Determining Antibody-Binding Site of Streptococcal Pyrogenic Exotoxin B to Protect Mice from Group A Streptococcus Infection

Streptococcal pyrogenic exotoxin B (SPE B), a cysteine protease, is an important virulence factor in group A streptococcal (GAS) infection. SPE B binds and cleaves antibody isotypes and further impairs the immune system by inhibiting complement activation. In this study, we examined the antibody-binding site of SPE B and used it to block SPE B actions during GAS infection. We constructed different segments of the spe B gene and induced them to express different recombinant fragments of SPE B. Using an enzyme-linked immunosorbent assay (ELISA), we found that residues 345–398 of the C-terminal domain of SPE B (rSPE B345–398), but not the N-terminal domain, was the major binding site for antibody isotypes. Using a competitive ELISA, we also found that rSPE B345–398 bound to the Fc portion of IgG. The in vitro functional assays indicate that rSPE B345–398 not only interfered with cleavage of antibody isotypes but also interfered with SPE B-induced inhibition of complement activation. Immunization of BALB/c mice using rSPE B345–398 was able to induce production of a high titer of anti-rSPE B345–398 antibodies and efficiently protected mice from GAS-induced death. These findings suggest that SPE B uses its C-terminal domain to bind the Fc portion of IgG and that immunization of mice with this binding domain (rSPE B345–398) could protect mice from GAS infection.


Introduction
Streptococcus pyogenes (group A streptococcus; GAS) is an important human pathogen that causes a variety of diseases, including pharyngitis, cellulitis, impetigo, scarlet fever, necrotizing fasciitis, puerperal sepsis, and streptococcal toxic shock syndrome (STSS) [1,2,3]. Despite intensive care with antimicrobial therapy, the mortality rate has remained high, as has the incidence postinfection sequelae, such as acute rheumatic fever [4]. Several virulence factors have been reported that contribute to evasion of host immunity by GAS. These factors consist of the cell surface M protein, M-like protein, the hyaluronic acid capsule, the streptococcal inhibitor of complement, and C5a peptidase [5,6,7,8,9], as well as secreted exotoxins and enzymes such as streptococcal pyrogenic exotoxin B (SPE B), IdeS (IgG-degrading enzyme of S. pyogenes), endo-b-N-acetylglucosaminidase (EndoS), and DNases [10,11,12,13,14]. SPE B, a cysteine protease secreted by most strains of GAS, has been reported to play a role in the pathogenesis of various diseases [10,15,16,17]. SPE B is expressed as a 40 kDa precursor zymogen (zSPE B) and is subsequently cleaved to a 28 kDa mature form protease (mSPE B) [10]. mSPE B directly interferes with host immunity by degrading complement molecules [18,19] and antibodies [10,20,21] and also digests other host proteins, such as fibronectin, vitronectin, kininogen, metalloprotease, prointerleukin-1b, and the urokinase receptor [22,23,24,25,26]. Another report has suggested that the covS mutation switches the M1T1 strain GAS phenotype from speB high /speA 2/ Sda1 low to the highly virulent speB 2/ speA + /Sda1 high phenotype [27]. Several reports still indicate that a speB mutant strain decreases resistance to neutrophil phagocytosis, dissemination to organs, and mortality in a mouse model [16,21,28]. Our previous study also indicates that SPE B and streptolysin (SLS) have a synergistic effect on GASmediated macrophage death and the resistance of GAS to immune cell-mediated killing and that SPE B plays a more important role than SLS in increasing the severity of GAS-induced skin lesions [29]. Clinical investigation indicates that high levels of SPE B protease activity are significantly associated with signs of STSS and with mortality. Patients with lower antibody levels against SPE B are more likely to succumb to invasive GAS disease [30]. Taken together, these reports indicate that SPE B is a critical virulence factor in GAS infection. SPE B has been known to digest free immunoglobulins, including IgG, IgA, IgM, IgE, and IgD [12], as well as antigenbound IgG [20,21]; hence, antibody-mediated neutralization and complement activation in GAS infection are impaired by SPE B. However, the exact antibody-binding site of SPE B has yet to be clearly defined. In this study, we demonstrated that SPE B uses its C-terminal domain, specifically amino-acid residues 345-398, to bind the Fc portion of IgG. Using a recombinant rSPE B 345-398 protein to block the binding between SPE B and antibody isotypes inhibited cleavage of antibodies by SPE B and SPE B-mediated inhibition of complement activation. Recombinant rSPE B 345-398 could potentially serve as a vaccine to protect mice from GASinduced death.

