Large-scale preparation of Shiga toxin 2 in Escherichia coli for toxoid vaccine antigen production



Hideyuki Arimitsu, Department of Microbiology, Fujita Health University School of Medicine, 1-98 Dengakugakubo, Toyoake, Aichi 470-1192, Japan. Tel: +81 562 93 2433; fax: +81 562 93 4003; email:


Enterohemorrhagic Escherichia coli (EHEC) causes hemorrhagic colitis, and in more severe cases, a serious clinical complication called hemolytic uremic syndrome (HUS). Shiga toxin (Stx)is one of the factors that cause HUS. Serotypes of Stx produced by EHEC include Stx1 and Stx2. Although some genetically mutated toxoids of Stx have been developed, large-scale preparation of Stx that is practical for vaccine development has not been reported. Therefore, overexpression methods for Stx2 and mutant Stx2 (mStx2) in E. coli were developed. The expression plasmid pBSK-Stx2(His) was constructed by inserting the full-length Stx2 gene, in which a six-histidine tag gene was fused at the end of the B subunit into the lacZα fragment gene of the pBluescript II SK(+) vector. An E. coli MV1184 strain transformed with pBSK-Stx2(His) overexpressed histidine-tagged Stx2 (Stx2-His) in cells cultured in CAYE broth in the presence of lincomycin. Stx2-His was purified using TALON metal affinity resin followed by hydroxyapatite chromatography. From 1 L of culture, 68.8 mg of Stx2-His and 61.1 mg of mStx2-His, which was generated by site-directed mutagenesis, were obtained. Stx2-His had a cytotoxic effect on HeLa cells and was lethal to mice. However, the toxicity of mStx2-His was approximately 1000-fold lower than that of Stx2-His. Mice immunized with mStx2-His produced specific antibodies that neutralized the toxicity of Stx2 in HeLa cells. Moreover, these mice survived challenge with high doses of Stx2-His. Therefore, the lincomycin-inducible overexpression method is suitable for large-scale preparation of Stx2 vaccine antigens.

List of Abbreviations:

Coomassie brilliant blue-R250


50% cytotoxic dose


cholera holotoxin


cholera toxin B subunit


enterohemorrhagic Escherichia coli




horseradish peroxidase


median lethal dose


heat-labile enterotoxin B subunit


minimum lethal dose


mutant Stx2


polyvinylidene difluoride




T-PBS containing 5% (w/v) skim milk


Shiga toxin


histidine-tagged Stx2


PBS containing 0.05% (v/v) Tween 20

Enterohemorrhagic Escherichia coli strains cause hemorrhagic colitis and a serious clinical complication called hemolytic uremic syndrome (HUS) that is characterized by hemolytic anemia, thrombocytopenia, and acute renal failure [1, 2]. Major causative factors of EHEC include two types of Stx, Stx-1 and Stx-2 (also referred to as Vero toxin-1 and Vero toxin-2, respectively), both of which consist of one A subunit (Stx1A and Stx2A) and five B subunits (Stx1B and Stx2B). At the amino acid sequence level, Stx1 is almost identical to Stx produced by Shigella dysenteriae 1, whereas Stx2 shares only 55% and 61% amino acid sequence identity with Stx1 in the A and B subunits, respectively. The B subunits bind to Gb3 on the eukaryotic cell membrane [3-5], whereas the A subunit functions as an RNA N-glycosidase that cleaves off a single adenine in the 28S rRNA component of the 60S ribosomal subunit, leading to cell death by inhibition of protein synthesis [6, 7]. Stx2 toxicity is reportedly greater than that of Stx1, because in mice the LD50 of Stx2 is lower than that of Stx1 [8], and in humans Stx2-producing strains generate more severe symptoms than do other strains [9-11].

