Staphylococcus aureus Host Range and Human-Bovine Host Shift

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Sept. 2011, p Vol. 77, No /11/$12.00 doi: /aem Copyright 2011, American Society for Microbiology. All Rights Reserved. Staphylococcus aureus Host Range and Human-Bovine Host Shift Olga Sakwinska, 1 * Marlyse Giddey, 1 Martine Moreillon, 1 Delphine Morisset, 1 Andreas Waldvogel, 2 and Philippe Moreillon 1 Department of Fundamental Microbiology, University of Lausanne, Lausanne, Switzerland, 1 and Institute Galli-Valerio, Lausanne, Switzerland 2 Received 2 February 2011/Accepted 23 June 2011 Staphylococcus aureus is a major agent of bovine mastitis. The concomitant emergence of pig-associated methicillin-resistant S. aureus (MRSA) in human carriage and infection requires a reexamination of the host range and specificity of human- and cow-associated S. aureus strains, something which has not been systematically studied previously. The genetic relatedness of 500 S. aureus isolates from bovine mastitis cases, 57 isolates from nasal carriage of farmers, and 133 isolates from nonfarmers was determined by amplified fragment length polymorphism (AFLP) analysis and spa typing. Multilocus sequence typing (MLST) was conducted on a subset of isolates to match AFLP clusters with MLST clonal complexes (CCs). This data set allowed us to study host range and host specificity and to estimate the extent of bovine-tohuman transmission. The genotype compositions of S. aureus isolates from farmers and nonfarmers were very similar, while the mastitis isolates were quite distinct. Overall, transmission was low, but specific genotypes did show increased cow-to-human transmission. Unexpectedly, more than one-third of mastitis isolates belonged to CC8, a lineage which has not been considered to be bovine mastitis associated, but it is well known from human carriage and infection (i.e., USA300). Despite the fact that we did detect some transmission of other genotypes from cows to farmers, no transmission of CC8 isolates to farmers was detected, except for one tentative case. This was despite the close genetic relatedness of mastitis CC8 strains to nonfarmer carriage strains. These results suggest that the emergence of the new bovine-adapted genotype was due to a recent host shift from humans to cows concurrent with a loss of the ability to colonize humans. More broadly, our results indicate that host specificity is a lineage-specific trait that can rapidly evolve. The main niche and the largest reservoir of Staphylococcus aureus are human nares. On average, every third person is colonized with this facultative pathogen, making ca. 2 billion people colonized worldwide. Despite the fact that only a very small fraction of the carriers ever develop staphylococcal disease, its global burden is remarkable (27). Steadily increasing rates of colonization and infections with methicillin-resistant S. aureus (MRSA) are of special concern (20) because of the limited therapeutic options for such infections. Because of their sheer population size, farm animals constitute a potentially enormous reservoir of any pathogen. Even though the largest reservoir of S. aureus is human nares, the second largest may well be cows. The prevalence of S. aureus in bovine mastitis was estimated to be 3 to 5% (28, 34). Given the world s milking cow population of 1.5 billion (31), up to 75 million cows can be infected worldwide. The widespread colonization of pigs with a specific lineage of MRSA, ST398 (2), and emerging infections in humans (10, 22, 32) showed that the animal reservoir of. S. aureus can have potentially serious consequences for human health care. Typically, human lineages of S. aureus are rarely found in animals, suggesting host range barriers (41, 42, 48). However, * Corresponding author. Present address: Nestlé Research Center, Vers-chez-les-Blanc, 1000 Lausanne 26, Switzerland. Phone: Fax: olga.sakwinska@unil.ch. Supplemental material for this article may be found at Published ahead of print on 8 July the rapidly emerging colonization of pig farmers with pigassociated MRSA (45, 47) shows that some lineages of S. aureus have a broader host range. A recent description of a novel MRSA variant found in both humans and cows suggests that ST398 is not the only lineage with an extended host range (9). Moreover, the host range of a pathogen is a dynamically evolving trait. For example, Lowder and colleagues described a recent (ca. 30 years) host shift of a particular S. aureus lineage from humans to chickens. The host shift was followed by global spread in chickens (24). While several authors compared S. aureus strains isolated from humans to those isolated from cows (e.g., see references 41, 42, and 48), the host range of bovine and human S. aureus strains and the extent of transmission between both species have not been systematically studied. Dedicated sampling efforts with a focus on farmers, the most likely transmission hot spots, are required. The time and space frames should be well defined, because isolates from human carriage show substantial geographical diversification, on both a lineage scale (clonal complex [CC]) (6, 36, 37) and a finer scale (29). Finally, a range of genotyping methods has to be applied to assess the identities of very closely related genotypes, which can be achieved by, for example, spa typing, as well as to obtain phylogenetic information, which is best done by, for example, multilocus sequence typing (MLST). In the present work we analyzed the genetic relatedness among S. aureus isolates from cases of bovine mastitis, S. aureus isolates from the nasal carriage of animal caretakers of infected animals (farmers), and isolates from urban healthy 5908

