Genotypes, Enterotoxin Gene Profiles, and Antimicrobial Resistance of Staphylococcus aureus Associated with Foodborne Outbreaks in Hangzhou, China

Staphylococcal food poisoning is an illness caused by the consumption of food that contains sufficient amounts of one or more enterotoxins. In the present study, a total of 37 S. aureus isolates were recovered from leftover food, swabs from a kitchen environment, and patient feces associated with four foodborne outbreaks that occurred in Hangzhou, southeast China, and were characterized by multilocus sequence typing (MLST), spa typing, pulse-field gel electrophoresis (PFGE), and antimicrobial susceptibility. Classical enterotoxin and enterotoxin-like genes were profiled by PCR analysis. ST6-t304 was the most common clone (40.54%), followed by ST2315-t11687 (32.43%). Six clusters (A to F) were divided based on PFGE patterns, and Clusters A and C were the most common types, constituting 86.49% of all isolates. Moreover, sea was the most frequently identified enterotoxin gene (81.08%), followed by the combination of seg–sei–selm–seln–sleo–selu and sec–sell (each 48.65%). Five isolates also harbored the exotoxin cluster sed–selj–ser. In addition, resistance to penicillin (97.30%), erythromycin (37.85), tetracycline (32.43%), clindamycin, gentamicin, and sulfamethoxazole (each 10.81%) was observed. Our research demonstrated the link between leftover foods and patients by molecular typing and detecting the profiles of enterotoxin or enterotoxin-like genes in human and food isolates. S. aureus maintains an extensive repertoire of enterotoxins and drug resistance genes that could cause potential health threats to consumers.


Introduction
Staphylococcus aureus is a well-known bacterial pathogen that causes a wide range of clinical infections from superficial skin and soft tissue infections to life-threatening septicemia [1]. S. aureus is also a significant cause of foodborne outbreaks, leading to an estimated 241,000 illnesses every year in the United States [2]. Staphylococcal food poisoning (SFP) was the third most frequent bacterial etiology of outbreaks in China after Vibro parahaemolyticus and Salmonella from 2011 to 2016 [3]. A total of 314 outbreaks of SFP were reported in China, involving 5196 illnesses and leading to one death between these six years. However, the true incidence of SFP could be underestimated due to the nontypical symptoms [4]. The prevalence of this disease makes its control important.
SFP is associated with enterotoxins (SEs) that are synthesized by S. aureus during food preparation and processing. SEs are highly stable and highly heat resistant. They are also resistant to proteolytic enzymes and low pH, enabling them to be fully functional in the gastrointestinal tract [5]. So far, 23 distinct SEs have been designated based on their antigenicity (SEA to SElY) [6]. Among them, five classical SEs (SEA, SEB, SEC, SED, and SEE) and more recently described nonclassical SEs, including SEG, SHE, SEI, SER, SES, and SET, exhibit emetic activity [7]. Related toxins that lack emetic activity or have not been tested for emetic activity are designated similarly as staphylococcal enterotoxin-like toxins (SEls) [8,9].
The emergence of methicillin-resistant S. aureus (MRSA) from livestock and companion animals [10,11] and the subsequent detection of enterotoxins produced by those isolates highlight the necessity of monitoring S. aureus clones from foodborne disease [12][13][14][15][16][17]. The aim of this study was to characterize S. aureus isolated from foodborne outbreaks in Hangzhou, southeast China, by testing a capacity for carrying genes encoding SEs, evaluating resistance properties, and determining pulse-field gel electrophoresis (PFGE), MLST, and spa typing.

Phenotypic Characterization and Identification of S. aureus Isolates
S. aureus isolates were isolated from various samples associated with four foodborne outbreaks: 14 fecal samples, 22 leftover items, and one cotton swab of a kitchen environment. The amount of S. aureus in food remnants was at least 10 5 CFU/g. Preliminary estimation was carried out by culture properties on a BP plate. Hemolytic activities were determined on sheep blood agar. Furthermore, a 278 bp (nuc gene) band detectable in a 1.5% agarose gel was observed in the PCR products of all 37 S. aureus isolates.

