Lytic phages display protective effects against soft rot-causing Pectobacterium sp.

ABSTRACT Soft rot caused by Pectobacterium sp. is responsible for significant losses in vegetable production worldwide. Methods for the effective control of soft rot are limited and primarily based on good agricultural practices. The use of bacteriophages as biocontrol agents appears to be a promising alternative to combat phytopathogens. In this study, we investigated the efficacy of lytic phages recovered from symptomatic cabbage tissues and environmental samples against soft rot caused by Pectobacterium carotovorum subsp. carotovorum. Three bacteriophage isolates, designated as vB_PcaP-A3, vB_PcaM-D1, and vB_PcaM-J3, were observed to effectively lyse P. carotovorum subsp. carotovorum. Phage vB_PcaP-A3 exhibited virion morphology similar to the members of the podovirus group, while phages vB_PcaM-D1 and vB_PcaM-J3 showed myovirus morphology based on transmission electron microscopy. The optimal multiplicity of infection (MOI) differed greatly among the three phages. All three phages survived incubations at 30, 40 and 50 °C and pH conditions ranging from 3.0 to 9.0, but were all inactivated at 60°C and at pH 12. Both monophage and cocktail preparations were effective in inhibiting the growth of P. carotovorum subsp. carotovorum in the in vitro challenge tests. In the potato slice assays, while treatment with cocktail preparations completely inhibited the development of soft rot in tissue slices, monophage treatments not only resulted in a significant reduction of tissue maceration in slices, but also showed protective effect against soft rot in tubers. Overall, these results demonstrate the efficacy of phages vB_PcaP-A3, vB_PcaM-D1, and vB_PcaM-J3 for the biocontrol of soft rot caused by P. carotovorum subsp. carotovorum.

Soft rot caused by Pectobacterium sp. is responsible for significant losses in vegetable production worldwide. Methods for the effective control of soft rot are limited and primarily based on good agricultural practices. The use of bacteriophages as biocontrol agents appears to be a promising alternative to combat phytopathogens. In this study, we investigated the efficacy of lytic phages recovered from symptomatic cabbage tissues and environmental samples against soft rot caused by Pectobacterium carotovorum subsp. carotovorum. Three bacteriophage isolates, designated as vB_PcaP-A3, vB_PcaM-D1, and vB_PcaM-J3, were observed to effectively lyse P. carotovorum subsp. carotovorum. Phage vB_PcaP-A3 exhibited virion morphology similar to the members of the podovirus group, while phages vB_PcaM-D1 and vB_PcaM-J3 showed myovirus morphology based on transmission electron microscopy. The optimal multiplicity of infection (MOI) differed greatly among the three phages. All three phages survived incubations at 30, 40 and 50°C and pH conditions ranging from 3.0 to 9.0, but were all inactivated at 60°C and at pH 12. Both monophage and cocktail preparations were effective in inhibiting the growth of P. carotovorum subsp. carotovorum in the in vitro challenge tests. In the potato slice assays, while treatment with cocktail preparations completely inhibited the development of soft rot in tissue slices, monophage treatments not only resulted in a significant reduction of tissue maceration in slices, but also

