THE LACK OF HIGH AFFINITY CALCIUM BINDING TO D-DOMAINS AND THE LACK OF PROTECTIVE EFFECT OF CALCIUM ON FIBRINOLYSIS*

Prolonged thrombin time was completely corrected by the addition of millimolar concentrations of calcium in a new abnormal fibrinogen, Osaka V. Analysis of lysyl endopeptidase digests of Aa-, B&, or y-chains by high performance liquid chromatography, and the following amino acid sequence analysis of relevant peptides revealed that about 50% of the y-chain has a replacement of y-arginine 375 by glycine. When fibrinogen was digested with plasmin in the presence of millimolar concentration of calcium, the amount of fragment Dl was about 50% of the normal control, and the rest was further cleaved to fragment Dz, DB, or De2 with an apparent M, of 62,000. Plasmic digestion of cross-linked fibrin in the presence of calcium resulted in the appearance of an abnormal fragment with an apparent M, of 123,000 as well as fragments D2, DS, and D62, concomitant with the decrease of D dimer. The ?-remnant of the abnormal fragment proved to be a cross-linked complex of the normal Dl y-remnant and residues 374-4061411 of the abnormal y-chain. The number of high affinity Ca2+-binding sites for the normal fibrinogen and fibrinogen Osaka V obtained by equilibrium dialysis was 2.88 (about 3) and 1.85, respectively, and that for the abnormal molecules was calculated as 0.9 (about 1) from their relative amounts in the samples, suggesting the lack of two Ca2+-binding sites in the D-domains. These data suggest that the normal structure of the COOH-terminal portion of the y-chain including residue 375 is required for the full expression of high affinity calcium binding to D-domains, the ability to be protected by calcium against plasmic digestion, and fibrin polymerization. During these studies, we found that the NH2-terminal amino acid of the y-remnant in fragments D or D dimer which were obtained after prolonged digestion with plasmin is y-Met*@.

Most of the abnormal fibrinogens have prolonged thrombin-clotting time and impaired polymerization of fibrin monomers, and some of them have been described to have additional abnormalities in fibrinolysis (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12). Fibrinogens Houston (2), Bern I (4), Haifa (8), Kyoto I (lo), and Vlissingen (11) exhibit accelerated fibrinolysis. The y-Asn308 -+ Lys exchange in fibrinogen Kyoto I causes a conformational change in the y-chain, accelerated cleavage of the y Lys356-Ala357 and y -L y~~~' -P h e~~~ bonds by plasmin, and the generation of a new plasmin cleavage site between Lys3OS and Gly309 with the normal calcium binding properties of fragment Dl (10). Fibrinogens Bern I (4), Haifa (8), and Vlissingen (11) have been reported to lack the ability to be protected by calcium against further attack by plasmin (13,14), but the relationship between the lack of protective effect and structural or functional abnormalities has not been fully studied for these abnormal fibrinogens.
In this report we describe a new y-chain variant in an abnormal fibrinogen, Osaka V, with a single-amino acid substitution of glycine for arginine at position 375 which is characterized by lack of high affinity calcium binding to Ddomains, lack of the ability to be protected by calcium against further plasmic digestion, and complete correction of defective fibrinogen clotting by calcium.

EXPERIMENTAL PROCEDURES
Materials-The reversed-phase HPLC' columns were from the following sources: TSK gel TMS-250 from Toyo Soda (Tokyo, Japan) and Cosmosil 5C18-P from Nakarai (Kyoto, Japan). Chelex 100 was obtained from Bio-Rad. Human plasminogen was purified as described (15). Streptokinase was obtained from AB Kabi, and contaminating albumin was removed with Blue Sepharose CL-GB (Pharmacia LKB Biotechnology Inc.). Lysyl endopeptidase and dithioerythritol were obtained from Wako Chemical Co. (Osaka, Japan), and 4vinyl-pyridine was purchased from Aldrich. Pharmalyte was obtained from Pharmacia, and Nonidet P-40 was purchased from Sigma. Affinity purified goat anti-mouse IgG horseradish peroxidase conjugate and peroxidase substrate (4-chloro-1-naphthol) were obtained from Bio-Rad. 45CaC12 (27.68 mCi/mg) and scintillation fluid, Biofluor, were from Du Pont-New England Nuclear. Polyvinylidene difluoride membranes (Clear Blot Membrane-P) were obtained from Atto (Tokyo, Japan).
