A recombinant carboxy-terminal domain of alpha-toxin protects mice against Clostridium perfringens



Masahiro Nagahama, Department of Microbiology, Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Yamashiro-cho, Tokushima 770-8514, Japan.

Tel: +81 088 622 9611; fax: +81 088 655 3051; email: nagahama@ph.bunri-u.ac.jp


Clostridium perfringens alpha-toxin (CP, 370 residues) is one of the main agents involved in the development of gas gangrene. In this study, the immunogenicity and protective efficacy of the C-terminal domain (CP251-370) of the toxin and phospholipase C (PLC; CB, 372 residues) of Clostridum bifermentans isolated from cases of clostridium necrosis were examined. The recombinant proteins were expressed as glutathione S-transferase (GST) fusion proteins. Antibodies that cross-reacted with alpha-toxin were produced after immunization with recombinant proteins including GST-CP251-370, GST-CP281-370, GST-CP311-370, CB1-372 and GST-CB251-372. Anti-GST-CP251-370, anti-GST-CP281-370 and anti-GST-CP311-370 sera neutralized both the PLC and hemolytic activities of alpha-toxin, whereas anti-CB1-372 and anti-GST-CB251-372 weakly neutralized these activities. Immunization with GST-CP251-370 and GST-CP281-370 provided protection against the lethal effects of the toxin and C. perfringens type A NCTC8237. Partial protection from the toxin and C. perfringens was elicited by immunization with GST-CP311-370 and CB1-372. GST-CP251-370 and GST-CP281-370 are promising candidates for vaccines for clostridial-induced gas gangrene.

List of Abbreviations

C. bifermentans phospholipase C

C. bifermentans

Clostridum bifermentans


colony forming units


Clostridium perfringens alpha-toxin

C. perfringens

Clostridium perfringens


glutathione S-transferase




optical density


PBS containing 0.05% (v/v) Tween 20


phospholipase C

Clostridium perfringens type A is the commonest organism isolated from patients with trauma-induced gas gangrene [1-3]. Once an infection has become established in traumatized tissues, acute invasion and destruction of adjacent healthy, living tissues can progress many cms per hour despite appropriate antibiotic therapy [4]. Shock and organ failure occur in 50% of infected patients and 40% of all infected patients die [4]. Gas gangrene occurs in elderly and diabetic patients and in patients who have undergone surgery of the gastrointestinal tract [5]. Unlike other soft tissue infections, gas gangrene also involves clostridial myonecrosis [4].

The pathogenesis of gas gangrene is largely attributable to production of potent exotoxins by the bacterium [2]. Of these, alpha-toxin, being a major contributor to the disease, has received particular attention [6-8]. Alpha-toxin causes hemolysis, platelet aggregation, contraction of blood vessels and the ileum and lethality [6, 7]. The toxin also exhibits PLC activity [7]. In the initial stages of infection, it may damage and reduce the blood supply to peripheral tissues, thus promoting the conditions required for spread of the infection [6]. Moreover, alpha-toxin contributes to rapid destruction of viable tissue by altering inflammatory and endothelial cell functions, stimulating microvascular occlusion and inhibiting tissue inflammatory responses [4]. Thus, alpha-toxin is a major virulence factor in the pathogenesis of gas gangrene.

The three-dimensional structure of alpha-toxin has two domains, the N-domain (1–250 residues) and the C-domain (251–370 residues) [9]. The N-domain contains the active site and is similar to PLC of Bacillus cereus [10]. The C-domain is similar to the C2 domain of intracellular eukaryotic proteins involved in vesicular transport and signal transduction [9, 11]. We have previously reported that alpha-toxin has two tightly bound zinc metals and an exchangeable divalent cation. Residues His-68, -126, and -136 bind an exchangeable divalent cation that is required for binding to membranes, His-148 and Glu-152 bind one zinc ion that is essential for the active site of the toxin and His-11 and Asp-130 tightly bind the other zinc ion, which is required for maintenance of the structure [12, 13]. We also reported that Tyr-57 and -65 play roles in penetration of the toxin into the bilayer of membranes and access of the catalytic site to sphingomyelin in membranes [14].

