Food poisoning outbreak in Tokyo, Japan caused by Staphylococcus argenteus

International Journal of Food Microbiology 262 (2017) 31–37

Contents lists available at ScienceDirect

International Journal of Food Microbiology journal homepage: www.elsevier.com/locate/ijfoodmicro

Food poisoning outbreak in Tokyo, Japan caused by Staphylococcus argenteus a,⁎

a

b

a

MARK

a

Yasunori Suzuki , Hiroaki Kubota , Hisaya K. Ono , Makiko Kobayashi , Konomi Murauchi , Rei Katoa, Akihiko Hiraia, Kenji Sadamasua a b

Department of Microbiology, Tokyo Metropolitan Institute of Public Health, 3-24-1 Hyakunin-cho, Shinjuku-ku, Tokyo 169-0073, Japan Laboratory of Zoonosis, Kitasato University School of Veterinary Medicine, 23-35-1 Higashi, Towada City, Aomori 034-8628, Japan

A R T I C L E I N F O

A B S T R A C T

Keywords: Accessory element Phylogenetic analysis Staphylococcal enterotoxin Whole genome sequencing

Staphylococcus argenteus is a novel species subdivided from Staphylococcus aureus. Whether this species can cause food poisoning outbreaks is unknown. This study aimed to investigate the enterotoxigenic activities of two food poisoning isolates suspected to be S. argenteus (Tokyo13064 and Tokyo13069). The results for phylogenic trees, constructed via whole genome sequencing, demonstrated that both isolates were more similar to a type strain of S. argenteus (MSHR1132) than any S. aureus strain. Moreover, the representative characteristics of S. argenteus were present in both strains, namely both isolates belong to the CC75 lineage and both lack a crtOPQMN operon. Thus, both were determined to be “S. argenteus.” The compositions of the two isolates' accessory elements differed from those of MSHR1132. For example, the seb-related Staphylococcus aureus pathogenicity island, SaPIishikawa11, was detected in Tokyo13064 and Tokyo13069 but not in MSHR1132. Both isolates were suggested to belong to distinct lineages that branched off from MSHR1132 lineages in terms of accessory elements. Tokyo13064 and Tokyo13069 expressed high levels of s(arg)eb and produced S(arg)EB protein, indicating that both have the ability to cause food poisoning. Our findings suggest that S. argenteus harboring particular accessory elements can cause staphylococcal diseases such as food poisoning, similarly to S. aureus.

1. Introduction Staphylococcus argenteus is a new species that was subdivided from Staphylococcus aureus in 2014 (Tong et al., 2015). The species epithet of S. argenteus reflects the color of its colony (silver) against S. aureus (golden) because this species is typified by the absence of production of the carotenoid pigment staphyloxanthin, which is encoded by the crtOPQMN operon (Holt et al., 2011; Xiong et al., 2015). Although some of the earliest reports mentioned that S. argenteus was linked to remote Aboriginal communities (Tong et al., 2010; Tong et al., 2013), it is currently recognized as having a global distribution between humans and animals (Dupieux et al., 2015; Jenney et al., 2014; Long et al., 2014; Thaipadungpanit et al., 2015; Tong et al., 2015). S. argenteus appears to be less virulent than typical S. aureus because it has a greater association with minor skin infections and a lesser association with serious and systemic infections such as sepsis (Holt et al., 2011). Moreover, it has not been reported that S. argenteus causes food poisoning, which is one of the representative staphylococcal diseases. Multilocus sequence typing (MLST) provides data that are well

suited for epidemiological and phylogenic studies of bacteria including staphylococci (Enright et al., 2000). S. argenteus was previously known as “S. aureus clonal complex (CC) 75” (Tong et al., 2013; Xiong et al., 2015), and the CC75 lineage (including ST75, ST850, ST883, and ST1223) is extremely distant from all other clones (Chantratita et al., 2016; Holt et al., 2011). MLST thus makes a clear distinction between S. aureus and S. argenteus. On the contrary, the original primers for MLST analysis of S. aureus cannot amplify several of the seven loci in S. argenteus (Ng et al., 2009); therefore, it is difficult to detect this species as a causative microbe using clinical specimens. Staphylococcal strains Tokyo13064 and Tokyo13069 were isolated from the feces of a patient and food and identified as the causative microbe in the same food poisoning outbreak in Tokyo, Japan in 2010. These strains formed white colonies opposed to yellow colonies (Fig. S1). The sequence types (STs) of these strains were not determined because aroE (shikimate dehydrogenase), which is one of the seven loci, was not amplified using the conventional MLST method (Fig. S2). Hence, it was suspected that these strains were S. argenteus and that they caused the food poisoning outbreak.

Abbreviations: ANOVA, Analysis of variance; BHI, Brain heart infusion; BLAST, Basic Local Alignment Search Tool; CC, Clonal complex; CDS, Coding sequence; ELISA, Enzyme-linked immunosorbent assay; MLST, Multilocus sequence typing; NGS, Next-generation sequencing; PCR, Polymerase chain reaction; SaPI, Staphylococcus aureus pathogenicity island; SE, Staphylococcal enterotoxin; SEB, Staphylococcal enterotoxin B; SFP, Staphylococcal food poisoning; SNP, Single nucleotide polymorphism; ST, Sequence type ⁎ Corresponding author. E-mail address: [email protected] (Y. Suzuki). http://dx.doi.org/10.1016/j.ijfoodmicro.2017.09.005 Received 6 February 2017; Received in revised form 8 June 2017; Accepted 10 September 2017 Available online 15 September 2017 0168-1605/ © 2017 Elsevier B.V. All rights reserved.

