|
|
||||||||
1 Department of Pediatrics and the World Health Organization Collaborating Center for Reference and Research on Streptococci, University of Minnesota, Minneapolis, MN 55455, USA
2 Centers for Disease Control and Prevention, Division of Bacterial and Mycotic Diseases, Respiratory Diseases Branch, 1600 Clifton Rd, Mailstop CO2, Atlanta, GA 30333, USA
Correspondence
Edward L. Kaplan
kapla001{at}umn.edu
Received 1 July 2005
Accepted 23 September 2005
Abbreviations: CDC, Centers for Disease Control and Prevention; GAS, group A streptococci; SOF, serum opacity factor; UMN, University of Minnesota.
| INTRODUCTION |
|---|
|
|
|---|
Successful studies depend on accurate subspecies characterization of GAS. Demonstration that M protein is an important GAS virulence factor and that antibody to this protein confers protection to the human host led to the acceptance of M typing as the preferred method for GAS characterization throughout most of the twentieth century.
The M-typing process is enhanced by inclusion of two other GAS-characterization methods: T-protein serotyping and detection of streptococcal serum opacity factor (SOF). However, even with utilization of T typing and SOF determination, scarcity of M antisera has prevented the identification of many GAS strains. This problem was more completely addressed in the mid 1990s by the development of emm-gene sequence typing, a system that provides a practical alternative to serotyping to identify M-protein type.
T serotyping and SOF determination add valuable parameters for strain identification. For example, it is well known that strains may share the same M/emm type, but may differ in clonal type. This is especially true when strains are recovered from diverse locations (Beall et al., 2000). Marked differences in T serotypes between strains sharing the same emm sequence type or M serotype almost invariably correlate to markedly different clonal types (Beall et al., 2000), and such differences could have potentially important pathogenetic significance. The most recent comprehensive review of the correlation of T pattern and SOF production with M type was published in 1993 (Johnson & Kaplan, 1993). Since then, the number of M/emm types has increased by more than 50 %, but important T-type/SOF correlation information has not been widely available. This report presents an updated and comprehensive correlation of T type and SOF/sof with M/emm type, based on analysis of more than 40 000 GAS isolates obtained worldwide during the past 50 years.
| METHODS |
|---|
|
|
|---|
T serotyping. T-agglutination serotyping was performed at the UMN following published protocols (Moody et al., 1965). Prior to 1990, the T-typing antisera used were obtained either from the Central Public Health Laboratories in London or from the CDC in Atlanta. Since 1990, T-typing antisera have been obtained from Sevapharma in Prague, and T-typing antisera used at the CDC were prepared there.
M serotyping. Streptococcal extracts for serogroup determination and for M typing were prepared at the UMN by slight modifications of the classical method of Lancefield (Johnson et al., 1996; Moody et al., 1965). Prior to 1972, M-type precipitin reactions were done by the glass capillary tube method (Swift et al., 1943); subsequent to that date, all M-typing reactions were performed by the Ouchterlony double-diffusion technique (Rotta et al., 1971). M-type-specific antisera used were either prepared in the UMN laboratory using standard methods (Johnson et al., 1996) or obtained from the CDC. Data presented from the CDC in this report were based entirely on emm typing; M typing was not performed during the period of time in which the 7500 isolates were tested.
Streptococcal SOF detection. Prior to 1988, detection of SOF was performed at the UMN by the agar plate method (Maxted et al., 1973). Since 1988, SOF detection has been performed in 96-well microplates, and the results are read spectrophotometrically (Johnson & Kaplan, 1988). Culture supernatants, acid extracts and SDS extracts were used as the source of SOF (Rehder et al., 1995). SOF detection was done at the CDC by the agar plate method using culture supernatants (Maxted et al., 1973).
sof-gene detection and sequencing. Routine sof gene-detection and sequencing performed at the CDC was based upon PCR and sequence analysis of a variable-length 450650-base PCR fragment using methods and primers described previously (Beall et al., 2000). Detection of the sof gene and sof sequencing were not done at the UMN during the time period of this report.
emm-sequence typing. emm-sequence typing has been in use at the CDC since 1995. More than 7500 emm-typing results obtained at the CDC are included in this report. Typing by emm sequencing was initiated at the UMN in 1997 and used in conjunction with M serotyping for characterization of GAS. Approximately 1400 of the 35 000 UMN results are based on emm typing. Preparation of GAS lysates, emm-gene PCR and emm sequencing were performed as previously described (Beall et al., 1996), with modifications at http://www.cdc.gov/ncidod/biotech/strep/protocols.htm.
