| Abstract |
In cross-feeding interactions, different strains or microbial species exchange usable products arising from the metabolism of the primal nutritional source. Cross-feeding interactions have been observed in several ecosystems. Furthermore, long-term evolution experiments on the bacterium Escherichia coli growing in a simple, single limited resource have shown the emergence of several subtypes with different phenotypes in the population maintained by cross-feeding interactions. Polymorphism and metabolic interactions can play an important role in the evolution of populations as they dynamically shape the fitness landscape allowing new phenotypes to evolve. Cooperative strategies in the form of cross-feeding may lead a population to better adaptation and more efficient exploitation of a given environment.
The availability of high-throughput data allows the mapping of cellular metabolism into a genome-scale metabolic network, which considers the set of almost all biochemical transformations that take place within the cell. Thus far, in the metabolic simulations, which describe bacterial growth based on the genome-scale metabolic reconstructions, cells are genetically identical.
In an attempt to improve our understanding of the evolution of metabolic diversity in simple environments and the mechanisms supporting cooperative behaviors, this work goes a step further from single-cell models; it develops the first genome-scale metabolic model capable of simulating a competitive life within cell communities, where different individuals co-grow, sense, shape and respond to a common, dynamic environment. The model aims to reveal communities composed of self-centered strains that exhibit group benefit because of their capability to better utilize the available resources than single strains. As proved analytically in this work, competition for the primal source alone in a simple and spatially homogeneous environment cannot lead a heterogeneous population to group benefit, supporting the hypothesis that other sources of heterogeneity such as by-production might play a critical role in growth efficiency. In addition to the metabolic model, a graph representation (diversity graph) is developed in order to reflect the mapping from the genetic to the metabolic variability with respect to by-production. The graph allows the efficient determination of strain communities with the potential to differently shape the environment and develop cross-feeding interactions. Several graph-theoretic measures are applied in order to reveal biologically insightful properties, to characterize the diversity graphs and to allow the direct comparison of the overall metabolic behavior of different mutants with respect to by-production under different growth conditions. The bacterium E. coli is used as a case study. Metabolic gene knockouts generate the pool of mutants among which potential cross-feeding interactions are examined.
The graph analysis suggests that the two acting processes towards stabilizing either the monomorphic (i.e. populations with a single mutant) or the polymorphic state are antagonistic and that among all potentially interacting communities probably only those consisting of mutants that are specifically adapted to the given environment are likely to evolve. It is observed that the metabolic capabilities of the mutants with respect to by-production are highly redundant. This property allows the efficient identification of all the potential interacting communities represented as cliques in the graphs. The growths of these communities are simulated in several growth conditions utilizing the developed genome-scale multi-competitor metabolic model. The growth simulations show that metabolic interactions are indispensable within strain communities in order to perform efficiently under conditions of resource competition. Strain communities can be beneficial even if not all of their pair-wise relations correspond to cross-feeding, which demonstrates the importance of exploring group-wise metabolic variability. Furthermore, it is observed that in several efficient cases co-growth provides immediate benefits to the competitors by increasing their growth rate. The existence of interacting heterogeneous populations capable of better exploiting a given growth medium than monocultures indicates that in some growth conditions, the involved metabolic pathways are coupled in a way that a single optimal mutant is incapable of fully utilizing the environment.
As complexity increases and as environments become more complex than the homogeneous medium of a single-limiting resource that was explored in this study, diversity might prove far more beneficial for the systems involved. The method presented in this work has many implications for research on the ecology of increasingly complex microbial communities in natural and engineered environments.
|
Computational study of the metabolic diversityof the bacterium Escherichia coli : from single cells to cell communities and efficient systems
