To isolate and characterize listeriaphages from seafood environments.
To isolate and characterize listeriaphages from seafood environments.
Listeriaphages (phages) isolated from seafood environments were distinguished by physical and biological techniques including restriction digestion of phage DNA. Three phages belonged to order Caudovirales and showed psychrotrophic characteristics. The phages had broad host ranges against 23 Listeria strains by productive infection or at least by adsorption. At 15 ± 1°C, adsorption rate constants of the three phages ranged from 8·93 × 10−9 to 3·24 × 10−11 ml min−1 across different Listeria monocytogenes strains. In indicator hosts, the mean burst sizes of phages LiMN4L, LiMN4p and LiMN17 were c. 17, 17 and 11 plaque-forming units (PFU) per cell, respectively, at 15 ± 1°C. The respective latent periods were c. 270 min for phages LiMN4p and LiMN17, whereas for phage LiMN4L, it was c. 240 min.
The three virulent psychrotrophic phages isolated from seafood-processing environments had broad host ranges and low productive replication. These characteristics suggest that the phages may be suitable as passive biocontrol agents against seafood-borne L. monocytogenes.
This is the first report on the isolation of autochthonous virulent listeriaphages from seafood-processing environments and information on single-step replication and adsorption characteristics of such listeriaphages.
Virulent listeriaphages isolated from different environments are now gaining acceptability in the control of Listeria monocytogenes in food systems. These include commercial phage products such as Listex™ containing phage P100 (Carlton et al. 2005; Anon 2006) and ListShield™, a six-phage cocktail (Pasternack and Sulakvelidze 2009). These virulent phages were isolated from a milk-processing plant in Germany (Carlton et al. 2005) and from Baltimore inner harbour water in USA (Pasternack and Sulakvelidze 2009), respectively. Loessner and Busse (1990) isolated a group of phages that included phage A511 from a waste water treatment plant. Listeriaphages have also been isolated from human waste water effluents (Dykes and Moorhead 2002), sheep faeces (Lee 2008; Bigot et al. 2011), silage (Hodgson 2000; Schmuki et al. 2012) and turkey processing plants (Kim et al. 2008). A listeriaphage designated as vB_LmoM_AG20 has also been isolated from the drains of a meat-processing plant in Ontario, Canada (Anany et al. 2011).
The broad host ranges of some of these phages indicate that they can be used to control L. monocytogenes contamination in processing plants (Loessner and Busse 1990; Carlton et al. 2005; Kim et al. 2008; Pasternack and Sulakvelidze 2009). Efficacy of phage lysis of L. monocytogenes has been investigated in broth cultures, food models and food packaging (Dykes and Moorhead 2002; Leverentz et al. 2003, 2004; Carlton et al. 2005; Guenther et al. 2009; Soni and Nannapaneni 2010a; Soni et al. 2010; Anany et al. 2011; Bigot et al. 2011; Guenther and Loessner 2011). Moreover, some phages have shown to be effective in dealing with Listeria cells adhered and biofilms formed on hard surfaces that simulate food contact surfaces (Roy et al. 1993; Soni and Nannapaneni 2010 b; Montanez-Izquierdo et al. 2011). The phages, however, need to be characterized including the single-step replication and adsorption capabilities in conditions under which phages are intended to be used (Gill 2010). This study reports, for the first time, the characteristics of three virulent phages isolated from seafood-processing environments with the potential to control L. monocytogenes in similar environments at low temperatures.
Twenty-one L. monocytogenes strains isolated from seafood or seafood-processing environments were obtained from different laboratories in New Zealand comprising two strains (SMAC91 and SSM91) from University of Otago, Otago (Bremer et al. 2002), one strain (CW2) from Cawthron Institute, Nelson, and 18 strains from Plant and Food Research Limited (PFR), Auckland (Cruz and Fletcher 2011) (Table 1). Two other Listeria species (L. ivanovii and L. innocua) obtained from in-house culture collection of Massey University were also used in this study. The typing characteristics of Listeria strains (serotype and subtype) used in this study are summarized in Table 1. Listeria strains were stored in cryovials (Cryobank-CRYO/M, Mast Diagnostics, Merseyside, UK) at −80°C as per supplier's instructions. Monthly working cultures were recovered by streaking the frozen cultures on trypticase soy agar (TSA; Difco, Sparks, MD, USA) plates and incubating at 30 ± 1°C for 18 h. The culture plates were stored at 4°C. Broth cultures were prepared by transferring colonies from monthly cultures of Listeria strains into trypticase soy broth (TSB; Difco) and incubating at selected temperatures described below. A suspension of virulent phage A511 (Loessner and Busse 1990), which was used as reference phage, was kindly provided by the Institute of Food, Nutrition and Health, Zurich, Switzerland.
