Studies on immunological assay of vitamin-K dependent factors. III. A double monoclonal immunoradiometric assay for factor IX antigen.

Two-site immunoradiometric assays (IRMAs) for factor IX antigen (IX:Ag) were developed using a monoclonal antibody (RFF-IX/1) on the solid-phase and either another monoclonal antibody (RFF-IX/4) or a human polyclonal inhibitor antiserum as tracer (M-M and M-I IRMA respectively). The lower sensitivity limits of these two assays for IX:Ag in normal reference plasma were 4 X 10(-4) (M-M IRMA) and 2 X 10(-4) (M-I IRMA) units/ml. In 20 samples of normal plasma, levels of factor IX coagulation activity (IX:C) and of factor IX antigen measured by both IRMAs were highly correlated. Mean values of approximately 1.0 units/ml were obtained in all three assays. In normal serum, IX:Ag levels were lower with means of 0.84 (M-M IRMA) and 0.83 (M-I IRMA) units/ml. 4/25 patients with haemophilia B were CRM neg., two were CRM + and the remaining 19 patients were CRMr variants. In two of these, IX:Ag was detectable by M-I IRMA whilst IX:C and IX:Ag measured by M-M IRMA were undetectable. In plasma from a fetus subsequently terminated on eugenic grounds, IX:C and IX:Ag by both M-M and M-I IRMA were undetectable. In warfarin-treated plasma (n = 12), the level of IX:C was low (mean 0.39 units/ml). The levels of IX:Ag measured by M-M IRMA (mean of 0.80 units/ml) and by M-I IRMA (0.70 units/ml) showed a discrepancy. M-M IRMA reflects the real amount of IX:Ag in warfarinized plasma because both monoclonal antibodies bind to epitopes distant from the light chain carboxylated region. Western blotting of denatured factor IX demonstrated that RFF-IX/1 binds an epitope that is lost after XIa activation. RFF-IX/4 binds the heavy chain. Antigen measured after activation but without denaturing showed loss of 60% reactivity after XIa activation but no change after RVV activation. These data indicate a binding site for RFF-IX/1 within the activation peptide (residues 146-180).