2 VOL. 77, 2011 HUMAN-BOVINE HOST SHIFT IN STAPHYLOCOCCUS AUREUS 5909 volunteers (nonfarmers), collected from the same geographical area. MATERIALS AND METHODS Sample collection in cows. S. aureus was isolated from bovine milk samples received by a diagnostic veterinary laboratory at the Institute Galli-Valerio. The samples were collected in Western Switzerland between June 2007 and July All samples containing S. aureus received by the diagnostic laboratory during this period were included in the study. The samples were collected in cases of suspected clinical or subclinical mastitis by veterinarians or farmers according to standard procedures. Subclinical mastitis is not associated with clinical signs in animals, but the milk quality is affected. Briefly, the teats were cleaned thoroughly and disinfected. After a short forestripping, a sample of milk was collected into sterile polypropylene tubes and sent to the laboratory. Milk samples from 133 farms were analyzed. Standard diagnostic procedures were used to detect the presence of S. aureus. Briefly, 10 ml of milk sample was centrifuged for 5 min at 1,400 g, and the supernatant was discarded. A sample of the pellet was streaked onto sheep blood agar and bromothymol blue lactose agar. S. aureus was identified by the presence of double hemolysis. If alpha-hemolytic colonies were present, the suspect colonies were restreaked onto specific chromogenic SAID plates (biomérieux, Geneva, Switzerland). Pellets of the samples containing S. aureus were frozen and subsequently processed at the University of Lausanne. The isolation of S. aureus was carried out by the plating of the pellets onto SAID plates. To detect very low numbers of bacteria, the rest of the sample was transferred into 5 ml of Bacto m Staphylococcus broth (Difco, BD, Allschwil, Switzerland) and incubated overnight at 37 C. The enrichment broth was then plated onto a SAID plate, and isolates were processed in the same manner as that for primary plates. Thus, two isolates per sample were collected (if available). To screen for MRSA, the enrichment broth was also plated onto MRSASelect (Bio-Rad, Reinach, Switzerland) selective chromogenic agar. Phenotypic screening for resistance to methicillin on MRSASelect plates was followed by a PCR screen for an internal fragment of the meca gene as described previously (38). Moreover, in order to establish the presence of snout and skin colonization by S. aureus in cows, 177 nostril swabs as well as 167 swabs of the groin were obtained. Sampling consisted of vigorous swabbing inside the nostrils (5 to 10 cm deep) and on the skin between the thigh and the udder using Amies agar transport swabs (Copan, Brescia, Italy). Sampling details are given in Table S1 in the supplemental material. The swabs were stored at 4 C and processed within 3 weeks. Each swab was vigorously rubbed in 1 ml Tris-EDTA (TE) buffer, and 100 l was inoculated onto a SAID plate. From this point on, the isolation procedure was identical to that used for the milk samples. Sample collection from farmers. A subset of owners of infected animals were contacted and invited to participate in the study. Ethical clearance was sought and obtained from the Ethical Committee of the University of Lausanne. One nasal swab was obtained from each farmer and processed in the same way as that described above for swabs from animals. Genotyping. DNA was extracted from 600- l to 1-ml cultures of bacterial isolates grown overnight in brain heart infusion broth as described previously (4). Amplified fragment length polymorphism (AFLP) analysis was performed as described previously (38), except that only one set of selective primers was used. Electropherograms of AFLPs were analyzed with GeneMapper software (Applied Biosystems, Carlsbad, CA). Bayesian phylogeny was constructed with Mr- Bayes (35). Multilocus sequence typing (MLST) analysis of selected isolates was performed as previously described (7). Sequence types (STs) were identified by consulting the S. aureus MLST database ( The repeat region of the spa gene was amplified as previously described (12, 40). The spa types were assigned with the online spa database ( The genotyping of the clfb locus was performed according to a double-locus sequence typing (DLST) scheme (23). Briefly, highly polymorphic fragments of the clfb gene were amplified and sequenced. Data analysis. AFLP analysis was conducted to establish the diversity and phylogenetic relationship among the S. aureus isolates from bovine mastitis, carriage in farmers, and carriage in nonfarmers. It was previously shown that clusters delimited by AFLP analysis correspond exactly to MLST clonal complexes (38). The AFLP data for 47 isolates from nonfarmer human carriage that had previously been genotyped by MLST (38) together with data obtained in the present study were used here to perform clustering. This allowed us to establish the identity of the majority of the clusters in the present study. Some clusters were