Dogs Are a Reservoir of Ampicillin-Resistant Enterococcus faecium Lineages Associated with Human Infections APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Apr. 2009, p. 2360–2365 Vol. 75, No. 8 0099-2240/09/$08.00�0 doi:10.1128/AEM.02035-08 Copyright © 2009, American Society for Microbiology. All Rights Reserved. Dogs Are a Reservoir of Ampicillin-Resistant Enterococcus faecium Lineages Associated with Human Infections� Peter Damborg,1* Janetta Top,2 Antoni P. A. Hendrickx,2 Susan Dawson,3 Rob J. L. Willems,2 and Luca Guardabassi1 Department of Veterinary Pathobiology, Faculty of Life Sciences, University of Copenhagen, Frederiksberg C, Denmark1; Department of Medical Microbiology, University Medical Center Utrecht, Utrecht, The Netherlands2; and National Zoonosis Centre, University of Liverpool, Neston, United Kingdom3 Received 2 September 2008/Accepted 10 February 2009 Ampicillin resistance is a marker for hospital-associated Enterococcus faecium. Feces from 208 dogs were selectively screened for the occurrence of ampicillin-resistant E. faecium (AREF). AREF was detected in 42 (23%) of 183 dogs screened in a cross-sectional study in the United Kingdom and in 19 (76%) of 25 dogs studied longitudinally in Denmark. AREF carriage was intermittent in all dogs studied longitudinally. Multilocus sequence typing of 63 canine AREF isolates revealed the presence of 13 distinct sequence types. Approximately 76% of the isolates belonged to hospital-adapted clonal complex 17 (CC17), including those of sequence types ST-78 and ST-192, which are widespread in European and Asian hospitals. Longitudinal screening of 18 healthy humans living in contact with 13 of the dogs under study resulted in the identification of a single, intermittent CC17 carrier. This person carried one of the sequence types (ST-78) recovered from his dog. Based on PCR and Southern hybridization analyses, the putative virulence gene cluster from orf903 to orf907 was widespread in canine AREF isolates (present in 97%), whereas orf2351 (present in 26% of isolates) and orf2430 (present in 31%) were strongly associated with CC17-related sequence types (P < 0.05). Surprisingly, esp and hyl were not detected in any of the isolates. The antimicrobial resistance profiles of canine AREF isolates generally differed from those previously described for clinical human isolates. The results indicate that dogs are frequent carriers of CC17-related lineages and may play a role in the spread of this nosocomial pathogen. The distinctive virulence and antimicrobial resistance profiles observed among canine AREF isolates raise interesting questions about the origin and evolution of the strains causing human infections. Enterococci are opportunistic pathogens and form part of the normal gastrointestinal flora in humans and animals. Over the last two decades, nosocomial infections caused by entero- cocci have emerged and their incidence has increased rapidly, first in the United States and recently in Europe (25, 26, 29). Although Enterococcus faecalis is the causative agent in most enterococcal infections, a shift toward infections caused by multidrug-resistant E. faecium has been noted in the last years, and presently, up to one-third of enterococcal infections in some countries are attributed to this species (17). This shift may be explained by changes in the patterns of antimicrobial usage, which may have resulted in the emergence of a distinct genogroup of hospital-associated ampicillin-resistant E. fae- cium (AREF) strains, currently labeled clonal complex 17 (CC17) (33). CC17 isolates are characterized by resistance to ampicillin and fluoroquinolones, as well as by the presence in most isolates of putative virulence genes encoding the entero- coccal surface protein (esp) and hyaluronidase (hyl) and five recently described open reading frames (ORFs; orf903, orf904.5, orf906.7, orf2351, and orf2430) encoding LPXTG surface pro- teins, which are found less frequently among other E. faecium lineages (15, 20, 27). Based on the results of multilocus sequence typing (MLST) (28) and amplified fragment length polymorphism analysis (34), E. faecium isolates of animal origin seem to be host specific and generally unrelated to human lineages of clinical importance. Prior to this study, AREF CC17 strains have been isolated only sporadically from animals, including pigs (2, 10) and more recently dogs (8). Following these unexpected find- ings, the present study was designed to investigate the preva- lence and shedding patterns of AREF in dogs. A cross-sec- tional study and two longitudinal studies involving a total of 208 dogs and 479 canine fecal samples were conducted in the United Kingdom and in Denmark, respectively. Canine iso- lates were characterized by MLST, antimicrobial susceptibility testing, and