Prevalence, Virulence, and Antimicrobial Resistance of Campylobacter spp. in Raw Milk, Beef, and Pork Meat in Northern Poland

The purpose of this study was to determine whether raw milk, unpasteurized dairy products, pork, and beef available for sale in the Kujawsko-Pomorskie and Wielkopolska regions in Poland are contaminated with Campylobacter spp. bacteria and may be a potential source of infection. For isolated strains, antibiotic susceptibility and the presence of genes responsible for virulence were examined. Material for research included 1058 food samples collected between 2014 and 2018 with 454 samples of raw milk and unpasteurized dairy products (milk from vending machines, milk from owners of dairy cows, cheese, milk cream) and 604 samples of raw meat (pork, beef). The results indicated that 9.3% of the samples were positive for Campylobacter spp., and Campylobacter jejuni was predominant in this study. Campylobacter bacteria was not found in milk collected from vending machines, as well as cheese and milk cream samples. Campylobacter was noted in 12.7% of beef samples, 11.8% of raw milk purchased from individual suppliers, and 10.9% of pork samples. Resistance to erythromycin (2.0%), azithromycin (3.1%), gentamicin (4.1%), tetracycline (65.3%), and ciprofloxacin (71.4%) was determined using the disc diffusion method. Furthermore, the prevalence of racR, sodB, csrA, virB11, cdtB, iam, and wlaN genes were examined using the PCR method. The sodB, csrA, and cdtB genes exhibited the highest detection rate, but none of the genes were identified in 100% of the isolates. Statistically significant differences between the presence of virulence marker genes, including for iam, racR, and csrA markers, were noted among different sources of the isolates. Differences in the distribution of iam, wlaN, and virB11 were also shown between C. jejuni and C. coli strains. As a result of the analysis, it has been concluded that unpasteurized milk, beef, and pork could be a sources of Campylobacter pathogens. Moreover, this study revealed virulent properties of Campylobacter isolated from such food products and high resistance rates to fluoroquinolones, which may represent difficulties in campylobacteriosis treatment.


Introduction
Increased interest in the Campylobacter bacteria has been observed for more than 30 years [1][2][3]. Campylobacteriosis is an infectious disease caused in humans by the bacteria belonging to the Campylobacter genus, especially C. jejuni and C. coli, and is the primary cause of human bacterial gastroenteritis worldwide [4]. Campylobacter organisms are currently the most common factor of foodborne infections caused by food intake, especially foods of animal origin [5,6]. Data from the European food safety authority (EFSA) show that approximately 200 thousand people per year in the countries of the European Union suffer from food infections caused by Campylobacter [7]. the invasion-associated marker virB11. WlaN was selected because it is involved in the expression of ganglioside mimics in Guillain-Barré syndrome. The sodB and csrA genes provide protection against oxidative stress [30,34].
The purpose of this study was to determine whether raw milk, unpasteurized dairy products, pork, and beef available for sale on the market in two regions in Northern Poland are contaminated with Campylobacter spp. and may be a potential source of infection. We also compared the occurrence of Campylobacter spp. pathogenic genes responsible for encoding virulence factors and antibiotic susceptibility in Campylobacter jejuni and Campylobacter coli positive samples isolated from different origins.

Sample Collection
A total of 1058 food samples of animal origin were collected in Northern Poland over a four-year period (2014)(2015)(2016)(2017)(2018). Materials for the study include 454 samples of raw milk and unpasteurized dairy products (cheese, cream) originating from Kujawsko-Pomorskie and Wielkopolska regions. Samples were purchased year-round directly from the owners of the dairy cows (n = 221), automatic vending machines (n = 113), and local markets (n = 120). We examined 604 samples of raw meat samples that were sold unpackaged (beef n = 347 and pork n = 257). The samples were purchased year-round from supermarkets and butcher shops located in the study area to provide the element of representativeness. The number of samples collected during each year of the study was as followed:  Table 1 presents all samples collected and tested over the study period. The samples were transported to the laboratory in cooler boxes on ice and analyzed immediately. Campylobacter spp. is the invasion-associated marker virB11. WlaN was selected because it is involved in the expression of ganglioside mimics in Guillain-Barré syndrome. The sodB and csrA genes provide protection against oxidative stress [30,34]. The purpose of this study was to determine whether raw milk, unpasteurized dairy products, pork, and beef available for sale on the market in two regions in Northern Poland are contaminated with Campylobacter spp. and may be a potential source of infection. We also compared the occurrence of Campylobacter spp. pathogenic genes responsible for encoding virulence factors and antibiotic susceptibility in Campylobacter jejuni and Campylobacter coli positive samples isolated from different origins.

