Toskon Vol. 29, No. 9, pp. 1097-I105, 1991 . Printed in dreat Brihio.
0041-0101/91 33 .00+ .00 ® 1991 Papmon Prer pk
SYSTEMIC HAEMORRHAGE IN RATS INDUCED BY A HAEMORRHAGIC FRACTION FROM BOTHROPS JARARACA VENOM A. S. ICAa~ucuTi,' R. D. G. TI~AxsroN,2 H. Dl?sMOrIDZ and R. A. Htrl-rorr' 'Department of Haematology, Royal Liverpool Hospital, Liverpool; ~ Liverpool School of Tropical Medicine, Liverpool, and'Haemophilia Centre, Royal Free Hospital, London, U .K . (Received 28 Jamrary 1991 ; accepted 1 May 1991)
A. S. ICAMICU~rc, R. D. G. T~AICSZ~oN, H. DE4MOND and R. A. HuTroN. Systemic haemorrhage in rats induced by a haemorrhagic fraction from Bothrops jararaca venom. Toxicon 29, 1097-1105.-The main haemorrhagc fraction of Bothrops jararaca venom, showing in vitro fibrinogenolytic activity and an inhibitory eßect on platelets aggregation induced by collagen, was studied in rats . Development of coagulopathy and/or haemorrhage was studied 2 hr after s.c . injection of batroxobin, B. jararaca whole venom and its haemorrhagic fraction. Incoagulable blood, together with low fibrinogen levels, were found only in rats injected with batroxobin and whole venom; thrombocytopenia alone was detected in rats given s.c . injections of haemorrhagin . Intravenous injection of low doses of haemorrhagin ( < 150 fig) resulted in significant thrombocytopenia, without any alterations in the blood coagulation mechanism. Severe damage to the vascular endothelium and skeletal muscle following s.c . injection of haemorrhagin together with signs of systemic haemorrhage in the kidneys, lungs and liver occurred . Levels of factor VIII and von Willebrand factor antigen were within the normal range in all animals. Serum levels of both whole venom and haemorrhagin were significantly correlated. This study confirms that B. jararaca haemorrhagin plays a vital role in systemic bleeding . INTRODUCTION
HUMAN envenoming by Bothrops jararaca in Brazil is frequently followed by severe coagulopathy associated with systemic bleeding . The venom of this species contains components capable of activating the blood coagulation mechanism (NAHAS et al., 1964, 1979); these are responsible for the development of coagulopethy. Other fractions, the haemorrhagins, are responsible for local haemorrhage in animals (MANDBLBAUM et al., 1976 ; ASSAKURA et al ., 1986). Haemorrhagins cause disruption of the collagenous basement membrane of the vascular endothelium causing leakage of blood components (OH3AKA, 1979). Several proteolytic properties have been attributed to haemorrhagins, such as collagenolytic and fibrinogenolytic activities ('I~r, 1988). However, despite several Address correspondence to : Dr Aura S . Kamiguti, University Department of Haematology, Dlincan Building, Royal Liverpool Hospital, Prescot Street, Liverpool L69 3BX, U .K. 1097
1098
A. S. KAMIGUTI et al.
studies on the pathogenesis of local haemorrhage produced by such components, very few investigations have been performed on the systemic effects of haemorrhagin . In order to elucidate the mode of action of the haemorrhagin, an investigation on the role of B. jararaca haemorrhagin and batrozobin, a thrombin-like enzyme supposedly not possessing haemorrhagic activity, on the haemostatic mechanism has been carried out using a rodent model. MATERIALS AND METHODS Haemorrhagic jroction of B. jararata venom Whole B. jararaca venom was dissolved in HPLC quality water and undissolved material was removed by brief centrifugation . The supernatant was then diluted 1 : 2 in 2 M ammonium sulphate and applied to a l ml phenyl superose hydrophobic interaction chromatography column on an FPLC system (Pharmacia Ltd, Uppsala, Sweden). Fractions were eluted with a linear gradient of ammonium sulphate in 50 pM Tris/HCI, pH 7.2 . The fraction containing more than 90% of venom haemorchagic activity was identified by the rat akin lesion method ~s:smN and Rte, 1983) and used for the studies. The fraction contained two major bands of similar mol. wt (58,000 and 56,000), the haemorchagin of 58,000 accounting for 70% of the total activity. The fraction obtained was tested for: (a) phoapholipase A activity using the indirect haemolytic assay of Bowewv and Ket,EtT~ (1957) and (b) coagulant activity on plasma and fibrinogen, using 200pl of humaa dusted plasma or 2 mg/ml bovine fibrinogen (Diagen Ltd., U.K .) plus 100 pl of the fractioa and recording the dotting time. Bovine thrombin (10 U/ml; Diagen Ltd, U.K.) was used, instead of the fraction, as control, (c) effect on platelet aggregation (Hoax and Cross, I%3), using 450 pl of human platelet-rich plasma (PRP), obtained by centrifuging a mixture of 9 parts blood to 1 part 3.8% triaodium citrate at 200 g for 10 min at 20-25°C. Fifty miaolitres of the fraction was added and the optical density of the mixture was monitored in a Payton dual channel aggregometer . As controls, 20 pM of adenosine diphosphate (ADP) or adrenaline (Sigma Ltd. U.K.), 12.5 mg/ml of ristocetin (Lundbeck Ltd, U.K.) and 20 pg/ml of equine tendon collagen (Hormon Chemie, Munich, Germany) were used instead of the fraction . (d) The effect on collagen was investigated using PRP: equal volumes of 40 pg,/ml of equine tendon collagen and saline or baemorchagic fraction were incubated at 37°C from 0-5 min; 50 Pl of the mixture was added to 450 Pl PRP and optical density changes were monitored in the aggregometer. (e) Fibrinogenolytic activity was estimated by incubating 200 ~l of fibrinogen or plasma with 50 lel of the fraction or of saline (control) in duplicate at 37°C for 5 min. One hundred microGtres of thrombin (100 U/ml) was added and the dotting time rnoorded. The clots thus formed were washed and the fibrinogen content estimated using the method of RerwoFr+ and MEtvz1E (1951) . In vivo experiments Male Spraguo-Dawley rats (220-340 g) were anaesthetised with an i.p. injection of a mixture of 0.5 ml/kg of Hypnorm (Jansaen Ltd., U.K.: 0.315 mg/ml sentanyl citrate; 10 mg/ml fluanisone) plus 3 mg/kg diazepam (C .P . Pharmaceuticals Ltd, Wrexham, U.K.). Groups of five animals were each injected s .c. into a shaved dorsal area with (a) whole B. jararaca venom (Instituto Batsman, Sao Paulo, Hrazil), (b) batroxobin (Pentapharm Ltd., Switzerland), (c) the major haemorchagic fraction isolated from B.jararaca venom and (d) saline (control). Two hours after administration, the anaethetiaed animals were killed by placing them in a CO= chamber for 2-3 min. After death animals were immediately bled by cardiac puncturs . Five other anaesthetised rata were bled in a similar manner to provide a pool of normal plasma . A further series of seven rats were injected i.v. into the tarsal vein with difftnent amounts (10 ltg, 25 pg, 75 ug, 150 Pg, 320 ltg, 650 Pg, 2000 pg) of haemorchagin. Blood coagulation
assays Whole blood dotting time was monitored by placing 2 ml of blood in a dean glass tube . This was left standing at room temperature for 20 min after which the presence/absence of a dot was recorded. Clot quality was also noted using the 1-5 grading system (Rte et al., 1%3) . Whole blood platelet counts were performed manually (Bxt:c~e and Caoxarre, 1950) using 1 ml of blood containing 10 pl of 10% potassium EDTA: other assays were carried out using plasma obtained after centrifugation at 2000 a for 10 min at 4°C of 3.6 ml blood mixed with 0.4 ml of 3.8% triaodium citrate. Duplicate fibrinogen assays were tarried out using the method of Rerrra~+ and Mt?rrzrE (1951) . Factor VIII coagulant activity was assessed by the one-stage assay (Dt?tvsox, 1976). The activated partial thromboplaatin time (APP'I') was determined using the ACL 300 ooagulometer (Instrumentation Laboratory, U.K .) . von Willebrand factor antigen (vWF-Ag) was estimated by enzyme-linked immunosorbent assay (ELISA), using 100i~l of plasma (1/50 and 1/100 dilutions) in a microtitre plate previously coated with 100 pl of 1/100 anti-human von Willebrand factor (vWF) serum (Dakopatts Ltd, U.K.) . Anti-human vWF IgG
Haemorrhagic Activity of Jararaca Venom
1099
peroxidase conjugate (100 itl, 1/200 dilution; Dakopatts Ltd, U .K .) was used to detect the reaction. A standard curve with normal pooled plasma was run in parallel .
