Amino acid sequence of a lectin-like protein from Lachesis muta stenophyrs venom

The primary structure of the lectin-like protein from Lachesis muta stenophyrs venom was deduced from analysis of the N-terminus and the sequence of peptides obtained after digestion with trypsin, Arg-C enzyme, Staphylococcus aureus V8 protease and endoproteinase Asp-N. Peptides generated by cleavage of the lectin with cyanogen bromide and o-iodosobenzoic acid were also sequenced. Comparison of the complete 135 amino acid residues sequence with those of the lectin from the venom of Crotalus atrox, with platelet coagglutinin from Bothrops jararaca beta-fragment and with the anticoagulant B protein chain from Trimeresurus flavoviridis venom, revealed 92, 46 and 29% identity, respectively. Significant homology was also found with C-type carbohydrate-recognition domain-like structures from invertebrate and vertebrate lectins. To our knowledge, this is the second known primary structure of a lectin-like protein from snake venom.     

Pharmacological aspects of mouse hind-paw oedema induced by Lachesis muta rhombeata venom

Pharmacological aspects of mouse hind-paw oedema induced by subplantar injections of Lachesis muta rhombeata (LMR) venom were investigated. The oedema induced by subplantar injections of 10 to 50 ng/g of LMR venom is dose dependent, with onset, peak and duration at 30, 60 and 180 min, respectively. Subplantar injection of 30 ng/g of Bothrops jararaca (BJ) venom induced oedema that has the same intensity as 30 ng/g of LMR venom but lasts for more than 4 h suggesting different time course. Systemic effects or haemorrhage were not observed with doses less than 50 ng/g. Oedema is not due to the presence of oedematogenic amines since dialysis did not change the oedema induced by 30 ng/g of LMR venom. Part of the oedema induced by LMR venom is due to a thermolabile fraction since pre-heating the venom at 100 degrees C for 15 min induced a significant reduction (56.19 +/- 6.8%) of the oedematogenic activity. The oedema induced by LMR venom is possibly induced by release of a pharmacological active substance at the site of injection. Histamine, arachidonate metabolites, nitric oxide and serotonin may play important roles in the oedematogenic effect of LMR venom since pre-treatment of mice with pyrilamine, indomethacin, dexamethasone, L-NAME and methysergide induced a significant reduction (49.86 +/- 10%; 51.06 +/- 5.9%; 77.66 +/- 3.6%; 73.30 +/- 6.1% and 93.77 +/- 2.8%, respectively) of the oedema formation. The present results demonstrate that the oedema induced by LMR and BJ venoms may be triggered and maintained by different pharmacological mechanisms. Since methysergide and L-NAME were the most active inhibitors of the oedema we can suggest that a link between serotonin release by the venom and a NO synthase activation may be an important step in the oedema formation induced by LMR venom.     

Purification, properties, and N-terminal amino acid sequence of a kallikrein-like enzyme from the venom of Lachesis muta rhombeata (Bushmaster)

An analytical procedure was developed to measure bromate residues in baked goods using a sequence of clean-up procedures followed by high performance liquid chromatography (HPLC) with a post-column reaction for oxidants. Deionized water was used to extract bromate from bread samples. The extract was treated with a C-18 solid phase extraction column to remove lipids, a cation exchange column with the silver cation to remove chloride, and an ultrafiltration membrane to remove proteins. Further treatment of the extract with the sodium form of a propylsulphonic acid ion exchange column was necessary to remove the silver that leached from the silver column. The method had a detection limit of 3 ng/g in baked goods. Recoveries of bromate from breads ranged from 73 to 86% at a fortified bromate level of 5-100 ng/g. Pullman-type white bread, produced by a sponge and dough method, was prepared in our laboratory for measurement of residual bromate. The dough was scaled in three different weights at different specific volumes (3.8, 4.1, 4.3), and samples of each of the three weights were baked for six different baking times ranging from 24 to 34 min. When bromate at a level of 25 mg/kg was added to flour, no residual bromate was detected in any of the samples, regardless of weight and baking time.     

