Subcellular platelet factor VIII antigen and von Willebrand factor

Subcellular membrane and granule fractions derived from human platelets contain factor VIIII antigen and von Willebrand factor activity but not factor VII procoagulant activity. Circulating platelets constitute a significant reservoir of plasma factor VIII antigen, containing approximately 15% of the amount of factor VIII antigen present in platelet-poor plasma. The antibiotic ristocetin, which aggregates human platelets in the presence of von Willebrand factor, nonspecifically precipitates platelet membrane factor VIII antigen. Thus normal platelets contain surface-bound as well as internally stored von Willebrand factor, a protein synthesized by endothelial cells which is necessary for normal platelet function in vivo.     

Double-antibody radioimmunoassay for factor VIII-related antigen.

A plasma protein required for the support of ristocetin-induced platelet aggregation was isolated from antihemophilic factor concentrate and radiolabeled with 125I. A double-antibody radioimmunoassay was developed, with use of specific rabbit anti-VIII related antigen serum and goat anti-rabbit globulin. The assay is sensitive, reproducible, and technically simple to perform. Values obtained in normal subjects ranged from 0.65 to 1.53 units, similar to our normal range for VIII coagulant activity (0.67-1.43 units). However, normal or increased values of VIII-related antigen were observed in VIII coagulant-deficient hemophiliacs. Also, concentrations of VII-related antigen significantly exceeded coagulant concentrations in several patients with liver disease or disseminated intravascular coagulation, or both. Of a broad selection of congenital coagulation disorders examined, only patients with von Willebrand's disease had decreased VIII-related antigen concentrations, and these corresponded to the lowered concentration of ristocetin cofactor in the patients. In three transfused patients, VII-related antigen values correlated with the concentration of the cofactor. Our results suggest that the radioimmunoassay of VIII-related antigen is a simple and valuable adjunct in the study of patients with clotting abnormalities.     

Active release of human platelet factor VIII-related antigen by adenosine diphosphate, collagen, and thrombin.

Platelet Factor VIII-related antigen (VIIIR:Ag) represents a significant proportion of the total circulating VIIIR:Ag pool. However, its participation in the events of primary hemostasis has not been shown. We now report that platelet-contained VIIIR:Ag is released from platelets by collagen, ADP and thrombin. The concentrations of these agonists, required for VIIIR:Ag release, are the same or lower than those required for release of serotonin, lysosomal enzymes, or fibrinogen. This release has the features of an energy-dependent secretory response because it is blocked by the metabolic inhibitors, antimycin A and 2-deoxy-D-glucose. The electrophoretic characteristics of the VIIIR:Ag released by collagen and ADP are similar to those of plasma VIIIR:Ag. However, thrombin-released platelet VIIIR:Ag differs from that of plasma in that the less anodal forms are relatively depleted. These differences do not appear to be the result of proteolytic degradation of platelet-derived VIIIR:Ag, but may reflect interactions between specific molecular forms of VIIIR:Ag and the platelet membrane. These studies suggest mechanisms by which platelet-contained VIIIR:Ag may contribute to the primary events of hemostasis.     

Evidence that the primary binding site of von Willebrand factor that mediates platelet adhesion on subendothelium is not collagen.

We have studied the binding of von Willebrand factor to extracellular matrices of endothelial cells and to the vessel wall of human umbilical arteries in relation to its function in supporting platelet adhesion. CLB-RAg 201, an MAb against von Willebrand factor, completely inhibits the binding of von Willebrand factor to collagen type I and type III. CLB-RAg 201 does not inhibit the binding of 125I-von Willebrand factor to extracellular matrices of endothelial cells, to smooth muscle cells, or to the subendothelium. CLB-RAg 201 partly inhibits platelet adhesion to these surfaces, but this directly affects the interaction between von Willebrand factor and platelets and is not due to inhibition of binding of von Willebrand factor to these surfaces. Another MAb, CLB-RAg 38, does not inhibit the binding of von Willebrand factor to collagen. CLB-RAg 38 completely inhibits the binding of von Willebrand factor to extracellular matrices. CLB-RAg 38 inhibits platelet adhesion to cellular matrices completely insofar as it is dependent on plasma von Willebrand factor. CLB-RAg 38 does not inhibit the total binding of von Willebrand factor to subendothelium, as there are too many different binding sites, but it completely inhibits the functional binding sites for von Willebrand factor that support platelet adhesion. The epitopes for CLB-RAg 38 and 201 on the von Willebrand factor molecule are located on different fragments of the molecule. These results indicate that von Willebrand factor binds to subendothelium and matrices of cultured cells by a mechanism that is different from that by which it binds to collagen.     

