PII: S0041-0101(97)00129-3
Toxicon, Vol. 36, No. 2, pp. 257±267, 1998 # 1998 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0041-0101/98 $19.00 + 0.00
PURIFICATION AND VARIABILITY IN THROMBINLIKE ACTIVITY OF BOTHROPS ATROX VENOM FROM DIFFERENT GEOGRAPHIC REGIONS R. A. CAVINATO,1 H. REMOLD2 and T. L. KIPNIS3* Harvard Medical School, Boston, MA, U.S.A.; 2Instituto Butantan, Sao Paulo, Brazil; and 3 LaboratoÂrio de Biologia do Reconhecer, Centro de BiocieÃncias e Biotecnologia, Universidade Estadual do Norte Fluminense, Av. Alberto Lamego 2000, CEP 28015-620, Campos dos Goytacazes, R.J. Brazil
1
(Received 13 May 1997; accepted 13 September 1997)
R. A. Cavinato, H. Remold and T. L. Kipnis. Puri®cation and variability in thrombin-like activity of Bothrops atrox venom from di erent geographic regions. Toxicon 36, 257±267, 1998.ÐBothrops atrox snake venoms from two di erent Amazon regions, i.e., Manaus, AM (380.6'400S; 6080.1'6.00W) and TucuruõÂ , PA (380.42'300S; 4980.41'450W), were analyzed with respect to the thrombin-like activity component by elution pro®le on gel-®ltration and reverse phase HPLC chromatography, electrophoretic mobility on SDSPAGE, and enzymatic activity on ®brinogen. Despite some individual discrepancies among venom specimens, the thrombin-like activity present in the Manaus pool was eluted earlier compared with the TucuruõÂ pool but its enzymatic speci®c activity on thrombin was lower (s.a. = 6.0) than that observed in the TucuruõÂ pool (s.a. = 134.0). However, the electrophoretic mobilities of the pools were similar, with most protein bands being concentrated around three main regions, i.e., protein bands with an apparent mr of 100 kDa, of 38±37 kDa and 30 kDa. However, no signi®cant di erences were observed in amidolytic activity on the synthetic substrate Tos.Gly-Pro-ArgpNa, and there was no correlation between thrombin-like and amidolytic activities. A 32 kDa protein endowed with thrombin-like activity and speci®c activity of 2444 recognized and neutralized by horse anti-B. atrox antivenom, was puri®ed by the successive use of gel ®ltration, electrofocusing and reverse phase HPLC. # 1998 Elsevier Science Ltd. All rights reserved
INTRODUCTION
Snake venoms consist of mixtures of several proteins some of which have enzymatic and or toxic activities. Once accidentally introduced into the tissues of a victim or injected into experimental animals, the venoms induce local in¯ammatory reactions and systemic symptoms, the severity of which depends on the composition and concen* To whom correspondence should be addressed. 257
258
R. A. CAVINATO et al.
tration of the venom (Kocholati et al., 1971). Preliminary epidemiological studies in the State of Acre, (Amazon Region, northern Brazil) have shown that the number of deaths is underestimated (Pierini et al., 1996). Among the factors contributing to the health problems caused by bites or stings by venomous animals mainly in impoverished areas such as the Amazon Region, is the inadequate use of neutralizing antivenom antibodies. These antibodies are frequently inadequately characterized as neutralizing the toxins present in the venoms. To improve treatment of snake envenomation in northern Brazil, Colombia and Equador, in addition to o cial measures to facilitate the supply of antivenoms, studies should be undertaken to allow manufacturing su cient amounts of antivenoms. In recent clinical trials performed in Colombia in order to compare the e cacy and the incidence of early antivenom reactions (EAR) after administration of horse-derived antivenoms in patients envenomed by Bothrops, it was observed that, despite the e ectiveness of the preparations in reversing the signs of intoxication, larger doses were required to treat moderate and severe envenomations, with the occurrence of EAR symptoms in all cases (Cardoso et al., 1993; Otero-Patino et al., 1997). In this study, most of the victims were presumably bitten by B. atrox and the most relevant systemic symptoms were related to alteration of blood coagulation, the reversal of nonclottable to clottable blood being the important sign recorded and used to monitor the e cacy of treatment. These data indicated that the antivenom antibody preparations although capable to improve the clinical symptoms either were of low speci®city or did not contain a su cient amount of speci®c antibodies to neutralize the toxic components present in the venoms. Several groups have shown that venoms from various snakes species assumed to belong to the genus Bothrops (Hoge, 1965) share many common antigenic components (Moura da Silva et al., 1991). Recent investigations using a combination of the classic systematics based on multivariate and mitochondrial DNA analyses, however, have shown that the Atrox complex di ers from the genus Bothrops (WuÈster et al., 1996; SalomaÄo et al., 1997). This fact, together with the epidemiological studies and clinical trials reported above, emphasizes the need for further information about the major toxic components of B. atrox venom. The objective of the present study is to isolate and characterize the thrombin-like component of B. atrox venom.
