Management of the bacterial pathogens Xanthomonas campestris pv. campestris and Pseudomonas syringae pv. porri in cabbage and leek production using novel bacteriophages

Bacterial plant diseases are an important cause of yield loss in agriculture since no disease control agents are currently available to combat them. In cruciferous vegetables belonging to the species Brassica oleraceae (e.g. cabbage, cauliflower and Brussels sprouts) and leek, the most important bacterial diseases are black rot of cabbage, caused by Xanthomonas campestris pv. campestris and bacterial blight of leek, caused by Pseudomonas syringae pv. porri. This study investigated the potential of biological control by bacteriophages, viruses that specifically infect bacteria, their replication resulting in the lysis of their bacterial host and the release of newly formed viral particles.

First, bacterial pathogens were isolated from symptomatic plants. The isolates were identified on the species level by comparing the partial sequence of a housekeeping gene commonly used for barcoding towards the sequence of reference strains in the NCBI GenBank collection. Identification on the pathovar level was achieved by pathogenicity testing. This resulted in the identification of 36 X. campestris pv. campestris isolates and 37 P. syringae pv. porri isolates, as the most prevalent bacterial pathogens on symptomatic plants, thus responsible for the economic losses in Brassica and leek production. These isolates were further characterized by analyzing their genomic diversity using BOX-PCR fingerprinting, dividing the P. syringae pv. porri isolates into two groups. Draft genome sequencing of a strain from each group indicated that differences could mostly be attributed to the presence of mobile genetic elements such as prophages and IS elements. In silico MLSA and ANIb phylogenetically classified the strains as most related to the rice pathogen P. syringae pv. oryzae, also infecting a monocotyl. Among the X. campestris pv. campestris isolates a larger diversity was noted according to differences in their GyrB sequence and BOX-PCR fingerprints. A collection was created with the bacterial isolates obtained in this study, supplemented with reference strains of related pathovars and other bacteria pathogenic on Brassica or leek from the LMG and CFBP collection.

In a second part, the bacterial collections were used to isolate phages from soil samples collected from fields with infected plants, using an enrichment strategy. Seven novel phages were isolated for X. campestris pv. campestris, named SoPhi1-7, and five novel phages were isolated for P. syringae pv. porri, named KIL1-5, supplemented with one h-mutant phage KIL3b selected for the extended host range. The novel phages were characterized to determine their microbiological suitability for phage therapy applications. Their morphology was determined by Transmission Electron Microscopy analysis, identifying SoPhi1, SoPhi2 and all the KIL-phages as members of the Myoviridae family. Host range analysis using the previously created bacterial collection allowed a determination of their specificity. All phages were deemed suitable for phage therapy applications because of their lytic potential and stability under temperature and pH conditions characteristic for agricultural fields. From selected phages, the genome was sequenced and analyzed to exclude the possibility of lysogeny and the presence of toxin-coding genes. Furthermore, phage SoPhi7, with a genome size of 315 kb, was found related to the giant phage PhiKZ, infecting Pseudomonas aeruginosa. Because of the potential transducing capacity of this phage, SoPhi7 should not be used for phage therapy applications until further research proves its safe use. The genomes also provided information for the taxonomic classification of the novel P. syringae pv. porri phages, proposing a new subclade, named KIL-like viruses, within the Felixounavirus genus.

In a third part of this dissertation, the in planta efficiency of the phages was tested. First, an assays on cauliflower transplants demonstrated the antibacterial potential of all X. campestris pv. campestris phages but SoPhi6. Furthermore, SoPhi3 and SoPhi5 significantly reduced symptom development. Next, pot trials were used to test different application strategies. When applied to the soil, phage SoPhi2 reduced symptom development in plants with inoculated leaves, indicating systemic movement of the phage. Significant reduced symptom development was obtained when spraying phage SoPhi2 on cauliflower leaves shortly before bacterial inoculation. Finally, field trials tested the potential of phages in an uncontrolled environment. Transplant treatment of cauliflower plants demonstrated that planting of wet plants in the field should be avoided as this resulted in a higher disease incidence. Furthermore, no disease control effect was found for a phage treatment prior to planting in an infested field. Spraying phages as a leaf treatment reduced symptom development in one trial. Phage assays in leek leaves demonstrated antibacterial potential of all P. syringae pv. syringae phages with KIL1, KIL2, KIL3 and KIL3b causing a significant reduction in lesion length. An infection was not obtained during the pot trials, conclusions can therefore not be made. In the field trials, treatment of leek transplants with phages prior to planting resulted in significant disease reduction in one trial. Spraying phages as leaf treatment reduced disease symptoms in two consecutive trials.

From the results of this study can be concluded that the isolated phages have potential in controlling black rot of cabbage and bacteria blight of leek. Further research is necessary to define the most effective formulation and application method to obtain a reliable phage therapy product.