Sublingual immunization with adenovirus F protein-based vaccines stimulates protective immunity against botulinum neurotoxin A intoxication

Abstract

Sublingual (s.l.) vaccination is an efficient way to induce elevated levels of systemic and mucosal immune responses. To mediate mucosal uptake, ovalbumin (OVA) was genetically fused to adenovirus 2 fiber protein (OVA-Ad2F) to assess whether s.l. immunization was as effective as an alternative route of vaccination. Ad2F-delivered vaccines were efficiently taken up by dendritic cells and migrated mostly to submaxillary gland lymph nodes, which could readily stimulate OVA-specific CD4⁺ T cells. OVA-Ad2F + cholera toxin (CT)-immunized mice elicited significantly higher OVA-specific serum IgG, IgA and mucosal IgA antibodies among the tested immunization groups. These were supported by elevated OVA-specific IgG and IgA antibody-forming cells. A mixed T_h-cell response was induced as evident by the enhanced IL-4, IL-10, IFN-γ and TNF-α-specific cytokine-forming cells. To assess whether this approach can stimulate neutralizing antibodies, immunizations were performed with the protein encumbering the β-trefoil domain of C-terminus heavy chain (Hcβtre) from botulinum neurotoxin A (BoNT/A) as well as when fused to Ad2F. Hcβtre-Ad2F + CT-dosed mice showed the greatest serum IgG, IgA and mucosal IgA titers among the immunization groups. Hcβtre-Ad2F alone also induced elevated antibody production in contrast to Hcβtre alone. Plasma from Hcβtre + CT- and Hcβtre-Ad2F + CT-immunized groups neutralized BoNT/A and protected mice from BoNT/A intoxication. Most importantly, Hcβtre-Ad2F + CT-immunized mice were protected from BoNT/A intoxication relative to Hcβtre + CT-immunized mice, which only showed ∼60% protection. This study shows that s.l. immunization with Ad2F-based vaccines is effective in conferring protective immunity.

Introduction

Mucosal surfaces serve as major portals of entry of pathogens and have evolved to provide the first line of defense (1, 2). Vaccinating these mucosal surfaces proves effective for stimulating protective immune responses when compared with conventional parenteral immunization methods (2). Moreover, often, CTL responses can be induced with mucosal immunization (3–5). Due to compartmentalization of the ‘common mucosal immune system’, selection of route of immunization is a major factor to facilitate favorable immune responses at the target site (6–9). For example, while oral immunization stimulates strong secretory (S) IgA responses in the small intestine, proximal colon and mammary glands, it is less capable of inducing these responses in the respiratory and the genitourinary tracts (2). In contrast, nasal immunization can stimulate protective immunity locally, as evidenced with live-attenuated influenza vaccines (10), as well as at distal mucosal sites (11, 12). An advantage of nasal immunizations is their ability to stimulate elevated SIgA responses in the airways and genitourinary tract, but they are less effective in providing intestinal immunity (13, 14). Nasal immunization is often favored because of its ease of administration, less antigen is required and it stimulates both humoral and cell-mediated immunity (2). Despite these advantages, there are safety concerns challenging the universality of nasal immunization in humans (13–15). One concern is antigen redirection into the central nervous system as a consequence of coadministered adjuvant that can result in an increased incidence of Bell’s Palsy in humans (16–19). Thus, alternative means to circumvent neuronal uptake of vaccines and/or adjuvants, either by vaccine formulation or by alternative routes of immunization, are needed.

The sublingual (s.l.) mucosa is a part of the oral mucosa and provides an attractive non-invasive alternative to nasal and oral immunizations. Despite the s.l. route of immunization having a number of advantages over other mucosal routes with regard to safety and ability to stimulate humoral and cell-mediated immunity, the lack of adjuvants suitable for human use imposes a major obstacle for adopting the s.l. route to immunize humans (20). The s.l. route is already approved for use in humans to deliver low-molecular weight drugs (21) and immunotherapeutics for the treatment of type I allergies, such as rhinitis and asthma (22–26). Importantly, there is no evidence of anaphylactic shock in treating allergies by the s.l. route in contrast to subcutaneous immunotherapies (27–30). For immunization, one study showed that s.l. administration of the model antigen, ovalbumin (OVA), in combination with the mucosal adjuvant, cholera toxin (CT), elicits elevated serum and mucosal antibodies as well as CTL responses (31). Furthermore, s.l. vaccination with an inactivated or live influenza virus vaccine is as effective as intra-nasal immunization and confers protection from lethal viral challenge with absence of neuronal redirection of antigen (32). In addition, s.l. immunization with non-replicating human papilloma virus virus-like particles induces antibody-forming cells (AFCs) and CTLs in female genital mucosa, and it is protective from lethal genital papilloma virus infection in mice (33).

Clostridium botulinum is an anaerobic Gram-positive spore forming bacterium (34). It produces a potent lethal toxin, botulinum neurotoxin (BoNT), and seven different serotypes (A–G) have been identified (34, 35). BoNTs exert their pathological effects by binding to peripheral cholinergic nerve endings and subsequently inhibiting acetylcholine release. Blocking acetylcholine release prevents muscle contraction and results in airway obstruction or paralysis of respiratory muscles (36, 37). BoNTs are initially synthesized as single polypeptide progenitors (∼150 kDa) and, following proteolytic cleavage, activate the toxin consisting of C-terminal heavy (H) chain (∼100 kDa) and N-terminal light (L) chain (∼50 kDa). L chain exerts the toxic effects via its zinc-dependent endoprotease activity, disabling the docking and fusion of acetylcholine-containing vesicles to the plasma membrane. H chain facilitates host receptor binding to ultimately translocate L chain (38, 39). In addition to neutralization of L chain activity (40), antibodies elicited to the host receptor-binding domain in the carboxy-terminus of H chain (Hc) are protective against BoNT intoxication (41–44). Treatment of botulism requires administration of antitoxin, which cannot reverse paralysis and can cause hypersensitivity in some recipients (45). Prophylactic immunization with a toxoid-based investigational vaccine is one option but is only available for laboratory workers or individuals at high risk to exposure; thus, we lack a botulinum vaccine (46). Even the current toxoid-based pentavalent vaccine has several shortcomings, including the ability to obtain pure toxoid devoid of culture contaminants, the loss of essential neutralizing epitopes by formalin treatment during toxoid preparation and the multiple doses required to sustain elevated levels of immune antibodies (46). Thus, recent efforts for botulism vaccines have been focused on using recombinant subunit approaches, circumventing the need for working with active toxin and eliminating loss of neutralizing epitopes attributed to formalin inactivation (47–50).

