|Home | About | Journals | Submit | Contact Us | Français|
Bacillus subtilis vaccine strains engineered to express either group A bovine or murine rotavirus VP6 were tested in adult mice for their ability to induce immune responses and provide protection against rotavirus challenge. Mice were inoculated intranasally with spores or vegetative cells of the recombinant strains of B. subtilis. To enhance mucosal immunity, whole cholera toxin (CT) or a mutant form (R192G) of Escherichia coli heat-labile toxin (mLT) were included as adjuvants. To evaluate vaccine efficacy, the immunized mice were challenged orally with EDIM EW murine rotavirus and monitored daily for 7 days for virus shedding in feces. Mice immunized with either VP6 spore or VP6 vegetative cell vaccines raised serum anti-VP6 IgG enzyme-linked immunosorbent assay (ELISA) titers, whereas only the VP6 spore vaccines generated fecal anti-VP6 IgA ELISA titers. Mice in groups that were immunized with VP6 spore vaccines plus CT or mLT showed significant reductions in virus shedding, whereas the groups of mice immunized with VP6 vegetative cell vaccines showed no difference in virus shedding compared with mice immunized with control spores or cells. These results demonstrate that intranasal inoculation with B. subtilis spore-based rotavirus vaccines is effective in generating protective immunity against rotavirus challenge in mice.
Group A rotaviruses are the most important cause of severe acute diarrhea in children less than 2 years old and have been estimated to be responsible for 362,000 to 592,000 (median, 440,000) deaths per year, primarily in the developing world (48). Because of the widespread nature of rotavirus disease and because the morbidity rates in developed and developing countries are similar, effective vaccines are considered key to their control. Although progress has been made in the development of live, attenuated oral vaccines, improved vaccines are still needed, particularly in developing countries where the burden of severe disease is the greatest but where the live oral vaccines have been the least effective to date (31, 33, 48). Although no longer marketed due to an association with intussusception (40), Rotashield (Wyeth-Ayerst), a rhesus-human reassortant rotavirus vaccine, which was 88% effective against severe diarrhea in rural areas of Venezuela (49), showed 18 to 46% efficacy in Peru and Brazil (31, 33). Two other live, attenuated oral vaccines have recently become available. Rotarix (GlaxoSmithKline), a human attenuated rotavirus vaccine, and RotaTeq (Merck), a human-bovine reassortant rotavirus vaccine, have performed well in some locales (13, 47), but in South Africa and Malawi, the efficacy was 40 to 80% (35). Other live, attenuated oral vaccines, such as vaccine 116E, based on rotaviruses obtained from asymptomatically infected newborns in Delhi, India, are also undergoing clinical trials (4, 5). Experience with previous candidate rotavirus vaccines, as well as vaccines against polio, cholera, and typhoid fever, has shown that the efficacy of live oral vaccines can be impaired in developing countries (47). In certain areas of India, the response to oral poliovirus vaccine has been so low that it has recently been recommended by the Indian Academy of Pediatrics that the oral vaccine be replaced with the injectable poliovirus vaccine (57).
Live bacterial vectors have also been used for oral immunization against rotavirus antigens. Oral immunization with recombinants of either Salmonella enterica serovar Typhimurium or Lactobacillus lactis expressing rotavirus virion protein 7 (VP7) generated immune responses after oral administration to mice, but no challenge studies were reported (50, 58).
A potential alternative to live oral immunization is live, intranasal inoculation. Attenuated Salmonella enterica serovar Typhi expressing tetanus toxin fragment C (TTFC) elicited protective immunity to tetanus toxin when the vaccine was given intranasally but not when the vaccine was given orally (22). Protection against tetanus toxin in mice nasally immunized with recombinant Lactococcus lactis or Lactobacillus plantarum expressing TTFC has also been shown (24, 42). Intranasal immunization was also effective for generating an immune response by an attenuated strain of Shigella flexneri expressing rotavirus VP4, but no challenge studies were done (34).
