Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
The complement system is composed of more than 30 soluble proteins, cell surface regulatory factors, and receptors that work together to accomplish a wide variety of functions. These are important not only in host defense and inflammatory responses but also in normal physiologic homeostasis. By facilitating proper disposal of apoptotic cells and antigens from infectious agents, complement plays an important role in preventing autoimmune and immune complex deposition diseases. During infection, complement fragments rapidly activate and attract leukocytes and macrophages and provide opsonins that facilitate clearance and killing of invading organisms. Complement also bridges the innate and adaptive immune systems in both directions. Besides its traditionally understood roles in amplifying the effector functions of antibodies, it also has important influences on antigen presentation and B-cell activation, giving it a key role in the afferent limb of antibody responses. As expected of a system with so many different physiologic activities, disorders of complement can have wide-ranging effects.
Complement activation involves a series of elegant biochemical processes by which individual circulating proteins are transformed into complex multisubunit enzymes, come together to form multichain membrane channels that insert into cells from the outside, provide address tags that direct trafficking of a wide variety of targets, and stimulate vigorous inflammatory responses. These transformations are mainly achieved by limited proteolytic cleavage and/or conformational changes. , At each stage of activation, newly revealed conformations are accompanied by new enzyme activities and/or markedly increased affinity for other components or cell surface receptors. Newly formed enzymes can cleave many molecules of the next component in the sequence, providing potent forward amplification. For this reason, the activation pathways are considered “cascades,” in analogy with the cascades of the clotting or contact activation systems. Understanding how the complement proteins are activated is the first step towards understanding this system.
Complement activation can be initiated by three major pathways. Because the complement system was originally discovered as a set of heat-labile serum proteins that could lyse erythrocytes sensitized with antibody, the antibody-dependent or “classical” pathway has been best studied. Although it is phylogenetically the newest pathway, its activation is easiest to describe and is presented first ( Fig. 120.1 ).
C1, its first component, is composed of 22 peptide chains arranged into three subcomponents termed C1q, r, and s. C1q itself contains six sets of trimers, each of which has a collagen-like stalk and a globular head , that can bind the Fc domain of immunoglobulin (Ig) G or IgM. Electron micrographs show the six stalks together in parallel and the globular heads all at the same end, giving the appearance of a bunch of tulips. Around the bunch of stalks are wrapped two molecules each of C1r and C1s, which are latent proteases. Classical pathway activation can be initiated when two or more of the six globular Ig-binding heads become attached to Fc domains of antigen-bound IgM or IgG molecules, but will be most efficient when all six are engaged by Fc domains in close proximity. Engagement of the heads of C1q induces internal rearrangements, leading to activation of latent proteolytic activity of C1r, which then cleaves and activates the C1s subunits. , Because of the requirement for two or more of the globular heads of C1q to be engaged for activation to begin, IgM, which has five Fc domains attached together, is a very efficient activator of the classical pathway. In contrast, many molecules of IgG must be bound to the surface of a target cell or bacterium for two or more IgG molecules to be close enough together to bridge a single C1q. The other classes of Igs—IgD, IgA, and IgE—do not bind to C1q and hence do not activate the classical pathway. C1s, when activated, has proteolytic activity and can also cleave synthetic esters in vitro; hence it is often referred to as C1 esterase . Its physiologic role is to carry out limited proteolytic cleavage of the next two components in the reaction sequence: C4 and C2. , (The components were numbered in the order in which they were discovered rather than the order in which they act; hence C4 comes before C2. 