Heterogeneity and phylogenetic relationships of community-associated methicillin-sensitive/resistant Staphylococcus aureus isolates in healthy dogs, cats and their owners


Chin Cheng Chou, School of Veterinary Medicine, National Taiwan University, No. 1, Section 4, Roosevelt Road, Taipei 10617, Taiwan. E-mail: chouchin@ntu.edu.tw


Aims:  To investigate the distribution of staphylococcal enterotoxin genes (se) and the molecular features of community-associated methicillin-sensitive/resistant Staphylococcus aureus (CA-MSSA/MRSA) isolates in the nostrils of healthy pets and their owners.

Methods and Results:  A total of 114 Staph. aureus isolates were identified from 1563 nasal swab samples, and CA-MRSA accounted for 20·2% (n = 23) of the total identified isolates. CA-MRSA isolates (91·3%, 21/23) harboured higher percentage of se than did CA-MSSA isolates (58·2%, 53/91) (P < 0·01), and the two highest se profiles of CA-MRSA were seb-sek-seq (42·9%, 9/21) and seb-sek-seq-sep (28·6%, 6/21). Of the MSSAs, 42·8% (39/91) were resistant to at least one antimicrobial drug and 8·8% (8/91) were multidrug resistant (MDR). We identified nine staphylocoagulase (SC) types (I–VIII and X) and three multilocus sequence types (ST59-MRSA-IV/V, ST-239-MRSA-V and ST241-MRSA-V). SC VII (23·4%, 22/94), a staphylococcal food poisoning isolate found mainly in Japan, and ST-59-MRSA-IV/V (85%, 17/20), a widespread CA-MRSA clone found mainly in Taiwan, both were the most predominant types. Phylogenetic analysis together with se and molecular characteristics obtained using pulsed-field gel electrophoresis showed that high levels of antimicrobial resistance and the se-carrying clone ST59-MRSA-IV/V-SC VII were all clustered in genogroup 5.

Conclusions:  The CA-MRSA clone of se-carrying-MDR-ST-59-IV/V-SC VII was identified predominantly in this study, and this clone might play a significant role in staphylococcal food poisoning in community settings.

Significance and Impact of the Study:  To our knowledge, this is the first study focussing on enterotoxin-carrying CA-MRSA/MSSA in pets and their owners, and the results support the future warnings in animal–human bond caused by CA-staphylococci in the commonwealth and the need to take cautions worldwide.


Coagulase-positive strains of Staphylococcus aureus (Staph. aureus) cause a wide range of infections, including endocarditis, osteomyelitis and abscesses (Brakstad et al. 1992). In the past few decades, methicillin-resistant Staph. aureus (MRSA) has gained global attention since becoming a major source of hospital-associated (HA) and community-associated (CA) infections (Vandenesch et al. 2003). MRSA carries the gene mecA, which resides in the staphylococcal chromosome cassette mec (SCCmec) (Katayama et al. 2000). Upon acquisition of SCCmec, methicillin-sensitive Staph. aureus (MSSA) transforms into MRSA, which is resistant to β-lactam antibiotics (Katayama et al. 2000). Many studies have examined the possibility of MRSA transmission between owners and their pets, including dogs, cats and horses. MRSA has emerged as a potential zoonotic pathogen (Morgan 2008; Cuny et al. 2010).

Some strains of Staph. aureus can produce staphylococcal enterotoxins (SEs), which are secretory proteins that interact with antigen-presenting cells and T lymphocytes to induce cellular proliferation and a high level of cytokine expression (Larkin et al. 2009). Five classical SEs (SEA–SEE) comprise the major determinant toxins responsible for staphylococcal food poisoning in humans worldwide (Le Loir et al. 2003). Some novel SEs, including SEG, SEH, SEI, SEP, SER, SES and SET, can cause vomiting in primates (Le Loir et al. 2003; Ono et al. 2008); the enterotoxin gene cluster (egc) that contains several SE genes (seg, sei, sem, sen and seo) is known to be a widely distributed mobile genetic element (Jarraud et al. 2001). However, the role of these new SEs in food poisoning has not been well characterized. The major harmful effects of Staph. aureus in humans are SE-induced staphylococcal food poisoning and the consequences of MRSA infection (Larkin et al. 2009). The distributions of classical and novel SEs vary depending on origin and location (Lämmler et al. 2000; Omoe et al. 2002). Therefore, studies on the distributions of Staph. aureus isolates and the phylogenetic relationships among different enterotoxins are necessary.

