Protection against necrotic enteritis in broiler chickens by co-administration of Enterococcus faecalis EF32 and recombinant Lactobacillus plantarum strain co-expressing FimA, FimB, NetB and Fba

Abstract

Necrotic enteritis (NE), caused by Clostridium perfringens(C. perfringen) type G carrying the NetB gene, represents a major infectious disease threatening the global poultry industry. Following the ban on in-feed antibiotic growth promoters, the incidence of NE has increased markedly, highlighting the urgent need for safe and effective prevention and control strategies. To address this challenge, we developed a novel probiotic-based oral immunization strategy.We constructed a recombinant Lactobacillus plantarum (L. plantarum) NC8 strain surface-displaying the FimB antigen [NC8(pYL515)] and orally co-immunized broiler chickens with the previously developed NC8(pYL590) strain co-expressing FimA, FBA, and NetB, together with Enterococcus faecalis EF32, to evaluate their protective efficacy against NE in broilers. In vitro, EF32 demonstrated a strong ability to promote the maturation of chicken bone marrow-derived DCs, as evidenced by the upregulation of CD83, CD86, CCL5, and CCR7 expression. Prior to challenge, the combination of live EF32 and multi-antigen recombinant strains established an enhanced systemic immunity (significantly elevated IgG), mucosal immunity (early activation of IgM and marked increase in sIgA) and cellular immunity (enhanced splenic lymphocyte proliferation and increased expression and secretion of IFN-γ and IL-4). In addition, the co-administration also significantly upregulated the mRNA levels of intestinal CCR7, CCR9, and CD86. Finally, in a broiler challenge model, this combined immunization strategy significantly alleviated intestinal lesions and mitigated growth suppression. Notably, live EF32 bacteria, as a key component exerting synergistic effects in this combined immunization strategy. Our findings demonstrate that combined immunization with recombinant L. plantarum co-expressing multiple antigens and live EF32 synergistically enhances the immunogenicity and protective efficacy of an oral vaccine, offering a promising antibiotic-free strategy for NE control.

Introduction

Necrotic enteritis (NE), an acute and often fatal intestinal disease of poultry caused by Clostridium perfringens (CP), damages intestinal cells and disrupts the intestinal ecosystem, leading to dysbiosis (Abd El-Hack et al., 2022). C. perfringens can be classified into multiple toxinotypes based on the different toxins it carries, among which type G C. perfringens has been identified as one of the major pathogens responsible for NE in broilers (Gu et al., 2019; Sarmah et al., 2021). The pathogenesis of NE is not determined by a single factor but is closely associated with multiple factors including co‑infection with coccidia, dietary nutrients, intestinal barrier integrity, microbial balance, and immune status (Waller et al., 2024). Clinical NE is characterized by acute onset and high mortality (up to 50%) (Song et al., 2022), whereas subclinical NE does not cause obvious clinical signs or death but leads to reduced feed intake, decreased weight gain, intestinal microbial dysbiosis, and significantly impaired feed conversion efficiency (Liu et al., 2025). With the increasing intensification of the poultry industry in China and the comprehensive implementation of antibiotic restriction policies, the global incidence of NE continues to rise, causing huge economic losses to poultry production (Emami and Dalloul, 2021). This has created an urgent need for the development of safe and effective antibiotic‑free strategies for the prevention and control of NE in the current poultry industry.

Currently, vaccination is considered the optimal strategy for controlling NE (Mot et al., 2014). Keyburn et al. first identified the NetB toxin and demonstrated its close association with avian NE, a discovery that shifted NE vaccine research from traditional whole-cell or toxoid strategies toward targeted immunization designs directed against key virulence factors (Keyburn et al., 2006). However, worldwide, only two vaccines based on CPA and NetB toxins have been commercialized, and they provide only partial protection (Waller et al., 2024). It can thus be observed that a single toxin antigen is insufficient to fully cover the complex pathogenic process of NE. In recent years, researchers have found that, in addition to NetB, antigens associated with adhesion, colonization, and surface structures, such as Fba, FimA, and FimB, also possess high application potential (Lepp et al., 2019; Wilde et al., 2019). Among these, FimB, as a core component protein of the pilus of type G C. perfringens strains, has important role in both NE pathogenesis and vaccine development (Lepp et al., 2019). FimB is a subunit of the adhesive pilus associated with NE pathogenicity, and immunization against pilus subunits can alleviate NE lesions to some extent (Lepp et al., 2019). Moreover, multi-component antigen immunization results in better post‑challenge survival and lower lesion scores (Yuan et al., 2022). Collectively, multiple studies indicate that the current development of NE vaccines can simultaneously cover multiple pathogenic links, including toxin neutralization, adhesion blockade, colonization intervention, and local immune activation, gradually shifting from a "single toxin neutralization" model toward a "multi‑antigen, multi‑mechanism combined protection" model. By targeting different links in the pathogenic chain through multiple targets, more comprehensive and durable immune protection can be achieved (Shamshirgaran and Golchin, 2025; Wang et al., 2022).

An ideal NE vaccine should incorporate multiple antigens targeting different mechanisms, molecular adjuvants, and be delivered via in ovo or oral routes (Waller et al., 2024). NE mainly occurs on the intestinal mucosa, and traditional parenteral immunization may not sufficiently induce effective local mucosal immunity (Waller et al., 2024), whereas oral vaccines face challenges such as antigen inactivation in the gastrointestinal tract, low delivery efficiency, and weak immunogenicity (De et al., 2024). Lactobacillus plantarum (NC8) is recognized as an ideal mucosal delivery vector, possessing excellent intestinal colonization ability and immunomodulatory functions (Qiao et al., 2021). It can efficiently deliver heterologous antigens and induce both local intestinal IgA and systemic IgG responses, making it particularly suitable for the prevention and control of intestinal infectious diseases (Cano-Garrido et al., 2015; Szatraj et al., 2017). In particular, the surface displayed system such as pgsA' anchoring sequence which could present antigens on the surface of NC8 directly and elicit specific mucosal immunity in the host (Cai et al., 2016), which exhibits stronger antigen presentation efficiency and immunogenicity compared with intracellular expression (Lee et al., 2006), thereby providing a reliable technological platform for multi-antigen co-delivery. In addition, some other surface anchoring sequences including Lp1261(Nguyen et al., 2016) and LysM (Xu et al., 2011) were also employed extensively in the design of probiotics based vaccine. To note, we have previously constructed a recombinant L. plantarum strain (pYL590) co-expressing FimA, FBA and NetB antigens through the above mentioned three anchoring sequences, with the properties of chicken dendritic cell targeting and lysosome escape, which could significantly decrease the disease rate to 50% after NE challenge (Gao et al., 2026).

