One-day-old mixed sex Cobb broilers (n = 120) purchased from a local producer (Curtis chicks, a division of Maple Lodge hatcheries, Guelph, Ca) were group housed throughout the experiment in specific-pathogen-free filtered air positive pressure rooms on floor pens with wood shaving. All chickens had ad libitum access to water and commercial feed. All animal experiments were approved (028–10,783 – ISOL and AUP 4328) by the Animal Care Committee of the University of Guelph and adhered to the guidelines for the use of animals. Experiments and analyses were performed according to the ARRIVE guidelines.

One-hundred milligrams of duodenum, jejunum and ileum intestinal contents, as collected on a weekly basis (24 h post rL. lactis inoculation) from five to six chickens, were dilute in 0.5 mL of ice-cold PBS. Mixtures were thoroughly vortexed and subsequently centrifuged (400 × g for 5 min). Supernatants were collected and tenfold serial dilution were performed in PBS. One hundred microliters of the dilution series were plated in triplicates on M17 agar supplemented with 5 μg/mL of chloramphenicol (Millipore-Sigma, Canada). Colonies were allowed to grow for 24 h in an anaerobic atmosphere (30°C and no shaking using a gas pack). The next day, colonies were counted to estimate recovery and persistence of rL. lactis.

Five-centimeter segments of the medial duodenum, jejunum, ileum and whole cecal tonsils were harvested from chickens and stored on ice in PBS containing penicillin (10 U/mL), and streptomycin (10 μg/mL). Each tissue was cut into 0.5 cm² segments and washed vigorously three times with PBS containing penicillin (10 U/mL), and streptomycin (10 μg/mL). Tissue samples were subsequently digested (collagenase type 1; 80 U/mL at 37°C for 20 min; Millipore-Sigma, Canada) in PBS containing penicillin (10 U/mL), and streptomycin (10 μg/mL). Whole tissue digests were applied onto 40-μm BD cell strainers (BD Biosciences, Canada), and crushed through using the rubber end of a 10 mL syringe plunger (Boodhoo et al., 2016). Duodenum, jejunum, ileum and cecal tonsils cell suspension were prepared by layering (2:1) onto histopaque 1,070 (Millipore-Sigma, Canada) density-gradient centrifugation and centrifuged at 2100 rpm (600 × G) for 20 min to allow the separation of mononuclear cells. Aspirated buffy coats were washed at 1500 rpm (400 × G) for 5 min in RPMI 1640 with penicillin (10 U/mL), and streptomycin (10 μg/mL). Mononuclear cells were suspended in complete RPMI cell culture medium; RPMI 1640 medium containing 10% fetal bovine serum (Millipore-Sigma, Canada), penicillin (10 U/mL), and streptomycin (10 μg/mL). Cell number and viability were calculated using a hemocytometer, and trypan blue exclusion method. Mononuclear cells were suspended in complete RPMI cell culture medium at a density of 5 × 10⁶ cells/ml and kept on ice.

Following a wash in FACS staining buffer (PBS with 0.2% BSA), duodenum, jejunum, ileum and cecal tonsils mononuclear cells, 5.0 × 10⁵ cells per well in triplicates, were counter stained in specific panels as APC (mouse anti-Kul01-FITC and mouse anti-major histocompatibility complex (MHC) II-PE), T cell (mouse anti-CD3ζ-PB, mouse anti-CD4-PE-CY7, mouse anti-CD8α-FITC and mouse anti-γδTCR-PE) and B cell (mouse anti-Bu1-PB, mouse anti-IgA-PE, mouse anti-IgY-FITC, mouse anti-IgM-APC) for 15 min at 4°C in FACS staining buffer. Cells were washed (400 × g for 5 min) with FACS staining buffer and incubated for a further 10 min at 4°C with 7-AAD-PE (BDTM Pharmigen, Canada). Mononuclear cells were washed and stored in 2% paraformaldehyde (PFA) in the fridge (4°C). 5.0 × 10⁵ cells were acquired on a FACS BD Canto II flow cytometer and data were processed by FlowJo V10 software. The full list of antibodies utilized is available in Table 2.

