This invention relates to bacteriophages. In particular, it relates use of bacteriophages for lysis of Salmonella bacteria. It also relates to compositions comprising bacteriophages for lysis of Salmonella bacteria.

The biocontrol of bacterial pathogens via the application of virulent bacteriophages (phages) has gained increasing credibility as an alternative to traditional antibiotic therapies. Phages are ubiquitous in the biosphere (an estimated 10 ^{3 1} particles), which places them as the most abundant biological entity on Earth. By outnumbering their bacterial counterparts by 10: 1 in a diverse range of environments, phages and their often virulent lifecycles are implicated in destroying half of the global bacterial population every 48 hours. Phage therapy has been proposed as a potential solution to deal with the problems posed by the increasing number of multidrug-resistant (MDR) bacterial pathogens. MDR bacteria can enter the human food chain from the use of antibiotics in farm animals, therefore in the EU, the use of antibiotics has been limited to therapeutic applications, and their use as growth promoters is banned. However, even in the absence of the selective pressure of antibiotics, swine reared in antibiotic-free production systems are continually exposed to persistent MDR S. Typhimurium. Bacteriophages offer the prospect of a sustainable alternative antimicrobial treatment against such pathogens since they are compatible with food use, with the flexibility that they can be applied therapeutically or for biosanitization purposes. In the USA, 'generally recognized as safe' (GRAS) status has been granted for the use of a number of phage products as biosanitization agents on ready-to-eat foods. Under EU legislation, phages are under consideration as 'processing aids' (Directive 89/107/EEC) and could be applied during the manufacturing process providing any treatment residues do not have any technological effect on the finished product. However, the responsibility for safety in this case lies with the manufacturer as Regulation (EC) No. 178/2002 states it is their responsibility to ensure the final product is safe for human consumption Virulent bacteriophages are natural predators of their bacterial hosts that complete their lifecycle by lysis of the infected bacterium, and it is these that are utilized in a therapeutic context. This is in contrast to temperate bacteriophages that can form a stable genetic relationship with the host during the process of lysogeny. The culmination of lysogeny is the integration of the bacteriophage genome into that of the host creating a stable genetic element known as a prophage. These undesirable traits are not associated with virulent bacteriophages which actively replicate at the expense of the bacterial population. Therefore, as long as the prevailing environmental conditions permit active infection/replication cycles, then bacterial numbers should decline whilst the phage population increases. Although bacterial resistance to infection is a well- documented phenomenon associated with phage predation, the use of a number of different phages in combination - a phage cocktail - can overcome this. Phage cocktails not only potentially provide a means to circumvent resistance to a single phage they also allow the treatment of multiple pathogens simultaneously.

Phage intervention strategies have been used to control various Salmonella serovars including Enteritidis and Typhimurium, with experiments highlighting their potential use for biosanitization and for phage therapy of infected animals. These studies have shown that phages can be effectively utilized against these pathogens. Most recently, a 2-3 logio CFU reduction of S. Typhimurium γ4232 was achieved following application of a phage cocktail designed to reduce S. Typhimurium γ4232 levels in artificially-infected market weight swine. An earlier study involving broiler chickens reported reductions of S. Typhimurium 4/74 (>2. 19 logio CFU reduction) and S . Enteritidis P 125109 (>4.2 logio CFU reduction) following phage application (Atterbury et al. , 2007).

Salmonella enterica serovar Typhimurium U288 is a MDR pathogen of livestock and has consistently been identified as the most prevalent serovar on UK pig production premises (VLA, 2009 ^a ). Also, several deaths have been documented following an S . Typhimurium U288 outbreak in elderly patients in Denmark during 2008. The antibiotic resistance profile of S . Typhimurium U288 covers a wide spectrum of the classes that are currently utilized by human and veterinary medicine. A core resistance to ampicillin (Am), chloramphenicol (C), streptomycin (S), sulphonamides (SU), tetracycline (T) and trimethoprim (TM) - AmCSSuTTm was observed in 76 % of isolates submitted to the Veterinary Laboratory Agency in the UK in 2008 (VLA, 2009 ^b ), and as such represents a reservoir of antibiotic resistance within pig production units. The resistance profile of S. Typhimurium U288 is similar to that of the significant human pathogen S. Typhimurium DTI 04 (AmCSSuT), which has been a major global cause for concern since its emergence a few decades ago.

It would be advantageous to provide new compositions comprising bacteriophages for lysing Salmonella serotypes Enteritidis and Typhimurium.

According to a first aspect the present invention provides an isolated bacteriophage selected from the group consisting of: OSH17 (deposit number NCIMB 42029 Salmonella typhimurium bacteriophage OSH17), OSH18 (deposit number NCIMB 42030 Salmonella typhimurium bacteriophage OSH18) and OSH19 (deposit number NCIMB 42031 Salmonella typhimurium bacteriophage OSH19) or a variant of one of said bacteriophages, wherein the variant retains the phenotypic characteristics of the parent bacteriophage.