International Journal of Food Microbiology 262 (2017) 31–37

Y. Suzuki et al.

electrophoresis.

In the present study, we conducted whole-genome analysis of the aforementioned two strains using next-generation sequencing (NGS) and compared their genomic structures and epidemiological properties to those of reported staphylococcal strains. Furthermore, we investigated whether these strains possessed the ability to cause staphylococcal food poisoning (SFP).

2.4. Construction of plasmids Plasmids and the primers for construction are presented in Tables S2 and S3. The plasmids pGTgyrB, pGTsodA, pGTftsZ, pGTrpoB, pGTseb, pGTsei, pGTsem, pGTsen, pGTseo, and pGTselu were constructed by cloning the individual polymerase chain reaction (PCR) products, which were amplified from the previously extracted gDNA, into pGEMT easy TA cloning vectors (Promega). E. coli DH5α was transformed with the cloning plasmids. Plasmids were extracted from the cultured cells using a QIAprep Spin Miniprep kit (QIAGEN) according to the manufacturer's instructions.

2. Materials and methods 2.1. Overview of the staphylococcal food poisoning outbreak that occurred in Tokyo, in 2010 In August 2010, two patients complained of vomiting, fever, and diarrhea 3 h after ingesting a box lunch prepared by a caterer at an industrial kitchen in Tokyo. Several other patients who ate the same box lunch also complained of vomiting, fever, and diarrhea. We examined fecal specimens from patients and caterer workers, and also examined suspected foods and swabs of the kitchen utensils and tableware in the caterer. We isolated 10 staphylococcal strains from the following specimens: fecal specimens from three patients and two workers, one suspected delicatessen food (contained eggplant, minced meat, and cheese), and four swabs each from a cold table, workbench, and empty lunch boxes. Bacterial identification was performed using a coagulase test from the staphylocoagulase antiserum kit (Denka Seiken, Tokyo, Japan). The detection and typing of the staphylococcal enterotoxin (SE) was performed by reversed passive latex agglutination for types A–E (Denka Seiken) according to the manufacturer's instructions. All 10 strains carried type VI coagulase and staphylococcal enterotoxin B (SEB). However, the colonies grown from the samples were not typical for S. aureus, because they formed white colonies instead of yellow colonies of S. aureus (Fig. S1). The investigation conducted by the health center, an administrative organization, demonstrated that the delicatessen cooked the food the day before the patients ate it and packed it in a lunch box without sufficient cooling; therefore, the food was left at a high temperature (at a temperature higher than room temperature in summer in Tokyo) for at least 12 h.

2.5. Genomic library preparation and whole genome sequencing The whole genomes of Tokyo13064 and Tokyo13069 were sequenced. For each gDNA, index-tagged libraries were prepared using a Nextera XT DNA Library Preparation Kit (Illumina, San Diego, CA, USA) and sequenced using Illumina MiSeq with 300-bp paired-end reads according to the manufacturer's instructions. To ensure that only the highest quality data were used for assembly, reads were trimmed and filtered using CLC Genomics Workbench 9.0 (QIAGEN) set to a minimum length of 100 bp and a quality score threshold of 30. These trimmed reads were then mapped to the S. argenteus MSHR1132 genome (Accession Number: FR821777) with mode “Map Reads to Reference.” Moreover, trimmed reads were de novo assembled via the “De Novo Assembly” mode with default settings. Each assembled contig was aligned to the MSHR1132 genome and combined as a single pseudocontig using CONTIGuator2 (http:// contiguator.sourceforge.net/), producing pcg13064 and pcg13069. 2.6. Phylogeny based on k-mer diversity A k-mer is a substring of length k from a read obtained through NGS, and counting the occurrences of all such substrings is a central step in many DNA sequence analyses. The reads of suspected S. argenteus isolates (Tokyo13064 and Tokyo13069) were phylogenetically compared with reference genomes including S. argenteus MSHR1132 using the diversity of k-mer as the index. Using the “Create K-mer Tree” program of Microbial Genomics Module in CLC Genomics Workbench 9.0, pairwise distances among Tokyo13064, Tokyo13069, and references were calculated using Feature Frequency Profile via Jensen–Shannon divergences, and the phylogenetic tree was constructed using the neighbor-joining method. The k-mer length was set to be 16 bases for this analysis. The sequences of the following S. argenteus strain and 10 S. aureus strains were used for comparative analysis: MSHR1132, MRSA252 (BX571856), MW2 (BA000033), N315 (BA000018), Mu3 (AP009324), COL (CP000046), NCTC8325 (CP000253), Newman (AP009351), TW20 (FN433596), T0131 (CP002643), and M013 (CP003166).

2.2. Bacterial strains and culture The bacterial strains used in this study are listed in Table S1. These strains were isolated from the feces of patients or foods suspected to have caused food poisoning outbreaks in Tokyo, Japan. At first, we performed the typing of the enterotoxin by reversed passive latex agglutination for SEA–SEE and confirmed that both Tokyo13064 and Tokyo13069 produced only SEB protein (Fig. S3). Staphylococcal strains were cultured overnight in brain heart infusion (BHI) broth (Becton Dickinson, Sparks, MD, USA) or BHI broth supplemented with 1% yeast extract (Becton Dickinson) at 37 °C with shaking (100 rpm) to extract genomic DNA (gDNA) or total RNA (tRNA), respectively. The staphylococcal strain cultures were grown in BHI broth supplemented with 1% yeast extract at 37 °C for 48 h with shaking (100 rpm) to assess SE production. Escherichia coli DH5α was purchased from Promega (Madison, WI, USA). The E. coli strain was cultured overnight in Luria–Bertani broth (Sigma, St. Louis, MO, USA) containing 50 μg/ml ampicillin (Wako Pure Chemicals, Osaka, Japan) at 37 °C with shaking for plasmid isolation.

