Background

Enterobacteriaceae, such as E.coli, K. pneumoniae and C. freundii, are important pathogens that cause human infections. Carbapenem antibiotics are used in the treatment of infections caused by multi-drug resistant Enterobacteriaceae. However, the emergence of Carbapenem-resistant Enterobacteriaceae (CRE) has been a serious challenge to manage in the clinic because of the rapid worldwide dissemination of multi-drug resistance [1]. As one main type of carbapenemases, New Delhi metallo-β-lactamase (NDM)is able to confer resistance to almost all β-lactams, including carbapenems, in Enterobacteriaceae. Since the first report of blaNDM-1, 17 variants of NDM enzymes (NDM-1 to NDM-17) have been identified among Gram-negative bacteria worldwide (http://www.ncbi.nlm.nih.gov/pathogens/submit_beta_lactamase/). Among NDM carbapenemases, New Delhi metallo-β-lactamase-5, first identified in an E. coli strain in the UK in 2011, attracted extensive attention because of increased resistance to carbapenems and broad-spectrum cephalosporins [2]. In addition, blaNDM-5 was reported to be carried in different incompatibility typing plasmids to transfer [3], such as IncF, IncN and IncX3. These plasmids are able to facilitate the dissemination of blaNDM-5 among the members of Enterobacteriaceae through horizontal gene transfer. NDM-5-producing isolates have been identified worldwide, such as in America [4], Australia [5], China [6], Denmark [7] and India [8]. Furthermore, NDM-5-positive strains were not only isolated from clinical specimens but also from animals, such as dogs [9], cats [10] and cows [11]. Worryingly, blaNDM-5 has also been identified in environmental samples [hospital sewage water [12] and urban river [13]], indicating its presence in the community. However, the dissemination mechanism of blaNDM-5 gene remains unclear.

In this study, we screened NDM-5-producing Enterobacteriaceae to elucidate the dissemination mechanism. In addition, to the best of our knowledge, this report is the first to isolate E. coli and C. freundii strains carrying blaNDM-5 from the same patient.

Methods

Bacterial strains

From Jun. 2016 to Sep. 2017, 224 carbapenem-resistant Enterobacteriaceae isolates, as determined by the agar dilution method according to the Clinical and Laboratory Standards Institute guidelines [14], were obtained from four hospitals in different locations in Zhejiang, China. In a retrospective study, common carbapenemase genes (blaKPC, blaIMP, blaVIM, blaOXA-48, and blaNDM) were amplified, and the positive products were sequenced; eleven NDM-5 producing strains were identified for further study. The NDM-5 producing strains were preliminarily identified by the VITEK 2 system (Sysmex-bioMérieux, Marcy l’Etoile, France) and further confirmed by whole genome sequencing. The characteristics of the isolates and related clinical data are shown in Table 1.

Table 1 Clinical characteristics

Antimicrobial susceptibility testing

Antimicrobial susceptibility testing was performed using broth microdilution method [14]. The antibiotics tested in this study were amikacin, aztreonam, cefepime, ceftazidime, ciprofloxacin, gentamicin, imipenem, minocycline, colistin and tigecycline. The results were analysed according to the CLSI guidelines [14], except tigecycline and colistin, for which the European Committee on Antimicrobial Susceptibility Testing breakpoints were used (http://www.eucast.org/clinical_breakpoints). E. coli ATCC 25922 was used as a quality control strain.

Bacterial genotyping

Pulsed-field gel electrophoresis (PFGE) was performed to analyse the clonal relatedness of the NDM-5 producing E. coli isolates according to the previous study [15]. Briefly, the isolates were digested by XbaI endonuclease, which was carried out with a CHEF-Mapper XA PFGE system (Bio-Rad, USA) with a 5–35 s linear ramp for 22 h at 6 V/cm and 14 °C. The PFGE profiles were analyzed with BioNumerics software (Applied Maths, Sint-Martens-Latern, Belgium). The Salmonella enterica serotype Braenderup H9812 was used as the size marker.

MLST was also performed for molecular typing. Bacterial genomic DNA was extracted from these isolates. Seven housekeeping genes of E. coli (adk, fumC, gyrB, icd, mdh, purA and recA), and K. pneumoniae (gapa, infb, mdh, pgi, phoe, rpob) were amplified by PCR, and the products were sequenced to analyse the ST.

