Erythrobacter litoralis DSM 8509 was used as host bacteria for phage isolation. In addition to the strain mentioned above, 16 bacterial strains were used for host range analysis (Table 1). All bacterial strains were cultivated in an RO medium (containing 1 g/L yeast extract, 1 g/L tryptone, and 1 g/L sodium acetate, pH 7.5) at 30°C (Yurkov et al., 1999).

Seawater samples were collected by filtering through a 0.22-μm filter membrane (Millipore, USA) from Sanggou bay in Weihai, China, in October 2018. Double-agar layer method was used to examine the presence of phages (Pajunen et al., 2000). Every single plaque was collected, plaque-purified four times, and then cultured in sodium chloride–magnesium sulfate (SM) buffer (100 mM NaCl, 50 mM Tris, 10 mM MgSO₄, and 0.01% gelatin, pH 7.5). Afterward, the isolated phages were stored at −80°C in SM buffer with several drops of chloroform.

The phage suspension was mixed with 1 L mid-log phase bacterial host culture and incubated at 30°C and 180 rpm in a shaker until fragmented and aggregated cells were observed. The phage particles were then harvested and purified following Zheng et al. (2018). Briefly, the phage–host culture medium was treated with DNase and RNase at the final concentration of 2 ng/L for 1 h at room temperature. Then NaCl concentration was adjusted to 1 M, and the lysates were incubated at 4°C for another 0.5 h. After centrifugation (8,000 × g), the supernatants were harvested and filtered through a 0.22-μm filter membrane (Millipore, USA). Then the filtrate was added with PEG 8000 (final concentration 10%, w/v), and placed at 4°C overnight. After centrifugation (10,000 × g, 45 min), the precipitate was resuspended in SM buffer and stored at 4°C overnight. Subsequently, the phages were purified using CsCl (gradient-density: 1.5 g/mL) gradient ultracentrifugation (Optima L-100 XP Ultracentrifugation, Beckman Coulter) at 200,000 × g, 4°C for 24 h. Then, CsCl was removed by dialysis against SM buffer.

Phage morphology was observed by using a transmission electron microscope (TEM). A 20 μl of purified phage suspension was deposited on a copper grid and negatively stained with 2% uranyl acetate for 10 min. Then the grid was loaded onto a 120 kV TEM (JEM-2100HC TEM, JEOL, Japan) for examination.

One-step growth experiments were conducted according to previous studies (Pajunen et al., 2000; Wu et al., 2007). Briefly, after the bacterial strain E. litoralis DSM 8509 was grown to mid-log phase (OD600 nm 0.3–0.5), the isolated phage was added to the bacterial culture at a multiplicity of infection (MOI) of 0.1, and the mixture was kept at 30 °C for 10 min for phage adsorption. After centrifugation for 5 min at 8,000 × g, the supernatant was removed, and cell pellets were resuspended in RO medium and cultured at 30°C. At selected time points (0, 20, 40, 60, 80, 100, 120, 140, 160, 180, 200, 220 min, post-infection), the triplicate samples were subjected to serial dilutions separately, and the eclipse period was determined by double-layer plaque assay method (Pajunen et al., 2000). After incubation at 30°C for 24 h, plaques were enumerated, and titers were expressed as PFU/ml (plaque forming units per mL) for each plate.

The host range of isolated phage was determined by spotting serial dilutions on double agar plates. Briefly, ~10⁷ PFU/ml phage suspension mixed with the potential host was added to each plate. Then the plates were incubated at 30°C for 24 h. The formation of phage plaques suggested the presence of lytic phages. A total of 17 marine bacterial strains were used for host range assessment, including 11 Erythrobacter strains, two Roseobacter strains, two Citromicrobium strains, one Lutibacterium strains, and one Halomonas strain.

The CsCl-purified phage was first treated with DNase I (1 μg/ml) and RNase A (1 μg/ml) for 30 min at 37°C, and then treated with 100 μg/ml proteinase K and 10% SDS for 2 h at 56°C. Afterward, the genomic DNA was extracted twice with phenol:isoamyl alcohol (25:24:1), followed by two washes of chloroform:isoamyl alcohol (24:1). After centrifugation for 10 min at 12,000 rpm, the aqueous phage was precipitated with isopropanol (1:1 v/v) and sodium acetate (10:1 v/v) for 1 h at −20°C. The sample was then centrifuged at 12,000 rpm for 15 min, and the precipitated DNA was subsequently washed two times with 70% and 100% ethanol, dried, and resuspended in water.