2.7. Phylogeny based on single-nucleotide polymorphisms Assembled contigs described in the “Genomic library preparation and whole genome sequencing” section were further compared with the reference genomes MSHR1132, MRSA252, MW2, N315, Mu3, COL, NCTC8325, Newman, TW20, T0131, and M013 based on single-nucleotide polymorphism (SNP) composition. First, assembled contigs and reference genome sequences were, respectively, split into 10,000-bp “reads” using Pyfasta (ver. 0.5.2) (Massachusetts Institute of Technology) and mapped to the MSHR1132 sequence as a reference using BWA-MEM (ver. 0.7.5a) (Li and Durbin, 2009). SNPs and indels were called using SAMtools (Li et al., 2009). After format conversion using VarScan (ver. 2.3.9) (Koboldt et al., 2009), indels were omitted using Vcftools (Danecek et al., 2011). The bases in which SNPs were

2.3. Preparation of gDNA gDNA was extracted from the cultured cells after treating the cells with lysostaphin (Wako Pure Chemicals) using a QIAamp DNA mini kit (QIAGEN GmbH, Hilden, Germany). The concentration of the extracted gDNA was determined using a QuantiFluor ONE dsDNA System (Promega), its purity was determined using a NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and its integrity was investigated by performing 0.8% agarose gel 32

International Journal of Food Microbiology 262 (2017) 31–37

Y. Suzuki et al.

quantifying the concentration of SEB in the culture were performed according to a method reported previously (Suzuki et al., 2015b). Capture and labeled antibodies for SEB were kindly supplied by Dr. Katsuhiko Omoe (Iwate University). Sandwich ELISA was performed in 96-well LumiNunc White plates (Thermo Fisher Scientific). The culture supernatant of the staphylococcal strain was treated overnight with 50% (v/v) normal rabbit serum (Thermo Fisher Scientific) at 4 °C to eliminate any nonspecific reactions due to protein A. The treated culture supernatant was then diluted 1000- to 100,000-fold in Can Get Signal immunoreaction enhancer solution 1 (TOYOBO). Capture antibody was diluted to 2 μg/ml in phosphate-buffered saline buffer and immobilized at 4 °C overnight. Anti-SEB antibody labeled with HRP was diluted 4800-fold in Can Get Signal immunoreaction enhancer solution 2 (TOYOBO). The luminescence signal was determined using the SpectraMax Paradigm (Molecular Devices, Sunnyvale, CA) and the SuperSignal ELISA Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific). The concentration of SE was determined by converting luminescence values to the corresponding concentrations using a standard curve (0.2–5 ng/ml) based on recombinant SEB protein purified as in previous reports (Hu et al., 2003). Unfortunately, we could not find the seb-harboring SFP isolates in the ATCC list. Therefore, we selected two strains, Tokyo12368 and Tokyo13001, as references; these were seb-harboring S. aureus strains isolated from the feces of SFP patients (Table S1).

detected in at least one sequence were connected to construct pseudogenes using the Python program “vcf2phhyloviz.py” (https://github. com/nickloman/misc-genomics-tools). If the SNPs were not located at the region commonly mapped to the MSHR1132 sequence among assembled contigs and reference genome sequences, the SNPs were omitted from this analysis. Pseudogenes were aligned, and the phylogenic tree was constructed using MEGA5 (Tamura et al., 2011) based on the neighbor-joining algorithm. 2.8. Comparative genomics Comparative genomics for circular whole genomes among MSHR1132, Tokyo13064, and Tokyo13069 was analyzed using the CGView Server (http://stothard.afns.ualberta.ca/cgview_server/index. html), which allowed sequence feature information to be visualized in the context of sequence analysis results. The CGView Server used the Basic Local Alignment Search Tool (BLAST) algorithm to compare the genome sequence to the comparison sequences. We demonstrated the homologous regions detected through tBLASTx comparisons of the sequence against reference itself, pcg13064, and pcg13069. The regions of the phage-related accessory elements such as prophages, Staphylococcus aureus pathogenicity islands (SaPIs), or genomic islands were identified in pcg13064 and pcg13069 using the PHAge Search Tool (Zhou et al., 2011) and BLAST. Putative boundaries of SaPIs (i.e., flanking direct repeats attL and attR) were identified using two-sequence BLASTn. Comparison of the sequences of accessory elements was facilitated using the Artemis Comparison Tool (Carver et al., 2005), which allowed an interactive visualization of BLASTn comparisons between genomes. A minimum score cutoff of 200 and a minimum percentage identity cutoff of 80% were adopted.

2.11. Accession numbers The accession numbers of run data are as follows: DRR083591 (Tokyo13064) and DRR083592 (Tokyo13069).

