Stabilization of factor VIII in plasma by the von Willebrand factor. Studies on posttransfusion and dissociated factor VIII and in patients with von Willebrand's disease.

In normal plasma, the ratio of the procoagulant activity of factor VIII (VIII(AHF)) to that of the von Willebrand factor activity (ristocetin cofactor, VIII(VWF)) or factor VIII antigen (VIII(AGN)) is approximately 1, but ratios > 1 (e.g., VIII(AHF) > VIII(VWF) or VIII(AGN)) may be observed in some patients with von Willebrand's disease and in the "late" posttransfusion plasmas of patients with this disorder. The lability of VIII(AHF) was studied by incubating plasma, diluted 1:10 in imidazole buffer pH 7.1, for 6 h at 37 degrees C. With normal plasmas, 77+/-12% (SD) of the original VIII(AHF) activity remained after incubation. VIII(AHF) was labile (e.g., 35-55% residual activity) in the "late" posttransfusion plasmas (VIII(AHF) >> VIII(VWF)) of a patient with von Willebrand's disease, but not in the "early" posttransfusion plasmas (VIII(AHF) approximately VIII(VWF)). VIII(AHF) was also labile in the (base-line) plasmas of three patients with von Willebrand's disease in whom the ratios of VIII(AHF) to VIII(VWF) were 4.4 to 8.1, but not in the plasmas of four other patients in whom the ratio was approximately 1. The electrophoretic mobility of factor VIII antigen was increased in two of the three patients with labile VIII(AHF). In both of these patients, and in the late posttransfusion plasmas, labile VIII(AHF) activity could be stabilized by the addition of purified von Willebrand factor (lacking VIII(AHF) activity) or by hemophilic plasma, but not by plasmas of patients with severe von Willebrand's disease. Thus, VIII(VWF) may serve to stabilize VIII(AHF) and this might explain the posttransfusion findings in von Willebrand's disease.

Dissociation of factor VIII in the presence of proteolytic inhibitors.

When gel filtration of factor VIII is performed with buffers of high ionic strength (1.0 M NaCl or 0.25 M CaCl2), the procoagulant activity elutes with proteins of relatively low molecular weight. It has been suggested that in the presence of proteolytic inhibitors, the procoagulant activity would appear at the void volume. To test this hypothesis, chromatography with buffers of high ionic strength was performed in the presence of benzamidine hydrochloride, soy bean trypsin inhibitor, heparin, DFP, and aprotinin. Under all of these conditions, the procoagulant activity continued to elute with proteins of low molecular weight. Similar findings were obtained after chromatographing cryoprecipitate prepared from the plasma of a normal subject who had received heparin. Thus, at present there is no direct evidence to suggest that proteolysis is involved in the dissociation of factor VIII by buffers of high ionic strength.

Subunit composition of plasma von Willebrand factor multimers: evidence for a non-proteolytic mechanism resulting in apparent increase in proteolytic fragments.

Plasma von Willebrand factor (vWf) consists of a series of multimers of different molecular sizes. When analyzed for subunit composition, plasma vWf contained a major polypeptide of approximately 225 kD, and at least three smaller proteolytic fragments. In the present study, the subunit composition of each of these multimers was analyzed using a two dimensional electrophoresis technique. Although nearly all the multimers, as separated by SDS 1.5% agarose gel electrophoresis, contained both the major polypeptide and the smaller fragments, the relative amount of the fragments varied among the multimers. The smaller multimers contained relatively more proteolytic fragments. After the larger multimers were depleted by incubation of plasma with formaldehyde-fixed platelets and ristocetin, the intact subunit decreased, while the amount of the proteolytic fragments remained relatively unchanged. The findings of this study are consistent with the scheme that plasma multimers undergo proteolytic cleavage after multimer assembly. Furthermore, an apparent increase in the amount of these proteolytic fragments may result from mechanisms unrelated to proteolysis, such as selective adsorption of the larger multimers onto platelets.

A new von Willebrand variant (type I, New York): increased ristocetin-induced platelet aggregation and plasma von Willebrand factor containing the full range of multimers.

