- Email: [email protected]
A Kinetic Assay to Determine Prothrombin Binding to Membranes
Thrombosis Research 92 (1998) 239–247
REGULAR ARTICLE
A Kinetic Assay to Determine Prothrombin Binding to Membranes Jose´ W.P. Govers-Riemslag, Lise Johnsen, Ramona J. Petrovan, Jan Rosing and Guido Tans Cardiovascular Research Institute Maastricht, Maastricht University, Maastricht, The Netherlands. (Received 28 May 1998 by Editor K. Stocker; revised/accepted 3 August 1998)
Abstract Activation of prothrombin by multisquamase, the prothrombin activator from the venom of Echis multisquamatus (Central Asian sand viper), is inhibited by membranes containing negatively charged anionic phospholipids. This inhibition appears to be due to the fact that the venom activator cannot activate membrane-bound prothrombin. Initial steady state rates of prothrombin activation by multisquamase in the presence of phospholipids appeared to depend on the fraction unbound prothrombin only and this phenomenon was used to quantitate binding of prothrombin to membranes of varying phospholipid composition. In this method, the initial rate of prothrombin activation by multisquamase is measured in the absence (total prothrombin) and in the presence of a procoagulant surface (rate depending only on free prothrombin) and from the difference in activation rates the amount of membrane-bound prothrombin is calculated. The validity of the method was established by determination of the binding paramAbbreviations: PC, 1,2-dioleoyl-sn-glycero-3-phosphocholine; PS, 1,2-dioleoyl-sn-glycero-3-phosphoserine; PA, 1,2-dioleoyl-sn-glycero-3-phosphatidic acid; PG, 1,2-dioleoyl-sn-glycero-3-phosphoglycerol; PE, 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine; SM, sphingomyelin; PI, phosphatidylinositol from bovine liver; p-NPGB, p-nitrophenyl-p9-guanidino-benzoate hydrochloride; S2238, D-Phe(pipecolyl)-Arg-pNA; PPACK, phenylalanyl-prolyl-arginine chloromethyl ketone; BCA, bicinchoninic acid; FPLC; fast protein liquid chromatography; BSA, bovine serum albumin. Corresponding author: J.W.P. Govers-Riemslag, Ph.D., Department of Biochemistry, Maastricht University, P.O. Box 616, 6200 MD Maastricht, The Netherlands. Fax: 131 (43) 3670988, E-mail: ,[email protected]..
eters for prothrombin binding to 100 mM phospholipid vesicles composed of 20 mole% phosphatidylserine and 80 mole% phosphatidylcholine. The binding parameters obtained were Kd50.84 mM and n50.021 mmoles prothrombin bound per mmole phospholipid which is in agreement with literature. Due to the nature of the measurement the method is especially suitable to quantitate binding of prothrombin at concentrations as low as 5 nM prothrombin. 1998 Elsevier Science Ltd. Key Words: Echis multisquamatus; Prothrombin activation; Prothrombin binding
T
he conversion of prothrombin into thrombin is one of the key events in blood coagulation. Under physiological conditions prothrombin is activated by the so-called prothrombinase complex consisting of the serine protease, factor Xa, and the nonenzymatic cofactors, factor Va, negatively charged phospholipids and Ca21 (for a recent review see [1]). Binding of prothrombin to the procoagulant membrane is essential for optimal activation and in the past years binding of prothrombin to membranes has been studied under a variety of reaction conditions [2–9]. Prothrombin can also be activated by so-called exogenous activators such as are present in snake venoms [10,11]. Recently, the purification of the activator present in the venom of Echis multisquamatus was reported [12,13]. The structural properties of this activator, which we have designated multisquamase, are different from ecarin, the wellknown prothrombin activator from the venom of
0049-3848/98 $–see front matter 1998 Elsevier Science Ltd. Printed in the USA. All rights reserved. PII S0049-3848(98)00144-3
240
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
Echis carinatus. Multisquamase resembles carinactivase, a second activator reported to be present in minor amounts in the venom of Echis carinatus [14]. Multisquamase is comprised of two subunits, a catalytic subunit of Mr558000 and a regulatory subunit of Mr523000. Like ecarin, multisquamase activates human prothrombin through cleavage at Arg320 resulting in meizothrombin formation [12, 13]. However, in contrast to ecarin, multisquamasedependent prothrombin activation is strongly influenced by experimental conditions such as ionic strength and the presence of Ca21 ions. Moreover, the presence of a fully carboxylated fragment 1 region in prothrombin appears to be essential for prothrombin activation by multisquamase since activation of prethrombin 1 (prothrombin des fragment 1) or prothrombin present in the plasma from patients on oral anticoagulant therapy is strongly diminished [12,13]. During the course of our studies, it became apparent that multisquamase-dependent prothrombin activation is inhibited in the presence of procoagulant membranes, a phenomenon also reported by Yamada et al. [12]. These observations prompted us to investigate the possibility to use multisquamase as a tool to quantitate the binding of prothrombin to procoagulant membranes.
1. Materials and Methods 1.1. Materials Dioleoyl-sn-glycero-3-phosphatidic acid (PA), sphingomyelin (SM), Hepes, Tris, EDTA, bovine serum albumin, and ovalbumin were purchased from Sigma Chemical Co. (St. Louis, MO). Dioleoyl-snglycero-3-phosphoserine (PS), dioleoyl-sn-glycero3-phosphocholine (PC), dioleoyl-sn-glycero-3-phosphoglycerol (PG), dioleoyl-sn-glycero-3-phosphoethanolamine (PE) were from Avanti Polar Lipids Inc., (Alabaster, AL). Phosphatidylinositol (PI) was a kind gift from Dr. E.M. Bevers, Department of Biochemistry, Maastricht University. The chromogenic substrate S2238 was supplied by Chromogenix (Mo¨lndal, Sweden). Ecarin, the prothrombin activator from Echis carinatus venom was from Pentapharm, Basel, Switzerland. Crude Echis multisquamatus venom was obtained from Latoxan (Rosans, France). Micro BCA protein assay kits were from Pierce (Rockford, IL). FPLC equipment and
column materials used for protein purification were purchased from Pharmacia (Uppsala, Sweden).
1.2. Proteins Human coagulation factors used in this study were isolated from fresh frozen plasma. Human prothrombin was purified according to DiScipio et al. [15]. Human thrombin was prepared from prothrombin activation mixtures as described by Pletcher and Nelsestuen [16]. Human Factor Va was purified as described previously [17]. Multisquamase was purified from the crude venom from Echis multisquamatus as described earlier [13]. Protein preparations were homogeneous and .95% pure as judged by SDS-PAGE according to Laemmli [18] and were stored at 2808C.
1.3. Protein Concentrations The concentration of multisquamase was determined with the micro BCA protein assay [19]. Prothrombin concentrations were determined after complete activation of prothrombin with ecarin and quantitation of meizothrombin with S2238 [20]. Factor Va was quantitated as described earlier [17].
1.4. Liposomes and Phospholipid Vesicle Preparations Phospholipid preparations (premixed in the desired composition in CHCl3/CH3OH (9:1 v/v)) were dried under a stream of N2 and suspended in a buffer containing 25 mM Hepes (pH 7.5) and 175 mM NaCl by vigorous vortexing in the presence of glass beads for 3 minutes. Centrifugable liposomes were obtained by centrifugation for 25 minutes at 900003g at 378C in a Beckman (Fullerton, CA, USA) TL100 ultracentrifuge after which the pellet was resuspended in buffer. The liposomes spun down quantitatively during subsequent centrifugation at 900003g at 378C for 25 minutes. Small unilamellar vesicles were prepared by sonication of liposomes for 10 min at 48C with a MSE (Loughborough, England) Soniprep150 ultrasonic disintegrator set at 8 mm peak to peak amplitude. Phospholipid concentrations were determined by phosphate analysis [21].
