Introduction

Methicillin-resistant Staphylococcus aureus (MRSA) spa-type t127, Sequence Type (ST) 1, was first reported as one of the three most prevalent MRSA lineages in Italian pig industry, present in 6% of the holdings surveyed in Italy in 2008 [1]. ST1 belongs to Clonal Complex (CC)1, a particularly successful lineage associated with human infections, which includes Panton-Valentine (PVL)-positive CA-MRSA also known as USA400 [2]. In Europe, PVL-positive ST1 CA-MRSA SCCmec V, was firstly reported in a Dutch patient (2005–2006) [3], then in an Italian patient (2007) after travel in the USA [4], and in Denmark [5]. In the United Kingdom, PVL-negative ST1, spa-type t127, SCCmec IVa isolates are among the most common CA-MRSA, often associated with injecting drug-users and homeless people [6]. Furthermore, ST1 spa-type t127 isolates have been reported as the sixth most prevalent clone, both MSSA and MRSA, isolated from human invasive infections in Europe [7,8].

As previously reported [9] spa-type t127 ST1 isolates of porcine origin shared a 75% similarity when tested by pulse field gel electrophoresis (PFGE) with isolates of human origin. They showed mainly differences in the SCCmec cassette and in some virulence and antimicrobial resistance marker genes.

In Italy, in the last five years, the CC1 lineage has also been increasingly detected in dairy cattle mastitis (Battisti, unpublished) [10, 11], in bulk milk from dairy cattle, and occasionally from goat milk and colonizing veal calves and small ruminants [12]. The aim of the study was to analyse the differences between isolates from food-producing animals and humans isolated in Italy and other European countries and to determine their genetic relatedness for epidemiological and risk assessment purposes.

Materials and Methods

A total of 83 ST(CC)1 S. aureus (71 MRSA and 12 MSSA) were studied (Fig 1). The majority (n = 65) were from Italy (including 30 MRSA from pigs, 18 MRSA from bovines, 8 MRSA from humans; 1 MSSA from bovine, 1 MSSA from pig, 1 MSSA from ovine, and 6 MSSA from human) and the remainders from Finland (6 MRSA from pigs), Spain (4 MRSA and one MSSA from pigs and 2 MSSA from humans), Denmark (4 MRSA from humans), Cyprus (1 MRSA from pigs).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Dendrogram of SmaI PFGE macrorestriction patterns, with selected virulence and antimicrobial resistance marker genes and elements of Sequence Type 1, Clonal Complex 1 MSSA and MRSA from animals and humans.

Legend: ^† LukF-PV(P83)-LukM; ^‡ vwb (COL+MW2); ^☼Ciprofloxacin resistance; ^§ PFGE profile. Notes: *SCCmecV&SCCfus; ** SCCmec IV (2B&5), subtype IVa.

https://doi.org/10.1371/journal.pone.0137143.g001

Isolates from animals (n = 63, 2007–2011) were collected from population-based studies in different primary productions (surveys in pigs, cattle), and passive laboratory surveillance programmes (mastitis in dairy cattle and sheep). The Italian isolates originated from farms located in Northern and Central Italy. For the purpose of the study, only one representative isolate per spa-type and ST per epidemiological unit (i. e. per holding) was included. The 20 isolates of human origin (8 MSSA and 12 MRSA, 1999–2011) investigated in this study were derived from previous laboratory surveillances [5,8,9,13], originated from infections, and were of unknown epidemiological relationships with animal farming [9]. Metadata of the isolates under study are included in S1 Table.

Most of isolates of animal origin (e. g. nasal swabs, milk) were cultured as previously described [1], with slight modifications. Selective enrichment cultures were plated on Oxacillin Resistance Screen Agar (ORSA, Oxoid, Basingstoke, UK). Suspect S. aureus colonies (denim blue colonies on ORSA) were subcultured in 5% sheep blood agar and incubated at 37°C for 24 h. In some cases, animal isolates from passive surveillance were detected by direct plating on 5% sheep blood agar or on selective agar (ORSA). S. aureus isolates were identified by means of standard techniques: colony morphology, Gram staining, catalase, coagulase tube test and further confirmed by PCR assay of the nuc gene as previously described [14].

In order to determine whether the difference between two proportions was significant (presence of genetic or phenotypic characteristics in isolates from swine vs isolates from ruminants and humans), the Fisher exact test was calculated by using the StatCalc utility of the Epi Info version 7.1.5 software (http://wwwn.cdc.gov/Epiinfo/7/index.htm).

Results

Discussion

This study represents the first wide molecular characterization of MRSA CC1 isolates from farm animals, and offers a further insight into relatedness and similarities with a set of CA-MRSA isolates from in humans.

Macrorestriction-PFGE clusters showed an impressive concordance with host of isolation and a series of other genetic markers (virulence, antimicrobial resistance) that are often carried by mobile genetic elements (MGEs), bacteriophages, and subject to loss or acquisition in relation to host adaptation process, useful for molecular epidemiology (Fig 1). Among these, SCCmec V is clearly a feature of porcine isolates, almost all in cluster I, while human and bovine MRSA isolates (cluster III) are characterized by SCCmec IVa. An exception is a single human MRSA SCCmec V (Staph 151), clustering with the porcine isolates, although also showing phage- or transposon- encoded genes typical of human adapted strains (Fig 1). Interestingly, we also report the finding of SCCmecV&SCCfus in two human isolates (from Italy and Denmark), an occurrence that has only been scarcely reported so far [25].