putative virulence gene profiling to assess the genetic relationship between human and canine AREF strains. MATERIALS AND METHODS Sampling. The occurrence of AREF in 183 dogs screened as part of a cross- sectional study in Cheshire, United Kingdom, in 2006 (32) and in 25 dogs studied longitudinally in the region of Zealand, Denmark, in 2007 was investigated. Fecal samples or swabs from freshly voided feces were collected in sterile containers and submitted to the laboratory by the dog owners. Samples were kept frozen at �80°C whenever bacteriological analysis could not be performed within 72 h after collection. Samples from Danish dogs were collected as part of two separate longitudinal studies investigating intrafamily bacterial transmission and antimi- crobial effects on fecal microflora, respectively. In the first longitudinal study (study A) (Table 1), samples from 18 human and 13 canine members of eight family households were obtained on 12 occasions over a 6-month period. In the second longitudinal study (study B) (Table 2), samples from 12 dogs with pyo- derma that were undergoing treatment with different �-lactam compounds (i.e., amoxicillin-clavulanic acid, cephalexin, and cefovecin) were collected at 14 time points over the course of 1 month. Human samples consisted of rectal swabs. The * Corresponding author. Mailing address: Department of Disease Biology, Faculty of Life Sciences, University of Copenhagen, Stig- bøjlen 4, 1870 Frederiksberg C, Denmark. Phone: 45 35332725. Fax: 45 35332755. E-mail: peda@life.ku.dk. � Published ahead of print on 20 February 2009. 2360 o n A p ril 5 , 2 0 2 1 a t C A R N E G IE M E L L O N U N IV L IB R h ttp ://a e m .a sm .o rg / D o w n lo a d e d fro m https://crossmark.crossref.org/dialog/?doi=10.1128/AEM.02035-08&domain=pdf&date_stamp=2009-04-15 http://aem.asm.org/ protocol for obtaining samples from human participants was approved by the Danish National Committee on Biomedical Research Ethics (license number H-KF-2007-0007). Bacterial isolation and identification. All fecal samples were streaked onto plates of Slanetz-Bartley agar (Oxoid, Basingstoke, United Kingdom) supple- mented with 32 �g/ml of ampicillin, and the plates were incubated for 48 h at 44°C. One putative AREF isolate from each culture-positive dog and human was randomly selected and identified by a species-specific PCR method (12). Isolates confirmed to be AREF were subjected to further analyses. As part of longitu- dinal study B, total and relative numbers of putative AREF bacteria in all culture-positive samples were determined in duplicate. One gram of feces was mixed in a stomacher with 9 ml of sterile Milli-Q water. Plate counts were then performed by streaking 10-fold dilutions of samples onto Slanetz-Bartley agar with and without ampicillin. Following 48 h of incubation at 44°C, putative AREF bacteria and total enterococci on selective and nonselective plates, re- spectively, were counted, and bacterial concentrations (in CFU per gram) were calculated based on the plate counts. Antimicrobial susceptibility. Antimicrobial susceptibility testing was per- formed by the disk diffusion method according to the CLSI breakpoints for isolates from humans (5). Disks with the following antimicrobials were used: ampicillin (10 �g), ciprofloxacin (5 �g), erythromycin (15 �g), gentamicin (120 �g), linezolid (30 �g), quinupristin-dalfopristin (15 �g), rifampin (5 �g), tetra- cycline (30 �g), and vancomycin (30 �g). MLST. One AREF isolate from each dog was chosen randomly and subjected to MLST according to the protocol described by Homan et al. (16). Three isolates collected from a single dog (A1) at different sampling times were se- lected to get information on AREF diversity over time. Alleles were analyzed and sequence types (STs) were assigned using the database available at http: //www.mlst.net and the software CLC Combined Workbench 3 (CLC bio A/S, Aarhus, Denmark). STs obtained for AREF isolates were analyzed and com- pared to the entries in the existing E. faecium MLST database by using the eBURST algorithm. New STs were classified as belonging to CC17 if they were single-locus variants of STs within this complex. Detection of putative virulence genes. The presence of the genes esp, hyl, orf903, orf905, orf907, orf2351, and orf2430 was investigated by PCR using the primers and conditions described in previous studies (15, 20, 27) with the excep- tion that template DNA was extracted using the DNeasy kit (Qiagen Inc., Venlo, The Netherlands). E. faecium E1162 (14) and E. faecium C68 (27) were included as positive controls. The PCR results were confirmed by Southern blot analysis. In brief, chromosomal DNA was digested with EcoRI for 3 h at 37°C and DNA fragments were separated overnight on a 1% agarose gel. Upon exposure