Abstract
Between July 2011 and May 2016, a total of 40 Staphylococcus aureus strains originating from 36 horses were confirmed as methicillin resistant (methicillin-resistant Staphylococcus aureus [MRSA]) in a university equine clinic. An additional 10 MRSA strains from 36 samples of clinic workers were obtained in October 2017. The first equine isolate represented the sequence type ST398, spa-type t011, and SCCmec IV. This isolate was resistant to a wide spectrum of antimicrobial agents. MRSA strains with the same genotype and with very similar resistance profiles were isolated on 21 more occasions from September 2013 to September 2014. A second outbreak occurred from May 2015 until May 2016. The first isolate in this second outbreak shared the same genotype, but was additionally resistant to chloramphenicol. The second isolate from August 2015 also showed resistance to rifampicin. The clone was isolated 18 times. Most of the human isolates shared the same genotype as the isolates from horses and their resistance patterns showed only slight differences. We can conclude that the MRSA-related cases at the Department and Clinic of Equine Medicine were all nosocomial infections caused by the same clonal lineage belonging to the clonal complex 398. The clonal complex 398 of equine origin is reported for the first time in Hungary. In addition, our observation of the emergence of new resistance to antimicrobial agents within the clonal lineage after treatment with antibiotics is of concern. Strict hygiene regulations have been introduced to lower the incidence of MRSA isolation and the related clinical disease.
Introduction
Methicillin-resistant Staphylococcus aureus (MRSA) infections are currently a serious challenge to equine medicine. 1 Infections are difficult to treat due to their resistance to several classes of antibiotics used in horses. In addition, both horses and people working with horses can be healthy carriers of MRSA, representing a reservoir for strains causing infections.2–5 Although the first reports of horse-related MRSA infections mostly consisted of sporadic cases in equine hospitals,6,7 nosocomial infections with this pathogen have become frequent all over the world during the past 20 years.1,8 The concomitant increase in case reports from private veterinary practices can be also related to this tendency. 9 Specific MRSA clones belonging to the same clonal complexes as those found in humans, including CC1, CC5, CC8, CC22, CC59, CC88, and CC398, have been identified in equine infection cases in many European countries.2,3,10–14 In July 2011, a horse hospitalized in the Department and Clinic of Equine Medicine (DCEM), Üllő, Hungary, developed a postoperative intraabdominal infection caused by MRSA. Since then, the number of MRSA-related clinical cases at the DCEM has increased year by year, raising the question of whether this was due to the nosocomial spread of a specific MRSA clone or to variable individual strains. Due to the lack of data on Hungarian MRSA of equine origin, neither scenario could be excluded and thus a meticulous investigation into the matter was needed. The aim of the study was to evaluate the nosocomial infections caused by MRSA in a clinical equine environment over a 3-year period by investigating the genetic variability and antimicrobial susceptibility of a conserved set of bacterial isolates. Another goal was to determine whether the clinic personnel harbored the same MRSA clones.
Materials and Methods
Samples were continuously collected from horses hospitalized in the DCEM for routine bacteriological culture between July 2011 and May 2016, and isolates that represented cefoxitin-resistant, hemolytic Staphylococcus spp. were entered into this study. 15 The sampling sites where cefoxitin-resistant, hemolytic Staphylococcus spp. were found included postoperative wound infections (n = 22), abscesses (n = 3), thrombophlebitis (n = 3), postoperative intraabdominal infections (n = 2), conjunctival infections (n = 2), pleuritis and pneumonia (n = 2), and routine nasal swabs or tracheal wash samples (n = 9).
In October 2017, a voluntary screening was performed on the clinic personnel to investigate the potential occupational risk groups for nasal carriage. Samples were collected from 36 out of 39 people, including veterinary internists (n = 6), surgeons (n = 6), technicians (n = 15), and other members of management who had no direct contact with horse patients (n = 9). All human nasal samples were collected from both nares on 1 day at the beginning of the shift, before entering the clinic's operative area, using nasal swabs premoistened with sterile physiological saline solution (0.85% NaCl).
All volunteers participated spontaneously in the study and provided written informed consent. The sampling protocol was supervised and approved by the Committee of Science and Research Ethics, Medical Research Council, Ministry of Human Capacities, Hungary, as a protocol that does not require medical intervention (No. 10214-2/2019/EKU).
