OUP user menu

Superantigen-like gene(s) in human pathogenic Streptococcus dysgalactiae, subsp. equisimilis: genomic localisation of the gene encoding streptococcal pyrogenic exotoxin G (speGdys)

Svea Sachse, Peter Seidel, Dieter Gerlach, Elisabeth Günther, Jürgen Rödel, Eberhard Straube, Karl-Hermann Schmidt
DOI: http://dx.doi.org/10.1111/j.1574-695X.2002.tb00618.x 159-167 First published online: 1 October 2002


Streptococcus pyogenes (GAS) causes about 90% of streptococcal human infections while group C (GCS) and G (GGS) streptococci can be pathogenic for different mammalians. Especially the human pathogenic GCS and GGS, Streptococcus dysgalactiae, subsp. equisimilis, account for 5–8% of the human streptococcal diseases like wound infections, otitis media, purulent pharyngitis and also streptococcal toxic shock syndrome. A defined superantigen so far was not identified in GCS and GGS strains. In the present investigation we screened DNA of GCS and GGS human isolates for the presence of genes for streptococcal pyrogenic exotoxins (spe) by hybridisation with probes that stand for the GAS genes speA, speC, speZ (smeZ), speH, speG, speI, speJ and ssa. In many GCS and GGS strains we found positive reactions with the probes speG, speJ and ssa, but not with the probes for the remaining genes under investigation. PCR amplification with subsequent sequence analysis of the PCR fragments revealed only the presence of the gene speG in GCS and GGS strains, while no DNA fragments specific for speJ and ssa could be amplified. Additionally, the upstream and downstream regions flanking speG in GGS strain 39072 were sequenced. Remarkable differences were found in the neighbourhood of speG between GAS and GGS sequences. Downstream of speG we identified in strain GGS 39072 two new open reading frames encoding proteins with no similarity to protein sequences accessible in the databases so far. In the compared GAS strains SF370 and MGAS8232, this segment, apart from some small fragments, had been deleted. Our analysis suggests that a gene transfer from GGS to GAS has preceded following deletion of the two genes orf1 and orf2 in GAS.

  • Superantigen
  • Streptococcal pyrogenic exotoxin G
  • Genomic organisation
  • Gene transfer
  • Streptococcus
  • Streptococcus dysgalactiae subsp. equisimilis

1 Introduction

Streptococcus pyogenes (GAS) causes about 90% of streptococcal human infections while group C (GCS) and G (GGS) streptococci can be pathogenic for different mammalians with species dependent preferences to distinct hosts. At humans GCS and GGS can be considered as commensals with exception of pathogenic variants like Streptococcus dysgalactiae, subsp. equisimilis. Those GCS and GGS account for 5–8% of human streptococcal diseases and were found in wound infections, otitis media and purulent pharyngitis [5,8,10,26]. A few case reports were published recently, describing the connection of GCS and GGS with streptococcal toxic shock syndrome [1416,20,27].

Although GCS and GGS have been isolated from patients with streptococcal toxic shock syndrome, the existence of superantigens in such strains could so far not be verified. Until now the production of superantigens in human pathogenic streptococci was only reported for GAS.

In this report, we analysed the DNA of GCS as well as GGS strains looking at the presence of genes encoding different streptococcal pyrogenic exotoxins (SPE) known from GAS. Hybridisation experiments were performed on immobilised chromosomal DNA of 24 GCS/GGS strains with spe probes produced by PCR from GAS template DNA. We searched for the existence of genes or gene fragments with similarity to speA, speC, speZ (smeZ), speH, speG, speI, speJ and ssa.