2.9. tRNA extraction and quantitative real-time RT-PCR

2.12. Statistical analysis

tRNA was extracted from Tokyo13064 and Tokyo13069. The cell bodies were harvested by centrifugation at 5000 ×g for 10 min, and the tRNA of the cells was then extracted using a RiboPure Bacteria kit (Thermo Fisher Scientific) according to the manufacturer's instructions. The amounts of tRNA were measured using a NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific). The quality of tRNA was confirmed by measuring the A260/A280 ratio and performing 1% agarose gel electrophoresis with no degradation of 23S or 16S ribosomal RNA. The extracted tRNA (2 μg) was reversely transcribed using random primers with the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) according to the manufacturer's instructions. The resultant cDNA was stored at −20 °C. The expression levels of enterotoxin genes were confirmed by qRT-PCR using SYBR Green Real-Time PCR Master Mix (TOYOBO, Osaka, Japan) and an ABI7500 Real-Time PCR System (Thermo Fisher Scientific) according to the manufacturer's instructions. We selected DNA gyrase subunit B (gyrB), superoxide dismutase (sodA), cell division protein (ftsZ), and RNA polymerase beta subunit (rpoB) as housekeeping genes for normalization. The primer sets were used as shown in Table S2. The standard curves for each gene were generated via serial dilution of the previously constructed plasmids to quantify their mRNA concentrations. A melting curve was used to confirm that the SYBR green-based amplicons were accurate. qRT-PCR was conducted with the following thermal cycling parameters: denaturation (95 °C for 1 min), 40 amplification cycles (95 °C for 15 s, 60 °C for 15 s, and 72 °C for 45 s with fluorescence reading during each extension step), and melting curve (60 °C–95 °C with a heating rate of 0.1 °C/s). The expression ratio of enterotoxin genes to those of gyrB, sodA, ftsZ, or rpoB was calculated to adjust for variations among the qRT-PCRs.

Statistical analyses were performed using Statcel2 software (The Publisher OMS Ltd., Saitama, Japan). SEB productions were analyzed initially by ANOVA and then by the Tukey–Kramer multiple comparison test.

3. Results 3.1. Overview of NGS analysis and phylogenetic tree analysis of Tokyo13064 and Tokyo13069 based on whole-genome sequencing The average read length of Tokyo13064 was 232 bp, and that of Tokyo13069 was 241 bp. In total, 96.9% and 96.8% of the trimmed reads of Tokyo13064 and Tokyo13069, respectively, were mapped to the MSHR1132 chromosome. These reads covered 95.3% and 95.4% of MSHR1132, respectively. The de novo assembly of the trimmed reads of Tokyo13064 and Tokyo13069 gave 14 and 10 contigs with > 1000 bp, respectively. These contigs represented an average coverage of 97.8–2017.9 and 192.7–2104.5 in the strain, respectively. No contig was matched to the pST75 plasmid, which was carried by MSHR1132. To confirm the phylogenetic positions of Tokyo13064 and Tokyo13069 relative to typical S. aureus and S. argenteus, we constructed two trees based on k-mer and SNPs using reported genomes or reads from NGS analysis. As shown in Fig. 1A, the result for k-mer tree analysis demonstrated that the genomic structures of both Tokyo13064 and Tokyo13069 were most similar to that of MSHR1132, which was the type strain of S. argenteus (Tong et al., 2015). A similar result was obtained by aligning the 245,800 SNP sites detected among the 13 analyzed genomes (Fig. 1B). MRSA252 was the most similar S. aureus strain to Tokyo13064 in the k-mer tree analysis, and the number of shared SNPs between Tokyo13064 and MRSA252 was 209,602 (of 245,800 sites), which is approximately 14-fold larger than the number shared between Tokyo13064 and MSHR1132 (14,882/245,800 sites).

2.10. Sandwich enzyme-linked immunosorbent assay Sandwich

enzyme-linked

immunosorbent

assay

(ELISA)

for 33

International Journal of Food Microbiology 262 (2017) 31–37

Y. Suzuki et al.

was 61,139 bp in size and inserted at the same site as the φSa3 (MSHR1132) prophage (see the yellow highlighted region of the outermost ring in Fig. 2). These novel prophages were most similar to the Staphylococcus phage StauST398-4 (NC_023499), and they included 59 coding sequences (CDSs) for phage component proteins and others (Fig. S4). The nucleotide sequence identity between the CDSs of novel prophages and homologous CDSs in other prophages was highly variable among the genes. A SaPI was detected at the 18′ site in both pseudocontigs. The sequences of this SaPI was completely identical to that of SaPIishikawa11 (see the orange highlighted region of the second and third outermost rings in Fig. 2), which is known as a seb-encoding SaPI (Sato'o et al., 2011). On the contrary, other insertable sites (8′, 9′, 19′, 44′, and 49′) were not occupied by any SaPI. In accordance with the recommendation from the International Nomenclature Committee for Staphylococcal Superantigen (Lina et al., 2004), molecular variants of enterotoxins harbored by S. argenteus were designated “S(arg)E” or “s(arg) e” with the appropriate alphabet in this report. The genomic islands νSAα and νSAβ were present at the same location in S. argenteus MSHR1132. The sequences of νSAα in pcg13064 and pcg13069 were identical and 30,590 bp in size, and they included 31 CDSs for nine superantigen-like proteins, nine lipoproteins, two restriction-modification system subunits, one nuclease, and several hypothetical proteins (Fig. 3). Compared with νSAα-typeV harbored by MSHR1132, there were some differences such as the number of CDSs for superantigen-like proteins and lipoproteins (Fig. 3). νSAβ in pcg13064 and pcg13069 was 28,185 bp in size, and it contained 22 CDSs (Fig. 3). This genomic island included an enterotoxin gene cluster (egc) region similar to egc in νSAβ-typeIV harbored by MSHR1132 (Holt et al., 2011) and egc3 in νSAβ-typeIII harbored by MRSA252 (Argudín et al., 2010), both of which encoded seg, sei, sem, sen, seo, and selu. These enterotoxin genes of νSAβ in pcg13064 and pcg13069 were identical to those of νSAβ-typeIV with the following percentages: s(arg) ei (99.5%), s(arg)em (99.4%), s(arg)en (96.7%), s(arg)eo (99.5%), and s (arg)elu (99.0%). The seg-like sequence encoded by νSAβ in pcg13064 and pcg13069 was 98.8% identical to that encoded by νSAβ-typeIV; however, there was a point mutation in the gene at the 436th base (“G” to “T”), resulting in alteration of the stop codon.

