A New Freshwater Cyanosiphovirus Harboring Integrase

Abstract

Pelagic cyanobacteria are key players in the functioning of aquatic ecosystems, and their viruses (cyanophages) potentially affect the abundance and composition of cyanobacterial communities. Yet, there are few well-described freshwater cyanophages relative to their marine counterparts, and in general, few cyanosiphoviruses (family Siphoviridae) have been characterized, limiting our understanding of the biology and the ecology of this prominent group of viruses. Here, we characterize S-LBS1, a freshwater siphovirus lytic to a phycoerythrin-rich Synechococcus isolate (Strain TCC793). S-LBS1 has a narrow host range, a burst size of ∼400 and a relatively long infecting step before cell lysis occurs. It has a dsDNA 34,641 bp genome with putative genes for structure, DNA packing, lysis, replication, host interactions, DNA repair and metabolism. S-LBS1 is similar in genome size, genome architecture, and gene content, to previously described marine siphoviruses also infecting PE-rich Synechococcus, e.g., S-CBS1 and S-CBS3. However, unlike other Synechococcus phages, S-LBS1 encodes an integrase, suggesting its ability to establish lysogenic relationships with its host. Sequence recruitment from viral metagenomic data showed that S-LBS1-like viruses are diversely present in a wide range of aquatic environments, emphasizing their potential importance in controlling and structuring Synechococcus populations. A comparative analysis with 16 available sequenced cyanosiphoviruses reveals the absence of core genes within the genomes, suggesting high degree of genetic variability in siphoviruses infecting cyanobacteria. It is likely that cyanosiphoviruses have evolved as distinct evolutionary lineages and that adaptive co-evolution occurred between these viruses and their hosts (i.e., Synechococcus, Prochlorococcus, Nodularia, and Acaryochloris), constituting an important driving force for such phage diversification.

Keywords: Synechococcus; cyanosiphovirus; freshwater; genome sequencing; lakes.

FIGURE 1

Transmission Electron Microscopy (TEM) images of Synechococcus phage S-LBS1, present as free particles outside of cells (A–C,E) or as immature progeny phage particles within the infected cell (D). The scale bar is 100 nm.

FIGURE 2

Genomic map of S-LBS1. Circles from outmost to innermost correspond to (i) predicted coding sequences on forward strand and (ii) reverse strand; (iii) defined genomic modules (structural and replication modules); and (iv) GC content plotted relative to the mean of G+C of genome at 60.2%. The black arrows show where the putative prophage attachment sites are located.

FIGURE 3

Schematic circos plot (A) displays line connections between homologous ORFs (BLASTp, acid amino identity >50%) in 16 cyanosiphoviruses (Supplementary Table S1). The lines are highlighted using orange for ORFs with homologs in S-LBS1, and light-blue if an ORF occurs in between two cyanosiphoviruses. Putative assigned functions of the ORFs are color-coded as indicated in the legend. To display similarity in gene content between two phages, a heatmap (B) was built on a matrix based on the percentage of genes shared between two phages (number of homologous genes shared between two phages/total genes of two phages). The dendogram on the left and on the top of the heatmap was clustered based on Bray–Curtis similarity. (C) Compares the genomes of five siphoviruses infecting Synechococcus sp. Pink panels connect homologous ORFs between two phages. The phylogenetic tree on the left was based on the alignment of DNA sequences of entire genomes using MAUVE (Darling et al., 2004). The color code is the same as in (A).

FIGURE 4

Unrooted Bayesian phylogenetic tree of 47 inferred amino-acid sequences of the terminase large subunit (terL), obtained from S-LBS1 and others phage representatives (Supplementary Table S2). Values shown at the nodes of the main branches are the Bayesian inference (BI) clade credibility and Maximum Likelihood (ML) bootstrap values, and are reported as BI/ML.

FIGURE 5

Prevalence of S-LBS1-like sequences in environmental viral metagenomic data. Fragment recruitment of reads from environmental viral metagenomic data (Supplementary Table S3) onto the genome of S-LBS1. Each horizontal line represents a read recruited from publicly available (A) Freshwater viral metagenomic datasets, or (B) marine viral metagenomic datasets. The length of each line represents the BLAST query coverage (percent of the query sequence that overlaps the subject sequence) and the y-axis position of each line indicates amino-acid identity (percent similarity) of the metagenomic reads to the S-LBS1 sequences over the length of the coverage area. The color codes for ORFs are as shown in Figure 3. The normalized total number of recruited reads for S-LBS1 (amino-acid identity >50%) and other phages across environments are shown in (C), respectively. The total number of hits to phage was normalized by dividing by the length of the phage genome (in kb) and the size of the database (number of reads recruited per kb of genome/size of the database in Gb), which provides a normalized measure to compare recruitments by different size contigs versus several metagenomic datasets.

FIGURE 6

RaxML phylogenetic tree of 107 inferred amino-acid sequences similar to Siphovirus-gp157 (pfam05565), obtained by blast against the NCBI nr reference protein database using Siphovirus-gp157 from S-LBS1 as the query. Points at the nodes show maximum-likelihood (ML) bootstrap values ranging from 0 (white) to 100 (black). Phylogenetic tree leaves were labeled by the name of the organism that contains this shiphovirus-gp157 sequence, followed by the NCBI accession number in parentheses. Phylogenetic tree leaves were also annotated by two columns of color strips; the first column distinguish the bacterial and viral shiphovirus-gp157 sequences; the second column differs the morphological types of viruses (myovirus, podovirus, siphovirus, and unclassified) in which the shiphovirus-gp157 sequences were found.