Abstract
Viruses are ubiquitous in the oceans, exhibiting high abundance and diversity. Here, we systematically analyze existing genomic sequences of marine prokaryotes to compile a Marine Prokaryotic Genome Dataset (MPGD, consisting of over 12,000 bacterial and archaeal genomes) and a Marine Temperate Viral Genome Dataset (MTVGD). At least 40% of the MPGD genomes contain one or more proviral sequences, indicating that they are lysogens. The MTVGD includes over 12,900 viral contigs or putative proviruses, clustered into 10,897 viral genera. We show that lysogens and proviruses are abundant in marine ecosystems, particularly in the deep sea, and marine lysogens differ from non-lysogens in multiple genomic features and growth properties. We reveal several virus-host interaction networks of potential ecological relevance, and identify proviruses that appear to be able to infect (or to be transferred between) different bacterial classes and phyla. Auxiliary metabolic genes in the MTVGD are enriched in functions related to carbohydrate metabolism. Finally, we experimentally demonstrate the impact of a prophage on the transcriptome of a representative marine Shewanella bacterium. Our work contributes to a better understanding of the ecology of marine prokaryotes and their viruses.
Similar content being viewed by others
Introduction
Viruses are ubiquitous in natural environments, and their extremely high abundance, diversity, and activity make them indispensable to various Earth’s ecosystems1. Viruses can be generally classified as virulent or temperate; the latter are those viruses that are capable of both lytic and lysogenic infections. During its lysogenic cycle, a temperate virus usually integrates into the host genome to form a provirus or prophage, the latter of which refers to a provirus integrated into the bacterial genomes2,3. Prophages are widespread in bacterial genomes4, and the prevalence of temperate virus-infected microbes, designated lysogens, varies in different environments5,6. Notably, multiple lines of evidence have suggested that temperate viruses and lysogens are prevalent in the ocean, which is the largest ecosystem on Earth7,8,9. Recently, Tuttle et al. analysed 1239 marine bacterial genomes and found that 17.7% of these genomes contained at least one prophage8. Furthermore, multiple viral genome-harbouring integrases have been identified from Mediterranean seawater, indicating that temperate viruses may be widely distributed in the deep sea10. Correspondingly, several temperate viruses have been isolated and members of which can infect dominant marine microbes, such as SAR11 and cyanobacteria11,12. Despite these important advances, most studies either had relatively small sample sizes or focused on culturable microorganisms, thus likely underestimating the occurrence of lysogeny in the oceans8.
Temperate viruses significantly influence a variety of physiological functions in host microorganisms, including DNA replication, gene transcription, protein expression, growth, motility, biofilm formation, and resistance to environmental stressors2,3,13,14. Importantly, the cryptic prophages, although they cannot produce intact phage particles, still confer significant benefits to the host15,16. Despite being a “time bomb” that can cause host cell death, temperate viruses are also pivotal for the survival of marine bacteria7,17,18,19,20. Led by the breakthrough finding that cyanophages possess photosynthesis genes21, researchers have identified and characterized a variety of auxiliary metabolic genes (AMGs) encoded by marine viruses22,23. Specifically, multiple AMGs were found in the Pacific Ocean Virome (POV) and Global Oceans Virome (GOV)24,25, and viral-encoded reverse dissimilatory sulfite reductase was discovered in hydrothermal vent plumes from the Lau and Guaymas basins, suggesting that viruses participate in oceanic sulfur cycling26. To date, although the genomic content and context of AMGs have been investigated in few marine viral clades, such as roseophages27, the occurrence and functional category of marine temperate virus-encoded AMGs warrant systematic assessment.
In general, although an increasing number of studies have suggested that lysogeny is widely distributed in the ocean and that temperate viruses may mediate various biogeochemical processes7,8,22, a comprehensive description and quantitative assessment of the genomic features and ecological functions of marine temperate viruses is currently lacking. Here, based on a large-scale analysis of marine prokaryotic genomes and their accompanying environmental parameters, we constructed a Marine Temperate Viral Genome dataset (MTVGD). Subsequently, we systematically assessed the distribution, diversity and potential ecological functions of marine temperate viruses in oceans worldwide.
Results and discussion
The distribution patterns of lysogens in global oceans
To construct the Marine Prokaryotic Genome dataset (MPGD), we extracted data from the NCBI GenBank database28, the Genomes from Earth’s Microbiomes (GEM)29 and Oceanic Trench Microbial Genomes (OTMG)30 datasets. The MPGD comprises 11,148 bacterial and 932 archaeal genomes with high quality (completeness > 80%, contamination < 5%); it covers 215 classes and 1793 genera (Supplementary Fig. 1, 2 and Data 1). Notably, the MPGD covers various ocean regions, including the Pacific, Atlantic and Indian Oceans (Fig. 1a), and the full ocean depth (from 0 to 11,094 m) (Fig. 1b). Among the genomes in the MPGD, 71.7% and 16.7% were derived from seawater and sediment, respectively. Of the seawater-derived genomes, the majority (79.2%) originated from the epipelagic zone, followed by the deep sea (14.7%) and mesopelagic zone (6.2%) (Fig. 1b). Of the marine sediment-derived genomes, the majority (75.7%) were from the deep sea. In total, we recovered 3351 genomes from the aphotic zone (water depth ≥ 1000 m), including 381 (11.4%) originating from the hadopelagic zone (water depth ≥ 6000 m), thus providing an important dataset to evaluate lysogens in the dark ocean.
a Geographic distribution of genomes in the MPGD. Each point represents a geographic site, and the colour indicates the genome derived environment: seawater, sediment, both or others (information unavailable). The map was drawn using the R package maps (3.4.1)126, in which the “world” data derived from the Natural Earth (v2.0) (https://www.naturalearthdata.com/). b Water depth distribution of genomes in the MPGD. Each circle in the scatter plots shows the water depth, and the circle size is proportional to the number of genomes found at that depth. The pie chart shows the proportions of genomes sampled from different ocean zones: the epipelagic (water depth: 0–200 m), mesopelagic (200–1000 m) and deep-sea (>1000 m) zones. c Ratios of lysogeny in different marine prokaryotic taxa. The bacterial and archaeal phylogenetic trees were constructed based on 120 and 122 concatenated marker proteins, respectively, using the maximum-likelihood algorithm. All branches were collapsed at the class level. Each pie chart corresponds to a class (with ≥ 20 genomes), showing the proportions of lysogenic and nonlysogenic genomes. d Comparison of lysogeny ratios between bacteria and archaea at the class level. n1: number of prokaryotic classes; n2: number of prokaryotic genomes contained in these classes. The significant difference between LyRs of bacterial and archaeal classes was determined by a two-sided Wilcoxon rank-sum test, and the P-value is shown above boxes. Each box represents the interquartile range (IQR), in which the middle line represents the median. The whiskers extend to 1.5 × IQR, and all contained data are shown as the individual points. e Occurrence of lysogeny in different ocean zones. Source data are provided as a Source Data file.
Subsequently, we identified and curated temperate viral sequences in the MPGD (Supplementary Data 2 and 3) to construct the MTVGD, which comprises 12,918 viral contigs (Supplementary Data 4). Notably, 1915 contigs could be recruited by the marine viromes (Supplementary Fig. 3), thereby providing evidence for their dynamic activity in the ocean. Containing 205,027 protein clusters (PCs), the viral contigs were clustered into 10,936 viral operational taxonomic units (vOTUs), showing high diversity of marine temperate viruses. Compared with proviruses in the current largest viral genome database IMG/VR v331, we found that 96.4% of vOTUs and 83.9% of PCs in the MTVGD were novel, thus identifying 16.7% and 11.5% more vOTUs and PCs, respectively, to contribute to the proviral genomes worldwide. Despite this progress, the viral PCs in the MTVGD were far from saturated (Supplementary Fig. 4), suggesting that unprecedented diversity of marine temperate viruses warrant further exploration in the future.
To place viruses with multiprotein-based related phylogeny into clusters, a total of 7964 viral sequences ≥ 10 kb were grouped by vConTACT2 (Supplementary Data 5); 4727 of them were allocated into 1,151 temperate viral clusters (tVCs) (Supplementary Data 6), in which only 8.7% were clustered with reference sequences. Notably, 5 tVCs contained viruses with genome sizes > 200 kb (Supplementary Fig. 5); these viruses probably belong to the huge phage clades32. Regarding taxonomic composition, only 13.6% and 4.2% of total proviral genomes could be identified at the class and family levels, respectively. Peduoviridae (43.1%), Kyanoviridae (24.3%) and Autographiviridae (16.4%) made up the majority of viral family assigned genomes (Supplementary Fig. 6).
In total, 4880 genomes in the MPGD contained at least one viral sequence (mean = 2.65), indicating a high lysogeny ratio (LyR, 40.4%) among marine microbes. This overall LyR was significantly higher than that in a previous report, and was consistent with their speculation that they may underestimate the true occurrence of lysogeny8. We acknowledge that the LyR could still be underestimated due to the majority of partial genomes in the MPGD, as it reached 46.5% and 56.1%, when calculated in genomes with 90% ~ 100% and 100% completeness, respectively (Supplementary Fig. 7).
Different taxa of marine microorganisms showed divergent LyRs (Fig. 1c, Supplementary Data 7), ranging from 8.0% (Thermococci) to 90.5% (Desulfovibrionia) at the class level. Bacteria showed significantly higher LyRs (mean = 41.2%) than archaea (mean = 28.5%) at the class level (P = 0.0412) (Fig. 1d). In terms of bacteria, Alpha and Gammaproteobacteria, the dominant microbes in the MPGD, showed LyRs of 46.7% and 46.6%, respectively. Compared to most classes, Desulfovibrionia, Desulfobacteria, Desulfobulbia and Bacilli displayed higher LyRs, while Cyanobacteriia and Bacteroidia displayed lower LyRs. As for archaea, Methanosarcinia and Bathyarchaeia exhibited relatively high LyRs, whereas Nitrososphaeria and Poseidoniia exhibited lower LyRs (Fig. 1c).
