Amyloid β-Protein Precursor Regulates Factor VIIa and the Factor VIIa–Tissue Factor Complex

Thrombosis Research 99 (2000) 267–276

ORIGINAL ARTICLE

Protease Nexin-2/Amyloid ␤-Protein Precursor Regulates Factor VIIa and the Factor VIIa–Tissue Factor Complex Fakhri Mahdi1, Alnawaz Rehemtulla2, William E. Van Nostrand3, S. Paul Bajaj4, and Alvin H. Schmaier1 Department of Internal Medicine, Division of Hematology and Oncology, 2Department of Radiation Oncology, University of Michigan, Ann Arbor, MI 48109-0724; 3Department of Medicine, State University of New York at Stony Brook, Stony Brook, NY; 4Department of Internal Medicine, Division of Hematology/Oncology, Saint Louis University Medical School, Saint Louis, MO 63110-0250, USA. 1

(Received 10 January 2000 by Editor C.W. Francis; revised/accepted 20 March 2000)

Abstract Protease nexin-2/amyloid ␤-protein precursor (PN-2/A␤PP) and its Kunitz protease inhibitory (KPI) domain were characterized as inhibitors of factor VIIa (FVIIa) and factor VIIa–tissue factor complex (FVIIa–TF). PN-2/A␤PP and KPI domain inhibited FVIIa with an apparent Ki of 1.1⫾0.2⫻ 10⫺7 M and 1.5⫾0.1⫻10⫺7 M, respectively. When soluble tissue factor (TF1–219) was present, there was increased FVIIa inhibition by PN-2/A␤PP or KPI domain (Ki⫽7.8⫾0.3⫻10⫺8 M and 6.8⫾0.6⫻10⫺8 M, respectively). When relipidated tissue factor (TF1–243) was present, the Ki of FVIIa inhibition by PN-2/A␤PP increased 4.7-fold further (Ki⫽1.65⫻ 10⫺8 M). PN-2/A␤PP complexed with FVIIa, as shown on gel filtration and solid phase binding assay. The apparent second-order rate constant of Abbreviations: AT, antithrombin; FVIIa–TF,: factor VIIa–tissue factor complex; FX, factor X; Fxa, factor Xa; PN-2/A␤PP, protease nexin-2/amyloid ␤-protein precursor; KPI domain, Kunitz protease inhibitory domain of PN-2/A␤PP; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; TF1–219, recombinant soluble tissue factor containing amino acids 1–219; TF1–243, recombinant relipidated tissue factor containing amino acids 1–243. Corresponding author: Alvin H. Schmaier, University of Michigan, 5301 MSRB III, 1150 W. Medical Center Drive, Ann Arbor, MI 48109-0640, USA. Tel: ⫹1 (734) 647 3124; Fax: ⫹1 (734) 647 5669; E-mail: [email protected].

inhibition of FVIIa by PN-2/A␤PP in the absence of TF1–219 (k″⫽7.5⫾4.0⫻104 M⫺1min⫺1) was less than that of the FVIIa–TF1–219 complex (k″⫽1.6⫾ 0.2⫻105 M⫺1min⫺1). Antithrombin in the absence of TF1–219 also had a lower apparent second-order rate constant of inhibition (k″⫽1.8⫾0.3⫻103 M⫺1min⫺1) than in its presence (k″⫽1.6⫾0.3⫻105 M⫺1min⫺1). In a mixture that included FVIIa, relipidated TF1–243 and factor X, PN-2/A␤PP or KPI domain had an IC50 at 65 and 250 nM, respectively; antithrombin and heparin (1 U/mL) had an IC50 of 12.8 nM. These data indicate that tissue factor promoted the inhibition of FVIIa by PN-2/A␤PP or KPI domain, but antithrombin was a better inhibitor of soluble FVIIa–TF in extrinsic tenase.  2000 Elsevier Science Ltd. All rights reserved. Key Words: Protease nexin-2/amyloid ␤-protein precursor; Kunitz protease inhibitor; Factor VIIa–tissue factor; Antithrombin; Factor Xa

H

emostatic reactions are controlled by protease inhibitors. Antithrombin III (AT), a serpin, is the major plasma inhibitor regulating most hemostatic enzymes especially ␣-thrombin [1]. Two other serpins, heparin cofactor II and protease nexin-1, are additional regulators of ␣-thrombin [2,3]. Within the last decade, tissue factor pathway inhibitor (TFPI), a Kunitz

0049-3848/00 $–see front matter  2000 Elsevier Science Ltd. All rights reserved. PII S0049-3848(00)00245-0

268

F. Mahdi et al./Thrombosis Research 99 (2000) 267–276

type serine protease inhibitor, has been characterized as an important regulator of factor VIIa–tissue factor (FVIIa–TF) in complex with factor X (FX) or factor Xa (FXa) [4–6]. Antithrombin, in the presence of heparin, has been characterized as an inhibitor of FVIIa, and the presence of tissue factor potentiates its inhibition [7,8]. Protease nexin-2/ amyloid ␤-protein precursor (PN-2/A␤PP), which has 38% sequence homology to TFPI, is another Kunitz-type serine protease inhibitor of hemostatic enzymes. PN-2/A␤PP is highly enriched in platelets and lymphocytes, which, when liberated in plasma, can achieve local concentrations of 30 nM [9]. It is the most potent inhibitor of factor XIa [10,11], although the serpin, C1 inhibitor, probably accounts for the majority of factor XIa inhibition in plasma [12]. PN-2/A␤PP has been described as an inhibitor of factor IXa in the intrinsic tenase complex and an inhibitor of FXa in the prothrombinase complex [13–15]. In plasma, the isolated Kunitz protease inhibitory (KPI) domain of PN-2/A␤PP is an equally potent inhibitor of a factor VII coagulant assay as it is of a factor X coagulant assay [13]. Further, studies using phage display of the KPI domain of PN-2/A␤PP suggest that more potent inhibitors of factor VIIa (FVIIa) can be prepared [16,17]. With this background, we investigated the ability of wild-type PN-2/A␤PP and KPI domain to inhibit FVIIa and the FVIIa–TF complex alone or in a factor X activation assay, and compared their degree of inhibition with AT.

