Negative interactions and virulence differences drive the dynamics in multispecies bacterial infections

Bacterial infections are often polymicrobial, leading to intricate pathogen-pathogen and pathogen-host interactions. There is increasing interest in studying the molecular basis of pathogen interactions and how such mechanisms impact host morbidity. However, much less is known about the ecological dynamics between pathogens and how they affect virulence and host survival. Here we address these open issues by co-infecting larvae of the insect model host Galleria mellonella with one, two, three or four bacterial species, all of which are opportunistic human pathogens. We found that host mortality was always determined by the most virulent species regardless of the number of species and pathogen combinations injected. In certain combinations, the more virulent pathogen simply outgrew the less virulent pathogen. In other combinations, we found evidence for negative interactions between pathogens inside the host, whereby the more virulent pathogen typically won a competition. Taken together, our findings reveal positive associations between a pathogen’s growth inside the host, its competitiveness towards other pathogens, and its virulence. Beyond being generalizable across species combinations, our findings predict that treatments against polymicrobial infections should target the most virulent species to reduce host morbidity, a prediction we validated experimentally. Importance There is increasing evidence that polymicrobial infections are common and that host morbidity is impacted by interactions between the co-infecting pathogens. In this study, we infected the larvae of the insect host Galleria mellonella with four opportunistic human pathogens as mono or multispecies bacterial infections in all possible combinations and followed host survival and pathogen interactions over time. We discovered that species followed the exact same rank order in terms of their virulence, their growth inside the host, and their competitive ability against co-infecting pathogens. As consequence of this consistent rank order, we found that the most virulent species determines host survival dynamics in mixed infections. Beyond revealing a generalizable predictive pattern for virulence outcomes, our findings also predict that treatments should target the most virulent species in polymicrobial infections, a prediction we validated experimentally.