The N-domain of alpha-toxin retains PLC activity but shows a loss of hemolytic activity and the C-domain influences the enzymatic activity of the N-domain and recognizes membrane phospholipids [6, 15]. We also reported that the C-domain is important for binding to membranes and hemolytic activity of alpha-toxin [7, 16]. Researchers have previously shown that, in mice, immunization with the C-domain generates antibodies that protect against the toxin [17]. Therefore, the C-domain of the toxin could be the principal immunogen of an alpha-toxin vaccine. In the present study, to determine which components of the central region of the C-domain of alpha-toxin act as immunogens, we investigated the efficacy of active immunization with various fragments (CP251-370, CP281-370 and CP311-370). Although C. perfringens is the main etiological agent of clostridial myonecrosis, C. bifermentans has also been isolated from necrotic tissues [18]. C. bifermentans produces PLC (CB, 372 residues), with 50% identity with alpha-toxin at the amino acid level [15]. We determined whether CB and the C-domain (CB251-372) could protect against alpha-toxin and these organisms.


Alpha-toxin and toxoid

Alpha-toxin was purified from culture supernatant of Bacillus subtilis, which contains an expression plasmid that carries the C. perfringens NCTC8237 alpha-toxin gene as described previously [12]. Toxoid of alpha-toxin was prepared by detoxification with 0.4% formaldehyde at 37°C for 2 days.


BALB/C female mice were purchased from Japan SLC (Hamamatsu, Japan). The mice were kept in our specific pathogen-free animal facility. For all experiments, 6-week-old-animals were used and maintained pathogen free prior to injection. Experiments were performed in accordance with institutional guidelines.

Expression and purification of recombinant proteins

Recombinant CP251-370, CP281-370 and CP311-370 were expressed as proteins fused with GST in Escherichia coli BL21 as described previously [16]. The CB gene was amplified by PCR with the chromosomal DNA from C. bifermentans KZ1012 as a template. DNA fragments encoding the CB1-372 and CB251-372 genes were amplified by PCR using primers that added BglII and SalI sites at the 5′- and 3′-ends of the products, respectively. The amplified fragments were subcloned into the corresponding sites of pGEX-5X-1. Recombinant CB1-372 and CB251-372 were expressed as described previously [16]. The purified recombinant proteins were analyzed by SDS–PAGE (Fig. 1). The protein assay was performed with Bradford protein assay reagent (Bio-Rad Lab, Tokyo, Japan).

Figure 1.

(a) Schematic representation of candidate vaccines. (b) SDS–PAGE analysis of recombinant proteins. Purified preparations (5 μg of protein) were subjected to SDS–PAGE (12.5% polyacrylamide gel) followed by staining with 1% (w/v) Coomassie brilliant blue. Lanes: 1, GST-CP251-370; 2, GST-CP281-370; 3, GST-CP311-370; 4, CB1-372; 5, GST-CB281-372.

Toxicity of recombinant proteins and alpha-toxin toxoid

To evaluate the mouse toxicity of recombinant proteins and alpha-toxin toxoids, the proteins (10 μg) were injected i.p. into groups of six female BALB/c mice (6 weeks old). The recombinant proteins and alpha-toxin toxoids injected were non-toxic for 2 weeks after the inoculation.

Immunization of mice

Female BALB/c mice were used at 6 weeks of age. Groups of 10 mice were injected subcutaneously with 5 µg of the recombinant proteins and alpha-toxin toxoid in complete Freund's adjuvant (Sigma–Aldrich, Tokyo, Japan). Booster injections were given 2 weeks later using the same amount of antigen but in incomplete Freund's adjuvant (Sigma–Aldrich).

Measurement of antibody titers

For sampling, serum from an immunized mouse was taken before the first injection (preimmune serum) and at 2 weeks after the second injection. Antibody titers were measured by ELISA as described previously [19]. Briefly, microtitration plates (Maxisorp; Thermo Scientific, Yokohama, Japan) were coated with alpha-toxin (5 µg/mL) in 0.05 M carbonate buffer (pH 9.6) and incubated at 37°C for 2 hr. The plates were blocked with PBS containing 1% (w/v) BSA and washed three times with PBST. The test sera, serially diluted (twofold) in PBS containing 0.1% BSA starting from 1:100, were added to triplicate wells (100 µL/well) for 1 hr at 37°C. The wells were washed five times with PBST. Goat anti-mouse (IgG) antibodies conjugated to horseradish peroxidase, obtained from Sigma–Aldrich, were diluted as recommended by the manufacturers and added to the wells. After incubation for 1 hr at 37°C and subsequent washes, the plates were developed with ortho-phenylenediamine (0.4 mg/mL) and H2O2 (6%; 0.4 µL/mL). Absorbance was measured at 492 nm in an ELISA reader and values are presented as endpoint titers. The highest dilution of a test serum with an OD greater than the mean OD of the preimmune serum (1:100 dilution) in the same group was considered the endpoint titer.

