P160: Model-based design of Saccharomyces cerevisiae for improved amino acid production

The production of amino acids which serve as precursors of many important small molecules is tightly regulated in S.cerevisiae by feedback inhibition. Here we discuss how a combination of computational modeling techniques, Optknock and Ensemble Modeling, can be used to make an effective design of engineered strains that improves the production of the required amino acid. Using Optknock as the strain design algorithm often results in a strategy that requires multiple genetic manipulations to be performed simultaneously in order to observe appreciable growth coupling. Before validating this strategy experimentally, it is desirable to know which of these suggested manipulations exercise a greater control on the productivity of the process. In order to predict this, we need to have knowledge of the kinetics of major reactions in the system. For this purpose we used Ensemble Modeling (EM) framework which predicts the dynamic behaviour of cells using the steady state flux data of wild type and mutant strains. We modified the original EM framework to account for transport of metabolites across the mitochondrial membrane of S.cerevisiae. Initial strain design is done using Optknock to force the flux through aromatic amino acid pathway. The kinetic parameter data estimated using Ensemble Modeling is then used to predict which of these manipulations are most important to improve the flux.   Furthermore, this knowledge of kinetic parameters can be used to predict the other major targets that can improve the productivity of process and are not predicted using Optknock.