Sample Collection
A total of 1058 food samples of animal origin were collected in Northern Poland over a four-year period (2014-2018). Materials for the study include 454 samples of raw milk and unpasteurized dairy products (cheese, cream) originating from Kujawsko-Pomorskie and Wielkopolska regions. Samples were purchased year-round directly from the owners of the dairy cows (n = 221), automatic vending machines (n = 113), and local markets (n = 120). We examined 604 samples of raw meat samples that were sold unpackaged (beef n = 347 and pork n = 257). The samples were purchased year-round from supermarkets and butcher shops located in the study area to provide the element of representativeness. The number of samples collected during each year of the study was as followed:  Table 1 presents all samples collected and tested over the study period. The samples were transported to the laboratory in cooler boxes on ice and analyzed immediately.

Isolation of Campylobacter
Isolation of Campylobacter spp. from meat and milk was conducted in compliance with the EN ISO 10272-1:2006 method with slight modification. Briefly, meat and cheese (25 g) were homogenized for 2 min in a stomacher with 225 mL buffered peptone water (Oxoid Limited, Basingstoke, United

Isolation of Campylobacter
Isolation of Campylobacter spp. from meat and milk was conducted in compliance with the EN ISO 10272-1:2006 method with slight modification. Briefly, meat and cheese (25 g) were homogenized for 2 min in a stomacher with 225 mL buffered peptone water (Oxoid Limited, Basingstoke, United Kingdom) in the sterile plastic bags. Then, 10 mL of the homogenate solution and 10 ml of raw milk was added to 90 mL of Bolton broth containing the Bolton broth selective supplement and 5% laked horse blood (Oxoid Limited, Basingstoke, United Kingdom) and incubated at 42 • C for 48 h under microaerobic conditions (Generbox microaer-BioMerieux, Marcy l'Etoile, France). Next, the bacterial suspension was spread onto CCDA plates (Oxoid Limited, Basingstoke, United Kingdom) and then incubated at 42 • C for 48 h under microaerobic conditions. Characteristic growth from the CCDA plates was transferred to a blood plate (Columbia agar containing 5% cattle blood, Oxoid Limited, Basingstoke, United Kingdom) and incubated overnight at 42 • C. Colonies suspected as Campylobacter spp. were examined for cell morphology, motility, and oxidase reactions. Putative Campylobacter colonies were frozen at −80 in Microbanks (Pro-Lab Diagnostics, Birkenhead, United Kingdom) until species differentiation using polymerase chain reaction (PCR).

Species Identification
The PCR method with specific primers, as previously described, was used for the purpose of identification of colonies as C. jejuni or C. coli [35,36]. Bacterial DNA lysates were prepared from fresh Campylobacter cultures using the boiling method, as previously described [37]. PCR was performed in a 25 mL volume containing 2. . PCR was performed using the cycling conditions specified by the original authors [35,36]. The amplified DNAs were analyzed by electrophoresis in a 1.5% agarose gel. Reference strains of C. jejuni (NCTC11322) and C. coli (NCTC11366) were used as a control strain.

Prevalence of Virulence Genes
The presence of racR, sodB, csrA, virB11, cdtB, iam, and wlaN genes was determined using the PCR method with the primers and cycling conditions, as previously described by Linton

Antimicrobial Resistance
Susceptibility to five antimicrobial agents was assessed by the disk diffusion method according to the clinical laboratory and standard institute [41] using Mueller-Hinton medium supplemented with 5% defibrinated horse blood (Oxoid, Basingstoke, United Kingdom). The antimicrobials tested included ciprofloxacin 5 µg, erythromycin 15 µg, gentamicin 10 µg, tetracycline 30 µg, and azithromycin 15 µg (Oxoid, Basingstoke, United Kingdom). In the cases when CLSI recommendations were not available for Campylobacters, CLSI guidelines for Enterobacteriaceae were followed. The strains that showed resistance to three or more classes of antimicrobial agents were considered multidrug-resistant (MDR) strains. Reference strains of C. jejuni (NCTC11322) and C. coli (NCTC11366) were used as a control strain.

