Molecular Epidemiology and Identification of a Staphylococcus aureus Clone Causing Food Poisoning Outbreaks in Japan

Abstract

Molecular characterization of isolates from staphylococcal food poisoning (SFP) outbreaks in Japan showed that the dominant lineage causing SFP outbreaks is clonal complex 81 (CC81), a single-locus variant of sequence type 1, coagulase type VII, positive for sea and/or seb, and positive for seh. Among various CC lineages producing staphylococcal enterotoxin A, CC81 showed the highest toxin productivity.

TEXT

Staphylococcal food poisoning (SFP), one of the most common food-borne diseases, results from the consumption of foods containing toxic amounts of staphylococcal enterotoxins (SEs) (1–4). SFP is associated with toxigenic Staphylococcus aureus strains that produce one or more members of a family of genes encoding heat-stable SEs. Recently, a superfamily of more than 23 different SEs and SE-like toxins (SEls) was studied for their biological activities (4–8). These bacterial toxins are also known as pyrogenic superantigens that stimulate polyclonal T-cell proliferation and can potentially cause toxic shock syndrome (1–4). The genes for SEs and SEls are harbored by various mobile genetic elements and/or genomic islands, including prophages, plasmids, S. aureus pathogenicity islands (SaPIs), and νSaβ. To date, in addition to the five classical types of SEs (SEA through SEE), 18 new types of SEs and SEls (SEG through SElX) have been described (4–8). Our recent study confirmed the emetic potential of SElK, SElL, SElM, SElN, SElO, SElP, and SElQ in the monkey, and these SEls were renamed SEK, SEL, SEM, SEN, SEO, SEP, and SEQ, respectively (9). Comparing SEs and SEls, SEA is considered the most important SE causing SFP. S. aureus is a common commensal bacterium of the skin and mucosal surfaces of humans, with estimates of 20% persistent and 60% intermittent colonization (10). Food handlers carrying enterotoxin-producing S. aureus in their nasal cavities or on their skin are important sources of food contamination during the cooking process (3). Contamination with S. aureus is believed to be associated primarily with improper handling of cooked or processed foods with improper storage under conditions that allow the growth of S. aureus and the production of SEs. Reports concerning the molecular epidemiology and genetic diversity of the isolates from SFP outbreaks are limited when clinical isolates are compared. In this investigation, we examined isolates from SFP outbreaks in Japan to study the molecular epidemiology of S. aureus causing SFP.

In this study, 371 S. aureus isolates from different prefectures in Japan were used. Forty-two of them were from SFP outbreaks from 1990 to 2010, and the other 329 were collected from the nasal cavities of healthy humans from 2005 to 2010. Bacterial culture conditions, DNA extraction, coagulase typing, SE/SEl genotyping, and multilocus sequence typing (MLST) were performed as described previously (11–14). SEA production was determined by ELISA according to our previous method (5), with some modification. Briefly, 10³ S. aureus cells were inoculated into broth medium. SuperSignal ELISA Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific), Wallac 1420 ARVO MX/Light (PerkinElmer), and Varioskan Flash (Thermo Fisher Scientific) were used for measurement of SEA concentrations. Typing of genomic elements by SaPI scanning was performed as previously described (11). In addition, we designed new primer sets for a putative truncated transposon harboring seh, the enterotoxin gene cluster (egc), and ϕSa3. For the seh transposon, SEHexS (5′-ACTAATGTCTCCATTGGTTGTTCTCTATTAAAAGCGCGAT-3′) and SEHexAS (5′-ATATCGTCTTTCGACACGTTGTAAGTGAAGCAGCTAGA-3′) were used. For egc, egcS (5′-TCTTAGAAGAGGATGGCTTCGAAACATTTACAGCG-3′) and egcAS (5′-GGTGGTATTGCGATCCCATTAAAAGATATTGAAGATTTAGA-3′) were used. ϕSa3 was divided into three parts as follows. Sa3RS (5′-GTGAGTTTCGTGATTTCAAAGGTTGGACTAAGATG-3′) and Sa3RAS (5′-GAC AATAGTGCCAAAGCCGAATCTAAGAAAGATGAT-3′), Sa3CS (5′-ATCTATCCACTCTTTCTCATCCATATCAATAGGTTTACG-3′) and Sa3CAS (5′-TACGCGTTTAAGCGAGTACAGAGAAAAGAAAACAATAGA-3′), and Sa3LS (5′-GAATCTTCAGATTGTGTATGTGTACCGATAACGT-3′) and Sa3LAS (5′-AACCCGTTGTATCCTTTAGTTTTAACTACTTCATC AAG-3′) were used to analyze the right, center, and left portions of the ϕSa3 phage, respectively. PCR-amplified products were analyzed by restriction enzyme (HindIII, EcoRV, or HaeIII) digestion.

