The genome and proteome of a virulent Escherichia coli O157:H7 bacteriophage closely resembling Salmonella phage Felix O1

Based upon whole genome and proteome analysis, Escherichia coli O157:H7-specific bacteriophage (phage) wV8 belongs to the new myoviral genus, "the Felix O1-like viruses" along with Salmonella phage Felix O1 and Erwinia amylovora phage φEa21-4. The genome characteristics of phage wV8 (size 88.49 kb, mol%G+C 38.9, 138 ORFs, 23 tRNAs) are very similar to those of phage Felix O1 (86.16 kb, 39.0 mol%G+C, 131 ORFs and 22 tRNAs) and, indeed most of the proteins have their closest homologs within Felix O1. Approximately one-half of the Escherichia coli O157:H7 mutants resistant to phage wV8 still serotype as O157:H7 indicating that this phage may recognize, like coliphage T4, two different surface receptors: lipopolysaccharide and, perhaps, an outer membrane protein.


Findings
Bacteriophages (phages) are promising potential alternatives to antibiotics as therapeutics to reduce carriage of pathogens by food animals, thus preventing the spread of organisms such as Escherichia coli O157:H7 along the food chain. Our research has shown that a cocktail of virulent phages can eliminate E. coli O157:H7 from experimentally infected calves [1,2]. Phage V8, isolated originally from sewage [3] was renamed wV8 in our laboratory to indicate that it was obtained from the National Microbiology Laboratory (Winnipeg), and was included in the phage cocktail due to its complementary host range on common phage types (PTs) of E. coli O157:H7. Here we report on the genome and proteome of phage wV8, noting its very close similarity to the Salmonella phage Felix O1 [4][5][6][7].
Phage wV8, purified as described below, was negatively stained with 1% (w/v) uranyl acetate for 20 s and the particles were observed using a LEO912AB and a Philips EM 300 transmission electron microscope operating at 100 kV and 60 kV, respectively. Phage wV8 is a member of the Myoviridae and is morphologically identical to Felix O1 and related phages. Viral particles were morphologically intact and generally have extended tails (Figure 1). Measurement of 20 particles indicated phage wV8 has a head 70.4 nm in diameter and a tail 112.8 × 16.8 nm long. These closely resemble those reported for phage Felix O1, in which the head measured 73 nm in diameter and the noncontracted tail was 113 × 17 nm long [8]. Phage wV8 has a neck of 7 × 7 nm, a collar disk of 10 × 2 nm, and four fibres of 40 × 2 nm that are generally folded along the tail, but may become unfolded in some particles. Tails have transverse striations of 3 nm periodicity, but sometimes present a pattern of overlapping subunits.
For host range studies, phage wV8 was tested for lytic activity on reference strains of 12 common E. coli O157:H7 PTs, the entire ECOR collection [9] and 12 Salmonella enterica serovars. Lytic activity on the reference E. coli O157:H7 PT strains and the Salmonella serovars was determined at multiplicities of infection (MOI) of between 0.001 and 10 in broth cultures in microplates incubated for 5 h at 37°C before inspection for complete lysis (no visible turbidity). Bacteria showing no visible lysis at any MOI were considered resistant to phage wV8, while those showing complete lysis at MOIs of 10 or less were considered sensitive to phage wV8. Strains of the ECOR collection were tested as freshly seeded bacterial lawns on agar plates spotted with 20 μl of diluted phage wV8 containing 10 4 -10 6 pfu. After incubation for 18 h at 37°C, strains showing >50% lysis were considered sensitive. Phage wV8 is highly specific for E. coli O157:H7 strains, completely lyses the 12 most common E. coli O157:H7 (PTs) isolated in Canada [10] and has no lytic activity against any of the Salmonella strains (Table 1).
Phage wV8 was propagated on E. coli strain EC990779 (ECOR strain 6, O173:H), precipitated from clarified lysates using polyethylene glycol 8000 and purified through two rounds of CsCl equilibrium gradient centrifugation [11]. The DNA was isolated as described by these authors and subjected to pyrosequencing at the National Microbiology Laboratory (Winnipeg, MB). Prior to annotation, the genome was opened immediately upstream of the rIIA gene so that it could be directly compared with the sequence of Felix O1. The genome was annotated using Kodon (Applied Maths, Austin, TX) and a variety of online tools http://molbiol-tools.ca including tRNAScan-SE [12] and ARAGORN [13] at their default setting. The GenBank accession number for this sequence is EU877232.  Tail fibre proteins from related phages typically show strong sequence similarity at the N-termini, where the protein associates with the phage tail plate. The carboxy termini, associated with receptor interaction, vary considerably. With Felix O1 and wV8, we see a completely different type of relationship: four regions of similarity separated by regions of dissimilarity, with both the C-and N-termini conserved (see Additional file 1) [14].
Since Felix O1 is LPS-specific [15], we analyzed wV8resistant mutants of E. coli O157:H7. An overnight broth culture of an E. coli O157:H7 strain was mixed with excess wV8 and incubated on plates for 24 h. Nine independent mutants were isolated and serotyped by the E. coli (VTEC) Reference Laboratory at the Laboratory for Foodborne Zoonoses. Approximately one-half of these still serotyped as O157:H7, while half were untypable (rough) indicating that this phage may recognize, like coliphage T4, two  Whole genome comparisons were made at the DNA level using Mauve [16] and Advanced Pipmaker [17] and at the protein level using CoreGenes [18]http:// binf.gmu.edu:8080/CoreGenes2.0/custdata.html. The latter program revealed that Felix O1 and wV8 share 92% of their proteins in common. Mauve analysis (Figure 3) reveals considerable sequence similarity between Felix O1 and wV8 with a few noticeable differences which centre at 11.9, 26.7 52.3, and 60 kb on the Felix O1 genome. The presence of heterologous sequences within these phage genomes is completely in accord with the evolution of the viruses via horizontal gene transfer [19].
Based upon an extensive analysis of relationships between prokaryotic viruses (Lavigne R, Summer EJ, Seto D, Mahadevan P, Nilsson AS, Ackermann H-W et al.: Classification of Myoviridae bacteriophages using BLASTP-tools: submitted) this level of similarity indicates that wV8 should be classified into the newly proposed genus,"Felix O1 viruses", along with Erwinia amylovora phage φEa21-4.

Conclusion
E. coli O157:H7-specific phage wV8 is a member of the Myoviridae and is closely related to the Salmonella-specific phage, Felix O1. Their tail fibre proteins show a unique pattern of sequence relationship.