Acute exposure to traffic-related air pollution alters antioxidant status in healthy adults

Abstract

Background.

Exposure to traffic-related air pollution is associated with an increased risk of cardiovascular and respiratory disease. Evidence suggests that inhaled pollutants precipitate these effects via multiple pathways involving oxidative stress.

Objective.

Postulating that a decrease in circulating antioxidant levels reflect an oxidative response, we investigated the effect of inhaled diesel exhaust (DE) on the ratio of reduced to oxidized glutathione (GSH/GSSG) in healthy adults, and whether pre-exposure antioxidant supplementation blunted this response. We also examined exposure-related changes in antioxidant/stress response leukocyte gene expression (GCLc, HMOX-1, IL-6, TGFβ) and plasma IL-6 levels.

Methods.

Nineteen nonsmoking adults participated in a double-blind, randomized, four-way crossover study. Each subject completed 120-min exposures to filtered air and DE (200 μg/m³), with and without antioxidant pretreatment. Antioxidant comprised 1000 mg ascorbate for 7 days and 1200 mg N-acetylcysteine 1 day prior to exposure, with 1000 mg and 600 mg, respectively, administered 2 hours prior to exposure. Whole blood glutathione was measured pre- and post-exposure; plasma IL-6 and mRNA expression were quantified pre, during and post exposure.

Results.

Diesel exhaust exposure was associated with significantly decreased GSH/GSSG (p=0.001) and a 4-fold increase in IL-6 mRNA (p=0.01) post exposure. Antioxidant pretreatment did not significantly mediate the effect of DE exposure on GSH/GSSG, though appeared to decrease the effect of exposure on IL-6 mRNA expression.

Conclusions.

Acute DE inhalation induced detectable oxidative effects in healthy adults, which were not significantly attenuated by the selected antioxidant pre-treatment. This finding supports the premise that oxidative stress is one mechanism underlying the adverse effects of traffic-related air pollution.

INTRODUCTION

Exposure to urban air pollution is associated with a range of adverse health effects, most notably increased cardiovascular disease and mortality.¹ Air quality is of particular concern in densely populated urban centers, where commercial transport and traffic congestion can drive pollutant levels well above national standards. Vehicle exhaust emissions predominately generate fine particulate matter (PM_2.5), a complex composite of organic and inorganic particles, metals, acids and other reactive species small enough to be inhaled into the deepest part of the lung. From there, respirable particles may initiate a number of pathophysiological pathways, including sensory receptor activation of the autonomic nervous system and induction of a pulmonary and systemic inflammatory response. ^{1, 2} Oxidative stress, an established risk factor for a number of inflammatory, cardiovascular and age-related conditions, plays an integral role in each of these pathways.³

Evidence suggests that inhaled pollutants precipitate these effects via multiple pathways involving oxidative stress. Previous controlled exposure studies have demonstrated strong associations between inhaled pollutants and markers of pulmonary oxidative stress,^4–7 yet few have directly examined circulating markers of an oxidative response. Postulating that oxidative stress is a mechanism underlying the pathophysiological effects of PM_2.5, we investigated the effect of an acute exposure to diesel exhaust (DE), the dominant source of urban PM_2.5, on the measurement of reduced to oxidized glutathione (GSH/GSSG and redox potential, E_h), in healthy adults. We also investigated whether pre-treatment with an antioxidant cocktail modified the effect of inhaled DE on antioxidant capacity and examined exposure-related changes in antioxidant/stress response leukocyte gene expression (GCLc, HMOX-1, IL-6, TGFβ) and plasma IL-6 levels.

MATERIALS AND METHODS

Twenty healthy adults participated in four exposure sessions in a randomized, crossover design. Qualification for enrollment included body mass index <30 kg/m², and no evidence or history of hypertension (blood pressure <130/85 mmHg), asthma, diabetes, hypercholesterolemia, cardiovascular illness, or other chronic medical condition, based on spirometry, fasting glucose, lipid panel, electrocardiogram and questionnaire. Age, sex, race/ethnicity, smoking, and medication use were self-reported. Subjects who were smokers or regularly exposed to second-hand smoke were excluded. Smoking status was confirmed by urinary cotinine (CAS-COT kit, Innovacon, Inc.; San Diego, CA) in samples collected prior to exposure. One subject was excluded from the analysis based on a positive cotinine screening (>200 ng/mL) leaving nineteen subjects in these analyses. Women were scheduled for exposures only during the first two weeks of a menstrual cycle, and pregnancy was ruled out prior to each exposure with a urine pregnancy test.