On the other hand, in poultry production, some probiotic candidates such as Enterococcus faecalis (E. faecalis) strains have been demonstrated to play important roles in immune response, modulation of intestinal microbiota, enhancement of intestinal barrier function, and improvement of growth performance (Hussain et al., 2024; Shehata et al., 2020). In our previous studies, we have successfully isolated a strain named E. faecalis YL-EF32 from the intestinal tract of healthy wild birds. Supplementation with E. faecalis YL-EF32 in a broiler NE model significantly alleviated NE-induced body weight loss and intestinal lesions, and significantly upregulated the expression of the tight junction protein-related gene Occludin and ZO-1, while also improving the intestinal microbial composition under NE conditions (Zhang et al., 2025a). Since that the administration of E. faecalis strain has been indicated as an efficient stimulator to enhance trained immunity by the presence of ribosomal protein S11, resulting in enhanced antigen presentation and improved influenza vaccine efficacy through TLR4-TET2 signaling-mediated ribosomal biogenesis inhibition, thereby shaping the enhanced MHC molecule expression phenotype in trained antigen-presenting cells (Zhang et al., 2025b). Therefore, our hypothesis was that whether the co-administration of E. faecalis strain in combination with recombinant L. plantarum strains could yield a better protection against NE challenge.

In this study, we firstly constructed a recombinant L. plantarum strain (pYL515) synthesizing FimB antigen on the surface, then the combination of L. plantarum strain (pYL515), L. plantarum strain (pYL590) and E. faecalis YL-EF32 were employed for oral administration in a broiler model to evaluate its immune response and protection against NE challenge. The results showed that the alive, but not inactivated E. faecalis YL-EF32 strain, could dramatically elevate the specific humoral and intestinal antibody production, yielding a significantly increased protection against NE challenge.

Materials and methods

Bacterial strains and plasmids

The bacterial strains and plasmids used in this study are listed in Table 1, E. faecalis YL-EF32 was isolated previously in our lab (Zhang et al., 2025a), Escherichia coli (E. coli) TOP10 for molecular cloning, Lactiplantibacillus plantarum (L. plantarum) NC8 as the live vaccine delivery vehicle, and a pathogenic C. perfringens type G (CPG) strain CP4 (kindly provided by Professor John F Prescott, University of Guelph, Canada) (Jiang et al., 2009) for subsequent challenge. Specifically, E. coli (TOP10) was cultured in LB broth at 37°C with shaking (220 rpm), E. faecalis and L. plantarum (NC8) strains were grown statically in MRS broth at 37°C, and the CP4 strain was maintained anaerobically in fluid thioglycollate medium at 37°C.Where applicable, the above media were supplemented with different antibiotics: 50 µg/mL chloramphenicol was added to the LB medium for E. coli (TOP10), and 10 µg/mL erythromycin was added to the MRS medium for L. plantarum (NC8). Plasmid pYL590 was previously constructed in our lab to synthesize FimB, Fba and NetB antigens on the surface of L. plantarum strain, with the ability of chicken DC targeting and lysosome escape.

Table 1.

Bacterial strains and plasmids.

Materials	Description	Source
Bacteria strains
E.coli Top10	E.coli host strain for regular cloning	Shenggong, China
E. faecalis YL-EF32	Isolated E. faecalis strain from wild bird	Lab stock(1)
L.plantarum NC8	Host strain, Plasmid-free, silage isolate	Thea Aukrust (1992)
Plasmids
pYL404	Synthesized pgsA’-fimB fragment, codon optimized for L. plantarum	Genewiz, China
pYL560	P23 promoter, prokaryotic vector in L. plantarum	Lab stock(2)
pYL590	pSIP409-based prokaryotic expression vector, express PgsA-FimB,LysM-Fba,LP1261-NetB anchoring protein.	Lab stock
pYL515	pSIP409-based prokaryotic expression vector (containing the PgsA’ anchoring sequence), expressing the protein FimB.	This study

Construction and expression of recombinant L. plantarum strains

Plasmid pYL560 which was a prokaryotic vector in L. plantarum with a P23 promoter was used as a parental vector. The pgsA’-fimB gene fragment was synthesized by Genewiz (China) for codon optimization in L. plantarum strain, then primers FimB-F-KpnI(GAGGTACCGAAAACACCGCGAACATTC)/ FimB-F-HindIII(CCCAAGCTTTTAGTTGGTATCGCCGGT) were used to amplify the pgsA’-fimB fragment and inserted into pYL560 by double enzyme digestion and T4 ligation, yielding pYL515 which synthesize surface displayed FimB antigen driven by a constitute P23 promoter. The confirmed plasmids from E. coli clones were then electroporated into competent L. plantarum NC8 cells (2.5 kV, 200 Ω, 25 μF) and spread onto MRS agar plates containing 10 µg/mL erythromycin and incubated anaerobically at 37°C for 48 hours for further selection. Finally, the recombinant L. plantarum NC8 (NC8/pYL515) was successfully isolated, and plasmid carriage of the FimB antigen was confirmed by colony PCR.

To verify the expression of the recombinant plasmid pYL515 in host strains, the recombinant L. plantarum NC8(pYL515) were inoculated into MRS medium containing erythromycin and grown to mid-log phase (OD600 ≈ 0.9) under static conditions at 37°C. Bacterial cells were then collected by centrifugation and the pellet was resuspended in suspension buffer (50 mM Tris, 1 mM EDTA, 6.7% sucrose) containing protease inhibitors and lysed by adding an equal volume of lysis buffer (200 mM NaOH, 1% SDS). Following incubation, the lysate was centrifuged and the supernatant was recovered and subjected to western blotting. In addition, the cell wall components were also extracted as described previously (Cai et al., 2016) and analyzed by western blotting. For immunoblotting, the protein marker and samples were separated by 12.5% SDS-PAGE and transferred onto nitrocellulose membranes. The membranes were then blocked with 5% skim milk in TBST at 4°C overnight. After blocking, the membranes were incubated with mouse anti-FimB polyclonal antibody (1:500) (lab stock) for 60 min at room temperature. Following washing, the membranes were incubated with HRP-conjugated goat anti-mouse IgG secondary antibody (TransGen, 1:5,000) for 60 min at room temperature. After incubation, signals were visualized using enhanced chemiluminescence (ECL) and captured with a chemiluminescence imaging system (GE Healthcare). Additionally, the synthesized FimA, Fba and NetB antigens by L. plantarum NC8(pYL590) were also confirmed using individual specific antibodies as described above.

