2.5. Genome characterization
Phage genome extraction and next-generation sequencing
The phage particles prepared by liquid propagation in TSBM were desalted using Millipore Amicon ultracentrifugal filter 10,000 NMWL (Merck, United States) at 10,000 x g , 4 oC for 15 min and concentrated by ultracentrifugation at 300,000 x g , 4oC for 3 h (Beckman Coulter, German). The pellets were resuspended in SM buffer. Phage genomic DNA was extracted using Phage DNA Extraction Kit (Cat. 46800, Norgen Biotek, Canada) following the manufacturer’s protocol. Quality and concentration of DNA were measured by Nanodrop (Colibri, German) and Quibit 4.0 (Thermo Scientific, United States). Purified genomic DNA (3.15 ng/µL) was subjected to library preparation and sequencing using Next Generation Sequencing System with Illumina Novaseq 6000 platform (Pair-end, 150; library construction size, 350 bp; data output, 1.0 GB, data quality, Q30 > 80) at KTEST company, Vietnam.
Phage genome assembly and annotation
Raw reads were filtered using Fastp v 0.20.1 with the qualified phred score ≥ Q25 and 8 bases trimming from 5’/3’ end (Chen et al., 2018). Host associated sequences were filtered out by mapping trimmed reads to the genome of A. hydrophila type strain (accession no. NZ_CP016990.1) using Bowtie2 v 2.3.4.3 (Langmead & Salzberg, 2012). Only unaligned reads were subjected to genome assembly using Unicycler v 0.4.8 (Wick et al., 2017) on the Galaxy web platform at usegalaxy.org (Afgan et al., 2016). Potential phage sequence was identified by submitting the assembled contigs to PHASTER web server (Arndt et al., 2016). The predicted phage sequence (assigned as ‘pAh6.2TG’ in this study) was annotated using Prokka v 1.14.6 with Viruses annotation mode (Seemann, 2014). The annotated phage genome was visualized using DNAplotter (Carver et al., 2009).
Phage taxonomic identification and phylogenetic reconstruction
Identification of phage species was carried using VICTOR web service (Meier-Kolthoff & Göker, 2017). VICTOR is a tool that perform pairwise genome comparison of prokaryotic viruses and automatically constructs phylogenomic trees using Genome-BLAST Distance Phylogeny method (GBDP) with the formula D0. This tool also classifies the virus at the species, genus and family level with the taxon boundaries estimating by OPTSIL program (Göker et al., 2009). Herein, only the genomes of the viruses belonging to the family Myoviridae (n = 91) were included in this genome comparison since pAh6.2TG was predicted as an unknownMyoviridae by PHASTER tool described in the above section.
In addition to genome comparison, the phylogenetic analyses based on the terminase large subunit (terL) and major capsid protein (MCP) amino acid sequences of pAh6.2TG and other related species (predicted by VICTOR) were also performed via PhyloSuite v1.2.2 (Zhang et al., 2020). Amino acid sequences were aligned using MAFFT (Katoh & Standley, 2013) and the maximum-likelihood trees were constructed using IQ-TREE (Nguyen et al., 2015) with 5,000 ultrafast bootstraps and best-fit model (LG+G4) estimated by ModelFinder. Phylogenomic tree, terL- and MCP-based trees were visualized using Phandango (Hadfield et al., 2018) and iTOL web tools (Letunic & Bork, 2019). Lastly, the protein sequence similarities between pAh6.2TG and the closest viral taxa were determined using CoreGenes3.5 web server with Blastp threshold score at 75 (Turner et al., 2013).
Effect of phage on Nile tilapia challenged with MDR A. hydrophila
Experimental fish
Healthy Nile tilapia (10.5 ± 4.7 g) obtained from a commercial tilapia hatchery in Thailand were acclimated for 2 weeks in dechlorinated tap water with aeration at 28 ± 1.0 oC before the experiments. The fish were fed with commercial tilapia feed (crude-protein 30%) at rate of about 3% of fish weight twice daily. Before starting the experiments, ten fish were randomly selected for bacterial isolation and found to be free of A. hydrophila . The experimental animal protocols were approved by Chulalongkorn University (Approval no. CU-IACUC 2031006).
