DNA adduct formation by mineral oils and their fractions as indicated by 32P-postlabelling: is adduct formation truly indicative of carcinogenic potential?

Oils of differing types, physical properties and carcinogenic activity were tested for ability to produce epidermal DNA adducts 24 h after application to the skin of mice, using the 32P-postlabelling method. Two studies were carried out, the first on three oils to identify the adduct-forming components, and the second on nine oils to investigate whether the nature of the oils affected their adduct-forming potential. In addition to the whole oils, fractions of the oils containing saturated hydrocarbons, 2-6-ring aromatic compounds and polar compounds were tested in both studies in proportion to their concentration in the original oil. In addition, a further examination of the aromatic fractions from two carcinogenic oils was carried out in the first study by testing 2-3-ring and 4-6-ring subfractions (the latter containing carcinogenic polycyclic aromatic compounds (PACs)). Results from the first study indicated that carcinogenic oils do produce adducts and that the adduct-forming components were mainly in the aromatic fraction. When, however, subfractions of the aromatic fraction were examined, it was found that slightly higher adduct levels were produced by the 2-3-ring aromatic fraction than with the 4-6-ring PAC fraction. This was contrary to expectation from published work on the skin carcinogenicity of oils and suggested that some non-carcinogenic PACs may produce adducts. The second study indicated that although most carcinogenic oils produced adducts, some non-carcinogenic oils can also do so.(ABSTRACT TRUNCATED AT 250 WORDS)