Barium Citrate Research Articles

Dicoumarol-induced 9-γ-carboxyglutamic acid prothrombin: isolation and comparison with the 6-, 7-, 8-, and l0-γ-carboxyglutamic acid isomers

The role of gamma-carboxyglutamic acid (Gla) in prothrombin function can be effectively evaluated by characterizing dicoumarol-induced, Gla-deficient prothrombin structural isomers. In addition to the isolation of 8-, 7-, 6-, 5-, 3-, 2-, 1-, and 0-Gla isomers, we have now purified a variant prothrombin containing 9(8.80) Gla residues by barium citrate adsorption, elution, and finally by DEAE-cellulose and immunoaffinity chromatographies. Agar gel electrophoretic mobilities of the 9-Gla isomer and its fragment 1 were slower than those of the respective 10-Gla (normal) prothrombin and fragment 1, both in the absence and presence of Ca(II). In the presence of Ca(II), both 9- and 10-Gla fragments 1 moved slower than 8- and 7-Gla fragments 1. However, in the absence of metal ions, 9- and 7-Gla fragments 1 migrated at the same rate, but slower than 10- and 8-Gla fragments. Similarly, the 9-Gla fragment 1 electrofocused cathodically to 10- and 8-Gla, but comparably with 7-Gla fragment 1. The 9-Gla fragment 1 exhibited a Ca(II)-induced 44% decrease in the intrinsic fluorescence, compared with a 40% decrease in that of 10-Gla; 8-Gla fragment 1 revealed only 23% quenching. Ca(II)-dependent anti-normal prothrombin antibodies are not specific for 10-Gla prothrombin, since only a twofold molar excess of the 9-Gla isomer was required to displace equal amounts of labeled normal prothrombin. The most critical Gla residue for influencing the functional, thrombin-generating properties of prothrombin appears to be the one present in the 9-Gla isomer but absent in the 8-Gla variant, since 9-Gla prothrombin possesses four times the normal coagulant activity (78 versus 20%) of the 8-Gla isomer.

Identification of Different Forms of Human C4b-Binding Protein Lacking β-Chain and Protein S Binding Ability

Human C4b-binding protein (C4BP) is a multimeric regulatory component of the complement system that circulates in plasma either as a free protein or in a noncovalent complex with the vitamin K-dependent protein S. The major form of C4BP is composed of seven identical alpha-chains (70 kDa) and one beta-chain (45 kDa). C4BP was purified from human plasma after barium citrate adsorption using anti-C4BP monoclonal antibody affinity chromatography. C4BP-high and low Mr forms were both obtained from the barium citrate precipitate and the supernatant. C4BP-high and low forms from the barium citrate precipitate were separated by sodium dodecylsulfate polyacrylamide slab gel electrophoresis and extracted with Triton X-100. Both forms contained the beta-chain as was demonstrated on sodium dodecylsulfate polyacrylamide slab gel electrophoresis under reduced conditions after silver-staining and with Western-blotting using monoclonal antibodies specific for the beta-chain. The C4BP-high and low forms demonstrated similar protein S binding affinity (KA: 3.18 x 10(8) and 3.21 x 10(8) M-1, respectively) in a C4BP-protein S binding assay and a protein S ligand blot using a peroxidase-conjugated monoclonal anti-protein S antibody. The barium citrate supernatant contained two forms of C4BP-high and one form of C4BP-low. One form of C4BP-high did contain the beta-chain and was capable of protein S binding (KA: 4.35 x 10(8) M-1). The two other forms of C4BP lacked the beta-chain and were unable to bind protein S.

Thrombin infusion in endotoxin-treated rabbits reduces the plasma levels of plasminogen activator inhibitor: evidence for a protein-C- mediated mechanism [see comments

