INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

O₃ and NO₂ are two of the most important inhaled oxidant pollutants found in indoor and outdoor atmospheres. Their toxic effects are a function of the gas concentration and exposure duration, though NO₂ is less toxic than O₃ at identical concentrations (1). Acute exposure to O₃ at levels near the National Ambient Air Quality Standard can cause respiratory symptoms, decrements in lung function, lower and upper airway inflammation, and epithelial cell injury. Studies in human beings using bronchoalveolar lavage (BAL) have shown O₃-induced increases in polymorphonuclear (PMN) cells, increases in soluble markers of inflammation and repair, and elevations of markers of epithelial permeability (2). Enhanced susceptibility to respiratory infection (1, 6, 7), with only mild inflammation (8), is the primary health effect caused by exposure to NO₂ at levels below 2 ppm. Individuals differ in their sensitivity to the health effects of oxidants with no correlation between changes in lung function and the intensity of the inflammatory response (1, 3, 5, 9). In mice, the susceptibility to changes in lung function, hyperreactivity, permeability, or inflammation appears to be controlled by separate genes (10). Although the basis for the differences in the sensitivity and the mechanisms involved have not been elucidated, variations in the ability to up-regulate antioxidants and antioxidant enzymes have been implicated (10, 12).

Current evidence suggests that reactive oxygen species (ROS), lipid ozonation products and NO₂-reactive absorption products, that are formed during reactive absorption are the major mediators of the effects of O₃ and NO₂ (16, 17). Only oxidants, which escape neutralization by antioxidants in epithelial lining fluid (ELF), can damage lung cells. Antioxidants, antioxidant enzymes (AOE), and glutathione (GSH) in ELF are the first line of defense of the lung against inhaled oxidants (15, 18). We have shown that ELF contains two selenium-dependent GPxs, cellular (cGPx) and extracellular (eGPx), each contributing approximately 50% to the total GPx activity (19). In the presence of GSH concentrations found in ELF, and taking into account the K_m of GPxs toward GSH, GPxs are able to detoxify H₂O₂ and lipid hydroperoxides (18, 20, 21). Primary bronchial epithelial cells, alveolar macrophages, and lung cell lines synthesize cGPx and eGPx and secrete eGPx (19). Animal studies have shown that exposures to O₃ and NO₂ increase cGPx activity in the lung tissue (1, 15, 22, 23). The eGPx response to inhaled pollutants has not been studied.

We hypothesized that the amount of GPxs in ELF is responsive to O₃ or NO₂ exposure and that the baseline level and the changes in ELF eGPx caused by O₃ and NO₂ exposure inversely correlate with the degree of inflammation. GPx activity and eGPx protein were determined in ELFs obtained from individuals who had varying degrees of susceptibility to O₃ and NO₂ exposure. The ELF samples were collected previously as part of other studies and analyzed for determinants of O₃ and NO₂ susceptibility (3, 5).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subject Selection

Informed written consent was obtained from each subject. The Research Subjects Review Board of the University of Rochester approved the studies.

Subject selection for O₃ exposure. Subject selection and study design were described in previous publications (3, 5). Briefly, the 25 volunteer subjects were healthy nonsmokers of both sexes, 25 ± 1 yr old (range 18-40 yr), who were screened to be "responders" or "nonresponders" as defined previously (3, 5).

Subjects for NO₂ exposure. Subjects were 12 males and 9 females, all healthy, lifetime nonsmokers between the ages of 18 and 40 yr (27.0 ± 4.3 yr), with no symptoms of upper respiratory infection for at least 6 wk prior to the study. All subjects participating in the study had normal spirometry.

Exposures

All exposures were carried out in a special exposure chamber whose characteristics have been previously described (24). Exposures were separated by at least 3 wk.

Exposure to O₃, and lavage processing protocols. The exposures and sample collections were performed as part of a previous study (3, 5). Each subject underwent a single exposure with exercise to filtered air and two identical exposures to 0.22 ppm O₃ gas for 4 h. BAL and blood samples were collected immediately after one of the O₃ exposures and 18 h after the second O₃ exposure. Exposures were double blinded and conducted in randomized order. O₃ was generated and monitored as previously described (5).

