Methicillin-resistant Staphylococcus aureus carriage among veterinary staff and dogs in private veterinary clinics in Hokkaido, Japan

Authors


ABSTRACT

To explore the prevalence and molecular characteristics of methicillin-resistant Staphylococcus aureus (MRSA) in veterinary medical practices, MRSA carriage was tested among 96 veterinarians (Vets), 70 veterinary technicians (VTs) and 292 dogs with which they had contact at 71 private veterinary clinics (VCs) in Hokkaido, Japan. MRSA isolates were obtained from 22 Vets [22.9%] and 7 VTs [10%]. The prevalence of MRSA among Vets was as high as that found in an academic veterinary hospital in our previous study. In contrast, only two blood donor dogs and one dog with liver disease (1.0%, 3/292) yielded MRSA. All MRSA-positive dogs were reared or treated in different VCs, in each of which at least one veterinary staff member carrying MRSA worked. Sequence types (ST) identified by multilocus sequence typing, spa types, and SCCmec types for canine MRSA isolates (ST5-spa t002-SCCmec II [from two dogs] or ST30-spa t021-SCCmec IV [from a dog]) were concordant with those from veterinary staff members in the same clinics as the MRSA-positive dogs, with which they had potentially had contact. Most MRSA isolates from veterinary staff were the same genotype (SCCmec type II and spa type t002) as a major hospital-acquired MRSA clone in Japan. The remaining MRSA was the same genotypes as domestic and foreign community-associated MRSA. Measures against MRSA infection should be provided in private VCs.

List of Abbreviations
CA-MRSA

community-associated MRSA

CI

confidence interval

HA-MRSA

hospital-acquired MRSA

MIC

minimal inhibitory concentration

MLST

multilocus sequence typing

MRSA

methicillin-resistant Staphylococcus aureus

ST

sequence types

VC

veterinary clinic

VT

veterinary technician

Vet

veterinarian

Methicillin-resistant Staphylococcus aureus is an important pathogen in both human and veterinary medicine [1, 2]. MRSA infection or carriage among companion animals has been recognized, especially in countries where MRSA is widespread in human hospitals [3]. Furthermore, MRSA carriage is potentially an occupational risk for veterinary personnel [4, 5]. Previously, in a study in an academic veterinary hospital, we found that veterinary staff members who are MRSA carriers can be a source of infection of animals, and that MRSA animal patients may subsequently transmit MRSA infection to other staff members [6]. It has also been reported that MRSA infections of three dogs and a cat were caused by the same MRSA clone as was isolated from a veterinary staff member in an Australian veterinary hospital [7].

In this study, we investigated the prevalence and molecular characteristics of MRSA isolates among Vets and VTs at 71 private VCs in Japan, where a majority of dogs and cats receive veterinary care. In addition, MRSA carriage in dogs that had been in contact with veterinary staff members was also tested.

MATERIALS AND METHODS

Sampling

Nasal swabs for the isolation of MRSA were collected using Seedswab γ No. 1 ‘‘Eiken” (Eiken Chemical, Tokyo, Japan) from 96 Vets and 70 VTs who were working in 71 private VCs in the Ishikari region around Sapporo in Hokkaido, Japan between April and June 2008. These 71 VCs represent 35% of the VCs (n = 202) owned by members of the Sapporo Veterinary Association in this region. Volunteers were recruited through the Sapporo Veterinary Association to provide samples for the isolation of MRSA. These veterinary staff members predominantly delivered veterinary care to dogs and cats. In addition, buccal mucosal samples were collected from 225 dogs that were examined and/or treated (including for vaccination and prevention of filariasis), 35 dogs that were reared at the VCs as blood donors, 13 dogs that were reared by veterinary staff members at home, and 19 dogs that attended the VCs for other non-medical care purposes, such as grooming or staying in a kennel. This study was approved by the Ethics Committee of the Graduate School of Dairy Science, Rakuno Gakuen University.