1. Materials and Methods 1.1. Proteins PN-2/A␤PP was purified from fibroblast culture media using techniques of heparin affinity chromatography as described previously [18]. On nonreduced and reduced sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), it was a single band at 110 kDa. The KPI domain of PN-2/A␤PP, which was provided by Dr. Steve Wagner, Salk Institute Biotechnology/Industrial Associates, La Jolla, CA, was produced in a recombinant yeast expression system and purified as previously described [19]. Purified PN-2/A␤PP and its KPI domain were functionally quantitated by their ability to neutralize active-site titrated trypsin

[18,19]. Human FVIIa was prepared as previously reported using pRc/CMV vector and 293 embryonic kidney cells [20]. It contained 9–11 ␥-carboxyglutamic acid residues/molecule. On nonreduced 12% SDS-PAGE it was a single band at 50 kDa, and when reduced with 2% ␤-mercaptoethanol, a doublet at 34 and 20 kDa. All FVIIa was activesite titrated for functional activity with AT using a modified procedure of Griffith et al. [21] as reported by Schmaier et al. [13] for factor IXa. AT and FXa-DEGR were purchased from Haematologic Technologies Inc., Essex Junction, VT. The amount of AT was determined by radial immunodiffusion (Calbiochem, San Diego, CA). Human factors X and Xa were purchased from Enzyme Research Laboratories, South Bend, IN. Human FX was 81 kDa nonreduced and 53 kDa and 22.5 kDa when reduced and electrophoresed on a 13% SDS-PAGE. FXa activated by Russel’s viper venom was purchased from Enzyme Research Laboratories. Soluble tissue factor, amino acids 1–219 (TF1–219), was provided by Dr. Tom Girard of Monsanto, St. Louis, MO. Recombinant relipidated tissue factor, amino acids 1–243 (TF1–243), containing the transmembrane domain, was provided by Dr. Bob Kelly of Genentech Inc., South San Francisco, CA. It was reconstituted in phospholipid vesicles as reported [22].

1.2. Determination of Complex Formation Between Factor VIIa and PN-2/A␤PP Complex formation between FVIIa and PN-2/ A␤PP was determined by two methods: gel filtration, and solid-phase binding assay, as previously reported [13]. Sephadex G-100 (Sigma) gel filtration of FVIIa (24 ␮g) or a mixture of FVIIa (24 ␮g) and PN-2/A␤PP (120 ␮g) was performed on a 1⫻100-cm column equilibrated in 0.01 M Tris, 0.5 M NaCl, pH 8.0 containing 5 mM CaCl2. The column that was packed freshly upon each use was run at room temperature at a flow rate of 30 mL/ h. Sample size was ⬍0.03 mL and 1 mL fractions were collected. The void volume was determined by gel filtration of blue dextran (Sigma). Molecular mass markers of human immunoglobulin G (160 kDa) and bovine serum albumin (68 kDa) also were used to characterize the column. Column fractions were assayed for FVIIa antigen by ELISA. A 0.5-mL aliquot of each fraction was freeze dried

F. Mahdi et al./Thrombosis Research 99 (2000) 267–276

in a concentrator (Speed-Vac; Savant Instruments, Farmingdale, NY). Each lyophilized column fraction was resuspended in 100 ␮L of 0.1 M Na2CO3, pH 9.6, and applied to a well of a microtiter plate. After overnight incubation at 37⬚C, the wells were washed with 0.01 M sodium phosphate, 0.15 M NaCl, pH 7.4, containing 0.05% Tween 20. The wells were blocked with 1% radioimmunoassay grade bovine serum albumin, washed again, and then incubated with a murine antihuman FVIIa antibody (2 ␮g/mL) (Haematologic Technologies Inc., Essex Junction, VT) followed by a rabbit antimouse antibody conjugated with alkaline phosphatase (Sigma) diluted 1:1000. The color reaction was initiated by the addition of p-nitrophenylphosphate, disodium (1 mg/mL) in 0.05 M Na2CO3, 1 mM MgCl2, pH 9.8. Complex formation between FVIIa and PN-2/ A␤PP was also demonstrated by solid-phase binding assay. Microtiter plates were coated with PN2/A␤PP (576 ng/well) in 0.1 M Na2CO3, pH 9.6 and then blocked with 1% radioimmunoassay grade bovine serum albumin (Sigma). After washing, the wells were incubated with FVIIa (50 ng) in 0.01 M Tris, 0.15 M NaCl, pH 7.4 containing 5 mM CaCl2 and 0.05% Tween-20 followed by an anti-FVIIa monoclonal antibody (1 ␮g/mL) (Haematologic Technologies Inc.). After further incubation and washing, a rabbit antimouse antibody conjugated with alkaline phosphatase was added. The color reaction was initiated by the addition of p-nitrophenylphosphate, disodium (1 mg/mL) in 0.05 M Na2CO3, 1 mM MgCl2, pH 9.8.

