Isolation, characterization, and comparative genomic analysis of vB_BviS-A10Y, a novel bacteriophage from mangrove sediments

Mangrove is among the most carbon-rich biomes on earth, and viruses are believed to play a significant role in modulating local and global carbon cycling. However, few viruses have been isolated from mangrove sediments to date. Here, we report the isolation of a novel Bacillus phage (named phage vB_BviS-A10Y) from mangrove sediments. Phage vB_BviS-A10Y has a hexameric head with a diameter of ~ 79.22 nm and a tail with a length of ~ 548.56 nm, which are typical features of siphophages. vB_BviS-A10Y initiated host lysis at 3.5 h postinfection with a burst size of 25 plaque-forming units (PFU)/cell. The genome of phage vB_BviS-A10Y is 162,435 bp long with 225 predicted genes, and the GC content is 34.03%. A comparison of the whole genome sequence of phage vB_BviS-A10Y with those of other phages from the NCBI viral genome database showed that phage vB_BviS-A10Y has the highest similarity (73.7% identity with 33% coverage) to Bacillus phage PBC2. Interestingly, abundant auxiliary metabolic genes (AMGs) were identified in the vB_BviS-A10Y genome. The presence of a β-1,3-glucosyltransferase gene in the phage genome supported our previous hypothesis that mangrove viruses may manipulate carbon cycling directly through their encoded carbohydrate-active enzyme (CAZyme) genes. Therefore, our study will contribute to a better understanding of the diversity and potential roles of viruses in mangrove ecosystems.

However, our knowledge regarding the role of viruses in mangrove ecosystems is considerably lacking [10], and few viruses have been isolated from mangrove and characterized [11]. Our previous virome study showed that viruses in mangrove sediment encode abundant auxiliary carbohydrate metabolism genes, most of which encode carbohydrate-active enzymes (CAZymes) with glycoside hydrolase activity, such as alpha-amylase and chitinase. These genes are critical for degrading complex carbohydrates in the mangrove sediment, suggesting that viruses manipulate carbon cycling directly by participating in the biomass recycling of complex polysaccharides [10]. However, all of these inferences come from bioinformatic analysis and need to be confirmed experimentally in a cultivated virus-host system. In this context, the isolation and characterization of virus (phage) from mangrove sediment can improve our understanding of virus diversity and virus-host interactions in mangrove and can help to elucidate the role of viruses in carbon cycling in mangrove ecosystems.

Morphology and life cycle of vB_BviS-A10Y
Phage plaques were found after suspended mangrove sediments (Qingmeigang, Sanya, China) were spread as a monolayer onto a 2216E agar plate and incubated at 25°C for three days. We isolated a colony with a plaque and purified the phage particles by polyethylene glycol precipitation (see Supplementary Materials for detailed experimental methods). The 16S rRNA sequence of the host showed 100% identity to that of Bacillus vietnamensis FJAT-46955, and the phage was designated as Bacillus phage vB_Bvi-A10Y. Transmission electron microscopy (TEM) observation showed that vB_BviS-A10Y has a long, flexible, noncontractile tail with an approximate length of 548.56 ± 2.32 nm and a hexameric head with an approximate diameter of 79.22 ± 0.95 nm ( Supplementary Fig. S1), which is the typical siphophage morphology.
To explore the life cycle of phage vB_BviS-A10Y, a one-step growth analysis was performed ( Supplementary  Fig. S2). We used purified virions to infect logarithmically growing host cells (OD 600 = 0.2) at a multiplicity of infection (MOI) of 0.01. vB_BviS-A10Y initiated host lysis at about 3.5 h postinfection, and the titer of the phage reached a plateau at 8.5 h postinfection. vB_BviS-A10Y exhibited a long latent and lytic phase with a relatively small burst size of about 25 PFU/cell ( Supplementary Fig. S2). Long latent and lytic phases have also been observed for phage SPβ, a prophage of B. subtilis 168, which has a latent phase of more than 90 min. The long latent phase of phage SPβ may be attributed to its complex regulation of lysis-lysogeny decisions upon infecting the host [12]. A recombinase gene potentially involved in control of lysogeny was identified in the genome of phage vB_BviS-A10Y (see below), suggesting that vB_BviS-A10Y is a temperate phage, and this would account for the long latent and lytic phases observed.

