Decarboxylation, CO2 and the reversion problem.

Decarboxylation reactions occur rapidly in enzymes but usually are many orders of magnitude slower in solution, if the reaction occurs at all. Where the reaction produces a carbanion and CO2, we would expect that the high energy of the carbanion causes the transition state for C-C bond cleavage also to be high in energy. Since the energy of the carbanion is a thermodynamic property, an enzyme obviously cannot change that property. Yet, enzymes overcome the barrier to forming the carbanion. In thinking about decarboxylation, we had assumed that CO2 is well behaved and forms without its own barriers. However, we analyzed reactions in solution of compounds that resemble intermediates in enzymic reaction and found some of them to be subject to unexpected forms of catalysis. Those results caused us to discard the usual assumptions about CO2 and carbanions. We learned that CO2 can be a very reactive electrophile. In decarboxylation reactions, where CO2 forms in the same step as a carbanion, separation of the products might be the main problem preventing the forward reaction because the carbanion can add readily to CO2 in competition with their separation and solvation. The basicity of the carbanion also might be overestimated because when we see that the decarboxylation is slow, we assume that it is because the carbanion is high in energy. We found reactions where the carbanion is protonated internally; CO2 appears to be able to depart without reversion more rapidly. We tested these ideas using kinetic analysis of catalytic reactions, carbon kinetic isotope effects, and synthesis of predecarboxylation intermediates. In another case, we observed that the decarboxylation is subject to general base catalysis while producing a significant carbon kinetic isotope effect. This requires both a proton transfer from an intermediate and C-C bond-breaking in the rate-determining step. This would occur if the route involves the surprising initial addition of water to the carboxyl, with the cleavage step producing bicarbonate. Interestingly, some enzyme-catalyzed reactions also appear to produce intermediates formed by the initial addition of water or a nucleophile to the carboxyl or to nascent CO2. We conclude that decarboxylation is not necessarily a problem that results from the energy of the carbanionic products alone but from their formation in the presence of CO2. Catalysts that facilitate the separation of the species on either side of the C-C bond that cleaves could solve the problem using catalytic principles that we find in many enzymes that promote hydrolytic processes, suggesting linkages in catalysis through evolution of activity.