Verification of anchored expression of FimB in recombinant L. plantarum NC8(pYL515)

Recombinant L. plantarum NC8(pYL515) cells from the logarithmic growth phase were harvested, washed, and fixed with 4% paraformaldehyde for 30 min at room temperature. After removal of the fixative by washing, the cells were resuspended in blocking buffer containing 3% BSA and incubated for 1 h at 37°C. The blocking buffer was then discarded, and the cells were incubated with diluted FimB polyclonal antibody at 4°C overnight (untransformed L. plantarum NC8 was used as a negative control). After washing to remove unbound primary antibody, the cells were incubated with FITC-conjugated goat anti-mouse IgG antibody(TransGen, 1:500) for 1 h at 37°C in the dark, followed by extensive washing and resuspension. A small volume of the resuspended cells was dropped onto a glass slide and observed under a microscope.

Chickens immunization and immunological analysis

All animal experiments in this study were approved by the Institutional Animal Care and Use Committee (IACUC) of Jilin Agricultural University (Approval No. [2025 10 10 008]), and were performed in strict accordance with the Guide for the Care and Use of Laboratory Animals. All experimental chickens were fed an antibiotic-free basal diet and randomly divided into six groups, including the L. plantarum NC8 group, the E. faecalis EF32 group, the NC8(515+590) co-immunization group, the deadEF32+NC8(515+590) group, the EF32+NC8(515+590) group, and the PBS group. Each group was housed separately in isolation (n=20/group). Additional 10 birds were also included without feeding probiotics and challenge as a normal control.

Immunization was initiated at 7 days of age as the primary immunization baseline, followed by a booster immunization 14 days later (at 21 days of age). Both immunizations were administered orally for three consecutive days, with a dose of 10⁹ CFU per bird per day. For the co-immunization groups, the two recombinant strains and the corresponding EF32 were mixed in equal proportions prior to oral administration. On days 14 (35 days of age) post boost immunization, four broilers from each group were randomly selected and euthanized for sample collection. Antigen-specific antibody levels in serum and intestinal lavage fluid were measured by ELISA as described previously (Jiang et al., 2015). Splenic lymphocyte proliferation activity was assessed using the CCK-8 assay. The secretion levels and mRNA transcription levels of IFN-γ and IL-4 in jejunal tissues and spleen were determined by ELISA and qRT-PCR, respectively. The production of sIgA in the jejunum and ileum sections were also determined by immunofluorescence analysis. In addition, jejunal tissues from the deadEF32+NC8(515+590) and EF32+NC8(515+590) groups were also subjected to transcriptome sequencing and qRT-PCR were performed to validate the relative mRNA expression levels of CCR9, CD28, and CD86.

NE challenge study

To evaluate the preventive protective efficacy induced by the vaccine, a challenge protection trial was initiated at 2 weeks post-boost (35 days of age). All birds were orally administered 10⁴ oocysts of Eimeria tenella at 32 days of age and switched to a high-protein induction diet (containing 36% fish meal) to predispose intestinal infection, followed by oral challenge with G-type C. perfringens CP4 strain at the dose of 4 × 10⁸ CFU/bird for five consecutive days. During the entire bacterial challenge period, clinical signs, mortality, and body weight data were recorded daily. The protective efficacy of the co-immunization was systematically evaluated by NE clinical lesion scoring and H&E staining of the jejunum. In detail, the isolated jejunal tissues were gently flushed with ice-cold PBS to remove intestinal contents, and approximately 1 cm segments were excised. After fixation and dehydration, the tissues were cleared in xylene, embedded in paraffin, and sectioned at a thickness of 4 μm. Following baking, the sections were stained, dehydrated through a graded ethanol series, cleared in xylene, and mounted with neutral balsam. Finally, histopathological changes in intestinal tissues were observed under a light microscope, and images were captured for analysis.

Enzyme-linked immunosorbent assay (ELISA) to determine antigen specific antibody titers

On days 14 post-boost immunization, serum and intestinal lavage fluid samples were collected (n = 4 per group) as described previously (Jiang et al., 2015). Intestinal samples were homogenized in ice-cold lavage buffer (PBS containing 0.05% Tween-20, 3% bovine serum albumin, and a protease inhibitor cocktail), and the supernatant was collected after centrifugation. Antigen-specific antibody levels in serum and intestinal lavage fluid were detected by indirect ELISA. Briefly, 96-well plates were coated with purified FimA, FimB, FBA and NetB proteins (1 μg/mL, 100 μL/well) overnight at 4°C. After blocking with PBST containing 2% bovine serum albumin overnight at 4°C, the plates were washed with PBST and then incubated with serially diluted serum (starting dilution 1:100) or intestinal lavage fluid samples (starting dilution 1:20) for 1 h at 37°C. Following another wash, the appropriate horseradish peroxidase (HRP)-conjugated goat anti-chicken IgG/IgA/IgM secondary antibody (Southern Biotech, dilution 1:5,000) was added according to the detection target. Color development was performed using ABTS substrate, and the reaction was stopped by adding 1% sodium dodecyl sulfate. The absorbance (OD value) of each well was measured at 405 nm.

Determination of antigen-specific splenic lymphocyte proliferation in chickens

To evaluate the antigen-specific proliferative response of lymphocytes, splenic lymphocytes were aseptically harvested from immunized broilers and prepared as single-cell suspensions. After purification by density-gradient centrifugation (850 rcf, 40 min), the cells were resuspended in complete RPMI-1640 medium (supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin) and seeded into 96-well plates at 2 × 10⁵ cells/well. Lymphocytes were stimulated with recombinant proteins (FimA, FimB, FBA or NetB proteins, 20 µg/mL each). Concanavalin A (ConA; 5 µg/mL) was used as a positive control, PBS as a negative control, and wells containing medium alone served as blank controls. After incubation for 44 h at 37°C in 5% CO₂, 10 µL of CCK-8 reagent was added to each well. Following an additional 4 h of incubation, the absorbance was measured at 450 nm. The stimulation index (SI) was calculated as: SI = (OD₄₅₀ of stimulated well – OD₄₅₀ of negative control well) / (OD₄₅₀ of negative control well – OD₄₅₀ of blank control well).