In an effort to minimize the relevant fraction of Hc responsible for stimulating neutralizing antibodies, our studies have shown that the ∼50-kDa β-trefoil (βtre) structure conserved among all BoNTs (51) contained within Hc, referred to as Hcβtre, possesses the ability to stimulate antibodies capable of neutralizing native BoNTs (52–54). Recently, we reported intra-nasal vaccines exploiting recombinant adenovirus 2 fiber protein’s (Ad2F) targeting capabilities since this protein is responsible for initial viral attachment to host epithelial cells and subsequently viral entry into the cells. Vaccines incorporating Ad2F greatly enhance onset of mucosal and systemic immunity (53, 55) as a consequence of vaccine retention in the nasal mucosa, resulting in greater immunogenicity (53, 55). Thus, we hypothesize that by exploiting Ad2F’s mucosal targeting property combined with Hcβtre’s strong immunogenicity potently enhances the stimulation of neutralizing antibodies.

Given the advantages of s.l. immunization, we queried whether Ad2F can be adapted for s.l. vaccine delivery. As such, Ad2F was tested using the model antigen, OVA, as part of the fusion protein, OVA-Ad2F. S.l. vaccination with OVA-Ad2 greatly enhanced OVA-specific systemic and mucosal antibody responses relative to OVA alone. In addition, when OVA-Ad2F was coadministered with mucosal adjuvant CT, OVA-specific immune responses were significantly more augmented than OVA + CT. To assess the functional attributes of this finding, BoNT challenge studies were performed in s.l.-vaccinated mice using the Hcβtre/A-Ad2F vaccine combined with CT adjuvant. The results showed 100% survival against BoNT/A intoxication, while mice dosed with Hcβtre + CT showed only ∼50% protection. These data strongly support the notion successful mucosal vaccine targeting enhances vaccines’ efficacy.

Methods

Animals and immunizations

BALB/c 6- to 8-week old female mice (Frederick Cancer Research Facility, National Cancer Institute, Frederick, MD, USA) and a breeding colony of DO11.10 TCR transgenic mice (The Jackson Laboratory, Bar Harbor, ME, USA) were maintained at Montana State University Animal Resources Center in individual ventilated cages under HEPA-filtered barrier conditions. All procedures were compliant with institutional policies for animal health and well-being.

Vaccines were prepared, as previously described (55), and given by the s.l. route similar to that previously described (32, 33). To prevent swallowing during s.l. immunization, mice were anaesthetized with 200 μl of ketamine/xylazine/acepromisine (430 μg/86 μg/28 μg in 100 μl; Vedco, St Joseph, MO, USA) cocktail. The s.l. administration volume was limited to 7 μl for enhanced absorption. For the OVA immunogenicity studies, 25 μg of OVA (tissue grade OVA; Sigma–Aldrich, St Louis, MO, USA), or 50 μg of OVA-Ad2F, representing an equimolar amount of OVA, was given three times either in the absence or in the presence of 2 μg of CT (List Biological Laboratories, Campbell, CA, USA) on days 0, 7 and 14. For the botulism studies, mice were given equimolar amount of vaccines on days 0, 7, 14, 21 and 42 corresponding to 25 μg of Hcβtre or 50 μg of Hcβtre-Ad2F per dose either without or with 2 μg of CT. For the vaccine uptake study, mice were given 50 μg of the recombinant protein, Red2-Ad2F, a genetic fusion between Red2 fluorescent protein and Ad2F (55), + 2 μg CT.

Adoptive transfer and measurement of in vivo T-cell proliferation

Splenic and head and neck lymph node (LN) (HNLN) CD4⁺ T cells were purified from DO11.10 mice using a mouse CD4⁺ negative isolation kit (Invitrogen, Oslo, Norway). Isolated (5 × 10⁶ per mouse) CD4⁺ T cells were adoptively transferred to naive BALB/c mice via tail vein, and 24 h later, recipient mice were s.l. vaccinated, as described above. Seven days later, total lymphocytes were isolated from spleens, submaxillary gland LNs (SMLNs), cervical LNs (CLNs) and parotid gland LNs (PRLNs), as previously described (43, 44), and >95% viability was determined by trypan blue exclusion. Purified lymphocytes (2.5 × 10⁵ per well) were re-suspended in 200 μl complete medium (CM) and restimulated with 1 μg ml⁻¹ OVA_323–339 peptide (Biosynthesis, Inc., Lewisville, TX, USA) for 3 days and cultures were pulsed with 1.0 μCi per well [³H]-thymidine during the last 12 h of culture. Incorporated radioactivity in harvested samples was measured, as previously described (56).

Purification of dendritic cells and in vitro T-cell proliferation

Mice were given 50 μg of OVA-Ad2F + 2 μg CT s.l., and 4 h later, dendritic cells (DCs) were purified from SMLNs using mouse plasmacytoid DC and mouse CD11c magnetic bead isolation kits (Miltenyi Biotec, Bergisch Gladbach, Germany). Purified DCs (2 × 10⁴ per well) were cultured in CM with DO11.10 CD4⁺ T cells (2 × 10⁵ per well) for 3 days in the presence or absence of OVA_323–339 peptide, and cultures were pulsed with 1.0 μCi per well [³H]-thymidine during the last 12 h of culture. Incorporated radioactivity in harvested samples was measured, as previously described (56).

