An Abnormal Fibrinogen Fukuoka II (Gly-B (cid:98) 15 3 Cys) Characterized by Defective Fibrin Lateral Association and Mixed Disulfide Formation*

A dysfibrinogenemia was attributable to a single amino acid substitution from glycine to cysteine at residue 15 of the B (cid:98) chain in a fibrinogen molecule designated as fibrinogen Fukuoka II. The fibrinogen Fukuoka II showed prolonged thrombin and reptilase times and impaired fibrinopeptide B release by throm- bin, resulting in abolition of fibrin monomer repolymerization under physiological conditions. Repolymeriza- tion of the des-(B (cid:98) 1–42)-fibrin monomers, however, was not distinguished from the normal pattern of des-(B (cid:98) 1–42)-fibrin monomers, suggesting that no other abnor- mality existed in fibrinogen Fukuoka II. Although an additional cysteine was substituted at residue 15 of the B (cid:98) chain, fibrinogen Fukuoka II had no free sulfhydryl group within the molecule. Instead, fibrinogen Fukuoka II formed a disulfide bond with cysteine, albumin, another mutated B (cid:98) chain within the same molecule, or intermolecular dimeric fibrinogen Fukuoka II. The mu- tation in fibrinogen Fukuoka II was the same as that in fibrinogen Ise published previously (Yoshida, Japanese boy was referred to us for in- vestigation of hypofibrinogenemia. The propositus and his family had no other clinical symptoms, and consanguinity was not found in his pedigree.

A dysfibrinogenemia was attributable to a single amino acid substitution from glycine to cysteine at residue 15 of the B␤ chain in a fibrinogen molecule designated as fibrinogen Fukuoka II. The fibrinogen Fukuoka II showed prolonged thrombin and reptilase times and impaired fibrinopeptide B release by thrombin, resulting in abolition of fibrin monomer repolymerization under physiological conditions. Repolymerization of the des-(B␤ 1-42)-fibrin monomers, however, was not distinguished from the normal pattern of des-(B␤ 1-42)-fibrin monomers, suggesting that no other abnormality existed in fibrinogen Fukuoka II. Although an additional cysteine was substituted at residue 15 of the B␤ chain, fibrinogen Fukuoka II had no free sulfhydryl group within the molecule. Instead, fibrinogen Fukuoka II formed a disulfide bond with cysteine, albumin, another mutated B␤ chain within the same molecule, or intermolecular dimeric fibrinogen Fukuoka II. The mutation in fibrinogen Fukuoka II was the same as that in fibrinogen Ise published previously (Yoshida, N., Wada, H., Morita, K., Hirata, H., Matsuda, M., Yamazumi, K., Asakura, S., and Shirakawa, S. (1991) Blood 77, 1958 -1963). Fibrinogen Ise, however, has been described as having prolonged thrombin time but normal reptilase time. Reasons for the discrepancy were not clear. Analysis of the B␤ 1-42 fragment showed that fibrinogen was heterogeneous at position 31 of the B␤ chain with respect to proline or hydroxyproline.
Fibrinogen molecules contain two copies each of three different polypeptide chains that are designated as A␣, B␤, and ␥. During clotting, the A␣ chain is cleaved by thrombin at position 16 to produce peptide A (fibrinopeptide A, FPA) 1 and the ␣ chain. Similarly, the B␤ chain is cleaved at position 14 to produce peptide B (fibrinopeptide B, FPB) and the ␤ chain. These cleavages and the resulting dissociation of the small peptides FPA and FPB from the molecule convert fibrinogen to fibrin (1,2). The six peptide chains in fibrinogen are held together by 29 disulfide bonds (1), and interchain disulfide bonds have important roles in the assembly and secretion of fibrinogen molecules during its biosynthesis (3,4). After thrombin cleavage of fibrinogen, many disulfide cross-linked, sixchain fibrin molecules assemble into large polymers. Immediately after assembly these noncovalent intermolecular complexes or polymers (fibrin clots) can be dissociated into the individual fibrin molecules in acetic acid, which are designated as fibrin monomers. Congenital abnormalities of fibrinogen molecules have been reported as dysfibrinogenemia (5,6). The amino-terminal region of the B␤ chain plays a pivotal role in the lateral associations of fibrin monomers (7)(8)(9). Synthetic peptides of the B␤ 15-42 region have inhibitory effects on polymerization of fibrin monomers (10).