to UV light for 5 min, the gel was washed for 15 min in 0.25 M HCl and then subjected to two separate 15-min washes in 0.4 M NaOH. DNA fragments were transferred onto a Hybond N� nylon membrane by vacuum blotting, and the membrane was fixated in 0.4 M NaOH for 2 min and neutralized in 5� SSC (1� SSC is 0.15 M NaCl plus 0.015 M sodium citrate). Membranes were hybridized overnight at 42°C with a 100-ng probe. Probes specific for each gene were amplified from the chromosomal DNA of E. faecium strain DO (accession no. AAAK00000000), except those for esp and hyl, which were amplified from E. faecium E1162 DNA and E. faecium C68 DNA, respectively. Probes were purified using the QIAquick TABLE 1. Longitudinal carriage of putative AREF strains in healthy dogs in longitudinal study A Dog identification no. ST of AREF isolate or resulta for sample from day: 1 2 3 4 11 18 25 48 90 120 150 180 A1 � � � � 78 � 19 � 396 � � � A2 � � � � � � � � � � � � A3 � � � � � � � � � � � � A4 NR NR NR NR NR NR NR � � 121 � � A5 � � � � � � � � � � � 78 A6 � � � � � � � � � 78 � � A7 � � � � � � � 78 � � � � A8 � � 78 � � � � � � � � � A9 � � � � � � � � 266 � � � A10 � � � � � � � � � � � 397 A11 � � � � � � � � � � � � A12 � � � � � � � 192 � � � � A13 � � � � � � � 78 � � � � a Samples from 13 healthy dogs (A1 to A13) from a family study were analyzed. Numbers refer to STs of isolates verified as AREF. �, putative AREF isolate, not characterized; �, no growth on Slanetz-Bartley agar with ampicillin; NR, sample not received. TABLE 2. Longitudinal carriage of putative AREF strains in dogs with pyoderma in longitudinal study B Dog identification no. ST of AREF isolate or resulta for sample from day: 1 2 3 4 6 8 10 12 14 16 18 20 22 29 B1b � 192 � � � � � � � � � � � � B2b � � � � � � � � � � � � � � B3b � � � � � � � � 78 � � � � NR B4c � 78 � � � � � � � � NR � � � B5c 192 � � � � � � � � � NR � � � B6c � NR - 19 � � � � � � � � � � B7c � � � � � � � � � � � � NR � B8c � � NR � � � � NR NR � � � NR NR B9d � 19 NR � � � � � � � � � � � B10d � 19 � NR � � � � � � � NR � � B11d NR 78 � � � � � � � NR � NR � NR B12d � � NR � � � 78 � � NR NR NR � � a Samples from 12 dogs (B1 to B12) with pyoderma being treated with �-lactams were analyzed. Numbers refer to STs of isolates verified as AREF. �, putative AREF isolate, not characterized; �, no growth on Slanetz-Bartley agar with ampicillin; NR, sample not received. b Dog was treated daily with amoxicillin-clavulanic acid tablets from day 1 until day 14. c Dog was treated daily with cephalexin tablets from day 1 until day 14. d Dog was treated by cefovecin injection on day 1. VOL. 75, 2009 ENTEROCOCCUS FAECIUM CLONAL COMPLEX 17 IN DOGS 2361 o n A p ril 5 , 2 0 2 1 a t C A R N E G IE M E L L O N U N IV L IB R h ttp ://a e m .a sm .o rg / D o w n lo a d e d fro m http://aem.asm.org/ PCR purification kit (Qiagen Inc.). Labeling with horseradish peroxidase, hy- bridization, and detection were done according to the instructions of the man- ufacturer of the enhanced chemiluminescence nucleic acid labeling kit (GE Healthcare, Diegem, Belgium). The ORFs orf905 and orf907 are not putative virulence genes since their products lack the typical N-terminal signal peptide sequence of cell wall-an- chored proteins (CWAPs). Isolates positive for these genes were therefore sub- jected to PCR amplification and sequencing of the regions at the orf904-orf905 and orf906-orf907 junctions to identify the possible existence of the merged orf904.5 and orf906.7 genes, which both encode potentially functional CWAPs (15). Statistical analyses. Geographical clustering of STs and statistically significant differences in the prevalence of virulence genes among CC17 and non-CC17 isolates were determined by comparing proportions in EpiCalc 2000 (version 1.02 [http://www.brixtonhealth.com]). Nucleotide sequence accession number. The novel DNA sequence identified in the junction region of orf906.7 has been deposited in the GenBank nucleotide sequence database under accession number FJ481924. RESULTS AREF prevalence in and patterns of shedding from dogs. A total of 479 fecal samples and swabs from dogs were received and analyzed. Presumptive AREF was isolated from 42 (23%) of the 183 dogs screened as part of the cross-sectional study in the United Kingdom. Among the 25 dogs evaluated longitudi- nally in Denmark, 19 dogs (76%) had at least one sample positive for a presumptive AREF strain during the study pe- riod. The patterns of shedding from most of the dogs appeared to be extremely variable, as indicated by the intermittent de- tection of AREF in their feces (Tables 1 and 2). AREF con- centrations in feces varied substantially among individuals and among samples collected from the same individual at different time points as part of longitudinal study