Collection of data. The T-pattern associations presented in this report are based on analysis of more than 42 000 strains (approx. 7500 CDC results and 35 000 UMN results). However, T-pattern frequency estimates were determined from the analysis of approximately 21 000 results that were available in easily quantifiable computerized databases; this subset included nearly 14 000 UMN results obtained since 1986 and more than 7500 isolates characterized at the CDC since 1995.
| RESULTS AND DISCUSSION |
|---|
|
|
|---|
T-pattern and SOF/sof information can be a valuable asset to efficient and cost-effective determination of emm-sequence type. In situations requiring large-scale emm-type identification of epidemiologically related GAS isolates, the sequencing materials and labour costs can be high, and the time required to generate results can be quite long. To improve sequencing efficiency and economy, restriction digests of emm amplicons from all strains to be characterized can be subjected to gel electrophoresis and the resulting enzyme restriction patterns compared. Strains with common T patterns, SOF production and emm-restriction patterns can then be grouped together, and emm sequencing can be performed on a small sample. The emm type of the group can then be deduced based on the result of the sequenced subset. Among epidemiologically associated isolates, common T pattern, SOF production and emm-gene restriction-enzyme patterns have been shown to quite accurately predict a mutually shared emm type (Beall et al., 1998; Espinosa et al., 2003; see also http://www.cdc.gov/ncidod/biotech/strep/protocols.htm).
T-pattern and SOF-production information is also useful for validating the accuracy of emm-typing results. The correlations shown in Table 1
, based on the analysis of large numbers of isolates from varied geographical areas and collected over a 50-year period of time, likely represent a majority of T-pattern/SOF/M-type/emm-type combinations expected to be encountered during routine studies. Results that deviate significantly from those in Table 1
, especially SOF/sofemm discordance, should be confirmed.
|
SOF production (and sof-gene detection) also provide useful additional strain information. If one performs phylogenetic analysis of aligned segments of all known deduced emm-gene products, these segments segregate almost entirely into sof-gene-positive and sof-gene-negative clusters (Beall et al., 2000; Facklam et al., 2002; McGregor et al., 2004; Teixeira et al., 2001; Whatmore et al., 1994). Also, sof-positive strains have a distinct pattern of genes in the neighbourhood of emm compared to that of sof-negative strains (McGregor et al., 2004). Several studies have also suggested that SOF production has a high correlation with M-protein class; strains which do not produce SOF tend to possess class I M proteins, and SOF-producing strains carry class II M proteins (Bessen et al., 1989; Bessen & Fischetti, 1990). It is becoming increasingly clear that the presence or absence of SOF/sof indicates fundamental differences in M-protein structures, and also indicates fundamental differences in mga-locus gene composition and arrangement. Both of these differences are likely to have profound, although as-yet undefined, biologic significance.
Finally, T-pattern and SOF information about a GAS strain, especially when combined with knowledge of M or emm type, provides an important link to information from studies published over the many decades of basic clinical and epidemiological GAS research, when these serological methods were the primary tools available for strain classification.
Factors affecting T-type interpretation
Several factors can cause difficulties in the interpretation of T-serotyping results. Many isolates carry combinations of T proteins, resulting in T-agglutination patterns. Several of these patterns are reproducible and widely recognized (e.g. 3/13/B3264, 5/27/44, 8/25/Impetigo 19). However, many GAS isolates react non-specifically with T antisera, leading to agglutination with many, or even all, sera. Unless these cross-reactions are removed by trypsinization, incorrect T-agglutination patterns will be obtained. Unfortunately, excessive trypsinization may remove true T-protein reactions and affect the final T pattern. For example, a strain with the very common 3/13/B3264 agglutination pattern may lose some, or even all, of the reactions making up that pattern depending on the intensity of trypsinization. Therefore, when comparing individual results with those in Table 1
or when interpreting T-typing results obtained by different persons, in different laboratories, or at different times, it is important to be aware that different combinations of common patterns may, in fact, represent the same type.
Factors affecting SOF determination
Determination of SOF production by group A streptococcal strains has been a widely used and valuable tool for strain characterization. SOF production occurs in approximately half of all known M/emm types, and this production correlates highly with specific M type (Top & Wannamaker, 1968) and emm type (Beall et al., 1996, 2000; Whatmore et al., 1994).
The sensitivity of tests used to determine SOF production varies considerably. It is most common to test for SOF in ToddHewitt broth-culture supernatants, but SOF can also be detected in Lancefield HCl extracts and in SDS extracts of streptococcal cells. The highest sensitivity is achieved by using several methods collectively, but, of the single tests, use of SDS extracts appears to give the lowest percentage of false-negative results (Rehder et al., 1995). The probability of a false-negative result is increased if only a single source of SOF is used.
