Cell Surface-Exposed Tetanus Toxin Fragment C Produced by Recombinant Bacillus anthracis Protects against Tetanus Toxin

Abstract

Bacillus anthracis, the causal agent of anthrax, synthesizes two surface layer (S-layer) proteins, EA1 and Sap, which account for 5 to 10% of total protein and are expressed in vivo. A recombinant B. anthracis strain was constructed by integrating into the chromosome a translational fusion harboring the DNA fragments encoding the cell wall-targeting domain of the S-layer protein EA1 and tetanus toxin fragment C (ToxC). This construct was expressed under the control of the promoter of the S-layer component gene. The hybrid protein was stably expressed on the cell surface of the bacterium. Mice were immunized with bacilli of the corresponding strain, and the hybrid protein elicited a humoral response to ToxC. This immune response was sufficient to protect mice against tetanus toxin challenge. Thus, the strategy developed in this study may make it possible to generate multivalent live veterinary vaccines, using the S-layer protein genes as a cell surface display system.

Bacillus anthracis, a gram-positive, spore-forming bacterium, is the causal agent of anthrax. The virulence of this extracellular pathogen depends on two exotoxins and a poly-γ-d-glutamic acid capsule, encoded by the plasmids pXO1 and pXO2, respectively (13, 26).

The toxinogenic Sterne strain is currently used as a live veterinary vaccine against anthrax (11). This attenuated strain, cured of pXO2, harbors pXO1 and synthesizes toxin components, including the protective antigen (PA; encoded by pag). Recent studies have shown that Sterne derivatives express genes encoding heterologous antigens in vivo under the control of the pag gene promoter. A recombinant Sterne strain, in which ibp, encoding the Ib component of iota toxin from Clostridium perfringens, is fused to pag, secretes Ib protein in vitro and induces a protective humoral response against C. perfringens and Clostridium spiroforme toxins (25). If hly, encoding the Listeria monocytogenes listeriolysin, is fused to pag, B. anthracis enters cells due to the production and secretion of listeriolysin. Such recombinants induce anti-Listeria CD8-mediated protective immunity (24). Thus, various types of immunity are induced by antigens secreted by B. anthracis. We investigated whether heterologous antigens, present at the bacterial surface, also mediate protection.

B. anthracis synthesizes two abundant surface proteins, EA1 and Sap, which form a surface layer (S-layer) (8). The genes encoding these proteins, eag and sap, respectively, have been cloned, sequenced, and found to be clustered on the chromosome (5, 18). We have shown that the cell wall-targeting domain of B. anthracis S-layer proteins consists of the three SLH (S-layer homology) motifs (14) located in the amino termini of these proteins (16). These SLH domains are sufficient to anchor the Bacillus subtilis levansucrase (Lvs), which is usually secreted, onto the cell wall. The exposed Lvs retains antigenicity and enzymatic activity (16). We used fragment C of tetanus toxin (ToxC) of Clostridium tetani as a heterologous antigen to extend this system for vaccination purposes. ToxC is the 50-kDa carboxy-terminal portion of the tetanus toxin responsible for binding to gangliosides (10, 12). It has been shown to protect against tetanus toxin in mice (9). We therefore investigated the production and anchoring of a chimeric SLH-ToxC protein and the protection that it mediated.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions.

Escherichia coli TG1 (23) was used as a host for derivatives of pUC19 (30) and pAT113 (28). E. coli JM83(pRK24) (27) was used for mating experiments. A derivative of a B. anthracis Sterne lethal factor mutant (RPL686) was used as a recipient strain (2). E. coli cells were grown in Luria broth (LB) or on LB-agar plates (19). B. anthracis cells were grown in brain heart infusion (BHI) medium (Difco Laboratories) or in SPY medium (5). Antibiotics were used at the following concentrations: for E. coli, ampicillin at 100 μg ml⁻¹, kanamycin at 40 μg ml⁻¹, and spectinomycin at 60 μg ml⁻¹; for B. anthracis, spectinomycin at 60 μg ml⁻¹.

DNA manipulations.

Plasmid extraction, endonuclease digestion, ligation, and agarose gel electrophoresis were carried out as described by Maniatis et al. (15) or Sambrook et al. (23). PCR amplification was carried out with Vent DNA polymerase (New England Biolabs), using a maximum of 25 amplification cycles with 200 ng of DNA as the template. PCR products were phosphorylated with T4 polynucleotide kinase according to the manufacturer’s recommendations.

Plasmid constructions.

