Development of a heat-stable and orally delivered recombinant M2e-expressing B. subtilis spore-based influenza vaccine

Abstract

Highly conserved ectodomain of influenza virus M2 protein (M2e) is an important target for the development of universal influenza vaccines. Today, the use of chemical or genetic fusion constructs have been undertaken to overcome the low immunogenicity of M2e in vaccine formulation. However, current M2e vaccines are neither orally delivered nor heat-stable. In this study, we evaluated the immune efficacy of an orally delivered recombinant M2e vaccine containing 3 molcules of M2e consensus sequence of influenza A viruses, termed RSM2e3. To accomplish this, CotB, a spore coat of Bacillus subtilis (B. subtilis), was used as a fusion partner, and heat-stable nonpathogenic B. subtilis spores were used as the carrier. Our results showed that CotB-M2e3 fusion had no effect on spore structure or function in the resultant recombinant RSM2e3 strain and that heterologous influenza virus M2e protein was successfully displayed on the surface of the recombinant RSM2e3 spore. Importantly, recombinant RSM2e3 spores elicited strong and long-term M2e-specific systemic and mucosal immune responses, completely protecting immunized mice from lethal challenge of A/PR/8/34(H1N1) influenza virus. Taken together, our study forms a solid basis for the development of a novel orally delivered and heat-stable influenza vaccine based on B. subtilis spore surface display.

Keywords:

• bacillus subtilis
• influenza virus
• M2e
• spores
• surface display

Abbreviations:

• ADCC, antibody-dependent cell-mediated cytotoxicity
• CDC, complement-dependent cytotoxicity
• CPE, cytopathic effect
• IAVs, influenza A viruses
• M2e, ectodomain of influenza virus M2 protein

Introduction

Influenza vaccination represents the cornerstone of influenza prevention. Universal influenza vaccines based on conserved viral antigens are believed to be a promising strategy for vaccine development based on their ability to induce cross-protective immunity against a broad-spectrum of influenza virus infection. In particular, the ectodomain of influenza virus M2 protein (M2e) is highly conserved among all influenza A viruses (IAVs) and is thus considered to be an attractive antigen target for developing universal influenza vaccines.¹ Protective antiviral immunity induced by M2e-based influenza vaccines was mediated by antibodies to the M2e antigen, leading to antibody-dependent, cell-mediated cytotoxicity (ADCC) and/or complement-dependent cytotoxicity (CDC).^2-4 However, it is difficult to induce potent protective immune response from the 23-amino acid M2e molecule alone, making it necessary to use some chemical or genetic M2e fusion constructs to a variety of carriers in order to improve immune efficacy. Accordingly, since 1999, a number of M2e vaccines, including those based on viral vectors, recombinant proteins or synthetic peptides, and virus-like particles,^5-9 have been developed using the above strategies and have shown protection against influenza A virus challenge. However, none of them is orally delivered or heat-stable.

Bacillus subtilis (B. subtilis) is regarded as nonpathogenic, being classified as a novel food which is currently used as a probiotic for both human and animal consumption.¹⁰ To date, bacterial spores have been applied as a novel surface display system to express heterologous proteins with functionality, among which the B. subtilis spore is the most widely used.^11,12 The B. subtilis spore offers resistance properties and can survive extreme temperature, desiccation and exposure to solvents and other noxious chemicals.¹³ These unique attributes make the B. subtilis spore an attractive vehicle for delivery of heterologous antigens. A variety of advantages have been shown in the utilization of B. subtilis spore as surface display, including high stability and good safety profile, as well as facile construction of recombinant spores containing heterologous genes. Since CotA, CotB, CotC, CotF and CotG are known as outer coat proteins of B. subtilis, they can be used as fusion partners to display heterologous proteins.¹⁴ Thus, by combining a needle-free delivery system with expression and display of heterologous antigens on the surface, recombinant B. subtilis spores have been utilized as useful tools against a number of parasites and bacterial pathogens.^11,12,15,16

In this study, we used CotB as a fusion partner to produce a recombinant B. subtilis strain that expressed influenza A virus M2e on its spore surface. We then tested the efficacy of this recombinant M2e-expressing B. subtilis spore-based vaccine in the induction of immunogenicity and protection against H1N1 influenza virus infection using oral delivery immunization.

