Frequency, Antimicrobial Resistance and Genetic Diversity of Klebsiella pneumoniae in Food Samples

This study aimed to assess the frequency of Klebsiella pneumoniae in food samples and to detect antibiotic resistance phenotypes, antimicrobial resistance genes and the molecular subtypes of the recovered isolates. A total of 998 food samples were collected, and 99 (9.9%) K. pneumoniae strains were isolated; the frequencies were 8.2% (4/49) in fresh raw seafood, 13.8% (26/188) in fresh raw chicken, 11.4% (34/297) in frozen raw food and 7.5% (35/464) in cooked food samples. Antimicrobial resistance was observed against 16 antimicrobials. The highest resistance rate was observed for ampicillin (92.3%), followed by tetracycline (31.3%), trimethoprim-sulfamethoxazole (18.2%), and chloramphenicol (10.1%). Two K. pneumoniae strains were identified as extended-spectrum β-lactamase (ESBL)–one strain had three beta-lactamases genes (blaSHV, blaCTX-M-1, and blaCTX-M-10) and one had only the blaSHV gene. Nineteen multidrug-resistant (MDR) strains were detected; the percentage of MDR strains in fresh raw chicken samples was significantly higher than in other sample types (P<0.05). Six of the 18 trimethoprim-sulfamethoxazole-resistant strains carried the folate pathway inhibitor gene (dhfr). Four isolates were screened by PCR for quinolone resistance genes; aac(6’)-Ib-cr, qnrB, qnrA and qnrS were detected. In addition, gyrA gene mutations such as T247A (Ser83Ile), C248T (Ser83Phe), and A260C (Asp87Ala) and a parC C240T (Ser80Ile) mutation were identified. Five isolates were screened for aminoglycosides resistance genes; aacA4, aacC2, and aadA1 were detected. Pulsed-field gel electrophoresis-based subtyping identified 91 different patterns. Our results indicate that food, especially fresh raw chicken, is a reservoir of antimicrobial-resistant K. pneumoniae, and the potential health risks posed by such strains should not be underestimated. Our results demonstrated high prevalence, antibiotic resistance rate and genetic diversity of K. pneumoniae in food in China. Improved control and prevention strategies are urgently needed.


Introduction
Klebsiella pneumoniae is a common opportunistic pathogen that causes human infections. It can be widely distributed not only in the respiratory and intestinal tracts of humans and animals but also in a variety of environments and vectors. This pathogen can cause pneumonia, respiratory tract infections, urinary system infections, septicemia and other diseases [1,2]. Antimicrobials have been widely used to treat K. pneumoniae infections in humans. However, increasing antimicrobial resistance, especially that mediated by extended-spectrum β-lactamases (ESBL), plasmid-borne AmpCs, and carbapenemases, has been reported in recent years and has become a serious problem [3][4][5].
Foodborne diseases caused by pathogenic bacteria constitute a serious threat to public health worldwide [6,7]. Until now, most investigations on foodborne bacteria focused on common foodborne pathogens, such as Salmonella, Campylobacter, Escherichia coli, Shigella, Listeria monocytogenes, Staphylococcus aureus, and Vibrio parahaemolyticus. In contrast, little information was obtained on foodborne K. pneumoniae as K. pneumoniae is generally not recognized as a foodborne pathogen. However, antimicrobial-resistant K. pneumoniae strains have been isolated from marketed fresh vegetables [8], shrimp in international trade [9], and farm-raised chicken [10]. A recent report showed that foodborne K. pneumoniae could cause a nosocomial outbreak [11]. Furthermore, several resistance genes in K. pneumoniae are located in transferable genetic elements that may be transferred to other bacteria. Thus, the potential contribution of K. pneumoniae to the resistance of clinically relevant bacteria is cause for concern.
The presence of antimicrobial-resistant K. pneumoniae strains in the food supply is alarming. Our objective was to assess the frequency of K. pneumoniae in food samples. We focused on the contamination rate in foods and on the characteristics of K. pneumoniae isolates. We characterized their antimicrobial resistance phenotypes, identified their antimicrobial resistance genes and analyzed their molecular subtypes.

Sample collection
A total of 998 food samples were collected in Shijiazhuang, a city of approximately 10 million inhabitants in eastern China, between April 2013 and July 2014. Those samples were used to isolate K. pneumoniae strains. The samples included 49 fresh raw seafood (fish, shellfish, shrimp) samples, 188 fresh raw chicken samples, 297 frozen raw food (meat, vegetables, flour and rice products) samples and 464 cooked food samples (meat, vegetables, flour and rice products). These samples were collected from different farms, supermarkets and restaurants distributed throughout the city. None of the samples were duplicated.

