Carriage of critically important antimicrobial resistant bacteria and zoonotic parasites amongst camp dogs in remote Western Australian indigenous communities

Camp dogs in indigenous communities in the Western Australian Kimberley Region, share the domestic environment with humans and have the potential to act as carriers of, and sentinels for, a wide range of zoonotic agents, including intestinal parasites and antimicrobial resistant bacteria. In this study, we investigated the carriage of extended-spectrum-cephalosporin-resistant (ESC-resistant) Escherichia coli, methicillin-resistant Staphylococcus aureus (MRSA) and species of hookworm and Giardia among camp dogs in remote Western Australian Aboriginal communities. A total of 141 canine faecal samples and 156 nasal swabs were collected from dogs in four communities of the Western Australian Kimberley region. Overall, ESC-resistant E. coli was detected in 16.7% of faecal samples and MRSA was isolated from 2.6% of nasal swabs. Of most significance was the presence of the community-associated Panton-Valentine leucocidin (PVL)-positive MRSA ST93 and ST5 clones and ESC-resistant E. coli ST38 and ST131. The most prevalent zoonotic intestinal parasite infection was Ancylostoma caninum (66%). The prevalence of Giardia was 12.1%, with the main genotypes of Giardia detected being dog specific assemblages C and D, which are unlikely to cause disease in humans.

All ESC-resistant E. coli isolates were resistant to two or more classes of antimicrobials (multidrug resistant) with universal resistance to ceftriaxone and ampicillin. Two of the isolates from EK2, (ST38 and ST131), and one ST3268 isolate from East Kimberley 3 (EK3) demonstrated resistance or intermediate resistance to ciprofloxacin. The ST38 and ST131 isolates were also resistant to trimethoprim/sulfamethoxazole. Two ST38 isolates, from different locations, were resistant to six or more antimicrobial classes and carried corresponding resistance genes. The two phenotypically and genotypically distinguishable ST2144 were isolated from the same animal with additional resistance to cefoxitin and amoxicillin-clavulanate (Table 1).
Carriage of MRSA. MRSA was isolated from 2.6% (4/156) of the nasal swabs, two isolates from EK1, one isolate from EK2 and one isolate from WK1. The four MRSA isolates were cefoxitin and penicillin resistant, but trimethoprim/sulfamethoxazole, tetracycline, erythromycin, ciprofloxacin and gentamicin susceptible ( Table 2). Two of the four MRSA isolates were identified as ST93. The remaining two isolates were identified as ST5 and ST872 ( Table 2). All isolates carried beta-lactam (blaZ), methicillin (mecA) and efflux pump (norA) genes. The Panton Valentine leucocidin (PVL) associated genes, lukF-PV and lukS-PV, were detected in 3 of 4 isolates (ST93 and ST5).
Carriage of zoonotic parasites. Based on molecular detection results of the species of interest, 82 dogs were infected with hookworm (A. caninum) alone, six dogs were infected with Giardia alone, and 11 dogs were infected with hookworm and Giardia combined. A total of 111 dogs were found to be infected with one or more parasites based on combined molecular and microscopic examination. Overall, hookworm infection was the most common with 93 of 141 dogs (66%) positive. The other helminths observed in the samples were Toxocara canis (4.3%, n = 6), Spirometra erinaceid (3.5%, n = 5), Taenia spp. (1.4%, n = 2), and Spirocerca lupi (0.7%, n = 1). Giardia was the second most prevalent protozoa in dogs (12.1%, n = 17) after Isospora spp. (12.7%, n = 18) followed by Sarcocystis spp. (9.9%, n = 14). The prevalence of Giardia and hookworm in each community in the Kimberley region is shown in Table 3. EK1 had the highest crude prevalence of hookworm (93.8%) and Giardia (18.8%).

Discussion
The current study provides information on the carriage of ESC-resistant E. coli, MRSA and zoonotic enteric parasites among camp dogs that are in close contact with Western Australian Aboriginal communities. ESC-resistant E. coli were carried by 16.7% of dogs with some of the isolates belonging to the globally disseminated ST131 and ST38 ESC-resistant pandemic clones. Additionally, camp dogs were colonized by PVL-positive CA-MRSA (ST93 and ST5) clones at a low prevalence (2.6%); and zoonotic intestinal parasites Giardia and Ancylostoma caninum were present at prevalences of 12.1% and 66% respectively. It should be noted that the prevalences reported are crude, given the opportunistic nature of sampling, and that the numbers of dogs present in each region are only estimates, with no valid enumeration data available.
