Integration of Integrons in res Sites

In a recent paper on the integrons related to In4, Partridge and colleagues (4) describe an interesting reconstruction, after the insertion of the In28 integron, of a res site located in the Tn1403 transposon. The authors also describe additional cases of integrons that are located in res sites (4). However, I feel that they could have more clearly distinguished between their original findings and those that have been previously published. Specifically, the authors report that the In1 integron of R46 is located in a “putative res site,” yet they make no reference to a study conducted by my colleagues and me that showed the presence of a functional res site overlapping the outer part of the 5′-CS of the In1 integron (6). This res site is recognized by the uvp1 resolvase, and the uvp1 gene is adjacent to In1. We also demonstrated that the resolution site in the 5′-CS can effectively replace the original uvp1 res site. Although there do exist differences between the uvp1 resolvase and “resP” (defined by Partridge and colleagues), these differences are negligible, in that they are limited to two consecutive amino acids, as deduced by a comparison of the sequences (accession numbers gi137185 and gi1688102, respectively). Furthermore, the upstream sequences, specifically, the res site and the 5′-CS of In1, are identical when comparing uvp1 and resP.

Two other papers have pointed out the association between the uvp1 resolvase and the In1 integron, thoroughly analyzing the role played by resolvases in determining the site of transposition (1, 3). Although examples from these reports (i.e., R388 and Tn21) are cited by Partridge and colleagues (4), there arises the problem created by the lack of a single nomenclature in the literature for these genetic elements (integrons and transposons), making it difficult to distinguish truly novel identifications of integron-resolvase associations.

Overall, the data reported over the last few years suggest that the locations of integrons and/or transposons close to resolvase genes are the result of targeted recombinations (1, 3, 6). However, the role of resolvases in determining these events remains to be fully understood. Some experiments clearly show that the presence of resolvase proteins is required for insertions to occur in res sites, although the resolvase catalytic activity does not seem to be necessary, at least in some cases (1). In the case of In28, the insertion was probably mediated by a transposase, since a 5-bp duplication surrounds the insert (4). However, in a recent study on a Tn21-like transposon, the insertion of an integron occurred within a region containing a putative res site without duplication of the target sequence, suggesting that there exist other mechanisms for this insertion (7).

Perhaps a comprehensive view of the characteristics of resolvases could suggest other possible mechanisms for the integron/transposon insertions in res sites. It has been recently demonstrated that the resolvases can effectively mediate DNA integration in a very specific target (5) and also fuse replicons bearing different res sites (2). The latter event produces a hybrid res site similar to the one described in R46 (6). On these grounds, it is conceivable that in some cases resolvases could directly integrate integrons in res sites.