|Home | About | Journals | Submit | Contact Us | Français|
In this study, we examine the potential of a combinatorial vaccine consisting of the lead-candidate antigens for the next generations of vaccines against anthrax (rPA) and plague (F1-V) with the specific objective of determining synergy or interference between the vaccine components when they are administered separately or together by both traditional parenteral immunization (SC) and mucosal immunization (IN) in the presence of appropriate adjuvants. The most significant findings of the study reported here are that 1) a combinatorial vaccine consisting of equal amounts of F1-V and rPA administered SC is effective at eliciting a robust serum and bronchoalveolar lavage (BAL) antigen-specific IgG and IgG1 response against both antigens in immunized animals, and when administered IN, a robust antigen-specific IgG2a response in the serum and BAL is also induced; 2) there were few instances where either synergy or interference was observed in the combined vaccine administered by either route and those differences occurred soon after the final immunization and were not sustained over time; 3) IN immunization was as effective as SC immunization for induction of antigen specific serum and BAL antibody responses using the same amount of antigen; 4) the IgG1/IgG2a ratios suggest a strongly biased Type 2 response following SC immunization, while IN immunization produced a more balanced Type 1/Type 2 response, and 5) the IgG1/IgG2a ratio was influenced by the route of immunization, the adjuvant employed, and the nature of the antigen. As with previously published studies, there were still detectable levels of circulating anti-F1-V and anti-rPA even six months post-primary immunization. These studies provide important insights into the development of new generation biodefense vaccines.
An ideal vaccine against potential agents of biological warfare/bioterrorism should be safe, easy to deliver, provide long-lasting protection, require only one or a few doses, and provide protective immunity against different agents. In an imminent or post-release bioterrorism event, it is possible that multiple vaccines will need to be administered simultaneously. In that case, vaccines that combine antigens from multiple different organisms would allow for the rapid immunization of large groups of individuals against multiple potential biological threats.
The first logical combination to examine for a combinatorial biodefense vaccine would be recombinant protective antigen (rPA) from Bacillus anthracis and the recombinant fusion protein, F1-V, from Yersinia pestis. B. anthracis and Y. pestis, the causative agents of anthrax and plague, respectively, are two select agents against which new vaccines are currently being developed [1-13]. Previously licensed vaccines against both diseases were reactogenic, required multiple doses, and in the case of plague, evidence indicates that the vaccine did not fully protect against the pneumonic form of the disease [1, 2, 14, 15]. For these reasons, there has been a great deal of interest in developing safer, more effective vaccines that could be used to easily immunize civilian and military personnel either before or after exposure to these agents.
Anthrax is caused by the Gram-positive bacterium B. anthracis. In nature, anthrax is a disease that herbivores or other mammals acquire after contact with soil contaminated with B. anthracis spores. Human disease results from contact with infected animals, contaminated animal products, or after exposure to accidentally or intentionally released spores of B. anthracis. There are three types of human anthrax infection: cutaneous, gastrointestinal and pulmonary. Although, each type can progress to fatal systemic anthrax, untreated pulmonary anthrax acquired by inhalation is the most severe form [16, 17].
The primary virulence factors of B. anthracis are thought to be anthrax toxin and a glutamic acid capsule, both of which are encoded by large plasmids (pXO1and pXO2). The anthrax toxin is a tripartite protein, composed of the protective antigen (PA), the Lethal Factor (LF) and Edema Factor (EF). These toxins conform to the AB model of bacterial exotoxins, where PA is the binding subunit and both LF and EF are alternative catalytic subunits. Lethal toxin is assembled from PA and LF and is believed to be primarily responsible for the acute effects observed in anthrax. Edema toxin is assembled from PA and EF and it is believed to be responsible for the edematous lesions at the site of infection. These toxins are taken into the cell after PA interacts with its cellular receptor, which leads to receptor-mediated endocytosis followed by toxin translocation into the cytosol. LF is a zinc metalloprotease that cleaves several mitogen-activated protein kinase kinases (MAP-KK) disrupting signal transduction pathways normally used to respond to invading pathogens [16, 18, 19] and inducing inflammatory changes in macrophages, leading to production of proinflammatory cytokines, activation of the oxidative burst pathway, and the release of oxygen intermediates. EF is a calmodulin-dependent adenylate cyclase that increases intracellular levels of cAMP [16, 20, 21]. PA is the main component of the two licensed human anthrax vaccines, Anthrax Vaccine Adsorbed (AVA) in the United States and Anthrax Vaccine Precipitated (AVP) in the United Kingdom, and previous studies have shown that a strong antibody response to PA is protective against anthrax disease . This protection could be mediated through several different mechanisms including inhibition of the activity of both lethal and edema toxin or through anti-spore activity, facilitating phagocytosis and spore killing, and inhibition of spore germination [22, 23].