Trimucytin: a collagen-like aggregating inducer isolated from Trimeresurus mucrosquamatus snake venom

OBJECTIVE: This study tested the hypothesis that gut stasis induced by parenteral morphine sulfate (MS) leads to enhanced bacterial translocation in rats on total parenteral nutrition (TPN). SUMMARY BACKGROUND DATA: TPN and MS are common adjuncts in the care of critically ill patients. TPN is known to provoke a variable degree of translocation. MS induces gut stasis with an accompanying bacterial overgrowth. The effect of these two treatments in combination on translocation is not known. METHODS: Rats were provided with central and subcutaneous lines for the continuous infusion of nutrients and drugs, respectively. Intestinal transit was assessed by the caudal movement of a fluorescent marker intubated into the proximal duodenum. Quantitative bacteriology was carried out from various segments of the gut and from ileocecal mesenteric lymph nodes (MLN), spleen, liver, and systemic blood obtained by cardia puncture on sacrifice at 96 hours. RESULTS: Transit was unchanged by TPN alone but prolonged when given in combination with MS. Bacterial overgrowth was also enhanced by MS and increased the bacterial translocation to MLN from 50% of animals with TPN, to 100% in those receiving both TPN and MS; the colony-forming units per MLN increased from 33 +/- 14 with TPN alone to 2079 +/- 811 (STD) with TPN plus MS. Furthermore, no bacteria were found at systemic sites with TPN alone, but in 93.3% of animals receiving TPN and MS. In a subgroup of rates provided with glutamine in TPN, the TPN plus MS effects on translocation were not reversed. CONCLUSIONS: These observations demonstrate the important role that morphine plays in promoting translocation, presumably by disrupting fasting motility and enhancing bacterial overgrowth.     

Thrombin-like enzyme from Lachesis muta muta venom: isolation and topographical analysis of its active site structure by means of the binding of amidines and guanidines as competitive inhibitors

A serine protease enzyme was purified from Lachesis muta muta venom, with 40% yield, by gel filtration on Sephadex G-100 and affinity chromatography on Sepharose-agmatin. Homogeneity of the enzyme preparation was demonstrated by sodium dodecyl sulfate polyacrylamide gel electrophoresis and the enzyme had a relative mol. wt of 45,000. The molar extinction coefficient at 280 nm was 62,127 (M x cm)-1. The enzyme hydrolysed Bz-Arg-Nan with Ks = 0.233 +/- 0.08 mM and kcat = 2.80 +/- 0.07 sec-1. All the amidines and guanidines tested for their inhibitory effect on thrombin-like enzyme behaved as competitive inhibitors of the enzyme with Ki values in the range 6.2 microM to 42.3 mM for amidines and 0.19 mM to 9.31 mM for guanidines. Dissociation constant values were analyzed in terms of the binding of the inhibitors with the subsite S1, the specificity pocket of the enzyme, Ki values were discussed in accordance with those for trypsin inhibition. beta-Naphthamidine was the strongest inhibitor, while guanidine was the weakest. The differences among the Ki values were interpreted in terms of the shape of the enzyme active site. For meta- and para-substituted benzamidinium ions a good correlation was found between log l/Ki and sigma Hammett values of the substituents. The substituent effects in the pi-electrons of the benzamidine ring were considered in the frame of Hückel molecular orbital theory. A model for the binding of p-benzamidine derivatives with the primary specificity S1 subsite of the enzyme active site was proposed.     

Comparative study of the venoms of three subspecies of Lachesis muta (bushmaster) from Brazil, Colombia and Costa Rica

A comparative study was performed on the pharmacology and biochemistry of venoms from three subspecies of Lachesis muta (L. m. stenophrys, L. m. muta and L. m. rhombeata) from Brazil, Colombia and Costa Rica. All venoms induced lethal, hemorrhagic, edema-forming, myotoxic, coagulant and defibrinating effects, showing also proteolytic and indirect hemolytic activities. The venoms of L. m. stenophrys from Costa Rica and L. m. muta from Cascalheira, Brazil, had the highest lethal and hemorrhagic activities and the venom of L. m. rhombeata showed the highest coagulant activity, whereas no significant differences were observed in myotoxic and edema-forming activities at most of the time intervals studied. In addition, venoms showed similar electrophoretic patterns on SDS-polyacrylamide gel electrophoresis. In conclusion, despite quantitative differences in toxic and enzymatic activities, together with subtle variations in electrophoretic patterns, our results indicate that experimental envenomation by these venoms induce a qualitatively similar pathophysiological profile.     

