Group G Streptococcal IgG Binding Molecules FOG and Protein G Have Different Impacts on Opsonization by C1q*

Recent epidemiological data on diseases caused by β-hemolytic streptococci belonging to Lancefield group C and G (GCS, GGS) underline that they are an emerging threat to human health. Among various virulence factors expressed by GCS and GGS isolates from human infections, M and M-like proteins are considered important because of their anti-phagocytic activity. In addition, protein G has been implicated in the accumulation of IgG on the bacterial surface through non-immune binding. The function of this interaction, however, is still unknown. Using isogenic mutants lacking protein G or the M-like protein FOG (group G streptococci), respectively, we could show that FOG contributes substantially to IgG binding. A detailed characterization of the interaction between IgG and FOG revealed its ability to bind the Fc region of human IgG and its binding to the subclasses IgG1, IgG2, and IgG4. FOG was also found to bind IgG of several animal species. Surface plasmon resonance measurements indicate a high affinity to human IgG with a dissociation constant of 2.4 pm. The binding site was localized in a central motif of FOG. It has long been speculated about anti-opsonic functions of streptococcal Fc-binding proteins. The presented data for the first time provide evidence and, furthermore, indicate functional differences between protein G and FOG. By obstructing the interaction between IgG and C1q, protein G prevented recognition by the classical pathway of the complement system. In contrast, IgG that was bound to FOG remained capable of binding C1q, an effect that may have important consequences in the pathogenesis of GGS infections.

reports from different parts of the world are accumulating demonstrating that they are emerging as important human pathogens (1)(2)(3)(4). Individual processes contributing to the immune defense against GGS are still elusive, and the knowledge about how GGS undermine and obstruct the actions of the immune system is fragmentary. Before the development of a specific immune response, the innate immune system constitutes a first line of defense against intruding bacteria. Part of the innate immune defense is the complement system that has two major functions. One is to kill bacteria by forming lytic complexes on the bacterial membrane, referred to as membrane attack complex. The other one is to facilitate the actions of the cellular immune response. The complement cascade produces anaphylatoxins to attract phagocytes; other complement reactions opsonize the bacteria, promoting extracellular bactericidal actions and phagocytosis (5)(6)(7)(8).
Bacteria have evolved diverse protective mechanisms to prevent the actions of the complement system. The membrane attack complex is thought to be inefficient in killing Grampositive bacteria due to their resistant cell wall (9). C5a-peptidase of Streptococcus pyogenes, which has a homologue in GGS, cleaves and thereby inactivates a potent anaphylatoxin (10,11). M proteins of S. pyogenes prevent opsonization by C4b by recruiting plasma proteins like fibrinogen or C4b-binding protein. A recent study suggests that fibrinogen bound to M protein interferes with the deposition of C4b2a, the C3 convertase of the classical pathway, on the bacterial surface (12). A recently defined fibrinogen-binding M-like protein of GGS referred to as FOG was shown to exert anti-phagocytic activity on neutrophils by forming fibrinogen aggregates (13). The complement cascade can be initiated by three different pathways; that is, the classical, the lectin, or the alternative pathway. A triggering step of the classical pathway is the binding of factor C1q to aggregated or antigen-bound IgG. Factor C1q together with the serine proteases C1r and C1s forms the C1 complex. The interaction with IgG induces conformational changes in C1q that lead to the activation of the serine proteases of the C1 complex. This sets off the classical pathway by cleaving C2 and C4, thus leading to the formation of opsonins and anaphylatoxins (5). Moreover, C1q itself acts as an opsonin, eliciting antibacterial actions of phagocytes (7,8,14).
GCS and GGS accumulate IgG on their surface by non-immune binding. A major protein that is responsible for the high capacity accumulation of IgG on GGS is protein G (15). Here we demonstrate for the first time that FOG also contributes in accumulating human IgG on the GGS surface and present a detailed characterization of the interaction between IgG and FOG. The influence of the IgG accumulations on recognition by human C1q was investigated. The data revealed an antiopsonic function of protein G. In contrast to protein G, FOG did not interfere with the interaction between IgG and C1q. The results suggest opposed functions of the IgG interactions of FOG and of protein G in human infections.