1.3. Measurement of Factor VIIa Actvity The enzymatic activity of FVIIa was determined by its ability to hydrolyze various substrates. FVIIa (50 nM) in 20 mM Tris, 0.14 M NaCl, pH 7.4 containing 0.1% bovine serum albumin and 5 mM CaCl2 was examined for its ability to hydrolyze the substrate methoxycarbonyl-D-cyclohexylglycyl-glycyl-arginine-p-nitroanilide (1.25 mM) (Spectrozyme fXa, American Diagnostica) [23]. When soluble TF1–219 was added to this reaction, it was added at 70 nM. When FXa-DEGR was added, it was added at 400 nM. With and without tissue factor, hydrolysis proceeded continuously after the addition of the substrate for 30 min at 20–25⬚C, and readings were obtained every 30 s. When FX (400 nM) was used

269

as the substrate, FVIIa was added at 0.05 nM, relipidated tissue factor (TF1–243) was added at 0.56 nM, and N-Tosyl-Gly-Pro-Arg-p-nitroanilide was used at 0.45 mM in the same buffer as above. In these assays, the tissue factor activity was titrated to the FVIIa such that the minimal amount of tissue factor that allowed for the maximal rate, and the extent of FVIIa activity was considered to be appropriate for the reaction. A concentration of relipidated TF1–243 (0.56 nM) 11.2 times that of FVIIa (0.05 nM) was used in the assay. When FX was the substrate of the FVIIa–TF1–243 complex, activation of FX proceeded for 15 min at 20–25⬚C, and the reaction was stopped by the addition of 10 mM EDTA. The formed FXa was then measured continuously for 20 min by addition of the substrate, N-TosylGly-Pro-Arg-p-nitroanilide after the reaction was quenched by the addition of 50% acetic acid.

1.4. Determination of the Inhibition of FVIIa or FVIIa–TF Inhibition studies on FVIIa or FVIIa–TF1–219 by PN-2/A␤PP, KPI domain, or AT were performed by preincubating the inhibitors with the enzyme for 5 min prior to the addition of Spectrozyme fXa. The inhibitors were always used in concentrations 2–10-fold M excess of the concentration of FVIIa. When heparin (Elkins-Sinn, Cherry Hill, NJ) was added, it was at a concentration of 1 U/mL. Preliminary investigations determined the minimal concentration of heparin to use in these reactions for maximal AT inhibition. When studies to determine the IC50 of the FVIIa–TF1–243 complex by PN-2/ A␤PP, KPI domain, or AT/heparin were performed using FX as the substrate, the inhibitors were added at concentrations from 0.25 to 1000 nM.

1.5. Determination of Substrate Cleavage Investigations sought to determine if one- to fourfold M excess FVIIa to PN-2/A␤PP resulted in proteolysis of the PN-2/A␤PP. In these experiments, 2 to 8 ␮M FVIIa were incubated with 2 ␮M PN-2/A␤PP in 0.02 M Tris, 0.15 M NaCl, pH 7.4 containing 5 mM CaCl2 at 22–25⬚C for 1 h. Following incubation, the samples were reduced and then electrophoresed on a 6% SDS-PAGE. Further investigations were performed to determine if PN2/A␤PP were a substrate of FVIIa–F1–219. Studies

270

F. Mahdi et al./Thrombosis Research 99 (2000) 267–276

determined if 50 to 200 nM FVIIa–TF resulted in proteolysis of 50 nM PN-2/A␤PP after incubation for 1 h at 23⬚C. After incubation, the samples were reduced and then electrophoresed on a 7.5% SDSPAGE. The protein on the SDS-PAGE was then electrotransferred onto nitrocellulose and detected with a primary antibody to PN-2/A␤PP. The protein on the nitrocellulose was detected by chemilluminesence of a second antibody conjugated with peroxidase [15].

1.6. Calculation of Kinetic Parameters and Constants The apparent Km and Vmax of human FVIIa activation of Spectrozyme fXa in the presence or absence of soluble TF1–219 (70 nM) were determined by measuring the rate of substrate hydrolysis (0.05 to 2.0 mM) by 50 nM FVIIa. For these experiments, Spectrozyme fXa was placed into solution strictly according to the manufacturer’s instructions. The apparent Km and Vmax of human FVIIa–TF1–243 activation of FX was determined by measuring the rate of FX activation (10 to 1000 nM) by 0.05 nM FVIIa in the presence of 0.56 nM TF1–243 in 0.02 M Tris, 0.14 M NaCl, pH 7.4 containing 0.1% bovine serum albumin and 5 mM CaCl2. The mean⫾SD of each point performed in triplicate were analyzed on a double reciprocal plot by linear regression. The apparent Km and Vmax were determined from the negative reciprocal or reciprocal of the x- and y-intercepts of double reciprocal plots, respectively. The nM of FXa formed was determined by comparing the level of hydrolysis seen in the present assay with the level of hydrolysis measured by known concentrations of human FXa utilized under identical assay conditions. The turnover numbers for FXa formation (kcat) was determined by the ratio of maximum rate of FXa formed (nM/ min) divided by the concentration (nM) of the forming enzyme in the complex (FVIIa–TF1–243 complex). The stoichiometry of FVIIa inhibition by PN-2/A␤PP and its KPI domain was determined by nonlinear regression as previously reported [13,21]. Briefly, FVIIa at 50 nM was added to 2 to 200 nM inhibitor (PN-2/A␤PP or its KPI domain) and the residual FVIIa activity was determined. The plotted x-intercept of the inhibitor concentration vs. the % inhibition of FVIIa activity indicated

the concentration of added inhibitor neutralizing the known amount of added FVIIa. The equilibrium inhibition constants (Ki) for PN2/A␤PP and its KPI domain were calculated as previously reported [18] by the procedure of Bieth [24]. This method of calculation for slow tight-binding inhibitors used the following equation: Ki,app⫽ {[(I)/(1⫺a)]⫺(E)}/(1/a) to yield an apparent Ki where (I) is the inhibitor concentration, (E) is the factor VIIa concentration, and a is the fractional factor VIIa activity after incubation with the inhibitor. The actual Ki was calculated using the subsequent equation: Ki⫽Ki,app/1⫹([S]/Km) where [S] is the concentration of the substrate and Km is the Michaelis constant for the FVIIa substrate or FVIIa–TF substrate (protease substrate) reaction [24]. The apparent second-order association rate constants (kassoc) for each of the inhibitors were calculated from the same data using the integrated second-order rate equation: k″⫽[(1/I⫺E)·ln E(I⫺ EI)/I(E⫺EI)]/t, where E is the FVIIa concentration, I is the inhibitor concentration, EI is the concentration of the FVIIa inhibitor complex, and t is the time of incubation in minutes [25]. The integrated second-order rate equation was used for this calculation because the concentration of the inhibitors was within 10-fold M excess to enzyme.