2 Materials and methods

2.1 Bacterial strains, plasmids, and culture conditions

In this study we investigated human clinical isolates of S. dysgalactiae subsp. equisimilis, of which 8 strains belonged to GCS and 16 strains to GGS. The patients suffered from suppurating lesions, otitis media and septic shock. These strains and the GCS strain H46A [7], the GAS strains M1 40/58, M5 Manfredo, NY5 and C203S came from the strain collection of the Streptococcal Laboratory of the University Jena, Jena, Germany. The strains were stored at −70°C in Todd–Hewitt broth supplemented with 20% calf serum and 20% glycerol or in lyophilised form. For isolation of total DNA streptococci were cultured at 37°C overnight in Todd–Hewitt broth. Escherichia coli strains JM109 [21] and EZ competent cells (Qiagen GmbH, Hilden, Germany) were used for subcloning experiments. The pGEM®-T Easy vector system (Promega Corporation, Madison, WI, USA) and p-Drive vector kit (Qiagen GmbH, Hilden, Germany) were used for T-cloning and following sequencing of PCR products.

2.2 DNA techniques

Isolation of plasmid DNA from E. coli, DNA restriction, ligation and related experiments were performed according to standard techniques [21] and as described by the manufacturers of cloning kits. Chromosomal streptococcal DNA was isolated according to Caparon and Scott [6] and purified by using the Invisorb Spin DNA Micro kit III (Invitek GmbH, Berlin-Buch, Germany). Briefly, streptococci were grown in 10-ml overnight cultures at 37°C. The bacteria were centrifuged and the sediments lysed for 3 h in 100 µl 0.05 M Tris–HCl buffer, pH 8.0, containing 10 µg ml−1 Mutanolysin (Sigma/Aldrich). Following, 1 mg proteinase K in 50 µl of lysis buffer was added and the lysate further incubated for 30 min at 37°C and for 15 min at 70°C. The mixture was centrifuged and 200 µl binding buffer, purchased with the kit, was added to the separated supernatant. The crude DNA then was applied to a spin column and finally purified according to the protocol of the manufacturers. The DNA concentration finally was adjusted to 1 µg µl−1.

2.3 Hybridisation

For dot hybridisation of chromosomal DNA, the denatured DNA was transferred to non-charged nylon membrane (Millipore Corporation) by using a vacuum dot-blot apparatus. Briefly, 5 µg DNA was mixed with 50 µl denaturing buffer (0.5 M NaOH, 1.5 M NaCl) and the solution was applied to the nylon membrane under vacuum. The membrane then was equilibrated with 6×SSC buffer (20×SSC: 0.3 M sodium citrate, 3 M NaCl, pH 7.0). Blocking and hybridisation were performed according to the ECL direct nucleic acid labelling and detection system from Amersham Life Science. The probes for all superantigen genes were synthesised by PCR from template DNA of GAS used in this study. Primers were designed on the basis of published GAS DNA sequences available in the NCBI database (Table 1). The PCR fragments were confirmed by sequencing (see below). Luminescence labelling, hybridisation and detection of the probes were performed with the ECL detection kit.

View this table:
Table 1

Primer sequences used to amplify the spe probes by PCR from DNA of different GAS strains

No.GenePrimer sequence
  • The primers for speC, speG, speH, speI, speJ and speZ (smeZ) were designed from the appropriate segments of the Streptococcal Genome Sequencing Project, strain SF370 ([9], accession number AE004092). Primers speA and ssa were deduced from the appropriate sequences with the accession numbers U40453 and U48792, respectively.

2.4 DNA sequencing

The PCR fragments from GAS DNA used as probes were directly sequenced with the appropriate PCR primers (Table 1). Before sequencing the PCR fragments amplified from GCS and GGS DNA, they were subsequently cloned into the vectors pGEM-T Easy or p-Drive. Progressive DNA sequencing to analyse the flanking region of the speG gene was performed with the DNA of the selected GCS strain 39072. Templates were produced by inverse PCR technique as described recently [12,25]. In the present case, the chromosomal DNA was digested with Eco RV and the fragments were ligated to circularise the DNA. The mixture was used as template for inverse PCR by using primers Q and R (Table 1) deduced from the speG sequence of GGS strain 39072 established in this study. Sequencing was performed using the BigDye Terminator sequencing kit and the Genetic Analyser ABI Prism 310 (Applied Biosystems).

