Molecular interactions between FX, SR-AI, and PTX2. Representative BLI graphs (Octet) of molecular binding between FX, SR-AI, and PTX2. Ten micrograms per milliliter of recombinant SR-AI (A-B) or purified PTX2 (C) were coated on the biosensor, and association with increasing concentrations (0.1 to 5 µM) of FX (A,C) or increasing concentrations (0.05 to 1 µM) of purified PTX2 (B) were measured over an association phase of 600 seconds. Results are expressed as the wavelength shift (in nm) generated by the binding of the different molecules. Apparent K_D values were calculated using the Michaelis-Menten equation (D). (E) Coimmunoprecipations of FX and PTX2 were performed in normal human plasma (lanes 1 and 3) and FX-deficient plasma (lanes 2, 4, and 5) supplemented or not with 10 µg/mL of FX (lane 5). Immunoprecipitates were analyzed by western blot using anti-FX and anti-PTX2 antibodies. (F) Coimmunoprecipitations of PTX2 were performed in FX-deficient plasma supplemented with 10 µg/mL of FX, FX_N181A mutant, or activated FXa. Immunoprecipitates or purified proteins (10 ng per lane) were analyzed with an anti-FX antibody by western blot. (G) FX (100 nM) preassociated (blue circles) or not (red circles) with 1 µM of PTX2 was incubated with Russel’s Viper Venom X (RVV-X) enzyme (left panel) or tissue factor/FVIIa (TF/FVIIa) and phospholipids (middle panel). Kinetic conversion of FX into FXa was monitored during a 30-minute (RVV-X) or 60-minute (TF/FVIIa) time course. Right panel represents effect of different PTX2 concentrations (0-52.5 μg/mL) on FX activation by the FIXa/FVIIIa complex in the presence of phospholipid vesicles. Concentrations were 42 nM FX, 20 pM FIXa, and 5 pM thrombin-activated FVIII, and FXa generation was allowed for 5 minutes. Graphs represent the mean ± SD of 3 different experiments.