In April 4, 2015, a 65-year-old female was hospitalized in the First Affiliated Hospital of Nanchang University, China, for a cerebral infarction. In addition, the patient has the following diseases: coronary atherosclerotic heart disease, hypertension, type 2 diabetes, and hyperlipoidemia. At that time, the white blood cell (WBC) count was 8.2 × 10⁹/L with 65.1% granulocytes and 19.8% lymphocytes. However, on April 12, her WBC count increased to 15.5 × 10⁹/L with unusual 87.5% granulocytes and 6.3% lymphocytes, accompanied by a high temperature, breathing difficulty, and brick red foamy blood sputum, and followed by respiratory distress syndrome. Furthermore, the patient developed pneumonia and bacteriuria that was empirically treated with ceftazidime, ciprofloxacin, and metronidazole. The clinical status of the patient worsened necessitating intubation and mechanical ventilation on day 8, and her transfer on day 11 to the intensive care unit and died at 4 am, on April 16.

Klebsiella pneumoniae strain 24835 was isolated from midstream urine specimen. The isolate was identified using the VITEK 2 (BioMérieux, France). Further identification of K. pneumoniae was performed by partially sequencing the 16S rRNA gene (Lane, 1991). In addition, the hypermucoviscous (HV) phenotype was tested by evaluating the formation of a viscous string (positive test, >0.5 cm in length) that was stretched using an inoculation loop (Shon et al., 2013).

Antimicrobial susceptibilities for the bla_NDM–5 positive isolates and transconjugants were initially tested using the VITEK 2 system (BioMérieux, France) and then were followed by measuring minimum inhibitory concentration (MIC) using the microbroth dilution method (for ampicillin/sulbactam, piperacillin/tazobactam, cefazolin, cefotetan, ceftazidime, cefepime, imipenem, ertapenem, ciprofloxacin, levofloxacin, gentamicin, amikacin, and aztreonam) and E-test (AB bioMérieux, Sweden) (for tigecycline), respectively. The standard microbroth methods were performed according to the guideline M07-A9 of the Clinical and Laboratory Standards Institute [CLSI] (2015). E-tests were conducted according to packet insert instructions using Mueller–Hinton agar (MHA). Fresh bacterial colonies taken directly from MHA plates that were incubated at 37°C for 16–20 h were re-suspended in sterile Mueller–Hinton broth to obtain a suspension of 0.5 McFarland turbidity. E. coli ATCC 25922 was used as the quality control. The MIC results were interpreted according to the CLSI guidelines M100-S25 (Clinical and Laboratory Standards Institute [CLSI], 2015). The Food and Drug Administration (FDA) breakpoint was used for tigecycline.

Double disc synergy tests (DDSTs) using imipenem/ethylenediaminetetraacetic acid (EDTA) and aztreonam and amoxicillin/clavulanic acid were employed to screen for metallo-β-lactamases and extended spectrum beta-lactamases (ESBLs), respectively. Isolate was screened by polymerase chain reaction (PCR) amplification using a panel of primers for the detection of MBLs, Klebsiella pneumoniae carbapenemases (KPCs), OXA-48, ESBLs, including the SHV, TEM, CTX-M, and IBC/GES enzymes, and plasmid-mediated AmpCs (D’Andrea et al., 2006; Queenan and Bush, 2007; Andrade et al., 2010).

Total genomic DNA of K. pneumoniae 24835 was extracted using QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol. The genomic DNA was sent to Shanghai Biotechnology Corporation for the de novo whole-genome sequencing. A combination of Illumina MiSeq (300 bp paired-end, 12 million reads) and PacBio (10 kb fragment library, 356,001 reads) sequencing data were used to assemble the genome using SPAdes version 3.1.0 (Nurk et al., 2013). The Prokka program was employed for annotating the genomic sequence. Gene prediction was carried out using Glimmer 3.02, tRNA prediction with tRNAscan-SE, and rRNA prediction with HMMER, while basic local alignment search tool (BLAST) searches were performed against several databases including CatFam, COG, NCBI RefSeq, and SEED.

The acquired antimicrobial resistance genes were identified by uploading assembled genomes to the Resfinder server v2.1¹. The other genes relating to resistance and virulence were detected using the mapping unit of CLC Genomics Workbench to map reads and/or by blasting assembled genomes to a pseudomolecule generated by concatenating a set of K. pneumoniae genes. Scaffolds with resistance-related and virulence genes were blasted against GenBank to identify their genetic location. The sequence data from K. pneumoniae 24835 was compared to the results of antimicrobial resistance and virulence to identify genetic factors that correspond to the observed mechanisms.

