Protective Effects of Bacteriophages against Aeromonas hydrophila Causing Motile Aeromonas Septicemia (MAS) in Striped Catfish

To determine the effectivity of bacteriophages in controlling the mass mortality of striped catfish (Pangasianodon hypophthalmus) due to infections caused by Aeromonas spp. in Vietnamese fish farms, bacteriophages against pathogenic Aeromonas hydrophila were isolated. A. hydrophila-phage 2 and A. hydrophila-phage 5 were successfully isolated from water samples from the Saigon River of Ho Chi Minh City, Vietnam. These phages, belonging to the Myoviridae family, were found to have broad activity spectra, even against the tested multiple-antibiotic-resistant Aeromonas isolates. The latent periods and burst size of phage 2 were 10 min and 213 PFU per infected host cell, respectively. The bacteriophages proved to be effective in inhibiting the growth of the Aeromonas spp. under laboratory conditions. Phage treatments applied to the pathogenic strains during infestation of catfish resulted in a significant improvement in the survival rates of the tested fishes, with up to 100% survival with MOI 100, compared to 18.3% survival observed in control experiments. These findings illustrate the potential for using phages as an effective bio-treatment method to control Motile Aeromonas Septicemia (MAS) in fish farms. This study provides further evidence towards the use of bacteriophages to effectively control disease in aquaculture operations.


Introduction
Striped catfish (Pangasianodon hypophthalmus) is one of the most important farmed fish species, especially in Vietnam, Thailand, Cambodia, Laos and, more recently, the Philippines and Indonesia [1].Vietnam supplied 90% of catfish production with a value of US$1.1 to 1.7 billion in 2015.Motile Aeromonas Septicemia (MAS), also called haemorrhage disease or red spot disease, causes great losses for farmers (up to 80% mortality) and presents in fish with clinical signs of haemorrhages on the head, mouth, and at the base of fins, a red, swollen vent, and the presence of pink to yellow ascitic fluid [2].Aeromonas hydrophila, Aeromonas caviae, and Aeromonas sobria species were often isolated from diseased catfish, and new species such as Aeromonas dhakensis and Aeromonas veronii were also reported by using molecular methods based on the sequencing of the rpoD gene [3].

Isolation and Characterization of Bacteriophages
The A. hydrophila-phage 2 (or Φ2) and A. hydrophila-phage 5 (or Φ5) were successfully isolated against the propagation hosts used (Figure 1 and Table 2).
Φ2 had an isometric head of 129 nm in diameter with a tail sheath 173 nm long and 15 nm wide.Φ5 was composed of: (i) an isometric head of 120 nm in diameter, (ii) a tail sheath of 198 nm in length and 15 nm in width.All of the phages had contractile tails (Figure 1 and Table 2).Both phages produced clear plaques with diameters of 0.1 mm (Figure 1).

Isolation and Characterization of Bacteriophages
The A. hydrophila-phage 2 (or Φ2) and A. hydrophila-phage 5 (or Φ5) were successfully isolated against the propagation hosts used (Figure 1 and Table 2).
Φ2 had an isometric head of 129 nm in diameter with a tail sheath 173 nm long and 15 nm wide.Φ5 was composed of: (i) an isometric head of 120 nm in diameter, (ii) a tail sheath of 198 nm in length and 15 nm in width.All of the phages had contractile tails (Figure 1 and Table 2).Both phages produced clear plaques with diameters of 0.1 mm (Figure 1).The genome size of the phage isolates was above 20 kb.The genomic material of the isolated phages was not digested by Mung bean nuclease and RNase A. Since Mung bean nuclease specifically cuts single-stranded nucleic acids of both DNA and RNA, it was concluded that the genomic DNA of both phages was double-stranded.RNA nucleic acids are degraded by RNase A, therefore, the nucleic acids of Φ2 and Φ5 were determined as double-stranded DNA (dsDNA) (Figure 2).The phages Φ2 and Φ5 belong to the Myoviridae family.

Host Range
Phage 2 and phage 5 were found to inhibit the growth of all A. hydrophila strains tested.None of the other 27 species was found to be susceptible to these phages (Table S2, Supplementary Materials).