To test the hypothesis that antioxidant prophylaxis may attenuate the oxidative effects of DE, subjects were given oral N-acetylcysteine (NAC), a cysteine precursor in GSH synthesis, and ascorbate (vitamin C) prior to exposure in a crossover design. This treatment delivers both an enzymatic and non-enzymatic antioxidant; the selected antioxidants have been shown to have a beneficial effect on markers of inflammation. The treatment consisted of 500 mg ascorbate or matched placebo every 12 hours for seven days prior to exposure, and 600 mg NAC or matched placebo every 12 hours one day prior to exposure. Subjects were administered an additional 1000 mg ascorbate and 600 mg NAC (or placebo) upon arrival at each session, approximately one hour before exposure commencement. The antioxidant and matched placebo were prepared by the Investigational Drug Service at the University of Washington Medical Center. Subjects were instructed not to consume supplemental vitamins or vitamin C-fortified food and drink the week prior to exposure. Plasma ascorbate, from blood drawn prior to each exposure session, was used to confirm compliance with antioxidant treatment.

Exposure System

The complete exposure system has been previously described in detail.⁸ In brief, DE was generated using a 2002 model turbocharged direct-injection 5.9-L Cummins B-series engine in a 100 kW generator set, running at steady state prior to the exposure (6BT5.9G6; Cummins, Inc., Columbus, IN). Load was maintained at 75% of rated capacity, using a load-adjusting load bank (Simplex, Springfield, IL), no. 2 undyed on-highway low sulfur diesel fuel and Valvoline 15W-40 crankcase oil (Lexington, KY). Carbon matrix and HEPA filtered (99.99% efficient) ambient air was used for filtered air (FA) sessions and for DE dilution. Emissions were diluted in two phases to achieve a 200 μg/m³ PM_2.5 concentration in the 116 m³ exposure room breathing zone (average 205.4 μg/m³; SD 5.4 μg/m³). PM_2.5 concentrations were assessed in real-time throughout each exposure using a tapered element oscillating microbalance (1400a PM_2.5, Rupprecht & Patashnick Co., Albany, NY) and were continuously adjusted, based on nephelometry measurements, with a feedback control system. The exposure room was maintained at a temperature of 20 to 21°C with 50% relative humidity. Based on multistage impactor-collected samples, the typical particle count was 2.8 × 10³ for FA and 5.3 × 10⁵ per cm³ for DE exposures, based on one-minute averages; DE particle mass median diameter was 0.080 μm. Nephelometers positioned within the exposure room were used to confirm spatial uniformity of particle concentrations. Average nitrogen dioxide concentrations during DE exposures were 35 ppb (approximately 1.5% of total NO_x). Concentrations of carbon monoxide averaged 0.30 ppm for FA and 0.80 ppm for DE.

Exposure Sessions

Each subject was exposed for 120 minutes, on separate days, in a placebo-controlled four-way crossover design, randomized to order of the four conditions: DE + placebo, FA + placebo, DE + antioxidant, FA + antioxidant. To reduce carry-over effect, a minimum 2-week washout period separated each session. Researchers followed set and scheduled protocols to limit any heterogeneity between sessions. Subjects were instructed to fast for 10 hours prior to exposure. Upon arrival at the University of Washington’s Clinical Research Center (CRC), vitals were recorded, an intravenous catheter was placed in the subject’s left arm and pre-exposure (baseline) blood was collected. Exposures began at approximately 9:00 am. Blood was collected again 90 minutes into the exposure (during exposure), and five hours post exposure. Following exposure, the subject received an identical defined composition meal (to limit any short-term dietary effects) and rested at the CRC for at least 6 hours until release. Subjects returned to the CRC the next morning, approximately 24 hours after baseline measurements were taken, for a final assessment and blood draw.

All subjects, researchers and technicians (other than the exposure engineer) were blinded to the exposure condition. Subject blinding was evaluated by asking the subject to estimate the level of DE (as high, medium, or none) during each exposure. The Human Subjects Division of the University of Washington approved subject consent forms and all study protocols.

