INTRODUCTION

Top Previous Next References

Streptococcus pyogenes (group A streptococcus) is an important species of gram-positive extracellular bacterial pathogens. Group A streptococci colonize the throat or skin and are responsible for a number of suppurative infections and nonsuppurative sequelae. As pathogens they have developed complex virulence mechanisms to avoid host defenses. They are the most common cause of bacterial pharyngitis and are the cause of scarlet fever and impetigo. The concept of distinct throat and skin strains arose from decades of epidemiological studies, in which it became evident that there are serotypes of group A streptococci with a strong tendency to cause throat infection, and similarly, there are other serotypes often associated with impetigo (62, 543). In the past, they were a common cause of puerperal sepsis or childbed fever. Today, the group A streptococcus is responsible for streptococcal toxic shock syndrome, and most recently it has gained notoriety as the "flesh-eating" bacterium which invades skin and soft tissues and in severe cases leaves infected tissues or limbs destroyed.

The group A streptococcus has been investigated for its significant role in the development of post-streptococcal infection sequelae, including acute rheumatic fever, acute glomerulonephritis, and reactive arthritis. Acute rheumatic fever and rheumatic heart disease are the most serious autoimmune sequelae of group A streptococcal infection and have afflicted children worldwide with disability and death. Group A streptococcal infections have recently been associated with Tourette's syndrome, tics, and movement and attention deficit disorders. This review will address the potential pathogenic mechanisms involved in poststreptococcal sequelae.

The Lancefield classification scheme of serologic typing distinguished the beta-hemolytic streptococci based on their group A carbohydrate, composed of N-acetylglucosamine linked to a rhamnose polymer backbone. Streptococci were also serologically separated into M protein serotypes based on a surface protein that could be extracted from the bacteria with boiling hydrochloric acid. Currently, more than 80 M protein serotypes have been identified, and a molecular approach to identification of emm (M protein) genes has been achieved. Vaccines containing the streptococcal M protein as well as other surface components are under investigation for prevention of streptococcal infections and their sequelae. This review will focus on the pathogenic mechanisms in group A streptococcal diseases and on new developments which have an impact on our understanding of group A streptococcal diseases in humans.

Although group A streptococci are exquisitely sensitive to penicillin, an unexplained resurgence of group A streptococcal infections has been observed since the mid-1980s (275). The first indication that infections due to S. pyogenes were on the rise was an outbreak of rheumatic fever which affected approximately 200 children during a 5-year period (531). From the mid-1980s to the 1990s, eight rheumatic fever outbreaks were documented in the United States, with the largest in Salt Lake City, Utah (17, 275, 531). Outbreaks were reported in Pennsylvania, Ohio, Tennessee, and West Virginia and at the Naval Training Center in San Diego, Calif. (17). A decline in rheumatic fever with a milder disease pattern had been observed in the previous decade (59). Therefore, the increased severity and the attack on middle-class families deviated from the past epidemiological patterns. Streptococcal M protein serotypes associated with the new outbreaks of rheumatic fever were M types 1, 3, 5, 6, and 18 (280).

In the late 1980s, streptococcal toxic shock syndrome, bacteremia, and severe, invasive group A streptococcal skin and soft tissue infections were reported in the United States and Europe (103, 212, 241, 275, 376, 498). Increased bacteremic infections were reported in Colorado, Sweden, and the United Kingdom (93a, 510). These severe and invasive diseases have high morbidity and mortality and may be linked to the emergence of certain serotypes or clonotypes. Although several different M protein serotypes have been isolated from severe, invasive streptococcal diseases, M protein serotype 1 has dominated (241, 245). Host susceptibility may be an important factor in production of toxic streptococcal syndrome. These aspects will be considered in this review. The resurgence of these serious infections and sequelae is strong evidence for increased awareness and surveillance of group A streptococci in the community.