Exposure to NO₂ and lavage processing protocols. The exposure protocols were designed to test worst-case exposure to NO₂ in homes with unvented sources of combustion. While performing exercise, the subjects were exposed for 3 h in the environmental chamber to air (control) and to two levels of NO₂, 0.6 and 1.5 ppm. Each exposure was administered in a random order and in a double blind fashion. Subjects exercised for 10 min of each 30 min at a level that was previously determined to achieve a minute ventilation (E) of approximately 40 L/min. Pulmonary function was measured before and immediately after each exposure. BAL and phlebotomy were performed 3.5 h after exposure (the time point at which the greatest reduction in inactivation of influenza virus in association with NO₂ exposure was previously observed [25]). NO₂ delivery and monitoring as well as pulmonary function tests have been described previously (6, 8).

Physiologic Testing

Methods for spirometry and determination of thoracic gas volume and specific airway conductance have been reported previously (3).

BAL

BAL procedures and processing were described elsewhere (5, 8). The fiberoptic bronchoscope was gently wedged in the subsegmental airway of the inferior segment of the lingula. Four 50-ml aliquots of sterile saline were sequentially instilled and immediately withdrawn (no dwelling time) under gentle suction. The lavage was repeated on the subsegmental airway of the right middle lobe. The same lingular and middle-lobe subsegments were entered during each subject's three bronchoscopies. The first returned aliquot (bronchial lavage) was not analyzed for GPxs. For the O₃ study, only the returns from the lingula were used for GPxs analysis, because the parameters for lung injury and inflammation were assayed only on these returns. For the NO₂ study, all assays were performed on pooled alveolar returns from the lingula and the right middle lobe.

The returns were placed on ice, and an aliquot was taken for total and differential cell counts. The rest of the lavage was immediately centrifuged at 300 × g for 6 min at 4° C, and the supernatant was frozen in aliquots at 70° C.

Total and Differential Cell Counts

Total cell counts and viability were performed as previously described and cell counts were expressed as cells/ml BAL (5, 6).

Measurement of GPx Activity

GPx activity was measured as previously described (19). For the activity measurement, the standard reaction mixture (at 37° C) contained 0.1 M tris (hydroxymethyl) aminomethane (Tris)-HCl, pH 8.0, 0.02 mM NADPH, 0.5 mM ethylenediaminetetraacetic acid (EDTA), 2 mM GSH, 1 U of glutathione reductase (GR), and 0.5-0.9 ml of cell-free BAL in a total volume of 0.99 ml. The reaction was started with the addition of t-BuOOH (10 µl, final concentration 70 µM). The oxidation of NADPH (at 340 nm) was measured against the reference that did not contain t-BuOOH using a spectrophotometer (Lambda 4B; Perkin Elmer, Norwalk, CT). One milliunit of GPx activity was defined as the oxidation of 1 nmol NADPH/min. In the assay described above, total GPx activity was determined. There was not enough BAL fluid to determine the contribution of cGPx and eGPx to the total GPx activity by immunoprecipitation of activity assay (19).

Determination of eGPx Protein by ELISA

eGPx protein was determined by a glutathione peroxidase ELISA kit (from CALBIOCHEM Biochemical & Immunochemical, San Diego, CA) according to the instructions of the manufacturer using a Spectra Max 340 microplate reader (Molecular Devices, Sunnyvale, CA). Normal human plasma and plasma from an anephric individual with a low eGPx level were used as the controls, because kidney is the major source for plasma eGPx (26). There were 9.0 and 2.3 µg/ml eGPx in the normal and anephric plasma, respectively. This corresponded to activities of 0.35 and 0.094 U/ml in normal versus anephric patient plasma, respectively. Because eGPx is the only GPx in plasma, the parallel decrease in activity and protein confirms that the ELISA assay used is specific to eGPx. cGPx protein could not be determined as there is no available kit to determine cGPx protein.

Measurements of Proteins in BAL

Albumin concentration was determined to provide an indication of changes in epithelial permeability and injury. Albumin was measured using modified antibody-capture ELISA as described previously (25). Interleukin 6 (IL-6) and IL-8 were measured using commercially obtained immunoassay kits (R&D Systems, Minneapolis, MN) by a microtiter plate reader (Model EL312; Bio-Tek Instruments, Winooski, VT).