FACS analysis

Mononuclear cell suspensions were prepared from spleens and HNLNs. Cells were stained for FACS analysis using conventional methods. Cells were analyzed using mAbs conjugated with Alexa 488 (mPDCA), PE-Cy7 (CD11c), eFluor 780 (CD11b), PerCP-Cy 5.5 (TCRβ) and eFluor 450 (MHC II) (all from eBioscience, San Diego, CA, USA).

Anti-OVA and anti-Hcβtre antibody ELISAs

To evaluate OVA- or Hcβtre-specific endpoint antibody titers, immune plasma and fecal extracts were obtained weekly, and nasal and vaginal wash samples were collected at termination of study. All samples were prepared and processed, as previously described (55). Anti-OVA (57) and anti-Hcβtre (53) endpoint antibody titers were determined by standard ELISA method, as previously described. Following washing, HRP-conjugated secondary goat-anti-IgG, -IgA, -IgG1, -IgG2a or -IgG2b antibodies (Southern Biotechnology Associates, Birmingham, AL, USA) were added to the wells and incubated for 90 min at 37°C. After washing, ABTS substrate (Moss, Inc., Pasadena, CA, USA) was added and incubated for 60 min at room temperature. Endpoint titers were defined as the highest reciprocal of dilution of sample, giving an absorbance at OD₄₁₅ >0.100 OD units above negative control.

OVA-specific B-cell ELISPOT

OVA-specific and total IgG and IgA AFCs were enumerated by B cell ELISPOT assay, as previously described (53, 55, 58).

Cytokine-specific ELISPOT

Total mononuclear cells from spleens, HNLNs and MLNs were re-suspended (5 × 10⁶ ml⁻¹) in a CM [RPMI 1640 containing 10% of FBS (Atlanta Biologicals, Lawrenceville, GA, USA) and the supplements (Invitrogen-Life Technologies, Grand Island, NY, USA): 2 mM l-glutamine, 100 U ml⁻¹ penicillin, 100 μg ml⁻¹ streptomycin, 1 mM sodium pyruvate and 0.1 mM non-essential amino acids] and re-stimulated with 10 μg ml⁻¹ OVA for 3 days at 37°C in a humidified chamber under 5% CO₂. After washing with CM, two different concentrations of cell suspension were prepared (2.4 × 10⁵ and 8 × 10⁴) and applied for IL-4, IL-6, IL-10, IL-17, IFN-γ and TNF-α-specific ELISPOT assay following previously described protocols (48, 53, 59). Cytokine-forming cells (CFCs) were enumerated using a dissection microscope (Leica).

BoNT mouse neutralization assay

To determine the efficacy of the BoNT/A vaccines, pooled immune plasma from each of four immunization groups (Hcβtre, Hcβtre + CT, Hcβtre-Ad2F or Hcβtre-Ad2F + CT) was evaluated 1 week after the last immunization. From each group, 200 μl of pooled immune plasma was diluted 1:10 or 1:40 in PBS + 0.2% gelatin and incubated for 1 h at room temperature with 2.5 or 5 LD₅₀ of BoNT/A (2.7 × 10⁸ MLD₅₀/mg, Lot no. A072210-01; Metabiologics, Madison, WI, USA), and then the antibody–toxin complex was injected intra-peritoneally into naive BALB/c mice (53, 55). Animals were observed hourly for the signs of BoNT intoxication, including difficulty breathing and lack of mobility. When signs of neuromuscular intoxication became evident, mice were euthanized in accordance with the Institutional Animal Care and Use Committee.

BoNT/A intoxication challenge assay

To determine the efficacy of the s.l.-immunized mice with the Hcβtre, Hcβtre + CT, Hcβtre-Ad2F or Hcβtre-Ad2F + CT vaccines 2 weeks after the last immunization, mice were intra-peritoneally challenged with 1000 or 5000 LD₅₀ of BoNT/A. Animals were observed hourly for the first day and then twice daily until day 9 post-challenge. When the signs of neuromuscular intoxication become evident, mice were euthanized in accordance with the Institutional Animal Care and Use Committee.

Statistical analysis

Results are expressed as means ± SE. To evaluate differences between variations in antibody titers, AFCs and CFCs, an analysis of variance followed by Tukey’s method was applied and discerned to the 95% confidence interval (CI). The Kaplan–Meier method was used to obtain the survival fraction following intoxication with BoNT/A. Adopting the Mantel–Haenszel log rank test, the P-values for statistical differences among surviving BoNT/A challenges with plasma from Hcβtre, Hcβtre + CT, Hcβtre-Ad2F and Hcβtre-Ad2F + CT-vaccinated mice or BoNT/A challenges of mice immunized with these vaccines were discerned at the 95% CI.

Results

S.l. immunization with OVA-Ad2F + CT primes SMLN CD4⁺ T cells and elevated antibody titers

To determine whether the s.l. vaccination with OVA by the epithelial cell targeting moiety, Ad2F, enhanced T-cell priming in HNLNs, OVA-specific transgenic CD4⁺ T cells obtained from DO11.10 mice were adoptively transferred to naive BALB/c mice. Mice were subsequently immunized by the s.l. route with equimolar amounts of OVA (25 μg) or OVA-Ad2F (50 μg), either without or with CT. Lymphocytes from spleens, SMLNs, CLNs and PRLNs were harvested 7 days after immunization and analyzed for their proliferative capacity by [³H]-thymidine incorporation. Notably, proliferation was only detected in SMLN (Fig. 1B) and not by splenic (Fig. 1A), PRLN (Fig. 1C) or CLN (Fig. 1D) lymphocytes. Equivalent OVA-specific proliferative responses were observed by lymphocytes from OVA + CT and OVA-Ad2F only immunized mice (Fig. 1B); however, OVA-Ad2F + CT-immunized mice showed the best proliferative responses (P ≤ 0.001). This evidence suggests the SMLN is the primary site of CD4⁺ T-cell activation following s.l. immunization.