Coagulation Studies of Plasma and Fibrinogen-Coagulation studies of plasma were performed according to standard procedures as described previously (16). Fibrinogen was purified from citrate-or citrate dextrose-plasma using lysine-Sepharose 4B chromatography, The abbreviations used are: HPLC, high performance liquid chromatography; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; EGTA, [ethylene-bis(oxyethylenenitrilo)]tetraacetic acid.

2753
This is an Open Access article under the CC BY license.

Calcium Binding
to Fibrinogen Osaka V with y -A r p 5 + Gly gelatin-Sepharose 4B chromatography, and fractionation by ammonium sulfates as previously described (16). Purified fibrinogen (10-20 mg/ml) in 0.3 M NaCl was stored at -80 "C until use. Release of fibrinopeptides A and B was examined by HPLC and by the rate of conversion of Aa-and BP-chains to a-and @-chains, respectively, on SDS-PAGE as described (16). Polymerization of preformed fibrin monomer was studied as described by Gralnick et al. (17) using their second method. Fibrin monomer (40 pg) in 30 pl of 20 mM acetic acid was mixed with 0.57 ml of 50 mM Tris-HCI, 71 mM NaCI, pH 7.4, containing various concentration of CaC12 and the absorbance at 350 nm was continuously monitored at room temperature.
?-Chain Digestion with Lysyl Endopeptidase-Reduced and pyridylethylated (18,19) fibrinogen was prepared as follows. Fibrinogen was dialyzed against 0.5 M Tris-HC1, 6 M guanidine HCI, 10 mM EGTA, pH 8.5, at 4 "C for 18 h, treated with dithioerythritol(O.9 mg/ mg fibrinogen), flushed with nitrogen, and incubated at 37 "C for 1 h. The mixture was treated with 4-vinyl-pyridine (3.1 mol/mol dithioerythritol) at room temperature for 2 h, dialyzed against water at 4 "C for 18 h, and further dialyzed against 50 mM Tris-HCI, 4 M urea, pH 9.0. HPLC separation of fibrinogen chains on a TSK gel TMS-250 reversed-phase HPLC column was performed as described (10). Purified y-chain dissolved in 50 mM Tris-HC1, 4 M urea, pH 9.0, was digested with lysyl endopeptidase (1 rg/nmol y-chain) at 37 "C for 18 h, and fractionated on a Cosmosil 5C18-P reversed-phase HPLC column. A 0.09% trifluoroacetic acid (solvent system A) and 0.09% trifluoroacetic acid in acetonitrile (solvent system B) gradient system was used as the eluant; a nonlinear gradient of 0-20% solvent system B for 1 h and 20-50% solvent system B for the following 1 h at a flow rate of 0.5 ml/min was employed, and the column effluent was monitored at 214 nm.
Plasmic Digestion of Fibrinogen and Cross-linked Fibrin-Fibrinogen (2.5 mg/ml) in 50 mM Tris-HC1, 0.135 M NaCI, pH 7.4, was incubated with various concentration of CaC12 or with 10 mM EGTA at 37 "C for 30 min and treated with 0.1 mg/ml human plasminogen and 3000 units/ml streptokinase for 18 h at 37 "C. Cross-linked fibrin was prepared by adding 5 NIH units/ml bovine thrombin and 0.2 units/ml human factor XI11 (20) to 1 mg/ml of fibrinogen in 50 mM Tris-HC1, 0.135 M NaC1, 5 mM CaCI2, pH 7.4, and incubating for 8 h a t 37 "C and for another 12 h at 4 "C. After squeezing out the crosslinked fibrin clots with a bamboo stick and wiping away water, 1 mg of fibrin clot was treated with 0.08 mg of plasminogen and 2400 units of streptokinase in 2 ml of 50 mM Tris-HCI, 0.135 M NaCI, 5 mM CaCI2, pH 7.4, at 37 "C for 18 h and further treated with the same amount of plasminogen and streptokinase for another 18 h. SDS-PAGE, Immunoblotting, Electroblotting, and Isoelectric Focusing-SDS-PAGE was performed as described previously (21) according to the method of Laemmli (22). Transfer of proteins from the polyacrylamide gel onto a nitrocellulose membrane (Western blotting) for immunoblotting was performed as described (10). Monoclonal antibody against the y-chain was used for identifying the ychain in the blots. The antigenic determinant is within the sequence spanning y-MetSs to Lys2** (10). Electroblotting of proteins for sequence analysis was performed according to the method of Matsudaira (23). About 1 nmol of protein was employed for the first SDS-PAGE, and transferred onto a polyvinylidene difluoride membrane. Isoelectric focusing was performed as described (24,25) with minor modifications. 6.5% gels contained 8% (v/v) Pharmalyte for pH 3-10, 3% (v/v) Nonidet P-40 and 8 M urea, and fibrinogen samples were pretreated with 1% SDS, 1% dithiothreitol, 4 M urea, 5% sucrose, and 8% Pharmalyte for 1 h at 37 "C. The anode and cathode electrode solution was 10 mM glutamic acid and 10 mM ethanolamine, respectively. Densitometric scanning of SDS-PAGE gels was done on a CS-930 dual-wavelength TLC scanner (Shimadzu, Kyoto, Japan) at 540 nm.
