Phage pAEv1: A potential biocontrol agent against Aeromonas veronii infections

ABSTRACT

Aeromonas veronii, a significant aquaculture pathogen, induces substantial mortality in farmed aquatic species. Given the therapeutic promise of phage-mediated bacterial lysis, we isolated A. veronii from contaminated aquaculture systems and characterized its biological properties. Concurrently, the phage pAEv1 was recovered from the same aquatic environment using the pathogen as a host. Biological characterization revealed an optimal multiplicity of infection (MOI) of 0.01, a 10-min latent period, and a titer of 1.3 × 10⁸ PFU/mL. pAEv1 exhibited notable stability, maintaining infectivity at temperatures ranging from −80°C to 55°C and pH levels of 5–9. Genomic analysis identified a circular double-stranded DNA genome, closely related to the Aeromonas phage pAEv1812 (ID: OL964754.1), with UniProtKB-based functional annotation categorizing encoded proteins into DNA modification and structural classes. COG and KEGG analyses further demonstrated gene enrichment in nucleic acid metabolism and purine biosynthesis, consistent with phage replication machinery. Crucially, virulence factor screening confirmed the absence of resistance or virulence genes, ensuring genetic-level biosafety. In vitro antimicrobial tests showed that phage pAEv1 effectively inhibited A. veronii. Additionally, in vivo experiments demonstrated that grass carp treated with pAEv1 had an 80% survival rate after 7 days of exposure to a double minimum lethal dose (MLD) of the bacterium.

Introduction

Aeromonas veronii, a Gram-negative, rod-shaped, facultatively anaerobic bacterial species, demonstrates ubiquitous distribution in aquatic ecosystems while posing significant infection risks to diverse hosts including teleost fish, amphibians, and humans [1–3]. As a globally distributed aquatic pathogen, A. veronii has incurred substantial economic losses in commercial aquaculture systems through endemic outbreaks and multidrug-resistant lineage proliferation [4,5]. A. veronii demonstrates a broad piscine host range, with confirmed infections in economically vital species including Pelteobagrus fulvidraco, Oreochromis niloticus, Micropterus salmoides, Ctenopharyngodon idella, and Hypophthalmichthys molitrix [6–9]. The pathogen induces epidemic ulcer syndrome and hemorrhagic septicemia, characterized by multifocal hemorrhagic/ulcerative lesions in fish tissues and organs, ultimately causing mass mortality in aquaculture systems [10,11].

Antibiotics remain the primary prophylactic and therapeutic agents against Aeromonas spp. infections in aquaculture. However, indiscriminate antibiotic use frequently results in residual drug accumulation and the development of multidrug-resistant bacterial strains, posing serious threats to aquatic ecosystems and public health [12–14]. Preena et al. identified A. veronii isolated from diseased Oreochromis niloticus as the most multidrug-resistant pathogen, exhibiting an antimicrobial resistance (MAR) index of 0.46 [15]. Similarly, Algammal et al. identified A. veronii strains exhibiting broad-spectrum antimicrobial resistance, harboring bla_TEM, bla_CTX-M, bla_SHV, tetA, aadA1, and sul1 resistance determinants [16]. Consequently, developing alternative control strategies to reduce antibiotic application is imperative, thereby mitigating the dual threats of multidrug-resistant bacterial emergence and drug residue accumulation in aquatic-derived food products.

Phages are obligate bacterial viruses characterized by strict host specificity, composed of a nucleic acid genome encapsulated within a protein capsid. They are classified into lytic phages and lysogenic phages [12,17]. Phages recognize specific bacterial hosts by detecting receptor molecules, including flagella, outer membrane protein C, lipopolysaccharide O antigen, vitamin B12 transporter, and ferrichrome outer membrane transporter, initiating infection processes that culminate in cellular lysis [18]. Phage therapy has garnered growing recognition as a therapeutic alternative, serving to mitigate antibiotic consumption, delay the development of antimicrobial resistance, and target multidrug-resistant bacterial pathogens [19]. Phage has demonstrated therapeutic efficacy against infections caused by A. veronii, Aeromonas hydrophila and Vibrio harveyi [20–22].

In this study, the pathogenic bacterium A. veronii was first isolated from aquaculture water, with subsequent biological identification and characterization. Using A. veronii as the host, phage pAEv1 was isolated, followed by biological characterization and genome-wide analysis to assess its safety. The antimicrobial capacity and therapeutic efficacy of pAEv1 were ultimately evaluated through in vitro and in vivo activity studies against A. veronii. This study holds significant value for elucidating the biological properties and lytic mechanism of phage pAEv1, and for advancing the application of phage therapy in preventing and controlling A. veronii infections in aquaculture.

Materials and methods

Materials

A. veronii (AEv1, PV991636) strains and phages (pAEv1, PX021363) were isolated from aquaculture water samples.

Isolation and culture of bacteria

Aeromonas screening using RS (Rimler-Shotts) medium. Water samples were uniformly spread on RS agar plates and incubated inverted at 28°C for 12 h. Three successive purification cycles using zonal streaking were conducted, with documentation of colony morphology and cultural characteristics. Isolated colonies were subsequently transferred into 30% glycerol-LB medium for cryopreservation at −80°C.

