Comparison of Antimicrobial Susceptibility Profiles of Thermotolerant Campylobacter spp. Isolated from Human and Poultry Samples in Georgia (Caucasus)

Antimicrobial resistance remains a public health concern globally. This study presents antimicrobial resistance by microdilution and genetic diversity by the whole-genome sequencing of Campylobacter spp. from human and poultry samples isolated in Georgia in 2020/2021. The major species in poultry samples was C. coli, while C. jejuni was preferentially isolated from human samples. Resistance against tetracycline was highest (100%) in C. coli from industrial chicken and lowest in C. jejuni from clinical isolates (36%), while resistance against ciprofloxacin varied from 80% in C. jejuni from backyard chicken to 100% in C. jejuni and C. coli from industrial chicken. The point mutations in gyrA (T86I) and tet (O) genes were detected as resistance determinants for (fluoro-)quinolone or tetracycline resistance, respectively. Ertapenem resistance is still enigmatic. All isolates displayed sensitivity towards erythromycin, gentamicin and chloramphenicol. Multi-resistance was more frequently observed in C. coli than in C. jejuni, irrespective of the isolation matrix, and in chicken isolates compared to human isolates, independent of the Campylobacter species. The Georgian strains showed high variability of multi-locus sequence types (ST), including novel STs. This study provides the first antibiotic resistance data from Campylobacter spp. in Georgia and addresses the need for follow-up monitoring programs.


Introduction
The emergence and spread of multi-resistant bacteria continues to be a global public health concern. In the European Economic Area (EEA), it was estimated that more than 670,000 diseases were caused by antimicrobial-resistant (AMR) bacteria yearly, with about 33,000 associated deaths [1].
Campylobacteriosis is a disease caused by thermotolerant Campylobacter spp. It is one of the four major causes of diarrhea worldwide, and is considered to be the most common cause of bacterial food-borne human gastroenteritis [2]. Campylobacter species are motile, curved, microaerobic, Gram-negative rods that commonly reside in the intestinal tract of many wild and domestic warm-blooded animals.
Although campylobacteriosis is mostly self-limiting, recent reports showed that a substantial proportion (31%) of reported Campylobacter infections have been treated with reared at small farms and households from all over the country and directly processed on the market slaughterhouse. Another 50 Campylobacter strains were isolated from samples collected at a medium-sized 'intensive-rear' poultry farm slaughterhouse, located at the eastern part of Georgia. In addition, 382 human stool samples had been previously collected from July 2020 to July 2021 at the Tbilisi Children Infectious Diseases Clinical Hospital from hospitalized children with diarrhea, from which 60 were positive for Campylobacter spp. [12]. Human stool samples were transported on Cary-Blair medium (Biolife Italiana srl, Milan, Italy) at cooling temperatures without microaerobic conditions and analyzed within 24 h. Chicken cecal samples were transported in plastic bags on ice and analyzed within 3-6 h after sampling.

Detection and Phenotypic Identification of Campylobacter spp.
Campylobacter detection was performed according to ISO 10272-1:2017 part C on modified charcoal cefoperazone deoxycholate agar (mCCDA) (Thermo Fisher Specialty Diagnostics Ltd., Hampshire, UK). For the clinical samples, Campylobacter Chromogenic agar Campylobacter (CHROMagar, France) was applied as an additional second selective medium to increase sensitivity [12]. Less than 20% of the clinical samples were also enriched with Preston broth (Biolife Italiana S.r.l., Milan, Italy) [13], but the results showed no enhanced detection [12].
Ceca were aseptically cut and the content mixed. One 1 µL loop of the cecal material was directly streaked on the mCCDA agar plate and distributed over the surface by using a fresh loop. The human stool samples were treated similarly but in addition to mCCDA a second selective plate was used in parallel. Incubation was performed at 42 • C in a microaerobic gas mixture consisting of 85% nitrogen, 10% carbon dioxide and 5% oxygen (LTD Argoni, Tbilisi, Georgia).
Suspicious colonies were sub-cultured on Columbia Blood Agar (ColbA; AES Laboratories, Bruz Cedex, France). Confirmation of colonies was initially performed applying the Biomerieux system ApiCampy (Biomerieux Inc, Marcy-l'Etoile, Lyon, France), consisting of 20 microtubes containing dehydrated substances. One half contained enzymatic tests and the other half substrates for assimilation or inhibition. In the latter, growth of bacteria is monitored. The specific pattern of growth and presence of enzymatic activity is used as read-outs for identification of bacteria. In addition, colonies were observed by microscopy after Gram-straining. All isolates were stored at −80 • C for further characterization.

