Automated Metabolic Rewiring Design for Optimized Succinate Production in Saccharomyces Cerevisiae

Succinic acid, or succinate, is the precursor of a wide range of consumer products and biotechnological applications. Hence, a sustainable microbial bioproduction of succinate by engineered Saccharomyces cerevisiae strain is one of the highest markets demands in the large-scale fermentation. In the perspective of novel design for competitive engineered yeast strains, in the present study we propose the application of a powerful pipeline based on non-linear optimization for in silico design of metabolic networks through specific gene deletions. The ad-hoc bio-inspired algorithm that we have developed uses a genome-scale model reconstruction of the metabolic network in S. cerevisiae and the distribution of metabolic fluxes predicted by flux balance analysis. Succinate production was obtained on a wide range of different media including anaerobic medium and with different combinations of carbon sources. Multiple in silico simulations of easy genetic manipulation of strains allow the selection of the most competitive ones within the approximated Pareto Fronts. The valuable gene deletions were iteratively and competitively selected through the constrained multi-objective optimization algorithm, which were run for a low number of objective function evaluations, followed by an ad-hoc post-processing policy keeping the significant changes for succinate production. Other constraints were also considered for the refinement of the pipeline, including the variability of the predictions, the allowable reduction on key fluxes within the metabolic network, and the role of essential genes. The resulting strains are scattered in the phenotypic space, with clusters of points showing similar growth and succinate production; most competitive strains can then be selected for further studies depending on other characteristics of interest, such as particular by-products or productivity. The potential of computational methodologies in design automation for metabolic engineering offers a powerful approach for improving the effective production of succinate and establishing a more efficient and much less time-consuming in vitro mutagenesis.