All experimental procedures used were approved by the China Agricultural University Animal Care and Use Committee. Bacterium culture media and agars were obtained from Beijing Land Bridge Technology Co. Ltd. (Beijing, China). Cell culture media, buffers and fetal bovine serum (FBS) were purchased from Gibco (Carlsbad, CA, USA). Epidermal growth factor (EGF) was purchased from BD Biosciences (San Jose, CA, USA). Bacterium and cell culture plastics were obtained from Corning Life Science (Tewksbury, MA, USA). All other chemicals, unless specified, were obtained from Sigma-Aldrich (St. Louis, MO, USA).

Lactobacillus acidophilus (CGMCC No. 1.1878) and L. fermentum (CGMCC No. 1.2029) were obtained from China General Microbiological Culture Collection Center (CGMCC, Beijing, China), and cultured in Mann-Rogosa-Sharpe (MRS) broth. The chicken C. perfringens type A field strain (CVCC No. 2030) was obtained from China Veterinary Culture Collection Center (CVCC, Beijing, China) and cultured in cooked meat medium. Both lactobacilli and C. perfringens were incubated at 37°C under anaerobic conditions. All bacterial experiments were performed with MRS broth, since this provided the best overall growth environment for all species (Allaart et al., 2011). For cell experiments, overnight-incubated bacterial cultures were centrifuged at 4,000 × g for 10 min, washed twice by sterile phosphate buffered saline (PBS, pH 7.4) and resuspended in cell culture medium. Bacterial concentration was estimated by measuring the absorbance at 600 nm and by relating the optical density to colony forming units (CFU, data not shown).

The procedures for bacterial co-culture experiments used in the current study were previously described by Jiang et al. (2014) with some modifications. Strains of lactobacilli and C. perfringens were cultured individually under anaerobic conditions in MRS broth and cooked meat medium, respectively, overnight to reach a stationary phase. Both cultures were centrifuged at 4,000 × g for 10 min. The pellet was then washed once with PBS and resuspended in fresh MRS broth with a concentration of 1 × 10⁷ CFU/mL. Thirty microliters of L. acidophilus and L. fermentum suspensions were individually inoculated into 3 mL of fresh MRS broth additionally containing 30 μL of C. perfringens suspension. Therefore, the ratio of Lactobacilli to C. perfringens was 1:1 in a 3.06 mL co-culture system. Another 30 μL suspensions of L. acidophilus, L. fermentum, and C. perfringens were individually added to the 3.03 mL of fresh MRS broth. This experiment was performed in triplicate. Cultures were anaerobically incubated at 37°C for 20 h. Following incubation, CFU of lactobacilli and C. perfringens were determined on MRS agar plates and tryptose-sulfite-cycloserine (TSC) agar plates, respectively. Meanwhile, the pH of cultures was measured.

The preparation of cell-free supernatant from lactobacilli cultures and agar well diffusion assay as well as broth culture inhibition assay was performed as previously described (Schoster et al., 2013), with some modifications. Briefly, lactobacilli suspensions were adjusted to an optical density of 0.220 at 600 nm and 1 mL of this suspension was inoculated into 9 mL of MRS broth. After 24 and 48 h of incubation, the lactobacilli cultures were centrifuged at 4,000 × g for 10 min and the supernatant was collected, filter-sterilized using a 0.22 μm membrane syringe filter and divided into two aliquots, one of which was neutralized to pH 6.20 using NaOH. Both aliquots were checked for absence of viable bacteria by plating on MRS agar and then frozen at −80°C for further use.

Overnight culture of C. perfringens was diluted 10-fold up to 10⁻² to achieve confluent growth. When the autoclaved TSC medium cooled down to about 55°C, 100 μL of bacterial dilution were mixed well with 20 mL TSC medium in a 10-cm plastic plate. Five 9-mm wells were made in each agar plate and 100 μL of lactobacilli supernatants were added to the four outer wells whereas sterile MRS broth was put into the central well (control well). When the effect of lactobacilli supernatants without neutralization was evaluated, MRS broth adjusted to pH 4.20 (using HCl) was added to the control well. When using neutralized supernatants, MRS broth at pH 6.20 was used instead. The presence of an inhibition zone >1 mm was assessed visually following 16 h of anaerobic incubation at 37°C. Each assay was performed in duplicate.

