| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Institute of Laboratory Medicine, Section of Medical Microbiology, Lund University, Lund, Sweden
2 National Institute of Research and Development for Microbiology and Immunology, Cantacuzino, Bucharest, Romania
3 German National Reference Center for Streptococci, Department of Medical Microbiology, University Hospital RWTH Aachen, Aachen, Germany
4 Department of Clinical Microbiology and Immunology, Lund University Hospital (USIL), Lund, Sweden
Correspondence
Aftab Jasir
aftab.jasir{at}med.lu.se
Received March 11, 2008
Accepted July 4, 2008
In 2002, the Romanian National Reference Laboratory was invited to join the Strep-EURO project to study invasive Streptococcus pyogenes infections. During 2003 and 2004, a total of 33 isolates recovered from invasive disease were received from eight Romanian counties. For comparison, 102 isolates from non-invasive disease, as well as a collection of 12 old invasive strains (isolated between 1967 and 1980) were included. All isolates were characterized by several methods: T and emm typing, presence of the fibronectin-binding protein F1 gene (prtF1), serum opacity factor (sof), and superantigen (SAg) genes (speA, speB, speC, speF, speG, speH, ssa and smeZ). The recent invasive isolates exhibited 19 emm-types, of which emm1, emm81, emm76, emm49 and emm78 covered 57 % of the strains. Furthermore, multilocus sequence typing analysis revealed nine new sequence types, corresponding to emm types 1, 12, 49, 81, 92, 100, 106 and 119. The non-invasive isolates comprised 24 different emm types with a predominance of emm1 and 12; the old invasive strains were of eight emm types, of which four were unique for this group. All isolates harboured speB and speF; smeZ was detected in all invasive strains, except for the emm49 and emm81 isolates. The majority of isolates from carriers, and patients with pharyngitis were prtF1 positive, most of these (14 strains) being emm12. High tetracycline resistance rates were noted among both invasive and control isolates (54 % and 35 %, respectively), whereas macrolide resistance rates were low (3 % and 5 %, respectively). Active and continuing surveillance is required to provide an accurate assessment of the disease burden and to provide epidemiological data on the character of isolates in Romania.
Abbreviations: CSF, cerebrospinal fluid; GAS, group A streptoccocci; iGAS, invasive group A streptococci; MLST, multilocus sequence typing; NF, necrotizing fasciitis; PFGE, pulsed-field gel electrophoresis; RFLP, restriction fragment length polymorphism; SAgs, superantigens; SOF, serum opacity factor; ST, sequence type; STSS, streptococcal toxic shock syndrome; VT, vir type.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSReferences |
|---|
The heterogeneity exhibited by the N terminus of the M-protein, the major GAS virulence factor (Bisno et al., 2003), is used in the M-serotyping technique (Lancefield, 1962). The emm gene (encoding the M-protein) is successfully targeted in sequence typing (Johnson et al., 2006). Another GAS surface protein, T-protein, was used as a basis for a crude, but widely used subdivision of strains (Moody et al., 1965). The variable presence of an apoproteinase serum opacity factor (SOF) enabled separation of GAS strains into SOF-positive or -negative. Several other molecular methods such as pulsed-field gel electrophoresis (PFGE) and multilocus sequence typing (MLST) were used to characterize GAS isolates epidemiologically (Bruneau et al., 1994; Enright et al., 2001). Finally, vir typing is also used to type GAS; this is based on the structural and sequence heterogeneity of the vir regulon (renamed the mga regulon) encoding some key virulence factors, including the M-protein family (Caparon & Scott, 1987; Gardiner et al., 1995).
Many extracellular virulence factors are also produced by GAS (Bisno et al., 2003), including several superantigens (SAgs), such as streptococcal pyrogenic exotoxins. These molecules have the capacity to activate a high number of T-cells unspecifically, leading to the release of excessive amounts of inflammatory cytokines and serious systemic effects (Bisno et al., 2003; Sriskandan et al., 2007).
