Growth inhibitory effect of bacteriophages isolated from western and southern coastal areas of Korea against Vibrio parahaemolyticus in Manila clams

Vibrio parahaemolyticus is one of the leading causes of seafood-related illnesses in Korea and the whole world. The growth inhibitory effect of V. parahaemolyticus-specific phages (Vpp) in clam samples was investigated. Six bacteriophages specific to V. parahaemolyticus were isolated by the agar overlay method using sediment and seafood samples from the western and southern coastal areas of Korea. Host range, restriction digestion pattern, and transmission electron microscope images of phage isolates were examined. The most effective phage, Vpp2, at a multiplicity of infection (MOI) of 10 showed 4.2 log CFU/mL reduction in V. parahaemolyticus growth at 30 °C and 6 log CFU/mL reduction after the incubation at 37 °C for 6 h. In a food application study, Vpp2 at an MOI of 100 demonstrated a 2.1 log CFU/mL reduction at 25 °C after a 24-h incubation in shell-less Manila clams spiked with V. parahaemolyticus. Additional studies are needed to characterize the newly discovered Vpp and their potential applications as biocontrol agents in food.


Introduction
Vibrio parahaemolyticus is a halophilic gram-negative bacterium that occurs in sea water and estuaries. It causes gastroenteritis with severe abdominal pain and diarrhea via the ingestion of raw or undercooked seafood and cross contamination from unsanitary cooking environments (Su and Liu 2007;Robert-Pillot et al. 2010). It is also one of the leading causes of seafood-related illnesses in Korea. From July to September, V. parahaemolyticus accounted for 14 of 18 total outbreaks (79 %) in 2010, 8 of 9 outbreaks (89 %) in 2011, 9 of 11 outbreaks (82 %) in 2012, 5 of 5 outbreaks (100 %) in 2013, and 7 of 7 outbreaks (100 %) in 2014 (KMFDS 2015); therefore, the species requires additional research and monitoring, particularly during the summer, as does V. vulnificus.
Various physical and chemical control techniques effectively inhibit V. parahaemolyticus growth; however, they could also cause changes in the physicochemical and sensory properties and the freshness of seafood and seafood products (Lin et al. 2005;Ren and Su 2006;Su and Liu 2007;Chae et al. 2009;Larsen et al. 2013;Rong et al. 2014). Furthermore, as many consumers in Korea and Asia prefer raw or partially cooked seafood, the continual development of novel control methods against V. parahaemolyticus contamination is needed.
Lytic bacteriophages are viruses that invade target bacteria, disrupt bacterial metabolism, and cause bacteria to lyse. Bacteriophage and phage-derived antimicrobials have many potential applications in the area of food safety, such as in phage therapy for livestock, biosanitation and biocontrol in the foodstuff industry, biopreservation in storage and marketing, and aquaculture (Hudson et al. 2005;García et al. 2008;Alagappan et al. 2010;Rong et al. 2014). Bacteriophages have important functions as alternative antibiotic agents against threats of antibiotic-resistant food-borne pathogenic bacteria, despite a number of future challenges, such as host range limitations, better knowledge of phage-host physiology, issues of consumer acceptance, and so on (Hudson et al. 2005;García et al. 2008). The initial approval for the application of phages specific to Listeria monocytogenes for food safety purposes (i.e., LISTEX P100 TM ) has accelerated research on the use of phage against other food pathogens (García et al. 2008). However, reports on food protection using phages and their lytic mechanism against V. parahaemolyticus are rare, except in the area of aquaculture. Applications of bacteriophages to seafood as non-thermal treatments are expected not only to inhibit the growth of Vibrio species, but also to maintain their desirable characteristics. The exploration and selection of appropriate bacteriophages are key factors in the success of phage-mediated control of bacterial infections (Rong et al. 2014).
The purpose of this study was to characterize V. parahaemolyticus-specific phages isolated from sediment and seafood samples and to examine the growth inhibitory effect of these phages against V. parahaemolyticus for use as biocontrol agents in foods.