Determination of ELF Volume in BAL

ELF volume in BAL was determined by the urea dilution technique to normalize for dilution of ELF by the saline in the BAL procedure. Urea concentration in the lavage and in plasma were determined using BUN Endpoint Kit (Sigma, St. Louis, MO) after having been obtained from the same individual at the time of the lavage. ELF dilution in BAL was calculated by the following formula: ELF dilution factor = urea concentration in plasma/urea concentration in lavage. One drawback of this method is that as a function of dwelling time during the lavage procedure and the volumes of the washes, urea might escape from the lung tissue into the lavage and lead to higher estimates of ELF volumes. Nevertheless, the urea dilution method is the best method known to estimate ELF volumes and is widely used to normalize components in ELF for comparison purposes (27, 28).

Data Handling and Statistical Methods

The primary analyses for the data from BAL were based on a two-way mixed model or repeated measures analysis of variance (ANOVA). For a small number of end points, analysis of covariance (ANCOVA) was performed. For each end point the dependent variable in the ANCOVA analyses was the difference between the measurement after exposure to air and exposure to ozone followed by assessment either early (immediately exposure) or late (18 h after exposure). Parallel analyses were run for both differences (early and late). The independent variables in the ANCOVA were age, sex, responder status ("responder," "nonresponder"), and both air GPx (or eGPx) and the difference between the air GPx (or eGPx) and the corresponding ozone exposure GPx (or eGPx). That is, we wished to determine whether either the baseline level or the change in GPx (or eGPx) after exposure to ozone would predict the corresponding change in the particular outcome measure. A residual analysis was included and outliers were removed for these analyses. In some cases, the residual analysis indicated that the variance was not constant, and a logarithmic transformation was employed to stabilize the variance. A level of 5% was required for significance. For comparison of means between two studies, the two-sample unequal variance t test was used.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ELF Dilution Factor as a Function of O₃ and NO₂ Exposures

The results presented below for GPx activity and eGPx protein measured in BAL were expressed per ml ELF using the correction factor obtained by the urea method (ELF dilution factor). When compared to air exposure (152.3 ± 8.4), the dilution factor was 84.1 ± 5.9 immediately after O₃ exposure and 97.6 ± 5.5 18 h after O₃ exposure (p = 0.0001). In the NO₂ study, the ELF dilution factor in air was 103.4 ± 5.2, in 0.6 ppm NO₂ it was 98.3 ± 5.8, and in 1.5 ppm NO₂ it was 95.8 ± 5.8. Thus, the dilution factor (or ELF volume recovered in BAL) changed with O₃ exposure but not with NO₂ exposure.

Total GPx Activity in BAL of O₃ versus Air-exposed Subjects

GPx activity measured in the BAL was corrected by the ELF dilution factor to estimate the actual activity in the original ELF. In Figure 1, values of GPx activity after air and O₃ exposures are presented for each individual. After air exposure, there was a substantial GPx activity variability among the subjects. O₃ exposure caused an immediate decrease in GPx activity in most of the subjects, which lasted, although to a lesser degree, for 18 h. As shown in Table 1 the average decrease in GPx activity immediately after O₃ exposure was 41%. The activity remained 31.3% lower than the activity in air (p = 0.0001, one outlier omitted). Although the activity was increased by 10% at 18 h postexposure compared with immediately after exposure, the increase was not significant. When the results were expressed per ml BAL (not normalized by the ELF dilution factor), there was no difference in GPx activity between air and ozone-treated groups (Table 1).

View larger version (27K): [in this window] [in a new window]   Figure 1.   GPx activity in ELF collected immediately (OE) or 18 h (OL) following air or O3 exposures. Each subject was exposed to air and two O3 exposures. BAL was collected either immediately (OE) or 18 h after exposure (OL). GPx activity and ELF dilution factor were determined. Results are expressed as mU/ml ELF. Basic GPx activity showed a large variability among the subjects. Most of the subjects showed a decline in the activity immediately after O3 exposure.

View larger version (29K): [in this window] [in a new window]   Figure 2.   eGPx protein in ELF collected immediately (OE) or 18 h (OL) following air or O3 exposures. Each subject was exposed to air and two O3 exposures. BAL was collected either immediately (OE) or 18 h after exposure (OL). eGPx protein and ELF dilution factor were determined. Results are expressed as µg/ml ELF. Basic eGPx protein concentration showed a large variability among the subjects. Most of the subjects showed a decline in eGPx protein immediately after O3 exposure.

eGPx Protein in BAL of O₃ versus Air-exposed Subjects

eGPx protein in BAL was determined by ELISA. In Figure 2, values of eGPx protein after air and O₃ exposures are shown for each individual. There was a wide variation in individual's eGPx protein level after air exposure. When considered as a group, when compared with eGPx protein in air, eGPx protein was decreased immediately after O₃ exposure by 40% and remained 29% lower 18 h postexposure (Table 1). Although there seems to be an increase in the protein 18 h after exposure when compared with eGPx levels immediately after secession of exposure, this increase was not significant. When the results were expressed per ml BAL, there was no difference in eGPx protein concentration between air and ozone-treated groups (Table 1).