Bacterial morphological observations

The bacterial cells were Gram-stained and subsequently examined using a light microscope. For ultrastructural analysis, cultures were first synchronized to the logarithmic growth phase. Cells were then harvested by centrifugation at 4000 rpm for 10 min and subjected to three PBS washing cycles. Following overnight fixation with 2.5% glutaraldehyde at 4°C, the samples were washed three additional times with PBS. Negative staining was performed using 2% phosphotungstic acid (pH 6.8) for 10 min at room temperature. Morphological features were ultimately examined using a Hitachi HT7800 transmission electron microscope (Hitachi High-Tech, Japan) operated at 80 kV.

Molecular biological identification

The bacterial genomic DNA was isolated, and universal primers 27F (5”-AGAGTTTGATCCTGGCTCAG-3‘) and 1492 R (5’-TACGGTTACCTTGTTACGACTT-3”) were used to amplify the 16S rRNA gene [23–25]. The amplified products were purified and sequenced by Sanger methodology. Nucleotide sequences were analyzed through the NCBI Basic Local Alignment Search Tool (BLAST; https://blast.ncbi.nlm.nih.gov/Blast.cgi) for similarity assessment, with threshold identity values set at ≥98% for species-level classification.

Physiological and biochemical determination

Physiological and biochemical characterization of the strains was performed using single-tube biochemical identification kits (Guangdong Huankai Microbial Technology Co., Ltd.). Includes the following 12 single-tube biochemical identification kits: Oxidase Test (029172), Spot Test Veoges-Proskauer Reagents (029050), OF (Hugh-leifson 075,640), Kovacs Indole Reagent (029030), Gelatine Biochemical Identification Tubes (075310), Hydrogen Sulfide Biochemical Identification Tubes (075300), Lysine decarboxylase biochemical identification tube (075280), Sucrose (075050), Rhamnose (075110), Arabinose (075060), Maltose (075030) and Esculin (075160). All experimental procedures and result interpretations strictly followed the manufacturer’s instructions provided in the single-tube biochemical identification kit manual.

Detection of bacterial haemolysis

Log-phase cultures were streaked onto 5% sheep blood agar plates (Thermo Scientific, Oxoid brand) and aerobically incubated at 28°C for 12 h. Hemolysis patterns were classified according to three characteristic phenotypes: β-hemolysis (complete erythrocyte lysis with clear zones), α-hemolysis (partial lysis with greenish discoloration), and γ-hemolysis (non-hemolytic). All determinations were performed by two independent observers with 100% inter-rater concordance across three biological replicates.

Determination of bacterial growth curve

Isolated colonies were selected and inoculated into 50 mL of sterile LB broth in 250 mL Erlenmeyer flasks. The cultures were incubated at 28°C with continuous shaking (120 rpm) using an inverter shaker. Optical density measurements were performed at 1 h intervals over a 12-h period. Bacterial growth was monitored by measuring the absorbance at 600 nm (OD₆₀₀) using a UV759S spectrophotometer. The spectrophotometer was blank-corrected with sterile LB medium prior to each measurement. Growth curves were constructed by plotting OD₆₀₀ values against incubation time, with three independent biological replicates being analyzed.

Determination of antibiotic sensitivity of bacteria

The bacterial suspension was adjusted to 1.5 × 10⁸ CFU/mL (0.5 McFarland standard) and evenly spread onto MH medium plates. Drug sensitivity tablets were affixed onto the agar surfaces, followed by incubation at 28°C for 12 h. The results were observed, and the inhibition zone diameters were measured using a vernier caliper. The antibiotic susceptibility was interpreted based on the inhibition zone diameters according to the Clinical and Laboratory Standard Procedure (CLSI) guidelines (CLSI M31-A3) [26].

Isolation and purification of phages

Water samples were filtered through a vacuum filtration system using sequentially arranged membrane filters (0.7 μm, 0.45 μm, and 0.22 μm pore sizes). The obtained filtrate was mixed with a culture medium at a 1:2 (v/v) ratio, followed by inoculation with isolated single colonies. The mixture was incubated in a constant-temperature inverter shaker at 28°C with 120 rpm agitation for 24 h.

The supernatant was subsequently centrifuged and filtered through a 0.22 μm membrane filter. Serial 10-fold dilutions were prepared using SM buffer. A 100 μL aliquot of phage dilution was combined with 900 μL of log-phase bacterial culture, and the mixture was allowed to adsorb for 10 min. This was then mixed with 9 mL of semi-solid LB medium.

The LB medium mixture was homogenized and plated onto LB agar plates. After solidification of the top agar layer, the plates were transferred to a 28°C incubator for overnight cultivation. Successive purification cycles were performed by selecting individual phage plaques until uniformly sized plaques were obtained, yielding isolated monophasic phage preparations.

Phage potency assay

Phage titers were determined by the double-layer agar plate technique [27]. Briefly, phage lysate (in SM buffer) was mixed with host bacterial suspension (1:1 ratio), incubated at 28°C for 20 min for adsorption, and combined with semi-solid LB agar (0.7%). The mixture was overlaid onto solidified LB agar plates, incubated at 28°C overnight, and plaques were counted. Titers were calculated as:

Phage titer (PFU/mL) = plaque count × dilution factor × 10

The arithmetic mean of triplicate measurements was recorded as the final titer value.