Fig. 1.

s.l. immunization with the model antigen, OVA, genetically fused to Ad2F (OVA-Ad2F) induced proliferation of adoptively transferred DO11.10 CD4⁺ T cells. Purified DO11.10 CD4⁺ T cells were adoptively transferred to naive BALB/c mice, and 24 h later, mice were s.l. immunized with an equimolar amount of OVA (25 μg), OVA + CT (25 + 2 μg), OVA-Ad2F (50 μg) or OVA-Ad2F + CT (50 + 2 μg). Seven days after challenge, total lymphocytes were isolated from spleens, SMLNs, PRLNs and CLNs and cultured in CM in the presence of OVA_323–339 peptide. Proliferation was analyzed by [³H]-thymidine incorporation. Data represent mean ± SEM of individual mice (n = 10). *P ≤ 0.001 versus OVA; ^†P ≤ 0.001 versus OVA + CT; ⁺P ≤ 0.001 versus OVA-Ad2F. (E) To assess OVA-specific IgG antibody titers, plasma from individual mice was collected on day 10 post-challenge. Data represent mean ± SEM. *P ≤ 0.001 represents significant differences versus OVA-immunized mice; ^†P ≤ 0.001 indicates statistical differences between versus OVA + CT-immunized mice; ⁺P ≤ 0.001 indicates statistical differences between OVA-Ad2F- and OVA-Ad2F + CT-immunized mice.

Evaluation of antibody production 10 days after immunization revealed the greatest anti-OVA endpoint antibody titers were obtained with mice vaccinated with OVA-Ad2F + CT (Fig. 1E). Moreover, mice vaccinated only with OVA-Ad2F (without adjuvant) also showed elevated anti-OVA antibody titers, and these were significantly enhanced (P ≤ 0.001) relative to mice vaccinated with OVA + CT (Fig. 1E). Collectively, these results show that s.l. vaccination with Ad2F fusion proteins offers an alternative means to vaccinate mucosally.

Ad2F mediates antigen uptake by DCs for delivery to SMLNs

As shown in Fig. 1(B), SMLNs appear to be the major site for T-cell activation following Ad2F-mediated s.l. immunization. To assess how antigen reaches SMLNs after s.l. Ad2F targeting, mice were s.l. immunized with Red2-Ad2F (55) + CT, and 1.5 h later, mononuclear cells from spleens, SMLNs, PRLNs and CLNs were obtained and analyzed by FACS for the location of Red2 fluorescence as well as associated cell type. Minimal to no Red2 fluorescence was detected in mononuclear cells from spleens, PRLNs and CLNs (data not shown). Otherwise, DCs (conventional, lymphoid and myeloid) in SMLNs showed elevated Red2 fluorescence (Fig. 2A), and this evidence further supports the contention that SMLNs are the major site of antigen deposition following s.l. Ad2F-targeted antigen delivery by DCs. To examine the T-cell proliferative capacity of the OVA-bearing DCs in SMLNs, mice were s.l. dosed with OVA-Ad2F + CT, and 4 h later, DCs were purified from SMLNs. Purified DCs (2 × 10⁴) were co-cultured with 2 × 10⁵ DO11.10 CD4⁺ T cells in the presence or absence of OVA_323–339. Significant elevation of T-cell proliferation was observed in the presence of CD11c⁺ DCs, even in the absence of OVA_323–339 when compared with T cells alone or co-cultured with mPDCA⁺ DCs (Fig. 2B). Thus, these studies show that DCs facilitate Ad2F-mediated antigen delivery to SMLNs following s.l. immunization.

Fig. 2.

Antigen uptake by DCs targets SMLNs following s.l. Ad2F immunization. (A) Mice were given Red2Ad2F + CT, and 1.5 h later, total lymphocytes from spleens, SMLNs, PRLNs and CLNs were harvested and analyzed for the presence of Red2 fluorescence. Fluorescence was only detected in (A) SMLNs but not in spleens, PRLNs and CLNs (data not shown) and associated only with MHC class II^high (conventional, lymphoid and myeloid) DCs, not plasmacytoid (mPDCA⁺) DCs. (B) DCs obtained from SMLNs of mice given s.l. OVA-Ad2F + CT induced the greatest T-cell proliferation. Mice s.l. immunized with OVA-Ad2F + CT, and 4 h later, DCs from SMLNs were co-cultured with T cells obtained from DO11.10 mice in the presence or absence of OVA_323–339. T-cell proliferation was measured by [³H]-thymidine incorporation. Data represent mean ± SEM and are representative of three experiments. *P ≤ 0.001 versus no DCs (none).

S.l. OVA-Ad2F targeting enhances systemic and mucosal antibody production

To ascertain OVA-Ad2F’s effectiveness in naive mice, groups of BALB/c mice were immunized by the s.l. route with OVA or OVA-Ad2F either without or with CT on days 0, 7 and 14. A kinetic analysis was performed, revealing OVA-Ad2F alone could effectively induce elevated systemic and mucosal antibodies, as opposed to OVA, which only stimulated marginal plasma IgG and IgA and SIgA antibody responses (Fig. 3A–C). An earlier onset of elevated plasma IgG antibody production was observed on day 14 with OVA-Ad2F-immunized mice than was observed with OVA alone or OVA + CT (Fig. 3A). However, as with mice coadministered with CT, both plasma IgA and SIgA responses were delayed until day 21 when differences among the immunization groups were observed, particularly, for mice immunized with OVA-Ad2F + CT, which showed the best IgA titers (Fig. 3B and C). Likewise, OVA-Ad2F + CT-immunized mice showed more rapid and robust plasma IgG responses than any of the immunization groups (Fig. 3A). Moreover, mice immunized with OVA-Ad2F + CT showed significantly greater IgG and IgA antibody responses than OVA + CT (Fig. 3A–C). Evaluation of IgG subclass antibodies responses revealed enhanced IgG1, IgG2a and IgG2b titers by the OVA-Ad2F + CT-immunized mice when compared with OVA + CT-immunized mice (Fig. 3D–F). Thus, these studies show that s.l. immunization with OVA-Ad2F, particularly when combined with adjuvant, can effectively stimulate elevated systemic and mucosal immune antibody titers.

Fig. 3.

