Genomic Characterization of mcr-1.1-Producing Escherichia coli Recovered From Human Infections in São Paulo, Brazil

Abstract

Polymyxins are one of most important antibiotics available for multidrug-resistant Gram-negative infections. Diverse chromosomal resistance mechanisms have been described, but the polymyxin resistance phenotype is not yet completely understood. The objective of this study was to characterize colistin resistant mcr-1-producing strains isolated from human infections over one year in a hospital setting (Hospital das Clínicas, São Paulo, Brazil). We isolated 490 colistin-resistant Gram-negative rods, of which eight were mcr-1.1-positive Escherichia coli, the only species with this result, indicating a low incidence of the mcr-1 production mechanism among colistin-resistant isolates. All mcr-1.1 positive isolates showed similarly low MICs for colistin and were susceptible to most antibiotics tested. The isolates showed diversity of MLST classification. The eight mcr-1.1-positive E. coli genomes were sequenced. In seven of eight isolates the mcr-1.1 gene is located in a contig that is presumed to be a part of an IncX4 plasmid; in one isolate, it is located in a contig that is presumed to be part of an IncHI2A plasmid. Three different genomic contexts for mcr-1.1 were observed, including a genomic cassette mcr-1.1-pap2 disrupting a DUF2806 domain-containing gene in six isolates. In addition, an IS1-family transposase was found inserted next to the mcr-1.1 cassette in one isolate. An mcr-1.1-pap2 genomic cassette not disrupting any gene was identified in another isolate. Our results suggest that plasmid dissemination of hospital-resident strains took place during the study period and highlight the need for continued genomic surveillance.

Keywords: Polymyxin; colistin; hospital dissemination; insertion sequence; plasmid.

FIGURE 1

Phylogenetic tree generated by program parsnp (core genome SNP tree). The tree was inferred with FastTree 2.1 (Price et al., 2009). The numbers above branches are lengths, measured in number of substitutions per site. Leaf labels: Isolate identification (Patient number, Genomic Context type). The core genome size was 3,812,930 bp. This resulted from multiplying the individual genome cluster coverage value by its size, adding up these results, and dividing by the total number of genomes in the core genome computation.

FIGURE 2

Genomic context #1 identified in EcHC891, Ec482, Ec483, Ec716, Ec1057 and Ec1177 isolates. The genomic cassette mcr-1.1-pap2 was found disrupting a pre-existing DUF2806-domain-containing gene, without any associated insertion sequence. Genomic coordinates refer to scaffold 3 of the Ec482 genome.

FIGURE 3

Genomic context #2 identified in the Ec721 isolate. A 698 bp Family 1 transposase was found upstream of the mcr-1.1 gene, on the opposite strand; otherwise, the genomic context is the same as the one shown in Figure 2. Genomic coordinates refer to scaffold 2 of the Ec721 genome.

FIGURE 4

Genomic context #3 identified in the Ec502 isolate. Genomic coordinates refer to contig Qabyss.discovar.masurca.spades.a5_22 of the Ec502 genome.