BIOCH1MIE, 1983, 65, 201-210.

Blood coagulation induced by Bothrops atrox venom : Idendification and properties of a factor X activator.
H616ne H O F M A N N , Claude D U M A R E Y and Cassian BON <> (Reu le 10-1-1983, acceptd le 9-2-1983).
R6sum6. Summary.

Unit~ des Venins, Institut Pasteur, 28, rue du Docteur-Roux, 75015 Paris, France.

L'action coagulante du venin de Bothrops atrox a ~t~ analys~e in vitro sur du fibrinogkne purifi~ et sur des plasmas normaux ou d~ficients en diff~rents facteurs de la coagulation sanguine. Cette ~tude a montr~ que le venin de Bothrops atrox a une action < < thrombin-like >> due ?tune ou des prot~ases ?t s~rine sensibles au DFP, mais que son action coagulante est en fait principalement due ~ des composants insensibles au DFP, qui activent la prothrombine et le facteur X (facteur Stuart). Nous avons purifi~ partiellement l'activateur du facteur X insensible au DFP du venin de Bothrops atrox. Darts des conditions natives, cette fraction purifi~e est constitude d'une prot~ine de masse mol~culaire 77.000 daltons, tandis que par ~lectrophorkse sur gel de polyacrylamide en presence de SDS elle contient principalement une cha?ne polypeptidique lourde (65.000) et un doublet l@er (12 - 13.000). L'activateur purifi( dont l'action coagulante est insensible au DFP, catalyse in vitro, de fa~on calcium-d@endante, la conversion du facteur X en facteur X,. Par ses propri(t~s mol~culaires il ressemble plus ~ l'activateur du venin de la vipkre de Russell qu'aux activateurs physioIogiques du facteur X.
Mots-cl6s : coagulation sanguine / aetivateur du faeteur X / venin de serpent : Bothrops atrox.

The coagulating activity of Bothrops atrox venom was investigated in vitro with purified fibrinogen, with normal plasma and with plasma deficient in various factors of the coagulation cascade. This study indicated that Bothrops atrox venom possesses a thrombin-like action caused by one or several serine proteases sensitive to DFP treatment, but that its clotting action is in fact mainly due to components insensitive to DFP which activate prothrombin and factor X (Stuart factor). We partially purified the DFP insensitive activator of factor X from Bothrops atrox venom and found it to be a protein of molecular weight 77,000. Analysis of the purified fraction by electrophoresis on polyacrylamide gel in the presence of SDS showed that it is mainly composed of one heavy polypeptide chain (65,000) and one light doublet (12 - 13,000). This activator is calciumdependent and catalyzes in vitro the conversion of purified factor X into factor Xa. By its molecular properties, it resembles the factor X activator from Vipera russellii venom rather than physiological activators.

Key-words : blood coagulation / factor X activator / snake venom : Bothrops atrox.

Introduction.

Fontana was probably the first to show in 1787 that snake venoms had the ability to clot human and animal blood, as reported by Weiss in 1969 [1]. In 1937, Eagle [2] showed that the venom of many snakes of the Crotalidae family coagulated citrated plasma. Further studies [3To whom all correspondence should be addressed.

6] showed that these venoms possess the ability to clot blood through a direct action on fibrinogen, similar to that of thrombin. Later on, the ~< thrombin-like >> enzymes from Agkistrodon rhodostoma [7, 8], Crotalus adamanteus [9] and Bothrops atrox [10-12] w e r e purified and their 15

202

H. Hofmann and coil.

10,000 - 20,000 [25-26]. The Russell's viper activator nicks the plasma factor X molecule in the presence of Ca 2+ ions, converting it into an active enzyme. The cleavage occurs at the same site as i~1 the physiological activation of factor X, but at variance with physiological activators, the venom activator is not inactivated by diisopropyl fluorophosphate (DFP) [25-32]. In the present study, we show that in addition to the serine proteases, batroxobin and thrombocytin, Bothrops atrox venom contains another protein which activates factor X and possesses several molecular and functional similarities with the factor X activator from Vipera russellii venom.

physical and enzymatic properties were characterized. Like thrombin, these enzymes are serine proteases which clot purified fibrinogen [9, 11, 12]. The thrombin-like enzyme of Bothrops atrox has been called batroxobin [11]. Unlike thrombin and crude Bothrops atrox venom, batroxobin does not activate blood platelets and factor X I I I [13, 16]. In fact, the complementary activity is due to another serine protease called thrombocytin which has very little clotting activity towards fibrinogen but readily activates factor X I I I and platelets [14-

