According to the Centers for Disease, Control and Prevention (CDC)
[9], foodborne infections produced by chicken, beef, pork and turkey are associated with
Campylobacter and
Salmonella presence meanwhile dairy products like raw milk and cheese are commonly infected by
Campylobacter,
Salmonella,
Escherichia and
Listeria. Most common bacteria infecting vegetables and fruits are
Salmonella,
Escherichia and
Listeria and most frequent pathogen producing foodborne illnesses in fishes and shellfishes are
Vibrio and
Salmonella.
2.1. Bacteriophages to Control Salmonella enterica
One in three foodborne outbreaks in the European Union in 2018 were caused by
Salmonella, being salmonellosis the second most commonly reported gastrointestinal infection in humans (91,857 cases reported) after campylobacteriosis (246,571)
[10].
Thung et al. studied the interaction of the bacteriophage SE07, isolated from retail meat samples, against
S. enterica serovar Enteritidis on different food matrices, such as fruit juice, fresh egg, beef and chicken meat. The reduction of the bacteria population in all of them was significant at 12 h (2.05 log CFU/mL, 1.98 log CFU/mL, 1.79 log CFU/mL, and 1.83 log CFU/mL, respectively), and after that time there was no further significant reduction
[11]. In 2018, Phongtang et al. evaluated the effect of P22 phage (ATCC 97541) against
S. enterica serovar Typhimurium in milk. This phage showed an inhibitory effect of more than 3 log UFC/mL reduction after 4 h
[12]. Bao et al. tested two lytic phages, vB_SenM-PA13076 (PA13076) and vB_SenM-PC2184 (PC2184), in chicken breast, pasteurized milk and Chinese cabbage. Phages were isolated from chicken sewage and infected
S. enterica serovar Enteritidis. PA13076 was able to infect 222 strains (71.4%) and PC2184 infected 298 strains (95.8%) out of 311 isolates tested. The two phages were rapidly inactivated at temperatures above 60 °C (PA13076) or 70 °C (PC2184). Interestingly, PA13076 reduced
Salmonella population in chicken breast, pasteurized milk and Chinese cabbage by 2 log, 2 log and 2.5 log UFC/mL, respectively, whereas PC2184 reduced bacteria population in chicken breast, pasteurized milk and Chinese cabbage by 3 log, 4 log and 3.5 log UFC/mL, respectively
[13].
The company Micreos Food Safety has developed the brand Phageguard S (
Table 1) based on phages Felix-O1a and S16 against
Salmonella enterica. This product is able to kill all
Salmonella serovars including those that are resistant to antibiotics and the 20 most virulent
Salmonella strains according to the United States Department of Agriculture (USDA). Phageguard S can reduce bacterial population by 1–3 log CFU/mL without affecting taste, odor or texture of foods. It is effective from 0 to 35 °C and its use is recommended as final treatment in spray or directly immersing food into the phage solution
[14]. Yeh et al. reported that the combination of phages S16 and Felix-O1a reduced
Salmonella on ground beef and pork by 1 and 0.8 log CFU/g, respectively
[15]. A recent study tested Phageguard S on lean pork, bacon and pork trims with good results. The product administration decreased
Salmonella population by 0.8–1.7 log CFU/cm
2 or g using 5 × 10
7 PFU/cm
2 or g of phages
[16].
Table 1. List of approved and commercially available bacteriophage products.
Company |
Phage Product |
Pathogen |
Micreos Food Safety (The Netherlands) |
PhageGuard Listex |
Listeria sp. |
PhageGuard S |
Salmonella enterica |
PhageGuard E |
Escherichia coli O157:H7 |
Intralytix (USA) |
ListShield |
Listeria monocytogenes |
SalmoFresh |
Salmonella enterica |
ShigaShield |
Shigella sp. |
EcoShield PX |
Escherichia coli |
Arm & Hammer (USA) |
Finalyse SAL |
Salmonella enterica |
Finalyse |
Recently, a bacteriophage cocktail was used against
Pseudomonas aeruginosa that produces rhinosinusitis in sheep. A mix of four phages was able to reduce biofilm biomass on frontal sinus mucosa at concentrations of 10
8–10
10 PFU/mL with no safety concerns
[42].
