Genomic Characterization of Antibiotic-Resistant Campylobacterales Isolated from Chilean Poultry Meat

Due to the lack of knowledge about Campylobacterales in the Chilean poultry industry, the objective of this research was to know the prevalence, resistance, and genotypes of Campylobacter, Arcobacter and Helicobacter in 382 samples of chicken meat purchased in Valdivia, Chile. The samples were analyzed using three isolation protocols. Resistance to four antibiotics was evaluated by phenotypic methods. Genomic analyses were performed on selected resistant strains to detect resistance determinants and their genotypes. A total of 59.2% of the samples were positive. Arcobacter butzleri (37.4%) was the most prevalent species, followed by Campylobacter jejuni (19.6%), C. coli (11.3%), A. cryaerophilus (3.7%) and A. skirrowii (1.3%). Helicobacter pullorum (14%) was detected by PCR in a subset of samples. Campylobacter jejuni was resistant to ciprofloxacin (37.3%) and tetracycline (20%), while C. coli and A. butzleri were resistant to ciprofloxacin (55.8% and 2.8%), erythromycin (16.3% and 0.7%) and tetracycline (4.7% and 2.8%), respectively. Molecular determinants were consistent with phenotypic resistance. The genotypes of C. jejuni (CC-21, CC-48, CC-49, CC-257, CC-353, CC-443, CC-446 and CC-658) and C. coli (CC-828) coincided with genotypes of Chilean clinical strains. These findings suggest that besides C. jejuni and C. coli, chicken meat could play a role in the transmission of other pathogenic and antibiotic-resistant Campylobacterales.


Introduction
The order Campylobacterales is widely known because of its pathogenic genera Campylobacter, Helicobacter, and to a lesser extent Arcobacter. However, other members are known to play ecologically important roles in diverse niches and environments [1].
Campylobacter jejuni and C. coli produce most of the human bacterial gastroenteritis cases worldwide, and C. jejuni is also associated with post-infection sequelae such as Guillain-Barre syndrome [2][3][4]. Meanwhile, A. butzleri, A. cryaerophilus and A. skirrowii have been associated with gastrointestinal and systemic diseases [5,6]. Regarding the genus Helicobacter, H. pylori is the most relevant species due to its association with gastritis, peptic ulcers, and gastric cancer [7]. However, in recent years entero-hepatic Helicobacter (EHH) species have become increasingly important as emerging pathogens and potential zoonotic agents due to their link to intestinal and hepatobiliary diseases [8].
Campylobacter jejuni/coli reservoirs include a wide variety of animals, such as farm animals, pets, and wildlife [2,9]. Zoonotic transmission can occur through the consumption of food of animal origin or by drinking contaminated water, but poultry meat has been attributed as their main route of transmission [2,10,11]. In the case of Arcobacter and Helicobacter, their transmission routes are not entirely clear, but there is growing evidence that it is like that of Campylobacter spp. [8,12].
The human disease caused by Campylobacter is generally self-limited, and antimicrobial treatment is recommended only in severe, persistent, or recurrent infections and in immunocompromised patients [13]. Macrolides and fluoroquinolones are the first line of treatment for gastrointestinal infections, and tetracyclines and aminoglycosides for systemic infections, whereas in Arcobacter the use of tetracyclines has been suggested [14].
In recent years, an increase in antimicrobial resistance has been observed in clinically relevant Campylobacterales isolated from both animal reservoirs and human samples worldwide [15][16][17]. Consequently, the World Health Organization (WHO) has classified Campylobacter spp. and Helicobacter pylori as high-priority bacteria for the study of new antimicrobials due to the high levels of fluoroquinolones and clarithromycin resistance reported, respectively [16].
In Chile, a noteworthy increase in resistance to ciprofloxacin (CIP) and tetracycline (TET) has been observed in clinical strains of Campylobacter spp. over the past two decades [17,18]. Likewise, resistance to CIP, TET, and Erythromycin (E) has been reported in isolates obtained from chicken and bovine meat [3,19]. Globally, it has been suggested that this increase in antimicrobial resistance is mainly due to the indiscriminate use of antibiotics in poultry and livestock production, to which is added the ability of these microorganisms to survive in these conditions and, therefore, endure over time, although there are also other potential resistance sources [20][21][22].
Based on the above data and the fact that the transmission of Campylobacter in the ecosystem is multidirectional, this microorganism represents a One Health challenge [23]. Unfortunately, this approach is still relatively early in Chile compared to other South American countries such as Brazil, and the study of Campylobacter epidemiology is little considered; therefore, its detection in various sources is limited and focused on certain research groups [18,24].
So far, most studies on resistance in the poultry industry have focused on C. jejuni and/or C. coli. However, considering that different Campylobacterales reside in the intestines of birds, it is necessary to assess how selective pressures affect them as a group. Bearing this in mind, and the fact that there is a lack of knowledge about Campylobacterales present in the Chilean poultry industry, this study aims to reveal the prevalence, antimicrobial resistance, and genetic diversity of Campylobacterales present in retailed chicken meat samples in Chile.
As shown in Table 2, Arcobacter spp. was able to grow in all isolation protocols, while Campylobacter spp. could only be recovered in Bolton broth and mCCDA plates incubated at 37 • C under microaerobic conditions (Protocol B).

