(48c) Characterizing the Biocatalytic Properties of Cytosolically Expressed Penicillin G Acylase from E. coli for the Synthesis of Cephalexin and Comparison with ?F24A Mutant

Beta-lactam antibiotics have made large contributions in the fight against bacterial infections ever since their discovery in 1928. Currently the industrial production of beta-lactam antibiotics is conducted via batch-wise chemical synthetic pathways consisting of multiple reaction steps, use of protecting groups, harsh solvents and very low temperatures, generating large amounts of hazardous waste. To approach this, a great deal of research has been conducted in developing a more efficient enzymatic synthetic processing pathway. The biocatalyst of interest is penicillin G acylase from E. coli (EcPGA). The combination of distinct acyl-donor moieties and nucleophilic beta-lactam moieties results in the synthesis of different semi-synthetic beta-lactam antibiotics. While wild-type EcPGA has been shown to have poor synthetic properties due to high product affinity, protein engineering on the binding pocket has altered the binding properties of substrates and products to enhance specificity.

We have improved the expression of wild -type EcPGA 20-fold to 45 mg/L culture. Additionally, we have characterized the synthetic properties of wild type EcPGA, βF24A, and Assemblase®, a commercial variant of EcPGA for cephalexin synthesis. We found that Assemblase® possessed marginally improved synthetic parameters over βF24A and wild-type enzymes. Higher product accumulation during long-term reaction experiments confirmed this observation. Next, wild type EcPGA was immobilized on various epoxy-functionalized supports to enable recycling of the biocatalyst between reactions. Approximately 30% of free enzymatic activity was retained after immobilization with decreased apparent activity attributed to diffusion limitations, as suggested by the decreasing apparent specific activity with increasing particle sizes. The synthesis to hydrolysis ratio decreased slightly, possibly due to internal product accumulation due to diffusion limitations. We are currently modeling the reaction and diffusion within the carrier pores using a reaction diffusion model to predict optimum enzyme loading to achieve high productivity while minimizing reduction of the synthesis to hydrolysis ratio.