HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 48, 38120-38126, December 1, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/48/38120    most recent M007174200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Blostein, M. D. Articles by Furie, B. C. Search for Related Content PubMed PubMed Citation Articles by Blostein, M. D. Articles by Furie, B. C. Social Bookmarking                      What's this?

The Gla Domain of Human Prothrombin Has a Binding Site for Factor Va^*

Mark D. Blostein, Alan C. Rigby, Margaret Jacobs, Bruce Furie, and Barbara C. Furie^§

From the Center for Hemostasis and Thrombosis Research, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston Massachusetts 02215

Received for publication, August 8, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The role of the Gla domain of human prothrombin in interaction with the prothrombinase complex was studied using a peptide with the sequence of the first 46 residues of human prothrombin, PT-(1-46). Intrinsic fluorescence measurements showed that PT-(1-46) undergoes a conformational alteration upon binding calcium; this conclusion is supported by one-dimensional ¹H NMR spectroscopy, which identifies a change in the chemical environment of tryptophan 41. PT-(1-46) binds phospholipid membranes in a calcium-dependent manner with a K_d of 0.5 µM and inhibits thrombin generation by the prothrombinase complex with a K_i of 0.8 µM. In the absence of phospholipid membranes, PT-(1-46) inhibits thrombin generation by factor Xa in the presence but not absence of factor Va, suggesting that PT-(1-46) inhibits prothrombin-factor Va binding. The addition of factor Va to PT-(1-46) labeled with the fluorophore sulfosuccinimidyl-7-amino-4-methylcoumarin-3-acetic acid (PT-(1-46)AMCA) caused a concentration-dependent quenching of AMCA fluorescence, providing direct evidence of a PT-(1-46)-factor Va interaction. The K_d for this interaction was 1.3 µM. These results indicate that the N-terminal Gla domain of human prothrombin is a functional unit that has a binding site for factor Va. The prothrombin Gla domain is important for interaction of the substrate with the prothrombinase complex.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Prothrombin, a vitamin K-dependent protein, is the zymogen form of thrombin, the final enzyme generated during blood coagulation. Prothrombin is activated to thrombin by the enzyme factor Xa in the presence of phospholipid membranes and factor Va. Although factor Xa is able to convert prothrombin to thrombin in solution, the presence of phospholipid membranes and factor Va results in a 100,000-fold acceleration in thrombin generation allowing this enzymatic reaction to proceed at a physiologically relevant rate (1). Phospholipid membranes decrease the apparent K_m of factor Xa for prothrombin approximately 100-fold, and factor Va increases the k_cat of factor Xa for prothrombin approximately 3000-fold (1).

Prothrombin is composed of five domains; an N-terminal -carboxyglutamic acid (Gla)¹ domain, an aromatic amino acid stack domain, two kringle domains, and a C-terminal serine protease domain. The N-terminal Gla domain contains 10 Gla residues that are responsible for the calcium binding properties of prothrombin. The Gla domain defines the binding of vitamin K-dependent proteins to phospholipid membranes (for review, see Refs. 2 and 3).

Upon binding calcium ions, prothrombin undergoes two sequential conformational alterations, one can be detected spectroscopically (4), and both can be detected with conformation-specific antibodies (5). The second conformational change, which is supported only by calcium or strontium ions, is necessary to allow prothrombin to bind phospholipid membranes (6). The crystal structure of bovine prothrombin fragment 1, which comprises the Gla, aromatic amino acid stack, and first kringle domains, has been solved both in the absence (7) and presence of calcium ions (8). The latter structure demonstrates that five calcium ions are chelated by six Gla residues (Gla⁷, Gla⁸, Gla¹⁷, Gla²⁶, Gla²⁷, Gla³⁰), rendering 4 of the calcium residues inaccessible to solvent. Two additional calcium ions are bound to Gla residues 15, 20, and 21, and remain exposed to solvent. The sequestration of calcium ions in the interior of the protein results in the burying of the N-terminal alanine and contributes to the structural stability of the amino-terminal domain (8). In human prothrombin, site-directed mutagenesis studies demonstrate that Gla¹⁶, Gla²⁶, and Gla²⁹ (equivalent to Gla¹⁵, Gla²⁷, and Gla³⁰ in bovine prothrombin) are critical for phospholipid binding (9).

The Gla domain of the vitamin K-dependent proteins serves functions in addition to mediating phospholipid binding. A mutation in the Gla domain of factor IX (Gly¹²  Arg) results in defective activation of factor X by the complete tenase complex in the presence but not absence of factor VIIIa (10). A mutation in tissue factor, the cofactor of factor VIIa, results in defective activation of factor X but not Gla-domainless factor X by the factor VIIa-tissue factor complex (11). Furthermore, computational protein-protein docking analysis of the factor VIIa-tissue factor-factor Xa complex suggests interactive sites between the Gla domain of factor Xa and tissue factor-factor VIIa (12). Although the N-terminal 12 residues of the Gla domain are important for phospholipid binding (13), the C-terminal region of the Gla domain may be involved in protein-protein interactions. A chimeric protein C molecule in which the Gla domain of the protein was replaced with the corresponding region of prothrombin renders activated protein C activity independent of its cofactor, protein S. This suggests that the Gla domain of protein C is required for protein S cofactor activity. This protein S-independent activity has been localized to residues 23-46 of the Gla domain of protein C (14, 15). Domains within prothrombin that have been shown to bind to factor Va include the kringle 2 domain (16) and the kringle 1 domain (17). However, the Gla domain of prothrombin may be important for factor Va binding when these proteins are assembled on membranes. Such a hypothesis is consistent with factor Va and prothrombin interacting through several contact sites located in different domains. This paradigm has been demonstrated in the factor VIIa-tissue factor complex, in which the crystal structure identifies multiple interactive sites between the two proteins, including a hydrophobic interaction between the C-terminal helix of the Gla domain of factor VIIa and the second fibronectin domain of tissue factor (18).

