A methicillin-resistant Staphylococcus aureus (MRSA) Sequence Type 8, spa type t11469 causing infection and colonizing horses in Italy

A Methicillin-resistant Staphylococcus aureus (MRSA) was isolated in Italy from a pathological sample of a mare presenting chronic purulent sinusitis and that had undergone frontal-sinus surgery three months before. Humans, horses, dogs and environmental samples were subsequently collected at the mare's stable and at the Veterinary Hospital, where the mare was operated/hospitalized, and screened for the presence of MRSA that was detected from other horses and from the environment at both sites. All the MRSA isolates belonged to clonal complex (CC)8, ST8-t11469-SCCmec-IVa, and showed similar phenotypic and genetic multidrug resistance patterns and macrorestriction-pulsed-field gel electrophoresis profiles. The only MRSA detected from humans was a CC1, ST1-t127-SCCmec-IVa. This paper represents the first report of a clinical MRSA infection in a horse in Italy. This study also supports the opinion that improper use of antibiotics and hospitalization/surgery can represent risk factors for MRSA colonization/infection in horses, and that the environment is among important sources for exposure.

INTRODUCTION

Staphylococcus aureus is involved in a wide variety of diseases in humans and animals and its pathogenicity is mainly related to a combination of genetic characteristics mediating virulence, invasive capacity, immune evasion and antibiotic resistance (Chua et al. 2014). Colonization and infections by different clonal complexes of methicillin-resistant S. aureus (MRSA) have been reported in companion animals such as cats, dogs and horses in several countries (Weese and van Duijkeren 2010; Boyen et al. 2013; Abdelbary et al. 2014), and zoonotic transmission from horses has been described (Boyen et al. 2013; Abdelbary et al. 2014). In horse farms, MRSA has been found circulating with prevalences ranging from 0.6% up to 4.7%, while higher prevalences (5.8%–12.0%) have been reported in horses admitted to veterinary hospitals (van Balen et al. 2014). As in humans, MRSA can colonize skin, nasal and oral mucosae of horses that can be healthy carriers. Nonetheless, clinical MRSA infections can occur, both as sporadic cases or outbreaks (Boyen et al. 2013), with risk factors for colonization/infection being hospitalization, surgery, wounds, previous MRSA infection or a history of MRSA isolation, antimicrobials exposure, chronic disease, or immunosuppression (Weese 2010; Maddox et al. 2015; Tirosh-Levy et al. 2015). Clinical signs can range from mild to rapidly fatal, with skin/soft tissue infections (including wound and surgical site infections) and joint infections being most common (Weese 2010; Maddox et al. 2015).

Genotypic characteristics of MRSA strains isolated from equids vary among countries (van Balen et al. 2014). In the last years, MRSA clonal complex (CC)398, and in particular the subclone named clade (C), has become the predominant lineage in clinically infected horses in Europe (Sieber et al. 2011; Abdelbary et al. 2014; Maddox et al. 2015), although different epidemic clones have also been described in horse populations (Weese et al. 2005, 2006; Walther et al. 2009), with CC8 being the most prevalent among them (Weese 2010; Maddox et al. 2015). Strains belonging to CC8 seem to be well adapted, not only to the human host, but also to the equine host, reflecting their extended host spectrum (Walther et al. 2009).

Even though in Italy, the presence of livestock associated (LA-)MRSA in humans and farmed animals have been widely documented (Pan et al. 2009; Battisti et al. 2010; Franco et al. 2011; Monaco et al. 2013; Cortimiglia et al. 2015), information about horses is still scarce (De Martino et al. 2010; Mallardo et al. 2013) and, at present, clinical MRSA infections in horses have not been reported.

The aims of this study were to describe the characteristics of an MRSA strain isolated from a clinically infected Italian horse and to report the field and laboratory investigations started after the MRSA isolation, conducted on epidemiologically related animals, humans and from their close environment.

MATERIALS AND METHODS

Case description

In August 2014, an MRSA isolate was detected at the Istituto Zooprofilattico Sperimentale del Lazio e della Toscana ‘M. Aleandri’ from a diagnostic horse sample. The sample was a sinus swab from a 15-year-old bay Belgian-mare, stabled in the province of Rome, with chronic purulent sinusitis and purulent discharge from a postoperative sinus surgical incision.

Following isolation, information was collected retrospectively in order to better figure out the case.