The identification of the gene encoding SOF, the sof gene, and the subsequent development of tests to detect this gene by PCR, have added another tool for GAS characterization (Beall et al., 2000; Rakonjac et al., 1995). There is a nearly complete correlation of sof-gene presence with SOF production, making this molecular test a practical alternative to classical phenotypic SOF determination (see Table 1
). Of interest is the observation that several SOF-negative isolates tested have been found to carry a sof gene (Beall et al., 2000; Rakonjac et al., 1995). Among SOF-negative strains analysed in the present report, only those of M/emm type 12 were found to have a sof gene, predicted to be truncated and not cell wall associated due to the presence of a frame-shift mutation (Jeng et al., 2003).
SOF inhibition serotyping as a surrogate for M or emm typing
Streptococci of different M/emm types produce immunologically distinct forms of SOF, and the resulting serum opacity reaction can be inhibited by anti-SOF antibodies in a type-specific manner (Top & Wannamaker, 1968). This type-specific inhibition of SOF was developed into a typing system with very high correlation to M type (Maxted et al., 1973). Many laboratories have used SOF-inhibition typing as a surrogate for M typing when appropriate M antisera were not available. However, recent studies have demonstrated that multiple sof-gene types do occur within a single M/emm type, and that a single sof sequence can be associated with more than one emm type (Beall et al., 2000). Therefore, although SOF-inhibition serotyping remains a useful GAS classification tool, it must be interpreted carefully, with the understanding that in some instances the SOF-inhibition type will not accurately predict M/emm type, especially when strains from diverse sources are analysed.
Recognition of new emm types
The development of emm typing has made GAS M-type identification accessible to virtually all laboratories with DNA-sequencing capability. Further, compared to the laborious serological process required for recognition of new M types (Facklam et al., 2002), recognition and differentiation of possible new emm types can be relatively simple. With many laboratories worldwide involved in emm typing, a need has arisen to establish criteria for and a process by which new emm-sequence types can be proposed and validated in an orderly and systematic manner. This was addressed in 1997 and 1999 by a Working Group composed of representatives of six international streptococcal reference centres (Facklam et al., 1999, 2000, 2002). This Working Group established guidelines for the recognition of new emm types, and they also agreed that the official, web-searchable database of emm-gene sequences would be maintained at the CDC (http://www.cdc.gov/ncidod/biotech/strep/emmtypes.htm).
Results and observations
The dramatic impact of emm typing becomes apparent when one considers that it took more than 50 years from the designation of M type 1 by Griffith in 1926 to the designation of M-81 in 1976 (Facklam & Edwards, 1979; Griffith, 1926). By contrast, only 6 years after demonstrating in 1996 that emm typing could be used as a reliable substitute for M typing, the number of formally recognized M/emm types rose to 124, an increase of more than 50 % (Facklam et al., 2002). A comprehensive list of T-type and SOF-production correlations with M type was published in 1993, several years prior to the introduction and widespread use of emm typing and the dramatic increase in recognized types (Johnson & Kaplan, 1993). Although limited information about the T types and SOF production associated with these new emm types can be found in scattered publications and on the internet, there has been no published, comprehensive source of this information. This report provides an updated compilation of current M and emm types, together with information about their corresponding T patterns and SOF production, as extracted from the records of two streptococcal reference laboratories and based on typing results of more than 40 000 GAS strains collected worldwide during the past 50 years.
Table 1
lists M and/or emm types, SOF production, sof-gene amplification results, the number of isolates analysed, and the T-agglutination patterns observed. Clarifying comments, when necessary, are included in the T-agglutination patterns' column or in the legend to Table 1
. Note that the numbers given in the No. of isolates' column represent approximately 21 000 results available in computerized databases maintained at the UMN since 1986 and at the CDC since 1995, and are only a subset of all results available. Therefore, the relative M-/emm-type frequencies may be biased toward more recent years. Also, because isolates sent to these two laboratories include outbreaks or surveys conducted in specific geographic areas, the frequency distribution may be influenced by selection bias.
Note that for completeness, M types 7, 10, 16, 20, 21, 35 and 45 are listed in Table 1
, even though they do not exist; either the reference strains for these types are not GAS or the M type is identical to another officially designated M type (Rotta, 1978).
