Effect-directed assessment of the bioaccumulation potential and chemical nature of Ah receptor agonists in crude and refined oils

Abstract

Recent studies have indicated that in addition to narcosis certain chemicals in crude oils and refined petroleum products may induce specific modes of action, such as aryl hydrocarbon receptor (AhR) agonism. The risks these toxic compounds pose to organisms depend on internal exposure levels, as driven by the chemicals' bioaccumulation potential. Information on this potential however is lacking, as the chemicals' identity mostly is unknown. This study showed that AhR agonists bioaccumulate from oil-spiked sediments into aquatic worms and persist in the worms for at least several weeks. Chemical fractionations of eight pure oils into saturates, aromatics, resins, and asphaltenes (SARA), followed by effect-directed analyses using in vitro reporter gene assays revealed that the agonists predominantly are aromatic and resin-like chemicals. Some of the compounds were easily metabolized in vitro, while others were resistant to biotransformation. HPLC-assisted hydrophobicity profiling subsequently indicated that the AhR-active chemicals had a high to extremely high bioaccumulation potential, considering their estimated logK(ow) values of 4 to >10. Most of the AhR agonism, however, was assigned to compounds with logK(ow) of 5-8. These compounds were present mainly in the mid to high boiling point fractions of the oils (C(14)-C(32) alkane range), which are usually not being considered (the most) toxic in current risk assessment. The fractionations further revealed considerable oil and fraction-dependent antagonism in pure oils and SARA fractions. The results of this study clearly demonstrate that crude oils and refined petroleum products contain numerous compounds that can activate the AhR and which because of their likely persistence and extremely high bioaccumulation potential could be potential PBT (persistent, bioaccumulative and toxic) or vPvB (very persistent and very bioaccumulative) substance candidates. Many chemicals were identified by GC-MS, but the responsible individual compounds could not be exactly identified in the complex mixtures of thousands of compounds. Because this obstructs a classical PBT risk assessment, our results advocate an adapted risk assessment approach for complex mixtures in which low concentrations of very potent compounds are responsible for mixture effects.