Characterization and genomic analysis of ValSw3-3, a new Siphoviridae bacteriophage infecting Vibrio alginolyticus.

Phage therapy has been considered a potential alternative to antibiotic therapy in treating bacterial infections. For controlling the vibriosis-causing pathogen Vibrio alginolyticus, well-documented phage candidates are still lacking. Here, we characterize a novel lytic Vibrio phage, ValSw3-3, based on its morphology, host range and infectivity, growth characteristics, stability under various conditions, and genomic features. Our results show that ValSw3-3 could be a potent candidate for phage therapy to treat V. alginolyticus infections due to its stronger infectivity and better pH and thermal stability than those of previously reported Vibrio phages. Moreover, genome sequence alignments, phylogenetic analysis, in silico proteomic comparison, and core gene analysis all support that this novel phage, ValSw3-3, and five unclassified phages form a clade distant from those of other known genera ratified by the ICTV. Thus, we propose a new viral genus within the Siphoviridae family to accommodate this clade, with ValSw3-3 as a representative member.

fied by the ICTV. Thus, we propose a new viral genus within the Siphoviridae family to accommodate this clade, with ValSw3-3 as a representative member. KEYWORDS Siphoviridae, Vibrio alginolyticus, bacteriophage, genomic analysis V ibrio alginolyticus is one of the most important opportunistic pathogens that can infect a variety of aquatic animals, and it has been implicated in several mass mortality cases in major aquaculture species, from fish to mollusks and crustaceans (1)(2)(3). This bacterium is also pathogenic to humans, causing skin and ear infections and acute gastroenteritis (4). Moreover, multidrug-resistant Vibrio strains are quickly emerging because of the overuse of antibiotics, and alternative approaches must be developed to control these pathogens. In this context, bacteriophages (or phages) have been considered potential therapeutic agents for use in the treatment of pathogenic Vibrio infections in aquaculture (5,6). Due to their high specificity for host bacteria and ability to proliferate, phages could be a potent and safe biocontrol agent.
For applications in biocontrol, a full characterization of phage genomes is required to rule out the presence of potentially harmful genes related to virulence or antibiotic resistance and to confirm their lytic or temperate nature (7). To date, approximately 300 phages infecting Vibrio hosts have been isolated, completely sequenced, and deposited in the National Center for Biotechnology Information (NCBI) genome database (as of 13 June 2019). Most of them are phages infecting V. cholerae, V. parahaemolyticus, and V. vulnificus. Until recently, the complete genomes of only a few V. alginolyticus phages had been reported, including those of four myoviruses (VP01 [1], PVA1 [8], phi-A318 [9], and pVa-21 [10]) and five podoviruses (ValKK3 [11], VEN [7], ⌽A318 [9], ⌽as51 [9], and IME 271 [12]). Lytic phage pVa-21 was isolated from seawater and showed potential as a biocontrol agent against V. alginolyticus because of its anti-planktonic and antibiofilm activities (10). In addition, V. alginolyticus phage VP01 exhibited therapeutic potential by effectively suppressing the growth of V. alginolyticus (1). Both phages belong to the Myoviridae family. On the other hand, the podovirus VEN can exclusively infect its original host, V. alginolyticus strain V2, but is unable to lyse other tested V. alginolyticus strains (1). Interestingly, no siphoviruses isolated using V. alginolyticus as the host strain have been reported with their complete genomic sequences in the literature.
In this study, we report the biological and genomic characterization of a novel lytic siphovirus, named ValSw3-3. This phage was isolated using a pathogenic V. alginolyticus strain from diseased shrimps as the host. To explore the novelty of this phage, phylogenetic analysis based on the major capsid protein and the terminase large subunit, genomic, comparative in silico proteomic analyses, and core gene analysis of ValSw3-3 and other related phages were carried out.

