Gene network visualization and quantitative synteny analysis of more than 300 marine T4-like phage scaffolds from the GOS metagenome

Abstract

Bacteriophages (phages) are the most abundant biological entities in the biosphere and are the dominant "organisms" in marine environments, exerting an enormous influence on marine microbial populations. Metagenomic projects, such as the Global Ocean Sampling expedition (GOS), have demonstrated the predominance of tailed phages (Caudovirales), particularly T4 superfamily cyanophages (Cyano-T4s), in the marine milieu. Whereas previous metagenomic analyses were limited to gene content information, here we present a comparative analysis of over 300 phage scaffolds assembled from the viral fraction of the GOS data. This assembly permits the examination of synteny (organization) of the genes on the scaffolds and their comparison with the genome sequences from cultured Cyano-T4s. We employ comparative genomics and a novel usage of network visualization software to show that the scaffold phylogenies are similar to those of the traditional marker genes they contain. Importantly, these uncultured metagenomic scaffolds quite closely match the organization of the "core genome" of the known Cyano-T4s. This indicates that the current view of genome architecture in the Cyano-T4s is not seriously biased by being based on a small number of cultured phages, and we can be confident that they accurately reflect the diverse population of such viruses in marine surface waters.

FIG. 1.

Synteny representations and I_synteny calculations. (A) Traditional arrow gene representations, with “core genes” in black and inserted novel ORFs in white. (B) Conversion of (A) into network representation (used by Cytoscape), with each line representing an occurrence/link between the respective genes/ORFs. (C) Reduction of (B), with the removal of non-core genes, showing either all (top) or a condensation (bottom) of the number of links. (D) Formulation of the Index of Synteny (I_synteny) (equation on right), which reports the proportion of links (L) to the left (in) and right (out) of a gene X that are to single sources (S) and targets (T). (E–G) Various examples of gene synteny, along with the corresponding I_synteny values and the total number of links (under the format L = n).

FIG. 2.

GOS myovirus scaffold content. Taxonomic distribution of the phage (A) and eubacterial hits (B).

FIG. 3.

GOS scaffolds containing the g23 MCP gene. Schematic of the scaffolds containing g23, with each gene/ORF (rectangles) colored to match its origin type, compared with the cultured Cyano-T4 genomes in this region. Those g23s that do not represent the majority origin type on the scaffolds or genomes are marked with an “X.” Due to space considerations, some of the genes/ORFs of known origin have been condensed (e.g., “x3”), and some ORFans have been omitted. The phylogenetic tree inset shows the relationships among the Cyano-T4 gp23 proteins, with T4 as the outgroup and the scale bar indicating 0.1 substitutions per site.

FIG. 4.

Cultured Cyano-T4 and GOS scaffold “core genes.” The 41 core genes of the cultured Cyano-T4 genomes (A) and the scaffolds (B) are presented, along with the I_synteny values for those genes <0.75. There are 162 distinct occurrences for the Cyano-T4s and 549 occurrences for the scaffolds. Note that the actual number is 609 for the latter due to large mate-pair gaps that break some core gene links (n = 18) and because scaffolds containing only a single core gene (n = 42) are not included (no links to other core genes).

FIG. 5.

All GOS scaffold genes/ORFs. All genes/ORFs on the scaffolds are presented, organized by their various origin types, where n is the number of genes and L the number of links to neighboring genes for each type.