The main strategy for the prevention of HUS is neutralization of toxins using specific antibodies. Two types of genetically mutated toxoids have been evaluated. One is the B subunit [12-14] and the other is attenuated holotoxin, which contains one or two mutations in the active center of the A subunit. The advantage of the B subunit vaccine is its safety, which is attributable to a total lack of the A subunit. On the other hand, genetically mutated holotoxoids are beneficial because they safely induce anti-A subunit antibody production. The enzymatic activity of the A subunit is reportedly reduced by mutations at position 167 (glutamic acid to glutamine), 170 (arginine to leucine), or both [15-18]. Additionally, a number of reports have shown that genetically attenuated holotoxins, such as mutant Stx1 [19, 20], mStx2 [20], mutant hybrid proteins [21], and mStx2e [22-24], are good candidates for vaccine antigens for prevention of Shiga toxemia. However, because the purification yields described in some reports are far too small for the practical use of these toxoids, overexpression and purification methods need to be developed for these antigen proteins.

We previously reported an overexpression method for production of recombinant CTB in E. coli [25]. In the expression plasmid, the entire CTB gene was inserted into the lacZα gene of a pBluescript II SK(+) vector with a Shine-Dalgarno sequence derived from the LTB of enterotoxigenic E. coli. Protein expression was induced only by cultivating the K12 derivative E. coli strain MV1184 transformed with the expression plasmid in CAYE broth containing lincomycin, which was originally identified as an antibiotic that prevents protein synthesis in gram-positive bacteria through inhibition of peptidyltransferase activity on the 50S ribosomal subunit [26]. Because this expression method has also been successfully applied to overexpression of CT [25], we reasoned that it would be applicable to overexpression of Stx, especially wild-type and mStx2.

In this paper, we present a lincomycin-inducible overexpression method for production of Stx2 and its mutant proteins. These proteins were expressed as histidine-tag fusion proteins at the C-terminal ends of the B subunits (Stx2-His and mStx2-His, respectively). We demonstrate the safety and antigenicity of mStx2-His as a vaccine antigen to protect mice from Shiga toxemia.


Construction of Stx expression plasmids

The expression plasmid for Stx2-His was prepared according to a previously published procedure for CT preparation [25]. The complete nucleotide sequence of the gene encoding Stx2 was PCR amplified using the genomic DNA of E. coli O157:H7 (which was an outbreak strain in Okayama, Japan in 1996) as template DNA and a set of two primers, LTB(SD)Stx2(EcoRI)-f and Stx2B(6 x His)HindIII-r. The forward primer included the SD sequence derived from LTB upstream from the start codon of the Stx2 gene and the reverse primer was a fusion of the end of the B subunit gene and six-histidine (6 x His)-coding sequences. The amplified product was cloned into the pCR2.1 vector (Invitrogen, Life Technologies, Carlsbad, CA, USA) and transformed into the E. coli strain TOP10F′. After confirming the sequence, the cloned DNA was extracted from the plasmid using restriction enzymes (EcoRI and HindIII) and then subcloned into the pBluescript II SK(+) vector (Stratagene, La Jolla, CA, USA) digested with the same enzymes. The expression plasmid for Stx2-His was named pBSK-Stx2(His). The expression plasmid of the attenuated mStx2-His was generated from pBSK-Stx2(His) by changing the glutamic acid at position 167 and the arginine at position 170 of the A subunit into glutamine and leucine, respectively, by site-directed mutagenesis using a QuikChange II Site-directed Mutagenesis Kit (Stratagene) and two primer sets: Stx2A(E167Q)-f and Stx2A(E167Q)-r; and Stx2A(E167Q + R170L)-f and Stx2A(E167Q + R170L)-r. All primer sequences used in this study are listed in Table 1 and the plasmid map for pBSK-Stx2(His) is shown in Figure 1.

Table 1. Sequences of primers used in this study
Primer nameSequence (5′–3′)
  1. The underlined sequence of each primer indicates the restriction enzyme recognition site.
  2. aUsed for preparation of pBSK-Stx2(His)
  3. bUsed for preparation of pBSK-mStx2(His).
Figure 1.

Schematic of the expression plasmid of histidine-tagged Shiga toxin 2 (Stx2-His). The full Stx2 gene was amplified with the SD sequence derived from LTB (indicated as GGA) upstream from the A subunit gene. Six-histidine coding sequences were fused to the 3′-end of the B subunit gene. The amplified product was treated with EcoRI and HindIII and then ligated into the lacZα fragment gene of pBluescript II SK(+), which was pretreated with EcoRI and HindIII, in a different reading frame with lacZα gene.