1,4 ) C1s cleaves the largest of the three chains of C4 at a single site, liberating a 9-kDa peptide (C4a) and leaving the larger fragment (C4b) with a transient ability to form a covalent bond with the target. The C4b takes on a new conformation in which it can bind C2 and promote its cleavage by C1s as well. Again, a small fragment is released and diffuses away, and the conformation and activity of the larger fragment are changed. With C2, the larger fragment is called C2a , which also has a newly exposed proteolytic active site. Together with C4b, which can also hold onto a C3 molecule, C2a forms an enzyme that can cleave and activate C3. This C4bC2a complex is therefore called C3 convertase . , This classical pathway C3 convertase, composed of one molecule of C4b covalently attached to the target and loosely bound to one molecule of C2a, is capable of cleaving many molecules of C3. C3 is homologous with C4, and on cleavage, its larger fragment (C3b) also transiently acquires the ability to bind covalently to the Ig molecule, surface, or target on which activation is occurring. , , , Many molecules of C3 can be cleaved by a single convertase, but the convertase eventually loses activity when the C2a subunit diffuses or is pushed away, and/or if C4b is degraded. Some of the newly cleaved C3b molecules diffuse away before they can bind to the surface. Many others bind to the surface at short distances around the convertase; these may serve to opsonize (facilitate phagocytosis of) the target. Some of the C3b molecules may deposit close enough to C4b2a to join with it and provide a binding site for C5 molecules, which will be cleaved by the proteolytic active site on C2a. , Thus the complex C4b2a3b serves as the classical pathway C5 convertase . Like C4 and C3, the cleavage of C5 liberates a small fragment (C5a), but unlike C4 and C3, C5 cannot form a covalent bond with the surface. The larger fragment, C5b, however, undergoes a conformational change that allows it to bind to C6. This starts formation of the lipophilic membrane attack complex (MAC), discussed later.
Activation by this route begins with the protein mannan -binding lectin (MBL) ( Fig. 120.2 , center ). Like C1q, MBL is a multimer of collagen-like stalks, each composed of three protein chains that also form a globular binding domain at one end. , In the case of MBL, however, the binding sites recognize polysaccharides containing mannose, hence its name. Because MBL, C1q, and other similar molecules have collagen-like domains and lectin-like binding sites for polysaccharides, they are often considered to be members of the same family of collagen-like lectins, or collectins . Lung surfactant proteins A and D are also members of this collectin family.
Different isoforms of MBL may contain between two and eight of the basic trimer subunits. Because many bacteria and fungi are coated with mannose-containing polysaccharides, the MBL pathway is considered part of the innate immune system, as is the alternative pathway, which is discussed later. Phylogenetically, the alternative and lectin pathways likely originated earlier than the classical pathway. In that sense, C1q may be seen as an adaptor protein that allowed this important part of the innate immune system to be recruited to enhance the activity of antibodies produced by the adaptive (or cognate) immune system. The ability of the lectin and alternative pathways to activate complement in the absence of antibody may be particularly important in the neonate, in whom the adaptive immune system is immature. Similarly, early in the course of infections, before antibodies and other specific effector mechanisms of the adaptive immune system are brought to bear, the lectin and alternative pathways provide an important innate system for distinguishing non-self from self. Like the binding of C1q to the Fc regions of IgG or IgM, the binding of MBL to mannose-containing polysaccharides leads to activation of latent protease subunits. In this pathway, the proteases that are analogous to C1r and C1s are termed MBL-associated serine proteases (MASPs). , , Two major MASPs have been identified and may circulate in the plasma as a large complex together with MBL and other minor protein components. When MBL is engaged with polysaccharides, MASP1 becomes activated and cleaves MASP2. , , Both MASP1 and MASP2 have been shown to be capable of cleaving C3, but some studies suggest that they actually act more like C1r and C1s in cleaving C4 and C2. , , Just as in the classical pathway, C2a and C4b generated by MASPs go on to form a C3 convertase, and the addition of a molecule of C3b confers C5 convertase activity on the complex. The physiologic importance of the lectin pathway is illustrated by an increased susceptibility to infections and also to autoimmune diseases like systemic lupus erythematosus (SLE) in patients with MBL deficiency.