Fifty microliters of C. perfringens suspension were added to 5 mL of fresh MRS broth and 1 mL of lactobacilli supernatants harvested after 48 h of incubation. MRS broth at pH 6.20 and pH 4.20 was used as controls for the supernatant with and without pH neutralization, respectively. Growth was measured spectrophotometrically at 0, 24 and 48 h after anaerobic incubation at 37°C. The differences of OD₆₀₀ measurements between 0 h of incubation and 24 or 48 h of incubation were calculated. Each assay was performed in triplicate.

Fifteen milliliters of C. perfringens cultures in a stationary phase were centrifuged at 4,000 × g for 10 min and washed once with PBS and resuspended in the same volume of fresh MRS broth. This suspension was divided into three equal parts. Two of them were supplemented with the bacterial pellets of L. acidophilus and L. fermentum, respectively. The suspension without lactobacilli pellets was used as the control. After 4 h of anaerobic incubation at 37°C, the numbers of CFU of C. perfringens were determined on TSC agar plates and the pH was measured. The cultures were centrifuged at 4,000 × g for 10 min, and α-toxin level in the supernatants was analyzed.

As lactobacilli could greatly decrease the pH of cultures, we examined α-toxin production at normal and lower pHs. Pellets from stationary culture of C. perfringens were resuspended in MRS broth set at pH 5.20 or pH 6.20 using HCl and incubated anaerobically at 37°C for 2 or 4 h. After each time period, the numbers of CFU of C. perfringens were determined and the pH of each culture was measured. The supernatants of cultures were collected for α-toxin assay. All experiments were performed in triplicate.

Pellets from stationary culture of C. perfringens were resuspended in the same volume of fresh MRS broth and incubated anaerobically at 37°C for 4 h. The supernatant of this culture were collected by centrifugation (4,000 × g, 10 min) and divided into three equal parts. Pellets of L. acidophilus and L. fermentum cultures were individually resuspended in two parts of this supernatant and incubated anaerobically at 37°C for 2 or 4 h. The culture without lactobacilli pellets was used as the control. At each time point, the pH of each culture was measured and the supernatants of cultures were sampled for α-toxin assay. This experiment was performed in triplicate.

The content of C. perfringens α-toxin in the supernatants of cultures was determined using antigenic ELISA kit (Bio-X Diagnostics, Rochefort, Belgium), according to the manufacturers' instructions. Each sample and the control antigen (supplied by the kit) were tested in both positive and negative wells. The data were expressed as a percentage of delta optical density of each sample (between positive and negative wells) to that of the control antigen.

The methods of isolating primary intestinal epithelial cells from chicken embryos were previously described in detail by Guo et al. (2015). Isolated cells were grown in a 1:1 mixture of Dulbecco's modified Eagles medium and Ham's F12 medium (DMEM/F12) supplemented with 2.5% FBS, 100 U/mL penicillin, 100 μg/mL of streptomycin, 20 ng/mL of EGF, 100 μg/mL of heparin sodium salt, 5 μg/mL of insulin. Cells were maintained in a humidified environment with an atmosphere of 5% CO₂ at 37°C.