MATERIAL AND METHODS
Venoms Individual samples were collected from 12 adult specimens of B. atrox from the Amazon Region: 6 from Manaus, Amazonas, Brazil (380.6'400S; 6080.1'6.00W) (NuÂcleo de Animais Pec° onhentos do Instituto de Medicina Tropical do Amazonas, Universidade do Amazonas), and 6 from TucuruõÂ , ParaÂ, Brazil (380.42'300S; 4980.41'450W) (LaboratoÂrio de Herpetologia, Instituto Butantan, SaÄo Paulo). The venoms of 5 snakes from Manaus were pooled and will be referred to hereafter as the Manaus-pool, while the venom non included in the pool was labeled M1. The venoms from TucuruõÂ were labeled T1, T2, T3, T4, T5 and T6, the TucuruõÂ -pool resulting from a mixture containing equal amounts of dry weight venom from each individual (T1±T6). The venoms were ®ltered through a 0.45 mm membrane, lyophilized, divided into 10 mg aliquots and stored at ÿ208C.
Bothrops atrox thrombin-like activity puri®cation
259
Puri®cation Chromatography on sephacryl S-100. Venom samples of 10 mg dry weight in 20 mM Tris±HCl bu er, pH 7.5, were applied to a Sephacryl S-100 (Pharmacia, Uppsala) column previously equilibrated with the same bu er. Elution was accomplished at a ¯ow rate of 30 ml/h, the protein contents were monitored spectrophotometrically at 280 nm and the thrombin-like activity was assayed. Preparative electrofocusing. Sephadex G-75 (Pharmacia, Uppsala) dehydrated with methanol was resuspended in 25 ml of a solution containing ampholyte in the pH range 4.5±6.5 (1:16) and 2.0 ml of ampholyte in the pH range of 2.0±9.0 plus 2.5 mg of the fraction from Sephacryl S-100 chromatography containing the thrombin-like activity were applied to an isoelectrofocusing plate. Isoelectrofocusing (700 V, 15 mA) was performed at 48C for 24 h. Thirty-six gel slices were then removed and transferred to 2.0 ml of deionized water, the pH was determined and the elution of the proteins completed by adjusting the pH to 7.5 and dialyzing the eluted samples against 20 mM Tris±HCl buffer, pH 7.5. The protein contents were spectrophotometrically determined at 280 nm and the thrombin-like activity was determined in the fractions. The fractions containing thrombin-like activity were pooled and concentrated using a Speed-Vac apparatus. HPLC chromatography. The material rich in thrombin-like activity obtained by preparative electrofocusing (140 mg) was applied to a reverse phase CLC-C8 column previously equilibrated with the eluting solution A (0.1% tri¯uoracetic acid) as starting bu er. The elution was performed in two steps: the ®rst consisted of a 30 min period at a ¯ow rate of 1.0 ml/min with a gradient of 0.5% starting bu er and 80% limit bu er (9 parts of acetonitrile:1 part of solution A) and the second consisted of a 5 min period with the limit bu er alone. Samples of 1.0 ml were collected and protein contents monitored at 280 nm. The fractions corresponding to protein peaks were pooled and dehydrated using a Speed-Vac apparatus, and the presence of thrombin-like activity was assayed. Analytical and assay methods SDS-PAGE. The samples were subjected to 10% acrylamide-SDS electrophoresis according to Laemmli (1970). Electrophoretic runs were carried out in a Mini-Protean Tm II equipment (Bio-Rad, CA) with a constant voltage of 100 V. The protein bands were stained either with Coomassie blue or with silver stain. Myosin (200 kDa), phosphorylase (105.1 kDa), bovine serum albumin (69.8 kDa), ovalbumin (43.3 kDa), carbonic anhydrase (28.3 kDa), lactalbumin (18.4 kDa) and lysozyme (34.3 kDa) (Gibco BRL, MD), were used as molecular mass markers. Zymogram. The proteins to be analyzed for proteolytic activity were applied to 10% acrylamide gel containing 1% ®brinogen. The electrophoretic runs were performed essentially as described for SDS-PAGE, except that the current used was 16 mA and the temperature 48C. The gels were then immersed in 2.5% triton at room temperature for 1 h, and next in 100 mM glycine bu er, pH 8.3, for 12 h at 378C, and then stained with Coomassie blue. The presence of proteolysis was determined visually as unstained bands. Protein assay. The protein content was determined by the method of Lowry et al. (1951).