To gain insight into the structure/function relationships of the Gla domain of prothrombin, we have synthesized and characterized the properties of a peptide comprising the Gla and aromatic amino acid stack domains of human prothrombin. The results demonstrate that this peptide binds calcium, binds to phospholipid membranes, and inhibits the prothrombinase complex. Using this synthetic Gla domain, we demonstrate a direct interaction between the Gla domain of human prothrombin and factor Va.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Human prothrombin, human factor IXa, human factor Xa, and human factor Va were purchased from Hematologic Technologies. Chromogenic substrate S-2238 was purchased from Diapharma Group Inc. Phosphatidylserine and phosphatidylcholine were obtained from Avanti Polar Lipids. Sulfosuccinimidyl-7-amino-4-methylcoumarin-3-acetic acid (AMCA) was purchased from Pierce. Rabbit IgG was purchased from Sigma.

Synthesis of PT-(1-46)-- PT-(1-46) was synthesized as described previously for factor IX-(1-47) (19) with the following modifications. The cleavage reaction was performed in trifluoroacetic acid/1,2-ethanedithiol/thioanisole/water/phenol (10:2.5:5:5:5, v/v) for 5 h at 25 °C. An intramolecular disulfide bond was formed by dialyzing the crude peptide (2 mg/ml) against 4 liters of 50 mM ammonium bicarbonate, pH 8.0. at 25 °C for 24 h. Following oxidation, solvent was removed from crude peptide by lyophilization. The crude, oxidized and deprotected peptide was purified by HPLC using an RP C18 column (Vydac, 250 × 21.5 mm). A linear gradient from 30 to 45% Buffer B (Buffer A: 0.1% trifluoroacetic acid, water; Buffer B: 0.1% trifluoroacetic acid, acetonitrile) over 60 min was employed. The column eluate was monitored at 214 nm and 280 nm. The amino acid sequence of the purified peptide was verified using an ABI Procise model 491 protein sequencer. The molecular mass of the peptide was determined by MALDI-TOF mass spectrometry on a Voyager linear MALDI-TOF spectrometer (PerSeptive Biosystems). The mass spectrometry analysis was performed with a nitrogen laser at 337 nm, employing either linear mode-positive or -negative ionizations. Amino acid analysis was performed on an acid hydrolysate using a Picotag system (Waters).

Determination of Free Sulfhydryl Groups of PT-(1-46)-- The percentage of free sulfhydryl groups remaining on PT-(1-46) after oxidation was determined by an Ellman assay (20). PT-(1-46) was dialyzed into 0.1 M sodium phosphate, pH 8.0, added to 1.8 mM 5,5'-dithio-bis(2-nitrobenzoic acid), and product formation was assayed by measuring the absorption at 412 nm using modified L-cysteine as a standard.

SDS-Polyacrylamide Gel Electrophoresis-- SDS-polyacrylamide gel electrophoresis was performed under non-reducing and reducing conditions according to Schagger and von Jagow (21). The peptides were visualized with Coomassie Blue.

Preparation of Phospholipid Vesicles-- Small unilamellar phospholipid vesicles (PC:PS, 60:40) were prepared by the method of Barenholz (22). Phospholipids in chloroform were dried at 45 °C under N₂, washed three times with methylene chloride, resuspended in 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, and sonicated in a bath sonicator until the solution cleared. The suspension of phospholipid vesicles was centrifuged, first at 160,000 × g for 30 min and then at 250,000 × g for 90 min in a Beckman L8-80M ultracentrifuge using a Ti70.1 rotor. The supernatant contained the small unilamellar phospholipid vesicles. Phospholipid concentrations were determined by phosphorus analysis (23).

Measurement of Phospholipid Binding-- The binding of PT-(1-46) to phospholipid vesicles was evaluated by 90° light scatter according to the method of Nelsestuen and Lim (24) on an SLM 8000C fluorescence spectrophotometer. The peptide was dissolved in 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 2 mM CaCl₂. Aliquots of the peptide were added to 3 ml of buffer containing 9.4 µM small unilamellar phospholipid vesicles (PC:PS, 60:40) to achieve the final peptide concentration shown. The sample was excited at a wavelength of 320 nm with a slit width of 4 nm, and emission was monitored at a wavelength of 320 nm with a slit width of 16 nm. Dissociation constants were calculated according to the equations of Lim et al. (24):

(Eq. 1)

where I is the scatter intensity of phospholipid plus protein corrected for increases in scatter intensity due to the protein itself. Io is the scatter intensity of phospholipid alone corrected for the decrease in intensity due to dilution with each addition of protein, and M2/M1 is the molecular weight ratio of the protein-phospholipid vesicle complex to phospholipid vesicles alone. The light scattering signal is related to the proportion of binding sites occupied by PT-(1-46):

