In vitro characteristics of Lactobacillus spp. strains isolated from the chicken digestive tract and their role in the inhibition of Campylobacter colonization

Abstract Campylobacter jejuni/coli infections are the leading cause of bacterial diarrheal illnesses in humans. Many epidemiological studies indicate that improperly prepared meat from chickens that carry a high load of Campylobacter in their intestinal tracts is the key source of human infections. LAB, mainly members of the Lactococcus and Lactobacillus genera, increasingly have been tested as vehicles for the delivery of heterologous bacterial or viral antigens to animal mucosal immune systems. Thus, the objective of this study was to isolate, identify, and characterize Lactobacillus spp. strains isolated from chickens bred in Poland. Their ability to decrease the level of bird gut colonization by C. jejuni strain was also analyzed. First, the influence of the different chicken rearing systems was evaluated, especially the effect of diets on the Lactobacillus species that colonize the gut of chickens. Next, selected strains were analyzed in terms of their anti‐Campylobacter activity in vitro; potential probiotic traits such as adhesion properties, bile and low pH tolerance; and their ability to grow on a defined carbon source. Given that improperly prepared chicken meat is the main source of human infection by Campylobacter, the selected strains were also assessed for their ability to inhibit Campylobacter colonization in the bird's intestine. These experiments revealed enormous physiological diversity among the Lactobacillus genus strains. Altogether, our results showed that L. plantarum strains isolated from the digestive tracts of chickens bred in Poland displayed some probiotic attributes in vitro and were able to decrease the level of bird gut colonization by C. jejuni strain. This suggests that they can be employed as vectors to deliver Campylobacter immunodominant proteins to the bird's immune system to strengthen the efficacy of in ovo vaccination.


