- Email: [email protected]
Structure and activation of the C1 complex of complement: unraveling the puzzle
Review
TRENDS in Immunology
Vol.25 No.7 July 2004
Structure and activation of the C1 complex of complement: unraveling the puzzle Christine Gaboriaud, Nicole M. Thielens, Lynn A. Gregory, Ve´ronique Rossi, Juan C. Fontecilla-Camps and Ge´rard J. Arlaud Laboratoire de Cristallographie et Cristalloge´ne`se des Prote´ines (CG, LAG and JCF-C) and Laboratoire d’Enzymologie Mole´culaire (NMT, VR and GJA), Institut de Biologie Structurale Jean Pierre Ebel, CEA-CNRS-Universite´ Joseph Fourier, 41, rue Jules Horowitz, 38027 Grenoble Cedex 1, France
C1, the multimolecular protease that triggers the classical pathway of complement, has a major role in the host defense against pathogens. It also participates in other biological functions, such as immune tolerance, owing to the ability of its binding subunit, C1q, to recognize abnormal structures from self, including apoptotic cells. Structural biology has been used over the past few years to elucidate the structure of its three subunits: C1q, C1r and C1s. These new advances have led to a comprehensive, three-dimensional model of C1 and provide insights into the mechanisms underlying its activation and the extraordinarily versatile recognition properties of its C1q subunit. The C1 complex of complement has long been recognized as an important component of antimicrobial host defense, owing to the known ability of its recognition subunit, C1q, to bind pathogens either directly or after their recognition by antibodies or C-reactive protein (CRP) [1,2]. Through activation of the classical complement pathway, this binding elicits a series of effector mechanisms aimed at limiting infection. This traditional portrait of C1 has been largely modified over the past few years by the discovery that C1q has the striking ability to recognize abnormal structures from self. Thus, there is growing evidence that b-amyloid fibrils [3,4] and the pathological form of the prion protein [5,6], are recognized by C1q. It is also established that C1q binds to, and induces clearance of, apoptotic cells [7], consequently acting as a key factor in immune tolerance [8]. These findings have led to a renewed interest in C1. C1 is a 790 kDa complex formed by the association of a recognition protein, C1q, and a Ca2þ-dependent tetramer comprising two copies of two proteases, C1r and C1s [1,9]. Binding of C1 to a target cell or molecule is mediated by C1q, and is thought to elicit a signal that triggers self-activation of C1r, which in turn converts proenzyme C1s into the highly specific protease that ultimately cleaves C4 and C2, thereby triggering the classical complement pathway. C1 function is Corresponding author: Ge´rard J. Arlaud ([email protected]). Available online 6 May 2004
regulated by C1 inhibitor, a member of the serine protease inhibitor (SERPIN) family, which controls both C1 activation and proteolytic activity [9]. C1q is a hexameric protein with the overall shape of a bouquet of tulips, comprising six heterotrimeric collagen-like triple-helical fibers that associate to form an N-terminal ‘stalk’. Because of interuptions in the repeating Gly– X – Y collagen sequence, these diverge to form six individual ‘stems’, each terminating in a heterotrimeric globular ‘head’ [10], which is involved in the recognition function of C1. C1r and C1s have homologous modular architectures (Figure 1c), with five non-catalytic protein modules preceding a chymotrypsin-like serine protease (SP) domain. Whereas the C-terminal complement control protein 1 (CCP1) – CCP2 – SP regions of C1r and C1s mediate their enzymatic properties, their N-terminal CUB1 [module originally found in C1r and C1s, uEGF (epidermal growth factor), bone morphogenetic protein 1]– EGF (epidermal growth factor) segments mediate the Ca2þ-dependent C1r– C1s interactions involved in assembly of the C1s –C1r– C1r– C1s tetramer. In addition, both segments contribute ligands for interaction between the tetramer and the individual collagen-like stems of C1q [11]. This domain structure of C1r and C1s, along with information provided by electron microscopy and neutron scattering analyses [12,13], have led to the concept that, in the C1 complex, C1s– C1r– C1r– C1s folds into a compact ‘8-shaped’ conformation, enabling contact between the catalytic regions of C1r and C1s, a prerequisite for C1s activation. This concept has provided the basis for most of the lowresolution C1 models proposed thus far [14 – 16]. X-ray crystallography has been used over the past few years to generate more detailed information about the structure of C1 at the atomic level. The dissection strategy used involves recombinant expression of modular segments from each C1 protein to characterize their function and solve their three-dimensional structure. The purpose of this Review is to describe these recent advances, which shed light on the threedimensional structure of C1 and the molecular mechanisms underlying its target recognition and catalytic functions.
www.sciencedirect.com 1471-4906/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.it.2004.04.008
Review
TRENDS in Immunology
369
Vol.25 No.7 July 2004
(a)
(b)
(c)
(d) CCP2 A
SP A C1rA
CUB1 EGF CUB2 CCP1 CCP2
SP
C1rB
CUB1 EGF CUB2 CCP1 CCP2
SP
C1s
CUB1 EGF CUB2 CCP1 CCP2
SP
CCP1 A
Catalytic site CCP2 B CCP1 B
(e)
(f)
SP B Activation site
(g)
CUB1
EGF
CUB2 CCP1
CCP2 SP
SP
CCP2 CCP1
CUB1
EGF
CUB2
TRENDS in Immunology
Figure 1. Three-dimensional structural model of the human C1 complex. (a) Side view and (b) bottom view of the C1q molecule. C1q is shown in the ‘open’ conformation proposed to mediate C1 activation. The C1q chains are colored blue (A), green (B) and red (C). The N-linked oligosaccharide attached to each globular head is represented in yellow. The arrows indicate the position of lysine residues A59 and B61 proposed to mediate interaction with C1r and C1s [18]. Most of the other lysines in the collagen stems carry O-linked disaccharides [19]. (c) Modular structure of C1r and C1s. Both proteases comprise an N-terminal CUB module, a Ca2þ-binding EGF-like module, a second CUB module, two CCP modules and a chymotrypsin-like SP domain (nomenclature and symbols are as defined in Ref. [20]). The color-coding used differentiates the C1rA and B molecules and also applies to parts (d– g). C1r and C1s are activated through cleavage of a single Arg– Ile bond, indicated by the arrows. N-linked oligosaccharides are represented by closed diamonds. (d) Homodimeric structure of the C1r catalytic region [adapted with permission from Ref. [23] (http://www.nature.com/ emboj/)]. (e) Side view of the C1 complex. (f) Bottom view of the tetramer alone. Labels indicate the location of individual modules within one C1r and one C1s subunit. (g) Bottom view of the C1 complex. C1 is depicted in the resting state, in which C1q is proposed to have a ‘closed’ conformation. Abbreviations: CCP, complement control protein; CUB, module originally found in C1r and C1s, uEGF (epidermal growth factor) and bone morphogenetic protein; SP, serine protease.
