The Ames Salmonella/microsomal activation mutagenesis assay has been modified to improve sensitivity and reproducibility to complex mixtures derived from the refining and processing of petroleum. Oil samples were dissolved in cyclohexane and subsequently extracted with dimethyl sulfoxide to produce aqueous compatible solutions which readily interact with tester bacteria. Also, the liver homogenate (S-9) and NADP cofactor concentrations were increased and hamster rather than rat liver S-9 was used. The initial slope of the dose response curve relating mutagenicity (revertants per plate) to the dose of extract added was used as an index of mutagenic activity; this slope was obtained through a computerized curve fitting procedure. The modified assay was used to rank 18 oil samples for mutagenic activity; this ranking correlates highly (r = 0.92) with potency rankings of the same samples previously determined from dermal carcinogenicity bioassays. Sensitivity and reproducibility of the assay are sufficient to permit routine use for detecting potential carcinogenic activity of individual refinery streams and blends which contain components boiling above 500 degrees F.

The Ames Salmonella/microsomal activation mutagenesis assay has been adapted to improve sensitivity to complex hydrocarbon mixtures produced by the refining of petroleum. Extraction of oil samples with dimethyl sulfoxide produces aqueous-compatible solutions that more easily interact with the tester bacteria. These extracts, therefore, produce higher revertant values than do equivalent volumes of oil delivered neat or dissolved in organic solvent. Parallel increases in the liver microsomal S-9 concentration further improve the sensitivity of the assay, allowing detection of mutagenicity in otherwise inactive samples. The effect of increased microsomal fraction from rodent liver is apparently attributable to the higher levels of activating enzymes rather than to the concomitant increase in the overall hydrophobicity of the test system. The modified assay has been used to rank thirteen petroleum-derived oils and a corn oil control for relative mutagenic activity. This ranking closely correlates (r = 0.97) with potency rankings of the same samples previously determined from dermal carcinogenicity bioassays.