Overall, the LyRs of prokaryotes in marine sediments (48.1%) were higher than those in seawater (38.4%). Interestingly, the LyRs of seawater-derived prokaryotes increased with water depth, reaching 52.8% in deep seawater, while they remained nearly constant in sediment-derived microbes, ranging from 46.1% to 48.7% (Fig. 1e). These trends were consistent in different ranges of genome completeness (Supplementary Fig. 7). We further found that microbes derived from deep-sea water and sediments exhibited significantly higher LyRs (Supplementary Fig. 8 and Data 8). Moreover, the average number of temperate viral sequences carried per lysogenic genome was higher in deep-sea water (mean = 3.01) and sediment (mean = 2.64) than in epipelagic seawater (mean = 2.36) and sediment (mean = 2.31) (Fig. 1e). In accordance with previous studies30,33,34, these results indicate that lysogeny is probably more prevalent in the deep sea.
Among the marine lysogens, 35.4% were polylysogens that contained multiple proviruses (Supplementary Fig. 9). Intriguingly, two polylysogens, Hyphomonas sp. IN11 and Dehalococcoidia sp. UBA6537 harboured 18 and 23 proviruses in their genomes, respectively (Supplementary Fig. 10). The incidence of polylysogeny among different marine environments showed a similar pattern to that of lysogeny (Fig. 1e and Supplementary Fig. 9), while lysogens in mesopelagic sediment showed the highest polylysogeny ratio (43.2%). Although the occurrence of polylysogeny was relatively low in deep-sea sediment (30.1%), polylysogens in that habitat seemed to harbour more temperate viruses (mean = 3.01) (Supplementary Fig. 9). Synthesizing these results, we propose that the lysogenic lifestyle is prevalent in global oceans, especially in the deep sea.
Genomic features and life-history traits of marine lysogens and temperate viruses
Given these findings, we then analysed the divergence of genomic features between the lysogenic and nonlysogenic groups (Supplementary Data 9). Marine lysogens often displayed significantly larger genome sizes, higher guanine-cytosine (GC) content and nitrogen atoms per residue side chain (N-ARSC), as well as lower protein coding density (PCD) and fewer carbon atoms per residue side chain (C-ARSC) than nonlysogens (Fig. 2a–c and Supplementary Fig. 11), while as exceptions, N-ARSC and C-ARSC showed opposite differences in archaea and several bacterial classes, such as Bacteroidia and Bacilli (Supplementary Fig. 11). Interestingly, divergences seemed to occur more frequently and markedly in seawater than sediment, and in bacteria than archaea. The largest differences in genome size and PCD between lysogens and nonlysogens occurred in surface seawater and deep-sea sediment (Supplementary Fig. 12).
a–c Box plots representing the genome size (a), protein coding density (b), and GC content (c), of marine lysogens and nonlysogens. All parameters of lysogens were calculated based on genomes with temperate viral sequences excluded. All the significant differences between lysogens and nonlysogens were determined by two-sided Wilcoxon rank-sum tests, and the P-values are shown above boxes. Each box represents IQR, in which the middle line represents the median. The whiskers extend to 1.5 × IQR, and all contained data are shown as the individual points. d Coloured and grey bars represent the percentage of fast growers (MDT < 5 h) in marine lysogens and nonlysogens, respectively. The marine prokaryotes were grouped based on ocean zone (left two panels) or host taxon (right two panels). Data groups in the (a–d) shared the same sample sizes (number of genomes in each group) which were shown at the top of the (a). Source data are provided as a Source Data file.
To determine whether marine lysogens have life-history traits that have differentiated from those of nonlysogens, we estimated the minimum doubling time (MDT) of marine prokaryotes, and calculated the percentage of fast growers (MDT < 5 h) in different groups (Fig. 2d and Supplementary Data 10). We found that lysogens generally had a higher percentage of fast-growers than nonlysogens. The exceptions were archaea and Bacteroidia, in which the percentage of fast-growers in lysogens was slightly lower than that in nonlysogens. Furthermore, among the fast growers, the MDT of bacterial lysogens was significantly lower (P = 3.7e-10) than that of nonlysogens; the opposite pattern was observed in marine archaea (Supplementary Fig. 13).
To further explore the extent to which genomic features were associated with the prevalence of marine lysogens, we performed a correlation analysis between the genomic features and LyRs (at the genus level) (Supplementary Data 11). We found that genome size, GC content and N-ARSC generally showed a positive correlation with LyR, while PCD and C-ARSC generally displayed a negative correlation with LyR (Supplementary Fig. 14). Notably, genome size was outstandingly significantly correlated with LyR across different marine environments and for almost all abundant prokaryotic taxa. These shared correlations suggest that marine lysogens have unique genomic features, and thus are physiologically and metabolically distinct from nonlysogens. Although this speculation has been confirmed in a few pure cultures of marine bacteria17,19,20,35, further high-throughput and multispecies data are needed.
In addition to marine lysogens, we further exploited the genomic differentiation of marine temperate viruses (Supplementary Data 12). The water depth-dependent evolution and differentiation of viral populations has previously been observed in the Mediterranean Sea, North Pacific Subtropical Gyre and hadal trenches10,30,36. Here we showed, interestingly, that the GC content of seawater-derived temperate viruses first increased and then decreased with increasing water depth, while the C-ARSC showed an opposite trend (Supplementary Fig. 15). This phenomenon is in accordance with a previous report of a microbial genomic transition zone below the photic zone36, suggesting that temperate virus-host coevolution influenced their genomic features. For sediment-derived temperate viruses, differences in genomic features were also associated with water depth but showed diverse trends (Supplementary Fig. 15). However, temperate viruses from different depth-stratified zones did not significantly differ in N-ARSC, suggesting that nitrogen source did not influence this divergence to the same extent.
Interaction networks of marine temperate viruses and prokaryotes
To identify ecologically important marine temperate viruses, we initially focused on enriched tVCs that encompassed ≥15 and ≥5 viral genomes in the seawater and sediment groups, respectively. Then, the distribution and host ranges of these tVCs were analysed; they were shown to be distinct in sediment and seawater and to differ in distribution in different ocean zones (Fig. 3a, b). Several tVCs (tVC 583_0, 168_0, and 955_0) were ubiquitous in all groups, suggesting that they may be cosmopolitan tVCs in the ocean. Then, to illustrate the marine temperate virus-host interactions, we included the enriched prokaryotic genera (with ≥20 and ≥5 genomes in the seawater and sediment groups, respectively) and resident tVCs in interaction networks (the complete networks are provided in Supplementary Figs. 16 and 17). Notably, the networks showed significantly more frequent virus-host interactions in seawater than in sediment (Supplementary Data 13). Subsequently, we identified the tVCs infecting ecologically important marine prokaryotic clades (Fig. 3c, d). Among the seven Pelagibacter-infecting tVCs, two unclassified tVCs, 784_0 and 577_0, showed frequent infections, which was represented by a high frequency of these temperate viruses present in the genomes of a host genus. tVC 784_0 contains the recently reported prophage PNP111, and shares homologous proteins with the highly abundant SAR11 virus HTVC010P; tVC 577_0 clustered with several previously reported SAR11 viruses (HTVC011P and HTVC019P)37 (Supplementary Fig. 18). Notably, in addition to SAR11, tVC 577_0 can infect multiple genera of Gammaproteobacteria. For archaeal viruses, two novel tVCs, 1190_0 and 1191_0, were found to frequently infect Bathyarchaeia (Fig. 3d), which is the critical driver of the carbon cycle in marine sediment38. Both of them carried head and tail proteins homologous to bacteriophages, and TerL homologous to archaeal virus HGTV-1 (Supplementary Fig. 19).
a, b Distribution and host ranges of temperate viral clusters (tVCs) in seawater (a) and sediment (b). For clarity, only tVCs with ≥15 (in seawater) or ≥5 (in sediment) viral genomes are shown in the heatmaps. The heatmaps show the number of viral genomes in tVCs derived from 3 depth-stratified ocean zones, which are hierarchically clustered by tVCs. The tVCs present in all 6 ocean zones are considered as putative cosmopolitan tVCs and are marked in green next to the tVC names. Each tVC is connected to the class of its host(s), and the transparency of the connecting lines is proportional to the number of infections. c, d Interaction networks of prokaryotic genera and tVCs in seawater (c) and sediment (d). For clarity, only prokaryotic genera with ≥20 genomes (in seawater) or ≥5 genomes (in sediment) are displayed in the network. The circles and hexagons represent host genera and tVCs, respectively, and the sizes are proportional to the numbers of genomes included. The coloured circles represent different bacterial classes or archaeal phyla, the inner letters mark ecologically relevant marine microbial hosts (at the genus level), and the yellow hexagons represent tVCs that infect multiple host genera. The numbers of infections are displayed as the shared edges and proportional to the transparency and width. The networks are visualized using the edge-weighted spring-embedded model, which places the host genera and tVCs sharing higher co-occurrence in closer proximity. Source data are provided as a Source Data file.
Although most tVCs show narrow host ranges (Supplementary Fig. 20), we identified 301 and 68 multigenera-infecting tVCs in seawater and sediment, respectively (Fig. 3c, d, Supplementary Figs. 16 and 17). Notably, tVC 409_0, which was classified in the family Peduoviridae, was highly enriched (Fig. 3a, b), infected 15 and 9 genera in seawater and sediment, respectively, and interacted with multiple Gammaproteobacteria hosts (Fig. 3c, d, Supplementary Fig. 21), which have been reported to prevail in diverse marine environments39,40. To conservatively evaluate broad-host-range viruses, we classified MTVGD into 10,897 temperate viral genera (tVG) using VIRIDIC41, and found 30 multiphylum-infecting and 37 multiclass-infecting tVG, represented by 5 tVG infecting Gammaproteobacteria and other classes (Supplementary Fig. 22). Some of the cross-phylum/class infectivity were further supported by host prediction based on virus-host homology of CRISPR spacer or nucleotide sequence (Supplementary Fig. 22). Host DNA can be carried by temperate viruses and transferred to other hosts via the process of transduction, which is an important mechanism of provirus-mediated horizontal gene transfer (HGT)42. Among diverse variables, the host range is a key variable that affects the ability of virus-mediated HGT43. Thus, we propose that tVCs/tVG with broad host ranges could potentially be important mediators of HGT in marine ecosystems. Finally, by analysing the co-occurrence of different tVCs in the same host, we observed significantly intertwined coinfections among tVCs, especially in seawater (Supplementary Fig. 23 and Data 14), indicating their high compatibility and potential mutual effects.
