Accelerating Decarboxylation Reactions through the Formation of Bicarbonate

Enzyme-catalyzed decarboxylation reactions are often many orders of magnitude faster than the corresponding uncatalyzed processes. In both cases, these reactions are typically represented as the unimolecular decomposition of a carboxylate to a carbanion and CO2. Attempts to understand the catalytic proficiency of decarboxylases generally focus on enzymatic factors that would stabilize the produced carbanion and ignore the released CO2. However, earlier studies have shown that decarboxylation is slowed by internal return of the CO2-carbanion pair. This thesis illustrates how both aqueous and enzymatic decarboxylation reactions can be accelerated by producing alternatives to CO2 that decrease the extent of reversion. This work was initiated by the unexpected observation of general base catalysis in the decarboxylation of mandelylthiamin (MTh), an accurate model of the pre-decarboxylation intermediate formed on benzoylformate decarboxylase (BFDC). This catalysis is inconsistent with the direct production of CO2 from the carboxylate. Instead, detailed kinetic analysis suggests a mechanism involving hydration of the carboxylate and subsequent base-catalyzed decomposition to produce HCO3-. Computational analysis of a model decarboxylation illustrates that this hydrolytic mechanism avoids the issue of internal return. This is consistent with HCO3- being less electrophilic and more water-soluble than CO2 as these properties would facilitate the separation of the carbanion-HCO3- pair. Kinetic and computational studies were also performed with numerous structural analogues of MTh. The results indicate that the hydrolytic mechanism is facilitated by an intramolecular hydrogen bond to the carboxylate of MTh. This interaction reduces the electron density on the carboxylate carbon and promotes hydration. A serine residue (S26) in BFDC is proposed to prevent internal return by trapping the nascent CO2 as a carbonate monoester. We hypothesized that an active site water could act as a surrogate nucleophile in BFDC mutants lacking S26. This would be analogous to the hydrolytic decarboxylation of MTh, resulting in the production of HCO3- upon carbon-carbon bond cleavage. Using a coupled assay with carbonic anhydrase, a number of BFDC variants were determined to release CO2. This indicates that either CO2 is produced in these decarboxylation reactions or that bicarbonate is converted to CO2 prior to its release from the active site.Ph.D