Towards the integration of bioanalytical tools into air pollution assessment, regulation and management

Risk assessment for air pollutants remains a complex task. In this project, we aimed to assess air pollutant risk through the combination of active air sampling, bioanalytical assessment and selective chemical analysis. To achieve this air samples were collected over several weeks from both indoor and outdoor sites and subjected to a range of bioanalytical assays. Both XAD (gas phase) and filter paper (particles) samples were tested for responses in genotoxicity, Ah-receptor response, estrogenicity and respiratory toxicity assays. The method of sampling, sample extraction and processing principally targeted semivolatile organic chemicals. Phase 1 of the study concerned the set-up of assays, evaluation of blanks and the validation of responses to known chemicals of concerns. Five endpoints were included in the evaluation including non-specific cytotoxicity, genotoxicity, Ah-receptor response, estrogenicity and respiratory immunotoxicity. Dose response assessments for test chemicals were carried out with all assays and standard operating procedures for the testing have been developed. Blank toxicity evaluation found that sampling matrices and the preparation of these requires some attention and it is important to use ultrapure solvent for the preparation of the sampling matrices. Phases 2 and 3 of the project focussed on the testing of actual samples collected at different sites in Queensland and Western Australia. During phase 2, the first evaluation of air samples showed that the assays could detect differences between different sites and in both the gas and particle phase. For example using extracts of filter papers collected from the indoor domestic site a genotoxic and estrogenic response was detectable whereas the extracts of the XAD fraction from the same site registered an effect in the respiratory assay. In the sample collected outdoors at the same location genotoxic and estrogenic responses in the XAD fraction were quantifiable. Phase 3 of the project tested samples collected at a wider range of sites in winter 2007. These sites included an indoor industrial location, two offices, two sites near major roadways, a residential area, light industry and a background site. Selected samples were also subjected to an aqueous extraction for examination of any ionic/water-soluble chemicals present. Results showed that overall, air samples produced a more genotoxic response in terms of direct acting chemicals (-S9) at both indoor and outdoor sites. In most cases the gas phase fraction of these samples was more toxic than the particle phase. The gas and particles phases of the samples collected at the indoor industrial site and a high traffic site were the most toxic. Interestingly the gas phase from the two office sites and the samples collected at the polluted sites produced similar genotoxic responses. Air samples collected at most sites also produced genotoxic responses for chemicals that required metabolic activation to damage DNA (+S9). However, in general, toxicity due to these chemicals was lower than that associated with direct acting chemicals. The genotoxic effect of chemicals requiring metabolic activation was detected in both the gas and particle phase of samples. The gas phase samples collected at the offices and at a high traffic site recorded the highest toxicity. However genotoxic effects were also detected in filter papers collected from three outdoor sites. These sites were all located near roadways. These results suggest that even though industrial and vehicular related National Research Centre for Environmental Toxicology Page 3 of 49 activities are associated with the release of genotoxic pollutants, there were also unidentified sources of such chemicals at the office sites. Air samples collected from all sites elicited an Ah receptor response in the CAFLUX bioassay. Results (expressed as TCDD equivalent air concentrations) showed that in contrast to genotoxicity assessed on the umu assay, the Ah-receptor activity was primarily associated with the particle phase of samples. Surprisingly, the sample collected at the high density residential site exceeded equivalent concentrations measured in samples collected at both high traffic sites. Further testing of two samples indicated that the response on CAFLUX was possibly due to chemicals other than dioxins and dioxin-like PCBs. All air samples showed a response in the E-screen bioassays for estrogenicity. Results were expressed as estradiol equivalency adjusted for air volume sampled. In all cases the gas phase of air samples was more estrogenic than the particle phase, usually by a factor 2 or more. The gas phase samples collected from indoors exhibited the highest estrogenicity of all samples. Both water and solvent extracts were tested on the respiratory toxicity assay. Water extracts of filter papers from both high traffic sites were significantly different from one of two field blanks. Solvent extracted samples were not significantly different from the corresponding field blanks. Further work is required to determine why certain samples, including the second field blank, tested as cytotoxic. This study demonstrates that we are exposed to a wide variety of chemical pollutants that have a range of potential health effects. This work has also identified that the gas phase is potentially an important pathway for exposure, particularly for genotoxic effects. Chemicals inhaled via the gas phase are more efficiently transported into the lung and adsorbed by tissue. Traditionally, investigations into the health risk associated with semivolatile organic pollutants in air have not separated the gas and particle phases. A comparison between the umu assay based benzo[a]pyrene (B[a]P) equivalencies and equivalences determined using chemical analysis showed that in the particle phase only 10% of chemicals and in the gas phase, less than 1% of chemicals responsible for the observed effect have been identified. Further work is required to identify the other chemicals that are contributing to the genotoxic response in samples. In contrast, CAFLUX results showed a significant correlation with the PAH analysis, suggesting the dominant component of the samples causing this response were PAHs. This area of research is rapidly producing useful techniques to expand the toolbox available for characterising air quality. Importantly this report shows that the results from such bioanalytical tests could be communicated in terms that were meaningful for air quality managers by using specific chemical equivalencies for comparison to guideline values. This approach has the added advantage of accounting for significantly more of the chemical constituents in a sample. As the comparison to PAH results showed, the traditional approach to air quality management which targets specific chemicals could underestimate the potential risk of exposure at certain sites including indoor locations where humans spend a significant amount of time