Analysis of BALF and Biopsy Material

Cytospin preparations were stained with Giemsa. Differential cell counts were assessed in all samples by one observer blinded to the exposure regimen.

Biopsies were fixed and embedded into paraffin by standard procedures and sections 2 to 3 µm thick were mounted on slides and dried at 60° C for 12 h. The slides were stained with specific antibodies to perform a differential cell count of the leukocytes within the mucosa. We used the following monoclonal antibodies: LCA (DAKO, Hamburg, Germany) against lymphocytes, NP57 (DAKO) against neutrophils, MAC387, KP1 (DAKO), and HAM56 (Enzo Diagnostics, New York, NY) against different populations of macrophages, EG2 (Pharmacia, Freiburg, Germany) against eosinophils and mast cell tryptase (DAKO) against mast cells. After inhibiting endogenous peroxide, nonspecific binding of antibodies was blocked with normal serum; primary antibodies were added and incubated overnight at room temperature. After incubation with a biotin-linked secondary antibody and with an avidin-biotin-peroxidase-complex for 30 min each, slides were stained with diaminobenzidin (DAB). The total submucosal area of the biopsy was measured using a computer-assisted videosytem (KONTRON, Munich, Germany); areas with glands or blood vessels were excluded. The number of stained cells was evaluated by light microscopy (magnification: ×400) and expressed as number of cells per mm² area.

Analysis of Soluble Components in BALF

The concentration of BALF total protein was determined using a commercial protein assay (BioRad, Munich, Germany). The analysis of leukotrienes, thromboxane B₂, and prostaglandins was performed as described previously (10). Concentrations of cytokines were measured by commercially available immunoassays (TNF-, IL-1, IL-6, IL-8, GM-CSF, IL-10 [Dianova Immunotech, Hamburg, Germany]). Furthermore, malondialdehyde, ortho-tyrosine, and para-tyrosine were measured as markers of oxidative stress (12) and glutathione, urate, ascorbate, and total thiols concentrations were determined as an index of antioxidant capacity (13). Glutathione peroxidase was measured with a commercially available immunoassay (Bioxytech, S.A.). For phospholipid analysis, BALF was extracted according to Bligh and Dyer (14). Following extraction, phospholipid classes were analyzed by TLC (15).

Statistical Analysis

Arithmetic mean values and standard deviations (SD) or standard errors of mean (SEM) were computed for symptoms, inflammation scores, and biopsy composition median values and quartiles. Comparisons between exposures were performed by Student's paired t test and by the Wilcoxon's signed-ranks matched-pairs test. Tests were performed two-tailed, and the Kolmogoroff-Smirnov test was used to check for normality of the data distribution. The level of statistical significance was set at p = 0.05. We did not use Bonferroni corrections for the multiplicity of comparisons. Instead, p values are given explicitly wherever possible.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Ventilation and Heart Rate

Mean (± SD) minute ventilation as determined at the screening visit via mouthpiece, was 6.8 ± 2.3 L/min during rest, with a corresponding heart rate of 80 ± 9. Mean minute ventilation during exercise was 26.2 ± 6.4 L/min, corresponding to a minute ventilation per body surface of 14.8 ± 2.1 L/(min · m²). Mean (± SD) heart rate during exercise was 115 ± 10 beats/ min. No significant differences in heart rate were observed between the exposures to filtered air or ozone. There was a small and statistically significant decrease in heart rate between Day 1 (117 ± 12) and Day 4 (113 ± 10) of the repeated exposures (p = 0.032).

Changes in Symptoms after Exposure

The median (quartiles) change in lower respiratory tract symptoms was 0 (0; 1) after FA, 2 (0; 3) after the single exposure to ozone, and 2 (0, 6), 2 (1, 3), 1 (0, 2), 0 (0, 1) on Days 1 to 4 of the repeated exposures to ozone. Lower respiratory tract symptoms were significantly increased after single exposure to ozone (p = 0.01) and after the first day of repeated ozone exposure as compared with FA (p = 0.001). On Days 3 and 4 of the repeated exposures, symptoms were significantly decreased as compared with those on Day 1 (p = 0.001, p = 0.0001).

Changes in Lung Function after Exposure

As compared with the FA exposure, the single ozone exposure caused a significant decrease in FEV₁ (Figure 1A) and VC (p = 0.0002, p = 0.0002). The decrease in FEV₁ and VC observed after each of the repeated exposures to ozone was different from zero on all days (p < 0.05, each). However, the decrease of FEV₁ and VC on Day 4 was smaller than the responses on the previous days (p < 0.01, each). The decrease of FEV₁ and VC observed after ozone exposure during screening, after single ozone exposure and on the first day of the repeated exposures were not statistically different. The baseline values of FEV₁ measured immediately before the start of the exposure on Days 2 to 4 during repeated exposure, showed a significant decrease as compared with the baseline value of Day 1 (p = 0.012, p = 0.010, p = 0.013) (Figure 1B).