EXPERIMENTAL PROCEDURES
Growth Conditions and Bacterial Strains-Strains G41 and G148 are throat isolates collected at the Department of Clinical Microbiology, Lund University Hospital, Sweden. Strain G45 was collected at the Royal Brisbane Hospital, Australia. G41 and G45 carry protein FOG on their surface, and G148 naturally lacks protein FOG. Streptococci were routinely grown in Todd-Hewitt (Difco) liquid medium in 5% CO 2 at 37°C.
If not indicated otherwise, protein G, human fibrinogen, human IgG, IgG subclasses, and papain-treated IgG fragments were purchased from Sigma, and C1q was from Calbiochem. Radiolabeling of proteins was carried out using Iodobeads (Pierce) following the manufacturer's protocol. Labeled proteins were separated from unincorporated iodine using PD-10 columns (Amersham Biosciences).
Slot Blots-Protein preparations were applied to polyvinylidene difluoride membranes using the Milliblot-D system. Membranes were blocked in PBS containing 1% bovine serum albumin and 0.25% Tween 20 (block buffer) for 3 ϫ 20 min at 22°C and then incubated with radiolabeled FOG (200,000 cpm/ml block buffer) for 3 h. Membranes were washed 3 ϫ 20 min in PBS containing 0.05% Tween 20, and bound ligand was detected using the Fuji Imaging System.
Surface Plasmon Resonance (SPR) Measurements-Protein interactions were studied in a BIAcore 2000 system (BIAcore AB) using 10 mM HEPES, 100 mM NaCl, pH 7.4, as the running buffer. A CM5 sensor chip was activated by a 4-min injection of 0.05 M N-hydroxysuccinimide, 0.2 M N-ethyl-N-(3-dimethylaminopropyl)-carbodiimide hydrochloride in water. Human IgG (11 mg/ml; Jackson ImmunoResearch) was diluted 1:1000 in 10 mM sodium acetate, pH 4.5. Injection of 25 l at a flow rate of 5 l/min led to immobilization of 1000 response units of human IgG. Residual reactive groups were inactivated by a 6-min injection of 1 M ethanolamine, 0.1 M NaHCO 3 , 0.5 M NaCl, 5 mM EDTA, pH 8.0. Interaction measurements were carried out at a flow rate of 60 l/min. Surface regeneration was achieved by injection of two 30-s pulses of 0.2% SDS in water. The BIAevaluation 3.0 software was used for further analysis of the data. Shown curves represent the difference between the signal of the IgG-coupled surface and of a deactivated control surface devoid of protein. They were further corrected by subtraction of the curve that was obtained after injection of running buffer alone. Buffer injection led to responses less than five response units.
Construction of the FOG Deletion and the Protein G Deletion Mutants-Isolation of chromosomal DNA from Streptococcus dysgalactiae subsp. equisimilis strain G45 was performed using the BIORAD Instagene Matrix solution as described by the manufacturer. All restriction and modifying enzymes were purchased from New England Biolabs. The sequence of the fog gene was derived from accession number AY600861 and served as basis for the selection of fog-specific oligonucleotides. The sequence of the protG gene was derived from accession number X06173. For construction of the FOG deletion mutant, a 697-bp internal fragment of fog was amplified by PCR using oligonucleotides Fog_mut1 (GCGGAGAATACATACGATA-GATGG) and Fog_mut2_MunI_overlap (ccgggcaattgccgggCT-AACTCTTGCTGCTTCTGAGC). For construction of the protein G deletion mutant, a 557-bp internal fragment of protG was amplified by PCR using oligonucleotides ProtG_ko (AAA-TCGTAGTTTCAAATTTGTCG) and protG_ko_MunI_overlap (ccgggcaattgccgggAAATCAGATAAGCCATCTGTTGC). For amplification of the aad9 gene (accession number M69221), oligonucleotides Spc1_MunI_overlap (cccggcaattgcccg-gATCGATTTTCGTTCGTGAATAC) and Spc2_AvrII_overlap (gcgcctaggcgcggcCCAATTAGAATGAATATTTCCC) were used. By application of overlap extension PCR (16), the fog and protG fragments were fused to the aad9 gene, respectively, and ligated into the cloning vector pCR2.1 (Invitrogen) generating pCR2.1-fog-aad9 and pCR2.1-protG-aad9. The fog-aad9 and the protG-aad9 cassettes were subsequently ligated into the thermosensitive shuttle vector pJRS233 (17) using vector-harboring restriction sites BamHI and XhoI, resulting in vector pISA14 and vector pIF2, respectively. E. coli DH5␣ clones carrying the recombinant plasmids were selected on spectinomycin as indicated above. Plasmid isolations were performed using the Qiagen Miniprep kit according to the manufacturer's recommendations.