Isolation and identification
A 25-g portion of each sample was suspended in 225 mL of buffered peptone water (BPW). The sample suspensions were incubated overnight at 36°C. A 1-mL aliquot of the pre-enrichment culture was added to 10 mL of selenite cystine broth (SC) and incubated overnight at 36°C. A loopful (10 μL) of SC was streaked directly onto Salmonella Shigella (SS) agar plates and incubated for 24 h at 36°C. Colorless, medium-sized, smooth and moist colonies were transferred to triple-sugar iron (TSI) agar plates. All the suspected K. pneumoniae isolates were identified using a BD Phoenix™-100 Automated Microbiology System (Becton, Dickinson and Company, Sparks, Maryland, USA).
A standardized international definition was used to define multidrug-resistant (MDR) bacteria [13]. MDR was defined as acquired non-susceptibility to at least 1 agent in 3 or more antimicrobial categories.

PCR amplification and sequencing
Genomic DNA was extracted using a QIAamp DNA minikit (Qiagen, Dusseldorf, Germany) or prepared by the boiling method. Antimicrobial resistance-associated genes were detected by PCR and sequenced using the primers listed in Table 1. The PCR was performed in a 50-μL reaction volume that contained 25 μL of Premix Taq TM (Takara, Dalian, China), 10 μM of each primer and 1 μL of sample DNA. The PCR conditions for the β-lactamase genes consisted of an initial denaturation at 95°C for 5 min, 35 cycles of denaturation at 95°C for 50 s, annealing at 56°C, 50°C or 60°C for 40 s and elongation at 72°C for 1 min, followed by a final extension at 72°C for 5 min, in a thermocycler (Labcycler, Senso, Germany). The PCR conditions for other resistance genes consisted of an initial denaturation at 95°C for 5 min, 30 cycles of denaturation at 94°C for 1 min, annealing at 55°C for 1 min and elongation at 72°C for 2 min, followed by a final extension at 72°C for 5 min, in a thermocycler. The PCR products were detected in a 1% agarose gel. Positive amplicons were sequenced on a PE Applied Biosystems ABI Prism 3730 instrument. The DNA sequences were annotated using the BLAST program (http://blast.ncbi.nlm.nih.gov) to identify the gene subtypes. Mutations in the gyrA and parC sequences of K. pneumoniae (reference GenBank accession numbers DQ673325 and NC009648 for gyrA and parC, respectively) were detected.

Pulsed-field gel electrophoresis (PFGE)
We used the 1-day, standardized PFGE protocol for K. pneumoniae [35]. Cell suspensions were placed in polystyrene tubes (Falcon; 12 × 75 mm), and their optical densities were adjusted to 3.8-4.0 using a Densimat photometer (BioMérieux, Marcy l'Etoile, France). Slices of K. pneumoniae agarose plugs were digested using 50 U of XbaI (Takara) per slice for 4 h at 37°C, and electrophoresis was performed using a CHEF-DRIII system (Bio-Rad Laboratories, Hercules, CA, USA). Electrophoresis was conducted with a switch time of 6 s to 36 s for 18.5 h, and images were captured using a Gel Doc 2000 system (Bio-Rad) and converted to TIFF files. The TIFF files were analyzed using BioNumerics version 5.1 software (Applied Maths, Kortrijk, Belgium). A similarity analysis of the PFGE patterns was performed by calculating the Dice coefficients (SD) [36] and clustering was performed using the unweighted-pair group method with average linkages (UPGMA).