260
R. A. CAVINATO et al.
Thrombin-like assay. One hundred microliters of total venom or of the puri®ed fractions were added to glass tubes (1.0 7.0 cm) containing 400 ml of puri®ed bovine ®brinogen (2.0 mg/ml). The preparation was incubated at 378C and the time of coagulation was determined. Coagulation is expressed as coagulation index (CI) obtained by the equation: CI = tÿ1 1000, where tÿ1=inverse of time. The minimum coagulating dose (MCD) is de®ned as the smallest amount of venom or puri®ed fraction capable of completely coagulating ®brinogen within 1 min. Assay for amidolytic activity. Ten microliters of venom or puri®ed fractions (10 mg) were added to glass tubes (1.0 7.0 cm) containing 0.15 mM of the synthetic substrate Tos-Gly-Pro-Arg-pNa (Sigma, St. Louis, MO) in 1.2 ml of 50 mM, Tris±HCl bu er, pH 8.0, and incubated at 378C for 1 h. The p-nitroaniline released was read spectrophotometrically at 405 nm and the results are expressed as mU/mg, where mU = nmol minÿ1. The inhibition assays were performed by incubating 5 mM of the inhibitors phenylmethylsulfonyl ¯uoride (PMSF) and Na-p-tosyl-L-lysyl chlormethyl ketone (TLCK) (Sigma) in 0.02 M Tris±HCl bu er, pH 8.0, with the venom or its puri®ed fractions for 15 min at 378C before being added to the mixture. The remaining coagulating activity was calculated as percent activity in relation to the control preparation incubated without inhibitors, which was considered to be 100% activity. In vitro neutralization. Five microliter samples of puri®ed B. atrox thrombin-like protein were incubated with di erent amounts of anti B. atrox venom immune serum (2.5, 5.0 and 10.0 ml sera) at 378C for 1 h. The immune precipitates were removed by centrifugation, the remaining thrombin-like activity was determined as described and percent inhibition was calculated as a function of the control of thrombin-like sample incubated with saline. Antisera. Hyperimmunized horse serum against B. atrox venom was supplied by the Instituto Butantan, SaÄo Paulo. Mouse immune sera against the B. atrox puri®ed thrombin-like protein were obtained by priming adult female BALB/c mice with 10 mg of the corresponding protein in 10 mg of Al(OH)3 as adjuvant, followed by three boosters at seven day intervals. Blood was collected and the sera were stored at ÿ208C until the time for use.
RESULTS
Heterogeneity of whole venom of B. atrox from Manaus and TucuruõÂ The venoms of B. atrox Manaus and TucuruõÂ and pools from Manaus and TucuruõÂ were compared in terms of electrophoretic mobility, chromatographic resolution and distribution of the thrombin-like activity and amidolytic enzymatic activity. Electrophoretic pro®le of the polypeptides present in the venom preparations Figure 1 shows the comparative electrophoretic pro®les of individual and pool samples of B. atrox venoms from TucuruõÂ (lanes 1±7) and from Manaus (lanes 8±9). All venoms present many bands around 100 kDa, 40±50 kDa and some bands with low mm; the individual samples of TucuruõÂ and Manaus pool (lanes 4 and 5) present strong
Bothrops atrox thrombin-like activity puri®cation
261
Fig. 1. Pro®le of Bothrops atrox venom obtained by SDS-PAGE. Twenty mg of venom from the following samples were used: (1) TucuruõÂ (sample 6); (2) TucuruõÂ (sample 5); (3) TucuruõÂ (sample 4); (4) TucuruõÂ (sample 3); (5 TucuruõÂ (sample 2); (6) TucuruõÂ (sample 1); (7) TucuruõÂ (pool); (8) Manaus (pool); (9) Manaus (sample 1).