The genome sequencing was performed on the Illumina Miseq platform with 2 × 250 bp paired-end reads, and CLC Genome Workbench software (43 × coverage) was used for genome assemblage.

The putative open reading frames (ORFs) were identified using the GeneMarkS online server (http://exon.gatech.edu/Genemark/genemarks.cgi) (Besemer et al., 2001), Glimmer 3.0 (http://ccb.jhu.edu/software/glimmer/index.shtml) (Delcher et al., 2007), and ORF Finder online server (https://www.ncbi.nlm.nih.gov/orffinder/). Genes were BLAST searched against non-redundant (nr) protein database of NCBI for functional annotation (https://blast.ncbi.nlm.nih.gov/Blast.cgi), with a cut-off E-value of 10⁻⁵. The tRNA encoding genes were identified with tRNAscan-SE 2.0 online server (http://lowelab.ucsc.edu/tRNAscan-SE/) (Lowe and Chan, 2016; Chan and Lowe, 2019).

To determine the viral taxonomic family of the isolated phage and its host taxonomic group, the phylogenetic tree was constructed using the Viral Proteomic Tree server (VipTree, https://www.genome.jp/viptree/) based on the genome-wide sequence similarities computed by tBLASTx (Nishimura et al., 2017).

To investigate the taxonomy position of phage vB_EliS-L02 in the siphovirus, phylogenetic analysis was conducted with Virus Classification and Tree Building Online Resource (VICTOR; https://ggdc.dsmz.de/victor.php) (Meier-Kolthoff and Göker, 2017) using the Genome-BLAST Distance Phylogeny (GBDP) method under settings recommended for prokaryotic viruses (Meier-Kolthoff et al., 2013). The phylogenetic tree was visualized by iTOL (https://itol.embl.de) (Letunic and Bork, 2021). Taxon boundaries at the genus and subfamily level were estimated with the OPTSIL program (Göker et al., 2009) with the recommended clustering thresholds (Meier-Kolthoff and Göker, 2017) and an F value (fraction of links required for cluster fusion) of 0.5 (Meier-Kolthoff et al., 2013).

Average nucleotide identity by orthology (OrthoANI) was obtained with OrthoFinder by using all-vs-all BLASTp analysis to determine the genomic similarity between two genomes (Emms and Kelly, 2015, 2019). vConTACT 2.0 was used to calculate the similarity scores between every pair of genomes selected in this study (Teeling et al., 2004; Bolduc et al., 2017). All the taxonomic information of the phages was expanded using BLASTp with a cut-off E-value of 10⁻⁵ and covering an alignment region of more than 50%.

For the maximum-likelihood phylogenetic trees of major capsid (ORF 9) and portal protein (ORF 6) construction, orthologs were obtained by BLASTp with nr protein database in NCBI, sequences alignments were constructed with Clustal W integrated with MEGA X, and run with MEGA X using “LG+G” model with 1,000 bootstrap replications (Kumar et al., 2018). For phoH (ORF 2) phylogenetic tree construction, reference amino acid sequences from viruses and hosts were gathered from NCBI GenBank and aligned using MUSCLE (Edgar, 2004). Then the phylogenetic tree was constructed using IQtree 2.0 with a bootstrap of 1,000 after evaluating the optimal amino acid substitution models (Nguyen et al., 2015), and visualized using iTOL (https://itol.embl.de) (Letunic and Bork, 2021).