GAS are known often to persist in the host after antibiotic treatment, potentially giving rise to recurrent infections or disseminating into tissues (Kaplan et al., 2006). This feature is linked to the ability of the bacteria to enter and persist within eukaryotic cells; one factor involved in this is fibronectin-binding protein (PrtF1/SfbI) (Neeman et al., 1998; Schwarz-Linek et al., 2006). The possible genetic linkage between this protein and macrolide resistance remains the subject of discussion (Facinelli et al., 2001).
Knowledge regarding epidemiology of invasive GAS (iGAS) disease in Romania is very limited. As part of the Strep-EURO initiative, the Romanian National Reference Laboratory for Streptococcal Infections started to collect isolates and clinical data from invasive infections during 2003–2004. The aim of the present national report was to compare microbiological and epidemiological characteristics of iGAS isolates with those of non-invasive GAS isolates obtained during the same period.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSReferences |
|---|
A detailed questionnaire, requesting data regarding demographic characteristics, clinical manifestation, predisposing factors, treatment and outcome after 7 days was distributed to clinicians. Information regarding antibiotic susceptibility to penicillin, erythromycin, clindamycin, tetracycline, chloramphenicol, ofloxacin, ciprofloxacin, spiramycin, vancomycin and quinupristin/dalfopristin was also requested.
In addition, a collection of 12 invasive GAS isolates recovered from 1967 to 1980 was included. The strains were from blood (six bacteraemia and one septic abortion isolate), peritoneum (three isolates) and CSF (one isolate), and four were from other sterile sites. Data on clinical presentation, risk factors and outcome for the corresponding cases were not available.
T- and emm-typing. Isolates were T-typed using commercial poly- and mono-specific T-antisera, according to the manufacturer's recommendation (Sevapharma) (Moody et al., 1965). For emm-typing, the N-terminal hypervariable portion of the emm gene was sequenced, as described previously (Beall et al., 1996). DNA sequence alignment and emm type assignment was performed using BLAST (http://www.cdc.gov/ncidod/biotech/strep/strepblast.htm) and sequences listed in the streptococcal emm sequence database at the Center for Disease Control and Prevention (ftp://ftp.cdc.gov/pub/infectious_diseases/biotech/tsemm).
vir typing. vir regulon typing was performed on invasive strains using a previously described protocol, with a forward primer that anneals upstream of the regulatory gene virR and a reverse primer that anneals in the scpA gene (Surdeanu et al., 2000). PCR products were digested with HaeIII and HinfI (Boehringer Mannheim and GIBCO). Restriction fragment length polymorphism (RFLP) profiles were examined using the Taxotron software package (Institut Pasteur, Paris, France); patterns were clustered by the single-linkage method with a fixed tolerance of 3.5 %. A Dice coefficient of 1 corresponds to 100 % similarity of RFLP profiles.
PFGE. All invasive isolates, as well as T1 and T12 non-invasive strains (determined using T-typing), were subjected to PFGE. Cultures in Brain Heart Infusion broth (Merck) were incubated for 18 h at 37 °C. DNA was extracted using techniques previously described (Stanley et al., 1995). Chromosomal DNA was digested with 40 U SmaI (Promega) and fragments were separated in a CHEF Mapper XA System (Bio-Rad) by electrophoresis at 200 V for 23 h at 13.5 °C with an initial pulse time of 10 s and a final pulse time of 35 s. Results were analysed using the Fingerprinting II software version 3.0 (Bio-Rad). Values for position tolerance and optimization were 1 % and 0.5 %, respectively.
MLST. All recent invasive strains (n=33) were typed by MLST as described previously (Enright et al., 2001); alleles and sequence types (STs) were identified using the MLST database (http://spyogenes.mlst.net/).