Horizontal gene transfer of pNDM-5 was evaluated by transconjugation assays using 10⁶ colony forming units (cfu) of 24835 as a donor and 10⁶ cfu of E. coli J53 as recipients. After 24 h incubation, transconjugants were selected on MHA containing sodium azide (100 μg/ml) and imipenem (1 μg/ml). The presence of bla_NDM–5 in transconjugants was confirmed using PCR and enterobacterial repetitive intergenic consensus polymerase chain reaction (ERIC-PCR) was used for further distinguishing transconjugants from the donor strain. Plasmid DNA that was prepared from the transconjugant using alkaline lysis was subjected to PCR-based replicon typing (Carattoli et al., 2005; Johnson et al., 2012). The resultant plasmids were annotated using the Prokaryotic 6 Genomes Automatic Annotation Pipeline (PGAAP) available at NCBI². The plasmid was compared to publicly available plasmid references using BLAST at GenBank³. The plasmid comparison and visualization were generated by Easyfig (version 2.2.2) (Sullivan et al., 2011).

Pulsed field gel electrophoresis (S1-PFGE) was performed to determine the number and size of plasmids carried by strain 24835 as described previously. Briefly, agarose plugs containing whole-cell DNA of strain 24835 were treated with 8 U of S1 nuclease (Fermentas, Thermo Scientific, Waltham, MA, USA) and the reaction was stopped by adding 0.5 M EDTA (pH = 8). PFGE was conducted with a 1% SeaKem Gold agarose gel (Lonza, Basal, Switzerland) using a CHEF DRII system (Bio-Rad, Hercules, CA, USA) at 14°C, with a 6-V/cm current and run times of 12 h at switch time of 5–40 s followed by 8 h at switch time of 3–8 s. MidRange I PFG Marker (NEB, Ipswich, MA, USA) was used for size estimation.

Multilocus sequence typing (MLST) on the K. pneumoniae were performed as previously described (Liu et al., 2015).

Human blood was obtained from 10 healthy individuals. Pooled serum was separated and stored in small volumes at −80°C until use. Serum killing assay was performed as previously described. An inoculums of 25 μL (adjusted to 10⁶ cfu/ml) prepared from the mid-log phase was diluted by 0.9% saline, and was added to 75 μL of pooled human sera contained in a 10 × 75 mm Falcon polypropylene tube (BD Biosciences, Franklin Lakes, NJ, USA). Serial dilutions were plated on MHA for 0, 1, 2, and 3 h to obtain colony counts.

Each strain was tested at least three times, and the mean results were expressed as percent inoculums. The results were expressed as percentage of inoculation and the responses in terms of viable counts were graded from 1 to 6, as previously described (Abate et al., 2012). A strain was defined as serum sensitive at grades of 1–2, intermediately sensitive at grades of 3–4, and resistant at grades of 5–6. K. pneumoniae ATCC 700603 and 19432 (NDM-1 positive) strains are used for comparison in the serum killing assay.

To determine the 50% lethal dose (LD50), pathogen-free, 6- to 8-week-old, male BALB/c mice were obtained from Nanchang University Animal Center. Six mice were used as a sample population for each bacterial concentration. Ten-fold serial dilution of cfu of K. pneumoniae was made from a starting concentration of 10⁶ cfu/ml -10 cfu/ml, and BALB/c mice were infected intraperitoneally with 0.1 ml of each concentration (Zhang et al., 2015). Symptoms and mortality rates were observed for 14 days. The exact inoculation dose was confirmed on Luria-Bertani (LB) agar and the LD50 was calculated as described by Reed and Muench (1938). Survival curves were assessed by Kaplan–Meier analysis and log-rank test. P < 0.05 was considered statistically significant. K. pneumoniae ATCC 700603 and 19432 (NDM-1 positive) strains are used for comparison in the mouse lethality assay.

The whole-genome shotgun project of K. pneumoniae 24835 has been deposited at GenBank under the accession number CP014004. The annotated sequences of p24835-NDM5 and p24835-CTXM15 have been submitted to GenBank nucleotide sequence database under accession numbers CP014005 and CP014006, respectively.

A carbapenem-resistant K. pneumoniae isolate (termed 24835) at >10⁶ cfu/ml was recovered from urine on April 13, 2015. The patient was isolated in a single room and treated according to national and local infection control guidelines. This included an extensive terminal disinfection of the sickroom after discharge of the patient. The patient had no recent history of travel outside. 24835 were resistant to all antibiotics recommended by CLSI for the susceptibility testing of Enterobacteriaceae, with both isolates exhibiting identical MIC values (Figure 1 and Supplementary Table S2). Phenotypic ESBL and AmpC screening by DDST was positive, and PCR analyses and subsequent sequencing revealed the presence of bla_TEM–1B, bla_SHV–28, and bla_CTX–M–15. The DDST with imipenem/EDTA was positive, suggesting the production of a metallo-β-lactamase, which was verified by PCR detection of bla_NDM–5.