Antimicrobial Resistance
Susceptibility to antimicrobials was assessed phenotypically in 280 isolates (118 Campylobacter spp. and 162 Arcobacter spp.). Campylobacter jejuni showed resistance to CIP (37.3%) and TET (20%), with 5.3% strains resistant to both antibiotics. Campylobacter coli was resistant to CIP (55.8%), E (16.3%), and TET (4.7%), in which 16.3% strains presented simultaneous resistance to CIP-E and 2.3% strain to CIP-TET. All Campylobacter isolates were susceptible to Gentamicin (GEN) ( Table 3).  Arcobacter butzleri was the only species of the genus with antimicrobial resistance confirmed by MIC, with 2.8% to CIP, 0.7% to E, and 2.8% to TET. All Arcobacter isolates were susceptible to GEN (Table 3). However, there were discrepancies between the results of the disk diffusion and minimum inhibitory concentration methods, where the resistant isolates only coincided in 50%, 33.3% and 5.6% for CIP, E and TET, respectively.

Molecular Mechanisms of Antibiotic Resistance
Molecular antibiotic resistance determinants detected after whole-genome sequencing and functional annotation on selected samples are shown in Figure 1. All Campylobacter spp. strains resistant to CIP and TET had the C257T mutation in the quinolone resistancedetermining regions (QRDR) of gyrA gene and the tet(O) gene, respectively. Meanwhile, the six C. coli strains resistant to E, presented the A2074G (2/6) or A2075G (4/6) mutation. The erm(B) gene, however, was detected in none of the strains. All Campylobacter strains presented the cmeABC gene. Moreover, 7 of 19 C. jejuni strains whose genome was sequenced and all C. coli had mutations in the cmeR gene. Additionally, resistance determinants were found for certain antibiotics that were not phenotypically tested. Some C. jejuni possessed the lnu(C) gene (6/19)  All A. butzleri strains that had resistance phenotype to CIP had the C254T mutation in the QRDR of gyrA gene. Moreover, in one A. butzleri strain resistant to E, we found a mutation in the areR gene, suggesting an overexpression of the AreABC pump. Interestingly, despite observing phenotypic resistance to TET, no known specific molecular determinant was found for this antibiotic. However, most (7/10) strains had blaOXA-464.

Virulence Genes
Although it was not one of the main objectives of the study, virulence determinants were investigated using whole-genome sequences. It was found that the main genes in Campylobacter are cadF, jlpA, porA, pebA, racR, dnaJ, pldA, ciaB, ceuE, iamb, and flaC. The cdtABC toxins genes were found almost exclusively in C. jejuni, while the genes associated with T6SS were found in C. jejuni and C. coli. ( Figure S1). In A. butzleri, the cadF, cj1349, mviN, ciaB, pldA and irgA, genes were found in all the isolates, while only some strains All A. butzleri strains that had resistance phenotype to CIP had the C254T mutation in the QRDR of gyrA gene. Moreover, in one A. butzleri strain resistant to E, we found a mutation in the areR gene, suggesting an overexpression of the AreABC pump. Interestingly, despite observing phenotypic resistance to TET, no known specific molecular determinant was found for this antibiotic. However, most (7/10) strains had blaOXA-464.