| INTRODUCTION
Microbiota present in human or animal intestinal tracts have enormous impact on the health of the host. Recent advances in technology, in particular progress made in the field of DNA sequencing combined with many bioinformatics methods has replaced culture-dependent methods with culture-independent strategies. Data from such analyses reveal details of microbiota diversity and their interactions with the host. Microbial communities of the chicken gastrointestinal tract (GIT), as well as the chicken microbiome, have recently been analyzed by many research group to better understand their influence on bird health and improve poultry industry productivity (for review see [Oakley, Lillehoj, et al., 2014]). The diverse chicken microbiota consists of over 900 bacterial species (Wei, Morrison, & Yu, 2013). Although the microbial communities found in different sections of the chicken GIT vary and their members belong to 13 phyla, the most dominant phyla are Firmicutes, Bacteroides, and Proteobacteria (Pan & Yu, 2014;Stanley, Hughes, & Moore, 2014). Kaakoush et al. (2014), who studied the microbiota of chickens from two farms, assigned them to four enterotypes dominated by varies phyla. Many factors influence the composition of the chicken microbiota; the most substantial factors are the chicken breeds, bird age, and farming practices (Kaakoush et al., 2014;Oakley, Buhr, et al., 2014;Schokker et al., 2015;Thibodeau et al., 2015).
Campylobacteriosis is the most frequently reported foodborne zoonosis worldwide (Kaakoush, Castano-Rodriguez, Mitchell, & Man, 2015). As chickens harbor a high load of Campylobacter jejuni in their GIT, improperly prepared chicken meat is the main source of human campylobacteriosis (Agunos, Waddell, Leger, & Taboada, 2014;Hermans et al., 2012;Kaakoush et al., 2015;Zambrano, Levy, Menezes, & Freeman, 2014). To date, many interventions to combat the chicken colonization by C. jejuni, such as introducing different hygiene barriers at farms, various nutritional strategies or using bacteriophages, have been examined with varying success (Meunier, Guyard-Nicodeme, Dory, & Chemaly, 2016;Newell et al., 2011;Wagenaar, French, & Havelaar, 2013). Given that members of the Lactobacillus genus are predominant in the chicken microbiota and that some of them can be classified as probiotics, their anti-Campylobacter activities have recently been intensively examined using different strategies (Chaveerach, Lipman, & van Knapen, 2004;Robyn, Rasschaert, Messens, Pasmans, & Heyndrickx, 2012;Santini et al., 2010). The aim of these studies was to manipulate the chicken microbiota in order to inhibit chicken digestive tract colonization by Campylobacter (Ghareeb et al., 2012;Santini et al., 2010;Willis & Reid, 2008). The studies showed that only a small percentage of Lactobacillus strains isolated from chicken intestinal tracts or from fecal samples are able to compete with Campylobacter. Although a lot of Lactobacillus isolates have demonstrated high anti-Campylobacter activity in vitro, many only slightly reduce bird colonization by the pathogen when administered in vivo, which demonstrates the complex mechanisms of interactions among microbiota constituents (Chaveerach et al., 2004;Robyn et al., 2012;Santini et al., 2010).
C. jejuni genetic diversity makes the issue even more complicated as inhibition of the pathogen's growth by a specific Lactobacillus isolate is dependent on the C. jejuni strain used in the experiment (Chaveerach et al., 2004;Santini et al., 2010). In some cases, probiotic formulations composed of several bacteria, even belonging to different genera, were proven to be the most effective (Ghareeb et al., 2012;Willis & Reid, 2008). The mechanisms by which Lactobacillus isolates inhibit Campylobacter growth in vivo are diverse and not fully understood. The source of the Lactobacillus strain appears not to be a critical factor, though the data are sometimes contradictory (Chaveerach et al., 2004;Lin, Yu, Jang, & Tsen, 2007;Nishiyama et al., 2014;Santini et al., 2010). Several mechanisms that potentially depend on both the pathogen and the probiotic may be involved in the competitive inhibition of C. jejuni by Lactobacillus. Those that have been proven to play a significant role in inhibition include adhesion to the intestinal epithelium of the host (Nishiyama et al., 2014;Spivey, Dunn-Horrocks, & Duong, 2014); coaggregation and autoaggregation (Tareb, Bernardeau, Gueguen, & Vernoux, 2013); production of antimicrobial compounds, such as organic acid, bacteriocin (Messaoudi et al., 2013;Neal-McKinney et al., 2012;Svetoch & Stern, 2010), and hydrogen peroxide (Mota et al., 2006); and immuno-modulatory effects (Messaoudi et al., 2012).
The objective of this study was to isolate, identify, and characterize the Lactobacillus spp. strains that constitute the chicken intestinal tract microbiota. The first aim of the study was to evaluate the influence of different rearing systems, especially, the dietary treatment of chickens, on Lactobacillus species colonizing their gut. Next, the selected strains were analyzed in terms of their anti-Campylobacter activity in vitro and their ability to colonize chicken's gut (adhesion properties, bile and low pH tolerance). Their metabolic properties, ability to grow on a defined carbon source, were also evaluated.
Given that improperly prepared chicken meat is the main source of human infection by Campylobacter, the ability of the selected strains to inhibit colonization of the bird's intestine by the pathogen was also examined.

| Bacterial strains media and growth conditions
Bacterial strains used in this study are listed in Table 1