Purification of Human Immunoglobulins
Normal human sera were donated by healthy volunteers. We obtained written informed consent from each person and approved by the ethics committee of E-Da Hospital. Protein Lagarose (Thermo) and protein A-agarose (Thermo) were used to purify human serum immunoglobulins. Ten milliliters of binding buffer containing 0.1 M phosphate and 0.15 M sodium chloride Figure 1. Binding of serum immunoglobulins and C192S. Human sera were passed through an affinity column immobilized with C192S, and then the binding ligands of C192S were eluted and the fraction 13 (F13) was collected. The contents in diluted human serum, different amount of purified human immunoglobulins, and F13 (1 mg) were verified using 12% SDS-PAGE (A) and blotted with goat anti-human IgG (B), IgM (C), or IgA (D). (E) 2 mM of purified C192S or BSA was coated in a 96-well ELISA plate, and binding of IgG, IgM, or IgA with C192S or BSA was detected by ELISA, as described in Materials and Methods. **P,0.01 compared with values determined for the BSA group. doi:10.1371/journal.pone.0055028.g001 (pH 7.2) was added to a protein L-agarose-packaged column. Normal human sera diluted 2-fold with binding buffer were passed through the protein L column. IgG, IgM, IgA, IgE, and IgD bound to protein L-agarose because of the ability of protein L to bind the k chain of immunoglobulins. After washing with binding buffer to remove unbound materials, 6 to 10 ml of the elution buffer containing 0.1 M glycine (pH 2.5) was added to elute the five immunoglobulin isotypes. The immunoglobulin mixture was then dialyzed using vivaspin 20 (GE Healthcare) with the binding buffer for protein A-agarose that contained 20% phosphatebuffered saline (PBS). The immunoglobulin mixture was passed through the protein A column to purify IgG, which was purified prior to use. Unbound immunoglobulin isotypes were also collected, and the contents were further examined by western blotting using anti-isotype antibodies. In this unbound mixture, only IgM and IgA were detected. The concentrations of IgE and IgD were below the level of detection for the assay. We prepared the IgM-IgA mixture for further use.

Cloning and Expression of SPE B Truncations
The recombinant SPE B and the C192S mutant lacking protease activity were prepared as described previously [19,31]. Briefly, the genomic DNA of GAS was extracted and the structural gene of SPE B was amplified using polymerase chain reaction (PCR) with the sense primer 59-GGATCCGGATCCCATCAT-CATCATCATCATGATCAAAACTTTGCTCGTAACGAA-39 and antisense primer 59-GGATCCGGATCCCTAAGGTTT-GATGCCTACAACAG-39. The PCR product was purified and then cloned into the BamH1 site of pET-21a vector. The wild-type construct was further used to produce C192S mutation using overlap extension PCR. Four pairs of specific primers with BamH1 and Xho1 restriction sites were designed to make the SPE B for speB 345-398 , 59-GCGGATCCGATGGTGCTGACGGACG-39 and 59-CCGCTCGAGAGGTTTGATGCCTACAACA-39. The PCR products were purified and then cloned into the BamH1 and Xho1 restriction sites of the pET-42a vector. The recombinant plasmids were transformed into Escherichia coli BL21(DE3) pLyS strains. Recombinant proteins that contained N-terminal glutathione S-transferase (GST) and six histidines at the C-terminal end were produced by growing cells at 37uC overnight in LB medium containing 100 mg of isopropyl-b-D-thiogalactoside, 25 mg of kanamycin, 16 g of Bacto-Peptone, 8 g of Bacto-yeast extract, and 16 g of NaCl in 1 liter. Proteins were then purified using Ni 2+chelating chromatography (GE Healthcare). The purified recombinant proteins were further examined by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and confirmed by western blotting using rabbit anti-SPE B serum (kind gift from Dr. Y. S. Lin, National Cheng Kung University Medical College, Taiwan.), and 2 mg of bovine serum albumin (BSA) (Sigma) was used as a control.