bands above 50 kDa; besides, an specimen of TucuruõÂ and the pool from TucuruõÂ present faint correspondent bands above 50 kDa (lanes 7 and 8). In order to verify the proteolytic enzymatic activity, SDS electrophoresis of the venom samples was repeated using polyacrylamide gels containing ®brinogen as substrate (Fig. 2). Unequivocal pro-
Fig. 2. Enzymatic activity of Bothrops atrox venoms on ®brinogen incorporated into polyacrylamide gel in the presence of SDS (zymogram). (1) TucuruõÂ (sample 6); (2) TucuruõÂ (sample 5); (3) TucuruõÂ (sample 4); (4) TucuruõÂ (sample 3); (5) TucuruõÂ (sample 2); (6) TucuruõÂ (sample 1); (7) TucuruõÂ (pool); (8) Manaus (sample 1); (9) Manaus (pool).
262
R. A. CAVINATO et al.
teolysis was observed in some samples (lanes 2, 4, 5, 6, 7 and 8). Except sample 6 the proteolytic activity was distributed in two close apparently related bands containing enzymatic activity restricted to the electrophoretic region of 28±40 kDa. When the proteolysis was assayed in the presence of PMSF a serine proteinase inhibitor, bands of proteolysis were not detected (data not shown). Gel ®ltration chromatography Equal amounts of protein of B. atrox venom were chromatographed on Sephacryl S100 and their coagulant activity on ®brinogen was assayed. The corresponding chromatographic pro®les are shown in Fig. 3. The chromatographic patterns of the pools and of the individual samples present 4±5 broad protein peaks. Thrombin-like activity was eluted at 187±195 ml of the e uent just before the ®rst protein peak (Manaus pool,
Fig. 3. Gel ®ltration of Bothrops atrox venoms on a Sephacryl S-100 column. Protein was monitored at 280 nm and clotting activity was assayed on bovine ®brinogen. (A) Bothrops atrox venom (Manaus pool); (B) Bothrops atrox venom (TucuruõÂ pool).
Bothrops atrox thrombin-like activity puri®cation
263
Table 1. Speci®c activity and yield of Bothrops atrox venom fractions Fractions
Protein total recovery (mg)
$
Venom Sephacryl% Eletrofocusing}
35 3.8 0.7
Clotting activity total recovery (%)
(U*)
100 11 2.0
6210 5165 1784
speci®c activity (%) 100 83 29
puri®cation ratio
(U/mg) 177 1341 2444
1 7.6 14
*1 U = enzyme quantity in 100 ml that clots 0.8 mg bovine ®brinogen at 378C within 60 s. 6.2 mg in 100 ml. %0.82 mg in 100 ml. }0.45 mg in 100 ml, that clots 0.8 mg bovine ®brinogen at 378C within 60 s. $
panel A) whereas the corresponding activity of the TucuruõÂ venom pool was eluted mainly at 225±237 ml (panel B) between the second and third protein peak. Individual samples were chromatographed under the same conditions for comparison. Thrombin like activity was eluted as observed in the corresponding pool (TucuruõÂ ), except for the M1 sample where the respective activity was eluted at 200±250 ml between the second and third protein peak (data not shown). Amidolytic activity Individual samples of B. atrox venom from Manaus and TucuruõÂ were analyzed for their ability to cleave the synthetic substrate Tos-Gly-Pro-Arg-pNa and to coagulate bovine ®brinogen. Table 1 