Several murine mastitis models have showed that phage therapy could be also used against
Staphylococcus aureus in bovine mastitis caused by microbial infection
[43][44][45]. A previous study published in 2006 by Gill et al. analyzed the efficacy of a 5-day treatment consisting of phage K administered intramammary in lactating Holstein cows with subclinical mastitis caused by
S. aureus. Three out of 18 animals were cured (16.7%) compared to none out of 20 cows of the negative control group (0%)
[46]. Despite some success, the low efficacy could be explained by the data of Gill et al. showing that incubation of
S. aureus with whey and bovine serum resulted in inhibition of phage K lysis. Accordingly, they concluded that proteins could block sterically the phage K attachment to the bacteria, suggesting that
S. aureus could be more resistant to phages in vivo in mastitis infections than in vitro experiments
[47].
Infections caused by
E. coli O157:H7 and treatment with phage therapy in ruminants have been already reviewed
[48], revealing that further understanding of phage administration, effective multiplicity of infection (MOI) and correct analysis of results are necessary in cattle phage therapy
[49][50][51]. In sheep, no significant reductions of
E. coli O157:H7 were found compared to controls when a single phage was administered after oral
E. coli inoculation
[52][53]. However, a mix of two phages reduced more than 99% the presence of
E. coli in the lower intestinal tracts of treated animals
[54]. In addition, a cocktail of eight phages reduced significantly fecal
E. coli O157:H7, although not in the rumen, after 24 h post phage administration
[55].
In piglet studies, phages were able to kill methicillin-resistant
S. aureus (MRSA) in vitro but no reduction was observed in the nasal mucosa in vivo or ex vivo
[56]. This fact emphasizes the importance of considering other factors that may counteract phage efficacy in vivo, such as reduced adherence or increased clearance by the animal fluids. However, experiments conducted in growing pigs showed that dietary supplementation with a commercial cocktail of phages against
Salmonella enterica,
S. aureus,
E. coli and
Clostridium prefringens was more efficient than probiotics as growth promoters
[57], improving food digestibility, daily weight gain and gain per feed, among other parameters.
The presence of wounds is relatively common in swine. An hydrogel containing phages against
Acinetobacter baumanii was used to reduce wound infections in an ex vivo model of pig skin, and achieved a 90% reduction in bacterial counts after only 4 h of treatment
[58].
Another study showed that seven phages isolated from pig farms in the United Kingdom were able to lyse all 68
Salmonella strains tested, including MDR ones, offering a valuable alternative to antimicrobials to reduce infections and food poisoning
[59].
Another recent review
[60] summarized the known phages infecting
Paenibacillus larvae. This spore-forming bacterium attacks honeybee larvae causing the American foulbrood, which is the most widespread and destructive of the honeybee brood diseases, being able to destroy an entire colony in just three weeks. Importantly, all known bacteriophages against
P. larvae to date are lysogenic. Despite that, studies of phage therapy in vitro and in hives have shown higher survival rates of treated groups including prophylactic benefits. Lack of success in some cases was attributed to the lysogenic nature of the phages or their inability to access the gut.
In aquaculture, the common carp has been used as a model to demonstrate the effectiveness of phage therapy against
Citrobacter freundii, using a single phage, IME-JL8. This bacterium belongs to the normal flora of fishes; however, it has been associated to systemic infection in common carp and other diseases in diverse fishes. Administration of phages into the carp decreased pro-inflammatory cytokines and protected the fish from infection when phages were administered one hour after bacteria inoculation, but not after 24 h, indicating that timing is relevant in phage therapy
[61]. Similarly, no adverse inflammatory response was induced by the ETP-1 phage in zebrafish (
Danio rerio), and twelve days of exposition to ETP-1 was able to increase survival from 18% in the control group up to 68% after infection with
Edwardsiella tarda bacteria
[62]. Another example can be found in the North African catfish (
Clarias gariepinus). Ulcerative lesions caused by
P. aeruginosa in North African catfish were reduced seven fold compared with untreated control after 8–10 days of treatment with a single phage
[63]. In addition, treatment with two different phages at MOI of 100 reached 100% of survival in Vietnamese striped catfish (
Pangasianodon hypophthalmus) infected with
Aeromonas hydrophila, which produces hemorrhagic septicemia, compared to 13% of survival in the control group
[64].