Isolation and identification of MRSA

For enrichment, swabs were inoculated into 5 mL of heart infusion broth (Nissui Pharmaceutical, Tokyo, Japan) containing 7.5% sodium chloride and incubated at 35 °C for 48 hr. Enrichment cultures were streaked out on CHROMagar MRSA (Kanto Kagaku, Tokyo, Japan) and incubated at 35 °C for 24 hr. Representative colonies colored light purple (presumptive MRSA), blue or white were selected from each sample and subcultured on nutrient agar (Nissui Pharmaceutical). Isolates were subjected to Gram staining, catalase testing and coagulase testing in tube tests with rabbit plasma (Eiken Chemistry, Tokyo, Japan). DNA was extracted from cultures using an InstaGene Matrix (Bio-Rad Laboratories, Tokyo, Japan). Panton-Valentine leukocidin (pvl) [8], mecA [9] and femA (specific to S. aureus) [10] genes were tested by PCR. mecA-positive isolates were stored in a Microbank (Pro-Lab Diagnostics, Richmond Hill, Canada) at −80 °C. Both mecA- and femA-positive isolates were identified as MRSA and subjected to the tests described below.

SCCmec typing, spa-typing and MLST

SCCmec typing was performed by PCR amplification of the mec (classes A–C) and ccr (types 1, 2, 3 and 5) regions [11]. Atypical elements were confirmed by DNA sequencing. In addition, the structure of SCCmec was determined using Oliveira's strategy [12]. All MRSA isolates were typed by DNA sequence analysis of the spa X region, as previously described [13]. MLST was conducted for representative MRSA isolates [14]. PCR products purified using a High Pure PCR Cleanup Micro Kit (Roche Diagnostics GmbH, Mannheim, Germany) were subsequently sent to FASMAC (Kanagawa, Japan) for DNA sequencing. spa types were determined with reference to the Ridom SpaServer (http://www.spaserver.ridom.de/), and the ST was determined from an MLST website (saureus.mlst.net/).

Antimicrobial susceptibility testing

Minimal inhibitory concentrations were determined by the broth micro-dilution method with a Frozen plate ‘‘Eiken” (Eiken Chemistry). The following antimicrobials were tested: oxacillin (breakpoint, 4 μg/mL), cefazolin (32 μg/mL), cefotiam (32 μg/mL), imipenem (16 μg/mL), streptomycin (64 μg/mL), kanamycin (64 μg/mL), gentamicin (16 μg/mL), arbekacin (not applicable), erythromycin (8 μg/mL), tetracycline (16 μg/mL), minocycline (16 μg/mL), chloramphenicol (32 μg/mL), ciprofloxacin (4 μg/mL), vancomycin (16 μg/mL), teicoplanin (32 μg/mL), quinupristin–dalfopristin (4 μg/mL) and linezolid (8 μg/mL). Breakpoints were adopted according to the Clinical and Laboratory Standards Institute guidelines [15]. For streptomycin and cefotiam, each intermediate MIC of bi-modal distribution was defined as the breakpoint in this study.

RESULTS

Isolation of MRSA

Methicillin-resistant Staphylococcus aureus isolates were obtained from 22 Vets [22.9%] and seven VTs [10%]. In addition, two blood donor dogs (from VC-39 and VC-57) and one dog patient (from VC-52) yielded MRSA isolates (Table 1). The MRSA-positive dog had liver problems and was being treated with ampicillin and enrofloxacin. This dog had repeatedly received various antimicrobials. MRSA isolates were also obtained from one of two veterinary staff in VC-39, one of two Vets in VC-52, and both a Vet and a VT in VC-57.