The use of Bacillus subtilis as a vehicle for vaccine antigen delivery is a relatively new approach to mucosal immunization (1, 16, 18, 43). The use of pathogenic bacteria as vectors has the disadvantage of requiring attenuation of the vector. Like the lactobacilli, B. subtilis is generally regarded as safe and is neither inherently pathogenic nor toxigenic to humans, animals, or plants (53). Moreover, B. subtilis is readily adaptable to genetic manipulation. Stable constructs can be integrated into the bacterial chromosome, making this bacterium a good candidate host for vaccine engineering.
Foreign antigens have been expressed in B. subtilis on the surface and inside vegetative cells and on the surfaces of spores. Both vegetative cells and spores have been used as delivery vectors (46), but the primary model used to date has been the spore form of B. subtilis displaying tetanus toxin antigen. A potential advantage of using spores as vectors is that bacterial spores are highly resistant to environmental stresses, such as extremes of heat, pH, desiccation, freezing and thawing, and radiation (41). Heterologous antigens displayed on the spore surface as a fusion product with spore coat proteins have been shown to elicit protective immune responses, for example, when spores displaying TTFC were given either orally or intranasally (16). Orally administered spores of B. subtilis can survive the gastrointestinal tracts of mice and may germinate in the intestine, but intestinal colonization is only for a brief period (6, 54). Thus, for oral immunization, several rounds of high doses of spores (≥1010 spores per dose) have been found to be necessary (1, 16, 43). Intranasal immunization yields similar levels of protection, with lower doses of spores expressing the antigen on the surfaces of spores (55).
In this report, we describe the preparation and application of B. subtilis spore and vegetative cell vaccines expressing group A rotavirus VP6, including proof-of-concept protection studies in a mouse challenge model. Intact rotavirus particles consist of triple-shelled capsids with outer, inner, and core layers. The outer capsid contains two viral proteins, VP4 and VP7. The inner capsid is formed by VP6, and the core consists of VP2, which encloses VP1 and VP3, along with the viral genome. The VP4 and VP7 proteins are the only two rotavirus proteins that are known to induce virus-neutralizing antibodies. Monoclonal antibodies to either VP4 or VP7 passively protect mice against rotavirus diarrhea. The major inner capsid protein VP6 plays a key role in the virion structure because of its interactions with both the outer capsid proteins VP4 and VP7 and the core protein VP2 (17). Rotavirus VP6 is antigenically conserved among group A rotaviruses (17). As a result, unlike with vaccines based on VP4 or VP7, serotype specificity should not be a problem with VP6-based vaccines, and effective vaccines based on VP6 could be expected to provide heterotypic protection against all group A rotaviruses. Our studies of VP6 DNA vaccines were the first to show protective immunity generated by a VP6-based vaccine (25, 26) and heterotypic protection against a murine rotavirus induced by a DNA vaccine expressing bovine VP6 (59). In subsequent studies, it was found that virus-like particles containing both VP2 and VP6 (VP2 is required for VLP formation) (45) or VP6 chimeric particles (7) also induced protection after treatment via the oral (45) or intranasal (7, 45) route. Monoclonal IgA antibodies to VP6 were also protective in a back-pack tumor model in mice (11). Here, we demonstrate that high titers of antibodies were raised in mice after intranasal administration of the B. subtilis vaccine strains and that protection against challenge with a murine rotavirus in the mice immunized with the spore vaccines but not in those immunized with the vegetative cell vaccines was obtained.
The B. subtilis strains used in this study were grown at 37°C in DS nutrient broth medium or in LB medium (19). The same media with addition of agar were used for growth of bacteria on plates. All B. subtilis strains were derivatives of strain 168 (trpC2). Escherichia coli strain JM107 (60) or strain BU1255 (dam-3 dcm-6 gal lac ara thr leu F+) (for experiments requiring unmethylated DNA) was used for isolation of plasmids and was grown in LB medium. The following antibiotics were used when appropriate: chloramphenicol, 2.5 μg/ml, or neomycin, 2.5 μg/ml, for B. subtilis strains and ampicillin, 50 μg/ml, or kanamycin, 25 μg/ml, for E. coli strains.