Examining Binding of SPE B to Immunoglobulins
One gram of the CNBr-activated sepharose 4B gel (Pharmacia Biotech) was suspended in 250 ml of 1 mM HCl for 30 min at 4uC. The gel was then washed with alternating 750 ml of 1 mM HCl, 100 ml of distilled water, and 300 ml of the reaction buffer containing 0.1 M NaHCO 3 and 0.5 M NaCl (pH 8.0). The gel was suspended in 15 ml of the reaction buffer (pH 8.0) and was incubated with 10 mg of C192S, the mutant SPE B that lacks protease activity, at 4uC overnight. The coupled gel was washed with 250 ml of 1 M glycine/reaction buffer (pH 8.0) to block unused activated sites. Thereafter, the coupled gel was washed three times with alternating 200 ml of borate buffer containing 0.1 M boric acid and 1 M NaCl (pH 8.5), 100 ml of distilled water, 200 ml of acetate buffer containing 0.1M sodium acetate and 1 M NaCl (pH 4.0), and 100 ml of distilled water. The gel was then suspended in PBS [19].
One milliliter of normal human serum was passed through a CNBr-activated sepharose 4B-packed affinity column that was immobilized with C192S. Buffer containing 50 mM Tris (pH 7.0) was used to wash non-binding materials. After several washes, binding ligands of C192S were eluted with 0.1 M glycine (pH 3.0), and different fractions were collected from the point when the signal appeared on the recording graph by measuring the absorbance at 280 nm [18,19]. The contents of fraction 13 (F13) and purified immunoglobulins were verified using 12% SDS-PAGE and then transferred to a PVDF membrane (Millipore). After blocking, blots were developed with goat anti-human IgG (Calbiochem), IgM (Abcam), or IgA (Abcam) antibody that was diluted 5000-fold with PBS. Blots were then hybridized using horseradish peroxidase-conjugated rabbit anti-goat IgG (Calbiochem), and the protein bands were visualized using enhanced chemiluminescence (Amersham Biosciences).

Detecting Cleavage of Immunoglobulins by SPE B
The purified human IgG or the IgM-IgA mixture (400 mg/ml) was incubated with 20 mg/ml of mSPE B or C192S at 37uC for 1, 2, or 18 h with 5 mM dithiothreitol (DTT)/0.1 mM EDTA that was used to activate SPE B protease activity [18,19]. In another experiment, the purified human IgG or the IgM-IgA mixture (40 mg/ml) was incubated with 2 mg/ml of mSPE B in the absence or presence of different concentrations of rSPE B 345-398 (10, 20, or 40 mg/ml) at 37uC for 1 h with 5 mM dithiothreitol (DTT)/ 0.1 mM EDTA. The reaction mixtures mentioned above were separated using 12% SDS-PAGE and then detected by western blot with goat anti-human IgG (Calbiochem), IgM (Abcam), or IgA antibody (Abcam).