Ascorbate assay

Blood was collected pre-exposure in a BD Vacutainer tube containing sodium heparin and centrifuged to separate plasma. 50 μl of meta phosphoric acid/dithioreitol (MPA/DTT) was added to 500 μl plasma. In this assay, ascorbic acid is oxidized to dehydroascorbic acid by ascorbic acid oxidase and dehydroascorbate is converted to the quinoxolone derivative by a fast reaction with 0-phenylenediamine at pH 6.5. Absorbance (340 nm) of the product is directly proportional to the ascorbic acid concentration. The assay was performed using an Olympus AU400 chemistry immuno analyzer (Center Valley, PA). Reagents were purchased from Sigma-Aldrich (St. Louis, MO).

Glutathione assay

Blood was collected pre-exposure and 5 hours-post exposure into a BD Vacutainer tube containing sodium heparin. Concentrations of total glutathione (GSH_T) and oxidized (GSSG) glutathione in whole blood were determined via enzymatic reaction with Ellman’s reagent (5,5′-dithiobis-2-nitrobenzoic acid [DTNB]) and spectrophotometric quantification at 412 nm using a Microplate Assay for GSH/GSSG (Oxford Biomedical Research; Oxford, MI). To prevent oxidation of GSH, samples for GSSG quantification were treated with a thiol-scavenging reagent prior to freezing at −80°C. Concentrations were determined based on an eight-point standard curve for GSH_T and GSSG. GSH/GSSG was calculated as GSH_T – 2GSSG / GSSG. The redox state values (E_h) were calculated using the Nernst equation: E_h = E_o + RT/nF ln [GSSG]/[GSH]², where E_o is the standard potential for the redox couple, R is the gas constant, T is the absolute temperature, n is the number of electrons transferred (2), and F is Faraday’s constant. The standard potential E_o for the GSH/GSSG couple used was −264 mV for pH 7.4.⁹

Genotyping

Effect modification by genotype for glutathione S-transferase mu 1 (GSTM1) was also included in our analysis, as the null polymorphism has been associated with decreased detoxification capacity and increased susceptibility to the effects of air pollution in other research.¹⁰ Blood was collected at screening in a BD Vacutainer tube containing the preservative sodium citrate. DNA was isolated from blood using Qiagen’s DNeasy kit (Valencia, CA) and genotyped for GSTM1 (deletion/no deletion) using a multiplex PCR, with ß-globin as a positive control.¹¹ Specific probes were 3’-labeled with TAMRA quencher dye; wild type and variant probes were 5’-labeled with 6-FAM and VIC reporter dye, respectively. Sequencing reactions were performed and analyzed on a 7900 Fast Real-Time PCR System (Applied Biosystems).

Gene expression

Blood was collected pre-exposure (baseline), during exposure, 5 hours post exposure and 24 hours from baseline in a BD Vacutainer CPT glass molecular diagnostics tube containing density gradient polymer gel and sodium citrate. Cells were pelleted by centrifugation and resuspended in TRIzol for RNA extraction. RNA was isolated from peripheral blood mononuclear cells (PMBCs) using an RNeasy kit (Qiagen); complementary DNA was synthesized with Oligo-dT primers and reverse transcriptase (Qiagen), using the same total RNA concentration for each sample. Expression levels were measured by quantitative real-time PCR (RT-qPCR) using Taqman primers and the ABI 7900 Sequence Detection System (Applied Biosystems). Samples were run in duplicate or triplicate. The threshold cycle C(t) was calculated using a standard curve; expression was normalized to the reference gene β-actin.

Plasma IL-6

Plasma IL-6 was assayed using the Quantikine High Sensitivity Human IL-6 enzyme linked immunosorbent assay (ELISA) kit (R&D Systems; Minneapolis, MN). As stated above, blood was collected at pre-exposure (baseline), 5 hours post exposure and 24 hours from baseline. Briefly, 100 μL of standard and sample were incubated on a 96-well microplate for two hours, followed by washing and two hours incubation with IL-6 conjugate. Substrate solution was added to the samples and incubated for one hour, followed by 30 minutes incubation of amplifier solution. Plates were read at 490 nm with the SpectraMax Plus (Molecular Devices; Sunnyville, CA) immediately after addition of stop buffer. Plasma IL-6 concentrations were log transformed for statistical analysis.

Statistical Analysis

Data were analyzed using a multivariate mixed model to assess the effect of DE exposure, antioxidant supplementation and an exposure x antioxidant interaction on outcome measures, by subject. Outcome values were baseline corrected by subtracting the same session’s pre-exposure measurement; the baseline corrected change for DE exposure was then compared with the analogous value in the FA exposure, to give a subject-specific measure of effect. Antioxidant vs. placebo values were evaluated similarly. The model was fit via linear mixed effects to estimate the mean effect, accounting for random variations in session-specific baselines by individual.

