Metabolic growth coupling strategies for in vivo enzyme selection systems

Whole-cell biocatalysis facilitates the production of a wide range of industrially and pharmaceutically relevant molecules from sustainable feedstocks such as plastic wastes, carbon dioxide, lignocellulose, or plant-based sugar sources. The identification and use of efficient enzymes in the applied biocatalyst is key to establishing economically feasible production processes. The generation and selection of favorable enzyme variants in adaptive laboratory evolution experiments using growth as a selection criterion is facilitated by tightly coupling enzyme catalytic activity to microbial metabolic activity. Here, we present a computational workflow to design strains that have a severe, growth-limiting metabolic chokepoint through a shared class of enzymes. The resulting chassis cell, termed enzyme selection system (ESS), is a platform for growth-coupling any enzyme from the respective enzyme class, thus offering cross-pathway application for enzyme engineering purposes. By applying the constraint-based modeling workflow, a publicly accessible database of 25,505 potential and experimentally tractable ESS designs was built for Escherichia coli and a broad range of production pathways with biotechnological relevance. Model-based analysis of the generated design database reveals a general design principle that the target enzyme activity is linked to overall microbial metabolic activity, not just the synthesis of one biomass precursor. Furthermore, the use of currently available genome-reduced strains or one preeminent carbon source does not significantly enable or improve growth-coupling of target enzymes. Most importantly, observed trade-offs between the predicted viability of ESSs, the design-inflicted metabolic perturbations, and the coupling strength suggest that a suboptimal coupling has benefits regarding the experimental implementation of ESSs and growth-coupling in general. Overall, the computed design database, which is accessible through https://biosustain.github.io/ESS-Designs/, and its analysis lay the foundation for generating valuable in vivo ESSs for a range of biotechnological applications.