Genotoxicity of complex PAH mixtures recovered from contaminated lake sediments as assessed by three different methods.

Although human exposure generally occurs to mixtures of chemicals, limited toxicological information is available to characterize the potential interactions of the components of environmental mixtures. This study was conducted to compare the genotoxicity of chemically characterized polycyclic aromatic hydrocarbon (PAH) mixtures using in vitro and in vivo techniques. A total of three extracts (E1-E3) were selected from sediment samples collected from a lake adjacent to an abandoned coal gasification site. Sediments were collected on a grid moving downstream and away from the most likely source of PAH contamination, with E1 collected closest to the shore, E2 at an intermediate distance, and E3 furthest from the shore. The sediment samples were extracted in methylene chloride and methanol, dried, and redissolved in an appropriate solvent for evaluation in a battery of genotoxicity assays. Samples were evaluated for their ability to produce point mutations in bacteria and DNA adducts in vitro without metabolic activation or in vivo. Samples were also analyzed using GC/MS. Sample E1 had both the highest concentration of benzo(a)pyrene (BP) (46.5 ppm) and carcinogenic PAHs and, using 32P-postlabeling, induced the highest adduct levels overall in vitro and in vivo. Sample E2, which had a BP concentration of 14 ppm, induced the greatest number of revertants in the bacterial mutagenicity assay. Sample E3, which had the lowest level of carcinogenic PAHs and BP, induced the lowest adduct levels. However, E3 was capable of inducing a positive genotoxic response in bacteria (with S9), although the slope of the response at lower doses was less than that of E2. The in vivo data showed that the major adduct formed by E1 and E2 was a BP adduct. This information could not have been obtained with the Salmonella or in vitro postlabeling tests. Among internal organs, the extracts of all three samples induced the greatest adduct levels in the lung, similarly to previous complex PAH mixtures studied. These data demonstrate the limitations of predicting genotoxic or carcinogenic potential based on chemical analysis or a single biological test. The results suggest that mixture interactions, cytotoxicity and metabolism are likely to have an influence on the potential of a complex mixture of chemicals to produce a carcinogenic effect. In addition, the concentration of genotoxic PAHs and both in vitro and in vivo DNA adduct formations were decreased with increasing distance from the shoreline. Copyright 1999 Wiley-Liss, Inc.