High Incidence and Endemic Spread of NDM-1-Positive Enterobacteriaceae in Henan Province, China

Abstract

The emergence and spread of New Delhi metallo-β-lactamase 1 (NDM-1)-producing carbapenem-resistant Enterobacteriaceae (CRE) present an urgent threat to human health. In China, the bla_NDM-1 gene has been reported mostly in Acinetobacter spp. but is rarely found in Enterobacteriaceae. Here, we report a high incidence and endemic spread of NDM-1-producing CRE in Henan Province in China. Sixteen (33.3%) of the 48 CRE isolates obtained from patients during June 2011 to July 2012 were positive for bla_NDM-1, and the gene was found to be carried on plasmids of various sizes (∼55 to ∼360 kb). These plasmids were readily transferrable to recipient Escherichia coli by conjugation, conferred resistance to multiple antibiotics, and belonged to multiple replicon types. The bla_NDM-1-positive CRE isolates were genetically diverse, and six new multilocus sequence typing (MLST) sequence types were linked to the carriage of NDM-1. Five of the isolates were classified as extensively drug-resistant (XDR) isolates, four of which also carried the fosA3 gene conferring resistance to fosfomycin, an alternative drug for treating infections by CRE. In each bla_NDM-1-positive CRE isolate, the bla_NDM-1 gene was downstream of an intact ISAba125 element and upstream of the ble_MBL gene. Furthermore, gene environment analysis suggested the possible transmission of bla_NDM-1-containing sequences from Acinetobacter spp. to Klebsiella pneumoniae and Klebsiella oxytoca. These findings reveal the emergence and active transmission of NDM-1-positive CRE in China and underscore the need for heightened measures to control their further spread.

INTRODUCTION

Carbapenems are the last-resort antibiotics against drug-resistant Gram-negative bacterial pathogens (1). However, carbapenem-resistant Enterobacteriaceae (CRE) are increasingly reported in health care-associated infections (HAIs) and are considered an urgent threat to human health by the Centers for Disease Control in the United States (2). Multiple genes conferring carbapenem resistance have been reported, which encode various types of carbapenemases (1). New Delhi metallo-β-lactamase 1 (NDM-1), an Ambler class B metallo-β-lactamase (MBL), hydrolyzes all β-lactams (including carbapenems) but monobactams and was first identified in Klebsiella pneumoniae and Escherichia coli isolated from a Swedish patient who was hospitalized in India in 2008 (3). Subsequently, the wide distribution of NDM-1-producing Enterobacteriaceae in India, Pakistan, and the United Kingdom was reported (4), and the Indian subcontinent was considered the main reservoir of NDM-1 producers (5). To date, the NDM-1-producing pathogens have been reported in more than 40 countries and have become a significant threat for public health worldwide (6).

In China, the bla_NDM-1 gene was first identified in four nonclonal Acinetobacter baumannii isolates (7). Subsequent studies found that Acinetobacter spp. were the main species carrying this gene, but the carriage of bla_NDM-1 was at a low frequency (<1.5%) (7–9). Carbapenem resistance in Enterobacteriaceae in China has been mainly associated with Klebsiella pneumoniae carbapenemases (KPCs), such as KPC-2, and MBLs (such as IMP-4) other than NDM-1 (10, 11). NDM-1-mediated carbapenem resistance in Enterobacteriaceae in China has been rarely reported; so far there have been only two confirmed cases of NDM-1-producing K. pneumoniae and E. coli infections in China (12, 13). A recent study detected a high infection rate (14.8%) of NDM-1-producing bacteria in clinical fecal samples from multiple hospitals in China, but none of the NDM-1-carrying bacteria belonged to Enterobacteriaceae (14). These observations suggested that NDM-1 was not prevalent in CRE isolates from China. However, in this study, we demonstrated the high incidence and endemic spread of NDM-1-producing Enterobacteriaceae in Henan Province in China. Additionally, we conducted molecular characterization of the bla_NDM-1-positive CRE isolates, revealing new molecular epidemiological features of CRE in China.

MATERIALS AND METHODS

Bacterial isolates, identification, and antimicrobial susceptibility testing.

