Isolation and characterization of the lytic bacteriophages and their application in combination with amoxicillin against Aeromonas dhakensis

2.1 Isolation of Aeromonas

To isolate Aeromonas species, 30 samples were collected from different sources, including fishponds, canal water, and rivers in Bangkok, Thailand. The samples were streaked onto an Aeromonas isolation medium (HiMedia, India) supplemented with ampicillin. The plates were incubated for 24 h at 30 °C. The dark green colonies resembling Aeromonas sp. were selected. Gram-negative bacteria capable of degrading nitrates to nitrites, glucose fermenters, oxidase, and catalase-positive isolates resembling the genus Aeromonas were selected for 16S rRNA gene sequencing analysis. Other biochemical tests were used to differentiate between Aeromonas genera. L-arabinose fermentation was also differentiated between A. hydrophila and A. dhakensis. Likewise, salicin fermentation allowed differentiation between A. hydrophila and A. dhakensis from A. hydrophila subsp. ranae (Beaz-Hidalgo et al., 2013).

2.2 Identification of Aeromonas spp

The genomic DNA obtained from the Aeromonas isolate underwent extraction utilizing the AccuPrep® Genomic DNA Extraction Kit (Bioneer, Korea) and was employed as templates for PCR amplification. The 16S rRNA gene was amplified using a pair of universal primers, 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-GGCTACCTTGTTACGACTT-3′) as described by Lane (1991). Subsequently, the sequenced fragments were compared to the GenBank database through the Basic Local Alignment Search Tool (BLAST), and phylogenetic trees were constructed using the neighbor-joining method in the MEGA 5.1 software package.

2.3 Antimicrobial susceptibilities of Aeromonas isolates

Aeromonas isolates were underwent MIC testing using MIC test strips, which included amoxicillin, chloramphenicol, doxycycline, gentamicin, and tetracycline (Liofilchem® MTS™, Italy). The interpretative criteria were in accordance with the Clinical and Laboratory Standards Institute (CLSI) VET04 guidelines (CLSI, 2020).

2.4 Phage isolation and detection

The isolated Aeromonas strains served as hosts for bacteriophage isolation, following the method described by Sunthornthummas et al. (2017). The presence of phages was determined using the double-layer agar plate method on NA medium. Phage plaques were counted following an overnight incubation at 30 °C and expressed as plaque-forming units (PFU/mL).

2.5 Electron microscopy

Phage morphology was visualized by transmission electron microscopy (TEM) using carbon-formvar-coated grids, 1% (w/v) uranyl acetate staining (pH 4.5), and a TECNAI 20 TWIN transmission electron microscope operating at 120 kV with a magnification of 120,000×.

2.6 Host-range determination and determination of optimal multiplicity of infection (MOI)

The host range of the isolated phages was determined using the spot test method. Other reference strains of Aeromonas were tested for susceptibility to phages. Bacterial sensitivity to the phage was indicated by the presence of a plaque at the spot. Additionally, a host strain suspension (10⁸ CFU/mL) in NB was mixed with the phage stock at four different ratios (0.01, 0.1, 1, and 10 PFU/CFU) to determine the optimal MOI. The ratio with the highest phage titer was considered the optimal MOI (Pringsulaka et al., 2011).

2.7 One-step growth curve experiments

A one-step growth curve for each phage isolate was performed as Sunthornthummas et al. (2017). The latent period, rise period, and burst size were calculated using the one-step growth curve (Adams, 1959).

2.8 pH and thermal stability

For the pH stability tests, NB was pre-adjusted to a range of pH values (pH 2.0–11.0). A phage suspension (10¹⁰ PFU/mL) was inoculated and incubated for 90 min at 30 °C. For thermal inactivation experiments, phage lysates (10¹⁰ PFU/mL) were subjected to heat treatment at 4, 30, 37, 45, 63, 72, and 100 °C in NB. The phage titer was determined using the double-layer agar plate method for both the pH stability tests and thermal inactivation experiments.