The alternative pathway, also known as the properdin pathway, shares with the lectin pathway the ability to activate complement without antibody (see Fig. 120.2 , right ). Alternative pathway activation is initiated by C3b, which may be formed by one of the other pathways, by nonspecific proteolytic cleavage of C3, or after the spontaneous activation of C3 by water molecules , , (see later). Most investigators visualize native C3 in the circulation as a kind of coiled spring, which holds a unique structure called a thioester in an internal hydrophobic pocket. , , , This highly unusual chemical structure is formed by the binding of a thiol on the side chain of a cysteine residue with the carboxyl on the side chain of glutamate three residues away in the largest (α) polypeptide chain. ,
C4 and α 2 -macroglobulin also contain this internal thioester. Cleavage of a short peptide (C3a) from the amino-terminal end of the α chain, distant from the thioester, causes conformational changes in which the coiled spring is released, the thioester is transiently exposed, and the carbon on the glutamate can transfer one of its bonds from the sulfur of the cysteine to another acceptor. , , , , It is the ability of C3b and C4b to transfer a carbon bond to a hydroxyl or amino group on another protein or to a sugar chain that allows covalent attachment. If the transfer to another protein or sugar does not occur within milliseconds, the bond will transfer to the hydroxyl group of a water molecule, whose other hydrogen atom will bind to the sulfur of the cysteine to create a free thiol (−SH).
Transfer of the internal thioester is greatly accelerated by the conformational changes that follow cleavage of the protein chain. , , However, water molecules may occasionally penetrate into the hydrophobic pocket and hydrolyze this bond even in intact molecules of C3. Once that occurs, the same sequence of conformational changes in the protein chains will follow, even though no cleavage fragment has been removed. Thus, the thioester can be visualized as an internal latch, which keeps the spring coiled. Hydrolysis of the thioester by water and the subsequent conformational changes occur spontaneously at a slow continuous rate; hence there is always some basal rate of generation of C3 molecules with hydrolyzed thioester bonds. These molecules have a conformation similar to that of C3b and are thus called C3b-like C3 molecules . , , , This constant low-grade “tick-over” of C3 into C3b-like molecules may be thought of as the idling of a car engine—the system is always running at a slow rate and is ready to accelerate at any time. This allows the alternative pathway to play an important role in recognition of foreign invaders as part of the innate immune system and also allows it to play a role in eliminating dead or damaged body cells. However, it also means that the system must be carefully regulated, as discussed in Control of Complement Activation.
C3 and C4 are highly homologous. Similarly, the alternative and lectin pathway C3 convertases are also highly homologous with the classical pathway convertase and are formed in an analogous way. , , Activation of the alternative pathway can be initiated when a molecule of “C3b-like C3,” or C3b itself, which acts like C4b in the classical pathway, binds with a molecule called factor B , which is homologous with C2. , The factor B is then cleaved by factor D, a proteolytic enzyme that seems to be constitutively activated by a third MASP. Factor D can act on factor B only when it is held in the proper configuration by C3b. Cleavage of factor B by D is analogous to cleavage of C2 by C1s, but the large fragment is named Bb. , , Like C2, factor B has latent protease activity. The Bb resulting from cleavage by factor D has its proteolytic active site exposed and can cleave C3 and C5. The C3b or C3b-like C3 molecule that initiated the pathway also serves the same role as C4b in the classical pathway C3 convertase, that is, holding additional molecules of C3 substrate in place, so they can be cleaved by Bb. Just as in the classical pathway, some of the newly cleaved C3b molecules will deposit sufficiently close to the C3bBb enzyme to join with it as a subunit that provides a binding site for C5, which can then be cleaved by Bb. Thus, the addition of this new C3b molecule changes the alternative and lectin pathway C3 convertase C3bBb into the alternative pathway C5 convertase C3bBb3b. , ,
The alternative pathway depends on an additional molecule called properdin , or factor P, which is not found in the classical or lectin pathways. , , The discovery of this molecule led to the elucidation of the alternative pathway, so it is sometimes called the properdin pathway . P stabilizes the alternative pathway convertases, which would otherwise lose activity when Bb, with its proteolytic active site, diffuses away. , , Thus, surfaces or targets that favor binding of P favor activation of the alternative pathway. Stabilization of the convertase may also occur pathologically in the presence of C3 nephritic factor, an autoantibody with properdin-like properties found in some patients with glomerulonepritis and in patients with gain-of-function mutations, which confer increased stability on the convertase (see below).
