THERAPEUTIC BACTERIOPHAGES
20220073885 · 2022-03-10
Inventors
Cpc classification
C12N7/00
CHEMISTRY; METALLURGY
C12N2795/10122
CHEMISTRY; METALLURGY
C12N1/04
CHEMISTRY; METALLURGY
C12N2795/10121
CHEMISTRY; METALLURGY
International classification
C12N7/00
CHEMISTRY; METALLURGY
Abstract
There is provided a bacteriophage of the Myoviridae family, comprising a genome of circularly permuted double-stranded DNA at a length of 200kbp or more; characterised in that the phage has any one or more of the following features: a) does not have a holin gene; and/or b) binds to the Salmonella lipopolysaccharide; and/or c) has only one tail fibre protein; and d) has a sequence with at least 95% sequence identity to SEQ ID NO. 7 and/or a sequence with at least 80% sequence identity to SEQ ID NO.8. Suitably, the bacteriophage may be used to treat salmonella, in particular in the treatment of salmonella in pigs and fowl.
Claims
1.-21. (canceled)
22. A bacteriophage of the Myoviridae family, comprising a genome of circularly permuted double-stranded DNA at a length of 200 kbp or more; characterised in that the phage has any one or more of the following features: a) does not have a holin gene; and/or b) binds to the Salmonella lipopolysaccharide; and/or c) has only one tail fibre protein; and d) has a sequence with at least 95% sequence identity to SEQ ID NO. 7 and/or a sequence with at least 80% sequence identity to SEQ ID NO.8.
23. The bacteriophage of claim 22, wherein the bacteriophage is able to lyse 5 or more Salmonella strains associated with fowl, pigs and/or cows.
24. The bacteriophage of claim 23, wherein the Salmonella strains comprise any one or more of the following serotypes: a) 25 S. typhimurium; b) 15 S. 4,12:i:-; c) 10 S. 4.5, 12:i:-; d) 10 S. bovismorbificans, and e) 10 S. derby.
25. The bacteriophage claim 22, having 80-100% sequence identity with any one of phages with deposit numbers NCTC 18080701; NCTC 18080702; NCTC 18080703; NCTC 18080704; NCTC 18080705; or NCTC 18080706.
26. A panel of bacteriophage, the panel comprising two, three, four or more of the bacteriophage of claim 22.
27. The panel of claim 26, comprising bacteriophages with deposit numbers NCTC 18080703 and 18080704.
28. A pharmaceutical composition comprising of any of the bacteriophage of claim 1.
29. A pharmaceutical composition comprising of any of the panel of claim 26.
30. A pharmaceutical composition comprising of any of the bacteriophage of the panel of claim 27.
31. Animal feed comprising any of the bacteriophage of claim 1.
32. Animal feed comprising the panel of bacteriophage of claim 5.
33. A method of treating or preventing a Salmonella infection in fowl, pigs or cows comprising, providing a composition for ingestion to said fowl, pigs or cows, wherein said composition comprises any of the bacteriophage of claim 1.
34. A method of treating or preventing a Salmonella infection in fowl, pigs or cows comprising, administering a composition to said fowl, pigs or cows, wherein said composition comprises any of the bacteriophage of claim 1.
35. A method of treating or preventing a Salmonella infection in fowl, pigs or cows comprising, providing a composition for ingestion to said fowl, pigs or cows, wherein said composition comprises any of the bacteriophage of the panel of claim 26.
36. A method of drying a bacteriophage solution to form a powder, the method comprising: a) mixing the bacteriophage with any one or more of the following; (i) a sugar; and/or (ii) a sugar alcohol; and/or (iii) an amino acid; and/or (iv) a polymer; and, b) drying the mixture.
37. The method of claim 36, wherein the sugar is a glucose-based sugar, and the amino acid is an aliphatic amino acid.
38. The method of claim 37, wherein said sugar is trehalose, said sugar alcohol is mannitol and said amino acid is leucine.
39. A bacteriophage composition comprising a phage, a sugar, a sugar alcohol, an amino acid and a polymer.
40. A heat stable phage comprising: a) Polymorphism 1 in a sequence with at least 80% sequence identity with SEQ ID NO. 1; and/or b) Polymorphism 2 in a sequence with at least 80% sequence identity with SEQ ID NO. 2; and/or c) Polymorphism 3 in a sequence with at least 80% sequence identity with SEQ ID NO. 2; and/or d) Polymorphism 4 in a sequence with at least 80% sequence identity with SEQ ID NO. 2.
41. A method of making a heat stable phage comprising mutating the nucleotide sequence of a phage genome at: a) the position of polymorphism 1 from bases a, g or c to t in a sequence with at least 80% sequence identity with SEQ ID NO. 1; b) the position of polymorphism 2 from bases a, t or c to g in a sequence with at least 80% sequence identity with SEQ ID NO. 2; c) the position of polymorphism 3 from bases a, g or t to c in a sequence with at least 80% sequence identity with SEQ ID NO. 2; and d) the position of polymorphism 4 from bases a, g or c to t in a sequence with at least 80% sequence identity with SEQ ID NO. 2, wherein the method optionally comprises a first step of introducing SEQ ID NO.1 and/or SEQ ID NO. 2 into the phage genome.
Description
DESCRIPTION OF THE FIGURES
[0186]
[0187] A. Multiple samples were sourced in the United Kingdom for phage isolation. (1) A nature reserve with wild boars in Hampshire (SPFM1-SPFM3), (2) a food processing plant in Essex (SPFM4-SPFM11), (3) pig farm in Warwick, Warwickshire (SPFM12-SPFM15), (4) pig farm in Hinckley, Leicestershire (SPFM16-SPFM20) and (5) a nature reserve with pigs in West Sussex (SPFM21-SPFM22). UK map sourced with permission from © Crown Copyright and Database Right [2017]. Ordnance Survey (Digimap Licence). B. All 22 isolated phages were myoviruses identified by transmission electron microscopy (TEM) and a representative micrograph is presented. The black bar represents 100 nm.
[0188]
[0189]
[0190]
[0191]
[0192]
[0193]
[0194]
[0195]
[0196]
[0197]
[0198]
EXAMPLES
[0199] Aspects and embodiments of the present invention will now be illustrated by way of example only and with reference to the following experimentation.
[0200] Data is presented on the characterisation of twenty-one Salmonella phages isolated in this study, referred to as SPFM phages. Phage killing activity was tested against prevalent and relevant circulating multi-drug resistant UK-pig Salmonella strains isolated from outbreaks in UK farms. The efficiency of plating and host range were determined for all SPFM phages. They were also sequenced, studied by phylogenetic analysis, compared to all previously sequenced Salmonella phages in the NCBI database, and their positively selected genes were identified. This is the first study to fully characterise and sequence a large collection of phages targeted against UK-pig-associated Salmonella strains.
[0201] Experimental Procedures
[0202] Bacterial Strains and Growth Conditions
[0203] In total 68 Salmonella strains were used in this study. Salmonella enterica subsp. enterica serovar Typhimurium SL1344 (accession number FQ312003) was used as a reference strain. The remaining 67 Salmonella enterica subsp. enterica strains, were isolated by the Animal and Plant health Agency (APHA) in Weybridge, UK from pigs in the UK between 2012 to 2015. From these strains the serotypes were: 22 S. typhimurium; 15 S. 4,12:i:-; 10 S. 4,5,12:i:-; 10 S. bovismorbificans and 10 S. derby. All strains were resistant to at least one of the following antibiotics: Nalidixic acid, Tetracycline, Neomycin, Ampicillin, Furazolidone, Ceftazidime, Sulfamethoxazole Trimethoprim, Chloramphenicol, Amikacin, Amoxicillin/clavulanic acid, Gentamicin, Streptomycin, Compound Sulphonamide, Cefotaxime, Apramycin and Ciprofloxacin.
[0204] All Salmonella strains were stored in 50% glycerol broth (Abtek Biologicals Ltd., UK) at −80° C. Strains were routinely grown on Xylose Lysine Deoxycholate (XLD) agar (Oxoid, UK) for 18 hours at 37° C. before being sub-cultured in NZCYM broth (Melford Biolaboratories Ltd, UK) for 18 hours at 37° C. at 100 rpm.
[0205] Phage Isolation, Purification and Propagation
[0206] All samples were collected from five geographical locations in the UK from July to December 2015. In total 15 samples were collected from a nature wild boar reserve in Hampshire; 10 samples from a food processing plant in Essex; 15 samples from a finishing pig farm in Warwickshire; 18 samples from pig farms in Leicestershire with piglets and finishing pigs; and seven samples from a nature reserve with wild pigs in West Sussex (
[0207] All samples were processed using the same enrichment procedure, 1 ml or 1 g of sample was mixed with 9 ml of NZCYM broth and 100 μl of an exponential growing Salmonella culture was added. To maximise phage isolation the same sample was aliquoted and enriched with 12 different MDR Salmonella host strains individually. Enrichments were incubated at 37° C. for 12 hours with shaking (at 100 rpm), after which samples were centrifuged at 4 000×g for 15 minutes at room temperature. The supernatant was filtered through 0.22 μm pore size syringe filters (Millipore, UK) and the filtered samples were stored at 4° C. until further use. Filtrates were screened for phage by the small drop plaque assay method (Mazzocco et al., 2009). Briefly, 10 μl of enriched sample was spotted on a Luria-Bertani (LB) 1% (w/v) agar plate (Fisher Scientific, UK) with NZCYM 0.5% (w/v) agar as the top layer, which was mixed with 100 μl of exponentially growing Salmonella culture. Plates were incubated at 37° C. for 18 hours and examined for phage lysis either by presence of clearing or phage plaques.