(Eq. 2)

where PT-(1-46):PL_n is the concentration of phospholipid binding sites occupied by peptide, PL_t is the total concentration (µM) of phospholipid, and n is the number of phospholipid monomers per binding site. Msat/M1 is the signal due to light scattering of PT-(1-46) bound to phospholipid when all the binding sites for PT-(1-46) are occupied. Dissociation constants were calculated using a simple bimolecular model analogous to that used by Gilbert et al. (25):

(Eq. 3)

The initial value of M2/M1 is adjusted to zero by subtraction of 1. Msat/M1 is similarly normalized. PT-(1-46)_t is the total concentration (µM) of peptide. The equilibrium binding constant, K_d, was determined by fitting the light scattering data to the above equation using non-linear regression analyses with SigmaPlot for Windows 4.0 (SPSS).

Metal-induced Quenching of Intrinsic Fluorescence-- Fluorescence quenching experiments were performed at 25 °C using an SLM 8000C fluorescence spectrophotometer. PT-(1-46) (4.4 µM) was dissolved in 20 mM Tris-HCl, pH 7.4, previously treated with Chelex 100. The sample was irradiated at 280 nm using a slit width of 4 nm, and the emission was monitored at 340 nm using a slit width of 16 nm. The titrations were performed by the addition of CaCl₂ to the final concentration indicated (26). At the completion of titration, the reversibility of metal-induced quenching was tested by the addition of EDTA.

NMR Spectroscopy-- The NMR sample contained 0.15 mM PT-(1-46) in 100 mM d₄-acetate, 1 M NaCl, at pH 5.5 with 10% D₂O as the deuterium lock signal. Prior to the CaCl₂ titration, this sample was heated to approximately 50 °C at a neutral pH of 7.0 in the presence of Chelex 100 to ensure that all trace metal ions were removed. Spectra were collected on a Bruker AMX-500 spectrometer with a proton frequency of 500.14 MHz. One-dimensional spectra were acquired at 25 °C with 16,000 real data points, 2,048 summed scans, and a spectral width of 12,045.1 Hz. Following data acquisition, each spectrum was processed by applying a squared sine bell window function shifted by 30°. The carrier frequency was set on the water resonance, which was suppressed using presaturation.

Kinetic Analysis-- The prothrombinase assay was a modification of that of Rosing et al. (1). The assay was performed in Titretek (ICN Biomedicals Inc., Aurora) microtiter plates at 25 °C. The reaction mixture (40 µl) contained 0.1 nM factor Xa, 5 nM factor Va, 1.25 µM phospholipid vesicles (PC:PS, 60:40), and varying concentrations of PT-(1-46) in 150 mM NaCl, 20 mM HEPES, pH 7.4, 2 mM CaCl₂, and 0.1% bovine serum albumin. The reaction was initiated by the addition of 20 µl of prothrombin (final concentration, 300 nM) and stopped 60 s later by adding 20 µl of buffer containing 150 mM NaCl, 20 mM HEPES, pH 7.4, 20 mM EDTA, and 0.1% bovine serum albumin. Thrombin activity was measured at 25 °C using the chromogenic substrate S-2238 (0.25 mg/ml) on a Molecular Devices enzyme-linked immunosorbent assay plate reader.

For assays performed in the absence of phospholipid membranes, the concentration of factor Xa was 1.7 nM when factor Va (1.56 nM) was present and factor Xa was 5 nM when factor Va was omitted. The concentration of prothrombin was 300 nM for both conditions. The reaction was stopped after 60 min. For assays performed in the absence of factor Va but in the presence of phospholipid membranes, the reaction time was 5 min. Reactant concentrations were 1 nM factor Xa, 1.25 µM phospholipid vesicles, and 300 nM prothrombin.

The K_i for PT-(1-46) inhibition of prothrombin activation by prothrombinase was determined graphically using the method of Dixon (27). Reactions were performed with 0.1 nM factor Xa, 5 nM factor Va, 1.25 µM phospholipid vesicles at three different concentrations of prothrombin, namely, 50, 75, and 100 nM in the presence of varying concentrations of PT-(1-46) as indicated. A K_i for PT-(1-46) inhibition of factor Xa (1 nM) activation of prothrombin in the absence of factor Va was calculated in a similar fashion except the concentrations of prothrombin were 100, 400, and 800 nM and phospholipid vesicles were 1.25 µM. A K_i for PT-(1-46) inhibition of factor Xa (5 nM) activation of prothrombin in the absence of phospholipid membranes was also determined. Reactant concentrations were 5 nM factor Xa, 25 nM factor Va, and 150, 300, or 450 nM prothrombin.

