Spread of bla(CTX-M-14) is driven mainly by IncK plasmids disseminated among Escherichia coli phylogroups A, B1, and D in Spain

Abstract

Since its first description in 2000, CTX-M-14 has become one of the most widespread extended-spectrum beta-lactamases in Spain. In the present Escherichia coli multilevel population genetic study involving the characterization of phylogroups, clones, plasmids, and genetic platforms, 61 isolates from 16 hospitalized patients and 40 outpatients and healthy volunteers recovered from 2000 to 2005 were analyzed. Clonal relatedness (XbaI pulsed-field gel electrophoresis [PFGE] type, phylogenetic group, multilocus sequence type [MLST]) was established by standard methods. Analysis of transferred plasmids (I-CeuI; S1 nuclease; restriction fragment length polymorphism analysis; and analysis of RNA interference, replicase, and relaxase) was performed by PCR, sequencing, and hybridization. The genetic environment of bla(CTX-M-14) was characterized by PCR on the basis of known associated structures (ISEcp1, IS903, ISCR1). The isolates were mainly recovered from patients in the community (73.8%; 45/61) with urinary tract infections (62.2%; 28/45). They were clonally unrelated by PFGE and corresponded to phylogenetic groups A (36.1%), D (34.4%), and B1 (29.5%). MLST revealed a high degree of sequence type (ST) diversity among phylogroup D isolates and the overrepresentation of the ST10 complex among phylogroup A isolates and ST359/ST155 among phylogroup B1 isolates. Two variants of bla(CTX-M-14) previously designated bla(CTX-M-14a) (n = 59/61) and bla(CTX-M-14b) (n = 2/61) were detected. bla(CTX-M-14a) was associated with either ISEcp1 within IncK plasmids (n = 27), ISCR1 linked to an IncHI2 plasmid (n = 1), or ISCR1 linked to IncI-like plasmids (n = 3). The bla(CTX-M-14b) identified was associated with an ISCR1 element located in an IncHI2 plasmid (n = 1) or with ISEcp1 located in IncK (n = 1). The CTX-M-14-producing E. coli isolates in our geographic area are frequent causes of community-acquired urinary tract infections. The increase in the incidence of such isolates is mostly due to the dissemination of IncK plasmids among E. coli isolates of phylogroups A, B1, and D.

FIG. 1.

Snapshot of clusters of E. coli populations from the E. coli MLST database with related and unrelated STs. The STs identified in this study are represented by circles: dark gray circles, founders of each clonal complex; light gray circles, subgroups of the founder; and black circles, all other STs. A node connection between two STs reflects a single-locus variation between them. This diagram does not show the genetic distance between unrelated STs and clonal complexes.

FIG. 2.

(Top) Protein sequence alignment of replication proteins amplified from IncI1 and IncK plasmids. White on black, invariant amino acids; black on dark gray, strongly conserved amino acids; black on light gray, similar amino acids; gray on white, weakly similar amino acids. (Bottom) Phylogeny of replicases from the IncI1 and IncK plasmids. A neighbor-joining tree was constructed by using the complete deletion and p-distance model and was tested with bootstrap values (1,000 replicates) and by use of the multiple alignments of complete replicase sequences of several IncI1, IncK, and IncL/M plasmids by using MEGA software (version 3.1) (27). The replicase of plasmid p29930 was used to root the tree. Replicases whose protein sequences are 100% identical to those amplified from samples of plasmids pRYC105, pRYC106, and pRYC108 are grouped by keys. The GenBank accession numbers of the proteins used in the phylogeny are as follows: R64, NP_863360; ColIb-P9, NP_052449; pNF1358, AAZ05341; pMU707, AAA98176; pO113, YP_308827; pSE11-1, YP_002296026; pETEC_73, YP_001451513; pSL476_91, YP_002043842; pCVM29188_101, YP_002039041; pTN38148, YP_002302249; pS12, BAD72945; pSERB1, AAT94184; pCTX-M3, NP_775034; pMU407.1, AAA87028; pEL60, NP_943262; p29930, CAD58558.

FIG. 3.

(Top) Sequence alignment of partial relaxase sequences amplified from CTX-M-14-carrying plasmids pRYC105 and pRYC108 with oligonucleotides specific for IncI1 and IncK plasmids. White on black, invariant amino acids; black on dark gray, strongly conserved amino acids; black on light gray, similar amino acids; gray on white, weakly similar amino acids. (Bottom) Phylogeny of clade MOB_P12 relaxases from IncI1 and IncK plasmids and an expanded view of clade MOB_P12 (19). A neighbor-joining tree was constructed by using the pairwise deletion and p-distance model and was tested with bootstrap values (1,000 replicates) and by use of the multiple alignment of the N-terminal 300 amino acids of relaxase proteins of several IncI1, IncK, and IncL/M plasmids by using MEGA software (version 3.1) (27). Relaxase TraI of IncP1-α plasmid RP4 was used to root the tree. Black lines, branches comprising relaxases of IncI1 plasmids; gray lines, relaxases of IncK plasmids. Relaxases whose protein sequences are 100% identical to those amplified from samples of plasmids pRYC105 and pRYC108 are grouped by keys. The GenBank accession numbers of the plasmid relaxases used in the phylogeny are as follows: R64, BAA78021; pSL476_91, YP_002043909; ColIb-P9, NP_052501; pNF1358, AAZ05365; pSE11-1, BAG80277; pETEC_73, YP_001451521; pCVM29188_101, YP_002039113; pSC138, YP_209402; pO113, YP_308733; pSERB1, AAT94234; pCTX-M3, AAN87675; pEL60, NP_943241; and RP4, CAA38336.