www.sciencedirect.com
370
Review
TRENDS in Immunology
A refined, three-dimensional model of C1 In addition to mediating the recognition function of the C1 complex, C1q also provides a scaffold for proper folding of the catalytic subunit C1s– C1r– C1r– C1s. Therefore, correct modeling of this protein is a prerequisite for building a precise three-dimensional model of C1. Resolution of the X-ray structure of the C1q globular head has revealed the arrangement of the three subunits (A, B, C clockwise, when the head is viewed from the top), enabling us to derive a three-dimensional model of the collagenlike triple-helix of C1q [17]. The resulting C1q model (Figure 1a, b) has several features relevant to its function. Thus, the fact that the B module of each globular head lies on the external part of the molecule, whereas the A and C modules are positioned inside, has direct implications for ligand recognition, as discussed later. Incidentally, this configuration places the N-linked oligosaccharide attached to each A module [10] in between two neighboring globular heads, a circle-like disposition that appears more appropriate than either external or internal locations, which would possibly interfere with ligand recognition. NMR spectroscopy [21] and X-ray crystallography [18,22 – 24] have been applied to several modular fragments from human C1r and C1s, leading to the resolution of a large part (67% and 72%, respectively) of the structure of these proteases. The proenzyme structure of the catalytic region of C1r [23], responsible for the initial activation step of C1, reveals a homodimeric head-to-tail assembly held together by interactions between the CCP1 module of one monomer and the SP domain of its counterpart, with a large opening in the center (Figure 1d). This region lies in the middle of the cone defined by the C1q stems (Figure 1g), and the structure determined by X-ray crystallography, which is trapped in the proenzyme form by means of a mutation at the Arg –Ile cleavage site, probably represents a resting state. The C1s catalytic regions, which mediate proteolytic activity of C1, occupy both ends of the C1s– C1r– C1r– C1s tetramer (Figure 1f). The X-ray structure of the C-terminal moiety at this region, comprising the CCP2 module and the SP domain, has been solved in the active form, showing that CCP2 is oriented perpendicularly to the surface of the SP domain and closely interacts with it by means of a rigid interface [22]. The structure of the CUB1 –EGF interaction region of C1s has also been determined recently [18], revealing a head-to-tail homodimer involving interactions between the CUB1 module of one monomer and the EGF module of its counterpart, and a structure that is strongly stabilized by Ca2þ ions. This structure was a major breakthrough because it led to a three-dimensional model of the C1r– C1s CUB1 –EGF heterodimer, which in the C1 complex connects C1r to C1s and mediates interaction with C1q. Based on this work and other known characteristics of C1 assembly (reviewed in Ref. [11]), a structural model of the C1q– C1r– C1s interface was derived [18], in which one of the triple-helical collagen stems of C1q fits into a groove along the transversal axis of the C1r– C1s CUB1– EGF heterodimer (Figure 1e). The major interaction involves ionic bonds between unmodified lysine residues located approximately half-way along the collagen-like stem of www.sciencedirect.com
Vol.25 No.7 July 2004
C1q (Figure 1a, b) and acidic residues contributed by an unusually large and highly acidic loop of the C1r EGF module, which appears to be mobile in the structure derived from NMR spectroscopy [21]. The described model provides, for the first time, precise information about the structure and location of the C1q – C1r– C1s interface, a keystone of C1 architecture, and can now be used as a guideline for further investigation of this interface. There are still some missing links in the C1s– C1r–C1r– C1s structure, namely the C1r CUB1 and CUB2 modules and the C1s CUB2– CCP1 segment, for which no structure is available. Nevertheless, reasonably accurate three-dimensional models of these modules can be obtained based on the structure of their counterparts in C1r (C1s CCP1) or C1s (C1r CUB1) and, for the CUB2 modules, on the X-ray structure of the CUB1– EGF– CUB2 segment of mannan-binding lectinassociated serine protease-2, a protease homologous to C1r and C1s [25]. Altogether, the information provided by structural biology enables us to generate a refined, threedimensional version of the low-resolution models initially proposed for the C1 complex [14 – 16] (Figure 1e– g). Of additional interest in the C1s CUB1– EGF X-ray structure [18] is that it sheds light on the role of Ca2þ ions in C1 assembly. In agreement with both the C1r and C1s EGF modules belonging to the Ca2þ-binding subset [26], the C1s CUB1– EGF structure shows a Ca2þ ion bound to each EGF module that stabilizes both the intra- and intermonomer interfaces. Unexpectedly, the structure reveals that a second Ca2þ ion is bound to, and stabilizes, the distal end of each CUB1 module. Because this latter Ca2þ ion is coordinated by three acidic residues that are strictly conserved in about two-thirds of the CUB repertoire, including the CUB1 module of C1r, it can be inferred that assembly of the C1s– C1r– C1r– C1s tetramer is stabilized by eight Ca2þ ions, that is, two in each CUB1– EGF pair. In addition, one Ca2þ ion is bound to each globular head of C1q [17] and therefore fourteen Ca2þ ions are expected to be contained in the C1 complex. A point that differentiated the macroscopic models initially proposed for C1 was whether C1s– C1r–C1r– C1s folds around two opposite C1q stems [15] or around two opposite pairs of stems [14,16]. Based on the size of the C1s domains relative to the space available in between two C1q stems, it can now be concluded that the folding process required to enable the C1s SP domain to reach the inside of C1 involves two neighboring stems on each side, one mediating the interface with C1r and C1s, the other providing an axis for the swing of the C1s SP domain. As discussed later, this feature has direct functional implications, notably on the accessibility of the C1s CCP1 module, which is required for C4 cleavage by C1s [27]. C1 activation: a random, mechanical process? In addition to revealing the details of their structure, X-ray analysis of the C1r and C1s catalytic regions also provides precise insights into the molecular mechanisms underlying C1 activation. From a functional standpoint, the most intriguing feature of the C1r catalytic region head-totail structure [23] is that the catalytic site of one monomer and the activation site of the other lie at opposite ends
Review
TRENDS in Immunology
of the dimer (Figure 1d). This configuration probably represents a resting state of the molecule, possibly designed to prevent spontaneous C1 activation. Indeed, it is not compatible with C1r self-activation, which requires cleavage of the susceptible Arg – Ile bond of each monomer by the catalytic Ser residue of its counterpart [28], implying a direct contact between the two SP domains through disruption of the head-to-tail structure seen in the crystal. This has led us to the proposal that the signal that triggers C1r activation in C1 is a mechanical stress transmitted through the C1q – C1r– C1s interface from a C1q stem to the C1r catalytic region when C1 binds to a target. This mechanism is consistent with the presence of a semi-flexible hinge in C1q at the point where the collagen stems join to form the stalk [29]. Thus, it is conceivable that multivalent binding of C1q through its globular heads to an irregular cluster of distant binding sites will make some of the stems swing away from the center of the C1 complex, thereby generating the activating signal. The interaction of the two SP domains is obviously rendered possible by the presence of the large opening in the center of the C1r catalytic region (Figure 1d). Furthermore, it is likely to be facilitated by inter-modular flexibility in C1r, particularly at the CUB2 – CCP1 and CCP1 – CCP2 junctions, as documented in the latter case for other CCP module pairs [30]. The activation process possibly also involves restrained flexibility at the CCP2 – SP interface, as revealed by a comparison of the three available X-ray structures containing this region [23,24]. Further insights into the C1r activation process arise from the crystal contacts observed in the zymogen and active C1r CCP2– SP structures [24]. Thus, in the wildtype active species, the SP domains are packed in such a way that they form an enzyme-product-like complex, probably similar to the one occurring on activation of one C1r molecule by its counterpart (Figure 2b). Several transient states, such as the one observed in the proenzyme CCP2 – SP crystal where the SP domains interact in a nearly symmetrical manner [24], are probably necessary to reach this final activation state. This scenario