1.7. Determination If There Is a Quarternary Complex Among FXa, FVIIa–TF, and PN-2/A␤PP One hundred microliters of purified FXa (217 nM) was coupled to microtiter plates in 0.1 M Na2CO3, pH 9.6 overnight at 37⬚C. After washing the cuvette wells with 0.01 M Tris, 0.15 M NaCl, pH 7.4, containing 0.05% Tween 20 and blocking with 1% radioimmunoassay grade bovine serum albumin (Sigma), the wells were incubated with 20 nM PN2/A␤PP in 0.01 M Tris, 0.15 M NaCl, pH 7.4 containing 5 mM CaCl2 and 0.05% Tween 20 in the absence or presence of 1–50 nM FVIIa–TF1–219. The presence of any PN-2/A␤PP on the cuvette wells was detected by anti-PN-2/A␤PP antibody (P21) at 1 ␮g/mL followed by a rabbit antimouse antibody conjugated with alkaline phosphatase. The color reaction was initiated by the addition of p-nitrophenylphosphate, disodium (1 mg/mL) in 0.05 M Na2CO3, 1 mM MgCl2, pH 9.8

F. Mahdi et al./Thrombosis Research 99 (2000) 267–276

Table 1. Inhibition of factor VIIa and Factor VIIa-soluble tissue factor complex Conditiona Factor VIIa Factor VIIa⫹TF1–219

PN-2/A␤PP KPI Domain Ki (M) 1.1⫾0.2⫻10⫺7b 7.8⫾0.3⫻10⫺8

1.5⫾0.1⫻10⫺7 6.8⫾0.6⫻10⫺8

a

FVIIa (50 nM) was incubated in 20 mM Tris, 0.14 M NaCl, pH 7.4 containing 0.1% bovine serum albumin and 5 mM CaCl2 in the absence or presence of TF1–219 (70 nM) in the presence of 200 to 500 nM PN-2/A␤PP or its KPI domain. After incubation for 5 min, the reaction was started by the addition of the substrate, methoxycarbonyl-D-cyclohexylglycyl-glycyl-arginine-p-nitroanilide (Spectrozyme fXa) (1.25 mM). Hydrolysis proceeded continuously for 30 min at 20–25⬚C before stopping the reaction with 50% acetic acid (see Materials and Methods). b The data presented are the mean⫾SD of ⬎26 determinations of FVIIa or FVIIa–TF1–219 complex and each inhibitor.

2. Results 2.1. PN-2/A␤PP Inhibits Factor VIIa FVIIa alone hydrolyzed methoxycarbonyl-Dcyclohexylglycyl-glycyl-arginine-p-nitroanilide with a Km of 0.284⫾0.034 mM (mean⫾SEM). When soluble TF1–219 was present, the Km remained similar at 0.315⫾0.02 mM. FVIIa hydrolysis of Spectrozyme fXa was linear for 75 min in the absence of soluble TF1–219 and 30 min in its presence. In the presence of soluble TF1–219, there was about 100fold increase in the rate of hydrolysis of the substrate. In the presence of a given concentration of PN-2/A␤PP, the degree of inhibition of FVIIa activity was constant over the 30 min of monitoring. This fact was true even if there were 0 to 60 min incubation of the enzyme with inhibitor before adding the chromogenic substrate. However, when the concentration of the inhibitor increased from 4- to 8-fold M excess, the degree of inhibition increased from 25 to 35%. PN-2/A␤PP inhibited FVIIa with a Ki of 1.1⫾0.2⫻10⫺7 M (Table 1). With KPI domain as an inhibitor, there also was a stable level of inhibition of FVIIa activity over 30 min of continuous monitoring. Similar to PN-2/A␤PP, when the KPI domain concentration was increased from 5- to 10-fold M excess, there was an increase in the degree of inhibition from 15 to 30%. The Ki of KPI domain inhibition of FVIIa was 1.5⫾0.1⫻ 10⫺7 M. Because the inhibition constant was similar for both PN-2/A␤PP and its KPI domain, the interaction between FVIIa with each form of the inhibitor must only have been through the Kunitz type

271

protease inhibitory domain. The stoichiometry of FVIIa inhibition by PN-2/A␤PP or its KPI domain was found to be 1:1, M:M. Further, high concentrations of the inhibitors, reduced FVIIa activity to zero, suggesting that the inhibitors were directed to the active site of FVIIa. Last, PN-2/A␤PP was not a substrate of FVIIa. Molar excess of FVIIa to PN-2/A␤PP (4:1 M to M) did not result in proteolysis of PN-2/A␤PP, as examined by SDS-PAGE (data not shown). When FVIIa was complexed with soluble TF1–219, the level of inhibition of FVIIa by both PN-2/A␤PP and KPI domain increased. At fourfold M excess PN-2/A␤PP or fivefold M excess KPI domain, there was about 40% inhibition of FVIIa–TF1–219 activity. At 10-fold M excess of both inhibitors, there was about 60% inhibition of FVIIa–TF1–219 activity. In all cases the degree of inhibition was stable over 30 min of continuous monitoring. The Ki of 4- to 10-fold M excess PN-2/A␤PP or KPI domain on the FVIIa–TF1–219 complex were also similar, but decreased over those seen with FVIIa alone, 7.8⫾0.3⫻10⫺8 M and 6.8⫾0.6⫻10⫺8 M, respectively (Table 1). When relipidated TF1–243 was substituted for soluble TF1–219, there was a 4.7-fold further increase in the degree of FVIIa inhibition by PN-2/A␤PP (Ki⫽1.65⫻10⫺8 M). Further, PN-2/ A␤PP was not a substrate of FVIIa–TF. Fourfold M excess FVIIa–TF1–219 did not result in proteolysis of PN-2/A␤PP when examined by an immunoblot of SDS-PAGE (data not shown).