RESULTS AND DISCUSSION
Biological features of ValSw3-3. By propagating ValSw3-3 on its host strain, Va-F4, plaques were formed on the bacterial lawn with a clear round morphology and a welldefined boundary (Fig. 1A, inset). The plaque size was approximately 2.24 Ϯ 0.14 mm in diameter after 12 h of incubation at 30°C. The morphology of the phage particles observed using transmission electron microscopy (TEM) showed that ValSw3-3 had an icosahedral head (67 Ϯ 4.0 nm) and a noncontractile sheathed tail (168 Ϯ 2.2 nm) (Fig.  1A), indicating that this phage belongs to the family Siphoviridae within the Caudovirales order.
The one-step growth curve of ValSw3-3 propagated on Va-F4 was plotted and is shown in Fig. 1B. The latent period of phage ValSw3-3 was approximately 15 min. The burst size was 95 Ϯ 2 PFU/cell (burst size is the total number of phages liberated at the end of one cycle of growth/number of infected bacteria). Although the burst size was not significantly higher, the latent period of ValSw3-3 was shorter than those of most previously reported Vibrio phages (1,13,14).
Spot tests showed that phage ValSw3-3 could infect multiple Vibrio species. Among the 34 strains (aside from Va-F4) tested in this study, phage ValSw3-3 was able to infect strains, including V. alginolyticus (Va-F10, Va-F2, and Va-F3), Vibrio species (Vs-F1, Vs-6, and Vs-9), V. parahaemolyticus (Vp-F7), and V. metschnikovii (Vm-4) ( Table 1). As expected, phage ValSw3-3 did not show infectivity in bacterial strains that do not belong to the Vibrio genus. The ability of ValSw3-3 to lyse multiple Vibrio strains highlighted that this phage could be explored for phage therapies against vibriosis caused by different Vibrio strains. Since ValSw3-3 has a broad host range spanning several species of Vibrio, determining the genes encoding tail fibers and other host specificity-related proteins in ValSw3-3 would be valuable.
In vitro lytic effect. The in vitro lytic effect of ValSw3-3 on its host bacteria was tested by infecting fresh V. alginolyticus cultures at the exponential phase (optical density at 600 nm [OD 600 ] of ϳ0.4). As shown in Fig. 2A, the killing curve indicated that ValSw3-3 was highly effective against the host at all MOIs tested in this study, and almost no bacterial growth was observed for the first 12 h after infection. The intense bactericidal activity indicates that ValSw3-3 is a potent candidate for use in in vivo phage therapy trials against V. alginolyticus. Generally, the regrowth of the bacterial population following phage infections implies the emergence of phage resistance (15). During our experiment, the regrowth of the bacterial population occurred slowly after infection for 13 h, but the OD 600 of the culture was below 0.35 even 30 h after infection (Fig. 2B).
Moreover, the killing curves, observed at different MOIs ( Fig. 2A), showed no observable difference in inhibitory effects after 5 h of infection, and even the lowest MOI resulted in suppressed host growth throughout the 30 h of the experiment (Fig. 2B). These results suggested that further increasing phage loads would lead to a limited positive effect on the infection of phages with strong lytic ability. Therefore, using heavy phage loads to restrain the emergence of phage resistance in bacteria is not an ideal strategy. Instead, further exploration for cocktails composed of ValSw3-3 and other phages could be a more practical approach to strengthen the antibacterial activity.
Genomic characteristics of ValSw3-3. The genome sequence of ValSw3-3 was obtained using an Illumina HiSeq 1500 sequencer platform; approximately 20,000 reads were obtained in total, with an average length of 150 bp. After de novo assembly based Ϫ Shrimp a a Strains isolated from the gut of sick shrimp in aquaculture (Shanwei, Guangdong Province, China). b Strains isolated from water in the Yangzi river (Wuhan, Hubei Province, China). c Strains isolated from the gut of sick shrimp in aquaculture (Shenzhen, Guangdong Province, China). d A total of 35 strains were tested during the host range analysis. The ability of ValSw3-3 to infect different Vibrio strains was evaluated on the basis of the formation of a lysis-cleared zone at a spot on a bacterial lawn classified into two categories, lysis (ϩ) and no lysis (Ϫ). e The host strain used for ValSw3-3 isolation. on overlap of Ն40 nucleotides with a minimum overlap identity of 90%, the whole genome was aligned as a single contig.
