Immune Responses Induced by Recombinant Bacillus Subtilis Expressing the Hemagglutinin Protein of H5N1 in chickens

To develop an effective, safe, and convenient vaccine for the prevention of highly pathogenic avian influenza (HPAI) H5N1, we have constructed a recombinant Bacillus subtilis strain (B.S.-HA) expressing the hemagglutinin (HA) protein. Then we evaluated the immune function in chicken bone marrow derived dendritic cells (BM-DCs), and the immune response after oral immunization. Our results show that the recombinant Bacillus subtilis B.S.-HA could be sampled by BM-DCs in vitro and increase the BM-DCs major histocompatibility complex (MHC) II phenotype. The weight, height of the small intestine villus, and lymphoid tissue area of the ileum increased significantly in B.S.-HA immunized chickens (P < 0.05 or P < 0.01). B.S.-HA induced the secretion of cytokines and the expression of Toll-like receptors in the trachea and small intestine (P < 0.05 or P < 0.01). In addition, B.S.-HA elevated the specific IgA titers in the trachea, IgG and HI antibody titers in serum (P < 0.05 or P < 0.01). Therefore, B.S.-HA provides a potential novel strategy and approach for developing an H5N1 vaccine.


Isolation and culture of chicken bone marrow derived dendritic cells (BM-DCs). Chicken
BM-DCs were generated as previous method 14 . Femurs and tibias of 4-6 week-old chickens were removed and isolated from the surrounding muscle tissue using sterile instruments. The bones were washed three times with 0.01 M PBS, and then both ends of the bones were cut with scissors in the dish. The marrow was flushed with 0.01 M PBS using a 10 ml syringe with a 0.45-mm-diameter needle. Clusters within the marrow suspension were disaggregated by vigorous pipetting. After one wash in PBS, the cells were suspended in 0.01 M PBS and loaded onto an equal volume of Histopaque-1119 (Sigma-Aldrich, UK) and centrifuged at 2500 rpm for 25 min. Cells at the interface were collected and washed twice with 0.01 M PBS.
Cells obtained from femurs and tibias were cultured at a final concentration of 2 × 10 6 cells/ml in six-well plates in the culture medium containing RPMI-1640 (Gibco, USA), 10% fetal bovine serum (FBS) (Wisent, CAN), 50 ng/ml recombinant chicken GM-CSF (Abcam, USA), 10 ng/ml IL-4 (Kingfisher, USA), 1 U/ml penicillin and 1 μ g/ml streptomycin, for 7 days at 37 °C and 5% CO 2 . Half of the medium was replaced with fresh complete medium at day 2 and day 4 to remove non-adherent cells (such as dead cells and granulocytes). Effects of the recombinant chicken GM-CSF and IL-4 on cell differentiation were recorded by observing cell morphology, clustering and cell growth. The cell cultures were photographed during 7 days of culture with a digital camera on an inverted microscope.
Purity and uptake phenotype assay. The chicken BM-DCs were stimulated with 0.01 M PBS, LPS (100 ng/ml), B.S. (10 7 cfu/well) and B.S.-HA (10 7 cfu/well) at 37 °C and 5% CO 2 at day 6. After 24 h, cells were harvested by gentle pipetting and washed with 0.01 M PBS. Then, the immature BM-DCs (0.5 × 10 6 cells/ml) were stained with 0.05 mg/ml of PE-conjugated mouse anti-human CD11c antibody (eBioscience, USA) and incubated for 20 min at room temperature. The purity of BM-DCs was evaluated based on the relative levels of CD11c expression. In addition, the mature BM-DCs were stained with 0.5 mg/ml of FITC-labeled mouse anti-chicken MHC II antibody (Abcam, USA) and then incubated for 20 min at room temperature, following washed with 0.01 M PBS and centrifuged at 1500 rpm for 5 min. DCs uptake was determined based on the relative levels of MHC II expression. After the final wash, cells were analyzed on a FACS Calibur (BD Bioscience, Cowley, UK).
Antigen uptake assay. The chicken BM-DCs (0.5 × 10 6 cells/ml) were incubated with DyLight 488-B.S. or DyLight 488-B.S.-HA (10 7 or 10 8 cfu/well) at 37 °C for 2 h or 3 h. The DCs were washed three times with PBS and then were analyzed by FACS. Immunization Schedule. The immunization schedule was described by other researchers 10 . In briefly, 10-day-old Hyline chickens were randomly divided into 4 groups of 30 chickens each and immunized. Nonimmunized control (PBS) chickens received oral administration of 150 μ l of 0.01 M PBS. Groups of chickens were immunized orally with one of the following: 10 10 cfu/kg B.S., 10 10 cfu/kg B.S.-HA, or 2 ml/kg IAIV. All chickens were immunized again one week after the first immunization. The body weights were detected from 6 chickens in each group every week.

