Physico-chemical determinants of in vitro particulate oxidative burden
KRYSTAL J. Godri (1,2), Ian S. Mudway (1), Frank J. Kelly (1), Roy M. Harrison (2), Maciek M. Strak (3), Maaike Steenhof (4), Paul H.B. Fokkens (3), A. John F. Boere (3), Daan L.A.C. Leseman (3), Kaas Meliefste (4), G. Hoek (4), Bert Brunekreef (4), Erik Lebret (3), Ilse Gosens (3), Flemming R. Cassee (3), Nicole A.H. Janssen (3)
(1) MRC-HPA Center for Environment and Health, King’s College London, London, United Kingdom, (2) Division of Environmental Health & Risk Management, University of Birmingham, Edgbaston, United Kingdom, (3) RIVM (National Institute for Public Health and the Environment), Bilthoven, the Netherlands, (4) IRAS (Institute for Risk Assessment Sciences), Utrecht University, Utrecht, the Netherlands
Abstract Number: 315
Preference: Platform Presentation
Last modified: November 9, 2009
Working Group: sq1
Background: The capacity of particulate matter (PM) to elicit inflammation in vivo has been proposed to be a function of its oxidative potential (OP): the ability of PM to directly or indirectly elicit reactive oxygen species generation in the lung. In this paradigm the imposition of oxidative stress in the lung following PM inhalation activates redox sensitive transcription factors, driving the expression of pro-inflammatory mediators. PM OP therefore reflects its content of catalytic transition metals and quinones, as well as other organics (PAHs) and biological (endotoxin) constituents that can induce oxidative stress indirectly via xenobiotic metabolism or the induction of inflammation respectively. The derivation of an expression that summarises the particulate OP is therefore appealing, as a way of integrating various PM compositional and physical characteristics into a single expression with pertinence to the observed biological response.
Objective: We examined the OP of size fractionated PM collected at microenvironments with contrasting local sources and identified the physical and chemical determinants of this metric.
Methods: PM$_2.5 and PM$_2.5-10 samples were collected at seven microenvironments with contrasting sources (n=45 per size fraction). Locations included traffic (urban intersection, carriageway, diesel), background (urban background, farm), and industrial (steel mill, harbour) sites. Six-hour sampling campaigns (9:00-15:00) were conducted with 4-10 visits per site. A high volume cascade impactor and micro-orifice cascade impactor were deployed to obtain samples for chemical composition and OP characterisation, respectively. The latter was quantified based on PM-induced antioxidant (ascorbate (AA) and glutathione (GSH)) oxidation from synthetic respiratory tract lining fluid (RTLF). Water soluble inorganic ions, total and soluble trace metals, elemental and organic carbon, polycyclic aromatic hydrocarbons and quinones were measured in all PM samples. Particle number concentration and surface area were also measured with 1 minute resolution. The determinants of PM OP, based on both AA and GSH oxidation were identified using stepwise multiple linear regression modelling with a backwards deletion approach.
Results: The total particulate OP was stratified across microenvironments with the greatest activities associated with the traffic sites (mean 55.6+/-2.2m$^-3) followed by the urban background site (39.3m$^-3), industrial sites (mean 29.0+/-5.4m$^-3) and the lowest at farm location (23.8m$^-3). Variation in PM OP based on AA oxidation from the RTLF was highly correlated with total trace metal concentrations for both size fractions (>95% total variance explained). Conversely, trace metal concentrations alone only accounted for 56% (PM$_2.5) and 68% (PM$_2.5-10) of the measured variation in OP based on GSH oxidation. Overall, OP based on GSH oxidation appeared to be driven by traffic emission components, specifically those related to non-tailpipe sources (Cu). In contrast, the OP based on AA oxidation appeared sensitive to emissions from both traffic (Cu, Fe) and fuel oil combustion sources (Ni, V).
Conclusions: Redox active and non-active metals were found to have the strongest correlations with the associated PM oxidative burden. OP based on AA and GSH oxidation was found to be sensitive to different panels of chemical species suggesting that a measure of total OP maybe best reflected by the sum of these two individual expressions.