The amino-terminal module of the C4b-binding protein beta-chain contains the protein S-binding site.

Human C4b-binding protein (C4BP) is composed of multiple α-chains associated with a single β-chain. Each chain is composed of homologous, tandemly arranged repeats of so-called short consensus repeats (SCRs). We have previously shown that the three SCR modules of the β-chain contain a high affinity binding site for anticoagulant vitamin K-dependent protein S. On the basis of experiments using synthetic peptides, residues 31-45 of the amino-terminal SCR (SCR-1) in the β-chain were suggested to be involved in protein S binding, but it is not known whether SCR-1 contains the entire protein S-binding site. To address this question, two different truncated forms of the β-chain (β1,2 and β2,3) were expressed in a prokaryotic expression system. The β1,2 construct (SCR-1 + SCR-2) contained the high affinity binding site for protein S in contrast to β2,3 (SCR-2 + SCR-3), which did not bind protein S. Unfortunately, it was not possible to express SCR-1 alone in this system. To further elucidate whether the protein S-binding site is fully contained in SCR-1 or whether SCR-2 is also required, recombinant α/β-chain chimeras were constructed. These chimeras were composed of α-chains with one, two, or three of the amino-terminal SCR modules replaced by the β-chain counterpart and were expressed in a eukaryotic expression system. All recombinant variants were retained within the cells and could be extracted in biologically active forms. The three α/β-chain chimeras bound protein S equally well, with a Ka of ∼2.3 × 108 ± 0.2 M−1 as compared with 2.1 × 108 ± 0.3 M−1 for plasma-purified C4BP. These results show that the entire protein S-binding site on C4BP is contained within β-chain SCR-1.

C4b-binding protein (C4BP) 1 is a high molecular weight glycoprotein that regulates both the complement system, by binding to complement component C4b, and the protein C anticoagulant system, through its interaction with protein S (1)(2)(3)(4)(5). Protein S is an anticoagulant vitamin K-dependent protein that functions as a cofactor to activated protein C in its degradation of coagulation factor Va and factor VIIIa (6). Approximately 70% of protein S in human plasma circulates in complex with C4BP. Only free protein S acts as an activated protein C cofactor, and deficiency of free protein S is associated with increased risk of thromboembolic disease (7). C4BP is composed of two kinds of subunits, the ␣and ␤-chains (Fig. 1A). Most of the C4BPs in human plasma contain seven ␣-chains and one ␤-chain, but other isoforms also exist (8). Each chain is composed of internally homologous repeats called short consensus repeats (SCRs), complement control proteins, or Sushi domains. The ␣-chain contains eight SCRs, and the ␤-chain contains three (9,10). Each SCR is composed of ϳ60 amino acids and contains several conserved residues including four cysteines, which form intradomain disulfide bridges (Cys 1 -Cys 3 and Cys 2 -Cys 4 ). In addition, each chain contains a carboxylterminal non-repeat region with two cysteines, which link the chains together through disulfide bridging. SCR modules are found in many complement and non-complement proteins. Some SCRs are involved in ligand binding, while others seem to function merely as spacers. Using a recombinant truncated ␤-chain that was expressed in a prokaryotic system, we previously showed that the protein S-binding site was located within the three SCR modules of the ␤-chain (11). A peptide comprising amino acids 31-45 of the C4BP ␤-chain (located in SCR-1) has been reported to inhibit the binding of protein S to C4BP (12). This peptide also bound directly to protein S with a K D that was ϳ60 times weaker than that of plasma C4BP-protein S binding (13). Furthermore, antibodies against this peptide inhibited the binding of protein S to C4BP. These results suggest that amino acids 31-45 contain an important part of the binding site for protein S, but it is not known whether this region of SCR-1 constitutes the entire binding site or whether other parts of the ␤-chain are involved in forming the binding site.
