A study on the structure and interactions of the C1 sub-components C1r and C1s in the fluid phase.

1. Both proenzyme and activated C1r, which are dimers at pH 7.4, dissociated into monomers at pH 5.0 (C1r) and 4.0 (C1r), as shown by the decrease of apparent molecular weight and of sedimentation coefficient, which was shifted from 7.1 S (dimer) to 5.0 S (monomer). 125I-labelling of C1r in the presence of lactoperoxidase occurred, for the dimer, 16-20% in the A chain and 80-84% in the B chain, whereas the distribution was 67.5% and 32.5%, respectively, for the monomer. It appears likely that the two monomers of C1r interact through their A chain and that the A and B chains are relatively independent from each other. 2. 125I-labelling of C1s in the presence of lactoperoxidase confirmed the calcium-dependent dimerization of this subcomponent. In the monomer, the B chain appears to be embedded in the A chain, as shown by the 125I- distribution in these chains, which was 5% and 95%, respectively. This changed after dimerization to 25% and 75%, respectively, which suggests that interactions occur through the A chain of each monomer and lead to an unfolding of the B chain. 3. C1r dimer and C1s monomer were found to interact in the absence of calcium to form a C1r2-C1s complex (7.7 S), whereas in the presence of calcium the two sub-components were associated into a C1r2-C1s2 complex (8.7S). It appears likely that the formation of this tetrameric complex involves both calcium-dependent, and calcium-independent binding forces, and that C1r and C1s interact through their respective A chain which, in the case of C1s, is hidden upon association.

Fluid phase activation of proenzymic C1r purified by affinity chromatography.

1. Proenzymic C1r was purified from human plasma in a two-step technique involving indirect affinity chromatography on Sepharose Ig anti-C1s. The capacity of C1r to monomerize at pH 5.0 and to redimerize at neutral pH was used for selective elution of C1r. The yield in purified C1r was 39% from plasma; no trace of contaminating serine proteases was detected from [3H]diisopropyl phosphorofluoridate labelling of C1r. 2. C14 was able to undergo a two-way autoactivation: an intramolecular catalytic process catalysed by proenzymic C1r itself and an intermolecular reaction catalysed by activated C1r formed in the process of the reaction. DFP (5mM) and C1 Inh at a C1 Inh/C1r ratio of 1:1 were effective on the solely intermolecular activation, leading to partial inhibition of the autoactivation from proenzymic C1r: C1r formed during the activation was titrated by the inhibitors. Calcium, high ionic strength or acid pH decreased C1r activation. The pH effect was characterized by a slowed-down reaction below pH 6.0 and no net influence at values as high as 10.5. The two types of activation developed similarly as a function of pH. 3. Peripheral iodination of C1r revealed differences in label distribution between proenzymic (A chain moiety 48%, B chain moiety 52%) and activated C1r (A chain 20%, B chain 80%). Two different conformational states of C1r were also suggested by 125I-labelling at different temperatures.

Purified proenzyme C1r. Some characteristics of its activation and subsequent proteolytic cleavage.

1. Upon incubation for 1 h at 37 degrees C, proenzymic C1r was activated by a proteolytic cleavage comparable to that observed in vivo; after reduction and alkylation, two fragments of apparent molecular weights 57 000 and 35 000 were evident on sodium dodecyl sulphate (SDS)-polyacrylamide gel electrophoresis. The activation kinetics were slightly sigmoidal and nearly independent of C1r concentration. They were characterized by a marked thermal dependence (activation energy = 45 kcal/mol). The reaction was inhibited by calcium and p-nitrophenyl-p'-guanidinobenzoate, but poorly sensitive to di-isopropyl phosphorofluoridate. The dependence of the activation rate on pH was unusual; it decreased progressively in the acid range (pH 4.5-6.5) which coincides with the dissociation of the C1r-C1r dimer. Above pH 6.5, the rate increased slightly and showed no clear maximum. These results are consistent with an intramolecular autocatalytic activation mechanism involving the pro-site of each subunit of the C1r-C1r dimer. 2. During a 5 h incubation period at 37 degrees C, C1r underwent two proteolytic cleavages which led to the successive removal of two fragments, alpha (35 000) and beta (7000-11 000) from each subunit, leaving a dimeric molecule of reduced size (Mr = 110 000; s20,w = 6.1 S). The proteolytic process was nearly independent of C1r concentration and characterized by a pH optimum at 8.5-9.0, and a high activation energy (36.8 kcal/mol). Calcium and p-nitrophenyl-p'-guanidinobenzoate, and also di-isopropyl phosphorofluoridate and benzamidine were inhibitors of this reaction. The product, C1r II, retained the original antigenic properties of C1r and a functional active site, but lost the capacity to bind C1s. These results are consistent with an autocatalytic intramolecular proteolysis mediated by the active site of each subunit of the C1r-C1r dimer.

Neutron scattering studies of subcomponent C1q of first component C1 of human complement and its association with subunit C1r2C1s2 within C1.

Neutron scattering studies are reported on subcomponent C1q of component C1 of human complement, and on C1, the complex of C1q with subunit C1r2C1s2. For C1q, the molecular weight was determined as 460,000. The radius of gyration at infinite contrast RC is 12.8 nm. The RC values for the proteolytically cleaved forms of C1q, namely the heads and the stalks, are 1.5 to 2 nm and 11 nm, respectively, and thus the axis-to-arm angle of C1q is estimated at 45 degrees. Neutron data for subunit C1r2C1s2 are published elsewhere. The neutron data on C1 lead to an RC value of 12.6 nm for proenzymic C1 and a molecular weight of 820,000. The wide-angle scattering curve of C1q exhibits a minimum at Q = 0.28 nm-1 and a maximum at 0.39 nm-1; on the addition of C1r2C1s2, this minimum disappears. The neutron data on C1 indicate that C1q and C1r2C1s2 have complexed with a large conformational change in one or both parts. No conformational changes can be detected on the activation of C1 by this method.

Covalent binding of non-proteolysed C3 to Jurkat T cells.

Purified C3 binds covalently to Jurkat T cells upon incubation at neutral pH. This binding does not appear to involve proteolysis of C3; it leads to high-molecular-weight associations, preferentially through ester linkages, which are disrupted upon incubation with hydroxylamine at alkaline pH. Part of the association also appears to involve disulfide links between C3 and Jurkat cells. Similarly, plasma membranes purified from these cells bind C3 with no evidence for proteolysis of C3. Binding of C3 appears to be "catalysed" by Jurkat cells, and is not due to the well-known spontaneous hydrolysis of C3. Binding of C3 involves hydrolysis of its thioester bond, as titratable--SH groups are available in soluble C3 after incubation of purified C3 with Jurkat plasma membranes; loss of C3 haemolytic activity confirms this finding. These observations give evidence for the binding of C3b-like C3 to Jurkat cells, conferring on these cells the potential to interact with other complement receptor-bearing cells such as B cells.

Involvement of the Zn-binding region of tetanus toxin in B and T recognition. Influence of Zn fixation.

Tetanus toxin contains a metal-binding site for zinc, located in its light chain. The sequence accounting for Zn fixation is part of a predicted amphipathic helical secondary structure and corresponds to a putative T cell epitope according to Rothbard and Taylor (EMBO J. 7, 93-100, 1988). In this paper, we analyse the antigenic properties of two synthetic peptides (233-248 = P12 and 225-243 = P13) containing the Zn binding sequence. Our results show that peptide P13 contains a B and T epitope. The B epitope seems to be immuno-dominant whether the T epitope is at least DR2 restricted. Zn binding on P13 leads to a decrease in its recognition by both antibodies and T lymphocytes.

C3b covalently associated to tetanus toxin modulates TT processing and presentation by U937 cells.