Determination of phage host spectra

Phage-host interaction profiles were characterized through the dot-drop assay protocol. Specifically, 200 μL of log-phase bacterial culture was uniformly plated onto LB agar plates. Subsequently, 10 μL of phage stock solution was aliquoted and applied onto the agar surfaces. The inoculated plates were allowed to equilibrate at ambient temperature for 10 min, followed by inversion and incubation at 28°C for 16–18 h in an incubator. Post-incubation, plates were examined under oblique illumination at 45°C angle for detection of lytic activity, with the presence of translucent zones (≥1 mm diameter) being recorded as positive interactions.

Determination of optimal multiplicity of infection by phage

Bacterial and phage suspensions were standardized to concentrations of 1 × 10⁸ CFU/mL and 1 × 10⁵ PFU/mL, respectively. Phage suspensions were inoculated at multiplicity of infection (MOI) ratios spanning 0.001 to 10 (logarithmic increments). Following a 10-min adsorption period at 28°C, the mixtures were transferred into 10 mL of LB broth and subjected to primary incubation under inverter shaker (120 rpm) at 28°C for 6 h. The cultures were subsequently processed through centrifugation (5000 rpm, 15 min), after which clarified supernatants were obtained through 0.22 μm membrane filtration. Serial 10-fold dilutions were prepared in SM buffer, with phage titers subsequently quantified using the standardized double-layer agar assay.

Phage one-step growth curve determination

The methodology was adapted from Zhao et al. with minor refinements to align with experimental requirements [28]. A 1 mL aliquot of standardized bacterial suspension (1 × 10⁸ CFU/mL) was mixed with phage inoculum at the predetermined optimal multiplicity of infection. The mixture was subjected to adsorption at 28°C for 20 min, followed by centrifugation (12,000 rpm, 2 min). The supernatant was carefully decanted, with the pellet being resuspended in 1.5 mL of pre-equilibrated LB broth (28°C). This washing procedure was repeated twice to ensure complete removal of unadsorbed phage particles. The final pellet was reconstituted in 10 mL of temperature-controlled LB medium (28°C) and homogenized by vortex mixing. The suspension was incubated under controlled conditions (28°C, 120 rpm) with sequential 300 μL samples being aseptically collected at 10-min intervals over a 150-min period. All samples were immediately filtered through 0.22 μm membranes and appropriately labeled. Phage titers were quantified via the standardized double-layer agar assay. Growth kinetics were graphically represented with a one-step growth curve, where temporal progression (min) was plotted on the abscissa against logarithmic phage titers (PFU/mL) on the ordinate.

Determination of temperature and pH tolerance of phage

Aliquots of phage lysate (1 mL) were distributed into sterile 1.5 mL microcentrifuge tubes and subjected to exposure at defined temperatures (−80°C, 4°C, 25°C, 35°C, 45°C, 55°C, 65°C, 75°C) for 1 h. Phage viability was quantified post-treatment using the standardized double-layer agar assay.

For pH stability assessment, LB broth was autoclaved (121°C, 15 min), and the pH was adjusted post-sterilization to values ranging from 2 to 11 using 0.22 μm filter-sterilized 1 M HCl or NaOH solutions [18]. Phage lysate was introduced into each pH-adjusted medium and incubated for 1 h under 28°C. The surviving phage titer was quantified via the double-layer agar plate method: the base layer consisted of pH-adjusted LB medium solidified with 1.5% agar, while the top layer contained a semi-solid medium (0.7% agar) mixed with the phage-host suspension. Serial dilutions were plated, incubated at 28°C for 12 h, and plaque-forming units (PFU) were enumerated.

Phage purification and concentration

A phage suspension (10 mL at 10⁸ PFU/mL) was combined with 30 mL of log-phase bacterial culture in 300 mL LB medium. The mixture was incubated under inverter shaker (120 rpm) at 28°C for 8 h, followed by primary centrifugation (5,000 rpm, 10 min). The clarified supernatant was filtered through a 0.22 μm membrane. DNase I and RNase A were supplemented to final concentrations of 1 μg/mL, with concurrent addition of NaCl (5.84 g/100 mL) under ice-bath conditions (0–4°C) for 1 h. The treated lysate was subjected to secondary centrifugation (7,000 rpm, 10 min) to pellet cellular debris. PEG8000 was gradually added to achieve 15% (w/v) concentration, followed by phage precipitation during ≥6 h ice-bath incubation. The precipitated phage particles were collected through centrifugation (7,000 rpm, 20 min, 4°C). The pellet was resuspended in sterile SM buffer to obtain concentrated phage preparation.

Phage electron microscopy

Phage morphology was visualized under standardized imaging conditions using a transmission electron microscope (HT7800, Hitachi, Japan). A 20 μL concentrated phage suspension was deposited onto copper grids and allowed to adsorb for 20 min under ambient conditions. Excess fluid was carefully wicked away using filter paper, followed by negative staining with 2% (w/v) phosphotungstic acid (pH 6.8) for 5 min. Residual stain was removed, and grids were air-dried prior to TEM examination at 80 kV accelerating voltage.