Genotyping
As shown in Figure 1, the 19 C. jejuni genome sequences were classified into 16 different STs, four of which had not previously been reported (2 by novel allele sequences and 2 by novel combinations of preexisting alleles), which were submitted to the Campylobacter jejuni/coli PubMLST database. The remaining twelve STs were grouped into nine different clonal complexes (CC-21, CC-48, CC-49, CC-257, CC-353, CC-354, CC-443, CC-446 and CC-658). In addition, five different STs were identified out of 14 C. coli, with four of them grouped in CC-828, while ST-1109 could not be assigned to any CC.
In A. butzleri, eight different genotypes were identified. According to the analysis in the Arcobacter spp. PubMLST, six of which had not previously been reported (3 by novel allele sequences and 3 by novel combinations of preexisting alleles). The remaining two STs corresponded to the ST-40 and ST-172. However, new allele/new ST numbers could not be assigned to these Arcobacter genes/strains, respectively, due to the lack of a curator for this species in the database at the submission date (February 2023) (Keith Jolley, personal communication).

Discussion
Although the existence of Campylobacter and taxonomically related organisms in chicken meat samples is well known, most of the studies worldwide have been focusing only on C. jejuni and, to a lesser extent, on C. coli, with other Campylobacterales rarely considered, even when they coexist in the same reservoir and are subject to the same selective pressures, due to exposure to antibiotics in the poultry industry [25].
In the current study, more than half of the chicken meat samples for sale in the city of Valdivia, Chile, were positive for different Campylobacterales. In total, six species were detected: C. jejuni, C. coli, A. butzleri, A. cryaerophilus, A. skirrowii and Helicobacter pullorum. This contrasts with previous studies, where up to four species for this type of sample have been reported [15,26]. As far as we are aware, this is the first study demonstrating a higher diversity of Campylobacterales coexisting in samples of retail chicken meat.

Prevalence and Distribution of Campylobacterales
The prevalence of Campylobacter spp. in chicken meat samples was 29.8%, with C. jejuni present in 19.6% and C. coli in 11.3%, figures that fall within the ranges reported worldwide [25]. Unlike the large number of studies carried out on chicken meat in North America and the European Union (EU), studies carried out in South America are scarce, and as such unrepresentative. Moreover, they are concentrated in just a few countries, mainly Brazil, followed by Argentina and Chile [3,27,28].
Some studies on poultry meat in Chile show Campylobacter spp. has a similar prevalence to Salmonella spp. [3,29]. Despite this, unlike Salmonella the diagnosis of Campylobacter is not included in the Chilean sanitary regulations for food [18]. This may be because of their infrequent diagnosis in clinical laboratories, in contrast to what happens in highincome settings such as the United States, Iceland, the Netherlands and New Zealand, where the high prevalence obtained from both clinical samples and poultry meat led to recommendations to control Campylobacter in processing plants and imported products [18,30]. However, despite the scarcity of clinical laboratory diagnoses, a progressive increase in its detection as a cause of diarrhea was observed in Chilean children under five years of age between 2013 and 2017, as well as in adults with diarrhea [18,31].
Studies on emerging Campylobacterales are very limited in South America. Indeed, this is the first report of Helicobacter spp. in chicken meat in this setting. Within the emerging Campylobacterales, Arcobacter spp. presented a prevalence of 40.8%, where the most isolated species was A. butzleri (37.4%), predominating even over the genus Campylobacter spp. (Table 2) [19,32,33]. This could be due to the capacity of A. butzleri to grow in a wide range of temperatures, in different atmospheres and in aquatic environments. As such, it may be better able to survive in poultry industry facilities and be found in high concentrations in this type of food [32,34]. Arcobacter butzleri has been identified as an emerging foodborne pathogen and has been reported from Chilean patients with gastroenteritis [35,36]. However, like other emerging pathogens, there are no regulations on its detection in food, and its diagnosis in clinical samples is not done routinely [6,37].
It was not possible to isolate Helicobacter spp. This could be because Helicobacter spp. was present in low concentrations and/or in a viable but non-culturable state or because the protocol used was not ideal for isolating it from chicken meat, even though the media used have allowed the isolation of several EHH species in other biological samples [8,38,39]. However, a frequency of 14% for H. pullorum was obtained by molecular detection. This figure is close to the range reported in chicken meat in Iran (16-49%) [40,41]. There is only one previous record of this species from the bile of a patient with chronic cholecystitis in Chile [42]. As such, this is the first report of H. pullorum in Chilean chicken meat.