Collection of Samples.
Blood samples were taken weekly from 6 chickens in each group after the first immunization and allowed to clot overnight at room temperature before serum was collected. Serum was separated by centrifugation and stored at − 20 °C for detection of specific IgG. The chickens were killed 1, 3, 5, and 7 weeks after the first immunization. Trachea and small intestine tissue samples were taken from 6 chickens in each group, and washed with 0.5 ml of 0.01 M PBS repeatedly. The suspensions were centrifuged at 5,000 × g for 10 min, collected, and stored at − 20 °C for specific IgA detection. The same location of the small intestine and ileum tissues from 6 chickens after 7 week immunization were fixed with Bonn's liquid. Chickens were killed at 3 days after the second immunization, trachea and small intestine tissues and their mucosal suspensions were also collected. The mucosal suspensions were stored at − 20 °C for detection of cytokines, and tissues stored in liquid nitrogen for TLRs expression by real-time quantitative PCR (RT-qPCR).
Hematoxylin-eosin staining assay. The same location of small intestine and ileum tissues were fixed with Bonn's liquid, embedded in paraffin and sectioned at 4 μ m thickness. Hematoxylin-eosin staining was applied to the paraffin sections 15 . Then visualized by OLYMPUS CX23. The height of small intestine villi in the same location of the small intestine and the lymph tissue area of ileum at ten different fields in each chicken were counted for the statistical analysis.
RNA isolation and quantitative RT-qPCR assay. Total RNA was extracted from trachea and small intestine tissues using a RNA extraction kit (Thermo Scientific) and subjected to reverse transcription with Prime Script RT-qPCR Kit (Takara, Dalian, CA), using an ABI 7500 instrument (Life Technologies) 16 . The data were reported as values normalized to a housekeeping gene (β -actin) to account for repeated measures. Specific primers were shown in Table 1.
Enzyme-linked immunosorbent assay. Specific IgG in serum and IgA in mucosal suspensions were determined by enzyme-linked immunosorbent assay (ELISA) as previously described 17 . In brief, ELISA plates were coated with the purified recombinant HA protein (our laboratory saved) in carbonate buffer (pH = 9.6) at 4 °C overnight. Plates were saturated with PBS containing 1% BSA at 37 °C for 2 h. The serum and mucosal suspensions of dilution degrees were added and incubated for 1 h at 37 °C. Subsequently, the HRP-conjugated goat anti-chicken IgG and goat anti-chicken IgA (BETHYL, CA) were added to the wells and incubated for an additional 1 h at 37 °C. Subsequently, a substrate solution containing o-phenylenediamine (OPD) and H 2 O 2 was added, the reaction was allowed to proceed for 15 min at room temperature before it was terminated by the stop solution. Finally, the absorbance was measured at 450 nm, using an automated ELISA reader (Molecular Devices, Shanghai, CA).

Hemagglutination inhibition assay.
To explore whether chicken generated avian influenza virus (AIV) neutralizing antibodies, the neutralization activity of serum antibody was determined by hemagglutination inhibition (HI) assays according to the Chinese National Standard of diagnostic techniques for highly pathogenic avian influenza and the Office International des Epizooties (OIE) standards (OIE, 2005).
T-cell proliferation assays. The splenic lymphocyte were isolated from the all group chicken spleens after 7 week immunization, and then labelled with CFSE (Invitrogen, USA). Cells were cultured in a lymphocyte culture medium of 2 × 10 5 cells per well in 24-well culture plates and stimulated by the purified recombinant HA protein (final concentration 10 μ g/ml) for 72 h at 37 °C and 5% CO 2 . Cell proliferation assays were detected by FACS. Non-stimulated cells were used as negative controls.
Statistical analysis. The results were expressed as the means ± the standard deviations (SD). Statistical significance was determined by oneway analysis of variance (ANOVA) followed by Dunnett's t test to evaluate variations between groups; a P value of < 0.05 was considered statistically significant.