One of the most dramatic features of complement activation is the conversion of individual, water-soluble, circulating proteins into a large multichain lipophilic assembly that can insert itself into plasma membranes, , forming a pore that spans both leaflets, compromising its barrier function. Insertion of just one of these channels into an erythrocyte allows enough water to rush into the cell to cause explosive lysis. This ability of a heat-labile fraction of plasma proteins to lyse erythrocytes sensitized by heat-stabile antibodies was recognized more than 100 years ago and led to the elucidation of the complement reaction sequence, because the liberated hemoglobin provided a convenient end point for assays. Arguably, however, complement-mediated lysis is not the most important function of the complement system in vivo, because nucleated cells can internalize or shed bits of membrane bearing C5 through C9 complexes. The transmembrane pore formed by the assembly of C5 through C9 is called the MAC and its formation begins with activation of C5 by proteolytic cleavage.
Regardless of the pathway by which C5 is cleaved, the remaining steps of activation of C6 through C9 follow the same sequence (see Fig. 120.2 ). Unlike the early activation pathways, however, activation beyond C5 and formation of the MAC do not involve proteolytic cleavages. C5 does not contain the same internal thioester as its homologues C3 and C4, and C5b cannot bind covalently to the target. However, just after cleavage, C5b gains the ability to interact with C6. Binding with C6 stabilizes and increases the lipophilicity of C5b, and the C5b6 complex can insert into lipid membranes at some distance from the C5 convertase or even on nearby cells. This phenomenon can lead to bystander lysis of normal cells in the vicinity of soluble immune complexes or bacteria on which complement is being activated. C5b6 can also bind a molecule of C7. The complex of C5b67 is highly lipophilic and, if formed in the fluid phase, will rapidly insert into plasma membranes, where it serves as a binding site for C8. Binding of C8 to C5b67 creates a complex that can disrupt the phospholipids of target cell membranes and can cause slow lysis of erythrocytes, even without the addition of C9. The C5b8 complex also induces conformational changes in soluble C9 molecules. , These include elongation and unfolding of the C9 molecules, exposing previously protected hydrophobic regions and potential sites for disulfide bonds between chains. , In turn, additional C9 molecules change conformation, polymerize and join the complex. Isolated C9 can also be induced to undergo these conformational changes and polymerize in vitro under carefully defined chemical conditions. C9 polymers may contain up to 12 molecules linked in dimers by disulfide bonds. Electron micrographs of these C9 polymers look very similar to MACs isolated from cells or artificial membranes that have been attacked by complement. They appear like grommets or donuts, with a top rim 15 to 20 nm in diameter sitting above a slightly narrower cylinder 15 to 16 nm in length, which span the plasma membrane. , The conformational changes that accompany C9 polymerization are recognized by monoclonal antibodies that do not bind to native, circulating C9 molecules. These antibodies to the C9 “neoantigen” formed upon activation may be used to identify sites of MAC deposition in tissue sections. Antibody sandwich assays against the soluble C5b-9 complex (sC5b-9) are now widely used in enzyme-linked immunosorbent assays of complement activity.
Besides the ability to lyse cells, complement activation liberates C3a and C5a, which are potent mediators of inflammation. The deposition of large amounts of C3b can opsonize, or target, our own cells for attack by neutrophils and/or destruction by macrophages in the reticuloendothelial system, just as it can opsonize invading microorganisms. The apparent dichotomy between complement’s role as a part of the innate immune system that is always turned on to detect foreign invaders and its potential for damage and destruction of host cells and tissues requires that a delicate balance be maintained at all times. This balance depends on a set of control mechanisms no less intricate than the activation pathways.