clearly delineated, but their identity was not established, as they did not contain isolates of the known sequence type. Several isolates per cluster were genotyped by MLST to establish the identity of these clusters as well as to confirm the identified clusters. The AFLP clusters are numbered after the most prominent (founder) sequence type of each MLST clonal complex. Although all isolates of the present study were genotyped by AFLP analysis, for visual clarity, the analysis presented in Fig. 1 includes a subset of all data. The AFLP profiles of all the isolates from farmers, including isolates from primary and enrichment plates, were obtained. For bovine mastitis isolates, this was done only in cases where transmission to or from farmers was suspected. Otherwise, only the isolate from the enrichment plate was analyzed. All statistical analyses were done with the R package (Foundation for Statistical Computing; RESULTS Genotype composition. Five hundred S. aureus bovine mastitis isolates from 133 farms were collected. One hundred sixty farmers from 64 of these farms provided nasal swabs. Only workers that were directly in contact with cows were included, which resulted in the sampling of one to eight persons per farm. Fifty-seven (36%) farmers were S. aureus carriers. For three nasal carriers, two distinct strains were found. AFLP analysis revealed a clear clustering of the isolates. The isolates from the farmers were quite diverse and belonged to the same range of clusters as that of the nonfarmer isolates (38), with the exception of a few isolates. Two of these isolates were genotyped by MLST and proved to be of ST1021, a sequence type which was not represented among nonfarmer isolates. Three other isolates dissimilar to nonfarmer isolates grouped with bovine mastitis isolates. Isolates from bovine mastitis were concentrated in a few distinct clusters. Many animal isolates clustered into cluster 8 (C8), which also contained isolates from cases of nonfarmer carriage. MLST of several bovine mastitis isolates established the identity of the clusters dominated by animal isolates as C20, C97, and C151 (Table 1). ST389 is a single-locus variant (SLV) of ST20, ST71 is an SLV of ST97, and ST504 is a double-locus variant (DLV) of ST151. Allelic differences corresponded to single or double mutations, consistent with diversification by point mutations, which is usual for S. aureus (7). Clusters 97 and 151 contained no human isolates. Cluster 20 contained isolates from cases of bovine mastitis and from farmers but no isolates from nonfarmers (Fig. 1 and Table 1 and see Table S2 in the supplemental material). spa typing provided an overview of the diversity among the collected isolates (Table 1). Many animal isolates have only two repeats, so no inferences about their relatedness can be made (25). spa types of C8 isolates consist of more repeats. Although we did not conduct formal analyses of their relatedness based on spa data, their close relatedness can be easily perceived by a comparison of spa repeats (Table 1). Genotype frequencies and distribution. Often, multiples samples were obtained per farm, and individual farms were sampled more than once. This introduces a bias to the estimate of the frequency of different genotypes. To estimate the unbiased frequency of different genotypes, only one (randomly selected) isolate from the first sampling time was considered. The genotype frequencies were calculated as the total number of isolates of a given genotype divided by the total number of isolates. The genotype compositions of S. aureus isolates from farmers and those from nonfarmers were very similar (Fig. 2), while the bovine mastitis isolates were clearly distinct. More than one-third of the bovine mastitis isolates belonged to C8. Many of these isolates were identical to nonfarmer isolates by

3 5910 SAKWINSKA ET AL. APPL. ENVIRON. MICROBIOL. FIG. 1. Clustering of selected AFLP data by a 50% consensus Bayesian tree. For clarity, only selected isolates are included: isolates from nonfarmers (n 47), all isolates from farmers (n 60), all bovine mastitis isolates which were genotyped by MLST (n 31), bovine isolates where human-to-animal transmission was suspected (n 13), and a random sample of other bovine mastitis isolates (n 30). Filled symbols indicate the isolates that were genotyped by MLST. The isolates for which human-to-bovine transmission was suspected are labeled with the farm number and M for bovine mastitis isolates or F for farmer isolates. The posterior probability (a proportion of trees supporting a given branch) is indicated in italics. AFLP analysis. Only one farmer carried a C8 isolate. However, the farmer isolate had a spa type (t008) different from that of bovine isolates (also see Transmission below). The geographic distribution of the farms as well as the distribution of the isolates within the farms are shown in Fig. 3. To avoid bias due to multiple sampling times, only the isolates collected at the first sampling were included (n 248). For 51 farms, multiple isolates were obtained. For samples from 21

