Prevalence of most common human pathogenic Campylobacter spp. in dogs and cats in Styria, Austria

Abstract The aim of this study was to determine the frequency of occurrence of most common human pathogenic Campylobacter species, Campylobacter jejuni (C. jejuni) and Campylobacter coli (C. coli), in dogs and cats in Styria, Austria. In the period from April 2010 to April 2012, 842 faecal samples from dogs and cats from Styria, Austria were examined for Campylobacter (C.) species (spp.). All samples were subjected to qualitative microbiological culture testing, and additionally, some of them have been studied using qualitative real‐time PCR. In microbiological culture, 5.9% of all samples investigated were C. spp. positive. With 3.1% out of positive samples, C. jejuni was the most common type. Campylobacter upsaliensis (C. upsaliensis) was detected only in 0.5% of the samples. The remaining positive samples (2.4%) were classified as C. species (sp.). C. coli could not be found in any of the samples. A higher prevalence of C. jejuni was found in kittens with 14.3% and in diarrhoeic dogs (7.4%) and cats (23.8%). The real‐time PCR revealed for dogs and cats together, 27% of C. jejuni‐positive faecal and 8% positive faecal swap samples. The obtained C. jejuni strains underwent antibiotic resistance testing using three different tests (agar diffusion, MIC testing and E‐test) with different numbers of antibiotics. From the antibiotics used in this study, several showed high test‐dependent resistance rates (cephalexin, cefovecin, kanamycin, sulfamethoxazole/trimethoprim, penicillin G, ciprofloxacin, enrofloxacin, marbofloxacin, nalidixic acid). Overall, the prevalence of C. spp. in this study was very low compared to others, with the exception of C. jejuni in kittens and diarrhoeic animals. The results of the real‐time PCR suggest that the rate of colonization of C. jejuni was actually higher than the results of the culture showed. As the resistance rates of C. jejuni isolates partly were very high, possible transmission of (multi‐) resistant C. jejuni strains to humans especially from kittens and diarrhoeic animals must be expected.


Introduction
Infections with Campylobacter (C.) species (spp.) are still the leading causes of acute bacterial gastroenteritis in industrialized countries. Human Campylobacteriosis shows the highest incidence among bacterial notifiable diseases in Europe since 2005. In 2015, the overall incidence in the EU was 65.5 confirmed cases/100 000 inhabitants. The rate in Austria 2015 was 73 cases/100 000 inhabitants. In 2015, 229 213 confirmed cases of disease were reported in the EU (Anonymous 2007(Anonymous , 2016. The most common clinical symptoms are diarrhoea, fever and abdominal pain. The infection is usually self-limiting and subsides without treatment within 1 week. In some cases, however, postinfectious complications such as Guillain-Barr e syndrome and reactive arthritis may occur (Allos 1997;Nachamkin & Blaser 2000;Moore et al. 2005).
The high incidence of Campylobacter infections in humans results in large costs for the public health system (Buzby et al. 1997;Mangen et al. 2004).
The main risk factors for Campylobacter infection include poultry meat and poultry meat products, contaminated drinking water or crops, raw or insufficiently heated milk and direct animal contact. Infections occur primarily due to poor kitchen hygiene when handling poultry meat and poultry meat products. Between 20 and 40% of all human, Campylobacter infections can be directly traced back to poultry meat and poultry meat products (Pebody et al. 1997;Nadeau et al. 2002;Vellinga & van Loock 2002) and 52-80% overall to sources which originate from the poultry sector (Mullner et al. 2009;Anonymous, 2010;van Gerve 2012).
In Austria, studies on the prevalence of Campylobacter spp. have been made primarily on livestock (Ziegler 1993a,b;Hein et al. 2003;Ursinitsch et al. 2005), with only one study on dogs (Balucinska 1995). Currently, small animals (dogs, cats) are considered as asymptomatic carriers of C. spp. However, there are also publications in which C. spp. was classified as primary or secondary pathogen which can trigger gastrointestinal symptoms in small animals (Fleming 1983;Burnens et al. 1992). A transfer from dogs and cats to humans or vice versa cannot be ruled out (Damborg et al. 2004). In particular, dog owners seem to have a significantly higher risk of infection with C. jejuni and coli from their pets (Mughini-Gras et al. 2013). Isolation of C. spp. in small animals succeeds very often. The most commonly detected species were Campylobacter upsaliensis (C. upsaliensis) 39-98%, C. jejuni 1.2-51.2% and C. coli 0-9.8% (Hald et al. 2004;Koene et al. 2004;Keller et al. 2007;Gargiulo et al. 2008;Acke et al. 2006Acke et al. , 2009aParson et al. 2010;Salihu et al. 2010;Parsons et al. 2011;Badlik et al. 2014;Procter et al. 2014;Giacomelli et al. 2015;Holmberg et al. 2015;Olkkola et al. 2015;Selwet et al. 2015;Bojani c et al. 2017). However, the isolation rates from faeces differ greatly depending on age, clinical signs, environment, concomitant diseases, infections with other enteropathogenic organisms, respective Campylobacter species, isolation method and design of the study (Torre & Tello 1993;Engvall et al. 2003;Bender et al. 2005;Wieland et al. 2005;Chaban et al. 2010).
The primary objective of this study was to determine the prevalence of the most pathogenic thermotolerant C. spp. for humans (C. jejuni, C. coli) in faeces of dogs and cats from Styria, Austria. Additionally, it should be examined whether the age of the animals or gastrointestinal diseases have an influence on the occurrence of these C. spp.
As a supplement, antibiotic susceptibility testing of C. jejuni isolates was carried out.