(a)
might apply to either C1r molecule, which might activate in any order because the two opposite C1q stems attached to the C1r– C1s interaction regions are expected to move in a random fashion, depending on whether their respective globular head is engaged in ligand recognition. C1s activation requires appropriate positioning of its Arg–Ile cleavage site within the active site cleft of the corresponding C1r molecule (Figure 2c). This might involve a movement of the C1s SP domain towards the interior of the C1 complex, or a movement of the C1r SP domain in the opposite direction, or both. Again, these shifts might be facilitated by the presence of flexible hinges at the CUB2– CCP1 and CCP1–CCP2 junctions in both C1r and C1s. It is tempting to postulate that C1s uses its CCP1–CCP2 interface to swing around a C1q stem. Once activated, the C1s SP domain is expected to move towards its protein substrates C4 and C2. In the case of C4, this would also enable proper positioning of the C1s CCP1 module relative to the active site, a probable prerequisite for efficient recognition and cleavage of this substrate [27]. As for C1r, it is likely that, once fully activated, its catalytic region folds back into a head-to-tail dimeric structure. This hypothesis is ˚ resolution of the strongly supported by the structure at 4 A active C1r catalytic region, which reveals an overall configuration strikingly similar to that of the resting proenzyme form [23]. This also appears to be consistent with our knowledge of the reaction of C1 inhibitor with activated C1, resulting in disassembly of the tetramer into two C1 inhibitor–C1s–C1r–C1 inhibitor complexes [23,31]. This scenario is largely hypothetical and obviously requires further experimental support, particularly with respect to the mechanisms that initiate activation on the target surface. Nevertheless, it seems clear that C1 activation is a mechanical process relying on displacement of the C1q stems on binding of C1 to an activating target. Owing to the random character of these movements, the C1 complex should not be viewed as a structurally homogenous molecule but as a mixture of conformational states differing in the activation status of C1r and/or C1s.
(b)
C1sA SP
371
Vol.25 No.7 July 2004
(c)
C1rA SP C1rB SP
C1sB SP
TRENDS in Immunology
Figure 2. Tentative scheme for C1 activation, as seen from bottom views of the complex. (a) C1s –C1r– C1r–C1s in the proenzyme, 8-shaped resting state. The sliding movement required to disrupt the resting head-to-tail C1r– C1r dimer is indicated by arrows. (b) S-shaped conformation featuring activation of a C1r SP domain by its counterpart. (c) Transient conformation featuring activation of C1sA by C1rB. C1q is shown in the ‘closed’ conformation in (a) and in the ‘open’ conformation in (b) and (c). Abbreviation: SP, serine protease. www.sciencedirect.com
Review
372
TRENDS in Immunology
Structural basis for the versatile recognition function of C1q One of the most striking properties of C1q lies in its ability to recognize an amazing variety of targets, including immunoglobulins, CRP, b-amyloid fibrils, the prion protein, lipid A, DNA, various microorganisms and apoptotic cells. There is no obvious structural feature shared by these diverse ligands, however, the fact that many polyionic structures are recognized by C1q [1], suggests that it might function as a charge pattern recognition molecule. The X-ray structure of the heterotrimeric C1q globular head [17] yields insights into its recognition properties. The three modules exhibit striking differences in their surface distribution of charged and hydrophobic residues. Therefore, the pseudo-threefold symmetry seen at the framework level disappears when surface patterns are considered (Figure 3a). Thus, individually, each of the three subunits can be expected to fulfill specific interaction properties. However, as illustrated in Figure 3a, the
(a)
(b)
(iii)
(ii)
(i)
(c)
Vol.25 No.7 July 2004
globular head forms a compact structure through the tight interaction of the three modules. Obviously, this type of structure also enables ligand recognition through residues contributed by two or even three subunits, thus providing a basis for the versatility of the recognition properties of C1q. One of the many possible recognition modes of C1q is illustrated by CRP, an acute phase protein that binds to phosphocholine groups of membrane phospholipids and is, in turn, recognized by C1q [2]. In the interaction model proposed on the basis of available structural and mutagenesis data [32– 34], the top of the C1q head, which is predominantly basic, is accommodated by the negatively charged central pore of C-reactive protein (Figure 3b), an interaction that involves the three C1q subunits. By contrast, the model proposed for interaction with IgG b12, a human IgG1 molecule, is an example of a recognition mediated by a single subunit of the C1q head (Figure 3c,d). Based on the X-ray structure of IgG b12 [35] and available information about the C1q– IgG interaction [36 –39], the model features binding of the equatorial region of subunit B of C1q at the Fab– Fc interface [17]. This raises the interesting possibility that C1q binds not only to the Fc domain but also to additional sites in the Fab arm, underscoring the possible role of the hinge region of IgG as a limiting factor of C1q recognition. Given the putative location of the C1q B module on the external part of the protein (Figure 1), the above model appears particularly well adapted to the recognition of IgG molecules within an immune complex network. Alternative recognition modes involving the internal A and/or C modules of C1q can be postulated, particularly in the case of the oligomeric IgM structure, in agreement with recent findings [40].
(d)
Fc
Fc
C1q
Fab
C1q Fab
TRENDS in Immunology
Figure 3. The versatile recognition function of C1q. (a) Top view of the C1q globular head, showing the different surface patterns of the three subunits, as illustrated by the distribution of basic (blue), acidic (red) and hydrophobic (yellow) residues. (i), (ii) and (iii) indicate A, B and C subunits. The lines indicate the approximate subunit boundaries. (b) Model of the C1q–CRP interaction. Two protomers of the pentameric CRP structure have been removed for clarity. (c,d) Model of the interaction between C1q and IgG1 b12. The IgG Fc domain and Fab arms are indicated. The C1q subunits are colored blue (A), green (b) and red (C). CRP (b) and IgG1 (c,d) are colored yellow, with functional residues identified by mutagenesis highlighted in different colors. Abbreviations: CRP, C-reactive protein. Adapted with permission from Ref. [17]. www.sciencedirect.com
Conclusions and perspectives The combined use of a dissection strategy and of structural biology techniques has led to a refined, three-dimensional model of the C1 complex of complement, providing for the first time a sound structural basis for the mechanisms underlying its activation process and its recognition properties. Understanding the details of the assembly and function of C1 at the atomic level will now require resolution of the remaining unknown three-dimensional structures within C1, and the use of site-directed mutagenesis to map the interaction sites within the complex and to further decipher the mechanisms involved in C1 activation. Detailed knowledge of the mechanisms underlying the extremely versatile recognition properties of C1q will also involve resolution of the structure of its globular head in complex with some of its ligands. Undoubtedly, a further key issue will be to investigate the molecular dynamic properties of C1, particularly the many areas of flexibility of its individual subunits, which enable the whole complex to mediate its finely tuned function. Acknowledgements We are indebted to Isabelle Bally, Beate Bersch, Monika Budayova-Spano, Claudine Darnault, Jordi Juanhuix, Monique Lacroix and David Pignol, who contributed to many of the studies reported in this review.
