Structure and function of snake venom toxins interacting with human von Willebrand factor

Toxicon 45 (2005) 1075–1087 www.elsevier.com/locate/toxicon

Structure and function of snake venom toxins interacting with human von Willebrand factor Taei Matsuia,*, Jiharu Hamakob a

Department of Biology, Fujita Health University School of Health Sciences, Fujita Health University, Toyoake, Aichi 470-1192, Japan b Medical Information Systems, Fujita Health University College, Toyoake, Aichi 470-1192, Japan Accepted 6 October 2004 Available online 22 April 2005

Abstract Hemostatic plug formation is a complex event mediated by platelets, subendothelial matrices and von Willebrand factor (VWF) at the vascular injury. Snake venom proteins have an excellent potency to regulate the interaction between VWF and platelet membrane receptors in vitro. Two protein families, C-type lectin-like proteins and Zn2C-metalloproteinases, have been found to affect platelet–VWF interaction. Botrocetin and bitiscetin from viper venom are disulfide-linked heterodimers with C-type lectin-like motif, and modulate VWF to elicit platelet glycoprotein Ib (GPIb)-binding activity via the A1 domain of VWF leading to the platelet agglutination. The crystal structures of botrocetin and bitiscetin together with complex from the VWF A1 domain indicate the following: (1) a central concave domain formed by two subunits of botrocetin or bitiscetin provides the binding site for VWF, (2) these modulators directly bind to the A1 domain of VWF in close proximity to the GPIb binding site, (3) both modulators induce no significant conformational change on the GPIb-binding site of the A1 domain but could provide a supplemental platform fitting for GPIb. These results suggest that the modulating mechanisms of these venoms are different from those performed by either antibiotic ristocetin in vitro or extremely high shear stress in vivo. Other modulator toxins include kaouthiagin and jararhagin, chimeric proteins composed of metalloproteinase, disintegrinlike and Cys-rich domains. These toxins cleave VWF and reduce its platelet agglutinating or collagen-binding activity. Kaouthiagin from cobra venom specifically cleaves between Pro708 and Asp709 in the C-terminal VWF A1 domain resulting in the decrease of the multimer structure of VWF. Recently a plasma proteinase, which specifically cleaves VWF into a smaller multimer, has been elucidated to be a reprolysin-like metalloproteinase with thrombospondin motif family (ADAMTS). This endogenous metalloproteinase (ADAMTS-13) specifically cleaves between Tyr842 and Met843 in the A2 domain of VWF regulating its physiological hemostatic activity. These VWF-binding snake venom proteins are suitable probes for basic research on platelet plug formation mediated by VWF, for subsidiary diagnostic use for von Willebrand disease or platelet disorder, and might be potently applicable to the regulation of VWF in thrombosis and hemostasis. Structural information of these venom proteins together with recombinant technology might strongly promote the construction of a new antihemostatic drug in the near future. q 2005 Elsevier Ltd. All rights reserved. Keywords: Botrocetin; Bitiscetin; von Willebrand factor; Kaouthiagin; ADAMTS-13; Platelet plug; GPIb

Abbreviations FVIII, coagulation factor VIII; FIX/X-BP, coagulation factors IX and X binding protein; GP, platelet glycoprotein; mAb, monoclonal antibody; SDS-PAGE, sodium dodecyl sulphate polyacrylamide gel electrophoresis; TTP, thrombotic thrombocytopenic purpura; VWD, von Willebrand disease; VWF, von Willebrand factor. * Corresponding author. Tel.: C81 562 93 2954; fax: C81 562 93 4595. E-mail address: [email protected] (T. Matsui). 0041-0101/$ - see front matter q 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.toxicon.2005.02.023

1076

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

1. Introduction Platelet plug formation is the first step for hemostasis at the vascular injury site followed by ignition of the coagulation cascade. Plasma von Willebrand factor (VWF) and its platelet membrane receptors, glycoproteins (GP) Ib and IIb/IIIa, are responsible for this first event (Sadler, 1998; Ruggeri, 2003). VWF circulates without interacting with platelets in a normal condition, but once subendothelial matrices such as collagens are exposed at the site of vascular damage, VWF immediately sticks to them under high shear stress and elicits an affinity to circulating platelets. It is not clear what makes VWF responsive to GPIb once absorbed into the subendothelial matrices. Conformational change of VWF molecules has been proposed as a candidate based on atomic force microscopic observation in that the immobilized VWF becomes an extended form upon exposure to the shear stress (Siedlecki et al., 1996), but conflicting data have been reported recently (Novak et al., 2002). Platelets start rolling on the immobilized VWF via its surface GPIb molecules like leukocyte does on inflamed endothelium (Somers et al., 2000), leading to the activation of cryptic GPIIb/IIIa molecules followed by the firm adhesion of platelets using GPIIb/IIIa (Savage et al., 1998). GPIIb/ IIIa is an integrin (aIIbb3) recognizing the Arg-Asp-Gly (RGD) sequence in fibrinogen or VWF, and those binding induce a platelet–platelet aggregation. In the nonphysiological conditions, platelet aggregation (or agglutination) is inducible by modulating VWF with a cofactor. Two major VWF-modulating venoms (coagglutinins), botrocetin and bitiscetin, have been extensively studied in detail. It has been proposed that these venom proteins might induce a conformational change on the GPIbbinding site of VWF to elicit a high affinity form mimicking an event which will occur in vivo, but the crystal analysis of the complex showed no significant conformational change in the GPIb-binding site on the VWF, evoking a new modulation mechanism for these venom modulators (Maita et al., 2003). In this review, we focus on VWF-activating modulators and briefly on VWF-cleaving proteinases purified from snake venoms. Other snake venom proteins affecting hemostasis and thrombosis by regulating the platelet function will be described elsewhere in this issue.

2. Structure and function of VWF VWF has two major physiological roles against bleeding. One is as a carrier for coagulation factor VIII (FVIII), and the other is as molecular glue for hemostatic plug formation. Deficiency or structural abnormality of VWF leads to a congenital bleeding disorder referred to as von Willebrand disease (VWD) (Ginsburg, 1999). VWF circulates in plasma at a concentration of 5–15 mg/ml as a series of multimers

comprised of a 270 kDa subunit. The mature subunit of human VWF consists of 2050 amino acids coded from 28 exons on chromosome 12p, and w15% saccharides with 12 N-linked and 10 O-linked sugar chains (Fujimura and Titani, 1993). It is quite unique that human VWF contains ABO blood group antigens in its N-linked sugar chains (Matsui et al., 1992). The VWF subunit shows the internal homology designated as A to D domains (D 0 -D3-A1-A2-A3-D4-B1-B2-B3-C1-C2) with distinct functions (Fig. 1). VWF binds FVIII and protects it from proteolytic degradation during plasma circulation. This FVIII-binding site of VWF is localized at the N-terminal D domains. The collagen binding site is separately present at the A1 (residues1 497–716) and A3 (residues 910–1111) domains. The A1 domain also has distinct binding sites for GPIb, sulfatide, antibiotic ristocetin and snake toxins. The crystal structures of the A1 and A3 domains have been elucidated (Huizinga, 1997; Celikel et al., 1998; Emsley et al., 1998). Both domains show a cuboid structure consisting of central six b strands flanked by six a helices and a single disulfide bond (Fig. 2). Interaction of the VWF A1 domain and platelet GPIb has recently been elucidated by crystal structural analysis of the complex (Huizinga et al., 2002). The N-terminal domain of GPIb (residues 1–290) shows a concave structure composed of a b hairpin (residues 1–14, b finger), eight Leu-rich repeats and a flanking region (residues 227–241, b-switch). The b-finger and b-switch interact with the bottom and side faces of the A1 domain, respectively, wrapping around one side of the A1. The small turn between b2 and b3 strands (residues 559–561) of the A1 has been assumed to be an interface for the GPIb (Celikel et al., 1998). The crystal structure shows that Thr240 in the b-switch interacts with Asp560-Gly561 by hydrogen bonding, suggesting that subtle conformational change around the Asp560-Gly561 might influence the GPIb-binding activity. VWF subunits are disulfide-linked in a head-to-head and tail-to-tail manner and synthesized by endothelial cells and megakaryocytes as a large multimer. The endothelial cells constitutively secrete VWF into plasma as ultra-large multimers of approximately 30,000 kDa. The hemostatic activity of VWF is affected by the multimer size: the highly multimeric structure is highly hemostatic due to its polyvalency. To regulate the hemostatic activity of VWF, it is specifically cleaved within the A2 domain (residues 717–909) by the plasma VWF-cleaving metalloprotease, producing a series of multimers ranging from 500 to 20,000 kDa (Furlan et al., 1998) (Fig. 3). This gradual degradation of VWF is physiologically very important since dysfunction or deficiency of the metalloproteinase causes thrombotic disorder such as thrombotic thrombocytopenic purpura (TTP) (Furlan et al., 1998; Tsai and Lian, 1998; Fujimura et al., 2002; Matsumoto et al., 2003). 1 The numbering of VWF residues in this review start at the first residue of the mature subunit.

