SIC (streptococcal inhibitor of complement) is a 31 kDa protein secreted by a few highly virulent strains of GAS (group A streptococci), predominantly by the M1 strain. Initially described as an inhibitor of the membrane attack complex of complement, it has turned out to be a polyfunctional inhibitor of the innate mucosal immune response. The SIC protein sequence contains three domains: an N-terminal SRR (short repeat region), followed by three longer tandem repeats [LRR (long repeat region)] and a C-terminal PRR (proline-rich region). SIC inhibits the antibacterial activity of a wide range of antimicrobial peptides and proteins: i.e. lysozyme, SLPI (secretory leucocyte proteinase inhibitor), LL-37, hNP-1 (human neutrophil peptide-1) and the human β-defensins 1, 2 and 3. Analysis of the functional properties of recombinant domains of SIC shows that binding and inhibition of lysozyme and human β-defensin-3 require the SRR+LRR, as does binding to SLPI. Complement inhibition is confined to the SRR. M12 GAS secrete a protein ‘distantly related to SIC’ (DRS). DRS contains a C-terminal PRR which is significantly similar to that of SIC, but it has no central LRR and the N-terminal SRR is very different. DRS inhibits human β-defensin-3, but has no effect on lysozyme, SLPI or complement.
In order to be successful pathogens, micro-organisms need to evade or to subvert the immune mechanisms of the host and this includes protecting themselves against the numerous antimicrobial peptides of the mucosal innate immune response. The mechanisms employed by micro-organisms are many and varied  but this account concentrates on a highly unusual strategy used by the human pathogen Streptococcus pyogenes, which secretes a polyfunctional inhibitor of many components of the mucosal innate response.
This inhibitor is known as the ‘streptococcal inhibitor of complement’ (SIC), reflecting the system where it was first discovered as an inhibitor of the membrane attack complex . The mechanism for this inhibition was subsequently found to be prevention of the insertion of C5b67 into the cell membrane, probably by blocking its transiently expressed membrane insertion site . Many normal plasma components (vitronectin, clusterin, lipoproteins and above all C8) share this activity, and this anti-complement activity of SIC is not thought to be important for its function in vivo.
SIC is a 31 kDa extracellular protein produced by a few highly virulent strains of GAS (group A streptococci), predominantly by the M1 strain, which tends to be associated not only with uncomplicated pharyngitis, but also with severe invasive disease. A similar sized protein, termed ‘distantly related to SIC’ (DRS), is produced by the M12 and M55 strains of GAS (which are also throat strains but are typically associated with acute post-streptococcal glomerulonephritis) . While DRS also binds to the individual complement components C6 and C7 by ELISA, and competes with SIC for C6 and C7 binding, it is unable to inhibit complement lysis .
The SIC protein sequence shows no marked similarity to that of other proteins so far described. It contains a number of distinct regions: an SRR (short repeat region) between residues 1 and 74, followed by three longer tandem repeats [LRR (long repeat region)] between residues 75 and 153 and a C-terminal PRR (proline-rich region), comprising the remainder of the protein from residues 154 to 273. The DRS protein contains a C-terminal PRR which is significantly similar to that found in M1 SIC, but it does not possess a central LRR and the N-terminal SRR only contains a few very short stretches that are similar to M1 SIC. In particular, the N-terminal region of DRS lacks ten out of the 15 tryptophan residues present in that of M1 SIC. All strains of DRS isolated in Europe so far contain two copies of the N-terminal region , whereas the drs genes analysed from M12 and M55 strains isolated in Australia contain only one copy of the N-terminal SRR . Both proteins are highly negatively charged, with a pI of 4.18 (SIC) and 4.47 (DRS).
Both SIC and DRS have been shown to be highly immunogenic [7,8] and extremely variable, with over 300 variants having been recorded for SIC . Sequencing of the sic gene from numerous M1 GAS isolates has shown that long repeats 1 and 2 are the least variable areas of the protein in terms of amino acid sequence (reviewed in ), although several isolates have been shown to contain variable numbers of long repeat 2. DRS is also polymorphic  and approximately six variants have been reported, although far fewer M12 and M55 isolates have been investigated.