S.l. immunization with OVA-Ad2F enhanced OVA-specific systemic and mucosal antibody production (A, B and C). BALB/c mice were immunized on days 0, 7 and 14 with OVA or OVA-Ad2F, either alone or in combination with CT. Plasma and fecal samples were collected weekly starting at day 14 post-primary immunization until day 42, and antibodies titers were evaluated by standard ELISA method. The genetic fusion of antigen with mucosal targeting molecule, Ad2F, enhanced systemic (A and B) and mucosal (C) antibody production, either in the presence or in the absence of CT when compared with mice immunized with OVA only. S.l. antigen targeting by Ad2F in combination with CT was synergistic in the production of systemic and mucosal antibodies. The statistical significance was not evident between OVA + CT- and OVA-Ad2F-immunized groups, except early in the IgG response. (D, E and F) S.l. immunization with OVA-Ad2F, either with CT or alone, elicited elevated IgG subclass antibodies. Data represent mean ± SEM (n = 10). *P ≤ 0.001 and **P < 0.05 represent significant differences versus OVA-immunized mice; ^†P ≤ 0.001 and ^††P < 0.05 represent statistical differences versus OVA + CT-immunized mice; ⁺P ≤ 0.001 and ⁺⁺P < 0.05 represent statistical differences between OVA-Ad2F- and OVA-Ad2F + CT-immunized mice.

S.l. immunization enhances antibody production in nasal and vaginal washes

To assess OVA-specific antibody titers in mucosal secretions, nasal and vaginal washes were performed 1 week after the last immunization. OVA-specific IgG and IgA antibodies in nasal and vaginal washes were significantly enhanced in mice s.l. immunized mice with OVA-Ad2F versus OVA alone (P ≤ 0.001; Fig. 4A and B). Mice immunized with OVA-Ad2F + CT elicited significantly higher mucosal IgG and IgA antibody titers than OVA + CT-immunized mice (P ≤ 0.001; Fig. 4A and B). Thus, s.l. immunization with OVA-Ad2F effectively stimulates both regional and distal mucosal tissues.

Fig. 4.

S.l. immunization with OVA-Ad2F induced elevated IgG and SIgA antibody secretion in nasal and gentiourinary tracts when compared with OVA (A and B). Significantly higher production of IgG and SIgA was also detected in combined administration of OVA-Ad2F with CT when compared with OVA + CT. There were no significant differences between OVA + CT- and OVA-Ad2F-immunized mice. (A) Vaginal and (B) nasal wash samples were collected 1 week after the third immunization and measured by OVA-specific IgG and IgA ELISA. Data are expressed as mean ± SEM (n = 15). *P ≤ 0.001 represents statistical differences in IgG and SIgA anti-OVA endpoint titers versus OVA-immunized mice; ^†P ≤ 0.001 represents significant differences versus OVA + CT-immunized mice; ⁺P ≤ 0.001 represents significant differences between OVA-Ad2F- and OVA-Ad2F + CT-immunized mice.

S.l. immunization with OVA-Ad2F enhances IgG and IgA AFC responses

To examine the source of antibody-producing B cells following s.l. immunization, mice were immunized as described above, and 1 week after their last immunization, OVA-specific B ELISPOT assay was performed on various lymphoid tissues, including spleens, HNLNs, MLNs, Peyer’s patch (PPs), intestinal lamina propria (iLP), reproductive tract (RT) and nasal passages (NPs) (Fig. 5). Elevated levels of OVA-specific IgG AFCs were detected in spleens, HNLNs, MLNs, iLP, RT and NPs from mice immunized with OVA-Ad2F and compared with those mice given OVA alone (P ≤ 0.001; Fig. 5A and B). Likewise, OVA-Ad2F + CT-immunized mice induced significantly greater IgG AFCs in spleens (P ≤ 0.001), HNLNs (P ≤ 0.001), iLP (P < 0.05), RT (P ≤ 0.001) and NPs (P < 0.05) than OVA + CT-immunized mice (Fig. 5A and B). S.l. immunization with OVA-Ad2F elevated IgA AFCs in spleens (P ≤ 0.001), HNLNs (P ≤ 0.001), MLNs (P < 0.05), PPs (P ≤ 0.001), iLP (P ≤ 0.001), RT (P ≤ 0.001) and NPs (P ≤ 0.001) compared with OVA-immunized mice (Fig. 5E and F). However, IgA AFC responses were only significantly enhanced (P < 0.05) in the HNLNs and NPs by the OVA-Ad2F + CT-immunized mice when compared with OVA + CT-immunized mice (Fig. 5E and F).

Fig. 5.

S.l. immunization with OVA-Ad2F enhances the number OVA-specific IgG and IgA AFCs in spleens and mucosal tissues. Spleens, head and neck lymph nodes, mesenteric LNs, PPs, iLP, RT and NPs were measured for (A, B, E and F) IgG and IgA OVA-specific and (C and D) total IgG and (G and H) IgA AFC responses among the four immunization groups: OVA, OVA-Ad2F, OVA + CT and OVA-Ad2F + CT. Mice were immunized as described in Fig. 3. S.l. immunization with OVA-Ad2F + CT greatly enhanced the number of IgG AFCs in spleens, HNLNs, PPs, iLP, RT and NPs; increased IgA AFCs were only detected in HNLNs and NPs when compared with OVA + CT-immunized mice. The significant difference between OVA-Ad2F and OVA + CT was only found in IgG AFCs obtained from HNLNs and iLP. The B ELISPOT analysis was conducted 7 days after the last of the three immunizations. Data represent the mean ± SEM of three separate experiments (n = 15). *P ≤ 0.001; **P < 0.05 represent significant differences in IgG and IgA AFC responses versus OVA-immunized mice; ^†P ≤ 0.001 and ^††P < 0.05 represent significant differences versus OVA + CT-immunized mice; ⁺P ≤ 0.001 and ⁺⁺P < 0.05 represent significant differences between OVA-Ad2F- and OVA-Ad2F + CT-immunized mice.

