Coexistence of mcr-1 and bla_NDM-1 in Escherichia coli from Venezuela

TEXT

Carbapenem-resistant Enterobacteriaceae (CRE) represent one of the most serious concerns for public health since they are susceptible to very few antibiotics, which converts the remaining compounds into last-resort agents (1). One last-resort antibiotic against CRE is colistin (polymyxin E) (2), which has been used in veterinary medicine since its discovery in 1949, mainly for the treatment of intestinal tract infections; it was initially restricted to ophthalmic and topical use in humans due to its toxicity (3). As a result of the limited therapeutic alternatives, in 2012 the World Health Organization included colistin on the list of critically important agents for human medicine (2).

Until recently, resistance to polymyxins had been identified only as chromosomally mediated mutations, which cannot be transferred between bacteria (3). In November 2015, however, a new plasmid-mediated colistin resistance mechanism, called MCR-1, was discovered (4). Since its first identification, mcr-1 from human, animal, food, and environmental origins has been widely reported (5–7). The coexistence of mcr-1 with a carbapenemase is especially worrisome, as therapeutic options in such cases are very limited. Currently, the carbapenemase New Delhi metallo-β-lactamase 1 (NDM-1) is broadly disseminated worldwide (8), although it has been scarcely described in South America (9, 10). In this work, we have detected mcr-1-positive isolates from different origins in Venezuela, as well as the coexistence of this resistance gene with bla_NDM-1.

Ninety-three samples from Cumana, Venezuela, were selected for their capacity to grow on MacConkey agar (Oxoid Ltd., Basingstoke, United Kingdom). The isolates were collected in August 2015 from human clinical fecal samples (16 samples), feces from dogs (8 samples), swine (17 samples), and poultry (16 samples), and sewage (36 samples) in different locations of Cumana. The presence of mcr-1 was screened for by PCR and Sanger sequencing (Secugen S.L., Madrid, Spain), using primers and conditions described previously (4). The two positive isolates (2.1%), BB1290 and BB1291, which were identified as Escherichia coli by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) (Bruker), were collected from fecal samples from a 43-year-old man and a pig, respectively, and displayed 100% identity to mcr-1 (4). Antimicrobial resistance was determined on the basis of MIC values determined using broth microdilution in microtiter plates (Sensititre EUVSEC; Trek Diagnostics, Inc., Westlake, OH) and interpreted following EUCAST guidelines (11). BB1290 and BB1291 exhibited multidrug-resistant profiles (Table 1).

BB1290 and BB1291 were sequenced (MiSeq; Illumina, San Diego, CA), which produced 100-bp single-end reads with 36× coverage (Life Sequencing S.L., Valencia, Spain); assembly was performed with SPADES (version 3.6.2), which produced 2,408 and 888 contigs, respectively. The data were used to characterize the strains according to antibiotic resistance genes, pathogenicity, serotypes, and plasmid incompatibility groups through the website of the Center for Genomic Epidemiology (http://www.genomicepidemiology.org). Moreover, the strains were typed by assigning alleles and sequence types (STs) from the Institute Pasteur multilocus sequence typing (MLST) website (http://bigsdb.web.pasteur.fr/ecoli/ecoli.html).

The human isolate BB1290 bore, in addition to mcr-1, a plethora of genes conferring resistance to beta-lactams (bla_NDM-1, bla_TEM-1, bla_ACT-15, bla_OXA-1, and bla_CTX-M-15), aminoglycosides [aadA5, aph(3′)-IIa, aacA4, aac(3)-IIa, strA, aadA15, and strB], fluoroquinolones [aac(6′)-Ib-cr and qnrB1], macrolides [mph(A) and erm(B)], phenicols (catB3, catA1, and floR), sulfonamides (sul1, sul2, and sul3), tetracycline [tet(B)], and trimethoprim (dfrA1, dfrA12, dfrA14, and dfrA17), which is in line with the extremely drug-resistant profile of this isolate. Incompatibility typing detected the presence of the replicons IncHI2, IncHI2A, ColBS512, IncI2, and IncFII. The strain belonged to ST19 and was assigned to serotype O100:H25, which is related to human enteropathogenic E. coli (EPEC) strains (12).

The animal sample harbored mcr-1, aadA1, aph(4)-Ia, aac(3)-VIa, bla_CTX-M-2, oqxB, sul1, tet(A), and dfrA14. Incompatibility group analysis showed the presence of the replicons IncFIB, IncI2, ColpVC, and Col8282. Remarkably, IncI2 was the only replicon shared by the two isolates. Furthermore, BB1291 belonged to a novel ST, ST452. In silico analysis assigned the isolate to serotype O17/O44:H34 and identified 684 pathogenic protein families, predicting the isolate as a human pathogen.

The 100-bp single-end reads were then mapped against the Chinese plasmid pHNSHP45, bearing mcr-1 (4), with Geneious (version 8.1.7) (13). The results showed that BB1290 and BB1291 harbored plasmids with 98% and 97% identity, respectively, with respect to pHNSHP45, albeit lacking the ISApl1 mobile element upstream of mcr-1. The absence of ISApl1 was confirmed by PCR analysis from ISApl1 to mcr-1 (data not shown).

Plasmid DNA extractions from BB1290 and BB1291 (QIAprep; Qiagen Inc., Chatswoth, CA) were transformed into E. coli HST08 (Stellar competent cells; TaKaRa Bio, Inc., Otsu, Japan), following the manufacturer's protocol, and were selected on brain heart infusion (BHI) agar containing colistin (2 mg/liter). BB1290T and BB1291T transformants obtained from the wild-type strains were positive for the mcr-1 gene and the IncI2 plasmid incompatibility group (PCR-based replicon typing [PBRT] kit; Diatheva). Resistance profiles of the transformants showed that the plasmids conferred resistance only to colistin (Table 1).

To the best of our knowledge, the two colistin-resistant E. coli isolates from Venezuela represent the first detection of mcr-1 in this country. The patient had no direct contact with animals, had not been treated with colistin, and had not recently traveled to other countries. However, only a very small number of samples have been tested, and no significant statements regarding transmission routes can be made.

One of the mcr-1-positive E. coli strains, BB1290, also harbored bla_NDM-1, revealing the coexistence of these two genes in the same isolate. This combination in a human pathogen is worrying, as it impedes the use of most last-resort antibiotics (14). In our case, BB1290 was still susceptible to tigecycline.

Control of mcr-1, its genetic platforms, and the bacteria implicated in its dissemination is essential. The collection of surveillance data from developing countries where the information is scarce, such as Venezuela, is crucial in order to establish accurate measures that eventually safeguard the effectiveness of last-resort antibiotics.