The isolated variant bacteriophage may be a variant of any one of the parent bacteriophages OSH17, OSH18 and OSH19 or felix 01. The variant bacteriophage may have at least 70 %>, at least 80%, at least 90%>, at least 95%>, at least 98%o or at least 99%> nucleotide sequence identity to one of the parent bacteriophages OSH17, OSH18, OSH19 or felix 01. Variants may include genetically modified versions of the deposited phages in which the genetic code is manipulated by means of, for example, genetic engineering or serial passage.

The bacteriophage may have lytic activity against one or more Salmonella strains. The bacteriophage may have lytic activity against Salmonella Enteritidis and/or Salmonella Typhimurium, for example Salmonella typhimurium U288 (deposited as NCIMB 42028 Salmonella typhimurium U288) or a wild type Salmonella strain (deposited as NCIMB 42027 Salmonella typhimurium wild type (Rawlings)). The term lytic activity may mean having activity against a target bacteria to lyse it. The lytic activity of a bacteriophage may be determined by standard techniques, for example by plaque assay. Phage titer may be determined as plaque forming units (pfu) The bacteriophage may have lytic activity against one or more Salmonella strains selected from the strains listed in table 1. The bacteriophage may have lytic activity against at least one, at least two, at least three, at least four, at least five, at least six, at least seven, at least eight, at least nine, at least ten, at least eleven or at least twelve Salmonella strains selected from the strains in table 1. The bacteriophage may have lytic activity against at least Salmonella. Typhimurium U288. The isolated bacteriophage may have lytic activity against one or more multi-drug resistant Salmonella strains, preferably Salmonella. Typhimurium U288. The term isolated may be considered to mean material that is removed from its original environment in which it naturally occurs. The purified isolated bacteriophage does not contain significant amounts of other bacteriophages.

The present invention further provides a bacteriophage cocktail comprising at least two different bacteriophages selected from OSH 17, OSH 18, OSH19 and felix 01 ).

Salmonella typhimurium wild type (Rawlings) is deposited as NCIMB 42027 Salmonella typhimurium U288 is deposited as NCIMB 42028

Salmonella typhimurium bacteriophage OSH17 is deposited as NCIMB 42029 Salmonella typhimurium bacteriophage OSH18 is deposited as NCIMB 42030 Salmonella typhimurium bacteriophage OSH19 is deposited as NCIMB 42031 Felix 01 , a member of the Myoviridae, was originally isolated by Felix and Callow (1943), British Medical Journal, Jul 31 ;2 (4308): 127-30.

The bacteriophage cocktail may comprise at least three different bacteriophages selected from OSH17, OSH18, OSH19 and felix 01. Bacteriophage cocktail may mean a combination comprising two phages of the invention, or variants thereof, each of which has been isolated from the environment from which they were originally found or have been produced by means or a technical process such as genetic engineering or serial passage techniques.

The bacteriophage cocktail may comprise at least four different bacteriophages selected from OSH17, OSH18, OSH 19 and felix 01. The bacteriophage cocktail may comprise two bacteriophages selected from: OSH 17, OSH18, OSH19 and felix 01.

The bacteriophage cocktail may comprise three bacteriophages selected from: OSH17, OSH18, OSH19 and felix 01

The bacteriophage cocktail may comprise OSH17, OSH 18, OSH19 and felix 01.

The present invention further provides a pharmaceutical composition comprising an isolated bacteriophage as described herein and a pharmaceutically acceptable carrier.

A pharmaceutical composition may mean a composition comprising a therapeutically effective amount of a phage, and a pharmaceutically acceptable carrier or diluent. The pharmaceutical composition may also comprise further ingredients. Pharmaceutical carriers may be sterile liquids, such as water, oils, saline solutions, buffers and solutions comprising dextrose or glycerol. Pharmaceutical excipients may comprise adjuvants. The composition may be in the form of a dry powder or water free concentrate.

An effective amount may be a dose of bacteriophage which provides the desired effect in the patient. The dose may be determined by one skilled in the art by the use of known techniques.

The present invention further provides a pharmaceutical composition comprising a phage cocktail as described herein and a pharmaceutically acceptable carrier. The present invention further provides an isolated bacteriophage as described herein or a bacteriophage cocktail as described herein for use in the treatment or prevention of Salmonella infection in an animal, preferably a mammal. The mammal may be a domestic animal, an animal intended for food production, for example the animal may be a pig, a sheep, a cow, a goat. The animal may be a fish or a shellfish.

