A rapidly expanding body of research has revealed that bacteria can engage in various social interactions. In-vitro experiments have shown that when cells cooperate through the secretion of "public good" molecules, non-producer individuals can sometimes free ride (i.e. "cheat") on molecules produced by others and increase in frequency (West et al., 2007). The fact that some of these shared molecules are important virulence factors led to the hypothesis that social interactions could be important during infections, and that social dynamics could be exploited to steer pathogen populations towards lower virulence. However, there is a pressing need to improve our understanding of how pathogens really interact within hosts if we want to successfully develop these approaches.

When I started my PhD almost four years ago, our research group was in the middle of developing two exciting lines of research: using C. elegans as a model host to study the evolution of virulence in P. aeruginosa (Granato et al., 2018), and tracking the expression of social traits using fluorescent reporters (Weigert and Kümmerli, 2017). My first project was a combination of these two approaches and aimed to study bacterial social interactions in living hosts. The journey to develop a new method is often very challenging, and includes the excitement of the first results, endless rounds of preliminary testing, strange results happening for no apparent reason, the frustration of feeling that you are not making progress. Nevertheless, in the end I managed to optimise the protocol, and a new automated microscope at the local microscopy facility made data acquisition much easier.

We used our newly developed live imaging system to (i) measure the expression of secreted virulence factors, which can be shared as public goods between bacterial cells; (ii) track host colonisation at different stages of infection, from exposure to stable colonisation of the host gut; and (iii) analyze competitions between virulence-factor-producing and non-producing strains in the host. We found that shareable siderophores and quorum sensing molecules are expressed during the infection, affect host colonisation, and benefit non-producer strains when mixed with producers. However, we also found that non-producers had limited cheating success within the host, and were unable to outcompete producers. 

Our results have implications for both the understanding of bacterial social interactions within the hosts, and therapeutic approaches that aim to manipulate social dynamics between strains for infection control. To further develop novel therapeutic approaches that seek to take advantage of cooperator-cheat dynamics to control infections, we highlight the importance of explicitly testing the impact of social behaviours within hosts. Characterizing social trait expression, temporal infection dynamics and physical interactions between genotypes within hosts are essential to fully understand whether social traits are exploitable in hosts.