Leveraging the viral-host interaction analysis, we revealed putative ecologically important tVCs/tVG, including those that (1) infect dominant marine microbial clades (such as SAR11, Gammaproteobacteria and Bathyarchaeia); (2) are enriched in multiple environments and thus potentially cosmopolitan in the ocean; and (3) feature extraordinarily wide host ranges and are speculated to be important in mediating genetic exchanges. Previously, multiphylum-infecting phages were isolated from Lake Michigan, and these phages are capable of infecting 5 strains belonging to 3 different bacterial phyla including Proteobacteria, Actinobacteria and Bacteroidetes44. Moreover, metagenome-based studies and CRISPR matches have found viral clusters linked to multiple hosts from different phyla45. Especially, in marine environments, broad-host-range viruses are reported to be distributed widely46,47. Recently, microbial domain-crossing host-viral interactions were illustrated in dense hydrothermal mats in the Guaymas Basin48. These findings suggested that despite currently no marine temperate viruses capable of infecting multiple phyla have been isolated, they are very likely to exist in the ocean. Although the vast genetic diversity of marine viruses has been documented in several fundamental datasets31,45,49, we believe that the aforementioned tVCs and their hosts warrant isolation and characterization, thereby revealing details of their life strategy and ecological functions.
Ecological influences of marine temperate viruses
To further assess how and to what extent temperate viruses affect marine microbes and ecosystems, we inspected their auxiliary metabolic genes (AMGs), integration sites and reprogramming effects. We identified a total of 302 AMGs in 239 marine temperate viruses (Fig. 4a and Supplementary Data 15) and predicted the 3D structures of all AMG-encoded proteins. Notably, 55.0% of them had homologous proteins in the Protein Data Bank (PDB) database with highly similar structures (Z-score > 20), indicating that they probably shared similar functions (Supplementary Data 15). Generally, AMGs displayed a niche-dependent distribution pattern, and the AMGs involved in carbohydrate-active enzyme (CAZy) were highly enriched, especially in deep-sea sediments (Fig. 4b). Moreover, the AMGs involved in photosynthesis were exclusively enriched (as expected) in epipelagic seawater, while the transporters encoding AMGs seem to be more preferred in the deep ocean. The CAZy-related AMGs encode diverse enzymes belonging to multiple classes, including glycoside hydrolase (GH, n = 11) and glycosyl transferase (GT, n = 23) (Fig. 4c). Next, we selected two GH-related AMGs, vgh5a and vgh16c, to experimentally test whether they encoded enzymatically active proteins. As a result, vGH5A showed hydrolytic activities of β-D-galactosidase, β-D-glucosidase and β-D-fucosidase (Supplementary Fig. 24), in accordance with the functional connectivity of these enzymatic activities50. However, vGH16C had no detectable activity under our experimental conditions (Supplementary Data 16). Notably, the predicted 3D structure of the vGH5A protein shared high homology with marine-derived β-glucanase (Z-score = 23.8) and soil-derived β-galactosidase (Z-score = 23.2) (Supplementary Fig. 24), which can hydrolyse β-(1,3;1,4)-glucans and xyloglucan, respectively51,52. These results suggested that AMG vgh5a may help microbial hosts hydrolyse polysaccharide substrates in the ocean.
a, b Heatmap showing the relative abundance of all AMGs (a) and the functionally categorized AMGs (b) encoded by marine temperate viruses in each group. The relative abundance was calculated as the average number of AMGs carried by per viral genome. The groupings are based on the ocean zone or host taxon. Functional categories of AMGs are annotated by DRAM-v82. The number of viral genomes contained in each group is shown at the top of the heatmap. CAZy Carbohydrate-Active enZYmes, MISC miscellaneous. c Composition of AMGs involved in CAZy. The asterisks indicate that AMGs were functionally characterized in this study. CBM carbohydrate-binding module, CE carbohydrate esterase, GH glycoside hydrolase, GT glycosyl transferase, PL polysacchaide lyase. Source data are provided as a Source Data file.
The integration of temperate viruses can substantially change the genomic structure and gene expression of hosts2. To examine the preference and potential influences of marine provirus integration, we identified the integration loci of 2794 temperate viruses in host genomes (Supplementary Data 17). Apart from 49.3% of loci located in noncoding regions, 16.5% and 40.9% of the analysed proviruses integrated into transfer RNAs (tRNAs) and protein-coding genes, respectively (Supplementary Fig. 25). The tRNAs Leu, Ser and Arg were integration hotpots. Strikingly, these hotspots were generally consistent with the dominant tRNAs encoded by marine temperate viruses (Supplementary Figs. 25–26 and Data 18), suggesting that integration site selection and the translational compensation of marine proviruses are intrinsically linked. Additionally, we found that the genetic regions related to two-component systems, transporters and transcription factors were preferentially selected for marine temperate virus integration (Supplementary Fig. 25).
Marine virus-induced metabolic reprogramming is related to virus-host genome complementarity35. To quantitatively assess this effect in oceans worldwide, we measured oligonucleotide frequency dissimilarity (d2*) and codon cosine distance between all temperate virus–host pairs and used these variables as indices of nucleotide and amino acid complementarity, respectively (Supplementary Data 19). Remarkably, both indices were negatively correlated with the LyR in different ocean zones and major prokaryotic taxa; the codon distance had more shared correlations, implying high dependency of temperate viruses on host translational machinery (Fig. 5a). We propose virus-host genome complementarity as a general influencing factor of lysogeny prevalence in microbial communities. Subsequently, we compared the two indices among different virus-host groups, and significant differences were observed between the deep and epipelagic ocean, as well as between the bacteria and archaea (Fig. 5b, c). In particular, virus-host complementarity reached the lowest level in the epipelagic seawater and deep-sea sediment groups. These results coincided with the increased divergences in several features between lysogens and nonlysogens in these two marine biomes (Fig. 2 and Supplementary Fig. 12), suggesting that temperate viruses living in these habitats can relatively heavily reprogramme host metabolism and alter host genomic features.
a Correlation between marine temperate virus-host complementarity and lyRs. Pearson’s correlation coefficients between the median d2*/codon distance and lyRs of genera (with ≥5 genomes) were calculated and are shown by the colour gradient of squares. Significant correlations were determined by two-sided tests and marked by white asterisks and P-values are represented by different sizes of asterisks and shown above squares. The slashes indicate statistical unachievability due to the small sample size (number of genera < 5). n1: number of host genera used for calculation of pearson’s correlation; n2: number of host genomes contained in the genera. b, c Distribution of temperate virus-host nucleotide (b) and amino acid complementarity (c) in different ocean zones or host taxa. The differences among groups were analysed by the two-sided Wilcoxon rank-sum test, and P-values of the significant differences are shown above boxes (other compared groups are shown in Supplementary Data 19). Each box represents the IQR, in which the middle line represents the median. The whiskers extend to 1.5 × IQR, and all contained data are shown as the individual points. The number of virus-host pairs contained in each group is shown at the top of the graphs. Source data are provided as a Source Data file.
Marine temperate phages are active and influence the transcriptome of host bacteria
To further validate the impacts of temperate viruses on hosts, the marine bacterium Shewanella psychrophila WP2 (hereafter referred to as WP2)53,54, which was isolated from deep-sea sediment in the western Pacific Ocean and harbours a prophage SP1 (MTVG_4770), was used as a representative temperate virus-host system. The boundary of SP1 was experimentally determined (Supplementary Fig. 27); its genome size is 42,608 bp, and it is integrated into the 5’ terminus of dusA and noncoding regions adjacent to the 3’ terminus of cheX (Fig. 6a). The excision and circularization of SP1 were verified, indicating its functionality as a complete prophage (Fig. 6b). Intriguingly, SP1 maintained an excision frequency of ~1% under all the tested conditions (Fig. 6c). The frequency is remarkably high compared with that of other prophages (ranging mostly between 10-7–10-4)15,17,55,56,57, and indicates that SP1 can be actively excised from WP2 genomes in in situ marine habitats.
a Genomic map of the prophage SP1 in the marine bacterium S. psychrophila WP2 (WP2). The arrows depict the location and direction of predicted proteins on the phage genomes, and the fill colours indicate different functional categories of genes, as indicated in the legend. b Verification of SP1 excision by PCR. The left schematic graph shows the process of SP1 excision, in which site-specific recombination occurs through the crossover between attL and attR sites to generate the SP1-deleted WP2 genome and a circular SP1 genome. The locations of the primer pairs used for verification are also shown. The right panel shows the electrophoresis of PCR products. The primer pairs and template DNA used for PCR are indicated for each lane, and the target bands are marked with an arrow. M, DNA size marker. Image of representative agarose gel from two independent experiments are shown. c Excision rates of SP1 in S. psychrophila WP2 under different treatments. The data represent the mean ± SD and are based on three biologically independent samples. MMC, Mitomycin C. d Graphic display of differentially expressed genes (DEGs) categorized by function in S. psychrophila WP2 after SP1 deletion. The transcriptome data represent three biologically independent samples for each strain (WP2 and WP2ΔSP1). Normalized differential expression levels (fold changes) are represented by heatmaps in boxes according to the scale bar (log2 scale) from most upregulated to most downregulated. The proteins encoded by the DEGs are shown in each box. Source data are provided as a Source Data file.
