Modeling bioreactors for the production of recombinant proteins in high-­‐cell density cultures of Escherichia coli

Abstract

The present work is focused on the study of recombinant protein production in high-­‐ cell density fed-­‐batch cultures of E.coli. In particular, the development of a model capable to predict Rhamnulose-­‐1-­‐Phosphate Aldolase (RhuA) production is the objective. Firstly, a qualitative and quantitative study about the variables involved in protein production has been made. This study has permitted the evaluation of the impact of the main experimental variables (inducer and biomass concentration at induction and the specific growth rate) on protein production (in mass and activity units). Using a Response Surface Methodology (RSM), that is a statistical methodology, it can be determined the optimal experimental conditions that conduce to a maximum in protein production, and set the operating working conditions. Secondly, because a deeper study about the importance of IPTG in inducible E.coli systems is needed, a model describing inducer uptake has been developed, calibrated and validated. IPTG uptake model has been developed in two steps: a) using a lacY deficient strain, non-­‐specific transport mechanisms have been modeled; b) in addition to non-­‐specific transport mechanisms, lactose permeases (specific transporting proteins for lactose –and IPTG) contribution has been added. It has been demonstrated that the model is capable to predict IPTG depletion from culture medium, not only for the model strain, but also for three different strains. Thirdly, a coupled model, composed by three different ones (biomass growth, IPTG uptake and protein production) has been proposed. In this case, a new protein production model has been presented, using as inputs the time evolution of the variables involved in the other two models. Protein production rate (expressed in mass) can be related to the amount of inducer bound to the repressor. The binding equilibrium depends on the intracellular concentration of IPTG along time, which is an output of the IPTG uptake model. Otherwise, biomass growth model is able to predict biomass concentration and the total volume into the bioreactor from the beginning of the batch phase. Finally, the protein production model, coupled with the IPTG uptake model, has been extended to the production of different proteins (Fructose-­‐6-­‐Phosphate Aldolase and ω-­‐Transaminase) using different expression systems. In this case, the expression system’s dependent parameters have been identified, and the model has demonstrated that, estimating those parameters is also capable to predict, properly, the protein production along time. To sum up, this work presents a new model, which contributes to the prediction of protein production in different inducible E.coli expression systems.

Date of Award	14 Nov 2014
Original language	English
Supervisor	Josep Lopez Santin (Director) & Carlos De Mas Rocabayera (Director)