The plasmids pISA14 and pIF2 were introduced into competent G45 bacteria by electroporation as described by McLaughlin and Ferretti (18). After selection of erythromycin-resistant GGS at 30°C and verification of the presence of the plasmid by PCR using vector-priming oligonucleotides M13 forward and M13 reverse, a temperature shift to 37°C led to transient inte-gration of pISA14 and pIF2 into the streptococcal genome at the designated fogI and protG locus, respectively. Correct insertion of the plasmids in the mutants, leading to interruption of the fog gene after amino acid 261, and interruption of the protG gene after amino acid 163 were confirmed by sequencing as well as Southern blot hybridization using an aad9-specific probe. Sequencing reactions were performed using BigDye Terminator Cycle sequencing kit (Applied Biosystems), and analysis of the obtained DNA sequences was performed using DNASTAR SeqManII Version 5.05 software.
Southern Blot Hybridization-Chromosomal DNA from GGS was digested with either BamHI, EcoRI, or HindIII, separated on a 0.7% agarose gel, and blotted onto positively charged nylon membrane (Macherey-Nagel). Digoxigenin labeling and detection of aad9-specific PCR probe was performed with the PCR DIG probe synthesis and detection kit (Roche Applied Science) according to the manufacturer's recommendations.
Binding Assay-Overnight cultures of bacteria were harvested by centrifugation at 1400 ϫ g for 10 min and washed with PBS containing 0.05% Tween 20. Binding of radiolabeled IgG and fibrinogen was performed as described (15,19).
Plasma and Protein Absorption-Overnight cultures of GGS were washed 3 times with PBS and resuspended to give a concentration of 0.5 g of bacteria/ml (wet weight/volume). A 100-l bacterial suspension was incubated with human blood plasma (10 l, 1 h, 37°C) or isolated C1q (2 g, 1 h, 37°C). In some experiments bacteria were pre-absorbed with human IgG (10 g, 30 min, room temperature) or Fc fragment of human IgG (2 g, 30 min, room temperature) and washed twice with PBS before the addition of C1q. After a final incubation, the bacteria were washed three times with PBS, and bound proteins were eluted with 30 l of 100 mM glycine, pH 2.0, for 15 min. Before SDS-PAGE, the samples were neutralized by the addition of 1.5 M Tris-HCl, pH 8.8.
SDS-Polyacrylamide Gel Electrophoresis and Immunoblotting-SDS-polyacrylamide gel electrophoresis was performed as described by Laemmli (20). Electrophoresis was performed under reducing conditions in a 12% separating gel. For immunoblots, the proteins were transferred to nitrocellulose and incubated with a dilution of the appropriate goat antiserum (anti-C1q, Calbiochem). Bound antibodies were detected using peroxidase-conjugated rabbit anti-goat IgG (Santa Cruz Biotechnology), 3-aminophthalhydrazide (1.25 mM), p-coumaric acid (225 mM), and 0.01% H 2 O 2 .
Mapping of Binding Regions on IgG Coupled to Immunobeads-As described in Frick et al. (21), Immunobeads (Immunobeads Reagent Activated Immunobead Matrix, Santa Ana, CA) were coupled with human IgG. Optimal binding of radiolabeled protein FOG was assessed by incubating beads and protein in Veronal-buffered saline (pH 7.35, 0.15 M NaCl, 0.1% gelatin) for 90 min at 22°C. Beads were then washed twice with 0.01 M EDTA in Veronal-buffered saline containing 0.05% gelatin, and the amount of radiolabeled protein (in cpm) bound to the beads was counted on a Wallac Wizard 1470 automatic gamma counter.
For competition experiments, increasing amounts of proteins G and FOG in 0.2 ml were used to compete for binding to 0.1 ml IgG-coupled immunobeads in the presence of radiola-beled FOG (ϳ12 ng or 30,000 cpm in 0.1 ml of Veronal buffer, pH 7.35, 0.15 M NaCl, 0.1% gelatin). Incubation and washing steps are described above.