B. Total and relative numbers of putative AREF bacteria varied from 10 CFU/g and less than 1% of the total enterococci up to 106 CFU/g and 89% of enterococci (data not shown). All presumptive canine AREF isolates selected for further pheno- and genotypic char- acterization were confirmed to be E. faecium by PCR. Overall, AREF was detected at least once in 61 (29%) of the 208 dogs studied. A total of 203 fecal samples were received from the 18 healthy family household members who were screened concur- rently with their dogs in longitudinal study A. Only one 10- year-old boy, living with dog A1 (Table 1), was positive for AREF on a single occasion. Antimicrobial susceptibility. Ampicillin resistance was con- firmed for all isolates. More than half of the 61 AREF isolates tested displayed intermediate or full resistance to ciprofloxacin (92% of isolates), erythromycin (97%), tetracycline (89%), and rifampin (54%). Lower prevalences of resistance toward cer- tain first- or second-line agents currently used for the treat- ment of enterococcal infections, such as gentamicin (5%), linezolid (3%), and streptogramins (2%), were observed. Re- sistance to vancomycin was not detected. Notably, only a small percentage of isolates were fully resistant to erythromycin (Table 3). MLST. MLST analysis of 63 canine AREF isolates (includ- ing 3 from the same dog) revealed the occurrence of 13 STs, including seven novel STs (ST-396 to ST-402) (Table 4). The most frequent clone was ST-78, which was found in 16 isolates (25%) and was significantly associated with Danish origin (P � 0.011). Another common clone (ST-398) was found solely in isolates from the United Kingdom (P � 0.007). Four STs (ST-19, ST-78, ST-121, and ST-192) had been associated pre- viously with CC17, and four of the novel STs (ST-397 to ST- 400) were single-locus variants of ST-78 and ST-192 and there- fore also considered to belong to CC17. Overall, 48 (76%) of the 63 canine AREF isolates belonged to CC17 (Table 4). The AREF strain isolated from a 10-year-old boy belonged to the same ST (ST-78) previously detected in the boy’s dog. MLST analysis of two other isolates from the same dog (A1) revealed the presence of AREF ST-19 and ST-396 on days 25 and 90, respectively (Table 1). The genetic relatedness of STs to those listed in the central database (http://www.mlst.net) is depicted in Fig. 1. Occurrence of putative virulence genes. The results obtained by PCR and Southern hybridization matched for all isolates. Table 5 shows the distributions of putative virulence genes in isolates classified as CC17 and non-CC17 strains. None of the 62 analyzed canine isolates carried esp or hyl. The ORFs orf903, orf905, and orf907 occurred simultaneously in 60 iso- lates (97%) and were not statistically associated with CC17. No variation in the sequence of the orf904-orf905 junction region relative to the publicly available E. faecium DO sequence was observed, indicating that orf904 and orf905 were not merged TABLE 3. Antimicrobial susceptibilities of 61 canine AREF isolates Antimicrobial No. (%) of isolates Susceptible Intermediately resistant Resistant Ciprofloxacin 5 (8) 41 (67) 15 (25) Erythromycin 2 (3) 48 (79) 11 (18) Gentamicin 58 (95) 0 3 (5) Linezolid 53 (87) 6 (10) 2 (3) Quinupristin-dalfopristin 52 (85) 8 (13) 1 (2) Rifampin 28 (46) 1 (2) 32 (52) Tetracycline 7 (11) 3 (5) 51 (84) Vancomycin 61 (100) 0 0 TABLE 4. Multilocus STs of 63 canine AREF isolates Genogroup ST No. with indicated ST among: United Kingdom isolates (n � 42) Danish isolates (n � 21c) CC17 isolates (n � 48) ST-19 2 4 ST-78 6 10b ST-121 0 1 ST-192 5 3 ST-397a 0 1 ST-398a 14b 0 ST-399a 1 0 ST-400a 1 0 Non-CC17 isolates ST-157 1 0 (n � 15) ST-266 9 1 ST-396a 0 1 ST-401a 2 0 ST-402a 1 0 a New ST. b This result shows a significant association between the ST and the country of origin (P � 0.05). c Three Danish isolates belonging to ST-78, ST-192, and ST-396 originated from the same dog, whereas other isolates were from distinct dogs. 2362 DAMBORG ET AL. APPL. ENVIRON. MICROBIOL. o n A p ril 5 , 2 0 2 1 a t C A R N E G IE M E L L O N U N IV L IB R h ttp ://a e m .a sm .o rg / D o w n lo a d e d fro m http://aem.asm.org/ into orf904.5 in any of the canine isolates. A more diverse picture was evident from the sequencing of the orf906-orf907 junction region; in 20 isolates, the sequence was identical to that in E. faecium DO; 3 isolates had a 101-bp oligonucleotide deletion before the stop codon of orf906, resulting in various premature stop codons; and 36 isolates had a 19-bp oligonu- cleotide insertion (TTTATAACCCGAATTCATC) just be- fore the orf906 stop codon, which resulted in a frameshift and the merging of orf906 and orf907 into orf906.7, encoding an intact CWAP. This 19-bp insertion was identical to one re- ported previously (15) and