Protein expression and purification

The pBSK-Stx2(His) plasmid was transformed into E. coli strain MV1184 (ara, Δ(lac-proAB), rpsL, thi (φ80lacZΔM15), Δ(srl-recA)306::Tn10 (tetr)/F′[traD36, proAB+, lacIq, lacZΔM15]). Each transformant was cultured in Luria–Bertani broth containing 50 μg/mL (final concentration) ampicillin overnight at 37°C. Next, 3 mL of culture was inoculated into 1 L of CAYE broth (2% casamino acids, 0.6% yeast extract, 0.25% NaCl, 0.871% K2HPO4 and 0.25% glucose) containing a 0.1% (v/v) trace salt solution (5% MgSO4, 0.5% MnCl2 and 0.5% FeCl3), 50 μg/mL of ampicillin, and 90 μg/mL of lincomycin (Pfizer, New York, NY, USA) and cultured for 48 hr at 30°C. The cells were collected by centrifugation (7600 g, 20 min) and sonicated in PBS (pH 7.4). After centrifugation (15,000 g, 90 min), the supernatant was applied to a 2 mL column of TALON metal affinity resin (Clontech, Mountain View, CA, USA) equilibrated with PBS, and then bound Stx2-His (or mStx2-His) was eluted by PBS containing 0.15 M imidazole. To remove the contaminated products of crude Stx2-His preparation, hydroxyapatite (Bio-Rad, Hercules, CA, USA) chromatography was conducted. Prior to chromatography, each crude preparation was dialyzed against 10 mM sodium phosphate buffer (pH 7.0) containing 1 M NaCl to avoid aggregation and then applied onto a hydroxyapatite column equilibrated with the same buffer. After collecting the unabsorbed fractions, the bound proteins were eluted with 0.4 M sodium phosphate buffer (pH 7.0). Unabsorbed Stx2-His was concentrated by applying it onto fresh TALON affinity resin and the final products were dialyzed in PBS. Throughout the purification process, insoluble proteins which were yielded during the dialyzing steps and storage period at −30°C were removed by centrifugation (15,000 g, 30 min). Protein concentrations were determined with DC protein assay reagent (Bio-Rad) using BSA as a standard.

Toxicity assays

The toxicity of each Stx2-His and EHEC-derived Stx2 (Nacalai Tesque, Kyoto, Japan) were evaluated in vitro and in vivo. For in vitro assays, a cytotoxicity assay was performed using HeLa229 cells according to the procedure published by Neri et al. [27]. For the toxin-neutralization assay, 20 pg/mL of EHEC-derived Stx2 was preincubated with an equal volume of 100-fold diluted sera from mice immunized with mStx2-His or PBS for 1 hr at 37°C. For the in vivo assays, each Stx2-His was serially diluted with PBS and 0.5 mL of each dilution injected intraperitoneally into at least five female ICR mice (6 weeks of age, Japan SLC, Hamamatsu, Japan). The animals were observed for 1 week and their deaths were recorded. The MLD was calculated from the dilution that killed all animals.

Vaccination schedule of mice

Ten micrograms of mStx2-His containing 0.05% (w/v) of aluminum hydroxide (which has been clinically used as an adjuvant) in 0.2 mL of PBS was injected s.c. twice at a 2-week interval into 25 female ICR mice (6 weeks of age). For a control group, PBS containing 0.05% (w/v) of aluminum hydroxide was injected into five mice instead of mStx2-His. Two weeks after the secondary immunization, the animals were tail bled to determine the specific serum antibody titer by ELISA. The mice immunized with mStx2-His were then divided into three groups that were intraperitoneally challenged with a 10-, 100-, or 1000-fold lethal doses of Stx2-His and their survivability was monitored for 1 week. All animal experiments were conducted according to the Guidelines for the Management of Laboratory Animals at Fujita Health University.

