Introduction

Pseudomonas spp. are aerobic, motile, and non-spore-forming bacilli widely found in the environment. In the food industry, Pseudomonas spp. are spoilage bacteria that cause problems in aerobically stored foods, particularly in foods with high water content and neutral pH, such as red meat, fish, poultry, milk, dairy products, vegetables, and fruits (Raposo et al, 2017). In particular, the P. fluorescens group, along with the psychrophilic P. fragi and P. putida, is involved in the spoilage of milk, meat, and fish at low storage temperatures because they can grow aerobically under chilled conditions (Papadopoulou et al. 2020). Additionally, Pseudomonas spp. form strong biofilms, higher-order structures composed of polysaccharides, polypeptides, and extracellular nucleic acids (Scales et al. 2014; Fanelli et al. 2021). Bacterial cells in biofilms are more resistant to physical and chemical stresses than planktonic bacterial cells, making it difficult to achieve sufficient bactericidal effects under normal sterilization conditions (Simões et al. 2010).

Drug-resistant bacteria are a growing problem worldwide. Bacteriophages (phages) have attracted considerable attention as a promising solution (Thiel 2004). Phages are viruses that infect bacteria, and unlike antibiotics, which have a broad antibacterial spectrum, they have high host specificity. They are harmless to humans and can selectively sterilize target host bacteria without affecting food characteristics (Sulakvelidze et al. 2001). These phages were shown to be effective against biofilms. Many phages, but not all, can produce polysaccharases or polysaccharide lyases and can access the biofilm interior through the water channel (Knecht et al. 2020; Sutherland et al. 2004). Therefore, it is useful in many fields, including clinical and food, as a novel antibacterial agent (Endersen and Coffey 2020). Although Pseudomonas phages have been known and studied for over half a century, studies have mainly focused on pathogenic bacteria (Ceyssens and Lavigne 2010) such as P. aeruginosa (Naknaen et al. 2023; Lister et al. 2009), P. plecoglossicida (Nishimori et al. 2000), and P. syringae (Hirano and Upper 2000). Pseudomonas spp. are most frequently observed in chilled food spoilage (Raposo et al, 2017), but few reports infecting such food spoilage Pseudomonas spp. (Tanaka et al. 2018; Lammens et al 2020).

In this study, we isolated Pseudomonas phages from food or soil samples using Pseudomonas spp. isolated from lettuce as hosts and investigated the various characteristics and the combined use of the phages to determine the optimal combination as a phage cocktail and examined the potential use of the phage cocktail as a non-thermal sterilization method, mainly in the food industry.

Materials and methods

Bacterial strains and culture media

Pseudomonas fluorescens groups No. 257 and No. 271 were isolated from lettuce leaves in our laboratory and identified as the P. fluorescens group (P. fluorescens group, P. fluorescens, or P. reactans) using the ID test NF-18 (Nissui Pharmaceutical Co., Ltd., Tokyo, Japan) and 16S rRNA sequencing using primers reported by Lane (1991). P. aeruginosa NBRC13275, P. alcaligenes NBRC14159, P. fluorescens NBRC14160, P. fragi NBRC3458, P. oleovorans NBRC13583, P. putida NBRC14164, and P. tolaasii NBRC15100 were obtained from the Biological Resource Center (NBRC). Bacterial strains grown on tryptic soy agar (TSA; Becton, Dickinson and Company, Sparks, MD, USA) were stored at 4 °C until use.

