Analysis of Drug Resistance Determinants in Klebsiella pneumoniae Isolates from a Tertiary-Care Hospital in Beijing, China

Background The rates of multidrug-resistant (MDR), extensively drug-resistant (XDR) and pandrug-resistant (PDR) isolates among Enterobacteriaceae isolates, particularly Klebsiella pneumoniae, have risen substantially worldwide. Methodology/Principal Findings To better understand the molecular mechanisms of drug resistance in K. pneumoniae, we analyzed the drug resistance determinants for K. pneumoniae isolates collected from the 306 Hospital, a tertiary-care hospital in Beijing, China, for the period of September 1, 2010-October 31, 2011. Drug susceptibility testing, PCR amplification and sequencing of the drug resistance determinants were performed. Conjugation experiments were conducted to examine the natural ability of drug resistance to disseminate among Enterobacteriaceae strains using a sodium azide-resistant Escherichia coli J53 strain as a recipient. Among the 223 consecutive non-repetitive K. pneumoniae isolates included in this study, 101 (45.3%) were extended-spectrum beta-lactamases (ESBLs) positive. The rates of MDR, XDR, and PDR isolates were 61.4% (n = 137), 22.0% (n = 49), and 1.8% (n = 4), respectively. Among the tested drug resistance-associated genes, the following ones were detected at relatively high rates bla CTX-M-10 (80, 35.9%), aacC2 (73, 32.7%), dhfr (62, 27.8%), qnrS (58, 26.0%), aacA4 (57, 25.6%), aadA1 (56, 25.1%). Results from conjugation experiments indicate that many of the drug resistance genes were transmissible. Conclusions/Significance Our data give a “snapshot” of the complex genetic background responsible for drug resistance in K. pneumoniae in China and demonstrate that a high degree of awareness and monitoring of those drug resistance determinants are urgently needed in order to better control the emergence and transmission of drug-resistant K. pneumoniae isolates in hospital settings.


Introduction
The emergence and rapid spread of drug-resistant Klebsiella pneumoniae isolates is becoming a serious antibiotic management problem and causing a great concern worldwide [1][2][3][4][5][6]. For example, by late 2009, the number of unique protein sequences for beta-lactamases exceeded 890 (http://www.lahey.org/Studies) [7]. There is an increasing recognition of isolates producing newer beta-lactamases including the extended-spectrum betalactamase (ESBL), carbapenem-hydrolyzing enzymes (e.g., K. penumoniae carbapenemase [KPC] types and the metallo-betalactamases [MBLs]) [8][9][10][11][12][13][14][15][16][17][18]. Since the production of newer betalactamases is frequently accompanied by broad-spectrum resistance, the ESBL positivity together with the existence of newer beta-lactamases should be monitored closely as the emergence of those highly drug-resistant K. pneumoniae strains will pose a serious impact on the remaining therapeutic options [19][20][21][22]. In a study based on the Tigecycline Evaluation and Surveillance Trial (TEST) global surveillance database, the rate of ESBL production was highest among the K. pneumoniae isolates collected in Latin America, followed by Asia/Pacific Rim, Europe, and North America (44.0%, 22.4%, 13.3% and 7.5%, respectively) [23]. Thus the potential of drug resistant K. pneumoniae to be a global health problem is great and more intensive surveillance and more in-depth investigation into the molecular mechanisms of drug resistance in K. pneumoniae isolates are necessary in order to provide information for the development of effective molecular diagnostic methods and novel drugs against K. pneumoniae infection.
In the face of increasing resistance among multidrug-resistant (MDR) gram-negative organisms for which no adequate therapeutic options exist, a joint initiative by the European Centre for Disease Prevention and Control (ECDC) and the Centers for Disease Control and Prevention (CDC) recently created a standardized international definitions for MDR, extensively drug-resistant (XDR) and pandrug-resistant (PDR) with an aim to enhance the comparability of data and promote better comprehension of the problem of highly drug-resistant bacteria [24]. MDR was defined as acquired non-susceptibility to at least one agent in three or more antimicrobial categories, XDR was defined as non-susceptibility to at least one agent in all but two or fewer antimicrobial categories (i.e. bacterial isolates remain susceptible to only one or two categories) and PDR was defined as non-susceptibility to all agents in all antimicrobial categories [24]. Though there are already many reports of drug-resistant K. pneumoniae worldwide, the extent of MDR, XDR and PDR K. pneumoniae isolates among patients is largely unknown. We thus in this study sought to determine the prevalence of MDR, XDR and PDR strains and to analyze the drug resistance determinants for K. pneumoniae isolates collected from patients being treated in the 306 Hospital, a tertiary-care hospital in Beijing, China, for the period of September 1, 2010-October 31, 2011 with an aim to better understand the current situation as well as the genetic background of the drug resistant K. pneumoniae isolates from hospital settings.