The present invention further provides a food or feed ingredient comprising an isolated bacteriophage according to the present invention or a bacteriophage cocktail according to the present invention, wherein the food or feed ingredient is able to cause lysis of Salmonella on or in a food or feed. A food or feed ingredient may be added to food or feed during manufacturing or may be added to feed just before consumption by a human or an animal. A food or feed ingredient may reduce the number of Salmonella bacteria in the food or feed. The food ingredient may reduce the number of Salmonella bacteria in the food or feed to zero or about zero.

The present invention provides a disinfectant composition comprising an isolated bacteriophage according the present invention or a bacteriophage cocktail according to the present invention. The disinfectant composition may be used to disinfect or decontaminate an animal before slaughter. This is advantageous because it reduces the likelihood of meat made from the animal being contaminated with Salmonella. The disinfectant composition may be suitable for oral administration to a human or an animal. The disinfectant composition may be suitable for administration onto the skin or hair of an animal or a human.

It is advantageous to disinfect the skin or hair of an animal before slaughter to reduce the risk of meat made from the animal being contaminated with Salmonella. An animal may be disinfected by administering a bacteriophage or bacteriophage cocktail orally or onto the skin before slaughter, for example about 5, about 10 or about 30 minutes before slaughter, or about 1 , about 2, about 5, about 12, about 36 or about 48 hours before slaughter of the animal. The present invention provides a use of a bacteriophage according to the present invention or a bacteriophage cocktail according to the present invention as a biological control agent. A biological control agent may be any agent that can kill or lyse Salmonella bacteria. The biological control agent may be formulated using suitable ingredients depending on what it is intended to be applied to. The biological control agent may also comprise ingredients to enhance the replication of the bacteriophages.

The present invention provides the use of a bacteriophage according to the present invention or a bacteriophage cocktail according to the present invention to reduce the number of viable Salmonella bacteria on or in an animal before or at the time of slaughter of the animal.

The present invention also provides the use of a bacteriophage according to the present invention or a bacteriophage cocktail according to the present invention to reduce the number of viable Salmonella bacteria on or in a food product or a feed product or on an animal carcass. The bacteriophage or bacteriophage cocktail may be within a food or feed product or on the surface of a food or feed product. The bacteriophage or bacteriophage cocktail may be mixed with a food or feed. The bacteriophage or bacteriophage cocktail may be applied to the outside of a food or feed product, for example by spraying, wiping or soaking.

The present invention also provides the use of a bacteriophage according to the present invention or a bacteriophage cocktail according to the present invention to reduce the number of viable Salmonella bacteria on or in an environment.

The environment may be any environment likely to harbour Salmonella bacteria. The environment may be a closed or an open environment. The environment may be, for example a building or a piece of equipment. The environment may, for example be an abattoir; a food or feed processing facility; food or feed processing equipment; food or feed storage equipment; a food or feed display area; a food or feed preparation area; or medical equipment. The present invention provides a method for lysing Salmonella bacteria comprising the step of contacting the Salmonella bacteria with a bacteriophage according to the present invention or a bacteriophage cocktail according to the present invenion.

The present invention provides a method for reducing the number of Salmonella bacteria comprising contacting the Salmonella bacteria with an isolated bacteriophage according to the present invenion or a bacteriophage cocktail according to the present invention and keeping the bacteriophage under conditions whereby they are able to replicate and lyse the bacteria..

There now follows by way of example only a detailed description of the present invention with reference to the accompanying drawings, in which; Figure 1 shows TEM images of OSH17 (A) at 87,000 x magnification -

Bar = 200 nm, OSH18 (B) at 160,000 x magnification - Bar = 100 nm, and OSH19 (C & D) at 135,000 x magnification - Bar = 100 nm.

Figure 2 shows Mean logl O recoverable S. Typhimurium U288 CFU/cm2 (■) and PCI PFU/cm2 (♦) (istandard deviation) following storage at 4°C (A) and -20°C (B).;

Multidrug-resistant Salmonella Typhimurium U288 is a significant pathogen of pigs, accounting for over half of all outbreaks on UK pig production premises. The potential of this serovar, and other Salmonella^, to enter the food chain during the slaughtering process requires that efforts be made to reduce the prevalence of these bacteria at both the pre- and post-harvest stages of production. A bacteriophage cocktail (PC I) capable of lysing various S. enterica serovars was designed using the broad host- range phage Felix 01 , and three phages isolated from sewage. PC I applied to pig skin experimentally-contaminated with U288 achieved significant reductions (P < 0.05) in Salmonella counts when stored at 4°C over 96 hours. Reductions of >1 log ₁₀ unit were observed when the ratio of phage applied was in excess of the bacterial concentration. The treatment was found to be effective at a multiplicity of infection (MOI) of 10 or above, with no significant reductions taking place when the MOI was less than 10. Under these conditions U288 counts of logio 4.1 - 4.3 CFU were reduced to undetectable levels following the application of PC I to pig skin (>99% reduction). These data suggest phage cocktails could be employed post- slaughter as a means to reduce Salmonella contamination of pig carcasses.

