Prevalence and distribution of pathogenic genes in Campylobacter jejuni isolated from poultry and human sources

Introduction: Campylobacter jejuni is one of the most common bacterial causes of human gastroenteritis. Despite its public health importance, the virulence factors and mechanisms underlying C. jejuni pathogenesis remain poorly understood and the relationships between these genes and the sources of the strains are not clear. We aimed to determine the virulence profiles of C. jejuni isolated from poultry and human cases of Campylobacteriosis. Methodology: A total of 50 strains of C. jejuni isolated from poultry and human cases of Campylobacteriosis were screened by polymerase chain reaction (PCR) for the presence of six pathogenic genes ( flaA , iam , wlaN , cdtA , cdtB , cdtC ). Results: A total of 40% (10/25) of the human isolates presented only one virulence marker. In contrast, 64% (16/25) of the poultry-derived strains showed four or five virulence markers. cdtA and flaA occurred more frequently in poultry-derived strains than in human strains. Ten different virulence profiles were observed among the human isolates and 11 among the poultry strains. Only four profiles were common to both sources: profiles 3 ( flaA , cdtA , cdtB , and cdtC ), 5 ( cdtA and cdtB ), 7 ( flaA and cdtB ), and 10 ( iam , flaA , cdtA , cdtB , and cdtC ). The human-derived strains had a higher Shannon diversity index (1.9396) and Simpson index (0.8367), indicating a more diversified population than found in poultry (1.7742 and 0.7333, respectively). We found variations in the genetic profiles of the circulating strains based on the isolation source and genes that are potentially pathogenic to humans were detected in poultry-derived strains.

Campylobacter species are usually present in the intestine of wild and domesticated animals. Poultry is the major reservoir and the primary source of transmission to humans [6,7]. Human cases of Campylobacteriosis are usually associated with the handling, preparation, and consumption of raw or undercooked chicken meat or cross-contamination with these products [8].
Despite its public health importance, the virulence factors and mechanisms underlying C. jejuni pathogenesis remain poorly studied in Latin-American countries [9]. Moreover, the relationships between these genes and the sources are unclear [10]. Campylobacter virulence factors are mostly associated with motility, chemotaxis, the adhesion to and colonization of intestinal epithelial cells, invasion and translocation capabilities, production of toxins and secreted proteins, and other mechanisms essential for bacterial survival [11].
The cytolethal distending toxin (CDT) consists of three subunits encoded by the cdtA, cdtB, and cdtC genes, which are genetically arranged as an operon [12]. CdtB is the active toxic unit. CdtA and CdtC are subunits required for CDT to bind on target cells and for the delivery of CdtB into the cell's interior [13]. Flagella motility is an important virulence marker for many bacterial pathogens, including C. jejuni, as it is necessary to establish infections [14]. Campylobacter strains carry one flagellum at each pole. Flagella are composed of several different proteins. The extracellular filament structural components produce two flagellins, FlaA and FlaB (encoded by flaA and flaB, respectively) that form the flagellar filament [11,14]. The Campylobacter strains responsible for GBS also carry wlaN, which encodes a β-1,3galactosyltransferase enzyme required for the production of sialylated lipooligosaccharide (LOSSIAL). Campylobacter LOSSIAL structures cross-react with gangliosides in peripheral nerves leading to clinical symptoms characteristic of this syndrome [15,16]. Invasion and host adaptation are influenced by several factors, for example, the expression of the invasion-associated marker (iam) is associated with more invasive Campylobacter strains [17].
Considering that Campylobacter is transferred from animals to humans, and that poultry serves as its main reservoir, knowing the genetic profile of circulating strains in broiler and human populations is important. Therefore, this study aimed to determine the virulence potential of C. jejuni isolated from poultry and humans.

