During the past 20 years contributions from many laboratories have led to the development of isolation procedures, delineation of primary structures, and more recently, to the expression of recombinant proteins associated with the coagulation cascade. In general, studies of coagulation proteins under defined conditions have demonstrated the prescience of Davie and Ratnoff and MacFarlane in their proposals of the coagulation cascade. The more recent discovery of thrombomodulin by Esmon et al has led to the identification and characterization of components of the vitamin K-dependent anticoagulant pathway. In this review we have attempted to analyze and compare the functional properties of each of the vitamin K-dependent enzyme complexes associated with the procoagulant and anticoagulant phases of blood clotting. Although dissimilarities exist, the vitamin K-dependent complexes have analogous requirements and appear to function with a common general mode of organization. Membrane-bound cofactors serve as anchoring sites for the appropriate membrane-binding enzymes. This process localizes the complex on the membrane surface and increases the catalytic efficiency for substrate utilization. Complex formation provides extraordinary improvements in the catalytic efficiency for the complexes as compared with their soluble enzyme components. Membrane- bound complexes provide a mechanism that can be regulated at a site by membrane presentation, zymogen activation, and cofactor activation or presentation. The kinetic constants obtained for the various coagulation reactions determined in vitro provide some insights into how these pathways may function in vivo. The catalytic efficiency (kcat/Km) for factor X activation by factor VIIIa/factor IXa is far in excess of the catalytic efficiency of activation of factor X by tissue factor/factor VIIa (Table 3). This may provide a rational interpretation for the observation that patients with hemophilia A and B bleed even though they appear to have an alternative pathway to factor X activation. In addition, tissue factor is not ordinarily presented by the vascular tissue that has direct access to blood. However, it appears that extravascular constitutive tissue factor is available once the blood vessel becomes disrupted. The efforts to identify the initiating reactions of the blood coagulation process have not been unambiguously successful. We conclude that factor VII is most likely a zymogen, just as are the other proenzymes of the blood clotting process. In addition, it is difficult to rationalize the importance of the intrinsic pathway of coagulation involving factor XII, prekallikrein, and high molecular weight kininogen since the congenital absence of any one of these factors does not result in abnormal bleeding.

No abstract

Because fibrin is commonly observed within arthritic joints, studies were undertaken to determine whether purified coagulation and fibrinolytic proteases degrade cartilage in vitro and to seek evidence for the activation of coagulation in arthritic joints through measurements of the levels of inhibitor-enzyme complexes and several other proteins associated with coagulation and fibrinolysis. The concentrations of 13 plasma proteins and complexes of thrombin and Factor Xa with antithrombin III were measured in synovial fluids recovered at the time of knee replacement surgery. All zymogens necessary to constitute the coagulation cascade were present. Thrombin and the combination of prothrombin plus prothrombinase induced proteoglycan release from both normal and arthritic cartilages. Factor Xa and plasmin induced release from diseased cartilage only, and urokinase, tissue plasminogen activator, and activated protein C were without effect at the levels used. At saturating levels of thrombin (-2.0 ,uM) 80% of the proteoglycan content of normal cartilage was released within 24 h. Thrombin, which is cationic, reversibly binds cartilage with Kd = 7.0±1.0 ,uM and B__ = 820±70 ng/mg of human cartilage. Levels of thrombin-antithrombin m complexes in synovial fluids and arthritis were 4-fold higher in osteo(OA) and 43-fold higher in rheumatoid(RA) than in controls (0.98 nM). Factor Xaantithrombin III complex levels were threefold lower in OA and fivefold higher in RA than in controls (0.24 nM). These elevated levels of enzyme-inhibitor complexes imply a history of activation of coagulation within the joint, especially in RA. Since thrombin degrades cartilage in vitro and had been generated in vivo, as inferred by the existence of thrombin-antithrombin m complexes, intraarticular activation of coagulation may both contribute to the pathology of arthritis and comprise a target for therapy and diagnosis. (J. Clin. Invest. 1994. 94:472-480.)

