Molecular characterization of multidrug-resistant Klebsiella pneumoniae isolates.

Klebsiella pneumoniae is an important cause of healthcare-associated infections worldwide. Selective pressure, the extensive use of antibiotics, and the conjugational transmission of antibiotic resistance genes across bacterial species and genera facilitate the emergence of multidrug-resistant (MDR) K. pneumoniae. Here, we examined the occurrence, phenotypes and genetic features of MDR K. pneumoniae isolated from patients in intensive care units (ICUs) at the First Affiliated Hospital of Xiamen University in Xiamen, China, from January to December 2011. Thirty-eight MDR K. pneumoniae strains were collected. These MDR K. pneumoniae isolates possessed at least seven antibiotic resistance determinants, which contribute to the high-level resistance of these bacteria to aminoglycosides, macrolides, quinolones and β-lactams. Among these isolates, 24 strains were extended-spectrum β-lactamase (ESBL) producers, 2 strains were AmpC producers, and 12 strains were both ESBL and AmpC producers. The 38 MDR isolates also contained class I (28/38) and class II integrons (10/38). All 28 class I-positive isolates contained aacC1, aacC4, orfX, orfX' and aadA1 genes. β-lactam resistance was conferred through bla SHV (22/38), bla TEM (10/38), and bla CTX-M (7/38). The highly conserved bla KPC-2 (37/38) and bla OXA-23(1/38) alleles were responsible for carbapenem resistance, and a gyrAsite mutation (27/38) and the plasmid-mediated qnrB gene (13/38) were responsible for quinolone resistance. Repetitive-sequence-based PCR (REP-PCR) fingerprinting of these MDR strains revealed the presence of five groups and sixteen patterns. The MDR strains from unrelated groups showed different drug resistance patterns; however, some homologous strains also showed different drug resistance profiles. Therefore, REP-PCR-based analyses can provide information to evaluate the epidemic status of nosocomial infection caused by MDR K. pneumoniae; however, this test lacks the power to discriminate some isolates. Thus, we propose that both genotyping and REP-PCR typing should be used to distinguish genetic groups beyond the species level.

Introduction

Klebsiella pneumoniae has been medically recognized as one of the most important opportunistic pathogens, causing hospital-acquired and healthcare-associated pulmonary system, urinary tract, circulating system and soft tissue infections worldwide (Podschun and Ullmann, 1998). However, K. pneumoniae has become a clinically important micro-organism, particularly in last two decades due to its tendency to develop antibiotic resistance and cause fatal outcomes (Nordmann ; Ko ).

In recent years, K. pneumoniae has been identified as a major cause of community-acquired pneumonia (CAP) and is responsible for approximately 10% of all hospital-acquired infections, ranking second among Gram-negative pathogens (Nordmann ). Studies performed in Asia have demonstrated that the fatality rate of K. pneumoniae-induced pneumonia in elderly people was 15% to 40%, equal to or even greater than that of Haemophilus influenzae (Molton ). The production of extended-spectrum β-lactamases (ESBL) in this organism contributes to the emergence and dissemination of K. pneumoniae infections. Effective anti-infective drugs, such as aminoglycosides, fluoroquinolones, and carbapenems, have been used to treat ESBL-producing K. pneumoniae infections (Sanchez ). In particular, the use of carbapenems against ESBL-producing Gram-negative micro-organisms poses a serious problem in the management of healthcare-associated infections because the abuse of antibiotics might lead to the emergence of carbapenem-resistant organisms. Notably, the selective pressure posed by the extensive use of antibiotics has facilitated the emergence of multidrug-resistant (MDR) K. pneumoniae. Furthermore, conjugational transmission of antibiotic resistance genes across bacterial species and genera has aggravated the problem of K. pneumoniae antibiotic resistance.