Bacterial strains
A total of 50 C. jejuni strains isolated from poultry (n = 25) and from human cases of Campylobacteriosis (n = 25) were used for this study. The poultry-derived strains were previously obtained from broiler carcass samples (cooled and frozen) from Brazilian poultry slaughterhouses according to the criteria described by the International Organization for Standardization (ISO 10272-1:2017). The human-derived strains were kindly provided by the Oswaldo Cruz Institute Foundation (Fiocruz, Brazil). All strains were isolated in 2012. The bacterial isolates were stored at −80 °C in brain-heart infusion broth (Oxoid, Hampshire, UK) supplemented with 15% glycerol (Synth, Diadema, Brazil). The bacterial strains were cultured on blood base agar (Oxoid) supplemented with 5% defibrinated sheep blood (Laborclin, Curitiba, Brazil) to reactivate them. The plates were incubated under microaerobic conditions using a gas tank [10% carbon dioxide (CO 2), 5% oxygen (O2), and 85% nitrogen (N2)] for 48 h at 42 °C.

DNA extraction
The DNA was extracted using an adapted protocol described by Borsoi et al. [18]. Briefly, an aliquot (1 mL each) of the bacterial cultures were boiled at 95 °C for 10 min. After centrifugation at 8,000 × g for 2 min, the supernatants were stored at -20 °C and used as template DNA. The genetic content of the isolates was confirmed by the polymerase chain reaction (PCR)based detection of 16S rRNA and mapA gene [19].

PCR primer design and amplification
C. jejuni isolates were screened for the presence of six pathogenic genes: flaA, iam, wlaN, cdtA, cdtB, and cdtC. The primers, PCR conditions, amplicon lengths, and references are listed in Table 1.
All PCR amplifications consisted of 10 × PCR Buffer [3 µL; 200 mM Tris-HCl (pH 8.4), 500 mM KCl], Taq thermostable DNA polymerase (2 U/µL), dNTPs (2.5 µL, 2.5 mM), and extracted template DNA (2 µL). For protocol 1 (cdtA and cdtC) the mixture (30 µL) consisted of MgCl2 (1 mM) and primers (0.5 µL each, 20 pmol). Protocol 2 (cdtB) was composed of a mixture (30 µL) with MgCl2 (1.5 mM) and primers (0.5 µL each, 20 pmol). For protocol 3 (iam), the mixture (25 µL) was composed of MgCl2 (1.5 mM) and primers (0.5 µL each, 20 pmol). Finally, for protocols 4 (flaA) and 5 (wlaN), the mixture consisted of MgCl2 (1.2 mM) and primers (0.5 µL each, 10 pmol). Sterile Milli-Q water was used to achieve the final reaction volume (25 µL or 30 µL). All amplification reactions were performed in a thermal cycler according to the cycling conditions described in Table 1. The DNA bands were stained with ethidium bromide to visualize the PCR products. Aliquots of 10 µL were subjected to electrophoresis in a 2% agarose gel in Tris-acetylated EDTA (TAE) buffer for 2 h at 100 V, viewed under an ultraviolet transilluminator, and photographed. The size of the PCR amplicons was determined through comparison with a 100-bp DNA ladder (Invitrogen, Carlsbad, CA, USA). C. jejuni ATCC 33560 and two strains of C. jejuni from our laboratory stock collection were used as positive controls. In all PCRs, a mixture of all constituents of the PCR mixed without the addition of extracted DNA was used as a PCR control.

Statistical analysis
The data were subjected to statistical analysis using the PASW Statistics software (IBM, Hong Kong). Descriptive statistics (frequency distribution) were used to determine the presence of the virulence genes according to the isolation source (poultry and human). Chi square (χ2) and Fisher's exact tests were used to compare the frequencies of virulence-associated genes within and between the isolation sources. Statistical significance was defined as p < 0.05, and the Bonferroni correction was applied to adjust the confidence intervals for multiple hypothesis testing; these adjusted p values are indicated in the tables. Shannon and Simpson diversity index values were determined based on virulence profiles. This index is calculated as the natural logarithm of the proportion of individuals in one particular group divided by the total number of individuals. Populations with higher indexes are considered more diverse [20]. Simpson diversity index values (D) were calculated using the following formula: where n represents the total number of bacterial strains with a particular virulence profile and N is the total number of bacterial strains from the respective source (human or poultry). The value of this index ranges from 0 to 1 with higher values indicating greater sample diversity [21].

