Identification of three novel superantigen-encoding genes in Streptococcus equi subsp. zooepidemicus, szeF, szeN, and szeP

Abstract

The acquisition of superantigen-encoding genes by Streptococcus pyogenes has been associated with increased morbidity and mortality in humans, and the gain of four superantigens by Streptococcus equi is linked to the evolution of this host-restricted pathogen from an ancestral strain of the opportunistic pathogen Streptococcus equi subsp. zooepidemicus. A recent study determined that the culture supernatants of several S. equi subsp. zooepidemicus strains possessed mitogenic activity but lacked known superantigen-encoding genes. Here, we report the identification and activities of three novel superantigen-encoding genes. The products of szeF, szeN, and szeP share 59%, 49%, and 34% amino acid sequence identity with SPEH, SPEM, and SPEL, respectively. Recombinant SzeF, SzeN, and SzeP stimulated the proliferation of equine peripheral blood mononuclear cells, and tumor necrosis factor alpha (TNF-α) and gamma interferon (IFN-γ) production, in vitro. Although none of these superantigen genes were encoded within functional prophage elements, szeN and szeP were located next to a prophage remnant, suggesting that they were acquired by horizontal transfer. Eighty-one of 165 diverse S. equi subsp. zooepidemicus strains screened, including 7 out of 15 isolates from cases of disease in humans, contained at least one of these new superantigen-encoding genes. The presence of szeN or szeP, but not szeF, was significantly associated with mitogenic activity in the S. equi subsp. zooepidemicus population (P < 0.000001, P < 0.000001, and P = 0.104, respectively). We conclude that horizontal transfer of these novel superantigens from and within the diverse S. equi subsp. zooepidemicus population is likely to have implications for veterinary and human disease.

FIG. 1.

(A) Superantigen characteristic signature sequences. The sAg PROSITE sequences and zinc binding motif are highlighted in red. The group 5 sAg-specific α3-β8 loop is shown in the red box. (B) Neighbor-joining tree showing phylogenetic relationships of known full-length sAgs. S. equi subsp. equi sAgs and novel S. equi subsp. zooepidemicus sAgs are highlighted in blue and red, respectively. The five main groups of sAgs are indicated (30, 31). The unrooted tree was based on the alignment of amino acid sequences using ClustalW (47) and constructed using MEGA 4 (45). The sAg abbreviations are indicated, followed by the relevant accession numbers. Percentages from 1,000 bootstraps supporting a given partitioning are indicated on the branches.

FIG. 2.

Functional activities of recombinant SzeF, SzeN, and SzeP in vitro. (A) Dose response. Equine PBMCs (2 × 10⁵) were cultured for 4 days in triplicate with the indicated concentrations of sAg and incubated with ³HT for 16 h before measurement of proliferation. (B) Kinetics of proliferation. Equine PBMCs (2 × 10⁵) were cultured with 0.125 μg/ml sAg in triplicate for 24 to 96 h. ³HT was added for the last 16 h of each period. The results are presented as the stimulation index. (C and D) Kinetics of TNF-α synthesis in equine PBMC culture supernatants (C) and IFN-γ synthesis after overnight culture (D) after stimulation with recombinant sAgs (0.125 μg/ml). Recombinant SeeH and SeeL (33) were used as negative and positive controls, respectively. IFN-γ synthesis was detected by flow cytometry after intracellular staining. The dashed line represents the threshold above which the response was considered positive (>2). The error bars represent standard deviations from the mean.

FIG. 3.

ClonalFrame analysis of MLST alleles of 26 S. equi subsp. equi and 165 S. equi subsp. zooepidemicus isolates and its relationship to the prevalence of sAg genes and mitogenic activity. Shown is a majority rules 50% consensus tree generated from 6 independent runs of ClonalFrame (16), each with 250,000 iterations, and imported into MEGA 4 (45). The genes examined were seeL, seeM, seeH, seeI, szeF, szeN, szeP, and gyrA. Mitogenic assays determined the abilities of different isolates to induce proliferation of equine PBMCs. The number of isolates representing each ST is indicated. STs in which all isolates contained the gene or possessed functional activity are shown in red, STs in which all isolates lacked the gene or functionality are shown in blue, and STs in which not all isolates contained the gene or functionality are colored yellow. The positions of S. equi subsp. equi strain 4047 and S. equi subsp. zooepidemicus strains H70, MGCS10565, and BHS5 are indicated.

FIG. 4.

Neighbor-joining tree showing phylogenetic relationships of szeF in 51 S. equi subsp. zooepidemicus isolates that screened positive. The unrooted tree was based on the alignment of nucleotide sequences using ClustalW (47) and constructed using MEGA 4 (45). Percentages from 1,000 bootstraps supporting a given partitioning are indicated on the branches. C indicates a predicted full-length SzeF. T, T2, and T3 (dotted boxes) indicate predicted truncated SzeF amino acid sequences (113, 22, and 154 amino acids long, respectively). The four groups of szeF (A, B, C, and D) are indicated.