Group A streptococci are the most common bacterial cause of pharyngitis and primarily affect school-age children 5 to 15 years of age (62). All ages are susceptible to spread of the organism under crowded conditions, such as those at schools and military facilities. Pharyngitis and its association with rheumatic fever are seasonal, occurring in the fall and winter (62, 506, 507). This is in contrast to pyoderma or skin infection, which occurs in the summer and can be associated with the production of acute glomerulonephritis (61). Organisms which colonize the skin can also colonize the throat, but streptococcal strains which commonly produce skin infections do not lead to rheumatic fever. Groups C and G can also cause pharyngitis and must be distinguished from group A organisms after throat culture (58, 62). Although they are not considered normal flora, pharyngeal carriage of group A streptococci can occur without clinical symptoms of disease.

Certain M protein serotypes, such as M types 1, 3, 5, 6, 14, 18, 19, and 24 of S. pyogenes, are found associated with throat infection and rheumatic fever. These pharyngitis-associated serotypes do not produce opacity factor as do M serotypes such as 2, 49, 57, 59, 60, and 61, which are associated with pyoderma and acute glomerulonephritis (60, 62, 506, 507; Facklam, personal communication). The skin and throat serotypes have been divided epidemiologically on the basis of (i) opacity factor production, (ii) the presence of the class I C repeat region epitope identified on M proteins by anti-M protein monoclonal antibody (MAb) 10B6, and (iii) emm (M protein) gene patterns A through E (50, 57).

Although usually associated with streptococcal throat infection, scarlet fever may occur due to infections at other sites (62). The group A streptococcal strain producing scarlet fever does so because it carries the genes for one or more of the streptococcal pyrogenic exotoxins A, B, and C. The genes for exotoxins A and C are encoded on a lysogenic temperate bacteriophage (66, 546), while exotoxin B is chromosomal. The pyrogenic exotoxins, currently known as streptococcal superantigens, are responsible for the rash, strawberry tongue, and desquamation of the skin seen in scarlet fever.

Group A streptococci which invade the skin and cause impetigo are different M protein serotypes from those that cause pharyngitis (50, 61, 506, 507). In addition, some of the skin strains are associated with production of acute poststreptococcal glomerulonephritis. The skin infections and nephritis are seasonal, usually occurring during the summer months and in temperate climates. The infection is limited to the epidermis, usually on the face or extremities, and is highly contagious (65). Streptococcal strains which cause pyoderma do not cause rheumatic fever. Staphylococci may be mixed with streptococci in impetigo, and thus the treatment of choice is not penicillin for penicillinase-producing staphylococci (65). Group A streptococcal strains may enter the skin through abrasions and other types of lesions to penetrate the epidermis and produce erysipelas or cellulitis. Erysipelas is a distinctive form of cellulitis with characteristically raised and erythematous superficial layers of the skin, while cellulitis affects subcutaneous tissues (65). Cellulitis may occur from infected burns or wounds. Both erysipelas and cellulitis can be caused by streptococcal groups A, B, C, and G.

Introduction. In 1987, Cone and colleagues described a toxic shock-like syndrome due to S. pyogenes (103). Later in 1989, Stevens described an unusually severe form of group A streptococcal disease which was similar to the staphylococcal toxic shock syndrome (498). Streptococcal toxic shock syndrome was characterized by hypotension and multiple organ failure. The Working Group on Severe Streptococcal Infections suggested a case definition of streptococcal toxic shock syndrome and necrotizing fasciitis (569a). Necrotizing fasciitis may accompany the toxic streptococcal syndrome. Group A streptococcal infection producing the toxic syndrome may occur in muscle and fascia and follow mild trauma with entrance of streptococci through the skin. In addition, group A streptococci may infect the vaginal mucosa and uterus, leading to severe disease or septicemia. Several excellent recent reviews on severe invasive streptococcal disease report on its features and pathogenesis (3, 290, 379, 498). The features of invasive streptococcal infections include hypotension and shock, multiple organ failure, systemic toxicity, severe local pain, rapid necrosis of subcutaneous tissues and skin, and gangrene (65). Predisposing factors include skin trauma, surgery, varicella, and burns.