The assay of bacterial adhesion to cells was performed as previously described (Martin and Smyth, 2010) with some modification. The isolated cells were seeded in 12-well cell culture plates at a density of 1.0 × 10⁶ cells/mL with a total volume of 1 mL. After 48 h of incubation, cells were gently washed with pre-warmed PBS. One milliliter of DMEM/F12 medium containing L. acidophilus or L. fermentum with multiplicity of infection (MOI) at 10 was added to wells. Cells without lactobacilli were used as the control. Following 3 h of anaerobic incubation, a few microliters of resuspended C. perfringens was inoculated into each well (MOI = 1) and incubated at 37°C under anaerobic condition for 1 h. All experiments were performed in triplicate. After incubation, the cells were washed twice with PBS and 200 μL of 0.05% trypsin was added to each well and incubated for 15 min at 37°C. An additional 800 μL of PBS was added to each well and mixed gently by pipetting. The number of C. perfringens released per well was determined by plating serial 10-fold dilutions of the cell suspension onto TSC plates. The percentage of adhering bacteria was calculated by dividing the number of bacteria recovered from wells with the total number of bacteria added to the wells. Meanwhile, the remaining cell suspension were centrifuged at 4,000 × g for 10 min and the supernatant was collected for cytotoxicity quantification. Lactate dehydrogenase (LDH), an enzyme present in the cytoplasm of cells, is quickly resealed upon damage of the plasma membrane. The LDH production was determined using commercially available kit (Nanjing Jiancheng Biological Product, Nanjing, Jiangsu Province, China) following the manufacturer's protocol.

Cells were seeded in 6-well cell culture plates at a concentration of 1.0 × 10⁶ cells/mL (2 mL per well). At 90% confluence, cells were pretreated with either L. acidophilus or L. fermentum at MOI of 1 for 2 h, and then stimulated with and without C. perfringens (MOI = 0.1) for 6 h. The DMEM/F12 medium and C. perfringens infection alone served as the negative and positive controls, respectively. Experiments were run in triplicate. After stimulation, the supernatant of cell culture medium was collected for the assay of LDH activity and cells were lysed for RNA extraction.

The methods of RNA extraction and quantitative real-time PCR were described in our previous publication (Guo et al., 2015). Total RNA of the cells was extracted using TRIzol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions. The concentration and purity of total RNA was determined by measuring its optical density at 260 and 280 nm, and the RNA integrity was assessed via agarose gel electrophoresis. One microgram of total RNA was reverse transcribed with PrimeScript^TM RT Master Mix (Perfect Real Time) (TakaRa, Dalian, Liaoning Province, China). All cDNA preparations were stored frozen at −20°C until further use. The qRT-PCR analysis was performed with the 7,500 fluorescence detection system (Applied Biosystems, Foster City, CA) using the SYBR® Premix Ex Taq^TM (TaKaRa). The primer pairs for the amplification of chicken IL-6, IL-8, lipopolysaccharide-induced tumor necrosis factor-alpha factor (LITAF), inducible nitric oxide synthase (iNOS), IL-1β, NOD1, TLR2.2, TLR4, NF-κB p65, and β-actin cDNA fragments, used as an endogenous reference gene, were used as listed in Table 1. To confirm amplification specificity, the PCR products from each primer pair were subjected to a melting analysis and subsequent agarose gel electrophoresis. Samples were run in triplicate in 96-well plates. Relative gene expression data were analyzed using the 2^−ΔΔCt method (Heid et al., 1996).

All data were analyzed with SPSS Version 17.0 (SPSS Inc., Chicago, IL). For the data of inflammatory responses, the comparison between negative and positive controls was conducted using an independent-samples t-test. The other data were subjected to one-way analysis of variance (ANOVA) and Tukey test was used for multiple comparisons between treatments. Significance was accepted at P < 0.05 and results were reported as means and standard error (SE).

Compared with culture alone, co-culture of C. perfringens with either L. acidophilus or L. fermentum decreased the number of C. perfringens after 20 h of incubation (P < 0.01; Figure 1A), but did not significantly affect the enumeration of both lactobacilli (Figure 1B). Unsurprisingly, lactobacilli growth, alone or together with C. perfringens, reduced the pHs of cultures in comparison with individual C. perfringens culture (P < 0.01; Figure 1C).

The data of agar well diffusion assay showed that lactobacilli supernatants from 24 to 48 h-cultures exhibited inhibition zones to the growth of C. perfringens (Figure 2A). However, when the pHs of supernatants were neutralized, the inhibition zones disappeared. Furthermore, the MRS broth with pH adjusted to 4.20 also exerted inhibitory effect on C. perfringens growth. Consistently, addition of both lactobacilli supernatants and MRS broth (pH 4.20) to C. perfringens cultures significantly decreased the OD₆₀₀ values (P < 0.01), which reflected the number of C. perfringens, following 24 h of incubation (Figure 2B). As the incubation time reached 48 h, the lactobacilli supernatants still suppressed the OD₆₀₀ values of cultures compared with the control (MRS broth at pH 6.20) (P < 0.01), but no significant differences were observed between the control and MRS broth at pH 4.20 (Figure 2C).

