Structural requirements for the complement regulatory activities of C4BP.

C4b-binding protein (C4BP) is a regulator of the classical complement pathway C3 convertase (C4bC2a complex). It is a disulfide-linked polymer of seven alpha-chains and a unique beta-chain; the alpha- and beta-chains are composed of eight and three complement control protein (CCP) domains, respectively. To elucidate the importance of the polymeric nature of C4BP and the structural requirements for the interaction between C4b and the alpha-chain, 19 recombinant C4BP variants were created. Six truncated monomeric variants, nine polymeric variants in which individual CCPs were deleted, and finally, four variants in which double alanine residues were introduced between CCPs were functionally characterized. The smallest truncated C4BP variant still active in regulating fluid phase C4b comprised CCP1-3. The monomeric variants were less efficient than polymeric C4BP in degrading C4b on cell surfaces. All three N-terminal CCP domains contributed to the binding of C4b and were important for full functional activity; CCP2 and CCP3 were the most important. The spatial arrangements of the first CCPs were found to be important, as introduction of alanine residues between CCPs 1 and 2, CCPs 2 and 3, and CCPs 3 and 4 resulted in functional impairment. The results presented here elucidate the structural requirements of individual CCPs of C4BP, as well as their spatial arrangements within and between subunits for expression of full functional activity.

Human C4b-binding protein (C4BP) 1 is a high molecular mass (570 kDa) plasma glycoprotein that efficiently inhibits the classical pathway of complement activation. Apart from preventing the assembly of the C3-convertase (C4b2a complex), it also accelerates the natural decay of the complex (1). In addition, C4BP binds C4b and serves as a cofactor to the plasma serine protease factor I in the cleavage of C4b both in fluid phase and when C4b is deposited on cell surfaces (2). C4BP belongs to a gene family of related proteins named the regulators of complement activation, which also includes factor H, complement receptors 1 and 2, membrane cofactor protein, and decay accelerating factor (3). Each of the proteins of the regulators of complement activation family binds C4b and/or C3b and is important for the inhibition of the classical and/or alternative pathways of complement activation. All these proteins contain variable numbers of tandemly arranged domains, which are denoted complement control protein (CCP) repeats. These domains are cysteine-rich and ϳ60 amino acid residues long, and each is composed of a hydrophobic core that is wrapped by ␤-strands (4). The major form of C4BP in plasma consists of seven identical ␣-chains (eight CCPs each) and one ␤-chain (three CCPs), and all of the chains are linked together by disulfide bridges (2,5). Electron microscopic analysis of C4BP demonstrated an octopus-like conformation, with the seven ␣-chains forming extended tentacles (6).
We have recently localized the C4b binding site to the interface between CCP1 and CCP2 of the ␣-chain and identified a cluster of positively charged amino acids that are important for the binding (7,8). This region of C4BP is also important for binding of C4BP to heparin (7), Bordetella pertussis (9), and M-proteins of Streptococcus pyogenes (10). The aim of the present study was to further characterize the structural requirements of C4BP for its binding to C4b and for its factor I cofactor activity. Recombinant C4BP lacking individual CCP domains and truncated monomeric C4BP variants were created and functionally characterized. The elongated conformation of ␣-chains and the fact that they are exclusively organized into repetitive, individually folding domains makes C4BP particularly suitable for this kind of studies. A total of 19 C4BP variants were studied, and based on their properties, we conclude that CCP2 and CCP3 are the most important domains, being crucial for binding of C4b and for the functional activity of C4BP. In addition, CCP1 was found to be important, as was the length of the linkers between the first four CCPs.

EXPERIMENTAL PROCEDURES
Proteins-Human plasma C4BP (11), C1 (12), C2 (13), C4 (14), and factor I (15) were purified as described in their respective references. C3 and C4b were purchased from Advanced Research Technologies. C1 and C2 were only functionally pure, i.e. they were devoid of other complement factors. C4BP, C4, C4b, C3, and factor I were at least 95% pure, as judged by Coomassie staining of proteins separated by polyacrylamide gel electrophoresis (PAGE) performed in the presence of SDS. All proteins were stored at Ϫ80°C. Protein concentrations were determined from absorbance at 280 nm or from amino acid analysis following 24 h hydrolysis in 6 M HCl. C4b and C4BP were labeled with 125 I using the chloramine T method. The specific activity was 0.4 -0.5 MBq/g of protein.
cDNA Clones for Recombinant Proteins-Full-length cDNA encoding human C4BP ␣-chain was cloned to pcDNA3 (Invitrogen), a eucaryotic expression vector. This was used as template for the mutagenesis. To create monomeric, truncated C4BP variants, stop codons were introduced after Trp 492 (CCP1-8), Lys 433 (CCP1-7), Gly 375 (CCP1-6), Glu 313 (CCP1-5), Glu 248 (CCP1-4), and Glu 187 (CCP1-3), using the Quik-Change kit (Stratagene). The recombinant, overlapping extension po-lymerase chain reaction technique was used to construct mutants lacking individual CCPs. The polymerase chain reaction products were cloned into HindII and NotI sites of pcDNA3. The amino acids deleted in each construct, and the sequences of the primers that were used for the polymerase chain reactions are given in Table I. All mutations were confirmed by automated DNA sequencing (PerkinElmer Life Sciences).
