The lipoprotein-associated coagulation inhibitor (LACI) is present in vivo in at least three different pools: sequestered in platelets, associated with plasma lipoproteins, and released into plasma by intravenous heparin, possibly from vascular endothelium. In this study we have purified the heparin-relesable form of LACI from post-heparin plasma and show that it is structurally different from lipoprotein LACI. The purification scheme uses heparin-agarose chromatography, immunoaffinity chromatography, and size-exclusion chromatography and results in a 185,000-fold purification with a 33% yield. Heparin- releasable LACI (HRL), as analyzed by sodium dodecyl sulfate- polyacrylamide gel electrophoresis, under reducing conditions, appears as a major band at 40 Kd and a minor band at 36 Kd. Immunoblot analysis suggests that the 36-Kd band arises from carboxyl-terminus proteolysis that occurs during the purification. HRL has a specific activity similar to that of HepG2 or lipoprotein LACI. HRL and lipoprotein LACI combine with lipoproteins in vitro while purified HepG2 LACI does not. I125-labeled HRL, injected into a rabbit, is cleared more slowly than I125-labeled HepG2 LACI, which may be due to attachment to lipoproteins in vivo. Preliminary evidence suggests that HRL is associated with vascular endothelium, possibly by attachment to glycosaminoglycans.

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.

Stimulation with thrombin or the calcium ionophore, A23187 caused human platelets to release a coagulation inhibitor similar to the Lipoprotein Associated Coagulation Inhibitor (LACI). This was documented functionally, with clotting assays measuring tissue factor inhibition and factor Xa inhibition, as well as immunologically, in a competitive immunoassay. The total amount of LACI released by 3 x 10(8) platelets after two hours stimulation was 7% to 8% of the amount found in 1 mL of serum. Half of the LACI was released by five minutes. The LACI was present in the platelet supernatant and was not associated with the platelet membrane or shed vesicles. The tissue factor and factor Xa inhibitory activities that were released were neutralized by preincubating the platelet supernatants with specific rabbit polyclonal anti-LACI IgG. On Western blot, platelet LACI appeared to run as a doublet with a molecular weight (mol wt) 45,000 to 47,000. Blood samples obtained from the site of a wound (template bleeding time) demonstrated a progressive increase in LACI concentration. A cDNA probe, derived from endothelial cell LACI cDNA, hybridized selectively to 4.0 and 1.4 kb transcripts in a preparation of platelet mRNA.

Factor V (Va) is essential for binding of factor Xa to the surface of platelets. After thrombin treatment, normal platelets release at least five times more factor Va activity than is required for maximal factor Xa binding. The concentration of factor V activity obtained after thrombin stimulation of 10(7) normal platelets is sufficient to allow half-maximal factor Xa binding to 10(8) platelets (10% normal, 90% factor-V deficient). Therefore, factor Va activity is not limiting in platelet-surface factor Xa binding and prothrombin activation in normal platelets; some other components limit the number of binding sites. We report studies of a patient (M.S.) with a moderate to severe bleeding abnormality whose platelets are deficient in the platelet-surface component required for the factor Va-factor Xa binding. The patient's platelet factor Va activity released after thrombin treatment is normal, but factor Xa binding is 20%-25% of control values at saturation. Abnormal prothrombin consumption in a patient with normal plasma coagulation factors and platelet function suggests a disorder in platelet-surface thrombin formation.

Blood coagulation is initiated when plasma factor VII(a) binds to its essential cofactor tissue factor (TF) and proteolytically activates factors X and IX. Progressive inhibition of TF activity occurs upon its addition to plasma. This process is reversible and requires the presence of VII(a), catalytically active Xa, Ca2+, and another component that appears to be associated with the lipoproteins in plasma, a lipoprotein-associated coagulation inhibitor (LACI). A protein, LACI(HG2), possessing the same inhibitory properties as LACI, has recently been isolated from the conditioned media of cultured human liver cells (HepG2). Rabbit antisera raised against a synthetic peptide based on the N-terminal sequence of LACI(HG2) and purified IgG from a rabbit immunized with intact LACI(HG2) inhibit the LACI activity in human serum. In a reaction mixture containing VIIa, Xa, Ca2+, and purified LACI(HG2), the apparent half-life (t1/2) for TF activity was 20 seconds. The presence of heparin accelerated the initial rate of inhibition threefold. Antithrombin III alpha alone had no effect, but antithrombin III alpha with heparin abrogated the TF inhibition. LACI(HG2) also inhibited Xa with an apparent t1/2 of 50 seconds. Heparin enhanced the rate of Xa inhibition 2.5-fold, whereas phospholipids and Ca2+ slowed the reaction 2.5-fold. Xa inhibition was demonstrable with both chromogenic substrate (S-2222) and bioassays, but no complex between Xa and LACI(HG2) could be visualized by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Nondenaturing PAGE, however, showed that LACI(HG2) bound to Xa but not to X or Xa inactivated by diisopropyl fluorophosphate. Thus, LACI(HG2) appears to bind to Xa at or near its active site. Bovine factor Xa lacking its gamma-carboxyglutamic acid-containing domain, BXa(-GD), through treatment with alpha-chymotrypsin, was used to further investigate the Xa requirement for VIIa/TF inhibition by LACI(HG2). LACI(HG2) bound to BXa(-GD) and inhibited its catalytic activity against a small molecular substrate (Spectrozyme Xa), though at a rate approximately sevenfold slower than native BXa. Preincubation of LACI(HG2) with saturating concentrations of BXa(-GD) markedly retarded the subsequent inhibition of BXa. The VII(a)/TF complex was not inhibited by LACI(HG2) in the presence of BXa(-GD), and further, preincubation of LACI(HG2) with BXa(-GD) slowed the inhibition of VIIa/TF after the addition of native Xa. The results are consistent with the hypothesis that inhibition of VII(a)/TF involves the formation of a VIIa-TF-XA-LACI complex that requires the GD of XA.(ABSTRACT TRUNCATED AT 400 WORDS).