https://orcid.org/0000-0002-9374-0811
Abstract
To characterize the clonal spread of carbapenem-resistant Klebsiella pneumoniae and Escherichia coli isolates between different healthcare institutions in Catalonia, Spain.
Antimicrobial susceptibility was tested by disc diffusion. MICs were determined by gradient diffusion or broth microdilution. Carbapenemase production was confirmed by lateral flow. PCR and Sanger sequencing were used to identify the allelic variants of resistance genes. Clonality studies were performed by PFGE and MLST. Plasmid typing, conjugation assays, S1-PFGE plus Southern blotting and MinION Oxford Nanopore sequencing were used to characterize resistance plasmids.
Twenty-nine carbapenem-resistant isolates recovered from three healthcare institutions between January and November 2016 were included: 14 K. pneumoniae isolates from a tertiary hospital in the south of Catalonia (hospital A); 2 K. pneumoniae isolates from a nearby healthcare centre; and 12 K. pneumoniae isolates and 1 E. coli isolate from a tertiary hospital in Barcelona (hospital B). The majority of isolates were resistant to all antimicrobial agents, except colistin, and all were NDM producers. PFGE identified a major K. pneumoniae clone (n = 27) belonging to ST147 and co-producing NDM-1 and CTX-M-15, with a few isolates also harbouring blaOXA-48. Two sporadic isolates of K. pneumoniae ST307 and E. coli ST167 producing NDM-7 were also identified. blaNDM-1 was carried in two related IncR plasmid populations and blaNDM-7 in a conjugative 50 kb IncX3 plasmid.
We report the inter-hospital dissemination of XDR high-risk clones of K. pneumoniae and E. coli associated with the carriage of small, transferable plasmids harbouring blaNDM genes.
Introduction
Carbapenem resistance has been rising very rapidly during recent decades, hence reducing the treatment options that are available to tackle infections caused by the main Gram-negative nosocomial pathogens, such as Acinetobacter baumannii, Pseudomonas aeruginosa and members of Enterobacterales, all of which are top-priority pathogens according to the global priority list of antibiotic-resistant bacteria from the WHO.1 The main mechanism of carbapenem resistance among these organisms is the production of carbapenem-hydrolysing β-lactamases, such as KPC, GES and OXA enzymes and MBLs (e.g. IMP, VIM and NDM).2 In particular, there is great concern regarding the dissemination of NDM-producing Gram-negative bacteria, since carriage of the blaNDM gene is usually associated with resistance to all β-lactam antibiotics, except monobactams, plus co-resistance to additional antibiotic families, such as quinolones and aminoglycosides.3,4 Since the initial identification of NDM-1 in a Swedish patient of Indian origin in 2008,5 NDM-producing Gram-negative bacteria have been reported worldwide and up to 29 different NDM allelic variants are recorded in the NCBI reference gene catalogue (last accessed 9 June 2020).3,4 Among Enterobacterales, NDM has been described in several species, but Escherichia coli and Klebsiella pneumoniae seem to be the most frequent hosts. In Spain, NDM was first described in 2011 in E. coli and only a few sporadic cases and outbreaks have been reported since.6,7 Here we have examined and characterized the clonal spread of NDM-producing K. pneumoniae between different healthcare institutions in Catalonia, Spain.
Materials and methods
Ethics
Bacterial samples studied here were recovered from clinical samples used for microbiological diagnosis at clinical microbiology laboratories. Informed consent was, therefore, not required. The protocol for this study was approved by the Ethics Committee on Clinical Research (CEIC) of the Hospital Clinic de Barcelona (HCB/2014/0499, HCB/2017/0923 and HCB/2017/0833).
Bacterial samples
Twenty-nine carbapenem-resistant isolates from three healthcare institutions were included: 14 K. pneumoniae isolates recovered from January to October 2016 from a tertiary hospital in the south of Catalonia (hospital A); 12 K. pneumoniae isolates and 1 E. coli isolate recovered from July to November 2016 at a tertiary hospital in Barcelona (hospital B); and 2 K. pneumoniae isolates recovered in September and October 2016, respectively, from a primary healthcare centre also in the south of Catalonia. Isolates were from surveillance and clinical samples. Identification of species was performed by MALDI-TOF MS. The clinical and microbiological data from all isolates and patients are provided in Table S1 (available as Supplementary data at JAC Online).
Susceptibility testing and resistance
Antimicrobial susceptibility was determined by disc diffusion in agar plates following EUCAST guidelines. Carbapenemase-producing Enterobacterales were selected according to EUCAST screening cut-off values for carbapenems.8 Production of KPC, OXA-48-like, VIM, IMP or NDM carbapenemases was detected using NG-Test® CARBA 5 (NG-Biotech, France). MICs were determined by gradient diffusion (AB-bioMérieux, Sweden), except for colistin MICs, which were determined by broth microdilution.9 Results were interpreted according to EUCAST guidelines.10 Isolates were categorized as MDR, XDR or pan-drug resistant according to ad hoc definitions.11E. coli ATCC 25922 was used for quality control.
The presence of genes encoding carbapenemases,7,12 ESBLs13 or 16S rRNA methyltransferases (armA and rmtA-rmtH)14 was investigated by PCR and Sanger sequencing. Alleles were determined through sequence alignment against the NCBI reference gene catalogue (PRJNA313047; last accessed 9 June 2020).
Epidemiology and molecular typing
Clonality was studied by PFGE using XbaI genomic digestions and a CHEFF-DRIII system (Bio-Rad, Spain).15 Molecular patterns were analysed with InfoQuest™FP-v.5.4 (Bio-Rad) and the unweighted pair group method with arithmetic mean to create dendrograms based on Dice’s similarity coefficient, using bandwidth tolerance and optimization values set at 1.5% and 1.2%, respectively. Isolates were considered within the same PFGE cluster (pulsotype) if their Dice similarity index was >85%.
