The Ecology and Evolution of Bacterial Social Interactions and How They Shape Microbial Communities and Interactions with Hosts

Microbes are social organisms. Their social repertoire ranges from mutually beneficial cooperative interactions with conspecifics, over to engaging in warfare with heterospecific competitors. Interactions among microbes are crucial for how microbial communities assemble and hence underlie many microbial-derived “macroscopic” effects, such as bioremediation, digestive functions of animal gut microbiomes, or nutrient cycling in soil ecosystems. Often, social interactions among microbesrequire the secretion of molecules into the extracellular environment, where these molecules might either aid or harm other microbes. One of the most widely studied microbial social trait is the production of siderophores, molecules secreted by bacteria to obtain otherwise inaccessible iron. Siderophores are at the interface of bacterial cooperation and competition: they can be shared among cells that have the matching receptor for uptake, but they can also be deployed to lock away iron from cells that do not possess a matching receptor. Siderophore basedcooperation has overwhelmingly been studied in the context of cheating (phenotypes whose bearers do not produce siderophores, but can import those produced by others), and siderophore-based competition has only recently been discovered. Thus, there are relevant knowledge gaps on the ecology and evolution of this trait. In my thesis, I study the relative importance of siderophore-based cooperation and competition, and how this trait evolves in different environments. Specifically, I performed two experimental studies on these topics, and one conceptual study on the evolution of microbial interactions when in host-associated communities. For my first experiment (Chapter 3), I used the model bacterial system Pseudomonas aeruginosa and its siderophore pyoverdine, and experimentally evolved populations of this microbe in nine environments that are expected to exert different selection pressures on siderophore-based cooperation. I manipulated either the amount of iron available or the viscosity of the growth medium, or both. Lower iron availability requires a high investment into pyoverdine production, making cooperation costly, whereas more available iron requires fewer siderophores, making cooperation cheaper. Environmental viscosity dictates the diffusion of cells and molecules, so that in well-mixed environments every individual can potentially interact with any another, whereas in viscous environments dispersal is hampered and individuals are more likely to interact with relatives. I found that cheating phenotypes can only invade and disrupt siderophore-based cooperation in iron-limited and well-mixed environments (i.e., where cooperation is costly and relatedness low). Conversely, I found that higher levels of pyoverdine-based cooperation are favoured in high environmental viscosity and/or high iron availabilities (i.e., respectively where the interaction rate between related individuals is highest or where cooperation is less costly). Moreover, I found that mutations in quorum sensing loci largely explain higher pyoverdine-based cooperation phenotypes. The results from this experiment suggest that: (i) the environmental conditions that favour the evolution of cooperation are wider than previously thought, and (ii) quorum-sensing communication is involved in the control of pyoverdine-based cooperation, a previously unknown regulatory link between two crucial social traits in P. aeruginosa. In my second experiment (Chapter 4), I investigated the role of pyoverdine-based competition for bacterial interspecific interactions. Pyoverdines vary in their chemical structure, and most strains produce only one type of pyoverdine and the corresponding receptors. Co-existing strains that produce different pyoverdine-receptor combinations will compete for the little iron available in the environment. Therefore, a given strain should outcompete another one if its pyoverdines prevent the competitor from accessing iron. Conversely, a given strain should be outcompeted if the competitor can use its iron-loaded siderophores. I tested these predictions using a combination of pyoverdine cross-feeding and invasion-from-rare experiments on a collection of natural Pseudomonas isolates and the aforementioned Pseudomonas aeruginosa as a focal strain. The results from these experiments support the hypothesis that invasion success is high if a strain can use pyoverdines produced by others, and is low if a strain cannot use another’s pyoverdines. These results demonstrate the importance of siderophores for interspecific bacterial competition and invasion success, and highlight their role in microbial community assembly. In the last chapter of my thesis (chapter 5), I zoom out, and look at how microbial cooperation and competition in host-associated microbial communities affect their animal hosts. In this review article, I introduce the similarities between social interactions among free-living and host-associated microbes, and describe the effects of the latter on host fitness. I identify how some hosts have evolved to target the social interactions of the microbes they harbour, and how this differs between hosts harbouring species-poor versus species-rich microbiotas. Then, I argue that the species-richness of a microbiota is a crucial predictor of the type of social interactions that occur within the community, with cooperation prevailing in simple microbiotas and competition in complex ones. I further argue that in species-poor microbiotas, hosts and microbes can more easily co-evolve, which might trigger the evolution of dependencies. In species-rich microbiotas, this is unlikely to happen because microbes are constantly in competition with one another. Social interactions are an important component of the life of a microbe. In this work, I have highlighted how one single type of molecule can mediate both intraspecific microbial cooperation and interspecific microbial competition. In a larger context, I have discussed how these social interactions can mediate host-symbiont relationships. In conclusion, to fully understand the evolution of a microbial social trait, an integrative view is required, where its effects on clonemates, other members of the community and (if applicable) the host is necessary