Enzyme-linked Immunosorbent Assay (ELISA)-type Binding Assay
The microtiter plate wells (NUNC) were coated with 2 mM of C192S or BSA (Sigma) in 50 ml of coating buffer containing 100 mM of NaHCO 3 and 100 mM of Na 2 CO 3 (pH 9.0) at 37uC for 1 h, and blocked with 200 ml of 5% skim milk (Sigma) at 4uC overnight. After several washes with 0.1% Tween-20/PBS (PBS-T), 50 ml of human serum diluted 100-fold with PBS was added into the wells and incubated at 4uC overnight. After several washes, 50 ml of peroxidase-conjugated goat anti-human IgG, IgM, or IgA antibody (Millipore) (1:10000) was added and incubated at 37uC for 1 h. Next, the TMB substrate (Vector Laboratories) was added, and the absorbance values were read at 650 nm. In a direct binding assay for Fc fragment, 50 ml of different concentrations of human Fc fragments of human IgG (Millipore) (0.5-5 mM) or purified human IgG (1 mM) were added to the C192S-or rSPE B 345-398 -coated wells. After that, peroxidase-conjugated goat anti-human IgG antibody (Millipore) (1:10000) was added and then the TMB substrate (Vector Laboratories) was developed, and the absorbance values were read at 650 nm.
Alternatively, for detection of immunoglobulins binding with SPE B recombinant proteins, the microtiter plate wells were coated with either 0.1 mg of purified human IgG or the IgM-IgA mixture in 50 ml of coating buffer (pH 9.0) at 37uC for 1 h. The plates were washed and then blocked with 200 ml of 1% BSA/PBS at 4uC overnight. After washing, 50 ml of the various SPE B recombinant proteins (2 mM) were added into the plates and incubated at 4uC overnight. After several washes with 0.1% PBS-T, mouse anti-GST monoclonal antibody (Abcam) (1:5000) was added and incubated for 1 h followed incubation by 100 ml of peroxidase-conjugated goat anti-mouse IgG antibodies (Calbiochem) (1:10000). After several wash with 0.1% PBS-T, the TMB substrate was added, and the absorbance values were read at 650 nm. The results of three experiments are represented and expressed as the mean 6 standard deviation.

Inhibition ELISA
The microtiter plate wells (NUNC) were coated with 2 mM of purified rSPE B 345-398 or BSA in 50 ml of the coating buffer containing 100 mM NaHCO 3 and 100 mM Na 2 CO 3 (pH9.0) at 37uC for 1 h. The plates were washed and then blocked with 200 ml of 1% BSA/PBS at 4uC overnight. Different concentrations of protein A or protein L (0.1-1 mg/ml) were add with purified human IgG (1 mg/ml) at varying molar ratio (0.5-5) and incubated at 37uC for 30 min, and was then added to rSPE B 345-398 or BSA-coated microtiter plates and incubated at 4uC overnight. After washing with 0.1% PBS-T, peroxidase-conjugated goat anti-human IgG antibody (Millipore) (1:10000) was added. Next, the TMB substrate (Vector Laboratories) was added, and the absorbance values were read at 650 nm. Relative binding activity was calculated as follows: activity = 100% 6 (A 650

Assessment of Complement Functional Activity
Complement activation through the classical pathway by human serum was detected by the Wielisa COMPL300 Total Complement Functional Screen kit (Wieslab AB) [19]. The kit provides strips of wells precoated with IgM for evaluation of the classical pathway. Following the kit instructions, the serum of a healthy individual, provided by the kit, was diluted 100-fold in specific buffers [19]. The diluted serum was then incubated with 20 mg/ml of mSPE B in the absence or presence of different concentrations (5, 10, or 20 mg/ml) of rSPE B 345-398 for 15 min in a 37uC water bath. Thereafter, mSPE B-treated human serum was added to strips and incubated for 1 h at 37uC. Alkaline phosphatase-conjugated anti-human C5b-9 antibody, provided by the kit, was added and incubated for another 30 min at room temperature. Finally, the substrate was added, and the absorbance values were read at 405 nm. In each assay, standard positive and negative control sera were provided in the kit. Complement activity was calculated as follows: activity = 100% 6 (A 405 (sample)-A 405 (negative control))/(A 405 (standard serum)-A 405 (negative control)) [18,19]. In each assay, samples, standard serum, and negative control serum were tested in duplicate at a fixed dilution. The percent inhibition was calculated as follows: 100% 6  Either PBS or 400 mg/ml of rSPE B 345-398 was mixed with an equal volume of Freund's complete adjuvant (Sigma-Aldrich), and then 0.25 ml of the emulsion was inoculated intraperitoneally into BALB/c mice. After that, 25 mg of rSPE B 345-398 emulsified with Freund's incomplete adjuvant (Sigma-Aldrich) was administered intraperitoneally every two weeks for a total of three boosts. The sera of mice were collected after immunization, and the titer of anti-rSPE B 345-398 antibody of each mouse was determined by ELISA, as previously described. After immunization, groups of 8 to 10 BALB/c mice were injected subcutaneously with 1 ml of air for two consecutive days to form an air pouch. Two days later, 0.2 ml of bacterial suspension containing 2.5 6 10 8 colony forming units (CFU) of S. pyogenes NZ131 was inoculated into the air pouch [29], and the animals were observed every day for a total of 14 days. During the observation time, the degree of skin lesion was photographed and measured using ImageJ 1.410 software (National Institutes of Health) [29]. The average damage area in each group was generated by examining the skin lesions from 8 to 10 mice.

