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

This Article Abstract Full Text (PDF) Supplemental Research Data 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 Maun, H. R. Articles by Lazarus, R. A. Search for Related Content PubMed PubMed Citation Articles by Maun, H. R. Articles by Lazarus, R. A. Social Bookmarking                   What's this?

Disulfide locked variants of factor VIIa with a restricted -strand conformation have enhanced enzymatic activity

¹ Departments of Protein Engineering, ² Protein Chemistry, and ³ Physiology, Genentech Inc., South San Francisco, California 94080, USA⁴ Institute for Biology III, University of Freiburg, D-79104 Freiburg, Germany

Reprint requests to: Robert A. Lazarus, Department of Protein Engineering, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, USA; e-mail: lazarus.bob{at}gene.com; fax: (650) 225-3734.

(RECEIVED September 1, 2004; FINAL REVISION January 21, 2005; ACCEPTED January 31, 2005)

	Abstract

TOP Abstract Introduction Results and Discussion Materials and methods Electronic supplemental material References

 
Proteolytic processing of zymogen Factor VII to Factor VIIa (FVIIa) is necessary but not sufficient for maximal proteolytic activity, which requires an additional allosteric influence induced upon binding to its cofactor tissue factor (TF). A key conformational change affecting the zymogenicity of FVIIa involves a unique three-residue shift in the position of -strand B2 in their zymogen and protease forms. By selectively introducing new disulfide bonds, we locked the conformation of these strands into an active TF•FVIIa-like state. FVIIa mutants designated 136:160, 137:159, 138:160, and 139:157, reflecting the position of the new disulfide bond (chymotypsinogen numbering), were expressed and purified by TF affinity chromatography. Mass spectrometric analysis of tryptic peptides from the FVIIa mutants confirmed the new disulfide bond formation. Kinetic analysis of amidolytic activity revealed that all FVIIa variants alone had increased specific activity compared to wild type, the largest being for variants 136:160 and 138:160 with substrate S-2765, having 670- and 330-fold increases, respectively. Notably, FVIIa disulfide-locked variants no longer required TF as a cofactor for maximal activity in amidolytic assays. In the presence of soluble TF, activity was enhanced 20- and 12-fold for variants 136:160 and 138:160, respectively, compared to wild type. With relipidated TF, mutants 136:160 and 137:159 also had an approximate threefold increase in their V_max/K_m values for FX activation but no significant improvement in TF-dependent clotting assays. Thus, while large rate enhancements were obtained for amidolytic substrates binding at the active site, macro-molecular substrates that bind to FVIIa exosites entail more complex catalytic requirements.

Keywords: coagulation factor VIIa; tissue factor; serine protease; zymogen; hemostasis; disulfide; allostery

Abbreviations: TF, tissue factor • FVIIa, Factor VIIa • TF•FVIIa, tissue factor•Factor VIIa complex • FVII, Factor VII • FIX, Factor IX • FIXa, Factor IXa • FX, Factor X • FXa, Factor Xa • sTF, soluble tissue factor comprising the extracellular domain, residues 1–219 • pNA, para-nitroanalide

Article and publication are at http://www.proteinscience.org/cgi/doi/10.1110/ps.041097505.

	Introduction

TOP Abstract Introduction Results and Discussion Materials and methods Electronic supplemental material References

 
Coagulation is the biological process of blood clot formation involving many different serine proteases as well as their essential cofactors and inhibitors (Nemerson 1988; Davie et al. 1991; Broze Jr. 1992; Davie 1995; Rapaport and Rao 1995; Giesen et al. 1999; Mann 2003). It is initiated by exposure of Factor VII (FVII) and Factor VIIa (FVIIa) to the membrane bound cofactor, tissue factor (TF), resulting in production of Factor FXa (FXa) and more FVIIa. The process is propagated upon production of Factor IXa (FIXa) and more FXa that, upon binding to their respective cofactors FVIIIa and FVa, form platelet bound complexes, ultimately resulting in the formation of thrombin and a fibrin clot. Thrombin also serves to further amplify coagulation by activation of cofactors such as FV and FVIII and zymogens such as Factor XI. Moreover, thrombin activates platelets leading to platelet aggregation, which is necessary for the formation of a hemostatic plug.

The initiation and subsequent regulation of coagulation is understandably complex since maintenance of hemostasis is crucial for survival (Mann 2003; Lawson and Murphy 2004). There is an exquisite balance between hemostasis and thrombosis. Serious clinical conditions involving aberrations in coagulation include deep vein thrombosis, myocardial infarction, pulmonary embolism, stroke, and disseminated intravascular coagulation in sepsis. There are also many bleeding coagulopathies where there is insufficient clot formation. These include hemophilia A (FVIII deficiency) or hemophilia B (FIX deficiency), where pro-coagulant therapy is required. The challenge in this therapeutic area is to operate in the narrow window between too much and too little coagulation.

The use of exogenous FVIIa as a therapeutic agent has been shown to provide hemostasis in patients with hemophilia A and B (Hedner 2001, 2004). It also has been used to treat bleeding in patients with liver disease, anticoagulation-induced bleeding, surgery, thrombocytopenia, thrombasthenia, von Willebrand disease, and other bleeding disorders (Midathada et al. 2004). The precise mechanism by which exogenously added FVIIa exerts its procoagulant effect is a matter of some debate, as evidence for both TF-dependent and TF-independent FVIIa clotting activity has been presented (Butenas and Mann 2003; Butenas et al. 2003a, b; Lisman and de Groot 2003; Monroe and Roberts 2003). Thus, FVIIa variants with higher enzymatic activity, either in the presence or absence of TF, are of interest.

Approaches to enhance the enzymatic activity of FVIIa stem from the observation that FVIIa has key allosterically linked regions involving the TF binding site, the active site, and the macromolecular substrate exosite (Ruf and Dickinson 1998). The structures of the TF•FVIIa complex (Banner et al. 1996; Zhang et al. 1999), FVIIa (Kemball-Cook et al. 1999; Pike et al. 1999; Dennis et al. 2000; Sichler et al. 2002), and the zymogen FVII (Eigenbrot et al. 2001) have provided a structural context for thinking about this allostery. Analyses of the structural differences between TF•FVIIa, FVIIa, and FVII have led to several recent advances. These have focused on the roles of allostery and zymogenicity of FVIIa in different states (Persson et al. 2001c; Petrovan and Ruf 2001 Petrovan and Ruf 2002; Toso et al. 2003). Further understanding has led to engineered variants with improved catalytic activity (Persson et al. 2001a,b, 2004; Persson and Olsen 2002; Soejima et al. 2002; Persson 2004). Recent studies have demonstrated improved procoagulant, antifibrinolytic, and hemostasis properties in models of hemophilia A (Lisman et al. 2003; Tranholm et al. 2003).