Following 4 h of incubation, L. acidophilus and L. fermentum did not significantly influence the number of C. perfringens in the co-culture systems (Figure 3A), but decreased the α-toxin secretion compared with C. perfringens cultured alone (P < 0.01; Figure 3B). What's more, L. fermentum showed a greater inhibitory effect on α-toxin production than L. acidophilus. The pHs of cultures exhibited similar changes as α-toxin production. Both lactobacilli reduced the pHs of co-cultures and L. fermentum induced a further decrease of pH compared with L. acidophilus (P < 0.01; Figure 3C).

As shown in Figure 3D, the number of C. perfringens was not significantly affected by MRS broth at a lower pH (5.20) compared with normal pH (6.20) after 2 and 4 h of culture. However, the α-toxin secretion was impaired by the MRS broth with pH 5.20 following incubation (P < 0.01; Figure 3E). As expected, the MRS broth at pH 5.20 maintained a lower level of pH along the treatment (P < 0.01; Figure 3F).

As shown in Figure 4A, the addition of L. acidophilus and L. fermentum to the supernatants from C. perfringens cultures significantly reduced the levels of α-toxin after 2 and 4 h of incubation (P < 0.01). And L. fermentum resulted in a greater reduction of α-toxin compared to L. acidophilus at 2 h of treatment. During incubation, both lactobacilli significantly decreased the pH of C. perfringens supernatants in contrast to the control (P < 0.01) and lower pH was detected in the L. fermentum-treated group in comparison with L. acidophilus treatment (P < 0.01; Figure 4B).

Compared with C. perfringens infection alone, L. acidophilus pretreatment significantly decreased the percentage of C. perfringens adhering to chicken intestinal epithelial cells (P < 0.05), but L. fermentum did not show inhibitory effect on the C. perfringens adhesion (Figure 5A). The levels of LDH released from cells were not significantly changed after 1 h of incubation with C. perfringens alone or together with 3 h of lactobacilli pretreatments (Figure 5B). The LDH levels in the bacterial treatments were also not significantly different from that in the control (cell medium alone) (P > 0.05), but lower than that in the positive control, in which all cells were lysed by 0.5% of Triton X-100 (P < 0.01).

The data of LDH level and cytokine mRNA expression were shown in Figure 6. Compared with the uninfected control, C. perfringens infection at MOI of 0.1 for 6 h significantly increased the LDH level released from cells (P < 0.01), which indicated the induction of cytotoxicity by infection. However, pre-incubation of L. acidophilus and L. fermentum at MOI of 1 for 2 h reduced LDH levels in the infected cells (P < 0.01). Consistently, the relative mRNA expression of IL-6, IL-8, iNOS, LITAF and IL-1β was up-regulated by the C. perfringens challenge (P < 0.01), but down-regulated in challenged cells by pretreatment with Lactobacilli (P < 0.01). In the uninfected cells, both lactobacilli significantly increased the IL-1β mRNA level (P < 0.01), and L. acidophilus pre-incubation tended to elevate the mRNA expression of IL-6 (P = 0.061) and IL-8 (P = 0.089).

To investigate the possible mechanism mediating the modulation of cytokine expression by lactobacilli pretreatment, the mRNA expression of receptors (TLR2.2 and NOD1) and NF-κB p65 was determined and the data were presented in Figure 7. The C. perfringens infection significantly up-regulated the TLR2.2 mRNA expression (P < 0.05) and tended to increase the mRNA levels of NOD1 (P = 0.057) and NF-κB p65 (P = 0.073). However, the relative expression of TLR2.2, NOD1, and NF-κB p65 in the infected cells was reduced by the preincubation of lactobacilli (P < 0.01). There was a tendency that L. acidophilus decreased TLR2.2 mRNA expression in the uninfected cells (P = 0.057).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.