Complement:Classical Secondary article Pathway Artice Contents Gerard Arlaud,CEA/Grenoble,Grenoble,France the Path Maurice G Colomb,CEA/Grenoble,Grenoble,France The classical pathway of complement isa major re mals and other ar montemninitem olecular roteases-C1 and the MBL/MASP c trigg oennaiepai thogenic Introduction remembered that its initial component Cl is also activated The classical pathw or system of directly by a remarkably diverse range of substances other than antibody(Table 1:Cooper,1985).These include some mate monmody-ccpenden Gr teases:CI and the mannan-binding lectin/MBL-associated Gram-posmven am-negative bactera.cemv serine protease(MBL/MASP)complex.These convert an cellular and subcellular structures from damag ed cells as well as several proteins,carbohydrates,lipids,polyanions, and other compounds such as sodi vation Both murate crystals.The tes the lead to the fo C-reactive protein,w pno further complex protease(C3 convertase)and the genera erse biological activates the classical pathway and thereby trigger elimination of the pathogens by phagocytic mechanisms In contrast,Cl act vation by peptide may ntribute the mer disease pathogenesis.There are no obvious structural Activators of the Pathway features shared by the diverse nonimmune Cl activators, ch activate CI more efficiently than immu Classical pathway activators ac rted to cells.efficiently activate the classical pathway after their regions(Figure 1),or both. ng Lectin pathway activators IgG subclasses is IgG3.followed by IgGl and IgG2. In addition to the well-characterized Cl-mediated initia whereas IgG4 exhibits very little reactivity.The Cl- tion of the classical pathway,a second activation route,the binding site in Ig has been oc cated w ithin the an ectin path way,w discovered in the early 1980 EceioneobleCand ctin by its Clg subunit and involves the heterotrimeric C- (MBL).a member of the collectin family (Turner,1996). terminal globular'heads'of the protein(Figure 1).Because Through its multiple C-terminal C-type lectin modules .the Cl binding afinity A) about 10000-fold less than do IG-containing number of bacteria (Salmonella montevideo. immune complexes. monocytogenes,Haemophilus influenzae and Neisseria ENCYCLOPEDIA OF LIFE SCIENCES/2001 Nature Publishing Group/www.els.net
Complement: Classical Pathway Ge´rard J Arlaud, CEA/Grenoble, Grenoble, France Maurice G Colomb, CEA/Grenoble, Grenoble, France The classical pathway of complement is a major recognition system in innate immunity that is found in mammals and other animal species. It triggers various mechanisms via two multimolecular proteases – C1 and the MBL/MASP complex – to eliminate pathogenic microorganisms and other antigens from blood and tissues. Introduction The classical pathway of complement is a major system of innate (nonantibody-dependent) immunity. It is triggered through activation of two different multimolecular proteases: C1 and the mannan-binding lectin/MBL-associated serine protease (MBL/MASP) complex. These convert an initial recognition signal into proteolytic activity, thereby initiating the ‘classical’ and ‘lectin’ routes of complement activation. Both routes then lead to the formation of a further complex protease (C3 convertase) and the generation of protein fragments triggering diverse biological activities, such as opsonization, endocytosis and inflammation. Activators of the Pathway Classical pathway activators A wide variety of pathogenic microorganisms, including a number of bacteria and viruses, as well as many infected cells, efficiently activate the classical pathway after their recognition by antibodies. Activation is triggered upon interaction of the serum C1 complex with antigen– antibody complexes or immune aggregates containing immunoglobulin G (IgG) or IgM. The most reactive of the IgG subclasses is IgG3, followed by IgG1 and IgG2, whereas IgG4 exhibits very little reactivity. The C1- binding site in Ig has been located within the Cg2 and Cm3 domains of the Fc regions of IgG and IgM, respectively. C1 binding to immunoglobulins is mediated by its C1q subunit and involves the heterotrimeric Cterminal globular ‘heads’ of the protein (Figure 1). Because of the hexameric structure of C1q, the C1 binding affinity to Ig-containing activators increases markedly with their state of aggregation: soluble monomeric IgG binds C1 about 10 000-fold less efficiently than do IgG-containing immune complexes. Although the classical pathway is often described as