Phage pAev1 genome determination

Phage nucleic acid was extracted using a viral genomic DNA isolation kit (Tengen Biotechnology Co., Ltd.). Quality-verified DNA samples were fragmented by random shearing to generate appropriately sized segments. The 3’-termini of these fragments were polyadenylated (“A-tailing”), enabling adapter ligation through complementary T-overhangs. A sequencing library was constructed via PCR amplification of adapter-ligated DNA fragments. Library preparations were processed through cluster generation and sequenced on an Illumina platform. Raw sequencing data quality was assessed using SOAPnuke (v2.0.5), with low-quality reads (Q-score < 20) being filtered to yield high-fidelity clean reads. Viral genome architecture was reconstructed through Circos visualization (v0.69–9), integrating sequence alignment data, functional annotations, and component analysis results.

Virus and species identification

Assembled contigs were subjected to CheckV database analysis for viral sequence identification, with sequencing depth profiles being generated through Samtools depth. Target viral contigs were subsequently identified through integration of contig completeness metrics (CheckV quality tiers), coverage depth values, and associated genomic parameters. Clean reads were aligned to selected viral contigs using BWA-MEM with minimum alignment length thresholds set at ≥80% of read length. Alignment files were processed through Samtools sort/index to generate coverage maps, visualized as depth distribution profiles across contig coordinates.

Gene prediction and functional analysis

Gene prediction on viral contigs was performed using Prokka software, with the quantity and length of identified coding sequences being systematically recorded. Predicted gene products were subjected to comparative analysis against the UniProtKB, KEGG, and COG databases to acquire functional annotations. These annotated viral sequences were processed through Virsorter2 for phage characterization, where scoring and classification were executed based on viral hallmark gene presence, homologous viral gene matches, and host-associated genetic elements prior to DRAM-compatible formatting. Phage genomic regions were annotated through DRAM pipeline analysis, with auxiliary metabolic genes (AMGs) being identified through conserved domain screening. Functional classification of AMGs was conducted according to the KEGG ontology, followed by categorical quantification based on metabolic pathway associations. Statistical summaries were generated to reflect AMG type distribution and biochemical process involvement.

Comparative proteomic phylogenetic analysis

Comparative proteomic phylogenetic analysis was conducted via the Viral Proteomic Tree (ViPTree) server (https://www.genome.jp/viptree; accessed 1 December 2024) [29], with comparisons against phage entries from the Virus-Host DB (https://www.genome.jp/virushostdb; accessed 1 December 2024) and NCBI GenBank. Phages exhibiting the highest tBLASTx scores (SG) and outgroup phages with minimal scores (SG) were identified from ViPtree output, combined with NCBI-derived top BLASTn hits, as input for VIRIDIC. Pairwise intergenomic similarities between phage PAEv1 and selected phages were subsequently computed using the VIRIDIC tool [30]. Additional phages filtered from the NCBI Virus database (filter criteria: Aeromonas host + complete RefSeq genomes) were analyzed using the CoreGenes 5.0 platform (https://coregenes.ngrok.io/; accessed 1 December 2024) to identify genus- and family-level conserved genes [31]. Terminase large subunit and major capsid protein sequences were employed for protein-based phylogenetic tree construction using MEGA 11 (https://www.megasoftware.net/; accessed 1 December 2024).

In vitro bacteriostatic assay of phage pAev1

Bacterial cultures were initiated through the inoculation of single colonies into 50 mL sterile medium, followed by the incubation at 28°C with inverter shaker (120 rpm) until a logarithmic growth phase was achieved. Cell densities were standardized to 1 × 10⁷ CFU/mL through dilution with sterile PBS. Six experimental conditions were established: Negative Control Group A (28 mL LB broth +1 mL bacterial suspension +1 mL PBS), Experimental Group B (MOI 10), Experimental Group C (MOI 1), Experimental Group D (MOI 0.1), Experimental Group E (MOI 0.01), and Experimental Group F (MOI 0.001). All test groups were subjected to continuous incubation under identical conditions (28°C, 120 rpm). Optical density measurements at 600 nm were recorded hourly using a UV759S spectrophotometer, with three independent biological replicates being performed for each experimental series.

In vitro bactericidal assay of phage pAev1

Bacterial cultures were initiated by inoculation of fresh single colonies into six aliquots of 100 mL LB broth followed by incubation under controlled conditions (28°C, 120 rpm inverter shaker). Cultivation was maintained until an optical density of 0.5 at 600 nm (OD₆₀₀) was achieved using a spectrophotometer calibrated with sterile medium blanks. Experimental groups were exposed to phage pAEv1 suspensions at predetermined optimal multiplicity of infection (MOI) values, with triplicate biological replicates (n = 3) being established for each treatment condition. Parallel control cultures (n = 3) were maintained under identical environmental parameters without phage supplementation. Continuous incubation was performed over 180 min at 28°C with synchronized agitation (120 rpm), during which turbidimetric measurements at 600 nm wavelength were systematically recorded at 30-min intervals using matched quartz cuvettes.