16].
It was also reported by Eagle [2] in 1937 that in addition to their thrombin-like action, Bothrops atrox, Bothrops jararaca and Bothrops mummifera venoms clot plasma by directly or indirectly converting prothrombin into thrombin. This observation was confirmed by several other investigators [17-19]. Later, Nahas et al. [20] showed however that Bothrops jararaca and Bothrops atrox venoms can also activate factor X (Stuart factor), thus resembling Russell's viper (Vipera russellii) venom [21]. While Nahas et al. [20] initially suggested that only one enzyme was responsible for thrombin-like activity and factor X activation, Denson and Rousseau [22] and Furukawa and Hayashi [23] showed later that the two clotting activities of Bothrops jararaca are catalyzed by two differents proteins. More recently Nahas et al. [24] showed that most of the 26 Bothrop~ venoms which have been tested possess a thrombin-like activity which was associated with direct activation of prothrombin, activation of factor X, or both, and suggested that factor X activators may play an important role in the coagulating properties of Bothrops venoms. Factor X is a plasma glycoprotein involved in both the intrinsic and extrinsic pathways of blood coagulation : in the extrinsic pathway factor X is activated by a complex of factor VII, a tissue factor and Ca 2+ ions, whereas the activation of factor X via the intrinsic pathway involves the factor IXa and factor VIIIa, Ca 2+ ions and phospholipids as cofactors. These physiological activators convert factor X into factor X.,, a serine protease, which catalyzes the activation of prothrombin into thrombin in the presence of factor V~L, Ca 2+ ions and phospholipids. The presence in Russell's viper venom of a protein component which activates factor X was initially described by Mac Farlane [21] ; this activating enzyme has been purified and its molecular and coagulating properties studied. It is a glycoprotein of molecular weight 80,000 constituted of two polypeptide chains : one heavy chain of 60,000 and one light doublet in the molecular weight range of

Materials and Methods.

Materials.
The venom from Bothrops atrox was from the Pasteur Institute ,s'tock and that from Vipera russellii w~s provided by Kabi-diagnostica (Amsterdam, The Netherlands). Normal citrated, platelet-poor, human plasma was the supernatant obtained after a 15 minutes centrifugation of human blood mixed with 1/10 volume of 3.8 per cent s.cdium citrate. A pool obtained from 10 to 15 donors was stored at --20C. Bo.vine plasma deficient in factor I1, factor V or factor VII, purified human factor X, cephalin and kaolin were obtained from Diagnostica-Stago (Asni6res, France). Human plasma deficient in factor X was provided by Dade (Diidingen, Switzerland). Thromboplastin was a gi.ft of Drs Houbouyan and Roussy (Haematology Department, Ambroise Par~ Hospital, Paris, France). It was prepared from human brain, according to Caen et al. [33] Bovine tl~rombin and bovine fibrinogen were provided by H.offmann-Laro,che (Basel, Switzerland) and Povite (Neuilly, France) respectively. The chromogenic substrate, BenzoyMle-Glu-Gly-Argp-nitreaniline-HC1 (S-2222), was purchased from Kabidiagnostica (Amsterdam, The Netherlands) ; DEAEcellulose (DE-52, mierogranular) from What~nann Biochemicals LTD ,(Springfield Mill, England),; Sephadex G-100 (superfine)from Pharmacia (Uppsala, Sweden); ethylene-glycol-bis (p-aminoetyl-ether)N-N'-tetraacetic acid (EGTA) and diisopropyl fiuorophos!0hate (DFP) from Sigma (St Louis, USA) ; and acrylamide and N-N'-methylene-bis-acrylamide from Flnka (Bachs, Switzerland). All other chemicaIs, salts and solvents were of the best available quality, obtained from Merck (Darmstadt, RFA) or frc~rn Prolabo (P~fis, France) .

Assays for clotting and ]actor X activities.

Thrombin-like activity was determined as described by Furukawa and Hayashi {23] with the .following modifications: 0.2 ml of Michaelis buffer ,(30 mM diethylmalonylurea, 44 mM sodium acetate, 108 mM sodium chloride adjusted to pH 7.3. with HC1) containing

BIOCHIMIE, 1983, 65, n 3.

C o a g u l a t i o n f a c t o r X activator of B o t h r o p s
0.4 per cent fibrinogen was preincubated at 37C for 2 minutes and the clotting time was measured after addition cf 0.2 ml of the studied sample, also preincubated at 37C. A linear relationship between the logarithm of clotting time and the logarithm of the concentration of purified bovin thrombin was obtained over the range 1-100 ~g lhrombin per ml, which enabled us to define one thrombin-like unit as the equivalent of the activity possessed by 1 ~xg ef purified bovin thrombin. The clotting activity of Bothrops atrox venom and fractions obtained during venom fraction~tion were estimated according to the procedure of Howell and as modified by Caen et al. [33], by measuring at 37C the clotting time o,f normal or deficient citrated plasma (0.1 ml) after addition of the sample to be tested (0.1 ml) and of distilled water or of a 25 mM CaCI~ solution (0.1 ml). In the presence of Ca 2 ions, we obtained a two slopes relation,ship between the logarithm of

atrox venom.