Confirmation of Campylobacter Species and Differentiation by Real-Time PCR Analysis
At the National Reference Laboratory for Campylobacter at BfR the 220 strains, from which 160 were derived from chicken and 60 from human sources, were re-cultured on ColbA for 48 h under microaerobic atmosphere. In case no growth or some contamination was obtained, a parallel enrichment in Bolton broth (Oxoid, Thermo Fisher Scientific Inc., Waltham, MA, USA) with 5% lysed defibrillated horse blood (Oxoid, Thermo Fisher Scientific Inc., Waltham, MA, USA) was streaked on mCCDA and incubated for another 48 h. Single suspected colonies were sub-cultured on ColbA and incubated 24 h under similar conditions.

Statistical Analyses
Isolates were categorized into susceptible and resistant, using the epidemiological cut-off values as mentioned in Section 2.4. The dependent variable was resistant vs. susceptible (reference category) to the antimicrobial in question. In addition to the individual antimicrobial, an outcome variable "2-3-fold resistance" was defined for an isolate resistant against two or three tested antimicrobials. This means that first, isolates were categorized according to their MIC and the epidemiological cut-off value (ECOFF) as sensitive or resistant towards every individual antimicrobial. Second, the number of resistances per isolate was counted and those with 2 or more resistances were defined as displaying "2-3-fold resistance".
Multiple logistic regression with forward selection was used to establish independent predictors for tetracycline resistance (variables of matrix source (human vs. chicken (reference category)) and bacterial species (C. coli vs. C. jejuni (reference category)) were included). A Nagelkerke R Square and a non-standardized beta coefficient (B) were calculated. An odds ratio with 95% confidence interval (CI) was calculated as an exponential of For all analyses, p-values of less than 0.05 were considered statistically significant. Statistical analyses were performed using SPSS (IBM SPSS Statistics for Windows, Version 21.0. Armonk, NY, USA: IBM Corp).

Collection of Campylobacter spp. Strains and Identification of Species
Campylobacter spp. isolates from chicken cecal content were obtained from February 2020 until September 2021. "Backyard" chicken samples aged between several days to one year were collected from chicken reared on small farms and in households all over the country and sold at a live market in Tbilisi. In addition, Campylobacter strains were isolated from samples collected at a medium-sized industrial poultry slaughterhouse, located at the eastern part of Georgia and supplying Tbilisi with fresh chicken meat. Those chickens were "standardized" with an age between 38 and 42 days. In addition, human stool isolates had been previously collected from hospitalized children with diarrhea from July 2020 to July 2021 [12]. Hence, the samples correlated in time and space. From a total of 220 isolates-160 derived from chicken and 60 from human sources (Supplementary Materials Table S1)-sixteen were non-culturable after transport to BfR. However, from these sixteen non-culturable samples, Campylobacter spp. were still detectable by real-time PCR in twelve of the enrichment inoculums, showing either C. coli Antibiotics 2022, 11, 1419 6 of 13 (4/12) or C. jejuni (3/12) in seven cases and mixed cultures of C. coli and C. jejuni in five cases (41%, n = 5/12).
Out of 204 strains re-cultured, 37.7% (n = 77) were identified as C. jejuni and 62.3% (n = 127) as C. coli applying real-time PCR [14]. The distribution of isolated species differed between human stool samples and cecal chicken samples (Figure 1).
from samples collected at a medium-sized industrial poultry slaughterhouse, located at the eastern part of Georgia and supplying Tbilisi with fresh chicken meat. Those chickens were "standardized" with an age between 38 and 42 days. In addition, human stool isolates had been previously collected from hospitalized children with diarrhea from July 2020 to July 2021 [12]. Hence, the samples correlated in time and space. From a total of 220 isolates-160 derived from chicken and 60 from human sources (Supplementary Material  Table S1)-sixteen were non-culturable after transport to BfR. However, from these sixteen non-culturable samples, Campylobacter spp. were still detectable by real-time PCR in twelve of the enrichment inoculums, showing either C. coli (4/12) or C. jejuni (3/12) in seven cases and mixed cultures of C. coli and C. jejuni in five cases (41%, n = 5/12).
Out of 204 strains re-cultured, 37.7% (n = 77) were identified as C. jejuni and 62.3% (n = 127) as C. coli applying real-time PCR [14]. The distribution of isolated species differed between human stool samples and cecal chicken samples (Figure 1).