Characteristics of Protein FOG Regarding IgG Binding-Pro-
tein FOG was shown to be a major surface protein that is common in clinical isolates of GGS (13,22). Recombinant FOG showed affinity for the immunoglobulin IgG (13), but the interaction is virtually uncharacterized. The primary structure of FOG does not contain any sequences with significant homology to the IgG binding motif of protein G, the major IgG binding molecule on the surface of GGS (15). To allow a first comparison between these IgG-binding proteins, recombinant FOG that represents the mature protein after sortase cleavage (amino acids (aa) 1-557, Fig. 1) was tested for binding to IgG of different animal species (Table 1), to different subclasses of human IgG ( Fig. 2A) and its papain fragments Fc and Fab (Fig.  2B). Radiolabeled FOG bound to immobilized human, equine, murine, monkey, guinea pig, porcine, and rabbit IgG in ligand blot experiments. A low but considerable signal was detected for canine IgG, but no binding was detectable for caprine, ovine, rat, or bovine IgG (Table 1). In contrast to protein G, which is known to bind all the four subclasses of human IgG (15), FOG bound to human IgG-subclasses IgG1, IgG2, and IgG4 but not IgG3 ( Fig. 2A). As opposed to a clear signal for binding to Fcfragment, the signal obtained for Fab fragment was faint (Fig.  2B). To further characterize the interaction with IgG, FOG was examined in SPR measurements for binding to immobilized human IgG, revealing a dissociation constant of 2.4 pM (Fig.  3A); thus, an interaction of high affinity.
Localization of the IgG Binding Region of FOG-It was described previously that a truncated form of FOG (aa 1-278 of the mature protein), which was referred to as FOG1-B, did not bind IgG, indicating that the C-terminal part (aa 279 -578) of the bacterial protein is crucial for the interaction with the immunoglobulin (13). To localize the responsible IgG binding region, suitable recombinant FOG fragments were expressed as glutathione S-transferase fusion proteins and tested in ligand blot experiments (Fig. 1) and SPR measurements (Fig. 3, B-G) for binding to human IgG. Both approaches gave identical results. In control experiments glutathione S-transferase did not show any binding to IgG either in ligand blots (data not shown) or in SPR measurements (Fig. 3H). In accordance with previous experiments (13), FOG1-B did not bind IgG (Fig. 3G). The minimal motif tested positive for IgG binding (aa 233-339) was the S-region of FOG (Fig. 3E). Truncation after amino acid 304 (B2-aa304) abolished the interaction completely (Fig. 3F). Constructs that started with amino acid 305 (aa305-S and aa305-C1) were also not capable of binding IgG (Fig. 1). For the interaction between the S-region and IgG, a dissociation constant of 890 pM was determined by SPR measurements. However, when the FOG fragment comprised the S-region together with flanking regions like the B2 domain (B2-S, Fig. 3D), the C-repeats C1 and C2 (S-C2; Fig. 3C), or both the B2 and the C-repeats (B2-C2; Fig. 3B), its affinity for IgG was significantly increased. This was judged by the dissociation constants measured by SPR of 610, 480, and 220 pM, respectively. These data imply that the flanking regions contributed to the structural integrity of the binding motif that was localized in the S-region of FOG.
FOG Accumulates IgG on the Bacterial Surface-To test whether FOG contributes to binding IgG to the bacterial surface, the fog gene of strain G45 was interrupted to generate a FOG-deficient mutant (G45⌬FOG). Insertion of the destructive DNA element was tested by Southern blot, and in contrast to lysates of the wild type strain, lysates of the mutant did not contain the protein FOG (results not shown). Binding of fibrinogen was abolished in G45⌬FOG (Fig. 4A), corroborating previous data about FOG-mediated fibrinogen binding to GGS (13). Additionally, by disrupting the prtG gene, a protein G-deficient mutant (G45⌬proteinG) was generated. When binding of radiolabeled human IgG to whole bacteria was examined (Fig. 4B), the wild type strain had the highest capacity for IgG. It bound 69 -80% of the labeled IgG. Strain G45⌬FOG had a clearly decreased capacity but was capable of binding more than 57% of the IgG even at the lowest bacterial concentration tested. Strain G45⌬proteinG had even lower capacity for IgG, but still, considerable binding of the immunoglobulin was measured (25-65%). The results indicate that FOG, like protein G, contributes substantially in accumulating IgG on the GGS surface.