203

racterizes the extrinsic pathway of coagulation was determined, according to the method o,f Quick, by incubating plasma (0.1 m l ) w i t h a thromboplastin solution (0.1-1 m g / m l ; 0.1 ml) at 3=7C for 2 minutes before performing the assay as recommended by C a e n e t al. [33]. The clotting activity of factor X, was determined by measuring the clotting time of 0.1 ml of h u m a n plasma congenitally deficient in factor X, after s~multaneous addition of CaCI~ 25 m M (0.1 ml) and of 0.1 ml Mlchaelis buffer containing the solution to be tested. The amidolytic activity of factor X~ was measured by recording the absorbance changes at 405 nm (1 cm pathlength) at 37'~C, in 1 ml assay medium containing 46 m M Tris-HC1 pH 8.3, 208 m M NaC1, 3.9 mM EGTA, 2.6 m~M CaCI~ and 0.8 m M chromogenic substra~e S-2222 (final concen~trations}. Activation of factor X was followed according to the procedure described by Di Scipio et al. [3:2] with the following modifications : Purified factor X (1 to 30 =~g per ml), was incubated at 37C for 1 minute in one volume of Michaelis solution buffered at pH 7.5 for determination of its clotting activity or at pH 8.3 for assaying its amidolytic activity. At zero time the samples to be tested for their factor X activating activity were added together with 5 m M CaC12 (final concentration). Aliquots were taken at different intervals and their clotting and amidolyfic aotivities were immediately measured.

300

100

LU

50

o \

IZ I.kO

20

Electrophoresis in polyacrylamide gels.

Polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS) were performed in 1.0 or 1.5 m m thick slab gels, according to the method of Laemmli [34], with the modifications described by Li and Ben [3'51. T h e following proteins were used as molecular weight standards : myosin from rabbR muscle (145,000), bovine serum albumin (67,000.)., catalase from bovine liver (62,50,0), chicken ovalbnmin ~(43,000)., aldolase (39,0,00) and lacti,codehydrogenase (35,000) from rabbit muscle, pig ,ehymotryps4n~ogen-A (25,000), soybean trypsin inhibitor (20,0,00,), and bovine ribonuclease (15,000).

8,

1 I
5

o \'. 3

~/.oo~oo 1

0.01

__

lOO [~g/ml]

VENOM

CONCENTRATION

FiG. 1. - - Coagulating activity o/ Bothrops atrox venom. The logarithm of clotting time, of a normal h u m a n plasma, measured in the absence (11--O ; I1--11) or in the presence of Ca 2+ ions ( O ~ ; n _ _ D ) i s plotted as function of the logarithm of the venom concentration. ( 0 - - 0 ; 0 - - - 0 ) native venom ( B - - B ; n__IB) DFPtreated venom. plasma clotting time and the logarittma of the concentration of Bothrops atrox venom .(figure 1). However, we used the linear part of the curve obtained at low venom concentrations to define one unit of venom clotting activity : one venom clotting unit being the activity of 1 ng o{ crude venom. This clotting time, which characterizes .the intrinsic pathway o.f coagulation, was also. determined, according to the method of Langdell, by incubating plasma (0.1 m l ) w i t h a mixture of cephalin and kaolin for 2 minutes at 37C before starting the coagulation assay as described by Caen et al. [33]. Similarly, the clotting time which cha-

Chemical modifications.
Inactivation of serine protease enzymes by D F P was performed by incubating the samples to be treated at room temperature for 16 hours (overnigh~t) with 25 m M D F P in 50 m M Tris-HC1, pH 7.5. The treated samples were then dialyzed twice for several hours against large volumes ot 5 m M Tris-HC1 pH 7.5.

Venom Jractionation. Bothrops atrox venom wa,s first fraction~ted by chromatography in D E A E cellulose (DE cellulose 52, microgranular) colunar~s a's described in detail by Dumarey et al. (in preparation) [36]. Briefly, 1.5 g of dried venom were dissolved in 15 ml of 50 mM sodium phosphate buffered a,t pH 7.5 and dialyzed against 1 1 of the same buffer. The crude v e n o m w a s then centrifuged 30 minutes at 8,000 g to remove the unsoluble material,
16

BIOCHIMIE, 1983, 65, n 3.