A linear 39,846-bp double-stranded DNA genome with a GC content of 43.1%, similar to that of its host (V. alginolyticus, 44%), was obtained. It has 97 promoter regions and 44 predicted transcription terminators ( Table 2). Among the eleven reported Vibrio phages belonging to siphovirus (NCBI database as of June 2019), the genome size of phage ValSw3-3 is 2,524 bp larger than the 37,324-bp genome of phage pYD38-B, while its GC content is equal to that of pYD38-B but not other Vibrio siphoviruses (Table 2). To identify the origin site of replication and the terminus point of the phage genome, cumulative GC skew was generated as described by Mackiewicz et al. (16). The results indicated that ValSw3-3 had a putative replication origin site at bp 11428 and a putative terminus location at bp 39742 (Fig. 3A), and the GC skew was consistent with the transcriptional directions of most of the open reading frames (ORFs). No tRNA genes were found in the genome of ValSw3-3, indicating that ValSw3-3 depends on the translation machinery of the host ( Table 2).
Analysis of the phage ValSw3-3 genome revealed 69 putative ORFs. The pVOG database was scanned for homologues of the proteins encoded by the predicted ORFs using BLASTP (E values of less than 10 Ϫ4 ). Since the genome sequence of ValSw3-3 significantly differed from those of other available phage genomes in NCBI (query coverage, 42%; identity, 88%), only 16% of the ORFs were related to proteins with known or predicted functions, and another 39% had hits of unknown functions. Thirty-eight ORFs showed different levels of similarity (28% to 96%) to their counterparts from 17 species (see Table S1 in the supplemental material) in the BLASTP analysis. Putative functions were assigned to 12 ORFs based on similarities in the amino acid sequences. Despite the low number of annotated ORFs, some representative  modules, such as those encoding proteins implicated in DNA replication (helicase), DNA metabolism (endodeoxyribonuclease, RecT recombination, exonuclease, and NADasparagine ribosyltransferase), DNA packaging (terminase large/small subunit), structural protein (capsid), and host lysis (lysozyme and hemagglutinin), were recognized in the ValSw3-3 genome (Fig. 3B). These functional modules constitute approximately 40% of the whole-genome length. The remaining ORFs showed less than 28% or no similarity to other proteins in the NCBI database. Pairwise comparison of the genomes of ValSw3-3 and its most closely related phages, phiHSIC (GenBank accession no. NC_006953.1) and P23 (MK097141.1) (17), was carried out. Although most of the ValSw3-3 ORFs (including hypothetical proteins) have no homologs with Listonella phage phiHSIC, the terminase large subunit and the major capsid proteins of these two phages share identities of over 85%. Moreover, there were no gene clusters related to lysogeny (e.g., Cro, CI, C2, C3, N, and Q) (18) in the genome, indicating that ValSw3-3 should be considered an obligate lytic bacteriophage that met the prerequisite for phage therapy candidates. Additionally, no antibiotic resistance genes or virulence factor-related genes were detected in the genome of ValSw3-3 using the ARDB (19) and VFDB (20) databases. Multiple-genome comparisons, phylogenetic analysis, and comparative genomics. ValSw3-3 is a siphovirus isolated using V. alginolyticus as the host. To assess the exact taxonomic position of