Fibrinogen Christchurch II (11), fibrinogen Seattle I (12,13), and fibrinogen IJmuiden (14) have a mutation from Arg to Cys at residue 14 of the B␤ chain. Dysfibrinogenemia with prolonged thrombin and reptilase times has been attributed to the impaired release of FPB from abnormal fibrinogen molecules by thrombin due to this Arg to Cys substitution. Fibrinogen with a deletion of the B␤ 9 -72 region, fibrinogen New York I, also shows the prolonged thrombin and reptilase times (15,16). However, fibrinogen Ise, an abnormal fibrinogen with a substitution of Gly to Cys at residue 15 of the B␤ chain shows a prolonged thrombin time but a normal reptilase time, although the release of FPB is strongly impaired (17). Thus, participation of the FPB release with polymerization of fibrin monomers in detail are still controversial.
In the present communication, we report an abnormal fibrinogen designated as fibrinogen Fukuoka II with a substitution of Gly to Cys at residue 15 of the B␤ chain, which is the same mutation as described for fibrinogen Ise. As shown for fibrinogen Ise (17), the B␤ chain of fibrinogen Fukuoka II was resistant to a low concentration of thrombin and clotting time with thrombin was prolonged. Unlike fibrinogen Ise, however, fibrinogen Fukuoka II showed a prolonged reptilase time and formed a disulfide complex with cysteine or albumin, an intramolecular disulfide bond formation, or an interchain disulfide bond with another fibrinogen Fukuoka II molecule at the mutated Cys residue.
Coagulation Studies-Blood was collected by venipuncture into 1 ⁄10 volume of 3.1% sodium citrate. Platelet-depleted plasma was prepared by centrifugation at 1,500 ϫ g for 10 min at 4°C. Samples were divided into aliquots and stored at Ϫ80°C until use. Fibrinogen concentrations in plasma were determined by the thrombin time method of Clauss (19) and by single radial immunodiffusion using chicken anti-human fibrinogen antibody. Thrombin and reptilase times were determined according to the method of Matsuda et al. (20). Prothrombin time and activated partial thromboplastin time were assayed according to Austen et al. (21).
Purification of Fibrinogen-Fibrinogen was purified from plasma of a propositus and his family and of normal healthy controls as described by Matsuda et al. (20). Purified fibrinogen concentration was determined by an extinction coefficient (E 280 1% ϭ 15.1) (22). The clottabilities of fibrinogen Fukuoka II and of normal fibrinogen as determined by the method of Laki (23) were 95% and 96%, respectively.
Preparation of Chicken Anti-human Fibrinogen Antibody-Purified fibrinogen in Freund's complete adjuvant was injected subcutaneously into a chicken several times. About 2 months after the first injection, yolk IgG was purified from eggs according to the method of Jensenius et al. (24). Monospecific anti-fibrinogen antibody was then obtained using purified fibrinogen that had been coupled to CNBr-activated Sepharose 4B.
Fibrinopeptide Release-Fibrinogen (1 mg/ml) was dissolved in TBS containing 0.05 M Tris-HCl (pH 7.4), 0.1 M NaCl, and 25,000 IU/ml aprotinin and was incubated with bovine thrombin (0.1 unit /ml) or reptilase (2 g/ml) at 37°C for 1 h. In some experiments, fibrinogen was incubated with a high concentration of thrombin (50 units/ml) at 37°C for 6 h followed by incubating at 4°C for another 12 h. The incubated fibrinogen solutions were boiled for 10 min and centrifuged at 10,000 ϫ g for 5 min. Fibrinopeptides (FPA and FPB) released into supernatant were analyzed by HPLC on a reversed-phase column of Cosmosil 5C18-P (4.6 ϫ 250 mm) with a linear gradient of 5-20% CH 3 CN containing 0.025 M ammonium acetate (pH 6.0) at a flow rate of 0.5 ml/min. (25).