Prevalence of Antimicrobial Resistance in Campylobacter Isolates.
All isolates were tested for their resistance to the six antimicrobials chloramphenicol, ciprofloxacin, ertapenem, erythromycin, gentamicin and tetracycline according to the European standardized EUCAMP3 plate format. Results from resistance testing are shown in Table 1. All tested strains were sensitive towards gentamicin, erythromycin and chloramphenicol. Resistance in both human and poultry isolates and in both bacterial species was highest against ciprofloxacin and tetracycline. From the isolates of backyard chicken, 25.8% were identified as C. jejuni (n = 25/97) and 74.2% (n = 72/97) as C. coli; in cecal samples from industrial chicken, C. coli was even more dominant with 90% (n = 45/50). In contrast, out of 57 clinical strains of children stool samples, 82.5% (n = 47/57) were identified as C. jejuni and 17.5% (n = 10/57) as C. coli ( Figure 1) [12].

Prevalence of Antimicrobial Resistance in Campylobacter Isolates
All isolates were tested for their resistance to the six antimicrobials chloramphenicol, ciprofloxacin, ertapenem, erythromycin, gentamicin and tetracycline according to the European standardized EUCAMP3 plate format. Results from resistance testing are shown in Table 1. All tested strains were sensitive towards gentamicin, erythromycin and chloramphenicol. Resistance in both human and poultry isolates and in both bacterial species was highest against ciprofloxacin and tetracycline.   Both human and poultry C. coli strains showed resistance against ertapenem-37% of the strains from backyard chicken, 60% of human isolates and 82% of industrial chicken strains, while C. jejuni isolates were fully susceptible to this antimicrobial. Among the ertapenem-resistant C. coli, 89% (n = 62) had a MIC value of 1 µg/mL, just above the current cut-off value, 10% (n = 7) displayed a MIC of 2 µg/mL and a single strain had a MIC of 4 µg/mL. From the strains with MIC values ≥2 µg/mL ETP, three were derived from human samples, four from backyard chicken and one from industrial chicken.
Overall, isolates of C. coli were less frequently fully susceptible (3/127, 2.4%) than isolates of C. jejuni (9/77, 11.6%), with each six strains isolated from backyard poultry and human samples and lack of susceptible strains among the industrial isolates ( Figure 2). 105 (83%) ECOFF, epidemiological cut-off for definition of resistance against antimicrobial substances (EU-CAST.org); R>, maximal MIC that represents sensitivity; any MIC exceeding this concentration is defined as resistant. Note that ECOFF for erythromycin and tetracycline differs for Campylobacter species. n, number of tested isolates; numbers in table represent numbers of resistant isolates; in brackets, percentage of resistant isolates.
Both human and poultry C. coli strains showed resistance against ertapenem-37% of the strains from backyard chicken, 60% of human isolates and 82% of industrial chicken strains, while C. jejuni isolates were fully susceptible to this antimicrobial. Among the ertapenem-resistant C. coli, 89% (n = 62) had a MIC value of 1 µ g/mL, just above the current cut-off value, 10% (n = 7) displayed a MIC of 2 µ g/mL and a single strain had a MIC of 4 µg/mL. From the strains with MIC values ≥2 µ g/mL ETP, three were derived from human samples, four from backyard chicken and one from industrial chicken.
Overall, isolates of C. coli were less frequently fully susceptible (3/127, 2.4%) than isolates of C. jejuni (9/77, 11.6%), with each six strains isolated from backyard poultry and human samples and lack of susceptible strains among the industrial isolates ( Figure 2).  Table 1 were counted per isolate and percentage of isolates with resistances against x-fold antimicrobial classes are depicted here.
C. coli were more likely resistant-compared to C. jejuni-against ciprofloxacin (OR 5.1, 95% CI 1.6-16.7) and tetracycline (OR 4.6, 95% CI 2.5-8.8). In addition, isolates from clinical samples were less likely resistant to tetracycline compared to chicken isolates (OR Isolates with x-fold resistance 0 1 2 3  Table 1 were counted per isolate and percentage of isolates with resistances against x-fold antimicrobial classes are depicted here. C. coli were more likely resistant-compared to C. jejuni-against ciprofloxacin (OR 5.1, 95% CI 1.6-16.7) and tetracycline (OR 4.6, 95% CI 2.5-8.8). In addition, isolates from clinical samples were less likely resistant to tetracycline compared to chicken isolates (OR 0.18, 95% CI 0.1-0.4). No statistically significant difference was observed for resistance to ciprofloxacin between human and poultry isolates.
Overall, C. coli was 18.5 times more likely resistant against two or more antibiotics compared to C. jejuni (OR 18.5, 95% CI 7.7-44.8). The same was observed in clinical isolates, where C. coli was 17.4 times more likely resistant to two or more antimicrobials than C. jejuni (OR 17.4, 95% CI 2.03-150.1); for poultry samples C. coli OR showed 7.9 times more probability to have resistance against two or more antibacterial agents compared to C. jejuni (OR 7.9,95% CI 2.6-24.6).
There was a significant association of multi-resistance probability with isolation source in C. jejuni strains. In particular, the probability of resistance against two or more antimicrobials for chicken isolates of C. jejuni was 4.5 times higher compared to human isolates (OR 4.5, 95% CI 1.7-12.1); however, we did not find a significant association between clinical and chicken isolates for C. coli species, probably due to low number of C. coli isolates from human stool samples. Additionally, no statistically significant difference was found for the presence of two or more resistances in C. jejuni or in C. coli isolates from industrial compared to backyard chicken.
Variables of bacterial species and isolates were subjected to logistic regression analysis to test association with resistance to two or more antimicrobials as dependent variables. Both variables were retained in the final model as independent variables. The Nagelkerke Antibiotics 2022, 11, 1419 8 of 13 pseudo R squared was 0.435 indicating that more than 43% of the variability of dependent variables is due to the independent variables model.
Multi-variate logistic regression was performed with two variables which showed significant association with tetracycline resistance. Both variables, bacterial species and sample sources were retained into final model as independent predictors. The regression model can explain more than 20% of the variation in the dependent variable (tetracycline resistance), according to the Nagelkerke pseudo R squared of 0.204. (Table 2). In other words, the predictive model, consisting of the variables "bacterial species" and "sample sources", can explain 20% of the variability of the dependent variable "tetracycline resistance". Alternatively, this means, that the remaining 80% of the variability of the dependent variable could be explained with variables, that were not measured within the study and/or are not identified as a possible predictor for the outcome variable. Nagelkerkes R squared 43% for the dependent variable "2-3-fold resistance" can be interpreted in the same way.