Ethics Statement
All of the investigation protocols in this study were approved by the institutional ethics committee of the 306 Hospital, Beijing, China. Written consent was given by the patients for their information to be stored in the hospital database and used for research. Permission for using the information in the medical records of the patients for research purposes was obtained from the 306 Hospital. The Institute ethics committee of the 306 Hospital reviewed that relevant ethical issues in this study were well considered.

Study Population, Bacterial Isolate Identification, and Drug Susceptibility Testing
This is a prospective surveillance study. Consecutive K. pneumoniae isolates were collected from unique patients being treated in the 306 Hospital in Beijing, China (which is a 1,100-bed tertiary-care hospital serving approximately 25,000 in-patients per year) for the period of September 1, 2010-October 31, 2011. In the case of duplicate patient samples, the first collected isolate was chosen. All strains were cultured in Luria-Bertani (LB) medium. The K. pneumoniae strains were confirmed by phenotypic tests and 16 S rDNA sequencing. Drug susceptibility testing (DST) for the K. pneumoniae strains was performed using the bioMérieux VITEK-2 AST-GN13 system following manufacturer's instructions. The following 18 drugs were tested: ampicillin (AMP), piperacillin/ tazobactam (TZP), ampicillin/sulbactam (SAM), cefazolin (CFZ), ceftriaxone (CRO), ceftazidime (CAZ), cefepime (FEP), cefotetan (CTT), ertapenem (ETP), imipenem (IMP), aztreonam (ATM), ciprofloxacin (CIP), levofloxacin (LVX), gentamicin (GM), tobramycin (TOB), amikacin (AMK), trimethoprim-sulfamethoxazole (SXT), furadantin (FD). The ESBLs were detected by the bioMérieux VITEK-2 AST-GN13 test (which is claimed to be a confirmatory ESBL test). In some cases, the ESBL positivity was further confirmed by the double disk diffusion method [25]. Escherichia coli strains ATCC 25922 and ATCC 35218, K. pneumoniae strain ATCC 700603 and Pseudomonas aeruginosa strain ATCC 27853 were used as quality control strains for the DST. Clinical records of patients from whom the K. pneumoniae isolates were obtained were reviewed retrospectively.

PCR Amplification and Sequencing
Genomic DNA was extracted using DNeasy Tissue kit (Qiagen; Valencia, CA, USA). Drug resistance-associated genes were detected by PCR and sequencing using 37 pairs of primers listed in Table 1. Direct sequencing of positive amplicons was conducted. The primers were synthesized by the Beijing Genomics Institute (BGI, China). PCR was performed in a 50-mL reaction mixture consisting of 5 mL of 106PCR buffer, 2.5 units of Taq DNA polymerase (Takara), 0.2 mM of dNTPs, 0.4 mM each of the primer, and 1 mL chromosomal DNA. All reaction mixtures were subjected to 30 cycles of 94uC for 1 min, 55uC for 1 min, and 72uC for 2 min. PCR products were purified and sequenced bidirectionally with the same primers used for PCR by the Beijing Genomics Institute (BGI, China). DNA sequences were annotated using the BLAST program at http://www.ncbi.nlm.nih.gov. Mutations in the gyrA and parC genes were identified by comparing the DNA sequences with gyrA and parC sequences of the K. pneumoniae (GenBank accession numbers DQ673325 and NC009648 for gyrA and parC, respectively).