Influence of Fc-binding GGS Proteins on the Interaction between IgG and the C1 Complex-To gain insight into the biological functions of the Fc-binding proteins of GGS, their interaction with the triggering component of the classical pathway of the complement system, the C1 complex, was examined. Because both FOG and protein G bind IgG to the surface of GGS (Fig. 4B) by interacting with its Fc-region, it was interesting to test if such surface-bound IgG is recognized by factor C1q or whether FOG or protein G interferes with C1q binding. To test binding of human C1q to whole bacteria, GGS isolates that were known to carry FOG and protein G (G41 and G45) or  protein G only (G148) were chosen. When comparative absorption experiments were performed, the FOG-positive strains G41 (Fig. 5) and G45 (see Fig. 7A), but not the FOG-negative strain G148 (Fig. 5), bound C1q from non-immune human blood plasma. The three subunits of C1q (25-30 kDa) were resolved and detected in Western blot using an anti-serum specific against human C1q. To dissect the system of occurring interactions, isolated C1q was added to the bacteria with or without preincubation with human IgG (Fig. 6A). Neither FOG-positive bacteria (G41) nor FOG-negative bacteria (G148) bound isolated C1q, indicating that direct binding of C1q to the GGS surface is of negligible magnitude. After preincubation, both GGS strains accumulated similar amounts of IgG, but only the FOG-positive strain G41 bound C1q. Experiments with the Fc fragment of IgG paralleled those results, demonstrating that a non-immune interaction with Fc-binding proteins is involved for the C1q binding (Fig. 6B, for Fc binding see also Fig. 2B). The quantity of bound C1q, however, was considerably lower, so that only the subunit with the strongest signal was detected in the Western blot.
To test FOG-dependent C1q binding in the same genetic background, experiments were carried out with strain G45 and its FOG-deficient mutant (G45⌬FOG). Incubated with nonimmune plasma, G45 but not G45⌬FOG bound C1q on its surface, unambiguously identifying FOG as the responsible protein (Fig.  7A). Consecutive incubation with IgG and C1q verified that the FOG-mediated C1q binding was IgG-dependent (Fig. 7A). Moreover, the protein G-deficient mutant (G45⌬protein G) retained binding of C1q to the bacterial surface, further emphasizing protein FOG as the responsible protein (Fig. 7B).
On the surface of G148 as well as on the FOG-deficient mutant G45⌬FOG, the major IgG binding molecule is protein G. The experiments, thus, demonstrate that binding of protein G to IgG efficiently prevents opsonization by C1q. On the sur-  1 and 2) and of FOG-negative strain G148 (lanes 3 and 4) were analyzed by Western blot using antiserum specific for C1q. Lanes 1 and 3 depict control experiments with buffer alone (Ϫ). Lanes 2 and 4 show eluates of GGS treated with non-immune human blood plasma (ϩ). M r standards are indicated to the left. Arrows to the right indicate the mobility of C1q subunits a, b, and c.  face of G41 and G45, however, a second population of IgG, bound via FOG, was present. This population was capable of binding C1q. Competition experiments with IgG coupled to immunobeads demonstrated that binding of radiolabeled FOG could be inhibited by the addition of protein G (Fig. 8A). This indicates an overlap in the targeted binding motifs on IgG and suggests that simultaneous binding of FOG and protein G to one heavy chain of IgG did not occur. In summary, the experiments demonstrate that FOG facilitates an IgG-dependent binding of C1q on GGS under non-immune conditions and even in the presence of other plasma proteins (e.g. fibrinogen), whereas protein G counteracts opsonization by the first component of the classical pathway, the C1 complex. We, therefore, propose the model depicted in Fig. 8B.