Campylobacter spp. Isolates Are Phylogenetically Diverse
We additionally analyzed forty Campylobacter strains by whole-genome sequencing, twenty derived from poultry and another twenty from human samples, approximately each ten C. jejuni and C. coli per matrix. The poultry isolates were both from backyard samples (n = 14) and from industrial chicken (n = 6). After de novo assembly of the raw reads, multilocus sequence type analysis (MLST, based on 7 housekeeping genes) and, for more precise resolution, the core-genome MLST (cgMLST) scheme based on the comparison of 1343 gene alleles was used for phylogenetic analysis. Missing cgMLST loci were pairwise ignored.
As expected, we obtained a high variability of multi-locus sequence types (ST, n = 24), including three strains with either unknown uncA allele and/or unknown ST-type. The Within the limited number of sequenced strains, we even found three sequence clusters. One of this clusters (ST-855) included four highly similar C. coli strains from industrial chicken, collected in June/July 2021 during three independent samplings, with maximal two cgMLST allele differences. Two further C. jejuni clusters with each two strains identified among the human isolates belonged both to ST-type 356 and were separated from each other by 226 allele difference. One of these clusters included two C. jejuni strains isolated from children in September and October 2021, harboring identical pairwise cgMLST. The other cluster included two C. jejuni strains isolated from children in July and September 2021.  Table S2 highlights new ST-types and their respective allelic combinations not previously reported in the PubMLST database as well as the metadata of the dataset. Figure 3. Whole-genome sequences of the isolates from chicken and human samples in Georgia displayed high variability. Minimum spanning tree of cgMLST analysis was based on 1343 core genes defined previously [20]. Missing alleles were pairwise ignored. Each colored circle with (C. jejuni) our without frame (C. coli) represents an ST-type of the 7 housekeeping genes MLST scheme as depicted in the inlay boxes per species. Numbers next to the connecting lines illustrate the number of allele differences analyzed by cgMLST between nearest neighbors. One new uncA allele and two new ST-types were found. More details, including all ST-types are shown in Supplementary Material Table S2.
Within the limited number of sequenced strains, we even found three sequence clusters. One of this clusters (ST-855) included four highly similar C. coli strains from industrial chicken, collected in June/July 2021 during three independent samplings, with maximal two cgMLST allele differences. Two further C. jejuni clusters with each two strains identified among the human isolates belonged both to ST-type 356 and were separated from each other by 226 allele difference. One of these clusters included two C. jejuni strains isolated from children in September and October 2021, harboring identical pairwise Figure 3. Whole-genome sequences of the isolates from chicken and human samples in Georgia displayed high variability. Minimum spanning tree of cgMLST analysis was based on 1343 core genes defined previously [20]. Missing alleles were pairwise ignored. Each colored circle with (C. jejuni) our without frame (C. coli) represents an ST-type of the 7 housekeeping genes MLST scheme as depicted in the inlay boxes per species. Numbers next to the connecting lines illustrate the number of allele differences analyzed by cgMLST between nearest neighbors. One new uncA allele and two new ST-types were found. More details, including all ST-types are shown in Supplementary Materials Table S2.
Eighteen Campylobacter isolates (45%) putatively carried plasmids (Supplementary  Materials Table S2), since contigs of the whole genome assembly were predicted as epichromosomal elements by Platon and BLAST analysis using the NCBI RefSeq plasmid database. All plasmids had at least 20% coverage of homology to known Campylobacter spp. plasmids (Supplementary Materials Table S3), except for BfR-CA-19911, which harbored a small plasmid without any match in the RefSeq database.

