Probing the interface between factor Xa and tissue factor in the quaternary complex tissue factor–factor VIIa–factor Xa–tissue factor pathway inhibitor Karin Carlsson 1 , Per-Ola Freskga ˚ rd 2, *, Egon Persson 3 , Uno Carlsson 1 and Magdalena Svensson 1 1 IFM-Department of Chemistry, Linko ¨ ping University, Linko ¨ ping, Sweden; 2 Protein Biotechnology, Novo Nordisk A/S, Novo Alle ´ , Bagsværd, Denmark; 3 Haemostasis Biology, Novo Nordisk A/S, Novo Nordisk Park, Ma ˚ løv, Denmark Blood coagulation is triggered by the formation of a complex between factor VIIa (FVIIa) and its cofactor, tissue factor (TF). TF–FVIIa is inhibited by tissue factor pathway inhibitor (TFPI) in two steps: first TFPI is bound to the active site of factor Xa (FXa), and subsequently FXa–TFPI exerts feedback inhibition of TF–FVIIa. The FXa-depend- ent inhibition of TF–FVIIa activity by TFPI leads to for- mation of the quaternary complex TF–FVIIa–FXa–TFPI. We used site-directed fluorescence probing to map part of the region of soluble TF (sTF) that interacts with FXa in sTF–FVIIa–FXa–TFPI. We found that the C-terminal region of sTF, including positions 163, 166, 200 and 201, is involved in binding to FXa in the complex, and FXa, most likely via its Gla domain, is also in contact with the Gla domain of FVIIa in this part of the binding region. Fur- thermore, a region that includes the N-terminal part of the TF2 domain and the C-terminal part of the TF1 domain, i.e. the residues 104 and 197, participates in the interaction with FXa in the quaternary complex. Moreover, comparisons of the interaction areas between sTF and FX(a) in the quater- nary complex sTF–FVIIa–FXa–TFPI and in the ternary complexes sTF–FVII–FXa or sTF–FVIIa–FX demonstra- ted large similarities. Keywords: fluorescence; local probing; protein–protein interactions; site-directed labeling. Complex formation between factor VIIa (FVIIa) and its cofactor tissue factor (TF) triggers blood coagulation. The TF–FVIIa complex activates factor X (FX) to factor Xa (FXa) and factor IX (FIX) to factor IXa (FIXa), which both contribute to the formation of thrombin and ulti- mately a fibrin clot. Tissue factor pathway inhibitor (TFPI) is the prominent physiological inhibitor of the TF–FVIIa complex [1]. It is a plasma proteinase inhibitor composed of three Kunitz-type domains [2]. TFPI inhibits TF–FVIIa in two steps: first, binding occurs between the second Kunitz domain of TFPI and the active site of FXa; thereafter, the first Kunitz domain of TFPI binds to TF-bound FVIIa, and FXa–TFPI thereby causes feedback inhibition of TF–FVIIa. FXa-mediated TFPI-induced inhibition of TF–FVIIa catalytic activity results in formation of the stable quaternary complex TF–FVIIa–FXa–TFPI [3]. In addition to down-regulating the procoagulant function of TF–FVIIa, the TF–FVIIa–FXa–TFPI complex is also crucial for the cell-surface redistribution of inhibited TF–FVIIa complexes into caveolae [4], and the rate of internalization of TF–FVIIa is increased in some cell types when part of the quaternary complex with FXa–TFPI [5]. The role of the third Kunitz domain of TFPI is not yet fully understood, although it is known that TFPI(1–161) lacking this region can inhibit TF–FVIIa in complex with FXa [6]. According to the X-ray crystallographic structure, FVIIa binds to the extracellular domain of TF in an extended conformation comprising contacts all the way from the lower part of the protease domain of FVIIa and the N-terminal domain of TF to the c-carboxyglutamic acid (Gla)-rich FVIIa module and the C-terminal part of TF [7]. Several attempts have been made using mutagenesis com- bined with functional studies to map ternary complexes that include TF–FVIIa/FVII and FX/FXa. The results of such investigations have suggested that an extended region of TF–FVIIa is involved in recognition of the macromolecular substrate FX. In addition to the catalytic cleft of the enzyme, the Gla domain of FVIIa has been implicated as an essential component in the activation of FX by TF–FVIIa, through either direct or indirect interaction [8–10]. It has also been proposed that an extensive area of the C-terminal domain of TF is involved in substrate recognition Correspondence to U. Carlsson, IFM-Department of Chemistry, Linko ¨ ping University, SE-581 83 Linko ¨ ping, Sweden. Fax: + 46 13281399, Tel.: + 46 13281714, E-mail: ucn@ifm.liu.se or M. Svensson, IFM-Department of Chemistry, Linko ¨ ping University, SE-581 83 Linko ¨ ping, Sweden. Fax: + 46 13281399, Tel.: + 46 13285686, E-mail: msv@ifm.liu.se Abbreviations: EGF1, first EGF-like domain; EGF2, second EGF-like domain; FVII/FVIIa, factor VII/factor VIIa; FX/FXa, factor X/factor Xa; Gla, c-carboxyglutamic acid; IAEDANS, 5-({[(2-iodo- acetyl)amino]ethyl}amino)naphthalene-1-sulfonic acid; PtdCho, phosphatidylcholine; PtdSer, phosphatidylserine; PD, protease domain; sTF, soluble tissue factor (1–219); TF, tissue factor; TF1, N-terminal domain of TF; TF2, C-terminal domain of TF; TFPI, tissue factor pathway inhibitor; TFPI(1–161), TFPI lacking the third Kunitz domain. *Present address: Nuevolution A/S, Rønnegade 8, Copenhagen, Denmark. (Received 7 February 2003, revised 15 April 2003, accepted 17 April 2003) Eur. J. Biochem. 