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

1077

Fig. 1. Schematic structure of human VWF subunit. VWF subunit is comprised of 2050 amino acid residues, 10 O-linked and 12 N-linked sugar chains, and shows an internal homology designated as A–D domains. VWF is rich in disulfide bond especially at the N- and C-terminal regions, but the A1 and the A3 domains have 1–2 bond. Binding sites for GPIb, botrocetin and bitiscetin are localized in the A1 domain. Arrows indicate cleavage sites for ADAMTS-13 and kaouthiagin.

3. VWF-modulating venoms 3.1. Snake venom coagglutinin VWF never interacts with platelets under static conditions, however, the antibiotic ristocetin and certain snake venoms (botrocetin and bitiscetin) are able to induce platelet agglutination artificially in vitro under static conditions. Ristocetin is a glycopeptide (with a molecular mass of about 2260 Da) synthesized by actinomycete Nocardia lurida and agglutinates both formalin-fixed and fresh platelets in the presence of VWF between the effective concentrations of 0.5–1.5 mg/ml (0.22–0.66 mM). This function of ristocetin is quite convenient for clinical subsidiary diagnosis of hemostatic disorder, but ristocetin is not responsive to some species such as canine plasma or platelets. Further, high concentration of ristocetin (!2 mg/ml) tends to precipitate plasma proteins such as fibrinogen by flocculating the molecules. To overcome this limitation, Read et al. (1978) screened snake venoms that induced the VWF-mediated platelet agglutination and found that the crude extracts of five species (Bothrops alternatus, B. jararaca, B. medusa, B. neuwiedii, and Bitis arietans) out of 73 venoms have this activity. The venom coagglutinin from rattlesnake B. jararaca referred to as botrocetin shows no coagulant or enzymatic activity, but it clearly agglutinates fixed or fresh platelets in the presence of VWF irrespective of mammal species. The effective concentration of botrocetin inducing VWF-dependent platelet agglutination is approximately 2–5 mg/ml (0.07–0.19 mM). Several groups have reported the purification of botrocetin (Andrews et al., 1989a; Fujimura et al., 1991a), and the complete amino acid sequence of botrocetin (two-chain botrocetin) has been determined (Usami et al., 1993) (Fig. 4). Monomeric botrocetin (one-chain botrocetin)

has been previously reported to have a low VWFmodulating activity, but later it was found to be identical to a C-terminal half of jararhagin (jararhagin C) and has no VWF-activating activity (Usami et al., 1994).

Fig. 2. Structure of the VWF A1 domain. Structure of the A1 domain is expressed as ribbon model. The VWF A1 domain has central six b strands (b1–b6) flanked by six a helices (a1–a6). Important residues for botrocetin binding (Arg629, Arg632, Arg636 and Lys667) and for bitiscetin binding (Arg632, Lys660, Gln666 and Lys673) are expressed as ball and stick model. The binding site for GPIb is located at the left side of the figure comprising of a turn between b2 and b3 strands, b3 strand and a3 helix. This figure was generated with MOLSCRIPT (Esnouf, 1999) referring to the Protein Data Bank entry 1OAK.

1078

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

Fig. 3. Cleavage of VWF by ADAMTS-13 and kaouthiagin. VWF is synthesized as ultra-large multimers disulfide-linked in a head-to-head and tail-to-tail manner. Multimeric VWF is cleaved at its A2 domains by plasma ADAMTS-13 under high shear stress in vivo, or by kaouthiagin in vitro into smaller multimers. Hemostatic activities of VWF such as platelet agglutinating or collagen binding potencies depend on the multimer size of VWF. N, N-terminal; C, C-terminal.

Botrocetin is a disulfide-linked heterodimer (approximately 25–27 kDa) comprised of similar a- (133 residues) and b-subunits (125 residues), each of which contains C-type lectin-like motif, although botrocetin shows neither sugar-binding nor Ca2C-dependent activities. Both subunits migrate at a close proximity on sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) under reduced conditions as a single band (approximately 14–15 kDa) but clearly are separated by HPLC on a reversed-phase column after reduction and S-pyridylethylation (Usami et al., 1993). Botrocetin a- and b-subunits show a high degree of similarity to the GPIb-binding proteins from B. jararaca (64 and 56% for a- and b-subunits, respectively, Kawasaki et al., 1996), mamushigin from Agkistrodon halys blomhoffii (54 and 55%, Sakurai et al., 1998), CHH-B from Crotalus horridus horridus (45 and 51%, Ward et al., 1996) or coagulation factor IX/X binding protein (FIX/X-BP) from Trimeresurus flavoviridis (66 and 51%, Atoda et al., 1991) (Fig. 4). Since botrocetin has a highly acidic pI of 4.6, it was suggested that it binds to the positively charged area on VWF and might induce the conformational change of the A1 domain by altering the electrostatic condition upon binding (Berndt et al., 1992). Bitiscetin was purified from the venom of viperide B. arietans as a second venom coagglutinin (Hamako et al., 1996). The primary structure was determined to be a heterodimer of the a (134 residues) and b-subunits (125 residues) with a C-type lectin-like motif (Matsui et al., 1997) (Fig. 4). The two subunits are clearly separated by SDS-PAGE under reduced conditions as two distinct bands (approximately 17 kDa a- and 14 kDa b-subunit). Each corresponding subunit shows 49 (a) and 42 (b), 47 (a) and

44 (b) % similarity to FIX/X-BP and botrocetin, respectively, indicating the relatively low similarity to botrocetin even though bitiscetin has the same function as botrocetin. Bitscetin has a positively charged nature with a pI of 9.1 in contrast to botrocetin. This property of bitiscetin suggested that it might bind to a different site on VWF from botrocetin, however, crystal structural analysis clearly indicated that both proteins bind to VWF A1 domain in very close proximity as described below. 3.2. Crystal structure of botrocetin and bitiscetin The 3D structures of botrocetin and bitiscetin were ˚ resolution, respectively, in elucidated at 1.8 and 2.0 A 2001 (Hirotsu et al., 2001; Sen et al., 2001) (Fig. 5). The overall structures of botrocetin and bitiscetin expectedly resembled other C-type lectin-like snake venom proteins such as FIX/X-BP (Mizuno et al., 1997), FX-BP from Deinagkistrodon acutus (Atoda et al., 1998), GP-Ib binding protein (flavocetin A) from flavoviridis (Fukuda et al., 1999) and GPVI binding protein (convulxin) from Crotalus durissus terrificus (Murakami et al., 2003). All these proteins are disulfide-linked heterodimer proteins, and both subunits are composed of two a helices and five to eight b strands, forming a globular domain with an extended loop domain (Figs. 4 and 5). Each loop domain of the subunit extends to a globular domain of the opposite subunit and embraces each other by hydrophobic interaction in addition to an inter-subunit disulfide bonding. This bow tie-like structure provides a central concave zone hinged by both subunits. This concave spot might form a ligand-binding site since the negatively