Effect modification was examined by GSTM1 genotype and pre-exposure GSH/GSSG. Descriptive analyses, paired t-tests and linear regression were used in the initial analyses. For expression analysis, we adjusted using the Benjamini and Hochberg method¹² for False Discovery Rate to account for multiple comparisons. Statistical significance was considered at α = 0.05. Results are reported as subject-specific change, or mean ± SE, unless otherwise noted. All statistical analyses were completed using Stata/IC v 12.1 (StataCorp, College Station, TX).

RESULTS

Subject characteristics are summarized in Table 1. Frequency of the GSTM1 null genotype (47%) in our study was similar to population distributions described in the literature, which range from 23–62% among studies conducted in the U.S.^13,14 Pre-exposure plasma ascorbate levels were significantly increased with antioxidant supplementation (0.9 ± .05 mg/dL vs 1.4 ± .08 respectively, p< 0.000), confirming subject compliance with treatment. Mean pre-exposure GSH/GSSG ratios were similar to those reported in other studies of healthy adults and did not differ by sex. Pre-exposure GSH/GSSG was, on average, increased by antioxidant treatment, however the effect was not significant (143; 95% CI: −56, 343; p=0.16). Redox potential did not differ by antioxidant treatment at baseline or following exposure to FA.

Table 1.

Characteristics of study participants

Characteristic
No.	19
Age, yrs	29 ± 9
Male, n (%)	14 (74)
Caucasian, n (%)	16 (84)
Body mass index	23 ± 2.2
SBP, mmHg	115 ± 13
DBP, mmHg	70 ± 9.7
Heart Rate, bpm	65 ± 9.0
Total cholesterol, mg/dL	163 ± 30
LDL	101 ± 20
HDL	46 ± 11
Triglycerides	79 ± 46
Glucose, mg/dL	90 ± 4.6
Ascorbate, mg/dL	.90 ±.05
GSTM1 +	10 (53%)

All values reported are mean (± SD) unless otherwise noted.

The mean effect of DE exposure (measured as change from baseline) on blood GSH/GSSG is shown in Figure 1. Compared with FA, GSH/GSSG ratio was significantly decreased with exposure to DE (p=0.003). This effect was not significantly attenuated by antioxidant pre-supplementation; (reduction in DE effect by 40; p=0.71) and when examined separately, the DE-related change in GSH/GSSG was similar in both placebo and antioxidant groups (Figure 1). GSTM1 genotype did not modify the effect of DE, with or without antioxidant pre-treatment. When examining the change using the redox potential (E_h), DE exposure elicits an increase in values (5.5 mV; p=0.06); there was no difference in the DE effect between antioxidant/placebo treatment groups (−0.2 mV; p=0.95; Figure 2).

Figure 1. Mean Change in GSH/GSSG Ratio from Baseline by Exposure and Antioxidant Treatment.

Baseline adjusted difference in GSH/GSSG 3 hours following 2-hour exposure to diesel exhaust (DE) and filtered air (FA). Mean subject specific effect relative to FA control, shown for exposures with placebo and with antioxidant pretreatment. Error bars represent 95% CI. * p=0.003 compared with FA control.

Figure 2. Mean Change in Redox Potential from Baseline by Exposure and Antioxidant Treatment.

Baseline adjusted difference in E_h following exposure to diesel exhaust (DE) and filtered air (FA). Error bars represent 95% CI.

In addition, we examined DE-related changes in the expression of select antioxidant, inflammatory, and stress response genes (GCLc, IL-6, TGFβ, HMOX-1) and plasma IL-6 concentrations. Compared to FA exposures, mRNA expression of IL-6 was significantly increased over four-fold (0.67, p=0.02, Figure 3) five hours post DE exposure start in placebo exposures. This effect was not detected with antioxidant pretreatment (0.16, p=0.58), however the interaction was not significant. There were no significant differences in IL-6 expression detected at other time points measured in treated or placebo groups. Plasma IL-6 protein levels mirrored changes in mRNA expression in the placebo group, but the effect detected did not reach statistical significance (Figure 4). DE exposure was not associated with expression changes in GCLc, HMOX-1, or TGFβ at any time points examined.

