The Pseudomonas syringae species comprises over 50 identified pathovars, which collectively infect almost all economically important crops (over 180 species) [1, 2], causing huge losses worldwide [3-6]. The use of copper compounds is the standard control method for diseases caused by P. syringae [7, 8]. However, this method has numerous limitations, such as soil contamination, phytotoxicity, a negative impact in plant microbiota, and the emergence and dissemination of copper resistant bacteria [1, 8-12]. An alternative control strategy against phytopathogens involves the use of bacteriophages or phages. These viruses specifically infect bacteria and can be targeted towards a particular species or even specific strains [13]. Biocontrol using phages presents many advantages compared to chemical control, such as being innocuous to eukaryotic cells, specificity, effectiveness against copper resistant bacteria, and overall low cost [13-17].
Here, we present the genome of Psxφ15, a lytic Pseudomonas phage isolated from the phyllosphere (leaves, stems and buds) of a plum tree located in La Aurora, Curacaví (Región Metropolitana, Chile) in June 2019. Samples were prepared by suspending them in tap water, filtered through 0.22 µm filter and then were aerobically enriched at 30°C using the P. syringae strain Ps15 (isolated from a cherry tree with bacterial canker symptoms) for 24 h in Brain Heart Infusion broth. The propagate was centrifuged at 7,000 g for 20 minutes and the supernatant was filtered through 0.22 µm filter. Five percent chloroform was then added to the filtrate. The resulting lysate was then plated with P. syringae Ps15 using the double overlay agar assay and incubated at 30°C for 24 h. Psxφ15 was purified by extracting a single plaque, which was then propagated and plated with P. syringae Ps15 using the double overlay agar assay. This purification process was done twice, obtaining different clear and circular plaques each time of both large (2 mm) and small sizes (<1 mm), which may be due to an unusual plaque polymorphism (Fig. 1A). This phenomenon has been reported previously for other phages, in which extraction of a single plaque resulted in the generation of multiple plaque sizes when inoculated on a double-layer agar plate with its host [18, 19]. A volume of a high titer lysate (109 PFU/mL) was used for transmission electron microscopy (TEM), using 2% uranyl acetate for negative staining. The TEM showed that Psxφ15 belongs to the Myoviridae family [20] with a capsid of 50-60 nm in diameter and a contractile tail of 100-110 nm in length (n = 6, Fig. 1B).
Phage DNA was isolated from 20 mL of a high titer lysate (109 PFU/mL) using the Qiagen lambda midikit (Promega, USA) according to the manufacturer’s instructions. The DNA integrity was visualized through a 0.8% agarose gel electrophoresis and its concentration was quantified using a NanoDrop™ Lite spectrophotometer. DNA was sequenced using the Illumina Novaseq 6000 sequencing platform obtaining around 500,000 paired-end 250 bp reads and a 30x depth of coverage through the MicrobesNG sequencing service (University of Birmingham, UK). Reads were quality and adapter trimmed using Trimmomatic version 0.30 [21] with a sliding window quality cutoff of Q15, and de novo assembly was performed using SPAdes version 3.7 [22], resulting in 2,242 contigs. Host DNA was removed by binning or grouping contigs arising from the same genome using CONCOCT version 1.1 [23] in the KBase website [24]. Each bin was analyzed using NCBI BLASTn [25], and those recognized as host-related were excluded. One bin, associated with a single contig, exhibited a 94.20% identity and a 90% query coverage with Pseudomonas phage VCM (GenBank accession number LN887844, accessed on 19 November 2023). This contig was identified as the genome of the Pseudomonas phage Psxφ15 and was selected for further analysis and annotation.
The genome of Psxφ15 comprises a double-stranded DNA sequence of 96,038 bp in length with a G+C content of 48.35%. Annotation using Pharokka version 1.6.0 [26] identified 201 coding sequences (CDS) with 73% (147 genes) classified as hypothetical proteins. Moreover, further functions were assigned using EggNOG-Mapper version 2.1.9 [27, 28] and RAST [29-31]. Additionally, each CDS was run through BLASTp, resulting in 55 CDS with an identified function, whilst the remaining CDS encoded hypothetical proteins. A circular genome plot for Psxφ15 was generated using the pharokka_plotter.py script from the Pharokka software illustrating the modular organization of the Psxφ15 annotation (Fig. 2), including “DNA/RNA and nucleotide metabolism”, “lysis”, “moron, auxiliary metabolic gene and host takeover”, “structural (head and packaging, connector, and tail)”, and “other functions” (Supplementary Table S1). No repressors or integrases were found in the genome through all annotation tools, confirming Psxφ15 as a virulent phage. Temperate phages are not recommended for biocontrol as these viruses additionally possess the lysogenic cycle, which may confer evolutionary advantages to the bacterial host through the transfer of beneficial genes [32, 33]. The annotation tools also revealed that there are 10 tRNA genes in the genome of Psxφ15. It has been suggested that tRNAs in phage genomes play a role in evading tRNA-targeting host defenses and are involved in higher fitness as well, therefore increasing phage virulence [34, 35]. Furthermore, it is of importance to determine the presence of bacterial toxins and/or antibiotic resistance genes in phage genomes, as biocontrol using these viruses may be hindered by the transfer of these genes through lysogenic conversion or transduction [36, 37]. The presence of genes coding for bacterial toxins was screened using VirulenceFinder 2.0 [38], while ResFinder version 4.4.3 [39] and CARD version 3.2.9 [40] were used to detect possible antibiotic resistance genes in the phage genome. All these bioinformatic tools were run with default parameters. This analysis showed that Psxφ15 lacks bacterial toxins and antibiotic resistance genes.
The host range of phage Psxφ15 was determined using four P. syringae pv. syringae (Pss), P. syringae pv. tomato DC3000 (Pst DC3000), P. syringae RAYR-BL, P. syringae pv. maculicola ES4326, other P. syringae (n = 29) and Pseudomonas spp. strains (n = 19) obtained from cherry tree canker samples and other sources, and several non-Pseudomonas strains (n = 36) (Supplementary Table S2). Ten µL of a 109 PFU/mL titer lysate was spotted onto double-layer agar plates containing the bacterium and incubated at 30°C for 24 h. A clear lysis area was formed for Pst DC3000 and one of the Pss strains, and a turbid halo with small (< 1mm) clear plaques was shown for the other three Pss strains. A turbid halo with no appreciable plaques was formed for P. syringae RAYR-BL. One environmental Pseudomonas spp. was susceptible to the phage, which needs further investigation to determine whether this strain is pathogenic or commensal. Other P. syringae, Pseudomonas spp. and non-Pseudomonas strains tested were not susceptible to Psxφ15 (Table 1).
Psxφ15 was isolated from a seemingly healthy plum tree and showed activity against pathogenic P. syringae strains such as Pst DC3000 and Pss, which are highly virulent to tomato plants and cherry trees, respectively. This is interesting to consider for further testing against other pathogenic P. syringae pathovars. The lytic cycle of this phage and lack of bacterial toxins and antibiotic resistance genes in its genome may make Psxφ15 suitable for the elaboration of a phage cocktail for the biocontrol of economically important P. syringae strains.