ARTICLE IN PRESS

Respiratory Medicine (2006) 100, 2227–2234

0954-6111/$ - sdoi:10.1016/j.r

�CorrespondiResearch LaborPlace, Edinburgfax: +44 131 65

E-mail addr

Permeability, inflammation and oxidant status inairspace epithelium exposed to ozone

Douglas Morrison, Irfan Rahman, William MacNee�

Respiratory Medicine Unit, Department of Medicine, University of Edinburgh, Medical School,Teviot Place, Edinburgh EH3 9YW, Scotland, UK

Received 16 December 2004; accepted 7 October 2005

KEYWORDSOzone;Permeability;Neutrophils;Oxidants;Anti-oxidants

Summary The aim of the study was to investigate possible mechanisms of epithelialinjury in normal subjects exposed to environmentally relevant concentrations of ozone.Fifteen healthy non-smoking subjects (M:F 12:3) were studied. Five of the 15 subjectswere exposed to filtered air, six were exposed to ozone 100 parts per billion (ppb) andseven were exposed to ozone 400ppb with 99mtechnetium labelled diethylene-triamine-penta-acetate (99mTc-DTPA) or bronchoalveolar lavage (BAL) performed 1 or 6h after

ee front matter & 2006med.2005.10.005

ng author. Respiratoratories, Wilkie Buildinh EH8 9AG, UK. Tel.: +1 1558.ess: wmacnee@ed.ac.u

exposure as indicated above. All the above studies were performed on differentoccasions at least 5 days apart. The subjects were exposed on each occasion for 1hduring intermittent exercise at a ventilation of 40 lmin�1. 99mTc-DTPA lung clearancedid not change after either level of ozone exposure, but neutrophils increased in BAL 6hafter exposure to 400ppb. Superoxide anion release from mixed BAL leucocytesdecreased 1h after 100ppb and 6h after 400ppb. Products of lipid peroxidation inepithelial lining fluid decreased both 1 and 6h after 400ppb. There was no change inanti-oxidant capacity or glutathione concentrations.

Ozone exposure did not increase epithelial permeability, but was associated withneutrophil influx into the airspaces, without evidence of increased oxidative stress.& 2006 Elsevier Ltd. All rights reserved.

Introduction

Concerns have been raised over man-made ozone inthe lower atmosphere and its possible adverse

Elsevier Ltd. All rights reserv

y Medicine, ELEGI, Coltg, Medical School, Teviot44 131 651 1435;

k (W. MacNee).

effects on health. Ozone has been shown toproduce increased respiratory epithelial perme-ability in vivo and in vitro.1,2 Ozone also producesinflux of neutrophils into the airspaces and evi-dence of inflammation in healthy humans3 and inasthmatics.4 Neutrophil influx has also been de-monstrated in animal studies of ozone exposure.5

Ozone is the second most powerful oxidantknown and is believed to cause damage tobiological tissues either by direct reaction and/or

ed.

ARTICLE IN PRESS

D. Morrison et al.2228

through the formation of free radicals and reactiveoxygen intermediates. Peroxidation of membranelipids is thought to be an important mechanism ofozone injury, resulting in oxidation of functionalgroups and loss of activity of biomolecules, includ-ing enzymes. Glutathione (GSH) is a vital intra- andextracellular thiol anti-oxidant. It plays a crucialrole against oxidative stress and inflammation andcan promote the repair of injured cells and tissues,enabling the lung to withstand further injury.6

Extracellular GSH is increased by ozone exposure7

and intracellular GSH is depleted.8

The lungs thus may be rendered more susceptiblefollowing ozone exposure to inhaled agents such asallergens, viruses and pollutants.9 Regulation of theoxidant anti-oxidant balance and in particular theanti-oxidant GSH may play an important role inmaintaining the integrity of the respiratory epithe-lium.

The aim of this study was to investigate epithelialpermeability, inflammation and oxidant anti-oxi-dant balance in the airspaces of healthy non-smokers exposed to ozone.

Methods

Subjects

Fifteen healthy non-smoking subjects (M:F 12:3)were recruited. None of the subjects had any

Table 1 Study design.