Introduction
Bacterial soft rot disease causes serious damage especially among vegetable and fruit crops worldwide (Bhat et al., 2010). Annual post-harvest losses due to bacterial soft rot vary from 15-30% of the harvested crops and at least 30-50% of the 65 million tons of vegetables in public markets (Bhat et al., 2010;Monjil et al., 2021;Rahman et al., 2012). Soft rot disease results from infections by several types of bacteria, but most commonly by members of the soft rot Pectobacteriaceae (SRP) including Pectobacterium carotovorum subsp. carotovorum, P. aroidearium, P. atrosepticum, P. brasiliense, P. betavasculorum and P. wasabiae, which were identified to cause soft rot both in storage and in transit (Adeolu et al., 2016;Agyemang et al., 2020;Charkowski, 2018;Czajkowski, 2015;Gardan et al., 2003;Nabhan et al., 2013;Pérombelon, 2002;Portier et al., 2019;Wallace, 1948). Pectobacterium species have the broadest host range among soft rot bacteria, and are considered as the leading cause of soft rot disease in various vegetable crops (Davidsson et al., 2013;Lim et al., 2017). During infection, SRPs produce enzymes like pectate lyase, cellulase and polygalacturonase, which are capable of breaking down the plant cell wall leading to plant tissue maceration and decay (Charkowski et al., 2012;Li et al., 2019;Mattinen et al., 2007;Muturi et al., 2019). There is still limited information on the predominant SRPs affecting vegetable crops in the Philippines. A previous report by Daengsubha and Quimio (1980) identified P. carotovorum subsp. carotovorum as the major cause of bacterial soft rot affecting crucifers, carrot, and tomato.
Methods used to control soft rot disease include cutting-off and disposal of infected tissues during or after harvest, wrapping with paper, treatment with hot water, and the use of bactericides, such as alum powder (aluminum potassium sulfate), formaldehyde, copper sulfate, and sodium or calcium hypochlorite (Bayogan et al., 2011;Bhat et al., 2010;Czajkowski et al., 2011;Liao et al., 2003). However, these methods only slow down the development of soft rot, and the use of chemical bactericides were reported to be associated with the detection of copper-and antibiotic-resistant phytopathogenic bacterial strains (Behlau et al., 2011;Cooksey, 1990;McManus et al., 2002;Petriccione et al., 2017;Roach et al., 2020;Sundin & Wang, 2018). Due to the limited success of the abovementioned control methods, and the currently growing global threat of antimicrobial resistance, there has been an increased interest in the use of bacteriophages or phages (viruses of bacteria) as biocontrol agents against pathogenic bacteria (Buttimer et al., 2017;Frampton et al., 2012;Jones et al., 2007;Jones et al., 2012).
Bacteriophages (phages) are viruses that infect and lyse bacterial cells. When environmental conditions are favourable for infection, phages can effectively reduce bacterial populations, as in the case of lytic phages that undergo rapid replication (Hadas et al., 1997;Sulakvelidze et al., 2001). Since phages have high host-specificity and have no direct negative effects on non-host cells, these allow the use of phage therapy to be a promising biocontrol method against soft rot caused by Pectobacterium sp. (Buttimer et al., 2017;Lim et al., 2017). The first Pectobacterium phage that was sequenced, PP1, was reported in 2012 and several additional novel phage isolates were subsequently reported (Buttimer et al., 2020;Kalischuk et al., 2015;D. H. Lee et al., 2012;Lim et al., 2013;Lim et al., 2014;Lim et al., 2015;Lim et al., 2017;Pedersen et al., 2020;Voronina et al., 2019). These bacteriophages vary especially in terms of morphology, lytic activity and host range but they all have the potential to control soft rot caused by Pectobacterium sp. as evidenced by the semi-in planta experiments performed in these reports.
In this study, three lytic phages isolated from symptomatic cabbage tissues and environmental samples, were characterised based on virion morphology, growth characteristics, and physico-chemical properties. The lytic activity and efficacy of the phages to prevent tissue maceration caused by P. carotovorum subsp. carotovorum in potato slices using monophage and cocktail suspensions were also determined. The effect of the timing of phage application on the efficacy of monophage treatments on whole potato tubers was also qualitatively assessed. Based on the in vitro and potato slice assays, we report the biocontrol potential of these lytic phages against bacterial soft rot caused by P. carotovorum subsp. carotovorum.

Bacterial strains and media
Pectobacterium carotovorum subsp. carotovorum BIOTECH 1752, obtained from the Philippine National Collection of Microorganisms (National Institute of Molecular Biology and Biotechnology, UPLB) was used as the primary host for phage isolation. A collection of 48 isolates of Pectobacterium spp. obtained from vegetable growing areas and market centers in the Philippines (Host-Pathogen Interaction Laboratory, IWEP, UPLB), and eight other representative strains from other genera (Microbiology Division, IBS, UPLB) were used for the host range assay (Supplementary Table 1). For routine tests and maintenance, bacteria were grown in Tryptone Soya Agar (TSA, HiMedia®, Mumbai, India) for 16-18 h at 30°C. For liquid preparations, bacterial cultures were grown in Tryptone Soya Broth (TSB, HiMedia®, Mumbai, India) for 12-18 h at 30°C with 125 rpm agitation.