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
1.5. Determination of the Rate of Prothrombin Activation by Multisquamase Human prothrombin was preincubated in a buffer containing 25 mM Hepes (pH 7.5), 175 mM NaCl, 2 mM CaCl2 in the absence or presence of phospholipids at 378C. After 4 minutes, activation was started by addition of appropriate amounts of multisquamase, the purified prothrombin activator from Echis multisquamatus, in the same buffer. After different time intervals the amount of prothrombin activated was quantitated in aliquots withdrawn from the activation mixtures as described earlier [22] or by measuring the amidolytic activity towards S2238 (in a total volume of 250 mL 50 mM Tris-HCl (pH 7.9), 175 mM NaCl, 20 mM EDTA, 0.5 mg/ml ovalbumin and 235 mM S2238) in a microtiterplate at 378C using a SLT 340C kinetic reader set in the dual wavelength mode at 405–492 nm. The amounts of activated prothrombin were calculated from a calibration curve made with known amounts of active sitetitrated thrombin. The kinetic parameters, Km and Vmax, of prothrombin activation by multisquamase were obtained by measurement of the initial rate of prothrombin activation (V) at varying prothrombin concentration (S) and fitting the V vs. S curve to the Michaelis-Menten equation (V5Vmax3S/(Km1S) using nonlinear least squares regression analysis.
1.6. Binding Analysis In this article, we show that in prothrombin-phospholipid mixtures membrane bound prothrombin is not available for activation by multisquamase. To quantitate bound and free prothrombin in such reaction mixtures the initial rate of prothrombin activation by multisquamase was determined both in the absence and presence of phospholipid. At prothrombin concentrations well below Km, i.e., conditions at which initial rates of activation are directly proportional to the prothrombin concentration, the concentrations free prothrombin PTfree and membrane-bound prothrombin PTbound are calculated from Equations 1 and 2: PTfree5(V1PL/V2PL)3PTtotal
(1)
PTbound5PTtotal2PTfree
(2)
in which V1PL is the rate of prothrombin activation determined in the presence of phospholipid and
241
V2PL is the rate of prothrombin activation determined in the absence of phospholipid. At prothombin concentrations .100 nM, PTfree and PTbound were calculated using Equations 3 and 4: PTfree5(V1PL3Km)/(Vmax2V1PL)
(3)
PTbound5PTtotal2PTfree
(4)
in which V1PL is the rate of prothrombin activation in the presence of phospholipid and Km and Vmax are the kinetic parameters of multisquamase-catalyzed prothrombin activation determined in the absence of phospholipids.
2. Results Multisquamase-dependent prothrombin activation requires Ca21-ions and activation rates are strongly dependent on the ionic strength [12,13]. The experiments presented in this paper were performed throughout in 25 mM Hepes (pH 7.5 at 378C), 175 mM NaCl, 2 mM CaCl2 and 5 mg/ml BSA as carrier protein to passivate reaction tubes for adsorption of reactants.
2.1. Inhibition of Prothrombin Activation by Multisquamase in the Presence of Membranes Figure 1A shows that increasing amounts of negatively charged phospholipid vesicles (PS/PC; 20/ 80; M/M) gradually decreased the initial rate of activation of 0.2 mM prothrombin by multisquamase until at 500 mM phospholipid vesicles approximately 10% of the rate measured in the absence of pospholipids remained. Time courses of prothrombin activation were linear with time (,2% of the prothrombin available was allowed to activate) both in the presence and absence of phospholipid (data not shown). This experiment indicates that the inhibition of multisquamase-catalyzed prothrombin activation is caused by the fact that multisquamase can only activate free prothrombin and not membrane-bound prothrombin. To verify whether the observed inhibition of the initial rate of prothrombin activation was indeed due to binding of prothrombin to the membrane we performed an experiment in which 0.2 mM prothrombin was incubated with varying amounts of centrifugable large multilamellar PS/PC vesicles. These incubation mixtures were divided into two portions. In one series, prothrombin bound to the
242
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
Fig. 1. Effect of phospholipids on prothrombin activation by multisquamase. (A) Initial rates of activation of 0.2 mM prothrombin by 15 ng/ml multisquamase were determined at 378C in the presence of varying amounts of phospholipid vesicles (PS/PC; 20:80; M/M) in a reaction mixture containing 25 mM Hepes (pH 7.5), 2 mM CaCl2, 5 mg/ml BSA, and 175 mM NaCl and expressed as percentage of the rate determined in the absence of phospholipid. (B) Prothrombin (0.2 mM) was incubated at 378C in a reaction mixture containing 25 mM Hepes (pH 7.5), 2 mM CaCl2, 5 mg/ml BSA, 175 mM NaCl and concentrations of large multilamellar phospholipid vesicles (PS/PC; 20:80; M/M) indicated in the figure. Aliquots of these incubation mixtures were centrifuged 25 minutes at 378C at 90,0003g and the remaining part of the incubation mixtures were kept at 378C without centrifugation. The initial rate of prothrombin activation was subsequently determined by addition of 7.5 ng/ml multisquamase to 100 ml of the noncentrifuged incubation mixtures (s) or to 100 ml supernatant of the centrifuged incubation mixtures (m). The rate of activation determined in the absence of phospholipids was taken as hundred percent.
liposomes was removed by centrifugation for 25 minutes at 900003g at 378C and the other series was kept at 378C during the experiment. The initial rate of prothrombin activation was subsequently determined in the supernatant of the centrifuged incubation mixtures as well as in the noncentrifuged incubation mixtures by the addition of equal amounts of multisquamase. In Figure 1B, it is shown that the initial rates of prothrombin activation were identical irrespective of whether the prothrombinliposome complexes had been removed (centrifuged tubes) or not (noncentrifuged). In a control experiment, it was shown that multisquamase did not bind to the liposomes. Thus, multisquamase can only activate prothrombin that is not bound to the lipid and multisquamase can therefore be used as a probe to quantitate prothrombin binding to negatively charged lipids. We have also determined the effect of phospholipids on ecarin-catalyzed prothrombin activation (data not shown). Since membrane-bound prothrombin is still susceptible to activation by ecarin, this venom prothrombin activator cannot be used to quantify prothrombin binding to phospholipids.
2.2. Determination of Binding Parameters Figure 2A shows the initial rate of multisquamasedependent prothrombin activation determined at
varying prothrombin concentrations both in the absence and in the presence of 100 mM negatively charged phospholipid vesicles (PS/PC; 20:80; M/M). The kinetic parameters obtained from the experiment performed in the absence of lipids were: Km50.72 mM prothrombin and Vmax51.45 nM prothrombin activated/minute/mg activator. Since the rate of prothrombin activation in the presence of 100 mM PS/PC vesicles (Figure 2A, filled triangles) is solely dependent on free prothrombin, the concentration of free prothrombin at each prothrombin concentration can be calculated from the observed velocity using the plot of V vs. S determined in the absence of lipid (Figure 2A, filled circles) as a calibration curve. Figure 2B shows the Scatchard plot obtained from the amounts of prothrombin bound and free calculated from the data shown in Figure 2A. The binding parameters obtained from the Scatchard plot were Kd50.84 mM and n50.021 mmoles/mmoles phospholipid.