Isolation, purification, and propagation of Pseudomonas phages

The Pseudomonas fluorescens group No. 257 strain was used as the bacterial host for phage isolation. Eighty-five food or soil samples were used for phage isolation. Food samples were obtained from 65 chickens, three beefs, six porks, one fish, and two vegetables purchased from various supermarkets in Fukuoka City, Japan. Eight soil samples were obtained from the Fruit Tree Research Institute of the Fukushima Prefectural Agricultural Research Center (Iizaka-cho, Fukushima City, Fukushima Prefecture, Japan). Each sample (50 g) was aseptically placed in a sterile stomacher bag and homogenized in 100 mL of tryptic soy broth (TSB; Becton, Dickinson and Company, Sparks, MD, USA) with 10 mM CaCl2 for 1 min. The homogenate was inoculated with 100 µL of the host culture and incubated at 30 °C for 24 h. Then, 10 mL of the culture was withdrawn and centrifuged at 12,000 × g at 4 °C for 20 min. The supernatants were filtered through a 0.45-µm pore size sterile membrane filter (Merck Millipore, Ireland). Filtrates were mixed at a 1:1 ratio with the host cell suspension and incubated for 1 h at 30 °C. After incubation, 200 µL of the mixture was added to 4 mL of molten top agar (TSB containing 0.5% (w/v) agar) at 55 °C, poured onto plates of TSA, and incubated at 30 °C for 24 h for plaque formation. An observed plaque was picked up, suspended in 1 mL of saline magnesium (SM) buffer (0.05 M Tris–HCl buffer, pH 7.5, containing 0.1 M NaCl, 0.008 M MgSO4, and 0.01% gelatin) and tenfold serially diluted with the same buffer. The phages were purified as follows. The phage dilution (100 µL) and 100 µL of the host culture were mixed and incubated at 30 °C for 1 h. The mixture was then added to 4 mL top agar at 55 °C, immediately poured onto TSA plates, and incubated overnight at 30 °C for plaque formation. This purification procedure was repeated thrice. Purified phages with high titers (> 108 PFU/mL) were stored at 4 °C until use. The phages were named to include information about phage morphology and host according to the proposal by Kropinski et al. (2009).

Host range determination

The host range of the purified phages was assessed using spot tests on the nine Pseudomonas strains listed in Fig. 1. The phage suspension (10 µL) was spotted onto molten top agar inoculated with each host bacteria and incubated at 30 °C for 24 h. Plaque formation was interpreted as lysis of the bacterial host, and no plaque was interpreted as no lysis.

Fig. 1
figure 1

Source of isolations was PCF1 from chicken fallopian tube, PCG1 from chicken gizzard; PCL1 from chicken liver; PCS1-7 from chicken skin, PCW2 from chicken wings; and PSP1 from soil. *Isolated from lettuce leaf

Heat map for host range of the phages against various Pseudomonas strains.

Full size image

Lytic activity test

To assess the effects of the isolated Pseudomonas phages on planktonic cells, P. fluorescens group No. 271 precultured at 30 °C in TSB was diluted with TSB to attain a cell concentration of approximately 107 CFU/mL. The cell suspension was inoculated with a phage suspension (107 PFU/mL) at a multiplicity of infection (MOI) of 1. For the control, the same volume of SM buffer was used instead of the phage suspension. The mixture was incubated at 30 °C for 24 h with shaking at 130 rpm. At 0, 4, 8, and 24 h of incubation, the viable counts were measured using the plating method using TSA plates after overnight incubation at 30 °C. The lytic activity tests were performed at 4 °C. The host cell suspension (107 CFU/mL) was inoculated with a phage suspension (109 PFU/mL) at an MOI of 100. The mixture was incubated at 4 °C for 96 h, with agitation at 130 rpm. Viable cell counts were measured by plating the assays on TSA. Colonies were counted after overnight cultivation at 4 °C followed by overnight cultivation at 30 °C.

To assess the effects of phage cocktails, equal volumes of different phage suspensions with 107 PFU/mL were mixed to prepare a phage cocktail (107 PFU/mL). A host cell suspension (107 CFU/mL) in TSB was inoculated with the phage cocktail at an MOI of 1. The mixture was incubated at 30 °C for 96 h, with agitation at 130 rpm. An aliquot of the mixture was withdrawn at a suitable interval, and viable counts were measured by the plating assay using TSA after overnight cultivation at 30 °C.

The combined effects of phages vB_PflP-PCS4 and/or vB_PflP-PCW2 on the viability of P. fluorescens group No. 271 and P. putida NBRC14164 were determined at 30 °C, 8 °C, and 4 °C. A bacterial cell suspension (107 CFU/mL) in TSB was inoculated with a single phage (109 PFU/mL) or a phage cocktail (109 PFU/mL) at an MOI of 100. An equal volume of SM buffer was used instead of the phage suspension used as a control. The mixtures were incubated at 30 °C, 8 °C, and 4 °C for 168 h with agitation at 130 rpm. Viable counts were measured by plating the cells on TSA. Colonies were counted after overnight incubation at 30 °C. In the case of the measurement of viable counts of bacteria treated with phages at 8 °C and 4 °C, plates were first incubated overnight at 8 °C and 4 °C, respectively, before overnight incubation at 30 °C.