270, 2576–2582 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03625.x [8,9,11–16], and the same area has been found to be important for interaction with FXa during FVII activation [17]. However, the interface between TF and FXa in the quaternary complex TF–FVIIa–FXa–TFPI has not been characterized as extensively as in the ternary complexes. Functional studies have identified only two residues in TF (K165 and K166) as being important for inhibition of the TF–FVIIa complex by FXa–TFPI [18]. Moreover, research has shown that the light chain of FXa facilitates the interaction between FXa–TFPI and TF–FVIIa [19], and that the Gla domain of FXa is essential for inhibition of TF–FVIIa activity by FXa–TFPI [20]. As a contact region between TF and FXa has not yet been thoroughly elucidated in TF–FVIIa–FXa–TFPI(1–161), we decided to map the interface between soluble TF (sTF) and FXa in this quaternary complex by introducing environmentally sensi- tive probes into specific positions in sTF (Fig. 1). These positions were selected, according to the X-ray structure of the sTF–FVIIa complex [7], not to be in direct contact with FVIIa. We produced a set of surface-exposed single-cysteine mutants of sTF, and we then covalently attached fluorescent labels to the cysteine sulfhydryl groups. The quaternary sTF–FVIIa–FXa–TFPI(1–161) complex was stabilized by adding phospholipid vesicles (phosphatidylcholine/phos- phatidylserine) to allow fluorescence measurements. With this powerful technique, local changes in polarity can be monitored at a resolution of individual amino acid residues during the formation of the complex, making it possible to map the contact region between sTF and FXa in the quaternary complex sTF–FVIIa–FXa–TFPI(1–161). Materials and methods Chemicals Phosphatidylcholine (PtdCho, 80%)/phosphatidylserine (PtdSer, 20%) were from Sigma (St. Louis, MO, USA) and vesicles were prepared essentially as described by just omitting the TF apoprotein [21]. The fluorescent label 5-({[(2-iodoacetyl)amino]ethyl}amino)naphthalene-1-sulfonic acid (1,5-IAEDANS) was purchased from Molecular Probes (Eugene, OR, USA). All other chemicals were of analytical grade. Proteins and labeling The QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA) was used to mutate in sTF and mutants were expressed in Escherichia coli and purified using Q Sepharose and FVIIa affinity chromatography as previously described [22]. The protein concentrations were calculated from absorption measurements using e 280 ¼ 37440 M )1 Æcm )1 for all sTF variants, except for sTF(Y156C) where e 280 ¼ 36160 M )1 Æcm )1 was used [23]. Light absorption measurements were performed on a Hitachi U-2000 spectrophotometer. Fluorescence labeling of the single cysteine mutants was carried out as previously described [22]. The isolation of human recombinant FVIIa [24], recombinant TFPI(1–161) [25], and preparation of FVIIa affinity matrix [26] have been described. Human FXa was purchased from Enzyme Research Laboratories (South Bend, IN, USA). Characterization of the cofactor properties of the sTF variants An amidolytic activity assay using 1 m M S-2288 (Chromo- genix, Mo ¨ lndal, Sweden) was performed to determine the affinity of the sTF variants for FVIIa and their ability to stimulate FVIIa. The activity of FVIIa was measured without and in the presence of various concentrations of sTF variant (2–320 n M )in50m M Hepes, pH 7.4, contain- ing 0.1 M NaCl, 5 m M CaCl 2 and 1 mgÆmL )1 bovine serum albumin. These measurements were performed twice. The concentration of sTF variant required for half-maximal enhancement of FVIIa activity was used to calculate the apparent dissociation constant. The maximal enhancement obtained with the sTF mutants was compared with that obtained with the wild-type cofactor and is hereafter referred to as the cofactor activity. The ability of the sTF variants to support FX activation in the presence of a membrane surface (PtdCho/PtdSer Fig. 1. Structure of the sTF–FVIIa complex [7]. sTF is displayed in space filling and FVIIa in C-alpha ribbon. The TF1 domain is in dark grey and the TF2 domain in light grey. The Gla domain of FVIIa is showningreen;EGF1inmagenta;EGF2inlightblue;PDindark blue. The sTF residues highlighted in red are located within the area of interaction with FXa in the sTF–FVIIa–FXa–TFPI(1–161) complex according to the fluorescence probing, whereas the two sTF residues in yellow gave no significant fluorescence response upon FXa–TFPI(1– 161) binding. Ó FEBS 2003 The sTF–FXa interface in sTF–FVIIa–FXa–TFPI (Eur. J. Biochem. 