Staphylocoagulase (SC), an extracellular protein of Staph. aureus encoded by the coa gene, is a virulent factor that causes plasma coagulation and is a major phenotypic determinant of Staph. aureus (Sakai et al. 2008). SC typing has been mentioned as a simple and rapid epidemiological tool for assessing Staph. aureus infection (Sakai et al. 2008). Most staphylococcal food poisoning is caused by serotype VII, whereas serotype II is the predominant MRSA strain in hospitals in Japan (Shimizu et al. 2000; Ishino et al. 2007). As of yet, neither the SEs nor the SC genotypes in Staph. aureus from healthy companion animals and their owners have been investigated together. In addition, the relationships among antibiogram, SC, SCCmec and SE in Staph. aureus of different origins among pets and their owners remain unclear.

Polymerase chain reactions (PCR) have been proven to be a useful tool for the rapid detection of Staph. aureus, MRSA, SC, enterotoxins and related genes (se) (Monday and Bohach 1999; Sakai et al. 2008). Pulsed-field gel electrophoresis (PFGE) is recommended for genotypic analysis of the whole genome of Staph. aureus strains of different origins, and multilocus sequence typing (MLST) has also been widely applied for population genetics of MRSA (Enright et al. 2000; Bonness et al. 2008; Cuny et al. 2010). Thus, the aims of this study were to identify the classical SEs and the 20 novel se genes, as well as antimicrobial susceptibility of Staph. aureus isolates collected from healthy companion animals and their owners, and to analyse the molecular phylogenetic relationship between the genotypes of enterotoxin, SC, SCCmec and MLST of the isolates according to PFGE patterns.

Material and methods

Sample collections

This study was conducted between July 2008 and November 2009. A total of 1563 nasal swabs were collected from companion animals (dogs and cats) and their owners. None of the participants had reported Staph. aureus infection. The samples were mainly collected from the National Taiwan University Veterinary Hospital in Taipei City and several private veterinary clinics in Taoyuan County (neighbour to Taipei City). The study was approved by the National Taiwan University Hospital Research Ethics Committee. The participants were asked to fill in a consent form, including a questionnaire, before sampling.

Bacterial isolation and identification

Sampling and isolation of Staph. aureus and MRSA from nasal swabs were performed as described previously (Hsieh et al. 2008). Briefly, Staph. aureus screening swabs were obtained by inserting an 11-cm-long cotton into the nares, with slight rotation to touch the wall of the nostril; swabs were then placed into 3-ml tryptic soy broth (TSB) transport medium with 6% NaCl (Difco, Sparks, MD, USA). After sampling, the swabs were shipped at 4°C for further isolation within 4 h. The sample of TSB was inoculated with an inoculation loop individually in brain heart infusion broth (BHIB; Difco, Detroit, MI, USA) at 37°C for 24 h for enrichment and then plated on Baird-Parker (BP) agar (Difco). The BP agar was incubated at 37°C for 48 h. The suspected colonies were identified as Staph. aureus by colony morphology, gram-staining and the tube coagulase test supplemented with rabbit plasma (Becton & Dickinson, Sparks, MD, USA). The identity of the isolates was confirmed by specific PCR primers of the nuc (thermonuclease) gene for the Staph. aureus and mecA (PBP2a) gene for MRSA. The primer sets for nuc and mecA were primer 1 (5′-GCGATTGATGGTGATACGGTT-3′) and primer 2 (5′-AGCCAAGCCTTGACGAACTAAAGC-3′) and MecA1 (5′-CTCAGGTACTGCTATCCACC-3′) and MecA2 (5′-CACTTGGTATATCTTCACC-3′), respectively (Brakstad et al. 1992; Sakoulas et al. 2001). The mecA-positive isolate was further confirmed by culturing on a selective agar containing 6 μg ml−1 oxacillin with 4% NaCl (CLSI, 2007). The Staph. aureus isolates and MRSA isolates were subsequently stored at −80°C until use.

