Using In Vitro Immunomodulatory Properties of Lactic Acid Bacteria for Selection of Probiotics against Salmonella Infection in Broiler Chicks

Poultry is known to be a major reservoir of Salmonella. The use of lactic acid bacteria has become one of successful strategies to control Salmonella in poultry. The purpose of this study was to select lactic acid bacteria strains by their in vitro immunomodulatory properties for potential use as probiotics against Salmonella infection in broiler chicks. Among 101 isolated lactic acid bacteria strains, 13 strains effectively survived under acidic (pH 2.5) and bile salt (ranging from 0.1% to 1.0%) conditions, effectively inhibited growth of 6 pathogens, and adhered to Caco-2 cells. However, their in vitro immunomodulatory activities differed significantly. Finally, three strains with higher in vitro immunomodulatory properties (Lactobacillus plantarum PZ01, Lactobacillus salivarius JM32 and Pediococcus acidilactici JH231) and three strains with lower in vitro immunomodulatory activities (Enterococcus faecium JS11, Lactobacillus salivarius JK22 and Lactobacillus salivarius JM2A1) were compared for their inhibitory effects on Salmonella adhesion and invasion to Caco-2 cells in vitro and their antimicrobial effects in vivo. The former three strains inhibited Salmonella adhesion and invasion to Caco-2 cells in vitro, reduced the number of Salmonella in intestinal content, spleen and liver, reduced the levels of lipopolysaccharide-induced TNF-α factor (LITAF), IL-1β, IL-6 and IL-12 in serum and increased the level of IL-10 in serum during a challenge study in vivo more efficiently than the latter three strains. These results suggest that in vitro immunomodulatory activities could be used as additional parameters to select more effective probiotics as feed supplements for poultry.


Introduction
Salmonella is one of the major causes of food-borne illnesses in humans. Poultry is known to be a major reservoir of Salmonella. Vaccination and bio-security have been adopted to control Salmonella in poultry production. In addition, lactic acid bacteria (LAB) based probiotics have been used to control Salmonella in poultry [1]. LAB probiotics mainly consist of Lactobacillus, of intestinal content from duodenum, jejunum, ileum or cecum were diluted with 10 ml phosphate-buffered saline (PBS, pH 7.4) and plated onto DeMan, Rogosa, and Sharpe (MRS) plates (Becton Dickinson, Mississauga, ON, Canada) for selecting lactic acid bacteria, [13]. The plates were grown at 37°C under anaerobic conditions (85% N 2 , 10% CO 2 , and 5% H 2 ) for 48 h, and individual colonies were selected and inoculated into MRS broth, cultured at 37°C under anaerobic conditions for 18 h and subcultured twice. The V3 region of the 16S rRNA gene was amplified by PCR using the specific primers and sequenced as described in a previous study [13] and sequence comparisons were performed using the Basic Local Alignment Search Tool (BLAST) program (http://www.ncbi.nlm.nih.gov/BLAST/).
Pathogenic bacteria used in this study to measure antibacterial activities of probiotics included Staphylococcus aureus ATCC 29213, Escherichia coli K88, 25922 and 1569, Salmonella Enteritidis ATCC 13076 and Salmonella Typhimurium ATCC 14082. To culture these bacterial strains, one colony of each strain was inoculated into 5 ml tryptic soy broth and incubated with shaking at 37°C for 12 h.

Culture of cell lines
Caco-2 cells (ATCC HTB-37), a human colon adenocarcinoma cell line, were maintained in Dulbecco's modified Eagle's medium (DMEM)/F12 media (Gibco, USA) supplemented with 10% (v/v) inactivated (30 min, 56°C) fetal bovine serum, 20 U/ml penicillin and 100 μg/ml streptomycin. Cells were cultured at 37°C in a 5% CO 2 /95% air atmosphere using a humidified CO 2 incubator. Cells were used at post-confluence after 21 days of culture. For adhesion assays as well as inhibition of intestinal cell adhesion and invasion by pathogens, monolayers of Caco-2 cells were prepared in 24-well tissue culture plates (Costar 3524, Corning Inc. NY, USA). Cells were seeded at a concentration of 5 × 10 4 cells/ml in the DMEM-F12 medium without penicillin and streptomycin.