204

H. Hofmann and coll.

at a flow rate of about 2.5 ml per hour. Fractions of 1.8 ml were collected and their absorbance at 280 nm was determined as well as their plasma clotting activity in the presence of Ca '~+ ions. The tubes containing the clotting activity were pooled (fraction III~) lyophilized, dialyzed, lyophilized again and stored as previously described. The Sephadex G-100 column was calibrated with standard globular proteins to determine the molecular weight of the active component according to the method of Andrews [37], as modified by Radvanyi and Bon [38]. The following pro.teins were used as standards: bovine serum albumin (67,000), soybean trypsin inhibitor (20,000) and bovine cytochrome C (13,000). The void volume and total column volume were identified by the elufion of dextran blue ( ~ 1 000,000) and 5' adenosine monophos~.hate (AM,P), respectively.

which represented less than 1 per cent by weight, and the supernatant was applied to the top of a 220 ml D E A E cellulose column (2.5 45 cm), previously equilibrated with ~the same buffer. Elution was performed at room temperature (about 20C) at a flow rate of 30 ml per hour with : i) 500 ml ot 50 mMI sodium phosphate pH 7.5 ; ii) 1000' ml of a linear NaC1 gradiene~ from 0 to 0.3 M in the same sodium phosphate buffer ; and finally, iii) with 260 ml of 50 m M sodium phosphate pH 7.5, 1 M NaC1. The column effluent was collected in 5 ml fractions. T h e absorbance of the fractions was measured at 2'80 n m and their clotting activity was determined in the presence and in the absence of C a 2+ ions. A s ,shown in figure 2 we pooled the frac-

3.0 ~

tt~

1.5 "~

I--

2.o ~

-1.o -~

1.o

o.s

I--

E
0 ~.._~ 5____00 ' , ~ , ,

-=
I--

1000

,t500

EFFLUENT FIG. 2. - - Chromatography o] Bothrops atrox venom in a D E A E cellulose column. 1.5 g of crude venom were dissolved in 15 ml of 50 m M sodium phosphate buffer pH 7.5 and applied to the top of the column (2.5 45 cm), after dialysis and elimination of undissolved material. Elution was performed as indicated in the experimental procedures. 5 ml fractions were

[ml] collected and their absorbance measured at 280 nm ( t - - - - Q ~ t ) . Two types of coagulating activities were distinguished by measuring clotting time of a normal h u m a n plasma in the absence ( A - - A ) and in the presence of Ca 2+ ions ( A - - i ) . They are expressed as equivalent units of crude venom per ml.

tions containing a C a 2+ independent (I and II) and a Ca ~+ dependent (III) clotting activity. Fractions I, II and III were concentrated by lyophilization, dialyzed twice against a large volume of 5 m M Tris-HC1 pH 7.5, and finally lyophilized. They were stored at - - 2 0 C . 50 mg of fraction HI were dissolved in about 5 ml of 5 m M Tris-HC1 pH 7.5 in order to be further fractionated by filtration in a Sephadex G-100 (superfine) column (300 ml ; 2 190 era) previously equilibrated with the same buffer. Elulion was performed at 4C,
BIOCHIM1E, 1983, f~i, n 3.

Results and

Discussion.

COAGULATING ACTIVITIES

OF Bothrops atrox VENOM. We investigated the coagulating activity of Bothrops atrox venom in vitro by measuring the clotring time of normal human plasma under various conditions. Figure 1 shows that in the absence of Ca ~+ ions, a linear relationship is obtained

Coagulation factor X activator of Bothrops atrox venom.

between the logarithm of clotting time and the logarithm of venom concentration, at least in the concentration range of 0.3 to 300 p,g per ml. In the presence of CaC12, the coagulating activity of Bothrops atrox venom is much more pronounced since it occurs at concentrations as low as 1 ng per ml. It also appears to be more complex since a two slopes relationship is observed in the double logarithmic representation of figure 1. This suggests that the clotting activity of Bothrops atrox venom is due either to several components which act at various levels of the coagulation cascade, or to a single component having a multiple action on clotting factors, or that the two possibilities exist simultaneously. Since Bothrops atrox venom is known to contain two serine proteases, batroxobin (or reptilase) which splits fibrinogen [11-13], and thrombocytin which activates factor X I I I and blood platelets [14-16], it was of interest to inactivate these two enzymes by D F P and to compare the TABLE I. Thrombin-like activity of Bothrops atrox venom