ST38 was identified as the major E. coli ST in the faecal samples collected in two community locations. ST2144 and ST131 were also identified for more than one isolate. The two ST2144 were isolated from the same animal in the West Kimberly community. This sample grew 3 morphologically distinguishable colony types, however whole genome sequencing showed that only two separate genotypes were present. ST131 carrying bla CTX-M genes, which was isolated from two animals located in the East Kimberley community, can exist as a globally disseminated multi-drug resistant pandemic extra intestinal pathogenic E. coli responsible for causing variety of extra intestinal infections in humans, including urinary tract infection and bacteraemia 31,32 . Significantly, ST131 and ST38 have previously been reported as causes of disease in various animals, including dogs [33][34][35][36] . The limitations of this study do not allow conclusions to be drawn on how the sampled animals acquired these infections. It could be hypothesized that they naturally circulate in dogs in these communities or alternatively they are spillover from the human population. However, these findings are of public health concern, given the possibility that these clonal types may be transferred from dogs to humans, and a larger scale study inclusive of human sampling may aid in determining the ecology of resistant E. coli in these populations.
PVL-positive ST93 -IV and PVL-negative ST5 -IV are community acquired (CA)-MRSA that have also been found in animals 37,38 . ST93-IV is the dominant CA-MRSA clone across Australia in humans, and has been associated with a range of skin and soft tissue infections, as well as severe invasive infections such as necrotizing pneumonia [38][39][40] . The three MRSA isolates harbored the beta-lactamase gene (blaZ), the penicillin-binding protein, PBP 2a gene (mecA) and the efflux pump gene (norA). This finding is of important public health significance in these populations, as these isolates are resistant to beta-lactam antibiotics which may be used for treatment of pneumonia,   skin and ear infections that are highly prevalent among Aboriginal communities 41,42 . As for E coli, a more detailed study to examine the ecology of these MRSA clones in Aboriginal communities and camp dogs is warranted. The prevalence of Ancylostoma caninum in dogs identified in the current study is similar to a previous report from the same area, completed in 1993 27 . The high prevalence of A. caninum increases the opportunity for spread of the infection to humans in the communities, which can cause cutaneous larva migrans 43 . Although A. caninum is a zoonotic agent, it is considered of minor public health significance as this species of hookworm rarely progresses past cutaneous infections 23 . Dog Health Programs in Aboriginal communities that were first introduced in the Kimberley region of Western Australia in 1992 44 , used the anthelmintic Ivermectin to reduce the prevalence of scabies and hookworm in dogs. Unfortunately, the treatment was only able to reduce the intensity of the infection but did not significantly diminish the prevalence of canine hookworm 44 . The failure of eradication of the parasite might be correlated to periodic treatment, however, more recently dogs in these communities have    received 3-monthly moxidectin treatments. As such these results are of concern, and may indicate anthelmintic resistance or heavy environmental contamination by dogs which have missed regular treatments. The prevalence of Giardia infection in dogs in this study (12.1%, 95% CI 7.7, 18.5) was similar to findings in the same region over 20 years ago (17%) 27 and to a national study of gastrointestinal parasites of dogs in Australia (9.3%, 95% CI 7.8-10.8) 45 . This study found that the genotype of Giardia from dogs in the region were mostly canine-specific Assemblages C and D. The zoonotic Giardia Assemblage A was only found in one sample, and it would appear that the likelihood of transmission of Giardia between dogs and humans in the Kimberley Region remains low.
The management of dogs is of paramount importance in minimizing the spread of zoonotic agents through these communities. Dogs in this study were able to roam freely, and scavenged on human waste. Access to materials such as human faeces has the potential for dogs to become infected with human associated bacterial clones and parasites, and maintain them in the community. Ongoing de-sexing and treatment clinics together with continuous client education regarding good husbandry practices and correct anthelmintic, antiprotozoal and antibiotic administration are also important to prevent recurrent infections.
In conclusion, this study demonstrates the carriage of antimicrobial resistant bacteria and zoonotic enteric parasites amongst camp dogs in remote Western Australian communities. The carriage of human associated MRSA (ST93 and ST5) and ESC-resistant E. coli (ST131 and 38) identified in this study is of particular importance, and requires further study to determine whether there is movement of CIA resistant bacteria from humans to animals and the potential for zoonotic transmission to humans.  Table 5.

Source of isolates. Work undertaken in this survey was approved by the Murdoch University Animal Ethics
Committee (Permit #408 and #R2876/16), with all experiments performed in accordance with relevant guidelines and regulations. Nasal swabs were collected from 156 dogs from five communities. Of these dogs, faecal samples could be collected from 141, with the remaining having empty rectums. Sampling was conducted for diagnostic purposes by the Murdoch University veterinary team undertaking a neutering operation and dog health program in the Kimberley region in three-time periods; June 2016, October 2016 and June 2017. Sample numbers were based solely on dogs entering the neutering programme. Only dogs which had not been previously neutered were sampled to prevent resampling the same individual. Faecal samples were collected into standard 70 ml plastic containers. Nasal swab samples were collected using swabs into charcoal media (Copan, Italy). Samples were stored at 4 °C until processed.