To determine the biological characteristics of SFP isolates, we first determined their coagulase types and enterotoxin genotypes. More than 70% of the SFP isolates were of coagulase type VII (Fig. 1A). In contrast, the isolates from nasal swabs showed various coagulase types. SE and SEl genotyping showed that all of the SFP isolates and 68.7% of the nasal swab isolates were positive for at least one of the se or sel genes or tst. The SFP isolates showed a significantly higher prevalence of sea (71.4%), followed by seh (54.8%), sek (52.4%), seq (52.4%), and seb (45.2%) (Fig. 1B). Among the SFP isolates, the dominant genotype was sea seb seh sek seq (28.6%), followed by sea seh sek seq (19.0%) (see Table S1 in the supplemental material). In contrast, the most frequent genotype among the nasal swab isolates was seg sei sem sen seo (18.8%), followed by sep alone (6.1%). To analyze the genetic backgrounds of the SFP isolates, MLST was performed. The data showed that clonal complex 81 (CC81) was the primary CC (54.8%) among the SFP isolates; while this CC was rare (2.7%) among the nasal swab isolates (Table 1). Comparison of the CC types and the se genotype showed two dominant genotypes of SFP isolates, sea seb seh sek seq and sea seh sek seq, that exclusively belonged to CC81 (Table 2). A summary of genotypes in the CC81 lineage is shown in Fig. 2. Genotypes sea seb seh sek seq, sea seh sek seq, and seb seh were present in both SFP and nasal swab isolates. Two genotypes, sec seg sei sel sem sen seo and seg sei sem sen seo, were found only among the nasal swab isolates. To further clarify the genomic characteristics of the isolates belonging to the CC81 lineages, we typed the genomic elements of all of the CC81 isolates (23 from SFP and 9 from nasal swabs) by PCR (11). This showed that CC81 can be classified into two subtypes, 1 and 2 (Table 3). All of the isolates belonging to subtype 1 were of coagulase type VII and positive for the 1.5′ seh transposon; whereas those of subtype 2 were of coagulase type VI and negative for the 1.5′ seh transposon. The restriction pattern of PCR products of the 9′ genomic element was subtype specific, assigned to type A for subtype 1 and type B for subtype 2 (see Fig. S1 in the supplemental material). Some isolates of subtype 1 were positive for other mobile elements such at 18′, 19′, 44′, and ϕSa3, as shown in Table 3. All subtype 1 isolates were negative for egc at 40′. Conversely, subtype 2 isolates were all positive for 40′ egc but negative for 19′, 44′, and ϕSa3. The enterotoxin profiles of subtypes 1 and 2 were totally different. All of the isolates belonging to subtype 1 were positive for sea, seb, or both; while all of the isolates belonging to subtype 2 were negative for both sea and seb. Of note, all CC81 SFP isolates were classified as subtype 1 and there was a significant difference between the SFP and nasal swab isolates in the rate of CC81 subtype 1 detection (Fig. 2, P = 6.6E-25), suggesting the CC81 subtype 1 is associated specifically with SFP outbreaks in Japan.

FIG 1.

Genetic typing of isolates from SFP outbreaks and nasal swabs. The percentages of isolates of the different coagulase types from SFP outbreaks (n = 42) and nasal swabs (n = 329) are shown. (A) Coagulase typing. NT, not typeable. (B) SE/SEl genotyping.

TABLE 1.

CC types of the strains isolated from SFP outbreaks and nasal swabs

CC type	No. (%) of isolates from:
	SFP outbreaks (n = 42)	Nasal swabs (n = 329)
CC5	3 (7.1)	28 (8.5)
CC6	2 (4.8)	6 (1.8)
CC7		3 (0.9)
CC8	4 (9.5)	27 (8.2)
CC12	2 (4.8)	27 (8.2)
CC15		42 (12.8)
CC20		13 (4.0)
CC25		9 (2.7)
CC30	1 (2.4)	17 (5.2)
CC59	2 (4.8)	4 (1.2)
CC81	23 (54.8)	9 (2.7)
CC96	1 (2.4)	5 (1.5)
CC97		7 (2.1)
CC101		2 (0.6)
CC121		14 (4.3)
CC188		41 (12.5)
CC398		3 (0.9)
CC508	4 (9.5)	62 (18.8)
CC509		5 (1.5)
Others		5 (1.5)

TABLE 2.