Preparation of spleen mononuclear cells
Spleens were harvested sterile from four adult Arbor Acres chickens, which were obtained from Yangling, Shaanxi, China. After euthanizing via cervical dislocation, the spleens were rinsed in 1 × Hanks' balanced salt solution (HBSS) and then minced with sterile scalpels. The tissue was further disrupted with the flat end of a 10-ml syringe plunger and filtered through a 40-μm nylon cell strainer to obtain a single-cell suspension. The suspension was then overlaid onto a Histopaque-1077 (Sigma, Oakville, ON, Canada) density gradient and centrifuged at 400 × g for 30 min. Mononuclear cells at the interface were collected and washed twice in 1 × HBSS and then suspended in RPMI 1640 (RPMI containing 10% fetal bovine serum, 2% chick serum, 0.146 g L-glutamine, 1.6 mM 2-mercaptoethanol). Cells were counted by the trypan blue dye exclusion assay before being resuspended in RPMI 1640 [12].

Resistance of LAB to acid and bile and antimicrobial activities of LAB
To evaluate acid resistance among the LAB strains, methods described by Tsai et al. [2] were used. Effects of bile salts on the growth of LAB cells were studied by a method modified from that of Osmanagaoglu et al. [14]. Briefly, for determining the acid tolerance, 150 μl of each culture containing about 10 8 −10 9 cfu/ml of LAB suspension was added to 4.85 ml MRS that had been adjusted to pH 2.5 by 0.1 N HCl. To test the bile salt resistance, the same inoculum was added to normal MRS supplemented with 0.1%, 0.3%, 0.5% or 1.0% bile salts (Sigma, Saint Louis, Missouri, USA). Each mixture was incubated at 37°C for 3 h. After incubation, viable bacterial counts were determined by plating serial dilutions (with PBS, pH 7.4) on the MRS agar under anaerobic conditions at 37°C for 48 h. These assays were performed in triplicate for each of 3 independent experiments.

Adhesion of LAB to intestinal cell lines
The adhesion study of LAB strains was performed by following the procedures described by Bianchi et al. [15] and Tsai et al. [2]. LAB strains were stained with fluorescein isothiocyanate (FITC, Sigma, USA) and kept in darkness at 37°C for 2 h, then washed three times with antibiotic-free PBS solution (pH 7.4) to remove the unlabeled FITC and resuspended with PBS. One hundred μl of each suspension (5×10 8 cfu/ml) was transferred to the 24-well multidish containing the Caco-2 cells and incubated for 2 h. After incubation, non-adherent bacteria were removed by washing three times with PBS. Two hundred μl of trypsin/EDTA-Na 2 were used to digest the cells and adherent bacteria for 5 min before 600 μl of PBS were added into the wells. After mixing, a 200 μl mixture containing cells and bacteria was added to 96 well plates. This fraction contained lysed bacteria attached to Caco-2 cells or within Caco-2 cells and it was reported as the adherent fraction. The fluorescence was read on a Perkin-Elmer LS55 Spectrophotometer (λ ex = 492 nm; λ em = 517 nm). Six independent experiments were performed for each microbial strain on Caco-2 cells. The results of adhesion assay were expressed as the adhesion percentage of adherent bacteria over added bacteria per well.