Determination of minimum lethal dose for AEv1 bacterial infection in grass carp

A single colony was inoculated into LB liquid medium and incubated at 28°C until an OD₆₀₀ value of 0.5 was attained. The bacterial culture was centrifuged at 6,000 rpm for 5 min under 4°C conditions, after which the supernatant was discarded and the pellet subjected to three consecutive washes with phosphate-buffered saline (PBS) of equivalent volume. Final resuspension was performed in sterile PBS. Uniformly healthy grass carp (15 cm average length) were randomly allocated into six experimental cohorts containing ten specimens each. Bacterial suspensions were administered via intraperitoneal injection at graded concentrations: Group 1 received 10⁴ CFU/fish, Group 2 10⁵ CFU/fish, Group 3 10⁶ CFU/fish, Group 4 10⁷ CFU/fish, and Group 5 10⁸ CFU/fish, while the control group received equivalent volumes of PBS. All cohorts were maintained under identical environmental parameters for 7 days, with daily monitoring of survival rates and clinical manifestations. The minimum lethal dose (MLD) of strain AEv1 for grass carp was defined as the lowest bacterial concentration inducing 100% mortality within the observation period.

Therapeutic protection assay of phage pAev1

Healthy grass carp were randomly allocated into three experimental cohorts containing ten specimens each, comprising one therapeutic group and two control groups. The therapeutic cohort received sequential intraperitoneal injections: initial administration of strain AEv1 suspension (1 × 10⁷ CFU/mL) followed by phage pAEv1 treatment at predetermined optimal multiplicity of infection (MOI) parameters (PFU/fish) one-hour post-infection. Negative control specimens were administered sterile phosphate-buffered saline (PBS) via identical injection protocol, while positive controls received bacterial challenge without subsequent phage intervention. All groups were maintained under standardized environmental conditions for seven consecutive days, with daily quantification of survival percentages and documentation of clinical manifestations to assess therapeutic efficacy.

Statistical analysis

Experimental data were expressed as mean ± standard deviation (Mean ± SD) and analyzed through one-way ANOVA using SPSS 26.0 software, with statistical significance defined at p < 0.05. Graphical representations were generated via GraphPad Prism 9. All experiments were conducted in triplicate.

Results

Isolation and identification and biological characterization of A. veronii

After isolation and purification, strain AEv1 formed round single colonies with white edges and yellow centers on LB solid medium. Microscopic observation revealed short red rod-shaped cells (Figure 1(A)), which were Gram-negative. Physiological and biochemical characterization results are presented in Figure 1(C). The strain tested positive for oxidase activity, gelatin liquefaction, hydrogen sulfide production, indole test, and V-P test. It demonstrated lysine decomposition capability but failed to metabolize rhamnose, sucrose, arabinose, or esculin. The O-F fermentation test classified it as F-type. 16S rRNA gene sequence analysis revealed that AEv1 exhibited 99% homology with isolates A. veronii (MF716720.1) as per the NCBI database. Moreover, phylogenetic analysis revealed that AEv1 was located on the same branch as A. veronii (Figure 1(B)). Combined with morphological, physiological and biochemical characteristics, it was tentatively identified as A. veronii. In addition, we conducted whole-genome sequencing of AEv1. The full gene sequences of the top five A. veronii and the top three A. hydrophila strains from the search results were downloaded from the NCBI database (Table S1), and the genome sequence of strain AEv1 was analyzed using ANI (Average Nucleotide Identity) and dDDH (digital DNA-DNA hybridization). ANI values greater than 95% and dDDH values greater than 70% indicate that the two strains are the same species of bacteria [32,33]. The results showed that AEv1 had ANI values greater than 95% and dDDH values greater than 70% with A. veronii strains (CF11FDAARGOS_632, HD6448, HD6454, 32286, and BIM B-1812) and lower than 95% and lower than 70% with the other selected strains (Table S2), which further indicated that the AEv1 strain is A. veronii.

Figure 1.

Identification of properties of AEv1. (A) Gram stain test (used distinguished two different kinds of bacteria). (B) Evolutionary tree of properties of AEv1. (C) Bacterial physiological and biochemical experiments. ‘+’ indicated positive; ‘-’ indicated negative. F indicated type of AEv1.

When cultured on blood agar plates, AEv1 exhibited a distinct zone of hemolysis as shown in Figure 2(A). The growth curve in Figure 2(B) demonstrates a rapid proliferation rate, with a logarithmic growth phase occurring between 2 and 8 hours. Transmission electron microscopy revealed that AEv1 was a bacillus possessing unipolar flagella and filamentous structures (Figure 2(C)). Antibiotic susceptibility testing (Figure 2(D)) indicated resistance or intermediate sensitivity to most tested antibiotics, with sensitivity observed to only a limited subset, confirming AEv1 as a multidrug-resistant A. veronii strain.

Figure 2.

Biological properties of AEv1. (A) Hemolysis activity, the black arrow indicates β-hemolysis. (B) Growth curve (C) Transmission electron microscopy observation. (D) Antibiotic susceptibility. Coloured boxes indicated areas where AEv1 is sensitive to antibiotics.

Isolation and purification of phage from A. veronii

A lytic phage specific to A. veronii AEv1, named pAEv1, was isolated from natural water bodies using the double-layer agar plate method. Following multiple purification cycles, pAEv1 was propagated overnight at 28°C, forming uniformly sized plaques (Figure 3(A)), with the phage stock solution demonstrating a titer of 1.3 × 10⁸ PFU/mL.

