A Multi-Scale Modeling Concept and Computational Tools for the Integrative Analysis of Stationary Metabolic Data

Abstract

Metabolic Engineering aims at the systematic analysis and targeted manipulation of the metabolism in biotechnologically utilized microorganisms [1,2]. Recently, consistent stationary in vivo data sets of intracellular metabolite concentrations, fluxes and specific enzyme activities have become available for this purpose [3,4]. For the integrative analysis of the metabolic data at hand, a novel multi-scale modeling concept, computeralgebraic methods and efficient numerical algorithms are proposed. The available metabolic data are typically afflicted with comparatively large measurement errors. Therefore, reliable comprehensive error estimations are essential for the reasonable interpretation of consecutive outcomes, such as simulation results.The concepts, methods and algorithms are first presented as universal methods and subsequently applied to the anaplerotic regulation in lysine-producing Corynebacterium glutamicum. A multi-scale model is set up, fitted to the available experimental data and validated by the prediction of further experiments. This model is capable of forecasting the quantitative effect of changes in the specific activity of anaplerotic enzymes, namely phosphoenolpyruvate carboxylase, pyruvate carboxylase and phosphoenolpyruvate carboxykinase, on lysine productivity and yield.