Table 1. Methicillin-resistant Staphylococcus aureus carriage among veterinary staff and dogs in Japan
Origin of sampleNo. +No. testedIsolation rate (%)CI95%
  • CI95%, 95% confidence interval.
  • Dog patients included those admitted for vaccination and prevention of filariasis.
  • Dogs brought to veterinary clinics for purposes other than veterinary care.
Vets2296(22.92%)14.95–32.62
VTs770(10%)4.12–19.53
Dog patients1225(0.44%)0.01–2.45
Blood donor dogs235(5.71%)0.70–19.16
Dogs owned by veterinary staff013 0–20.58
Healthy dogs019 0–14.59

Molecular characteristics of MRSA isolates

SCCmec and spa types for 32 MRSA isolates from veterinary staff members and dogs, and STs according to MLST of 16 MRSA isolates, are shown in Table 2. One MRSA isolate from each genotype based on SCCmec subtype (by Oliveira's strategy) and spa type was selected for MLST. Moreover, all isolates obtained from VCs where MRSA was isolated from a dog were tested by MLST.

Table 2. Molecular characteristics and antimicrobial resistance of 32 MRSA isolates from veterinary clinics in Japan
VC No.OriginSCCmecLocus amplifed by the Oliveira strategyST by MLSTspa typeAntimicrobial resistance pattern
  • Dog-2 in VC-57 and VC-39 were blood donor dogs.
  • Dog-5 in VC-52 was a dog patient with liver disease.
  • §The center of the ccrA2-ccrB genes was deleted.
  • The inserted gene was detected in class B mec complex gene. CEZ, cefazolin; CPFX, ciprofloxacin; CTM, cefotiam; EM, erythromycin; GM, gentamicin; IPM, imipenem; KM, kanamycin; MINO, minocycline; MPIPC, oxacillin; ND, not done; SM, streptomycin; TC, tetracycline.
57Vet-1IIkdp + dcs + pUB110 + mecI5t002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
 VT-1IIkdp + dcs + pUB110 + mecI5t002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
 Dog-2IIkdp + dcs + pUB110 + mecI5t002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
52Vet-1IIkdp + dcs + pUB110 + mecI5t002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
 Dog-5IIkdp + dcs + pUB110 + mecI5t002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
39Vet-1IVdcs30t021MPIPC, SM, KM, GM, EM
 Dog-2IVdcs30t021MPIPC, CEZ, SM, KM, GM, EM
43Vet-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
 VT-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
70Vet-2II§kdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, IPM, KM, EM, TC, MINO, CPFX
 VT-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
50VT-1IIdcs + mecI2690t002MPIPC, CEZ, KM, GM, EM, TC, CPFX
 Vet-1IVdcs380t1852MPIPC, KM, GM, EM, CPFX
 Vet-2IVdcsNDt1852MPIPC, CEZ, KM, GM, EM, CPFX
 VT-2IVdcsNDt1852MPIPC, KM, GM, CPFX
33Vet-1IIkdp + dcs + pUB110 + mecI5t067MPIPC, CEZ, KM, EM, TC, CPFX
 Vet-2IIkdp + dcs + pUB110 + mecINDt067MPIPC, CEZ, CTM, IPM, KM, EM, CPFX
5Vet-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, IPM, KM, EM, CPFX
18Vet-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
40Vet-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, SM, KM, EM, TC, MINO, CPFX
41Vet-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, KM, EM, TC, MINO, CPFX
56Vet-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, KM, EM, CPFX
67VT-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, IPM, KM, EM, TC, CPFX
68Vet-1IIkdp + dcs + pUB110 + mecINDt002MPIPC, CEZ, CTM, IPM, KM, EM, CPFX
10Vet-1IIkdp + dcs + mecI5t002MPIPC, CEZ, CTM, KM, GM, EM, TC, MINO, CPFX
30Vet-1IIkdp + dcs + mecINDt002MPIPC, CEZ, EM, TC, CPFX
61Vet-1IIkdp + dcs + mecINDt002MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, CPFX
37Vet-1IIkdp + dcs + pUB110 + mecI5t045MPIPC, CEZ, CTM, IPM, KM, GM, EM, TC, MINO, CPFX
19Vet-1IIkdp + dcs + mecI5t458MPIPC, CEZ, CTM, IPM, EM, CPFX
24Vet-1IIkdp + dcs + mecI5t8023MPIPC, KM, GM, EM, TC, MINO, CPFX
29Vet-1IVdcs + pUB11072t324MPIPC, CEZ, CTM, KM, EM
64VT-1IVdcs1516t1852MPIPC, KM, GM, CPFX