Southern blot analysis and conjugation experiments

To determine the plasmid location of the blaNDM-5 gene, genomic DNA digested with S1-nuclease (TaKaRa, Japan) was electrophoresed on a CHEF-mapper XA pulsed-field gel electrophoresis (PFGE) system (Bio-Rad, USA) for 18 h at 14 °C with run conditions of 6 V/cm and pulse times from 2.16 s to 63.8 s. The DNA fragments were transferred to a positive-charged nylon membrane (Millipore, USA) and then hybridized with a digoxigenin-labeled NDM-5-specific probe. An NBT/BCIP color detection kit (Roche, Germany) was then used to detect the fragments. The Salmonella enterica serotype Braenderup H9812 was used as the size marker.

A filter-mating experiment was performed between the blaNDM-5-positive isolates and rifampicin-resistant E. coli C600 as the recipient strain [15]. Transconjugants were selected on Mueller-Hinton agar plates containing 500 mg/L rifampicin and 100 mg/L ampicillin. PCR sequencing and antimicrobial susceptibility testing of the transconjugants were subsequently carried out to confirm whether the plasmid was successfully transferred to the recipient.

Plasmids analysis

Plasmid extraction and analysis was performed as previously described [15]. Briefly, the plasmid DNA of strains was extracted using a QIAamp DNA MiniKit (Qiagen, Valencia, CA, USA) following the manufacturer’s recommendations. The plasmids were sequenced on an Illumina-Hiseq™ 2000 (Illumina Inc., San Diego, U.S.A) platform with 2 × 100 bp paired-end reads. Sequence reads were assembled using CLC Genomics Workbench software package (CLC Bio 8.0). Gaps of a representative plasmid were closed by standard PCR and Sanger sequencing according to previous study [16]. The RAST (Rapid Annotation using Subsystems Technology) annotation website server (http://rast.nmpdr.org/rast.cgi) was then used to annotate the genomes of the plasmid. The circular map of the pEC463-NDM5 plasmid was generated using the CGview server [17]. A comparison of pEC463-NDM5 and three related plasmids was performed with EasyFig 2.2.2 [18]. The rested plasmid sequences were mapped to the representative plasmid sequence with CLC genomics workbench version 8.0.

Incompatibility typing of the blaNDM plasmid was performed by PCR-based replicon typing [19, 20] and was further identified with the help of PlasmidFinder-1.3 server (https://cge.cbs.dtu.dk/services/PlasmidFinder/).

In addition, plasmid stability was determined [3]. Briefly, the blaNDM-5-positive isolates were individually streaked out in the MH agar, incubated at 37 °C for 24 h, and then transferred to a fresh MH agar. After repeating this procedure for 12 days, 12 individual colonies were randomly selected. Subsequently, the blaNDM-5 gene was screened by PCR and sequenced.

Nucleotide sequence accession number

The complete sequence of the plasmid pEC463-NDM5 (accession number MG545911), is deposited at DDBJ/EMBL/GenBank.

Results and discussion

Isolate characteristics and antimicrobial susceptibility testing

Among the 224 CRE isolates, 137 isolates were KPC-2 carbapenemase producers, eleven isolates were NDM-5 carbapenemase producers, four isolates carried blaIMP-1 gene, two isolates carried blaVIM-1 gene and two isolates carried blaNDM-1 gene. In addition, 68 isolates exited other unknown mechanism of carbapenem-resistance.

In this study, eleven NDM-5-producing isolates were further identified, including nine E. coli, one K. pneumoniae and one C. freundii. These isolates were all recovered from hospitalized patients. These patients were aged between 16 and 85 years, with an average age of 55 years, had different severities of illness (Table 1), and all had previously received broad-spectrum antibiotics. Notably, with both E. coli (EC418) and C. freundii strains (CF418) were isolated from the feces of one patient from haematology department. This patient was found to be a carrier of blaNDM-5-positive strains. In contrast, the other patients from whom blaNDM-5-carrying strains were isolated from blood, pus, ascites, urine or sputum were symptomatic. In addition, these patients had no recent history of travel or hospitalization abroad.