MLST was performed according to the Pasteur scheme for K. pneumoniae and the Achtman scheme for E. coli.16,17
Plasmid analysis
Transferability of blaNDM was studied by biparental conjugation in broth medium using azide-resistant E. coli J53 as the recipient. Transconjugant strains (TC1–4) were selected on LB agar plates containing 1 mg/L imipenem and 100 mg/L sodium azide. Plasmid profiling was performed by S1-nuclease digestion followed by PFGE and Southern hybridization with digoxigenin-labelled probes against blaNDM, blaOXA-48 and blaCTX-M.
Plasmid incompatibility groups were identified using the PBRT-2.0 kit (Diatheva, Italy).18 Classification of IncR plasmids into IncR1 or IncR2 arbitrary groups was performed by PCR using the following primers: IS26_Rev, 5’-ggcactgttgcaaagttagcg-3’; ISAba125_Rev, 5’-caaacatgaggtgcgacag-3’; Tn5403Int_Fwd, 5’-ggtttgcgtgacatcacttcg-3’; and Tn5403Int_Rev, 5’-ccgtgagtgtggctttagag-3’. Plasmids belonged to the IncR1 group if the PCR was negative upon using primers IS26_Rev with ISAba125_Rev, but positive when combining IS26_Rev with Tn5403Int_Rev and Tn5403Int_Fwd with ISAba125_Rev (3 and 4.3 kb, respectively). Plasmids belonged to the IncR2 group if positive to the IS26_Rev and ISAba125_Rev primer combination (1.7 kb), but negative for the other two primer pairs.
Genomic DNA extracted using the Wizard Genomic DNA purification kit (Promega, Spain) was sequenced using MinION sequencing (Oxford Nanopore, UK), in accordance with the manufacturer. Base-calling was done with Guppy-v3.0.3 and demultiplexing with qcat-v1.1.0 (https://github.com/nanoporetech/qcat). FASTQ files were mapped using Minimap2-v2.17 against plasmids from Enterobacterales.19 Mapping reads were assembled with Flye-v2.5 (https://github.com/fenderglass/Flye). Annotation was done with Prokka-v.1.12 combined with BLASTP/BLASTN searches against the UniProtKB/Swiss-Prot and RefSeq databases.20
ResFinder (https://cge.cbs.dtu.dk/services/ResFinder/), PlasmidFinder (https://cge.cbs.dtu.dk/services/PlasmidFinder/) and ISFinder (https://www-is.biotoul.fr/) were used to identify antimicrobial resistance genes, mobile elements and plasmid replicons. Gene organization diagrams were drawn using SnapGene® Viewer-v5.1.2 (https://www.snapgene.com/) and CGViewAdvanced-v.0.0.1.21 Sequence comparisons were graphically displayed using Kablammo.22
FASTQ files of isolates HA-2, HA-3, HA-4, HB-377 and HB-536 have been deposited into the NCBI Sequence Read Archive (SRA) under accession numbers SRR11828896, SRR11828895, SRR11828894, SRR18228893 and SRR11828892, respectively; BioProject PRJNA6346391. Annotated plasmid sequences are available as Supplementary data at JAC Online.
Results
Bacterial isolation and PFGE
Overall, 24 different patients were involved in the study, being either colonized (n = 7) or infected (n = 17), and typically presented multiple comorbidities (mostly hepatic, pancreatic and cardiovascular diseases). Five of the infected patients died (29%). The most common treatment for infected patients was the administration of carbapenems together with tigecycline and/or colistin (Table S1).
In January 2016, a carbapenem-resistant NDM-producing K. pneumoniae isolate (HA-3) was recovered at a tertiary hospital (hospital A) in the province of Tarragona, Spain, from a urine sample of a patient admitted to the internal medicine ward who had just been transferred from a tertiary hospital in Barcelona (hospital D). From April to June 2016, four additional NDM-producing K. pneumoniae isolates were recovered from the surveillance and clinical samples of three different patients admitted to the internal medicine ward of hospital A.
At the beginning of July, an NDM-positive patient from hospital A was transferred to the liver ICU of a second university hospital in Barcelona (hospital B). The patient was isolated and enhanced barrier precautions were implemented upon admission. Nevertheless, 11 additional NDM-producing K. pneumoniae isolates and 1 NDM-producing E. coli isolate were recovered from different wards from July to November.
During the same period, NDM-producing K. pneumoniae continued to disseminate at hospital A and eight new isolates were reported in both the internal medicine and surgery wards, two of them recovered from a newly admitted patient transferred again from hospital D. Furthermore, two additional NDM-producing K. pneumoniae isolates were also reported in September and October 2016, respectively, from a primary healthcare centre close to hospital A. Infection control measures and active screening of both carriers and environmental samples were intensified in all centres during this period and, at hospital B, hydrogen peroxide vaporizers were even used to decontaminate those wards involved in the outbreak. No additional isolates were recovered after November 2016 (Figure 1).
Temporal and spatial distribution of NDM-producing isolates recovered from patients at three healthcare institutions in Catalonia from January to November 2016. Coloured boxes show the different wards hosting the patients at the time of isolation. Bacterial isolates are represented by an inverted triangle (K. pneumoniae) or a circle (E. coli). Colour codes indicate carriage of different combinations of β-lactamases and IncR or IncX3 plasmids. Isolates recovered from patients referred from hospital D are shown with an asterisk. The transfer of patients between the three healthcare centres is represented with a dotted line. NS indicates isolates selected for Nanopore sequencing. Superscript numbers indicate isolates recovered from the same patient (Table S1). This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.