A total of 48 CRE (imipenem [MIC, ≥4 μg/ml] or meropenem [MIC, ≥2 μg/ml]) isolates, including 6 E. coli, 33 K. pneumoniae, 1 Klebsiella oxytoca, 5 Enterobacter cloacae, and 3 Citrobacter freundii isolates, were collected from diagnostic laboratories in three hospitals located in the middle (Zhengzhou, n = 40), western (Sanmenxia, n = 4), and southern (Zhumadian, n = 4) regions of Henan Province (north-central China). These isolates were obtained from June 2011 to July 2012. All isolates were identified using the Vitek 2 system (bioMérieux, France) and 16S rRNA gene sequencing. Antimicrobial susceptibilities for the bla_NDM-1-positive isolates and transconjugants were initially tested using the Vitek 2 system (bioMérieux, France) and then were followed by measuring the MIC using the broth microdilution method (for ampicillin-sulbactam, piperacillin-tazobactam, cefazolin, cefotetan, ceftazidime, cefepime, imipenem, ertapenem, ciprofloxacin, levofloxacin, gentamicin, amikacin, and aztreonam), the Etest (AB bioMérieux, Sweden) (for chloramphenicol, tetracycline, tigecycline, and colistin), and the agar dilution method (for fosfomycin), respectively. The standard microbroth and agar dilution methods were performed according to the guideline M07-A9 of the Clinical and Laboratory Standards Institute (CLSI) (15). Etests 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 to 20 h were resuspended in sterile Mueller-Hinton broth to obtain a suspension with McFarland standard 0.5 turbidity. E. coli ATCC 25922 was used as the quality control. The MIC results were interpreted according to the CLSI guideline M100-S22 (15). The 2013 European Committee on Antimicrobial Susceptibility Testing breakpoints were used (available at http://www.eucast.org/clinical_breakpoints/) for colistin and tigecycline.

Detection of resistance determinants.

The modified Hodge test and the imipenem-EDTA double-disk synergy test were performed according to the CLSI guidelines to detect carbapenemase activity (15). PCR and nucleotide sequencing were employed to screen for the presence of carbapenemase-encoding genes, including bla_OXA-48-_like, bla_IMP, bla_VIM, bla_KPC, and bla_NDM. In addition, extended-spectrum β-lactamase (ESBL) genes, plasmid-mediated AmpC genes, 16S rRNA methyltransferase genes, and fosfomycin resistance determinants were also detected by use of the methods described previously (16–20).

Bacterial genotyping.

Pulsed-field gel electrophoresis (PFGE) of XbaI-digested genomic DNA of bla_NDM-1-positive CRE and reference marker Salmonella H9812 was performed using a contour-clamped homogeneous electric field (CHEF)-Mapper XA PFGE system (Bio-Rad, USA) with a 5- to 35-s linear ramp for 22 h at 6 V/cm at 14°C. The running buffer was 0.5× Tris-boric acid-EDTA (TBE) without thiourea. The gel image was captured using a Bio-Rad Universal Hood II gel imaging system (Bio-Rad, USA). The dendrograms were constructed from the PFGE data by the unweighted-pair group method with arithmetic mean (UPGMA) with the Dice coefficient using InfoQuest FP software version 4.5 (Bio-Rad Laboratories, USA). Multilocus sequence typing (MLST) for clinical K. pneumoniae and E. coli isolates was performed following the methods described previously (21, 22). The allelic profiles and sequence types (STs) were assigned using online databases (see http://www.pasteur.fr/recherche/genopole/PF8/mlst/Kpneumoniae.html for K. pneumoniae and http://mlst.warwick.ac.uk/mlst/dbs/Ecoli for E. coli).

Conjugation and plasmid analysis.

Conjugative assays were performed according to the method described previously (23). The bla_NDM-1-positive CRE served as the donors, while E. coli J53 (sodium azide resistant) was used as the recipient strain. Transconjugants were selected on Mueller-Hinton (MH) agar supplemented with sodium azide (100 μg/ml) and imipenem (1 μg/ml). The presence of the bla_NDM-1 gene and other resistance genes in transconjugants were identified using PCR and DNA sequencing, and antimicrobial susceptibility was also determined.