Statistical analysis
SPSS software (version 15.0) was used to statistically analyze the data. Categorical variables were compared using the Fisher's exact test. A P value <0.05 was considered to be statistically significant. Antimicrobial susceptibility patterns of the K. pneumoniae isolates Antimicrobial susceptibility testing was conducted for the 99 K. pneumoniae isolates, and detailed information on the resistance rates to all of the tested antimicrobials is listed in Table 2. The highest resistance rate was observed for AMP, which reached 92.3% (n = 92), followed by resistance to TE (n = 31; 31.3%), SXT (n = 18; 18.2%), C (n = 10; 10.1%), and 12 other antimicrobials with resistance rates under 10.0%. There was no resistance noted to carbapenems (IPM, MEM). Notably, the resistances to 7 antimicrobials (GEN, CTX, FEP, ATM, SAM, CIP, or LVX) were detected only among fresh raw chicken isolates. Furthermore, the rate of resistance to 5 antimicrobials (CZO, PRL, SXT, C and TE) in fresh raw chicken isolates was significantly higher than in isolates from other types of samples (P<0.05). Two K. pneumoniae strains were detected as ESBL-producing; both were from fresh raw chicken samples. Nineteen MDR strains were detected among 99 K. pneumoniae isolates. The proportions of MDR strains in different samples were 50.0% (14/28), 11.4% (4/35), 2.9% (1/34) and 0% (0/4), in fresh raw chicken, cooked food samples, frozen raw food and fresh raw seafood, respectively. The proportion of MDR isolates from fresh raw chicken samples was significantly higher than that from other types of samples (P<0.05).
Antimicrobial resistance determinants of the K. pneumoniae isolates According to the results of antimicrobial susceptibility testing, 2, 16, 4 and 5 strains were selected to analyze ESBL genes, folate pathway inhibitor genes, fluoroquinolone resistance genes and aminoglycoside resistance genes, respectively. For the 2 ESBL strains, 8 β-lactamase genes and 5 AmpC genes were amplified. As shown in Table 3, 1 strain carried bla SHV , bla CTX-M-1 and bla CTX-M-10 , and the other carried bla SHV . No bla CTX-M-9 , bla CTX-M-14 , bla DHA , bla TEM , ba CMY , bla ACT , or bla FOX genes were detected in these isolates.
Eighteen isolates that showed trimethoprim-sulfamethoxazole resistance were selected for folate pathway inhibitor gene (dhfr) testing; 6 of the isolates were positive for dhfr. All of the 6 dhfr-positive isolates were isolated from fresh raw chicken, whereas no isolates from frozen raw food or cooked food samples tested were positive for dhfr.

PFGE and MLST analysis of K. pneumoniae isolates
All the 99 isolates of K. pneumoniae were analyzed by PFGE, and 91 different PFGE patterns were obtained, with similarity values of 47.1% (Fig 1). Eighty-five (85.9%) isolates showed unique PFGE patterns. No dominant pattern was identified among these isolates. Only 6 patterns included more than 1 isolate. Two isolates of KPX01.CN0357 were isolated from fresh raw chicken samples collected from the same market at the same time; 3 isolates of KPX01. CN0404 were isolated from cooked food samples collected from the same restaurant at the same time; 2 isolates of KPX01.CN0365 were isolated from cooked food samples collected from different restaurants at the same time; and 2 isolates of KPX01.CN0355 were isolated from fresh raw chicken samples collected from different markets at the same time. The 2 ESBL-producing isolates showed different PFGE patterns. An MLST analysis of these 2 isolates produced two unique STs (ST1651, ST1652; Table 3); neither has been previously reported.