205

the intrinsic pathway. In both cases, low concentrations of venom (about 10 p.g) reduced the amidolytic activity of purified thrombocytin and crude venom (results not shown), it did not significantly modify the capacity of the unfractionated venom to clot normal human plasma in vitro, both in the presence and in the absence of Ca ~ ions. These observations indicate : firstly, that DFP-sensitive enzyme(s), in particular batroxobin, is(are) responsible for the thrombin-like activity of Bothrops atrox v e n o m ; secondly that the thrombin-like activity of the crude venom is not significantly responsible for its coagulating activity on normal plasma ; and thirdly that the main coagulating activity of Bothrops atrox venom is Ca 2 dependent and insensitive to D F P treatment. The coagulating activity of Bothrops atrox ven o m was also investigated by using more specific plasma clotting times in a very significant manner, these effects being markedly more pronounced in the presence of Ca -~* ions than in their absence (table II). DFP-treatment of the venom reduced only slightly its clotting activity in the absence of Ca 2+ ions but not in their presence, thus confirming the observations previously described. Therefore, in addition to batroxobin and thrombocytin, the venom of Bothrops atrox contains one or several components, insensitive to DFP-treatment, which activates the common pathway of the coagulation cascade or simultaneously acts in a similar manner on the extrinsic and intrinsic pathways. The coagulating activity has been further studied by using plasma deficient in either factor I I (prothrombin), factor V, factor X (Stuart) or factor VII. In particular, the clotting time of a plasma deficient in factor I I (prothrombin), measured in the presence of brain thromboplastin was significantly longer than that of a normal plasma (table II). It was markedly reduced by low doses (a few lxg per ml) of native Bothrops atrox venom but not by DFP-treated venom, even at high doses (up to 50 p,g per ml), thus confirming the existence of a thrombin-like enzyme, such as batroxobin, in the venom. Conversely we observed that low doses (a few Ixg per ml) of native or DFP-treated venom were equally active in reducin~ the clottin~ times of plasma deficient either in factor V, in factor X or in factor V I I (table II). This confirmed the existence in Bothrops atrox venom of at least one DFP-insensitive component which activates p r o t h r o m N n directly or indirectly', or both. This conclusion promoted us to fractionate the venom to characterize this (these) component (s) and identify its (their) site of action in the coagulation cascade,

and fractions obtained during purification. Thrombin-like DFP activity treatment (U/rag)
Venom -260 + 1 Fraction I -630 + 10 Fraction II -420 + IO Fraction III -18 + 1 Fraction III~ -5 + 0.5 Reptilase -740 + 4 Thrombocytin -30 + 1 The thrombin-like activity was determined by measuring the clotting time of a solution of purified fibrinogen in the presence o.f the indicated fraction and comparing it with that obtained in the presence of purified thrombin, as described in the experimental procedures, and expressed as thrombin-like unit per mg of protein. assays, which characterize either the extrinsic or coagulating activity of DFP-treated venom with that of native venom. Although DFP-treatment almost completely abolished the fibrinogen clotting activity of purified reptilase and crude Bothrops atrox venom (table I), as well as the assays, which characterize either the extrinsic or

BIOCHIMIE, 1983, 65, n 3.

io

T A B L E II.

Coagulating activities of B o t h r o p s

a t r o x venom and fractions.

Clotting as:say performed in the presence of : Thromboplastin N~rmal 0 no coagulatiun 33 52 15 23 15 52 ~600 ~600 >600 ~600 240 210 220 240 ~600 3>600 ~600 ~600 140 n.d. 25 25 13 75 240 210 170 170 105 480 35 30 35 15 24 32 19 70 ~600 ~600 ~600 ~600 215 360 89 70 75 15 30 15 70 115 115 95 94 55 70 45 45 38 6 mM 155 0 n, coagulation 6 mM 195 0 6 mM no coagulation 310 Deficient in factor II Deficient in factor V Deficient in factor X 0 no c,,agutation 35 66 16 25 15 60 ~600 ~600 ~600 ~600 Nbrmal 6 mM 58 20 24 23 25 15 25 25 25 26 26 ~600 15 ~600 17 ~600 20 ~600 20 65 20 8 28 20 25 14 50 21 35 19 0 6 mM no,'q~ag,dati~u 36

Cephalin and kaolin

"~

Plasma

l)ef[cient in factor VII 6 mM 250 40 39 40 53 30 3I n.d. n.d. 24 25 0 no coa{tulatlun 100 125 52 65 55 108 n.d. n.d, ~600 ~600

CaCl~ Control

Venom (9.7 l~g/ml) D F P treated Venom (9.7 ~g/ml)

Fraction I (9.4 ;~g/ml) D F P treated Fraction I (9.4 ~g/ml)

Fraction II (24 ~g/ml) D F P treated Fraction II (24 p,g/ml)

16

Fraction III (8.7 g,g/ml) DI3P treated Fraction III (8.7 ~g/ml)

Fraction III~ (2.7 ~g/ml) D F P treated Fraction III~ (2.7 gg/ml)

The coagulating activities of Bothrops atrox venom and of its partially purified fractions obtained as explained in the text were investigated at the indicated concentrations before and after D F P treatment. Blood clotting times were measured in the presence of either cephalin and kaolin, or h u m a n brain thromboplastin, using normal and deficient plasma. The values are given in seconds.

Coagulation factor X activator of B o t h r o p s a t r o x venom.

