Regulation of factor V and factor V-short by TFPIα: Relationship between B-domain proteolysis and binding

Figures

• 

  Figure 1. FV species and quality control.A, schematic representation of FV species used in this study. The B-domain, defined by residues 710 to 1545 lies between the HC and LC. FV-1033 has B-domain residues 1034 to 1491 deleted and FV-short has residues 756 to 1458 deleted. Thrombin cleavage sites are indicated as are AR1, AR2, and the BR. B, purified FV species (300 nM) before (−) and after (+) treatment with thrombin (2 nM) or RVV-V (2 nM for FV-short-QR), were subject to SDS-PAGE (4–12% gradient) under reducing conditions and stained with Coomassie Blue R-250. Bands corresponding to the HC and LC are indicated on the right. FV mutants in which specific thrombin cleavage sites were changed from Arg to Gln are indicated.

• 

  Figure 2. FV-short is constitutively active like FVa.A, thrombin generation was measured in FV-DP supplemented with 1.5 nM FV-short, FV-short-QQ, or FVa and initiated with 0.4 pM TF. The data show the average of three independent experiments with standard deviation. B, prothrombin activation by prothrombinase assembled with FVa or different FV-short mutant. The (∗) refers to FV-short mutants precleaved prior to introduction into the assay. The initial velocity of thrombin generation was determined at increasing concentrations of prothrombin and data were fit using the Michaelis–Menten equation using GraphPad Prism (v.9.0). For the data in B, the fitted constants and the standard errors are: FVa, KM = 0.48 ± 0.04 μM, Vmax = 204.0 ± 6.0 nM thrombin/min; FV-short, KM = 0.31 ± 0.02 μM, Vmax = 173.4 ± 4.2 nM thrombin/min; FV-short-QQ, KM = 0.34 ± 0.03 μM, Vmax = 172.4 ± 4.5 nM thrombin/min; FV-short-RQ∗, KM = 0.49 ± 0.03 μM, Vmax = 175.3 ± 4.0 nM thrombin/min; and FV-short-QR∗, KM = 0.46 ± 0.04 μM, Vmax = 222.7 ± 7.3 nM thrombin/min. The data are representative of 3 to 6 independent experiments. The average value for the kinetic constants and standard deviations are summarized in Table 2.

• 

  Figure 3. TFPIα-BR regulates thrombin generation initiated with FV-short. Thrombin generation was measured in FV-DP supplemented with 1.5 nM FVa (A–B), FV-short (C and D), or FV-short-QQ (E and F). A, C and E are thrombin generation curves with increasing amounts of TFPIα-BR initiated with 0.4 pM TF. B, D, F and G represent the change in thrombin peak height at different TFPIα-BR concentrations initiated 0.4 pM TF, while H employs different initiators. Cofactor species used in the assay are indicated on the graph. The (∗) refers to FV-short mutants precleaved prior to introduction into the assay. For A, C, and E, data are a single representative experiment that was performed three times. For B, D, F, G, and H, the data show the average thrombin peak with standard deviation from three independent experiments. In these graphs, the data were fit using a two-phase exponential decay model.

• 

  Figure 4. TFPIα-BR and TFPIα inhibits prothrombin activation using FV-short but not FVa. The initial velocity of thrombin generation was determined at increasing concentrations of TFPIα-BR (A) or TFPIα (B). Prothrombinase was assembled using limiting amounts of the cofactor species. In B, an excess of an anti-K2 antibody to block FXa inhibition was included in the assay. The (∗) refers to FV-short mutants precleaved prior to introduction into the assay. The data were fit to the equation for competitive inhibition using GraphPad Prism (v. 9.0). The calculated inhibition constants and the standard errors are: for A) FV-short, Ki = 19.8 ± 0.1 nM; FV-short-QQ, Ki = 21.5 ± 0.1 nM; FV-short-RQ∗, Ki = 86.1 ± 14.1 nM; for data in B) FV-short, Ki = 0.14 ± 0.03 nM; FV-short-QQ, Ki = 0.10 ± 0.02 nM; FV-short-RQ∗, Ki = 0.26 ± 0.03 nM. No inhibition constant could be calculated for FVa or FV-short-QR∗. The data are representative of 3 to 5 independent experiments. The average value for the inhibition constants and standard deviations are summarized in Table 2.