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

1079

Fig. 4. Comparison of the amino acid sequence of botrocetin and bitiscetin. Amino acid sequences (a- and b-subunits) of botrocetin, bitiscetin and other C-type lectin-like snake venom proteins (jararaca GPIb-binding proteins (GPIb-BP), mamushigin, flavocetin A and coagulation factor IX/X binding protein (FIX/X-BP)) are aligned. Numbers indicate the residue number of botrocetin. The secondary structure is expressed as shaded bars (a helix) and black arrows (b strand) at the top referring to Sen et al. (2001). Residues in circle indicate the important residues for the binding to VWF, and residues in square indicate the residues forming a positively charged patch.

charged Gla domain of FX interacts with a positively charged patch in the central concave portion of FX-BP, revealed by the crystal analysis of the complex (Mizuno et al., 2001) (Fig. 5). Interestingly, surface electro potential showed that both botrocetin and bitiscetin have a negatively charged patch at the central concave region, which was not found in other C-type lectin-like venom proteins. This negatively charged region is formed from the cluster of acidic amino acid residues of both subunits. Though bitiscetin shows a basic nature (pI 9.1), mainly due to the b-subunit (rich in basic amino acid residues), acidic residues are assembled in the central concave zone. Thus, despite the relatively low amino acid sequence similarity between botrocetin and bitiscetin, the surface potential of the putative ligand binding sites resemble each other. Bitiscetin, however, also has a positively charged patch on the b subunit near the concave zone, which is not present in botrocetin, suggesting another possibility that bitiscetin interacts with a negatively charged region of VWF (Hirotsu et al., 2001).

3.3. Binding site of botrocetin and bitiscetin on VWF Botrocetin and bitiscetin specifically bind to VWF but not to GPIb or FIX/X in spite of their structural similarities to GPIb- or FIX/X-binding venom proteins (Hamako et al., 1996). The binding activity of proteolytic VWF fragments to botrocetin (Howard et al., 1989; Andrews et al., 1989b) indicated that the A1 domain of VWF corresponds to the botrocetin-binding site. Before the 3D structure of the A1 domain was resolved, the A1 domain had been speculated as a large loop hinged by a single disulfide bond (Fujimura et al., 1996) (Fig. 1). It was suggested that the anionic botrocetin binds to an inner cationic region of the A1 loop model, whereas the cationic ristocetin binds to the outer anionic hinge regions of the A1 domain (Girma et al., 1990; Berndt et al., 1992). Monoclonal antibody (mAb) NMC-4 raised against VWF specifically inhibited both GPIb- and botrocetin-bindings to VWF suggesting that these binding sites are in a close proximity (Fujimura et al., 1992). Three discontinuous parts of the A1 domain were suggested to be a botrocetin-binding site by an analysis using a series of

1080

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

Fig. 5. Structures of botrocetin, bitiscetin and FX-BP. Ribbon representation of a-(upper) and b-(lower) subunits of botrocetin, bitiscetin and FX-BP. Two subunits embrace each other with extending loop and globular domains. The ligand binding sites (central concave zone) are oriented to the right of the figure as shown by FX-BP complexed with Gla domain (Gla domain is expressed by space filling model). Figures were generated with RasMol (Sayle and Milner-White, 1995) referring to the Protein Data Bank entry 1IJK (botrocetin), 1JWI (bitiscetin) and 1IOD (FX-BP and Gla domain complex).

synthetic peptides spanning the A1 domain (Sugimoto et al., 1991). Another approach using Ala-scanning mutagenesis introduced to the charged amino acid residues in the A1 domain indicated that selective basic amino acid residues (Arg629, Arg632, Arg636 and Lys667) play a critical role in botrocetin-binding (Matsushita and Sadler, 1995; Matsushita et al., 2000), since these mutant VWF with point mutation showed a decreased botrocetin-binding (Fig. 2). Those former three residues (Arg629, Arg632 and Arg636) were also essential for the binding to NMC-4, which was coincident with the results showing that NMC-4 inhibits the binding of botrocetin to VWF (Fujimura et al., 1991b). In 1998, crystal structures of the A1 domain and the complex with NMC-4 (Fab) were elucidated and the binding sites for GPIb and botrocetin were suggested, respectively (Celikel et al., 1998; Emsley et al., 1998; Cruz et al., 2000). These critical amino acid residues for botrocetin-binding are located in the a4 helix facing the outside of the A1 domain and overlapping the binding surface of NMC-4 (Fig. 2). In bitiscetin, the binding site on VWF had been suggested to be different from that of botrocetin since bitiscetin showed a positively charged nature like ristocetin. Obert et al. (1999) found that bitiscetin did not bind to the mutant VWF lacking the A3 domain and that bitiscetin competitively inhibited the binding between collagen and VWF. Since the A3 domain contains a major collagenbinding site, bitiscetin might remotely induce a conformational change on the A1 domain by binding to the A3 domain as collagen might do in vivo. The surface potential of bitiscetin indicates that bitiscetin contains a positively

charged patch in the b-subunit near the negatively charged patch at the concave structure suggesting that the former patch might interact with the negatively charged surface of the A3 domain (Hirotsu et al., 2001). However, botrocetin competitively inhibited the bitiscetin-binding to VWF, and Ala-scanning mutagenesis of the A1 domain indicated that Arg632, Lys660, Gln666 and Lys673 residues, present on the a5 and the a6 helix in the A1 domain, were essential for the binding of bitiscetin (Fig. 2). These results suggested that bitiscetin binds to the A1 domain distinctly but is in close proximity to the botrocetin-binding site (Matsui et al., 2002). Bitiscetin showed binding activity to VWF even after SDS-PAGE but lost its activity if bitiscetin was reduced to subunits, suggesting that there is a structural requirement of intra- or inter-subunit disulfide bonding for the VWFbinding activity (Hamako et al., 1996). Western blot analysis using anti-botrocetin mAb BCI-7, which inhibits the binding between botrocetin and VWF, indicates that BCI-7 reacted to the reduced a-subunit but not to the b-subunit (Fujimura et al., 1996), suggesting that botrocetin a-subunit interacts with VWF. In bitiscetin, mAbs recognizing the a-subunit inhibited the platelet agglutination induced by bitiscetin or ristocetin without inhibiting the binding between bitiscetin and VWF, whereas the mAbs recognizing the bitiscetin b-subunit did not show any effect on the platelet agglutination (Matsui et al., 2002). These results suggest that mAb against the bitiscetin asubunit could have caused VWF to have been unable to access GPIb due to the steric hindrance.

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

1081

partially overlaps the binding site of mAb NMC-4 (Fig. 2), which is consistent with the result that NMC-4 also inhibits the binding of bitiscetin to VWF (Matsui et al., 2002). The botrocetin and bitiscetin binding sites on VWF overlap at the a4 and a5 helices, but interestingly, the direction of the long-axis of bitiscetin is almost perpendicular to that of botrocetin (Fig. 6). The Kd values for botrocetin- and bitiscetin-binding to VWF A1 domain are estimated as 12 (Miura et al., 2000) and 2 nM (Maita et al., 2003), respectively. 3.4. Modulation mechanism of botrocetin and bitiscetin

Fig. 6. Structure of the A1 domain complexed with botrocetin and bitiscetin. Botrocetin and bitiscetin (expressed by backbone model) bind to the VWF A1 domain in a close site but perpendicularly. The A1 domain (expressed by space filling model) is oriented so as the GPIb-binding site to face the front of the figure. Asp560 and Gly561 residues in the b2–b3 turn are expressed in dark gray. The basic residues forming a positively charged patch in botrocetin and bitiscetin (Fig. 4) are expressed by stick model in black.