Purification of Recombinant Proteins-Human kidney 293 cells (ATCC number 1573-CRL) were transfected with the various C4BP constructs using Lipofectin, according to the manufacturer's instructions (Life Technologies, Inc.). The neomycin analogue, G418, at a concentration of 400 g/ml was used for selection of transfected cells. Colonies of cells showing the highest expression levels, as judged by immunoblotting of the expression medium, were expanded in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 3.4 mM glutamine, 100 units/ml of penicillin and streptomycin, and 400 g/ml G418. Medium from transfected cells was collected in 3-day periods, during which time, the cells were cultured in serum-free Opti-MEM Glutamax. It was stored at Ϫ20°C until ϳ4 liters of medium was collected. The medium was centrifuged for 30 min at 5000 ϫ g to remove cell debris and applied at a flow of 90 ml/h on an affinity column with mAb 104 (directed against CCP1) or mAb 67 (directed against CCP4) coupled to Affi-Gel 10 (2.6 ϫ 12 cm; Bio-Rad). The column was washed with Tris-buffered saline, 1 M NaCl, and the recombinant protein was subsequently eluted with 3 M guanidinium chloride and immediately dialyzed extensively against Tris-buffered saline. All preparative work was done at 4°C. The exact concentrations of all mutants were determined from the amino acid composition analysis after 24 h hydrolysis in 6 M HCl. In some cases, the proteins separated by SDS-PAGE were transferred into Immobilon membranes and stained with Coomassie, and their N-terminal amino acid sequences were determined using an Applied Biosystems 475A sequencer.
Circular Dichroism-The mutants were dialyzed against 10 mM sodium phosphate (pH 7.4) before analysis. Approximately 120 g of each mutant was analyzed in the far UV region (185-250 nm). The resolution was 1 nm, the sensitivity was 20 millidegrees, the speed was 10 nm/ min, and every spectrum presented is an average of four measurements.
Competition Assay for Binding of C4b-To immobilize C4b, 50 l of C4b (10 g/ml) in 75 mM sodium-carbonate, pH 9.6, were added to microtiter plate wells and left overnight at 4°C. The wells were washed three times with washing buffer (50 mM Tris-HCl, 0.15 M NaCl, 0.1% Tween, pH 7.5) and then incubated at room temperature with 200 l of quench solution (washing buffer supplemented with 3% fish gelatin). After another three washes, 125 I-labeled wild type recombinant C4BP was added (20 kcpm/well), together with the various unlabeled proteins diluted in 50 mM Tris-HCl, 150 mM NaCl, 0.1% Tween 20, 0.1% bovine serum albumin, pH 7.5. The samples were incubated for 4 h at room temperature, the wells were washed five times, and the amount of radioactivity bound in each well measured in a gamma counter.
C4b Degradation Assay-C4BP (200 nM) was mixed with 250 nM C4b, 60 nM factor I, and trace amounts of 125 I-labeled C4b in 50 mM Tris-HCl, 150 mM NaCl, pH 7.4. The samples were incubated for 1.5 h at 37°C, and the reaction was terminated by the addition of SDS-PAGE sample buffer with reducing agent (dithiothreitol). The samples were then incubated at 95°C for 3 min and applied to a 10 -15% gradient SDS-PAGE. The separated proteins were visualized and quantified using a PhosphorImager (Molecular Dynamics).
Prevention of C3-convertase Assembly by C4BP-Sheep erythrocytes were washed twice with DGVB ϩϩ (2.5 mM veronal buffer, pH 7.3, containing 70 mM NaCl, 140 mM glucose, 0.1% gelatin, 1 mM MgCl 2 , and 0.15 mM CaCl 2 ). They were suspended at a concentration of 10 9 cells/ml and incubated for 20 min at 37°C with an equal volume of amboceptor (Boehringverke, diluted 1:3000 in DGVB ϩϩ ). The erythrocytes were then washed twice with ice-cold DGVB ϩϩ and resuspended to a concentration of 10 9 cells/ml. C1 was added dropwise to a final concentration of 4 g/ml, and the mixture was incubated for 20 min at 30°C. The cells were washed twice with ice-cold DGVB ϩϩ buffer and then incubated for 20 min at 30°C with 2 g/ml of C4, which resulted in the generation of EAC14 cells. The EAC14 cells were incubated in DGVB ϩϩ containing C2 alone (approximately 8 g/ml) or together with C4BP mutants. The mixtures were kept at 30°C with constant shaking, and 50-l aliquots were removed at time intervals and added to 100 l of guinea pig serum diluted 1:30 in GVB-EDTA (2.5 mM veronal buffer, pH 7.3, containing 100 mM NaCl, 0.1% gelatin, 40 mM EDTA) in order to measure residual C3-convertase activity. After 60 min of incubation at 37°C, the samples were centrifuged, and lysis of the erythrocytes was determined spectrophotometrically (A 405 nm).
