O4-Methyl, -ethyl, or -isopropyl substituents on thymidine in poly(dA-dT) all lead to transitions upon replication.

In a previous paper, we reported that O4-methyl dTTP can be incorporated into poly(dA-dT) in place of thymidine without distortion of the helical structure, but on replication it could behave as deoxycytidine and misincorporate dGTP. Only weak interactions are possible for any O4-modified T X A pair. While O4-alkyl T X G pairing should be favored, experiments to detect the ability of Escherichia coli DNA polymerase I (pol I) to utilize the triphosphate as dCTP were ambiguous. dTTPs with larger alkyl groups (ethyl, isopropyl) have now been synthesized and tested for their recognition as dTTP by pol I. Enhanced steric hindrance could be expected, particularly for O4-isopropyl dTTP, which has a three-carbon branched chain. However, both compounds behaved qualitatively like O4-methyl dTTP, being incorporated into poly(dA-dT) and then directing deoxyguanosine misincorporation by pol I. Quantitative comparisons of mutagenicity were not possible because of the finding that, unlike polymers made with O4-methyl dTTP, those made with ethyl or isopropyl dTTP were resistant to hydrolysis by using a variety of nucleases. The frequent misincorporations of dGTP would be expected to produce transitions in vivo. O4-ethyldeoxythymidine is very poorly repaired in vivo, which would also be expected for repair of O4-isopropyldeoxythymidine. Therefore, under suitable conditions, these particular carcinogen products are likely to be initiators of carcinogenesis.