Immunoglobulins (Igs) play critical roles in immune defence against infectious disease. They elicit potent elimination processes such as triggering complement activation and engaging specific Fc receptors present on immune cells, resulting in phagocytosis and other killing mechanisms. Many important pathogens have evolved mechanisms to subvert or evade Ig-mediated defence. One such mechanism used by several pathogenic bacteria features proteins that bind the Ig Fc region and compromise engagement of host effector molecules. Examples include different IgA-binding proteins produced by Staphylococcus aureus, Streptococcus pyogenes, and group B streptococci, all of which interact with the same interdomain region on IgA Fc. Since this region also forms the interaction site for the major human IgA-specific Fc receptor CD89, the bacteria are able to evade CD89-mediated clearance mechanisms. Similar disruption of Ig effector function by pathogen Ig-binding proteins is evident in other species. Remarkably, all the Ig-binding proteins studied in detail to date are seen to target the CH2–CH3 domain interface in the Ig Fc region, suggesting a common mode of immune evasion. A second Ig subversion mechanism that has evolved independently in numerous pathogens involves proteases that cleave Ig molecules within their hinge regions, uncoupling the antigen recognition capability of the Fab region from clearance mechanisms elicited by the Fc region. The emerging understanding of the structural basis for the recognition of Igs as substrates for these proteases and as interaction partners for Ig-binding proteins may open up new avenues for treatment or vaccination.
Immunoglobulin effector function and protection against infectious disease
Immunoglobulins (Igs) play a key role in protection against infectious disease. By virtue of the structural flexibility of their component Fab and Fc regions and their multi-functionality, Igs are able to prevent infection via a range of mechanisms. Binding of the Ig Fab region to antigenic structures on invading pathogens blocks interactions with host cells and tissues and directly stops infection. Similarly, binding by the Ig Fab region to protein toxins produced by many bacteria prevents their interaction with cellular receptors, thereby preventing disease. In addition to this direct blocking effect mediated by the Fab region, Igs are able, through interactions involving their Fc regions, to recruit powerful effector molecules that can elicit a variety of potent clearance mechanisms. The precise effector molecules recruited are dependent on the particular class and subclass of Ig involved . Here, we will focus on IgG and IgA, the two main Ig classes in serum. IgA is also the major Ig class in mucosal secretions.
Human IgG, and particularly IgG1 and IgG3 subclasses, when suitably arranged on the surface of a pathogen, can bind to C1q, the first component of the classical pathway of complement. The CH2 domain of the IgG region is critical for binding of C1q, with the core of the interaction site lying across the upper surface of this domain . It is now appreciated that elements lying down the outer edge of the Fc region with contributions from the CH3 domain also play a role in enabling assembly of hexamers of IgG molecules such that C1q can readily bind . This event leads to activation of the complement cascade and ultimately killing of the pathogen through lysis or phagocytosis.
IgG can also recruit a family of specific Fc receptors, namely FcγRI (also known as CD64), FcγRII (CD32), and FcγRIII (CD16), which collectively are expressed on phagocytes such as monocytes, macrophages, and neutrophils and other immune cells . These receptors bind the Fc region of IgG molecules that have clustered on pathogen surfaces and trigger elimination mechanisms including phagocytosis, antibody-mediated cellular cytotoxicity, and release of activated oxygen species and enzymes which attack and kill the foreign cell that has been recognised. The human IgG subclasses differ in the affinity for each FcγR, with IgG1 and IgG3 generally displaying the highest affinity.
IgA, in contrast, cannot activate the classical complement pathway and much of its effector function is thought to result from interaction with the IgA-specific Fc receptor, FcαRI (CD89), found on neutrophils, macrophages, monocytes, eosinophils, and some other immune cell types. Mutagenesis experiments on IgA implicated two loops (Leu257–Gly259 in the CH2 domain and Pro440–Phe443 in the CH3 domain), lying at the interface between the CH2 and CH3 domains of the IgA Fc region, in interaction with FcαRI (Figure 1A) [5,6]. Solving of the X-ray crystal structure of the complex of IgA Fc and the extracellular domains of FcαRI later revealed that this site localisation was correct .
Co-localisation of sites bound by IgA-binding proteins with those bound by host effector molecules.