270) 2577 vesicles, total phospholipid concentration 5 l M ) was studied by mixing 0.1 n M FVIIa, 200 n M FX, and 200 n M sTF variant in the buffer above. After a 10-min incubation, FXa generation was stopped by excess EDTA and quantified by the addition of 0.5 m M S-2765 (Chromogenix, Mo ¨ lndal, Sweden). Each measurement was duplicated. Fluorescence measurements Fluorescence emission spectra were recorded on a Hitachi F-4500 spectrophotometer with a thermostated cell com- partment at a constant temperature of 23 °C. All measure- ments were carried out using a 0.5-cm quartz cell and the slits were set to 5 nm for both excitation and emission. Fluorescence emission spectra for the IAEDANS-labeled protein were recorded in the wavelength region 400–600 nm after excitation at 350 nm. A spectrum was collected for each labeled sTF-variant at a concentration of 0.3 l M . Another spectrum was recorded for a sample containing FVIIa in a 1.5-fold molar excess over labeled sTF. TFPI(1)161) and FXa were incubated for 30 min in the presence of 5 m M CaCl 2 , followed by the addition of sTF and FVIIa. A spectrum was recorded for this mixture with the final concentrations of 0.3 l M sTF, 0.45 l M FVIIa, 0.45 l M FXa, and 2.25 l M TFPI(1–161). All reaction mixtures contained 50 l M PtdCho/PtdSer in 50 m M Hepes, 0.15 M NaCl, 5 m M CaCl 2 ,pH7.5,and the samples were incubated for 30 min. Each scan was reproduced at least three times and the first derivative was used to find the wavelength of the fluorescence emission maximum. Results We created eight sTF variants, each with one position mutated to a cysteine. We labeled the variants with the fluorescent probe IAEDANS so that we could monitor the binding of FXa by studying changes in the fluorescence emission spectra upon AEDANS-sTF–FVIIa–FXa– TFPI(1–161) complex formation. In this study, we focused on the Stokes’ shifts caused by alterations in the environ- ment around the probe when it came in contact with the surface of FXa. Detection of such changes provides more reliable results compared to detection of small changes in fluorescence intensity, because the wavelength shifts of the emission maximum do not depend on changes in concen- trations. The effects on sTF-dependent FVIIa activity and activation of FX as a result of mutation and labeling of sTF were also investigated. The ability of IAEDANS-labeled sTF variants to support FVIIa activity We performed a set of experiments to examine the impact of mutation and labeling of sTF on FVIIa binding and enhancement of FVIIa activity. The ability of the IAEDANS-labeled sTF variants to stimulate FVIIa was assessed and the K d values for FVIIa binding to AEDANS-sTF were calculated (Table 1). These data show that the AEDANS-sTF mutants maintained virtually normal binding to FVIIa and normal cofactor activity (Table 1). Effects on fluorescence spectra upon FVIIa and TFPI(1–161) binding We conducted a series of measurements to ascertain whether any fluorescence emission shifts accompanied FVIIa binding to AEDANS-sTF, that is, we examined possible changes in the environment surrounding the fluorescent probe as a result of AEDANS-sTF–FVIIa complex formation (Table 1). Three of the labeled sTF variants (Y156C, S163C and K201C) showed a change in emission spectra and in the rest of the mutants the probe was not significantly affected by the sTF–FVIIa complex formation. As the purpose of our study was to map the interaction between FXa and sTF in the quaternary sTF–FVIIa–FXa–TFPI(1–161) complex, we also investi- gated whether TFPI(1)161) contributed to the emission changes sensed by the probe after addition of the FXa– TFPI(1–161) complex to AEDANS-sTF–FVIIa. In this control experiment, we used a concentration of TFPI(1– 161) that was 250 times higher than the IC 50 value that Hamomoto et al. [6] reported when using TFPI(1–161) to inhibit FVIIa bound to relipidated TF, which assured stoichiometric binding of TFPI(1–161) to AEDANS- sTF–FVIIa. Comparison of the emission spectra of AEDANS-sTF–FVIIa with the corresponding spectra after addition of TFPI(1–161) alone demonstrated that the spectral shifts detected upon