Bacteriophages have two well characterized life cycles; lytic and lysogenic. Most phages are capable of undergoing both, these are referred to as temperate. The factors that influence a temperate phage’s decision of whether to enter the lytic or lysogenic cycle are hotly debated; however, it is clear that the decision has an enormous impact on bacterial communities.

The Hynes’ lab has been working on ways to bias the phage lysis/lysogeny decision to better understand the underlying mechanism. Dormant phages are canonically induced out of the bacterial genome in response to the activation of the SOS pathway for DNA repair. This is often metaphorically likened to phages ‘abandoning a sinking ship’. Several DNA-damaging antibiotics are known to induce phages. We found combining temperate phages with SOS-inducing antibiotics improved their effectiveness as bactericidal agents by inducing any formed lysogens. Curiously, we found other antibiotics that decreased the frequency of lysogeny by biasing the initial lysis/lysogeny decisions. These recent findings support our hypothesis that this decision is far more complex than simply the activation of the SOS response by DNA damage.

To explore this, we have conducted high-throughput screening of 3921 bioactive compounds exposed to wild type E. coli and E. coli carrying temperate phage, HK97. ‘Hits’ are classified as differences in cell density from phage-driven bacterial lysis in the presence of the compound between bacteria with and without dormant phage. We found 37 compounds causing phage mediated bacterial killing. A few compounds were randomly selected and verified in lab through growth curve analysis and measuring end point phage concentration. Verified ‘hits’ included expected fluroquinolones and beta-lactams, known activators of the SOS response, as well as unexpected compounds such as Prozac, a commonly prescribed SSRI, and berberine, a plant alkaloid used in Aryuvedic medicine. These commonly consumed compounds may be resulting in phage-driven effects through biasing the lysis/lysogeny decision.

We are currently repeating the HTS using a range of concentrations as well as in a bacterial host unable to activate the SOS response, to determine which interactions depend upon that bacterial pathway and which are operating independently. To identify mechanisms involved in non-SOS activating compounds, we will expose phage + compound to an E. coli mutant library and identify which host pathways are needed to generate observed effects. This will enable possible identification of new bacterial stress pathways.

Our screening to date has been able to identify phage mediated killing with exposure to novel, commonly consumed, compounds. This is the first systematic approach to identifying compounds influencing the lysis/lysogeny decision and will likely reveal many new mechanisms by which phages can guide their decision-making behaviours. These, in turn, can be used to exploit phages and manipulate bacterial populations to improve the therapeutic application of phages.