Active and inactive (zymogen-like) forms of serine pro-teases exist in an equilibrium (Huber and Bode 1978), which is thought to favor the inactive state in the case of FVIIa (Higashi et al. 1996). Although not found in FVIIa, the zymogen-like form of the protease may even have some catalytic activity in some cases (Boose et al. 1989; Lijnen et al. 1990; Pasternak et al. 1998). In fact, FVIIa activity is not optimal until it binds to its cofactor TF, shifting the equilibrium to the active form of FVIIa (Butenas et al. 1993; Neuenschwander et al. 1993; Higashi et al. 1996). In addition, certain residues on FVIIa that contact TF have major effects on TF-dependent activity (Dickinson et al. 1996; Dickinson and Ruf 1997; Persson et al. 2001c).

Comparisons between the zymogen FVII and TF•FVIIa structures have revealed typical conformational differences in the serine protease activation domain (Eigenbrot et al. 2001; Eigenbrot 2002; Eigenbrot and Kirchhofer 2002), comprising the N terminus and the c140s, c180s, and c220s loops; chymotrypsinogen numbering is used throughout. However, a major conformational change in the TF binding region of the protease domain was also observed, due to an unexpected three-residue shift in -strand B2 relative to strand A2 (Fig. 1). Here, residues Thr151 to Val160 were shifted toward the C terminus relative to FVIIa, even though the main-chain H-bond interactions between -strands B2 and A2 are essentially identical in the zymogen FVII and TF•FVIIa structures.

View larger version (26K): [in this window] [in a new window]   Figure 1. FVIIa disulfide lock strategy. The registration of strands A2 and B2 in the TF•FVIIa-like active enzyme state is shown. The link between cysteine pairs is depicted in bold lines that were introduced at residue pairs that could form a disulfide only in the strand registration of the TF•FVIIa-like active state. The distances between C atoms in Å for the TF•FVIIa-like registration and (zymogen FVII) registrations are noted in the table with arrows pointing toward the engineered disulfide residue pair. The Leu-X-Val-Leu-X-Val residues important for reregistration in the zymogen and TF•FVIIa-like conformations (Eigenbrot et al. 2001) are depicted in bold ovals for both registrations. Hydrogen bonds are depicted as dashed lines.

	Results and Discussion

TOP Abstract Introduction Results and Discussion Materials and methods Electronic supplemental material References

 
Expression and purification of FVIIa mutants
The use of the specifically engineered mammalian expression vector pCMV.PD5.IRES-GFP facilitated and accelerated the generation of stable pools of cells expressing high yields of FVII variants (~1 µg/mL culture media). Comparing this yield with previous transient transfections, an improvement of 100- to 1000-fold was found. All mutants were initially purified from serum-free media using an sTF affinity column, thus verifying that the different FVII mutants were properly folded and contained a competent TF binding site. Proteins were eluted with 10 mM EDTA, since the binding of FVII to TF is highly Ca²⁺ dependent. In order to ensure a high degree of purity, further contaminants were removed by size exclusion. All mutants were purified as zymogens as indicated by a single band by SDS-PAGE under both nonreducing and reducing conditions; a representative gel is shown for mutant 139:157 (Fig. 2, lanes A,B).

View larger version (38K): [in this window] [in a new window]   Figure 2. Representative SDS-PAGE gels of FVII and FVIIa mutants. Purified FVII mutant 139:157 was run on SDS-PAGE gels under both nonreduced (A) and reduced (B) conditions. FVIIa mutant 139:157 activated by FXa as described in Materials and Methods was run on a reduced SDS-PAGE gel (C) where heavy and light chains are indicated. All mutants were expressed as zymogen and remained intact. Molecular mass markers are shown in kDa. All other mutants basically gave the same results from SDS-PAGE analysis under these conditions.

In order to characterize the kinetics of the designed FVIIa mutants, K_m and V_max values were first determined for amidolytic activity in the presence of sTF with a variety of chromogenic substrates, including S-2765, Spectrozyme fXa, Chromozym t-PA, and S-2288 (Table 2). In the presence of sTF, mutants 136:160 and 138:160 had 20.3- and 12.0-fold respective increases in S-2765 specific activity compared to wild type due to altered K_m and V_max values. The same mutants also had 8.8- and 4.0-fold increase in specific activity with Spectrozyme fXa. However, the activity for these mutants was greatly reduced using Chromozym t-PA and S-2288 as substrates (Table 2). In general, changes in activity for FVIIa mutants 137:159 and 139:157 were more moderate with all substrates.

View larger version (15K): [in this window] [in a new window]   Figure 3. Kinetics of FVIIa mutants with S-2765 amidolytic activity. Representative individual kinetic analysis for amidolytic activity of S-2765 with 30 nM wild-type FVIIa (•) and 30 nM FVIIa mutants (all normalized by active site titration) 136:160 (), 137:159 (), 138:160 (), and 139: 157 (). Data for the Michaelis-Menten plot were fit to a hyperbolic equation using Kaleidagraph, from which values for Km and Vmax were derived; independent determinations were performed in triplicate.

View larger version (17K): [in this window] [in a new window]   Figure 4. Amidolytic activity of FVIIa mutants. The fold increases in amidolytic activity (Vmax/Km) for FVIIa disulfide-locked variants in the absence of sTF relative to wild type are shown. Data for mutants with S-2765 is shown in black and with Spectrozyme fXa in gray; the fold increase is shown above the respective column.

Overall, the data demonstrate that locking the A2 and B2 -strand registration in FVIIa can lead to variants with significantly higher enzymatic activity, especially in the absence of TF. It is challenging to understand the substrate specificity across the various FVIIa disulfide locked variants. All substrates contain Arg in the P1 position. Substrates S-2765, Spectrozyme fXa, and Chromozym t-PA have a Gly at the P2 position, whereas S-2288, which is relatively poor substrate for the mutants in the absence of sTF, has a Pro at P2. The different substrate effects are most likely due to changes at the P3 position where S-2765, Spectrozyme fXa, Chromozym t-PA, and S-2288 have D-Arg, D-cyclohexylglycyl, D-Phe, and D-Ile, respectively.