2.3. Effect of Different Anionic Phospholipids on the Binding of Prothrombin To test the general applicability for the use of multisquamase as a probe to quantitate prothrombin binding to membranes the experiment shown in Figure 1 was repeated with a variety of lipid membranes (Figure 3). Almost all membranes inhibited
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
243
Fig. 2. Prothrombin activation by multisquamase in the absence and presence of phospholipid vesicles as a function of the prothrombin concentration. (A) Varying amounts of prothrombin were activated by 6 ng/ml multisquamase at 378C in a reaction mixture containing 25 mM Hepes (pH 7.5), 2 mM CaCl2, 5 mg/ml BSA, and 175 mM NaCl, in the absence (d) or presence (m) of 100 mM phospholipid vesicles (PS/PC; 20:80; M/M). Prothrombin activation was determined as described under Materials and Methods, The kinetic parameters of prothrombin activation by multisquamase in the absence of phospholipids were obtained by fitting the data by using nonlinear least squares regression. Bound and free prothrombin at each prothrombin concentration in the presence of phospholipids (Figure 2A) were subsequently calculated as described under Materials and Methods. (B) Scatchard plot of prothrombin binding to phospholipid vesicles. The amounts of bound and free prothrombin in the presence of phospholipids were determined from the initial rate of prothrombin activation shown in panel A as described under Materials and Methods. The solid line shown in the Scatchard plot represents the best fit obtained by nonlinear least squares regression analysis. The binding parameters obtained were Kd50.84 mM and n50.021 mmoles prothrombin bound per mmole phospholipid.
multisquamase-dependent prothrombin activation and the extent of inhibition varied with the phospholipid composition of the membrane in a manner consistent with the known dependency of prothrombin binding on the nature of the polar head group in anionic phospholipid-containing lipid membranes. Table 1 summarizes the data obtained at 150 mM phospholipid. The strongest prothrombin binding, approximately 75% at 150 mM vesicles, was observed in the presence of vesicles that mimic the outer leaflet of the plasma membrane of activated platelets or with vesicles containing PS/PC/PE (20:60:20; M/M/M). Membranes containing PS/PC (20:80; M/M) or PA/PC (20:80; M/M) were only marginally less effective but membranes containing 10% PS showed a clearly reduced prothrombin binding. On vesicles containing 20% PG only weak binding was observed, whereas vesicles which contained only PC did not bind prothrombin (Figure 3, Table 1).
2.4. Detection of Binding at Low Prothrombin Concentrations (Effect of Factor Va) At prothrombin concentrations well below the Km5 0.72 mM for prothrombin, the initial rate of multi-
squamase-dependent prothrombin activation is linearly dependent on the amount of prothrombin available for activation. This, together with the fact that concentrations as low as 0.1 nM (meizo)thrombin are easily detectable makes multisquamase eminently suitable for detection of prothrombin binding at concentrations as low as 10 nM without the risk of interfering with the binding equilibrium by lowering the free ligand concentration through consumption of free prothrombin. This is illustrated by the experiment shown in Figure 4. In this experiment, the binding of 10 nM prothrombin to 25 mM vesicles composed of PS/PC (5:95; M/M), PS/PC (20:80; M/M), or PS/PC/PE/ SM/PI (10:39:27:19:5; M/M/M/M/M) was determined both in the absence and in the presence of 20 nM factor Va. Factor Va has been reported to promote the binding of prothrombin to negatively charged membranes [8]. In the absence of phospholipid, factor Va did not influence multisquamase-dependent prothrombin activation (data not shown) whereas addition of factor Va in the presence of lipid resulted in an increased inhibition of prothrombin activation. This indicates that factor Va stimulated the binding of prothrombin to the membrane (Figure 4). Although the absolute additional binding to the poor prothrombin binding vesicles
244
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
Fig. 3. The effect of phospholipid vesicles of varying composition on multisquamase-dependent prothrombin activation. Initial rates of prothrombin activation by 15 ng/ml multisquamase were determined at 378C in a reaction mixture containing, 0.2 mM prothrombin, 25 mM Hepes (pH 7.5), 2 mM CaCl2, 175 mM NaCl, 5 mg/ml BSA and amounts of phospholipid vesicles as indicated in the figure. The composition of the different phospholipid vesicles used were: m - m 100% PC, r - r PG/PC (20:80; M/M)), s s PS/PC (10:90; M/M), j - j PA/PC (20:80; M/M), d - d PS/PC (20:80; M/M), h - h; PS/PC/PE (20:60:20; M/M/M), and n - n; PS/PC/PE/SM/PI (10:39:27:19:5; M/M/M/M/M). Rates of prothrombin activation determined in the absence of lipid were taken as 100%.
(5:95; PS/PC) was lowest, the relative increase in binding to these vesicles was threefold as compared with the modest 1.5- and 1.2-fold increase observed for the other vesicles (Figure 4). Increasing the amounts of factor Va, however, did not result in a further enhancement of binding (not shown).
3. Discussion In this article, we show that the prothrombin activator purified from the venom of Echis multisquamatus, designated here as multisquamase, can be used as a tool to quantitate prothrombin binding to procoagulant membranes. The purification of multisquamase has been reported by us [13] and by Yamada et al. [12,23]. Multisquamase (designated multactivase by Yamada et al. [12]) has been shown to require the fragment 1 domain and the posttranslational carboxylation of the glutamic acid residues to g-carboxyglutamic acid (gla) to efficiently
convert prothrombin into meizothrombin [12,13]. In agreement with the data reported by Yamada et al. [12], we observed that the presence of negatively charged phospholipids inhibits multisquamasedependent prothrombin activation. In the present article, we show that binding of prothrombin to the lipid effectively blocks the activation and that multisquamase can only activate free prothrombin. This property of multisquamase was used to develop a method to quantitate the binding of prothrombin to phospholipids. The validity of the method is established by the fact that the parameters determined for binding of prothrombin to 100 mM 20:80 (M/M) PS/PC vesicles (Kd50.84 mM and n50.021 mmoles/mmole phospholipid) are in agreement with literature [2,4,6,24] and by the fact that the binding of prothrombin to phospholipid vesicles of different composition showed a variation in binding completely in accordance with the known variation in binding afinities of prothrombin for such membranes [2,4,6]. The method described here is not restricted to quantitate binding of human prothrombin to membranes but can also be used with bovine prothrombin. In a number of cases, the key experiments were repeated with bovine prothrombin in which case also only free prothrombin appears to be activated by multisquamase (data not shown). Due to the nature of the measurement, determination of initial rates of activation of prothrombin available for activation (i.e., free prothrombin), multisquamase is especially useful for quantitation of prothrombin binding at prothrombin concentration well below the Km of 0.7 mM for prothrombin. Under these conditions, changes in the rate of prothrombin activation are directly proportional to changes in the concentration of free prothrombin and the ratio of activation rates determined in the presence and absence of surface is a direct measurement of the fraction of free prothrombin. Since (meizo)thrombin concentrations of 0.1 nM are easily detectable using S2238, free prothrombin concentrations as low as 5 nM can be detected without interfering with the binding equilibrium (prothrombin free remains constant throughout the time course of the activation). Interestingly, this should make the use of multisquamase valuable in those cases where binding of prothrombin to planar phospholipid bi- or monolayer is studied since with
245
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
Table 1. Initial rate of multisquamase-dependent prothrombin activation in the absence and presence of 150 mM phospholipid vesicles of different composition Initial activation rate (nM PT-activated/minutes) Vesicle composition PC PG/PC PS/PC PA/PC PS/PC PS/PC/PE PS/PC/PE/SM/PI
(100%) (20:80; M/M) (10:90; M/M) (20:80; M/M) (20:80; M/M) (20:60:20; M/M/M) (10:39:27:19:5; M/M/M/M/M)
Without lipid
With lipid
% PTbound
5.50 5.13 5.40 5.33 5.18 5.28 5.29
5.51 4.85 3.91 2.21 2.01 1.42 1.67
— 7% 34% 65% 67% 78% 74%
Initial rates of prothrombin activation by 15 ng/ml multisquamase were determined at 378C in a reaction mixture containing 0.2 mM prothrombin and 150 mM phospholipid vesicles of varying composition in 25 mM Hepes (pH 7.5), 2 mM CaCl2, 175 mM NaCl, and 5 mg/ml BSA. The concentration of prothrombin available for activation was calculated from the rate of prothrombin activation determined using Equations 3 and 4 of the Materials and Methods and the kinetic parameters (Km50.72 mM and Vmax521.75 nM prothrombin activated per minute; see also Results). The percentage of prothrombin-bound was then calculated taking the amount of prothrombin available in the absence of phospholipid as 100%.