If it is assumed that only half of the total GPx activity in ELF is due to eGPx (19), the specific activity of eGPx in ELF, regardless of O₃ exposure, was 40 U/mg. This value is in agreement with the specific activity determined for purified eGPx ([21] and unpublished results).

Total GPx Activity in BAL of NO₂ versus Air-exposed Subjects

As shown in Figure 3, the basic levels of GPx activity in air were variable among subjects. There was no difference in GPx activity between the air and the 0.6 or the 1.5 ppm NO₂-exposed subjects (145.5 ± 13.4, 156.7 ± 23, and 126.4 ± 12.9 mU/ml, respectively). Expressing the activity per ml BAL did not change this outcome (data not shown).

View larger version (25K): [in this window] [in a new window]   Figure 3.   GPx activity in ELF collected 3.5 h following air or NO2 exposures. Each subject was exposed to air, and two NO2 exposures (0.6 and 1.5 ppm). BAL was collected 3.5 h after exposure. ELF dilution factor was measured by the urea dilution method and GPx activity was determined. Results are expressed as mU/ml ELF. Basic GPx activity showed a large variability among the subjects. No difference in GPx activity was found between air and NO2-exposed groups.

The measured baseline GPx activity in air-exposed subjects in this study group (145.47 ± 13.42 U/ml ELF) was lower (by unequal variance t test p = 0.02) than the measured baseline GPx activity in the air-exposed subjects in the O₃ exposure study group (223.56 ± 24.4 U/ml ELF). GPx activity in control plasma assayed in parallel in both studies was of equal value. Currently we have no explanation for the difference in ELF dilution factor and GPx activity between air-exposed subjects in the O₃ and in the NO₂ studies. It might be attributed to the different alveolar compartments sampled (lingula versus combined), or to the various sampling hours (morning for O₃ study and afternoon for the NO₂ study).

eGPx Protein in BAL of NO₂ versus Air-exposed Subjects

The basic concentration of eGPx in air was variable among subjects (Figure 4). There was no difference in eGPx protein between the air and the 0.6 or the 1.5 ppm NO₂-exposed subjects (2.45 ± 0.28, 2.50 ± 0.33, and 2.30 ± 0.28 µg/ml ELF, respectively). Expressing the protein concentration per ml BAL did not change this outcome (data not shown).

View larger version (27K): [in this window] [in a new window]   Figure 4.   eGPx protein in ELF collected 3.5 h following air or NO2 exposures. Each subject was exposed to air, and two NO2 exposures (one to 0.6 ppm and one to 1.5 ppm). BAL was collected 3.5 h after exposure. ELF dilution factor was measured by the urea dilution method and eGPx protein was determined by ELISA and expressed as µg/ml ELF. Basic eGPx protein concentration showed a large variability among the subjects. No difference in eGPx protein was found between air and NO2-exposed groups.

eGPx protein concentrations in all the groups in this study were comparable to eGPx protein concentrations in the air-exposed subjects in the O₃ study (2.68 ± 0.25 µg/ml ELF). Because baseline values for GPx activity were lower in the NO₂ study than in the O₃ study, the specific eGPx activity in ELF of the subjects from the NO₂ study was lower (27.5-31.3 U/mg compared with 40 U/mg). A possible explanation for this discrepancy might be that the relative amounts of eGPx and cGPx are variable in the various compartments sampled in the two studies (lingula versus combined). Consequently, our assumption of equal amounts of eGPx and cGPx on which the calculations were based might not be accurate.

Pulmonary function, airway inflammation, and epithelial permeability of these subjects were measured and described in other studies (3, 5, 29) and are briefly summarized below.

Pulmonary Function

As described previously (3), the subjects did not differ with regard to age, baseline FEV₁, or E during exposure. Decrements in FEV₁ after O₃ exposure of "responders" returned to near baseline at 18 h.