Concentrations of factor VIII-related antigen and factor XIII during open heart surgery

Plasma levels of factor VIII-related antigen (fVIIIRA) and factor XIII S and A subunits (fXIIIS, fXIIIA) were assayed by counterimmunoelectrophoresis before, during, and after cardiopulmonary bypass (CPB) in patients with coronary artery and valvular heart disease to define the basis for clinical and laboratory abnormalities of hemostasis occurring in this form of surgery. During CPB, concentrations of fXIIIA dropped in both patient groups but returned to preoperative levels promptly after pump removal. In contrast, fVIIIRA and fXIIIS, which are not incorporated into the clot, remained unchanged even during fluid administration. These data provide evidence of a transient consumption coagulopathy as a feature of CPB. Hemodilution probably plays a secondary role in these changes.     

Immunoradiometric measurement of the factor VIII procoagulant antigen.

A fluid-phase immunoradiometric assay has been developed which identifies an antigen on the Factor VIII (antihemophilic factor) procoagulant protein. This sensitive and quantitative assay is not influenced by levels of Favor VIII-related antigen (von Willebrand factor) or other plasma proteins. There is a close correlation of procoagulant activity and immunologically detectable protein in normal and von Willebrand's disease plasmas. In contrast, several different patterns have been identified in hemophilic plasmas. Neither procoagulant activity nor procoagulant antigen is detectable in plasmas from patients with severe classic hemophilia. Patients with mild and moderate hemophilia have either comparable plasma concentrations of procoagulant activity and procoagulant antigen or relatively greater levels of immunologically detectable protein.     

Effect of insulin treatment in streptozocin-induced diabetic rats on in vitro platelet function and plasma von Willebrand factor activity and factor VIII-related antigen

Diabetes mellitus is associated with altered platelet function and endothelial damage, but their relationship remains unclear. We examined the effect of short-term metabolic control with insulin in 14- and 28-day streptozocin-induced diabetic rats on alterations in in vitro platelet aggregation and serotonin release. Endothelial damage was assessed by plasma concentrations of von Willebrand factor activity (VIIIR:WF) and factor VIII-related antigen (VIIIR:Ag). Insulin was administered for 5 or 7 days at 9 or 21 days, respectively, after streptozocin. Enhanced platelet aggregation responses to adenosine diphosphate (ADP) and thrombin occurred after both durations of diabetes. Insulin therapy returned ADP-induced, but not thrombin-induced, responses to normal. Enhanced thrombin-induced platelet release of serotonin occurred at both times. Collagen-induced platelet release was enhanced in 28-day diabetic rats. Insulin therapy returned these responses to normal. Plasma concentrations of VIIIR:WF and VIIIR:Ag were elevated in 28-day, but only VIIIR:WF was elevated in 14-day diabetic rats. Insulin therapy reduced the elevated levels of VIIIR:Ag in 28-day diabetic rats, but had little effect on either parameter after the shorter duration of diabetes. In summary, Enhanced platelet aggregation and increased release of serotonin occur shortly after the induction of diabetes by streptozocin in adult rats. These platelet changes precede alterations of endothelial function, as determined by plasma VIIIR:WF and VIIIR:Ag levels. Platelet changes respond more rapidly to insulin therapy than do endothelial changes in diabetic rats. The duration of diabetes before insulin therapy does not affect these relationships.(ABSTRACT TRUNCATED AT 250 WORDS)     

Interaction of factor VIII antigen in hemophilic plasmas with human antibodies to factor VIII.

By utilizing a simple modification of previous immunological assays, we have demonstrated that most, if not all, hemophilic plasmas contain antigen reactive with human antibodies directed against Factor VIII procoagulant activity (VIIIc). Antibodies developing in a nonhemophiliac patient and in a hemophiliac patient gave similar results. The VIIIc antigen so identified was removed from hemophilic plasmas with immobilized rabbit antibody which reacted with normal VIIIc and von Willebrand's disease antigen. These data suggest that there are greater antigenic similarities between normal and hemophilic Factor VIII than previously thought.     

Fibronectin and factor VIII-related antigen in liver cirrhosis and acute liver failure

The liver is involved in the turnover of fibronectin in two different ways: hepatic synthesis contributes substantially to the plasma fibronectin pool, while Kupffer-cells, performing an important role of the reticuloendothelial system, remove fibronectin opsonized material from the circulation. In 45 patients with histologically confirmed liver cirrhosis and six patients with acute liver failure due to intoxication we determined fibronectin concentration in plasma by electroimmunoassay and additionally measured factor VIII-related antigen, which is a large glycoprotein not synthesized in the liver. Fibronectin levels in plasma were decreased in liver cirrhosis. This decrease was correlated with the extent of porto-caval collateral circulation. Very low levels were found in patients with acute liver failure. Factor VIII-related antigen levels were greatly increased as a function of the hepatic insufficiency. Between both parameters there was a significant inverse correlation. It is concluded that the simultaneous determination of both proteins provides reliable information about the remaining liver function.     

Molecular basis of factor VIII inhibition by human antibodies. Antibodies that bind to the factor VIII light chain prevent the interaction of factor VIII with phospholipid.

Most antibodies to factor VIII have recently been shown to react with discrete regions of the factor VIII light chain (within the C2 domain) and/or the factor VIII heavy chain (within the amino-terminal segment of the A2 domain). The mechanism by which these antibodies, usually designated "factor VIII inhibitors," interfere with factor VIII function has been examined by determining their effect on factor VIII binding to a phospholipid. Factor VIII-phosphatidylserine binding was prevented by all seven factor VIII inhibitors that had strong factor VIII light chain reactivity and reduced by two inhibitors with weak anti-light chain reactivity. None of four inhibitors with heavy chain reactivity prevented factor VIII-phosphatidylserine interaction, though a partial reduction (less than 50%) was noted for the intact IgG preparations. However, when Fab' fragments were substituted, no detectable reduction in factor VIII-phosphatidylserine binding was noted for the anti-heavy chain inhibitors and complete inhibition was retained by the anti-light chain inhibitors. These data suggest that a subset of factor VIII inhibitors, those that bind to light chain determinants, inactivate factor VIII by preventing its effective interaction with phospholipid.     

Evidence that production of platelet fibrinogen is synchronous with platelet production in the turpentine-induced acute phase response

Fibrinogen is synthesized by both hepatocytes and megakaryocytes, and available evidence indicates that cellular production of this protein is controlled by a single gene. We evaluated platelet and plasma fibrinogen production in rats injected with turpentine to induce the acute phase response. After turpentine injection, plasma fibrinogen levels, as expected, rose to more than double the baseline values within 48 hours and then declined to the upper limit of the normal range in 6 days. When levels were expressed as a percentage of total plasma protein, a rise form 4.2% to 10.0% occurred. Thrombocytosis was not observed until day 4; platelet counts returned to the normal range by day 8. The mean fibrinogen content per platelet did not change significantly (P greater than 0.05) during this period of observation. Thus, in contrast to the abrupt rise in the level of plasma fibrinogen, changes in the blood level of platelet fibrinogen paralleled changes in the platelet count. We interpret these findings to indicate that platelet fibrinogen production is synchronized with platelet production by megakaryocytes and thus suggest that fibrinogen production is regulated differently in these cells than it is in hepatocytes.     

Prediction of factor VIII inhibitor development in the SIPPET cohort by mutational analysis and factor VIII antigen measurement

Essentials

• A residual factor VIII synthesis is likely to be protective towards inhibitor (INH) development.
• Mutation type‐inhibitor risk association was explored in 231 patients with severe hemophilia A.
• A 2‐fold increase in INH development for in silico null vs. non‐null mutations was found.
• A 3.5‐fold increase in INH risk for antigen negative vs. antigen positive mutations was found.