A receptor critical for IgA function at mucosal surfaces is the polymeric Ig receptor (pIgR). This receptor carries dimeric forms of IgA out across the epithelial barrier at mucosal surfaces, delivering this secretory form of IgA to its site of action. pIgR has five extracellular Ig-like domains, and available evidence indicates that domains 1 and 5 are essential for binding to dimeric IgA. Domain 1 first interacts non-covalently with the CH3 domain of IgA Fc [8,9]. Residues 402–410 lying in an exposed loop of the CH3 domain have been especially implicated [9,10]. Residues Phe411, Val413 and Thr414 appear to play central roles, with contributions from close-lying residue Lys377 and the CH3 interface loop Pro440–Phe443 (Figure 1B) . The latter loop is also involved in the FcαRI interaction.
It is worth noting that both IgG and IgA can also be recognised by the tripartite motif-containing protein 21 (TRIM21), which is present in almost all cell types and engages pathogens opsonised with IgG, IgA, or IgM that have entered the cell cytoplasm. TRIM21 binds at the Ig Fc domain interface .
In both IgG and IgA, the two Fab regions and the single Fc region are separated by a flexible hinge region, the length and sequence of which varies with Ig subclass. Despite this considerable variability, the hinge appears to play a common function, namely to impart flexibility to the molecule and act as a spacer between the Fab and Fc regions. The flexibility, coupled with the spacer action, allows Ig molecules to interact with two antigen molecules spaced at a range of distances apart and simultaneously to engage effector molecules such as Fc receptors. It is clear that without the hinge this capability would be compromised, resulting in low efficiency of antibody effector function. However, the exposed nature of the hinge necessary to impart this flexibility and spacer action comes at a price, that of susceptibility to proteolysis. As will be discussed later, this susceptibility has been exploited by numerous pathogens as a means to circumvent Ig-mediated immune protection.
Bacterial Ig-binding proteins
Several important bacterial pathogens have evolved proteins that bind specifically to Igs. Sometimes these so-called Ig-binding proteins have additional capabilities, for instance affinity for key complement or serum proteins, that enhance their ability to counteract host immune defence processes. Bacteria which produce such proteins are associated with major diseases and include (a) Staphylococcus aureus, an invasive pathogen responsible for skin and soft tissue infections, bacteraemia, sepsis, and endocarditis, and one of the most frequent causes of human bacterial infection, (b) Streptococcus pyogenes, a common cause of acute sore throat, skin infections, and life-threatening toxic-shock syndrome, with possible sequelae including rheumatic fever or acute glomerulonephritis, and (c) group B streptococci, the most common cause of life-threatening bacterial infection in newborn babies.
S. aureus produces both IgG- and IgA-specific Ig-binding proteins. Of the former, perhaps the best known, as a result of its wide use in laboratories as a means to detect or isolate IgG, is protein A (SpA). S. aureus also produces the so-called Staphylococcal binder of Ig (Sbi), which binds IgG and is found mainly extracellularly . A further IgG-specific protein produced by S. aureus is Staphylococcal superantigen-like protein 10 (SSL-10), one of a family of exoproteins with some structural similarity to superantigens but with no superantigenic activity. Among other binding partners, SSL-10 binds IgG , while another member of this family, SSL-7, binds to IgA .
Of the Ig-binding proteins produced by S. pyogenes, here we will focus on the IgA-binding proteins Arp4 , and Sir22, which also binds IgG . These proteins are members of the M protein family, which comprises coiled-coil proteins implicated in virulence. Certain strains of group B streptococci produce β protein, an IgA-binding protein that is unrelated to the IgA-binding proteins of S. pyogenes , and which will also be considered here.
Examples of perturbation of human IgA effector function by IgA-binding proteins
Analysis of the impact of swapping of Fc domains between IgA and IgG, and of targetted mutations, on interactions with Arp4, Sir22, and β protein, revealed that all three IgA-binding proteins interact with similar sites lying at the interface of the two IgA Fc domains . Residue Leu257 in the CH2 domain and residues Pro440 and Phe443 in the CH3 domain have been predicted to play key roles in binding to these IgA-binding proteins (Figure 1C). These binding sites overlap with that for human FcαRI, and the IgA-binding proteins have been shown to inhibit FcαRI binding and FcαRI-mediated elimination mechanisms . Such receptor blockade provides an effective means for the bacteria to evade IgA-mediated clearance mechanisms normally elicited following FcαRI engagement.
Strikingly, the unrelated SSL-7 protein from S. aureus has been shown to bind IgA Fc in the interdomain region also [20,21]. The solved X-ray structure of the IgA Fc–SSL-7 complex revealed similar overlap with the FcαRI interaction site (Figure 1D) . SSL-7 inhibits FcαRI binding to IgA, providing another example of evasion of the IgA immune response through targeting of a host receptor interaction site .
