Phages and their satellites encode hotspots of antiviral systems

Abstract

Bacteria carry diverse genetic systems to defend against viral infection, some of which are found within prophages where they inhibit competing viruses. Phage satellites pose additional pressures on phages by hijacking key viral elements to their own benefit. Here, we show that E. coli P2-like phages and their parasitic P4-like satellites carry hotspots of genetic variation containing reservoirs of anti-phage systems. We validate the activity of diverse systems and describe PARIS, an abortive infection system triggered by a phage-encoded anti-restriction protein. Antiviral hotspots participate in inter-viral competition and shape dynamics between the bacterial host, P2-like phages, and P4-like satellites. Notably, the anti-phage activity of satellites can benefit the helper phage during competition with virulent phages, turning a parasitic relationship into a mutualistic one. Anti-phage hotspots are present across distant species and constitute a substantial source of systems that participate in the competition between mobile genetic elements.

Keywords: abortive infection; bacterial immunity; genetic diversity; inter-viral competition; mobile genetic elements; phage defense; phage satellite.

Graphical abstract

Figure 1

A diversity of genetic systems encoded on P4-like phages in E. coli (A) Genomic visualization of the P4 reference genome and P4-like satellites in five E. coli strains, highlighting genetic diversity between psu and int genes, including known anti-phage systems. Genome accession numbers and positions are shown on the left. The DNA sequence of the cos-proximal region in these strains is highlighted with conserved sequences in blue and variable sequences in red. (B) Systematic analysis of genetic systems encoded between psu and int identified in 26% (5,251/20,125) of analyzed E. coli genomes. The pie chart shows the proportion of loci encoding each of the 30 most abundant systems (Table S1B), shown as gene cassettes colored by protein domains. Validated systems providing phage defense (Figure 2A) are highlighted with a dashed linker. The system from the P4 reference genome is highlighted with a gray background. When a system comprises accessory genes, different variants are shown with the percentage of each occurrence. RT, reverse-transcriptase; REase, restriction-endonuclease; RM, restriction-modification; HEPN, higher eukaryotes and prokaryotes nucleotide-binding; AIPR, abortive infection phage resistance; TIR, Toll/interleukin-1 receptor; HAD, haloacid dehydrogenase-like; SIR2, sirtuin; DUF, domain of unknown function; TM, transmembrane domain. See also Figures S1 and S3 and Tables S1A and S1B.

Figure 2

P4-encoded hotspots encode a variety of anti-phage systems Phage resistance heatmap of the validated defense systems shows the mean fold resistance of three independent replicates against a panel of eight phages (key resources table). Genes are colored by protein family. Genome accession numbers are provided in Table S2A. Systems are under the control of their native promoters, with the exception of PARIS-1, which was only active when expressed from a ptet promoter (pFD237) in the presence of anhydrotetracycline (aTc, 0,5 μg/mL). Defense was measured at 37°C, with the exception of the RT-nitrilase + 1TM system, which was measured at room temperature. RT, reverse-transcriptase; TIR, Toll/interleukin-1 receptor; HAD, haloacid dehydrogenase-like; DUF, domain of unknown function; TM, transmembrane helix. See also key resources table, Figure S2, and Tables S2A and S3.

Figure 3

PARIS is triggered by a phage-encoded anti-restriction protein (A) Time course experiment with cells harboring a control plasmid or a PARIS-encoding plasmid. Cells were kept uninfected or were infected with T7 at a high or low multiplicity of infection (MOI) once cells reached OD ∼0.2. Each curve shows the mean of three technical replicates with the standard deviation shown as a transparent area. (B) Serial dilutions of a high titer lysate of T7 or T7^OcrF54V spotted on MG1655. (C) Transformation efficiency of a plasmid expressing a green fluorescent protein (GFP) or T7 Ocr protein. Bars show the mean of four independent replicates. The p value of a two-sided Mann-Whitney test is shown. (D) Transformation efficiency of methylated or unmethylated plasmid DNA in MG1655 cells expressing a GFP, wild-type Ocr, or Ocr^F54V. Barplot shows the mean of three independent replicates shown as black dots. (E) Serial dilutions of a high titer lysate of T7or T7^Ocr(ΔT35) spotted on DH10B (top) or on MG1655, expressing an aTc-inducible dCas9 and a sgRNA targeting EcoKI (bottom), representative of three independent replicates. (F) Current model for the defense activity of PARIS. See also Figure S4 and Tables S4A and S4B.

Figure 4

A diversity of genetic systems encoded on P2-like phages in E. coli (A) Visualization of genomic regions from five E. coli strains containing a P2-like prophage, highlighting genetic diversity between Q and A genes, including known anti-phage defense systems. Genome accession numbers and positions are shown on the left. (B) Systematic analysis of genetic systems encoded between gpA and gpQ from P2-like phages. The pie chart shows the 30 most abundant systems classified by prevalence and shown as gene cassettes colored by protein domains (not to scale). When a system comprises accessory genes, different variants are shown, with the percentage of each occurrence on the left. A validated system providing phage defense (Figure 5A) is highlighted with a dashed linker. NLR, NOD-like receptor; SIR2, sirtuin; DUF, domain of unknown function; TM, transmembrane helix; SMC, structural maintenance of chromosome; REase, restriction-endonuclease. See also Figure S1 and Tables S5A and S5B.

Figure 5

P2-encoded hotspot includes diverse anti-phage systems (A) Phage resistance heatmaps of the validated defense systems show the median fold resistance of three independent replicates against a panel of eight phages (key resources table). Genes are colored by protein family. Genome accession numbers are provided in Table S2A. HAD, haloacid dehydrogenase-like; HP, hypothetical protein; DUF, domain of unknown function. (B) Lysogenization of E. coli C with P2-like phage AC1 protects against phage λ and LF82_P8. (C) Description of the system found in P2-like phage AC1. TIR, Toll/interleukin-1 receptor; TM, transmembrane helix. (D) The candidate defense system from phage AC1 was cloned and introduced into E. coli C. The cloned system recapitulates the defense phenotype of the lysogen. (E) The AC1-encoded system provides protection against P2-like relatives. See also Figure S5.

Figure 6

Antiviral hotspots mediate inter-viral competition (A) Defense activity of the TIR-NLR system against five P2-like phages. Phage dilutions were spotted on a bacterial lawn of E. coli C encoding the TIR-NLR system or a GFP control. Representative of three independent replicates. (B) The TIR-NLR system affects both P2 and P420 propagation. E. coli C cells carrying the P420 plasmid, a kanamycin resistant variant of P4 (Kahn and Helinski, 1978), were infected by P2 in the presence of the TIR-NLR system or a GFP control. Titers of P2 and P420 were measured in the lysate (STAR Methods). (C) PARIS favors P2 during co-infection with LF82_P8. Titers of P2 and LF82_P8 were measured after co-infection at a 1:1 ratio and MOI ∼ 0.01 of E. coli C cells expressing PARIS or a GFP control (left) (STAR Methods). In parallel, titers of P2 and LF82_P8 were also measured after infection by a single phage (right). (D) Colony-forming units were measured after single or co-infection by P2 and/or LF82_P8 in E. coli C cells expressing PARIS or a GFP control. Bar plots show the mean of three independent replicates, each shown as a black dot. See also Figure S6.

Figure 7

Hotspots for anti-phage systems encoded on other prophage genomes (A and B) Genomic view of hotspots encoded on prophages from Vibrionales (A) and Bacilliales (B). Phage genes are shown with different shades of blue. Gray shades show the percentage of identity between homologous proteins from different genomes. Genome accession numbers and positions are shown on the left. See also Table S6.