Airway Inflammation and Epithelial Permeability

As previously described for these subjects (5, 29), "responders" and "nonresponders" showed evidence of airway inflammation and injury with O₃ exposure. PMN influx into BAL at the 18 h time point was six times higher than after air exposure. IL-6 and IL-8 increased by up to 10-fold immediately after O₃ exposure and correlated with the late increase in PMN. Albumin concentration, an indicator of epithelial injury, increased with O₃ exposure and reached maximal levels 18 h after exposure. When the concentration in the lavage of the above-mentioned proteins was expressed per ml ELF (instead of per ml BAL), the results obtained were similar except for IL-8, which was not significantly altered with ozone exposure (data not shown).

There were no significant effects of NO₂ exposure on lung function tests. Also, when compared with air-exposed subjects, no changes in PMN or epithelial permeability markers were found in alveolar lavage after NO₂ exposure.

Relationships between GPx Activity and eGPx Protein Level and Pulmonary Function, Inflammation, and Epithelial Permeability in Response to O₃ Exposure

ANCOVA was used to examine the possible role of GPxs in the changes of lung function, inflammation, and epithelial permeability in response to O₃. GPxs baseline activity and eGPx baseline protein level in air and changes in these parameters in response to O₃ were the independent variables; changes in lung function, indicators of inflammation and permeability, were the dependent variables. As is shown in Figure 5, there was a significant inverse relationship between the baseline (air) level of eGPx protein and the increase in PMN 18 h after O₃ exposure (p = 0.04). Thus, lower levels of eGPx in ELF may be a predictor of increased susceptibility to ozone-induced inflammation.

View larger version (10K): [in this window] [in a new window]   Figure 5.   ANCOVA with baseline (air) eGPx protein as the independent variable and the increase in PMN 18 h after exposure to O3 as the dependent variable. Shown is the association between the increase in PMNs adjusted for covariates included in an ANCOVA analysis versus baseline eGPx. The PMN values have been adjusted for all other independent variables in the model, which included baseline eGPx as one of the predictors (see METHODS).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The main findings of this study are that in the majority of subjects GPx activity and the eGPx protein level in ELF decreased and remained low for at least 18 h post-O₃ exposure. In addition, individuals who had lower baseline concentration of the antioxidant enzyme eGPx in their ELF had higher PMN 18 h after exposure to O₃ at near ambient concentrations. Because GPxs in the presence of glutathione (in concentrations found in ELF) might be responsible for neutralization of hydrogen peroxide and lipid hydroperoxides, the decrease in GPxs activity can lead to oxidative damage to the cells in the air-lung interface (15, 18, 20).

The baseline amounts of GPxs varied by as much as 600% between individuals while O₃ exposure decreased GPxs in most individuals by 40%. The decrease was significant as determined by ANOVA and suggests that the capacity to scavenge hydroperoxides might be impaired in most of the subjects.

The mechanisms involved in decreasing GPx activity and eGPx protein in ELF are unclear. ROS formed by reactive absorption of O₃ or NO₂ exposure initially consume low-molecular-weight antioxidants like uric acid, ascorbic acid, and glutathione. Thereafter, or in parallel, proteins in the ELF are oxidized and depleted (18, 30, 31). Therefore, the most probable explanation for the loss of GPx activity and eGPx protein (epitope recognized by the antibody) is their oxidation. However, there is a flaw in this explanation. Total GPx activity represents combined cGPx and eGPx activity, each contributing 50% of the total (19). The percentage decline in GPx activity (40%) was similar to the percentage decline in eGPx protein (40%), which indicates that eGPx and cGPx were inactivated to the same extent. eGPx and cGPx are selenium-dependent enzymes. ROS originating from O₃ oxidize the ionized selenol in the active site of cGPx, and thereby inactivate the enzyme (32). We believe that there is no available information on the response of eGPx to ROS. In addition, cGPx is more labile than eGPx (e.g., heat inactivation) (33). Therefore, it is unlikely that cGPx and eGPx are inactivated by ROS to the same extent. A more plausible explanation is that more cGPx than eGPx was inactivated, but cGPx was replenished and eGPx was not. The source for the cytoplasmic cGPx could be lysis of damaged cells, which do not contain stores of eGPx; the latter is secreted as soon as it is formed (19). An alternative explanation to the disappearance of 40% of GPxs could be that O₃, but not NO₂, caused GPxs to move out of the lavageable compartment.

