Group A streptococcus: Virulence factors and pathogenic mechanisms

INTRODUCTION — Group A Streptococcus (GAS), also known as Streptococcus pyogenes, causes a broad range of infections and complications. The interaction between the host and this important pathogen will be reviewed here. The wide variety of GAS infections and postinfectious sequelae are discussed separately. (See related topics.)

GENERAL CHARACTERISTICS — GAS is a facultative, gram-positive coccus that grows in chains. It causes an array of infections involving the respiratory tract and soft tissues ranging in severity from mild to severe. The only known reservoirs are the skin and mucous membranes of the human host. The pathogenic mechanisms underlying these infections are poorly understood, largely because each is the culmination of highly complex interactions between the human host defense mechanisms and specific virulence factors of the organism.

MICROBIOLOGIC CHARACTERISTICS — GAS requires complex media for optimal growth in the laboratory. It grows best in an environment of 10 percent carbon dioxide and produces colonies on blood agar plates surrounded by a zone of complete (beta) hemolysis. Colonies are typically 0.5 to 1.0 mm in diameter, although some strains grow as larger, translucent-appearing or "mucoid" colonies due to abundant production of the hyaluronic acid capsular polysaccharide. (See 'Capsule' below.)

Classification schemes — The Lancefield classification divides streptococci into types A through O based upon reactions of specific antisera with acid-extractable cell wall carbohydrate antigens [1]. GAS has also been subdivided based upon serotyping of surface-expressed M and T antigens. M-typing using specific antisera has been largely supplanted by emm-typing, that is, sequencing the variable region of the emm gene. Increasingly, whole genome sequencing is used for definitive identification of outbreak-associated strains [2,3]. (See 'M and M-like proteins' below.)

VIRULENCE FACTORS — A number of different cell-surface molecules and secreted products of GAS have been identified as virulence factors (figure 1) [4].

Cell-surface molecules

M and M-like proteins — M protein, encoded by the emm gene, is a dimeric, coiled-coil structure consisting of a hypervariable N-terminus, the sequence of which determines the M type, and four regions of repeating amino acids (A to D). At its C-terminus, M protein contains an LPXTG motif through which it is covalently linked to the bacterial cell wall [5]. Antibodies to N-terminal region confer type-specific immunity. For many M types, regions within the B, C, or D repeats bind one or more plasma proteins such as fibrinogen, plasminogen, immunoglobulin G (IgG), and/or complement regulatory proteins such as factor H, factor H-like protein, and C4b-binding protein.

M protein protects the organism against phagocytosis by polymorphonuclear leukocytes [6], although this resistance to phagocytosis can be overcome by type-specific antibodies [1,7,8]. Binding of complement regulatory proteins also contributes to resistance to phagocytosis by limiting deposition of opsonins such as C3b on the bacterial surface. (See 'Anti-phagocytic properties' below.)

More than 200 different M protein types of GAS have been described, and new M types are constantly emerging. Nontypeable strains of GAS may express minute amounts of M protein, express a totally new M type that is not reflected in the current typing protocols, or may lack M protein altogether. As an example, 60 nontypeable strains isolated from the upper respiratory tract demonstrated resistance to phagocytosis in normal serum or plasma, suggesting that these strains do in fact produce M protein [9]. This could result from mutations within the emm gene, leading to production of novel M proteins that do not cross-react with standard typing sera.

Heterogeneity within the emm gene has been demonstrated among M-1 strains [10]. In one study, approximately half of randomly collected human serum samples opsonized all seven M-1 strains tested [11]. The remaining samples opsonized only some M-1 strains, suggesting that opsonic antibody may be strain specific rather than M protein–type specific [11].

These limitations of serologic typing for surface M proteins with available polyclonal sera led to development of an alternate means of M type identification by emm typing; investigators designed primer pairs that allowed amplification by polymerase chain reaction (PCR) of the emm genes of all available M type reference strains [12]. Sequence analysis of specific PCR products was used to accurately deduce emm genes corresponding with the majority of the known M serotypes. This was possible because the highly variable N-terminus of the M protein that confers antigenic specificity is encoded by a variable 5' region of the emm gene.

Vaccine implications — No vaccines for GAS are commercially available for use in humans. However, several vaccines are under study in mouse and nonhuman primates and some have progressed through early-stage human clinical trials [13]. Multiple GAS antigens have been evaluated for vaccine potential.