Pyrogenic exotoxins and superantigens in invasive disease. Several virulence factors of group A streptococci are likely to be involved in the pathogenesis of toxic shock, invasion of soft tissues and skin, and necrotizing fasciitis. These virulence factors are the extracellular pyrogenic exotoxins A, B, and C as well as newly discovered exotoxins and superantigens such as exotoxin F (mitogenic factor) and streptococcal superantigen (SSA) (366, 394, 395). In addition, several new superantigens with strong mitogenic activity have recently been reported as SpeG, SpeH, SpeJ, SmeZ, and SmeZ-2 (274, 434). Details about the mitogenic toxins are described in a separate section under virulence factors. These new data point to the fact that there are a large number of superantigens which may play a role in toxic streptococcal syndrome. All of these toxins act as superantigens which interact with major histocompatibility complex (MHC) class II molecules and a limited number of V regions of the T-lymphocyte receptor to activate massive numbers of T cells nonspecifically. The activation liberates large amounts of interleukins as well as other inflammatory cytokines such as tumor necrosis factor and gamma interferon (171, 221, 395). The pyrogenic exotoxins are potentially responsible for at least some of the manifestations of toxic streptococcal syndrome. Kotb and colleagues demonstrated evidence for selective depletion of T cells expressing V1, V5.1, and V12 in patients with streptococcal toxic shock syndrome, further supporting the hypothesis that the streptococcal superantigens play an important role in disease pathogenesis (544). Further evidence also suggests that streptococcal isolates from toxic streptococcal syndrome induce a Th1 rather than a Th2 cytokine response, which is characteristic of superantigens (393).

Pyrogenic exotoxin B. Pyrogenic exotoxin B is an extracellular cysteine protease which has been shown to cleave fibronectin and vitronectin (288), extracellular matrix proteins, and human interleukin-1 into the active form of the molecule (287). Therefore, the protease may be important in inflammation, shock, and tissue destruction. Humans with a diverse range of invasive disease (erysipelas, cellulitis, pneumonia, bacteremia, septic arthritis, toxic shock syndrome, and necrotizing fasciitis) all produced elevated levels of antibodies against streptococcal pyrogenic exotoxin B following infection (214).

M protein serotypes. M protein serotypes are nonrandomly represented among invasive-disease-causing strains (91, 100, 102, 369, 377, 378, 381, 389, 482). In clinical reports and epidemiological studies of invasive and toxic streptococcal diseases, M types 1, 3, 11, 12, and 28 have frequently been reported, with M1 and M3 being the most common. Other serotypes have occasionally been reported. Using restriction fragment length polymorphism (RFLP) as detected in pulsed-field gel electrophoresis, two subclones from invasive disease were identified by RFLP type and multilocus enzyme electrophoretic type (377, 378). The M1T1 genotype was recovered from both invasive disease and uncomplicated pharyngitis, suggesting that host factors as well as the streptococcal strain play a role in the development of severe disease (369, 379). The M1T1 genotype was studied during a recent epidemic in Sweden; the conclusion was that M1T1 did not appear to be clonal, since it had genetic diversity downstream of the emm-1 gene and both genes for erythrogenic toxins A and B exhibited clonal variation (389). This was in contrast to findings reported by Cleary and others that M1 may be clonotypic (100). Recent data reported by Stockbauer and colleagues (500) show that invasive M1 organisms are a heterogeneous array of related organisms that differ by variation in the streptococcal inhibitor of complement.

Animal models of invasive soft-tissue infection. Wessels and colleagues have developed a murine model of human invasive soft-tissue infection (16). The animals challenged with wild-type M3 streptococci developed a spreading soft-tissue necrosis, with bacteremia and death. Animals challenged with acapsular or M protein-deficient M3 mutants did not develop the spreading, necrotizing disease but developed a localized abscess. In further studies, mutants of the M3 strain from which the cysteine protease gene was deleted caused the same necrosis and spreading disease as the wild type. Therefore, in the murine model of soft-tissue infection described by Ashbaugh et al. (16), the capsule and M protein play a major role in development of the spreading necrotic lesion.