Histologica! study Tissues from the site of injection (skin and underlying muscle), from the area opposite to the site, liver, kidney and lung, were removed and fixed for light microscopy in 10% buffered formol saline, and for electron microscopy in 3% buffered glutaraldehyde, following the standard procedure for killing the animals 2 hr after the original injection .
Assay of venom antigen and hatmorrhagin levels Whole venom assays were carried out on serum samples using ELISA following the method of Ti~AESPON et al. (1977) with the modifications described by Ho et al. (1986). Assays for venom haemorrhagin were carried out by using plates coated with the IgG fraction of B. jararaca and using an anti-B. jararaca haemorrhagin alkaline phosphatase conjugate (1/250 dilution) for detection . Quantification of B. jararaca venom and B . jararaca haemorrhagin levels were carried out by reference to a standard curve containing 500 ng/ml-0 .1 ng/ml of venom and haemorrhagin respectively .
RESULTS
The haemorrhagc fraction isolated from B. jararaca venom (accounting for about C~12% of the whole venom) produood haemorrhage when injected s.c. in vivo . The haemorrhagic fraction did not contain any phospholipase A activity, and it failed to clot plasma or fibrinogen (data not shown) ; neither did it possess any aggregating effects on PRP (Fig. 1). When haemorrhagin (1 mg/ml) was added to collagen, the ability of collagen to aggregate platelets was completely abolished even without prior incubation (Fig. 2). Even at a dilution of 1/4 in saline, the haemorrhagc fraction was still capable of delaying the effects of collagen on PRP, although at this dilution it did not prevent aggregation. When the haemorrhagic fraction was incubated with fibrinogen, the mixture was unclottable by thrombin due to complete destruction of fibrinogen (Table 1). Conversely, incubation of haemorrhagin with plasma did not affect the thrombin time and fibrinogen levels were normal (Table 1) possibly due to the various proteinase inhibitors in plasma which may be inhibiting the fibrinogenolytic activity of haemorrhagin . Haemorrhagin did not cause hydrolysis of fibrin clots after 24hr incubation indicating no fibrinolytic activity .
1~0. I . EFPEGT OF BOt1VOpJ jararOCa HAHIIORRHAOIN ON 1LAlE1L~r AOORHOATION. Light transmiaion changes were recorded in an aggregometer after addition of: (A) ADP; (H) adt+enalin; (G) collagen; (D) ristocetin; and (E) haemorrhagin to human platelet-rich plasma.
110 0
A. S. ICAMIGUTI et al.
FYO. 2.
F.FFecr OF
BOtIfIOpJ,%4laraCa
HAE1(OARHAßIN ON Tfd PLATEIEr-AOßR00ATnQ0 AC17vrrY OF COLLACfiN .
Collagen (40 pg/ml) was incubated with saline or baemorrhagin (1 mg/ml) in a 1 : 1 mixture for 0-5 min . Fifty microlitres of the mixture was added to 450 Id of platelet-rich plasma and changea of light transmission recorded. The final concentration of collagen was 2 pg/ml. Haemorrhagin+collagen: (A) o min.; (B) 1 min.; (C) 2 min.; (D) 5 min.; (>~ saline+collagen, 5 min .; and (F) haemorrhagirr (diluted 1 :4)+collagen.
Only rats injected with 750 hg of B. jararaca venom or 500 hg batroxobin s.c. (Groups C and D respectively) had unclottable blood (grade 5; Table 2). All rats presented with local haemorrhage except animals of group A and D. No deaths were recorded within 2 hr after administration of the venom or its fractions. The levels of B. jararaca venom detected in the serum of envenomed rats correlated in all groups with the amount injected s.c. (r = 0.852, 12 d.f., p = 0.001), except in Group D (batroxobin; Table 2). The small amount of haemorrhagin recordod in this group is probably due to contamination of the batroxobin preparation by haemorrhagin. Amounts of haemorrhagin injected, cornlated highly with the haemorrhagin detected both in the serum (r = 0.741, 13 d.f., p < 0.01) and with whole venom (r = 0.927, 13 d.f., p < 0.001). TAHI~
I . F»uen~oe~vol .rnc
Incubation mixture Fibrinogen+haemorrhagin Fibrinogen+aalirre Plasma+haemorrhagirr Plasma+saline
AC17vITY oF
B. faroraca
xAeecoxxxAalN
Clotting time (sec)
Fibrinogen (g/liter)
NC 15 .4-16.0 10.0-12.0 11 .0-12.0
0 4 .0-5 .5 2.5-2 .5 2.5-3 .5
NC : rro clot. After S min incubation of 20011 of plasma or fibrinogen with 5014 of haemorrhagin or saline, 10014 of thrombin was added and the clotting time was recorded. Fibrinogen was assayed in the incubation mixtlus . Two determinations are shown.
Haemonhagic Activity of Jararaca Venom TABLE
2.