Bacterial Isolation and detection. For MRSA isolation, swabs were plated onto Brilliance MRSA Agar (ThermoFisher Scientific) and incubated overnight at 37 °C. Colonies resembling MRSA were subcultured onto 5% Sheep Blood Agar (Edwards Media). Screening for ESC-resistance was performed by incubating the faecal samples onto Brilliance ESBL Agar (ThermoFisher Scientific) and incubating overnight at 37 °C. Colonies resembling ESBL E. coli were sub-cultured onto 5% Sheep Blood Agar (Edwards Media). If more than one colony morphology was identified on a plate an isolate from each colony type was taken. Identification of all isolates was conducted using a Bruker microflex MALDI-TOF.
Antimicrobial susceptibility testing. Isolates underwent susceptibility testing via disc diffusion according to the Clinical Laboratory Standards Institute (CLSI) Performance Standards for antimicrobial disk susceptibility tests M02-A12 46 . MRSA were tested using the following seven antimicrobials: trimethoprim/sulfamethoxazole, tetracycline, cefoxitin, erythromycin, penicillin, ciprofloxacin and gentamicin. E. coli isolates were tested using the following 12 antimicrobials: Trimethoprim/Sulfamethoxazole, tetracycline, cefoxitin, ceftriaxone, gentamicin, chloramphenicol, ampicillin, streptomycin, imipenem, ciprofloxacin, amoxicillin-clavulanate and meropenem. Zone diameter results were categorized as susceptible, intermediate and resistant using the clinical interpretative criteria specified in CLSI performance standard VET01-S3 47 . If interpretive criteria was not present in VET01-S3, CLSI performance standard M100-S25 was used 48 .
Detection of resistance genes. DNA was extracted from isolates using a MagMax DNA multi sample kit (ThermoFisher Scientific) as per manufacturer's instructions with the modification to omit the RNAse treatment step. Library preparation was performed using an Illumina NexTera XT library preparation kit as per  Parasite egg identification and DNA extraction. One gram of each faecal sample was examined for the presence of parasite eggs using flotation in saturated zinc sulphate, followed by examination under a light microscope. Briefly, approximately 1 g of faeces was mixed with 9 mL of zinc sulphate solution (specific gravity 1.18) in a 10 mL centrifuge tube. The tube was centrifuged at 900 xg for five minutes with no brake. Additional zinc sulphate solution was added to form a positive meniscus and a cover slip was placed on the top of the tube. After approximately 5 minutes the cover slip was placed onto a slide and examined for the presence of parasites at 100× and 400× magnification. DNA was extracted directly from all faecal samples using a Bioline Isolate II Faecal DNA Kit (Bioline), as per the manufacturer's instructions. DNA extracts were stored at −20 °C until required.
Polymerase chain reaction. DNA extracts were subject to qPCR for identification of Giardia, conventional PCR for differentiating Ancylostoma species and conventional PCR for genotyping Giardia. An approximately 380 bp section of internal transcribed spacer-2 (ITS-2) region of Ancylostoma spp. was amplified using a protocol modified from Smout et al. 50 . Primers used in this assay are listed in The presence of Giardia in all samples were screened at the glutamate dehydrogenase (gdh) locus using a quantitative PCR (qPCR) procedure previously described by Yang et al. 51 . Conventional PCR amplification of the glutamate dehydrogenase (gdh) and the triose phosphate isomerase (tpi) locus was conducted on all samples found positive for Giardia on qPCR. An approximately 733 bp portion of the gdh gene was obtained using formerly published primers 52,53 . For this nested PCR, primers GDHeF and GDH 2 were used in the primary reaction, and primers GDHiF and GDH 4 were used in the secondary reaction (Table 2). PCR reaction volume for each sample in both primary and secondary PCRs was 20 µL, containing 10 µL GoTaq ® Green Master Mix (Promega,USA), 0.25 µM of each primer, 4 µL nuclease-free water and 5 µL of template genomic DNA. Cycling conditions for primary PCR were 1 cycle of 94 °C for 2 min, followed by 35 cycles of 94 °C for 30 s, 55 °C for 30 s and 72 °C for 60 s with a final extension of 72 °C for 7 min and a 12 °C hold. The cycling conditions for the secondary PCR were similar to the primary PCR, except the annealing temperature, which was 52 °C. PCR of the tpi locus utilised a nested PCR protocol developed by Sulaiman et al. (2003) with slight modifications 54 . Primary and secondary primers are shown in Table 6   with a final extension of 72 °C for 7 min. The amplified DNA products from the gdh and tpi PCR were visualized on a 1.5% agarose gel containing SYBR ® Safe DNA gel stain.
PCR products of the Ancylostoma spp., Giardia gdh and Giardia tpi reactions were excised from gels and purified using the Wizard ® SV Gel and PCR Clean-Up System kit (Promega, USA) before DNA sequencing. DNA sequencing was performed at the Australian Genome Research Facility (Perth, WA). Following screening by qPCR for Giardia, only samples which were positive upon Sanger sequencing on the gdh and/or tpi assays were considered as confirmed positives. Ancylostoma positive status was also based on Sanger sequence positive PCR results. Data availability. The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.