Recently crystal structures of the A1 domain complexed with botrocetin (Fukuda et al., 2002) and bitiscetin (Maita et al., 2003) have been elucidated (Fig. 6). The binding surface of VWF for botrocetin contains Arg632, Arg636, Gln639, Arg663 and Lys667 in the a4 and the a5 helix as expected from Ala-scanning results (Matsushita et al., 2000), and these polar residues interact with botrocetin Glu95(b), Glu107(a), Lys49(a), Asp86(b) and Asp89(b) residues, respectively (Fukuda et al., 2002) (Fig. 4). Further, Tyr45 (a) and Tyr91 (b) also interact with Gln668 and Gln661 in the A1 domain, respectively, and a number of Lue residues provide hydrophobic contacts near the center of the interface. Botrocetin binds via its concave region with comparable contributions from the a (43%) and b (57%) subunits. In bitiscetin, helix a5 of the A1 domain is positioned near the center of the concavity of bitiscetin but interacts mainly with the b-subunit, and helix a4 makes contact with the a-subunit (Maita et al., 2003). Structural data are consistent with Ala-mutagenesis studies since Arg632 (a4), Lys660 (a5), Glu666 (a5) and Lys673 (a5) residues, which are important for the binding to bitiscetin (Matsui et al., 2002), are all involved in hydrogen bond interactions with bitiscetin Tyr64(a), Glu96(a), Trp111(b)CLys117(b) and Glu22(b), respectively (Fig. 4). Further, hydrogen bond interactions between His106 (a) and Gln639, between Gln110 (b) and Arg663, between Lys20 (b) and both Gln686 and Glu689, and between Arg115 (b) and Glu666, Lys667 and Ala669 were observed. Hydrophobic interactions were also observed between the bitiscetin b-subunit (Leu58, Val64, Lue65, Phe104 and Ile109) and the A1 domain (Pro655, Leu659, Ile662 and Val676) (Maita et al., 2003). The binding site of bitiscetin on the A1 domain

It has been proposed that botrocetin and bitiscetin modulate VWF susceptible to GPIb by evoking an allosteric structural change on the GPIb-binding domain of the VWF. The crystal structure of the A1 complexed with the GPIba fragment indicated that Asp560 interacts with the side chain of Thr240 in GPIba (Huizinga et al., 2002), suggesting that the conformation of Asp560-Gly561 could change upon GPIba binding. In VWD type 2B, which shows an increased affinity toward GPIb resulting in a spontaneous platelet aggregation, many single amino acid substitutions have been found at the bottom face of the A1 domain (Ginsburg, 1999). In this type 2B mutation, VWF has lost the regulation for GPIb-binding (gain-of-function). These mutation sites are far from the GPIb-binding site on the A1 domain, suggesting that the mutation remotely up-regulates the affinity for the GPIb-binding through conformational change within the domain. The crystal structure of the mutant A1 domain containing I546V substitution, one of the type 2B VWD mutations, has been compared with the wild type A1 domain. The mutation propagates a small change at Asp560 and Gly561 residues in the turn between b2 and b3 strands in the A1 domain which plays a key role in interacting with GPIb (Celikel et al., 2000; Fukuda et al., 2002). Although it is not sufficient to conclude that this small conformational change could explain the cause of all type 2B VWDs, these results suggest that the binding of botrocetin and bitiscetin also provides some conformational changes on the A1 domain, especially on the GPIb-binding surface. However, contrary to expectation, crystal structures of the A1 domain complexed with botrocetin and bitiscetin have indicated that there are no significant conformational changes on the GPIb-binding site of the A1 domain before and after binding with botrocetin or bitiscetin (Fukuda et al., 2002; Maita et al., 2003). Fukuda et al. (2002) further found that the above small change induced by I546V mutation was switched to the wild type conformation by complexing with botrocetin. These contradictory observations indicate that the toxin-binding does not influence the structure of the surface residues corresponding to up-regulation of the GPIbbinding, suggesting that the modulation mechanism of VWF by snake venoms is different from those performed in vivo at high-shear stress or by ristocetin in vitro.

1082

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

It has been found that the shear-dependent interaction between VWF and GPIb more closely correlates with ristocetin-dependent binding rather than with botrocetininduced binding under static conditions by using a panel of specific antibodies (Dong et al., 2001b). The N-terminal domain of the GPIb close to the b-switch contains three sulfated Tyr residues (Tyr276, Tyr278 and Tyr279), and the proteolytic fragment or mutant GPIb deleting these sulfated residues causes the loss of VWF-binding activity in the presence of botrocetin, but it has normal activity in the presence of ristocetin (Ward et al., 1996; Dong et al., 2001a). The crystal structure of the A1 and GPIba complex showed that the sulfated region is disordered indicating that it is flexible and flanking (Huizinga et al., 2002). The importance of the anionic region of the GPIba for the botrocetin-induced VWF-binding suggests that it plays an essential role in the high affinity binding of GPIba to botrocetin-A1 complex even though botrocetin does not bind to GPIb by itself. Maita et al. (2003) found that both bitiscetin and botrocetin provide a positively charged patch close to the respective anionic region of the GPIba by ternary complex modeling (Fig. 6). The positive patch of bitiscetin is composed of five Lys residues of the a-subunit (Lys17, Lys20, Lys21, Lys61 and Lys120), whereas that of botrocetin is formed by Arg98 (a), Lys101 (a), Lys102 (b), Lys107 (b), Trp109 (b), Arg115 (b) and Lys117 (b) (Fig. 4). They suggested that bitiscetin- or botrocetin-induced binding of GPIb to VWF depends on electrostatic interactions between the anionic region of GPIb (sulfated Tyr residues) and a complementary electropositive spot contributed by the venom protein in the VWF A1-bitiscetin or VWF-botrocetin complex (Fig. 7). Although GPIb without the N-terminal anionic region has not been

examined for its effect on the VWF-bitiscetin complex, it is conceivable that the bitiscetin and botrocetin complexed with the A1 domain provide a supplemental platform suited for GPIb rather than inducing an allosteric conformational change on the A1 domain. This new hypothesis is in good agreement with the results that GPIb without the sulfated Tyr residues failed to bind VWF in the presence of botrocetin (Dong et al., 2001a). Further crystallographic analysis using the ternary complex of VWF A1-GPIbbotrocetin/bitiscetin might clearly elucidate the modulation mechanism of these venom proteins in the near future.

4. VWF-cleaving venoms Another group of proteins interacting with VWF are the snake venom metalloproteinases (SVMPs). These belong to the metzincin family that has a Zn2C-binding motif of HEXXHXXGXXH. Two such class PIII SVMPs, jararhagin from B. jararaca (Kamiguti et al., 1994) and kaouthiagin (Hamako et al., 1998) from Naja naja kaouthia have been reported to cleave human VWF. Although mocarhagin from N. mocambique mocambique, which has sequence homology with jarahagin, blocks VWF-dependent platelet aggregation, it cleaves GPIba near the sulfated Tyr residues (Ward et al., 1996). Complete amino acid sequences of jararhagin from cDNA (Paine et al., 1992) and kaouthiagin (Ito et al., 2001) indicate that they belong to the Reprolysin family of metalloproteinases which included the ADAMs family of metalloproteinases, consisting of metalloproteinase, disintegrin-like and Cys-rich domains (Bjarnason and Fox, 1995; Jia et al., 1996; Matsui et al., 2000) (Fig. 8).

Fig. 7. Hypothetical model of the VWF-modulation by botrocetin and bitiscetin. Botrocetin and bitiscetin bind to the A1 domain of VWF with their negatively charged concave domain. The complex shows no significant conformational change on the GPIb-binding site but provides a comfortable platform for GPIba since positively charged patch near the concave zone locates close to the anionic region (sulfated Tyr cluster) of the GPIba. In this model, VWF complexed with botrocetin and bitiscetin could make a firm binding with N-terminal of GPIba, with the help of these electrostatic interactions. This model is modified from Maita et al. (2003).

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

Fig. 8. Domain structures of kaouthiagin and ADAMTS-13. Kaouthiagin (401 residues) is a P-III type reprolysin comprising of metalloproteinase (Met), disintegrin-like (Dis) and Cys-rich (CR) domains. Both kaouthiagin and ADMTS-13 contain a conserved Zn2C-binding motif (HEXXHXXGXXH sequence) in the metalloproteinase domain, and also have an RGD sequence in the Cys-rich domain. In addition, ADAMTS-13 has propeptides (P), thrombospondin type 1 motif (Ts), spacer (Sp) and CUB domains.