Antibiotic Resistance
All 37 stains from the food poisoning samples were resistant to one or more kinds of antibiotic. Penicillin (PEN) resistance was the most commonly observed resistance among the tested isolates except for one, followed by erythromycin resistance (37.84%). A total of 29.73% of the isolates (11/37) were only resistant to penicillin, followed by the combination of erythromycin-clindamycin-gentamicin-sulfamethoxazole-penicillin (4/37, 10.81%).

Molecular Genotyping of S. aureus
To achieve more reliable genotyping results and enhance the convenience of comparing different studies, we performed MLST, spa typing, and PFGE on the 37 isolates. The allelic profile of seven housekeeping genes revealed eight sequence types (STs), namely, ST5, ST6, ST72, ST188, ST573, ST2250, ST2315, and ST3055 (Table 1). ST6 was the most common ST and was identified in 15 (40.54%) of the isolates from 2 individual outbreaks. Twelve isolates (32.43%) from only 1 outbreak were identified as ST2315. Spa typing of all of the isolates yielded nine kinds ( Table 1). The top two were t304 and t11687, the corresponding partners of ST6 and ST2315, respectively. It is interesting that isolates with the same ST (ST5) could be divided into two different spa types. The 37 isolates were also typed by PFGE ( Figure 1). Using an 80% similarity cut-off, the strains were grouped into six clusters, designated by the letters A to F. Clusters A and C were the dominant types, constituting 86.49% of all isolates. Two individual strains in Outbreak 1 exhibited the same ST and spa type (ST188-t189) but presented different PFGE types (D1 and D2) with 85.7% similarity. ST6 was also identified in Outbreaks 3 and 4, with the same spa type (t304) and almost 100% similarity to the PFGE type. All ST5 strains were classified as Cluster A but with two different spa types (t1228 and t002).   (14), and swabs from the environment (1). b The 27 foodborne isolates were assigned to 12 strains c PEN, penicillin; ERY, erythromycin; TE, tetracycline; CD, clindamycin; GEN, gentamicin; SXT, sulfamethoxazole.

Prevalence of Enterotoxin or Enterotoxin-Like Genes
As shown in Table 2, the majority of the isolates in this study contained sea gene (81.08%); fewer isolates carried the combination of seg-sei-selm-seln-selo-selu and sec-sell (each 48.65%). Only five isolates harbored the exotoxin cluster sed-selj-ser. None of the isolates tested positive for the see, selk, selp, and selq genes. The number of enterotoxin genes ranged from 1 to 11.

Discussion
SFP is one of the most prevalent causes of foodborne illness worldwide. Genetic information of those isolates typically differs between countries and even between cities in the same country. In the current study, 37 foodborne-related S. aureus isolates from Hangzhou, China, were assigned to 12

Prevalence of Enterotoxin or Enterotoxin-Like Genes
As shown in Table 2, the majority of the isolates in this study contained sea gene (81.08%); fewer isolates carried the combination of seg-sei-selm-seln-selo-selu and sec-sell (each 48.65%). Only five isolates harbored the exotoxin cluster sed-selj-ser. None of the isolates tested positive for the see, selk, selp, and selq genes. The number of enterotoxin genes ranged from 1 to 11.

