UC San Diego

Organisms must adapt to fluctuating environmental conditions to ensure survival. This dissertation investigates the metabolic adaptations to optimize resource allocation under stress in single and sporulating cells, as well as microbial communities, employing advanced computational models to elucidate these complex processes. First, we highlighted the biosynthetic costs of various biomass components, emphasizing the significant impact of high molecular weight lipids and the free energy contributions of specific amino acids on growth. Second, we focused on Bacillus subtilis, a gram-positive bacterium with significant industrial applications. We developed a ME-model, iJT964-ME, that integrates gene expression and metabolic reactions to predict enzyme production and biomass composition variations under stress conditions. This model demonstrated enhanced predictive capabilities, aiding in optimizing protein production strategies and simulating stress-induced metabolic shifts. Third, we delved into the metabolic differentiation during the sporulation of Bacillus subtilis. Utilizing a novel two-cell ME-model, SporeME2, we mapped metabolic exchanges between the mother cell and forespore, elucidating the nutrient dependency of the forespore on the mother cell. Model-guided experiments identified key metabolic interactions, providing insights into the proteome-wide metabolic differentiation essential for spore formation. Fourth, we addressed the challenges in reconstructing ME-models due to their complexity and resource intensity. We introduced coralME, an automated pipeline that accelerates ME-model reconstruction. This tool significantly reduced reconstruction time, facilitating the creation of functional ME-models for 21 diverse bacteria. The application of coralME to update existing ME-models and reconstruct new ones showcases its potential to advance next-generation systems biology research. Finally, we applied coralME to the human gut microbiome, reconstructing 495 ME-models to analyze dietary effects and integrate multi-omics data. Simulations revealed the impact of nutrient availability on bacterial growth and metabolism, providing mechanistic insights into microbial interactions and secretory profiles in health and disease contexts, particularly in inflammatory bowel disease (IBD). Collectively, this dissertation underscores the power of integrating ME-models with experimental data to understand and predict microbial metabolic adaptations in single cells and microbial communities. The developed tools and methodologies pave the way for future research in metabolic engineering, bioproduction optimization, and microbiome-based therapeutic interventions.