| Isolation and identification of lactic acid bacteria
Twenty stool samples from back-yard chickens or cloacal swabs from 25 commercial broiler chickens were streaked onto selective MRS agar and incubated overnight in a microaerobic atmosphere at 37°C.
After 24 hr incubation, colonies were randomly picked from the plates and subcultured two or three times on fresh MRS agar plates. The pure LAB cultures were then kept in MRS broth supplemented with 30% (v/v) glycerol and frozen at −80°C until further analysis.
Preliminary identification consisted of Gram staining to assess the morphology of cells by optical microscopy. Next, amplification of the 16S rDNA gene was performed for all gram positive strains.
For this, chromosomal DNA was isolated from overnight cultures of selected strains. Pellets were resuspended in TES (25 mmol/L sucrose, 50 mmol/L Tris HCl pH 8.0, 10 mmol/L EDTA) containing lysozyme (20 mg/ml, Sigma) and mutanolysin (500 u/ml, A&A Biotechnology). After incubation for 1 hr at 37°C under constant shaking, chromosomal DNA from the bacterial strains was isolated using a commercial kit (A&A Biotechnology, Poland).
Polymerase chain reactions (PCR) were performed with Color op-tiTaq (EURX) or HotStar HiFidelity Polymerase (Qiagen) under standard conditions. For amplification of the 16S rDNA gene, universal primers F27 (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492 (5′-GGTTACCTTGTTACGACTT-3′) were used (McDonald, Kenna, & Murrell, 1995) with an expected PCR product size of 1.5 kb. PCR products were purified using the Clean-up Concentrator kit (A&A Biotechnology, Poland). Synthetic oligonucleotide synthesis and DNA sequencing of the PCR product was performed by Genomed S.A. (Warsaw, Poland), and then the nucleotide sequences were analyzed using BLAST against the nucleotide database at the NCBI website.

| C. jejuni inhibition assays in vitro
The C. jejuni inhibition assay was performed on BA plates containing 5% horse blood (Oxoid) inoculated with 100 μl of Campylobacter cultures (C. jejuni 12 or C. jejuni 81-176), prepared by suspending the microorganisms in 0.9% NaCl and adjusting the density to 0.5 McFarland standard (~1 × 10 8 CFU/ml). Plates were incubated for 2 hr at 37°C under microaerobic conditions. Overnight cultures of lactobacilli were centrifuged, and to determine if bacteriocins contributed to inhibition, part of supernatant was neutralized to neutral pH with 6 N NaOH. Untreated supernatants and supernatants neutralized to pH 7 were subsequently filter sterilized (0.22 μm) and spotted onto MRS agar plates inoculated with Campylobacter. Plates were incubated for 24-48 hr at 37°C under microaerobic conditions. Growth inhibition was evaluated as inhibition around the spotted Lactobacillus supernatant.

| Detection of lactic acid production
Detection of lactic acid was carried out as described earlier by Neal-McKinney et al. (2012). The concentrations of d-and l-lactate in the cell-free supernatants were measured using stereo-specific d-and llactate assay kits (Megazyme, Wicklow, Ireland). The measurements for selected strains were performed in triplicate for reproducibility.
Assays were performed with pure solutions of d-and l-lactate to ensure that the kits were stereo-specific.

| Tolerance to acid and bile
Tolerance to low pH and bile content was assessed as described by Bujnakova et al. (Bujnakova, Strakova, & Kmet, 2014), with minor modifications. Briefly, 100 μl lactobacilli from overnight cultures (37°C, MRS broth, 5% CO 2 ) diluted to OD 660 nm of 1.0 was inoculated into 10 ml of following test solutions: MRS broth control, MRS broth adjusted to pH 2.5, and MRS broth containing 0.5% oxgall (Difco). As a control, broth without inoculation was used. For survival under the different conditions, samples were taken at 6 hr and 100 μl cultures were plated on MRS agar after appropriate dilution. Colonies were counted after 48 hr incubation at 37°C. The survival rate was calculated as the log 10 value of CFU/ml.

| Sugar fermentation pattern
To determine sugar fermentation profiles of the Lactobacillus spp., API 50 CH kit was used according to the manufacturer's instructions (BioMerieux, France). The resulting fermentation pattern was inspected following anaerobic incubation at 37°C after 48 hr.
Fermentation of a carbohydrate was detected by acid production, as demonstrated by a color change in the pH indicator present in the medium. Each strain was tested in three independent experiments.

| Adhesion to bare polystyrene
Adhesion of bacterial cells to polystyrene (PS) was tested on the 96-well microtiter PS plates (Thermo Fischer Scientific Nunc A/S, Denmark) using a modification of the method already described To calculate the adherence ratio of bacteria to PS, the value of absorbance of bacteria was divided by the absorbance of the control sample (PBS, only). Bacteria were characterized as strongly adherent (A ≥ 6), moderately adherent (6 > A ≥ 3), weakly adherent (3 > A > 2), and nonadherent (A ≤ 2).