Acceleration of Decay Rate of C3-convertase by C4BP-EAC142 cells were generated by incubating EAC14 cells at 10 9 cells/ml with an equal volume of C2 diluted in DGVB ϩϩ (final C2 concentration, approximately 20 g/ml). The cells were incubated for 5 min (equal to the experimentally determined T max ), centrifuged at 4°C, and resuspended in DGVB ϩϩ (prewarmed to 30°C) that contained various concentrations of the C4BP mutants. The mixtures were kept shaking at 30°C, and 50-l aliquots were removed at time intervals and mixed with 100 l of guinea pig serum diluted 1:30 in GVB-EDTA. After 1 h at 37°C, the samples were centrifuged, and the amount of erythrocyte lysis was determined spectrophotometrically.
Effect of Wild Type and Monomeric ␣-Chain on Decay of C3-convertase-The EAC14 cells that were used in this experiment were prepared essentially as described above but with somewhat different concentrations of proteins. The amboceptor was used at a 1:1000 dilution, and the C1 concentration was 20 g/ml. In an initial experiment, several batches of C3-convertases were prepared with concentrations of C4 ranging from 0.5 to 32 g/ml. The deposited C4b was detected by flow cytometry using a monoclonal antibody against C4c (Quidel). In this experiment, the cells were incubated for 1 h with a 1:300 dilution of the  Trp 492  5Ј-CCC AAG TGT GAG TGG TAG ACC CCC GAA GGC TGT  CCP1-7  Expressed Asn 1 -Lys 433  5Ј-GCC CCT CAA TGT AAA TAG CTG TGC CGG AAA CCA  CCP1-6  Expressed Asn 1 -Gly 375  5Ј-ACA CCA TCA TGT GGA TAG ATT TGC AAT TTT CCT  CCP1-5  Expressed Asn 1 -Glu 313  5Ј-TAC CAA GGA TGT GAG TAG TTA TGT TGC CCT GAA  CCP1-4  Expressed Asn 1 -Glu 248  5Ј-CCT CCT GCT TGT GAG TAG AAT AGT TGT ATT AAT  CCP1-3 Expressed antibody and then washed twice in DGVB ϩϩ . A fluorescein isothiocyanate-conjugated rabbit anti-mouse antibody (Dakoppats) was used during the second incubation, and the cell-associated fluorescence was then detected in a flow cytometer (Becton Dickinson). Based on the results of the C4 titration, two batches of C3-convertases were prepared with 1 or 32 g/ml of C4. EAC142 cells were generated by incubating EAC14 cells at 2 ϫ 10 8 cells/ml with an equal volume of C2 diluted in DGVB ϩϩ (final C2 concentration, approximately 20 g/ml). The cells were incubated for 5 min, centrifuged at 4°C, and resuspended in DGVB ϩϩ containing different concentrations of wild type C4BP or the CCP1-8 mutant. The mixtures were kept shaking at 30°C, and 50-l aliquots were removed after 5 min and either added to 100 l of cold guinea pig serum diluted 1:30 in GVB-EDTA or placed on ice. The aliquots that were mixed with the guinea pig serum were incubated at 37°C for 1 h, and the amount of erythrocyte lysis was determined spectrophotometrically after centrifugation. To the aliquots that were placed on ice, C3 (40 g/ml) was added, and the samples were then incubated for 30 min at 30°C. After washing, a monoclonal anti-C3 antibody (1:500, Dakoppats) was added and allowed to bind to the cells for 30 min at 30°C. After additional washing, the cells were incubated with fluorescein isothiocyanate-conjugated rabbit anti-mouse antibodies and deposited C3 was detected in a flow cytometer. Heparin Affinity of Recombinant C4BP Variants-Approximately 10 g of wild type C4BP or each of the C4BP variants lacking individual CCP domains was applied to a 5-ml heparin-Sepharose column (Hi-Trap, Amersham Pharmacia Biotech), which was equilibrated in 20 mM Tris-HCl, pH 7.4, containing 1 mg/ml bovine serum albumin. The flow rate was 1.5 ml/min. The column was washed with 30 ml of the starting buffer, and the proteins were eluted with a linear gradient of NaCl from 0 to 0.5 M NaCl at a flow of 1.5 ml/min. C4BP in collected fractions (0.2 ml) was detected by enzyme-linked immunosorbent assay. Conductivity of the eluate was continuously monitored (the column was attached to Ä KTA explorer liquid chromatography system (Amersham Pharmacia Biotech)).