All the available anamnestic information obtained from two different referring clinicians, and from the University Veterinary Hospital where surgery was performed, are summarized below. In May 2014, the mare had undergone a surgery procedure at a University Veterinary Hospital located in a different region from the stable of origin (roughly 200 km from Rome), for the removal of cyst on the left sinus in which the laboratory of the University Veterinary Hospital had detected Streptococcus equi subsp. zooepidemicus. The cyst was considered the probable cause of a chronic unilateral nasal discharge that in the past had been empirically treated with various and not further specified antimicrobials. Before surgery, clinical signs were depression, nasal airway obstruction, facial swelling, ocular and nasal discharge. Following surgery, the mare was reported to be treated for about four months with various antimicrobials, including fourth-generation cephalosporins and a combination of penicillin and streptomycin. Even following MRSA detection, despite the laboratory result, the mare was still treated with fourth-generation cephalosporins for 10 days, and with sulfonamides and trimethoprim for another 10 days. Finally, the mare was subjected for about four months to cycles (for one month daily, for two months every other day, and for one month twice a week) of nasal irrigation and sinus rinse with amikacin and/or local disinfectants, such as chlorhexidine or iodine solutions.

Samples collection following MRSA isolation

After the first MRSA isolation from the infected mare (August 2014, day 0) several samples were collected over a period of about 10 months in order to follow the evolution of the clinical case, the possible colonization of in-contact humans and of epidemiologically related animals, and to assess possible environmental contamination. The samples collected from the infected mare, and the clinical signs observed at the time of sampling are reported in Table 1.

Other samples were collected at the stable of origin of the clinical case 10, 180 and 300 days after the first MRSA isolation. Overall, 15 horses (all the horses present in the stable), 4 in-contact dogs and various animal and human in-contact surfaces/materials were sampled (Table 2).

Nasal swabs from 16 persons in close contact with the horses, including the two stable horse caretakers, the owner of the infected mare and its veterinarian, were also taken on different occasions (between time 1 and time 3 whenever available) (Table 2).

Two months after the first MRSA isolation from the infected mare, and five months after it underwent surgery, an investigation was also arranged and conducted at the University Veterinary Hospital, where the mare had been operated and hospitalized. On that occasion, nasal swabs were taken from the six horses present in the ‘Horse unit’, all of them recently admitted for non-infectious orthopedic problems, and from four veterinarians that had been in close contact with the infected mare (Table 2). Environmental samples from human and animal in-contact surfaces/materials/medical equipment were also collected (Table 2).

At every sampling occasion, horse and dog samples (anterior nares or visible wounds/skin lesions) were taken by using cotton-tipped swabs (Amies Agar Gel with Charcoal, Laboindustria s.p.a., Italy). Humans’ samples were taken by means of cotton-tipped swabs from both anterior nares.

For environmental samples, dry electrostatic cloths (for large surfaces) and sterile pre-moistened cotton-tipped swabs (for smaller surfaces) were used. Open and large size spaces/surfaces of a same section/room were sampled and examined as a pool.

All human samples were obtained on a voluntary basis, and the same applied for animal samples, since in Italy MRSA infection in animals is not a notifiable disease.

All procedures followed were in accordance with ethical standards and all participants gave oral informed consent to participate to the study.

MRSA isolation and identification

Within few hours after collection, samples were tested using MRSA selective media as previously described (Alba et al. 2015). Briefly, selective enrichment cultures (Mueller-Hinton broth supplemented with 6.5% NaCl) were plated on Oxacillin Resistance Screen Agar (ORSA, Oxoid, Basingstoke, UK). Suspect S. aureus colonies were subcultured and identified by biochemical and molecular assays as previously described (Battisti et al. 2010). Detection of mecA gene was performed by PCR assay as previously described (Battisti et al. 2010).

Antimicrobial susceptibility testing

All mecA positive isolates were further tested for phenotypic susceptibility by the broth micro-dilution method in 96-well microtitre plates (Trek Diagnostic Systems, Westlake, OH, USA). Results (Minimum Inhibitory Concentrations, MICs) were interpreted according to both the European Committee on Antimicrobial Susceptibility Testing (EUCAST; http://www.eucast.org) epidemiological cut-offs and clinical breakpoints (CB), or in absence of EUCAST CB, according to the Clinical Laboratory Standard Institute (CLSI) CB (e.g. for kanamycin and sulfamethoxazole as described by Feltrin et al. 2015). Epidemiological cutoffs were only used for tiamulin and streptomycin, since no CB have been made available by both EUCAST and CLSI. The following drugs were tested: penicillin, cefoxitin, ciprofloxacin, chloramphenicol, clindamycin, erythromycin, gentamicin, kanamycin, streptomycin, linezolid, quinupristin/dalfopristin, fusidic acid, mupirocin, rifampicin, tetracycline, tiamulin, sulfamethoxazole, trimethoprim and vancomycin.