Plasminogen activator inhibitors (PAIs) play a pivotal role in the control of fibrinolysis. The mechanisms regulating the plasma levels of PAI(s) are still unknown. We report here that the infusion of bovine thrombin (1 U/kg/min, over 60 minutes) in rabbits treated with 0.5 microgram/kg endotoxin (to induce an increase in circulating fast- acting PAI) causes a marked reduction of PAI (50% of preinfusion value), as indicated by functional assay and reverse fibrin autography. Moreover, blood clots prepared from samples obtained after thrombin infusion lysed faster than preinfusion clots when exposed, in vitro, to tissue plasminogen activator. Donor-receiver transfusion experiments showed that the half-life of circulating PAI activity was shorter in thrombin-infused rabbits than in controls (4.1 minutes versus 7.4 minutes), suggesting an accelerated clearance. As expected, thrombin infusion resulted also in activation of protein C (PC). The following observations suggest a close relationship between PC activation and PAI reduction. (1) Infusion of thrombin in rabbits made deficient in vitamin K-dependent plasma proteins by warfarin treatment did not result in modification of PAI activity. (2) Treatment of the latter animals with a barium citrate eluate (PE) of rabbit plasma restored both the anticoagulant and profibrinolytic response to thrombin. (3) Short infusion of thrombin-activated PE (containing activated PC, PCa), but not of unactivated PE, induced both anticoagulation and reduction of PAI activity. In vitro, incubation of PAI-rich rabbit serum with thrombin-activated PE and phospholipids resulted in a progressive disappearance of PAI activity with a t1/2 of 30 minutes. However, this slow inactivation rate does not fully explain the results obtained in vivo. Our data suggest that thrombin infusion in rabbits causes a reduction of circulating PAI activity and that activation of PC is the intermediary mechanism involved in this phenomenon.

Thrombin Infusion in Endotoxin-Treated Rabbits Reduces the Plasma Levels of Plasminogen Activator Inhibitor: Evidence for a Protein-C-Mediated Mechanism

Plasminogen activator inhibitors (PAIs) play a pivotal role in the control of fibrinolysis. The mechanisms regulating the plasma levels of PAI(s) are still unknown. We report here that the infusion of bovine thrombin (1 U/kg/min, over 60 minutes) in rabbits treated with 0.5 microgram/kg endotoxin (to induce an increase in circulating fast-acting PAI) causes a marked reduction of PAI (50% of preinfusion value), as indicated by functional assay and reverse fibrin autography. Moreover, blood clots prepared from samples obtained after thrombin infusion lysed faster than preinfusion clots when exposed, in vitro, to tissue plasminogen activator. Donor-receiver transfusion experiments showed that the half-life of circulating PAI activity was shorter in thrombin-infused rabbits than in controls (4.1 minutes versus 7.4 minutes), suggesting an accelerated clearance. As expected, thrombin infusion resulted also in activation of protein C (PC). The following observations suggest a close relationship between PC activation and PAI reduction. (1) Infusion of thrombin in rabbits made deficient in vitamin K-dependent plasma proteins by warfarin treatment did not result in modification of PAI activity. (2) Treatment of the latter animals with a barium citrate eluate (PE) of rabbit plasma restored both the anticoagulant and profibrinolytic response to thrombin. (3) Short infusion of thrombin-activated PE (containing activated PC, PCa), but not of unactivated PE, induced both anticoagulation and reduction of PAI activity. In vitro, incubation of PAI-rich rabbit serum with thrombin-activated PE and phospholipids resulted in a progressive disappearance of PAI activity with a t1/2 of 30 minutes. However, this slow inactivation rate does not fully explain the results obtained in vivo. Our data suggest that thrombin infusion in rabbits causes a reduction of circulating PAI activity and that activation of PC is the intermediary mechanism involved in this phenomenon.

Dicoumarol-induced prothrombins containing 6, 7, and 8 γ-carboxyglutamic acid residues: isolation and characterization

Isolation and characterization of gamma-carboxyglutamic acid (Gla) deficient prothrombins induced by Warfarin or dicoumarol are useful for studying the role of specific Gla residues in prothrombin. In addition to 7-Gla prothrombin, we have isolated two more atypical prothrombins from the barium citrate eluate, one containing 6.11, and the other, 7.85 Gla residues, presumably 6- and 8-Gla prothrombins. The actual Gla content of the 7-Gla isomer was 7.05. Each of the 6-, 7-, and 8-Gla variants showed a single component by agar or dodecyl sulfate polyacrylamide gel electrophoresis. When agar gel electrophoresis was performed in calcium, each of the variants moved more rapidly than normal (10-Gla) prothrombin. In the presence of EDTA, the 8-Gla isomer exhibited the fastest mobility, equivalent to that of normal prothrombin, followed by 7-, and then 6-Gla variants. The physiological activities of the isomers were found to be 18 to 23% for 8-, 6 to 8% for 7-, and 2 to 3% of normal prothrombin for 6-Gla variant. Prothrombin fragment 1, derived from 8-Gla prothrombin, exhibited 23% of calcium-induced fluorescence quenching, compared with 40% for 10-Gla and 8% or less for 7- and 6-Gla fragments 1. Competition radioimmunoassay data show that calcium-dependent anti (normal) prothrombin polyclonal antibodies are not specific for 10-Gla prothrombin, since the 7- and 8-Gla isomers were able to displace radiolabeled (125I) normal prothrombin.

