Contribution of Each of Four Superantigens to Streptococcus equi-Induced Mitogenicity, Gamma Interferon Synthesis, and Immunity

Streptococcus strains. S. equi strain 4047 was originally isolated in 1990 from a submandibular abscess of a New Forest pony and has been maintained in the culture collection of the Animal Health Trust (Newmarket, United Kingdom). Twenty-eight isolates of S. equi (7364, JKS 225, 7325, 7171, 303, 3155, JKS 063, JKS 043, 7329, 3682, CF32, 1351, SA, 8229, 7326, 3156 7325, 4047, 7344, 3154, JKS044, 181063, 1350, 1931, 7060, 7140, and JKS 559, all sequence type 179 [ST-179], and 7329 [ST-151]) were used. Twenty-two of these 28 S. equi isolates have been shown to contain seeI, seeH, seeL, and seeM by quantitative PCR (qPCR) (10). Twenty-one isolates of S. zooepidemicus (5845 [ST-45], H70 [ST-1], 3512 [ST-143], 8250 [ST-146], 4859 [ST-119], 5770 [ST-106], 4895 [ST-119], 5936 [ST-106], 2410 [ST-144], 8295 [ST-104], 8575 [ST-97], BHS41 [ST-10], 6458 [ST-82], D14a [ST-2], 2958 [ST-178], 5622 [ST-106], 8275 [ST-104], 5768 [ST-112], 8301 [ST-104], 4863 [ST-108], and 4887 [ST-108]) were used in this study. These S. zooepidemicus isolates belong to 16 different sequence types (as defined by multilocus sequence typing [MLST] [37]), from which representatives strains were screened for the presence of seeI, seeH, seeL, and seeM (10). Furthermore, all 21 S. zooepidemicus isolates were directly screened for the presence of these four superantigens. seeI and seeH were found to be absent in all S. zooepidemicus isolates. The isolates 5936, 5770, 4895, and 5622 belong to sequence type 106, which has previously been shown to contain some strains that possess seeL and seeM (10). These four isolates have been screened and found to possess seeL and seeM. Details of all isolates are available on the online MLST database (http://pubmlst.org./szooepidemicus/ ; accessed 13 July 2009).

Production of recombinant S. equi sAgs.The genes encoding the mitogenic toxins were cloned as glutathione S-transferase (GST) fusions using pGEX-3X and the primers listed in Table 1. The cloned fragments corresponded to codons 26 to 259 (seeI), 33 to 236 (seeH), 27 to 259 (seeL), and 36 to 262 (seeM), and in each case the DNA encoding the signal peptide was omitted. PCR products were generated using strain 4047 DNA and Phusion DNA polymerase (New England Biolabs). Purified PCR products were cut with either BamHI and SmaI (seeI and seeH) or BamHI and EcoRI (seeL and seeM), ligated into the pGEX-3X vector cut with the appropriate restriction enzymes, and transformed into Escherichia coli DH10B, and transformants were selected at 37°C on 2×YT (yeast extract-tryptone) plates containing ampicillin (100 μg/ml). For expression, cultures (10 ml) were grown overnight at 37°C in 2×YT with ampicillin (100 μg/ml), diluted 1/10 the next day into 100 ml of 2×YT, grown for 1 h, induced by the addition of 0.5 mM isopropyl-β-d-thiogalactopyranoside, and grown for a further 4 h at 28°C. Cells were harvested by centrifugation and lysed, and fusion protein was recovered using glutathione-Sepharose beads (Amersham). Factor Xa (Amersham) was then used to cleave the recombinant proteins from the GST tag. The concentrations of purified recombinant sAgs were quantified, and the proteins were stored at −70°C in 50% glycerol. The purity of the proteins was assessed by polyacryamide gel electrophoresis and staining with Coomassie blue. All of the recombinant sAgs migrated at the expected sizes and were shown to be greater than 99% pure.