To investigate the impacts of SP1 on the host, we deleted SP1 from the WP2 genome to obtain the WP2ΔSP1 strain (Supplementary Fig. 27). Although SP1 did not significantly influence the growth of WP2 (Supplementary Fig. 28), it had a significant influence on the transcriptome of WP2 under stimulated in situ environmental conditions (Fig. 6d and Supplementary Fig. 29). Overall, 56 genes were differentially expressed between WP2ΔSP1 and WP2 (Supplementary Data 20). Notably, the transcriptional levels of 7 genes (cysDNHIGJK) involved in assimilatory sulfate reduction were significantly upregulated in WP2 compared with WP2ΔSP1 (Fig. 6d). In addition, the genes involved in serine metabolism, including cysE, sdaA, and eamB, were upregulated, and the transcription levels of genes participating in cell motility, the type III secretion system and extracellular hydrolysis were higher in WP2 than in WP2ΔSP1. The 4 downregulated genes were responsible for polyhydroxybutyrate (PHB) synthesis and disulfide bond formation. These results suggest that marine temperate viruses can significantly alter a variety of host physiological activities, including carbon, nitrogen and sulfur metabolism.
Virus-infected cells, termed virocells, exhibit a reprogrammed metabolism and therefore physiologically differ from uninfected cells58,59. Specifically, Pseudoalteromonas phage PSA-HP1, isolated from surface seawater, drastically alters the central carbon and energy metabolism of its host35. In addition, filamentous phage SW1 significantly influences the growth, motility and transcriptome of its deep-sea bacterial host S. piezotolerans WP319,20. Shewanella strains are ubiquitous in the ocean owing to their versatile metabolism60,61. Although SP1 is currently unclassified, this actively excising prophage resembles multiple temperate phages among marine Shewanella species (Supplementary Fig. 30), and the recruitment analysis showed that the SP1-like prophages are prevalent in the Pacific Ocean (Supplementary Fig. 31). Therefore, SP1-infected virocells (WP2) and nonvirocells (WP2ΔSP1) are likely to widely coexist in natural marine environments. Our comparative transcriptome analysis of virocells and nonvirocells under simulated benthic marine conditions (high pressure and low temperature) provided experimental evidence for the effects of deep-sea temperate viruses on prokaryotic hosts, data that were previously scarce.
Pioneering global viromic studies have revealed the remarkable abundance, diversity, geographic distribution and ecological roles of virioplankton25,49,62. However, the lack of a comprehensive investigation of temperate viruses that reside in marine lysogens hinders a complete understanding of marine viruses. In this study, we systematically revealed the distributions of lysogens and temperate viruses in global oceans (Supplementary Fig. 32), and described the viral diversity as well as underlying genomic features associated with marine lysogens. By examining virus-host interactions, we identified potentially ecologically important tVCs. Combining large-scale data analysis and experimental evidence showing divergences between lysogenic virocells and nonlysogenic cells, we suggest that lysogenic virocells play distinct roles in ecological processes. Taken together, these findings indicate the critical roles of marine temperate viruses in rewiring host metabolism, shaping microbial communities and participating in biogeochemical cycles.
Methods
Construction of the Marine Prokaryotic Genome dataset
We collected prokaryotic genomes and accompanying biosample information from the NCBI GenBank (v244) database28, GEM29 and OTMG30 datasets. To screen for marine-derived genomes, corresponding biosamples were preliminarily filtered by keywords, including “marine”, “ocean”, “sea” and “pelagic”. Combined with available references and geographic coordinates, we manually curated the biosamples; only marine-derived biosamples were retained. Next, these candidate marine biosamples were further purified based on the following criteria: (1) excluding biosamples collected at the border between marine and terrestrial environments, including the coast, beach, seashore, estuary and intertidal zone, due to strong terrestrial influences on microbes inhabiting these ecosystems;63,64,65 and (2) excluding host-associated biosamples, such as algae-, coral reef-, eukaryotic plankton- and mammal-associated samples, because the genomes of parasitic/symbiotic microbes are highly influenced by their hosts and differ widely from free-living microbes66,67,68,69. In addition, the information for each genome-derived sample, including sample type (seawater or sediment) and water depth, was adequately collected and curated, and genomes were classified as epipelagic (0–200 m), mesopelagic (200–1000 m) or deep sea (>1000 m), according to the depth at which they were sampled. If this information was unavailable, the genomes were labelled “not available” (NA).
CheckM (v1.1.3)70 was used to evaluate the quality of marine prokaryotic genomes, and only high-quality genomes (completeness > 80% and contamination < 5%) were retained. For species-level descriptions, MUMmer-based71 average nucleotide identities (ANIs) among all the genomes were calculated using pyani (v0.2.7);72 genomes with > 95% ANIs and 60% alignment length were grouped into prokaryotic operational taxonomic units (OTUs)29,73. Moreover, genomes in the same OTU and samples were dereplicated.
Taxonomic classification of all genomes was performed by GTDB-Tk (v1.3.0)74 using the “classify_wf” pipeline. Briefly, 120 and 122 marker proteins in the bacterial and archaeal genomes, respectively, were identified, concatenated and aligned75,76. Then, the maximum-likelihood placement of a genome was determined in the GTDB-Tk reference tree77. Data on the placement in the reference tree, relative evolutionary divergence, and/or ANI to reference genomes were combined to classify a given genome74. The phylogenetic tree of marine prokaryotic genomes was constructed by FastTree2 (v2.1.10)78 based on the maximum-likelihood algorithm and visualized by iTOL (v6)79.
Construction of the Marine Temperate Virus Genome dataset
First, VirSorter2 (v2.2.2)80 was used to predict viral sequences in the curated Marine Prokaryotic Genome dataset with the setting “--include-groups dsDNAphage, ssDNA -min-length 5000 -min-score 0.5”. The mode of searching double-stranded DNA (dsDNA) and single-stranded DNA (ssDNA) viruses was appropriate for recovery of temperate viruses, the minimum score (0.5) was chosen for maximal sensitivity, and the minimum length (5,000 bp) was necessary for the downstream steps. Second, CheckV (v0.8.1)81 was used to control the quality of the viral sequences, through steps such as removing nonviral sequences, trimming nonviral genomic regions and evaluating quality. Third, the CheckV-trimmed sequences were passed through VirSorter2 again and formatted to serve as input to DRAM-v (v1.2.4)82 for viral annotation. Retained viral sequences were curated according to the widely recognized empirical screening criteria based on the information of viral and host gene counts, score, hallmark gene counts and sequence length83. The sequences requiring manual checking were further curated according to summarized criteria based on detailed viral annotation from DRAM-v82. Specifically, the viral sequences > 200 kb were screened based on a highly confident VirSorter2 score (cut-off of 0.9) and strictly curated, resulting in 25 sequences being discarded, because long sequences with lower scores frequently contained dense nonviral proteins and are normally considered false positives80,83. In addition, we excluded 15 viral sequences annotated as φX174 that were commonly used as quality control in Illumina sequencing84.
Assessment of the prevalence of lysogens
Lysogeny ratios (LyRs) for each taxon and environment were calculated to determine the prevalence of lysogens. The LyRs in different prokaryotic taxa were examined at the superkingdom, class, and genus levels (Supplementary Data 7). In the statistical analyses, only genera with ≥5 genomes and classes with ≥20 genomes in the MPGD were analysed. Polylysogeny and coinfection of temperate viruses were defined as the detection of multiple distinct temperate viruses in a single prokaryotic host84. However, the frequency of polylysogeny can be overestimated if a single viral genome is misassembled and/or split into several contigs. To avoid overestimation, we adopted rigorous criteria to determine polylysogeny and coinfection, according to the previous study84. Briefly, the large unit of the terminase (TerL) was considered as a single-copy marker protein (SCMP) in Caudovirales and other unclassified viruses84. In addition, the zonular occludens toxin (Zot) was considered as a SCMP in Inovirales85. Only viral genomes containing SCMPs were included when calculating the polylysogeny ratio.
Viral clustering, taxonomic assignment and network analysis
Prodigal (v2.6.3)75, with default parameters, was used to determine the open reading frames (ORFs) of temperate viral genomes. All the viral proteins were clustered by CD-HIT (v4.8.1)86 using 60% identity, 80% coverage of the shorter sequence and the recommended parameter of ‘-g 1 -n 4 -d 0’24. They were also coclustered with viral proteins of the IMG/VR provirus database (v3)31, using the same parameters, to enable protein family comparison.
We used pyani (v0.2.7)72 to group temperate viral genomes sharing 95% MUMmer-based71 ANIs with at least 85% length coverage into temperate viral OTUs (tvOTUs), at approximately the species level87. They were also coclustered with genomes from the IMG/VR provirus database31 with the same parameters for vOTU comparison. Next, viruses ≥10 kb were clustered into temperate viral clusters (tVCs) using vConTACT2 (v2.0);88 in brief, all-to-all BLASTp, similarity calculation and Markov clustering were performed with the default parameters88. The reference viral genomes provided by vConTACT2 were coclustered with query genomes for taxonomic prediction. A query genome was assigned the taxonomy of clustered reference genomes, in which the taxonomy occupying the highest proportion was chosen. To supplement this analysis, the taxonomy of viral genomes not assigned in the vCONTACT2 pipeline and < 10 kb in size was further predicted for taxonomy by a “majority-rules approach”49. In detail, all of their proteins were subjected to Diamond (v0.9.14.115)89 for BLASTp alignment against RefSeq Virus (v207);90 a viral genome was considered to belong to a viral family if ≥50% of the proteins were assigned to that family with a bitscore ≥50. The host range of a tVC was calculated as the number of different host taxa its included viruses infected.