Subject Exposure type

Filtered air Ozone 100 ppb

1 h post 1 h post 6

1 DTPA BAL23 DTPA BAL B4 BAL DTPA B56 DTPA BAL B7 BAL DTPA B8 BAL DTPA9

1011 DTPA BAL12 BAL DTPA B13 DTPA BAL1415 BAL DTPA

Fifteen subjects in total were studied. Five of the 15 subjects werwere exposed to ozone 400 ppb ozone with 99mTc-DTPA or BAL pabove studies were performed on different occasions at least 5 d

significant medical problems, nor any history ofrespiratory infection within 6 weeks of the study.

Study design

Details of the study design are given in Table 1.

Exposure

Subjects were exposed to filtered air (FA) or ozonemeasured in parts per billion (ppb).

Studies performed

The clearance rates of 99mtechnetium labelleddiethylene-triamine-penta-acetate (99mTc-DTPA)from the lungs and bronchoscopy and bronchoal-veolar lavage (BAL) were performed.

Exercise protocol

A standard exercise protocol on a bicycle erg-ometer was used (15min exercise, 15min rest,15min exercise and 15min rest for a total of 1 h).The level of exercise was set to achieve aninspiratory minute ventilation (VI) of 40 lmin�1 asassessed in a previous study breathing FA.

Ozone 400 ppb

h post 1 h post 6 h post

BAL DTPABAL DTPA BAL DTPA

AL DTPAAL DTPA

DTPA BAL BAL DTPAAL DTPAAL DTPA

BAL DTPABAL DTPADTPA BAL

AL DTPA

BAL DTPA BAL DTPA

e exposed to FA, six were exposed to ozone 100 ppb and sevenerformed 1 or 6 h after exposure as indicated above. All theays apart.

ARTICLE IN PRESS

Ozone and airspace epithelium 2229

Carboxyhaemoglobin (COHb)

COHb (IL 282 Co-oximeter Instrumentation Labora-tory, Lexington, MA, USA) was measured prior torecruitment and on each visit prior to study toconfirm the history of non-smoking.

Spirometry

Spirometry (Vitalograph Ltd., Buckingham, Eng-land) was performed before and immediately afterall exposures to FA or ozone and either 1 or 6 h afterexposure.

Ozone exposure

Compressed air was purified and filtered throughPurafil (Purafil Corporation, Sutton Coldfield, Eng-land) and activated carbon filters (Spirax MonnierIX1, Spirax Sarco, Cheltenham, England), whichremoved oxides of nitrogen and hydrocarbons andthen through a HEPA filter (Spirax Monnier IC3A). Itwas warmed and humidified and passed alongTeflon tubing at 140 lmin�1 to a perspex face maskand shield worn by the subject. All fittings incontact with ozone were made of either stainlesssteel or Teflon. Ozone was produced by passingmedical grade oxygen through an electric arcozonator Type BA. 023 (Wallace and Tiernan,Tonbridge, Kent, England) and was injected intothe air stream at a sufficient distance from thesubject to ensure good mixing of the ozone. Asample of the air ozone mixture was drawn fromthe mask and monitored continuously by an ozoneanalyser (Series 49-100/103, Thermoelectron Cor-poration, Warrington, England), calibrated withreference to a primary ozone source (Institute ofTerrestrial Ecology, Edinburgh University, Scot-land).

99mTc-DTPA lung clearance

99mTc-DTPA lung clearance is a measure of airspaceepithelial permeability. Each subject inhaled 1200megabecquerels (MBq) of nebulised 99mTc-DTPAfrom an Ultravent nebuliser (Mallinkrodt MedicalLtd., Petten, Holland) at a radioactive concentra-tion of 1.8 GBq/2ml. This was prepared fromsodium pertechnetate (99mTc) injection obtainedfrom an Amertec II generator (Code MCC20) and aPentetate II kit (Code N108), both supplied byAmersham International (Aylesbury, England). Themass median aerodynamic diameter (7SD) of theparticles generated was 0.5970.04 mm with ageometric standard deviation of 1.7970.14 mm, as