Expression and Characterization of Recombinant C4BP
Variants-To define which of the eight CCP domains in the ␣-chain of C4BP are functionally involved in the regulation of C3convertase activity, 19 recombinant C4BP variants were cre-ated. In series A of C4BP variants (Fig. 1), stop codons were introduced after CCP3, CCP4, CCP5, CCP6, CCP7, or CCP8, which resulted in the generation of six truncated, monomeric C4BP ␣-chains. In series B, each of the CCPs was individually deleted, resulting in eight variants. In addition, in one deletion variant, both CCP1 and CCP2 were removed. In series C of C4BP variants, the linkers between CCPs were modified by insertion of two alanine residues. All recombinant C4BP molecules were expressed in stable cell lines and purified from culture media by affinity chromatography employing one of two monoclonal antibodies: mAb 104 directed against CCP1 or mAb 67 recognizing CCP4 of ␣-chain. The purified truncated forms of C4BP (series A) were analyzed by SDS-PAGE and by Western blotting using mAb 104 (Fig. 2). All six C4BP mutants were recognized by mAb 104, which is known to react with the CCP1 domain. The two mutants CCP1-3 and CCP1-4 both migrated as double bands, under reducing as well as under nonreducing conditions. The two bands of each variant were transferred to polyvinylidene difluoride membranes after electrophoresis and subjected to N-terminal amino acid sequencing. Both bands in the two C4BP variants yielded identical amino acid sequences corresponding to the N terminus of mature C4BP. The size heterogeneity was due to different degrees of glycosylation because after N-glycosidase F digestion, both CCP1-3 and CCP1-4 migrated as single, sharp bands (not shown).
The purified recombinant, polymeric forms of C4BP lacking single CCP domains (series B) and mutants having Ala residues introduced between the CCP domains (series C) were analyzed by SDS-PAGE under both reducing and nonreducing conditions (Fig. 3). All of the variants of both series B and C formed polymeric C4BP similarly to wild type C4BP, and the molecular weight differences between the C4BP variants on nonreduced gels were those that could be expected from the introduced mutations. This suggests that the polymerization process was unaffected by the introduced mutations. The expression levels of all of the recombinant C4BP variants were similar to that of wild type C4BP, which is consistent with proper folding during synthesis. The mutants were also probed with a panel of monoclonal antibodies directed against ␣-chain of C4BP (mAbs 67, 70, 92, 96, 102, and 104). The mutants were immobilized in wells of microtiter plate and incubated with increasing concentrations of antibodies. Bound antibodies were detected with a rabbit anti-mouse antibody conjugated with horseradish peroxidase. Approximately 50% binding to immobilized wild type C4BP was observed at 0.4 -1 nM concentration of each antibody (not shown). Similar results were obtained for all variants of C4BP except when the deleted CCP domain contained antibody epitope. We mapped antibody epitopes to CCP1 (mAbs 70, 92, 96, 102, and 104) and CCP4 (mAb 67). In general, the mutations did not affect the recognition of the C4BP by the monoclonal antibodies, suggesting that the overall conformation of the C4BP variants was correct. Circular dichroism analysis of C4BP variants yielded very similar spectra, once again confirming that mutagenesis did not cause folding changes. Results are presented as strength of the signal relative to the lowest point (millidegrees/s) measured (Fig. 4). We also analyzed temperature-induced unfolding of our mutants (by collecting circular dichroism spectra at increasing temperatures). Up to 90°C, we found polymeric wild type C4BP to be resistant to unfolding (not shown). All polymeric C4BP variants shared this temperature resistance. The monomeric CCP1-8 demonstrated changes in ellipticity at 230 nm at ϳ70°C. These changes were reversible as the normal spectra returned after overnight incubation of the protein at 4°C.
