(264f) Engineering Co-Dependance of Microbial Communities for Biochemical Production

The global increase in environmental awareness, combined with the declining costs of DNA technology, has made biochemical production a serious alternative to conventional chemical synthesis. Despite industrial successes, significant efforts are still aimed at improving production titers to expand the range of commercially available products. To that effect, alternative production methods and tools, such as biosensors and metabolic models, have been developed over the past two decades. One promising method involves dividing synthetic metabolic pathways among different strains. Research has shown substantial improvements for biomolecules such as naringenin and 1,3-propanediol using this approach. Communities can better mitigate the yield-reducing impacts of allosteric enzymes by spatially separating enzymes among different strains. Furthermore, balancing enzymatic fluxes can be easily achieved by modulating the biomass concentration of each strain. However, new challenges may arise within communities, as the inclusion of a new population can lead to competition, and unbalanced growth rates can result in community collapse.

Several strain design algorithms have been proposed to aid in designing stable synthetic communities. However, these often rely on pre-selected strategies and do not exhaustively explore all possibilities. Therefore, we propose a new method that utilizes the minimal cut sets algorithm to enumerate all knockout interventions that could lead to co-dependence. Using this method, we calculated 10,378 strategies for a community model consisting of two E. coli genome-scale stoichiometric models. These strategies ensure that both strains are unable to grow unless all community members are growing. The method also extends to larger communities of three members. Additionally, the algorithm successfully computes solutions for bioproduction purposes by specifying production targets for the community. The solution ensures that glucose uptake is coupled with naringenin production and prevents retroactive inhibition by dividing the pathway among the two strains. By addressing these issues, the calculated interventions ensure co-dependence and high production yields while avoiding problems associated with allosteric enzymes.

Our developed method is the first to allow for the complete enumeration of all intervention strategies that ensure co-dependence. It is unbiased and only requires a list of metabolites known to be readily importable by the selected strain. In nature, communities can achieve diverse functions and grow on complex substrates. Unlocking their potential for bioproduction requires better control of each member to prevent competition and instability. Our algorithm can enable the facile design of communities for biochemical production