Duodenum, jejunum and ileum intestinal microbial genomic DNA extraction was performed using QIAamp® Fast DNA Stool Mini Kit (Qiagen, Canada) as recommended by the manufacturer. Briefly, 1 mL of InhibitEX buffer was added to 200 mg of duodenum, jejunum and ileum content. After a brief vortex (1 min) and centrifugation (20,000 × g for 1 min), supernatants were collected and added to tubes containing 25 μL of Proteinase K. AL buffer was added to the mixtures and incubated for 10 min at 70°C. Following the addition of 600 μL of 100% Ethanol, lysate-mixtures were briefly vortexed (1 min). The lysate-mixtures were applied to the spin columns, centrifuged (20,000 × g for 1 min) and washed with AW buffers in 2 steps. Eluted DNA (20,000 × g for 1 min) were quantified using a NanoDrop® ND-1000 spectrophotometer (NanoDrop Technologies, Wilmington, DE) and adjusted to 40 ng/μl for SYBR green qRT-PCR. Plates were read and analyzed by Light Cycler 480 II (Roche Diagnostics GmbH Mannheim, Germany). Total bacterial population was determined relative to the universal primers over control samples. Primer sequences and annealing temperatures are outlined in Table 2.

Graph Pad Prism 8 for windows was utilized to generate graphs and perform statistical analysis. All data are presented as mean + SD and analyzed by Wilcoxon and Mann Whitney non-parametric as well as two-way ANOVA was used to test significance. Results were considered statistically significant at p < 0.05 (*).

Lactococcus lactis bacteria were transformed with the pNZ8142 plasmid (rL. lactis) to enable detection and enumeration in intestinal contents post inoculation. The experimental timeline is illustrated in Figure 1A. The result demonstrates that chickens infected with C. perfringens had lower NE lesions if they had been orally inoculated with a rL. lactis (10*8 CFU; n = 10) compared to the PBS-treated infected group (n = 10) (Figure 1B). Lesion scoring, as listed in Table 1, was performed from the proximal end to the distal end of the small intestine (Figure 1C). Inoculation of rL. lactis prior to infection with C. perfringens inhibited fibrin formation, necrotic foci and ulcers. However, erythematous mucosa was still evident in these birds. It should be noted that chickens receiving rL. lactis or PBS without infection with C. perfringens showed no intestinal lesions. The quantification of rL. lactis in intestinal contents, collected from at least 6 chicks from each group (24 h post oral gavage), demonstrated that rL. lactis pass through and can survive in the medial duodenum (Figure 1D), jejunum (Figure 1E) and ileum (Figure 1F). Positive bacterial colonies were considered as the rL. lactis carrying the pNZ8124 plasmid providing resistance to chloramphenicol. Taken together, the results indicate that the presence of rL. lactis in the intestinal contents is associated with a reduction in necrotic lesions induced by C. perfringens.