Generation of S. equi mutants containing superantigen deletions.In order to generate single deletion mutants lacking seeH, seeI, seeL, or seeM (ΔH, ΔI, ΔL, or ΔM), double deletion mutants lacking seeH and seeI (ΔHI) or seeL and seeM (ΔLM), a triple deletion mutant lacking seeI, seeL, and seeM (ΔLMI), and a quadruple deletion mutant lacking seeH, seeI, seeL, and seeM (ΔLMHI) of S. equi strain 4047 by allelic replacement, PCR products flanking the sequences to be deleted were generated using Vent DNA polymerase (New England Biolabs) and the relevant combination of primer pairs listed in Table 1. The corresponding PCR products were then digested with the restriction endonucleases EcoRI and HindIII (5′ product) and HindIII and SalI (3′ product) and ligated into the EcoRI- and SalI-digested pG+Host9:ISS1 plasmid (19) in a three-way ligation to form the different deletion constructs, pGΔsAg. Engineering of a HindIII site into primers as part of the cloning strategy results in the introduction of a non-sAg DNA sequence at the site of the deletion. The different pGΔsAg plasmids were transformed into E. coli TG1repA⁺, and transformants were selected at 37°C on LB plates containing erythromycin (150 μg/ml).

Allelic replacement mutagenesis.Transformation of the encapsulated strain 4047 with the different pGΔsAg plasmids was achieved using a modification of the method described by Simon and Ferretti (27), as described by Hamilton et al. (8). Allelic replacement of the desired sAg gene(s) was performed as previously described for a ΔprtM mutant (8), and the deletions were confirmed by PCR and DNA sequencing. Absence of sAg expression in S. equi mutants was confirmed after RNA extraction by using an RNeasy midikit (Qiagen) and analysis by quantitative reverse transcription-PCR (qRT-PCR) with the primer pairs listed in Table 1. Supernatant samples were collected from overnight cultures (stationary phase) and filter sterilized. S. equi mutants displayed no growth defect or phenotypic modification.

Blood donors and convalescent-phase sera.Residual blood samples from the Animal Health Trust sera archive from several horses or ponies prior to and after strangles diagnosis, including Welsh Mountain ponies, thoroughbreds, Gypsy cobs, Shetland ponies, and donkeys, were used in this study.

Purification of equine PBMC and proliferation assay.Equine and asinine PBMC were purified from blood by centrifugation on a Ficoll density gradient and were cultured in complete medium which consisted of RPMI 1640 medium containing 10% heat-inactivated fetal calf serum (FCS), 2 mM l-glutamine, 55 μM 2-mercaptoethanol, and 1% penicillin-streptomycin (100 U/ml and 100 μg/ml, respectively) (all supplied by Sigma-Aldrich). PBMC were stimulated in vitro with S. equi supertantigens at the indicated concentrations. Equine PBMC proliferation was quantified by either incorporation of [³H]thymidine (³HT)or carboxyfluorescein diacetate succinimidyl ester (CFSE; Sigma-Aldrich) staining as follows. For [³H]thymidine incorporation, 200-μl aliquots of equine PBMC at 10⁶ cells/ml were cultured in triplicate with S. equi sAgs or medium alone as a negative control in a 96-well plate. After 3 days of incubation, 0.5 μCi of [³H]thymidine (Amersham Biosciences) was added to the culture for 8 h. Plates were stored at −20°C and thawed immediately prior to harvesting on a Filtermate harvester (Perkin-Elmer, United Kingdom) and counted on a TopCount NXT apparatus (Perkin-Elmer). Equine PBMC proliferation is expressed as the index of stimulation (SI), which was calculated as follows: (experimental cpm)/(control cpm). For CFSE staining, 10⁷ equine PBMC were incubated with 5 μM CFSE in phosphate-buffered saline (PBS) for 20 min, washed three times in PBS plus 10% FCS, and kept in culture for 24 h prior to stimulation with sAgs or medium alone. Cell cultures were analyzed by flow cytometry.