In previous studies using a nonhuman primate model of Protein C (PC) activation in vivo, immunoblotting showed substantial amounts of activated PC (APC) in a high molecular weight complex with what was presumed to be a previously unrecognized APC binding protein. This APC complex can also be formed in citrated plasma in vitro. It is of low electrophoretic mobility, sodium dodecyl sulfate (SDS) stable, with an apparent Mr of 320 Kd. Its purification from human plasma was accomplished using barium citrate adsorption, sequential polyethylene glycol (PEG) precipitations, diethylaminoethyl sepharose chromatography, AcA-34 gel filtration, and zinc-chelate affinity chromatography. This was monitored by subjecting the fractions to nondenaturing polyacrylamide gel electrophoresis (PAGE), transfer to polyvinylidene-difluoride membranes, and probing with 125I-labeled human APC. The purified APC-binding protein was homogeneous by SDS-PAGE with an Mr of 275 Kd. Its identity as alpha 2-macroglobulin (alpha 2M) was demonstrated immunochemically. Complex formation between alpha 2M and APC was found to be almost completely inhibited by EDTA, but to a lesser extent by citrate. Complex formation could also be prevented by active site inhibition with D-Phenylalanyl-L-Prolyl-L-Arginine- Chloromethyl Ketone (PPACK) or pretreatment of alpha 2M with methylamine. Incubation of APC (33 nmol/L) with alpha 2M (1 mumol/L) resulted in time-dependent inhibition of APC anticoagulant activity when measured using an activated partial thromboplastin time based APC assay. These data show that alpha 2M binds and inhibits APC in vitro and the interaction is both metal-ion and active-site dependent, requiring functionally intact alpha 2M. As the complexes formed in vitro comigrate electrophoretically with those observed in vivo after PC activation, it is suggested that alpha 2M is a physiologically relevant inhibitor involved in the processing of APC in vivo.

A model of Protein C (PC) activation in vivo was used to investigate the complexing of activated PC (APC) with its plasma inhibitors, PC inhibitor (PCI) and alpha 1-antitrypsin (alpha 1AT). Chimpanzees were infused with a bolus of activated factor X (F.Xa) together with vesicles of phosphatidylcholine and phosphatidylserine (PCPS). Pre- and post-infusion plasma samples were analyzed using enzyme linked immunosorbent based assays (ELISA) for PC and APC complexes, and immunoblotting of PC from nondenaturing polyacrylamide gel electrophoresis. Within 2 minutes of infusion, a 60% decrease in nonactivated PC zymogen (PCz) levels was observed. This coincided with a precipitous drop in plasma activities of cofactors VIIIa and Va. In contrast, total PC antigen (PCt) levels decreased by only 1%, indicating APC generation. Complexes of APC with both PCI and alpha 1AT were observed on immunoblots, and further identified and quantified using a sandwich ELISA employing antibodies to both PC and these inhibitors. The distribution of APC between these two inhibitors varied with the dose of F.Xa/PCPS infused. At a dose of F.Xa/PCPS of 24.05 pmol and 37.70 nmol/kg, respectively, an initial spike of APC generation, associated with decreases in the levels of factors VIIIa and Va, was noted but dissipated over the next 30 minutes. During this period, APC/inhibitor complexes appeared with the levels of APC-PCI and APC-alpha 1AT reaching 8.5 nmol/L and 2.2 nmol/L by 30 minutes, respectively. In contrast, at a higher dose of F.Xa/PCPS of 36.60 pmol and 56.30 nmol/Kg respectively, complexes of APC-alpha 1AT appeared rapidly and reached a level of 6 nmol/L by 30 minutes postinfusion, whereas APC-PCI complexes were only present at a concentration of 3.4 nmol/L by this time. Additional experiments with lower doses of F.Xa/PCPS suggest that PCI is the preferred inhibitor of APC, but as the availability of this inhibitor becomes limiting, alpha 1AT plays an increasingly crucial role as a secondary inhibitor of endogenously generated APC. Moreover, evidence is presented suggesting the existence of additional inhibitor(s) of APC that may have a role similar to alpha 1AT.