Monomeric C4BP␣-Chains Equally Active as Polymeric C4BP in C4b Degradation of Fluid Phase-It has been shown that each C4BP can bind four C4b molecules strongly and two additional C4b more weakly (16). To investigate whether isolated, monomeric ␣-chains were fully active in regulation of fluid phase C3 convertase, we compared the CCP1-8 variant with wild type C4BP. The abilities of the two proteins to inhibit the formation of C3 convertase, to accelerate its decay, and to act as cofactors to factor I were investigated. In a fluid phase C4b degradation assay, increasing concentrations of C4BP or CCP1-8 were incubated with C4b, factor I, and trace amounts of 125 I-labeled C4b and then subjected to SDS-PAGE and autoradiography. To estimate the cofactor activity, the C4d bands were quantified by densitometry (Molecular Dynamics), and the C4d formation was plotted against the concentration of added protein (Fig. 5A). Both C4BP and CCP1-8 were active as factor I cofactors; ϳ100 nM C4BP was needed to reach 50% of C4d release, whereas 350 nM CCP1-8 was required to obtain a similar effect. The 3-4 times higher CCP1-8 concentration needed agrees well with the presence of four strong C4b binding sites in polymeric C4BP, whereas CCP1-8 only binds one C4b. The results suggest that CCP1-8 is equally efficient as intact C4BP in the degradation of fluid phase C4b, implying that there is no functional cooperativity between the ␣-chains of polymeric C4BP during C4b degradation in fluid phase. Polymeric C4BP Is More Efficient Than CCP1-8 in Inhibition of Cell-bound C3-convertase-The efficiency of monomeric CCP1-8 to inhibit surface bound C3-convertases was compared with that of polymeric C4BP, thus testing the hypothesis that the polymeric nature of C4BP is important for regulation of surface-bound C3-convertases, in particular at high density of C3-convertases. In initial experiments, several batches of EAC14 cells were prepared using C4 concentrations ranging from 0.5 to 32 g/ml. Deposited C4b was detected with fluorescein isothiocyanate-labeled monoclonal antibody against C4b using flow cytometry. At the highest concentration of C4 (32 g/ml), the density of deposited C4b was at least 8-fold higher than that obtained using 1 g/ml of C4 (Fig. 5B). EAC14 cells prepared at 1 g/ml C4 and at 32 g/ml C4 were used in the functional evaluation of C4BP and CCP1-8. The decay of C3convertases was measured by two means, i.e. by the classical hemolytic assay (Fig. 5C) and by measuring C3b deposition on the cell surface using flow cytometry and fluorescein isothiocyanate-labeled C3-antibody (Fig. 5D). In both assays, polymeric C4BP was much more effective than CCP1-8 independently of the convertase density. Moreover, it was observed that lower concentrations of the C4BP were needed to reach similar relative inhibition at high density of C3-convertases as compared with low density, suggesting that the inhibition efficiency of polymeric C4BP increases with increased density of C3-convertases.
Binding Site for C4b on C4BP Involving CCP1-3-To evaluate the ability of all C4BP variants to bind C4b, a competition assay was used in which the various C4BP variants were allowed to compete with 125 I-labeled wild type C4BP for binding to immobilized C4b (Fig. 6). In the absence of competitor, 25-35% of the added 125 I-labeled C4BP bound to the immobilized C4b, binding that was competed out by unlabeled C4BP (half-maximal inhibition observed at 5 nM C4BP). The ability of C4BP variants lacking CCP1, CCP2, CCP3, or CCP1-2 to compete was severely impaired, indicating that CCP1-3 is crucial for C4b binding. In particular, CCP2 was found to be important because deletion of this CCP completely abolished the C4b binding activity. In addition, the spacing between the CCPs was found to be important for the integrity of the C4b binding site, as illustrated by the disruptive consequence of alanine insertions between CCPs 1 and 2, CCPs 2 and 3, and CCPs 3 and 4. Deletion of CCP7 also resulted in 10-fold decreased ability to compete with C4BP binding to C4b.
Factor I Cofactor Activity of C4BP Depending on CCP1-3-To elucidate whether the impaired C4b binding was matched by a decrease in factor I cofactor activity, the C4BP variants were tested in the C4b degradation assay (Fig. 7). During degradation of C4b, factor I cleaves two peptide bonds in C4b. Cleavage of Arg 1217 -Asn 1318 gives rise to a small, 13-kDa peptide named ␣4. The second cleavage at Arg 937 -Asn 938 results in release of the 45-kDa C4d fragment, which was quantified in the assay. The first three CCPs were found to be important for expression of factor I cofactor activity, and even the shortest monomeric form of C4BP (CCP1-3) was active. When ⌬CCP1 was analyzed, the amount of generated C4d was decreased by 60% as compared with the wild type C4BP. However, it was noteworthy that most of the ␣-chain of C4b was gone after the incubation, suggesting that ⌬CCP1 retains its ability to serve as cofactor for the cleavage of Arg 1217 -Asn 1318 bond. Cleavage of this site releases the small ␣4 fragment, and the remainder of the ␣-chain co-migrates FIG. 5. Functional activity of polymeric wild type C4BP and monomeric ␣-chains. A, C4b degradation assay. Samples containing 250 nM C4b, trace amounts of 125 I-labeled C4b, 60 nM factor I, and 200 nM wild type C4BP or CCP1-8 were incubated at 37°C for 2 h. They were then mixed with a sample buffer containing reducing agent and heated, and the proteins were separated by 10 -15% SDS-PAGE. The gels were then subjected to autoradiography, and the intensities of bands corresponding to C4d were estimated by densitometry. B, deposition of C4b. EAC1 cells were incubated with various concentrations of C4, and the deposited C4b was detected by flow cytometry after incubation with a monoclonal antibody against C4b. C, decay rate of the C3-convertase measured in hemolytic assay. EAC142 prepared with 1 or 32 g/ml of C4 was incubated in DGVB ϩϩ containing increasing concentrations of an inhibitor. After 5 min of incubation, the remaining activity of C3-convertase was determined by incubation with guinea pig serum and measurement of released hemoglobin from lysed erythrocytes. D, decay rate of the C3-convertase measured by flow cytometry. The experiment was done as in C except that the activity of C3-convertase was measured as deposition of C3b on the cell surface. C3b was detected by flow cytometry after incubation with a specific antibody. 100% of C3 deposition was set in the absence of an inhibitor.