A series of B. subtilis promoter-containing vectors that can be integrated by a double-crossover event at the sacA locus of the B. subtilis chromosome was created based on plasmid pSac-Kan (39). First, pSac-Kan was digested with BseRI and BglII, blunt ended with the Klenow fragment of DNA polymerase I, and self-ligated to create pBB1364. The latter plasmid lost the 0.21-kb BseRI-BglII fragment containing one of the duplications within the original plasmid (the BglII site was reconstituted in the final construct). In several cloning steps, PCR fragments containing the strong B. subtilis Pspacp1/2 promoter, a modified form (see below) of the semisynthetic Pspac promoter (61), and the bovine and murine VP6-coding sequences, were inserted into pBB1364. The modified 0.27-kb Pspac fragment was synthesized in two steps, using pAG58 (28) as a template and mutagenic oligonucleotides as primers. As a result, two site-directed mutations were introduced, p1 [A(+1)G] and p2 [C(−12)T]; these mutations increased the strength of the promoter (data not shown). The p1 [A(+1)G] form was described previously (30).
The 1.23-kb PCR fragments containing the bovine or murine VP6-coding regions were synthesized using pCR.2.1-VP6bov (obtained from L. J. Saif, Ohio State University, Wooster, OH) or pBluescript-VP6mur (GenBank accession no. U36474; obtained from H. B. Greenberg, Stanford University, Palo Alto, CA), respectively, as templates and oligonucleotides oBB251 (5′-AAAAAACTAGTTTAATTAAAGGAGGAATTCAAAATGGATGTCCTGTACTC) and oBB232 (5′-GGTTAGAGCTCTATCATTTGACAAGCATGCTTC) as primers (the restriction sites are underlined, and the initiation codon is in bold). The oBB251 primer contained the ribosome-binding site, the ATG codon, and the intervening sequence of the B. subtilis gsiB gene. The oligonucleotides used as PCR primers introduced four silent mutations into the extreme ends of the VP6mur PCR fragment. The bovine and murine VP6 PCR fragments were cloned in pBB1375 to place them under the control of the Pspacp1/2 promoter and create pBB1428 and pBB1432, respectively. The bovine VP6 gene was obtained from the Indiana strain of bovine rotavirus (GenBank submission pending). The sequencing of the original murine VP6 clone, obtained from EDIM virus strain EW, showed that the gene contained 2 small inversions, GC instead of CG and TA instead of AT at positions 377 to 378 and 1186 to 1187, respectively, compared to the MRU36474 GenBank entry.
Plasmids pBB1428 and pBB1432 were introduced into the chromosome of B. subtilis by transformation through a double-crossover recombination event identified by screening for loss of a sacA::cat marker (39) due to its replacement by the plasmid-borne kan marker (Fig. (Fig.1).1). Strains BB2666 and BB2667 contained the sacA::(Pspacp1/2-VP6bov kan) and sacA::(Pspacp1/2-VP6mur kan) constructs, respectively. Strain BB2643 contained the sacA::kan mutation (39) (Fig. (Fig.1)1) and served as a control for strains BB2666 and BB2667.
Methods for agarose gel electrophoresis, use of restriction and DNA modification enzymes, DNA ligation, PCR, and electroporation of E. coli cells were as described previously (51). Isolation of chromosomal DNA and transformation of B. subtilis cells by chromosomal or plasmid DNA was as described previously (3). All cloned PCR-generated fragments were verified by sequencing.