TF binding to FVIIa disulfide-locked variants
The effects of the mutations in FVIIa upon binding to sTF were determined by surface plasmon resonance (Table 4). In this assay, wild-type FVIIa had a K_D of 5.2 nM, in good agreement with data previously reported (Kelley et al. 2004). All of the disulfide locked mutants were somewhat impaired in their ability to bind to sTF, with 138:160 having the most significant loss in binding of 12-fold. All mutants had moderately slower association rates and slightly faster dissociation rates. The lower affinity for sTF was not predicted, since both FVII and FVIIa bind with comparable affinities (Kelley et al. 2004). A complex set of allosteric interactions involving TF, the FVIIa active site and macro-molecular binding sites complicates any interpretation (Ruf and Dickinson 1998).

View larger version (21K): [in this window] [in a new window]   Figure 5. Relative TF-dependent clotting of FVIIa mutants in FVII-deficient plasma. Relative clotting times were normalized to the clotting time in FVII-deficient plasma. Data is shown for wild-type FVIIa (•) and FVIIa mutants 136:160 (), 137:159 (), 138:160 (), and 139:157 (). The average data from three independent determinations were fit by a four-parameter fit using Kaleidagraph; the error as standard deviation is shown.

Comparison with other approaches
The mechanism underlying the zymogenicity of FVIIa and its dependence on cofactor TF for a full activity is only partially understood (Ruf and Dickinson 1998; Eigenbrot 2002; Eigenbrot and Kirchhofer 2002; Petrovan and Ruf 2002). Several engineered variants of FVIIa have been reported, revealing specific residues or regions on the protein that are important for its interactions with TF. To date, all reported mutations that improve enzymatic activity have been found either in the -carboxyglutamic acid (Gla) domain, resulting in FVII variants with significantly enhanced membrane binding properties and FX activation activity (Nelsestuen et al. 2001; Harvey et al. 2003), or one of three allosteric regions in the protease domain: the TF binding region, the macromolecular substrate binding exosite, and residues in the catalytic cleft (Persson et al. 2001a,b,c, 2004; Persson and Olsen 2002; Persson 2004). Alterations in one of these regions can have effects on the others, indicating that there is a direct yet complex set of interactions involved.

Previous strategies addressing FVIIa zymogenicity and engineering of rate enhancements for zymogen-like FVIIa have primarily involved substitution of residues from other intrinsically more active serine proteases. Met156 has been reported to be a determinant of FVIIa zymogenicity, since mutation to Gln, which is found at this position in FIX, resulted in three- and ninefold enhanced FVIIa amidolytic and proteolytic activity, respectively (Petrovan and Ruf 2001). Based upon a comparison of the free and TF-bound FVIIa structures (Pike et al. 1999), changing Leu163 to Val increased FVIIa activity three- to fourfold, presumably due to movement of the -helix comprising residues 165–170 into an orientation more akin to that found in FXa or thrombin (Persson et al. 2001a). Additional residues targeted to stabilize this helix also increased activity, presumably by inducing a conformation similar to that found upon TF binding (Persson et al. 2004). Mutations nearby the same -helix also increased activity as found when the 170 loop in FVIIa was replaced by a shorter one from trypsin; replacement of the 99 loop with the corresponding sequence from trypsin also resulted in more active FVIIa variants and also broadened substrate specificity (Soejima et al. 2001, 2002). Mutation of Lys188 to Ala, designed to minimize repulsion of the positively charged N terminus forming its salt bridge with Asp194 also resulted in FVIIa rate enhancement (Persson et al. 2001b). The three-residue motif Val21, E154, and M156 in FVIIa was replaced by Glu, Arg, and Lys, respectively, found in thrombin and FIX, which also resulted in enhanced FVIIa activity (Persson et al. 2001b). Mutations of these three "zymogenicity-determining" residues have been studied more extensively, resulting in the conclusion that there is a complex set of interactions that stabilize the active conformation of FVIIa, but not zymogen FVII (Petrovan and Ruf 2002). Further studies have elucidated the molecular properties of these mutations (Persson and Olsen 2002).

While this manuscript was under review, a paper describing a somewhat similar strategy appeared (Olsen et al. 2004). These authors concluded that altering the disulfide link between Cys22 and Cys27 to a new one between Cys22 and Cys157 to restrict -strand conformation did not lead to mutants with improved activity. However, the residues chosen to impart the conformational restraint differed from those we chose to use. Thus, engineering mutants with enhanced activity greatly depends upon the specific residues used to lock the disulfides, requiring a somewhat empirical approach. This is not really surprising, since not all of our mutants exhibited improved activities either.

Conclusions
We sought to enhance the enzymatic activity of FVIIa by engineering a new disulfide bond to restrict -strand conformational changes. Due to uncertainties associated with introducing two Cys mutations and subsequent oxidation, we designed and produced several mutants. This approach was successful in so far as the cysteines engineered into -strands A2 and B2 of FVIIa indeed formed disulfides, and the resulting disulfide-locked mutants had greatly enhanced amidolytic activity, at least for some of the substrates. Most notable was the fact that we were able to eliminate the role of sTF as a cofactor, thus achieving the goal of mimicking a TF•FVIIa-like conformational state with FVIIa itself. However, by locking the disulfides to favor a FVIIa enzyme-like state, we may have altered the structural plasticity such that undesired conformational restraints in areas linked to enzymatic activity were introduced, as previously discussed in other serine proteases (Gráf 1995; Szabó et al. 1999, 2003; Reyda et al. 2003). This may be the case with macromolecular substrates, where only slightly enhanced activity was found. The goal of engineering a TF-independent FVIIa with the potential to initiate clotting as efficiently as FVIIa in complex with TF is daunting. Nonetheless, such an engineered FVIIa might have advantageous properties as a therapeutic agent in certain clinical scenarios.

