Characterization of an IncN2-type blaNDM-1-carrying plasmid in Escherichia coli ST131 and Klebsiella pneumoniae ST11 and ST15 isolates in Thailand

Sir, The prevalence of New Delhi metallo-β-lactamase-1 (NDM-1), encoded by the blaNDM-1 gene, has been increasing among various Gram-negative bacteria.1blaNDM-1 has been shown to reside in various plasmid incompatibility (Inc) types.1 Recently, a new Inc type, IncN2, has been identified in Escherichia coli and Klebsiella pneumoniae and proposed to acquire the blaNDM-1-carrying region by transposition.2,3 In Thailand, NDM-1 producers have become an emerging issue,4 and yet the characteristics of the blaNDM-1-carrying plasmid have not been elucidated. Here we report the genetic study of plasmids harbouring blaNDM-1 from domestic isolates. We examined three clinical isolates, one E. coli (ECS01) and two K. pneumoniae (KPS01 and KPS03), which were obtained from urine samples of independent Thai patients at two unrelated hospitals during June to August 2012 (IRB approval no. Si 454/2009). No previous hospitalization, antimicrobial exposure or healthcare-related history was documented for any patient. MIC values were determined using Etest (bioMérieux, France). According to the CLSI guideline,5 all isolates were resistant to all tested β-lactam agents, including cefoxitin, ceftazidime, imipenem, meropenem, doripenem and piperacillin/tazobactam, as well as ciprofloxacin with high MIC values (Table 1). Both K. pneumoniae isolates were resistant to gentamicin, but only KPS03 was resistant to amikacin. E. coli ECS01 remained susceptible to both gentamicin and amikacin. Multilocus sequence typing was performed to assign sequence type (ST) using primers and amplification conditions as recommended for E. coli (http://mlst.ucc.ie) and K. pneumoniae (http://www.pasteur.fr/mlst). E. coli ECS01 belonged to ST131 and K. pneumoniae KPS01 and KPS03 were ST11 and ST15, respectively. MICs of various antimicrobial agents for ECS01, KPS01 and KPS03 and their corresponding blaNDM-1 transformants MICs of various antimicrobial agents for ECS01, KPS01 and KPS03 and their corresponding blaNDM-1 transformants blaNDM-1 was identified on a plasmid, extracted by alkaline lysis from these isolates by PCR sequencing according to the protocol reported previously.5 The Inc type of blaNDM-1-carrying plasmids was determined by multiplex PCR-based replicon typing using primers and conditions as previously published.6 Forward and reverse primers to target the IncN2 variants were designed as N2-F: 5′-TAGCCTTCGGACAGGGTGAG-3′ and N2-R: 5′-ACGTTCGCCTGGATTTCATC-3′, respectively. All blaNDM-1-carrying plasmids were matched with the IncN2-type plasmid. The blaNDM-1-carrying plasmids were determined for their transferability by electroporation using E. coli TOP10 (Invitrogen, USA) as described elsewhere. The NDM transformants generated from ECS01, KPS01 and KPS03 parents were designated as ECS01-NDM, KPS01-NDM and KPS03-NDM, respectively. The blaNDM-1-carrying plasmids from each strain were designated as pNDM-ECS01, pNDM-KPS01 and pNDM-KPS03, respectively. Southern blotting hybridization using a specific blaNDM-1 probe illustrated that a blaNDM-1-carrying plasmid, approximately 40 kb in size, was presented in both parents and transformants (data not shown). Transformants showed decreased susceptibility to all tested β-lactam agents, including carbapenems, but remained susceptible to gentamicin and amikacin (Table 1). To further characterize the genetic structure of the blaNDM-1-carrying plasmid, pNDM-ECS01 was selected for whole DNA sequencing using the Nextera DNA library kit (Illumina, USA) according to the manufacturer's directions, and data were generated on MiSeq (Illumina). Annotation and sequence analysis were performed using CLC Genomics Workbench (version 6.5.1). The plasmid was 41 190 bp and consisted of 50.8% GC content. Fifty-six open reading frames were predicted encoding 40 coding sequences for known proteins, 3 truncated proteins and 13 hypothetical proteins (see Supplementary Data). Our plasmid was most closely related to the 41 187 bp identical plasmids pTR3 and pTR4 (GenBank numbers JQ349086 and JQ349085, respectively), which were recently reported as IncN2-type blaNDM-1-carrying plasmids in K. pneumoniae ST1 and ST273 from two patients in Singapore,3 except for three nucleotide insertions, C, A and T, at positions 1152, 9230 and 41119, respectively, relative to the start of repA gene. In addition, pNDM-ECS01 was also closely related to the 35 947 bp p271A, another IncN2-type blaNDM-1-carrying plasmid identified in an E. coli ST101 isolate from Australia.7 A 5243 bp region at the position 3492–8734 on pNDM-ECS01 was absent in p271A. This region contained the conserved upstream repeat-controlled regulon normally found in IncN plasmid and was related to bacterial conjugation efficiency. The remaining pNDM-ECS01 shared 99.98% homology to p271A. The complete DNA sequence of pNDM-ECS01 was deposited to GenBank database (KJ413946). NDM-1 producers have gained serious attention due to their high-level resistance to carbapenems. Genetic studies have shown that blaNDM-1 is commonly located on a plasmid that could be easily disseminated via horizontal transfer. We report here the characterization of IncN2-type blaNDM-1-carrying plasmids from clinical isolates in Thailand. Unlike the previous reports, this study first demonstrated the acquisition of IncN2-type plasmid-containing blaNDM-1 in highly virulent E. coli ST131 and outbreak-related drug-resistant K. pneumoniae ST11 and ST15 clones.8–10 The first IncN2-type blaNDM-1-carrying plasmid from a patient in Australia had a link to the Indian subcontinent since he was transferred from a hospital in Bangladesh.7 Our cases and cases in Singapore had no link to the Indian subcontinent and indeed had no history of traveling abroad. This suggests the potential for international multiclone spread of blaNDM-1 gene in this plasmid backbone. This should be an alert for continuous multinational surveillance of blaNDM-1-carrying isolates to appropriately control these highly resistant bacteria. This study was supported by the Thailand Research Fund through the Royal Golden Jubilee PhD Program (grant no. PHD/0232/2552 to T. N. and P. K.), and the Office of the Higher Education Commission and Mahidol University under the National Research University Inititative (to P. K.). P. K. was also supported by the ‘Chalermphrakiat’ Grant, Faculty of Medicine Siriraj Hospital, Mahidol University, during this study. None to declare. This work was partially presented as a poster at the ASID Gram-negative Superbugs Meeting held by the Australasian Society for Infectious Diseases, Gold Coast, Australia, 2013 (Poster 2). We greatly appreciate Dr Amornrut Leelaporn, Dr Preecha Montakantikul and Dr Pornpan Koomanachai for their valuable advice.