Staphylococcus (S.) intermedius and S. pseudintermedius are both Staphylococcus intermedius group (SIG) members. These microorganisms are opportunistic pathogens that cause skin and nosocomial infections in dogs and cats [32]. S. pseudintermedius is the predominant SIG species in Korea and until recently was misclassified as S. intermedius [30,33]. These pathogens rarely cause human disease although they do have zoonotic potential [5,6,26]. Increased identification of methicillin-resistant SIG (MRSIG) isolates has emphasized the importance of public health precautions for veterinary staff and pet owners [7,11,26].

Many current studies have evaluated methicillin-resistant characteristics and the staphylococcal cassette chromosome mec (SCCmec) of SIG isolates, particularly S. pseudintermedius. Although type I to type VIII SCCmec types have been identified, there is no significant studies have been conducted to examine SCCmec types isolated from veterinary hospitals located nationwide in Korea. SCCmec types may produce staphylococcal exotoxins including staphylococcal enterotoxins (SEs), exfoliative toxins (ETs) including S. intermedius exfoliative toxin (SIET), and toxic shock syndrome toxin 1 (TSST 1); these toxins are associated with atopic dermatitis, mastitis, food poisoning, canine pyoderma, and chronic otits [1,3,10,14,29]. SIET (encoded by the siet gene), first described by Terauchi et al. [27], plays a potential role in the pathogenesis of canine pyoderma and chronic otitis [31]. There may also be a greater chance of SIG isolates to colonization and/or expand in veterinary staff, companion animals, and hospital environments. Therefore, this study is focused on the genetic identification of staphylococcal exotoxins, SCCmec types, and genetic relatedness by pulsed-field gel electrophoresis (PFGE) of the SIG isolates in Korea.

Samples of S. pseudintermedius (n = 167) and S. intermedius (n = 11) were isolated; their identity was confirmed by Gram-staining and biochemical testing such as coagulase, DNase production and hydrolysis [33]. Polymerase chain reaction (PCR) was carried out with primers targeting S. intermedius nuclease and 16S rRNA genes [2]. To differentiate S. pseudintermedius from S. intermedius and S. delphini, PCR-restriction fragment length polymorphism analysis was performed using pta gene-specific primers (pta_f1:5'-AAA GAC AAA CTT TCA GGT AA-3', and pta_r1: 5'-GCA TAA ACA AGC ATT GTA CCG-3'), and the restriction enzyme MboI (New England Biolabs, USA) [3]. Samples were collected from eight veterinary hospitals from four different regions in Korea between November 2006 and January 2010 as previously described [33]. PCR analysis to detect the methicillin-resistance gene (mecA) was performed with gene-specific primers as previously described [32]. PCR was also performed to amplify genes encoding SEA (sea), SEB (seb), SEC (sec), SED (sed), SEE (see), SEG (seg), SEH (seh), SEI (sei), TSST 1(tsst-1) and ETs (eta, etb, etd, and siet) using primers and PCR conditions used in other previous studies [12,16,19,20,24,25]. FRI913, FRI 361, FRI 472, FRI 569, MNHOCH, and RN4220 strains were used as either positive or negative controls for superantigen genes. A superantigen gene obtained in this study was further analyzed by DNA sequencing using the Vector NTI Align X program (Invitrogen, USA).

All MRSIG isolates including 49 methicillin-resistant S. pseudintermedius (MRSP) and three methicillin-resistant S. intermedius (MRSI) were subsequently classified as SCCmec type I to VIII by either single or multiplex PCR assays using protocols in other previous studies including primers and PCR conditions [4,9,13,15,17,18]. Genomic DNA plug kits (Bio-Rad, USA) with SmaI (New England Biolabs, USA) digestion were used for the PFGE analysis of 39 S. pseudintermedius isolates from a private referral veterinary hospital collected at different times (eight isolates from November 2006, 24 isolates from April 2008, four isolates from June 2009, and three isolates from October 2009) and six S. intermedius isolates (all collected during April 2008). PFGE was performed according to protocol used in previous study [8] and the manufacture's instruction (Bio-Rad, USA) except for an initial pulse time of 5 sec and final time of 40 sec for 21 h [8].

Analysis of the SIG isolates (n = 178) for SEs and TSST 1 genes showed that only a single S. pseudintermedius isolate from a veterinary staff member (H1-23) harbored the enterotoxin C (sec) gene (Fig. 1). Subsequent DNA sequencing analysis (Fig. 2) revealed that the amplified sec gene the canine type C SE gene (sec_caine). A total of 166 out of 178 SIG isolates (155 S. pseudintermedius and 11 S. intermedius) isolated from veterinary staff members (n = 38, 95%), companion animals (n = 107, 93%), and veterinary hospitals (n = 21, 91.3%) harbored siet originally detected in S. intermedius. However, other ET genes such as eta, etb, and etd, known to originate from S. aureus, were not identified.