Figure 3.

Morphological observations of phage pAEv1. (A) Spot plate observation. (B) Transmission electron microscopy observation. The white line segment indicated that the scale of the map is 100 nm.

Transmission electron microscopic observations of phage

As shown in Figure 3(B), transmission electron microscopy revealed that phage pAEv1 exhibited a short tail structure.

Analysis of phage host profiles

Host range analysis of pAEv1 is demonstrated in Table 1. The phage exhibits narrow host specificity, lysing only its original host strain and an additional A. veronii isolate obtained from Chinese soft-shelled turtles.

Table 1.

The host range of pAEv1.

Bacterial species	Bacterial sources	pAEv1 Efficiency	Bacterial species	Bacterial sources	pAEv1 Efficiency
A. hydrophila	Standard strain (CICC)	NO	A.veronii	Standard strain (CICC)	YES
A. hydrophila	grass carp	NO	A.veronii	grass carp	NO
A. hydrophila	Crucian carp	NO	A.veronii	Crucian carp	NO
A. hydrophila	Soft-shell Turtle-1	NO	A.veronii	Soft-shell Turtle-1	YES
A. hydrophila	aquaculture organism-1	NO	A.veronii	aquaculture organism-1	NO
A. hydrophila	Pelophylax nigromaculatus	NO	A.veronii	Soft-shell Turtle-3	NO
A. hydrophila	Soft-shell Turtle-2	NO	A.veronii	Soft-shell Turtle-2	NO
A. hydrophila	aquaculture organism-2	NO	A.veronii	aquaculture organism-2	NO

“YES” indicates that the bacterium is capable of being lysed by the phage.

Determination of optimal phage infection multiplicity

As demonstrated in Figure 4(A), phage pAEv1 exhibited an optimal MOI of 0.01, achieving maximum progeny production with a titer of 1.0 × 10⁸ PFU/mL when applied at this infection ratio.

Figure 4.

Biological properties of pAEv1. (A) Determination of optimal MOI. Different letters indicate significant differences (p <0.05). (B) One-step growth curve analysis. (C) Phage tolerance various temperatures. (D) Phage tolerance to different pH levels.

One-step growth curve of phage

The one-step growth curve demonstrates phage growth dynamics by tracking progression from host infection through replication to eventual stabilization. This characteristic triphasic pattern comprises latency, lysis, and stabilization phases. Figure 4(B) reveals that phage pAEv1 exhibited a 10-min latency period prior to lysis phase initiation, achieving stabilization within 90 min post-infection.

Determination of phage tolerance to different temperatures and pH

Phage pAEv1 stability was evaluated across temperature and pH gradients to assess environmental tolerance for potential applications. Figure 4(C) demonstrates maintained viability at low temperatures, with significant attenuation above 55°C and residual activity persisting at 75°C, indicating superior thermal tolerance. Figure 4(D) reveals optimal stability at neutral pH, while complete viability loss occurred under extreme acidity (pH 2.0) or alkalinity (pH 12.0), though moderate pH variations retained partial activity.

Phage whole genome

High-throughput sequencing data underwent host genome decontamination followed by length-based filtering to yield clean reads (Table S3 and S4). Assembly validation confirmed read usage metrics through methodological verification (Table S5). Comparative genomic analysis revealed Aeromonas phage pAEv1812 as the nearest relative (Table S6), while overlapping terminal sequences confirmed circular genome configuration (Table S7). Predicted gene features are cataloged in Table S8. UniProtKB/Swiss-Prot annotation identified conserved structural domains, with phage-specific functional validation conducted for 12 target genes (Table S9).

Protein sequences were aligned against COG database entries using BLASTp (v2.9.0+), with best-hit matches (E-value < 1e-5) retained for functional annotation (Figure 5(A)). Predominant viral functions comprised replication/recombination/repair systems and mobilome components (prophages, transposons), alongside transcription-related and cell cycle regulation machinery. KEGG pathway analysis revealed viral strategies for cell cycle manipulation and genetic repair mechanisms. Database annotations demonstrated KEGG’s utility for elucidating biological system interactions across cellular and organismal levels. Minimum E-values from BLAST alignments were extracted per sequence, with distribution patterns visualized in Figure 5(B). Annotated proteins were categorized by functional class into DNA modification enzymes and structural components. The absence of additional conserved domain annotations suggested phylogenetic novelty of this phage strain. Integrated analysis incorporating viral identification, coding sequence prediction, and COG annotation enabled Circos visualization of the genomic architecture (Figure 6). Functional classification revealed eight metabolic modules, ten cellular processing/signaling systems, five information storage/processing units, and three uncharacterized elements.

Figure 5.

Functional annotation of phage pAev1. (A) COG functional annotation of pAev1. (B) KEGG functional annotation of pAev1, including pathway level 1, pathway level 2 and pathway level 3.

Figure 6.

Phage PAEv1 genome visualization mapping.