Toxin gene detection. The isolates were tested for the presence of the speA, speB, speC, speF, speG, speH and ssa genes by PCR using primers described by Luca-Harari et al. (2008). All seven pairs of primers were used in an initial multiplex PCR, using the following conditions: denaturing for 2 min at 95 °C, 35 cycles of 1 min at 94 °C, 1 min at 50 °C and 1.5 min at 72 °C, followed by a final elongation step for 7 min at 72 °C. In addition, presence of the gene of streptococcal mitogenic exotoxin, smeZ, was traced by a single PCR, using primers covering its known allelic variation (Darenberg et al., 2007).
sof and prtF1 gene detection. The sof gene was detected using methods and primers described previously (Beall et al., 2000), yielding amplicons of 560–700 bp. These were further digested with DdeI (Promega) and restriction fragments were separated on a 4 % agarose gel (MetaPhor, FMC Bio Products). Amplification of prtF1 was performed using published primers (Neeman et al., 1998), resulting in amplicons of 100–600 bp which were visualized on a 3 % agarose gel (MetaPhor, FMC Bio Products).
Antimicrobial susceptibility testing. In vitro susceptibility to antibiotics was tested by disk diffusion on 5 % blood agar, following the instructions provided by the Clinical and Laboratory Standards Institute (CLSI). Minimal inhibitory concentrations were determined by broth microdilution, on Mueller–Hinton agar plates supplemented with 5 % (v/v) defibrinated sheep blood according to the guidelines of the CLSI (2005). Erythromycin-resistant strains were assigned the constitutive (cMLS), inducible (iMLS) or efflux-mediated (M) macrolide-resistance phenotype on the basis of the double-disk (erythromycin and clindamycin) test, as described by Seppala et al. (1993).
Antibiotic resistance gene detection. Multiplex PCRs were used for detection of tetracycline and macrolide resistance genes. Tetracycline-resistant isolates were tested for the presence of the ribosome protection genes tetM and tetO, using previously described primers (Olsvik et al., 1995; Roberts et al., 1993). Macrolide-resistant isolates were tested for the ermA, ermB and mefA genes, using published primers (Nielsen et al., 2004; Sutcliffe et al., 1996).
|
RESULTS AND DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSReferences |
|---|
The age of the patients with invasive GAS infections varied between 2 months and 83 years. However, two age groups prevailed: 0–9 years (n=11, 33 %) and 40–49 years (n=9, 27 %). There was no gender preponderance. The age of carriers and patients with non-invasive GAS varied from 2 months to 69 years, with a male to female ratio of 1.2 : 1.
Clinical presentations in patients with invasive infections included cellulitis (n=10), NF (n=4), bacteraemia without focal symptoms (n=3) and osteomyelitis (n=2); the remaining 14 patients had arthritis, meningitis, bronchopneumonia, peritonitis, cervical adenitis, erysipelas or abscess. In total, nine STSS cases were reported (three of which were fatal); eight of these were adults and one was a 2-month-old baby with septic infection of the umbilical cord stump.
Underlying conditions were reported for 21 of the total 33 invasive cases (64 %); these included skin lesions (11 patients, 5 of whom developed STSS), immunosuppressive therapy (3 cases), hospital-acquired infection (one patient who also developed STSS), drug abuse (one case) and chickenpox (one case). In four cases, the patients had underlying conditions other than those listed in the questionnaire (malignancy, alcoholism, traumatism of superior right limb and malnutrition).
All fatal cases (n=3) occurred in patients with STSS, giving a case fatality rate (CFR) of 33.3 among this group.
Distribution of T and emm types
A large number of T and emm types was found (27 and 34, respectively) (Table 1
). Predominant T types among the invasive strains were T12, T1 and T14, and among non-invasive isolates T1 and T12 were predominant, with these two types covering 54 % of the control isolates.
|
Among non-invasive isolates, 24 emm types were detected; emm1 accounted for 25 % of the cases, followed by emm12 (23 %). Other common types were emm75, 76, 81 and 95, representing around 20 % of the isolates. The old invasive strains exhibited eight emm types, of which emm5, emm23, emm50/62 and emm77 were exclusively found among these isolates.