Review
TRENDS in Immunology
Vol.25 No.7 July 2004
References 1 Cooper, N.R. (1985) The classical pathway of complement: activation and regulation of the first complement component. Adv. Immunol. 37, 151 – 216 2 Szalai, A.J. et al. (1999) C-reactive protein: structural biology and host defense function. Clin. Chem. Lab. Med. 37, 265– 270 3 Rogers, J. et al. (1992) Complement activation by b-amyloid in Alzheimer disease. Proc. Natl. Acad. Sci. U. S. A. 89, 10016 – 10020 4 Tacnet-Delorme, P. et al. (2001) b-amyloid fibrils activate the C1 complex of complement under physiological conditions. Evidence for a binding site for Ab on the C1q globular regions. J. Immunol. 167, 6374 – 6381 5 Mabbott, N.A. et al. (2001) Temporary depletion of complement component C3 or genetic deficiency of C1q significantly delays onset of scrapie. Nat. Med. 7, 485 – 487 6 Klein, M.A. et al. (2001) Complement facilitates early prion pathogenesis. Nat. Med. 7, 488 – 492 7 Navratil, J.S. et al. (2001) The globular heads of C1q specifically recognize surface blebs of apoptotic vascular endothelial cells. J. Immunol. 166, 3231– 3239 8 Botto, M. and Walport, M.J. (2002) C1q, autoimmunity and apoptosis. Immunobiology 205, 395 – 406 9 Arlaud, G.J. et al. (2001) Structural biology of C1: dissection of a complex molecular machinery. Immunol. Rev. 180, 136 – 145 10 Kishore, U. and Reid, K.B.M. (2000) C1q: structure, function, and receptors. Immunopharmacology 49, 159 – 170 11 Arlaud, G.J. et al. (2002) Structural biology of the C1 complex of complement unveils the mechanisms of its activation and proteolytic activity. Mol. Immunol. 39, 383 – 394 12 Strang, C.J. et al. (1982) Ultrastructure of the first component of human complement: electron microscopy of the crosslinked complex. Proc. Natl. Acad. Sci. U. S. A. 79, 586 – 590 13 Perkins, S.J. et al. (1984) Neutron scattering studies of subcomponent C1q of first component C1 of human complement and its association with C1r2C1s2 within C1. J. Mol. Biol. 179, 547 – 557 14 Weiss, V. et al. (1986) Functional model of subcomponent C1 of complement. J. Mol. Biol. 189, 573 – 581 15 Arlaud, G.J. et al. (1987) A functional model of the human C1 complex. Immunol. Today 8, 106 – 111 16 Schumaker, V.N. et al. (1987) Activation of the first component of complement. Annu. Rev. Immunol. 5, 21 – 42 17 Gaboriaud, C. et al. (2003) The crystal structure of the globular head of complement protein C1q provides a basis for its versatile recognition properties. J. Biol. Chem. 278, 46974 – 46982 18 Gregory, L.A. et al. (2003) X-ray structure of the Ca2þ-binding interaction domain of C1s. Insights into the assembly of the C1 complex of complement. J. Biol. Chem. 278, 32157 – 32164 19 Reid, K.B.M. (1979) Complete amino acid sequences of the three collagen-like regions present in subcomponent C1q of the first component of human complement. Biochem. J. 179, 367 – 371 20 Bork, P. and Bairoch, A. (1995) Extracellular modules. Trends Biochem. Sci. 20 (Suppl), C03 21 Bersch, B. et al. (1998) Solution structure of the epidermal growth factor (EGF)-like module of human complement protease C1r, an atypical member of the EGF family. Biochemistry 37, 1204– 1214 22 Gaboriaud, C. et al. (2000) Crystal structure of the catalytic domain of
www.sciencedirect.com
23
24
25
26 27
28
29
30
31
32 33
34 35
36
37 38
39
40
373
human complement C1s: a serine protease with a handle. EMBO J. 19, 1755– 1765 Budayova-Spano, M. et al. (2002) The crystal structure of the zymogen catalytic domain of complement protease C1r reveals that a disruptive mechanical stress is required to trigger activation of the C1 complex. EMBO J. 21, 231 – 239 Budayova-Spano, M. et al. (2002) Monomeric structures of the zymogen and active catalytic domains of complement protease C1r: further insights into the C1 activation mechanism. Structure 10, 1509– 1519 Feinberg, H. et al. (2003) Crystal structure of the CUB1 – EGF – CUB2 region of mannose-binding protein associated serine protease-2. EMBO J. 22, 2348 – 2359 Campbell, I.D. and Bork, P. (1993) Epidermal growth factor-like modules. Curr. Opin. Struct. Biol. 3, 385 – 392 Rossi, V. et al. (1998) Baculovirus-mediated expression of truncated modular fragments from the catalytic region of human complement serine protease C1s. Evidence for the involvement of both complement control protein modules in the recognition of the C4 protein substrate. J. Biol. Chem. 273, 1232 – 1239 Lacroix, M. et al. (2001) Assembly and enzymatic properties of the catalytic domain of human complement protease C1r. J. Biol. Chem. 276, 36233 – 36240 Poon, P.H. et al. (1983) Conformation and restricted segmental flexibility of C1, the first component of human complement. J. Mol. Biol. 168, 563 – 577 Kirkitadze, M.D. and Barlow, P.N. (2001) Structure and flexibility of the multiple domain proteins that regulate complement activation. Immunol. Rev. 180, 146– 161 Sim, R.B. et al. (1979) C1 inhibitor-dependent dissociation of human complement component C1 bound to immune complexes. Biochem. J. 179, 449 – 457 Shrive, A.K. et al. (1996) Three-dimensional structure of human C-reactive protein. Nat. Struct. Biol. 3, 346 – 354 Thompson, D. et al. (1999) The physiological structure of human C-reactive protein and its complex with phosphocholine. Struct. Fold. Des. 7, 169 – 177 Agrawal, A. et al. (2001) Topology and structure of the C1q-binding site on C-reactive protein. J. Immunol. 166, 3998 – 4004 Saphire, E.O. et al. (2001) Crystal structure of a neutralizing human IgG against HIV-1: a template for vaccine design. Science 293, 1155– 1159 Idusogie, E.E. et al. (2000) Mapping of the C1q binding site on rituxan, a chimeric antibody with a human IgG1 Fc. J. Immunol. 164, 4178– 4184 Idusogie, E.E. et al. (2001) Engineered antibodies with uncreased activity to recruit complement. J. Immunol. 166, 2571 – 2575 Hezareh, M. et al. (2001) Effector function activities of a panel of mutants of a broadly neutralizing antibody against human immunodeficiency virus type 1. J. Virol. 75, 12161 – 12168 Marques, G. et al. (1993) Arginine residues of the globular regions of human C1q involved in the interaction with immunoglobulin G. J. Biol. Chem. 268, 10393 – 10402 Kishore, U. et al. (2003) Modular organization of the carboxylterminal, globular head region of human C1q A, B, and C chains. J. Immunol. 171, 812– 820
TRENDS in Immunology
Vol.25 No.7 July 2004
Structure and activation of the C1 complex of complement: unraveling the puzzle Christine Gaboriaud, Nicole M. Thielens, Lynn A. Gregory, Ve´ronique Rossi, Juan C. Fontecilla-Camps and Ge´rard J. Arlaud Laboratoire de Cristallographie et Cristalloge´ne`se des Prote´ines (CG, LAG and JCF-C) and Laboratoire d’Enzymologie Mole´culaire (NMT, VR and GJA), Institut de Biologie Structurale Jean Pierre Ebel, CEA-CNRS-Universite´ Joseph Fourier, 41, rue Jules Horowitz, 38027 Grenoble Cedex 1, France