MDR K. pneumoniae was first reported in the United States, followed by Europe, South America, and Asia (Ko ; Pfaller ; Winokur ; Yigit ). At present, infections caused by MDR K. pneumoniae have become a major problem, as few antibiotics are available, resulting in higher morbidity, longer hospitalization, increased mortality rates, and excessive health care costs compared with infections associated with antibiotic-susceptible micro-organisms (Correa ; Ma and Wang, 2013). However, the prevalence of antibiotic-resistant bacteria significantly varies according to region, country, and susceptible population, as the seriousness of this problem is significantly associated with the measures applied to control the spread of drug-resistant bacteria (Ko ). The aim of this study was to characterize the drug resistance phenotypes and molecular antibiotic resistance mechanisms of MDR K. pneumoniae strains isolated from patients in intensive care units (ICUs). The relatedness of the strains was also investigated using the repetitive-sequence-based PCR (REP-PCR)-based DiversiLab system.

Materials and Methods

Bacterial identification and antimicrobial susceptibility testing

A total of 38 MDR K. pneumoniae strains were isolated from sputum, bronchial-alveolar perfusate, blood, intraperitoneal drainage, pleural drainage of patients and surfaces of objects (respirators, bedrails and bedclothes) in ICUs from January to December 2011. Thirty nosocomial infection strains were identified according to the Diagnostic Criteria for Nosocomial Infection established by the Ministry of Health of China. Five strains (K33, K34, K35, K36 and K37) were isolated from ventilators, two strains (K19 and K20) were isolated from bedrails, and one (K31) strain was isolated from bedclothes. Bacterial identification and antimicrobial susceptibility testing were performed using the Vitek2 semi-automated system with GN13 and AST13 cards (bioMérieux Inc., Durham, NC), respectively. The minimum inhibitory concentrations (MICs) for 28 antimicrobial agents for MDR isolates were confirmed using the broth microdilution method. In addition, polymyxin B and tigecycline susceptibilities were determined through Etests (bioMérieux Inc., Durham, NC). Multidrug resistance was defined as non-susceptibility to at least one agent in three or more antimicrobial categories according to the breakpoints recommended by the Clinical and Laboratory Standards Institute (CLSI), excluding tigecycline and polymyxin B. For these studies, K. pneumoniae ATCC 700603 and Pseudomonas aeruginosa ATCC 27853 were used as controls. Disks of cefotaxime (30 μg), cefotaxime/clavulanic acid (30/10 μg), ceftazidime (30 μg), and ceftazidime/clavulanic acid (30/10 μg) were used to confirm ESBL production of these K. pneumoniae isolates using the Double Disc Synergy Test (DDST) according to the recommendations of the CLSI. For the phenotypic identification of AmpC-producing isolates, the DDST was applied using disks containing ceftazidime (30 μg) and ceftazidime/3-aminophenylboronic acid (30/300 μg). A ≥ 5-mm enhancement of the zone of inhibition around the ceftazidime disk was considered AmpC-positive. The three-dimensional test was also used to confirm AmpC-producing isolates.

DNA extraction

Genomic DNA from MDR strains was prepared for PCR and genetic analysis using the UltraClean® Microbial DNA Isolation Kit (MO BIO Labs Inc., Solana Beach, CA) according to the manufacturer’s instructions. Briefly, MDR strains were grown to the desired optical density and were collected after centrifugation at 10,000 × g for 60 s. The bacterial cells were resuspended in bead solution and transferred to a bead-beating tube containing microbeads and proteinase K. Subsequently, the cells were lysed through mechanical disruption using a specially designed MO BIO Vortex Adapter on a standard vortex. The supernatant was loaded onto a Spin Filter mini column and washed twice with Wash buffer, and the genomic DNA was eluted with DNA-free Tris buffer.

Resistance determinants detection

The genes involved in the antibiotic resistance of MDR strains were screened through PCR assays as previously described (Perez-Perez and Hanson, 2002; Wen, 2010; Mak, 2009), except that the PCR conditions used to detect bla KPC and bla NDM-1, included a pre-denaturing step at 94 °C for 5 min, followed by 30 cycles at 94 °C for 30 s, 55 °C for 30 s, and 72 °C for 60 s, with polishing at 72 °C for 10 min. The PCR products were purified using the QIAGEN Gel Extraction Kit (Qiagen Inc., Hilden, Germany) and were subjected to sequencing using the same primers used for PCR amplification. The resistance genes and the primers used in the present study are listed in Table 1.

