European Journal of Clinical Microbiology & Infectious Diseases

, Volume 29, Issue 5, pp 585–589

Molecular markers for discriminating Streptococcus pyogenes and S. dysgalactiae subspecies equisimilis

Authors

  • D. J. McMillan
    • Bacterial Pathogenesis LaboratoryThe Queensland Institute of Medical Research (QIMR)
    • Griffith Medical Research CollegeGriffith University and the Queensland Institute of Medical Research (QIMR)
  • T. Vu
    • Bacterial Pathogenesis LaboratoryThe Queensland Institute of Medical Research (QIMR)
  • P. V. Bramhachari
    • Department of Microbiology and Cell BiologyIndian Institute of Science
  • S. Y. Kaul
    • Bacterial Pathogenesis LaboratoryThe Queensland Institute of Medical Research (QIMR)
    • Department of MicrobiologyKEM Hospital
  • A. Bouvet
    • Université Paris Descartes, Service de Microbiologie-Hygiène, Hôtel Dieu AP-HP, Centre National de Référence des Streptocoques LA-SGA-A
  • M. S. Shaila
    • Department of Microbiology and Cell BiologyIndian Institute of Science
  • M. G. Karmarkar
    • Department of MicrobiologyKEM Hospital
    • Bacterial Pathogenesis LaboratoryThe Queensland Institute of Medical Research (QIMR)
Article

DOI: 10.1007/s10096-010-0899-x

Cite this article as:
McMillan, D.J., Vu, T., Bramhachari, P.V. et al. Eur J Clin Microbiol Infect Dis (2010) 29: 585. doi:10.1007/s10096-010-0899-x

Abstract

Given the increasing aetiological importance of Streptococcus dysgalactiae subspecies equisimilis in diseases which are primarily attributed to S. pyogenes, molecular markers are essential to distinguish these species and delineate their epidemiology more precisely. Many clinical microbiology laboratories rely on agglutination reactivity and biochemical tests to distinguish them. These methods have limitations which are particularly exacerbated when isolates with mixed properties are encountered. In order to provide additional distinguishing parameters that could be used to unequivocally discriminate these two common pathogens, we assess here three molecular targets: the speB gene, intergenic region upstream of the scpG gene (IRSG) and virPCR. Of these, the former two respectively gave positive and negative results for S. pyogenes, and negative and positive results for S. dysgalactiae subsp. equisimilis. Thus, a concerted use of these nucleic acid-based methods is particularly helpful in epidemiological surveillance to accurately assess the relative contribution of these species to streptococcal infections and diseases.

Introduction

Streptococcus pyogenes and S. dysgalactiae subspecies equisimilis are closely related Gram-positive bacteria that colonise the skin and respiratory tract of humans. S. pyogenes infection is associated with a wide range of diseases, which include relatively benign and common diseases (such as bacterial pharyngitis, impetigo and scarlet fever), potentially fatal invasive diseases (such as necrotising fasciitis and toxic shock syndrome) and rheumatic heart disease [1]. Together, these diseases are estimated to kill half a million people each year [2]. S. dysgalactiae subsp. equisimilis, which occupies the same tissue sites as S. pyogenes, was previously considered to be a less pathogenic organism and causes infections opportunistically [3]. S. dysgalactiae subsp. equisimilis expresses many of the same virulence factors (including the M-protein, C5a peptidase, streptokinase, fibronectin binding proteins and plasminogen receptors) as those expressed by S. pyogenes [49] and can cause a similar spectrum of diseases in humans as S. pyogenes [3, 10, 11].

In many populations, the S. dysgalactiae subsp. equisimilis colonisation rate in the throat and the rate of its recovery from pharyngitis exceed that of S. pyogenes [12]. The relative contribution by these species to the disease burden could vary considerably between communities and populations [13, 14]. There is a paucity of information on streptococcal disease burden in many resource-poor nations [15]. The lack of definitive and rapid tests to differentiate these species and the perception among clinicians that S. dysgalactiae subsp. equisimilis is generally non-pathogenic may also have contributed to this paucity. Recent studies on invasive disease due to beta-haemolytic streptococci of groups other than A and B concluded distinct clinical manifestations in some populations [1618]. These studies recommended increased awareness of S. dysgalactiae subsp. equisimilis as a human pathogen and emphasise the importance of the identification of beta-haemolytic streptococci at the species level.

It is generally difficult to distinguish S. pyogenes and S. dysgalactiae subsp. equisimilis on a morphological basis, as both species are usually large colony forming with similar haemolysis patterns. These and other beta-haemolytic streptococci are traditionally differentiated on the basis of differences in carbohydrate on the bacterial surface and sensitivity to bacitracin [19]. Additional diagnostics such as the PYR reaction (for L-pyrrolidonyl-β-naphthylamide aminopeptidase) may be used as confirmatory tests, but are not routine procedures in many clinical microbiology laboratories. Whereas the group A carbohydrate is found in S. pyogenes, it has also been found in S. dysgalactiae subsp. equisimilis in rare cases [5, 20]. Positive reactivity to group C or G antigen is given by S. dysgalactiae subsp. equisimilis, S. dysgalactiae subsp. dysgalactiae, S. equi, S. canis and S. anginosus [5, 20, 21]. The bacitracin sensitivity test can sometimes give false-positive results [19]. While the PYR test has excellent specificity and sensitivity in discriminating S. pyogenes from S. dysgalactiae subsp. equisimilis, other bacteria commonly found in the same habitat may also give a positive reaction [22], and the test performed less well in routine diagnostic laboratory settings, sometimes giving weaker reactions [23]. In order to provide additional distinguishing parameters that could be used to discriminate these species, we assess here three molecular targets.