Analysis of cell phenotype and IFN-γ synthesis.For the analysis of cell phenotype, a total of 10⁶ cells were stimulated in vitro with sAgs at the indicated concentrations or medium alone as negative controls. After 3 days of culture at 37°C, surface labeling was performed with antibodies that recognize equine CD4 (clone HB61A), equine CD5 (clone HB19a; CD5 is not expressed on equine B lymphocytes [12, 18]), and equine CD8α (clone 73/6.9.1), used at 1 μg per 10⁶ PBMC. Control isotype antibodies were obtained from DakoCytomation (Glostrup, Denmark). Bound antibodies were recognized by rhodamine-phycoerythrin (RPE)-conjugated goat anti-mouse IgG antibody (Southern Biotechnology). For measurement of IFN-γ synthesis by equine peripheral blood lymphocytes (PBL), a total of 10⁶ fresh PBMC were stimulated in vitro with sAgs at the indicated concentrations or medium alone as a negative control. Brefeldin A (BFA; BD GolgiPlug from BD Biosciences) was added at the indicated times after the stimulation (1 μl/ml final concentration, according to the manufacturer's recommendations). After overnight incubation, expression of CD5 or CD8 was determined using the mouse anti-equine CD5 antibody (clone HB19a, IgG2a isotype; VMRD, Pullman, WA) or the mouse anti-equine CD8α antibody (clone 73/6.9.1, IgG3 isotype; VMRD). Fluorescein isothiocyanate (FITC)-conjugated rabbit anti-mouse IgG2a (Serotec) or FITC-conjuguated goat anti-mouse IgG3 (Southern Biotechnology) was used as secondary antibody. Cells were fixed in 3% (vol/vol) formaldehyde in PBS prior to detection of equine IFN-γ by intracellular cytokine staining (ICC) using a biotinylated antibody specific for equine IFN-γ (clone CC302; Serotec) and streptavidin-RPE (Serotec) in permeabilization buffer (PBS with 0.5% bovine serum albumin and 0.2% saponin; Sigma-Aldrich). For single-color staining to detect intracellular IFN-γ, unbiotinylated clone CC302 antibody and FITC-conjugated goat anti-mouse IgG antibody (Jackson ImmunoResearch) were used. IFN-γ detection was performed in duplicate or triplicate. The gate used to select PBL during the analysis was based on forward and side scatter (FSC and SSC) characteristics and CD5 staining (17). The percentage of sAg-specific IFN-γ synthesis was calculated according to the following formula (22): (percent sAg-stimulated IFN-γ synthesis) − (percent medium-stimulated IFN-γ synthesis). The IFN-γ SI was calculated as follows: (experimental IFN-γ synthesis)/(control IFN-γ synthesis).

Measurement of sAg-specific antibody response by ELISA and neutralizing activity.Microtiter plates (96-well Immunlon 2HB; Dynex Technologies) were coated overnight at 4°C with 100 μl of recombinant S. equi sAgs at a concentration of 2 μg per ml in carbonate-bicarbonate buffer (0.05 M, pH 9.6; Sigma-Aldrich). Plates were washed four times with 400 μl of PBS plous 0.05% (vol/vol) Tween 20 (PBS-T). The plates were blocked with 300 μl of PBS-T plus 5% (vol/vol) Marvels skimmed milk per well for 1 h at 37°C and then washed four times with 400 μl of PBS-T. Serum samples (100 μl) diluted 1/800 in PBS were added to the wells. Plates were incubated at 37°C for 1 h and washed again four times with 400 μl of PBS-T. Detection of superantigen-specific bound antibodies was performed with 100 μl of horseradish peroxidase (HRP)-conjugated goat anti-horse IgG (Bethyl Laboratories) diluted 1/600 in PBS-T plus 1% (vol/vol) Marvels skimmed milk. Plates were incubated for 1 h at 37°C and then washed six times with 400 μl of PBS-T. HRP was detected by incubation with 100 μl of 3,3′,5,5′-tetramethylbenzidine (TMB) microwell peroxidase substrate (Kirkegaard & Perry Laboratories) at room temperature for 10 min. The reaction was stopped by the addition of 100 μl of H₂SO₄ (0.18 M). Results are expressed as the optical density (OD) at 450 nm. For the SeeI- and SeeH-specific enzyme-linked immunosorbent assays (ELISAs), the cutoff values (1 and 0.5 OD, respectively) were defined using receiver operator characteristic (ROC) curves. For SeeL- and SeeM-specific ELISAs, the cutoff values (0.25 and 0.68 OD, respectively) were defined as the 2-fold average day zero absorbance (sample prior to S. equi infection). To measure neutralizing activity of equine sera, 5 μl of serum was incubated for 1 h with 10 μl of culture supernatant or recombinant sAg. After this period of incubation, 185 μl of equine PBMC culture suspension was added (final volume, 200 μl). The percentage of neutralization was calculated according to the following formula: 100 − [(proliferation in the presence of convalescent-phase serum × 100)/(proliferation in the presence of preinfection serum)].

Statistical analyses.Statistical analyses were performed with the StatGraphics Plus program for Windows. Analysis of variance (ANOVA) was used to test the significance of data between groups (Tukey's honest significant difference test or Bonferroni procedure with a confidence interval of 95%). The level of significance was set at a P value of <0.05. If variances between each group were not homogenous (P < 0.05), data were log transformed to allow ANOVA to be applied.