Bacterial strains and characterization
All reference strains were donated from the Korea Agricultural Culture Collection, Korea Culture Center of Microorganisms, and Korea Collection for Type Cultures. V. parahaemolyticus reference strains, ATCC 27969 and 17802, and V. parahaemolyticus isolates were grown at 37°C on Tryptic Soy Agar (TSA) and Tryptic Soy Broth (TSB) supplemented with 1 % NaCl.

Vibrio parahaemolyticus-specific phage preparation
Six V. parahaemolyticus-specific lytic bacteriophages were isolated using the agar overlay method (Baross et al. 1978;Comeau et al. 2005). One hundred microliters of the 24-h enrichment culture (TSB) filtered through a 0.45-lm membrane was mixed with 100 lL of the host strain (reaching an optical density of 1.0 at 600 nm from the overnight culture) and 3 mL of 0.7 % soften TSA. The mixture was then immediately overlaid on the TSA. The plates were incubated at 37°C overnight, and the number of plaques were counted and expressed as plaque-forming units per milliliter (PFU mL -1 ). V. parahaemolyticus ATCC 27969 and 17802, and V. parahaemolyticus isolates were used as hosts for the assays. A single plaque was selected and resuspended in SM buffer (5.8 g of NaCl, 2 g of MgSO 4 Á7H 2 O, 50 mL of 19 phosphate buffered saline, pH 7.8 in 1 L), after which it was centrifuged at 10,0009g for 10 min at 4°C and filtered through a 0.45lm membrane to remove intact bacteria and bacterial debris. Three successive single-plaque isolations were performed to obtain the pure phage stock. The stocks were stored at 4°C with the addition of 1 % chloroform for future studies.

Host range determination of phages
The lytic spectrum of phage based on the extent of host range was tested as follows. One hundred microliters of a phage sample [multiplicity of infection (MOI), 0.001] was added to 100 lL of bacterial overnight culture, and the agar overlay method was used. The presence of phage plaques as indicators of bacterial lysis was checked after 18-24 h of incubation.

Restriction digestion pattern of the phage genome
The phage DNA was extracted using the Genomic DNA Extraction Kit (Qiagen Tissue and Blood Kits, Chatsworth, CA, USA) according to the manufacturer's instructions. After DraI (TaKaRa, Shiga, Japan) digestion, the DNA fragments were separated by electrophoresis in a 0.8 % agarose gel in TAE buffer at 80 V for 40 min.

Phage morphology
The phage suspension (10 9 PFU/mL) was centrifuged with a 40 and 10 % glycerol gradient at 35,000 rpm for 90 min at 4°C. After dialysis against SM buffer, phage samples were deposited on formvar-coated copper grids (Ted Pella, Inc., USA) fixed with 2.5 % glutaraldehyde for 2 h and washed twice with phosphate buffer (pH 7.4). All phages were examined using a Tecnai F20 Cryo transmission electron microscope operated at 60 kV at a magnification of 920,000 after they were negatively stained with 2 % uranyl acetate.

V. parahaemolyticus growth inhibition by the phage in culture
After 100 mL of TSB with 1 % NaCl was inoculated with fresh overnight bacterial culture to reach 1 9 10 4 CFU/ mL, 100 lL of purified phage suspension at an MOI of 0.1, 1, and 10 was added to the bacterial culture and incubated with shaking for 12 h at 25, 30, and 37°C. One milliliter of culture medium was collected every 2 h and used to count the colonies grown on plate count agar with 2 % NaCl. The growth inhibitory effect of the phage was measured by a comparison with the CFU of the control group, which was not treated with phage.
V. parahaemolyticus phage application to spiked seafood samples Live Manila clams were purchased from a local market and transferred to the laboratory immediately. Fresh shell-less Manila clam samples were washed with three volumes of sterile water three times, and 10 g of sample was placed into a sterile stomach plastic bag with membranes. The samples were spiked with 100 lL of diluted overnight host bacterial cells (1 9 10 4 CFU/mL) and then 100 lL of each purified phage suspension at an MOI of 1, 10, or 100 was applied to the contaminated samples, which were incubated for 24 h at 4, 15, and 25°C. To measure the viable cells in each treatment, an aliquot of 1 mL was collected at 0, 2, 4, 6, and 24 h. After adding 99 mL of alkaline peptone water, serially diluted samples were plated on a TCBS agar plate and incubated for 24 h until the green-blue colonies appeared at 36°C.