Figure 3. Mean Diesel Exhaust Effect on Changes in Gene Expression.

Mean difference between change (from pre-exposure) in normalized leukocyte mRNA expression with DE compared with FA control for exposures with antioxidant (hatched) pre-treatment and without (placebo, solid) for each time point measured (exp=during exposure; 5 = 5 hrs post; 24 = 24 hours from baseline). Error bars represent 95% CI. * p=0.02 compared with FA control.

Figure 4. Mean Diesel Exhaust Effect on Changes in IL-6 protein levels.

Mean difference between change (from pre-exposure) in plasma IL-6 protein with DE exposure compared with FA control, with placebo and antioxidant pre-treatment for each time point measured (exp=during exposure; 5 = 5 hrs post; 24 = 24 hours from baseline). Error bars represent 95% CI.

As with prior experiments in our lab¹⁵, subject blinding to DE exposure type was effective; only three subjects identified the correct exposure more than twice in their four visits.

DISCUSSION

In healthy adults, an acute exposure to DE significantly decreased the ratio of reduced to oxidized glutathione in whole blood, and increased circulating PBMC gene expression of the pro-inflammatory cytokine IL-6, in the period following exposure. Antioxidant pretreatment did not attenuate the DE-effect on GSH/GSSG, however it notably reduced the exposure-related increase in IL-6 expression. These findings support the hypothesis that oxidative stress is one mechanism underlying the pathophysiological effects of exposure to traffic-related air pollution, and suggest that acute exposures may reduce the capacity of an essential cellular defense system.

Oxidative stress occurs when reactive oxidants, most commonly oxygen and nitrogen species (herein referred to as ROS), exceed the antioxidant capacity of the cell.¹⁶ PM_2.5 may induce oxidative stress directly, via reactive electron donors, or indirectly, as a secondary effect of inflammatory signaling or target cell interactions (i.e. altered mitochondrial function).^17–19 Under conditions of surplus, ROS can induce alterations in redox-sensitive functions and signaling, leading to the disruption of normal physiological responses.^20,21 For example, an excess of superoxide can severely diminish levels of the vasodilator nitric oxide (NO), an important regulator of blood pressure, endothelial homeostasis, coagulation, and adhesion. In addition, ROS can induce the release of vasoactive (e.g. angiotensin II) and pro-inflammatory mediators, circulating factors known to promote cardiovascular risk-related mechanisms underlying hypertension, cardiac hypertrophy, and atherosclerosis.^22,23,24

Previous controlled exposure studies have found that inhaled pollutants induce pulmonary oxidative stress, ^4–6,25 however, few have examined markers of a systemic response. We previously reported differential regulation of stress response genes in healthy subjects exposed to 200 μg/m³ DE, but found no change in urinary markers of oxidative stress among subjects with metabolic syndrome exposed to the same conditions.²⁶ Multiple population and occupational cohort studies have reported associations linking inhaled PM with markers of protein, lipid and DNA oxidation,^27–30 outcomes that also have been observed in controlled exposures³¹ and, more consistently, in cellular and animal research.^32–37 Animal studies on exposure to fine and ultrafine PM have specifically detected enhanced GSH oxidation, as well as protein and endothelial nitric oxide synthase (eNOS) S-glutathionylation, which inhibits vascular bioavailability NO.^38,39 Controlled DE inhalation studies have also reported subclinical effects attributed to oxidative stress, including endothelial dysfunction, blood pressure elevation, and reduced NO bioavailability,^40–42 outcomes which have been independently correlated with GSH levels in both healthy adults and in smokers.⁴³

The ratio of reduced to oxidized glutathione is an accepted biomarker of systemic antioxidant capacity and redox status in humans.⁴⁴ A shift in the GSH/GSSG balance has been associated with altered cell signaling pathways and increased susceptibility to both acute and chronic disease, including the development and progression of atherosclerosis.^45,46 Our results are in line with multiple findings that women living in areas with higher residential air pollutants had lower blood GSH levels;^47,48 however unlike chronic GSH depletion, which is associated with sustained oxidative stress, transient drops in the thiol are believed to reflect a strong antioxidant defence.^44,49 Similarly, a panel study conducted prior to, during, and after the Beijing Olympics found measures of total antioxidant status decreased with increased levels of air pollution.⁵⁰ The DE-effect we observed indicates that inhaled PM_2.5 directly or indirectly induces an oxidative response, precipitating the fall in GSH/GSSG. Importantly, this finding suggests that DE exposure may influence or exacerbate the effects of other endogenous or exogenous compounds detoxified by GSH.