Results
The occurrence of virulence genes according to the isolation source is shown in Table 2. Only one strain (poultry origin) did not present any virulence markers. However, no strain presented all of the virulence markers. Among the human isolates, 40% (10/25) presented only one virulence marker. In contrast, 64% (16/25) of the poultry-derived strains presented four or five virulence markers. Statistical analysis indicated that the isolation sources differed significantly (p < 0.05) regarding cdtA and flaA. For both genes, C. jejuni isolated from poultry presented higher frequencies of these pathogenicity-associated genes than the strains isolated from humans. Among the poultry-derived strains, wlaN and iam were significantly less frequent (p < 0.05) than the other genes. Among the human isolates, the frequency of cdtB was significantly higher (p < 0.05) than the other genes.

Discussion
C. jejuni is one of the most reported causes of gastroenteritis worldwide [1]. However, most studies analyzing Campylobacter occurrence and pathogenicity focus on European and North American isolates [22]. The absence of an internal monitoring program adopted by poultry companies and national governments, especially in developing countries, is likely responsible for the lack of more precise data in these regions, including South America [23,24]. The pathogenic potential of C. jejuni strains isolated from Brazilian poultry has increased over the last decade [25]. Recently, C. jejuni strains isolated from poultry in Brazil were characterized according to their zoonotic risk [26]. However, studies comparing virulence factors in human and poultry sources in South American countries remain scarce. Therefore, studies that identify and monitor pathogenic markers in C. jejuni isolated from this region are extremely important for the development of effective management strategies.
Campylobacter spp. pathogenicity depends on the presence of virulence factors that differ among isolates of different origin [27]. In the present study, Campylobacter strains isolated from poultry presented more virulence markers than those obtained from humans. We expected that human-derived strains would present more virulence marks, as previously reported [28], because these isolates were obtained from people presenting Campylobacteriosis, rather than presumably healthy chickens.
The presence of the wlaN gene is associated with the occurrence of GBS in patients with Campylobacteriosis [16]. This gene was only detected in the poultry-derived strains in the present study, which was unexpected. Although previous studies have detected this gene in human isolates in other countries [28,29], studies analyzing Brazilian samples also did not detect wlaN [30]; neither did these studies find a higher occurrence in C. coli strains than in C. jejuni [31]. The wlaN gene is associated with GBS. However, it is not essential, and other genes (e.g., cgtB) have also been associated with the occurrence of this syndrome [16,28]. The presence of wlaN in poultry isolates poses challenges to public health because the consumption of undercooked chicken meat is the most common transmission route of C. jejuni to humans [16].
The genes associated with bacterial motility are usually essential for the development of Campylobacter infection [28]. The flaA gene was significantly more common in the poultry-derived strains (80%) than in the human ones (20%). This result was not expected, since human-derived strains were isolated from persons suffering from Campylobacteriosis, and the poultryderived strains were collected from carcasses of presumably healthy broilers. This differs from previous studies, which found this gene to be highly conserved among C. jejuni, independent of the strain's origin [26,28,32]. Similarly, other Brazilian studies found a high frequency (> 80%) of flaA among human and poultry isolates [25,33], which was not confirmed in the present study. The expression of all three cdt genes is required to maximize the effect of CDT toxin [10,34]. In this study, one strain from each source did not present any of the cdt genes. Van Deun et al. [35] reported that not all Campylobacter strains produce toxins. In contrast, 75% (15/20) of poultry and 45% (9/20) of the human-derived strains possessed all three genes. Our results differ from previous reports which indicate that these genes occur with similar [10] or even a higher frequency in humanderived strains [36]. Furthermore, more pronounced CDT production is associated with human-derived strains [10,35]. Differences in cytotoxicity levels between isolates obtained from humans and poultry have also been previously described [10,35].
