Antibiotic Resistance in Campylobacter spp. Isolated from Broiler Chicken Meat and Human Patients in Estonia

Poultry meat is considered the most important source of Campylobacter spp. Because of rising antimicrobial resistance in Campylobacter spp., this study investigated the antimicrobial resistance of Campylobacter isolates from fresh broiler chicken meat originating from the Baltic countries sold in Estonian retail settings. Additionally, human clinical isolates obtained from patients with Campylobacter enteritis in Estonia were analysed. The aim of this study was to investigate the susceptibility of Campylobacter spp. to nalidixic acid, ciprofloxacin, tetracycline, streptomycin, erythromycin and gentamicin. The broth microdilution method with the EUCAMP2 panel was used for MIC determination and antimicrobial mechanisms were analysed using WGS data. A total of 46 Campylobacter strains were analysed, of which 26 (42.6%) originated from Lithuanian, 16 (26.2%) from Latvian, and 4 (6.6%) from Estonian fresh broiler chicken meat. In addition, 15 (24.6%) Campylobacter strains of patients with campylobacteriosis were tested. The antimicrobial resistance patterns of Campylobacter spp. isolated from fresh broiler chicken meat samples of Estonian, Latvian and Lithuanian origin collected in Estonian retail, and from patients with Campylobacter enteric infections, were determined. A total of 46 (75%) of the isolates tested were C. jejuni and 15 (25%) were C. coli. Campylobacter resistance was highest to nalidixic acid (90.2% of strains) and ciprofloxacin (90.2%), followed by tetracycline (57.4%), streptomycin (42.6%) and erythromycin (6.6%). All strains were sensitive to gentamicin. Additionally, antimicrobial resistance genes and point mutations were detected in 27 C. jejuni and 8 C. coli isolates previously assigned as resistant with the phenotypic method. A high antibiotic resistance of Campylobacter spp. in Lithuanian- and Latvian-origin broiler chicken meat and Estonian clinical isolates was found. Similar antibiotic resistance patterns were found for broiler chicken meat and human Campylobacter isolates.


Introduction
Campylobacter is small spiral Gram-negative bacterial pathogen and the main agent of campylobacteriosis, which is the most common zoonotic disease in the EU transmitted to humans directly from animals, or through the foodborne route [1,2]. Recent source attribution analysis has revealed the prevalent role of broiler chickens as the cause of human Campylobacter infections in the Baltic countries [3]. Broilers and broiler chicken meat are the most important sources of Campylobacter infections in humans [4][5][6]. Disease caused by Campylobacter spp. is generally mild and self-limited. Nevertheless, it can cause severe systematic infection in children/elderly people and humans with immunosuppression, and also-in very rare cases-Guillan-Barré syndrome [7,8]. Such occasions often require therapy with first-line antibiotics such as fluoroquinolones, e.g., ciprofloxacin, and macrolides, e.g., erythromycin [9][10][11][12][13][14]. Over time, Campylobacter has acquired resistance to these antibiotics which are considered critically important for the treatment of Campylobacter infections [11,15,16]. The increasing resistance of thermophilic Campylobacter spp. to antibiotics could lead to detrimental effects on public health [2,[17][18][19]; therefore, the World Health Organization has classified many of these antimicrobials as critically important for human medicine [20].
Previous studies [21,22] have identified different levels of Campylobacter-contaminated poultry meat in the Baltic states, and the proportion of antimicrobial-resistant Campylobacter spp. strains in poultry meat has been found to be very high in both Lithuania and Latvia [23,24]. This is affecting public health and needs to be addressed.
This study aimed to determine the proportions of antimicrobial-resistant Campylobacter strains from fresh broiler chicken meat of Estonian, Latvian and Lithuanian origin at the Estonian retail level, and among strains of human clinical infections. Additionally, resistance pheno-and genotypes were determined and compared. The results of this work will make it possible to assess the trends in antimicrobial resistance over a long period and determine related public health risks.