The need to control alternative pathway activation becomes apparent when one considers the fact that the alternative pathway can be initiated whenever and wherever C3b is formed, whether by any of the activation pathways, by nonspecific action of a serine protease, or even when C3b-like molecules are formed spontaneously as the thioester is lysed by water. , , The alternative pathway is often visualized as an amplification loop that can feed forward whenever C3b is formed (see Fig. 120.2 ). The danger of uncontrolled alternative pathway activation is illustrated by observations in a patient with congenital homozygous deficiency of the complement regulatory protein factor H (see later). This patient lacked detectable C3 activity in his serum because most of the C3 was consumed as fast as it was produced, by continuous nonspecific activation of the alternative pathway. The lack of C3 caused increased susceptibility to infection. Incongruously, some of the C3b deposited on his red blood cells caused a mild but chronic hemolytic anemia.
First of all, the activity of complement is limited by the low circulating concentrations of most of its components. However, C3, C4, and other components are acute-phase reactants, whose synthesis by hepatocytes and other types of cells increases rapidly in response to cytokines, hormones, and other signals. Thus, local synthesis and changes in vascular permeability determine the amounts of the components available at any site or time. The activation cascades are inherently limited by the instability of the multisubunit convertases. In these bimolecular and trimolecular complexes, the proteolytic active sites are on the C2a and Bb fragments, which are not covalently bound to the target or to the other subunits of the convertases. These proteolytically active subunits have relatively weak, noncovalent interactions with C4b and/or C3b and can easily dissociate and diffuse away. The mechanism by which P facilitates activation of the alternative pathway is by holding together the two subunits of the noncovalent C3bBb enzyme so that it will maintain activity. , In contrast, dissociation of the proteolytically active subunit (C2a or Bb, respectively, in the classical/lectin vs. alternative pathways) causes loss, or decay , of the convertase activity. Decay may occur spontaneously, or it may be accelerated by proteins that bind to the covalently attached member of the convertase and push away the loosely held enzymatically active subunit. Several proteins have this function; some are soluble and some are membrane bound ( Table 120.1 ). Many of these proteins share homology with each other, as expected because they all bind to C3, C4, or both, which are themselves homologous. These proteins belong to the regulators of complement activation (RCA) family and are encoded by genes in a cluster on chromosome 1q32. , Members of this family are all composed of variable numbers of short homologous repeats, each of which has 60 to 65 amino acids and two internal disulfide bonds, which give it a double-looped structure. These are further grouped into long homologous repeats, which form the binding sites for C4b and C3b. , A prototypic member of this family is called decay accelerating factor (DAF) (CD55). , , DAF is bound to the plasma membrane by a glycolipid anchor, which is believed to increase its mobility in the membrane, to better prevent attack of complement on our own cells. The major function of DAF is to bind to C3b and C4b molecules and push off Bb or C2a. Besides the ability to bind to and push off (or block binding of) the active subunits of convertases, some of these regulatory proteins can also facilitate or serve as cofactors for degradation of the C4b or C3b to which they bind. A circulating protease, called C3b/C4b inactivator (factor I), can cleave C3b and C4b into forms that no longer bind other components of the convertases. Factor I can cleave C3b and C4b only in the presence of an additional cofactor that renders C3b and C4b susceptible to cleavage. , Soluble cofactors include C4-binding protein (C4bp) and complement factor H (CFH, formerly termed β-1-H globulin ). Although it can bind to C3b, DAF does not have cofactor activity. However, several other membrane proteins have this kind of cofactor activity, including complement receptor type 1 (C3b receptor, CR1, CD35) and a separate membrane cofactor protein (MCP) (CD46). DAF and MCP bind only to C3b and C4b that have been deposited on the same cell on which they reside, so they are considered important in protecting our own cells from becoming activators or targets of complement. C4bp, CFH, and CR1 can also bind to C3b on other cells or circulating complexes, so they have an important role in processing these complexes. The enzymatically inactive fragments are termed iC4b and iC3b , respectively. The latter still serves as an important opsonin, however, by binding to its own receptor, termed CR3 .