Typing of emm genes has important implications for GAS virulence and vaccine development. In one report of 5400 cases of invasive GAS infection, emm types 1, 3, and 12 were independent risk factors associated with mortality [14]. The emm types in a proposed 26-valent vaccine [15,16] accounted for 79 percent of isolates, for 85 and 88 percent of isolates implicated in necrotizing fasciitis and streptococcal toxic shock syndrome (TSS), respectively, and for 79 percent of deaths. The authors calculated that the proposed vaccine could prevent up to 50 and 63 percent of invasive GAS infections among older adults and children, respectively. However, the distribution of these M types may not be uniform in all regions, and serotypes may shift over time, as has been observed with the pneumococcal conjugate vaccine [17-19]. In a subsequent randomized trial, a 30-valent M-protein vaccine was well tolerated in healthy adult human volunteers and elicited strong immunologic responses to 24 of the 30 component M-antigens [20].

A study demonstrated that convalescent sera from patients with skin and soft tissue infection contained opsonic antibody against not only the M type of the causative strain but also to a cluster of genetically related M types of GAS [21]. This finding suggests that fewer peptides of M protein may be required to provide protective antibodies in vaccine preparations. Using a structure/function-based computational approach, a method was developed to predict antibody cross-reactive immunogenicity among M-protein peptides [22]. Using this approach, a six-peptide vaccine could elicit significant cross-reactive antibody against 67 percent of non-vaccine strains [22].

Lastly, the epidemiologic observation that pharyngitis is less frequent in adults compared with children suggests that immunity to GAS develops through natural exposure; it is unclear whether protective immunity depends exclusively or primarily on antibodies to specific M proteins or whether antibodies and/or T cell responses to other GAS antigens also play a role.

Toxins and other secreted virulence factors — Secreted molecules thought to contribute to pathogenesis include the hemolytic toxins streptolysin O (SLO) and streptolysin S (SLS); enzymes such as hyaluronidase, streptokinase, nicotinamide-adenine dinucleotidase, deoxyribonucleases; and pyrogenic exotoxins.

Hemolytic toxins — The zone of complete or beta hemolysis surrounding GAS colonies on blood agar is due to the hemolytic action of SLS (picture 1). The hemolytic activity of SLO is inactivated by atmospheric oxygen but can be demonstrated by stabbing the GAS inoculum below the surface of the agar in strains that lack SLS.

Streptolysin O — SLO belongs to a family of oxygen-labile, thiol-activated cytolysins. In sublytic concentrations, SLO rapidly and profoundly activates other cell types such as platelets and endothelial cells, contributing to perfusion deficits that lead to ischemic destruction of tissue [23]. There is substantial amino acid homology between SLO and thiol-activated cytolysins from other gram-positive bacteria [24,25]. Thiol-activated cytolysins bind to cholesterol on eukaryotic cell membranes where cytolysin monomers oligomerize and insert into the cell membrane resulting in cell lysis via a colloid-osmotic mechanism. Free cholesterol inhibits both toxicity in isolated myocytes and hemolysis of red blood cells in vitro. In situations in which serum cholesterol is high (such as nephrotic syndrome), antistreptolysin O (ASO) titers may be falsely elevated because both cholesterol and anti-ASO antibody in serum neutralize SLO. (See "Lipid abnormalities in nephrotic syndrome".)

ASO titers vary with age, season, and geography [26]. Healthy elementary school children frequently have titers of 200 to 300 Todd units per mL, while asymptomatic pharyngeal carriers tend to have very low titers, just above detectable levels. Following streptococcal pharyngitis, the antibody response peaks at about four to five weeks. Nonsuppurative complications such as rheumatic fever and poststreptococcal glomerulonephritis generally develop during the second or third week of illness. Therefore, it may be useful to perform one titer when a nonsuppurative complication is first suspected and subsequently repeat the titer approximately two weeks later. Antibody titers fall off rapidly in the next several months and, after six months, have a slower decline.

About 80 percent of patients with acute rheumatic fever or poststreptococcal glomerulonephritis demonstrate a rise in ASO titer; however, the degree of ASO titer elevation does not correlate with severity of disease. In patients with suspected rheumatic fever or glomerulonephritis but with an undetectable ASO titer, prompt testing for other antistreptococcal antibodies such as anti-DNase B (detectable for six to nine months following infection), streptokinase, and antihyaluronidase should be performed. About 90 percent of patients with acute rheumatic fever or poststreptococcal glomerulonephritis have at least one positive result if antibodies to two antigens are evaluated; about 95 percent have a positive result if responses to three antigens are evaluated.