Treatment. Treatment of toxic and severe invasive disease with antibiotics is not always effective, and mortality can exceed 50% (149). The failure of penicillin to treat severe invasive streptococcal infections successfully is attributed to the phenomenon that a large inoculum reaches stationary phase quickly and penicillin is not very effective against slow-growing bacteria (497, 499). Treatment with clindamycin in experimental models of fulminant streptococcal infections appears to be more efficacious than penicillin, but this has not yet been demonstrated in humans. Clindamycin acts on protein synthesis rather than on cell wall synthesis.

IDENTIFICATION OF THE ORGANISM: OLD AND NEW TECHNIQUES

Top Previous Next References

Description and Clinical Microbiology

Throat culture. The group A streptococci have long been recognized as the Streptococcus sp. associated with acute pharyngitis. In a positive throat culture, group A streptococci appear as beta-hemolytic colonies among other normal throat flora which are usually alpha- or nonhemolytic on 5% sheep blood agar. S. pyogenes may appear as highly mucoid to nonmucoid; colonies are catalase negative. Optimal recovery of group A streptococci may be achieved by use of blood agar plates containing sulfamethoxazole-trimethoprim to inhibit some of the normal flora and growth under anaerobic conditions to enhance streptolysin O activity (303). Throat culture is still recognized as the most reliable method for detecting the presence of group A streptococci in the throat (170). Presumptive identification of the beta-hemolytic group A streptococci relies on susceptibility to bacitracin or a positive pyrrolidonylarylamidase test (170).

Lancefield group. The Lancefield serological grouping system for identification of streptococci is based on the immunological differences in their cell wall polysaccharides (groups A, B, C, F, and G) or lipoteichoic acids (group D) (303). The group A carbohydrate antigen is composed of N-acetyl--D-glucosamine linked to a polymeric rhamnose backbone. Confirmation of S. pyogenes is done by highly accurate serological methods, such as the Lancefield capillary precipitin technique and the slide agglutination procedure, which utilize standardized grouping antisera (170). These methods, including Streptex on primary plates (24 h) or subculture (48 h), would confirm group A streptococci. For this reason, rapid tests which screen for the presence of group A streptococci in the throat have been developed and are popular in the clinical setting (170, 303). Facklam has recently reviewed the currently available group A screening tests and discusses their sensitivity and specificity in comparison with the conventional methods (170). It is beyond the scope of this review to describe the many tests available for identification of group A streptococci from throat swabs. However, the most rapid tests take 5 to 30 min and use some form of nitrous acid or enzymatic extraction of the group A carbohydrate (170). Fluorescent-antibody and genetic probe tests can be performed directly on throat swabs but are not easily adapted to the clinical setting. Once extracted, the group A carbohydrate antigen is detected by one of four methods, including slide agglutination, enzyme-linked immunosorbent assay (ELISA), optical immunoassay, and a modified one-step ELISA procedure (170).

M protein and T typing: development of a molecular biology approach. Streptococcal M protein, which extends from the cell membrane of group A streptococci, has been used to divide S. pyogenes into serotypes. Quite a number of years ago, Lancefield designed a serotyping system for the identification of the M protein serotypes (317). The method consisted of treating group A streptococci grown in Todd-Hewitt broth with boiling 0.1 N HCl. This method extracted the group A carbohydrate, M protein, and cell wall, and the clarified extract was used in capillary precipitin tests to determine the M protein serotype with standardized typing sera. The N-terminal region of the M protein has been demonstrated to contain the type-specific moiety and is recognized by specific typing sera in the precipitin test (28, 179, 271, 318). There were several difficulties with M serotyping, including ambiguities in the results, discovery of new M types, difficulty in obtaining high-titered antisera against opacity factor-positive strains, and the availability and high cost of preparing high-titered antisera for all known serotypes (170). Currently, more than 80 different serotypes of M protein have been identified (170).