2.2. PN-2/A␤PP Complexes with Factor VIIa To confirm that PN-2/A␤PP associated with FVIIa to inactivate it, investigations were performed by gel filtration and solid phase binding assay to study complex formation between these two proteins. Purified immunoglobulin G (Mr⫽160 kDa) and blue dextran (data not shown) gel filtered identically at fraction 32, characterizing the void volume of the column [13]. Bovine serum albumin gel filtered at fraction 38 [13]. Purified FVIIa gel filtered as a sharp peak at fraction 44 (Figure 1A). This finding indicated that FVIIa alone did not gel filter as a void volume protein on this column. When FVIIa was incubated with purified PN-2/A␤PP, the FVIIa antigen peak shifted to the left to fraction 38. The leading edge of the FVIIa antigen came off the column at fraction 32, indicating a higher

272

F. Mahdi et al./Thrombosis Research 99 (2000) 267–276

Fig. 1. Complex formation between FVIIa and PN-2/A␤PP. (A) Sephadex G-100 in 0.01 M Tris, 0.5 M NaCl, pH 8.0 containing 5 mM CaCl2 was used for gel filtration of FVIIa (24 ␮g) alone (䊐) or a mixture of FVIIa (24 ␮g) and PN-2A␤PP (120 ␮g [䉫]). One mL fractions were collected. The data presented are a single experiment of two performed. (B) Solid-phase binding assay between PN-2/A␤PP and FVIIa. Purified PN-2/A␤PP (576 ng) was linked to microtiter plate wells (see Materials and Methods). After blocking with bovine serum albumin, purified FVIIa (50 ng), antibody to FVIIa, and a secondary antibody to detect the antibody to FVIIa was added in sequential order. In various wells, one of the components was excluded. “No ABPP” represents wells where no PN-2/A␤PP were linked to the cuvette wells. “No FVIIa,” wells where no FVIIa was added; “No Anti-FVIIa,” wells where no antibody to FVIIa was added; “No 2nd Ab,” wells where no secondary antibody to detect the anti-FVIIa antibody was added; “ALL,” wells where every component was added. The data presented represent the mean⫾SEM of six experiments.

molecular mass complex between FVIIa (53 kDa) and PN-2/A␤PP (110 kDa) that gel filtered in the void volume fraction. Further studies were performed to determine if FVIIa forms a complex with PN-2/A␤PP as detected by solid phase binding assay. When PN-2/A␤PP was coupled to microtiter plate wells, FVIIa specifically bound to the PN-2/ A␤PP as detected by an antibody to FVIIa, followed by a secondary antibody conjugated with alkaline phosphatase (Figure 1B). These combined studies indicated that FVIIa and PN-2/A␤PP formed a complex characteristic of Kunitz-type inhibitors [13,15].

2.3. Comparison of factor VIIa Inhibition by PN-2/A␤PP and Antithrombin Investigations were next performed to determine the comparative inhibitory abilities of PN-2/A␤PP vs. AT and heparin (1 U/mL) on isolated FVIIa and FVIIa–TF1–219 complex (Table 2). When FVIIa or FVIIa–TF1–219 complex was incubated with 2–10fold M excess PN-2/A␤PP, its second-order association rate constants (kassoc) were 7.5⫾4.0⫻104 M⫺1min⫺1 and 1.6⫾0.2⫻105 M⫺1min⫺1, respectively. The addition of heparin (1 U/mL) to PN-2/ A␤PP did not influence its ability to inhibit FVIIa–

F. Mahdi et al./Thrombosis Research 99 (2000) 267–276

Table 2. Comparison of inhibition of factor VIIa and factor VIIa-soluble tissue factor complex Inhibitor PN-2/A␤PPa PN-2/A␤PP⫹heparin Antithrombin⫹heparin KPI domain

FVIIa FVIIa–TF1–219 k″(M⫺1min⫺1) 7.5⫾4.0⫻104b – 1.8⫾0.3⫻103 3.6⫾0.7⫻104

1.6⫾0.2⫻105 1.1⫾0.01⫻105 1.6⫾0.3⫻105 1.4⫾0.04⫻105

a FVIIa was added at 50 nM; TF1–219 was added at 70 nM; PN-2/A␤PP, KPI domain, and AT were added at 2–10-fold molar excess to the FVIIa. Heparin was at 1 U/mL when added. Each inibitor when added was incubated with the enzyme for 5 min prior to starting the reaction by the addition of the chromogenic substrate. The assay was performed as described in Table 1. b Each value is the mean⫾SD of seven or more determinations.

TF1–219 complex (k″⫽1.1⫾0.01⫻105 M⫺1min⫺1). In the absence of tissue factor, the second-order rate constant of AT and heparin inhibition (1 U/mL) of FVIIa was less than that of PN-2/A␤PP (k″⫽1.8⫾0.3⫻103 M⫺1min⫺1). In preliminary studies, heparin at 1 U/mL was found to be the minimal concentration that gave maximal inhibition of AT under the conditions of the reaction. The presence of tissue factor increased the inhibitory secondorder rate constant of the FVIIa–TF1–219 complex by AT and heparin 89-fold (k″⫽1.6⫾0.3⫻105 M⫺1min⫺1). The KPI domain inhibited FVIIa (k″⫽ 3.6⫾0.7⫻104) and FVIIa–TF1–219 complex (k″⫽ 1.4⫾0.04⫻105 M⫺1min⫺1) similar to PN-2/A␤PP. These combined studies indicated that tissue factor increased the inhibition by PN-2/A␤PP and KPI domain of FVIIa and FVIIa–TF1–219 complex as previously reported for AT and heparin [7,8]. Further, in the absence of tissue factor, PN-2/A␤PP and KPI domain were better inhibitors of FVIIa than AT. In the presence of tissue factor, these three inhibitors appeared to be equipotent at concentrations of the inhibitor up to 10-fold M excess, provided that heparin at 1 U/mL was present for AT.

