New methods and models for constraint-based modeling of the microbial metabolism

Constraint-based metabolic modeling has been an important technique in broadening and deepening understanding of microbial metabolism. This thesis describes new models and methods which extend previous techniques in constraint-based modeling. Constraint-based modeling uses mass balances, flux capacity, and reaction directionality constraints to predict fluxes through metabolism, but kinetic rate laws have not been extensively used. In the first study, we formulated and solved an in vivo kinetic parameter estimation problem using multi-omic data sets for Escherichia coli. The kinetic model was then used to calculate the maximum and minimum possible flux values for individual reactions. Incorporating these kinetically-derived flux limits into the constraint-based metabolic model improved predictions for uptake and secretion rates and intracellular fluxes. Computational strain design techniques employ constraint-based modeling to determine metabolic and regulatory network changes which are needed to improve chemical production. These methods identify deletions, additions, downregulations, and upregulations of metabolic genes that will increase biological production of a desired metabolic product. In the second study, we proposed a new strain design method with continuous modifications (CosMos) that provides strategies for deletions, downregulations, and upregulations of fluxes that will lead to the production of the desired products. We found that the method was able to find strain design strategies that required fewer modifications and had larger predicted yields than strategies from previous methods. Streptomyces is a genus of gram-positive, spore-forming anaerobic actinobacteria. The genus contains prolific antibiotic producers, and a majority of the world's antibiotics that are produced by microbes originate from Streptomyces species. In a third study, we created genome-scale metabolic models for two streptomycetes, Streptomyces avermitilis and Streptomyces coelicolor, and used these models to investigate their unique catabolic and antibiotic production capabilities using constraint-based modeling. Based on growth phenotypes of the species on a variety of carbon sources, catabolic pathways were rediscovered in Streptomyces that were only recently described in distantly-related soil bacteria. The new constraint-based models can be used in future studies of S. coelicolor and S. avermitilis to enhance production of antibiotics and other constraint-based modeling analysis of the species.