	Materials and methods

TOP Abstract Introduction Results and Discussion Materials and methods Electronic supplemental material References

 
Mutant design
The designed disulfide links were engineered by seeking one residue each in -strands A2 and B2 (Fig. 1), where a disulfide might reasonably form without large changes to the direction of the vectors from C to C, i.e., without changing the direction in which the side chain was projected. Among several possible selections, we noted that residues Ser136, Leu137, Val138, and Ser139 from -strand A2 and Val157, Asn159, and Val160 from -strand B2 are the middle section of main-chain–main-chain H-bonds between the two -strands in both zymogen and enzyme structures. Disulfide links from each of these positions in A2 to positions in B2 were evaluated visually for steric conflicts and judged to have accessible conformations consistent with formation of engineered covalent links. Thus, potential disulfide links that would favor the active TF•FVIIa registration were generated to form disulfide-locked mutants of FVIIa designated 136:160, 137:159, 138:160, and 139:157, reflecting the position of the new disulfide bond.

Mutagenesis and construction of plasmids
The wild-type FVII expression plasmid pRKCT31 was used as a starting template for Kunkel mutagenesis (Kunkel et al. 1987) to generate the various mutant-encoding plasmids. The following primers were used in combination to introduce the indicated mutations: VII-S136C (5'-GCTGACCAAGCAGAAGCGCAC-3'), VII-L137C (5'-GCCGCTGACGCATGAGAAGCG-3'), VII-V138C (5'-GCCCCAGCCGCTGCACAATGAGAAGCG-3'), VII-S139C (5'-GCCCCAGCCGCAGACCAATGA-3'), VII-V157C (5'-CA CGTTGAGGCACATGAGCTC-3'), VII-N159C (5'-CCGGGG CACGCAGAGGACCAT-3'), VII-V160C (5'-CAGCCGGGGGC AGTT GAGGAC-3').

Two primers were used in one Kunkel mutagenesis reaction to introduce mutations to make four different FVII variants. The entire cDNAs encoding the double mutants were verified by DNA sequencing to exclude the presence of unwanted mutations. The cDNAs encoding the various FVII double mutants were cloned into the EcoR1 and HindIII sites of the mammalian expression vector pCMV.PD5.IRES-GFP, which was derived from vector pCMV.DI.tPA (Lucas et al. 1996) by introducing IRES-GFP downstream of the target gene. Plasmids were prepared by using the QIAprep spin miniprep kits (Qiagen).

Cell culture, transfection, selection, and expression
DP12 cells (1–1.5 x 10⁶ of CHO K1 DUX B11 (DHFR^–) (Lucas et al. 1996) were seeded on a 100-mm dish in 10 mL DP12 media (F12/DMEM low glucose media containing 10% FBS [Sigma], 1% glutamine, 100 µg/mL penicillin, 250 µg/mL streptomycin [Invitrogen] 1 mM HEPES [pH 7.2], and thymidine 5 µg/mL [GHT]) 24 h before transfection. Transfection media (1.2 mL of HG DMEM without FBS) were mixed with 36 µL FuGENE 6 (Roche Applied Science) in a sterile tube and incubated for 5 min at room temperature. pCMV.PD5.IRES-GFP expression plasmid (12 µg) encoding FVII mutant was added and incubated for 15 min at room temperature. FuGENE 6/plasmid mixture was added dropwise to DP12 cells and incubated for 48 h at 37°C. Transfected cells were split after 48 h and maintained in DP12 media containing 10 µg/ mL puromycin to select for stable transfectants.

Stable transfectants were then sorted by FACS on a Beckman Coulter Epics Elite Flow Cytometer for cells having the top 5% in fluorescence intensity due to the GFP reporter. Cells were maintained for expression in DP 12 media including 10 µg/mL puromycin. FVII variants were expressed from stable cell pools in serum-free media containing trace elements, 10 µg/mL human insulin, and 6 µg/mL vitamin K (Aquamephyton, Merck) at 32°C. Medium containing secreted FVII variant was harvested after 7 d of incubation.

Purification of FVII mutants
An FVII affinity column was prepared by immobilizing 13 mg of soluble tissue factor (sTF) (Kelley et al. 1995) on a 1-mL HiTrap NHS-activated HP column (Amersham Biosciences) following the manufacturer’s instruction. Harvested tissue culture media was sterile filtered and brought to 5 mM CaCl₂ and 20 mM Tris (pH 8) before loading at 1 ml/min onto the immobilized sTF column, previously equilibrated with wash buffer (20 mM Tris [pH 8], 5 mM CaCl₂, 135 mM NaCl, and 2 mM benzamidine). The column was washed with 10 column volumes of wash buffer and eluted with 5 column volumes of 20 mM Tris (pH 8), 150 mM NaCl, 10 mM EDTA, and 2 mM benzamidine. The eluate was concentrated and subjected to size exclusion on a Superdex 200 Tricor column (Amersham Biosciences) for further purification in running buffer (20 mM Tris [pH 8], 300 mM NaCl, 10 mM EDTA) at a flow rate of 0.5 mL/min. Fractions containing FVII variants were pooled and concentrated.

Characterization of activated FVIIa mutants
To activate FVII mutants, proteins were mixed with 1/10 (w/w) biotinylated FXa (Roche Applied Science) and brought to 1.5 mL final volume in 50 mM Tris (pH 8), 100 mM NaCl, 5 mM CaCl₂. Following incubation for 4 h at room temperature, biotinylated FXa was removed with Streptavidin beads as suggested in the manufacturer’s protocol.

All FVIIa variants were analyzed by SDS-PAGE in nonreduced or reduced form; samples were reduced by addition of 1 µL of 14.3 M -mercaptoethanol (Sigma) to sample and boiling for 3 min prior to SDS-PAGE analysis on a 4%–20% Tris-Glycine Novex gel followed by staining with Coomassie Blue. Protein concentrations were determined by amino acid analysis and OD₂₈₀ with an extinction coefficient of (1.34 g/L)^–1 x cm^–1. Amino acid analysis confirmed the calculated extinction coefficient was accurate to determine the protein concentration by OD₂₈₀. All FVIIa variants were active site titrated using the Kunitz domain inhibitor TF7I-C, quantified by active site titrated trypsin, as described to determine the concentration of active sites (Dennis and Lazarus 1994; Seymour et al. 1994).

