Craig Billington

Bacteriophages as biocontrol agents in food

Bacteriophages possess attributes that appear to be attractive to those searching for novel ways to control foodborne pathogens and spoilage organisms. These phages have a history of safe use, can be highly host specific, and replicate in the presence of a host. Campylobacter, Salmonella, and Listeria monocytogenes and various spoilage organisms have responded to phage control on some foods. However, the use of phages as biocontrol agents is complicated by factors such as an apparent requirement for a threshold level of host before replication can proceed and by suboptimal performance, at best, at temperatures beneath the optimum for the host. This review is a summary of the information on these issues and includes brief descriptions of alternative phage-based strategies for control of foodborne pathogens.

Phage inactivation of foodborne pathogens on cooked and raw meat

Phages infecting Salmonella Typhimurium PT160 and Campylobacter jejuni were added at a low or high (10 or 10(4)) multiplicity of infection (MOI) to either low or high (<100 or 10(4)cm(-2)) densities of host bacteria inoculated onto raw and cooked beef, and incubated at 5 and 24 degrees C to simulate refrigerated and room temperature storage. Counts of host bacteria were made throughout the incubation period, with phages being counted at the first and last sampling times. Host inactivation was variable and depended on the incubation conditions and food type. Significant host inactivations of the order of 2-3 log(10)cm(-2) at 5 degrees C and >5.9 log(10)cm(-2) at 24 degrees C were achieved compared to phage-free controls using the Salmonella phage under optimal conditions (high host cell density and MOI). These results alongside those already published indicate that phages may be useful in the control for foodborne pathogens.

Influence of host and bacteriophage concentrations on the inactivation of food-borne pathogenic bacteria by two phages.

To use bacteriophages (phages) to control food-borne pathogenic bacteria, it will be necessary to determine the conditions allowing optimal activity. To start exploring these conditions, a Salmonella phage (P7) and a Campylobacter phage (Cj6) were incubated with their respective hosts at 24 degrees C for up to 2 h at varying phage and host cell concentrations, and surviving host cells were enumerated. A quadratic polynomial equation was fitted to the inactivation data and contour maps of inactivation against log(10) phage and host concentrations were plotted. Inactivation of Salmonella by P7 seemed to be independent of the host concentration, with close to 100% inactivation occurring at a phage concentration of around 5 x 10(8) PFU mL(-1). For Campylobacter phage Cj6, there appeared to be an interaction of both phage and host concentrations. The data obtained were largely consistent with prior work indicating that, at low host cell concentrations, the proportion of cells killed is independent of the host cell concentration. The data indicate that biocontrol of pathogens present in low numbers in liquid foods is achievable, given a sufficiently high concentration of added phages, and that it is not necessary to know the concentration of pathogens to achieve this.

Control of Listeria monocytogenes growth in a ready-to-eat poultry product using a bacteriophage.

A bacteriophage (phage) that infected strains of the species Listeria monocytogenes as well as Listeria ivanovii and Listeria welshimeri, but not Listeria grayi or Listeria innocua, was isolated from sheep faeces. The phage had a contractile tail and an icosohedral head indicating that it was a myovirus, and was morphologically similar to phage A511. At 30 °C, phages added at 5.2 × 10⁷ PFU ml⁻¹ prevented the growth in broth of L. monocytogenes present at approximately twice this concentration for 7 h, but re-growth occurred such that the concentration after 24 h incubation was similar in both control and phage-treated cultures. At the same temperature, but on the surface of vacuum-packed ready-to-eat chicken breast roll, there was an immediate 2.5 log₁₀ CFU cm⁻² reduction in pathogen concentration following addition of phages and then re-growth. However, at a temperature reflecting that at which a chilled food might be held (5 °C), this re-growth was prevented over 21 days incubation. The data suggest a dose-dependent rapid reduction in pathogen concentration followed by no continued phage-mediated effect. These results, alongside other published data, indicate that a high concentration of phages per unit area is required to ensure significant inactivation of target pathogens on food surfaces.

Use of a bacteriophage to inactivate Escherichia coli O157:H7 on beef.

A number of outbreaks of Escherichia coli O157:H7 infections involving beef have been reported. Options for controlling bacterial pathogens in raw foods are limited, but one is to use bacteriophages (phages). We describe the isolation and characterisation of phage FAHEc1, which infects E. coli O157, and its ability to kill its host in vitro and on beef. The phage belonged to the family Myoviridae and lysed 28 of 30 E. coli O157 (:H7, :HNM and :H not specified) isolates, only one other non-O157 E. coli serotype (O162:H7), and none of the other 13 bacterial species tested. The phage did not contain stx1, stx2, eae or ehxA virulence genes as assessed by PCR. An approximate 4 log₁₀ inactivation of E. coli O157:H7 occurred at 5 °C in the presence of phage FAHEc1 at >10⁷ PFU/ml in broth in vitro. On thinly sliced beef pieces incubated at 37 °C, a > 2.7 log₁₀ reduction occurred with 3.2 × 10⁷ PFU/4 cm² meat piece. At lower phage concentrations (10³-10⁴ PFU/4 cm² piece) phage replication occurred on beef at 37 °C. When the phage was applied to beef pieces under conditions simulating hot boning and conventional carcass cooling, inactivation of E. coli O157:H7 of approximately 2 log₁₀ was measured under optimal conditions with phages applied at 3.2 × 10⁷ PFU/4 cm² meat piece.