Statistical Analysis
Statistical analysis was performed using the Statistica 10.0 program (StatSoft, Cracow, Poland, 2011). Statistical differences in the prevalence of virulence genes and antimicrobial resistance between individual samples and between C. jejuni and C. coli strains were analyzed using the Chi square test. Holm-Bonferroni correction was applied for multiple comparisons. For small sample sizes, Yates' correction was also used. A significance level of p = 0.05 was accepted. p-values of < 0.05 were considered statistically significant.

Prevalence of Campylobacter spp. in Milk, Pork, and Beef
During our 4-year study, 1058 samples from dairy products, fresh pork, and beef meat were analyzed ( Table 1). The results have indicated the presence of Campylobacter spp. in 98 (9.3%) of the samples. Frequency of C. jejuni in the examined samples was 76.5%, C. coli was found in 23.5% of the analyzed samples. The highest prevalence of Campylobacter spp. (12.7%) was noted in beef samples followed by raw milk from farmers (11.8%). Sample contamination with Campylobacter bacteria was the lowest in the pork meat group (10.9%). The occurrence of Campylobacter in the samples obtained from pork chops was slightly higher than in minced pork samples. Not all types of food exhibited Campylobacter spp. contamination. Raw, unpasteurized milk obtained from vending machines, as well as cheese and milk cream purchased from retail stores, were not contaminated with Campylobacter bacteria. Campylobacter isolates recovered from milk exclusively included C. jejuni, while both species were detected at the same frequency (50%) in pork meat products. The Chi square test revealed that C. jejuni isolates were significantly more frequently isolated than C. coli isolates in beef meat samples (p < 0.5).

Prevalence of Virulence Genes
Overall, the occurrence of virulence markers varied among different sources of the isolates, and significant differences were found for genes associated with invasiveness (iam), adherence, colonization (racR), and stress response (csrA). Overall, the most common virulence markers were sodB (90.8%), cdtB (88.8%), and csrA (71.4%) genes, which are associated with stress response and toxin production ( Table 2).
High levels of C. jejuni virulence genes were noted in milk, with the exception of racR, wlaN, and virB11 genes, which were present at relatively low levels. Among all Campylobacter spp. isolates obtained from beef, genes responsible for toxin production, stress response, and colonization were the most common. The majority of genes were found at high levels in Campylobacter spp. isolated from pork meat (csrA, sodB, cdtB, and racR). Low levels of the pathogenic markers virB11 and iam were noted in Campylobacter isolates from pork and beef meat. Significant differences in the occurrence of iam, wlaN, and virB11 genes were detected between C. jejuni and C. coli isolates (Table 3). * Statistically significant differences between the presences of virulence marker genes among different sources of the isolates have been identified. The following differences were identified for the presence of iam from milk and beef meat (p < 0.0000), iam from milk and pork meat (p = 0.0014), racR from milk and beef meat (p = 0.0048), racR from milk and pork meat (p = 0.0118), csrA from milk and pork meat (p = 0.0207), and csrA from pork and beef meat (p = 0.0207).