4 VOL. 77, 2011 HUMAN-BOVINE HOST SHIFT IN STAPHYLOCOCCUS AUREUS 5911 TABLE 1. Genotyping results for bovine mastitis isolates a CC/AFLP No. of isolates spa type spa repeats Sequence type(s) (no. of isolates) b 1 3 t ND 7 2 t ND t t t (2) 17 t , t (2) 29 t (2) 119 t (3), t (2) 15 2 t ND t t (2) 1 t ND 45 1 t t (3) 1 t t ND t t (2), t t c 7 t Total 500 a ND, not determined. The most numerous groups are indicated in boldface type. b If more than one isolate per spa type was genotyped by MLST, the number of genotyped isolates is given in parentheses. c MRSA. farms (41%), more than one genotype of S. aureus was detected. No obvious spatial clustering was detected. Within-farm prevalence. It appeared that farms with the largest number of isolates tended to contain C8 isolates, while C151 dominated farms with a single isolate (Fig. 3). To quantify this observation, the estimate of the minimum within-farm prevalence of each cluster was calculated. The number of animals that carried an isolate belonging to a given cluster was divided by total number of lactating cows on this farm. This estimate represents minimum prevalences, because some of the infected animals could have been overlooked. Only the isolates from the first available sampling time were taken into account (Fig. 4). The prevalence was slightly but significantly higher for cluster 8 {F (4, 132) 2.7; P 0.05 (determined by one-way analysis of variance [ANOVA])}. Persistence. For 35 animals, multiple (usually 2) isolates were obtained, separated by at least a 3-week time interval. C8 isolates were found on the second sampling time or replaced another genotype in 21 out of 35 cases (60%), C20 isolates were found on the second sampling time or replaced another genotype in 4 cases (11%), C97 isolates were found on the second sampling time or replaced another genotype in 3 cases (9%), and C151 isolates were found on the second sampling time or replaced another genotype in 7 cases (20%). This appeared to be equivalent to the respective frequencies of these genotypes. Thus, there was no indication that one of the genotypes rapidly displaced the other genotypes. MRSA. MRSA was found in samples from 3 farms, including 2 samples from farmers and 6 from cases of bovine mastitis. In addition, MRSA was detected in one more mastitis sample after enrichment and plating on a selective medium. Methicillin-sensitive S. aureus (MSSA) (C151) was also found in the same sample. All MRSA isolates belonged to cluster 398. No discrepancies were found among data for phenotypic detection on plates and PCR screening for the meca gene. Transmission. S. aureus-colonized farmers were found on 36 farms, allowing a comparison of the genotypes of isolates from cases of bovine mastitis and human carriage. On 3 farms, the isolates from cases of bovine mastitis and farmer carriage were identical according to 3 markers used. On 2 more farms, the isolates from cases of bovine mastitis and farmer carriage were very similar but not identical (Table 2 and Fig. 1). S. aureus from cow snouts and groins. S. aureus was found in 1 out of 167 groin samples and 1 out of 177 snout samples (see Table S1 in the supplemental material). The genotype of the groin isolate was typically bovine (t164; C20). The cow snout

5 5912 SAKWINSKA ET AL. APPL. ENVIRON. MICROBIOL. FIG. 2. Genotype composition of S. aureus isolates from cases of bovine mastitis and from human carriers. One isolate per farm and sampling time is included. The colors represent clusters defined by AFLP analysis. isolate was of spa type t008 and C8. In fact, this sample originated from farm 34, where the only case of human colonization with a C8, spa type t008 strain was found (Table 2). The nasal isolate of the cow was thus identical to the nasal isolate of the farmer but slightly different from the mastitis isolate of the same animal. DISCUSSION Bovine-adapted CC8 strains. Strains belonging to clonal complex 8 constituted more than one-third of S. aureus bovine mastitis isolates from Western Switzerland. Given their regional prevalence and wide distribution over the study area, corroborated by the estimated high within-farm prevalence, there was little doubt that these strains represent a truly bovine-adapted genotype. Similar strains were previously rarely reported for bovine mastitis. The number and quality of past studies of bovine mastitis in Europe make it unlikely that a massive occurrence of CC8 in bovine mastitis would have been missed. We propose that bovine-adapted CC8 strains probably originated in Switzerland or nearby and, moreover, might be in the process of spreading across Europe. Several years ago, a spa type identical to the one found in the present study (t2953) was detected in 6 out of 50 isolates from Central Switzerland (26). More recently, very similar strains were quite common in Eastern France, only 150 km away from the Swiss study area (39). It is possible that similar strains were also found very recently in the Netherlands. van den Borne and colleagues previously found 10% of bovine mastitis isolates of ST8, but no spa typing was performed (43). The previous absence of CC8 from isolates from cases of bovine mastitis, its predominantly local distribution, and its extreme genetic closeness to human isolates imply a recent host shift and the consequent emergence as a major bovine mastitis genotype. A collection of healthy carriage isolates from the same small region of Western Switzerland (38) was used here as a reference to determine whether the isolates from farmers were typical or unusual and thus to assess the likelihood of bovineto-human