We report three members of a family who had reduced levels of plasma von Willebrand factor (vWF) and increased ristocetin-induced platelet aggregation (RIPA) (aggregation of platelet-rich plasma with ristocetin at a concentration of 0.45 mg/mL), as previously reported in type IIB and pseudo-von Willebrand's disease (vWD). However, in contrast to the latter two disorders in which the larger vWF multimers are absent in plasma, the entire range of vWF multimers was observed in the patients' plasma after sodium dodecyl sulfate-agarose gel electrophoresis, and all vWF multimers (including the largest) were present in the same proportion as in normal plasma and type I vWD. Thus, despite increased RIPA, the levels and multimeric pattern of vWF in this family's plasma were indistinguishable from those in type I vWD in which RIPA is usually decreased. Addition of ristocetin to the patients' platelet-rich plasma resulted in the removal of vWF (and, more selectively, of the large multimers) at lower concentrations of ristocetin than normal, as in type IIB and pseudo-vWD. The defect in the patients was localized to their vWF, which had an enhanced capacity for aggregating washed normal platelets in the presence of low concentrations of ristocetin and for aggregating pseudo-vWD platelets (in the absence of ristocetin). Both glycoproteins (GP) Ib and IIb-IIIa were involved in the enhanced aggregation response. RIPA (at low ristocetin concentrations) in the patients' platelet-rich plasma was abolished by a monoclonal antibody (AP1) to GPIb and was markedly reduced by monoclonal antibodies (10E5 and LJP9) that block adenosine diphosphate and thrombin-induced binding of vWF and fibrinogen to GPIIb-IIIa but was unaffected by an antibody (LJP5) that only blocks vWF binding. Partial inhibition of the initial aggregation slope (and complete inhibition of second phase aggregation) was achieved with creatine phosphate/creatine phosphokinase. EDTA blocked second-phase aggregation but was without effect on the initial slope. The findings in this family combine some features of both type I vWD (normal pattern of vWF multimers in plasma) and type IIB vWD (increased RIPA) and further demonstrate the increasing complexity of the structure-function relationships in vWD.

The destabilization of factor VIII by a vitamin K dependent protein.

Since a vitamin K dependent protein, protein C, can inactivate factor VIII, a study was undertaken to determine if the level and stability of factor VIII in plasma are influenced by such a protein. Factor VIII lability was determined by incubating citrated plasma, diluted 1:10 and 1:20 in pH 7 X 2 imidizole buffer, for 6 h at 37 degrees C. Normal plasma had a mean factor VIII of 98 +/- 61 U/100 ml. The amount of factor VIII remaining after 6 h of incubation was 68 +/- 14% of the original factor VIII level. In warfarinized patients, factor VIII (218 +/- 65 U/100 ml) and VWF:AGN (331 +/- 102 U/100 ml) were elevated (P less than 0.001). Following incubation, their residual activity was 103 +/- 20% of the original factor VIII level. In samples taken after warfarin was discontinued, normal factor VIII lability returned, while plasma levels of factor VIII and VWF:AGN remained elevated. Similarly, in the plasma of a vitamin K deficient patient, increased factor VIII stability was also evident; lability was restored following vitamin K replacement. We conclude that factor VIII stability is determined in part by a vitamin K dependent protein. In clinical states in which this protein is functionally absent, factor VIII is elevated and more stable.

Postoperative thrombocytopenia in type IIB von Willebrand disease.

We report studies of a large kindred with type IIb von Willebrand disease and manifestations of thrombocytopenia. While only one member of the family was thrombocytopenic routinely, three members of the family who underwent various surgical procedures demonstrated thrombocytopenia and platelet clumping postoperatively. Platelet clumps were found on peripheral blood smear only in the immediate postoperative specimens and did not appear to be a technical artifact. In the one patient who received no preoperative prophylactic therapy, postoperative plasma specimens showed the transient appearance of high molecular weight von Willebrand factor multimers. These results support the hypothesis that surgery, or some related aspect such as stress, led to the release of high molecular weight multimers, resulting in platelet clumping and removal from the circulation, and subsequent thrombocytopenia. Thrombocytopenia under conditions of stress may be a more common manifestation of type IIb vWd than is currently appreciated.