occurred more commonly in non- CC17 isolates (86%) than in CC17-isolates (51%) (Table 5). In contrast to the almost ubiquitously present orf903-to-orf907 cluster, orf2351 and orf2430 were present in only 15 isolates (25%) and 18 isolates (30%), respectively, and both genes were significantly (P, 0.037 and 0.015) associated with CC17. Fur- thermore, orf2430 was specific to ST-78 and its two single-locus variants, ST-121 and ST-397. DISCUSSION We describe for the first time the widespread occurrence of hospital-associated AREF lineages in dogs. Remarkably, two of the STs most frequently isolated from dogs (ST-78 and ST-192) are among the most common AREF lineages causing infections in European and Asian hospitals (3, 18, 29, 31). This finding was surprising considering the general percep- tion that E. faecium strains are host specific and cluster ac- cording to the species of origin (19, 34). Approximately one in every four dogs harbored AREF CC17 bacteria; hence, dogs seem to be an important reservoir for these bacteria of medical interest. On the contrary, only 1 of the 18 healthy humans tested was found to be positive for E. faecium. One previous study failed to detect AREF in healthy humans despite the use of selective isolation media (9). However, the human carriage rate observed needs to be confirmed on a larger scale, since the low community prevalences of AREF among healthy people (0 to 6%) reported in other studies (1, 2, 11, 23) might be influ- enced by the use of nonselective isolation methods. AREF CC17 has spread rapidly in hospitals across the world (25, 29, 33). The widespread occurrence of ST-78, ST-192, and other CC17-related clones in dogs is worrisome since these animals may provide a vehicle for the spread of AREF among humans. In line with this hypothesis, AREF isolates displaying the same ST (ST-78), virulence level, and resistance pattern were obtained from a dog and a boy living within the same household. We were informed that the boy had close contact with the dog and that this contact included frequent kissing FIG. 1. Clustering of the MLST profiles as depicted by the eBURST algorithm. The 13 MLST profiles identified for canine isolates are indicated by arrows in the figure, which is based on MLST profiles of 1,449 E. faecium isolates from the central database (http://www.mlst.net). Each ST is represented as a node; the size of each node indicates the relative frequency of a particular ST. Each line indicates a single-locus difference between STs. TABLE 5. Distribution of putative virulence genes among 61 canine AREF isolates Gene Occurrence (no. �% of isolates with gene) in: CC17 group (n � 47) Non-CC17 group (n � 14) esp 0 0 hyl 0 0 orf903b 47 (100) 12 (86) orf904.5 0 0 orf906.7 24 (51) 12 (86) orf2351 15 (32)a 0 orf2430 18 (38)a 0 a This result shows that the gene is significantly associated with CC17 isolates (P � 0.05). b All isolates positive for orf903 also contained orf905 and orf907. VOL. 75, 2009 ENTEROCOCCUS FAECIUM CLONAL COMPLEX 17 IN DOGS 2363 o n A p ril 5 , 2 0 2 1 a t C A R N E G IE M E L L O N U N IV L IB R h ttp ://a e m .a sm .o rg / D o w n lo a d e d fro m http://aem.asm.org/ and petting. Such a close relationship may have enhanced the opportunity for the transmission of the strain between the dog and the boy. Possible links between canine and human patho- genic enterococci have been addressed previously by other authors; genetic similarities between vancomycin-resistant E. faecium isolates of canine and human origins were observed by Willems et al. (34) using amplified fragment length polymor- phism; a vancomycin-resistant E. faecalis isolate in a case of canine mastitis in New Zealand, described by Manson et al. (22), displayed a pulsed-field gel electrophoresis profile indis- tinguishable from that of a common human pathogenic clone in New Zealand. The putative virulence gene content of canine AREF iso- lates differed considerably from that usually observed in CC17 isolates from human infections. In particular, two putative virulence genes associated with CC17, esp and hyl (3, 6, 30), were completely absent among canine AREF isolates. Simi- larly, orf2351 and orf2430 occurred less frequently (in �30% of isolates) than previously reported for human vancomycin-re- sistant and -sensitive CC17 isolates ( 70% of which carried the genes) (15). The difference between canine and human clinical isolates may reflect an evolutionary multistep process during which E. faecium CC17 sequentially acquired a number of virulence and antibiotic resistance properties before becom- ing the most successful hospital-adapted lineage. As previously suggested by Leavis et al. (19), ampicillin resistance is likely one of the first properties that was acquired by a diverse group of clones and lineages currently included