Group A streptococcus (GAS) strains differ widely in their encapsulation degree, and those with an exuberant capsule production have a mucoid morphology when are cultured on blood agar plates. The expression of the capsule has been related to the virulence of the strain since it was first recognized [1, 2]. In the United States temporal association between a high incidence of rheumatic fever cases and an increased presence of mucoid S. pyogenes strains was observed [3]. More concretely, the rheumatogenic ability has been associated to a limited number of emm types such as emm1, 3, 5, 6, or 18 which often were mucoid [1, 3] although this pattern is not held in all parts of the world, like in Hawaii [4]. Furthermore, these emm types, and particularly emm1 and emm3 types, are the most prevalent emm types found among specimens causing invasive GAS disease worldwide [5, 6]. By performing blood agar cultures for S. pyogenes detection, phenotypic characteristics like mucoid colony morphology can be recognized. In fact, recently the advisability of detecting mucoid isolates in clinical practice has been emphasized in the literature [1].

In 2009, a sudden emergence of encapsulated S. pyogenes isolates was observed among clinical samples received at the microbiology laboratory of Hospital Donostia, (Gipuzkoa), in Northern Spain, where circulation of mucoid S. pyogenes strains was highly infrequent in the past. The aim of this work was to describe molecular characteristics of mucoid isolates causing this outbreak and to compare them to a sample of non-mucoid S. pyogenes isolates obtained in the same period of time.

In 2009, 1571 S. pyogenes isolates were detected among samples received at the microbiology laboratory of Hospital Donostia. Of them, 132 were mucoid, representing the 8.4% of all isolates of that year, while in the previous four years only 9 mucoid isolates were detected. Circulation of mucoid isolates through 2009 is represented in figure 1.

Crater et al. [9] demonstrated that capsule production is greatest in the early logarithmic growth phase in broth cultures and that its expression is lost when the stationary growth phase is reached. In our experience, mucoid colonies showed a wet and glossy shape in blood agar plates (figure 2), although few of them were already dried showing a colapsed, rough and "matte" morphology different to that found in non-mucoid colonies. Incubation of blood agar plates in an atmosphere enriched with 5% of CO₂ enhanced capsule expression. All mucoid isolates were susceptible to erythromycin, clindamycin, tetracycline and levofloxacin antimicrobials showing the following MICs: erythromycin MIC < 0.5 μg/mL; clindamycin MIC < 0.5 μg/mL; tetracycline MIC < 4 μg/mL; and levofloxacin MIC < 2 μg/mL.

Overall, 98.5% of all mucoid isolates (130/132) belonged to the emm3.1 type. The two isolates with a different emm type were characterized as emm5.46 and emm1.0. In a high percentage of mucoid isolates the emm gene was difficult to amplify. Probably the vast amount of hyaluronic acid could have interfered in DNA extraction as after following specific CDC recommendations for mucoid strains, a complete emm gene amplification success was obtained. PFGE revealed 4 different patterns: patterns A and B, with a similarity of 73% between them for emm3 mucoid isolates (figure 3), pattern C for the emm1 isolate and pattern D for the emm5 isolate. All emm3 mucoid isolates, independently of the PFGE pattern, belonged to the ST15, the emm1 isolate to the ST28 and the emm5 isolate to the ST99. All emm3 isolates were non T-typeable. This finding was unrelated to the mucoid capsule because among 89 non-mucoid emm3 isolates obtained in the last 4 years, only 1 could be T-typed as T3 (data not shown). The circulation of a mucoid S. pyogenes clone (emm3.1/ST15) in 2009 contrasts with the results obtained in the past, when only a few mucoid isolates representing a varied sample of emm and sequence types were detected in the community.

Twenty four out of 144 non-mucoid isolates analyzed in 2009, belonged to the emm3 type, which shared the same molecular and susceptibility characteristics (emm3.1/ST15) as the emm3 mucoid isolates, although belonged only to the PFGE pattern A. The remaining 120 non-emm3 non-mucoid isolates showed percentages of non-susceptibility in variable proportions (19.2% to erythromycin, 5.8% to clindamycin, 5.8% to tetracycline and 1.7% to levofloxacin). The only difference between emm3 non-mucoid and mucoid isolates was the degree of hyaluronate encapsulation.

Among mucoid isolates, 87.9% (n = 116) were obtained from throat samples. Age distribution of patients affected by the mucoid clone was similar to that among patients infected by non-mucoid S. pyogenes isolates: 84.1% (n = 111) and 80.2% (n = 1154) were children under 14 years old respectively (p = 0.30). The main clinical manifestations were pharyngitis (n = 55), scarlet fever (n = 30) and otitis (n = 13), similar to those found among non-mucoid isolates (n = 1439), although they also caused 4 cases of invasive disease. The emm3 type has been ranked among the most prevalent emm-types isolated from patients with severe disease worldwide and has been related to high case fatality rates [5, 6]. In this study four emm3 mucoid isolates were obtained from sterile body sites (3 from blood and 1 from pleural fluid), two of them associated with streptococcal toxic shock syndrome, 1 with arthritis, and 1 with celullitis with a fatal outcome. Among non-mucoid isolates (n = 1439), only 3 cases of invasive disease were registered (one emm3.1, one emm75, and one emm87). These data revealed that invasive disease was more often caused due to mucoid isolates (3%) compared to non-mucoid isolates (0.21%, p = 0.001) in 2009 in our hospital. Independently of the mucoid feature, the emm3 type was detected in 71.4% of all invasive GAS disease of that year.

We have found few reports investigating mucoid S. pyogenes colony phenotype in Spain. Bosch et al. [10] reported 11 out of 31 invasive S. pyogenes isolates obtained between 1995 and 1998 showing a mucoid appearance, although only three were heavily mucoid. Sierra et al. [11] investigated the presence of mucoid isolates among invasive GAS infections in injecting drug users in Barcelona, finding that all isolates were non-mucoid. By contrast, in other parts of the world, like in the United States, mucoid GAS circulation has been more frequently reported. Veasy et al. [3] found an association between mucoid strains circulation and an increase of acute rheumathic fever incidence. Rheumatic fever is considered a rare disease in Europe [12] and no cases were produced in our region in the last ten years. Nevertheless, after the characterization of this clone which merges a traditionally virulent serotype (emm3) and the mucoid feature, pediatricians and other physicians were alerted in order to prevent the re-emergence of rheumatic fever into an area previously free of this disease. We have investigated the epidemiology of S. pyogenes in our community since the 90's, and to our knowledge emm3 S. pyogenes isolates circulate every year but without showing a mucoid appearance (data not shown). Why these emm3 S. pyogenes isolates express the mucoid phenotype is under research.

To our knowledge, this is the first outbreak of mucoid S. pyogenes isolates described in the community in southern Europe, and reminds us of the importance of performing a comprehensive phenotypic and genotypic characterization of circulating strains. It is of special concern that the same isolate merged two important virulence factors: the emm-type, emm3, and the hiperproduction of capsule observed as a mucoid morphology, which could have enhanced the ability to cause severe disease.