S.l. immunization with OVA-Ad2F stimulates a mixed T_h-cell response

To discern the supportive T_h cells responsible for the observed antibody responses subsequent s.l. OVA-Ad2F immunization, mice were immunized, as described above, and 1 week after their last immunization, splenic, HNLN and MLN lymphocytes were evaluated for their cytokine production by cytokine-specific T-cell ELISPOT for IL-4, IL-6, IL-10, IL-17, IFN-γ and TNF-α. Mucosal targeting by Ad2F clearly stimulated enhanced T_h1- and T_h2-cell responses, particularly, by the OVA-Ad2F + CT-immunized mice (Fig. 6). Relative to their OVA-immunized mice counterparts, OVA-Ad2F-immunized mice with or without CT showed marked elevations in IFN-γ- and TNF-α-producing CFCs in spleens, HNLNs and MLNs (Fig. 6A and B). In addition, they showed significantly elevated numbers of IL-4- and IL-10-producing CFCs in the spleens and HNLNs but not as much in the MLNs (Fig. 6C and D). Evaluation of T_h17 cells paradoxically showed enhanced IL-6- and IL-17-producing CFCs in the MLNs and minimal CFC responses by spleens and HNLNs from mice immunized with OVA-Ad2F without or with CT relative to their OVA-immunized control groups (Fig. 6E and F). Thus, these results show s.l. vaccine targeting by Ad2F stimulates mixed T_h1- and T_h2-cell responses supportive of the observed increases in OVA-specific antibody titers.

Fig. 6.

S.l. immunization with OVA-Ad2F enhanced the number of (A) IFN-γ, (B) TNF-α, (C) IL-4 and (D) IL-10 CFCs in spleens and HNLNs when compared with OVA. In MLNs, only IL-10 and TNF-α CFCs were increased. Coadministration of CT with OVA induced higher production of (A–F) IFN-γ, TNF-α, IL-4, IL-10, IL-6 and IL-17 in various tissues when compared with OVA + CT. Cytokine ELISPOT assays were conducted 1 week after the third immunization. Data represent mean ± SEM of three separate experiments of individual mice (n = 15). *P ≤ 0.001 and **P < 0.05 represent significant differences versus OVA-immunized mice; ^†P ≤ 0.001 and ^††P < 0.05 represent significant differences versus OVA + CT-immunized mice; ⁺P ≤ 0.001 and ⁺⁺P < 0.05 represent significant differences between OVA-Ad2F- and OVA-Ad2F + CT-immunized mice.

S.l. immunization with Hcβtre-Ad2F also stimulates elevated IgG and IgA antigen-specific antibodies

To evaluate whether s.l. immunization with other Ad2F-based vaccines can elicit stimulate neutralizing antibodies, additional studies were performed with our botulinum vaccines to assess protection against BoNT/A challenge. Groups of mice were immunized with equimolar amounts of Hcβre or Hcβtre-Ad2F without or with CT on days 7, 14, 21 and 42, and plasma and fecal samples were monitored until day 63. Mice immunized with Hcβtre-Ad2F without or with CT showed significant elevations (P < 0.05) in their plasma IgG and IgA anti-Hcβtre antibody titers relative to their OVA-immunized control groups (Fig. 7A and B). Notably, mice immunized with Hcβtre only showed poor plasma antibody responses, as opposed to mice immunized with Hcβtre-Ad2F only, which induced modest IgG and IgA anti-Hcβtre titers. Enhanced production of SIgA was observed following s.l. administration with Hcβtre-Ad2F without or with CT relative to Hcβtre-only immunized mice (Fig. 7C). Interestingly, the Hcβtre-Ad2F-only mice showed elevated SIgA and in some instances as much as the groups co-immunized with CT. Analysis of IgG subclass responses revealed Hcβtre-Ad2F-immunized mice induced significantly higher Hcβtre-specific IgG1, IgG2a and IgG2b antibodies, either in the presence or in the absence of CT compared with their counterparts (Fig. 7D–F). In addition, s.l. immunization with Hcβtre + CT also enhanced IgG1, IgG2a and IgG2b antibodies, implying that Hcβtre is effective when combined with adjuvant and given by this route (Fig. 7D–F). Thus, elevated Hcβtre-specific antibody production can be induced upon s.l. immunization with our BoNT vaccines.

Fig. 7.

Systemic and mucosal anti-Hcβtre antibody production was enhanced by s.l. immunization with Hcβtre genetically fused to Ad2F. (A to C) Hcβtre-Ad2F + CT elicited elevated antibody titers relative to Hcβtre-Ad2F alone. (D–F) Hcβtre-specific IgG subclass responses were enhanced by s.l. targeting with Ad2F, either in the presence or in the absence of CT when compared with Hcβtre- or Hcβtre-Ad2F. BALB/c mice were s.l. immunized on days 0, 7, 14, 21 and 42 with Hcβtre or Hcβtre-Ad2F, either in the presence or in the absence of CT. Plasma (A) IgG, (B) IgA, (C) fecal IgA and (D–F) IgG subclasses antibody titers against Hcβtre were determined by standard ELISA method. Data represent mean ± SEM (n = 10). *P ≤ 0.001 and **P < 0.05 represent significant differences versus Hcβtre-immunized mice; ^†P ≤ 0.001 and ^††P < 0.05 represent significant differences versus in Hcβtre + CT-immunized; ⁺P ≤ 0.001 and ⁺⁺P < 0.05 represent significant differences between Hcβtre-Ad2F- and Hcβtre-Ad2F + CT-immunized mice.

Mouse BoNT neutralization assay shows that plasma from Hcβtre or Hcβtre-Ad2F + CT protects against BoNT/A intoxication

To determine if the induced antibodies were neutralizing, mice were immunized, as described in Fig. 7, but required an additional booster on day 42 post-primary immunization. Fourteen days after their last immunization, individual plasma was collected, diluted 1:40, tested for their ability to neutralize native 5.0 LD₅₀ of BoNT/A and administered into naive BALB/c mice to assess their neutralization capacity. Plasma collected from mice given Hcβtre or Hcβtre-Ad2F alone showed poor protection, although Hcβtre-Ad2F-immunized mice protected ∼25% of the challenged mice, while naive or Hcβtre-immunized succumbed within 16 h (Fig. 8A). Otherwise, mice challenged with BoNT/A plus pooled plasma from either adjuvanted Hcβtre or Hcβtre-Ad2F-dosed mice were fully protected (Fig. 8A).