Conjugation Experiments, Plasmid Analysis, and MLST Analysis
Transfer of resistance genes by conjugation experiments were carried out in LB broth using clinical isolates as donors and the E. coli J53AzR as the recipient as described previously [26]. Cultures of donor and recipient cells in logarithmic phase (0.5 mL each) were added to 4 mL of fresh LB broth and incubated overnight without shaking. Transconjugants were selected on LB plates containing 100 mg/mL sodium azide for counterselection and 100 mg/mL ampicillin to select for plasmid-encoded resistance. The drugs tested were purchased from Sigma Chemical Co. Plasmid DNA from the K. pneumoniae donor strains and E. coli transconjugants were prepared using the Plasmid Maxprep Kit (Vigorous Biotechnology, Beijing, China) and were separated on 0.7% agarose gels. Genotyping was determined by MLST analysis. MLST with seven genes (gapA, infB, mdh, pgi, phoE, rpoB and tonB) was performed on isolates according to the protocol described on the K. pneumoniae MLST website (www.pasteur.fr/ mlst). Sequence types (STs) were assigned by using the MLST database (www.pasteur.fr/mlst/Kpneumoniae.html).

Statistical Analysis
SPSS software (version 15.0) was used for data analysis. Categorical variables were compared with the chi-square test or Fisher's exact test. A p value of ,0.05 was considered to be statistically significant.

Demographic and Clinical Characteristics of the Patients
From September 1, 2010 to October 31, 2011, a total of 223 non-repetitive patients at the 306 Hospital who had K. pneumoniae isolates were subjected to DST using 18 antibiotics. Among which, 137 (61.4%) were MDR isolates, 49 (22.0%) were XDR isolates, 4 (1.8%) were PDR isolates, and 33 (14.8%) were other types of isolates. The proportion of the male and female were 73.5% (n = 164) and 26.5% (n = 59), respectively. Sixty-eight (30.5%) of the patients were Beijing residents and the rest were from other provinces of China (non-Beijing residents). The median (6SD) age of the patients was 74.0620.3 years (range 1.0-98.0 years). The majority of the patients were from medical ward (97, 43.5%) and intensive care unit (75, 33.6%). The main source of the specimens was sputum (168, 75.3%). The proportion of the ESBL positive cases was 45.3% (n = 101). The proportion of XDR (42,41.6%) and PDR (4, 4.0%) cases was significantly higher among patients whose isolates were ESBL positive as compared with those whose isolates were ESBL negative. In addition, the proportion of MDR cases (83, 68.0%) and other types of cases (32, 26.2%) was significantly higher in patients with ESBL-negative isolates than that observed for XDR (7,5.7%) and PDR cases (0). The detailed information on relevant demographic and clinical characteristics of the study population is summarized in Table 2.

Characteristics of Carbapenem-resistant K. pneumoniae Isolates
Among the 223 isolates, 16 were detected to be carbapenemresistant and were used for further characterization. Most of the patients from whom the isolates obtained were male 87.5% (14/ 16) and aged patients (all were 56 years old or above). Among the 14 patients whose treatment outcome information was available, 6 died. Although carbapenemase genes were detected only in 7 of the 16 isolates, the majority of them exhibited resistance to a high number of drugs and contained a variety of corresponding drug resistance-associated genes. The proportion of MDR, XDR and PDR were 12.5% (n = 2), 62.5% (n = 10), 25.0% (n = 4), respectively. The more detailed characteristics of those carbapenemresistant isolates are shown in Table S1.