|Listeria strainsa||% EOP of listeriaphage|
|Strain code||Typing characteristics||LiMN4L||LiMN4p||LiMN17|
|Listeria monocytogenes strains|
|15JO5||1/2a or 3a||382||79||96||114||89||96||103|
|CW2||1/2a or 3a||872||93||96||88||104||84||87|
|18AO1||1/2a or 3a||22||91||95||89||96||72||87·3|
|22BO5||1/2a or 3a||652||77||91||92||97||45||67|
|22BO7||1/2a or 3a||872||78||91||98||93||45||67|
|18GO1||1/2a or 3a||8342||57||87||101||94||65||77|
|19CO5||1/2a or 3a||846||89||100||76||79||75||80|
|17AO2||1/2a or 3a||702||76||91||91||102||54||94|
|15AO7||1/2a or 3a||652||41||89||101||55||81||119|
|19EO1bc||1/2a or 3a||5946||<10−3||<10−3||<10−2||<10−2||<10−3||<10−3|
|18FO9||1/2a or 3a||5946||<10−3||<10−3||85||94||<10−3||<10−3|
|19CO9||1/2a or 3a||872||100||100||100||100||100||100|
|19EO3bc||1/2a or 3a||5132||<10−2||<10−2||<10−2||<10−2||<10−2||<10−2|
|19DO9bc||1/2a or 3a||5942||<10−5||<10−5||<10−3||<10−3||<10−7||<10−7|
|19DO3c||4b,4d or 4e||2312||<10−3||<10−3||<10−1||<10−1||<10−7||<10−7|
|19EO5||4b,4d or 4e||6331||67||72||77||79||46||68|
|19CO7||4b,4d or 4e||6331||66||74||78||80||46||65|
|18DO7||1/2b, 3b or 7||3527||<10−7||<10−7||<10−7||<10−7||<10−7||<10−7|
|18D05||1/2b,3b or 7||3527||<10−7||<10−7||<10−7||<10−7||<10−7||<10−7|
|Listeria ivanovii c||–||–||<10−3||<10−3||<10−2||<10−2||<10−7||<10−7|
|Listeria innocua c||–||–||<10−3||<10−3||<10−3||<10−3||<10−3||<10−3|
Field samples were collected from different locations (interior wall of draining receptacle of the milliscreen in the final dewatering section, floor, floor drains and interior wall of final sump prior to the discharge of screened fisheries effluent) of a fish waste treatment unit of a fish-processing plant in Nelson, New Zealand in January 2011. The samples were collected using 25 cm2 swabs (BSN code 71786-05, Propax brand, BSN Medical, Hamburg, Germany) premoistened with 5 ml of transport medium [1 g of peptone (Merck KGaA, Darmstadt, Germany), 5 ml of 10% sodium thiosulfate (Scharlau Chemie S.A., Barcelona, Spain), 10 g of Tween 20 (Ajax Finechem Pty Ltd, Sydney, NSW, Australia), 1 l of ultrapure water] over 50 cm2 on surfaces of sampling locations.
The sample swabs in plastic bags were transported in ice to the laboratory at the Cawthron Institute (Nelson, New Zealand) within 2 h. Each swab sample was placed in a glass bottle containing 10 ml of 10% (w/v) meat extract (ME) (Merck), vortex-mixed (WISE Mix VM-10, Daihaw Scientific company Ltd, Seoul, Korea) for 30 s to expel gas bubbles and incubated at 4 ± 1°C for c. 18 h (Hurst et al. 1991; Khan et al. 2002; Mark R Liles, personal communication). The samples were sonicated (70 W, 43 kHz and 80% power; Ultrasonic power for 160-HT, Soniclean Pty Ltd, Thebarton, SA, Australia) for 2 min, incubated at room temperature (21 ± 1°C) for 1 h, mixed by vortex for 30 s and the liquid squeezed out from the swab using forceps, centrifuged at 3000 g for 10 min, filtered [0·45-μm cellulose acetate syringe filter (MSR, Shanghai, China)] and then diluted with TSB supplemented with 2 mmol l−1 of calcium chloride (Ca2Cl2) (Scharlau Chemie) to give a final concentration of c. 0·6% (w/v) ME. The sample filtrates (10 ml) were inoculated with 100 μl of exponential-phase (c. 5 × 108 CFU ml−1) L. monocytogenes strains 19CO9 and 18AO1 (indicator hosts) separately, incubated at 25 ± 1°C for 48 h and centrifuged at 3000 g at 4°C for 20 min, and the supernatants were treated with chloroform (CHCl3) (Analytical Grade, Merck) at 1% (v/v) for c. 1 h in the dark to destroy old host cells and other potential contaminant bacteria passed through filtrate into sample enrichment (Walakira et al. 2008). The bacteria-free sample enrichment (lysate) was assayed by the double-layer agar (DLA) plate method (Kropinski et al. 2009) using a TSAYE [TSB plus 0·6% yeast extract (Scharlau Chemie)] overlay containing 0·4% agar (Difco) plus CaCl2 (2 mmol l−1) and TSAYE base layer (1% agar). The plates were examined for phage plaques formed at 25 ± 1°C after 18–48 h.