The DNA fragment encoding ToxC was amplified from total DNA extracted from C. tetani CN655 (22) by PCR. A PstI site (underlined) was introduced at the 5′ end with oligonucleotides TOXC3 (5′- CCAATTCCATTTTCTCTGCAGAAAAATCTGGATTGTT-GGGTTGAT-3′) and TOXC2 (5′-CATGCCATGGTCATGAACATATCAATCTGTTTA-3′). The amplified fragment was inserted into the HincII site of pUC19. The resulting plasmid, pRTOX1, was cut with XbaI and SacI and ligated to the XbaI-SacI fragment of pSAL322 (18), giving rise to pRTOX2. pRTOX50 was obtained by inserting the partially digested 3.9-kb PstI-SacI fragment of pRTOX2 between the NsiI and SacI sites of pSLH5 (16). The insert of pRTOX50 was thus equivalent to that of p5SacB (16), in which the sacB gene was replaced by the DNA fragment encoding ToxC, upstream from a spectinomycin resistance cassette, and a DNA fragment corresponding to the 3′ end of eag (encoding EA1). This construct therefore contained a translational fusion between the SLH domain of EA1 and ToxC (Fig. 1).

FIG. 1.

Schematic representation of the fusion protein, SLH-ToxC. The portion of the protein corresponding to the cell wall-targeting domain (SLH domain) of EA1 is shaded; that corresponding to ToxC is hatched. The spacer between the two polypeptides consists of the four underlined amino acids (see Materials and Methods).

Construction of the recombinant SLH-ToxC strain.

The recombinant suicide plasmid pRTOX50 was transferred from E. coli to B. anthracis by heterogramic conjugation as described by Trieu-Cuot et al. (27). Allelic exchange was carried out as described by Pezard et al. (20, 21), selecting for spectinomycin resistance.

RPL-ToxC was obtained by allelic exchange between the recombinant DNA from plasmid pRTOX50 and the chromosomal eag gene, encoding EA1, from the Sterne lethal factor mutant derivative strain. RPL-ToxC is thus EA1⁻ SLH-ToxC⁺.

Protein analysis.

B. anthracis cells were grown to an optical density at 600 nm (OD₆₀₀) of ∼3 in BHI medium and washed in 20 mM Tris-HCl (pH 8.0). Bacterial pellets were sonicated, and proteins in culture supernatants were precipitated with 10% trichloroacetic acid. The equivalent of 100 μl of culture for each fraction was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in a 12% gel. Separated proteins were transferred to nitrocellulose sheets by using the Bio-Rad Trans-Blot system. Blots were probed with polyclonal anti-EA1 (1:50,000 dilution) or anti-ToxC (1:20,000 dilution) serum and developed by using the Amersham ECL Western blotting analysis system, with the secondary antibody diluted 1:20,000.

Immunofluorescence.

Bacteria grown on BHI agar plates were washed in 20 mM phosphate-buffered saline (PBS; pH 8.0), applied to a micro-coverglass slide, air dried, and fixed by incubation for 5 min with 80% acetone (−20°C). The bacteria were washed twice with PBS and incubated for 1 h with mouse polyclonal anti-EA1 (1:100 dilution) or rabbit polyclonal anti-ToxC (1:100 dilution) antibodies. Bacteria were washed twice with PBS-bovine serum albumin (BSA; 1%, wt/vol), and the primary antibodies were detected by incubating slides for 1 h with either rhodamine-conjugated goat anti-rabbit immunoglobulin G (IgG) or fluorescein isothiocyanate-conjugated rat anti-mouse IgG (1:50 dilution). The bacteria were washed four times with PBS-BSA and mounted in Fluoprep (Bio Mérieux SA, Marcy l’Etoile, France).

Immunization and challenge procedures.

Swiss mice (6 to 10 weeks old, female) were supplied by IFFA-CREDO (L’Arbesle, France). Animals (seven per group) were immunized on day 0 (single dose), days 0 and 32 (two doses), or days 0, 32, and 47 (three doses) by subcutaneous injection of 10⁸ (first injection) and 10⁷ bacilli (further injections) in 0.5 ml of saline (0.15 M NaCl) complemented with 0.02% saponin. The strains were grown in SPY medium at 37°C to an OD₆₀₀ of ∼0.7. In these conditions, the synthesis of the anthrax toxin components is not induced, but the SLH-ToxC hybrid protein is expressed. The monitoring of anti-anthrax toxin antibody titers thus reflects the development of B. anthracis in vivo.

Tetanus toxin was prepared as previously described (1), and its activity was determined by establishing its 100% lethal dose in Swiss mice. On day 32 (single immunization), 47 (two immunizations) or 62 (three immunizations), the animals were challenged with a lethal dose of tetanus toxin (three times the 100% lethal dose). Survival was monitored for 1 week to estimate protection. For each challenge, groups of five mice immunized with control strain RPL686 were injected with tetanus toxin to check that animals received a lethal dose. These mice always died within 3 days.