Antimicrobial Resistance and Virulence
There is a need to explain the increase in antibiotic resistance observed in clinical Campylobacter spp. strains over the last two decades in Chile [17,18]. While it has been attributed to the use of antimicrobials in the poultry industry, its cause is not known with certainty [18].
The results obtained show that the pattern of antibiotic resistance in Campylobacter spp. analyzed coincides with what has been reported in chicken meat and clinical strains worldwide [3,43]. According to phenotypic and genomic analyses, all Campylobacter strains resistant to CIP presented the C257T mutation in gyrA gene. This mutation is also associated with a gain in the fitness of Campylobacter, which would imply that strains with these mutations persist over time, even when the antibiotic is discontinued [44,45]. On the other hand, Campylobacter resistance to TET was similar to what had previously been reported for clinical strains, chicken meat and bovine liver in Chile [3,19]. All strains phenotypically resistant to TET possess the tet(O) gene in their genome [46]. Among the Campylobacter species, only C. coli presented resistance to E, which was associated with the presence of the A2074G and A2075G mutations in the 23S rRNA. These mutations generate a cost in the fitness of C. jejuni, but not in C. coli, which could explain the results obtained in this study [47,48]. The erm (B) gene was not detected in the genomes analyzed, though it has already been reported in strain from chicken skin samples from Peru and China [49,50].
All Campylobacter genomes contained cmeABC genes, which generate an efflux system or pump that contributes to Campylobacter intrinsic resistance to various antimicrobials, including CIP, E, and TET [51,52]. This pump has already been reported in Chile in Campylobacter spp. strains of bovine origin, as well as in clinical strains [19,53]. However, seven of the 19 C. jejuni genomes and all of C. coli (n = 14) presented mutations in the cmeR gene, which is associated with a greater expression of this multidrug efflux pump. Despite this, a difference in MICs was not observed between those that did and did not present the mutation [54]. To our knowledge, this is the first time that mutations in the cmeR gene is reported in Chilean strains.
Additionally, genomic analysis allowed us to detect antibiotic-resistance genes that were not phenotypically tested here, such as lnu(C), cfr(C) and blaOXA-61 [55,56]. Meanwhile, only the C. coli genomes presented molecular determinants of resistance to aminoglycosides (aph(3 )-III and aadE-Cc). Even though both are related to resistance to aminoglycosides, none of them confers resistance to gentamicin, in accordance with the phenotypic results obtained in this study [57].
Genomic analysis of resistance determinants has the potential to accurately predict resistance phenotypes [57]. As such, 27.3% of these Campylobacter spp. strains could be classified as multidrug-resistant (MDR), that is, strains with resistance to three or more classes of antibiotics, which coincides with the MDR reported in clinical samples in Chile [53]. This figure is lower than that obtained in other countries such as India (54.4%) and China (93.7%), in which a considerable increase in strains resistant to antibiotics has been observed over the years [58,59]. However, it is similar to other South American countries, like Brazil (13%) [60].
Similar phenotypic patterns of antimicrobial resistance reported in Campylobacter were observed for A. butzleri. The percentage of strains resistant to TET is within the range reported worldwide, while lower values were obtained for CIP and E [61]. A discrepancy was observed between the resistance results by DD and MIC in Arcobacter spp., mainly to TET. Since Arcobacter spp. does not have cut-off values in CLSI or EUCAST, most studies use the criteria and breakpoints listed for C. jejuni/C. coli in CLSI M45 [33,62]. However, the data obtained in this, and previous studies would suggest that Campylobacter cut-off values are not reliable for this microorganism and confirm the need for having own and standardized cut-off values for a trustworthy interpretation [63].
All the sequenced A. butzleri strains resistant to CIP carried the C254T mutation in the gyrA gene [63]. Additionally, all of them presented the adeF gene (with a low percentage of identity), associated with resistance to fluoroquinolones and tetracyclines [64]. A total of 70% of the A. butzleri genomes presented the previously reported blaOXA-464 gene, which is associated beta-lactams resistance [63,65]. Moreover, for the first time we report a mutation in the areR gene, involving an overexpression of the AreABC pump, which would correlate with the resistance to E observed in a strain, which could be supported with further gene expression analysis [66,67]. Meanwhile, the presence of the tet(O), tet(A) or tet(W), which generate resistance to TET, were not found in these strains. As we currently have limited knowledge of molecular determinants of resistance in Arcobacter, the presence of an unknown gene conferring phenotypic resistance cannot be ruled out, which highlights the need to investigate them, more so given that this is a naturally transformable bacterium, and therefore, could acquire an antibiotic resistance gene by horizontal transfer [68].
Even though the same antimicrobials used in human medicine are not used in the poultry industry [18,69], those applied in avian production belong to the same families, as in the case of enrofloxacin, tylosin, and oxytetracycline, which are antibiotics belonging to the fluoroquinolones, macrolides and tetracyclines families, respectively, and authorized for use in poultry according to the Chilean registry of veterinary medicines [69]. It is important to know about the local use of these antibiotics as it has previously been shown that the use of enrofloxacin for the treatment of broiler chickens, generated resistance to CIP in 100% C. jejuni, obtaining a MIC ≥ 32 µg/mL, which persists after the end of treatment [70]. In the case of tylosin, its use in therapeutic concentrations has generated resistance to E in 33.3% of C. coli and 7.9% of C. jejuni, a situation that in some cases generates highly resistant mutants (MIC > 512 µg/mL) [71,72]. Significantly higher resistance to TET, meanwhile, was seen when broilers were treated with oxytetracycline, and this increased resistance to TET has even been reported in Salmonella spp. isolated from chicken meat in Chile, with a range of 95.4-100% resistance to this antibiotic [29,73,74].
On the other hand, although it was not one of the main objectives of the study, virulence determinants were also detected. In Campylobacter spp., the virulence genes found are consistent with those previously reported in Chile in different types of samples, where those associated with adhesion and invasion predominate. It was also found that C. jejuni presents a greater number of virulent genes than C. coli [3,53]. We have also reported here on virulence genes detected in the genomes of A. butzleri, which is an emerging pathogen that presents several virulence determinants homologous to genes present in Campylobacter spp. (Figure S2).