This study involves the application of a phage cocktail targeted to reduce S. Typhimurium U288 levels on artificially- contaminated pig skin. The S. Typhimurium U288 strain was identified during screening of a pig production farm known to be contaminated with Salmonella (H. Davies, Personal communication). The phage cocktail comprises four distinct anti- Salmonella phage - OSH17, OSH18, OSH19, and Felix 01.

Materials & methods

2.1 S. Typhimurium U288 stock

S. Typhimurium U288 was cultured and maintained on solid media containing NZCYM broth (Fisher Scientific, UK) supplemented with 1.2 % [wt/vol] Bacteriological Agar No. l (Oxoid, UK). Working stock plates were produced by streaking out a loopful of S. Typhimurium U288 15 % glycerol stock (stored at -80°C) onto NZCYM agar. All working stock plates were inverted and incubated overnight at 37°C, prior to storage at 4°C. For the selection and enumeration of S. Typhimurium U288, XLD agar (Oxoid, UK) containing 50 μg/mL kanamycin (Fisher Scientific, UK) was used throughout.

2.2 Phage isolation

OSH 17, OSH18, and OSH19 were identified during screening of sewage effluent using S. Typhimurium WT (Rawlings) and S. Typhimurium DT I 04 (WT) as hosts. For phage isolation, sewage effluent was filtered through 0.2 μιη Minisart filters (Sartorius Biotech, Germany) and the filtrate collected in sterile universals and stored at 4°C until required. For phage extraction from solid matter such as pig faeces, samples were used to create 10 % (wt/vol) solutions in SM buffer. To dissociate phage, the solution was incubated over night at 4°C with shaking and then processed as described above. For lawn preparation, Salmonella 10 mL NZCYM broth cultures were prepared and incubated overnight at 37°C with shaking. Following this, 50 iL of overnight culture was used to seed fresh 10 mL NZCYM broth (10 mM MgS0 ₄ ) which was then incubated for 2 hours at 37°C with shaking. To 5 mL molten (tempered to ~ 50°C) NZCYM top agar (0.6 % Bacteriological Agar No. l) 500 μΙ_, of the required Salmonella was added, followed by 500 μΙ_, filtrate, and the mixture was poured onto NZCYM agar plates. The plates were left to set on the bench for 20 minutes before being inverted then incubated overnight at 37°C. Any plaques identified were picked using sterile pipette tips and resuspended in 500 iL SM buffer (50 mM Tris-HCl [pH7.5], 100 mM NaCl, 8 mM MgS0 ₄ :7H ₂ 0, 0.01 % gelatin, pH 7.5), incubated at 37°C for 1 hour, then serial diluted in SM buffer. A 25 volume of each dilution was then added to Salmonella/NZCYM top agar, and lawns were prepared as described above. This process was repeated at least three times for each phage in order to obtain single clonal isolates.

Novel phage isolates were then propagated to obtain high titre stocks as described by Kocharunchitt et al., (2009), International Journal of Food Microbiology 128, 453-459. Briefly, NZCYM broth cultures of the required propagating host were prepared as described above. Following overnight incubation, 250 of culture was added to 25 mL NZCYM broth (10 mM MgS0 ₄ ) in conical flasks. To each flask 25 μί of the required phage was added, and flasks were then incubated overnight at 37°C with shaking. Following this, the lysate was drawn up using sterile syringes then passaged through 0.2 μιη Minisart filters. Aliquots of each phage stock were then 10-fold serial diluted in SM buffer, followed by spotting of triplicate 10 μΐ, drops of each dilution on Salmonella lawns prepared as described above. Plates were enumerated the following day. All phage stocks were stored at 4°C until required.

The efficacy of each phage isolate against a wide panel of Salmonella serovars was determined as follows. Salmonella top agar lawns were prepared as described above. To each Salmonella/NZCYM top agar lawn, 20 μΐ, volumes of logio7 PFU/mL dilutions of each purified phage stock were applied. This approach delivered a routine test dilution of 10 ⁵ PFU per spot. Following a sufficient drying period, plates were inverted and incubated overnight at 37°C. The following day, the degree of lysis of each phage was observed and the data obtained is presented in Table 1. The Salmonella^ used during host-range determination encompass isolates from human and animal disease outbreaks, as well as NCTC strains available for purchase. However, of note are the significant human (S. Typhimurium and S. Enteritidis serovars) and pig pathogens (S. Derby, S. Kedougou, and S. Typhimurium).