Sample processing, isolation, and identification of MRSA
Bacterial cultures of the equine samples were obtained using routine diagnostic procedures at the Diagnostic Laboratory, Department and Clinic of Production Animal Medicine, University of Veterinary Medicine Budapest, Üllő, Hungary, on Columbia agar plates containing 5% sheep blood (Biolab, Budapest, Hungary) after 24–48 hr incubation at 37°C. Individual colonies were subcultured under identical conditions to yield a pure culture of the bacteria. The primary identification of pure cultures consisted of Gram staining and catalase and oxidase tests. 15
The prediction of methicillin resistance was routinely tested in case of each presumptive Staphylococcus sp. isolate by disc diffusion using a cefoxitin disk as recommended in the standards of the Clinical and Laboratory Standards Institute (CLSI). 16 Cefoxitin-resistant strains that produced an alpha- or both an alpha- and beta-hemolysis on Columbia sheep blood agar were reported as hemolytic methicillin-resistant Staphylococcus sp. and were frozen at −80°C. S. aureus was distinguished from non-S. aureus isolates by subsequent identification by matrix-assisted laser desorption/ionization/time-of-flight (MALDI TOF) mass spectrometry (Bruker Daltonik GmbH, Leipzig, Germany) at the Institute of Veterinary Bacteriology, Vetsuisse Faculty, University of Bern, Bern, Switzerland. Non-S. aureus isolates were excluded from further investigation.
Human nasal swabs underwent a previously described MRSA pre-enrichment protocol, using Muller-Hinton broth (Biolab) with 6.5% NaCl. 3 After incubation (24 hr at 37°C), a loopful (∼10 μL) of the pre-enrichment medium was spread onto chromogenic agar plates for selective isolation (Brilliance MRSA 2 Agar; Oxoid, Basingstoke, United Kingdom). Pure cultures of presumptive MRSA colonies were stored at −80°C until further evaluation. Cefoxitin-resistant strains of both equine and human origin were confirmed as MRSA by MALDI TOF mass spectrometry for species identification and by the presence of the mecA gene by PCR. 17 The total number of staphylococci in the survey period was determined retrospectively, according to the case reports of the Diagnostic Laboratory, Üllő.
Antimicrobial susceptibility testing and resistance genes
The minimal inhibitory concentration (MIC) of 19 selected antibiotics were determined by microdilution in Mueller-Hinton broth using custom Sensititre susceptibility plates EUST (Thermo Fisher Scientific, Waltham, MA; MCS Diagnostics BV, Swalmen, The Netherlands) as recommended by the CLSI. 18 The resistance breakpoints from the CLSI document M100 for human isolates were used for clindamycin (≥4 mg/L), tetracycline (≥16 mg/L), penicillin (≥0.25 mg/L), chloramphenicol (≥32 mg/L), kanamycin (≥64 mg/L), quinupristin/dalfopristin (≥4 mg/L), vancomycin (≥16 mg/L), gentamicin (≥16 mg/L), trimethoprim (≥16 mg/L), erythromycin (≥8 mg/L), ciprofloxacin (≥4 mg/L), cefoxitin (≥8 mg/L), linezolid (≥8 mg/L), and sulfamethoxazole (≥512 mg/L) as indicators of a possible acquired resistance and not for clinical purposes. 19
For fusidic acid (>1 mg/L) and rifampicin (>0.5 mg/L), the resistance breakpoints of the European Committee on Antimicrobial Susceptibility Testing were used. 20 For streptomycin, mupirocin, and tiamulin, for which no breakpoints exist, the isolates were further evaluated for the presence of a resistance mechanism if the MIC was higher than the lowest concentration measured (≤0.5 mg/L for tiamulin and mupirocin and ≤4 mg/L for streptomycin).
Phenotypic resistance was confirmed by the genetic identification of the mechanism. Antibiotic resistance genes were identified using a microarray (AMR+ve-5.1 tubes; Alere Technologies GmbH; Jena, Germany) capable of detecting up to 117 different antibiotic resistance genes known to be present in Gram-positive bacteria.21,22 Mutations of topoisomerase GyrA and GrlA and mutations in the RNA polymerase RpoB were investigated by amino acid analysis of translated DNA sequences obtained by PCR and Sanger sequencing.23–26
Genotyping of MRSA
Genetic relatedness was confirmed by the use of variable numbers of tandem repeats (VNTRs) typing, spa-typing, and multilocus sequence typing (MLST).27–29 The spa type was analyzed using the Ridom StaphType software (Ridom StaphType; Ridom GmbH, Würzburg, Germany). The sequence types (STs) were determined using the Multilocus Sequence Typing home page (http://saureus.mlst.net).