Enzyme-linked immunosorbent assay

Flat-bottom, 96-well plates were coated with 1 μg/100 μL of Stx2-His overnight at 4°C. After washing the plates three times with T-PBS, each well was blocked using 200 μL of S-PBS for 1.5 hr at 37°C. After washing the plates three times, 100 μL of immunized or untreated (normal) mice sera serially diluted with S-PBS was added to the plates and incubated for 1 hr at 37°C. The plates were washed three times and incubated with 100 μL of HRP-conjugated anti-mouse IgG goat Immunoglobulin (Jackson ImmunoResearch, West Grove, PA, USA) for 1 hr at 37°C. After washing the plates, the wells were reacted with 100 μL of citrate buffer (pH 5.0) containing 0.04% (w/v) o-phenylenediamine and 0.02% (v/v) hydrogen peroxide for 30 min at 37°C. The reaction was stopped by the addition of 100 μL of 1 M H2SO4 and the absorbance measured at 492 nm using a microplate reader (Tecan, Mannedorf, Switzerland). The absorbance value for each sample was compared with that of normal serum at the same dilution, and the antibody titer was determined as a reciprocal of the highest dilution with the lowest positive difference of the 1.5 × absorbance value of normal serum subtracted from the 1 × absorbance value of each sample.

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis and western blot analysis

Cell lysates from transformants were prepared using previously described methods [25]. The sample proteins were resolved on a 15% polyacrylamide gel. The gel was stained with CBB-R250 or electroblotted onto a PVDF membrane using the iBlot gel transfer system (Invitrogen). The membrane was incubated with blocking buffer (S-PBS) for 1 hr and reacted for 1 hr with mouse antiserum immunized with mStx2-His or rabbit antiserum immunized with Stx2 toxoid prepared from the culture supernatant of EHEC, each of which was diluted with T-PBS containing 5% (w/v) BSA. The membrane was incubated for 1 hr with HRP-conjugated anti-mouse IgG goat Immunoglobulin (Jackson ImmunoResearch) or anti-rabbit immunoglobulin porcine immunoglobulin (Dako, Copenhagen, Denmark), each of which was diluted with blocking buffer. Specific bands were detected with the Immobilon Western Chemiluminescent HRP Substrate (Millipore, Billerica, MA, USA) using an LAS4000 image analyzer (Fujifilm, Tokyo, Japan). All reactions were carried out at room temperature and the membranes were washed three times with T-PBS for 5 min before each reaction.

The N-terminal amino acid sequence of each subunit on the PVDF membrane stained with CBB-R250 was determined with a pulsed-liquid phase protein sequencer (model Procise 491HT; Applied Biosystems, Life Technologies, Carlsbad, CA, USA).

Statistical analysis

The antibody titers in the mStx2-His and adjuvant groups were statistically compared by Student's t-test.


Expression and purification of Stx2-His

To effectively purify large amounts of wild-type and mStx2, we constructed Stx2-expression plasmids in which we fused a six-histidine-coding gene to the 3′ end of the B subunit gene. We confirmed expression of Stx2-His, which has common antigenicities with EHEC-derived Stx2, in the MV1184 strain cultivated in CAYE broth in the presence of lincomycin by western blot analysis using anti-Stx2 rabbit serum (Fig. 2a), although the molecular mass of the histidine-tagged B subunit (lane 3) estimated according to electric mobility was somewhat higher than that of the EHEC-derived Stx2B subunit (lane 1).

Figure 2.

Expression of histidine-tagged Shiga toxin 2 (Stx2-His) in lysates from the E. coli MV1184 strain transformed with pBSK-Stx2(His) and SDS-PAGE profile of purified Stx2-His. (a) The transformants were cultured in CAYE broth in the presence (lane 3) or absence (lane 2) of lincomycin for 48 hr at 30°C. Each lane was loaded with total protein lysate corresponding to 1.5 × 108 cells, and Stx2-His was detected by anti-Stx2 rabbit serum. Five-hundred nanograms of EHEC-derived Stx2 (purchased from Nacalai Tesque) was loaded in lane 1 as a positive control. Four microliters of molecular weight size marker for western blot (MagicMark, Invitrogen) was loaded in the left lane. (b) Each 5 μg of Stx2-His and mStx2-His proteins was resolved on a 15% polyacrylamide gel and stained with Coomassie brilliant blue R-250. A molecular weight size marker was loaded on the left side of the gel. The N-terminal amino acid sequence of each band is shown in parentheses.