Mass spectrometry analysis of FVIIa mutants
Mass spectrometry was used to confirm the presence of the additionally introduced disulfide bond. FVII (100 µg of mutant or wild type) was incubated with fivefold molar excess of iodoacetamide (Sigma) in 50 mM ammonium bicarbonate (pH 7.5) for 15 min at room temperature in the dark in order to alkylate all free cysteines. After alkylation, FVII was digested with 2.5 µg trypsin (Promega) in 50% acetonitrile at 37°C overnight. The entire digest mixture was analyzed in the oxidized and reduced state (addition of -mer-captoethanol) by mass spectrometry to identify peptide masses that correlated with the disulfide-linked peptides from -strands A2 and B2.

Nonreduced peptides were analyzed by orthogonal MALDI-TOF MS (QSTAR XL; Applied Biosystems) and capillary HPLC electrospray ion trap tandem mass spectrometry. MALDI samples were prepared by 1:1 mixture with -cyano-4-hydroxycinnamic acid (Agilent Technologies) and 1 µL applied to the sample probe and dried under ambient conditions. For LC-MS analysis, sample aliquots were injected onto 75 µm id Picofrit capillary columns (New Objective Inc.), packed with 9 cm of C18 resin (5 µm; Michrom Bioresources). Peptides were eluted directly into the microelectrospray source of an LCQ Deca XP-plus mass spectrometer (Thermo Electron) with a gradient of 0%–40% acetonitrile in 0.1% acetic acid, 0.005% TFA at a flow rate of 200 nL/min. The mass spectrometer performed MS and MS/MS scans in a data-dependent experiment; full mass range MS scans were followed by collision-induced dissociation (CID) scans of the three most intense ions detected. A dynamic exclusion list prevented any precursor ion from being subjected to CID more than twice. The disulfide-linked peptides of interest were identified by examination of the mass spectral data. Reconstructed ion chromatograms were plotted for the doubly and triply charged ions of each anticipated peptide dimer. The corresponding CID spectra were then interpreted, matching observed fragment ions to those predicted for each peptide.

Disulfide-linked peptides were reduced for 1 h at 37°C with 1 mM DTT. The reduced peptide mixture (1 µL) was diluted with 1 µL of 2,5-DHB matrix (2,5-dihydroxybenzoic acid, Agilent), spotted onto a stainless steel maldi plate, and allowed to air dry at room temperature. MALDI-TOF mass spectrometry was performed on a Voyager-DE STR instrument (Applied Biosystems) operated in reflection mode with delayed extraction.

The extent to which Gla domain glutamic acid residues were post-translationally modified to -carboxyglutamic acids was determined by MALDI-TOF mass spectrometry as described above with the following differences: Digestion with trypsin was carried out after reduction of disulfides with 10 mM DTT and was stopped after 2 h. The MALDI matrix used in this experiment was a saturated solution of 5-methoxysalicylic acid (Tokyo Kagei Kogyo Co., Ltd.) in 60% acetonitrile/0.1% TFA. MALDI-TOF mass spectrometry was performed in the linear mode of the Voyager-DE STR with delayed extraction.

FVIIa amidolytic activity assay
The amidolytic activity of FVIIa and the FVIIa mutants were measured using chromogenic substrates Chromozym t-PA; N-methylsulphonyl-D-Phe-L-Gly-L-Arg-pNA (Roche Applied Science), S-2288; H-D-Ile-L-Pro-L-Arg-pNA, S-2765; Z-D-Arg-L-Gly-L-Arg-pNA where Z is a benzoyl group (DiaPharma), and Spectrozyme fXa; methoxycarbonyl-D-cyclohexylglycyl-L-Gly-L-Arg pNA (American Diagnostica). FVIIa and FVIIa mutants (30 nM) alone or in the presence of sTF (10 nM FVIIa, 250 nM sTF for S-2288 and Chromozym t-PA; 30 nM FVIIa, 100 nM sTF for S-2765 and Spectrozyme fXa) were incubated with varying concentrations of chromogenic substrates (ranging from 10 mM to 2µM) in a final volume of 100 µL containing 100 mM HEPES (pH 7.8), 140 mM NaCl, 0.1% PEG-8000, 0.02% Tween-20, and 5 mM CaCl₂. The absorbance of released pNA was monitored at 405 nm on a SpectraMax Plus³⁸⁴ microplate reader (Molecular Devices) at ambient temperature. Conversion to µmol/min was calculated using µmol pNA/min = (0.0417) x (mOD₄₀₅/min); the conversion factor was determined with a standard curve of pNA in 100 µL of the same buffer. Initial rate data were fitted to the Michaelis-Menten equation using Kaleidagraph (Synergy Software) and K_m and V_max values were determined from the averages of three independent determinations.

FVIIa proteolytic activation assay
FVIIa or FVIIa mutant (1 nM) and 0.4 nM TF_1–243 relipidated in phosphotidylcholine/phosphotidylserine (PC/PS) vesicles, 70/30 was mixed with varying concentrations of FX (1000 nM to 0.5 nM) in a final volume of 100 µL containing 20 mM HEPES (pH 7.4), 150 mM NaCl, 5 mM CaCl₂, and 0.5 mg/mL BSA. After a 3-min incubation the reaction was quenched with 40 mM EDTA; controls were run to determine that rates were linear with different quench times, indicating that substrate depletion did not occur. Spectrozyme fXa was added to yield a final concentration and volume of 0.5 mM and 200 µL, respectively. The amount of generated FXa was determined by monitoring the OD₄₀₅/min on a SpectraMax Plus³⁸⁴ microplate reader at ambient temperature. Initial rate data were fitted to the Michaelis-Menten equation using Kaleidagraph and K_m and V_max determined from the averages of three independent determinations.

Due to the high K_m value for FX in the absence of negatively charged phospholipid vesicles and TF, the activity of FVIIa and mutants were measured at one fixed FX concentration as described above. The proteolytic activity was tested with 100 nM FVIIa alone, 10 nM FVIIa with 100 nM sTF, and 10 nM FVIIa with 0.5 mM PC/PS (70/30) phospholipid vesicles. S-2765 was used as a chromogenic substrate to determine the amount of FXa generated. Initial rates were obtained as above and used to calculate the relative activity of the FVIIa mutants compared to wild type. Background activity of FVIIa mutants toward S-2765 was subtracted prior to comparison.