Analysis of B␤ 1-42-C. atrox proteinase III digests of fibrinogen were heated at 56°C for 10 min and centrifuged at 10,000 ϫ g for 10 min. An aliquot of B␤ 1-42 fragments was reduced and S-pyridylethylated as described previously (28). Pyridylethylated B␤ 1-42 fragments and the nonreduced fragments were fractionated by HPLC on a reversed-phase column of TSK gel ODS-120T (4.6 ϫ 250 mm) with a linear gradient of 20 -30% CH 3 CN containing 0.1% trifluoroacetic acid. Analysis of B␤ 1-42 fragments was done by SDS-PAGE (7-20% acrylamide) (29) under either reducing or nonreducing conditions. Electron Microscopy-Fibrin clots (2 mg/ml) produced by reptilase (2 g/ml) were dissolved in acetic acid, diluted with 25-fold TBS, and mounted on a glow discharged carbon/formvar-coated grid. After incubation for 15 min at room temperature, the samples were negatively stained with 2% (w/v) uranyl acetate. Electron microscopy was performed using a Joel JEM 2000EX transmission electron microscope (Japan Electron Optics Laboratory Co. Ltd., Tokyo, Japan). Fibrinogen Fukuoka II and normal fibrinogen were diluted to 20 g/ml in TBS containing 50% glycerol, sprayed onto mica discs, and rotary shadowed with carbon platinum at an angle of 3 degrees. The rotary shadowed sample was floated on distilled water, loaded on a glow discharged carbon/formvar-coated grid, and examined by transmission electron microscopy.
DNA Amplification and Sequence Analysis-Genomic DNA was isolated from peripheral blood cells according to Hoar (30). Exon II of the fibrinogen B␤ chain was amplified by polymerase chain reaction with a sense primer (5Ј-TGAGGGAATTCGAATAGTTACATTCC-3Ј) and an antisense primer (5Ј-CTTTCTCTAGATGAGAGAGCCACCACT-3Ј) (31). The polymerase chain reaction products were inserted into pUC19 plasmid and transfected into competent cells derived from E. coli strain DH5␣. Plasmids were isolated from the transformed E. coli by the method of Sambrook et al. (32). Following alkaline treatment of plasmids, the inserts were sequenced using the dideoxynucleotide chain termination method (33).
Analytical Procedures-Total protein concentrations were determined by the method of Lowry et al. (34). Immunoblot analysis was performed after blotting SDS-PAGE gels to Immobilon polyvinylidene difluoride membranes using the method of Towbin et al. (35). Free sulfhydryl groups of fibrinogen were determined using Ellman's procedure (36). Amino acid analysis of B␤ 1-42 fragments was performed using a Picotag system (Waters, Millipore Corp.) according to the method of Heinrikson and Meredith (37). After treatment with pyroglutamate aminopeptidase, B␤ 1-42 fragments were sequenced using a gas phase sequenator (Applied Biosystems, model 476A, Foster, CA), and the phenylthiohydantoin-derivatives were identified using an on-line Applied Biosystems 120A phenylthiohydantoin-derivative analyzer.
Case Report-A 1-year-old Japanese boy was referred to us for investigation of hypofibrinogenemia. The propositus and his family had no other clinical symptoms, and consanguinity was not found in his pedigree.