B. subtilis strains grown overnight on DS agar plates were used to inoculate 4-liter cultures in DS medium (19). After incubation with shaking (200 rpm) at 37°C for 48 h, the mixture of spores and nonsporulating bacteria was harvested by centrifugation, washed with sterile deionized water, treated with egg white lysozyme (1 mg/ml) to kill nonsporulating cells, washed once with sterile deionized water, and stored at 4°C in sterile water. Spores were titrated by direct counting using a Petroff-Hauser chamber and by comparing the colony-forming abilities observed before and after a sample was heated to 80°C for 10 min. The typical yield was 3 × 1011 spores per liter of initial culture. Vegetative B. subtilis cells were prepared for use in immunization by growth at 37°C in LB medium until the absorbance at 600 nm reached 0.8 to 1.0. After being harvested by centrifugation, the cell pellets were resuspended at 50% (wt/vol) in phosphate-buffered saline (PBS). The vegetative cells used for inoculation were freshly prepared or reconstituted after storage as a lyophilized powder. We estimated the amounts of VP6 antigen to be 600 ng per dose for spores and 800 ng/dose for vegetative cells, as determined by immunoblotting.
Samples of vegetative cells were grown in LB medium to an optical density at 600 nm (OD600) of 0.7, concentrated 40 fold in 50 mM Tris-HCl, pH 8, containing 0.01 M phenylmethylsulfonyl fluoride (PMSF), sonicated, and clarified by centrifugation. Samples of supernatant fluids were boiled in denaturing buffer (0.06 M Tris, pH 6.8, 2% SDS, 10% glycerol, 0.1 M dithiothreitol [DTT]) for 5 min, and cell debris was removed by centrifugation. Samples of spores were harvested from plates, washed with water, and resuspended in water to give an OD600 of 96. Samples were boiled in denaturing buffer for 5 min to release spore-associated proteins and centrifuged to remove spores and cell debris. Additionally, samples of spores were diluted 1/5 in 50 mM Tris, pH 8, and sonicated to release spores from unlysed mother cells. The spores were pelleted by centrifugation, resuspended in denaturing buffer, and treated as described above. Protein samples were subjected to SDS-PAGE and transferred to Immobilon-P membranes (Millipore Corp.). Membrane blots were blocked with 2% bovine serum albumin (BSA) in TBST (50 mM Tris, pH 8, 0.15 M NaCl, 0.1% Tween 20) for 2 h, rinsed briefly with TBST, incubated with rabbit polyclonal anti-SA11 (simian rotavirus) antibody that we had previously prepared, diluted 1:5,000 in TBST, rinsed 4 times with TBST, incubated with goat anti-rabbit IgG conjugated with horseradish peroxidase (KPL; 1:10,000 in TBST), rinsed 4 times with TBST, treated with detection reagents (Amhersham ECL Plus Western blotting detection system), and exposed to HyBlot CL autoradiography film for 5 s to visualize protein bands.
BALB/c or DBA/2 female mice were purchased from Charles River Laboratories (Wilmington, MA). All animal handling was done in accordance with Tufts University IACUC protocols. Groups of five 6- to 8-week-old BALB/c or DBA/2 female mice were inoculated via the intranasal route with either spores or vegetative cells of the B. subtilis recombinants. Mice were inoculated with spore suspensions (1 × 109 CFU) or with 5 × 108 vegetative cells per mouse on days 0, 14, and 28. For nasal dosing, spores or vegetative cells in a volume of 20 μl were administrated using a plastic pipette tip without anesthesia. Blood samples were taken on days −1, 13, 27, and 41 postvaccination, and their sera were analyzed for anti-VP6 titers by an enzyme-linked immunosorbent assay (ELISA). The adjuvants used were cholera toxin (CT) (Sigma Chemical Co.) and an attenuated E. coli toxin, the mutant heat-labile toxin LT(R192G) (mLT), coinoculated with the vaccines. mLT has a single nucleotide substitution within the toxin A subunit coding sequence, leading to replacement of the arginine residue at position 192 with glycine. mLT was created as described previously (15). In preliminary experiments, the adjuvantivity of mLT was evaluated at different concentrations (1, 5, and 10 μg). Five micrograms and 10 μg of mLT showed similar adjuvantivity levels, but not 1 μg. Therefore, either 5 μg or 10 μg/dose of mLT was coinoculated with the vaccines. There were no abnormal signs in the mice (e.g., respiratory distress, weight loss, loss of appetite, lethargy, or ruffled fur) after immunization or throughout the duration of the experiments.