Neutralization of biological activities of alpha-toxin in vitro

Neutralization of the phospholipase C (egg-yolk phospholipid hydrolyzing) and hemolytic activities of alpha-toxin by antisera was assessed. Negative control serum was obtained from a preimmune mouse. Pooled serum was serially diluted in microtiter plates and dilutions preadsorbed with 100 ng of alpha-toxin. Preadsorbed mixtures were coincubated with sheep erythrocytes or egg-yolk emulsion at 37°C for 1 hr. Hemolytic and PLC activities were determined as described previously (12, 13).


Clostridia perfringens NCTC8237 was used for challenging mice. To prepare an inoculum, the bacteria were grown in brain–heart infusion broth, and 4 hr log-phase cultures pelleted by centrifugation, washed twice in PBS, and diluted to 109  CFUs per mL, based on a standard plot of turbidity (measured at 650 nm) against the number of CFUs per mL. The bacterial concentration of the inoculum was verified by duplicate plating of serial dilutions on blood agar plates.

Toxin and organism challenges

For mice injected i.p., 50% lethal doses of the toxin and C. perfringens NCTC8237 were 100 ng and 107 cells, respectively. Groups of 10 immunized animals were administered i.p. injections of alpha-toxin (1 μg) or C. perfringens NCTC8237 (108 cells) in PBS. The mice were closely observed for 24 hr for development of symptoms and the time to death was recorded.


Antibody response to recombinant proteins

Recombinant proteins containing CP251-370, CP281-370, CP311-370, CB1-372 and CB251-372 were expressed in and purified from Escherichia coli. To generate antibodies to the recombinant proteins, mice were immunized with GST, GST-CP251-370, GST-CP281-370, GST-CP311-370, CB1-372 and GST-CB251-372 in Freund's complete adjuvant for the first immunization and Freund's incomplete adjuvant for the booster immunization. Serum titers were analyzed 2 weeks after the second immunization. The appearance of circulating antibodies directed against alpha-toxin was monitored by ELISA (Table 1). No antibodies to alpha-toxin were detected in the serum of mice immunized with GST and the serum before immunization. In mice immunized with GST-CP251-370 and GST-CP281-370, strong antibody responses against alpha-toxin were detected. The magnitude of the antibody responses elicited by GST-CP251-370 was equivalent to that induced by alpha-toxin toxoid. On the other hand, GST-CP311-370 and CB1-372 reduced the titer of specific antitoxin antibody. Mice mounted very low antibody responses to the toxin after immunization with GST-CB251-372. No significant antitoxic responses were detected in mice immunized with CP251-370 and CP281-370 under our experimental conditions (data not shown).

Table 1. Anti-alpha-toxin titers following parenteral immunization
  1. n.d., not detected.
Alpha-toxin toxoid1:64,000

In vitro neutralization

After samples had been taken from the immunized animals 2 weeks after the second immunization, the ability of sera from animals immunized with the recombinant proteins to neutralize, in vitro, the biological activities associated with alpha-toxin was analyzed (Table 2). Antiserum against alpha-toxin toxoid inhibited the PLC and hemolytic activities of alpha-toxin, up to a final dilution of 360. Antisera to GST-CP251-370 and GST-CP281-370 neutralized PLC activity of alpha-toxin, up to a final dilution of 360. Similarly, antisera to GST-CP251-370 and GST-CP281-370 inhibited the hemolytic activity of alpha-toxin, up to a final dilution of 360 and 240, respectively. Antisera against GST-CP311-370 and CB1-372 weakly inhibited the two activities of alpha-toxin, compared with antisera to GST-CP251-370 and GST-CP281-370. Serum raised against GST-CB251-372 very weakly neutralized these activities.

Table 2. In vitro neutralization of alpha-toxin activity
AntiserumNeutralization titer
Phospholipase C activityHemolytic activity
GST< 15< 15
Alpha-toxin toxoid360360

Protection against the toxin

To determine the level of protection afforded by immunization with recombinant proteins, immunized mice were challenged with alpha-toxin (Table 3). When injected with 1 μg of purified alpha-toxin, all the control mice died within 3.2 hr. Similarly, the GST-immunized controls died rapidly (mean time to death, 3.5 hr). All mice immunized with GST-CP251-370, GST-CP281-370 and alpha-toxin toxoid were protected against the toxin (100% survival) and showed no signs of intoxication. On the other hand, the animals immunized with GST-CP311-370 and CB1-372 exhibited 60% and 40% survival, respectively, showing partial protection against the toxin. Protection of GST-CB251-372-immunized mice against alpha-toxin was weak (10% survival).

