Crystal structure of C5b-6 suggests structural basis for priming assembly of the membrane attack complex

Abstract

The complement membrane attack complex (MAC) forms transmembrane pores in pathogen membranes. The first step in MAC assembly is cleavage of C5 to generate metastable C5b, which forms a stable complex with C6, termed C5b-6. C5b-6 initiates pore formation via the sequential recruitment of homologous proteins: C7, C8, and 12-18 copies of C9, each of which comprises a central MAC-perforin domain flanked by auxiliary domains. We recently proposed a model of pore assembly, in which the auxiliary domains play key roles, both in stabilizing the closed conformation of the protomers and in driving the sequential opening of the MAC-perforin β-sheet of each new recruit to the growing pore. Here, we describe an atomic model of C5b-6 at 4.2 Å resolution. We show that C5b provides four interfaces for the auxiliary domains of C6. The largest interface is created by the insertion of an interdomain linker from C6 into a hydrophobic groove created by a major reorganization of the α-helical domain of C5b. In combination with the rigid body docking of N-terminal elements of both proteins, C5b becomes locked into a stable conformation. Both C6 auxiliary domains flanking the linker pack tightly against C5b. The net effect is to induce the clockwise rigid body rotation of four auxiliary domains, as well as the opening/twisting of the central β-sheet of C6, in the directions predicted by our model to activate or prime C6 for the subsequent steps in MAC assembly. The complex also suggests novel small molecule strategies for modulating pathological MAC assembly.

FIGURE 1.

Structure of the C5b-6 complex. A, stereo view of the C5b-6 complex shown as Cα chain trace. C5b is depicted in blue, green, and yellow; C6 is depicted in yellow, orange, and red. B, C5b-6 shown as molecular surface representation, colored as in A. The left view is the same as in A, and the right view is rotated by 90° about a vertical axis. C, domain organizations of C5 and C6, colored as in A and B. N-Linked carbohydrates are indicated by black hexagons, and disulfide linkages are indicated by brackets. The C5 convertase cleavage site in C6, which leads to the loss of C5a and formation of C5b, is indicated by an arrow.

FIGURE 2.

Structural differences between C5b and C5 versus C3b and C3. A, side by side comparison of C5b and C5, overlaid on the MG scaffold. The large (50 Å) motion of the C5d domain is illustrated by overlaying its C5d location onto the C5 structure (right panel). The major binding site for C6 (the TS3-CCP1 linker) is located in C5d and is indicated by a black helix in both C5 and C5b, illustrating how it is cryptic in C5 and becomes exposed in C5b. B, a similar comparison of C3b and C3, illustrating the larger movement of the C3d domain compared with C5d; the extension of the CUB-C3d linker; and the distinct final location of C5d, which packs against the body of MG1 domain. The thioester site is shown as a black ball, illustrating its analogous location and exposure compared with the C6 binding site in C5d (black helix).

FIGURE 3.

Structure of the C5b-C6 interface. A, overview of the binding site of C6 with C5b, illustrating the major elements of the interface (with the exception of the domain-exchanged FIM-C345C interaction, shown in Fig. 3E). The major interaction between the TS3-CCP1 linker is only partly visible in this view and is shown in detail in Fig. 4. B, close-up of the interface between the LR (magenta cartoon; side chains colored by atom-type; disulfide bonds in yellow; Ca²⁺ ion is shown as a black ball) and TS2 domains (in blue) of C6 with the base of the MG scaffold of C5b (shown as molecular surface). The top of the MACPF domain is shown as a green ribbon. C, close-up of the interface between the CCP modules (yellow cartoon with atom-colored side chain sticks and selected residues labeled) of C6 with C5b, shown as a molecular surface colored by surface potential. Red, negative; blue, positive; white, neutral. CCP1 packs into a large groove formed by the juxtaposition of the C5d, CUB, and MG2 domains. CCP2 makes limited contacts, exclusively with C5d. D, same as C, but the view is orthogonal (about vertical axis), revealing the tight packing at the base and loose packing at the top of the pocket. E, a three-dimensional slice through the crystal of C5b-6 showing the interchange of C-terminal arms across a 2-fold axis of symmetry (black lens at center) that enables binding of FIMs from one C5b-6 complex with the C345C domain from the symmetry-related complex. F, close-up view of the FIM2-C345C interaction. C345C is shown as colored ribbons (colored as in E), the FIMs as electrostatic potential surfaces.