Amino Acid Sequence Analysis-Amino acid sequence analysis was performed by automated Edman degradation (Applied Biosystems model 470A Protein Sequencer to which a model 120A phenylthiohydantoin analyzer was connected).
Calcium Binding to Fibrinogen-Equilibrium dialysis to characterize the Ca2+ binding properties of fibrinogen was performed essentially according to the method of Van Ruijven-Vermeer et al. (26) as described previously (10). The buffer used (50 mM Tris-HCI, 0.135 M NaCl, pH 7.6) was filtered through a Chelex 100 column, and the dialysis bags were pretreated according to the method of Marguerie et al. (27). Fibrinogen (3 mg/ml) was added with 5 mM EGTA and dialyzed against the above buffer with 3 mM EGTA, followed by extensive dialysis against the above buffer without EGTA at 4 "C. 0.1 ml of EGTA-treated fibrinogen (2.4 mg/ml) was dialyzed against 20 ml of buffer containing Ca2+ at concentrations from 1 to 100 p~ at 25 "C for 48 h. Each vessel contained 2 pCi of 4sCa2+. After dialysis, 50-pl aliquots of the materials inside and outside the dialysis bag were mixed with 20 ml of scintillation fluid and counted in an LSC-700 liquid scintillation system (Aloka, Tokyo, Japan). Scatchard analysis was performed assuming M, 340,000 for fibrinogen.
Case Report-The propositus (F. U.), 44-year-old female, was admitted to a local hospital to undergo a tonsillectomy for chronic tonsillitis. During the routine hematological study, it was discovered that she had hypofibrinogenemia (110 mg/dl) by the thrombin time method, but the immunologic method showed a normal concentration of plasma fibrinogen (400 mg/dl). Other coagulation studies including one-stage prothrombin time, activated partial thromboplastin time, factor XIII, antithrombin 111, plasminogen, as-plasmin inhibitor, fibrin degradation products in serum, platelet count, and bleeding time were all within the normal range. Tonsillectomy was done without any complication. The propositus and her family members had no history of thrombosis or hemorrhage. Her sister whose plasma was the only one available for analyses in her family members had the normal concentration of plasma fibrinogen both by the thrombin time method and by the turbidimetric method.

RESULTS
Effect of Calcium on Fibrinogen Clotting-The plasma fibrinogen concentration of the propositus evaluated by the thrombin time method was lower than that by the turbidimetric or immunologic method as shown in Table I. The thrombin time of the plasma in the absence of calcium was prolonged but that in the presence of calcium was almost normal. These abnormalities were always recognized over 3 years of investigations, which suggests that the propositus has a congenitally abnormal fibrinogen, although no affected family members have been found yet. The purified fibrinogen from the propositus also had a prolonged thrombin or Repti-lase@ time in the absence of calcium but that in the presence of calcium was entirely normal ( Table I). This was verified by calcium-induced correction of the defective polymerization of fibrin monomers. Polymerization of preformed fibrin monomers which were prepared by the addition of thrombin to fibrinogen was defective in the absence of calcium ( Fig. 1, 0 mM), but the increasing concentration of calcium gradually revised the defect, resulting in the entirely normal polymerization of fibrin monomers in the presence of 5 mM CaClz ( Fig.  1, 5 mM). Defective polymerization of fibrin monomers prepared by the addition of Reptilase" was also corrected by calcium (data not shown). Such complete correction of defective fibrinogen clotting by calcium strongly suggested the presence of calcium-related abnormalities in the propositus fibrinogen. Release of fibrinopeptides A and B was normal (data not shown).