Results
The construction of recombance plasmids and the analysis of fusion protein expression. The  Fig. 1A). These aggregates grew in size and were found to be floating or loosely adherent on day 7 (Fig. 1B), following B.S., B.S.-HA, and positive lipopolysaccharide (LPS) stimulation. Many individual cells and peripheral aggregates displayed a large veiled or dendritic appearance indicating maturation (Fig. 1C). When chicken bone marrow cells were cultured under the same conditions without GM-CSF or IL-4, no cell aggregates were observed, and only a few live cells remained in the plates by day 7 (Fig. 1D). Flow cytometry analysis of the cultured cells harvested on culture day 7 is shown in Fig. 1E. The immature DCs expressed high levels of cell surface and putative CD11c molecules.  The changes of body weight, small intestinal villi height, and lymphoid tissue area. B. subtilis has been used as a feed additive to improve the microorganismal environment in the intestine 18 . We weighed the chickens after they developed immunity (Fig. 4).

The proliferation of spleen lymphocytes in vitro.
To examine the proliferative response of splenic lymphocytes primed with the HA protein in vitro, splenic lymphocytes were isolated from chickens 7 weeks after immunization. After the splenic lymphocytes were stimulated by HA protein in vitro, their proliferative index was increased markedly after immunization with B.S.-HA (P < 0.01) (Fig. 6).
The changes of AIV-specific immune responses. Specific IgA antibody levels were measured 1, 3, 5, and 7 weeks after the first immunization. High levels of mucosal specific IgA antibody were observed in the tracheal Specific immunity can be effectively promoted by cellular and innate immunities. Serum IgG antibody levels were detected 1-7 weeks after the first immunization. Serum antibody levels increased (P < 0.01; Fig. 7B), peaking at 3-5 weeks after immunization with B.S.-HA compared with immunization with PBS or B.S. In addition, the IAIV induced higher levels of IgG antibodies at weeks 3 and 4.
The serum antibody titer was detected by the HI assay (Fig. 7C). Serum titers from group that immunized with B.S.-HA increased at 3 weeks and peaked at 5 weeks after the first immunization. In addition, IAIV induced a higher titer 4 weeks after the first immunization. These results show that oral administration of B.S.-HA effectively induced mucosal immunity, a systemic immune response, and resistance to the infection by AIV to a certain extent.

Discussion
B. subtilis is a well-characterized spore-forming bacterium widely used to express heterologous proteins and deliver antigens 8,[19][20][21] . B. subtilis is a probiotic microorganism in broilers and laying hens [22][23][24] . Previous studies have indicated that dietary supplementation of B. subtilis exerts a beneficial effect on intestinal microbiota and gut morphology thereby enhancing growth performance and improving the feed conversion ratio in animals 18 . These effects of B. subtilis are probably due to its ability to produce amylase, protease, lipase and amino acids 25 , which could increase the efficiency of digestion and absorption of nutrients. Researchers generally agree that the improved growth performance observed in response to probiotics is associated with healthy modulation of the gut community composition 26 . In addition, B. subtilis has been used as an adjuvant to stimulate the immune response 27 .
DCs are the most potent professional antigen-presenting cells and play a crucial role in linking of innate and adaptive immunities 28 . B. subtilis recruits more DCs, migrates to mesenteric lymph nodes, and induces an immune response. Our previous experiments showed that B. subtilis increases the MHC II phenotype in pig BM-DCs in vitro and stimulates the uptake ability of DCs 29 . In this study, recombinant B. subtilis was sampled by BM-DCs in vitro, and the percentage of BM-DCs containing B. subtilis increased with the quantity of bacteria used. Interestingly, the quantity of the recombinant bacteria in DCs decreased as the incubation time extended. It reported that recombinant HA proteins from H5N1 influenza viruses are capable of activating mouse DCs activation and suppressing endocytosis 30 . After the B.S.-HA sampled by chicken BM-DCs a period of time, the exposed HA protein inhibited the endocytosis of chicken BM-DCs. TLRs are required to activate innate and adaptive immune responses 31,32 . B. subtilis is recognized by TLR2 or TLR4 and augments mucosal and systemic responses to intranasal antigens in mice 33,34 . B. subtilis could activate the nuclear factor kappa B pathway by binding to TLRs and induced production of primary T helper type 1 (Th1) cytokines such as IL-1, IFN-γ , and IL-12 35    According to the "common mucosal immune system" theory, antigen-sensitized precursor B and T lymphocytes generated at one mucosal site (such as the gut) can be detected at anatomically remote and functionally distinct compartments (such as the respiratory mucosa) 37 . In addition, avian influenza H5N1 has been detected in the nasal cavity mucosa, lungs, cloaca, and serum of aerosol-infected chickens 38 . Secretory IgA antibodies are the major contributors to humoral mucosal immunity against an influenza viral infection 39