Streptolysin S — SLS is a cytolysin on the surface of GAS that lyses red blood cells and likely other eukaryotic cells; its role in pathogenesis is uncertain. A study in mice provided evidence that SLS is a critical determinant of pain during invasive soft tissue infection and may interfere with host defense. SLS stimulated peripheral sensory nerves to secrete the neuropeptide calcitonin gene-related peptide, which inhibited neutrophil influx and bacterial killing [27].

Other enzymes implicated in virulence

●Hyaluronidase – Hyaluronidase is an extracellular enzyme that hydrolyzes hyaluronic acid in connective tissues and may facilitate the spread of infection along fascial planes. Anti-hyaluronidase titers rise following S. pyogenes infections, especially those involving the skin.

●Streptokinase – Streptokinase is produced by all strains of GAS and is located in the extracellular milieu. Plasminogen-binding sites are found on the surface of GAS strains [28]. Once plasminogen is bound, streptokinase proteolytically converts bound plasminogen to active plasmin. Inhibition of the coagulation cascade by this mechanism has been suggested to contribute to the spread of GAS infection in deep tissues. Streptokinase also may play a role in poststreptococcal glomerulonephritis.

●Nicotinamide-adenine dinucleotidase – Streptococcus pyogenes nicotinamide-adenine dinucleotidase (NADase; also called SPN or NAD glycohydrolase) is an extracellular enzyme that is produced by many strains of GAS [29]. In vitro studies suggest that NADase is translocated from the extracellular space into host cells in a process dependent on SLO [30]. Once inside the host cell, NADase is thought to augment SLO-mediated cytotoxicity by degrading intracellular NAD stores [31]. Not all strains of GAS produce enzymatically active NADase [32]. Emergence and global spread of an invasive clone of M type 1 GAS in the 1980s was associated with acquisition of high-level production of active NADase and SLO [2].

●Deoxyribonucleases A, B, C, and D – Expression of deoxyribonuclease (DNase) in vivo, especially DNase B, elicits production of anti-DNase antibody following either pharyngeal or skin infection.

●Two enzymes that may contribute to pathogenesis by interfering with neutrophil migration to the site of infection are C5a peptidase (ScpA), which cleaves the chemotactic complement component C5a, and SpyCEP, a protease that inactivates the chemotactic cytokine interleukin-8 [33,34].

Pyrogenic exotoxins — Streptococcal pyrogenic exotoxins include Spe A, C, and G to M, streptococcal superantigen (SSA), and streptococcal mitogenic exotoxin Z_n (SMEZ_n). These proteins induce lymphocyte blastogenesis, potentiate endotoxin-induced shock, induce fever, suppress antibody synthesis, and act as superantigens [35]. Alternate names include scarlatina toxins and erythrogenic toxins reflecting their association with scarlet fever (see 'Cytokine induction' below). The streptococcal superantigens are encoded by bacteriophages, except for SpeG and SMEZ, which are encoded on the bacterial chromosome [36]. The mechanism for control of exotoxin production is likely to be important, since it is well established that strain production of SpeA has varied over time. In addition, point mutations in the speA gene have been correlated with dramatic changes in the potency of SpeA toxin [37].

Strains capable of producing SpeA and other pyrogenic exotoxins have been associated with scarlet fever and with streptococcal TSS [38,39]. SpeB is a cysteine protease and whether it also has superantigen activity is controversial. Although all strains of GAS have the gene for speB, not all strains produce the SpeB protein. In addition, the quantity of protease produced among SpeB-producing strains varies greatly [36,38,40,41].

The mere presence of virulence factors may be less important than the dynamics of their production in vivo. For example, it has been demonstrated that approximately 40 and 75 percent of strains from patients with necrotizing fasciitis and streptococcal TSS produced SpeA or SpeB, respectively [42]. The quantity of SpeA was higher for strains from patients with streptococcal TSS than for strains from patients with noninvasive disease [42].

Structural components

Capsule — Most strains of S. pyogenes produce a capsular polysaccharide composed of hyaluronic acid; some strains form large mucoid colonies on blood agar as a result of abundant capsule production. The mucoid colony phenotype is characteristic of M type 18 isolates, in which hyperproduction of hyaluronic acid is caused by a mutation in the gene encoding RocA, a regulatory protein [43,44].

Highly encapsulated strains have been shown to resist phagocytosis and are virulent in mice, whereas capsule–deficient mutants are attenuated in experimental infection models and susceptible to killing by phagocytes [45]. These observations suggest that hyaluronic acid capsule, like M protein, confers resistance to phagocytosis.

The hyaluronic acid capsule can impair bacterial adherence to host cells by masking cell wall-associated adhesins; however, the hyaluronic acid capsule itself can mediate attachment to skin or oropharyngeal keratinocytes through binding to CD44 [46,47].