The etiologic agent of plague is the Gram-negative bacterium Yersinia pestis. The natural route of transmission of Y. pestis from one host to another is either directly through infected respiratory droplets or via a flea vector. Plague is endemic in some regions of the world and outbreaks occasionally occur as a consequence of natural disasters. Y. pestis is also a concern as one of the microorganisms with potential for use against civilian or military populations as a biological warfare agent. In that case, the pneumonic form of plague would be the most likely outcome. This form of plague is particularly devastating because of the rapidity of onset, the high mortality, and the rapid spread of the disease. Both live attenuated and killed plague vaccines have been used in man, although questions remain about their safety and relative efficacy, especially against the pneumonic form of infection . For these reasons development of improved vaccines against plague is a high priority. F1 is a capsular protein located on the surface of the bacterium and the V antigen is a component of the Y. pestis Type III secretion system. These antigens have been shown to induce a protective immune response in mice both separately and as a fusion protein [5, 12]. The plasmid encoding the recombinant F1-V codes for the entire structural F1 gene minus its stop codon fused at its carboxy-terminus to the structural V gene .
In this study, we examine the potential of a combinatorial vaccine consisting of rPA and F1-V with the specific objective of determining synergy or interference between the vaccine components when they are administered separately or together by both traditional parenteral immunization and mucosal immunization in the presence of appropriate adjuvants.
Groups of 8-10 week old female BALB/c mice (Charles River Laboratories) were immunized intranasally (IN) or subcutaneously (SC) with 5 μg rPA, 5 μg F1-V, or 5 μg each of rPA and F1-V three times at weekly intervals. Mice immunized IN received the vaccine formulation admixed with 5 μg of the mucosal adjuvant LT(R192G) [3, 4, 25-34] in a final volume of 8 μl in one nostril following brief exposure to Isofluorane. Mice immunized SC received the vaccine formulation adsorbed to 0.19 mg of an aluminum hydroxide adjuvant (2.0% Alhydrogel batch no. 3275; Superfos Biosector, Vedbaek, Denmark) in a final volume of 100 μl injected beneath the skin between the shoulder blades. rPA was obtained from List Biological Laboratories, Inc. F1-V was obtained from the United States Army Medical Research Institute of Infectious Diseases [5, 24] or purified from the same construct in our laboratory . LT(R192G) was prepared in our laboratory by galactose-affinity chromatography as previously described .
Five animals from each group were sacrificed by CO2 inhalation one week after the last immunization and then at monthly intervals for a total of 6 months. Blood was collected from each animal by cardiac puncture. Bronchoalveolar lavage (BAL) was obtained by exposing the trachea and making a small incision into which an 18-gauge needle was inserted and secured. Through this cannula the lungs were repeatedly lavaged by slowly injecting and withdrawing 1 ml of phosphate buffered saline (PBS) supplemented with a protease inhibitor cocktail (Roche Laboratories).
Individual serum and BAL samples were examined for anti-rPA and anti-F1-V total IgG, IgG1, and IgG2a antibodies using a quantitative ELISA as described previously . For quantitative analysis, concentrations of sample total IgG, IgG1, and IgG2a were determined by non-linear regression from a standard curve of mouse myeloma IgG1 or IgG2a (Sigma Chemical Co., Saint Louis, MO, USA) serially diluted as a standard on each ELISA plate. The results are expressed as the mean concentration±S.E.M. Ratios of IgG1 to IgG2a were determined by dividing the antigen-specific concentration of IgG1 by the concentration of IgG2a for each sample. The value for each group is presented as the mean ratio.
Statistical analysis was performed using a one-way analysis of variance followed by a Bonferroni Multiple Comparisons post-test on selected groups. P values of ≤ 0.05 were considered significant. P values were inserted in the figures above any results that were found to be statistically significantly different. Statistical analysis was performed using Prism 4 software (GraphPad Inc.)
The purpose of this group of experiments was to compare the serum and BAL anti-F1-V and anti-rPA responses when the antigens were administered separately or in combination subcutaneously. Mice were immunized SC three times at weekly intervals with F1-V or rPA or F1-V+rPA adsorbed to alum. Groups of five animals from each group were sacrificed beginning one week after the final immunization and monthly thereafter for five additional months. Antigen-specific antibody responses were determined by ELISA and a comparison was made between the mean response of the group receiving the antigen alone and the group receiving the antigens in combination. Three possible outcomes were considered: synergy (an increased antigen-specific response when the antigens were combined compared to the response observed when the antigen was administered alone), interference (a decreased antigen-specific response when the antigens were combined compared to the response observed when the antigen was administered alone), or no effect (no significant change in the antigen-specific response when the antigens were combined compared to the response observed when the antigen was administered alone). Total IgG, IgG1, and IgG2a responses were measured in both the systemic and lung compartments of immunized and naïve mice. There were no detectable antigen specific responses in either the serum or lungs of the naïve mice.