Plasma contains eGPx (400 mU/ml) in higher concentrations than ELF (34). The main source for plasma eGPx is the kidney (26). eGPx in ELF decreased despite an expected increase in its concentration because of the influx of plasma proteins. The latter is manifested by the increase in albumin concentration in ELF (29). Unless eGPx in plasma is also inactivated by the O₃ treatment, even more eGPx protein might have been depleted from ELF than what is evident merely by comparing its concentration in ELF of air and O₃-exposed individuals.

The decrease in eGPx protein and GPx activity after O₃ exposure was evident when the data were expressed per ml ELF rather than per ml BAL. It is critical to correct the concentrations of ELF components for the actual ELF volume when their concentrations are compared before and after injury to the lung, because the injury causes changes in the actual ELF volume as a result of exudative processes (27). Signs of increased exudative processes were evident only after O₃ exposure (29). As expected, ELF volume was higher (ELF dilution factor lower) after O₃ exposure compared with air exposure. Because of the lack of differences in GPxs with O₃ treatment when the results were expressed per ml BAL, the following possibility should be considered: it cannot be ruled out that the decreased concentrations of GPx activity and eGPx protein in ELF were not because their total amount was decreased, but because of the increase in ELF volume. Even if this is the case, our results show that the lung cells are not able to secrete more eGPx to maintain the normal concentration of eGPx in ELF.

In this study, we used a sequential BAL procedure and estimated the dilution factor of ELF in BAL by the urea method. This lavage procedure causes an overestimation of ELF volume because of the equilibration of urea between the tissue and the urea-free saline instilled. Therefore, the concentrations of both eGPx protein and GPx activity are most likely underestimated under the conditions tested (27). The underestimate should be equal under all conditions studied, provided that there was no greater leakage of urea into BAL in the O₃-exposed then in the air-exposed individuals. Urea is a small molecule that should freely equilibrate between the fluid compartments of the body regardless of whether the cellular barrier is injured or not (28). Therefore, for a given individual, we do not expect a higher influx of urea during the lavage procedure performed after O₃ versus air exposure. This assumption should be taken cautiously because several animal studies showed that O₃ exposures at higher concentrations and longer durations than were used in this study might cause urea influx (35).

There was a marked variability among individuals in the degree of the O₃-induced changes in GPx activity and eGPx concentration. This variability was due mostly to the marked variability in the baseline GPx activity and eGPx protein as measured after air exposure. Whether this variability in GPx changes in response to O₃ is genetically controlled needs to be elucidated. Changes in ELF albumin (a marker of lung permeability due to cell damage) with O₃ exposure were not correlated with either baseline or changes in eGPx protein or GPx activity, which indicates no association between eGPx and a marker of cell injury. There was also no association between eGPx protein and GPx activity and "responders" versus "nonresponders" status, which suggests no association between GPxs and changes in lung function. In contrast, the association between eGPx protein baseline level and the increase in PMN after O₃ exposure suggests that baseline eGPx levels may predict the intensity of the inflammatory response. Whether lower baseline ELF eGPx contributes to other health effects by other inhaled oxidants and in more sensitive populations (asthmatics) needs additional investigation.

The association of eGPx baseline concentration and the inflammatory response, and the dissociation between eGPx concentration and the degree of cell injury or lung function changes, agrees with the segregation of the types of injury obtained in strains of mice with variable susceptibility to oxidants (10, 12, 36).

NO₂, in concentrations used in the current study, did not reduce GPx activity and eGPx protein and did not cause alveolar inflammation and epithelial hyperpermeability. This is probably because NO₂ is a weaker oxidant than O₃. In addition, the genetic mechanisms underlying the susceptibility of mice to O₃ and to NO₂ are different (13).

In conclusion, GPx activity in ELF of healthy nonsmokers was decreased immediately by 40% and remained low 18 h after O₃ exposure. This decrease was due, in part, to depletion of eGPx because eGPx protein declined by 40%. In contrast, NO₂ exposure at two concentrations found indoors did not alter GPx activity and eGPx protein level. Thus, O₃, a strong oxidant, decreased both total GPx activity and eGPx protein level in ELF, while NO₂, a weaker oxidant did not. These results suggest that exposure to ambient levels of O₃ might deplete eGPx in ELF for as long as 18 h. The importance of eGPx in the pathology of O₃ exposure was evident from the association between lower baseline eGPx concentrations in ELF and increased PMN after O₃ exposure.