ABSTRACT
Shiga toxin 1 (Stx1) of enterohemorrhagic Escherichia coli O157:H7 was cloned, and four mutant Stx1s were constructed by site-directed mutagenesis with PCR. The wild-type and mutant Stx1s with amino acid replacements at positions 167 and 170 of the A subunit were purified by one-step affinity chromatography with commercially available Globotriose Fractogel, and the mutant Stxs were used for the immunization of mice. The mutant toxins were nontoxic to Vero cells in vitro and to mice in vivo and induced the immunoglobulin G antibody against the wild-type Stx1, which neutralized the cytotoxicity of Stx1. The induced antibody titers depended on the mutation at position 170 of the A subunit. The mice immunized with the mutant Stx1s were protected against a challenge of approximately 100 times the 50% lethal dose of the wild-type Stx1, suggesting that the mutant toxins are good candidates for toxoid vaccines for infection by Stx1-producing E. coli.
Enterohemorrhagic Escherichia coli (EHEC), which produces Shiga toxin (Stx), is a causative agent of one of the serious foodborne infectious diseases occurring in developed countries, including the United States, Canada, European countries, and Japan (11, 25). The infection is usually due to contaminated meats or vegetables but is occasionally transmitted by water or even by person-to-person contact (8). Most incidents of the disease are sporadic, but some massive outbreaks have been reported, such as the outbreak in a primary school in Sakai, Japan (28) and an outbreak in the United States caused by transmission by hamburgers (24). The disease is associated with diarrhea, hemorrhagic colitis, and, at a certain frequency, hemolytic-uremic syndrome (HUS). HUS is the most serious complication of the infection, with a high mortality rate among children and elderly people (13). The main virulence factors of EHEC, which cause systemic infections such as HUS, are considered to be two kinds of toxins, Shiga toxin 1 (Stx1) and Stx2 (1).
To control outbreaks caused by EHEC and to reduce mortality due to HUS, a safe and effective vaccine is required. Many efforts have been made by several research groups to develop a live vaccine (2, 20), a cell component vaccine (6), a polysaccharide-conjugated vaccine (15, 16), and a B subunit or toxoid vaccine (10, 19, 21) of Stx. A glutaraldehyde-inactivated Stx was shown to have good protective efficacy in rabbits (3, 18). Mutant Stx1 and Stx2 toxins constructed by site-directed mutagenesis in the active center of the A subunit were reported by two groups (7, 26). They were antigenic in rabbits (5, 26) and induced an antibody against the wild-type toxins. The mutant Stx2s were applied to a porcine vaccination, and its ability to prevent edema disease was demonstrated (4, 7). On the basis of these reports, mutant Stxs with amino acid replacements in the active center of A subunit were proved to be good candidates for the vaccine. However, a limited number of amino acid replacements or protection experiments using mutant toxins were examined in the studies so far reported, and further study is required to develop a vaccine suitable for practical use for humans.
To examine other possible mutations for the present study, we constructed four different mutant Stx1s (including a mutation reported by another group [26]), purified them by quick one-step affinity chromatography, and compared their antigenicities and protective abilities with the lethal toxicity of Stx1 in mice.
MATERIALS AND METHODS
Bacterial strains and culture conditions. E. coli O157:H7 strain 147, which produces only Stx1, was provided by K. Tamura (National Institute of Infectious Diseases, Tokyo, Japan). E. coli DH5α (Table 1) was used as a host strain for the wild-type and mutant Stx production. E. coli strain cultures were grown in antibiotic medium 3 (Difco Laboratories, Detroit, Mich.) at 37°C for 16 h with shaking.