Campylobacter Isolates
In total, 61 Campylobacter isolates were studied, of which 46 (75%) were C. jejuni and 15 (25%) C. coli. From the broiler chicken meat isolates, 34 (73.9%) were C. jejuni and 12 were (26.1%) C. coli. Among the clinical isolates, 12 (80%) and 3 (20%) were C. jejuni and C. coli, respectively. All isolates were obtained from a previous study on Campylobacter spp. in Estonia [25]. The samples consisted of all Campylobacter strains isolated from Estonian (n = 4), Latvian (n = 16) and Lithuanian (n = 26) fresh broiler chicken meat products in Estonian retail from our previous study [25]. Additionally, isolates from human patients (n = 15) originating from ambulatory and hospitalized patients from north and northeastern Estonia were obtained from East-Tallinn Central Hospital and Rakvere Hospital were included. Campylobacter detection from broiler chicken meat samples was performed according to ISO 10272-1:2017 [26], as described by Tedersoo et al. [25]. In brief, for Campylobacter spp. detection, 10 g of broiler chicken meat sample was transferred into 90 mL Preston enrichment broth and incubated in microaerobic conditions at 41.5 ± 0.5 • C for 24 h. Then, a 10 µ loopful of Preston enrichment broth was inoculated onto mCCD agar medium (Oxoid Ltd.; Basingstoke, Hampshire, UK). All plates were incubated in microaerobic conditions at 41.5 • C ± 0. 5 • C for 48 h. Typical Campylobacter colonies on mCCDA plates were streaked on Columbia blood agar (Oxoid Ltd.; Basingstoke, Hampshire, UK), which were incubated for 24 h at 41.5 • C ± 0.5 • C. Additional confirmation tests included Gram staining, motility analysis, and oxidase and catalase tests were performed.

Antimicrobial Susceptibility Testing
Minimal inhibitory concentration (MIC) values for nalidixic acid, ciprofloxacin, tetracycline, streptomycin, erythromycin and gentamicin were determined by using the broth microdilution method with the EUCAMP2 panel (TREK diagnostic Systems Ltd., East Grinstead, UK) in accordance with the manufacturer's protocols.
The cut-off values recommended by the European Committee on Antimicrobial Susceptibility Testing were used for C. jejuni and C. coli in accordance with the European Commission implementing decision 2013/652/EU on the monitoring and the reporting of antimicrobial resistance in zoonotic and commensal bacteria. C. jejuni was assigned resistant when MIC values equated to: erythromycin > 4 µg/mL, ciprofloxacin > 0.5 µg/mL, tetracycline > 1 µg/mL, streptomycin > 4 µg/mL, nalidixic acid > 16 µg/mL or gentamicin > 2 µg/mL. C. coli was assigned resistant when MIC values equated to: erythromycin > 8 µg/mL, ciprofloxacin > 0.5 µg/mL, tetracycline > 2 µg/mL, streptomycin > 4 µg/mL, nalidixic acid > 16 µg/mL or gentamicin > 2 µg/mL. Analyses were performed in the Veterinary and Food Laboratory of Estonia, which is also the national reference laboratory.

Whole-Genome Sequencing and Analysis of Resistant Genes
Molecular analysis, whole-genome sequencing and bioinformatics were performed as described by Tedersoo et al. [25]. In brief, the sequencing was carried out on an Illumina NextSeq500 System (Illumina, Inc.; San Diego, CA, USA) using the NextSeq 500/550 High Output Kit v2.5 (300 Cycles) in paired-end 2 × 151 bp mode. All genome sequences were submitted to the C. jejuni/C. coli multilocus sequence typing (MLST) database [27].
Antimicrobial resistance genes and point mutations of C. jejuni (n = 27) and C. coli (n = 8) isolates previously assigned as resistant with the MIC test were detected in the subset of isolates. MIC-sensitive campylobacters were not included in the analysis and only genotypic resistance mechanisms corresponding to phenotypic AMR were identified and reported in present study. AMRFinderPlus v3.10.23 with database v2021-12-21.1 (downloaded 3 March 2022) was used according to the default settings, except for the organism "Campylobacter" and the "plus" options [28,29]. Genes with coverage of less than 80% were not included in the analysis.

