The influence of automobile exhausts on mutagenicity of soils: contamination with, fractionation, separation, and preliminary identification of mutagens in the Salmonella/reversion assay and effects of solvent fractions on the sister-chromatid exchanges in human lymphocyte cultures and in the in vivo mouse bone marrow micronucleus assay.

To test the assumption that automobile exhausts contribute to soil mutagenicity, two soils with low levels of mutagenic activities were exposed to traffic exhausts at a heavily charged junction of German motorways (Autobahnen) for 3, 7, 10, 13, 17, 21, and 26 weeks. Indeed, in the presence of a metabolic activation system from rat liver (S9), an average increase of 8 and 9 (4 and 12) revertants per gram per week was found in Salmonella typhimurium TA 98 (TA 100). In the absence of S9, meaningful measurements were impossible on account of a concurrent dose dependent increase of toxicity. No correlation between the increase of mutagenicity and the contents of polycyclic aromatic hydrocarbons (PAH) could be detected. In another series, soils sampled at the roadside and at distances of 10 and 50m of five roads near Mainz expressed 10-20-fold higher mutagenicity (revertants per gram) under identical test conditions as compared with the average of agricultural soils. Toxic effects, however, again confounded the results and no correlation between the distance from roads and the levels of mutagenicity could be demonstrated. Subsequently, Soxhlet-extraction with the solvent sequence dichloromethane, acetone, and toluene/diethylketone was found to be an optimum procedure for soils at roadsides. The mass balance of solvent fractionation of such soils revealed that <2% each belonged to organic acids and bases, approximately 4% to fractions designed polar neutrals, approximately 8% to polar aromatics, approximately 7% to dichloromethane solubles, and approximately 79% to cylohexane solubles, among them approximately 63% acetone soluble compounds. The major part of mutagenicity (55-65%) was present in the fraction of polar aromatics, followed by polar neutrals and the acetone subfraction of cyclohexane solubles ( approximately 10% each) summarizing the results obtained with S. typhimurium TA 98, TA 98NR, YG 1021, YG 1024, TA 100, YG 1026, and YG 1029 with and without addition of S9. The modified tester strains, either deficient in nitroreductase (TA 98NR) or overproducing nitroreductase (YG 1021, 1026) or O-acetyl-transferase (YG 1024, 1026), indicated a major contribution of nitroarenes to soil mutagenicity. With respect to mutagenic PAH, high pressure liquid chromatography (HPLC) revealed that >90% of dibenz[a,h]anthracene (4.18mg/kg soil), benzo[a]pyrene (1.96mg), benzofluoranthenes (0.14mg), and benz[a]anthracene (0. 18mg) were present in the acetone subfraction of cyclohexane solubles. Concentrations and mutagenic activities, however, did not correlate. Additional preparative and analytical HPLC of the solvent fractions of polar neutrals and polar aromatics, resulted in the tentative identification of 2-nitrofluorene. Analysis of the vertical profile of soil revealed an increase of mutagenicity per gram from the surface to a maximum at 5-15cm depth and a subsequent decrease with very little activity remaining deeper than 35cm. In human lymphocyte cultures, the fraction of polar aromatics, 0.01-0. 3microg/ml, induced 11.27+/-4.76-20.70+/-6.19 sister-chromatid exchanges (SCE) per cell in the absence of S9 (solvent control: 10. 16+/-4.83 SCE per cell) and 12.77+/-6.53-17.87+/-4.93 SCE per cell in the presence of S9 (solvent control: 8.37+/-3.92 SCE per cell). However, no activities could be detected in the fractions of polar neutrals and non-polar neutrals. Again, negative results were obtained in the in vivo mouse bone marrow micronucleus assay at 2000mg/kg p.o. with all fractions.