Adsorption Rate of Phages and One-Step Growth Curve
The number of free phages in suspension decreased over time, as illustrated in the adsorption curve (Figure 3a).At 40 min, the percentage of Φ2-infected bacteria was over 90%.
The one-step growth experiment (Figure 3b) results revealed that the latent period and burst size of Φ2 were 10 and 213 PFU per infected host cell, respectively.

Host Range
Phage 2 and phage 5 were found to inhibit the growth of all A. hydrophila strains tested.None of the other 27 species was found to be susceptible to these phages (Table S2, Supplementary Materials).

Adsorption Rate of Phages and One-Step Growth Curve
The number of free phages in suspension decreased over time, as illustrated in the adsorption curve (Figure 3a).At 40 min, the percentage of Φ2-infected bacteria was over 90%.
The one-step growth experiment (Figure 3b) results revealed that the latent period and burst size of Φ2 were 10 and 213 PFU per infected host cell, respectively.

Host Range
Phage 2 and phage 5 were found to inhibit the growth of all A. hydrophila strains tested.None of the other 27 species was found to be susceptible to these phages (Table S2, Supplementary Materials).

Adsorption Rate of Phages and One-Step Growth Curve
The number of free phages in suspension decreased over time, as illustrated in the adsorption curve (Figure 3a).At 40 min, the percentage of Φ2-infected bacteria was over 90%.
The one-step growth experiment (Figure 3b) results revealed that the latent period and burst size of Φ2 were 10 and 213 PFU per infected host cell, respectively.

Inactivation of Aeromonas Species in Vitro
The bacterial concentration (OD550nm values) of the uninfected control (only A. hydrophila N17) increased continuously during 18 h of incubation.In contrast, during the infection with 2 at MOI 1, MOI 0.1, and MOI 0.01 bacterial growth began to be inhibited at 1, 2, and 2.5 h, respectively, and the inhibition was maintained up to 8 h (Figure 4a).Then, the bacterial concentration increased as a consequence of the development of phage-resistant A. hydrophila cells.
The lowest OD550nm value was 0.177 ± 0.023 after 4 h of incubation of Φ5 at MOI 0.1.There was a significant decline in the bacterial concentration (MOI 0.01, 0.1, and 1) in the first 3 h, followed by low level stabilization in the next 1, 2, and 4 h for MOI 1, 0.1, and 0.01, respectively (Figure 4b).Then, the bacterial concentration underwent a turnaround because of the development of phage-resistant A. hydrophila cells.

Inactivation of Aeromonas Species in Vitro
The bacterial concentration (OD 550nm values) of the uninfected control (only A. hydrophila N17) increased continuously during 18 h of incubation.In contrast, during the infection with Φ2 at MOI 1, MOI 0.1, and MOI 0.01 bacterial growth began to be inhibited at 1, 2, and 2.5 h, respectively, and the inhibition was maintained up to 8 h (Figure 4a).Then, the bacterial concentration increased as a consequence of the development of phage-resistant A. hydrophila cells.
The lowest OD 550nm value was 0.177 ± 0.023 after 4 h of incubation of Φ5 at MOI 0.1.There was a significant decline in the bacterial concentration (MOI 0.01, 0.1, and 1) in the first 3 h, followed by low level stabilization in the next 1, 2, and 4 h for MOI 1, 0.1, and 0.01, respectively (Figure 4b).Then, the bacterial concentration underwent a turnaround because of the development of phage-resistant A. hydrophila cells.

Inactivation of Aeromonas Species in Vitro
The bacterial concentration (OD550nm values) of the uninfected control (only A. hydrophila N17) increased continuously during 18 h of incubation.In contrast, during the infection with 2 at MOI 1, MOI 0.1, and MOI 0.01 bacterial growth began to be inhibited at 1, 2, and 2.5 h, respectively, and the inhibition was maintained up to 8 h (Figure 4a).Then, the bacterial concentration increased as a consequence of the development of phage-resistant A. hydrophila cells.
The lowest OD550nm value was 0.177 ± 0.023 after 4 h of incubation of Φ5 at MOI 0.1.There was a significant decline in the bacterial concentration (MOI 0.01, 0.1, and 1) in the first 3 h, followed by low level stabilization in the next 1, 2, and 4 h for MOI 1, 0.1, and 0.01, respectively (Figure 4b).Then, the bacterial concentration underwent a turnaround because of the development of phage-resistant A. hydrophila cells.