transmission. Given that this previous study showed that a majority of unrelated human carriers had a unique S. aureus strain, we consider that a finding of identical isolates in farmers and cows from the same farm is very unlikely to arise from anything other than transmission. Strikingly, we found almost a complete lack of transmission of highly human-similar and prevalent CC8 strains from cows to the farmers. The lack of transmission strongly suggests that the host shift was accompanied by the loss of determinants (genetic factors) necessary for human carriage. Our study should have detected any transmission, and indeed, we identified some highly likely cases of both human-to-cow and cow-to-human transmission (see below). Our discovery of a lineage that has very recently lost the determinants necessary for inhabiting humans may permit the precise identification of such genetic determinants. A humanto-ruminant host jump was recently proposed for small-ruminant-specific CC133 (11), but it most likely occurred hundreds of years ago. Lowder and colleagues (24) described a recent host shift from humans to chickens, but it was not entirely clear whether some transmission back to humans still occurred among newly chicken-adapted S. aureus strains. A very recent study (9) described an MRSA lineage (ST130) found both in humans and in cows. Other bovine-adapted strains. In the present study we have found genotypes similar to those previously considered typical of bovine mastitis (13, 14, 30, 41), namely, C151 and C97. Our findings emphasize the fact that the existence of globally distributed genotypes of bovine-associated S. aureus, such as CC151 or CC97, does not mean that they are the most prevalent ones at any given location. The existence of substantial geographic structure is evident and reminiscent of the population structure of human carriage isolates (36). For example, in Brazil ST126 was the most common ST (30), while in Norway ST133 was predominant (18). These genotypes show a close relationship neither to each other nor to CC151 or CC97. CC20 had not been considered a typical bovine-adapted genotype, but increasing numbers of studies reported it as being common (France [39], Southern Germany [26], Japan [14], and South Korea [17]). The consequences of the genetic variability of S. aureus for the pathogenesis and treatment of bovine mastitis remain to be explored. In general, we have found only a very moderate level of transmission of S. aureus between farmers and cows. Although we cannot completely exclude an unknown common source or chance findings, 5 out of 57 farmers most likely acquired their strain from the animals. Three farmers carried isolates of C20, which were very frequent in cows, including their own animals, but were never found among nonfarmers in this region. Human carriage and infection with strains belonging to C20 were previously reported in Europe (7, 36) and China (6) but rather infrequently ( 5%). The only two farmers colonized with MRSA carried isolates

6 VOL. 77, 2011 HUMAN-BOVINE HOST SHIFT IN STAPHYLOCOCCUS AUREUS 5913 of ST398. Pigs colonized with MRSA of ST398 were present on this farm (A. Oppliger et al., unpublished data), and because ST398 is a typical pig-associated genotype, the transmission from pigs to both humans and cows appears to be the most FIG. 4. Within-farm prevalence of isolates from different clusters. Averages and standard deviations are shown. The colors indicate clusters as defined by AFLP analysis. likely transmission path. Because the two animal species were not housed together, humans are suspected to be the vector of transmission. The isolation of the same MRSA strain from the farmer and the cows was described previously (19). The discovery of the colonization potential of MRSA of ST398 in humans and emerging infections (10, 21, 22) made it clear that even low levels of transmission and only to persons who are in direct contact with animals can have significant consequences. Concerning other cases of suspected human-to-cow transmission, yet another farmer had a C15 isolate very similar though not identical to the ones isolated from his two cows. The fact that this genotype was not found in any other bovine mastitis samples suggests transmission from the farmer to the cows. We have confirmed that C151 appears to be confined to cases of bovine mastitis, and to our knowledge, it has never been reported for human carriage or infection. We found no human cases of carriage or infection of isolates of C97, although this genotype has been reported for isolates from cases of human carriage and infection in Europe and Algeria (7, 36). Only a few ( 2%; 10 of 513) bovine mastitis isolates had a genotype that can be considered typically human (C1, C7, C15, FIG. 3. Approximate geographic position (based on postal codes) of the studied farms and the genotypes of S. aureus bovine mastitis isolates. Colors represent the genotypes of the isolates, as defined by AFLP analysis. The size of the circle indicates the number isolates from a given farm. The smallest circles indicate a single isolate. The data for the first available sampling time are shown for each farm. The circles representing farms where the farmers were sampled are indicated by thicker black lines. The farms where MRSA was found at any sampling time are indicated by thick purple lines.