an antibody-dependent defence mechanism, it should be remembered that its initial component C1 is also activated directly by a remarkably diverse range of substances other than antibody (Table 1; Cooper, 1985). These include some Gram-positive and Gram-negative bacteria, certain viruses (including human immunodeficiency virus, HIV), cellular and subcellular structures from damaged cells, as well as several proteins, carbohydrates, lipids, polyanions, and other compounds such as sodium urate crystals. The acute-phase C-reactive protein, which recognizes phosphocholine groups on pneumococci and other pathogens, activates the classical pathway and thereby triggers elimination of the pathogens by phagocytic mechanisms. In contrast, C1 activation by fibrils of the b-amyloid peptide may contribute, through the development of uncontrolled inflammation, to the progression of Alzheimer disease pathogenesis. There are no obvious structural features shared by the diverse nonimmune C1 activators, some of which activate C1 more efficiently than immunoglobulins. In most cases, C1 binding to these activators occurs through the C1q subunit, and has been reported to involve either its globular ‘heads’, or its collagen-like regions (Figure 1), or both. Lectin pathway activators In addition to the well-characterized C1-mediated initiation of the classical pathway, a second activation route, the ‘lectin pathway’, was discovered in the early 1980s. It is triggered by a C1-like complex protease in which the recognition function is mediated by mannan-binding lectin (MBL), a member of the collectin family (Turner, 1996). Through its multiple C-terminal C-type lectin modules (Figure1),MBL recognizes arrays of oligosaccharides at the surface of a wide spectrum of microorganisms, including yeasts (Candida albicans), viruses (HIV, influenza A), and a number of bacteria (Salmonella montevideo, Listeria monocytogenes, Haemophilus influenzae and Neisseria meningitidis). There is increasing evidence that the lectin pathway of complement activation plays an important role Article Contents Secondary article . Introduction . Activators of the Pathway . Components . Functioning of the Pathway . Regulation of the Pathway . Protein Biosynthesis and Deficiencies . Phylogeny ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net 1
CUBEC CUBC@C①Ser Pr☐wMp.2 VA Ser Pr☐ C2(Factor B) 0 Macroglobulin family Macroglobulin family C345Cc3,c4(C5) Ser Pr -binding lectin teins C6/C7 (FIMAC)m LA low grey.and coiled-coil struc ner,whereas the pro Table 1 Activators of the classical pathway of complement in immune defence,particularly during the phase of primary contact with microorganisms. Immune activators Antigen-antibody complexes containing IgM or IgG No strains (e coli salmonella klebsiellay Mycoplasma:Mycoplasma pneumoniae Components Various viruses (Sindbis.N wcastle disease.Epstein-Barr rus)and A characteristic feature shared by most complement i.Trypanosoma bruce protein motifsor modules (Figure Table2). Proteins:C-reactive protein,myelin,B-amyloid peptide. These belong to different types which may be recognized at serum amyloid P component the sequence level by specific consensus sequences and have Oasatart.andioolcthandc the ability to mediate pro S. tein n.DNA) The recognition proteins Clq and MBL are both Modified from Cooper(1985). oligomers of trim regions prolonged by s but distinet A.B andCchains.Because ofrepeating Gly-Xaa-Yaa tripletsin ENCYCLOPEDIA OF LIFE SCIENCES/ ublishing Group/www.els.net
in immune defence, particularly during the phase of primary contact with microorganisms. Components A characteristic feature shared by most complement components is the occurrence in their structure of contiguous protein motifs or modules (Figure 1; Table 2). These belong to different types which may be recognized at the sequence level by specific consensus sequences and have probably evolved from a common ancestor by gene duplication. Protein modules fulfil specific functions, often the ability to mediate protein–protein interactions. The recognition proteins C1q and MBL are both oligomers of trimeric subunits, comprising collagen-like regions prolonged by C-terminal globular domains, with overall structures resembling a bunch of tulips (Figure 1). C1q is a hexamer of three homologous but distinct A, B and Cchains. Because of repeating Gly-Xaa-Yaa triplets in F M C1r, C1s MASP-1, MASP-2 CP CP Ser Pr CP CP C2 (Factor B) CP Ser Pr CUB EG CUB VA C345C C3, C4 (C5) SR Ser