1101
BLOOD COAGULATION DATA AND VENOM LEVELS IN RATS AFTER AD1QNIaTRATION OF B.,%araraCa VEN01(, BATAOXOBIN AND HAEMORAHAGRd
Clot Group Brade A
1
B
1
C
5
D
5
E
I
F
1
G
1
Platelets ( x 10'/1)
Fibrinogen (gJliter)
APPT (sec)
F VIII (UJdI)
vWF (U/dl)
977 (t45) 610+ (t49) 148+ (t25) 1001 (t69) 898 (f93) 711+ (t13) 493+ (t50)
3.67 (t0.32) 1.37+ (t0.21) 0 .23+ (t0.07) 0.47+ (t0.20) 2.93 (t0.19) 3.40 (t0.24) 3.25 (t0.37)
20.4 (t0 .6) -
72 (t 11) 64 (t6) 54 (t8) 83 (t14) 67 (t9) 62 (t13) 92 (t10)
116 (t6) 90 (t6) 82 (t8) I03 (t5) 93 (t9) 100 (t6) 132 (t9)
NC NC 19.5 (t0.4)
Venom (ng/ml) 0 81 (t 11) 874 (t167) 64 (t4) 760 (t151) 1770 (t98) 3260 (t220)
Haemonhagin (ng/ml) 0 5.6 (t 1 .4) 41 (fll) 5.6 (tl .l) 50 (t~ 74 (t2) 87 (t~
Group (A) 0.9+/+ NaCI ; (B) 100 ug B.Jararaca venom; (C) 7501+8 B. jarmaca venom; (D) 500 pg batroxobin ; (E) 320 Rg haemonhagin; (F) 650 Pg haemonhagin ; (G) 2000Ieg haemonhagin . Clot quality (grades 1-5) are recorded according to RED et al. (1963) . 'Significantly different from Group A. NC : no clot ; APTT: activated partial thromboplastin time; vWF: von Willebrand factor (antigen). Results are mean f S.E .M . (n = 5)
Amounts of purified B.jararaca venom haemorrhagin detected in the same sertun samples are also given in Table 2. A high degree of correlation was observed between the venom and haemorrhagin levels detected in the samples (r = 0.892, 35 d.f., p < 0.001). Platelet numbers were significantly decreased in rats given both venom and haemorrhagin (Groups B, C, F, G; Table 2). However, fibrinogen levels were decreased in rats injected with venom and batroxobin only (Table 2). Levels of factor VIII and vWF-Ag were within normal ranges for all groups (Table 2). ,coo soc
0
t 0
t000
l000
!IAll10RlWAOIN (YSl FIG . 3. THAOI~OCYTOFEPifA nvDVCED BY B. jararaca HAHIIORRHAGIN ua RATS. Different doses of baemorrhagin were administered i.v . and platelets were counted after 2 hr. Data from individual rata . Normal rat platelet count was 977145 x 109/liter (n ~ 5) . +Fibrinogen levels at these doses were < 0.23 gjliter (normal levels : 3.67 f0.32 g/liter; n = 5) .
1102
A . S . KAMIGUTI et aJ.
Intravenous injection of haemorrhagin induced a dose-related thrombocytopenia (Fig. 3), together with a decrease in fibrinogen levels . At doses between 75-150 pg, the haemorrhagin caused a fall in platelet numbers without causing a lowering of fibrinogen levels. Histologically, vascular endothelial damage was observed in all tissues examined from rats injected with both B. jararaca venom and haemorrhagin . Extravascular erythrocytes were also observed in animals injected with haemorrhagin and whole venom. When very high doses of haemorrhagin were administered, extensive muscle breakdown was observed (Fig. 4G) both at the site and, to a lesser degree, in the area of contralateral muscle . Endothelial cell changes were especially obvious in the kidney (Fig. 4D), the lung and in the muscle closest to the area of injection (Fig. 4C). DISCUSSION
Spontaneous bleeding at the site of the bite and/or systemically are symptoms frequently observed in Bothrops envenoming in man. Coagulopathy, due to the consumption of clotting factors, is another common feature of envenoming; this is associated with the presence of a thrombin-like enzyme and procoagulant components (activators of factor X and/or prothrombin), as well as a platelet-aggregating component (botrocetin) in Bothrops venom. Thrombin-like enzymes are frequently associated with Bothrops venoms (Ki osusrrzxY and Koxlc, 1936; STOCg>~t and B~.ow, 1975); these are the main cause of the coagulopathy observed in envenomed patients (RoB131~ .D, 1971 ; KAbucu~rl et al., 1986; MARUYAMA et al., 1990). Botrocetin, isolated from B. jararaca venom, induces platelet aggregation via the high molecular weight multimers of von Willebrand factor (REAn et al., 1978). These multimers are decreased in thrombocytopenic rats following administration of botrocetin ($ANDSIt3 et al., 1988). Haemorrhagins present in snake venoms are professes which are responsible for local haemorrhage and systemic blooding in envenomed patients (OxsAxw, 1979; WAltltxt .r . et al., 1977). Recently Ko1tNAi.>zc and VoRtovA (1990) suggested that i.v. injection of isolated B. riper haemorrhagin into rats does not damage the vessel wall; they concluded that the haemorrhagic diathesis in one bitten patient developed only during disseminated intravascular coagulation (DIC) . However, spontaneous systemic bleeding has been recorded in patients without coagulopathy following B. jararaca envenoming in man (K~ucU~rl et al., 1991). The present study proves that haemorrhagin induces systemic bleeding . Isolated B. jararaca haemorrhagin, without producing DIC in rats, has a definite systemic effect inducing a decrease in platelets numbers; this, therefore, results in the alteration in one of the factors that contributes to the haemostatic mechanism. The systemic effect of haemorrhagin on platelets or the induced thrombocytoponia may be one of the causes of bleeding . As haemorrhagin does not either aggregate platelets or possess any coagulant activity, thrombocytopenia, following s.c. inoculation of haemorrhagin, may be sequential to the primary damage of the vascular endothelium by haemorrhagin. A lesion of the vascular endothelium promotes the primary response of platelet adhesion mediated by fibrinogen and von Willebrand factor, followed by aggregation and consequent formation of a platelet plug in order to stop haemorrhage at the site of injury (Hvz-roN, 1989). In fact, several occluded damaged capillaries containing platelet aggregates were found following haemorrhagin administration. Similar findings occur after i.v. administration of
Haemonhagic Activity of Jararaca Venom
FiG. 4. ULTRASTRUCTURAL CIiANG173 IN MUSCLE AND KIDNEY 2 HR AFTER S .C. ADl~TISfRAT10N OF 2 MO PUAIFIID B. jAraraCa HAEMORRHAGIN IN RATS. A, normal muscle ; B, normal kidney ; C, muscle following s.c . injection of haemorrhagin; D, kidney following s .c. injection of haemorrhagin. e, erythrocyte; ec, vascular endothelial cell; m, muscle; rte, renal tubular epithelial cell ; c, collagen ; arrow, indicates damage to endothelial cell. Scale corresponds to 1 Em1.
1103
1104
A. S. KAMIGUTI et al.
B.jararaca haemonhagin, although at non-physiologically high i.v. doses it also induces hypofibrinogenaemia which is probably due to its fibrinogenolytic effects. Apart from incoagulable blood, batroxobin (thrombin-like enzyme) administration produced no other alterations, supporting the fact that coagulopathy due to hypofibrinogenaemia alone does not constitute a risk of spontaneous bleeding . The effect of hacmonhagin on the vascular endothelium is most probably related to the disruption of the basement membrane collagen of the vessel wall. In vitro, B. jararaca haemonhagin possessed powerful collagenolytic activity; it completely inhibited the effects of collagen on platelet aggregation. Similar effects wen observed with haemonhagin isolated from Vipers palaestinae venom, which possessed fibrinogenase activity and inhibited platelet aggregation (GRO~rro et al., 196 . Thrombocytopenia, induced by B. jararaca haemonhagin, is not related to von Willebrand factor (vWF), as normal levels were recorded in all groups (Table 2); this contrasts with the observation of high vWF levels in envenomed patients (KAMIGUTI et al., 1991). It is highly likely that the normal vWF levels recorded in the rat experiments were because the animals were purposely not subjected to stress due to the anaesthetic; obviously, bitten patients are naturally stressed both due to the effort of the bite (whether or not it be due to a bite by a venomous or non-venomous snake) and to the effect of being admitted to hospital . B. jararaca haemonhagin caused rapid disruption of capillary endothelium together with damage to skeletal muscle cells both systemically and locally. The local aspects were similar to the findings of OWNBY et al. (1990) who studied the pathogenesis of haemorrhage induced by bilitoxin from Agkistrodon bilineatus bilineatus venom in mice . This study confirms the importance of B. jararaca haemonhagin in inducing systemic bleeding in rats; clinical observations suggest that haemonhagin acts similarly in human victims. Adowwledgementr-We gatefully acraowkdge the expert technical assistance of Mise C. Rwnoctt and Miss P. Ptix, Animal Unit, and Mrs A. Bnoc[eruve, Histology Unit, Liverpool School of Tropical Medicine . Financial support was provided by the British Council and the European Economic Community.
