Cooperation in microbial communities and their biotechnological applications

Abstract

Microbial communities are increasingly utilized in biotechnology. Efficiency and productivity in many of these applications depends on the presence of cooperative interactions between members of the community. Two key processes underlying these interactions are the production of public goods and metabolic cross-feeding, which can be understood in the general framework of ecological and evolutionary (eco-evo) dynamics. In this review, we illustrate the relevance of cooperative interactions in microbial biotechnological processes, discuss their mechanistic origins and analyse their evolutionary resilience. Cooperative behaviours can be damaged by the emergence of 'cheating' cells that benefit from the cooperative interactions but do not contribute to them. Despite this, cooperative interactions can be stabilized by spatial segregation, by the presence of feedbacks between the evolutionary dynamics and the ecology of the community, by the role of regulatory systems coupled to the environmental conditions and by the action of horizontal gene transfer. Cooperative interactions enrich microbial communities with a higher degree of robustness against environmental stress and can facilitate the evolution of more complex traits. Therefore, the evolutionary resilience of microbial communities and their ability to constraint detrimental mutants should be considered to design robust biotechnological applications.

Figure 1

A. Interactions based on shared public goods. Some cells (cooperators, shown in black edge) produce an enzyme required to split a substrate into digestible products. Other cells (cheats, shown in grey), do not produce the enzyme but take advantage of the public goods produced by the others. B. Interactions based on cross‐feeding. Some cells in the community excrete metabolites that can be taken up by other cells giving rise to a web of interactions.

Figure 2

A. Metabolic interactions that can take place in a population. Cells can exchange metabolites that are required to support each other's growth in a mutualistic interaction (left). One of the cells can use a metabolite excreted by another cell, favouring in this way the metabolism of the producer through the pathways leading to the excretion (centre). When the metabolites excreted have an inhibitory effect on the producer (e.g. because they lead to thermodynamic equilibrium), the relationship with a degrader cell of the inhibitory metabolite is mutually beneficial and known as syntrophy (right). B. Dynamic modelling of the evolution of Flux Balance Analysis models. Cells can be modelled as metabolic networks exchanging metabolites with other cells in the population. In this abstraction each cell is represented by a FBA model. These models can replicate over time and also evolve, producing populations composed by models with different constrains for uptake and secretion of metabolites. C. Dynamic analysis of model genealogy. The frequency of each model in the population changes over time being the darkest bars the most abundant models. Due to mutations, new models arise and they are represented as new branches in the phylogeny. Plot redrawn from Großkopf et al. (2016).

Figure 3

Mechanisms to Preserve Cooperation in Cellular Communities. A. A structured environment can facilitate cooperation. The figure shows the growth of fluorescently labelled colonies (cooperators in red, cheaters in green) of S. cerevisiae (Figure from (Van Dyken et al., 2013)). Cooperative cells produce invertase that breaks down sucrose into digestible glucose and fructose. Non‐producers cells (cheaters) have a fitness advantage because they do not produce invertase but can access glucose. In unstructured environment (liquid culture) cooperators decline. However, in a spatial environment (obtained by spotting a droplet of mixed cooperator/cheater cultures onto solid medium) cooperators can spread over cheaters. The diffusion of cells leads to the formation of discrete sectors – cooperator sectors are more productive than cheater sectors and expand radially faster. B. Eco‐evo dynamics can preserve cooperation in communities of S. cerevisiae (redrawn from (Sanchez and Gore, 2013)). Red circles represent cooperative cells (invertase producers), green circles represent cheaters (non‐producers). Below a certain cooperator density, there is little glucose available. Cooperative cells grow at a slow rate on the little amount of glucose they can retain, while cheater cells grow more slowly (it is crucial that cooperators have preferential access to the glucose). Above a certain cooperator density, both cooperators and cheaters grow at a fast rate because of the large pool of available glucose, but cheaters grow faster as they do not have the burden of producing invertase. Such density‐dependent selection favours cooperators at low densities and cheaters at high densities, which leads to the stable coexistence of cooperative and cheating yeast cells. C. Regulation of public good production can preserve cooperation in a meta‐population model in which the population is transiently divided in sub‐populations (figure from (Cavaliere and Poyatos, 2013)). In‐silico simulations present two possible successful types of regulation against cheaters: positive plasticity (top row) in which cooperators constraint cheaters by stopping the production of public good when cheaters appear (a) and fully restarting only when cheaters have disappeared (b) and negative plasticity (bottom row) in which cooperators produce permanently low amounts of public good which helps controlling cheaters invasion (c). Thick arrows denote the cellular decision to produce (P) or not produce (nP) the public good. The success of the regulation is coupled to the heterogeneity (variance) of the subpopulations, i.e., positive plasticity transiently modifies the variance while negative plasticity keeps a relatively constant heterogeneity (variance shown in (d) correspond to trajectories (b) and (c) respectively).

Figure 4

A. Division of labour in microbial populations. Colonies of Pseudomonas fluorescens Pf0‐1 are composed by cells with two different morphologies known as mucoid and dry that can evolve from each other due to a single mutation (left picture). Colonies composed by a mixture of the two phenotypes expand faster allowing cells to colonize larger regions in shorter periods of time compared to colonies composed by each of the individual phenotypes. The two morphotypes occupy different regions of the colony as shown when labelled with fluorescent reporters (centre). Dry cells (in red) exhibit a radial distribution growing on top of the mucoid (in green). Confocal microscopy reveals that the edge of the colony (right picture) displays a distinct spatial organization in which mucoid cells form a thin strip at the very edge. The differentiation and spatial segregation allows the distribution of labour in the population: Mucoid cells produce a lubricant polymer at the edge, whereas dry cells sit behind and push both of them along. The cooperation of these two phenotypes results in a fast growing colony. Pictures have been reproduced from Kim et al. (2016). B. Engineered populations can improve bioprocesses. Two strains are combined to carry out the synthesis of a product of interest (red pentagons) that cannot be produced using each of the strains individually. The process involves that one strain produces an intermediate (the yellow pentagon) that is used by the other to synthesize the final product. If the two strains compete for the same resources (e.g. carbon source shown by the blue hexagon; left panel) the population with the lower fitness under those conditions will eventually collapse. However, when the two populations are engineered so that one grows at the expenses of the other (e.g. through cross‐feed or syntrophy shown by the purple triangle), the two populations cooperate (centre panel) and the synthesis of the product of interest takes place for a longer period of time resulting in higher yields (right panel). Panels inspired by Zhou et al. (2015).