Antimicrobial Resistance
The results for antimicrobial resistance in relation to sample origin (milk, pork, beef) and species (C. jejuni or C. coli) are shown in Tables 4 and 5. Overall, most of the strains were resistant to ciprofloxacin and tetracycline. Ciprofloxacin is one of the two antimicrobials regarded as critically important for treatment of Campylobacter infections in humans, and very high (71.4%) resistance levels were reported in all Campylobacter isolates. Resistance to tetracycline was confirmed in 65.3% of the strains. The lowest antimicrobial resistance rates were noted for erythromycin, azithromycin, and gentamicin (2.0%, 3.1%, and 4.1%, respectively).
Tetracycline resistance was the most common in Campylobacter stains isolated from milk (77.0%). Ciprofloxacin and gentamicin resistance rates were estimated as 65.4% and 3.8%, respectively. All isolates from milk were susceptible to azithromycin and erythromycin.
Resistance to ciprofloxacin was the most common in isolates from beef meat (63.4%) followed by tetracycline (59.1%). Campylobacter isolates obtained from beef (2.3%) exhibited the lowest antimicrobial resistance rates to erythromycin and azithromycin.
Among the identified C. coli strains from pork samples, no strains sensitive to ciprofloxacin were obtained. In total, 64.3% of strains obtained from pork were resistant to tetracycline. Resistance to gentamicin was detected in three Campylobacter pork isolates (10.7%), and resistance to azithromycin and erythromycin was estimated as 7.2% and 3.6%, respectively. Significant differences in Campylobacter ciprofloxacin resistance were noted between C. coli isolated from pork and beef samples. One strain of C. coli obtained from a pork sample presented simultaneous resistance to ciprofloxacin, erythromycin, and azithromycin. * Statistically significant differences between antimicrobial resistance to ciprofloxacin among C. coli isolates from beef and pork were detected (p = 0.0407).