Determination of gene transcription and protein secretion levels of IFN-γ and IL-4 in splenic lymphocytes and jejunal tissues following antigen stimulation

The single-cell suspension prepared as above was seeded into 6-well plates at 3 × 10⁶ cells/well and then supplemented with complete medium containing FimA, FimB, FBA, or NetB proteins (20 μg/mL, final volume 1 mL/well). After incubation for 72 h at 37°C in 5% CO₂, the cells were centrifuged at 8,000 × g for 2 min at 4°C, and the cell pellets and culture supernatants were collected separately, subjected to RT-qPCR and ELISA assays, respectively.

In pre-coated 96-well ELISA plates (Kete, Jiangsu, China), 50 μL of standard, test culture supernatant (in duplicate), and blank control (dilution buffer) were added per well. After incubation at 37°C for 30 min, 50 μL of HRP-labeled detection secondary antibody was added, followed by incubation at 37°C for 30 min in the dark. Chromogenic substrate was then added, and after full color development, the reaction was stopped. The absorbance was measured immediately, and standard curves were plotted to calculate the concentrations of IFN-γ (pg/mL) and IL-4 (ng/L) in the culture supernatant.

To complement these data at the transcriptional level, gene expression analysis was performed in parallel on the collected cell pellets. Total RNA was extracted from using the TRIzol method. A total of 1 µg of RNA was used to remove genomic DNA, and cDNA was reverse-transcribed using the BeyoRT™ First Strand cDNA Synthesis Kit (Beyotime, China) in a 20 µL reaction system (42°C for 60 min, followed by 80°C for 10 min). Quantitative real-time PCR was performed using the SYBR Green dye method in a 20 µL reaction mixture containing cDNA template, 2× BeyoFast SYBR Green premix, and gene-specific primers for IL-4 (chIL4-F/R), IFN-γ (chIFN-γ-F/R), CCR7 (chCCR7-F/R), and the reference gene β-actin (chβ-actin-F/R) (sequences listed in Table 2). The cycling conditions were as follows: 95°C for 2 min for initial denaturation, followed by 40 cycles of 95°C for 15 s and 60°C for 10 s. Melting curve analysis was performed by gradually increasing the temperature from 60°C to 95°C at a rate of 0.3°C/s. Cycle threshold (Ct) values were acquired using an Applied Biosystems 7500 real-time PCR system, and the relative expression levels of target genes were calculated using the 2⁻^ΔΔCt method. All data were normalized using the reference gene (β-actin). In addition, the jejunal tissues were also collected at necropsy aseptically and analyzed by RT-qPCR as described above.

Table 2.

Primers used in this study for qRT-PCR.

Primers	Sequences
chβ-actin	F: GAGAAATTGTGCGTGACATC
	R: CCTGAACCTCTCATTGCCA
chIL-4	F: TGAATGACATCCAGGGAGAG
	R: GGCTTTGCATAAGAGCTCAG
chIFN-γ	F: AGCTGACGGTGGACCTATTATT
	R: GGCTTTGCGCTGGATTC
chCCR7	F: GCTTTCAGCCTTCCTCTCAT
	R: GACCTCCTTCTTCTCACACATC
chCD83	F: CTGGGAGTTGGCGACAGAATAG
	R: TTGAAAGGTGGTGTTGGAGAG
chCD86	F: TACCTTGGCCAGGAAAAACA
	R: ATACTGCCCCTCATCCACAA
chCCL5	F: CCGTGTGCTGCTTCAACTAT
	R: CTTCCTGGTGATGAACACAACT
chCCR9	F: GTGCCTCCCTGAGATCATGT
	R: TGTGCTTTTGGCATCTTTTG
chCD28	F: GTGGCAGTGACGGGACTTCTTG
	R: CCTTGTTCTTCTGGTGAGGTGGATG

1. Zhang G, Zahra A, Yang T, Guo Q, Sun Y, Zhang Y, Gao Y, Zhang Y, Wang M, Gong J, Huang H, Wang Z, Wang C, Jiang Y. 2025. Enterococcus faecalis strains derived from wild bird provide protection against Clostridium perfringens challenge in locally-sourced broilers. Front Vet Sci 12:1601605.

2. Yang T, Wang T, Zhang G, Guo Q, Zhang Y, Gao Y, Sun Y, Zhang Y, Wen X, Lu X, Xu Y, Wang Y, Du L, Huang H, Zou S, Wang Z, Wang C, Jiang Y. 2026. Oral administration of recombinant Lactobacillus plantarum co expressing FimB, CnaA, NetB and FBA antigens targeting chicken dendritic cells provides effective protection against necrotizing enteritis in broilers. Poult Sci 105:106889.

Determination of intestinal immunoglobulin A (IgA) secretion levels by immunofluorescence

Jejunal and ileal tissues were collected from experimental animals at 14 days post-booster immunization, fixed in 4% paraformaldehyde, paraffin-embedded, and sectioned at a thickness of 3 μm. After deparaffinization and rehydration, antigen retrieval was performed by heating the sections in citrate buffer (pH 6.0). Following blocking with PBS containing 5% bovine serum albumin, the sections were incubated with mouse anti-chicken IgA antibody (SouthernBiotec, 1:200) overnight at 4°C in the dark. FITC-conjugated goat anti-mouse IgG antibody (TransGen, China, 1:500) was used as the secondary antibody and incubated for 1 h at room temperature. Cell nuclei were counterstained with DAPI for 5 min. Stained sections were visualized and imaged using a fluorescence microscope. The mean fluorescence intensity of IgA-positive cells was quantified using Image J software, with five random fields analyzed per sample.