2.9 Whole genome sequencing and computational analyses

2.10 A. dhakensis growth inhibition by single phage and phage cocktail in vitro

Phage therapy was divided into two treatments: pre- and post-treatment. In the pre-treatment experiment, phages or phage cocktail were added before inoculation with A. dhakensis AM (1 × 10⁸ CFU/mL), resulting in MOIs of 0.1, 1, and 10. In the post-treatment experiment, A. dhakensis AM suspensions (1 × 10⁸ CFU/mL) were inoculated into NB and incubated for 3 h. Equal volumes of phages or phage cocktail were added at MOIs of 0.1, 1, and 10. Both treatments were performed in a 250 mL Erlenmeyer flask containing 50 mL of NB at 200 rpm and incubated at 30 °C for 48 h. For each assay, two control samples were set: the bacterial control and the phage control. The bacterial control was inoculated with A. dhakensis but not phages, and the phage controls were inoculated with phages but not bacteria. The control and test samples were incubated under the same conditions. Aliquots of the test samples and their controls were sampled at 0, 6, 12, and 24 h of incubation. In all assays, phage titer was determined in triplicate using the double-layer agar plate method. The bacterial concentration was determined in triplicate in the NA medium. Three independent experiments were performed for each condition.

2.11 A. dhakensis growth inhibition by phage cocktail and antibiotics combination

The inhibitory effects of the selected two-phage cocktail with effective MOIs in combination with antibiotics at sub-MIC (1/2 MIC) were determined as previously described. In the pre-treatment experiment, a combination of the selected two-phage cocktail with effective MOIs and amoxicillin at sub-MIC was added before inoculation with A. dhakensis AM (1 × 10⁵ CFU/mL). In the post-treatment experiment, A. dhakensis AM suspensions (1 × 10⁵ CFU/mL) were inoculated into NB and incubated for 3 h. Equal volumes of the selected three-phage cocktail with effective MOIs and amoxicillin at sub-MICs were added. Only the phage cocktail and antibiotics at MIC were also administered in both pre-and post-treatment. Phage and bacterial counts were determined in NB in two different volumes: 200 μL in 96-well microtiter plates and 20 mL in 250 mL Erlenmeyer flasks. The latter was incubated on an orbital shaker with a shaking speed of 200 rpm. After incubation at 30 °C, the aliquots of each sample and their controls were collected every 6 h for 48 h and were serially diluted to determine viable bacteria (CFU/mL) in NA plates incubated for 24 h at 30 °C.

2.12 Statistical analysis

Statistically significant differences in all experiments were determined by one-way analysis of variance (ANOVA), and post-hoc Tukey’s test was applied to illustrate significant differences between bacterial concentrations between treatment groups over time. A p-value < 0.05 was considered to indicate statistical significance. SPSS statistical software package (version 13.0) was used for all analyses.

3.1 Aeromonas isolation and identification

Three of the 40 isolates from 30 collection sites were preliminarily identified as Aeromonas by biochemical tests. These Aeromonas strains were then used as hosts for phage isolation. However, only the Aeromonas isolate AM was able to isolate phages using the enrichment technique. The isolated AM was further characterized using 16S rRNA gene sequencing, revealing a 99% identity with A. dhakensis. The neighbor-joining tree indicated that strain AM was most closely related to A. dhakensis (Fig. 1). The biochemical tests of A. dhakensis AM are shown in Table 1. To distinguish A. dhakensis from A. hydrophila subsp. hydrophila and A. hydrophila subsp. ranae, the results confirmed that strain AM was negative for L-arabinose and positive for salicin fermentation, confirming its classification as A. dhakensis. Furthermore, for more comprehensive characterization, a whole genome sequence analysis of strain AM was included in this study.

Close

Fig. 1

Phylogenetic tree of Aeromonas spp. based on the 16S rRNA gene using neighbor-joining method. Bootstrap values (%) of 1000 replicates are represented on the branches.