Pr Factor I Macroglobulin family AT Macroglobulin family L A L A CLECT CLECT CLECT CLECT CLECT CLECT C1q C1q C1q C1q C1q C1q C1q Mannan-binding lectin A B C S — C = O Figure 1 Modular structures of the complement classical pathway proteins. The nomenclature and symbols used for protein modules are those defined by Bork and Bairoch (1995). AT, anaphylatoxin; CP, complement control protein (CCP) module; CUB, module found in complement C1r/C1s, Uegf, and bone morphogenetic protein; C345C, complement proteins C3/C4/C5 C-terminal module; EG, epidermal growth factor (EGF)-like module; FM, factor I/ membrane attack complex proteins C6/C7 (FIMAC) module; LA, low-density lipoprotein receptor class A (LDLRA) module; SR, scavenger receptor cysteine-rich (SRCR) module; Ser Pr, serine protease domain. VA, von Willebrand factor type A (VWFA) module. Heterotrimeric associations of ‘C1q’ modules (A, B, C), and homotrimeric associations of C-type lectin (CLECT) modules are found at the C-terminal end of the collagen ‘arms’ of C1q and mannan-binding lectin, respectively. Unlabelled portions of the molecules represent connecting segments or sequence areas with no known homology to other proteins. Areas of collagen-like structure are grey, and coiled-coil structures are shown in black. C1q only occurs as a hexamer, whereas mannanbinding lectin is present in serum under multiple oligomeric forms. Arrows indicate peptide bonds cleaved on activation of proteolytic enzymes. In the case of C3 and C4, the only cleavage shown is that mediated by C3 convertase and C1s or MASP-2, respectively. The location of the internal thioester group of C3 and C4 is indicated. The relative sizes of the proteins are approximate. Table 1 Activators of the classical pathway of complement Immune activators Antigen–antibody complexes containing IgM or IgG Nonimmune activators Various bacterial strains (E. coli, Salmonella, Klebsiella) Mycoplasma: Mycoplasma pneumoniae Various viruses (Sindbis, Newcastle disease, Epstein–Barr virus) and retroviruses (HIV) Parasites: Schistosoma mansoni, Trypanosoma brucei Cellular and subcellular membranes Proteins: C-reactive protein, myelin, b-amyloid peptide, serum amyloid P component Oligosaccharides and polysaccharides Lipids (lipid A, cardiolipin) Polyanions (heparin, DNA) Modified from Cooper (1985). Complement: Classical Pathway 2 ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net
Complement:Classical Pathway Table 2 Components of the classical and lectin activation pathways of human complement Carbohydrates Protein Chains (content,w/w:type)(mg L) Biological function Cla 459300 A:275508.3% 80 Recognition unit of the Cl complex B:25 200 glucosylgalactosyl C:23800 and N-linked MBL 4570000 2540055 1.5 Recognition unit of the MBL/MASP CIr 1730005 34 Activation unit of the CI complex 79800 31 MASP-I 86000 Activation unit of the MBL/MASP complex B28150 6 MASP-2 74150 47650 none B26500 ND Active component of the MBL/MASP C4 205000 a970007.00% CI substrate.yields active fragments C4a B:75 000 N-linked and C4h t33000 102000 C2a:7000015.9% 20 CI substrate;yields active fragments C2a C2b:30 000 N-linked and C2b C3 185000 1100005.9% 1300 Substrate of the C3-convertase:yields active B:75 000 N-linked fragments C3a and C3b a six collas triple all belong to the chy as a 'stalk'and then.due to esidue interruptions in the Gly-Xaa-Yaa motif,diverge to form trypsin-like specificity.Howeve the serine protease imately merges r,Cls and MA lack the 'histidine pop',a d own m MAD differs from Cla in that it occurs in serum as a mixture of oligomers ranging from dimers to hexamers,made up of a serine protease domain and the preceding CCP modules, gle type ofcha A further difference is that theco and th latter have waeepiepatemh d fr drate-recogn CUB-E calcium-dependent (C-type)lectin modules held together high-affinity Cabinding site.These regions mediate the by a coiled-coil of helices called the'neck'region. Ca dependent CIr-Cls interactions involved in the and mbly of the Cls CIr-CIr Cls catalytic subunit of C a3945 part in Cl assem exhibit the same type of modular organizati on (Figure 1) factor B.its functional homologue in the alterative Each comprises.from the N-terminus.two CUB modules comple nent pathway,consisting of,from the N-terminus les, single von Willebrand factor type a pair me protease (F Lys bond in the N-terminal region of the VWFA module ation.through cleavage of a which splits the molecule into the N-terminal fragment chains Nand'(the semne proteae a.The serine conserve ENCYCLOPEDIA OF LIFE SCIENCES/2001 Nature Publishing Group/www.els.net