207

FRACTIONATION OF Bothrops atrox VENOM, IDENTIFICATION OF A FACTOR X ACTIVATOR

Bothrops atrox v e n o m was f r a c t i o n a t e d by c h r o m a t o g r a p h y in D E A E cellulose c o l u m n as

presence or absence of cephalin a n d k a o l i n or of b r a i n t h r o m b o p l a s t i n ) even after D F P t r e a t m e n t (table II). F r a c t i o n I I I also r e d u c e d the clotting time of p l a s m a deficient in factor V I I m e a s u r e d in the presence of t h r o m b o p l a s t i n . It had h o w e v e r
O I O
o

T
'
1.0

(b] 0

:~. 'I

i
o

I 60
g:

100
I-tO --

2
20

uJ

0.5~
J
o L/LL
0
//

i! I \
i ----.~Z,,,
100 ]][1

_>
1
IO <

-10 6 < _1 Ud _1 O

,~"
.,n

,,.....,. .,. . / \
200

.......... i
300

~" I ,

tY~_,,

,",,,,

,Io

IS
O

]ff2]~'3

]K4

1ff5

]][6

EFFLUENT

Fio. 3. - - Gel filtration o/fraction 111 in a Sephadex G-IO0 column. The filtration was performed in a Sephadex G-I'00 column (2 10.0 cm) equilibrated with 100 mM N~CI, 50 mM Tris-HC1 pH 7.5. Fractions of 1.8 ml were collected, their absorbance at 280 nm was measured (O O) and their coagulatir~g activity deterd e s c r i b e d b y D u m a r e y et al. (in p r e p a r a t i o n ) [36]. A s shown in figure 2, nine c o m p o n e n t s were identified a c c o r d i n g to their p l a s m a coagulating activity a n d to their aggregating a n d anti-aggregating a c t i o n on b l o o d platelets. W h i l e fractions I to V I I all s h o w e d a coagulating activity, fractions I to I I I were the m o r e p o t e n t a n d have been studied in detail b y the m e t h o d s d e s c r i b e d a b o v e for the c r u d e venom. T h e clotting p r o p e r t i e s of fraction I and I I were similar to those of crude v e n o m : t h e y h a d a t h r o m b i n - l i k e activity, sensitive to D F P - t r e a t m e n t , w h i c h was r e s p o n s i b l e for their clotting action o n p l a s m a deficient in factor I I ; t h e y also c o n t a i n e d an active c o m p o n e n t which was not sensitive to D F P a n d directlv or indirectly activated factor I I ( p r o t h r o m b i n ) (table II). O n the other hand, fraction I I I c o n t a i n e d a very l o w t h r o m b i n - l i k e activity, c o r r e s p o n d i n g to less t h a n 5 p e r cent of that of c r u d e Bothrops atrox v e n o m (table I). F r a c t i o n I I I r e d u c e d efficiently the clotting .time o f n o r m a l h u m a n p l a s m a ( m e a s u r e d in the
BIOCHIM1E, 1983, 65, n 3.

(ml] mined in the presence of Ca 2 ions (A A) as described in the legend of figure 2. The column was calibrated with globular proteins (O--O) as indicated in the experimental section: (a) dextran blue ( > 1,000,000), (b) bovine serum albumin (67,000), (c) soybean trypsin inhibitor (20,00'0), (d) bovine cytochrome C (13~000) and (e) 5' adenosine monopho~sphate.
co o

~._%2oo
I-100 (3 "' 60 re"
\o

o%\

~,
.) O

20

Fro. 4. - - Polypeptide analysis o/ fraction HI~ by polyacrylamide gel electrophoresis in the presence of SDS. A sample containing 50 p~g of fraction lII2 was incubated for 2 minutes at 100C in 20 ~i .of 150 mM TrisHC1 pH 8.0 containing 4 per cent S'DS and 2 per cent [3-mercaptoethanol. It was then applied to the top of the gel and electrophoresis was performed as described in the experimental section. The polypeptide bands were stained with Coomassie Brilliant Blue.

208

H. Hofmann and coll.

major components of fraction IIIe were one polypeptide of molecular weight 65 - 67,000 and two polypeptides of 12 and 13,000. Since the apparent mass of the native protein responsible for the coagulating action of fraction IIIe was 77,000 daltons (figure 3), it appears that it could consist of one heavy polypeptidic chain of 65,000 --3,000 and one or two light chain of 12 or 13;000 --- 2,000. Figure 4 also shows that, in addition to these two main components, fraction 1112 contained several minor components of 20, 30 and 40 000 which amount to less than 10 per cent of the total protein and probably represent inactive contaminants.

very little or no coagulating action on plasma samples deficient in factors of the common pathway of the coagulation cascade, factor II, factor V or factor X (table II). These results led us to conclude that the coagulant activity was mainly due to the activation of factor X (Stuart factor). The factor X activator present in fraction III, was DFP-resistant and was calcium-dependent since fraction III failed to clot citrated normal plasma Fraction III was further fractionated by filtration in a Sephadex G-100 column. As shown on figure 3, its clotting activity eluted as a single peak (fraction III~_) having an apparent molecular

>-

300
i
200

e ~

t/
.
o2

;
'

~
p0

100 ~////A
Z.St _
5 5 10 15 20

ti
,01

~'
(J

F.
Q

~
25

INCUBATION

TiME

{mn]

FIG. 5. - - Kinetics of activation of purified human fac-

tor X by fraction IlL.