Sampling and shipment
In the period between April 2010 and April 2012, 842 samples (498 dogs, 344 cats) were examined for Campylobacter (C.) species (spp.). They consisted of 756 faecal swabs (442 dogs, 314 cats) and 70 faecal samples (51 dogs, 19 cats) (Table 1), which were gathered from the rectum of the animals by practicing veterinarians from Styria during routine investigations using sterile cotton swab and placed in a hermetically sealed tube containing nutrient medium (Amies W, Switzerland; Sterilin Ltd., Newport Gwent, UK). The extracted faecal samples containing at least 1 g of faeces were placed in sterile stool tubes with spoon (76 9 20 mm; Sarstedt, Germany), sealed and protected by a screw top vessel (85 9 30 mm; Sarstedt, Germany).
Until shipping, the samples were refrigerated at a temperature from 2 to 8°C. A maximum of 4 days was set between sampling and arrival at the institute. The samples were shipped together with a submission form by a messenger in an uncooled padded envelope.
Moreover, 16 faecal samples which were obtained from autopsies in our own institute were examined on the day of removal.
The samples were divided into two groups based on history, one group of animals suffering from gastrointestinal symptoms and another group not suffering this symptoms (healthy or having other diseases).
Qualitative microbiological culture method (ISO 10272-1/2006) For sample collection and selected microbiological culture, simple and rapid methods (faecal swabs, direct plating on mCCDA agar) were chosen, to keep effort and costs low and to be able to examine as many samples as possible. In literature, there was no clear indication which sample material or which cultural detection method would be the best of all. Only the combination of several methods seems to provide best results (Koene et al. 2004;Acke et al. 2006Acke et al. , 2009a. After arrival at the institute, the samples were stored until further processing at refrigerator temperature (2-8°C). Further processing of the samples was carried out on the same day. First, the required selective agar plates (mCCDA agar) were brought to room temperature. The faecal swabs were streaked directly and faecal samples with a laboratory wire loop on the mCCDA agar. Then, two dilutional streaks with an annealed laboratory wire loop were made. The plates were incubated for 48 h at 37°C under microaerophilic conditions (CampyGen CN0035A, 3.5 L; Oxoid, Hampshire, UK) either in a glove box (Scholzen Microbiology Systems AG, Switzerland MC 1G) or an anaerobic box (Anaero-Pack Rectangular Jar, 2.5/7 L; Mitsubishi Gas Chemical Company Inc., Japan). A portion of the faecal swabs (n = 84) was placed after the smear in a Preston enrichment broth (PEB) and was incubated for 24 h at 37°C under microaerophilic conditions. The faecal samples were either immediately DNA extracted or frozen at À20°C until processing.
Three drops (100 lL) of the enrichment broth were pipetted on a selective agar plate mCCDA (Campylobacter blood-free selective agar base, CM0739; Oxoid/UK, CCDA selective supplement, SR0155; Oxoid) and CASA (Campylobacter Selective Agar 20 BT 90; Chemunex AES, France), two dilutional streaks with an annealed laboratory wire loop were made and the plates were incubated under microaerophilic conditions at 37°C for 48 h. In addition, from some of the samples (n = 75), 1 mL supernatant of the enrichment broth was taken out and transferred into a 1.5 mL Eppendorf tube for DNA extraction. These samples were frozen at À20°C till processing. DNA extraction was performed using QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. In this method, the hippuricase gene served as target gene, which only occurs in C. jejuni. The C. jejuni strain typing using real-time PCR is more reliable than with the biochemical Hippuric acid reaction, because hippuricase negative strains (≤10%), in which there is no expression of the hippuricase gene, will also be detected. The remaining hippuricase negative respectively double negative strains were further differentiated by PCR (Wang et al. 2002;Jensen et al. 2005).