REFERENCES A~ute~, M. T., Rmc~n., A. P. and 11~IANDBZ.BAt11I, F. R (1986) Comparison of immunological, biochemical and biophysical propertid of three haemorrhagic factors isolated from t1u venom of Bothrops fararmca (Jararaca) . Toxicat 24, 943-946. HoaN, G. V. R and Cstass, M. J. (1%3) The aggregation of blood platelets. J. Physiol. 168, 178-195. Boweurt, H. G. and Ketane, U. (1957) Chromatography of rattlesnake venom: a separation of three phosphodiestetases . Biochim. biophys. Acts 24, 61931 . Bnnnc,~a, G. and Caortrt~, E. P. (1950) . Morphology and enumeration of human blood platelets. J. Appl. Physiol. 3, 365-377. Dwsox, K. W. E. (1976) Appendix II. Techniques . In: Hrontu Blood Coagulation, Haemostasir and Thrombosis . (Btaas, R. Ed.) pp. 670-769. Oxford: Blad~well Scientific Publications. Gaorro, L., Mostoz, C., D$ Vams, A. and Got.nat .ux, N. (1%7) Isolation of Viptra palestinae hemorrhagin and distinction between ib hemorrhagic and proteolytic activities. Biochim. biophys. Acts 133, 356-362. Ho, M., W~~nn~ .* , M. J., WNta~.t D. A., Hmwttt.t D. and Vot.tlay A. (1986). A critical reappraisal of the use of enzymo-Gnked immunosorbent assays in the study of snake bite . Toxicon 24, 211-221 . HvnnN, R. A. (1989). Nonmal haemostasis. In : Ptutgradttate Hatmatology, (Hoa~+ea~uvn, A. V. and Lt:w~, S. M. Eds) pp . 560-597. Oxford: Heinemann. Ktiaatrn, A. S., MATS[JNA(iA, S., Srm, M., S~vo-MeartNS, I. S. and Nen~s, L. (1986) Alterations of the blood coagulation system after accidental inoculation by Bothrops fararaca venom. Brazilian J. med. Biol. Res. 19, 199-204.
Haemorrhagic Activity of Jararaca Venom
1103
Kesaavn, A. S., Ceanoso, J. L. C., Tl~ALSPON, R. D. G., Setvo-MeaTU~a, I. S., Hvn~ox, R. A., Ruosr~w, F. P., Weaxeu D. A. and Iiev, C. R. M. (1991) Coagulopathy and haemorrhage in human victims of Bothrops jarararn envenoming in Brazil . Toxlcon 29, %1-972 . Kwausrtz¢tr, D. and Korno, P. (1936) Hinchemische Studien Ober die Gitte der Schlangengattung Bothrops. III . Die Trennung der gerinnungsfordernden Substanz von Hothropotoxin and den Obrigen Sekrctbestandteilen . Naunyn-ScGmkdelxrgs Arcbs . Exp. Patlutl . Pbarnrakol. 181, 387-398. KORNALnC, F. and Vortt.ove, Z. (1990) Non-specific therapy of a haemorrhagic diathesis after a bite by a young Botbrops riper (Bathe armarilla) : a case report . Toxicon 28, 1497-1501 . MAxne~ .setnc, F. R., RHCxi., A. P. and Asseetrxe, M. T. (1976) Some physical and biochemical characteristics of HF2, one of the haemorrhagc factors in the venom of Bothrops jararaca. In: Anônal, Plant mid Microbial Toxins, (O~sexe, A., HeYw~n, K. and Sewer, Y. Eds) pp . 1(1-121. London: Plenum Press. McRVVeYe, M., Ke~atrrt, A. S., Cetenoso, J. L. C., Sexo-McArnvs, I. S., Catrn:nvscr, A. M., Setnnseo, M. L., Moaexe, P., Toanr, S. C., Axroiao, L. C., Mtxeae, H. and Kta.gv, E. M. A. (1990) Studies on blood coagulation and fibrinolysis in patients bitten by Bothrops jararaca (Jararaca). Throm . Hatmas . 63, 449-d53. Nettes, L., Dwsox, K. W. E and MecFAty exe, R G. (1964) A study of the coagulant activity of eight snake venoms. T)womb. Diath . Hanrarrlutg. 12, 335-367. Netus, L., Keaaotrrt, A. S. and Beams, M. A. R (1979) Thrombin-like and factor X-activator components of Bothrops snake venoma Tbromb. Harnas. 41, 314-328. Ot~es:w, A. (1979) Hemorrhagic, tucrotizing and edema-forming effects of snake venoms. In : Snake Venons, pp. 480-346 (Lee, C. Y. Ed.). Herlin : Springer. Owxav, C. L., Naet, T., Lxet, K. and $UütHAAA, H. (1990) Pathogenesis of haemorrhage induced by bilitoxin, a haemorrhagc toxin isolated from the venom of the common cantil . (Agkistrodon bilintatus blUneatusJ . Toxicon 28, 837-846. Relt~oarF, O. D. and Mexzn:, C. (1931) . A new method for the determination of fibrinogen in small samples of pLumut. J. Lab . Clôr. Med. 37, 316-320. Reeu, M. S., St>~t~, R. W. and BRII~ilS, K. M. (1978) Venom coagglutinin: an activator of platelet aggregation dependent on von Willebrand factor. Proc. Nato . Aced. Scl. U.S.A . 7S, 4514518. R®, H. A., Cxwrr K. E. and Tax, P. C. (1%3) Prolonged oosgulation defect (defibrination syndrome) in Malayan viper bite . Lancet 1, 62126. Ra®v~n, G. (1971) . Symptomatology, pathology and treatment of snakebites in South America. In: Venomous Animals and their Vennms, (Bum W. and BUC[IBY, E. E. Eds) pp. 345-384. N.Y.: Academic Press. SeHn®es, W. E., Reen, M. S., R~ntcs:, R L., Gewus, J. B. and Ban~ous, K. M. (1988) Thrombotic thrombocyfopenia with von Willebrand deficiency induced by botrocetin: an animal model. Lab. !newt. 99, 443-d32. Stogy K. F. and Beat.ow, G. H. (1973). Characterisation of Deflbrase. In : Drftbrinkrung nit thrombin8hnlichen Schlm4ge4glfttnzymtn. Acktuelk Problems in der Anglologie, (IVIAR7ßd, M. and Scxoor, W. Eds.) Vol. XXVI, pp . 452. Horn: Hens Huber. 7~eKSroN, R. D. G. and Rem, H. A. (1983) . Development of simple standard assay procedures for the characterization of snake venons . Bull. W.H.O. 61, 949956 . T~KSroN, R. D. G., Lt .oYn-Joxes, M. J. and Rem, H. A. (1977) . Micro-ELISA for detecting and assaying snake venom and venom antibody. Lancet 11, 639641 . Tu, A. T. (1988). Hemorrhagic professes from snake venoms : Review . In : Henwtasis and Animal Venons, (PtaY~Q H. and Mwau~t~n, F. S. Eds.) pp . 425-~43. New York: Dekker. WeateeiL, D. A., DAVIIx10N, N. McD., Gxewwoon, B. M., Ou~ou, L. D., Pore, H. M., WersJtrs, B. J. and PRENTiCE, C. R. M. (1977) . Poisoning by bites of the saw-scaled or carpet viper (Echis carinates) in Nigeria. Quart. J. Med. 46, 332.