Export to PPT

3.2 Antimicrobial susceptibility of A. dhakensis AM

The MICs of six antimicrobial agents against A. dhakensis AM were evaluated. Notably, amoxicillin had the highest MIC of 64 µg/mL among the antibiotics, a value significantly higher than the Clinical and Laboratory Standards Institute (CLSI) MIC breakpoints (>8 µg/mL) (data not shown). Considering the limited availability of information regarding the susceptibility profiles of A. dhakensis, our findings provide valuable insights into the antibiotic susceptibility of A. dhakensis AM. Specifically, our results demonstrate that A. dhakensis AM displays susceptibility to chloramphenicol, doxycycline, and gentamicin. Traditionally, Aeromonas have shown susceptibility to a range of antimicrobial agents, including 4th-generation cephalosporins, aminoglycosides, fluoroquinolones, tetracycline, and trimethoprim-sulfamethoxazole (Aravena-Roman et al., 2012). It is important to highlight that only a limited selection of antimicrobial agents, including oxytetracycline, amoxicillin, sulfadimethoxine/ormetoprim, and enrofloxacin, have been approved for use in aquaculture in Thailand (Baoprasertkul et al., 2012). Our results underscore the resistance rate to amoxicillin, in line with the report by Aravena-Roman et al. (2011), which noted that only 1.6% of 193 Aeromonas isolates were susceptible to amoxicillin. Recognizing the unique resistance pattern of amoxicillin against the target bacteria, we chose to incorporate amoxicillin at sub-MIC levels for our evaluation of synergism between the antibiotics and the phage cocktail. This choice was driven by the need to explore alternative treatment strategies given the observed resistance and to assess the potential of phages in complementing amoxicillin's limited efficacy in addressing A. dhakensis AM infections.

3.3 Genomic features of A. dhakensis AM

The in silico genome of A. dhakensis AM comprises one circular chromosome of 4,884,279 bp with a G + C content of 61.9% (Fig. 2). The genome contained 4256 coding DNA sequences (CDSs). We emphasized the antimicrobial resistance genes and virulence factors corresponding to the main bacterial virulence determinants. Antimicrobial resistance genes were identified in the genome of A. dhakensis AM (Table S1). Virulence factor genes were identified in the genome of A. dhakensis AM (Table S2). Several typical toxin-encoding genes have been identified, such as aerolysin, hemolysin, and exotoxin. Five prophages were identified in the genome (Table S3), and no plasmids were found during genome analysis.

Close

Fig. 2

Genome features of A. dhakensis AM. Circular representation of the following characteristics are shown from the outside to the center of the diagram. Circle 1: coding sequence (CDS) on the reverse strand, circle 2: coding sequence (CDS) on the forward strand, circle 3: GC contents, circle 5: GC skew values (GC skew + shown in green, GC skew- shown in pink).

Export to PPT

3.4 Phage isolation, purification and phage morphology

Four phages, designated as vB_AdhS_TS3, vB_AdhM_TS9, vB_AdhM_DL, and vB_AdhS_M4 were isolated using A. dhakensis AM as the host. These phages exhibited clear plaques with diameters ranging from 1.7 to 2.0 mm (Fig. 3). The electron micrographs revealed distinct morphologies for the isolated phages. Phage vB_AdhS_TS3 exhibited an icosahedral head of approximately 75.2 nm and a contractile tail with a length of 225.3 nm, while phage vB_AdhS_M4 had an icosahedral head of approximately 64.8 nm and a tail length of 185.4 nm. In contrast, phages vB_AdhM_DL and vB_AdhM_TS9 displayed different morphologies, with phage vB_AdhM_DL possessing an icosahedral head with dimensions of 50.4 nm and a tail length of 210.4 nm, and phage vB_AdhM_TS9 featuring an icosahedral head of approximately 85.1 nm and a shorter tail measuring 101.4 nm (Fig. 3). In a study by Bai et al. (2019), it was reported that among 51 complete genome sequences of Aeromonas phages in GenBank, the majority of Aeromonas phages were classified into different families. However, it's important to note that the ICTV's updated classification emphasizes that these morphological categories do not hold formal taxonomic significance in the classification of phages.