Statistical analysis
Results of the experiment performed in triplicate are presented as mean ± SD. The statistical significance among treatments was analyzed using a Student's t test. P values of less than 0.05 were considered statistically significant.

Results and discussion
Isolation of Vibrio parahaemolyticus-specific phages Bacteriophages specific to V. parahaemolyticus were isolated using sediment and seafood samples collected in the western and southern coastal areas from March to September of 2013. Seafood samples, such as oysters and clams, salt-fermented seafood, and other sources were collected from marine environments or from local fish markets. The 6 presumptive phages specific to each V. parahaemolyticus isolate, which were obtained from pure subcultures using three successive single-plaque isolations, were acquired : V. parahaemolyticus phage (Vpp) 1 from an oyster in Gwangyang, Jeonnam, Vpp2 from sediment in Ganghwa, Incheon, Vpp3 from a scallop in Tongyeong, Geongnam, Vpp4 from a warty sea squirt in Tongyeong, Geongnam, Vpp5 from sediment in Dangjin, Chungnam, and Vpp6 from sediment in Siheung, Gyeonggi, Korea.

Characterization of Vibrio parahaemolyticus isolates
Eight V. parahaemolyticus strains were isolated from seawater and clam samples using selective media, TCBS/TSI agar, or CHROMagar Vibrio, and were identified biochemically using the VITEK 2 systems (Table 1). The species-specific genes of toxR and tl were detected in all isolates, including two V. parahaemolyticus reference strains (ATCC 27969 and 17802). However, none of the isolates were verified as pathogenic; the virulence genes tdh and trh genes were not detected (Table 1). Only V. parahaemolyticus reference strain ATCC 17802 had the virulent trh gene. These results confirmed the results of previous reports indicating that most V. parahaemolyticus isolates from environmental and seafood samples are tdhand trh-negative and are not pathogenic; 1-6 % of samples contain tdh-and trh-positive isolates (Lee et al. 2008;Rosec et al. 2009). According to the results of an antimicrobial resistance test, V. parahaemolyticus was not resistant against fifteen antibiotics, but was resistant to ampicillin/sulbactam.

Characterization of Vibrio parahaemolyticus-specific phages
The lytic spectrum of phages was investigated against 3 Gram-positive and 22 Gram-negative strains (Table 2). Six phages showed no lytic activity against major food-borne pathogens, such as Escherichia coli O157:H7, Salmonella strains, Staphylococcus aureus, Listeria monocytogenes, and Bacillus cereus. All 6 phages showed lytic activity against Vibrio species (i.e., either the reference strains or isolates), and had a narrow lytic spectrum. Only Vpp2 had lytic properties against V. cholera and V. mimicus as well as V. parahaemolyticus. The restriction analysis of the phage genome using DraI revealed that the pairs Vpp1, Vpp2, Vpp3, Vpp4, Vpp5, and Vpp6 showed similar patterns (Fig. 1). However, the host range analysis of the three phage pairs with similar pattern revealed differences between the two phages in each pair (Table 2). Vpp1 and Vpp2 were isolated from different sources in distant locations; they had similar restriction digestion patterns, but divergent host ranges, similar to the pattern observed for Vpp5 and Vpp6. In an analysis of phage morphology based on transmission electron microscopy (Fig. 2), Vpp2 belonged to Podoviridae with short extensions or Siphoviridae with tails cut via operational mistakes (Fig. 2(A)). Vpp3 and Vpp 4 belonged to Siphoviridae, and were characterized by a flexible tail (Fig. 2(B), (C)). Vpp6 also belonged to Siphoviridae and exhibited an elongated head (Fig. 2(D)). Gram positive