0041-0101/91 33 .00+ .00 ® 1991 Papmon Prer pk
SYSTEMIC HAEMORRHAGE IN RATS INDUCED BY A HAEMORRHAGIC FRACTION FROM BOTHROPS JARARACA VENOM A. S. ICAa~ucuTi,' R. D. G. TI~AxsroN,2 H. Dl?sMOrIDZ and R. A. Htrl-rorr' 'Department of Haematology, Royal Liverpool Hospital, Liverpool; ~ Liverpool School of Tropical Medicine, Liverpool, and'Haemophilia Centre, Royal Free Hospital, London, U .K . (Received 28 Jamrary 1991 ; accepted 1 May 1991)
A. S. ICAMICU~rc, R. D. G. T~AICSZ~oN, H. DE4MOND and R. A. HuTroN. Systemic haemorrhage in rats induced by a haemorrhagic fraction from Bothrops jararaca venom. Toxicon 29, 1097-1105.-The main haemorrhagc fraction of Bothrops jararaca venom, showing in vitro fibrinogenolytic activity and an inhibitory eßect on platelets aggregation induced by collagen, was studied in rats . Development of coagulopathy and/or haemorrhage was studied 2 hr after s.c . injection of batroxobin, B. jararaca whole venom and its haemorrhagic fraction. Incoagulable blood, together with low fibrinogen levels, were found only in rats injected with batroxobin and whole venom; thrombocytopenia alone was detected in rats given s.c . injections of haemorrhagin . Intravenous injection of low doses of haemorrhagin ( < 150 fig) resulted in significant thrombocytopenia, without any alterations in the blood coagulation mechanism. Severe damage to the vascular endothelium and skeletal muscle following s.c . injection of haemorrhagin together with signs of systemic haemorrhage in the kidneys, lungs and liver occurred . Levels of factor VIII and von Willebrand factor antigen were within the normal range in all animals. Serum levels of both whole venom and haemorrhagin were significantly correlated. This study confirms that B. jararaca haemorrhagin plays a vital role in systemic bleeding . INTRODUCTION
HUMAN envenoming by Bothrops jararaca in Brazil is frequently followed by severe coagulopathy associated with systemic bleeding . The venom of this species contains components capable of activating the blood coagulation mechanism (NAHAS et al., 1964, 1979); these are responsible for the development of coagulopethy. Other fractions, the haemorrhagins, are responsible for local haemorrhage in animals (MANDBLBAUM et al., 1976 ; ASSAKURA et al ., 1986). Haemorrhagins cause disruption of the collagenous basement membrane of the vascular endothelium causing leakage of blood components (OH3AKA, 1979). Several proteolytic properties have been attributed to haemorrhagins, such as collagenolytic and fibrinogenolytic activities ('I~r, 1988). However, despite several Address correspondence to : Dr Aura S . Kamiguti, University Department of Haematology, Dlincan Building, Royal Liverpool Hospital, Prescot Street, Liverpool L69 3BX, U .K. 1097
1098
A. S. KAMIGUTI et al.
studies on the pathogenesis of local haemorrhage produced by such components, very few investigations have been performed on the systemic effects of haemorrhagin . In order to elucidate the mode of action of the haemorrhagin, an investigation on the role of B. jararaca haemorrhagin and batrozobin, a thrombin-like enzyme supposedly not possessing haemorrhagic activity, on the haemostatic mechanism has been carried out using a rodent model. MATERIALS AND METHODS Haemorrhagic jroction of B. jararata venom Whole B. jararaca venom was dissolved in HPLC quality water and undissolved material was removed by brief centrifugation . The supernatant was then diluted 1 : 2 in 2 M ammonium sulphate and applied to a l ml phenyl superose hydrophobic interaction chromatography column on an FPLC system (Pharmacia Ltd, Uppsala, Sweden). Fractions were eluted with a linear gradient of ammonium sulphate in 50 pM Tris/HCI, pH 7.2 . The fraction containing more than 90% of venom haemorchagic activity was identified by the rat akin lesion method ~s:smN and Rte, 1983) and used for the studies. The fraction contained two major bands of similar mol. wt (58,000 and 56,000), the haemorchagin of 58,000 accounting for 70% of the total activity. The fraction obtained was tested for: (a) phoapholipase A activity using the indirect haemolytic assay of Bowewv and Ket,EtT~ (1957) and (b) coagulant activity on plasma and fibrinogen, using 200pl of humaa dusted plasma or 2 mg/ml bovine fibrinogen (Diagen Ltd., U.K .) plus 100 pl of the fractioa and recording the dotting time. Bovine thrombin (10 U/ml; Diagen Ltd, U.K.) was used, instead of the fraction, as control, (c) effect on platelet aggregation (Hoax and Cross, I%3), using 450 pl of human platelet-rich plasma (PRP), obtained by centrifuging a mixture of 9 parts blood to 1 part 3.8% triaodium citrate at 200 g for 10 min at 20-25°C. Fifty miaolitres of the fraction was added and the optical density of the mixture was monitored in a Payton dual channel aggregometer . As controls, 20 pM of adenosine diphosphate (ADP) or adrenaline (Sigma Ltd. U.K.), 12.5 mg/ml of ristocetin (Lundbeck Ltd, U.K.) and 20 pg/ml of equine tendon collagen (Hormon Chemie, Munich, Germany) were used instead of the fraction . (d) The effect on collagen was investigated using PRP: equal volumes of 40 pg,/ml of equine tendon collagen and saline or baemorchagic fraction were incubated at 37°C from 0-5 min; 50 Pl of the mixture was added to 450 Pl PRP and optical density changes were monitored in the aggregometer. (e) Fibrinogenolytic activity was estimated by incubating 200 ~l of fibrinogen or plasma with 50 lel of the fraction or of saline (control) in duplicate at 37°C for 5 min. One hundred microGtres of thrombin (100 U/ml) was added and the dotting time rnoorded. The clots thus formed were washed and the fibrinogen content estimated using the method of RerwoFr+ and MEtvz1E (1951) . In vivo experiments Male Spraguo-Dawley rats (220-340 g) were anaesthetised with an i.p. injection of a mixture of 0.5 ml/kg of Hypnorm (Jansaen Ltd., U.K.: 0.315 mg/ml sentanyl citrate; 10 mg/ml fluanisone) plus 3 mg/kg diazepam (C .P . Pharmaceuticals Ltd, Wrexham, U.K.). Groups of five animals were each injected s .c. into a shaved dorsal area with (a) whole B. jararaca venom (Instituto Batsman, Sao Paulo, Hrazil), (b) batroxobin (Pentapharm Ltd., Switzerland), (c) the major haemorchagic fraction isolated from B.jararaca venom and (d) saline (control). Two hours after administration, the anaethetiaed animals were killed by placing them in a CO= chamber for 2-3 min. After death animals were immediately bled by cardiac puncturs . Five other anaesthetised rata were bled in a similar manner to provide a pool of normal plasma . A further series of seven rats were injected i.v. into the tarsal vein with difftnent amounts (10 ltg, 25 pg, 75 ug, 150 Pg, 320 ltg, 650 Pg, 2000 pg) of haemorchagin. Blood coagulation
assays Whole blood dotting time was monitored by placing 2 ml of blood in a dean glass tube . This was left standing at room temperature for 20 min after which the presence/absence of a dot was recorded. Clot quality was also noted using the 1-5 grading system (Rte et al., 1%3) . Whole blood platelet counts were performed manually (Bxt:c~e and Caoxarre, 1950) using 1 ml of blood containing 10 pl of 10% potassium EDTA: other assays were carried out using plasma obtained after centrifugation at 2000 a for 10 min at 4°C of 3.6 ml blood mixed with 0.4 ml of 3.8% triaodium citrate. Duplicate fibrinogen assays were tarried out using the method of Rerrra~+ and Mt?rrzrE (1951) . Factor VIII coagulant activity was assessed by the one-stage assay (Dt?tvsox, 1976). The activated partial thromboplaatin time (APP'I') was determined using the ACL 300 ooagulometer (Instrumentation Laboratory, U.K .) . von Willebrand factor antigen (vWF-Ag) was estimated by enzyme-linked immunosorbent assay (ELISA), using 100i~l of plasma (1/50 and 1/100 dilutions) in a microtitre plate previously coated with 100 pl of 1/100 anti-human von Willebrand factor (vWF) serum (Dakopatts Ltd, U.K.) . Anti-human vWF IgG
Haemorrhagic Activity of Jararaca Venom
1099
peroxidase conjugate (100 itl, 1/200 dilution; Dakopatts Ltd, U .K .) was used to detect the reaction. A standard curve with normal pooled plasma was run in parallel .