Close

Fig. 3

Plaques and TEM images of phage vB_AdhS_TS3 (A), vB_AdhS_M4 (B), vB_AdhM_DL (C), and vB_AdhM_TS9 (D). Scale bar = 100 nm.

Export to PPT

3.5 Host range determination

All phages were infected only with A. dhakensis and did not infect other Aeromonas spp., such as A. hydrophila, A. caviae, A. sobria, A. trota, or A. veronii (data not shown). Bacteriophage vB_AdhM_TS3 and vB_AdhM_TS9 are the broadest host range phage, able to infect A. dhakensis in five out of the six strains tested.

3.6 Optimal multiplicity of infection determination (MOI) and one-step growth curve

Phage vB_AdhS_TS3, vB_AdhM_DL, vB_AdhS_M4 and vB_AdhM_TS9 generated a maximum titre of 9.68 ± 0.05, 9.94 ± 0.05, 10.41 ± 0.06 and 8.85 ± 0.25 PFU/mL when infected at an optimal MOI of 10 (data not shown). The one-step growth curve of the phages revealed latent periods of approximately 40, 30, 50, and 30 min for vB_AdhS_TS3, vB_AdhM_DL, vB_AdhS_M4, and vB_AdhM_TS9, respectively. The burst sizes for these phages were estimated as 1380, 1280, 253, and 630 PFUs/infected cells, respectively (data not shown). Among the four phages, phage vB_AdhS_M4 had the longest latent period, smallest burst size, and narrowest host range. Therefore, we selected the other three phages, vB_AdhS_TS3, vB_AdhM_DL, and vB_AdhM_TS9, for further studies.

3.7 pH and thermal stability

All phages were resistant to a wide range of pH values after 2 h of incubation, and the optimum range was pH–6–8 (data not shown). No plaques were seen at pH 2. Regarding thermal stability, the phages maintained their stability relatively well after a 60-minute incubation at 4 °C, 25 °C, 30 °C, and 37 °C but were sensitive to higher temperatures (data not shown).

3.8 Whole genome sequencing and computational analyses

The genome size of vB_AdhS_TS3 was 115,560 bp with a G + C content of 41.10% (Fig. 4). The open reading frames (ORFs) of vB_AdhS_TS3 were identified with a total of 272 predicted ORFs. Among these genes, 58 were predicted to have known functions (Table 2, Fig. 4), and 184 ORFs were predicted to encode hypothetical proteins. Similarly, the genome size of vB_AdhM_TS9 was 115,503 bp, with a G + C content of 35.34%, and encoded 199 proteins. Out of 199 ORFs, 136 ORFs were hypothetical, whereas only 63 ORFs predicted functions (Table 2, Fig. 4). The genome size of vB_AdhM_DL was 42,388 bp, with a G + C content of 34.43% and 79 proteins, respectively. Of the 79 encoded proteins, Only 29 out of 79 encoded predicted functions, whereas 50 ORFs were hypothetical (Table 2, Fig. 4). We did not find an ORF encoding a protein with known toxins, antibiotic-resistant genes (ARGs), virulent factors (VFs) of bacterial origin, or lysogenic markers such as integrase, recombinase, repressor/anti-repressor protein, and excisionase in all three phage genomes.

Close

Fig. 4

Genomic characterization of three bacteriophages targeting A. dhakensis AM, A) vB_AdhS_TS3 B) vB_AdhM_TS9 C) vB_AdhM_DL. Circles from outermost to innermost correspond to predicted genes (BLASTp, nr database, E value of 〈10⁻⁵) on the forward strand, reverse strand, and GC content.