Statistical Analysis
MS Excel 2010 software (Microsoft Corporation; Redmond, WA, USA) was used to record the results. The Chi-squared test was used to test for statistically significant associations between the antimicrobial resistance of Campylobacter spp. in fresh broiler chicken meat from different sources. The results were considered statistically significant for p values of ≤0.05.

Results
The results showed that a total of six (9.8%) isolates were sensitive to all the tested antibiotics: four isolates from Estonian-origin chicken meat and two human Campylobacter strains. Campylobacter isolates of broiler chicken meat origin showed the highest resistance to quinolones, tetracycline and streptomycin. In addition, clinical Campylobacter isolates were found to be most resistant against the same antibiotics. All Campylobacter strains were sensitive to gentamicin (Table 1). Significant differences (p < 0.001) were found in nalidixic acid, ciprofloxacin and tetracycline resistance among Estonian versus Latvian and Lithuanian Campylobacter isolates from fresh broiler chicken meat. The Estonian broilerchicken-meat-origin Campylobacter isolates were significantly (p < 0.001) less resistant to fluoroquinolones than those strains which originated from Latvian and Lithuanian broiler chicken meat and Estonian human patients. There were no differences detected in streptomycin, erythromycin or gentamicin resistance between Campylobacter broiler chicken meat isolates of Estonian versus Latvian and Lithuanian origin. Resistance in the human isolates and broiler chicken meat isolates of Latvian (p = 0.13) and Lithuanian (p = 0.06) origin did not differ significantly. A total of 55 (90.2%) isolates were resistant to one or more antibiotics: 10 (16.4%) were resistant to one antibiotic, 28 (45.9%) were resistant to at least two antibiotics not belonging to the same group of antimicrobials (fluoroqinolones and quinolone (ciprofloxacin, nalidixic acid), macrolides (erythromycin), tetracyclines (tetracycline) and aminoglycosides (streptomycin, gentamicin)), and 17 (27.9%) isolates were resistant to three or more antibiotics not belonging to the same group. The proportion of isolates resistant to C. jejuni and C. coli was 87% and 100%, respectively. Antimicrobial resistance to one or more antimicrobial was significantly higher (p < 0.001) in the Campylobacter isolates from the broiler chicken meat of Latvian and Lithuanian origin compared to that of Estonian origin. It was found that 27.9% of isolates were multidrugresistant, of which 11 isolates (18.0%) were of Lithuanian and 2 (3.3%) of Latvian broiler chicken meat origin, and 4 (6.6%) were from Estonian human patients. Multidrug resistance was defined as resistance to three or more antibiotics not belonging to the same group.
All Latvian and Lithuanian isolates originating from broiler chicken meat were resistant to fluoroquinolones.
The resistance phenotypes of Campylobacter isolates are presented in Table 2. The most prevalent antimicrobial resistance pattern was Cip/Nal/Tet, with 55.3% and 21.7% in human and chicken meat isolates, respectively. Other common resistance phenotypes were Cip/Nal/Tet/Str and Cip/Nal/Str. Table 2. Campylobacter-resistant phenotypes.