Examples of perturbation of human IgG effector function by IgG-binding proteins
SpA is present on the surface of S. aureus and binds to the interdomain region of IgG Fc, lying between the CH2 and CH3 domains. Although this interaction site lies some distance away from the binding site for host FcγRs in the lower hinge region , SpA attachment to the IgG Fc interdomain region compromises IgG binding to FcγRs on host cells  and helps to protect staphylococci from phagocytic killing . It sequesters IgG on the bacterial surface forming a ‘shield’, preventing further recognition by the immune system . In addition, SpA can bind to the Fab domain of VH3-type B-cell receptors and cross-link them, resulting in proliferative supraclonal expansion and apoptotic collapse of the activated B cells . Through the use of SpA mutants that were unable to bind IgG, it has been shown in a mouse model that the ability of SpA to bind IgG Fc is essential for S. aureus escape from host immune surveillance .
In contrast with SpA's role at the cell surface, Sbi appears to be able to occur both extracellularly and anchored to the cell envelope . The former location allows it to neutralise the function of antibodies in the extracellular space, favouring survival of the bacterium . It has been noted that the two N-terminal domains of Sbi share sequence similarity with SpA. Conservation of key interacting residues suggests that Sbi binds IgG in a similar manner to SpA .
Turning to SSL-10, mutagenesis studies indicate that this protein most probably binds to the outer face of the Cγ2 domain in close proximity to both the FcγR- and C1q-binding sites . It blocks the binding of IgG1 to FcγRs and phagocytosis of IgG1-opsonised bacteria , as well as inhibiting IgG-mediated activation of the classical complement pathway .
A common theme of targeting Ig Fc interdomain regions
Analysis of protein interaction sites on Ig Fc regions has led to an appreciation of the frequent co-localisation or overlap of binding sites. It is remarkable that many pathogens had evolved means to target the same Fc interdomain regions utilised by several host receptors. Thus, the same region at the IgA CH2–CH3 interface acts as a multifunctional interaction site, binding host receptors FcαRI, pIgR, and Fcα/μR (a receptor specific for IgA and IgM expressed in the tonsil and gastrointestinal tract) , and Ig-binding proteins from S. aureus and streptococcal species.
Similarly in IgG, the CH2–CH3 interdomain region acts as a common binding site for both host receptors and proteins produced by various pathogens. Thus, host proteins TRIM21, responsible for control of antibody-coated pathogens in the intracellular environment , and FcRn, which determines IgG turnover and transport , both bind here, as do bacterial proteins SpA from S. aureus and protein G produced by certain streptococcal strains.
This theme is evident not only across different Ig classes (IgA and IgG) and different pathogens, but also different host species. Notably, Streptococcus equi subspecies equi, the causative agent of strangles, a highly infectious respiratory disease in horses, produces an IgG-binding protein known as FgBP that interacts with the IgG CH2–CH3 interface and blocks IgG effector function .
The Fc interdomain region appears to be particularly suited for protein–protein interactions. Indeed, in the case of IgG, it has been recognised as one of a small number of regions on the antibody surface that are intrinsically suitable for such interactions [32,33]. Presumably, it has been conserved as a site for host receptors, and in an elegant example of convergent evolution, numerous pathogens have co-opted this region as an interaction site for their own proteins, as an effective means to subvert the Ig response. Indeed, taking the IgA–FcαRI interaction as an example, phylogenetic and diversity analysis has highlighted an ‘arms race’, in that the interaction has evolved under selective pressure imposed by pathogen IgA-binding proteins [34,35].
Bacterial proteases that target Igs
As an alternative means to evade Ig responses, various bacterial pathogens have evolved proteases that are capable of specifically cleaving Igs in a manner that perturbs normal Ig effector function. For example, several pathogenic bacteria produce proteases that are capable of cleaving IgG within the lower hinge of the CH2 domain, thereby disrupting the very region that is critical for binding to host FcγR. The glutamyl endopeptidase V8 (GluV8) of S. aureus and Ig-degrading enzyme (IdeS) of S. pyogenes are prime examples . GluV8 cleaves N-terminal to Leu234, whereas IdeS cleaves between Glu236 and Glu237 (Figure 2A). Both cleavage sites lie at the heart of the key interaction site for all FcγR. It has been noted that cleavage of just one of the heavy chains in the lower hinge region is sufficient to render IgG incapable of engaging FcγRs, emphasising the benefit to pathogens to be gained by precise targeting of this region .
Ig proteases and evasion of the Ig response.