Molecular characterization

All MRSA isolates were genotyped by spa-typing, multilocus sequence typing and by typing/subtyping of the staphylococcal cassette chromosome mec (SCCmec) as previously reported (Battisti et al. 2010). A total of 12 MRSA selected isolates, representative of the different types of samples/sources tested (Tables 1 and 2), were also characterized by micro-array analysis using the S. aureus Genotyping DNA Microarray (Alere Technologies GmbH, Germany) according to the manufacturer's instructions. A further characterization of the selected isolates involved SmaI pulsed-field gel electrophoresis (PFGE), performed as previously described (Alba et al. 2015). DNA PFGE patterns were analyzed using the BioNumerics 7 software (Applied Math, St-Martens-Latem, Belgium). Pair-wise similarities were calculated using the Dice coefficient, with tolerance and optimization both set at 1.5%. The algorithm UPGMA (Unweighted Pair Group Method using Arithmetic averages) was used for constructing dendrograms.

RESULTS

After the first isolation (day 0), the infected mare tested constantly positive for MRSA during the following seven months of sampling from both nares, skin and sinus wound swabs, even during antibiotic treatment (Table 1). A complete clinical signs remission was obtained about eight months from the first MRSA detection, around four months after the suspension of any antimicrobial therapy. Two sets of samples were taken after the mare healed in order to assess the carrier status, and both tested negative (Table 1). In the stable of origin of the mare, MRSA was also detected from the nasal swab of a horse caretaker, from a single nasal swab of an in-contact horse (horse A), and from environmental samples, but only from surfaces mostly in contact with positive horses (Table 2).

At the University Veterinary Hospital, MRSA was isolated from the nasal swabs of three-sixth hospitalized horses, and from the environment, including surfaces likely to come into contact with both humans and horses. MRSA was not isolated from any in-contact person, neither from any medical equipment (Table 2).

All the MRSA isolates from the infected mare, from the other sampled horses and from the different environmental samples, both at the mare's stable and at the University Veterinary Hospital (Tables 1 and 2), belonged to CC8, sequence type 8 (ST8), spa type t11469, SCCmec type IVa, and showed an identical antimicrobial resistance (AMR) pattern, being microbiologically and clinically resistant not only to beta-lactams, but also to clindamycin, erythromycin, gentamicin, kanamycin, tetracycline, trimethoprim and sulfamethoxazole, and microbiologically resistant to streptomycin (MIC = 64 mg L^{−1}). The 12 selected MRSA isolates subjected to micro-array analysis (Tables 1 and 2) showed the same profile, harboring relevant virulence genes, such as seb, sek, seq of the enterotoxin cluster, the superantigen/enterotoxin-like gene setC/selx, and the haemolysin gene without the phage-insertion (un-disrupted hlb). They also carried the same AMR genes, and beside beta-lactam resistance genes (mecA and blaZ), they were also positive for tet(M), erm(C), aacA-aphD, dfrA and fosB genes, involved respectively in tetracycline, macrolide, aminoglycoside, trimethoprim and fosfomycin resistance.

These MRSA isolates also showed an indistinguishable macrorestriction-PFGE profile, except three isolates presenting a single discordant band (≥96% similarity). No CC8 isolates were detected from human samples. The MRSA isolate detected from the nasal swab of the horse caretaker belonged to CC1, ST1, spa type t127, SCCmec type IVa, and presented genetic and phenotypic characteristics different from those of the CC8 MRSA isolates (data not shown).