Results
Forty MRSA strains were identified from 43 cefoxitin-resistant, hemolytic Staphylococcus sp. strains between July 2011 and May 2016 at the DCEM. The MRSA strains originated from different skin and soft tissue infection sites or from routine nasal swabs and tracheal wash samples of 36 equine patients. The three cefoxitin-resistant non-S. aureus strains consisted of two S. haemolyticus strains and one S. sciuri-like strain, which were both isolated from the nasal swab samples of two horses. These latter three isolates were excluded from further evaluation.
The isolation frequency and main characteristics of the total of 87 horse-related staphylococci obtained during the survey period are presented in Fig. 1. An additional 10 MRSA strains were isolated from nasal swabs of 36 clinic staff members in October 2017. Of these staff members, two were veterinary internists, six were technicians, and two were from the management team of the DCEM.

Occurrence and frequency of equine MRSA ST398-t011 and other Staphylococcus sp. isolates during the survey period. Each column represents the number of isolates per month, different column patterns indicate different phenotypic or resistance patterns:
Staphylococcus sp. isolates (not further investigated),
cefoxitin-resistant, hemolytic, non-S. aureus Staphylococcus sp. (not further investigated),
MRSA isolates resistant to penicillin, cefoxitin, tetracycline, trimethoprim, streptomycin, gentamicin, kanamycin, and ciprofloxacin,
MRSA isolates additionally resistant to chloramphenicol,
MRSA isolates susceptible to chloramphenicol and streptomycin,
MRSA isolates additionally resistant to chloramphenicol and rifampicin,
period without MRSA isolates,
inclusion of chloramphenicol and rifampicin into the antimicrobial treatment protocol. MRSA, methicillin-resistant Staphylococcus aureus.
Frequency, genotype, and antimicrobial susceptibility of equine isolates
All the equine MRSA isolates belonged to ST398, spa type t011, and contained the same SCCmecIV element. In addition, they all exhibited the same VNTR profile, thus confirming clonality. The strains differed according to their resistance profile, which changed remarkably over time (Fig. 1; Table 1). The first MRSA clone identified exhibited resistance to penicillin (blaZ), cefoxitin (mecA), trimethoprim (dfrK), tetracycline [tet(M)], streptomycin (str), gentamicin and kanamycin [aac(6′)-Ie-aph(2′′)-Ia], and ciprofloxacin (GyrA-Ser83Leu and GrlA-Tyr80Phe). This clone was detected once in July 2011 and only reappeared 2 years later in September 2013 in a case of conjunctival infection, and was then associated with 21 more infections over the next year until September 2014 (Fig. 1). All, but two of the isolates from the first outbreak showed the same resistance pattern (Table 1). One isolate from July 2014 had additional resistance to chloramphenicol (MIC > 64 mg/L, [catpC221]) and the last isolate of the first outbreak was susceptible to streptomycin (MIC = 8 mg/L, str negative). No MRSA strains were isolated within the next 8 months.
Genetic Characteristics and Antibiotic Resistance Profile of Methicillin-Resistant Staphylococcus aureus Isolates from Horses and Humans in Hungary
The MIC breakpoints determining resistance were those recommended for S. aureus in CLSI supplement, 19 except for rifampicin for which breakpoint from European Committee on Antimicrobial Susceptibility Testing was used. 20
NA, no breakpoint was available for streptomycin and MIC was given for strains that did not contain the str gene, while strains containing an str gene had MIC >32 mg/L.
One isolate had both His481Asn and Ser529Tyr mutations in the RpoB protein sequence.
S, susceptible to the antibiotic.
VNTR, variable-number tandem repeat analysis; MLST, multilocus sequence typing; SCCmec, Staphylococcus cassette chromosome mec; PEN, penicillin; TET, tetracycline; TMP, trimethoprim; GEN, gentamicin; KAN, kanamycin; CIP, ciprofloxacin; STR, streptomycin; RIF, rifampicin; CHL, chloramphenicol; ERY, erythromycin; CLI, clindamycin; MIC, minimal inhibitory concentration.