Complement protein C3, like C4 and alpha 2-macroglobulin (alpha 2M), is a potentially bivalent ligand: (1) its proteolytic fragment, C3b, is able to interact covalently with antigens, and (2) this bound fragment is able to interact non-covalently with specific complement receptors of antigen presenting cells (APC). The formation of antigen-C3b complexes frequently occurs in vivo at inflammatory sites during the early stages of an immune response. Tetanus toxin (TT)-C3b covalent complexes, prepared from purified proteins, were used to study how C3b association influences the handling of TT by U937 cells used as APC. TT-specific T cell proliferation following TT-C3b processing was observed at a concentration when TT alone was inefficient. Whereas TT pinocytic uptake was low, TT-C3b uptake, through the help of complement receptor CR1, was three times higher. Free TT was rapidly transported to the lysosomes where it was proteolysed, whereas TT-C3b complexes were first retained in the endosomes and underwent only limited proteolysis. While the ester link of the complexes was fairly stable in the endosomes, it was gradually hydrolysed in the lysosomes with ensuing efficient proteolysis of the two proteins. This reflects the fact that associated C3b escorts TT during intracellular trafficking in the APC, and influences antigen processing. A triple role of C3b escorting antigen residues at the level of antigen uptake, routing, and proteolysis inside U937 cells, thus modulating antigen-dependent T cell proliferation.

Comparative study of the fluid-phase proteolytic cleavage of human complement subcomponents C4 and C2 by C1s and C1r2-C1s2.

The C3 convertase of the classical pathway of complement is composed of fragments C4b and C2a resulting from cleavage of C4 and C2 by activated C1. The limited proteolysis of these two different substrates by the same protease, C1s, has been studied in the fluid phase using purified proteins. The turnover numbers of C2 and C4 cleavage by C1s were affected to different extents, depending on whether C1s was alone or associated with C1r or with monoclonal antibodies to C1s. The binding of C2 to C4 favours the proteolysis of C2 by C1s, as revealed by the use of I2-treated C2.

C1r serine proteinase of human complement: a case of intramolecular autolytic activation.

This paper presents a short review of our contribution to the knowledge of the structure and function of human C1r, the activation unit of C1, the first component of the classical pathway of complement. On the basis of the domain structure of C1r, a model accounting for its autolytic activation mechanism is proposed. We suggest that this represents the basic mechanism of C1 function.

[Role of the complement C3 protein in the control of the specific immune response]. / Rôle de la protéine C3 du système complémetnaire dans le contrôle de la réponse immunitaire spécifique.

Whereas complement system was usually considered as a member of innate defence, one of its components (C3) is now thought to facilitate acquired immunity. This role is due first to its capacity to covalently bind to antigens and secondly to the interactions of its proteolytic fragments with different receptors expressed on most cells involved in the acquired immune response. After activation in the plasma, C3 is proteolysed in fragments which possess various biological activities, as a modification in cell activities occurred after binding to cell surface receptors. Injection of low amount of antigen results in a modified immune response in C3 deficient animals with a decrease in the level of specific antibodies and an absence of IgM/IgG switch. One of the fragments of C3 (C3d) and its receptor (CR2) seem particularly important : knock out animals in C3 or CR2 have similar phenotypes. Mice with a deficit in CR2 restricted to the B lymphocytes present a strong reduction in the number and the size of germinal centers. Moreover, expression of CR2 on follicular dendritic cells is necessary for the generation of a strong memory response. C3 is also involved in the control of the intracellular processing of the antigen as the use of covalent complex (C3b-antigen) instead of free antigen increases the amount of stable MHC class II molecules at the antigen presenting cells surface. In summary, C3 fragments increase cell to cell interactions, induce intracellular signalling after binding to their receptors and increases intracellular processing of antigens. A better knowledge of the different roles of C3 may be useful to modify the immune response and to promote the immune memory, a domain where C3 seems particularly important.

Diamine-induced dissociation of the first component of human complement, C1.

Lysine has been shown to inhibit spontaneous and antibody-dependent C1 activation. This paper demonstrates that lysine does not prevent autoactivation of purified C1r. 20 mM lysine, 1,2-diaminoethane, 1,3-diaminopropane, 1,4-diaminobutane or 1,5-diaminopentane are able to dissociate C1 into its two entities, C1q and the calcium-dependent C1r2-C1s2 complex. Ig-ovalbumin insoluble complexes bearing C1 are also dissociated by lysine and the above-mentioned diamines used at the same concentration: C1q remains bound to the complexes whereas the C1r2-C1s2 complex is partially solubilized. The effect of lysine or diamines is not due to a competition with calcium for calcium-binding sites, as increasing concentrations of calcium even slightly increase the dissociation due to the amines. The dissociative effect is dependent on the carbon chain length of the diamines, with an optimum for 1,3-diaminopropane. It is also dependent on the relative 'cis-position' of the amino groups in the diamines. Polyamines such as spermine and spermidine are also able to dissociate C1 with even a higher efficiency than lysine and putrescine. Thus, a diamine-induced 'structural inhibition' of C1 is demonstrated, of potential interest for a pharmacological control of complement activation.

Proteolysis of C3 on U937 cell plasma membranes. Purification of cathepsin G.

Covalent binding of C3 fragments to U937 cell membranes involved a cell surface-associated proteolytic activity. Two proteases able to cleave C3 were purified from U937 plasma membranes. Purification involved solubilization of the membranes and ion exchange chromatography. One of the purified proteases was identified as elastase, based upon a substrate specificity for benzyloxycarbonylalanine-o-nitrophenyl ester and complete inhibition by elastatinal and methoxysuccinyl-alanyl-alanyl-prolyl-valyl-chloromethyl-ketone. The other protease (m.w. 28,000) is cathepsin G, as deduced from the amino acid composition, the amino-terminal sequence, and the substrate specificity for succinyl-alanyl-alanyl-phenylalanine-p-nitroanilide. These two lysosomal proteases are present on the U937 cell surface, as confirmed by immunofluorescence analysis. Plasma membrane elastase and cathepsin G from U937 cells cleave C3 into C3a- and C3b-like fragments; further incubation leads to C3c- and C3dg-like fragments, as judged from SDS-PAGE analysis of the digests. Sequencing of the C3b-like fragment purified by reverse phase chromatography indicates that initial cleavage of C3 by purified cathepsin G occurs at two positions in the amino-terminal part of the alpha-chain, at a Arg-Ser bond located between residues 748 and 749 and at a Leu-Asp bond between residues 751 and 752. These proteases are, thus, able to generate, on the U937 surface, active fragments of C3, which are likely to be involved in cell-protein and cell-cell interactions.

Domain structure and associated functions of subcomponents C1r and C1s of the first component of human complement.

The serine protease subcomponents of the activated form of the first component of human complement (C1), C1r and C1s, were observed by electron microscopy after the native proteins and their limited proteolysis products, obtained from autolytic cleavage (C1r) or from incubation with plasmin (C1s) were rotary shadowed. At the monomeric level, both C1r and C1s comprised two globular domains, a smaller interaction domain (corresponding to the NH2-terminal half of the A chain, alpha, and responsible for calcium binding and C1r-C1s interaction) and a larger catalytic domain (corresponding to the COOH-terminal part of the A chain, gamma, disulfide-linked to the B chain and bearing the serine protease active site). The two globular domains are linked by a connecting strand, beta. The (C1r)2 dimer appeared as a "croissant"-like association, where the two monomers interact through their catalytic domains. On the basis of the domain structure of C1r and C1s, a model of the calcium-dependent C1s dimer is proposed, in which the two monomers interact through their NH2-terminal interaction domains; in the same way, a model of the C1s-(C1r)2-C1s catalytic subunit of C1 is presented, in which (C1r)2 forms a core, its distal interaction domains interacting with the corresponding domains of C1s.

Autoactivation of human complement subcomponent C1r involves structural changes reflected in modifications of intrinsic fluorescence, circular dichroism and reactivity with monoclonal antibodies.

Autoactivation of C1r is closely correlated with an irreversible increase of its intrinsic fluorescence. The activation and the fluorescence increase of C1r are accelerated on addition of activated C1r. Ca2+, di-isopropyl phosphorofluoridate and C1 inhibitor, which all inhibit, although to different extents, C1r activation, inhibit in parallel the fluorescence increase. C1r activation is blocked at pH 4.0-5.0, whereas it is accelerated at pH 10.5; under the same conditions the fluorescence increase shows parallel effects. No such fluorescence increase is observed during C1s activation by trace amounts of C1r. Far-u.v. circular-dichroism spectra of C1r indicate 73 and 78% of unordered form in both the proenzyme and the activated species respectively. The slight changes observed on activation are not restricted to C1r, as comparable results are obtained for proenzyme and activated C1s. C1r activation appears thus to involve structural changes leading to an 'activated state' distinct from the 'proenzyme state'. Monoclonal antibody to activated C1r is poorly reactive with proenzyme C1r, a finding that also supports this hypothesis.