measured by a seven stage cascade impactor. Aflow of 12 lmin�1 O2 was used to generate theaerosol which subjects inhaled for 2min, whilesupine and wearing a noseclip, during normal tidalbreathing, to prevent proximal deposition throughturbulent air flow. Respiratory rate was countedduring the 2min inhalation period. Subjects werethen imaged supine using a Siemens gamma camerapositioned posteriorly with a 140 keV low energy,all purpose collimator (Siemens plc., Bracknell,England) linked to a Bartec computer (BartecMedical Systems Ltd., Farnborough, England) anda Unix Sun workstation (Sun Microsystems Inc.,Camberley, England) with Micas System V software(Nodecrest Ltd., Byfleet, England). Counts wereacquired in 30 s time frames for 30min at aresolution of 128� 128 pixels. An intravenousinjection of 20MBq 99mTc-DTPA at a concentrationof 50MBq/2.5ml was given at 20min to correct forbackground activity.

Image analysis was performed using an inter-renal background region of interest, normalised forarea and corrected for 99mTc decay.10 No activitywas seen to accumulate in the region of the thyroidgland. Monoexponential lung clearance was ob-served and the time for lung activity to fall to 50%of the initial value (t50) was calculated by linearregression analysis of the first 20min of thecorrected lung curves. All correlation coefficientswere greater than 0.92 for the semi-logarithmictime–activity curves.

The stability of the 99mTc-DTPA was confirmed bythin layer chromatography of the residual 99mTc-DTPA in the nebuliser and urine.

Bronchoscopy and BAL

BAL (240mls) and BAL fluid (BALF) processing wereperformed according to a standard technique.11

Cell counts in BALF were carried out with ahaemocytometer and viability was ascertained byexclusion of Trypan blue. Cell differentials wereperformed on cytospin preparations (Shandon,Pittsburgh, PA, USA) stained with Diff Quik (MerzDade, Switzerland).

Biochemical assays

Superoxide anion generation by mixed BAL leuco-cytes (2.5� 105 cells) was measured by the super-oxide dismutase (SOD) inhibitable reduction offerricytochrome c using a Pye Unicam 8700 seriesUV/VIS spectrophotometer.12

The level of plasma or BALF lipid peroxides wasmeasured as thiobarbituric acid reactive sub-

ARTICLE IN PRESS

cts.

Post

0h

Post

1h

Age

Pre

Post

0h

Post

6h

70.3

4.07

0.3

3.97

0.3

76.3

112.07

7.3

108.87

4.9

70.4

5.07

0.4

4.87

0.4

70.3

4.27

0.4n¼

44.37

0.3

34.674.1

4.27

0.4

4.27

0.4

4.17

0.4

72.8

110.87

3.5

108.47

2.8

113.47

2.6

116.67

2.0

110.27

1.4

70.5

4.97

0.5

5.07

0.5

4.97

0.5

4.97

0.6

4.77

0.5

70.2

4.17

0.4n¼

44.07

0.3

37.273.7

4.17

0.2

3.97

0.2

3.87

0.2

73.3

9979.3

9478.2

1077

6.5

103.47

8.5

100.07

7.5

70.3

5.27

0.5

4.87

0.4

5.17

0.3

4.97

0.4

4.77

0.3

before(pre)an

dafter(post0,

1an

d6h)

expo

sure

toozon

e0,

100or

400ppbfor1hduringinterm

ittent

exercise

atVI

espirom

etry

orag

ein

anyof

thegrou

psstud

ied,no

rweretherean

ysign

ifica

ntdifferenc

esin

spirom

etry

afterozon

e

D. Morrison et al.2230

stances (TBARS).13 EDTA (Sigma Chemical Co.,Poole, England) at a final concentration of 1mMwas added to BALF prior to storage at �70 1C.

The anti-oxidant capacity of plasma and BALFwas measured as the Trolox equivalent anti-oxidantcapacity (TEAC) as described previously.14

Reduced and oxidised glutathione (GSSG) in BALFand mixed BAL leucocytes (2� 106 cells) wasmeasured according to the method of Tietze withslight modifications.15

Serum albumin was assayed by a standardtechnique in the hospital biochemistry department.Albumin in BALF was measured by an immunoturbi-dometric method (Boehringer Mannheim, Lewes,England). Concentrations of BALF constituentswere estimated in epithelial lining fluid (ELF) usingthe albumin method, assuming its level in ELF to be10% of the plasma albumin level.16

Statistics

The effects of 100 and 400 ppb ozone werecompared with 0 ppb, separately for the 1 and 6 hstudies, by one-way analysis of variance with a posthoc Scheffe test. Data from the 1 and 6 h studieswere compared by independent t-test. 99mTc-DTPAlung clearance for the right and left lungs and pre-and post-values for lipid peroxidation and TEACwere compared by paired t-test. Correlation wasperformed by a two-variable parametric analysis.

