Radical mechanisms for 1,5-dihydroflavin reduction of carbonyl compounds.

A discussion of the expected mechanism of radical reduction of carbonyl compounds by dihydroflavin as it relates to the structure of the carbonyl compound is provided. Factors which must be taken into account are the free energies of formation of the radical anions (CR2O), THE ACIDITY OF THE CONJUGATE ACID OF THE CARBONYL FUNCTION (R2C negative charge OH), and the stability of the carbanion species of the product [(-)CR2(OH)]. It is proposed that in those instances where the product alcohol can flavin radical to CR2O or CR2OH occurs, otherwise H transfer is the terminal step for dihydroflavin reductions. The free energy of formation of the radical species CH2OH obtained by acid-catalyzed electron transfer from dihydroflavin (Fred) to formaldehyde is shown to be less than the experimentally determined free energy of activation (increment F is not equal to exp) for the reduction of formaldehyde by dihydroflavin (i.e., Fred plus CH2O leads to Fox plus CH2OH). Therefore, the radical pair composed of dihydroflavin radical species serve as an intermediate in the reduction. Our proposed mechanism is: Fred plus CH2O ka[H3O] forms k-a[H2O] Frad CH2OH leads to kc Fox plus CH3OH. The value of kaah/k-a has been obtained from the calculated standard potential Eo' for Fred plus CH2O plus H forms Frad plus CH2OH. Assuming ka to represent a two-step process (i.e., Fred plus CH2O plus H k'a forms k'-a Fred plus CH2OH kb forms k-b Frad CH2OH, where ka equals (k'akb)/(k'-ak-b), the value of k'-a will equal 10-10 M-minus 1 sec-minus 1 and kb 10-9 M-minus 1 sec-minus 1. From k'a/k'-a and kb there can be computed the expected value of increment F is not equal to calc as a function of pH. Comparison of increment F is not equal to calc to increment F is not equal to exp reveals that increment F is not equal to calc varies from increment F is not equal to exp by only about 2kcal mol-minus 1 (8.4 kJ mol-minus 1) between pH 5 and 9. Similar considerations establish that radical intermediates should serve eminently well in dihydroflavin reduction of ethyl pyruvate, pyruvic acid, etc. In these cases, 1 e transfer should compete with H transfer to yield the carbanions as the immediate products. Similar comparisons suggest that dihydronicotinamide reduction proceeds via RPyH plus C negative charge O forms RPyH plus C-0minus +H forms -H RPyH plus C-OH forms RPy plus HC-OH.