Serological diagnosis of group A streptococcal infections is based on immune responses against the extracellular products streptolysin O, DNase B, hyaluronidase, NADase, and streptokinase, which induce strong immune responses in the infected host (507). Anti-streptolysin O (ASO) is the antibody response most often examined in serological tests to confirm antecedent streptococcal infection. Todd developed the assay for ASO antibodies by 1932 (520). An increase in the ASO titer of 166 Todd units is generally accepted as evidence of a group A streptococcal infection. In previous studies it has been shown that infants are born with maternal levels of antistreptococcal antibodies and that infants develop streptococcal infections after the first year of life. ASO antibodies may not demonstrate a detectable rise in 1- to 3-year-olds, who have had few previous group A streptococcal infections (349). At <2 years of age, >50% of the patients had ASO titers of <50 Todd units, and none of the patients had titers above 166 (349). In the same study, older school-age children developed higher ASO titers. All five of the extracellular streptococcal enzymes may become significantly elevated over normal levels during a streptococcal infection. Although the ASO titer is the standard serological assay for confirmation of a group A streptococcal infection, assay of several of the enzymes enhances the chance for a positive test if the patient did not produce high levels of antibody against one or more of the extracellular enzymes. In general, the titers of antibodies against the extracellular products parallel each other; however, exceptions may be seen in infections with pyoderma or nephritogenic strains, when the anti-DNase B titers have been found to be a reliable indicator of streptococcal infection (507). Infection of the skin does not always elicit a strong ASO response.

Confirmation of a group A streptococcal infection is imperative for the diagnosis of rheumatic fever, since most often streptococci cannot be cultured from the pharynx. In rheumatic fever, approximately 80% of the patients will have an elevated ASO titer (>200 Todd units) at 2 months after onset (61). If an anti-DNase B or antihyaluronidase assay is performed on sera from these patients, the number of patients with at least one positive antistreptococcal enzyme titer rises to 95%. Thus, most acute rheumatic fever cases demonstrate an elevated ASO titer with some exceptions which require more than one antibody test to detect previous group A streptococcal infection. The streptozyme test was developed some years ago as a hemagglutination assay for the detection of multiple antibodies against extracellular products such as anti-streptolysin O, anti-DNase B, antihyaluronidase, antistreptokinase, or anti-NADase, and it is used clinically in some laboratories as an additional diagnostic test (61).

PATHOGENESIS: INTERACTION BETWEEN HOST AND PATHOGEN

Top Previous Next References

Adherence and Colonization

The adhesion process involves multiple group A streptococcal adhesins reported by several investigators as detailed below and described in excellent reviews (19, 108, 226, 227). Adherence has been outlined in these reviews as an initial weak interaction with the mucosa which is followed by a second adherence event which confers tissue specificity and high-avidity adherence (227). In addition, one could speculate that the presence of multiple adhesins in strains could give them the advantage of more avid adherence and potentially enhanced virulence. Although not yet well understood, environmental factors expressed in a particular body site may be important cues for expression of adhesins important for colonization of a tissue-specific site. It is possible that movement of streptococci from the mucosa or skin into deeper tissues may be facilitated by specialized adhesion mechanisms. Recent studies indicate that adherence to particular types of host cells may induce localized cytokine production and inflammatory responses (110, 536).