Determination of latent period and burst size of phages

One-step growth curve experiments were performed according to the methods described by Son et al. (2018) and Masuda et al. (2021) with some modifications. The host strain P. fluorescens group No. 271 in the stationary phase of growth and the vB_PflP-PCS4 or vB_PflP-PCW2 phage solutions were mixed at final concentrations of 108 CFU/mL and 106 PFU/mL in 500 mL of TSB (MOI of 0.01) and incubated at 30 °C for 5 min. Mixed solutions were centrifuged at 4 °C and 12,000 × g for 3 min, and the supernatant was discarded to remove excess phage particles. The pellets were resuspended in 10 mL of prewarmed TSB at 30 °C, and suspensions were incubated at 30 °C for 40 min. Every 5 min from the start of incubation, 200 mL of suspension was transferred into a 1.5-mL Eppendorf tube and centrifuged at 4 °C and 12,000 × g for 3 min. After centrifugation, supernatants were transferred into new tubes and kept at 4 °C as phage solutions until they were used as templates for phage quantification using quantitative PCR (qPCR). The burst size of each phage was calculated by dividing the phage particle number at the end of a single infectious cycle by the phage particle number at time zero, using data from three independent experiments.

In this study, the number of phage particles in each phage solution was quantified using a plaque-forming assay and qPCR with each phage solution as a template, as previously described (Peng et al. 2018; Masuda et al. 2021). qPCR was performed using the Mx3000P real-time PCR system (Stratagene, La Jolla, CA, USA) and Thunderbird SYBR qPCR master mix (TOYOBO Co., Ltd., Osaka, Japan) according to the manufacturer’s instructions. Phage solutions were first quantified using a plaque-forming assay to prepare standard solutions (104–107 PFU/mL). Standard curves for both vB_PflP-PCS4 and vB_PflP-PCW2 phages were prepared using these phage solutions as templates and the primers PCS4-11629Fw/PCS4-11733Rv and PCW-15735Fw/PCW2-15836Rv (Table S1).

pH and thermal stability test

The effect of pH on phage activity was determined at 25 °C. The pH of the SM buffer was adjusted with 0.1 M HCl and 0.1 M NaOH at pH values ranging from 3 to 12. Then, 20 µL of phage suspension (1 × 107 PFU/mL for vB_PflP-PCS4 and 1 × 106 PFU/mL for vB_PflP-PCW2) was added to 380 μL aliquots of SM buffer with different pH values and incubated at 25 °C for 24 h. For the thermal stability test, 200 µL of the same phage suspensions was incubated at 4 °C, 30 °C, 40 °C, 50 °C, 60 °C, 70 °C, and 80 °C for 1 h. After pH and thermal treatments, the number of surviving phages was determined using a plaque-forming assay.

Genetic analysis

Nucleic acids were extracted from newly purified phages vB_PflP-PCS4 and vB_PflP-PCW2 suspensions (109 PFU/mL). The phages were lysed using proteinase K at 70 °C for 10 min. After lysis, DNA was purified using a Phage DNA Isolation Kit (NORGEN, Cat. 46800, 46850 ON, Canada). The genomic DNA was sequenced using an Illumina HiSeq system (HiSeq-PE150). The read sequences were assembled de novo using the VelvetOptimizer version 1.2.10 (Zerbino and Birney 2008). The assembly gaps were closed using GMcloser (Kosugi et al. 2015). The assembled genome was annotated using the Prokka 1.14.6 software (Seemann 2014). Open reading frames (ORFs) were determined using the NCBI ORF Finder (https://www.ncbi.nlm.nih.gov/orffinder/). The function of each ORF was predicted by homology searching using NCBI BLASTP (Ramsay et al. 2000). The complete genome sequences of phages vB_PflP-PCS4 and vB_PflP-PCW2 have been deposited in the GenBank database under accession numbers OK094519.1 and OM471789, respectively. The absence of tRNA, virulence, and toxic genes was confirmed using tRNAscan-SE 2.0 (Chan et al. 2021; Lowe and Chan 2016) and VirulenceFinder 2.0 (Clausen et al. 2018; Joensen et al. 2014; Malberg Tetzschner et al. 2020).

Statistical analysis

Data points from all experiments represent the mean of results from at least three independent experiments, and error bars indicate standard deviations. All data were subjected to Student’s t-test to ensure statistically significant differences at P < 0.05.

Results

Isolation and purification of Pseudomonas phages from meat and soil samples

In the current study, 85 food and soil samples were tested, and 12 lytic phages were isolated using a lettuce-derived P. fluorescens group strain as a host. Among them, seven phages were isolated from chicken skin, one from chicken wing, one from chicken gizzard, one from chicken liver, one from chicken fallopian tubes, and one from soil.