Table 1

Primer sequences used in this study

Primer	Sequence	Target gene	Source or reference
KPC-F	ATGTCACTGTATCGCCGTC	bla KPC	This study
KPC-R	TTACTGCCCGTTGACGCC	bla KPC	This study
NDM-1-F	TGCATTGATGCTGAGCGGGTG	bla NDM-1	This study
NDM-1-R	ATCACGATCATGCTGGCCTTG	bla NDM-1	This study
IMP-F	GGAATAGAGTGGCTTAAYTCTC	bla IMP	(Wen and Mi, 2010)
IMP-R	CCAAACYACTASGTTATCT	bla IMP	(Wen and Mi, 2010)
VIM-F	GATGGTGTTTGGTCGCATA	bla VIM	(Wen and Mi, 2010)
VIM-R	CGAATGCGCAGCACCAG	bla VIM	(Wen and Mi, 2010)
EBC-F	TCTATAAGTAAAACCTTCACCG	bla EBC	(Perez-Perez and Hanson, 2002)
EBC-R	CCAGGTATGGTCCAGCTTGAG	bla EBC	(Perez-Perez and Hanson, 2002)
ACC-F	AACAGCCTCAGCAGCCGGTTA	bla ACC	(Perez-Perez and Hanson, 2002)
ACC-R	TTCGCCGCAATCATCCCTAGC	bla ACC	(Perez-Perez and Hanson, 2002)
FOX-F	TTCGAGATTGGCTCGGTCAGC	bla FOX	(Perez-Perez and Hanson, 2002)
FOX-R	CAAAGCGCGTAACCGGATTGG	bla FOX	(Perez-Perez and Hanson, 2002)
DHA-F	AACTTTCACAGGTGTGCTGGGT	bla DHA	(Perez-Perez and Hanson, 2002)
DHA-R	CCGTACGCATACTGGCTTTGC	bla DHA	(Perez-Perez and Hanson, 2002)
CIT-F	TGGCCAGAACTGACAGGCAAA	bla CIT	(Perez-Perez and Hanson, 2002)
CIT-R	TTTCTCCTGAACGTGGCTGGC	bla CIT	(Perez-Perez and Hanson, 2002)
TEM-F	ACAGCGGTAAGATCCTTGAGAG	bla TEM	(Wen and Mi, 2010)
TEM-R	GAAGCTAGAGTAAGTAGTTCG	bla TEM	(Wen and Mi, 2010)
SHV-F	ACCTTTAAAGTAGTGCTCTGC	bla SHV	(Wen and Mi, 2010)
SHV-R	CACCATCCACTGCAGCAGCTG	bla SHV	(Wen and Mi, 2010)
CTX-M-F	ATGGTTAAAAAATCACTGCGYCAGTTC	bla CTX-M	(Wen and Mi, 2010)
CTX-M-R	TCACAAACCGTYGGTGACGATTTTAGCCGC	bla CTX-M	(Wen and Mi, 2010)
OXA-F	CTGTTGTTTGGGTTTCGCAAG	bla OXA	(Wen and Mi, 2010)
OXA-R	CTTGGCTTTTATGCTTGATG	bla OXA	(Wen and Mi, 2010)
PER-F	GCCTGACGATCTGGAACC	bla PER	(Wen and Mi, 2010)
PER-R	GATACTGCACCTGATCATC	bla PER	(Wen and Mi, 2010)
VEB-F	GCGTTATGAAATTTCCGATTG	bla VEB	(Wen and Mi, 2010)
VEB-R	CAACATCATTAGTGGCTGCTG	bla VEB	(Wen and Mi, 2010)
GES-F	ATGCGCTTCATTCACGCAC	bla GES	(Wen and Mi, 2010)
GES-R	CTATTTGTCCGTGCTCAGG	bla GES	(Wen and Mi, 2010)
1Int-F	ATCATCGTCGTAGAGACGTCGG	Class I Integron	(Wen and Mi, 2010)
1Int-R	GTCAAGGTTCTGGACCAGTTGC	Class I Integron	(Wen and Mi, 2010)
2Int-F	GCAAATAAAGTGCAACGC	Class II Integron	(Wen and Mi, 2010)
2Int-R	ACACGCTTGCTAACGATG	Class II Integron	(Wen and Mi, 2010)
3Int-F	GCAGGGTGTGGACGAATACG	Class III Integron	(Wen and Mi, 2010)
3Int-R	ACAGACCGAGAAGGCTTATG	Class III Integron	(Wen and Mi, 2010)
qnrA-F	ATTTCTCACGCCAGGATTTG	qnrA	(Mak JK, 2009)
qnrA-R	GATCGGCAAAGGTTAGGTCA	qnrA	(Mak JK, 2009)
qnrB-F	GATCGTGAAAGCCAGAAAGG	qnrB	(Mak JK, 2009)
qnrB-R	ACGATGCCTGGTAGTTGTCC	qnrB	(Mak JK, 2009)
qnrS-F	ACGACATTCGTCAACTGCAA	qnrS	(Mak JK, 2009)
qnrS-R	TAAATTGGCACCCTGTAGGC	qnrS	(Mak JK, 2009)
strA-F	ATGATGTCTAACAGCAAACTG	strA	(Mak JK, 2009)
strA-R	TCAACCCCAAGTAAGAGG	strA	(Mak JK, 2009)
strB-F	ATGGGGTTGATGTTCATGCCGC	strB	(Mak JK, 2009)
strB-R	CTAGTATGACGTCTGTCGCAC	strB	(Mak JK, 2009)
aphA1-F	ATGAGCCATATTCAACGGG	aphA1	(Mak JK, 2009)
aphA1-R	TCAGAAAAACTCATCGAGCATC	aphA1	(Mak JK, 2009)
gyrA-F	GCGATGTCGGTCATTGTTGG	gyrA	This study
gyrA-R	CCGAACTGGTCACGGATCAG	gyrA	This study
gyrB-F	CTCCGTCTCCGTACAGGATGAC	gyrB	This study
gyrB-R	TGTGATAGCGCAGTTTATCC	gyrB	This study