To examine any possible association between necrotic lesions and the abundance of microbial groups, intestinal contents (n = 6 per time points) were analyzed to determine phylogenetic relations in the rL. lactis treated chickens compared to control/PBS treated chickens prior to and post C. perfringens infection (Figure 2). Temporal changes in microbial relative content were used to estimate the abundance of specific bacterial phylum: Bacillota (Firmicutes), Bacteroidota (Bacteroides), Lactobacillaceae (Lactobacillus), Pseudomonadota (γ-proteobacteria) and Actinomycetota (Actinobacteria) to universal primers at 36 h post oral gavage with rL. lactis. The assessment of intestinal contents indicated that a diverse microbial community was established and could be modified by rL. lactis treatment. The results demonstrate site specific and temporal increases (p ≤ 0.01) in the Bacteroidota, Bacillota, and Lactobacillaceae prior to infection. There was a significant increase in the relative abundance of Bacteroidota (Figure 2A), Bacillota (Figure 2B), and Lactobacillaceae (Figure 2C) in jejunum (p ≤ 0.01) and ileum (p ≤ 0.01) at week 2 and 3 compared to that observed in duodenum. However, Actinomycetota (Figure 2D) relative number was significantly (p ≤ 0.005) more abundant in jejunum and ileum content at week 3 in comparison to that in duodenum. No changes in Pseudomonadota were observed after rL. lactis treatment alone in comparison to the control group (Figure 2E). Post C. perfringens infection, treatment with rL. lactis led to a decrease in the Bacillota but an increase in Bacteroidota, Lactobacillaceae, Pseudomonadota and Actinomycetota. The results also demonstrate that infection alone with C. perfringens (day 24) led to a significant (p ≤ 0.01) increase in duodenum, jejunum and ileum abundance of the Bacillota (Figure 2B), Actinomycetota (Figure 2D) and Pseudomonadota (Figure 2E). However, prior treatment with rL. lactis and subsequent infection with C. perfringens (day 24) resulted in a significant (p ≤ 0.005) decrease in duodenum, jejunum and ileum abundance of the Bacillota (Figure 2B). In contrast, prior treatment with rL. lactis and subsequent infection with C. perfringens (day 24) resulted in a significant (p ≤ 0.005) increase in duodenum, jejunum and ileum intestinal content of the Bacteroidota (Figure 2A), Lactobacillaceae (Figure 2C), Actinomycetota (Figure 2D) and Pseudomonadota (Figure 2E).

To identify the effects of rL. lactis treatment on modulation of mucosal immune parameters, we analyzed induction of innate responses, such as temporal expression of IL-6, and IL-8 to β-actin as well as frequency of intestinal mononuclear phagocytes. Treatment with rL. lactis prior to infection with C. perfringens led to a time- and site-specific dependent increase in expression of IL-6 (Figure 3A), and IL-8 (Figure 3B). Over time (week 1 vs. week 3), IL-6 expression was significantly decreased in the duodenum and jejunum of rL. lactis treated chickens (Figure 3A). There was a significant (p ≤ 0.05) decrease in expression of IL-6 (Figure 3C), and IL-8 (Figure 3D) in the duodenum and jejunum of chickens that were pre-treated with rL. lactis and infected with C. perfringens in comparison to C. perfringens only infected chickens. In contrast, infection with C. perfringens led to a significant increase (p ≤ 0.05) in expression of IL-8 (Figure 3D) when comparing to rL. lactis treated only and rL. lactis and C. perfringens infected groups in the ileum.

As IL-6 and IL-8 can be produced by mononuclear phagocytes, we analyzed the frequencies of Kul01 + MHC-II+ cells in different experimental groups using flow cytometry (Figure 4A). The results demonstrate a significant increase in frequency of the Kul01 + MHC-II+ cells, following rL. lactis treatment, in the medial duodenum (p ≤ 0.0001), and jejunum (p ≤ 0.05) but not in the ileum and cecal tonsil when compared to the control chickens (Figure 4B). Infection with C. perfringens led to a significant decrease in frequency of the Kul01 + MHC-II+ cells in duodenum (p ≤ 0.05) and jejunum (p ≤ 0.05) compared to the control or the rL. lactis treated and infected chickens (Figure 4C). There was a significant (p ≤ 0.0001) increase in frequencies of Kul01 + MHCII+ cells in ileum of the C. perfringens infected chickens compared to the control or the rL. lactis treated only groups (Figure 4C). The increase in mucosal Kul01 + MHC-II+ cells was sustained post infection with C. perfringens (Figure 4C).