7 5914 SAKWINSKA ET AL. APPL. ENVIRON. MICROBIOL. Farm TABLE 2. Characteristics of isolates for which transmission between cows and farmers was suspected a No. of isolates shared/no. of isolates available Cows Farmers AFLP cluster AFLP identical DLST clfb spa type spa repeats 4 5/14 1/3 20 Yes 218 t /14 2/2 20 Yes 218 t /8 2/6 398 Yes 245 t /2 15 Yes 6 t /4 15 New 6 t /18 8 Type 1 b New 1 t /18 8 Type 1 3 t /1 (nasal) 1/1 8 Type 2 b 3 t a Boldface type facilitates comparisons among spa repeats. b Two out of 58 fragments differ between type 1 and type 2. C45, C101, and C188). It is common to find a low percentage of typically human isolates among bovine mastitis isolate collections. Contamination of bovine mastitis samples with S. aureus carried by humans cannot be excluded. We consider that these cases represented a probably transient acquisition from human carriage, although transient infection and true contamination are hard to distinguish. Are cows colonized with S. aureus? We found S. aureus only very rarely in the samples from groin skin and snout. Our results imply that the intramammary reservoir is sufficient for the transmission and, thus, the persistence of bovine S. aureus, as was suggested previously (15). However, some studies reported much higher rates of isolation of S. aureus from external body sites as well as the environment (1). Remarkably, the only case of isolation of S. aureus from cow snout concerned the farm where the potential transmission of a C8 isolate to the farmer might have occurred. This suggests that nasal colonization, even though rare, can be of importance for interspecies transmission. It is very likely that details of animal husbandry and hygiene are important factors determining the presence of S. aureus outside the mammary gland and in the environment. However, even though such a reservoir might be epidemiologically relevant at some locations, cows do not appear to be colonized by S. aureus in the same manner as that for humans or pigs, with colonization in this sense meaning that the facultative pathogen can maintain itself in the host population without causing any disease. MRSA of ST398 in Switzerland, February MRSA of ST398 has been reported for cases of bovine mastitis from a number of European countries, including Switzerland (8, 16, 26, 44). Our samples were collected at the beginning of 2008, thus pushing back the first date of detection of this genotype in bovine mastitis isolates in Switzerland. Unexpectedly, in only one case could we find a likely transmission source of this genotype, namely, the presence of colonized pigs on the farm. In 3 other cases (farms), no potential transmission route could be envisioned. The question about the spread of this genotype remains open. Conclusion. Many prevalent human diseases most likely originated from domesticated animals (46). Despite the pronounced host specificity of staphylococcal lineages, rare host shifts are to be expected. Here, we found evidence for a recent host shift from humans to cows in clonal complex 8, a common staphylococcal lineage and frequent host for the staphylococcal cassette chromosome mec (SCCmec) genomic island. Molecular determinants of this host shift are under investigation. Many clinically important MRSA clones belong to CC8 (e.g., see reference 3). The acquisition of the SCCmec genomic island and, thus, the emergence of MRSA from MSSA apparently occurred on multiple independent occasions in this genetic background (33), indicating a general suitability of CC8 as the host for SCCmec (5). The emergence of bovine-associated CC8 MRSA, besides the problems related to the treatment of bovine mastitis, would create a reservoir of MRSA alarmingly close to the most virulent human S. aureus strains. ACKNOWLEDGMENTS We are grateful to Patrick Boujon and Patricia Gabioud for help with sample collection; Marc Robinson-Rehavi and Sebastién Moretti for bioinformatic help with spa typing; Antoine Guisan and Anne Dubuis for help with preparation of the map; Harald Brüssow, Gregory Resch, and Ian Sanders for comments; and veterinarians and farmers for participation. This work was supported by grant 3200B to P.M. and a Marie Heim Vögtlin grant (PMPDA ) to O.S., both from the Swiss National Science Foundation. REFERENCES 1. Capurro, A., A. Aspan, H. E. Unnerstad, K. P. Waller, and K. Artursson Identification of potential sources of Staphylococcus aureus in herds with mastitis problems. J. Dairy Sci. 93: de Neeling, A. J., et al High prevalence of methicillin resistant Staphylococcus aureus in pigs. Vet. Microbiol. 