with the ␤-chain. The small ␣4 fragment could be quantified after overexposure of the gels. Almost 80 -90% of ␣4 fragment, as compared with the amount formed using the wild type C4BP, was generated. Thus, deletion of CCP1 results in the specific loss of C4b-cleavage at Arg 937 -Asn 938 . Similar results were obtained using the C4BP variants in which two alanines had been inserted between CCP1 and CCP2 or between CCP3 and CCP4. In contrast, C4BP variants lacking both CCP1 and CCP2 were completely inactive (Table II). Similarly, C4BP variants lacking CCP2 and CCP3 were inactive as cofactors to factor I. Complete loss of function was also observed in variants having two alanine residues inserted between CCP2 and CCP3, whereas insertions of the alanines between CPP3 and CCP4 yielded an intermediate effect (60% inhibition). The AACCP4/CCP5 mutant was found to be equally active as the wild type C4BP. The factor I cofactor activity was completely inhibited by mAb 104, a monoclonal antibody that was directed against CCP1.
CCP1-3 Is Sufficient for Prevention of C3-convertase Assembly-C4BP is able to prevent the assembly of the classical C3-convertase. A hemolytic assay was used to elucidate whether disruption of the C4b binding site in C4BP correlated with impaired ability to prevent convertase assembly. EAC14 cells were mixed with C2 alone or together with C4BP and incubated at 30°C. Aliquots were drawn at intervals and mixed with guinea pig serum that was a source of C3 and the terminal complement components. The samples were incubated at 37°C for 1 h, and the degree of erythrocyte lysis was determined. The results were expressed as Z, which represents the number of C142 sites formed and equals the negative natural logarithm of (1 -% lysis). In the absence of C4BP, the convertase was efficiently assembled with maximal activity (T max ) observed after 5 min (this Z value was considered as 100%). Recombinant wild type C4BP (44 nM) efficiently prevented the convertase assembly, decreasing the Z value to 42% (Table III). The monomeric forms of C4BP were tested at 6-fold higher molar concentrations than C4BP yielding similar molar levels of C4b binding sites. All monomeric C4BP variants prevented assembly of the C3-convertase, and CCP1-4 and CCP1-5 were even more active than C4BP (Table III). In contrast, deletion mutants ⌬CCP1, ⌬CCP2, ⌬CCP3, and ⌬CCP1-2, as well as alanine insertion variants AACCP1/2 and AACCP2/3, demonstrated decreased activity, suggesting that CCP1-3 contains the site responsible for inhibition of C3-convertase assembly.  ) were incubated with 250 nM C4b, 60 nM factor I, and trace amounts of 125 I-labeled C4b for 1.5 h at 37°C. Immediately afterward, a sample buffer with reducing agent was added, samples were heated at 95°C, and the proteins were separated by SDS-PAGE (10 -15% gradient gel). The gel was dried and subjected to autoradiography. In one sample, the CCP1-8 mutant was preincubated with an excess of mAb 104. As a control, factor I was omitted in the incubation mixture. B, mutants lacking individual CCP domains.

TABLE II
C4b degradation assay Intensities of bands corresponding to the C4d fragment, released from C4b after cleavage by factor I in the presence of C4BP (Fig. 7), were determined by densitometry and are represented as mean values of three determinations Ϯ S.D.