View larger version (11K): [in this window] [in a new window]   View larger version (11K): [in this window] [in a new window]   Figure 1.   (A) Changes of FEV1 immediately after exposure to filtered air or ozone as compared with baseline values measured before the exposure. (B) Comparison of absolute volume of FEV1 measured before the exposure. Mean values and SEM are presented. *p < 0.05, **p < 0.01.

Differential Cell Counts in BALF

Data on the cellular composition of BALF are presented in Table 2. The percentage of macrophages was significantly lower (p = 0.005) and the percentage of lymphocytes significantly higher (p = 0.003) after the single ozone exposure as compared with FA exposure (Figure 2B). The same was true when values were compared with those obtained after repeated ozone exposure (p = 0.02 and p = 0.03). The percentage (p = 0.004) (Figure 2A) and the absolute number of neutrophils (p = 0.02) as well as the absolute number of lymphocytes (p = 0.02) were significantly elevated after single ozone exposure as compared to FA.

View larger version (12K): [in this window] [in a new window]   View larger version (12K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 2.   Percentage of neutrophils (A) and lymphocytes (B) in BALF 20 h after exposure. Number of neutrophils (C ) and lymphocytes (D) per mm2 in mucosal biopsies. Bronchitis score (E ) and hypervulnerability score (F ) evaluated during bronchoscopy (scale 0-3). Mean values and SEM are presented. *p < 0.05, **p < 0.01.

Differential Cell Counts in Biopsies

The single exposure to ozone did not cause statistically significant changes in differential cell counts as compared with FA exposure. However, after repeated exposure to ozone the number of neutrophils per mm² was significantly (p = 0.048) elevated as compared with FA exposure (Table 3 and Figure 2C).

Macroscopic Scores of Airway Inflammation

A significant increase in the ratings of bronchitis (p = 0.018, p = 0.027) and hypervulnerability (p = 0.045, p = 0.007) of the bronchial mucosa was observed after repeated ozone exposures as compared with single ozone or filtered air exposure, respectively (Table 4 and Figures 2E and 2F). Similarly, the score for erythema of the bronchial mucosa was increased after repeated ozone exposure as compared with FA (p =0.03).

Analysis of Soluble Components of BALF

After single and repeated exposure to ozone, the concentration of total protein was significantly (p = 0.0002, p = 0.003) increased as compared with exposure with FA (Table 5). The concentrations of IL-8 and IL-6 were increased after single ozone exposure (p < 0.0001, p = 0.0002) as well as after repeated exposure (p = 0.006, p = 0.005). IL-10 could only be detected in samples obtained after repeated ozone exposure. The ratio of ortho-tyrosine to tyrosine was increased after both single and repeated ozone exposure (p < 0.0001, each) (Table 6). Furthermore, the concentration of malondialdehyde was significantly increased after the single ozone exposure (p = 0.019). The level of reduced glutathione was increased after single (p = 0.02) and repeated ozone exposure (p = 0.001), whereas increased levels of uric acid were detected only after single ozone exposure (p = 0.03). The changes in the concentrations of surfactant phospholipids in BALF (Table 7) after ozone exposures were not statistically significant. The increase in the concentration of phosphatidylserine after single ozone exposure was only significant (p < 0.006) as compared to the repeated exposure.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our data demonstrate a persisting or even increased inflammation of airway mucosa after repeated exposures to ozone, despite attenuation of some inflammatory markers in BALF and irrespective of the adaptation of lung function.

The inflammatory airway response to single ozone exposures has been extensively studied (1). Ozone causes an influx of neutrophils and lymphocytes into the airways and an increase in the levels of inflammatory mediators and cytokines (16). Although, at the same time, it produces an acute deterioration in lung function, until now there is no convincing evidence that inflammation as detectable in BALF, biopsies, or induced sputum is related to the observed functional changes or respiratory tract symptoms (17, 20, 21). In the present study, the results obtained after a single exposure to ozone were in line with the previous observations. However, a comparison of the magnitude of the effects of ozone between our study and previous studies reveals rather weak effects in the present study, which may partially be explained by the lower rate of exercise ventilation as compared with that in other studies (4, 16), thereby resulting in a lower dose of ozone delivered to the airways. The low ventilation rate as well as considerations of feasibility also prompted us to include only a single air exposure, as we considered it unlikely that exercise per se, at the level chosen, would cause significant airway inflammation.