Twenty-five MRSA isolates [25/32, 78.1%] and seven MRSA isolates [7/32, 21.9%] were classified as SCCmec type II and IV, respectively. Absence of pUB110 and/or kdp was detected by the Oliveira strategy in six SCCmec type II MRSA isolates (Table 2). One MRSA isolate (S270) from a Vet of VC-70 had an atypical type 2 ccr complex. The PCR amplicon of this ccr gene was approximately 600 bp using primers α2 and βc, whereas a 937 bp product was obtained from isolates with a typical type 2 ccr complex. Strain S270 had a deletion of 326 bp (from position 64,942 to 65,267 in GenBank no. BA000018) in the center of the ccrA2-ccrB genes as compared with MRSA strain N315. The remaining sequence of the ccr genes (349 bp of ccrB and 157 bp of ccrA2) was identical to that of N315. Therefore, strain S270 was classified as SCCmec type II. Strain S93 (from Vet-1 of VC-29) harbored the larger class B mec complex, which was identified as an atypical class B mec complex of MRSA (GenBank no. EF596937). This MRSA isolate was classified as SCCmec type IVA because it contained pUB110 (Table 2).

Twenty-five SCCmec type II MRSA isolates were divided into spa type t002 [20/25, 80%; including two canine isolates], t067 [2/25, 8%], t045 [1/25, 4%], t458 [1/25, 4%] and new t8023 (26–23-17–12-17–16-17–12-17–16) [1/25, 4%]. Ten SCCmec type II MRSA isolates were typed as ST5 (1–4-1–4-12–1-10), regardless of spa type (t002, t045, t067, t458 and t8023). One SCCmec type II MRSA isolate with spa type t002 was typed as a new type of ST2690 (1–4-1–4-12-11-10). Two SCCmec type II MRSA isolates from a dog patient in VC-52, and a blood donor dog in VC-57 were ST5 with spa type t002 (Table 2).

Two SCCmec type IV MRSA isolates from a Vet and a blood donor dog of VC-39 had a t021 spa type and were classified as ST30. Four SCCmec type IV MRSA isolates with t1852 spa type were obtained from veterinary staff members of VC-50 and VC-64. Of these, one isolate from VC-50 was typed as ST380 (3–3-61-42–4-4–3) and one isolate from VC-64 as ST1516 (3–3-1-42–4-4–3). The SCCmec type IVA isolate (S93) had a t324 spa type was typed as ST72. No MRSA isolates harbored the pvl gene.

Antimicrobial resistance

The antimicrobial resistance patterns of the 32 MRSA isolates are shown in Table 2. All MRSA isolates were susceptible to arbekacin (MIC range, 0.25–4 μg/mL), chloramphenicol (4–16 μg/mL), vancomycin (0.25–2 μg/mL), teicoplanin (≤0.125 to 2 μg/mL), quinupristin–dalfopristin (0.25 μg/mL) and linezolid (1–4 μg/mL). All SCCmec type IV MRSA isolates were susceptible and 17/25 SCCmec type II MRSA isolates [68%] were resistant to imipenem. All four MRSA isolates that had spa type t1852 were resistant to ciprofloxacin, whereas the other SCCmec type IV MRSA isolates were susceptible to this drug.

DISCUSSION

The rate of MRSA found in this study among Vets [22.92%, CI95% 14.95–32.62%] in 22 of 71 private VCs is similar to that among Vets in an academic veterinary hospital reported in our previous study [23.5% and 25%] [6]. Thus, the prevalence of MRSA carriers among Vets in both these studies in Japan is higher than that in Denmark [3.9%] [16] and Scotland [3.1%] [17].