Table 1. Lytic spectrum of FelixOl, OSH17, OSH18, and OSH19 on Salmonellae

Phage

Salmonella FelixO l OSH 17 OSH 18 OSH l

S. Agama — (+) (+) —

S. Amina — — — —

S. Amsterdam — — — —

S. Atlanta NCTC 9986 — — — —

S. Burielly NCTC 8745 +++ — — —

S. Derby WT +++ (++) — —

S. Enteritidis SA025 PT4 +++ +++ +++ (++)

S. Enteritidis SA029 +++ +++ +++ (++)

S. Enteritidis WT (Harrison) — +++ +++ (++)

S. Enteritidis WT (Hood) — +++ +++ (++)

S. Enteritidis WT (Platten) — +++ +++ (++)

S. Hadar WT — — — —

S. Infantis NCTC 6903 — — — —

S. Kedougou ΒΡΦ (+++) — — —

S. Kedougou PI — — — —

S. Kubacha WT — — — (++)

S. Montevideo NCTC 5747 — — — —

S. Montevideo WT — — — —

S. Muenster/Orion — — — —

S. Senftenburg WT — — — —

S. Thompson NCTC 2252 — — — —

S. Tobga WT — — — —

S. Typhimurium DT I 04 +++ (+++) (++) (++)

S. Typhimurium LT2 +++ (++) (++) +++

S. Typhimurium WT (Rawlings) +++ +++ +++ (++)

S. Typhimurium WT (Turner) — (++) (++) —

S. Typhimurium U288 (++) (++) (++) (+++)

S. Virchow WT _ +++ - confluent lysis, ++ - semi- confluent lysis, + - individual plaques, ( ) - opalescent lysis

2.3 Phage cocktail 1 (PCI)

High titre stocks of OSH17, OSH18, OSH19, and Felix 01 were prepared and titrated as described in Section 2.2. The purified high titre phage stocks were subsequently used to make a 20 mL phage cocktail (PC I) in SM buffer with a combined titre of 10 ⁸ PFU/mL. Any subsequent dilutions of PC I were made in SM buffer and stored at 4°C until required.

2.4 Sizing of phage genomes

Pulse field gel electrophoresis (PFGE) was used to determine phage genome sizes. Briefly, 1 mL of phage stock (109 - 101 1 PFU/mL) was centrifuged for 2 hours at 34,900 x g using a JA18.1 rotor and Beckman J2-21 centrifuge. Following this the supernatant was carefully decanted and the pellet resuspended in 50 μΐ _^ phage stock. To each tube, 10 μΐ _^ of 10 mg/mL proteinase K (Sigma Aldrich, UK) was added followed by mixing with an equal volume of molten 1.2 % PFGE grade agarose (Biorad, USA) in TE (10 mM Tris-HCl [pH 7.5], 1 mM EDTA) buffer. The suspension was thoroughly mixed before being dispensed into PFGE plug moulds (Biorad, USA) and allowed to solidify. Each plug was then incubated in 1 mL lysis buffer (50 mM Tris-HCl [pH 8], 50 mM EDTA, 1 % N-lauroyl sarcosine, 100 μg/mL proteinase K) at 55°C overnight. Plugs were then washed 3 times for fifteen minutes in 1 mL wash buffer (20 mM Tris-HCl [pH 8], 50 mM EDTA) at 55°C. A 3 mm slice from each plug was loaded onto a 1 % PFGE agarose gel (100 mL 1 x TAE buffer) along with 0.1-200 Kb and 50- 1000 Kb DNA pulse markers (Sigma Aldrich, UK), and the lanes sealed with molten 1 % PFGE grade agarose. A CHEF-DR II system was used to run the gel with a switch time of 10-30 seconds over 17 hours at 6 V/cm, with 1 x TAE as the running buffer (circulating at 14°C). After electrophoresis the gel was stained for several hours with ethidium bromide (1 μg/mL) and visualized with a Biorad Image Capture (Biorad, USA). A standard curve of migration of marker DNA was used to approximately calculate the genome size of each phage. 2.5 Transmission electron microscopy (TEM) of phage

Freshly-prepared high titre phage suspensions of each phage were sedimented at 34,900 x g for 2 hours (4°C). Following centrifugation, the supernatant was decanted and each phage pellet was washed twice with 0.1M ammonium acetate for 1 hour at 25,000 x g. The wash solution was discarded and 2 mL SM buffer added to each centrifuge tube. Phage pellets were recovered following overnight incubation at 4°C with gentle shaking. A small drop of washed phage suspension was spotted onto a carbon-coated copper mesh grid and allowed to sit for 3 minutes. Excess phage suspension was then removed with filter paper. For negative staining one drop of phosphotungstic acid [pH 7.4] was added to each grid, and excess stain was removed one minute later with filter paper. Each grid was then covered and allowed to dry for 15 minutes. Images were taken with a Fei Tecnai Biotwin TEM (Fei Company, USA). 2.6 Porcine sample preparation

Skin sections from each 25 day-old pig (7.5 to 8 Kg) carcasses were removed post-mortem using an ethanol flame-sterilized scalpel. Large sections of pig skin were then cut into approximately 4 cm ² pieces which were subsequently placed in sterile 95 mm Petri dishes (Sarstedt, Germany). A SALMOTYPE ^® Pig Screen ELISA (Labor Diagnostik, Germany) was performed as per manufacturer's instructions to confirm that the pigs were serologically Salmonella negative at slaughter.