2.4. Inhibition Studies of FVIIa–TF1–243 Complex in the Presence of Factor X Investigations next proceeded to determine the inhibition of FVIIa–TF1–243 complex by PN-2/A␤PP and KPI domain when factor X was the substrate. Initial studies characterized the Km and Vmax of FX activation by the FVIIa–TF1–243 complex using the relipidated form of TF1–243. The Km was found to be 0.13⫾0.034 ␮M (mean⫾SD) with a Vmax of 8.9⫾0.5

273

nM/min. These values have a kcat of 178 min⫺1 and a kcat/Km ratio of 1369 ␮M⫺1min⫺1. These results using relipidated TF1–243 were similar to those found by Ruf et al. [26] and Huang et al. [27]. Soluble TF1–219 alone was not able to support factor X hydrolysis under the conditions of the assay (data not shown). The FVIIa–TF1–243 complex activated factor X at a 2423-fold better rate than the chromogenic substrate methoxycarbonyl-D-cyclohexylglycyl-glycy-arginine-p-nitroanilide. Next, studies were performed to determine the relative potency of AT, PN-2/A␤PP, or KPI domain to block FX activation by FVIIa–TF1–243 by ascertaining their IC50 values. The IC50 of AT in the presence of heparin (1 U/mL) to inhibit FX activation was 12.8 nM. PN-2/A␤PP was the next most potent inhibitor of FX activation in this assembly with an IC50 at 65 nM. The KPI domain was the weakest inhibitor of FVIIa–TF1–243 in the presence of FX/Xa, achieving an IC50 only at 250 nM inhibitor (data not shown). Further investigations were performed to determine if the presence of factor X/Xa alters the ability of PN-2/A␤PP or its KPI domain to inhibit FVIIa– TF. Increasing concentrations of FVIIa–TF1–219 (1–50 nM) were able to compete PN-2/A␤PP from binding to FXa linked to a microtiter plate (Figure 2). This information indicated that in inhibition reactions containing both enzymes, a single molecule of PN-2/A␤PP or its KPI domain would be directed to either FXa or FVIIa–TF, but not both enzymes. Investigations next were performed to determine if the presence of an inactivated form of FXa, FXa-DEGR, would alter the inhibition of FVIIa–TF from that seen with a soluble FVIIa–TF alone. In the presence of 400 nM FXa-DEGR, the inhibition of FVIIa–TF by KPI domain of PN-2/ A␤PP was found to be 6.5⫾1.1⫻10⫺8 M, a value similar to that seen in the absence of FX. Likewise, in the presence of FXa-DEGR, the Ki of FVIIa–TF inhibition by PN-2/A␤PP was 6.2⫻10⫺8 M. These data also indicated that the presence of FX/Xa did not alter the degree of inhibition of FVIIa–TF by PN-2/A␤PP or its KPI domain.

3. Discussion These investigations indicate that PN-2/A␤PP and KPI domain are inhibitors of FVIIa and the FVIIa–TF complex. Under the conditions of the

274

F. Mahdi et al./Thrombosis Research 99 (2000) 267–276

Fig. 2. The ability of FVIIa–TF1–219 to compete PN-2/A␤PP from binding to FXa. FVIIa–TF1–219 (0–50 nM) was mixed with PN-2/A␤PP (20 nM) before the inhibitor was incubated in microtiter plate cuvettes that were coated with 217 nM FXa. PN-2/A␤PP bound to immobilized FXa was detected by an antibody to PN-2/A␤PP followed by a second antibody conjugated with alkaline phosphatase (see Materials and Methods). The data presented are the % inhibition of PN-2/A␤PP binding to the FXa after being incubated with increasing concentration of FVIIa–TF1–219. The figure is the mean⫾SEM of three independent experiments.

assays, PN-2/A␤PP and KPI domain are better inhibitors of FVIIa in the presence of tissue factor (Table 1). These data are consistent with that of Dennis and Lazarus, who found that KPI domain was a better inhibitor of FVIIa–TF than FVIIa alone [16,17]. Our studies also indicate that like factor IXa, PN-2/A␤PP and KPI domain are active site-directed inhibitors of FVIIa and the FVIIa–TF complex. At infinite concentrations, the residual activity of factors IXa and VIIa in the presence of these inhibitors is zero [14]. This result is different from the inhibition of factor Xa seen with these inhibitors [15]. At high concentrations of the inhibitor to factor Xa, there is residual activity of the factor Xa, suggesting that the inhibitor is not fully directed to the active site of factor Xa [15]. In comparison studies with AT and heparin in the absence of TF1–219, KPI domain and PN-2/A␤PP were 20–42-fold more potent (Table 2). As previously reported, our studies show that the presence of TF1–219 potentiates the inhibition of FVIIa by AT in the presence of heparin [7,8]. The presence of tissue factor also potentiates the inhibition of FVIIa by PN-2/A␤PP and its KPI domain. In