To further characterize the C4BP-protein S interaction, two different truncated forms of the ␤-chain were expressed in a prokaryotic expression system and tested for their protein S binding capacity. A recombinant protein composed of SCR-1 ϩ SCR-2 bound protein S equally well as plasma-purified C4BP, whereas a construct containing SCR-2 ϩ SCR-3 did not bind protein S. However, it was not possible to further localize the protein S-binding site in this system. We were unable to express SCR-1 by itself, and due to folding problems, the system was unsuitable for site-directed mutagenesis. Attempts to express truncated or intact ␤-chain in various eukaryotic expression systems were also unsuccessful. 2 Instead, recombinant C4BP cDNA chimeras composed of C4BP ␣-chains with SCR-1, SCR-1 ϩ SCR-2, or SCR-1 ϩ SCR-2 ϩ SCR-3 replaced by corresponding SCRs from the ␤-chain were constructed and expressed in eukaryotic cells. Protein S binding studies using these recombinant proteins showed that ␤-chain SCR-1 contains the entire protein S-binding site and that neither ␤-chain SCR-2 nor SCR-3 contributes to the high affinity binding between C4BP and protein S.

Proteins
C4BP (14) and protein S (15,16) were purified from human plasma. Concentrations of purified proteins were estimated by measuring absorbance at 280 nm. The extinction coefficients (⑀ 1 cm 1% ) used were 14.1 for intact C4BP (17) and 9.5 for protein S (18), whereas we chose to use 10 as the extinction coefficient for the recombinant proteins that were expressed in the prokaryotic expression system. The monoclonal antibody HPC4 was a kind gift from Dr. C. T. Esmon (Oklahoma Medical Research Foundation). The polyclonal antibodies against intact C4BP were prepared and characterized as described previously (19). Monoclonal antibody 36 (mAb 36) was raised against the three SCRs of the C4BP ␤-chain, and its epitope has been shown to be located in the third SCR of the ␤-chain, 2 whereas mAb 67 and mAb 96 recognized the C4BP ␣-chain. Protein S and C4BP were radiolabeled with 125 I (Amersham Corp.) using IODO-BEAD TM (Pierce) and purified as described previously (19).
Eukaryotic Expression-The cDNA chimeras were constructed by overlapping PCR using full-length ␣and ␤-chain cDNAs cloned into the pBluescript vector as templates (9,10,22). Eight different ␣/␤oligonucleotides were synthesized and used in the amplification steps: 1, GCTGTTCTTGGCAATTGTCCAGAGCTTCTT; 2, TCCTGGGT-GTCTGCAGTGGCCCAAGCGGCA; 3, TGGAGGAGGCTTACAGTC-CCTACTTTTGCA; 4, ACAGGTGATTTTTTCGCAGACTGGAAGT-GC; 5, AGGAAGCTCTGGACAATTGCCAAGAACAGC; 6, TGC-CGCTTGGGCCACTGCAGACACCCAGGA; 7, TGCAAAAGTAGG-GACTGTAAGCCTCCTCCA; and 8, GCACTTCCAGTCTGCGAAAA-AATCACCTGT ( Fig. 2A). Boldface letters represent ␤-chain oligonucleotides, with the remainder being ␣-chain oligonucleotides. In addition, universal and reverse primers (Pharmacia Biotech Inc.) were used. The resulting constructs were called ␤1␣, ␤2␣, and ␤3␣. To construct ␤1␣, universal primer and oligonucleotide 5 were used to amplify the ␣-chain leader peptide, and oligonucleotides 1 and 2 were used to amplify ␤-chain SCR-1 (Fig. 2B). These two DNA fragments, containing overlapping tags, were then annealed and amplified using universal primer and oligonucleotide 2. The carboxyl-terminal part of the ␣-chain was amplified using oligonucleotide 6 and reverse primer. Finally, the fragment composed of the ␣-chain leader peptide and ␤-chain SCR-1 was annealed to the amplified carboxyl-terminal part of the ␣-chain, and the resulting construct was amplified using universal and reverse primers. The other two constructs, ␤2␣ and ␤3␣, were prepared in a similar way, but with oligonucleotide 2 being replaced by oligonucleotides 3 and 4, respectively, and oligonucleotide 6 by oligonucleotides 7 and 8, respectively. The chimeric part of the ␤1␣ construct was isolated after HindIII and SnaBI cleavage, ligated to ␣-chain cDNA in the pBluescript vector cut with the same enzymes, and sequenced. Similar procedures were repeated for the ␤2␣ and ␤3␣ constructs, but with SnaBI being replaced by BlnI. Finally, the constructs were cloned into HindIII-and NotI-cleaved pcDNA3 vector (Invitrogen).