Safety of invasive procedures in patients with the coagulopathy of liver disease.

To assess the need for prophylactic fresh frozen plasma in patients with chronic liver disease before procedures, we followed 39 consecutive patients with prothrombin times (PT) of 15-29 s who had 71 invasive procedures. A total of 57 procedures was done in 30 patients receiving no blood product support, while 12 patients received blood products within 12 h of 14 procedures. Only three of the latter group received fresh frozen plasma prophylactically to improve the PT. The two groups were similar in the severity of their liver disease. There were nine surgical procedures as well as paracenteses, thoracenteses, lumbar punctures, and central venous line placement. Three bleeding episodes occurred. Two of the bleeding episodes required no further treatment. Because of the low incidence of bleeding and the ease in controlling the bleeding once it occurred, fresh frozen plasma is not recommended for prophylaxis of an elevated prothrombin time.

Subendothelial deposition of von Willebrand's factor requires the presence of endothelial cells.

The presence of FVIII/VWF in human aortic subendothelium has been previously established. The present study was undertaken to determine the origin of this FVIII/VWF. Indirect immunofluorescence microscopy was employed to detect FVIII/VWF in rabbit aortas, after passage of a balloon catheter. Immediately after this de-endothelialization, FVIII/VWF was found to be present directly luminal to the internal elastic lamina. However, in the subsequent 5 days to 2 months, no FVIII/VWF was found on the luminal surface of exposed, nonre-endothelialized neointima. After endothelial cell regrowth, however, FVIII/VWF was again seen on the luminal surface. Thus these studies reveal that the deposition of subendothelial FVIII/VWF in neointima requires the presence of endothelial cells and that relatively trivial amounts of plasmatic FVIII/VWF are deposited on neointimal connective tissue after de-endothelialization.

Monitoring heparin therapy--a role for the chromogenic assay.

Heparin therapy was evaluated by simultaneous determinations of the activated partial thromboplastin time (APTT), and plasma heparin levels by chromogenic assay. These tests showed good correlation (r = 0.73) in patients on heparin alone. However, in the presence of concurrent warfarin administration, r fell to 0.37, reflecting an additional warfarin effect upon the APTT. Furthermore, the studies confirmed the increased heparin requirement in patients with pulmonary embolization.

Long-term culture of human endothelial cells.

Human umbilical vein endothelial cells can be grown in vitro for 28 passages (CPDL 58) in Medium 199 supplemented with newborn bovine serum and a partially purified growth factor derived from bovine brain. Newborn bovine serum is superior to fetal bovine serum for the proliferation of human umbilical vein endothelial cells seeded at low density in the presence of the growth factor. The endothelial cells, which can be passaged every 7 to 10 d at a 1-to-5 split ratio, retain their morphological and biochemical characteristics. The proliferation of cells seeded at low density (10(3)/cm2) is proportional to the concentration of the growth factor present in the medium. The growth factor, which has an isoelectric point between 5.0 and 5.5, can support cell proliferation at reduced serum concentrations; half-maximal growth is achieved in medium containing the growth factor and 3% serum. The brain endothelial cell growth factor does not stimulate DNA synthesis significantly in cultures of human skin fibroblasts.

Concentration-dependent dissociation of factor VIII in 1 M NaCl.

Plasma, cryoprecipitate, Hemofil, and human factor VIII concentrate were dissolved in 1.0 M NaCl and chromatographed on Bio-Gel A-5m. With high concentrations of factor VIII the activity eluted as a symmetrical peak in the void volume; with a low factor VIII concentration the procoagulant activity was retarded. Dilution curves were performed for several human factor VIII concentrates. When the concentration of factor VIII was decreased, elution patterns showed a gradual transition from a peak in the void volume to a peak with a Ve/Vo of 1.7. Cryoprecipitate exhibited a similar behavior in 1.0 M NaCl, but the percent dissociation was greater than expected at high concentrations of factor VIII. When gel filtration was performed with 0.25 M CaCl2, significant dissociation occurred at all concentrations of factor VIII tested. The behavior of factor VIII in 1.0 M NaCl closely fit a theoretically derived curve for the dissociation of a protein from its binder. We conclude that the dissociation of factor VIII in 1 M NaCl is dependent on the concentration and purification of the procoagulant protein.