Isolation and purification of bacteriophages
Bacteriophages were isolated from rhizosphere soil, symptomatic cabbage tissues, and sewage water in Los Baños, Laguna, Philippines (Czajkowski et al., 2014). Cabbage samples showing symptoms of soft-rot disease were sourced from a local market and the soil and sewage water samples were provided by Envirokonsult Equipment and Services, Inc. (Bay, Laguna, Philippines). Twenty grams (or mL) of each sample was homogenised in 20 mL saline-magnesium (SM) buffer (MgSO 4 , NaCl, CaCl 2 , and Tris-HCl, pH 7.5) and incubated at 30°C for 12-16 h. Five hundred μL of log phase (5-6-h subculture prepared from an overnight culture) P. carotovorum subsp. carotovorum culture containing approximately 10 8 cells mL −1 (0.5 McFarland Standard) was added to each enrichment set-up and was incubated as above. Enrichment cultures were centrifuged at 3,000 rpm for 15 min and the supernatants were chloroform-treated in 1:20 chloroform/supernatant ratio for at least 10 min. The resulting lysates were screened for the presence of bacteriophages by spot test assay . Overlay mixtures, containing 4.5 mL soft agar (TSA), 0.5 mL 20 mM CaCl 2 , and 300 µL of log phase P. carotovorum subsp. carotovorum culture (0.5 MFS) were poured on Petri plates containing pre-solidified TSA. Five μL of the phage lysates were spotted on the overlay and the spots were absorbed into the agar for 10-15 min prior to incubation of plates at 30°C for 12-16 h. Zones of lysis on the bacterial lawn indicated the presence of phages.
Phage purification was done using the double layer plaque assay (Kropinski et al., 2009). Ten-fold dilutions were prepared from each positive lysate. In separate sterile 1.5-mL microcentrifuge tubes, 300 μL of diluted lysate suspensions were mixed with 300 μL log phase culture of P. carotovorum subsp. carotovorum (0.5 MFS) in TSB. These were pre-incubated for 10 min at 4°C to allow adsorption of the phages. Molten overlay (4.5 mL) mixed with 0.5 mL 20 mM CaCl 2 was added to the suspensions and the resulting mixtures were poured onto Petri plates containing pre-solidified TSA. The overlays were allowed to solidify for 15-20 min and the plates were incubated at 30°C for 12-16 h. Well-isolated plaques were picked using sterile pipette tips and the phages were suspended in 500 µL SM buffer. Bacteriophage isolates were purified by successive single plaque isolation for at least five times or until uniform plaque morphology was observed.

Plaque formation using double layer plaque assay
Plaque morphologies were characterised after a series of double layer plaque assays as described above (Jurczak-Kurek et al., 2016). The transparency and morphology of the plaques formed by each bacteriophage isolate was described and plaque diameter was measured using a digital caliper.

Bacteriophage lysate preparation and titration
Double layer plaque assay plates with uniform plaques were flooded with 3-5 mL of SM buffer to obtain the bacteriophage lysate. Plates were gently shaken for 15-20 min and the overlay-SM buffer mixtures were collected and centrifuged at 8,000 rpm for 10 min. The supernatants were chloroform treated in 1:20 chloroform/lysate ratio and were maintained at 4°C until titration. Lysates were subjected to double layer plaque assay (or spot test) for titration as described above. Phage titer was determined and expressed as plaque forming units (PFU mL −1 ).

Determination of virion morphology using transmission electron microscopy (TEM)
High titer lysates (at least 10 10 PFU mL −1 ) of the three phage isolates were submitted to the Electron Microscopy Laboratory of the Research Institute for Tropical Medicine (Muntinlupa City, Philippines). The phage suspension was placed on carbon-coated copper grids, and 2% aqueous uranyl acetate was added for three min to negatively stain the virions. Phages were visualised by transmission electron microscopy (TEM; JEOLJEM-1220 Electron Microscope), and the images were scanned with 800 ms exposure, 1.4 gain, and 1.0 bin with no sharpening, normal contrast, and 1.00 gamma.