Histologica! study Tissues from the site of injection (skin and underlying muscle), from the area opposite to the site, liver, kidney and lung, were removed and fixed for light microscopy in 10% buffered formol saline, and for electron microscopy in 3% buffered glutaraldehyde, following the standard procedure for killing the animals 2 hr after the original injection .
Assay of venom antigen and hatmorrhagin levels Whole venom assays were carried out on serum samples using ELISA following the method of Ti~AESPON et al. (1977) with the modifications described by Ho et al. (1986). Assays for venom haemorrhagin were carried out by using plates coated with the IgG fraction of B. jararaca and using an anti-B. jararaca haemorrhagin alkaline phosphatase conjugate (1/250 dilution) for detection . Quantification of B. jararaca venom and B . jararaca haemorrhagin levels were carried out by reference to a standard curve containing 500 ng/ml-0 .1 ng/ml of venom and haemorrhagin respectively .
RESULTS
The haemorrhagc fraction isolated from B. jararaca venom (accounting for about C~12% of the whole venom) produood haemorrhage when injected s.c. in vivo . The haemorrhagic fraction did not contain any phospholipase A activity, and it failed to clot plasma or fibrinogen (data not shown) ; neither did it possess any aggregating effects on PRP (Fig. 1). When haemorrhagin (1 mg/ml) was added to collagen, the ability of collagen to aggregate platelets was completely abolished even without prior incubation (Fig. 2). Even at a dilution of 1/4 in saline, the haemorrhagc fraction was still capable of delaying the effects of collagen on PRP, although at this dilution it did not prevent aggregation. When the haemorrhagic fraction was incubated with fibrinogen, the mixture was unclottable by thrombin due to complete destruction of fibrinogen (Table 1). Conversely, incubation of haemorrhagin with plasma did not affect the thrombin time and fibrinogen levels were normal (Table 1) possibly due to the various proteinase inhibitors in plasma which may be inhibiting the fibrinogenolytic activity of haemorrhagin . Haemorrhagin did not cause hydrolysis of fibrin clots after 24hr incubation indicating no fibrinolytic activity .
1~0. I . EFPEGT OF BOt1VOpJ jararOCa HAHIIORRHAOIN ON 1LAlE1L~r AOORHOATION. Light transmiaion changes were recorded in an aggregometer after addition of: (A) ADP; (H) adt+enalin; (G) collagen; (D) ristocetin; and (E) haemorrhagin to human platelet-rich plasma.
110 0
A. S. ICAMIGUTI et al.
FYO. 2.
F.FFecr OF
BOtIfIOpJ,%4laraCa
HAE1(OARHAßIN ON Tfd PLATEIEr-AOßR00ATnQ0 AC17vrrY OF COLLACfiN .
Collagen (40 pg/ml) was incubated with saline or baemorrhagin (1 mg/ml) in a 1 : 1 mixture for 0-5 min . Fifty microlitres of the mixture was added to 450 Id of platelet-rich plasma and changea of light transmission recorded. The final concentration of collagen was 2 pg/ml. Haemorrhagin+collagen: (A) o min.; (B) 1 min.; (C) 2 min.; (D) 5 min.; (>~ saline+collagen, 5 min .; and (F) haemorrhagirr (diluted 1 :4)+collagen.
Only rats injected with 750 hg of B. jararaca venom or 500 hg batroxobin s.c. (Groups C and D respectively) had unclottable blood (grade 5; Table 2). All rats presented with local haemorrhage except animals of group A and D. No deaths were recorded within 2 hr after administration of the venom or its fractions. The levels of B. jararaca venom detected in the serum of envenomed rats correlated in all groups with the amount injected s.c. (r = 0.852, 12 d.f., p = 0.001), except in Group D (batroxobin; Table 2). The small amount of haemorrhagin recordod in this group is probably due to contamination of the batroxobin preparation by haemorrhagin. Amounts of haemorrhagin injected, cornlated highly with the haemorrhagin detected both in the serum (r = 0.741, 13 d.f., p < 0.01) and with whole venom (r = 0.927, 13 d.f., p < 0.001). TAHI~
I . F»uen~oe~vol .rnc
Incubation mixture Fibrinogen+haemorrhagin Fibrinogen+aalirre Plasma+haemorrhagirr Plasma+saline
AC17vITY oF
B. faroraca
xAeecoxxxAalN
Clotting time (sec)
Fibrinogen (g/liter)
NC 15 .4-16.0 10.0-12.0 11 .0-12.0
0 4 .0-5 .5 2.5-2 .5 2.5-3 .5
NC : rro clot. After S min incubation of 20011 of plasma or fibrinogen with 5014 of haemorrhagin or saline, 10014 of thrombin was added and the clotting time was recorded. Fibrinogen was assayed in the incubation mixtlus . Two determinations are shown.
Haemonhagic Activity of Jararaca Venom TABLE
2.
1101
BLOOD COAGULATION DATA AND VENOM LEVELS IN RATS AFTER AD1QNIaTRATION OF B.,%araraCa VEN01(, BATAOXOBIN AND HAEMORAHAGRd
Clot Group Brade A
1
B
1
C
5
D
5
E
I
F
1
G
1
Platelets ( x 10'/1)
Fibrinogen (gJliter)
APPT (sec)
F VIII (UJdI)
vWF (U/dl)
977 (t45) 610+ (t49) 148+ (t25) 1001 (t69) 898 (f93) 711+ (t13) 493+ (t50)
3.67 (t0.32) 1.37+ (t0.21) 0 .23+ (t0.07) 0.47+ (t0.20) 2.93 (t0.19) 3.40 (t0.24) 3.25 (t0.37)
20.4 (t0 .6) -
72 (t 11) 64 (t6) 54 (t8) 83 (t14) 67 (t9) 62 (t13) 92 (t10)
116 (t6) 90 (t6) 82 (t8) I03 (t5) 93 (t9) 100 (t6) 132 (t9)
NC NC 19.5 (t0.4)
Venom (ng/ml) 0 81 (t 11) 874 (t167) 64 (t4) 760 (t151) 1770 (t98) 3260 (t220)
Haemonhagin (ng/ml) 0 5.6 (t 1 .4) 41 (fll) 5.6 (tl .l) 50 (t~ 74 (t2) 87 (t~
Group (A) 0.9+/+ NaCI ; (B) 100 ug B.Jararaca venom; (C) 7501+8 B. jarmaca venom; (D) 500 pg batroxobin ; (E) 320 Rg haemonhagin; (F) 650 Pg haemonhagin ; (G) 2000Ieg haemonhagin . Clot quality (grades 1-5) are recorded according to RED et al. (1963) . 'Significantly different from Group A. NC : no clot ; APTT: activated partial thromboplastin time; vWF: von Willebrand factor (antigen). Results are mean f S.E .M . (n = 5)
Amounts of purified B.jararaca venom haemorrhagin detected in the same sertun samples are also given in Table 2. A high degree of correlation was observed between the venom and haemorrhagin levels detected in the samples (r = 0.892, 35 d.f., p < 0.001). Platelet numbers were significantly decreased in rats given both venom and haemorrhagin (Groups B, C, F, G; Table 2). However, fibrinogen levels were decreased in rats injected with venom and batroxobin only (Table 2). Levels of factor VIII and vWF-Ag were within normal ranges for all groups (Table 2). ,coo soc
0
t 0
t000
l000
!IAll10RlWAOIN (YSl FIG . 3. THAOI~OCYTOFEPifA nvDVCED BY B. jararaca HAHIIORRHAGIN ua RATS. Different doses of baemorrhagin were administered i.v . and platelets were counted after 2 hr. Data from individual rata . Normal rat platelet count was 977145 x 109/liter (n ~ 5) . +Fibrinogen levels at these doses were < 0.23 gjliter (normal levels : 3.67 f0.32 g/liter; n = 5) .