SC typing

All isolates were analysed for SC genotype I–VIII by multiplex PCR assay consisting of two panels of primers (A and B) as described by Sakai et al. (2008). Type III, IV, VII and VIII were identified in Panel A, and type I, II, V and VI were identified in Panel B. The femA gene was used as a specific internal positive control for detecting Staph. aureus. Types IX and X were detected following the description of Hirose et al. (2010).

SCCmec typing and MLST

For the mecA-positive isolates, SCCmec typing was performed by the multiplex PCR approach (Milheiriço et al. 2007; Higuchi et al. 2008; Chen et al. 2009), and the SCCmec type I–VIII was determined according to the mec class, the J region and the ccr element. MLST was performed for MRSA isolates according to the protocol described by Enright et al. (2000). The results of sequence types (ST) were assigned using the Staph. aureus database from the MLST website (http://www.mlst.net).

Detection of classical SEs by RPLA agglutination assay

The presences of the classical SEs A, B, C, D and E were analysed by SET-reversed passive latex agglutination using a commercial detection kit (SET-RPLA®; Denka Seiken, Tokyo, Japan). Agglutination was observed using transmitted light through the bottom of the plate after 16–18 h of incubation.

PCR assays for the detection of 20 se genes

Twenty se genes were detected according to six published panel protocols. Panel I for sea, seb, sec, sed and see followed Wang et al. (2002); panel II for seg, seh, sei and sej followed Monday and Bohach (1999); panel III for sel, sem, sen, seo and seq followed Smyth et al. (2005); panel IV for sek followed Smyth et al. (2005); panel V for sep, ser and seu followed Chiang et al. (2008) and panel VI for ses and set followed Ono et al. (2008). The reference strains Staph. aureus ATCC 13565 (harbouring the sea gene), ATCC 19095 (harbouring the sec, seh, seg and sei genes), ATCC 23235 (harbouring the sed, seg, sei and sej genes), ATCC 14458 (harbouring the seb gene) and ATCC 27664 (harbouring the see gene) were included as positive controls for PCR analysis.

PFGE analysis

PFGE was performed as described by Hsieh et al. (2008). Briefly, all isolates were grown in 5 ml of BHIB for 24 h at 37°C. SmaI (New England Biolabs Inc., Beverly, MA) was used for the digestion of genomic DNA. Electrophoresis was performed with the CHEF-DR III system (Bio-Rad Laboratories, Hemel Hempstead, UK) through a 1% SeaKem Gold agarose gel (Bio-Rad) under the following conditions: an initial switch time of 5 s to a final switch time of 40 s, a running time of 21 h, a gradient of 6 V cm−1 and a temperature of 14°C. The gels were stained with ethidium bromide (0·5 μg ml−1), and the images were acquired in the TIFF format. The band patterns were analysed using the BioNumerics software (Applied Maths NV, Sint-Martens-Latem, Belgium). Similarity values were computed using the Dice coefficient, and clustering was performed by the unweighted pair group method with arithmetic mean (McDougal et al. 2003). The phylogenetic dendrogram patterns were compared among the SC, SE and SCCmec types.