Assays for LITAF and IL-12 production by spleen mononuclear cells after stimulation with LAB strains
Spleen cells were cultured in triplicate, at a density of 1×10 6 cells/ml of RPMI-1640 medium without penicillin or streptomycin, in 24-well tissue culture plates. LAB cells were centrifuged at 8,000 g for 5 min and the pellet was resuspended in RPMI-1640 medium containing spleen cells to a final concentration from 10 8 to 10 9 cfu/ml. Lipopolysaccharide (10 μg/ml) from E. coli O26:B6 (Sigma, USA) was used as a positive control. After 24 h and 48 h, LITAF and IL-12 produced in these culture supernatants were analyzed.
Cytokines were measured using an enzyme-linked immunosorbent assay (ELISA) method. Ninety-six-well Immuno-Maxisorp plates (Nunc) were coated with polyclonal antibodies for LITAF and IL-12 (1: 1,000) in coating buffer (0.05 M Carbonate Buffer, pH 9.6) overnight at 4°C. Plates were blocked and washed. Culture medium was added to the plates and they were incubated for 2 h at room temperature. Plates were then washed again, and goat-anti-rabbit secondary antibody conjugated horseradish peroxidase (1: 20,000) was added, followed by incubation for 1 h at room temperature and another wash. The chromogenic reactions were developed with the 3, 3', 5, 5'-tetramethylbenzidine substrate at 37°C for 30 min. The reactions were terminated with 50 μl of 2N H 2 SO 4 and the absorbance at A 450 nm was measured. Equivalent levels of LITAF and IL-12 were calculated by comparison with reference curves generated using LITAF and IL-12 standards. The results were expressed as the concentration of the cytokines in the culture medium (ng/ml).

Inhibition of Salmonella adhesion and invasion of Caco-2 cells by LAB strains
Each LAB strain (5×10 7 cfu per well) was added to Caco-2 cells in a fresh tissue culture medium without Penicillin-Streptomycin and incubated at 37°C for 2 h in a 5% CO 2 /95% air atmosphere incubator before 2 h of incubation with 100 μl of FITC labeled Salmonella (5×10 8 cfu/ml). Salmonella were labeled with Fluorescein isothiocyanate (FITC, Sigma, USA) in darkness at 37°C for 2 h. After incubation with FITC labeled Salmonella, non-adherent bacteria were washed away three times with PBS.
To determine the relative adhesion of Salmonella to Caco-2 cells, one 24-well multidish was added with 200 μl trypsin to digest the cells and adherent bacteria for 5 min before adding 600 μl of PBS. After mixing, 200 μl of mixture containing cells and bacteria were added to each well of 96-well plates to measure the strength of fluorescence. To determine the relative invasion of Salmonella to Caco-2 cells, another 24-well multidish was lysed with 200 μl of 1% Triton X-100 for 10 min before trypsin digestion. Six independent experiments were performed for each LAB strain.
The relative adhesion or invasion of S. Enteritidis ATCC13076 to Caco-2 cells was expressed as a percentage using the following formula: relative of adhesion or invasion = 100 x A1/A2, where A1 and A2 were the percentages of adhesion or invasion by S. Enteritidis ATCC13076 in the presence and absence of LAB strains, respectively.