S. equi sAg-induced lymphoproliferation in vitro.CFSE-stained equine PBMC were cultivated in the presence of recombinant S. equi sAgs or medium alone as negative controls for 3 days before being analyzed by flow cytometry (Fig. 1A). SeeI, SeeL, and SeeM repeatedly induced proliferation of equine PBL in vitro. Superantigen-induced proliferation was dose dependent (Fig. 1B), with activity observed at very low concentrations (e.g., 50 fg/ml for SeeM). SeeM possessed the most potent activity, with a median half-maximal proliferation response (P₅₀) of 0.9 pg/ml (n = 6). SeeI and SeeL presented median P₅₀ values of 36 pg/ml (n = 4) and 92 pg/ml (n = 5), respectively. Proliferation of equine PBMC was detected from the second day of culture in the presence of sAgs (Fig. 1C). A peak of proliferation was reached at the third, fourth, and beyond the fifth day of culture with SeeM, SeeL, and SeeI, respectively. The phenotype of superantigen-activated PBL was analyzed after 4 days of culture. CD5⁺ CD4⁺ T lymphocytes represented the majority of stimulated PBL (Fig. 1D).

• Open in new tab
• Download powerpoint

FIG. 1.

Cell proliferation induced by S. equi superantigens. (A) Equine PBMC labeled with CFSE and cultured for 3 days in the presence of medium alone (gray) or S. equi sAgs at 0.125 μg/ml (open). A peak in M₁ indicates a fluorescence decrease due to cell division. (B) Dose response. A total of 2 × 10⁵ equine PBMC were cultured for 4 days in triplicate with the indicated concentration of sAg and incubated with ³HT for 16 h before measurement of proliferation. (C) Kinetics of proliferation. A total of 2 × 10⁵ equine PBMC were cultured with 0.125 μg/ml sAg in triplicate for 1 to 5 days, with incorporation of ³HT during the last 16 h of each time period. Results are presented as the stimulation index. (D) Representative cell phenotypes of S. equi sAg-activated PBL. Equine PBMC were cultured with 0.125 μg/ml sAg for 4 days prior to analysis. CD5, CD4, and CD8 lymphoblast staining levels after SeeM stimulation are shown as representative phenotypes of S. equi sAg-activated cells.

Recombinant SeeH did not stimulate proliferation of PBMC purified from Welsh mountain ponies (Fig. 1A) or the other pony and horse breeds examined in this study (i.e., thoroughbred [n = 2], Gypsy cob [n = 1], and Shetland pony [n = 1]) (Table 2). However, PBMC purified from six of eight donkeys responded to SeeH stimulation. As previously observed with equine PBMC, SeeI, SeeL, and SeeM also induced the proliferation of asinine PBMC.

Recombinant S. equi sAgs (SeeI, SeeL, and SeeM) and S. equi culture supernatants stimulated IFN-γ synthesis in vitro.Equine PBMC were stimulated for 24 h with S. equi recombinant sAgs, and IFN-γ synthesis was measured by flow cytometry. SeeI, SeeL, and SeeM stimulated the synthesis of IFN-γ in vitro (Fig. 2A). IFN-γ⁺ PBL were mostly of the CD5⁺ CD4⁺ phenotype (Fig. 2B). SeeM possessed the most potent activity, with a median half-maximal IFN-γ response (I₅₀) of 0.41 ng/ml (n = 4) (Fig. 2C), SeeL and SeeI presented an I₅₀ of 2.24 ng/ml (n = 6) and 4.8 ng/ml (n = 5), respectively. IFN-γ synthesis was detected as early as 6 h after S. equi-sAg stimulation and reached a maximum between 24 and 48 h. The percentage of PBL synthesizing IFN-γ after stimulation with S. equi recombinant sAgs decreased after 48 h (data not shown). SeeH did not induce IFN-γ synthesis by equine PBMCs.

• Open in new tab
• Download powerpoint

FIG. 2.

IFN-γ synthesis induced by S. equi superantigens. (A) A total of 1 × 10⁶ equine PBMC were incubated overnight with medium alone or 0.125 μg/ml sAg. Density plots are representative of at least three experiments. (B) Representative cell phenotypes of IFN-γ⁺ PBL after stimulation with S. equi sAg. Equine PBMC were cultured with 0.125 μg/ml sAg for 4 days prior to analysis. CD5, CD4, and CD8 lymphoblast staining after SeeM stimulation is shown as a representative phenotype of S. equi sAg-activated cells. (C) Dose response of IFN-γ synthesis. A total of 1 × 10⁶ equine PBMC were incubated overnight with the indicated concentrations of sAgs. IFN-γ synthesis was detected by flow cytometry after intracellular staining.