the presence of TF1–219, PN-2/A␤PP, KPI domain, and AT were equipotent inhibitors of FVIIa provided that 1 U/mL heparin is present for AT. Our studies also show that when FX was the substrate, the relationship between FVIIa–TF1–243 and each of the inhibitors does not really change. The IC50 of PN-2/A␤PP (65 nM) and KPI domain (250 nM) of FVIIa–TF1–243 were similar to their Kis determined in the absence of FX (78 nM, 68 nM, respectively). The IC50 for KPI domain is fourfold higher than that determined with PN-2/A␤PP. Although this difference may be within accepted variation, it is not entirely clear why there was this difference. There was no objective data to indicate that the whole molecule interacted better at higher concentrations than the smaller KPI domain. When FXa-DEGR was present, KPI domain and PN-2/ A␤PP inhibited FVIIa–TF1–219 (Ki⫽66 and 62 nM, respectively) the same as that seen when FXaDEGR was absent. These results were expected, because inhibition of an enzyme should not vary depending upon the substrate. Under experimental conditions that mimic a physiologic situation, the inhibitors were compared in the FX activation assay. AT, in the presence of heparin, was a fivefold more potent inhibitor than PN-2/A␤PP. At first glance, AT, due to its plasma concentration, might appear to be the major plasma inhibitor of FVIIa. However, the effectiveness of AT as an inhibitor of FVIIa requires pharmacologic concentrations of heparin (⭓1 U/mL), i.e., a nonphysiologic situation. Alternatively, PN-2/A␤PP is an effective inhibitor of FVIIa in the presence of TF1–219 and in the absence of heparin (Ki⫽7.8⫻10⫺8 M). At sites where platelets are activated, concentrations equal to its Ki to inactivate FVIIa and FXa are achieved [9,10]. In regions of the body where PN-2/A␤PP is located, it could also regulate FVIIa–TF activation of FIX. Last, it needs to be emphasized that neither PN-2/A␤PP nor AT under any circumstance are as potent inhibitors of FVIIa–TF as tissue factor pathway inhibitor [28]. PN-2/A␤PP is the most potent known inhibitor of factor IXa [13,14]. Our studies indicate that in tenase with factors IXa and VIIIa, PN-2/A␤PP is an effective inhibitor of factor IXa with a Ki of 2.5⫻ 10⫺8 M [14]. Although, PN-2/A␤PP kinetically is a very potent inhibitor of factor XIa (Ki⫽10⫺11 M), its relative importance in plasma has yet to be clarified, because C1 inhibitor, probably due to its

F. Mahdi et al./Thrombosis Research 99 (2000) 267–276

plasma concentration, would appear to be the most important inhibitor of factor XIa [11,12,18]. However, recent preliminary data for our laboratory indicate that factor XIa bound to endothelial cell membranes is only inhibited by PN-2/A␤PP and the KPI domain and not C1 inhibitor or ␣1 antithrypsin [29]. PN-2/A␤PP and the KPI domain also are reasonably potent inhibitors of FXa in the prothrombinase complex (Ki⫽1.3⫻10⫺8 M) [15]. Their inhibitory ability does not exceed AT and heparin, which is present in plasma in abundance. However, PN-2/A␤PP does not require heparin to inhibit FVIIa–TF1–219,and is not influenced by the absence or presence of FX in solution with a Ki (7.8⫻ 10⫺8 M). Kunitz-type protease inhibitors appear to be important regulators at critical juncture points in the hemostatic system because TFPI is the major regulator of FX activation by the FVIIa–TF1–243 complex, and PN-2/A␤PP, when present, is a potential regulator of FIX activation by FVIIa–TF and FX activation by FIXa in the presence of FVIIIa. This notion is especially important, considering the fact that PN-2/A␤PP is not an inhibitor of ␣-thrombin [13,18]. The ability of PN-2/A␤PP and KPI domain to inhibit factors XIa, IXa, Xa, and VIIa, suggest that it or its congeners could be anticoagulants of FXa formation and activity. Further, its additional ability to inhibit plasmin and kallikrein suggests that this class of compounds has features similar to aprotinin [30]. Investigations by Dennis and Lazarus [16,17] suggest that by using competitive phage selection with immobilized FVIIa–TF complex and soluble factor XIa, phage with high affinity for factor XIa were removed, and variants were selected with improved inhibitory ability to factors VIIa and Xa, plasmin, and kallikrein. These data suggest that a protease inhibitor of this class with its selective spectrum of inhibition may have the potential to be developed as an anticoagulant for special situations like cardiopulmonary bypass or extracorporeal circulation. Recent studies indicate that a recombinant form of the KPI domain was as good an anticoagulant as standard heparin in a rabbit venovenous extracorporeal circulation model [31]. This work was supported in-part by grant HL49566 to W.E.V.N. and A.H.S.

275

References 1. Rosenberg RD, Bauer KA. The heparin– antithrombin system: a natural anticoagulant mechanism. In: Colman RW, Hirsh J, Marder VJ, Salzman EW, editors. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. Philadelphia: J.B. Lippincott Co.; 1994. pp 837–60. 2. Tollefsen DM, Majerus DW, Blank MK. Heparin cofactor II. Purification and properties of a heparin-dependent inhibitor of thrombin in human plasma. J Biol Chem 1982;257:2162–9. 3. Baker JB, Low DA, Simmer RL, Cunningham DD. Protease-nexin: a cellular component that links thrombin and plasminogen activator and mediates their binding to cells. Cell 1980; 21:37–45. 4. Girard TJ, Warren LA, Novotny WF, Likert KM, Brown SG, Miletich JP, Broze GJ Jr. Functional significance of the Kunitz-type inhibitory domains of lipoprotein-associated coagulation inhibitor. Nature 1989;338:518–20. 5. Broze GJ Jr, Girard TJ, and Novotny WF. Regulation of coagulation by a multivalent Kunitztype inhibitor. Biochemistry 1989;29:7539–46. 6. Huang Z-F, Wun T-C, Broze GJ Jr. Kinetics of factor Xa inhibition by tissue factor pathway inhibitor. J Biol Chem 1993;268:26950–5. 7. Lawson J, Butenas S, Ribarik N, Mann KG. Complex-dependent inhibition of factor VIIa by antithrombin III and heparin. J Biol Chem 1993;268:767–70. 8. Rao LVM, Rapaport SI, Hoang AD. Binding of factor VIIa to tissue factor permits rapid antithrombin III/heparin inhibition of factor VIIa. Blood 1993;81:2600–7. 9. Van Nostrand WE, Schmaier AH, Farrow JS, Cines DB, Cunningham DD. Protease nexin2/amyloid beta-protein precursor in blood is a platelet-specific protein. Biochem Biophys Res Commun 1991;175:15–21. 10. Van Nostrand WE, Schmaier AH, Farrow JS, Cunningham DD. Protease nexin-II (amyloid beta-protein precursor): a platelet alpha-granule protein. Science 1990;248:745–8. 11. Smith RP, Higuchi DA, Broze GJ Jr. Platelet coagulation factor XIa-inhibitor, a form of Alzheimer amyloid precursor protein. Science 1990;248:1126–8. 12. Wuillemin WA, Minnema M, Meijers JCM, Roem D, Eerenberg AJM, Nuijens JH, ten