such membranes much lower dissociation constants have been reported [5,9]. At higher prothrombin concentrations it is necessary to obtain the kinetic parameters for activation (Vmax and Km) to enable calculation of free
Fig. 4. Effect of factor Va on prothrombin binding to phospholipid vesicles. Prothrombin (10 nM) was incubated at 378C in the absence (open bars) or presence of 20 nM factor Va (shaded bars) in a reaction mixture containing 25 mM Hepes (pH 7.5), 2 mM CaCl2, 175 mM NaCl, 5 mg/ ml BSA containing 25 mM phospholipid vesicles of lipid composition as indicated in the figure. After 4 minutes, prothrombin activation was started by adding 7.5 ng/ml multisquamase and the initial rate of prothrombin activation and the amount of membrane-bound prothrombin was determined as described under Materials and Methods.
prothrombin from the observed initial rate of activation. The kinetic parameters determined (Vmax5 1.45 mM prothrombin activated per minute per microgram activator and Km50.7 mM prothrombin) are in agreement with the kinetic parameters reported earlier by us [13]. At concentrations well above Km activation rates become independent of the prothrombin concentration. Thus, multisquamase cannot be used to detect prothrombin binding at free prothrombin concentrations higher than 2 mM. Moreover, since prothrombin activation by multisquamase is strongly dependent on Ca21ions [12,13] and ionic strength [13] care has to be taken that the experimental conditions are kept constant. The method described in this paper may offer a welcome addition to other methods for the determination of protein binding to phospholipid such as quasielastic light scattering [2,4,6], ellipsometry [5], or prothrombin binding to monolayers [9]. Since binding is usually detected by measurement of the protein mass adsorbed to the membrane it may be difficult to determine prothrombin binding in the presence of other lipid binding components. In such a case, as illustrated in the experiment shown in Figure 4, prothrombin binding can still be quantitated with the method described here. Further experiments are underway to establish the use of multisquamase to screen for prothrombin binding in more complex systems containing platelets (or even whole blood) or cultured cells.
246
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
This work was supported by Program Grant 900-526-192 from the Dutch Organisation for Scientific Research (NWO). R.J. Petrovan was supported by the Deutsche Akademie der Naturforscher “Leopoldina”. L. Johnsen was supported via the International Federation of Medical Student’s Association (IFMSA).
References 1. Davie EW. Biochemical and molecular aspects of the coagulation cascade. Thromb Haemostas 1995;74:1–6. 2. Nelsestuen GL, Broderius M. Interactions of prothrombin and blood-clotting factor X with membranes varying composition. Biochemistry 1977;16:4172–7. 3. Nelsestuen GL, Lim TK. Equilibria involved in prothrombin and blood-clotting factor X-membrane binding. Biochemistry 1977;169:4164–71. 4. Wei GJ, Bloomfield VA, Resnick RM, Nelsestuen GL. Kinetic and mechanistic analysis of prothrombin membrane binding by stopped flow light scattering. Biochemistry 1982;21: 1949–59. 5. Kop JMM, Cuypers PA, Lindhout T, Hemker HC, Hermens W. The adsorption of prothrombin to phospholipid monolayers quantitated by ellipsometry. J Biol Chem 1984;259:13993–8. 6. Cutsforth GA, Whitaker RN, Hermans J, Lentz BR. A new model to describe extrinsic protein binding to phospholipid membrane of varying composition: application to human coagulation proteins. Biochemistry 1989;28: 7453–61. 7. Giesen PL, Willems GM, Hemker HC, Stuart MC, Hermens WT. Monitoring of unbound protein in vesicle suspensions with off-null ellipsometry. Biochim Biophys Acta 1993;1147: 125–31. 8. Billy D, Willems GM, Hemker HC, Lindhout T. Prothrombin contributes to the assembly of the factor Va- factor Xa complex at phosphatidylserine-containing phospholipid membranes. J Biol Chem 1995;270:26883–9. 9. Ellison EH, Castellino FJ. Adsorption of bovine prothrombin to spread phospholipid monolayers. Biophys J 1997;72:2605–15. 10. Rosing J, Tans G. Inventory of exogenous prothrombin activators. Thromb Haemost 1991; 65:627–30.
11. Rosing J, Tans G. Structural and functional properties of snake venom prothrombin activators. Toxicon 1992;30:1515–27. 12. Yamada D, Morita T. Purification and characterization of a Ca21-dependent prothrombin activator, multactivase, from the venom of Echis multisquamatus. J Biochem 1997;122: 991–7. 13. Petrovan RJ, Govers-Riemslag JWP, Nowak G, Hemker HC, Rosing J, Tans G. Purification and characterization of multisquamase, the prothrombin activator present in Echis multisquamatus venom. Thromb Res 1997;88:309–16. 14. Yamada D, Sekiya F, Morita T. Isolation and characterization of carinactivase, a novel prothrombin activator in Echis carinatus venom with a unique catalytic mechanism. J Biol Chem 1996;271:5200–7. 15. DiScipio RG, Hermodson MA, Yates SG, Davie EW. A comparison of human prothrombin: Factor IX (Christmas factor); factor X (Stuart factor); and protein S. Biochemistry 1977;16:698–706. 16. Pletcher CH, Nelsestuen GL. The rate-determining step of the heparin-catalyzed antithrombin/thrombin reaction is independent of thrombin. J Biol Chem 1982;257:5342–5. 17. Nicolaes GAF, Tans G, Thomassen MCLGD, Hemker HC, Pabinger I, Varadi K, Schwarz HP, Rosing J. Peptide bond cleavages and loss of functional activity during inactivation of factor Va and factor VaR506Q by activated protein C. J Biol Chem 1995;270:21158–66. 18. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–5. 19. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC. Measurement of protein using bicinchoninic acid [published erratum appears in Anal Biochem 1987 May 15;163:279]. Anal Biochem 1985; 150:76–85. 20. Tans G, Janssen Claessen T, Hemker HC, Zwaal RFA, Rosing J. Meizothrombin formation during factor Xa-catalyzed prothrombin activation Formation in a purified system and in plasma. J Biol Chem 1991;266:21864–73. 21. Bo¨ttcher CJF, van Gent CM, Pries C. A rapid
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
and sensitive sub-micro phosphorus determination. Anal Chim Acta 1961;24:203–7. 22. Rosing J, Zwaal RFA, Tans G. Formation of meizothrombin as intermediate in factor Xacatalyzed prothrombin activation. J Biol Chem 1986;261:4224–8. 23. Yamada D, Sekiya F, Morita T. Prothrombin
247
and factor X activator activities in the venoms of Viperidae snakes. Toxicon 1997;35:1581–9. 24. Lu YF, Nelsestuen GL. Dynamic features of prothrombin interaction with phospholipid vesicles of different size and composition: Implications for protein-membrane contact. Biochemistry 1996;35:8193–8200.
