Emergence of mcr-1 in Raoultella ornithinolytica and Escherichia coli Isolates from Retail Vegetables in China

INTRODUCTION

Colistin is generally considered the last resort antibiotic for treating infections caused by extensively drug-resistant (XDR) Enterobacteriaceae (1, 2). The emergence of the plasmid-mediated colistin resistance gene, mcr-1, has therefore gained a lot of attention (3). This gene has spread worldwide and been detected in several genera of Enterobacteriaceae (Escherichia, Klebsiella, Salmonella, Shigella, Enterobacter, Kluyvera, and Citrobacter) (4–6). Although MCR-1-producing bacteria are mainly detected in animal sources, environmental samples (river water and sewage) with strains harboring mcr-1 have also been reported (4–9). Vegetables, which might be contaminated with bacteria via sewage irrigation, have also been reported to carry bacteria harboring mcr-1 (8). We investigated the occurrence of mcr-1 in Enterobacteriaceae isolated from retail vegetables in Guangzhou, China.

Between May 2015 and August 2016, 916 fresh vegetable samples (239 lettuce heads, 218 tomatoes, 176 carrots, 264 cucumbers, and 19 bean sprouts) were collected from 41 farmers' markets and 12 supermarkets in Guangzhou, China. Chromogenic Brilliance extended-spectrum β-lactamase (ESBL) agar plates (CHROMagar Microbiology, Paris, France) were used for isolation of ESBL producers which were identified by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) (Shimadzu, Japan) and some were confirmed by 16S rRNA sequencing. A total of 244 ESBL-producing Enterobacteriaceae isolates (175 Klebsiella pneumoniae, 23 E. coli, and 46 others) were recovered from 216 samples. The MICs of 10 antibiotics were determined by an agar dilution or a broth microdilution method. The presence of mcr-1 and mcr-2 in this strain collection was determined by PCR and sequencing using the primers described previously (1, 10). Our results showed that six of the 244 (2.5%) ESBL-producing isolates (four E. coli and two Raoultella ornithinolytica) were positive for mcr-1 (Table 1). The prevalence of mcr-1 in the 23 ESBL-producing E. coli isolates was 17.4%. Four ESBL producers (one Klebsiella pneumonia isolate and three Enterobacter species isolates) were resistant to colistin but were negative for mcr-1 and mcr-2.

In addition, 26 E. coli isolates were obtained from the 916 samples by using MacConkey agar plates without antibiotics and were identified by biochemical methods and 16S rRNA sequencing. Three (11.5%) of them were positive for mcr-1.

All nine mcr-1-carrying isolates showed resistance to colistin and florfenicol (Table 1). In addition, 6 and 5 isolates exhibited resistance to cefotaxime and fosfomycin, respectively. By PCR, all mcr-1-positive isolates carried the floR gene, and 6, 5, and 4 of them carried the bla_CTX-M (4 bla_CTX-M-14 and 2 bla_CTX-M-55), fosA3, and oqxAB genes, respectively (Table 1).

The clonal relatedness among the mcr-1-positive isolates was determined using pulsed-field gel electrophoresis (PFGE) and multilocus sequence type (MLST) analysis (the second only for E. coli) (http://www.warwick.ac.uk/mlst/). Six of the mcr-1-positive E. coli isolates were clonally unrelated, while two E. coli and the two R. ornithinolytica strains were related (Table 1). PFGE after S1 nuclease digestion and hybridization revealed that mcr-1 was located on 30- to 250-kb plasmids in 7 isolates (five E. coli and 2 R. ornithinolytica), but was located on the chromosome in the two ST206 E. coli strains (Table 1). A conjugation experiment using streptomycin-resistant E. coli C600 as recipient and selection with colistin (2 μg/ml) and streptomycin (3000 μg/ml) yielded 5 transconjugants from five of the seven mcr-1-positive E. coli isolates. In 2 transconjugants, fosA3, floR, and bla_CTX-M-14 were cotransferred with mcr-1. The replicon types of the mcr-1-harboring plasmids were determined using the protocol provided in the Plasmid MLST Database (http://pubmlst.org/plasmid/); two IncHI2/ST3, two IncI2, and one IncX4 plasmids were identified (Table 1). The genetic context of mcr-1 was investigated using primers: CLR3-F (CGAAGCACCAAGACATCA) and CLR3-R (CCACAAGAACAAACGGACT); Mhp-F (TTGCCAGATTTGCTACTGT) and ISAp-R (TTTCTCGCTCGTTTAT TGTA). ISApI1 was located upstream of mcr-1 in all the isolates except E. coli TS62CTX and 6HS20E. Three isolates (6BS21eCTX, TF33E, and TF10E) also carried ISApI1 downstream of mcr-1.

Recent studies have implied that fresh produce might be a possible route for the spread of resistance genes in the community (8, 11). This study documented the presence of mcr-1 in Enterobacteriaceae isolates from lettuce and tomato. Although the occurrence of mcr-1 in isolates from vegetables was lower than that in isolates from food animals and animal foods in China (3), the presence of MCR-1-producing bacteria in vegetables represents a threat to human health, as fresh vegetables such as tomato and lettuce are often consumed raw. Of note, the plasmids (IncHI2/ST3, IncX4, and IncI2) accounting for the spread of mcr-1 found in this study were similar to those found in food animals from China (5, 12–14) and all mcr-1-positive isolates carried the florfenicol (an antimicrobial agent used only in animals) resistance gene floR. Vegetables are often in direct contact with the soil and water, which are known to be important reservoirs of antimicrobial resistance genes shared between animals, humans, and the environment (15). In addition, vegetables might be fertilized with manure and wastewater from antimicrobial-treated livestock. Thus, the MCR-1-producers found in vegetables might have originated from animals.

Here, we reported the emergence of mcr-1 and fosA3 in R. ornithinolytica, belonging to a genus in Enterobacteriaceae that is closely related to Klebsiella (16). R. ornithinolytica is usually found in aquatic, soil, and botanical environments. It is also an unusual pathogen associated with community-acquired infections. However, cases of R. ornithinolytica infection might have been underrecognized due to its misidentification as a Klebsiella species in clinical laboratories that use conventional phenotypic methods (17). Recently, the number of reported R. ornithinolytica infections in humans has grown, probably due to the introduction of mass spectrometry and the use of molecular identification techniques in clinical microbiology laboratories (16, 17). Carbapenem resistance genes have been detected in R. ornithinolytica in China (18); thus, the presence of mcr-1 in R. ornithinolytica, as well as fosA3 and bla_CTX-M-14, will lead to the emergence of multidrug-resistant R. ornithinolytica strains.

In summary, the results of this study indicated that fresh vegetables constituted a possible route for the spread of MCR-1-producing Enterobacteriaceae species. The presence of MCR-1-producing organisms in the food supply and ready-to-eat foods is alarming and represents a serious risk for human health. To improve food safety and consumer health, appropriate measures such as proper disposal of animal excrement before use as fertilizers and improvement of the quality of irrigation water need to be taken. Further studies are required for evaluating the prevalence of antimicrobial resistance genes in vegetables in China and other countries.