Results

Construction and chromosomal integration of CotB-M2e gene fusion

To obtain recombinant B. subtilis spores expressing M2e of influenza A viruses on their surface, the CotB gene and its promoter were used to construct translational fusion. The CotB gene in fusion frame was encoded for N-terminal 275 amino acids of CotB of wild-type B. subtilis PY79 to avoid potential stability problems of genetic constructs caused by C-terminal 3 27-amino-acid repeats of CotB. As shown in Figure 1A, the consensus sequence of M2e of human influenza A viruses was fused to CotB-based fusion frame and was added to 3 linker-chained tandem copies to overcome the low immunogenicity of monomeric M2e. The CotB-M2e fusion was integrated on the B. subtilis PY79 chromosome at the thrC locus by double-crossover recombination events as described in Materials and Methods. The confirmed clone, RSM2e3, was used for further analysis.

Figure 1. Strategy for chromosomal integration of CotB-M2e gene fusion and detection of M2e expression on recombinant RSM2e3 spores. (A) Strategy for chromosomal integration of CotB-M2e gene fusion. Dark blocks in the CotB-M2e frame indicate genes that encoded linkers connecting 3 tandem copies of M2e. Arrows indicate direction of transcription. (B) Western blot analysis of M2e protein expression in purified spores of RSM2e3 (lane 1) and PY79 (lane 2) using M2e-specific mAb. Molecular weight marker is shown on the left. (C) Flow cytometric analysis of the spore surface. Open histogram shows that the purified spores reacted with M2e-specific mAb, while reaction with isotype control antibody is shown as filled histogram.

Expression of influenza A virus M2e protein on the recombinant RSM2e3 spore surface

Spore coat proteins were extracted to detect heterologous M2e expression by Western blot, using M2e-specific monoclonal antibody (mAb). As shown in Figure 1B, a clear band corresponding to the size of CotB-M2e fusion protein (∼37 kDa) was revealed in the samples containing CotB-M2e, but not in those with wild-type PY79 control, indicating the successful expression of heterologous M2e protein on the recombinant RSM2e3 spore surface and the specificity of expressed fusion protein to the consensus sequence of M2e protein of influenza A viruses.

Flow cytometry was then performed to detect the expression of M2e on the spore surface. As shown in Figure 1C, strong fluorescence signals were captured in the recombinant RSM2e3 spores stained with M2e-specific primary antibody and anti-mouse IgG-FITC secondary antibody, while recombinant spores stained with isotype antibody only showed background fluorescence signals. As expected, wild-type PY79 spores showed no fluorescence reactivity in the presence of the test antibodies. This result suggests that M2e protein was displayed on the surface of recombinant RSM2e3 spores.

Morphology characterization and species identification of recombinant RSM2e3

To determine if the recombinant strain had any morphological changes from its parental PY79 strain, both bacteria and their spores were examined by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). SEM showed similar oval or bar shapes of recombinant RSM2e3 and wild-type PY79 bacteria (Fig. 2A, D). TEM revealed clear exine and intine structures of endospores (Fig. 2B, E) and purified spores (Fig. 2C, F) of RSM2e3 and wild-type PY79. When grown in Difco sporulation medium, the recombinant strain RSM2e3 and its parental strain PY79 exhibited comparable sporulation and germination efficiencies, and their spores were equally resistant to chloroform and lysozyme treatment (data not shown), suggesting that the presence of the CotB-M2e fusion did not significantly affect spore structure and function.

Figure 2. Electron microscopy of recombinant RSM2e3 bacteria and spores. Scanning electron microscopy visualized dimensional topographies of bacterial cells (A), RSM2e3; (D), PY79. Transmission electron microscopy revealed ultrastructures of sporulating bacteria with endospores and purified spores (B and C), RSM2e3; (E and F), PY79. OC, outer coat; IC, inner coat.