Resistance-associated genes
Resistance-associated genes detected in phenotypic resistant isolates, n/N a (%) Resistance-associated genes detected in phenotypic susceptible isolates, n/N b (%)

Resistance-associated genes detected in all isolates, n/N c (%)
Antipseudomonal penicillins + beta -lactamase inhibitors, penicillins + beta -lactamase inhibitors, 1st and 2nd generation cephalosporins, 3rd and 4th generation cephalosporins, cephamycins (n = 204) 208) (Figure 1). In addition, the transconjugant for TZSKP-82 harbored new plasmid with different size than that in the donor strain. The phylogenetic tree based on the MLST analysis results for the isolates is shown in Figure 2. Seven different STs were identified for those 12 isolates. Three isolates (TZSKP-1, 9, and 82) belonged to ST15, three isolates (TZSKP-13, 15, and 17) belonged to ST11, two isolates (TZSKP-228 and 245) belonged to ST218, and the rest of the isolates had unique STs. The epidemiological links were further determined for the patients from whom the clustered isolates were obtained.

Discussion
Worldwide emergence and dissemination of ESBL and carbapenemase genes among Enterobacteriaceae, especially in K. pneumoniae isolates, poses a considerable threat to public health. The major goal of this study was to evaluate the current situation and genetic background of drug-resistant K. pneumoniae isolates from patients in hospital settings. The highest and lowest resistance rates were observed for penicillins (99.6%) and carbapenems (7.2%), respectively. The rates of MDR, XDR and PDR isolates observed in this study are alarmingly high. This could cause difficulty in treating K. pneumoniae-associated infections since fewer and fewer effective drugs are available for treating those highly drug-resistant isolates. We also found that the proportion of MDR and other types of cases was significantly higher in patients with ESBL-negative isolates than that for XDR and PDR cases. This could be partially explained by the fact that the ESBL-positive isolates are normally resistant to many drugs, leaving only a few effective drugs available for treatment, which could lead to further resistance to those drugs. Indeed, this study further reveals that resistance to most of the drugs was found to be associated with ESBL positivity. Infections due to those ESBL positive and highly resistant strains are reported to be associated with higher morbidity and mortality rates [19][20][21][22], thus globally coordinated surveillance of epidemiology of those resistant isolates are warranted.
Another goal of this study was to evaluate the correlation between resistance phenotypes and the genetic determinants clinical K. penumoniae isolates, so as to give a ''snapshot'' of the background of those resistant isolates. A striking feature of this study is the large number of antibiotic resistance-associated genes detected in the examined isolates. We also found that while some of the previously reported resistance-associated genes were indeed detected at relatively higher rates among corresponding phenotypic resistant isolates, some others were detected in very low proportion of the phenotypic resistant isolates, suggesting the existence of unknown drug resistance mechanisms such as reduced permeability of the outer membrane or up-regulated unknown efflux pumps in some clinical isolates [27,28]. In addition, some of Table 5. Percentage of non-beta-lactamase antibiotics resistance-associated genes detected in K. pneumoniae isolates.