S1-PFGE and Southern blotting were conducted to estimate sizes of bla_NDM-1 plasmids (24). Briefly, whole-cell DNA of clinical and transconjugant strains in agarose gel plugs was treated with S1 nuclease (TaKaRa, Dalian, China), then separated by PFGE under the following conditions: 0.5× Tris-borate-EDTA, 1% agarose solution for 18 h at 6 V/cm and 14°C, with a pulse angle of 120° and the pulse time linearly ramped from 2.16 s to 63.8 s. Linear plasmids generated by S1-PFGE were transferred to nylon membranes (Millipore, USA), hybridized with digoxigenin-labeled bla_NDM-1-specific probes, and detected using a nitroblue tetrazolium-5-bromo-4-chloro-3-indolylphosphate (NBT-BCIP) color detection kit (Roche Applied Sciences, Germany) according to the recommendations of the supplier. Plasmid replicons were determined using the PCR-based replicon typing method (25).

The genetic environment surrounding bla_NDM-1 was investigated by PCR mapping and sequencing, and the Acinetobacter lwoffii plasmid of pNDM-BJ01 (accession no. JQ001791) and E. coli plasmid of pBJ01 (GenBank accession no. JX296013) were used as the references. PCR primers were designed from the reference sequences and are listed in Table S1 in the supplemental material. The locations of the primers are also shown (see Fig. 3).

FIG 3.

Comparison of the bla_NDM-1 gene environments identified in this study with others published previously. (A) bla_NDM-1-surrounding sequences in A. baumannii 161/07 (GenBank accession no. HQ857107), A. lwoffii pNDM-Bj01 (JQ001791), and non -baumannii Acinetobacter spp. plasmids (JQ080305). (B and D) the bla_NDM-1 gene environments identified in this study; the names of the isolates are listed on the right of each panel.(C, E and F) bla_NDM-1-surrounding sequences in E. coli BJ01 (JX296013), E. coli pNDM-HK (HQ451074), and E. coli pNDM-1_Dok01 (AP012208), respectively. The boxed arrows indicate the positions and directions of transcription of the genes. The gray-shaded areas represent regions sharing >99% DNA identity. Positions of the primers used for PCR mapping are indicated by arrows.

Nucleotide sequence accession number.

The sequence described in this paper has been submitted to GenBank under the accession no. KF985036 (K. oxytoca isolate ZMD8).

RESULTS AND DISCUSSION

Identification of bla_NDM-1-positive isolates from hospital patients.

Among the 48 CRE, 16 (16/48 [33.3%]) were identified as bla_NDM-1 positive, including 6 E. coli, 4 K. pneumoniae, 1 K. oxytoca, 3 E. cloacae, and 2 C. freundii isolates, which were obtained from specimens of blood (n = 7), urine (n = 6), and sputum (n = 2) and a burn site (n = 1) (Table 1). Additionally, 26 K. pneumoniae isolates harbored bla_KPC-2, 2 isolates (1 K. pneumoniae and 1 E. cloacae) carried bla_IMP-4, and the remaining isolates did not contain the carbapenemase genes (bla_NDM, bla_KPC, bla_VIM, bla_IMP, and bla_OXA-48-like) screened in this study. The 16 bla_NDM-1-positive CRE were from three hospitals in three different cities in Henan Province. The clinical data associated with the 16 isolates were summarized in Table 1. These patients were diagnosed with different clinical diseases, but none of the patients had a history of foreign travel. Notably, 5 patients, including 3 young children, died of infections. All of the patients who died, except an infant, had a positive blood culture for CRE and two of the patients had septicemia (Table 1). Thus, the death rate among the patients infected with a bla_NDM-1-positive isolate was 31%, which is higher than previously reported rates in confirmed cases associated with NDM-1-producing bacteria worldwide (26). These results revealed the high prevalence of bla_NDM-1-positive CRE in the hospitals of Henan Province and their association with a high mortality rate.

TABLE 1.