[0208] For phage purification individual plaques was picked with 1 μl loops, mixed with 500 μl of SM buffer with gelatin (100 mM NaCl, 8 mM MgSO.sub.4.7H.sub.2O, 50 mM Tris-CI and 0.01% (w/v) gelatin) and centrifuged at 21,000×g for 10 minutes. The resultant supernatant was used for the next round of phage purification by the double agar overlay plaque assay (Kropinski et al., 2009) and the process was repeated seven times to produce clonal phage stocks.
[0209] Increased volumes of purified phage lysates were made by mixing exponential growing liquid cultures of SL1344 (10′ CFU/ml) of Salmonella infected with 107 PFU/ml phages in NZCYM broth at 37° C. with shaking (100 rpm) for 6 hours. Phage cultures were centrifuged at 4,200×g for 15 minutes, the supernatant was filtered with 0.22 μm pore size filters and phage lysates were stored at 4° C. To determine phage titre, phage lysate was serially diluted 10-fold and the small drop plaque assay method (Mazzocco et al., 2009) was used on LB 1% agar plates. Final phage titres were expressed as PFU/ml.
[0210] Transmission Electron Microscopy
[0211] Phage lysates were concentrated before Transmission electron microscopy (TEM) analysis by centrifugation at 21,000×g for 1 hour, the pellet was resuspended with 0.1 M ammonium acetate (Fisher Scientific, UK), centrifuged at 21,000×g for 1 hour and resuspended with a 0.1 M ammonium acetate solution. The highly concentrated phages (101 PFU/ml) were negatively stained with 1% uranyl acetate (w/v) for 10 seconds and applied to 3 mm carbon coated copper grids (Agar Scientific Ltd, UK). Phages were examined with an EOL 1220 (Joel UK LtD, UK) ran at 80 kV and images were acquired by SIS Megaview III camera with analysis software (Olympus Soft Imaging Solutions, Germany) (Ackermann 2009). TEM analysis was conducted by Dr Ali Ali and Natalie Allcock, Core Biotechnology Services, University of Leicester, UK.
[0212] Phage Host Range Analysis and Efficiency of Plating
[0213] The host range of individual phages was determined by the small drop plaque assay method (Mazzocco et al., 2009) on different Salmonella enterica subsp. enterica serotypes and incubated for 18 hours at 37° C. Plates were examined for either bacterial lysis via clearing or plaques or for no infection and average observations were noted from three biological replicates, each with three technical repeats.
[0214] Efficiency of plating (EOP) was conducted on two representative MDR strains from five Salmonella serotypes: S. typhimurium; S. 4,12:i:-; S. 4,5,12:i:-; S. bovismorbificans and S. derby. EOP was also established on the phages propagation host S1344. For EOP the small drop plaque assay method was used and phage lysates were 10-fold serially diluted and spotted onto bacterial lawns (Kutter, 2009). Average PFU/ml EOP values were calculated from three biological replicates each with three technical repeats. Principle component analysis was done for the EOP values of all phages, on all strains screened using the prcomp function in the R base package (R Core Team, 2017) and plotted as a biplot with the autoplot function from the ggplot2 package (Wickham, 2009).
[0215] Phage DNA Extraction, Sequencing and Annotation
[0216] High titre phage lysate (10.sup.10 PFU/ml) was used to extract DNA using a revised phenol-chloroform-isoamyl method as previously described (Nale et al., 2016). Following the extraction the final DNA pellet was dissolved in 5 mM Tris HCl, quantified using the Qubit fluorometer with the Qubit double-stranded HS kit (Thermo Scientific, UK) and sequenced by the Illumina MiSeq platform. NexteraXT libraries were prepared according to the manufacturer's instructions, using 1 ng of input DNA. Reads were trimmed with Sickle v1.33 (Joshi and Fass, 2011) prior to assembly with SPAdes v v3.9.1 (Bankevich et al., 2012). Annotation was carried out with Prokka v1.11 (Seemann 2014) using a custom database constructed from all phage proteins (December, 2017) and hmmscan to identify pVOGs (Grazziotin et al., 2017). The accession numbers assigned to the phage genomes are listed in Table 1. There was an error in uploading the phage genomes to European Nucleotide Archive (ENA) and the phage names are incorrect. Even so, the accession numbers listed in Table 1 are correct and allows to distinguish between the phages. ENA are currently resolving the issue and the correct files with the right names can be found via this link http://s3.climb.ac.uk.info/Sinfo/SPFM_genome.tar,Qz
[0217] Calculating ANI Between Salmonella Phage Genomes
[0218] 158 genomes from Salmonella-infecting phages were collected from both phages newly sequenced in this study and phages sequenced from previous studies. The average nucleotide identity, as defined by (Goris et al., 2007), was measured between all pairwise combinations of phage genomes using the BLASTN alignment option in the pyani package (Pritchard et al., 2016) and plotted in an interactive heatmap using heatmaply (Galili et al., 2017).
[0219] Protein Orthologue Clustering
[0220] Phage protein sequences annotated by Prokka were clustered into orthologous groups using the COG triangles algorithm (Kristensen et al., 2010) implemented in the get_homologs.pl script (Vinuesa and Contreras-Moreira, 2015) using a BLASTP e-value threshold of 1e-5. The Jaccard distance between each phage genome was based on gene presence/absence as determined from COG clustering of phage protein sequences. The Jaccard distance was calculated using the base R dist function, a dendrogram was constructed from the distances using the base R hclust function, and the dendrogram was plotted using the ggdendro package in R (Vries and Ripley, 2016). Ortholog clustering data is shown in Supplementary Table 4 as published in Thanki AM, Brown N, Millard AD, Clokie MRJ. Genomic Characterization of Jumbo Salmonella Phages That Effectively Target United Kingdom Pig-Associated Salmonella Serotypes. Front Microbiol. 2019; 10:1491. Published 2019 Jul. 2. doi:10.3389/fmicb.2019.01491.
[0221] Postulating a Phylogenomic Tree Based on Core Genes
[0222] COGs that contained protein sequences from the Seoulvirus genus phage genomes were used to construct codon-aware alignments of the corresponding nucleotide sequences with MUSCLE v3.8.31 (Edgar, NAR 32(5)) and the pal2nal.pl script (Suyama et al., 2006). SNPs from the codon-aware nucleotide alignments were extracted with snp-sites (Page et al., 2016) and used to construct a phylogenomic tree using FastTree v.2.1.10 SSE3 (Price et al., 2010) with the generalized time-reversible model of nucleotide evolution.
[0223] Measuring Positive Selection
[0224] A test for positive selection was done at each codon of each orthologue cluster using the HyPhy package (Pond et al., 2005). In particular, a protein alignment of each orthologue cluster from MUSCLE v3.8.31 was used to guide construction of a codon-aware alignment of the nucleotide sequences from each cluster. Recombination breakpoints in each codon-aware alignment were detected using the GARD algorithm (Pond et al., 2006) implemented in the HyPhy package and partitioned alignments and trees were constructed on either side of each breakpoint. These partitions were tested for positive selection at each site with the Bayesian FUBAR algorithm (Murrell et al., 2013) implemented in the HyPhy package, which returns the posterior probability that each site is under positive selection (that is, the dN/dS is greater than 1). As the sample size is too small to allow accurate estimation of dN/dS at individual codons, the posterior probability values are given alongside the dN/dS values. The hypothesis test is sensitive enough to detect positive selection and the PSRF/N effective ratio was maintained below 0.006, so that the posterior probability of positive selection was converging on a solution.
[0225] Results
[0226] Several in vitro and in vivo studies have shown that phages can significantly reduce Salmonella spp. numbers in food settings and in pigs and can therefore offer a viable option for reduction of Salmonella spp. in pigs to improve food safety. However, to date not all published studies have fully characterised their phages in terms of infectivity and by sequencing. Sequencing is useful for phages used in therapy because it can confirm that the phages are obligately lytic, that they do not carry toxin genes, that they do not belong to particular phage genera known to facilitate the spread of antibiotic resistance genes from the host (such as the Felixo1virus genus), and to help rationally optimize phage therapy cocktails. The inventors have characterised 21 new lytic jumbo SPFM Salmonella phages isolated from environmental sources targeted again Salmonella spp commonly associated with pigs. All SPFM phages have broad-host ranges and high EOP on clinically relevant UK MDR Salmonella strains isolated from pigs and could be utilised for phage therapy and can help to improve food safety.
[0227] Phage Isolation, Plaque Morphology and Propagation
[0228] Samples for phage isolation were collected between July and December 2015 from a food processing plant, from wild and domestic pigs and boars in the UK. The 65 samples were screened and 21 phages were isolated, which are named SPFM1 to SPFM22. Phage SPFM18 was excluded from analysis due to its incomplete genome sequence. 15 SPFM phages were isolated on five different Salmonella enterica serotype Typhimurium strains and the remaining seven on Salmonella enterica serotype 4,12:i:-strains. Three phages originated from wild boar faeces, eight from the food processing plant, and ten from finishing pig and piglet faeces obtained from farms in Warwickshire, Hinckley and West Sussex (
[0229] In terms of plaque morphology, phages SPFM9, SPFM10 and SPFM11 were isolated on S. typhimurium strains and produced clear −1 mm in diameter plaques and the remaining 18 phages produced clear plaques of ˜0.5 mm in diameter. All phages were propagated on S. typhimurium SL1344 to produce high titre (10.sup.10 plaque forming units (PFU)/ml) stocks. The main motivation for this is that the original isolation strains spontaneously release prophages, which complicates downstream characterisation, whereas prophage release from SL1344 was never observed.