Although we purified Stx2-His proteins from the extract of MV1184 transformed with pBSK-Stx2(His) using TALON affinity resin, we also confirmed multiple contaminants by SDS–PAGE (data not shown). Therefore, we tried using hydroxyapatite chromatography to eliminate contaminants. However, most of the proteins aggregated during dialysis in 10 mM sodium phosphate buffer without NaCl, which is generally used as the initial binding buffer for hydroxyapatite (data not shown). For this reason, we dialyzed the proteins that were eluted from the TALON resin against the same buffer containing 1 M NaCl and then applied them to a hydroxyapatite column. We collected Stx2-His proteins in the unabsorbed fractions.

As shown in Figure 2b, purified Stx2-His and mStx2-His showed 35 kDa (A subunit; Stx2A) and 11.6 kDa (B subunit; Stx2B-His) bands. The N-terminal amino acid sequence of each subunit was identical to that of the EHEC-derived Stx2, which was reported by Jackson et al. [28]. The means of the final yield of Stx2-His and mStx2-His from 1 L of culture in CAYE broth were 68.8 and 61.1 mg, respectively.

Biological activities of Stx2-His and mStx2-His

To confirm that the recombinant Stx2-His proteins have toxic activities, we used in vitro and in vivo assays. We calculated the CD50 of Stx2-His protein to HeLa229 cells at 1.8 pg/mL, which is similar to that of EHEC-derived Stx2 (2.5 pg/mL), whereas the CD50 of mStx2-His was considerably higher (585 ng/mL). On the other hand, the intraperitoneal MLD of Stx2-His in adult mice (6 weeks of age) was 100 ng, whereas that of mStx2-His was considerably higher (100 μg), indicating that the activities of these mutant toxins are close to non-hazardous when administered at vaccination dosages.

Vaccine effect of mStx2-His in mice

To confirm the effect of mStx2-His as a vaccine antigen, we immunized ICR mice s.c. with 10 μg of mStx2-His containing aluminum hydroxide as a practical adjuvant for vaccine. No mice died of or were weakened by the immunization. As shown in Figure 3a, the IgG antibody titers in mice that were immunized twice with mStx2-His were significantly higher (mean ± SEM 2,206,250 ± 335,643, range 156,250–3,906,250) than those of mice immunized with adjuvant alone (titers of all five were < 10). The neutralizing activities of these antibodies were confirmed by an in vitro neutralization assay using 10 pg/mL of EHEC-derived Stx2 (corresponding to a 20.9% survival concentration in HeLa229 cells). No sera derived from immunized mice with PBS neutralized the toxicities (mean ± SEM of survival rate 25.7 ± 0.4%), whereas the sera derived from immunized mice with mStx2-His neutralized the toxicities (mean ± SEM survival rate 70.3 ± 7.0%).

Figure 3.

Specific antibody (IgG) responses in the sera of mice that had received two s.c. immunizations with mStx2-His. (a) Antibody titers of the sera were determined by ELISA against Stx2-His. Data are represented as mean ± SEM values. *P < 0.0001. (b) Survival rate of mice immunized with mStx2-His against lethal doses of Stx2-His challenge. Each mouse immunized with mStx2-His (open symbols) or PBS (closed circles, n = 5) was intraperitoneally inoculated with 10- (both circles, n = 8), 100- (open triangle, n = 8), or 1000-fold (open squares, n = 9) MLD Stx2-His on day 0.

To investigate the degree of protection conferred by antibodies that were induced in mice by immunization with mStx2-His, we divided the mice into three groups and challenged them with different lethal doses of wild-type Stx2. In this study, we used Stx2-His to challenge mice with high lethal doses of purified toxin on the assumption that a large amount of toxin protein was needed. As shown in Figure 3b, all the mice immunized with mStx2-His survived a challenge of Stx2-His at 10- and 100-fold MLD (1 and 10 μg/mouse, respectively) for at least 1 week with no symptoms, whereas only three of nine mice survived a challenge of 1000-fold MLD (100 μg/mouse). All of the mice immunized with adjuvant alone succumbed to a challenge with 10-fold MLD within 3 days.


It is crucial to consider the following three points if toxoids are to have clinical utility. First, the toxoid itself must not be hazardous to humans and animals. Second, the toxoid should induce sufficient antibody production to neutralize an excess amount of wild-type toxin. Third, it should be possible to prepare large amounts of the toxoid antigen easily and cheaply. Taken together, these factors necessitate use of the overexpression method for preparation of antigenic proteins.

