HerbertWM and Beyond: A Multi-Faceted Approach to Understanding and Engineering Bacteriophage Host Range

In order for bacteriophage to be useable as reliable and efficient tools in synthetic biology, real-world problems, and other scientific applications, there must be a way to screen efficiently for viruses when targeting specific hosts, especially on the strain-level; in the real world, bacteria constantly mutate and evolve, developing anti-phage defense mechanisms and diverging significantly from model strains used in the lab. Additionally, there is a dire need for further characterization and engineering of phage and phage-satellite circuits in lesser-studied hosts. This thesis contributes to the field of bacteriophage research through a multi-faceted approach to host range prediction and phage engineering. By manually compiling a comprehensive dataset of all publicly available E. coli and Mycobacteria phage host range assays, this work presents the first infectivity prediction model to incorporate confirmed incompatibility features alongside compatibility and homology metrics. The E. coli model demonstrates consistent strain-level accuracy, addressing the high-resolution specificity required for targeting dynamic bacterial populations beyond static lab strains. Preliminary engineering efforts using the pYS2/pYS1 system lay the groundwork for modifying phage HerbertWM, offering a promising strategy to refine its host range and develop a controllable lysogenic switch for targeting Mycobacteria. Lastly, the characterization and initial engineering of phagelet genomes provide both practical tools and foundational insights into these understudied genetic elements, opening new avenues for synthetic biology applications.