A STRUCTURED, MINIMAL PARAMETER MODEL OF THE CENTRAL NITROGEN-METABOLISM IN SACCHAROMYCES-CEREVISIAE - THE PREDICTION OF THE BEHAVIOR OF MUTANTS

In order to enable future pathway engineering of a complex system, suc h as the nitrogen metabolism in yeast, mathematical modelling tools ha ve to be developed. The stoichiometric and biochemical characteristics of the glutamate and glutamine nodes (the Central Nitrogen Metabolism , CNM) are qualitatively known. Quantitative knowledge about the dynam ics of the network lacks and needs to be developed for metabolic repro gramming. A model-based-experiment approach is proposed in which the d evelopment of a model initiates new experiments of which the results t hen improve the model. As a first step in this iterative system identi fication cycle, recent experimental data, both qualitative and quantit ative, obtained from defined studies on the CNM of the yeast Saccharom yces cerevisiae have been translated into an initial mathematical mode l. The model approach is based on a combination of Flux Analysis and s imple enzyme kinetics. The model is constructed using nonlinear Ordina ry Differential Equations and regulation of the synthesis and activity of key enzymes of the CNM is included. The parameters of the model ar e estimated with a constrained Least Squares algorithm using the stead y-state and dynamic pulse data of a glutamine limited continuous cultu re. The resulting model describes a continuous culture of a wild-type strain correctly and in general the trends of the dynamic behaviour af ter both glutamine and ammonia pulses to this culture are good. Inclus ion of countercurrent reactions and compartmentation in the model is e ssential for the descriptive quality of the model under dynamic condit ions. It is clear that more experimental work is needed. The model ind icates that the GOGAT/Glutamine Synthetase (GS) pathway plays a more i mportant physiological stabilizing role in yeast than is generally ass umed. New, model-based, experiments have to investigate the function o f GOGAT, especially under dynamic conditions. Also redox cofactors and ATP have to be measured. The resulting model is validated with data o f similar experiments with a GS-mutant. The quality of the prediction of the behaviour of the mutant is comparable to the descriptive proper ty, which is a very promising result, taking into account the limited dataset compared to the system complexity. (C) 1998 Academic Press Lim ited.