The analysis of the 28 K. pneumoniae isolates by PFGE revealed the presence of two different pulsotypes, A and B (Figure 2). Most of the isolates were highly related and clustered together into pulsotype A, while pulsotype B contained a single strain recovered at hospital B (HB-536). Pulsotype A could be further subdivided into clusters A1 and A2 (20 and 7 isolates, respectively), differing by only one band and sharing 96% similarity (Figure 2). Isolates from pulsotype A1 were recovered from all three centres, while those from the A2 pulsotype were exclusively from hospital A.
PFGE dendrogram of NDM-producing K. pneumoniae isolates from three healthcare centres in Catalonia (HA, hospital A; HB, hospital B; and HC, primary healthcare centre). Other, carriage of additional β-lactamases; PT, pulsotype; PG, plasmid incompatibility group of plasmids harbouring blaNDM; NS, isolates selected for Nanopore sequencing. Braces indicate classification to the corresponding PFGE cluster or pulsotype. Isolates were included in the same pulsotype if their Dice similarity index was ≥85%. Superscript numbers indicate isolates recovered from the same patient (Table S1). This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.
Antimicrobial resistance
Antimicrobial susceptibility testing showed that 25 out of the 27 pulsotype A K. pneumoniae isolates were XDR, only remaining susceptible to colistin. Interestingly, the two pulsotype A1 isolates recovered from the primary healthcare centre (HC-15 and HC-16) were also susceptible to all aminoglycosides and, therefore, considered MDR. The single K. pneumoniae isolate from pulsotype B (HB-536), as well as the NDM-producing E. coli isolate (HB-543), were also MDR, remaining susceptible to either amikacin, fosfomycin and colistin, or to gentamicin, amikacin, tobramycin, fosfomycin, tigecycline and colistin, respectively (Table S1).
PCR screening confirmed carriage of blaNDM in all K. pneumoniae and E. coli isolates, as well as blaCTX-M-group 1 in all K. pneumoniae isolates. K. pneumoniae isolates from pulsotype A harboured the blaNDM-1 allelic variant, but both the K. pneumoniae isolate from pulsotype B (HB-536) and the single NDM-producing E. coli isolate (HB-543) carried blaNDM-7. The β-lactamase gene blaTEM-1 was also identified in these two isolates. DNA sequencing also confirmed the blaCTX-M-15 allelic variant among K. pneumoniae isolates. Of note, the rmtF gene was detected in the 25 XDR K. pneumoniae isolates from pulsotype A1.
All isolates were negative for blaKPC, blaVIM and blaIMP, but four K. pneumoniae isolates from pulsotype A2 also carried the blaOXA-48 gene, encoding a class D carbapenem-hydrolysing β-lactamase (Figure 2). Interestingly, isolates co-carrying OXA-48 and NDM were recovered at hospital A after the outbreak had been extended to hospital B, and the first patient at hospital A with NDM/OXA-48 (HA-4) represents a second referral from hospital D, the original source of the outbreak at hospital A (Figure 1).
Molecular typing and plasmid analysis
MLST studies identified isolates from pulsotype A as belonging to ST147, while the NDM-7-producing K. pneumoniae (pulsotype B) and E. coli isolates belonged to ST307 and ST167, respectively.
The transferability of putative plasmids harbouring blaNDM-1, blaCTX-M-15, blaOXA-48 and blaNDM-7 was analysed by conjugation using K. pneumoniae producing NDM-1/OXA-48 or NDM-7 as donors. Four different types of transconjugant E. coli strains were obtained: TC1, carrying blaNDM-1, blaCTX-M-15 and blaOXA-48; TC2, carrying blaNDM-1 and blaCTX-M-15; TC3, carrying just blaOXA-48; and TC4, that only acquired blaNDM-7. None of the transconjugant strains acquired the rmtF gene. TC1 and TC2 transconjugants acquired resistance to all β-lactam antibiotics, but not to aminoglycosides, while TC3 transconjugants only showed reduced susceptibility to β-lactams and TC4 transconjugants acquired resistance to all β-lactams, except aztreonam. The MIC values and molecular characteristics for representative isolates and transconjugant strains are shown in Table 1.
Antimicrobial susceptibility and molecular characterization of representative K. pneumoniae and E. coli isolates and their corresponding transconjugant E. coli strains (using azide-resistant E. coli J53 as the recipient)