Toxin gene detection
Overall, 20 patterns of SAg genes were found (Table 2). All tested strains harboured the chromosomally encoded genes speB and speF. Similarly, smeZ was detected in all invasive strains, except for the emm49 and emm81 isolates. The other tested genes were found at different frequencies, ranging from 0 to 4 per isolate. In particular, speA was detected in 20 % of the invasive isolates and in 38 % of controls. This gene was common in emm1 strains (five of seven invasive isolates and all controls). In contrast, speC, present in 57 % of recent invasive and 45 % of control isolates, was frequently detected among emm12 strains (77 %). In addition, all emm12 strains were found to harbour speH, whereas only approximately 10 % were ssa positive.
|
sof and prtF1 gene detection typing
In total, 21 of 33 new invasive strains (64 %), 5 old invasive (42 %) and 56 non-invasive (59 %) strains were positive for the sof gene. In general, the restriction profiles of DdeI amplicons correlated with the emm type, though with some exceptions. Thus, one emm81 isolate, distinct from the other emm81 strains, exhibited a profile identical to that of an emm84 strain. The two emm92 and the single emm22 strain showed identical profiles (Fig. 1a), whereas the single emm25 strain had the same restriction profile as three of the emm75 isolates (Fig. 1b).
|
). Eighty per cent of control strains, of 18 different emm types, were prtF1 positive, whereas those of eight other emm types were negative. The majority of isolates from patients with pharyngitis and from throat carriers were prtF1-positive (Table 3), most of these (14 strains) being emm12. This might account for the ability of these isolates to adhere to and internalize into pharyngeal epithelial cells (Hanski et al., 1992; Molinari et al., 1997) and is supported by a recent study which demonstrated that PrtF1/SfbI was seldom encoded by isolates causing invasive disease but was encoded by 35 % of those from pharyngotonsillitis (Bianco et al., 2006). In the present study, many of the invasive isolates, as well as superficial isolates, were of type T12; the presence of prtF1 may thus be primarily correlated with the type rather than disease localization.
|
). Each of three other HaeIII VTs were divided into two HinfI types (VT119 and VT64; VT81 and VT69; VT6a and VT6b). Generally, there was a better agreement between PFGE, emm types and VTs determined using combined HinfI and HaeIII, rather than only HaeIII RFLP of the vir amplicon. Since the restriction profiles appeared to be emm-type specific, VTs were assigned corresponding to emm types (Fig. 2). However, as mentioned above, in two cases, the same VT was obtained for isolates of different emm-types (VT9/49 was common to one emm9 and four emm49 isolates, and VT25/75 was shared by one emm25 and two emm75 strains). The region targeted in vir-typing contains some important virulence genes, among them emm, scpA and sic (Akesson et al., 1996; Caparon & Scott, 1987; Gardiner et al., 1995). The technique, though not widely used, was shown to be discriminatory and, with few exceptions, provides a close correlation to M/emm typing (Gardiner et al., 1998).
|
, Fig. 2). Thus, one emm9 and four emm81 isolates showed pattern C, whereas both emm25 and emm75 isolates grouped under pattern D. Interestingly, recent isolates of emm1 and emm49 showed different PFGE patterns compared with old isolates; however, they were still about 52 and 83 % related, respectively, within the same general clade. The two recent emm95 isolates were unrelated, grouping under different PFGE patterns. The invasive T12 isolates were of three different emm types (emm12, 76 and 81); the two ‘old’ emm12 invasive strains present 65 % similarity whereas the recent emm12 invasive isolate (Rou6) is more closely related to the old invasive isolate 629 (similarity 84 %); the last two types grouped accordingly into two PFGE patterns. All T1 control strains were emm1. They presented PFGE pattern A, with the exception of two strains that presented pattern A1 and one remaining strain that presented pattern A2. Pattern A1 lacked a single band of pattern A. Pattern A2 had three bands that were different from pattern A. According to the criteria for interpreting DNA restriction patterns produced by PFGE, the three patterns shared by all contemporary T1 emm1 isolates, irrespective of their origin, were closely related. Control isolates of T12 were emm12, but four isolates were emm76. PFGE patterns of T12 emm12 isolates belonged to the pattern G and closely related subtypes G1–G4 (clustered at 86 % similarity); a single control strain belonged to subtype G5 (clustered at 76 % similarity with G and G1–G4). Five T12 emm12 strains were SmaI resistant, as they had macrolide resistance M phenotype. The use of isoschizomer XmaI failed to give a digestion profile. The T12 emm76 isolates presented PFGE pattern B, as did the majority of emm76 invasive isolates, with the exception of a single isolate which presented a pattern that differed by two bands from pattern B, designated pattern B2 (data not shown).