Culture supernatants from 28 S. equi isolates were screened for stimulation of IFN-γ synthesis and mitogenic activity when cultured in the presence of equine PBMC. Culture supernatants from S. zooepidemicus isolates (n = 18) that were negative for mitogenic activity were used as controls. As illustrated in Fig. 3, all S. equi isolates stimulated IFN-γ synthesis after 24 h of incubation. The percentage of IFN-γ⁺ PBL induced by S. equi culture supernatants was similar to the IFN-γ response induced by 0.125 μg/ml of recombinant SeeM. The average IFN-γ SI for S. equi supernatants was 8.5 ± 2.3 (n = 28) and was 1.3 ± 0.7 for S. zooepidemicus supernatants (n = 18). Repartition of the IFN-γ SI induced by S. zooepidemicus culture supernatants (n = 18) revealed a cutoff value of 3.1 (upper limit of distribution containing 99% of results) above which results were considered positive. Four culture supernatants from S. zooepidemicus isolates, which were positive for mitogenic activity, were also found to be positive for IFN-γ synthesis (Fig. 3B).

• Open in new tab
• Download powerpoint

FIG. 3.

IFN-γ synthesis induced by S. equi culture supernatants. (A) Typical density plot of 1 × 10⁶ equine PBMC incubated overnight with medium alone, S. zooepidemicus H70, or S. equi (isolate JKS044) culture supernatants or with 0.125 μg/ml SeeM. IFN-γ synthesis was detected by flow cytometry after intracellular staining. (B) Biparametric histogram of cell proliferation and IFN-γ synthesis induced by culture supernatants (1/20) of 28 S. equi isolates (open circles) and 22 S. zooepidemicus isolates (black circles). For cell proliferation, 2 × 10⁵ equine PBMC were cultured for 4 days in triplicate with culture supernatants and incubated with ³HT for 16 h before measurement of proliferation. For IFN-γ synthesis, 1 × 10⁶ equine PBMC were incubated overnight with culture supernatants. IFN-γ synthesis was detected by flow cytometry after intracellular staining. Dotted lines represent thresholds above which the response was considered positive (>2 for cell proliferation and >3.1 for IFN-γ synthesis).

Deletion of sAg genes abrogates mitogenic activity and IFN-γ synthesis induced by S. equi culture supernatants.Equine PBMC proliferation and IFN-γ synthesis (Fig. 4A and B, respectively) were measured after stimulation with culture supernatants from wild-type strain 4047 (control) or S. equi mutants with superantigen deletions. As illustrated in Fig. 4, the impact of single sAg deletion (ΔI, ΔH, ΔL, or ΔM) on S. equi supernatant-induced cellular responses was limited, with only the seeL deletion resulting in a statistically significant inhibition of lymphoproliferation and IFN-γ synthesis (19.1 ± 12.9 and 53.3 ± 16.4, respectively). The combined deletion of seeL and seeM (ΔLM) significantly increased the inhibition of both lymphoproliferation and IFN-γ synthesis (62.8 ± 17.8 and 84 ± 12, respectively), compared to single deletion alone or the seeH and seeI double deletion strain (ΔHI). Deletion of seeI, seeL, and seeM (ΔILM) or all sAg genes (ΔIHLM) entirely abrogated mitogenic activity and IFN-γ synthesis induced by S. equi culture supernatants compared with wild-type strain 4047.

• Open in new tab
• Download powerpoint

FIG. 4.

Box-and-whisker plot histograms of cell proliferation (A) and IFN-γ synthesis (B) induced by S. equi culture supernatants after sequential deletion of sAgs. (A) For cell proliferation, 2 × 10⁵ equine PBMC were cultured for 4 days in triplicate with culture supernatants (1/20) and incubated with ³HT for 16 h before measurements of proliferation. Results are expressed as the percentage of proliferation inhibition compared with the mitogenic activity of wild-type S. equi strain 4047. (B) For IFN-γ synthesis, 1 × 10⁶ equine PBMC were incubated overnight with culture supernatants (1/20). IFN-γ synthesis was detected by flow cytometry after intracellular staining. Results are expressed as the percent inhibition of IFN-γ synthesis compared with wild-type strain 4047. Differences between groups are indicated with a letter code (a to d). For example, a and b indicate a statistical difference (P < 0.05) between the two groups, and a group noted with ab is not statistically different from a group noted a or b (P > 0.05). The most relevant P values are indicated. The rectangle represents 50% of the observed response for each group, the horizontal line indicates the median of the group, and the cross indicates the average of the group. Squares indicate outlier results. n indicates the number of samples measured per group.

