Identification of superantigen genes speM, ssa, and smeZ in invasive strains of beta-hemolytic group C and G streptococci recovered from humans


*Corresponding author. Tel.: +1 (404) 639 1237; Fax: +1 (404) 639 3123, E-mail address:


Group C and G Streptococcus dysgalactiae subspecies equisimilis (GCSE and GGSE) cause a substantial percentage of invasive disease caused by beta-hemolytic streptococci. To determine whether Streptococcus pyogenes superantigen (SAg) genes commonly exist within these organisms, 20 recent invasive GCSE and GGSE human isolates and one group G Streptococcus canis human isolate were tested for the presence of SAg genes speH, speJ, speL, speM, ssa and smeZ by polymerase chain reaction (PCR). Prior to this work, sequence-based evidence of the speM, ssa, and smeZ genes in GCSE, GGSE, and S. canis had not been documented. Eleven of the 21 isolates were PCR-positive for the presence of one to two of the SAgs speM, ssa, or smeZ, with four of these isolates carrying ssa+speM or ssa+smeZ. No isolate was positive for speH, speJ and speL. All six ssa-positive GGSE strains harbored the ssa3 allele, previously only found among S. pyogenes strains. All three smeZ-positive GGSE isolates carried one of two smeZ alleles previously only found within S. pyogenes, however the single S. canis isolate carried a new smeZ allele. All five GCSE and GGSE speM-positive isolates harbored a newly discovered speM allele. The identification of these SAgs within S. dysgalactiae subsp. equisimilis and S. canis with identical or near-identical sequences to their counterparts in S. pyogenes suggests frequent interspecies gene exchange between the three beta-hemolytic streptococcal species.


Group C and G streptococci (GCS and GGS) are commonly regarded as commensals because they are often found in association with the normal flora of human skin, pharynx and the intestine [1,2]. In contrast, the closely related group A streptococcus (Streptococcus pyogenes or GAS) is a human pathogen that commonly causes a wide range of non-invasive and invasive infections. However, there has been an increasing number of reports showing the isolation of GCS and GGS from human infections with clinical manifestations similar to infections caused by GAS and in some instances GCS and GGS have been implicated in severe invasive infections such as necrotizing fasciitis and toxic shock syndrome [2,3–9]. Recent preliminary data from Active Bacterial Core Surveillance (ABCs), part of the Emerging Infections Program Network in the state of Georgia, suggests that group G and C streptococci or Streptococcus dysgalactiae subspecies equisimilis (GGSE and GCSE respectively) constitute more than 60% of invasive GGS and GGC human infections, and account for approximately 5% of the invasive burden due to beta-hemolytic streptococci in this ABCs region (unpublished data).

GCSE and GGSE express several virulence determinants initially identified in GAS. They express a hyaluronic acid capsule, anti-phagocytic M protein, streptolysins O and S, streptokinase, C5a peptidases, and binding proteins for fibronectin, plasminogen, and immunoglobulin [10]. Besides these factors, GAS strains express a large number of distinct, but structurally related streptococcal pyrogenic exotoxins (SPEs) [10–12]. These effectors belong to the family of superantigens (SAgs) [13–15]. These SAgs are capable of binding to the major histocompatibility complex II molecules on antigen presenting cells and T cell receptors [14], leading to the release of large amounts of cytokines. These cytokines have been shown to play major roles in fever, tissue destruction, shock, hypotension, and organ failure associated with invasive streptococcal infections [12,15–18].

Most reported SAg genes are located on integrated phage genomes, a situation that is putatively conducive to genetic transfer among different GAS emm types as well as other streptococcal species [19–22]. Presently, several SAg genes of GAS have been described [11,12,19,23–27]. GAS pyrogenic exotoxin A, C and G determinants have been found in human pathogenic GGSE [20,28], and SAg genes with high homology to speH, speI, and speL of GAS have been identified in group C Streptococcus equi[22,29]. Since SAg genes appear to be randomly distributed among streptococcal isolates and other related species, it appeared likely that other SAg genes are present within GGSE and GCSE. Therefore we screened a set of 20 diverse S. dysgalactiae subsp. equisimilis invasive isolates and a single Streptococcus canis human invasive isolate (the only one available to us from Georgia surveillance at this time) for the presence of six GAS SAgs that have not been previously documented among these two species.