Expression and Characterization of Recombinant Proteins
Prokaryotic Expression-The truncated ␤-chains were expressed in the bacterial strain JA221 as described previously (11). Briefly, 1 ml of an overnight culture of the transformed bacteria in LB medium containing 50 g/ml ampicillin was used to inoculate 25 ml of the same medium and grown to A 600 ϭ 0.8. This culture was then grown in 1 liter of LB medium to the same cell density. Protein expression was induced by the addition of isopropyl-1-thio-␤-D-galactopyranoside (final concentration, 1 mM), and incubation was continued for 12 h at room temperature. After centrifugation of the bacteria (10,000 ϫ g, 30 min, room temperature), the periplasmic extract was prepared by hypotonic shock of the bacterial pellet in 100 ml of H 2 O for 30 min at 4°C and centrifugation as described above. The periplasmic extract and the supernatant obtained after the first centrifugation were mixed and supplemented with 20 mM Tris-HCl (pH 7.5), 10 mM benzamidine, 1 mM phenylmethylsulfonyl fluoride, 0.02% NaN 3 , 2 mM CaCl 2 , and 50 mg/ liter soybean trypsin inhibitor. Glutathione (reduced form; 1 mM) was added to facilitate the formation of correctly folded recombinant proteins. The recombinant proteins were purified using monoclonal antibody HPC4 coupled to Affi-Gel as described previously (11).
Pulse-Chase Analysis-A pulse-chase experiment was performed to check for expression and secretion of recombinant ␤1␣. A pulse containing 300 Ci of [ 35 S]Cys and [ 35 S]Met (Amersham Corp.) in Dulbecco's modified Eagle's medium was added to each 100-mm plate containing 90% confluent cells and left for 30 min. The cells were lysed either directly after the pulse or after a 240-min chase with unlabeled complete medium. The medium and the cell lysates were preincubated for 2 h at 4°C with polyclonal rabbit IgG (3 g) and protein A-Sepharose (50 l, 40 mg/ml) (Pharmacia Biotech Inc.) to reduce nonspecific binding. The protein A-Sepharose was then pelleted by centrifugation for 2 min; a polyclonal antibody to C4BP was added; and the mixture was Each oval represents an SCR module. The shaded modules are those that were expressed in the prokaryotic system. incubated overnight at 4°C. The next day, protein A-Sepharose was added at the same concentrations as above, and after another 2-h incubation, the immunocomplexes were pelleted by centrifugation for 2 min. The pellet was washed three times in ice-cold lysis buffer and dissolved in sample buffer. After SDS-PAGE of the samples, the gel was dried, and the labeled proteins were detected using a PhosphorImager TM SI (Molecular Dynamics, Inc.) (24,25).
Characterization of Recombinant Proteins by Monoclonal Antibodies-Microtiter wells were coated with three different monoclonal antibodies (mAb 36, mAb 67, and mAb 96) (50 l, 10 g/ml) in 75 mM sodium carbonate buffer (pH 9.6) (coating buffer) at 4°C overnight. The next day, the wells were washed three times in Tris-buffered saline (50 mM Tris-HCl (pH 7.5), 0.15 M NaCl) containing 0.1% Tween 20 (washing buffer). The plates were then quenched with 1% bovine serum albumin in Tris-buffered saline for 30 min at room temperature; after washing, plasma-purified C4BP and the recombinant proteins in their cell lysate were added at a concentration of 1 g/ml, and the plates were incubated for 2 h at room temperature. After incubation, the plates were washed; incubated with biotinylated mAb 67; washed again; and then developed using streptavidin, biotinylated horseradish peroxidase, and 1,2-phenylendiamine (Dakopatts AB) according to the manufacturer's instructions.

Determination of Concentrations of Recombinant ␣/␤-Chain Chimeras
Microtiter wells were coated with mAb 67 (50 l, 10 g/ml) in coating buffer, washed, and quenched as described above. After washing, a trace amount of 125 I-labeled C4BP was added together with increasing amounts of unlabeled C4BP (0.017-35 nM) or recombinant proteins (cell lysate) in Tris-buffered saline (final volume, 50 l). After overnight incubation, the wells were washed, and the bound radioactivity was measured in a ␥-counter. The concentration of recombinant C4BP was calculated against the standard of plasma-purified C4BP. which were used to amplify the ␤-chain, contained tags complementary to the ␣-chain, whereas oligonucleotides 5-8, which were used to amplify the ␣-chain, contained tags complementary to the ␤-chain. Restriction enzyme sites used in the cloning are also marked. B, flow chart of ␤1␣ construction. Intact ␣and ␤-chains in pBluescript were used as templates in PCR 1, 2, and 4, whereas the products from PCR 1 and 2 were used as templates in PCR 3, and the products from PCR 3 and 4 were used as templates in the final amplification, PCR 5. C, schematic representation of intact ␣and ␤-chains and the three recombinant chains. Each circle represents an SCR module. White circles, ␣-chain; shaded circles, ␤-chain.