Effect of Calcium on Plasmic Digestion of Fibrinogen-No abnormalities were shown by SDS-PAGE of purified propos-  itus fibrinogen before and after reduction (not shown). Plasmic digestion of the normal fibrinogen in the presence of 1 mM CaCl, resulted in the formation of fragments E and Dl with very little breakdown products of fragments Dz, DS, and DGz ( Fig. 2 A , lane 1 ). Fragment DG2 is composed of an ctremnant with an apparent M , of 12,000, a p-remnant with an apparent M , of 40,000, and a y-remnant with an apparent M , of 12,000 (10). In contrast, that of the propositus fibrinogen resulted in the formations of fragments E, a reduced amount of Dl, and considerable amounts of Ds and DG2 (Fig. 2 A , lane 2), concomitant with a reduced amount of the fragment Dl yremnant (lane 6 ) compared to that of the normal control (lane 5). Densitometric scans of SDS-PAGE gels were performed and the amount of propositus fragment Dl was compared to that of the normal control using fragment E as the internal standard. The amount of propositus fragment Dl was 47% of normal, suggesting that the remaining 53% of propositus fibrinogen was further digested by plasmin even in the presence of calcium. These data suggest the presence of abnormal molecules with defective y-chains in the propositus fibrinogen. Such a degradation of fragment Dl (cleavage of the y-L y~" ' -T h r~~~ bond and y-LyS"6-Ala357 bond) (28) was observed in the range of 0.1-10 mM calcium, in which the ratio of smaller fragments (D3 or Ds2) to fragment D2 increased with the reduction of calcium concentration (Fig. 2B, lanes  6-8). In this range of calcium concentrations, fragments Dl and E were the main plasmic digests in the normal fibrinogen ( Fig. 2B, lanes 2-4), but small amounts of fragments D2 and Ds were also observed a t a calcium concentration of 10 PM (Fig. 2B, lane 1 ). Plasmic digestion of fibrinogen in the presence of EGTA showed no substantial difference between the propositus fibrinogen and the normal fibrinogen ( Fig. 2 A ,  lanes 3 , 4 , 7, and 8).
Lysyl Endopeptidase Cleavage of y-Chain-SDS-PAGE of reduced and pyridylethylated fibrinogen and the HPLC separation profile of fibrinogen chains showed no differences between the propositus fibrinogen and normal fibrinogen (not shown). The HPLC elution pattern of the lysyl endopeptidase-cleaved y-chains of propositus fibrinogen (Fig. 3, OV) revealed a decrease in one peak, the one designated as OVN, compared to the corresponding normal peak, N N , and the appearance of an abnormal peak, designated as OVA. The ratio of the area of OVN to that of NN was 0.54:1.00, meaning a decrease of the normal peptide OVN to 54% on the assumption that areas of peptides with the same elution time can be comparable and the amounts applied are strictly equivalent.
The ratio of the area of OVN to that of OVA was 0.480.52. As shown in Table 11, the amino acid sequence of OVN corresponded to residues 374-380 of the normal y-chain, and that of OVA was the same as OVN except for the substitution of glycine for arginine a t residue 375. If the assumption that peptides OVN and OVA have about the same extinction coefficient is allowed, the ratio of the amount of the normal y-chain to the abnormal (mutant) y-chain in the whole propositus fibrinogens will be 0.480.52. We could not find the other abnormal peptide peaks in lysyl endopeptidase digests of propositus y-chain. The HPLC elution pattern of the lysyl endopeptidase digest of the propositus Actor BP-chains was the same as that of the normal control (data not shown). Thus, fibrinogen Osaka V was demonstrated to be from a heterozygous dysfibrinogenemia in which the abnormal ychain, y-Osaka V, has a substitution of glycine for arginine OVN is the peak with the same elution time as N N (the corresponding normal peak); OVA is the new peak.