2.7 Survival and recovery of S. Typhimurium U288 on artificially- contaminated pig skin stored under fresh (4°C) and frozen (-20°C) storage conditions

To determine the effect of fresh (4°C) and frozen (-20°C) storage conditions on the survival and recovery of S. Typhimurium U288 on artificially-contaminated pig skin, a small-scale sampling regime was prepared. Briefly, triplicate 4 cm ² pieces of pig skin were inoculated with logio 5.4 CFU/cm ² S. Typhimurium U288 and the inoculum was uniformly spread over each sample using sterile disposable spreaders. A drying period of 1 hour was allowed before the first set of triplicate pieces of pig skin were added to 20 mL volumes of MRD (Oxoid, UK) in sterile filter stomacher bags (Seward, UK). At this point all remaining samples were stored under the appropriate storage conditions (4°C or -20°C) until required. Samples were subjected to stomaching for 3 mins at 300 rpm using a Seward Lab Blender 400. Serial dilutions of the stomachate were then performed using MRD, prior to spread plating triplicate 100 volumes of each dilution onto XLD agar (50 μg/mL kanamycin). Following overnight incubation at 37°C, mean levels of S. Typhimurium U288 recovered were recorded and standard deviations calculated. This process was repeated every 24 hours for the duration of the experiment. The data obtained was used to plot a graph (Figures 2A & 2B) of mean logio CFU/cm ² recovered (± SD) of S. Typhimurium U288 over time, for fresh and frozen storage conditions. 2.8 Survival and recovery of PCI on artificially-contaminated pig skin stored under fresh (4°C) and frozen (-20°C) storage conditions

A series of 4 cm ² pieces of pig skin were inoculated with logio 6.4 PFU/cm ² of PCI which was then spread out evenly using sterile disposable spreaders. A drying period of 1 hour was allowed before the samples were transferred to the appropriate storage conditions (4°C or -20°C). To recover PCI from pig skin sections, samples were placed in sterile filter stomacher bags containing 20 mL SM buffer prior to stomaching for 1 min at 300 rpm (Seward Lab Blender 400). One mL of stomachate was removed and filtered through 0.2 μιη Minisart filters, and the filtrate collected in a sterile tube. The filtrate was then serial diluted in SM buffer and plaque assays were performed as follows (adapted from Santos et al, 2009). Molten 5 mL NZCYM/5 % glycerol top agar (0.6 % Bacteriological Agar No. l) tempered to ~ 50°C was inoculated with 100 μL S. Typhimurium U288/NZCYM broth overnight culture. Diluted filtrate (100 μL) was then added to the top agar and the mixture poured onto NZCYM/5 % glycerol plates. Each plate was rotated gently to ensure an even bacterial lawn, allowed to dry on the bench for 20 minutes, then inverted and incubated overnight at 37°C. Following overnight incubation, recoverable mean PFU (± SD) were calculated for the triplicate samples. This process was repeated every 24 hours for the duration of the experiment. The data obtained was used to plot a graph (Figures 2A & 2B) of mean logio PFU/cm ² recovered (± SD) of PCI over time, for fresh and frozen storage conditions. 2.9 Bacteriophage-mediated biosanitization of artificially-contaminated pig skin

A 3 x 3 matrix (Table 2) of PFU against CFU was designed to cover a range of multiplicities of infection (MOIs). MOI is used to describe the ratio of phage to bacteria and in this instance ratios ranging from 0.01 to 10,000 were employed. The matrix was then used to analyze the effects of administering various phage titres (10 ⁷ , 10 ⁵ , and 10 ⁴ PFU) to a range of S. Typhimurium U288 concentrations (10 ⁶ , 10 ⁴ , and 10 ³ CFU). Sampling took place at three separate time points over a 96 hour period (1 hour, 42 hours, and 96 hours) with the initial sampling taking place 1 hour post-inoculation. Triplicate 4 cm ² pig skin samples were prepared as described in Sections 2.7 and 2.8 with S. Typhimurium U288 being administered to the samples first, followed by PCI 30 minutes later. Control samples of non phage-treated S. Typhimurium U288 (10 ⁶ , 10 ⁴ , 10 ³ CFU) were also prepared to allow comparisons with the phage-treated samples. Following phage inoculation a 30 minute drying period was allowed prior to samples being placed in storage at 4°C. Skin sections were then processed to determine the number of S. Typhimurium U288 in each sample, as described above. The data obtained was logio transformed and analyzed using a one-tailed t test from the data analysis kit in Microsoft Excel 2007. 3. Results

