The canonical bacteriophage is obligately lytic: the virus infects a bacterium and hijacks cell functions to produce large numbers of new viral particles which burst from the cell. Viruses of this type are well-studied, but there exist a wide range of coexisting virus lifestyles that are less well understood. Temperate viruses exhibit both a lytic cycle and a latent (lysogenic) cycle, in which viral genomes are integrated into the bacterial host. Meanwhile, chronic (persistent) viruses use cell functions to produce more viruses without killing the cell; chronic viruses may also exhibit a latent stage in addition to the productive stage. Here, we use a mathematical model and experimental data from Pseudomonas aeruginosa to study the ecology of these competing viral strategies. Understanding the ecology of P. aeruginosa and its phages is critical to controlling bacterial infections in humans, especially when bacteria are resistant to antibiotics. Our analysis reveals the conditions under which certain antibiotics can be effectively used in synergy with naturally occurring phages to treat P. aeruginosa infections in immunocompromised humans.
Results/Conclusions
Using our model of virus competition for bacterial hosts, we find three distinct results. First, low lysogen frequencies provide competitive advantages for both virus types; however, chronic viruses maximize steady-state density by eliminating lysogeny entirely, while temperate viruses exhibit a non-zero ‘sweet spot' lysogen frequency. From the perspective of infection control, non-zero lysogen frequency supplies a population of bacteria that can be easily killed by inducing the latent viruses. Next, viral steady-state density maximization leads to coexistence of temperate and chronic viruses, explaining the presence of multiple viral strategies in natural environments, including humans. Finally, the natural presence of chronic viruses in humans inhibits control of bacterial infections because chronic viruses exclude secondary infection by lytic viruses but do not kill host bacteria. In fact, our model reveals that increasing the antibiotic dosing frequency counterintuitively increases the bacterial load when a large fraction of the bacteria develop antibiotic-resistance.