As seen in Fig. 1A, there was no significant difference in the serum anti-F1-V IgG response at any time point over the six months of the experiment between animals immunized with F1-V alone (closed bars) and those immunized with F1-V in combination with rPA (open bars). Similarly, with the exception of the first time point, the anti-rPA IgG response in animals immunized with rPA alone was not significantly different from the anti-rPA IgG response observed in animals immunized with the combination vaccine (Fig. 1B). At the first time point, one week following the last immunization, there appeared to be a synergistic effect for rPA in the combined vaccine, which diminished over time. The differences observed in anti-rPA responses between animals immunized with rPA alone when compared to animals immunized with the combinatorial vaccine at the second time point were not statistically significant. It is also important to note that the serum anti-rPA IgG response is consistently greater than the serum anti-F1-V IgG response in these animals, especially in the combined vaccine.
In the BAL, the anti-F1-V IgG (Fig. 1C) and anti-rPA IgG (Fig. 1D) responses were never significantly different between the animals immunized SC with either single antigen alone or with the two antigens in combination (no effect). The differences between the anti-F1-V IgG and anti-rPA IgG responses observed in the BAL were not as pronounced as in the serum.
We also performed an isotype analysis of the serum and BAL anti-F1-V and anti-rPA responses following immunization with each antigen alone or with the combined vaccine. Understanding the effect of combining vaccines on the isotype distribution could be particularly important in plague vaccines, since previous studies by Williamson et al.  have shown that anti-F1 and anti-V IgG1 responses correlate significantly with protection. Figs. 2A and 2B examine the serum anti-F1-V and serum anti-rPA IgG1 responses, respectively, in animals immunized SC with each antigen alone or in combination. As seen in Fig. 2A, the only significant difference observed in the anti-F1-V serum IgG1 response was at the two month time point with the F1-V alone response being significantly greater than the response observed with the combined vaccine (interference). This effect was not observed at the first time point or subsequent to the second time point. Interestingly, the anti-rPA IgG1 response observed in animals at two months was significantly greater in the combined vaccine than the response observed in animals immunized with rPA alone (synergy). As with the anti-F1-V IgG1 response, this effect was not observed at the first time point or subsequent to the second time point.
As seen in Fig. 2C, there was also a significantly greater BAL anti-F1-V IgG1 response in animals immunized with F1-V alone when compared to animals immunized with the combined vaccine (interference) immediately following the final immunization. This effect was maintained through the second month but was absent thereafter as the BAL anti-F1-V IgG1 response to the combined vaccine continued to increase at the third month. Despite the apparently greater BAL anti-rPA responses observed in animals receiving the combined vaccine compared to animals immunized with rPA alone (Fig. 2D), there were no statistically significant differences in the responses of these two groups.
We next examined the serum anti-F1-V and anti-rPA IgG2a responses in animals immunized SC with F1-V or rPA alone or in combination (data not shown). In these studies, there were no significant differences in the serum anti-F1-V IgG2a responses observed in animals immunized with F1-V alone and the responses observed in animals immunized with the combined vaccine at any time point throughout the six months of the study. The only significant difference observed in the serum anti-rPA IgG2a response was at the first time point immediately following the last immunization, at which point the anti-rPA response in the combined vaccine was significantly greater than the response seen in animals immunized with rPA alone (synergy).
Importantly, no anti-F1-V or anti-rPA IgG2a was detectable in the BAL fluids of animals immunized SC with either antigen alone or in combination, despite the presence of significant antigen-specific IgG and IgG1 in these fluids, suggesting a strong Type 2 bias in the BAL following SC immunization with these antigens. As will be seen below, this was not the case following IN immunization.
The purpose of this group of experiments was to compare the serum and BAL anti-F1-V and anti-rPA responses when the antigens were administered separately or in combination intranasally. Mice were immunized IN three times at weekly intervals with F1-V or rPA or F1-V+rPA admixed with the mucosal adjuvant LT(R192G) and groups of five animals from each group were sacrificed beginning one week after the final immunization and monthly thereafter for five additional months. Antigen-specific antibody responses were determined by ELISA and a comparison was made between the mean response of the group receiving the antigen alone and the group receiving the antigens in combination.