indicates that all the venoms tested were active against both bovine ®brinogen and the synthetic substrate. The intensity of these activities varied randomly in the tested samples. Puri®cation of thrombin-like activity from B. atrox venom Thrombin-like activity was puri®ed from M1 venom by sequential gel-®ltration chromatography on Sephacryl S-100, preparative electrofocusing and HPLC reverse chromatography. Table 2 summarizes the results of the recovery, speci®c activity and puri®cation ratios after the last procedure. The puri®ed protein contained a speci®c activity of 2444 vs 177 for the nonpuri®ed venom preparation, representing a 14-fold puri®cation. This protein was rechromatographed by HPLC reverse phase chromatography Table 2. Direct clotting activity on bovine ®brinogen (MDC-F)* and amidolytic activity of Bothrops atrox venoms from Manaus and TucuruõÂ DMC-F (mg/ml)
venoms Bothrops Bothrops Bothrops Bothrops Bothrops Bothrops Bothrops Bothrops Bothrops
atrox atrox atrox atrox atrox atrox atrox atrox atrox
(Manaus-pool) (Manaus 1) (TucuruõÂ -pool) (TucuruõÂ 1) (TucuruõÂ 2) (TucuruõÂ 3) (TucuruõÂ 4) (TucuruõÂ 5) (TucuruõÂ 6)
1830 62 82 313 47 52 2060 58 733
Speci®c activity (U/mg) 6 177 134 35 234 211 5 190 15
Amidolytic activity (mU/mg) 440 290 300 360 250 90 260 320 330
1 U = enzyme quantity in 100 ml that clots 0.8 mg bovine ®brinogen at 378C within 60 s. mU/mg = m/mol/ min/mg of venom considering the extinction coe cient of pNa to be 10 000. *MDC-F: Minimum clotting dose on ®brinogen. Quantity of venom that clots 0.8 mg bovine ®brinogen at 378C within 60 s.
264
R. A. CAVINATO et al.
Fig. 4. SDS-PAGE pro®le of the thrombin-like enzyme resulting from rechromatography by reverse phase HPLC.
and the single protein peak obtained was analyzed by SDS-PAGE electrophoresis. A major protein band of 32.3 kDa was detected and a very faint band of 43.3 kDa was also observed (Fig. 4). By isoelectrophocusing the pI was between 5.2±5.6. This puri®ed protein was further analyzed in terms of its enzymatic and antigenic properties. In addition to showing activity in terms of bovine ®brinogen coagulation, the puri®ed thrombin-like protein was capable of cleaving the synthetic substrate Tos-Gly-Pro-Arg-pNa (data not shown). The puri®ed protein was also recognized and its enzymatic activities were blocked by antibodies present in serum from horses hyperimmunezed with a mixture containing venom of seven species of Bothrops (Table 3). Table 3. Antibody neutralizing clotting activity of thrombin-like enzyme on bovine ®brinogen sera Thrombin-like fraction + 0.15 M NaCl Mouse pre-immune serum Mouse anti-total venom serum Horse anti-Bothrops serum Mouse anti-thrombin-like serum
% inhibition 0 14 81 100 87
20 ml of sera were incubated with 1 ml of puri®ed thrombin-like fraction for 60 min. at 378C (20 ml serum/ml puri®ed thrombin). After this period the supernatants were assayed as described. % inhibition was calculated as a function of the control of the thrombin-like sample incubated with 0.15 M NaCl.