Phage Treatment of Infected Fish
The negative control 1 (fishes with no injection) and negative control 2 (fishes injected with the growth medium filtered to remove bacterial cells) showed no mortality of catfish (Figure 5), indicating that the uninfected, control medium did not have any detrimental effect on fish health.
Catfish in the positive control groups (infected with A. hydrophila N17) that were not treated with bacteriophages started to die at a constant rate starting from post-infection day two, with a cumulative mortality rate of 81.67 ± 2.36% (Figure 5).
In contrast, the fish treated with the phages showed lower mortality rates at each different MOI (p < 0.01).While no mortality was observed in the groups treated with MOI 100, the cumulative mortalities in the other groups were 45% (MOI 1) and 68.33 ± 2.36% (MOI 0.01) at the end of the eight-day experiment (Figure 5).

Phage Treatment of Infected Fish
The negative control 1 (fishes with no injection) and negative control 2 (fishes injected with the growth medium filtered to remove bacterial cells) showed no mortality of catfish (Figure 5), indicating that the uninfected, control medium did not have any detrimental effect on fish health.
Catfish in the positive control groups (infected with A. hydrophila N17) that were not treated with bacteriophages started to die at a constant rate starting from post-infection day two, with a cumulative mortality rate of 81.67 ± 2.36% (Figure 5).
In contrast, the fish treated with the phages showed lower mortality rates at each different MOI (p < 0.01).While no mortality was observed in the groups treated with MOI 100, the cumulative mortalities in the other groups were 45% (MOI 1) and 68.33 ± 2.36% (MOI 0.01) at the end of the eightday experiment (Figure 5).

Discussion
The findings of this study demonstrate that the examined Aeromonas spp.were resistant to multiple antibiotics and were thus able to cause high mortality rates in catfish in Vietnam, in spite of the use of various antibiotic treatments.In the bacteriophage treatments, however, 2 and 5 were able to lyse all tested A. hydrophila strains, displaying strong inhibition also of the virulent A. hydrophila strains carrying many virulence genes.Therefore, Φ2 and 5 are promising candidates for the application of a phage therapy to control Aeromonas infection in catfish.
Phage 2 and 5 were found to belong to the Myoviridae family, and our findings are in line with those of Ackermann [10] who indicated that 33 of a total of 43 Aeromonas phages he investigated were tailed and belonged to the Myoviridae family.Recently, other Aeromonas phage studies against different Aeromonas species by Haq et al. [11], Jun et al. [12], and Kim et al. [13] also reported that all phages they identified belonged to the Myoviridae family.Therefore, Myoviridae family members are most likely to be abundant in natural environments.
There was a correlation between the diameter of the plaques observed and the latent period and burst size for the A. hydrophila phage.The 2 had a short latent period (10 min), and these findings are in line with another study conducted by Anand et al. [14] who found that Aeromonas phage BPA 6 had a latent period of 10 min and a burst size of 244 PFU/cell.
The different MOI of Φ2 and 5 caused different bacterial growth patterns.The higher the MOI value, the sooner phage-resistant bacterial cells appeared.A similar result was noted by Kim et al. [13] for the phage PAS 1 against an Aeromonas salmonicida strain, indicating that bacterial resistance appeared after 3, 6, and 24 h at MOIs 10, 1, and 0.1, respectively.
Several Aeromonas phages, such as Aeh1, Aeh2, AH1 have also been reported [12,15,16].However, there have been few reports demonstrating the successful use of phages for the treatment of Aeromonas infections in catfish.The treatment of catfish by an intraperitoneal (IP) injection