Antirotavirus antibody responses were measured by ELISA as we have previously described in detail (59). In brief, a hyperimmune rabbit anti-SA-11 rotavirus (SA-11 is a simian rotavirus) serum diluted 1:10,000 was used to coat ELISA plates for 1 h at 37°C. The coated plates were washed and then blocked for 1 h at 37°C with a solution of 5% BSA (wt/vol) in 0.01 M PBS, pH 7.2. The plates were washed, rotavirus antigen (crude culture supernatant from SA11 rotavirus-infected MA104 cells) was added, and the plates were incubated for 1 h at 37°C. Various dilutions of serum or fecal samples from mice were added to the ELISA plates and incubated for 1 h at 37°C. Affinity-purified antibody peroxidase-labeled goat anti-mouse IgA-IgG-IgM (H+L) (KPL, Gaithersburg, MD) was added at 1 μg/ml and incubated for 1 h at 37°C. The plates were washed five times. TMB (3,3′,5,5′-tetramethylbenzidine; KPL, Gaithersburg, MD) peroxidase substrate was added and incubated for 10 min and the reaction stopped with 1 N HCl. The A492 of samples was measured with a plate reader spectrophotometer (Labsytems, Inc., Franklin, MA). The absorbance values of preimmune sera were used as reference blanks. Dilution curves were drawn for each serum sample and for endpoint titers, represented as the reciprocal values of the last dilution that gave an optical density of ≥0.1, expressed as the means ± standard errors (SE) of results from all animals subjected to the same vaccine regimen.
Serum rotavirus-specific IgG subclass responses were measured with the same experimental procedure, but using peroxidase-conjugated rabbit anti-mouse IgG1 and IgG2a. Fecal or serum IgA and IgG antibodies to a murine strain (EDIM EW) rotavirus were determined by use of IgA- or IgG-specific peroxidase-labeled antiglobulin in an indirect ELISA (27). Ten percent (wt/vol) stool suspensions were made in 0.01 M phosphate-buffered saline, pH 7.1, with 1% BSA and 0.05% Tween 20 added.
Three weeks following the last immunization, mice were challenged (by oral gavage) with 100 50% infective dose (ID50) equivalents per mouse of a murine strain (EDIM EW) of rotavirus. Virus shedding in feces was determined with a commercial ELISA (Rotaclone, Meridian Bioscience, Inc., Cincinnati, OH) for detecting rotavirus antigen.
Antibody titers and standard deviations were calculated with the Microsoft Excel and Sigmaplot programs. Analysis of variance (ANOVA) was applied for comparison of the differences among experimental groups.
The expression of VP6 in B. subtilis was confirmed by Western blot analysis. Extracts of vegetative cells of strain BB2666, which expresses bovine VP6, were subjected to SDS-PAGE. The 45-kDa protein was detected with polyclonal anti-SA11 (simian rotavirus) antibody, and its size is that expected for rotavirus VP6 (Fig. (Fig.2).2). A protein of the same size was also detected in the crude culture supernatant from SA11 rotavirus-infected MA104 cells used as a positive control (Fig. (Fig.2).2). These results indicate that the 45-kDa protein expressed in BB2666 vegetative cells is rotavirus VP6. Rotavirus VP6 was also detected in spore preparations and was mostly found associated with the spores themselves.
Serum antibody titers generated by immunization with the spore vaccines are shown in Fig. Fig.3A.3A. For these studies, animals were immunized intranasally with either BB2666 (expressing bovine VP6) or BB2667 (expressing murine VP6) spores given with or without CT as an adjuvant. The best responses were obtained for mice that received CT adjuvant coinoculated with the spores. To determine protection, mice were challenged with the EDIM EW strain of murine rotavirus, as described above. The spore vaccine expressing murine rotavirus VP6 administered with CT as an adjuvant gave almost complete protection, as indicated by a large reduction in viral shedding in feces (Fig. (Fig.3B).3B). The adjuvanted spore vaccine expressing bovine rotavirus VP6 also gave substantial protection, although the shedding in the spore vaccine expressing bovine rotavirus VP6 was significantly elevated compared to that in the spore vaccine expressing murine rotavirus VP6.