| Strain | Bacterial species | MIC (mg/L) | Pulsotype | ST | NDM | Other | rmtF | PG | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| IPM | MEM | FOX | FEP | CTX | CAZ | ATM | GEN | AMK | TOB | CIP | LVX | TGC | CSTa | FOF | ||||||||
| HA-3 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.25 | 64 | A2 | ST147 | 1 | CTX-M-15 | + | R1 |
| HA-2 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.125 | 48 | A1 | ST147 | 1 | CTX-M-15 | + | R2 |
| HB-377 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 3 | 0.25 | 48 | A1 | ST147 | 1 | CTX-M-15 | + | R2 |
| HA-4 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.125 | 64 | A2 | ST147 | 1 | OXA-48 CTX-M-15 | + | R1/L |
| HC-16 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | 2 | 2 | 6 | >32 | >32 | 4 | 0.25 | 48 | A1 | ST147 | 1 | CTX-M-15 | – | R2 |
| HB-536 | Kp | >32 | >32 | >256 | 256 | >32 | >256 | 48 | 24 | 4 | 8 | >32 | >32 | 4 | 0.25 | 12 | B | ST307 | 7 | TEM-1 CTX-M-15 | – | X3 |
| HB-543 | Ec | >32 | >32 | >256 | >256 | >32 | >256 | >256 | 0.25 | 1 | 0.38 | >32 | >32 | 0.75 | 0.25 | 0.38 | NA | ST167 | 7 | TEM-1 | – | X3 |
| TC1 | Ec | 16 | 4 | >256 | 32 | >32 | >256 | 32 | 0.125 | 0.5 | 0.38 | 0.125 | 0.25 | 0.25 | 0.125 | 1 | NA | NA | 1 | OXA-48 CTX-M-15 | – | R1/L |
| TC2 | Ec | 12 | 2 | >256 | 16 | >32 | >256 | 12 | 0.25 | 0.5 | 0.25 | 0.125 | 0.38 | 0.38 | 0.125 | 1 | NA | NA | 1 | CTX-M-15 | – | R1 |
| TC3 | Ec | 0.5 | 0.19 | 4 | 0.64 | 0.5 | 0.19 | 0.094 | 0.38 | 1.5 | 0.25 | 0.08 | 0.023 | 0.19 | 0.125 | 1 | NA | NA | – | OXA-48 | – | L |
| TC4 | Ec | >32 | >32 | >256 | 8 | >32 | >256 | 0.125 | 0.38 | 1.5 | 0.25 | 0.012 | 0.016 | 0.125 | 0.125 | 1 | NA | NA | 7 | – | – | X3 |
| J53AzR | Ec | 0.19 | 0.016 | 4 | 0.023 | 0.016 | 0.047 | 0.125 | 0.064 | 0.75 | 0.094 | 0.008 | 0.016 | 0.094 | 0.125 | 0.094 | NA | NA | – | – | – | NA |
| Strain | Bacterial species | MIC (mg/L) | Pulsotype | ST | NDM | Other | rmtF | PG | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| IPM | MEM | FOX | FEP | CTX | CAZ | ATM | GEN | AMK | TOB | CIP | LVX | TGC | CSTa | FOF | ||||||||
| HA-3 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.25 | 64 | A2 | ST147 | 1 | CTX-M-15 | + | R1 |
| HA-2 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.125 | 48 | A1 | ST147 | 1 | CTX-M-15 | + | R2 |
| HB-377 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 3 | 0.25 | 48 | A1 | ST147 | 1 | CTX-M-15 | + | R2 |
| HA-4 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.125 | 64 | A2 | ST147 | 1 | OXA-48 CTX-M-15 | + | R1/L |
| HC-16 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | 2 | 2 | 6 | >32 | >32 | 4 | 0.25 | 48 | A1 | ST147 | 1 | CTX-M-15 | – | R2 |
| HB-536 | Kp | >32 | >32 | >256 | 256 | >32 | >256 | 48 | 24 | 4 | 8 | >32 | >32 | 4 | 0.25 | 12 | B | ST307 | 7 | TEM-1 CTX-M-15 | – | X3 |
| HB-543 | Ec | >32 | >32 | >256 | >256 | >32 | >256 | >256 | 0.25 | 1 | 0.38 | >32 | >32 | 0.75 | 0.25 | 0.38 | NA | ST167 | 7 | TEM-1 | – | X3 |
| TC1 | Ec | 16 | 4 | >256 | 32 | >32 | >256 | 32 | 0.125 | 0.5 | 0.38 | 0.125 | 0.25 | 0.25 | 0.125 | 1 | NA | NA | 1 | OXA-48 CTX-M-15 | – | R1/L |
| TC2 | Ec | 12 | 2 | >256 | 16 | >32 | >256 | 12 | 0.25 | 0.5 | 0.25 | 0.125 | 0.38 | 0.38 | 0.125 | 1 | NA | NA | 1 | CTX-M-15 | – | R1 |
| TC3 | Ec | 0.5 | 0.19 | 4 | 0.64 | 0.5 | 0.19 | 0.094 | 0.38 | 1.5 | 0.25 | 0.08 | 0.023 | 0.19 | 0.125 | 1 | NA | NA | – | OXA-48 | – | L |
| TC4 | Ec | >32 | >32 | >256 | 8 | >32 | >256 | 0.125 | 0.38 | 1.5 | 0.25 | 0.012 | 0.016 | 0.125 | 0.125 | 1 | NA | NA | 7 | – | – | X3 |
| J53AzR | Ec | 0.19 | 0.016 | 4 | 0.023 | 0.016 | 0.047 | 0.125 | 0.064 | 0.75 | 0.094 | 0.008 | 0.016 | 0.094 | 0.125 | 0.094 | NA | NA | – | – | – | NA |
Kp, K. pneumoniae; Ec, E. coli; IPM, imipenem; MEM, meropenem; FOX, cefoxitin; FEP, cefepime; CTX, cefotaxime; CAZ, ceftazidime; ATM, aztreonam; GEN, gentamicin; AMK, amikacin; TOB, tobramycin; CIP, ciprofloxacin; LVX, levofloxacin; TGC, tigecycline; CST, colistin; FOF, fosfomycin; NDM, presence of blaNDM-1 or blaNDM-7; Other, presence of additional bla genes; rmtF, presence of the rmtF gene; PG, plasmid incompatibility group of plasmids carrying blaNDM and blaOXA-48; NA, not applicable.
Colistin MICs were determined by broth microdilution.