Detection of Antimicrobial Resistance Genes
Whole-genome sequencing analysis also revealed several resistance genes, responsible for the observed phenotypes. The presence of the tet(O) gene, which mediates resistance to tetracycline, was detected in all tetracycline-resistant strains (70%, n = 28/40). The most common mutation in the gyrA gene (T86I) was identified in all ciprofloxacin-resistant isolates (90% (n = 36/40)). The presence of bla OXA-61 family genes (OXA-193, OXA-452, OXA-460, OXA-461, OXA-489, OXA-594), which confer resistance to beta-lactams, was observed in 75% (n = 30/37) of strains. In addition, we found the aadE-Cc gene in three C. coli, putatively conferring streptomycin resistance. Streptomycin and ampicillin are not part of EUCAMP3 plate format, so the phenotype was not confirmed. The AMRFinderPlus database also annotated the mutation 50S_L22_A103V of the L22 ribosomal protein as a putative resistance marker for macrolide resistance in 30% (n = 12/37) of the strains; however, all isolates were sensitive towards erythromycin. The resistance mechanism against ertapenem is still unknown. According to Platon prediction, all resistance determinants were chromosomally located.

Discussion
EU countries have made significant strides in developing and implementing national monitoring plans on antimicrobial resistance [6]; however, in Georgia, monitoring programs are still lacking.
Our study results on antibiotic resistance in Georgian Campylobacter spp. isolates from chicken show similarities to the AMR data profiles of Campylobacter spp. in EU member states. In particular, both C. jejuni and C. coli from poultry sources in the EU exhibited high resistance against (fluoro-)quinolones and tetracycline, which is in line with our data [6,23,24]. However, notably, the resistance rate to ciprofloxacin and tetracycline was 100% in isolates from industrial poultry samples in Georgia, while in backyard chicken and in human isolates Campylobacter strains displayed slightly lower resistance against both antimicrobials. Comparing multi-resistance in C. jejuni or C. coli in industrial versus backyard chicken, no significant difference could be found. Interestingly, all isolates were sensitive towards gentamicin, chloramphenicol and erythromycin.
Use of (fluoro-)quinolones was shown to be the major risk factor for ciprofloxacin resistance in Campylobacter spp. on broiler farms [25]. However, it was shown that the gyrA mutation, conferring resistance against (fluoro-)quinolones, can also contribute to a fitness increase in C. jejuni in poultry depending on the strain background [26]. The clonal spreading of (fluoro-)quinolone-resistant clones was suggested to occur in Europe [27], although the contribution of whether the resistance was selected through (fluoro-)quinolone use in individual countries and/or transmission between countries is still unclear [28]. Moreover, the differences in resistance rates between the bacterial species from the same source and, therefore, the same antimicrobial exposure indicated that antimicrobial use alone cannot explain differences in resistance profiles of C. jejuni and C. coli [29]. C. coli from the same matrix exhibited higher resistance than C. jejuni towards multiple antimicrobials tested [29]. The reason for this phenomenon is still unclear. (Fluoro-)quinolones are among WHOs "Highest Priority Critically Important Antimicrobials" (HPCIA) [30]. Increases in resistance to (fluoro-)quinolones in Campylobacter spp. are of concern, as resistance in Campylobacter from animals has been shown to be associated with resistance of Campylobacter from human infections [6]. When Georgian isolates were compared according to their origin, the chicken C. coli or C. jejuni isolates were each significantly more resistant towards two and three classes of antimicrobials than the human strains. This might hint to additional infection routes other than cross-contamination from preparing fresh chicken meat and/or direct contact to animals on chicken farms in Georgian children suffering from campylobacteriosis. In addition to the preparation of poultry meat and contact with poultry animals, contact with sand in a sandbox with putative contact to animal feces such as that from dogs and wild animals was also identified in a German study as risk factor positively associated with a Campylobacter infection for children under 5 years of age [3].
