CRISPR-Cas systems in the marine actinomycete Salinispora: linkages with phage defense, microdiversity and biogeography

Abstract

Background: Prokaryotic CRISPR-Cas systems confer resistance to viral infection and thus mediate bacteria-phage interactions. However, the distribution and functional diversity of CRISPRs among environmental bacteria remains largely unknown. Here, comparative genomics of 75 Salinispora strains provided insight into the diversity and distribution of CRISPR-Cas systems in a cosmopolitan marine actinomycete genus.

Results: CRISPRs were found in all Salinispora strains, with the majority containing multiple loci and different Cas array subtypes. Of the six subtypes identified, three have not been previously described. A lower prophage frequency in S. arenicola was associated with a higher fraction of spacers matching Salinispora prophages compared to S. tropica, suggesting differing defensive capacities between Salinispora species. The occurrence of related prophages in strains from distant locations, as well as spacers matching those prophages inserted throughout spacer arrays, indicate recurring encounters with widely distributed phages over time. Linkages of CRISPR features with Salinispora microdiversity pointed to subclade-specific contacts with mobile genetic elements (MGEs). This included lineage-specific spacer deletions or insertions, which may reflect weak selective pressures to maintain immunity or distinct temporal interactions with MGEs, respectively. Biogeographic patterns in spacer and prophage distributions support the concept that Salinispora spp. encounter localized MGEs. Moreover, the presence of spacers matching housekeeping genes suggests that CRISPRs may have functions outside of viral defense.

Conclusions: This study provides a comprehensive examination of CRISPR-Cas systems in a broadly distributed group of environmental bacteria. The ubiquity and diversity of CRISPRs in Salinispora suggests that CRISPR-mediated interactions with MGEs represent a major force in the ecology and evolution of this cosmopolitan marine actinomycete genus.

Figure 1

cas3 phylogeny and CRISPR repeat diversity. Condensed maximum likelihood phylogeny of cas3 nucleotide sequences reveals clades corresponding to Cas array subtype. The two major clades delineate known (I-E, I-C, I-B) and previously undescribed (I-U) subtypes. The order of genes for each subtype is displayed on the right. Gene annotations in parentheses designate hypothetical proteins with low identity to those indicated. In the I-B arrays, cas8b was replaced by a larger gene related to cst1. The total number of each array subtype among the 75 genomes is shown in the condensed nodes. Six arrays were missing several genes and therefore excluded from the analysis. Nodal support values (● above 80%, ▲ 100%) were obtained by 1000 bootstrap replicates (see Additional file 2 for the full tree including strain names and bootstrap values). Consensus repeat sequences in the associated CRISPR loci (indicated in gray) were specific to each array subtype and mostly showed considerable conservation.

Figure 2

Streptomyces SV1-related prophage content in Salinispora spp. A) Comparison of genes from six intact, SV1-related prophages in Salinispora genomes with the SV1 type phage (gray; with GenBank Gene IDs in parentheses) by color-coding percent sequence similarities over a 12400 bp region within SV1. B) History of SV1 encounters in S. tropica CNR-699 as reflected in multiple spacer matches in a single array. Incomplete matches are indicated by asterisks, perfect matches by encircled asterisks with sequences shown above.

Figure 3

Subclade-specific spacer content in S. arenicola phylotype B. Alignment of spacers from a homologous CRISPR locus present in all four S. arenicola phylotype B strains. The phylogenetic relationships among the four strains are depicted on the right. Spacers are indicated by rectangles arranged from the oldest (trailer end; right) to the most recent spacers (leader end; left). Vertically aligned spacers are identical and separated by subclade-specific groups of spacers (1, 2, 3), which were likely deleted from the respective other lineage or selectively acquired.

Figure 4

Biogeographic patterns among Salinispora spacers and prophages. A) Distribution of spacers unique to certain locations (SC = Sea of Cortez, HW = Hawaii, FJ = Fiji, PY = Palmyra, BH = Bahamas, PA = Palau) and biomes; B) Location-specific spacer variants reflected in single nucleotide polymorphisms (marked in red) within a conserved 41 nt spacer in strains from Hawaii (3 strains with 3 variants), Fiji (4 strains with 1 variant) and Palau (1 strains with 1 variant); C) Maximum likelihood phylogeny (1000 bootstrap replicates) of SV1-related prophages in genomes from geographically distant Salinispora strains, showing location-specific SV1 lineages.