DISCUSSION
Binding of immunoglobulins is a key event in immune defense being crucial for the development of humoral immune response and contributing to innate immunity. Such host responses greatly depend on antigen binding. Streptococci have evolved a variety of proteins for non-immune binding of IgG. The role of these interactions is thought to be inactivation of IgG, but the consequences of such binding are widely understudied. On the bacterial surface non-immune binding of immunoglobulins and the mode of binding may have an important influence on the recognition by Fc receptors on phagocytes as well as on recognition by the first step of the classical pathway of the complement system, the C1 complex. The latter point has been addressed in this study. One major IgG-binding protein of GGS is protein G (15). Because it has been shown that soluble protein G does not interfere with the binding of subcomponent C1q to IgG (23), a role of protein G in protection against the classical pathway was arguable. However, the inter-  . Non-immune IgG binding to GGS. A, in vitro competition assay of the interaction between radiolabeled FOG and IgG coupled to Immuno-beadsா. Displacement curves were obtained by plotting the amount of unlabeled inhibitor against the observed reduction in cpm bound to the beads. FOG (black diamonds) or protein G (gray boxes) were added as inhibitors. B, model of non-immune IgG binding to GGS. The two complexes depicted, IgG bound to FOG to the left and IgG bound to protein G to the right, represent two populations of IgG bound to the surface of FOG-carrying GGS. Only the FOG-bound IgG is capable of binding to C1q (arrow); IgG bound to protein G is not (crossed out arrow). The model does not exclude that simultaneous binding of FOG and protein G to one IgG molecule may occur (not shown) but excludes their binding to the very same heavy chain. The shape of FOG and protein G shown in this figure is not based on structure data and completely notional.
actions in solution may not necessarily represent those occurring at the streptococcal surface where protein G is arrayed on the cell wall of GGS with a defined orientation and spacing and where protein G is present as a full-length protein. Also, it cannot be excluded that co-factors on the bacterial surface contribute to the ability of protein G to arrest IgG. Comparing protein G and FOG-expressing strains (G41, G45) with strains deficient for FOG (G148, G45⌬FOG) revealed that the IgG population attached to the GGS surface by non-immune binding to protein G does not promote or even allow considerable binding of C1q to the bacterial surface. Although contribution of hitherto unknown co-factors cannot be excluded, the presented results clearly indicate that interaction with protein G abolishes the capability of IgG to bind C1q. We, therefore, conclude that they reflect a function of protein G, which is to prevent activation of the classical pathway of complement. Protein G interacts with the interface region of Fc, which connects the C␥2 with the C␥3 domain of IgG (24 -26). The site of C1q-interaction is localized in the C-terminal part of the C␥2 domain of IgG (27). The close vicinity of these sites may allow protein G to interfere sterically, preventing the interaction with C1q. Competition between FOG and protein G for IgG binding indicates that the binding site for FOG is also in or close to the interface region but distant enough not to interfere with binding of C1q to the C␥2 domain.
M-like protein FOG is found on a large number of clinical GGS isolates from human patients (13,22). Like protein G, FOG can be an abundant surface component of GGS. Its mode of IgG binding differs clearly from that of protein G, which manifests itself in allowing the interaction between FOGbound IgG and C1/C1q. To our knowledge this is the first report about non-immune binding of IgG to bacteria that can lead to recruitment of this early component of the classical pathway of complement to the bacterial surface.
Unlike protein G, which binds all subclasses of human IgG (15), FOG binds the subclasses IgG1, IgG2, and IgG4 but not IgG3 ( Fig. 2A). Notably, IgG1 was found to be efficient in binding of C1q and C1 activation. This suggests that FOG-bound IgG1 triggers the complement cascade, which at first sight appears disadvantageous for the streptococci. A protective effect of FOG in a whole blood phagocytosis assay (13) indicates that anti-phagocytic actions of FOG overcome possible detrimental effects of IgG and C1q binding. Among those anti-phagocytic actions is the formation of FOG/fibrinogen aggregates that exert an anti-phagocytic effect on neutrophils, as reported recently (13). Also recently, it was described that fibrinogen, bound by M5 to the surface of S. pyogenes, reduces the amount of classical pathway C3 convertase on the bacterial surface, diminishing deposition of downstream components of the complement system (12). A similar function of FOG is conceivable.
The majority of GGS isolates that cause human infections are equipped with M or M-like proteins (28). Many possess FOG or homologues of FOG (13,22). Together with protein G such strains possess two differently specialized tools to handle IgG. The role of protein G is to act as an anti-opsonin that blocks an important interaction of the innate immune system. Protein FOG contributes to accumulating IgG molecules on the surface of GGS but does not interfere with the interaction between IgG and C1. A potential initiation of the classical pathway due to FOG may not be a threat for the bacteria since FOG is a potent anti-phagocytic factor (13). Production of anaphylatoxins due to a triggered complement cascade and possible induction of inflammatory responses, however, may play a crucial role in the pathogenesis of GGS infections.