In-situ product recovery from a Baeyer-Villiger monooxygenase catalysed bioconversion

In-situ product recovery (ISPR) has been investigated for an industrially important Baeyer-Villliger monooxygenase catalysed bioconversion in order to enhance the productivity of the bioconversion process. A recombinant whole cell biocatalyst, Escherichia coli TOP 10 pQR239, expressing cyclohexanone monooxygenase from Acinetobacter calcoaceticus, was studied and the reaction catalysed was the conversion of the ketone substrate, bicyclo[3.2.0]hept-2-en-6-one, to the lactone products, (1R,5S)-3-oxabicyclo[3.3.0]oct-6-en-2-one (ee = 99%) and (1S,5R)-2- oxabicyclo[3.3.0]oct-6-en-3-one (ee = 94%). Initially the bioconversion was characterised with respect to the properties of the substrate and products, the E.coli TOPIO pQR239 biocatalyst and the interactions between them. The E.coli TOPIO pQR239 biocatalyst was found to have a maximum specific activity of 111 μmolmin-1g-1 at the optimum conditions of 37 °C and pH 7. Inhibition of the E.coli TOPIO pQR239 biocatalyst by the ketone substrate was observed at concentrations greater than 0.8 gl-1 with half maximal activity occurring at a ketone substrate concentration of 2.2 gL-1. Inhibition of the E.coli TOPIO pQR239 biocatalyst by the lactone products of the reaction was observed to be severe with half maximal activity occurring at a product concentration of just 1.1 g1-1 . The specific activity of the E.coli TOPIO pQR239 biocatalyst was found to decrease with a half life of 8.9 hours when the E.coli TOPIO pQR239 biocatalyst was maintained in a stirred tank reactor (STR) under optimal conditions in the absence of both the ketone substrate and the lactone products. This loss in specific activity of the E.coli TOP10 pQR239 biocatalyst was temperature dependent. A 2 1 fed-batch bioconversion process was designed based on a STR configuration, in order to overcome substrate inhibition effects. Several 2 1 fed-batch bioconversions were subsequently performed at varying substrate feed rates (0.8-3.2 gl-1hr-1). The maximum product concentration achieved was 3.6 gl-1, and the maximum space-time yield obtained was 1.85 gl-1hr-1. A high conversion of the ketone substrate could be obtained by varying the substrate feed rate to the STR during the process. A simple unsteady state mathematical model was developed to describe the fed-batch bioconversion process based on parameters determined from the earlier characterisation studies. This model was found to accurately predict the rate of production of the lactone products up to a substrate feed rate of 2.2 gl-1hr-1. For the implementation of ISPR with this process, adsorption of the lactone products onto a non-functionalised adsorbent was investigated. This ISPR technique was chosen in order to exploit the hydrophobicity of the lactone products as the basis for their separation from the bioconversion medium. A range of polymeric and carbonaceous adsorbents were initially characterised with respect to the adsorption of the ketone substrate and the lactone products. Adsorption isotherms were determined for each adsorbent for substrate and product adsorption separately in buffer, and substrate and product adsorption together in bioconversion medium. In addition the technique of molecular imprinting was studied to determine whether an adsorbent selective for the lactone products could be created. High throughput screening approaches to adsorbent characterisation were also investigated. On the basis of the adsorption isotherms the adsorbent, Amberlite XAD4 was chosen for the implementation of ISPR with the fed-batch bioconversion process. This adsorbent had a maximum capacity for the lactone products of 31 mggads-1 in bioconversion medium in the presence of the ketone substrate. This adsorption capacity was too low to facilitate the direct addition of adsorbent in the STR so ISPR was implemented via an external adsorption column through which bioconversion medium was recirculated. The E.coli TOP 10 pQR239 biocatalyst was retained in the STR by a microfiltration membrane unit. When the fed-batch ISPR bioconversion process was operated at a substrate feed rate of 1.20 gl-1hl-1 the product concentration obtained was increased by 85% to 5.8 gl-1. The mathematical model developed to describe the fed-batch bioconversion process was extended to incorporate ISPR and was found to accurately predict the performance of the process. In principle the generic approach to the selection and design of an ISPR bioconversion process described in this thesis can be applied to any bioconversion process while the reactor model will enable the rapid optimisation of reactor operating conditions.