INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Ozone is a photochemical oxidant pollutant formed as a result of a series of complex chemical reactions involving oxides of nitrogen (NOx), volatile organic compounds (VOCs), and sunlight, making it predominantly a summer pollutant. Previous human studies have shown that short-term exposure to ozone at concentrations between 0.08 and 0.40 ppm can induce changes in lung function, such as decreases in FEV₁, FVC, TLC, and IC, increase airway resistance (Raw) and symptoms such as fatigue, inspiratory chest pain, and breathlessness, and cause acute inflammation in the airways of healthy humans (1). The inflammation is characterized by an influx of polymorphonuclear leukocytes (PMNs), and is associated with an increase in total protein, albumin, complement C3a, interleukin-6 (IL-6), IL-8 and granulocyte-macrophage colony-stimulating factor (GM-CSF) in bronchoalveolar lavage fluid (BALF) (4). Recent studies have also highlighted the lack of a correlation between lung function responses and airway inflammation, suggesting that these responses are mediated by different pathways (5, 6). The mechanisms underlying the functional responses to short-term ozone exposure are not well understood. The major mechanism of ozone-induced reduction in FEV₁ and FVC is a reduction in IC. It has been suggested that the early functional responses described previously could be mediated by prostanoids and leukotrienes generated by activation of the arachidonate pathway as a result of damage to bronchial epithelium (7, 8). The inspiratory chest discomfort following ozone exposure is attributable to neurally mediated involuntary inhibition of inspiration, perhaps caused by stimulation of C-fibers (9).

In addition, animal studies have shown that short-term exposure to ozone induces bronchial hyperresponsiveness, and that this phenomenon is partly mediated by the release of tachykinins such as substance P (SP) and neurokinins (10). Toxins such as cigarette smoke or mechanical and chemical irritants are known to induce release of SP from sensory nerves in the airways (11). Neuropeptides such as SP and calcitonin gene-related peptide (CGRP) can induce bronchoconstriction and vasodilation, and increase mucus secretion in the airways to varying degrees; additionally SP has several proinflammatory properties, including the priming of PMNs in an inflammatory response, increasing the respiratory bursts in PMNs, the ability to induce chemotaxis to PMNs and eosinophils, and the ability to cause mast-cell degranulation (12).

Only one human in vivo study has addressed the role of neuropeptides in ozone-induced airway inflammation. In this study, Hazbun and colleagues showed that short-term exposure to 0.25 ppm ozone was associated with an increase in the levels of SP in bronchial washings, although the investigators did not report any association between SP and PMNs in the BALF (19).

The main objective of the present study was to test the hypothesis that short-term exposure of healthy subjects to ozone leads to bronchial epithelial damage and stimulates subepithelial sensory nerves, resulting in the release of neuropeptides including SP and CGRP into the airways. A further purpose of the study was to determine whether the release of these neuropeptides into the airways contributes to the development of early bronchoconstrictor responses and recruitment of PMNs in the airways through the neuropeptides proinflammatory properties.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Twelve subjects (10 males and two females; 27.6 ± (6.2) yr of age, [mean ± SD]) were recruited into the study. All of the subjects had a detailed medical history recorded and underwent a physical examination prior to recruitment. Written informed consent was obtained from all of the subjects, and the study was approved by the Southampton University and Hospitals Joint Ethics Committee. All of the study subjects were healthy nonsmokers without a history of allergies, asthma, or any respiratory infection for at least 6 wk prior to the study days. None of the subjects received any dietary antioxidant supplementation, antiinflammatory drugs, or antihistamines during and at least 3 wk prior to the challenge days. Paired samples were available for analysis from nine subjects. Subject 1 could not tolerate bronchoscopy on Day 2 (an ozone-exposure day) of the study, and the procedure was abandoned. Subject 3 had a 39% decrease in FEV₁ after Day 1 (an ozone-exposure day), was not considered suitable for undergoing bronchoscopy, and was withdrawn from the study. Subject 9 was withdrawn because she became distressed on Day 1 (a day of exposure to filtered air [FA]) during recovery following bronchoscopy. She was observed overnight in the hospital and discharged on the following day, and her general practitioner was informed.