Optimal multiplicity of infection (MOI)
The optimal ratio of phage particles to the number of bacterial cells, also called the optimal multiplicity of infection (MOI) was determined for the three phage isolates. P. carotovorum subsp. carotovorum was infected with each phage isolate at four different PFU mL −1 /cell mL −1 ratios (MOI): 0.01, 0.1, 1.0, and 10. Following overnight incubation at 30°C, the suspensions were chloroform-treated (1:20) for 10 min and centrifuged at 8,000 rpm for 10 min. The supernatants were subjected to double layer plaque assay to determine phage titer. The MOI demonstrating the highest phage titer or PFU mL −1 was considered optimal.
2.2.6. Latent period and phage burst size One-step growth experiments were conducted to determine phage latent period and burst size following previously described protocol (Czajkowski et al., 2014). Cells of P. carotovorum subsp. carotovorum at log phase were harvested by centrifugation at 12,000 rpm for 15 min at 10°C, and were resuspended in TSB to an OD 600 of 0.5 (c. 5 × 10 8 CFU mL −1 ). Subsequently, 100 μL of 20mM CaCl 2 and 100 μL of the phage isolate were added into the bacterial culture at MOI of 0.1 (10 7 PFU mL −1 /10 8 CFU mL −1 ). The phages were allowed to adsorb by incubation at 4°C for 10 min and the suspensions were further incubated at 30°C for 10 min. The suspensions were centrifuged at 12,000 rpm for 15 min at 10°C to remove unadsorbed phages. The resulting pellet was resuspended in 10 mL fresh TSB. The mixtures were incubated at 30°C for 60 min with agitation at 125 rpm. Aliquots were taken from the mixtures every 10 min and the titer was determined using the Spot Test method. The burst size was calculated using the formula: burst size = PFU final − PFU initial PFU initial and the experiment was repeated twice with the same set-up and the results were averaged.

Adsorption rate
Phage adsorption rate was determined following previously described protocol (Czajkowski et al., 2014). One mL of P. carotovorum subsp. carotovorum (ca. 10 8 CFU mL −1 ) was infected with each phage isolate by adding 100 μL of phage lysate (ca. 10 8 pfu mL −1 ) to reach MOI 0.1 and the resulting suspensions were incubated at 30°C for 20 min. After 0 (control), 1, 2, 4, 10 and 20 min, 50-μL aliquots were taken from each suspension and added to 450 μL TSB. Ten-fold dilutions in SM buffer were prepared and the residual titer of the unadsorbed phages were determined using the spot test method as described above. The experiment was repeated twice with the same set-up and the results were averaged. Phage adsorption was calculated as follows: %adsorption = control titer − residual titer residual titer x100.
2.2.8. Thermal resistance and pH stability Thermal resistance of the three phages was determined at 30, 40, 50 and 60°C. Phages (10 8 PFU mL −1 ) in SM buffer were incubated for 1 h at each temperature condition and the surviving PFU was determined using double layer plaque assay (Czajkowski et al., 2014).
The effect of pH on phage stability was determined using phages (10 8 PFU mL −1 ) in SM buffer within a pH range of 3.0-12.0. Phage lysates (10 8 PFU mL −1 ) were added to SM buffer, adjusted with 2N HCl or 3N NaOH to pH 3, 7, 9, and 12 (Lim et al., 2013). The experiment was carried out by incubating the phage suspensions at 4°C for 16-24 h and the surviving PFU was determined using double layer plaque assay.

Host range determination
The host ranges of the three phage isolates was determined using a collection of 48 Pectobacterium spp. isolates and eight representative strains from different genera (Supplementary Table 1) using the spot test assay. Pectobacterium spp. were isolated in the years 2017-2019 from symptomatic tissues of various vegetable crops.
For the spot test assay, bacterial cultures were grown in TSB at 30°C for 12-16 h with agitation at 125 rpm. From the resulting bacterial cultures, 300 μL (0.5 MFS) of each was inoculated into the overlay agar and was subsequently poured into a prepared TSA plate and allowed to solidify at room temperature for 30 min. Five μL of the phage lysate (10 9 PFU mL −1 ) was spotted onto the top agar layer, and the plates were allowed to dry at room temperature for 10-15 min. Following overnight incubation at 30°C, plates were inspected for lysis zones.