Target antimicrobial category
Resistance-associated genes Resistance-associated genes detected in phenotypic resistant isolates, n/N a (%) Resistance-associated genes detected in phenotypic susceptible isolates, n/N b (%)   Table 6. Transmissibility of drug resistance of ESBL positive MDR, XDR and PDR K. pneumoniae isolates by conjugation.
K. pneumoniae isolates Resistance profile of K. pneumoniae isolates a,b Resistance-associated genes detected in K. the resistance-associated genes were also detected in a sizable proportion of the phenotypic susceptible isolates, suggesting that individual resistance gene alone is not sufficient to cause resistance phenotype and only when some of them were accumulated can the resistance become detectable in the clinical isolates. Rapid detection of genetic determinants associated with drug resistance in clinical K. pneumoniae isolates is crucial for appropriate antimicrobial therapy and infection control measures. We detected relatively high percentage of previously reported genes associated with resistance to beta-lactams [2,10,[29][30][31][32], fluoroquinolones [28,[33][34][35][36][37], aminoglycosides [38,39], and folate pathway inhibitors [15] in K. pneumoniae isolates. CTX-M-type beta-lactamase genes (such as bla CTX-M-14 and bla CTX-M-15 ) have been reported to be prevalent worldwide [40][41][42]. For example, a recent study from China reported that among the 21 K. pneumoniae isolates from 1270 specimens collected in a prospective multi-center study in eight teaching hospitals in China from June to December in 2007, 3 were detected to have bla CTX-M-14 (3,14.3%) [40]. Another study conducted in Scotland showed that 16 of the 219 (7.3%) clinical isolates of K. pneumoniae collected in 2006 and 2007 at the Royal Infirmary of Edinburgh, Scotland had bla CTX-M-15 [41]. In the present study, the highest rate of CTX-M-type beta-lactamase genes was observed for bla CTX-M-10 (35.9%), followed by bla CTX-M-1 (16.6%) , bla CTX-M-14 (16.6%) , and bla CTX-M-15 (15.2%) . The rates of the non-ESBL SHV-type beta-lactamase genes bla SHV-1 and bla SHV-11 genes were 24.7% and 21.1%, respectively in this study, which were relatively lower compared to that reported by some previous studies conducted in other regions. For example, according to a study conducted in Korea, the rates of the bla SHV-1 and bla SHV-11 genes among K. pneumoniae isolates collected from May to July, 2002 were 35% (50/142) and 62% (62/142), respectively [43]. Another study from Brazil reported that 55.8% (29/52) of the K. pneumoniae isolates collected in Recife, PE, Brazil during 1998 to 2005 harbored the bla SHV genes [44]. Thus, the prevalence of some beta-lactamase genes such as the bla SHV genes could be greatly variable geographically and timewise.
The more recently reported carbapenemases genes (such as bla IMP , bla VIM , bla NDM , plasmid-mediated clavulanic acid-inhibited class A beta-lactamases genes such as bla KPC , and the class D betalactamase gene bla OXA-48 ) were rarely detected or undetected in this study. Carbapenemases increasingly have been reported in Enterobacteriaceae in the past decade. KPC carbapenemases have been reported in the United States and then worldwide [6,8]. VIM and IMP metallo-beta-lactamases also have been reported in many regions of the world, with a higher prevalence in southern Europe and Asia [1,5,[45][46][47]. Carbapenemases of the oxacillinase-48 type (OXA-48) have been identified mostly in Mediterranean and European countries and in India [48,49]. Although the worldalarming New Delhi metallo-beta-lactamase-1 (NDM-1) was not detected in this study, it has been detected worldwide since it was first identified in India and its variants have emerged [13,[50][51][52]. Thus resistance caused by those recently emerging beta-lactamases is still worrisome and needs continuous monitoring. The association of alterations in gyrA (gene encoding for GyrA subunit of DNA gyrase) and parC (gene encoding for ParC subunit of DNA topoisomerase IV) with fluoroquinolone resistance in K. pneumoniae is still not clear. Some studies suggested that in K. pneumoniae, DNA gyrase A is a primary target of quinolones and that ParC alterations play a complementary role in the development of higher-level fluoroquinolone resistance [53,54], while a study reported that hypermutation in K. pneumoniae is uncommon and does not contribute to accumulation of gyrA mutations or directly to ciprofloxacin resistance [55]. We identified 3 types of gyrA mutations including the previously reported C248T (Ser83Phe) and A260C (Asp87Ala) [53] and the unreported T247A (Ser83Ile). No mutations in parC were detected in this study. The plasmidencoded 16 S rRNA methylases armA and rmtB has emerged as a new mechanism of resistance to aminoglycosides, and the concomitant presence of armA or rmtB with bla CTX-M type betalactamase genes, especially the group 1 (CTX-M-3 and CTXM-15) or group 9 (CTX-M-14), among amikacin-resistant ESBLproducing K. pneumoniae isolates was reported in Taiwan and Belgium [56,57]. In this study, both armA and rmtB were detected (5.8% and 3.6%, respectively). One isolate was found to harbor both armA and rmtB genes, and consistent with previous reports, the armA and rmtB genes were coexistent with at least one of the bla CTX-M type beta-lactamases tested in this study.