1102
A . S . KAMIGUTI et aJ.
Intravenous injection of haemorrhagin induced a dose-related thrombocytopenia (Fig. 3), together with a decrease in fibrinogen levels . At doses between 75-150 pg, the haemorrhagin caused a fall in platelet numbers without causing a lowering of fibrinogen levels. Histologically, vascular endothelial damage was observed in all tissues examined from rats injected with both B. jararaca venom and haemorrhagin . Extravascular erythrocytes were also observed in animals injected with haemorrhagin and whole venom. When very high doses of haemorrhagin were administered, extensive muscle breakdown was observed (Fig. 4G) both at the site and, to a lesser degree, in the area of contralateral muscle . Endothelial cell changes were especially obvious in the kidney (Fig. 4D), the lung and in the muscle closest to the area of injection (Fig. 4C). DISCUSSION
Spontaneous bleeding at the site of the bite and/or systemically are symptoms frequently observed in Bothrops envenoming in man. Coagulopathy, due to the consumption of clotting factors, is another common feature of envenoming; this is associated with the presence of a thrombin-like enzyme and procoagulant components (activators of factor X and/or prothrombin), as well as a platelet-aggregating component (botrocetin) in Bothrops venom. Thrombin-like enzymes are frequently associated with Bothrops venoms (Ki osusrrzxY and Koxlc, 1936; STOCg>~t and B~.ow, 1975); these are the main cause of the coagulopathy observed in envenomed patients (RoB131~ .D, 1971 ; KAbucu~rl et al., 1986; MARUYAMA et al., 1990). Botrocetin, isolated from B. jararaca venom, induces platelet aggregation via the high molecular weight multimers of von Willebrand factor (REAn et al., 1978). These multimers are decreased in thrombocytopenic rats following administration of botrocetin ($ANDSIt3 et al., 1988). Haemorrhagins present in snake venoms are professes which are responsible for local haemorrhage and systemic blooding in envenomed patients (OxsAxw, 1979; WAltltxt .r . et al., 1977). Recently Ko1tNAi.>zc and VoRtovA (1990) suggested that i.v. injection of isolated B. riper haemorrhagin into rats does not damage the vessel wall; they concluded that the haemorrhagic diathesis in one bitten patient developed only during disseminated intravascular coagulation (DIC) . However, spontaneous systemic bleeding has been recorded in patients without coagulopathy following B. jararaca envenoming in man (K~ucU~rl et al., 1991). The present study proves that haemorrhagin induces systemic bleeding . Isolated B. jararaca haemorrhagin, without producing DIC in rats, has a definite systemic effect inducing a decrease in platelets numbers; this, therefore, results in the alteration in one of the factors that contributes to the haemostatic mechanism. The systemic effect of haemorrhagin on platelets or the induced thrombocytoponia may be one of the causes of bleeding . As haemorrhagin does not either aggregate platelets or possess any coagulant activity, thrombocytopenia, following s.c. inoculation of haemorrhagin, may be sequential to the primary damage of the vascular endothelium by haemorrhagin. A lesion of the vascular endothelium promotes the primary response of platelet adhesion mediated by fibrinogen and von Willebrand factor, followed by aggregation and consequent formation of a platelet plug in order to stop haemorrhage at the site of injury (Hvz-roN, 1989). In fact, several occluded damaged capillaries containing platelet aggregates were found following haemorrhagin administration. Similar findings occur after i.v. administration of
Haemonhagic Activity of Jararaca Venom
FiG. 4. ULTRASTRUCTURAL CIiANG173 IN MUSCLE AND KIDNEY 2 HR AFTER S .C. ADl~TISfRAT10N OF 2 MO PUAIFIID B. jAraraCa HAEMORRHAGIN IN RATS. A, normal muscle ; B, normal kidney ; C, muscle following s.c . injection of haemorrhagin; D, kidney following s .c. injection of haemorrhagin. e, erythrocyte; ec, vascular endothelial cell; m, muscle; rte, renal tubular epithelial cell ; c, collagen ; arrow, indicates damage to endothelial cell. Scale corresponds to 1 Em1.
1103
1104
A. S. KAMIGUTI et al.
B.jararaca haemonhagin, although at non-physiologically high i.v. doses it also induces hypofibrinogenaemia which is probably due to its fibrinogenolytic effects. Apart from incoagulable blood, batroxobin (thrombin-like enzyme) administration produced no other alterations, supporting the fact that coagulopathy due to hypofibrinogenaemia alone does not constitute a risk of spontaneous bleeding . The effect of hacmonhagin on the vascular endothelium is most probably related to the disruption of the basement membrane collagen of the vessel wall. In vitro, B. jararaca haemonhagin possessed powerful collagenolytic activity; it completely inhibited the effects of collagen on platelet aggregation. Similar effects wen observed with haemonhagin isolated from Vipers palaestinae venom, which possessed fibrinogenase activity and inhibited platelet aggregation (GRO~rro et al., 196 . Thrombocytopenia, induced by B. jararaca haemonhagin, is not related to von Willebrand factor (vWF), as normal levels were recorded in all groups (Table 2); this contrasts with the observation of high vWF levels in envenomed patients (KAMIGUTI et al., 1991). It is highly likely that the normal vWF levels recorded in the rat experiments were because the animals were purposely not subjected to stress due to the anaesthetic; obviously, bitten patients are naturally stressed both due to the effort of the bite (whether or not it be due to a bite by a venomous or non-venomous snake) and to the effect of being admitted to hospital . B. jararaca haemonhagin caused rapid disruption of capillary endothelium together with damage to skeletal muscle cells both systemically and locally. The local aspects were similar to the findings of OWNBY et al. (1990) who studied the pathogenesis of haemorrhage induced by bilitoxin from Agkistrodon bilineatus bilineatus venom in mice . This study confirms the importance of B. jararaca haemonhagin in inducing systemic bleeding in rats; clinical observations suggest that haemonhagin acts similarly in human victims. Adowwledgementr-We gatefully acraowkdge the expert technical assistance of Mise C. Rwnoctt and Miss P. Ptix, Animal Unit, and Mrs A. Bnoc[eruve, Histology Unit, Liverpool School of Tropical Medicine . Financial support was provided by the British Council and the European Economic Community.
