Plasmids allow the movement of genetic material, including antimicrobial resistance genes, between bacterial species and genera. They frequently mediate resistance to multiple antimicrobials and can result in the acquisition by a pathogen of resistance to all or most clinically relevant antimicrobials. Unfortunately, there are still large gaps in our understanding of how new multi-resistance plasmids evolve. Five Australian clinical institutions collaborated in this study of multi-resistance plasmids in clinical isolates of Escherichia coli. We characterized 72 resistance plasmids in terms of the antimicrobial resistance profile they conferred, their size and their incompatibility group. Restriction fragment length polymorphisms were used to determine the genetic relationships between the plasmids. Relationships between the host cells were determined using multi-locus enzyme electrophoresis. A lack of correlation between the evolutionary history of the host cells and their plasmids suggests that the horizontal transfer of resistance plasmids between strains of E. coli is common. The resistance plasmids were very diverse, with a wide range of resistance profiles and a lack of discrete evolutionary lineages. Multi-resistance plasmids did not evolve via the co-integrative capture of smaller resistance plasmids; rather, the roles of recombination and the horizontal movement of mobile genetic elements appeared to be most important.
INTRODUCTIONSince their discovery in the 1950s (Watanabe & Fukasawa, 1960), antimicrobial resistance plasmids have been increasingly associated with both Gram-positive and Gramnegative bacterial infections. Plasmid-associated resistance genes have been discovered for a majority of known antimicrobials, including the quinolones and fluoroquinolones (Hawkey, 2003;Neu, 1992), and it is not uncommon for a single plasmid to simultaneously mediate resistance to five or six antimicrobials. This ability to sequester multiple resistance genes is of particular concern to modern medicine. Levin (1995) made two predictions regarding the evolution of multi-resistance plasmids. Where two incompatible plasmids are simultaneously selected for, he argues that new plasmids will arise by transposition and predicts that the position of resistance genes in otherwise identical plasmids will be highly variable. Where selection for the co-transfer of compatible resistance plasmids is involved, he argues that new multi-resistance plasmids will arise through cointegration and predicts the occurrence of resistance genes in plasmids with multiple replicons that can be identified as the co-integrates of other plasmids.Both co-integration and transposition have been implicated in empirical studies of plasmid evolution (Berg et al., 1998; Bradley et al., 1986;Guessouss et al., 1996;Mitsuhashi et al., 1977;Schwarz et al., 1996;Sohail & Dyke, 1995;Venkatesan et al., 2001;Woodward et al., 1990). However, a range of other mechanisms, including recombination and the acquisition of integron cassettes, have also been observed...