Formation of DNA adducts from oil-derived products analyzed by 32P-HPLC

The aim of this study was to investigate the genotoxic potential of DNA adducts and to compare DNA adduct levels and patterns in petroleum vacuum distillates, coal tar distillate, bitumen fume condensates, and related substances that have a wide range of boiling temperatures. An in vitro assay was used for DNA adduct analysis with human and rat S-9 liver extract metabolic activation followed by P-32-postlabeling and P-32-high-performance liquid chromatography (P-32-HPLC). For petroleum distillates originating from one crude oil there was a correlation between in vitro DNA adduct formation and mutagenic index, which showed an increase with a distillation temperature of 250 degreesC and a peak around a distillation point of approximately 400 degreesC. At higher temperatures, the genotoxicity (DNA adducts and mutagenicity) rapidly declined to very low levels. Different petroleum products showed a more than 100-fold range in DNA adduct formation, with severely hydrotreated base oil and bitumen fume condensates being lowest. Coal tar distillates showed ten times higher levels of DNA adduct formation than the most potent petroleum distillate. A clustered DNA adduct pattern was seen over a wide distillation range after metabolic activation with liver extracts of rat or human origin. These clusters were eluted in a region where alkylated aromatic hydrocarbons could be expected. The DNA adduct patterns were similar for base oil and bitumen fume condensates, whereas coal tar distillates had a wider retention time range of the DNA adducts formed. Reference substances were tested in the same in vitro assay. Two- and three-ringed nonalkylated aromatics were rather low in genotoxicity, but some of the three- to four-ringed alkylated aromatics were very potent inducers of DNA adducts. Compounds with an amino functional group showed a 270-fold higher level of DNA adduct formation than the same structures with a nitro functional group. The most potent DNA adduct inducers of the 16 substances tested were, in increasing order, 9,10-dimethylanthracene, 7,12-dimethylbenz[a]anthracene and 9-vinylanthracene. Metabolic activation with human and rat liver extracts gave rise to the same DNA adduct clusters. When bioactivation with material from different human individuals was used, there was a significant correlation between the CYP 1A1 activity and the capacity to form DNA adducts. This pattern was also confirmed using the CYP 1A1 inhibitor ellipticine. The P-32-HPLC method was shown to be sensitive and reproducible, and it had the capacity to separate DNA adduct-forming substances when applied to a great variety of petroleum products.