Binding of FVIIa mutants to sTF by surface plasmon resonance
The effects of the mutations in FVIIa upon binding to sTF were determined by surface plasmon resonance measurements on a Biacore 3000 instrument (Biacore). Soluble TF was immobilized on a CM5 sensor chip surface by coupling through free amino groups. The carboxylated dextran matrix was first activated with a mixture of N-hydroxysuccinimide (NHS) and 1-ethyl-3-(1-dimethylamino-propyl)-carbodiimide (EDC) using a protocol provided by the manufacturer. A 20-µL injection of 50 µg/mL of sTF in 10 mM sodium acetate (pH 4) at a flow rate of 5 µL/min resulted in the immobilization of 750 resonance units above baseline. Unreacted NHS was blocked by injection of 35 µL 1 M ethanolamine. The affinity of FVIIa variants for sTF was calculated from binding kinetics to immobilized sTF. For a 1:1 binding interaction, A + B AB, where K_D = k_off/k_on. The dissociation rate constant (k_off) was determined by analyzing the response curve observed upon return to buffer flow for 6 min after saturation with various concentrations of FVIIa. Association rate constants (k_s) were calculated by using a series of seven FVIIa concentrations ranging from 6.125 nM to 400 nM in twofold increments. One hundred microliters of each sample was injected and k_on was determined from the concentration dependence of k_s. A flow rate of 5 µL/min was employed for all kinetics measurements with buffer containing 20 mM Tris (pH 7.5), 100 mM NaCl, 5 mM CaCl₂, 0.05% Tween 20, and 0.01% NaN₃. The sensor chip surface was regenerated by elution of bound FVIIa with an injection of 50 mM EDTA. Kinetic constants were determined by nonlinear regression analysis using software supplied by the manufacturer.

Clotting activity in human FVII-deficient plasma
FVIIa and FVIIa mutants were diluted to a concentration of 5 µg/mL directly into FVII-deficient plasma from three different donors, lots 523b1 and N2521 (George King Bio-Medical) and lot 707/045 (American Diagnostica), all having <1% FVII. Each stock was further diluted with additional FVII-deficient plasma to a final concentration range of 5 µg/mL to 5 pg/mL FVIIa in the plasma. In an ACL 6000 coagulometer (Beckman Coulter), one part plasma ± FVIIa was mixed with two parts Innovin (Dade) pro-thrombin time reagent (recombinant human tissue factor with phospholipids and CaCl₂). Clot formation was detected optically and time to clotting measured. Clotting time (seconds) was compared to mean clotting time of FVII-deficient plasma alone, which had a clotting time of ~90 sec, and plotted as a fractional reduction in clotting time versus FVIIa concentration.

	Electronic supplemental material

TOP Abstract Introduction Results and Discussion Materials and methods Electronic supplemental material References

 
Figure legend: Trypsin digestion of FVIIa produced a Gla domain peptide (Tryptic peptide: ANAFLXLRPGSLXRXCKXXQCS FXXARXIFK where X = Gla) containing nine of the 10 sites of potential Gla modification. Representative MALDI-TOF analysis shows similar patterns and extents of modification between the wild-type and S139C:V157C mutant. Other mutants displayed the same pattern.

	Footnotes

 
Supplemental material: see www.proteinscience.org

	Acknowledgments

 
We acknowledge Martin Roberge for initial constructs, Lynne Giere and Craig Crowley for assistance with mammalian cell culture expression, Paul Moran for phospholipid vesicles and relipidated TF, and Bob Kelley for assistance with surface plasmon resonance and the DNA synthesis and sequencing groups. We also thank Bob Kelley and Daniel Kirchhofer for helpful discussions.

	References

TOP Abstract Introduction Results and Discussion Materials and methods Electronic supplemental material References

 
Banner, D.W., D’Arcy, A., Chène, C., Winkler, F.K., Guha, A., Konigsberg, W.H., Nemerson, Y., and Kirchhofer, D. 1996. The crystal structure of the complex of blood coagulation factor VIIa with soluble tissue factor. Nature 380: 41–46.[CrossRef][Medline]

Boose, J.A., Kuismanen, E., Gerard, R., Sambrook, J., and Gething, M.-J. 1989. The single-chain form of tissue-type plasminogen activator has catalytic activity: Studies with a mutant enzyme that lacks the cleavage site. Biochemistry 28: 635–643.[CrossRef][Medline]

Broze, Jr., G.J. 1992. Tissue factor pathway inhibitor and the revised hypothesis of blood coagulation. Trends Cardiovasc. Med. 2: 72–77.[CrossRef]

Butenas, S. and Mann, K.G. 2003. Response: Mechanism of action of high-dose factor VIIa. Arterioscler. Thromb. Vasc. Biol. 23: 10.[Free Full Text]

Butenas, S., Ribarik, N., and Mann, K.G. 1993. Synthetic substrates for human factor VIIa and factor VIIa-tissue factor. Biochemistry 32: 6531–6538.[CrossRef][Medline]

Butenas, S., Brummel, K.E., Bouchard, B.A., and Mann, K.G. 2003a. How factor VIIa works in hemophilia. J. Thromb. Haemost. 1: 1158–1160.[CrossRef][Medline]

———. 2003b. Influence of factor VIIa and phospholipids on coagulation in "acquired" hemophilia. Arterioscler. Thromb. Vasc. Biol. 23: 123–129.[Abstract/Free Full Text]

Davie, E.W. 1995. Biochemical and molecular aspects of the coagulation cascade. Thromb. Haemost. 74: 1–6.[Medline]

Davie, E.W., Fujikawa, K., and Kisiel, W. 1991. The coagulation cascade: Initiation, maintenance, and regulation. Biochemistry 30: 10363–10370.[CrossRef][Medline]

Dennis, M.S. and Lazarus, R.A. 1994. Kunitz domain inhibitors of tissue factor–factor VIIa. I. Potent inhibitors selected from libraries by phage display. J. Biol. Chem. 269: 22129–22136.[Abstract/Free Full Text]

Dennis, M.S., Eigenbrot, C., Skelton, N.J., Ultsch, M.H., Santell, L., Dwyer, M.A., O’Connell, M.P., and Lazarus, R.A. 2000. Peptide exosite inhibitors of factor VIIa as anticoagulants. Nature 404: 465–470.[CrossRef][Medline]

Dickinson, C.D. and Ruf, W. 1997. Active site modification of factor VIIa affects interactions of the protease domain with tissue factor. J. Biol. Chem. 272: 19875–19879.[Abstract/Free Full Text]

Dickinson, C.D., Kelly, C.R., and Ruf, W. 1996. Identification of surface residues mediating tissue factor binding and catalytic function of the serine protease factor VIIa. Proc. Natl. Acad. Sci. 93: 14379–14384.[Abstract/Free Full Text]