To conservatively evaluate host ranges, we further clustered viruses into temperate viral genera (tVG) using VIRIDIC41 based on a nucleotide identity cut-off of 70%, according to the ICTV standard91. The tVG with cross-class/phylum/superkingdom host range were further checked by host prediction based on virus-host homology of CRISPR spacer and nucleotide sequence, which were previously demonstrated to show high precision92. Prokaryotic genomes from the RefSeq database (Release 206, n = 216,865) and the MPGD (n = 12,080) were used for the prediction. The CRISPRs of prokaryotic genomes were predicted using the CRISPR Recognition Tool (CRT)93 with optimized parameters as previously described94, and CRISPR clusters with less than three spacers were discarded. The CRISPR spacers were aligned against the viral contigs using BLASTn and matches with 100% coverage and <2 mismatches were retained95. For nucleotide homology, viral contigs were aligned against prokaryotic genomes using BLASTn with the cutoff of bitscore >50, e-value < 1e-3, identity > 70% and match length > 2500 bp96.
The virus-host interaction network analysis was based on the number of infections by a given tVC in a given prokaryotic genus, which was calculated as the number of temperate viral genomes in a given tVC present in a given host genus. In the network, prokaryotic genera and tVCs were represented as distinct nodes, and the number of infections was represented as shared edges. Similarly, the coinfection network analysis was based on the number of coinfections with tVCs, and only tVCs carrying SCMPs were included. In this network, tVCs were represented as nodes, and the number of coinfections of two tVCs was represented as shared edges. Protein sharing network analysis of SAR11 viruses was performed by vConTACT288. All of the networks were visualized and analysed by Cytoscape (v3.7.0)97 using the edge-weighted spring-embedded model.
Characterization of genomic features and maximal growth rates
Protein coding density was calculated as the number of predicted proteins per kb genome. Codon relative frequencies were calculated using the “uco” function in the R package seqinr35. The script “get_gc_and_narsc.py” was used to calculate the GC content and nitrogen/carbon atoms per residue side chain (N/C-ARSC) of the genomes98. For analysis of microbial genomic features, we excluded temperate viral sequences from microbial genomes and calculated the genomic features of these “pure” host genomes. The maximal growth rates of prokaryotes were estimated by the R package gRodon (v1.0) based on genome-wide codon usage statistics99. The genes encoding ribosomal proteins, which are typically highly expressed in prokaryotes, were identified by BLASTp alignment against bacterial ribosomal proteins (br01610) in the Kyoto Encyclopedia of Genes and Genomes (KEGG) database100 with e-value < 1e-10, identity > 30% and coverage > 50%. The Minimal doubling time (MDT) was calculated using the “predictGrowth” function in the gRodon package, with data from whole genes and genes encoding ribosomal proteins as the input. If a genome was incomplete, the function was set to the “partial” mode. As fast-growing copiotrophs in the ocean normally exhibit a maximum growth rate of >1 d-1, and MDT prediction appeared to be accurate only up to 5 h in practice99,101, only prokaryotes with a predicted MDT < 5 h were considered fast growers in the study.
The nucleotide and amino acid dissimilarity between each virus-host pair were calculated using the method previously reported35. Briefly, oligonucleotide frequency dissimilarity d2* at k = 6 was adopted to determine the nucleotide distance between a virus and host, using VirHostMatcher102 with default parameters, and a cosine distance between the codon frequency vectors of the virus and host (Dc(phage, host)) was used to describe their amino acid dissimilarity35.
Identification of putative auxiliary metabolic genes (AMGs)
All of the marine viral genomes were annotated by the DRAM-v workflow82. Then, each viral gene was assigned metabolic flags and an auxiliary score, determined by its location on the viral sequence relative to other viral hallmarks or viral-like genes identified by VirSorter103. According to the usage guidelines, only viral genes with a metabolic flag’M’ and a confidence score of 1–3 were considered as candidate AMGs82. To further screen high-confidence AMGs, we manually curated the candidate AMGs to retain those with unambiguous virus-like genomic contexts, referring to conservative criteria reported previously96. Specifically, a conserved set of AMGs was required to have viral structural genes, terminases or integrases in genomic regions both downstream and upstream of the gene. In addition, we manually deleted the candidate AMGs that are probably involved in viral functions, including DNA methylase- and nucleotide metabolism-related genes. Gene annotations of viral genomes containing AMGs are displayed in Supplementary Data 15 in which the putative AMGs, viral structural genes, terminases and integrases are marked. In AMG distribution analyses, a relative abundance of an AMG category was calculated as the average number of included AMGs carried by per viral genome, which was dereplicated according to their gene IDs.
Analysis of integration sites of marine temperate viruses
After a temperate virus integrates into a host genome, the provirus usually ends with two attachment sites, attL and attR, which display as short repeated sequences104. To ensure reliability, only temperate viral genomes with both two flanking host sequences which contained ≥5 ORFs were included in the integration site analysis. In line with previous studies105,106, we scanned perfect repeated sequences within 1000 bp inside the viral boundaries, as determined by VirSorter2 and/or CheckV, and the length of att sites ranged from 2–145 bp, with a bias toward longer repeats. The host sequences upstream of attL and downstream of attR were reconnected to recover the integration loci, and then subjected to a BLASTx89 alignment against the KEGG database107 online with an e-value cut-off of 1e-10. In addition, tRNAscan-SE (v2.0.3)108 was used to predict tRNA in the integration loci, as well as in all temperate viral genomes. The “-B” and “-A” modes were combined for the bacterial and archaeal tRNA prediction. The context of the integration loci was examined by analysing 5 flanking host genes; these genes were annotated by BLASTp89 alignment against the KEGG database107 online with an e-value cut-off of 1e-10. Only temperate viral genomes with ≥1 flanking host sequences which contained ≥5 ORFs were included in this analysis.
Statistical analysis
The two-sided Wilcoxon rank-sum test109 was adopted to evaluate all the significance of differences between groups using the Python function “ranksums” from Scipy (v1.5.0)110. Pearson correlation analysis111 was used to determine the relationship between LyRs and genomic features at the genus level via the Python function “pearsonr” (two-sided test) from SciPy (v1.5.0);110 only the genera with ≥5 genomes were included in this analysis.
Read recruitment analysis
Read recruitment of marine temperate viruses in the Pacific Ocean Virome (POV)24 was performed by BLASTn alignment as previously reported112, using an e-value ≤ 10-5, identity ≥ 95% and hits >50 bp. The relative abundance was calculated as the number of mapped reads per kilobase of temperate viral genome per billion reads in each metagenome. Recruitment analysis of SP1-like viruses was performed by BLASTn with an e-value cut-off of ≤ 10−3 against metagenomic raw reads from the POV samples with water depths <1000 m and ≥1000 m.
Functional characterization of viral glycoside hydrolases
The vgh5A gene (MTVG_4775_21) encoding a viral glycoside hydrolase was synthetized and cloned into the pET28a plasmid (Saiheng Biotechnology, Shanghai, China), and the recombinant plasmid was transformed into E. coli BL21 (DE3) cells. The transformant was selected on LB medium containing kanamycin and confirmed by PCR and DNA sequencing. E. coli BL21 (DE3) cells containing the vGH5A expression vector were grown in LB broth with 50 μg/ml kanamycin at 37 °C. vGH5A expression was induced by the addition of IPTG (0.2 mM) when the OD600 reached 0.6, and the culture was then incubated at 15 °C for 16 h. The cells were collected by centrifugation, resuspended in binding buffer [500 mM NaCl, 10% glycerol, and 20 mM Tris–HCl (pH 8.0)], and sonicated on ice. The cell extract was clarified by centrifugation at 10,000 × g for 45 min at 4 °C. Ni-NTA Sefinose (TM) Resin 6FF (Sangon Biotech, Shanghai, China) was used to purify His-tagged vGH5A according to the manufacturer’s instructions. The protein was eluted in elution buffer [500 mM NaCl, 10% (v/v) glycerol, 300 mM imidazole, and 20 mM Tris–HCl (pH 8.0)]. The purity of the protein was confirmed by SDS-PAGE (12% acrylamide) with visualization using Coomassie Brilliant Blue R-250, and the concentration of the purified protein was determined using a Nanodrop 2000 spectrophotometer (Thermo Scientific, Waltham, USA).
The enzyme activity of glycoside hydrolysis was determined using the p-nitrophenyl (pNP) and 3,5-dinitrosalicylic acid (DNS) methods as previously described113,114. For the pNP method, pNP-β-D-xylopyranoside (pNP-βXyl) (Haohong Biomedical Technology, Shanghai, China), pNP -β-D-mannopyranoside (pNP-βMan) (Macklin Biochemical Technology, Shanghai, China), pNP -β-D-galactopyranoside (pNP-βGal) (Macklin), pNP -β-D-fucopyranoside (pNP-βFuc) (Macklin), pNP -α-L-arabinofuranoside (pNP-αAra) (Solarbio Science & Technology, Beijing, China) and pNP -β-D-glucopyranoside (pNP-βGlc) (Biosynthesis Biotechnology, Beijing, China) were used as substrates, and the enzymatic activity measurement was performed in final volumes of 210 μl of 50 mM sodium phosphate buffer (pH 7.0) containing 1 mM p-nitrophenyl-glycosides and 10 μg recombinant vGH5A. After incubation at 50 °C for 60 min, 20 μl of Na2CO3 (2 M) was added to stop the reaction and stabilize the chromophore in its anionic 4-nitrophenolate form. The enzyme activity was determined colorimetrically at 410 nm by measuring the released p-nitrophenol using a Synergy H1 microplate reader (BioTek, Winooski, USA). For the DNS method, polysaccharides including barley β-glucan (Macklin), carboxymethylcellulose (Titan Biotech, Shanghai, China), lichenan (Megazyme, Bray, Ireland), xyloglucan (Megazyme), and extracts of sargassum (Ruiyuanze Biotech, Gansu, China), porphyra (Ruiyuanze), kelp (Ruiyuanze), bladderwrack (Ruiyuanze), and locust bean gum (Macklin) were used as substrates. The enzymatic activity measurement was performed in final volumes of 50 μl of 50 mM sodium citrate buffer (pH 7.0) containing 1% (w/v) polysaccharide substrate and 10 μg recombinant vGH5A. Following incubation at 30 °C for 60 min, the reaction mixture was heated at 100 °C for 5 min to stop the reaction. and stabilize the chromophore in its anionic 4-nitrophenolate form. The enzyme activity was determined colorimetrically at 540 nm by measuring the liberated glucose using a Synergy H1 microplate reader (BioTek, Winooski, USA). For all the enzymatic assays, the reaction mixtures without enzyme and with the heat-inactivated enzyme were used as the blank and the negative control, respectively. The same procedures were used for the functional characterization of another viral glycoside hydrolase vGH16C (MTVG_6627_10), except that vgh16C was cloned into the pMAL-c2x vector (Takara Bio, Dalian, China) using the ClonExpress II One Step Cloning Kit (Vazyme Biotech, Nanjing, China), and the recombinant plasmid was transformed into E. coli Rosseta (DE3) cells.