2.5 Database searches and sequence analysis

The EMBL and NCBI databases were accessed and searched by computer. Nucleic acid and protein sequences were analysed by the programs provided in the database home pages.

Nucleotide sequences reported in this paper appear in the above named databases under the accession numbers AJ294849 and AJ489606.

3 Results

3.1 Hybridisation of GCS and GGS DNA with spe probes derived from GAS DNA

The DNA of 24 GCS and GGS strains as well as of four GAS control strains was immobilised to nylon membranes in dot spots and analysed with the gene probes for speA, speC, speZ/smeZ, speG, speJ, ssa, speH and speI synthesised by PCR from DNA of GAS strains (Fig. 1). The probes speA, speC, speZ, speH and speI did not react with the DNA spots of any of the investigated GCS/GGS strains. These probes only stained the DNA of those GAS control strains which carry the appropriate gene (Fig. 1). However, probes speG, spej and ssa hybridised with DNA spots of several GCS and GGS strains. PCR was run to prove whether the chromosome of positive GCS/GGS strains could carry genes corresponding to speG, speJ and ssa. The primers in these experiments were the same as used for amplification of probes from GAS DNA (Table 1). In Fig. 2 the PCR results are shown for two relevant GGS strains and one GAS control strain. PCR fragments of the same size were only obtained with the primer pair A and B with specificity for the GAS gene speG, below called speGGAS. These primers amplified with GAS DNA as well as with GCS and GGS DNA fragments at 0.7 kb, the size of speGGAS with 705 bp. The primer pairs for the genes ssa and speJ, however, did not amplify an appropriate PCR fragment with GGS DNA but with GAS DNA the control band could be amplified (Fig. 2). Moreover, following PCR experiments with additional primers deduced from inner sequence regions of the GAS genes speJ and ssa generated some PCR fragments with GGS DNA templates, but those were not specific for speJ or ssa (not shown). Sequencing of such PCR fragments revealed that the primers amplified other sections of the GCS/GGS chromosome rather than those related to ssa or speJ. These results indicate that the sequence regions in GCS/GGS strains which hybridised with probes ssa or speJ have partially been deleted or were considerably different to the primer binding sites of the corresponding segments in GAS.

Figure 1

Dot-blot hybridisation of the GAS probes speA, speC, smeZ/speZ, speG, speJ, ssa, speH and speI with chromosomal DNAs of GCS/GGS strains and control GAS strains immobilised in dots on nylon membranes. Only the probes speG, ssa, and speJ hybridised with the DNA of individual GCS and GGS strains.

Figure 2

PCR amplification of speG, ssa and speJ from template DNAs of GAS control strain 40/58 and two of the GGS strains positive in the appropriate hybridisation experiment. The primer pairs specific for the three GAS genes from Table 1 were used.

3.2 Sequence analysis of speGdys from selected GGS strains

The PCR fragment of one selected GGS strain, strain 39072, amplified with the primers A and B specific for speGGAS, was further analysed. The purified band was ligated into the vector pGEM®-T Easy and transformed in E. coli strain JM109. The insert was sequenced from both ends starting with the primers T7 and SP6 located at this vector. The sequence revealed an open reading frame encoding a protein which we named SPEGdys. The deduced amino acid sequence showed homology to SPEGGAS (Fig. 3). The amino acid sequences of SPEGdys of GGS strain 39072 show significant differences to the SPEGGAS protein of GAS strain SF370 mainly in the middle region between the positions 117 and 149 (Fig. 3).

Figure 3

Alignment of the encoding amino acid sequence of SPEGdys (accession number AJ294849) of GGS strain 39072 with SPEGGAS of GAS strain SF370 ([9], accession number AE006489). Signal sequences (SS) of both proteins were theoretically calculated by the program SignalP of Nielsen et al. [17].

