Comparative genomics of IncL/M-type plasmids: evolution by acquisition of resistance genes and insertion sequences.

IncL/M-type plasmids R446b and R471a were originally isolated from Morganella morganii (formerly Proteus morganii) as the first members of the IncL/M group of multidrug resistance (MDR) plasmids (1, 2). IncL/M plasmids are now commonly identified among environmental and clinical isolates (3–5). This group of plasmids can be considered an emerging threat since it has been increasingly identified as a source of broad-spectrum -lactam resistance, encoding either the metallo-lactamase NDM-1 or the class D carbapenemase OXA-48 and also being responsible for the dissemination of extended-spectrum -lactamase genes such as blaCTX-M-3 (5–8). In this study, we analyzed the sequence of plasmid pNDM-OM, harboring the blaNDM-1 gene, that was recovered from a clinical Klebsiella pneumoniae isolate from the Sultanate of Oman, providing new insights into the evolution of MDR IncL/M-type plasmids. The complete sequencing workflow of the Illumina Genome Analyzer IIx system (Illumina Inc., San Diego, CA) was performed by the DNAVision company (Gosselles, Belgium). Plasmid pNDM-OM was 87,185 bp in size with an average GC content of 52% and contained 98 open reading frames. It possessed a complete array of genes involved in replication, conjugation, and partition (see Table S1 in the supplemental material). The plasmid architecture observed in pNDM-OM was similar to that of other IncL/M plasmids previously sequenced (Fig. 1). It exhibited a very high gene synteny with plasmid pNDM-HK, which also encodes NDM-1 with only very few differences, consisting of one to four nucleotide changes/deletions at three different locations together with the lack of two insertion sequences identified in pNDM-HK, namely, IS26 and IS186 (Fig. 1). In addition, part of the tnpA gene of Tn2 (563 bp) in pCTX-M-3 and pNDM-OM was missing in pNDM-HK. Overall, in silico analysis of the GenBank databases revealed that IncL/M plasmids have evolved through the acquisition of resistance genes and insertion sequences. The 60,145-bp plasmid pEL60, isolated from Erwinia amylovora, can be considered the typical IncL/M backbone since it does not possess any resistance gene or insertion sequence (see Table S1 in the supplemental material) (Fig. 1 and 2) (4). Plasmid pOXA-48a differed from pEL60 by the following two main additional elements: (i) the integration of composite transposon Tn1999 carrying the blaOXA-48 carbapenemase gene in the tir gene that encodes a transfer inhibitory protein and (ii) a recombination event resulting in the exchange of three genes within the tra operon (namely, traX, traY, and excA) (8). Plasmid pCTX-M3 differs from pEL60 (9) by the acquisition of the ISEcp1-blaCTX-M-3-carrying transposon and by the acquisition of a large resistance region constituted by transposon Tn2 in which a Tn1548-like transposon, carrying the 16S rRNA methylase armA gene, was inserted. The structure of pCTX-M360 can be considered an intermediate between plasmids pEL60 and pCTX-M3 since it possesses the typical IncL/M backbone with the insertion of ISEcp1-blaCTX-M-3 and the insertion of a native Tn2 (Fig. 2) (9). On the other hand, plasmids pNDM-HK and pNDM-OM possess a synteny similar to that of pCTX-M3, including the insertion of Tn2 and the insertion of a Tn1548-like transposon within the Tn2 resolvase gene, but a likely recombination event resulted in the insertion of the NDM module within the Tn1548-borne class 1 integron, leading to the replacement of the integrase gene and the gene cassette array by that module. The integration hot spot we identified among the IncL/M-type plasmids was located between the replication locus and the trbC gene that encodes part of the Trb transfer operon (Fig. 2). First, transposon Tn2 carrying the blaTEM-1 -lactamase gene was inserted within this region, and then IS26-mediated insertions may have occurred, leading to a deletion of the resolvase gene and the 5= end of the Tn2 transposase gene. These events gave rise to a complex structure carrying many resistance determinants, including those compromising the activity of carbapenems, broadspectrum cephalosporins, aminoglycosides, quinolones, trimethoprim, and sulfonamides (6–8, 10–12). The acquisition of the blaNDM-1 gene in pNDM-OM resulted from a recombination event leading to an exchange of a class 1 integron within the Tn1548 transposon by the NDM module. This acquisition was probably recent, since the resistance loci of pNDM-OM, pNDMHK, and pCTX-M-3 derive from a common ancestor (Fig. 1). A second integration hot spot has been identified near the pemI/ pemK locus, in which the ISEcp1-blaCTX-M-3 transposon was inserted. This transposon, conferring resistance to broad-spectrum cephalosporins, was widely distributed among clinical isolates (5). Surprisingly, pNDM-OM and pNDM-HK did not harbor the ISEcp1-blaCTX-M transposon despite the fact that they are very likely derivatives of pCTX-M3 based on the synteny and phylogenetic analysis of the transfer operon. This further strengthens the evolutive potential of IncL/Mtype plasmids that constitute currently efficient vectors for emerging and clinically relevant resistance genes. Nucleotide sequence accession number. The nucleotide and protein sequences corresponding to plasmid pNDM-OM have been deposited in GenBank under accession number JX988621.