TABLE-US-00013 TABLE 1 Summary of isolated phages with their respective isolation host, source in the UK and genome information. Coding Genome sequences GC content Accession Phage Isolation host serotype Source.sup.a size (bp) (CDS) tRNA (%) number.sup.b SPFM S. Typhimurium 1 242,624 307 1 48.78 LR535901 1 SPFM S. 4, 12:i:- 1 240,111 260 1 48.62 LR535921 2 SPFM S. Typhimurium 1 240,198 257 1 48.62 LR535920 3 SPFM S. Typhimurium 2 240, 197 257 1 48.62 LR535902 4 SPFM S. Typhimurium 2 240,194 287 1 48.88 LR535903 5 SPFM S. Typhimurium 2 240,198 305 1 48.62 LR535905 6 SPFM S. Typhimurium 2 240,197 290 1 48.62 LR535904 7 SPFM S. 4, 12:i:- 2 240,197 297 1 48.84 LR535906 8 SPFM S. Typhimurium 2 240,197 298 1 48.62 LR535907 9 SPFM S. Typhimurium 2 240,197 285 1 48.62 LR535908 10 SPFM S. Typhimurium 2 240,197 257 1 48.62 LR535909 11 SPFM S. Typhimurium 3 240,197 305 1 48.62 LR535911 12 SPFM S. 4, 12:i:- 3 241, 405 261 1 48.57 LR535910 13 SPFM S. 4, 12:i:- 3 240,197 289 1 48.61 LR535912 14 SPFM S. Typhimurium 3 239,951 289 1 48.63 LR535913 15 SPFM S. 4, 12:i:- 4 233,195 249 1 48.63 LR535915 16 SPFM S. 4, 12:i:- 4 239,842 258 1 48.64 LR535914 17 SPFM S. 4, 12:i:- 4 240,197 258 1 48.62 LR535916 19 SPFM S. Typhimurium 4 236,956 296 1 48.62 LR535917 20 SPFM S. Typhimurium 5 240,196 259 1 48.62 LR535919 21 SPFM S. Typhimurium 5 240,196 259 1 48.62 LR535918 22 .sup.aSamples source numbers relate to FIG. 1: (1) A nature reserve with wild boars in Hampshire, (2) a food processing plant in Essex, (3) pig farm in Warwick, Warwickshire, (4) pig farm in Hinckley, Leicestershire and (5) a nature reserve with pigs in West Sussex. .sup.bAll genome sequences can be found via this link: http://s3.climb.acuk/Sinfo/SPFM genome.ta ngz.
indicates data missing or illegible when filed
[0230] Phage Morphology
[0231] TEM analysis revealed the 21 phages have isometric heads and contractile tails and so were classed as members of the Myoviridae family within Caudovirales (
[0232] Host Range Analysis
[0233] The efficacy of SPFM phages was tested on 67 MDR Salmonella enterica subsp. enterica strains isolated from pigs. All 67 strains are representatives of the top five UK pig associated Salmonella serotypes, namely S. typhimurium, S.4,12:i:-, S.4,5,12:i:-, S. bovismorbificans and S. derby. All phages have a wide lytic spectrum of activity, and each lyses over 80% of strains (
[0234] Efficiency of Plating (EOP)
[0235] In order to determine how efficient phage infection is on clinically relevant strains, and thus decide which phages should ultimately make up a phage cocktail, the EOP for the SPFM phages was carried out on two representative MDR pig strains from Salmonella serotypes S. typhimurium, S.4,12:i:-, S.4,5,12:i:-, S. bovismorbificans and S. derby. The EOP of phages on their propagation host, S. typhimurium SL1344 was included as a control. To identify patterns within the EOP data set a principle component analysis (PCA) was used, which revealed that the phages cluster into two distinctive groups (
[0236] Genome Characterisation of SPFM Phases
[0237] The SPFM phage set was sequenced using the Illumina MiSeq platform. All 21 phages have linear, circularly permuted dsDNA genomes ranging from 233 Kb to 242 Kb and encode between 258 to 307 coding sequences (CDS) (Table 1). As all genomes are larger than 200 Kb they are classified as jumbo phages and all are predicted to be lytic based on the absence of known lysogeny associated genes.
[0238] Despite the variation in SPFM genome size, they all encode one tRNA, have average GC contents of 48.5% (Table 1), an average gene length of 0.860±0.010 Kb, gene densities of 1.074/kb and gene coding regions constitutes 93% of their genomes. As the phages are genetically similar in architecture as well as content, a representative genome map of SPFM1 is shown in
[0239] Other genes were identified in all SPFM phage genomes, which could potentially alleviate their dependency on their bacterial host during infection. Genes of particular interest include dihydrofolate reductase, thymidylate synthase and thymidylate kinase, all of which are predicated to be used for folate synthesis and radical S-adenosylmethionine (SAM) genes involved in enhancing host metabolism during phage infection (Lee et al., 2011). In addition, DNA adenine methylase gene was identified in SPFM phages that could provide defence against the hosts' restriction modification systems.
[0240] Hierarchical Cluster Analysis of Isolated SPFM Phages
[0241] Genomes were compared using pairwise local alignment with nucleotide BLAST (NCBI, 1988) and it was observed all SPFM phages were genetically very similar to each other with ˜95% ANI. To gain an insight into variation within the genomes a dendrogram was built based on the presence or absence of 46 shared accessory genes (
[0242] Comparison of SPFM Phages to Previously Sequenced Salmonella Phages
[0243] To determine how similar SPFM phages are to previously sequenced Salmonella phages deposited in Genbank (until December 2017), an all-versus-all comparison analysis against 158 fully sequenced Salmonella phage genomes was conducted (
[0244] Seoulvirus Genus Phage Cluster
[0245] To determine how closely related the SPFM phages are to the Salmonella phages SPN3US and SEGD1, a phylogenetic analysis was constructed (
[0246] Phage Genes Under Positive Selection
[0247] To determine which genes are under evolutionary selection pressure, genes under positive selection were identified within the genus Seoulvirus cluster. This was carried out by determining the ratio of nonsynonymous over synonymous substitution rates (dN/dS) of pairwise comparison of core orthologues of the 21 SPFM phages, SPN3US and SEGD1 (Supplementary Table 5 as published in Thanki AM, Brown N, Millard AD, Clokie MRJ. Genomic Characterization of Jumbo Salmonella Phages That Effectively Target United Kingdom Pig-Associated Salmonella Serotypes. Front Microbiol. 2019; 10:1491. Published 2019 Jul. 2. doi:10.3389/fmicb.2019.01491). The data presented in Table 2 illustrates the posterior probability values above >0.900 for the predicted genes under positive selection and their orthologue cluster number. The analysis predicted that 33 genes are under positive selection. 22 of these genes are putative virion structural proteins, one is a putative endodeoxyribonuclease RusA, two encode thymidylate synthase and eight are hypothetical proteins. For the putative virion structural proteins under positive selection, HHpred (Zimmermann et al., 2017) was used to determine if the structural proteins are involved in phage tail or capsid assembly. The program predicted with over 75% probability that they were putative baseplate wedge proteins and thus likely to be involved in phage tail assembly. In addition, the orthologue cluster number 201 (relates to the orthologue cluster number in Supplementary Table 5) had 98.26% probability hits to C-terminal pectate lyase domain, which is also part of phage tail fiber formation. Not all genes under positive selection could be assigned a function but it can be predicted the following hypothetical proteins with orthologue cluster numbers 106, 108, 129, 130 and 211 could potentially be putative virion structural proteins due to their localisation close to other putative structural proteins on the genome.
[0248] Similarly, the hypothetical protein with orthologue cluster number 92 is positioned close to where the DNA replication and transcription genes are clustered and so likely involved in this function.
