Pathogenic Lineage of mcr-Negative Colistin-Resistant Escherichia coli, Japan, 2008–2015

evolutionarily distinct from other AIVs (5). This H5N5 strain is a contemporary reassortant virus related to North American and Eurasian strains. The positive animals we identified originated from a single location on the Antarctic Peninsula, which suggests recent introduction of this AIV H5N5 in the colonies sampled. Antarctica is refuge for most penguin colonies, including the near-threatened emperor penguins. Previous reports suggested that AIV could have caused Adélie penguin chick death (3). Four positive samples (including the sequenced virus) were obtained from juvenile chinstrap penguins that were weak, depressed, and possibly ill (i.e., they had ruffled feathers, lethargy, and impaired movement). Thus, additional studies are warranted to assess the health and conservation status of resident bird species and potential pathologic effects of AIV. These data provide novel insights on the ecology of AIV in Antarctica. Our findings also highlight the need for increased surveillance to understand virus diversity on this continent and its potential contribution to the genetic constellation of AIV in the Americas.

We tested 514 E. coli isolates obtained from clinical specimens taken at Sapporo Clinical Laboratory Inc. (Sapporo, Japan) and Sapporo Medical University Hospital in Japan during 2008-2009 (1) and 2015, respectively. Samples were processed according to Clinical and Laboratory Standards Institute guidelines (2). Identification of O25b:H4-ST131, O25b, H4, and ST131 were determined as described previously (1). For identification of the H30Rx subclone of O25b:H4-ST131, H30 was determined by PCR using a specific primer set (3), R was determined according to ciprofloxacin MIC, and x was determined by detecting 2 singlenucleotide polymorphisms, as previously described (4).
Four E. coli isolates exceeded the colistin resistance breakpoint (>2 mg/mL) (Table). None of the patients from whom the E. coli isolates were derived had a history of colistin treatment. Three of the 4 colistin-resistant isolates belonged to a pandemic lineage, O25b:H4-ST131-H30R, which has been isolated from urinary tract and bloodstream infections (3,4). The frequency of colistin-resistant ST131 E. coli isolates among O25b:H4-ST131 was 2.2%. This lineage is fluoroquinolone resistant and is frequently resistant to β-lactams because it possesses CTX-M-type extendedspectrum β-lactamase genes (1,3,4).
The colistin-resistant isolates reported were resistant to fluoroquinolones, and 1 (SME296) was resistant to cephalosporins (due to expression of bla CTX-M-14 ). Another colistin-resistant E. coli isolate (SME222) belonged to O18-ST416, which is also known as an extraintestinal pathogenic E. coli (5), although this lineage has not previously been reported to exhibit colistin resistance.
The colistin-resistant E. coli isolates we identified were sensitive to carbapenems and aminoglycosides, including amikacin, whereas previously it was reported that some E. coli ST131 isolates exhibited resistance to carbapenems by possessing carbapenemases, such as NDM-1 and KPC-2; the NDM-1-possessing ST131 isolate also exhibited resistance to amikacin (6,7). Thus, these findings may affect future antimicrobial choices because of the clonal dominance, multidrug resistance, and pathogenicity of the isolates.
Recent studies reported a plasmid-mediated colistin resistance gene, mcr-1, in various countries (8). In addition, a novel plasmid-mediated colistin resistance gene, mcr-2 (76.7% nucleotide identity to mcr-1), was found in E. coli isolates in Belgium (9). These genes encode a phosphoethanolamine transferase family protein, which modifies the lipid A component of lipopolysaccharide (8,9). The colistin-resistant E. coli isolates we identified did not possess mcr-1or mcr-2, although the MICs for colistin were the same as or higher than that of the transconjugant of a mcr-1-harboring plasmid in an E. coli ST131 isolate (4 mg/L) reported by Liu et al. (8). Thus, these colistin-resistant isolates may have other colistin resistance mechanisms. For example, modification of lipid A with 4-amino-4-deoxy-l-arabinose or phosphoethanolamine, caused by chromosomal mutations in mgrB, phoPQ, and pmrAB genes, might occur and could be responsible for the resistance. This polymyxin-resistance mechanism is seen in Enterobacteriaceae; however, other novel mechanisms are also conceivable.
In conclusion, we report colistin resistance in a major global-spreading extraintestinal pathogenic E. coli strain, O25b:H4-ST131-H30R, in Japan. This strain acquired colistin resistance without carrying a plasmid bearing the mcr gene. Clarifying the colistin-resistance mechanisms in these isolates is necessary if we are to forestall the emergence of multidrug (including  colistin)-resistant O25b:H4-ST131-H30R. The worstcase scenario is the global spread of this isolate, which has acquired resistance to the last-line antimicrobial drug, colistin.