DISCUSSION

In Italy, MRSA has been so far reported at a low prevalence (0.5%–0.6%) among staphylococcal strains isolated from nasal swabs of healthy horses (De Martino et al. 2010; Mallardo et al. 2013) and from conjunctival swabs (8.7%) of healthy donkeys (Foti, Fisichella and Giacopello 2013). To the best of the author's knowledge, the present paper describes for the first time the occurrence of a clinical MRSA infection in an Italian horse. Despite multiple courses of therapy, clinical signs and detectable infection in the infected mare lasted for as long as eight months. In a previous study, horses with clinical infections showed persistent nasal carriage for up to 711 days, with a median of 143 days (Bergström et al. 2013). In horses, MRSA eradication not only represents a tough challenge in cases of multidrug-resistant (MDR) MRSA infections, due to the limited therapeutic options, but also creates several management problems (Weese et al. 2005). In our case, recommendations of applying biosecurity and hygiene practices were given at the mare's stable, but were only partially implemented. Nevertheless, no other clinical infections and only a transient colonization in a single horse was observed during the 10 months survey, confirming that MRSA can be present, not necessarily associated with outbreaks (Van Balen et al. 2014).

The genotypic characterization demonstrated that all the 12 fully characterized MRSA isolates from the horses and the environment, both at the mare's stable and at the Veterinary Hospital, belonged to the same lineage and showed the same phenotypic and genetic MDR pattern, macrorestriction-PFGE and micro-array profile.

The MRSA isolates identified in this study belonged to ST8, a Sequence Type often associated with MRSA colonization and infection in horses in both Europe and North America (Weese 2010). In particular, in Europe, spa types t008, t020, t036, t064 and t451, within CC8, have been the most frequently reported (Sieber et al. 2011). Interestingly, spa type t11469 has a repeat motif similar to spa type t064 (although it lacks the last three repeats) and was previously reported only in poultry in Nigeria (Amos, Josiah and Busayo 2013). The MRSA strain here described showed genetic characteristics very similar to those of three equine MRSA ST8 spa type t064 isolates previously reported in Germany (Walther et al. 2009). In particular, all these isolates lacked some virulence markers of human adaptation, such as the beta-converting phage-borne immune evasion cluster genes sak, scn and sea, but carried the seb, sek, seq enterotoxin cluster, usually harbored on a pathogenicity island (SaPI1). This latter gene cluster is infrequently found in CC8 human clinical isolates (Holtfreter et al. 2007), although it was already described in some of the ‘Archaic’ CC8 clone isolates (Li et al. 2009). The genetic mosaic of this Italian horse strain is also very similar to that of the German ST8 isolates in the acquired resistance genes towards beta-lactams, tetracyclines, aminoglycosides, macrolides, lincosamides, streptogramins B (MLSb), dihydrofolate reductase inhibitors and fosfomycin. Such an extended MDR pattern may have contributed to their success in the hospital/healthcare environment (Tenover et al. 2008). Moreover, the high similarity in gene content of both this Italian strain and the German isolates with the human ST8 SCCmec type IV Hannover EMRSA (Walther et al. 2009), suggests that a risk of spill back to humans cannot be ruled out. These ST8 horse isolates are also similar to the PVL-negative, ACME-negative epidemic ST8 USA500, but maintain the SaPI-encoded toxin genes. In this study, however, a transmission of the MRSA ST8 strain to humans or dogs was not demonstrated, and the only MRSA isolated from human samples was a CC1, ST1, spa type t127, SCCmec type IVa, a major lineage in humans and livestock in Italy (Battisti et al. 2010; Franco et al. 2011; Alba et al. 2015; Cortimiglia et al. 2015), also detected in milk and dairy products (Carfora et al. 2015).

In conclusion, this study indicates that a ST8 MRSA strain infection can represent a hazard for equine health, and supports the opinion that inappropriate administration of antimicrobials and hospitalization/surgery represent risk factors for MRSA colonization and infection in horses (Weese 2010; Maddox et al. 2015; Tirosh-Levy et al. 2015). The detection of MRSA from environmental samples collected both at the mare's stable and at the Veterinary Hospital also indicates that environmental contamination could play an important role in the dissemination and maintenance of MRSA strains overtime (Weese 2010; van Balen et al. 2014), highlighting the importance of applying effective biosecurity and infection control practices at both the stable- and hospital-level.

The Authors wish to thank Carmela Buccella, Gessica Cordaro, Paola Di Matteo, Valentina Donati, Fabiola Feltrin, Manuela Iurescia, Angela Ianzano, for their outstanding technical assistance.

FUNDING

This work was partially supported by the EMIDA ERA-NET (FP7- KBBE) European Union first transnational research call funding, Project ‘LA-MRSA’ (acronym title), ID 68.

Conflict of interest. None declared.

REFERENCES