Determination of intrinsic enzyme reaction kinetics using thermodynamic activities

Solvent selection plays a major role in economically competitive biocatalytic reaction systems. Until now, solvent selection relies on trial and error, since thermodynamic phenomena and kinetic effects during reactions are poorly understood. To allow for rational medium engineering, the effects of organic solvents on biocatalytic reaction systems have to be identified. These effects can be divided into (I) changes of the thermodynamic activities of the reactants and (II) changes of the intrinsic enzyme kinetic parameters. By combining high-throughput experimentation with reaction kinetic modeling and thermodynamic prediction, this work aims for the discrimination of thermodynamic and kinetic effects on oxidoreductase reaction in water-miscible monophasic organic solvents, in order to provide a basis for a rational choice of these solvents. The work can be divided into three packages. First, a thorough quantitative inactivation study is applied on Candida parapsilosis carbonyl reductase (CPCR2) and on Lactobacillus brevis alcohol dehydrogenase (LbADH) as model catalysts. Possible inactivation phenomena are quantified on microliter scale with respect to their effect on kinetic assays in order to derive a suitable mathematical inactivation model. For CPCR2, the results demonstrate that interface interactions and dimer dissociation are the main reasons for inactivation resulting in a complex inactivation scheme. For LbADH, tetramer dissociation has been observed at low protein concentration. This can be mathematically described by a first-order inactivation model. Second, reliable experimentation in micro-scale enzyme assays is required. Thus, typical problems occurring in MTPs such as temperature deviations or evaporation were evaluated. Internal thermal gradients within a polystyrene MTP of up to 2.2° C and even higher deviations from the set temperature were observed, which caused a variation in the corresponding enzyme activity of up to 20% and the deviations in relation to experiments in cuvettes by up to 40%. Subsequently, the reproducibility and precision of enzyme kinetic assays in different microliter scale systems were thoroughly analyzed. The initial reaction rates increased systematically from polystyrene MTPs to quartz MTPs to quartz cuvettes, whereas the experimental errors decreased in the same order (18 to 2%). While the microkinetic parameters vary up to an order of magnitude between different systems, the corresponding macrokinetic parameters lie in the same range for all systems varying only by up to a factor of 2 to 3.Third, MTBE was selected as model solvent providing reasonable LbADH stability and effect on thermodynamic activity as predicted using the COSMO-RS model. Kinetic experiments at both identical and purposefully different thermodynamic activities were used for kinetic modeling. Although MTBE seems to have an overall positive effect on the enzymatic reaction if examined on basis of concentration, solvation-corrected data suggest a negative effect of increasing MTBE concentration on the enzyme. It is assumed that MTBE hinders the intrinsic enzyme catalytic steps and compromises the performance of the enzyme, which is also indicated by the kinetic parameters. Overall, the importance of rational analysis and understanding of the experimental environment in biocatalysis is highlighted in this work. Important steps were taken to develop a proof of concept for kinetic modeling based on thermodynamic activities. This might provide better understanding of the behavior of biocatalysts in organic solvents. Solvent-enzyme interactions might be further elucidated rationally and, thereby, lead to a rational medium engineering