ASSESSMENT OF POLYCYCLIC AROMATIC HYDROCARBON BIOAVAILABILITY FROM SOIL USING THE JUVENILE SWINE MODEL

Polycyclic aromatic hydrocarbons (PAHs) are common soil contaminants due to their lipophilic nature which limits partitioning to water or air. Soil properties such as organic carbon can affect PAH release from soil, and thus affect PAH bioavailability of ingested soil. Risk assessment of PAHs in soil generally assumes equal bioavailability of PAHs ingested in soil compared to PAHs ingested in reference dose media, leading to environmental cleanup guidelines that are potentially too conservative. This research intended to use the juvenile swine model to assess PAH bioavailability from impacted soil to better inform bioavailability estimates for risk assessment. This was done by assessing PAH bioavailability from single and repeated exposure to PAHs in different spiked exposure media, assessing PAH bioavailability from soil collected from PAH impacted sites, and assessing biomarkers of exposure and effect following PAH exposure. The effect of exposure duration on bioavailability was assessed because people are usually chronically exposed to PAHs, rather than acutely exposed, as most bioavailability studies are performed, and chronic exposure may lead to increases in xenobiotic metabolizing enzymes and transporters which may affect bioavailability. This research found that exposure duration did not significantly affect anthracene and benzo[a]pyrene bioavailability (p>0.075), but exposure media did (p<0.004). These results suggest that exposure medium has a more important effect on bioavailability than exposure duration, and also bioavailability calculated from a single exposure is appropriate for use in risk assessment. Bioavailability from 24 naturally impacted soils was assessed to determine which soil characteristics had the greatest effect on PAH bioavailability. Area under the curve (AUC) measurements for benzo[a]pyrene (BaP) and anthracene in swine blood after oral exposure from a soil matrix for benzo[a]pyrene and anthracene in soils had s very poor relationship with soil concentrations in soils collected from impacted sites (r2<0.15), but a very strong relationship with soil concentrations from spiked artificial soils (r2<0.95). As spiked soils had much higher concentrations of PAH, these results suggest there is a point of departure in soil concentrations where internal exposure becomes linearly related to soil concentration. Point of departure modeling indicates that this point occurs at soil PAH concentrations greater than 1,900 mg kg-1. Thus, risk assessment can assume a constant exposure to PAHs at soil concentrations lower than the point of departure. Comparison of terminal rate constants from intravenous (IV) exposure to PAHs and oral exposure to PAHs in a soilmatrix suggest that flip-flop kinetics occur in swine, where absorption occurs at a slower rate than elimination. Flip-flop kinetics likely explains the lack of relationship between real world soil concentrations and area under the curve measurements as absorption is the rate limiting step of elimination. Biomarkers of exposure and effect were assessed in swine liver and ileum tissue, as well as blood following single and subchronic exposure to PAHs to determine if relationships could be drawn between exposure magnitude and duration and biomarker formation. Biomarkers included cytochrome P450 (P450) 1A1, 1A2, and 1B1 expression and activity as biomarkers of exposure and DNA adducts, carbonylated proteins, and micronucleated reticulocytes as biomarkers of effect. Biomarkers of exposure were not affected by exposure magnitude or duration, indicating that they would serve best as exposure markers rather than indicators of bioavailability or other effects. However, DNA adduct and protein carbonyl formation was significantly affected by exposure duration (p