122: Diep, B. A., et al Complete genome sequence of USA300, an epidemic clone of community-acquired meticillin-resistant Staphylococcus aureus. Lancet 367: Elphinstone, M. S., G. N. Hinten, M. J. Anderson, and C. J. Nock An inexpensive and high-throughput procedure to extract and purify total genomic DNA for population studies. Mol. Ecol. Notes 3: Ender, M., N. McCallum, R. Adhikari, and B. Berger-Bachi Fitness cost of SCCmec and methicillin resistance levels in Staphylococcus aureus. Antimicrob. Agents Chemother. 48: Fan, J., et al Biogeography and virulence of Staphylococcus aureus. PLoS One 4:e Feil, E. J., et al How clonal is Staphylococcus aureus? J. Bacteriol. 185: Fessler, A., et al Characterization of methicillin-resistant Staphylococcus aureus ST398 from cases of bovine mastitis. J. Antimicrob. Chemother. 65: García-Álvarez, L., et al Meticillin-resistant Staphylococcus aureus with a novel meca homologue in human and bovine populations in the UK and Denmark: a descriptive study. Lancet Infect. Dis. [Epub ahead of print.] doi: /s (11) Golding, G. R., et al Livestock-associated methicillin-resistant Staphylococcus aureus sequence type 398 in humans, Canada. Emerg. Infect. Dis. 16: Guinane, C. M., et al Evolutionary genomics of Staphylococcus aureus

8 VOL. 77, 2011 HUMAN-BOVINE HOST SHIFT IN STAPHYLOCOCCUS AUREUS 5915 reveals insights into the origin and molecular basis of ruminant host adaptation. Genome Biol. Evol. 2: Harmsen, D., et al Typing of methicillin-resistant Staphylococcus aureus in a university hospital setting by using novel software for spa repeat determination and database management. J. Clin. Microbiol. 41: Hasman, H., et al spa type distribution in Staphylococcus aureus originating from pigs, cattle and poultry. Vet. Microbiol. 141: Hata, E., et al Genetic variation among Staphylococcus aureus strains from bovine milk and their relevance to methicillin-resistant isolates from humans. J. Clin. Microbiol. 48: Haveri, M., M. Hovinen, A. Roslof, and S. Pyorala Molecular types and genetic profiles of Staphylococcus aureus strains isolated from bovine intramammary infections and extramammary sites. J. Clin. Microbiol. 46: Huber, H., S. Koller, N. Giezendanner, R. Stephan, and C. Zweifel Prevalence and characteristics of meticillin-resistant Staphylococcus aureus in humans in contact with farm animals, in livestock, and in food of animal origin, Switzerland, Eurosurveillance 15: Hwang, S. Y., Y. K. Park, H. C. Koo, and Y. H. Park spa typing and enterotoxin gene profile of Staphylococcus aureus isolated from bovine raw milk in Korea. J. Vet. Sci. 11: Jorgensen, H. J., T. Mork, D. A. Caugant, A. Kearns, and L. M. Rorvik Genetic variation among Staphylococcus aureus strains from Norwegian bulk milk. Appl. Environ. Microbiol. 71: Juhasz-Kaszanyitzky, E., et al MRSA transmission between cows and humans. Emerg. Infect. Dis. 13: Klein, E., D. L. Smith, and R. Laxminarayan Hospitalizations and deaths caused by methicillin-resistant Staphylococcus aureus, United States, Emerg. Infect. Dis. 13: Kock, R., et al Prevalence and molecular characteristics of methicillinresistant Staphylococcus aureus (MRSA) among pigs on German farms and import of livestock-related MRSA into hospitals. Eur. J. Clin. Microbiol. Infect. Dis. 28: Krziwanek, K., S. Metz-Gercek, and H. Mittermayer Methicillinresistant Staphylococcus aureus ST398 from human patients, Upper Austria. Emerg. Infect. Dis. 15: Kuhn, G., P. Francioli, and D. S. Blanc Double-locus sequence typing using clfb and spa, a fast and simple method for epidemiological typing of methicillin-resistant Staphylococcus aureus. J. Clin. Microbiol. 