Targeting of the hinge region is also seen in IgA. Many clinically important pathogenic bacteria produce proteinases that specifically cleave the extended hinge region of human IgA1 (Figure 2B). In general, the bacteria that produce these so-called IgA1 proteases are associated with and/or colonise mucosal surfaces, and the capability to evade the predominantly IgA responses at mucosal sites surely affords them a considerable advantage. The fact that they have evolved several times over, and fall into very different protease families such as serine proteases and metalloproteinases, supports such a notion. Among other diseases, the bacteria in question are responsible for life-threatening meningitis (Neisseria meningitidis, Haemophilus influenza, and Streptococcus pneumoniae), as well as major infections of the genital tract (Neisseria gonorrhoeae) and oral cavity (Streptococcus sanguis, Streptococcus mitis, and Streptococcus oralis). Each enzyme cleaves at a specific post-proline site within the IgA1 hinge (Figure 2B). Since the equivalent region is lacking in IgA2, this subclass is resistant.
Cleavage of IgA1 within the hinge region separates the antigen-binding capability of the Fab arms from the effector function capability of the Fc region. Thus, IgA1 protease action can drastically reduce the ability of IgA1 to elicit phagocytosis and other killing mechanisms. Using S. pneumoniae as an example, strains differing in the ability to produce IgA1 protease have been compared. Such experiments showed that production of IgA1 protease and the resultant sabotage of IgA effector function result in increased bacterial survival in vivo, offering a convincing demonstration of the advantage conferred to the bacterium in secreting an IgA1 protease .
To recognise IgA1 as a substrate, clearly IgA1 proteases must recognise elements within the IgA1 hinge itself. However, there is a body of evidence to suggest that the context of the hinge is also important for efficient recognition and cleavage by some of these proteases [39–41]. Analysis of a series of hinge mutants revealed a requirement for the susceptible bond to be positioned at a suitable position relative to the Fc for cleavage to occur . Moreover, some IgA1 proteases demonstrate a requirement for the presence of elements within the Fc region of IgA1 for cleavage to proceed [39,42]. In yet another example of a pathogen co-opting host–receptor-binding sites, the N. meningitidis type 2 enzyme, for example, shows a requirement for residues Pro440–Phe443, the same loop in the CH2–CH3 interface involved in binding to FcαRI and pIgR (Figure 2C) . In a further example, it appears that the Haemophilus influenzae type 2 enzyme needs other Cα3 residues implicated in binding to pIgR, before it will recognise an IgA1 molecule as a substrate (Figure 2C) . Interestingly, modelling based on the solved X-ray crystal structure of an H. influenzae IgA1 protease is consistent with involvement of the Fc region . It has been proposed that binding of Fc by the protease stabilises a particular IgA1 conformation such that the IgA1 hinge may access the protease's active site and cleavage may then occur .
It is hoped that a better understanding of the molecular basis of Ig-targeted evasion strategies employed by pathogens may highlight opportunities for intervention or for improved vaccine design strategies. In summing up, perhaps it is worth highlighting one or two examples to illustrate where such insights may be beginning to bear fruit.
As a first example, the finding that a SpA mutant that is unable to bind IgG Fc elicits adaptive responses that protect against recurrent infection has led to the mutant being proposed as a protective antigen for the development of a staphylococcal vaccine .
As a second example, inhibitors of IgA1 proteases should be useful in deciphering the role of these proteases in bacterial colonisation and virulence, enabling investigation into their potential as antibacterial targets [44,45]. Indeed, strategies designed to counteract bacterial protease defence against host Igs are emerging and showing some promise . Moreover, an improved understanding of the molecular basis for recognition of IgA as a substrate is likely to inform new ways to probe IgA1 protease activity in patient samples. Indeed, probes for IgA1 protease activity have been developed, which may prove useful in screens for further IgA1 protease inhibitors . Another potentially beneficial outcome of better understanding of the molecular basis of IgA1 protease action is the use of IgA1 proteases as therapeutic candidates . Encouragingly, IgA1 protease treatment has been reported to reverse pathology in animal models of IgA nephropathy, the most common form of glomerulonephritis that is characterised by deposition of IgA1-containing immune complexes in the kidney [49,50], suggesting that IgA1 proteases may represent a novel treatment for this common disease.
The Author declares that there are no competing interests associated with this manuscript.
The author thanks colleagues for their invaluable contributions to collaborative research cited here, in particular Melanie Lewis, Bernard Senior, Richard Pleass, Gunnar Lindahl, Thomas Areschoug, James Dunlop, Catherine Anderson, Ana Pinheiro, and Pedro Esteves.