Eigenbrot, C. 2002. Structure, function and activation of coagulation FVII. Curr. Protein Peptide Sci. 3: 287–299.[CrossRef][Medline]

Eigenbrot, C. and Kirchhofer, D. 2002. New insight into how tissue factor allosterically regulates factor VIIa. Trends Cardiovasc. Med. 12: 19–26.[CrossRef][Medline]

Eigenbrot, C., Kirchhofer, D., Dennis, M.S., Santell, L., Lazarus, R.A., Stamos, J., and Ultsch, M.H. 2001. The factor VII zymogen structure reveals reregistration of -strands during activation. Structure 9: 627–636.[Medline]

Giesen, P.L.A., Rauch, U., Bohrmann, B., Kling, D., Roque, M., Fallon, J.T., Badimon, J.J., Himber, J., Riederer, M.A., and Nemerson, Y. 1999. Blood-borne tissue factor: Another view of thrombosis. Proc. Natl. Acad. Sci. 96: 2311–2315.[Abstract/Free Full Text]

Gráf, L. 1995. Structural basis of serine protease action: The fourth dimension. In Natural sciences and human thought (ed. R. Zwilling), pp. 139–148. Springer-Verlag, Heidelberg, Germany.

Harvey, S.B., Stone, M.D., Martinez, M.B., and Nelsestuen, G.L. 2003. Mutagenesis of the -carboxyglutamic acid domain of human factor VII to generate maximum enhancement of the membrane contact site. J. Biol. Chem. 278: 8363–8369.[Abstract/Free Full Text]

Hedner, U. 2001. Recombinant factor VIIa (Novoseven®) as a hemostatic agent. Semin. Hematol. 38(Suppl. 12): 43–47.[CrossRef][Medline]

———. 2004. Dosing with recombinant factor VIIa based on current evidence. Semin. Hematol. 41(Suppl. 1): 35–39.[CrossRef][Medline]

Higashi, S., Matsumoto, N., and Iwanaga, S. 1996. Molecular mechanism of tissue factor-mediated acceleration of factor VIIa activity. J. Biol. Chem. 271: 26569–26574.[Abstract/Free Full Text]

Huber, R. and Bode, W. 1978. Structural basis of the activation and action of trypsin. Acc. Chem. Res. 11: 114–122.[CrossRef]

Kelley, R.F., Costas, K.E., O’Connell, M.P., and Lazarus, R.A. 1995. Analysis of the factor VIIa binding site on human tissue factor: Effects of tissue factor mutations on the kinetics and thermodynamics of binding. Biochemistry 34: 10383–10392.[CrossRef][Medline]

Kelley, R.F., Yang, J., Eigenbrot, C., Moran, P., Peek, M., Lipari, M.T., and Kirchhofer, D. 2004. Similar molecular interactions of factor VII and factor VIIa with the tissue factor region that allosterically regulates enzyme activity. Biochemistry 43: 1223–1229.[CrossRef][Medline]

Kemball-Cook, G., Johnson, D.J.D., Tuddenham, E.G.D., and Harlos, K. 1999. Crystal structure of active site-inhibited human coagulation factor VIIa (des-Gla). J. Struct. Biol. 127: 213–223.[CrossRef][Medline]

Kunkel, T.A., Roberts, J.D., and Zakour, R.A. 1987. Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods Enzymol. 154: 367–382.[Medline]

Lawson, J.H. and Murphy, M.P. 2004. Challenges for providing effective hemostasis in surgery and trauma. Semin. Hematol. 41(Suppl. 1): 55–64.[CrossRef][Medline]

Lijnen, H.R., Van Hoef, B., Nelles, L., and Collen, D. 1990. Plasminogen activation with single-chain urokinase-type plasminogen activator (scu-PA). Studies with active site mutagenized plasminogen (Ser⁷⁴⁰ Ala) and plasmin-resistant scu-PA (Lys¹⁵⁸ Glu). J. Biol. Chem. 265: 5232–5236.[Abstract/Free Full Text]

Lisman, T. and de Groot, P.G. 2003. Mechanism of action of recombinant factor VIIa. J. Thromb. Haemost. 1: 1138–1139.[CrossRef][Medline]

Lisman, T., de Groot, P.G., Lambert, T., Rojkjaer, R., and Persson, E. 2003. Enhanced in vitro procoagulant and antifibrinolytic potential of superactive variants of recombinant factor VIIa in severe hemophilia A. J. Thromb. Haemost. 1: 2175–2178.[CrossRef][Medline]

Lucas, B.K., Giere, L.M., DeMarco, R.A., Shen, A., Chisholm, V., and Crowley, C.W. 1996. High-level production of recombinant proteins in CHO cells using a dicistronic DHFR intron expression vector. Nucleic Acids Res. 24: 1774–1779.[Abstract/Free Full Text]

Mann, K.G. 2003. Thrombin formation. Chest 124: 4S–10S.[Abstract/Free Full Text]

Midathada, M.V., Mehta, P., Waner, M., and Fink, L.M. 2004. Recombinant factor VIIa in the treatment of bleeding. Am. J. Clin. Pathol. 121: 124–137.[CrossRef][Medline]

Monroe, D.M. and Roberts, H.R. 2003. Mechanism of action of high-dose factor VIIa: Points of agreement and disagreement. Arterioscler. Thromb. Vasc. Biol. 23: 8–9.[Free Full Text]

Nelsestuen, G.L., Stone, M., Martinez, M.B., Harvey, S.B., Foster, D., and Kisiel, W. 2001. Elevated function of blood clotting factor VIIa mutants that have enhanced affinity for membranes. Behavior in a diffusion-limited reaction. J. Biol. Chem. 276: 39825–39831.[Abstract/Free Full Text]

Nemerson, Y. 1988. Tissue factor and hemostasis. Blood 71: 1–8.[Free Full Text]

Neuenschwander, P.F. and Morrissey, J.H. 1994. Roles of the membrane-interactive regions of Factor VIIa and tissue factor. J. Biol. Chem. 269: 8007–8013.[Abstract/Free Full Text]

Neuenschwander, P.F., Branam, D., and Morrissey, J.H. 1993. Importance of substrate composition, pH and other variables on tissue factor enhancement of factor VIIa activity. Thromb. Haemost. 70: 970–977.[Medline]