From phage-positive plates, individual plaques with a clear centre were transferred using a pipette tip into 1 ml saline–magnesium buffer containing gelatine (SM) [100 mmol l−1 NaCl (Merck), 50 mmol l−1 Tris–HCl (Sigma-Aldrich, Steinheim, Germany), 8 mmol l−1 of MgSO4 (Merck), 0·001% of gelatine solution (LabChem, Auburn, NSW, Australia) at pH 7·5] and stored at 4°C. The phage isolates were streaked individually using a tooth pick on base agar plates (Serwer et al. 2004), and overlay containing L. monocytogenes 19CO9 was poured on the phage streaks, and plates (DLA method) were incubated at 25 ± 1°C for 24 h. The selected phage plaques were purified over four cycles by directly streaking the phages from a well-separated plaque on fresh agar plate as mentioned above. Purified phage plaques were picked by boring the agar using a filter tip (Labcon filter tips, Interlab, Wellington, New Zealand) of c. 1·5 mm diameter. The agar plug was transferred into a labelled glass bottle containing 2 ml of SM and 50 μl of CHCl3 and stored in the dark at 4°C.
A DLA plaque assay was performed for each phage isolate on two types of overlays [0·4% agar and 0·1% agarose (Sigma-Aldrich)] with base layer of TSA containing 1% agar (Serwer et al. 2004) using L. monocytogenes 19CO9 as indicator host. Plaque morphology (appearance and diameter) was observed on both overlay types after incubation at 25 ± 1°C for 24–48 h.
Phages (new isolates and A511) were propagated separately by the DLA method using L. monocytogenes 19CO9 in 2-l flasks (Sillankorva et al. 2004) at 25 ± 1°C. The phage lysate was centrifuged at 8000 g for 30 min at 4°C (Biofuge, Primo R, Heraeus, Conquer scientific, San Diego, CA, USA). The supernatant was treated with CHCl3 (c. 200 ml lysate per 0·5 ml CHCl3) for c. 18 h and then centrifuged at 8000 g for 30 min. About 50 ml of clear phage lysate was ultracentrifuged (SS-34 rotor, Sorvall RC6 + Centrifuge, Thermoscientific Sorvall, Frankfurt, Germany) at 25 000 g for 6 h (Ackermann 2009a). The phage pellet was suspended in 1 ml of SM for 18 h at 4°C followed by agitation at 22 ± 1°C with shaking (60 rev min−1) for 1–2 h in the dark. The concentrated phage suspension was transferred into glass bottles and stored with a few drops of CHCl3 at 4°C. The titre of the phage stock was assayed over a 10-fold dilution series in SM using DLA plate method as described above.
The phages were discriminated based on their dependency on free Ca2+ in growth medium by DLA plaque assay using both agar layers supplemented with the same concentration of citrate (trisodium citrate; Biolab Ltd, Mulgrave, Vic., Australia) at either 0·5 or 1 mol l−1 (Olsen et al. 1968). Control DLA plates containing agar without citrate supplements were also inoculated with phages. Triplicate plates were inoculated for each treatment, and plates were incubated at 25 ± 1°C for 24–48 h. The efficiency of plating (EOP) of phage isolates was calculated with respect to plaque counts on noncitrate plates (Kutter 2009).
Volumes (2 ml) of phage lysates containing c. 106 PFU ml−1 were heated at 50 ± 1 and 60 ± 1°C in thin-walled narrow glass vials (Ø, 1·5 cm and height 6 cm) using the method described by Ackermann et al. (1978). Viable phage titre was assessed by DLA plaque assay (detection limit = 1 log10 PFU ml−1). Three independent experiments were carried out.
Phage lysate (c. 109 PFU ml−1) was purified in 0·1 mol l−1 of ammonium acetate (Scharlau Chemie) by ultracentrifugation (25 000 g) for 2 h (Ackermann 2009a), and the phage pellet was suspended in c. 0·5 ml of SM as previously described. The purified phage suspension (>1010 PFU ml−1) was adsorbed onto the carbon-coated copper grids stained with 2% aqueous uranyl acetate solution. The stained phages were then observed by transmission electron microscopy (TEM) (Philips CM12; FEI Company, Eindhoven, Netherlands) (Zink and Loessner 1992), and the images with scale bars were photographed (Model 792 Bioscan Inc., Poway, CA, USA). Dimensions of the phage images were measured using Image J 1.45s software (National Institute of Health, Bethesda, MD, USA).