Ethical permissionEthical permission was obtained from the localmedical ethics committee and all subjects gaveinformed written consent.

Table

2Effectsof

ozon

eon

spirom

etry

inno

rmal

subje

Ozone

(ppb)

Spirom

etry

Age

Pre

0FE

V1(l)

35.474.7

4.0

FEV1%predicted

112.2

FVC(l)

4.9

100

FEV1(l)

35.074.3

4.4

FEV1%predicted

109.6

FVC(l)

5.2

400

FEV1(l)

33.272.2

4.4

FEV1%predicted

104.6

FVC(l)

5.4

Age

andspirom

etry

(mea

n7SE)in

15he

althyno

n-sm

okingsubjects

40lm

in�1.The

rewereno

sign

ifica

ntdifferenc

esbetwee

nbaselin

exposure.

Results

Subject characteristics and spirometry

There were no significant differences betweenbaseline spirometry or age between the five groupsof subjects (Table 2). No significant effects of ozoneon spirometry were seen immediately, 1 or 6 h afterexposure. The maximum fall in the mean FEV1 of0.4 l was produced by 400 ppb immediately and 1 hafter exposure. FVC followed the same patternwith a similar magnitude of change.

99mTc-DTPA lung clearance

99mTc-DTPA lung clearance did not change signifi-cantly after exposure to ozone 100 or 400 ppb ateither 1 or 6 h after exposure (Fig. 1).

ARTICLE IN PRESS

100

75

50

25

01 hour 6 hours

t50

(min

s)

Time after ozone exposure

Air

100 ppb

400 ppb

Figure 1 99mTc-DTPA lung clearance (t50 mean7SE, n ¼ 5in each group) in healthy non-smoking subjects after 1and 6 h exposure to ozone 0, 100 or 400 ppb for 1 h duringintermittent exercise at VI 40 lmin�1. There was nosignificant difference between 0, 100 and 400 ppb norbetween the 1 and 6 h studies.

Table 3 Effects of ozone inhalation on bronch-oalveolar lavage neutrophil counts.

Percentage Number (� 106)

1 h post0 ppb 0.870.2 0.1170.04100 ppb 0.670.2 0.0670.03400 ppb 0.870.5 0.1170.07

6 h post100 ppb 0.670.1 0.0870.03400 ppb 2.470.7� 0.3170.09y

BAL neutrophils (% and absolute numbers � 106) in 15healthy non-smoking subjects (mean7SE) 1 and 6 h afterexposure to ozone 0, 100 or 400 ppb for 1 h duringintermittent exercise at VI 40 lmin�1.�Po0:05 compared with neutrophils after exposure for

1 h to air and neutrophils 6 h after exposure to 100 ppb O3.yPo0:05 compared with neutrophils 6 h after 100 ppb

ozone.

Ozone and airspace epithelium 2231

Bronchoscopy and BAL

There were no significant differences between thegroups in BAL recovery, total number of cellsobtained or their viability, which was greater than90% in all five groups. The total number of cellsobtained did not increase following ozone. Inhala-tion of ozone 400 ppb resulted in an increasedpercentage (FA, 0.870.2%, 400 ppb, 2.470.7%,Po0:05) and number (FA, 0.1170.04� 106 cells,400 ppb, 0.3170.09, Po0:05) of neutrophils in BAL6 h after exposure (Table 3).

BALF albumin did not change significantly follow-ing ozone, although 1 h after the inhalation of400 ppb the concentration rose from 29.473.6 to40.773.2 mgml�1. After 6 h it had fallen to34.774.7 mgml�1. There were no changes inplasma albumin. There was a statistically signifi-cant effect of ozone overall (Po0:05) on ELFvolume, but otherwise it followed the same patternas BALF albumin increasing from 1.1070.11 to1.6470.11mls (P ¼ 0:057) 1 h after the inhalationof 400 ppb. Six hours later the level had fallen to1.3270.18mls.