Serological studies.

Mice were bled from the retro-orbital plexus on day 32 (single immunization), on days 32 and 47 (two immunizations), or on days 32, 47, and 62 (three immunizations) to obtain a serum sample prior to further boosting or challenge. Antibody titers (total mouse Ig) were determined as previously described (21, 25), by an enzyme-linked immunosorbent assay (ELISA). Microplates were coated with 25 μg of anatoxin, 100 ng of Sap protein, or 100 ng of PA per well. The ELISA antibody titer was defined as the serum dilution at which the absorbance at 492 nm was 0.5. The mean antibody titers were geometric means. Student’s t test was used to compare geometric mean titers of antibody to a given antigen.

RESULTS AND DISCUSSION

Construction of B. anthracis recombinant strain RPL-ToxC synthesizing SLH-ToxC.

A DNA fragment encoding an SLH domain, consisting of the three SLH motifs of the EA1 S-layer protein, was fused to that encoding ToxC, which is normally secreted (Fig. 1; see also Materials and Methods).

The Sterne strain is commonly used as a veterinary vaccine, and Sterne derivatives have been shown to give similar levels of protection in animal models, without killing the experimental animals at high concentration (21). Thus, the chimeric gene was introduced, by allelic exchange, into the chromosome of a Sterne derivative mutant producing an inactive lethal factor. The recombinant gene was expressed under the control of the eag promoter. The hybrid protein SLH-ToxC contained a spacer of four extra amino acids between the two fused polypeptides: a proline and a glycine were introduced to promote the correct folding of the two domains. The insertion of these amino acids and the design of the C-terminal limit of the third SLH motif were based on the results obtained in in vitro experiments on SLH motifs and the synthesis and anchoring of a protein that is normally secreted, Lvs from B. subtilis, using SLH-Lvs fusions (16). A methionine and a glutamine resulted from the ligation of two restriction sites to give an in-frame fusion (Fig. 1).

Characterization of the RPL-ToxC strain.

The recombinant strain RPL-ToxC was grown in vitro. Samples from crude extracts and culture supernatants were subjected to SDS-PAGE and Western blotting, using sera against EA1 and ToxC (Fig. 2). Anti-EA1 antibodies gave a positive signal with extracts from the control strain (Fig. 2A, lanes 1 and 2). No 94-kDa protein reacted with the extracts from RPL-ToxC, indicating that the strain is EA1⁻ and that allelic exchange had therefore occurred at the predicted locus (Fig. 2A, lanes 3 and 4). However, a faint signal, corresponding to the size predicted for the hybrid protein SLH-ToxC (about 70 kDa), was detected in the crude fraction of the recombinant strain (Fig. 2A, lane 3). This was probably due to the presence of antibodies raised against the SLH domain of EA1 in the polyclonal serum. The anti-ToxC antibodies did not cross-react with any polypeptides in the control strain (Fig. 2B, lanes 1 and 2), indicating that the serum was specific in the experimental conditions used. A strong signal corresponding to the size expected for the hybrid protein was detected in the crude extract of the recombinant strain (Fig. 2B, lane 3). This protein was processed to give a polypeptide of about 50 kDa, the apparent molecular mass of ToxC. The cleavage product, which was present primarily in the culture supernatant (Fig. 2B, lane 4), was not recognized by anti-SLH antibodies (data not shown) and was therefore probably the ToxC moiety of the hybrid protein. Similar results were obtained with a hybrid SLH-Lvs protein (16). Thus the chimeric protein was processed to give mature Lvs.

FIG. 2.

Western blot analysis of B. anthracis synthesizing SLH-ToxC. Protein samples equivalent to 100 μl of culture at an OD₆₀₀ of ∼3 of the parental strain (lanes 1 and 2) and of RPL-ToxC (lanes 3 and 4) were subjected to SDS-PAGE and transferred to nitrocellulose. Lanes 1 and 3, crude extract; lanes 2 and 4, culture supernatant. The blots were probed with anti-EA1 polyclonal serum (A) and with anti-ToxC polyclonal serum (B).

Immunofluorescence experiments showed that the recombinant strain did not express EA1 (compare Fig. 3C and D) and that the fusion protein, SLH-ToxC, was located on the cell surface of the recombinant bacilli (compare Fig. 3A and B).

FIG. 3.

Detection of SLH-ToxC at the cell surface of B. anthracis by immunofluorescence staining of vegetative cells. RPL-ToxC bacilli (A and C) and the parental strain bacilli (B and D) were stained with either anti-EA1 antibodies conjugated with fluorescein isothiocyanate (C and D) or anti-ToxC antibodies conjugated with rhodamine (A and B).