2Materials and methods

2.1Bacterial isolates

A set of 21 consecutive M protein gene (emm) typeable group C and G beta-hemolytic streptococcal isolates (GCS and GGS) were obtained from invasive streptococcal infections in metropolitan Atlanta identified by the Georgia ABCs system from mid 2002 through early 2003. One hundred-twenty GAS isolates (invasive and non-invasive) recovered during the past five years from a variety of human sources and geographic sites (within and outside of the USA) were used to verify that the SAgs could be readily screened among GAS. All isolates used in this study were identified to the species level by previously described phenotypic methods [30].

2.2emm typing

emm typing and emm subtype determination were performed for all isolates as previously described [31,32].

2.3Polymerase chain reaction (PCR) amplification and sequencing of SAg genes

Crude chromosomal DNA extracts were prepared using mutanolysin and hyaluronidase as previously described [31]. Extracts were screened by PCR for SAg genes using primers indicated in Table 1. PCR products were sequenced with an ABI 3100 Genetic Analyzer using the same primers that were used for amplification.

Table 1.  Primers used for the amplification of SAg genes from DNA of group A, C and G streptococcal isolates
  1. Primer sequence pairs were designed from streptococcal genome sequencing projects (accession numbers AE004092, NC002737 and NC003485).

GenePrimer sequence

2.4Flanking sequence of the smeZ gene in S. canis isolate

The 5′ and 3′ ends of the smeZ structural gene and additional sequence flanking smeZ within the single S. canis isolate were obtained by single primer PCR as described previously [33]. Briefly, a denaturation at 94°C for 4 min was followed by 20 cycles of denaturation at 94°C for 30 s, annealing at 50°C for 30 s, and elongation at 72°C for 3 min. This was followed by 30 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s and extension at 72°C for 2 min; then, a final extension step at 72°C for 7 min. Primer 5′-GAC AAT AAA TCT TTC AAT ATA GAT-3′ (anneals to bases 643–666 of smeZ) was used for a single PCR amplification of the smeZ 3′ region and downstream region which was sequenced with primer 5′-GGGCATTTAGATATAGAAATTGACT-3′ (anneals to bases 673–679 of smeZ). The 5′ end of smeZ and upstream region was amplified with primer 5′-AGC ATA TCT AAC ATC AAG TTT C-3′ (anneals to smeZ complement, bases 395–416) and sequenced with primer 5′-AGT CAC TAA GTT ATG ACT GGT-3′ (bases 372–392 of smeZ). The sequence data obtained from single primer PCR-generated template was reconfirmed using sequence obtained from conventional PCR products generated with two primers from strain 4074-03 chromosomal DNA.

2.5Sequence analysis

New nucleotide sequences not found in the GenBank were deposited in NCBI database under the accession numbers: AY344227 (S. canis smeZ25 allele and flanking region), AJ557010 (S. dysgalactiae subsp. equisimilis; speM4 allele), AJ560765 (speM1 allele), AJ560766 (speM2 allele), AJ560767 (speM3 allele) and AJ560764 (speL1 allele).

3Results and discussion

GCSE and GGSE are not commonly regarded as invasive human pathogens. However, the increasing frequency of recovery of these organisms from human invasive disease cases, and the discovery of SAgs speA, speC, and speG in GGSE [20,28] led us to screen for additional SAgs within GGSE and GCSE. Since most known SAg genes appear to be located on mobile elements that can be transferred between different streptococcal species, it suggests that additional SAgs found in GAS may be found in GCS or GGS.

3.1speH, speJ, and speL

All 21 GGS and GCS were PCR-negative for the presence of these speH, speJ, and speL (Table 1) using primers that anneal to structural gene sequences. This indicates that either these genes do not exist within these isolates, or that these isolates lack spe gene sequences highly similar to one or both of the two primers in the primer sets used.