The DNA of phages including A511 as a control (Loessner et al. 1994) was extracted from high-titre phage lysates (>109 PFU ml−1) as described by Pickard (2009) and individually digested with Cla1 and Sac1 (NEB, New England Biolabs, Ipswich, MA, USA) restriction enzymes at 37 ± 1°C for 1·5 h. Genomic profiles of digested DNA phage were separated in a 0·7% agarose gel (Sigma-Aldrich) followed by staining with 1 μg ml−1 ethidium bromide (Sigma-Aldrich) for 30 min. HindIII digest of phage lambda (λ) DNA (NEB) was used as a molecular ladder. Images were taken using a fluorescence imaging system (Gel Doc™ 1000 Bio-Rad, Inc., Hercules, CA, USA). Three replicates were performed for each restriction enzyme.
Using the DLA method (TSA of 0·4% agar overlay and 1% agar base), the host range of each phage strain was determined by lysis zone formation on the spots of 20-μl phage drops (drop test) or plaques (plaque assay) in host lawns of 23 Listeria strains at 15 ± 1 and 25 ± 1°C (Carlson 2005; Kutter 2009). The EOP of a phage using a test Listeria strain was calculated with respect to plaque counts formed by L. monocytogenes 19CO9 (Kutter 2009). The test host strains in which phages did not form visible plaques in plaque assays were also assessed in overlays supplemented with 5 mg l−1 of ampicillin (ampicillin sodium salt; Sigma-Aldrich), which increases infection vigour in the host causing visible plaques on host lawns (Santos et al. 2009). Each host was tested in duplicate in three independent experiments.
Exponential-phase cultures of L. monocytogenes 19CO9, 19DO3, 18DO5 and 18DO7 strains (c. 107–108 CFU ml−1) were centrifuged at 8000 g for 10 min, resuspended in phosphate-buffered saline (PBS) [0·008 mol l−1 Na2HPO4 (Merck), 0·001 mol l−1 NaH2PO4.H2O (Merck) and 0·145 mol l−1 NaCl; pH 7·5] and then mixed with each phage (c. 104–105 PFU ml−1) suspended in PBS (Gallet et al. 2009). The coculture was then incubated at 15 ± 1°C, and 1-ml volumes were withdrawn into chilled plastic tubes containing CHCl3 (100 μl) at 1 to 2-min intervals, mixed by vortexing for 20 s, incubated at room temperature in the dark for at least 1 h and then centrifuged at 8000 g for 10 min. The unadsorbed phage titres of samples were determined by the DLA method using L monocytogenes 19CO9. Initial counts of each host strain cultures were enumerated using the standard spread plate technique with plate incubation at 30 ± 1°C for 48 h. The k value of phage was calculated based on the equation given by Hyman and Abedon (2009).
Three phages were assessed for plaque formation by performing DLA method as described in the host range experiment between 4 and 37°C using exponential-phase broth cultures of indicator strain (L. monocytogenes 19CO9) grown at 4 ± 1°C for 20 day and at 15 ± 1, 25 ± 1 and 35 ± 1 or 37 ± 1°C for 48, 24 and 18 h, respectively. The plaque counts were obtained by incubating the inoculated plates at each temperature for a further 2–3 day more than the incubation time used to grow the exponential-phase broth cultures at respective temperatures. Six replicate plates were prepared at each temperature. The EOP of phage at each temperature was calculated with respect to plaque counts at 25 ± 1°C.
Single-step replication of three phages was determined at 15 ± 1, 25 ± 1 and 35 ± 1°C. Briefly, a 10 ml exponential culture of L. monocytogenes 19CO9 (c. 108 CFU ml−1) was infected with each phage to reach an input multiplicity of infection value of 0·1 and mixed well. The phage adsorption (initiation of the infection) in the coculture was allowed at 15 ± 1°C for 10 min and at 25 ± 1 or 35 ± 1°C for 5 min with shaking (60 rev min−1). The coculture was immediately diluted 10-fold with pre-incubated TSB at 15 ± 1, 25 ± 1 or 35 ± 1°C and mixed well to prepare 300-ml volumes of 10−4 and 10−5 dilutions. The diluted coculture was swirled (60 rev min−1) throughout the sampling period at 25 ± 1 and 35 ± 1°C, while the coculture was incubated at 15 ± 1°C statically and shaken well before each sampling. Two series of samples were withdrawn from 10−4 and 10−5 dilutions of a coculture at 25 ± 1 or 35 ± 1°C in every 10 min and at 15 ± 1°C in every 30 min. A series of samples was assayed for plaques directly by the DLA method, while the other set was immediately treated with CHCl3 (1% v/v) and then assayed for plaques (Carlson and Miller 1994). Single-step growth parameters were calculated following the method described by Carlson and Miller (1994) and Hyman and Abedon (2009).