the phage, phylogenetic analysis was performed based on the amino acid sequences of the terminase large subunit (21) and major capsid protein (22)  (Fig. 3C). These six phages formed a branch clearly distinct from other branches corresponding to known genera and ratified by the ICTV. However, according to the phylogenetic tree based on the major capsid protein, ValSw3-3 was only clustered with Listonella phage phiHSIC (NC_006953.1) (23,24) and Vibrio phage jenny 12G5 (HQ632860.1) (Fig. 3D). To confirm this finding, we carried out comparative analysis of the whole-genome sequences of the same phages used in the phylogenetic analyses. Similar to the results of the major capsid protein analysis, ValSw3-3 grouped with the five phages and shared lower similarity with other phages (Fig. 4). Among the six ratified genera represented in the phylogenetic tree, the branch containing ValSw3-3 was most closely related to the genus Brussowvirus. To further clarify the relationship between ValSw3-3 and Sfi11virus phages, we performed a phylogenetic analysis (Fig. 5A) and pairwise genomic comparisons (Fig. 5B) between ValSw3-3 and other species within the Brussowvirus genus. The results indicated that ValSw3-3 had a distant evolutionary relationship with Sfi11virus phages, confirming the novelty of ValSw3-3. All five phages clustered with ValSw3-3 in the phylogenetic analysis are siphoviruses, but none of them have been classified at the genus level to date. Therefore, to facilitate a more detailed classification, we conducted an in silico pairwise comparison of the proteomes of ValSw3-3 and the five phages, as well as 18 other distantly related phages that were well classified and ratified by the ICTV. This analysis revealed five similarity groups (Fig. 6A). Typically, phages within a genus are defined as sharing more than 35% of their genes (25,26). The ratios of shared genes between ValSw3-3 and Vibrio phage P23 (GenBank accession no. MK097141.1), Vibrio phage pYD8-B (NC_021561.1), and Vibrio phage 2E1 (KX507045.1) were above the threshold value, namely, 52.2% (36/69), 39.1% (27/69), and 36.2% (25/69). Although ValSw3-3 shared only 34.8% (24/69) and 30.4% (21/69) of its homologous genes (amino acid identity, 69% on average) with phiHSIC (NC_006953.1) and 12G5 (HQ632860.1), respectively, these three phages were grouped together in both phylogenetic trees. Overall, the results of proteome comparisons were consistent with the results of phylogenetic analysis. Complete genome sequence comparisons of the five phages Listonella phage phiHSIC (NC_006953.1), Vibrio phage P23 (MK097141.1), Vibrio phage pYD8-B (NC_021561.1), Vibrio phage 2E1 (KX507045.1), and Vibrio phage 12G5 (HQ632860.1) with ValSw3-3 further confirmed different similarity regions located throughout the phage genome, and the most similar regions were from bp 6500 to 7000, bp 8800 to 10500, and bp 14300 bp to 15800 (Fig. 6B). DNA sequence identity is an essential parameter for identifying new members of the viruses classified by the ICTV, and those with genome sequence identity of less than 50% can be defined as a new genus (27). Given that the highest sequence query coverage was merely 42%, observed when the genome of phage phiHSIC was used as the reference, our analyses allow us to propose a novel genus to accommodate phage ValSw3-3 in the family Siphoviridae (an application has been submitted to the ICTV).