Fig. 1. Nucleotide divergence among Staphylococcus aureus strains and Staphylococcus argenteus strains. (A) The neighbor-joining tree was generated on the basis of the k-mer distribution among Tokyo13064, Tokyo13069, S. aureus MRSA252, MW2, N315, Mu3, COL, NCTC8325, Newman, TW20, T0131, M013, and S. argenteus MSHR1132. (B) The neighbor-joining tree was generated on the basis of the single-nucleotide polymorphisms (SNPs) shared among the aforementioned 13 strains. The tree was inferred from the alignment of 245,800 SNPs detected among all S. aureus and S. argenteus genomes and the pseudocontigs of Tokyo13064 and Tokyo13069. The tree was inferred with 1000 bootstrap replicates. Bar length = number of substitutions.

3.3. The mRNA expression levels of the S(arg)E genes To identify the causative enterotoxin(s), the mRNA expression levels of the S(arg)E genes harbored by Tokyo13064 and Tokyo13069 were examined. As shown in Fig. 4, similar results were confirmed using gyrB, sodA, ftsZ, and rpoB as internal controls. s(arg)eb was expressed by both strains. Conversely, low levels of transcription of s(arg)ei, s(arg)em, s(arg)en, s(arg)eo, and s(arg)lu, which are collectively termed “egc-related S(arg)E genes,” were detected in both strains (Fig. 4).

3.2. General characteristics of the Tokyo13064 and Tokyo13069 genomes 10 of 14 contigs of Tokyo13064 and 8 of 10 contigs of Tokyo13069 were combined to construct a single pseudocontig using the MSHR1132 genome sequence as a reference (pcg13064 and pcg13069, see the second and third outermost rings in Fig. 2). All of the uncombined contigs included regions encoding transfer RNA or ribosomal RNA. Both pcg13064 and pcg13069 possessed the alleles of ST1223, which was one of STs of CC75 (Chantratita et al., 2016; Holt et al., 2011). The crtOPQMN operon was not detected in these pseudocontigs. From the BLAST results (Fig. 2), several elements harbored by MSHR1132 were not detected in pcg13064 and pcg13069. Staphylococcal cassette chromosome mec (SCCmec), host specificity defective genes, and the clustered regularly interspaced short palindromic repeat (CRISPR)/CRISPR-associated (cas) genes (CRISPR/cas) were not found (see the red highlighted region of the outermost ring in Fig. 2). Moreover, an integrative conjugative element (ICE6013MSHR1132) and transposons were also missing (see the green or pink highlighted region of the outermost ring in Fig. 2). Both the pseudocontigs possessed the same novel prophage, which

3.4. Detection of S(arg)EB production To address the toxicity of food poisoning caused by Tokyo13064 and Tokyo13069 expressing s(arg)eb, the production of S(arg)EB or SEB was measured in the two isolates and two reference S. aureus strains (Tokyo12368 and Tokyo13001) isolated from past food poisoning outbreaks. Both S. argenteus isolates produced the S(arg)EB protein at levels exceeding 40 μg/ml. Compared with the SEB production levels of the reference S. aureus strains, no significant differences were observed (P > 0.05, Fig. 5). 4. Discussion S. argenteus has recently been described as a novel species of Staphylococcus. It was characterized as belonging to the CC75 lineage, which is extremely distinct from other clones of S. aureus, as well as by the lack of staphyloxanthin production. S. argenteus has been 34

International Journal of Food Microbiology 262 (2017) 31–37

Y. Suzuki et al.

Fig. 2. Comparison of the MSHR1132 genomic sequence to the pseudocontig sequences of Tokyo13064 and Tokyo13069. The outermost ring shows each region of mobile genetic elements harbored by MSHR1132. The next ring shows the pseudocontigs of Tokyo13064 (pcg13064: outside) and Tokyo13069 (pcg13069: inside). The assembled contig numbers are designed in each ring. The next two rings comprising colored arrowheads show features extracted from the MSHR1132 genome GenBank file. The next three rings indicate the positions of BLAST hits detected through tblastx comparisons of the sequence against itself (blue ring) and two pseudocontigs (broken red and green rings for pcg13064 and pcg13069, respectively). The height of each arc in the BLAST results rings is proportional to the percent identity of the hit. Overlapping hits appear as darker arcs. The two subsequent rings show the GC content and GC skew of the MSHR1132 genome, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

almost all Staphylococcus strains ranges from 2.4 to 3.0 Mbp, the results of nucleotide divergence of Tokyo13064 and Tokyo13069 are in agreement with those of a previous comparison between S. argenteus (MSHR1132) and S. aureus. Moreover, the representative characteristics for S. argenteus were present in both Tokyo13064 and Tokyo13069, namely both strains belong to the CC75 (ST1223) lineage and lack the crtOPQMN operon. These findings indicate that both Tokyo13064 and Tokyo13069 belong to “S. argenteus.” The CC75 lineage strains including MSHR1132 would have an ancestral phenotype that is less impacted by the accessory element (Holt et al., 2011). The accessory elements harbored by MSHR1132 but not Tokyo13064 and Tokyo13069 include an SCCmec element, CRISPR/cas genes, an ICE element, and a transposon, whereas SaPIishikawa11 is detected in the latter but not in the former (Fig. 2). In terms of the number or type of accessory elements, it is possible that Tokyo13064 and Tokyo13069 are collateral lineages that branched off from MSHR1132 lineages. SEs are exotoxins that cause staphylococcal food poisoning in