Multiple adhesins. In early studies of bacterial adherence, Ellen and Gibbons suggested that an adhesin was associated with the M protein fimbriae on the surface of the group A streptococci (168, 169). Shortly thereafter, Beachey and Ofek published the first paper describing lipoteichoic acid (LTA) as an adhesin (27). Their data suggested that M protein was not the adhesin, but that LTA, an amphipathic molecule, was the adhesin for buccal epithelial cells. In these first studies of LTA, experiments characterizing adhesion were established (27). Antibody against LTA on the group A streptococci blocked adhesion to epithelial cells, and LTA, when reacted with epithelial cells, saturated the epithelial cell adhesin and inhibited adherence. Later, fibronectin was identified as the epithelial cell receptor binding LTA (488). In these studies, fibronectin was found to inhibit adhesion of group A streptococci to epithelial cells, and antifibronectin was found to block the binding of streptococci to epithelial cells. The LTA molecule was found to act as an adhesin by reacting with molecules on the streptococcal surface, such as the M protein, through its negatively charged polyglycerol phosphate backbone and positively charged residues of surface proteins. The lipid moiety of LTA projected outward and interacted with fatty acid-binding sites on fibronectin and epithelial cells (397). Evidence suggested that LTA accounted for approximately 60% of adhesion to epithelial cells, indicating that other adhesins were involved in the adherence of group A streptococci to epithelial cells. In a recent review (226), Hasty and Courtney state that at least 11 adhesins have been described for group A streptococci, including M protein (90, 136, 168, 169, 401, 540), LTA (106, 109, 111, 398, 488, 489), protein F/Sfb (223, 224), a 29-kDa fibronectin-binding protein (105), glyceraldehyde-3-phosphate dehydrogenase (412, 563), a 70-kDa galactose-binding protein (201, 535), a vitronectin-binding protein (526), a collagen-binding protein (533), serum opacity factor (310), a 54-kDa fibronectin binding-protein, FBP54 (109), and the hyaluronate capsule (549). Several extracellular host cell proteins have been implicated in attachment or adherence to group A streptococci, including fibronectin (105, 488), fibrinogen (463), collagen (533), vitronectin (526), a fucosylated glycoprotein (541), and integral membrane proteins including CD46, the membrane cofactor protein on keratinocytes (400), and CD44, the hyaluronate-binding receptor on keratinocytes (471). Table 1 summarizes the streptococcal adhesins and host receptors described previously.

In the past few years, new evidence suggested that group A streptococci not only adhere to epithelial cells but also invade them (324). LaPenta and colleagues demonstrated that group A streptococci have the potential to invade human epithelial cells at frequencies equal to or greater than classical intracellular bacterial pathogens, such as Listeria and Salmonella spp. (324). This initial report generated considerable interest and was confirmed by several laboratories (187, 211, 261, 365). Figure 1 illustrates invasion of epithelial cells by group A streptococci (187). High-frequency invasion requires expression of M protein (118) and/or fibronectin-binding proteins such as SfbI (261, 365). Both the M1 protein and SFbI are considered invasins because latex beads coated with either protein are efficiently internalized by epithelial cells. In addition, mutations in genes that encode these proteins reduce the capacity of streptococci to invade cultured cells. Most recent work has demonstrated high-frequency intracellular invasion of epithelial cells by the M1 serotype of group A streptococci (160). The investigation demonstrated cytoskeletal rearrangements within the cells during the invasion process.

View larger version (115K): [in this window] [in a new window]   FIG. 1.   Electron micrographs demonstrating the attachment and internalization of streptococci by human cultured pharyngeal cells. Group A streptococci were observed to associate with microvilli upon initial contact with the pharyngeal cells. Membrane extension occurred during the internalization process. Surface interaction can be seen between the pharyngeal cell and streptococcus. Intracellular streptococci were found ingulfed in cytoplasmic vacuoles. Magnifications: A, ×12,700; B, ×24,300. (Reprinted from reference 187 with permission from the publisher.)

Fibronectin (118) and high-affinity fibronectin-binding proteins (406) trigger invasion by engagement of 51 integrin receptors on epithelial cells. Group A streptococci can invade human cells by other mechanisms. Laminin has been shown to bind group A streptococci (515) and will induce ingestion of M1 streptococci, independent of serum and fibronectin. Small peptides with the RGD amino acid sequence stimulate uptake in the absence of serum and M protein. The fact that group A streptococci evolved multiple routes to the interior of epithelial cells is a strong indication that intracellular invasion plays an important role in their pathogenesis. The most direct evidence that intracellular invasion is more than a laboratory phenomenon comes from studies of patients with recurrent tonsillitis. Failure to eradicate streptococci from the throat in pharyngotonsillitis occurs in approximately 30% of cases (348, 421). Osterlund and colleagues showed that tonsils excised from such individuals harbored intracellular group A streptococci (403). Others report that strains isolated from carriers are exceptionally able to invade HEp-2 cells in vitro (364). The reason for invasion of host cells is not entirely clear, although the streptococci may find the intracellular environment to be a good place to avoid host defense mechanisms. Therefore, internalization of streptococci may lead to carriage and persistence of streptococcal infection. In further studies, an association of the presence of the fibronectin-binding gene prtF1 with streptococcal strains from antibiotic treatment failures was found (383). In the treatment failures, 92.3% of the strains contained the prtF1 gene, while 29.6% of strains from eradicated infections did not.

