Genome Analysis of the Carbapenem- and Colistin-Resistant Escherichia coli Isolate NRZ14408 Reveals Horizontal Gene Transfer Pathways towards Panresistance and Enhanced Virulence

LETTER

Colistin is used as a last-resort antibiotic for treating infections caused by multidrug-resistant members of the family Enterobacteriaceae, in particular for carbapenem-resistant isolates. Resistance to colistin in Enterobacteriaceae species has been attributed to chromosomal mutations (1), but recently a plasmid-encoded colistin resistance gene, mcr-1, was described in isolates from different sources and in many countries worldwide (2, 3). The emergence of plasmid-encoded colistin resistance is critical, as its transfer could render carbapenem-resistant bacteria pandrug resistant, resulting in virtually untreatable pathogens.

We recently reported the first mcr-1/bla_KPC-2-encoding carbapenem-resistant Escherichia coli isolate (NRZ14408) from a patient with a wound infection (4). In order to characterize the isolate, we sequenced the whole genome to completion (see the supplemental material).

E. coli NRZ14408 harbored a chromosome of 5,344,876 bp with a GC content of 50.65%, four plasmids of about 12 to 238 kbp (p14408_M, p14408_1, p14408_2, and p14408_3; see Table S1 in the supplemental material), and two phages of 41 and 19 kbp (14408_1, 14408_2; see Table S1 in the supplemental material). E. coli NRZ14408 is an O7:H6 isolate of multilocus sequence type 362 (ST362) and phylogenetic group D, which is associated with extraintestinal infections (5).

Plasmid p14408_3 harbored no antibiotic resistance genes but did have colicin R genes (cra, cri, and crl) and the mobC gene for mobilization of the plasmid. The mcr-1 gene was flanked by two ISApl1 elements and located on an IncHI2 plasmid (p14408_M) with high similarity to other mcr-1-encoding IncHI2 plasmids (pHNSHP45-2 [accession no. KU341381.1], pSA26-MCR-1 [KU743384.1], and pS38 [KX129782]). p14408_M differed from these plasmids by the presence of five additional insertion sequences (ISKpn11, ISKpn12, ISAba14, and two copies of IS629; Fig. 1) and a different set of antibiotic resistance genes (see Table S2 in the supplemental material). Plasmids p14408_1 and p14408_2 harbored the same 36,152-bp region, including nine antibiotic resistance genes (see Table S1 in the supplemental material) and a large portion of the IncN plasmid conjugation machinery (see Fig. S2 in the supplemental material). It displayed a unique bla_KPC-2 cassette [IS26, aac(3)-IId, ISCfr1, bla_KPC-2, bla_TEM-1B, and ISKpn16], present only in the IncN plasmid pCF8698_KPC2, which was isolated during an Enterobacteriaceae bla_KPC-2 outbreak in southern Hesse in Germany (6). The region was flanked by two IS26 insertion sequences in p14408_1, indicating an IS26-mediated transfer (7, 8).

FIG 1.

Whole plasmid comparison of three mcr-1-encoding IncHI2 plasmids, p14408_M, pHNSHP45-2 and pSA26-MCR-1, performed using BRIG (9). Mobile elements are marked in blue color; genes are marked with black arrows.

The genome of E. coli NRZ14408 displayed a distinct distribution of antibiotic resistance and virulence genes. Antibiotic resistance genes were located exclusively on plasmids (see Table S1 in the supplemental material), whereas virulence genes were located solely on the chromosome (see Table S3 in the supplemental material). Analysis of other E. coli ST362 isolates from 1952 to 2014 (see Table S4 in the supplemental material) harboring virulence determinants or antibiotic resistance genes provided further evidence supporting this observation (see Fig. S2 in the supplemental material). In addition, the total number of virulence genes was generally higher in ST362 isolates harboring antibiotic resistance genes than in antibiotic-susceptible isolates (see Fig. S3 in the supplemental material).

As ST362 is associated with extraintestinal infections, it may be a representative of those E. coli clones where increased antibiotic resistance together with enhanced virulence is being selected through the use of injudicious antibiotic regimens.

Accession number(s).

The closed genome and plasmid sequences obtained in this study were deposited in the European Nucleotide Archive under accession numbers LT599825 to LT599831.