Characteristics of bla_NDM-1-positive CRE isolates

Isolatea	Clinical features				Additional resistance determinantsb			MLSTc	Plasmid type carrying blaNDM-1/plasmid size (kb)
	Patient age/sex	Specimen	Diagnosis/wardd	Outcome	β-Lactamases	16S rRNA methylase	Others
EC-01	76 yr/female	Urine	Pyelonephritis complicated with diabetes/endocrinology	Discharge	−	−	−	ST1237	Untypeable/230
EC-03	58 yr/male	Blood	Severe acute pancreatitis/ICU	Death	TEM-1, CTX-M-55, CMY-30	RmtB	FosA3	ST361	A/C/180
EC-13	6 days/male	Blood	Neonatal sepsis/NICU	Discharge	TEM-1, CMY-30	−	−	ST40	FIB/310
EC-24	56 yr/female	Urine	Diabetes and urinary tract infections/endocrinology	Discharge	CTX-M-15	−	FosA3	ST205	A/C/230
EC-SMX3	42 yr/male	Blood	Multiple injuries/EICU	Discharge	TEM-1, CTX-M-15, CMY-30	−	−	ST410	I1/60
EC-SMX5	63 yr/female	Sputum	Lung cancer/oncology	Discharge	CTX-M-15, CMY-30	−	−	ST361	A/C/260
KP-07	72 yr/female	Blood	Intracranial hemorrhage associated with cerebral infections/neurosurgery	Discharge	−	−	−	ST11	Untypeable/55
KP-09	1 yr/male	Urine	Multiple contusions as a result of a car accident/PICU	Discharge	TEM-1	RmtB	−	ST889	A/C/245
KP-40A	10 days/male	Blood	Neonatal sepsis/NICU	Death	TEM-1, CTX-M-15	−	−	ST966	A/C/−e
KP-41	9 mo/female	Blood	Septicemia/PICU	Death	−	−	−	ST113	N/55
KO-ZMD8	75 yr/male	Urine	Nephrotic syndrome/nephrology	Discharge	−	−	−	ND	Untypeable/55
ECL- ZMD10	49 yr/male	Wound	Extensive burns/burn unit	Discharge	−	ArmA	FosA3	ND	Untypeable/360
ECL- ZMD12	21 yr/female	Blood	Severe aplastic anemia/hematology	Death	−	ArmA	−	ND	A/C/55
ECL-36	15 days/male	Sputum	Lung infections and asphyxia/NICU	Death	MIR-2	−	−	ND	A/C/160
CF-SMX4	67 yr/male	Urine	Nephropathy/nephrology	Discharge	CMY-73	−	−	ND	Untypeable/55
CF-25	62 yr/male	Urine	Cerebral hemorrhage and lung infection/neurosurgery	Discharge	CMY-73	ArmA	FosA3	ND	A/C/170

^aEC, E. coli strains; KP, K. pneumoniae strains; KO, K. oxytoca strains; ECL, E. cloacae strains; CF, C. freundii strains.

^bResistance markers that are cotransferred with bla_NDM-1 by conjugation are underlined. Minus signs indicate negative results. PCR screening included bla_TEM; bla_SHV; bla_OXA-1-like; bla_CTX-M groups 1, 2, 8, 9, and 26; bla_OXA-48-like; bla_IMP; bla_VIM; bla_KPC; bla_NDM; 6 groups of bla_AmpC β-lactamase genes; and fosfomycin resistance genes fosA, fosB, fosC, and fosX as well as the armA, rmtA-rmtE, and npmA 16S methylase genes.

^cSTs that are newly identified harboring bla_NDM-1 in this study are underlined. MLST, multilocus sequence typing; ND, not determined.

^dICU, intensive care unit; NICU, newborn ICU; EICU, emergency ICU; PICU, pediatric ICU.

^eS1-PFGE and Southern blotting failed to detect the bla_NDM-1 plasmid in KP-40A and its transconjugant, although the bla_NDM-1 gene was detected in them by PCR and sequencing.

Antimicrobial susceptibility patterns and resistance determinants.

All of the NDM-1-positive CRE isolates were resistant to carbapenems and cephalosporins but susceptible to colistin (MICs of ≤2 mg/liter) (Table 2). Molecular characterization showed that most of the E. coli (5/6) and half of the K. pneumoniae isolates (2/4) harbored ESBL genes, AmpC genes, or both (Table 1). Other carbapenemase-encoding genes, including bla_KPC, bla_VIM, bla_IMP, and bla_OXA-48-like, were not detected in any of the NDM-1-positive CRE isolates. Six isolates (EC-24, KP-07, KP-09, KP-40A, KP-41, and ZMD12) demonstrated tigecycline resistance according to the EUCAST clinical breakpoint, with MICs of ≥2 μg/ml (Table 2). More significantly, five isolates, including EC-03, CF-25, ECL-ZMD10 (susceptible only to tigecycline and colistin), EC-24 (susceptible only to amikacin and colistin), and ECL-ZMD12 (susceptible only to fosfomycin and colistin), were identified as extensively drug-resistant (XDR) bacteria (nonsusceptible to ≥1 agent in all of the 17 but ≤2 antimicrobial categories, including glycylcyclines and polymyxins used for treatment of infections caused by Enterobacteriaceae) according to the definitions described by Magiorakos et al. (27). In addition, four isolates (EC-03, EC-24, ECL-ZMD10, and CF-25) harbored a plasmid-mediated fosfomycin resistance gene, fosA3. This gene was found only in CTX-M-producing E. coli isolated from Asian countries (28), and our finding represents the first report of NDM-1-producing CRE harboring the fosA3 gene encoding fosfomycin resistance. Generally, the resistance rate of fosfomycin in E. coli was low (1% to 3%) worldwide, and fosfomycin is also used as an alternative choice for treating infections caused by ESBL-producing and even carbapenemase-producing Enterobacteriaceae (29). However, the occurrence of fosfomycin resistance in NDM-1-producing CRE observed in this study will further limit clinical therapeutic options.