|
MLST
MLST is the most recently developed typing method, which allows distinctions among and within GAS populations. Compared to emm-typing, this method has the ability to index variation that has accumulated relatively slowly (McGregor et al., 2004). Notably, despite the small number of invasive isolates (33), 23 different STs were detected; of those, 5 new allelic variants of housekeeping genes (Table 4), as well as 9 new sequence types (ST 434–442) were identified. As expected, the same emm type yielded similar STs; thus, four emm1 isolates were of ST28, whereas the remaining one was a single-locus variant of ST28 (recently designated ST440). Similarly, out of four emm81 isolates, three showed ST341 and the remaining one was a single locus variant, designated ST434. Though STs are often emm type-restricted, there are studies showing that some particular STs can occur in more than one emm type (McGregor et al., 2004). However, since the numbers of isolates with the same emm type were limited in our study, it is difficult to draw conclusions regarding the discriminatory power of MLST within a single emm type.
Among different methods used for characterization of S. pyogenes, T- and emm-typing together will provide a good, quick and relatively inexpensive pool for the identification of isolates. However, for clonal identification and relationship analysis, PFGE and MLST will be required. While PFGE can reveal differences in total DNA, MLST targets specific genes with allelic variation. As stated above, vir typing might offer additional information but is not recommended for routine epidemiology. Since the number of isolates in our study was limited, any conclusion on the role of prtFI as an epidemiological marker will require further study.
Antimicrobial susceptibility
All invasive isolates were fully susceptible to penicillin, clindamycin, chloramphenicol, ofloxacin, ciprofloxacin, spiramycin, vancomycin and quinupristin/dalfopristin. Only one invasive and five non-invasive isolates were macrolide-resistant (3 % and 5 %, respectively). The invasive isolate showed the iMLS phenotype (and harboured ermA), whereas four out of five non-invasive strains were of M phenotype (mefA) and one showed cMLS resistance (ermB). These low macrolide resistance rates were similar to those found in Scandinavia (Littauer et al., 2006; Luca-Harari et al., 2008), but different from those in other European countries (Bingen et al., 2004; Creti et al., 2007).
In contrast, tetracycline resistance was noted among 20 (54 %) invasive and 36 (35 %) control isolates, comprising 24 different emm types. The determinant tetM was detected in 83 % and 97 % of resistant invasive and control strains, respectively. None of the five emm1 invasive strains and only one out of 30 emm1 non-invasive strains was resistant. A low percentage (10 %) of tetracycline-resistance was also found among emm12 non-invasive isolates. Since tetracycline is not the drug of choice in the treatment of GAS infections, the high rate of resistance among Romanian isolates is not readily explainable. However, the extensive use of tetracycline in the treatment of a variety of human and veterinary infections might have contributed to the dissemination of resistance among GAS isolates.
Since changes in the epidemiology of GAS infections, in particular the increase of severe cases, are noted worldwide, active surveillance is of high priority in the process of improving prophylactic measures and treatment strategies. However, the small number of isolates collected in the present study precludes an exact picture of the actual situation of invasive GAS disease in Romania. The limitations of surveillance (e.g. low clinical awareness, lack of network of laboratories and insufficient reporting strategies) need to be addressed in order to obtain a real image of the national incidence of these diseases. Epidemiological studies analysing the distribution of circulating emm types have clearly become increasingly important for the development of multivalent M-protein vaccines. This is one of few reports from Eastern Europe that combines clinical information and microbiological findings of invasive and non-invasive GAS infections.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSReferences |
|---|
|
References |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONACKNOWLEDGEMENTSReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