Epidemiology
At the time of writing, the genotypes of only 145 Chilean strains of Campylobacter spp. were available in the PubMLST database, which were obtained mainly from clinical samples [17,53]. This is around 200 and 300 times less in comparison with the USA and the United Kingdom, respectively, where the epidemiology of Campylobacter spp. has been extensively investigated. However, it is similar to the amount of data from South American countries such as Brazil and Ecuador. This is the first Chilean study reporting Campylobacter MLST genotypes of chicken meat strains. The C. jejuni strains were grouped into clonal complexes (CC) previously reported in clinical strains isolated in Valdivia (CC-21, CC-48, CC-257 and CC-353) and Santiago (CC-21, CC-48, CC-49, CC-257, CC-353, CC-443, CC-446 and CC-658) [17,53]. One poultry isolate corresponded to ST-3874 belonging to CC-354, which has been previously reported in a bovine liver isolate in the same geographical area (Valdivia city) and has also been isolated from clinical samples in the USA and Canada [19,75]. A total of 21% of the C. jejuni strains (4/19) clustered in CC-353, being the most prevalent CC, a genotype known to cause human infections and colonize broiler products, as well as its association with resistance to quinolones [76][77][78].
Campylobacter coli presents a particular genetic population structure, in which at least three main clades have been described to date [79]. In this study, all the C. coli strains belonged to clade 1 ( Figure S3), that are commonly isolated from farm animals and human gastroenteritis cases, unlike clade 2 and 3, which are mainly isolated from environmental sources [53]. Most of the isolates clustered in CC-828 (92.9%), in line with previously reported genotypes of clinical strains in Chile and the worldwide distribution of C. coli in birds and patients with diarrhea [53,80]. The only ST of C. coli that did not belong to any CC was the ST-1109, which has previously been isolated from animals and humans. These associations suggest that domestic broiler meat is likely an important source of antibiotic-resistant Campylobacter in Chile. Additionally, 21% (4/19) of the C. jejuni genomes did not match with previously reported STs, which correspond to four new STs (12300, 12301, 12302, 12303) found in this study.
The genotypes obtained in A. butzleri were only identified at ST level (genotypes did not cluster into clonal complexes), and corresponded to ST-40 and ST-172, which have only been described in samples associated with poultry (chicken, turkey, and poultry environment) in Nigeria, Thailand, and the USA. A total of 70% (7/10) of the genomes analyzed did not coincide with previously reported STs, corresponding to six potential new STs, which cannot be assigned yet due to the lack of a curator in the PubMLST database, as mentioned above. Due to the few studies carried out on this emerging pathogen and the late development of their MLST protocol compared to Campylobacter spp., knowledge of the epidemiology of A. butzleri is even more limited [81]. In fact, the data reported here correspond to the first MLST genotypes of A. butzleri in South America, which could provide a basis for understanding its distribution when data on clinical strains in the area are available.