Fig. 8.

S.l. immunization with Hcβtre + CT or Hcβtre-Ad2F + CT neutralizes BoNT/A and s.l. immunization with Hcβtre-Ad2F + CT confers complete protection against intra-peritoneal challenge with BoNT/A. (A) Plasma from mice on day 56 post-primary immunization (Fig. 7A) was evaluated for their BoNT-neutralizing capacity in a mouse neutralization assay. Pooled plasma from each immunization group was diluted 1:40 in PBS containing 0.2% gelatin and incubated with 5 LD₅₀ BoNT/A for 60 min before intra-peritoneal injection into naive BALB/c mice. All mice given naive plasma mixture with BoNT/A succumbed within 8 h. In contrast, pooled plasma obtained from Hcβtre + CT- or Hcβtre-Ad2F + CT-immunized mice neutralized BoNT/A and conferred complete protection from BoNT/A intoxication. The survival fraction derived from the Hcβtre-dosed plasma and BoNT/A mixture-treated mice was not statistically different from naive plasma-dosed group. Immune plasma from Hcβtre-Ad2F-dosed mice showed 25% survival rate, which was statistically different from naive plasma-treated mice; n = 4–6 mice per group. *P = 0.005 and **P < 0.05 for survival fractions from treatment groups compared with mice given naive plasma. (B and C) BALB/c mice from Fig. 7A were challenged intra-peritoneally on day 63 with (B) 1000 or (C) 5000 LD₅₀ BoNT/A and monitored for survival for 9 days. Only the (B and C) Hcβtre-Ad2F + CT-immunized mice showed 100% survival, and the Hcβtre + CT-treated groups showed ∼57% survival. The survival rates for the (B) non-adjuvanted groups were ≤25%, and P-values were determined by the Kaplan–Meier method. Survival fractions compared with BoNT/A-challenged naive mice were obtained, and significance was determined: *P ≤ 0.001 and **P < 0.05 versus naive (PBS-dosed) mice.

S.l. immunization with Hcβtre-Ad2F + CT conferred complete protection against BoNT/A intoxication

To directly assess whether mice immunized with our Hcβtre-based vaccines can confer protection against BoNT/A challenge, the same mice described in Fig. 7, on day 63, were challenged intra-peritoneally with 1000 or 5000 LD₅₀ of BoNT/A. The Hcβtre-Ad2F + CT-immunized mice were completely protected from 5000 to 1000 LD₅₀ challenge during the experimental period in contrast to naive mice, which succumbed to intoxication within 3 h (Fig. 8B and C). In the absence of CT, s.l. immunization with Hcβtre failed to show protective efficacy against 1000 LD₅₀ challenge, and mice immunized with Hcβtre-Ad2F alone showed only 20% survival (Fig. 8B). In the presence of CT, Hcβtre-dosed mice showed better protection, but it was still less effective than Hcβtre-Ad2F + CT-immunized mice (Fig. 8B and C). Thus, these studies show that s.l. targeting with our Hcβtre-Ad2F in the presence of CT confers complete protection against BoNT/A intoxication and performs better than the vaccine lacking the Ad2F-targeting moiety.

Discussion

The efficiency of s.l. delivery of drugs and small molecular weight molecules has been well established for uptake into the blood stream to avoid hepatic metabolism and digestion by gastric acid (21, 60). In fact, targeting the s.l. mucosa in humans is currently used for treating type I allergies (22–26). The allergen-specific immune suppression might be partially due to s.l. mucosa-specific Langerhans cell-like DCs, which have been suggested to have intrinsic tolerogenic properties with constitutive elevated expression of the high-affinity IgE receptor, Fc_ϵRI (61). The production of IL-10, TGF-β and indoleamine 2-dioxygenase by s.l. DCs as a result of IgE engagement with its receptor on DCs’ surface may lead to the down-regulation of inflammatory T-cell proliferation (61–63). Furthermore, there is growing evidence suggesting that the s.l. route, when vaccine antigens are coadministered with the mucosal adjuvant CT, induces a similar magnitude of mucosal and systemic immune responses when compared with nasal immunizations (31). Importantly, s.l. immunization has shown no neuronal redirection of antigen or adjuvant (31, 32). When assessing whether our previously defined nasal epithelium targeting molecule, Ad2F (53, 55), would also be as effective when applied to the s.l. mucosa, herein this study, we found s.l. administration of OVA-Ad2F alone effectively induced mucosal and systemic antibody responses in the absence of coadministered adjuvant with a similar magnitude to s.l. administered OVA when given with the potent mucosal adjuvant, CT. When combined with adjuvant, our BoNT vaccine, Hcβtre-Ad2F, showed marked enhancement in stimulating neutralizing antibodies to protect against BoNT intoxication.

In our previous study, Ad2F was shown to bind the nasal epithelium to facilitate antigen uptake and stimulate robust antibody responses when given nasally to mice (53, 55) and rabbits (54). Thus, based on these previous observations, we hypothesized that the Ad2F fusion vaccines could be applied by the s.l route to stimulate elevated mucosal and systemic antibody responses. Initial work assessed whether s.l. OVA-Ad2F could prime transgenic OVA-specific CD4⁺ T cells. Naive BALB/c mice were adoptively transferred with DO11.10 CD4⁺ T cells, and mice subsequently were primed with OVA-Ad2F alone or with co-administered CT. As a result, OVA-specific IgG antibodies were induced to greater levels in descending order to OVA-Ad2F, OVA + CT or to OVA. In a similar fashion, elevated T-cell proliferative responses were only observed with lymphocytes obtained from SMLNs of OVA + CT, OVA-Ad2F and OVA-Ad2F + CT-dosed mice, and this evidence suggests SMLNs are the primary site for T-cell activation following s.l. immunization. This was further supported by Red2 fluorescence detected in SMLN DCs 1.5 h after s.l. immunization. Thus, these findings support the notion that SMLNs are the site where OVA-bearing DCs ultimately migrate for initial T-cell activation following s.l. immunization, as others have found (31).

