3D structure of the C3bB complex provides insights into the activation and regulation of the complement alternative pathway convertase

Abstract

Generation of the alternative pathway C3-convertase, the central amplification enzyme of the complement cascade, initiates by the binding of factor B (fB) to C3b to form the proconvertase, C3bB. C3bB is subsequently cleaved by factor D (fD) at a single site in fB, producing Ba and Bb fragments. Ba dissociates from the complex, while Bb remains bound to C3b, forming the active alternative pathway convertase, C3bBb. Using single-particle electron microscopy we have determined the 3-dimensional structures of the C3bB and the C3bBb complexes at approximately 27A resolution. The C3bB structure shows that fB undergoes a dramatic conformational change upon binding to C3b. However, the C3b-bound fB structure was easily interpreted after independently fitting the atomic structures of the isolated Bb and Ba fragments. Interestingly, the divalent cation-binding site in the von Willebrand type A domain in Bb faces the C345C domain of C3b, whereas the serine-protease domain of Bb points outwards. The structure also shows that the Ba fragment interacts with C3b separately from Bb at the level of the alpha'NT and CUB domains. Within this conformation, the long and flexible linker between Bb and Ba is likely exposed and accessible for cleavage by fD to form the active convertase, C3bBb. The architecture of the C3bB and C3bBb complexes reveals that C3b could promote cleavage and activation of fB by actively displacing the Ba domain from the von Willebrand type A domain in free fB. These structures provide a structural basis to understand fundamental aspects of the activation and regulation of the alternative pathway C3-convertase.

Fig. 1.

Electron microscopy and 3D reconstruction of C3b and C3bB(Ni²⁺). (A) Reference-free 2D averages obtained for the data set containing images of single molecules of C3b. These averages reveal a characteristic “L” shape evocative of the C3b crystal structure. (B) Reference-free 2D averages of C3bB(Ni²⁺) display a bulky appearance compatible with fB binding to C3b. (C) Front view of the 3D structure of C3b derived from the EM data at a resolution of 28 Å (gray density). The atomic structure of C3b (PDB file 2i07) has been fitted within the EM map and displayed in purple with the C345C and CUB domains highlighted in orange and blue, respectively. (D) Several views of the 3D structure of C3bB(Ni²⁺) at 27 Å resolution (gray density). Fitting of the atomic structure of C3b (PDB file 2i07) allows the assignment of specific regions of the EM map to specific C3b domains. Some densities of the 3D reconstruction cannot be accounted by C3b (asterisks) and correspond to C3b-bound fB.

Fig. 2.

Open conformation of fB within the C3bB(Ni²⁺) complex.(A) Views of the 3D reconstruction of C3bB(Ni²⁺). C3b in the complex has been colored in orange whereas fB is shown in green. (B) The structure of inactive fB (PDB file 2OK5) shows a compact conformation (Left). This structure was filtered to ≈25 Å resolution and was compared with the EM reconstruction (Top Left). C3b-bound fB was extracted by calculating the difference map between the EM density for C3bB(Ni²⁺) and the fitted atomic structure of C3b (Right), and was found to display a more open conformation. SP, vWA, and SCR1–3 domains have been colored in pale blue, green, and yellow, respectively. (C) Fitting of the atomic structure of fB into the structure of C3bB(Ni²⁺) complex. Factor B structure was divided in 2 segments corresponding to Bb and Ba fragments, and fitted separately within the density assigned to fB in the complex.

Fig. 3.

Structural insights in the assembling of the proconvertase. (A) Representation of an atomic model for fB within the C3bB(Ni²⁺) complex. For clarity only the C345C domain of C3b is represented. Color codes for domains as in Fig. 2. Specific residues have been highlighted, representing them as space-filled amino acids. These include D279 (known to affect proenzyme formation), K323 (known to affect regulation by DAF), and Q34 (to label the N terminus of the Ba fragment). The vWA α1 helix (contributing to the C3b-binding region) and vWA α4/5 helix (implicated in the DAF/CR1 binding site) are highlighted in blue and red ribbons, respectively. The N terminus of the C3b α′ chain (α′NT) is depicted with space-filled amino acids. (B) A side view of the structural model of the C3bB(Ni²⁺) complex where the atomic structure of C3b (PDB file 2i07) is also represented in purple color. Color codes as in Figs. 1 and 2.

Fig. 4.

Three-dimensional structure of the C3bBb convertase. (A) A reference-free 2D average corresponding to a side view of the C3bBb complex, where the vWA and SP domains appear projecting from the C3b structure. (B) Two views of the C3bBb complex revealing that the density assigned to the SCR1–3 domains in the structure of C3bB(Ni²⁺) is missing. This reconstruction represents a 3D average, where the density of the SP domain is blurred because of conformational flexibility. (C) The flexibility of the vWA-SP cassette projecting from the C3bBb complex is reflected in the 2D reference-free averages of the data. Whereas some averages show a good definition of 3 dots of density corresponding to the C345C, vWA and SP domains (i), others reveal some blurring of this area (ii), while maintaining the definition of the C3b molecule.