The one-electron adducts of pyrimidine bases in low-temperature glasses and their significance to gamma-irradiated frozen aqueous DNA: An EPR study.

By the use of model compounds such as pyrimidine bases, nucleosides and nucleotides in low-temperature aqueous glasses, the primary reduction product of gamma-irradiated frozen aqueous DNA has been examined using EPR spectroscopy. Ionisation of frozen aqueous solutions of lithium chloride, sodium chloride, methanol, ethylene glycol and sodium hydroxide generates positive 'holes' and mobile electrons. In the absence of suitable solutes, such as DNA bases, both species are trapped within the glass. In the presence of pyrimidine derivatives, however, the mobile electrons react to form n*-anions. For thymine and uracil derivatives the n*-anions, as monitored by EPR spectroscopy, give rise to anisotropic doublets in both H2O and D2O matrices (chapters 2 and 4 respectively). For cytosine derivatives, however, although doublets are observed in D2O, triplets are detected in H2O systems corresponding to an extra 12G splitting (chapter 3). This splitting is assigned to a proton attached to the exocyclic amino nitrogen atom (N4) and is taken as clear proof of heteroatom protonation of the radical anion of cytosine. No direct EPR evidence for the protonation states at 77K of either the thymine or uracil n*-anions is obtained in any of the glasses. On warming these systems, however, protonation of the n*-anions of both thymine and uracil at a carbon atom position (C6) is observed giving rise to C6 H-adduct radicals. No such species is detected for cytosine on annealing. Exposure of frozen aqueous DNA to 60Co gamma-rays at 77K gives electron-loss and -gain centres localised on the bases. The electron-gain centres are believed to be a mixture of pyrimidine n*-anions (Py-, i.e. C- + T-). This assumption is based on the fact that, on annealing, the Py-. doublet is only partially converted into the 5,6- dihydrothymine-5-yl radical TH, the resulting radical having a completely characteristic octet EPR spectrum. The results suggest that ca. 36% of the doublet is due to T-centres, the remainder (64%) being assigned to C-centres. It is argued that ejected electrons move through stacked bases, becoming trapped at cytosine or thymine depending upon the relative rates at which C- and T- are protonated to give C-(H+), protonated at N3 (not N4), and T-(H+), probably protonated on a carbonyl oxygen. For this to be correct then interconversion between C- and T on annealing is unlikely.