Engineering microbial consortia for controllable outputs

Abstract

Much research has been invested into engineering microorganisms to perform desired biotransformations; nonetheless, these efforts frequently fall short of expected results due to the unforeseen effects of biofeedback regulation and functional incompatibility. In nature, metabolic function is compartmentalized into diverse organisms assembled into robust consortia, in which the division of labor is thought to lead to increased community efficiency and productivity. Here we consider whether and how consortia can be designed to perform bioprocesses of interest beyond the metabolic flexibility limitations of a single organism. Advances in post-genomic analysis of microbial consortia and application of high-resolution global measurements now offer the promise of systems-level understanding of how microbial consortia adapt to changes in environmental variables and inputs of carbon and energy. We argue that, when combined with appropriate modeling frameworks, systems-level knowledge can markedly improve our ability to predict the fate and functioning of consortia. Here we articulate our collective perspective on the current and future state of microbial community engineering and control while placing specific emphasis on ecological principles that promote control over community function and emergent properties.

Figure 1

Example of how competition and conflict between different metabolic processes could promote the division of metabolic labor. In this hypothetical pathway, an enzyme (ES) transforms a substrate (S) into an intermediate (I). A second enzyme (EI) then transforms the intermediate into a product (P). These two enzymes may compete for the same intracellular resource, such as cellular space, co-factors for enzyme activity or building blocks for biosynthesis. If both enzymes are contained within the same cell and competition is asymmetric (for example, the enzyme for the first step has preference over the enzyme for the second step), then this would lead to the accumulation of the intermediate. If each enzyme is contained within different cells, then competition is lost, thus preventing one enzyme from outcompeting the other and reducing the accumulation of the intermediate.

Figure 2

(a) Interactions and community behaviors can be manipulated by modifying the abiotic environment (Momeni et al., 2013). Vertical cross sections of communities of two non-mating S. cerevisiae strains, one requiring lysine and releasing adenine (red) and one requiring adenine and releasing lysine (green). The two strains engaged in competition (top) if both adenine and lysine are exogenously provided by the agarose medium, commensalism if, for example, lysine but not adenine is supplied (middle), and cooperation if neither lysine nor adenine is supplied (bottom). Different interactions lead to different spatial organization in experiments and simulations (scale bar, 100 μm). (b) Community behaviors can be manipulated by modifying the spatial organization of cells (Kim et al., 2011). A sufficiently thick ‘shell' of R. metallidurans can be deposited around a ‘core fiber' of S. chlorophenolicum, using microfluidics. R. metallidurans protects S. chlorophenolicum by reducing the toxic Hg(II) to Hg(0), and S. chlorophenolicum can in turn detoxify PCP.

Figure 3

Conceptual design of a self-assembling autonomous, synthetic microbial consortium built through photoautotroph–heterotroph interactive partnership. Genetically tractable photoautotrophic organisms, such as cyanobacteria, can sustain and drive an engineered consortium through metabolic exchange reactions (black arrows), which include photosynthetic production of O₂ and organic C; these serve as respiration substrates for the heterotrophic module(s) to generate CO₂. This creates opportunities for metabolic engineering using interdependent modules, whose functional outputs and interactions can be programed for additional control via synthetic regulatory circuitry (blue arrows).