Cell wall — The cell wall is comprised of a peptidoglycan backbone with integral lipoteichoic acid components. The main function of these components is structural stability, although the exact function of lipoteichoic acid is unknown. Lipoteichoic acid may play a role in pathogenesis by facilitating the adherence of GAS to pharyngeal epithelial cells [48]. Peptidoglycan is capable of activating the alternative complement pathway [49,50].

Binding proteins

Ig-binding M-like proteins — GAS produces a family of proteins that share structural similarities to M proteins. Various M and M-like proteins bind immunoglobulins including IgG, IgM, and IgA in a nonopsonic manner. In addition, some of these proteins bind complement regulatory proteins such as C4b-binding protein, factor H, factor H-like protein, or CD46, and these interactions may interfere with effective opsonophagocytic killing of GAS by inhibiting complement activation [51,52].

Protein F — Binding of GAS to fibronectin appears to enhance the adherence of GAS to epithelial surfaces. Protein F is a protein with two fibronectin-binding domains [53]. High carbon dioxide concentrations increase the expression of protein F. Thus, protein F, and perhaps other fibronectin-binding proteins, might play an important role in the adhesion of GAS to mucosal or skin surfaces [54].

Other factors

Streptococcal inhibitor of complement — Streptococcal inhibitor of complement (SIC) is an extracellular protein of 305 amino acid residues that inactivates the complement membrane attack complex [55]. SIC is found only in M types 1 and 57. The sic gene is located in the mga regulon of M type 1 GAS, directly adjacent to the emm gene.

Via production of SIC, GAS can evade destruction by the membrane attack complex (C5 to C9) generated by either the alternative or classical complement pathway. SIC can also inhibit the antibacterial activity of chemokines, which is part of innate immunity against GAS [56]. (See 'Role in innate immunity' below.)

Opacity factor — The opacity factor (OF) has been found largely in GAS but has also been detected in group G and C streptococci, Streptococcus dysgalactiae, and Streptococcus equisimilis [57]. OF is a type-specific lipoprotein lipase whose role in pathogenesis is unknown. There is evidence to suggest that the presence of OF correlates with the arrangement of specific emm genes. Specifically, OF is associated with M types that are largely skin strains [58].

PATHOGENIC MECHANISMS

Anti-phagocytic properties — M protein contributes to invasiveness by impeding phagocytosis of streptococci by human polymorphonuclear leukocytes [1]. Conversely, type-specific antibody against the M protein enhances phagocytosis [1]. Specific antibodies induced by infection with a particular M type confers resistance to challenge with GAS of that M type [1]. It has been shown that GAS protease cleaves the terminal portion of the M protein, rendering the organism more susceptible to phagocytosis by normal serum but more resistant to phagocytosis in the presence of type-specific antibody [59].

While M-1 and M-3 strains have been the most frequent isolates from cases of streptococcal toxic shock syndrome (TSS), many other M types have also been recovered from such cases, including some nontypeable strains. M types 1 and 3 are also commonly isolated from asymptomatic carriers and patients with pharyngitis or mild scarlet fever [40,60].

Mechanisms of fever induction — Pyrogenic exotoxins induce fever and contribute to shock by lowering the threshold to exogenous endotoxin [61]. Streptococcal pyrogenic exotoxins induce human mononuclear cells to synthesize tumor necrosis factor (TNF)-alpha, interleukin (IL)-1, and IL-6 [62-65], suggesting that TNF could mediate the fever, shock, and organ failure observed in patients with streptococcal TSS [39]. SpeA, SpeC, and SSA have been associated with outbreaks of scarlet fever since 2011 in the United Kingdom, Hong Kong, and mainland China [41,66,67].

Cytokine induction — There is strong evidence suggesting that streptococcal pyrogenic exotoxins SpeA, SpeB, and SpeC act as superantigens and stimulate T cell responses through their ability to bind to both the class II MHC complex of antigen-presenting cells and the V beta region of the T cell receptor [68]. The net effect is induction of T cell proliferation (via an IL-2 mechanism) with concomitant production of cytokines (eg, IL-1, TNF-alpha, TNF-beta, IL-6, interferon-gamma) that mediate shock and tissue injury. SpeA has been demonstrated to induce both TNF-alpha and TNF-beta from mixed cultures of monocytes and lymphocytes [69], supporting the role of lymphokines (TNF-alpha) in shock associated with strains producing SpeA. (See 'Pyrogenic exotoxins' above.)

A digest of M protein type 5 can also stimulate T cell responses by this mechanism [70]. Quantitation of such V beta T cell subsets in patients with acute streptococcal TSS demonstrated deletion rather than expansion, suggesting that perhaps the lifespan of the expanded subset was shortened by a process of apoptosis [71]. In addition, the subsets deleted were not specific for speA, speB, speC, or mitogenic factor, suggesting that perhaps an as yet undefined superantigen may play a role in streptococcal TSS [71]. (See 'Pyrogenic exotoxins' above.)

It has been suggested that host genetic factors play a role in these cytokine responses to streptococcal superantigens; patients with the DRB1*1501/DQB1*0602 leukocyte antigen class II haplotype displayed a muted cytokine response and were less likely to develop severe systemic disease due to the organism [72].

Cytokine induction by other mechanisms may also contribute to the genesis of shock and organ failure:

●Peptidoglycan, lipoteichoic acid [73], and killed bacteria [69,74] are capable of inducing TNF-alpha production by mononuclear cells in vitro [61,74,75]. Exotoxins such as SLO are also potent inducers of TNF-beta and IL-1 beta.

●The cysteine protease SpeB has the ability to cleave pre–IL-1 beta to release IL-1 beta [76].

●SLO and SpeA together have additive effects in the induction of IL-1 beta by human mononuclear cells [69].

Induction of cytokines in vivo is likely the cause of shock associated with streptococcal TSS, and SLO, SpeA, SpeB, and SpeC as well as cell wall components are potent inducers of TNF and IL-1 [42]. In addition, the cysteine protease activity of SpeB may play an important role in pathogenesis by the release of bradykinin from endogenous kininogen and by activating metalloproteases involved in coagulation [77].

Role in innate immunity — Certain cytokines are involved in innate immunity against infection. One such group is the CXC chemokines CXCL9, CXCL10, and CXCL11, as illustrated by the following observations [56]:

●In vitro, all the chemokines have antibacterial activity against GAS.

●Tonsillar fluid from patients with GAS pharyngitis contains high amounts of CXCL9, and inflamed pharyngeal epithelium produce large amounts of CXCL9 in the presence of GAS. Inhibition of CXCL9 expression reduces antibacterial activity at the surface of inflamed pharyngeal cells.

●GAS of the M1 serotype secretes a protein (streptococcal inhibitor of complement) that reduces the antibacterial activity of the chemokines. (See 'Streptococcal inhibitor of complement' above.)

DISEASE MECHANISMS OF POSTINFECTIOUS SEQUELAE — The pathogenesis of acute rheumatic fever and poststreptococcal glomerulonephritis is discussed elsewhere. (See "Acute rheumatic fever: Epidemiology and pathogenesis" and "Poststreptococcal glomerulonephritis".)

SUMMARY

●Group A Streptococcus (GAS) is a gram-positive coccus that causes an array of infections involving the respiratory tract and soft tissues ranging in severity from mild to severe. The pathogenic mechanisms underlying these infections are not completely understood, largely because each is the culmination of highly complex interactions between the human host defense mechanisms and specific virulence factors of the organism. (See 'Introduction' above.)

●The Lancefield classification divides streptococci into types A through O based upon serologic reactions of specific antibodies with acid-extractable carbohydrate antigens of cell wall material. Rapid sequencing of the gene encoding M protein provides a rapid, definitive means of subdividing GAS (Streptococcus pyogenes) isolates into emm types. (See 'Classification schemes' above.)

●GAS have been subdivided based upon serotyping of surface-expressed M and T antigens. M-typing using specific antisera has been largely supplanted by emm-typing, that is, sequencing the variable region of the emm gene. More than 200 emm types have been described. M protein protects the organism against opsonophagocytic killing by polymorphonuclear leukocytes in the absence of type-specific antibody. Various vaccine strategies target M protein with differing success. (See 'M and M-like proteins' above.)

●Streptolysin O is a cytolysin released by the bacteria; the host mounts antibody against this antigen that can be used as a diagnostic tool. Nonsuppurative complications such as rheumatic fever and poststreptococcal glomerulonephritis generally develop during the second or third week of illness. Therefore, it may be useful to perform one titer when a nonsuppurative complication is first suspected and subsequently repeat the titer approximately two weeks later. (See 'Streptolysin O' above.)

●Streptococcal pyrogenic exotoxins are virulence factors with capacity to induce lymphocyte blastogenesis, potentiate endotoxin-induced shock, induce fever, suppress antibody synthesis, and act as superantigens. The genes are transmitted by bacteriophage. (See 'Pyrogenic exotoxins' above.)