REGULAR ARTICLE
A Kinetic Assay to Determine Prothrombin Binding to Membranes Jose´ W.P. Govers-Riemslag, Lise Johnsen, Ramona J. Petrovan, Jan Rosing and Guido Tans Cardiovascular Research Institute Maastricht, Maastricht University, Maastricht, The Netherlands. (Received 28 May 1998 by Editor K. Stocker; revised/accepted 3 August 1998)
Abstract Activation of prothrombin by multisquamase, the prothrombin activator from the venom of Echis multisquamatus (Central Asian sand viper), is inhibited by membranes containing negatively charged anionic phospholipids. This inhibition appears to be due to the fact that the venom activator cannot activate membrane-bound prothrombin. Initial steady state rates of prothrombin activation by multisquamase in the presence of phospholipids appeared to depend on the fraction unbound prothrombin only and this phenomenon was used to quantitate binding of prothrombin to membranes of varying phospholipid composition. In this method, the initial rate of prothrombin activation by multisquamase is measured in the absence (total prothrombin) and in the presence of a procoagulant surface (rate depending only on free prothrombin) and from the difference in activation rates the amount of membrane-bound prothrombin is calculated. The validity of the method was established by determination of the binding paramAbbreviations: PC, 1,2-dioleoyl-sn-glycero-3-phosphocholine; PS, 1,2-dioleoyl-sn-glycero-3-phosphoserine; PA, 1,2-dioleoyl-sn-glycero-3-phosphatidic acid; PG, 1,2-dioleoyl-sn-glycero-3-phosphoglycerol; PE, 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine; SM, sphingomyelin; PI, phosphatidylinositol from bovine liver; p-NPGB, p-nitrophenyl-p9-guanidino-benzoate hydrochloride; S2238, D-Phe(pipecolyl)-Arg-pNA; PPACK, phenylalanyl-prolyl-arginine chloromethyl ketone; BCA, bicinchoninic acid; FPLC; fast protein liquid chromatography; BSA, bovine serum albumin. Corresponding author: J.W.P. Govers-Riemslag, Ph.D., Department of Biochemistry, Maastricht University, P.O. Box 616, 6200 MD Maastricht, The Netherlands. Fax: 131 (43) 3670988, E-mail: ,[email protected]..
eters for prothrombin binding to 100 mM phospholipid vesicles composed of 20 mole% phosphatidylserine and 80 mole% phosphatidylcholine. The binding parameters obtained were Kd50.84 mM and n50.021 mmoles prothrombin bound per mmole phospholipid which is in agreement with literature. Due to the nature of the measurement the method is especially suitable to quantitate binding of prothrombin at concentrations as low as 5 nM prothrombin. 1998 Elsevier Science Ltd. Key Words: Echis multisquamatus; Prothrombin activation; Prothrombin binding
T
he conversion of prothrombin into thrombin is one of the key events in blood coagulation. Under physiological conditions prothrombin is activated by the so-called prothrombinase complex consisting of the serine protease, factor Xa, and the nonenzymatic cofactors, factor Va, negatively charged phospholipids and Ca21 (for a recent review see [1]). Binding of prothrombin to the procoagulant membrane is essential for optimal activation and in the past years binding of prothrombin to membranes has been studied under a variety of reaction conditions [2–9]. Prothrombin can also be activated by so-called exogenous activators such as are present in snake venoms [10,11]. Recently, the purification of the activator present in the venom of Echis multisquamatus was reported [12,13]. The structural properties of this activator, which we have designated multisquamase, are different from ecarin, the wellknown prothrombin activator from the venom of
0049-3848/98 $–see front matter 1998 Elsevier Science Ltd. Printed in the USA. All rights reserved. PII S0049-3848(98)00144-3
240
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
Echis carinatus. Multisquamase resembles carinactivase, a second activator reported to be present in minor amounts in the venom of Echis carinatus [14]. Multisquamase is comprised of two subunits, a catalytic subunit of Mr558000 and a regulatory subunit of Mr523000. Like ecarin, multisquamase activates human prothrombin through cleavage at Arg320 resulting in meizothrombin formation [12, 13]. However, in contrast to ecarin, multisquamasedependent prothrombin activation is strongly influenced by experimental conditions such as ionic strength and the presence of Ca21 ions. Moreover, the presence of a fully carboxylated fragment 1 region in prothrombin appears to be essential for prothrombin activation by multisquamase since activation of prethrombin 1 (prothrombin des fragment 1) or prothrombin present in the plasma from patients on oral anticoagulant therapy is strongly diminished [12,13]. During the course of our studies, it became apparent that multisquamase-dependent prothrombin activation is inhibited in the presence of procoagulant membranes, a phenomenon also reported by Yamada et al. [12]. These observations prompted us to investigate the possibility to use multisquamase as a tool to quantitate the binding of prothrombin to procoagulant membranes.
1. Materials and Methods 1.1. Materials Dioleoyl-sn-glycero-3-phosphatidic acid (PA), sphingomyelin (SM), Hepes, Tris, EDTA, bovine serum albumin, and ovalbumin were purchased from Sigma Chemical Co. (St. Louis, MO). Dioleoyl-snglycero-3-phosphoserine (PS), dioleoyl-sn-glycero3-phosphocholine (PC), dioleoyl-sn-glycero-3-phosphoglycerol (PG), dioleoyl-sn-glycero-3-phosphoethanolamine (PE) were from Avanti Polar Lipids Inc., (Alabaster, AL). Phosphatidylinositol (PI) was a kind gift from Dr. E.M. Bevers, Department of Biochemistry, Maastricht University. The chromogenic substrate S2238 was supplied by Chromogenix (Mo¨lndal, Sweden). Ecarin, the prothrombin activator from Echis carinatus venom was from Pentapharm, Basel, Switzerland. Crude Echis multisquamatus venom was obtained from Latoxan (Rosans, France). Micro BCA protein assay kits were from Pierce (Rockford, IL). FPLC equipment and
column materials used for protein purification were purchased from Pharmacia (Uppsala, Sweden).
1.2. Proteins Human coagulation factors used in this study were isolated from fresh frozen plasma. Human prothrombin was purified according to DiScipio et al. [15]. Human thrombin was prepared from prothrombin activation mixtures as described by Pletcher and Nelsestuen [16]. Human Factor Va was purified as described previously [17]. Multisquamase was purified from the crude venom from Echis multisquamatus as described earlier [13]. Protein preparations were homogeneous and .95% pure as judged by SDS-PAGE according to Laemmli [18] and were stored at 2808C.
1.3. Protein Concentrations The concentration of multisquamase was determined with the micro BCA protein assay [19]. Prothrombin concentrations were determined after complete activation of prothrombin with ecarin and quantitation of meizothrombin with S2238 [20]. Factor Va was quantitated as described earlier [17].
1.4. Liposomes and Phospholipid Vesicle Preparations Phospholipid preparations (premixed in the desired composition in CHCl3/CH3OH (9:1 v/v)) were dried under a stream of N2 and suspended in a buffer containing 25 mM Hepes (pH 7.5) and 175 mM NaCl by vigorous vortexing in the presence of glass beads for 3 minutes. Centrifugable liposomes were obtained by centrifugation for 25 minutes at 900003g at 378C in a Beckman (Fullerton, CA, USA) TL100 ultracentrifuge after which the pellet was resuspended in buffer. The liposomes spun down quantitatively during subsequent centrifugation at 900003g at 378C for 25 minutes. Small unilamellar vesicles were prepared by sonication of liposomes for 10 min at 48C with a MSE (Loughborough, England) Soniprep150 ultrasonic disintegrator set at 8 mm peak to peak amplitude. Phospholipid concentrations were determined by phosphate analysis [21].