To evaluate the distribution of the phage vB_EliS-L02 in different environments, the ORF amino acid sequences were used for homologs recruitment from the Pacific Ocean Virome (POV) database. The reads were recruited by tBLASTn with a cut-off of E ≤ 10⁻⁵, an alignment value ≥ 30, and a score value ≥30, as previously described (Lu et al., 2017). Then the genome of the isolated phage was mapped to the Global Ocean Viromes 2.0 (GOV 2.0) (Gregory et al., 2019) using minimap2 (2.17–r941) (Li, 2018). Reads Per Kilobase per Million mapped reads (RPKM) was used for comparing the relative abundance between different viruses (Mortazavi et al., 2008), and the depth of coverage was calculated using CoverM v0.6.0 (https://github.com/wwood/CoverM) with the following parameters: -m rpkm–min-read-percent-identity 0.95–min-read-aligned-percent 0.75) (Li et al., 2021).

The nucleotide sequence of the phage vB_EliS-L02 was submitted to the GenBank database under access no. OL955261.

Phage vB_EliS-L02 was isolated from surface water in Sanggo bay in October 2018 using the double-agar layer method and found to form small, clear, and round plaques on the lawn of Erythrobacter. TEM analysis revealed that phage vB_EliS-L02 was a siphovirus, the diameter of the long prolate head was 159.4 ± 4.4 nm, and the length of the tail was 240.0 ± 5.1 nm (Figure 1A). The morphology of phage vB_EliS-L02 was remarkably distinct from the only reported Erythrobacter phage vB_EliS-R6L (Lu et al., 2017), and it is similar to a few Cbk-like phages, namely, Roseobacter phage DSS3P8 (Zhan et al., 2016), Phaeobacter phage MD18 (Urtecho et al., 2020), Caulobacter phiCbk-like phage (Gill et al., 2012), and Sphingobium phage Lacusarx (Nielsen et al., 2017).

The one-step growth curve of phage vB_EliS-L02 showed that the latent period lasted for ~150 min (Figure 1B). Then a rapid increase of PFU lasted from 150 to 200 min, with the burst size reaching ~45 PFU per infected cell. Chloroform sensitivity tests showed that phage vB_EliS-L02 was insensitive to chloroform, indicating that the phage was most likely unenveloped.

To determine the host range of phage vB_EliS-L02, 17 bacterial strains belonging to the genera Erythrobacter, Roseobacter, Citromicrobium, Lutibacterium, and Halomonas were used. Phage vB_EliS-L02 could only infect E. litoralis DSM 8509, Erythrobacter JL 475, and E. longus DSM 6997, suggesting a relatively narrow host range (Table 1). The phylogenetic analysis of Erythrobacter strains showed a close evolutionary relationship between the three strains that could be infected by phage vB_EliS-L02 as mentioned above (Zheng et al., 2016), suggesting that phage vB_EliS-L02 may only prefer to infect a specific Erythrobacter clade.

Phage vB_EliS-L02 contained a 150,063-bp double-stranded linear DNA genome. The G + C content was 59.43%, lower than the 65.2% of its host Erythrobacter (GenBank accession no. NZ_CP017057) and 66.5% of the other Erythrobacter phage vB_EliS-R6L (Lu et al., 2017). The genomic features of phage vB_EliS-L02 were then compared with the other 11 Cbk-like phages and showed that the most closely related phage was Sphingobium phage Lacusarx (Nielsen et al., 2017; Table 2).

Twenty-nine tRNA (20 unique) were identified in the genome of phage vB_EliS-L02. A high number of tRNAs were detected in the genomes of other jumbo phages, such as Cbk-like and T4-like phages (Skliros et al., 2016; Yang et al., 2021). These intrigued tRNAs might relate to phage integration in the prokaryotic chromosomes and usage of codons, which could lead to higher translational efficiency of phage genes than that of their hosts (Bailly-Bechet et al., 2007; Delesalle et al., 2016), especially at the late stages of infection when the host rRNA pool might be degraded (Yang et al., 2021). Moreover, the carried tRNAs in jumbo phages could prolong the replication period amid host cell shutdown, and reduce the selection pressure under the long infection period (Yang et al., 2021).