within CC17. Canine AREF isolates may therefore represent an early evolutionary ancestor of human clinical CC17 strains, which may have evolved and adapted to hospital environments by acquiring virulence genes such as esp and hyl. Alternatively, human AREF strains may be ancestors of canine strains, which may have evolved by the loss of these putative virulence factors outside hospital settings. The observation that the orf903-to- orf907 gene cluster was present in almost all canine AREF isolates, irrespective of their genetic backgrounds, suggests that the acquisition of this gene cluster may have been an early event in the evolutionary development of CC17. It is notewor- thy that the orf906.7 fusion, resulting in a potentially functional CWAP with similarities to the biofilm enhancer in enterococ- cus 3 protein of E. faecalis, was predominant in non-CC17 isolates and generally occurred more frequently in canine AREF isolates (60%) than previously reported for human AREF isolates (24%) (15). This finding suggests that the gene cluster became partly obsolete upon the transmission of E. faecium to humans and may have a minor role in the patho- genesis of human E. faecium infections. Despite the atypical putative virulence gene profile of canine isolates, the virulence and thereby the potential of canine isolates to cause human infections should not be underestimated, since esp and hyl occur only in a proportion of human AREF CC17 strains (7, 27, 35) and may therefore not be necessary for establishing infection in the human host. In addition, orf2351 and orf2430 occurred more or less specifically in ST-78, suggesting that the most common ST in dogs may also be one of the most patho- genic. Data on shedding patterns were obtained by the longitudinal studies conducted on 25 dogs in Denmark. Although 76% of the Danish dogs carried AREF at least once during the study period of 1 to 6 months, none of the dogs were found to be positive at all sampling times (Tables 1 and 2). This result suggests that the enterococcal flora in dogs is subject to fre- quent shifts and that most dogs are only transiently colonized with AREF or that AREF colonization dropped below the detection level. Plate counts performed as part of longitudinal study B indicated a high degree of variation in the concentra- tions of AREF bacteria among and within dogs screened at different time points. This variability suggests that the inclusion of a preenrichment step to enhance the detection of AREF in dogs may be advisable. Interestingly, dogs appear to shed dis- tinct AREF strains over time, as indicated by dog A1’s being colonized by three distinct STs (ST-19, ST-78, and ST 396) over a period of 3 months (Table 1). The levels of resistance of canine AREF strains to some of the most clinically relevant antimicrobials were relatively low (Table 3). Most importantly, vancomycin resistance was not detected and high-level gentamicin resistance was rare (occur- ring in 5% of isolates). Both vancomycin and gentamicin are first-line drugs for the treatment of enterococcal diseases, ei- ther separately or in combination with �-lactams (4). Two of the most frequently used second-line drugs, linezolid and streptogramins, were also included in our panel, and the ma- jority of isolates ( 85%) were susceptible to both agents. Ca- nine isolates displayed some atypical resistance patterns in comparison with data previously reported for CC17 isolates from human infections, including vancomycin-resistant and -sus- ceptible variants (6, 13, 24). In particular, the prevalences of macrolide and streptogramin resistance were low (18 and 2%, respectively) whereas the frequency of tetracycline resistance was unexpectedly high (82%). High-level resistance to fluoro- quinolones was not as common as that reported previously for human CC17 isolates (21). Altogether, these differences sup- port the notion that human and canine AREF strains may represent two distinct bacterial populations, despite the ge- netic similarities observed by MLST. In conclusion, healthy dogs are frequent carriers of human hospital-associated AREF CC17. Dogs may therefore play a role in the spread of this nosocomial pathogen in the commu- nity, and a risk of zoonotic transfer exists, as indicated by the possible case of transmission between a boy and his dog. Al- though the distinct putative virulence gene profiles suggest that canine isolates represent early evolutionary ancestors of hu- man pathogenic strains, further research is needed to assess the virulence of canine strains in comparison with that of human strains and, more generally, to quantify the magnitude of this possible emerging zoonotic problem. The Centers for Disease Control and Prevention have stated immunocompro- mised groups, for example, people with human immunodefi- ciency virus infection, organ transplant patients, and young children, to be at risk for infection with canine zoonotic agents (http://www.cdc.gov/healthypets/animals/dogs.htm). The pro- fessional use of pets to promote the recovery of patients (pet therapy) may pose a risk to such patients if the dogs are not previously screened for the presence of AREF and other zoo- notic pathogens. The occurrence of AREF in dogs and other domestic animals could be addressed by national programs for the surveillance of antimicrobial resistance in animals in order to explore the importance of the animal reservoir in the evo- lution of human hospital-associated enterococci. 