Challenge study
The challenge study was performed by following the procedures described by Chen et al. [4] with few modifications. One hundred ninety-two hatched 1-d-old healthy Arbor Acres male broilers were obtained from a commercial hatchery at Xianyang, Shaanxi, China. Chicks were reared in two layer metal cages, with an average stocking density of 16.7 birds per square meter, and the brooding temperature was 31 to 33°C throughout the experiment. Chicks had free access to water and a commercial starter diet without supplementation of antibiotics. The chicks were randomly divided into 8 groups (6 repeats/group, 4 chicks/repeat): In the negative control group (group 1), the chicks were only given sterile PBS buffer (pH 7.4) (0.2 ml/chick) via the intragastric route once every day throughout 4 experimental days. In the positive control group (group 2), chicks were challenged with Salmonella on d4 (0.2 ml/chick, 10 8 cfu per 0.2 ml) and were given sterile PBS buffer (pH 7.4) (0.2 ml/chick) during d1-d3. For treatment groups 3-5, 3 LAB strains with higher in vitro immunomodulatory properties (L. plantarum PZ01, L. salivarius JM32 and P. acidilactici JH231) were used, while 3 LAB strains with lower in vitro immunomodulatory properties (E. faecium JS11, L. salivarius JK22 and L. salivarius JM2A1) were used for treatment groups 6-8. For treatment groups 3-8, chicks were gavaged with LAB (0.2 ml/chick, 10 9 cfu per 0.2 ml) once every day for 3 consecutive days, then challenged with Salmonella Enteritidis ATCC 13076 on d4 (0.2 ml/chick, 10 8 cfu per 0.2 ml).
Broiler chickens were euthanized via cervical dislocation. Samples for blood, spleens, livers and cecum contents of six randomly selected chicks in each group were taken at 1, 3 and 5 days post Salmonella challenge. Salmonella Enteritidis ATCC 13076 was selected for the challenge study, due to the invasive characteristic previously described by Dawoud et al. [17].
Assessment of LITAF, IL-1β, IL-6, IL-10 and IL-12 in chick serum and enumeration of the Salmonella cells invaded in chick liver and spleen, and colonized in the cecum For all groups, blood samples were collected from the carotid artery of 6 chicks at 1, 3 and 5 days post Salmonella challenge. Blood serum was obtained after incubation for 1 h at room temperature followed by 2,000 g for 10 min. Sera were stored at -80°C until tested. The cytokines LITAF, IL-1β, IL-6, IL-10 and IL-12 were measured by the ELISA method with preparation of specific antibodies of LITAF, IL-1β, IL-6, IL-10 and IL-12 in our laboratory. Samples for spleens, livers and cecum contents of six randomly selected chicks in each group were taken at 1, 3 and 5 days post Salmonella challenge. Salmonella that had invaded the spleens and livers and colonized in the cecum of chicks were enumerated by cfu method [4]. Spleens and livers were homogenized and without serial dilution, and cecum contents were diluted with PBS (pH 7.4). All the samples were incubated on selected Brilliant Green agar (Difco) containing 50 μg/ml novobiocin (Sigma, USA) for 24 h at 37°C before counting the number of Salmonella.

Statistical analysis
All results were expressed as mean ± SD from at least three independent experiments. Statistical analysis was performed using the SPSS for Windows version 17.0 (Chicago, IL, USA). Data were subjected to one-way ANOVA and, where appropriate, the Scheffe test was used for comparison of means. Differences were considered to be statistically significant when the P value was <0.05.

Results
Resistance of LAB to acid and bile, antimicrobial activity and in vitro adhesion assay to Caco-2 cells Among the 101 isolated LAB strains, 13 strains were better than the remaining 88 LAB strains to survive following exposure to acid (pH 2.5) or bile salts for 3 h (Table 1). These 13 bacteria had a survival rate of over 77% after 3 h at pH 2.5. With the presence of 1.0% of bile salts, the survival rates were more than 60.0% for these 13 strains. Bile salts at the 0.1% level had no significant effect on bacterial survival. As shown in Table 2, the above-mentioned 13 LAB strains were able to inhibit the growth of the pathogenic bacteria. All 13 LAB strains had strong inhibition against Escherichia coli except JM31, JK22, JK231 and JS11 which exhibited mild inhibition against ATCC K88. Similarly, all 13 LAB strains had strong inhibition against Salmonella and Staphylococcus aureus except JM 241 and JH231 which exhibited mild inhibition against Staphylococcus aureus and Salmonella Typhimurium. All 13 tested strains were able to adhere to Caco-2 cells with different adhesion activities (Table 3). JS233, JM41, PZ01 and JK21V showed the strongest adherence to Caco-2 cells (from 19.67 to 10.94%), while JS11 was the least effective one to adhere to Caco-2 cells (2.24 ± 0.24%).
LITAF and IL-12 production by spleen mononuclear cells in response to recall antigen stimulation Except JM31, JM241 and JM14, the remaining strains enhanced LITAF production more than the positive control at 48 h post-treatment (P<0.05), especially PZ01 (57.81 ng/ml), JM32 (32.64 ng/ml) and JH231 (4.44 ng/ml). Only JM31 and JM41 produced more LITAF than the positive control group after co-incubation with spleen mononuclear cells for 24 h (P<0.05) (Fig 1A). At 24 h post-treatment, Pediococcus spp. (JH231, JS233 and JM241) did not increase IL-12 production, while the other LAB strains significantly induced IL-12 production in comparison with the LPS positive control group (P<0.05). All 13 strains except JM241 enhanced IL-12 production more than LPS after incubation for 48 h (P<0.05). The highest producers of IL-12 were PZ01 (724.85 ng/ml), JM32 (427.63 ng/ml) and JH231 (22.25 ng/ml) (Fig 1B). These same 3 strains were also the highest producers of LITAF. Thus, they were retained for further assays. Three other strains (JS11, JK22 and JM2A1) were randomly selected to represent lower producers of LITAF and IL-12 for subsequent assays. The remaining 7 LAB strains were not studied further.