Host range of isolated phages

The lytic activities of the isolated phages against Pseudomonas spp. are listed in Fig. 1. All the phages were active against lettuce-derived P. fluorescens strains and P. putida NBRC14164. Some phages were active against P. alcaligenes NBRC14159, P. fragi NBRC3458, P. oleovorans NBRC13583, and P. tolaasii NBRC15100. None of the phages were active against P. aeruginosa NBRC13275 and P. fluorescens NBRC14160. The phages were grouped into six groups according to their host range patterns.

Phages vB_PflP-PCS4, vB_Pfl?-PCS1–3, and vB_Pfl?-5–7 were isolated from different chicken skin samples. Among them, vB_Pfl?-PCS1, 2, and 7 showed the same host range pattern A, and vB_Pfl?-PCS3 and vB_Pfl?-PCS6 showed the same host range pattern D. Since the phages were isolated from the different chicken samples purchased at the different supermarket at different day, it seems that they were different phages. Genomic DNA sequencing is required to confirm that these phages showing the same host range pattern are different phages.

Lytic activity of purified phages at 30 °C and 4 °C

The lytic activity of the 12 phages was determined at different temperatures. The lytic activity of the phages at 30 and 4 °C is shown in Fig. 2. At 30 °C, phages vB_Pfl?-PCS2, vB_Pfl?-PCS3, vB_Pfl?-PCS5, vB_Pfl?-PCS6, vB_PflP-PCW2, and vB_Pfl?-PSP1 largely decreased the viable counts of the lettuce-derived P. fluorescens group No. 271 strain by 3 to 4 log at 4 or 8 h of incubation. At 4 °C, the phages vB_Pfl?-PCL1 and vB_Pfl?-PSP1 decreased the viable count by approximately 2.5 log after 48 and 24 h of incubation, respectively.

Fig. 2
figure 2

Effects of phages on viability of P. fluorescens group No. 271 at 30 °C and 4 °C. P. fluorescens group No. 271 suspension in TSB was inoculated with phage suspension at an MOI of 1 and 100 at 30 °C and 4 °C, respectively. For the control, an SM buffer was used. The mixtures were incubated at 30 °C and 4 °C with shaking, and viable counts were determined. Symbols: ○ , control without phage; , vB_Pfl?-PCF1; , vB_Pfl?-PCG1; , vB_Pfl?-PCL1;●, vB_Pfl?-PCS1; , vB_Pfl?-PCS2; , vB_Pfl?-PCS3; , vB_PflP-PCS4; , vB_Pfl?-PCS5; ▲, vB_Pfl?-PCS6; ●, vB_Pfl?-PCS7; , vB_PflP-PCW2; , vB_Pfl?-PSP1

Full size image

Lytic activity of phage cocktails at 30 °C

Based on the host range patterns and lytic activities at different temperatures, vB_Pfl?-PCL1 (pattern C), vB_PflP-PCS4 (pattern C), vB_Pfl?-PCS5 (pattern E), vB_PflP-PCW2 (pattern F), and vB_Pfl?-PSP1 (pattern D) were selected to test the combined effects of the phages. Among the two phages (pattern C) with the widest host range, vB_Pfl?-PCL1 and vB_PflP-PCS4 were selected because of the strongest lytic activity at 4 °C and strong lytic activity both at 30 °C and 4 °C, respectively. Among three phages (pattern D), vB_Pfl?-PSP1 was selected since it showed relatively strong lytic activity both at 30 °C and 4 °C than those of the other two phages. Phage vB_Pfl?-PCS5 was selected for cocktail testing, because the phage (pattern E) showed unique and the widest host range different from those of pattern C phages. Phage vB_PflP-PCW2 (pattern F) was selected for cocktail testing, because the phage showed a unique host range. In contrast, phage vB_Pfl?-PCS2 showed very strong lytic activity at 30 °C among the four phages (pattern A) showing the second widest host range. The host range was covered by the phages with patterns C and D. Although phage vB_Pfl?-PCG1 (pattern B) showed unique and the widest host range like patterns C and E, the lytic activity was not strong both at 30 °C and 4 °C.