Antimicrobial susceptibility testing

Antimicrobial susceptibility testing was performed in Mueller-Hinton agar using agar dilution according to the M7-MIC method (CLSI 2007). Resistance and susceptibility breakpoints were tested using the following antimicrobials: cefoxitin (32 and 8 μg ml−1), chloramphenicol (32 and 8 μg ml−1), clindamycin (4 and 0·5 μg ml−1), erythromycin (8 and 0·5 μg ml−1), gentamicin (16 and 4 μg ml−1), oxacillin (4 and 2 μg ml−1), rifampin (4 and 1 μg ml−1), tetracycline (16 and 4 μg ml−1) and vancomycin (4 μg ml−1). The Staph. aureus ATCC 29213 was used for quality control, and the results were defined as susceptible (S), intermediate (I) and resistant (R) according to CLSI criteria (CLSI 2007). Multidrug resistance (MDR) was defined as simultaneous resistance to at least two antimicrobials used in this study.

Data analysis

The ratios of MSSA and MRSA isolates between owners and pets, differences in the PFGE clusters and participant information were assessed using the chi-square test. P values of <0·05 were considered statistically significant.


Prevalence of Staphylococcus aureus and MRSA

Overall, 787 human subjects and 776 dogs and cats were sampled for Staph. aureus. A total of 114 Staph. aureus isolates were recovered and 94 of them (82·5%, 94/114) were from human samples (Table 1). Twenty-three (20·2%) isolates were MRSAs and the rest were MSSAs (Table 1). Geographical distribution showed that area A (Taipei City) had the highest isolation rate of Staph. aureus (both from human and pet isolates) among the sampled areas; the remaining 16 (14%, 16/114) isolates, including three MRSA isolates, had no accompanying geographical information, and thus, they were grouped as area D (Table 1). The isolation rate of MRSA in Taoyuan County (45·5%, 5/11) was significantly higher than the rates in New Taipei City (17·9%, 7/39) and Taipei City (16·7%, 8/48) (P  < 0·05).

Table 1.   The origin of samples and prevalence
Urban areasNo. of samplesStaphylococcus aureus positive (%)MRSA positive (%)Total (%)
  1. A, Taipei City; B, New Taipei City; C, Taoyuan County; D, background unavailable; MRSA, methicillin-resistant Staph. aureus.

AHuman29639 (13·2)7 (17·9)48 (42·1)
Pet2929 (3·1)1 (11·1)
BHuman28533 (11·6)7 (21·2)39 (34·2)
Pet2806 (2·1)0 (0)
CHuman11410 (8·8)5 (50)11 (9·6)
Pet1131 (0·9)0 (0)
DHuman9212 (13·0)3 (25)16 (14)
Pet914 (4·4)0 (0)
Total 156311423 

Genotyping of Staphylococcus aureus, and SCCmec typing and MLST of MRSA isolates

All 114 Staph. aureus isolates were rendered typeable following digestion with restriction enzyme SmaI, and 97 pulsotypes (PTs) were identified. The PTs were further grouped into six clusters (1–6) in a phylogenetic dendrogram generated by the BioNumerics software (Fig. 1). Majority of the MRSA isolates were grouped in cluster 5 (73·9%, 17/23) and the remaining six isolates were grouped in clusters 1 (n = 3), 2 (n = 1) and 6 (n = 2) (Fig. 1). Among the 23 MRSA isolates, only the SCCmec types IV (43·5%, n = 10) and V (34·8%, n = 8) were found and the remaining five isolates were nontypeable (NT). The MLST results showed that ST59-MRSA (85%, 17/20), including one pet isolate (41K39), was the most dominant clone. The three remaining isolates were ST239-MRSA-V (31H75), ST241-MRSA-V (32H15) and ST89-MRSA-NT (41H35).

Figure 1.

 Phylogenetic dendrogram of 114 analysed Staphylococcus aureus isolates was typed by Pulsed-field gel electrophoresis (PFGE). Six clusters (1–6) were grouped according to their enterotoxin genotypes and phylogenetic relationships by BioNumerics software. NT: nontypeable; Cef, cefoxitin; Chl, chloramphenicol; Cli, clindamycin; Ery, erythromycin; Gen, gentamicin; Oxa, oxacillin; Tet, tetracycline; Rif, rifampin. *H: human isolates; D, K, C and F: pet isolates.