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
1.5. Determination of the Rate of Prothrombin Activation by Multisquamase Human prothrombin was preincubated in a buffer containing 25 mM Hepes (pH 7.5), 175 mM NaCl, 2 mM CaCl2 in the absence or presence of phospholipids at 378C. After 4 minutes, activation was started by addition of appropriate amounts of multisquamase, the purified prothrombin activator from Echis multisquamatus, in the same buffer. After different time intervals the amount of prothrombin activated was quantitated in aliquots withdrawn from the activation mixtures as described earlier [22] or by measuring the amidolytic activity towards S2238 (in a total volume of 250 mL 50 mM Tris-HCl (pH 7.9), 175 mM NaCl, 20 mM EDTA, 0.5 mg/ml ovalbumin and 235 mM S2238) in a microtiterplate at 378C using a SLT 340C kinetic reader set in the dual wavelength mode at 405–492 nm. The amounts of activated prothrombin were calculated from a calibration curve made with known amounts of active sitetitrated thrombin. The kinetic parameters, Km and Vmax, of prothrombin activation by multisquamase were obtained by measurement of the initial rate of prothrombin activation (V) at varying prothrombin concentration (S) and fitting the V vs. S curve to the Michaelis-Menten equation (V5Vmax3S/(Km1S) using nonlinear least squares regression analysis.
1.6. Binding Analysis In this article, we show that in prothrombin-phospholipid mixtures membrane bound prothrombin is not available for activation by multisquamase. To quantitate bound and free prothrombin in such reaction mixtures the initial rate of prothrombin activation by multisquamase was determined both in the absence and presence of phospholipid. At prothrombin concentrations well below Km, i.e., conditions at which initial rates of activation are directly proportional to the prothrombin concentration, the concentrations free prothrombin PTfree and membrane-bound prothrombin PTbound are calculated from Equations 1 and 2: PTfree5(V1PL/V2PL)3PTtotal
(1)
PTbound5PTtotal2PTfree
(2)
in which V1PL is the rate of prothrombin activation determined in the presence of phospholipid and
241
V2PL is the rate of prothrombin activation determined in the absence of phospholipid. At prothombin concentrations .100 nM, PTfree and PTbound were calculated using Equations 3 and 4: PTfree5(V1PL3Km)/(Vmax2V1PL)
(3)
PTbound5PTtotal2PTfree
(4)
in which V1PL is the rate of prothrombin activation in the presence of phospholipid and Km and Vmax are the kinetic parameters of multisquamase-catalyzed prothrombin activation determined in the absence of phospholipids.
2. Results Multisquamase-dependent prothrombin activation requires Ca21-ions and activation rates are strongly dependent on the ionic strength [12,13]. The experiments presented in this paper were performed throughout in 25 mM Hepes (pH 7.5 at 378C), 175 mM NaCl, 2 mM CaCl2 and 5 mg/ml BSA as carrier protein to passivate reaction tubes for adsorption of reactants.
2.1. Inhibition of Prothrombin Activation by Multisquamase in the Presence of Membranes Figure 1A shows that increasing amounts of negatively charged phospholipid vesicles (PS/PC; 20/ 80; M/M) gradually decreased the initial rate of activation of 0.2 mM prothrombin by multisquamase until at 500 mM phospholipid vesicles approximately 10% of the rate measured in the absence of pospholipids remained. Time courses of prothrombin activation were linear with time (,2% of the prothrombin available was allowed to activate) both in the presence and absence of phospholipid (data not shown). This experiment indicates that the inhibition of multisquamase-catalyzed prothrombin activation is caused by the fact that multisquamase can only activate free prothrombin and not membrane-bound prothrombin. To verify whether the observed inhibition of the initial rate of prothrombin activation was indeed due to binding of prothrombin to the membrane we performed an experiment in which 0.2 mM prothrombin was incubated with varying amounts of centrifugable large multilamellar PS/PC vesicles. These incubation mixtures were divided into two portions. In one series, prothrombin bound to the
242
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
Fig. 1. Effect of phospholipids on prothrombin activation by multisquamase. (A) Initial rates of activation of 0.2 mM prothrombin by 15 ng/ml multisquamase were determined at 378C in the presence of varying amounts of phospholipid vesicles (PS/PC; 20:80; M/M) in a reaction mixture containing 25 mM Hepes (pH 7.5), 2 mM CaCl2, 5 mg/ml BSA, and 175 mM NaCl and expressed as percentage of the rate determined in the absence of phospholipid. (B) Prothrombin (0.2 mM) was incubated at 378C in a reaction mixture containing 25 mM Hepes (pH 7.5), 2 mM CaCl2, 5 mg/ml BSA, 175 mM NaCl and concentrations of large multilamellar phospholipid vesicles (PS/PC; 20:80; M/M) indicated in the figure. Aliquots of these incubation mixtures were centrifuged 25 minutes at 378C at 90,0003g and the remaining part of the incubation mixtures were kept at 378C without centrifugation. The initial rate of prothrombin activation was subsequently determined by addition of 7.5 ng/ml multisquamase to 100 ml of the noncentrifuged incubation mixtures (s) or to 100 ml supernatant of the centrifuged incubation mixtures (m). The rate of activation determined in the absence of phospholipids was taken as hundred percent.
liposomes was removed by centrifugation for 25 minutes at 900003g at 378C and the other series was kept at 378C during the experiment. The initial rate of prothrombin activation was subsequently determined in the supernatant of the centrifuged incubation mixtures as well as in the noncentrifuged incubation mixtures by the addition of equal amounts of multisquamase. In Figure 1B, it is shown that the initial rates of prothrombin activation were identical irrespective of whether the prothrombinliposome complexes had been removed (centrifuged tubes) or not (noncentrifuged). In a control experiment, it was shown that multisquamase did not bind to the liposomes. Thus, multisquamase can only activate prothrombin that is not bound to the lipid and multisquamase can therefore be used as a probe to quantitate prothrombin binding to negatively charged lipids. We have also determined the effect of phospholipids on ecarin-catalyzed prothrombin activation (data not shown). Since membrane-bound prothrombin is still susceptible to activation by ecarin, this venom prothrombin activator cannot be used to quantify prothrombin binding to phospholipids.
2.2. Determination of Binding Parameters Figure 2A shows the initial rate of multisquamasedependent prothrombin activation determined at
varying prothrombin concentrations both in the absence and in the presence of 100 mM negatively charged phospholipid vesicles (PS/PC; 20:80; M/M). The kinetic parameters obtained from the experiment performed in the absence of lipids were: Km50.72 mM prothrombin and Vmax51.45 nM prothrombin activated/minute/mg activator. Since the rate of prothrombin activation in the presence of 100 mM PS/PC vesicles (Figure 2A, filled triangles) is solely dependent on free prothrombin, the concentration of free prothrombin at each prothrombin concentration can be calculated from the observed velocity using the plot of V vs. S determined in the absence of lipid (Figure 2A, filled circles) as a calibration curve. Figure 2B shows the Scatchard plot obtained from the amounts of prothrombin bound and free calculated from the data shown in Figure 2A. The binding parameters obtained from the Scatchard plot were Kd50.84 mM and n50.021 mmoles/mmoles phospholipid.