Species identification was subsequently performed to confirm the recombinant RSM2e3 as B. subtilis, using the API CH 50 test that generates species identification as a probability percentage. The results shown in Table 1 identified RSM2e3 as B. subtilis. To further determine the relatedness of the strains at the genetic level, we sequenced the entire 16S rRNA gene from the RSM2e3 strain. Phylogenetic analysis revealed RSM2e3 to be a member of the B. subtilis subgroup, having homology with B. subtilis strain PY79 (GenBank accession no. AF142577) (Fig. 3).

Table 1. Species designations

Strain	API score	16S rRNA identification ^c
	% ID ^a	T ^b
RSM2e3	89.6 B.subtilis	0.3	B.subtilis
PY79	92.4 B.subtilis	0.61	B.subtilis

^a Percent similarity, i.e., how closely the profile corresponds to the taxon relative to all the other taxa in the database.

^b The T index, i.e., how closely the profile corresponds to the most typical set of reactions for each taxon.

^c Species identification by sequencing of the entire 16S rRNA genes and analysis for homologous sequences, as shown in Figure 3.

Figure 3. Phylogenetic relationship of Bacillus probiotic bacteria. The relatedness of recombinant RSM2e3 and 14 typical commercial Bacillus probiotics based on analysis of 16S rRNA is shown. GenBank accession number of each selected strain is bracketed following its name. RSM2e3 is shown in bold. Bacillus subgroup to which selected probiotics belonged is indicated to the right of the phylogenetic tree.

Recombinant RSM2e3 spores induced strong systemic immune responses in immunized mouse sera and spleens

To evaluate the ability of recombinant spores to induce systemic humoral immune responses, we immunized mice with RSM2e3 spores and tested M2e-specific IgG and subtypes in immunized mouse sera. As shown in Figure 4A, RSM2e3 induced elevated M2e-specific IgG antibody response, reaching the highest titer of 1:12,800 at week 17 post-1st immunization. In addition, IgG1 (Th2), but not IgG2a (Th1), antibodies specific to M2e were also significantly high in RSM2e3-immunized mice (Fig. 4B), indicating that RSM2e3 induced a biased Th2-associated antibody response. In contrast, wild-type PY79 spores and PBS controls did not elicit M2e-specific IgG antibody responses or subtypes in vaccinated mice (Fig. 4). The above results suggest that recombinant RSM2e3 spores were capable of inducing long-term systemic humoral immune responses in mice through the oral route.

Figure 4. Detection of systemic humoral immune responses by ELISA in mice immunized with RSM2e3 spores. Mice were vaccinated orally with RSM2e3 spores, wild-type PY79 control spores or PBS control. (A) M2e-specific IgG antibodies were tested using mouse sera collected at 0, 5, 7, 9, 11, 13, 15, and 17 weeks post-immunization. (B) IgG1 and IgG2a subtypes were tested using mouse sera from 17 weeks post-immunization. The titers were expressed as the endpoint dilutions that remained positively detectable. The data are presented as mean ± SE of 8 mice per group.

To further assess the capability of RSM2e3 to induce systemic cellular immune responses, splenocytes from mice at 17 weeks post-1st immunization were collected and detected for IL-4- and IFN-γ-producing T cells by cell-surface marker and intracellular cytokine staining, followed by flow cytometric analysis. As shown in Figure 5, although RSM2e3 failed to induce the secretion of INF-γ-producing T cells, it did elicit a high frequency of IL-4-producing T cells in CD4⁺ T cell populations after stimulation of the immunized splenocytes with M2e peptide. However, wild-type PY79 and PBS controls did not elicit INF-γ- or IL-4-producing T cell responses in the immunized mice.

Figure 5. Detection of systemic cellular immune responses in mice immunized with RSM2e3 spores. M2e-specific T cell responses were tested by flow cytometric analysis in mouse splenocytes 17 weeks post-immunization. At least 10,000 CD3⁺/CD4⁺ lymphocytes were first gated, and frequencies of IFN-γ⁺ and IL-4⁺ cells were then analyzed and indicated as percentages of CD3⁺/CD4⁺ T cells. The graphs are presented as mean value of 5 independent experiments. Numbers in the upper right corner of each graph represent the frequencies of IFN-γ- or IL-4-producing CD4⁺ T cells. Wild-type PY79 and PBS are controls.