Proteins isolated from the venom of the common tiger snake (Notechis scutatus scutatus) promote hypotension and hemorrhage

Notechis scutatus scutatus venom contains several toxic acidic proteins called HTa-i which promote hypotension and hemorrhage in mice. They have apparent mol. wts in the 18,000-21,000 range, i.v. LD50 values between 0.5 and 1.5 micrograms/g, and no detectable phospholipase, arginine esterase, proteolytic or hemolytic activities. A polyclonal antibody raised against HTg binds to other purified proteins, suggesting that they are isoforms of the same protein. Many other elapid crude venoms contain proteins which recognize the polyclonal antibody raised against HTg. Crotalid and viperid crude venoms do not recognize this antibody, although some of their component proteins are known to exhibit hypotensive and hemorrhagic activities. A combination of gel-filtration on Sephacryl S-200, cation-exchange and anion-exchange chromatography allows isolation of the N. s. scutatus proteins in high purity. They are the first hypotension-inducing proteins to be purified from an Australian elapid.     

A study on the venom yield of venomous snake species from Argentina

A study on the venom yield of snakes from Argentina over a three year period was carried out on adult specimens of Bothrops alternatus (n = 74); Bothrops neuwiedii (n = 127); Bothrops ammodytoides (n = 30); Bothrops moojeni (n = 14); Bothrops jararaca (n = 14); B. jararacussu (n = 6); Crotalus durissus terrificus (n = 120) and Micrurus spp. (n = 6) as well as with 12 specimens of newborn C. d. terrificus kept in captivity. While for each species there was a positive correlation between venom yield and number of snakes milked, the correlation with the snake's body weights after individual milkings was even better, suggesting that the size of the snakes is more important in determining the venom yield than the number of snakes milked or the specimen's sex. Individual milkings indicated that, in addition to the snake size, when the amount of venom is normalized per 100 g body weight there is a species specific difference in venom yield. It follows the order B. jararacussu > B. moojeni approximately = B. jararaca approximately = B. alternatus > B. neuwiedii> Micrurus spp approximately = B. ammodytoides> C. d. terrificus. Although the venom yield per 100 g body weight of newborn C. d. terrificus specimens is 2-fold higher than that of adults, no correlation was observed between venom yield and body weight.     

Practical applications of snake venom toxins in haemostasis

Snake venom toxins have an established role in the coagulation laboratory for the assay of haemostatic parameters and a potential role for therapeutic treatment of thrombotic disorders. In the laboratory, snake venom thrombin-like enzymes (SVTLEs) are used for the assay of fibrinogen and detection of fibrinogen breakdown products and dysfibrinogenaemias. Importantly, because SVTLEs are not inhibited by heparin, they can be used for assaying antithrombin III and other parameters in samples which contain heparin. Prothrombin activators occur in many snake venoms and these have become established in the assay of prothrombin, in the study of dysprothrombinaemias and in the preparation of meizothrombin and non enzymic forms of prothrombin. Russell's viper (Daboia russelli) venom contains a number of useful compounds including toxins which can be used to assay blood clotting factors V, VII, X, platelet factor 3 and lupus anticoagulants (LA). More recently, activators from the taipan, Australian brown snake and saw-scaled viper have been used to assay LA. Proteins C and S can be measured by means of protac, a fast acting inhibitor from Southern copperhead snake venom and von Willebrand factor can be studied with botrocetin from Bothrops jararaca venom. The disintegrins, a large family of Arg-Gly-Asp (RGD)-containing proteins found in snake venoms, show great potential for the study of platelet glycoprotein receptors, notably, GPIIb/IIIa and Ib, and in the treatment of arterial thrombotic disease. Established SVTLEs used in clinical practice include ancrod and defibrase although success with these agents has been limited. A further group of enzymes under consideration as thrombolytic agents are the fibrinogenases.     

Inhibition of a snake venom hemorrhagic metalloproteinase by human and rat alpha-macroglobulins

Jararafibrase I is a hemorrhagic metalloproteinase purified from Bothrops jararaca venom, which induces local hemorrhage by degrading the basement membrane components. The present study was undertaken to investigate the inhibition of jararafibrase I by human and rat serum proteinase inhibitors. The proteolytic activity of jararafibrase I was completely inhibited by human and rat sera. In particular, rat serum displayed a greater inhibitory capacity. The inhibitory capacities of both sera were dependent on alpha-macroglobulins. SDS-PAGE analysis revealed that jararafibrase I formed complexes with alpha-macroglobulins that were present in normal sera. The proteolytic activity of jararafibrase I was completely inhibited by alpha1-macroglobulin and murinoglobulin in rat serum, and by human alpha2-macroglobulin. The inhibition molar ratios of alpha-macroglobulin/jararafibrase I were 1.5 for rat alpha1-macroglobulin and human alpha2-macroglobulin, and 2.4 for rat murinoglobulin. SDS-PAGE under reducing conditions demonstrated that the bait region of human alpha2-macroglobulin and rat murinoglobulin was cleaved by jararafibrase I. The bait region cleavage sites were identified as being situated at the 696Arg-697Leu peptide bond in human alpha2-macroglobulin, and at the 686Ala-687Val peptide bond in rat murinoglobulin.     