Discussion
Most of the available studies are mainly concerned with the prevalence of Campylobacter in poultry as a main source of human campylobacteriosis. The number of studies discussing Campylobacter contamination in other meat types is limited in the literature. In our study, we emphasized that Campylobacter contamination in retail dairy and meat products other than poultry also raise concern, especially given the high resistance profile of milk, beef, and pork Campylobacter isolates.
In this study, 28 out of 257 (10.9%) of pork meat samples were positive for Campylobacter spp., which corroborates findings reported in Poland [15]. Other reports on the prevalence of  [42][43][44][45]. Pork and beef samples were negative for Campylobacter in a Canadian study [16]. These differences may be due to several variables, including geographical location, different sanitary conditions on farms, or slaughter practices. Most of the authors indicate that the main species isolated from pigs or retail pork products was C. coli, and a similar result is reported in our study [10,42].
The prevalence of Campylobacter spp. in retail beef meat products in our study is similar to previous research in Poland that reported Campylobacter contamination in 10.1% of fresh beef [15]. Similar levels of contamination were revealed in studies of Kashoma et al., 2016-9.5%, and in Pakistan, where the occurrence of Campylobacter was estimated in 15.5% of beef by Nisar et al., 2018 [10,46]. Campylobacter was present in 78% of beef livers in Iran [44]. Previous studies in Poland reported that 10% and 30% of bovine and pig carcasses, respectively, were positive for Campylobacter spp. [13].
Recent studies clearly indicate that the pork and beef may be contaminated with Campylobacter and constitute a potential source of campylobacteriosis infection in humans. To protect consumers, there is a need for greater realization of food safety programs "from the farm to the consumer", further risk assessment, and consumer education.
Campylobacter spp. was isolated from 11.8% of retail raw milk samples in this study, which is similar to levels obtained by Hussaina et al., 2007-10.2%, Kashoma et al., 2016-13.4%, and Raeisi et al., 2017-8.7% [10,12,47]. A lower rate of Campylobacter prevalence in raw milk samples was previously reported in Poland in the study by Wysok, Wiszniewska, Uradziński, and Szteyn, 2011 [22]. In our study, no bacteria was isolated from vending machines. Despite a relatively low prevalence of Campylobacter spp. in retail raw milk samples, many authors suggest that raw milk could represent the second most common source of human campylobacteriosis, after chicken meat [10,22,48]. Organic and raw food is becoming increasingly popular, given their natural health properties, but consumers should also be aware of the risk associated with consuming unpasteurized milk [49]. Consumption of raw milk is inherently risky because the milk has not been treated to inactivate pathogens.
Data on the occurrence of Campylobacter virulence markers in food samples are mainly related to the detection of virulence genes in Campylobacter isolated from poultry. The number of studies on the prevalence of genes related to virulence among isolates of Campylobacter spp. from milk, beef, and pork is limited. In this study, among the 98 Campylobacter strains, the most frequently detected genes were sodB, cdtB, and csrA genes. Research conducted by Modi, Brahmbhatt, Chatur, and Nayak (2015) showed that the cdtB gene (involved in toxin production) was detected in all C. jejuni isolates from raw milk samples [48]. Similarly, the increased prevalence of the cdtB gene was demonstrated in C. jejuni isolates obtained from pork liver [50]. The low prevalence of the cdtB gene in C. coli stains (5.9%) in the same study is noteworthy. The cdtB gene was noted in 60% of Campylobacter isolates from chicken, pork, and turkey meat samples in studies performed by Lapierre et al. [45].
The available literature provides a few works on the detection of sodB and csrA genes in milk, pork, or beef in Campylobacter isolates [27,51]. Most of the authors underline their crucial role in campylobacteriosis pathogenesis. SodB exhibited a high prevalence among poultry meat and human Campylobacter strains analyzed in the study of Wieczorek, Wołkowicz, and Osek, 2018 [51]. The csrA gene was present in 87.3% of C. jejuni isolates form broiler meat and in 97.7% from bovine in the studies of Gonzalez-Hein, Huaracán, García, and Figueroa, 2013 [27]. The authors suggested an important regulatory role of this gene in the stress responses in C. jejuni pathogenesis. Our study revealed significant differences in the distribution of csrA genes between different Campylobacter sources.
In the current study, racR, which was selected as pathogenic gene responsible for adherence and colonization, was statistically detected more often in pork or beef Campylobacter isolates than those recovered from milk. Similar to the results of this study, Bardon et al. detected racA at high frequencies in Campylobacter pork liver samples [50]. Greater than 90% of C. jejuni isolates from chicken harbor the racR gene in the studies of Datta. Niwa and Itoh, 2003 [40]. Moreover, Wysok and Wojtacka, 2018, reported a high prevalence of the racR gene in isolates obtained from cattle and swine [14].
The iam gene is a virulence marker examined in the present study that determines the invasiveness of Campylobacter spp. The prevalence of iam differed between both species and sources. The iam gene was present in 84.6%, 35.7%, and 25% of Campylobacter obtained from milk, pork, and beef, respectively. A high frequency of this invasion marker was previously detected in Campylobacter spp. from food samples, as reported by Rizal et al., 2010 [52]. On the other hand, Modi, Brahmbhatt, Chatur, and Nayak (2015) identified the iam gene in 14.3% of C. jejuni isolates from milk [48]. In the study by Bardon et al. (2017), the iam sequence was exclusively found in C. coli isolates from pork liver, which is consistent with the present study wherein iam was primarily detected in the C. coli group [50]. Given that the iam sequence appeared at different frequencies between different species and sources, it can be concluded that the role of this marker in the pathogenesis of Campylobacter spp. infections is unclear. Further studies need to be performed to explain the causes of such differentiation.