FXa–TFPI(1–161) binding to AEDANS-sTF–FVIIa did not originate from TFPI(1–161) (data not shown). Effects on fluorescence spectra upon FXa binding In the sTF–FVIIa complex, positions 156, 163 and 166 in sTF are adjacent to the Gla domain of FVIIa, and positions 200 and 201 are situated near EGF1 and the so-called hydrophobic stack, which is a linker between the Gla domain and EGF1 of FVIIa (Fig. 1). All five of the mentioned residues are located in the TF2 domain [7]. Three of our mutants labeled in this region, sTF(S163C), Table 1. Functional binding of IAEDANS-labeled sTF to FVIIa and fluorescence emission shifts caused by formation of the AEDANS-sTF– FVIIa complex. The K d(app) for wt-sTF is 7 n M [37]. The cofactor activity is set to 100% for wt-sTF. sTF variant K d(app) a (n M ) Cofactor activity a (%) Dk Fmax b (nm) E99C 10.6 ± 0.6 90 ± 3 ) 1.5 ± 0.0 L104C 3.8 ± 0.5 98 ± 6 + 0.3 ± 0.1 Y156C 4.0 ± 0.2 105 ± 2 ) 4.4 ± 0.2 S163C 3.8 ± 0.7 112 ± 6 ) 8.7 ± 0.1 K166C 8.3 ± 0.4 112 ± 5 + 0.8 ± 0.3 T197C 7.1 ± 1.0 93 ± 4 ) 0.4 ± 0.5 R200C 5.0 ± 0.6 100 ± 1 ) 0.3 ± 0.6 K201C 8.9 ± 0.9 90 ± 3 ) 15.0 ± 0.2 a Calculated from amidolytic activity stimulation. The standard error of the mean is given. b Dk Fmax ¼ k Fmax (AEDANS-sTF– FVIIa) – k Fmax (AEDANS-sTF). The calculated pooled standard deviations are also given. 2578 K. Carlsson et al. (Eur. J. Biochem. 270) Ó FEBS 2003 sTF(K166C), and sTF(R200C), displayed a blue shift of the emission upon binding of FXa in the AEDANS-labeled sTF–FVIIa–FXa–TFPI(1–161) complex (Fig. 2; Table 2). On the other hand, FXa binding to AEDANS-sTF(K201C) resulted in a red shift of the spectrum, probably due to interaction with a charged group in FXa. The attached label in position 156 did not sense any significant change in the environment when FXa was associated with sTF in the quaternary complex. Residues E99 and L104 are located in the TF1 domain, and T197 is situated in the N-terminal portion of the TF2 domain, i.e. close to the TF1 domain at the level of EGF1 in FVIIa in the sTF–FVIIa complex (Fig. 1). Binding of FXa caused blue shifts in the fluorescence emission spectra of AEDANS-sTF(L104C) and AEDANS-sTF(T197C) but did not affect the emission spectrum of AEDANS- sTF(E99C) upon formation of the quaternary complex (Fig. 2; Table 2). Fig. 2. Fluorescence emission spectra of the IAEDANS-labeled sTF variants upon complex formation. In complex with FVIIa (grey line) and when part of the AEDANS-sTF–FVIIa– FXa–TFPI(1–161) complex (black line). Representative spectra are shown. Table 2. Fluorescence emission shifts occurring upon formation of the AEDANS-sTF–FVIIa–FXa–TFPI(1–161) complex and effects of mutation and labeling of sTF on sTF–FVIIa-catalyzed activation of FX. sTF variant Dk Fmax a (nm) FX activation rate (v mutant /v wt ) AEDANS-sTF b sTF b E99C ) 0.8 ± 0.2 0.93 ± 0.03 0.90 ± 0.01 L104C ) 4.8 ± 0.3 0.86 ± 0.04 1.00 ± 0.07 Y156C ) 1.0 ± 0.2 0.11 ± 0.02 0.37 ± 0.04 S163C ) 3.4 ± 0.2 0.01 ± 0.00 0.01 ± 0.00 K166C ) 5.0 ± 0.5 0.03 ± 0.02 0.01 ± 0.01 T197C ) 5.6 ± 0.5 0.57 ± 0.05 0.66 ± 0.02 R200C ) 6.3 ± 0.6 0.32 ± 0.03 0.23 ± 0.00 K201C + 4.6 ± 0.2 0.13 ± 0.01 0.78 ± 0.04 a Dk Fmax ¼ k Fmax [AEDANS-sTF–FVIIa–FXa–TFPI(1–161)] – k Fmax (AEDANS-sTF–FVIIa). The calculated pooled standard deviations are also given. b The standard error of the mean is given. Ó FEBS 2003 The sTF–FXa interface in sTF–FVIIa–FXa–TFPI (Eur. J. Biochem. 270) 2579 The ability of the sTF variants to support FX activation We used a two-stage amidolytic assay to assess the ability of the labeled and unlabeled sTF variants to support FVIIa- catalyzed activation of FX. In Table 2, the cofactor activity of the individual sTF variants is expressed as a ratio between the rates of FX activation with the mutated form of sTF and wild-type sTF, respectively, bound to FVIIa. Replacement of the wild-type residue by Cys decreased the FX activation rate for all of the studied positions in sTF except sTF(E99C) and sTF(L104C) which retained normal cofactor activity. The effect on the FX activation rates as a result of attachment of the fluorescent probe IAEDANS was also monitored by comparing the data with the corresponding values for the unlabeled sTF variant (Table 2). Linking IAEDANStosTF(R200C)resultedinaslightrecoveryof FX activation ability, whereas labeling of sTF(Y156C) and sTF(K201C) led to a significantly decreased ability to stimulate activation of FX. Labeling of the remaining variants did not significantly affect the cofactor activity. The rates of FX activation with IAEDANS-labeled sTF(Y156C), sTF(S163C), sTF(K166C), sTF(R200C), and sTF(K201C) were markedly decreased (to 1–30% of the activation