Bothrops atrox thrombin-like activity puri®cation
265
DISCUSSION
Information from the Word Health Organization for South America indicates that 90% of venomous snake bites are caused by Bothrops of several species (WHO, 1981; Kamiguti and Cardoso, 1989). The high incidence of snake bites in the Amazon Region, a zone with large reserves of tropical rain forest, have allowed the development of an important biodiversity including over one hundred snakes species. Most bites are caused by B. atrox. Regional epidemiological studies in Colombia have demonstrated a 6% frequency of sequelae and a 6% death rate (Otero et al., 1992a,b). However, reports from the NuÂcleo de Animais Pec° onhentos do Instituto de Medicina Tropical do Amazonas, Universidade do Amazonas the incidence of snake bites is not so high and the death rate is far below around 0.7% (personal communication from P. Buhrnheim). It has also been reported that the venom of B. atrox, like the venoms of all Bothrops species, is endowed with proteolytic, de®brinating, haemorrhagic, myotoxic, edemaforming and indirect hemolytic activities. The intensity of each activity per se or the relative concentration of each in individual samples of venom vary according to the region where the snakes are captured (Otero et al., 1992b). This information demonstrates the need to supply these vast impoverished regions with antivenoms consisting of an antibody mixture capable neutralizing the toxins found in the snake venom. To manufacture these antivenoms, correct knowledge of the snake venom heterogeneity and detailed characterization at both the biological and molecular levels of the components responsible for the induction of symptoms and tissue lesions in bite victims are necessary. The objective of the present study was to isolate and characterize the thrombin-like protein found in B. atrox venom. This component is responsible together with factor X activator (Nahas et al., 1964) and prothrombin activator (Nahas et al., 1979) and ®brinogenases (Kamiguti et al., 1988) in blood uncoagulability, low levels of ®brinogen and ®brinolysis, the hallmark e ects of B. atrox venom. Based on ®brinogen coagulation, gel ®ltration on Sephacryl S-100, electrophoretic analysis by SDS-PAGE and visualization of the electrophoretic bands with proteolytic activity, B. atrox venoms can be divided into two groups: (1) venoms with more intense ®brinogen coagulating activity, ®ltered between dextran blue and BSA, demonstrating a 32 kDa band with ®brinogen-proteolytic activity (venoms T2, T3, T5 and M1), and (2) venoms with lower ®brinogen-coagulating activity eluted from the column after dextran blue, with no apparent 32 kDa proteolytic band (T4, T6 and M-pool). In addition, the pool of venoms collected at TucuruõÂ showed two distinct bands with ®brinogen-coagulating activity. These results suggest, therefore, that at least the thrombin-like activities present in the pools from Manaus and TucuruõÂ are distinct, a variability probably due to the di erent distribution of the snakes. However, since the activity of the venoms also varied individually, a possible e ect of the age of the snake and of the housing conditions prevalent at the time of venom collection cannot be ruled out. Interregional di erences among other biological activities in snake venoms have been described. For instance, B. asper venom from the Atlantic coast contains compounds with higher activity than the corresponding venom from the Paci®c coast (Aragon-Ortiz and Gubensek, 1981). Some enzymatic activities and local and systemic e ects produced by B. atrox varied according the site in Colombia where the snakes were captured (Otero et al., 1992a,b). Also, signi®cant di erences in coagulating activity between B. atrox from Brazil and Peru, the former being more potent, have been reported (Sanchez et al., 1991). These di erences, however, may not simply be due to environmental e ects, but also to factors inherent in the snake itself. Bonilla et al. (1973), Kamiguti and
266
R. A. CAVINATO et al.