REFERENCES A~ute~, M. T., Rmc~n., A. P. and 11~IANDBZ.BAt11I, F. R (1986) Comparison of immunological, biochemical and biophysical propertid of three haemorrhagic factors isolated from t1u venom of Bothrops fararmca (Jararaca) . Toxicat 24, 943-946. HoaN, G. V. R and Cstass, M. J. (1%3) The aggregation of blood platelets. J. Physiol. 168, 178-195. Boweurt, H. G. and Ketane, U. (1957) Chromatography of rattlesnake venom: a separation of three phosphodiestetases . Biochim. biophys. Acts 24, 61931 . Bnnnc,~a, G. and Caortrt~, E. P. (1950) . Morphology and enumeration of human blood platelets. J. Appl. Physiol. 3, 365-377. Dwsox, K. W. E. (1976) Appendix II. Techniques . In: Hrontu Blood Coagulation, Haemostasir and Thrombosis . (Btaas, R. Ed.) pp. 670-769. Oxford: Blad~well Scientific Publications. Gaorro, L., Mostoz, C., D$ Vams, A. and Got.nat .ux, N. (1%7) Isolation of Viptra palestinae hemorrhagin and distinction between ib hemorrhagic and proteolytic activities. Biochim. biophys. Acts 133, 356-362. Ho, M., W~~nn~ .* , M. J., WNta~.t D. A., Hmwttt.t D. and Vot.tlay A. (1986). A critical reappraisal of the use of enzymo-Gnked immunosorbent assays in the study of snake bite . Toxicon 24, 211-221 . HvnnN, R. A. (1989). Nonmal haemostasis. In : Ptutgradttate Hatmatology, (Hoa~+ea~uvn, A. V. and Lt:w~, S. M. Eds) pp . 560-597. Oxford: Heinemann. Ktiaatrn, A. S., MATS[JNA(iA, S., Srm, M., S~vo-MeartNS, I. S. and Nen~s, L. (1986) Alterations of the blood coagulation system after accidental inoculation by Bothrops fararaca venom. Brazilian J. med. Biol. Res. 19, 199-204.
Haemorrhagic Activity of Jararaca Venom
1103
Kesaavn, A. S., Ceanoso, J. L. C., Tl~ALSPON, R. D. G., Setvo-MeaTU~a, I. S., Hvn~ox, R. A., Ruosr~w, F. P., Weaxeu D. A. and Iiev, C. R. M. (1991) Coagulopathy and haemorrhage in human victims of Bothrops jarararn envenoming in Brazil . Toxlcon 29, %1-972 . Kwausrtz¢tr, D. and Korno, P. (1936) Hinchemische Studien Ober die Gitte der Schlangengattung Bothrops. III . Die Trennung der gerinnungsfordernden Substanz von Hothropotoxin and den Obrigen Sekrctbestandteilen . Naunyn-ScGmkdelxrgs Arcbs . Exp. Patlutl . Pbarnrakol. 181, 387-398. KORNALnC, F. and Vortt.ove, Z. (1990) Non-specific therapy of a haemorrhagic diathesis after a bite by a young Botbrops riper (Bathe armarilla) : a case report . Toxicon 28, 1497-1501 . MAxne~ .setnc, F. R., RHCxi., A. P. and Asseetrxe, M. T. (1976) Some physical and biochemical characteristics of HF2, one of the haemorrhagc factors in the venom of Bothrops jararaca. In: Anônal, Plant mid Microbial Toxins, (O~sexe, A., HeYw~n, K. and Sewer, Y. Eds) pp . 1(1-121. London: Plenum Press. McRVVeYe, M., Ke~atrrt, A. S., Cetenoso, J. L. C., Sexo-McArnvs, I. S., Catrn:nvscr, A. M., Setnnseo, M. L., Moaexe, P., Toanr, S. C., Axroiao, L. C., Mtxeae, H. and Kta.gv, E. M. A. (1990) Studies on blood coagulation and fibrinolysis in patients bitten by Bothrops jararaca (Jararaca). Throm . Hatmas . 63, 449-d53. Nettes, L., Dwsox, K. W. E and MecFAty exe, R G. (1964) A study of the coagulant activity of eight snake venoms. T)womb. Diath . Hanrarrlutg. 12, 335-367. Netus, L., Keaaotrrt, A. S. and Beams, M. A. R (1979) Thrombin-like and factor X-activator components of Bothrops snake venoma Tbromb. Harnas. 41, 314-328. Ot~es:w, A. (1979) Hemorrhagic, tucrotizing and edema-forming effects of snake venoms. In : Snake Venons, pp. 480-346 (Lee, C. Y. Ed.). Herlin : Springer. Owxav, C. L., Naet, T., Lxet, K. and $UütHAAA, H. (1990) Pathogenesis of haemorrhage induced by bilitoxin, a haemorrhagc toxin isolated from the venom of the common cantil . (Agkistrodon bilintatus blUneatusJ . Toxicon 28, 837-846. Relt~oarF, O. D. and Mexzn:, C. (1931) . A new method for the determination of fibrinogen in small samples of pLumut. J. Lab . Clôr. Med. 37, 316-320. Reeu, M. S., St>~t~, R. W. and BRII~ilS, K. M. (1978) Venom coagglutinin: an activator of platelet aggregation dependent on von Willebrand factor. Proc. Nato . Aced. Scl. U.S.A . 7S, 4514518. R®, H. A., Cxwrr K. E. and Tax, P. C. (1%3) Prolonged oosgulation defect (defibrination syndrome) in Malayan viper bite . Lancet 1, 62126. Ra®v~n, G. (1971) . Symptomatology, pathology and treatment of snakebites in South America. In: Venomous Animals and their Vennms, (Bum W. and BUC[IBY, E. E. Eds) pp. 345-384. N.Y.: Academic Press. SeHn®es, W. E., Reen, M. S., R~ntcs:, R L., Gewus, J. B. and Ban~ous, K. M. (1988) Thrombotic thrombocyfopenia with von Willebrand deficiency induced by botrocetin: an animal model. Lab. !newt. 99, 443-d32. Stogy K. F. and Beat.ow, G. H. (1973). Characterisation of Deflbrase. In : Drftbrinkrung nit thrombin8hnlichen Schlm4ge4glfttnzymtn. Acktuelk Problems in der Anglologie, (IVIAR7ßd, M. and Scxoor, W. Eds.) Vol. XXVI, pp . 452. Horn: Hens Huber. 7~eKSroN, R. D. G. and Rem, H. A. (1983) . Development of simple standard assay procedures for the characterization of snake venons . Bull. W.H.O. 61, 949956 . T~KSroN, R. D. G., Lt .oYn-Joxes, M. J. and Rem, H. A. (1977) . Micro-ELISA for detecting and assaying snake venom and venom antibody. Lancet 11, 639641 . Tu, A. T. (1988). Hemorrhagic professes from snake venoms : Review . In : Henwtasis and Animal Venons, (PtaY~Q H. and Mwau~t~n, F. S. Eds.) pp . 425-~43. New York: Dekker. WeateeiL, D. A., DAVIIx10N, N. McD., Gxewwoon, B. M., Ou~ou, L. D., Pore, H. M., WersJtrs, B. J. and PRENTiCE, C. R. M. (1977) . Poisoning by bites of the saw-scaled or carpet viper (Echis carinates) in Nigeria. Quart. J. Med. 46, 332.