induced by wild-type sTF; Table 2). Similar results have been obtained by Kirchhofer et al. [13] when using mutants with primarily Ala substitutions to achieve functional mapping of the sTF–FX interaction in the sTF– FVIIa–FX ternary complex. Compared with wild-type sTF, AEDANS-sTF(E99C) and AEDANS-sTF(L104C) were nearly as effective in activating FX, and sTF(T197C) was somewhat less efficient (i.e. it induced 57% of the activation obtained with wild-type sTF; Table 2). Discussion Formation of the TF–FVIIa–FXa–TFPI complex is an important event in the regulation of coagulation, because it inhibits the function of TF–FVIIa. The quaternary complex mediates translocation of the inhibited TF–FVIIa complex into caveolae [4], and, in some types of cells, the presence of TFPI influences the rate of internalization and degradation of TF–FVIIa [5]. However, no crystallographic data have been obtained on the large sTF–FVIIa–FXa–TFPI com- plex. From functional studies only limited results have been published regarding the binding area between sTF and FXa in this complex. Regarding the ternary sTF–FVIIa–FX complex other investigators have performed functional analyses to investi- gate the binding region between sTF and FX. Kirchhofer and coworkers [13] reported mutagenesis of surface-exposed residues of sTF and the effect of the modified sTF–FVIIa complex on FX activation. In another study [17], these authors also examined the interaction between FXa and sTF–FVII in a similar way by observing the impact of mutation in sTF on activation of FVII by FXa. The aim of the present work is to map in more detail the interaction area between sTF and FXa in the quaternary complex than previously carried out. This will also allow a comparison of the binding interface between sTF and FXa in the quaternary and ternary complexes. We applied a site- directed fluorescence labeling approach that can determine whether the labeled positions are involved in the interaction between the proteins in the complex. The main advantage of this experimental strategy is that it permits direct mapping of the contact area; by comparison, in methods such as alanine scanning, structural effects caused by mutagenesis can affect the function without the mutation necessarily being located in the binding interface. We selected eight amino acid residues in sTF for individual cysteine replacement and used the cysteines as handles to which we attached the fluorescent probe IAEDANS. The residues were chosen on the basis of X-ray data [7], which showed that they were not in direct contact with FVIIa in the sTF–FVIIa complex. Accord- ingly, the AEDANS-sTF mutants retained virtually normal affinity for FVIIa (Table 1). For five of the mutants, essentially no change in emission spectra could be detected upon formation of the AEDANS-sTF–FVIIa complex (Table 1). However, three of the mutants, AEDANS- sTF(Y156C), AEDANS-sTF(S163C), and AEDANS- sTF(K201C), showed significantly altered emission spectra upon complex formation. This means that the introduced fluorophore was affected by FVIIa association with sTF. However, in this case it is unlikely that the probe is located in the interaction area between FVIIa and sTF, as the binding strength between these two proteins was not decreased for the labeled variants (Table 1). In accordance with this, we have previously observed rather large changes in K d for sTF–FVIIa complexes when containing AEDANS labels within the contact interface [27]. Thus it is likely that the probes in positions 156, 163 and 201 are just pointing towards the border between sTF and FVIIa. Based on our fluorescence emission shift data, it can be seen that part of the C-terminal region of sTF (i.e. positions 163, 166, 200 and 201) interacts with FXa in the sTF– FVIIa–FXa–TFPI(1–161) complex (Figs 1 and 2; Table 2). This agrees well with what Rao and Ruf [18] have suggested for position 166 in TF when part of the quaternary complex based on activity measurements. Interestingly, the observed interaction pattern between sTF and FXa in the quaternary complex seems to be similar in this region to the one proposed for the ternary sTF–FVIIa–FX and sTF–FVII– FXa complexes based on the results of functional studies [11,13,17]. Combining the fluorescence shifts of the AEDANS-sTF–FVIIa and AEDANS-sTF–FVIIa–FXa– TFPI(1–161) complexes (Tables 1 and 2) gives further information that indicates that FXa is in direct contact with or is located very close to the border of the Gla domain of FVIIa around position 163 and the hydrophobic stack around position 201. This is implied because, according to the AEDANS-sTF–FVIIa data, the reporter