Shear stress enhances the proteolysis of von Willebrand factor in normal plasma.

While von Willebrand factor (vWF) is secreted from endothelial cells as a very large polymer, it circulates as a series of multimers that are reducible to a 225-kD polypeptide and three proteolytic fragments of 189, 176, and 140 kD. Cleavage at the Tyr-842/Met-843 bond of the vWF polypeptide creates the 140- and 176-kD fragments. In the process of understanding vWF multimer formation, the role of shear stress in vWF proteolysis was investigated in this study. A shear-rate-dependent loss of the largest multimers was observed when normal plasma was perfused through long capillary tubings achieving shear rates normally encountered in the circulation. The shear-dependent vWF change was not observed when purified vWF or normal plasma containing calcium chelator EGTA or EDTA was perfused. As the large multimers decreased, an increase in the smaller multimers, including 200- and 350-kD bands, was detected. Elution and immunoblotting studies with peptide-specific antibodies LJ-7745 and VP-1 showed that the 200-kD band was a dimer of the 140-kD fragment, whereas the 350-kD band was a dimer of the 176-kD fragment. When analyzed after disulfide bonds were reduced, sheared plasma showed an increase in the 176- and 140-kD fragments, but not the 189-kD fragment. Finally, shearing of purified vWF enhanced its proteolytic cleavage when it was subsequently incubated with the cryosupernatant fraction of normal plasma or with cathepsin G, a leukocyte granule serine protease. These results show that shear stress is capable of enhancing the susceptibility of vWF to proteolytic cleavage. It promotes vWF proteolysis in normal plasma at a site that generates the 140-kD/176-kD fragments, leading to a decrease in multimer size. Shear stress might be involved in modulating the size of vWF in the circulation.

Oral contraceptives and platelet aggregation.

There are conflicting data in the literature concerning the effects of oral contraceptives on platelet aggregation. We studied 26 healthy women who used oral contraceptives for at least three months. The control group consisted of 25 healthy women of childbearing age, who were receiving no medications. No differences in platelet aggregation in response to adenosine diphosphate, epinephrine, and collagen were found between the two groups.

Multimeric composition of endothelial cell-derived von Willebrand factor.

The multimeric composition of human endothelial cell (EC)-derived von Willebrand factor (vWF) was studied using SDS-agarose gel electrophoresis and autoradiography. Two multimers were found in lysates prepared from confluent cultures of human umbilical vein endothelial cells. The smaller multimer had a molecular weight (mol wt) of approximately 950 Kd, while the second was larger than those seen in plasma. When electrophoresis was performed using the discontinuous buffer system of Ruggeri and Zimmerman, the small multimer consisted of a single band migrating with the slowest-moving component of the corresponding plasma triplet. The large EC-vWF multimer was detected in culture media conditioned with EC monolayers for ten minutes. It remained the only multimer in media conditioned for up to three days. Calcium ionophore A23187 increased the amount of the large vWF multimer released into the culture media, but did not change its multimeric composition. The small multimer was never detected in the EC-conditioned media. These findings suggest that (1) a large, fully polymerized multimer of vWF is released from the ECs, while the small multimer probably represents a major intermediate component in the process of multimerization, and (2) plasma vWF multimers are probably generated from the large endothelial vWF after it is released into the circulation.

Suppression of fibroblast proliferation in vitro and of myointimal hyperplasia in vivo by the triazolopyrimidine, Trapidil.