Conformational changes in human prothrombin as detected by antibody populations.

The amino-terminal peptides of human prothrombin corresponding to residues 1-51 and 52-156 have been isolated from a thrombin digest of prothrombin fragment 1. The products of digestion were purified by means of barium citrate and ammonium sulfate precipitations, followed by gel filtration and hydroxyapatite chromatographies. They were identified by their molecular sizes as well as their amino acid compositions. Peptides 1-51 (F1A) and 52-156 (F1B) were used as affinity ligands for the isolation of antibody populations from antisera that were elicited against human prothrombin or prothrombin fragment 1. These antibody populations displayed restricted specificity for the respective ligands as shown by competitive radioimmunoassays. They were used to study the conformational changes in prothrombin and fragment 1. The F1A-specific antibody populations detected a conformational change which is stabilized by calcium ions and which has a transition midpoint at approximately 0.2 mM calcium ion concentration. The F1B-specific antibody populations identified a different conformational change which is destabilized by calcium ions and which has a transition midpoint at approximately 0.5 mM calcium.

A moderate form of hemophilia B is caused by a novel mutation in the protease domain of factor IXVancouver

A genomic phage library was constructed using lymphocyte DNA from a patient with cross-reacting material-positive, moderately severe hemophilia B. The library was screened by using a full-length factor IX cDNA as a hybridization probe. DNA sequence analysis of the factor IX exons and intron/exon junctions revealed a single point mutation at nucleotide 31,311 of the gene. This mutation occurs in the protease domain of factor IXa and changes the codon for isoleucine 397 (ATA) to a threonine codon (ACA). The resulting abnormal protein has been named factor IXVancouver. Factor IXVancouver was isolated from the patient's plasma by barium citrate adsorption, affinity chromatography on a Ca2+-dependent antibody bound to agarose, and anion-exchange chromatography. On gel electrophoresis, the purified protein exhibited a normal molecular weight and a normal pattern of activation cleavages with bovine factor XIa. Kinetic studies on the purified protein indicated that the Km of factor IXaVancouver for human factor X was 3.4 times higher than that of normal factor IXa. The kcat of factor IXaVancouver was 12.5% of the kcat of normal factor IXa. Structural models of the protease domain of human factor IXa and of factor IXaVancouver were constructed, based on the homology of factor IXa with related serine proteases of known structure. The factor IXaVancouver model suggests that hydrogen bonding between the side chain hydroxyl group of threonine 397 and the carbonyl oxygen of tryptophan 385 reduces the ability of factor IXaVancouver to bind factor X in a configuration favoring catalysis.

Monoclonal Antibodies Directed to the Calcium-Free Conformation of Human Protein S

Four mouse hybridomas secreting monoclonal antibodies specific for human protein S (PS) have been generated. The antibodies, all of the IgG1 subclass, were designated S2, S3, S8, and S10. In a fluid phase radioimmunoassay, the binding of monoclonal antibodies to PS was about 30% greater in the presence of EDTA and totally inhibited in presence of Ca2+. Using the same technique, we performed displacement curves of 125I-labeled PS by purified PS, thrombin-cleaved PS, normal plasma, plasma from a patient on warfarin therapy, and plasma from a patient with no free PS and only PS bound to C4b-binding protein. The slopes of the curves show that the monoclonal antibodies reacted equally with all the tested forms of PS indicating that the antigenic site(s) to which the monoclonal antibodies are directed are present and exposed in free and bound PS, in thrombin-cleaved PS, and in the coumarin form of the protein. Each EDTA-dependent antibody, immobilized on Sepharose 4B-CNBr was used to purify PS from the barium citrate-absorbed, ammonium sulphate-soluble fraction of plasma. The fraction eluted from the immunoabsorbent with a buffer containing 4 mmol/l CaCl2 and analysed by SDS-PAGE, contained two bands, one migrating with conventionally purified PS and the other with purified C4b-binding protein. Homogeneous PS was obtained by chromatography of the barium citrate adsorbate on a DEAE-Sephadex column. The protein peak containing the bulk of PS was subsequently applied to the immunoadsorbent and eluted with 4 mmol/l CaCl2.(ABSTRACT TRUNCATED AT 250 WORDS)