The above results suggest that recombinant RSM2e3 spores induced strong, long-term systemic humoral IgG antibody responses and IL-4-secreting cellular immune responses in sera and splenocytes of orally immunized mice, indicating the importance of RSM2e3 spore-induced humoral and cellular immune responses for the protection of immunized mice from subsequent influenza virus challenge.

Recombinant RSM2e3 spores induced strong local mucosal immune responses in immunized mice challenged with H1N1 influenza virus

To evaluate whether recombinant RSM2e3 spores could elicit local mucosal immune responses, secretory IgA (sIgA) antibodies were tested by ELISA in intestine and lung washes collected from mice 5 d post-challenge with A/PR/8/34(H1N1) influenza virus. As shown in Figure 6, RSM2e3 induced strong M2e-specific sIgA antibody responses in both lung wash and, particularly, intestine wash, while PY79 and PBS controls were unable to elicit the induction of sIgA antibodies in tested samples. Similar to systemic humoral and cellular immune responses, as noted above, these data confirm that RSM2e3 spore-induced local mucosal immune response is also important for the protection of immunized mice from subsequent influenza virus challenge.

Figure 6. Detection of mucosal antibody responses by ELISA in mice immunized with RSM2e3 spores. M2e-specific sIgA antibodies elicited by RMS2e3 spores were tested using mouse intestine and lung washes 5 d post-A/PR/8/34(H1N1) virus challenge. Control mice were administered with wild-type PY79 spores or PBS. The data are presented as mean ± SE of 5 mice per group.

Recombinant RSM2e3 spores induced protective immunity against H1N1 influenza virus challenge

To test the ability of M2e protein-expressing RSM2e3 to induce protection from influenza virus infection, RSM2e3-immunized mice were challenged with A/PR/8/34(H1N1) 17 weeks after the initial immunization, tested for virus titers at day 5 post-challenge, and observed daily for 14 d for body weight changes and survival rate. Results from mouse lung specimens revealed a significantly lower level of virus titers in the mice immunized with RSM2e3 than those administered with wild-type PY79 spores and PBS (Fig. 7A), confirming the ability of RSM2e3 spores to suppress virus replication. Further observation of body weight changes and survival rate revealed that the body weight of mice immunized with RSM2e3 slightly decreased after virus challenge, but then kept increasing 10 d post-challenge, while the body weight of PY-79 and PBS control mice continued to decrease during the observation period (Fig. 7B). In addition, all mice immunized with RSM2e3 spores survived challenge by A/PR/8/34(H1N1) influenza virus to maintain a 100% survival rate, while 90% of the mice in the control groups were dead (Fig. 7C). The above data suggest that RSM2e3 completely protected all vaccinated mice against H1N1 influenza virus challenge.

Figure 7. Cross-protective immunity of recombinant RSM2e3 spores against H1N1 virus challenge. Wild-type PY79 and PBS were used as the controls. (A) Virus titers in RSM2e3-immunized mouse lung tissues tested at 5 d post-A/PR/8/34(H1N1) virus challenge. The data are expressed as mean ± SE of viral titers (Log₁₀TCID₅₀/g) of lung tissues from 5 mice per group. The lower limit of detection is 1.5 Log₁₀TCID₅₀/g of tissues, as indicated by the dotted line. Percentage of body weight change (%) (B) and survival rate (%) (C) of RSM2e3 spore-immunized mice after challenge with A/PR/8/34(H1N1) influenza virus are shown.** indicates significant difference (P < 0.01) between RSM2e3 group and PY79 or PBS group.

Discussion

Human infections from current and other potential influenza viruses require the development of influenza vaccines in large quantities and easy administration. However, routinely prepared live attenuated- and inactivated virus-based vaccines need strict biosafety conditions with the requirement of adjuvants for injection, limiting their use for oral immunization and long-term storage.^17,18

On the other hand, since spores can resist heat, bacterial spore-based vaccines have the greatest advantages for long-term usage without losing stability.¹⁹ Moreover, spores themselves can have adjuvant effects; therefore, no adjuvants are required for immunization, making oral immunization feasible.²⁰ With easy production and storage, high safety profile, and simple immunization with sufficient immunogenicity, B. subtilis spore-based vaccines are expected to be a good choice for prevention of influenza virus infections.