Ozone Exposure

All subjects were exposed to 0.2 ppm ozone and filtered air on two separate occasions, with an interval of at least 6 wk between exposures. Exposures were conducted in a randomized, double-blinded, crossover control fashion. The total exposure time was 2 h, during which the subjects were made to exercise (E = 30 L/min) on a bicycle ergometer and to rest, in 15-min alternating cycles. Exposures were conducted with a purpose-built ozone-exposure system in our laboratory. Ozone was produced with a generator (Triogen Ozone System, Glasgow, Model NVF 1/20 DE; manufactured by Ordup Maskin, Copenhagen, Denmark) capable of producing ozone at 5 to 30 mg/h. Ozone was delivered through a mask fitted with a one-way valve to prevent the admixture of inspired and expired air. The ozone concentration was measured every 20 s at the mask by means of a probe connected to an ultraviolet (UV) photometric ozone analyzer (Model 427; Rotork, Oxon, UK) with a precision of 0.02 ppm and a linearity of ± 0.01 ppm. A standard RS232 output cable was used to link the analyzer to a computer in order to display the ozone concentration on the computer screen at 20-s intervals. These values were averaged during the challenge period (0.199 ± 0.009 ppm [mean ± SD] for the nine subjects who successfully completed the study).

We used hospital compressed air as a source from which to generate ozone/FA. In order to ensure that all the air used was free of contamination, a series of filters was incorporated into the system before this air was fed into the generator to produce ozone. The filters used were as follows:

1. Filter Regulator (Model AW2000). This filters particles down to 5 µm diameter, and extracts water.

2. Mist separator (Model AFM2000-02D). This filters particles down to 0.3 µm and has a strong vortex action that precipitates any remaining oil and water.

3. Micro mist separator (Model AFD2000-02). This increases oil and particle extraction down to 0.01 µm.

4. Sofnofil filter (99.9% efficiency). After particle, water, and oil filtration, and air supply was fed through a sealed chamber filled with Sofnofil granules to remove any active gaseous pollutant species including ozone, oxides of nitrogen, and sulfur dioxide.

5. The FA was then passed through an odor filter (SMC Model AFM200-4AG). This filter can remove any particulate matter down to 0.01 µm, and its activated charcoal element is capable of removing any organic species still present.

The FA was then fed into the ozone generator. The concentration of ozone could be adjusted by regulating the flow of ozone into the generator together with the flow of FA into a mixing chamber fitted between the ozone generator and the mask. Through the use of a humidifier (Clinical Humidifier Model MR 730; Fisher and Paykel, Auckland, New Zealand) incorporated into the system, a relative humidity of 40 to 60% and a temperature of 25° C were achieved at the mask. E was measured with the pneumotachographs incorporated into the breathing circuit. The pressure decrease across the stainless steel mesh of the pneumotachograph was transmitted to a pressure transducer via a pair of tubes. A bipolar analogue signal from the transducer is then converted to a digital value, which is transmitted via a dedicated link to the computer. The computer program then averages these values over a sampling window of 0.5 s.

Bronchoscopy

Fiberoptic bronchoscopy (BFIT20; Olympus, Tokyo, Japan) was performed 6 h after the exposures. All subjects were premedicated prior to the procedure with 4% lignocaine spray delivered to the nasal airways and postpharyngeal wall, and with 0.6 mg atropine and 2 to 10 mg midazolam both given intravenously. The bronchoscope was introduced either via the nose or the mouth, depending upon which was more comfortable for the subject. Local anaesthesia of the upper airways and lower airways was achieved with 4% and 1% lignocaine, respectively. Bronchoalveolar lavage (BAL) was done with prewarmed (37° C) saline (160 ml) instilled into from the right upper lobe and the left upper lobe on study Days 1 and 2, respectively. Bronchial biopsies were done with fenestrated forceps from the anterior main carina, subcarinae of the middle and right lower lobes, posterior main carina, subcarinae of the lingula, and left lower lobe bronchi after the first and second exposures, respectively.

Immunohistochemistry for Neuropeptides

Tissue was processed according to the standard protocol described previously (20). Briefly, one or two biopsy specimens were fixed immediately by immersion in 1% paraformaldeyde solution in 0.1 M phosphate buffer, pH 7.4, for 3 to 4 h. After washing in 15% (wt/vol) sucrose in phosphate-buffered saline (PBS: 0.01 M phosphate buffer, pH 7.4; 0.15 M NaCl), tissue was embedded in ornithyl carbamyltransferase (OCT) medium (Miles, Elkhart, IN) on a small cork mat and frozen. Sections (6 µm) were cut in a cryostat, taken up on glass, and dried for 1 h at room temperature before staining with hematoxylin and eosin (H&E) or immunohistochemical staining.