2.3. Effect of Different Anionic Phospholipids on the Binding of Prothrombin To test the general applicability for the use of multisquamase as a probe to quantitate prothrombin binding to membranes the experiment shown in Figure 1 was repeated with a variety of lipid membranes (Figure 3). Almost all membranes inhibited
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
243
Fig. 2. Prothrombin activation by multisquamase in the absence and presence of phospholipid vesicles as a function of the prothrombin concentration. (A) Varying amounts of prothrombin were activated by 6 ng/ml multisquamase at 378C in a reaction mixture containing 25 mM Hepes (pH 7.5), 2 mM CaCl2, 5 mg/ml BSA, and 175 mM NaCl, in the absence (d) or presence (m) of 100 mM phospholipid vesicles (PS/PC; 20:80; M/M). Prothrombin activation was determined as described under Materials and Methods, The kinetic parameters of prothrombin activation by multisquamase in the absence of phospholipids were obtained by fitting the data by using nonlinear least squares regression. Bound and free prothrombin at each prothrombin concentration in the presence of phospholipids (Figure 2A) were subsequently calculated as described under Materials and Methods. (B) Scatchard plot of prothrombin binding to phospholipid vesicles. The amounts of bound and free prothrombin in the presence of phospholipids were determined from the initial rate of prothrombin activation shown in panel A as described under Materials and Methods. The solid line shown in the Scatchard plot represents the best fit obtained by nonlinear least squares regression analysis. The binding parameters obtained were Kd50.84 mM and n50.021 mmoles prothrombin bound per mmole phospholipid.
multisquamase-dependent prothrombin activation and the extent of inhibition varied with the phospholipid composition of the membrane in a manner consistent with the known dependency of prothrombin binding on the nature of the polar head group in anionic phospholipid-containing lipid membranes. Table 1 summarizes the data obtained at 150 mM phospholipid. The strongest prothrombin binding, approximately 75% at 150 mM vesicles, was observed in the presence of vesicles that mimic the outer leaflet of the plasma membrane of activated platelets or with vesicles containing PS/PC/PE (20:60:20; M/M/M). Membranes containing PS/PC (20:80; M/M) or PA/PC (20:80; M/M) were only marginally less effective but membranes containing 10% PS showed a clearly reduced prothrombin binding. On vesicles containing 20% PG only weak binding was observed, whereas vesicles which contained only PC did not bind prothrombin (Figure 3, Table 1).
2.4. Detection of Binding at Low Prothrombin Concentrations (Effect of Factor Va) At prothrombin concentrations well below the Km5 0.72 mM for prothrombin, the initial rate of multi-
squamase-dependent prothrombin activation is linearly dependent on the amount of prothrombin available for activation. This, together with the fact that concentrations as low as 0.1 nM (meizo)thrombin are easily detectable makes multisquamase eminently suitable for detection of prothrombin binding at concentrations as low as 10 nM without the risk of interfering with the binding equilibrium by lowering the free ligand concentration through consumption of free prothrombin. This is illustrated by the experiment shown in Figure 4. In this experiment, the binding of 10 nM prothrombin to 25 mM vesicles composed of PS/PC (5:95; M/M), PS/PC (20:80; M/M), or PS/PC/PE/ SM/PI (10:39:27:19:5; M/M/M/M/M) was determined both in the absence and in the presence of 20 nM factor Va. Factor Va has been reported to promote the binding of prothrombin to negatively charged membranes [8]. In the absence of phospholipid, factor Va did not influence multisquamase-dependent prothrombin activation (data not shown) whereas addition of factor Va in the presence of lipid resulted in an increased inhibition of prothrombin activation. This indicates that factor Va stimulated the binding of prothrombin to the membrane (Figure 4). Although the absolute additional binding to the poor prothrombin binding vesicles
244
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
Fig. 3. The effect of phospholipid vesicles of varying composition on multisquamase-dependent prothrombin activation. Initial rates of prothrombin activation by 15 ng/ml multisquamase were determined at 378C in a reaction mixture containing, 0.2 mM prothrombin, 25 mM Hepes (pH 7.5), 2 mM CaCl2, 175 mM NaCl, 5 mg/ml BSA and amounts of phospholipid vesicles as indicated in the figure. The composition of the different phospholipid vesicles used were: m - m 100% PC, r - r PG/PC (20:80; M/M)), s s PS/PC (10:90; M/M), j - j PA/PC (20:80; M/M), d - d PS/PC (20:80; M/M), h - h; PS/PC/PE (20:60:20; M/M/M), and n - n; PS/PC/PE/SM/PI (10:39:27:19:5; M/M/M/M/M). Rates of prothrombin activation determined in the absence of lipid were taken as 100%.
(5:95; PS/PC) was lowest, the relative increase in binding to these vesicles was threefold as compared with the modest 1.5- and 1.2-fold increase observed for the other vesicles (Figure 4). Increasing the amounts of factor Va, however, did not result in a further enhancement of binding (not shown).
3. Discussion In this article, we show that the prothrombin activator purified from the venom of Echis multisquamatus, designated here as multisquamase, can be used as a tool to quantitate prothrombin binding to procoagulant membranes. The purification of multisquamase has been reported by us [13] and by Yamada et al. [12,23]. Multisquamase (designated multactivase by Yamada et al. [12]) has been shown to require the fragment 1 domain and the posttranslational carboxylation of the glutamic acid residues to g-carboxyglutamic acid (gla) to efficiently
convert prothrombin into meizothrombin [12,13]. In agreement with the data reported by Yamada et al. [12], we observed that the presence of negatively charged phospholipids inhibits multisquamasedependent prothrombin activation. In the present article, we show that binding of prothrombin to the lipid effectively blocks the activation and that multisquamase can only activate free prothrombin. This property of multisquamase was used to develop a method to quantitate the binding of prothrombin to phospholipids. The validity of the method is established by the fact that the parameters determined for binding of prothrombin to 100 mM 20:80 (M/M) PS/PC vesicles (Kd50.84 mM and n50.021 mmoles/mmole phospholipid) are in agreement with literature [2,4,6,24] and by the fact that the binding of prothrombin to phospholipid vesicles of different composition showed a variation in binding completely in accordance with the known variation in binding afinities of prothrombin for such membranes [2,4,6]. The method described here is not restricted to quantitate binding of human prothrombin to membranes but can also be used with bovine prothrombin. In a number of cases, the key experiments were repeated with bovine prothrombin in which case also only free prothrombin appears to be activated by multisquamase (data not shown). Due to the nature of the measurement, determination of initial rates of activation of prothrombin available for activation (i.e., free prothrombin), multisquamase is especially useful for quantitation of prothrombin binding at prothrombin concentration well below the Km of 0.7 mM for prothrombin. Under these conditions, changes in the rate of prothrombin activation are directly proportional to changes in the concentration of free prothrombin and the ratio of activation rates determined in the presence and absence of surface is a direct measurement of the fraction of free prothrombin. Since (meizo)thrombin concentrations of 0.1 nM are easily detectable using S2238, free prothrombin concentrations as low as 5 nM can be detected without interfering with the binding equilibrium (prothrombin free remains constant throughout the time course of the activation). Interestingly, this should make the use of multisquamase valuable in those cases where binding of prothrombin to planar phospholipid bi- or monolayer is studied since with
245
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
Table 1. Initial rate of multisquamase-dependent prothrombin activation in the absence and presence of 150 mM phospholipid vesicles of different composition Initial activation rate (nM PT-activated/minutes) Vesicle composition PC PG/PC PS/PC PA/PC PS/PC PS/PC/PE PS/PC/PE/SM/PI
(100%) (20:80; M/M) (10:90; M/M) (20:80; M/M) (20:80; M/M) (20:60:20; M/M/M) (10:39:27:19:5; M/M/M/M/M)
Without lipid
With lipid
% PTbound
5.50 5.13 5.40 5.33 5.18 5.28 5.29
5.51 4.85 3.91 2.21 2.01 1.42 1.67
— 7% 34% 65% 67% 78% 74%
Initial rates of prothrombin activation by 15 ng/ml multisquamase were determined at 378C in a reaction mixture containing 0.2 mM prothrombin and 150 mM phospholipid vesicles of varying composition in 25 mM Hepes (pH 7.5), 2 mM CaCl2, 175 mM NaCl, and 5 mg/ml BSA. The concentration of prothrombin available for activation was calculated from the rate of prothrombin activation determined using Equations 3 and 4 of the Materials and Methods and the kinetic parameters (Km50.72 mM and Vmax521.75 nM prothrombin activated per minute; see also Results). The percentage of prothrombin-bound was then calculated taking the amount of prothrombin available in the absence of phospholipid as 100%.