The antimicrobial susceptibility testing results showed that the blaNDM-5-positive isolates were resistant to carbapenems, third-generation cephalosporins, and cefperazone/sulbactam. These isolates were also resistant to fluoroquinolones (81.8%), aztreonam (36.4%), amikacin (36.4%), nitrofurantoin (45.4%) and tigecycline (18.2%). All isolates were susceptible to colistin. E.coli EC122 and K. pneumoniae KP387 strains were both resistant to tigecycline, suggesting that increased resistance phenotypes of blaNDM-5-postive isolates are increasing in clinics. In addition, other β-lactamase genes, such as those encoding CTX-M-24, CTX-M-55, CMY-42, were also frequently detected in various blaNDM-5-positive E. coli strains (Fig. 1). Gene encoding SHV-1 and CMY-26 were detected in the K. pneumoniae KP387 and C. freundii CF418 strains, respectively.

Fig. 1
figure 1

The dendrogram is based on the similarity of PFGE patterns from 9 blaNDM-5 positive clinical E. coli isolates. The right illustrates results from MLST, hospitals and β-lactamase gene(s)

Fig. 2
figure 2

S1-digested plasmid DNA and southern blot hybridization of blaNDM-5 positive isolates. Bands in A with arrows pointing to them showed positive signals in Southern blot hybridization with the NDM-5 probe. M = Salmonella serotype Braenderup strain H9812 molecular marker. 1 = K. pneumoniae KP387; 2 = E. coli EC135; 3 = E. coli EC463; 4 = E. coli EC734; 5 = E. coli EC144; 6 = E. coli EC122; 7 = E. coli EC418; 8 = C. freundii CF418; 9 = E. coli EC310; 10 = E. coli EC611; 11 = E. coli EC126

Our recent studies showed that blaNDM-5 was able to coexist in the same isolate with tigecycline and colistin resistance phenotypes, thereby generating strains that approached pan-resistance. For example, blaNDM-5 was not only identified in high-level tigecycline resistance E. coli strains [21], but also coexisted in the same strain with the transferrable colistin resistance gene mcr-1 [15]. It is clear that generating strains results in so-called “superbug” isolates and accelerating entery into a “postantibiotic” era [22].

Genetic relatedness

MLST and PFGE experiments were performed to analyse the clonal relatedness of blaNDM-5-positive isolates because NDM-5 producers are infrequently isolated worldwide. According to the MLST results, nine blaNDM-5-postive E. coli isolates were grouped into 9 different sequence types. In accordance with the MLST results (Fig. 1), the different PFGE patterns confirmed that the seven E. coli isolates are not clonally related to each other even though some of the strains were collected from the same hospital. Strains EC122 and EC144 own similar the PFGE profiles, but the two strains have different sequence type and different resistance genes. Furthermore, core genome multi-locus sequence typing (cg-MLST) analysis in our study showed the blaNDM-5-positive isolates were not clonal relatedness (Additional file 1: Figure S1). In addition, the K. pneumoniae KP487 isolate belongs to ST182.

A previous study collected 11 NDM-5-producing E. coli strains from 7 hospitals in various locations in China from 2013 to 2014, and found that ST167 E. coli strains in clinical settings exhibited close linkages with the blaNDM-5 gene [23]. Our previous study also showed that high-level tigecycline resistance E. coli strains carrying blaNDM-5 also belonged to the ST167 clonal lineage [21], indicating that the ST167 sequence type is an important reservoir of blaNDM-5 in China. However, the diversity of MLST and PFGE types in the present study showed that the blaNDM-5 gene has been carried in other STs E. coli isolates from 2016 to 2017. Moreover, the blaNDM-5 gene was detected in the K. pneumoniae and one C. freundii strains, indicating that this gene has further disseminated in Enterobacteriaceae. Note that NDM-5-related outbreak has been reported [24, 25]. Although no genetic association was found between our blaNDM-5-positive isolates with other strains, the widespread dissemination of blaNDM-5 in recent years in Enterobacteriaceae highlights the need for extensive attention.

