Proteolytic properties of single-chain factor XII: a mechanism for triggering contact activation

Abstract

When blood is exposed to variety of artificial surfaces and biologic substances, the plasma proteins factor XII (FXII) and prekallikrein undergo reciprocal proteolytic conversion to the proteases αFXIIa and α-kallikrein by a process called contact activation. These enzymes contribute to host-defense responses including coagulation, inflammation, and fibrinolysis. The initiating event in contact activation is debated. To test the hypothesis that single-chain FXII expresses activity that could initiate contact activation, we prepared human FXII variants lacking the Arg353 cleavage site required for conversion to αFXIIa (FXII-R353A), or lacking the 3 known cleavage sites at Arg334, Arg343, and Arg353 (FXII-T, for "triple" mutant), and compared their properties to wild-type αFXIIa. In the absence of a surface, FXII-R353A and FXII-T activate prekallikrein and cleave the tripeptide S-2302, demonstrating proteolytic activity. The activity is several orders of magnitude weaker than that of αFXIIa. Polyphosphate, an inducer of contact activation, enhances PK activation by FXII-T, and facilitates FXII-T activation of FXII and FXI. In plasma, FXII-T and FXII-R353A, but not FXII lacking the active site serine residue (FXII-S544A), shortened the clotting time of FXII-deficient plasma and enhanced thrombin generation in a surface-dependent manner. The effect was not as strong as for wild-type FXII. Our results support a model for induction of contact activation in which activity intrinsic to single-chain FXII initiates αFXIIa and α-kallikrein formation on a surface. αFXIIa, with support from α-kallikrein, subsequently accelerates contact activation and is responsible for the full procoagulant activity of FXII.

Figure 1.

Recombinant contact factors. (A) Schematic diagrams of contact factors showing noncatalytic (white boxes) and catalytic (shaded boxes) domains. Positions of active site serine residues are indicated by black bars. Sites of proteolysis during activation are indicated by arrows, with black arrows indicating sites of cleavage required for full protease activity. FXII is an 80-kDa polypeptide that may be cleaved at 3 locations. Cleavage after Arg353 converts FXII to αFXIIa. Cleavage of αFXIIa after Arg334 separates the noncatalytic and catalytic domains, forming βFXIIa. The importance of cleavage after Arg343 is not clear. The FXII noncatalytic domains are the fibronectin type 2 (F2), epidermal growth factor (EGF), fibronectin type 1 (F1), and kringle (K) domains, and a proline-rich region (PRR). PK is a 93-kDa polypeptide that is cleaved after Arg371 to form α-kallikrein (α-Kal). A second cleavage after Arg140 produces β-kallikrein (β-Kal). FXI is a homodimer of 80-kDa polypeptides. It is converted to FXIa by cleavage after Arg369. The noncatalytic portions of PK and FXI contain 4 apple domains, designated A1 to A4. (B) Coomassie blue–stained nonreducing SDS-PAGE of purified FXII (left panel) and PK (right panel) containing ∼2-μg samples per lane. Positions of molecular mass standards in kilodaltons are shown to the left of the images.

Figure 2.

FXII autoactivation and activation by kallikrein. (A) Plasma FXII (200 nM) was incubated in standard buffer without Poly-P (−Poly-P) or in the presence of 70 μM Poly-P (+Poly-P). At the indicated times, samples were removed from the reaction into reducing sample buffer, size fractionated on a 12% polyacrylamide-SDS gel and stained with Coomassie blue. Positions of standard for single-chain FXII and the heavy chain (HC) and light chain (LC) of αFXIIa are shown at the right of each panel. (B) Chromogenic substrate assay for FXIIa activity for the fractions shown in the gel in panel A. Reactions were run in the presence (△) or absence (♦) of 70 μM Poly-P. (C) Recombinant FXII species (200 nM) incubated in the absence of an activator (control, left column), in the presence of 70 μM Poly-P (center column), or in the presence of 50 nM α-kallikrein without a surface (right column). At the indicated times, samples were removed into reducing sample buffer. Samples were size fractionated by SDS-PAGE, followed by western blot analysis with a polyclonal anti-human FXII IgG. For panels A and C, positions of standards for FXII and the heavy chain (HC) and light chain (LC) of αFXIIa are indicated at the right of each image. (D) Recombinant FXII species (200 nM) were incubated in the presence of 70 μM Poly-P and 200 μM S-2302. Changes in OD 405 nm were continuously monitored on a microplate reader. (E) FXII (200 nM) was incubated with 50 nM α-kallikrein for 120 minutes at 37°C in the absence of a surface. Kallikrein was inhibited with IgG H03 and FXIIa cleavage of S-2302 (200 μM) was measured.

Figure 3.

Reciprocal activation of FXII and PK in the absence of a surface. (A-C) FXII (200 nM) and PK (200 nM) species were incubated at 37°C. At indicated time points, samples were removed into reducing (A,C) or nonreducing (B) sample buffer, size fractionated by SDS-PAGE, and analyzed by western blot using (A,C) polyclonal IgG to FXII (XII) or PK or (B) monoclonal IgGs that preferentially recognize the activated forms of FXIIa (D06) and kallikrein (H03). (A-B) Reciprocal activation of plasma FXII and PK. (C) Activation of PK-WT by recombinant FXII species. For panels A and C, positions of standards for FXII (XII) and the heavy chain (HC) and light chain (LC) of FXIIa; and standards for PK, the heavy chain and light chain of α-kallikrein, and a fragment of the heavy chain of β-kallikrein (β) are indicated at the right of each image. For panel B, positions of standards for αFXIIa, βFXIIa, α-kallikrein (α-kal), and β-kallikrein (β-kal) are indicated on the right.

Figure 4.