Immunostaining for Neuropeptides

The avidin-biotin complex (ABC) method, as described previously (20), was used to detect nerves and neuropeptides. Endogenous peroxidase was inhibited by soaking tissue sections in 0.3% (vol/vol) hydrogen peroxide in methanol for 20 min at room temperature. After rinsing in PBS, sections were incubated with goat serum diluted to 1:10 to block nonspecific binding of antibodies, and were then incubated overnight at 4° C with primary antiserum (Table 1). After washing in three changes (5 min each) of PBS, the sections were incubated with biotinylated goat antirabbit IgG (Vector Laboratories, Peterborough), diluted 1:100 for 1 h, and washed again. ABC (Vector Laboratories) was prepared according to the manufacturer's instructions and incubated with the sections for 60 min at room temperature. After washing in PBS, the sections were then incubated for 5 min in a chromogen solution containing glucose oxidase, 3,3'-diaminobenzidine, and nickel ammonium sulfate (Sigma Chemical Co., Poole, UK) to amplify the reaction product (20). Omission of primary or secondary antisera was included as a method of providing a control for each biopsy. Immunostains of sensory nerves for protein gene peptide (PGP) 9.5, SP, and CGRP are shown in Figure 1a through c.

View larger version (194K): [in this window] [in a new window]   View larger version (188K): [in this window] [in a new window]   View larger version (182K): [in this window] [in a new window]   Figure 1.   (A through C ) Sequential sections showing immunostaining of sensory nerves (stained black) for PGP 9.5 (A), SP (B), and CGRP (C ) in the bronchial submucosa. Original magnification: ×207.

Nerve Quantification

Measurements of nerve area were made with a Symphony image analyzer (Seescan, Cambridge, UK) as described previously (21, 22). Briefly, nerves in the submucosa were quantified separately by image analysis in at least four random fields, using interactive thresholding to delineate stained nerves and measure their area as a percentage of the specific tissue-compartment area.

Immunohistochemistry for PMNs in Biopsies

Two biopsies were immediately fixed in acetone containing protease inhibitors, and were processed into glycol methacrylate, using our standard laboratory protocol for this, as described previously (23). Details relating to the monoclonal antibody used to identify neutrophil elastase are described in Table 1. Immunostaining was done with the ABC method as described previously (23).

Quantification of PMNs in Biopsies

Quantification of positively staining cells was done with image analysis as previously described (24). Briefly, positively staining cells were counted separately in the epithelium and submucosa, with the results expressed as cells/mm and cells/mm² of epithelium and submucosa, respectively. The length of the epithelium and area of submucosa were determined through computer-assisted image analysis with color-vision software (Improvision, Birmingham, UK).

Total and Differential Cell Counts

BAL samples were filtered through a 100-µm filter (Becton-Dickinson, Oxford, UK), centrifuged at 300 × g for 10 min and washed once in sterile PBS. Cells were counted in a Neubauer hemocytometer (BDH, Dagenham, UK), and an aliquot was separated for cytocentrifuge preparation. Slides were prepared with a Shandon cytospin device (Shandon Southern Instruments, Runcorn, UK). The slides were air dried and differential cell counts were made after staining with a rapid Giemsa stain (HemaGurr; BDH, Poole, UK), with at least 400 cells per slide being counted.

Lung Function Tests

Lung Function was measured at baseline (0 h) and 15 min after exposures on challenge days. Spirometry was done with a hand-held heated-mesh pneumotachograph (IOS Master Screen; Jaeger GmbH, Würzberg, Germany). After a practice maneuver, the subject performed three conventional forced spirometric maneuvers. The best reading was considered for later analysis.