Oxidants and anti-oxidants

Superoxide anion production by mixed BAL leuco-cytes was reduced in general following inhalationof ozone 1 or 6 h after exposure. This wasstatistically significant following 100 ppb 1 h after

exposure and 6 h after exposure to 400 ppb (Fig. 2).There were no significant differences in plasmaproducts of lipid peroxidation between pre- andpost-exposure levels. There were no significantdifferences between the groups for lipid peroxida-tion in BALF. In ELF, there was a significantreduction in products of lipid peroxidation both 1and 6 h following 400 ppb (1 h post,34.471.4 mmol l

�1

, Po0:05, 6 h post,44.474.2 mmol l�1, Po0:05) compared to the con-trol group (49.975.2 mmol l

�1

).Plasma anti-oxidant capacity showed no biologi-

cally relevant changes following ozone exposure,compared to the pre-exposure values, although theslight increase from 1.2070.18 to1.2470.17mmol l�1 1 h following 400 ppb ozonewas statistically significant. Anti-oxidant capacitydid not change significantly in BALF or ELF followinginhalation of ozone at 100 or 400 ppb either 1 or 6 hafter exposure. Reduced GSH in mixed BAL leuco-cytes, BALF or ELF were also unchanged followinginhalation of ozone at 100 or 400 ppb, either 1 or6 h after exposure, although there was a trend incells, BALF and ELF for GSH to decrease 6 h after400 ppb ozone. There were no significant changes inGSSG in cells, BALF or ELF following ozone. Thelevels were close to the limit of detection of theassay.

There were no significant correlations between99mTc-DTPA clearance and any of the variablesmeasured in particular with the percentage ornumber of neutrophils, BALF albumin, ELF volume,O2� production or products of lipid peroxidation in

ELF.

ARTICLE IN PRESS

5

10

15

20

25

01 hour 6 hours

(nm

ol 2

.5 x

107

cells

/120

min

s)S

up

erox

ide

rele

ase

Time after ozone exposure

Air

100 ppb

400 ppb

* **

Figure 2 Superoxide anion (O2�) release from mixed BAL

leucocytes in healthy non-smoking subjects (mean7SE,n ¼ 5 in each group) 1 and 6 h after exposure to ozone 0,100 or 400 ppb for 1 h during intermittent exercise at VI of40 lmin�1. �Po0:05 and ��Po0:01 compared with breath-ing air for 1 h.

D. Morrison et al.2232

Discussion

Ozone has been shown to produce increasedrespiratory epithelial permeability in vivo and invitro.1,2 In ozone exposed rats, tracers localise inthe intercellular spaces of epithelial cells, althoughnot in tight junctions, and in endocytic vesicles inall parts of the epithelial cells.17 In a subsequentstudy, total protein (TP), alkaline phosphatase (AP)and fibronectin in BALF showed a similar timerelated increase suggesting a role for inflammationand injury in increased epithelial permeability. Theearlier return of AP levels to baseline compared toTP was felt by the authors to suggest repair ofinjured cells but a delay in the formation ofepithelial junctions.2 In confluent canine bronchialepithelial cells, ozone increased epithelial perme-ability and this was partly prevented by vitamins Eand A and phalloidin which produces actin poly-merisation.1 The cytoskeleton may, thus, be apotential target for oxidant stress and ozone.18

The levels of ozone used in our study were chosento equate with ambient levels one might find in theUK (100 ppb) and in the US (400 ppb). Most studiesto date have used ozone concentrations which arenot relevant to the UK and are in fact often higherthan those encountered in the US, particularlywhen one considers the intense level of exerciseused in many studies. In our study, there were nosignificant effects of ozone on spirometry. This is incontrast to Nightingale et al.,19 who found thatexposure to 400 ppb ozone for 2 h during inter-mittent exercise for 20 out of each 30min at 50Wreduced lung function. The concentration andduration of ozone exposure used in our study did

not result in increased epithelial permeability to99mTc-DTPA. Frampton et al.,20 however, showed anincrease in epithelial permeability following ex-posure to 220 ppb ozone for 4 h during intermittentexercise for 20 out of each 30min at 25 lmin�1m�2.In our study, there was a significant effect of ozoneoverall on ELF volume suggesting the possibility ofan increase in epithelial permeability. There was,however, no significant effect of ozone on BALFalbumin another measure of epithelial permeabil-ity. Overall the results of the present study may beregarded as reassuring and indicate that the effectsof ozone are dose related, this being a product ofconcentration, duration of exposure and minuteventilation.