Antimicrobial susceptibility and molecular characterization of representative K. pneumoniae and E. coli isolates and their corresponding transconjugant E. coli strains (using azide-resistant E. coli J53 as the recipient)
| Strain | Bacterial species | MIC (mg/L) | Pulsotype | ST | NDM | Other | rmtF | PG | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| IPM | MEM | FOX | FEP | CTX | CAZ | ATM | GEN | AMK | TOB | CIP | LVX | TGC | CSTa | FOF | ||||||||
| HA-3 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.25 | 64 | A2 | ST147 | 1 | CTX-M-15 | + | R1 |
| HA-2 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.125 | 48 | A1 | ST147 | 1 | CTX-M-15 | + | R2 |
| HB-377 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 3 | 0.25 | 48 | A1 | ST147 | 1 | CTX-M-15 | + | R2 |
| HA-4 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.125 | 64 | A2 | ST147 | 1 | OXA-48 CTX-M-15 | + | R1/L |
| HC-16 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | 2 | 2 | 6 | >32 | >32 | 4 | 0.25 | 48 | A1 | ST147 | 1 | CTX-M-15 | – | R2 |
| HB-536 | Kp | >32 | >32 | >256 | 256 | >32 | >256 | 48 | 24 | 4 | 8 | >32 | >32 | 4 | 0.25 | 12 | B | ST307 | 7 | TEM-1 CTX-M-15 | – | X3 |
| HB-543 | Ec | >32 | >32 | >256 | >256 | >32 | >256 | >256 | 0.25 | 1 | 0.38 | >32 | >32 | 0.75 | 0.25 | 0.38 | NA | ST167 | 7 | TEM-1 | – | X3 |
| TC1 | Ec | 16 | 4 | >256 | 32 | >32 | >256 | 32 | 0.125 | 0.5 | 0.38 | 0.125 | 0.25 | 0.25 | 0.125 | 1 | NA | NA | 1 | OXA-48 CTX-M-15 | – | R1/L |
| TC2 | Ec | 12 | 2 | >256 | 16 | >32 | >256 | 12 | 0.25 | 0.5 | 0.25 | 0.125 | 0.38 | 0.38 | 0.125 | 1 | NA | NA | 1 | CTX-M-15 | – | R1 |
| TC3 | Ec | 0.5 | 0.19 | 4 | 0.64 | 0.5 | 0.19 | 0.094 | 0.38 | 1.5 | 0.25 | 0.08 | 0.023 | 0.19 | 0.125 | 1 | NA | NA | – | OXA-48 | – | L |
| TC4 | Ec | >32 | >32 | >256 | 8 | >32 | >256 | 0.125 | 0.38 | 1.5 | 0.25 | 0.012 | 0.016 | 0.125 | 0.125 | 1 | NA | NA | 7 | – | – | X3 |
| J53AzR | Ec | 0.19 | 0.016 | 4 | 0.023 | 0.016 | 0.047 | 0.125 | 0.064 | 0.75 | 0.094 | 0.008 | 0.016 | 0.094 | 0.125 | 0.094 | NA | NA | – | – | – | NA |
| Strain | Bacterial species | MIC (mg/L) | Pulsotype | ST | NDM | Other | rmtF | PG | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| IPM | MEM | FOX | FEP | CTX | CAZ | ATM | GEN | AMK | TOB | CIP | LVX | TGC | CSTa | FOF | ||||||||
| HA-3 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.25 | 64 | A2 | ST147 | 1 | CTX-M-15 | + | R1 |
| HA-2 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.125 | 48 | A1 | ST147 | 1 | CTX-M-15 | + | R2 |
| HB-377 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 3 | 0.25 | 48 | A1 | ST147 | 1 | CTX-M-15 | + | R2 |
| HA-4 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | >256 | >256 | >256 | >32 | >32 | 2 | 0.125 | 64 | A2 | ST147 | 1 | OXA-48 CTX-M-15 | + | R1/L |
| HC-16 | Kp | >32 | >32 | >256 | >256 | >32 | >256 | >256 | 2 | 2 | 6 | >32 | >32 | 4 | 0.25 | 48 | A1 | ST147 | 1 | CTX-M-15 | – | R2 |
| HB-536 | Kp | >32 | >32 | >256 | 256 | >32 | >256 | 48 | 24 | 4 | 8 | >32 | >32 | 4 | 0.25 | 12 | B | ST307 | 7 | TEM-1 CTX-M-15 | – | X3 |
| HB-543 | Ec | >32 | >32 | >256 | >256 | >32 | >256 | >256 | 0.25 | 1 | 0.38 | >32 | >32 | 0.75 | 0.25 | 0.38 | NA | ST167 | 7 | TEM-1 | – | X3 |
| TC1 | Ec | 16 | 4 | >256 | 32 | >32 | >256 | 32 | 0.125 | 0.5 | 0.38 | 0.125 | 0.25 | 0.25 | 0.125 | 1 | NA | NA | 1 | OXA-48 CTX-M-15 | – | R1/L |
| TC2 | Ec | 12 | 2 | >256 | 16 | >32 | >256 | 12 | 0.25 | 0.5 | 0.25 | 0.125 | 0.38 | 0.38 | 0.125 | 1 | NA | NA | 1 | CTX-M-15 | – | R1 |
| TC3 | Ec | 0.5 | 0.19 | 4 | 0.64 | 0.5 | 0.19 | 0.094 | 0.38 | 1.5 | 0.25 | 0.08 | 0.023 | 0.19 | 0.125 | 1 | NA | NA | – | OXA-48 | – | L |
| TC4 | Ec | >32 | >32 | >256 | 8 | >32 | >256 | 0.125 | 0.38 | 1.5 | 0.25 | 0.012 | 0.016 | 0.125 | 0.125 | 1 | NA | NA | 7 | – | – | X3 |
| J53AzR | Ec | 0.19 | 0.016 | 4 | 0.023 | 0.016 | 0.047 | 0.125 | 0.064 | 0.75 | 0.094 | 0.008 | 0.016 | 0.094 | 0.125 | 0.094 | NA | NA | – | – | – | NA |
Kp, K. pneumoniae; Ec, E. coli; IPM, imipenem; MEM, meropenem; FOX, cefoxitin; FEP, cefepime; CTX, cefotaxime; CAZ, ceftazidime; ATM, aztreonam; GEN, gentamicin; AMK, amikacin; TOB, tobramycin; CIP, ciprofloxacin; LVX, levofloxacin; TGC, tigecycline; CST, colistin; FOF, fosfomycin; NDM, presence of blaNDM-1 or blaNDM-7; Other, presence of additional bla genes; rmtF, presence of the rmtF gene; PG, plasmid incompatibility group of plasmids carrying blaNDM and blaOXA-48; NA, not applicable.