group from these positions was pointing towards FVIIa, and association of FXa gave rise to an additional spectral shift, thus both FVIIa and FXa should have been in contact with the fluorophore. Based on the structural similarity between FVIIa and FXa [28–30], as well as similar orientation of FVIIa and FXa relative to the membrane [7,31–33], it is plausible that the Gla domain of FXa is involved in binding to the C-terminal domain of sTF. As we found that FXa is bound next to the Gla domain of FVIIa, it is highly likely that the Gla domains of FVIIa and FXa interact with each other in the quaternary complex. Interestingly, previous reports indicate that the Gla domain of FVIIa/FVII also plays a 2580 K. Carlsson et al. (Eur. J. Biochem. 270) Ó FEBS 2003 role in recognition of FX/FXa in the ternary complexes [8–10,34]. On the other hand, the fluorescent label linked to position 156 in the C-terminal part of sTF between positions 163 and 201 (Fig. 1) was not affected by incorporation of FXa into the AEDANS-sTF–FVIIa–FXa–TFPI(1)161) com- plex. Hence, FXa appears to be positioned some distance from the FVIIa Gla domain, probably over the side chain of Tyr156 in sTF, as indicated by substrate activation meas- urements in this study (Table 2). Similar measurements on the ternary sTF–FVIIa–FX and sTF–FVII–FXa complexes indicate that this residue indeed is involved in the FX/FXa contact area [13,17]. Residues E99, L104, and T197 are located very close to each other in an area of the C-terminal part of the TF1 domain and the N-terminal portion of the TF2 domain (Fig. 1). Mutation and labeling of positions 99 and 104 had little effect on the FX-activating ability of the AEDANS- sTF–FVIIa complex. Mutation and labeling of position 197 resulted in a somewhat lower rate of FX activation (Table 2). In addition, when using AEDANS-sTF(E99C), we found no significant change in emission upon formation of the AEDANS-sTF–FVIIa–FXa–TFPI(1)161) complex, indicating that position 99 in sTF is not involved in the binding of FXa. In contrast, for both AEDANS- sTF(L104C) and AEDANS-sTF(T197C), formation of the AEDANS-sTF–FVIIa–FXa–TFPI(1)161) complex resulted in alterations in the environment of the probe, detected as shifts in the emission spectra. These shifts imply that positions 104 and 197 are involved in binding to FXa in the quaternary complex, although the two residues, in particular 104, are not essential for activation of the substrate FX. FXa and FX seem to bind to sTF in a similar way in the ternary sTF–FVII–FXa/FX complexes [13,17,35]. Provided that in this region FX in the ternary complex does not bind differently than FXa in the quaternary sTF–FVIIa–FXa–TFPI(1–161) complex the results of our activity experiments and our direct fluores- cence binding studies for position 104 are contradictory. This emphasizes the risk of arriving at definitive conclusions concerning the binding interface between sTF and FXa based solely on data provided by activity studies, especially when the active site is located far from the site of mutation. Thus, we have demonstrated that the region of inter- action between sTF and FXa in sTF–FVIIa–FXa– TFPI(1)161) involves not only the C-terminal part of sTF, but also an area in the N-terminal part of TF2 (around position 197) and a region in the C-terminal part of TF1 (around position 104) (Fig. 1). However, the entire C-terminal part of TF1 is not involved in the FXa interaction area, because position 99 does not participate in the binding. This region is found at the level of EGF1 of FVIIa in the sTF–FVIIa complex. However, in the previ- ously mentioned studies by Kirchhofer et al. [13,17], these positions were not probed, hence they might also be involved in the interaction region in the ternary complexes sTF–FVII–FXa and sTF–FVIIa–FX. The structural analogy between FVIIa and FXa (dis- cussed above) in combination with our fluorescence meas- urement results suggests that, similar to the Gla domain, EGF1 of FXa is involved in binding to sTF in the quaternary sTF–FVIIa–FXa–TFPI(1–161) complex. In this context, it should be noted that also for the ternary TF–FVIIa–FX complex it has been demonstrated that FX interactswithTFusing,inpart,theEGF1domain[36]. To conclude, we have described the interaction region between sTF and FXa in the quaternary complex, at the level of specific residues, and we have also shown major similarities regarding this interaction area in the quaternary and ternary complexes. Increased knowledge about the interface between FX/FXa and TF in these complexes would facilitate the design of small molecules with anti- thrombotic effects. In any case, potential contacts between sTF and FXa at the level of EGF2 and the PD of FVIIa in the sTF–FVIIa–FXa–TFPI(1–161) complex remain to be identified. Acknowledgements This work was supported by the Swedish Research Council (MS, UC) and Magnus Bergvall Stiftelse (MS). We thank Helle Bak for technical assistance. References 1. Golino, P. (2002) The inhibitors of the tissue factor: factor VII pathway. Thromb. Res. 106, V257–V265. 