Virus satellites are widespread subcellular entities, present both in eukaryotic and in prokaryotic cells. Their modus vivendi involves parasitism of the life cycle of their inducing helper viruses, which assures their transmission to a new host. However, the evolutionary and ecological implications of satellites on helper viruses remain unclear. Here, using staphylococcal pathogenicity islands (SaPIs) as a model of virus satellites, we experimentally show that helper viruses rapidly evolve resistance to their virus satellites, preventing SaPI proliferation, and SaPIs in turn can readily evolve to overcome phage resistance. Genomic analyses of both these experimentally evolved strains as well as naturally occurring bacteriophages suggest that the SaPIs drive the coexistence of multiple alleles of the phage-coded SaPI inducing genes, as well as sometimes selecting for the absence of the SaPI depressing genes. We report similar (accidental) evolution of resistance to SaPIs in laboratory phages used for Staphylococcus aureus typing and also obtain the same qualitative results in both experimental evolution and phylogenetic studies of Enterococcus faecalis phages and their satellites viruses. In summary, our results suggest that helper and satellite viruses undergo rapid coevolution, which is likely to play a key role in the evolution and ecology of the viruses as well as their prokaryotic hosts. Satellites are defined as viruses that have a life cycle dependent on a helper virus. Thus, they can be considered as parasites of parasites. In addition to their fascinating life cycle, these widespread infectious elements, present both in eukaryotic and prokaryotic cells, have a dramatic role in virulence by controlling the symptoms induced by their eukaryotic helper viruses or by encoding key bacterial virulence genes. While satellites can play an important role in the ecology of the viruses they parasitise, the evolutionary impact on their helper viruses is unclear. Here we show that staphylococcal pathogenicity islands (SaPIs), an example of a virus satellite, are a major selective force on the viruses (bacteriophages) they parasitise. Using both bioinformatic and experimental evolution data we have been able to confirm that pathogenicity islands are a major selective pressure enhancing the diversity of both genes and gene content in Staphylococcus aureus phages. Since SaPIs exploit the life cycle of their helper phages to enable their rapid replication and promiscuous spread, these strategies are mechanisms that reduce SaPI interference, thus facilitating the infectivity and dissemination of the helper phages in nature.

Summary: Staphylococcus aureus pathogenicity islands (SaPIs) are a group of related 15–17 kb mobile genetic elements that commonly carry genes for superantigen toxins and other virulence factors. The key feature of their mobility is the induction of SaPI excision and replication by certain phages and their efficient encapsidation into specific small‐headed phage‐like infectious particles. Previous work demonstrated that chromosomal integration depends on the SaPI‐encoded recombinase, Int. However, although involved in the process, Int alone was not sufficient to mediate efficient SaPI excision from chromosomal sites, and we expected that SaPI excision would involve an Xis function, which could be encoded by a helper phage or by the SaPI, itself. Here we report that the latter is the case. In vivo recombination assays with plasmids in Escherichia coli demonstrate that SaPI‐coded Xis is absolutely required for recombination between the SaPI att_L and att_R sites, and that both sites, as well as their flanking SaPI sequences, are required for SaPI excision. Mutational analysis reveals that Xis is essential for efficient horizontal SaPI transfer to a recipient strain. Finally, we show that the master regulator of the SaPI life cycle, Stl, blocks expression of int and xis by binding to inverted repeats present in the promoter region, thus controlling SaPI excision.

Targeting conserved and essential processes is a successful strategy to combat enemies. Remarkably, the clinically important Staphylococcus aureus pathogenicity islands (SaPIs) use this tactic to spread in nature. SaPIs reside passively in the host chromosome, under the control of the SaPI-encoded master repressor, Stl. It has been assumed that SaPI de-repression is effected by specific phage proteins that bind to Stl, initiating the SaPI cycle. Different SaPIs encode different Stl repressors, so each targets a specific phage protein for its de-repression. Broadening this narrow vision, we report here that SaPIs ensure their promiscuous transfer by targeting conserved phage mechanisms. This is accomplished because the SaPI Stl repressors have acquired different domains to interact with unrelated proteins, encoded by different phages, but in all cases performing the same conserved function. This elegant strategy allows intra- and inter-generic SaPI transfer, highlighting these elements as one of nature’s most fascinating subcellular parasites.