Sample Collection
A total of 382 broiler meat packages were obtained from 14 supermarkets in the city of Valdivia (southern Chile), between June 2021 and January 2022. The meat came from the three main supply chains in Chile. This study was conducted by simple random sampling. The sample size was calculated using Working in Epidemiology "http://www.winepi.net/ (accessed on 10 March 2021)", based on a seroprevalence of 46%, an error of 5%, and a 95% confidence interval [3]. The samples were transported refrigerated in an airtight bag to the laboratory, where they were analyzed within six hours.

Isolation of Campylobacterales
Three protocols were used for the isolation of different Campylobacterales genera. Protocol A, aimed at the isolation of Arcobacter, consisted of enriching 10 g of meat in 90 mL of Arcobacter broth, supplemented with CAT (cefoperazone, amphotericin B and teicoplanin), homogenized for 1 min in a Stomacher bag and then incubated at 30 • C for 48 h in aerobic conditions. Then, 200 µL of broth was transferred to the surface of a Millipore membrane filter (diameter pore 0.45 µm) for passive filtration on Columbia Agar supplemented with 5% sheep blood, and allowed to filter for 30 min [82]. Finally, the filter was removed, and the plates were incubated at 30 • C for 48 h under aerobic conditions. Plates that were negative at 48 h were incubated for up to 5 days.
Protocol B, aimed at the isolation of Campylobacter, consisted of the enrichment of 10 g of meat in Bolton broth, supplemented with 5% sheep blood, homogenized for 1 min in a Stomacher bag and then incubated for 48 h at 37 • C under microaerobic condition using Anaerocult © C (Merck Millipore). Next, 50 µL of this broth was streaked in mCCDA, which was incubated for 48 h at 37 • C under microaerobic conditions. Plates that were negative at 48 h were incubated for up to 5 days.
Protocol C1 and C2, aimed at the isolation of Helicobacter, consisted of adding 10 g of the sample to 90 mL of PBS 1X, and homogenizing it for 1 min in a Stomacher bag. This was then centrifuged, discarding the supernatant and resuspending the pellet of which 400 µL was used to carry out passive filtration as described above (Protocol C1). Another 50 µL was streaked in Columbia agar supplemented with 5% sheep blood and CAT antibiotic supplement. Both media were incubated under microaerobic conditions at 37 • C for 48 h (Protocol C2). Cultures that were negative at 48 h were incubated up to 7 days.
Suspicious colonies, with Gram characteristics of Campylobacterales, were cultured on Columbia agar supplemented with 5% blood sheep to obtain a pure culture. They were stored at −80 • C in BHI broth with 20% glycerol.

Identification
DNA extraction was performed from fresh bacterial culture using the boiling method [83]. Colonies obtained by protocols B and C were identified using multiplex PCR (mPCR) for Campylobacter [84]. Samples that resulted negative for the Campylobacter genus and isolates obtained from protocol A underwent PCR for Arcobacter genus, with the subsequent mPCR analysis for Arcobacter species [85,86].

Antimicrobial Susceptibility Testing
Antimicrobial susceptibility to CIP, E, TET, and GEN was assessed. The qualitative disk diffusion method was performed, for which the breakpoints of the Clinical and Laboratory Standards Institute (CLSI) M45 (2016) were used [87]. Staphylococcus aureus ATCC 25923 was used as a quality control strain. For GEN interpretation, the CLSI M100 (2020) guidelines for Enterobacterales were followed [88]. The minimal inhibitory concentration (MIC) was then determined using MIC Test Strip (Liofilchem) in the resistant isolates or in those with intermediate resistance.