3.3 Sequence analysis of the genomic regions flanking speGdys

In extended experiments the genomic regions flanking the speGdys gene at the 5′ and 3′ ends were sequenced. The speGdys has an Eco RV site almost at the end of the gene. The chromosomal DNA of GGS strain 39072 was digested with Eco RV and inverse PCR was performed according to Triglia et al. [25]. The primers Q and R (Table 1) were used to analyse the 5′ flanking region (Fig. 4). The 3′ flanking region was sequenced after PCR amplification with primers S and T (Fig. 4). These primers were derived from the genomic sequence of GAS strain SF370 [9]. Both PCR fragments were ligated into the vector p-Drive and transformed in E. coli strain EZ. The plasmids carrying the inserts were analysed by walking sequencing. The whole sequence of 3930 bp obtained from GGS strain 39072 (accession number AJ489606) was compared with the genomic region of the GAS strains SF370 [9] and MGAS8232 [24] carrying speGGAS. Fig. 4 shows that all three strains differ to each other in the upstream and downstream regions flanking speG. Upstream of speGdys in GGS strain 39072 the intergenic region and the following open reading frame encoding a hypothetical protein showed 75% and 66% homology, respectively, to the same segment in GAS strain SF370. The comparison of the same region between the GAS strains SF370 and MGAS8232 revealed a more different situation, because strain MGAS8232 carries an additional open reading frame between the genes hyp. protein and speGGAS, encoding a transposase. In the region downstream of speG the difference between the GGS and GAS strains became more pronounced. While here both analysed GAS strains were almost identical, the GGS strain 39072, however, carried two additional open reading frames in its chromosome between the two genes speGdys and pgi (Fig. 4). The gene pgi encodes a glucose-6-phosphate isomerase found in many bacteria [9]. Database analysis of orf1 and orf2 in strain 39072 revealed no significant alignment with known sequences, therefore no function of the encoding proteins could be established so far. The intergenic sequences in this region were different and show similarity only in small segments to GAS. This can be impressive illustrated by a gene deletion in orf1. Smoot et al. [24] found downstream from speGGAS in GAS strain MGAS8232 an open reading frame encoding a small hypothetical peptide of 39 amino acids which is also present in GAS strain SF370. We call this gene here orf1t which means truncated orf1. The alignment of the DNA sequences of orf1 of GGS strain 39072 (618 bp) with orf1t (120 bp) of GAS leads to the following result. (i) The first 21 bp of orf1t in GAS were identical with those of orf1 in GGS. (ii) The remaining 96 bp of orf1t showed about 80% identity to the end of the gene orf1 (illustrated by left arrow in Fig. 4). (iii) In conclusion, orf1t represents a truncated gene which has to be descended from orf1. So it can be speculated that the region downstream from speGGAS in GAS came from GGS via horizontal gene transfer. After the transfer process from GGS to GAS the region including orf1 and orf2 was deleted in GAS with remains of orf1t but complete loss of orf2 (Fig. 4). Considering additional segments, the intergenic region upstream from pgi was found to be completely different between the GGS strain 39072 and the GAS strains except a short part of 89 bp immediately before the start of pgi (Fig. 4, small black rectangle in front of pgi) which also supports a preceding gene transfer.

Figure 4

Illustration of the genomic region of superantigen SPEGdys in GGS strain 39072 (below) compared with the gene arrangements in the GAS strains SF370 [9] and MGAS8232 [24]. Abbreviations: hyp. protein — gene encoding a hypothetical protein [9]; trpase — gene encoding a transposase [24]; orf1t — gene encoding a hypothetical peptide representing a truncated orf1. The sequenced region of GGS strain 39072 is marked off with *–*. The small horizontal arrows → indicate primers used to start sequencing speGdys flanking regions. Broken line arrows ↖ indicate gene transfer direction.