On the other hand, MRSA isolates were obtained from only three of 292 dogs (two blood donor dogs and one dog patient with liver problems). The three MRSA-positive dogs were treated or reared in three different VCs and at least one MRSA human carrier was detected in each of these VCs. STs based on MLST, spa types and SCCmec types of the dog MRSA isolates were concordant with those of the MRSA isolate(s) from veterinary staff member(s) in the same VC. We therefore suggest that MRSA can be transmitted between veterinary staff and dogs.

The percentage of MRSA carrier among dogs (1.03%, CI95% 0.02–2.97%) in our study is similar to reported percentages in previous studies in Japan [1.75%] [18], the UK [3.2%] [19] and Ireland [1.1%] [20]. MRSA carriage in dogs appears to be rare, despite the opportunities for dogs to have contact with MRSA carriers in veterinary medical practices in Japan. In our study, we collected only buccal mucosal samples from dogs in order to detect MRSA. However, Hanselman et al. reported detecting S. aureus in both nasal and rectal swab samples from only one of 19 dogs [5.3%] that carried S. aureus in nasal and/or rectal areas [21]. Another study in which Staphylococcus pseudintermedius was investigated reported that the most frequent staphylococcal carriage sites in dogs were the perineum [66%] and the mouth [65%], followed by the nose [27%] [22]. Therefore, the percentage of MRSA carrier among dogs in our study may be lower than the true prevalence.

Twenty-five of 32 MRSA isolates from VCs were classified as SCCmec type II, of which 20 had a t002 spa type. SCCmec type II and spa type t002 have often been observed in Japanese human hospitals [23-25]. Most MRSA isolates from an academic veterinary hospital in Japan were also classified as SCCmec type II and had spa type t002 [6]. The present study shows that the HA-MRSA clone has spread widely among veterinary medical practices, including primary care practices, in Japan.

In one VC (VC-50), three of four veterinary staff members who were subjected to nasal sampling carried spa type t1852 MRSA clone with SCCmec type IV. This MRSA clone potentially circulated between veterinary staff members in this VC. Previously, MRSA isolates in Germany and Japan were registered as spa type t1852 in SpaServer. One of the spa type t1852 MRSA isolates in our study was typed as ST380 (3–3-61-42–4-4–3), a ST type that had also been reported in a 9-year-old boy in Aichi, Japan as a CA-MRSA in S. aureus MLST site [26].

A specific MRSA clone (SCCmec type IV, ST30, spa type t021) was identified in both a Vet and a blood donor dog in the same VC (VC-39). MRSA of the same genotype has previously been reported in Norway [27] and Spain [28]. ST72 MRSA with a spa type t324 was also detected in a Vet; this was classified as SCCmec IVA and contained an atypical class B mec complex. The same genotype had also been detected in hospitals and community in Korea [29, 30]. All MRSA isolates detected in this study exhibited molecular characteristics identical to HA-MRSA or CA-MRSA among humans in Japan or abroad. Although information concerning overseas travel among veterinary staff members was not obtained, it is thought that a CA-MRSA clone was introduced into the Japanese community, including veterinary medical practices, from abroad. Otsuka et al. mentioned that multiple MRSAs, including genetically typed CA-MRSA, circulate in the community in Japan [31].

All MRSA isolates were susceptible to antimicrobials used to treat MRSA or vancomycin-resistant enterococcal infections; these include vancomycin, teicoplanin, arbekacin, quinupristin–dalfopristin and linezolid, all of which are important antimicrobials in human medicine.

In conclusion, there is high prevalence of MRSA carriers among Vets in primary care veterinary clinics in Japan. The molecular characteristics of canine MRSA isolates were concordant with those of MRSA isolates from veterinary staff members who had contact with MRSA-positive dogs. Without exception, human MRSA carrier(s) were identified in the same VCs as those in which the MRSA-positive dogs were reared or treated. Measures against MRSA infection should be implemented in private VCs.

ACKNOWLEDGMENTS

We thank the veterinary staff members for providing samples and Dr. T. Nomura of Sapporo Veterinary Association for assisting us with collecting these samples. This work was supported in part by JSPS KAKENHI Grant Number 23790671.

DISCLOSURE

The authors declare that they have no conflicts of interest.

Ancillary