Abstract
The growing threat of antibiotic resistant bacteria has prompted a reconsideration of phage therapy, or the use of bacteriophage viruses to treat patient infections. This century old practice still holds promise as one of the few alternatives to treat drug-resistant infections. While positive patient outcomes are frequently reported, treatment failures still regularly occur with little explanation of the possible reasons for failure. Current practice applies phages with little investigation into phage efficacy other than knowing the phage can infect the target bacteria. Conditions inside patients, however, present novel challenges for phage replication and may be contributing to poor treatment outcomes. This dissertation addresses this knowledge gap by investigating possible in vivo conditions that may be impacting the efficacy of phage therapy. Using in vitro simulation of two key conditions, prior antibiotics use and nutrient limitation, we investigate how these conditions may be affecting phage replication through inhibition the bacterial hosts upon which phages rely.
We first focus on prior use of antibiotics, since current standards of care dictate phage therapy is only considered after failure of traditional drug treatment. We investigate the effects of sub-lethal two different antibiotics on the growth of E. coli phage T7. Observing both kanamycin and chloramphenicol inhibit the growth of T7, we next investigate if this phage can be evolved to better tolerate these conditions. Comparing T7 evolved on cells exposed to one of these two drugs, as well as no-drug cells, we see fitness gains for all evolved phage lines compared to the ancestor phage. Further analysis found fitness gains were not specific to the environment of evolution (drug), but found adaption to generic conditions such as cells used, media, etc., conferred a fitness advantage in drug-treated cells.
Next we compare three common E. coli phages, T4, T5, and T7, on their ability to grow on nutrient-deprived cells. We focus specifically on the limitation of iron, as sequestration of this vital growth factor is a common biological strategy to combat pathogenic bacteria. We limit iron via addition of a chemical chelator, 2-2’-dipyridl (DIP) to LB broth and allow cells to grow until cessation of exponential growth was achieved (lag-phase). Overnight growth in three different concentrations of DIP results in differential final cell titer coordinating with the volume of DIP added. As a comparison we also include untreated overnight (stationary-phase) cells. While T4 and T5 are unable to grow in all concentrations of DIP as well as stationary phase cells, T7 experiences increases in plaque-forming units (PFUs) of five orders of magnitude in the lowest and highest DIP concentrations, and grows moderately well in untreated overnight cells. Interestingly, in the intermediate DIP concentration, T7 has slightly negative growth, exhibiting unexpected bimodality. The adsorption constant, k, for all phages is decreased in all overnight cells, but is too high to explain the lack of growth observed in T4 and T5. Further analysis of attached versus free PFUs present in the media over two hours suggests suspension of replication after infection may be occurring for T4, while attachment issues paired with abortive infection may explain the lack of growth for T5. Although different mechanisms may be at play, T4 and T5’s lack of growth on slow/non-growing cells makes them poor candidates for phage therapy targeting bacteria in these growth states. Of these three T7 hold the most therapeutic potential.
Finally, we evaluate a novel invertebrate model for studying phage therapy- the tobacco hornworm (Manduca sexta). Animal models are essential for determining if principles established in vitro translate in vivo. Using this model, we assess the efficacy of three different phages in the treatment of Pseudomonas aeruginosa infections. All phage treatments signifi-cantly improve survival rates compared to bacteria-only controls, although no differences in efficacy are observed between phages. Additionally, phage treatment does not significantly af-fect worm weight, either at day 4 or at pupation.
In comparison with established invertebrate models, such as waxworms, we find hornworms provide a useful and informative system for studying phage therapy. Together, these findings highlight the importance of considering how in vivo conditions may be affecting phage efficacy and incorporating these conditions into in vitro studies.