Labeling of PT-(1-46) with AMCA-- To label PT-(1-46) with AMCA-Sulfo-N-hydroxysuccinimide, 100 µg of peptide was dissolved in 200 µl of 150 mM NaCl, 50 mM sodium borate, pH 8.5, and 1 mM CaCl₂. AMCA (10 µl, 0.67 mg/ml in water) was added, and the mixture was incubated at 25 °C for 60 min shielded from light. The reaction was stopped with 100 µl of 0.2 M Tris-HCl, pH 7.4, and the resulting product was dialyzed against Tris-buffered saline, pH 7.4, to remove any free AMCA. PT-(1-46)AMCA (150 µg) in 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, was added to a gel filtration column (D-Salt Extracellulose Desalting Column from Pierce, molecular weight cutoff 5000) and eluted with 150 mM NaCl, 50 mM Tris-HCl, pH 7.4. Fractions (0.5 ml) were collected, and the absorbance at 340 nm was measured. After labeling and dialysis, PT-(1-46)AMCA was subjected to analytical HPLC on a C18 column (Vydac, 2.2 × 25 cm) employing a linear gradient of 35-45% Buffer B (buffer A: 0.1% trifluoroacetic acid, water; Buffer B: 0.1% trifluoroacetic acid acetonitrile) over 30 min. Peaks were detected with a fluorescence detector (Beckman) using an excitation wavelength of 350 nm and an emission wavelength of 450 nm. Collected peaks were analyzed by MALDI-TOF mass spectrometry on a Voyager linear MALDI-TOF spectrometer.

Measurement of Factor Va Binding to PT-(1-46)AMCA-- Binding of factor Va to PT-(1-46)AMCA was monitored by quenching of AMCA fluorescence upon addition of factor Va using a Photon Technology International fluorescence spectrophotometer (Barnegat, NJ). The excitation wavelength was 350 nm with a slit width of 1 nm; emission scans were monitored from 400 to 500 nm with a slit width of 8 nm. Aliquots of factor Va in Tris-buffered saline, pH 7.4, 2 mM CaCl₂, 0.001% Tween 80 were added to 0.1 µM PT-(1-46)AMCA in the same buffer. Correction was made for changes in fluorescence signal due to the addition of buffer alone. Factor Va was added in aliquots of 0.25-0.5 µM up to a concentration of 3 µM. At the end of the titration, excess unlabeled PT-(1-46) was added to assess the reversibility of the fluorescence change. As a negative control, rabbit IgG was added to 0.1 µM PT-(1-46) up to a concentration of 3 µM. The binding affinity of factor Va for PT-(1-46) was calculated from fluorescence quenching data using a bimolecular equilibrium model as described previously (28).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

A peptide consisting of the -carboxylated N-terminal 46 residues of human prothrombin, PT-(1-46), was synthesized by solid phase peptide synthesis using N-(9-fluorenyl)methoxycarbonyl/N-methylpyrrolidone chemistry (Fig. 1). Following synthesis, the peptide was air-oxidized at pH 8.0 at room temperature for 24 h and purified using reverse phase HPLC. Automated Edman degradation revealed that the synthetic peptide had the correct sequence; amino acid analysis also confirmed the expected amino acid composition of the peptide. The peptide had the expected decarboxylated molecular mass of 5268 daltons (decarboxylation occurs during ionization) determined by MALDI-TOF mass spectrometry in linear mode-negative ionization.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Amino acid sequence of human PT-(1-46). This peptide has the sequence of the Gla and aromatic amino acid stack domains of human prothrombin. Gla residues are shaded; the disulfide bond is indicated by a black bar.

An Ellman assay for free sulfhydryls of the oxidized, purified peptide revealed that the molar ratio of cysteine to half-cystine was less than 10%. The purified peptide was subjected to SDS-gel electrophoresis in the presence and absence of 10% -mercaptoethanol. A single band migrating with a molecular weight of about 6000 was visible in the oxidized form of the peptide, confirming that dimerization or multimerization had not occurred during disulfide bond formation (Fig. 2).

View larger version (90K): [in this window] [in a new window]   Fig. 2.   SDS-gel electrophoresis of synthetic PT-(1-46). Each lane contains 6 µg of purified and oxidized PT-(1-46). Lane 1, reduced; lane 2, non-reduced. MW, molecular weight markers.

To determine whether PT-(1-46) undergoes a conformational alteration upon binding calcium ions, we measured the effect of increasing CaCl₂ concentration on PT-(1-46) intrinsic fluorescence. When irradiated at a wavelength of 280 nm, PT-(1-46) has an emission maximum at 340 nm. The single tryptophan residue at position 41 accounts for this intrinsic fluorescence. The intrinsic fluorescence of PT-(1-46) was quenched with increasing CaCl₂ concentration, reaching a maximal quenching of 55% (Fig. 3). Half-maximal quenching was observed at 0.2 mM CaCl₂. This effect was reversed by addition of EDTA. These results indicate that PT-(1-46) undergoes a change in its tertiary structure upon binding calcium ions, similar to that observed with factor IX-(1-47) (19), prothrombin fragment 1 (29), and prothrombin (4).

View larger version (9K): [in this window] [in a new window]   Fig. 3.   CaCl2-induced quenching of PT-(1-46) intrinsic fluorescence. Aliquots of CaCl2 were added to 4.4 µM PT-(1-46) in Tris-HCl, pH 7.4, to yield the indicated CaCl2 concentration. The emission spectrum was monitored at 340 nm after excitation at 280 nm. Data are corrected for the effect of dilution.