Inhibition of pathogen adhesion and invasion to Caco-2 cells by LAB
The competitive inhibition of adhesion and invasion of Salmonella Enteritidis to Caco-2 cells by 6 probiotic strains was shown in Fig 2A and 2B, respectively. Strain PZ01 displayed the strongest inhibition of Salmonella Enteritidis adhesion and invasion to Caco-2 cells (Fig 2). On average, 3 strains with higher in vitro immunomodulatory properties reduced adhesion by

Inhibitory effects of selected LAB strains on the invasion and colonization of Salmonella in vivo
In order to determine whether LAB strains with higher in vitro immunomodulatory properties had better inhibitory effects against Salmonella in vivo, a Salmonella challenge study was performed. As shown in Table 4, all 6 strains significantly decreased Salmonella Enteritidis in cecum content on sampling time points, compared with the positive control (P<0.05). On average, 3 strains with higher in vitro immunomodulatory properties reduced the bacterial counts by 88.37%, 94.93% and 96.63%, while 3 strains with lower in vitro immunomodulatory properties reduced the bacterial counts only by 61.52%, 73.12% and 76.24% on d1, d3 and d5 post-infection, respectively. Salmonella was detected in livers and spleens from d1 to d5 post-infection for the positive control group. Three strains with higher in vitro immunomodulatory properties prevented  The relative adhesion or invasion of S. Enteritidis ATCC13076 to Caco-2 cells was expressed as a percentage using the following formula: relative adhesion or invasion = 100 x A1/A2, where A1 and A2 were the percentages of adhesion or invasion by S. Enteritidis ATCC13076 in the presence and absence of LAB strains, respectively. Each value represents the mean value ± SD from six trials. Different letters above bars indicate significant differences among treatments within each sampling day (P<0.05).
invasion of Salmonella into livers and spleens, with no viable Salmonella detected on d1, d3 and d5. Three strains with lower in vitro immunomodulatory properties reduced invasion of Salmonella in livers and spleens, especially for d1 with no viable Salmonella detected except for strain JK22 in the spleen. On d3, JS11 significantly reduced invasion of Salmonella in livers and spleens, compared to the positive control. Other 2 strains also significantly reduced invasion of Salmonella in livers but they had no effects in spleens. For d5, only JS11 and JM2A1 significantly reduced invasion of Salmonella in livers (P<0.05), while all three strains with lower in vitro immunomodulatory properties had not significantly reduced invasion of Salmonella in spleens (P>0.05).
Assay of the cytokines LITAF, IL-1β, IL-6, IL-10 and IL-12 in chick serum Blood samples were assayed for pro-inflammatory cytokines LITAF, IL-1β, IL-6, IL-12 and anti-inflammatory cytokine IL-10 (Fig 3). For the positive control group, the highest expression levels of LITAF, IL-1β and IL-12 were observed at 1 day post-infection and gradually decreased afterwards. The opposite was true for IL-6 with increasing concentrations after the infection. However, there was no change of IL-10 after infection. In comparison with the positive control, three strains with higher in vitro immunomodulatory properties (PZ01, JM32 and JH231) significantly reduced LITAF, IL-1β, IL-6 and IL-12 and increased IL-10 levels at all 3 time points. Three strains with lower in vitro immunomodulatory activities (JS11, JK22 and JM2A1) also significantly reduced levels of LITAF, IL-1β, IL-6 and IL-12 mainly on d1 and d3 but did not affect IL-10 levels in comparison with the positive control. The abilities of PZ01, JM32 and JH231 to reduce LITAF, IL-1β, IL-6 and IL-12 and to increase IL-10 were significantly greater than those JS11, JK22 and JM2A1.