Olsen, O.H., Nielsen, P.F., and Persson, E. 2004. Prevention of strand movement into a zymogen-like position does not confer higher activity to coagulation factor VIIa. Biochemistry 43: 14096–14103.[CrossRef][Medline]

Pasternak, A., Liu, X., Lin, T.-Y., and Hedstrom, L. 1998. Activating a zymogen without proteolytic processing: Mutation of Lys15 and Asn194 activates trypsinogen. Biochemistry 37: 16201–16210.[CrossRef][Medline]

Persson, E. 2004. Variants of recombinant factor VIIa with increased intrinsic activity. Semin. Hematol. 41(Suppl. 1): 89–92.[CrossRef][Medline]

Persson, E. and Olsen, O.H. 2002. Assignment of molecular properties of a superactive coagulation factor VIIa variant to individual amino acid changes. Eur. J. Biochem. 269: 5950–5955.[Medline]

Persson, E., Bak, H., and Olsen, O.H. 2001a. Substitution of valine for leucine 305 in factor VIIa increases the intrinsic enzymatic activity. J. Biol. Chem. 276: 29195–29199.[Abstract/Free Full Text]

Persson, E., Kjalke, M., and Olsen, O.H. 2001b. Rational design of coagulation factor VIIa variants with substantially increased intrinsic activity. Proc. Natl. Acad. Sci. 98: 13583–13588.[Abstract/Free Full Text]

Persson, E., Nielsen, L.S., and Olsen, O.H. 2001c. Substitution of aspartic acid for methionine-306 in factor VIIa abolishes the allosteric linkage between the active site and the binding interface with tissue factor. Biochemistry 40: 3251–3256.[CrossRef][Medline]

Persson, E., Bak, H., Ostergaard, A., and Olsen, O.H. 2004. Augmented intrinsic activity of Factor VIIa by replacement of residues 305, 314, 337 and 374: Evidence of two unique mutational mechanisms of activity enhancement. Biochem. J. 379: 497–503.[CrossRef][Medline]

Petrovan, R.J. and Ruf, W. 2001. Residue Met156 contributes to the labile enzyme conformation of coagulation factor VIIa. J. Biol. Chem. 276: 6616–6620.[Abstract/Free Full Text]

———. 2002. Role of zymogenicity-determining residues of coagulation factor VII/VIIa in cofactor interaction and macromolecular substrate recognition. Biochemistry 41: 9302–9309.[CrossRef][Medline]

Pike, A.C.W., Brzozowski, A.M., Roberts, S.M., Olsen, O.H., and Persson, E. 1999. Structure of human factor VIIa and its implications for the triggering of blood coagulation. Proc. Natl. Acad. Sci. 96: 8925–8930.[Abstract/Free Full Text]

Rapaport, S.I. and Rao, L.V.M. 1995. The tissue factor pathway: How it has become a "Prima Ballerina." Thromb. Haemost. 74: 7–17.[Medline]

Reyda, S., Sohn, C., Klebe, G., Rall, K., Ullmann, D., Jakubke, H.D., and Stubbs, M.T. 2003. Reconstructing the binding site of factor Xa in trypsin reveals ligand-induced structural plasticity. J. Mol. Biol. 325: 963–977.[CrossRef][Medline]

Ruf, W. and Dickinson, C.D. 1998. Allosteric regulation of the cofactor-dependent serine protease coagulation factor VIIa. Trends Cardiovasc. Med. 8: 350–356.[CrossRef][Medline]

Seymour, J.L., Lindquist, R.N., Dennis, M.S., Moffat, B., Yansura, D., Reilly, D., Wessinger, M.E., and Lazarus, R.A. 1994. Ecotin is a potent anticoagulant and reversible tight-binding inhibitor of factor Xa. Biochemistry 33: 3949–3958.[CrossRef][Medline]

Sichler, K., Banner, D.W., D’Arcy, A., Hopfner, K.P., Huber, R., Bode, W., Kresse, G.B., Kopetzki, E., and Brandstetter, H. 2002. Crystal structures of uninhibited factor VIIa link its cofactor and substrate-assisted activation to specific interactions. J. Mol. Biol. 322: 591–603.[CrossRef][Medline]

Soejima, K., Mizuguchi, J., Yuguchi, M., Nakagaki, T., Higashi, S., and Iwanaga, S. 2001. Factor VIIa modified in the 170 loop shows enhanced catalytic activity but does not change the zymogen-like property. J. Biol. Chem. 276: 17229–17235.[Abstract/Free Full Text]

Soejima, K., Yuguchi, M., Mizuguchi, J., Tomokiyo, K., Nakashima, T., Nakagaki, T., and Iwanaga, S. 2002. The 99 and 170 loop-modified factor VIIa mutants show enhanced catalytic activity without tissue factor. J. Biol. Chem. 277: 49027–49035.[Abstract/Free Full Text]

Szabó, E., Böcskei, Z., Náray-Szabó, G., and Gráf, L. 1999. The three-dimensional structure of Asp189Ser trypsin provides evidence for an inherent structural plasticity of the protease. Eur. J. Biochem. 263: 20–26.[Medline]

Szabó, E., Venekei, I., Böcskei, Z., Náray-Szabó, G., and Gráf, L. 2003. Three dimensional structures of S189D chymotrypsin and D189S trypsin mutants: The effect of polarity at site 189 on a protease-specific stabilization of the substrate-binding site. J. Mol. Biol. 331: 1121–1130.[CrossRef][Medline]

Toso, R., Bernardi, F., Tidd, T., Pinotti, M., Camire, R.M., Marchetti, G., High, K.A., and Pollak, E.S. 2003. Factor VII mutant V154G models a zymogen-like form of factor VIIa. Biochem. J. 369: 563–571.[CrossRef][Medline]

Tranholm, M., Kristensen, K., Kristensen, A.T., Pyke, C., Rojkjaer, R., and Persson, E. 2003. Improved hemostasis with superactive analogs of factor VIIa in a mouse model of hemophilia A. Blood 102: 3615–3620.[Abstract/Free Full Text]

Zhang, E., St. Charles, R., and Tulinsky, A. 1999. Structure of extracellular tissue factor complexed with Factor VIIa inhibited with a BPTI mutant. J. Mol. Biol. 285: 2089–2104.[CrossRef][Medline]

CiteULike