AP SITES ARE NOT SIGNIFICANTLY INVOLVED IN MUTAGENESIS BY THE (-ANTI DIOL EPOXIDE OF BENZO[A]PYRENE - THE COMPLEXITY OF ITS MUTAGENIC SPECIFICITY IS LIKELY TO ARISE FROM ADDUCT CONFORMATIONAL POLYMORPHISM())

In previous work, mutations induced by the (+)-anti diol epoxide of be nzo[a]pyrene [(+)-anti-B[a]PDE] were scored in the supF gene of the Es cherichia coli plasmid pUB3 [Rodriguez & Loechler (1993) Biochemistry 32, 1759]. pUB3 was reacted with (+)-anti-B[a]PDE and then either (1) transformed immediately into E. coli or (2) heated at 80-degrees-C for 10 min prior to transformation. Heating only released a small fractio n of adducts (approximately 5%) and did not significantly affect the m utagenic pattern at most sites in supF. However, at the major base sub stitution hotspot, G115, principally G-->T mutations (87%) were obtain ed prior to heating, while after heating, G-->T mutations decreased (4 5%) and G-->A (21%) and G-->C (33%) mutations became more prevalent. O ne model for this result is that prior to heating a heat-labile adduct at G115 causes one pattern of mutagenesis, but after heating the labi le adduct is hydrolyzed to an apurinic site (AP site), which causes a second mutational pattern. To test this, a role for AP sites generated from labile adducts by heating at 80-degrees-C for 10 min is investig ated. It is shown that when plasmid pUB3 contains 22 (+)-anti-B[a]PDE adducts, 0.6% (or fewer) are converted to AP sites as determined in an assay based upon the action of an AP-endonuclease. In a separate line of investigation not involving (+)-anti-B[a]PDE adducts, mutation fre quency (MF) per AP site is estimated. (In these experiments, AP sites were introduced into pUB3 by the classic procedure of heating at 70-de grees-C/pH 5.0 to hydrolyze purines. In fact, the majority of mutants induced by this procedure probably arose via cytosine deamination to u racil and not via AP sites, based upon three criteria, including that approximately 75% of the mutations were GC-->AT.) Given the number of AP sites formed from (+)-anti-B[a]PDE adducts and the estimate of MF/A P site, we conclude that < approximately 2% of the 115 base substituti on mutations induced in supF by (+)-anti-B[a]PDE can be attributed to AP sites when the adducted plasmid was heated at 80-degrees-C for 10 m in prior to transformation. (With the unheated adducted plasmid, this limit is < approximately 1%). This makes it unlikely that AP sites are significantly involved in (+)-anti-B[a]PDE mutagenesis, including at G115, where 29% of base substitution mutants were found. The most like ly alternative model for the heat-induced changes in mutational patter n at G115, as well as the complexity of the mutagenic spectra of (+)-a nti-B[a]PDE in general, is that a single adduct can adopt multiple con formations, each of which can cause different kinds of mutations, and factors such as heating and DNA sequence context can influence adduct conformation.