Survey of Carbapenemase-Producing Enterobacteriaceae in Companion Dogs in Madrid, Spain

TEXT

The emergence and worldwide spread of carbapenemase-producing Enterobacteriaceae (CPE) have significantly increased in recent years, mainly in Klebsiella pneumoniae (1–3). Most carbapenemases are encoded on mobile genetic elements (4), which usually contain other antimicrobial resistance genes.

CPE and other carbapenemase-producing bacteria, which seemed limited to humans, have also been described in animals, particularly livestock. The detection of Pseudomonas aeruginosa with VIM-2 and Acinetobacter baumannii with OXA-23 and OXA-58 in livestock animals from Lebanon was recently reported (5). Several VIM-1 carbapenemase-producing Escherichia coli and Salmonella isolates were recovered from pigs in Germany (6, 7), Acinetobacter spp. harboring OXA-23 was isolated from cattle in France (8), and A. baumannii and Acinetobacter lwoffii isolates producing NDM-1 were detected in pig and chicken sources in China (9). In addition, NDM-1-producing Salmonella enterica subsp. enterica serovar Corvallis was reported in Germany from a wild bird (10). Only rarely have CPE been found in pets, specifically in three studies that detected OXA-48-producing E. coli and K. pneumoniae from dogs in Germany (11), NDM-1-producing E. coli from dogs and cats in the United States (12), and OXA-23-producing Acinetobacter spp. from horses in Belgium (13).

Since there is very little information about CPE in companion animals, we describe here a study aimed at assessing the prevalence of CPE in rectal swab samples from companion dogs.

Between October 2014 and January 2015, 160 rectal swab samples from different dogs who attended a veterinary hospital in Fuenlabrada (Madrid, Spain) were collected, without any inclusion or exclusion criteria, by inserting a swab into the rectum while rotating it. The swab was then transported to the laboratory in Amies medium. At the time of swab collection, 54% of the animals were under antibiotic treatment. The drugs used as monotherapy were, in order of frequency, amoxicillin-clavulanic acid (18.6%), cephalexin (15.1%), metronidazole (9.3%), marbofloxacin (8.1%), sulfadoxine-trimethoprim (SxT) (7%), ampicillin (4.7%), amoxicillin (3.5%), cefazolin (1.2%), nitrofurantoin (1.2%), and azithromycin (1.2%). Antibiotic combinations included metronidazole plus cephalexin (7%), amoxicillin-clavulanic acid plus marbofloxacin (5.8%), cephalexin plus marbofloxacin (3.5%), metronidazole plus marbofloxacin (3.5%), metronidazole plus marbofloxacin plus cephalexin (3.5%), metronidazole plus SxT (2.3%), amoxicillin-clavulanic acid plus metronidazole (1.2%), amoxicillin-clavulanic acid plus SxT (1.2%), marbofloxacin plus azithromycin (1.2%), and ampicillin plus marbofloxacin (1.2%).

Samples were inoculated on a selective chromogenic medium biplate (chromID Carba Smart, bioMérieux), which consists of an associated medium to screen for OXA-48 on one side and other carbapenemases on the other side. Plates were then incubated at 37°C for 24 h in aerobic atmosphere. Susceptibility testing and identification were performed on every isolate with growth on this medium by using the Wider system (Francisco Soria Melguizo).

Phenotypic carbapenemase characterization was conducted, according to the susceptibility testing results, on isolates that matched EUCAST screening criteria (14) by use of the combination disk test (Rosco), which includes meropenem (10 μg) alone and combined with boronic acid, cloxacillin, and dipicolinic acid. A temocillin disk (30 μg) was used as a phenotypic indicator of OXA-48.

In addition, isolates with growth on chromID Carba Smart but with negative results for carbapenemase production were further investigated for the presence of extended-spectrum β-lactamases (ESBL) and/or AmpC-type enzymes according to EUCAST guidelines (14). The phenotypic confirmation method used was the combination disk test, which included a disk containing cefotaxime (30 μg) alone and in combination with clavulanic acid, cloxacillin, and cloxacillin-clavulanic acid (Rosco). Furthermore, the double-disk synergy test, which included a disk containing cefepime and another one containing clavulanic acid, was performed to confirm the presence of ESBL in Enterobacteriaceae with inducible chromosomal AmpC.

The presence of genes encoding carbapenemases (bla_OXA-48, bla_KPC, bla_VIM, bla_IMP, and bla_NDM) (15), ESBLs (bla_TEM, bla_SHV, and bla_CTX-M) (16), and plasmid-mediated AmpC (bla_CMY, bla_FOX, bla_MOX, bla_DHA, bla_EBC, and bla_ACC) (17) was determined by use of PCR and DNA sequencing assays. The sequences obtained were compared with those available in the GenBank and Lahey Clinic (http://www.lahey.org/Studies/) public databases.

Multilocus sequence typing (MLST) was performed for all carbapenemase-producing isolates with the Institut Pasteur scheme (http://www.pasteur.fr/recherche/genopole/PF8/mlst/Kpneumoniae.html).