The calcium-induced structural perturbation of PT-(1-46) was assessed using one-dimensional ¹H NMR spectroscopy (Fig. 4). The peptide was initially rendered metal-free using Chelex 100. The proton resonances associated with the metal-free conformer of PT-(1-46) suggest that the peptide is partially structured, with the spectral dispersion associated with the amide protons shifted by greater that 0.15 parts per million (ppm) from their random coil values (30). However, the -protons associated with the malonate-like side chains of the 10 Gla residues were spectrally degenerate within a limited frequency envelope in this metal-free conformer (data not shown). To assess the ability of the peptide to bind calcium ions, a series of one-dimensional spectra were collected at increasing molar equivalents of CaCl₂ to peptide. Following the addition of CaCl₂, the proton frequencies showed increased spectral dispersion in all regions of the spectrum, including the amide proton region (Fig. 4). Many of the amide protons were shifted downfield (>0.25 ppm), the -protons became spectrally dispersed, and many of the -protons were shifted upfield from their random coil values indicative of increased helical structure at concentrations of calcium ions in excess of 40 equivalents (30, 31). The presence of two resonances for the indole proton (N₁) of tryptophan 41 (9.9 and 10.1 ppm) in the one-dimensional spectra illustrates that the metal-free and calcium-induced structures are in slow exchange on the NMR time scale. The metal-free resonance remains sharp, suggesting that the correlation time (_c) of this conformer is shorter. The calcium-induced resonance is shifted upfield and broadened, identifying the presence of a new structure that places the tryptophan moiety in an altered chemical environment. At elevated calcium ion concentrations, this resonance broadens due to self-aggregation of the peptide.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   One-dimensional 1H NMR spectrum of PT-(1-46). The sample contained 0.15 mM Chelex-treated PT-(1-46), 100 mM d4-acetate, 1 M NaCl, and 10% D2O at pH 5.5 in H2O. Each spectrum was acquired with 2048 scans at 25 °C. Each one-dimensional spectrum shows the aromatic and amide regions of PT-(1-46), following the addition of CaCl2 to the indicated final concentrations.

We determined that PT-(1-46) binds to phospholipid membranes in a calcium-dependent manner analogous to prothrombin. Binding was measured by the relative 90° light scattering technique. When increasing amounts of PT-(1-46) were added to phospholipid vesicles (PC:PS, 60:40) in the presence of 2 mM CaCl_2, saturable binding was observed (Fig. 5). A dissociation constant, K_d, of 0.5 ± 0.2 µM with an n value of 50 phospholipid monomers per binding site was calculated by fitting the data to a bimolecular equilibrium model. The binding was reversible following the addition of excess EDTA.

View larger version (10K): [in this window] [in a new window]   Fig. 5.   Binding of PT-(1-46) to phospholipid membranes. The binding of PT-(1-46) to phospholipid vesicles was measured by 90° light scattering. Aliquots of PT-(1-46) were added at the indicated concentrations to 9.4 µM phospholipid vesicles (PC:PS, 60:40), 2 mM CaCl2, 150 mM NaCl, 50 mM Tris-HCl, pH 7.4. Emission was monitored at 320 nm after excitation at 320 nm. M2/M1 is the molecular weight ratio of the peptide:phospholipid vesicles complex to phospholipid vesicles alone and is equal to the square root of the ratio of scatter intensity of phospholipid plus peptide over the scatter intensity of the phospholipid alone.

When PT-(1-46) was incubated with prothrombinase in the presence of the substrate prothrombin, thrombin generation was inhibited. Thrombin generation was inhibited 70% at a concentration of 40 µM PT-(1-46). Higher concentrations of peptide could not be tested due to aggregation (Fig. 6A). PT-(1-46) inhibition of prothrombin activation by prothrombinase fits a model of competitive inhibition. A K_i of 0.8 µM was measured (Fig. 6B). To further elucidate the mechanism by which PT-(1-46) inhibits prothrombin activation, the ability of PT-(1-46) to inhibit prothrombin cleavage by factor Xa in the presence of phospholipid membranes and absence of factor Va was measured. The K_i for inhibition of prothrombin activation under these conditions was 2.0 ± 0.5 µM, similar to the K_i of inhibition for the prothrombinase complex (Fig. 7).

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Inhibition of prothrombinase complex activity by PT-(1-46). A, the ability of PT-(1-46) to inhibit thrombin generation on phospholipid membranes was measured. PT-(1-46), at the indicated concentrations, was incubated with 0.1 nM factor Xa, 5 nM factor Va, and 1.25 µM phospholipid vesicles (PC:PS, 60:40) in 150 mM NaCl, 20 mM HEPES, pH 7.4, 2 mM CaCl2 at 25 °C. Prothrombin (300 nM) was added, and thrombin generation was measured after 60 s. B, PT-(1-46) was incubated with 0.1 nM factor Xa, 5 nM factor Va, 1.25 µM phospholipid vesicles, and prothrombin at three different concentrations of prothrombin, 50 nM () 75 nM (), and 100 nM (). PT-(1-46) inhibited prothrombin turnover as a competitive inhibitor with a Ki of 0.8 µM.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   PT-(1-46) inhibition of thrombin generation on phospholipid membranes. PT-(1-46), at the indicated concentrations, was incubated with 1 nM factor Xa and 1.25 µM phospholipid membranes in 150 mM NaCl, 20 mM HEPES, pH 7.4, 2 mM CaCl2 at 25 °C.