Colistin MICs were determined by broth microdilution.
These results suggested that ST147 K. pneumoniae isolates carried blaNDM-1 and blaCTX-M-15 in the same plasmid, while blaOXA-48 and rmtF were likely to be located in two separate plasmids. In contrast, blaNDM-7 and blaCTX-M-15 were apparently located in different plasmids in the ST307 K. pneumoniae isolate HB-536. To corroborate these results, the location of the blaNDM-1, blaCTX-M-15,blaOXA-48 and blaNDM-7 genes was investigated by S1-PFGE and Southern blotting. As shown in Figure 3(a and b), DNA probes specific for blaNDM and blaCTX-M-15 hybridized with the same bands in NDM-1-producing K. pneumoniae (ST147) isolates and the corresponding transconjugant strains (TC1 and TC2), while the same probes hybridized with different bands in the NDM-7-producing K. pneumoniae (HB-536) isolate and the corresponding transconjugant strain (TC4) and only the blaNDM probe hybridized against the NDM-7-producing E. coli isolate (HB-543). In addition, DNA probes against blaOXA-48 hybridized with a smaller band in an isolate co-producing NDM-1/OXA-48 (HA-4) and the same band was also identified in the corresponding OXA-48 (TC3) or NDM-1/OXA-48 (TC1) transconjugant strains (Figure 3c).
S1-PFGE and Southern hybridization of K. pneumoniae or E. coli isolates harbouring different combinations of β-lactamases. (a) Hybridization with the blaNDM probe. (b) Hybridization with the blaCTX-M-15 probe. (c) Hybridization with the blaOXA-48 probe. Braces indicate molecular detection of different bla genes. Transconjugants are shown in blue. Arrows indicate weak bands. Of note, mirroring double bands were attributed to incomplete cleavage by the S1 nuclease so that other conformations of the plasmid rather than the linear one are also seen. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.
PBRT showed that K. pneumoniae isolates producing NDM-1 or NDM-1/OXA-48 were positive for IncR and IncFIIS, or IncR, IncFIIS and IncL plasmid replicons, respectively, while the corresponding E. coli transconjugants presented the following plasmid replicons: TC1 (NDM-1/OXA-48), positive for IncR and IncL; TC2 (NDM-1), positive for IncR; and TC3 (OXA-48), positive for IncL. The K. pneumoniae isolate producing NDM-7 (HB-536) presented IncX3, IncFIB-KN and IncFIIk plasmid replicons, but only the IncX3 replicon was transferred to the E. coli transconjugant (TC4). Likewise, the single E. coli clinical isolate producing NDM-7 also harboured an IncX3 plasmid replicon (Table 1).
Plasmid sequences
K. pneumoniae isolates initially selected to characterize the plasmids harbouring blaNDM and blaOXA-48 genes included: the first NDM-1-producing isolates at both hospital A and hospital B (HA-3 and HB-377, respectively); the first NDM-1/OXA-48-producing isolate that was recovered at hospital A (HA-4) [all three of them belonging to ST147 (Figures 1 and 2)]; and the ST307 K. pneumoniae isolate producing NDM-7 (HB-536).
Genomic sequencing corroborated previous results showing that ST147 K. pneumoniae isolates harboured both the blaNDM-1 and blaCTX-M-15 genes in a single plasmid belonging to the IncR incompatibility group (Figures S1 to S3). There were, however, some interesting differences. The HA-3 and HA-4 isolates, both recovered at hospital A from patients referred from hospital D, shared almost identical plasmids of approx. 74 kb (pNDM-HA3 and pNDM-HA4, respectively) that carried blaNDM-1 within a Tn3000 transposon. In both plasmids the upstream IS3000 element was partially replaced by a full-length Tn5403, as previously described.6 The Tn3000 transposon was in turn flanked by two IS26 elements in reverse orientation (Figure 4). The bulk of the remaining plasmid backbone shared high similarity (>99% average identity) with p48896_1 (CP024430), an IncR plasmid carrying blaCTX-M-15 (but not blaNDM) that was recently recovered from an ST147 K. pneumoniae isolate in Pakistan.23
Schematic drawing showing the genetic elements surrounding the blaNDM genes in pNDM-HA3, pNDM-HB377 and pNDM-HB536, as well as the blaOXA-48 gene in pOXA48-HA4. Arrows are proportional to the lengths of the genes and are orientated in the direction of transcription. Red arrows represent resistance genes, orange arrows represent full-length transposon-related genes and ISs, blue arrows represent partial or truncated transposon-related genes and ISs, and purple arrows are used for the remaining set of genes. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.
Additional resistance genes included the aminoglycoside resistance gene aph(3’)-Ia, as well as qnrB, involved in low-level quinolone resistance.24 On the other hand, plasmid pNDM-HB377 (from the first isolate recovered at hospital B) presented a slightly smaller IncR plasmid of approx. 67 kb, also containing both blaNDM-1 and blaCTX-M-15. The genetic structures surrounding blaNDM-1 in pNDM-HB377 were similar to those of pNDM-HA3 and pNDM-HA4, except for a 5494 bp sequence containing Tn5403 that was missing upstream of blaNDM-1 (Figure 4). In addition, pNDM-HB377 presented an inversion of a 52 kb region containing blaCTX-M-15, the IncR repB gene and mobC, involved in plasmid mobilization. The region containing the aminoglycoside resistance gene aph(3’)-Ia was also missing (Figure 5).
Graphical comparison of plasmids pNDM-HA3 and pNDM-HB377 in either: (a) parallel orientation or (b) antiparallel orientation (reverse complement sequence of pNDM-HB377). The shaded stripes show regions shared between the two plasmids. The locations of the ltrA sequences flanking a 52 kb inversion are shown. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.
