Enzymatic analysis of isomeric trithymidylates containing ultraviolet light-induced cyclobutane pyrimidine dimers. II. Phosphorylation by phage T4 polynucleotide kinase

Phage T4 polynucleotide kinase (EC 2.7.1.78) proved incapable of catalyzing the phosphorylation of thymidylyl-(3'----5')-thymidine containing either a cis-syn-cyclobutane pyrimidine dimer (d-T less than p greater than T) or a 6-4'-pyrimidin-2'-one]pyrimidine photoproduct (d-Tp]-T), and similarly the UV-modified compounds of (dT)3 bearing either photoproduct at their 5'-end (d-T less than p greater than TpT and d-Tp]TpT). In contrast, the 3'-structural isomers of these trinucleotides (d-TpT less than p greater than T and d-TpTp]T) were phosphorylated at the same rate as the parent compound. These phosphorylatable lesion-containing oligonucleotides are quantitatively released from UV-irradiated poly(dA):poly(dT) by enzymatic hydrolysis with snake venom phosphodiesterase and alkaline phosphatase (Liuzzi, M., Weinfeld, M., and Paterson, M. C. (1989) J. Biol. Chem. 264, 6355-6363). By combining this digestion regimen with phosphorylation by polynucleotide kinase and gamma-32P]ATP, pyrimidine dimers were quantitated at the fmol level following exposure of poly(dA):poly(dT) and herring sperm DNA to biologically relevant UV fluences. The rate of dimer induction in the synthetic polymer, approximately 10 dimers/10(6) nucleotides/Jm-2, was in close agreement with that obtained by conventional methods. Dimers were induced at one-fourth of this rate in the natural DNA. Further treatment of the phosphorylated oligonucleotides derived from irradiated herring sperm DNA with nuclease P1 released the labeled 5'-nucleotide, thus permitting analysis of the nearest-neighbor bases 5' to the lesions. We observed a ratio for pyrimidine-to-purine bases of almost 6:1, implicating tripyrimidine stretches as hotspots for UV-induced DNA damage.