Antibiotic Resistance Phenotype a,b
Campylobacter spp. Number of Strains (%)
Altogether, 28 C. jejuni and 7 C. coli isolates, of which 29 were of broiler chicken meat origin and 6 were of human origin, previously assigned as resistant with the MIC test, were sequenced and all their antimicrobial resistance genes and point mutations are presented in Table 4. In total, 29 Campylobacter isolates from broiler chicken meat and 6 Campylobacter isolates of human origin showed resistance to quinolones, and all contained a point mutation T86I in the gyrA gene. The genotypic antibiotic resistance against tetracyclines (tetO) was 62% and 83% in broiler chicken meat (n = 18) and human isolates (n = 5), respectively. A total of 52% of broiler chicken meat isolates (n = 15) showed resistance against aminoglycosides and macrolides. The resistance against macrolides in human isolates was 50% (n = 3).   Altogether, 28 C. jejuni and 7 C. coli isolates, of which 29 were of broiler ch origin and 6 were of human origin, previously assigned as resistant with th were sequenced and all their antimicrobial resistance genes and point mutati sented in Table 4. In total, 29 Campylobacter isolates from broiler chicken meat pylobacter isolates of human origin showed resistance to quinolones, and all point mutation T86I in the gyrA gene. The genotypic antibiotic resistance aga clines (tetO) was 62% and 83% in broiler chicken meat (n = 18) and human iso respectively. A total of 52% of broiler chicken meat isolates (n = 15) showe against aminoglycosides and macrolides. The resistance against macrolides in lates was 50% (n = 3).
Microorganisms 2022, 10, x FOR PEER REVIEW 5 of 11  Altogether, 28 C. jejuni and 7 C. coli isolates, of which 29 were of broiler chicken meat origin and 6 were of human origin, previously assigned as resistant with the MIC test, were sequenced and all their antimicrobial resistance genes and point mutations are presented in Table 4. In total, 29 Campylobacter isolates from broiler chicken meat and 6 Campylobacter isolates of human origin showed resistance to quinolones, and all contained a point mutation T86I in the gyrA gene. The genotypic antibiotic resistance against tetracyclines (tetO) was 62% and 83% in broiler chicken meat (n = 18) and human isolates (n = 5), respectively. A total of 52% of broiler chicken meat isolates (n = 15) showed resistance against aminoglycosides and macrolides. The resistance against macrolides in human isolates was 50% (n = 3).   Altogether, 28 C. jejuni and 7 C. coli isolates, of which 29 were of broiler ch origin and 6 were of human origin, previously assigned as resistant with th were sequenced and all their antimicrobial resistance genes and point mutati sented in Table 4. In total, 29 Campylobacter isolates from broiler chicken meat pylobacter isolates of human origin showed resistance to quinolones, and all point mutation T86I in the gyrA gene. The genotypic antibiotic resistance aga clines (tetO) was 62% and 83% in broiler chicken meat (n = 18) and human iso respectively. A total of 52% of broiler chicken meat isolates (n = 15) showe against aminoglycosides and macrolides. The resistance against macrolides in lates was 50% (n = 3). Table 4. Comparison of C. jejuni and C. coli phenotypic and genotypic antibiotic resist ing mechanisms, patterns, sources and origin.

Antibiotic (Class)
Phenotype/Genotype (n/n)   Altogether, 28 C. jejuni and 7 C. coli isolates, of which 29 were of broiler chicken meat origin and 6 were of human origin, previously assigned as resistant with the MIC test, were sequenced and all their antimicrobial resistance genes and point mutations are presented in Table 4. In total, 29 Campylobacter isolates from broiler chicken meat and 6 Campylobacter isolates of human origin showed resistance to quinolones, and all contained a point mutation T86I in the gyrA gene. The genotypic antibiotic resistance against tetracyclines (tetO) was 62% and 83% in broiler chicken meat (n = 18) and human isolates (n = 5), respectively. A total of 52% of broiler chicken meat isolates (n = 15) showed resistance against aminoglycosides and macrolides. The resistance against macrolides in human isolates was 50% (n = 3).   ERY-erythromycin. Bold indicates concurrence between genotypic and phenotypic resistance. b j-C. jejuni; c-C. coli. c One isolate also had the aad9 gene. d 50S L22 (A103V) modification was detected in 14 erythromycin MIC-sensitive isolates.