such membranes much lower dissociation constants have been reported [5,9]. At higher prothrombin concentrations it is necessary to obtain the kinetic parameters for activation (Vmax and Km) to enable calculation of free
Fig. 4. Effect of factor Va on prothrombin binding to phospholipid vesicles. Prothrombin (10 nM) was incubated at 378C in the absence (open bars) or presence of 20 nM factor Va (shaded bars) in a reaction mixture containing 25 mM Hepes (pH 7.5), 2 mM CaCl2, 175 mM NaCl, 5 mg/ ml BSA containing 25 mM phospholipid vesicles of lipid composition as indicated in the figure. After 4 minutes, prothrombin activation was started by adding 7.5 ng/ml multisquamase and the initial rate of prothrombin activation and the amount of membrane-bound prothrombin was determined as described under Materials and Methods.
prothrombin from the observed initial rate of activation. The kinetic parameters determined (Vmax5 1.45 mM prothrombin activated per minute per microgram activator and Km50.7 mM prothrombin) are in agreement with the kinetic parameters reported earlier by us [13]. At concentrations well above Km activation rates become independent of the prothrombin concentration. Thus, multisquamase cannot be used to detect prothrombin binding at free prothrombin concentrations higher than 2 mM. Moreover, since prothrombin activation by multisquamase is strongly dependent on Ca21ions [12,13] and ionic strength [13] care has to be taken that the experimental conditions are kept constant. The method described in this paper may offer a welcome addition to other methods for the determination of protein binding to phospholipid such as quasielastic light scattering [2,4,6], ellipsometry [5], or prothrombin binding to monolayers [9]. Since binding is usually detected by measurement of the protein mass adsorbed to the membrane it may be difficult to determine prothrombin binding in the presence of other lipid binding components. In such a case, as illustrated in the experiment shown in Figure 4, prothrombin binding can still be quantitated with the method described here. Further experiments are underway to establish the use of multisquamase to screen for prothrombin binding in more complex systems containing platelets (or even whole blood) or cultured cells.
246
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
This work was supported by Program Grant 900-526-192 from the Dutch Organisation for Scientific Research (NWO). R.J. Petrovan was supported by the Deutsche Akademie der Naturforscher “Leopoldina”. L. Johnsen was supported via the International Federation of Medical Student’s Association (IFMSA).
References 1. Davie EW. Biochemical and molecular aspects of the coagulation cascade. Thromb Haemostas 1995;74:1–6. 2. Nelsestuen GL, Broderius M. Interactions of prothrombin and blood-clotting factor X with membranes varying composition. Biochemistry 1977;16:4172–7. 3. Nelsestuen GL, Lim TK. Equilibria involved in prothrombin and blood-clotting factor X-membrane binding. Biochemistry 1977;169:4164–71. 4. Wei GJ, Bloomfield VA, Resnick RM, Nelsestuen GL. Kinetic and mechanistic analysis of prothrombin membrane binding by stopped flow light scattering. Biochemistry 1982;21: 1949–59. 5. Kop JMM, Cuypers PA, Lindhout T, Hemker HC, Hermens W. The adsorption of prothrombin to phospholipid monolayers quantitated by ellipsometry. J Biol Chem 1984;259:13993–8. 6. Cutsforth GA, Whitaker RN, Hermans J, Lentz BR. A new model to describe extrinsic protein binding to phospholipid membrane of varying composition: application to human coagulation proteins. Biochemistry 1989;28: 7453–61. 7. Giesen PL, Willems GM, Hemker HC, Stuart MC, Hermens WT. Monitoring of unbound protein in vesicle suspensions with off-null ellipsometry. Biochim Biophys Acta 1993;1147: 125–31. 8. Billy D, Willems GM, Hemker HC, Lindhout T. Prothrombin contributes to the assembly of the factor Va- factor Xa complex at phosphatidylserine-containing phospholipid membranes. J Biol Chem 1995;270:26883–9. 9. Ellison EH, Castellino FJ. Adsorption of bovine prothrombin to spread phospholipid monolayers. Biophys J 1997;72:2605–15. 10. Rosing J, Tans G. Inventory of exogenous prothrombin activators. Thromb Haemost 1991; 65:627–30.
11. Rosing J, Tans G. Structural and functional properties of snake venom prothrombin activators. Toxicon 1992;30:1515–27. 12. Yamada D, Morita T. Purification and characterization of a Ca21-dependent prothrombin activator, multactivase, from the venom of Echis multisquamatus. J Biochem 1997;122: 991–7. 13. Petrovan RJ, Govers-Riemslag JWP, Nowak G, Hemker HC, Rosing J, Tans G. Purification and characterization of multisquamase, the prothrombin activator present in Echis multisquamatus venom. Thromb Res 1997;88:309–16. 14. Yamada D, Sekiya F, Morita T. Isolation and characterization of carinactivase, a novel prothrombin activator in Echis carinatus venom with a unique catalytic mechanism. J Biol Chem 1996;271:5200–7. 15. DiScipio RG, Hermodson MA, Yates SG, Davie EW. A comparison of human prothrombin: Factor IX (Christmas factor); factor X (Stuart factor); and protein S. Biochemistry 1977;16:698–706. 16. Pletcher CH, Nelsestuen GL. The rate-determining step of the heparin-catalyzed antithrombin/thrombin reaction is independent of thrombin. J Biol Chem 1982;257:5342–5. 17. Nicolaes GAF, Tans G, Thomassen MCLGD, Hemker HC, Pabinger I, Varadi K, Schwarz HP, Rosing J. Peptide bond cleavages and loss of functional activity during inactivation of factor Va and factor VaR506Q by activated protein C. J Biol Chem 1995;270:21158–66. 18. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–5. 19. Smith PK, Krohn RI, Hermanson GT, Mallia AK, Gartner FH, Provenzano MD, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC. Measurement of protein using bicinchoninic acid [published erratum appears in Anal Biochem 1987 May 15;163:279]. Anal Biochem 1985; 150:76–85. 20. Tans G, Janssen Claessen T, Hemker HC, Zwaal RFA, Rosing J. Meizothrombin formation during factor Xa-catalyzed prothrombin activation Formation in a purified system and in plasma. J Biol Chem 1991;266:21864–73. 21. Bo¨ttcher CJF, van Gent CM, Pries C. A rapid
J.W.P. Govers-Riemslag et al./Thrombosis Research 92 (1998) 239–247
and sensitive sub-micro phosphorus determination. Anal Chim Acta 1961;24:203–7. 22. Rosing J, Zwaal RFA, Tans G. Formation of meizothrombin as intermediate in factor Xacatalyzed prothrombin activation. J Biol Chem 1986;261:4224–8. 23. Yamada D, Sekiya F, Morita T. Prothrombin
247
and factor X activator activities in the venoms of Viperidae snakes. Toxicon 1997;35:1581–9. 24. Lu YF, Nelsestuen GL. Dynamic features of prothrombin interaction with phospholipid vesicles of different size and composition: Implications for protein-membrane contact. Biochemistry 1996;35:8193–8200.