Kaouthiagin specifically cleaves VWF between Pro708 and Asp709 in the C-terminal hinge region of the A1 domain and decreases the multimeric structure of VWF resulting in the loss of both collagen binding and platelet agglutinating activities of VWF (Fig. 3). This proteolytic activity of kaouthiagin is Ca2C- and Zn2C-dependent, but the binding of kaouthiagin to VWF is not affected by EDTA treatment of kaouthiagin. Further, Western Blot analysis indicates that the VWF-binding activity of kaouthiagin is not influenced by SDS-treatment but is lost after reduction of kaouthiagin (Hamako et al., 1998). Jararhagin cleaves not only VWF but also fibrinogen and the b1 subunit of platelet integrin a2b1, although the cleaving site on VWF has not yet been determined (Kamiguti et al., 1996). Since these are chimeric proteins composed of distinct domains, they produce a complex effect on the hemostasis. It is well known that disintegrins, an RGD containing short protein purified from snake venoms, interferes the interaction between fibrinogen and integrin aII bb3 (GPIIb/IIIa) resulting in the inhibition of fibrinogen-mediated platelet aggregation (McLane et al., 1998). It has been proven that metalloproteinase, disintegrin-like and Cys-rich domain of jararhagin can each interact with the a2b1 collagen receptor of the platelet even though each has no RGD sequence (Pentikainen et al., 1999; Kamiguti et al., 2003; Nymalm et al., 2003). Kaouthiagin uniquely has an RGD sequence in the Cys-rich domain (Fig. 8). Synthetic circular peptides of disintegrin-like domain and Cys-rich domain show an inhibitory effect on the collagen- or ADP-induced platelet aggregation, suggesting that kaouthiagin and jararhagin affect hemostasis by interfering with both platelet aggregation mediated by fibrinogen or collagen receptor (integrins) and VWF function. Some metalloproteinases from snake venom have been reported to activate VWF to induce platelet aggregation (De Luca et al., 1995), but their modulation mechanisms have not been elucidated. VWF is synthesized as an ultra-large multimer and degraded into smaller multimers by an endogenous metalloproteinase under high shear stress (Tsai, 1996).

1083

Recently, this endogenous VWF-cleaving metalloproteinase was elucidated to be a family of ADAMTS (a disintegrinlike and metalloproteinase with thrombospondin type 1 motif) (Fujikawa et al., 2001; Soejima et al., 2001; Dong et al., 2002). The VWF-cleaving proteinase thus named ADAMTS-13 specifically cleaves VWF between Tyr842 and Met843 at the A2 domain, producing a heterogeneous multimer sized VWF (Fig. 3). The physiological significance of ADAMTS-13 involves the regulation of the hemostatic activity of circulating VWF since the hemostatic activity of VWF strictly depends on its multimer size. In chronic relapsing TTP or Upshaw-Schulman syndrome (congenital TTP) patients, a deficient activity of ADAMTS13 is observed (Furlan and Lammle, 1999; Tsai et al., 2001; Fujimura, 2003). ADAMTS-13 contains Cys-rich domain followed by spacer, seven thrombospondin type 1 motifs and two CUB domains (Fig. 8). A deletion study has indicated that Cys-rich and spacer domains are essential for VWF-cleaving activity of the metalloproteinase domain suggesting that these domains function as a VWFrecognizing site (Soejima et al., 2003). ADAMTS-13 contains an RGD sequence in its Cys-rich domain as kaouthiagin does (Fig. 8), but this signal does not contribute to VWF-cleaving activity since mutated protein with RGE substitution normally cleaves VWF. The biological function of the RGD sequence, thrombospondin type 1 repeats and CUB domain of ADAMTS-13 requires further investigation. Kaouthiagin cleaves VWF near the sialylated Oglycosidically linked sugar chain cluster between the A1 and the A2 domain resembling mocarhagin in that mocarhagin also tends to cleave the anionic region of GPIba. ADAMTS-13 requires high shear stress or low-ionic strength or high concentrations of urea at the digestion of VWF in vitro (Furlan et al., 1996; Fujimura et al., 2002), but kaouthiagin cleaves VWF in a normal saline conditions (Hamako et al., 1998). The kaouthiagin clevage site may be located at the hinge region between the A1 and the A2 domains (Fig. 1), whereas the ADAMTS-13 cleaving site may be at the carboxyl terminal of b4 strand exposed at the top of the A2 domain from the secondary structure prediction (Perkins et al., 1994). Since the A2 domain has no disulfide linkage (Fig. 1), the tertiary structure of the A2 domain could be more flexible and susceptible to shear stress compared to the A1 and A3 domains. The A2 domain plays a key role in proteolytic degradation since many point mutations in this domain have been found in the type 2A VWD that lacks large multimers due to increased proteolytic degradation of VWF (Sadler and Ginsburg, 1993). It is interesting that the reprolysin-like enzymes (metzincin family) play an important role in physiological hemostatic control in vivo, while there was no homologue of botrocetin-like modulators of mammalian origin except for the 130 kDa component immuno-reactive to anti-botrocetin mAb which was extracted from human vein (Katayama et al., 1995). Crystal structures of this metalloproteinase

1084

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

and the complex with VWF have not yet been reported at present. Elucidation of the detailed structure-function relationship of these metalloproteinases must wait for further investigation.

the development of the molecular design of anti-hemostatic drugs in the near future.

Acknowledgements 5. Perspective Snake venom is a natural library for screening valuable bioactive substances for hemostasis, thrombosis and neuropathology. These snake venom cofactors described are useful for clinical evaluation or subsidiary diagnosis of bleeding disorders as well as for basic investigation of the molecular mechanism of platelet plug formation induced by VWF and platelets. However, investigators often encounter difficulties in preparing homogeneous proteins from the venom powder of the same species, since the contents of the bioactive substances vary among individual commercial lots. It is well known that diet, geographic origin, sex and age of the snake influence the composition of the venom (Chippaux et al., 1991; Daltry et al., 1996; Sasa, 1999). Further, some venoms contain competitive components: B. jararaca contains not only botrocetin but also GPIbbinding protein which inhibits the botrocetin-induced platelet agglutination (Fujimura et al., 1995). Some lots of B. jararaca venom show strong activity of GPIb-BP, which apparently conceals the activity of botrocetin present if any. It is necessary to fractionate crude venom extracts to check for each potential activity, even if the crude extract shows no apparent activity. There are also some isoforms derived from several amino acid substitutions (Matsui et al., 1997; Monteiro et al., 1998). Although it is not clear why snake venoms contain so many proteins which are structurally similar but do show different specificities like immunoglobulins, it seems likely that C-type lectin-like snake venom proteins also evolve faster like snake venom phospholipases or serine proteases under adaptive pressure to acquire new physiological activities (Deshimaru et al., 1996; Ogawa et al., 1996). Because of this microheterogeneity due to the original venom lot or the protein structure itself, it will be preferable to develop recombinant proteins with a constant activity rather than preparing them from natural resources. These recombinant modulators will be widely prevalent for clinical tests and also for basic research as a more sensitive and reliable standard reagent instead of ristocetin. A recombinant technique also implies a possibility of producing dominant-negative botrocetin and bitiscetin by introducing point mutations, which have the ability to bind VWF without inducing GPIb-binding. The specificity of kaouthiagin toward VWF suggests a possibility that it may applicable for the degradation of ultra-large VWF found in TTP as a kind of substitute for endogenous ADAMTS-13. Further structural and functional analysis of the VWFaffecting snake venom proteins may lead to and accelerate

We thank Mr Ronald G. Belisle for editing the manuscript.