Characterization of a fibrinogen-clotting enzyme from Trimeresurus stejnegeri venom, and comparative study with other venom proteases

It has been well documented that drug-associated cues are important for the development and expression of drug tolerance. The Pavlovian conditioning analysis of tolerance emphasizes the importance of a drug-associated cues to tolerance by equating a drug administration with a learning trial. According to this analysis, tolerance should be subject to external inhibition, the disruption of a conditional response by a novel stimulus. We previously reported that tolerance to the ataxic effect of ethanol was attenuated by a novel strobe/noise presentation (31). In this article we report evidence of a compensatory CR in rats tolerant to the ataxic effect of ethanol as tested on the tilting plane. Both the compensatory CR and tolerance were disrupted by the presentation of a novel strobe/noise stimulus providing converging evidence that the attenuation of tolerance by a novel stimulus results from external inhibition of Pavlovian conditioning. The disruption of ethanol tolerance and the conditional response mediating tolerance was also apparent when the novel omission of the strobe/noise stimulus was used as the external inhibitor in rats made tolerant to ethanol with the stimulus on. Finally, we have shown that the disruptive effect of a novel stimulus on ethanol tolerance is decreased when there is a 10-day delay between the final tolerance development session and testing, demonstrating that the interval between training and testing is important when assessing associative tolerance.     

Use of snake venom fractions in the coagulation laboratory

The G1 cyclin, cyclin D1, has been implicated in the development of human and mouse tumors. Here we describe immunohistochemical analyses of cyclin D1 for a large panel of mouse B cell tumors. In addition, we characterize cyclin D1 expression in a series of cultured cell lines that represent transformed B cells at different stages of development. Immunohistochemical analysis showed that for low-grade lymphomas, cyclin D1 was expressed by 83% of centroblastic centrocytic (CBCC) and 14% of small lymphocytic lymphomas (SLL). For high-grade tumors, 28% of B lymphoblastic and 23% of centroblastic tumors expressed cyclin D1, while all immunoblastic lymphomas were negative. Studies of RNA and protein prepared from cultured B lineage tumors showed that cyclin D1 was expressed by all pre-B and most B cell tumors but not by cell lines representative of late B cell differentiation or by plasma cells. Expression of cyclin D1 in the lymphomas was not associated with alterations in the genomic structure of the Fis-1 (Bcl-1) common proviral integration site or cyclin D1 itself or with cell growth activity as assessed by expression of proliferating cell nuclear antigen (PCNA).     

Characterization of the major metalloprotease isolated from the venom of the northern pacific rattlesnake, Crotalus viridis oreganus

Rattlesnake venoms typically contain several different metalloproteases, some of which are hemorrhagic toxins. Metalloproteases contribute significantly to the often severe necrotic changes in tissues following envenomation, and these prominent components are important to the predigestive role of venoms. Venom of the northern Pacific rattlesnake (Crotalus viridis oreganus) contains at least five distinct metalloproteases, and the dominant protease (trivial name, CVO protease V) has been isolated and characterized as being a single polypeptide chain acidic protein with a molecular mass of 61 kDa and a pH optimum of approximately 9.0. It catalyzes the hydrolysis of several protein substrates, including casein, and is inhibited by metal chelators such as EDTA, EGTA and 1,10-phenanthroline but not by serine protease inhibitors such as PMSF. Calcium is present at a molar ratio of approximately 1:1, but, unlike other described venom metalloproteases, this protease does not appear to contain zinc. Caseinolytic activity is not significantly inhibited by citrate (at pH 9.0) at levels up to 2.0 mM; at 100 mM citrate (at pH 9.0) more than 65% of activity is retained. It is partially inhibited by nanomolar concentrations of ATP, but higher amounts (micromolar) do not result in further inhibition of activity. The protease shows fibrinolytic and fibrinogenolytic activity, but is only weakly hemorrhagic in rats. When stored in solution for long periods it undergoes autolytic degradation. This protease or a homolog appears to be present in venoms from several rattlesnake species but is not present in venoms from juvenile C.v. oreganus. The presence of this component in venoms from adult Pacific rattlesnakes is responsible for the age-related increase in metalloprotease activity of the crude venom.     