considered to have attenuated virulence compared with that of S. aureus because there were only a few reports describing the isolation of S. argenteus from patients with a serious staphylococcal disease (Holt et al., 2011), and the prevalence of virulence factors was notably lower in S. argenteus (Chantratita et al., 2016). This is, to the best of our knowledge, the first study to demonstrate that S. argenteus caused a food poisoning outbreak via whole-genome sequencing and enterotoxin expression analyses. The two phylogenetic trees constructed using NGS data illustrated that the genome structures of Tokyo13064 and Tokyo13069 were similar to that of MSHR1132 (Fig. 1), which was the type strain of S. argenteus (Tong et al., 2015). The numbers of SNPs shared between the pcg13064 and MRSA252 genomes and between the pcg13064 and MSHR1132 genomes were approximately 209,602 and 14,882, respectively. Some researchers reported that S. argenteus was an early branching lineage because of the approximately 10% nucleotide divergence compared with any S. aureus lineage (Holt et al., 2011; Ng et al., 2009). Considering the fact that the size of the whole genome in

Fig. 3. Comparison of the structures of νSaα-typeV and νSaβ-typeIV to those harbored by Tokyo13064 and Tokyo13069. Comparisons were performed using the Artemis Comparison Tool with a minimum score cutoff of 200 and a minimum percentage identity cutoff of 80%. Forward matches are colored in red. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

35

International Journal of Food Microbiology 262 (2017) 31–37

Y. Suzuki et al.

Fig. 4. Enterotoxin gene expression profile of each of the Staphylococcus argenteus isolates obtained during staphylococcal food poisoning outbreaks. (A) Tokyo13064 and (B) Tokyo13069 are shown. The amount of s(arg)e mRNA was normalized to those of gyrB, sodA, ftsZ, or rpoB mRNA as an internal control. Experiments were performed in triplicate, and data are presented as mean ± standard deviation.

Tokyo13064 and Tokyo13069 harbored s(arg)eb, which is located in SaPIishikawa11, as well as s(arg)ei, s(arg)em, s(arg)en, s(arg)eo, and s (arg)elu, which are located in νSaβ (Figs. 2 and 3). One study detected seb in all ST1223 S. argenteus strains isolated from the noses of villagers in the Amazonian forest (Ruimy et al., 2010). The other study conducted in Thailand demonstrated that only 1 of 58 S. argenteus isolates, which included ST2250 (54 of 58 isolates), ST2854 (2 of 58), ST1223 (1 of 58), and ST2198 (1 of 58), harbored seb (Chantratita et al., 2016). These facts and the results of this study suggest that the ST1223 lineage is more susceptible to the acquisition of seb-related SaPIs than other S. argenteus lineages. Interestingly, both Tokyo13064 and Tokyo13069 lacked CRISPR/cas genes present in the MSHR1132 genome, implying that the acquisition might be affected by differences in the type or presence/absence of “defense islands” such as a CRISPR/cas system. The results of qRT-PCR demonstrated that a significant transcription of s(arg)eb occurred, whereas transcription of egc-related S(arg)E genes was not generally detected in either isolate (Fig. 4). Previous studies conducted using S. aureus isolates demonstrated that small amounts of the transcriptomes of strains harboring egc-related SE genes such as seg and sei were detected in laboratory medium, indicating that these strains produce low levels of these SE proteins (Derzelle et al., 2009; Pocsfalvi et al., 2008; Sato'o et al., 2015; Suzuki et al., 2015a). These facts suggested that egc-related SEs were less likely to cause food poisoning outbreaks. In the present study, the expression of each S(arg)E gene is also in agreement with the findings of previous studies, indicating that S(arg)EB was a causative toxin of the food poisoning outbreak in Tokyo. SEB is known to be the second or third most common SE associated with food poisoning, and a dosage of 20–25 μg of SEB is sufficient to cause food poisoning (Raj and Bergdoll, 1969). In the present study, both S. argenteus isolates produced significant amounts (> 50 μg/ml) of

Fig. 5. Amounts of S(arg)EB produced by Staphylococcus argenteus isolates originating from staphylococcal food poisoning outbreaks. Experiments were performed in triplicate, and data are presented as mean ± standard deviation. Statistical analysis was performed using analysis of variance (ANOVA) followed by the Tukey–Kramer multiple comparison test. N.S. = not significant.

humans, which results in symptoms such as vomiting and nausea (Balaban and Rasooly, 2000; Dinges et al., 2000). Many enterotoxin genes are encoded in accessory elements such as prophages (e.g., sea, sep, sek), plasmids (e.g., sed, selj, ser), transposons (e.g., seh), νSaβs (e.g., seg, sei, sem), and SaPIs (e.g., seb, sec, sel) (Argudín et al., 2010; Malachowa and DeLeo, 2010), indicating that horizontal transfer of the accessory elements is an important phenomenon for acquiring enterotoxingenic pathogenicity. In the present study, the genomes of 36