| Campylobacter colonization experiment
Experiments were performed on Hy-line chickens hatched and reared under controlled conditions from the day of hatch. The chickens were kept under controlled light (L:D 12:12) and temperature (32 ± 2°C during first week and 24 ± 2°C thereafter) conditions, with free access to the standard food and water. Chickens were confirmed to be culture-negative for Campylobacter by cloacal swabbing. All animal ex- Plates that were culture-negative at 48 hr were reincubated for an additional 48 hr. This procedure permits detection of 10 3 CFU/g of cecal contents.

| Isolation and identification of Lactobacillus strains
The Lactobacillus genus includes more than 200 species (http:// www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?id=1578, Thus, in this study, we decided to search for Lactobacillus strains with potential probiotic activity among strains isolated from chickens breeding in Poland and we asked whether different chicken rearing systems, especially dietary treatment, have an impact on Lactobacillus species colonizing their gut (Bjerrum et al., 2006;Pan In total, 107 Lactobacillus strains were isolated: 64 from fecal samples of two privately owned "back-yard" chicken farms and 43 from cloaca of birds reared in seven commercial poultry farms specializing in production of consumer eggs. The hens (Lohman Brown) from farms were at laying age, 22-50 weeks old. The chickens were fed a commercially available feed for laying hens that consists of raw materials (wheat, triticale, corn, chalk, vegetable oils, vitamins, amino acids, and microelements). This feed does not contain growth-promoting antibiotics or coccidiostats.
Chicken stool samples were streaked onto selective MRS agar and incubated within an anaerobic atmosphere overnight at 37°C.
Pure cultures were obtained. Lactobacillus genus membership was confirmed by molecular analysis (PCR amplification of the intergenic chromosomal DNA region between the 16S rDNA and 23S rDNA and evaluation of the sizes of the DNA fragments). To determine the species of the isolates, the 16S rDNA genes were amplified, and the PCR products were sequenced and analyzed using BLAST against the nucleotide database at the NCBI website. The strains were also characterized by morphological and microscopic observations ( Figure S1).
Generally, the most abundant Lactobacillus species was L. salivarius, regardless of chicken breeding methods. L. salivarius (68%) and L. agilis (12%) were more prevalent than other species of Lactobacillus genus obtained from the farm chickens. The species distribution among all the Lactobacillus species isolated from fecal samples of "back-yard" chickens was more diverse. L. salivarius (29%) was the dominant species, followed by L. reuteri (24%) and L. plantarum (23%). Other

