Thirty-three Pa strains (isolated at the Scottish Crop Research Institute, Dundee, UK) were screened for the presence of ECA29 and ECA41 by duplex PCR (Fig. 1). Primers were designed to amplify an internal region of the prophage, and the presence of a product does not necessarily indicate that the complete prophage is present. Three strains were found to harbour both prophages, and ECA41 alone was present in a further four. In NVP-AUY922 mw none of the strains tested could ECA29 be found alone. These results suggest that acquisition of the prophages occurred relatively recently, because only a fraction of the strains carry them, and that ECA41 may be more ancestral. Upon excision from the bacterial

genome, prophage DNA circularizes to form a replication-competent intermediate. To detect such molecules in Pa, outward-reading primers at each end of the prophages were designed. No circularization product was obtained in the case of ECA29. In contrast, primers specific to ECA41 did yield an amplicon across

a circularization junction. The PCR products were ligated into pBlueScript and 12 unique clones were sequenced. Surprisingly, additional DNA sequences of bacterial origin could be detected between the prophage termini. These insertion sequences ranged in length from 6 to 179 bp, and originated from diverse locations within the genome (Table 2). Unexpectedly, DNA from one region of ECA29 was captured by ECA41 excision in three independent clones. Sequence comparison of the two prophages

demonstrated that this locus was found within a region Y-27632 clinical trial of 1.4 kb located in the centre of ECA29, which showed Megestrol Acetate high similarity to a region at the 3′ terminus of ECA41. This region, displaying 79% identity at the nucleotide level between the two prophages, encompasses the genes ECA2607 and the 3′-end of ECA2608 in ECA29, and ECA3742 and the 3′-end of ECA3741 in ECA41. ECA2607 and ECA2608 are predicted to encode components of the phage tail fibres – genes that are known to be frequent sites of recombination (Sandmeler, 1994). This suggests that random transposition of ECA41 can be enhanced by site-specific recombination. This is in contrast to transposition by the well-studied phage Mu, which has been described as entirely random, although hot spots for integration may exist (Paolozzi & Ghelardini, 2006). In some cases, excised bacterial DNA originated from the right-hand side of the ECA41 phage in its annotated location, demonstrating that imprecise excision occurs at this end. Excision at the left-hand end was precise and was always found to occur at base 4144429. The data presented above demonstrate that ECA41 is able to excise from the genome while simultaneously excising bacterial DNA originating from a variety of genomic locations. This suggests that ECA41 transposes from its original position to random target sites before excision. These features are reminiscent of phage Mu and suggest that ECA41 could be mutagenic.