Discussion
SFP is one of the most prevalent causes of foodborne illness worldwide. Genetic information of those isolates typically differs between countries and even between cities in the same country. In the current study, 37 foodborne-related S. aureus isolates from Hangzhou, China, were assigned to 12 strains by molecular typing techniques. MLST provided a comparison of the nucleotide sequences of seven housekeeping genes. MLST has been widely used for epidemic analyses of S. aureus related to hospital infections and foodborne diseases since the first report of MLST use for S. aureus [18]. In South Korea, ST1, ST59, and ST30 strains were the clones that were most frequently associated with SFP [19]. ST45 and ST81 were more closely related to SFP outbreaks in Japan [20][21][22]. ST45 and ST5 have been identified in SFP outbreaks in Europe [23,24]. In this study, S. aureus ST6-t304 was the predominant lineage isolated from two separate outbreaks. Previous SFP reports occurred in Shenzhen, Ma'anshan, and Xi'an, and it was demonstrated that ST6-t701 was the most common type in China. It is interesting that ST6-t304 is the ancestor of ST6-t701, which had one spa repeat (r25) based on BURP analysis results [25,26]. In addition, another two dominant lineages in our study, namely, ST188-t189 and ST5-t002, were reported in other SFP outbreaks and food samples in China [25][26][27][28], suggesting that additional studies are needed to confirm the temporal relationship of these strains.
In this study, PFGE was used in the analysis of the molecular epidemiological diversity of the strains, especially in relation to the analysis of short-term epidemiology due to high discrimination. Based on PFGE analysis, all outbreaks were caused by multiple clones (A to F), which was identical to the results from the MLST methods and spa typing. In Outbreaks 1 and 3, two similar PFGE patterns (C1 and C2) were detected with a similarity of 90.3%, whereas the spa and MLST types remained indistinguishable. According to the criteria proposed by Tenover et al., a PFGE pattern with a 4-6-band difference can be regarded as possibly related to outbreak strains [29]. Four obvious bands were detected between Strains 1 and 10. A period of approximately 10 months may be the reason for simple insertions or deletions of DNA. Outbreaks 3 and 4 may have been caused by the same epidemiological strain, which was supported by the fact that Strains 9 and 11 had the same MLST, spa types, and PFGE patterns. Although these two outbreaks occurred in two different districts, the period between the outbreaks was less than two months. The ST188-t189-D1 clone had the same ST and spa type as the ST188-t189-D2 clone, while several bands were detectable between these two strains, indicating some large-scale changes in the accessory genome. This situation was also found in the isolates ST5-t002-A1 and ST5-t002-A2. In Outbreak 2, there were two spa types (t002 and t1228) with the same PFGE pattern (A1). According to the spa repeat analysis, there was only one repeat difference between these two clones, suggesting that these strains may belong to the same clone and differed only due to the recombination of select genes [25]. Based on the above comparison, PFGE is proposed to be more effective in describing the strain population.
SFP is associated with one or more families of genes that code for heat-stable enterotoxins. This study demonstrated that the sea gene was the dominant enterotoxin (81.08%) in S. aureus strains recovered in this study. Additionally, sea was described to be the most prevalent gene recovered from not only clinical samples but also food-and foodborne-related samples in other studies [25,26,30,31]. Yan et al. reported that SFP caused by sea alone may be unique in Shenzhen. Fortunately, Strains 9 and 11 in our study also harbored only sea genes. The ST188-t189-D2 clone only had the seb gene, which was exactly similar to the strain from the studies conducted by Yan et al. and Song et al. [25,27] in comparison with MLST, spa typing, and PFGE bands. Nearly all se and sel genes are associated with mobile genetic elements, including plasmids, prophages, S. aureus pathogenicity island (SaPI), and genomic island νSa or next to the staphylococcal cassette chromosome (SCC) [23]. In our study, the combination of seg, sei, selm, seln, selo, and selu was detected at a relatively high frequency, forming an operon encoded by enterotoxin gene cluster 2 (egc2) in the vSaβ genomic island inserted at specific loci in the chromosome [23]. Genotype sec-sell and sed-selj-selr could be a combination of Type II vSa3 and pIB485, respectively, which was also observed in this study. Strains 2 and 4 or Strains 7 and 8 harbored the same MLST and spa type and almost the same PFGE bands, but the distribution of se/sel genes was largely distinct. This finding suggested that toxin gene transfer between S. aureus plays a considerable role in genetic diversity [7].
The majority of the isolates in this study were resistant to one or two antibiotics. The high rate of resistance to penicillin was in agreement with previous studies of S. aureus isolates from food samples or foodborne outbreaks in both China and other countries [25,26,32,33]. There was only one base difference in spa sequence among ST5-t002 and ST5-t1228, which were found to be multidrug-resistant to five antibiotics. ST5-t002 in the studies conducted by Li et al. and Argudin et al. was also identified as resistant to at least three drugs [25,26], indicating that a deep investigation may be needed to explain the relationship between genomic characterization and multidrug resistance.