C1, the multimolecular protease that triggers the classical pathway of complement, has a major role in the host defense against pathogens. It also participates in other biological functions, such as immune tolerance, owing to the ability of its binding subunit, C1q, to recognize abnormal structures from self, including apoptotic cells. Structural biology has been used over the past few years to elucidate the structure of its three subunits: C1q, C1r and C1s. These new advances have led to a comprehensive, three-dimensional model of C1 and provide insights into the mechanisms underlying its activation and the extraordinarily versatile recognition properties of its C1q subunit. The C1 complex of complement has long been recognized as an important component of antimicrobial host defense, owing to the known ability of its recognition subunit, C1q, to bind pathogens either directly or after their recognition by antibodies or C-reactive protein (CRP) [1,2]. Through activation of the classical complement pathway, this binding elicits a series of effector mechanisms aimed at limiting infection. This traditional portrait of C1 has been largely modified over the past few years by the discovery that C1q has the striking ability to recognize abnormal structures from self. Thus, there is growing evidence that b-amyloid fibrils [3,4] and the pathological form of the prion protein [5,6], are recognized by C1q. It is also established that C1q binds to, and induces clearance of, apoptotic cells [7], consequently acting as a key factor in immune tolerance [8]. These findings have led to a renewed interest in C1. C1 is a 790 kDa complex formed by the association of a recognition protein, C1q, and a Ca2þ-dependent tetramer comprising two copies of two proteases, C1r and C1s [1,9]. Binding of C1 to a target cell or molecule is mediated by C1q, and is thought to elicit a signal that triggers self-activation of C1r, which in turn converts proenzyme C1s into the highly specific protease that ultimately cleaves C4 and C2, thereby triggering the classical complement pathway. C1 function is Corresponding author: Ge´rard J. Arlaud ([email protected]). Available online 6 May 2004
regulated by C1 inhibitor, a member of the serine protease inhibitor (SERPIN) family, which controls both C1 activation and proteolytic activity [9]. C1q is a hexameric protein with the overall shape of a bouquet of tulips, comprising six heterotrimeric collagen-like triple-helical fibers that associate to form an N-terminal ‘stalk’. Because of interuptions in the repeating Gly– X – Y collagen sequence, these diverge to form six individual ‘stems’, each terminating in a heterotrimeric globular ‘head’ [10], which is involved in the recognition function of C1. C1r and C1s have homologous modular architectures (Figure 1c), with five non-catalytic protein modules preceding a chymotrypsin-like serine protease (SP) domain. Whereas the C-terminal complement control protein 1 (CCP1) – CCP2 – SP regions of C1r and C1s mediate their enzymatic properties, their N-terminal CUB1 [module originally found in C1r and C1s, uEGF (epidermal growth factor), bone morphogenetic protein 1]– EGF (epidermal growth factor) segments mediate the Ca2þ-dependent C1r– C1s interactions involved in assembly of the C1s –C1r– C1r– C1s tetramer. In addition, both segments contribute ligands for interaction between the tetramer and the individual collagen-like stems of C1q [11]. This domain structure of C1r and C1s, along with information provided by electron microscopy and neutron scattering analyses [12,13], have led to the concept that, in the C1 complex, C1s– C1r– C1r– C1s folds into a compact ‘8-shaped’ conformation, enabling contact between the catalytic regions of C1r and C1s, a prerequisite for C1s activation. This concept has provided the basis for most of the lowresolution C1 models proposed thus far [14 – 16]. X-ray crystallography has been used over the past few years to generate more detailed information about the structure of C1 at the atomic level. The dissection strategy used involves recombinant expression of modular segments from each C1 protein to characterize their function and solve their three-dimensional structure. The purpose of this Review is to describe these recent advances, which shed light on the threedimensional structure of C1 and the molecular mechanisms underlying its target recognition and catalytic functions.
www.sciencedirect.com 1471-4906/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.it.2004.04.008
Review
TRENDS in Immunology
369
Vol.25 No.7 July 2004
(a)
(b)
(c)
(d) CCP2 A
SP A C1rA
CUB1 EGF CUB2 CCP1 CCP2
SP
C1rB
CUB1 EGF CUB2 CCP1 CCP2
SP
C1s
CUB1 EGF CUB2 CCP1 CCP2
SP
CCP1 A
Catalytic site CCP2 B CCP1 B
(e)
(f)
SP B Activation site
(g)
CUB1
EGF
CUB2 CCP1
CCP2 SP
SP
CCP2 CCP1
CUB1
EGF
CUB2
TRENDS in Immunology
Figure 1. Three-dimensional structural model of the human C1 complex. (a) Side view and (b) bottom view of the C1q molecule. C1q is shown in the ‘open’ conformation proposed to mediate C1 activation. The C1q chains are colored blue (A), green (B) and red (C). The N-linked oligosaccharide attached to each globular head is represented in yellow. The arrows indicate the position of lysine residues A59 and B61 proposed to mediate interaction with C1r and C1s [18]. Most of the other lysines in the collagen stems carry O-linked disaccharides [19]. (c) Modular structure of C1r and C1s. Both proteases comprise an N-terminal CUB module, a Ca2þ-binding EGF-like module, a second CUB module, two CCP modules and a chymotrypsin-like SP domain (nomenclature and symbols are as defined in Ref. [20]). The color-coding used differentiates the C1rA and B molecules and also applies to parts (d– g). C1r and C1s are activated through cleavage of a single Arg– Ile bond, indicated by the arrows. N-linked oligosaccharides are represented by closed diamonds. (d) Homodimeric structure of the C1r catalytic region [adapted with permission from Ref. [23] (http://www.nature.com/ emboj/)]. (e) Side view of the C1 complex. (f) Bottom view of the tetramer alone. Labels indicate the location of individual modules within one C1r and one C1s subunit. (g) Bottom view of the C1 complex. C1 is depicted in the resting state, in which C1q is proposed to have a ‘closed’ conformation. Abbreviations: CCP, complement control protein; CUB, module originally found in C1r and C1s, uEGF (epidermal growth factor) and bone morphogenetic protein; SP, serine protease.