Two theories have been proposed for the role of internalization of group A streptococci in disease pathogenesis. It has been suggested to potentially play a role in the carriage and persistence of streptococci, as stated above. Second, studies suggest that internalization may lead to invasion of deeper tissues (324), while other studies have found that low virulence was associated with internalization (472). Perhaps both theories are correct depending on the virulence and properties of the invading bacterium and whether the invasion is of the throat or skin epithelium. It is also possible that internalization of group A streptococci by host epithelial cells represents successful containment of the pathogen by the host. This hypothesis is supported by the observation that poorly encapsulated strains are internalized most efficiently but are relatively avirulent in infection models (472). Future study of these and other questions about intracellular invasion will no doubt yield new and unexpected clues to the role of internalization in the pathogenesis of group A streptococci.

It is well established that group A streptococci are antiphagocytic due to surface exposed M protein and hyaluronic acid capsule (368, 551). Two mechanisms have been proposed to explain the antiphagocytic behavior of M-positive streptococci. One mechanism is the binding of factor H, which inhibits the activation of the complement pathway (246). Factor H is a regulatory component of the complement pathway, which inhibits the deposition of soluble C3b. Factor H binds to the C repeat region of the M proteins, and deletion of the C1 and C2 repeat regions reduces factor H binding (418).

The antiphagocytic behavior of group A streptococci is also mediated by the binding of fibrinogen to the surface of M protein (555-557). Fibrinogen binding to the surface of group A streptococci blocks the activation of complement via the alternate pathway and greatly reduces the amount of C3b bound to streptococci, which therefore reduces phagocytosis by polymorphonuclear leukocytes (247). Type-specific M protein antibodies overcome this effect by binding to the exposed N-terminal M protein epitopes. This results in activation of the classical complement pathway, deposition of C3b, and subsequent phagocytosis. Figure 2 illustrates opsonization of group A streptococci by M type-specific antibody and complement.

View larger version (21K): [in this window] [in a new window]   FIG. 2.   How the immune system recognizes group A streptococci and uses opsonization by complement and type-specific antibody against M protein or any other surface molecule capable of generating opsonic antibody. Fc receptors shown on macrophages bind to the antibody Fc region, inducing phagocytosis and killing of the streptococci.

In addition to the antiphagocytic properties of the M protein, other surface molecules contribute to resistance to phagocytosis by the group A streptococcus. Recent studies have shown that mutations in Mrp (M-related protein) affect resistance to phagocytosis compared with wild-type strains, and in fact, insertion mutations in any one of the genes emm-49, enn-49, and mrp-49 resulted in reduced resistance to phagocytosis in human blood and purified polymorphonuclear leukocytes (PMNs) (425). From these results Podbielski suggested that resistance to phagocytosis depended on the cooperative effects of all three genes. Further studies by Ji, Cleary, and colleagues confirmed this hypothesis by demonstrating that failure to produce all three M-like proteins, M49, Mrp, and Enn-49, reduced resistance to phagocytosis but did not alter the persistence of streptococci at the oral mucosa (266).

It was found that C5a-activated PMNs were able to kill M-positive streptococci. It has been shown previously that C5a alters the clearance and trafficking of group A streptococci during infection (265). By inactivating C5a, C5a peptidase blocks chemotaxis of PMNs and mononuclear phagocytes to the site of infection (265, 552), while the M protein is limiting the deposition of complement on the surface of the streptococci (260). Therefore, the multiple mechanisms involved in resistance to phagocytosis in a bacterium where antiphagocytic behavior is essential for survival may have only recently been appreciated.

The group A streptococci are covered with an outer hyaluronic acid capsule (298), while the group A carbohydrate antigen and the type-specific M protein are attached to the bacterial cell wall and membrane, as shown in Fig. 3. Both the M protein and the capsule are considered virulence factors conferring antiphagocytic properties upon the streptococcal cell (179, 189, 368).