Vibrio sp. produce mortality in bivalve larvae and bacteriophages could be used as biocontrol agents in oyster hatcheries. Two different approaches have been described to solve this problem. The first consists on direct phage treatment comprising two phages, which diminished mortality rates from 77.9% in the control group to 28.2% after just 24 h of incubation
[65]. However, the second approach focuses on decontaminating microalgae as vectors for
Vibrio sp. infection of larval cultures. Phage administration in microalgae resulted in significant reduction of
Vibrio sp. within 2 h, suggesting that feeding larvae with decontaminated microalgae could be a promising preventive method to avoid infection of bivalve larvae
[66]. Curiously, in 2019, a study using a heterologous expression vector was performed against
Vibrio parahaemolyticus. The yeast
Pichia pastoris X-33 expressed the phage endolysin Vplys60 from bacteriophage qdv001 and the enzyme was shown to inhibit biofilm formation and to reduce mortality rates for the crustacean
Artemia franciscana [67]. In other studies, a phage treatment with two phages against
Vibrio anguillarum infection was effective at 72 h in zebrafish larvae
[68], and a cocktail of three phages isolated from sewage showed host specificity against eight
Vibrio coralliilyticus strains and a
Vibrio tubiashii strain, obtaining a decrease of over 90% in
V. coralliilyticus compared to the untreated control
[69].
These studies reveal that current results are more promising in aquaculture than in farms. More studies are needed to clarify the real sanitary and economic potential of phage-based therapies in the food industry. It is possible that, as it happens in humans, better results could be obtained by mixing phages and antibiotics due to the synergistic effect.
Table 2. Summary of reviewed studies using phage therapy in animals.
Animal |
Infection/Colonization |
Bacteria |
Phage Therapy |
Outcome |
References |
Chicken |
Salmonellosis and colibacillosis |
S. enterica serovar Kentucky and Escherichia coli O119 |
Siphoviridae (107 PFU) against serovar Kentucky and Podoviridae (10 PFU) against Escherichia coli orally |
Reduction of mortality from 30% to 0% in treated group |
[41] |
Sheep |
Rhinosinusitis |
Pseudomonas aeruginosa |
Cocktail of 4 phages (Pa193, Pa204, Pa222, and Pa223) at 108–1010 PFU/mL |
Reduction of biofilm biomass on sinus mucosa |
[42] |
Cow |
Subclinical mastitis |
Staphylococcus aureus |
Phage K (1011 PFU) intramammary infusions for 5 days |
3/18 cows were cured compared to 0/20 of control group |
[46] |
Sheep |
Gut |
Escherichia coli O157:H7 |
Oral phage KH1 (1011 PFU) or DC22 (1013 PFU) |
No reduction of strain O157:H7 |
[52][53] |
Sheep |
Gut |
Escherichia coli O157:H7 |
Cocktail of CEV1 (T4-like) and CEV2 (T5-like) orally |
Reduction >99% of Escherichia coli in the lower intestinal tract |
[54] |
Sheep |
Gut |
Escherichia coli O157:H7 |
Cocktail of 8 phages orally |
Reduction of fecal Escherichia coli O157:H7, but not in the rumen, 24 h after phage administration |
[55] |
Pig |
Nasal colonization |
MRSA V0608892/1 strain |
P68 (Podovirus) and K* 710 (Myovirus) in gel |
No reduction observed in the nasal mucosa |
[56] |
Pig |
Prevention |
Salmonella enterica, Staphylococcus aureus, Escherichia coli and Clostridium prefringens |
Cocktail of phages orally |
Compared to probiotics, phages had better results as growth promoters, improving digestibility, daily weight gain and gain per feed |
[57] |
Escherichia coli | O157:H7 |
Omnilytics (USA) |
BacWash |
Salmonella enterica, Escherichia coli O157:H7 |
AgriPhage |
Xanthomonas campestris, Pseudomonas syringae |
APS Biocontrol Ltd. (UK) |
Biolyse-PB |
Erwinia sp., Pectobacterium sp., Pseudomonas sp. |
Proteon Pharmaceuticals SA (Poland) |
Bafasal |
Salmonella enterica |
Bafador |
Pseudomonas sp., Aeromonas sp. |
FINK TEC GmbH (Germany) |
Secure Shield E1 |
Escherichia coli |
Brimmedical (Georgia) |
PYO Phage |
Staphylococcus sp., Escherichia coli, Streptococcus sp., Pseudomonas sp., Proteus sp. |
Intesti Phage |
Shigella sp., Salmonella enterica, Staphylococcus sp., Proteus sp., Escherichia coli, Pseudomonas aeruginosa |
SES Phage |
Staphylococcus sp., Enteropathogenic serotypes of Escherichia coli, Streptococcus sp. |
