Synthesis of antihemophilic factor antigen by cultured human endothelial cells

Antihemophilic factor (AHF, Factor VIII) antigen has been demonstrated in cultured human endothelial cells by immunofluorescence studies using monospecific rabbit antibody to human AHF. Control studies with cultured human smooth muscle cells and human fibroblasts were negative. By radioimmunoassay it was demonstrated that cultured human endothelial cells contain AHF antigen which is released into the culture medium. Cultured smooth muscle cells and fibroblasts did not have this property. Cultured endothelial cells incorporated radioactive amino acids into high molecular weight, AHF antigen-rich protein fractions prepared from the culture media, 7% of the radioactive amino acid counts incorporated into this material were precipitated by globulin prepared from rabbit anti-AHF whereas normal rabbit globulin precipitated only 1.5% of the counts. Although cultured endothelial cells actively synthesize AHF antigen, AHF procoagulant activity was not detected in the culture medium. Studies seeking a basis for the lack of procoagulant activity have not clarified this deficiency, but they have established that exogenous AHF procoagulant activity is not inactivated by the tissue culture system.

Quantitative assay of a plasma factor deficient in von Willebrand's disease that is necessary for platelet aggregation. Relationship to factor VIII procoagulant activity and antigen content

In a previous paper, we showed that the abnormality of ristocetin-induced platelet aggregation in platelet-rich plasma in 10 patients with von Willebrand's disease could be corrected by a factor in normal plasma that was present in the same fractions as factor VIII procoagulant activity (antihemophilic factor, AHF, VIII(AHF)) when prepared by chromatography on Bio-Gel 5 M (Bio-Rad Laboratories, Richmond, Calif.). This observation suggests that patients with this disorder are deficient in a plasma factor, associated with the factor VIII molecule, that is necessary for normal platelet function. In the present paper, we describe, an assay for this factor, the von Willebrand factor (VIII(VWF)), based on the observation that a log-log relationship exists between the amount of ristocetin-induced aggregation of washed, normal platelets and the concentration of normal plasma present in the test system. We assayed the activity of VIII(VWF) as well as antihemophilic factor procoagulant activity (VIII(AHF)) and factor VIII antigen (VIII(AGN)) in 15 patients with von Willebrand's disease and 20 normal subjects. A highly significant correlation (r approximately 0.80) between VIII(VWF) and both VIII(AHF) was found in normal subjects and in patients with von Willebrand's disease. This finding, in addition to the observation that agarose gel chromatography fractions that have VIII(AHF) procoagulant activity also have VIII(VWF) activity, strongly suggests that the von Willebrand factor is associated with the factor VIII molecule. VIII(VWF) in normal plasma was not inhibited by human anti-VIII, and VIII(VWF) levels were normal in hemophilic plasma. Thus, the VIII(VWF) site on the factor VIII molecule appears to be different from that determining VIII(AHF). Finally, the activity of VIII(VWF) appeared to correlate better with the bleeding time than either VIII(AHF) or VIII(AGN). This suggests that VIII(VWF) assayed in this study may be the "anti-bleeding factor" that is deficient in von Willebrand's disease. These findings are consistent with a decreased synthesis of the factor VIII molecule in von Willebrand's disease and suggest the possibility of additional abnormalities of the site on the molecule that determines the activity of VIII(VWF).

Tissue factor activity in lymphocyte cultures from normal individuals and patients with hemophilia A

The procoagulant material of lymphocytes has been characterized as tissue factor. Lymphocytes stimulated with phytohemagglutinin or the purified protein derivative of the tubercle bacillus developed procoagulant activity with incubation in tissue culture. While this material corrected the prolonged clotting time of factor VIII (AHF) deficient plasma, we have shown, utilizing a sensitive radioimmunoassay, that no AHF antigen was present in the cell cultures. Further, we have demonstrated this material to be tissue factor by coagulation techniques and immunological cross-reactivity. The published data regarding factor VIII synthesis is reviewed in light of these observations and comments are made regarding the role of the lymphocyte procoagulant.

The significance of erythrocyte antigen site density. II. Hemolysis

The importance of antigen site density has been studied by means of a model passive hemolysis system using red cells coupled with sulfanilic acid groups. Relative site numbers were estimated from the covalent linkage of sulfanilic acid-(35)S to red cell membrane protein, and the effective antigen site number was determined with (125)I-labeled rabbit IgG anti-sulfanilic acid (anti-S). Immune hemolysis was demonstrated for red cells which had greater than a threshold number of antigen sites, the value of which was different for normal human cells (80,000 sites/cell), cells from a patient with paroxysmal nocturnal hemoglobinuria (PNH) (40,000 sites/cell), and sheep red blood cells (RBC) (15,000 sites/cell). Cells with antigen site densities below these values did not hemolyze when tested with 1 mg/ml purified rabbit IgM anti-S. 2-8 times greater antigen site densities were required to obtain hemolysis with IgG anti-S. Above the threshold value, hemolysis titers were proportional to the antigen site number until maximal values were obtained. The greater hemolytic efficiency of IgM antibody was demonstrated in this system, and it was established that the magnitude of the difference was related to the test cell antigen site density. These data, taken with previously reported hemagglutination studies, have been used to develop a general classification of immune hemolysis and hemagglutination based on antigen site density and antibody class. It is suggested that the heterogeneity of blood group systems is caused by differences in the site separation of erythrocyte membrane antigens.

The significance of erythrocyte antigen site density. I. Hemagglutination

The importance of antigen site density has been studied by means of a model passive hemagglutination system using human red cells coupled with sulfanilic acid groups. Relative site numbers were estimated from the covalent linkage of sulfanilic acid-(35)S to red cell membrane protein and the effective antigen site number was determined with (125)I-labeled rabbit IgG antisulfanilic acid. Cells which had fewer than 20,000 antigen sites per cell were not agglutinated. As greater numbers of sulfanilic groups were coupled to the red cells, the agglutination titers increased to maximum values with red fanilic groups were coupled to the red cells, the agglutination titers of purified IgM antibody were 10-20 times greater than IgG antibody when preparations with the same protein concentration were compared, but this difference was not noted when IgG antibody was measured by antiglobulin reactions. These findings emphasize the need to consider differences in antigen site density when comparing blood group systems. They are consistent with the hypothesis that those blood group antigens which have a very low site number will not be detected by IgG antibodies in saline hemagglutination determinations.

Heterogeneity of rabbit IgM antibody as detected by C'1a fixation

The C'1a-fixing properties of purified rabbit IgM anti-benzenearsonate antibody were determined. When tested with sheep erythrocytes to which hapten had been coupled by diazo linkage, the number of C'1a molecules fixed was 21% of the number of IgM antibody molecules bound to the erythrocyte surface. This was not due to loss of C'1a-fixing capacity during the purification procedure. Preparative electrophoresis of the antibody concentrated C'1a-fixing molecules in the anodal region so that antibody fractions with greater C'1a-fixing capacity were obtained. The demonstration that C'1a fixation is a property of a subpopulation of IgM molecules provides evidence for previously unrecognized micro-chain heterogeneity.