rL. lactis treatment prior to infection with C. perfringens led to temporal and spatial changes in expression of IL-13 (Figure 5A), IL-17 (Figure 5B), IL-18 (Figure 5C), IL-22 (Figure 5D), IL-12p40 (Figure 5E) and IFN-γ (Figure 5F) to β-actin from week 1 onwards in all sites along the intestine. IL-17 (Figure 5B), and IFN-γ (Figure 5F) mRNA transcripts were significantly (p ≤ 0.01) upregulated in the jejunum of rL. lactis treated group but not in the duodenum and ileum. In contrast, IL-13 (Figure 5A) and IL-18 (Figure 5C) mRNA transcripts were significantly (p ≤ 0.05) upregulated in the ileum of rL. lactis treated group. There was a significant (p ≤ 0.05) decrease in expression of intestinal IL-13 (Figure 5G) and IL-18 (Figure 3I) in chickens that were pre-treated with rL. lactis and subsequently infected with C. perfringens in comparison with C. perfringens only infected chickens. Post- C. perfringens infection, there was a significant increase in intestinal IL-13 (p ≤ 0.001) (Figure 5G) and IL-18 (p ≤ 0.05) (Figure 5I) expression in both rL. lactis treated only and rL. lactis treated and subsequently C. perfringens infected chickens compared to control chickens. Expression of IL-17 was significantly decreased (p ≤ 0.001) in the ileum only and not in duodenum and jejunum in both rL. lactis treated only and rL. lactis treated and subsequently C. perfringens infected chickens compared to control chickens (Figure 5H). Both IL-12p40 (Figure 5K) and IFN-γ (Figure 5L) intestinal mRNA transcripts were significantly increased (p ≤ 0.001) in rL. lactis treated and C. perfringens infected chickens when compared to C. perfringens only infected chickens. No change in IL-22 (Figure 5J) was observed post-infection with C. perfringens.

Alterations in intestinal cytokines can affect intestinal cellular composition, such as abundance of T cells. FACS analysis was used to characterize intestinal mononuclear cells along the intestinal tissue (Figure 6A). The frequencies of CD3ζ + CD8α + γδ T cells within CD3ζ + γδ T cell population are presented. In control chickens, the ileum was found to have higher frequencies of CD3ζ + CD8α + γδ T cells in comparison with the duodenum, jejunum and cecal tonsils. The results also demonstrate that treatment with rL. lactis led to a significant increase in the frequency of CD3ζ + CD8α + γδ T cells in the duodenum (p ≤ 0.05), ileum (p ≤ 0.0001) and cecal tonsils (p ≤ 0.05) as assessed by two-way ANOVA when compared to control chickens (Figure 6B). Infection with C. perfringens also led to a significant increase in the frequency of CD3ζ + CD8α + γδ T cells in duodenum (p ≤ 0.01), ileum (p ≤ 0.05) and cecal tonsils (p ≤ 0.005) (Figure 6C). Similarly, rL. lactis treatment followed by infection with C. perfringens increased the frequency of these cells in duodenum (p ≤ 0.005), jejunum (p ≤ 0.0001) and cecal tonsils (p ≤ 0.01) CD3ζ + CD8α + γδ T cells when compared to control chickens (Figure 6C).

The frequencies of αβ T cells were also assessed by FACS analysis. A negative gating strategy, CD3ζ + γδ- T cells, was used to identify CD3ζ + αβ + T cells that are either express CD4+ or CD8α + (Figure 7A). CD3ζ + CD8α + αβ T cells (Figure 7D) made up the majority of the αβ T cells and were more abundant (on average 2:1 ratio) than CD3ζ + CD4+ αβ T cells (Figure 7B). The abundance of intestinal tissue CD3ζ + CD8α + αβ T cells was further amplified following a prolonged treatment with rL. lactis. The results demonstrate that treatment with rL. lactis led to a significant increase in the frequency of CD3ζ + CD8α + αβ T cells in the duodenum (p ≤ 0.01) and ileum (p ≤ 0.01) when compared to the control chickens (Figure 7D). In contrast, treatment with rL. lactis led to a significant decrease in the frequency of intestinal CD3ζ + CD4+ αβ T cell when compared to the control chickens (Figure 7B). Infection with C. perfringens also altered the frequency of CD4+ and CD8α + αβ T cells. Infection with C. perfringens resulted in a significant decrease in the frequency of CD3ζ + CD4+ αβ T cells across all segments of the intestine (Figure 7C). Conversely, infection significantly increased the frequency of CD3ζ + CD8α + αβ T cells across all segments of the intestine when compared to the rL. lactis treated and infected chickens (Figure 7E). However, the combination of rL. lactis treatment and subsequent C. perfringens infection resulted in the increase of both CD3ζ + CD4+ (Figure 7C) and CD3ζ + CD8α + (Figure 7E) αβ T across the intestine when compared to the infected or control chickens, respectively.