• 

  Figure 5. TFPIα binds FV-short but not FV. FV species (A) and TPFIα (B) alone or premixed at 1 μM each (C–H) were loaded onto a Sephacryl S-200 column in HBS/Ca2+, pH 7.4. Collected fractions were assessed for TFPIα by ELISA (right y-axis; black star symbol and dashed black line) or total protein as measured by absorbance at 280 nm (left y-axis). FV species used in the assay are indicated on the graph. The (∗) refers to FV-short mutants precleaved prior to introduction into the assay. The data are representative of at least two independent experimental runs.

• 

  Figure 6. TFPIα-BR binds to FV-short with high affinity.A and C, reaction mixtures containing 20 nM OG488-TFPIα-BR and 50 μM PCPS in assay buffer were titrated with increasing concentration of FV species. The (∗) refers to FV-short mutants precleaved prior to introduction into the assay. The lines are drawn following analysis to independent, noninteracting sites. B, unlabeled TFPIα-BR or TFPIα was titrated into reaction mixtures containing 25 nM OG488-TFPIα-BR, 20 nM FV-short-QQ, and 50 μM PCPS. Equilibrium binding constant was determined assuming a stoichiometry of 1 mol of FV-short-QQ/mol of BR peptide. Data are representative of 2 to 4 similar experiments.

• 

  Figure 7. Cleavage of FV facilitates binding to TFPIα-BR. FV species (40 nM each; A, D, G, J and M) were added to a cuvette containing 50 μM PCPS and 40 nM OG488-TFPIα-BR in assay buffer. In A, D, G, J and M, the steady-state fluorescence anisotropy change was measured over time following the addition of thrombin. At the same time intervals, aliquots were collected over time and subjected to SDS-PAGE/western blotting on 3 to 8% gradient gels under reducing conditions, and the FV fragments were visualized using fluorescently labeled antibodies directed against the FV-LC (green) and the FV-HC (red) (B, E, H, K and N). FV species corresponding to each band are schematically shown on the left of the blots: HC (red bar), B-domain (gray bar), LC (green bar), and BR (blue bar). Symbols to the right of the blot (B, E, H, K, N) representing FV fragments were plotted as normalized intensity on the graphs to the right and match in color and shape (C, F, I, L, O). The data are representative of 2 to 3 similar experiments.

• 

  Figure 8. The endogenous FV BR and TFPIα-BR delay FV cleavage by thrombin. Twenty nanometer PD-FV (A) or FV-short in the absence (B) or in the presence (C) of 1 μM TFPIα-BR in assay buffer was incubated with thrombin (0.2 nM). Samples collected over time (indicated on blots in minutes) were subject to SDS-PAGE/western blotting on 3 to 8% gradient gels under reducing conditions and visualized using a mixture of fluorescently labeled antibodies against the FV-LC (green) and FV-HC (red). Species corresponding to each band are drawn on the left side: HC (red bar), B-domain (gray bar), BR (blue bar). LC (green bar). D, normalized intensity of the formation of the LC for PD-FV or FV-short in the presence or absence of TFPIα-BR was plotted as function of time. Data show the average of three independent experiments with standard deviation.

• 

  Figure 9. Properties of different forms of FV and regulation by TFPIα.A, FV is a procofactor and is maintained in an inactive state by the concerted action of the BR and AR2. These functional landmarks likely interact with AR1 and block the ability of FXa to bind on the HC/LC region. Due to the unavailability of AR2, TFPIα does not bind FV and is not regulated by it directly. B, cleavage of FV at Arg709 by thrombin or FXa has little to no influence on weakening the interaction of BR with AR2. C, further cleavage at Arg1018 disrupts BR binding and generates a FV species (FVAR2) that has cofactor-like activity, which can bind TFPIα and be regulated by it. D, cleavage at Arg1545 to generate FVa, releases AR2, and eliminates TFPIα binding. E, since FV-short lacks the BR yet retains AR2, it is cofactor-like FVa, but can bind and be regulated by TFPIα.