In these studies, animals immunized IN with either F1-V or rPA, alone or in the combined vaccine, developed high levels anti-F1-V and anti-rPA serum and BAL IgG and IgG1. Importantly, there were no significant differences in antigen-specific IgG responses between groups of animals immunized with either antigen alone or in combination by this route. In contrast to the absence of detectable BAL anti-F1-V or anti-rPA IgG2a when the antigens were administered SC, there was a substantial BAL anti-F1-V IgG2a response and anti-rPA IgG2a response when the antigens were administered IN in combination with LT(R192G), suggesting a more balanced Type 1/Type 2 response. We also observed that the serum anti-rPA IgG2a response was consistently greater than the serum anti-F1-V IgG2a response following IN immunization.
The purpose of this comparison was to determine the antigen-specific IgG1/IgG2a ratios of the serum and BAL anti-F1-V and anti-rPA responses when the antigens were administered alone or in combination. We specifically addressed whether changing the formulation of the vaccine or the route of immunization would alter IgG1/IgG2a ratio for each antigen. As seen in Fig. 3, there were no significant differences in the antigen-specific IgG1/IgG2a ratios in the serum (Fig. 3A) or BAL (Fig. 3B) of animals immunized SC when comparing either F1-V or rPA alone to the combined vaccine. Moreover, all SC immunized animals demonstrated a highly Type 2 polarized response, as reflected in high IgG1/IgG2a ratios (>10:1). This was especially true for BAL responses in animals immunized SC (IgG1/IgG2a ratios >1,000:1) owing to the lack of a BAL IgG2a response following SC immunization. By contrast, there were both antigen-specific and route-specific differences observed in animals immunized IN in the presence of LT(R192G); as seen in Fig. Fig.3C3C and and3D,3D, all IN immunized animals had a more balanced Type 1 /Type 2 response (IgG1/IgG2a ratios 0.1 – 10). Interestingly, the serum anti-F1-V response following IN immunization was skewed towards a Type 2 bias while the response against rPA was skewed towards a Type 1 bias (Fig. 3C). These observations are important because they demonstrate that not only the route of immunization, or likely the adjuvant included in the vaccine formulation, but also the nature of the antigen itself can influence the immunologic outcome. Following IN immunization, the BAL response to both antigens skewed towards a Type 1 bias (Fig. 3D), possibly indicating a difference in the ability of the different antibody subtypes to transudate across the lung epithelia or a difference in the different subtypes within the lung environment.
Combining vaccines to decrease the number of immunizations and to increase vaccine coverage is not a new concept in vaccine development. Combination vaccines have been used very successfully for decades in both pediatric and traveler's applications. As such, the possible implications of simultaneously administering multiple vaccines have been an issue in these fields as well . While combined vaccines have been successful in the past, as indicated by the success of the DTaP (diphtheria, tetanus, and acellular pertussis) and MMR (measles, mumps, rubella) vaccines as well as others , it has been shown that certain combinations of vaccines can either enhance or diminish the immune response to co-administered antigens. Laboratory tests and clinical trials in these areas have generally shown no deleterious effect , but occasionally immunological interference between individual components of different combined vaccines [37-41], and rarely an enhanced immune response to co-administered vaccines, is observed. For example, Gil et al.  demonstrated that immunization with a hepatitis A vaccine alone or in combination with a yellow fever vaccine resulted in similar seroconversion rates and levels of geometric mean titers against each vaccine. By contrast, Frey et al.  demonstrated that combining a Hepatitis A vaccine with a Hepatitis B vaccine was not as effective as administering the two vaccines by separate but concurrent immunization. These investigators observed a diminished response against the hepatitis B component of the combined vaccine, characterized by lower seroprotection rates and geometric mean titers, as compared to when the vaccines were administered concurrently. The researchers also indicated that the seroprotection observed after separate but concurrent immunization was lower than that seen in other studies when the hepatitis B vaccine was given completely independently of the hepatitis A vaccine. These studies as well as many others, illustrate that immunological interference can occur not only by administering vaccine admixed in the same syringe, but also by administering the two vaccines at the same time in separate anatomical sites, and that all new combinations of vaccines need to be thoroughly tested before use in a clinical setting.