Transcriptome sequencing

To comprehensively and systematically elucidate the molecular mechanisms underlying the immune adjuvant effects of heat-inactivated versus live E. faecalis EF32 at the transcriptome level, high-throughput transcriptome sequencing analysis was performed using the intestinal samples from the deadEF32+NC8(515+590) and EF32+NC8(515+590) groups. Firstly, principal component analysis (PCA) and sample correlation heatmaps were used to assess the reproducibility and global expression patterns among groups. Subsequently, volcano plots were employed to screen for differentially expressed genes (DEGs) between the two groups. Furthermore, KEGG pathway enrichment analysis was conducted to systematically dissect the pathways involved in immune response, cell signal transduction, and intestinal barrier function, thereby elucidating the impact of heat inactivation on the immune adjuvant effects of EF32 and the host immune regulatory mechanisms.

To validate the transcriptome sequencing results, the expression levels of immune-related genes CCR9, CD28, and CD86 (primer sequences listed in Table 2) in the deadEF32+NC8(515+590) and EF32+NC8(515+590) groups were detected by RT-qPCR as described above. Using β-actin as the reference gene, the relative expression levels of target genes were calculated using the 2⁻ΔΔCt method. Three replicates were set for each sample, and the experiments were conducted according to the manufacturer's instructions.

In vitro effects of E. faecalis EF32 culture supernatant on the expression of maturation-related genes in chicken bone marrow-derived dendritic cells (chBMDCs)

Immature chBMDCs were prepared according to an established protocol (Wu et al., 2010) with modifications. Briefly, bone marrow was aseptically flushed from the tibiae and femurs of 7-day-old broiler chickens using RPMI-1640 basal medium to collect bone marrow cells. Mononuclear cells were isolated by density-gradient centrifugation using a poultry-specific lymphocyte separation medium (Solarbio, China). The cells were resuspended in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS), 1% penicillin-streptomycin, 10 ng/mL recombinant chicken GM-CSF, and 10 ng/mL recombinant chicken IL-4 (KingFisher Biotech). Cells were then cultured at 37°C in a 5% CO₂ atmosphere, with half-medium changes performed every other day. After 7 days, immature DCs (imDCs) were collected and subjected to co-incubation with the supernatant of EF-32 strain. In detail, the culture supernatants of EF-32 strain were collected after a overnight culture of MRS medium. After centrifuge at 12,000 g for 15 min at room temperature, the collected supernatants were filtered with a 0.45 um filter and inoculated into the imDCs cells at the final percentage of 15%. Then the cells were collected at 12 h, 24 h, and 36 h post incubation and total RNA was extracted from cell pellets using the TRIzol method. Then gene-specific primers for CD83 (chCD83-F/R), CD86 (chCD86-F/R), CCL5 (chCCL5-F/R), CCR7 (chCCR7-F/R), and the reference gene β-actin (chβ-actin-F/R) (sequences listed in Table 2) were employed for RT-qPCR to determine the differentiation and maturation of chBMDC. A control group receiving only regular complete medium was also included as negative control.

Statistical analysis

GraphPad Prism software (version 9) was used for plotting and statistical analysis. One-way analysis of variance (ANOVA) was used to compare differences between groups $*\mathrm{P}<0.05;**\mathrm{P}<0.01;***\mathrm{P}<0.001;****\mathrm{P}<0.0001$.

Results

Construction of recombinant strain expressing pgsA'-FimB fusion protein

In this study, the recombinant plasmid pYL515 was constructed to synthesize the surface-displayed FimB protein via pgsA’ anchoring sequence, whereas plasmid pYL590 was constructed previously in our laboratory to display FimA, Fba, and NetB proteins on the surface of L. plantarum strain via pgsA’, LysM and Lp1261 anchoring sequences, respectively (Fig. 1A). To validate antigen expression and cellular localization, Western blotting and immunofluorescence analyses were performed. The FimB antigen (38.6 kDa, including the anchoring sequence) was detected in the cell pellet fractions of recombinant L. plantarum NC8(pYL515), as well as in the cell wall extracts of NC8(pYL515) (Fig. 1B-C). The surface displayed FimB was also confirmed by immunofluorescence assay (Fig. 1D), after incubation of recombinant L. plantarum NC8(pYL515) with anti-FimB primary antibody and FITC-labeled fluorescent secondary antibody, green fluorescent bacteria were observed under a fluorescence microscope, whereas no specific fluorescent signal was detected in the NC8 negative control strain. On the other hand, the fusion proteins FimA (54.45 kDa), Fba (43.23 kDa), and NetB (42.24 kDa) were also observed in the cell wall extractions of the NC8(pYL590) strain (Fig. 1E-G), consistent with expectations. No corresponding bands were detected in the empty vector control group.

Fig. 1.

Plasmid construction and validation of protein expression. (A) Schematic diagrams of plasmids. pYL515 is the recombinant plasmid constructed in this study, carrying the FimB antigen gene; pYL590 is a laboratory-preserved plasmid containing the FimA, NetB, and Fba antigen genes.(B-C) Western blot analysis of FimB expression in L. plantarum NC8 (B), and its localization in the cell wall protein fraction (C). (D) The anchoring effect of proteins expressed by pYL515 was verified using confocal laser scanning microscopy (CLSM).(E-G) Western blot validation of the antigens FimA (E), Fba (F), and NetB (G) expressed by pYL590 in L. plantarum NC8.

Co-immunization with E. faecalis EF32 and recombinant L. plantarum NC8 (pYL515 + pYL590) enhanced serum and intestinal antigen-specific antibody levels

To evaluate the effect of oral immunization with EF32 combined with multi-antigen recombinant strains on the humoral immune response in chickens, serum samples were collected on day 14 post-booster immunization (Fig. 2A). Specific IgG levels against FimA, FimB, NetB, and Fba were detected by ELISA. The results showed that the production of serum IgG could be noticed at day 14 post boost immunization, especially the FimB and NetB specific antibodies in G6 group (Fig. 2B). To note, the increased production of NetB specific IgG antibody could also be observed in either G4(mixed NC8 strains) or G5 (dead EF32 plus mixed NC8 strains) groups. On the other hand, no obvious generation of FimA and Fba IgG antibody could be observed among all the groups. In general, co-administration of EF32, especially live EF32, enhanced the levels of various antigen-specific IgGs, with more pronounced enhancement for FimB and NetB.

Fig. 2.

Animal study protocol and determination of antigen specific antibodies. (A) Schematic timeline of immunization, coccidiosis priming, and C.perfringens challenge. (B)Serum IgG. (C) Intestinal IgA. (D) Intestinal IgM. (E)Intestinal IgG. Data are presented as mean ± SD (n = 4 birds/group);*P < 0.05, ***P < 0.001, ****P < 0.0001 vs. PBS control group and NC8 control group.