We first wanted to determine whether speM was likely to be widely disseminated among GAS strains. Examination of 120 GAS isolates comprising 28 emm types revealed that 11 isolates representing six types (emm types 4 [one of 10 tested], 6 [one of six tested], 18 [two of two tested], 22 [one of eight tested], 28 [one of five tested], 41 [three of three tested]) were positive for both speL and speM. The speM sequence found within our two type emm18 isolates was identical to the single allele previously described [27]. In addition, we identified three new speM alleles from GAS strains (Fig. 1, alleles 1–3). In all 11 speM-positive GAS isolates speL and speM were closely linked, since flanking primers (to the type M18 open reading frames (ORFs) spyM18_1237 and spyM18_1240) amplified fragments of the appropriate size (Fig. 2) which were shown by nested PCR to contain the speL and speM genes. Smoot et al. [27,34] found speM within all 33 type emm18 isolates tested, but did not find speM within individual isolates of emm types 1–6, 11, 12, 22, 28, 49, 77, and 89. It is not surprising that these authors did not find speM within types emm4, emm22, and emm28 since we found that the majority of isolates within these types were negative for speM. In agreement with these authors, we did not detect speM or speL within emm types 11 (three isolates), 12 (12 isolates), 77 (one isolate), and 89 (four isolates).

Figure 1.

Four alleles of speM gene found during this work. Numbers along the upper portion of the figure indicate the codons; amino acids are indicated with bold letters. All nucleotides different from those in the speM allele (speM18) of GAS serotype M18 strain MGAS8232 [27] are shown. Alleles speM18, speM1 (accession number AJ560765), speM2 (accession number AJ56060766) and speM3 (accession number AJ560767) are from GAS isolates, while speM4 (accession number AJ557010) is from GCSE and GGSE.

Figure 2.

PCR analysis indicating that speL flanks speM in diverse GAS strains but not GCS and GGS study isolates. Primers annealing to the ORFs flanking speL and speM genes in S. pyogenes serotype M18 were used for the amplification. Lanes 1–5, GCS and GGS isolates; lanes 6–16, GAS isolates from different emm types. The emm types of GAS isolates are indicated in parentheses. All GCS and GGS study isolates were additionally negative for speL-specific PCR.

It has previously been shown that speM and speL are located on a prophage-like element in type emm18 GAS [27]. Five of 21 (24%) GCS and GGS isolates (one GCSE and four GGSE) were speM PCR-positive, yielding a 610 bp product (corresponding to bases 50–660 of the GAS 710 base structural gene), however attempts to amplify speL in all 21 GGS and GCS isolates were unsuccessful using either speL-specific primers or phage-specific primers flanking the GAS speL and speM genes. The five speM-positive isolates identified in this study yielded amplicons with an identical partial speM sequence (consisting of eight signal sequence codons and 196 codons encoding mature protein). This speM sequence (allele 4 in Fig. 1) was highly similar to the full-length speM (consisting of 24 signal sequence codons and 216 mature protein codons) from strain MGAS8232 [34] (four base changes: three conservative substitutions (14 Val to Leu, 39 L to I, 141 (R to K) and one synonymous substitution). speM4 from these five S. dysgalactiae subsp. equisimilis isolates was also found to be 97.4% identical to a previously reported speM allele from S. dysgalactiae subsp. dysgalactiae (designated sdm by these authors) [35].


Previous studies demonstrated that ssa is widely spread among S. pyogenes and was found within about 12.5% of 80 distinct clonal lineages [36]. This group [36] additionally found that ssa was absent from 79 GGS and GCS, although the species of these GGS and GCS were not communicated in this report. We found that six of 21 GGS and GCS (all S. dysgalactiae subsp. equisimilis) were PCR-positive for the presence of ssa and harbored a sequence identical to bases 46–769 of the 780 base ssa3 allele that is widely distributed among diverse GAS strains [36].