We also tested another adjuvant shown to be effective in rotavirus studies, mLT. As shown in Fig. Fig.4A,4A, high-titer serum antibodies were induced by the spore vaccines when coinoculated with mLT. The results shown in Fig. Fig.4B4B demonstrate that the vaccines were highly effective, almost completely blocking rotavirus infectivity.
The serum antibodies generated by immunization with the adjuvanted vegetative cell vaccine in BALB/c mice are shown in Fig. Fig.5A.5A. All of the mice showed antibody responses, but the levels of these responses were generally lower than those obtained with the spore vaccines, despite the fact that the amount of VP6 antigen per dose was about 5 times larger for the vegetative cell preparations than for the spores (Fig. (Fig.3A3A and and4A4A and data not shown). To determine protection, mice were challenged with the EDIM EW strain of murine rotavirus. None of the mice showed any significant reduction in fecal shedding of virus compared with controls (Fig. (Fig.5B).5B). Coadministration of control spore preparations (i.e., spores of a strain that does not express VP6) with the vegetative cells expressing bovine or murine VP6 was also tried. Serum antibody responses were again obtained, but protection was not (data not shown).
Because the antibody titers obtained with the vegetative cell vaccines were lower than those obtained with the vaccines given as spores, the vegetative cell vaccines were tested with another strain of mice, DBA/2. The vaccines had been lyophilized and reconstituted before use to ensure a consistent source of inoculum. Most B. subtilis vegetative cells were ruptured, and few viable cells were detected (<103 cells) after lyophilization and reconstitution. The serum antibodies generated by immunization with the vegetative cell vaccine in DBA/2 mice are shown in Fig. Fig.6A.6A. The DBA/2 mice showed higher levels of antibody response than those obtained previously for BALB/c mice (Fig. (Fig.5A),5A), with antibody titers similar to those obtained with spore vaccines (Fig. (Fig.3A)3A) that proved to be protective. However, after challenge with EDIM EW rotavirus, none of the mice showed any reduction in fecal shedding of virus compared with controls (Fig. (Fig.6B6B).
T helper cell responses (Th1 or Th2) were estimated by the ratios of IgG1 and IgG2 antibodies in sera from mice immunized with the bovine VP6 B. subtilis vaccines. The results are shown in Table Table1.1. Except for the spore vaccine given without adjuvant, which had a bias toward a Th2-type response, the ratios indicated a bias toward a Th1-type response. Serum IgA antibodies were also analyzed, but the absorbance values were as low as those obtained with the preimmune sera used as reference blanks, indicating that the serum IgA level was minimal.
Fecal rotavirus-specific IgG and IgA antibodies in BALB/c mice were examined before challenge. IgG was clearly detected in feces of mice that had been immunized with the spore vaccine but was barely above the background level in samples immunized with the vegetative cell vaccine. As seen in Fig. Fig.7,7, IgA was detected in the feces of mice that had been immunized with the spores, indicating that an intestinal mucosal antibody response had been induced. No rotavirus-specific IgA was detected in stools of mice immunized with the vegetative cell vaccine. No IgG or IgA antibodies against rotavirus VP6 were detected in the feces of control mice.