Krishna et al.3 exposed healthy humans to0.2 ppm for 2 h during intermittent exercise at VE30 lmin�1 and found a three-fold increase inthe proportion of neutrophils in BALF 6 h afterexposure. This was positively correlated withaccompanying increased concentrations of IL-8and Gro-alpha. Basha et al.4 studied non-smokingasthmatics and healthy volunteers exposed toozone 0.2 ppm and FA for 6 h with intermittentexercise at 5 lmin�1 (l VC�1). There was a signifi-cantly greater increase in the percentage andnumber of neutrophils per millilitre of BALF 18 hafter exposure in the asthmatics, than in thecontrol subjects exposed to ozone. This wasassociated with significantly greater increases inIL-8 and IL-6.

In our own studies, we found that the totalnumber of cells recovered in BAL did not change.However, the percentage and number of neutro-phils increased 6 h after the inhalation of 400 ppbozone. Our data are thus in keeping with that ofSchelegle et al.21 where neutrophil influx peaked at6 h in pooled BAL. In our study, in contrast to that ofSchelegle and co-workers, there was no increase inthe neutrophil BAL count 1 h after exposure.However, Schelegle and colleagues found a sig-nificant increase at 1 h in the first or proximalaliquot of BAL. Had we not pooled our aliquots ofBAL then we may also have seen neutrophil influxbeginning at 1 h.

In our studies, immediately after ozone expo-sure, cell viability fell to 85% compared to 92% incontrol cells. It recovered rapidly by 4 h andsubsequently rose to above 90%, although itremained below that of the control cells. This fallsuggests a rapid and direct toxic effect of ozone.

Following a study of rats inhaling 0.2–0.8 ppmozone for 7 h daily for up to 4 days, Donaldsonet al.22 concluded that epithelial injury at theconcentrations used was likely to be a direct effectof the ozone. In rat trachea following 3 h exposure

ARTICLE IN PRESS

Ozone and airspace epithelium 2233

to 0.8 ppm ozone, permeability to 99mTc-DTPA andBAL, it was suggested that the neutrophil influxoccurred as a consequence of the increasedpermeability, rather than causing the increasedpermeability.23 Ozone-induced permeability inrats, exposed to 0.8 ppm for 2 h, was inhibited bypretreatment with cyclophosphamide, FPL 55712(a leukotriene D4 antagonist) and indomethacin.5

Neutrophils from ozone exposed rats have beenshown to alter epithelial resistance in vitro, butonly when stimulated with PMA or FMLP.24 Theyhave also been shown to have a synergistic effecton ozone-induced airway epithelial injury whenperfused in the isolated perfused rat lung.25

Concomitant changes in permeability and re-cruitment of inflammatory cells in the lung follow-ing ozone exposure have been shown varying withozone concentration, exposure duration, mode ofanalysis, animal species, physical activity, preg-nancy and repeated exposure.9 Tumour necrosisfactor may play a role in the increased permeabilityand inflammation.26 It would seem likely thereforethat ozone produces an initial direct injury to cellmembranes, which may result in increased perme-ability, possibly also affecting macrophages, result-ing in increased release of chemotactic stimuli forneutrophils. The consequent secondary inflamma-tory response may result in further injury andincreased permeability. However, we were unableto convincingly demonstrate increased epithelialpermeability in normal subjects exposed to envir-onmentally relevant concentrations of ozone.

There are relatively few data on the effect ofozone inhalation on the oxidant/anti-oxidant bal-ance in the airspaces in humans. We found thatalthough the production of superoxide anion wasincreased in the control group, as a result ofexercise, ozone inhalation decreased superoxideanion production from BAL leucocytes. This effectof ozone supports data from previous animalstudies.22 Increased ozone-induced neutrophilicinflammation has, however, been shown in extra-cellular SOD null mice.27

Ozone inhalation has been shown to increaselipid peroxidation in humans although at a higherdose than in our study.20,28 We found no significantchanges in plasma or BALF lipid peroxides. In BALF,and in ELF, the control levels were very muchhigher than those we measured in a previous studyin resting subjects. In ELF, there was a significantreduction in TBARS both 1 and 6 h after 400 ppbozone. This decrease could be related to the anti-oxidant properties of albumin or may be adilutional effect in BALF.