The genetic marker iam has been associated with the adherence and invasion of HEp-2 cells in vitro, and is usually positively associated with in vitro invasion [10,38]. Furthermore, a correlation between the clinical occurrence of diarrhea and the isolation of Campylobacter strains that adhere to and invade HEp-2 cells has been established [32]. In this study, a lower number (4/40) of C. jejuni strains presented this gene, regardless of the isolation source. Among the poultry strains, only one (4%) was positive for iam, and three human isolates (12%) presented this gene. The poultry strains analyzed in this study had a lower occurrence of iam than those evaluated in previous reports [39,40]. These results show that the occurrence of this virulence marker depends not only on the isolated Campylobacter species but also on the isolation source [26]. According to Carvalho et al. [40], the absence of iam in invasive strains supports the existence of function polymorphisms, high heterogeneity in the iam locus, and the contribution of additional loci.
In the present study, a variety of virulence profiles were identified, but only four (3, 5, 7, and 10) were shared by the human and poultry sources. Interestingly, profiles 5, 7, and 10 were not common in either of the two isolation sources. Profile 3 was the most common (52%) among the poultry-derived strains, and profiles 1 and 2 among the human isolates (56%).
These virulence profiles were also used to determine the diversity among C. jejuni strains. The Shannon diversity index is a statistical indicator that assumes all groups are represented in a sample and that they are randomly sampled. Populations with higher indexes are considered more diverse [20]. In this study, the human-derived strains had a higher Shannon diversity index (1.9396) than those isolated from poultry (1.7742), indicating a more diversified population in the former. Similarly, Simpson index values were greater for the human-derived strains (0.8367) than their poultry counterparts (0.7333). These results differed from previous studies, which found that poultry isolates were more diverse than the human ones [39]. A possible reason for the difference found in the present study is that the poultry isolates were obtained from a single Brazilian state (RS, Brazil), and, therefore, the strain types may be restricted. On the contrary, although the human-derived strains were also obtained from a single institution, they were isolated from human samples received from across the country. However, it is noteworthy that previous studies have found that, despite the evidence that virulenceassociated genes in C. jejuni are widely dispersed in both species, poultry isolates presented a high occurrence of virulence markers [10,41,42]. A previous study has evaluated the genotypic relationship between human and poultry strains of C. jejuni in Brazil and observed that human isolates presented limited virulence capacity when compared to their poultry counterparts, and were different in molecular typing. They also observed that the ability to cause GBS was similar for both strains [41].
Previous studies have demonstrated that C. jejuni isolated from Brazilian poultry presented a high level of diversity of circulating genotypes, since no clones were observed in strains isolated in the same period (2012) of the isolates from this study [43]. In addition, a concern regarding this high frequency of virulence genes in poultry isolates is their stability in chicken juice through conditions that mimic the transmission route and their transmission through host models, which demonstrate the potential transmission of C. jejuni from food to human [44]. It is possible that the overlap observed in distribution of virulence-associated genes among human and chicken isolates suggests that Campylobacteriosis may be linked with chicken meat [22]. However, their relationship with clinical severity in humans and the expression of virulence factors should be further investigated.
Studies that identify virulence traits are crucial to better understand the risk of Campylobacteriosis associated with different strains and the zoonotic potential of animal strains. However, despite the high prevalence of pathogenic factors found in the present study, it is difficult to predict the in vivo virulence of C. jejuni strains during human infection [10,28]. Thus, the detection of virulence-associated genes should be complemented by gene expression analyses and by determining the pathogenicity through in vivo models [10].

Conclusions
This study has shown that there are variations in the genetic profiles of the circulating strains based on the isolation source (poultry or human) and that genes potentially pathogenic in humans were detected in poultry-derived strains. In this study, poultry strains presented more virulence markers, making these strains potentially capable to adapt to the environment, invade, and cause disease in humans. Our findings support the potential risk of transmitting highly virulent C. jejuni from chicken meat to human. Their relationship with clinical severity in humans and the expression of virulence factors warrant further investigation.