Several instances in which different, previously designated M types were found to share identical or near-identical 5' emm sequences have been described (Beall et al., 1996, 2000; Facklam & Beall, 1997; Whatmore et al., 1994). M-type pairs with shared emm types described in Table 1
are Lancefield's M-27 and M-77 (note that Griffith's type 27, designated 27G in the table, has a unique emm sequence), M-38 and M-40, M-44 and M-61, and M-50 and M-62. M types 65 and 69 have nearly identical emm sequences. Two of these pairs, 27L/77 and 44/61, share identical emm-sequence types. In addition, limited experimentation indicates that representative reference strains of these pairs and clinical isolates of corresponding T types test positively with sera prepared against both members of the pairs (Beall et al., 2000). M types 38 and 40 have been reported to have identical corresponding 5' emm sequences, but yet react type-specifically with M-38 and M-40 antisera (Beall et al., 1996). The explanation for this possible lack of concordance between emm sequence and M specificity is unknown at this time.
Conclusions
The development of emm typing has brought a powerful tool to the study of GAS. It allows characterization of essentially all GAS, with results that can be directly related to the extensive history of GAS research when M typing was the primary tool available, and the relative ease of characterization has resulted in a dramatic increase in the number of identified M and emm types. Even with this powerful molecular tool, T-agglutination type and SOF-production information continue to be important and relevant, providing the potential for improved efficiency, validation of results, and increased discrimination of clonal differences within emm types. However, dissemination of T-pattern and SOF-production information has not kept pace with developments resulting from emm-sequencing technology. The T-pattern/SOF correlations provided in this report, covering not only the large number of new types established as a result of emm-typing technology but also updated information about classical types, will provide an important resource for epidemiologists and laboratory scientists as they seek to further unravel the still existing mysteries about the pathogenesis of group A streptococcal infections and their suppurative and non-suppurative sequelae.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. Menon, C. Lloyd, B. Malathy, V. Sakota, D. Jackson, and B. Beall emm type diversity of {beta}-haemolytic streptococci recovered in Chennai, India J. Med. Microbiol., April 1, 2008; 57(4): 540 - 542. [Full Text] [PDF] |
||||
![]() |
R. L. Woodbury, K. A. Klammer, Y. Xiong, T. Bailiff, A. Glennen, J. M. Bartkus, R. Lynfield, C. Van Beneden, B. W. Beall, and for the Active Bacterial Core Surveillance Team Plasmid-Borne erm(T) from Invasive, Macrolide-Resistant Streptococcus pyogenes Strains Antimicrob. Agents Chemother., March 1, 2008; 52(3): 1140 - 1143. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Luca-Harari, K. Ekelund, M. van der Linden, M. Staum-Kaltoft, A. M. Hammerum, and A. Jasir Clinical and Epidemiological Aspects of Invasive Streptococcus pyogenes Infections in Denmark during 2003 and 2004 J. Clin. Microbiol., January 1, 2008; 46(1): 79 - 86. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Bidet, C. Courroux, C. Salgueiro, A. Carol, P. Mariani-Kurkdjian, S. Bonacorsi, and E. Bingen Molecular Epidemiology of the sil Streptococcal Invasive Locus in Group A Streptococci Causing Invasive Infections in French Children J. Clin. Microbiol., June 1, 2007; 45(6): 2002 - 2004. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Perez-Trallero, M. Montes, B. Orden, E. Tamayo, J. M. Garcia-Arenzana, and J. M. Marimon Phenotypic and Genotypic Characterization of Streptococcus pyogenes Isolates Displaying the MLSB Phenotype of Macrolide Resistance in Spain, 1999 to 2005 Antimicrob. Agents Chemother., April 1, 2007; 51(4): 1228 - 1233. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Neal, B. Beall, K. Ekelund, B. Henriques-Normark, A. Jasir, D. Johnson, E. Kaplan, M. Lovgren, R. R. Reinert, Members of the Strep-EURO Study Group and Internat, et al. International Quality Assurance Study for Characterization of Streptococcus pyogenes J. Clin. Microbiol., April 1, 2007; 45(4): 1175 - 1179. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Lizano, F. Luo, and D. E. Bessen Role of Streptococcal T Antigens in Superficial Skin Infection J. Bacteriol., February 15, 2007; 189(4): 1426 - 1434. [Abstract] [Full Text] [PDF] |
||||
![]() |
Z. Kratovac, A. Manoharan, F. Luo, S. Lizano, and D. E. Bessen Population Genetics and Linkage Analysis of Loci within the FCT Region of Streptococcus pyogenes J. Bacteriol., February 15, 2007; 189(4): 1299 - 1310. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Siljander, M. Toropainen, A. Muotiala, N. P. Hoe, J. M. Musser, and J. Vuopio-Varkila emm typing of invasive T28 group A streptococci, 1995-2004, Finland. J. Med. Microbiol., December 1, 2006; 55(Pt 12): 1701 - 1706. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |