Ficolins: innate immune recognition proteins for danger sensing

Las ficolinas son proteínas de defensa que forman oligómerosa partir de tallos homólogos al colágeno y dominios semejantesa fibrinógeno. Son capaces de sentir señales de peligro talescomo patrones moleculares asociados patógenos o a células apoptóticas.En el hombre, las ficolinas L y H se han caracterizado enel suero, mientras que la ficolina M es secretada por células monocíticas.Al igual que la lectina de unión a manano (“mannan-bindinglectin”, MBL), pueden asociarse a las serina-proteasas asociadasa MBL e iniciar la vía de activación de complemento de laslectinas, un importante sistema efector de la inmunidad innatahumoral. También pueden actuar como opsoninas, incrementandola eliminación de sus dianas por fagocitosis. Estudios estructuralesrecientes muestran que la ficolina L es una proteína dereconocimiento versátil, capaz de unir moléculas acetiladas y carbohidratosneutros por medio de sitios de unión diferentes, mientrasque la ficolina H posee un único sitio de unión con una especificidadmás restringida hacia los carbohidratos neutros. Los estudiosfilogenéticos revelan que las ficolinas han sido conservadasen el proceso evolutivo, apoyando la hipótesis de que el sistemade complemento primitivo era un sistema de opsonización basadoen lectinas, y ponen de relieve la importancia de las proteínasde reconocimiento de carbohidratos en la inmunidad innata

Ficolins are oligomeric defence proteins assembled from collagen-like stalks and fibrinogen-like domains that are able to sensedanger signals such as pathogen- or apoptotic cell-associated molecularpatterns. In humans, L- and H-ficolins have been characterizedin serum whereas M-ficolin is secreted by monocytic cells.Like mannan-binding lectin (MBL), they are able to associate withMBL-associated serine proteases and to trigger activation of thelectin pathway of complement, a major effector system of humoralinnate immunity. They can also act as opsonins to enhance clearanceof their targets by phagocytosis. Recent structural studieshave shown that L-ficolin is a versatile recognition protein ableto bind acetylated molecules and neutral carbohydrates throughdifferent binding sites, whereas H-ficolin has a single binding sitewith a more restricted specificity for neutral carbohydrates. Phylogeneticstudies reveal that ficolins have been conserved throughevolution, supporting the hypothesis that the primitive complementsystem was a lectin-based opsonic system, and emphasizingthe essential role of carbohydrate recognition proteins in innateimmunity

Structural biology of C1.

The classical complement pathway is a major element of innate immunity against infection, and is also involved in immune tolerance, graft rejection and various pathologies. This pathway is triggered by C1, a multimolecular protease formed from the association of a recognition protein, C1q, and a catalytic subunit, the calcium-dependent tetramer C1s-C1r-C1r-C1s, which comprises two copies of each of the modular proteases C1r and C1s. All activators of the pathway are recognized by the C1q moiety of C1, a process that generates a conformational signal that triggers self-activation of C1r, which in turn activates C1s, the enzyme that mediates specific cleavage of C4 and C2, the C1 substrates. Early work based on biochemical and electron microscopy studies has allowed characterization of the domain structure of the C1 subcomponents and led to a low-resolution model of the complex in which the elongated C1s-C1r-C1r-C1s tetramer folds into a compact, figure-of-8-shaped conformation upon interaction with C1q. The strategy used over the past decade was based on a dissection of the C1 proteins into modular segments to characterize their function and solve their three-dimensional structure by X-ray crystallography or NMR spectroscopy. This approach allows deep insights into the structure-function relationships of C1, particularly with respect to the assembly of the C1 complex and the mechanisms underlying its activation and proteolytic activity.

Evaluation and clinical interest of mannan binding lectin function in human plasma.

The mannan binding lectin (MBL) plays a major role in innate immunity through its ability to activate complement upon binding to carbohydrate arrays on the surface of various microorganisms. The question of a possible association of the MBL structural gene polymorphism and the oligomeric state of MBL was poorly documented. For these reasons, it appears difficult to evaluate MBL in blood patients on the only basis of protein contents, even in combination with MBL genotyping. This study reports a method to calculate a specific activity for circulating MBL, that relies on: (i) the availability of purified MBL; and (ii) a simplified MBL activity assay based on complement activation. The three-step MBL purification from human plasma reported here is characterized by a highly purified MBL, that occurs in two different oligomeric forms. The results on the specific activity of these forms show that the higher oligomeric forms of MBL have the ability to induce C4 cleavage more efficiently than the corresponding lower oligomers. The usefulness of this approach is illustrated by its potential interest in the biological exploration of certain pathology, for example in the follow-up of chronic hepatitis C. Further investigation is needed to establish whether MBL specific activity (MBLsa) is correlated to the polymorphic state of the molecule. The relative simplicity of the test described here allows better investigation on the relationship between MBL biological activity and its genotype.

Beta-amyloid fibrils activate the C1 complex of complement under physiological conditions: evidence for a binding site for A beta on the C1q globular regions.

Previous studies based on the use of serum as a source of C have shown that fibrils of beta-amyloid peptides that accumulate in the brain of patients with Alzheimer's disease have the ability to bind C1q and activate the classical C pathway. The objective of the present work was to test the ability of fibrils of peptide Abeta1-42 to trigger direct activation of the C1 complex and to carry out further investigations on the site(s) of C1q involved in the interaction with Abeta1-42. Using C1 reconstituted from purified C1q, C1r, and C1s, it was shown that Abeta1-42 fibrils trigger direct C1 activation both in the absence of C1 inhibitor and at C1 inhibitor:C1 ratios up to 8:0, i.e., under conditions consistent with the physiological context in serum. The truncated peptide Abeta12-42 and the double mutant (D7N, E11Q) of Abeta1-42 did not yield C1 activation, providing further evidence that the C1 binding site of beta-amyloid fibrils is located in the acidic N-terminal 1-11 region of the Abeta1-42 peptide. Binding studies performed using a solid phase assay provided strong evidence that C1q interacts with Abeta1-42 fibrils through its C-terminal globular regions. In contrast to previous studies based on a different experimental design, no significant involvement of the C1q collagen-like domain was detected. These findings were confirmed by additional experiments based on C1 activation and C4 consumption assays. These observations provide direct evidence of the ability of beta-amyloid fibrils to trigger activation of the classical C pathway and further support the hypothesis that C activation may be a component of the pathogenesis of Alzheimer's disease.

The role of the individual domains in the structure and function of the catalytic region of a modular serine protease, C1r.

The first enzymatic event in the classical pathway of complement activation is autoactivation of the C1r subcomponent of the C1 complex. Activated C1r then cleaves and activates zymogen C1s. C1r is a multidomain serine protease consisting of N-terminal alpha region interacting with other subcomponents and C-terminal gammaB region mediating proteolytic activity. The gammaB region consists of two complement control protein modules (CCP1, CCP2) and a serine protease domain (SP). To clarify the role of the individual domains in the structural and functional properties of the gammaB region we produced the CCP1-CCP2-SP (gammaB), the CCP2-SP, and the SP fragments in recombinant form in Escherichia coli. We successfully renatured the inclusion body proteins. After renaturation all three fragments were obtained in activated form and showed esterolytic activity on synthetic substrates similar to each other. To study the self-activation process in detail zymogen mutant forms of the three fragments were constructed and expressed. Our major statement is that the ability of autoactivation and C1s cleavage is an inherent property of the SP domain. We observed that the CCP2 module significantly increases proteolytic activity of the SP domain on natural substrate, C1s. Therefore, we propose that CCP2 module provides accessory binding sites. Differential scanning calorimetric measurements demonstrated that CCP2 domain greatly stabilizes the structure of SP domain. Deletion of CCP1 domain from the CCP1-CCP2-SP fragment results in the loss of the dimeric structure. Our experiments also provided evidence that dimerization of C1r is not a prerequisite for autoactivation.

Substrate specificities of recombinant mannan-binding lectin-associated serine proteases-1 and -2.

Mannan-binding lectin (MBL)-associated serine proteases-1 and 2 (MASP-1 and MASP-2) are homologous modular proteases that each interact with MBL, an oligomeric serum lectin involved in innate immunity. To precisely determine their substrate specificity, human MASP-1 and MASP-2, and fragments from their catalytic regions were expressed using a baculovirus/insect cells system. Recombinant MASP-2 displayed a rather wide, C1s-like esterolytic activity, and specifically cleaved complement proteins C2 and C4, with relative efficiencies 3- and 23-fold higher, respectively, than human C1s. MASP-2 also showed very weak C3 cleaving activity. Recombinant MASP-1 had a lower and more restricted esterolytic activity. It showed marginal activity toward C2 and C3, and no activity on C4. The enzymic activity of both MASP-1 and MASP-2 was specifically titrated by C1 inhibitor, and abolished at a 1:1 C1 inhibitor:protease ratio. Taken together with previous findings, these and other data strongly support the hypothesis that MASP-2 is the protease that, in association with MBL, triggers complement activation via the MBL pathway, through combined self-activation and proteolytic properties devoted to C1r and C1s in the C1 complex. In view of the very low activity of MASP-1 on C3 and C2, our data raise questions about the implication of this protease in complement activation.

Assembly and enzymatic properties of the catalytic domain of human complement protease C1r.

The catalytic properties of C1r, the protease that mediates activation of the C1 complex of complement, are mediated by its C-terminal region, comprising two complement control protein (CCP) modules followed by a serine protease (SP) domain. Baculovirus-mediated expression was used to produce fragments containing the SP domain and either 2 CCP modules (CCP1/2-SP) or only the second CCP module (CCP2-SP). In each case, the wild-type species and two mutants stabilized in the proenzyme form by mutations at the cleavage site (R446Q) or at the active site serine residue (S637A), were produced. Both wild-type fragments were recovered as two-chain, activated proteases, whereas all mutants retained a single-chain, proenzyme structure, providing the first experimental evidence that C1r activation is an autolytic process. As shown by sedimentation velocity analysis, all CCP1/2-SP fragments were dimers (5.5-5.6 S), and all CCP2-SP fragments were monomers (3.2-3.4 S). Thus, CCP1 is essential to the assembly of the dimer, but formation of a stable dimer is not a prerequisite for self-activation. Activation of the R446Q mutants could be achieved by extrinsic cleavage by thermolysin, which cleaved the CCP2-SP species more efficiently than the CCP1/2-SP species and yielded enzymes with C1s-cleaving activities similar to their active wild-type counterparts. C1r and its activated fragments all cleaved C1s, with relative efficiencies in the order C1r < CCP1/2-SP < CCP2-SP, indicating that CCP1 is not involved in C1s recognition.

Functional determinants of the Epstein-Barr virus protease.

Herpesvirus proteases are essential for the production of progeny virus. They cleave the assembly protein that fills the immature capsid in order to make place for the viral DNA. The recombinant protease of the human gamma-herpesvirus Epstein-Barr virus (EBV) was expressed in Escherichia coli and purified. Circular dichroism indicated that the protein was properly folded with a secondary structure content similar to that of other herpesvirus proteases. Gel filtration and sedimentation analysis indicated a fast monomer-dimer equilibrium of the protease with a K(d) of about 60 microM. This value was not influenced by glycerol but was lowered to 1.7 microM in the presence of 0.5 M sodium citrate. We also developed an HPLC-based enzymatic assay using a 20 amino acid residue synthetic peptide substrate derived from one of the viral target sequences for the protease. We found that conditions that stabilised the dimer also led to a higher enzymatic activity. Through sequential deletion of amino acid residues from either side of the cleavage site, the minimal peptide substrate for the protease was determined as P5-P2'. This minimal sequence is shorter than that for other herpesvirus proteases. The implications of our findings are discussed with reference to the viral life-cycle. These results are the first ever published on the EBV protease and represent a first step towards the development of protease inhibitors.

Structural biology of C1: dissection of a complex molecular machinery.

The classical pathway of complement is initiated by the C1 complex, a multimolecular protease comprising a recognition subunit (C1q) and two modular serine proteases (C1r and C1s) associated as a Ca2+-dependent tetramer (C1s-C1r-C1r-C1s). Early studies have allowed identification of specialized functional domains in these proteins and have led to low-resolution models of the C1 complex. The objective of current studies is to gain deeper insights into the structure of C1, and the strategy used for this purpose mainly consists of dissecting the C1 components into modular fragments, in order to solve their three-dimensional structure and establish the structural correlates of their function. The aim of this article is to provide an overview of the structural and functional information generated by this approach, with particular emphasis on the domains involved in the assembly, the recognition function, and the highly specific proteolytic properties of C1.

Interaction properties of human mannan-binding lectin (MBL)-associated serine proteases-1 and -2, MBL-associated protein 19, and MBL.

The mannan-binding lectin (MBL) activation pathway of complement plays an important role in the innate immune defense against pathogenic microorganisms. In human serum, two MBL-associated serine proteases (MASP-1, MASP-2) and MBL-associated protein 19 (MAp19) were found to be associated with MBL. With a view to investigate the interaction properties of these proteins, human MASP-1, MASP-2, MAp19, as well as the N-terminal complement subcomponents C1r/C1s, Uegf, and bone morphogenetic protein-1-epidermal growth factor (CUB-EGF) segments of MASP-1 and MASP-2, were expressed in insect or human kidney cells, and MBL was isolated from human serum. Sedimentation velocity analysis indicated that the MASP-1 and MASP-2 CUB-EGF segments and the homologous protein MAp19 all behaved as homodimers (2.8-3.2 S) in the presence of Ca(2+). Although the latter two dimers were not dissociated by EDTA, their physical properties were affected. In contrast, the MASP-1 CUB-EGF homodimer was not sensitive to EDTA. The three proteins and full-length MASP-1 and MASP-2 showed no interaction with each other as judged by gel filtration and surface plasmon resonance spectroscopy. Using the latter technique, MASP-1, MASP-2, their CUB-EGF segments, and MAp19 were each shown to bind to immobilized MBL, with K:(D) values of 0.8 nM (MASP-2), 1.4 nM (MASP-1), 13.0 nM (MAp19 and MASP-2 CUB-EGF), and 25.7 nM (MASP-1 CUB-EGF). The binding was Ca(2+)-dependent and fully sensitive to EDTA in all cases. These data indicate that MASP-1, MASP-2, and MAp19 each associate as homodimers, and individually form Ca(2+)-dependent complexes with MBL through the CUB-EGF pair of each protein. This suggests that distinct MBL/MASP complexes may be involved in the activation or regulation of the MBL pathway.

Crystal structure of the catalytic domain of human complement c1s: a serine protease with a handle.

C1s is the highly specific modular serine protease that mediates the proteolytic activity of the C1 complex and thereby triggers activation of the complement cascade. The crystal structure of a catalytic fragment from human C1s comprising the second complement control protein (CCP2) module and the chymotrypsin-like serine protease (SP) domain has been determined and refined to 1.7 A resolution. In the areas surrounding the active site, the SP structure reveals a restricted access to subsidiary substrate binding sites that could be responsible for the narrow specificity of C1s. The ellipsoidal CCP2 module is oriented perpendicularly to the surface of the SP domain. This arrangement is maintained through a rigid module-domain interface involving intertwined proline- and tyrosine-rich polypeptide segments. The relative orientation of SP and CCP2 is consistent with the fact that the latter provides additional substrate recognition sites for the C4 substrate. This structure provides a first example of a CCP-SP assembly that is conserved in diverse extracellular proteins. Its implications in the activation mechanism of C1 are discussed.

Grafting an RGD motif onto an epidermal growth factor-like module: chemical synthesis and functional characterization of the chimeric molecule.

A novel protein was engineered by inserting the GRGDS motif of fibronectin within the 14-residue loop of the EGF-like module from human complement protease C1r. The resulting chimeric EGF-RGD module (52 residues, three disulfide bridges) was assembled by automated solid-phase synthesis using the t-Boc strategy. Using reduced/oxidized glutathione, the EGF-RGD module was folded as efficiently as the natural C1r-EGF module, resulting in formation of the appropriate disulfide bridge pattern as shown by mass spectrometry and N-terminal sequence analyses of thermolytic fragments. Circular dichroism and NMR measurements provided further indication that introduction of the GRGDS motif had no significant effect on the folding. Using Chinese Hamster Ovary (CHO) cells bearing the integrin receptors specific for fibronectin and vitronectin, EGF-RGD was shown to induce cell adhesion via the introduced GRGDS motif. Cell binding was inhibited specifically and efficiently by the synthetic peptide GRGDSP and by fibronectin, and to a much lesser extent by vitronectin, whereas the monoclonal antibody PB1 directed to the alpha5 subunit of alpha5beta1 integrin had no effect. The ability of EGF-RGD to trigger significant cell spreading and intracellular signaling was also demonstrated using immunofluorescence and confocal microscopy.

Mapping of the interaction between the immunodominant loop of the ectodomain of HIV-1 gp41 and human complement protein C1q.

The human immunodeficiency virus type 1 transmembrane envelope glycoprotein gp41 has been previously shown to activate the C1 complex of human complement through direct interaction with its C1q subunit. The major interaction site has been located within the gp41 immunodominant region (residues 590-620), and a synthetic peptide overlapping residues 601-613 of gp41 (sequence GIWGCSGKLICTT) was shown to inhibit binding of gp41 to C1q in vitro (Thielens, N.M., Bally, I.M., Ebenbichler, C.F., Dierich, M.P. & Arlaud, G.J. (1993) J. Immunol. 151, 6583-6592). The ectodomain of gp41 (s-gp41) was secreted from the methylotrophic yeast Pichia pastoris and purified by immunoaffinity chromatography. Enzymatic deglycosylation of the recombinant s-gp41 was necessary to allow its in vitro interaction with C1q. A solid-phase competition assay was used to monitor the effect of mutant peptides derived from segment 601-613 of gp41 on the binding of deglycosylated s-gp41 to C1q. Whereas mutation of Ser606 had no effect, replacement of Ile602, Trp603, Lys608, Leu609 and Ile610 by Ala abolished the ability of the resulting peptides to inhibit binding of s-gp41 to C1q, suggesting that these residues participate in the interaction between gp41 and C1q. These findings are discussed in the light of a structural model of the immunodominant loop of gp41. It is proposed that the recognition of gp41 by C1q is driven by hydrophobic interactions, and that the sites of gp41 responsible for interaction with gp120 and C1q partly overlap.

A coil-helix instead of a helix-coil motif can be induced in a chloroplast transit peptide from Chlamydomonas reinhardtii.

A synthetic peptide MQVTMKSSAVSGQRVGGARVATRSVRRAQLQV corresponding to the 32 amino acid chloroplast transit sequence of the ribulose bisphosphatase carboxylase/oxygenase activase preprotein from Chlamydomonas reinhardtii, required for translocation through the envelope of the chloroplast, has been characterized structurally using CD and NMR under the same experimental conditions as used previously for the 32 amino acid presequence of preferredoxin from the same organism [Lancelin, J.-M., Bally, I., Arlaud, G. J., Blackledge, M., Gans, P., Stein, M. & Jacquot, J.-P. (1994) FEBS Lett. 343, 261-266]. The peptide is found to undergo a conformational transition in aqueous 2,2,2-trifluoroethanol, characterized by three turns of amphiphilic alpha-helix in the C-terminal region preceded by a disordered coil in the N-terminal region. Compared with the preferredoxin transit peptide, the helical and coiled domains are arranged in the reverse order along the peptide sequence, but the positively charged groups are distributed analogously as well as the hydrophobic residues within the amphiphilic alpha-helix. It is proposed that such coil-helix or helix-coil motifs, occasionally repeated, could be an intrinsic structural feature of chloroplastic transit peptides, adapted to the proper translocase and possibly to each nuclear-encoded chloroplast preproteins. This feature may distinguish chloroplastic transit sequences from the other organelle-targeting peptides in the eukaryotic green alga C. reinhardtii, particularly the mitochondrial transit sequences.

Structure and functions of the interaction domains of C1r and C1s: keystones of the architecture of the C1 complex.

C1r and C1s, the proteases responsible for activation and proteolytic activity of the C1 complex of complement, share similar overall structural organizations featuring five nonenzymic protein modules (two CUB modules surrounding a single EGF module, and a pair of CCP modules) followed by a serine protease domain. Besides highly specific proteolytic activities, both proteases exhibit interaction properties associated with their N-terminal regions. These properties include the ability to bind Ca2+ ions with high affinity, to associate with each other within a Ca2+-dependent C1s-C1r-C1r-C1s tetramer, and to interact with C1q upon C1 assembly. Precise functional mapping of these regions has been achieved recently, allowing identification of the domains responsible for these interactions, and providing a comprehensive picture of their structure and function. The objective of this article is to provide a detailed and up-to-date overview of the information available on these domains, which are keystones of the assembly of C1, and appear to play an essential role at the interface between the recognition function of C1 and its proteolytic activity.

In vitro analysis of complement-dependent HIV-1 cell infection using a model system.

Previous studies based on the use of human serum as a source of C have provided evidence for the C-dependent enhancement of cell infection by HIV-1. The present study was undertaken to distinguish C from other serum factors and to identify the proteins and the mechanisms involved in C-dependent cell infection by HIV-1. The classical C activation pathway was reconstituted from the proteins C1q, C1r, C1s, C4, C2, C3, factor H, and factor I; each were purified to homogeneity. A mixture of these proteins at physiological concentrations was shown to reproduce the ability of normal human serum to enhance the infection of MT2 cells by HIV-1 at low doses of virus. This enhancing effect was abolished when heat-inactivated serum and C2- or C3-depleted serum were used, and was restored upon addition of the corresponding purified proteins. A mixture of two synthetic peptides corresponding to positions 10-15 and 90-97 of human C receptor type 2 (CD21) as well as soluble CD4 both inhibited the C-dependent infection process. These data provide unambiguous evidence that HIV-1 triggers a direct activation of the classical C pathway in vitro and thereby facilitates the infection of MT2 cells at low doses of virus. These findings are consistent with a mechanism involving increased interaction between the virus opsonized by C3b-derived fragment(s) and the CD21 cell receptors and subsequent virus entry through CD4 receptors.

The N-terminal CUB-epidermal growth factor module pair of human complement protease C1r binds Ca2+ with high affinity and mediates Ca2+-dependent interaction with C1s.

The Ca2+-dependent interaction between complement serine proteases C1r and C1s is mediated by their alpha regions, encompassing the major part of their N-terminal CUB-EGF-CUB (where EGF is epidermal growth factor) module array. In order to define the boundaries of the C1r domain(s) responsible for Ca2+ binding and Ca2+-dependent interaction with C1s and to assess the contribution of individual modules to these functions, the CUB, EGF, and CUB-EGF fragments were expressed in eucaryotic systems or synthesized chemically. Gel filtration studies, as well as measurements of intrinsic Tyr fluorescence, provided evidence that the CUB-EGF pair adopts a more compact conformation in the presence of Ca2+. Ca2+-dependent interaction of intact C1r with C1s was studied using surface plasmon resonance spectroscopy, yielding KD values of 10.9-29.7 nM. The C1r CUB-EGF pair bound immobilized C1s with a higher KD (1.5-1.8 microM), which decreased to 31.4 nM when CUB-EGF was used as the immobilized ligand and C1s was free. Half-maximal binding was obtained at comparable Ca2+ concentrations ranging from 5 microM with intact C1r to 10-16 microM for C1ralpha and CUB-EGF. The isolated CUB and EGF fragments or a CUB + EGF mixture did not bind C1s. These data demonstrate that the C1r CUB-EGF module pair (residues 1-175) is the minimal segment required for high affinity Ca2+ binding and Ca2+-dependent interaction with C1s and indicate that Ca2+ binding induces a more compact folding of the CUB-EGF pair.

Structural and functional studies on C1r and C1s: new insights into the mechanisms involved in C1 activity and assembly.

C1r and C1s, the enzymes responsible for the activation and proteolytic activity of the C1 complex of complement, are modular serine proteases featuring similar overall structural organizations, yet expressing very distinct functional properties within C1. This review will initially summarize available information on the structure and function of the protein modules and serine protease domains of C1r and C1s. It will then focus on the regions of both proteases involved in: (i) assembly of C1s-C1r-C1r-C1s, the Ca(2+)-dependent tetrameric catalytic subunit of C1; (ii) expression of C1 catalytic activities. Particular emphasis will be aid on recent structural and functional studies that provide new insights into the complex mechanisms involved in the assembly, activation, and proteolytic activity of C1.

Evolutionary conserved rigid module-domain interactions can be detected at the sequence level: the examples of complement and blood coagulation proteases.

Several extracellular modular proteins, including proteases of the complement and blood coagulation cascades, are shown here to exhibit conserved sequence patterns specific for a particular module-domain association. This was detected by comparative analysis of sequence variability in different multiple sequence alignments, which provides a new tool to investigate the evolution of modular proteins. A first example deals with the proteins featuring a common complement control protein (CCP) module-serine protease (SP) domain pattern at their C-terminal end, defined here as the CCP-SP sub-family. These proteins include the complement proteases C1r, C1s and MASPs, the Limulus clotting factor C, and the proteins of the haptoglobin family. A second example deals with blood coagulation factors VII, IX and X and protein C, all featuring a common epidermal growth factor (EGF)-SP C-terminal assembly. Highly specific motifs are found at the connection between the CCP or EGF module and the activation peptide of the SP domain: [P/A]-x-C-x-[P/A]-[I/V]-C-G-x-[P/S/K] in the case of the CCP-SP proteins, and C-x-[P/S]-x-x-x-[Y/F]-P-C-G in the case of the EGF-SP proteins. Each motif is strictly conserved in the whole sub-family and it is detected in no more than one other known protein sequence. Strikingly, most of the conserved residues specific to each sub-family appear to be clustered at the interface between the SP domain and the CCP or EGF module. We propose that a rigid module-domain interaction occurs in these proteins and has been conserved through evolution. The functional implications of these assemblies, underlined by such evolutionary constraints, are discussed.