Studies performed by Griffin et al. have shown that separate but concurrent parenteral administration of a recombinant F1+V vaccine and an rPA vaccine results in no significant reduction in the immune response toward rF1 or rV and no decrease in protection against subcutaneous Y. pestis challenge. In addition, those studies showed that while concurrent administration of rPA did not significantly influenced the polarity of the immune response toward rF1 or rV, concurrent immunization with a live vaccine that typically stimulates a Type 1 response did result in a reversal of the polarity of the immune response toward each antigen, from a primarily IgG1- to IgG2a-biased antibody response. In spite of the switch in the polarity of the immune response to both rF1 and rV, these animals were fully protected against parenteral plague challenge. Those studies focused primarily on the effect on the immune responses to rF1 and rV and protection against plague challenge when the proteins and other biodefense vaccines were administered simultaneously by parenteral injection. Those studies did not include any effects on the primary immunogens contained in the concurrently administered vaccines or the effects of administering the vaccines by non-parenteral immunization.
The primary objective of the study reported here was to determine if the distribution, polarity, magnitude, or duration of the antibody response to rPA and F1-V were altered in a combined vaccine when compared to the response to the individual components administered singly. Since a primary consideration for development of an effective vaccine against these and other potential biological warfare agents is aerosol delivery [1, 42], we measured BAL fluids as well as serum samples for the presence of antigen specific antibodies. Antigen(s) were administered either subcutaneously adsorbed to an aluminum hydroxide adjuvant, the only adjuvant currently approved for human use in the United States, or mucosally in combination with LT(R192G), a derivative of the heat-labile toxin produced by some enteropathogenic strains of Escherichia coli and potent mucosal adjuvant in a variety of animal models and humans [3, 4, 25-34].
The most significant findings of the study reported here are that 1) a combinatorial vaccine consisting of equal amounts of F1-V and rPA administered SC is effective at eliciting a robust serum and bronchoalveolar lavage (BAL) antigen-specific IgG and IgG1 response against both antigens in immunized animals, and when administered IN, a robust antigen-specific IgG2a response in the serum and BAL is also induced; 2) there were few instances where either synergy or interference was observed in the combined vaccine administered by either route and those differences occurred soon after the final immunization and were not sustained over time; 3) IN immunization was as effective as SC immunization for induction of antigen specific serum and BAL antibody responses using the same amount of antigen; 4) the IgG1/IgG2a ratios suggest a strongly biased Type 2 response following SC immunization, while IN immunization produced a more balanced Type 1/Type 2 response, and 5) the IgG1/IgG2a ratio was influenced by the route of immunization, the adjuvant employed, and the nature of the antigen. As with previously published studies, there were still detectable levels of circulating anti-F1-V and anti-rPA even six months post-primary immunization.
The mechanisms behind any differences in the magnitude and polarity of the immune response to co-administered antigens may reflect the nature of the antigens themselves or differential processing following different routes of immunization. For instance, the uniformly greater response to rPA when compared to F1-V following SC immunization, and less so following IN immunization, indicates that rPA is more immunogenic when administered parenterally in the context of an aluminum hydroxide based adjuvant. This could indicate that the anti-inflammatory properties associated with the V antigen from Y. pestis are active in the F1-V fusion. Although it is has not been established whether or not V still possess its anti-inflammatory properties as an F1-V fusion, Nakajima et al. showed that fusion between the N-terminal portion of a staphylococcal protein A and the C-terminus of LcrV (PAV) still induced IL-10 and suppressed TNF-α and IFN-γ secretion when injected into normal Swiss Webster mice . A functional anti-inflammatory region of V could result in enhanced IL-10 production and suppression the immune response toward F1- V when administered SC. With the exception of the serum IgG2a response against each antigen, the magnitude of the anti-rPA and anti-F1-V responses following IN immunization were similar and did not favor rPA over F1-V, confirming that the route of immunization, or adjuvant employed, also plays a role in determining immunological outcome. This was also clear for BAL anti-rPA and anti-F1-V IgG2a responses, which developed following IN, but not SC, immunization.
In and of itself, the observation that it is possible to immunize with a combinatorial vaccine combining the lead-candidate antigens for the next generations of vaccines against anthrax and plague is both promising and important. In practical terms alone, especially in an imminent or post-release bioterrorism event, the ability to administer a combinatorial vaccine would greatly improve national preparedness. It remains to be determined if the combinatorial vaccine is as protective against aerosol challenge as the antigens administered individually.
This study was supported by Public Health Service grant AI055013 from the National Institute for Allergy and Infectious Diseases (J.D.C.). This research was performed by A.B.D. while on appointment as a U.S. Department of Homeland Security (DHS) Fellow under the DHS Scholarship and Fellowship Program, a program administered by the Oak Ridge Institute for Science and Education (ORISE) for DHS through an interagency agreement with the U.S. Department of Energy (DOE). ORISE is managed by Oak Ridge Associated Universities under DOE contract number DE-AC05-00OR22750. All opinions expressed in this paper are the authors' and do not necessarily reflect the policies and views of DHS, DOE, or ORISE.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.