TABLE 2.

Antibiotic susceptibilities of bla_NDM-1-positive CRE and their transconjugants

Isolatea	Antibioticb susceptibility (μg/ml) to:
	SAM	TZP	CFZ	CTT	CAZ	FEP	IPM	ETP	CIP	LVX	GEN	AMK	SXT	ATM	CHL	TET	FOF	TGC	CST
EC-01	>256	>256	>256	>256	>256	>256	8	16	2	2	64	16	40	8	256	256	8	0.75	0.5
EC-03	>256	>256	>256	>256	>256	256	>32	32	>32	>32	>256	>256	>320	256	256	256	512	0.5	0.5
EC-13	>256	>256	>256	>256	>256	>256	4	16	8	2	32	16	>320	64	256	256	8	0.38	1
EC-24	>256	>256	>256	>256	>256	>256	>32	32	>32	>32	32	8	>320	128	256	256	512	3	1
EC-SMX3	>256	>256	>256	>256	>256	>256	4	8	>32	>32	64	128	>320	256	4	2	32	0.5	2
EC-SMX5	>256	>256	>256	>256	>256	>256	32	16	>32	>32	32	<2	>320	128	256	256	8	0.5	2
KP-07	>256	>256	>256	>256	>256	>256	8	>32	>32	>32	<1	<2	>320	256	128	2	8	3	1
KP-09	>256	>256	>256	>256	>256	>256	>32	4	1	1	>256	>256	>320	32	256	256	32	3	1
KP-40A	>256	>256	>256	>256	>256	>256	>32	16	<0.25	<0.25	64	<2	>320	>256	256	256	32	6	1
KP-41	>256	>256	>256	>256	>256	>256	>32	>32	<0.25	<0.25	<1	<2	>320	<1	6	2	8	3	0.5
KO-ZMD8	>256	>256	>256	>256	>256	>256	8	16	<0.25	0.5	32	<2	>320	128	256	256	32	0.5	0.5
ECL-ZMD10	ND	64	ND	ND	>256	>256	8	32	>32	>32	>256	>256	>320	256	256	256	128	1	1
ECL-ZMD12	ND	>256	ND	ND	>256	>256	8	32	>32	>32	>256	>256	>320	>256	256	256	32	3	2
ECL-36	ND	>256	ND	ND	>256	>256	32	>32	<0.25	<0.25	32	<2	>320	256	8	4	8	2	1
CF-SMX4	ND	>256	ND	ND	>256	>256	>32	>32	8	4	16	<2	<20	256	6	128	8	1	1
CF-25	ND	>256	ND	ND	>256	>256	>32	>32	>32	>32	>256	>256	>320	64	48	256	512	0.75	0.5
E. coli transconjugant strains
EC-01-J53	>256	64	>256	>256	>256	32	8	16	<0.25	<0.25	32	<2	<20	4	256	256	8	1	0.5
EC-03-J53	>256	32	>256	>256	>256	8	>32	8	<0.25	<0.25	>256	>256	>320	128	256	2	512	0.19	1
EC-13-J53	>256	64	>256	32	>256	4	4	4	<0.25	<0.25	16	<2	>320	32	128	128	8	0.25	1
EC-24-J53	>256	>256	>256	>256	>256	>256	32	16	<0.25	<0.25	8	<2	<20	2	256	256	8	0.19	1
EC-SMX3-J53	>256	>256	>256	>256	>256	>256	4	8	8	16	32	32	>320	64	4	0.75	8	0.19	1
EC-SMX5-J53	>256	64	>256	>256	>256	16	32	16	<0.25	<0.25	32	<2	<20	<1	256	8	16	1	0.5
KP-07-J53	>256	>256	>256	>256	>256	>256	8	16	8	16	<1	<2	>320	128	24	2	8	0.5	1
KP-09-J53	>256	64	>256	4	>256	2	4	4	<0.25	<0.25	<1	<2	<20	<1	8	256	32	2	1
KP-40A-J53	>256	>256	>256	32	>256	8	8	4	<0.25	<0.25	32	<2	>320	128	256	256	8	1	1
KP-41-J53	>256	>256	>256	>256	>256	8	8	16	<0.25	<0.25	<1	<2	40	<1	4	2	8	0.75	1
KO-ZMD8-J53	>256	>256	>256	>256	>256	>256	8	16	<0.25	<0.25	8	<2	>320	128	256	256	16	0.75	1
ECL-ZMD10-J53	>256	64	>256	>256	>256	>256	4	16	32	16	>256	>256	>320	128	256	256	128	1.5	0.5
ECL-ZMD12-J53	>256	>256	>256	>256	>256	>256	8	32	32	32	>256	16	<20	>256	256	256	32	3	1
ECL-36-J53	>256	64	>256	>256	>256	>256	32	>32	<0.25	<0.25	32	<2	>320	<1	8	8	8	1.5	1
CF-SMX4-J53	>256	64	>256	32	>256	>256	8	16	<0.25	<0.25	2	<2	<20	2	4	128	8	1	1
CF-25-J53	>256	64	>256	16	>256	2	8	4	<0.25	<0.25	32	<2	<20	4	48	128	128	0.25	1
EC J53	<2	<4	<4	<4	<1	<1	<1	<0.5	<0.25	<0.25	<1	<2	<20	<1	8	2	2	0.25	0.5

