Mol Gen Genet (1988) 213:487-490 © Springer-Verlag 1988 ISl-mediated mobility of the aerobactin system of pColV-K30 in Escherichia cob" Victor de Lorenzo*, Marta Herrero*, and J.B. Neilands Department of Biochemistry, University of California, Berkeley, CA 94720, USA Summary. Genes determining the high affinity iron trans- port system mediated by the siderophore aerobactin are flanked in the enterobacterial plasmid pColV-K30 by in- verted repeats of IS/ sequences, suggesting that the aero- bactin genes are part ofa transposon. To study this possibil- ity, the entire region between the two IS/ sequences was cloned as an 18 kb HindIII-BamHI restriction fragment in pUC8 giving plasmid pMO1. A number of derivatives of pMO1, in which aerobactin genes were tagged with a kana- mycin resistance gene, were prepared in order to assess the ability of both ISis to promote the formation of cointe- grates with pCJl05, an F derivative devoid of insertion sequences. Mating-out assays indicated that both flanking ISis were active in cointegrate formation at detectable fre- quencies. In some cases, the cointegrates could be resolved, the final result being a transposition-like event for the entire aerobactin system. Key words: pColV - Aerobactin Transposition IS/ Cointegration Introduction Aerobactin is an hydroxamate-type siderophore first found to be produced by Aerobacter aerogenes (Gibson and Ma- grath 1969). Since the discovery of its production by Escherichia coli strains carrying the virulence plasmid pColV-K30 (Warner et al. 1981), systematic surveys in clini- cal isolates have shown a significant incidence of aerobactin production in enterobacteria causing different types of in- fections in humans (Montgomerie et al. 1984; Bindereif and Neilands 1985; Valvano et al. 1986; Carbonetti et al. 1986). Aerobactin production has been found in virtually all path- ogenic enterobacteria (Cercenado etal. 1986; Martinez et al. 1987). The genetic organization of the aerobactin determinants of plasmids pColV-K30 (Carbonetti and Williams 1984; Roberts et al. 1986a; Ford et al. 1986; de Lorenzo et al. 1986) and pColV-K311 (Gross et al. 1984, 1985)ofE. coli has been completely ascertained and found to be virtually identical in both cases. The aerobactin system spans about 8 kb of DNA and consists of a cluster of five genes (iucABCD iutA) arrayed in an operon (Fig. 1). The same genetic arrangement has been detected by DNA hybridiza- * Present address: Universit6 de Gen6ve, D6partement de Biochi- mie M~dicale, Centre Medical Universitaire. 9, Avenue de Cham- pel, CH-1211 Gen6ve 4, Switzerland Offprint requests to. J.B. Neilands tion in virtually all aerobactin-producing strains. In addi- tion to pColV-type plasmids, aerobactin production has been frequently found to be encoded by non-pColV plas- mids and, in many cases, is chromosomally determined (McDougall and Neilands 1984; Marolda et al. 1987; Ro- berts et al. 1986b). While the genetic determinants for aerobactin produc- tion are highly conserved among the different strains which carry them, the flanking regions have been observed to vary to different extents. In the prototypical plasmid pColV- K30, aerobactin genes are flanked by inverted IS1 insertion sequences and two replication regions (McDougall and Nei- lands 1984; Perez-Casal and Crosa 1984, 1987). Similar DNA environments have been found surrounding the aero- bactin genes of other pColV plasmids, although changes were observed outside of the approx. 18 kb bracketed by the insertion sequences and one of the two replication re- gions was occasionally missing (Waters and Crosa 1986). In other instances, one of the two insertion sequences and/ or replication regions is missing. Finally, aerobactin genes can be found devoid of any flanking IS1 or replication regions (Roberts et al. 1986b; Marolda et al. 1987). In order to explain the conservation of aerobactin gene sequences in the midst of variable regions, it has been pro- posed that in pColV plasmids the region bracketed by ISis and replication regions could form part of a transposon (McDougall and Neilands 1984; Perez-Casal and Crosa 1984; Bindereif and Neilands 1985) or of a larger" aerobac- tin replication unit" (Waters and Crosa 1986). The cases where these elements are missing (Roberts et al. 1986b; Marolda et al. 1987), have been claimed to be evidence for an extinct transposon (Roberts et al. 1986b). However, fail- ures to show transposition of the system in laboratory-con- ditions have been reported (Roberts et al. 1986b; Waters and Crosa 1986). In this paper, we address the issue of whether IS1 flank- ing aerobactin genes of pColV-K30 are competent to pro- mote replicon fusion and, if so, whether that ability may result in the mobility of the aerobactin system. Genetic evi- dence shown here indicates that both ISls are indeed able to induce formation of cointegrates which, upon resolution, lend a transposition-like character to the aerobactin genes. Materials and methods Media, strains and plasmids. Standard LB medium (Miller 1972) was supplemented, when necessary, with 15 gg/ml tet- racycline (Tc), 60 lag/ml streptomycin (Sin), 50 gg/ml kana- mycin (Kin), 40 tag/ml choramphenicol (Cm), and 50 ~g/ml ampicillin (Ap) for monocopy and 200 ~tg/ml for multicopy