Location of the bla NDM-5 gene

S1-PFGE followed by Southern blot demonstrated that the blaNDM-5-positive strains were all located on plasmids of the same size(~ 46 Kb) (Fig. 2). The filter mating experiments were carried out to confirm the transferability of these blaNDM-5 plasmids. Nine of the 11 isolates tested could successfully transfer their carbapenem-resistant phenotype to E. coli strain C600 (Table 2). In addition,incompatibility plasmid classification showed that all the blaNDM-5 plasmids belonged to the IncX3-type plasmid. IncX3 plasmids might have played an important role in mediating the horizontal transmission of the blaNDM gene. This possibility has been supported by the results of several studies [6, 2629]. In this study, blaNDM-5 was carried by the IncX3 plasmids. Moreover, 81.8% (9/11) of isolates carrying this type plasmid were able to transfer carbapenem-resistant phenotype. However, conjugation experiments of E. coli EC126 and EC135 strains were not performed because these two strains were resistant to rifampin. To date, IncX3 plasmids carrying blaNDM-5 have been reported worldwide [3, 22, 23]. Therefore, our present study further supplements those previous studies. In addition, we isolated E. coli and C. freundii strains carrying blaNDM-5 from a single patient. These blaNDM-5-carrying plasmids had very similar sequences (99% coverage and 98% similarity), indicating probable horizontal transfer of blaNDM-5 between E. coli and C. freundii strains by one same plasmid. In addition, the plasmid stability experiments showed that the blaNDM-5-positive plasmids were all stable in these isolates. After 12 rounds of subculture in MH agar without antibiotic addition, the randomly selected strains all carried the blaNDM-5 gene and a plasmid identical to their parental isolate in size. Overall, it is important for the IncX3 type plasmid to play an important role in the further dissemination of blaNDM-5 in Enterobacteriaceae. Therefore, it is imperative that effective measures be taken immediately to control the spread of this plasmid.

Table 2 Antibiotic susceptibility of NMD5-producing isolates and their transconjugants

Plasmid sequence analysis of bla NDM-5

The entire plasmid sequence was obtained to better characterize the blaNDM-5-positive plasmid. Sequence analysis showed that the plasmid was 46,145 bp in length (Fig. 3a). The blaNDM-5 gene was preceded by IS3000, ISAba125 and IS5, and followed by bleMBL, trpF, dsbC, IS6 and ISkox3.No other antimicrobial resistance genes were detected in this plasmid.

Fig. 3
figure 3

Plasmid analysis of pEC463-NDM5. Schematic map of plasmid p pEC463-NDM5 (a), comparative analysis of three blaNDM-5-carrying IncX3 plasmids (b). The putative open reading frames are shown as arrowheads orrods (less than 130 amino acids). The gene name is shown near the corresponding arrowhead or rod. The depthof shading is indicative of the percentage BLASTN match, as indicated on the bottom

Further sequence alignments based on BLAST revealed that the plasmid sequences showed almost identical nucleotide sequences with those of the previously reported IncX3 plasmids pNDM-MGR194 of K. pneumoniae MGR-K194 in India [8]. The plasmid pNDM-MGR194 carrying blaNDM-5 was reported in 2015 in India, which was considered to play an important role in the dissemination of the blaNDM-5 gene because pNDM-MGR194-like plasmid was highly similar to those plasmids reported in China [3], Australia [5] and Denmark [7]. In addition, most of the blaNDM-5-carrying plasmids reported in China belonged to the IncX3-type and were identical or near-identical to pNDM-MGR194-like plasmid (Table 3). In this study, identification of the IncX3-type pNDM-MGR194-like plasmid in E. coli of different STs, K. pneumoniae and C. freundii strains indicated that this plasmid could mediate inter- and intra-species transfer of blaNDM-5. This possibility was further supported by our conjunction experimental data in vitro. Moreover, this plasmid carried in E. coli and C. freundii strains was isolated from faeces sample of a single patient at the same time, providing strong evidence that this plasmid could mediate blaNDM-5 dissemination in Enterobacteriaceae. Overall, our results revealed that IncX3-type pNDM-MGR194-like plasmids facilitate the rapid dissemination of blaNDM-5 among Enterobacteriaceae in China.

Table 3 Detailed information of the blaNDM-5-habouring plasmids reported in the NCBI database

Conclusions

We report a near-term epidemiological study demonstrating the further dissemination of Enterobacteriaceae with the blaNDM-5 gene in China. Our work provides evidence that the IncX3-type plasmid played an important role in the dissemination of blaNDM-5 in Enterobacteriaceae. In addition, to the best of our knowledge, this report is the first to isolate E. coli and C. freundii strains carrying blaNDM-5 from a single patient. Close surveillance is urgently needed to monitor the further spread of NDM-5-producing isolates.