CHEMISTRY OF PEROXIDIC TETRAHEDRAL INTERMEDIATES OF FLAVIN

By means of pulse radiolysis 4a-peroxy intermediates of normal and 5-alkylated flavins were produced and the kinetics of their decay into flavin and the corresponding hydroperoxide was investigated as a function of the pH. The neutral and proton-catalyzed breakdown of the 4a-intermediates of 5-alkylated flavinium cations on the one hand and of 5-protio flavins on the other was very similar. It was concluded that the rate-determining step in the neutral decomposition of normal flavin 4a-peroxides is a heterolysis along the C(4a)-O bond which is catalyzed by water as a general acid. The species initially produced consist of a N(5)-protonated flavinium cation, a neutral hydroperoxide, and a hydroxide ion. The process is completed by rapid deprotonation of the flavinium cation to yield the neutral flavin. By combination of kinetic and thermodynamic data determined in this and other laboratories, the energetics of the autoxidation of 1,5-dihydroflavin was resolved into individual steps. The proton-catalyzed breakdown of flavin 4a-peroxides is initiated by a proton-assisted expulsion of neutral hydroperoxide leaving behind the N(5)-protonated flavinium cation. The attenuation of proton catalysis with decreasing pH indicates thermodynamic protonation of the 4a-intermediates around pH 3. The site of protonation is presumably the N(5) or the N(10) atom. The hydroxide ion catalyzed breakdown of the 4a-species is best interpreted by assuming the rate-determining step to be deprotonation of the N(5)-H site followed by rapid explusion of the hydroperoxide anion and neutral flavin. This picture demands the microscopic pK(a) of the N(5)-H group to be below 17. The possible role of enzymes in stabilizing the 4a-intermediates against breakdown into flavin and hydroperoxide is discussed. It is suggested that an apolar, hydrophobic pocket may be the chief stabilizing factor. In such an environment, the transition state for heterolysis and homolysis may approach each other. Finally, the bond strength of the peroxidic O-O bond was calculated from recent thermodynamic data. This bond turns out to be weaker (< 26 kcal/mol) than the O-O bond in any known linear peroxide. From the finding that the O-O bond is weaker than the C(4a)-O bond it is argued that, in sufficiently hydrophobic enzymes, monooxygenation may be initiated by homolysis of the O-O bond. It is suggested that the comparable strengths of the C(4a)-O and O-O bonds may be the prime reason for the versatility of flavin enzymes.