Summary

Background

The type of F8 mutation is the main predictor of inhibitor development in patients with severe hemophilia A. Mutations expected to allow residual synthesis of factor VIII are likely to play a protective role against alloantibody development by inducing immune tolerance. According to the expected full or partial impairment of FVIII synthesis, F8 variants are commonly classified as null and non‐null.

Objectives

To explore the mutation type–inhibitor risk association in a cohort of 231 patients with severe hemophilia A enrolled in the Survey of Inhibitors in Plasma‐Product Exposed Toddlers (SIPPET) randomized trial.

Methods

The genetic defects in these patients, consisting of inversions of intron 22 (n = 110) and intron 1 (n = 6), large deletions (n = 16), and nonsense (n = 38), frameshift (n = 28), missense (n = 19) and splicing (n = 14) variants, of which 34 have been previously unreported, were reclassified according to two additional criteria: the functional effects of missense and splicing alterations as predicted by multiple in silico analyses, and the levels of FVIII antigen in patient plasma.

Results

A two‐fold increase in inhibitor development for in silico null mutations as compared with in silico non‐null mutations (hazard ratio [HR] 2.08, 95% confidence interval [CI] 0.84–5.17) and a 3.5‐fold increase in inhibitor development for antigen‐negative mutations as compared with antigen‐positive mutations (HR 3.61, 95% CI 0.89–14.74] were found.

Conclusions

Our findings confirm an association between the synthesis of minute amounts of FVIII and inhibitor protection, and underline the importance of investigating the residual FVIII antigen levels associated with causative variants in order to understand their clinical relevance.

     

Response to fibrinolytic activity and factor VIII-related antigen to stimulation with desmopressin in hyperlipoproteinemia

Impairment of fibrinolysis is supposed to contribute to CVD. In 38 hyperlipoproteinemic patients, known to be at risk for early CVD, fibrinolytic activity was measured before and after stimulation with DDAVP. A negative correlation was found between serum triglyceride levels and fibrinolytic activity, both before and after DDAVP. A subnormal activity was invariably found when serum triglyceride concentration was above 8 mmol/L. The defect can be attributed to low levels of extrinsic plasminogen activator. High cholesterol levels were not associated with impairment of fibrinolysis. Fibrinolytic activity and response to DDAVP were lowest in those patients with hypertriglyceridemia who also had a tendency to develop hyperchylomicronemia. (type V/IV). The low fibrinolytic activity in this type of hyperlipoproteinemia cannot be explained by obesity. Factor VIII was higher than normal in most patients with hyperlipoproteinemia; the level increased after stimulation with DDAVP in every patient. This imbalance between coagulation and fibrinolysis might increase the risk of CVD.     

Evidence that endotoxin is the cyclic 3':5'-GMP--promoting factor in erythropoietin preparations.

Since Ep preparations are contaminated with endotoxin, the possibility that the latter might be the factor in crude Ep which increases cGMP levels in rat fetal liver cells was examined. Endotoxin produced a striking elevation of cGMP in rat fetal liver cells without affecting cAMP levels or heme synthesis. Absorption with Limulus lysate of more than 99% of the endotoxin in a crude Ep preparation caused a parallel decrease in the cGMP-promoting activity without reduction of heme synthetic potency. It is concluded that endotoxin is the component of crude Ep which increases cGMP levels in rat fetal liver. The precise role of elevated cGMP in the action of endotoxin on cells and the universality of this effect remain to be determined.     

Time course of changes in in vitro platelet function and plasma von willebrand factor activity (VIIIR:WF) and factor VIII-related antigen (VIIIR:Ag) in the diabetic rat