AlphaFold (v2.1)115 was used to predict the 3D structures of the AMGs-encoded proteins. Distance matrix alignment (DALI, online version, http://ekhidna2.biocenter.helsinki.fi/dali/)116 was used to calculate the structure similarity between AMG-encoded proteins and proteins in the Protein Data Bank (PDB) database117. The structure of vGH5A was visualized by PyMOL (v2.0)118.
Culture conditions and growth assays
All bacterial strains and plasmids used in this study are listed in Supplementary Table 1. The Shewanella strains were cultured in modified 2216E marine media (2216E) (5 g/l tryptone, 1 g/l yeast extract, 0.1 g/l FePO4, 34 g/l NaCl) and shook at 220 rpm at 4 °C and 15 °C; stainless steel pressure vessels were used for cultivation at high hydrostatic pressure. Escherichia coli strains were incubated in lysogeny broth (LB) media (10 g/l tryptone, 5 g/l yeast extract, 10 g/l NaCl) supplemented with 50 μg/ml DL-α, ε-diaminopimelic acid (DAP) at 37 °C. For solid media, agar-A (Bio Basic Inc., Ontario, Canada) was added at 1.5% (w/v). The antibiotic chloramphenicol (Cm) (Sigma, St. Louis, USA) was added to the media at final concentrations of 25 μg/ml and 12.5 μg/ml for E. coli and Shewanella, respectively, when needed. The growth of the Shewanella and E. coli strains was determined using turbidity measurements at 600 nm with a spectrophotometer (UV-2550, Shimadzu, Kyoto, Japan).
Construction of the SP1 deletion strain
The SP1 prophage deletion mutant was constructed by a recombination knock-out method19. Briefly, the upstream and downstream fragments flanking both ends of SP1 were amplified with PCR primer pairs (Supplementary Table 2) which were synthesized by Sangon Biotech (Shanghai, China). These two fragments were used as templates in a second fusion PCR, resulting in a fusion fragment flanking the boundary of SP1. Then, the PCR product was cloned into the suicide plasmid pRE112, yielding pRE112-SP1. This plasmid was transformed into E. coli WM3064 and then into WP2 by two-parent conjugation. The transconjugant was selected by chloramphenicol resistance and verified by PCR. The WP2 strain with pRE112-SP1 inserted into the chromosome was plated on 2216E agar medium supplemented with 10% sucrose. A successful prophage deletion mutant was screened for and confirmed by PCR and DNA sequencing.
Determination of the SP1 excision frequency
Quantitative PCR (qPCR) was used to determine the frequency of prophage excision under different conditions as previously described56. Specifically, the number of WP2 genomes was determined based on the single-copy reference gene rho. The number of genomes devoid of SP1 prophage was determined using primers (Sangon Biotech, Shanghai, China) flanking the prophage, which only produce PCR products when the prophage is removed.
RNA isolation and RT-qPCR
The S. psychrophila WP2 strains were inoculated into 2216E media, after which the culture was collected and immediately frozen in liquid nitrogen when the cells reached the exponential phase. Total RNA was isolated with a TRI reagent-RNA isolation kit (Molecular Research Center, Cincinnati, USA). The RNA samples were treated with DNase I at 37 °C for 1 h to remove DNA contamination, and then the purified RNA was reverse transcribed to complementary DNA (cDNA) by a RevertAid First Strand cDNA Synthesis Kit (Fermentas, Maryland, USA). The primer pairs used to amplify the selected genes for RT-qPCR were designed using Primer Express software (v3.0.1) (Applied Biosystems, CA, USA). PCR cycling was conducted using a StepOnePlus real-time PCR system (Thermo Fisher Scientific) in 20 μl reaction mixtures that included 1 × SYBR Green I Universal PCR Master Mix (Thermo Fisher Scientific), 0.5 μM of each primer, and 1 μl of cDNA template.
Transcriptome analysis
Strand-specific transcriptome sequencing was performed at Magigene Biotechnology Co., Ltd. (Guangdong, China). First, ribosomal RNA (rRNA) was removed using an Epicentre Ribo-Zero rRNA Removal Kit (Epicentre, Madison, WI, USA), and a cDNA library was prepared with a NEBNext Ultra II Directional RNA Library Prep Kit for Illumina (NEB, Ipswich, MA, USA) according to the manufacturer’s instructions119. The initial quantification of the library was carried out using a Qubit Fluorometer (Life Technologies, Carlsbad, CA, USA), and the insertion fragment size of the library was determined with an Agilent 2100 Bioanalyser (Agilent Technologies, Palo Alto, CA, USA). The effective concentration of the library was accurately quantified via qPCR (effective concentration > 2 nM). The different libraries were pooled together in a flow cell according to the effective concentration and the target offline data volume. After clustering, the Illumina HiSeq sequencing platform (Illumina, San Diego, USA) was used for paired-end sequencing. The raw data were filtered and evaluated by fastp software (v0.19.7)120, after which the clean reads were mapped to the S. psychrophila WP2 genome (NZ_CP014782.1) by HISAT software (v2.1.0)121. RSEM (v1.3.1)122 was used to calculate the read counts per sample, and the sequencing results were evaluated in terms of quality, alignment, saturation, and distribution of reads on the reference genome by DEGseq (v1.36.0)123. Gene expression was calculated by the number of reads mapped to each gene using the fragments per kilobase per million mapped reads (FPKM) method124 and analysed by edgeR (v3.20.2)125. The DEGs were identified according to the following standards: a false discovery rate (FDR) <0.05 and an FPKM fold change ≥2 between two samples. For each strain, three biologically independent samples were used for the RNA-seq analysis. The transcriptomic data were validated via RT-qPCR analysis, and a high correlation coefficient (R2 = 0.9593) was revealed (Supplementary Fig. 29), indicating that the transcriptomic data were reliable and could be used for follow-up analysis.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
All assembled prokaryotic genomes (n = 12,080) used in this study were collected from publicly available databases including the NCBI GenBank (v244)28, GEM29 and OTMG30. The nucleotide sequences of marine temperate viral genomes (n = 12,918) were deposited in the Supplementary Data 21. The transcriptomic data from the current study was deposited in the NCBI SRA under project ID PRJNA1014337. Source data are provided with this paper.
Code availability
All custom scripts, network and phylogenetic tree files used in this study are available on Github at https://github.com/kaixinbei132/Marine-temperate-viruses and the Zenodo repository (https://doi.org/10.5281/zenodo.8289908).
References
Dance, A. The incredible diversity of viruses. Nature 595, 22–25 (2021).
Feiner, R. et al. A new perspective on lysogeny: prophages as active regulatory switches of bacteria. Nat. Rev. Microbiol. 13, 641–650 (2015).
Howard-Varona, C., Hargreaves, K. R., Abedon, S. T. & Sullivan, M. B. Lysogeny in nature: mechanisms, impact and ecology of temperate phages. ISME J. 11, 1511–1520 (2017).
Touchon, M., Bernheim, A. & Rocha, E. P. Genetic and life-history traits associated with the distribution of prophages in bacteria. ISME J. 10, 2744–2754 (2016).
Kim, M.-S. & Bae, J.-W. Lysogeny is prevalent and widely distributed in the murine gut microbiota. ISME J. 12, 1127–1141 (2018).
Jarett, J. K. et al. Insights into the dynamics between viruses and their hosts in a hot spring microbial mat. ISME J. 14, 2527–2541 (2020).
Paul, J. H. Prophages in marine bacteria: dangerous molecular time bombs or the key to survival in the seas? ISME J. 2, 579–589 (2008).
Tuttle, M. J. & Buchan, A. Lysogeny in the oceans: lessons from cultivated model systems and a reanalysis of its prevalence. Environ. Microbiol. 22, 4919–4933 (2020).
Jiang, S. C. & Paul, J. H. Seasonal and diel abundance of viruses and occurrence of lysogeny bacteriocinogeny in the marine environment. Mar. Ecol. Prog. Ser. 104, 163–172 (1994).
Mizuno, C. M., Ghai, R., Saghaï, A., López-García, P. & Rodriguez-Valera, F. Genomes of abundant and widespread viruses from the deep ocean. mBio 7, e00805–e00816 (2016).
Morris, R. M., Cain, K. R., Hvorecny, K. L. & Kollman, J. M. Lysogenic host-virus interactions in SAR11 marine bacteria. Nat. Microbiol. 5, 1011–1015 (2020).
McDaniel, L., Houchin, L. A., Williamson, S. J. & Paul, J. H. Lysogeny in marine Synechococcus. Nature 415, 496 (2002).
Bondy-Denomy, J. & Davidson, A. R. When a virus is not a parasite: the beneficial effects of prophages on bacterial fitness. J. Microbiol. 52, 235–242 (2014).
Harrison, E. & Brockhurst, M. A. Ecological and evolutionary benefits of temperate phage: what does or doesn’t kill you makes you stronger. Bioessays. 39, 201700112 (2017).
Wang, X., Kim, Y. & Wood, T. K. Control and benefits of CP4-57 prophage excision in Escherichia coli biofilms. ISME J. 3, 1164–1179 (2009).
Wang, X. et al. Cryptic prophages help bacteria cope with adverse environments. Nat. Commun. 1, 147 (2010).