Potato slice assays
2.4.1. Phage biocontrol using monophage and cocktail treatments on tissue slices The efficacies of monophage and cocktail treatments against soft rot caused by P. carotovorum subsp. carotovorum were evaluated using a potato slice assay . Potato tubers (cultivar Igorota) were surface sterilised by soaking in 70% ethanol for 10 min, washed with sterile distilled water, and air dried under the laminar flow hood. The tubers were cut into 7.0 mm-thick transverse disks using sterile knives and three wells (5-mm diameter) were made on each slice using a sterile cork borer. For monophage treatments, log phase culture of P. carotovorum subsp. carotovorum containing approximately 10 8 cells mL −1 (0.5 MFS) was mixed with the phage lysates at optimal MOI. For cocktail preparations, low (0.001) and high (100) MOIs were used for the experiment. Three potato slices were used for each treatment. Wells were filled with 20 µL of the bacteria-phage mixtures, 20 µL of sterile SM buffer as negative control, and 20 µL of P. carotovorum subsp. carotovorum (0.5 MFS) as positive control. The diameter of the rotten tissues around the wells inoculated with bacteria and wells coinoculated with bacteria and phages were measured after incubation at 30°C for 72 h. Treatment means were compared using one-way ANOVA with Tukey's post-test (α = 0.05).
2.4.2. Protective and curative potential of monophage treatments against soft rot A potato tuber assay was conducted to determine the protective and curative effect of monophage treatments at various application time points relative to P. carotovorum subsp. carotovorum inoculation following a previously described protocol by Czajkowski et al. (2014). Potato tubers (cultivar Igorota) were washed in running water to remove soil and other organic particles, surface-sterilised with 70% ethanol for 10 min, rinsed with sterile water for three times, and air dried under laminar flowhood. For each tuber, two wells (8 mm diameter) were made using flame-sterilised cork borer. Two set-ups were prepared for the experiment: (1) advanced phage introduction (API) treatment, and (2) delayed phage introduction (DPI) treatment. For both treatments, log phase culture of P. carotovorum subsp. carotovorum containing approximately 10 7 cells mL −1 was used and the phage lysate was added to the wells at MOI 10. In the API treatment, wells were filled with 25 μL of the phage suspension at 0, 1, 3, 8, 15 and 30 h prior to the inoculation of 25 μL of the bacterial suspension. In the DPI treatment, wells were filled with 25 μL of the phage suspension at 0, 1, 3, 8, 15 and 30 h after bacterial inoculation. Three potato tubers were used per time point for each of the API and DPI treatment. For the control set-ups, 25 μL aliquots of sterile water was used as negative control while 25 μL of P. carotovorum subsp. carotovorum suspension was used as positive control. The tubers were placed in a humid box and the degree of tissue maceration was qualitatively assessed (scored as +/-) after incubation at 30°C for 72 h.

Isolation of lytic bacteriophages
Bacteriophages vB_PcaP-A3, vB_PcaM-D1 and vB_PcaM-J3 were isolated from rhizosphere soil, symptomatic tissues, and sewage water, respectively, using P. carotovorum subsp. carotovorum as enrichment and isolation host. The three phages were named following the unified phage naming system proposed by Kropinski et al. (2009). All three phages formed clear plaques typical for lytic phages, with varying sizes (Table 1).

Characterization of bacteriophages
All three phages exhibited strong adsorption to P. carotovorum subsp. carotovorum, and approximately 90% of the phage particles adsorbed to the bacterial cells within five min ( Figure 2). vB_PcaP-A3 and vB_PcaM-D1 showed similar infection kinetics with a latent period of 10 min and a burst size of 142-151 progeny phages/cell, whereas vB_PcaM-J3 showed a longer latent period of 20 min and a smaller burst size of 42 progeny phages/cell  (Table 1, Figure 2(A,B)). Among the three phages, vB_PcaP-A3 has the lowest optimal MOI (0.01), followed by vB_PcaM-D1 (1.0) and vB_PcaM-J3 (10.0) ( Table 1). All of the three phages retained infectivity after incubation at 30°C, 40°C, and 50°C for 1 h, and in the pH range 3-9 for 16-24 h (Figure 2(C)).