The presence of the new identified genes orf1 and orf2 seems to be common in human pathogenic GCS and GGS strains carrying speGdys. PCR analysis of additional GCS and GGS strains with primers T and S revealed that those strains which were positive for speG delivered a PCR fragment. With the speGdys carrying GCS and GGS strains, a 2.5-kb PCR fragment could be amplified while with the speGdys negative strains no band was obtained. A 1.0-kb fragment was obtained with the GAS control strain M1 40/58 (Fig. 5A). The comparison of the different molecular sizes and partial sequencing of the fragments showed that those GCS and GGS strains which possess the gene speGdys subsequently in 3′ direction carry the genes orf1 and orf2. The shorter 1.0-kb fragment from the GAS strain 40/58 (Fig. 5A) lacked both genes just like the two GAS strains M1 SF370 and MGAS8232 of the sequencing projects [9,24]. A second PCR with the same strains used in Fig. 5A was performed after replacement of primer S with a primer starting in front of orf1 but keeping reverse primer T (Fig. 5B). Sequence analysis of the PCR fragments at 1.6 kb confirmed that the GCS and GGS strains, which were analysed in this study, possessed the two genes orf1 and orf2 regardless whether they were positive for speGdys or not. The very small fragment which was amplified with the DNA of control strain M1 40/58 did not show specificity for the intergenic region of GAS between speGGAS and pgi (Fig. 5B).

Figure 5

A: PCR analysis of the genomic region between speG and pgi in selected GCS/GGS strains positive and negative for speG in hybridisation experiments from Fig. 1. The bands were amplified with primers S (starting in speG), and T (reverse out of pgi). The PCR bands at 2.5 kb appear with speG positive strains and represent the region speGdys-orf1-orf2-pgi. With the speG negative strains no PCR fragment was obtained. The GAS control strain A 40/58 lacks the two genes seen at the smaller band at 1000 kb. B: PCR analysis with the strains from (A) with the modification that primer S was replaced by the following primer, 5′-CGTATTATGTTGAGATTTTG-fw, which binds shortly in front of the start codon of orf1 (compare Fig. 4), and primer T was kept. The fragment at 1.6 kb indicates that orf1 and orf2 are common in the investigated GCS and GGS strains, regardless whether they possess speG or not.

4 Discussion

In this study we used DNA hybridisation to screen chromosomal DNAs of several human pathogenic GCS/GGS strains for the presence of genes with similarity to superantigen genes in GAS. We could not identify genes related to the superantigens SPEA, SPEC, SPEZ (SMEZ), SPEH, and SPEI. However, with DNA probes of the genes speG, ssa, and speJ we found a positive reaction with DNA from several GCS and GGS strains. But the presence of the genes ssa and speJ in the analysed GCS/GGS strains could not be verified by PCR. With unequivocal evidence, however, we identified a gene encoding a SPEG-like protein (SPEGdys) in 15 of the 24 investigated GCS and GGS strains. Recombinant SPEGGAS from GAS has been recently characterised and shown to be a potent superantigen [19]. Until now, only few reports described mitogenicity of culture filtrates [3] or isolated protein G [18] of group G streptococci. However, the mitogenic component derived from S. dysgalactiae strains at that time could not be defined. Our finding of a gene encoding the amino acid sequence similar to SPEG in human pathogenic group C and G streptococci shows that those strains have the potential to produce superantigens and would explain their isolation from patients with sepsis and toxic shock.