TABLE-US-00014 TABLE 2 Genes under positive selection within the Seoulvirus phage cluster, which includes phages SPN3US, SEGD1 and 21 SPFM phages. Ortholog Cluster Posterior dN/dS cluster predicted Reannotation of Codon probability omega number protein predicted protein.sup.a number (dN/dS).sup.b value.sup.c 92 Hypothetical 627 0.9098 0.1179 protein 100 Putative 553 0.9223 0.0648 virion structural protein 100 Putative Putative 762 0.9192 0.0648 virion baseplate wedge structural protein protein 106 Hypothetical 253 0.9332 0.1172 protein 106 Hypothetical 246 0.9019 0.3001 protein 108 Hypothetical 55 0.9404 0.2553 protein 108 Hypothetical 35 0.9402 0.2553 protein 109 Putative 110 0.9462 0.284 virion structural protein 129 Hypothetical 200 0.9479 0.0417 protein 130 Hypothetical 62 0.9697 0.212 protein 139 Putative 209 0.9397 0.2496 endodeoxyri bonuclease RusA 190 Putative Putative DNA 226 0.9545 0.1129 virion binding protein structural protein 192 Putative 39 0.9463 0.4287 virion structural protein 192 Putative 41 0.9449 0.4545 virion structural protein 192 Putative 73 0.9265 0.4287 virion structural protein 194 Putative 207 0.9315 0.1397 virion structural protein 196 Putative 337 0.9448 0.4848 virion structural protein 196 Putative 421 0.9333 0.2123 virion structural protein 196 Putative 124 0.9033 0.237 virion structural protein 199 Putative 429 0.9469 0.0105 virion structural protein 201 Putative Putative pectate 271 0.9783 0.3591 virion lyase domain structural protein 204 Putative 210 0.9143 0.6042 virion structural protein 205 Putative 292 0.9244 0.2598 virion structural protein 211 Hypothetical 27 0.9413 0.1693 protein 219 Putative 241 0.9167 0.1225 virion structural protein 222 Putative Putative 981 0.9272 0.1589 virion baseplate wedge structural protein protein 222 Putative Putative 314 0.9190 0.1806 virion baseplate wedge structural protein protein 223 Putative Putative short tail 207 0.9747 0.2195 virion fiber structural protein 234 Putative 38 0.9253 0.001 virion structural protein 281 Thymidylate 89 0.9236 0.1395 synthase 281 Thymidylate 92 0.9019 0.0625 synthase 289 Putative 118 0.9018 0.0805 virion structural protein 302 Putative 72 0.9652 0.9065 virion structural protein .sup.aThe tool HHPred (Zimmermann et al. 2017) was used to reannotate protein sequences and presented are annotations with over 75% probability. .sup.bThe sample size is too small to allow accurate estimation of dN/dS at individual codons, but the hypothesis test (the posterior probability) is sensitive enough to detect positive selection.
[0249] The inventors have characterised 21 new Salmonella phages. All phages could infect representative MDR strains isolated from pigs. Seven virulent phages were identified that could infect 100% of Salmonella isolates tested. A possible explanation of the high infectivity of these phages could be that as the phages infect S. typhimurium, they can also infect its monophasic variants, which are genetically closely related (Moreno et al., 2013). It is likely the phages are using the same bacterial receptor to attach to strains from these serotypes. These seven phages appear to be ideal for therapeutic use based on their host range and efficiency of plating.
[0250] Two phages SPFM9 and SPFM11 could only infect ˜80% of the strains screened. Additionally, both phages produced turbid clearing on a fifth of the strains, which could indicate potential lysogeny as turbid clearing is often a characteristic of temperate phages (Gallet et al., 2011). Although, sequence analysis confirmed both phages had no known lysogeny modules and are likely to be lytic phages, as only ˜30% of genes have been assigned with a known function, unknown lysogeny modules could exist. A possible explanation to why turbid clearing was observed could be that the bacterial strains screened may have been partly resistant to the phage so only a sub population of cells were infected, which resulted in turbid clearing (Bull et al., 2014). An alternative explanation could be the phages are inducing a prophage within these strains, which could have produced the turbid clearing (Campoy et al., 2006). Although turbid clearing was observed, SPFM9 and SPFM11 were able to infect and replicate on the 11 strains screened for EOP analysis, which indicates lytic infection. Due to potentially incomplete lysis and issue of resistance or induction of prophages, phages SPFM9 and SPFM11 would not be good candidates for therapy (Chan et al., 2013; Abedon et al., 2017).
[0251] To further characterise SPFM phages and to narrow down which have ideal traits for use therapeutically, EOP analysis was conducted (Mirzaei and Nilsson, 2015). The data revealed a group of phages that had no differences in EOP across all representative strains from five different serotypes and as there was no difference in infectivity between the strains, these phages would appear to be ideal candidates for therapeutic application. In particular phages SPFM14, SPFM15 and SPFM17 are good candidates, as they could infect all representative strains from the dominant UK Salmonella serotypes and had high EOP's on the strains screened.
[0252] All isolated SPFM phages had genomes greater than 233 Kb and will significantly add in numbers and diversity to the ˜170 jumbo phage genomes available on NCBI (Yuan and Gao, 2017). The genomes of the SPFM phages were also bigger in size in comparison to other Salmonella phages, such as the myovirus vB_SalM_SJ_3 that has a genome size of 162,910 bp (Wall et al., 2010; Zhang et al., 2010; Saez et al., 2011; Zhang et al., 2014) and the podovirus UAB_78 that has a genome size of 48,110 bp (Bardina et al., 2016). Jumbo phages are rare to isolate and are not frequently isolated by conventional methods, which can be biased towards smaller genome size phages (Hillyard et al., 2016; Saad et al., 2018; Salmond and Fineran, 2015; Serwer et al., 2009). However, it can be argued, from the total phages described to date, approximately 2% are jumbo so it could be jumbo phages are truly rare as stated previously and are not actually underrepresented.
[0253] All SPFM phages were genetically similar to each other, even though different Salmonella strains were used for enrichment and samples from a variety of environmental sources were collected (Jurczak-Kurek et al., 2016). The SPFM phages do differ in SNP's, which could be present in host-interacting proteins. This could affect the attachment kinetics of the phage to the cell surface, leading to changes in host specificity (Switt et al., 2013). This potentially could explain why differences were observed in host range and EOP between the SPFM phages. A similar observation was described with 90% genetically identical Pseudomonas phages and likewise SNP's lead to phenotypic differences between the phages (Ceyssens et al., 2011). It should be noted that phages SPFM5, SPFM9, SPFM10 and SPFM11 share the same SNPs, which could suggest they represent clones with SNP's. The SNPs could have been induced by the propagation host, which was different from the host the phages were originally isolated on.
[0254] SPFM phages were compared to all sequenced Salmonella phages, and they clustered with other known jumbo Salmonella phages SPN3US (Lee et al., 2011) and SEGD1, isolated in South China from chicken faeces and in Korea respectively. It is very interesting that the SPFM phages where isolated in the UK but cluster and are genetically similar to phages isolated in a different continent. Similarly 87% genetically similar Pseudomonas phages were isolated from different countries in the US and Europe (Ceyssens et al., 2011).
[0255] The large genome sizes of jumbo phages, allows for the carriage of numerous genes not present in smaller genome sized phages 17), such as the six RNAP beta subunits that all SPFM phages have. The multiple RNAP subunits of phage SPN3US have been extensively studied in recent publications l., 2017) and are very similar to RNAP beta subunits of phiKZ-like phages. A further three RNAP subunits were predicted recently in phage SPN3US by the construction of amber mutants of phage genes. All three predicted subunits were also identified in SPFM phages: nvRNAP β′, vRNAP β present in the C terminus and vRNAP β′ present in the C terminus (Table A). Furthermore, the predicted three RNAP subunits are part of the core genes shared between the phages. Presence of multiple RNAP beta subunits is consistent with other sequenced jumbo phages, such as the seven RNAP beta subunits that have been identified bioinformatically in V. coralliilyticus phage BONAISHI; Ralstonia solanacearum phages RP12 and CD8RP31. Overall the presence of extra genes in jumbo phages could reduce their dependence on their bacterial host for essential proteins associated with the phage lifecycle and consequently could help broaden the phage host range. This could explain why all SPFM phages can infect multiple clinically relevant Salmonella isolates, which could make SPFM phages ideal candidates for phage therapy.
[0256] Within the Seoulvirus genus, genes under positive selection were identified from the core genes shared by the phages. Genes under positive selection included host-interacting proteins, such as two putative virion structural proteins predicted to be baseplate wedge proteins and involved in the formation of tail fibers, both of which are involved in binding of phages to bacterial cells. These host-interacting proteins have to adapt to different bacterial hosts, which could explain why they are under positive selection. These results could also give a rational explanation in the differences observed in host range and EOP between the SPFM phages. Other phage studies have also identified host-interacting proteins as being under positive selection, such as gene gp6 that encodes the baseplate protein and likely to be involved in host specificity.sup.are. Further genetic and mutation studies are needed to characterise phage genes under positive selection to understand their importance.
[0257] The inventors study has described and characterised 21 genetically similar lytic jumbo phages that can lyse Salmonella strains commonly associated with UK pigs. Comprehensive host range analysis and EOP identified a number of phages that would be ideal candidates for phage therapy to improve food safety.
[0258] The work described below identifies the best phage or phage cocktails that maximally reduce Salmonella both in vitro and in vivo.
[0259] Methods and Materials
[0260] Bacterial Strains and Growth Conditions
[0261] Two Salmonella enterica subsp. enterica serovar Typhimurium strains were used in this study: SL1344 (accession number FQ312003) a laboratory reference strain and strain S01160-12 phage type DT193 isolated by the Animal and Plant Health Agency (Weybridge, UK) from a pig in 2012 (Card et al. 2016). S01160-12 is a multi-antibiotic resistant strain, resistant to tetracycline, neomycin, ampicillin, sulfamethoxazole, chloramphenicol, gentamicin, streptomycin, compound sulphonamide and apramycin. Both Salmonella isolates were routinely grown on Xylose Lysine Deoxycholate (XLD) agar (Oxoid, UK) for 18 hours at 37° C. and for liquid cultures inoculated in NZCYM broth (Melford Biolaboratories Ltd, UK) and grown for 18 hours at 37° C. at 100 rpm.
