The genetic basis for medium‐dependent bacteriophage infection

Abstract

While broad spectrum antibiotics have served a major role in combatting bacterial pathogens over the past decades, we are currently witnessing the rise of antibiotic‐resistant pathogenic strains. Moreover, the harmful effects of broad spectrum antibiotics on the natural human microbiome have been demonstrated in numerous cases. For these reasons, bacteriophages have been recently gaining attention as promising candidates for treatment of bacterial‐related diseases. Bacteriophages are natural viruses that infect and often kill specific bacterial strains; it is the phages’ high accuracy and specificity, alongside their natural origin, that highlights their value as an alternative treatment for diseases with bacterial involvement.However, a major challenge in screening for bacteriophages relevant for treatment of human diseases is the phages’ medium‐dependent infection trait. Several factors have been suggested to take a role in explaining this discrepancy, including the mechanism of infection, the spatial structure of the phage capsid, the biochemical characteristics of the bacterial cell wall, and other phage life‐cycle features. However, a comprehensive model for characterizing the media‐range for phage infection is still missing. This uncertainty often translates into multiple laborious screening rounds, since common practice in vitro assays are typically based on either liquid or solid medium infection, whereas the state of matter relevant for in vivo application is often hard to determine; thus, one must screen the same phages using multiple assays.In attempt to lay the groundwork for such a model, in this work we set out to determine the genetic‐basis for medium‐dependent infection of pathogenic Klebsiella pneumoniae phages. To this end we obtained 285 K. pneumonia strains, that span across the entire domain of known K. pneumonia genotypes. Subsequently we obtained 50 phages that were able to propagate on at least some of these hosts. For each phage‐host pair we determined whether infection occurs in two different assays: in liquid culture, and on solid agar plates. As expected, the results show that liquid medium enables more phage infections compared to solid; this is traditionally associated with the higher mobility enabled by the liquid. However, cases where a phage was able to infect its host in solid, but not in liquid, were observed as well. Finally, by analyzing genomic differences between the phages in the experiment we were able to discern several genetic features that correlate with the observed phenotype. Further exploration of these genetic features are the basis for a constructing a model for characterizing the media‐range for phage infection.