In addition to T cells, B cells were also characterized in the present study (Figure 8). As expected, three major B cell subsets expressing either IgM, IgY or IgA immunoglobulins were found in the intestine (Figure 8A). The intestine was mainly populated by IgA+ (Figure 8B) followed by IgY+ B cells (Figure 8C) while IgM+ B cells (Figure 8D) made up a minority portion of duodenum, jejunum and ileum Bu1+ B cells. In contrast, the cecal tonsils are mainly populated (p ≤ 0.0001) by IgM+ B cells (Figure 8D) followed by IgY+ (Figure 8C) and IgA+ B cells (Figure 8C). The results demonstrated that repeated treatment with rL. lactis over several weeks resulted in the expansion of B cells that are IgA+ within the duodenum (p ≤ 0.05), jejunum (p ≤ 0.005) and ileum (p ≤ 0.0001) and IgY+ within the jejunum (p ≤ 0.0001) and ileum (p ≤ 0.0001) when compared to control chickens. There were no changes in the frequency of IgM+ B cells within the duodenum, jejunum and ileum as a result of rL. lactis.

Infection with C. perfringens resulted in further expansion of IgA+, IgY+ and IgM+ Bu1+ B cells across the intestine (Figure 9). The results demonstrate that infection with C. perfringens led to a significant increase in the frequency of both IgA+ (p ≤ 0.0001) (Figure 9B) and IgY+ (p ≤ 0.0001) (Figure 9C) Bu1+ B cell within the duodenum, jejunum and cecal tonsils when compared to the control chickens. The combination of prolonged rL. lactis pre-treatment followed by infection with C. perfringens resulted in a significant increase of both IgA+ (p ≤ 0.0001) (Figure 9B) and IgY+ (p ≤ 0.0001) (Figure 9C) Bu1+ B cell frequencies within the duodenum, jejunum and ileum when compared to control chickens. The frequency of IgM+ Bu1+ B cell frequency was significantly increased in the jejunum (p ≤ 0.005) and ileum (p ≤ 0.0001) of chickens that were previously treated with rL. lactis and subsequently infected with C. perfringens or in C. perfringens only infected chickens (Figure 9D).

To determine the immune regulatory effects of rL. lactis, temporal expression of IL-10, TGF-β, CTLA-4 and CD40 was assessed to β-actin (Figure 10). Prior treatment with rL. lactis led to an increase in expression of CD40 (Figure 10A), TGF-β (Figure 10B), IL-10 (Figure 10C) and CTLA-4 (Figure 10D) in the duodenum, jejunum and ileum. Infection with C. perfringens in those chickens that had prior treatment with rL. lactis resulted in a significant (p ≤ 0.01) increase in CD40 expression in comparison to C. perfringens infected only chickens (Figure 10E). CD40 (Figure 10E) expression levels were significantly (p ≤ 0.01) decreased in the duodenum, jejunum and ileum post infection with C. perfringens. However, CTLA-4 expression was only decreased in the ileum post infection with C. perfringens (Figure 10H). IL-10 mRNA expression level was significantly (p ≤ 0.01) decreased in chicken pretreated with rL. lactis and subsequently infected with C. perfringens, similar to rL. lactis treated only chickens, when compared to infected only chickens (Figure 10G). In contrast, infection with C. perfringens resulted in a significant (p ≤ 0.01) increase in IL-10 mRNA transcripts (Figure 10G) in all intestinal segments studied. No changes in induction of TGF-β mRNA transcripts was observed (Figure 10F).