Ligand Blotting-After quenching with fish gelatin, the membranes were incubated with 125 I-labeled human protein S in 2 mM CaCl 2 for 120 min. After washing in washing buffer containing 2 mM CaCl 2 for 3 ϫ 10 min, the signals were detected and quantified using a PhosphorImager.

Binding Assays
Direct Binding-Recombinant proteins expressed in the prokaryotic system were tested for direct binding to protein S essentially as described previously (11). ␤1,2, ␤2,3, or plasma-purified C4BP was immobilized in microtiter wells, and the wells were quenched as described above. 125 I-Labeled protein S was then allowed to compete with increasing concentrations of unlabeled protein S (1.9 -120 nM) for binding to the immobilized proteins, and the amount of bound protein S was measured in a ␥-counter. Direct binding of the recombinant ␣/␤-chain chimeras to protein S was measured in the following way. mAb 67 (50 l, 10 g/ml) was immobilized in microtiter wells; the wells were quenched; plasma-purified C4BP or cell lysates containing the various recombinant proteins (1 g/ml; final volume, 50 l) were added; and the wells were incubated for 2 h. The rest of the assay was identical to that described for measuring direct protein S binding to ␤1,2 and ␤2,3.
Competition Assay-The ability of the recombinant proteins to displace protein S binding to immobilized C4BP was tested essentially as described previously (11). Plasma-purified C4BP was immobilized in microtiter wells. After quenching with 1% bovine serum albumin, increasing amounts of plasma-purified C4BP (0.03-400 nM) or of the various recombinant proteins (0.0008 -1000 nM) were added together with a trace amount of 125 I-labeled protein S. After an overnight incubation, the amount of bound protein S was detected as described above.
Protein S Binding to ␤1,2 and ␤2,3-In a ligand blot assay using 125 I-protein S, unreduced ␤1,2 gave a positive binding signal, whereas ␤2,3 did not bind protein S (Fig. 3B). Reduction of recombinant ␤1,2 was followed by a total loss of protein S binding capacity. To compare protein S binding to the recombinant truncated ␤-chain with that to plasma-purified C4BP, two different binding assays were performed. In the first assay, immobilized ␤1,2 bound to protein S with high affinity and with a half-maximum binding of 9 nM (as compared with 5 nM for   4. Protein S binding to ␤1,2 and ␤2,3. A, direct binding assay. Increasing concentrations of protein S were allowed to compete with a trace amount of 125 I-labeled protein S for binding to immobilized plasma-purified C4BP or to the recombinant proteins. Each point represents the mean value from three different experiments. Binding is expressed as percent of the maximum binding observed in each experiment. B, competition assay. Immobilized plasma-purified C4BP was allowed to compete with increasing concentrations of C4BP, ␤1,2, or ␤2,3 in fluid phase for binding to 125 I-labeled protein S. Each point represents the mean value from three different experiments. 100% binding was estimated in the absence of fluid-phase competitor. E, wells coated with plasma-purified C4BP; Ⅺ, wells coated with ␤1,2; f, wells coated with ␤2,3. plasma-purified C4BP), whereas ␤2,3 did not bind protein S (Fig. 4A). In the second assay, ␤1,2 completely inhibited the binding of protein S to C4BP, whereas ␤2,3 did not interfere with the C4BP-protein S interaction (Fig. 4B).