Prevalence of enterotoxin genes and classical enterotoxins, SEA–SEE, in MRSA

Seventy-four isolates (64·9%, 74/114) harboured more than one se gene, and a total of 15 genotypes of enterotoxin were recovered (all except see, sej, ser, ses and set) (Fig. 1). The major genotypes of enterotoxin recovered were seb (41·9%, 31/74), followed by sek (40·5%, 30/74) and seq (40·5%, 30/74) (Table 2). Human isolates contained more enterotoxin genes (70·2%, 66/94) compared with pet isolates (40%, 8/20) (< 0·05). MRSA isolates (91·3%, 21/23) harboured more enterotoxin genes than MSSA isolates (58·2%, 53/91) (< 0·01) (Table 2). The profiles seb-sek-seq (42·9%, 9/21) and seb-sek-seq-sep (28·6%, 6/21) were the most commonly found in MRSA isolates, and those of sep (18·9%, 10/53), seb-sek-seq (17%, 9/53) and sea (15·1%, 8/53) were the most frequently found in MSSA isolates (Table 2). Pet isolates harboured genotypic enterotoxins were found sporadically. Twelve of the MRSA isolates were characterized by RPLA as containing enterotoxin type A (two isolates) and type B (10 isolates), corresponding to enterotoxin genes sea and seb, respectively (Fig. 1). However, the other six MRSA isolates containing classical enterotoxin genotypes (five seb and one sed) showed no detectable RPLA enterotoxins.

Table 2.   Distribution of se genes and staphylocoagulase (SC) genotypes in MRSA and MSSA isolates
No. of isolatesse gene*SC genotype (no. of isolates)Origins
  1. NT, Nontypeable; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-sensitive Staph. aureus.

  2. *Bold type = commonly associated with food poisoning.

 2Not detectedIV (1), NT (1)2 
 2akqIV (1), VII (1)2 
 9bkqII (1), VII (4), NT (4)9 
 6bkqpII (2), VII (4)51
 1dNT (1)1 
 1gimnoNT (1)1 
 1moI (1)1 
 1pV (1)1 
 38Not detectedIII (2), IV(2), V (14), VI (5), X (10), NT (5)2612
 8aIV (3), II (2), III (1), NT (2)8 
 1akqVII (1)1 
 2ahkqX (1), VII (1)2 
 3agmnouIV (3)3 
 5bIV (1), V (3), NT (1)41
 9bkqVII (6), II (2), NT (1)81
 1bkqpIII (1)1 
 1bimnouV (1)1 
 1chlII (1)1 
 2cgilmnouVII (2)2 
 1cilmnouVII (1)1 
 1dNT (1) 1
 1giNT (1) 1
 3gimnouII(1), VIII (2)21
 2hX (1), II (1)2 
 1mnouVIII (1) 1
 10pII (1), VII (2), III (6), NT (1)91
 1uNT (1)1 
 114  9420

Prevalence and distribution of SC typing

Ninety-four of the 114 Staph. aureus isolates, including 16 MRSA and 78 MSSA, could be typed for SC; the rest were NT (Table 2). The predominant SC types, VII (23·4%, 22/94) and V (20·2%, 19/94), made up 43·6% of the typeable isolates, whereas other SC types (I–IV, VI, VIII and X) were found sporadically. The SC types V and VII were the dominant typeable ones (50%, 7/14) of the pet isolates.

Antimicrobial susceptibility

Of the 114 Staph. aureus isolates, 32·5% (37/114) were resistant to erythromycin, and more than 70% of the isolates were susceptible to other antimicrobials. All 114 isolates were susceptible to vancomycin (Table 3). Human isolates had a higher percentage of antimicrobial resistance than did the pet isolates (P  < 0·01) (Table 3). The MRSA isolates were all MDR and expressed more than 95% resistance to oxacillin, clindamycin and erythromycin, followed by more than 50% resistance to chloramphenicol and tetracycline (Table 3). MSSA isolates were uniformly susceptible to oxacillin, cefoxitin and gentamicin. However, 42·8% (39/91) of MSSA isolates were resistant to a single antimicrobial and 8·8% (8/91) of MSSA isolates were MDR (Fig. 1).