Relationships between se/sel types and CCs

se/sel type	No. (%) of isolates from SFP outbreaksa
	CC5	CC6	CC8	CC12	CC30	CC59	CC81	CC96	CC508
sea		2 (4.8)	2 (4.8)
sea seb seh sek seq							15 (28.6)
sea seb sek seq						2 (4.8)
sea sec sel								1 (2.4)
sea sec sel tst-1
sea sed selj ser			1 (2.4)
sea seg sei sek sen seq tst-1
sea seg sei sem sen seo	1 (2.4)
sea seg sei sen tst-1					1 (2.4)
sea seg sen
sea seg sen tst-1
sea seg sen sep tst-1
sea seh sek seq							5 (19.0)
sea sei sen tst-1
sea seq tst-1
seb seh							3 (7.1)
seb sep				2 (4.8)
seg sei sem sen seo									3 (7.1)
seg sei sem sen seo sep	1 (2.4)								1 (2.4)
seg sei selj sem sen seo ser	1 (2.4)
selj ser			1 (2.4)

^aTotal n = 42.

FIG 2.

Relationship between MLST and SE/SEl types. Percentages of CC81 and its enterotoxin profiles among isolates from SFP and nasal swabs. Isolates harboring sea seb seh sek seq, sea seh sek seq, and seb seh were found in SFP and nasal swab samples (subtype 1). Isolates harboring sec seg sei sel sem sen seo and seg sei sem sen seo were present only in nasal swab samples (subtype 2). A chi-square test comparing CC81 subtype 1 isolates from SFP and nasal swab samples was performed.

TABLE 3.

Genetic backgrounds of two subtypes of CC81 lineages

Locus	Genomic element(s)	Presencea of or version found in:
		Subtype 1 (n = 30)b	Subtype 2 (n = 2)c
1.5′	seh transposon	+ [30]	−
9′	Restriction pattern	Type A (no SEs/SEls) [30]	Type B (no SEs/SEls) [2]
18′	SaPIs	SaPIishikawa11 (seb) [12]	Putative SaPIs (sec sel) [1]
19′	SaPIs	SaPIhhms2 (no SEs/SEls) [2]	−
40′	egc	−	+ (seg sei sem sen seo) [2]
44′	SaPIs	SaPIno10 (seb) [7]	−
44′ (φSa3)	Prophage	φSa3mw2 (sea sek seq) [25]	−

^a+, positive; −, negative. In brackets is the number of isolates positive.

^bOrigins: SFP, n = 23; nasal swabs, n = 7. Coagulase type, VII; se/sel type, sea seb seh sek seq.

^cBoth isolates were from nasal swabs. Coagulase type, VI; se/sel type, sec seg sei sel sem sen seo.

CC81 (ST81) is a single-locus variant of CC1 (ST1), to which the typical clinical S. aureus lineages belong, e.g., MW2 and MSSA476 (15, 16). Because CC81 harboring sea is the major lineage of SFP isolates, we further analyzed and compared the SEA production of isolates of the various CC types. As shown in Fig. 3, sea-harboring CC81 isolates produced significantly larger quantities of SEA than other sea-harboring CC types did (Fig. 3). The SEA prophage genotype has been reported to influence SEA production (17), where S. aureus isolates were divided into high and low SEA production types based on prophage genotypes. Cao et al. reported that the genotype ϕSa3mw2 was a high SEA production type (18), and the sea-harboring CC81 isolates in this study also carry ϕSa3mw2. The data suggest the high toxin production by the CC81 lineage may be one of the reasons they cause SFP outbreaks more frequently than other isolates.

FIG 3.

SEA production by S. aureus isolates of various CCs. SEA production was evaluated by sandwich ELISA (5). For CC81, 10 isolates were assayed; for others, 2 isolates were assayed. Each isolate was assayed for toxin production in three independent experiments. The data are means ± standard errors. The SEA production of CC81 was compared to that of other CCs isolates by using the Student t test and the Holm method. ***, P < 0.001.

To our knowledge, this is the first report showing that CC81 lineage subtype 1 is closely related to SFP outbreaks in Japan. Genome sequencing and characterization of the mechanism of SEA production by the CC81 lineage are necessary to understand the high production of SEA and its strong association with SFP.