Activation of C1.

The first component of complement, C1, is a calcium-dependent complex of two loosely interacting subunits: C1q, responsible for the binding of activators to C1; C1r2-C1s2, which supports the autoactivation potential of C1, together with the proteolytic activity of activated C1- on its two substrates, C4 and C2. Isolated dimeric C1r2 is able to autoactivate through an intradimer cross-proteolysis; this capacity is lost when C1r2 is associated with two molecules of C1s inside the calcium-dependent C1r2-C1s2 subunit; this capacity is again observed in reconstituted C1. A model for reconstituted soluble C1 is proposed, based on electron microscopy, neutron diffraction, ultra-centrifugation, various biochemical findings, as well as functional properties of C1 or of its subcomponents. The flexible rod-like structure of C1r2-C1s2 is folded around two arms of C1q, with the catalytic domains of C1r and C1s inserted inside the cone defined by the C1q stalks. Activation of C1 which, in vivo, is controlled by C1 inhibitor, can be achieved by various activators, such as immune complexes; it appears to result from the suppression of a negative control and resides in a positive modulation of the intrinsic autocatalytic potential of C1r inside C1.

Secretion, cleavage and binding of complement component C3 by the human monocytic cell line U937.

Secretion of complement component C3 by U937 cells was studied. Preliminary evidence for a cell-associated proteolytic activity specific for C3 is given, as well as for a covalent-like binding of C3 fragments to the cell membranes. Secretion of C3, in the presence of 10 ng of phorbol 12-myristate 13-acetate/ml, is 120-140 ng/10(6) cells per 24 h on the third day after addition of the activator. As shown by SDS/polyacrylamide-gel electrophoresis, the intracellular pro-C3 (200 kDa) and the extracellular secreted C3 (alpha-chain 110 kDa and beta-chain 75 kDa) are identical with the forms of C3 previously characterized from human serum. Incubation of U937 cells in the presence of exogenous radiolabelled C3 shows that membrane-bound proteinase(s), not related to the classical-pathway or the alternative-pathway C3 convertases, is (are) able to cleave C3; this cleavage leads to the binding of the resulting C3 fragments to the cell membrane through reaction of membrane acceptors with the carbonyl group of C3 revealed after disruption of the intramolecular thioester bond. The proteolysis appears to be fairly specific to C3, as C4, which also possesses an intramolecular thioester bond, is not cleaved and does not bind to the cells. p-Nitrophenyl p'-guanidinobenzoate (1 mM) and di-isopropyl phosphorofluoridate (2 mM) are potent inhibitors of the proteolysis, whereas soya-bean trypsin inhibitor (1 mM), leupeptin (0.1 mg/ml) and 1,10-phenanthroline (1 mM) were ineffective. Immunological characterization of the cell-bound C3 fragments with monoclonal antibodies shows an evolution of the proteolysis of the fragments from iC3b to C3dg epitopes. Extraction of membrane-bound fragments by detergent, followed by SDS/polyacrylamide-gel electrophoresis, shows two fragments, of 43 kDa and 46 kDa, with C3dg-like characteristics.

Prolongation of cell cycle transit time and the presence of non-cycling cells in human lymphoblastoid cells cultured under adverse conditions.

The growth characteristics of B lymphocytes infected with Epstein-Barr virus (lymphoblastoid cells) have been investigated by flow cytometric analysis of DNA content and by estimation of cell culture doubling times. It was found that the manipulative procedures involved in the cell cycle analysis resulted in a slowing of the growth rate. This slowing of growth was brought about by the prolongation of cell cycle transit times and by the entry of cells into a short-lived non-cycling pool. The entry of a proportion of the cells into the non-cycling pool may be the normal response of lymphoblastoid cells to non-optimal conditions. The non-cycling cells survived in culture with a T 1/2 of approximately 30-60 hr and continued to secrete immunoglobulin. Their surface transferrin receptors were considerably reduced, which suggests that the failure to divide may have resulted from a failure of growth factor receptors to reach a threshold value following mitosis.