Discussion
Foodborne bacteria are widely studied, but research on K. pneumoniae is scarce. In a previous survey conducted in the United States, 53 (16.1%) MDR K. pneumoniae strains were isolated from 330 farm-raised frozen shrimp that were imported from Thailand to the United States [9]. In another survey focusing on fresh vegetables in Spain, 9 K. pneumoniae strains were obtained from 160 vegetables, among which 1 (0.6%) was an MDR strain [8]. The major goal of this study was to evaluate the current frequency and antimicrobial resistance of K. pneumoniae strains in fresh raw seafood, fresh raw chicken, frozen raw food and cooked food samples in China. The incidence of K. pneumoniae in these samples was 9.9%, which shows that contamination of food with K. pneumoniae is common in this region of China. Furthermore, 19 (1.9%) MDR K. pneumoniae strains were isolated in this study, representing a lower percentage than that reported by Nawaz et al. [9] but a higher percentage than that reported by Falomir et al. [8].
The highest isolation rates for K. pneumoniae and MDR K. pneumoniae strains in this study were observed for fresh raw chicken samples. Fresh raw chicken is an important reservoir of antimicrobial-resistant K. pneumoniae. A recent study focused on retail raw chicken demonstrated that β-lactamases and ESBLs were emerging and prevalent in foodborne Salmonella in China [38]. Profitable chicken farms demand the extensive usage of antimicrobials to inhibit  infectious diseases. However, the use of antimicrobials in these ecosystems may select for antimicrobial-resistant microorganisms. Hence, chicken meat may be a reservoir of antimicrobialresistant bacteria such as K. pneumoniae, which constitutes a public health concern. The resistance mechanisms of K. pneumoniae include the production of β-lactamases (including ESBLs and plasmid-mediated AmpCs) and carbapenemases, the production of biological membrane formation factors, the loss of outer membrane proteins, and antimicrobial efflux [39][40][41]. In this study, we investigated a total of 99 K. pneumoniae isolates from food; 4 strains were ESBL, and 1 strain produced an AmpC beta-lactamase. ESBL-producing K. pneumoniae strains have been shown to have a significant impact on the treatment options and clinical outcomes of patients. Likewise, they have been shown to cause higher morbidity and mortality [42][43][44]. Currently, ESBLs and AmpCs are the predominant β-lactamases that mediate Gram-negative bacterial resistance to new broad-spectrum β-lactam antimicrobials. ESBLs are mainly encoded by plasmids, whereas AmpCs are mainly encoded on the chromosome. The CTX-M type is the major phenotype of domestic ESBLs; it reportedly predominates worldwide, followed by the SHV type [45][46][47]. Recently, Enterobacteriaceae carrying bla CTX-M -type genes were isolated from chicken in several countries [38,[48][49][50]. In this study, both ESBLproducing K. pneumoniae strains carried the bla SHV gene, and one strain carried bla CTX-M genes. One strain isolated in this study carried at least 4 ESBL-associated genes, i.e., coexisting bla CTX-M-1 , bla CTX-M-10 , and bla SHV . The coexistence of multiple bla CTX-M -type genes in K. pneumoniae isolates was also reported in previous studies [51,52]. The detection of ESBL-producing strains and the coexistence of several ESBL-associated genes in the same isolates pose a serious epidemiological, clinical and public health threat.
Quinolones are broad-spectrum antimicrobial agents that have been widely used in clinical medicine and for raising food-producing animals (such as chicken in China). The isolation and characterization of ciprofloxacin-and levofloxacin-resistant K. pneumoniae from fresh raw chicken samples is corresponding to that fluoroquinolones have been used in chicken farms. Among the 7 plasmid-encoded fluoroquinolone resistance-associated genes, the aac(6')-Ib-cr enzyme, qnrB, qnrA, and qnrS were the most prevalent plasmid-mediated mechanisms of quinolone resistance, as previously reported [29]. Several studies have 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 [30,53]. In contrast, one study reported that hypermutation in K. pneumoniae is uncommon and does not contribute to the accumulation of gyrA mutations or directly to ciprofloxacin resistance [54]. Sequence analysis of the gyrA gene in K. pneumoniae isolates from fresh raw chicken in this  study identified 3 types of gyrA mutations (encoding Ser83Ile, Ser83Phe, and Asp87Ala substitutions) and 1 type of parC mutation (encoding Ser80Ile). These point mutations were previously reported in clinical K. pneumoniae cases and may be responsible for mediating resistance In the column of "forms of drug-resistant", "No" means not resistant to all drugs, "R" means non-susceptibility to at least 1 agent in one or two antimicrobial categories, "MDR" means non-susceptibility to at least 1 agent in 3 or more antimicrobial categories. to fluoroquinolones [9,51,53,55]. The amino acid substitutions at positions 83 (Ser to Phe) and 80 (Ser to Ile) in gyrase A resemble a substitution that confers fluoroquinolone resistance in Salmonella spp. [56,57]. Among the aminoglycosides resistance-associated genes, the aacA4, aacC2, and aadA1 genes were detected in this study. These 3 genes were the major aminoglycosides resistance genes among clinical K. pneumoniae cases reported at a hospital in China [51], which suggests that the resistance genes in clinical strains may come from foodborne strains. PFGE is a useful tool to reveal genotypic characteristics and to trace the reservoirs of infectious pathogens, the rates of transmission and the mechanisms of infectious diseases. The variety of different PFGE strain patterns in this study was unexpected. Except for 6 groups of isolates that had identical PFGE patterns, all other isolates showed unique PFGE patterns. Furthermore, the 2 ESBL-producing isolates showed different PFGE patterns and MLST types. These results reflected a high genetic diversity of foodborne K. pneumoniae isolates.
There were three limitations of this study. Firstly, this study had not included agricultural antimicrobial use data from the regions which supplied food to the farms, supermarkets, and restaurants from where we sampled food items. The second one was that not compared these foodborne K. pneumoniae isolates to clinical isolates from the same region. Furthermore, the use of agar SC and SS would underestimate the frequency of K. pneumoniae isolated, because these two seletive medium are designed to inhibit other microorganisms than Salmonella spp. and Shigella spp.
In conclusion, our results indicate that food, especially fresh raw chicken, is a reservoir of antimicrobial-resistant K. pneumoniae. They may have the potential to become a public health risk. Thus, our study demonstrates that improved monitoring and prevention strategies are urgently needed to better control the emergence and transmission of antimicrobial-resistant K. pneumoniae isolates.