2. Broze, G.J. Jr, Girard, T.J. & Novotny, W.F. (1990) Regulation of coagulation by a multivalent Kunitz-type inhibitor. Biochemistry 29, 7539–7546. 3. Girard, T.J., Warren, L.A., Novotny, W.F., Likert, K.M., Brown, S.G., Miletich, J.P. & Broze, G.J. Jr (1989) Functional significance of the Kunitz-type inhibitory domains of lipoprotein-associated coagulation inhibitor. Nature 338, 518–520. 4. Sevinsky, J.R., Rao, L.V.M. & Ruf, W. (1996) Ligand-induced protease receptor translocation into caveolae: a mechanism for regulating cell surface proteolysis of the tissue factor-dependent coagulation pathway. J. Cell. Biol. 133, 293–304. 5. Iakhiaev, A., Pendurthi, U.R., Voigt, J., Ezban, M. & Rao, L.V.M. (1999) Catabolism of factor VIIa bound to tissue factor in fibroblasts in the presence and absence of tissue factor pathway inhibitor. J. Biol. Chem. 274, 36995–37003. 6. Hamamoto, T., Yamamoto, M., Nordfang, O., Petersen, J.G.L., Foster,D.C.&Kisiel,W.(1993)Inhibitorypropertiesoffull- length and truncated recombinant tissue factor pathway inhibitor (TFPI). Evidence that the third Kunitz-type domain of TFPI is not essential for the inhibition of factor VIIa- tissue factor com- plexes on cell surfaces. J. Biol. Chem. 268, 8704–8710. 7. Banner, D.W., D’Arcy, A., Che ` ne, C., Winkler, F.K., Guha, A., Konigsberg, W.H., Nemerson, Y. & Kirchhofer, D. (1996) The crystal structure of the complex of blood coagulation factor VIIa with soluble tissue factor. Nature 380, 41–46. 8. Huang, Q., Neuenschwander, P.F., Rezaie, A.R. & Morrissey, J.H. (1996) Substrate recognition by tissue factor-factor VIIa. Evidence for interaction of residues Lys165 and Lys166 of tissue factor with the 4-carboxyglutamate-rich domain of factor X. J. Biol. Chem. 271, 21752–21757. 9. Ruf, W., Shobe, J., Rao, S.M., Dickinson, C.D., Olson, A. & Edgington, T.S. (1999) Importance of factor VIIa Gla-domain residue Arg-36 for recognition of the macromolecular substrate factor X Gla-domain. Biochemistry 38, 1957–1966. 10. Martin, D.M.A., O’Brien, D.P., Tuddenham, E.G.D. & Byfield, P.G.H. (1993) Synthesis and characterization of wild-type and variant c-carboxyglutamic acid-containing domains of factor VII. Biochemistry 32, 13949–13955. 11. Ruf, W., Miles, D.J., Rehemtulla, A. & Edgington, T.S. (1992) Tissue factor residues 157–167 are required for efficient Ó FEBS 2003 The sTF–FXa interface in sTF–FVIIa–FXa–TFPI (Eur. J. Biochem. 270) 2581 proteolytic activation of factor X and factor VII. J. Biol. Chem. 267, 22206–22210. 12. Rehemtulla, A., Ruf, W., Miles, D.J. & Edgington, T.S. (1992) The third Trp-Lys-Ser (WKS) tripeptide motif in tissue factor is associated with a function site. Biochem. J. 282, 737–740. 13. Kirchhofer, D., Lipari, M.T., Moran, P., Eigenbrot, C. & Kelley, R.F. (2000) The tissue factor region that interacts with substrates factor IX and factor X. Biochemistry 39, 7380–7387. 14. Roy, S., Hass, P.E., Bourell, J.H., Henzel, W.J. & Vehar, G.A. (1991) Lysine residues 165 and 166 are essential for the cofactor function of tissue factor. J. Biol. Chem. 266, 22062–22066. 15. Ruf, W., Miles, D.J., Rehemtulla, A. & Edgington, T.S. (1992) Cofactor residues lysine 165 and 166 are critical for protein sub- strate recognition by the tissue factor–factor VIIa protease com- plex. J. Biol. Chem. 267, 6375–6381. 16. Dittmar, S., Ruf, W. & Edgington, T.S. (1997) Influence of mutations in tissue factor on the fine specificity of macro- molecular substrate activation. Biochem. J. 321, 787–793. 17. Kirchhofer, D., Eigenbrot, C., Lipari, M.T., Moran, P., Peek, M. & Kelley, R.F. (2001) The tissue factor region that interacts with factor Xa in the activation of factor VII. Biochemistry 40, 675–682. 18. Rao, L.V.M. & Ruf, W. (1995) Tissue factor residues Lys165 and Lys166 are essential for rapid formation of the quaternary com- plex of tissue factor-VIIa with Xa-tissue factor pathway inhibitor. Biochemistry 34, 10867–10871. 19. Girard, T.J., MacPhail, L.A., Likert, K.M., Novotny, W.F., Miletich, J.P. & Broze, G.J. Jr (1990) Inhibition of factor VIIa- tissue factor coagulation activity by a hybrid protein. Science 248, 1421–1424. 20. Warn-Cramer,B.J.,Rao,L.V.M.,Maki,S.L.