Descriptive statistical analyses were performed using Microsoft Office Excel 2007 (Microsoft Office, Redmond, WA, USA) and SAS univariate procedure (SAS 9.1 version, SAS Institute, Cary, NC, USA). One-way anova was performed on the different characteristics or treatments to determine significance (P < 0·05), and significant treatments were separated using post hoc Duncan's multiple range test (SAS 9.1).
When field samples were phage positive, plaques were visible in lawns of both indicator host strains (L. monocytogenes 19CO9 and 18AO1). The phage plaques were consistently larger on lawns of L. monocytogenes 19CO9 than on lawns of 18AO1, and the plaques on the former lawn were selected for further studies. Based on different plaque characteristics (diameter and appearance), two types of phage isolates were identified on the same sample plate originating from the milliscreen drain receptacle. A phage from a plaque (Ø c. 1·0–1·5 mm) was designated as LiMN4L, while the other phage from another plaque (Ø c. 0·7–1·0 mm) was designated as LiMN4p (Table S1). Another phage, named LiMN17, was derived from a plaque (Ø c. 0·5–1·0 mm) picked from a sample plate originating from the final sump (Table S1). Plaques were transferred separately into SM and purified by restreaking.
The three phages formed plaques with narrow haloes, which looked similar for all three phages on a 0·4% agar overlay after incubation at 25 ± 1°C for 48 h (Table S1). The three phages produced comparatively larger plaques with haloes of different appearance on 0·1% agarose overlay plus 1% agar base (Fig. 1). With these double layers, phage LiMN4p produced plaques consisting of one halo (Fig. 1b), while phages LiMN4L and LiMN17 formed plaques with additional secondary haloes (Fig. 1a,c). However, all three phages produced secondary haloes in the 0·1% agarose overlay when a 1·5% agar base was used (data not shown). Phage A511 also showed plaque morphology similar to LiMN4p on 0·1% agarose overlay (data not shown).Considering the plaque characteristics and different locations of the field samples, the three phages LiMN4L, LiMN4p and LiMN17 were characterized further.
Three phages were discriminated at the highest level in 1 mol l−1 citrate agar compared with 0·5 mol l−1 agar (Fig. 2). Phages LiMN4p and LiMN17 showed the lowest (c. 54%) and highest (c. 85%) EOP values, respectively (P < 0·05) at 1 mol l−1 of citrate concentration (Fig. 2). The % EOP values of LiMN4L at both citrate concentrations were not significantly different (P > 0·05) (Fig. 2).
At 50 ± 1°C, phage LiMN17 was undetectable (<1 log10 PFU ml−1) after 10 min of incubation, while phage LiMN4p was reduced by c. 3-log units. Comparatively, c. 17, 12 and 7% of phage LiMN4L survived at 50 ± 1°C for 10, 30 and 60 min, respectively. Phage LiMN4L was recovered at levels of c. 9, 0·04 and <0·0001% at 60 ± 1°C after 10, 30 and 60 min, respectively. Phages LiMN4p and LiMN17 were more heat labile and did not survive (<1 log10 PFU ml−1) after 10 min at 60 ± 1°C (Table S2).
According to the TEM images, each phage was comprised of an icosahedral head and a long, rigid, contractile tail with tail fibres. All three therefore belonged to the order Caudovirales and family Myoviridae (Ackermann 2009b) (Fig. 3). Dimensions of the phage were calculated over at least 20 phage particles, which were within the interquartile range (between Q1 and Q3), and three phages were different in dimensions. The mean head diameter of phage LiMN4L was significantly larger (P < 0·05) than phages LiMN4p and LiMN17 (Fig. 3). Phage LiMN4p had a significantly longer tail than the other two phages (P < 0·05). The tail width of phage LiMN4L was narrower (P < 0·05) than phages LiMN4p and LiMN17 (Fig. 3).
Restriction digest profiles of phages LiMN4L, LiMN4p and LiMN17 with Sac1 (Lanes 2–4) were compared with phage A511 in lane 5 (Fig. 4). Profiles given by Sac1 digestion confirmed that phage LiMN4p was different from the other two phages and from phage A511, but phage LiMN4p showed some relatedness to phage A511 (Fig. 4). Phages LiMN4L and LiMN17 had similar enzyme digestion profiles. Phage LiMN4p was also distinguished from the other two using the Cla1 enzyme (data not shown).