Comparative genomic analysis of PWJ1_01 (phage pAEv1) with reference phages using ViPTree produced a circular proteomic tree, demonstrating its clustering within an unclassified viral family that infects Pseudomonadota-class bacterial hosts (Figure 7(A)). Top-scoring phages were selected for detailed rectangular tree construction, revealing PWJ1_01’s phylogenetic relationship with Casjensviridae family members (Figure 7(B)). Among them, PWJ1_01 is on the same branch as Aeromonas phage BUCT551 (NC_052986), which is a confirmed member of the family Casjensviridae, suggesting that PWJ1_01 may also belong to this family. Whole-genome comparison of PWJ1_01 with closest relatives Aeromonas phages BUCT551, LAh_7, and vB_AhyS-A18P4 showed conserved genomic patterns (Figure 7(C)).

Figure 7.

Proteomic phylogeny and whole-genome alignment on ViPTree. (A) Circular proteomic tree of aeromonas phage PWJ1_01 and related RefSeq phage genomes on NCBI. (B) Rectangular proteomic tree of top matches to phage PWJ1_01. (C) Whole-genome alignment of Aeromonas phage PWJ1_01 with the closest match Aeromonas phage BUCT551, LAh_7, and vB_AhyS-A18P4.

Genomic pairwise analysis using VIRDIC generated 21 clusters at the species level and 5 clusters at the genus level (Table S11). PWJ1_01 exhibited the highest genomic similarity to Aeromonas phage_pAEv1812 (OL964754.1), Aeromonas phage BUCT551 (NC_052986), and Aeromonas phage_phiA034 (OP792756.2) (Figure 8). Phylogenetic relationships were further investigated through MEGA11-based analyses (https://www.megasoftware.net/, accessed on 30 November 2024) using terminase large subunit (Figure 9(A)) and major capsid protein (Figure 9(B)) sequences. Phylogenetic trees of the terminase large subunit and major capsid protein revealed that PWJ1_01 clustered on the same branch as Aeromonas phage_pAEv1812 (OL964754.1), Aeromonas phage BUCT551 (NC_052986), and Aeromonas phage_phiA034 (OP792756.2), further supporting PWJ1_01 as a novel member of the family Casjensviridae.

Figure 8.

VIRDIC heatmap for PWJ1_01 and closely related phage genomes based on BLASTn top hits.

Figure 9.

Phylogenetic analysis of Aeromonas phage PWJ1_01 and selected phages on the amino acid sequences of the terminase large subunit (A) and the major capsid protein (B). The phylogenetic trees were constructed using MEGA 11 software by neighbor-joining (NJ) method and 10,000 replications of bootstrap. Values at nodes indicated bootstrap support.

In vitro bacteriostatic and bactericidal activity of phage pAev1 against A. veronii

The in vitro inhibition assay is shown in Figure 10(A), where the OD₆₀₀ value of the A. veronii control group reached a steady state of approximately 1.450 at 10 h. In contrast, bacterial growth was inhibited in the phage-treated group, with OD₆₀₀ values remaining lower than the control until 5 h when values began to increase. This delayed increase may result from emerging phage-resistant strains or bacterial evasion mechanisms such as quorum sensing and motility.

Figure 10.

Antibacterial activity of phage pAEv1. (A) in vitro bacterialstatic assay. (B) in vitro bactericidal assay.

The in vitro bactericidal assay is shown in Figure 10(B), where the OD₆₀₀ value of the control group increased to 1.402 after 180 min, while the pAEv1-treated group showed a significant decrease to 0.232.

Therapeutic protection activity of grass carp after phage pAEv1 treatment

An infection model of A. veronii in grass carp was successfully established. Mortality was monitored for 7 days post intraperitoneal injection with bacterial concentrations ranging from 10⁴ to 10⁸ CFU/tail. The control group received an equivalent volume of sterile PBS (Figure 11(A)), maintaining 100% survival throughout the observation period (Figure 11(C)). Complete mortality occurred in all fish injected with ≥10⁷ CFU/tail, establishing the minimum lethal dose (MLD) as 10⁷ CFU/tail for A. veronii in this model.

Figure 11.

Therapeutic protection activity of grass carp after phage pAEv1 treatment. (A) Schematic diagram of the grass carp infection model experiment. (B) Schematic diagram of the grass carp treatment experiment. (C) Survival of grass carp after injection with various concentrations of A. veronii. Different letters indicate significant differences (p <0.05). (D) Survival rate of grass carp after phage pAEv1 treatment. Different letters indicate significant differences (p <0.05).

Therapeutic protection activity is shown in Figure 11(B), where grass carp challenged with 2×MLD of A. veronii received phage pAEv1 administration 1 h post-infection. The pAEv1-treated group demonstrated an 80% survival rate after 7 days (Figure 11(D)).

Discussion

A. veronii is an opportunistic pathogen responsible for hemorrhagic septicemia and ulcerative diseases in aquaculture, posing significant threats to aquatic ecosystems and food safety [34]. The widespread misuse of antibiotics in aquaculture has exacerbated environmental pollution, accelerated the emergence of multidrug-resistant bacterial strains, and increased risks of antibiotic residues in aquatic products [35]. These challenges underscore the urgent need for sustainable alternatives to conventional antibiotics.

