Is "Frank" DNA-strand breakage via the guanine radical thermodynamically and sterically possible?

Using the reduction potential of one-electron oxidized guanosine in water a nd the pK(a) values of the radical and of the parent, the NI-H bond energy of the 2'-deoxyguanosine moiety is determined to be (94.3 +/- 0.5)kcal mol( -1). Using the DFT method, the energy of the N1-centered guanosine radical is calculated and compared with those of the C1 '- and C4 ' -radicals forme d by H-abstraction from the 2'-deoxyribose moiety of the molecule. The resu lt is that these deoxyribose-centered radicals appear to be more stable tha n the N1-centered one by up to 3kcal mol(-1) Therefore, H-abstraction from a 2'-deoxyribose C-H bond by an isolated guanosine radical should be thermo dynamically feasible. However, if the stabilization of a guanine radical by intrastrand pi-pi interaction with adjacent guanines and the likely loweri ng of the oxidation potential of guanine by interstrand proton transfer to the complementary cytosine base are taken into account, there is no more th ermodynamic driving force for I-I-abstraction from a deoxyribose unit. As a further criterion for judging the probability of occurrence of such a reac tion in DNA, the stereochemical situation that a DNA-guanosine radical face s was investigated utilizing X-ray data for relevant model oligonucleotides . The result is that the closest H-atoms from the neighboring 2'-deoxyribos e units are at distances too large for efficient reaction. As a consequence , H-abstraction from 2 ' -deoxyribose by the DNA guanine radical leading su bsequently to a "frank" DNA strand break is very unlikely. The competing re action of the guanine radical cation with a water molecule which eventually yields 8-oxo-2'-deoxyguanosine (leading to "alkali-inducible" strand break s) has thus a chance to proceed.