Antibiotic sensitivity testing
Antibiotic sensitivity testing was performed on 13 C. jejuni strains. For this, three different tests were used with various antibiotics.

Agar diffusion test (AGES/IVET Graz)
The bacterial strain to be tested was streaked out on a Mueller-Hinton agar plate with 5% sheep blood (Biom erieux). After that, antibiotic test plates with a defined concentration were applied on the agar plates and incubated under microaerophilic conditions 48 AE 2 h at 37 AE 1°C. The inhibition zones were measured and evaluated according to CLSI (Clinical and Laboratory Standards Institute) as sensitive, intermediate or resistant.

MIC testing (AGES/IMED Graz)
For the implementation of the MIC testing, the strains to be tested were plated out on Columbia blood agar (Biom erieux) and incubated under microaerophilic conditions 44 AE 4 h at 37 AE 1°C. This culture was used to prepare a suspension according to McFarland 0.5. From this suspension, 50 lL was transferred in Mueller-Hinton broth with TES/lysed Horse Blood (TREK Diagnostic Systems; Thermo Scientific, Waltham, MA, USA). The resistance determination was performed using the Sensititre â system (TREK Diagnostic Systems; Thermo Scientific), a technique for determining the MIC value. For this purpose, dehydrated AB gradients were applied to the wells of a microtiter plate, with the appropriately prepared bacterial suspension inoculated (with automatically inoculator) and incubated 44 AE 4 h at 37 AE 1°C under microaerophilic conditions. For the evaluation of the microtiter plates, SensiTouch â system was used. In this system, step by step, each AB gradient was accessed. The following antibiotics were used: ampicillin, amoxycillin/clavulanic acid, chloramphenicol, ciprofloxacin, colistin, erythromycin, gentamycin, imipenem, nalidixic acid, neomycin, streptomycin, tetracycline.

E-test (AGES/IVET Graz)
The E-test (Biom erieux) is a quantitative method for the determination of antibiotic susceptibility of bacteria. An Epsilon test strips, which was coated with a defined, ascending concentration of an antibiotic, was placed on Mueller-Hinton agar with 5% sheep blood (Biom erieux) after plating out the bacterial strain. Then, the plate was incubated for 48 AE 2 h at 37°C under microaerophilic conditions, the MIC value (point of intersection between the E-test strips and the germ growth boundary) was read. This test was only used for enrofloxacin.