Coagulation factor X-binding protein from Deinagkistrodon acutus venom is a Gla domain-binding protein

Factor IX/factor X-binding protein (IX/X-bp) is an anticoagulant isolated from the venom of Trimeresurus flavoviridis (habu snake) and binds predominantly to factor IX. In this study, we isolated IX/X-bp-like proteins from the venom of Deinagkistrodon acutus (hundred pace snake) with binding characteristics different from those of IX/X-bp. The complete amino acid sequence and binding characteristics of the main anticoagulant protein, named X-bp, were investigated. The concentrations of X-bp at half-maximal binding to solid-phase factors X and IX were 0.4 and 3 nM, respectively. The binding of X-bp to solid-phase factor X was inhibited by 50% by 6- and 9-fold excess concentrations of factor X and Gla domain (GD) peptide 1-44, respectively, but was not influenced by GD peptide 1-41 and Gla domainless factor X. X-bp bound two Ca2+ ions per molecule with Kd values of 16 +/- 0.7 (mean +/- SE, n = 6) and 103 +/- 10 microM. X-bp was a heterodimer of C-type lectin-like subunits. The 16 kDa chain (A chain) consisted of 129 amino acid residues and was 68% identical to the sequence of the A chain of IX/X-bp. The 15 kDa chain (B chain) consisted of 123 amino acid residues and was 87% identical to IX/X-bp. Three-dimensional model construction from the known fold of IX/X-bp showed that amino acid residues different from those of IX/X-bp are mostly on the molecular surface. Some of these are concentrated on a part of the concave surface which is considered to be the coagulation factor-binding site, presumably acting as a discriminator for ligand binding. These results indicated that X-bp isolated from D. acutus venom was a GD-binding protein, and the C-terminal region of GD peptide was critical for folding of the peptide.     

Crystal structures of acutolysin A, a three-disulfide hemorrhagic zinc metalloproteinase from the snake venom of Agkistrodon acutus

Acutolysin A alias AaHI, a 22 kDa hemorrhagic toxin isolated from the snake venom of Agkistrodon acutus, is a member of the adamalysin subfamily of the metzincin family and is a snake venom zinc metalloproteinase possessing only one catalytic domain. Acutolysin A was found to have a high-activity and a low-activity under weakly alkaline and acidic conditions, respectively. With the adamalysin II structure as the initial trial-and-error model, the crystal structures were solved to the final crystallographic R-factors of 0. 168 and 0.171, against the diffraction data of crystals grown under pH 5.0 and pH 7.5 conditions to 1.9 A and 1.95 A resolution, respectively. One zinc ion, binding in the active-site, one structural calcium ion and some water molecules were localized in both of the structures. The catalytic zinc ion is coordinated in a tetrahedral manner with one catalytic water molecule anchoring to an intermediate glutamic acid residue (Glu143) and three imidazole Nepsilon2 atoms of His142, His146 and His152 in the highly conserved sequence H142E143XXH146XXGXXH152. There are two new disulfide bridges (Cys157-Cys181 and Cys159-Cys164) in acutolysin A in addition to the highly conserved disulfide bridge Cys117-Cys197. The calcium ion occurs on the molecular surface. The superposition showed that there was no significant conformational changes between the two structures except for a few slight changes of some flexible residue side-chains on the molecular surface, terminal residues and the active-site cleft. The average contact distance between the catalytic water molecule and oxygen atoms of the Glu143 carboxylate group in the weakly alkaline structure was also found to be closer than that in the weakly acidic structure. By comparing the available structural information of the members of the adamalysin subfamily, it seems that, when lowering the pH value, the polarization capability of the Glu143 carboxylate group to the catalytic water molecule become weaker, which might be the structural reason why the snake venom metalloproteinases are inactive or have a low activity under acidic conditions.     

Myotoxic phospholipases A from snake venom, Pseudechis colletti, producing myoglobinuria in mice

2 proteins producing myoglobinuria in mice were isolated from the venom of the Australian elapid snake Pseudechis colletti and identified as phospholipases A showing close similarities in amino acid composition to a similarly acting enzyme from a sea snake venom (Enhydrina schistosa).     