RESULTS
Fibrinopeptide Release-Coagulation studies on plasma and purified fibrinogen from the propositus and his family are shown in Table I. In the propositus and his father, the fibrinogen concentration determined by thrombin time assay was much lower than that determined by the immunological assay. The reptilase as well as thrombin clotting times for plasma and purified fibrinogen were also prolonged with or without calcium ions. After digesting fibrinogen molecules of the propositus and his father with a low concentration of thrombin (0.1 unit/ml) at 37°C for 1 h, the release of FPB from fibrinogen was about 50% of that from normal fibrinogen, whereas the release of FPA from the propositus molecule was normal (Fig. 1a). The retarded release of FPB, however, was restored to almost 80% of normal when the molecule was extensively digested with a high concentration of thrombin; under these conditions fibrinogen molecules were cleaved at multiple additional sites in the A␣ and B␤ chains (Fig. 1b). Essentially the same results were obtained using fibrinogen from the father of the propositus (data not shown). In subsequent experiments, plasma and purified fibrinogen molecules from the father were used unless otherwise noted. The abnormal fibrinogen in this study is designated hereafter as fibrinogen Fukuoka II. Release of FPA by reptilase instead of thrombin was not different between normal fibrinogen and fibrinogen Fukuoka II (Fig. 1c).
Fibrin Monomer Repolymerization-Normal fibrinogen and fibrinogen Fukuoka II were digested with thrombin or reptilase to form fibrin clots, and fibrin monomers were obtained by dissolving the fibrin clots in acetic acid as described under "Experimental Procedures." In contrast to normal fibrin monomers, Fukuoka II fibrin monomers started to repolymerize after a longer lag time with a decreased maximum rate and a lower ultimate turbidity at physiological ionic strength as well as at low ionic strength ( ϭ 0.04 -0.14) (Fig. 2, a and b). When fibrin monomers obtained with reptilase instead of thrombin were compared, Fukuoka II fibrin monomers also showed a slow maximum rate of repolymerization and a decreased ultimate turbidity at all ionic strengths examined ( ϭ 0.04 -0.14) (Fig. 2, c and d). At physiological ionic strength ( ϭ 0.14), Fukuoka II fibrin monomers obtained with either thrombin or reptilase were most extensively impaired (Fig. 2, b and d).
These observations were consistent with the prolonged thrombin and reptilase times of propositus plasma and purified fibrinogen Fukuoka II. C. atrox proteinase III cleaves the B␤ chain of fibrinogen at the carboxyl side of Arg-42 (9). Des-(B␤ 1-42)-fibrinogens prepared from either normal fibrinogen or fibrinogen Fukuoka II Fibrinogen was incubated with a low concentration of thrombin (0.1 units/ml) at 37°C for 1 h (a), with a high concentration of thrombin (50 units/ml) at 37°C for 6 h followed by incubating at 4°C for another 12 h (b), and with reptilase (2 g/ml) at 37°C for 1 h (c). The released FPA and FPB were analyzed by HPLC on a reversed-phase column of Cosmosil 5C18-P (4.6 ϫ 250 mm) with a linear gradient of 5-20% CH 3 CN (dotted line) containing 0.025 M ammonium acetate (pH 6.0), at a flow rate of 0.5 ml/min (25). by pretreating with C. atrox proteinase III were digested with reptilase to form fibrin clots. Repolymerization of normal des-(B␤ 1-42)-fibrin monomers did not occur at physiological ionic strength but did occur at lower ionic strengths, although the ultimate turbidity was lower than when uncleaved fibrin monomers were used (Fig. 2, c and e). Repolymerization of Fukuoka II des-(B␤ 1-42)-fibrin monomers was identical to normal des-(B␤ 1-42)-fibrin monomers (Fig. 2, e and f), suggesting that the susceptible fibrinogen Fukuoka II structural abnormalities are located within the first 42 residues of the B␤ chain.
Fibrin Structure Determined by Electron Microscopy-The ultrastructure of repolymerized fibrin monomers at physiological ionic strength was examined. In repolymerized normal fibrin monomers produced by reptilase, the clot matrix was predominantly composed of thick branching striated fibers with few thin fibrils (Fig. 3a). In contrast to normal fibrin monomers, the clot matrix of fibrin monomers Fukuoka II contained mainly thin fibrils with only a few thick fibers (Fig. 3b).
DNA Sequence Analysis-Using genomic DNA from the propositus as well as the father, we amplified exon II of B␤ chain, which codes for amino acids from B␤ 9 to 72, inserted it into a pUC19 plasmid, and sequenced it. As shown in Fig. 4, a single base substitution was found. This mutation changed the codon GGT (Gly) to TGT (Cys) at the 15th residue of B␤ chain. The normal sequence was also observed, indicating that fibrinogen Fukuoka II is heterozygous.