Genomic analysis of vB_BviS-A10Y
We assembled the vB_BviS-A10Y genome sequence from Illumina high-throughput sequencing data. The genome is a linear double-stranded DNA with a length of 162,435 bp and a GC content of 33.49% (Table 1). A total of 225 putative genes encoding proteins longer than 50 amino acids were identified in the vB_BviS-A10Y genome, using MetaGen-eMark (v3.38) software. Of these, 43 genes were annotated using the UniProtKB/Swiss-Prot ViralZone non-redundant protein database, accounting for 19.11% of the genes in the genome. This is consistent with previous observations that most phage genes are not functionally annotated [13]. The coding regions within the vB_BviS-A10Y genome are tightly packed, and approximately 88.48% of the genome consists of coding regions. The average gene density was calculated to be 1.385 genes per kb, which was similar to that of most phage genomes. As shown in the phage genome map, we observed two main regions of the phage genome that are transcribed in opposite directions ( Fig. 1).
As shown in Fig. 1, 60.46% (26/43) of the annotated genes are associated with conventional viral functions, including DNA replication, structure and morphogenesis, lysis, regulation of transcription, and control of lysogeny. Abundant genes involved in DNA recombination, replication, and repair were identified, and most of them are located in the left arm of the genome. These genes include RecD-like DNA helicase YrrC (gp2), single-stranded-DNA-specific exonuclease RecJ (gp84), HNH endonuclease (gp132), DNA topoisomerase IV (gp133 and gp134), T4 RNA ligase (gp147), endonuclease (gp149), DNA ligase (gp153), RNA ligase (gp156 and gp157), DNA polymerase III alpha subunit (gp173), and DNA binding protein HU2 (gp226). Only three genes involved in structure and morphogenesis were identified, and these were associated with tail morphogenesis, including tail fiber protein (gp185 and187) and tape measure protein (gp190). Genes for two important proteins involved in host cell lysis were identified in the phage vB_BviS-A10Y genome, namely, holin (gp174) and endolysin (gp182). Additionally, we identified a gene involved in control of lysogeny, gp199 (recombinase), suggesting that vB_BviS-A10Y may enter a lysogenic cycle under certain conditions. Although phages are highly dependent on the host's translation machinery, tRNA genes are frequently identified in their genomes [14]. The presence of tRNA genes in the phage genome can compensate for differences in codon and/ or amino acid usage between the phage and host, thereby minimizing host dependence and expanding the host repertoire [15,16]. Using tRNAscan-SE 2.0 [17], 13 tRNA genes were identified in the genome of vB_BviS-A10Y: tRNA-Asn-GTT, tRNA-Gly-GCC, tRNA-Glu-TTC, tRNA-Pro-TGG, tRNA-Met-CAT, tRNA-Cys-GCA, tRNA-Leu-TAG, tRNA-Thr-TGT, tRNA-Ser-TGA, tRNA-Ser-GCT, tRNA-Gly-TCC, tRNA-Arg-ACG, and tRNA-Arg-TCT. The tRNA genes in phages are always clustered, as is the case in the vB_BviS-A10Y genome. The tRNA genes may contribute to the translational efficiency of unique genes, such gp53 was predicted to encode nicotinamide phosphoribosyltransferase, which was previously proposed to play a role in nucleotide scavenging in virus-infected cells [22]. gp140 was predicted to encode deoxyuridine 5'-triphosphate nucleotidohydrolase, which may prevent the incorporation of uracil into DNA, thereby assuring an appropriate supply of deoxyribonucleotides to achieve a high rate of DNA synthesis [23]. The large number of AMG and tRNA genes in the vB_BviS-A10Y genome may improve the efficiency of viral replication and broaden the role that the phage plays in the host's fitness during infection.