V. parahaemolyticus growth inhibition by the phage
Vibrio parahaemolyticus phage-mediated growth inhibition was examined in culture medium at two different temperatures, 30 and 37°C (at an MOI of 0.1-10). In a preliminary study of the six phages, Vpp2 and Vpp4 had the highest lytic activity and were selected for further analyses. Both Vpp2 and Vpp4 at an MOI of 10 exhibited greater inhibition of V. parahaemolyticus at 37 than 30°C (Fig. 3). When Vpp2 was incubated in the presence of its host, Vp-KF4, significant decreases of 6 log CFU/mL (P \ 0.001) at 37°C and 4.2 log CFU/mL (P \ 0.01) at 30°C were observed after 6 h. When Vpp4 was incubated in the presence of its host, VP-KF3, a significant decrease of 4 log CFU/mL (P \ 0.01) at 37°C was observed after 4 h. There were no dramatic differences in CFU between the Vpp4-treated group and the control group after 6 h. Vpp2 and Vpp4 each have lytic properties against the pathogenic strain V. parahaemolyticus ATCC 17802 based on the host range study (Table 2). Using the same host, Vp-KF4, the two-phage mixture (Vpp1 and Vpp2) was not more effective than each single phage with respect to lytic activity in the culture medium (data not shown). Only Vpp2 was used for the food application study described belows because it showed the highest activity among the phages. When Vpp2 was added to Vp-KF4-spiked clam samples at MOIs of 1, 10, and 100, its growth inhibitory effect was dose-dependent (data not shown). V. parahaemolyticus growth in clam samples treated with Vpp2 at an MOI of 100 revealed a significant reduction of 2.1 log CFU/mL (P \ 0.01) at 25°C until 24 h, but there was no effect of Vpp2 treatment on V. parahaemolyticus growth at 4°C (Fig. 4). The phage had a much lower inhibitory effect on V. parahaemolyticus growth in clam samples than in culture medium. However, the effect of phage application on clam samples persisted for 24 h, but the effect in culture medium reached a maximum at 6 h and then gradually decreases (Fig. 4). The mode of phage treatment on the surface of food, such as pipetting or spraying, could influence the lytic activity of the phage because complex factors in the food matrix affect the relationship between host and phage. This might explain the differences in the effects of phages on bacterial growth inhibition between culture conditions and food matrixes.
The key to successful biocontrol is to isolate promising phages that exhibit high potency and persistent lytic activity. Rong et al. (2014) found that depuration of infected oysters at 16°C with an 0.1 MOI was optimal for reducing V. parahaemolyticus by 2.35-2.76 log CFU/mL. According to Jun et al. (2014), when a phage was surfaceapplied to the flesh of oysters after V. parahaemolyticus inoculation, bacterial growth was inhibited by 6 log CFU/ mL for 12 h. Interestingly, this result implied a lack of group at 37°C (MOI of 10). Data are expressed as mean ± SD in triplicate. *P \ 0.05, **P \ 0.01, ***P \ 0.001: significantly different from the phage-untreated control group Fig. 4 Growth inhibition of phage, Vpp2 against Vibrio parahaemolyticus at different temperature in clam samples. Con A: control group at 4°C; Con B: control group at 15°C; Con C: control group at 25°C. MOI 100A: phage-treated group at 4°C (MOI of 100); MOI 100B: phage-treated group at 15°C (MOI of 100); MOI 100C: phage-treated group at 25°C (MOI of 100). Data are expressed as mean ± SD in triplicate. *P \ 0.05, **P \ 0.01, ***P \ 0.001: significantly different from the phage-untreated control group phage resistance overtime. This was not consistent with our results, presumably owing to the different phages that were used. Additional research on host-phage relationships is necessary to improve our understanding of phage resistance. Further food application studies should be performed to control Vibrio species in seafood using phage mixtures or phage endolysin to enhance the effectiveness of whole single phages.