Core gene analysis. Putative ortholog analysis was performed with a set containing all the proteins predicted from the 24 phage (ValSw3-3, P23, 2E1, phiHSIC, pYD8-B, 12G5, and 18 other distantly related phages) genomes to screen for core genes. A total of 2,500 protein sequences from the 24 phage genomes were grouped into 384 clusters. As expected, due to phage diversity, no cluster contained genes from all phage genomes (Fig. 7A). Therefore, we identified the genes in each ortholog group and determined the numbers of shared core clusters among groups. The results showed that the core genes among most clusters were annotated as terminase large subunit, major capsid protein, tail protein, and DNA polymerase. This reflected the conservative nature of the genes encoding these proteins and corroborated the feasibility of phylogenetic analyses of these phages based on these proteins. The results also revealed large numbers of hypothetical proteins shared within each of the clusters, which suggested that very little has been done to functionally characterize phage proteins, even though many of the phages with their genomes sequenced may have great potential in biomedical applications. In total, the ValSw3-3 genome contained 69 clusters, among which 47 were shared with phiHSIC, P23, pYD8-B, 2E1, or 12G5 (Fig. 7B). Seven clusters commonly shared with phiHSIC, P23, pYD8-B, 2E1, and 12G5 were annotated as a portal protein, NAD-asparagine ribosyltransferase head completion adaptor, coil-containing protein, and three hypothetical proteins (Fig. 7C). Interestingly, the two clusters (portal protein and NAD-asparagine ribosyltransferase) of ValSw3-3 were also shared with vB_VpaS_MAR10 (GenBank accession no. JX556418.1), a temperate Vibrio parahaemolyticus phage. Furthermore, a sequence similarity search of the ORF of NAD-asparagine ribosyltransferase in UniProt or the NCBI database revealed that this protein was encoded only by Vibrio phage genomes, except for Listonella phiHSIC. It should be noted that Listonella pelagia is not found to be distinct from the Vibrio genus based on the genomic and phenotypic data, and members of this taxon were formerly named Vibrio pelagia (23,28). As for the other shared clusters in the ValSw3-3 genome, there were ten clusters shared with four out of the five phages and nine clusters with three. Clusters annotated as major capsid protein, tail tape measure protein, head closure protein, and four hypothetical proteins were shared with phiHSIC, P23, pYD8-B, and 2E1, while putative lysozyme was shared with 12G5 but not 2E1. Structural proteins, such as the major capsid protein (29), tail tape measure protein (30), head protein (31), and terminase large subunit (32), are widely used for phage phylogenetic analysis due to their conserved sequences. Notably, the terminase large subunit of ValSw3-3 was not included in the commonly shared clusters by phiHSIC, P23, pYD8-B, 2E1, 12G5, and ValSw3-3 but only grouped with phiHSIC and 12G5. This finding suggested that it is rational to use various strategies when assessing the taxonomy of phages. Lysozymes (endolysins) are highly diverse proteins both enzymatically and structurally and now are one of the most promising families of alternative antimicrobials (33,34).
Conclusions. In summary, this study demonstrates that ValSw3-3, isolated using a V. alginolyticus strain (GenBank accession no. MH298559) as the host, is a novel lytic phage of the family Siphoviridae. As characterized by determining its host range, one-step growth curve, and lytic activity, ValSw3-3 exhibits strong lytic activity and great potential as a novel antibacterial agent for the biological control of vibriosis in aquaculture. Moreover, ValSw3-3 is stable at up to 50°C and pH ranging from 4 to 10, which makes it suitable for potential use in seawater environments. Phylogenetic analyses and core gene analysis suggest that a new viral genus can be proposed to accommodate ValSw3-3. Although a large portion of clusters still are annotated as hypothetical proteins, several relatively conserved genes were identified by the analysis of core genes, including those encoding terminase large subunit, capsid protein, and tail protein. Furthermore, a unique gene encoding NAD-asparagine ribosyltransferase in the ValSw3-3 genome was found to be of specificity in the core gene analysis, and it could serve as a marker gene of Vibrio phages. For further understanding of ValSw3-3, our work will focus on the functional characterizations needed to enable biomedical applications of this phage in various contexts.