their N-terminal region, these form six collagen triple helices which associate as a ‘stalk’ and then, due to interruptions in the Gly-Xaa-Yaa motif, diverge to form six ‘arms’. Each arm ultimately merges into a C-terminal globular ‘head’ region formed from the heterotrimeric association of protein modules of the C1q type. MBL differs from C1q in that it occurs in serum as a mixture of oligomers ranging from dimers to hexamers, made up of a single type of chain. A further difference is that the collagen triple helices of MBL merge into carbohydrate-recognition domains formed from the homotrimeric association of calcium-dependent (C-type) lectin modules held together by a coiled-coil of a helices called the ‘neck’ region. The proteases associated with C1q and MBL – C1r and C1s in the classical pathway and MASP-1 and MASP-2 in the lectin pathway – share 39–45% sequence identity and exhibit the same type of modular organization (Figure 1). Each comprises, from the N-terminus, two CUB modules surrounding a single epidermal growth factor (EGF)-like module, a pair of contiguous complement control protein (CCP) modules, and a serine protease domain. All four proteases are single-chain proteins in their proenzyme form and are split upon activation, through cleavage of a single Arg–Ile bond, into disulfide-linked polypeptide chains A and B (the serine protease domain). The latter all belong to the chymotrypsin-like family and have an Asp residue in their substrate-binding pocket, indicative of trypsin-like specificity. However, the serine protease domains of C1r, C1s and MASP-2 lack the ‘histidine loop’, a disulfide bridge present in all known mammalian serine proteases including MASP-1. The regions of C1r and C1s responsible for catalytic activity comprise both the serine protease domain and the preceding CCP modules, and the latter have been shown to participate in the recognition by C1s of its substrate C4. The N-terminal CUB-EGF module pairs of C1r and C1s each contain one high-affinity Ca2+-binding site. These regions mediate the Ca2+-dependent C1r–C1s interactions involved in the assembly of the C1s–C1r–C1r–C1s catalytic subunit of C1 and also take part in C1 assembly. C2 exhibits the same type of structural organization as factor B, its functional homologue in the alternative complement pathway, consisting of, from the N-terminus, three CCP modules, a single von Willebrand factor type A (VWFA) module, and a serine protease domain (Figure 1). Activation by C1s occurs through cleavage of a single Arg– Lys bond in the N-terminal region of the VWFA module, which splits the molecule into the N-terminal fragment C2b and the C-terminal catalytic fragment C2a. The serine protease domain of C2 lacks the highly conserved Table 2 Components of the classical and lectin activation pathways of human complement a Molecular weight of the hexameric form of MBL b Molecular weight of the C1r–C1r dimer c Not determined Protein Molecular mass (Da) Carbohydrates (content, w/w; type) Serum concentration (mg L–1 Molecule Chains ) Biological function C1q 459 300 A: 27 550 B: 25 200 C: 23 800 8.3% glucosylgalactosyl and N-linked 80 Recognition unit of the C1 complex MBL 457 000a 25 400 5.5% O-linked 1.5 Recognition unit of the MBL/MASP complex C1r 173 000b A: 55 300 B: 31 200 9.5% N-linked 34 Activation unit of the C1 complex C1s 79 800 A: 52 150 B: 27 650 6.1% N-linked 31 Active component of C1; cleaves C4 and C2 MASP-1 86 000 A: 57 850 B: 28 150 10.3% N-linked 6 Activation unit of the MBL/MASP complex MASP-2 74 150 A: 47 650 B: 26 500 none NDc Active component of the MBL/MASP complex C4 205 000 α: 97 000 β: 75 000 γ: 33 000 7.0% N-linked 600 C1 substrate; yields active fragments C4a and C4b C2 102 000 C2a: 70 000 C2b: 30 000 15.9% N-linked 20 C1 substrate; yields active fragments C2a and C2b C3 185 000 α: 110 000 β: 75 000 5.9% N-linked 1300 Substrate of the C3-convertase; yields active fragments C3a and C3b Complement: Classical Pathway ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net 3