Statistical Analyses

Wilcoxon's matched paired signed rank test was used to compare the immunoreactivity of sensory neuropeptides, results of quantification of neutrophil-elastase-positive cells in biopsy specimens, and total and differential cell counts in BALF. Student's t test was used to compare the percentage change in FEV₁ and FVC from baseline in the first and second exposures. Spearman's rank correlation was used to test the association between the percentage of PMNs and epithelial cells in BALF and SP immunoreactivity, and the percentage change in FEV₁ from baseline with SP immunoreactivity following ozone exposure. A value of p < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Although nine subjects successfully completed the study, only eight paired biopsy and cytospin specimens were considered suitable for final analysis. Immunostained nerves were seen in submucosa, smooth muscle, submucosal glands, blood vessels, and epithelium. Since there was a considerable degree of variability in the amount of smooth muscle, blood vessels, and submucosal glands in the mucosal biopsies, we quantified the immunostaining results in the submucosa where the distribution of sensory nerves was more consistent (Figure 1a through c). There was a significant decrease in immunoreactivity to SP following exposure to ozone (Table 2, Figure 2a). However, there was no difference in the immunoreactivity to CGRP and PGP 9.5 (Table 2, Figure 2b and c) in the first versus the second exposure. In addition, we found increased levels of PMNs (p = 0.07) and ciliated epithelial cells (p = 0.05) in BALF after ozone exposure (Table 3, Figure 3a and b). The significant decrease in macrophage levels in BALF was most likely a secondary shift due to the increase in ciliated epithelial cells and PMNs.

View larger version (21K): [in this window] [in a new window]   View larger version (18K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Figure 2.   (A through C ) Comparison of immunonoreactivity for sensory nerves in the bronchial mucosa between the two exposures. Horizontal lines represent medians.

View larger version (15K): [in this window] [in a new window]   View larger version (17K): [in this window] [in a new window]   Figure 3.   (A and B) Comparison of levels of PMNs and epithelial cells in BALF between the two exposures. Horizontal lines represent medians.

No differences were seen in neutrophil-elastase-positive cells in submucosa on the two exposure days (Table 4). One biopsy was considered unsuitable for quantification owing to an excessive amount of smooth muscle in the sections, thus leaving eight paired biopsy specimens for final analysis. Among the eight paired biopsy specimens, intact epithelium suitable for quantification was present in only four specimens. Therefore, statistical analysis of PMN influx was done for the submucosa only. No significant changes were observed in FEV₁ and FVC following ozone exposure (Table 5). However, a significant inverse correlation was seen between SP immunoreactivity and percentage change in FEV₁ after ozone exposure (r = 0.95, p < 0.0001; Figure 4a). A significant negative correlation was observed between SP immunoreactivity in the biopsies and the percentage of PMNs in the BALF (r = 0.86, p = 0.01; Figure 4b) following ozone exposure. A similar correlation was observed between ciliated epithelial cells and SP immunoreactivity (r = 0.8, p = 0.03; Figure 4c).

View larger version (14K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   View larger version (15K): [in this window] [in a new window]   Figure 4.   (A) Correlation between percent change in FEV1 ([baseline FEV1 on ozone day  post-ozone FEV1] × 100/baseline FEV1 on ozone day) and SP immunoreactivity in bronchial mucosa following ozone exposure. (B) Correlation between PMN levels in BALF and SP immunoreactivity in submucosa following ozone exposure. (C ) Correlation between ciliated-epithelial-cell levels in BALF and SP immunoreactivity in bronchial mucosa following ozone exposure.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The data produced by the study show that short-term exposure to 0.2 ppm ozone decreases the density of SP-immunoreactive subepithelial sensory nerves in the airways, as well as inducing an early inflammatory response and damage to the bronchial epithelium at 6 h after exposure.

We chose to sample the airways at 6 h after exposure because the peak airway inflammatory response to ozone is known to occur at this time (25). We applied immunohistochemical techniques to evaluate the effects of ozone directly on the sensory nerves in the bronchial mucosa, and found a significant decrease in SP immunoreactivity in the submucosa following exposure to ozone, suggesting that there is a release of SP into the airways. Although we did not measure SP levels in the BALF, the observed decrease in SP immunoreactivity is most likely to represent release of this neuropeptide at some time during the 6 h after ozone exposure, and the time required for synthesis of new SP in the jugular/nodose ganglia and its transport down to the nerve terminals. Although a decrease in peptide-immunoreactive nerves could theoretically be caused by damage to or destruction of the nerves themselves, this is unlikely, since there was no significant change in the density of nerves estimated in sequential sections stained for PGP 9.5, the anatomic marker of nerve tissue.

The neuropeptide SP is believed to mediate neurogenic inflammation, the animal studies have highlighted the role of SP in ozone-induced bronchial hyperresponsiveness (BHR) (10, 26). SP has a wide array of biologic properties, such as increasing microvascular permeability, releasing inflammatory mediators from various cell types, and stimulating inflammatory cellular activities including lymphocyte proliferation, mast-cell degranulation, and PMN activation (14). It has also been shown that the release of SP precedes its possible functional effects in humans.