The relationship between platelet abnormalities and vessel wall changes in diabetes is not known. We have examined the time course of alterations in in vitro platelet function and endothelial damage, as assessed by measurement of plasma levels of von Willebrand factor (VIIIR:WF) and factor VIII-related antigen (VIIIR:Ag), in streptozotocin-induced diabetic rats. Platelet aggregation and the platelet release reaction in response to ADP, thrombin, and collagen were measured in suspensions of washed platelets prepared from rats 3, 7, 14, or 28 days after induction of diabetes and in control animals. Platelets from diabetic animals showed enhanced aggregation response to ADP as early as 3 days after induction of diabetes and became hyperresponsive to thrombin after 7 days, compared to control platelets. Thrombin-induced release of serotonin was greater in platelets from diabetic animals at 14 days. Collagen-induced responses were not different at any time studied. VIIIR:WF was determined by ristocetin-induced platelet agglutination time in gel-filtered platelets, and VIIIR:Ag was determined by immunoelectrophoretic technique. VIIIR:WF and VIIIR:Ag were significantly enhanced in plasma from rats at 28 days after induction of diabetes and VIIIR:Ag was enhanced in plasma from rats at 14 days after induction of diabetes, but at the earlier times studied, neither were different from values in plasma from control-treated rats. Changes in VIIIR:WF and VIIIR:Ag therefore occurred later than the changes in platelet function. Plasma cholesterol concentrations were not significantly different at any of the times studied, but plasma triglyceride concentrations were significantly increased at 3 days and remained increased with further durations of diabetes. This may have contributed to the observed platelet and vessel wall changes. If these in vitro alterations reflect in vivo behavior, then platelet alterations occur before vessel wall changes and therefore do not appear to be a consequence of such changes in experimental diabetes mellitus.     

Subunit structure of factor VIII antigen synthesized by cultured human endothelial cells.

Cultured human endothelial cells were labeled with (3H)leucine, and the radioactive Factor VIII antigen present in the postculture medium was isolated by double anitbody immunoprecipitation and characterized by sodium dodecyl sulfate-polyacrylamide gel electrophoresis after reduction with dithiothreitol. The Factor VIII antigen synthesized by cultured endothelial cells was found to contain the same single polypeptide subunit (mol wt 225,000) present in plasma Factor VIII antigen. These results suggest that in vivo, the endothelial cell is a major site of synthesis of circulating Factor VIII antigen.     

Immunofluorescent localization of factor VIII-related antigen, fibrinogen, and several other plasma proteins in hemostatic plugs in humans.

Factor VIII-related antigen and its low ionic strength subunits were demonstrated in endothelium and hemostatic plugs obtained from punch biopsies of bleeding time wounds according to Mielke. No IgG, IgA, IgM, C4-C3, albumin, or prothrombin was demonstrated in the cryostat sections of hemostatic plugs. Antifibrinogen stained the hemostatic plug and fibrin fibers along the edges of the skin wound. Antiplatelet actomyosin stained the hemostatic plug and the endothelium and pericytes of blood vessels. These results suggest that factor VIII is present in the hemostatic plug not by trapping of plasma but by close association with or presence within the blood platelets. The positive staining of a hemostatic plug in which most platelets have undergone the release reaction indicates that antifactor VIII may be useful for the detection of platelet thrombi in tissue of patients with suspected diffuse intravascular coagulation.     

Evidence that serum calcium oxalate supersaturation is a consequence of oxalate retention in patients with chronic renal failure.

Serum oxalate rises in uremia because of decreased renal clearance, and crystals of calcium oxalate occur in the tissues of uremic patients. Crystal formation suggests that either uremic serum is supersaturated with calcium oxalate, or local oxalate production or accumulation causes regional supersaturation. To test the first alternative, we ultrafiltered uremic serum and measured supersaturation with two different methods previously used to study supersaturation in urine. First, the relative saturation ratio (RSR), the ratio of the dissolved calcium oxalate complex to the thermodynamic calcium oxalate solubility product, was estimated for 11 uremic (before and after dialysis) and 4 normal serum samples using a computer program. Mean ultrafiltrate oxalate predialysis was 89 +/- 8 microM/liter (+/- SEM), 31 +/- 4 postdialysis, and 10 +/- 3 in normals. Mean RSR was 1.7 +/- 0.1 (predialysis), 0.7 +/- 0.1 (postdialysis), and 0.2 +/- 0.1 (normal), where values greater than 1 denote supersaturation, less than 1, undersaturation. Second, the concentration product ratio (CPR), the ratio of the measured calcium oxalate concentration product before to that after incubation of the sample with calcium oxalate monohydrate crystal, was measured in seven uremic and seven normal serum ultrafiltrates. Mean oxalate was 91 +/- 11 (uremic) and 8 +/- 3 (normal). Mean CPR was 1.4 +/- 0.2 (uremic) and 0.2 +/- 0.1 (normal). Predialysis, 17 of 18 uremic ultrafiltrates were supersaturated with respect to calcium oxalate. The degree of supersaturation was correlated with ultrafiltrate oxalate (RSR, r = 0.99, r = 29, P less than 0.001; CPR, r = 0.75, n = 11, P less than 0.001). A value of ultrafiltrate oxalate of 50 microM/liter separated undersaturated from supersaturated samples and occurred at a creatinine of approximately 9.0 mg/dl.     