Hanada (1985) and Furtado et al. (1991) have provided data indicating that the coagulating activity tends to decrease as the snake reaches adulthood. In agreement with these observations there are also clinical reports indicating that the blood incoagulability observed in victims of snake bites is more persistent when the bite was in¯icted by young snakes (Rosenfeld et al., 1959; Kamiguti et al., 1986; Ribeiro and Jorge, 1989; Kouyoumdjian and Polizelli, 1989). The role of the origin of the snakes in the composition of the venom suggested above can be better addressed by the analysis of individual components. When an attempt is made to combine practical and research aspects, individual venom components should be studied. The thrombin-like activity of B. atrox venom, one of the mediators of a vast range of local and systemic e ects observed after envenomation, was chosen for analysis in the present study. The two-step-procedure delineated in this communication, i.e., gel ®ltration on Sephacryl S-100 followed by isoelectrofocusing in the 2.0 to 9.0 pH range, yielded a homogeneous 32.3 kDa protein with thrombin-like activity. The protein obtained had a speci®c activity of 2444 compared to the 177 speci®c activity of whole venom, representing a 13.7-fold puri®cation. SDS-PAGE showed that this protein was concentrated in one 32.3 kDa protein band endowed with ®brinogen coagulating activity and amidolytic activity. This activity was blocked by antibothropic horse serum, indicating that the trombin-like protein is both antigenic and immunogenic for horses. This puri®cation was done with venom of the sample M1 that shows only one proteolitic band on SDS-PAGE electrophoresis as seen in Fig. 2. Recent results obtained in our laboratory show that the puri®cation of samples of TucuruõÂ presenting the two proteolitic bands results also in two protein bands of 32 kDa that by terminal amino-acid sequencing are identical between themselves and present high homology to thrombinlike (Petretski et al., manuscript in preparation) isolated from B. moogeni. Since these animals are extensively used to produce antivenoms for therapeuthic purposes, this information indicates that antivenom can be produced against the thrombin-like activity of B. atrox. The method for the puri®cation of the thrombin-like component from B. atrox venom described here can be used to compare the corresponding protein from various specimens and therefore to study regional variability with respect to this component. The thrombin-like component is probably one important mediator of local and systemic tissue lesions and should be well represented both quantitatively and qualitatively in the antigenic mixtures used to immunize horses. Likewise, this component, due to its biological similarity to a functional counterpart in mammalian plasma, may represent an important molecule warranting detailed studies aiming at the discovery of structural features leading to the development of a model for the construction of speci®c inhibitors. AcknowledgementsÐWe thank Mauro Sucupira for technical assistance, Dr Paulo Buhrnheim, NuÂcleo de Animais Pec° onhentos do Instituto de Medicina Tropical do Amazonas, Universidade do Amazonas) for providing the venoms from Manaus, and Dra M. Fatima Furtado (LaboratoÂrio de Herpetologia, Instituto Butantan) for providing the venoms from TucuruõÂ . RAC is the recipient of a CAPES scholarship, and TLK is the recipient of a CNPq grant.
REFERENCES Aragon-Ortiz, F. and Gubensek, F. (1981) Bothrops asper venom from the Atlantic and Paci®c zones of Costa Rica. Toxicon 19, 797±805. Bonilla, C. A., Faith, M. R. and Minton, S. A. (1973) L-amino acid oxidase, phosphodiesterase, total protein and other properties of juvenile timber rattlesnake (C.h. horridus) venom at di erent stages of growth. Toxicon 11, 301±303.
Bothrops atrox thrombin-like activity puri®cation
267
Cardoso, J. L. C., Fan, H. W., Franc° a, F. O. S., Jorge, M. T., Leite, R. T., Nishioka, S. A., Avila, A., SanoMartins, I. S., Tomy, S. C., Santoro, M. L., Chudzinski, A. M., Castro, S. C. B., Kamiguti, A. S., Kelen, E. M. A., Hirata, M. H., Mirandola, R. M. S., Theakston, R. D. G. and Warrel, D. A. (1993) Randomized comparative trial of three antivenoms in the treatment of envenoming by lance-headed vipers (Bothrops jararaca) in SaÄo Paulo. Brazil. Q. J. Med. 86, 315±325. Furtado, M. F. D., Maruyama, M., Kamiguti, A. S. and Antonio, L. C. (1991) Comparative study of nine Bothrops snake venoms from adult female snakes and their o spring. Toxicon 29, 219±226. Hoge, A. R. (1965) Preliminary account on neotropical Crotalinae (Serpentes