Cleavage of S-2302 and PK by FXII species. (A) FXII-WT (WT), FXII-R353A, FXII-T (T), or FXII-S544A (200 nM) were incubated with S-2302 (200 μM) in standard buffer in the absence of a surface. Continuous formation of amidolytic activity was monitored at 405 nm. (B-C) Varying concentrations of S-2302 (1.6-1000 μM) were incubated with (B) 5 nM αFXIIa or (C) 200 nM FXII-T in the absence of a surface. Changes in OD 405 nm/minute were measured on a plate reader and converted to pNA generated per minute. Each point is a mean ± 1 SD for 3 separate experiments. (D) Same as reactions described in panel A except that incubations were conducted in the presence of a surface (PTT-A reagent, 10% final volume). Note that PTT-A reagent only increases the amidolytic activity of FXII-WT, compared with reactions in panel A because, of the 4 FXII species tested, it is the only 1 that is cleaved after Arg353 and has an active site serine at residue 544. (E) PK-WT (200 nM) was incubated with 200 nM FXII-WT, FXII-R353A, FXII-T, FXII-S544A, or control vehicle (C). Generation of kallikrein (or kallikrein and αFXIIa in the case of FXII-WT) was continuously monitored by cleavage of S-2302 (200 μM). (F-G) Varying concentration of PK were incubated with (F) 25 pM αFXIIa or (G) 15 nM FXII-T in the absence of a surface. Kallikrein generation was determined by measuring the rate of S-2302 (200 μM) cleavage. Each point is a mean ± 1 SD for 3 separate experiments. (H) Generation of kallikrein from PK-WT (200 nM) was followed by continuous monitoring of S-2302 (200 μM) cleavage in the presence or absence of a surface (70 μM Poly-P), and in the presence or absence of 200 nM FXII-T (T). (I) Kinetic parameters for cleavage of S-2302 and PK by αFXIIa or FXII-T determined from the curves in panels B, C, F, and G.

Figure 5.

Activation of FXII and FXI by single-chain FXII. (A) Western blots of a mixture of FXII-T (200 nM) and FXII-S544A (200 nM) in the absence of a surface (No Surface), or in the presence of PTT-A reagent (25% final volume) or Poly-P (70 μM). At indicated times, samples were removed into nonreducing sample buffer, size fractionated by SDS-PAGE, and analyzed by western blot using IgG D06 which recognizes formation of the FXIIa active site. The bottom row shows results for FXII-S544A and FXII-T incubated separately with 70 μM Poly-P. (B) FXI-S557A (30 nM) was incubated with 200 nM FXII-WT, FXII-T, or FXII-S544A in the absence (top row) or presence (bottom row) of 70 μM Poly-P. At indicated times, samples were removed into reducing sample buffer, size fractionated by SDS-PAGE, and analyzed by western blot using a goat-anti-human FXI polyclonal IgG.

Figure 6.

Surface-initiated clotting and thrombin generation in human plasma. (A) Clotting times in an aPTT assay for pooled normal plasma (PNP), FXII-deficient plasma (FXII DP), or FXII-deficient plasma supplemented with recombinant FXII species (FXII added to FXII DP). Each symbol indicates 1 clotting time and the horizontal bars indicate means for each group ± 1 SD. (B) Thrombin generation in FXII-deficient plasma supplemented with PTT-A reagent in the absence of FXII (No XII dotted line) or in the presence of plasma FXII (dashed line) or FXII-WT (solid line). (C) Thrombin generation in FXII-deficient plasma supplemented with PTT-A reagent and FXII-WT (solid line), FXII-T (dashed line), or FXII-S544A (dotted line). (D) Reactions for FXII-WT and FXII-T shown in panel C run in the presence of CTI or anti-FXI IgG 01A6. For all panels, results represent means for 3 runs.

Figure 7.

Mouse carotid artery thrombosis model. FXII-deficient C57Bl/6 mice were infused with 100 μL of PBS (C) or 100 μl of PBS containing 25 μg of FXII-WT, FXII-R353A, FXII-T, or FXII-S544A. Thrombus formation was induced by application of 2 pads saturated with 5% FeCl₃ to opposite sides of the carotid artery for 3 minutes. Flow through the artery was recorded for 30 minutes. The percentages of animals with occluded arteries 30 minutes after FeCl₃ application are shown in the bar graph (n = 9 for each bar). Representative plasma samples from test mice were analyzed by western blot to make certain that FXII was still in the circulation at the end of the study. Each number indicates a separate animal. C, FXII control; DP, FXII-deficient mouse plasma; NP, normal mouse plasma.

Figure 8.

Comparison of tPA and FXII structures. Shown are stick diagrams of S1 pocket structures with hydrogen bonds and electrostatic interactions shown as dotted lines (purple). In panels A and B, the position of the oxyanion hole is indicated by the juxtaposed blue spheres that represent the nitrogen atoms of Ser195 and Gly193. (A) Single-chain tPA active S1 pocket crystal structure (pdb:1BDA) is shown with Asp194 stabilized by the salt bridge formed with Lys156 (indicated by + and − symbols), and also by hydrogen bonds with the main-chain nitrogens of Gly142 and Cys191. The cyan stick figure represents the side chain of the arginine P1 residue of the tPA inhibitor dansyl-Glu-Gly-Arg-chloromethylketone. (B) Homology model (SWISS-MODEL) of the S1 pocket of FXII-T based on the tPA crystal structure where Gln156 forms a hydrogen bond to the Asp194 carboxylate group. The side chain shown in cyan represents the P1 arginine of a substrate (PK or FXI). (C) Crystal structure (pdb:XDE) showing the inactive zymogen conformation of FXIIc where the oxyanion hole is absent (all figures prepared with PyMOL Molecular Graphics System, version 1.8; Schrödinger, LLC).