Evidence from oocyte expression that the erythrocyte water channel is distinct from band 3 and the glucose transporter.

It has been proposed that the mercurial-sensitive water transporter in mammalian erythrocytes is the anion exchanger band 3 (AE1) and/or the glucose transporter, band 4.5 (GLUT1). Using a functional assay for water channel expression in Xenopus oocytes (Zhang, R., K. A. Logee, and A. S. Verkman. 1990. J. Biol. Chem. 265:15375-15378), we compared osmotic water permeability (Pf) of oocytes injected with water, reticulocyte mRNA, AE1 mRNA, and GLUT1 mRNA. Injection of oocytes with 5-50 ng of in vitro-transcribed AE1 mRNA had no effect on Pf, but increased trans-stimulated 36Cl uptake greater than fourfold in a dinitro-disulfonic stilbene (DNDS)-inhibitable manner. Injection with 1-50 ng of in vitro-transcribed GLUT1 mRNA increased 3H-methylglucose uptake greater than 15-fold in a cytochalasin B-sensitive manner and increased Pf from (3.7 +/- 0.4) x 10(-4) cm/s (SE, n = 16, 10 degrees C) in water-injected oocytes up to (13 +/- 1) x 10(-4) cm/s (n = 18). Both the increments in sugar and water transport were inhibited by cytochalasin B (25 microM) and phloretin (0.2 mM); neither was inhibited by 0.3 mM HgCl2. In oocytes injected with 50 ng of rabbit reticulocyte mRNA, the Pf of (18 +/- 2) x 10(-4) cm/s (n = 18) was reduced to (4.0 +/- 0.6) x 10(-4) cm/s (n = 10) by HgCl2, but was not inhibited by DNDS (0.4 mM), cytochalasin B or phloretin. Coinjection of reticulocyte mRNA with antisense oligodeoxyribonucleotides against AE1 or GLUT1 did not affect Pf, but inhibited completely the incremental uptake of 36Cl or 3H-methylglucose, respectively. Expression of size-fractionated mRNA from reticulocyte gave a 2-2.5-kb size for water channel mRNA, less than the 4-4.5-kb size for the Cl transporter. These results provide evidence that facilitated water transport in erythrocytes is mediated not by bands 3 or 4.5, but by distinct water transport protein(s).     

Bernard-Soulier syndrome: a new platelet glycoprotein abnormality. Its relationship with platelet adhesion to subendothelium and with the factor VIII von Willebrand protein.

A decreased platelet adhesion to rabbit aorta subendothelium (Baumgartner technique) is confirmed in the Bernard Soulier (giant platelet) syndrome. Electron microscope techniques using a purified antibody against Factor VIII/von Willebrand protein, revealed an apparently normal presence of the Factor VIII/von Willebrand protein on the Bernard Soulier platelets. Electrophoretic characterization of the major protein and glycoprotein components of the Bernard Soulier platelets following sodium dodecyl sulfate solubilization indicated a relatively normal protein content but suggested a reduced content of the 155,000 molecular weight major platelet glycoprotein. This was confirmed by a reduced release of high molecular weight acidic glycopeptides following incubation of washed Bernard Soulier platelets with trypsin. It is proposed that this abnormality may be related to the previously reported reduced sialic acid content and the reduced electrophoretic mobility of the Bernard Soulier platelets and that a glycoprotein reduced or abnormal in the Bernard Soulier platelets is necessary for the normal adhesion of platelets to subendothelium.