Ozone causes oxidation of sulphydryl groups, e.g.GSH and those found in proteins, amines, alcohol

groups, aldehydes and other functional groups inproteins, enzymes, nucleic acids and membranes.Studies have shown that acute ozone exposureresults in an increase in extracellular GSH7 anddepletion and oxidation of intracellular GSH.8

We studied the total, functional anti-oxidantcapacity of plasma, BALF and ELF following ozoneexposure. No statistically significant changes oc-curred. Similarly, no significant changes occurred inreduced or oxidised glutathione levels in mixed BALleucocytes, BALF or ELF. The control values, notonly for glutathione but also for anti-oxidantcapacity, were generally lower than in a previousstudy in which measurements were made at rest.10

In summary, we have shown no effect followingthe inhalation of ozone 400 ppb for 1 h on epithelialpermeability during moderate intermittent exer-cise in normal subjects. There was a significantincrease in the percentage and number of neutro-phils in BALF 6 h after 400 ppb ozone. Despite this,the levels of superoxide production from mixed BALleucocytes and products of lipid peroxidation in ELFdecreased. There was, therefore, no evidence ofincreased oxidant stress accompanying the neutro-phil influx into the airspaces.

Acknowledgements

This study was supported by the British LungFoundation and the Chest, Heart and StrokeAssociation (Scotland). The authors wish to thankJan Stolk, Department of Pulmonology, Leiden forproviding the Teflon tubing and the perspex facemask and shield, the Institute of OccupationalMedicine, Edinburgh for the loan of the ozonegenerator and analyzer, Alison Langridge, Xiao YangLi and David Brown, Rayne Laboratory, MedicalSchool, Edinburgh University for technical assis-tance. Statistical advice was given by Dr. TJ Peters,Senior Lecturer in Public Health Medicine, Depart-ment of Social Medicine, University of Bristol,Bristol.

References

1. Yu X-Y, Takahashi N, Croxton TL, Spannhake EW. Modulationof bronchial epithelial cell barrier function by in vitro ozoneexposure. Environ Health Perspect 1994;102:1068–72.

2. Bhalla DK, Gupta SK, Reinhart PG. Alteration of epithelialintegrity, alkaline phosphatase activity, and fibronectinexpression in lungs of rats exposed to ozone. J ToxicolEnviron Health A 1999;57:329–43.

3. Krishna MT, Madden J, Teran LM, et al. Effects of 0.2 ppmozone on biomarkers of inflammation in bronchoalveolarlavage fluid and bronchial mucosa of healthy subjects. EurRespir J 1998;11:1294–300.

ARTICLE IN PRESS

D. Morrison et al.2234

4. Basha MA, Gross KB, Gwizdala CJ, Haidar AH, Popovich Jr J.Bronchoalveolar lavage neutrophilia in asthmatic andhealthy volunteers after controlled exposure to ozone andfiltered purified air. Chest 1994;106:1757–65.

5. Bhalla DK, Daniels DS, Luu NT. Attenuation of ozone inducedairway permeability in rats by pretreatment with cyclopho-sphamide, FPL 55712 and indomethacin. Am J Respir CellMol Biol 1992;7:73–80.

6. Rahman I, MacNee W. Regulation of redox glutathione levelsand gene transcription in lung inflammation: therapeuticapproaches. Free Radical Biol Med 2000;28:1405–20.

7. Blomberg A, Mudway IS, Nordenhall C, et al. Ozone-inducedlung function decrements do not correlate with early airwayinflammatory or antioxidant responses. Eur Respir J1999;13:1418–28.

8. Van der Zee J, Dubbelman TM, Raap TK, Van Steveninck J.Toxic effects of ozone on murine L929 fibroblasts. Enzymeinactivation and glutathione depletion. Biochem J1987;242:707–12.