www.sciencedirect.com
370
Review
TRENDS in Immunology
A refined, three-dimensional model of C1 In addition to mediating the recognition function of the C1 complex, C1q also provides a scaffold for proper folding of the catalytic subunit C1s– C1r– C1r– C1s. Therefore, correct modeling of this protein is a prerequisite for building a precise three-dimensional model of C1. Resolution of the X-ray structure of the C1q globular head has revealed the arrangement of the three subunits (A, B, C clockwise, when the head is viewed from the top), enabling us to derive a three-dimensional model of the collagenlike triple-helix of C1q [17]. The resulting C1q model (Figure 1a, b) has several features relevant to its function. Thus, the fact that the B module of each globular head lies on the external part of the molecule, whereas the A and C modules are positioned inside, has direct implications for ligand recognition, as discussed later. Incidentally, this configuration places the N-linked oligosaccharide attached to each A module [10] in between two neighboring globular heads, a circle-like disposition that appears more appropriate than either external or internal locations, which would possibly interfere with ligand recognition. NMR spectroscopy [21] and X-ray crystallography [18,22 – 24] have been applied to several modular fragments from human C1r and C1s, leading to the resolution of a large part (67% and 72%, respectively) of the structure of these proteases. The proenzyme structure of the catalytic region of C1r [23], responsible for the initial activation step of C1, reveals a homodimeric head-to-tail assembly held together by interactions between the CCP1 module of one monomer and the SP domain of its counterpart, with a large opening in the center (Figure 1d). This region lies in the middle of the cone defined by the C1q stems (Figure 1g), and the structure determined by X-ray crystallography, which is trapped in the proenzyme form by means of a mutation at the Arg –Ile cleavage site, probably represents a resting state. The C1s catalytic regions, which mediate proteolytic activity of C1, occupy both ends of the C1s– C1r– C1r– C1s tetramer (Figure 1f). The X-ray structure of the C-terminal moiety at this region, comprising the CCP2 module and the SP domain, has been solved in the active form, showing that CCP2 is oriented perpendicularly to the surface of the SP domain and closely interacts with it by means of a rigid interface [22]. The structure of the CUB1 –EGF interaction region of C1s has also been determined recently [18], revealing a head-to-tail homodimer involving interactions between the CUB1 module of one monomer and the EGF module of its counterpart, and a structure that is strongly stabilized by Ca2þ ions. This structure was a major breakthrough because it led to a three-dimensional model of the C1r– C1s CUB1 –EGF heterodimer, which in the C1 complex connects C1r to C1s and mediates interaction with C1q. Based on this work and other known characteristics of C1 assembly (reviewed in Ref. [11]), a structural model of the C1q– C1r– C1s interface was derived [18], in which one of the triple-helical collagen stems of C1q fits into a groove along the transversal axis of the C1r– C1s CUB1– EGF heterodimer (Figure 1e). The major interaction involves ionic bonds between unmodified lysine residues located approximately half-way along the collagen-like stem of www.sciencedirect.com
Vol.25 No.7 July 2004
C1q (Figure 1a, b) and acidic residues contributed by an unusually large and highly acidic loop of the C1r EGF module, which appears to be mobile in the structure derived from NMR spectroscopy [21]. The described model provides, for the first time, precise information about the structure and location of the C1q – C1r– C1s interface, a keystone of C1 architecture, and can now be used as a guideline for further investigation of this interface. There are still some missing links in the C1s– C1r–C1r– C1s structure, namely the C1r CUB1 and CUB2 modules and the C1s CUB2– CCP1 segment, for which no structure is available. Nevertheless, reasonably accurate three-dimensional models of these modules can be obtained based on the structure of their counterparts in C1r (C1s CCP1) or C1s (C1r CUB1) and, for the CUB2 modules, on the X-ray structure of the CUB1– EGF– CUB2 segment of mannan-binding lectinassociated serine protease-2, a protease homologous to C1r and C1s [25]. Altogether, the information provided by structural biology enables us to generate a refined, threedimensional version of the low-resolution models initially proposed for the C1 complex [14 – 16] (Figure 1e– g). Of additional interest in the C1s CUB1– EGF X-ray structure [18] is that it sheds light on the role of Ca2þ ions in C1 assembly. In agreement with both the C1r and C1s EGF modules belonging to the Ca2þ-binding subset [26], the C1s CUB1– EGF structure shows a Ca2þ ion bound to each EGF module that stabilizes both the intra- and intermonomer interfaces. Unexpectedly, the structure reveals that a second Ca2þ ion is bound to, and stabilizes, the distal end of each CUB1 module. Because this latter Ca2þ ion is coordinated by three acidic residues that are strictly conserved in about two-thirds of the CUB repertoire, including the CUB1 module of C1r, it can be inferred that assembly of the C1s– C1r– C1r– C1s tetramer is stabilized by eight Ca2þ ions, that is, two in each CUB1– EGF pair. In addition, one Ca2þ ion is bound to each globular head of C1q [17] and therefore fourteen Ca2þ ions are expected to be contained in the C1 complex. A point that differentiated the macroscopic models initially proposed for C1 was whether C1s– C1r–C1r– C1s folds around two opposite C1q stems [15] or around two opposite pairs of stems [14,16]. Based on the size of the C1s domains relative to the space available in between two C1q stems, it can now be concluded that the folding process required to enable the C1s SP domain to reach the inside of C1 involves two neighboring stems on each side, one mediating the interface with C1r and C1s, the other providing an axis for the swing of the C1s SP domain. As discussed later, this feature has direct functional implications, notably on the accessibility of the C1s CCP1 module, which is required for C4 cleavage by C1s [27]. C1 activation: a random, mechanical process? In addition to revealing the details of their structure, X-ray analysis of the C1r and C1s catalytic regions also provides precise insights into the molecular mechanisms underlying C1 activation. From a functional standpoint, the most intriguing feature of the C1r catalytic region head-totail structure [23] is that the catalytic site of one monomer and the activation site of the other lie at opposite ends
Review
TRENDS in Immunology
of the dimer (Figure 1d). This configuration probably represents a resting state of the molecule, possibly designed to prevent spontaneous C1 activation. Indeed, it is not compatible with C1r self-activation, which requires cleavage of the susceptible Arg – Ile bond of each monomer by the catalytic Ser residue of its counterpart [28], implying a direct contact between the two SP domains through disruption of the head-to-tail structure seen in the crystal. This has led us to the proposal that the signal that triggers C1r activation in C1 is a mechanical stress transmitted through the C1q – C1r– C1s interface from a C1q stem to the C1r catalytic region when C1 binds to a target. This mechanism is consistent with the presence of a semi-flexible hinge in C1q at the point where the collagen stems join to form the stalk [29]. Thus, it is conceivable that multivalent binding of C1q through its globular heads to an irregular cluster of distant binding sites will make some of the stems swing away from the center of the C1 complex, thereby generating the activating signal. The interaction of the two SP domains is obviously rendered possible by the presence of the large opening in the center of the C1r catalytic region (Figure 1d). Furthermore, it is likely to be facilitated by inter-modular flexibility in C1r, particularly at the CUB2 – CCP1 and CCP1 – CCP2 junctions, as documented in the latter case for other CCP module pairs [30]. The activation process possibly also involves restrained flexibility at the CCP2 – SP interface, as revealed by a comparison of the three available X-ray structures containing this region [23,24]. Further insights into the C1r activation process arise from the crystal contacts observed in the zymogen and active C1r CCP2– SP structures [24]. Thus, in the wildtype active species, the SP domains are packed in such a way that they form an enzyme-product-like complex, probably similar to the one occurring on activation of one C1r molecule by its counterpart (Figure 2b). Several transient states, such as the one observed in the proenzyme CCP2 – SP crystal where the SP domains interact in a nearly symmetrical manner [24], are probably necessary to reach this final activation state. This scenario