Four phage cocktails were prepared by removing one phage from the 5-phage cocktail. The phage cocktails were tested for lytic activity against the P. fluorescens group No. 271 strain. The effects of 4- and 5-phage cocktails on the viability of the P. fluorescens group No. 271 strain at MOIs of 1 and 30 °C in TSB are shown in Fig. 3. All cocktails decreased the viable counts to below the lower limit of detection at 1 h. The viable count was less than the lower limit of detection for 8 h in all cocktail treatments. Viable cell counts did not increase 72 h after treatment with the 5-phage cocktail and the phage cocktail without vB_Pfl?-PSP1. The viable cell count increased after 48 h in the presence of phages without vB_Pfl?-PCL1. Viable counts increased to the same level as those of the control in the presence of phage cocktails without vB_PflP-PCS4, PCS5, or vB_PflP-PCW2. These results suggest that vB_Pfl?-PSP1 and vB_Pfl?-PCL1 have little or no effect on suppressing the growth of resistant bacteria in combination with other phages.

Fig. 3
figure 3

Effects of 4- and 5-phage cocktails on the viability of P. fluorescens group No. 271 at 30 °C. P. fluorescens group No. 271 suspension in TSB was inoculated with phage cocktails at an MOI of 1. For the control, an SM buffer was used. Four-phage cocktails were prepared by removing one phage from a 5-phage cocktail (vB_Pfl?-PCL1 + vB_PflP-PCS4 + vB_Pfl?-PCS5 + vB_PflP-PCW2 + vB_Pfl?-PSP1). The mixtures were incubated at 4 °C with shaking, and viable counts were determined. Symbols: ○, control without phages; , 5-phage cocktail; ●, cocktail without vB_PflP-PCW2; ■, cocktail without vB_PflP-PCS4; , cocktail without vB_Pfl?-PCS5; , cocktail without vB_Pfl?-PCL1; , cocktail without vB_Pfl?-PSP1

Full size image

According to the results, a phage cocktail with three phages, vB_PflP-PCW2, vB_PflP-PCS4, and PCS5 and 2-phage cocktails consisting of two of the three phages were tested. The effects of 2- and 3-phage cocktails on the viability of P. fluorescens group No. 271 at MOIs of 1 and 30 °C in TSB are shown in Fig. 4. In all the cells treated with the phage cocktails, the viable counts of P. fluorescens decreased to less than the lower limit of detection at 1 h and did not increase for 8 h from the start of incubation. However, the count increased to > 6 log CFU/mL after treatment with the phage cocktail without vB_PflP-PCW2. The viable count was the lowest after treatment with the phage cocktail without PCS5. These results indicate that the phage cocktail consisting of phages vB_PflP-PCS4 and vB_PflP-PCW2 showed the highest performance against P. fluorescens group No. 271 among the phage cocktails tested. These two phages were selected for characterization and genomic DNA sequencing.

Fig. 4
figure 4

Effects of 2- and 3-phage cocktails on the viability of P. fluorescens group No. 271 at 30 °C. P. fluorescens group No. 271 suspension in TSB was inoculated with phage cocktails at an MOI of 1. For the control, an SM buffer was used. Two-phage cocktails were prepared by removing one phage from a 3-phage cocktail (vB_PflP-PCS4 + vB_PflP-PCW2 + vB_Pfl?-PSP1). The mixtures were incubated at 30 °C with shaking, and viable counts were determined. Symbols: , control without phages; , 3-phage cocktail; , cocktail without vB_PflP-PCW2; , cocktail without vB_PflP-PCS4; , cocktail without vB_Pfl?-PSP1

Full size image

Characterization of phages vB_PflP-PCS4 and vB_PflP-PCW2

Phages vB_PflP-PCS4 and vB_PflP-PCW2 formed plaques with large and small halos, respectively (Fig. 5a, b). A one-step growth experiment determined the latent time and burst size of phages vB_PflP-PCS4 and vB_PflP-PCW2. As shown in Fig. S1, phage vB_PflP-PCS4 had a latent time of 10 min and a burst size of 15 ± 2.0 PFU/bacterial cell. Phage vB_PflP-PCW2 had a latent time of 15 min and a burst size of 12 ± 0.3 PFU/bacterial cell. Both phages had small burst sizes; however, vB_PflP-PCS4 had a short lytic cycle. According to the results of the thermal stability test (Fig. 5c), the titers of phages vB_PflP-PCS4 and vB_PflP-PCW2 did not change after heating at 50 °C for 1 h but decreased gradually with increasing heating temperature from 60 to 80 °C (Fig. 5c). The effect of pH on the titer of the phages is shown in Fig. 5d. The phages were stable for 24 h at pH values ranging from 3 to 11. Phage vB_PflP-PCW4 was stable at pH 2.

