Varga, Edward George; (1997) Modelling of the downstream processing of a recombinant intracellular enzyme. Doctoral thesis (Ph.D), UCL (University College London).

Text Modelling_of_the_downstream_pr.pdf Download (9MB)

Abstract

This thesis examines the generation and verification of robust and predictive models for key unit operations in the downstream processing of a recombinant intracellular enzyme. Computer models are becoming increasingly used in the biotechnology industry to assess process feasibility, ensure appropriate specification and achieve efficient operation of bioprocesses. With increasing pressure to bring bioproducts to market in the shortest possible time and at low cost, it is essential that these models can be generated early in the development cycle with a minimum of time and expense. Regulatory concerns over the use of recombinant systems also require a highly consistent product and a large degree of confidence in any final design. Models must therefore be verified against real process data in order to demonstrate and improve their validity. The specific operations under study were cell disruption by high pressure homogenisation, enzyme fractionation and concentration by ammonium sulphate precipitation, and cell harvesting, debris removal and precipitate separation by disc stack centrifugation. Existing models based on the recovery of alcohol dehydrogenase (ADH) from natural bakers' yeast were applied to the processing of a protein engineered form of the enzyme (pe-ADH) and their validity assessed by comparison with experimental data obtained from small amounts of fermented recombinant material. Differences were observed in the processing of the natural and protein engineered enzyme and, although parameter re-evaluation was generally required to maintain model accuracy, the models were found to be generic in nature. Individual unit models were then arranged and linked in a modular fashion to produce a whole process model. Verification of the process model based on recombinant system parameters against pilot-scale data showed sufficient accuracy for use in feasibility studies (±30%). The inability to describe precipitant breakage in the disc stack centrifuge feed-zone was the major source of model inaccuracy. The work highlighted the utility of scale-down systems as an aid to rapid and cost- effective model development. The importance of experimental verification in order to provide confidence in model accuracy and scalability was stressed, as was the need for modelling to be an integrated activity during bioprocess development.

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as JSONXMLCSV

Archive Staff Only

View Item