[0262] Phage Propagation and Titration
[0263] Phages SPFM2, SPFM4, SPFM10, SPFM14, SPFM17 and SPFM19 were propagated. Briefly exponential growing liquid cultures (10.sup.7 CFU/ml) of Salmonella SL1344 growing in NZCYM broth were infected with 10.sup.7 plaque forming units (PFU)/ml phages and incubated at 37° C. with 100 rpm for 6 hours. Phage cultures were centrifuged at 4,200×g for 15 minutes, the supernatant was filtered with a 0.22 μm filter and phage lysates were stored at 4° C. To determine phage titre, phage lysate was serially diluted 10-fold and the small drop plaque assay method (Mazzocco et al., 2009) was used on LB 1% agar plates with a bacterial lawn of SL1344. Final phage titres were expressed as PFU/ml.
[0264] One Step Growth Assay
[0265] Exponential cultures of SL1344 at optical density (OD.sub.600) 0.2 and cell density of 107 CFU/ml were mixed with phage at a multiplicity of infection of 0.01, i.e. 1:0.01 bacteria cell to phage ratio. Phage was allowed to adsorb for 5 minutes at 37° C. and unbound phages were removed by centrifuging at 4200×g for 10 minutes. The pellet was re-suspended with NZCYM broth and incubated at 37° C. with shaking at 100 rpm. Aliquots were taken every 10 minutes for 1.5 hours and the small drop plaque assay method was used to determine the PFU/ml count (Mazzocco et al., 2009). Three biological replicates were performed each with three technical repeats.
[0266] Phage Receptor Analysis
[0267] Sangryeol Ryu kindly provided S. typhimurium SL1344 strains with deletions in the flagellar production gene (flgK); the gene that encodes vitamin B12 uptake outer membrane protein (butB); and the gene involved in the LPS-related O-antigen production (rfaL). The mutant strains were constructed using the lambda red recombination method (Shin et al. 2012). To determine if the phages use either of the three as their host receptor, phage lysates were diluted 10-fold and plated on the mutant strains using the small drop plaque assay method (Mazzocco et al. 2009). Three biological replicates were performed each with three technical repeats.
[0268] Temperature and pH Stability Assays
[0269] Heat stability of phage lysates was tested by exposing the phages to a range of temperatures of 4, 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100° C. for one hour. Phages were incubated in the Polymerase Chain reaction (PCR) machine. For pH studies, 100 μl of phage lysate were added to 900 μl of SM buffer adjusted at a range of pH's from 1 to 14. To determine phage titre after exposure to different temperatures and pH's, phage lysate was serially diluted 10-fold and the small drop plaque assay method was used on LB 1% agar plates with a bacterial lawn of SL1344 (Mazzocco et al. 2009). Final phage titres were expressed as PFU/ml and three biological replicates were performed with three technical repeats.
[0270] In Vitro Bacterial Killing Assay with Phage Cocktails
[0271] Killing assays were conducted with phages alone and with phage cocktails consisting of two, three and four phages mixed at equal volumes. For the different killing assay phage cocktail combinations cultures of S01160-12 were grown to an optical density (OD.sub.600) of 0.2 at which point phage or phages were added at a MOI of 100. In addition, there were two controls: one was a bacterial control, which only included SL1344 and the other was a phage control, which only included the phage(s). Both the control and test samples aliquots were incubated at 37° C. at 100 rpm and aliquots were taken at time-points: 0, 1, 2, 3, 4, 5 and 6 hours. For each time-point bacterial concentrations (CFU/ml) were determined on LB 1% agar plates and phage concentrations (PFU/ml) by the small drop plaque assay method (Mazzocco et al., 2009). For both CFU/ml and PFU/ml three technical replicates were conducted and plates were incubated for 16 hours at 37° C. Three biological replicates were performed for each killing assay. Killing assay graphs were plotted using GraphPad prism 6 and paired t-tests were used to determine the significance of data, where P≤0.5 were considered significant.
[0272] Testing Efficacy of Phage Cocktails in the In Vivo Galleria mellonella Infection Model
[0273] Preparation of G. mellonella
[0274] Larvae were purchased from Live Food UK Ltd. (Rooks Bridge, UK), stored at 4° C. and used within 5 days. For all in vivo experiments larvae that weighed approximately 0.25 to 0.30 g were chosen and were surface sterilised with cotton swabs dipped in 70% ethanol.
[0275] Salmonella Infected G. mellonella Treated with Phage Cocktails
[0276] For all larvae phage therapy studies a previously described method was used (Nale, Chutia, et al. 2016). Both the Salmonella strain S01160-12 at a final concentration of 105 CFU and phages cocktails with a final concentration of 107 PFU were administered to the larvae via the oral route. The strain S01160-12 was prepared by first growing a liquid culture of S01160-12 containing 109 CFU/ml, which was then centrifuged at 4,200×g for 10 minutes. The supernatant was discarded and the pellet was resuspended in 0.1 M phosphate-buffered-saline (PBS) and centrifuged at 4,200×g for 10 minutes. The supernatant was discarded again, the pellet was resuspended in PBS and diluted 10-fold in PBS to 105 CFU/ml. larvae was administered 10 μl of bacterial culture and phage cocktails via the oral route using a Hamilton syringe pump. Phage cocktails were either administered prophylactically, 1 hour prior to infection or simultaneously with Salmonella (all regimens detailed in Table 3). After which, larvae were incubated at 37° C. for 3 days. Every 24 hours, survival rate of larvae were monitored and larvae were sacrificed to determine CFU and PFU counts. For this larvae were stored at −20° C. for 3 hours, dissected to remove their haemolymph, which was suspended in 1 ml of PBS and vortexed for 30 seconds (Nale et al., 2016). For the PFU counts the samples were centrifuged at 5,000×g for 5 minutes. The filtrate was then diluted 10-fold and titred by the small drop plaque assay method (Mazzocco et al. 2009). For the CFU counts the suspension was diluted with 10-fold dilutions and spot tested on XLD agar plates. Salmonella colonies which were recovered after exposure to phage cocktails were picked using 5 μl loops, streaked onto XLD agar plates and re-streaked three times on XLD agar. Colonies were inoculated into NZCYM broth and grown for 2 hours at 37° C. at 100 rpm. The culture was then used to make lawns to determine their sensitivity to phages via spot tests (Mazzocco et al., 2009). Phages at a concentration of 10.sup.8 PFU/ml were spotted. To determine efficiency of plating of the phages on the isolated colonies, phage lysates were diluted 10-fold and plated using the small drop plaque assay method (Mazzocco et al. 2009).
[0277] For all treatment groups 60 larvae were used in total and 20 larvae were assessed for survival and dissected every 24 hours. If the larvae were unresponsive to touch and changed colour from light brown to black they were considered dead. Similarly, for the control groups 60 larvae were used and 20 were scarified every 24 hours to determine CFU and PFU counts in their haemolymphs. Control larvae included in all experiments were: healthy larvae; larvae administered with 107 PFU phages via the oral route; larvae administered with 105 CFU S01160-12 via the oral route; and larvae administered PBS (Table 3). All controls were monitored for survival every 24 hours and their gastrointestinal tracts were dissected for CFU and PFU counts.
TABLE-US-00015 TABLE 3 Experimental setup to evaluate the effectiveness of phage therapy for the treatment of larvae infected with Salmonella. Groups.sup.1 Regimens.sup.2 Controls C-1 Uninfected C-2 PBS only C-3 Administered 2-phage cocktail only C-4 Administered 3-phage cocktail only C-5 Administered 4-phage cocktail only C-6 Salmonella only Prophylactic phage treatments P-1 Orally administered 2-phage cocktail one hour prior to infection P-2 Orally administered 3-phage cocktail one hour prior to infection P-3 Orally administered 4-phage cocktail one hour prior to infection Co-infection studies Col-1 Orally administered 2-phage cocktail simultaneously with Salmonella Col-2 Orally administered 3-phage cocktail simultaneously with Salmonella Col-3 Orally administered 4-phage cocktail simultaneously with Salmonella .sup.1in each group 60 larvae were used for the study. Every 24 hours 20 larvae were checked for survival and culled to enumerate PFU and CFU counts. .sup.2The phage cocktails were: 2-phages SPFM10-SPFM14; 3-phages SPFM4-SPFM10-SPFM19; and 4-phages SPFM2-SPFM10-SPFM14-SPFM19.
[0278] Survival Curves
[0279] For all experiment three biological replicates were conducted and survival data was plotted using Graphpad Prism version 6 (GraphPad Software Inc, USA) with the Kaplan-Meier method. The differences in survival rates were assessed using the Log-rank (Mantel-Cox) test.
[0280] Bioinformatics Analyses
[0281] Phages were previously deposited to ENA and their accession numbers are for SPFM2 (LR535921), SPFM4 (LR535902), SPFM10 (LR535908), SPFM14 (LR535912), SPFM17 (LR535914) and SPFM19 (LR535916). Annotated phage genomes can also be found via this link http://s3.climb.ac.uk/Sinfo/SPFM cenome.tar.gz. For SNP analysis frequency above 90% and with low p values was considered but when the frequency was lowered further SNP's were identified.
[0282] Results
[0283] Phage Burst Size
[0284] Phage growth curves were conducted for six SPFM phages. Phages SPFM10 and SPFM17 had large burst sizes of 175 and 162 PFU/per cell and short latent periods of 20 minutes. SPFM2, SPFM4, SPFM14 and SPFM19 had burst sizes of 165, 162, 162 and 153 PFU/per cell respectively and longer latent periods of 30 minutes (Table 4).