Fig. 5
figure 5

Characteristics of phages vB_PflP-PCS4 and vB_PflP-PCW2. Plaque morphology of phages a vB_PflP-PCS4 and b vB_PflP-PCW2 on the host P. fluorescens group No. 271 strain. Phages vB_PflP-PCS4 (○) and vB_PflP-PCW2 (△) were incubated in modified TSB broth at c different temperatures for 1 h and d different pH values at 25 °C for 24 h. The experiments were performed in triplicate, and error bars indicate standard errors of the mean. Bar: 10 mm

Full size image

The genome maps of phages vB_PflP-PCS4 and vB_PflP-PCW2 are shown in Fig. 6. Phage vB_PflP-PCS4 had a circular double-stranded DNA genome composed of 39,191 bp, with an overall G + C content of 58% (Fig. 6a). In the vB_PflP-PCS4 genome, 46 ORFs were identified; however, no tRNA, virulence, or toxic genes were identified using tRNAscan-SE 2.0 and VirulenceFinder2.0. A BLAST search indicated that the vB_PflP-PCS4 phage genome had 97.8% similarity to the P. fluorescens phage UNO-SLW1 complete genome (accession code NC_047873.1). Among these, 26 ORFs (56.5%) were assigned known functions, and 20 (43.5%) were assigned hypothetical functions, including one unmatched hypothetical protein. The results of the genome annotation are presented in Table S2. These proteins were categorized into four groups: (i) phage structure and packaging (connector protein, major capsid protein, tail spike protein, DNA encapsidation protein, collar protein, and tail protein); (ii) DNA replication (DNA polymerase, RNA polymerase, exonuclease, and primase/helicase); (iii) host lysis (N-acetylmuramoyl-L-alanine amidase, holin, and RZ lysis protein); and (iv) hypothetical protein. Phage vB_PflP-PCS4 had an N-acetylmuramoyl-L-alanine amidase endolysin gene (gpp19). The highest identity was found for the gene encoding lysin of P. fluorescens phage UNO-SLW1 (accession code YP_009793708) with 87.58% identity. Therefore, gpp19 in vB_PflP-PCS4 appears to encode a novel endolysin.

Fig. 6
figure 6

Genome maps of phage a vB_PflP-PCS4 and b vB_PflP-PCW2. Arrows: green, phage structure and packaging; orange, DNA replication; yellow, host lysis; red–purple, hypothetical protein. The accession numbers of vB_PflP-PCS4 and vB_PflP-PCW2 are OK094519.1 and OM471789, respectively

Full size image

As shown in Fig. 6b, phage vB_PflP-PCW2 had a circular double-stranded DNA genome composed of 40,117 bp, with an overall G + C content of 59%. In the vB_PflP-PCW2 genome, 48 ORFs were identified; however, no tRNA, virulence, or toxicity genes were identified. A BLAST search indicated that the vB_PflP-PCW2 phage genome had 98.04% similarity with the P. syringae pv. Actinidiae phage CHF7 complete genome (accession code MN729596.1). The results of the genome annotation are listed in Table S3. Among them, 24 ORFs (50%) were assigned known functions, and 24 (50%) were assigned hypothetical functions, including one unmatched hypothetical protein. Phage vB_PflP-PCW2 also contained an N-acetylmuramoyl-L-alanine amidase endolysin gene (gpp27); however, the nucleotide sequence of this gene showed 100% similarity to that of the lysozyme of Pseudomonas phage CHF19 (accession code QHB47983.1). P. fluorescens phage UNO-SLW1 (Lu et al. 2017) and P. syringae pv. Actinidiae phage CHF7 (Flores et al. 2020) were reported to be the Podoviridae family. According to the high identity in the nucleotide sequences of the genomic DNAs obetweenvB_PflP-PCS4 and P. fluorescens phage UNO-SLW1 and vB_PflP-PCW2 and P. syringae pv. Actinidiae phage CHF7, these phages seem to belong to the Podoviridae family.

Effects of phage cocktail with phages vB_PflP-PCS4 and vB_PflP-PCW2 on viability of P. fluorescens group No. 271 and P. putida NBRC14164 in broth at 30 °C, 8 °C, and 4 °C