Export to PPT

3.9 Effect of single in pre- and post-treatment to control A. dhakensis AM growth

The lytic effect of individual phages on the growth of A. dhakensis AM was evaluated at different MOIs. Both pre-and post-treatment, the maximum cell decrease for all phages was observed during 6–12 h of incubation at all MOIs compared with the uninfected bacterial control. The pre-treatment with phages vB_AdhS_TS3, vB_AdhM_DL, and vB_AdhM_TS9 reduced the maximum bacterial count by 5.40, 6.67 and 3.91 log CFU/mL, respectively, after 6 h of incubation. In post-treatment, the maximum inactivation was achieved at 12 h with the log reduction number of 4.68, 5.25 and 4.43 log CFU/mL, respectively. The growth of bacteria cultured with phages decreased remarkably depending on the regrowth of bacteria at 48 h in all treatments (Fig. 5). When the phages were incubated in the presence of the host, the phages gradually increased and then became stable over 48 h of incubation. Based on the maximum inhibition, the combination of two phages as a phage cocktail in pre- and post-treatment with optimal MOIs was chosen, as shown in Fig. 6.

Close

Fig. 5

The log reduction in A. dhakensis number in pre- and post-treatment using single phage vB_AdhS_TS3, vB_AdhM_DL and vB_AdhM_TS9. The data were expressed as mean ± SD. All assays were carried out in triplicates.

Export to PPT

Close

Fig. 6

The log reduction in A. dhakensis number in pre- and post-treatment groups using phage cocktail vB_AdhS_TS3, vB_AdhM_DL and vB_AdhM_TS9. The data are expressed as mean ± SD. All assays were carried out in triplicates.

Export to PPT

3.10 Effect of phage cocktail in pre- and post-treatment to control A. dhakensis AM growth

The effectiveness of the phage cocktail in the reduction of A. dhakensis AM is shown in Fig. 6. Cocktail 3, composed of phages vB_AdhM_TS9 and vB_AdhM_DL, was more effective against A. dhakensis AM than the other cocktails. Upon pre-treatment, the maximum inactivation with cocktail 3 (vB_AdhM_TS9 (MOI 1) + vB_AdhM_DL (MOI 0.1)) was 5.08 ± 0.51 log CFU/mL after 6 h of incubation compared with uninfected control. In post-treatment, the maximum reduction with cocktail 3 (vB_AdhM_TS9 (MOI 1) + vB_AdhM_DL (MOI 1)) was 4.71 ± 0.49 log CFU/mL after 12 h of incubation when compared with those of the bacterial control. Bacterial regrowth was observed at 24 h in all treatments. The phage alone was constant throughout the experiment. While phage cocktails hold promise in preventing the emergence of phage-resistant mutants, it's essential to acknowledge that prolonged incubation of phages and bacteria may lead to the development of phage-resistant strains (Malik et al., 2021). Another challenge in phage therapy is the high specificity of phages for their target bacteria. Each bacterial strain often requires a specific phage, and identifying the right phage for a particular infection can be a time-consuming process. In urgent or novel situations, this may not always be feasible. Thus, we have investigated the combination of phages with antibiotics as a strategy to mitigate potential limitations and expand the scope of treatment.

