Cytosine Catalysis of Nitrosative Guanine Deamination and Interstrand Cross-Link Formation

Abstract

Effects are discussed of the anisotropic DNA environment on nitrosative guanine deamination based on results of an ab initio study of the aggregate 3 formed by guaninediazonium ion 1 and cytosine 2. Within 3, the protonation of 2 by 1 is fast and exothermic and forms 6, an aggregate between betaine 4 (2-diazonium-9H-purin-6-olate) and cytosinium ion 5. Electronic structure analysis of 4 shows that this betaine is not mesoionic; only the negative charge is delocalized in the π-system while the positive charge resides in the σ-system. Potential energy surface exploration shows that both dediazoniation and ring-opening of betaine 4 in aggregate 6 are fast and exothermic and lead irreversibly to E-11, the aggregate between (E)-5-cyanoimino-4-oxomethylene-4,5-dihydroimidazole E-10 and 5. The computed pair binding energies for 3, 6, and E-11 greatly exceed the GC pair binding energy. While 1 can be a highly reactive intermediate in reactions of the “free nucleobase” (or its nucleoside and nucleotide), the cyanoimine 10 emerges as the key intermediate in nitrosative guanine deamination in ds-DNA and ds-oligonucleotides. In essence, the complementary nucleobase cytosine provides base catalysis and switches the sequence of deprotonation and dediazoniation. It is argued that this environment-induced switch causes entirely different reaction paths to products as compared to the respective “free nucleobase” chemistry, and the complete consistency is demonstrated of this mechanistic model with all known experimental results. Products might form directly from 10 by addition and ring closure, or their formation might involve water catalysis via 5-cyanoamino-4-imidazolecarboxylic acid 12 and/or 5-carbodiimidyl-4-imidazolecarboxylic acid 13. The pyrimidine ring-opened intermediates 10, 12, and 13 can account for the formations of xanthosine, the pH dependency and the environment dependency of oxanosine formation, the formation of the classical cross-link dG(N ²)-to-dG(C2), including the known sequence specificity of its formation, and the formation of the structure-isomeric cross-link dG(N1)-to-dG(C2).