View larger version (29K): [in this window] [in a new window]   FIG. 3.   Diagram of the group A streptococcal cell covered with an outer hyaluronic acid capsule and the group A carbohydrate, consisting of a polymer of rhamnose with N-acetylglucosamine side chains. Streptococcal M protein extends from the cell wall and is anchored in the membrane. (Reprinted from reference 119 with permission from the publisher.)

Hyaluronic acid capsule. The group A streptococcal capsule is composed of a polymer of hyaluronic acid containing repeating units of glucuronic acid and N-acetylglucosamine (509). Synthesis of the polymer involves the products of three genes, hasA, hasB, and hasC, which are all in the same operon (161). The hasA gene encodes hyaluronate synthase (152, 163); hasB encodes UDP-glucose dehydrogenase (162); and hasC encodes UDP-glucose pyrophosphorylase (112). The expression of the has genes is transcriptionally controlled. The three has genes are transcribed as a single message from the promotor upstream of hasA (113). However, only hasA and hasB are required for capsule expression in group A streptococci (15).

M protein. The group A streptococcal M proteins have been studied extensively since their discovery (317-320). Two excellent comprehensive reviews by Fischetti describe the characteristics of the M protein molecule in detail (177, 179). The M protein is a major surface protein and virulence factor of group A streptococci, with more than 80 distinct serotypes identified. The amino-terminal region extends from the surface of the streptococcal wall, while the carboxy-terminal region is within the membrane. The M protein is anchored in the cell membrane by the LPSTGE motif identified by Fischetti and colleagues (183). The M protein extends from the cell surface as an alpha-helical coiled-coil dimer which appears as fibrils on the surface of group A streptococci (419) (Fig. 4).

View larger version (94K): [in this window] [in a new window]   FIG. 4.   Electron micrograph of group A M type 24 streptococci. The surface fimbriae or hairlike projections indicate the presence of M protein on the surface of the streptococci.

View larger version (10K): [in this window] [in a new window]   FIG. 5.   The A, B, C, and D repeat regions of M protein, with the protein anchor and pepsin cleavage site shown. The A repeat region varies between serotypes and contains the highly variable serotype-specific amino acid sequence of M protein at the N terminus. The B repeat region varies from serotype to serotype, while the C repeat region contains a conserved sequence shared among all of the serotypes. The anchor region contains the LPXTGX motif required to anchor gram-positive proteins in the cell membrane. (Reprinted from reference 119 with permission from the publisher.)

emm-like genes and the emm gene superfamily. Genes related to the M protein gene (emm) are called the M gene superfamily and include immunoglobulin-binding proteins, M-related proteins, and M proteins. These proteins may possess functional properties of immunoglobulin binding or antiphagocytic behavior. According to Hollingshead and colleagues (238), more than 20 genes have been identified in the emm gene superfamily, in which the genes shared greater than 70% DNA sequence identity at their 5' ends (238). The highly conserved identity is found within three distinct domains in the cell-associated region of the M protein molecule. These domains include domain H, which may serve to anchor the protein in the membrane; the peptidoglycan-associated domain; and the cell wall-associated domain. The domains are all highly conserved among emm gene products compared with the other regions of the molecules which are surface exposed (238). These domains do not include the C repeat region of M proteins, which can be divided into class I and class II M proteins depending on the presence of the class I epitope detected by MAb 10B6 and others (50, 52). Phylogenetic analysis of the region has revealed four major lineages, designated subfamilies SF1 through SF4, that contain differences in the peptidoglycan-spanning domain of the M or M-like protein (53, 238, 239). Five major emm chromosomal patterns of the subfamily genes were identified based on the number and arrangement of the emm subfamily genes (56). These subfamily gene arrangement patterns were designated A through E. A given strain has one, two, or three emm or emm-like genes that are tandemly arranged on the chromosome near the positive transcriptional regulator called the multiple gene regulator of group A streptococci (mga). The emm gene patterns display several phenotypes. Table 2 summarizes the patterns of emm genes associated with skin and throat infections.