| Preliminary characteristics of isolated strains: anti-Campylobacter activity in vitro, lactic acid production, and adhesion properties
Subsequent stepwise analysis was performed using Lactobacillus strains isolated from fecal samples of two privately owned "back-yard" farms, where chickens were allowed access to outdoor areas. The rationale behind this decision was based on two assumptions. First, of all these chickens were never given antibiotics as feed additives at any stage of life. It has been documented that antibiotics may influence the composition of chicken microflora for long periods (Pan & Yu, 2014;Wise & Siragusa, 2007). Second, their food, though difficult to describe in detail, was more diverse than that used on commercial poultry farms.
The preliminary experiments evaluated the inhibitory effect of the Lactobacillus isolates on the growth of Campylobacter using spot titer assays for 64 Lactobacillus strains, where two C. jejuni strains (81176 and 12) of different origins were the target strains. C. jejuni 81176 is human isolate that is widely used in pathogenesis studies, and it has gone through many laboratory passages since it was first isolated (Korlath, Osterholm, Judy, Forfang, & Robinson, 1985). C. jejuni 12 is fresh chicken isolate obtained in our lab. First we checked whether cell-free supernatants (CFS) of analyzed Lactobacillus strains displayed anti-Campylobacter activity. The pH of the culture supernatants ranged from 3 to 5 (predominantly 4). We found that anti-Campylobacter activity was species dependent. Members of L. reuteri species did not exhibit an antagonistic effect on Campylobacter growth. Members of L. curvatus, L. ingluviei, L. kitasatonis, and L. oris also did not inhibit Campylobacter growth, though these results should be treated with caution due to the small number of strains analyzed. Among the strains of L. plantarum and L. salivarius species analyzed, some differences in the size of the inhibition halos were observed (Table S1). Similar results were obtained for both Campylobacter strains.
F I G U R E 1 Pie charts represent relative abundance of Lactobacillus species isolated from cloaca of commercial broiler chickens (a) and from stools of backyard chickens (b) Next, to find the chemical nature of the CFS component responsible for Campylobacter growth inhibition, the CFSs were neutralized before using them for spot titer assays. We found that none of the neutral supernatants demonstrated an antagonistic effect on Campylobacter, independent on the pathogen strains, suggesting that the inhibition was due to acid production.
To shed more light on the mechanism by which Lactobacillus isolates exert their action, the levels of lactic acid present in the supernatants of the cultures of all Lactobacillus strains were evaluated.
In general, we found that L. reuteri strains produced relatively lower amount of lactic acid when compared to L. plantarum or L. salivarius.
However, the supernatants of the L. reuteri culture reveal a pH of about 4, suggesting production of the different organic acids by L. reuteri isolates. Of four examined L. johnsonii strains, two generated high levels of lactic acid. In the case of other Lactobacillus species (L. agilis, L. crispatus, L. curvatus, L. ingluviei, L. kitasatonis, and L. oris), too few strains were examined to make a conclusion concerning the ability of a particular species to generate lactic acid. However, in general and similar to the L. reuteri strains, they produced mild level of lactic acid that did not translate into antagonistic effect against Campylobacter.
It is postulated that highly adhesive probiotic bacteria have the greatest beneficial effects on host health and they should, at least transiently, colonize the host gut (Bermudez-Brito et al., 2012). The adhesion microtiter plate assay is a well-known and simple method for a preliminary assessment of strain adhesion ability (Radziwill-Bienkowska et al., 2014), which, as we observed in our previous studies, can correlate with its adhesiveness to biotic surfaces Radziwill-Bienkowska et al., 2014. In this study, we checked the adhesive properties of 64 Lactobacillus strains isolated from chicken gastrointestinal tracts and compared them with highly adhesive or nonadhesive L. rhamnosus strains (GG or LOCK 0908, respectively). The overwhelming majority of bacteria demonstrated adhesion to PS as determined by an adherence ratio higher than 2 (Table S1). The strength of this ability varied between strains and, in some cases, even exceeded the value determined for the positive control strain L. rhamnosus GG (adherence ratio 9.38). Among the three species most widely represented in this test (L. plantarum, L. reuteri, and L. salivarius), L. plantarum seem to have the most consistent adherence since as many as ~60% of all strains efficiently adhered to PS. However, most of them adhered moderately, and only two strains were strongly adherent. Of all L. salivarius strains analyzed, 58% adhered moderately or strongly. L. reuteri strains were slightly less adherent: half adhered either moderately (six strains) or strongly (two strains; Table S1).
On the basis of the preliminary characteristics, 11 strains were chosen for detailed examination in subsequent detailed in vitro experiments. The selection was not obvious as their activity against Campylobacter, lactic acid production and adhesion properties revealed a high level of diversity.
The main selection criterion was the strong ability to inhibit Campylobacter growth. The selected strains of L. salivarius and L. plantarum differ in respect to the two other qualities. Two of L. salivarius strains (PA17D and PA18C) generate high levels of lactic acid but each exhibits a different level of adhesion to PS. L. salivarius PA14C, by contrast, is characterized by low production of lactic acid and high adhesion. The differences among the chosen L. plantarum strains were related to their adherence capacity to PS. L. agilis and L. crispatus were added to the list due to their strong ability to inhibit Campylobacter growth. L. reuteri strains were included in consideration of their abundance in the microbiota of "back yard" chickens, although they showed no anti-Campylobacter activity.

