THE PHOTOREACTIVITY OF PYRIMIDINE‐PURINE SEQUENCES IN SOME DEOXYDINUCLEOSIDE MONOPHOSPHATES AND ALTERNATING DNA COPOLYMERS

— A reversed-phase HPLC system has been developed which separates the common nucleo-bases from the 6-methylimidazo[4,5-b]pyridin-5-one (6-MIP) produced on acid hydrolysis of a thymine-adenine photoadduct (TA*) that is formed between adjacent thymine and adenine bases in UV-irradiated polydeoxyribonucleotides. By measuring the relative amounts of adenine and 6-MIP in acid hydrolysates, this system has been used to investigate how polynucleotide conformation affects the yield of TA* in poly(dA-dT) irradiated at 254 nm. The photoreactivity of other pyrimidine-purine sequences has been examined with the deoxydinucleoside monophosphates d(TpI) and d(m5CpA) and with the alternating DNA copolymers poly(dA-dU), poly(dI-dC), poly(dG-dC) and poly(dA-dC).poly(dG-dT). Samples were irradiated at 254 nm in aqueous solution and in ice, and at wavelengths >290 nm with acetone as photosensitizer. A photoproduct resembling TA*, and giving 6-MIP on acid hydrolysis, was isolated from d(TpI) irradiated at 254 nm in solution or in ice; d(m5CpA) was comparatively unreactive. Acid hydrolysates of the irradiated DNA copolymers were screened by HPLC and by TLC and paper electrophoresis, for the presence of imidazo[4,5-b]pyridin-5-one, 6-MIP, or other species attributable to specific photoproduct formation. By this criterion, however, none of the copolymers showed evidence of significant photoreactivity in either their single- or double-stranded conformational states. The formation of mixed pyrimidine-purine photoadducts in DNA is therefore probably restricted to T-A doublets.