Functional prediction of phage vB_EliS-L02 genome showed 231 ORFs, accounting for 63.6% of the total genome (Supplementary Table 1). Each identified ORF was characterized using BLAST analysis in the NCBI database, 116 ORFs were predicted as functional ORFs, and 50 ORFs were annotated as functional domain categories. Further analysis showed that four functional modules were classified within the functional ORFs, including phage packing, phage structure protein, host lysis, and DNA replication, recombination, and repair functions (Supplementary Table 2). In addition, ORFs 21, 108, and 148 were predicted to encode DUF2460, DUF2958, and DUF1508 domain-containing proteins, respectively, with unknown functions (Supplementary Table 2).

Four genes were associated with phage packing, including portal protein (ORF 6) and tail tape measure protein (ORF 20), which were located at the anterior of the genome, and terminase small subunit (TerS) (ORF 189) and terminase large subunit (TerL) (ORF 190), which were located at the posterior of the genome. Portal protein plays essential roles in phage packing, such as procapsid assembly, screening mutations during packing, tail assembly, and DNA passaging into the capsid (Isidro et al., 2004). Tail tape measure protein dictates the tail length of the tailed bacteriophages and could facilitate DNA transit to the cell cytoplasm during infection (Mahony et al., 2016). The terminase encoded by phage comprises two subunits, TerS and TerL, and could initiate head packaging and process viral DNA into procapsids (Feiss and Rao, 2012).

Eight phage structure–related genes were predicted in the genome, located at the anterior position of the genome. Two genes (ORFs 8 and 13) were predicted to encode virions and phage structure proteins. Major capsid protein (ORF 9) and minor capsid protein (ORF 11) were identified in the genome; the former is the major structural component of icosahedral virus particles, and the latter is believed to enhance the stability of capsid structure by forming protein–protein interactions (Thammatinna et al., 2020). Four ORFs were related to tube structure, including major tail tube protein (ORF 17), which is used for viral genome delivery to the host cells (Davidson et al., 2012); tail fiber protein (ORF 25), which is responsible for host-range determination (Garcia-Doval and van Raaij, 2012); and two tail proteins (ORFs 24 and 26).

Five genes were involved to host lysis, of which, two genes were related to host adhesion, including ORFs 22 and 27 that are contained in F5/8 type C domain and LamG domain, respectively. These two domains were widely distributed in jumbo phages and host cells that play roles in adhesion (Iyer et al., 2021). ORF 23 (NlpC/P60 family protein) and ORF 28 (1,6-anhydro-N-acetylmuramyl-L-alanine amidase, AmpD) are related to host cell wall lysis, in which NlpC/P60 could degrade the peptidoglycan and destabilize the cell wall of hosts (Anantharaman and Aravind, 2003), and AmpD could cleave the cell wall peptidoglycan as an endolysin (Catalao et al., 2013). In addition, ORF 74 was predicted to be rIIA lysis inhibitor that can delay lysis when sensing externally related phage attacking the host (Abedon, 2019).

A total of 16 genes were related to DNA replication, recombination, and repair functions. Seven genes participated in phage DNA replication, recombination, and repair functions directly, and most of them were located in the upstream region of the phage vB_EliS-L02 genome. Two types of DNA polymerases were predicted, including PolC-type DNA polymerase III (ORF 55) and thermostable DNA polymerase I (ORF 60). ORFs 40 (RNA polymerase alpha subunit C-terminal domain protein) and 72 (RNA polymerase-associated protein) are contained in a DNA-dependent RNA polymerase that can help establish the virus at the early stage after infection (Ahn et al., 1994; Iyer et al., 2021). Three ORFs were related to DNA assembly, identified as ribonucleoside-diphosphate reductase (NrdA) subunit beta (ORFs 36 and 37) and subunit alpha (ORF 38), which are tightly coupled to DNA biosynthesis. NrdA catalyzes the first committed step in the biosynthesis of the deoxyribonucleoside triphosphates and could be allosteric effectors that regulate the DNA replication of the phage (Brown and Reichard, 1969; Tseng et al., 1988). Nine ORFs were involved in the regulation of DNA replicating, recombining, and repairing. ORF 32 encodes a DNA ligase that could participate in the processes of DNA replication, recombination, and repair. ORF35 encodes an HTH-type transcriptional regulator, which is usually found to be a DNA-binding domain in the transcriptional regulator, and associated with regulatory metabolism in gene expression (Yu et al., 2019). ORF 47, codifying for an ATP-dependent RecD-like DNA helicase, is implicated in DNA repair (Köppen et al., 1995). ORF 73 encodes a transposase, which may relate to the regulation of gene expression. ORF 77 encodes tyrosine recombinase that mediates unidirectional site-specific recombination between two DNA recognition sequences by mediating strand cleavage through catalyzing tyrosine (Groth and Calos, 2004). Other than that, tyrosine recombinases could protect genome integrity in phages by promoting post-replicative segregation of plasmids and circular chromosomes during cell division, and drive the movement of mobile genetic elements which can promote the phages' integration and transfer in bacterial genomes (Smyshlyaev et al., 2021). ORF188 encodes crossover junction endodeoxyribonuclease (RusA), which is involved in the viral processes of DNA replication, recombination, and repair (Mandal et al., 1993). Two AAA family ATPase-encoded genes (ORF 85 and 117) were predicted, which should participate in DNA replication and recombination (M Iyer et al., 2021). In addition, a tRNA repair-related ORF (ORF 85) was identified in phage vB_EliS-L02 genome, encoding RNA-splicing ligase that participates in tRNA supplementary by ligating cleaved tRNAs (Burroughs and Aravind, 2016).