Host range
Spot test assay revealed that the three phages showed antibacterial activity specific for P. carotovorum subsp. carotovorum. None of the three phages were able to lyse the 48 Pectobacterium spp. isolates and the eight representative strains from other genera.

Phage activity in vitro
Killing curves established for vB_PcaP-A3, vB_PcaM-D1 and vB_PcaM-J3 showed that all three phages have strong lytic activity and were highly effective against P. carotovorum subsp. carotovorum at optimal MOI (Figure 3(A)). The bacterial cell density remained less than 0.05 (OD 600 ) throughout the 8-h observation period. The absence of any increase in OD 600 reading also suggests that there is no observed phage resistance development of the host, at least during the observation period (Muturi et al., 2019). These results indicate that monophage treatments of vB_PcaP-A3, vB_PcaM-D1 and vB_PcaM-J3 at optimal MOI were equally effective in inhibiting the growth of P. carotovorum subsp. carotovorum in vitro.
For the cocktail treatments, low MOI (0.001) and high MOI (100) were used in the in vitro challenge tests. Each phage was used at the same concentration and the phage cocktails consisting of two or three phages were combined and evaluated. At MOI of 0.001, the killing curves of the phage cocktails composed of two or three phages maintained its bacterial inactivation with an OD 600 < 0.05 for 8 h (Figure 4(A)). At low MOI, it appears that a phage cocktail containing any of the three phages effectively controlled bacterial growth during the 8-h observation period. At MOI 100, killing curves showed that all cocktail treatments were highly effective against P. carotovorum subsp. carotovorum (Figure 4(A)). The bacterial density remained less than 0.05 (OD 600 ) throughout the duration of the experiment.
3.1.6. Phage-Mediated biocontrol using monophage and cocktail treatments 3.1.6.1. Monophage and cocktail treatments reduced soft rot symptoms in tissue slices. The average diameter of rotting tissue was significantly lower in all monophage treatments than in the set-up inoculated with P. carotovorum subsp. carotovorum only (Figure 3(B-D)). The set-up co-inoculated with P. carotovorum subsp. carotovorum and vB_PcaP-A3 exhibited the lowest diameter of rotting tissue among the three treatments (Figure 3(B-D)). Phage vB_PcaP-A3 was able to reduce potato tuber tissue maceration by at least 60%, while phages vB_PcaM-D1 and vB_PcaM-J3 were able to reduce tissue maceration by at least 50% of that observed for the control potato slices inoculated with P. carotovorum subsp. carotovorum only (Figure 3(B-D)). All of the phage cocktail treatments completely prevented tissue maceration at low and high MOI (Figure 4(B-D)).
3.1.6.2. Advanced phage introduction can protect tubers from soft rot. The potential use of phages as a post-harvest disease control strategy was assessed using potato tubers artificially inoculated with P. carotovorum subsp. carotovorum treated with monophage suspensions applied at different time points. Results of the whole tuber assay showed that in the API treatment, monophage suspensions of the three phages applied on to the wound up to 30 h before bacterial inoculation provided protection against soft rot disease development (Figure 3(E)). In the DPI treatment, a 3-h and 8-h delay in the application of monophage suspensions of vB_PcaM-J3, and vB_PcaP-A3 and vB_PcaM-D1, respectively, following bacterial inoculation resulted in disease development (Figure 3(E)).