276

13.

14.

15.

16.

17.

18.

19.

20.

21.

F. Mahdi et al./Thrombosis Research 99 (2000) 267–276

Cate H, Hack CE. Inactivation of factor XIa in human plasma assessed by measuring factor XIa- protease inhibitor complexes: major role for C1-inhibitor. Blood 1995;85:1517–26. Schmaier AH, Dahl LD, Rozemuller AJM, Roos RAC, Wagner SL, Chung R, Van Nostrand WE. Protease nexin-2/amyloid beta protein precursor. A tight-binding inhibitor of coagulation factor IXa. J Clin Invest 1993; 92:2540–5. Schmaier AH, Dahl LD, Hasan AAK, Cines DB, Bauer KA, Van Nostrand WE. Factor IXa inhibition by protease nexin-2/amyloid betaprotein precursor on phospholipid vesicles and cell membranes. Biochemistry 1995;34:1171–8. Mahdi F, Van Nostrand WE, Schmaier AH. Protease nexin-2/amyloid beta-protein precursor inhibits factor Xa in the prothrombinase complex. J Biol Chem 1995;270:23468–74. Dennis MS, Lazarus RA. Kunitz domain inhibitors of tissue factor-factor VIIa. I. Potent inhibitors selected from libraries by phage display. J Biol Chem 1994;269:22129–36. Dennis MS, Lazarus RA. Kunitz domain inhibitors of tissue factor-factor VIIa. II. Potent and specific inhibitors by competitive phage selection. J Biol Chem 1994;269:22137–44. Van Nostrand WE, Wagner SL, Farrow JS, Cunningham DD. Immunopurification and protease inhibitory properties of protease nexin-2/amyloid beta-protein precursor. J Biol Chem 1990;265:9591–4. Wagner SL, Siegel RS, Vedick TS, Raschke WC, Van Nostrand WE. High level expression, purification, and characterization of the Kunitz-type protease inhibitor domain of protease nexin-2/amyloid beta-protein precursor. Biochem Biophys Res Commun 1992;186:1138–45. Billington JS. Construction, expression, and isolation of factor VII and factor VIIEGFI. M.S. Thesis. St. Louis, MO: St. Louis University; 1994. Griffith MJ, Breitkreutz L, Trapp H, Breit E, Noyes CM, Lundblad RL, Roberts HR. Characterization of the clotting activities of structurally different forms of activated factor IX. Enzymatic properties of normal human factor

22.

23.

24. 25. 26.

27.

28.

29.

30.

31.

IXa alpha, factor IXa beta, and activated factor IX Chapel Hill. J Clin Invest 1985;75:4–10. Sabharwal AK, Birktoft JJ, Gorka J, Wildgoose P, Petersen LC, Bajaj SP. High affinity Ca(2⫹)-binding site in the serine protease domain of human factor VIIa and its role in tissue factor binding and development of catalytic activity. J Biol Chem 1995;270:15523–30. Ruf W, Miles DJ, Rehemtulla A, Edginton TS. Mutational analysis of receptor and cofactor function of tissue factor. Methods Enzymol 1993;222:209–23. Bieth JG. In vivo significance of kinetic constants of protein proteinase inhibitors. Biochem Med 1984;32:387–97. Gigli I, Mason JW, Colman RW, Austin FW. Interaction of plasma kallikrein with the C1 inhibitor. J Immunol 1970;104:574–81. Ruf W, Rehemtulla A, Edgington TS. The integrity of the cysteine 186-cysteine 209 bond of the second disulfide loop of tissue factor is required for binding of factor VII. J Biol Chem 1991;266:10294–9. Huang Q, Neuenschwander PF, Rezaie AR, Morrissey JH. Substrate recognition by tissue factor-factor VIIa. Evidence for interaction of residues Lys165 and Lys166 of tissue factor with the 4-carboxyglutamate-rich domain of factor X. J Biol Chem 1996;271:21752–7. Broze GJ Jr, Miletich JP. Isolation of the tissue factor inhibitor produced by HepG2 hepatoma cells. Proc Natl Acad Sci USA 1987;84: 1886–90. Rojkjaer R, Schmaier AH. Regulation of kallikrein, factor XIa and factor XIIa activity on endothelial cell membranes. Thromb Haemost 1999;Suppl:33. Van Nostrand WE, Schmaier AH, Siegel RS, Wagner SL, Raschke WC. Enhanced plasmin inhibition by a reactive center lysine mutant of the Kunitz-type protease inhibitor domain of the amyloid beta-protein precursor. J Biol Chem 1995;270:22827–30. Annich G, White T, Damm D, Zhao Y, Mahdi F, Meinhardt J, Rebello S, Lucchesi B, Bartlett RH, Schmaier AH. Recombinant Kunitz protease inhibitory domain of the amyloid ␤- protein precursor as an anticoagulant in venovenous extracorporeal circulation in rabbits. Thromb Haemost. 1999;82:1474–81.