Genetic relatedness determination

The genetic similarities of the MDR K. pneumoniae strains were further analyzed using REP-PCR, a semi-automated PCR technique for the amplification of the regions between the non-coding repetitive sequences in bacterial genomes. REP-PCR was performed using the DiversiLab Klebsiella Kit (bioMérieux Inc., Durham, NC) according to the manufacturer’s instructions. The PCR products were separated on an Agilent B2100 Bioanalyzer using DNA LabChip reagents (Agilent Technologies Inc., Palo Alto, CA). The data were uploaded to the web-based DiversiLab system for further analysis. All results, including dendrograms, scatterplots, electropherograms, and virtual gel images, were illustrated using DiversiLab software (v.r.3.3.40). Pearson’s correlation coefficient and the unweighted-pair group method were used to determine the genetic similarity of the tested isolates. Strains with similarities > 98% are indistinguishable and designated as one pattern, and strains with similarities between 95% and 98% are homologous and designated as one group (Tenover, 2009; Marchaim, 2011).

Results

In vitro antimicrobial susceptibility

All 38 MDR K. pneumoniae strains showed resistance to aminoglycosides, macrolides, quinolones and β-lactams in the semi-automated systems. All isolates yielded MIC values ≥ 128 mg/L for cefotaxime and ceftazidime; ≥ 8 mg/L for imipenem, ertapenem and meropenem; ≥ 128 mg/L for kanamycin and amikacin; and ≥ 16 mg/L for ciprofloxacin. All MDR isolates were susceptible to colistin (≤ 1 mg/L), tigecycline (≤ 1 mg/L) and fosfomycin (≤ 64 mg/L). Among the 38 MDR K. pneumoniae isolates, 89.5% (34/38) were confirmed as ESBL producers, while 36.8% (14/38) were AmpC producers, and 31.5% (12/38) produced both ESBL and AmpC β-lactamases. The susceptibility characteristics of the MDR K. pneumoniae are shown in Tables 2 and 3.