Discussion
As stated in the European Union Regulation No. 1831/2003, antimicrobials as growth promoters in food animal production have been banned since 2006 [30]. Antimicrobials are still used intensively in poultry for therapy and infection prophylaxis, which has caused the spread of resistant strains to humans [31,32]. However, some countries are showing positive trends, for example, antimicrobial use in poultry in Scandinavian countries is generally low. Denmark has declared carbapenems, third-and fourth-generation cephalosporins, fluoroquinolones and colistin as 'critically important', and the use of these antimicrobials is restricted [33]. Based on the report of DANMAP [33], cephalosporins and colistin are not used in Danish poultry production and the use of fluoroquinolones is close to zero.
The situation in Swedish broiler production is very good since the use of antibiotics is infrequent [34]. Consequently, the prevalence of resistant bacteria isolated from animals in Sweden is low [35,36].
FINRES-Vet [37] reports that the occurrence of antibiotic-resistant Campylobacter spp. from broilers has been at a low level. Compared to previous years, in 2020, the proportion of quinolone-resistant isolates dropped and the resistance to tetracycline, erythromycin, gentamicin or streptomycin remained low.
The annual NORM-VET 2020 report showed improvements of antimicrobial resistance in Norway [38]. As stated in this report, the prevalence of antimicrobial resistance among C. jejuni isolates from broilers is low; 90.8% of isolates were susceptible to all tested antimicrobials. Although the isolates were commonly resistant to quinolones and streptomycins, there were no multidrug-resistant isolates detected [38].
The rapid spread of antimicrobial resistance has been identified across the world and it is associated with the use of antimicrobials [39]. According to the European Food Safety Authority and the European Centre for Disease Prevention and Control (EFSA and ECDC) [40], in 2019, ciprofloxacin resistance in human Campylobacter isolates was high to extremely high (at the EU level it was 61.5% and 61.2% for C. jejuni and C. coli, respectively). The resistance to erythromycin was low (1.5% and 12.9% for C. jejuni and C. coli, respectively). C. coli erythromycin resistance was extremely high in Portugal (73.1%). The tetracycline resistance proportions were 47.2% and 66.9% for C. jejuni and C. coli, respectively [40]. According to EFSA and ECDC [40], the resistance to gentamycin in 2019 was low. In China, the prevalence of resistance in Campylobacter from human patients to ciprofloxacin, tetracycline and nalidixic acid is very high (89.7%, 74.6% and 69.0%, respectively), due to the extensive use of these antimicrobials without prescription [41]. In Ireland, compared to the early 2000s, tetracycline resistance among Campylobacter spp. in broilers has risen by approximately 10% [42]. In Portugal and Spain, Campylobacter spp. resistance to tetracycline in broilers is high: between 90 and 100% [43,44]. In a study conducted in Lithuania, Aksomaitiene et al. [23] found that C. jejuni isolates from human clinical cases were most frequently resistant to ciprofloxacin (88.1%), but all human isolates were sensitive to gentamicin and erythromycin. In our study, all the Estonian-origin broiler chicken meat Campylobacter isolates (n = 4) were sensitive to all of the studied antimicrobials. The small number of isolates was related to the very low Campylobacter prevalence (1.8% from 163 samples) in Estonian-origin broiler chicken meat [25]. The most frequently observed resistance (86.7%) of human strains was against ciprofloxacin and nalidixic acid. This high antimicrobial resistance among human strains probably indicates that ciprofloxacin and nalidixic acid would not be suitable for human Campylobacter infection treatment. In Estonia, the first choice of antibiotic treatment for human patients with severe Campylobacter infection is azithromycin, followed by ciprofloxacin as the alternative choice. In the present study, the proportions of Campylobacter isolates from fresh broiler chicken meat that were resistant to ciprofloxacin and erythromycin, all of Latvian and Lithuanian origin, were 91.3% and 8.7%, respectively. According to the EFSA and ECDC [40], in 2019, the highest levels of resistance in broiler meat were for nalidixic acid and ciprofloxacin (64-90%), and also for tetracycline (43-53%). A previous Estonian study by Mäesaar et al. [45] found similarly high proportions of fluoroquinolone resistance among Latvian (87.5%) and Lithuanian (84.8%) Campylobacter isolates originating from broiler chicken meat. In the present study, fluoroquinolone resistance of Latvian and Lithuanian isolates originating from broiler chicken meat was 100%, which probably reflects the wide use of these antibiotics in poultry production in these countries. The use of synthetic fluoroquinolone (enrofloxacin) for the treatment of