hydrophobic N-terminal sequence generated during acti- vation of other zymogens including CIr.Cls and the of the active con- th →C4a substituted for the Asp residue normally present in the C1INH C4b◆Factor I binding pocket. m0d1 of C contains a (C2) hialgm the (metal which likely media ◆C2b hindingofa mo c46c2 C3an ay C3a 1 Th maior part of the C3 and C4 precursor molecules is accounted for by polype tide sequences showing homology with acroglobulin Lytic pathway betwe thes three proteins Factor I hich fo cide ch Figure 2 Classical and lectin pathways of c mplement Activation of the cysteine and a glutamine residue in the sequence f a m ctin vay is ini (Figure 1).C3 and C4 also have in common SA(AT)m arboh drat at the surface of a m sin ment of C3 and CA c)m red a ents C4b and C xhibiti g a rea ve thi action or (C1 INH) of the sen e in the towards MASP-1 and MASP-2 is no bridges.Activation of C3 by C3 convertase.and of C4 by C1 involves cleavage of a single arginyl bond at the C terminus of the A odule,thereby le ase .The isolat -CIs complex is an ng fragments ed Cis-CIr-CIr ar y of domain with width of um and a contour length of 51-59nm.The CIr-CIr dime Functioning of the Pathway rms the core of the tetramer,with its catalytic regions in the of a Cls molecule In this extended c h ion catalytic regions of Cls are located at both ends of the and the MBL/MASH x enzymes tetramer. us su app studies have indicated that the ability to cleave the same protein substrates C4 and C2. conditions.it was shown more recently that the dissocia thereby allowing formation of a further complex protease tion constant of human Cl at room ter mperature is in the 6a2 ing fragments endowed with picomolar range,indicating that virtu lly all of the Clq serum is as h C1 and the MBL/MASP complex ed and de of the CIr-CIssubunit.Several linesofevidenceindicatethat the -dependent complex of molecular ma ass about containing three sites of Clq respon ated wit for the interaction with CIs-CIr the co th (the rat of the ke region c subunits.Clq nit)and the Co inked Cl complex reveals that the tetra mer forms a dependent tetramer Cis-CIr-CIr-CIs (the catalytic compact mass centred on the Clg 'arms'.between the 4 ENCYCLOPEDIA OF LIFE SCIENCES/2001 Na
hydrophobic N-terminal sequence generated during activation of other zymogens including C1r, C1s and the MASPs, indicating that assumption of the active conformation in fragment C2a is achieved through a particular mechanism. Another unique feature is that, despite the known trypsin-like specificity of C2a, a Ser residue is substituted for the Asp residue normally present in the binding pocket. The VWFA module of C2 contains a free thiol group and the MIDAS (metal ion-dependent adhesion site) motif described in integrins, which likely mediates binding of a Mg2+ ion. Both sites appear to be involved in the assembly of the C3 convertase C4bC2a. C3 and C4, as well as component C5 of the lytic pathway, are synthesized as single-chain precursors with similar overall structures (Figure 1). The major part of the C3 and C4 precursor molecules is accounted for by polypeptide sequences showing homology with a2-macroglobulin. Sequence homology between these three proteins is particularly high in the region containing an internal thioester bond, which forms between the side chains of a cysteine and a glutamine residue in the sequence ...CysGly-Glu-Gln... (Figure 1). C3 and C4 also have in common a short anaphylatoxin (AT) motif in their central part and a complement C3/4/5 C-terminal (C345C) module. Proteolytic processing of C3 and C4 involves intracellular excision of a basic tetrapeptide at the N-terminal end of the AT module, and a further, two-step removal of a 26- residue sequence in the case of C4. The circulating mature proteins comprise two chains (a, b) in the case of C3 and three chains (a, b, g) in the case of C4, all linked by disulfide bridges. Activation of C3 by C3 convertase, and of C4 by C1 involves cleavage of a single arginyl bond at the Cterminus of the AT module, thereby leading to the release of the corresponding fragments C3a and C4a (Figure 2). Functioning of the Pathway Triggering of the classical pathway may occur through activation of two different multimolecular proteases, C1 and the MBL/MASP complex. These complex enzymes comprise distinct but homologous subunits, and appear to share similar structural and functional features, including the ability to cleave the same protein substrates C4 and C2, thereby allowing formation of a further complex protease (C3 convertase) and generating fragments endowed with biological activities (Figure 2). C1 and the MBL/MASP complex C1 is a Ca2+-dependent complex of molecular mass about 790 000 Da containing three subcomponents (C1q, C1r, C1s) in a 1:2:2 molar ratio. It is assembled from two subunits, C1q (the recognition subunit) and the Ca2+- dependent tetramer C1s–C1r–C1r–C1s (the catalytic subunit). The isolated C1s–C1r–C1r–C1s complex is an elongated, reversed ‘S’-shaped molecule comprising a linear array of domains, with an average width of 3–4 nm and a contour length of 51–59 nm. The C1r–C1r dimer forms the core of the tetramer, with its catalytic regions in the centre, each of its interaction regions being connected in a Ca2+-dependent manner to the corresponding region of a C1s molecule. In this extended conformation, the catalytic regions of C1s are located at both ends of the tetramer. Although early studies have indicated that C1q and C1s–C1r–C1r–C1s interact weakly under physiological conditions, it was shown more recently that the dissociation constant of human C1 at room temperature is in the picomolar range, indicating that virtually all of the C1q, C1r and C1s in serum is associated in C1. The strength of the association increases markedly when the temperature is raised, and decreases upon activation of the C1s–C1r– C1r–C1s subunit. Several lines of evidence indicate that the sites of C1q responsible for the interaction with C1s–C1r– C1r–C1s are located within the collagen-like region of the protein, and electron microscopy of the chemically crosslinked C1 complex reveals that the tetramer forms a compact mass centred on the C1q ‘arms’, between the Alternative pathway Opsonization Lytic pathway Classical pathway Lectin pathway C4a 1 and 2 C2 C2b C4b C4 C1r2 C1s2 MA SP C1q MBL Factor I C4b C2a C3-convertase C3 C3b Factor I C3a C1 INH Figure 2 Classical and lectin pathways of complement. Activation of the classical pathway is triggered by direct or antibody-dependent recognition of a microorganism by C1q, whereas the lectin pathway is initiated by interaction of the mannan-binding lectin (MBL) with arrays of carbohydrates at the surface of a microorganism. Proteins expressing proteolytic activity are shown in red and proteolytic cleavages are indicated by red arrows. Fragments C4b and C3b exhibiting a reactive thioester group are green, whereas small fragments generating inflammatory reactions are yellow. C1 inhibitor (C1 INH), a member of the serine protease inhibitor (serpin) family, controls both C1 activation and C1 proteolytic activity. Its reactivity towards MASP-1 and MASP-2 is not fully established. Factor I cleaves C4b and C3b, preventing formation of the convertases and generating C3b fragments endowed with important biological activities. Complement: Classical Pathway 4 ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net
Complement:Classical Pathway activation n of Can h str The natu of the signl that triggers Cl activation remains,howeve a complex pro s an s tetramer may adopt a heir N-terminal CUB-EGF inter reions.which oth C n therefore represent key elements of the architecture of thereby gain access to C4 and C2.the substrates of active macromolec lar C1 everal of the low-resoionCmodpropos structure and of complement activation.Although MBL (Figure3).This conformation brings the catalytic regionsof CIr and CIs into the same vicinity and thereby allows the dicate that MBL/MASP ally equivalent to CIr (MASP-1)and CIs (MASP-2) CIr.Although the molecular mechanisms involved in CI However,further analyses of the stoichiometry of the o Cupports an overal propertie and cear picture of fuctio are C3 convertase Proteolytic complex.hei of C4 by CI or the MBL/MASP nd Th reactive with amino and hydroxyl groups,with differences due to C4 polymorph- ism.The two isoforms of human C4,C4A and C4B,mainl i00 e nature at pos pCAA Ca Ing or th CA sotype involves a single-step reaction between the thioester and a sotype rea amino-nucleophi ich the two-step midazole intermediate.Native 4p is thus able to form amide or ester bonds or reacts readily with water.The eactivity of activated C4isextremely ort- the hal The above characteristics are also valid for c3 but this Figure3 C1 complex of complement.CIr,and C1s are blue, re ghthe com protein only exhibits reactivity of the C4B type.Overall. ytic regions of Cr and C1s resp sible A14 indiscriminate.the occurrence of regulatory proteins on the c normal host cells prevents complement deposition and amplif ENCYCLOPEDIA OF LIFE SCIENCES/2001 Nature Publishing Group /www.els.net