Viperidae). Mem. Inst. Butantan 32, 109±184. Kamiguti, A. S. and Hanada, S. (1985) Study of the coagulant and proteolytic activities of newborn Bothrops jararaca venom. Toxicon 23, 580±587. Kamiguti, A. S., Silva, M. V. and Cardoso, J. L. C. (1986) Des®brinac° aÄo do sangue no envenenamnto acidental por serpentes Bothrops jararaca. Rev. Soc. Med. Trop. 28, 383±390. Kamiguti, A. S., Cardoso, J. L. C., Maruyama, M., Sano-Martins, I., Morena, P., Chudzinski, A. M., Tomy, S. C., Antonio, L. C., Santoro, M. L. and Kelen, E. M. A. (1988) Studies on blood coagulation and ®brinolysis in patients bitten by Bothrops jararaca snakes. Abstracts of the 5th Encontro Brasileiro de Hemostasia e Trombose, Florianopolis, p. 14. Kamiguti, A. S. and Cardoso, J. L. C. (1989) Haemostatic changes caused by the venoms of south american snakes. Toxicon 27, 955±963. Kocholati, W. F., Ledford, E. B., Daylu, J. G. and Billings, T. A. (1971) Toxicity and some enzymatic properties and activities in the venoms of Crotalinae, Elapidae and Viperidae. Toxicon 9, 31±138. Kouyoumdjian, J. A. and Polizelli, C. (1989) Acidentes ofõ dicos causados por Bothrops moojeni: correlac° aÄo do quadro clõ nico com o tamanho da serpente. Rev. Inst. Med. Trop. SaÄo Paulo 31, 84±90. Laemmli (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 256, 680±684. Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951) Protein measurement with the Folin phenol reagent. J. Biochem. 193, 165±275. Moura da Silva, A. M., DIÂmpeÂrio Lima, M. R., Nishikawa, A. K., Stephano, M. A., Santos, M. C., Dias da Silva, W. and Ueda, C. M. P. M. (1991) Antigenic cross reactivity of venoms obtained from snake of genus Bothrops. Toxicon 29, 783±786. Nahas, L., Denson, K. W. L. and MacFarlane, R. G. (1964) A study of the coagulant action of eight snake venoms. Thromb. Haemost. 12, 355±359. Nahas, L., Kamiguti, A. S. and Barros, M. A. R. (1979) Thrombin-like and factor X activator components of bothrops snake venoms. Thromb. Haemost. 41, 314±319. Pierini, S. V., Warrell, D. A., De Paulo, A. and Theakston, R. D. G. (1996) High incidence of bites and stings by snakes and other animals among rubber tappers and Amazonian Indians of the Jurua Valley, Acre state, Brazil. Toxicon 34, 225±236. Otero, R., Osorio, R. G., Valderrama, R. and Giraldo, C. A. (1992a) Efectos farmacoloÂgicos y enzimaÂticos de los venenos de serpientes de Antioquia y Choco (Colombia). Toxicon 30, 611±620. Otero, R., Tobon, G. S., Gomez, L. F., OsoÂrio, R. G., Valderrama, R., Hoyos, D., Urreta, J. E., Molina, S. and Arboleta, J. J. (1992b) Accident o®dico en Antioquia y Choco. Aspectos clõ nicos y epidemioloÂgicos (marzo de 1989±febrero de 1990). Acta Med. Colomb. 17, 229±249. Otero-Patino, R., Cardoso, J. L. C., Higashi, H. G., NunÄez, V., Diaz, M. F., Toro, M. F., Garcia, M. E., Sierra, A., Garcia, L. F., Moreno, A. M., Medina, M. C., CastanÄedas, N., Silva-Diaz, J. F., Murcia, M., CardenÄas, S. Y., Dias da Silva, W. and the Regional Group on Antivenom Therapy Research (REGATHER). (1997). A randomized blinded comparative trial of three antivenoms in Bothrops snake bites in Uraba, Colombia. Am. J. Trop. Med. Hyg., in press. Ribeiro, L. A. and Jorge, M. T. (1989) Alterac° aÄo do tempo de coagulac° aÄo sanguinea em pacientes picados por serpentes Bothrops jararaca adulta e ®lhote. Rev. Hosp. Clin. Med. Univ. SaÄo Paulo 44, 143±145. Rosenfeld, G., Hampe, O. G. and Kelen, E. M. A. (1959) Coagulant and ®brinolytica activity of animal venom: determination of coagulant and ®brinolytic index of di erent species. Mem. Inst. Butantan 29, 143± 163. SalomaÄo, M. G., WuÈster, W., Thorpe, R. S. and Touzet, J. M. (1997) DNA evolution of South American pitvipers of the genus Bothrops (Reptilia:Serpentes:Viperidae). Symp. Zool. Soc. London 70, 89±98. Sanchez, E. F., Freitas, T. V., FerreiraAlves, T. L., Velarde, D. T., Diniz, M. R., Cordeiro, M. N., AgostiniCotta, G. and Diniz, C. R. (1991) Biological activities of venoms from South American snakes. Toxicon 30, 95±103. World Health Organization (1981) Progress in the characterization of venoms and standardization of antivenoms. WHO O set Publ. 58, pp. 1±44. WuÈster, W., Thorpe, R. S. and Porto, G. (1996) Systematics of the Bothrops atrox complex (Reptilia:Serpentes:Viperidae) in Brazil: a multivariate analysis. Herpetologica 52, 263±271.