TABLE-US-00016 TABLE 4 Burst sizes of SPFM phages. Phage burst Salmonella Latent size per phage period.sup.1 bacterial cell.sup.1 SPFM2 30 165 SPFM4 30 162 SPFM10 20 175 SPFM14 30 162 SPFM17 20 169 SPFM19 30 153 .sup.1Average results are presented from three biological replicates, each with three technical repeats.
[0285] Temperature and pH Stability of Phages
[0286] To determine the sensitivity of phages SPFM2, SPFM4, SPFM10, SPFM14, SPFM17 and SPFM19 to heat, phages were exposed to temperatures 4, 10, 20, 30, 40, 50, 60, 70, 80 and 90° C. for 1 hour (
[0287] Spray Drying SPFM Phages
[0288] Prior to spray drying phages were concentrated using 50 ml Amicon Ultra-concentrators with a 10 kDa cutoff (Millipore, UK), 15 ml of phage lysate was added to the concentrators and spun at 5,000×g for 20 minutes. The flow-through was discarded and a further 15 ml of phage lysate was added to the concentrators. Concentrated phage collected above the filter was titred and if titres were approximately 10.sup.10 was used for spray-drying experiments. For phage spray drying a 200 ml volume of 4% trehalose (Oxoid, UK) excipient solution dissolved in ultra-pure water at pH 7.5 was mixed with 0.2 ml concentrated phage at an approximate titre of 5×10.sup.10 PFU/ml. The excipient-phage solution was spray dried using a laboratory scale LabPlant Spray Dryer (UK) with a two-fluid nozzle for atomisation with an orifice diameter of 0.5 mm. A constant feed rate of 3 ml/min was used for all runs with an atomising airflow of 6 l/min. The drying inlet temperature was heated to temperatures 80° C., 85° C., 90° C. or 100° C. and the corresponding outlet temperatures varied from 40° C. to 60° C. Dried powder phages were passed through the cyclone, collected in 100 ml glass bottles and stored at 4° C. till use. To determine phage titre 0.05 g of dried powder phage was suspended in 500 μl phage suspension buffer, diluted 10-fold and titered by plaque assays.
[0289] Temperature stability of phage at different inlet temperatures used for spray drying.
TABLE-US-00017 Inlet Outlet temperature temperature Phage titre loss (log.sub.10 PFU/g) (° C.) (° C.) SPFM2 SPFM4 SPFM10 SPFM14 SPFM17 SPFM19 80 −1.39 −3.01 −0.02 −0.35 −1.80 −1.49 85 −1.73 −3.16 −0.14 −0.81 −1.33 −1.33 90 −1.17 −3.15 −0.20 −1.07 −2.40 −2.16 100 −2.00 −3.15 −0.16 −0.75 −3.13 −3.01
[0290] Spray Drying SPFM Phages
[0291] All six phages were spray dried with the sugar excipient trehalose. During the spray drying process phages were exposed to thermal stress as heated air was used to dry the phage-trehalose liquid to powder. To evaporate the liquid, high inlet drying temperatures (can be manually set) were used, to which the phage-trehalose solution was only exposed to for milliseconds. Subsequently the powder was collected and exposed to a lower outlet temperature, which was dependent on the inlet temperature set. In this study phages were subjected to different inlet drying temperatures of 80, 85, 90 and 100° C. to optimise the drying process that resulted in minimal phage titre reduction.
[0292] Regardless of the inlet temperature used during the drying process, there was at least 1 log 10 PFU/g phage titre loss for SPFM2, SPFM4, SPFM17 and SPFM19. SPFM2 and SPFM4 were more stable when the inlet temperature was reduced to 80° C. but for SPFM4 at all inlet temperatures the phage titre loss was significant between 3.00 to 3.16 log 10 PFU/g. Phage titre loss of SPFM17 and SPFM19 were lowest with inlet temperature of 85° C. and 90° C. respectively. Phages SPFM10 and SPFM14 were the most stable from the six, with minimal phage titre loss of 0.02 and 0.35 log 10 PFU/g when 80° C. was used as the inlet temperature. Further titre losses were noted when the inlet temperature was increased for SPFM14 but with SPFM10 phage titre loss remained below 0.2 log 10 PFU/g. For all phages the final yield of powdered dried phages recovered was consistently low between 0.2-0.5 g and consequently phage recovery was between 5-13% for all inlet temperatures tested.
[0293] All six SPFM phages used in this study are 95% genetically similar to each other but yet phenotypically they are have different characteristics as observed from their individual and when combined killing activity against S. typhimurium in vitro and in vivo. To further investigate why phage combinations SPFM10-SPFM14; and SPFM4, SPFM10 and SPFM19 were the best, phenotypic analysis of all six phages was conducted by determining the temperature stability of phages. Consistent with previously characterised phages, SPFM2, SPFM4, SPFM14 and SPFM19 were stable up to 50° C. When phages are exposed to higher temperatures irreversible damage to phage proteins can occur, which leads to subsequent loss of phage infectivity. Interestingly despite this phages SPFM10 and SPFM17 are both heat stable and retained activity up to 70° C. Though with phage SPFM17 there was a significant reduction in phage titre after incubation at temperatures above 50° C. In comparison with phage SPFM10 there was no drop in phage titre up to 80° C. and even after being exposed to 90° C. for one hour over 55% of phage activity remained. Phage SPFM10 has increased thermal stability at extreme temperatures and significantly more stable than the reported E. coli phage vB_Eco4M-7 that only retained 20.5% of phage activity after being exposed to 95° C. for 5 minutes. It has been hypothesised that within the phage capsid the formation of disulphide cross-links could be involved in stabilising phages during extreme temperatures, which has been confirmed with RNA phage PP7. Whether disulphide cross-links in capsids of phages SPFM10 and SPFM17 are contributing to increased heat-stability is unknown.
[0294] Moreover it is likely there is a correlation between the heat stable of SPFM10 and its increased stability during spray drying. Phage SPFM10 is very resilient naturally and when combined with the excipient trehalose the phages still retains activity when exposed to the highest inlet temperature of 100° C. Surprisingly phage SPFM14, though not exceptional heat stable, was very stable during spray drying. In contrast, it would be predicted phage SPFM17 could withstand spray drying as it is naturally heat stable but there were up to 3 log 10 PFU/g reductions at the highest inlet temperature tested, similar to SPM2, SPFM4 and SPM19. This study highlights phage stability during spray drying is highly dependent on the phage used, even when the phages tested are genetically similar and are all myoviruses and thus have the same morphology. It was previously hypothesised myoviruses do not spray dry well due to their delicate virion structures as their long tails may contract or their capsids can separate under stress. This could explain why with phages SPM2, SPFM4, SPFM17 and SPM19 there was loss in titre, which was also observed for Staphyloccus phage Romulus for which a 2.5 log 10 reduction was noted. Exceptions to the hypothesis are jumbo phages SPFM10, SPFM14 and Pseudomonas phage KZ.
[0295] Phages were exposed to range of pH's to determine their sensitivity to pH (
[0296] Identifying Phage Receptor
[0297] Efficiency of plating was screened on different receptor mutant strains of SL1344 to determine the bacterial receptor the phages were using to bind to Salmonella (Table 5). The phages were tested on strains with deletions in the gene that encodes vitamin B12 uptake outer membrane protein (SL1344 AbutB) and deletions in the flagellar production gene (SL1344 ΔflgK). On both strains there were no significant changes in plaquing efficiencies when compared to wild type SL1344. However on strain SL1344 ΔrfaL, which has deletions in the gene involved in the LPS-related O-antigen production all phages were unable to infect and plaquing efficiencies were 0. Based on this data we predict all six phages are using the O antigen of LPS as the host receptor.