Table 3.   Antimicrogram of 114 Staphylococcus aureus isolates categorized by human and pet, and MRSA and MSSA
AntimicrobialHuman (n = 94)Pet (n = 20)MRSA (n = 23)MSSA (n = 91)
  1. MRSA, methicillin-resistant Staph. aureus; MSSA, methicillin-sensitive Staph. aureus.

  2. *Data given in percentages in each category.


The correlation between the SC genotype, SE genes, MLST, antibiogram and genotypes from PFGE

Thirteen of 19 isolates (68·4%) in cluster one harboured se genes, and a high prevalence of the egc-related genes was detected (69·2%, 9/13) (Fig. 1). Most of the predominant SC type VII isolates, including seven ST-59-MRSA-IV/V isolates, were grouped in cluster 5 (68·2%, 15/22), and a high percentage of the isolates carried se genotype profiles [seb-sek-seq (73·3%, 11/15), seb-sek-seq-sep (26·7%, 4/15)]. These isolates were MDR and showed relatively high resistance to erythromycin (80%, 12/15), chloramphenicol (73·3%, 11/15) and oxacillin (60%, 9/15) (Fig. 1). Eighty percent of the MDR isolates (8/10) of MSSA were also grouped in cluster 5 (Fig. 1). By contrast, clusters 2 (major SC type X) and 4 (major SC type V) had significantly low detectable levels of se genes than the other clusters (< 0·05).


In this study, we collected 20 and 94 Staph. aureus isolates from healthy companion animals and their owners in the urban areas of northern Taiwan, respectively. Wertheim et al. (2005) reported that 20% (range 12–30%) of humans are persistent Staph. aureus nasal carriers. Our results showed that Staph. aureus was detected in 12% of human samples, similar to Wertheim’s report. On the other hand, Staph. aureus was detected only in 3% of our pet samples. These results agreed with the report of Boost et al. (2008) that the carriage of Staph. aureus is <10% in dog’s nares. However, the isolation rate of MRSA from pet isolates (5%, 1/20) was higher than those in other studies: 0·5–2% in Brazil, Hong Kong and UK (Lilenbaum et al. 1998; Boost et al. 2008; Loeffler et al. 2010). The results showed that 64·9% of Staph. aureus isolates harboured se genes, similar to the 67·9% of isolates reported by Nashev et al. (2007). The human origins of se (70·2%) were much lower than the se prevalence (>99%) reported by Varshney et al. (2009), who used a collection of clinically well-characterized isolates from wounds and blood of hospitalized patients. In addition, ser was reported to be the most prevalent se gene (82–96%) by Varshney et al. (2009), whereas none of our isolates contained this se gene. The most frequently observed se genotypes (seb, sek and seq) were associated with Staph. aureus pathogenicity islands 3, and thus, they were detected simultaneously in Staph. aureus strains (Subedi et al. 2007).

The most common se gene profile in our isolates from healthy humans was seb-sek-seq (19·1%). This profile is statistically different from the most commonly reported gene profile egc in studies in Germany (63·7%) (Nashev et al. 2007) and the Netherlands (73·4%) (van Belkum et al. 2006). These results indicate variability in the occurrence of se genotypes in different origins and geographical areas. According to the PFGE phylogenetic dendrogram, 83% of the isolates in cluster 5 were found to share the enterotoxin genotype combinations of seb-sek-seq or seb-sek-seq-sep. The toxin gene elements may act as the mobile fragment, but some enterotoxins might spread clonally (Mørk et al. 2010). Thus, our results support the varied distribution of SEs from different origins (Lämmler et al. 2000; Omoe et al. 2002; Nashev et al. 2007). For RPLA tests, 66·7% (12/18) of the MRSA isolates in this study harbouring the se genes (sea, seb and sed) were enterotoxin positive and the remaining six isolates were RPLA negative. The reasons for the absence of enterotoxin expression could be due to the level of toxin production being below the detection limit of the RPLA assay, or the absence of enterotoxin production during a specific time period or under certain conditions (Smith et al. 1983).