Similarity to other known bacteriophages
A comparison of the whole genome sequence of phage vB_BviS-A10Y with those of other phages in the NCBI viral genome database showed that phage vB_BviS-A10Y had the highest similarity (73.7% identity with 33% coverage) to Bacillus phage PBC2 (KT070867.1), a siphophage infecting Bacillus cereus. A whole-genome-based phylogenetic analysis showed that vB_BviS-A10Y clustered with Bacillus phage vB_BcoS-136 (MH884508.1) (Supplementary Fig. S3). Bacillus phage vB_BcoS-136 is a haloalkaliphilic phage infecting Bacillus cohnii isolated from Lake Elmenteita [24]. To compare the genome sequences of vB_BviS-A10Y, Bacillus phage PBC2, and Bacillus phage vB_BcoS-136, a multi-genome alignment was made using Mauve software (Fig. 2) and EasyFig software (Supplementary Fig. S4). As shown in Fig. 2, some subregions of the vB_BviS-A10Y genome have a high degree of sequence similarity to Bacillus phage PBC2 and Bacillus phage vB_ BcoS-136. Moreover, the gene order of vB_BviS-A10Y is similar to that of Bacillus phage PBC2. Seventeen AMGs were identified in the vB_BviS-A10Y genome ( Table 2), only seven of which are also present in the PBC2 and vB_ BcoS-136 genomes, namely, nicotinamide phosphoribosyltransferase, guanylate kinase, deoxyuridine 5'-triphosphate nucleotidohydrolase, polynucleotide kinase, dihydrofolate reductase, thymidine kinase, and aspartate phosphatase. The gene encoding nicotinamide phosphoribosyltransferase (colored in blue in Fig. 2) was identified in all three of these genomes but located at different positions.

Conclusion
Phage vB_BviS-A10Y is a novel siphophage that infects B. vietnamensis isolated from mangrove sediments. Our study provides the first genome sequence of a phage from mangrove sediment. These data will help to reveal the viral diversity and virus-host interactions in mangrove as AMGs [18]. We identified many AMGs in the vB_BviS-A10Y genome, as summarized in Table 2.
gp49 was predicted to encode a β-1,3-glucosyltransferase that was identified as a CAZyme of the GT2 family according to the CAZy database classification. CAZymes were identified previously in viruses in bovine rumen [19], activated sludge systems of wastewater treatment plants [20], thawing permafrost peatlands [21], and mangrove sediments [10]. The presence of a CAZyme gene in the vB_BviS-A10Y genome, the first genome of a phage from mangrove sediment ever sequenced, supports our previous hypothesis that CAZymes may be widespread in viruses in mangrove sediment and may play an important role in carbon cycling in mangrove ecosystems [10]. Further studies should be carried out to examine the enzymatic activity of the gp49 protein and its role in virus-host metabolic interactions. Dihydrofolate reductase (gp168) is involved in folate biosynthesis and was identified previously in bovine rumen viruses [19]. Dihydrofolate reductase, together with other AMGs, may redirect carbon to the one-carbon pool by folate to boost viral replication [19]. A large proportion of AMGs encoded by vB_BviS-A10Y are involved in nucleotide metabolism, including nicotinamide phosphoribosyltransferase (gp53), guanylate kinase (gp128), ribonucleotide reductase (glutaredoxin-like protein nrdH, gp136, gp137, gp138, and gp139), deoxyuridine 5'-triphosphate nucleotidohydrolase (gp140), polynucleotide kinase (gp150), thymidylate synthase (TS, gp166), and thymidine kinase (TK, gp171). These AMGs are among the most common AMGs that are essential for viral DNA replication [20]. For example, ecosystems. Interestingly, abundant AMGs were identified in the phage genome, including a GT2-class CAZyme gene, thereby suggesting that viruses may play important roles in carbon cycling in a mangrove ecosystem.

Statements and declarations
Conflict of interest The authors declare that they have no competing interests.  Each of the three genomes is displayed horizontally with the homolo-gous regions outlined in the same color and connected by lines. The level of sequence similarity between homologous regions is indicated by a similarity plot within the colored blocks, where the height of the plot is proportional to the average nucleotide sequence similarity.