ABSTRACT: Nitrite has recently been recognized as a storage form of NO in blood and as playing a key role in hypoxic vasodilation. The nitrite ion is readily reduced to NO by hemoglobin in red blood cells, which, as it happens, also presents a conundrum. Given NO's enormous affinity for ferrous heme, a key question concerns how it escapes capture by hemoglobin as it diffuses out of the red cells and to the endothelium, where vasodilation takes place. Dinitrogen trioxide (N(2)O(3)) has been proposed as a vehicle that transports NO to the endothelium, where it dissociates to NO and NO(2). Although N(2)O(3) formation might be readily explained by the reaction Hb-Fe(3+)+NO(2)(-)+NO?Hb-Fe(2+)+N(2)O(3), the exact manner in which methemoglobin (Hb-Fe(3+)), nitrite and NO interact with one another is unclear. Both an "Hb-Fe(3+)-NO(2)(-)+NO" pathway and an "Hb-Fe(3+)-NO+NO(2)(-) " pathway have been proposed. Neither pathway has been established experimentally. Nor has there been any attempt until now to theoretically model N(2)O(3) formation, the so-called nitrite anhydrase reaction. Both pathways have been examined here in a detailed density functional theory (DFT, B3LYP/TZP) study and both have been found to be feasible based on energetics criteria. Modeling the "Hb-Fe(3+)-NO(2)(-)+NO" pathway proved complex. Not only are multiple linkage-isomeric (N- and O-coordinated) structures conceivable for methemoglobin-nitrite, multiple isomeric forms are also possible for N(2)O(3) (the lowest-energy state has an N-N-bonded nitronitrosyl structure, O(2)N-NO). We considered multiple spin states of methemoglobin-nitrite as well as ferromagnetic and antiferromagnetic coupling of the Fe(3+) and NO spins. Together, the isomerism and spin variables result in a diabolically complex combinatorial space of reaction pathways. Fortunately, transition states could be successfully calculated for the vast majority of these reaction channels, both M(S)=0 and M(S)=1. For a six-coordinate Fe(3+)-O-nitrito starting geometry, which is plausible for methemoglobin-nitrite, we found that N(2)O(3) formation entails barriers of about 17-20 kcal mol(-1) , which is reasonable for a physiologically relevant reaction. For the "Hb-Fe(3+) -NO+NO(2) (-) " pathway, which was also found to be energetically reasonable, our calculations indicate a two-step mechanism. The first step involves transfer of an electron from NO(2)(-) to the Fe(3+)-heme-NO center ({FeNO}(6)) , resulting in formation of nitrogen dioxide and an Fe(2+)-heme-NO center ({FeNO}(7)). Subsequent formation of N(2)O(3) entails a barrier of only 8.1 kcal mol(-1) . From an energetics point of view, the nitrite anhydrase reaction thus is a reasonable proposition. Although it is tempting to interpret our results as favoring the "{FeNO}(6)+NO(2)(-) " pathway over the "Fe(3+)-nitrite+NO" pathway, both pathways should be considered energetically reasonable for a biological reaction and it seems inadvisable to favor a unique reaction channel based solely on quantum chemical modeling.