Model‐based characterization of an amino acid racemase from Pseudomonas putida DSM 3263 for application in medium‐constrained continuous processes

Abstract

The amino acid racemase with broad substrate specificity from Pseudomonas putida DSM 3263 was overproduced and characterized with respect to application in an integrated multi-step process (e.g., dynamic kinetic resolution) that--theoretically--would allow for 100% chemical yield and 100% enantiomeric excess. Overexpression of the racemase gene in Escherichia coli delivered cell free extract with easily sufficient activity (20-50 U mg(-1) total protein) for application in an enzyme membrane reactor (EMR) setting. Model-based experimental analysis of a set of enzyme assays clearly indicated that racemization of the model substrates D- or L-methionine could be accurately described by reversible Michaelis-Menten kinetics. The corresponding kinetic parameters were determined from progress curves for the entire suitable set of aqueous-organic mixtures (up to 60% methanol and 40% acetonitrile) that are eligible for an integrated process scheme. The resulting kinetic expression could be successfully applied to describe enzyme membrane reactor performance under a large variety of settings. Model-based calculations suggested that a methanol content of 10% and an acetonitrile content of 20% provide maximum productivity in EMR operations. However product concentrations were decreased in comparison to purely aqueous operation due to decreasing solubility of methionine with increasing organic solvent content. Finally, biocatalyst stability was investigated in different solvent compositions following a model-based approach. Buffer without organic content provided excellent stability at moderate temperatures (20-35 degrees C) while addition of 20% acetonitrile or methanol drastically reduced the half-life of the racemase.

Integration of simulated moving bed chromatography and enzymatic racemization for the production of single enantiomers

“…Hence, the design and optimization of a fully integrated production system requires a detailed characterization and parameterization of all involved units as a function of interoperable solvent (desorbent) compositions. In the case of the enzymatic racemization, function of the amino acid racemase (AAR) in water/methanol solvent systems was already demonstrated (Bechtold et al, 2007b). We also reported the suitability of teicoplanin aglycone, a macrocyclic glycopeptide as a CSP for the SMB separation of amino acids in a similar solvent environment (Bechtold et al, 2006;Fuereder et al, 2012).…”

“…Pseudomonas putida DSM 3263 has been described before (Bechtold et al, 2007b). To obtain an enzyme preparation, cells were harvested by centrifugation and resuspended in aqueous buffer (50 mM NH 4 Ac, 40 µM pyridoxal-5-phosphate (PLP), adjusted to pH7 with ammonium hydroxide solution) to a concentration of 100 g cell wet weight (CWW) L -1 .…”

“…The parameters for the bi-Langmuir isotherm model have been determined previously using the perturbation method at 25% methanol concentration and 22 o C (Fuereder et al, 2014), as have been the parameters for the reversible Michaelis-Menten kinetics (Bechtold et al, 2007b). Enzyme concentration and the dilution factor δ were fixed at the value given in Table 1 unless stated otherwise.…”

“…For the proof of principle a constant conversion close to racemic concentration was targeted which was achieved by loading the EMR with an excess of AAR (> 6,000 U). The large excess of enzyme was applied in order to realize an experimentally simple solution to compensate for enzyme degradation that was previously reported for this enzyme under elevated methanol concentrations (Bechtold et al, 2007b) and thus to ensure constant conversion over time. The proof of concept-run was performed for 33 hours yielding virtually invariant process performance over the entire time period.…”

Model-based cost optimization of a reaction–separation integrated process for the enzymatic production of the rare sugar d-psicose at elevated temperatures

“…Assuming reversible Michaelis-Menten (rMM) kinetics, the K mS and K mP for substrate and product were first estimated using progress curve experiments (Bechtold et al, 2007). Each experiment consisted of 6 individual progress curves for the forward and backward reaction, respectively, recording the conversion over time from D-fructose to D-psicose and vice versa.…”