Improving Methanosarcina genome-scale models for strain design by incorporating cofactor specificity and free energy constraints

Genome-scale metabolic models have the potential to revolutionize synthetic biology by informing the development of modified organism strains with enhanced production of desired compounds. A major caveat to this dogma is the lack of extensive experimental data for validating the models of less-studied organisms, relegating the confidence in the capabilities of these models to a much lower level than could potentially be attained with more comprehensive knowledge of organism metabolism. The process of enhancing the predictive capabilities of such models requires an iterative process of ongoing improvement to better reflect established knowledge of organism-specific biology. This work reports modifications to models of the metabolic networks of two methane-producing Methanosarcina, the iMB745 model of M.acetivorans, and the iMG746 model of M.barkeri. With these modifications, I integrated new experimental data and resolved cofactor specificity for reactions that utilize NAD and NADP. I also developed a new method for adding additional to these models by including free energy data for exchange reactions, thus allowing use of experimental data to restrict model predictions to flux distributions that satisfy the second law of thermodynamics. The updated models are each a more extensively curated body of data than the original models that more accurately reflect wet lab observations than previous iterations. I tested the updated models for their utility as metabolic engineering tools in two ways: (1) by creating models of mutants predicted to consume ethanol and pyruvate, two substrates with immense potential as carbon sources; and (2) by searching for potential knockout targets to enhance methane production in silico. These efforts to improve the models of the M.acetivorans and M.barkeri metabolic networks can serve as a blueprint for understanding cofactor specificity in a range of organisms, and the novel approach for integrating thermodynamics-based constraints by adding free energy data serves as a general tool for improving the constraint based approach to simulating the function of metabolic networks