MATERIALS AND METHODS
Isolation and identification of bacterial pathogens. Bacterial strains used in this work were isolated from diseased shrimps and water in different aquafarms (Shenzhen, Guangdong province, China; Shanwei, Guangdong province, China; and Wuhan, Hubei province, China). Vibrio strains were isolated using thiosulfate citrate bile salt-sucrose agar medium (TCBS agar; Haibo, China) and further incubated at ambient temperature (28°C) in 2216E medium (5 g peptone and 2 g yeast extract per liter). Taxonomy assignment of these strains was achieved by 16S rRNA gene sequence analysis (35). Briefly, microbial DNA samples were used as templates for the amplification of 16S rRNA gene fragments using the universal primers 27f and 1492r. The reaction mixture consists of 10 ng of template DNA, 2.5 U of DNA polymerase, 5 l of 10ϫ PCR amplification buffer (100 mM Tris-HCl, 500 mM KCl), 200 M deoxynucleoside triphosphates, 1.5 mM MgCl 2 , and 10 pmol of a primer. The mixture described above first was denatured for 1 min at 98°C, followed by 30 PCR cycles of denaturation (10 s at 98°C), annealing (30 s at 56°C), and extension (40 s at 72°C). Finally, another extension was executed at 72°C for 2 min. The PCR products were sequenced by a Sanger dideoxy sequence platform (BGI Gene Co., Ltd., China).
Phage isolation, purification, and host range determination. One of the bacterial isolates, V. alginolyticus strain Va-F4 (GenBank accession no. MH298559), caused high mortality rates (53.1%) of shrimps during the challenge test in our previous study (36). Using this strain as the host bacterium, phage ValSw3-3 was isolated from sewage from an aquaculture farm in Shanwei, China. The isolation step was performed as described previously (37,38).
The host range of ValSw3-3 was determined by spot testing and confirmed by the double-layer agar method. Briefly, the test strains listed in Table 1 were individually cultured to an OD 600 of about 0.5, an aliquot of 350 l of each culture was individually added to 5 ml of prewarmed semisolid liquid medium, and then the ingredients were mixed and poured onto individual plates containing solid medium to form a double-layer plate. Ten microliters of the purified phage solution (10 11 PFU/ml) was dropped onto the overlaid top agar. The plates then were incubated at 30°C and examined for the presence of a lysis zone after 12 h.
One-step growth curve. To determine the one-step growth curves of the phage, a modification of the method of Sasikala and Srinivasan (1) was used. In brief, 6 ml of V. alginolyticus (GenBank accession no. MH298559) cells (OD 600 of 0.5 to 0.6) was centrifuged at 8,000 rpm for 5 min. The cell pellets were resuspended in 300 l of 2216E liquid medium, and then 600 l of phage suspension was added to yield a multiplicity of infection (MOI) of 0.01. After adsorption for 10 min at 30°C, the mixtures were centrifuged at 12,000 rpm for 10 min to remove any unadsorbed phage particles prior to resuspending the cell pellets in 6 ml of 2216E liquid medium. The samples then were shaken and incubated at 30°C, and aliquots were collected at different intervals over a 60-min period. These aliquots were immediately diluted, and phage titers were determined using the double-layer agar method (39). Three independent experiments were performed for each assay.
pH and thermal stability assays. Environmental factors, including pH and temperature, were tested according to the methods described by Kim et al. (10). For pH stability tests, 100 l filter-sterilized phage samples (10 8 PFU/ml) was inoculated in 1 ml SM buffer (100 mM NaCl, 8 mM MgSO 4 , 5 mM Tris-HCl, 0.01% gelatin, pH 7.5), adjusted to pH 2 to 12 with 1 M NaOH and 1 M HCl, and incubated at 30°C for 1 h statically. For thermal stability tests, 1 ml filter-sterilized phage samples (10 8 PFU/ml) was incubated statically under the indicated condition (4°C, 10°C, 20°C, 30°C, 37°C, 42°C, 50°C, and 65°C) for 1 h. After the treatment, aliquots were collected and assayed to determine the number of surviving plaqueforming units.