Of the four smeZ-positive GGS and GCS isolates, two were found to carry smeZ2, one isolate carried smeZ17, and the single S. canis isolate (4074-03) carried the new smeZ25 allele (Table 2). This allele differs from its closest match, smeZ17, by three unique nucleotide substitutions relative to sequence encoding the processed protein (codon 36 synonymous, codon 143 Met to Thr, codon 152 Ser to Asn). The predicted processed forms of SmeZ17 and SmeZ25 demonstrate the rather wide degree of allelic variation inherent to this determinant [37] since they share only about 94% amino acid sequence identity with SmeZ2 and have 10–11 substitutions relative to SmeZ2 distributed between residues 23 and 177 of the 210 amino acid protein.

Table 2. S. pyogenes SAg genes among invasive strains of group C and G Streptococcus recovered from humans
  1. All isolates are S. dysgalactiae, subsp. equisimilis except isolate 4074-03 which is a S. canis. Isolates 4012-03 and 6317-02 are GCS, while the other isolates are GGS.

Isolate no.emm typeSourcespeH, J, LsmeZspeMssa

Since smeZ is known to be a chromosomal determinant in GAS [11,34,37], additional sequence flanking the smeZ25 structural gene was obtained from the single S. canis isolate employing single primer PCR. No ORFs greater than 30 codons were found within the 415 bases downstream of smeZ25. The sequence of bases 1–252 immediately downstream of the smeZ stop codon was 94.4% identical to corresponding non-annotated sequences downstream of smeZ in the types M1, M3, and M18 genomes (see accession numbers AE006622, AE014169, and AE010110 respectively). Bases 253–415 did not show high levels of significant identity to any known sequences. The sequence corresponding to the signal sequence and putative promoter region was highly similar to corresponding sequences from the M1, M3, and M18 genome sequences. The 197 base sequence immediately upstream of smeZ25 was nearly identical to corresponding sequences from the M1, M3, and M18 genomes. Bases 198–306 showed no significant homologies to streptococcal sequences. These flanking sequences are consistent with integration of a 1080 bp fragment of S. pyogenes chromosomal DNA between coordinate base positions −252 and +198 relative to the smeZ structural gene in this S. canis isolate.


Evidence has suggested that S. pyogenes SAgs play key roles in the development of severe invasive diseases such as toxic shock and necrotizing fasciitis [10]. Additionally, the correlation between the recoveries of speL- and speM-positive GAS isolates from acute rheumatic fever (ARF) patients, and the presence of speL and speM antibodies in these patients suggests that these SAgs play role(s) in this disease [27]. Increasing reports of invasive GCS and GGS recovered from humans and reports of the presence of GAS SAg genes in these organisms indicates the possibility that more virulent forms of these organisms have emerged through the acquisition of SAg genes from S. pyogenes. Eleven of the 21 invasive GGS and GCS screened during this study were positive for the presence of one or more of these genes, with four of them positive for the presence of two SAg genes formerly thought to be exclusive to S. pyogenes. These data suggest that the transfer of these genes from S. pyogenes to S. dysgalactiae subsp. equisimilis has occurred frequently.

Previous to this work, sequence-based evidence demonstrating the presence of ssa, smeZ and speM genes in GGSE and GCSE had not been provided. The ssa gene was previously detected by DNA hybridization among human isolates of GCSE and GGSE [28] but attempts to further confirm these results using PCR were unsuccessful. Here we have reported the initial identification of the ssa, smeZ and speM genes of GGSE and GCSE isolated from humans. Additionally, we have demonstrated a new smeZ allele in a human isolate of S. canis, demonstrating the wide dissemination of the GAS smeZ gene within yet another streptococcal species.


E.I.I. was supported by an American Society for Microbiology/National Center for Infectious Disease Postdoctoral fellowship. We thank the Georgia Department of Human Resources, Wendy Baughman and the Active Bacterial Core Surveillance sites/Emerging Infections Program (see, and Microbiology Laboratories in the Metropolitan Atlanta Area for providing the group C and G streptococcal isolates. We thank Varja Sakota, Zhongya Li, and Bob Gertz for laboratory assistance.