TABLE-US-00018 TABLE 5 Titres of Salmonella phages on wild-type and receptor mutant strains of SL1344. Phage titres (log.sub.10 PFU/ml) Wild-type SL1344 SL1344 SL1344 Phages SL1344 ΔbutB ΔflgK ΔrfaL SPFM2 8.91 8.87 8.80 0 SPFM4 8.20 8.14 8.27 0 SPFM10 8.55 8.40 8.59 0 SPFM14 8.60 8.74 8.56 0 SPFM17 8.42 8.36 8.42 0 SPFM19 8.50 8.49 8.56 0
[0298] Infection Dynamics of Six SPFM Phages In Vitro
[0299] Phages SPFM2, SPFM4, SPFM10, SPFM14, SPFM17 and SPFM19 alone and as two, three and four cocktail combinations were tested, to determine the optimal combination that caused the highest reduction of a multi-antibiotic resistant S. typhimurium strain. Killing was assessed over a six hour time-course and compared to the uninfected control. Data presented in
[0300] Bacterial Killing Assays with Single Phage Suspensions
[0301] At MOI 100 when phages SPFM2, SPFM4, SPFM10, SPFM14, SPFM17 and SPFM19 were added alone, they caused significant reductions 1 hour post infection (p≤0.05). However maximum bacterial cell lysis was observed 2 hours post-infection by reductions of 1.77, 2.36, 2.15, 2.43, 2.32 and 1.52 log 10 CFU/ml, respectively in comparison to the uninfected culture (
[0302] Bacterial Killing Assays with Two Phage Cocktails
[0303] Fifteen different two phage cocktails were tested and for all combinations there were reductions of S. typhimurium by 1 and 3 log 10 CFU/ml 1 hour post-infection in comparison to the uninfected culture (
[0304] Bacterial Killing Assays with Three Phage Cocktails
[0305] In comparison to single phages and two phage cocktails, the eleven three phage combinations tested cleared bacterial liquid cultures more rapidly 1 hour post-infection (p≤0.01), by approximately 2.5 log 10 CFU/ml (
[0306] Bacterial Killing Assays with Four Phage Cocktails
[0307] Rapid bacterial lysis that caused 1.5-2.5 log 10 CFU/ml reduction in S. typhimurium was observed 1 hour post-infection for all four, four phage cocktails (p≤0.01) tested. From the four, three combinations: SPFM10-SPFM14-SPFM17-SPFM19, SPFM2-SPFM10-SPFM14-SPFM19 and SPFM2-SPFM4-SPFM10-SPFM19 were the most effective at reducing bacterial counts 2 hour post-infection by 3.51, 3.40 and 3.38 log 10 CFU/ml, respectively (
[0308] Infection Dynamics of Phage Cocktails In Vivo
[0309] To determine if the efficacy results observed in vitro translate in vivo, efficacy of phage cocktails, SPFM10 and SPFM14 (two-phage); SPFM4, SPFM10 and SPFM19 (three-phage); SPFM2, SPFM10, SPFM14 and SPFM19 (four-phage) were tested in a Galleria mellonella infection model. Larvae were infected with the MDR S. typhimurium S01160-12 strain. The activity of prophylactic treatment of phage cocktails and phages administered simultaneously to Salmonella were assessed (
[0310] Prophylactic Treatment of Phage Cocktails
[0311] Phage cocktails were administered prophylactically, 1 hour prior to infection to determine their efficacy in treating Salmonella infection (
[0312] Co-Infection of Phage Cocktails
[0313] Phage cocktails and Salmonella were simultaneously orally administered to larvae and monitored every 24 hours over 72 hours (
[0314] Control Larvae Groups
[0315] Control larvae groups C-1, C-2, C-3, C-4 and C-5 had survival rates of 100% after 72 hours and no Salmonella was detected in these groups. Phages were only recovered from groups C-3, C-4 and C-5 and phage counts were conducted every 24 hours for 72 hours (
[0316] Sensitivity of Recovered Salmonella Colonies after Exposure to Phage Cocktails in a Larvae Model
[0317] 168 Salmonella colonies were recovered from larvae after 72 hours, 56 colonies after being exposed to the 2-phage cocktail, 56 for 3-phage cocktail and 56 for 4-phage cocktail. These colonies were tested for their susceptibility to individual phages within their respective cocktails. All colonies remained sensitive to the individual phages within the cocktail that they were exposed to, which was based on whether a clear zone of lysis was observed. 10 Salmonella colonies from each phage treatment were selected and the efficiency of plating of the individual phages within the cocktails was determined and compared to the wild-type S. typhimurium S01160-12 strain (
[0318] Differences in plaquing efficiencies were observed for some screened colonies, such as for 3/10 colonies isolated after being exposed to the 2-phage cocktail, the plaquing efficiency of SPFM14 was reduced by 10-19%. In contrast the efficiency of SPFM10 on 4/10 colonies had improved by 10-26%. For colonies isolated after being exposed to the 3-phage cocktail, SPFM4 and SPFM10 on 3/10 colonies the plaquing efficiencies of SPFM19 was reduced by ˜15%. However, on 4/10 colonies the plaquing efficiencies of SPFM4 and SPFM10 had increased by over 10%. On colonies exposed to the 4-phage cocktail activity of SPFM10 and SPFM19 was reduced by ˜15% on 3/10 colonies but SPFM14 had higher plaquing efficiency consistently on all 10 colonies.
[0319] Genetic Differences Between the Six SPFM Phages
[0320] The Salmonella phages isolated were 95% similar to each other based on average nucleotide identity. Single nucleotide polymorphisms (SNP) analysis was conducted and the heat-stable phage SPFM10 was compared to SPFM2, SPFM4, SPFM14, SPFM17 and SPFM19. SNP data is presented in Table 6 and overall very few SNP's were identified between the phages. The consistent SNP's that were identified across all five phages was at position 70,890, part of the gene annotated as hypothetical protein. This was the only SNP difference identified between phages SPFM10 and SPFM14. In contrast when SPFM10 was compared to SPFM2, SPFM4, SPFM17 and SPFM19 a total of 3, 3, 5 and 2 SNP's in phage genes were identified respectively. All SNP's identified were in hypothetical proteins, expect for a SNP difference between SPFM10 and SPFM17 at genome position 57985 annotated as chromosome partition protein Smc. Furthermore the phages were compared to SPN3US and SEGD1, two jumbo phages. Approximately over 6900 SNP's were identified, and at least one SNP was present in all annotated genes.
TABLE-US-00019 TABLE 6 Phage SPFM10 was compared to SPFM2, SPFM4, SPFM14, SPFM17 and SPFM19 to identify SNP's between the phages. Salmonella Position in VarFreq phage the genome Protein annotation (%) P value SPFM2 70890 Hypothetical protein 100 1.05E−38 74294 Hypothetical protein 100 1.70E−43 742295 Hypothetical protein 100 4.28E−44 SPFM4 70890 Hypothetical protein 100 2.20E−144 74291 Hypothetical protein 99.75 1.34E−238 74295 Hypothetical protein 100 3.22E−247 SPFM14 70890 Hypothetical protein 100 5.14E−16 SPFM17 57985 Chromosome partition 100 1.73E−49 protein Smc 70890 Hypothetical protein 100 2.77E−66 74291 Hypothetical protein 100 6.66E−94 74295 Hypothetical protein 100 1.06E−92 113292 Hypothetical protein 100 1.10E−65 SPFM19 70890 Hypothetical protein 98.86 1.55E−50 72949 Hypothetical protein 100 1.74E−52
[0321] Salmonella spp. is a major animal and human food safety concern and due to the rise in infections caused by MDR Salmonella strains alternatives to antibiotics are needed to tackle the infection. Phages could provide an alternative. Twenty-one, jumbo phages were identified that could kill representative strains of the prevalent serovars associated with UK pigs. Six phages from the collection, were selected based on their broad host ranges, as identified herein and as deposited as discussed herein. All six phages had large burst sizes and short latent periods, which are one of the characteristics of virulent phages that would be ideal for therapy. A large burst size, also increases the probability of contact between phages and their target pathogen, which is essential for infection. Furthermore, the large burst size could mean phages are eliminating bacteria faster than they can replicate, which could potentially reduce risk of phage-resistant bacteria being selected. Other jumbo phages have been isolated that have similar short latent periods of under 40 minutes and large burst sizes of over 140 phages per infected cell, which includes Klebseilla phage vB_KleM-RaK2 (Simoliunas et al. 2013) and Serratia phage MAMA1 (Matilla and Salmond 2014).
[0322] All six phage are 95% genetically similar to each other and bind to same bacterial receptor but the pH and temperature stability data highlighted that there are phenotypic differences between the phages. The pH stability data showed significant reductions in phage titres at pH 3 for three out of the six phages. Furthermore, although all phages were stable at 50° C., consistent with other characterised phages, only phages SPFM10 and SPFM17 retained activity after being exposed to 70° C. However, phage SPFM17 showed significant reductions in phage titre after incubation at temperatures above 50° C. In comparison with phage SPFM10 there was no drop in phage titre up to 80° C. and even after being exposed to 90° C. for one hour and over 55% of phage activity remained. This is very unusual as typically when phages are exposed to high temperatures it can lead to irreversible damage to phage proteins, which subsequently leads to loss of phage infectivity (Ahmadi et al. 2017). The natural thermal stability of SPFM10 is rare and is significantly more stable than the reported E. coli phage vB_Eco4M-7 that only retained 20.5% of phage activity after being exposed to 95° C. for 5 minutes (Jurczak-Kurek et al. 2016). Within the phage capsid, the formation of disulphide cross-links could be involved in stabilising phages during extreme temperatures. Disulphide cross-links in capsids of phages SPFM10 and SPFM17 may be contributing to increased heat-stability.
[0323] Ideal phage or phage cocktails for use in therapy, are those which are effective at eliminating multiple subgroups of the target bacterial species and are able to delay occurrence of phage-resistant mutants. SPFM phage combinations: as single, two, three and four-phages were tested in vitro and it was determined which combination could rapidly lyse a representative S. typhimurium MDR isolate (Agency 2014). All phages when tested individually caused a 1 log 10 CFU/ml reduction in bacterial numbers after just 1 hour. Re-growth of Salmonella was immediately evident after 1 hour and by 6 hours Salmonella counts of phage infected cultures were similar to uninfected controls, with no significant differences between either. These results are consistent with previous killing assay conducted with single phages against Salmonella.