Materials and methods

S. pyogenes and S. dysgalactiae subsp. equisimilis isolates were collected from two different geographical regions, India and Australia. The three isolates with unusual characteristics were from Europe. The bacteria were grown on Todd–Hewitt agar supplemented with 2% horse blood. Group carbohydrates were determined using the Prolex Streptococcal Grouping Latex Kit (Pro-Lab Diagnostics). The PYR test was performed according to the manufacturer’s instructions (Becton Dickinson). DNA was extracted using purification kits (QIAGEN Inc.). The polymerase chain reaction (PCR) conditions were as follows. For speB: denaturation 94°C/30 s, annealing 55°C/30 s, extension 72°C/90 s; forward and reverse primers were GGTTCTGCAGGTAGCTCTCG and TGCCTACAACAGCACTTTGG. For the intergenic region upstream of the scpG gene (IRSG): the conditions were the same as above and the primers were CAACACATAACCACCTTCTGGA and TTGCAAGTGCGTCACAAGAT. The primers and conditions for virPCR were as described previously [24, 25]. Sequence typing (emm typing) was performed as described by Beall et al. [26] and sequencing the gene for rRNA was carried out after amplifying with universal primers [27].

Results and discussions

The rationales for the selection of the targets are as follows. Comparative microarray analysis showed that the gene encoding streptococcal cysteine protease, speB, found ubiquitously in S. pyogenes isolates [28, 29], is absent in S. dysgalactiae subsp. equisimilis [30]. Furthermore, a speB orthologue has not been reported in other bacteria. Hence, the PCR primers targeting this gene are expected to be specific for S. pyogenes. Indeed, a PCR for speB has been suggested as a diagnostic test for the S. pyogenes aetiology of necrotising fasciitis [31]. The remaining two targets are based on the sequence differences in S. pyogenes and S. dysgalactiae subsp. equisimilis upstream of the gene for C5a peptidase (scpA and scpG, respectively). The physical locations of scpA, emm (gene for M protein) and the regulator mga in the S. pyogenes genome are close to each other and are amplifiable by PCR (virPCR; [24, 25]). Depending on the architecture of this chromosomal region, the amplicon arising from virPCR is about 4–7 kb. All S. dysgalactiae subsp. equisimilis isolates possess emm and scpG genes [10, 32]. However, previous studies [7, 33] revealed that the emm and scpG genes in the S. dysgalactiae subsp. equisimilis genome are not physically linked and the sequences immediately upstream of scpG exhibit mosaic structure [34]. However, the virPCR designed for S. pyogenes also gave a product with S. dysgalactiae subsp. equisimilis, but the amplicon was only about 2 kb [3335], suggesting that the primer binding sites may be conserved. We sequenced the virPCR products from diverse S. dysgalactiae subsp. equisimilis strains and found that an open reading frame for a hypothetical protein within the IRSG is highly conserved in all of the strains (Fig. 1) and is absent in S. pyogenes. This allowed us to design S. dysgalactiae subsp. equisimilis-specific IRSG PCR primers.
https://static-content.springer.com/image/art%3A10.1007%2Fs10096-010-0899-x/MediaObjects/10096_2010_899_Fig1_HTML.gif
Fig. 1

Organisation of the intergenic region upstream of scpG (IRSG) of Streptococcus dysgalactiae subsp. equisimilis. The virPCR primers (dashed arrows) and the IRSG primers (solid arrows) are shown. The length virPCR amplifiable region is somewhat variable, but in more than 98% of S. dysgalactiae subsp. equisimilis isolates, it is about 2 kb. Within this region, the segment defined by IRSG PCR primers is highly conserved and targets a hypothetical protein

S. pyogenes (n = 58) and S. dysgalactiae subsp. equisimilis (n = 109) isolates in this study belong to 41 and 36 distinct emm (sub)types, respectively, suggesting that they are highly diverse (Table 1). The collection also represents multiple isolates of the same emm type. We have shown previously that horizontal gene transfers mediated by phages and conjugative transposons between these species are ongoing events and occur with greater frequencies among isolates from highly endemic regions [35]. Therefore, it is all the more necessary to show species-specific distribution of the target genes among the isolates from highly endemic regions. Clonal diversity and two distinct geographical sources of isolates, one of them (India) being highly endemic for streptococcal infection, fulfil the requirement in this regard.
Table 1

Molecular and biochemical tests carried out on 167 S. pyogenes and S. dysgalactiae subsp. equisimilis isolates from Australia and India