Since the development of genome sequencing techniques, phage genomes were found to acquire some ecologically important genes from their hosts or other phages by horizontal gene transfer (Lawrence et al., 2002). These genes are usually referred to as auxiliary metabolic genes (AMGs), which are thought to increase viral replication (Hurwitz and U'Ren, 2016) and affect the metabolisms of hosts (Thompson et al., 2011). In this study, 14 ORFs were predicted to be AMGs within the phage vB_EliS-L02 genome.

ORF 2 encodes phoH gene, which belonged to a Pho regulon, and is speculated to play a role in regulating the phosphate uptake and metabolism under a phosphate-limiting condition (Wanner, 1996; Hsieh and Wanner, 2010; Luo et al., 2020). Phosphorus, which serves as a major element for nucleotide biosynthesis, is always a limiting factor in the ocean, especially for oligotrophic waters (Paytan and McLaughlin, 2007). Thus, proteins related to phosphorus acquisition such as phoH, pstS, and phoA, which may assist in phosphorus acquisition during viral infection (Hsieh and Wanner, 2010), are ubiquitous in microorganisms in the ocean. phoH has been used as a marker gene for phage diversity analysis, and the result showed a wide distribution pattern and universal existence across prokaryotes and phages (Goldsmith et al., 2011). In our study, phoH was also used as a marker protein to investigate evolution relationships.

Four ORFs were affiliated with pyrimidine metabolism. ORF 31 encodes FAD-dependent thymidylate synthase (ThyX), which could catalyze dUMP to form dTMP in the presence of reduced nucleotides and oxidized flavin adenine dinucleotide (FAD). The Thy–dUMP complexes with bound FAD show high efficient NAD(P)H oxidase activity and can affect DNA metabolism (Graziani et al., 2004). ORF 44 encodes nucleotide pyrophosphohydrolase (MazG). MazG homologs are implicated as programmed cell death regulators via reducing the levels of the central alarmone molecule (p)ppGpp in Escherichia coli (Gross et al., 2006), while preferring dGTP and dCTP as its substrates and play a role in deoxyribonucleotides hydrolyzing in the phage (Rihtman et al., 2019). ORF 46 encodes deoxynucleotide monophosphate kinase (dNMP kinase), which is an enzyme used for the rapid synthesis of the DNA for phage, catalyzing the phosphorylation of deoxynucleoside monophosphates to the respective diphosphates (Mikoulinskaia et al., 2003). ORF 69 contains a cytidine and deoxycytidylate deaminase zinc-binding region. Deoxycytidylate deaminase (dCMP deaminase) catalyzes the deamination of deoxycytidine 5'-monophosphate nucleotide (dCMP) to uridylate (dUMP), providing the major flux of dUMP production (Maley and Maley, 1990).