2364 DAMBORG ET AL. APPL. ENVIRON. MICROBIOL. o n A p ril 5 , 2 0 2 1 a t C A R N E G IE M E L L O N U N IV L IB R h ttp ://a e m .a sm .o rg / D o w n lo a d e d fro m http://aem.asm.org/ ACKNOWLEDGMENTS We thank Nina B. Thomsen for processing samples from dogs in longitudinal study B. We also thank Nicola J. Williams at the National Zoonosis Centre in Liverpool for fruitful collaboration and for providing access to the fecal samples collected in Cheshire, United Kingdom. REFERENCES 1. Aarestrup, F. M., Y. Agerso, P. Gerner-Smidt, M. Madsen, and L. B. Jensen. 2000. Comparison of antimicrobial resistance phenotypes and resistance genes in Enterococcus faecalis and Enterococcus faecium from humans in the community, broilers, and pigs in Denmark. Diagn. Microbiol. Infect. Dis. 37:127–137. 2. Biavasco, F., G. Foglia, C. Paoletti, G. Zandri, G. Magi, E. Guaglianone, A. Sundsfjord, C. Pruzzo, G. Donelli, and B. Facinelli. 2007. VanA-type en- terococci from humans, animals, and food: species distribution, population structure, Tn1546 typing and location, and virulence determinants. Appl. Environ. Microbiol. 73:3307–3319. 3. Bonora, M. G., D. Olioso, C. G. Lo, and R. Fontana. 2007. Phylogenetic analysis of vancomycin-resistant Enterococcus faecium genotypes associated with outbreaks or sporadic infections in Italy. Microb. Drug Resist. 13:171– 177. 4. Cetinkaya, Y., P. Falk, and C. G. Mayhall. 2000. Vancomycin-resistant en- terococci. Clin. Microbiol. Rev. 13:686–707. 5. Clinical and Laboratory Standards Institute. 2005. Performance standards for antimicrobial susceptibility testing; 15th informational supplement, M100–S15. Clinical and Laboratory Standards Institute, Wayne, PA. 6. Coque, T. M., R. J. L. Willems, J. Fortún, J. Top, S. Diz, E. Loza, R. Cantón, and F. Baquero. 2005. Population structure of Enterococcus faecium causing bacteremia in a Spanish university hospital: setting the scene for a future increase in vancomycin resistance? Antimicrob. Agents Chemother. 49:2693–2700. 7. Coque, T. M., R. J. L. Willems, R. Cantón, C. R. Del Campo, and F. Baquero. 2002. High occurrence of esp among ampicillin-resistant and vancomycin- susceptible Enterococcus faecium clones from hospitalized patients. J. Anti- microb. Chemother. 50:1035–1038. 8. Damborg, P., A. H. Sørensen, and L. Guardabassi. 2008. Monitoring of antimicrobial resistance in healthy dogs: first report of canine ampicillin- resistant Enterococcus faecium clonal complex 17. Vet. Microbiol. 132:190– 196. 9. Del Campo, R., P. Ruiz-Garbajosa, M. P. Sánchez-Moreno, F. Baquero, C. Torres, R. Cantón, and T. M. Coque. 2003. Antimicrobial resistance in recent fecal enterococci from healthy volunteers and food handlers in Spain: genes and phenotypes. Microb. Drug Resist. 9:47–60. 10. De Leener, E., A. Martel, E. M. De Graef, J. Top, P. Butaye, F. Haesebrouck, R. J. L. Willems, and A. Decostere. 2005. Molecular analysis of human, porcine, and poultry Enterococcus faecium isolates and their erm(B) genes. Appl. Environ. Microbiol. 71:2766–2770. 11. Duh, R. W., K. V. Singh, K. Malathum, and B. E. Murray. 2001. In vitro activity of 19 antimicrobial agents against enterococci from healthy subjects and hospitalized patients and use of an ace gene probe from Enterococcus faecalis for species identification. Microb. Drug Resist. 7:39–46. 12. Dutka-Malen, S., S. Evers, and P. Courvalin. 1995. Detection of glycopep- tide resistance genotypes and identification to the species level of clinically relevant enterococci by PCR. J. Clin. Microbiol. 33:1434. 13. Ghoshal, U., A. Garg, D. P. Tiwari, and A. Ayyagari. 2006. Emerging van- comycin resistance in enterococci in India. Indian J. Pathol. Microbiol. 49:620–622. 14. Heikens, E., M. J. Bonten, and R. J. L. Willems. 2007. Enterococcal surface protein Esp is important for biofilm formation of Enterococcus faecium E1162. J. Bacteriol. 189:8233–8240. 15. Hendrickx, A. P., W. J. van Wamel, G. Posthuma, M. J. Bonten, and R. J. L. Willems. 2007. Five genes encoding surface-exposed LPXTG proteins are enriched in hospital-adapted Enterococcus faecium clonal complex 17 iso- lates. J. Bacteriol. 