Bacterial strains and plasmids
Construction of mutant Stx1.A DNA region including the stx1 gene in E. coli O157:H7 strain 147 was amplified by PCR with the primers 5′-CTACGCATGCTGTTAAGGTTGCAGCTCTC-3′ and 5′-CACTGTCGACGCCCTGACCACATCGTAG-3′, ligated with the cloning vector pUC118 at SphI and SalI sites, and introduced into E. coli DH5α by transformation. Site-directed mutagenesis was carried out with four primers harboring HindIII sites (underlined), corresponding to the active center of the Stx1 A subunit: 5′-GTAAAGCTTGAGCTGTCACAGT-3′ for replacement of E by Q at position 167 [A167(E→Q)], 5′-GTAAAGCTTTAGCTGTCACAGT-3′ for A167(E→K), 5′-CTGAAGCTTTAGGTTTTCGCA-3′ for A170(R→G), and 5′-CTGAAGCTTTACCTTTTCGCA-3′ for A170(R→L). Amplified DNA fragments were digested with HindIII, ligated, and further ligated to pUC118 to construct plasmids to produce four kinds of mutant Stx1 (Table 1).
Purification of wild-type and mutant Stx1s.Cultures of E. coli strains harboring plasmids for producing the wild-type or mutant Stx1 were grown in 200 ml of medium and harvested by centrifugation. The bacterial cells were suspended in 0.1% polymyxin B in phosphate-buffered saline (PBS) (pH 7.4) and incubated at 37°C for 60 min. The bacterial cells were removed by centrifugation and filtration with a 0.45-μm-pore-size membrane, and the crude toxin preparation thus obtained was applied to a small column (1 or 2 ml) of Globotriose Fractogel (IsoSep AB, Tullinge, Sweden). After nonabsorbed proteins were washed out with 15 ml of PBS, Stx was eluted with 15 ml of 4 M MgCl2 in PBS, dialyzed against PBS, and concentrated by ultrafiltration to a volume of approximately 500 μl. The purified Stx was divided into aliquots and stored at −80°C until use.
SDS-PAGE.Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed using 15% acrylamide gel and the buffer system of Laemmli (17). The gel was stained by Coomassie brilliant blue to visualize protein bands.
Cytotoxicity.Vero cell cultures grown in minimal essential medium (MEM) (Nissui Pharmaceutical Co., Tokyo, Japan) supplemented with 5% fetal bovine serum (FBS) (Gibco BRL, Grand Island, N.Y.), 1.05 g of NaHCO3/liter, 2.92 g of l-glutamic acid/liter, and 100 U of penicillin G (5% FBS-MEM)/ml were dispensed to each well of a 96-well culture plate (4 × 104 cells/well) and incubated at 37°C for 16 h with 5% CO2. After the culture medium was changed to the medium containing the wild-type or mutant Stx1s, the cell cultures were grown under the same conditions for 72 h. The cells were then washed twice with PBS and incubated with 5% FBS-MEM containing 10% Alamar blue (Biosource, Camarillo, Calif.) for 2 h, and the absorbance at 570 nm and 600 nm was measured to calculate the cell viability and 50% cytotoxic dose (CD50) (13). The triplicate assays were repeated three times, and mean values were used for CD50 calculation.
Mouse lethality.BALB/c mice were purchased from Charles River Japan, Inc. (Yokohama, Japan) and used after 1 week (at the age of 7 weeks). Mice were administered intraperitoneal doses of serial 10-fold dilutions (five mice for each dose) of purified wild-type or mutant Stx1s dissolved in 0.2 ml of PBS. Survival of the mice was observed for 4 days, and the 50% lethal dose (LD50) was calculated by the method of Kärber (12). All animal experiments were done in accordance with the Kitasato University Guidelines for Animal Care and Experimentation.
Immunization of mice.Mice (Japan Charles River; 7 weeks old) were subcutaneously administered 10 μg of mutant Stx1 emulsified with an equal amount of Freund's complete adjuvant (Sigma Chemical Co., St. Louis, Mo.) and booster doses with the same dose of mutant Stx1 emulsified with an equal volume of Freund's incomplete adjuvant (Sigma Chemical Co.) by the same route on days 21 and 35. Sera were collected 2 weeks after the last immunization. In another experiment, the immunized mice were used for the Stx1 challenge experiments.