Discussion
Phage-mediated biocontrol of soft rot disease is a promising and an up-and-coming technology that has been widely explored worldwide. To our knowledge, this is the first report on the isolation and characterization of bacteriophages infecting P. carotovorum subsp. carotovorum in the Philippines, which explores some aspects that need to be considered for potential formulation of a postharvest biocontrol product against soft rot disease including phage stability, bioactivity in tissues, and application timing.
The latent period and burst size of phages vB_PcaP-A3, vB_PcaM-D1, and vB_PcaM-J3 were determined based on one-step growth curve experiments. The phages were further characterised, and the adsorption rate, optimal MOI, and stability on varying temperature and pH levels were also determined. Of the three phage isolates, vB_PcaP-A3 has the lowest optimal MOI of 0.01, a latent period of 10 min and a calculated burst size value of 142 PFU/cell. Phage vB_PcaM-D1 has an optimal MOI of 0.1, latent period of 10 min and a burst size of 151 PFU/cell. vB_PcaM-J3 has an optimal MOI of 10, longest latent period of 30 min, and the lowest calculated burst size of 42 PFU/cell. Since phages displaying shorter latent periods and higher burst sizes are reported to provide better efficiency than phages with longer latent periods in cultures where the density of the host is high, vB_PcaM-J3 showed less promising growth characteristics compared to vB_PcaP-A3 and vB_PcaM-D1 (Abedon et al., 2003). All three phages, however, showed strong adsorption to their host evidenced by the observed 90% phage adsorption within 5 min after incubation. When their stabilities on varying temperature and pH levels were tested, all three phages survived incubations at 30°C-50°C and at pH 3-9, but were inactivated at 60°C and at pH 12. These are important environmental factors in biological control applications. Since temperatures under field conditions do not exceed 50°C and pH levels usually range from pH 3.3-4.3, results indicate that vB_PcaP-A3, vB_PcaM-D1 and vB_PcaM-J3 can be suitable for possible field applications (Navarrete et al., 2017).
All three phages showed specificity to their isolation host, P. carotovorum subsp. carotovorum. Phages are generally known to have narrow host ranges, typically limited to strains within a certain species of their bacterial host (Ross et al., 2016). This can allow the formulation of phage cocktails targeting bacterial species of interest only, which may be a certain bacterial plant pathogen or a specific bacterium in a microbial community that negatively impacts plant growth or health conditions. Developing a phage-based biocontrol product that can eliminate multiple genera, or even different related species could be challenging, but combining phages specific to target hosts may overcome this problem (Buttimer et al., 2017;Frampton et al., 2012). The specificity of phages also reduces the risk of harming the natural microbiota, especially those nontarget bacteria that are potentially beneficial, and it limits the emergence of phage-resistant bacteria (Loc-Carrillo & Abedon, 2011).
In recent years, there were numerous attempts to systematise the single and cocktail application of phages to tubers and vegetable crops. For example, the first bacteriophage of P. carotovorum subsp. carotovorum that was sequenced, PP1, exhibited effective protection against its host in lettuce (Lim et al., 2013). In another study, two other phages namely ϕPD10.3 and ϕPD23.1 were applied together in both whole potato tubers and slices against a number of Pectobacterium and Dickeya isolates. Tissue maceration was significantly reduced by 80% and 95% in slices and whole tubers, respectively . Two further studies revealed effective inhibition of SRPs in potato tubers using monophage formulations (J. Lee et al., 2017;Smolarska et al., 2018). In this study, the efficacy of the three bacteriophages, vB_PcaP-A3, vB_PcaM-D1, and vB_PcaM-J3, in both monophage and cocktail suspensions (combinations of two and three phages) were assessed through in vitro and semi-in planta bacterial challenge tests, and we found that both monophage and phage cocktails completely inactivated the growth of P. carotovorum subsp. carotovorum. However, phages in cocktail suspensions were more effective over individual application in terms of their protective effect against soft rot by P. carotovorum subsp. carotovorum as indicated by the complete maceration reduction in both MOI (0.001 and 100) tested in the study. Recent reports of phage cocktail formulations against P. carotovorum subsp. carotovorum have also described efficient inhibition of bacterial growth both in vitro and in vivo (Carstens et al., 2019;Muturi et al., 2019;Zaczek-Moczydłowska et al., 2020). The use of higher MOI in phage formulations has also been reported to increase the chance for the bacteriophages to infect all host cells, thus effectively killing all host cells after one round of replication (Alves et al., 2016;Thomas, 2001).