CCP1-3 and Decay Acceleration-
The hemolytic assay was also used to elucidate whether the structural requirements for the decay accelerating property of C4BP were the same as those for C4b binding. For this purpose, EAC142 cells were generated by incubating EAC14 cells with C2 for 5 min (T max ). The cells were then centrifuged and resuspended in DGVB ϩϩ (control) or DGVB ϩϩ containing C4BP or C4BP variants. The C3-convertase decayed to 50% of its initial activity after ϳ15 min in the presence of buffer alone (not shown). Recombinant wild type C4BP enhanced the rate of decay. Dose-response curves were constructed for wild type C4BP and all of the C4BP variants. To compare the efficacy of the various proteins, the concentrations needed to accelerate 5 min decay down to 60% of control were estimated. We found that all monomeric forms of C4BP, including the CCP1-3 variant, accelerated decay of the classical pathway C3-convertase (Table III). The hypothesis that CCP1-3 is responsible for acceleration of C3-convertase decay was further supported by the observed low or absent activity displayed by ⌬CCP1, ⌬CCP2, ⌬CCP3, ⌬CCP1-2, as well as AACCP1/2 and AACCP2/3.
Heparin Binding Site Confined to CCP1-3-The interaction between C4BP and C4b can be inhibited by heparin, suggesting that the C4b and heparin binding sites overlap (17). To assess the capacity of the C4BP variants reported here to bind to heparin, heparin affinity chromatography was used. Recombinant wild type C4BP eluted from the column as a single peak (93.6%) at 27 mS/cm (Fig. 8). The heparin binding ability of C4BP was compromised by the removal of CCP2 and by insertion of two alanines between CCP1 and CCP2. In contrast to the dramatic effects on C4b binding, deletion of CCP3 and CCP1 had only minor effects on heparin binding, suggesting that CCP2 is the most important for the interaction. DISCUSSION The purpose of this study was to delineate complement regulatory sites within the eight CCP modules of the C4BP ␣-chain. Several different regions of the ␣-chains have been suggested to be important for binding of C4b. In the first report concerning this subject, an N-terminal 48-kDa ␣-chain fragment, formed by chymotrypsin digestion, was found to bind C4b and to express factor I cofactor activity (18,19). This agreed well with electron microscopy images of C4BP-C4b complexes, which demonstrated binding of C4b at the peripheral end of each C4BP tentacle (6,20). Some later reports strengthened the concept that the three most N-terminal CCPs are necessary for binding of C4b (21)(22)(23). We recently demonstrated a cluster of positively charged amino acids at the interface between CCP1 and CCP2 to be crucial for binding of C4b and for the ability of C4BP to regulate the C3 convertase (7,8,10). We have now obtained additional data related to the structure-function relationships of C4BP based on studies of a panel of recombinant C4BP variants. Systematic deletions of CCPs from the C-terminal end of the ␣-chain in combination with several functional assays showed that all recombinant proteins that included CCP1-3 had functional activity. Furthermore, specific deletions of CCP2 and CCP3 entirely abolished the ability of C4BP to bind C4b and to inhibit the C3-convertase. Deletion of CCP1 had a significant effect; however, it did not entirely destroy functional activity of the ␣-chain. We therefore concluded that CCP2 and CCP3 are the most crucial domains for the functional activity of C4BP and that CCP1 also plays an important role. The pertinence of the native configuration of the three N-terminal CCPs was shown by mutants in which the relevant four CCPs were maintained but where the distance between them was increased by the introduction of two additional alanines. Variants having alanines inserted between CCP2 and CCP3 completely lacked biological activity, whereas the activities of the AACCP1/2 and AACCP3/4 variants were significantly impaired. Similar results were obtained in the C4b degradation assay, in which the ability of mutant proteins to act as factor I cofactors was assessed.
Heparin interacts with several complement proteins, but the TABLE III Inhibition of classical pathway C3-convertase by C4BP variants A hemolytic assay was employed; cellular intermediates were prepared using purified C1, C4, and C2; and lytic sites were developed with guinea pig serum. For prevention of assembly, EAC14 were incubated in buffer containing C2 and C4BP. At timed intervals, aliquots were transferred to guinea pig serum diluted in EDTA to measure the activity of formed C3-convertase. Z represents the number of C142 formed. For decay acceleration, EAC142 were incubated in DGVB ϩϩ alone or with C4BP. A dose-response curve was performed with wild type C4BP and each mutant. Shown is the protein concentration required for inhibition of convertase down to 60% of value obtained in the absence of an inhibitor. NA, no activity: less than 20% inhibition was reached at 80 nM protein. Data  Recombinant wild type C4BP or its mutants were applied on 5 ml of heparin-Sepharose and eluted with a gradient of NaCl from 0.0 to 0.5 M; 0.2-ml fractions were collected. C4BP was detected in fractions by an enzymelinked immunosorbent assay, and the conductivity of the eluate was measured continuously. affinity of the binding of heparin to C4BP is higher than the affinities of heparin for other complement proteins (24). We have shown previously that heparin is able to block the interaction between C4b and C4BP and that Arg 39 , Arg 64 , and Arg 66 play key roles in heparin binding (7). The same electropositive cluster of amino acids was also shown to be involved in binding of C4b. Analysis of the homology-based three-dimensional model of the whole ␣-chain of C4BP implied that the electropositive cluster between CCP1 and CCP2 is the only area of the ␣-chain that presents characteristics of a heparin binding site (7). In the present study, we found that the ability of C4BP to bind heparin, assessed by affinity chromatography, is strongly compromised by the removal of CCP2 and by insertion of two alanines between CCP1 and CCP2. Furthermore, the deletion of CCP1 and CCP3 had minor effects on the affinity for heparin. These results suggest that the binding site for heparin is located to CCP1-3, and CCP2 is the most crucial for the interaction. It appears that interaction sites for C4b and for heparin overlap to a large extent. In addition, both interactions are very sensitive to ionic strength and governed by charge-charge interactions.
