32P-adduct assay: comparative recoveries of structurally diverse DNA adducts in the various enhancement procedures.

A 32P-adduct assay for the measurement of low levels (1 adduct per 10(7) nucleotides) of binding of carcinogens to DNA has been reported previously. In this procedure, DNA is enzymatically hydrolyzed to 3'-monophosphates of normal nucleosides and adducts, which are 5'-32P-labeled by T4 polynucleotide kinase and [gamma-32P]ATP. Labeled adducts are resolved by TLC. Enrichment of adducts by extraction in 1-butanol [Gupta, R.C. (1985) Cancer Res., 45, 5656] or digestion with nuclease P1 [Reddy, M.V. and Randerath, K. (1986) Carcinogenesis, 7, 1543] prior to 32P-labeling, however, increased the sensitivity of detection for many adducts to a level of 1 per 10(9-10) nucleotides, although adduct recovery particularly in the latter assay depended on the chemical nature of adducts. We have now compared recoveries for greater than 70, different carcinogen-DNA adducts of known and unknown chemical nature in the two enrichment procedures as well as in a new procedure in which polynucleotide kinase is substituted for nuclease P1. When compared with the butanol extraction procedure, arylamines (such as 2-aminofluorene, 2-aminophenanthrene, 2-naphthylamine, 4-aminobiphenyl, 4-azoaminobenzene and N'-acetylbenzidine) bound to the C8 position of guanine were lost almost completely (0.2-4% recovery) in the nuclease P1-mediated assay, but the presence of a polar group in the aromatic amine moiety (such as 2-acetylaminofluorene, 2-acetylamino-phenanthrene and methyl-4-azoaminophenyl) rendered similar recovery. In contrast, aromatic amines (2-amino-phenanthrene, 2-acetylaminophenanthrene, 2-acetylaminofluorene and methyl-4-azoaminobenzene) and polycyclic aromatic hydrocarbons (benzo[a]pyrene, bromomethylbenzanthracene and benzanthracene) bound to the exocyclic positions of guanine or adenine showed extensive or as complete recovery in the nuclease P1 procedure as in the extraction procedure. Some of the unknown presumably polar adducts showed a lower recovery (30-70%) in the butanol procedure as compared to the nuclease P1 enrichment. The recovery pattern of most adducts examined in the polynucleotide kinase-enrichment assay was essentially the same as found in nuclease P1-mediated assay, except that overall lower values were obtained. Our data suggest that a given DNA sample should be analyzed by different versions of the 32P-adduct assay, particularly, DNA of specimens of humans exposed to low levels of unknown carcinogens. The observation that chemical structure of an adduct may be detrimental in its recovery in the enzyme- and extraction-mediated enrichment procedures may serve as a probe in the structural characterization of adducts of unknown carcinogens.