Pneumococcal prophages are diverse, but not without structure or history

Abstract

Bacteriophages (phages) infect many bacterial species, but little is known about the diversity of phages among the pneumococcus, a leading global pathogen. The objectives of this study were to determine the prevalence, diversity and molecular epidemiology of prophages (phage DNA integrated within the bacterial genome) among pneumococci isolated over the past 90 years. Nearly 500 pneumococcal genomes were investigated and RNA sequencing was used to explore prophage gene expression. We revealed that every pneumococcal genome contained prophage DNA. 286 full-length/putatively full-length pneumococcal prophages were identified, of which 163 have not previously been reported. Full-length prophages clustered into four major groups and every group dated from the 1930-40 s onward. There was limited evidence for genes shared between prophage clusters. Prophages typically integrated in one of five different sites within the pneumococcal genome. 72% of prophages possessed the virulence genes pblA and/or pblB. Individual prophages and the host pneumococcal genetic lineage were strongly associated and some prophages persisted for many decades. RNA sequencing provided clear evidence of prophage gene expression. Overall, pneumococcal prophages were highly prevalent, demonstrated a structured population, possessed genes associated with virulence, and were expressed under experimental conditions. Pneumococcal prophages are likely to play a more important role in pneumococcal biology and evolution than previously recognised.

Figure 1. Graph depicting the number of prophages in each category: partial, putatively full-length or full-length.

Bars represent the number of prophages relative to the length of the prophage sequence that was identified.

Figure 2. Description of nucleotide sequence similarity, phylogenetic clustering and shared genes among the 66 representative full-length prophages identified in the pneumococcal genome dataset.

(a) Heat map depicting the percentage nucleotide sequence identity shared between pairs of full-length prophage sequences. Groups of similar prophages are marked A-E and correspond to the clusters seen in part b. (b) Clusters of prophage sequences based upon nucleotide sequence identity. Bootstrap values are marked on branches of each cluster. (c) Venn diagram depicting genes unique to each prophage cluster and genes shared between clusters.

Figure 3. Nucleotide sequence alignments of representative full-length prophages from clusters A and B.

The coloured bar at the top of each cluster indicates the mean pairwise nucleotide sequence identity over all pairs in the column: bright green = 100% identity; green-brown = <100% but >30% identity; and red = <30% identity. Prophage genes are coloured based on putative or known function. The names of each prophage are given, followed by the year of isolation (in brackets) of the oldest known pneumococcus harbouring that prophage.

Figure 4. Nucleotide sequence alignments of representative full-length prophages from clusters C, D and E.

The coloured bar at the top of each cluster indicates the mean pairwise nucleotide sequence identity over all pairs in the column: bright green = 100% identity; green-brown = <100% but >30% identity; and red = <30% identity. Prophage genes are coloured based on putative or known function. The names of each prophage are given, followed by the year of isolation (in brackets) of the oldest known pneumococcus harbouring that prophage.

Figure 5. Illustration of the number of different full-length and putatively full-length prophages identified in each of the major pneumococcal clonal complexes.

Clonal complexes are labelled on the y-axis followed by brackets containing the number of pneumococcal genomes within that complex and the years of isolation of those pneumococci. See Table 2 and S2 for details of the specific prophages identified within each clonal complex.

Figure 6. Heat maps describing the results of the RNA-seq experiment.

Prophage genes are depicted by rows and differential expression levels at each of five time points are presented in columns. An asterisk to the left of a cell indicates a statistically significant differential level of expression (p < 0.05). Prophage gene expression levels are given for two full-length prophages, Sp195_1 (a) and Sp195_2 (b), and one partial prophage sequence, IPPX439 (c). Mitomycin C was added to the broth culture after 3 h of incubation.