Discussion
This study was the first to evaluate effects of LAB on LITAF and IL-12 expression by the spleen mononuclear cells of chickens at the protein level. The reason we chose spleen mononuclear cells was because lactic acid bacteria induced significantly more cytokines in spleen cells than in cecal tonsil cells of chickens [11]. LITAF and IL-12 levels detected in our study were generally higher than those from previous studies using mouse macrophages RAW 264.7 cells [2,3]. TNF-α production was significantly increased by lactic acid bacteria (LAB) strains after their co-culturing with RAW 264.7 cells [2]. A further study from the same research group indicated that viable and heat-killed LAB strains, either individually or in mixture, were able to induce the release of TNF-α and IL-12 from RAW 264.7 cells [3]. Considering the species specialization and higher sensitivities of spleen mononuclear cells from chickens than RAW 264.7 cells, it seems reasonable to use spleen mononuclear cells from chickens to evaluate in vitro immunoregulatory activities of lactic acid bacteria for poultry production.
In this study, we demonstrated that cytokines, LITAF, IL-1β, IL-6, IL-10 and IL-12, were involved in the immunity of Salmonella-infected chicks. Three strains with higher in vitro immunomodulatory properties (PZ01, JM32 and JH231) reduced LITAF, IL-1β, IL-6 and IL-12 and increased IL-10 more efficiently than three other strains with lower in vitro immunomodulatory activities (JS11, JK22 and JM2A1). TNF-α is a member of a group of cytokines that stimulate the acute phase reaction in mammal. Although TNF-α has not been found nor described in the chicken genome, LITAF, which is the regulator for TNF-α expression in mammal [18], has been shown to play an important role in the intestinal inflammatory response in chicken [19]. IL-12 is produced by inflammatory myeloid cells and influences the development of T H 1 cell responses [20]. Similarly, IL-1β is also a major mediator of inflammation and is produced by monocytes, tissue macrophages, enterocytes and other cells [21]. These three cytokines (LITAF, IL-1β and IL-12) indicate an early inflammatory response [22]. The concentrations of LITAF, IL-1β and IL-12 at the protein level were the highest at d1 and gradually decreased afterwards in the positive control group. Similar changes have been reported at the mRNA level by Chen et al. [4]. IL-6 is a multifunctional cytokine. IL-6 gradually increased after the Salmonella challenge, similar to the previous findings at the mRNA level for chickens [4,23] and at the protein level for mice [3]. In comparison with the positive control, three strains with higher in vitro immunomodulatory activities (PZ01, JM32 and JH231) significantly increased IL-10 levels, while three strains with lower in vitro immunomodulatory activities (JS11, JK22 and JM2A1) did not affect IL-10 levels. In addition, this study showed that three strains with higher in vitro immunomodulatory properties (PZ01, JM32 and JH231) reduced the levels of Salmonella Enteritidis recovered from chick livers, spleens and cecal contents more efficiently than three strains with lower in vitro immunomodulatory activities (JS11, JK22 and JM2A1).
The selected six LAB strains for the in vivo study showed consistent tolerance to acid and bile salts in vitro, suggesting that these six LAB strains could survive the gastrointestinal tract and function effectively [24,25]. Moreover, these six LAB strains were able to inhibit the growth of the pathogenic bacteria. Antimicrobial activities of all six strains might be associated with acidic metabolites such as acetic acid, lactic acid [26] and organic acid [27], or bacteriocins [28,29] and proteinaceous substances [30]. Finally, these six LAB strains were biologically safe due to negative haemolytic activities (data not shown).
In conclusion, compared with strains E. faecium JS11, L. salivarius JK22 and L. salivarius JM2A1 with lower in vitro immunomodulatory properties, strains L. plantarum PZ01, L. salivarius JM32 and P. acidilactici JH231 with higher in vitro immunomodulatory activities were more effective to reduce Salmonella counts in cecal content and decease invasion of Salmonella into livers and spleens. These results suggest that in vitro immunomodulatory activities could be used as additional parameters to select more effective probiotics for poultry.