3.1 Selection and characterization of bacteriophage

A number of novel phage isolates were tested to identify potential candidates for use as biosanitization agents against Salmonella. The ability of each bacteriophage to lyse a panel of Salmonella in vitro allowed the selection of three phages - OSH17, OSH18, and OSH19 (Table 1). These phage displayed broad activity against a range of Salmonella, and along with Felix 01 (noted for it's anti-Salmonella properties) were selected for use in a phage cocktail (PC I). All three OSH phage were found to be capable of lysing a number of different serovars: < SH17- 12/28, < SH18 - 1 1/28, and OSH19 - 1 1/28. The genome sizes of the component bacteriophage of PC I were then estimated by PFGE: OSH 17, OSH18, and OSH 19 were found to be ~ 40 Kb, ~ 48.5, and ~ 155 Kb, respectively, whilst an estimate of 86.1 Kb for Felix 01 was consistent with the sequenced genome of 86, 155 bp (Whichard et al., 2010). TEM images of OSH17 and OSH18 (Figure 1) revealed a siphovirus type structure (icosahedral heads and flexible non-contractile tails), whereas OSH19 had structure indicative of the Myoviridae family (icosahedral head with contractile tail) that includes T4 and the T4-like phages. Phage isolates not selected for this study were discarded on the following criteria: the inability to clear plaque on relevant hosts; deviations from the efficiency of plating using the routine test dilution of 10 ⁷ PFU/mL.

3.2 Recovery of S. Typhimurium U288 and PCI from pig skin stored at 4°C and -20°C

The survival and recovery of S. Typhimurium U288 on experimentally- contaminated pig skin sections stored at 4°C and -20°C was monitored over a 9-day period. S. Typhimurium U288 recovery from an initial inoculum of logio 5.4 CFU/cm ² was followed by a minor decrease in recoverable U288 on day 2-3, and thereafter remained relatively stable throughout the remainder of the experiment (Figures 2A and B). Evidently any S. Typhimurium U288 that remain post-slaughter or that re-contaminate the skin surface will remain a source of contamination throughout retail, and therefore a potential hazard to the consumer. The stability and viability of the bacteriophage preparation was also monitored over a 9-day period at 4°C and -20°C. From an initial titre of logio 6.4 PFU/cm ² , the phage fell to logio 4.2 PFU/cm ² one hour post-inoculation, and thereafter remained stable with recoverable phage titres ranging between logio 4.1 - 4.4 PFU/cm ² . 3.3 Bacteriophage biosanitization of pig skin

PCI was applied to a range of S. Typhimurium U288 concentrations on pig skin at various MOIs (Table 2), and monitored over a 96-hour period at 4°C. The data obtained was used to analyze reductions on phage-treated skin sections, as compared with untreated S. Typhimurium U288-only controls. S. Typhimurium U288 could not be detected below 10 ² CFU without prior enrichment. Therefore, XLD (50 μg/mL kanamycin) plates containing no recoverable S. Typhimurium U288 during sampling were assumed to be <2 logio CFU. Table 3 shows the S. Typhimurium U288 counts recovered throughout the duration of the experiment with control inoculums ± standard deviations.

Table 2. A 3 x 3 matrix used to create a range of MOIs to monitor the therapeutic effect of PCI on 4 cm ² pig skins artificially-contaminated with S.

Typhimurium U288.

S. Typhimurium Phage inoculum (PFU)

U288 inoculum & MOI

(CFU) 10 ' 10= 10 ⁴

10 ^b 10 0.1 0.01

10 ⁴ 1000 10 1

10 ³ 10,000 100 10

Table 3. Mean logio CFU counts (±standard deviation) of S. Typhimurium U288 recovered from experimentally-contaminated 4 cm ² pig skin sections of control and PCI treated samples (ND=not detectable; *P>0.01 and **P>0.001 show significant differences compared to control values).

The application of PC I produced significant reductions 1 h post- inoculation, with the largest reduction recorded the combination of 10 ⁷ PFU PC I and 10 ⁴ CFU S. Typhimurium U288 (MOI 1000). This combination resulted in a 1.2 logio CFU reduction (logio 3.5 ± 0.1 CFU) when compared with untreated controls (logio 4.7 ± 0.2 CFU). In this instance a 92.6 % reduction in recoverable S. Typhimurium U288 CFU (P < 0.01) was observed from the sampled pieces of pig skin. A 1.0 logio CFU reduction resulted from an MOI of 10 (10 ⁵ PFU/10 ⁴ CFU) reducing S. Typhimurium U288 numbers from logio 4.7 ± 0.2 CFU (untreated controls) to 3.7 ± 0.2 CFU recovered (phage-treated samples) that corresponds with a 92.1 % reduction of recoverable S. Typhimurium U288 CFU (P < 0.01). By comparison MOIs of 100 and 1000 with 10 ³ CFU inoculums produced reductions of 92.1 % and 91.9 % respectively in the recoverable S. Typhimurium U288 counts (P < 0.05). No reductions were observed on any of the samples with an MOI of less than 10.