3.11 A. dhakensis growth inhibition by phage cocktail and antibiotics combination

To establish the phage-antibiotic synergy (PAS) effect, we determined the bacterial inactivation by three combinations of phage cocktails with amoxicillin at sub-MIC (32 μg/mL) in different volumes (200 µL and 20 mL). In the presence of amoxicillin and phage alone, the antibiotic- and phage-resistant variants rapidly grew after 6 h of incubation. In the pre-treatment, the combination of phage cocktail 1 or 2 with amoxicillin at sub-MIC resulted in complete inhibition during 48 h and 12 h in a volume of 200 µL and 20 mL, respectively (Fig. 7). At a volume of 20 mL, a significant reduction in bacterial numbers was observed when treated with a combination of phage cocktail 1 or 2 and sub-MIC amoxicillin at 48 h of incubation (p < 0.05). After post-treatment, the combination of phage cocktail 1 or 2 with amoxicillin at sub-MIC resulted in complete inhibition for 48 h in 200 µL (Fig. 7). However, only partial inhibition was observed after 12 h at a volume of 20 mL. Bacterial regrowth gradually increased after 12 h, and no significant reduction in viable bacteria was observed after 48 h of incubation compared to the phage cocktail of antibiotics alone. In this study, the bacterial concentration in this treatment (1 × 10⁵ CFU/mL) was much higher than in natural bacterial contamination. Moreover, this study was performed in a higher volume of medium (20 mL), which may reduce the interaction between phages and/or antibiotics before reaching the bacteria. However, phage cocktails 1 and 2 decreased the CFU 1.2–1.7 log CFU/mL compared to the control and other groups treated individually after incubation for 48 h. Our study strongly suggests that the synergistic antibacterial effects of antibiotics and phages should be performed in the early stages when the bacterial number is low. The first use of the phage–antibiotic synergy (PAS) strategy was described by Comeau et al. (2007). Sublethal concentrations of antibiotics may help lytic bacteriophages reproduce rapidly and promote their antibacterial effects. Additionally, in combination with antibiotics, phages have multiple mechanisms to augment antibiotic effectiveness. They can break down bacterial biofilms using phage enzymes such as depolymerases and lysins, rendering bacteria more susceptible to antibiotics (Liu et al., 2022b). Additionally, this combination therapy can reduce the likelihood of bacterial resistance development to both phages and antibiotics (Segall et al., 2019). Our study underscores the potential of phage-based approaches in combating A. dhakensis and demonstrates the efficacy of combination therapy utilizing phage cocktails and sublethal antibiotic concentrations. Furthermore, the ability of phages to target antibiotic-resistant strains, which are often challenging to treat with antibiotics alone, adds to the value of this approach. By focusing on A. dhakensis, we provide insights that extend to the broader challenge of antimicrobial resistance, emphasizing the importance of exploring innovative strategies to combat this critical global health issue.

Close

Fig. 7

Effect of phage cocktail and amoxicillin combination at 1/2 MIC against A. dhakensis AM. The bar graph represents the bacterial concentration (log CFU/mL), and the line graph represents the phage concentration (log PFU/mL). The data are expressed as mean ± SD. All assays were carried out in triplicates. Each lowercase label corresponds to a significantly different (p < 0.05) bacterial concentration within each time point. Capital letters denote significantly distinct (p < 0.05) bacterial concentrations and time points compared to each other time point within the same conditions.

Export to PPT

Close

Fig. 7

Effect of phage cocktail and amoxicillin combination at 1/2 MIC against A. dhakensis AM. The bar graph represents the bacterial concentration (log CFU/mL), and the line graph represents the phage concentration (log PFU/mL). The data are expressed as mean ± SD. All assays were carried out in triplicates. Each lowercase label corresponds to a significantly different (p < 0.05) bacterial concentration within each time point. Capital letters denote significantly distinct (p < 0.05) bacterial concentrations and time points compared to each other time point within the same conditions.

Export to PPT

Our study demonstrates that phage-based approaches are an attractive way to inactivate A. dhakensis in vitro. The cocktail of three different bacteriophages (phage vB_AdhS_TS3, vB_AdhM_DL and vB_AdhM_TS9) revealed promising in vitro lytic activity on A. dhakensis. Furthermore, the combination therapy using phage cocktails and antibiotics showed greater promise compared with either therapy alone. Moreover, combination therapy can also prevent the development of resistant mutants that would otherwise develop rapidly when exposed to antibiotics or phages. This demonstrates that using phages as an adjuvant with a sublethal concentration of antibiotics is an effective therapeutic strategy.