Statistics
Statistical analysis was performed using ANOVA. Differences were considered significant at P,0.05. For the mouse model, the survival curves were compared for significance using the log-rank test.

C192S-bound Antibodies Present in Human Sera
To examine the binding of SPE B with antibodies in human sera, the SPE B mutant protein C192S was used because of the instability of the active form of SPE B (mSPE B) [18,19]. Human sera were passed through the affinity column packed with C192S immobilized on sepharose 4B. Different fractions were collected during the elution and their contents were analyzed by western blotting using anti-human IgG, IgM or IgA. When compared with protein L-purified human IgG, IgM and IgA, we found that fraction 13 (F13) contained human IgG, IgM, and IgA and that the amount of IgG in the F13 fraction is most abundant, IgM is less, and IgA is least abundant (Figure 1A-D). No IgD or IgE was found in the F13 fraction (data not shown) because of the low concentration of both isotypes in human sera. We observed similar results with an ELISA assay; the serum IgG and IgM binded more strongly to the C192S than IgA did ( Figure 1E). These results indicate that C192S was able to efficiently bind IgG, IgM, and IgA in human sera ( Figure 1).

SPE B Cleaved IgG, IgM, or IgA
The purified human IgG, or the IgM-IgA mixture were incubated with the active form mSPE B or C192S mutant at a ratio of 20:1 (w/w) for different times, and the reaction mixture was separated using SDS-PAGE and blotted with anti-human IgG, IgM, or IgA antibody. After the 1 h incubation, the amount of IgG and IgM were significantly reduced than IgA after mSPE B but not C192S treatment (Figure 2A). The cleavage of antibodies by mSPE B, regardless of the isotype, was almost complete after the 18 h incubation ( Figure 2B). These results suggest that active SPE B was able to bind serum IgG, IgM, or IgA and cleave them effectively.    Figure 3A, and their antigenicity was confirmed by western blotting with anti-SPE B antibody ( Figure 3B).

SPE B Bound Serum Antibodies by its C-terminal Domain
The antibody binding motif of SPE B was further determined with these recombinant SPE B fragments by ELISA assay. The rSPE B 146-398 , representing mSPE B, was able to bind IgG or the IgM-IgA mixture but with different binding capacities (Figure 4). Regardless of IgG or the IgM-IgA mixture, the binding activity of rSPE B 146-280 , which contains the N-terminal domain of mSPE B, had a poor binding capacity to antibody isotypes. Dissecting the C-terminal domain of mSPE B, we found that rSPE B 345-398 had the strongest binding activity to IgG or the IgM-IgA mixture ( Figure 4A and 4B). However, rSPE B 281-358 manifested stronger binding activity to IgG than N-terminal rSPE B146-280 ( Figure 4A).These results suggest that the C-terminal domain, especially amino-acid residues 345-398, of SPE B is the major site for antibody binding.

