The Croonian Lecture, 1980 - The complex proteases of the complement system

Abstract

The assembly and activation of the early components of complement, after their interaction with antibody--antigen complexes, are described in terms of the structures of the different proteins taking part. C1q, a molecule of unique half collagen--half globular structure, binds to the second constant domain of the antibody molecules through its six globular heads. A tetrameric complex of C1r$_{2}$-C1s$_{2}$ binds to the collagenous tails and leads to formation of the serine-type proteases C$\overline{1}$r and C$\overline{1}$s. C$\overline{1}$s activates C4, which forms a covalent bond between its $\alpha ^{\prime}$ chain and the Fab section of the antibody. C2 is also activated by C$\overline{1}$s and associates with the bound C$\overline{4}$ molecule to form C$\overline{42}$, a labile protease that activates C3, but which loses activity as the C$\overline{2}$ peptide chains dissociate from C$\overline{4}$. C2, by analogy with factor B, the equivalent component of the alternative pathway of activation, appears to be a novel type of serine protease with a similar catalytic site but different activation mechanism to the serine proteases that have been described previously. The assembly of the proteases on the aggregated antibody can be summarized as shown on figure 14. The globular heads of the hexameric C1q molecule interact with the C$_{\text{H}}$2 domains of several Fc sections of the aggregated antibody molecules. This interaction appears to be principally ionic as it can be prevented by high salt concentration (Burton et al. 1980). The C1r$_{2}$-Ca$^{2+}$-C1s$_{2}$ tetramer binds to the collagenous tails of the C1q, possibly to the N-terminal end, which contains the interchain disulphide bonds before the collagenous sequence starts. The binding of the tetramer is through C1r and it is possible that the binding site is formed jointly by the collagenous tails of C1q and the Fc section of the antibody. C1s does not bind unless C1r and Ca$^{2+}$ are present and may therefore interact only with C1r, not with C1q nor antibody. Activation of C1r follows the binding of the C1r-C1s tetramer to C1q bound to the antibody, probably by exposure of a proteolytic site in the unsplit C1r peptide chain. This site in C1r catalyses the formation of C$\overline{1}$r, which in turn activates C1s. It is C$\overline{1}$s that activates C4 and C2. The C4 molecule forms a covalent bond with the Fab part of the antibody, probably through reaction of an active acyl group in the $\alpha ^{\prime}$ chain with the Fd section of the heavy chain. This postulated acyl group also reacts readily with water and productive binding to the Fab occurs only if activation of C4 occurs close to the aggregates by C$\overline{1}$s bound to the antibody in C$\overline{1}$. The activated C4 may react with itself to form dimers and higher aggregates and also with cell surface components. The C2, activated by C$\overline{1}$s, binds non-covalently to the bound C4, to from C3 convertase. This enzyme has a half life of about 0.35 min, activity being lost by the spontaneous dissociation of C2a, the section of the molecule carrying the proteolytic site. It is probable that the whole C2 interacts with C4b, and that hydrolysis of the activation bond by C$\overline{1}$s occurs on the bound C2. This causes formation of the proteolytic site in the C2a fragment and, as C2a has a lower affinity for the C4b, also initiates dissociation. C2a, by analogy with Bb, is likely to be a serine protease as all the invariant residues from the catalytic site of this class of protease are present in the expected relative positions. However, there are about 300 additional amino acid residues N-terminal to the catalytic chain found in all other serine protease investigated so far and the typical N-terminal sequence is missing, suggesting that the activation mechanism is different. As C2a cannot hydrolyse C3 except when in association with C4b it is probable that C3 interacts simultaneously with C4b and C2a. This would be in agreement with the data of Kerr (1980), who showed that the half life of the C3 convertase appears to be increased several-fold in the presence of excess C3, i.e. the substrate stabilizes the C4b--C2a complex. C3 convertase may also release an active acyl group in the C3$\alpha ^{\prime}$ chain, leading to covalent bond formation, perhaps with the Fab section of the antibody or perhaps with C4b bound to the Fab. The attachment of C3b facilitates hydrolysis of C5 by the proteolytic site in C2a. It is a complex picture, in which the key features appear to be: (i) the attachment of protease to the antibody aggregate through a linking protein, either C1q for C$\overline{1}$r and C$\overline{1}$s or C4 for C$\overline{2}$; (ii) the necessity for two or three proteins to be associated before proteolytic activation of the next component in the chain can occur. Attachment has the advantage that it localizes the proteases formed and permits a temporary excess over the high concentration of proteolytic inhibitors in the blood and lymph and hence enables the cascade to continue. It also has the advantage that proteolytic damage of adjacent tissues is minimized. The bringing together of several distinct components for one catalytic activity is probably responsible for the extreme specificity of several of the proteases and this will reduce non-specific proteolytic damage. As in several cases, dissociation is rapid; this offers an additional method of control. Several features described here, such as the C1q structure, covalent bond formation and the necessity for several proteins to associate to achieve effective biological activity, appear unusual or even unique to the complement system. There is, however, already evidence to suggest that similar features will be found in cell surface receptors and in other biological activities of cell membranes.