32P-postlabeling of bile components: bulky adduct-like behavior in polyethyleneimine-cellulose thin layer chromatography.

The 32P-postlabeling assay has been used widely in carcinogen-DNA adduct analysis because of its sensitivity and reproducibility. Cloned T4 polynucleotide kinase (PNK), routinely used in this assay, phosphorylates the 5'-OH groups of adducted nucleotides in the presence of [gamma-32P]ATP. However, as an exception to this property, PNK has been reported to phosphorylate non-adducted carcinogen metabolites, such as tetrol derivatives of benzo[a]pyrene and chrysene. Also, PNK phosphorylates both 5'-OH and 3'-OH groups of safrole-adducted deoxydinucleoside monophosphates having an unmodified purine in the 3'-position. In the present study we show that T4 PNK catalyzed the transfer of [32P]phosphate from [gamma-32P]ATP to rat bile components or purified bile acids (derivatives of 3 alpha-hydroxy-5 beta-cholanic acid) in the absence of nucleic acids or nucleases. However, labeling of the bile acids appeared over 100,000-fold less efficient than labeling of 2'-deoxyadenosine-5'-monophosphate. There was no reaction in the absence of bile components or PNK. Dehydrocholic acid, which lacks hydroxyl groups, was resistant to phosphorylation. On polyethyleneimine-cellulose TLC maps, 32P-labeled rat bile extract gave an array of non-polar radioactive spots which resembled carcinogen-DNA adducts, while 32P-labeled purified bile acids each gave a single spot. These 32P-labeled products liberated 32Pi upon incubation with prostatic acid phosphatases. Two of the radioactive spots obtained from rat bile were identified as phosphorylated taurocholic and taurodeoxycholic acids by co-chromatography with 32P-labeled standards. These findings demonstrate for the first time that PNK is able to phosphorylate natural products other than nucleotides and further emphasize the need to rule out contamination with bile acids and possibly other bulky/hydrophobic alcohols when analyzing DNA samples by 32P-postlabeling.