374-380"
" It was not possible to quantitate recovery. Corresponding residues of the y-chain.
at residue 375, constitutes about 50% of the whole y-chains and has a COOH-terminal region probably not protected by calcium against plasmic digestion. Abnormal Plasmic Digests of Cross-linked Fibrin in the Presence of Calcium-Propositus fibrinogen showed normal cross-linking abilities of its a-and y-chains (data not shown). Cross-linked fibrin was digested with plasmin in the presence of calcium and analyzed on SDS-PAGE. In the normal control (Fig. 4, lane 1  Thr-Gly-Trp-Tyr-Ser (residues 374-378 of ?-Osaka V) and Met-Leu-Glu-Glu-Ile (residues 89-93 of the normal y-chain) were obtained. A small amount of Tyr-Glu-Ala-Ser-Ile (residues 96-100 of the normal y-chain) was also noted. The amino acid sequences of the first 3 residues of the normal y-y dimer remnant (Fig. 4, lane 3 ) analyzed after electroblotting were Met-Leu-Glu (residues 89-91). A small amount of Tyr-Glu-Ala (residues 96-98) was also noted. The sequences of the first 3 residues of the normal Dl y-remnant (Fig. 2 A , lane 5 ) were Met-Leu-Glu (residues 89-91). These data demonstrate that the abnormal y-remnant is a cross-linked complex of the normal Dl-y-remnant (residues 89-411) and residues 374-406/411 of y-Osaka V. Thus, it was found that the COOHterminal region of y-Osaka V in cross-linked form is not protected by calcium against further plasmic digestion, which results in the appearance of abnormally cleaved y-remnant (Abn./y). At the same time, the y-Arg:",'+Gly exchange was confirmed by two different methods. The sequences of the first 3 residues of the y-remnants in fragment Dee or DX9 (10) (see Fig. 2-4, lane 3 ) were also Met-Leu-Glu.
Defective Calcium Binding to Fibrinogen Osaka V-Equilibrium dialysis was performed to examine Ca2+ binding to fibrinogen. A Scatchard analysis of the data (Fig. 5) showed that the number of Ca'+-binding sites for the normal fibrin-ogen and fibrinogen Osaka V is 2.88 (about 3) and 1.85, respectively, with comparable dissociation constants of high affinity binding (3.2 and 3.4 p~, respectively).
As Fibrinogen Osaka V is heterozygous, it is important to know the relative amounts of normal and mutant fibrinogen in the samples to assess the number of calcium-binding sites obtained. Replacement of Arg (basic amino acid) by Gly (neutral) at residue 375 of ?-Osaka V led us to analyze fibrinogen by isoelectric focusing, which was the only method able to differentiate the abnormal y-chain from the whole ychains. The normal y-chain was mainly resolved into three lines (Fig. 6, lanes 2 and 5) as previously reported (25). However, the amounts of these three lines from the propositus y-chains were reduced, and another three lines with increased negative charge were detected as shown by the schematic representation of Fig. 6 (lanes 4 and 6). Although too closely moving lines did not allow us to measure the intensity of each line densitometrically, the relative amounts of normal and mutant y-chains in the propositus y-chains seemed to be almost equivalent (Fig. 6, lanes 1 and 3).
If 52% of the propositus fibrinogen were abnormal molecules (Fig. 3, OVA) and the remaining 48% of normal molecules had the same number of Ca2+-binding sites as the normal control, the number of Ca2+-binding sites of the abnormal molecules is calculated as 0.9 ((1.85 -2.88 X 0.48)/0.52), suggesting that the abnormal molecule in fibrinogen Osaka V lacks 1.98 (about 2) high affinity Ca"-binding sites.