Table 2

MICs for the 38 K. pneumoniae isolates, determined through agar dilution

Strain No.	MIC (mg/L)a
	AMC	SAM	TZP	CAZ	CRO	CTT	FEP	ATM	ETP	IPM	MEM	AMK	TOB	CIP	LVX
K25, K32	> 128	> 256	128	128	> 256	> 256	32	> 128	8	16	16	128	128	16	32
K7, K9, K22, K24, K1, K17, K19, K16, K15, K13, K31	> 128	> 256	> 256	> 128	128	> 256	32	> 128	16	32	16	128	> 128	32	64
K20, K21, K14	128	128	256	> 128	> 256	128	128	> 128	16	32	32	256	> 128	16	64
K33, K18, K8	> 128	> 256	> 256	> 128	> 256	> 256	128	> 128	> 32	> 32	> 32	> 256	64	32	32
K4, K36, K12, K6	> 128	> 256	> 256	> 128	> 256	> 256	64	> 128	8	16	16	128	64	32	32
K5, K4, K34, K11	> 128	> 256	> 256	> 128	> 256	> 256	128	> 128	16	32	32	> 256	128	32	> 64
K35, K23	> 128	> 256	> 256	> 128	> 256	> 256	256	> 128	8	16	16	256	128	16	64
K10, K37	> 128	> 256	> 256	> 128	> 256	> 256	64	> 128	32	> 32	> 32	> 256	> 128	> 32	> 64
K38, K2	> 128	> 256	> 256	> 128	> 256	> 256	128	> 128	16	32	32	256	128	32	16
K28, K29	> 128	> 256	> 256	> 128	> 256	> 256	32	> 128	16	> 32	32	> 256	> 128	16	> 64
K26, K27, K30	> 128	> 256	> 256	> 128	> 256	> 256	128	> 128	8	16	16	> 256	> 128	32	32

AMC, amoxicillin/clavulanate; SAM, ampicillin/sulbactam; TZP, piperacillin/tazobactam; CAZ, ceftazidime; CRO, ceftriaxone; CTT, cefotetan; FEP, cefepime; AET, aztreonam; ETP, Ertapenem; MEM, meropenem; IPM, imipenem; AMK, amikacin; TOB, tobramycin; KAN, kanamycin; CIP, ciprofloxacin; LVX, levofloxacin.

The MICs of the following antibiotics were identical for all 38 isolates: cefotaxime ≥ 128 mg/L; cefazolin ≥ 128 mg/L; ampicillin ≥ 512 mg/L; piperacillin ≥ 128 mg/L; cefazolin ≥ 128 mg/L; cefuroxime ≥ 128 mg/L; gentamycin ≥ 128 mg/L; kanamycin ≥ 64 mg/L; chloramphenicol ≥ 32 mg/L; sulfafurazole ≥ 128 mg/L; norfloxacin ≥ 128 mg/L; nitrofurantoin ≥ 128 mg/L.

Table 3

Resistance determinants detected through the PCR screening of 38 K. pneumoniae isolates

Strain No.				Resistance determinants in K. pneumoniae
	ESBL	AmpC	Integron	bla SHV	bla CTX-M	bla TEM	bla OXA	bla DHA	bla CIT	bla VIM	bla KPC	gyrA mutation	qnrB	aacC1	aacC4	aadA1	strB
K25, K32	+	−	Class 2	−	+	−	−	−	+	−	+	+	−	−	−	−	+
K7, K9, K22, K24, K1, K17, K19, K16, K15, K13, K31, K34	+	−	Class 1	+	−	−	−	−	−	−	+	+	−	+	+	+	−
K20, K21, K14, K33, K18, K8, K4, K36, K12, K6	+	−	Class 1	−	−	+	−	−	−	−	+	−	+	+	+	+	−
K35, K23, K10, K5, K4, K11	+	+	Class 2	+	−	−	−	+	−	−	+	+	−	−	−	−	+
K38, K2	+	+	Class 2	+	−	−	−	+	−	−	+	+	+	−	−	−	−
K29	+	+	Class 1	−	−	−	+	−	−	+	−	−	+	+	+	+	−
K26, K27, K38, K28, K37	+	+	Class 1	−	+	−	−	+	−	−	+	+	−	+	+	+	−

+ and − indicate the presence and absence, respectively, of a particular resistance determinant.