The GSH/GSSG redox potential (E_h) has also been used to demonstrate oxidative stress associated with disease. Our findings show that GSH/GSSG redox is increased over 5 mV following DE inhalation, confirming changes in levels of oxidation reflected in the ratios above. The effect modeled as redox potential clearly mirrors data shown as GSH/GSSG. Research has reported that GSH/GSSG redox is oxidized by more than 30 mV in association with type 2 diabetes and 9 mV – in line with our estimates – in smokers compared to nonsmokers.³⁵

Though antioxidant treatment (or free-radical scavengers) have been shown to attenuate PM-related insults experimentally,^32,51 findings in human studies have been far from conclusive. In a controlled exposure to concentrated ambient PM, Brook and colleagues found exposure-related increases in blood pressure were unchanged by pretreatment with 2000 mg vitamin C, in line with our results.⁵² Other studies have shown dietary antioxidant intake to be inversely associated with systolic blood pressure⁵³ and with urinary biomarkers of oxidative stress⁵⁴ in persons exposed to higher residential air pollutants, but not at levels sufficient to ameliorate all effects attributable to pollutant exposure.

While higher baseline dietary antioxidant levels are strongly correlated with better health,^55–57 clinical studies have generally been unable to reproduce these effects using antioxidant supplementation.^58–61 Proposed reasons for this disparity have included insufficient dosing and/or method of delivery, non-specific targeting of the antioxidant, and differences in baseline antioxidant capacity. The latter proposes that supplementation is effective only for those with low or reduced baseline capacity, a theory noted specifically for NAC, and one that may explain our findings.^62,63 NAC contributes to intracellular levels of GSH via free reduced cysteine following cleavage of the acetyl group; its antioxidant properties have therefore been attributed to increased GSH capacity. For this reason it has been proposed that, barring GSH depletion, NAC has limited benefit.^62,64 Baseline GSH/GSSG for subjects in our study was well within healthy levels, and likely sufficient to handle a short-term exposure to DE. Other clinical studies similarly have found ascorbic acid and NAC supplementation confer no benefit in the absence of disease, which may underscore the minor effect of treatment in our study population.⁶⁵ Indeed, in our laboratory, antioxidant supplementation did not prevent—and appeared to enhance—the vasoconstrictive effects of DE.⁶⁶ The effectiveness of supplementation in this study may also have been limited by the route of administration; intestinal N-deacetylation and first-pass metabolism diminish the bioavailability of oral NAC.⁶⁷ Pre-exposure GSH/GSSG following antioxidant supplementation did not differ significantly from those in the placebo arm.

It is likely that the effectiveness of supplementation depends both on the baseline capacity of endogenous defenses and the amount or duration of the oxidative insult. Subjects with existing health conditions, reduced antioxidant levels, or otherwise compromised detoxification capacity may have a more pronounced response to both exposure and to supplementation.⁶⁸

We did detect a DE-related increase in leukocyte expression of the proinflammatory cytokine IL-6. IL-6 plays a fundamental role in immune regulation, hematopoiesis, and the acute phase response,⁶⁹ and chronically elevated levels have been identified as a marker of increased cardiovascular risk.⁷⁰ Furthermore, experimental research has demonstrated that PM-induced IL-6 release promotes a prothrombotic state, in conjunction with ROS and adenylyl cyclase.⁷¹ Interestingly, our data suggest antioxidant supplementation may modify the DE-induced increase in expression of IL-6, as stratification by treatment shows mean exposure-related changes were over four fold lower with antioxidant pre-treatment than with placebo. While we detected a similar pattern in plasma IL-6 levels in the placebo group, the effect was not statistically significant. This may have been due to our sample collection time: based on the mRNA data, peak plasma levels likely occurred between our 5 hour and 24 hour collection times. Alternatively, changes in gene expression induced by DE may reflect a response, which, in healthy adults, is not sufficient to induce large alterations in protein production. Positive associations with short-term PM_2.5 levels have been most commonly reported in studies of vulnerable subgroups, such as the elderly and those with comorbid conditions.^72–75 We also did not detect any differences in expression of the other antioxidant response genes tested. This may be attributed to timing: the sampling time we selected did not mirror the transcriptional response time. IL-6 is one of the first cytokines released in circulation.⁷⁶ Ideally, we would have testing a larger number of genes multiple times over a longer period. Or, as postulated above for IL-6 protein, the oxidative effects of exposure may not have been sufficient to induce upregulation of these genes systemically in a healthy population.