| Lactic acid production
First, the production of lactic acid by the lactobacilli was assessed in detail ( Figure 2). The highest producers were two strains -L. crispatus 11D F I G U R E 2 The concentrations (g/L) of d-and l-lactate in the cell-free supernatants from Lactobacillus culture. The concentration were measured using stereo-specific d-and l-lactate assay kits (Megazyme, Wicklow, Ireland)  Goffin et al., 2005). Thus resistance to vancomycin was also checked using M.I.C. Evaluator (Oxoid). We found that all but one (L. crispatus) of the strains were resistant to a high concentration of vancomycin (>256 μg/ml).

| Carbohydrate fermentation profiles of Lactobacillus sp
Utilization of many oligosaccharides from different categories appears to be an ubiquitous feature of lactobacilli (Ganzle & Follador, 2012).
We analyzed the ability of selected strains (members of the L. salivarius, plantarum, reuteri, and agilis species) to grow on media containing various sugars as a carbon source, using the API test (Figure 3) l-arabinose, l-fucose, l-rhamnose, methyl αd-glucopyranoside, methyl αd-mannopyranoside, N-acetylglucosamine, salicin, starch, and d-melibiose) was found to be species-dependent. The profiles for each Lactobacillus strain are summarized in Figure 3.

| Osmotic stress resistance
The ability to resist high salt concentrations is important for probiotic bacteria, since they must survive and grow in the gastrointestinal tract where the environment has an osmolarity equivalent to 0.3 mol/L NaCl (Chowdhury, Sahu, & Das, 1996;West et al., 1996).
In this study, the isolated Lactobacillus spp. showed a variable capacity to survive different concentrations of NaCl. The level of bacterial osmo-tolerance was more species-than strain-dependent, but none of the isolates could grow in 1.36 mol/L NaCl. The strongest resistance (up to 1 mol/L) was found for all L. plantarum isolates, whereas the resistance of L. salivarius depended on the strain. L. agilis, L. crispatus, and L. reuteri strains showed the weakest resistance, with only faint tolerance to 0.3 mol/L NaCl (Figure 4).

| Tolerance to acid and bile salt
The capacity of Lactobacillus strains to act as probiotics is also determined by the ability to survive in the low pH of the stomach and in the high concentration of bile salt of the gastrointestinal tract (Bull, Plummer, Marchesi, & Mahenthiralingam, 2013). In our studies, we decided to check the ability of chosen strains to tolerate pH 2.5 and 0.5% bile salt (

| Oral administration of selected Lactobacillus strains to chickens -impact on colonization of chickens by Campylobacter
Of the 11 Lactobacillus strains characterized in detail, five were tested in vivo in chicken experiments (Table 3) F I G U R E 4 Tolerance of Lactobacillus isolates to different concentration of NaCl. Black circle -full resistance, B&W circleweak resistance, white circle -lack of resistance 1 × 10 8 CFU/g of cecal content). The reduction in C. jejuni cecal content due to L. plantarum application was apparent 4 days after challenge. In the case of L. plantarum PA18A, the median reduction was 1 log 10 . In the case of L. plantarum PA20A, the mean concentration of C. jejuni cells 4 days after challenge was similar to that observed for control group. However, distinct differences appeared among birds in this group. Two of six chickens were colonized at a low level (about 1 × 10 5 and 1 × 10 6 CFU/g of cecal content).