A second MRSA outbreak occurred in May 2015. The first MRSA ST398-t011-SCCmecIV isolate from this outbreak (May 2015–May 2016) had additional resistance to chloramphenicol (MIC > 64 mg/L, [catpC221]) and the second isolate was also resistant to rifampicin (RpoB-His481Asn). This clone was recovered from 17 other infections in the year following the second outbreak. During that time, only one MRSA ST398-t011-SCCmecIV was found to have a different resistance pattern; it was susceptible to chloramphenicol (MIC < 8 mg/L) and streptomycin (MIC = 8 mg/L). Of note, the emergence of MRSA ST398-t011-SCCmecIV exhibiting additional resistance to chloramphenicol and rifampicin became frequent after these two antibiotics were implemented in the MRSA therapy protocols (Fig. 1). It should be noted that the owners of all horses that received chloramphenicol and/or rifampicin declared in the passports for their horses that the animals were excluded from slaughter for human consumption.
From the 10 MRSA strains isolated from human nasal swabs, 7 had the genotype ST398-t011-SCCmecIV and shared the same VNTR pattern demonstrated in MRSA strains of equine origin. Only slight differences could be observed among the strains’ resistance profiles (Table 1). Based on genotyping, three human isolates differed both from other human and equine strains. One was isolated from a technician and had the genotype ST398-t1580-SCCmecIV, but otherwise showed the same wide-scale resistance pattern as the aforementioned genotype. A third genotype, ST541-t1250-SCCmecV, was isolated from two members of management who worked occasionally in the stables, but not directly with horses. Both isolates shared common VNTR and antimicrobial resistance patterns markedly differing from those observed in the case of the ST398 strains (Table 1).
Discussion
Forty MRSA strains isolated between July 2011 and May 2016 from clinical specimens and nasal swabs from hospitalized horses were analyzed in detail to evaluate the possibility of nosocomial infection. All equine isolates shared the same profile using VNTR typing, a method that has been shown to be consistent with pulsed-field gel electrophoresis typing and MLST for outbreak analysis. 27 They also belonged to the same clonal lineage ST398, which has not been previously reported in horses from Hungary, but has been found widespread in many European countries.8,30
As an equine nosocomial pathogen, ST398 MRSA was first reported in an Austrian university veterinary hospital in 2006. Over the past 10 years, it has become prevalent in many European equine health care settings, either newly emerging or successfully replacing other MRSA lineages.2,3,13,31–33 Although the clonal lineage was initially associated with swine production,34,35 later studies revealed a specific horse clinic-related lineage with spa-type t011, SCCmec-type IV, and gentamicin resistance.2,3 The outbreaks at the DCEM could also be attributed to a clonal lineage of t011-SCCmecIV MRSA isolates, which were all resistant to gentamicin, further supporting the global spread of a specific nosocomial clone within horse clinical environments.
In this horse setting, human asymptomatic carriage is known to represent a potential risk for human-to-horse-to-human transmission and horse infection with livestock-associated MRSA.3,4,7,10,14 To investigate human nasal carriage and identify risk groups within the staff, a voluntary sampling session was performed in October 2017, which revealed that carriers consisted mainly of internists and technicians, and none of the surgeons were positive for MRSA. The difference could possibly be explained by the nature and frequency of patient contact and contaminated areas.3,36,37 Internists and technicians are in close daily contact with the horses and work in the stables and examination area of the DCEM. Surgeons usually meet their patients for short-term periods (e.g., examinations) and mostly treat them in the clean operating halls.
One of the two distinct genotypes found in the human samples was from a technician. The ST and SCCmec type corresponded to that of equine origin ST398 and SCCmecIV, respectively. The spa type (t1580; 08-02-25-34-24-25) differs in only one repeat from t011 (08-16-02-25-34-24-25). The resistance patterns of the two genotypes had only minor differences (Table 1) and thus it can be deduced that this new spa type emerged from the common t011 type by losing one repeat. Similar variations among nosocomial isolates have been reported. 3 The other genotype ST541-t1250 markedly differed according to its genetic characteristics and resistance pattern as well. This genotype also belongs to the CC398 lineage and is rarely isolated from swine and human samples.38,39 The two carriers had occasionally worked in the operative area of the DCEM and in the stables of the Department and Clinic of Production Animal Medicine carrying out maintenance. This suggests a possible origin of the strains and at the same time warns of the risk of interspecies cross-contamination with CC398. 30
To overcome the nosocomial spread of MRSA, hand hygiene seems to be the key factor in stopping transmission; usage of disposable gloves, regular handwashing, and disinfection between patients can dramatically minimize the incidence of cases once MRSA becomes resident at a health care unit.4,40–42 Breaching strict hygiene regulations might also have contributed to the repeated outbreaks and long persistency of MRSA in the DCEM. Based on the results of our study, hygienic measures have been reinforced for all staff members and students by the mandatory use of gloves and masks, while treating horses, as well as hand disinfection. Decolonization of humans was not envisaged, since it has been shown that this might fail and recolonization occurs rapidly after treatment. 4
Horse-related subpopulations of MRSA ST398 seem to be highly clonal throughout Europe. 1 The primary source of MRSA introduction to the DCEM cannot be traced back for several reasons. Since the pathogen can colonize both humans and animals, neither route of introduction could be excluded. The lack of previous reports on horse-related MRSA ST398 in Hungary does not exclude the possibility of its former residence in the Hungarian equine population or even in other Hungarian equine health care settings. Staff members can also potentially transfer the agent between patients, not only within a clinic but also between veterinary institutions,4,41 since staff and student exchanges with other equine hospitals are common in university institutions.