Platelet derived growth factor (PDGF) is a mitogen capable of stimulating the proliferation of both vascular smooth muscle cells (SMC) and fibroblasts in tissue culture. Recently, it was reported that this in vitro stimulatory effect is significantly reduced by Trapidil, a triazolopyrimidine. We confirm this, and also that similar inhibition occurs in vivo, in balloon catheter deendothelialized aorta of male, Sprague-Dawley rats treated with Trapidil. Two groups of rats, of 8 animals each, were treated daily with oral doses of either 45 mg or 90 mg Trapidil. A third group of 9 control animals were treated identically, but received only the diluent used to dissolve the drug. ADP or collagen-induced platelet aggregation, and endothelial cell regrowth were unaffected by Trapidil administration, but the degree of myointimal hyperplasia was significantly reduced in all animals receiving the drug. Thus, Trapidil seems to possess the selective ability to alter the SMC proliferative response which normally follows deendothelialization.

Factor VIII/von Willebrand factor in subendothelium mediates platelet adhesion.

The present studies were undertaken to determine whether factor VIII/von Willebrand factor (vWF) present in the vessel wall (in addition to that in plasma) may mediate the attachment of platelets to subendothelium. Subendothelium from everted rabbit aorta was exposed to human citrated blood flowing through an annular perfusion chamber at 40 mL/min (wall shear rate of 2,600 s-1 for five minutes). The vessel segments were incubated at 37 degrees C for one hour with various dilutions of either goat-anti-rabbit factor VIII/vWF serum or an IgG fraction prepared from the serum. Control segments were incubated with serum or IgG from a nonimmunized goat. Values of platelet contact (C), platelet adhesion (C + S), and thrombus formation (T) on the subendothelium were evaluated by a morphometric technique. Compared with vessels incubated with fractions prepared from a normal goat, a significant decrease in platelet adhesion (C + S), ranging from 45% to 65%, was observed on vessels incubated with various dilutions (1:5 to 1:50) of either serum or IgG fractions of goat-anti-rabbit factor VIII/vWF. A similar decrease in platelet adhesion was observed with vessels incubated with an F(ab')2 fragment against rabbit factor VIII/vWF prepared in the goat. When goat-anti-rabbit factor VIII/vWF IgG was added to rabbit blood (1:75 dilution), platelet adhesion was reduced to the same extent (65%) on normal rabbit vessels and on vessels pre-incubated with goat-anti-rabbit factor VIII/vWF. Immunofluorescence studies revealed the presence of rabbit factor VIII/vWF in the subendothelium of rabbit aorta and the continued binding of the goat-anti-factor VIII/vWF antibodies on subendothelium during the perfusion studies. No uptake of human factor VIII/vWF on the rabbit subendothelium was observed by this immunologic technique; human factor VIII/vWF was found to be entirely associated with the attached human platelets. Thus, factor VIII/vWF in the vessel wall may mediate platelet attachment to subendothelium in a manner similar to that of plasma factor VIII/vWF.

Sub-aggregatory doses of catecholamines prevent prostacyclin-induced inhibition of platelet aggregation.

To assess the effect of subaggregatory concentrations of catecholamines on the antiaggregatory effect of prostacyclin (PGI2), platelets from normal human volunteers were exposed sequentially in vitro to epinephrine (less than or equal to 50 nM)- or norepinephrine (less than or equal to 1 microM) followed by PGI2 and adenosine diphosphate (ADP). Platelets thus pretreated did not manifest the normal inhibitory response to PGI2, aggregating to a similar extent as platelets exposed to ADP alone. This effect was unaffected by aspirin but was abolished by exposure to phentolamine. Catecholamine pretreatment similarly blocked the PGI2-induced increase in intracellular cyclic AMP, an effect which was also reversed by phentolamine. These data suggest that platelets exposed in vivo to elevated catecholamine concentrations, such as are seen clinically during myocardial infarction, might be similarly unresponsive to endogenous PGI2.

Pseudothrombocytopenia masking true thrombocytopenia.

Pseudothrombocytopenia owing to platelet clumping is usually associated with blood specimens anticoagulated with EDTA. It may also be seen if specimens possessing IgM cold agglutinins are processed at room temperature. A patient with a temperature-independent, EDTA-independent agglutinin is reported whose pseudothrombocytopenia was masking true thrombocytopenia. A technique for blood collection when evaluating similar cases is described.