Infection with S. equi induced a sAg-specific antibody response in vivo.Convalescent-phase sera have been shown to contain antibodies to both SeeI and SeeH. Furthermore, horses that recover from strangles or are immunized with SeeI are resistant to the pyrogenic effects of SeeI (2, 5). The antibody response to all four S. equi sAgs was measured by ELISA in seven horses during strangles outbreaks (Fig. 5). The initial sera (day zero) were taken from healthy horses and were negative for the presence of S. equi sAg-specific antibodies (with the exception of horse number 7, which presented low levels of SeeH- and SeeL-specific antibodies). It is assumed that these horses had either not yet been naturally infected with S. equi or were at an early stage of natural infection (within a few days) at day zero. While the precise date of infection is not known, these horses were diagnosed with S. equi infection a few days after the initial serum sample had been taken. SeeI- and SeeH-specific antibody levels rose quickly above the background cutoff thereafter. The peak of the antibody response for each horse was between 20 and 80 days after the first sample. The antibody responses to SeeL and SeeM were less homogenous, with three and two horses that did not seroconvert to SeeM and SeeL, respectively.

• Open in new tab
• Download powerpoint

FIG. 5.

Kinetics of S. equi sAg-specific antibody responses in vivo. The antibody responses to all four S. equi sAgs were quantified by ELISA using sera taken from seven horses during strangles outbreaks. The initial sera (day zero) were taken from healthy horses with no signs of disease. These horses were diagnosed with strangles infection a few days after the initial serum samples were taken. Dotted lines represent the threshold above which the antibody response was considered positive.

Convalescent-phase sera have limited neutralizing sAg activity in vitro.Equine PBMC were cultivated with recombinant SeeI, SeeL, and SeeM (concentration close to 10 times the calculated P₅₀) that were incubated with sera (5 μl) purified from six horses prior to and after S. equi infection (convalescent-phase sera). Convalescent-phase sera consistently reduced SeeI-induced proliferation (0.5 ng/ml) with a positive correlation (R² = 0.84; n = 11) between the SeeI antibody titer and the percentage of neutralization (Fig. 6A). The abilities of convalescent-phase sera to neutralize recombinant SeeL- and SeeM-induced lymphoproliferation (1 ng/ml and 10 pg/ml, respectively) varied between individuals, as illustrated in Fig. 6B and C. SeeL- and SeeM-specific antibodies were measured in horses 2 to 5 after S. equi infection. Only sera from horses 2 and 5 showed some levels of neutralizing activity. No correlations were found between SeeL or SeeM antibody titers and the percentage of neutralization (R² = 0.24 and 0.55, respectively). Convalescent-phase sera from horses 3 to 5 were further tested against culture supernatants from wild-type strain 4047 and ΔHI and ΔLM deletion mutants. Inhibition of SeeI-induced lymphoproliferation by convalescent-phase sera was confirmed when culture supernatant from ΔLM S. equi mutants was used (diluted 1/20) to stimulate equine PBMC (data not shown). Convalescent-phase serum from horse number 5 was the only sample to significantly inhibit proliferation induced by all culture supernatants tested (data not shown).

• Open in new tab
• Download powerpoint

FIG. 6.

Quantification of sAg-neutralizing activity of convalescent-phase horse sera (1 to 3 different time points per horse). Equine sera purified from horses numbers 1 to 6 (prior to and after S. equi infection) were incubated for 1 h with recombinant SeeI (A), SeeL (B), or SeeM (C) prior to culture with 2 × 10⁵ equine PBMC. The final concentrations of SeeI, SeeL, and SeeM were 0.5 ng/ml, 1 ng/ml, and 10 pg/ml, respectively. Proliferation levels induced by recombinant sAgs incubated with sera purified prior to S. equi infection and convalescent-phase sera were compared after 4 days of culture. The percentages of proliferation inhibition (open columns) are expressed on the left y axes. Superantigen-specific antibody titers (gray columns) are expressed on the right y axes. Dotted lines represent the thresholds above which the antibody response was considered positive.