Our study further confirmed the notion that patients infected with carbapenem-resistant K. pneumoniae isolates normally have worse treatment outcome. In addition, the conjugation results suggest that certain ESBL genes (such as bla CTX-M-14 ), aac(69)-Ib-cr and aacA4 were frequently co-transmitted and co-selected in MDR, XDR and PDR isolates and can be naturally transferred to susceptible E. coli strains by conjugation. Five transconjugants contained plasmids with the same size as those in their respective donors. Nevertheless, plasmids of the same sizes found in both donor and recipient isolates could not guarantee that the resistance transfer was plasmid-mediated. A subsequent DNA-DNA hybridization experiment with probes made by the respective resistance genes is warranted to show that the plasmids of the same sizes do carry the same resistance genes. On the other hand, we noticed that plasmids were not identified in some of the donor and recipient isolates. This could be explained by the existence of some other non-plasmid-mediated mechanisms involved in the occurrence and transfer of drug resistance in those isolates. For example, the drug resistance-associated genes could be carried on chromosomally located transposons and integrons. Although the isolate TZSKP-28 was detected to be ESBL-positive, only the non-ESBL bla SHV-11 gene was detected in it. After conjugation experiments, the recipient strain became multidrug resistant and again only the bla SHV-11 was detected. We also noticed that the beta-lactamase genes found in the transconjugants of the isolates TZSKP-1 and TZSKP-40 (bla SHV-11 or bla TEM-1 ) were not ESBLs, either. This result suggested that some other mechanisms may be involved in causing the ESBL positivity and MDR phenotypes in those isolates. As horizontal transmission event can result in the acquisition of multidrug resistance by wild-type strains, thus this could presumably contribute to the rapid increase in the prevalence of multidrug resistance among clinical bacteria. Another important aspect for infection control is to know whether there is a clonal spread among the highly drug-resistant isolates. Relatively diverse genotypes were identified for those 12 isolates used in conjugation analysis by MLST analysis. Three clusters (which belonged to ST15, ST11, and ST218, respectively) each consisted of 2 or 3 isolates were detected and the epidemiological links were not observed for the patients from whom the clustered isolates were obtained. Further investigation of the transmission patterns of a larger sample of drug-resistant K. pneumoniae isolates by MLST analysis is warranted and which is currently underway.
Emerging plasmid-encoded ESBLs and carbapenemases are increasingly reported worldwide [58]. Carbapenemase production encoded by genes located on mobile genetic elements is typically accompanied by genes encoding resistance to other drug classes, and are frequently located on the same mobile DNA elements such as integrons, which act as bacterial recombination systems that mediate the capture and expression of gene cassettes and are considered as the primary mechanism for antibiotic resistance gene acquisition among bacteria and are frequently associated with transposons and conjugative plasmids (http://integrall.bio. ua.pt 2009) [59,60]. In this study, class 1 integrons were detected in a relatively high percentage of the isolates. Further sequencing analysis of class 1 integrons and gene cassette arrays is currently undergoing in the lab.
In summary, our results indicate that there is a high prevalence and possible transmission of MDR, XDR and PDR K. pneumoniae isolates among hospitalized patients. In addition, our data give a ''snapshot'' of the complex genetic background responsible for drug resistance in those highly drug-resistant K. pneumoniae isolates. Thus our study demonstrate that a high degree of awareness and monitoring of those drug resistance determinants are urgently needed in order to better control the emergence and transmission of drug-resistant K. pneumoniae isolates in hospital settings.