Development of modified R2 Pyocins for the control of Erwinia amylovora

This thesis project focuses on the development of alternative strategies to antibiotics for the control of fire blight, a serious disease affecting plants of the Rosaceae family, especially apple, pear, and quince. The causative agent, Erwinia amylovora, is a Gram-negative bacterium that enters plant tissues through flowers or wounds and causes extensive necrosis, compromising water and nutrient transport. Due to the ban on streptomycin use in several countries, including Switzerland, there is a growing need for new biocontrol approaches. In this context, this work aimed to engineer R-type tailocins—specifically, R2 pyocins—to target E. amylovora by introducing binding components from the lytic bacteriophage L1. R2 pyocins, naturally produced by Pseudomonas aeruginosa, are phage tail-like bactericidal proteins. They do not contain genetic material, do not replicate, and are characterized by high specificity, mediated by receptor-binding proteins (RBPs). To redirect their activity, native RBPs were replaced with those from phage L1, a lytic Podoviridae known for its activity against E. amylovora. This is partly due to the presence of a tail spike protein (TSP) with depolymerase activity against amylovoran, an extracellular polysaccharide involved in bacterial virulence. To evaluate the potential of phage L1 and its interaction with E. amylovora, a detailed analysis was conducted. After identifying the most susceptible strain, phage extraction and titration were performed, followed by a host range assay on 21 E. amylovora strains. The results confirmed good lytic activity of the phage on several isolates. Interestingly, a CFBP strain lacking cellulose showed higher sensitivity, suggesting that cellulose may influence the exposure or masking of bacterial receptors. For this reason, the study also investigated the ability of phage L1 to bind the bacterial surface, even in the absence of lysis, to better understand receptor accessibility. In parallel, two modified pyocin constructs were designed. In these constructs, the C-terminal domain of the RBP was derived from phage L1, while the N-terminal domain was compatible with the R2 pyocin structure. Coding fragments and adapter sequences were amplified by PCR and assembled using Gibson Assembly. The constructs were validated by Sanger sequencing and transformed into P. aeruginosa Δprf15, a strain lacking the native RBP. The engineered pyocins were tested against a subset of representative E. amylovora strains, selected based on their sensitivity to phage L1. Although the first two constructs did not show higher activity than the wild-type pyocin, sequencing and functional assays confirmed proper assembly and expression of the engineered proteins. A third construct, currently under testing, has been designed and may show improved activity. Overall, this study demonstrates the feasibility of a modular strategy to redirect pyocins toward phytopathogenic bacteria and provides a foundation for future development of sustainable and targeted biocontrol tools.