Genome replication dynamics of a bacteriophage and its satellite reveal strategies for parasitism and viral restriction

Abstract

Phage-inducible chromosomal island-like elements (PLEs) are bacteriophage satellites found in Vibrio cholerae. PLEs parasitize the lytic phage ICP1, excising from the bacterial chromosome, replicating, and mobilizing to new host cells following cell lysis. PLEs protect their host cell populations by completely restricting the production of ICP1 progeny. Previously, it was found that ICP1 replication was reduced during PLE(+) infection. Despite robust replication of the PLE genome, relatively few transducing units are produced. We investigated if PLE DNA replication itself is antagonistic to ICP1 replication. Here we identify key constituents of PLE replication and assess their role in interference of ICP1. PLE encodes a RepA_N initiation factor that is sufficient to drive replication from the PLE origin of replication during ICP1 infection. In contrast to previously characterized bacteriophage satellites, expression of the PLE initiation factor was not sufficient for PLE replication in the absence of phage. Replication of PLE was necessary for interference of ICP1 DNA replication, but replication of a minimalized PLE replicon was not sufficient for ICP1 DNA replication interference. Despite restoration of ICP1 DNA replication, non-replicating PLE remained broadly inhibitory against ICP1. These results suggest that PLE DNA replication is one of multiple mechanisms contributing to ICP1 restriction.

Figure 1.

Model of ICP1 infection in PLE(−) and PLE(+) V. cholerae. ICP1 injects its DNA into V. cholerae; prior to DNA replication, ICP1 activity leads to PLE activation and excision. ICP1 DNA replication is reduced in the PLE(+) cell where the PLE replicates to high copy. Finally, the cell lyses and releases infectious particles. No ICP1 particles and a low number of PLE transducing particles are released from the PLE(+) cell.

Figure 2.

PLE robustly replicates following infection while altering ICP1 replication. (A and B) Genomes per million (GPM) of total DNA mapping to the V. cholerae large (VC I) and small (VC II) chromosomes, ICP1, and the PLE across an infection time course in PLE(−) (A) and PLE(+) (B) V. cholerae. Samples were taken at 4, 8, 12 and 16 min post-infection, data show the average and standard deviation of three independent experiments. (C) Percent reads coverage plots across the ICP1 genome during PLE(−) (top) and PLE(+) (bottom) infection. For each time point, the percent reads coverage across the genome for three biological replicates was determined. The average percent reads coverage is shown as a black line, while standard deviation appears as dark gray shading around the line. The GC skew (right axis) is shown as light gray shading.

Figure 3.

A single PLE-encoded ORF and a noncoding region are necessary for PLE replication. (A) A representation of the PLE genome (top) with average reads coverage of PLE 8 min post ICP1 infection plotted below. The percent reads coverage was determined for three biological replicates and is shown as a black line, while standard deviation appears as dark gray shading around the line. The GC skew (right axis) is plotted as light gray shading. Gene clusters mutated for analysis are labelled. (B) PLE’s noncoding region (NCR) between orf5 and orf7, with repeat sequences shown as arrows. Repeat sequences share colors for each repeat type, and are designated as repeats 1, 2, 3 or 4. Regions of the NCR deleted for analysis in (C) are shown. Panels C–E: replication of PLE mutants 20 min post-infection with ICP1 as assessed by qPCR. Replication efficiency is relative to a wild-type PLE control. (C) Replication of ΔNCR mutants. (D) Replication of PLE gene cluster knockouts. (E) Replication of individual gene knockouts of the ORFs contained in cluster 7–14.

Figure 4.

RepA drives PLE replication in the presence of ICP1. (A) RepA complementation of PLE ΔrepA as assessed by qPCR. PLE fold copy increase 20 min post-infection is shown in different combinations with ICP1 and the inducer of the complementation construct. (B) A diagram of the midiPLE construct used to assess the minimal requirements for PLE replication (not to scale). Attachment sites, the PLE integrase, and the noncoding region (NCR) are present along with a kanamycin resistance gene (kanR). (C) Replication of a RepA complemented ΔrepA strain and midiPLE 20 min post ICP1 infection. The replication of these strains was compared to a wild-type PLE control, and the relative replication is displayed as a percentage above the bars.

Figure 5.

PLE RepA is a RepA_N family protein. (A) Cartoon representation of the PLE RepA-NTD dimer. Monomers are differently colored in yellow and white. The N and C termini of the monomers are colored blue and red respectively. (B) Alignment of the NTD dimers of PLE RepA and pTZ2162 RepA in light blue and dark grey, respectively, depicted in cartoon representations (RMSD = 4.527 over 176 residues). (C) Surface view of PLE RepA-NTD dimer in light blue aligned with pTZ2162 RepA-NTD dimer in dark gray bound to substrate DNA. (D) Electrostatic potential map, turned 90 degrees as (C), of PLE RepA-NTD dimer aligned to the pTZ2162 RepA-dsDNA bound structure. Positive (blue) and negative (red) charges are indicated on the surface. (E) Binding iterons for the RepA initiators of pTZ2162 and pSK41 are shown alongside repetitive sequences found in the putative PLE origin of replication. Direct repeats are denoted with an arrow, while the sequence comprising inverted sub-repeats is boxed. Sequence for the minus strand for PLE is shown to make the central poly-A tract apparent.

Figure 6.

Repeat 3 serves as the PLE ori iteron sequence. (A) The nucleotide sequences of 5′ fluorescently labelled dsDNA probes used to test RepA specificity. Probe sequence was derived from the repetitive sequences found in PLE noncoding region 3 (NCR3). Inverted sub-repeats are boxed. (B) An electrophoretic mobility shift assay using the probes in (A). The * denotes a RepA(+) no DNA control. Additional replicates of this experiment are shown in Supplementary Figure S9. (C) The replication of PLE mutants with deletions spanning the repeat 3 (R3) and repeat 4 (R4) regions 20 min post-infection with ICP1 as assessed by qPCR. Replication efficiency is relative to a wild-type PLE control.

Figure 7.

Loss of replication impairs PLE anti-phage activity. (A) ICP1 replication in wild-type and mutant PLE(+) strains as assessed by qPCR. Replication efficiency is relative to ICP1 infection of PLE(−) V. cholerae 20 min post-infection. (B) Replication of ICP1 as assessed by qPCR in RepA complemented midiPLE and ΔrepA PLE infection relative to an un-complemented midiPLE control. (C) Efficiency of center of infection (EOCI) for ICP1 on wild-type PLE and non-replicating PLE mutant hosts. ECOI is relative to a PLE(−) permissive control strain. The dashed line indicates the threshold at which the number of output phage is equal to the number of input phage. Above the dashed line output has a larger value, below the dashed line input has a larger value.

Figure 8.

Non-replicating PLE still alters ICP1’s replication profile. (A) Genomes per million of ICP1 in PLE(−), PLE(+), and PLE ΔrepA cultures at 8 and 16 min post-infection. Values shown are the means of three biological replicates. (B) Percent reads coverage profile of ICP1’s genome in ΔrepA PLE infection at 8 min (top) and 16 min (bottom) post-infection. (C) Percent reads coverage profile of ICP1’s genome in PLE (−), PLE (+), and ΔrepA PLE hosts at 16 min post infection. The ICP1 genome has been rotated so that it is centered around the putative rolling circle replication origin. For each reads profile plot in (B) and (C), the -average percent reads coverage across the genome for three biological replicates is shown as a black line, while standard deviation appears as dark gray shading around the line. The GC skew (right axis) is plotted as light gray shading.