189:8321–8332. 16. Homan, W. L., D. Tribe, S. Poznanski, M. Li, G. Hogg, E. Spalburg, J. D. van Embden, and R. J. L. Willems. 2002. Multilocus sequence typing scheme for Enterococcus faecium. J. Clin. Microbiol. 40:1963–1971. 17. Iwen, P. C., D. M. Kelly, J. Linder, S. H. Hinrichs, E. A. Dominguez, M. E. Rupp, and K. D. Patil. 1997. Change in prevalence and antibiotic resistance of Enterococcus species isolated from blood cultures over an 8-year period. Antimicrob. Agents Chemother. 41:494–495. 18. Ko, K. S., J. Y. Baek, J. Y. Lee, W. S. Oh, K. R. Peck, N. Lee, W. G. Lee, K. Lee, and J. H. Song. 2005. Molecular characterization of vancomycin-resis- tant Enterococcus faecium isolates from Korea. J. Clin. Microbiol. 43:2303– 2306. 19. Leavis, H. L., M. J. Bonten, and R. J. L. Willems. 2006. Identification of high-risk enterococcal clonal complexes: global dispersion and antibiotic resistance. Curr. Opin. Microbiol. 9:454–460. 20. Leavis, H. L., R. J. L. Willems, J. Top, E. Spalburg, E. M. Mascini, A. C. Fluit, A. Hoepelman, A. J. de Neeling, and M. J. Bonten. 2003. Epidemic and nonepidemic multidrug-resistant Enterococcus faecium. Emerg. Infect. Dis. 9:1108–1115. 21. Leavis, H. L., R. J. L. Willems, J. Top, and M. J. Bonten. 2006. High-level ciprofloxacin resistance from point mutations in gyrA and parC confined to global hospital-adapted clonal lineage CC17 of Enterococcus faecium. J. Clin. Microbiol. 44:1059–1064. 22. Manson, J. M., S. Keis, J. M. Smith, and G. M. Cook. 2003. Characterization of a vancomycin-resistant Enterococcus faecalis (VREF) isolate from a dog with mastitis: further evidence of a clonal lineage of VREF in New Zealand. J. Clin. Microbiol. 41:3331–3333. 23. Novais, C., T. M. Coque, J. C. Sousa, and L. V. Peixe. 2006. Antimicrobial resistance among faecal enterococci from healthy individuals in Portugal. Clin. Microbiol. Infect. 12:1131–1134. 24. Novais, C., T. M. Coque, M. J. Costa, J. C. Sousa, F. Baquero, and L. V. Peixe. 2005. High occurrence and persistence of antibiotic-resistant entero- cocci in poultry food samples in Portugal. J. Antimicrob. Chemother. 56: 1139–1143. 25. Oteo, J., O. Cuevas, C. Navarro, B. Aracil, and J. Campos. 2007. Trends in antimicrobial resistance in 3469 enterococci isolated from blood (EARSS experience 2001–06, Spain): increasing ampicillin resistance in Enterococcus faecium. J. Antimicrob. Chemother. 59:1044–1045. 26. Rice, L. B. 2001. Emergence of vancomycin-resistant enterococci. Emerg. Infect. Dis. 7:183–187. 27. Rice, L. B., L. Carias, S. Rudin, C. Vael, H. Goossens, C. Konstabel, I. Klare, S. R. Nallapareddy, W. Huang, and B. E. Murray. 2003. A potential viru- lence gene, hylEfm, predominates in Enterococcus faecium of clinical origin. J. Infect. Dis. 187:508–512. 28. Top, J., R. J. L. Willems, H. Blok, M. de Regt, K. Jalink, A. Troelstra, B. Goorhuis, and M. J. Bonten. 2007. Ecological replacement of Enterococcus faecalis by multiresistant clonal complex 17 Enterococcus faecium. Clin. Mi- crobiol. Infect. 13:316–319. 29. Top, J., R. J. L. Willems, S. van der Velden, M. Asbroek, and M. J. Bonten. 2008. Emergence of clonal complex 17 Enterococcus faecium in The Neth- erlands. J. Clin. Microbiol. 46:214–219. 30. Vankerckhoven, V., T. van Autgaerden, C. Vael, C. Lammens, S. Chapelle, R. Rossi, D. Jabes, and H. Goossens. 2004. Development of a multiplex PCR for the detection of asa1, gelE, cylA, esp, and hyl genes in enterococci and survey for virulence determinants among European hospital isolates of En- terococcus faecium. J. Clin. Microbiol. 42:4473–4479. 31. Werner, G., I. Klare, and W. Witte. 2007. The current MLVA typing scheme for Enterococcus faecium is less discriminatory than MLST and PFGE for epidemic-virulent, hospital-adapted clonal types. BMC Microbiol. 7:28. 32. Westgarth, C., G. L. Pinchbeck, J. W. Bradshaw, S. Dawson, R. M. Gaskell, and R. M. Christley. 2007. Factors associated with dog ownership and con- tact with dogs in a UK community. BMC Vet. Res. 3:5. 33. Willems, R. J. L., J. Top, M. van Santen, D. A. Robinson, T. M. Coque, F. Baquero, H. Grundmann, and M. J. Bonten. 2005. Global spread of vanco- mycin-resistant Enterococcus faecium from distinct nosocomial genetic com- plex. Emerg. Infect. Dis. 11:821–828. 34. Willems, R. J. L., J. Top, N. van den Braak, A. van Belkum, H. Endtz, D. Mevius, E. Stobberingh, A. van den Bogaard, and J. D. van Embden. 2000. Host specificity of vancomycin-resistant Enterococcus faecium. J. Infect. Dis. 182:816–823. 35. Woodford, N., M. Soltani, and K. J. Hardy. 2001. Frequency of esp in Enterococcus faecium isolates. Lancet 358:584. VOL. 75, 2009 ENTEROCOCCUS FAECIUM CLONAL COMPLEX 17 IN DOGS 2365 o n A p ril 5 , 2 0 2 1 a t C A R N E G IE M E L L O N U N IV L IB R h ttp ://a e m .a sm .o rg / D o w n lo a d e d fro m http://aem.asm.org/