BALB/c mice s.l. immunized with OVA-Ad2F, either in the presence or in the absence of CT, elicited elevated plasma OVA-specific IgG and IgA antibody titers as well as elevated SIgA antibody titer in various mucosal secretions. These responses were supported by elevated numbers of OVA-specific AFCs in various mucosal tissues. Importantly, even in the absence of CT, s.l. immunization with recombinant OVA-Ad2F protein induced systemic antibody responses as great as mice immunized with OVA + CT in contrast to marginal responses by mice immunized with OVA alone. Moreover, s.l. OVA-Ad2F immunization alone stimulated elevated SIgA antibodies in vaginal and nasal washes, and these IgA antibodies were of a similar magnitude to s.l. OVA + CT-immunized mice. These observations suggest Ad2F possibly possesses adjuvant-like activity in addition to its ability to bind host cells or possibly this activity is a consequence of host binding, which in turn could reduce the amount of adjuvant required to stimulate protective immunity. Current studies are assessing these possibilities. S.l.-targeted OVA delivery combined with CT stimulated a mixed T_h1- and T_h2-cell response, as evidenced by the enhanced IFN-γ, TNF-α, IL-4 and IL-10. Similar mixed T_h-cell responses were previously found when using Ad2F-based vaccines (53, 55). Collectively, these observations indicate targeted s.l. vaccine delivery by Ad2F could be a promising alternative for mucosal immunization capable of stimulating both mucosal and systemic immune responses.

BoNTs are the most toxic agents for humans (64), and in the presence of neutralizing antibodies, botulism can be prevented. Active immunization or treatment with antitoxins can prevent further intoxication (45, 46). While only available for high-risk individuals, the current formalin-inactivated pentavalent BoNToxoid vaccine has several shortcomings. BoNToxoids have declining potency that requires multiple boosters to maintain neutralizing activity (64). This is also evident in experimental animal studies showing a lack of or poor efficacy against BoNT challenge (54, 55, 64). One attribute accountable for its poor performance is the use of formalin, which cross-links many of the neutralizing epitopes and diminishes the ability to stimulate a protective antibody response (65). Thus, alternative vaccines for botulism are needed, and current efforts are seeking to develop prophylactic vaccines using non-toxic subunit and DNA vaccine approaches to render elevated and sustained neutralizing antibodies (47–50).

Along these lines, significant efforts have focused on using the Hc from the various BoNT serotypes since these possess a fair number of neutralizing epitopes (52). Antibodies induced to Hc can confer protective immunity against lethal BoNT challenges, as evidenced in various animal and non-human primate models (41, 42, 49). Adapted Hc vaccines in various vaccine vectors, including adenovirus and attenuated Salmonella, also have been found to confer protective immunity against lethal BoNT challenge in rodent models (43, 44). Although Hc-encoding DNA vaccine approaches have been used, these elicit lower neutralizing antibody titers and are relatively inadequate when compared with purified Hc protein vaccines (65).

A mucosal BoNT vaccine would be advantageous by providing multiple levels of protection since the most likely method of disseminating BoNTs would be by aerosolization (66). Thus, intra-nasal or s.l. route of delivery is the most attractive means to stimulate protective immunity in the respiratory mucosa. Our previous studies demonstrate that intra-nasal targeting with Hcβtre-Ad2F, which contains the neutralizing epitope for BoNT/A, elicits strong mucosal and systemic antibody responses, and the inclusion of CT adjuvant further enhances protective immunity against lethal BoNT challenge (54, 55). These studies suggest the use of Ad2F for mucosal targeting, rather than intact Ad, is still effective in stimulating protective immunity. Likewise, in this current study, s.l. targeting by Ad2F also proved effective in stimulating protective antibodies against BoNT/A and was mediated by inducing elevated mucosal and systemic anti-Hcβtre antibodies by Hcβtre-Ad2F. Interestingly, s.l. Hcβtre-Ad2F, in the absence of CT, induced considerable mucosal and systemic antibodies, albeit, still less IgG responses than Hcβtre + CT-immunized mice but with similar SIgA titers. However, when coadministered with CT, Hcβtre-Ad2F induced more potent mucosal and systemic antibody titers than mice immunized with equimolar concentrations of Hcβtre + CT.

Despite the elevations in antibody titers, neutralizing antibodies must also parallel these titers (54); in this regard, s.l. immunization with either Hcβtre or Hcβtre-Ad2F in the presence of CT elicited robust systemic and mucosal antibody responses. When assessed in a mouse neutralization assay, plasma from both groups showed similar neutralization potential since both could resist 5.0 LD₅₀ challenge. However, the true potency of protection was evident from the studies in which the immunized mice were directly challenged. As such, only the mice immunized with Hcβtre-Ad2F + CT were protected against 5000 LD₅₀ challenge, while only 60% survival was observed with Hcβtre + CT-immunized mice. Although some neutralizing antibodies were induced by mice immunized with Hcβtre-Ad2F only, this vaccine did require coadministered adjuvant for optimal protection, as evidenced by the mouse neutralization assay and by direct BoNT challenge.

In summary, these data show s.l. Ad2F-based vaccines provide a promising non-invasive alternative to intra-nasal immunization still capable of eliciting protective antibodies against BoNT by retaining the vaccine sufficiently in the mucosal epithelium for subsequent delivery to regional LNs by DCs. Such vaccine delivery allows for improved antibody responses supported by mixed T_h1 and T_h2 cells. Importantly, these findings suggest an adjuvant-like activity imparted by Ad2F and further implicate the significance of mucosal vaccine targeting as a possible means to replace adjuvant use.

Funding

This work was supported by Public Health Service grants (AI-078938); Rocky Mountain Research Center of Excellence, National Institute of Health (U54 AI-06537); and in part by Montana Agricultural Experiment Station and U.S. Department of Agriculture Formula Funds. The Department of Immunology and Infectious Diseases’ flow cytometry facility was in part supported by National Institute of Health/National Center for Research Resources Center of Biomedical Research Excellence (P20 RR-020185).