To evaluate the effects of combined oral immunization on local mucosal immune responses in the chicken small intestine, small intestinal samples were collected on day 14 post-booster immunization, and antigen-specific sIgA, IgM and IgG levels were detected by ELISA. The results showed that NetB-specific sIgA levels were significantly increased in groups G5 and G6 at day 14(P < 0.05), Fba-specific sIgA levels were significantly (P < 0.01) and extremely significantly (P < 0.001) increased in groups G5 and G6, respectively, compared with the PBS group, and FimA- and FimB-specific sIgA levels showed increasing trends in all immunized groups relative to group G1, with the most evidenced increase observed in group G6 (Fig. 2C). Regarding to small intestinal mucosal IgM, the response mainly occurred in the early phase post-booster and no significant differences were observed among groups (Fig. 2D). The productions of mucosal IgG were also determined and the results showed that the FimB and Fba specific IgG antibodies could be observed at 14 days post-boost immunization, in particular, all the four different specific IgG antibody titers in G6 showed significantly increased compared to G1 group (Fig. 2E). Collectively, these results indicate that oral co-immunization with EF32 and multi-antigen-expressing recombinant strains enhances local antigen-specific intestinal immune responses in the chicken small intestine. Moreover, the live EF32 co-immunization group (G6) showed overall better efficacy than the inactivated EF32 co-immunization group (G5).

Co-administration stimulated elevated splenic cellular proliferation

Splenic lymphocyte proliferation assays were performed using four recombinant antigen proteins and the non‑specific mitogen ConA at 14 days post boost immunization. The results showed that chickens immunized with the EF32+NC8(pYL590+pYL515) vaccine (group G6) exhibited a significantly higher proliferation index upon FimB and ConA stimulation compared with the PBS control group(G1) (P < 0.05), and the NetB protein also dramatically elevated the proliferation results in G6 group, even though no statistic difference could be observed (Fig. 3A).

Fig. 3.

Detection of splenic lymphocyte proliferation and intracellular production of cytokines. (A) Splenocyte proliferation index was determined using the CCK-8 assay following in vitro stimulation with recombinant proteins (FimA, FimB, NetB, Fba) and concanavalin A (ConA) at day 14 post-immunization. (B) Relative mRNA expression levels of IL-4 and IFN-γ in splenic lymphocytes. (C) Concentrations of IL-4 and IFN-γ in splenic lymphocyte culture supernatants. (D) Relative mRNA expression levels of IL-4, IFN-γ, and CCR7 in intestinal tissues. Data are presented as mean ± SD (n = 4 per group). *P < 0.05, **P < 0.01, ***P < 0.001 vs. G1 and G2 groups.

The presence of live EF32 strain dramatically increased the intracellular production of cytokines

Subsequently, intracellular cytokine productions were determined by qRT‑PCR and ELISA. The results of qRT-PCR showed that group G6 significantly increased the mRNA levels of the Th2‑type cytokine IL‑4 (P < 0.01) and the Th1‑type cytokine IFN‑γ(P < 0.05) compared with G1 group(Fig. 3B). To note, the mRNA levels of both cytokines were also dramatically increased in G6 compared with all the other groups, even no statistic differences were observed. A similar trend was also observed for IL‑4 and IFN‑γ secretion in splenocyte culture supernatants, which indicated that production of both IL-4(P < 0.001) and IFN‑γ(P < 0.05) in the supernatants were significantly elevated compared with G1 group(Fig. 3C). To further confirm the effects of live EF32 in vivo, the intestinal samples from all the immunized birds were subjected to qRT-PCR and the results also demonstrated that the co-administration of live EF32 and recombinant L. plantarum strains significantly increased the mRNA levels of both IL‑4 (P < 0.01) and IFN‑γ(P < 0.001)compared with G1 group, whereas the inactivated EF32 did not. On the other hand, a very interesting results was that the presence of live EF32 alone was also able to elevate the mRNA levels of both cytokines (Fig. 3D).

Intestinal sIgA levels in the jejunum and ileum sections were significantly elevated after co‑immunization

To evaluate the effect of combined oral immunization on local mucosal immune responses in the intestinal tract, jejunal and ileal tissues were collected from each group at day 14 post‑booster immunization, and generation of secretory IgA (sIgA) was detected by immunofluorescence. The results showed that, in both jejunal and ileal samples, the sIgA immunofluorescence signals in all immunized groups were enhanced compared with those in groups G1 and G2 groups. Among them, group G6 exhibited the strongest fluorescence signal, with a broader and more continuous positive distribution in both jejunal and ileal tissues (Fig. 4A, B). Further quantitative analysis of mean fluorescence intensity (MFI) revealed that in both the jejunum and ileum, the sIgA fluorescence intensity in group G6 was significantly higher than that in groups G1 and G2 (P < 0.05). In comparison, groups G3, G4, and G5 also showed certain increasing trends, but their overall intensities were lower than that of group G6 (Fig. 4C, D). These results suggest that oral administration of EF32 or multi‑antigen recombinant strains alone can enhance local intestinal sIgA immune responses to some extent, whereas the combination of live EF32 with the recombinant strains exhibits a more evidenced effect, indicating that live EF32 may have a synergistic effect on the mucosal immunity induced by the recombinant strains.

Fig. 4.

Enhanced mucosal IgA secretion determined by immunofluorescence staining. (A, C) Immunofluorescence staining and quantitative analysis of IgA (green) in ileal tissue sections. (B, D) Immunofluorescence staining and quantitative analysis of IgA (green) in jejunal tissue sections. Cell nuclei are counterstained with DAPI (blue). MFI means fluorescence intensity. Data are presented as mean ± SD; * P < 0.05 vs. PBS control group and NC8 control group.