Statistical methods
For statistical analysis of associations, the free software R (3.4.0) was used. To check for significant associations, an exact Fisher test was performed.
The number of positive samples in dogs was lower (5.6%; n = 28), measured by the total number of samples of this species, compared with cats 22 (6.4%). The number of positive C. jejuni samples in dogs (n = 11, 2.2%) was less than in cats (n = 15, 4.4%) ( Table 1). These differences were not statistically significant (a = 0.05). No statistically significant differences in regard to any species were found.
A comparison between mCCDA-and CASA agar (n = 86) revealed 2.3% C. spp.-positive samples on both agars at direct plating of faeces. When comparing direct plating on mCCDA-versus enrichment in Preston broth and striking out on mCCDA agar (n = 84), 9.5% and 6% were found positive, respectively.
The results of the antibiotic susceptibility testing of 13 C. jejuni strains are shown in Figs 1-3.
Possible reasons for this difference could be our focus on the most common human pathogenic species C. jejuni and C. coli which led to a study design not ideal for the detection of Campylobacter upsaliensis. Further reasons are the lower sensitivity of rectal swabs, in some cases a delay in plating out and the time allowed for in vitro growth.
The most frequently detected Campylobacter species was C. jejuni, the second most common type was C. upsaliensis. These results stand in contrast to most of the above-mentioned publications. The isolation rate of C. spp. in faeces of dogs and cats differed primarily in relation to the age of the animals (Torre & Tello 1993;Engvall et al. 2003;Hald et al. 2004;Bender et al. 2005;Wieland et al. 2005;Acke et al. 2009a,b,c;Holmberg et al. 2015), the predominant Campylobacter species (Hald et al. 2004;Wieland et al. 2005;Acke et al. 2009a,b,c) and the isolation method used (Koene et al. 2004;Acke et al. 2006Acke et al. , 2009a. Testing for the presence of C. jejuni with real-time PCR, 27% positive faecal samples and 8% positive faecal swabs were found. This suggests that the colonization rate was actually higher. Chaban et al.  In dogs, no noticeable age-dependent differences in Campylobacter positives between young animals under 1 year and adults were found. However, not a single specific C. spp. could be found in young dogs. Among the adult dogs, the amount of C. jejuni and C. upsaliensis-positive samples was also very low. Selwet et al. (2015) found more Campylobacter spp. positive in young dogs (60%) than in adults (38.9%). Hald et al. (2004) have created an accurate excretion pattern for thermotolerant C. spp. and found that the rate of Campylobacter carriers in dogs during development increases from 60% in young animals of 3 months to nearly 100% in animals of 12 months and falls back to 67% at the age of 24 months. Balucinska (1995)identified 45.3% C. spp.-positive animals among young dogs and only 26.6% C. spp. positives among adult dogs in Austria. In diarrhoeic dogs, the number of C. spp. as well as C. jejuni-positive animals was higher than the total positives.
Very striking was the difference in juvenile cats below 1 year of age compared to the total number of C. spp.-positive cats, while the percentage of C. spp.positive adult cats was well within the range of the total positives. It should be emphasized that there were without exception only C. jejuni-positive animals among positive kittens. This result was statistically significant (P = 0.034). Bender et al. (2005) found 30% Campylobacter-positive among juveniles but only 3% positive cats among adults. Gargiulo et al. (2008) revealed in his studies of stray cats 27.7% C. jejunipositive adult and only 2.1% positive juvenile cats.
The group of diarrhoeic cats had the highest proportion of C. spp positives. All positive diarrhoeic cats were positive for C. jejuni. Acke et al. (2006Acke et al. ( , 2009b noticed in dogs and cats with diarrhoea, especially in those under 6 months, high prevalence of C. spp. The differences between diseased and healthy animals as well as between the age groups were largely insignificant. In the animals with diarrhoea, C. jejuni was the most abundant species (Acke et al. 2009b). Balucinska (1995) found 31.6% C. spp. samples positive on diarrhoeic and 29.9% in healthy dogs in Austria.
In this study, when comparing the methods, matching results were mainly found in gyrase inhibitors (cipro-, enro-, marbofloxacin, nalidixic acid) and tetracyclines. The high rates of resistance in gyrase inhibitors and tetracyclines are comparable to those of the above-mentioned recent literature. This indicates a certain transmission risk of (multi) resistant, potentially human pathogenic C. jejuni strains from dogs and cats to humans.