Characterization of an RNA-a protein complex isolated from the RNA bacteriophage M12

The properties of an RNA-A protein complex isolated from the RNA bacteriophage M12 are described. The molar ratio of RNA to A protein in the complex is estimated to be 1:1. In sucrose gradients, the complex sediments like free RNA molecules. In contrast to RNA alone, which can only infect spheroplasts, the RNA-A protein complex infects intact Escherichia coli cells and produces infectious progeny particles like the original phage. Evidence is presented that the infection of the host cells by the complex takes place via F pili. All of the infectivity disappears if the ionic bonds of RNA to A protein in the complex are dissociated in 0.5 M sodium chloride buffer at 37 degrees C. Furthermore, the kinetics of complex dissociation and loss of infectivity are the same, implying that the binding of A protein to the RNA is a prerequisite for infectivity on intact host cells.     

Characterization of a second lysine decarboxylase isolated from Escherichia coli

We report here on the existence of a new gene for lysine decarboxylase in Escherichia coli K-12. The hybridization experiments with a cadA probe at low stringency showed that the homologous region of cadA was located in lambda Kohara phage clone 6F5 at 4.7 min on the E. coli chromosome. We cloned the 5.0-kb HindIII fragment of this phage clone and sequenced the homologous region of cadA. This region contained a 2,139-nucleotide open reading frame encoding a 713-amino-acid protein with a calculated molecular weight of 80,589. Overexpression of the protein and determination of its N-terminal amino acid sequence defined the translational start site of this gene. The deduced amino acid sequence showed 69.4% identity to that of lysine decarboxylase encoded by cadA at 93.7 min on the E. coli chromosome. In addition, the level of lysine decarboxylase activity increased in strains carrying multiple copies of the gene. Therefore, the gene encoding this lysine decarboxylase was designated Idc. Analysis of the lysine decarboxylase activity of strains containing cadA, ldc, or cadA ldc mutations indicated that ldc was weakly expressed under various conditions but is a functional gene in E. coli.     

Isolation and partial characterization of a prothrombin-activating enzyme from the venom of the Australian rough-scaled snake (Tropidechis carinatus)

A prothrombin activator from the venom of Tropidechis carinatus has been isolated by means of gel filtration and benzamidine-based affinity chromatography, a novel use of the latter technique. Two bands possessing prothrombinase activity were obtained from the affinity chromatography procedure and designated A1 and A2. The bulk of the enzyme activity was recovered in peak A2 which represented 27-31% of the starting activity and a 14-16-fold purification. The venom contained, in total, around 5% by weight of the two isoforms of the prothrombin activator. The two fractions were electrophoretically similar on polyacrylamide electrophoresis, migrating with a mol. wt of 64,500 under native conditions and as a single band of 41,500 under reducing conditions. The prothrombinase was dependent on factor Va, phospholipid and calcium ions for its activity and is, thus, a member of the type II class of prothrombinases requiring such co-factors. The enzyme did not possess any phospholipase activity nor did it cleave the substrates N-alpha-benzoyl-L-arginine-p-nitroanilide (BAPNA), N-benzoyl-L-tyrosine ethyl ester (BTEE), azocollagen or azocasein, indicating a lack of amidolytic, esterolytic and broad-spectrum protease activity.     

Insights into the mechanism of haemorrhage caused by snake venom metalloproteinases

Local and systemic haemorrhage are common consequences of crotaline and viperine envenoming. Several studies carried out using purified toxins have indicated that local haemorrhage can be attributed to a distinct class of venom metalloproteinases. Analyses of their cDNAs predict multi-domain enzymes, with an N-terminal metalloproteinase domain, a disintegrin-like domain and a Cys-rich C-terminus. Haemorrhagic metalloproteinases are responsible for degrading proteins of the extracellular matrix and they also have cytotoxic effects on endothelial cells. However, to date very few investigations have been carried out on the effects of venom haemorrhagic metalloproteinases on components of the haemostatic system. We describe here the effects of a high molecular weight haemorrhagic metalloproteinase, jararhagin, from the venom of a South American pit viper Bothrops jararaca, on platelet and plasma components involved in haemostasis. Jararhagin, which is not inhibited in plasma, causes the loss of the platelet collagen receptor alpha 2 beta 1 integrin (gpIa/IIa or VLA-2) and degrades the adhesive plasma protein von Willebrand factor. Alterations of these haemostatic components are known to result in bleeding. This suggests that venom haemorrhagic metalloproteinases, in addition to causing local bleeding, may also contribute to systemic haemorrhage.