The highly conserved influenza virus M2e protein has been reported as a vaccine target for development of universal influenza vaccines.^21,22 However, the small M2e monomeric molecule alone cannot induce sufficient immune responses, thus requiring its fusion with other components to improve immune efficacy.^23,24 We have previously shown that a tetra-branched multiple antigenic peptide vaccine (M2e-MAP) containing 4 molcules of M2e from the H5N1 virus confers cross-protective immunity against lethal challenges of divergent strains of H5N1 virus, including clade1, A/Vietnam/1194/2004, and clade 2.3.4, A/Shenzhen/406H/06, as well as a heterosubtypic strain of pandemic 2009 H1N1 virus.^25,26 We have also found that a recombinant protein containing 3 molcules of such M2e fused with activation associated protein-1 (ASP-1) adjuvant (M2e3-ASP-1) provides significant cross-clade protection against the above H5N1 viruses.²⁷ These findings suggest that M2e-based influenza vaccines containing 3 or more conserved M2e molecules fused with other components are sufficient to prevent divergent influenza virus infections.

Therefore, in this study, we fused 3 molcules of the M2e consensus sequence of human influenza A viruses (H1, H2, and H3) to CotB, a spore coat of B. subtilis PY79, and integrated the resultant CotB-M2e gene fusion to the B. subtilis PY79 chromosome to generate a recombinant RSM2e3 strain. After sporulation, M2e protein was expressed and displayed on the recombinant spore surface. Morphological examination and species identification proved that the recombinant RSM2e stain maintained the characteristics of its parental wild-type PY79 strain.

M2e exposed on the recombinant spore surface should be in a biologically active form because of the significant immunogenicity elicited by recombinant RSM2e3 spores. The spore can protect the exogenous proteins on its surface from degradation and act as a good adjuvant for these antigens. Therefore, recombinant RSM2e3 spores themselves have self-adjuvanticity, and no additional adjuvant is needed for the RSM2e3 spore-based vaccine formulation. The above results are consistent with previous reports showing self-adjuvanticity of spore-based vaccine candidates against bacterial and parasitic pathogens.^16,20 The potential adjuvant function of B. subtilis spores might be partially due to that these spores can serve as a ‘natural’ microparticle that likely expresses ligands recognized by innate pattern recognition receptors, and thus instruct and augment polyvalent immune responses.²⁸

Although prime-boost immunization is required for the B. subtilis spore-based vaccines to induce sufficient immune responses, repeated immunizations will not affect subsequent induction of elevated immune responses by spore vaccines.²⁹ Here we showed that repeated immunizations of mice with RSM2e3 spores elicited increasing titers of influenza virus M2e-specific IgG antibodies, with IgG1 (Th2) responses, as well as strong cellular immune responses, indicating the ability of RSM2e3 spores to promote specific humoral and cellular immune responses in vaccinated mice. The fact that RSM2e3 spores maintained a long-term antibody response in the immunized mice can be explained by the ability of spores to continuously stimulate the production of immune responses in intestines. In a previous study, it was reported that spores displaying tetanus toxin fragment C (TTFC) generated specific antibodies to both the antigen in the primary oral inoculation sites and those displayed on the resporulated spores in intestines, inducing strong and long-term immune responses.³⁰ It has been shown that the intestine constitutes an important organ for the mucosal immune system by maintaining the ability to elicit mucosal immune responses.³¹ The orally immunized RSM2e3 spores in this study stimulated mice to generate high titers of sIgA, particularly in intestines, confirming their ability to induce strong mucosal immune responses. These RSM2e3-induced M2e-specific humoral and mucosal immune responses have been proven to protect mice from challenges of H1N1 influenza virus, suggesting the ability of the recombinant spores to induce protective immunity.