Co-immunization conferred enhanced protection against NE challenge

After NE challenge, broilers in the PBS group (G1) showed the slowest body weight gain, whereas the administration of NC8(G2) or EF32(G3) alone somehow alleviated the body weight gains. On the other hand, almost similar body weight gains were observed in both mixed L. plantarum group(G4) and deadEF32 mixed group (G5), whereas the mixed live EF32 group (G6) showed the best performance compared to all the other groups (Fig. 5A). Similar results could also be observed regarding to the intestinal lesion scores (Fig. 5B). In detail, an average lesion score of 3.429 was observed in G1 group, whereas the administration of NC8(G2) or EF32(G3) alone alleviated the lesion scores to 2.429 and 1.571, respectively, indicating that both strains could somehow provide some protection against NE challenge. On the other hand, dramatically decreased lesion scores were observed in both G4 and G5 groups, with an average score of 0.571 and 0.833, respectively. Interestingly, the most evident decreased lesion score could be observed in G6 group, with an average lesion score of 0.4. In addition, the H&E staining of jejunal samples were consistent with the intestinal lesion scores. The G1 group showed typical necrotic enteritis pathological changes, such as villus atrophy and fragmentation, epithelial necrosis and shedding, and marked inflammatory cell infiltration. In contrast, the co‑immunization groups (G4–G6), especially group G6, exhibited relatively intact villus structures and significantly reduced inflammatory responses (Fig. 5C).

Fig. 5.

Protection against C. perfringens challenge. (A) Mean body weight changes in each group monitored for 5 days post-challenge. (B) Intestinal lesion scores of each group following challenge, assessed using a standardized 0–6 scoring system. (C)Representative hematoxylin and eosin (H&E) staining images of jejunal tissue sections from each group. Scale bars are indicated in the original images. Data are presented as mean ± SD; P < 0.05, *P < 0.01, **P < 0.01 vs. PBS control group and NC8 control group.

All these results indicated that oral co‑immunization with EF32 and multi‑antigen recombinant strains effectively alleviated intestinal tissue damage, improved growth performance, and reduced the severity of NE lesions following challenge. Among them, the live EF32 co‑immunization group provided the best protective efficacy, which was superior to that of the recombinant strains alone or the inactivated EF32 co‑immunization group.

EF32 culture supernatant induces chBMDC maturation and upregulates the expression of chemotaxis-related molecules

Since the obvious adjuvant effects of live EF32 strain on the recombinant L. plantarum strains, an interesting hypothesis was that whether the culture supernatants of EF32 could stimulate the humoral immune response. In case of that, the culture supernatants were added into the isolated chBMDC cells and the differentiation and maturation markers were determined by qRT-PCR analysis. The results showed that, compared with the PBS group, EF32 culture supernatant modulated the expression of molecules associated with chBMDC differentiation (CD86) and maturation (CD83, CCL5 and CCR7) (Fig. 6). Notably, at 12 h, the mRNA transcription levels of CD86(P<0.01), CCL5(P<0.01), and CCR7(P<0.05) in the EF32 group were significantly increased compared with those in the PBS group, and CD83 mRNA also showed an upward trend, indicating that EF32 supernatant promoted the transition of chBMDCs toward an activated state at the early stage of stimulation (Fig. 6A). The stimulatory effect of EF32 supernatants on chBMDCs remained active until 24 h, with particularly prominent induction of CD86 and CCL5 (Fig. 6B). On the other hand, the relative mRNA levels of maturation marker CD83 in EF32 treated chBMDCs were significantly increased compared with the mock control(P<0.01) at 36 h, whereas the other three markers appeared to be similar, without no statistic difference (Fig. 6C), indicating that the supernatants of EF32 strains could stimulate the differentiation and maturation of chBMDCs as time elapsed.

Fig. 6.

Transcriptional levels of maturation markers and chemokines in chicken bone marrow-derived dendritic cells after stimulation with EF32 supernatant.(A) Relative mRNA expression levels of CD83, CD86, CCL5, and CCR7 in cells cultured for 12 h.(B) Relative mRNA expression levels of CD83, CD86, CCL5, and CCR7 in cells cultured for 24 h.(C) Relative mRNA expression levels of CD83, CD86, CCL5, and CCR7 in cells cultured for 36 h. Cells were treated with EF32 supernatant (EF32 group) or PBS (control group). Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001 vs. PBS control group.

To further confirm these effects in vivo, the chicken intestinal samples were collected at 14 days post boost immunization and subjected to qRT-PCR assay targeting to CCR7, and the results demonstrated that the presence of EF32 alone (G3) could somehow elevate the mRNA levels of CCR7, whereas the combination of both EF32 and recombinant L. plantarum strains could increase the mRNA levels further, especially the live EF32 combination group (G6) (Fig. 6D), in consistent with the elevated CCR7 levels observed in vitro.

Transcription assay indicated the live EF32 combined with multi‑antigen recombinant strains enhanced mucosal immune responses in chickens

The intestinal samples from G5(dead EF32 mixed group) and G6(live EF32 mixed group) were collected and subjected to transcription analysis. Principal component analysis (PCA) revealed a clear separation between the inactivated bacteria group (dEF32Mix,G5)and the viable bacteria group (LEF32Mix, G6), with only minor overlap, indicating that viable bacteria treatment significantly altered the global gene expression profile in the host intestine. Meanwhile, biological replicates within each group clustered closely, demonstrating good reproducibility and high reliability of the experimental data. This result provides a prerequisite for subsequent differential gene screening and functional enrichment analysis, and validates the effectiveness of the experimental design (Fig. 7A).

Fig. 7.

Transcriptome sequencing and immune-related gene expression in the intestine of chickens immunized with live or dead EF32 combined with recombinant L. plantarum strains.(A) Principal component analysis (PCA) of transcriptomes between the live EF32 group (LEF32Mix) and dead EF32 group (dEF32Mix).(B) Correlation analysis of gene expression levels between the two groups.(C) Volcano plot showing differentially expressed genes (DEGs) between LEF32Mix and dEF32Mix (red: upregulated; blue: downregulated; gray: not significant).(D) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of DEGs.(E) Relative mRNA expression levels of immune-related genes (CCR9, CD28, CD86) in the intestine.Data are presented as mean ± SD; *P < 0.05, ns: not significant vs. G5 group.

The correlation heatmap further indicated good intra‑group reproducibility and differences in sample characteristics between these groups (Fig. 7B). Differential analysis revealed a certain number of differential features between the two groups (Fig. 7C), and functional enrichment was mainly involved in immune‑ and barrier‑related pathways such as the intestinal immune network for IgA production, T cell receptor signaling pathway, tight junction, as well as various metabolic pathways (Fig. 7D). Then qRT‑PCR results showed that, compared with G5 group, the mRNA transcription levels of CCR9 and CD86 in the intestine of group G6 were significantly increased (P < 0.05), while CD28 showed an increasing trend but did not reach statistical significance (Fig. 7E). To note, the presence of CCR9 was a critical factor for the migration of cells to gut-associated lymphoid tissue, which was necessary for lymphocytes homing and generation of mucosal immune response (Deng et al., 2025).