Conclusions
Our study demonstrated the SE/SEl profiles and genetic diversity of S. aureus isolated from foodborne outbreaks in Hangzhou, China. The detection of not only classical SEs but also new SE/SEls should be considered for the diagnostic analysis of SFP outbreaks due to the frequency of their observation in our study. Molecular genotyping methods, including MLST, PFGE, and spa typing, were useful for investigating epidemiological relatedness and even contaminated food sources, especially for outbreaks occurring in the same city.

Bacterial Isolates
Isolates were collected from the fecal specimens of individual patients, swabs from the environment, and food samples associated with four foodborne outbreaks in Hangzhou, Zhengjiang Province, China, during 2015-2016. A total of 37 S. aureus isolates were identified according to the methods described in GB 4789.10-2010 Food Microbiological Examination: S. aureus. In general, test isolates were cultured in tryptic soy broth, Baird-Parker agar, BHI, and tryptic soy agar (TSA). The suspected single clones were then tested by Gram staining, coagulase testing, hemolytic activity (determined on sheep blood agar), and catalytic reactions using the Vitek 2 compact system. Furthermore, all isolates were investigated for the presence of the nuc gene (S. aureus species specific). Bacterial stocks of each isolate were maintained at −80 • C in tryptic soy broth (TSB) containing 20% glycerol (v/v). All the isolates were thawed and subcultured in TSA for 18-24 h prior to use.

Extraction of Genomic DNA
S. aureus isolates were grown overnight at 37 • C in BHI broth. Genomic DNA was extracted using a bacterial genomic DNA extraction kit (DNeasy Blood and Tissue Kit, Qiagen Inc., Redwood City, CA, USA) according to the manufacturer's instructions.

Detection of 23 SE/SEl Genes
The primer sets used to detect the sea, seb, sec, sed, and see genes were those described by Becker et al. [34], seg, seh, sei, selj, selk, selm, seln, and selu were those described by Tang et al. [31], selo, selp, selq, sell, and ser were those described by Omoe et al. [35]. PCR was performed using a commercial PCR kit (Takara, Kusatsu, Japan) in a 25 µL volume containing 2.5 µL of buffer, 2 µL of dNTP (2.5 mM), 0.125 µL of Taq DNA polymerase (1 U/µL), 0.5 µL of each primer set (10 µM), and 1 µL of template DNA. PCR was performed on a Mycyler thermal cycler (Bio-Rad, Hercules, CA, USA) using the following steps: initial denaturation at 95 • C for 5 min, 30 cycles of 95 • C for 30 s, each annealing temperature for 30 s and 72 • C for 45 s, and a final extension at 72 • C for 5 min. The PCR products were separated with electrophoresis in 1.5% agarose gel to verify the expected size of the amplicons.

PFGE
PFGE of the S. aureus isolates was performed in a CHEF Mapper system (Bio-Rad Laboratories, Hercules, CA, USA) as described previously [36] with some modifications. In this study, DNA from Salmonella choleraesuis serotype Branderup H9812 digested with XbaI (New England Biolabs Inc., Ipswich, MA, USA) was included as a molecular size marker. The banding patterns were analyzed with Denmark BioNumerics, version 6.6 (Applied Maths, Sint-Martens-Laterm, Belgium) with a 1% optimization and a band-matching tolerance of 1%.

MLST and Spa Typing
Genomic DNA, extracted as indicated above, was used in PCR amplifications for the spa type and seven housekeeping genes (arcC, aroE, glpF, gmk, pta, tpi, and yqiL) listed on the S. aureus MLST website [37]. The allelic profile of an S. aureus isolate appearing as an array of seven allele numbers was achieved by sequencing (Shangong, Shanghai, China). The Ridom Spa Server [38] and the MLST website [37] were used to assign spa types and sequence types (ST), respectively.