Fish survivability and sample collection
This experiment aimed to investigate whether lytic phage treatment improves survivability of Nile tilapia challenged with a pathogenic MDRA. hydrophila BT14. A total of 258 fish were randomly divided into six groups with 2 replicate tanks per each group (Figure S1): Group 1 was exposed to culture medium without phage (no Ah + no phage); Group 2 was exposed to bacteria without phage (Ah + no phage); Group 3 was exposed to culture medium and phage pAh6.2TG at multiplicity of infection (MOI) = 0.1 (no Ah + phage 0.1); Group 4 was exposed to culture medium and treated with phage at MOI = 1.0 (no Ah + phage 1.0); Group 5 was challenged with A. hydrophila and treated with phage at MOI = 0.1 (Ah + phage 0.1); Group 6 was challenged with A. hydrophila and treated with phage at MOI = 1.0 (Ah + phage 1.0).
In bacterial challenge groups (2, 5 and 6), 1 L of MDR A. hydrophila BT14 (approx. 8 × 108 CFU/mL) was added to 50 L water to reach a final concentration of approx. 2 × 107 CFU/mL. Groups 5 and 6 tanks had 2 and 20 mL of phage pAh6.2TG (approx. 8.5 × 1010 PFU/mL) added to reach a final concentration of approx. 2 × 106 and 2 × 107 PFU/mL, respectively. Group 2 tank had 20 mL of SM buffer without phage added. The mixtures in groups 2, 5 and 6 were maintained at 29 ± 1.0 °C with aeration for 3 h. In culture medium exposure groups (1, 3 and 4), 1 L of TSB was added to 50 L water. Groups 3 and 4 tanks had 2 and 20 mL of phage pAh6.2TG (approx. 8.5 × 1010 PFU/mL) added, respectively. After 3 h, the fish were transferred to all groups, maintained at 29 ± 1°C with aeration for 14 days. In order to investigate the effect of phage on the concentration of A. hydrophila in rearing water, a volume of 25 mL water from groups 2, 5 and 6 were sampled at 3, 24 and 48 h after exposure with phage. A volume of 1 mL water was centrifuged at 4 oC, 10.000 x g , for 5 min. The supernatant were collected and diluted in SM buffer to measure concentration of phage by Plaque Assay method (Jun et al., 2013). The pellet was washed 1 time and re-suspended in 1 mL of PBS buffer. Bacterial concentration was then enumerated by conventional plate count method using RS supplemented with Novobiocin (Harrigan & McCance, 2014). In order to investigate the effect of phage on the concentration of A. hydrophila in liver, two fish from groups 2, 5 and 6 were sampled at 24, 48 and 72 h after exposure with phage. The fish were necropsied, and 0.1 g of live tissue was collected and homogenized in a microtube containing 900 µL of SM buffer. The samples were then centrifuged at 10.000 x g , for 5 min. The supernatant and pellet were used for respective phage and bacterial enumeration same as above.
The remaining fish were observed daily for 14 days, and mortality was recorded. Representative moribund or freshly dead fish were collected for bacterial re-isolation using RS supplemented with Novobiocin as described above. The RPS was calculated according to the formula described by Ellis (1988): RPS = (1 - % mortality in challenge / % mortality in control) * 100.
Determination of serum antibody by the enzyme-linked immunosorbent assay (ELISA)
For the comparison of specific antibody (IgM) levels against A. hydrophila between experimental groups, blood samples of 5 surviving fish in each tank (10 fish/group) were collected at the end of the experiment (day 14). Sera were collected after centrifugation at 8,000 x g for 15 min, stored at -20 oC until used. ELISA assay was carried out following the protocol described by Dien et al. (2021).
Statistical Analysis
Percent survival data from the challenge experiments was analyzed by the Kaplan-Meier method and differences among groups were tested using a log-rank test, p -values of 0.05 or less were considered to be statistically significant. Enumeration of A. hydrophilaconcentration and phage titer in rearing water and fish liver samples was analyzed by ANOVA. Dunnett post-hoc test was used to measure specific differences between pairs of mean. The OD450nmreadings from the indirect ELISA assay were analyzed using a Kruskal-Wallis test. Multiple comparison analyses were performed by Bonferroni test. All statistical analyses were performed using SPSS Software ver22.0 (IBM Corp., USA).
