Section: New Results

Beware batch culture: Seasonality and niche construction predicted to favor bacterial adaptive diversification

Participants: Charles Rocabert, Carole Knibbe, Guillaume Beslon

The evolution of stable bacterial cross-feeding interactions is often considered as the first step toward bacterial speciation in sympatry. It is thus important to study the conditions favoring the emergence and the stabilization of cross-feeding interactions in well-mixed environments. Experimental evolution in laboratory, where fast organisms are replicated for thousands of generations in controlled conditions, provides important insights on this question. Indeed cross-feeding is commonly observed in batch cultures or in chemostat. However, the reasons why cross-feeding interactions become stable and lead to monophyletic ecotypes remain unclear. Because laboratory experiments are a long and costly process, we explored this question by evolving digital organisms in artificial systems mimicking the conditions of wet experiments.

Models of digital evolution helped a lot to decipher the evolution of cross-feeding interactions. However, the evolution of real microorganisms implies the interaction of a wide range of biological structures and levels, while those models often include only two or three levels, limiting their ability to mimic real experiments. In this work, we developed a new multi-scale model of digital evolution, integrating a complex and realistic genotype-to-phenotype mapping (including a metabolic network) and a complex environment (that links the organism’s metabolic networks together, opening the possibility for cross-feeding). This model has been developed under the European project EvoEvo, and has been inspired by previous models developed by the Beagle team, and in the Theoretical Biology and Bioinformatics group at Utrecht University. By mimicking laboratory experiment setups and running simulations for tens of thousands of generations, we were able to recover ecological dynamics similar to those found in real experiments.

In batch culture, like in the Long Term Evolution Experiment (LTEE), it is accepted that the seasonality generated by the serial transfers triggers the maintenance of cross-feeding interactions on the long-term by favoring niche construction and specialization. In chemostat, cross-feeding interactions are observed and seem stable for a few hundreds of generations but the reasons of their stability remain unclear. Thanks to our model, we were able to observe stable cross-feeding interactions reproducing the same properties as those observed in the LTEE. We then showed that seasonal conditions found in batch cultures are essential for the maintenance of stable ecotypes on the long-term, since it produces conditions for niche construction and stable cross-feeding. In chemostat conditions, the absence of seasonality and competitive exclusion precludes any stabilization of emerging cross-feeding interactions. Finally, we proposed to consider a cross-feeding interaction to be stable only if interacting ecotypes undergo independent periodic selection events on the long-term. Stable cross-feeding interactions could then be considered as premises to speciation in sympatry.

This work is the result of an enriching collaboration between the Beagle team (Charles Rocabert, Carole Knibbe, Guillaume Beslon), and microbiologists from the TIMC-IMAG in Grenoble (Jessika Consuegra, Dominique Schneider). It has been published in the renowned journal PLoS Computational Biology in January 2017. This work is of interest for the fields of evolutionary biology, microbiology but also for computer science. Indeed our findings also suggest that digital evolution is a useful tool to study bacterial evolution, and that the use of models integrating a complex genotype-to-phenotype mapping and complex interactions between digital organisms and their environment is important to accurately study real biological systems thus appealing for further fruitful transdisciplinary collaborations.

Publication: [21].