&Rapaport,S.I. (1988) Modifications of extrinsic pathway inhibitor (EPI) and FXa that affect their ability to interact and to inhibit factor VIIa/ tissue factor: evidence for a two-step model of inhibition. Thromb. Haemost. 60, 453–456. 21. Rao, L.V.M., Williams, T. & Rapaport, S.I. (1996) Studies of the activation of factor VII bound to tissue factor. Blood 87, 3738–3748. 22. Owenius, R., O ¨ sterlund, M., Lindgren, M., Svensson, M., Olsen, O.H., Persson, E., Freskga ˚ rd, P O. & Carlsson, U. (1999) Prop- erties of spin and fluorescent labels at a receptor–ligand interface. Biophys. J. 77, 2237–2250. 23. Gill, S.C. & von Hippel, P.H. (1989) Calculation of protein extinction coefficients from amino acid sequence data. Anal. Biochem. 182, 319–326. 24. Thim, L., Bjoern, S., Christensen, M., Nicolaisen, E.M., Lund- Hansen, T., Pedersen, A.H. & Hedner, U. (1988) Amino acid sequence and posttranslational modifications of human factor VIIa from plasma and transfected baby hamster kidney cells. Biochemistry 27, 7785–7793. 25. Petersen, J.G.L., Meyn, G., Rasmussen, J.S., Petersen, J., Bjørn, S.E., Jonassen, I., Christiansen, L. & Nordfang, O. (1993) Char- acterization of human tissue factor pathway inhibitor variants expressed in Saccharomyces cerevisiae. J. Biol. Chem. 268, 13344– 13351. 26. Freskga ˚ rd, P O., Olsen, O.H. & Persson, E. (1996) Structural changes in factor VIIa induced by Ca 2+ and tissue factor studied using circular dichroism spectroscopy. Protein Sci. 5, 1531–1540. 27. Owenius, R., O ¨ sterlund, M., Svensson, M., Lindgren, M., Persson, E., Freskga ˚ rd, P O. & Carlsson, U. (2001) Spin and fluorescent probing of the binding interface between tissue factor and factor VIIa at multiple sites. Biophys. J. 81, 2357–2369. 28. Pike, A.C.W., Brzozowski, A.M., Roberts, S.M., Olsen, O.H. & Persson, E. (1999) Structure of human factor VIIa and its implications for the triggering of blood coagulation. Proc. Natl Acad. Sci. USA 96, 8925–8930. 29. Padmanabhan, K., Padmanabhan, K.P., Tulinsky, A., Park, C.H., Bode, W., Huber, R., Blankenship, D.T., Cardin, A.D. & Kisiel, W. (1993) Structure of human des (1–45) factor Xa at 2.2 A ˚ resolution. J. Mol. Biol. 232, 947–966. 30. Brandstetter, H., Ku ¨ hne, A., Bode, W., Huber, R., von der Saal, W., Wirthensohn, K. & Engh, R.A. (1996) X-ray structure of active site-inhibited clotting factor Xa. Implications for drug design and substrate recognition. J. Biol. Chem. 271, 29988–29992. 31. Husten, E.J., Esmon, C.T. & Johnson, A.E. (1987) The active site of blood coagulation factor Xa. Its distance from the phospholipid surface and its conformational sensitivity to components of the prothrombinase complex. J. Biol. Chem. 262, 12953–12961. 32. McCallum, C.D., Hapak, R.C., Neuenschwander, P.F., Morris- sey, J.H. & Johnson, A.E. (1996) The location of the active site of blood coagulation factor VIIa above the membrane surface and its reorientation upon association with tissue factor. A fluorescence energy transfer study. Biol. Chem. 271, 28168–28175. 33. McCallum,C.D.,Su,B.,Neuenschwander,P.F.,Morrissey,J.H. & Johnson, A.E. (1997) Tissue factor positions and maintains the factor VIIa active site far above the membrane surface even in the absence of the factor VIIa Gla domain. A fluorescence resonance energy transfer study. J. Biol. Chem. 272, 30160–30166. 34. Ruf, W., Kalnik, M.W., Lund-Hansen, T. & Edgington, T.S. (1991) Characterization of factor VII association with tissue factor in solution. High and low affinity calcium binding sites in factor VII contribute to functionally distinct interactions. J. Biol. Chem. 266, 15719–15725. 35. Baugh, R.J., Dickinson, C.D., Ruf, W. & Krishnaswamy, S. (2000) Exosite interactions determine the affinity of factor X for the extrinsic Xase complex. J. Biol. Chem. 275, 28826–28833. 36. Zhong, D., Bajaj, M.S., Schmidt, A.E. & Bajaj, S.P. (2002) The N-terminal epidermal growth factor-like domain in factor IX and factor X represents an important recognition motif for binding to tissue factor. J. Biol. Chem. 277, 3622–3631. 37. O ¨ sterlund, M., Owenius, R., Carlsson, K., Carlsson, U., Persson, E., Lindgren, M., Freskga ˚ rd, P O. & Svensson, M. (2001) Prob- ing inhibitor-induced conformational changes along the interface between tissue factor and factor VIIa. Biochemistry 40, 9324–9328. 2582 K. Carlsson et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . Probing the interface between factor Xa and tissue factor in the quaternary complex tissue factor factor VIIa factor Xa tissue factor pathway inhibitor Karin. 200 and 201, is involved in binding to FXa in the complex, and FXa, most likely via its Gla domain, is also in contact with the Gla domain of FVIIa in this