Discussion
The findings of this study demonstrate that the examined Aeromonas spp.were resistant to multiple antibiotics and were thus able to cause high mortality rates in catfish in Vietnam, in spite of the use of various antibiotic treatments.In the bacteriophage treatments, however, Φ2 and Φ5 were able to lyse all tested A. hydrophila strains, displaying strong inhibition also of the virulent A. hydrophila strains carrying many virulence genes.Therefore, Φ2 and Φ5 are promising candidates for the application of a phage therapy to control Aeromonas infection in catfish.
Phage Φ2 and Φ5 were found to belong to the Myoviridae family, and our findings are in line with those of Ackermann [10] who indicated that 33 of a total of 43 Aeromonas phages he investigated were tailed and belonged to the Myoviridae family.Recently, other Aeromonas phage studies against different Aeromonas species by Haq et al. [11], Jun et al. [12], and Kim et al. [13] also reported that all phages they identified belonged to the Myoviridae family.Therefore, Myoviridae family members are most likely to be abundant in natural environments.
There was a correlation between the diameter of the plaques observed and the latent period and burst size for the A. hydrophila phage.The Φ2 had a short latent period (10 min), and these findings are in line with another study conducted by Anand et al. [14] who found that Aeromonas phage BPA 6 had a latent period of 10 min and a burst size of 244 PFU/cell.
The different MOI of Φ2 and Φ5 caused different bacterial growth patterns.The higher the MOI value, the sooner phage-resistant bacterial cells appeared.A similar result was noted by Kim et al. [13] for the phage PAS 1 against an Aeromonas salmonicida strain, indicating that bacterial resistance appeared after 3, 6, and 24 h at MOIs 10, 1, and 0.1, respectively.
Several Aeromonas phages, such as Aeh1, Aeh2, AH1 have also been reported [12,15,16].However, there have been few reports demonstrating the successful use of phages for the treatment of Aeromonas infections in catfish.The treatment of catfish by an intraperitoneal (IP) injection illustrated significant protective effects, which increased the relative percentages of the survival rates observed for fish compared to the controls when the MOI increased.Our study revealed that in the MOI-100 experiment the relative percentage survival was 100%.The study of Jun et al. [12] showed that the relative percentage survival of fish treated with A. hydrophila phages pAh6-C and pAh1-C was 16.67 ± 3.82% and 43.33 ± 2.89%, respectively, when the fish were injected with the bacterium (2.6 × 10 7 CFU/fish).However, the labour-intensive and time-consuming mode of delivery of bacteriophages can constitute a disadvantage for the treatment of fish by IP injection in catfish farms.Therefore, further studies should be conducted into whether phage treatments are effective when an on-farm oral method of administration is evaluated.With the use of bioreactors, large volumes of bacteriophages can be produced for bacteriophage incorporation into fish feed.Moreover, the survival of phages and their persistent survival on or in fish, as well as in phage-coated feed preparations should be studied under different environmental factors (e.g., temperature, salt concentration) to determine whether phages are able to persist and effectively reduce Aeromonas spp.levels in fish farms.In conclusion, this study demonstrates that phage treatment of Aeromonas spp.might be an effective tool to improve the survival of farmed catfish affected by MAS.

Aeromonas Species
Bacterial isolates stored at the Research Institute for Aquaculture No. 2 (Ho Chi Minh City, Vietnam) and the ATCC type strains of the pathogens are listed in Table S2.Isolates were previously obtained from diseased catfish in farms in the south of Vietnam (Table S2).

Prophage Induction
In order to choose an Aeromonas species as a propagation host for phage isolation, A. hydrophila N17 was subjected to a prophage induction test.The Aeromonas species was cultured in 10 mL fresh Luria-Bertani (or LB) broth (Sigma-Aldrich, St. Louis, MO, USA) and incubated at 30 • C on an orbital shaker operating at 150 rpm until reaching an OD 550nm of 0.2.Mitomycin C (Sigma-Aldrich) was added to a final concentration of 1 µg/mL and 5 µg/mL, and again the bacterial suspension was incubated at 30 • C on an orbital shaker operating at 150 rpm.The cell density of the bacteria (OD 550nm ) was monitored every 1 h for a 6 h period.At the end of the incubation, the suspension was centrifuged at 10,000 g for 15 min and filtered through a nitrocellulose filter (0.45 µm, Merck Millipore, Burlington, MA, USA) before spotting the filtrate onto an agar plate seeded with the host bacterium to confirm the presence of viable phage particles.A significant decrease in the cell density (OD 550nm ) suggested that prophages were released [17,18].
The antimicrobial susceptibility of Aeromonas species is usually recorded using Enterobacteriaceae breakpoints [20].Susceptible (S), intermediate resistance (I), and resistant (R) were evaluated according the criteria given in the Performance Standards for Antimicrobial Susceptibility Testing M100-S21 (2017, Table 2A-1, pages 33-39) [19].Multi-antibiotic resistance (MAR) was recorded when the bacteria resisted to three or more antibiotics [21].