International Journal of Food Microbiology 262 (2017) 31–37

Y. Suzuki et al.

Genome Biol. Evol. 3, 881–895. Hu, D.L., Omoe, K., Shimoda, Y., Nakane, A., Shinagawa, K., 2003. Induction of emetic response to staphylococcal enterotoxins in the house musk shrew (Suncus murinus). Infect. Immun. 71, 567–570. Jenney, A., Holt, D., Ritika, R., Southwell, P., Pravin, S., Buadromo, E., Carapetis, J., Tong, S., Steer, A., 2014. The clinical and molecular epidemiology of Staphylococcus aureus infections in Fiji. BMC Infect. Dis. 14, 160. Koboldt, D.C., Chen, K., Wylie, T., Larson, D.E., McLellan, M.D., Mardis, E.R., Weinstock, G.M., Wilson, R.K., Ding, L., 2009. VarScan: variant detection in massively parallel sequencing of individual and pooled samples. Bioinformatics 25, 2283–2285. Li, H., Durbin, R., 2009. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754–1760. Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., Marth, G., Abecasis, G., Durbin, R., 1000 Genome Project Data Processing Subgroup, 2009. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079. Lina, G., Bohach, G.A., Nair, S.P., Hiramatsu, K., Jouvin-Marche, E., Mariuzza, R., International Nomenclature Committee for Staphylococcal Superantigens, 2004. Standard nomenclature for the superantigens expressed by Staphylococcus. J. Infect. Dis. 189, 2334–2336. Long, S.W., Beres, S.B., Olsen, R.J., Musser, J.M., 2014. Absence of patient-to-patient intrahospital transmission of Staphylococcus aureus as determined by whole-genome sequencing. MBio 5, e01692–14. Malachowa, N., DeLeo, F.R., 2010. Mobile genetic elements of Staphylococcus aureus. Cell. Mol. Life Sci. 67, 3057–3071. Ng, J.W., Holt, D.C., Lilliebridge, R.A., Stephens, A.J., Huygens, F., Tong, S.Y., Currie, B.J., Giffard, P.M., 2009. Phylogenetically distinct Staphylococcus aureus lineage prevalent among indigenous communities in northern Australia. J. Clin. Microbiol. 47, 2295–2300. Pocsfalvi, G., Cacace, G., Cuccurullo, M., Serluca, G., Sorrentino, A., Schlosser, G., Blaiotta, G., Malorni, A., 2008. Proteomic analysis of exoproteins expressed by enterotoxigenic Staphylococcus aureus strains. Proteomics 8, 2462–2476. Raj, H.D., Bergdoll, M.S., 1969. Effect of enterotoxin B on human volunteers. J. Bacteriol. 98, 833–834. Ruimy, R., Angebault, C., Djossou, F., Dupont, C., Epelboin, L., Jarraud, S., Lefevre, L.A., Bes, M., Lixandru, B.E., Bertine, M., El Miniai, A., Renard, M., Bettinger, R.M., Lescat, M., Clermont, O., Peroz, G., Lina, G., Tavakol, M., Vandenesch, F., van Belkum, A., Rousset, F., Andremont, A., 2010. Are host genetics the predominant determinant of persistent nasal Staphylococcus aureus carriage in humans? J Infect Dis 202, 924–934. Sato'o, Y., Omoe, K., Ono, H.K., Nakane, A., Hu, D.L., 2011. A novel comprehensive analysis method for Staphylococcus aureus pathogenicity islands. Microbiol. Immunol. 57, 91–99. Sato'o, Y., Hisatsune, J., Nagasako, Y., Ono, H.K., Omoe, K., Sugai, M., 2015. Positive regulation of staphylococcal enterotoxin H by rot (repressor of toxin) protein and its importance in clonal complex 81 subtype 1 lineage-related food poisoning. Appl. Environ. Microbiol. 81, 7782–7790. Suzuki, Y., Kobayashi, M., Matsushita, S., Uehara, S., Kato, R., Sato'o, Y., Ono, H.K., Sadamasu, K., Kai, A., Kamata, Y., 2015a. Detection of the staphylococcal enterotoxin D-like gene from staphylococcal food poisoning isolates over the last two decades in Tokyo. J. Vet. Med. Sci. 77, 905–911. Suzuki, Y., Kubota, H., Sato'o, Y., Ono, H.K., Kato, R., Sadamasu, K., Kai, A., Kamata, Y., 2015b. Identification and characterization of novel Staphylococcus aureus pathogenicity islands encoding staphylococcal enterotoxins originating from staphylococcal food poisoning isolates. J. Appl. Microbiol. 118, 1507–1520. Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., Kumar, S., 2011. MEGA5: Molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28, 2731–2739. Thaipadungpanit, J., Amornchai, P., Nickerson, E.K., Wongsuvan, G., Wuthiekanun, V., Limmathurotsakul, D., Peacock, S.J., 2015. Clinical and molecular epidemiology of Staphylococcus argenteus infections in Thailand. J. Clin. Microbiol. 53, 1005–1008. Tong, S.Y., Lilliebridge, R.A., Bishop, E.J., Cheng, A.C., Holt, D.C., McDonald, M.I., Giffard, P.M., Currie, B.J., Boutlis, C.S., 2010. Clinical correlates of Panton-Valentine leukocidin (PVL), PVL isoforms, and clonal complex in the Staphylococcus aureus population of Northern Australia. J. Infect. Dis. 202, 760–769. Tong, S.Y., Sharma-Kuinkel, B.K., Thaden, J.T., Whitney, A.R., Yang, S.J., Mishra, N.N., Rude, T., Lilliebridge, R.A., Selim, M.A., Ahn, S.H., Holt, D.C., Giffard, P.M., Bayer, A.S., Deleo, F.R., Fowler Jr., V.G., 2013. Virulence of endemic nonpigmented northern Australian Staphylococcus aureus clone (clonal complex 75, S. argenteus) is not augmented by staphyloxanthin. J. Infect. Dis. 208, 520–527. Tong, S.Y., Schaumburg, F., Ellington, M.J., Corander, J., Pichon, B., Leendertz, F., Bentley, S.D., Parkhill, J., Holt, D.C., Peters, G., Giffard, P.M., 2015. Novel staphylococcal species that form part of a Staphylococcus aureus-related complex: the nonpigmented Staphylococcus argenteus sp. nov. and the non-human primate-associated Staphylococcus schweitzeri sp. nov. Int. J. Syst. Evol. Microbiol. 65, 15–22. Xiong, Y.Q., Yang, S.J., Tong, S.Y., Alvarez, D.N., Mishra, N.N., 2015. The role of staphylococcal carotenogenesis in resistance to host defense peptides and in vivo virulence in experimental endocarditis model. Pathog. Dis. 73 (pii:ftv056). Zhou, Y., Liang, Y., Lynch, K.H., Dennis, J.J., Wishart, D.S., 2011. PHAST: a fast phage search tool. Nucleic Acids Res. 39, W347–352.