References Andrews, R.K., Booth, W.J., Gorman, J.J., Castaldi, P.A., Berndt, M.C., 1989a. Purification of botrocetin from Bothrops jararaca venom. Analysis of the botrocetin-mediated interaction between von Willebrand factor and the human platelet membrane glycoprotein Ib-IX complex. Biochemistry 28, 8317–8326. Andrews, R.K., Gorman, J.J., Booth, W.J., Corino, G.L., Castaldi, P.A., Berndt, M.C., 1989b. Cross-linking of a monomeric 39/34-kDa dispase fragment of von Willebrand factor (Leu-480/Val-481-Gly-718) to the N-terminal region of the alpha-chain of membrane glycoprotein Ib on intact platelets with bis(sulfosuccinimidyl) suberate. Biochemistry 28, 8326– 8336. Atoda, H., Hyuga, M., Morita, T., 1991. The primary structure of coagulation factor IX/factor X-binding protein isolated from the venom of Trimeresurus flavoviridis. Homology with asialoglycoprotein receptors, proteoglycan core protein, tetranectin, and lymphocyte Fc epsilon receptor for immunoglobulin E. J. Biol. Chem. 266, 14903–14911. Atoda, H., Ishikawa, M., Mizuno, H., Morita, T., 1998. Coagulation factor X-binding protein from Deinagkistrodon acutus venom is a Gla domain-binding protein. Biochemistry 37, 17361–17370. Berndt, M.C., Ward, C.M., Booth, W.J., Castaldi, P.A., Mazurov, A.V., Andrews, R.K., 1992. Identification of aspartic acid 514 through glutamic acid 542 as a glycoprotein Ib-IX complex receptor recognition sequence in von Willebrand factor. Mechanism of modulation of von Willebrand factor by ristocetin and botrocetin. Biochemistry 31, 11144–11151. Bjarnason, J.B., Fox, J.W., 1995. Snake venom metalloendopeptidases: reprolysins. Methods Enzymol. 248, 345–368. Celikel, R., Varughese, K.I., Madhusudan, M., Yoshioka, A., Ware, J., Ruggeri, Z.M., 1998. Crystal structure of the von Willebrand factor A1 domain in complex with the function blocking NMC-4 Fab. Nat. Struct. Biol. 5, 189–194. Celikel, R., Ruggeri, Z.M., Varughese, K.I., 2000. von Willebrand factor conformation and adhesive function is modulated by an internalized water molecule. Nat. Struct. Biol. 7, 881–884. Chippaux, J.P., Williams, V., White, J., 1991. Snake venom variability: methods of study, results and interpretation. Toxicon 29, 1279–1303. Cruz, M.A., Diacovo, T.G., Emsley, J., Liddington, R., Handin, R.I., 2000. Mapping the glycoprotein Ib-binding site in the von willebrand factor A1 domain. J. Biol. Chem. 275, 19098–19105. Daltry, J.C., Wuster, W., Thorpe, R.S., 1996. Diet and snake venom evolution. Nature 379, 537–540.

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087 De Luca, M., Ward, C.M., Ohmori, K., Andrews, R.K., Berndt, M.C., 1995. Jararhagin and jaracetin: novel snake venom inhibitors of the integrin collagen receptor, alpha 2 beta 1. Biochem. Biophys. Res. Commun. 206, 570–576. Deshimaru, M., Ogawa, T., Nakashima, K., Nobuhisa, I., Chijiwa, T., Shimohigashi, Y., Fukumaki, Y., Niwa, M., Yamashina, I., Hattori, S., Ohno, M., 1996. Accelerated evolution of crotalinae snake venom gland serine proteases. FEBS Lett. 397, 83–88. Dong, J., Ye, P., Schade, A.J., Gao, S., Romo, G.M., Turner, N.T., McIntire, L.V., Lopez, J.A., 2001a. Tyrosine sulfation of glycoprotein I(b)alpha. Role of electrostatic interactions in von Willebrand factor binding. J. Biol. Chem. 276, 16690– 16694. Dong, J.F., Berndt, M.C., Schade, A., McIntire, L.V., Andrews, R.K., Lopez, J.A., 2001b. Ristocetin-dependent, but not botrocetin-dependent, binding of von Willebrand factor to the platelet glycoprotein Ib-IX-V complex correlates with sheardependent interactions. Blood 97, 162–168. Dong, J.F., Moake, J.L., Nolasco, L., Bernardo, A., Arceneaux, W., Shrimpton, C.N., Schade, A.J., McIntire, L.V., Fujikawa, K., Lopez, J.A., 2002. ADAMTS-13 rapidly cleaves newly secreted ultralarge von Willebrand factor multimers on the endothelial surface under flowing conditions. Blood 100, 4033–4039. Emsley, J., Cruz, M., Handin, R., Liddington, R., 1998. Crystal structure of the von Willebrand factor A1 domain and implications for the binding of platelet glycoprotein Ib. J. Biol. Chem. 273, 10396–10401. Esnouf, R.M., 1999. Further additions to MolScript version 1.4, including reading and contouring of electron-density maps. Acta Crystallogr., D Biol. Crystallogr. 55 (Pt 4), 938–940. Fujimura, Y., 2003. Is Upshaw-Schulman syndrome congenital thrombotic thrombocytopenic purpura or hemolytic-uremic syndrome? Yes to both. J. Thromb. Haemost. 1, 2457–2458. Fujimura, Y., Titani, K., 1993. Structure and function of von Willebrand factor. In: Thomas, D. (Ed.), Haemostasis and Thrombosis, vol. 1. Churchill Livingstone, Edinburgh, pp. 379– 395. Fujimura, Y., Titani, K., Usami, Y., Suzuki, M., Oyama, R., Matsui, T., Fukui, H., Sugimoto, M., Ruggeri, Z.M., 1991a. Isolation and chemical characterization of two structurally and functionally distinct forms of botrocetin, the platelet coagglutinin isolated from the venom of Bothrops jararaca. Biochemistry 30, 1957–1964. Fujimura, Y., Usami, Y., Titani, K., Niinomi, K., Nishio, K., Takase, T., Yoshioka, A., Fukui, H., 1991b. Studies on anti-von Willebrand factor (vWF) monoclonal antibody NMC-4, which inhibits both ristocetin- and botrocetin-induced vWF binding to platelet glycoprotein Ib. Blood 77, 113–120. Fujimura, Y., Miyata, S., Nishida, S., Miura, S., Kaneda, M., Yoshioka, A., Fukui, H., Katayama, M., Tuddenham, E.G., Usami, Y., et al., 1992. The interaction of botrocetin with normal or variant von Willebrand factor (types IIA and IIB) and its inhibition by monoclonal antibodies that block receptor binding. Thromb. Haemost. 68, 464–469. Fujimura, Y., Ikeda, Y., Miura, S., Yoshida, E., Shima, H., Nishida, S., Suzuki, M., Titani, K., Taniuchi, Y., Kawasaki, T., 1995. Isolation and characterization of jararaca GPIb-BP, a snake venom antagonist specific to platelet glycoprotein Ib. Thromb. Haemost. 74, 743–750.