SDS-PAGE and Immunoblot Analysis of Purified Fibrinogen-Purified fibrinogens from a normal control and the affected individuals were analyzed by SDS-PAGE using nonreducing conditions. Gels were stained with Coomassie Blue (Fig.  5a). Normal fibrinogen showed two bands with molecular masses of 340,000 and 300,000 daltons, corresponding to high molecular mass fibrinogen and low molecular mass fibrinogen (38) (lane 1 in Fig. 5a). Fibrinogen Fukuoka II contained two additional minor bands with molecular masses of 400,000 and 700,000 daltons as well as several faint bands (lanes 2 and 3 in Fig. 5a). After digesting with C. atrox proteinase III, the des-(B␤ 1-42)-fibrinogen Fukuoka II showed only two bands, which migrated slightly faster than the high or low molecular mass fibrinogen (lane 4 in Fig. 5a). Immunoblot analysis using an anti-fibrinogen antibody showed that the additional slow moving bands as well as the major bands were fibrinogen (lanes 2 and 3 in Fig. 5b) and that these slow moving bands disappeared when des-(B␤ 1-42)-fibrinogen Fukuoka II was formed (lane 4 in Fig. 5b). Immunoblot analysis using anti-albumin antibody showed that the band of high molecular mass at 400,000 daltons contained albumin (lanes 2 and 3 in Fig. 5c). The other band of molecular mass at 700,000 daltons was determined to be a dimer of fibrinogen molecules by eluting the band followed by reanalysis using SDS-PAGE after reducing the sample (data not shown). Thus, the replacement of Gly with Cys at residue 15 of the B␤ chain in fibrinogen Fukuoka II results in the formation of albumin-fibrinogen complexes as well as dimeric fibrinogen complexes through a disulfide bond involving the mutated Cys residue.
The albumin-fibrinogen complex was selectively removed from an aliquot of purified fibrinogen Fukuoka II using an immobilized anti-human albumin antibody column. At physiological ionic strength, repolymerization of the fibrinogen Fukuoka II was not improved after removing the albuminfibrinogen complex (data not shown).
Titration of Free Sulfhydryl Groups in Normal Fibrinogen and Fibrinogen Fukuoka II Molecules-The number of free sulfhydryl groups in fibrinogen molecules were examined using Ellman's procedure (36). Less than 0.05 mol of -SH group/mol of fibrinogen (0.05 Ϯ 0.01, mean Ϯ S.D., n ϭ 3) was observed in normal fibrinogen as well as fibrinogen Fukuoka II, indicating that the mutated Cys at residue 15 of the B␤ chain in fibrinogen Fukuoka II is disulfide cross-linked to a sulfhydryl group in other molecules including albumin, cysteine, and fibrinogen Fukuoka II itself (see below).
Analysis of B␤ 1-42-Normal fibrinogen or fibrinogen Fukuoka II molecules were digested with C. atrox proteinase III, and the B␤ 1-42 fragments were separated from the precipitable des-(B␤ 1-42)-fibrinogen molecules by heating at 56°C for 10 min. The supernatant was analyzed using SDS- PAGE followed by staining with Coomassie Blue and immunoblotting using anti-human albumin antibody (Fig. 6). Both undigested and digested samples contained several bands above 67,000 daltons, which must be due to contamination in the purified fibrinogen fractions (lanes 1-4 in Fig. 6, a and b). Residual unprecipitable fibrinogen (lanes 1 and 4 in Fig. 6a) or des-(B␤ 1-42)-fibrinogen molecules (lanes 2 and 3 in Fig. 6a) remained at the top of the gel under nonreducing conditions. When normal fibrinogen was digested with C. atrox proteinase III, the B␤ 1-42 fragment was detected by SDS-PAGE as a single band at 4,500 daltons under either reducing or nonreducing conditions (lane 2 in Fig. 6, a and b). Two new bands of 9,000-and 65,000-dalton molecular mass were observed when fibrinogen Fukuoka II was digested with C. atrox proteinase III under nonreducing conditions (lane 3 in Fig. 6a). The 9,000dalton band disappeared under reducing conditions (lane 3 in Fig. 6b). Immunoblot analysis using anti-human albumin antibody confirmed that the protein at 65,000-dalton molecular mass was albumin (lane 3 in Fig. 6, c and d).