Liu, X. et al. Symbiosis of a P2-family phage and deep-sea Shewanella putrefaciens. Environ. Microbiol. 21, 4212–4232 (2019).
Yu, Z. et al. Filamentous phages prevalent in Pseudoalteromonas spp. confer properties advantageous to host survival in Arctic sea ice. ISME J. 9, 871–881 (2015).
Jian, H., Xiao, X. & Wang, F. Role of filamentous phage SW1 in regulating the lateral flagella of Shewanella piezotolerans strain WP3 at low temperatures. Appl. Environ. Microbiol. 79, 7101–7109 (2013).
Jian, H., Xiong, L., Xu, G. & Xiao, X. Filamentous phage SW1 is active and influences the transcriptome of the host at high-pressure and low-temperature. Environ. Microbiol. Rep. 8, 358–362 (2016).
Mann, N. H., Cook, A., Millard, A., Bailey, S. & Clokie, M. Bacterial photosynthesis genes in a virus. Nature 424, 741 (2003).
Hurwitz, B. L. & U’Ren, J. M. Viral metabolic reprogramming in marine ecosystems. Curr. Opin. Microbiol. 31, 161–168 (2016).
Warwick-Dugdale, J., Buchholz, H. H., Allen, M. J. & Temperton, B. Host-hijacking and planktonic piracy: how phages command the microbial high seas. Virol. J. 16, 15 (2019).
Hurwitz, B. L. & Sullivan, M. B. The Pacific Ocean Virome (POV): a marine viral metagenomic dataset and associated protein clusters for quantitative viral ecology. PLoS One 8, e57355 (2013).
Roux, S. et al. Ecogenomics and potential biogeochemical impacts of globally abundant ocean viruses. Nature 537, 689–693 (2016).
Anantharaman, K. et al. Sulfur oxidation genes in diverse deep-sea viruses. Science 344, 757–760 (2014).
Huang, X., Jiao, N. & Zhang, R. The genomic content and context of auxiliary metabolic genes in roseophages. Environ. Microbiol. 23, 3743–3757 (2021).
Sayers, E. W. et al. GenBank. Nucleic Acids Res. 49, D92–D96 (2021).
Nayfach, S. et al. A genomic catalog of Earth’s microbiomes. Nat. Biotechnol. 39, 499–509 (2021).
Jian, H. et al. Diversity and distribution of viruses inhabiting the deepest ocean on Earth. ISME J. 15, 3094–3110 (2021).
Roux, S. et al. IMG/VR v3: an integrated ecological and evolutionary framework for interrogating genomes of uncultivated viruses. Nucleic Acids Res. 49, D764–D775 (2021).
Al-Shayeb, B. et al. Clades of huge phages from across earth’s ecosystems. Nature 578, 425–431 (2020).
Luo, E., Aylward, F. O., Mende, D. R. & DeLong, E. F. Bacteriophage distributions and temporal variability in the ocean’s interior. mBio 8, e01903–e01917 (2017).
Williamson, S. J. et al. Lysogenic virus–host interactions predominate at deep-sea diffuse-flow hydrothermal vents. ISME J. 2, 1112–1121 (2008).
Howard-Varona, C. et al. Phage-specific metabolic reprogramming of virocells. ISME J. 14, 881–895 (2020).
Mende, D. R. et al. Environmental drivers of a microbial genomic transition zone in the ocean’s interior. Nat. Microbiol. 2, 1367–1373 (2017).
Zhao, Y. et al. Abundant SAR11 viruses in the ocean. Nature 494, 357–360 (2013).
Zhou, Z., Pan, J., Wang, F., Gu, J. & Li, M. Bathyarchaeota: globally distributed metabolic generalists in anoxic environments. FEMS Microbiol. Rev. 42, 639–655 (2018).
Hoshino, T. et al. Global diversity of microbial communities in marine sediment. Proc. Natl Acad. Sci. USA 117, 27587–27597 (2020).
Nishimura, Y. & Yoshizawa, S. The OceanDNA MAG catalog contains over 50,000 prokaryotic genomes originated from various marine environments. Sci. Data 9, 305 (2022).
Moraru, C., Varsani, A. & Kropinski, A. M. VIRIDIC-A novel tool to calculate the intergenomic similarities of prokaryote-infecting viruses. Viruses 12, 1268 (2020).
Canchaya, C., Fournous, G., Chibani-Chennoufi, S., Dillmann, M.-L. & Brüssow, H. Phage as agents of lateral gene transfer. Curr. Opin. Microbiol. 6, 417–424 (2003).
Touchon, M., Sousa, J. A. M. D. & Rocha, E. P. Embracing the enemy: the diversification of microbial gene repertoires by phage-mediated horizontal gene transfer. Curr. Opin. Microbiol. 38, 66–73 (2017).
Malki, K. et al. Bacteriophages isolated from Lake Michigan demonstrate broad host-range across several bacterial phyla. Virol. J. 12, 164 (2015).
Paez-Espino, D. et al. Uncovering earth’s virome. Nature 536, 425–430 (2016).
Jonge, P. A. D., Nobrega, F. L., Brouns, S. J. J. & Dutilh, B. E. Molecular and evolutionary determinants of bacteriophage host range. Trends Microbiol. 27, 51–63 (2019).
Chiura, H. X. Generalized gene transfer by virus-like particles from marine bacteria. Aquat. Microb. Ecol. 13, 75–83 (1997).
Hwang, Y., Roux, S., Coclet, C., Krause, S. J. E. & Girguis, P. R. Viruses interact with hosts that span distantly related microbial domains in dense hydrothermal mats. Nat. Microbiol. 8, 946–957 (2023).
Gregory, A. C. et al. Marine DNA viral macro- and microdiversity from pole to pole. Cell 177, 1109–1123 (2019).
Zeng, Y.-c, Li, Y.-T., Gu, Y.-j & Zhang, S.-z Purification and characterization of a strictly specific β-D-fucosidase from aspergillus phoenicis. Arch. Biochem. Biophys. 298, 226–230 (1992).
Larsbrink, J. et al. A complex gene locus enables xyloglucan utilization in the model saprophyte Cellvibrio japonicus. Mol. Microbiol. 94, 418–433 (2014).
Lafond, M. et al. The quaternary structure of a glycoside hydrolase dictates specificity toward β-Glucans. J. Biol. Chem. 291, 7183–7194 (2016).
Xiao, X., Wang, P., Zeng, X., Bartlett, D. H. & Wang, F. Shewanella psychrophila sp. nov. and Shewanella piezotolerans sp. nov., isolated from west Pacific deep-sea sediment. Int. J. Syst. Evol. Microbiol. 57, 60–65 (2007).
Xu, G., Jian, H., Xiao, X. & Wang, F. Complete genome sequence of Shewanella psychrophila WP2, a deep-sea bacterium isolated from west Pacific sediment. Mar. Genom. 35, 19–21 (2017).
Liu, X. et al. Physiological function of rac prophage during biofilm formation and regulation of rac excision in Escherichia coli K-12. Sci. Rep. 5, 16074 (2015).
Zeng, Z. et al. Cold adaptation regulated by cryptic prophage excision in Shewanella oneidensis. ISME J. 10, 2787–2800 (2016).
Li, Y. et al. Excisionase in Pf filamentous prophage controls lysis‐lysogeny decision‐making in Pseudomonas aeruginosa. Mol. Microbiol. 111, 495–513 (2019).
Rosenwasser, S., Ziv, C., Creveld, S. G. V. & Vardi, A. Virocell metabolism: metabolic innovations during host-virus interactions in the ocean. Trends Microbiol. 24, 821–832 (2016).
Forterre, P. The virocell concept and environmental microbiology. ISME J. 7, 233–236 (2013).
Fredrickson, J. K. et al. Towards environmental systems biology of Shewanella. Nat. Rev. Microbiol. 6, 592–603 (2008).
Lemaire, O. N., Méjean, V. & Iobbi-Nivol, C. The Shewanella genus: ubiquitous organisms sustaining and preserving aquatic ecosystems. FEMS Microbiol. Rev. 44, 155–170 (2020).
Brum, J. R. et al. Patterns and ecological drivers of ocean viral communities. Science 348, 1261498 (2015).
Kaestli, M. et al. Spatial and temporal microbial patterns in a tropical macrotidal estuary subject to urbanization. Front. Microbiol. 8, 1313 (2017).
Sipler, R. E. et al. Microbial community response to terrestrially derived dissolved organic matter in the coastal arctic. Front. Microbiol. 8, 1018 (2017).
Erazo, N. G. & Bowman, J. S. Sensitivity of the mangrove-estuarine microbial community to aquaculture effluent. iScience 24, 102204 (2021).
Caporaso, J. G., Knight, R. & Kelley, S. T. Host-associated and free-living phage communities differ profoundly in phylogenetic composition. PLoS One 6, e16900 (2011).
Raimundo, I. et al. Functional metagenomics reveals differential chitin degradation and utilization features across free-living and host-associated marine microbiomes. Microbiome 9, 43 (2021).
Kuwahara, H. et al. Reduced genome of the thioautotrophic intracellular symbiont in a deep-sea clam, Calyptogena okutanii. Curr. Biol. 17, 881–886 (2007).
Salerno, J. L., Bowen, B. W. & Rappe, M. S. Biogeography of planktonic and coral-associated microorganisms across the Hawaiian Archipelago. FEMS Microbiol. Ecol. 92, fiw109 (2016).
Parks, D. H., Imelfort, M., Skennerton, C. T., Hugenholtz, P. & Tyson, G. W. CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes. Genome Res. 25, 1043–1055 (2015).
MarcËais, G. et al. MUMmer4: A fast and versatile genome alignment system. PLoS Comput. Biol. 14, e1005944 (2018).
Pritchard, L. pyani v0.2.8: Average Nucleotide Identity (ANI) and Related Measures for Whole Genome Comparisons (San Francisco, CA, 2019).
Varghese, N. J. et al. Microbial species delineation using whole genome sequences. Nucleic Acids Res. 43, 6761–6771 (2015).