After transfection of the cells, none of the recombinant proteins were detected in the medium. A pulse-chase experiment was performed in which cells transfected with ␤1␣, intact ␣-chain, or intact expression vector (mock-transfected cells) were labeled with [ 35 S]Met and [ 35 S]Cys and then chased with unlabeled medium as described under "Materials and Methods" (Fig. 5). After immunoprecipitation, the recombinant proteins were analyzed by SDS-PAGE, and it was confirmed that the chimeric recombinant proteins were captured inside the cells and not exported to the medium. Directly after the pulse, both ␤1␣ and recombinant intact ␣-chain were found only in the cell lysate. After a 4-h chase experiment with unlabeled medium, a small amount of intact ␣-chain was detected in the cell lysate (most of it was found in the medium), whereas recombinant ␤1␣ was present only in the cell lysate. Like intact ␣-chain, ␤1␣ appeared as two distinct bands on unreduced SDS-PAGE. In both cases, these bands probably correspond to molecules containing seven and eight disulfide-bridged chains. Neither cell lysate nor medium from mock-transfected cells contained any C4BP.
The cell lysates were also subjected to SDS-PAGE and Western blotting and were developed with a polyclonal antibody against the ␣-chain to confirm that the bands detected in the pulse-chase experiment corresponded to recombinant C4BP (Fig. 6). The reduced recombinants were of approximately the same size as the plasma-purified C4BP ␣-chain (70 kDa).
To investigate whether the recombinant proteins were correctly folded, they were tested for binding to three different monoclonal antibodies against C4BP in an enzyme-linked immunosorbent assay system. mAb 67, which recognized all three recombinant proteins, and mAb 96, which recognized only ␤1␣, were both raised against intact C4BP. A third monoclonal antibody, mAb 36, which has earlier been shown to bind to SCR-3 of the ␤-chain, 2 reacted only with the recombinant protein containing this module, i.e. ␤3␣.
To determine the concentrations of ␤1␣, ␤2␣, and ␤3␣ in the cell lysates, 125 I-labeled plasma-purified C4BP was allowed to compete with the recombinant proteins for binding to mAb 67, and the concentrations were calculated using a standard of plasma-purified C4BP. From each 100-mm dish, 1-20 g of recombinant protein was obtained.
Protein S Binding to ␣/␤-Chain Chimeras ␤1␣, ␤2␣, and ␤3␣-The protein S binding capacity of the recombinant ␣/␤chain chimeras was analyzed using a ligand blot assay (Fig. 7). Cell lysates containing ␤1␣, ␤2␣, or ␤3␣ were subjected to SDS-PAGE. The proteins were transferred to a membrane and incubated with 125 I-protein S. Plasma-purified C4BP, purified recombinant C4BP composed of intact ␣-chains, and cell lysate from mock-transfected cells were used as controls. Bound 125 Iprotein S was quantified using a PhosphorImager, and it was found that the signals from the three recombinant ␣/␤-chains were equally strong and approximately four times stronger than the signal derived from plasma-purified C4BP, whereas neither recombinant intact ␣-chain nor cell lysate from mocktransfected cells bound protein S. Reduced proteins did not bind protein S (data not shown).
To confirm that ␤-chain SCR-1 contains the entire protein S-binding site, both a direct binding assay and a competition assay were performed. Protein S binding to immobilized ␤1␣, ␤2␣, and ␤3␣ was saturable and of high affinity. Scatchard analysis of the data showed that the affinity constants for protein S binding to the recombinant chimeras and to plasmapurified C4BP were almost identical (K a ϭ 2.3 ϫ 10 8 Ϯ 0.2 and 2.1 ϫ 10 8 Ϯ 0.3 M Ϫ1 , respectively) ( Fig. 8A) (26). In the competition assay, all three recombinant proteins were equally efficient (and on a molar basis, approximately four times more efficient than plasma-purified C4BP) in displacing binding of protein S to immobilized plasma-purified C4BP, whereas cell lysate from mock-transfected cells did not interfere with the C4BP-protein S interaction (Fig. 8B).