These findings suggested a slightly different plasmid population between NDM-producing K. pneumoniae isolates from hospitals A and B. To corroborate such differences, specific PCR primers (see the Materials and methods section) were designed to differentiate between pNDM-HA3/HA4-like plasmids (hereafter IncR1 group) and pNDM-HB377-like plasmids (hereafter IncR2 group). PCR screening showed that all NDM-1-producing K. pneumoniae isolates from hospitals B and C carried blaNDM-1 in a plasmid from the IncR2 group, while isolates from hospital A carried blaNDM-1 in plasmids from either group (Figure 2). Sequence analysis of yet another IncR2 plasmid (pNDM-HA2), but from an isolate at hospital A, also yielded a plasmid of roughly 67 kb, almost identical to pNDM-HB377 (Figure S4). Interestingly, isolates carrying the IncR2 group plasmid belonged to pulsotype A1, while isolates harbouring the IncR1 group belonged to pulsotype A2 (Figure 2).
The immediate genetic structures downstream of blaNDM-1 were identical in these plasmids and comprised genes commonly associated with blaNDM genes in Enterobacterales and Acinetobacter spp. (Figure 4).25,26
In addition, the HA-4 isolate harboured a second plasmid of approx. 64 kb carrying blaOXA-48 and belonging to IncL (pOXA48-HA4) that was highly similar (>99% average identity and 100% query coverage) to the IncL pUR17313-1 plasmid (KP061858) sequenced in Portugal from an Enterobacter cloacae clinical isolate (Figure S5).27,blaOXA-48 was located within Tn1999.2 (Figure 4) and no additional antimicrobial resistance genes were identified in pOXA48-HA4.28
On the other hand, the ST307 K. pneumoniae isolate HB-536 carried blaNDM-7 (but not blaCTX-M-15) in an IncX3 plasmid of 50.5 kb (pNDM-HB536) highly similar (>98% average identity and 99% query coverage) to the pAR_0162 plasmid recovered from E. coli (CP021682) (Figure S6). The sequence upstream of blaNDM-7 comprised several partial or complete ISs and transposons (again Tn5403 and IS3000) among which stood out the presence of an ISAba125 that was truncated by an IS5 element. The region downstream of blaNDM-7 also presented the canonical ble, trpF and dsbC genes, but the adjacent cutA1 gene was partially deleted upon the insertion of an IS26 element (Figure 4). The blaNDM-7 gene was the only antimicrobial resistance gene present in plasmid pNDM-HB536.
Discussion
NDM was first reported in Spain in 2011 from an E. coli isolate in a Spanish traveller returning from India.7 Since then, NDM-producing Enterobacterales have slowly crawled into the Spanish healthcare system, initially associated with imported sporadic cases, then autochthonous and, more recently, causing small hospital outbreaks.25,29–35
A recent nationwide study comparing the genome sequences of NDM-producing K. pneumoniae and E. coli isolates in Spain also suggested the inter-hospital and inter-regional spread of a few clonal lineages.6
Here we report the inter-hospital dissemination of NDM-1-producing K. pneumoniae isolates between at least three healthcare settings in Catalonia from January to October 2016, with some isolates also co-producing OXA-48. A single clone carrying both blaNDM-1 and blaCTX-M-15 in the same transferable IncR plasmid and belonging to the high-risk clone ST147 was responsible for such dissemination. Nevertheless, three distinct populations of ST147 isolates were identified according to the number and types of plasmids carried.
The first population clustered together into pulsotype A2 was associated with the isolate from the index case patient at hospital A, probably acquired from a previous hospital admission (hospital D), and carried blaNDM-1 within an IncR1 plasmid of approx. 74 kb characterized by the presence of a Tn5403 insertion upstream of blaNDM-1 (Figure 2). The second population is a subset of the previous one, ST147 K. pneumoniae isolates with the IncR1 plasmid and belonging to pulsotype A2, but co-producing OXA-48 from an IncL plasmid. This population appeared at hospital A after a second patient transfer from hospital D. These two populations only disseminated among patients from hospital A (Figures 1 and 2). Interestingly, OXA-48-producing K. pneumoniae in Spain is commonly found in many hospitals and associated with several lineages, including ST405, ST15 and ST101, but, to our knowledge, not ST147.12,36 Most likely, the identification of OXA-48 associated with ST147 in this study reflects the acquisition of the IncL plasmid from OXA-48-producing strains previously circulating at hospital D, the original source of the outbreak.
The third population comprises ST147 K. pneumoniae isolates carrying blaNDM-1 and blaCTX-M-15 in an IncR2 plasmid of approx. 67 kb, highly similar to IncR1, except for a 52 kb inversion and the loss of a region upstream of blaNDM-1 (Figures 4 and 5).
IncR2 plasmids may have resulted from homologous recombination through ltrA sequences at an IncR1 plasmid that developed into this particular inversion, since ltrA genes were located flanking the inverted region. Isolates within this population belonged to pulsotype A1 and appeared first at hospital A, but further spread to other centres through the referral of patients (Figures 1 and 2).
K. pneumoniae ST147 is commonly associated with carbapenem resistance and is considered one of the main high-risk clones of MDR K. pneumoniae currently disseminating worldwide.37,38 In Spain, NDM-1-producing K. pneumoniae ST147 was first reported in 2018 causing a small outbreak at a tertiary hospital in Tenerife.34 More recently, a retrospective surveillance study has reported sporadic IncR NDM-1-producing ST147 K. pneumoniae isolates in the south of Catalonia during the summer of 2016, as well as causing another small outbreak in Galicia in 2015.6 The genetic structures surrounding blaNDM-1 in those isolates matched that of IncR1 plasmids in the present study, although the plasmid was slightly larger (110 kb) and lacked blaCTX-M-15. The index case in our study was referred from another hospital within Catalonia and, hence, it is not clear whether these ST147 isolates are related to previous outbreaks in Spain.