‘stalk’ and the globular ‘heads’ (Strang et al., 1982). Solution scattering studies provide further support to the hypothesis that the elongated C1s–C1r–C1r–C1s tetramer folds into a more compact conformation upon interaction with C1q. The binding of the tetramer to C1q appears to be a complex process involving ionic bonds and multiple sites contributed by both C1r and C1s and mainly located in their N-terminal CUB-EGF interaction regions, which therefore represent key elements of the architecture of macromolecular C1. Several of the low-resolution C1 models proposed so far are based on the assumption that the C1s–C1r–C1r–C1s tetramer adopts a compact, ‘8’-shaped conformation in C1 (Figure 3). This conformation brings the catalytic regions of C1r and C1s into the same vicinity and thereby allows the two-step C1 activation to proceed, namely (1) autoactivation of the C1r–C1r dimer through mutual cleavage of each monomer by its neighbour, and (2) C1s activation by active C1r. Although the molecular mechanisms involved in C1 activation and control are not fully understood, current knowledge supports an overall scheme involving: (1) the ability of C1 to activate spontaneously, reflecting the intrinsic property of C1r to autoactivate; (2) the occurrence in serum of a permanent negative control of C1 activation exerted by C1 inhibitor; and (3) the release of this control upon interaction of C1 with strong activators. The nature of the signal that triggers C1 activation remains, however, an intriguing question. Once C1 activation has occurred, the C1s–C1r–C1r–C1s tetramer may adopt a more relaxed, open ‘S’-shaped conformation in which both C1s catalytic regions protrude outside the C1q molecule and thereby gain access to C4 and C2, the substrates of active C1. Little information is available on the structure and function of the MBL/MASP complex that triggers the lectin pathway of complement activation. Although MBL was initially found associated in human serum with a single MASP species (MASP-1), more recent studies indicate that the catalytic component of the MBL/MASP complex contains two distinct serine proteases, probably functionally equivalent to C1r (MASP-1) and C1s (MASP-2). However, further analyses of the stoichiometry of the MBL/MASP-1/MASP-2 complex and of the structure and enzymatic properties of MASP-1 and MASP-2 are required to get a clear picture of its assembly, function and control. C3 convertase The thioester bond and the covalent binding of C4b and C3b Proteolytic cleavage of C4 by C1 or the MBL/MASP complex, and of C3 by C3 convertase leads to concomitant exposure of their internal thioester bond. The acyl group of the thioester in C4 becomes highly reactive with amino and hydroxyl groups, with differences due to C4 polymorphism. The two isoforms of human C4, C4A and C4B, mainly differ by the nature of the amino acid at position 1106 (about 100 residues C-terminal of the thioester), which is an Asp in C4A and a His in C4B. Binding of the C4A isotype involves a single-step reaction between the thioester and amino-nucleophiles, whereas the C4B isotype reacts with hydroxyl-nucleophiles in a two-step mechanism in which the His contributes to an acylimidazole intermediate. Native C4b is thus able to form amide or ester bonds or reacts readily with water. The reactivity of activated C4 is extremely short-lived: the halflife of activated C4A is about 10 s and that of C4B less than 1 s (Dodds et al., 1996). The above characteristics are also valid for C3, but this protein only exhibits reactivity of the C4B type. Overall, this B type reactivity appears to be the most frequent, leading to formation of ester links between C3b or C4b and acceptor sites. Although binding of C3b or C4b is indiscriminate, the occurrence of regulatory proteins on normal host cells prevents complement deposition and amplification on their surfaces, providing self/nonself discrimination. C1q C1s C1r Figure 3 C1 complex of complement. C1q, C1r, and C1s are blue, red and green, respectively. Upon assembly of the complex, the elongated C1s–C1r–C1r–C1s tetramer is thought to bend around two opposite collagen ‘arms’ of C1q, resulting in a compact, distorted ‘8’-shaped conformation. The catalytic regions of C1r and C1s responsible for C1 activation and proteolytic activity towards C4 and C2 are located inside the cone defined by the C1q arms, whereas the interaction regions involved in both the assembly of the C1s–C1r–C1r–C1s tetramer and its interaction with C1q lie outside the collagen ‘arms’ of C1q. Modified from Arlaud et al. (1987). Complement: Classical Pathway ENCYCLOPEDIA OF LIFE SCIENCES / & 2001 Nature Publishing Group / www.els.net 5