Protein a | Abbreviation | Molecular Weight | No. of Short Consensus Repeats (SCRs) | Dissociation of C3 and C5 Convertases | Factor I Cofactor Activity on | ||
---|---|---|---|---|---|---|---|
Alternative | Classical | C3b | C4b | ||||
Factor H, β 1H | H | 150,000 | 20 | + | − | + | − |
C4-binding protein | C4bp | 570,000 (Octamer) | 8 | − | + | − | + |
Decay accelerating factor (CD55) | DAF | 70,000–80,000 | 4 | + | + | − | − |
Membrane cofactor protein (CD46) | MCP | 45,000–70,000 | 4 | − | − | + | + |
Complement receptor type 1 C3b/C4b receptor (CD35) | CR1 | 205,000–250,000 | 22–30 | + | + | + | + |
Complement receptor type 2 C3d receptor (CD21) | CR2 | 145,000 | 15/16 | − | − | − | − |
a Factor H and C4bp are soluble. All others are membrane proteins indicated by CD number in parentheses.
From the viewpoint of a C3b molecule, regulation of the alternative pathway might be seen as a competition between binding of factor B, which would lead to additional activation, and factor H, which would lead to inactivation. , , The enzymes that execute those actions, factor D and factor I, respectively, both circulate in their active forms. An important determinant of the fate of any given C3b molecule is the chemical nature of the surface on which it is bound, which plays a critical role in influencing the binding of B versus H. Surfaces that are rich in sialic acid, like our own cells, favor the binding of H and thus promote the action of I. , These surfaces are thus poor activators of the alternative pathway. In contrast, many bacteria and cells from some other species of mammals lack sialic acid. On these surfaces, binding of B is favored. Those cells are good activators of the alternative pathway. Thus, the binding of P and H versus B serves to distinguish non-self from self, even in the absence of a specific antibody or T-cell receptors. Interestingly, bacteria such as K12 Escherichia coli and type III group B Streptococcus have adapted by adding terminal sialic acid residues to their surfaces or capsular polysaccharides, making them dangerous pathogens, particularly for newborns who lack specific antibody. Related bacteria without the sialic acid are much less virulent. Antibody molecules provide good acceptor sites for C3b deposition, because B is favored over H, and bound C3b is protected against inactivation. This allows the alternative pathway to strongly and rapidly amplify initial signals generated by antibodies and the classical pathway. Besides their coating with sialic acid, most of our own cells are protected from amplification of the alternative pathway by DAF and MCP. Some cell types, including erythrocytes, are also similarly protected by CR1.
The importance of these protein-protein interactions that regulate the alternative pathway, particularly on our own cells, is illustrated by the occurrence of the atypical hemolytic-uremic syndrome (aHUS) in patients in whom the alternative complement pathway escapes from control because of loss-of-function mutations in factor H, factor I, or MCP, or gain-of-function mutations in C3 or factor B. More than 300 disease-causing mutations in these proteins have been described. Clinically, aHUS includes the triad of microangiopathic hemolytic anemia, thrombocytopenia, and renal failure. Rarely, aHUS may present in the first weeks or months of life, even without infectious diarrhea or another trigger. Exogenous triggers are more typical in hemolytic-uremic syndrome caused by toxin-producing E. coli. There have even been reports of perinatal asphyxia due to anemia caused by aHUS. The damage to the red cells and consumption of platelets are believed to be consequences of excessive complement activation and MAC formation on the endothelial cells. Endothelial cell damage in the mother may occur when complement control protein mutations are present in women who develop antiphospholipid or other autoantibodies. This type of endothelial damage likely contributes to many cases of preeclampsia; the HELLP syndrome (hypertension; hemolysis, elevated liver enzymes, and low platelets; intrauterine growth restriction (IUGR); and even some cases of recurrent fetal loss (see below).
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here