To date, many M2e-based universal influenza vaccines have been reported and some of them have been tested in clinical trials. However, these M2e-based vaccines have some limitations for better practical application in addition to their limited protection.²² Compared with currently developed M2e-based vaccine candidates, RSM2e3 spores-based vaccine has unique characteristics. Taking advantage of heat-stable, needle-free, and oral delivery features of B. subtilis spores, RSM2e3 spores maintained the ability to induce long-term mucosal immune responses, representing the future trend of developing vaccines against influenza virus, a mucosal pathogen infecting humans via the mucosal route. It is noted that recombinant spore-based vaccines might be challenged in consideration of the size limitation, or requirement of complexed structures or glycosylation of the antigens for antigen presentation.²⁹ Nevertheless, RSM2e3 spores presented effective protection against heterologous influenza virus challenge, regardless of the amino acid difference in the M2e of challenging A/PR/8/34(H1N1) influenza virus and the M2e in RSM2e3 (one amino acid change at position 21), suggesting the potential for further development of RSM2e3 spore candidate as an effective mucosal universal influenza vaccine.

In sum, this study utilizes bacterial spores as a delivery vector to display conserved M2e protein of human influenza A viruses on the spore surface, thereby generating a bacterial spore vaccine with high immunogenicity and sufficient protective immunity. Our data provide a novel means of developing safe, effective, and stable universal vaccines against divergent influenza virus infections. The recombinant spore-based vaccines could also be used as a promising approach to develop vaccines against other viral pathogens.

Materials and Methods

Ethics Statement

All of the procedures involving animals were approved by the Institutional Animal Care and Use Committee (IACUC) of the Beijing Institute of Microbiology and Epidemiology (Permit number: BIME 2013–17). The animal studies were conducted in strict accordance with the recommendations set forth in the Guide for the Care and Use of Laboratory Animals.

Construction of recombinant strain

The B. subtilis wild-type strain used in this work is PY79, a standard prototrophic laboratory strain.³² To obtain a CotB-based gene fusion, CotB gene and its promoter were amplified from PY79 chromosomal DNA using forward and reverse primer CotBF1 and CotBR1, respectively (Table 2). The 1,108-bp amplified product was ligated to pMD18-T (TaKaRa, Japan) and then cleaved with BglII (CotBF1) and HindIII (pMD18-T) for insertion into pDG1664³³ (cleaved with BamH I and HindIII), making a recombinant pDG1664-CotB plasmid. Subsequently, a synthetic gene (GenScript, NJ) encoding 3 linker (GGGGS)-chained tandem copies of M2e consensus sequence (SLLTEVETPIRNEWGCRCNDSSD) derived from human influenza A viruses²² was fused to the downstream of CotB gene frame on pDG1664-CotB by restriction sites of BamH I and HindIII, producing a recombinant pDG1664-CotB-M2e3 plasmid. The inserted sequences were confirmed by nucleotide sequencing.

Table 2 Synthetic oligonucleotides

Oligonucleotide	Sequence (5’—3’)^a
CotBF1	GAGATCTACGGATTAGGCCGTTTGTC
CotBR1	TGGATCCGGATGATTGATCATCTGAAG
thrC-Fi	GAAGGGAACGGTTGGAGC
thrC-Ri	CGGGAACAGTGACAGAGAAC

^a The restriction site is underlined.

Plasmid pDG1664, which allowed inserted gene fusion at the thrC (threonine) locus, enabled selection for Erm^r33 (Fig. 1A). Therefore, plasmid pDG1664-CotB-M2e3 was linearized by digestion with Pst I and used to transform competent cells of PY79 strain according to a previously described 2-step transformation procedure.³⁴ Erm^r (1μg/ml) clones resulted from double-crossover recombination (Fig. 1A) and were tested by PCR using chromosomal DNA as template and oligonucleotides thrC-Fi and thrC-Ri (Table 2) as primers for amplification. The 3,676-bp PCR product indicates the occurrence of correct recombination events based on the presence of an additional part of gene fusion carrying CotB and 3 M2e copies. One confirmed clone was named RSM2e3 (Fig. 1A).

Preparation of spores

Sporulation of either PY79 or RSM2e3 was performed in Difco sporulation medium using the exhaustion method.³⁵ After 40-hour incubation at 37°C, spores were collected, washed several times, and purified by lysozyme treatment, as previously described by Nicholson and Stelow.³⁵ After final suspension in sterile phosphate-buffered saline (PBS), spores were treated at 65°C for 1 h to ensure the absence of viable vegetative cells and then stored in aliquots (1 × 10¹¹ spores/ml) at −70°C until use. Spore counts were determined by serial dilution and plate counting.