Comparing to other MRSA ST398-t011 of equine origin reported so far, the resistance pattern of the isolates from the DCEM is strikingly wide and became wider with time (Table 1). Resistance to tetracycline encoded by tet(M) or rarely tet(K) is a common characteristic of LA-MRSA CC398 in general, 43 whereas gentamicin resistance [aac(6′)-Ie-aph(2′′)-Ia] is usual in the equine clinic-adopted subpopulations within the clonal complex. 1 Resistance to trimethoprim is also widespread among equine nosocomial lineages,4,31,34,43 whereas fluoroquinolone and streptomycin resistance are reported with lower frequencies with a variable genetic background.4,31,44 To date, only a few studies have included the susceptibility testing of chloramphenicol and rifampicin when investigating MRSA of equine origin and these studies reported a sporadic occurrence of resistance to either of them. 33
Although resistance to chloramphenicol already appeared in one isolate in July 2014 in the DCEM, it only became rife, along with rifampicin resistance, soon after these two drugs were introduced into the clinical treatment protocol in the autumn of 2014 (Fig. 1). In the highly clonal isolates of the DCEM, the newly emerging phenicol resistance could be attributed to catpC221, which may have been acquired from the resident Gram-positive flora of the clinic in response to the chloramphenicol selective pressure.45,46
The newly emerging resistance to rifampicin was due to nonsynonymous, substitution-forming mutations in the β-subunit of the RNA polymerase gene (rpoB). All of the rifampicin-resistant strains exhibited at least one amino acid substitution in the RpoB, mainly His481Asn (Table 1). A recent analysis of a worldwide collection of more than 7,000 S. aureus full genomes revealed that the His481Asn mutation is the most prevalent among rifampicin resistance-forming mutations and most likely to emerge under rifampicin selective pressure in vivo. 47 Despite this, very little is known about the occurrence of rifampicin resistance in MRSA of animal origin.48,49
The combination of chloramphenicol and rifampicin was introduced as last resort in the case of MRSA infections at the DCEM. In this study, we report a simultaneous emergence of resistance to both antimicrobial agents within a short timeframe, presumably due to the massive selective pressure of the antibiotic usage and possible inappropriate dosage. However, dosage records were not available to provide more data to explain the emergence of chloramphenicol and rifampicin resistance. The decreased number of therapeutic possibilities highlighted the importance of prudent antibiotic usage and urged the adaptation of alternative preventive and therapeutic methods in the clinic.
Conclusions
The first isolate of MRSA ST398-t011 at the Department and Clinic of Equine Medicine was detected in 2011. Over the following 6 years, MRSA strains were involved in mild to fatal clinical cases and gained resistance to further antimicrobial agents. According to previous documented observations, an early and meticulous characterization of the first few isolates could have helped us to install more efficient preventive measures. Nevertheless, our survey led to the implementation of routine representative screening of patients, monitoring changes in the antibiotic profile of MRSA isolates, and the implementation of strict hygiene regulations in an effort to lower the incidence of MRSA isolation and related clinical disease.
Footnotes
Acknowledgments
The authors thank Kerstin Cotting and Christian Strauss for their assistance and advice. Part of the study was performed during a traineeship of E.A. at the Institute of Veterinary Bacteriology of the University of Bern, Switzerland, in September 2016. The study was partially supported by the Institute of Veterinary Bacteriology, University of Bern, and by the European Union, and was co-financed by the European Social Fund (grant agreement no. EFOP-3.6.3-VEKOP-16-2017-00005, project title: Strengthening the scientific replacement by supporting the academic workshops and programs of students, developing a mentoring process) through the grant of E.A.
Disclosure Statement
No competing financial interests exist.
