Metals and oxidative potential in urban particulate matter influence systemic inflammatory and neural biomarkers: A controlled exposure study

Abstract

Background: Oxidative stress and inflammation are considered to be important pathways leading to particulate matter (PM)-associated disease. In this exploratory study, we examined the effects of metals and oxidative potential (OP) in urban PM on biomarkers of systemic inflammation, oxidative stress and neural function.

Methods: Fifty-three healthy non-smoking volunteers (mean age 28 years, twenty-eight females) were exposed to coarse (2.5-10 μm, mean 213 μg/m³), fine (0.15-2.5 μm, 238 μg/m³), and/or ultrafine concentrated ambient PM (<0.3 μm, 136 μg/m³). Exposures lasted 130 min, separated by ≥2 weeks. Metal concentrations and OP (measured by ascorbate and glutathione depletion in synthetic airway fluid) in PM were analyzed. Blood and urine samples were collected pre-exposure, and 1-h and 21-h post exposure for assessment of biomarkers. We used mixed-regression models to analyze associations adjusting for PM size and mass concentration.

Results: Results for metals were expressed as change (%) from daily pre-exposure biomarker levels after exposure to a metal at a level equivalent to the mean concentration. Exposure to various metals (silver, aluminum, barium, copper, iron, potassium, lithium, nickel, tin, and/or vanadium) was significantly associated with increased levels of various blood or urinary biomarkers. For example, the blood inflammatory marker vascular endothelia growth factor (VEGF) increased 5.3% (95% confidence interval: 0.3%, 10.2%) 1-h post exposure to nickel; the traumatic brain injury marker ubiquitin C-terminal hydrolase L1 (UCHL1) increased 11% (1.2%, 21%) and 14% (0.3%, 29%) 1-h and 21-h post exposure to barium, respectively; and the systemic stress marker cortisol increased 1.5% (0%, 2.9%) and 1.5% (0.5%, 2.8%) 1-h and 21-h post exposure to silver, respectively. Urinary DNA oxidation marker 8‑hydroxy‑deoxy‑guanosine increased 14% (6.4%, 21%) 1-h post exposure to copper; urinary neural marker vanillylmandelic acid increased 29% (3%, 54%) 1-h post exposure to aluminum; and urinary cortisol increased 88% (0.9%, 176%) 1-h post exposure to vanadium. Results for OP were expressed as change (%) from daily pre-exposure biomarker levels after exposure to ascorbate-related OP at a level equivalent to the mean concentration, or for exposure to glutathione-related OP at a level above the limit of detection. Exposure to ascorbate- or glutathione-related OP was significantly associated with increased inflammatory and neural biomarkers including interleukin-6, VEGF, UCHL1, and S100 calcium-binding protein B in blood, and malondialdehyde and 8-hydroxy-deoxy-guanosine in urine. For example, UCHL1 increased 9.4% (1.8%, 17%) in blood 21-h post exposure to ascorbate-related OP, while urinary malondialdehyde increased 19% (3.6%, 35%) and 8-hydroxy-deoxy-guanosine increased 24% (2.9%, 48%) 21-h post exposure to ascorbate- and glutathione-related OP, respectively.

Conclusion: Our results from this exploratory study suggest that metal constituents and OP in ambient PM may influence biomarker levels associated with systemic inflammation, oxidative stress, perturbations of neural function, and systemic physiological stress.

Figure 1.

Percentage change (95% confidence interval) in blood biomarkers 1-hr and 21-hr post exposure to AA-related OP at the mean exposure level (1A), and to GSH-related OP at levels above the limit of detection (1B). Percent change was calculated using the equation: [(post-exposure value minus pre-exposure value)/pre-exposure value]*100. Models were adjusted for age, sex, BMI, season, type of CAPs, mass concentration. *, p<0.05.

Figure 2.

Percentage change (95% confidence interval) in urinary biomarkers 1-hr and 21-hr post exposure to AA-related OP at the mean exposure level (2A), and to GSH-related OP at levels above the limit of detection (2B). Percent change was calculated using the equation: [(post-exposure value minus pre-exposure value)/pre-exposure value]*100. Models were adjusted for age, sex, BMI, season, type of CAPs, mass concentration. *, p<0.05.