Most published reports agree with the C4b binding site being located within CCPs 1-3, but other regions of the ␣-chain have been implicated to be involved in C4b binding as well. Thus, proteolytic fragments of C4BP were used to map the C4b binding and factor I cofactor activities to distinct regions spanning CCPs 6 -7 and 3-6, respectively (25). Furthermore, a monoclonal antibody directed against CCP6 was found to block binding of C4b to C4BP (26). These results were difficult to reconcile with the observation that mouse C4BP lacking CCP5-6 binds human C4b (27). Recently, it was shown that a cryptic binding site for C3b is located near the C terminus of the ␣-chain (28). This site was only exposed in recombinant, monomeric, and cell-bound C4BP, which makes its physiological significance unclear. However, such a cryptic C3b binding site may also have the ability to interact with C4b, which could explain some of the disagreement in published reports. Interestingly, we found that C4BP lacking CCP7 displayed 10-fold lower apparent affinity for immobilized C4b in a competition assay. The ⌬CCP7 mutant was also a bad inhibitor of convertase assembly, and it did not accelerate decay of C3-convertase. However, the mutation did not affect the ability of C4BP to serve as a cofactor to factor I.
For the other complement regulators, complement receptor 1, membrane cofactor protein, decay accelerating factor, and factor H, the involvement of individual CCP domains in complement regulatory function and binding of C4b/C3b have been investigated. In the case of decay accelerating factor (four CCPs), it was shown that the classical pathway C3 convertase regulatory function resides within CCP2 and CCP3, whereas regulation of the alternative pathway requires CCP1, CCP2, and CCP3 (29). In membrane cofactor protein (four CCPs), sites for C4b/C3b interaction have been mapped primarily to CCP2, CCP3, and CCP4 (30,31). In factor H (20 CCPs), there are three C3b binding sites localized to CCP1-4 (32-35), CCP12-14, and CCP19 -20 (36). Complement receptor 1 (28 CCPs) is organized into four repeats, each consisting of seven CCP units. Full ligand binding (C4b and C3b) and functional activity require the first four CCPs in each repeat (37)(38)(39)(40). Taken together, our present analysis of C4BP and reports on other complement regulators suggest that a basic C3b/C4b binding unit consists of three or four CCP domains.
C4BP is the only complement regulator that is composed of multiple, identical subunits, and we aimed at elucidating the functional importance of the multimeric structure of C4BP. It has been suggested that the interaction between one binding site on C4BP and a single molecule of C4b on the cell surface may be very weak and that the multiple interactions are needed for efficient inhibition of cell-bound C3-convertases by C4BP (41). Indeed, we found that the polymeric wild type C4BP was much more efficient as inhibitor of C3-convertase than the monomeric ␣-chain. Furthermore, it appears that inhibition efficiency of wild type C4BP increases with the density of C3-convertase. This is most probably due to a fact that densely deposited C4b molecules will allow binding of several ␣-chains of C4BP simultaneously and will result in high affinity for the cell surface.
Our future plans include elucidation of the three-dimensional structure of CCP1-3 of the ␣-chain by NMR spectroscopy. Also, the mutants used in this study can be employed for determination of binding sites for other ligands of C4BP, such as streptococcal M-proteins and B. pertussis. We have recently reported that the binding site for porins of Neisseria gonorrhoeae is localized to CCP1 of C4BP (42). This interaction is very relevant physiologically because it confers serum resistance to N. gonorrhoeae. The structural requirements for the functional activities of C4BP are of interest because C4BP is the major regulator of the classical pathway of complement, and it may have therapeutic potential. It has been shown recently that inhibition of the complement system in animal models of rheumatoid arthritis and reperfusion injury has considerably improved survival of the test animals (43,44).