The results presented demonstrate that B. subtilis-based rotavirus VP6 vaccines are effective in generating protective immunity against rotavirus challenge in the adult mouse. Protective immunity was obtained with the B. subtilis spore-based vaccines when CT or mLT was added but not when spores were given in the absence of adjuvant. This is consistent with previous studies indicating that an adjuvant is necessary to obtain high levels of protection with rotavirus VP6 antigen-based vaccines (7, 27). In our studies, mLT was more effective as an adjuvant than CT. Although mLT is attenuated and nontoxic compared with native LT or with CT, recent studies have shown that enzymatically inactive nontoxic mutant LT adjuvants may cause transient side effects (Bell's palsy) when given intranasally (32), and therefore, other adjuvants are likely to be needed when vaccines such as the ones we describe here are applied clinically. Although there was a significant difference in rotavirus shedding between bovine VP6 and murine VP6 adjuvanted with CT, the shedding being elevated in mice immunized with bovine VP6 (Fig. (Fig.3B),3B), vaccines derived from bovine rotavirus genes were as protective in mice as those derived from murine rotavirus genes when mLT was the adjuvant used (Fig. (Fig.4B).4B). These results demonstrate that the VP6-based vaccines were not species specific. Serum antibody responses were obtained with vegetative B. subtilis cells expressing VP6, but no protection was obtained with or without mLT or CT. This was true both in BALB/c mice and in DBA/2 mice. DBA/2 mice were tested because these mice tend to give Th1 responses, whereas BALB/c mice tend to give Th2 responses (23), which could affect the response to challenge virus. The DBA/2 mice generated higher serum antibody titers than those obtained for BALB/c mice. Protection with VP6-based vaccines has also been obtained for rabbits (8) and gnotobiotic lambs (56). An exception is gnotobiotic pigs, which have not shown protection after vaccination with VP6-based vaccines given alone (62). The gnotobiotic pig may respond differently to certain rotavirus vaccines. For example, inactivated-whole-rotavirus vaccines given parenterally generated neutralizing antibody responses in gnotobiotic pigs but did not give protection (63), whereas protection was obtained after parenteral injection of inactivated rotavirus in rabbits (10) or gnotobiotic lambs (56).
Protective neutralizing antibodies are directed against the two outer layer proteins, VP4 and VP7, proteins that are involved in viral attachment and entry into cells. Because antibodies to rotavirus VP6 do not have traditional in vitro neutralizing activities, the mechanism of protection by VP6-expressing spores in this study and other VP6-based vaccines remains to be determined. One possibility that has been proposed is that VP6-based vaccines work by intracellular interference with rotavirus assembly in enterocytes during transcytosis of VP6-specific IgA from the lamina propria to the lumen (11). In support of this theory, protection in mice immunized with VP2/6 virus-like particles proved to depend on the J chain, consistent with a transcytosis mechanism (52). In another study, however, protection was generated by chimeric VP6 proteins in J-chain knockout mice (37). It has been suggested that a possible reason for these different results is that the structure of the VP6 epitope recognized by the anti-VP6 IgA is probably crucial for intracellular neutralization to occur (11). In addition, intranasal immunization with VP6 antigen induces only a minority of rotavirus-specific B cells with an intestinal homing profile (44). Other studies indicate that CD4+ T cells are the only lymphocytes needed for protection after intranasal delivery of VP6 (38). It is apparent that various types of VP6-based vaccines can protect against an intestinal rotavirus challenge, whatever the mechanism, even when the vaccine is given intranasally. The findings that protection against challenge was achieved only in animals in which fecal IgA was detected lend support to the theory that intracellular interference with rotavirus assembly during transcytosis may explain the immunity, but more-extensive studies will need to be done before we can reach a definitive conclusion.
It is not clear why the spore vaccines induced protective immunity whereas vegetative cell vaccines expressing the same VP6 antigen did not, whether in BALB/c or DBA mice (which generated higher serum antibody titers than those obtained for BALB/c mice). Serum IgG levels were similar in the groups vaccinated with the spore forms and those vaccinated with the vegetative forms (serum IgA was also analyzed, but the levels were below the detection limit), indicating the same level of systemic immunization between spore and vegetative forms. However, fecal IgG and IgA were significantly increased in the group vaccinated with the spore form but not in that vaccinated with the vegetative form, implying a crucial role for the spore in generating protective immunity in this system. We believe that it is unlikely that an immunomodulating effect of the spores themselves can explain this phenomenon, since coadministration of control spores with vegetative cells of the vaccine strain did not induce protective immunity (data not shown). Thus, it appears that direct association of the VP6 antigen with the spore form or the manner in which the antigen is presented is essential for the observed protection. The recombinant B. subtilis strains used in this work express VP6 from a constitutive promoter active in vegetative cells. Spores of such strains may present antigens to the animal immune system by germination in the nasopharynx, by entrapment during spore coat assembly and consequent surface exposure, by association with the spore surface after mother cell lysis, or by degradation in macrophages of spores containing cytoplasmic antigen. There is little or no replication of either spores or vegetative cells. The vaccines are given intranasally, not orally, so except for a possibly small amount that may be inadvertently swallowed, there is no concern about possible germination in the gut. Further, in related experiments with TTFC-expressing B. subtilis strains, germination-deficient strains still induced immune responses equal to those induced by germination-proficient strains and there was no detectable replication in nares in a viability test (unpublished data).