Striking is that swine MRSA isolates from Finland, show a high degree of similarity (subcluster III.c), and although with a SCCmec V, they seem more closely related to the human-bovine cluster (grouped in subcluster III.b) than the porcine isolates from Italy and Spain (cluster I). Conversely, they lacked all other acquired resistance genes, except for tet(K) and blaZ (Fig 1), and fluoroquinolone resistance, which is a typical feature of porcine multidrug-resistant MRSA CC1 isolates from Italy and from Spain. This is probably explained by a low antimicrobial usage in Finnish pig industry as shown in the ESVAC Reports [26]. Additionally, among the specific genetic markers of the Italian and Spanish multidrug-resistant MRSA of porcine origin is the presence of pleuromutilin resistance, mostly attributed to the vga(A) gene and also contributing to combined resistance to streptogramins A and lincosamides [27], in agreement to what has been previously reported in porcine MRSA [9]. Interestingly, resistance to quaternary ammonium compounds, commonly used in sanitization procedures in pig holdings [28] was another feature of Italian porcine isolates (cluster I), and was detected in one Danish human isolate only (cluster III).

Typically, all isolated studied including those from cattle/sheep, lack the SAPI-encoded bovine vwb gene, an important trait of host-specific pathogenicity, and this may be considered as one of the indications of a recent transfer of this lineage to food-producing animals, and especially to the ruminant hosts. As for the enterotoxin gene profile, seh appears to be constitutive of the ST(CC)1, irrespective of the host of origin, as previously observed [9]. Additionally, the pattern sea-seh-sek-seq is a feature typical of human isolates, since it has been found in only one IEC-positive MSSA isolate of animal origin, from a dairy cattle herd, clustering with human isolates (cluster II).

Most interestingly, the vast majority of isolates from dairy cattle (and sheep), shared a lot of genetic features usually found in typical human-associated CC1 clones. Amongst these are the immune evasion cluster (IEC) genes sak and scn, located on β-haemolysin-converting bacteriophages [29] and sea. Indeed, among the sak- or scn-positive isolates, ten were also positive for the enterotoxin sea gene, of which n = 9 isolated from humans (3 MRSA and 6 MSSA), while only one (MSSA) from dairy cattle. Similarly, almost all isolates from cattle and sheep often carried sat and aphA3 resistance genes associated with Tn5405-like elements, to be considered as further markers of a human origin [30,31]. Another characteristic of the human-bovine profile was the absence of erm(A) mediated macrolide-lincosamide resistance, and the rare occurrence of fluoroquinolone resistance, which is an epidemiological marker for Italian and Spanish porcine isolates.

All these features may be suggestive of a recent adaptation to the bovine (ruminant) host of human-associated CC1 strains, compared to the porcine cluster. A possible pathway of human-to-cattle exchange may be direct contact between farm workers and animals or indirect exposure through farm environment. Exposure of dairy animals is likely to result in colonization and even infection, since it is known that ST1 MRSA is capable of causing mastitis in cattle and mastitis/colonization in small ruminants [12]. A longitudinal study on a single case of chronic mastitis in a dairy cow [10], reported shedding over time, with immune response indicating acute infection, and neutrophils as the most represented cell population. Indeed, the first case of ST1 spa-type t127 SCCmec IVa mastitis in cattle was reported in Hungary in 2002, and showed PFGE-indistinguishable isolates amongst dairy cattle and one of the workers at the farm [32]. Unfortunately no further molecular typing was performed on these strains.

In conclusion, all MRSA CC1 isolates studied, irrespective of their host (animal or human) origin possess several virulence genes and resistance genes towards major classes of antimicrobials, and a variety of other pathogenicity factors. In this respect, MRSA in farm animals, either of bovine or swine origin, are to be considered a hazard for the community, and may represent a serious therapeutic challenge in case of invasive infections in humans.

Additionally, the finding that most of ST(CC)1 MRSA and MSSA associated with cattle in Italy, besides being multidrug-resistant and showing very high genetic relatedness (>90–100% PFGE similarity) to human isolates, also possess some of the virulence and immunomodulatory (e. g. IEC cluster) marker genes associated with full capability of colonizing and infecting humans, is of further concern. Since this clone has been demonstrated causing mastitis in cattle, with shedding of the pathogen over the lactation cycle, dairy farming may represent a source of exposure for the community also through unpasteurized milk or dairy products. With these genetic characteristics, spill-over from food-producing animals in high-prevalence geographic areas should be taken seriously. Prevention efforts should include actions to be taken at farm-level, aimed at minimizing exposures in the community and in categories directly related to farm animal industry.

Supporting Information

Acknowledgments

This work was supported by grants from the EMIDA ERA-NET (FP7-KBBE) first transnational research call, Project “LA-MRSA” (acronym title) ID 68.

The Authors wish to thank Paola Di Matteo, Angela Ianzano, Manuela Iurescia, Tamara Cerci for outstanding technical assistance.