^aAll of the bla_NDM-1-positive isolates were multidrug-resistant (MDR) strains, and the XDR isolates are highlighted in bold type. EC, E. coli strains; KP, K. pneumoniae strains; KO, K. oxytoca strains; ECL, E. cloacae strains; CF, C. freundii strains. For the transconjugants, all were E. coli J53 harboring plasmids from the respective clinical isolates.

^bSAM, ampicillin-sulbactam (1/0.5–256/128) (the numbers in parentheses indicate the test range [μg/ml] for each agent); TZP, piperacillin-tazobactam (0.5/4–256/4); CFZ, cefazolin (0.5–256); CTT, cefotetan (0.03–256); CAZ, ceftazidime (0.03–256); FEP, cefepime (0.015–256); IPM, imipenem (0.06–32); ETP, ertapenem (0.004–32); CIP, ciprofloxacin (0.004–32); LVX, levofloxacin (0.008–32); GEN, gentamicin (0.25–256); AMK, amikacin (0.5–256); ATM, aztreonam (0.06–256); CHL, chloramphenicol (0.016–256); TET, tetracycline (0.016–256); FOF, fosfomycin (0.25–512); TGC, tigecycline (0.016–256); CST, colistin (0.016–256). The MICs of trimethoprim-sulfamethoxazole (SXT) were obtained by the Vitek 2 system. ND, not determined (E. cloacae and C. freundii are intrinsically resistant to SAM, CFZ, and CTT).

Genotyping.

PFGE was successfully performed with 4 E. coli, 3 K. pneumoniae, 1 K. oxytoca, 3 E. cloacae, and 2 C. freundii isolates, and their patterns were completely different (Fig. 1). Three isolates, including EC-24, EC-SMX5, and KP-40A, were nontypeable by PFGE, as their XbaI-digested genomic DNA failed to yield distinct bands despite multiple efforts to repeat the experiment. MLST was performed for all of the bla_NDM-1-positive E. coli and K. pneumoniae isolates, since the two species of Enterobacteriaceae were the major ones carrying the bla_NDM-1 gene. Five MLST types were identified among the six E. coli isolates. Similarly, four MLST types were identified among the four K. pneumoniae isolates (Table 1). Two E. coli isolates (EC-03 and EC-SMX5) obtained from two different hospitals located in geographically separated areas (Zhengzhou and Sanmenxia) shared the same ST type (ST361), suggesting they were clonally related. Overall, our data showed that clonally diverse NDM-1-producing Enterobacteriaceae contributed to the dissemination of bla_NDM-1in Henan Province.

