Assessment of mutagenic efficiency of two carcinogen-modified nucleosides, 1,N6-ethenodeoxyadenosine and O4-methyldeoxythymidine, using polymerases of varying fidelity.

Terminal deoxynucleotidyl transferase (TdT) was used to prepare copolymers of dA and 1,N6-ethenodeoxyadenosine (epsilon dA). When used as templates for Escherichia coli DNA polymerase I (Pol I) and compared with poly (dA), normal dTTP incorporation was not significantly affected by the presence of 7% epsilon dA. dGTP misincorporation was only slightly increased and occurred about once for every 500 epsilon dA residues. The error-prone polymerase from avian myeloblastosis virus (AMV reverse transcriptase) increased this error rate 5- to 20-fold to a maximum of 1 dG/25 epsilon dA. No dCTP misincorporation was detected with either polymerase. In transcription with E. coli DNA-dependent RNA polymerase, no errors were revealed by nearest neighbor analysis. Poly (dA) treated with chloroacetaldehyde under conditions producing the same proportion of epsilon dA (without the hydrated form) as the synthesized template behaved in the same manner with a similar low level of misincorporation of dG. Such treatment of alternating poly d(A-T) caused structural changes indicative of crosslinks but did not alter its template properties. Increasing the amount of epsilon dA in either synthesized or modified polymers greatly decreased the template activity without increasing the error rate. It is suggested that epsilon dA generally does not prevent dT incorporation but behaves as a bulky lesion which is bypassed. In contrast to the low mutagenic efficiency of epsilon dA, O4-methyldeoxythymidine (m4dT), in copolymers with dA, directed the misincorporation of 1 dG/12 m4dT with Pol I and 1 dG/3 m4dT with reverse transcriptase. Nearest neighbor analysis of transcripts showed the incorporation of 1 dG/12 m4dT. These data are in agreement with the previous reported mutagenicity of m4dT in alternating poly d(A-T, m4T).