Discussion

Currently, the development of an effective oral vaccine against NE still needs to overcome challenges such as insufficient mucosal immune responses and low efficiency of local antigen delivery in the intestinal tract. In this study, four different protective antigens were selected, including FimB, NetB, FimA, and FBA proteins, taking advantages of surface-displayed antigens on lactic acid bacteria which was designed to enhance the stable display of antigens on bacterial surface and improve their accessibility to the host mucosal immune system. It should be noted that the NetB, FimA, and FBA antigens were previously constructed in our laboratory as recombinant L. plantarum (pYL590) (unpublished data). On this basis, a recombinant L. plantarum (pYL515) capable of surface-displaying the FimB antigen was further constructed (Fig. 1). The two recombinant L. plantarum strains were then orally co-immunized with either live or inactivated E. faecalis EF32 strain to investigate the protective efficacy of this co-immunization strategy against NE.

Co-immunization with live EF32 and multi-antigen-expressing recombinant L. plantarum strains not only induced higher levels of serum antigen-specific IgG but also effectively promoted the establishment of mucosal sIgA, IgM and IgG antibody responses in the small intestines (Fig. 2). SIgA plays an important role in blocking pathogen adhesion, limiting colonization, and neutralizing toxins. The marked enhancement of sIgA observed 14 days post-booster in this study is likely one of the key mechanisms underlying the protective efficacy of the co-immunization strategy. Immunofluorescence results further confirmed that live EF32 combined with the recombinant strains significantly increased sIgA-positive expression levels in the jejunum and ileum (Fig. 4), supporting that this co-immunization strategy establishes a more effective mucosal immune barrier in the intestinal tract. In addition to humoral and mucosal immune responses, this co-immunization strategy significantly enhanced cellular immune responses: after antigen re‑stimulation, the live EF32 co‑immunization group exhibited significantly increased splenic lymphocyte proliferation, along with elevated expression and secretion of IFN-γ and IL-4 in splenic lymphocytes and their culture supernatants (Fig. 3). This suggests that the strategy could induce a relatively balanced Th1/Th2‑type immune status. This is highly consistent with the summary by Emami and Dalloul on host–pathogen interactions in NE, i.e., that effective control depends not only on antibodies but also on local inflammatory regulation, barrier protection, and a more coordinated immune network (Emami and Dalloul, 2021). Although the immune intensities induced by different antigens were not entirely consistent, the challenge protection results indicated that this co‑immunization strategy was more favorable for combating C. perfringens infection (Fig. 5).

Another key finding was that the adjuvant effect of live EF32 was significantly superior to that of inactivated EF32 in most cases. Compared with the recombinant strain alone(G4) or the inactivated EF32 co‑immunization group(G5), the live EF32 co‑immunization group(G6) exhibited remarkable elevated immune response and protection, including serum IgG (Fig. 2B), intestinal mucosal sIgA (Fig. 2C), splenic lymphocyte proliferation (Fig. 3A), cytokine expression (Fig. 3B, C), as well as body weight recovery and lesion alleviation after challenge (Fig. 5). This suggests that EF32 is more likely to act as an active factor with immune‑adjuvant function in this co‑immunization regimen, especially the unknown components of culture supernatants, instead of EF32 bacterial itself.

Based on the above results, to explore the mechanism by which EF32 enhances immune effects, an in vitro chBMDCs experiment was performed. The results confirmed that the culture supernatant of EF32 could elevate the transcription levels of CD83, CD86, CCL5, and CCR7 mRNA in chBMDCs at the early stage of stimulation. Among them, CCR7 as a chemokine receptor, plays a crucial role in dendritic cells migration and homing to lymph node and initiation of subsequent adaptive immunity (Pang et al., 2026). These findings suggest that the secreted products of EF32 can promote the activation of dendritic cells and enhance their functional status related to antigen presentation, chemokine signal formation, and migration (Fig. 6). To further confirm the adjuvant effects of EF32 in vivo, the intestinal samples after oral immunization were analyzed by transcription assay. Interestingly, the live EF32 strongly stimulated the transcription of some immune related genes, such as the intestinal immune network for IgA production, T cell receptor signaling pathway, tight junction, and endocytosis, accompanied by significant upregulation of CCR9 and CD86 (Fig. 7). Notable, the intestinal homing molecule CCR9 on lymphocytes has been reported to be related with lymphocyte activation, the upregulation is usually considered to enhance lymphocyte homing to the intestinal mucosa and a subsequent boost in mucosal immunity (Deng et al., 2025), in consistent with our results in which the presence of EF32 significantly increased the production of sIgA by ELISA (Fig. 2C) and immunofluorescence analysis (Fig. 4).

Conclusion

In this study, a recombinant L. plantarum candidate vaccine based on the FimB antigen was successfully constructed. Co‑immunization of broilers with this strain together with the previously constructed L. plantarum (pYL590) expressing FimA, Fba, and NetB antigens, as well as a wild bird isolated E. faecalis EF32 strain, significantly enhanced protection against G‑type C. perfringens through stimulation of humoral, intestinal and cellular immune response, providing a new approach for the development of novel oral vaccines against chicken NE.

Author contributions

Yao Xu was responsible for the overall experimental design, major experimental implementation, data analysis and collation, and the drafting of the original manuscript. Yuxi Zhang and Huanyu Liu participated in the optimization of experimental methodologies and data validation. Xin Lu, Xin Wen, Tong Wang, Linhai Du and Yijie Wang assisted in recombinant strain construction, animal experiments and sample detection. Haibin Huang, Zhannan Wang and Chunfeng Wang performed data statistical analysis and graphical visualization. Yanlong Jiang designed the research project, undertook overall supervision of the study, secured the supporting research funds, revised and polished the manuscript, and finalized the version for submission. All authors have read and approved the final manuscript, and agree to assume full accountability for the accuracy and integrity of all research contents.

Ethics declaration

This study was conducted in accordance with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines. This study was approved by the Jilin Agricultural University. (Approval No. 2025 10 10 008)

Disclosures

The authors declare that they have no competing interests.