The C-terminal Domain of SPE B Bound the Fc Portion of IgG
We chose IgG, the most abundant antibody in human serum, to examine its binding motif with the C-terminal domain of SPE B (rSPE B 345-398 ). We used protein L and protein A to compete with the Fab and Fc portions of IgG binding to rSPE B 345-398, respectively. The results indicate that protein A but not protein L could inhibit IgG binding to rSPE B 345-398 ( Figure 5A) and that the competitive inhibition was dose-dependent for protein A ( Figure 5B). Even at 2 mg per ml, the molar ratio of protein L/IgG was 10, protein L still could not interfere with IgG binding to rSPE B 345-398 (data not shown). Moreover, a direct ELISA binding assay of purified IgG Fc fragment with rSPE B 345-398 or C192S also indicate that SPE B uses its C-terminal domain to bind the Fc portion of IgG ( Figure 5C).

rSPE B 345-398 Interferes with Cleavage of Immunoglobulins and Inhibition of Complement Activation by SPE B
The active form of mSPE B can efficiently cleave IgG, IgM, or IgA within 1 h of incubation (Figure 2A). To confirm that the amino acid sequence 345-398 is the major binding site of mSPE B, rSPE B 345-398 was used as an inhibitor to interfere with mSPE B binding to antibody isotypes. As shown in Figure 6, rSPE B 345-398 could effectively inhibit cleavage of antibody isotypes by mSPE B in a dose-dependent manner. Our previous study indicates that mSPE B is able to impair the classical pathway of complement  activation [19]. To examine whether rSPE B 345-398 was able to interfere with mSPE B -mediated inhibition of classical complement activation, a simple ELISA-based assay for the classical pathway was used [19]. The results show that when the dose of rSPE B 345-398 was increased up to 20 mg/ml, mSPE B-mediated inhibition of classical complement activation dropped from 85% to 59% (Figure 7), indicating that rSPE B 345-398 could restore complement activation even when mSPE B is present.

Immunization of rSPE B 345-398 could Protect Mice from GAS Infection
Immunization of BALB/c mice using rSPE B 345-398 was able to induce production of a high titer of anti-rSPE B 345-398 antibodies ( Figure 8A). To examine whether anti-rSPE B 345-398 antibodies possess neutralizing activity to block mSPE B actions, we infected control mice and rSPE B 345-398 -immunized mice with a lethal dose of GAS strain NZ131 (2.5 6 10 8 CFU/mouse) through the air-pouch route, which mimics local necrotizing fasciitis [29], and monitored survival. The results indicate that control mice all died within 4 days after GAS infection, while 80% of rSPE B 345-398immunized mice were still alive 14 days after GAS infection ( Figure 8B). Moreover, the skin lesions induced by GAS infection were smaller in rSPE B 345-398 -immunized mice (81621 mm 2 ) compared to control mice (187638 mm 2 ). These results indicate that rSPE B 345-398 could be used as a vaccine to block mSPE B actions and efficiently protect mice from GAS infection.