| DISCUSSION
A wide array of carriers, both living and nonliving, has been devel- We have shown that in ovo chicken immunization with liposomes containing Campylobacter hybrid protein rCjaAD results in a moderate protection against colonization . The same hybrid protein expressed on the surface of L. lactis also exerted a moderately positive effect, when used for chicken vaccination (Kobierecka, Olech, et al., 2016). We postulate that the efficacy of in ovo immunization with nonliving carriers may be increased by a boost and immunization after hatching with live vectors carrying Campylobacter antigens known to induce a mucosal immune response.
The main drawback of the L. lactis delivery vehicle is the lack of longlasting colonization of the chick digestive tract. Using Lactobacillus spp. that can consistently colonize, or at least persist for an extended period in the bird intestine and display probiotic properties, should result in more efficient induction of the immune response than that exerted by L. lactis. Lactobacillus strains may be used to decrease the number of Campylobacter colonizing the chicken digestive tract, but at the same time, they can be employed as a vector to deliver Campylobacter immunodominant proteins to the bird immune system. The Campylobacter jejuni inhibition assay was performed on Blood Agar plates containing 5% horse blood (Oxoid) inoculated with 100 μl of Campylobacter cultures. Campylobacter growth inhibition by supernatant of Lactobacillus strains is marked as +. Bacterial adhesion was determined using the technique described by Radziwill-Bienkowska et al. (2014) in three independent experiments. Bacteria were characterized as strongly adherent (A ≥ 6), moderately adherent (6 > A ≥ 3), weakly adherent (3 > A > 2), and nonadherent (A ≤ 2). The adherence ratio for strain L. rhamnosus GG (positive control) was 9.38 and the ratio for nonadhesive L. rhamnosus LOCK 0908 (negative control) was 1.05.
The concentrations (g/L) of d-and l-lactate in the cell-free supernatants were measured using stereo-specific d- Indeed, L. plantarum is commonly found in many different ecological niches, such as vegetables, meat, fish, dairy products as well as in the gastrointestinal tracts (Siezen & van Hylckama Vlieg, 2011 (Ferain et al., 1996;Goffin et al., 2005). This phenotype is seen mostly in L. plantarum and some other heterofermentative lactobacilli.
The only strain sensitive to vancomycin was L. crispatus, which similar to L. plantarum produces high levels of d-lactate and belongs to the facultatively heterofermentative group of lactobacilli. Explanation of this vancomycin sensitivity requires more research, as the mechanism that determines vancomycin resistance was studied in detail only in L.
Anti-Campylobacter activity expressed in in vitro experiments has been documented for some Lactobacillus strains of different origins (Ghareeb et al., 2012;Robyn et al., 2012;Santini et al., 2010).
Although many Lactobacillus strains have been examined for their anti-Campylobacter activity in vitro, only limited studies to date have been undertaken to confirm this effect in vivo. Our knowledge about interaction among constituents of intestinal microbiota is rather limited. Thus, the data arising from in vitro experiments, although they provide useful information about strain physiology, have to be treated with caution and only as indication for planning further experiments.
To evaluate anti-Campylobacter activities in vivo, potential probiotic Lactobacillus species have been administered as mono-or multispecies feed additives. The studies have differed in their experimental designs and the data obtained through in vitro examination were not always consistent with those provided by animal studies (Robyn, Rasschaert, Hermans, Pasmans, & Heyndrickx, 2013;Robyn et al., 2012;Santini et al., 2010). The main differences among in vivo experiments concern the length of time of formula administration and C. crispatus species, but the two strains appear to differ in terms of their physiology, as based on type and amount of lactate production. Of five species tested in this work, lactobacilli strains that are members of L.
plantarum appeared to be the most effective. Other strains did not display anti-Campylobacter activity, though it is not known whether a more frequent and longer application period would improve their efficacy.
Overall, the present data show that L. plantarum strains isolated from digestive tracts of chicken bred in Poland displayed some probiotic attributes in vitro and were able to decrease the level of bird gut colonization by C. jejuni strain. However, further experiments are needed to understand details concerning their interaction with C. jejuni, as well as with the host, which should help to improve the schema of its application. Additionally, recent experiments in our lab show that L. plantarum strains are susceptible to genetic manipulation, so they may be used as potential delivery vectors for C. jejuni antigens.

ACKNOWLEDGMENTS
We thank Dr. Jeffrey Hansen for critical reading of the manuscript.