Abstract
Numerous studies have considered infections as pairwise interactions between a single pathogen and its host, sometimes leading to an incomplete picture of infectious processes. In this work, we focused on more complex types of interactions that arise because infections are usually heterogeneous. More precisely, we have investigated two main issues: (I) how pathogen transmission is impacted by phenotypic heterogeneity which arises within the pathogen population during the infection, and (ii) how do pathogens interact with the bacterial community which is naturally associated to the host before infection? To assess these questions, we have been studying Xenorhabdus nematophila, an insect-killing bacterial pathogen which is transmitted by a nematode vector, Steinernema carpocapsae. One interesting feature of X. nematophila is that it produces different sub-populations during the course of an infection, each one having distinctive phenotypic features (e.g. one form produces antibiotics and is mobile, while the other does not produce antibiotics nor flagella). In this work, we first tried to identify the molecular mechanisms responsible for this diversification of phenotypes, and tested if phenotypic heterogeneity in X. nematophila has some adaptive value. We showed that some of these phenotypic forms were mutants, which seem to be under strong positive selection during infection. We also showed, however, that these mutants impair nematodes reproduction, which in turn reduces transmission. Therefore, the dynamics of phenotypic heterogeneity in X. nematophila seems to be determined by contradictory short-term and long-term selective pressures. A second interesting feature of X. nematophila is that it produces a lot of antimicrobial compounds which should allow it to dominate the bacterial community inside the insect it has killed. This can be key to ensure the re-association of X. nematophila with its nematode vector inside the insect cadaver. We investigated the bacterial composition of the microbial communities present in insects cadavers after infection by X. nematophila. We found that despite the numerous antibiotics it is able to secrete, X. nematophila is far from dominating microbial community after host death. It rather cohabits with microorganisms from the microbiota of both the insect host and the nematode vector. This raises numerous questions about the impact of these other microorganisms on Xenorhabdus-Steinernema interactions, and therefore on their potential influence on how this mutualistic association has evolved.