Further analysis of the B␤ 1-42 fragments using either reducing or nonreducing conditions was performed by reversedphase HPLC (TSK gel ODS-120T, 4.6 ϫ 250 mm) using a linear gradient of 20 -30% CH 3 CN containing 0.1% trifluoroacetic acid. Using nonreducing conditions, the normal and mutated B␤ 1-42 fragments were separated into two peaks (B1 and B2) and seven peaks (B3-B9), respectively (Fig. 7a, i and ii). Using reducing conditions, the B␤ 1-42 fragment of fibrinogen Fukuoka II was separated into four peaks (B10 -B13) (Fig. 7a,  iii), whereas the elution pattern of normal B␤ 1-42 fragments were not changed by reduction. The separated peaks were analyzed on SDS-PAGE. B␤ 1-42 fragments in peaks B1 through B6 appeared as a single band on SDS-PAGE at 4,500 daltons under either nonreducing or reducing conditions (Fig.  7, b and c). In contrast, peaks B7, B8, and B9 showed bands at 9,000 daltons under nonreducing conditions that shifted to 4,500 daltons under reducing conditions (Fig. 7, b and c).
Peaks B1-B9 were sequenced on a gas phase sequenator after treating with pyroglutamate aminopeptidase (Table II). These results showed that the sequence of B␤ 1-42 in peak B2 was the normal sequence (1), but the B␤ 1-42 in peak B1 had hydroxyproline instead of proline at residue 31 in the B␤ chain as previously observed by Henschen (39). B␤ 1-42 in peaks B4 and B6 had the same normal sequence as those in peaks B1 and B2, respectively. B␤ 1-42 in peaks B3 and B5 had the Gly to Cys mutation at residue 15 as deduced from the genomic DNA analysis. These results are consistent with the genomic DNA analysis, which shows that the propositus and his father are heterozygous (Fig. 4).
The B␤ 1-42 fragments from peaks B10 and B12 confirmed that they were the mutated fragments, and those from peaks B11 and B13 had the normal sequence. Judging from the peak heights of B10 -B13 in Fig. 7a (iii), we concluded that fibrinogen Fukuoka II was expressed at almost the same level as normal fibrinogen in these heterozygosis patients.
Cysteine was shown to disulfide cross-link to fibrinogen Fukuoka II in an equilimolar ratio. The results of amino acid composition and of sequence analysis of all B␤ 1-42 fragments were consistent with the exception of peaks B3 and B5. Amino acid composition of those two fragments indicated that an additional Cys residue was associated with B␤ 1-42 in peaks B3 and B5 (Table III). The additional Cys was lost after reducing those samples with dithiothreitol (see B3Ј and B5Ј in Table  III). Thus, a cysteine-fibrinogen Fukuoka II complex was formed through a disulfide bond at the mutated Cys of residue 15.