In the process of constructing the CTB expression plasmid in our previous study [25], we confirmed that the SD sequence derived from LTB worked well for expression of the Vibrio cholerae derived CTB gene in E. coli without any obvious toxicities. Therefore, we expected that the SD sequence of LTB would be perfectly applicable to expression of Stx2, which has a gene derived from E. coli itself. Furthermore, the conventional purification method for Stx2 is very cumbersome, because to obtain 440 μg of Stx2 from 12 L of culture supernatant, several purification steps (ammonium sulfate precipitation, DEAE-cellulose column chromatography, repeated chromatofocusing column chromatographs, repeated high performance liquid chromatographs) are needed, leading to significant protein loss [29]. Therefore, we constructed expression plasmids for Stx2 as histidine-tagged proteins to aid in the purification process. Western blot analysis using the anti-Stx2 antibody confirmed that the transformants expressed Stx2-His in the presence of lincomycin. Furthermore, the presence of a band of A subunit, which was crudely purified by TALON affinity chromatography, in the SDS–PAGE analysis of Stx2-His confirmed that the A subunit formed holotoxin complex with histidine-tagged B subunits. We attempted to eliminate contaminants from the crude Stx2-His preparations by hydroxyapatite chromatography because this chromatography method is effectively in purifying recombinant CT from other contaminants, including free CTB complexes [25]. However, prior to performing chromatography, the dialysis process in 10 mM sodium phosphate buffer without NaCl, which we used as the initial binding buffer for hydroxyapatite chromatography, caused irreversible aggregation of Stx2-His, indicating that histidine-tagged Stx2 is denatured into an insoluble form under low-salt buffered conditions.

The molecular mass of Stx2B-His, estimated by SDS–PAGE, was somewhat higher than that deduced from the amino acid sequence (8.6 kDa including 6 x His), despite the fact that the N-terminal modification of each subunit corresponded to that observed in previous studies [28]. However, we confirmed that non-tagged Stx2, which was expressed in the transformant using an expression plasmid (pBSK-Stx2) prepared by site-directed mutagenesis of pBSK-Stx2(His), had the same electric mobility as EHEC-derived Stx2 (shown in Supporting Information), indicating that the observed increase in molecular mass of Stx2B-His might be attributable to a characteristic of histidine-tag fusion proteins that causes delayed electric mobility. Purified Stx2-His showed cytotoxic activity against HeLa229 cells and was lethal to mice, whereas the mutant toxin displayed decreased toxicity, as described in previous reports [15-17, 20, 21], even in the presence of 6 x His.

To investigate whether mStx2-His is available as a vaccine antigen, we immunized mice s.c. with aluminum hydroxide. The mice immunized with mStx2-His produced serum toxin-neutralizing antibodies and survived a challenge with 10- and 100-fold MLD Stx2-His, whereas more than half the mice died when challenged with 1000-fold Stx2-His. The range of specific antibody titers among mice immunized with mStx2-His was similar (156,250–3,906,250), suggesting that small differences in antibody production by mStx2-His-immunized mice that were challenged with 1000-fold MLD Stx2-His could ultimately determine whether they lived or died. Although mStx2-His vaccination did not confer sufficient protection to mice to withstand challenge with 1000-fold MLD Stx2-His, vaccination did completely protect mice from challenge with 100-fold MLD, leading us to conclude that there was sufficient evidence for mStx2-His as a vaccine antigen. In this study, we could not use EHEC-derived Stx2 to challenge the mice because this would have required a large amount of toxin. Although we confirmed the in vitro neutralization effect of anti-mStx2-His sera against EHEC-derived Stx2, we have yet to confirm the in vivo neutralization effect of the antisera against a large amount of EHEC-derived Stx2.

In summary, we succeeded in overexpressing wild-type and mStx2-His to be employed as a vaccine antigen to protect mice from Shiga toxemia. The method described in this study is cost effective and suitable for large-scale preparation of toxoid vaccine.


This work was supported, in part, by grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan and Health and Labour Sciences Research Grants for Research on global health issues from the Ministry of Health, Labor and Welfare, Japan.


The authors declare no conflicts of interest or financial support.