To test our hypothesis that the Gla domain of prothrombin, in addition to binding to phospholipid membranes, binds directly to factor Va, we assessed the ability of PT-(1-46) to inhibit prothrombin activation by factor Xa and factor Va in the absence of phospholipid. As indicated in Fig. 8, PT-(1-46) inhibits at least 70% of the thrombin generation in the absence of phospholipid membranes and in the presence of factor Va. A K_i of 14.9 ± 1 µM was determined for inhibition of this reaction by PT-(1-46). In the absence of both phospholipid membranes and factor Va, no inhibition was observed (Fig. 8). PT-(1-46) is able to inhibit the generation of thrombin in the absence of phospholipid membranes, and this inhibition is dependent on the presence of factor Va. These results suggest a prothrombin-factor Va interaction mediated, in part, by the Gla domain of prothrombin.

View larger version (15K): [in this window] [in a new window]   Fig. 8.   PT-(1-46) inhibition of thrombin generation by factor Xa and factor Va. The ability of PT-(1-46) to inhibit thrombin generation in solution was measured as follows. PT-(1-46), at the indicated concentrations, was incubated with 1.56 nM factor Va and 1.7 nM factor Xa in 150 mM NaCl, 20 mM HEPES, pH 7.4, 2 mM CaCl2 at 25 °C. Prothrombin (300 nM) was then added, and thrombin generation was measured (). PT-(1-46), at the indicated concentrations, was incubated with 5 nM factor Xa in 150 mM NaCl, 20 mM HEPES, pH 7.4, and 2 mM CaCl2 at 25 °C. Prothrombin (300 nM) was then added, and thrombin generation was measured ().

To provide a probe for measurement of direct interaction between PT-(1-46) and factor Va, PT-(1-46) was labeled with the fluorophore AMCA (excitation maximum 350 nm, emission maximum at 450 nm). This fluorophore selectively labels free primary amines, including lysine residues and the N terminus of a polypeptide. To prevent labeling of the N terminus of PT-(1-46) and loss of function (32), calcium ions were included in the labeling buffer to protect the N-terminal alanine (33). Analysis of the labeled peptide by reverse phase HPLC and mass spectrometry revealed that the preponderance of labeled peptide had a molecular mass of 5491 daltons corresponding to derivatization with a single AMCA moiety. The fluorescence emission spectrum of PT-(1-46) after excitation at 350 nm is shown in Fig. 9A. Addition of 3 µM factor Va to PT-(1-46)AMCA results in a fluorescence quenching of 30%. This quenching is completely reversed by addition of 20-fold excess of unlabeled PT-(1-46). IgG at 3 µM has no effect on PT-(1-46)AMCA fluorescence.

View larger version (16K): [in this window] [in a new window]   Fig. 9.   Binding of factor Va to PT-(1-46)AMCA. A, factor Va (3 µM) was added to 0.1 µM PT-(1-46)AMCA in 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 2 mM CaCl2, 0.001% Tween 80. Excitation was at 350 nm: the emission spectrum was obtained from 400 to 500 nm. PT-(1-46)AMCA emission prior to factor Va addition (solid line), PT-(1-46)AMCA emission after the addition of 3 µM factor Va(), PT-(1-46)AMCA emission after the addition of 3 µM IgG (x), PT-(1-46)AMCA emission after the addition of 2 µM of unlabeled PT-(1-46) to 3 µM factor Va and 0.1 µM PT-(1-46)AMCA(). B, aliquots of factor Va were added to 0.1 µM PT-(1-46)AMCA in 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 2 mM CaCl2, 0.001% Tween 80. Data are plotted as a percentage of quenching of emission at 450 nm after excitation at 350 nm. Non-linear regression curve fitting was employed as described under "Experimental Procedures."

Direct binding between the Gla domain of prothrombin and factor Va was demonstrated in experiments in which factor Va was added to PT-(1-46)AMCA. Addition of factor Va resulted in a concentration-dependent quenching of emission at 450 nm after excitation at 350 nm (Fig. 9A). Data were corrected for sample dilution and fitted to a bimolecular model of factor Va binding to PT-(1-46)AMCA (Fig. 9B). A binding constant, K_d, of 1.3 ± 0.5 µM, was determined.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

During blood coagulation, prothrombin is converted to thrombin by factor Xa in complex with factor Va and phospholipid membranes. This enzyme complex is known as prothrombinase. Phospholipid membranes are key components for the conversion of prothrombin to thrombin to occur at a physiologically relevant rate. The membrane binding portion of prothrombin resides in the N-terminal domain, which contains -carboxyglutamic acid and is referred to as the Gla domain (2). The Gla domain may also mediate protein-protein interactions within the prothrombinase complex as has been demonstrated in other vitamin K-dependent coagulation proteins (10, 11, 14, 15, 18). To study Gla domain-mediated protein-protein and protein-membrane interactions within the prothrombinase complex, a 46-residue peptide comprising the Gla domain and aromatic amino acid stack of human prothrombin was chemically synthesized. This peptide was anticipated to be the minimum size required to confer the membrane binding properties of the intact protein (19, 34, 35). The synthesized peptide had the correct molecular mass, amino acid composition, and disulfide bonding pattern. In the current work we have established that PT-(1-46) exhibits the properties associated with the Gla domain in intact prothrombin and prothrombin fragment 1. PT-(1-46) undergoes a calcium ion-induced conformational transition as measured by quenching of intrinsic fluorescence and one-dimensional ¹H NMR analysis. The peptide binds reversibly to phospholipid vesicles containing acidic phospholipid. Furthermore, the peptide inhibits activation of prothrombin by prothrombinase. Thus PT-(1-46) is a useful tool for assessing Gla domain-mediated protein-phospholipid and protein-protein interactions.