respiratory and gastrointestinal infections in poultry has been shown to induce fluoroquinolone resistance in Campylobacter spp. [46]. Similarly to the results of the present study, Aksomaitiene et al. [23] found that all C. jejuni isolates from broiler products from Lithuanian retail settings were resistant to ciprofloxacin. Meistere et al. [24] reported that Latvia has one of the highest proportions of fluoroquinolone resistance among C. jejuni (97.5%) in broilers. Furthermore, Kovalenko et al. [47] found a high proportion of Campylobacter isolates from Latvian broilers resistant to fluoroquinolones (100%), ciprofloxacin (100%), nalidixic acid (87.9%) and streptomycin (39.6%). In the present study, fluoroquinolone resistance among human isolates was 86.7%, and 91.3% in broiler meat. Mäesaar et al. [45] found resistance to fluoroquinolones to be higher for humans (67.9%) than for broilers (60.2%). Multidrug-resistant strains were co-resistant to nalidixic acid and ciprofloxacin. Several studies in Canada and the USA have reported Campylobacter spp. ciprofloxacin resistance in up to 47% of Campylobacter strains [18,48,49]. In addition to high fluoroquinolone resistance among broiler chicken meat isolates, the present study observed high tetracycline resistance among Lithuanian broiler chicken meat isolates (76.9%) and high streptomycin resistance among Latvian and Lithuanian broiler chicken meat isolates: 68.8% and 42.3%, respectively. Tetracycline resistance among human isolates was 80.0%, which matched with the tetracycline resistance found among Lithuanian broiler chicken meat isolates (76.9%). In Table 4, the phenotypic resistance pattern and related genotypic mechanisms (gene, point mutation) are shown. All phenotypic resistance found for aminoglycosides, macrolides, quinolones and tetracyclines determined with the MIC test had corresponding genotypic antibiotic resistance mechanisms. In previous studies, aad9, aadE, aadE-Cc and aph(3 )-IIIa resistance genes [50][51][52][53] associated with aminoglycoside resistance were detected from all isolates with corresponding phenotypic resistance. For tetracycline resistance only, the tetO [54] gene was detected from isolates showing phenotypic resistance to tetracycline in the MIC test. The latter has also been found in several other studies [54]. Two point mutations in 23S (A2075G) and gyrA (T86I) genes associated with erythromycin and quinolone resistance [52,55] in campylobacters were also found in our study. In addition, a previous study conducted in Estonia found gyrA (T86I) mutation in quinoloneresistant C. jejuni ST5 isolates [56]. Additionally, 50S ribosomal protein L22 modification (A103V) [31] was detected in 14 isolates with no corresponding phenotypic erythromycin (macrolide) resistance found. The majority of isolates with matching geno-and phenotypic resistance had high MIC, and often exceeded the maximum concentration ranges.
High Campylobacter resistance in chicken meat can be a key risk factor for the treatment of severe human campylobacteriosis cases in Estonia. The high proportions of resistance and similar antimicrobial pheno-and genotypes found from imported broiler chicken meat products and for Estonian human clinical isolates indicate that the consumption of imported broiler chicken meat might pose the risk of Campylobacter to the Estonian population.
The application of a vertically integrated management system and strict biosecurity and biosafety measures at all levels of broiler chicken production may be the reason for the very low Campylobacter prevalence and counts as well as low antimicrobial resistance among Campylobacter strains isolated from the Estonian-origin fresh broiler chicken meat.

Conclusions
Among the Campylobacter strains isolated from broiler meat in 2018-2019, a total of 90.2% were resistant to one or more kind of antibiotics. Multidrug resistance was found in 27.9% of isolates. Campylobacter isolates from Estonian fresh chicken meat were sensitive to all of the tested antibiotics. Isolates of Latvian and Lithuanian origin were 100% resistant to one or more of the antibiotics, and 86.7% of the Estonian human strains were resistant to one or more of the antibiotics. There was high antibiotic resistance in Campylobacter spp. in Lithuanian and Latvian isolates from fresh broiler chicken meat in the Estonian retail market. There was also a high antibiotic resistance in Campylobacter spp. of human origin. This suggests that broiler chicken meat poses a potential risk to humans as it is well known that broiler chicken meat is a main source of human campylobacteriosis. To minimize the emergence of Campylobacter resistance, it is crucially important to follow common policies and implement good practices on antimicrobial usage at the farm level. Resistant bacteria in the food production chain can easily reach the consumer and pose a serious risk to public health.  Institutional Review Board Statement: Not applicable.