Table 3. Protection of immunized mice against challenges with alpha-toxin
ImmunogenNo. of survivors/total no. of mice% SurvivalMean time to death (hr)
  1. N/A, not applicable.
Alpha-toxin toxoid10/10100N/A

Protection against organisms

To investigate the possibility that the recombinant proteins could also provide protection against experimental gas gangrene, groups of mice were exposed to live C. perfringens (1 × 108 organisms) (Table 4). Control mice and GST-immunized mice died within 8 hr of the challenge. The GST-CP251-370, GST-CP281-370 and alpha-toxin toxoid vaccines protected all the mice. The GST-CP311-370 and CB1-372 groups showed 50% and 40% protection. GST-CB251-372 provided little protection.

Table 4. Protection of immunized mice against challenges with C. perfringens
ImmunogenNo. of survivors/total no. of mice% SurvivalMean time to death (hr)
  1. N/A, not applicable.
Alpha-toxin toxoid10/10100N/A


The present study demonstrates that immunization of mice with GST-CP251-370 or GST-CP281-370 provides protection against the lethal effects of alpha-toxin or C. perfringens. On the other hand, GST-CP311-370-immunized mice were only partially protected. These results indicate that the C-terminal region (CP281-370) of alpha-toxin contains enough epitopes to develop an effective immunogen and vaccine against the toxin.

This vaccine offers several clear advantages over formol toxoid: it is much more easily prepared and, because it is free from formaldehyde, it is likely to be safer and less reactogenic. In this respect, we did not observe any adverse side effects during the immunization schedule used.

The C-domain of alpha-toxin is an eight-stranded antiparallel β-sandwich containing a calcium binding sites. Calcium ions are reportedly essential for the binding of alpha-toxin to membranes [7, 8]. We have reported that acrylodan-labeled C-domain variant (S365C) binds to membranes and exhibit a marked blue shift, indicating internalization of the C-domain into the hydrophobic environment in membranes [16]. GST-CP251-370 and GST-CP281-370, derived from the C-domain of alpha-toxin, are immunogenic, generating antibodies that react with holotoxin, suggesting that this part of the toxin is correctly folded. We first demonstrated that GST-CP281-370 in the C-domain was clearly the most effective candidate vaccine tested, closely followed by GST-CP251-370. Anti-GST-CP281-370 serum neutralized PLC and the hemolytic activity of alpha-toxin. Obviously, neutralization of toxin-induced hemolysis is important for protection against death caused by the toxin. On the other hand, GST-CP311-370 did not show an immune response to alpha-toxin, suggesting that the structural conformation of the C-domain may also be important for stimulating a protective immune response. These data indicate that CP281-370 contains the dominant protective epitope of alpha-toxin. The strong level of protection achieved in these experiments suggests that CP281-370 of alpha-toxin plays an important role in the molecular mechanism of pathogenicity involving alpha-toxin. The structural arrangement may make the epitopes in CP281-370 the most prominent for recognition by immune effector cells. From these findings, we concluded that GST-CP281-370 is a strong immunogen, an effective vaccine against alpha-toxin in mice, and required for full protection.

The C. bifermentans PLC (CB) shows only 50% amino acid identity with alpha-toxin. Its C-terminal domain shows 42% identity with that of alpha-toxin. In spite of this, immunization with CB1-372 provided protection against the lethal effects of alpha-toxin or C. perfringens in a mouse model. Antibodies raised against CB1-372 cross-reacted with alpha-toxin in vitro. Immunization with GST-CB251-372 was less effective in generating high-titer or neutralizing antibodies. These findings indicate that alpha-toxin and C. bifermentans PLC share antibody-neutralizing epitopes. We are the first demonstrate to that recombinant CB1-372 elicits partial protection against the toxin.

In conclusion, we have shown that a C-terminal fragment of alpha-toxin, CP251-370 or CP281-370, when fused to GST, could prove useful for protection against clostridial myonecrosis and diseases in which the C. perfringens alpha-toxin is a major virulence factor.


We thank M. Mukai for technical assistance. This work was supported by a grant-in-aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan, MEXT.SENRYAKU, 2012.


The authors have no financial conflicts of interest.