Background:

Microbes often exist in complex communities containing a variety of species in a spatially defined network. However, extrapolating information from laboratory measurements on pure monocultures to predict the behavior of complex communities is difficult. This is due to a variety of factors, including metabolic exchange between different members of a community – the production by one microbe of a bioactive molecule that can negatively or positively regulate the growth of another. Examples of interspecies communication include shared quorum sensing autoinducers [Federle 2003] and metabolic cross-feeding that regulates pathogenesis of invasive microbes in the intestinal microbiota [Pacheco 2012, Ng 2013]. As these interactions are emergent properties of communities of microbes, the measurement and modeling of microbe-microbe interactions and the effects of these interactions on growth dynamics of multiple members of a community is key to understanding, and potentially manipulating, microbial communities. These two papers present new methods, models, and concepts governing microbial community dynamics.

 

Technical approach:

Both papers utilize high-throughput microscopic examination of fluorescent strains of bacteria, followed by automated image analysis, to measure cell numbers, density, or size. Hsu et al develop a microdroplet based method (MINI-Drop), where small numbers of fluorescently labeled cells (expressing different colors for different strains of interest) are encapsulated in droplets using an oil-and-culture mixture method, resulting in a large number of droplets randomly containing one or multiple cells. These “micro-communities” are grown independently, and a computer vision method is used to count cells within each droplet after a defined growth period. This enables counting of each constituent strain in the final community. Comparison of droplets containing only a single strain versus mixed communities enables measurement of the growth effect of a partner strain on a particular microbe. The authors then proceed through a number of two- and three-strain experiments in multiple different growth conditions predicted to result in positive or negative interactions (cooperation or competition).

Dal Co et al focus on a single model community, using two auxotrophic E. coli strains (unable to synthesize proline and tryptophan) that can cross-feed to support cooperative growth. The authors use a microfluidics setup to encapsulate a mixed community within a single chamber, with fresh medium flowed in at a constant rate. The entire chamber was imaged over an extended growth period, and each individual cell was tracked computationally to measure its growth rate. Because all cells in an area could be imaged, the authors were able to measure the growth rate of all cells as well as the composition of its neighbors.

 

Key findings:

• MINI-Drop enables measurements of multiple community dynamics and validates a probabilistic cell growth model.
  Hsu et al measure interactions in multiple model communities showing a variety of interaction structures. MINI-Drop enables measurement of interactions in both cooperative and competitive settings, and reveals higher-order interactions in a three-member community that only emerge when all three members of the community are present, not predictable from the behavior of any pairwise combination of strains. Finally, they model cell growth through a discrete-time Markov model and show that for both positive and negative interactions the Markov model can predict growth, suggesting probabilistic models are capable of accurately predicting community growth states.
• Individual cells in a microbially community are heavily influenced by the composition of their immediate spatial neighbors, and this interaction is driven by nutrient uptake.
  Dal Co et al reveal the growth rates of individual cells are highly dependent on the fraction of their complementary partner (producing the required nutrient) in a small distance – depending on the strain, 3 to 12 µm, which is only a few cell lengths. By modeling the cell growth based on known biochemical parameters, the authors also show that the dominant factor driving this short interaction range is the uptake of nutrients by other cells. That is, essentially, because cells are actively taking up nutrients, even freely diffusible molecules in solution are removed by the presence of cells and therefore producers of essential nutrients must be close to a cell for it to grow best.

 

Importance:

These two preprints describe important models and techniques for measuring cell growth in mixed communities. Hsu et al develop MINI-Drop, a practical method to measure, with high numbers of replicates, communities that exhibit complex interaction networks including emergent properties. Dal Co et al demonstrate that the physical location of an individual cell within a community, and the composition of its immediate neighbors, can be critical in determining cell fate. Together, these preprints lay the ground for further high-throughput investigation of complex community dynamics.

 

Moving forward:

• Both approaches are currently limited by the requirement for fluorescently labeled bacteria. It will be interesting to see how advances in computer vision and automated cell tracking will enable tracking of a broader range of cells, including genetically intractable species that are difficult to label, as well as representing a broader range of potential metabolic interactions and requirements.
• Both techniques rely on pre-growth of the individual strains in isolation, followed by mixing and measurement. In complex communities, it is possible to imagine that molecules required for metabolic exchange (e.g. nutrient transporters) may depend on environmental signals present only in the mixed community or in particular environments. How do the expression kinetics of these molecules change the models the authors describe? This may be of particular interest in situations of ecological perturbation, such as antibiotic recovery or probiotic consumption in the intestinal microbiota.

 

References:

• Federle M.J. et al. Interspecies Communication in Bacteria. J Clin Invest (2003) 112(9)
• Pacheco A.R. et al. Fucose sensing regulates bacterial intestinal colonization. Nature (2012) 492(7427)
• Ng K.M. et al. Microbiota-liberated host sugars facilitate post-antibiotic expansion of enteric pathogens. Nature (2013) 502(7569)

 

doi: https://doi.org/10.1242/prelights.8120

(No Ratings Yet)