Detection and characterization of integrons

PCR was used to assess the presence of integrons in the MDR K. pneumoniae strains as previously described (Mak, 2009). Class I and II integrons were detected in 73.7% (28/38) and 26.3% (10/38) of the strains, respectively. No class III integrons were detected. Subsequent PCR amplification and sequencing revealed that all 28 class I integron-positive K. pneumoniae isolates contained aacC1, which confers a high level of gentamicin resistance, aacC4, which encodes amikacin and tobramycin resistance, aadA1, which confers spectinomycin and streptomycin resistance, and two unknown open reading frames, orfX and orfX’.

Aminoglycoside resistance

All 38 MDR isolates were resistant to streptomycin, while only 28 class I integron-positive isolates contained aadA1, aacC1 and aacC4. Therefore, the other two streptomycin resistance genes, strA and strB, were screened through PCR. Eight class II integron-positive K. pneumoniae isolates containing the strB gene were detected, whereas no isolates containing the strA gene were observed. The control strain K. pneumoniae ATCC 700603 was completely sensitive to both spectinomycin and streptomycin; this strain also lacked aadA1, strA and strB genes.

Genetic basis of β-lactam resistance

The PCR analyses demonstrated that ESBL-producing MDR K. pneumoniaestrains were rich in β-lactamase genes, such as bla SHV(22/38), bla TEM (10/38) and bla CTX-M (7/38) (Table 3). Non-ESBL producers also harbored at least one β-lactamase gene. bla SHV-12 and bla CTX-M-9 genes were highly prevalent, accounting for 42% and 13%, respectively, of class A β-lactamase genes, while bla CTX-M-1 was not detected. The bla DHA gene was detected in 11 ESBL-positive strains and 2 ESBL-negative strains, representing the most prevalent AmpC gene. Notably, the bla DHA gene was detected in all qnr-positive strains and 5 qnr-negative strains (Table 2 and 3). PCR analysis revealed that 37 carbapenem-resistant isolates harbored KPC carbapenemase. Sequencing and alignment analyses revealed the presence of a highly conserved bla KPC-2 allele in KPC-positive isolates, and bla OXA-23 and bla VIM were only detected in KPC-negative isolates. bla CMY, bla ACC, bla FOX, bla PER, bla VEB, bla GES, bla IMP and bla SPM were not detected in any of the clinical isolates examined.

Quinolone resistance is conferred through a gyrA site mutation or the qnrB gene

All 38 MDR isolates showed resistance to norfloxacin (MIC ≥ 256 mg/L) and ciprofloxacin (MIC ≥ 16 mg/L). To identify the quinolone resistance determinants, three subtypes of qnr genes (qnrA, qnrB, and qnrS) and two subtypes of gyr genes (gyrA and gyrB) were screened through PCR and sequencing analyses. Among the 38 MDR K. pneumoniae isolates, 13 isolates contained the plasmid-mediated quinolone resistance determinant qnrB, and 27 isolates contained the DNA gyrase gene gyrA, containing a nucleotide mutation in the quinolone-resistance determining region (QRDR). This mutation results in an amino acid change of S83L, which confers quinolone resistance in many bacteria. The gene qnrB was present in isolates that did not contain the gyrA mutation, and high rates of the qnrB4 subtype were detected among qnrB-positive isolates (10/12). The plasmid-mediated qnrB gene was also detected among 27 gyrA mutation strains of the two isolates K38 and K2. The qnrS gene was not detected in any of the MDR K. pneumoniae strains. The control strain K. pneumoniae ATCC 700603, which did not carry the gyrA mutation or any qnr gene, was susceptible to quinolones. These findings suggest that the S83L gyrA site mutation conferred quinolone resistance to 27 of 38 isolates, and qnrB conferred quinolone resistance to the other 11 isolates. qnrA, qnrS and mutations in gyrB were not detected in any of the 38 MDR K. pneumoniae isolates.