In this study, we isolated and characterized the lytic phage pAEv1 from aquaculture water samples. Biological characterization revealed that phage pAEv1 exhibited typical features of a short latent period (10 min) and sustained lytic activity with a burst time of 80 min. The phage demonstrated optimal therapeutic efficacy at an MOI of 0.01, suggesting high amplification efficiency under controlled conditions [36]. However, its environmental tolerance was restricted to low temperatures and neutral pH, with viability declining under extreme conditions – a limitation common to many phages that must be considered for field applications [37].

Whole-genome sequencing of pAEv1 revealed a 0.23 Mb double-stranded DNA genome with 56.17% GC content, showing the closest homology (72.077%) to Aeromonas phage pAEv1812 (GenBank: OL964754.1) (Table S6). Based on the latest classification framework (release 38) of the International Committee on Taxonomy of Viruses (ICTV) [38] and phylogenetic analysis, phage PAEv1 is proposed to be a novel member of the family Casjensviridae. Functional annotation confirmed the absence of virulence or antibiotic resistance genes, genetically validating its biosafety for therapeutic use [39]. Notably, the genome encoded only nine annotated proteins, primarily involved in DNA modification and structural assembly, suggesting pAEv1 represents a novel phage lineage with streamlined genetic machinery. COG and KEGG pathway analyses further linked its metabolic activity to nucleic acid processing and purine metabolism, likely supporting rapid virion assembly during infection.

Our study showed that phage pAEv1 at different MOIs inhibited bacterial growth for approximately 5 h; interestingly, after this period, the bacteria resumed gradual growth, presumably the bacteria became resistant to the phage. We sequenced the whole genome of the bacterium AEv1 and analyzed its CRISPR-Cas system (Table S10). The results revealed that eight clusters of regularly interspaced short palindromic repeats (CRISPR) were detected in the genome of AEv1, with 31 distinct spacer regions (spacers) and 19 short repeat sequences (DRs) identified (Figure S1). The CRISPR-Cas system serves as the core immune mechanism of bacteria against invasion by exogenous genetic material (e.g. phages or plasmids), and its structural and functional features are directly related to the host’s phage resistance [40]. The wide distribution of CRISPR clusters (eight clusters) and high diversity of spacer regions (31 spacers) in the strain suggest that it may be chronically exposed to a complex aquatic phage environment, developing immune memory through the capture of phage gene fragments and thereby enhancing recognition and clearance of specific phages [41]. In addition, phage pAEv1 treatment significantly decreased bacterial concentration, indicating its potent bactericidal activity. Meanwhile, the survival rate of the pAEv1-treated group reached 80% 7 days after a 2 × MLD attack, suggesting the potential for treating A. veronii infections. Phages are increasingly being used to prevent and treat bacterial infections in aquatic animals. Liang et al. [42] identified a novel phage, vB_AhaP_PT2, which was effective in treating carp enteritis caused by A. hydrophila-TPS without side effects. Similarly, phage vB_Va_ZX-1 treatment protected Chinese mitten crab larvae from Vp24 infection, with a survival rate of up to 90% after 7 days, whereas the untreated control died rapidly within 3 days [43]. Phage therapy for A. veronii infections has also been studied. Wu et al. [12] evaluated 11 phages for the control of A. veronii infections in Pelteobagrus fulvidraco by immersion and found that two phages, ZPAV-18 and ZPAV-25, provided excellent prophylaxis rates of 85.7% and 71.4%, respectively, and treatment rates of 85.7% and 85.7%, respectively. However, the narrow host range of pAEv1—limited to specific A. veronii genotypes including the standard strain, our isolate, and Soft-shell Turtle-1 strain (Table 1) – highlights a critical constraint for broad-spectrum applications. This strain specificity, while advantageous for targeted therapy in localized outbreaks, may restrict its utility in diverse aquaculture settings harboring heterogeneous A. veronii populations. To address this limitation, future efforts should prioritize combinatorial strategies such as phage cocktails or engineered phage hybrids targeting complementary bacterial receptors. For example, the use of phage mixtures was effective in protecting the structure and bacterial community of the gut of Carassius auratus gibelio and was effective against A. veronii infection of Carassius auratus gibelio [20].

Conclusions

In conclusion, our findings position phage pAEv1 as a strain-specific therapeutic candidate against A. veronii infections in controlled environments. While its narrow host range and environmental sensitivity necessitate cautious deployment, these limitations also underscore the value of precision phage therapy in mitigating antibiotic resistance. Further validation of phage-host adaptation mechanisms and large-scale trials in polymicrobial aquaculture systems will be essential to translate this potential into sustainable practices.

Funding Statement

This research was funded by the National Natural Science Foundation of China [32073020, 32201960].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Arrive statement

The study adhered to the ARRIVE guidelines in the methods section of the manuscript.

Data availability statement

The data that support the findings of this study are openly available in Figshare at https://doi.org/10.6084/m9.figshare.29791340.v1 [44].

Ethical statement

The project has applied to the Ethics Committee of Hunan Agricultural University under License No. 430520 (application date 20 February 2024).

List of Abbreviations

MOI

The optimal multiplicity of infection

pAEv1

Aeromonas veronii Phage

MLD

A double minimum lethal dose

AEv1

Aeromonas veronii

SM bufer

Sodium chloride magnesium sulfate bufer

PWJ1–01

phage pAEv1

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21505594.2025.2548627