Class 1 integron structures, the integrase gene intI1 and the variable regions, were screened by PCR amplification and DNA sequencing (18). Conjugation experiments were performed with carbapenemase-producing K. pneumoniae isolates with the kanamycin-azide-resistant E. coli HB101 as a recipient. Putative transconjugants were selected on Mueller-Hinton agar plates containing kanamycin (100 μg/ml), azide (160 μg/ml), and cefotaxime (4 μg/ml). Plasmids were classified according to their incompatibility groups by a PCR-based replicon-typing scheme (19). Pulsed-field gel electrophoresis (PFGE) with S1 nuclease digestion of whole-genomic DNA (S1-PFGE) was used to detect plasmids as previously described (20).

Three Enterobacteriaceae from three different dogs grew on the selective chromogenic medium biplate: two K. pneumoniae (named 1 and 2) and one Enterobacter cloacae. Only K. pneumoniae 1 produced carbapenemase, and the remaining isolates were considered false positives on the chromID Carba Smart medium. Susceptibility testing results are listed in Table 1. The combination disk test showed a profile compatible with CPE in the K. pneumoniae 1 isolate confirmed by PCR and sequencing as a metallo-β-lactamase VIM-1, belonging to the new sequence type (ST) 2090, not related to any ST previously described. This isolate, also resistant to trimethoprim-sulfamethoxazole, tobramycin, and fosfomycin, was detected in a dog from an animal shelter receiving no medication at the time of sampling or in the previous month. Genes encoding ESBLs or plasmid-mediated AmpC were not detected in this isolate.

The bla_VIM-1 gene was carried on a class 1 integron in the following cassette combination: intI1 (integrase gene)-bla_VIM-1-aac(6′)-Ib (tobramycin resistance gene, also called aacA4)-dfrB1 (trimethoprim resistance gene)-aadA1 (streptomycin resistance gene)-catB2 (chloramphenicol resistance gene)-qacEδ1/sul1 (quaternary ammonium compounds resistance gene/sulfonamide resistance gene). Carbapenem-nonsusceptible E. coli transconjugant was obtained from VIM-1-producing K. pneumoniae isolates. Both the isolate from surveillance and the transconjugant carried a plasmid of ∼48 kb that was untypeable by PCR. In addition, the transconjugant was positive for class 1 integron and bla_VIM. VIM-1-producing E. coli transconjugant also presented reduced susceptibility to tobramycin and trimethoprim-sulfamethoxazole but not to fosfomycin.

The K. pneumoniae 2 strain was further investigated, as because the results showed that it seemed to harbor an ESBL and a plasmidic AmpC enzyme, confirmed as CTX-M-15 and DHA, respectively. It belonged to a dog that was being treated with marbofloxacin and amoxicillin-clavulanate.

The double-disk synergy test with cefepime and amoxicillin-clavulanic acid was positive for the E. cloacae strain, suggesting the ESBL production confirmed as SHV-12 by sequencing. No genes encoding carbapenemases were detected by PCR. The dog was receiving cephalexin and metronidazole at the time of sampling.

The isolation of multidrug-resistant Enterobacteriaceae in companion animals is an emerging phenomenon. Few studies describe colonization or infection of companion animals with these organisms, but the number seems to be increasing. According to a recent review (21), and similar to humans (22), CTX-M-type ESBLs are the most prevalent, particularly CTX-M-1, -14, and -15, although SHV-12 and TEM-52 have also been described, mostly in E. coli and K. pneumoniae strains from cats and dogs; CMY-2 and DHA are the most common AmpC β-lactamases described so far (21). Fortunately, carbapenemases are exceptionally rare in companion animals (11–13). In veterinary medicine, β-lactam antibiotics are the most widely used antibiotics for bacterial infections (21). In dogs, first-generation cephalosporins and amoxicillin-clavulanic acid are among the most commonly prescribed (23). Nevertheless, the use of carbapenems is not currently licensed, although it has been reported (24). Therefore, resistance to carbapenems is not routinely evaluated in animal isolates, so it seems likely that its prevalence is underestimated (11).

However, although the use of carbapenems is not common practice, the presence of genes encoding carbapenemases in mobile genetic elements that can also include other antibiotic resistance genes means that the use of other antimicrobials may contribute to coselection (6). The detection of genes conferring resistance to aminoglycosides, trimethoprim, and sulfonamides in the same class 1 integron with bla_VIM-1 can facilitate that coselection phenomenon.

The great increase and spread of CPE from human sources and its recent isolation from animals may reflect an emerging problem in human and veterinary medicine, as interspecies transmission may occur between humans and companion animals within the same household (25). Our results support this hypothesis, as the class 1 integron and plasmid carrying bla_VIM-1 detected in this study are among the most frequent mobile genetic elements associated with bla_VIM-1 dissemination in humans (26).

In Spain, the isolation of CPE in humans has significantly increased in recent years. According to a multicenter Spanish study, including 83 hospitals, OXA-48 carbapenemase is currently the most prevalent in Spain, followed by VIM-1, which appears much less frequently (27). The estimated overall prevalence of infection by carbapenemase-producing K. pneumoniae was 1.7% (range per hospital, 0% to 11.6%), and 23 (30.7%) hospitals had a prevalence of >1% (27). In the present study, we found a 0.6% CPE prevalence in companion dogs.

Emergence and spread of CPE in companion animals may represent an underestimated threat to public health. Our findings support probable human-to-animal transference of carbapenemases genes and reveal the need for active surveillance studies in companion animals.