Location of .gamma.-carboxyglutamyl residues in partially carboxylated prothrombin preparations

Prothrombin contains 10 gamma-carboxyglutamyl (Gla) residues in the N-terminal (fragment 1) domain of the protein. Following anticoagulant administration, a spectrum of undercarboxylated, physiologically less active forms of prothrombin is secreted into bovine or human plasma. The sites of undercarboxylation in these prothrombin species have now been investigated. Plasma containing a mixture of partially carboxylated forms of prothombin was obtained from a dicoumarol-treated bovine, and three pools of partially carboxylated (four, six, or eight Gla) species were purified by adsorption onto barium citrate and barium oxalate, ammonium sulfate fractionation, and chromatography. Fragment 1 obtained from these variants was equilibrated with 3H2O and heated in a dry state to decarboxylate Gla and incorporate 3H into the resulting Glu residues. This peptide was then sequenced by Edman degradation, and the specific radioactivity of PTH-Glu was determined for each potential Gla-containing site. Data obtained from normal prothrombin fragment 1 fit a linear model when the log of specific activity of PTH-Glu was plotted against the cycle number. Analysis of the 80% variant showed a decrease in carboxylation only in the last two Gla residues, while data obtained from the 60% variant indicated a general decrease in carboxylation from the most amino- to the more carboxyl-terminal Gla residues. In the 40% Gla variant, all but the most amino-terminal of the Gla residues appeared to be undercarboxylated. These data indicate that the gamma-carboxylation of glutamyl residues in prothrombin does not occur randomly but instead with preferential carboxylation of the most amino-terminal Gla residues. When carboxylation is limited, the impairment of carboxylation is more severe at the more carboxyl-terminal residues.

Factor X-activating activity in normal and malignant colorectal tissue

The factor X-activating activity (FXAA) of homogenates from human colorectal tumours and corresponding normal colonic mucosa from the same patients was assessed with a specific chromogenic substrate technique. FXAA was detected in all normal and tumour tissue tested, but was significantly higher in tumour tissue. The procoagulant activity was inhibited by DFP, but was unaffected by iodoacetamide and mercuric chloride. FXAA was largely abolished by prior incubation of both normal and tumour tissue homogenates with a rabbit anti-human factor VII serum, but was greatly enhanced by the addition of purified factor VII. FXAA was partially adsorbed on to aluminium hydroxide and almost completely abolished by treatment with barium citrate. It is concluded that the FXAA of both normal and malignant colorectal tissue is the result of tissue factor-factor VII interaction.

Differential effects of activation of prothrombin by streptokinase compared with urokinase and tissue-type plasminogen activator (t-PA)

We have previously reported paradoxical prothrombotic effects manifest by elevations of fibrinopeptide A (FPA) after administration of streptokinase to patients with acute myocardial infarction. To characterize mechanisms responsible and their dependence on streptokinase (SK) as opposed to other activators of the fibrinolytic system, the present study was performed to compare effects of streptokinase, tissue-type plasminogen activator (t-PA), and urokinase on plasma and on purified prothrombin in concentrations similar to those achieved pharmacologically. The effects of plasmin were assessed to determine the extent to which the elevations of FPA seen could be attributed to activation of plasminogen. Elevations of FPA were observed after incubation of each of the activators with citrated plasma at 37°C for 60 minutes. However, they were most marked with streptokinase (64.5 ± 4.6 pmol FPA/ml (mean ± SE) with 100 IU SK/ml, and 77.6 ± 5.0 pmol/ml with 500 IU SK/ml). Elevations of FPA induced by streptokinase were attenuated by 100 IU/ml heparin [15.2 ± 1.9 pmol/ml after 100 IU of SK (p < 0.001 compared with results with streptokinase without heparin)). Human plasmin, 2.5 CTA/ml, caused changes similar to those induced by streptokinase. The minimal elevations of FPA induced by t-PA or urokinase (< 10 pmol/ml without heparin) were not significantly attenuated by heparin. Incubation of barium citrate adsorbed plasma (vitamin K factor depleted) with streptokinase markedly attenuated elevation of FPA. Addition of prothrombin (1.5 μM) and streptokinase (100 IU/ml) to the barium citrate adsorbed plasma elicited elevations of FPA similar to those induced by streptokinase in citrated plasma. Amidolytic activity with the “thrombin” substrate H-D-phenylalanyl-Lpipecolyl-L-arginine-p-nitroanilide dihydrochloride (S-2238) was evident when streptokinase, plasminogen (0.24 μM), and prothrombin (1.5 μM) were incubated in buffer. Thus, concentrations of streptokinase that are low in terms of therapeutic blood levels activate prothrombin in plasma, likely due to activation of plasminogen. Neither tissue-type plasminogen activator nor urokinase in pharmacologically comparable concentrations increase thrombin activity appreciably perhaps because of less intense activation of plasminogen. Consideration of the prothrombotic effects observed maybe relevant to selection of specific agents for therapeutic thrombolysis, to appropriate titration of dose, and to the need for the use of heparin conjointly with particular activators of the fibrinolytic system such as streptokinase.