(a)
might apply to either C1r molecule, which might activate in any order because the two opposite C1q stems attached to the C1r– C1s interaction regions are expected to move in a random fashion, depending on whether their respective globular head is engaged in ligand recognition. C1s activation requires appropriate positioning of its Arg–Ile cleavage site within the active site cleft of the corresponding C1r molecule (Figure 2c). This might involve a movement of the C1s SP domain towards the interior of the C1 complex, or a movement of the C1r SP domain in the opposite direction, or both. Again, these shifts might be facilitated by the presence of flexible hinges at the CUB2– CCP1 and CCP1–CCP2 junctions in both C1r and C1s. It is tempting to postulate that C1s uses its CCP1–CCP2 interface to swing around a C1q stem. Once activated, the C1s SP domain is expected to move towards its protein substrates C4 and C2. In the case of C4, this would also enable proper positioning of the C1s CCP1 module relative to the active site, a probable prerequisite for efficient recognition and cleavage of this substrate [27]. As for C1r, it is likely that, once fully activated, its catalytic region folds back into a head-to-tail dimeric structure. This hypothesis is ˚ resolution of the strongly supported by the structure at 4 A active C1r catalytic region, which reveals an overall configuration strikingly similar to that of the resting proenzyme form [23]. This also appears to be consistent with our knowledge of the reaction of C1 inhibitor with activated C1, resulting in disassembly of the tetramer into two C1 inhibitor–C1s–C1r–C1 inhibitor complexes [23,31]. This scenario is largely hypothetical and obviously requires further experimental support, particularly with respect to the mechanisms that initiate activation on the target surface. Nevertheless, it seems clear that C1 activation is a mechanical process relying on displacement of the C1q stems on binding of C1 to an activating target. Owing to the random character of these movements, the C1 complex should not be viewed as a structurally homogenous molecule but as a mixture of conformational states differing in the activation status of C1r and/or C1s.
(b)
C1sA SP
371
Vol.25 No.7 July 2004
(c)
C1rA SP C1rB SP
C1sB SP
TRENDS in Immunology
Figure 2. Tentative scheme for C1 activation, as seen from bottom views of the complex. (a) C1s –C1r– C1r–C1s in the proenzyme, 8-shaped resting state. The sliding movement required to disrupt the resting head-to-tail C1r– C1r dimer is indicated by arrows. (b) S-shaped conformation featuring activation of a C1r SP domain by its counterpart. (c) Transient conformation featuring activation of C1sA by C1rB. C1q is shown in the ‘closed’ conformation in (a) and in the ‘open’ conformation in (b) and (c). Abbreviation: SP, serine protease. www.sciencedirect.com
Review
372
TRENDS in Immunology
Structural basis for the versatile recognition function of C1q One of the most striking properties of C1q lies in its ability to recognize an amazing variety of targets, including immunoglobulins, CRP, b-amyloid fibrils, the prion protein, lipid A, DNA, various microorganisms and apoptotic cells. There is no obvious structural feature shared by these diverse ligands, however, the fact that many polyionic structures are recognized by C1q [1], suggests that it might function as a charge pattern recognition molecule. The X-ray structure of the heterotrimeric C1q globular head [17] yields insights into its recognition properties. The three modules exhibit striking differences in their surface distribution of charged and hydrophobic residues. Therefore, the pseudo-threefold symmetry seen at the framework level disappears when surface patterns are considered (Figure 3a). Thus, individually, each of the three subunits can be expected to fulfill specific interaction properties. However, as illustrated in Figure 3a, the
(a)
(b)
(iii)
(ii)
(i)
(c)
Vol.25 No.7 July 2004
globular head forms a compact structure through the tight interaction of the three modules. Obviously, this type of structure also enables ligand recognition through residues contributed by two or even three subunits, thus providing a basis for the versatility of the recognition properties of C1q. One of the many possible recognition modes of C1q is illustrated by CRP, an acute phase protein that binds to phosphocholine groups of membrane phospholipids and is, in turn, recognized by C1q [2]. In the interaction model proposed on the basis of available structural and mutagenesis data [32– 34], the top of the C1q head, which is predominantly basic, is accommodated by the negatively charged central pore of C-reactive protein (Figure 3b), an interaction that involves the three C1q subunits. By contrast, the model proposed for interaction with IgG b12, a human IgG1 molecule, is an example of a recognition mediated by a single subunit of the C1q head (Figure 3c,d). Based on the X-ray structure of IgG b12 [35] and available information about the C1q– IgG interaction [36 –39], the model features binding of the equatorial region of subunit B of C1q at the Fab– Fc interface [17]. This raises the interesting possibility that C1q binds not only to the Fc domain but also to additional sites in the Fab arm, underscoring the possible role of the hinge region of IgG as a limiting factor of C1q recognition. Given the putative location of the C1q B module on the external part of the protein (Figure 1), the above model appears particularly well adapted to the recognition of IgG molecules within an immune complex network. Alternative recognition modes involving the internal A and/or C modules of C1q can be postulated, particularly in the case of the oligomeric IgM structure, in agreement with recent findings [40].
(d)
Fc
Fc
C1q
Fab
C1q Fab
TRENDS in Immunology
Figure 3. The versatile recognition function of C1q. (a) Top view of the C1q globular head, showing the different surface patterns of the three subunits, as illustrated by the distribution of basic (blue), acidic (red) and hydrophobic (yellow) residues. (i), (ii) and (iii) indicate A, B and C subunits. The lines indicate the approximate subunit boundaries. (b) Model of the C1q–CRP interaction. Two protomers of the pentameric CRP structure have been removed for clarity. (c,d) Model of the interaction between C1q and IgG1 b12. The IgG Fc domain and Fab arms are indicated. The C1q subunits are colored blue (A), green (b) and red (C). CRP (b) and IgG1 (c,d) are colored yellow, with functional residues identified by mutagenesis highlighted in different colors. Abbreviations: CRP, C-reactive protein. Adapted with permission from Ref. [17]. www.sciencedirect.com
Conclusions and perspectives The combined use of a dissection strategy and of structural biology techniques has led to a refined, three-dimensional model of the C1 complex of complement, providing for the first time a sound structural basis for the mechanisms underlying its activation process and its recognition properties. Understanding the details of the assembly and function of C1 at the atomic level will now require resolution of the remaining unknown three-dimensional structures within C1, and the use of site-directed mutagenesis to map the interaction sites within the complex and to further decipher the mechanisms involved in C1 activation. Detailed knowledge of the mechanisms underlying the extremely versatile recognition properties of C1q will also involve resolution of the structure of its globular head in complex with some of its ligands. Undoubtedly, a further key issue will be to investigate the molecular dynamic properties of C1, particularly the many areas of flexibility of its individual subunits, which enable the whole complex to mediate its finely tuned function. Acknowledgements We are indebted to Isabelle Bally, Beate Bersch, Monika Budayova-Spano, Claudine Darnault, Jordi Juanhuix, Monique Lacroix and David Pignol, who contributed to many of the studies reported in this review.