The effects of a phage cocktail with phages vB_PflP-PCS4 and vB_PflP-PCW2 at an MOI of 100 were determined on the viability of lettuce-derived P. fluorescens group No. 271 and P. putida NBRC14164 strains at 30 °C, 8 °C, and 4 °C (Fig. 7). At 30 °C, the viable count of P. fluorescens group No. 271 decreased to the lower limit of detection (less than 1 log CFU/mL) after 1 h, and the bacterial regrowth was inhibited for 144 h (Fig. 7). At 8 °C and 4 °C, against P. fluorescens group No. 271, the phage cocktail did not completely inhibit the regrowth of P. fluorescens, but the viable counts were lower by 2.2 and 3.5 log, respectively, compared with those of the control at 168 h (Fig. 7). In contrast, at all temperatures, the viable count of P. putida NBRC14164 decreased to the lower limit of detection after 1 h, and bacterial regrowth was inhibited after 168 h in the presence of the phage cocktail (Fig. 7). Therefore, the combined use of phages vB_PflP-PCS4 and vB_PflP-PCW2 was more effective against P. putida NBRC14164 than against lettuce-derived Pseudomonas spp. No.271 at 30 °C, 8 °C, and 4 °C.

Fig. 7
figure 7

Effects of phage cocktail composed of vB_PflP-PCS4 and vB_PflP-PCW2 on the viability of Pseudomonas strains at different temperatures. P. fluorescens group No. 271 (○,●) and P. putida NBRC14164 (△,▲) were incubated in the absence (open symbols) and presence (closed symbols) of the phage cocktail (vB_PflP-PCS4 and vB_PflP-PCW2) at an MOI of 100

Full size image

Discussion

Food spoilage is a serious global problem in developed and developing countries. Pseudomonas spp. produce many thermotolerant lipolytic and proteolytic enzymes, reducing the quality and shelf life of raw and processed foods. Food spoilage caused by Pseudomonas spp. occurs in milk, dairy products, meat, fish, water, fruits, and vegetables (Papadopoulou et al. 2020; Singh, 2017). In this study, we isolated and characterized phages against food-spoilage pseudomonads for the biocontrol of bacteria in foods during cold storage. Twelve Pseudomonas phages were isolated from a soil sample and different chicken samples obtained on different days at different supermarkets and were grouped into six groups according to their host range patterns. Phages vB_PflP-PCS4, vB_Pfl?-PCS1–3, and vB_Pfl?-PCS5–7 were isolated from different chicken skin samples. Among them, vB_Pfl?-PCS1, 2, and 7 showed the same host range pattern (A), and vB_Pfl?-PCS3 and 6 showed the same host range pattern (D). Since the phages were isolated from the different chicken samples purchased at the different supermarket at different day, it seems that they were different phages. Genomic DNA sequencing is required to confirm that these phages showing the same host range pattern are different phages.

Among Pseudomonas phages, some phages, such as P. tolaasii phage Bf7 (Sajben-Nagy et al. 2012), have been reported to have wide host specificity, infecting multiple Pseudomonas spp., including P. agarici, P. costantinii, P. fluorescens, P. fluorescens, P. putida, and P. tolaasii. In this study, phage vB_PflP-PCS4 and vB_PflP-PCW2 infected 77.8% and 44.4% of the Pseudomonas strains selected from food-spoilage pseudomonads, respectively (Fig. 1). The combined use of these two phages did not extend the host range but enhanced lytic activity and effectively retarded or inhibited regrowth of the phage-resistant population in both strains of the P. fluorescens group and P. putida at 30 °C, 8 °C, and 4 °C, although the single phage did not retard the regrowth of bacteria. Since these phages seem to have great potential for the biocontrol of pseudomonads during cold storage of food, the effect of the phage cocktail on the viability and growth of pseudomonads in various foods during cold storage will be reported in the near future. In actual food application tests, it is necessary to conduct tests with wide ranges of MOIs to determine the optimal phage application conditions.

Phage vB_PflP-PCS4 genome showed 97.8% similarity to P. fluorescens phage UNO-SLW1 genome. However, the tail fiber gene showed 53.55% identity to that of Pseudomonas phage UNO-SLW1, being the lowest similarity to the known gene among the whole genes in phage vB_PflP-PCS4 (Table S2). The phage tail fiber plays a very important role in binding to the host bacterial phage receptor. This result suggests that vB_PflP-PCS4 seems to be a novel phage with a different host range from the previously reported Pseudomonas phages. Phage vB_PflP-PCW2 genome had 98.04% similarity with the P. syringae pv. Actinidiae phage CHF7 complete genome. In contrast to phage vB_PflP-PCS4, the tail fiber gene showed high similarity to that of Pseudomonas phage CHF21, but low similarity in tail fiber assembly chaperone gene (67.23% identity to tail fiber assembly chaperone of Pseudomonas phage vB_PsyP_3MF5) and DNA-directed RNA polymerase gene (61.36% identity to hypothetical protein of Pseudomonas phage ALEA) and HNH endonuclease gene (58.39% identity to hypothetical protein of Pseudomonas phage CHF1) (Table S3), suggesting phage vB_PflP-PCW2 seems to be a novel phage with similar host range of known phages.

