Vaccine design, dynamics, and evolution in bacteriophage

Vaccines are a cornerstone of public heath and are our most effective method of preventing infectious disease. Live attenuated vaccines (LAVs) provide potential applications for improving vaccine efficacy, safety, and delivery. However, LAVs as they exist today are imperfect. They tend to be genetically unstable and occasionally revert to wild-type virulence. They require direct application to each individual, which is inefficient. In this dissertation, I use bacteriophage to explore the application of engineered attenuated viruses as transmissible vaccines to improve vaccine delivery, and examine the evolutionary stability of a rationally designed viral attenuation strategy.

First, I develop and anaylyze an empirical system that uses a non-lethal therapeutic virus (phage f1) to protect a host (E. coli) cell population from a lethal virus (phage Qβ). Computer simulations and experimental assays are used to test the feasibility and predictability of the system. Dynamics for two scenarios are tested, differing in whether the therapeutic virus is introduced before or after the lethal virus. In general, observed dynamics are consistent with those predicted by the simulation model. When the therapeutic virus is introduced first, it spreads infectiously without any appreciable change in host growth. When introduced after the lethal virus, host abundance is depressed at the time of therapeutic virus application and initial spread of the therapy is brief, with the subsequent rise in protected hosts coming primarily from reproduction of hosts already protected.

Next, I propose, model, and experimentally test an engineered two-component transmissible vaccine system in which phage M13 and a mobilizable gene therapy phagemid vector are used to spread through and genetically modify an E. coli population to resist the lethal phage T5. I show that computational simulations and genetic modification of the M13 helper phage improves the dynamic properties facilitating spread of the gene therapy vector. Overall, computational simulations agree with observed dynamics suggesting that such such models can be useful guides in designing optimal systems for transmissible vaccines.

Finally, I examine the efficacy of promoter knockout as a viable strategy for genetically stable viral attenuation in phage T7. Promoters for two highly expressed structural genes in T7 were knocked out either individually or together and attenuation and adaptation was quantified through fitness, DNA sequencing, and RNA sequencing. Initial knockout strains behaved broadly as expected; fitness was reduced and transcript abundance was knocked down for genes downstream of our promoter mutations. After hundreds of generations of adaptation, fitness and RNA expression increase modestly, but remain significantly under wild-type levels. DNA sequencing also reveals no common molecular mechanism for adaptation across replicate adaptations, supporting the use of promoter knockout as a method of genetically stable viral attenuation.