DISCUSSION
Heterozygous abnormal fibrinogen Osaka V is characterized by the correction of defective fibrinogen clotting with physiological concentrations of calcium; lack of protective effect of calcium on fibrinogen or cross-linked fibrin against further plasmic digestion; defective calcium binding to high affinity sites; and a single-amino acid substitution of glycine for arginine at position 7-375. This substitution can arise from a point mutation involving a single nucleotide change in the codon (CGG) responsible for position 7-375 (31): the codon is most likely altered from CGG to GGG. A considerable number of abnormal fibrinogens are known with defective ychains in which the molecular abnormalities have been elucidated, but their abnormalities have been limited to the @ 1 2 3 4 5 6 FIG. 6. Isoelectric focusing of reduced fibrinogen. Lanes 2 and 5, normal control; h e s 1, 3, 4, and 6, fibrinogen Osaka V. yN, normal y-chain bands; y-OV, abnormal y-chain bands in fibrinogen Osaka V. The cathode and anode are indicated. Note that the abnormal y-chain bands are difficult to see because they are too closely located to the normal ones. A schematic diagram (lunes [4][5][6] is also shown. sequence spanning y-ArgiS to Asp33" (10,11,(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44). It has been shown that a fibrin y-chain polymerization site resides in the COOH-terminal portion (45-47) and that the native tertiary y-chain structure is necessary for the expression of the polymerization site (28,48). These reports are supported by the discovery of fibrinogen Osaka V with a single amino acid substitution at position 7-375.
Human fibrinogen has three high affinity calcium-binding sites (49)(50)(51)(52). Two of them are located in the two D-domains (10,47,53), especially in residues 311-336 of the y-chain (521, and the third site is located in the NHderminal disulfide knot (54, 55). The existence of low affinity binding sites in human fibrinogen is also suggested (52,56). The role of calcium bound to each site is not clarified yet. 2.88 (about 3) of Ca"-binding sites for the normal fibrinogen (Fig. 5) is in good agreement with previous reports (49)(50)(51)(52). In intact fibrinogen Osaka V, 1.85 binding sites were obtained. The ratio of the relative amounts of the normal fibrinogen to the mutant fibrinogen in whole fibrinogen Osaka V was about 1:l which was obtained by isoelectric focusing of reduced fibrinogen (Fig. 6) and HPLC of lysyl endopeptidase digests of y-chains (Fig. 3). 0.9 (about 1) of Ca2+-binding site and the lack of 1.98 (about 2) of binding sites calculated for the mutant fibrinogen molecule in fibrinogen Osaka V will probably be presumed to be caused by a lack of high affinity calcium binding to ychains in D-domains. Fibrinogen Osaka V is the first case of an abnormal fibrinogen with defective calcium binding which is not the result of Aa-chain degradation (51). Defective calcium binding caused by a single amino acid substitution has been reported only in mutant chicken skeletal myosin light chain (57).
The mechanism by which calcium ion enhances the polymerization rate of normal fibrin monomers is still unresolved, although potentiative effects of calcium on the binding of NHa-terminal peptides of fibrin a-or &chains to fibrinogen have been advanced as one of the explanations (58). Defective fibrinogen clotting is partially corrected via the binding of physiological concentrations of calcium in most abnormal fibrinogens, but its almost complete correction is very rare and has been observed only in three cases, fibrinogens Baltimore I11 (y-Asn:"" + Ile) (40,59), Milano I (y-AspT3' + Val) (43, 60) and Bern I (4). The mechanism for such a correction despite the presence of abnormal molecules in these three fibrinogens and fibrinogen Osaka V remains unknown. Thrombin-induced aggregation of fibrinogen Bern I was markedly delayed a t 10 p~ calcium, where only the high affinity calcium-binding sites of normal fibrinogen are occupied, but was normal a t 5 mM calcium, where the molecule is saturated with calcium (4), and such was also the case with fibrinogens Milano I (60) and Osaka V (Fig. 1). Importantly, normal aggregation of fibrinogen at millimolar concentration of calcium was not affected by the presence (fibrinogen Baltimore 111) (40) or absence (fibrinogen Osaka V) of high affinity calcium binding to D-domains. Aggregation of normal fibrin was also significantly enhanced by increasing the calcium concentration from 0.1 to 1 mM (61) or from 20 p M to 5 mM (52). These observations suggest that enhancement of fibrin polymerization is due to calcium bound not to the high affinity binding sites in D-domains but to some low affinity binding sites.