Avian thrombocyte thrombospondin

A high molecular weight glycoprotein (450,000) was obtained from thrombin-treated duck thrombocytes by barium citrate adsorption technique followed by heparin-agarose affinity chromatography. Amino acid composition (high number of acidic amino acids and cystine) as well as carbohydrate contents (1.3 per cent hexosamine, 0.9 per cent sialic acid and 1.5 per cent hexose) showed similarity to mammalian platelet thrombospondin.

Propeptide of human protein C is necessary for .gamma.-carboxylation

Protein C is one of a family of vitamin K dependent proteins, including blood coagulation factors and bone proteins, that contains gamma-carboxyglutamic acid. Sequence analysis of the cDNAs for these proteins has revealed the presence of a prepro leader sequence that contains a pre sequence or hydrophobic signal sequence and a propeptide containing a number of highly conserved amino acids. The pre region is removed from the growing polypeptide chain by signal peptidase, while the pro region is subsequently removed from the protein prior to secretion. In the present study, deletion mutants have been constructed in the propeptide region of the cDNA for human protein C, and the cDNAs were then expressed in mammalian cell culture. These deletions included the removal of 4, 9, 12, 15, 16, or 17 amino acids comprising the carboxyl end of the leader sequence of 42 amino acids. The mutant proteins were then examined by Western blotting, barium citrate adsorption and precipitation, amino acid sequence analysis, and biological activity and compared with the native protein present in normal plasma. These experiments have shown that protein C is readily synthesized in mammalian cell cultures, processed, and secreted as a two-chain molecule with biological activity. Furthermore, the pre portion or signal sequence in human protein C is 18 amino acids in length, and the pro portion of the leader sequence is 24 amino acids in length. Also, during biosynthesis and secretion, the amino-terminal region of the propeptide (residues from about -12 through -17) is important for gamma-carboxylation of protein C, while the present data and those of others indicate that the carboxyl-terminal portion of the propeptide (residues -1 through -4) is important for the removal of the pro leader sequence by proteolytic processing.

Human Plasma Protein Z Antigen: Range in Normal Subjects and Effect of Warfarin Therapy

In contrast to the other well-studied vitamin K-dependent proteins that circulate in plasma, protein Z antigen is much more variable. The concentration in plasmas collected in EDTA from 455 normal, healthy donors is normally distributed with a mean of 2.9 micrograms/mL (46 nmol/L) and a SD of 1.0 microgram/mL (95% interval of 32% to 168% of the mean). No significant correlation to age or sex could be detected. In comparison, the concentration of protein C antigen measured with the same type of assay on the same 455 samples has a log normal distribution with a mean of 4.0 micrograms/mL (65 nmol/L) and a 95% interval of 70% to 138% of the mean. Also in marked contrast to other plasma vitamin K-dependent proteins, the total protein Z antigen level is extremely low in patients on stable warfarin therapy (range 1% to 16% of normal). Moreover, even though greater than 95% of the antigen in normal plasmas adsorbs to barium citrate (a crude reflection of the presence of gamma-carboxyglutamic acid (Gla) residues), in the patients taking warfarin almost all of the small amount of the antigen failed to adsorb, suggesting that virtually no protein Z had its full complement of Gla residues. Total protein C antigen in the same 25 patients averaged 53% of normal (34% to 72%) and 54% (average) of the total remaining antigen still adsorbed to barium citrate. The concentration of protein Z antigen in the plasma of a normal individual given a loading dose of warfarin fell at an initial rate of approximately 20% a day, indicating a plasma half-life (t1/2) of 2 to 3 days.