FIGURE 4.

The TS3-CCP1 of C6 linker interface with the C5d domain of C5. A, structure of the linker region of C6 (ribbon main chain and side chains as sticks, with atomic coloring) inserted into a groove on the surface of C5d (colored by electrostatic potential). The regions (TS3, L1, and L2) are labeled as in C. B, stereo superposition of the linker region of C6 (cyan) onto a cartoon (cylindrical helices and main chain loops) overlay of C5d (red) and C5 (gray helices, green loops), showing the multiple changes in helical elements and loops, as well as a shift in helical register between residues 993 and 1010, that together create the major groove observed in A. Arrows point to the connections with the CUB domain. C, sequence alignments of C6 (upper sequence) and C7 (lower sequence) TS3-CCP1 linker regions among vertebrate orthologs, including the cartilaginous fish, the shark. In C6, the first loop at the top of TS3 is discontinuous with the adjacent linker sequence, but joined by a disulfide bond (Cys⁵⁵⁶–Cys⁵⁹⁰ in human). The major interaction sites are boxed, revealing strong conservation of sequence, as well as length between the delimiting cysteines. Homologous cysteines in C6 and C7 are linked by blue arrows. C7 also has conserved motifs, but they are distinct from those in C6. Notably, the hydrophobic motif (FSIM in human C6) that inserts into C5d is absent in C7. Furthermore, the lengths of the intracysteine segments are different: S1 is longer in C7, whereas S2 is much shorter. The numbering of human C6 sequence does not include the leader peptide.

FIGURE 5.

Conformational changes in C6 induced by complex formation with C5b. A, ribbon diagram of the C6 (with the CCP modules and FIMs removed for clarity), indicating the three segments: lower (L), upper (U), and regulatory (R) that we previously defined as subdomains that rotate as rigid bodies about two hinge points (one at the bend in the β-sheet, the second at the end of the Linchpin helix), based on comparisons of C6 and C8 (14). B, unliganded C6 (red) and C5b-bound C6 (cyan) overlaid on their upper segments, which are identical within experimental error (root mean square deviation for 125 Cα = 0.51 Å). The upper panel, which is viewed from the “outside” of the nascent pore, reveals a downward shift of TS3 by ∼5 Å, and a concerted rotation of TS1, TS2, and TS3, three elements of the regulatory segment (the fourth element, the EGF domain, cannot be seen in this view. The largest rotation (∼20°) occurs in TS1, leading to a shift of ∼20 Å at its tip. Smaller rotations occur nearer to the top of the TS2, which is constrained by its contact with the MG scaffold. The lower panel is an “inside” view of C6 (i.e., looking from what will become the lumen of the pore. The counterclockwise twist of the β-sheet (∼10°) occurs about hinge point 1, at the bend in the β-sheet (labeled H1), and is similar but smaller to that observed in C8α (see Ref. 14). The CH1–3 elements rotate in concert with the sheet. C6 is red, and C5b-6 C6 is cyan. C, comparison between C6 (in red), C6 in the C5b-6 complex (C6′ in cyan), and C8β (in yellow), aligned on the upper segment (the root mean square difference for the overlaid upper segments of C6 and C8β is 0.8 Å for 92 on Cαs). The view is similar to that in B (upper panel) but cut away to reveal the rotation of the EGF domains in concert with the downward motion of the TS3 domains. The rotation is about the second hinge point (H2). The total rotation of the EGF domain is ∼18°, in the direction of the incoming recruit to the nascent pore. Note that liganded C6 (C6′) lies about halfway between unliganded C6 and C8β (∼8° rotation); note also the intermediate bend at the base of the Linchpin helix. C8β provides one model of the “open” conformation (see Ref. 14). Thus, C5b appears to prime C6 by inducing conformational changes toward the open conformation.