When the activation was followed by measuring factor X amidolytic activity (triangles), the reaction medium contained : 6.4 ~Lg/ml of purified human factor X, 1.2 txg/ml of fraction III2, 55 mM Tris-HC1 pH 8.3, 250 m M NaC1 and 5 m M CaCL ( A - - A ) or not (,~__A), The reaction was performed at 37C and was initiated by adding a sample of fraction III,.. It was stopped by taking 450 ~xl aliquots at the indicated times and immediately diluting them with 500 ~1 of 38 mM Tris-HC1 pH 8.3, 170 m M NaC1 ; 7.5 m M E G T A and 1.6 mM S-2222. The amidolytic activity was determined by measuring absorbance changes at 405 nm as a function of

time. The arrow indicates the maximal amidolytic activity obtained under the same conditions using Vipera russellii venom as a source of factor X activator. When the activation of factor X was fallowed by measuring its c.oagulating activi,ty (circles) in the presence ( I - - I ) er in the absence (O--O) of Ca 2+ ions, the experimental conditions were the same except that 15 Ixg/ml of factor X and 0.55 ~g/ml of fraction III~ were used. At the indicated time 10 ~I aliquots were diluted 100-fold and immediately assayed for their capacity to clot human plasma deficient in factor X. It is expressed as the difference between the clotting times observed in the presence and absence of activator.

weight of about 77,000 -+ 5,000. The coagulating properties of fraction ILL., were analyzed in detail and found to be very similar to those of fraction III except that fraction IIIe was about tenfold more efficient than fraction III (table II). The polypeptide composition of fraction III2 was analyzed by polyacrylamide gel electrophoresis in the presence of 1 per cent SDS after denaturation and reduction. Figure 4 shows that the
BIOCHIMIE, 1983, 65, n 3.

ACTIVATION OF PURIFIED HUMAN FACTOR X BY FRACTION IIIe.

Figure 5 shows that low concentrations (less than 1 t~tg per ml) of fraction III2 rapidly activate (in less than 5 minutes) and without latency both the amidolytic and the coagulating activities of purified human factor X. This observation unambiguously confirms the conclusion, obtained by measuring the clotting times of deficient plasma,

Coagulation [actor X activator o[ Bothrops atrox venom. that a factor X activator is responsible for the coagulating activity of fraction III2. The fact that fraction III2 promoted the irreversible activation of the totality of the factor X present in the assay, when added at a mass ratio of i : 30 (figure 5) suggests a catalytic mechanism. The factor X activator from Bothrops atrox venom is not however a serine protease since, at variance with the physiological activators, it is insensitive ot DFP treatment and strictly requires the presence of Ca 2+ ions to convert factor X into factor X,~

209

tions which inactivated all the serine proteases that we tested, confirming that the activator does not belong to that class of enzymes. On the other hand, we found that the Bothrops atrox factor X activator is active only in the presence of Ca 2+ ions. It differs from the physiological DFP-sensitive, activators of factor X, but resembles the factor X activators which have been isolated from Vipera russellii venom [25] and identified in Bothrops ]araraca venom [23]. The Bothrops atrox factor X activator appears as a useful tool in the understanding of coagulation mechanisms and of venom toxicity.

Conclusion.
The present investigation showed that the venom of Bothrops atrox exerts its coagulating activity on plasma in vitro in a complex manner, due to several components, acting at various levels of the reaction cascade which leads to clotting. We observed in particular an active thrombin-like activity characterized by its ability to clot purified fibrinogen and plasma deficient in factor II (prothrombin), in the presence and in the absence of Ca -+ ions. This activity is, at least in part, due to batroxobin, a serine protease which resembles thrombin [10-12]. We observed however that the thrombin-like enzymes, as well as the serine proteases of Bothrops atrox venom, were not essential for its coagulating activity since treatment with D F P did not significantly reduce its ability to clot normal plasma. We also showed, in agreement with the conclusion of Nahas et al. [24], that the venom of Bothrops atrox can convert prothrombin into tbrombin in vitro and activate factor X (Stuart factor). This coagulating activity was found to be insensitive to DFP, indicating that it is not due to serine proteases. The venom of Bothrops atrox therefore contains a coagulating activator which differs from the two enzymes so far identified : batroxobin [10-12] and thrombocytin [14-16]. We fractionated Bothrops atrox venom by chromatography on D E A E cellulose and by gel filtration on S~phadex G-100 and identified a coagulating fraction (fraction III2) which accounts for the factor X activation properties of the venom. The factor X activator of Bothrops atrox venom is a protein of 77,000 molecular weight, constituted of one heavy polypeptide chain (65~000) probably associated with one or two light polypeptides (12 to 13,000). Its activity was not affected by treatment with DFP, under condiBIOCHIM1E, 1983, 65, n 3.

Acknowledgments.
The authors are indebted to Dr. B. B. Vargaftig for encouragements and Drs. L. Houbouyan and J. Roussy for help]ul discussions concerning coagulation assays.