S(arg)EB proteins compared with the level of SEB production by the reference S. aureus strains that were isolated during previous food poisoning outbreaks (Fig. 5). Although SEB production in a laboratory medium cannot be simply applied to that in a food environment due to differences in pH, nutrients, water activity, and other variables, this fact strongly indicated that S. argenteus strains encoding SEB can cause food poisoning. In conclusion, the present study uncovered evidence that S. argenteus was a causative agent of food poisoning. Comprehensive analysis of genomic structures illustrated that two food poisoning isolates (Tokyo13064 and Tokyo13069) lacked the crtOPQMN operon and that they were more similar to the type strain of S. argenteus than S. aureus strains, thus indicating that both are “S. argenteus.” In contrast, these S. argenteus were considered to be distinct lineages that branched off from lineages of the type strain in terms of accessory elements. Both S. argenteus isolates can cause food poisoning because they express significant levels of s(arg)eb and S(arg)EB in laboratory medium. Our findings suggest that S. argenteus supplemented by particular accessory elements possesses the potential to cause staphylococcal diseases that are not limited to food poisoning, similarly as S. aureus does. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ijfoodmicro.2017.09.005. Acknowledgements This study was supported by the Research Program on Emerging and Re-emerging Infectious Diseases from the Japan Agency for Medical Research and Development (grant number: 16fk0108119j0001). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. We thank members of the food poisoning laboratory (at the Tokyo Metropolitan Institute of Public Health) for kindly providing the staphylococcal strains used in this study. References Argudín, M.Á., Mendoza, M.C., Rodicio, M.R., 2010. Food poisoning and Staphylococcus aureus enterotoxins. Toxins (Basel) 2, 1751–1773. Balaban, N., Rasooly, A., 2000. Staphylococcal enterotoxins. Int. J. Food Microbiol. 61, 1–10. Carver, T.J., Rutherford, K.M., Berriman, M., Rajandream, M.A., Barrell, B.G., Parkhill, J., 2005. ACT: the Artemis Comparison Tool. Bioinformatics 21, 3422–3423. Chantratita, N., Wikraiphat, C., Tandhavanant, S., Wongsuvan, G., Ariyaprasert, P., Suntornsut, P., Thaipadungpanit, J., Teerawattanasook, N., Jutrakul, Y., Srisurat, N., Chaimanee, P., Anukunananchai, J., Phiphitaporn, S., Srisamang, P., Chetchotisakd, P., West, T.E., Peacock, S.J., 2016. Comparison of community-onset Staphylococcus argenteus and Staphylococcus aureus sepsis in Thailand: a prospective multicentre observational study. Clin. Microbiol. Infect 22 (458.e11-9). Danecek, P., Auton, A., Abecasis, G., Albers, C.A., Banks, E., DePristo, M.A., Handsaker, R.E., Lunter, G., Marth, G.T., Sherry, S.T., McVean, G., Durbin, R., 1000 Genomes Project Analysis Group, 2011. The variant call format and VCFtools. Bioinformatics 27, 2156–2158. Derzelle, S., Dilasser, F., Duquenne, M., Deperrois, V., 2009. Differential temporal expression of the staphylococcal enterotoxins genes during cell growth. Food Microbiol. 26, 896–904. Dinges, M.M., Orwin, P.M., Schlievert, P.M., 2000. Exotoxins of Staphylococcus aureus. Clin. Microbiol. Rev. 13, 16–34. Dupieux, C., Blondé, R., Bouchiat, C., Meugnier, H., Bes, M., Laurent, S., Vandenesch, F., Laurent, F., Tristan, A., 2015. Community-acquired infections due to Staphylococcus argenteus lineage isolates harbouring the Panton-Valentine leucocidin, France, 2014. Euro Surveill. 20 (pii), 21154. Enright, M.C., Day, N.P., Davies, C.E., Peacock, S.J., Spratt, B.G., 2000. Multilocus sequence typing for characterization of methicillin-resistant and methicillin-susceptible clones of Staphylococcus aureus. J. Clin. Microbiol. 38, 1008–1015. Holt, D.C., Holden, M.T., Tong, S.Y., Castillo-Ramirez, S., Clarke, L., Quail, M.A., Currie, B.J., Parkhill, J., Bentley, S.D., Feil, E.J., Giffard, P.M., 2011. A very early-branching Staphylococcus aureus lineage lacking the carotenoid pigment staphyloxanthin.

37