Because Gram-positive organisms are protected against complement-mediated lysis by the presence of their peptidoglycan cell walls and hyaluronic acid capsules, because there is no complement on the normal mucosal surface (which is one of the most frequent initial sites of GAS colonization), and because C567 uptake inhibition is not likely to be important anyway, it seemed likely that SIC had other functions that conferred advantage on the invading bacteria. Lukomski et al.  investigated the ability of a wild-type SIC-positive and a mutant SIC-negative strain to colonize the mouse throat and showed that colonization by the SIC-negative strain was significantly impaired during the first 4 days post-inoculation, showing that SIC did indeed act as a virulence factor during the early stages of infection.
Studies were therefore performed to see whether SIC (and, subsequently, DRS as well) interacted with any of the antimicrobial proteins and peptides of the innate immune system found on mucosal surfaces [12,13]. Lysozyme, the most abundant and ubiquitous antimicrobial protein in humans, was studied first. SIC was found to precipitate HEL (hen egg lysozyme) quite dramatically at room temperature (20°C), even though quantitative binding studies showed the binding to be of low affinity and to inhibit the antimicrobial and catalytic functions of both human and HEL at a molar ratio of approx. 1:3 . This binding was highly temperature-dependent and largely enthalpy-driven.
The second candidate molecule to be studied was SLPI (secretory leucocyte proteinase inhibitor), another important antibacterial peptide that SIC bound at a much higher affinity and with paradoxical temperature dependence, showing the reaction to be largely entropy-driven, and totally inhibited its ability to kill GAS at an SIC/SLPI molar ratio of 1:2 . SLPI is a dual function molecule comprising two very similar domains, one of which contains the enzyme inhibitory activity and the other the antimicrobial activity . Interestingly, SIC had no effect on the anti-proteinase activity of SLPI located in the C-terminal domain of the protein, implying that its action is specifically against the N-terminal antimicrobial domain of SLPI . Since publication of that study, SIC has been shown to bind only to the isolated N-terminal domain by ELISA. Recent work has shown that DRS is unable to inhibit the antibacterial activity of SLPI against M1 GAS, or to inhibit the catalytic activity of HEL .
Further studies have shown that SIC and DRS also bind to and inhibit the antibacterial action of a number of smaller antimicrobial peptides, all of which are able to kill GAS at micromolar concentrations. Frick et al.  showed that both SIC and DRS bound to and inhibited the cathelicidin LL-37: SIC at a molar ratio of approx. 1:2 (SIC/LL-37) and DRS somewhat less efficiently at 1:1. Both proteins also inhibited the antibacterial action of the human α-defensin hNP-1 (human neutrophil peptide-1) equally well at approx. 1:1 . These experiments were all performed in low ionic strength buffer (10 mM Tris, pH 7.5, and 5 mM glucose), which facilitates ionic interactions between oppositely charged proteins.
We have studied the activity of SIC with the human β-defensins  both at low ionic strength (10 mM Tris) and in 10 mM phosphate buffer (pH 7.2), which more closely resembles the ionic strength of saliva, i.e. the environment encountered by streptococci in the throat. hBD-2 (human β-defensin) and hBD-3 killed M1 GAS in both buffers (although more efficiently at low ionic strength) but hBD-1 was only able to kill M1 GAS at low ionic strength and a 7-fold higher concentration was required compared with that of hBD-3. However, hBD-1 was able to kill an M6, SIC-negative, strain of GAS in phosphate buffer, but the reason for its greater susceptibility is not known. SIC inhibited the antibacterial activity of hBD-2 against M1 GAS equally well in both buffer systems – 100% inhibition being achieved at a molar ratio of approx. 1:3 SIC/defensin. Inhibition of hBD-3 was more efficient at low ionic strength: at a molar ratio of 1:1 in 10 mM phosphate and 1:2 in 10 mM Tris. SIC also inhibited hBD-1 at a molar ratio of 1:2 at low ionic strength . DRS also inhibits hBD-3, but less efficiently than SIC . It seems improbable that the inhibition seen only at low ionic strength is of importance in vivo.