Two ORFs were linked to nicotinate and nicotinamide metabolism, including ORF 102 (Nicotinate phosphoribosyltransferase) and ORF 103 (bifunctional NMN adenylyltransferase/nudix hydrolase). These two enzymes belonged to pyridine nucleotide adenylyltransferase (PNAT) family and catalyze the biosynthesis of nicotinamide adenine dinucleotide (NAD⁺) (Singh et al., 2002), an essential cofactor for innumerable metabolic processes. This NAD⁺ biosynthesis belongs to the NAD⁺ salvage pathway, which could be found in large phage genomes (Miller et al., 2003; Yamada et al., 2010; Lee et al., 2017), and provided more substrates for DNA replication and metabolic regulation (Lee et al., 2017).

Moreover, other AMGs involving metabolic regulation were detected in the phage vB_EliS-L02 genome, which provides relaxed pathways for viral metabolism. ORF 34 encodes a 5'-nucleotidase that catalyzes the hydrolysis of phosphate esterified at carbon 5' of the ribose and deoxyribose portions of nucleotide molecules (Zimmermann, 1992). ORF 83 encodes a proteasome, which plays an important role in the regulation of protein degradation, cell cycle progression, DNA repair, apoptosis, and receptor signaling in host cells, and may enhance the efficiency of viral infection (Ros and Kempf, 2004; Volcy and Dewhurst, 2009). ORF 92 encodes a metallophosphoesterase that belongs to phosphatases and could participate in various regulatory metabolisms including mediating the dephosphorylation of certain proteins and phosphodiester bonds to regulate viral transcription (Cohen and Cohen, 1989). Peptidyl-tRNA hydrolase (ORF 109) is identified in the genome, which is broadly distributed in nature. Peptidyl-tRNA hydrolase can rapidly clear peptidyl-tRNA that may lead to cell death with accumulation by cleaving the ester bond between the peptide and tRNA, and plays a critical role in protein biosynthesis in the host cells (Das and Varshney, 2006; Sharma et al., 2014). ORF 130 belonged to the nucleotidyltransferase domain that participates in the repair of the 3′ terminus of defective tRNA molecules in cells (Shan et al., 2008). ORF 134 encodes thioredoxin (TRX), which is known as a hydrogen donor for ribonucleotide reductase (Sengupta and Holmgren, 2014). ORF 166 was predicted to be membrane protease (FtsH), which participates in stress response and protein quality control, and regulates the circuits of membrane proteins regulation in host cells (Langklotz et al., 2012). Otherwise, FtsH is related to the late host lysis step of the phage (Roces et al., 2013).

To gain a better understanding of the evolutionary dynamics of phage vB_EliS-L02, the phylogenetic tree was constructed with genome sequences by using the Viral Proteomic Tree server (VipTree) (Figures 2A,B). Phage vB_EliS-L02 was clustered with the Caulobacter viruses that belonged to Cbk-like phage within the family Siphoviridae, and this result consisted of the morphological analysis. Subsequently, we analyzed our whole genome sequence using BLASTn search on GenBank, and the results showed a low query coverage (<7%), suggesting that phage vB_EliS-L02 is a new phage member compared with the published phage genomes. Then, we compared the similarity of phage vB_EliS-L02 and other 11 Cbk-like phages (Figure 2C). The results showed that the most similar phage was Sphingobium phage Lacusarx, which belonged to the genus Lacusarxvirus within the family Siphoviridae, with an average nucleotide identity (ANI) value of 68.91%. This ANI value was lower than the mechanistic demarcation criterion (70%) for phage genera (Turner et al., 2021). To determine the taxonomic position of phage vB_EliS-L02 in Cbk-like phages, whole genome–based phylogenetic analysis was conducted for phage vB_EliS-L02 and other 19 Cbk-like phages (Figure 2D). The results showed that phage vB_EliS-L02 was clustered with Sphingobium phage Lacusarx and was distant from subfamily Dolichocephalovirinae. Integrating with OrthoANI analysis, phage vB_EliS-L02 should be classified as a new viral genus, named Eliscbkvirus, and could form a subfamily with the genus Lacusarxvirus, named Dolichocephalovirinae-II.