WlaN was selected as a pathogenic gene responsible for the ganglioside mimicking Guillain-Barré syndrome. A limited number of studies have reported wlaN gene detection in Campylobacter obtained from meat or milk. The occurrence of this gene in Campylobacter strains from pork, beef, and milk was estimated at approximately 50% in this study, and significant differences were between C. jejuni and C. coli strains (p < 0.05). Raeisi et al. (2017) described wlaN detection exclusively in C. coli isolates recovered from raw milk samples [12]. In the study by Lapierre et al. (2016), C. jejuni strains isolated from meat showed increased wlaN gene frequency compared with C. coli, which was also observed in C. jejuni isolates from pork in our study [45].
The least prevalent gene in our study was virB11 (20.4%). This gene was a statistically more often detected gene in C. coli isolates. The plasmid-associated virulence marker virB11 was not detected in isolates from milk in Raeisi et al.'s studies [12]. In Bardon et al.'s study (2017), the virB11 gene was only confirmed in C. coli isolates obtained from pork liver with a low detection rate-5.9% [50].
Resistance to antimicrobials in Campylobacter spp. isolated from beef and pork has not previously been examined in Poland. Data on antibiotic profiling showed that Campylobacter isolated from pig and cattle carcasses at slaughter in Poland were most frequently resistant to quinolones-57.1% and tetracycline-51.4% [53]. The prevalence of Campylobacter isolated from red meat being highly resistant to ciprofloxacin and tetracycline (88.8% and 66.7%, respectively) was confirmed by Raeisi et al. [12]. Lapierre et al. (2016) demonstrated that all C. coli strains from pork liver were resistant to ciprofloxacin and tetracycline, whereas C. jejuni isolated from bovine meat were sensitive to ciprofloxacin [45]. According to an EFSA report (2019) in 2017, more than half of the tested Campylobacter isolates from either pigs or cattle exhibited resistance to fluoroquinolones. Considering that the C. coli isolates from pig meat were tested for susceptibility in 2016, resistance to ciprofloxacin (76.5%) and tetracycline (85.3) was very high. Increasing trend of resistance to tetracycline in C. coli isolated from pig meat was observed in Portugal (66.7% in 2016-100% in 2017) [7].
Raw milk C. jejuni isolates in our study were also highly resistant to ciprofloxacin and tetracycline. Increased resistance of Campylobacter from raw milk to this antimicrobial agent (85.8%) was previously recorded in Poland [22]. Similar data regarding the isolation of Campylobacter species from milk that were highly resistant to ciprofloxacin and tetracycline (greater than 85%) were reported by Modi, Brahmbhatt, Chatur, and Nayak, 2015 [48]. In contrast, Kashoma et al. (2016) reported that only 9.3% and 11.8% (depending on the method) of Campylobacter isolated from milk samples and beef carcasses, respectively, exhibiting resistance to ciprofloxacin [10]. Similar rates for ciprofloxacin resistance were detected in the US [54]. An outbreak of fluoroquinolone-resistant C. jejuni infections associated with raw milk consumption in United States of America was recently described by Burakoff et al., 2018 [55].
Reports from all over the world recently noted emerging resistance to macrolides in clinical, environmental, animal, and food Campylobacter samples [7,56,57]. Campylobacter isolates in our study were mostly sensitive to the optimal antibiotics used for Campylobacter infection (especially in children), but we confirmed single resistance for macrolides and gentamicin, which was not previously described in Campylobacter isolated from pork, beef, or milk in Poland. Noteworthy is the increasing trend in gentamicin resistance in C. jejuni and C. coli isolates from broilers, cattle, pigs, and their products, reported by EFSA in 2017 (percentages for gentamicin resistance in C. jejuni from cattle and C. coli from pigs were higher than 65% in Croatia in 2017) [7].
The isolation of Campylobacter bacteria resistant to azithromycin in raw milk was previously described by Kashoma et al. [10]. The percentage of strains resistant to azithromycin and erythromycin was significant in a study by Noormohamed and Fakhr (2013), especially in C. coli recovered from beef liver samples [40]. In contrast, Raeisi et al. (2017) reported that Campylobacter recovered from milk and cattle samples were sensitive to erythromycin [12].
One C. coli strain isolated from a pork sample in our study was resistant to three antibiotics simultaneously and expressed putative virulence genes. These data raise public health concerns, especially given that antimicrobial resistance was observed for the drug of choice for clinical treatment of campylobacteriosis. Multi-drug resistant patterns were previously described for milk and beef Campylobacter isolates in the studies of Kashoma et al., 2016, andRaeisi et al., 2017 [10,12].

Conclusions
We reported on Campylobacter contamination of unpasteurized milk, pork, and beef sold in retail in Poland that may represent potential sources of infection. High resistance rates for fluoroquinolones, and emergence of MDR isolates from pork sample are reported. Moreover, a high level of resistance to ciprofloxacin and tetracycline among C. jejuni and C. coli species indicate the reduced clinical utility of these antibiotics for the treatment of patients. There is also a need for further monitoring of food products in relation to possible transmission of resistant Campylobacter to humans. The present study is the first in Poland to assess the frequency of genes responsible for virulence at different stages of pathogenesis among strains of Campylobacter isolated from food of animal origin, such as milk, beef, and pork. In this study, the number of strains with the key virulence factors was significant; however, differences in the frequency of genes between different sources and species of Campylobacter were also described, which should be further verified.