9. Bhalla DK. Ozone-induced lung inflammation and mucosalbarrier disruption: toxicology, mechanisms, and implica-tions. J Toxicol Environ Health B Crit Rev 1999;2:31–86.

10. Morrison D, Rahman I, Lannan S, MacNee W. Epithelialpermeability, inflammation, and oxidant stress in the airspaces of smokers. Am J Respir Crit Care Med1999;159:473–9.

11. Walters EH, Gardiner PV. Bronchoalveolar lavage as aresearch tool. Thorax 1991;46:13–618.

12. Johnston Jr RB, Keele Jr BB, Misra HP, et al. The role ofsuperoxide anion generation in phagocytic bactericidalactivity. Studies with normal and chronic granulomatousdisease leukocytes. J Clin Invest 1975;55:1357–72.

13. Yagi K. A simple fluorometric assay for lipoperoxide in bloodplasma. Biochem Med 1976;15:212–6.

14. Rahman I, Morrison D, Donaldson K, MacNee W. Systemicoxidative stress in asthma, COPD, and smokers. Am J RespirCrit Care Med 1996;154:1055–60.

15. Rahman I, Li XY, Donaldson K, Harrison DJ, MacNee W.Glutathione homeostasis in alveolar epithelial cells in vitroand lung in vivo under oxidative stress. Am J Physiol1995;269:L285–92.

16. Ward C, Duddridge M, Fenwick J, et al. Evaluation ofalbumin as a reference marker of dilution in bronchoalveolarlavage fluid from asthmatic and control subjects. Thorax1993;48:518–22.

17. Bhalla DK, Crocker TT. Tracheal permeability in rats exposedto ozone. An electron microscopic and autoradiographicanalysis of the transport pathway. Am Rev Respir Dis1986;134:572–9.

18. Bellomo G, Mirabelli F, Vairetti M, Iosi F, Malorni W.Cytoskeleton as a target in menadione-induced oxidativestress in cultured mammalian cells. 1. Biochemical andimmunocytochemical features. J Cell Physiol1990;143:118–28.

19. Nightingale JA, Rogers DF, Fan Chung K, Barnes PJ. No effectof inhaled budesonide on the response to inhaled ozone innormal subjects. Am J Respir Crit Care Med2000;161:479–86.

20. Frampton MW, Pryor WA, Cueto R, Cox C, Morrow PE, UtellMJ. Ozone exposure increases aldehydes in epithelial liningfluid in human lung. Am J Respir Crit Care Med1999;159:1134–7.

21. Schelegle ES, Siefkin AD, McDonald RJ. Time course of ozoneinduced neutrophilia in normal humans. Am Rev Respir Dis1991;143:1353–8.

22. Donaldson K, Brown GM, Brown DM, Slight J, MacLaren W,Davis JMG. Characteristics of bronchoalveolar lavage leuco-cytes from the lungs of rats inhaling 0.2–0.8 ppm of ozone.Inhalation Toxicol 1993;5:149–64.

23. Bhalla DK, Young C. Effects of acute exposure to O3 on rats:sequence of epithelial and inflammatory changes in thedistal airways. Inhalation Toxicol 1992;4:17–31.

24. Bhalla DK, Rasmussen RE, Daniels DS. Adhesion and motilityof polymorphonuclear leucocytes isolated from the blood ofrats exposed to ozone: potential biomarkers of toxicity.Toxicol Appl Pharmacol 1993;123:177–86.

25. Joad JP, Bric JM, Pino MV, Hyde DM, McDonald RJ. Effects ofozone and neutrophils on function and morphology of theisolated rat lung. Am Rev Respir Dis 1993;147:1578–84.

26. Bhalla DK, Reinhart PG, Bai C, Gupta SK. Amelioration ofozone-induced lung injury by anti-tumor necrosis factor-alpha. Toxicol Sci 2002;69:400–8.

27. Jonsson LM, Edlund T, Marklund SL, Sandstrom T. Increasedozone-induced airway neutrophilic inflammation in extra-cellular-superoxide dismutase null mice. Respir Med2002;96:209–14.

28. Montuschi P, Nightingale JA, Kharitonov SA, Barnes PJ.Ozone-induced increase in exhaled 8-isoprostane in healthysubjects is resistant to inhaled budesonide. Free Radical BiolMed 2002;33:1403–8.