In further experiments we analysed the DNA region flanking speGdys in the GGS strain 39072. Our sequence data revealed the existence of a genomic segment which shows partial similarity to, but also clear differences from, the speGGAS region in GAS. In the upstream region of speG the organisation up to the first flanking gene is nearly identical in strains GGS 39072 and GAS SF370. In both strains, an intergenic segment and a gene encoding a strongly related hypothetical protein are linked. The sequence comparison of the genomic segment upstream of speG revealed homologies of 87% (speG), 75% (intergenic region) up to 66% (hypothetical protein) between the strains GGS 39072 and GAS SF370. In this region upstream of speG the GAS strain MGAS8232 is more different from the two other strains analysed, because it carries an additional gene encoding a transposase inserted between the genes speGGAS and hypothetical protein[9,24]. Differences were found in the region flanking speG at the 3′ end which impressively illustrates horizontal gene transfer between GGS and GAS. In strain GGS 39072 we identified two open reading frames orf1 and orf2 between the genes speGdys and pgi. The sequences of orf1 and orf2 did not align with sequences found in the databases and encode proteins with unknown function. Our PCR experiments showed that these two genes seem to be present in the majority of GCS and GGS strains independently whether they carry speGdys or not. A detailed analysis revealed that, in contrast to the GCS and GGS strains, in the GAS strains the genes orf1 and orf2 were deleted. While the gene orf2 was completely removed, the orf1 was truncated in GAS to a small open reading frame (orf1t) encoding a short peptide of 39 amino acids with unknown function. It is interesting that the complete protein derived from orf1 of the GGS strain shows homology to the first seven amino acids of the N-terminus and the remaining 32 amino acids of the C-terminal end of the short GAS peptide. The middle part of orf1 from GGS was deleted in GAS resulting in the gene fragment orf1t. Therefore we speculate that the gene transfer of genes downstream of the genomic region of speG took place from GGS to GAS followed by partial deletion of orf1 and complete removal of orf2 in GAS.

Kalia et al. [13] postulated gene transfer in the opposite direction from GAS to GCS and GGS mainly for housekeeping genes. A recent study in our group showed gene transfer between GCS/GGS and GAS in the region of the mga regulon. The following evolution resulted in low changes in the emm genes but in considerable differences of the intergenic regions and also in the genes mga of GAS and, the mga-like gene, mgc of GCS/GGS [12].

From the summary of our and the results of other investigators it can be assumed that gene transfer will take place in both directions, because often GAS and GCS/GGS can be isolated at the same time from the same human host.

Recently, Anzai and coworkers demonstrated mitogenic activity in culture filtrates of Streptococcus equi in horse white blood cells [1]. Now, the presence of superantigen genes, namely the genes speH and speI, first identified in the chromosome of GAS strain M1 SF370 [9,19] was described in horse pathogenic group C streptococci, S. equi[2]. The DNA sequences of the both adjacent genes were nearly identical in GAS and S. equi, which also demonstrates gene transfer between different host specific streptococcal species. These two genes we did not find in the human pathogenic GCS and GGS strains investigated in the present study.

As already mentioned S. pyogenes (GAS) mainly causes human infections while group C and G streptococci can be pathogenic for different mammalians with species dependent preferences to distinct hosts. On humans GCS and GGS mostly can be considered as commensals with exception of pathogenic variants. Virulence factors, found in GAS, however, were also described in different GCS and GGS strains. It has been shown that human isolates can express M proteins [4,8,11,23] and streptolysin O [28]. We show here that human pathogenic GCS and GGS strains can carry superantigen genes. This implicates that GCS and GGS have potential tools like GAS to resist host defence mechanisms like M protein and streptolysin O [11,22,23,28] and may be potential producers of some superantigen. Expression studies of SPEG in such strains are in progress.


We acknowledge the excellent technical assistance of Mrs. Roswitha John and Ms. R. Wagner for reviewing the manuscript. The study was supported by Grant B307–01022 of the Interdisciplinary centre (IZKF) of the University of Jena and the Thüringer Forschungsministerium TMWFK.

group A, C or G streptococci
gene encoding streptococcal pyrogenic exotoxin or superantigen G (SPEG) of GGS
gene encoding streptococcal pyrogenic exotoxin or superantigen G of GAS


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
View Abstract