REP-PCR and genetic diversity

To characterize the genetic diversity of the 38 MDR K. pneumoniaeisolates with similar multidrug resistance profiles, strain typing based on REP-PCR was performed using genomic DNA. Five REP-PCR groups and sixteen patterns were observed (Figure 1). K7, K9, K22, K24, K1, K11, K17, K19, K16, K15, K13, and K31 showed high similarity (> 95%) in group II, designating the type 3 REP-PCR pattern. Moreover, the strains in group II pattern III, except for K19 (isolated from a bedrail) and K33 (isolated from bedclothes), were isolated from ICU patients. All group II strains, except K11 and K2, possessed class I integrons and aphA. Therefore, REP-PCR lacked the power to discriminate MDR K. pneumoniae strains, suggesting that a combination of genotypic data and REP-PCR types should be used to group the 38 isolates.

Figure 1

Dendrograms and virtual gel images of the MDR K. pneumoniaeisolates identified using web-based DiversiLab software

Discussion

The incidence of MDR K. pneumoniae infections has increased during the last decade, reflecting the extensive use of antimicrobial agents (Wu ). MDR K. pneumoniae is considered as a significant health problem because of limited options for antibiotic treatment. In this study, we conducted the phenotypic characterization and investigated the molecular mechanisms of antibiotic resistance for MDR K. pneumoniae strains isolated from ICUs. REP-PCR-based homology analyses provide information concerning the epidemic status of nosocomial infections caused by MDR K. pneumoniae.

In this study, 89.5% of the MDR K. pneumoniae strains isolated from ICU patients during a 12-month period were ESBL producers. This percentage was much higher than those reported in previous studies concerning other types of bacteria in China (Huang ; Wang ; Yong-Hong, 2012; Chen ). The ESBL-positive MDR K. pneumoniae strains possessed at least seven resistance determinants conferring resistance phenotypes to nine classes of antimicrobials. This finding was troubling, as most clinical therapeutic regimens are restricted to a maximum of 4 classes. No significant associations among ESBL-producing strains, antibiotic susceptibilities and integron carriage profiles were observed. However, isolates with similar resistance determinant profiles showed similar antibiotic resistance phenotypes. Notably, class I integrons were predominantly detected among ESBL-positive K. pneumoniaeisolates (82.4%), consistent with the results of previous Asian studies. However, this percentage is slightly higher than those reported in Australian and American studies and much higher than those reported in European studies (Wu ; Rao ; Yao ; Machado ; Jiang ). According to the results obtained in the present study, 26.3% of the MDR K. pneumoniaeisolates and 15.7% of the ESBL-producing isolates harbored class II integrons. Several studies have also reported the distribution of class II integrons among the Enterobacteriaceae family, including Escherichia coli, Salmonella and Enterobacter cloacae, and Acinetobacter baumannii in China (Zhou ; Gu ; Li ). The co-existence of class I and II integrons was not observed in any of the MDR K. pneumoniae strains. To our knowledge, there has only one report describing the co-existence of class I and II integrons among Enterobacteriaceae in China (Li ).

The prevalence and wide distribution of plasmid-mediated quinolone resistance determinants has been demonstrated among clinical isolates in China. In the present study, plasmid-mediated quinolone resistance determinants (qnr genes) conferred quinolone resistance in 13 K. pneumoniae strains. Furthermore, at least one ESBL subtype (CTX-M, TEM and SHV) or AmpC β-lactamase (DHA) was present in qnr-positive K. pneumoniae isolates (Table 3). Previous studies have also shown that most ESBL genes are located in transposons and self-transferable plasmids, which typically co-exist with plasmid-mediated quinolone resistance determinants (Jiang ; Hassan ; Park ; Han ). This phenotype might reflect the excessive use of cephalosporins and quinolones and the absence of strict antimicrobial policies in medical care facilities. The most frequent ESBL genotype among the qnr-positive K. pneumoniae strains was SHV-12, a result that was not consistent with results showing that SHV-7 was more common in qnr-positive isolates (Jiang ).