Growth curve experiment. An in vitro lysis assay for phage ValSw3-3 on V. alginolyticus (GenBank accession no. MH298559) was determined at various MOIs (10 Ϫ6 , 10 Ϫ3 , 10 Ϫ2 , and 10 Ϫ1 ). The measurement was performed in sterile 100-well plates using a growth curve measuring instrument (110001-892; BioScreen C; Growth Curves USA) by monitoring the OD 600 (optical density measurements at a wavelength of 600 nm). Briefly, wells were loaded with 200 l of freshly prepared culture of the host bacteria. The plate was placed in the instrument and incubated at 30°C with orbital shaking. The initial purified phage stock solution was prepared, and the phage titers were adjusted to desired concentrations by serial dilution. Cultures were infected with equal volumes of phage lysate at 4 different MOIs when the bacterial culture reached the exponential phase (OD 600 of ϳ0.4). For each assay, two control samples were included: the bacterium-only control, which was inoculated with V. alginolyticus (MH298559) only, and the phage control, which was inoculated with the phage lysate without bacteria. The growth curves were monitored in real time over 30 h, and OD 600 measurements were recorded every 5 min. For all phage infection assays, the titer of each initial phage inoculum had been determined previously using the double agar layer method. Three independent assays were carried out for each assay.
Phage DNA extraction, genome sequencing, and assembly. The concentrated phage particles were treated using DNase I and RNase A (New England BioLabs) to remove bacterial nucleic acids. The phage genomic DNA then was extracted using a lambda bacteriophage genomic DNA rapid extraction kit (DN22; Aidlab, China) by following the manufacturer's protocol and sequenced using an Illumina HiSeq 1500 sequencer platform (Annoroad Gene Technology Co., Ltd., China). The high-quality filtered reads were assembled using SOAP de novo with default parameters (40). The complete genome sequence then was manually inspected.
Genome analysis, phylogenetic analysis, and core gene analysis. Open reading frames (ORFs) were predicted with GeneMark.hmm (http://exon.gatech.edu/GeneMark/index.html) (41,42). The ORFs were annotated using the BLASTP algorithm with the nonredundant (nr) protein database of the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov), with an E value cutoff of 10 Ϫ4 . The visualization of comparisons of multiple phage genome sequences was performed using BLAST Ring Image Generator (BRIG) with default settings (43). The putative promoter sequences were predicted using the online tools Softberry (http://www.softberry.com/berry.phtml) (44) and Promoter Scan (https://www-bimas.cit.nih.gov/molbio/proscan). The presence of virulence genes was searched with the Virulence Search program (http://www.hpa-bioinfotools.org.uk/pise/virfactfind_small.html). Rho-independent transcription terminators were also predicted using the ARNold program (http://rna .igmors.u-psud.fr/toolbox/arnold). tRNAs carried by the phage genomes were detected using the protein ARAGORN and tRNAscan-SE (http://lowelab.ucsc.edu/cgi-bin/tRNAscan-SE2.cgi) (45). Comparative genomic analysis was conducted using EasyFig 2.1. (46). To determine the taxonomy of the isolated phages, phylogenetic analysis based on the terminase large subunit was carried out using MEGA 5.02 software (http://www.megasoftware.net) with the neighbor-joining method and 1,000 bootstrap replications (47). Pairwise similarity comparisons of each predicted protein were performed using BLASTP and visualized using the R, version 3.4.1, gplots package (48). Furthermore, core gene analysis was carried out with the established pipeline orthoMCL (49) (https://orthomcl.org/orthomcl/), and the resulting network was visualized with Cytoscape software (version 3.6.1; http://cytoscape.org/), using an edge-weighted spring-embedded model, which places the phage genomes that share more protein clusters (which clusters proteins based on similarity) in closer proximity in the display. Moreover, the proteins shared by the target phage (ValSw3-3) with others were also visualized in the Venn diagram constructed with the aid of the online software DrawVenn (http://bioinformatics.psb.ugent.be/webtools/Venn/). Statistical analysis. The SPSS software package, version 12.0, was used for all statistical analyses. Statistically significant differences in all of the experiments were determined using the Kruskal-Wallis test with the h multiple-comparison test at P values of Ͻ0.01.
Data availability. The complete genome sequence of the phage ValSw3-3 is available in GenBank under accession number MG676223. The 16S rRNA gene sequences are also available in the NCBI GenBank database under the accession numbers listed in Table 1.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.3 MB.