Structure of the C1 complex of complement

In PNAS, Mortensen et al. (1) propose a model of the C1 complex of complement derived from small-angle X-ray scattering (SAXS) and electron microscopy (EM) analyses that contradicts previously published models and suggests an intermolecular activation process. This proposal is largely derived from a conjectural structure of C1r₂C1_s2, the catalytic unit of C1, in which the serine protease (SP) domains of both C1r and C1s are located at the periphery of the tetramer. Such an arrangement is clearly not consistent with the multiple biochemical, biophysical, and structural studies carried out so far (reviewed in ref. 2).

Thus, the X-ray structures (2) showing that C1r dimerizes through interactions between the SP domain of one monomer and the CCP1 domain of its counterpart are fully consistent with ultracentrifugation (3) and EM analyses (4, 5) of the C1r CCP1-CCP2-SP domains, indicating that these domains form homodimers. The dimeric X-ray structures are therefore unlikely to result from crystal packing interactions, as implied by Mortensen et al. (1). In contrast, the central core of the C1r₂C1_s2 model proposed by Mortensen et al. (1) is not supported by experimental evidence in solution (6). Clearly, all EM studies performed previously are consistent with a model in which the SP domains of C1s lie at both ends of the C1s-C1r-C1r-C1s tetramer (5) and the SP domains of C1r lie in the center, with C1s–C1r interactions being mediated by their CUB1-EGF-CUB2 domains (4–7).

It is puzzling that most of the C1r₂C1_s2 EM images shown in figure 3 of ref. 1 exhibit rather straight structures with split ends, in sharp contrast to the typical S-shaped structures obtained by other groups (refs. 5, 7 and our own observations). A possible explanation for this discrepancy is that only selected 2D class-averaged images are shown. Because of the modular structures and flexibility of C1r and C1s, superposition of prominent conformations comprising flexible moieties during particle averaging may result in split ends.

That the isolated tetramer is highly extended is well established from EM (5, 7), ultracentrifugation (7), and neutron scattering studies (8). The radius of gyration (Rg) determined by the latter technique (17 nm) was even significantly higher than the 13-nm value measured by Mortensen et al. (1). However, association of C1r₂C1_s2 with C1q (Rg = 12.8 nm) resulted in a value of only 12.6 nm for C1 (8), providing a strong indication that the tetramer folds into a more compact conformation in the complex. Finally, it should be stressed that EM analysis of C1 was reported to require prior stabilization of the complex by chemical cross-linking (9) and showed that C1r₂C1_s2 formed a compact mass centrally located on C1q, in agreement with neutron scattering. Using another strategy to stabilize C1 through interaction with hexameric immune complexes, it was not possible to resolve the position of the proteolytic arms of C1r and C1s in isosurface density EM maps of the averaged antibody–C1 complexes (10). In contrast, the interaction domains of C1r and C1s were found to cluster in a planar molecular platform inside C1, consistent with their interaction with C1q lysines B61 and C58 (2).

Further appropriate investigations are required to reveal the details of the structure and activation of C1.