SCCmec typing of the 49 MRSP isolates identified 38 type V isolates, one type IV isolate, and 10 non-identifiable isolates. Three MRSI isolates were characterized as type V isolates (Table 1). PFGE analysis showed that SIG isolates from a private referral animal hospital (collected between 2006 and 2009) recovered from veterinary staff, companion animals, and the environment had the same band patterns (Fig. 3). Six S. pseudintermedius isolates (two from companion animals collected in September 2006, three from veterinary staff members, and one from companion animals collected in April 2008) and two S. pseudintermedius isolates (one from a veterinary hospital environment isolated in April 2008, and one from a companion animal isolated in June 2009) showed the same band patterns (Fig. 3).

Since SIG is closely related to S. aureus, studies have been performed to determine whether SIG has adapted the eta, etb and etd toxins from S. aureus by PCR test with specific primers [16,24,25]. None of these toxins were detected. However, 166 out of 178 (93.3%) SIG isolates harbored the siet toxin originating from S. intermedius [27]. This result and those of other studies [22,31] imply that the majority of SIG isolates harbor the siet gene. Although the siet gene was present in 93.3% of the SIG isolates (from 108 dogs and two cats) in this study, only 14 dogs had a history of various skin disease including allergy and prolonged inflammation lesion in skin (n = 13) or otitis (n = 1) (data not shown). Therefore, other factors such as the general health of an animal and existence of other SIG virulence factors may play an important role in outbreaks of various kinds of skin disease or otitis.

The sec gene was detected in a single isolate in the present study, which was identified as sec_canine by DNA sequencing. However, this isolate was isolated from a veterinary staff member, and no additional SEC_canine isolates were identified in the veterinary hospital where this individual worked. This suggested that the isolate might be transmitted from an in- or outgoing companion animal with which the veterinary staff was in contact. The low incidence of toxins in this study could be secondary to the small number of isolates collected from companion animals that a history of skin disease or otitis. Only 11.3% of S. pseudintermedius in a previous study had exotoxins although all samples were taken from patients diagnosed with pyoderma or chronic otitis and referred to a veterinary teaching hospital [30]. This result suggested that SE toxins may be associated with pyoderma and chronic otitis.

A total of 52 out of 178 SIG isolates harbored the mecA gene. Although the majority of SCCmec types were type V (78.9%), one isolate was type IV. In a previous study [10], 23 isolates (85.2%), three isolates (11.1%) and one isolate (3.7%) from veterinarians, staff, students, companion animals and environment in the veterinary hospitals were determined to be an MRSP hybrid SCCmec type I~II, type V, or non-identifiable, respectively. A previous European and North American study [21] identified 75 hybrid SCCmec type II~III isolates (72.8%), two type III isolates (1.9%), six type IV isolates (5.8%), 14 type V isolates (13.6%), four type VII isolates (3.9%), and two non-identifiable isolates (1.9%) from diseased and healthy dogs in veterinary diagnostic laboratories of different countries. This demonstrated that the majority of MRSIG isolates in Korea harbor the SCCmec type V whereas the hybrid type II~III is the main SCCmec type found in veterinary hospitals in Japan, Europe, and North America.

In the present study, PFGE analysis of the 39 S. pseudintermedius and six S. intermedius isolates from a private referral veterinary hospital (collected during November 2006, April 2008, June 2009, and October 2009) showed that 20 isolates (lines 1~6, lines 7~9, lines 10~15, lines 16~18, and lines 19~20) had the same band patterns. Moreover, some isolates obtained on different sampling dates showed the same band patterns (six isolates from lines 12~17, and two isolates from lines 21~22). These results suggest potential contamination or expansion of S. pseudintermedius and S. intermedus isolates among veterinary staff, companion animals, and veterinary hospital environments, and colonization by these specific strains for more than 13 or 18 months in the same hospital.

In conclusion, eta, etb, etd genes were not detected but siet toxin was found in 166 isolates in the current study. PFGE analysis results of isolated from H animal hospital showed that S. pseudintermedius isolates collected in over a period of 13 and 18 months from veterinary staff, companion animals, and the hospital environment had the same band patterns. S. pseudintermedius infections in humans [5,26,28], spread of MRSP populations [6,23,29], and association of SIG with canine pyoderma and chronic otitis [3,14,29] has been previously reported. Therefore, SIG, especially MRSIG, may have significant clinical implications for companion animals with skin infections or chronic otitis that is of concern for veterinary staff, companion animal owners, and healthy companion animals.