Review
TRENDS in Immunology
Vol.25 No.7 July 2004
References 1 Cooper, N.R. (1985) The classical pathway of complement: activation and regulation of the first complement component. Adv. Immunol. 37, 151 – 216 2 Szalai, A.J. et al. (1999) C-reactive protein: structural biology and host defense function. Clin. Chem. Lab. Med. 37, 265– 270 3 Rogers, J. et al. (1992) Complement activation by b-amyloid in Alzheimer disease. Proc. Natl. Acad. Sci. U. S. A. 89, 10016 – 10020 4 Tacnet-Delorme, P. et al. (2001) b-amyloid fibrils activate the C1 complex of complement under physiological conditions. Evidence for a binding site for Ab on the C1q globular regions. J. Immunol. 167, 6374 – 6381 5 Mabbott, N.A. et al. (2001) Temporary depletion of complement component C3 or genetic deficiency of C1q significantly delays onset of scrapie. Nat. Med. 7, 485 – 487 6 Klein, M.A. et al. (2001) Complement facilitates early prion pathogenesis. Nat. Med. 7, 488 – 492 7 Navratil, J.S. et al. (2001) The globular heads of C1q specifically recognize surface blebs of apoptotic vascular endothelial cells. J. Immunol. 166, 3231– 3239 8 Botto, M. and Walport, M.J. (2002) C1q, autoimmunity and apoptosis. Immunobiology 205, 395 – 406 9 Arlaud, G.J. et al. (2001) Structural biology of C1: dissection of a complex molecular machinery. Immunol. Rev. 180, 136 – 145 10 Kishore, U. and Reid, K.B.M. (2000) C1q: structure, function, and receptors. Immunopharmacology 49, 159 – 170 11 Arlaud, G.J. et al. (2002) Structural biology of the C1 complex of complement unveils the mechanisms of its activation and proteolytic activity. Mol. Immunol. 39, 383 – 394 12 Strang, C.J. et al. (1982) Ultrastructure of the first component of human complement: electron microscopy of the crosslinked complex. Proc. Natl. Acad. Sci. U. S. A. 79, 586 – 590 13 Perkins, S.J. et al. (1984) Neutron scattering studies of subcomponent C1q of first component C1 of human complement and its association with C1r2C1s2 within C1. J. Mol. Biol. 179, 547 – 557 14 Weiss, V. et al. (1986) Functional model of subcomponent C1 of complement. J. Mol. Biol. 189, 573 – 581 15 Arlaud, G.J. et al. (1987) A functional model of the human C1 complex. Immunol. Today 8, 106 – 111 16 Schumaker, V.N. et al. (1987) Activation of the first component of complement. Annu. Rev. Immunol. 5, 21 – 42 17 Gaboriaud, C. et al. (2003) The crystal structure of the globular head of complement protein C1q provides a basis for its versatile recognition properties. J. Biol. Chem. 278, 46974 – 46982 18 Gregory, L.A. et al. (2003) X-ray structure of the Ca2þ-binding interaction domain of C1s. Insights into the assembly of the C1 complex of complement. J. Biol. Chem. 278, 32157 – 32164 19 Reid, K.B.M. (1979) Complete amino acid sequences of the three collagen-like regions present in subcomponent C1q of the first component of human complement. Biochem. J. 179, 367 – 371 20 Bork, P. and Bairoch, A. (1995) Extracellular modules. Trends Biochem. Sci. 20 (Suppl), C03 21 Bersch, B. et al. (1998) Solution structure of the epidermal growth factor (EGF)-like module of human complement protease C1r, an atypical member of the EGF family. Biochemistry 37, 1204– 1214 22 Gaboriaud, C. et al. (2000) Crystal structure of the catalytic domain of
www.sciencedirect.com
23
24
25
26 27
28
29
30
31
32 33
34 35
36
37 38
39
40
373
human complement C1s: a serine protease with a handle. EMBO J. 19, 1755– 1765 Budayova-Spano, M. et al. (2002) The crystal structure of the zymogen catalytic domain of complement protease C1r reveals that a disruptive mechanical stress is required to trigger activation of the C1 complex. EMBO J. 21, 231 – 239 Budayova-Spano, M. et al. (2002) Monomeric structures of the zymogen and active catalytic domains of complement protease C1r: further insights into the C1 activation mechanism. Structure 10, 1509– 1519 Feinberg, H. et al. (2003) Crystal structure of the CUB1 – EGF – CUB2 region of mannose-binding protein associated serine protease-2. EMBO J. 22, 2348 – 2359 Campbell, I.D. and Bork, P. (1993) Epidermal growth factor-like modules. Curr. Opin. Struct. Biol. 3, 385 – 392 Rossi, V. et al. (1998) Baculovirus-mediated expression of truncated modular fragments from the catalytic region of human complement serine protease C1s. Evidence for the involvement of both complement control protein modules in the recognition of the C4 protein substrate. J. Biol. Chem. 273, 1232 – 1239 Lacroix, M. et al. (2001) Assembly and enzymatic properties of the catalytic domain of human complement protease C1r. J. Biol. Chem. 276, 36233 – 36240 Poon, P.H. et al. (1983) Conformation and restricted segmental flexibility of C1, the first component of human complement. J. Mol. Biol. 168, 563 – 577 Kirkitadze, M.D. and Barlow, P.N. (2001) Structure and flexibility of the multiple domain proteins that regulate complement activation. Immunol. Rev. 180, 146– 161 Sim, R.B. et al. (1979) C1 inhibitor-dependent dissociation of human complement component C1 bound to immune complexes. Biochem. J. 179, 449 – 457 Shrive, A.K. et al. (1996) Three-dimensional structure of human C-reactive protein. Nat. Struct. Biol. 3, 346 – 354 Thompson, D. et al. (1999) The physiological structure of human C-reactive protein and its complex with phosphocholine. Struct. Fold. Des. 7, 169 – 177 Agrawal, A. et al. (2001) Topology and structure of the C1q-binding site on C-reactive protein. J. Immunol. 166, 3998 – 4004 Saphire, E.O. et al. (2001) Crystal structure of a neutralizing human IgG against HIV-1: a template for vaccine design. Science 293, 1155– 1159 Idusogie, E.E. et al. (2000) Mapping of the C1q binding site on rituxan, a chimeric antibody with a human IgG1 Fc. J. Immunol. 164, 4178– 4184 Idusogie, E.E. et al. (2001) Engineered antibodies with uncreased activity to recruit complement. J. Immunol. 166, 2571 – 2575 Hezareh, M. et al. (2001) Effector function activities of a panel of mutants of a broadly neutralizing antibody against human immunodeficiency virus type 1. J. Virol. 75, 12161 – 12168 Marques, G. et al. (1993) Arginine residues of the globular regions of human C1q involved in the interaction with immunoglobulin G. J. Biol. Chem. 268, 10393 – 10402 Kishore, U. et al. (2003) Modular organization of the carboxylterminal, globular head region of human C1q A, B, and C chains. J. Immunol. 171, 812– 820