Browsing by Author "Reuss, Matthias (Prof. Dr.-Ing)"

Now showing 1 - 1 of 1

Open Access
Metabolic egineering of the valine pathway in corynebacterium glutamicum : analysis and modelling
(2007) Magnus, Jørgen Barsett; Reuss, Matthias (Prof. Dr.-Ing)
The functionality of the intracellular reaction network in a Corynebacterium glutamicum valine production strain was investigated with special focus on the valine / leucine biosynthesis pathway. The aim was to gain a quantitative understanding of the behaviour of the reaction network. The methods required to do so were developed, and enzyme targets for the further optimisation of the investigated strain were identified. The intracellular metabolite concentrations were observed during a transient state by performing a glucose stimulus experiment. A mathematical model describing the in vivo reaction dynamics of the valine / leucine pathway was developed and a metabolic control analysis was performed based on the data from the stimulus experiment and the dynamic model. The thermodynamic driving forces in the valine / leucine pathway were analysed. The optimal procedure for the stimulus experiment with respect to obtaining a useful data set for the modelling and analysis was identified. Samples were taken at sub-second intervals and the concentrations of 26 metabolites from the valine / leucine pathway and the central metabolism were measured. A very fast response to the stimulus was observed in most intracellular metabolites with for example a 3-fold increase in the pyruvate concentration within one second. The connectivities of the metabolites around the ketoisovalerate branchpoint were investigated using a time series analysis. The kinetic model consisted of a system of differential equations defined by setting up material balances on the metabolites. The model can simulate the concentrations and fluxes in the valine and leucine pathway accurately during the transient state. The implementation of a model selection criterion based on the second law of thermodynamics was demonstrated to be essential for the identification of realistic and unique models. Other, alternative methods of setting up a kinetic model were also investigated. The alternative models included a mechanistic model of the valine / leucine pathway and a large linlog model of the whole metabolism of the strain. The mechanistic model was not capable of simulating the measured concentrations due to the limitations of its elasticities. The instability of the whole cell model made it inappropriate for a metabolic control analysis and further interpretation. However, the simulation of the whole metabolism of the strain provides a proof of concept for the whole cell modelling approach and shows in which direction metabolic modelling will develop in the future. Both data driven and model based methods were used to analyse the control hierarchy in the valine / leucine pathway. In addition, predictions of the effect of changes in the enzyme levels were made based on the model. In an optimisation study the enzyme levels were optimised with respect to the valine flux. Based on the acquired understanding of the behaviour of the reaction network the following targets for further strain development were identified: 1. Overexpression of the valine translocase 2. Implementation of an inhibition resistant AHAS enzyme and possibly further overexpression. 3. Removal of the overexpression of the gene coding for DHAD on the plasmid to save the cell the burden of overproducing this enzyme which has negligible influence on the valine flux. 4. Modification of the central carbon metabolism to increase pyruvate availability. The identification of the targets for strain development demonstrates the usefulness of a kinetic model in metabolic engineering and in the general understanding of metabolic control. The concentration data and the kinetic model were used to analyse the thermodynamic driving force, i.e. the reaction affinity, in the valine / leucine pathway. The concept of a reaction resistance was introduced to relate the driving force to reaction rate in analogy with Ohm’s law. This provided a new angle of analysing metabolic networks. The linear relation between reaction rate and affinity which apply for uni-uni reactions can not be assumed to be valid for bi-bi reactions operating far from equilibrium. The theory of metabolic control analysis was extended to include also the reaction potential and the reaction resistance. Reactions far from equilibrium are controlled almost entirely through the changes in the resistance while reactions closer to equilibrium are also affected by changes in the affinity.