Novel Lineage of Methicillin-Resistant Staphylococcus aureus, Hong Kong

To determine whether spa type of methicillin-resistant Staphylococcus aureus in pigs belonged to sequence type (ST) 398, we analyzed nasal swabs from pig carcasses at Hong Kong markets in 2008. ST9 belonging to spa type t899 was found for 16/100 samples, which indicates that a distinct lineage has emerged in pigs.

M ethicillin-resistant Staphylococcus aureus (MRSA) has long been recognized as an important hospital pathogen and in recent years has emerged in the community. Increasing numbers of reports have concerned MRSA in animals. Multilocus sequence typing (MLST) has shown widespread dissemination of sequence type (ST) 398 among pigs in the Netherlands (1). This MRSA lineage has subsequently been reported in several other countries and animal species. Compelling microbiologic and epidemiologic evidence indicates that persons living or working on farms, especially pig farms, have an increased risk for colonization or infection with ST398 (2). In the present study, MRSA isolates obtained from slaughtered pigs in Hong Kong were characterized genotypically and compared with ST398.

The Study
Nasal swab specimens collected by using Transwabs (Medical Wire Ltd, Corsham, UK) were collected from 100 pig carcasses at 2 wet markets in Hong Kong on 5 separate days over a 7-week period in 2008. Cross-contamination was minimized by selecting carcasses with intact nasopharyngeal tracts and by instructing the butchers taking part in the project to avoid causing damage to the nares when cutting up the heads. The frontal section of each snout was cleaned with 75% alcohol before swabbing the nasal mucosa up to 8 cm into the nares. Nasal swabs were enriched in brain-heart infusion broth (Oxoid Ltd, Basingstoke, UK) with 7% NaCl at 37 o C for 48 h and then injected into man-nitol salt agar (Oxoid) supplemented with 6 μg/mL oxacillin, for 24 h. Presumptive S. aureus colonies were tested for heat-stable nuclease (DNase) and coagulase production, and isolates positive for both were confi rmed to the species level by latex agglutination (Staphaurex Plus, Murex Diagnostics Ltd, Dartford, UK). Antimicrobial drug sensitivity testing was performed by using disk diffusion following Clinical and Laboratory Standards Institute recommendations (3). Methicillin resistance was confi rmed by disk diffusion using cefoxitin (30 μg) and mecA PCR detection. All MRSA isolates were characterized by pulsed-fi eld gel electrophoresis (PFGE) by using SmaI (4), staphylococcal chromosome cassette (SCC) mec typing (5), and spa typing (6) with Ridom StaphType 1.4.1 software (www.ridom.de/ staphtype). Two isolates representative of distinct PFGE patterns and SCCmec types were analyzed by MLST (7), and the remaining isolates were characterized by singlelocus (aroE) sequencing.
MRSA was isolated from 16 samples collected on 4 of the 5 sampling days. In contrast to ST398, which has the characteristic of being nontypeable by PFGE using SmaI (1,2), the 16 MRSA isolates were typeable. They displayed 6 PFGE patterns; 2 predominant types (A1 and B1) were associated with SCCmec types IV and V, respectively (Table  1). Both PFGE types were ST9 according to MLST analysis. The 4 remaining patterns were either closely related to A1 (A2, A3, and A4) or possibly related to B1 (B2) according to the criteria of Tenover et al. (8). All isolates belonged to spa type t899 and harbored the ST9-associated aroE allele 3, which differs from that in ST398 (allele 35) by multiple mutations. Porcine MRSA ST9 isolates were negative for Panton-Valentine leukocidin genes and resistant to a broader range of antimicrobial agents than that previously described for MRSA ST398 isolated from pigs in the Netherlands (1). Twelve isolates displayed a typical multiple resistance pattern, including resistance to chloramphenicol, ciprofl oxacin, clindamycin, cotrimoxazole, erythromycin, gentamicin, and tetracyline. The remaining 4 isolates were additionally resistant to fusidic acid ( Table 1). All isolates were negative for Panton-Valentine leukocidin and susceptible to vancomycin and linezolid.
A search of the scientifi c literature and the Internet for information about the frequency of S. aureus ST9 in humans and animals indicated that ST9 is a clone of porcine origin. In 2005, Armand-Lefevre et al. (9) reported that ST9 was the most prevalent ST of methicillin-susceptible S. aureus (MSSA) isolated from pig farmers and infected pigs in France but not from a control group of persons without occupational contact with pigs. In 2007, an erythromycinresistant MSSA ST9 clone belonging to spa type 337 was found to be endemic on a farm in Denmark (10). A clinical ST9 isolate of porcine origin carrying the multidrug resistance gene cfr, associated with linezolid resistance, has  (12). Eighteen spa-type t899 isolates have been submitted to the Ridom SpaServer (http://spaserver2.ridom.de): 10 from Germany, 7 from the Netherlands, and 1 from Belgium. All submissions were recorded as MRSA, but unfortunately, the origins of the isolates and the MLST types were not reported. Although MRSA t899 has been previously associated with ST398 isolates from pigs (13) and from participants at a conference on pig health (14), the repeat succession of this spa type is completely different from those of other ST398related spa types and similar to those of spa types related to ST9 (Table 2). That the same spa type occurs in both ST9 and ST398 is surprising because the MLST allelic sequences of these 2 S. aureus lineages are unrelated (Table  2). However, the occurrence of the same spa types in dis-tant lineages has been previously reported (15) and could have resulted from either convergent evolution or genetic recombination.

Conclusions
Our study of MRSA colonization of commercial pigs in Asia provides evidence that methicillin resistance has emerged in a porcine S. aureus lineage other than ST398. It appears that ST9 has achieved methicillin resistance through multiple acquisitions of SCCmec, as indicated by the recovery of distinct PFGE and SCCmec types. A combination of the results of literature and database searches indicates that ST9 is associated with pig farming and, although it is found infrequently, this ST has been isolated from infected persons worldwide.
Several studies have previously investigated the prevalence of MRSA nasal carriage in pigs sampled immediately after slaughter (2) or at the farm of origin (13). In our study, samples were collected at wet markets, because it not possible to access pigs at the single slaughterhouse in Hong Kong or at the farm sites of origin because >90% of slaugh-  Feb 22  3  t899  IVb  A1  3  ST9  G29  Mar 11  4  t899  IVb  A2  3  ST9 ‡  55  Mar 27  1  t899  IVb  A1  3  ST9 ‡  56  1  t899  V  B1  3  ST9  57  1   ter pigs are raised in mainland China and delivered by train directly to the slaughterhouse in Hong Kong. Notably, the previously reported human infection in China with MRSA-ST9 occurred in Guangzhou, the province closest to Hong Kong, where most of the pigs originate. The colonization rate determined in our study represents the level of contamination immediately prior to sale of pig meat to consumers. Although pig heads are rarely available in European and North American markets, because these parts of the animal are generally centrally processed, homemade soup using the pig's nose is commonly consumed in Hong Kong; this gastronomic tradition may increase the risk for zoonotic transmission of MRSA. Further epidemiologic studies are needed to determine the rates of colonization and infection with MRSA and MSSA ST9 both in personnel exposed to pigs and in the community.