RESULTS
Prophage induction, phage isolation, and morphology
Although three doses of Mitomycin C (250, 500, and 1,000 ng/mL) were used for prophage induction, no plaque was detected, indicating thatA. hydrophila BT09 did not contain prophage and was suitable as a bacterial host for lytic phage isolation. Subsequently, a phage, designated pAh6.2TG, was isolated from a freshwater sample. Phage pAh6.2TG produced medium, clear, and round plaques with diameter of 1.3 ­ 1.8 mm (Figure 1A-B) after 16 h of incubation. TEM morphology examination showed that the phage had an icosahedral head with 59.6 ± 2.5 nm diameter (n = 3) and a contractile tail which was 137 ± 10.2 nm in length and 20.2 ± 2.7 nm in diameter (n = 3) (Figure 1C-D). Based on the morphological features, phage pAh6.2TG was initially classified to the Myoviridae family.
Host range and specificity of phage pAh6.2TG
Among all bacterial isolates tested, pAh6.2TG showed lytic activity against 10/17 A. hydrophila isolates (Table 1) of which 8 isolates were MDR (Table S1). In contrast, no lytic activity was observed against other fish bacterial pathogens including A. veronii , A. schubertii, E. ictaluri , S. agalactiae as well as two probiotic bacteria L. fermentum , and L. plantarum (Table 1).
Stability of phage pAh6.2TG at different environmental conditions
Stability of pAh6.2TG at different temperatures (4 to 40oC) is shown in Figure 2A. Similar percentages of viable phage were detected after 1 h (96 ± 0.55 – 99.6 ± 0.08%) and 24 h (93 ± 0.23 – 98.6 ± 0.17%) of incubation, indicating that pAh6.2TG is a relatively thermostable phage.
Phage pAh6.2TG was stable (93.5 ± 1.69 – 97 ± 0.87%) at pH 7, 9 and 11 (Figure 2B). However, the phage pAh6.2TG was not stable at low pH. At pH 5, 93 ± 0.24% phage remained viable after 1h, and decreased sharply to 32.7 ± 0.44% (from 7.88 to 2.58 ± 0.06 logPFU/mL) after 24 h. At pH 3, only 15.2 ± 1.47% (1.19 ± 0.2 logPFU/mL) of phage was still viable after 1 h and reduced to undetectable level at 24 h (Figure 2B).
Phage pAh6.2TG was relatively stable at a wide range of salinity (0 – 40 ‰), with 95.9 ± 1.35 – 99 ± 0.19 % and 94.2 ± 0.29 – 99 ± 0.32 % viable after 1 and 24 h, respectively (Figure 2C).
In fish-rearing water (30 ± 1 oC, pH 6.9, 0% NaCl) spiked with phages, percentage of stability at 1 and 3 h were 99.5 ± 0.15% and 98.6 ± 0.11%, respectively. After 24 and 48 h, phage titer decreased slightly to 91.9 ± 0.85% and 91.3 ± 0.5%, equivalent to 6.52 ± 0.07 and 6.47 ± 0.03 logPFU/mL, respectively.
Genome characterization of pAh6.2TG phage
Based on the assembly graph generated by Unicycler software, pAh6.2TG was predicted to contain a circular genome with a length of 51,780 bp, a GC content of 52.48%, encoding 65 putative open reading frames (ORFs) (Table S2) without tRNA genes (Figure 3). According to bioinformatics prediction, pAh6.2TG genome consists of three main functional modules: i) phage structure and DNA packaging (major capsid protein, baseplate protein, tail fiber protein, and terminase subunit), ii) DNA metabolism and replication (RNA polymerase, DNA polymerase, DNA helicase, 5’-3’ exonuclease, DNA ligase, and Ribonucleoside-diphosphate reductase large subunit), and iii) host lysis (cell wall hydrolase).
The closest phage taxonomic classification of pAh6.2TG toward other 91Myoviridae phages in the public database revealed thatAeromonas phage pAh6.2TG and PVN02 (accession no. LR813619) were classified as the identical species with 99.33% identity. The result also showed total 64/65 ORFs were homologous between pAh6.2TG and PVN02 (97.3 - 100 % nt. identity), except for ORF03 that showed the highest homology (70%) to another Aeromonas phage pAh6-C (Table S2). Phylogenetic analysis based on whole genome (Figure 4A-B), major capsid protein sequence (Figure 4C), and terminase large subunit sequence (Figure 4D) confirmed high homology of phage pAh6.2TG and phage PVN02. In addition, pAh6.2TG was closely related to the Aeromonas phage pAh6-C (accession no. KJ858521), Shewanella phage Spp001 (accession no. NC023594), and Shewanella phage SppYZU05 (accession no. NC047824) (Figure 4).
In vivo challenge results