REFERENCES. 1. Weiss, H. J., Allan, S., Davidson, E. a Kochwa, S. (1969) Am. J. Med., 47, 625-629. 2. Eagle, H. (19'3.7) J, Exp. Med., 65, 613-629. 3. Eagle, H. ,~ Harris, T. N. (1937) J. Gen. Physiol., 20, 543-560. 4. Kellaway, C. H. (1937) Bull. John Hopkins Hosp., 60, 1-4. 5. Cesari, E. a Bcquet, P. (1937) Annls. Inst. Pasteur, Paris, 58, 6-25. 6. Vellard, J. (1938) Annls. Inst. Pasteur, Paris, 60, 511-548. 7. Esnouf, M. P. a Tunnah, G. W. (1967) Brit. J. Haematol., 13, 581-590. 8. Hatton, M. W. C. (1973)Biochem. J., 131, 799-807. 9. Marldand, F. S. ~ Damu3, P. S. (1971) J. Biol. Chem., 246, 6460-6473. 10. Stocker, K. & Egberg, N. (1973) Throm. Diath. Haemorrh., 1, 361-370. I1. Stocker, K. & Barlow, G. H. (1976) Methods Enzytool., 46, 2:14-222. 12. Holleman, W. H . . ~ Weiss, L. J. (1976) J. Biol. Chem., 251, 1663-1669. 13. Niewiarowski, S., Regoeczi, E., Stewart, G. J., Senyi, A. F. a Mustard, J. F. (1972) J. Clin. lnvest., 51, 685-7'00. 14. Niewiarowski, S., Kirby, E. P. a Stocker, K. (1977) Thromb. Res., 10, 863-869. 15. Kirby, E. P., Niewiarow~ki, S., Stocker, K., Kettner, C., Shaw, E. & Brudzynski, T. M. (1979) Biochemistry, 18, 3564-3570. 16. Niewiarowski, S., Kirby, E. P., Brudzynski, T. M. Stocker, K. (1979) Biochemistry, 18, 3570-3577. 17. Janzky, B. (950) Archs. Biochem., 28, 139-140. 18. Breda, R., Bernardi, R. ~ Sala, F. (1951) Giorn. Clin. Med., 82, 882. 19. Mi.chl, H. (1954) Monatsh. Chem., 85, 1240-1247. 20. Nahas, L., Denson, K. W. E. & Mac Farlane, R. G. (1964) Thromb. Diath. Haemorrh., 12, 35521. Mac Farlane, R. G. (1961) Br. J. Haemat., 7, 496-" 511.

210

H. H o [ m a n n and coll.
31. Furie, B. C., Furie, B., Gottlieb, A. J. a Williams, W. J. (1974) Biochim. Biophys. Acta, 365, 121132. 32. Di Scipio, R. G., Hermodson, M. A..~ Davie, E. W. (1977) Biochemistry, 16, 5253-5260. 33. Caen, J., Larr~eu, M. J. ~ Samama, M. (1975) in: <<H~mostase. M~thodes d' exploration et diagnostic pratique ~> (Caen, J., Larrieu, M. J. and Samama, M., eds) E~pansion scientifique fran~aise, Paris. 34. Laemmli, U. K. (1970) Nature, Lond., 227, 680685. 35. Li, Z. Y. a Bon, C. (1983) J. Neurochem., 40, 33834.9. 36. Dumarey, C., Bon, C. ~ Vargaftig, B. B. (in preparation). 37. Andrews, P. (1'964)Biochem. J., 91, 222-233. 38. Radvanyi, F. R. e Bon, C. (1982) J. Biol. Chem., 257, 12616-12623.

22. D errson, K. W. E. ~ Rousseau, W. E. (1970) Toxicon, 8, 15-19. 23. Furukawa, Y. a Hayashi, K. (1977) Toxicon, 15, 107-114. 24. Nahas, L., Kamiguti, A. a Barros, A. R. (1979) Thrombos. Haemostas., 41, 3q4-328. 25. KMel, W., Hermodson, M. A. ~ Davie, E. W. (1976) Biochemistry, 15, 4901-4906. 26. Amphlett, W. G., Byrne, R. Castellino, F. J. (1'9'82) Bigchemistry, 21, 125-132. 27. Williaans, W. J. ~ Esnouf, M. P. (1962) Biochem. L, 84, 52-62. 28. Schiffman, S., Theodo~-, I. ~ Rapaport, S. I. (1969) Biochemistry, 8, 1397-1405. 29. Jackson, C. M., Gordon, J. G. a Hanahan, D. J. (1971) Biochim. Biophys. Acta, 252, 255-261. 30. Fujikawa, K., Legaz, M. E. ~ Davie, E. W. (1972) Biochernistry, 11, 48,92-4899.

BIOCHIM1E, 1983, 65, n 3.