1085

Fujimura, Y., Kawasaki, T., Titani, K., 1996. Snake venom proteins modulating the interaction between von Willebrand factor and platelet glycoprotein Ib. Thromb. Haemost. 76, 633–639. Fujikawa, K., Suzuki, H., McMullen, B., Chung, D., 2001. Purification of human von Willebrand factor-cleaving protease and its identification as a new member of the metalloproteinase family. Blood 98, 1662–1666. Fujimura, Y., Matsumoto, M., Yagi, H., Yoshioka, A., Matsui, T., Titani, K., 2002. Von Willebrand factor-cleaving protease and Upshaw-Schulman syndrome. Int. J. Hematol. 75, 25–34. Fukuda, K., Mizuno, H., Atoda, H., Morita, T., 1999. Crystallization and preliminary X-ray studies of flavocetin-A, a platelet glycoprotein Ib-binding protein from the habu snake venom. Acta Crystallogr., D Biol. Crystallogr. 55, 1911–1913. Fukuda, K., Doggett, T.A., Bankston, L.A., Cruz, M.A., Diacovo, T.G., Liddington, R.C., 2002. Structural basis of von Willebrand factor activation by the snake toxin botrocetin. Structure (Camb) 10, 943–950. Furlan, M., Lammle, B., 1999. von Willebrand factor in thrombotic thrombocytopenic purpura. Thromb. Haemost. 82, 592–600. Furlan, M., Robles, R., Lamie, B., 1996. Partial purification and characterization of a protease from human plasma cleaving von Willebrand factor to fragments produced by in vivo proteolysis. Blood 87, 4223–4234. Furlan, M., Robles, R., Galbusera, M., Remuzzi, G., Kyrle, P.A., Brenner, B., Krause, M., Scharrer, I., Aumann, V., Mittler, U., Solenthaler, M., Lammle, B., 1998. von Willebrand factorcleaving protease in thrombotic thrombocytopenic purpura and the hemolytic-uremic syndrome. N. Engl. J. Med. 339, 1578– 1584. Ginsburg, D., 1999. Molecular genetics of von Willebrand disease. Thromb. Haemost. 82, 585–591. Girma, J.P., Takahashi, Y., Yoshioka, A., Diaz, J., Meyer, D., 1990. Ristocetin and botrocetin involve two distinct domains of von Willebrand factor for binding to platelet membrane glycoprotein Ib. Thromb. Haemost. 64, 326–332. Hamako, J., Matsui, T., Suzuki, M., Ito, M., Makita, K., Fujimura, Y., Ozeki, Y., Titani, K., 1996. Purification and characterization of bitiscetin, a novel von Willebrand factor modulator protein from Bitis arietans snake venom. Biochem. Biophys. Res. Commun. 226, 273–279. Hamako, J., Matsui, T., Nishida, S., Nomura, S., Fujimura, Y., Ito, M., Ozeki, Y., Titani, K., 1998. Purification and characterization of kaouthiagin, a von Willebrand factorbinding and -cleaving metalloproteinase from Naja kaouthia cobra venom. Thromb. Haemost. 80, 499–505. Hirotsu, S., Mizuno, H., Fukuda, K., Qi, M.C., Matsui, T., Hamako, J., Morita, T., Titani, K., 2001. Crystal structure of bitiscetin, a von Willebrand factor-dependent platelet aggregation inducer. Biochemistry 40, 13592–13597. Howard, M.A., Greco, T., Coghlan, M., 1989. Cleavage of human von Willebrand factor by porcine pancreatic elastase. Blood 74, 673–681. Huizinga, E.G., Martijn van der Plas, Kroon, J., Sixma, J.J., Gros, P., 1997. Crystal structure of the A3 domain of human von Willebrand factor: implications for collagen binding. Structure 5, 1147–1156. Huizinga, E.G., Tsuji, S., Romijn, R.A., Schiphorst, M.E., de Groot, P.G., Sixma, J.J., Gros, P., 2002. Structures of glycoprotein Ibalpha and its complex with von Willebrand factor A1 domain. Science 297, 1176–1179.

1086

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087

Ito, M., Hamako, J., Sakurai, Y., Matsumoto, M., Fujimura, Y., Suzuki, M., Hashimoto, K., Titani, K., Matsui, T., 2001. Complete amino acid sequence of kaouthiagin, a novel cobra venom metalloproteinase with two disintegrin-like sequences. Biochemistry 40, 4503–4511. Jia, L.G., Shimokawa, K., Bjarnason, J.B., Fox, J.W., 1996. Snake venom metalloproteinases: structure, function and relationship to the ADAMs family of proteins. Toxicon 34, 1269–1276. Kamiguti, A.S., Slupsky, J.R., Zuzel, M., Hay, C.R., 1994. Properties of fibrinogen cleaved by Jararhagin, a metalloproteinase from the venom of Bothrops jararaca. Thromb. Haemost. 72, 244–249. Kamiguti, A.S., Hay, C.R., Theakston, R.D., Zuzel, M., 1996. Insights into the mechanism of haemorrhage caused by snake venom metalloproteinases. Toxicon 34, 627–642. Kamiguti, A.S., Gallagher, P., Marcinkiewicz, C., Theakston, R.D., Zuzel, M., Fox, J.W., 2003. Identification of sites in the cysteine-rich domain of the class P-III snake venom metalloproteinases responsible for inhibition of platelet function. FEBS Lett. 549, 129–134. Katayama, M., Nagata, S., Hirai, S., Miura, S., Fujimura, Y., Matsui, T., Kato, I., Titani, K., 1995. Fibronectin and 130 kDa molecule complex mimics snake venom botrocetin-like structure potentially modulating association between von Willebrand factor and vascular vessel wall. J. Biochem. (Tokyo) 117, 331– 338. Kawasaki, T., Fujimura, Y., Usami, Y., Suzuki, M., Miura, S., Sakurai, Y., Makita, K., Taniuchi, Y., Hirano, K., Titani, K., 1996. Complete amino acid sequence and identification of the platelet glycoprotein Ib-binding site of jararaca GPIb-BP, a snake venom protein isolated from Bothrops jararaca. J. Biol. Chem. 271, 10635–10639. Maita, N., Nishio, K., Nishimoto, E., Matsui, T., Shikamoto, Y., Morita, T., Sadler, J.E., Mizuno, H., 2003. Crystal structure of von Willebrand factor A1 domain complexed with snake venom, bitiscetin: insight into glycoprotein Ibalpha binding mechanism induced by snake venom proteins. J. Biol. Chem. 278, 37777–37781. Matsui, T., Titani, K., Mizuochi, T., 1992. Structures of the asparagine-linked oligosaccharide chains of human von Willebrand factor. Occurrence of blood group A, B, and H(O) structures. J. Biol. Chem. 267, 8723–8731. Matsui, T., Hamako, J., Suzuki, M., Hayashi, N., Ito, M., Makita, K., Fujimura, Y., Ozeki, Y., Titani, K., 1997. Complete amino acid sequence of bitiscetin, a novel von Willebrand factor modulator protein, purified from snake venom of Bitis arietans. Res. Commun. Biochem. Cell Mol. Biol. 1, 271–284. Matsui, T., Fujimura, Y., Titani, K., 2000. Snake venom proteases affecting hemostasis and thrombosis. Biochim. Biophys. Acta 1477, 146–156. Matsui, T., Hamako, J., Matsushita, T., Nakayama, T., Fujimura, Y., Titani, K., 2002. Binding site on human von Willebrand factor of bitiscetin, a snake venom-derived platelet aggregation inducer. Biochemistry 41, 7939–7946. Matsumoto, M., Kokame, K., Soejima, K., Miura, M., Hayashi, S., Fujii, Y., Iwai, A., Ito, E., Tsuji, Y., Takeda-Shitaka, M., Iwadate, M., Umeyama, H., Yagi, H., Ishizashi, H., Banno, F., Nakagaki, T., Miyata, T., Fujimura, Y. 2004. Molecular characterization of ADAMTS13 gene mutations in Japanese patients with Upshaw-Schulman syndrome. Blood 103, 1305– 1310.