The predominant type of SC from our data was type VII (23·4%). Mizumachi et al. (2011) reported that type VII was responsible for the majority of the cases of food poisoning in Japan. Type II, the predominant MRSA strain in Japanese hospitals (Ishino et al. 2007), was also detected in our isolates; only 13·3% (3/23) of the isolates were identified as MRSA. The significance of these SC isolates is not clear and warrants further investigation. Moon et al. (2007) reported that there was no significant correlation between the SC genotypes and their specific SE types. However, a relatively high percentage of type VII isolates with seb-sek-seq or seb-sek-seq-sep genotypes was noted in the present study, which differed from the data reported by Mizumachi et al. (2011). This difference may be due to species specificity, and different countries may have unique prevalence patterns for SC and SE types (Omoe et al. 2002; Saei et al. 2009).

Human isolates in this study were resistant to erythromycin (37%), tetracycline (33%) and clindamycin (31%). These results are higher than the resistances to erythromycin (14%), tetracycline (2%) and clindamycin (6%) reported by Tavares et al. (2010) and the resistances to erythromycin (23%), tetracycline (16%) and clindamycin (6%) reported by Hamdan-Partida et al. (2010). These differences may be due to the variations in antibiotic use in different areas of the world (Agustín et al. 2005). Furthermore, some resistant isolates in this study were phylogenetically related. Therefore, we suspect that the gene elements conferring antimicrobial resistance are highly conserved among clones.

Molecular analysis showed that most of the se-carrying MRSA isolates (16/23) contained the type IV or type V SCCmec elements that are typical for CA-MRSA strains. Thus, we hypothesized that these strains might play a significant role in staphylococcal food poisoning in community settings. However, future studies are necessary to confirm this hypothesis. CA-MRSA is generally susceptible to non-β-lactam antibiotics (Chua et al. 2011). In this study, nevertheless, SCCmec IV and SCCmec V were present in all MDR phenotypes characterised by clindamycin and erythromycin resistance. ST239-MRSA and ST241-MRSA (a single-locus variant of ST239), which are resistant to seven antimicrobials, displayed high MDR and were the predominant HA-MRSA clones in Taiwan and were also detected in the present study; they belonged to SCCmec V, but not SCCmec III. These results agree with the suggestion by Borghi et al. (2010) that the barrier between CA-MRSA and HA-MRSA starts to blur.

Upon combining the results from the PFGE phylogenetic dendrogram patterns and the antimicrobial susceptibility tests, ST59-MRSA-IV/V (belonging to SC VII) from cluster 5 had the highest percentage of MDR isolates. In Taiwan, ST59-MRSA-IV/V is the predominant clone of CA-MRSA (Wang et al. 2009). Our study found that approx. 100% of ST59-MRSA-IV/V isolates were resistant to clindamycin and erythromycin, similar to a previous report by Wang et al. (2009). This finding suggests that high levels of antimicrobial resistance and se-carrying clones of ST59-MRSA-IV/V may become a potential hazard in a community setting. Furthermore, two isolates (12H9 and 12C9) from a dog and its owner having the same PTs and se genes (Fig. 1) indicated that cross-transmission of this clone between owners and their companion animals might be possible.

Our results showed the clustering characteristics of the Staph. aureus enterotoxin genes, SC and antimicrobial resistance in various regions. The clone of se-carrying-MDR-ST-59-IV/V-SC VII was identified predominantly in the present study. Thus, the potential threat of this clone in a community setting should not be ignored. Future studies are needed to assess the risk factors of this clone and its impact on humans and their pets.


The authors acknowledge the Animal Hospital of the National Taiwan University for supplying samples and also the National Health Research Institutes for PFGE assistance.