PT-(1-46) binds calcium ions and undergoes a change in tertiary structure as monitored by fluorescence spectroscopy and one-dimensional ¹H NMR spectroscopy. Intrinsic fluorescence is quenched with half-maximal quenching observed at 0.2 mM CaCl₂; this value is similar to that previously observed for prothrombin fragment 1 (4, 29, 36). The one-dimensional ¹H NMR spectra of PT-(1-46) acquired in the current study demonstrate a calcium-induced conformational reorientation of tryptophan 41. These results are supported by a comparison of the crystal structure of the calcium-bound and apo forms of bovine prothrombin fragment 1, which demonstrate reorientation of tryptophan 42 (equivalent to tryptophan 41 in human prothrombin) relative to the disulfide loop at residues Cys¹⁸-Cys²³ in the presence of calcium ions (8). The results of another study with N-terminal fragments of prothrombin (a mixture of residues 1-44 and 1-41) obtained with chymotryptic digestion of prothrombin found results similar to ours (33).

One of the major functions of the Gla domain of the vitamin K-dependent proteins is to support binding of these proteins to phospholipid membranes. The phospholipid binding properties of N-terminal fragments of several vitamin K-dependent blood coagulation proteins generated by either proteolysis or chemical synthesis indicate that both the Gla and aromatic amino acid stack domains are required to achieve binding similar to that of the native protein (19, 33, 35). Although these domains are encoded by separate exons, they work together as a functional unit. Our results with PT-(1-46) support this notion. We show that PT-(1-46) binds to phospholipid membranes in a calcium-dependent fashion with a dissociation constant, K_d, of 0.5 ± 0.2 µM, similar to that of prothrombin. It has been hypothesized that the second disulfide loop (Cys⁴⁸-Cys⁶¹) of the first kringle domain of bovine prothrombin is involved in membrane binding by forming a scaffold with residues Cys⁴⁸-Arg⁵⁵ of the Gla domain on which the remainder of the Gla domain anchors (8, 37). The results of our study and those of others (33) refute the importance of the second disulfide loop, and hence, the first kringle domain, for membrane binding. Our study supports the concept that only the Gla domain of prothrombin is required for protein-membrane binding, in a manner analogous to that of other vitamin K-dependent proteins. Nevertheless, prothrombin is unique among the vitamin K-dependent proteins in that it has two kringle domains instead of two EGF domains. The EGF domains of other blood-clotting proteins bind calcium and may play a role in stabilizing the structure of the Gla domain, as has been shown for factor X (38). It remains unclear whether the kringle domains bind calcium and play a similar role in prothrombin. However, site-directed mutagenesis studies of human prothrombin in which one or both kringle domains are deleted demonstrate that the kringle domains are not required for phospholipid binding (39). Furthermore, the crystal structure of prothrombin fragment 1 in the presence of calcium does not reveal many close contacts between the Gla and kringle 1 domains, although van der Waals interactions between Arg⁵⁵ and Gla¹⁵/Gla²⁰ were identified (8). An additional site-directed mutagenesis study of the Gla domain of human prothrombin is in agreement that the interactions between the prothrombin Gla domain, and kringle 1 domains are not important for phospholipid binding. Mutations of residues Gla¹⁴ and Gla¹⁹ (equivalent to Gla¹⁵ and Gla²⁰ in bovine prothrombin) do not result in altered phospholipid binding (9).

The physiological relevance of the prothrombin-like properties of PT-(1-46) is further emphasized by its ability to inhibit the prothrombinase reaction on membrane surfaces. This inhibition is likely due to inhibition of prothrombin binding to phospholipid membranes, because the K_i (0.8 µM) is very similar to the K_d (0.5 µM) of PT-(1-46) binding to phospholipid membranes. Furthermore, the peptide inhibits prothrombin turnover in the absence of factor Va but in the presence of phospholipid membrane, with a similar K_i (2 µM).

PT-(1-46) inhibits prothrombin activation by factor Xa-factor Va but not by factor Xa alone, suggesting that the prothrombin Gla domain participates in the interaction of prothrombin with factor Va. The K_i for PT-(1-46) inhibition of prothrombin activation by factor Xa-factor Va is 14.9 µM. This K_i of inhibition is greater than the K_d of binding of PT-(1-46)AMCA for factor Va, 1.3 µM. It is known that kringle 2 and possibly kringle 1 also constitute binding sites on prothrombin for factor Va (16, 17). Because domains other than the Gla domain contribute to the binding energy for interaction of prothrombin with factor Va, it is not surprising that the K_i for inhibition of prothrombin activation by factor Xa-factor Va is higher than the K_d for PT-(1-46) binding to factor Va. Although PT-(1-46) does not inhibit activation of prothrombin by factor Xa alone, we cannot rule out the possibility that the presence of factor Va induces interaction between the prothrombin Gla domain and factor Xa. Disruption of such an interaction would also be expected to contribute to the K_i. Analytical ultracentrifugation studies of prothrombin and factor Va interactions in solution reveal that prethrombin 1 and prothrombin bind equally (10 µM), implying that the Gla and kringle 1 domains do not bind to factor Va (40). On the other hand, another study demonstrated that prothrombin fragment 1 and kringle 1 peptides inhibit prothrombin turnover by factor Xa in solution in the presence of factor Va. Prothrombin fragment 1 is a more effective inhibitor than kringle 1, which is consistent with our observations (17).