Rotary Shadowing of Fibrinogen Molecules-Fibrinogen molecules were directly observed by electron microscopy. Examination of normal fibrinogen molecules revealed the usual tridomainal structures (Fig. 8a), whereas in addition to the tridomainal structures, extra globular domains situated near the fibrinogen E domain and dimers of the tridomainal structures were observed in fibrinogen Fukuoka II (Fig. 8b). It is reasonable to assume that the extra globular domain structure and the dimer of tridomainal structure correspond to the albumin-fibrinogen complex as reported in fibrinogen Dusart (40) and the dimers of fibrinogen Fukuoka II. Statistic analysis on electron micrographs showed that about 5% (10 out of 200) and FIG. 6. SDS-PAGE and immunoblot analysis of B␤ 1-42 fragments. Normal fibrinogen or fibrinogen Fukuoka II (60 g of protein each) was incubated with or without C. atrox proteinase III (1 g/ml) at 37°C for 2 h. The reaction was stopped by heating at 56°C for 10 min and centrifuged at 10,000 ϫ g for 5 min. B␤ 1-42 fragments in the supernatant were analyzed by SDS-PAGE (7-20% acrylamide) under nonreducing (a and c) and reducing conditions (b and d). a and b, Coomassie Blue staining. c and d, immunoblot analysis with antihuman albumin IgG. Lanes 1, normal fibrinogen without C. atrox proteinase III treatment; lanes 2, B␤ 1-42 from normal fibrinogen; lanes 3, B␤ 1-42 from fibrinogen Fukuoka II; lanes 4, fibrinogen Fukuoka II without C. atrox proteinase III treatment. HPLC fractions (b and c). a, B␤ 1-42 from normal fibrinogen (i), B␤ 1-42 from fibrinogen Fukuoka II (ii), and reduced S-pyridylethylated B␤ 1-42 from fibrinogen Fukuoka II (iii) were analyzed on a TSK gel ODS-120T reversed-phase column (4.6 ϫ 250 mm) with a linear gradient of 20 -30% CH 3 CN (dotted line) containing 0.01% trifluoroacetic acid. B1-B9 were analyzed by SDS-PAGE (7-20% acrylamide) under nonreducing (b) and reducing conditions (c). 2% (4 out of 200) of fibrinogen Fukuoka II molecules formed albumin-fibrinogen complexes and intermolecular dimeric fibrinogen complexes, respectively. This proportion was consistent with the density of bands on SDS-PAGE (Fig. 5). DISCUSSION A congenital heterozygous abnormal fibrinogen designated as fibrinogen Fukuoka II is characterized by prolonged thrombin and reptilase times (Table I), retarded release of FPB, normal release of FPA (Fig. 1), and disrupted repolymerization of fibrin monomers (Figs. 2 and 3). Genomic DNA analysis on exon II of B␤ chain indicated that fibrinogen Fukuoka II had a single base substitution in the codon normally coding for Gly at residue 15 of the B␤ chain. This substitution changed the codon from GGT (for Gly) to TGT (for Cys) (Fig. 4), and amino acid sequencing of B␤ 1-42 peaks confirmed the mutation (Table  II). Normal repolymerization of des-(B␤ 1-42)-fibrin monomers of fibrinogen Fukuoka II was recovered (Fig. 2, e and f), indicating that the abnormal repolymerization behavior of intact fibrinogen Fukuoka II is due to the mutation at residue 15 of the B␤ chain.

FIG. 7. HPLC elution profiles of B␤ 1-42 fragments (a) and SDS-PAGE analysis of the
Titration of sulfhydryl groups showed that fibrinogen Fukuoka II contained no detectable free sulfhydryl groups, indicating that the mutated Cys residue forms a disulfide bond with a sulfhydryl group from other compounds. SDS-PAGE and immunoblot analysis showed that this mutated Cys residue could form a disulfide bond with a sulfhydryl group in albumin as shown in other abnormal fibrinogens (14,40). In addition, dimeric fibrinogen Fukuoka II was also formed via a intermolecular disulfide bond (Fig. 5). Rotary shadowing of the affected fibrinogen molecules indicated that 5 and 2% of fibrinogen Fukuoka II fraction formed albumin-fibrinogen complexes and intermolecular dimeric fibrinogen complexes, respectively. This means that approximately 10 and 4% of the mutated fibrinogen forms albumin-fibrinogen and dimeric fibrinogen complexes, because the examined patients were heterozygous for this mutation.