At day 3 , treatment with an MOI of 10 (10 ⁷ PFU/10 ⁶ CFU) resulted in a logio 1.3 CFU reduction in recoverable S. Typhimurium U288 counts (logio 5.0 ± 0.3 CFU) compared with untreated controls (logio 6.3 ± 0.1 CFU) that is equivalent to a 92.3 % reduction (P < 0.001). The application of PC I at an MOI of 1000 (10 ⁷ PFU/10 ⁴ CFU) resulted in a reduction of logio 1.4 CFU (96.1 %) from the control count of logio CFU 4.3 ± 0.1 (P < 0.01). Notably the combination of 10 ⁴ PFU and 10 ³ CFU inoculated on pig skin resulted in no detectable S. Typhimurium U288 on any of the triplicate samples, implying a >logio 2 CFU reduction compared to the control value of logio CFU 4.0 ± 0.2. As with day 1 , no significant reductions were observed when MOIs of less than 10 were used.

Day 5 had a number of reductions of logio 1.0 CFU or greater. No S. Typhimurium U288 could be recovered from two combinations with initial inoculums of 10 ³ CFU and PC I corresponding with MOIs of 10 and 100. The bacterial counts on these skin sections fell below the level of detection implying reductions >99 % compared with the untreated controls (logio CFU 4.3 ± 0.2). Consistent with the results of the previous time point only the sample containing 10 ⁷ PFU produced a significant reduction (P>0.001) with the 10 ⁶ CFU S. Typhimurium U288 inoculums, which corresponds with an MOI of 10. All samples treated with an MOI of less than 10 showed no significant reductions in S. Typhimurium U288 counts recovered compared to controls.

4. Discussion

The data obtained from this phage therapy trial provides a proof of principle that the application of a suitable phage cocktail (PC I ) can reduce levels of S. Typhimurium U288 (the most prevalent serovar found in pigs) on artificially-contaminated pig skins. A number of factors were identified during the trials which may be of significance during future studies. The use of MOIs in excess of the bacterial concentration appears to be of great relevance to the outcome of the treatment. In each instance where a significant reduction in S. Typhimurium U288 was observed, PC I was administered in excesses ranging from 10 - 10,000. Reductions to below detectable levels were observed on a number of occasions when the MOI was between 10 and 100 and the initial S. Typhimurium U288 inoculum was 10 ³ CFU. PC I shows greater efficacy against low levels of S. Typhimurium U288 contamination, as might be anticipated for S. Typhimurium contaminations of post process pig carcasses. Moreover the application of phage preparations like PC I would complement biosecurity measures targeted to reduce the exposure of the consumer and the health of pigs in meat production. Pigs can be susceptible to low levels of Salmonella exposure, for example, it is reported that >10 ³ S. Typhimurium HL 10969 cells were required to induce acute salmonellosis infection in pigs.

In summary, the application of a phage cocktail produced significant reductions of S. Typhimurium U288 on experimentally-contaminated pig skin; however this appears to be linked to an MOI in excess of the target bacterium. When applied at an MOI of 10 or above, PC I is capable of reducing S. Typhimurium U288 up to 2 logio units over the course of 96 hours. Little or no reductions were observed when PC I was applied at an MOI of 1 or less. This suggests that when PC I is used as a decontamination agent the therapeutic effect observed is passive. Under these conditions the initial phage dose is sufficiently in excess of the target bacterium population to cause reductions without the need for the bacteriophage to replicate and complete their life cycle. In contrast active therapy involves phage infection/replication cycles to reduce the target bacterium. One of the major advantages of passive therapy is a reduction in the likelihood of the development of bacterial resistance. Over the course of active therapy, initial low numbers of resistant bacteria may rise in abundance eventually replacing the original susceptible population. This process has been well-documented during in vitro and in vivo studies. However, the 'single-hit' approach utilized here greatly limits opportunities for the target bacterium to develop resistance. Due to the commercial storage conditions adopted in this study (4°C), the growth of Salmonella are greatly reduced or halted, which will either prevent or at least slow phage replication. The low temperature required for meat storage does not impede the passive action of the bacteriophage. This study has set the ground work for in vivo trials to reduce S. Typhimurium U288 levels in pigs and their production environments via the application of bacteriophage as therapeutic and biosanitization agents. Of the bacteriophage examined here OSH19 and felixOl have broad spectrum activities, and conform to the virulent phage families that are preferred for commercial application. Further in vitro phage therapy work will involve the application of bacteriophage to reduce levels of S. Typhimurium U288 and S. Typhimurium DTI 04 on various foods and fomites.