Detection of M2e protein expression in recombinant spores

Western blot and flow cytometry analysis were applied to detect expression of M2e protein in recombinant RSM2e3 spores. Spore coat proteins were extracted from suspensions of spores using an SDS-DTT extraction buffer, as described in detail elsewhere.³⁵ Extracted proteins were separated by 12% Tris-Glycine SDS-PAGE gels and transferred to nitrocellulose membranes. After blocking with 5% non-fat milk in PBST for 4 h at room temperature, the blots were incubated overnight at 4°C with M2e-specific mAb 14C2 (1:400, Santa Cruz Biotechnology, Dallas, Texas) and washed 5 times. The blots were then incubated with horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (1:5,000, Invitrogen) for 1 h at room temperature. Signals were visualized with ECL Western blot substrate reagents and Amersham Hyperfilm (GE Healthcare).

For flow cytometry analysis, a total of 10⁵ purified spores were washed in PBS for 3 times and incubated with either M2e-specific or isotype mAb at 37°C for 2 h. After 3 washes in PBS, the spores were incubated with fluorescein isothiocyanate (FITC)-conjugated anti-mouse immunoglobulin G (IgG) at 37°C for 1 h. Spores were finally resuspended in 1 ml of PBS following 3 washes, and at least 10⁴ spores were acquired by Guava easyCyte 6HT-2L flow cytometer (Merck Millipore) and analyzed for M2e expression on the spore surface using FlowJo software (Tree Star, Inc.).

Scanning and transmission electron microscope analysis

Cultured RSM2e and PY79 bacteria were harvested, fixed with 3% glutaraldehyde at 4°C overnight, and then dehydrated with a graded ethanol (50%, 70%, 90%, and 100%). Following subsequent critical-point drying and sputter coating, samples were processed for examination under scanning electron microscope (Hitachi, Japan). For transmission electron microscopy, sporulating bacteria and purified spores of RSM2e3 and PY79 were collected and fixed in glutaraldehyde at 4°C overnight, followed by post-fixing in 1% osmium tetroxide at 4°C for 2 h. Afterwards, samples were dehydrated using a series of graded ethanol (50%, 70%, 90%, and 100%) and then embedded in Epon 812. Ultrathin sections were mounted on 230-mesh copper grids, stained with 1% uranyl acetate and lead citrate, and finally examined under a transmission electron microscope (Philips, Netherlands).

Phylogenetic analysis

Genomic DNA of recombinant bacteria RSM2e3 was extracted, followed by amplifying the entire 16S rRNA using the 16S rDNA Bacterial Identification PCR Kit in strict accordance with the manufacturer's instructions (TaKaRa, Japan). The 1,400-bp PCR product was sequenced using oligonucleotides provided in the kit. The genes for 16S rRNA from RSM2e3 and 14 typical commercial Bacillus probiotics (Genbank AF142577, AB018485, AB018487, AF142574, X64465, AF013121, X55059, AB021192, D16266, D16281, X76436, X76446, X76440, and X76441) were aligned using the Mega 5.2 program,³⁶ and the alignment was used to construct Neighbor-Joining trees using the phylogeny module in the Mega 5.2 program.³⁶ The tree represents similarities between the 16S rRNA genes.

Biochemical tests

The API 50 CH kit (bioMérieux), comprising 49 unique biochemical tests appropriate for Bacillus species, was used for biochemical testing of the RSM2e3 strain, as described in the manufacturer's instructions. The complete test was performed 3 times.