FIG 1.

PFGE-based dendrograms showing the genetic relationships of 13 bla_NDM-1-positive CRE isolates. EC-24, EC-SMX5, and KP-40A failed to yield bands by XbaI-PFGE and were not included.

This study linked for the first time six new STs (ST40, ST205, and ST1237 [E. coli] and ST113, ST889, and ST966 [K. pneumoniae]) to the production of NDM-1. E. coli ST410 and K. pneumoniae ST11 were frequently detected in clinical isolates of NDM-1-producing E. coli and K. pneumoniae from different countries, suggesting that they play an important role in the dissemination of bla_NDM-1 (30). In addition, the NDM-1-producing E. coli ST361 isolate was also found in New Zealand (31).

Plasmid analysis.

S1-PFGE and Southern blot analysis showed that the presence of the bla_NDM-1 gene in the 16 CRE isolates was located on diverse plasmids, with sizes ranging from ∼55 to ∼360 kb. Furthermore, half of the bla_NDM-1 plasmids belonged to plasmid replicon type IncA/C (Table 1 and Fig. 2). This broad-host-range plasmid was most frequently reported for carrying bla_NDM-1 and other resistance genes, and it is widely disseminated among Gram-negative bacteria worldwide (6, 32).

FIG 2.

Detection of bla_NDM-1- (I) and fosA3- (II) carrying plasmids by PFGE and Southern hybridization. (I) S1-PFGE (top) and Southern blotting (bottom) with bla_NDM-1 specific probe. Lanes M, marker (Salmonella H9812); lane 1, EC-01-J53; lane 2, EC-SMX5-J53; lane 3, EC-SMX3; lane 4, KP-41; lane 5, KP-40A (untypeable); lane 6, KP-09; lane 7, KP-07; lane 8, KO-ZMD8; lane 9, ECL-ZMD10; lane 10, ECL-ZMD12; lane 11, ECL-36; lane 12, CF-SMX4; lane 13,CF-25; lane 14, EC-03-J53; lane 15, EC-24-J53; lane 16, EC-13. The arrows indicate the locations of the plasmids hybridized to the NDM-1 probe. (II) S1-PFGE (left) and Southern blotting (right) with fosA3-specific probe. Lane M, marker (Salmonella H9812); lanes 1, ECL-ZMD10-J53; lanes 2, CF-25-J53; lanes 3, EC-03-J53.

Conjugative assays revealed that all of the bla_NDM-1 plasmids were successfully transferred to E. coli J53 from the 16 donors by conjugation. The 16 transconjugants all showed resistance to carbapenems and cephalosporins (Table 2). Moreover, the bla_NDM-1 plasmids in all the transconjugants remained stable after 10 passages in the absence of imipenem selection. In addition, resistance genes such as bla_TEM-1, bla_CTX-M-15/55, bla_CMY-30, rmtB, armA, and fosA3 were cotransferred to E. coli J53 with the bla_NDM-1 gene in several isolates (Table 2). Interestingly, we found that the bla_NDM-1 and fosA3 genes were carried by distinct IncA/C plasmids in EC-03 (∼180 kb for bla_NDM-1 and ∼90 kb for fosA3) and CF-25(∼170 kb for bla_NDM-1 and ∼310 kb for fosA3) (Fig. 2), but the two plasmids in each isolate were cotransferred to the E. coli J53 recipient, suggesting that both were mobile or one of them acted as a helper plasmid for the other to transfer. In contrast to the above two isolates, ECL-ZMD10 harbored a large plasmid (∼360 kb) of an untypeable replicon type, which carried both bla_NDM-1 and fosA3 and was transferred to E. coli J53 by conjugation (Fig. 2 and Table 2). Several recent studies demonstrated that the fosA3 gene was associated with bla_CTX-M genes and carried by plasmids belonging to different replicon types, including F, N, B/O, and I1 in E. coli and K. pneumonia (33–35). In our study, this gene was found to be carried by a conjugative bla_NDM-1 plasmid in a species (E. cloacae) other than E. coli and K. pneumonia, implying the further spread of fosA3 among Enterobacteriaceae. Moreover, the E. coli J53 recipient became an XDR strain after accepting the ECL-ZMD10 plasmid by conjugation, indicating multiple resistance determinants might be carried by this bla_NDM-1 plasmid. Additional studies are ongoing to characterize this plasmid. In vivo interspecies dissemination of IncA/C plasmids carrying bla_NDM-1was described in previous studies (23, 36). However, all of the NDM-1 plasmids identified in the 16 bla_NDM-1-positive CRE isolates were different either in size or replicon type, suggesting that insertion elements or transposons may play important roles in the mobilization of bla_NDM-1.

