(686c) Cellular Economics of Metabolite Synthesis Determine Ratios of Microbial Trading Partners

Metabolite cross-feeding is central to the stability, resilience, and productivity of microbial communities, spanning from chronic wounds to nutrient cycling in soils. Most naturally occurring microorganisms live in complex consortia encapsulated in self-produced polymer matrices known as biofilms. The complexity of these natural systems limits the ability to dissect and quantify important design parameters. Synthetic ecology can be used to assemble consortia with known functions that can decode the rules of consortia assembly and metabolite exchange.

Here, we used synthetic ecology to study the cellular economics of metabolite exchange between two obligate cross-feeding Escherichia coli strains. The consortium was comprised of two engineered strains, an arginine secreting strain that cannot catabolize the sugar lactose (L^-R⁺), and an arginine auxotroph which can catabolize lactose and secrete organic byproducts including pyruvate (L⁺R^-). The metabolic cost to produce each exchanged metabolite varied between strains. Arginine is a large, energy dense amino acid that is metabolically expensive to synthesize but the required cross feeding flux is relatively small. On the other hand, pyruvate is a relatively inexpensive molecule to synthesis but the flux to maintain the consortium is much higher than the flux required for arginine.

The consortium was grown as planktonic and biofilm cultures which resulted in substantially different ratios of the strains which is interpreted as distinct exchange costs. Planktonic cultures consistently exhibited a 95:5 ratio of organic acid producer cells (L⁺R^-) to arginine producer cells (L^-R⁺) regardless of inoculation ratio. Conversely, biofilm cultures consistently had an aggregate 50:50 cell ratio of the strains. The biofilm cultures were spatially analyzed using laser microdissection microscopy and qPCR; higher cell frequencies for the arginine producer (L^-R⁺) were located at the air interface which had higher O₂ concentrations based on microelectrode measurements. Diffusion limitation of O₂ shifted the strain frequencies along the vertical axis.

The experimental data was analyzed using in silico systems biology to test possible hypotheses explaining consortium behavior. Differences in O₂ availability altered the metabolic cost of producing pyruvate but not arginine and the experimental cell ratios reflected these changes. The predictions were further tested using isothermal microcalorimetry experiments which varied the availability of electron donor (lactose) and electron acceptor (O₂). Reducing electron acceptor shifted the consortium in a predictable manner toward the lactose positive, organic acid producer (L⁺R^-). The cost to produce pyruvate increased as the availability of O₂ decreased and the consortium required a larger base of strain L⁺R^- to support the consortium. The predictable changes in cross fed metabolite costs high lights the potential to design and fine tune cross feeding consortia for enhanced biocatalytic function. In summary, synthetic consortia with engineered interactions can decode complex natural systems by reducing the number of unknowns and enhancing the ability to design, engineer, and control consortia with desired properties.