Animal Immunization and Virus Challenge

Groups of 40 female BALB/c mice at 4–6 weeks were administered orally on day 1, 2 and 3, respectively, with 1.5 × 10¹⁰ spore/mouse of PBS-suspended RSM2e3 or wild-type PY79 control. Mice injected with PBS were included as blank control. Mice were then boosted for 3 continuous days at week 4, with the last boost at week 5 post-1st immunization. Sera were collected at week 5, 7, 9, 11, 13, 15, and 17 after the 1st immunization to detect M2e-specific IgG antibodies and subtypes, and splenocytes were isolated at week 17 to detect cellular immune responses. Twenty mice/group were intranasally (i.n.) infected with 10³ tissue culture infectious dose (TCID₅₀) of A/PR/8/34(H1N1) influenza virus, respectively, at week 17 post-1st vaccination. Ten infected mice in each group were observed daily for 14 d for clinical symptoms, including body weight changes, and survival rates were calculated. On day 5 post-challenge, intestine and lung washes were obtained from 5 mice in each group to examine sIgA, and another 5 infected mice in each group were sacrificed to detect viral titers in lungs.

ELISA

M2e-specific IgG, subtypes and sIgA antibodies were detected by ELISA in mouse sera and lung and intestine washes, as previously described.³⁷ Briefly, 96-well ELISA plates were respectively precoated with synthesized M2e peptide (SLLTEVETPIRNEWGCRCNDSSD) (SBS, China) at 4°C overnight and blocked with 3% BSA at 37°C for 1 h. Serially diluted mouse sera or washes were added to the plates and incubated at 37°C for 1 h, followed by 5 washes in PBST. Bound antibodies were respectively incubated with HRP-conjugated anti-mouse IgG, IgG1, IgG2a, or IgA (1:5,000, Santa Cruz Biotechnology) at 37°C for 1 h. The reaction was visualized by substrate 3,3’,5,5’-tetramethylbenzidine (TMB) (Invitrogen) and stopped by 1N H₂SO₄. The absorbance at 450 nm (A450) was measured by ELISA plate reader (Tecan, San Jose, CA).

Intracellular Cytokine Staining and Flow Cytometry Analysis

T cell responses in immunized mice were detected by intracellular cytokine staining followed by flow cytometry analysis, as previously described.³⁸ Briefly, splenocytes (10⁷) were stimulated with or without M2e peptide for 8 h at 37°C with 5% CO₂ in the presence of GolgiPlug™ containing brefeldin A (1 μl/ml; BD Biosciences, San Jose, CA). The cells were fixed using a Cytofix/Cytoperm™ PlusKit in accordance with manufacturer's protocols (BD Biosciences) and stained directly with conjugated anti-mouse-CD3 (FITC), anti-mouse-CD4 (PerCP), anti-mouse-IL-4 (PE), and anti-mouse-IFN-γ (PE) (BD Biosciences) for 30 min at 4°C. The stained cells were detected using Guava easyCyte 6HT-2L flow cytometer (Merck Millipore), and the data were analyzed by FlowJo software (Tree Star, Inc..).

Virus Titer Detection

Virus titers were detected in challenged mouse lungs, as previously described.³⁹ Briefly, lung tissues from euthanized mice were aseptically removed and homogenized in minimal essential medium (MEM) plus antibiotics to achieve 10% (w/v) suspensions of lungs. Serial dilutions of samples were added in quadruplicate to monolayers of MDCK cells preplated in 96-well cell culture plates and allowed to absorb for 2 h at 37°C with 5% CO₂. Supernatants were then removed and replaced with MEM plus antibiotics, and MDCK cells were incubated as above for 72 h. Virus cytopathic effect (CPE) was observed daily, and virus titer was determined by HA test. For the HA test, 50 μl of 0.5% turkey red blood cells were added to 50 μl of cell culture supernatant and incubated at room temperature for 30 min. Wells containing HA were scored as positive. The virus titer was calculated by the Reed-Muench method and expressed as Log₁₀TCID₅₀/g of lung tissues.

Statistical Analysis

Statistical analysis was performed using the GraphPad Prism 5 software. Data were presented as mean ± standard error (SE). Statistical significance between different vaccination groups was calculated by the Student's t test. The significance between survival curves was analyzed by Kaplan-Meier survival analysis with log-rank test. Values of P < 0.05 were considered significant.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Funding

This study was supported by the Innovation Grant of Beijing Institute of Microbiology and Epidemiology (2012CXJJ016), and the grant from the National Institute of Allergy And Infectious Diseases of the National Institutes of Health (R21AI111152).