Inhibition experiments have also been performed using recombinant SIC proteins (in the pBAD Thio vector) containing the three separate domains of the protein, and combinations of the SRR+LRR or LRR+PRR, in order to find out which region of SIC is responsible for inhibition of the antimicrobial activity of hBD-3 and of lysozyme, and for inhibition of complement lysis . Similar experiments were also performed using purified elastase-generated proteolytic fragments of native SIC which spanned broadly similar areas of SIC to the recombinant proteins. The results clearly defined the binding sites within this polyfunctional protein. A clone containing the SRR+LRR gave 100% inhibition of hBD-3 at equimolar and 1:2 molar ratios (clone/hBD-3) and approx. 80% inhibition at a 1:4 molar ratio. This activity is similar to that of native SIC. More surprisingly, a clone containing the C-terminal PRR gave rather variable inhibition (up to 50%) of hBD-3 at a molar ratio of 1:1. A clone containing the LRR+PRR inhibited no better than that containing PRR alone. No inhibition of hBD-3 activity was shown by the SRR, LRR or the thioredoxin fusion partner clones alone.
ELISA showed that binding of SIC to hBD-3 and SLPI required the intact molecule, or the SRR+LRR, or a similar proteolytic fragment of the native protein (; B.A. Fernie-King, D.J. Seilly and P.J. Lachmann, unpublished work). This proteolytic fragment, which contained part of the SRR together with two of the long repeats, also inhibited the catalytic activity of HEL, albeit less well than the whole molecule.
Inhibition of complement-mediated lysis was demonstrated, in two different assay methods, to be confined solely to the SRR region of SIC. Interestingly, a proteolytic fragment comprising the N-terminal half of the SRR showed no inhibition whatsoever of complement lysis, implying either that the whole of that region is required for optimum function or that the binding point is in the vicinity of the elastase digestion site and had thus been destroyed. Binding to the purified complement components C6 and C7 was also confined to the SRR . A summary of all of the inhibitory functions is shown in Table 1.
|M1 SIC||M12 DRS|
|M1 SIC||M12 DRS|
Data from .
Binding by ELISA.
Another study reported a totally different property for SIC, i.e. the ability to inhibit the adherence of GAS to cultured human epithelial cells. Using the same mutant SIC-negative strain and parental wild-type SIC-positive strain used in the mouse throat colonization study described above , Hoe et al.  observed that the mutant adhered to A549 human lung epithelial cells significantly better than did the wild-type. Pre-incubation of A549 cells with purified SIC inhibited the subsequent adherence of both the mutant and non-SIC-producing GAS strains to these cells, indicating that SIC was interacting in some way with the host cell. Electron microscopy revealed that SIC was rapidly internalized by the cells, appeared to induce cell flattening and loss of microvilli and was distributed throughout the cytoplasm. Binding of SIC was localized to the intracellular proteins ezrin and moesin (which mediate the interaction of the actin-based cytoskeleton with the plasma membrane) and specifically with the C-terminal region of both proteins. The area of SIC responsible for this binding was subsequently found to be the PRR . The same area of DRS also binds to ezrin by ELISA, but less strongly than the PRR of SIC . Hoe et al.  suggest that binding of SIC to ezrin somehow affects cellular mechanisms necessary for GAS adhesion, internalization and killing, and that SIC enhances bacterial survival by enabling the pathogen to avoid the intracellular environment. However, over the years, there have been a number of papers demonstrating that GAS can survive in the intracellular environment (for instance ), so this interpretation of these results is somewhat problematic.
In conclusion, SIC and DRS would appear to be multifunctional virulence factors of certain strains of GAS which protect the bacteria from the antimicrobial action of several components of the innate immune system present on mucosal surfaces.
Antimicrobial Peptides: Mediators of Innate Immunity in the Development of Anti-Infective, Therapeutic and Vaccination Strategies: Focused Meeting held at New Royal Infirmary, Edinburgh, U.K., 21 November 2005. Organized and edited by J.-M. Sallenave and J. Govan (Edinburgh, U.K.).
We thank the British Heart Foundation for their generous support of this work.