Carbapenems have been considered as last resort treatments against infections caused by MDR Gram-negative organisms. However, Klebsiella has developed an efficient carbapenem resistance mechanism, known as KPC (Klebsiella pneumoniaecarbapenemase) (Naas ). In China, KPC-producing K. pneumoniae was first identified in 2007 (Leavitt ; Cuzon ). Since then, the prevalence of KPC-producing K. pneumoniae has increased to epidemic proportions, particularly in large and public city hospitals. An increase in carbapenem-resistant strains has recently been reported in China (Cai ), likely reflecting the horizontal and vertical transmission of KPC-encoding genes, overcrowded environments and insufficient infection control measures in these hospitals. In the present study, sequencing and alignment revealed the presence of the bla KPC-2 allele in 37 MDR K. pneumoniaeisolates. The bla KPC-2 genes amplified from these isolates were highly conserved, and no nucleotide differences were observed. However, the KPC-negative isolate was also OXA-23-positive. To our knowledge, few studies have revealed the contribution of bla OXA-23 to carbapenem resistance in other bacteria, with the exception of Acinetobacter baumannii. Because the bla OXA-23 gene identified in the KPC-negative K. pneumoniae isolate showed 100% identity to Acinetobacter baumannii bla OXA-23 (data not shown), we speculated that bla OXA-23 was transferred from A. baumanniithrough mobile genetic elements, which have been implicated in the dissemination of antimicrobial resistance through horizontal transmission. Previous studies have shown that integrons contribute to the prevalence of transferable extended-spectrum cephalosporin resistance through Enterobacteriaceae (Pisney ; Magiorakos ). Therefore, an in-depth investigation is required to study the mechanism of reduced susceptibility and to predict the clinical efficiency of carbapenem drugs.

The results of the present study demonstrated that the streptomycin resistance of class I integron-positive K. pneumoniae isolates was conferred through the aadA1 gene cassette. However, the widespread streptomycin resistance gene strB was only detected in class II integron-positive isolates. So far, the genes strA and strB are considered as a gene pair, comprising a mobile genetic element of the conjugative antibiotic resistance plasmid, which confers high-level streptomycin resistance in bacteria (Chiou et al., 1995; Tauch et al., 2003). We did not detect the strA gene in any of the isolates examined; thus, it is likely that the strB gene is an element of class II integrons. Currently, we are examining the genetic elements of class II integrons and the mechanism of streptomycin resistance in two other class II integron-positive strains, which have been demonstrated as strB-negative.

To provide evidence for the prevention and control of nosocomial infections, the genotypic similarities of MDR K. pneumoniae strains were investigated using the REP-PCR-based DiversiLab system. Five different groups and sixteen patterns of MDR K. pneumoniae were isolated from ICU patients. The MDR K. pneumoniae isolated from the ventilators and bed rails shared genetic similarities with the strains isolated from patients. The homology between some strains isolated from ventilators and bedrails suggested that the equipment or object surfaces in the ICU had persistent MDR K. pneumoniae colonization. Although REP-PCR is a powerful tool for strain typing, with resolution beyond the species level, this technique lacked the power to discriminate group II strains, as all of the MDR strains in group II, except K11 and K2, possessed class I integrons and aphA. Furthermore, the K. pneumoniae strains in the same clone group could possess different resistance phenotypes and different drug resistance determinants. Therefore, REP-PCR fingerprinting should be combined with other detection methods to distinguish genetic groups beyond the species level.

In conclusion, MDR K. pneumoniae is becoming a serious problem in ICUs, with many strains developing resistance to most available antibiotics. The results of this study provide evidence for appropriate surveillance and outbreak investigations. We propose that infection control measures and strict antimicrobial stewardship policies should be applied to reduce the selective pressure that inevitably favors the emergence and epidemic of MDR strains and to increase the therapeutic usefulness of these antibiotics.