In a study on immune responses of mice to TTFC, Barnes et al. (2) found that B. subtilis spores coadministered by intranasal immunization with TTFC augmented specific IgA to TTFC both in the local respiratory and in the distal vaginal mucosa and increased antigen-specific IgG antibody in draining lymph nodes and blood. Coadministration of B. subtilis spores intranasally with TTFC decreased the amount of IgG1, resulting in less of a Th2-type response (IgG1/IgG2a ratios of 4, compared to 8 without spores). IgG2a and IgG1 have been used as indicators of the induction of Th1 and Th2 responses, respectively; thus, the IgG1/IgG2a ratio can help to define the T-cell phenotype induced by vaccination (9). Th1 responses are generally associated with cell-mediated immunity, whereas Th2 responses are more associated with humoral immunity. The type of T helper cell response that we obtained, as determined by IgG1 and IgG2a ratios, indicated a bias toward a Th1-type response in either spore or vegetative cell vaccines coadministered with mLT as an adjuvant. Chimeric rotavirus VP6 vaccines given intranasally with mLT (7) showed some bias toward Th2-type responses (IgG1/IgG2a ratios of 3.4). Engineered B. subtilis spores expressing TTFC given orally also gave responses biased toward the Th2 type (IgG1/IgG2a ratios of approximately 4) (55). In studies with other bacteria, both the commensal Lactobacillus murinus and the pathogen Streptococcus pyogenes were found to stimulate a Th1 response when given intranasally (12). Th1 responses are thought to generate stronger cytotoxic T cell responses and higher levels of gamma interferon (IFN-γ) and interleukin-2 (IL-2), factors that may contribute to viral clearance, than Th2 responses. The clearance of rotavirus infection in mice can be mediated by immune CD8+ T lymphocytes, and this clearance can occur in the absence of virus-specific antibodies (14, 20, 21, 36). In our studies, no significant differences in the types of Th responses were seen with vegetative cells or spores.
The immune responses and protection that we obtained here are the first for a bacterial spore-based vaccine against a viral pathogen. The use of B. subtilis-based vaccines is a new approach to immunization against rotavirus which offers potential solutions to the difficulties that may be encountered, especially in some areas of the developing world, with current live, orally administered rotavirus vaccines and, as these vaccines are not based on infectious viruses, may be safer as well (29). Because the vaccines described here are not given orally and do not require viral replication, there should be no interference by maternal antibodies in breast milk, by naturally occurring rotavirus infections, or by other enteric virus infections. The use of B. subtilis spores presenting specific rotavirus proteins also offers an additional approach to investigation of immune responses to rotavirus infection and other factors involved in immunity. Moreover, our results justify the evaluation of the candidate vaccines and the strategy of mucosal (and oral) immunization in additional experimental models.
This study was supported by a grant (FNIH 1137) from the Grand Challenges in Global Health program of the Bill and Melinda Gates Foundation, and this grant was administered by the Foundation for the National Institutes of Health.
We thank the following for their helpful suggestions and guidance during the course of this work: Nirmal Ganguly, Jawaharlal Institute of Post Graduate Medical Education and Research; Dennis J. Kopecko, Center for Biologics Evaluation and Research, FDA; Richard M. Losick, Harvard University; and Alan Shaw, VaxInnate Corp.
Published ahead of print on 1 September 2010.