Genomatica has developed an integrated platform for the rapid development of next-generation cell-based factories for the production of small molecules, combining the use of state-of-the-art computational design tools with complementary experimental methodologies. The platform is designed to accommodate all types of chemical products, bridging various market segments, including both those endogenous to the organism and those which require introduction of novel pathways into the host strain. An automated model generating pipeline allows the rapid construction of genome-scale models for any microbe of industrial interest, leveraging information from our growing collection of high quality models. Constraint-based modeling techniques are used in conjunction with various data types to provide an accurate picture of intracellular fluxes in the wild-type or current production strain. OptKnock design algorithms are then implemented to predict metabolic engineering strategies that couple product formation to growth of the organism; i.e., the organism must produce the compound of interest in order to grow most efficiently. A complementary experimental approach, evolutionary engineering, is applied next which uses controlled selection pressure to optimize strain fitness and growth rate following genetic manipulations. In addition to achieving superior product yield, strains generated by this approach are suitable for continuous bioprocessing, due to their inherent genetic stability. As validation of this integrated platform, we will present an example of a microbial strain engineered for the growth-coupled production of a metabolite of commercial interest. A non-intuitive knockout strategy was designed and implemented, resulting in a moderate level of production. To increase the rates of both product formation and growth, these strains were then subjected to adaptive evolution using a specially designed machine to provide frequent serial dilutions. As a result of the evolution product yield increased 4-fold, to near-theoretical levels. The combined computational and experimental approach presented here has the potential to significantly expedite the efficient metabolic engineering of both traditional and nontraditional industrial organisms for continuous bioprocess applications.

Extended Abstract Status: Not Uploaded