Genome analysis of vB_PflP-PCS4 and vB_PflP-PCW2 revealed no tRNA, virulence, or toxicity genes, suggesting that they can be used in food. Moreover, phages vB_PflP-PCS4 and vB_PflP-PCW2 harbored a peptidoglycan lytic exotransglycosylase gene (EC:4.2.2. n1) as the tail spike protein and N-acetylmuramoyl-L-alanine amidase (EC 3.5.1.28) as the endolysin. Phage vB_PflP-PCS4 had a gene encoding a novel N-acetylmuramoyl-L-alanine amidase endolysin (gpp19), which showed 87.58% identity to the lysin of P. fluorescens phage UNO-SLW1 (accession code YP_009793708). In addition, the RZ lysis protein or spanin (gpp43) of phage vB_PflP-PCS4 showed 97.95% identity to the putative Rz lysis protein of Pseudomonas phage UNO-SLW1 (accession code YP_009793733.1) and low identity (61.54%) to the lysis protein of Pseudomonas phage PFP1 (YP_009804024.1). RZ lysis protein or spanin is a bacteriolytic enzyme required for disrupting the outer membrane and is present in most gram-negative bacteriophages (Summer et al. 2007). Spanins are cationic antimicrobial peptides (Holt et al. 2021). Based on these facts, further characterization of vB_PflP-PCS4 lysis proteins expressed as recombinant proteins is needed to apply the enzymes to control gram-negative bacteria in food matrices.

Lytic activity tests of phages at low temperatures showed that phages PCL1 and PSP1 showed strong lytic activity at 4 °C (Fig. 3). Some Pseudomonas phages have been reported to be active at low temperatures, such as P. fluorescens phage VW-6S and VW-6B (Xiang et al. 2018). Moreover, considering the possibility that the surface structure and metabolism of bacteria are different between low and optimal growth temperatures (Moreno and Rojo, 2014), the lytic activity of the phage cocktail was investigated at 4 °C and 8 °C (Fig. 7). Even at low temperatures, the phage cocktail showed strong lytic activity and retarded or inhibited the regrowth of P. fluorescens group No. 271 and P. putida, suggesting its effectiveness in extending the shelf life of food under cold storage. Further experiments will be performed to investigate the effects of this phage cocktail on the viability and regrowth of various food-spoilage pseudomonads in food matrices at low temperatures.

The combination of vB_PflP-PCS4 and vB_PflP-PCW2 phages caused rapid lysis of the host bacterium after 1 h of incubation at 30 °C, 8 °C, and 4 °C (Figs. 5 and 7). In general, the combined effect of multiple phages is explained by differences in receptor-binding proteins (RBPs), and many phage cocktails have been designed focusing on differences in host range or RBPs. However, some reports have shown that the efficacy of phage cocktails differs depending on the MOI, temperature, incubation time, and phage-derived lysis proteins (Niu et al. 2021). Lysis proteins, such as tail spike proteins and endolysins, are characterized by their ability to lyse bacterial cells more rapidly and cause less bacterial resistance than phages (Schmelcher et al. 2012). Therefore, combining different endolysins and tail spike proteins derived from phages vB_PflP-PCS4 and vB_PflP-PCW2 caused rapid lysis and suppressed bacterial regrowth. In future studies, these proteins derived from phages will be produced as recombinant proteins, and the mechanism underlying the rapid lysis of bacterial cells by the phage cocktail will be investigated.

Conclusions

Among the 12 Pseudomonas phages isolated from chicken meat and soil samples, a phage cocktail consisting of phages vB_PflP-PCS4 and vB_PflP-PCW2 showed strong lytic activity against lettuce-derived Pseudomonas spp. and P. putida NBRC14164 at low temperatures. The absence of tRNA, virulence, or toxicity genes in the genomic DNA of these phages suggests that they are safe for use in food. The phage cocktail seems potentially effective as an antimicrobial agent against the psychrotrophic spoilage Pseudomonas spp. in food. Genomic DNA sequence analysis indicated that vB_PflP-PCS4 harbors a novel endolysin gene. Further research is necessary to reveal the mechanisms underlying the combined effects of phages vB_PflP-PCS4 and vB_PflP-PCW2, characterize the novel endolysin, and evaluate the lytic activity and growth inhibition of phage-resistant bacteria using the phage cocktail in food.