Chaumeil, P.-A., Mussig, A. J., Hugenholtz, P. & Parks, D. H. GTDB-Tk: a toolkit to classify genomes with the genome taxonomy database. Bioinformatics 36, 1925–1927 (2020).
Hyatt, D. et al. Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinform. 11, 119 (2010).
Potter, S. C. et al. HMMER web server: 2018 update. Nucleic Acids Res. 46, W200–W204 (2018).
Matsen, F. A., Kodner, R. B. & Armbrust, E. V. pplacer: linear time maximum-likelihood and bayesian phylogenetic placement of sequences onto a fixed reference tree. BMC Bioinform. 11, 538 (2010).
Price, M. N., Dehal, P. S. & Arkin, A. P. FastTree 2—approximately maximum-likelihood treesfor large alignments. PLoS One 5, e9490 (2010).
Letunic, I. & Bork, P. Interactive Tree Of Life (iTOL) v5: an online tool for phylogenetic tree display and annotation. Nucleic Acids Res. 49, W293–W296 (2021).
Guo, J. et al. VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses. Microbiome 9, 37 (2021).
Nayfach, S. et al. CheckV assesses the quality and completeness of metagenome-assembled viral genomes. Nat. Biotechnol. 39, 578–585 (2021).
Shaffer, M. et al. DRAM for distilling microbial metabolism to automate the curation of microbiome function. Nucleic Acids Res. 48, 8883–8900 (2020).
Guo, J., Vik, D., Pratama, A. A., Roux, S. & Sullivan, M. Viral Sequence Identification SOP with VirSorter2. https://doi.org/10.17504/protocols.io.bwm5pc86 (2021).
Roux, S., Hallam, S. J., Woyke, T. & Sullivan, M. B. Viral dark matter and virus-host interactions resolved from publicly available microbial genomes. eLife 4, e08490 (2015).
Roux, S. et al. Cryptic inoviruses revealed as pervasive in bacteria and archaea across Earth’s biomes. Nat. Microbiol. 4, 1895–1906 (2019).
Fu, L., Niu, B., Zhu, Z., Wu, S. & Li, W. CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics 28, 3150–3152 (2012).
Roux, S. et al. Minimum Information about an Uncultivated Virus Genome (MIUViG). Nat. Biotechnol. 37, 29–37 (2019).
Jang, H. B. et al. Taxonomic assignment of uncultivated prokaryotic virus genomes is enabled by gene-sharing networks. Nat. Biotechnol. 37, 632–639 (2019).
Buchfink, B., Xie, C. & Huson, D. H. Fast and sensitive protein alignment using DIAMOND. Nat. Methods 12, 59–60 (2015).
O’Leary, N. A. et al. Reference sequence (RefSeq) database at NCBI: current status, taxonomic expansion, and functional annotation. Nucleic Acids Res. 44, D733–D745 (2016).
Turner, D., Kropinski, A. M. & Adriaenssens, E. M. A roadmap for genome-based phage taxonomy. Viruses 13, 506 (2021).
Coutinho, F. H. et al. RaFAH: host prediction for viruses of bacteria and archaea based on protein content. Patterns 2, 100274 (2021).
Bland, C. et al. CRISPR recognition tool (CRT): a tool for automatic detection of clustered regularly interspaced palindromic repeats. BMC Bioinform. 8, 209 (2007).
Huntemann, M. et al. The standard operating procedure of the DOE-JGI Metagenome Annotation Pipeline (MAP v.4). Stand. Genom. Sci. 11, 17 (2016).
Nayfach, S. et al. Metagenomic compendium of 189,680 DNA viruses from the human gut microbiome. Nat. Microbiol. 6, 960–970 (2021).
Emerson, J. B. et al. Host-linked soil viral ecology along a permafrost thaw gradient. Nat. Microbiol. 3, 870–880 (2018).
Shannon, P. et al. Cytoscape a software environment for integrated models of biomolecular interaction networks. Genome Res. 13, 2498–2504 (2003).
Getz, E. W., Tithi, S. S., Zhang, L. & Aylward, F. O. Parallel evolution of genome streamlining and cellular bioenergetics across the marine radiation of a bacterial phylum. mBio 9, e01089–01018 (2018).
Weissman, J. L., Hou, S. & Fuhrman, J. A. Estimating maximal microbial growth rates from cultures, metagenomes, and single cells via codon usage patterns. PNAS 118, e2016810118 (2021).
Kanehisa, M., Furumichi, M., Sato, Y., Ishiguro-Watanabe, M. & Tanabe, M. KEGG: integrating viruses and cellular organisms. Nucleic Acids Res. 49, D545–D551 (2021).
Kirchman, D. L. Growth rates of microbes in the oceans. Annu. Rev. Mar. Sci. 8, 285–309 (2016).
Ahlgren, N. A., Ren, J., Lu, Y. Y., Fuhrman, J. A. & Sun, F. Alignment-free d2* oligonucleotide frequency dissimilarity measure improves prediction of hosts from metagenomically-derived viral sequences. Nucleic Acids Res. 45, 39–53 (2017).
Roux, S., Enault, F., Hurwitz, B. L. & Sullivan, M. B. VirSorter: mining viral signal from microbial genomic data. PeerJ 3, e985 (2015).
Groth, A. C. & Calos, M. P. Phage integrases: biology and applications. J. Mol. Biol. 335, 667–678 (2004).
Akhter, S., Aziz, R. K. & Edwards, R. A. PhiSpy: a novel algorithm for finding prophages in bacterial genomes that combines similarity- and composition-based strategies. Nucleic Acids Res. 40, e126–e126 (2012).
Tang, K. et al. Prophage tracer: precisely tracing prophages in prokaryotic genomes using overlapping split-read alignment. Nucleic Acids Res. 49, e128 (2021).
Kanehisa, M., Sato, Y. & Morishima, K. BlastKOALA and GhostKOALA: KEGG Tools for functional characterization of genome and metagenome sequences. J. Mol. Biol. 428, 726–731 (2016).
Chan, P. P., Lin, B. Y., Mak, A. J. & Lowe, T. M. tRNAscan-SE 2.0: improved detection and functional classification of transfer RNA genes. Nucleic Acids Res. 49, 9077–9096 (2021).
Wilcoxon, F. Individual comparisons of grouped data by ranking methods. J. Econ. Entomol. 39, 269–270 (1946).
Virtanen, P. et al. SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat. Methods 17, 261–272 (2020).
Benesty, J., Chen, J., Huang, Y. & Cohen, I. Noise Reduction in Speech Processing. Vol. 5 (Springer, 2009).
Zaragoza-Solas, A., Rodriguez-Valera, F. & López-Pérez, M. Metagenome mining reveals hidden genomic diversity of pelagimyophages in aquatic environments. mSystems 5, e00905–e00919 (2020).
Reichenbach, T. et al. Structural and biochemical characterization of the Cutibacterium acnes exo-β-1,4-mannosidase that targets the N-glycan core of host glycoproteins. PLOS ONE 13, e0204703 (2018).
Liberato, M. V. et al. Molecular characterization of a family 5 glycoside hydrolase suggests an induced-fit enzymatic mechanism. Sci. Rep. 6, 23473 (2016).
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Holm, L. in Structural Bioinformatics: Methods and Protocols (ed Gάspάri, Z.) Ch. 3 (Springer Science+Business Media, 2020).
Burley, S. K. et al. RCSB Protein data bank: powerful new tools for exploring 3D structures of biological macromolecules for basic and applied research and education in fundamental biology, biomedicine, biotechnology, bioengineering and energy sciences. Nucleic Acids Res. 49, D437–D451 (2021).
Schrodinger, LLC. The PyMOL Molecular Graphics System, Version 1.8. http://www.pymol.org. (2015).
Jian, H. et al. The origin and impeded dissemination of the DNA phosphorothioation system in prokaryotes. Nat. Commun. 12, 6382 (2021).
Chen, S., Zhou, Y., Chen, Y. & Gu, J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 34, i884–i890 (2018).
Kim, D., Langmead, B. & Salzberg, S. L. HISAT: a fast spliced aligner with low memory requirements. Nat. Methods 12, 357–360 (2015).
Li, B. & Dewey, C. N. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinform. 12, 323 (2011).
Wang, L., Feng, Z., Wang, X., Wang, X. & Zhang, X. DEGseq: an R package for identifying differentially expressed genes from RNA-seq data. Bioinformatics. 26, 136–138 (2010).
Trapnell, C. et al. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat. Biotechnol. 28, 511–515 (2010).
Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 26, 139–140 (2010).
Becker, R. A. & Wilks, A. R. (eds) AT&T Bell Laboratories Statistics Research Report (AT&T Bell Laboratories, 1995).
Acknowledgements
This work was financially supported by the National Natural Science Foundation of China (No. 42176095 to H.J. and 41921006 to X.X.), the National Key R&D Program of China (No. 2021YFF0501302 to H.J.), the Oceanic Interdisciplinary Program of Shanghai Jiao Tong University (No. SL2021PT201 to H.J.), and the Hainan Provincial Joint Project of Sanya Yazhou Bay Science and Technology City (No. 2021JJLH0057 to H.J.).
Author information
Authors and Affiliations
Contributions
H.J. conceived and designed the research; Y.Y., Q.S., S.T., and Q.T. collected and curated genomic and metadata; Y.Y. performed the bioinformatic and statistical analysis; S.L., Y.H., and X. Lei performed the microbiological experiments; YC.W., J.W., and M.Z. helped in statistical analysis; Y.Y. and H.J. analysed and interpreted the data; Y.Y. and H.J. wrote the manuscript; X.X., YZ.W., X. Liu, and Y.Z. provided useful comments to improve the manuscript; H.J. supervised the project. All the authors reviewed the results and approved the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Source data
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Yi, Y., Liu, S., Hao, Y. et al. A systematic analysis of marine lysogens and proviruses. Nat Commun 14, 6013 (2023). https://doi.org/10.1038/s41467-023-41699-4
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41467-023-41699-4
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
