A necessary first step in modeling the metabolic reaction network is a systematic procedure for determining the different reaction sequences which connect two metabolites. A software system (MPS) has been designed on the principles of artificial intelligence in order to address the problem of analysis and synthesis of metabolic pathways leading from one carbon-containing metabolite to another. MPS can be used to predict on a qualitative basis the effects of adding or deleting enzymatic activities to or from the cellular environment, to extract information about metabolic regulation, and to direct experiments in metabolic engineering. The main principles that have been used for the development of MPS are described along with case studies demonstrating the capabilities and potential applications of such a software system. The examples will examine carbon catabolic pathways and amino acid biosynthetic pathways. The catabolic pathways example concentrates on the conversion of glucose 6-phosphate to pyruvate. The output from MPS, which synthesized the classical catabolic pathways along with possible variations, leads to the identification of required genotypes/sets of enzymes that convert glucose 6-phosphate to pyruvate with different ATP and NAD(P)H coupling. The amino acid examples refer to the production of L-alanine from pyruvate and the identification of alternative pathways that perform this bioconversion.
The appropriate use of controllable promoters and plasmid origins of replication provide an opportunity for identifying operating strategies that maximize productivity of unstable recombinant cultures. This is demonstrated by first developing a kinetic model for product formation in recombinant cultures that exhibit both segregational and structural instability, and subsequently by identifying operating conditions that maximize productivity with respect to the base case (uncontrollable promoter and plasmid origin of replication).
Finally, an approach is being made to examine the validity of traditional-macroscopic relationships for the description of equilibrium cellular processes. A model is described that addresses these processes from a statistical point of view and results of this model are compared with analogous results obtained from traditional methods applied to a subset of the reactions that characterize the control mechanisms of the lac promoter.