Whole-Genome Sequencing
DNA of a set of representative antibiotic-resistant-strains (19 C. jejuni, 14 C. coli and 10 A. butzleri) were extracted using an Easy-DNA TM gDNA Purification Kit (Invitrogen), and quantified by spectrophotometry (NanoQuant-Infinite M200, Tecan) and fluorometry (Qubit 3.0 fluorometer). The sample libraries were prepared using the Illumina DNA Prep kit and IDT 10 bp UDI indices, sequenced on an Illumina NextSeq 2000 at the Microbial Genomics Sequencing Center (SeqCenter, Pittsburgh, PA, USA). Sequence files were evaluated using FastQC v.0.11.9 before and after trimming [89]. Reads were trimmed using Trimmomatic v.0.39 to discard sequences with per base sequence quality score < 28 [90]. Two assemblers (SPAdes v3.13.1 and Unicycler v0.4.8) were used with default parameters and without reference-guided options. The assessment of draft genome assembly quality was done using the 3C criterion (contiguity, completeness, and correctness; Table S1) [91][92][93]. Assembly sequences were kept at the contig level with a minimum size of 1000 bp. The draft genomes were investigated for the presence of genes of antimicrobial resistance, which were identified using Comprehensive Antibiotic Resistance Database (CARD) and ResFinder [94]. Additionally, the sequences of the QRDR-gyrA gene from ciprofloxacinresistant mutant strains in A. butzleri were analyzed in comparison with parental strains to identify point mutations. The detection of virulence determinants was carried out using the Virulence Factor Database (VFDB) and a local BLAST alignment (identified with > 90% identity and > 60% coverage). Genome assemblies were also mapped to PubMLST [95].

Molecular Detection of Helicobacter Pullorum
A subset of 50 samples, processed by protocol C, underwent DNA extraction using Cells and Tissue DNA Isolation Kit (Norgen Biotek, Thorold, ON, Canada), which were subjected to a specific PCR targeting to the cdtB gene of Helicobacter pullorum [96]. This was confirmed by Sanger sequencing and BLASTn analysis.

Phylogenomic Tree
The genome sequences from this study were uploaded to the Type (Strain) Genome Server (TYGS) (https://tygs.dsmz.de), for a whole genome-based taxonomic analysis. TYGS used the Genome-BLAST Distance Phylogeny method (GBDP) to compare wholegenome sequences. The phylogenomic tree was viewed and edited using iTOL v.4 [97].

Statistical Analysis
Prevalence among all Campylobacterales isolates in this study was assessed with Cochran's Q test, while McNemar's tests was used to determine any differences between the prevalence of Campylobacter and Arcobacter species. In both tests a p-value of <0.05 was considered significant. The analysis was performed using the statistical software R v4.1.1.

Conclusions
To our knowledge, this is the first study where three genera of the order Campylobacterales have been studied in parallel in samples of retail chicken meat using phenotypic and genomic approaches, showing a high prevalence and diversity of bacterial species along with high antibiotic resistance and virulence potential. In fact, emerging species such as Arcobacter spp. were identified with the highest prevalence. In addition, Helicobacter pullorum was reported for the first time in chicken meat in a South American country. Additionally, the concordance between the MLST genotypes of previously reported clinical strains of Campylobacter with those from chicken meat, as well as the presence of the same pattern of antibiotic resistance, leads us to suggest that poultry production in Chile bears part of the responsibility for the increase in antibiotic resistance observed in human campylobacteriosis. Therefore, epidemiological surveillance of Campylobacterales should be a priority for the Chilean food industry, and the governmental institutions should implement specific control measures to assess the contamination in chicken meat of national origin.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/antibiotics12050917/s1, Table S1. Comparison of contiguity and annotation of genomes assemblies by different approaches.; Figure S1. Phylogenetic tree based on whole-genome sequencing generated with Genome Blast Distance Phylogeny (GBDP). The presence or lack of virulence genes in antibiotic-resistant Campylobacter spp. is indicated by the presence or absence of red blocks, respectively. The new STs are in bold type. Bar, 0.01.; Figure S2. Phylogenetic tree based on whole-genome sequencing generated with Genome Blast Distance Phylogeny (GBDP). The presence or lack of virulence of antibiotic-resistant Arcobacter butzleri is indicated by the presence or absence of red block, respectively. NST: New Sequence Type. NST are in bold type. Bar, 0.01.; Figure S3. Phylogenetic tree based on whole-genome sequencing generated with Genome Blast Distance Phylogeny (GBDP). This shows the clade to which the C. coli strains isolated in this study belong. RM2228 belongs to clade 1, H055260513 to clade 2, and BIGS0008 to clade 3. Bar, 0.01.

Data Availability Statement:
The datasets presented in this study can be found in NCBI under Bioproject PRJNA932782.