Thus, although SP is detectably increased in BALF immediately after ozone exposure (19), the peak inflammatory response to ozone occurs at 6 h (25), and is detectable for up to 18 h after ozone exposure (4). The results obtained in our study show that by 6 h after ozone exposure there is no neutrophilia in biopsy specimens whereas there is a trend toward increased PMNs in BALF, and continued evidence of local SP release from nerves.

The absence of neutrophilia in biopsy specimens, together with a trend toward an increase in PMNs in BALF, suggests that the transit of PMNs across the airway mucosa could have been rapid. To our knowledge, only one biopsy study has investigated PMN influx in bronchial biopsy specimens from healthy humans. In that study, Aris and colleagues (5), using morphometric analysis, reported PMN influx in the bronchial mucosa at 18 h after ozone exposure. Although the concentration of ozone used in their study was the same as in our study, they employed a higher E, longer periods of exercise (50 min/h), and longer duration of total exposure, all resulting in a greater dose of ozone to the airways. The disparity in the results could in part be explained by the differences in protocol and in the intrinsic ozone responsiveness of the subjects in the two studies.

The inverse correlation between neural SP immunoreactivity and the percentage of PMNs in BALF at this time point (6 h) suggests that SP may contribute to the recruitment and activation of neutrophils in ozone-induced airway inflammation. Animal studies have further demonstrated that ablation of the sensory nervous system or its neutopeptide content with the neurotoxin capsaicin blocks this response (26). In guinea pigs, pretreatment with capsaicin prevented the airway responsiveness and inflammation induced by ozone exposure, whereas in control (vehicle-treated) animals, ozone caused significant bronchial-wall edema and a marked increase in bronchial responsiveness (10).

Although in vitro studies have highlighted a number of proinflammatory properties of SP in relation to PMNs, there are still no reports indicating the instillation of SP, even in the presence of inhibitors of its degradation, causes PMN influx into BALF. Therefore, it remains plausible that the decrease in SP immunoreactivity in biopsies and the PMN influx in BALF observed in our study could represent two separate responses to ozone that could be mechanically unrelated.

Other factors resulting from ozone exposure may act to enhance the effects of SP. We found increased numbers of ciliated epithelial cells in BALF at 6 h after exposure to 0.2 ppm ozone, suggesting that the ozone was causing epithelial damage or fragility. This has been demonstrated previously in animals and with higher concentration of ozone (3 ppm) (10, 26), but never in humans. The damage to bronchial epithelium leading to epithelial-cell desquamation could enhance local antidromic release of SP by exposing intraepithelial sensory nerves (27). Removal of the epithelium could also prolong the effects of the released peptide by removing the major source of neutral endopeptidase, an enzyme that causes SP catabolism and inactivation, thereby enhancing neurogenic inflammation (28).

Furthermore, our finding of an inverse correlation between SP immunoreactivity and ciliated-epithelial-cell levels in BALF following ozone exposure is further evidence for enhanced release of the peptide following epithelial-cell damage.

Although SP and CGRP are often colocalized in the sensory nerves of the airways, no decrements in CGRP immunoreactivity similar to those in SP immunoreactivity were seen. Thus, the decrease in SP immunoreactivity in the absence of a similar change in CGRP may represent a specific response to ozone.

The absence in our study of significant changes in FEV₁, which were seen in some previous studies, could be explained by the smaller sample size, shorter duration of exposure, and relatively milder exercise protocol employed in our study. It has been suggested that an individual's response to ozone is a function of ozone concentration (c), duration of exposure (t), and ventilatory rate (v) (c × t × v) (31). However, the significant inverse correlation between SP immunoreactivity and percentage change in FEV₁ after ozone exposure suggests that SP could contribute to the development of bronchoconstriction in responding subjects.

Future studies will be needed to address the role of neuropeptides in the exacerbation of mucosal inflammation in asthmatic airways following exposure to ozone. However, short-term exposure of humans to 0.2 ppm ozone causes epithelial damage and appears to stimulate subepithelial sensory nerves to release SP into the airways. We conclude that the development of neutrophilic bronchitis following ozone exposure is at least partly due to neurogenic inflammation.