Challenge with Stx1.At 2 weeks after the last immunization, the immunized mice were challenged intraperitoneally with the wild-type Stx1 at serial 10-fold dilutions and observed for 4 days, and LD50 values were calculated as described above.
ELISA.Enzyme-linked immunosorbent assays (ELISA) were used for the detection of immunoglobulin G (IgG) antibodies to Stx1. The purified wild-type Stx1 was dispensed to a 96-well plate (0.2 μg/well) and incubated at 4°C for 16 h for adsorption. Using peroxidase-conjugated goat anti-mouse IgG (Sigma Chemical Co.) as the secondary antibody, ELISA was performed as described in a previous paper (14). Triplicate assays were repeated three times, and standard deviations were calculated.
Western blotting.The ability of IgG antibodies induced by mutant Stx1s to recognize the wild-type Stx1 was examined by Western blotting assays. The purified wild-type Stx1 (5 μg) was applied to SDS-PAGE and electrotransferred to a nitrocellulose membrane filter after electrophoresis. Protein bands were detected by the sera prepared from mice immunized with mutant Stx1s and alkaline phosphatase-conjugated rabbit anti-mouse IgG (Jackson Immune Research Laboratories, West Grove, Pa.) as described elsewhere (27). The sera and the secondary antibody were diluted 200 and 1,000 times before use, respectively. A nonimmunized mouse serum was used as control.
Neutralization experiments.The wild-type Stx1 in 5% FBS-MEM was mixed with twofold serially diluted mouse serum and incubated at 37°C for 30 min. The final concentration of Stx1 was adjusted to 2 times the CD50. The mixture of Stx and the sera was dispensed to Vero cells (4 × 104 cells/well) in a 96-well plate, and the plate was incubated at 37°C for 72 h. The cell viability was measured by an Alamar blue assay as described above.
RESULTS
Yield and homogeneity of purified toxins.The purification took about 30 h after harvest of bacterial cells, and the yield of Stx1 was approximately 1 μg from 1 ml of culture broth. The purified wild-type and mutant Stx1s were analyzed by SDS-PAGE to check the homogeneity. As shown in Fig. 1, the wild-type Stx1 and the mutant Stx1s (STX111, STX112, STX 113, and STX114) were separated to two distinct protein bands with apparent molecular weights of 32,000 and 7,700, which correspond to the A and B subunits of Stx1, respectively. Compared with the A subunit of the wild-type Stx1, the A subunit bands of mutant Stx1s were thinner, suggesting that the association of A and B subunits is weaker in mutant toxins than in the wild-type toxin and that parts of the A subunits are lost during purification.
SDS-PAGE of the purified wild-type and mutant Stxs. The wild-type and mutant toxins (10 μg) were applied to a 15% polyacrylamide gel. MW, molecular weight standards; lane 1, wild-type Stx1; lane 2, STX111; lane 3, STX112; lane 4, STX113; lane 5, STX114.
Toxicity of the mutant Stx1s.Vero-cell cytotoxicity of the mutant Stx1s was assayed by comparison with that of the wild-type Stx1. Neither decrease of viability nor change of cell morphology was observed for the cells mixed with 10 μg of the mutant Stx1s/ml, which showed that the CD50 values for these mutant toxins were more than 10 μg/ml. On the other hand, the CD50 of the wild-type Stx1 was determined to be 17 pg/ml.
The results of the mouse lethality assay are shown in Table 2. All mice administered 100 ng to 10 μg of the mutant Stx1s (STX111, STX112, STX113, and STX114) survived, indicating that the LD50 values for these mutant toxins were more than 10 μg/mouse. In contrast, all mice administered 1 μg of the wild-type Stx1 died, and the LD50 for the wild-type Stx1 was calculated as 200 ng/mouse (10 μg/kg of body weight). This value was higher than that for the wild-type Stx2 (approximately 1 ng/mouse) cloned, purified, and assayed by the same method used for Stx1.