In our study, two sporadic isolates of K. pneumoniae and E. coli producing NDM-7 and belonging to ST307 and ST167, respectively, were identified in one of the hospitals. They were recovered from two patients that overlapped in the same ward and both carried blaNDM-7 in an IncX3 plasmid of approx. 50 kb, thus suggesting horizontal transfer. E. coli and K. pneumoniae isolates harbouring blaNDM-7 in similar IncX3 plasmids have also been described in Spain, mainly in Madrid, where they were associated with the dissemination of K. pneumoniae ST437 and Enterobacter hormaechei in 2013 and 2016, respectively.6,35 In Catalonia, NDM-7 has been reported twice in sporadic ST679 (2013) or ST399 (2015) E. coli isolates associated with IncX4 or IncX3 plasmids, respectively, both of which were recovered from patients that had just returned from Pakistan.29,39 In our study we could not relate either of these two patients to recent travel, although one of them was originally from Pakistan. Nevertheless, it is important to highlight that K. pneumoniae ST307 is also one of the main MDR K. pneumoniae high-risk clones and E. coli ST167 is considered an epidemic clone commonly associated with the worldwide dissemination of IncX3 plasmids harbouring blaNDM-5 and blaNDM-7.37,40
We acknowledge several limitations in our study. First, WGS was performed using a single Nanopore approach, while the use of a hybrid approach would have allowed for additional and more accurate comparisons with isolates from previous studies. Also, only a handful of isolates were sequenced and sequence similarity was, therefore, assumed for the remaining isolates on the basis of PFGE, MLST and conventional PCR data. Unfortunately, further WGS studies were beyond our possibilities. Finally, the index case and the spread of NDM-producing K. pneumoniae at hospital D was not known, although we are aware that an investigation is currently ongoing at the hospital and the results will be made available in due course.
Overall, the results presented here report the dissemination of XDR and MDR high-risk clones of K. pneumoniae and E. coli among different hospitals in Catalonia, whose success in the clinical setting is likely related to the carriage of small, transferable plasmids, harbouring blaNDM genes that provide resistance to last-resort antimicrobials. In Spain, such XDR clones are increasingly being reported in outbreak situations and autochthonous dissemination, and our results reinforce previous findings suggesting the national spread of NDM-producing K. pneumoniae associated with a few clonal lineages.
Acknowledgements
We thank the team of curators of the Institute Pasteur MLST and whole-genome MLST databases for curating the data and making them publicly available at http://bigsdb.pasteur.fr/.
We thank the clinical and laboratory staff of the participating hospitals for obtaining clinical samples and the initial identification of microorganisms.
Other members of the MERCyCAT Study Group
Pepa Pérez Jove, Emma Padilla and Mónica Ballestero-Téllez (Catlab, Centre analítiques Terrassa AIE), Yuliya Zboromyrska, Miguel Ángel Benítez, Raquel Clivillé, Sabina González and Iolanda Calvet (Consorci del Laboratori Intercomarcal de l'Alt Pendès, l'Anoia i el Garraf), Carmen Gallés (Corporació de Salut del Maresme i la Selva), Goretti Sauca (Hospital de Mataró), Carmina Martí-Sala and Ma Angeles Pulido (Hospital General de Granollers), Anna Vilamala (Hospital General de Vic), Araceli González-Cuevas, (Hospital General del Parc Sanitari Sant Joan de Déu), Amadeu Gené (Hospital Sant Joan de Déu de Barcelona), Gloria Trujillo and Joan Lopez Madueño (Hospital Sant Joan de Déu de Manresa), Xavier Raga (Hospital Sant Pau I Santa Tecla), Frederic Gómez, Ester Picó and Carolina Sarvisé (Hospital Universitari Joan XXIII de Tarragona) and Xesca Font (Hospital Universitari Sant Joan de Reus).
Funding
This study was supported by: Plan Nacional de I+D+i 2013–2016, Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Economía y Competitividad, Spanish Network for Research in Infectious Diseases (REIPI RD16/0016/0010); the 2017 call for Strategic Action on Health (PI17/01932), co-financed by European Development Regional Fund ‘A way to achieve Europe’ and the operative programme Intelligent Growth 2014–2020; and grant 2017 SGR 0809 from the Departament d’Universitats, Recerca i Societat de la Informació, of the Generalitat de Catalunya. M.M.-A. and C.C. were supported by grants FPU 14/06357 and FPU 13/02564, respectively, from the Spanish Ministry of Education, Culture and Sports. I.R. was supported by the Department of Health, Generalitat de Catalunya, grant SLT002/16/00349. We also acknowledge support from the Spanish Ministry of Science, Innovation and Universities through the ‘Centro de Excelencia Severo Ochoa 2019–2023’ programme (CEX2018-000806-S) and support from the Generalitat de Catalunya through the CERCA programme.
The funders had no role in the study design, data collection, analysis and interpretation of data, decision to publish or preparation of the manuscript.
Transparency declarations
None to declare.
Author contributions
All authors critically revised the manuscript for intellectual content and read and approved the final manuscript.
Supplementary data
Table S1, annotated plasmid sequences and Figures S1 to S6 are available as Supplementary data at JAC Online.
References
EUCAST. EUCAST Guidelines for Detection of Resistance Mechanisms and Specific Resistances of Clinical and/or Epidemiological Importance. Version 2.0.
EUCAST. Recommendations for MIC Determination of Colistin (Polymyxin E) as Recommended by the Joint CLSI-EUCAST Polymyxin Breakpoints Working Group.
EUCAST. Breakpoint Tables for Interpretation of MICs and Zone Diameters. Version 10.0.
Author notes
Marta Marí-Almirall, Clara Cosgaya and Cristina Pitart contributed to this work.
Other members of the MERCyCAT Study Group are listed in the Acknowledgements section.