Human plasma protein Z antigen: range in normal subjects and effect of warfarin therapy

In contrast to the other well-studied vitamin K-dependent proteins that circulate in plasma, protein Z antigen is much more variable. The concentration in plasmas collected in EDTA from 455 normal, healthy donors is normally distributed with a mean of 2.9 micrograms/mL (46 nmol/L) and a SD of 1.0 microgram/mL (95% interval of 32% to 168% of the mean). No significant correlation to age or sex could be detected. In comparison, the concentration of protein C antigen measured with the same type of assay on the same 455 samples has a log normal distribution with a mean of 4.0 micrograms/mL (65 nmol/L) and a 95% interval of 70% to 138% of the mean. Also in marked contrast to other plasma vitamin K-dependent proteins, the total protein Z antigen level is extremely low in patients on stable warfarin therapy (range 1% to 16% of normal). Moreover, even though greater than 95% of the antigen in normal plasmas adsorbs to barium citrate (a crude reflection of the presence of gamma-carboxyglutamic acid (Gla) residues), in the patients taking warfarin almost all of the small amount of the antigen failed to adsorb, suggesting that virtually no protein Z had its full complement of Gla residues. Total protein C antigen in the same 25 patients averaged 53% of normal (34% to 72%) and 54% (average) of the total remaining antigen still adsorbed to barium citrate. The concentration of protein Z antigen in the plasma of a normal individual given a loading dose of warfarin fell at an initial rate of approximately 20% a day, indicating a plasma half-life (t1/2) of 2 to 3 days.

A plasma protein C activity assay suitable for a clinical laboratory. Its use to measure activity in hereditary and acquired deficiency states.

The authors describe the technic and standardization of an assay to measure plasma protein C activity feasible for clinical laboratory use. It is modified from an assay of Francis and Patch (Thromb Res 1983; 32:605-613) to enhance protein C recovery in the barium citrate eluate, to eliminate the steps of addition and neutralization of heparin, and to use only commercially available reagents. The normal range for plasma protein C in the assay is 72-130% (+/- 2 SD) (0.7-1.30). Hepatocellular disease lengthening the prothrombin time by 3-4 seconds was associated with plasma protein C activity of 25% (0.25) to 35% (0.35). Although the assay is thought to accurately measure protein C activity in patients taking warfarin, one cannot evaluate such patients for hereditary functional protein C deficiency because treatment with warfarin will itself reduce the ratio of protein C activity to antigen. The assay can be used in a patient receiving heparin if the heparin is removed as described.

Dysfunctional Activated Protein C (PC Cádiz) in a Patient with Thrombotic Disease

The partial characterization of a dysfunctional protein C (PC), provisionally named "PC Cádiz", in a 45-year-old male patient suffering from recurrent venous thrombosis is described. The only defect found in laboratory assays for haemostasis and hepatic function was a half normal level of both amidolytic and anticoagulant protein C activity, measured by different functional assays that use thrombin-thrombomodulin complex and a snake venom to activate protein C. Protein C antigen was always found to be within normal levels. Two young daughters of the propositus were found to have the same defect. Double-crossed immunoelectrophoresis, performed in the presence and absence of Ca2+ in the first dimension, showed no clear differences between patient and control PC. PC adsorption to barium salts was also found to be normal. Measurement of the PC activation peptide in the barium citrate eluates after PC activation showed no significant differences between patient and 10 normal controls, the concentration of this peptide being very similar to that of PC zymogen in the same eluates before PC activation. These results indicate that this abnormal PC is able to be normally activated by thrombin-thrombomodulin complex but does not exhibit serine protease activity, probably due to a defect in the PC molecule near the active site center.

Purification of six human vitamin K-dependent proteins in a single chromatographic step using immunoaffinity columns.