We also did not detect differences in the response to DE exposure by GSTM1 genotype. GSTM1 deletion has been associated with increased susceptibility to the effects of long-term air pollution exposure: studies have reported greater risk of lung cancer,⁷⁷ reduced lung function,^78,79 insulin resistance,⁸⁰ inflammation,⁸¹ and alterations in heart rate variability.⁸² However, with the exception of ozone,⁸³ short-term and controlled exposure studies have more commonly reported outcomes to be independent of GSTM1 genotype, in line with our results.^84–86 Interestingly, Bhattacharjee and colleagues recently found GSTM2 overexpression compensated for GSTM1 deletion, noting similar levels of plasma GST enzymatic activity in both positive and null individuals.⁸⁷ We too detected no difference in baseline GSH/GSSG by GSTM1 genotype, with or without antioxidant supplementation. This compensatory mechanism may explain the discrepancy between controlled inhalation and epidemiological studies: a healthy adult may be able to adequately manage an intermittent acute exposure, irrespective of this deletion. Alternatively, it is possible that we would find a different result with a larger sample size, or that the increased risk detected in other studies involved additional gene-gene or gene-environment interactions not present in our study population, or not examined in this analysis.^88–90 While our findings suggest that inhaled pollutants reduce antioxidant capacity, it is likely that exposure provokes other pathophysiologic mechanisms in addition to or associated with oxidative stress.

Though our experimental model is designed to replicate exposure to traffic-related air pollution, the exposures do not reflect the complexity and variability of pollutants in urban air. Our facility does, however, produce consistent, replicable exposures to the most common traffic-related air pollutant, at a standardized composition and concentration. Concentrations used in our controlled exposures are above those commonly experienced in the U.S., however, average PM_2.5 levels in many major cities worldwide regularly exceed such concentrations.⁹¹ It is also important to note that our findings reflect only an acute response, and not the effects of long-term, repetitive exposures. Furthermore, the decrease in GSH/GSSG we report represents the mean response to DE inhalation in young, healthy adults; individual responses within our subject pool varied considerably, and our findings are likely to significantly underrepresent the risk of exposure in susceptible populations.

This controlled inhalation study demonstrates that an acute exposure to DE is associated with a reduction in the GSH/GSSG ratio consistent with shift in balance toward an oxidized state, along with an increase in IL-6 expression in healthy adults. Antioxidant supplementation may modify the DE-induced increases in expression of IL-6. These findings support the theory that exposure to traffic-related air pollution can promote or exacerbate oxidative stress.

Table 2.

Baseline adjusted change in mRNA expression from filtered air.

Gene	Timepoint	Placebo (95% CI)	Antioxidant (95% CI)
GCLc	5 hrs	−.0004 (−.003, .002)	−.001 (−.006, .003)
	24 hrs	.002 (−.002, .005)	−.004 (−.008, .001)
TGFB	5 hrs	−.075 (−.022, 0.007)	.002 (−0.013, .016)
	24 hrs	−.015 (−.034, 0.005)	.001 (−.016, .014)
HMOX	Exposure	−.029 (−.19, .13)	.04 (−.05, .13)
	5 hrs	.004 (−.05, .06)	.08 (.005, .16)
	24 hrs	.02 (−.05, .09)	.05 (−.04, .13)
IL-6	Exposure	−.27 (−.63, .09)	−.05 (−.45, .35)
	5 hrs	.67 (.14, 1.2)*	.16 (−.22, .53)
	24 hrs	−.14 (−.36, .07)	−.24 (−.40, −.09)

Mean difference between change (from pre-exposure baseline) in normalized leukocyte mRNA expression with DE compared with FA control for each time point measured (Exposure=during exposure; 5 hrs = 5 hrs post exposure; 24 hrs = 24 hours from baseline).

^*p=0.02 compared with FA control.

• Traffic-related air pollution is associated with a number of adverse health effects.

• One proposed mechanism is oxidative stress.

• Randomized controlled cross-over human study using inhaled diesel exhaust.

• Diesel exhaust decreased reduced glutathione and increased IL-6 expression.

• Pretreatment with oral antioxidants did not mitigate effects of diesel inhalation.