The three isolated phages infected many of the Listeria strains (Table 1) but did not produce plaques in lawns of strains 19EO1, 19EO3 and 19DO9 using the standard DLA method. However, at 104–108 PFU ml−1, they formed pinpoint-sized plaques with these strains in agar overlays supplemented with ampicillin (0·05 mg l−1). The minute plaques were difficult to enumerate exactly; therefore, the EOP of the three phages was estimated to be in the range of c. 10−2–10−6 (Table 1). Phage LiMN17 did not form plaques by plaque assay, but phage drops (>109 PFU ml−1) resulted in clear lytic zones on the spots of original phage drops on lawns of L. monocytogenes SMAC, 19DO9, 19DO3 and L. ivanovii strain. Three phages did not form plaques on the lawns of 18DO5 and 18DO7 strains, but formed hazy lytic zones (‘host kill zones’), which were not completely clear lytic zones and contained turbidity throughout the zones on the spots of original phage drops (>109 PFU ml−1) on the lawns of 18DO5 and 18DO7 strains (Carlson 2005; Kutter 2009) compared with blank control drops containing SM only (Table 1).
Phage LiMN4L, however, at a titre of >109 PFU ml−1, formed 1–2 plaques per plate on the lawns of strains 18DO5 and 18DO7. Phages LiMN4L and LiMN4p infected both L. ivanovii and L. innocua, while phage LiMN17 was found to infect only L. innocua (Table 1). Phages LiMN4L, LiMN4p and LiMN17 were distinguished from each other by the EOP with L. monocytogenes SMAC91 and 18FO9. Phage LiMN4p showed c. 94 and 85% EOP with both strains, respectively, at 15°C. The EOPs of phage LiMN4L with two strains (SMAC91 and 18FO9) were c. 78 and <0·001%, respectively, at 15°C. Phage LiMN17 was unable to make plaques with strain SMAC91 even with titres >109 PFU ml−1 at both 15 and 25°C (Table 1). Based on EOP range of 10−5–100%, the host ranges of phages LiMN4L, LiMN4p and LiMN17 were 21/23 (c. 91%), 21/23 (c. 91%) and 17/23 (c. 74%), respectively (Table 1), at 15°C.
The k values of three phages on L. monocytogenes 19CO9, 19DO3, 18DO5 and 18DO7 ranged between c. 1 × 10−9 and 3 × 10−11 ml min−1 (Table 2). The three phages showed varying% EOP values between <10−7 and 100% with the four strains (Table 1). All three phages had k values >10−9 ml min−1 with the indicator strain 19CO9 giving EOP of 100% (Tables 1 and 2). The phage LiMN17 did not infect (<10−7% EOP) strain 19DO3 (Table 1). However, LiMN17 showed high adsorption rate (10−9 ml min−1) on 19DO3 similar to 19CO9 (Table 2). Therefore, % EOP-based host range (74%) of phage LiMN17 was broadened to 18/23 (78%) based on the high adsorption rates obtained for an additional strain (L. monocytogenes 19DO3). The k values of three phages were low (10−10–10−11 ml min−1) on 18DO5 and 18DO7 with which EOP values were <10−7% (Table 2).
|Listeria monocytogenes strain||k (ml min−1) of listeriaphage (Meanb ± SEMc)|
|19CO9a||(8·93a ± 0·46) × 10−9||(3·16c ± 0·97) × 10−9||(2·42d ± 0·27) × 10−9|
|19DO3||(1·44e ± 0·27) × 10−9||(7·35b ± 0·18) × 10−9||(1·56e ± 0·43) × 10−9|
|18DO5||(5·26f ± 0·70) × 10−11||(4·10f ± 0·06) × 10−10||(1·17f ± 0·15) × 10−10|
|18D07||(7·54f ± 2·12) × 10−11||(4·82f ± 1·16) × 10−11||(3·24f ± 0·63) × 10−11|
Phages LiMN4L and LiMN4p produced plaques at 4, 15, 25, 30 and 35°C, and the % EOPs of both phages, however, remained <50 and <4 × 10−4 at 30 and 35°C, respectively (Table S3). Neither phage formed plaques but did form lysis zones on host lawns at 37°C. Phage LiMN17 formed plaques between 4 and 30°C and formed lysis zones with phage drops (≥4 × 106 PFU ml−1) on host lawns at 35 and 37°C. Phages LiMN4L and LiMN4p adsorbed to host at 37°C, while phage LiMN17 adsorbed at 35 and 37°C even though they did not make productive infections in the preliminary trials of single-step replication experiments.