Then, comparative genomic analysis was performed between phage vB_EliS-L02 and typical Cbk-like phages, including Sphingobium phage Lacusarx (GenBank accession number NC_041927), Roseobacter phage DSS3P8 (KT870145), and Phaeobacter phage MD18(MT270409) by BLASTx. Phage vB_EliS-L02 genome showed relatively higher similarity with Sphingobium phage Lacusarx and was significantly different from other Cbk-like phage genomes (Figure 3). The shared ORFs between phage vB_EliS-L02 and phage Lacusarx was 27.3%, including 23 functional domains, with identities ranging from 35.5 to 79.2%. And only 16 genes were shared by phage vB_EliS-L02 and phage Lacusarx with identities higher than 50%, accounting for 15.5% of the total length of phage vB_EliS-L02 genome, including one structure protein (major capsid protein, the amino acid identity of 79.2%, calculated by BLASTp), one phage packaging–related protein (terminase-like family protein, 66.6%), one host lysing–related protein (F5/8 type C domain protein, 51.7%), 11 DNA replication related proteins (52.0–70.3%), and two nicotinate and nicotinamide metabolism–related proteins (51.0–64.8%). Considering only a few ORFs with higher identities were found between phage vB_EliS-L02 and the closed related phage, the result verified that phage vB_EliS-L02 represents a new phage genus.

To further validate the novelty of phage vB_EliS-L02, the phylogenetic trees were constructed by using major capsid protein (MCP), portal protein (Figures 4A,B), and phoH (Figure 5). The former two are ubiquitous and conserved in phage genomes, and are usually used as gene markers for phylogenetic analysis, while the latter one is widely distributed in viruses and hosts, and has been proved to be an effective signature gene for phage diversity analysis in many environments (Goldsmith et al., 2011; Wang et al., 2016; Li et al., 2019). Results of three phylogenetic trees all confirmed that phage vB_EliS-L02 showed a long evolutionary distance from other known phages, and represented a very new virus taxon. Results of phylogenetic trees based on MCP and portal protein showed that Sphingobium phage Lacusarx was the only phage clustered together with phage vB_EliS-L02. And, these two phages were clustered with some Roseobacter phages according to the phylogenetic tree of portal protein, suggesting a relatively close evolutionary distance among them (Zhan et al., 2016; Urtecho et al., 2020). Moreover, the evolutionary relationships, based on phoH among vB_EliS-R6L, other viruses, and their hosts, revealed that phage vB_EliS-L02 formed a distinctly separate cluster, with the closest relationship with Sphingobium phage Lacusarx. And the phylogenetic tree of phoH also showed that these two phages had far genetic distances from other isolated phages and bacteria.

Through metagenome recruit, a total of 4,834 reads within 49 ORFs were obtained. The results showed that the homologs were distributed in diverse ocean environments, including coastal waters, open oceans, and the intermediate waters between them, and most commonly detected in the coastal waters (76.6%), followed by open oceans (13.4%) and intermediate waters (10.0%). Further analysis showed a similar distribution pattern of phage vB_EliS-L02 homologs by using ORF 9 (Major capsid protein), ORF 6 (Portal protein), and ORF 2 (phoH). This distribution profile is consistent with the host Erythrobacter (Shiba and Simidu, 1982; Jiao et al., 2007).

Then, the genome of phage vB_EliS-L02 was mapped to the Global Ocean Viromes data set (GOV2.0), and five viral ecological zones (VEZs) were defined, including Arctic (ARC), Antarctic (ANT), bathypelagic (BATHY), temperate and tropical epipelagic (EPI), and temperate and tropical mesopelagic (MES), as per a previous study (Figure 6; Wang et al., 2021). The number of databases for each ecological environment was normalized before RPKM calculation. The results showed high abundances of Pelagibacter phage, the Puniceispirillum phage HMO-2011 that infects SAR116 clade, and Cyanophages in almost all VEZs, and Alteromonas phages were only abundant in MES, which were similar to some previous studies (Hurwitz and Sullivan, 2013; Kang et al., 2013; Zhao et al., 2013; Wang et al., 2021). All Cbk-like phages showed low abundance in detected VEZs, and vB_EliS-L02 were only detected in EPI and MES (Figure 6), indicating that Cbk-like phages had low abundance in the ocean.