Matsushita, T., Sadler, J.E., 1995. Identification of amino acid residues essential for von Willebrand factor binding to platelet glycoprotein Ib. Charged-to-alanine scanning mutagenesis of the A1 domain of human von Willebrand factor. J. Biol. Chem. 270, 13406–13414. Matsushita, T., Meyer, D., Sadler, J.E., 2000. Localization of von willebrand factor-binding sites for platelet glycoprotein Ib and botrocetin by charged-to-alanine scanning mutagenesis. J. Biol. Chem. 275, 11044–11049. McLane, M.A., Marcinkiewicz, C., Vijay-Kumar, S., WierzbickaPatynowski, I., Niewiarowski, S., 1998. Viper venom disintegrins and related molecules. Proc. Soc. Exp. Biol. Med. 219, 109–119. Miura, S., Li, C.Q., Cao, Z., Wang, H., Wardell, M.R., Sadler, J.E., 2000. Interaction of von Willebrand factor domain A1 with platelet glycoprotein Ibalpha-(1-289). Slow intrinsic binding kinetics mediate rapid platelet adhesion. J. Biol. Chem. 275, 7539–7546. Mizuno, H., Fujimoto, Z., Koizumi, M., Kano, H., Atoda, H., Morita, T., 1997. Structure of coagulation factors IX/X-binding protein, a heterodimer of C-type lectin domains. Nat. Struct. Biol. 4, 438–441. Mizuno, H., Fujimoto, Z., Atoda, H., Morita, T., 2001. Crystal structure of an anticoagulant protein in complex with the Gla domain of factor X. Proc. Natl Acad. Sci. USA 98, 7230– 7234. Monteiro, R.Q., Dutra, D.L., Machado, O.L., Carlini, C.R., Guimaraes, J.A., Bon, C., Zingali, R.B., 1998. Bothrops jararaca snakes produce several bothrojaracin isoforms following an individual pattern. Comp. Biochem. Physiol., B Biochem. Mol. Biol. 120, 791–798. Murakami, M.T., Zela, S.P., Gava, L.M., Michelan-Duarte, S., Cintra, A.C., Arni, R.K., 2003. Crystal structure of the platelet activator convulxin, a disulfide-linked alpha4beta4 cyclic tetramer from the venom of Crotalus durissus terrificus. Biochem. Biophys. Res. Commun. 310, 478–482. Novak, L., Deckmyn, H., Damjanovich, S., Harsfalvi, J., 2002. Shear-dependent morphology of von Willebrand factor bound to immobilized collagen. Blood 99, 2070–2076. Nymalm, Y., Puranen, J.S., Nyholm, T.K., Kapyla, J., Kidron, H., Pentikainen, O.T., Airenne, T.T., Heino, J., Slotte, J.P., Johnson, M.S., Salminen, T.A., 2004. Jararhagin-derived ‘RKKH1’peptides induce structural changes in alpha 1 I domain of human integrin alpha 1 beta 1. J. Biol. Chem. 279, 7962–7970. Obert, B., Houllier, A., Meyer, D., Girma, J.P., 1999. Conformational changes in the A3 domain of von Willebrand factor modulate the interaction of the A1 domain with platelet glycoprotein Ib. Blood 93, 1959–1968. Ogawa, T., Nakashima, K., Nobuhisa, I., Deshimaru, M., Shimohigashi, Y., Fukumaki, Y., Sakaki, Y., Hattori, S., Ohno, M., 1996. Accelerated evolution of snake venom phospholipase A2 isozymes for acquisition of diverse physiological functions. Toxicon 34, 1229–1236. Paine, M.J., Desmond, H.P., Theakston, R.D., Crampton, J.M., 1992. Purification, cloning, and molecular characterization of a high molecular weight hemorrhagic metalloprotease, jararhagin, from Bothrops jararaca venom. Insights into the disintegrin gene family. J. Biol. Chem. 267, 22869–22876. Pentikainen, O., Hoffren, A.M., Ivaska, J., Kapyla, J., Nyronen, T., Heino, J., Johnson, M.S., 1999. ‘RKKH’ peptides from the snake

T. Matsui, J. Hamako / Toxicon 45 (2005) 1075–1087 venom metalloproteinase of Bothrops jararaca bind near the metal ion-dependent adhesion site of the human integrin alpha(2) Idomain. J. Biol. Chem. 274, 31493–31505. Perkins, S.J., Smith, K.F., Williams, S.C., Haris, P.I., Chapman, D., Sim, R.B., 1994. The secondary structure of the von Willebrand factor type A domain in factor B of human complement by Fourier transform infrared spectroscopy. Its occurrence in collagen types VI, VII, XII and XIV, the integrins and other proteins by averaged structure predictions. J. Mol. Biol. 238, 104–119. Read, M.S., Shermer, R.W., Brinkhous, K.M., 1978. Venom coagglutinin: an activator of platelet aggregation dependent on von Willebrand factor. Proc. Natl Acad. Sci. USA 75, 4514– 4518. Ruggeri, Z.M., 2003. Von Willebrand factor, platelets and endothelial cell interactions. J. Thromb. Haemost. 1, 1335–1342. Sadler, J.E., 1998. Biochemistry and genetics of von Willebrand factor. Annu. Rev. Biochem. 67, 395–424. Sadler, J.E., Ginsburg, D., 1993. A database of polymorphisms in the von Willebrand factor gene and pseudogene. For the Consortium on von Willebrand Factor Mutations and Polymorphisms and the Subcommittee on von Willebrand Factor of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Thromb. Haemost. 69, 185–191. Sakurai, Y., Fujimura, Y., Kokubo, T., Imamura, K., Kawasaki, T., Handa, M., Suzuki, M., Matsui, T., Titani, K., Yoshioka, A., 1998. The cDNA cloning and molecular characterization of a snake venom platelet glycoprotein Ib-binding protein, mamushigin, from Agkistrodon halys blomhoffii venom. Thromb. Haemost. 79, 1199–1207. Sasa, M., 1999. Diet and snake venom evolution: can local selection alone explain intraspecific venom variation?. Toxicon 37, 249– 252 author reply 253–260. Savage, B., Almus-Jacobs, F., Ruggeri, Z.M., 1998. Specific synergy of multiple substrate-receptor interactions in platelet thrombus formation under flow. Cell 94, 657–666. Sayle, R.A., Milner-White, E.J., 1995. RASMOL: biomolecular graphics for all. Trends Biochem. Sci. 20, 374. Sen, U., Vasudevan, S., Subbarao, G., McClintock, R.A., Celikel, R., Ruggeri, Z.M., Varughese, K.I., 2001. Crystal structure of the von Willebrand factor modulator botrocetin. Biochemistry 40, 345–352. Siedlecki, C.A., Lestini, B.J., Kottke-Marchant, K.K., Eppell, S.J., Wilson, D.L., Marchant, R.E., 1996. Shear-dependent changes in the three-dimensional structure of human von Willebrand factor. Blood 88, 2939–2950.

1087

Soejima, K., Mimura, N., Hirashima, M., Maeda, H., Hamamoto, T., Nakagaki, T., Nozaki, C., 2001. A novel human metalloprotease synthesized in the liver and secreted into the blood: possibly, the von Willebrand factor-cleaving protease?. J. Biochem. (Tokyo) 130, 475–480. Soejima, K., Matsumoto, M., Kokame, K., Yagi, H., Ishizashi, H., Maeda, H., Nozaki, C., Miyata, T., Fujimura, Y., Nakagaki, T., 2003. ADAMTS-13 cysteine-rich/spacer domains are functionally essential for von Willebrand factor cleavage. Blood 102, 3232–3237. Somers, W.S., Tang, J., Shaw, G.D., Camphausen, R.T., 2000. Insights into the molecular basis of leukocyte tethering and rolling revealed by structures of P- and E-selectin bound to SLe(X) and PSGL-1. Cell 103, 467–479. Sugimoto, M., Mohri, H., McClintock, R.A., Ruggeri, Z.M., 1991. Identification of discontinuous von Willebrand factor sequences involved in complex formation with botrocetin. A model for the regulation of von Willebrand factor binding to platelet glycoprotein Ib. J. Biol. Chem. 266, 18172–18178. Tsai, H.M., 1996. Physiologic cleavage of von Willebrand factor by a plasma protease is dependent on its conformation and requires calcium ion. Blood 87, 4235–4244. Tsai, H.M., Lian, E.C., 1998. Antibodies to von Willebrand factorcleaving protease in acute thrombotic thrombocytopenic purpura. N. Engl. J. Med. 339, 1585–1594. Tsai, H.M., Li, A., Rock, G., 2001. Inhibitors of von Willebrand factor-cleaving protease in thrombotic thrombocytopenic purpura. Clin. Lab. 47, 387–392. Usami, Y., Fujimura, Y., Suzuki, M., Ozeki, Y., Nishio, K., Fukui, H., Titani, K., 1993. Primary structure of two-chain botrocetin, a von Willebrand factor modulator purified from the venom of Bothrops jararaca. Proc. Natl Acad. Sci. USA 90, 928–932. Usami, Y., Fujimura, Y., Miura, S., Shima, H., Yoshida, E., Yoshioka, A., Hirano, K., Suzuki, M., Titani, K., 1994. A 28 kDa-protein with disintegrin-like structure (jararhagin-C) purified from Bothrops jararaca venom inhibits collagen- and ADP-induced platelet aggregation. Biochem. Biophys. Res. Commun. 201, 331–339. Ward, C.M., Andrews, R.K., Smith, A.I., Berndt, M.C., 1996. Mocarhagin, a novel cobra venom metalloproteinase, cleaves the platelet von Willebrand factor receptor glycoprotein Ibalpha. Identification of the sulfated tyrosine/anionic sequence Tyr-276-Glu-282 of glycoprotein Ibalpha as a binding site for von Willebrand factor and alpha-thrombin. Biochemistry 35, 4929–4938.