Analysis by HPLC of the B␤ 1-42 peptides from fibrinogen Fukuoka II produced seven peaks (B3-B9), whereas normal fibrinogen produced only two peaks (B1 and B2). Peak B1 contained hydroxyproline instead of proline at normal residue 31 of the B␤ chain, and peak B2 contained proline at this position (Table II). When multiple individuals were examined, the ratio of peak B1 to peak B2 ranged from 0.2 to 0.7. B␤ 1-42 in peaks B7 and B9 were homodimers of the mutated B␤ 1-42 fragment that contained hydroxyproline and proline at position 31, respectively (Table II). Peak B8 was a heterodimer of B␤ 1-42 that contained hydroxyproline or proline. These data show that normal fibrinogen molecules have a random distri- bution of proline or hydroxyproline at position 31 of the B␤ chain. The functional significance of this heterogeneity is not currently understood. Peaks B3, B5, B7, B8, and B9 contained the Gly to Cys mutation at position 15 of the B␤ chain (Table II). B␤ 1-42 in peaks B3 and B5 were disulfide cross-linked to cysteine (Table  III). Peaks B7-B9 were B␤ 1-42 dimers cross-linked by disulfide bonds through the mutated Cys (Fig. 7, b and c) and indicated that two fibrinogen Fukuoka II molecules were crosslinked to form a 700,000-dalton dimer. A rough estimation from Fig. 7a (ii) shows that the amount of fibrinogen in peaks B3 and B5 was almost equal to that in peaks B7-B9, indicating that the amount of cysteine-fibrinogen complex is similar to the sum of intermolecular cross-linked dimeric fibrinogen molecules and intramolecular cross-linked fibrinogen molecules. Under reducing conditions, all peaks of B3, B5, and B7-B9 were reduced into two peaks B10 and B12 (Fig. 7a, iii). The peak heights of B10 and B12 were almost equal to the heights of peaks B11 and B13, indicating that normal fibrinogen and fibrinogen Fukuoka II are equally expressed in the heterozygous patients. Taken together with statistic analysis on rotary shadowing electron micrograph, we estimated that the proportions of the cysteine-fibrinogen Fukuoka II complex, the in-tramolecular cross-linked fibrinogen Fukuoka II, the albuminfibrinogen Fukuoka II complex, and the intermolecular crosslinked dimeric fibrinogen were approximately 45, 40, 10, and 5%, respectively. In fibrinogen Fukuoka II, almost half of the mutated molecules complexed with cysteine.
The same mutation designated as fibrinogen Fukuoka II has been reported as fibrinogen Ise (17). However, fibrinogen Ise showed prolonged thrombin time with normal reptilase time and defective release of FPB by thrombin but normal repolymerization of fibrin monomers pretreated with reptilase. Contrary to fibrinogen Ise, no repolymerization occurred from des-(A)-fibrin monomer Fukuoka II at physiological ionic strength (Figs. 2 and 3). Despite the same mutation from Gly to Cys at residue 15 of the B␤ chain, analytical data on the fibrinogen Ise and Fukuoka II mutants were not consistent with each other. The most important differences are the effects of the mutated B␤ chain on repolymerization, i.e. repolymerization of fibrinogen Fukuoka II was disrupted as long as the B␤ chain associated with the original fibrinogen molecule. The bulky aminoterminal region of the B␤ chain sterically disturbed the repolymerization of fibrinogen Fukuoka II but did not affect the structure of the A␣ chain, because the release of FPA by thrombin or reptilase was not depressed (Fig. 1c) as in the case of fibrinogen Ise. Steric hindrance of fibrin monomer repolymerization by the amino-terminal region of mutated B␤ chain has been reported. Abnormal fibrinogens with a substitution of Arg to Cys at position 14 of the B␤ chain (11,13,14) show prolonged reptilase time and impaired repolymerization of des-(A)-fibrin monomers as described here for fibrinogen Fukuoka II, but these properties were not observed for fibrinogen Ise (17). Another difference between fibrinogen Ise and fibrinogen Fukuoka II is that the mutated Cys in fibrinogen has been reported in the free sulfhydryl form (17), whereas our results show that all mutated Cys residues are disulfide-bonded to other molecules. The basis for the discrepancies between fibrinogen Ise and fibrinogen Fukuoka II described above is not clear but may involve additional mutations in the fibrinogen Ise molecule that have not yet been determined.