|Listeriaphage||Temperature (°C)||Single-step growth parameters|
|Latent period (min)||Rise period (min)||Eclipse period (min)||Burst size (PFU per cell)|
|LiMN4L||15||240||150||150||17 ± 5|
|25||90||40||60||30 ± 4|
|35||130||30||NIa||5 ± 2|
|LiMN4p||15||270||90||150||17 ± 4|
|25||90||35||60||37 ± 3|
|35||130||30||NIa||4 ± 2|
|LiMN17||15||270||120||180||11 ± 1|
|25||90||40||60||20 ± 1|
Three phages were isolated from thick scum samples collected on swabs from different locations of a seafood-processing waste treatment unit. The three phages were eluted from the thick scum and incubated in 10% ME to facilitate desorption of phages by intercalating the small protein molecules between phage particles and hard surfaces, thereby weakening the mutual attractive electrostatic forces (Wait and Sobsey 1983). In the next analytical step, the sample extract was diluted with TSB to c. 0·6% ME in order to avoid any potential concentration effect by 10% ME. Some phages are dependent on cofactors such as divalent cations (c. 1–10 mmol l−1) for their attachment on host or intracellular growth (Twest and Kropinski 2009). Therefore, the sample enrichment was supplemented with Ca2+ (2 mmol l−1). In isolation protocols for other listeriaphages, c. 1·25–10 mmol l−1 Ca2+ was used (Hodgson 2000; Kim et al. 2008; Lee 2008).
Phage isolates were examined for various physical properties to obtain more distinguishing characteristics. Large plaques of phages can be obtained by using low concentration (0·3%) of agar (Kropinski et al. 2009). Pseudomonas chlororaphis-specific phages (201Ф2-1 and 201Ф2-2), which had different dimensions of about twofold in both head diameter and tail length, were discriminated based on plaque size in low strength gel (0·1% agarose) (Serwer et al. 2004). However, by following the same method, the three phages of the present study could not be distinguished based on plaque size in 0·1% agarose (Fig. 1 and Table S1). The TEM images of the three phages showed significant (P < 0·05) differences in the dimensions although they were less than the two-fold differences reported by Serwer et al. (2004). Nevertheless, use of agarose at low concentration permitted discrimination of phages by the appearance of plaque haloes (Fig. 1).
Loessner and Busse (1990) also reported large plaques of Listeria-specific phage A513 containing one or more zones of secondary lysis. In phage infections, lysis of the host cell wall is mediated by phage endolysins. Endolysins are produced in excess during phage replication (Wang 2006) and degrade the cell wall of Gram-positive bacteria, inclusive from the exterior of the cell wall because they lack an outer envelope (Schmelcher et al. 2012). Therefore, free endolysins may have diffused out of the plaque and lysed cell walls of neighbouring uninfected cells resulting in haloes around the plaques in the soft agar overlay in the present experiments. Similarly, listeriaphage endolysin (ply 118) expressed in E. coli 109(DE3), diffused from the dead E. coli colonies and lysed the adjacent cells of L. monocytogenes 1001-lawn giving haloes in the agar overlay (Loessner et al. 1995). The difference in appearance of haloes between phage LiMN4p and the other two phages (Fig. 1) might be due to differences in heterogeneity of phage endolysins.
A group of Pseudomonas-specific phages was distinguished in agars containing citrate (different concentrations), which is capable of depleting free Ca2+ ions in the growth medium (Olsen et al. 1968). Using this methodology, the three phages had significantly different EOP values (Fig. 2). Uniquely, phage LiMN17 had high EOPs at both 0·5 and 1 mol l−1 citrate levels indicating its reduced dependency on free Ca2+ compared with the other two phages. Alternatively, in adsorption assays, the adsorption rate (k) of phage LiMN17 was lower in the presence of Ca2+, while the k values of phages LiMN4L and LiMN4p were increased at 5 and 10 mmol l−1 levels of Ca2+, respectively (data not shown). Based on the percentage survival of phage at 60°C, a group of phage isolates of Pseudomonas was distinguished into two groups (Olsen et al. 1968). One group contained the psychrotrophic phages that seemed to diminish to nondetectable levels at 60°C after 10 min. The other contained mesophilic phages of which over 40% survived after 30 min at 60°C (Olsen et al. 1968). The three phages in the present study were more heat labile because only LiMN4L survived up to c. 10% at 60°C for 10 min, and the other two phages reduced to nondetectable levels in 10 min. Similar to the phages discussed here, phages A511, P100 (Klumpp et al. 2008) and FWLLm1, FWLLm3 and FWLLm5 (Lee 2008) also belong to family Myoviridae.