Using different novel phage isolates against a number of SRP model strains, our work and the several studies mentioned above have provided preliminary evidence of the biocontrol potential of bacteriophages in controlling bacterial soft rot. However, a comprehensive understanding of their bioefficacy in postharvest applications is still limited and our work contributes to this pool of information by determining the efficient timing of phage treatment on tissues. Preliminary experiments conducted to assess possible timing strategies in employing phage treatments to control post-harvest soft rot revealed the protective effect of introducing phages prior to bacterial inoculation (Figure 3(E)). In each monophage treatment, advanced phage introduction (API) treatments of up to 30 h were observed to protect the tubers from developing the disease (Figure 3(E)). With API, we hypothesise that, considering that the phage number applied is higher than or equal to the optimal MOI, the phages were able to rapidly kill the inoculated bacterial population, thus suppressing both the bacterial growth and the emergence of visible symptoms of the disease.
In contrast, a shorter window of time wherein the phage treatment seemed to offer effective protection from the disease was observed in the delayed phage introduction (DPI) treatment. This observation, in line with the hypothesised phage-bacterium interaction in the API treatment, seemed to suggest that a critical bacterial population relative to the phage population was required to effect noticeable disease manifestation. It was previously reported that quorum sensing regulates the coordination of the synthesis of virulence factors, including the production of plant cell wall degrading enzymes by Pectobacterium spp, which links the observed longer protective window of API treatment and the reduced protective window observed in DPI treatment (Pollumaa et al., 2012).
Furthermore, taking both into account the differences in the characteristics of the phages and the MOI used in this experiment (MOI 10.0, relative to bacterial population at time 0), it follows that, for the DPI treatment, a shorter protective window shall be observed for vB_PcaM-J3, which has the longest latent period (30 min), lowest burst size (42 CFU/cell) and highest MOI (10.0). That is, with the advanced introduction of the bacterial population of at least 3 h, the bacterial population seemed to have grown enough such that the vB_PcaM-J3 population is not at its optimal MOI anymore. In line with this is the observed longer protective window afforded by vB_PcaP-A3 and vB_PcaM-D1 that have lower optimal MOIs of 0.01 and 1.0 respectively, which are both lower than the MOI used in this experiment, and higher burst sizes of 151 and 142 CFU/cell, respectively. These characteristics may have also provided a buffering effect to account for the bacterial growth in DPI treatment.
Taken together, these results are suggestive of a potential post-harvest soft rot disease control strategy that may maximise the benefit of phage treatment by introducing efficiently lytic phages (ie. short latent period, larger burst size, and low optimal MOI) at a high MOI at the earliest time point after harvest. By doing so, while a small population of the soft rot-causing bacteria may have been inadvertently introduced during the process of harvesting, this population can be effectively controlled thus preventing bacterial growth, disease manifestation, and cross-contamination.
However, we are also aware that the use of higher MOI could also possibly increase the emergence of phage-resistant bacterial variants (Chen et al., 2018). This is due to the rapid evolution of the bacterial population from the dominance of phage-sensitive clones to phage-resistant clones when exposed to the selection pressure of lytic phages (Castillo et al., 2015;Hyman & Abedon, 2010;Middelboe, 2000;Middelboe et al., 2001). This implies that the use of heavy phage loads is an important selection factor that should be taken into account in phage therapy and that lower MOIs should also be used when testing protective effect and bacterial inactivation by phages, particularly in actual plant tissues (Chen et al., 2018). To this end, further experimental studies to optimise the phage treatment parameters and to identify the minimum effective phage number for potential field and post-harvest application are to be conducted. A detailed assessment of the sequence of the whole genome of the characterised phages in this report would also be a necessary step moving forward. Elucidation of the genomic sequences of these phage isolates would allow phylogenetic relationship analysis that would shed light on their taxonomic relationships with other previously reported phages from other countries. More importantly, by analyzing the sequences of these phage isolates, the presence of genes related to lysogeny, antimicrobial resistance, or toxin production might also be detected, which are all important considerations for the development of a potential phage-based biocontrol treatment against soft rot disease caused by Pectobacterium spp.