Isolation and lytic activity of the Listeria bacteriophage endolysin LysZ5 against Listeria monocytogenes in soya milk.

The endolysin gene (lysZ5) from the genome of the Listeria monocytogenes phage FWLLm3 was cloned in Escherichia coli and characterized. Comparative sequence analysis revealed that lysZ5 resembled the murein hydrolase ply511 encoded by L. monocytogenes phage A511. The encoded protein LysZ5 had a predicted molecular mass of 35.8 kDa and was expressed in E. coli as an N-terminal fusion protein of 41.5 kDa. Addition of purified fusion protein to lawns of indicator bacteria showed that LysZ5 could lyse L. monocytogenes, Listeria innocua and Listeria welshimeri, but not Staphylococcus aureus or Enterococcus faecalis. The purified protein was able to kill L. monocytogenes growing in soya milk, with the pathogen concentration reduced by more than 4 log₁₀ CFU ml⁻¹ after 3 h incubation at 4 °C. As far as we know, this is the first report of a Listeria phage endolysin to control pathogens in soya milk and to demonstrate endolysin activity in foods at refrigeration temperatures. Moreover, LysZ5 may also be useful for biocontrol in other ready-to-eat foods.

Isolation and characterization of phages infecting Bacillus cereus

Aim:  To isolate and characterize bacteriophages (phages) that infect the foodborne pathogen Bacillus cereus.

Effect of phage and host concentration on the inactivation of Escherichia coli O157:H7 on cooked and raw beef

A previously described phage infecting Escherichia coli O157:H7 was added to raw and cooked beef pieces at concentrations ranging from 101–108 plaque forming units/cm2 to either low (<100 CFU/cm2) or high (104 CFU/cm2) concentrations of host bacterial cells. Incubation for up to 24 h was performed at 5℃ and 24℃ to simulate refrigerated and room temperature storage/temperature abuse. Surviving bacteria were enumerated during the incubation period, with phages being counted at the first and last sampling times. Significant reductions of E. coli O157:H7 of the order of >4 log10 CFU/cm2 at both temperatures could be achieved compared to phage-free controls. There was a trend for greater inactivation to occur with increasing phage concentration. While re-growth of surviving cells occurred in nearly all samples incubated for 24 h at 24℃, these conditions are not typical of those experienced by perishable foods. It was concluded that phages can be used to reduce the concentration of a bacterial pathogen on meat, but the concentration of phages needs to be high (>4–5 log10 plaque forming units/cm2) for reductions to occur. A concentration of the order 8 log10 plaque forming units/cm2 was needed to achieve a 4 log10 CFU/cm2 reduction.

Listeriaphages and coagulin C23 act synergistically to kill Listeria monocytogenes in milk under refrigeration conditions

Bacteriophages and bacteriocins are promising biocontrol tools in food. In this work, two Listeria bacteriophages, FWLLm1 and FWLLm3, were assessed in combination with the bacteriocin coagulin C23 to inhibit Listeria monocytogenes. Preliminary results under laboratory conditions demonstrated that both antimicrobials act synergistically when they were applied in suboptimal concentrations. The combined approach was further assessed in milk contaminated with 5×10(4) CFU/ml L. monocytogenes 2000/47 and stored at 4 °C for 10 days. When used alone, phage FWLLm1 added at 5×10(6) PFU/ml, FWLLm3 at 5×10(5) PFU/ml and coagulin C23 at 584 AU/ml kept L. monocytogenes 2000/47 counts lower than the untreated control throughout storage. However, when used in combination, inhibition was enhanced and in the presence of FWLLm1 and coagulin C23, L. monocytogenes 2000/47 counts were under the detection limits (less than 10 CFU/ml) from day 4 until the end of the experiment. Resistant mutants towards phages and coagulin C23 could be obtained, but cross-resistance was not detected. Mutants resistant to FWLLm3 and coagulin C23 were also recovered from surviving colonies after cold storage in milk which may explain the failure of this combination to inhibit L. monocytogenes. Remarkably, the fraction of resistant mutants isolated from the combined treatment was lower than that from each antimicrobial alone, suggesting that synergy between bacteriocins and phages could be due to a lower rate of resistance development and the absence of cross-resistance.

Inhibition of Listeria monocytogenes by Enterococcus mundtii isolated from soil.

Two bacterial isolates with inhibitory activity against Listeria monocytogenes and Enterococcus faecalis were obtained from soil. Genotypic and phenotypic characterization identified them as Enterococcus mundtii, a species whose ability to compete with L. monocytogenes is relatively unexplored compared to other members of the genus. The thermal stability of the inhibitory factor and its sensitivity to proteolytic enzymes indicate that it is most likely a bacteriocin. Both isolates grew at comparable rates to L. monocytogenes at 5 °C and 10 °C in vitro. One isolate killed L. monocytogenes when it reached concentrations of 10(6)-10(8) CFU ml(-1). Minimum inocula of 10(6) and 10(5) CFU ml(-1) of E. mundtii were required to reduce and maintain L. monocytogenes concentrations beneath the level of detection at 5 °C and 10 °C, respectively. In situ experiments at 5 °C showed that E. mundtii inhibited the growth of L. monocytogenes on vacuum-packed cold smoked salmon during its four week shelf life. E. mundtii could, therefore, control the growth of L. monocytogenes at low temperatures, indicating a potential application in controlling this pathogen in chilled foods. To control growth of Listeria, the concentration of E. mundtii needs to be high, but it is possible that a purified bacteriocin could be used to achieve the same effect.