Lethality of wild-type and mutant Stx1s
Induction of anti-Stx1 antibody.An anti-Stx1 IgG antibody was induced in the sera by immunization with the mutant Stx1s. After the first booster injection on day 21, the ELISA titers were found to rise and were further enhanced by the second booster. As shown in Fig. 2, the antibody was detected on day 49 in the sera of all mice immunized with one of the four kinds of mutant Stx1s, although the titers were different among the mutant toxins. The higher titers were detected in the sera of the mice immunized with STX112 and STX114, indicating that these mutant toxins maintained the antigenicity of the intact Stx1 more rigorously than the other two mutant toxins, STX111 and STX113.
Anti-Stx1 antibody titers induced by immunization with the mutant Stx1s. Groups of mice (n = 5) were immunized subcutaneously with 10 μg of the mutant Stx1s as described in Materials and Methods. Levels of Stx1 IgG antibodies were determined by ELISA. The serum dilution at which the difference of absorbance between the sample and the control became less than 0.2 was defined as the antibody titer. The data are expressed as means ± standard deviation.
Western-blotting experiments revealed that the induced polyclonal antibody responded mainly to the A subunit of Stx1 (Fig. 3, lanes 2 to 5). However, faint but definite staining for B subunit bands was also observed, suggesting that the induced anti-Stx1 antibodies also responded to a monomeric form of the B subunit.
Response to the subunits of Stx1 by the sera from the mutant Stx1-immunized mice. The wild-type Stx1 (5 μg) was subjected to SDS-PAGE and stained by Coomassie brilliant blue (lane 1). The wild-type Stx1 (5 μg) was subjected to a Western blotting assay. The sera from the mice immunized with STX111 (lane 2), STX112 (lane 3), STX113 (lane 4), or STX114 (lane 5) or with nonimmunized serum (lane 6) were used as primary antibodies, and alkaline phosphatase-conjugated rabbit anti-mouse IgG was used as a secondary antibody.
Neutralization of the cytotoxicity of Stx1.Vero cell cytotoxicity of the wild-type Stx1 was neutralized by incubation with the anti-Stx1 antiserum. At a 32-fold dilution, the sera from the mice immunized with STX112 and STX114 neutralized more than 80% of the cytotoxicity (Fig. 4). The neutralization ability of the sera from the mice immunized with STX111 and STX113 was lower than that of the other two sera, which correlated well with the antibody titers in these sera.
Neutralization of the Stx1 cytotoxicity by the mouse antisera. The wild-type Stx1 serum and the sera from the mice immunized with the mutant Stx1s were incubated at 37°C for 30 min before the cytotoxicity assay. ⧫, STX111; ▪, STX112; Δ, STX113; ○, STX114; ▿, nonimmunized.
Protection against Stx1 challenge.Mice immunized with STX112 or STX114 were challenged by the wild-type Stx1. As shown in Table 3, all immunized mice survived after a challenge of Stx1 as high as 10 μg. These mice showed neither paralysis nor any sign of toxicity due to Stx1, and they survived for 6 months after the challenge. In contrast, all control mice injected with PBS alone died after the challenge of 1 μg of Stx1 (Table 3). Two out of five control mice survived at a dose of 100 ng, but they showed slight paralysis of the hind legs. These results clearly demonstrated that the mice were protected against the challenge of the wild-type Stx1 by vaccination with two of the mutant toxins, STX112 and STX114.
Protection of immunized mice against Stx1 challenge
DISCUSSION
To assay the effect of mutation in a protein toxin molecule or evaluate the efficacy of vaccination by a toxoid, preparation of a highly purified toxin is required. For the present study, several methods of affinity chromatography were examined using Gb3 (23) or P1 antigenic glycoprotein (22). Among them, the method using commercially available Globotriose Fractogel was the easiest to perform. By using this method, wild-type and mutant Stx1 and also Stx2 can be purified in a short time without any elaborate skills or equipment and with good yield.
