PubMed1995
Conjugative transposons are important determinants of antibiotic resistance, especially in gram-positive bacteria. They are remarkably promiscuous and can conjugate between bacteria belonging to different species and genera. Transposon-promoted conjugation may be similar to F plasmid-promoted conjugation, as it appears that only one strand of the transposon DNA is transferred from donor to recipient. The recent determination of the entire nucleotide sequence of Tn916 allowed us to make specific predictions about the possible function of different open reading frames and the position of a (hypothetical) origin of transfer. The mechanism of recombination during conjugative transposition differs from that of other transposons, as shown by the absence of a duplication of the target sequence upon integration. The current model for recombination postulates that staggered double-stranded cleavages occur at each end of the transposon. One DNA strand is cut six bases from the end of the transposon, and the other strand is cut immediately adjacent to the end. The ends of the excised transposon are then ligated to form a circular intermediate with a six-base heteroduplex. Staggered cleavages of the circular intermediate and the target DNA allow the transposon to insert into the target, where it is flanked by heteroduplex regions that are resolved by replication. All hosts examined contain preferential target sites: these are not specific sequences but apparently consist of bent DNA. The site-specific recombinases encoded by conjugative transposons belong to the integrase family. Like phage lambda integrase, the integrase of Tn916 has two DNA-binding domains that recognize different sequences, one within the ends of the element and one that includes target DNA. The affinity of Tn916 integrase for target sites correlates with the frequency of integration into a particular site. The similarity between conjugative transposons and phage lambda is striking and suggests that both use the same mechanism of recombination. In lambda, however, recombining sites must be homologous. Homology may be necessary because of branch migration, which is thought to occur during recombination. In conjugative transposition, the recombining sites are nearly always different, and therefore branch migration probably does not occur. This review presents a speculative model for the alignment of the ends of Tn916 during excision that was adapted from one recently proposed for lambda.