The synthesis and efficient anchoring of SLH-ToxC at the cell surface of B. anthracis thus supported our previous results with the B. subtilis Lvs (16). In addition, our results suggest that such an expression system could be used with various antigens.

Protective immunity against tetanus toxin induced by B. anthracis RPL-ToxC.

We first investigated whether RPL-ToxC elicited a humoral response to ToxC after one, two, or three immunizations. Mice were immunized on day 0 (single dose), days 0 and 32 (two doses), or days 0, 32, and 47 (three doses) by subcutaneous injection, and antibody titers were determined by ELISA (Table 1). Antibody titers were very low following a single immunization but increased significantly after a booster (2,100 versus <100, P = 0.017) (experiment B). A second booster further significantly increased the antibody titers (4,200 versus 1,850, P = 0.0014) (experiment C). The very low antibody response against ToxC is also described for other systems (3, 4, 29). The relatively weak titers of antibody against ToxC compared with those against other B. anthracis antigens such as anthrax toxin components (20) can be tentatively explained by the low amount of ToxC produced by B. anthracis (less than 1 μg per 10⁹ bacteria). A humoral response was also elicited against Sap and PA, as previously described (17, 18, 21). The titers of antibody against Sap and PA also increased after a booster but were lower after a second booster, for reasons which remain unclear (Table 1). As the bacilli were grown in a medium that does not induce anthrax toxin component synthesis, the anti-PA antibody titers demonstrated that B. anthracis developed in vivo. The increases in the various antibody titers after boosting are consistent with the induction of a memory response.

TABLE 1.

Antibody responses before tetanus toxin challenge

Mouse	Geometric mean antibody titer (total mouse Ig)a
	Expt A, day 32b			Expt B				Expt C
	Anatoxin	Sap	PA	Day 32, anatoxin	Day 47			Day 32, anatoxin	Day 47, anatoxin	Day 62
					Anatoxin	Sap	PA			Anatoxin	Sap	PA
1	<100	150	12,800c	<100	2,400	37,000	200,000	<100	1,200	12,800	9,600	150,000
2	100	150	25,000	<100	4,800	200,000	400,000	150	2,400	3,200	25,000	100,000
3	<100	1,200	25,000	0	4,800	37,000	40,000	<100	1,600	9,600	25,000	100,000
4	<100	800	6,400c	150	1,600	75,000	800,000	<100	600	1,200	4,800	75,000
5	0	150	4,800c	<100	200	200,000	400,000c	<100	800	6,400	25,000	100,000
6	<100	150	6,400c	150	6,400	25,000	600,000	<100	3,200	1,600	9,600	50,000
7	100	150	6,400	0	1,600	800	400,000	<100	9,600	4,800	50,000	100,000
Mean titer	<100	250	10,000	<100	2,100	36,000	300,000	<100	1,850	4,200	16,500	90,000

^aAntianatoxin antibody content of sera from mice immunized once (experiment A), twice (experiment B), or three times (experiment C) with bacilli containing RPL-ToxC was determined by ELISA. Anti-Sap and anti-PA antibody contents of the sera were also determined for each mouse prior to challenge. In experiments A, B, and C, 43, 85, and 100%, respectively, of mice were protected. 

^bDay of bleeding. 

^cMouse died following the challenge. 

There is substantiated evidence that ToxC is immunogenic and induces protection against challenge with the toxin, if injected alone (6, 7) or delivered via attenuated bacteria (3, 4, 29). Therefore, we challenged immunized mice with tetanus toxin. The survival rates of mice vaccinated with RPL-ToxC correlated with the humoral response to the anatoxin (Table 1). The recombinant strain, RPL-ToxC, gave protective immunity after a single injection (43%) despite weak antibody titers against ToxC. The level of protection increased to 100% after two boosters; 100% protection was also achieved after three injections in the absence of saponin (data not shown). A humoral response was obtained simultaneously against the secreted protein PA and the surface-exposed antigen SLH-ToxC, giving protection against tetanus toxin. Our expression system seems to be valid for veterinary vaccinal purposes since it allows high protection (85%) after only two immunizations and maximal protection after three immunizations, which is comparable to results obtained with other live recombinant bacteria expressing ToxC, such as Vibrio cholerae or Lactococcus lactis (4, 29). In addition, it allows stable expression of heterologous antigens since the recombinant DNA fragment encoding a translational fusion is integrated into the chromosome. Therefore, it is tempting to assume that the sap gene, encoding the second S-layer protein of B. anthracis, could be used simultaneously to drive the expression of other heterologous antigens. Thus, multivalent B. anthracis live vaccines, presenting surface-exposed antigens engineered according to the strategy described above and secreted antigens, through the gene encoding toxin components, could be used.