The major human vitamin K-dependent proteins were purified from plasma using immunoadsorbents made with antibodies specific for each protein. Monoclonal antibodies to Factor VII, Factor IX, Factor X, Protein C, and Protein S were prepared from mice immunized with isolated vitamin K-dependent antigens. Purified monoclonal antibodies and a purified burro polyclonal anti-prothrombin immunoglobulin were individually coupled to Sepharose and used in a tandem series of columns to purify each of the vitamin K-dependent proteins from eluates of barium citrate precipitates of plasma. The proteins were eluted from the columns by sodium thiocyanate and retained functional activity following dialysis. Prothrombin, Factor VII, Factor IX, Factor X and Protein C were essentially homogeneous as judged by NaDodSO4-PAGE; Protein S was isolated as a Protein S-C4b binding protein complex. These results indicate the utility of monoclonal antibody immunoadsorbents for purifying the human vitamin K-dependent proteins and represent a considerable simplification over other purification schemes.

Physicochemical characterization of human S-protein and its function in the blood coagulation system.

S-protein, the main inhibitor of the assembly of the membrane attack complex of complement, was isolated from human plasma by a simple purification procedure, which includes barium citrate adsorption, ammonium sulphate precipitation, chromatography on DEAE-Sephacel and Blue Sepharose and gel filtration on Sephacryl S-200. The homogeneous protein (sedimentation coefficient 4.6 S) was obtained in approx. 5% yield relative to its concentration in plasma, which was found to be 0.3-0.5 mg/ml. The final product did not cross-react with antisera against complement proteins or other proteinase inhibitors of human plasma. On polyacrylamide-gel electrophoresis in the presence of sodium dodecyl sulphate, S-protein migrated as a single-chain band with an apparent Mr of 74000 under non-reducing conditions and as a doublet of Mr 78000 and 65000 upon reduction. In plasma or serum S-protein also existed in two forms of corresponding Mr values, as was evidenced by an immunoblot enzyme-linked immunosorbent assay technique. S-protein was found to be an acidic glycoprotein with 10% (W/W) carbohydrate content and several isoelectric points in the range pH 4.75-5.25, and it contained one free thiol group per molecule of protein. The functional properties of S-protein in the complement system were demonstrated by its ability to inhibit complement-dependent cell lysis in a concentration-dependent manner (Ki 0.6 microM) and by its incorporation into the nascent SC5b-7 complex. A new function for S-protein could be revealed in the blood coagulation system. The slow progressive inhibition of thrombin by antithrombin III was not affected by S-protein, whereas the purified protein interfered with the fast inactivation of thrombin clotting as well as amidolytic activity by antithrombin III-heparin complex. The acceleration of this inhibition reaction by heparin was counteracted by S-protein, indicating the ability of S-protein to neutralize heparin activity.

Monoclonal anti-human factor VII antibodies. Detection in plasma of a second protein antigenically and genetically related to factor VII.

Several murine monoclonal anti-human Factor VII antibodies were produced using hybridoma technology. Two noncompetitive monoclonal antibodies were used to examine by Western blotting the Factor VII cross-reactive material (CRM) in normal human plasma and three commercially available congenitally Factor VII-deficient plasmas, and to construct a facile "sandwich" immunoassay for plasma Factor VII. A second, previously undescribed, form of Factor VII CRM was detected in human plasma, which on Western blotting stained with an apparent intensity 5-8% that of Factor VII. This glycoprotein, tentatively called VII*, has a molecular weight 4,500 D less than Factor VII, lacks detectable Factor VII functional activity, does not bind to barium citrate, and is not recognized by a monoclonal antibody that recognizes Factor VII but not alpha-chymotrypsin-treated Factor VII. VII* was not proteolytically produced from Factor VII during in vitro coagulation or after infusion of human Factor VII into rabbits. As determined by Western blotting, the human hepatoma cell line, HepG2, cultured in the presence of vitamin K, secreted relatively greater levels of VII* in proportion to VII (75%) than that found in human plasma. Warfarin treatment of HepG2 cells decreased the quantity of VII secreted by 77%, whereas it only inhibited the secretion of VII* by 14%. Immunologic studies of the plasmas from a patient on chronic warfarin therapy and an individual given a short course of high dose warfarin therapy corroborated the in vitro synthetic studies obtained with HepG2 cells. The data are consistent with the production of VII* by posttranslational, proteolytic, modification of VII, that, at least in the HepG2 cells studied, occurs intracellularly. However, other mechanisms for the production of VII*, in particular, alternative RNA splicing of the transcript from a single gene, cannot be excluded.