A total of 4,001,934 (1,208,584,068 bp) paired-end reads were generated using Illumina-HiSeq 2500. Using the PacBio RS II platform, 64,416 (941,657,914 bp) raw reads were produced. The complete genome of 24835 consists of a 5,387,996-bp circular chromosome and two plasmids of 101,030 and 46,161 bp in size. The genomic features of 24835 and the reference strains are summarized in Figure 1. Based on a Prokka analysis, 5191 putative open reading frames (ORFs) and 84 RNA genes on the circular chromosome (Figure 1; Supplementary Table S1), 108 putative ORFs on plasmid 1, and 49 putative ORFs on plasmid 2 were identified.

Nineteen antibiotic resistance genes were identified in the genome of 24385 strain. The β-lactam resistance genes in the 24385 genome were bla_TEM–1B and bla_SHV–28 in the chromosome, bla_CTX–M–15 in plasmid 1, and bla_NDM–5 in plasmid 2 (Supplementary Figure S1). The clinical features of K. pneumoniae infections depend on the virulence factors expressed by the infecting strain (Yu et al., 2007). A BLAST search was performed against virulence factors database (VFDB) to identify 120 virulence factors harbored by the 24385 strain. The 120 virulence genes of 24385 strain were classified into 31 the following categories: Iron uptake (32 genes), Immune evasion (12 genes), Endotoxin (11 genes), Adherence (11 genes), Fimbrial adherence determinants (8 genes), Toxin (7 genes), Antiphagocytosis (6 genes), Regulation (5 genes), Acid resistance (3 genes), Anaerobic respiration (2 genes), Cell surface components (2 genes), and Secretion system (2 genes). Among the 120 virulence genes identified, 8 genes were lipopolysaccharide (LPS)-related genes and 4 genes were capsular polysaccharide-related.

bla_NDM–5 was successfully transferred to E. coli J53, suggesting its localization on a plasmid. Strikingly, the conjugation frequency was high. Plasmid p24835-NDM5 was readily transferred from K. pneumoniae 24835 to E. coli J53Az^R with a frequency of (2.4 ± 0.9) × 10^-4 per donor cell at 37°C.

Transconjugants expressing NDM-5 were non-susceptible to all tested β-lactam antibiotics, including aztreonam, suggesting the presence of other β-lactamases besides NDM-5 on the plasmid (Supplementary Table S2). Transconjugants were positive for bla_CTX–M–15, but negative for bla_SHV–28, confirming the results of phenotypic testing. NDM-carrying transconjugants were furthermore sensitive to gentamicin and amikacin, indicating the absence of aminoglycoside resistance genes. Plasmid p24835-NDM5 carrying the bla_NDM–5 gene was 46,161 bp in length, with an average GC content of 52.04% (Figure 2A). It possessed 114 ORFs, including those corresponding to the replication, conjugation, and antibiotic resistance modules. Its scaffold corresponded to that of IncX3-type plasmids, commonly identified among enterobacterial isolates. In particular, it showed similarities with plasmid pNDM-MGR194, which was recently identified at the origin of the bla_NDM–1 gene acquisition in a K. pneumoniae isolate in India (Krishnaraju et al., 2015). Comparative analysis of the genetic contexts of bla_NDM–5 in this IncX3 plasmid and the previously reported plasmids revealed some differences (Figure 2B).

MLST showed that 24835 isolate was ST14. The strain was confirmed as serotype K2 and aerobactin-kfu-rmpA positive (Supplementary Figure S2). The identified virulence determinants may have contributed to the infection and or colonization of K. pneumoniae 24835 in the urinary tract. Furthermore, the isolate grew as positive HV phenotype with a mucoviscous string >5 mm in length from the colony on the blood agar plate (Supplementary Figure S2), however, the fact was observed after initial isolation of the bacterium, and, thereafter, the characteristic decreased.

Serum killing resistance (grade 6) was found in strain 24835. The other two isolates (ATCC 700603 and 19432) showed serum sensitive (grade 1 or 2) (Figure 3). Mouse lethality assay was used to study the virulence in vivo. The results showed that LD50 of >10⁶ cfu, which indicates low virulence, was identified in a NDM-1-producing strain 19432 and ATCC 700603. In contrast, strain 24835 was shown hypervirulent, as all mice injected with this strain died within 7 days at a low concentration (Figure 4). LD50 of strain 24835 was 105.2 cfu (<10³ cfu), exhibiting a high virulence level of “+++” according to previously described, which is in line with studies from Wei et al. (2016) who also reported an LD50, 120 cfu for highly virulent KPC-producing strain. This shows that the NDM-5-containing clinical isolate 24835 was more virulent than ATCC 700603 and the other, NDM-1 positive, K. pneumoniae strain 19432.