Isolation and Characterization of Bacteriophages
Phages were isolated from water samples from the Saigon River in the south of Vietnam against A. hydrophila N17 and they were purified following the methods described by Jun et al. [12].
For transmission electron microscopy (TEM): A 200 mesh copper grid was immersed in 40 µL of phage solution for five min before fixing the phage with glutaraldehyde solution (1%) for five min.Then, the phage samples were negatively stained with 5% (w/v) uranyl acetate and observed by TEM (JEOL JEM-1010) operating at a voltage of 80 kV at the Vietnam National Institute of Hygiene and Epidemiology.The phage morphology was determined using the criteria of the International Committee on Taxonomy of Viruses (ICTV) (http://www.ictvonline.org/)and Ackermann et al. [25].
Phage genomic DNA extraction and restriction analyses: Phage genomic DNA was extracted using the Phage DNA Isolation Kit (Norgen Biotek Corp, Thorold, Canada).The nature of the nucleic acids was determined by digestion with Mung bean nuclease and RNase A (ThermoFisher Scientific, Waltham, MA, USA) as per the manufacturer's protocols.The genomic DNA phages were digested using the restriction enzymes: EcoRV, EcoRI, Ncol, SalI, MspI, XmnI, and KpnI, as per the manufacturer's instruction (ThermoFisher Scientific).The DNA fragments were then electrophoresed at 120 V for 40 min.

Host Range
The method was adapted from Le et al. [23] and Goodridge et al. [26] with some modifications described below.The Aeromonas spp.(Table S2) were incubated overnight.Then, a 100 µL aliquot of each Aeromonas spp.culture (optical density of 0.5 at 550 nm) was spread on brain heart infusion agar (BHIA) (OXOID, UK) and dried for 20 min in a biological safety cabinet Class II.The host range of the phage was determined by pipetting 10 µL of phage preparation (~10 8 PFU/mL) on lawn cultures of the strains.The plates were observed for the appearance of clear zones after incubation at 30 • C after 18 h.

Adsorption Rate of Phages
Phage adsorption was studied using the method described previously [27].A phage solution was added to 100 mL of log-phase growing Aeromonas hydrophila N17 culture (×10 7 CFU/mL) in LB broth to get a final MOI of 0.1.The mixture was incubated at 30 • C.An aliquot of 1 mL was collected from the sample every two min over a period of 60 min.The sample was then centrifuged at 4000 g for 15 min, and then the supernatant was diluted with SM buffer + 1% chloroform (http://cshprotocols.cshlp.org/content/2006/1/pdb.rec8111.full?text_only=true).Then, the titers of unabsorbed free phages in the supernatant were determined by the double-layer agar technique, and the results were recorded as percentages of the initial phage counts.The percentages of free phages and the adsorption rates were calculated following the formula of Haq et al. [11].

One-Step Growth Curve
The phage and bacteria were prepared in the same way as in the adsorption method described above.At 40 min, when the adsorption rate was maximal, the mixture was further incubated at 30 • C with 150 rpm.Samples were collected every 5 min for 120 min and phage titers were determined by the double-layer agar technique.Then, the latent period and burst size were calculated [28].
rate of A. hydrophila-phage 2

Figure 5 .
Figure 5. Cumulative mortality rates (%) of striped catfishes obtained in challenging experiments using A. hydrophia N17 and the phage cocktail at the different MOIs (0.01, 0.1, and 1).The ratio of Φ2 to Φ5 in a phage cocktail was 1:1.

Figure 5 .
Figure 5. Cumulative mortality rates (%) of striped catfishes obtained in challenging experiments using A. hydrophia N17 and the phage cocktail at the different MOIs (0.01, 0.1, and 1).The ratio of Φ2 to Φ5 in a phage cocktail was 1:1.

Table 1 .
Antibiogram profile of the Aeromonas hydrophila strains tested.

Table 2 .
Characteristics of bacteriophages against A. hydrophila strains.

Table 2 .
Characteristics of bacteriophages against A. hydrophila strains.