The reduction of lung function responses in repeated ozone exposures has been termed "adaptation" or "tolerance" and has already been demonstrated in 1979 (22). In accordance with most of the literature, we also observed the adaptation of lung function responses after repeated exposures to ozone. Throughout the repeated exposures, the particular nature of the functional response was maintained, which is characterized by similar reductions in FEV₁ and VC. In contrast to the reduction of acute effects, there was a persistent although small decrease in baseline lung function as measured before the exposure, thereby suggesting that there are different time scales of the functional responses to ozone, which may rely on different mechanisms. It may be speculated that the persistent effects are related to airway inflammation, whereas it is known that the acute effects are not (23). In parallel to the acute lung function responses and in line with previous observations, the lower airway symptoms experienced during repeated exposures to ozone showed the highest scores on the first day and declined to baseline levels over the consecutive ozone exposures. Despite these parallel changes, there were only weak or no correlations between symptoms and lung function.

There is some information on inflammatory changes after repeated exposures to ozone (4, 5). Markers in BALF such as neutrophil numbers and the concentrations of IL-6 and IL-8 have been found to be decreased as compared with single exposures; this is supported by our data. In contrast to the previous data, however, the levels of IL-6, IL-8, and some markers of oxidative stress and antioxidant defense were still statistically significantly increased as compared with baseline values obtained after filtered air. Interestingly, all studies found that the levels of total protein remained to be elevated after repeated exposures, thereby indicating ongoing cellular damage irrespective of the attenuation of cellular inflammatory responses within the airways. This discrepancy underlines the potential value of the findings that we obtained in the analysis of macroscopic inflammatory scores obtained during bronchoscopy and of the analysis of biopsies. We considered these parameters as potentially relevant, since they should be more closely related to persistent effects of ozone than to changes within the airway lumen that are detectable in BALF. Indeed, our data demonstrated that airway mucosal inflammation was even significantly increased after repeated exposures. The highest scores for bronchitis, erythema, and hypervulnerability and the largest number of neutrophils within biopsies occurred after repeated ozone exposures. These findings are comparable with the results of animal studies where similar increases in the number of neutrophils have been observed (6, 7). Interestingly, Aris and coworkers (17) reported an increase in neutrophil numbers in bronchial mucosal biopsies even after a single exposure to ozone; however, their protocol included a markedly higher proportion of exercise and a higher ventilation rate.

Malondialdehyde, a marker of lipid peroxidation, and ortho-tyrosine, a marker for OH-radical attack, have not been previously analyzed after repeated ozone exposures. Increased concentrations of malondialdehyde occurred after single ozone exposures only, thereby suggesting that during repeated ozone exposures the antioxidant defense might be more effective. The fact that the concentrations of GSH were highest after repeated exposures, stands in favor of this interpretation. Increased levels of GSH, however, could not only result from an upregulated antioxidant defense but also from ozone-induced increases in cellular permeability and the release of intracellular GSH (13). The concentration of ortho- tyrosine was increased after both single and repeated ozone exposures, which suggests that either this marker indicated acute effects of the ozone exposure the day before lavage, or that a maximum level of ortho-tyrosine as compatible with tyrosine turnover and repair had been reached.

Because lung surfactant components may be altered by inhaled oxidative gases such as nitrogen dioxide (15), we also analyzed the distribution of phospholipids in BALF. There were no obvious changes that could be related to single or repeated ozone exposures, and it might be that either the ozone concentration was too low or that 4 d of repeated exposure were too short to influence surfactant metabolism.

To our knowledge, there are no data available so far on the effect of ozone on IL-10. This cytokine was detectable in the BALF of six of 23 subjects, but only after repeated ozone exposure. In vitro studies have shown that IL-10 has the capacity to downregulate the production of proinflammatory cytokines and chemokines by activated monocytes, polymorphonuclear leukocytes, and eosinophils (24). This suggests that IL-10 could play a role in the reduction of ozone-induced airway inflammation within the airway lumen.

Persistent inflammation may cause airway remodeling (25) and in animals, chronic exposure to ozone results in structural alterations of airway mucosa (26). In addition to these chronic effects, ozone is capable of enhancing the airway response to allergen (2, 27). There are also epidemiologic data which suggest that ozone might also increase the susceptibility to virus infections (28). Preliminary data from an ongoing study suggest that repeated exposure to an ambient air concentration of ozone may increase the early as well as the late phase response to allergen and the percentage of sputum granulocytes (29). Currently, however, it can only be speculated whether the ongoing airway inflammation as demonstrated in the present study is responsible for these effects either by a better access of allergen to the tissue or by an enhanced immune response.

In conclusion, our data demonstrate that repeated exposures to ozone increase airway mucosal inflammation as detectable from macroscopic scores and the number of bronchial mucosal neutrophils. In contrast, cellular and biochemical markers of airway inflammation and oxidative stress were either not further increased or attenuated after repeated ozone exposures. These data suggest that, despite adaptation of lung function, airway mucosal damage persists or even increases. This finding might be the basis for detrimental effects of ozone other than acute lung function responses.