Using the purified toxin preparations, it was confirmed that mutations at positions 167 (E→Q) and 170 (R→L) in the A subunit fully inactivated the Vero cell cytotoxicity and mouse lethality, as reported previously by other groups (26). Moreover, three additional mutant Stx1s were constructed which harbored different combinations of amino acid replacements at the same positions. All four mutant toxins were nontoxic and were characterized as candidates for toxoid vaccines.
The mutant Stx1s were all immunogenic and induced the IgG antibody against the wild-type Stx1, although the ELISA titers differed among the mutant toxins. The titers of the antibody induced by the immunization with STX111 and STX113 were lower than those of the antibody induced by the immunization with STX112 and STX114 (Fig. 2). At the moment we do not know the precise biochemical reasons for these phenomena, but it can be assumed that the replacement of R by G at position 170 of the A subunit changes the conformation of the Stx1 molecule and reduces the antigenicity of the mutant toxins.
The induced polyclonal antibody strongly recognized the A subunit of Stx1 separated by SDS-PAGE, but the signal for the B subunit was very weak. The IgG induced by the immunization with the mutant toxin might recognize only the conformational structure of B subunits but not the monomer. Using a chemiluminescence method, therefore, we reexamined the reactivity of the antibody and confirmed the finding (presented in Fig. 3) that the antibody also recognized the monomer of the B subunit.
The neutralization activity of the antibody induced by STX111 was very low (Fig. 4), but the antibody showed the binding ability to the B subunit of the wild-type Stx1 to be equivalent with those induced by STX112 and STX114 (Fig. 3). These results seem to indicate that the neutralization of the cytotoxicity is mainly due to the antibody directed to the A subunit, although we have to check the antibody binding ability to the B subunit pentamer to confirm this idea.
Since the induced antibody neutralized the cytotoxicity of Stx1 in vitro, an in vivo protective effect against the wild-type Stx1 was expected. In challenge experiments, the protection of mice by vaccination with two of the mutant toxins, STX112 and STX114, was clearly demonstrated. The vaccinated mice remained healthy for 6 months, even after injection of the wild-type Stx1 at a dose 100 times the LD50. The antibody titers induced by the immunization with STX112 and STX114 did not decrease quickly, and 6 months after the immunization the titers were still about 1/10 of the highest values. Although we did not perform challenge experiments using these aged mice, the level of the antibody suggests that the protective effect remained for several months.
As demonstrated in this study as well as other studies (5, 7, 26), the mutant Stx with the mutation in the A subunit is a good candidate for a vaccine, because it retains the antigenicity of both A and B subunits with highly reduced toxicity. The preparation of the mutant Stx is not complicated once the mutation is introduced, which is an advantage compared with a glutaraldehyde-inactivated Stx (3). In the immunization in this study, Freund's adjuvant was used. In the preliminary experiments, however, the antibody was induced by immunization with these mutant toxins without any adjuvant. These results suggest the use of these toxoids as a vaccine against human infection by Stx-producing E. coli. Since several conjugated vaccines composed of protein toxins and O polysaccharides of E. coli lipopolysaccharide have been reported (9, 15, 16), the mutant Stx1s constructed in this study can also be used as a vaccine in combination with the O polysaccharides of Stx-producing E. coli or Shigella dysenteriae type 1.
ACKNOWLEDGMENTS
We thank C. Sugawara, J. Yamamoto, and K. Kudo (The Kitasato Institute) for excellent technical assistance and T. Sasaki and K. Komiyama (The Kitasato Institute) for useful discussion. K. Tamura and H. Watanabe (NIID, Japan) are also acknowledged for providing E. coli strains.
This study was supported in part by a grant from the 21st Century COE Program, Ministry of Education, Culture, Sports, Science and Technology (MEXT).
FOOTNOTES
- Received 13 November 2002.
- Returned for modification 17 January 2003.
- Accepted 24 February 2003.
- Copyright © 2003 American Society for Microbiology