Group carbohydrate

Number of isolates

Number of emm types or subtypes

PYR test

IRSG PCR

speB PCR

virPCR size (kb)

∼2

3

4

>4–7

A

58

41

+

+

0

0

1

57

G

98

31

+

98

0

0

0

C

11

5

+

9

1

1

0

All of the strains listed in Table 1 were beta-haemolytic on blood agar plates and possessed group A, C or G carbohydrate. The isolates expressing group A carbohydrate were PYR-positive, and the isolates expressing the group C or G antigen were PYR-negative. All of the isolates presumptively identified as S. pyogenes based on the above reactions were positive for the speB amplicon, whereas none of the S. dysgalactiae subsp. equisimilis isolates were positive for this reaction. The IRSG assay also proved highly discriminatory, as all S. dysgalactiae subsp. equisimilis isolates tested produced the expected product of about 500 bp, and none of the S. pyogenes isolates gave this amplicon. All PCRs were performed with appropriate positive controls and template-negative controls. Thus, speB and IRSG PCRs together unequivocally distinguish S. pyogenes and S. dysgalactiae subsp. equisimilis.

As reported earlier, the virPCR gave products with both the S. pyogenes and S. dysgalactiae subsp. equisimilis isolates in this study (Table 1). As expected, DNA from all of the S. pyogenes isolates gave approximately 4–7-kb product. By contrast, the reaction gave approximately 2-kb products from 107 of the 109 S. dysgalactiae subsp. equisimilis isolates. One S. dysgalactiae subsp. equisimilis isolate yielded a 3-kb product and another a 4-kb product. Both of these isolates expressed group C carbohydrate. Sequence analysis of these variant virPCR products (data not shown) revealed insertions within the mosaic regions upstream of the C5a peptidase gene. However, the region of the hypothetical open reading frame targeted by the IRSG primers is conserved (Fig. 1).

Ongoing cross-species horizontal gene transfers [33, 35, 36] could give rise to isolates with mixed characteristics. In this study, three isolates with mixed characteristics were tested. Two isolates (2005-0193 and 2006-0098) that were S. dysgalactiae subsp. equisimilis as judged by negative reaction to the PYR test but showed agglutination reaction with the group A reagent were also tested by PCRs (Table 2). Both were positive for IRSG PCR and negative for speB PCR. The size of the virPCR amplicons were 2 kb. We further confirmed the identity of these isolates as S. dysgalactiae subsp. equisimilis by 16S rRNA gene sequencing (data not shown). Another isolate (2007-0217) was S. pyogenes as shown by PYR positivity, group A carbohydrate reactivity and rRNA gene sequence. However, 2007-0217 belonged to emm stg1750, a type normally found in S. dysgalactiae subsp. equisimilis, suggesting horizontal transfer of the emm gene. This isolate was speB-positive and IRSG-negative (Table 2). Furthermore, the virPCR product is 7 kb long and revealed physical proximity of the emm locus to the scpA gene (data not shown).
Table 2

Molecular and biochemical tests carried out on three unusual S. pyogenes and S. dysgalactiae subsp. equisimilis isolates

Strain name

Group carbohydrate

emm type

PYR test

IRSG PCR

speB PCR

virPCR (kb)

2005-0193

A

stg6

+

∼2

2006-0098

A

stg6

+

∼2

2007-0217

A

stg1750

+

+

∼7

Recently, a real-time PCR and a low-density microarray were developed to distinguish S. pyogenes and S. dysgalactiae subsp. equisimilis. These assays are based on a common target for both species and another one specific for S. pyogenes [37]. The authors recognised that their diagnostic system does not have a positive test for S. dysgalactiae subsp. equisimilis. By choosing specific targets for each of the species (speB and IRSG PCRs for S. pyogenes and S. dysgalactiae subsp. equisimilis, respectively), we have overcome this limitation. It may be noted that none of the molecular markers thus far described differentiate between group G and C carbohydrate expressing S. dysgalactiae subsp. equisimilis, suggesting that the acquisition of differences in group carbohydrate may have preceded that of molecular markers.

In summary, the three PCRs proposed here offer excellent discriminatory power in identifying common beta-haemolytic streptococcal human pathogens. The tests could be particularly useful in epidemiological surveillance to accurately assess the relative contribution of S. pyogenes and S. dysgalactiae subsp. equisimilis to streptococcal infections and diseases. Current clinical management of infections owing to these species are not different, as both have remained sensitive to penicillin. However, resistance to fluoroquinolone in S. pyogenes has been found to be horizontally acquired from S. dysgalactiae [38, 39]. Horizontal gene transfers have the ability to rapidly change in population structure, which, in turn, may require tailored management in the future. In the light of these observations, we believe that it would be prudent to identify the definitive aetiology of streptococcal infections and diseases.

Acknowledgements

This work was funded through an Australian Government National Health and Medical Research Council (NHMRC) Program Grant, the National and International Research Alliances Program of Queensland, Australia.

Copyright information

© Springer-Verlag 2010