These studies provide the first demonstration of a direct interaction between the Gla domain of prothrombin and factor Va when these proteins bind in solution. Given that kringle 1 (17) and kringle 2 (16) have also been shown to bind to factor Va, a model of prothrombin-factor Va interaction likely involves multiple contact sites in these proteins. This paradigm is analogous to that of the factor VIIa-tissue factor complex in which the crystal structure reveals multiple contact sites, including an interaction between the C-terminal helix of the Gla domain of factor VIIa and the second fibronectin domain of tissue factor (18). A similar model has also been proposed for the ternary complex of factor Xa-factor VIIa-tissue factor (12). Further study is required to identify the factor Va contact sites in the prothrombin Gla domain and to determine whether this protein-protein interaction occurs when factor Va and prothrombin assemble on phospholipid membranes.

	ACKNOWLEDGEMENTS

We thank Dr. Michael Parniak for use of his Photon Technologies International (Barnegat, NJ) fluorescence spectrophotometer and Dr. Nicholas Sluis-Cremer for his technical assistance.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants HL-42443 and -18834 and by a fellowship from the Medical Research Council of Canada (to M. D. B.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Current address: Lady Davis Institute for Medical Research, Sir Mortimer B. Davis-Jewish General Hospital, McGill University, Montreal, Quebec H3T IE2, Canada.

^§ To whom correspondence should be addressed: Center for Hemostasis & Thrombosis Research, Research East Rm. 319, Beth Israel Deaconess Medical Ctr., P. O. Box 15732, Boston, MA 02215. Tel.: 617-667-0620; Fax: 617-975-5505; E-mail: bfurie1@caregroup.harvard.edu.

Published, JBC Papers in Press, September 5, 2000, DOI 10.1074/jbc.M007174200

	ABBREVIATIONS

The abbreviations used are: Gla, -carboxyglutamic acid; AMCA, sulfosuccinimidyl-7-amino-4-methylcoumarin-3-acetic acid; HPLC, high pressure liquid chromatography; MALDI-TOF, matrix-assisted laser desorption time of flight; ppm, parts per million.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				P. Kamath and S. Krishnaswamy Fate of Membrane-bound Reactants and Products during the Activation of Human Prothrombin by Prothrombinase J. Biol. Chem., October 31, 2008; 283(44): 30164 - 30173. [Abstract] [Full Text] [PDF]

				S. Tran, E. Norstrom, and B. Dahlback Effects of Prothrombin on the Individual Activated Protein C-mediated Cleavages of Coagulation Factor Va J. Biol. Chem., March 14, 2008; 283(11): 6648 - 6655. [Abstract] [Full Text] [PDF]

				M. A. Bukys, P. Y. Kim, M. E. Nesheim, and M. Kalafatis A Control Switch for Prothrombinase: CHARACTERIZATION OF A HIRUDIN-LIKE PENTAPEPTIDE FROM THE COOH TERMINUS OF FACTOR Va HEAVY CHAIN THAT REGULATES THE RATE AND PATHWAY FOR PROTHROMBIN ACTIVATION J. Biol. Chem., December 22, 2006; 281(51): 39194 - 39204. [Abstract] [Full Text] [PDF]

				C. A. Kretz, A. R. Stafford, J. C. Fredenburgh, and J. I. Weitz HD1, a Thrombin-directed Aptamer, Binds Exosite 1 on Prothrombin with High Affinity and Inhibits Its Activation by Prothrombinase J. Biol. Chem., December 8, 2006; 281(49): 37477 - 37485. [Abstract] [Full Text] [PDF]

				R. Toso and R. M. Camire Role of Hirudin-like Factor Va Heavy Chain Sequences in Prothrombinase Function J. Biol. Chem., March 31, 2006; 281(13): 8773 - 8779. [Abstract] [Full Text] [PDF]

				F. Saller, B. O. Villoutreix, A. Amelot, T. Kaabache, B. F. Le Bonniec, M. Aiach, S. Gandrille, and D. Borgel The {gamma}-carboxyglutamic acid domain of anticoagulant protein S is involved in activated protein C cofactor activity, independently of phospholipid binding Blood, January 1, 2005; 105(1): 122 - 130. [Abstract] [Full Text] [PDF]

				S. Yegneswaran, R. M. Mesters, J. A. Fernandez, and J. H. Griffin Prothrombin Residues 473-487 Contribute to Factor Va Binding in the Prothrombinase Complex J. Biol. Chem., November 19, 2004; 279(47): 49019 - 49025. [Abstract] [Full Text] [PDF]

				H. Deguchi, S. Yegneswaran, and J. H. Griffin Sphingolipids as Bioactive Regulators of Thrombin Generation J. Biol. Chem., March 26, 2004; 279(13): 12036 - 12042. [Abstract] [Full Text] [PDF]

				D. O. Beck, M. A. Bukys, L. S. Singh, K. A. Szabo, and M. Kalafatis The Contribution of Amino Acid Region Asp695-Tyr698 of Factor V to Procofactor Activation and Factor Va Function J. Biol. Chem., January 23, 2004; 279(4): 3084 - 3095. [Abstract] [Full Text] [PDF]