Discussion
Several reports suggest that SPE B contributes to increasing GAS invasion [10,16,17,29], degrading fibronectin, vitronectin, and fibrinogen [22,32], cleaving urokinase plasminogen activator receptor [26], activating matrix metalloprotease [24], generating active IL-1b from its precursor [25], damaging endothelial cells [16], and reducing the phagocytic activity of U937 monocytic cells Figure 8. Immunization of rSPE B 345-398 protects mice from GAS-induced death. (A) The immunization protocol was described in Materials and Methods. Sera were collected, and the titer of anti-rSPE B 345-398 antibody was detected by ELISA, as described in Materials and Methods. (B) The survival rates in PBS-or rSPE B 345-398 -immunized mice after infection with NZ131 GAS strains. Groups of 8 to 10 BALB/c mice were immunized with PBS (n = 8) or rSPE B 345-398 (n = 10) and were then inoculated via the air pouch route with 2.5 610 8 CFU of GAS strain NZ131. The survival curves were compared for significance using the log-rank test for the rSPE B 345-398 versus the PBS group (P,0.01). doi:10.1371/journal.pone.0055028.g008 [33]. Our previous studies using an air pouch animal model show the correlation of SPE B with GAS-induced tissue damage, bacterial dissemination and bacteremia [16,29]. In addition, SPE B has been known to directly interfere with killing by the immune system through cleavage of the Fc portion of antigen-bound IgG [20,21], properdin [18], and complement C3 [19], thereby contributing to bacterial evasion of antibody-and complementmediated opsonophagocytosis.
Previous studies indicate that opsonizing antibodies specific for M and M-like proteins of GAS are capable of providing protection against GAS infections [34]. Through antibody-induced complement activation followed by opsonization by C3b and its cleavage fragment iC3b, complement receptor 3 (CR3; CD11b/CD18) on phagocytes efficiently takes up and eliminates GAS [35]. SPE B cleaved GAS-bound opsonizing antibodies resulting in failure of complement deposition as well as GAS elimination, enhancing bacterial dissemination. In this study, we showed that the Cterminal domain amino acid residues 345-398 was the major binding site of SPE B with immunoglobulins ( Figure 4) using an ELISA binding assay. The results of inhibition ELISA indicated that protein A could inhibit IgG binding to rSPE B 345-398 ( Figure 5A and 5B). The results of the Fc fragment-direct binding assay also indicated that rSPE B 345-398 bound the Fc portion of IgG; even the binding activity of the Fc fragment of IgG to rSPE B 345-398 was weaker than that to the zymogen form C192S ( Figure 5C). The recombinant protein, rSPE B 345-398 , could efficiently block cleavage of immunoglobulins by mSPE B (Figure 6) and interfere with mSPE B-induced inhibition of complement activation (Figure 7). These functional assays indicate that the C-terminal domain of SPE B is the actual binding site for immunoglobulins. This finding is consistent with the study reported by Wang et al., which showed, by nuclear magnetic resonance spectroscopy, that the flexible C-terminal loop of SPE B may play an important role in controlling substrate binding, resulting in its broad substrate specificity [36].
Clinical investigations indicate that GAS not only induces acute invasive infection but also causes immune-mediated post-infectious sequelae, such as acute rheumatic fever and post-streptococcal glomerulonephritis [37,38]. Molecular mimicry between group A streptococcus and host antigens has an important role in the development of post-streptococcal sequelae, including glomerulonephritis and rheumatic heart disease. Luo et al. indicate that anti-SPE B antibodies exhibit characteristics of autoantibodies, which cross-react with endothelial cells, causing apoptotic cell death in the mouse heart valve. They further identify the SPE B antigenic epitope recognized by anti-SPE B antibodies, indicating that the major epitope of anti-SPE B antibodies (mAb 10G) is localized to amino-acid residues 296-310 of SPE B [39]. Amino-acid residues 296-310, the probable epitope of the autoantigen, were distinct from amino-acid residues 345-398 of SPE B. Even rSPE B 281-358 seemed to manifest binding activity with IgG ( Figure 4A). We excluded the probable epitope of the autoantigen and based on the results of the ELISA binding assay and functional assays, we used the recombinant protein rSPE B 345-398 to immunize BALB/c mice and found that rSPE B 345-398 -immunized mice could effectively protect mice from skin lesions and GAS NZ131induced death (Figure 8).
Using rSPE B 345-398 to immunize mice, we did not find characteristics of glomerulonephritis in immunized mice ( Figure  S1). The serum creatinine (CRTN) and blood urea nitrogen (BUN) levels of rSPE B 345-398 -immunized group (CRTN: 0.4660.05 mg/dl; BUN: 20.562.38 mg/dl) were similar to the control group (CRTN: 0.4460.04 mg/dl; BUN: 21.862.48 mg/ dl). Those results indicate that immunization of rSPE B 345-398 was able to protect mice from GAS infection but not to induce autoimmune-like symptoms. Furthermore, six synthetic peptides which contain C-terminal region extended from residues 281 to 398 of SPE B were examined by IgG binding assay. ELISA results indicated the C-terminal motifs of SPE B, Gly 346 -Gly 360 and Ala 376 -Pro 398 , both mediated the binding of IgG. Immunization with synthetic peptides either Gly 346 -Gly 360 or Ala 376 -Pro 398 protected mice against challenge with a lethal dose of GAS (unpublished data). In this study, we used the NZ131 GAS strain to infect mice, while the protective effects of rSPE B 345-398 on other GAS clinical isolates need further investigation.