Genetic environments of bla_NDM-1.

The genetic environments surrounding the bla_NDM-1 gene were detected by PCR mapping and DNA sequencing. The bleomycin resistance gene ble_MBL, followed by a truncated trpF gene, was located immediately downstream of the bla_NDM-1 gene in all of the 16 CRE isolates (Fig. 3). The three-gene cluster (bla_NDM-1-ble_MBL-ΔtrpF) was highly conserved and was also previously reported in E. coli plasmids (pNDM-HK, pBJ01, and pNDM_Dok01) and A. lwoffii plasmids (pNDM-BJ01 and pNDM-BJ02) (37) (Fig. 3). It was reported that bla_NDM-1 and ble_MBL are under the control of the same promoter in front of bla_NDM-1, and this structure may facilitate the spread of NDM-1 when under the selective pressure of bleomycin-like molecules in the environment (38). The entire ISAba125 element was detected immediately upstream of the bla_NDM-1 gene in each of the 16 isolates. It has been documented that ISAba125 provides the −35 region of the promoter sequence for bla_NDM-1 in all reported cases (37). For two K. pneumonia isolates (KP-09 and KP-41) and one K. oxytoca isolate (KO-ZMD8), a 5,859-bp DNA segment was sequenced and it contained eight genes, including ISAba125, bla_NDM-1, ble_MBL, ΔtrpF, dsbC, cutA1, groES, and groEL (Fig. 3). This DNA segment is highly homologous (>99%) to previously identified sequences in Acinetobacter spp. from China (9, 39) (Fig. 3), suggesting the possible transmission of bla_NDM-1-containing sequences between Acinetobacter spp. and Enterobacteriaceae. Considering that bla_NDM-1 was mostly reported in Acinetobacter spp. in China, our finding suggests that Acinetobacter may have served as a reservoir for the dissemination of NDM-1 toward Enterobacteriaceae.

The sporadic emergence of NDM-1-producing E. coli and K. pneumoniae in mainland China has been reported in two recent studies (12, 13). Additionally, several cases of NDM-1-producing CRE reported from Hong Kong and Taiwan were also epidemiologically linked to mainland China (40–42). Here, we report our study of a cluster of bla_NDM-1-positive CRE from patients in Henan Province. Together, these findings strongly suggest that NDM-1-producing CRE is spreading in mainland China. Compared to previously reported studies, our work revealed several important findings on CRE in China. First, a high incidence (33.3%) of NDM-1-producing Enterobacteriaceae was observed among the examined CRE, which is much higher than previously realized and suggests that bla_NDM-1-positive CRE is endemic in Henan Province. Second, none of the patients carrying NDM-1-producing CRE had a history of foreign travel, suggesting that the NDM-1-carrying CRE were endogenous to the region and likely originated locally due to antibiotic selection and spread of the resistance gene. Third, most of the NDM-1-producing E. coli and K. pneumoniae isolates identified in this study belong to new MLST sequence types that have not been reported previously, suggesting the active spread of bla_NDM-1 among Enterobacteriaceae in China. These findings provide new insights into the epidemiology of CRE in China.

In summary, our study demonstrated a high incidence of NDM-1-producing CRE from patients with different clinical diseases in Henan Province. These NDM-1-positive isolates were genetically diverse and carried plasmids of various sizes that were readily transferred by conjugation. Our results suggest that mainland China is becoming an endemic reservoir for NDM-1-producing Enterobacteriaceae. In addition, the emergence of XDR Enterobacteriaceae carrying conjugative bla_NDM-1 plasmids coharboring other resistance determinants such as fosA3 is alarming, as spread of these isolates will seriously limit options for clinical treatment in future. Thus, enhanced efforts are urgently needed to control the further spread of NDM-1-producing Enterobacteriaceae.