DISCUSSION
Human C4b-binding protein forms a high affinity, noncovalent complex with anticoagulant vitamin K-dependent protein S (3)(4)(5). We previously demonstrated that the protein S-binding site was contained within the three SCR modules of the ␤-chain (11). In peptide binding experiments, amino acids 31-45, located in SCR-1 of the ␤-chain, were found to constitute an important part of the binding site for protein S (12). Homology modeling (based on the NMR structures of SCR-15 and SCR-16 of factor H) suggested that the ␤-chain was mainly covered by a negative contour, but that an area in the carboxyl-terminal region of SCR-1 (including amino acids 31-45) had a positive electrostatic potential (13). A hydrophobic patch that was accessible to the solvent was also found in this region. Several hydrophobic residues are also found in SCR-3, however; and the homology modeling study did not rule out the possibility that some of these residues are in close contact with SCR-1 and thus are able to interact with protein S. In the study presented here, two truncated forms of the ␤-chain (lacking SCR-1 and SCR-3, respectively) were expressed in a prokaryotic expression system, and it was shown that deletion of SCR-3 had no effect on protein S binding, whereas deletion of SCR-1 resulted in a total loss of protein S binding. The binding assays showed that protein S bound to C4BP with somewhat higher affinity than to ␤1,2. However, the protein S binding of ␤1,2 was equally strong compared with that of a recombinant protein composed of all three ␤-chain SCRs described earlier, and the difference compared with plasma-purified C4BP could probably be explained by the fact that not all of the recombinant proteins produced in the prokaryotic expression system were correctly folded (11). Attempts to express SCR-1 by itself in this system or intact or truncated variants of the ␤-chain in various eukaryotic expression systems have been unsuccessful. 2 To investigate whether SCR-1 contained the entire protein S-binding site, overlapping PCR was used to construct recombinant ␣-chains with one, two, or three of the amino-terminal SCRs replaced by the ␤-chain counterpart. Cells transfected with these cDNA constructs were unable to export the recombinant proteins from the cells. A possible explanation for these difficulties in expressing and exporting the ␤-chain is that the ␤-chain contains information important for the regulation of the amount of ␤-chain-containing C4BP that is produced by the cell or that the presence of ␣-chain SCR-1 is a prerequisite for export of the protein. The polymeric form of the recombinant proteins had a molecular weight comparable to that of plasmapurified C4BP, which shows that the polymerization process is independent of the three amino-terminal SCRs of the ␣-chain. The constructs were recognized by three different monoclonal antibodies that bound only to unreduced C4BP, thus strongly suggesting that the recombinant proteins were correctly folded. The recombinant ␣/␤-chain chimeras contained one ␤-chain SCR-1 on each ␣-chain and therefore provided seven or eight protein S-binding sites on each molecule, as compared with the single protein S-binding site on plasma-purified C4BP. This FIG. 7. Ligand blotting. Plasma-purified C4BP, recombinant intact ␣-chain, and cell lysate from the transfected cells were applied to 5% SDS-polyacrylamide gel, transferred to a membrane, and subjected to ligand blotting using 125 I-protein S. Lane 1, plasma-purified C4BP; lane 2, recombinant intact ␣-chain; lane 3, cell lysate from mock-transfected cells; lanes 4 -6, cell lysates from cells transfected with ␤1␣, ␤2␣, and ␤3␣, respectively.
FIG. 8. Protein S binding to ␤1␣, ␤2␣, and ␤3␣. A, results of the direct binding assay performed as described under "Materials and Methods." Each value represents the average of three determinations. The data were plotted according to Scatchard (26). E, plasma-purified C4BP; f, ␤1␣. B/F, concentration of bound protein S divided by the concentration of unbound protein S; B, concentration of bound protein S. B, competition between the recombinant proteins and plasma-purified C4BP for protein S binding. Increasing concentrations of fluid-phase C4BP or fluid-phase recombinant proteins (cell lysates) competed with immobilized C4BP for binding a trace amount of 125 I-labeled protein S. 100% binding was estimated in the absence of fluid-phase competitor. Results are shown of three different binding experiments using plasma-purified C4BP (E), ␤1␣ (f), ␤2␣ (ࡗ), and ␤3␣ (å). was reflected in the binding studies, in which the recombinant proteins bound approximately four times more protein S than did plasma-purified C4BP. The observation that all protein S-binding sites on the chimeric molecules were not available is consistent with results of studies of the C4b-C4BP interaction that showed that, even though each C4BP molecule contains six or seven binding sites for C4b, only four of them are occupied at physiological ionic strength, whereas binding of additional C4b is probably sterically hindered (27). The binding studies performed in this work clearly showed that the aminoterminal SCR of the ␤-chain (SCR-1) contains the entire protein S-binding site. As we have earlier been unable to express the ␤-chain satisfactory, further characterization of the protein S-binding site using site-directed mutagenesis has not been possible. The ␤1␣ chimeric protein that we have described here is a new way of expressing the SCR-1 module from the ␤-chain, which will be useful in determining which amino acids in SCR-1 are involved in the C4BP-protein S interaction.