45: Lowder, B. V., et al Recent human-to-poultry host jump, adaptation, and pandemic spread of Staphylococcus aureus. Proc. Natl. Acad. Sci. U. S. A. 106: Mellmann, A., et al Based Upon Repeat Pattern (BURP): an algorithm to characterize the long-term evolution of Staphylococcus aureus populations based on spa polymorphisms. BMC Microbiol. 7: Monecke, S., P. Kuhnert, H. Hotzel, P. Slickers, and R. Ehricht Microarray based study on virulence-associated genes and resistance determinants of Staphylococcus aureus isolates from cattle. Vet. Microbiol. 125: Moreillon, P., and Y. A. Que Staphylococcus aureus (including staphylococcal toxic shock), p In G. L. Mandell, J. E. Bennett, and R. Dolin (ed.), Mandell, Douglas and Bennett s principles and practice of infectious diseases, 7th ed. Elsevier, New York, NY. 28. Moret-Stalder, S., et al Prevalence study of Staphylococcus aureus in quarter milk samples of dairy cows in the Canton of Bern, Switzerland. Prev. Vet. Med. 88: Nubel, U., et al Frequent emergence and limited geographic dispersal of methicillin-resistant Staphylococcus aureus. Proc. Natl. Acad. Sci. U. S. A. 105: Rabello, R. F., et al Multilocus sequence typing of Staphylococcus aureus isolates recovered from cows with mastitis in Brazilian dairy herds. J. Med. Microbiol. 56: Raney, T The state of food and agriculture 2009: livestock in the balance. Food and Agriculture Organization of the United Nations, Rome, Italy. 32. Rasigade, J. P., F. Laurent, P. Hubert, F. Vandenesch, and J. Etienne Lethal necrotizing pneumonia caused by an ST398 Staphylococcus aureus strain. Emerg. Infect. Dis. 16: Robinson, D. A., and M. C. Enright Evolutionary models of the emergence of methicillin-resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 47: Roesch, M., M. G. Doherr, W. Scharen, M. Schallibaum, and J. W. Blum Subclinical mastitis in dairy cows in Swiss organic and conventional production systems. J. Dairy Res. 74: Ronquist, F., and J. P. Huelsenbeck MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: Ruimy, R., et al Comparisons between geographically diverse samples of carried Staphylococcus aureus. J. Bacteriol. 191: Ruimy, R., et al The carriage population of Staphylococcus aureus from Mali is composed of a combination of pandemic clones and the divergent Panton-Valentine leukocidin-positive genotype ST152. J. Bacteriol. 190: Sakwinska, O., et al Genetic diversity and ecological success of Staphylococcus aureus strains colonizing humans. Appl. Environ. Microbiol. 75: Sakwinska, O., et al Link between genotype and antimicrobial resistance in bovine mastitis-related Staphylococcus aureus strains, determined by comparing Swiss and French isolates from the Rhone Valley. Appl. Environ. Microbiol. 77: Shopsin, B., et al Evaluation of protein A gene polymorphic region DNA sequencing for typing of Staphylococcus aureus strains. J. Clin. Microbiol. 37: Smith, E. M., et al Multilocus sequence typing of intercontinental bovine Staphylococcus aureus isolates. J. Clin. Microbiol. 43: Sung, J. M. L., D. H. Lloyd, and J. A. Lindsay Staphylococcus aureus host specificity: comparative genomics of human versus animal isolates by multi-strain microarray. Microbiology 154: van den Borne, B. H. P., et al Host adaptation of bovine Staphylococcus aureus seems associated with bacteriological cure after lactational antimicrobial treatment. J. Dairy Sci. 93: Vanderhaeghen, W., et al Methicillin-resistant Staphylococcus aureus (MRSA) ST398 associated with clinical and subclinical mastitis in Belgian cows. Vet. Microbiol. 144: van Loo, I., et al Emergence of methicillin-resistant Staphylococcus aureus of animal origin in humans. Emerg. Infect. Dis. 13: Wolfe, N. D., C. P. Dunavan, and J. Diamond Origins of major human infectious diseases. Nature 447: Wulf, M., and A. Voss MRSA in livestock animals an epidemic waiting to happen? Clin. Microbiol. Infect. 14: Zadoks, R. N., et al Comparison of Staphylococcus aureus isolates from bovine and human skin, milking equipment, and bovine milk by phage typing, pulsed-field gel electrophoresis, and binary typing. J. Clin. Microbiol. 40: