Pathogenicity Factors in Group C and G Streptococci

ABSTRACT

Initially recognized zoonoses, streptococci belonging to Lancefield group C (GCS) and G (GGS) were subsequently recognised as human pathogens causing a diverse range of symptoms, from asymptomatic carriage to life threatening diseases. Their taxonomy has changed during the last decade. Asymptomatic carriage is <4% amongst the human population and invasive infections are often in association with chronic diseases such as diabetes, cardiovascular diseases or chronic skin infections. Other clinical manifestations include acute pharyngitis, pneumonia, endocarditis, bacteraemia and toxic-shock syndrome. Post streptococcal sequalae such as rheumatic fever and acute glomerulonephritis have also been described but mainly in developed countries and amongst specific populations. Putative virulence determinants for these organisms include adhesins, toxins, and other factors that are essential for dissemination in human tissues and for interference with the host immune responses. High nucleotide similarities among virulence genes and their association with mobile genetic elements supports the hypothesis of extensive horizontal gene transfer events between the various pyogenic streptococcal species belonging to Lancefield groups A, C and G. A better understanding of the mechanisms of pathogenesis should be apparent by whole-genome sequencing, and this would result in more effective clinical strategies for the pyogenic group in general.

INTRODUCTION

The pyogenic streptococci of Lancefield groups C and G were initially recognized as a cause of animal infections long before they were even considered as agents of human disease, even though they are widely distributed in animals and humans. They comprise a heterogeneous complex of streptococcal species that act as causative agents of a spectrum of diseases ranging from mild pharyngitis to skin infection to life-threatening systemic infections associated with high mortality rates. In this article we provide an overview of the various group C and group G streptococcal species, the diseases they cause, and the major pathogenicity factors that contribute to their virulence (Table 1).

TABLE 1.

Pathogenicity factors of group C and G streptococci

Pathogenenicity factors	Organism	Reference(s)
Fibronectin binding proteins
FnbA	S. dysgalactiae	59
FnbB	S. dysgalactiae	60
GfbA	S. dysgalactiae	64
FNZ/FNE	S. zooepidemicus, S. equi	61, 65
FNZ2/FNEB	S. zooepidemicus, S. equi	66, 69
SFS	S. zooepidemicus, S. equi	70
M-like proteins	S. dysgalactiae S. equisimilis	91, 105
FOG	S. dysgalactiae	93, 94
DemA	S. dysgalactiae	105
SzM/SeM	S. zooepidemicus, S. equi	76, 95–97, 101
SzPSe/SzP	S. zooepidemicus, S. equi	102
Se18.9	S. equi	76, 100
ScM	S. canis	104
Others
C5a peptidase	S. dysgalactiae	108
SeCEP/SzoCEP	S. zooepidemicus, S. equi	112
Immunoglobulin binding proteins
Protein G	S. dysgalactiae	116
MIG	S. dysgalactiae	120
MAG	S. dysgalactiae	122
ZAG	S. zooepidemicus	123
IgG-endopeptidases
IdeE/IdeE2	S. equi, S. zooepidemicus	129
IdeE2/IdeZ2	S. equi, S. zooepidemicus	130
IdeP	S. phocae	133
Toxins
Streptokinase	S. dysgalactiae, S. equisimilis	134
Streptolysin O	S. dysgalactiae, S. equisimilis	145
Streptolysin S	S. dysgalactiae, S. equisimilisS. equi, S. zooepidemicus	152, 154
Superantigens
SpeGdys/Spegg/SpeG	S. equisimilis, S. canis	158, 159
SpeH	S. equi	167
SpeI	S. equi	167
SpeK	S. equi, S. zooepidemicus, S. equismilis	158, 168, 169
SpeL	S. equi, S. zooepidemicus, S. equismilis	158, 168, 169
SpeM	S. equisimilis	173
ssa	S. equisimilis	173
SDM/SpeM	S. dysgalactiae	171
SpeA	S. equismilis	170
SpeC	S. equismilis	170
SzeN	S. zooepidemicus	170
SzeP	S. zooepidemicus	170
SzeF	S. zooepidemicus	170

TAXONOMY AND IDENTIFICATION

Taxonomic classification of group C streptococci (GCS) and group G streptococci (GGS) has always proven to be a complex issue. However, extensive taxonomic studies over the past few years have distinguished most of the veterinary pathogens belonging to Lancefield groups C and G from the human pathogens. Previously, GCS and GGS were divided into the following species: Streptococcus dysgalactiae subsp. equisimilis (only GCS, human pathogen), Streptococcus dysgalactiae subsp. dysgalactiae (GCS, animal pathogen), Streptococcus equi (animal GCS), Streptococcus zooepidemicus (animal and human GCS), Streptococcus canis (animal GGS), the Streptococcus anginosus group, and Streptococcus phocae (1). GCS and GGS of human origin are now considered to constitute a single subspecies, S. dysgalactiae subsp. equisimilis. Therefore, the current taxonomy characterizes the species as follows: S. dysgalactiae is divided into the subspecies S. dysgalactiae subsp. equisimilis and S. dysgalactiae subsp. dysgalactiae (hereafter referred to in this article as S. equisimilis and S. dysgalactiae). S. equi is divided into the subspecies S. equi subsp. equi, S. equi subsp. zooepidemicus, and S. equi subsp. ruminatorum (hereafter referred to as S. equi, S. zooepidemicus, and S. ruminatorum). However, it has been suggested that S. equi may simply be regarded as a subclone of S. zooepidemicus (2). On the basis of genetic evidence, S. dysgalactiae, S. equi, and S. canis are more closely related to each other than to the S. anginosus group and constitute species with the “large-colony” colony phenotype. S. phocae is a new species expressing the group C antigen thus far only isolated from seals (3, 4). Some of these species may also contain strains which express the Lancefield group A or group F antigens (Table 2).

TABLE 2.

Species of Lancefield group C and G streptococci^a

Species	Known sources	Lancefield group
S. dysgalactiae subsp. dysgalactiae	Animals	C
S. dygalactiae subsp. equisimilis	Humans, animals (rare)	C, G
S. equi subsp. equi	Animals	C
S. canis	Animals, humans (rare)	G
S. equi subsp. zooepidemicus	Animals, humans	C
S. phocae	Animals (seal)	C
S. anginosus group	Humans	A, C, F, G

^a From reference 1.

HUMAN DISEASE AND DIAGNOSIS

GCS and GGS can cause a wide range of diseases from mild to severe, among both human and animal populations. GCS and GGS generally colonize the human respiratory, gastrointestinal, and genitourinary tracts, with an estimated <4% asymptomatic pharyngeal carriage rate in adults (5–7). Human invasive infections caused by GCS and GGS have been associated with underlying conditions, such as diabetes, cardiovascular diseases, and chronic skin conditions (8–10). While S. dysgalactiae is considered an animal pathogen, S. equisimilis is almost exclusively a human pathogen, with increasing prevalence and overlaps in clinical manifestations with group A Streptococcus (GAS). S. equisimilis infections include acute pharyngitis, pneumonia, endocarditis, cellulitis, peritonitis, septic arthritis, bacteremia, and toxic shock syndrome (11–23). Like GAS, S. equisimilis has also been linked to the poststreptococcal sequelae rheumatic heart disease, and in high endemic areas of rheumatic fever, carriage rates of GCS/GGS have been found to be higher than those of GAS (24, 25).

Infections with S. zooepidemicus and S. canis, normally considered zoonotic species, have also been reported in humans (26–29), including severe infective endocarditis (30).

Generally, person to person transmission of GCS/GGS occurs via respiratory droplets or skin contacts, but other zoonotic vehicles of transmission such as unpasteurized milk products are also possible (26, 27, 31).

S. dysgalactiae predominantly resides in domestic animals such as cattle, sheep, cats, and dogs that are either healthy carriers of the bacterium or go on to develop diseases such as pneumonia, arthritis, septicemia, and abscesses, particularly bovine mastitis. S. equi, a pathogen primarily restricted to horses and donkeys, causes strangles, which is a highly contagious disease characterized by purulent discharges from the respiratory tract and the development of abscesses (32). This organism is not considered to be part of the normal flora, because of its close association with disease. In contrast, S. zooepidemicus is an opportunistic commensal that colonizes mucosal surfaces and causes rhinopharyngitis, pneumonia (33), endometritis, neonatal septicemia, and wound infections in horses (34) as well as disease in other domestic animals such as cattle, sheep, pigs, and chickens. Although uncommon, S. zooepidemicus is also responsible for a range of zoonotic infections in humans (26, 27).

The small-colony phenotype of group C and G streptococci is expressed by the S. anginosus group, formerly known as Streptococcus milleri. The group contains three species, S. anginosus, Streptococcus constellatus, and Streptococcus intermedius (1), and forms a distinct subclade by phylogenomic analysis (35).

Strains belonging to the S. anginosus group express group antigens F, C, A, and G or no antigen. These strains are also likely to be non-beta-hemolytic. Members of these species are recognized as common commensal organisms of the human oral cavity, gastrointestinal tract, and genitourinary tract. They are also associated with abscess formation in the mouth and other body sites (36, 37) as well as pharyngitis (38) and endocarditis (39, 40). It has also been reported that S. anginosus, along with Streptococcus mitis and Treponema denticola, can be isolated from esophageal cancer tissue and by initiating inflammation in the cancerous tissue has a role in the development or progression of these cancers (41, 42).

Several assays are available to detect, classify, and genetically describe GCS and GGS. Routine microbiological diagnosis follows the same guidelines that are in use for the identification of the other beta-hemolytic streptococci (43). The beta-hemolytic streptococci isolates are divided into large and small colonies forming groups based on the growth on sheep blood agar: the large-colony-forming group is “pyogenic.” The small-colony-forming species make up the anginosus group and are not referred to as GCS or GGS, even though they may cross-react with C or G sera. Lancefield agglutination tests are still used to group beta-hemolytic streptococci into the Lancefield groups. Fermentation tests are used to classify GCS as S. equisimilis or S. zooepidemicus and are based on a spectrum of biochemical tests which are now available commercially (44). More recently, matrix-assisted laser desorption ionization–time of flight mass spectrometry has been described as a rapid alternative for the identification of streptococci (45). In addition, multilocus sequence analysis of seven housekeeping genes (46) and other sequence-based assays are performed to analyze the streptococcal phylogenies. In particular, sequencing of the emm gene, which encodes the hyper-variable surface protein M, and 16s rRNA genes is undertaken for phylogenetic analysis of these streptococci based on sequence similarities (47, 48).

Next-generation sequencing is increasingly being used for genetic characterization, phylogenetic analyses, virulence and pathogenicity profiles, and antibiotic resistance analyses (22, 49). S. equisimilis and Streptococcus pyogenes (or GAS) share a high genetic similarity (∼72%) and similar virulence genes, suggesting a common evolutionary origin and genetic exchange (22, 25, 50).

MECHANISMS FOR ADHERENCE

S. equisimilis causes a spectrum of disease in humans similar to that caused by S. pyogenes, and molecular studies have demonstrated virulence determinants that are almost identical. Putative virulence determinants of S. equisimilis include adhesins, toxins, and factors that are essential for dissemination in human tissues and for interference with the host immune responses.

Adhesion of microorganisms to host tissues represents a critical phase in the development of infection. It is therefore unsurprising that microorganisms have evolved dedicated mechanisms for attachment and adherence to extracellular matrix components of the host (51, 52). Of these components, the high-molecular-weight glycoprotein fibronectin appears to be the major attachment target of Gram-positive cocci, including GGS and GCS. Fibronectin itself is responsible for substrate adhesion of eukaryotic cells via specific cell surface factors of the integrin family. It also specifically interacts with other matrix components, such as collagen, fibrin, and sulfated glycosaminoglycans, demonstrating that this molecule fulfills multifunctional roles within the extracellular network (53). Being present in the extracellular matrix of most tissues, as well as in plasma and other body fluids in its soluble form, fibronectin represents an exquisite target for bacteria to exploit the cell attachment properties of this molecule by linking the pathogen to specific target cells. Epithelial cells of the human upper respiratory tract are bathed in secretions containing fibronectin in its soluble form. Once bound to the bacterial surface, it enables the pathogen to attach and subsequently colonize the primary site of infection.

Fibronectin binding proteins (FBPs) were first identified in Staphylococcus aureus and S. pyogenes (54), and approximately 11 have been defined in S. pyogenes, highlighting their importance for this pathogen (55). Binding by these bacteria to eukaryotic cells via fibronectin is also an important preliminary event prior to the invasion of these cells (56–58). The FBPs from streptococci and staphylococci share a common architecture, with a putative signal sequence at the N terminus and a wall- and membrane-spanning region. The major fibronectin binding domains are located within the C-terminal part of the proteins and are composed of 3 to 5 repetitive units that consist of 35 to 37 amino acid residues and bind to the 29-kDa N-terminal fragment of fibronectin. Further binding studies of PrtF/SfbI from S. pyogenes and FnBPA in S. aureus also indicate the presence of a secondary fibronectin binding site upstream of the repeat regions (54).

The first FBPs to be identified in group C and G streptococci were FnbA and FnbB from S. dysgalactiae (59). FnbA can opacify serum and bind fibrinogen, like the serum opacity factor of S. pyogenes (60). Both FnbA and FnbB bind fibronectin through C-terminal repeat regions (51).

Three FBPs have been defined in S. zooepidemicus: FNZ, FNZ2, and SFS. Although the modular organization of FNZ, an S. zooepidemicus FBP, is similar to those from other species, homology at the amino acid level is weak (61). Fibronectin binding is mediated through a repeat region and a second upstream region containing the amino acid motif LAGESGET. This motif is also present in the secondary binding domain of SfbI (62), where it acts independently of the repeat region in binding to fibronectin (63). This domain is also present in GfbA, the homologue of SfbI in S. dysgalactiae (64). fne is the homologous gene to fnz in S. equi (65). However, FNE is unique in that a single nucleotide deletion in the fne gene has resulted in the truncation of the protein, so with the loss of the cell-wall binding motif, it is secreted into the surrounding environment. The truncated FNE also lacks the classic fibronectin binding repeats found in the C-terminal region of FNZ but is able to bind fibronectin through the amino-half domain, also found in FNZ. Consequently, fibronectin binding at the bacterial surface is much lower for S. equi than S. zooepidemicus, although both have an additional cell wall-anchored FBP: FNEB and FNZ2 (66). FNE also binds to collagen, and the protein affects the interstitial fluid pressure in vivo (67). The molecular structure of the interaction between FNE and fibronectin has been reported (68). FNZ2 also has both collagen and fibronectin binding properties (69). SFS of both S. equi and S. zooepidemicus strains contains a signal peptide but no cell wall binding motifs or traditional fibronectin binding motifs. SFS inhibits binding between fibronectin and collagen. If SFS were to bind fibronectin attached to the bacterial surface through FNEB/FNZ/FNZ2, this might inhibit binding of fibronectin to collagen, which could have several physiological consequences (70). The molecular interaction between SFS and fibronectin has been reported (71).

Some of the FBPs and collagen binding proteins identified in S. pyogenes are chromosomally located within the fibronectin and collagen binding proteins and T antigen encoding (FCT) locus. There are at least nine types of FCT regions in S. pyogenes that vary by gene content and order, but within this region are the genes required to make the pili. The pilus has been shown to contribute to the formation of biofilm and mediate adherence to host cells (72–75). Whole-genome sequence analysis of group C and G streptococci has also identified potential FCT regions. One putative pilin locus, FimI, exists within the genome of S. equi strain 4047 (76). Although no pili structures have been identified for S. equi, the proteins of FimI are expressed during growth and have the potential to affect adherence in vivo (77). The gene encoding the collagen binding protein CNE of S. equi (78) is present in this FimI locus. A homologous locus was also found in S. zooepidemicus isolates, as well as an additional one or two potential pili loci (76, 79). Two FCT regions were also found in S. equisimilis, which share high levels of genetic identity to FCT regions from different S. pyogenes genotypes, suggesting multiple horizontal gene transfer events (22, 80).

An alternative way for group C and G streptococci to adhere to host cells is via binding to other extracellular matrix molecules, including fibrinogen, vitronectin, laminin, collagen, and plasminogen (81, 82). The M-protein-like fibrinogen binding protein of G streptococci (FOG) can act as an adhesin by binding collagen IV (83) and has been shown to also bind collagen I fibrils in dermis in vivo (84).

Vitronectin is a multifunctional serum protein that affects the humoral immune system by binding to and inhibiting the complement membrane attack complex (85) and is also a major matrix-associated adhesive glycoprotein that regulates blood coagulation. The ability of group C and G streptococci specifically to interact with vitronectin (86) and mediate the adherence to both epithelial and endothelial cells (87, 88) was demonstrated some time ago. However, a specific vitronectin binding protein has never been identified in streptococci.

ANTIPHAGOCYTIC FACTORS

A major requirement of pathogenic streptococci is to be able to resist phagocytosis. The streptococcal M protein, first identified and characterized in S. pyogenes, is the major antiphagocytic factor (89). It is a multidomain surface-exposed molecule that forms a coiled-coil secondary structure with significant irregularities that in the B-repeat region are essential for the fibrinogen binding properties (90). Binding of complement factors, fibrinogen, and inhibition of C3b deposition on the bacterial surface are mechanisms by which the M-protein can inhibit opsonization of the organism by the alternative complement pathway, thus evading the host’s nonspecific immune defense mechanism.

Several M-like proteins have been identified in group C and G streptococci. Protein MG1, the first group G streptococcal M-like protein characterized on the molecular level (91), exhibits typical structural and biological features of M proteins, such as coiled-coil structure and the ability to generate type-specific opsonizing antibodies. Protein MG1 shares highly homologous sequences with the C-terminal repeat region of class I M proteins, which are frequently associated with rheumatic fever. M proteins of GGS are also responsible for conferring resistance to phagocytosis (92).

In human strains of GGS, FOG is critical for bacterial pathogenesis through its antiphagocytic action of binding fibrinogen (93). It also has the ability to bind IgG subclasses IgG1 and IgG2, although FOG-bound IgG1 can trigger the complement cascade through C1/C1q activation. It is unclear whether this is detrimental to the bacteria, because FOG remains protective against phagocytosis (94).

The antiphagocytic protein SeM (FgBP) appears to be the predominant M-like protein on the surface of S. equi, capable of binding fibrinogen and IgG4 and IgG7 subclasses (95–97). SeM confers resistance to phagocytosis, but this also requires the presence of the hyaluronic acid capsule (98). Like the M protein of S. pyogenes, SeM is variable at the N-terminal region, suggesting selective pressure (99). Interestingly, this does not appear to be the case for two other M-like proteins expressed by S. equi: Se18.9 and SzPSe (99). The antiphagocytic protein Se18.9 is commonly found in strains of S. equi, but rarely in strains of S. zooepidemicus (76), and acts by binding to fibrinogen and the complement regulatory protein, factor H (100). SzP proteins are also antiphagocytic and are expressed by both S. equi and S. zooepidemicus. Whole-genome sequence analysis of S. zooepidemicus strain H70 identified a potential M-protein homologue in SzM that resembles SeM in its secondary structure, which consists of a predicted C-terminal coiled-coil (76, 101). The N-terminal of SeM is unique to S. equi and binds fibrinogen (97, 102). SzM can also bind fibrinogen but cannot bind IgG (103). The M protein of S. canis specifically binds the Fc region of IgG in a nonopsonic manner (104).

DemA is an S. dysgalactiae protein with homology to FgBP identified by screening of a phagemid expression library (105). The mature DemA protein is 54 kDa in size, contains a signal sequence and a cell wall binding domain, and is predicted to have a coiled-coil secondary structure. DemA shows greatest homology to the FgBP at the C-terminal end in a region that does not participate in fibrinogen binding. The amino acid motif VSKDLADKL present with the repeat units of both DemA and FgBP suggests that the sequence may have an important biological function. DemA is able to bind IgG from various animal sources, reminiscent of type IIa Ig-binding proteins of S. pyogenes in a domain distinct from the fibrinogen binding domain. Nucleotide sequencing of the demA locus identified an open reading frame upstream of demA that is homologous to mga, a positive regulator of M-protein expression in S. pyogenes (105).

C5a peptidase of S. pyogenes (ScpA) is well known to specifically cleave and inactivate the chemoattractant C5a, but it has also recently been shown to cleave the complement factor C3 and the chemoattractant C3a (106), expanding its impact on immune evasion. ScpA has also been identified as an adhesion factor enabling bacteria to adhere to endothelial and epithelial cells (106). Homologues of ScpA have been identified in group B streptococci (ScpB) and group C and G streptococci isolated from humans but not from GGS from animals (107, 108). Both S. equi and S. zooepidemicus also carry homologous C5a peptidase genes (scpE and scpZ, respectively). However, the C5a peptidase gene in S. equi is reported to be a pseudogene (76). Unlike in S. pyogenes, where scpA is located in the mga regulon that includes the emm-gene, the scp genes of group B, C, and G streptococci are associated with a transposon, suggesting that lateral genes transfer between these species (107, 109).

The S. pyogenes cell wall envelope proteinase, SpyCEP, is another protease that contributes to the prevention of phagocytosis through the specific cleavage of interleukin-8 and other chemokines to prevent the activation and migration of neutrophils (110, 111). Similar enzymes have been identified in S. equi (SeCEP) and S. zooepidemicus (SzoCEP) which share 98% identity to each other and 59% identity to SpyCEP (108). SeCEP has been shown to cleave both human and equine IL-8, and vaccination with a recombinant portion of SpyCEP prevented bacterial dissemination in a murine model of S. equi invasive infection, suggesting an important contribution to disease (112).

IMMUNOGLOBULIN BINDING AND INACTIVATING PROTEINS

Streptococcal protein G is a surface molecule associated with the majority of group C and G streptococcal isolates of human origin. Protein G interactions with immunoglobulins and other host proteins have been the subject of detailed reviews (113–115). Protein G is defined as a type III IgG Fc receptor and interacts with a wide species range of immunoglobulins, as well as human serum albumin, kininogen, and α₂-macroglobulin. Protein G exhibits a modular structure in which the binding sites for IgG are located in the C-terminal repeat region (114–116). The central A/B-repeat region constitutes the binding domain for serum albumin, and the N-terminal E region is responsible for interacting with the native form of α₂-macroglobulin (82). In contrast to human pathogenic strains of GGS that exclusively bind to the native (slow) form of α₂-macroglobulin via protein G, animal-derived isolates of bovine and equine origin bind the proteinase-complexed (fast) form of the molecule. The B1 domain of protein G involved in immunoglobulin binding consists of an α-helix and four β-strand sheets. This domain has been used as a model structure in numerous biochemical studies examining protein folding, protein interaction, and synthetic protein design (117–119).

Two protein G-related proteins, MIG and MAG, from two mastitis-causing S. dysgalactiae strains (120) and MAG (121, 122) and ZAG (123) from S. zooepidemicus have been characterized on the molecular level. Like protein G, MAG and ZAG exhibit serum albumin and type III Fc receptor activity, whereas the protein MIG lacks albumin-binding activity. However, MAG is able to bind to immunoglobulins from a greater number of animal sources than protein G. MAG, ZAG, and MIG also bind to the fast form of α₂-macroglobulin. The α₂-macroglobulin binding region in MIG is not homologous to those of protein G, MAG, ZAG, and GRAB. In phagocytosis assays a MIG isogenic mutant strain of S. dysgalactiae was not as resistant to opsonization by bovine neutrophils as the parental strain (124). MIG has also been shown to bind to bovine immunoglobulin A (125) and can inhibit bacterial internalization into host cells (126).

As well as IgG binding proteins, group C and G streptococci can express variants of the S. pyogenes IdeS and EndoS, which are IgG-degrading enzymes. While IdeS cleaves the hinge region of IgG (127), EndoS removes core IgG glycans (128). Both S. equi and S. zooepidemicus express two forms of IgG-endopeptidases, IdeE/IdeE2 and IdeZ/IdeZ2, respectively, that are capable of cleaving IgG from several mammalian species (129, 130), although IdeE2/IdeZ2 cleave horse IgG with greater efficiency than IdeE/IdeZ (130). EndoS homologues have also been identified in S. equi (EndoSe) and S. zooepidemicus (EndoSz), and they share 86 to 89% sequence identity to each other and 70% identity to EndoS of S. pyogenes (131). EndoSd has also been found in S. equisimilis with IgG hydrolyzing activity (132), and IdeP was identified in the group C S. phocae subsp. phocae that affects marine animals, although its function has not been confirmed (133).

ENZYMES AND TOXINS

After colonization, adherence, and evasion of host immune responses, the dissemination of pathogenic streptococci in tissues is regarded to be an important step for the onset and development of an invasive disease. One of the factors involved in this process is streptokinase, a protein found in groups C, G, and A streptococci (134). The formation of a streptokinase/plasminogen complex results in the exposure of the plasminogen active site, which then catalyzes the conversion of other plasminogen molecules into plasmin (135). Plasmin is a key serine protease in the fibrinolytic system that is able to break down tissue barriers, thereby enabling the dissemination of streptococci. M proteins coordinately interact with the secreted streptokinase by binding either fibrinogen (136) or plasminogen (137). A high level of variation exists within the streptokinase gene, ska, and alleles of ska have been associated with tissue tropisms and differing levels of plasminogen activation (138). There is some overlap of alleles between human strains of group C and G streptococci and GAS (139). Streptokinases have been isolated from both human and animal group C and G isolates and have specificity for the plasminogens of their respective hosts (140). Thus, streptokinases from human group A, C, and G streptococci isolates have greater homology to each other than to animal isolates of group C and G streptococci. A study by Caballero et al. (58) found the amino acid homology between streptokinases from human and animal S. equisimilis isolates to be only 35%. Homology between the streptokinase from S. equisimilis of equine and porcine origins was only 21%. The ability of streptokinases to cleave plasminogen from specific species may therefore be a critical factor in determining the host range of individual streptococcal strains (141). Invasive human strains of S. equisimilis have been found to express higher levels of streptokinase, and this can drive virulence in a murine model of invasive necrotizing fasciitis, suggesting an important role for streptokinase in disease (142, 143).

Streptolysin O (SLO) is the prototype of a family of thiol-activated cytolysins produced by the genus Streptococcus and by other Gram-positive bacteria, including Bacillus, Clostridium, and Listeria species (144). The genes coding for SLO of group C and G streptococci (S. dysgalactiae) are almost identical to that of S. pyogenes (145). SLO homologues have not been described in S. equi. SLO is able to disrupt the cytoplasmic membrane of several eukaryotic cell types that include erythrocytes, leukocytes, macrophages, platelets, and epithelial cells. Separate from its pore-forming activity, SLO also acts to translocate NADase into host cells (146), which contributes to cytotoxicity by depleting energy stores. Other functions of these two streptococcal toxins include limiting neutrophil responsiveness and potentiating bacterial survival, replication, and persistence inside keratinocytes, epithelial cells, and macrophages, which may protect GAS from the immune response and antimicrobial therapy (147–151).

Streptolysin S (SLS) is another cytolysin secreted by groups A, C, and G streptococci, including the animal-pathogenic S. equi species (152). SLS belongs to a distinct group of hemolytic toxins that are characterized by their resistance to oxidation and sensitivity to trypan blue. SLS activity results in damage to membranes of various cell types as well as subcellular organelles (153). In contrast to the 57-kDa SLO, SLS is a small, 57-amino acid protein. The gene encoding SLS, sagA, is part of a locus containing nine open reading frames which contain significant homology with genes from bacteriocin loci (154, 155). In S. equisimilis, SLS expression is under the control of both the covRS and fasCAX two-component regulatory systems, which in S. pyogenes have been shown to be involved in the regulation of multiple virulence factors (156). In a mouse infection model, S. equisimilis expressing SLS proliferates and induces necrotic lesions at the site of infection, whereas SLS-negative strains do not, suggesting that SLS is an important factor in the development of necrotizing fasciitis (154).

Streptococcal pyogenic toxins (spes) are superantigen genes capable of cross-linking the major histocompatibility complex class II on antigen-presenting cells to the T cell receptor, leading to proliferation of T cells and substantial release of inflammatory cytokines. Eleven superantigen genes have been identified in S. pyogenes, and they are thought to drive the development of scarlet fever and toxic shock syndrome, the latter contributing to high mortality rates following necrotizing fasciitis. Recent work has also identified a role for superantigens in upper respiratory tract infection (157). Homologues of the GAS superantigens have been identified in group C and G streptococci (158). The most commonly found superantigen in S. equisimilis is speG (spegg or speG^dys) (159, 160), and genomic analysis indicates that S. pyogenes and S. equisimilis speG genes are orthologues that are direct descendants from a common ancestor gene (161). The gene for speG can also be found in S. canis (158). The role of S. equisimilis speG in humans is unclear because the presence of this gene does not correlate with disease severity and it does not confer mitogenic activity toward human mononuclear cells, although it can stimulate bovine T cells (160, 162, 163). The streptococcal superantigens speA, speC, speJ, speK, speH, speL, speM, and ssa are found infrequently in S. equisimilis and other human group C and G streptococci (160, 164–166).

Homologues of S. pyogenes superantigens speH and speI can be carried by strains of S. equi (167), and homologues of speK and speL can be carried by both S. equi and S. zooepidemicus, although because of confusion with the nomenclature, speK homologues were originally termed speL_se/seeL/szeL, and speL homologues were termed speM_se/seem/szeM (160, 167–169). Three additional superantigen genes, termed szeN, szeP, and szeF, have been identified in S. zooepidemicus only and are capable of stimulating equine peripheral blood mononuclear cells (170). S. dysgalactiae-derived mitogen shows homology to speM of S. pyogenes and can stimulate human mononuclear cells (160, 171). The majority of the superantigens found in S. pyogenes and group C and G streptococci are associated with prophages which may drive the lateral transfer of these virulence factors between streptococcal species (76, 172, 173).

CONCLUSIONS

Many recent studies have highlighted the increasing number of systemic infections caused by S. equisimilis, particularly among immunocompromised individuals and specific populations, and this therefore suggests that this species will gain even more clinical importance in the future. High nucleotide similarities among virulence genes and their association with mobile genetic elements support the hypothesis of extensive horizontal gene transfer events between streptococcal species of the pyogenic group. A better understanding of the mechanisms of pathogenesis will hopefully be revealed by whole-genome sequencing, and this may result in more effective clinical strategies for the pyogenic group of streptococci in generally.