EnkoPhagum |
Salmonella enterica, Shigella sp., Enteropathogenic serotypes of Escherichia coli, Staphylococcus sp. |
Fersisi Phage |
Staphylococcus sp., Streptococcus sp. |
Mono-phage |
Staphylococcus sp., Escherichia coli, Streptococcus sp., Enterococcus sp., Pseudomonas aeruginosa, Proteus sp. |
2.2. Bacteriophages to Control Listeria monocytogenes
L. monocytogenes is peculiar due to its ability of growing at refrigerated temperatures (2–8 °C). Guenther et al. showed the effect of the lytic A511 phage to control
L. monocytogenes in different ready to eat foods. In liquid samples as chocolate milk and mozzarella cheese brine, this phage was able to reduce the
L. monocytogenes population below the detection limit, while in solid samples (hot dogs, sliced turkey meat, smoked salmon, seafood, sliced cabbage, and lettuce leaves) the reduction was above 5 log units
[17]. Another phage used to control
L. monocytogenes in food products and food processing environments is P70, a phage known to have a broad host range infecting
Listeria sp. serovars 1/2a, 1/2b, 1/2c, 4a, 4c, 4d, 4e, 5, 6a and 6b with results over 62% of lysis
[18].
Currently, there are two products based on phages approved in the US to be used in food industry against
Listeria (
Table 1). The United States Food and Drug Administration and the U.S. Department of Agriculture approved ListShield
TM (Intralytix, Baltimore, MA, USA) as a food additive for ready-to-eat meat and poultry products, usually as a spraying or dipping suspension
[19]. ListShield
TM is a mixture of six lytic phages targeting
L. monocytogenes that does not affect the organoleptic quality of foods and does not produce adverse effects on commensal microbiota
[20]. Gutierrez et al. tested the product ListShield
TM on Spanish dry-cured ham and the surfaces that are commonly used in food industry and obtained a 100% lysis of
L. monocytogenes strains examined. In dry-cured ham, the reduction of bacterial population was of 3.5 log units after 14 days of incubation at 4 °C. Moreover, ListShield was effective in removing 72 h biofilms formed on stainless steel surfaces by most of the assayed strains after four hours of treatment at 12 °C
[21]. A recent study also tested the effectiveness of ListShield
TM in chicken breast. The phage treatment reduced the bacterial population 0.84 log CFU/mL when it was applied alone and 2.04 log CFU/mL in combination with UV-C treatment during storage for 72 h without significant differences in colour, pH or food quality
[22].
The second formulation approved in the USA is LISTEX™P100 (Micreos Food Safety, Wageningen, The Netherlands), a brand composed of bacteriophage P100 produced to control
L. monocytogenes. This product has been shown to reduce at least 3.5 log units on soft cheese
[23]. Soni et al. demonstrated its activity on fresh channel catfish fillets (
L. monocytogenes reduction between 1.4 and 2.0 log CFU/g at 4 °C, 10 °C, and 22 °C)
[24], raw salmon (bacterial reduction of 1.8, 2.5, and 3.5 log CFU/g from initial bacterial loads of 2, 3, and 4.5 log CFU/g, respectively, at 4° and 22 °C)
[25], and on soft cheese (with initial bacterial reduction of 2–4 log CFU/cm
2 at 4 °C, but subsequent bacterial regrowth reported)
[26]. Also this bacteriophage has been tested to reduce
L. monocytogenes biofilms on stainless steel coupon surfaces resulting in high elimination of biofilm mass in all
L. monocytogenes strains tested
[27]. In 2017, the effect of this product was tested in sushi. Promising results were obtained in assays with initial 6-log CFU/g of bacteria and 8-log PFU/g of phage inoculation at 22 °C; a maximum reduction of 4.44 log CFU/g was achieved when the product was inoculated directly in sashimi samples, compared with the control group
[28]. LISTEX™P100 has also been tested in soft cheeses achieving a reduction of more than 2 log CFU/mL
[29]. Recently, the effect of the phage P100 in combination with the antimicrobial peptide pediocin PA-1 and mild high hydrostatic pressure was evaluated as a new method to eradicate
Listeria from milk. The combination decreased immediately the
L. monocytogenes population, although in a few cases a regrowth during the storage process was encountered
[30].