Furthermore, our study showed a high prevalence of C.coli in comparison to C.jejuni from poultry samples, which was untypical in a number of countries even in the Caucasus region [6,[31][32][33]. However, there are other studies that identified a higher prevalence of C. coli than C. jejuni in swab samples from farms and neck skins at slaughter in Italy [34] or some alterations of species distribution depending on the stage of broiler production [35]. A long-term study over seven years showed a gradual decrease in the prevalence of C. jejuni and a concomittant increase in C. coli in cecal samples from chicken in China [36], while in Malaysia both species were frequently isolated from different broiler parts [37].
One explanation for different species distribution might be age and race of the chicken, which is not likely in our study, since we obtained a similar species distribution from backyard chicken of different age and industrial chicken with standardized rearing period of 38-42 days. Our results may additionally hint at the fact that initially, we might have isolated mixed cultures of both C. jejuni and C. coli in some cases, since PCR results of inoculums identified the presence of both species, which in turn could not be recultivated together.
All tested isolates from Georgia were sensitive towards erythromycin and gentamicin, which was similar for isolates in the EU. Erythromycin resistance in Campylobacter isolates from human cases of campylobacteriosis and from broilers in sixteen EU member states was either absent or detected at very low levels in C. jejuni, but was observed at higher levels in C. coli isolates. Overall, erythromycin resistance was reported in 10% (2020) and 12.9% (2019) of human isolates and 4.4% of broiler isolates. Combined resistance to both ciprofloxacin and erythromycin, which is considered critical for the treatment of campylobacteriosis, was reported to be 8.9% (2020) and 10.4% (2019) in isolates from humans and 4.1% in broilers. In 2020, EU countries reported low prevalence of gentamicin resistance [6]. Data from C. jejuni and C. coli of human and animal origin in 2019-2020 showed very high to extremely high levels of resistance to (fluoro-)quinolones, which are also critically important antimicrobial agents (CIAs) for the treatment of Campylobacter infections in humans [30]. WGS of isolates, especially those with multi-drug resistance, highlevel resistance to erythromycin or ciprofloxacin, or resistance to gentamicin or ertapenem, is strongly recommended in order to decipher the antimicrobial resistance determinants involved, their genetic location, and the potential for horizontal transmission [38].

Conclusions
Preventive and control activities in Georgia are still limited concerning the monitoring and antimicrobial susceptibility profiling of thermotolerant Campylobacter spp. Our first national study showed similar AMR patterns of thermotolerant Campylobacter spp. strains isolated in Georgia to those reported by the European Union. In particular, resistances against (fluoro-)quinolones and tetracycline were high and should be considered in local therapeutic protocols for severe human cases. Antimicrobial resistance and the prevalence of thermotolerant Campylobacter spp. in animals, food and humans need further approaches in order to gain a representative picture of concurrent strains in the Caucasian region.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/antibiotics11101419/s1, Table S1: Complete sample list, including phenotype of antibiotic resistance; Table S2: WGS data overview and antibiotic resistance of tested samples; Table S3: WGS data-plasmid annotation.