[0324] Phage cocktails of two, three and four phage combinations were tested and bacterial lysis was improved to approximately 3 to 4 log 10 CFU/ml reductions after 2 hours and re-growth of S. typhimurium was less prominent. In total thirty-six different phage combinations were tested in vitro. The phage cocktails SPFM10-SPFM14 and SPFM2-SPFM10 were the most effective at lysing S. typhimurium and would be ideal candidates to be used in phage therapy. Similarly in previous studies three (Bardina et al. 2012), six (Albino et al. 2014) and ten (Zhang et al. 2010) phage cocktails could reduce Salmonella numbers by approximately 2 log 10 CFU/ml. This is unlike the 4 log 10 CFU/ml reduction observed with the best two phage cocktails in the present study. The in vitro data highlights that two jumbo phage cocktails are effective at quickly lysing Salmonella and both combinations were able to lyse 100% of the representative isolates of the top UK pig associated serotypes screened. Interestingly both cocktails also include the natural heat stable phage SPFM10, which could possibly indicate it is more virulent than the other five phage. Possible explanations to why bacterial killing was not improved with three or four phage cocktails could be that as phages were mixed by equal volumes, phages were being diluted to achieve the final MOI, and with that phage virulence could have been diluted. Furthermore, as the phages are competing for the same receptor site on the Salmonella cells, one phage could have outcompeted the others in the cocktail. The phage that outcompeted the others could be one which reproduces more quickly (has a short latent period and large burst size), and as the other phages in the cocktail are being outcompeted it would reduce overall efficacy of the cocktail. Assessing different phage combinations is important to determine the optimal mix at reducing the target pathogen.
[0325] The efficacy of phage cocktails to clear an MDR Salmonella strain was tested in the in vivo larvae infection model to compare phage efficiency in vitro and in vivo. Three cocktails were tested: SPFM10-SPFM14; SPFM4-SPFM10-SPFM19; and SPFM2-SPFM10-SPFM14-SPFM19 and were administered prophylactically and co-infected. Similar to the in vitro data, phage cocktail SPFM10-SPFM14 was the best at clearing infection via both treatments, it significantly improved survival of larvae and after 72 hours Salmonella was not recovered from over 65% of larvae. The results provide further evidence that this cocktail could be ideal for therapy. Similar efficacy of phage cocktails to clear their target bacteria has been observed in other larvae studies, such as a six phage cocktail administered at an MOI of 100, 2-hours prior to infection with Pseudomonas aeruginosa was able to improve survival of larvae by 80% after 24 hours (Beeton et al. 2015). Another independent study showed a different six cocktail improved survival of larvae by 30% when infected with a virulent P. aeruginosa strain (Forti et al. 2018). Despite the lower efficacy of the three and four phage cocktail in vitro, both cocktails improved survival of larvae and Salmonella clearance but the three-phage cocktail was more efficient when co-infected. The results highlight in vitro killing data does not consistently correlate in vivo and screening efficacy of phage cocktails in larvae infection models could be an improved method of assessing the virulence of phage cocktails (Seed and Dennis 2009; Nale, Chutia, et al. 2016; Abbasifar et al. 2014; Manohar, Nachimuthu, and Lopes 2018). To our knowledge this is the first study to assess activity of jumbo phages against MDR S. typhimurium in a larvae infection model and the first study to use jumbo phages.
[0326] Colonies isolated after being exposed to the three different phage cocktails in the larvae infection model remained sensitive to the individual phages with the cocktail, which shows we did not isolate any phage resistant Salmonella cells. Furthermore, phages still had high plaquing efficiencies on the recovered colonies and could infect by the same degree when compared to the wild-type strain.
[0327] Due to the phenotypic differences between the six SPFM phages, SNP analysis was conducted in identify the genetic differences between the phages. When the heat stable phage SPFM10 was compared to the other five SPFM phages, there are very few SNPS's between them. Across all five phages there was a SNP in a hypothetical protein at genome position 70890 and the SNP is present in a protein that is under positive selection and naturally susceptible to mutation. Further analysis with HHpred predicted that the gene has hits to C-terminal pectate lyase domain (98% probability) and is likely involved in phage tail fibre formation. The pectate lyase domain function is predicted to be to cleave glycoside bonds and could be involved in degrading polysaccharide on the surface of Salmonella, therefore could play an important role in the phage host range. Furthermore, this SNP could be playing a role in heat stability as the only difference between SPFM10 and SPFM14, and the latter phage is only stable at 50° C. vs heat stability of SPFM10 till 90° C. Further SNP's between SPFM10 and phages SPFM2, SPFM4, SPFM17 and SPFM19 were identified in another predicted pectate lyase protein, which was also shown to be under positive selection. The protein is likely to be a host interacting gene, it must be prone to mutation. Only in SPFM17 a further two SNP's were identified in the chromosome partition protein and in a hypothetical protein when it was compared to SPFM10. SPFM17 is also resistant to high temperatures till 70° C. and the addition of these two SNP's could potentially be playing a role in stabilising as they are unique to the phage. Again further mutagenesis analysis is needed to determine how important the presence of both SNP's are in heat stability.
[0328] The work described below describes the infectivity of phages against the top serotypes associated with UK chickens.
[0329] Experimental Procedures
[0330] Bacterial Strains and Growth Conditions
[0331] In total 18 Salmonella strains were examined in this study. All Salmonella enterica subsp. enterica strains, were isolated by the Animal and Plant health Agency (APHA) in Weybridge, UK from chickens in the UK between 2015 to 2017 (Table 7). From these strains the serotypes were: 3 S. 13,23:i:-; 5 S. enteritidis; 4 S. infantis; 3 S. Ohio and 3 S. Seftenberg.
TABLE-US-00020 TABLE 7 Chicken Salmonella strains. Strain S/L-Number number Serovar (APHA label) Farm 1 S. 13,23:i:- S03363-15 Hatchery 2 S. 13,23:i:- S03360-15 Hatchery 3 S. 13,23:i:- S03362-15 Hatchery 4 S. Enteritidis S03579-16 Broiler 5 S. Enteritidis S02733-17 Hatchery 6 S. Enteritidis S03467-16 Layer hen 7 S. Enteritidis S02455-17 Turkey 8 S. Enteritidis S02481-17 Turkey 9 S. Infantis S00794-17 Broiler 10 S. Infantis S02954-16 Broiler 11 S. Infantis S03726-16 Broiler 12 S. Infantis S04700-15 Layer hen 13 S. Ohio S01770-17 Broiler 14 S. Ohio S01146-17 Broiler 15 S. Ohio L01476-17 Broiler 16 S. Seftenberg L01512-17 Hatchery 17 S. Seftenberg L01508-17 Hatchery 18 S. Seftenberg S02722-17 Broiler
[0332] Phage Isolation, Purification and Propagation
[0333] Increased volumes of purified phage lysates were made by mixing exponential growing liquid cultures of SL1344 (10.sup.7 CFU/ml) of Salmonella infected with 10.sup.7 PFU/ml phages in NZCYM broth at 37° C. with shaking (100 rpm) for 6 hours. Phage cultures were centrifuged at 4,200×g for 15 minutes, the supernatant was filtered with 0.22 μm pore size filters and phage lysates were stored at 4° C. To determine phage titre, phage lysate was serially diluted 10-fold and the small drop plaque assay method (Mazzocco et al., 2009) was used on LB 1% agar plates. Final phage titres were expressed as PFU/ml.
[0334] Phage Host Range Analysis and Efficiency of Plating
[0335] The host range of individual phages was determined by the small drop plaque assay method (Mazzocco et al., 2009) on different Salmonella enterica subsp. enterica serotypes and incubated for 18 hours at 37° C. Plates were examined for either bacterial lysis via clearing or plaques or for no infection and average observations were noted from three biological replicates, each with three technical repeats.
[0336] Host Range Analysis
[0337] The efficacy of SPFM2, 4, 10, 14, 17 and 19 phages were tested on 18 Salmonella enterica subsp. enterica strains isolated from chickens. All 18 strains are representatives of the top UK pig associated Salmonella serotypes, namely S. 13,23:i:-, S. enteritidis, S. infantis, S. Ohio and S. Seftenberg. All phages could infect all S. enteritidis strains. Phage SPFM2, SPFM14 and SPFM19 produced turbid clearing on all the S. infantis strains and could not infect representative strains of serotype S. Seftenberg. Phages SPFM4 and SPFM17 could lyse all S. infantis strains and could only lyse 1/3 of the S. Seftenberg strains.
TABLE-US-00021 TABLE 8 Host range of SPFM phages on Salmonella chicken isolates and data presented is the average of three replicates. The strains numbers relate to the numbers listed in Table 7. 0 1 2 No Turbid Complete infection clearing lysis Strains 1 2 3 4 5 6 7 8 9 SPFM2 0 1 0 2 2 2 2 2 1 SPFM4 1 0 0 2 2 2 2 2 2 SPFM10 1 1 1 2 2 2 2 2 2 SPFM14 0 0 0 2 2 2 2 2 1 SPFM17 1 0 0 2 2 2 2 2 2 SPFM19 0 0 0 2 2 2 2 2 1 Strains 10 11 12 13 14 15 16 17 18 SPFM2 1 1 1 0 2 1 0 0 0 SPFM4 2 2 2 1 2 2 0 0 2 SPFM10 2 1 2 2 2 2 0 0 0 SPFM14 1 1 1 0 2 1 0 0 0 SPFM17 2 2 2 2 2 2 0 0 2 SPFM19 1 1 1 1 2 1 0 0 0
[0338] Although the invention has been particularly shown and described with reference to particular examples, it will be understood by those skilled in the art that various changes in the form and details may be made therein without departing from the scope of the present invention.