The Cla1 and Sac1 DNA restriction digestion profiles of listeriaphage A511 (Loessner et al. 1994), P35 (Dorscht et al. 2009), P100, 20422-1 and 805405-1 (Kim et al. 2008) were different from the phages studied here. From restriction enzyme digestion profiles, phage LIMN4p was distinguishable from phages LiMN4L and LiMN17 (Fig. 4), and these latter two may be closely related. The host range of phages is suggested as a medium level criterion to distinguish new phage isolates (Ackermann et al. 1978). Based on the host ranges presented here, the three phages are different from each other. Santos et al. (2009) demonstrated productive infections by triggering the formation of visible plaques on host lawns using agar supplemented with different antibiotics at suboptimal levels. In this study, three phages formed tiny visible plaques on lawns of some host strains in agar containing ampicillin (Table 1). The adsorption of a phage to host strains may be used to define its host range (Hyman and Abedon 2010). The results of the adsorption assay therefore indicate that the low levels of EOP of phages in different host strains did not directly correlate with the adsorption rates and that they might be influenced by other phenomena such as blocking of the uptake of the phage genome, restriction modification, abortive infection, reduced infection vigour and interference of phage dissemination (Hyman and Abedon 2010). However, further studies are required to investigate the viable counts of cultures immediately after phage adsorption.
The magnitude of the k value is a function of the structure of adsorption appendages of both phage and bacteria in addition to the extrinsic parameters such as cofactors, temperature and viscosity of medium (Gallet et al. 2009; Hyman and Abedon 2009). Phages with side tail fibres show high k values in the range of c. 10−8–10−9 ml min−1 (Gallet et al. 2009). The three novel phages of this study also had side tail fibres (Fig. 3) and showed high k values on some host strains, while low k values (10−10–10−11 ml min−1) were observed on other strains (Table 2). Therefore, significantly lower k values (P < 0·05) of the same phages may be attributed to the cell wall characteristics of particular host strains. The low adsorption rates of E. coli-infecting phages may be related to the presence of phage receptors at low densities on the host cell wall (Schwartz 1976; Kasman et al. 2002).
The phages used in this trial were isolated at 25°C to target phages favouring growth at low temperatures (Olsen et al. 1968), because the phages were intended to kill Listeria at low temperatures in seafood-processing plants. Similarly, listeriaphage P35 (ФLMUP35) is reported to have been isolated and propagated at room temperature (Hodgson 2000; Dorscht et al. 2009). There are reports of other psychrotrophic nonlisteriaphages, such as those that infect Pseudomonas spp. and Brochothrix thermosphacta, which have been isolated using similar low-temperature protocols (Olsen 1967; Olsen et al. 1968; Greer 1983). Not surprisingly, the three phage isolates investigated in this study showed productive infections between 4 and 30°C, while phages had low productivity or were nonproductive at 35°C (Table S3). Phage LiMN17 did not infect host cells productively at 35°C, unlike the other two phages (Table S3). Contrary to the results of the present study, Kim and Kathariou (2009) reported that listeriaphages isolated by analysing the samples at 37°C had the highest host ranges and high EOPs at temperatures >30°C. The three phages replicated optimally at 25°C. At 15°C, three phages replicated at low threshold for longer latent periods (240–270 min) than latent time (90 min) at 25°C (Table 3). Loessner et al. (1995) reported that the latent time of phages A511 and A118 was 55 and 65 min, respectively, at 30°C. Nevertheless, currently, detailed information on the single-step replication of listeriaphages is not abundant. Phage LiMN17 may be more psychrotrophic, as demonstrated by its inability to make productive infections at 35°C, unlike the other phages. The lowest survival at 60°C, high EOP value at low temperatures and single-step replication at <35°C may be attributed to psychrotrophic characteristics of three phages adapted to low-temperature habitats. In conclusion, the three psychrotrophic phages isolated from seafood-processing premises had broad host ranges against seafood-borne L. monocytogenes strains at low temperature and showed low burst size. The results suggest that three virulent phages may be suitable as decontaminating agents under ambient conditions of seafood-processing plants using passive biocontrol strategies, which will be investigated in future studies.
The authors acknowledge New Zealand King Salmon for funding the project; Foundation for Research, Science and Technology, New Zealand, for providing a scholarship [Contract No.: NZKX0902] to Geevika Ganegama Arachchi; and Cawthron Institute, Nelson, for providing laboratory facilities as well as technical support. The authors wish to thank Professor P. Bremer at Otago University, Dunedin; Mr G. Fletcher at Plant & Food Research Ltd, Auckland; Ms. Anneke van Laanen, Cawthron Institute, Nelson; and Prof. Martin J. Loessner at Institute of Food, Nutrition and Health, Switzerland, for providing bacterial strains and bacteriophage A511 for this project. Mr. Ron Fyfe at Cawthron Institute in Nelson, Mr. Richard Smith, ex-company mentor in New Zealand King Salmon, Mr. Graeme Fox in Sealord Group Ltd in Nelson, Mrs. Karen Wittington in New Zealand King Salmon Company, Dr. Andrew Cridge and Ms. Rachel Liu at Massey University for providing technical support. Authors also acknowledge the anonymous reviewers.
No conflict of interest declared.