The 'safety catch' on the trigger for formation of the membrane attack complex - a structural analysis

The job of the body's immune system is to detect and to then destroy invading cells, such as bacteria, which threaten our health and can ultimately kill us. This proposal is concerned with potent weapons employed by the body to destroy bacterial cells, but which sometimes go astray and damage our own tissues. The ability of components within human blood to burst open foreign cells was first recognised almost a century ago and represented one of the great early discoveries of immunology. Today, the eleven proteins involved (called C1, C2, C3 etc. up to C9 and factors B and D) have been thoroughly characterised. The sequence in which they act (roughly starting with C1 and culminating with C9) is also well established. The grand finale is the self-assembly on cell membranes of a ring of C9 proteins that punches a hole (or pore) right through the membrane. Many thousands of such pores can form on a cell resulting in leakage of its contents, an influx of water, swelling and rupture ('cytolysis'). It is clearly vital to minimise the number of such pores that form on our own cells. Unfortunately this does not always happen. Many diseases are associated with failure to control MAC formation on human cells and there is a lot of interest in developing inhibitors of the MAC as potentially powerful drugs. One way in the body controls MAC assembly is by ensuring that formation of the pore itself can occur only after C6 and C7 have assembled on C5b to form a 'pre-lytic' complex. The key point is that C5b is not normally present in the blood in substantial quantities. Its creation (by conversion from inactive C5) is triggered by detection of bacteria or other hazardous particles. In this proposal we want to take a very close look at the role played by C5b in nucleating MAC formation. In particular we will focus our efforts on the proteins C7 and C6. We suspect that C7, like C8 and C9, contains the molecular equivalent of a flick-knife fitted with a safety catch. Release of the safety catch results in the flipping out of a blade like structure that effectively cuts through the membrane. One proposal is that C5b is necessary for release of the safety catch in C7 and we wish to investigate whether this is true and if it is, then we want to find out precisely how this comes about. This will involve taking a very detailed look at the structure of the safety catch and its mode of interaction. In our model, the safety catch consists of a swinging arm of C7 that interacts with itself to maintain MACPF in a state of latency, or with C5b to release the MACPF. We propose to dissect the arm, into its component domains and study their structures, how flexibly they are attached to one another and how the arm is disposed relative to the MACPF domain in C7. These insights will be useful in the future design of small molecules designed to prevent the safety catch from releasing. Such molecules would be the starting points for therapeutic compounds in clinical settings where dangerous levels of the MAC form on human cell surfaces.

The self-assembly of a membrane-penetrating pore (the membrane-attack complex or MAC) from its soluble plasma protein components C5b, C6, C7, C8 and C9 is the best known, least well understood, and most destructive outcome of the complement system. Deficiencies increase susceptibility to bacterial infection; nonetheless, MAC assembly is a candidate for therapeutic intervention where need to modulate inflammation outweighs accompanying elevated risks of infection. The sequence of MAC-forming events was established > 25 years ago, while more recently structure determination of the C8alpha MACPF domain revealed the molecular 'flick knife' potential of this domain-type. But molecular mechanisms initiating MAC assembly are poorly understood. We shall interrogate a hypothetical mechanism for assembly of the C5b.C6.C7 (C5b67) complex that associates with membranes and recruits bilayer-penetrating C8 and C9 components. We will focus on a putative molecular 'safety catch'. This consists of an array of C-terminal modules in C7 that, we think, prevents the C7 MACPF domain undergoing the critical and irreversible transition to a membrane-associative form. In our model, release of the safety catch is contingent upon relocation of C-terminal modules from MACPF to a binding site on C5b (in the context of C5b67 complex) thus ensuring MAC assembly can occur only following activation of C5 to C5b. We will test key predictions of our model regarding the 3D solution structure and flexibility of the C-terminus of C7 and its relationship with the MACPF domain. Results will be extrapolated to the close homologue, C6. Chemical cross-linking and small-angle X-ray scattering will guide modelling of C7 and C6 substructures and their complexes with C5b. This data will illuminate the steps that couple complement amplification with the lytic terminal pathway. It will provide sound foundations upon which to select the optimal target for therapeutic reagents.

We will continue to work with the Edinburgh University Press Office and the University's technology transfer company (ERI Ltd) to ensure the results of our research on the membrane attack complex reach both the public and relevant commercial audiences. Where appropriate we will file for patent protection as we have done with our work on expression of factor H. Furthermore, the results of the proposed research may help build the foundations needed to establish the C5b-C6/C7 interaction as a target for the powerful array of in-house ligand discovery platforms (ultrahigh-throughput screening, peptide hit morphing etc.) housed within our new Centre for Translational and Chemical Biology. There is a very clear unmet need for therapeutic interventions that limit complement action. As was illustrated in Fig. 1 (Part 1a) the last proteolytic step in the amplification/activation phase of the pathway is the cleavage of C5 into the anaphylatoxin C5a and the MAC-initiating C5b fragment. Binding of nascent C5b (C5b*) to C6 and then to C7 initiates self-assembly of the MAC, which has the structure C5b1C61C71C81C9n, where n is between about 10 and 16. It is important to note that all the activation products responsible for stimulation of potentially beneficial cellular immune responses are already generated before MAC assembly is initiated. Unregulated or inappropriate complement activation is a, if not the, major contributor to a variety of pathologies including age-related macular degeneration. Inhibitors that block specifically the terminal pathway are advantageous since they avoid compromise of the more beneficial complement functions. Furthermore, long-term MAC inhibition as might be needed in chronic conditions may have only mild side-effects, based on the observation that the consequences of genetic deficiencies in MAC proteins are manageable. Finally, blocking incorporation of C6 or C7 into the nascent MAC is the most desirable goal of all because C5b and C5b-6 do not damage the plasma membrane, whereas even the C5b67 intermediate can induce inflammatory responses.