Abstract
The current risk assessment of Polycyclic Aromatic Hydrocarbon mixtures is inadequate. The genotoxicity of a PAH-mixture cannot be assumed to be a sum of individual compounds, as interactions are likely to occur. Therefore, a battery of tests was developed employing an in-vitro and in-silico physiologicallybased pharmacokinetic (PBPK) modelling to monitor such interactions at the metabolism level.
Precision-cut rat liver slices incubated with binary mixtures of soil PAHs were assessed for their ability to modulate the activity of bioactivating/detoxifying
enzymes. In addition, non-carcinogenic PAHs were investigated, as these binary mixtures have been reported to potentiate tumour formation in mice. Binary mixtures of PAHs in combination with benzo(a)pyrene showed to synergistically modulate CYP1A1 activity, even those that do not individually induce this enzyme; while some inducers such as benz(a)anthracene and dibenzo(a,l)pyrene displayed antagonism. None of the individual PAHs investigated enhanced CYP3A activity. A similar observation was noted for microsomal-epoxide hydrolase. UDP-glucuronosyltransferase was induced by almost all individual PAHs, while binary mixtures had a varied impact on its modulation. Co-incubation of PAHs with B(a)P had no impact on glutathione-S-transferase, while quinolone-oxidoreductase-1 was significantly antagonised by several PAHs.
Interaction with the Ah-receptor was investigated as a potential underlying mechanism of the observed PAH-interactions employing H1L75c3 mouse hepatoma cells. Most mixtures appeared to regulate the receptor antagonistically. While some degree of correlation was noted between the receptor affinity, CYP1A1, and quinolone-oxidoreductase-1 activities for individual PAHs, no such relationship was observed for mixtures.
The human PBPK model was also used to investigate PAH interactions at the CYP1A1 activity-level using GastroPlus. Concurrent oral exposure to benzo(a)pyrene and benz(a)anthracene displayed antagonism, thus confirming the experimental observations. The identified set of biomarkers, namely CYP1A1, quinolone-oxidoreductase-1, and UDP-glucuronosyltransferase, along with Ah-receptor can be used in PAH-mixture risk assessment; and should further be incorporated into the interaction-PBPK model to build a complete bioactivation/detoxification prediction model.