It has been shown that calcium has a protective effect in the plasmic digestion of fibrinogen (13), and purified fragment Dl is degraded to fragments Da and Da by plasmin if bound calcium is chelated with EGTA (47, 62). Dang et al. (52) showed that this protective effect is obtained by 20 pM calcium during the digestion of fibrinogen and proposed that this effect is due to calcium bound to the high affinity Ddomain sites. 47% of propositus fibrinogen remained as fragment Dl during plasmic digestion in the presence of calcium, but the remaining 53% was further digested (Fig. 2). If susceptibility of the normal and mutant fibrinogen to plasmin was not modified by each other, 53% of abnormally digested fibrinogen will probably reflect 52% of abnormal molecules with y-Osaka V (Fig. 3, OVA) which lacks high affinity calcium binding to D-domains. The lack of protective effect of calcium on the plasmic degradation of abnormal molecules in fibrinogen Osaka V will thus be caused by lack of high affinity calcium binding to the abnormal D-domains. The lack of protective effect has been reported in fibrinogens Bern I (4), Haifa (8), and Vlissingen (ll), and defective calcium binding is presumed except for fibrinogen Haifa.
On the other hand, calcium bound to low affinity sites also seems to have some effects on the plasmic degradation of fibrinogen. Millimolar concentration of calcium are required for the full expression of the protective effect under the experimental condition of Haverkate and Timan (13). Five mM calcium exerted a partial protective effect on fibrinogen Bern I while 1 mM calcium had no effect at all (4). In fibrinogen Osaka V, the ratio of smaller fragments decreased with increase of calcium concentration in the range of 0.1-10 mM (Fig. 2B). These phenomena will not be explainable without considering the role of calcium bound to low affinity sites.
Special attention should be paid to fragment D2 which has been shown to have one high affinity calcium binding site, although the dissociation constants are somewhat larger than that of fragment Dl (47, 62). If this is the case, fragment D2 produced by plasmic digestion of abnormal fibrinogen which lacks abnormal residues 374-411 of y-Osaka V should regain calcium binding ability. Nieuwenhuizen et al. (62) reported that the conversion of fragment Dz to D3 stops on addition of excess (over EGTA) calcium during the plasmic digestion of fragment Dl; and Purves et al. (14) also showed that the plasmic digestion of fragment D2 as well as D-dimer in the presence of EGTA is inhibited by readdition of excess calcium to a free calcium concentration of 5 mM, although these effects could have been caused by calcium bound to low affinity sites. We tried to purify fragment Dz from the abnormal molecule, but this was left for future analysis as it requires much more purified fibrinogen Osaka V. If fragment Dz of the abnormal molecule which contains y -G l r~~" -M e t~~~ has a high affinity calcium-binding site, abnormal residues 374-411 in y-Osaka V will serve as an inhibitor of calcium binding to intact y-Osaka V or affect conformations required for calcium binding. The conformational change of fibrinogen on the partial reduction of disulfide bonds including that between y -C y~~'~ and 7 -c~~~~' leads to a lack of protective effect of calcium against plasmic digestion (63), but completely reduced and carboxyamidated y-chain binds calcium in the presence of M2+ (52), which suggests that the relationship between calcium binding and protective effect is much more complicated than expected.
The lack of protective effect of calcium on the plasmic degradation of the abnormal molecule in fibrinogen Osaka V led us to investigate the plasmic digestion of cross-linked fibrin in the presence of calcium. If y-y dimers composed of the normal y-chain and y-Osaka V in cross-linked fibrin were digested with plasmin in the presence of calcium, cleavage of the and bonds in the y-Osaka V portion will occur despite cross-linking, and the appearance of abnormally cleaved y-remnant (Abn./y) which is a crosslinked complex of the normal fragment Dl y-remnant and residues 374-4061411 of y-Osaka V should be expected, and, in fact, was obtained (Fig. 4, lunes 4 and 7). This is particularly important because such an abnormal fragment which does not usually exist as a stable end product from cross-linked fibrin (Fig. 5 of Ref. 14) cannot be demonstrated in the plasmic digestion of fibrinogen (fragments D). In addition, this strongly supports the hypothesis that the abnormal 7chain with a single amino acid substitution, y-Osaka V, is not protected by calcium against further plasmic digestion.
The main NHz-terminal amino acid of the normal Dl y-remnant, the normal y-y dimer remnant, the normal Dl y-remnant in the abnormally cleaved y-remnant and the y-remnant of fragment D62 or DS9 is y-Mets9, which is in agreement with the findings of Marder et al. In summary, fibrinogen Osaka V with y-Ar$I5 + Gly provides insights into the tertiary y-chain structure necessary for the expression of calcium-binding sites and the role of calcium bound to high affinity or low affinity binding sites in fibrin polymerization and fibrinolysis.