Ohanin a novel protein from king cobra (ophiophagus hannah) venom

Chapter I Literature Review 1.2 VENOMOUS SNAKES Approximately 1300 snake species are venomous Hider et al.. Generally, snake PLA2 enzyme is a single chain polypeptide of approximately 1

Chapter I Literature Review

1.2 VENOMOUS SNAKES

Approximately 1300 snake species are venomous (Hider et al 1991; Stafford 2000) The

evolution of the venomous form, however, was much more recent, possibly as late as the Miocene period (less than 30 million years ago) (Harris 1991) Venomous snakes are

usually defined as those that have venom glands and specialized venom conducting fangs which enable them to inflict fatal bites upon their victims (Klemmer 1968)

1.2.1 Classification and distribution of venomous snakes

The systematic classification of both venomous and non-venomous snakes still presents many problems Most taxonomists and authorities would only recognize 11 to 13 distinct families However, venomous snakes are generally identified in only five families They are the Elapidae, Hydrophiidae, Viperidae, Crotalidae and Colubridae (Harris 1991) It is interesting to note that snakes are widely distributed on all continents in the world except Antarctica, New Zealand, Madagascar, Ireland, Greenland, the Azores and Canaries (Phelps 1981) They have successfully evolved into efficient predators and colonized various habitats from mangrove swamps, estuaries, freshwater lakes, streams, dunes, grasslands to forests (Garl and Roger 1989) The classification and distribution of venomous snakes in the world are shown in Table 1.1

1.2.2 King cobra (Ophiophagus hannah)

King cobra, also known as Ophiophagus hannah (Figure 1.1), belongs to the Elapidae

family It is the longest venomous snake in the world King cobra has an average size ranging from 10 to 12 feet, but sometimes can grow up to 18 feet (5.5 meter) long (Zhao 1990) It is widely distributed in the northern parts of India, southern China (Hainan, Fujian, Guangdong, Hainan, Guangxi, Guangzhou and Hong Kong) and southeast Asia (Malaysia, Indonesia, Burma, Thailand, Philippines and Singapore) (Ganthavorn 1971;

Chapter I Literature Review

Zhao 1990) King cobra is generally found in dense or open rainforests, as well as mangrove swamps, bamboo thickets, savannas and even around human settlements

Its genus name, Ophiophagus, means snake eater, with ‘ophis’ and ‘phagein’

representing ‘snake’ and ‘to eat’, respectively in Latin Hence king cobra preferentially feeds on snakes and small reptiles These preys sometimes can be as huge as 10 feet in length In addition to snakes, it also feeds on mice, rats, birds, frogs and fishes The king cobra kills the prey by injecting a lethal amount of venom with its fangs It then swallows its preys as a whole It hunts both during the day and night time

King cobra yields an average of 420 mg of crude venom in dry weight per milking (Ganthavorn 1971) The LD50 in mouse is ~1.2 to 3.5 mg/kg via intravenous injection (Mebs 1989) The relatively low toxicity of king cobra’s venom is compensated

by the large amount of venom produced and injected into the preys each time King cobra’s venom shows predominantly haemotoxic and neurotoxic effects The clinical manifestations upon envenomation are: drowsiness, stupor, ptosis, dysarthria, dysphagia and general muscular weakness (Ganthavorn 1971) In severe envenomation, impairment

of cardiovascular function can occur (Reid 1968; Wetzel and Christy 1989)

Chapter I Literature Review

Figure 1.1 King cobra (Ophiophagus hannah) Photo is reprinted with the permission of

Mr Peter Mirtschin from Venom Supplies Pty Ltd., Australia

1.3 SNAKE VENOMS

Snake venoms are secretory products of venom glands (Oron and Bdolah 1973) Typical venom glands consist of three major cell types, namely basal cells, conical mitochondria-rich cells and secretory cells Venom is only produced by secretory cells in the glands (Oron and Bdolah 1978) It is further carried from the glands to the fangs by the ducts that flow through the accessory glands The function of the accessory glands is to prevent wasteful flow of the secretions Venom production appears to be regulated by the glands themselves and is independently of neural control

Venom proteins are used mainly to immobilize and kill the preys and predators as well as to support the digestion of the food swallowed by the snake (Aird 2002) The composition of venom components varies with the time of secretion into the glands For example, venom that is freshly secreted into the glands has a different composition than

venom that has been allowed to mature (De Lucca et al 1974; Kochva and Gans 1965) It

should be noted also that the variation in the population or individual, age, diet, geographical distribution and climate, can easily influence the venom’s composition quantitatively and qualitatively even within the same species (Sasa 1999)

The composition of the venoms also differs between different families of venomous snakes For example, elapid and hydrophid venoms are rich in neurotoxic proteins and peptides They have been known to induce effects at the nervous systems (Chang 1979) On the contrary, crotalid and viperid venoms are rich in proteinases These proteinases, such as the serine proteinases and metalloproteinases, tend to cause

Chapter I Literature Review

hemolytic effects and are largely responsible for the necrosis following the snake bite However, in general, the closer the phylogenetic relationships of the snakes, the more similar are the venom properties and compositions (Tu 1996)

Snake venom proteins have evolved to target different tissues, organs and physiological systems Hence, a diversity of symptoms arises after a snake bite which will ultimately lead to failure of multiple tissues, organs and systems and often death

(Torres et al 2003) Some of the major clinical symptoms are intense localized pain, loss

of consciousness, drowsiness, headache, vomiting, inflammation, bleeding, shock, hemorrhage, necrosis and muscular paralysis (Campbell 1979; Efrati 1979; Reid 1979; Russell 1979)

1.3.1 Compositions and properties of snake venoms

Although snake venoms have always been of great interest for studies, it is only in the recent years serious attempts have been made to fractionate individual venoms These studies have shown that snake venoms consist of proteins as well as non-protein components The minor, non-proteinaceous components of snake venoms are metals, lipids, nucleotides, carbohydrates and amines The proteinaceous components, which consist of ~ 90 to 95 % of the total dry weight of the venom, can be further grouped as

enzymatic and non-enzymatic peptides and proteins (Hider et al 1991)

The major enzyme groups found in snake venoms include phospholipases A2 (PLA2), serine proteinases, metalloproteinases, phosphodiesterases, acetylcholinesterase,

L-amino acid oxidases, glycosidase, hyaluronidase and nucleotidases (Torres et al 2003)

Generally, enzymes in the venom have molecular mass ranging from 13,000 Da to 150,000 Da Most of these are hydrolases and possess a digestive role There are also over 1000 non-enzymatic venom proteins that have been characterized They are grouped into three-finger toxins, serine proteinase inhibitors, C-type lectin-related proteins, disintegrins, helveprins/ CRISPs, waprins, sarafatoxins, nerve growth factors, natriuretic

peptides and bradykinin-potentiating peptides (Kini 2002; Mochca-Morales et al 1990; Torres et al 2003; Yamazaki et al 2003) The first category of non-enzymatic venom

peptides and proteins has a molecular mass around 1,000 Da to 25,000 Da and are rich in disulfide bonds Therefore, they are robust and are relatively stable once isolated Another category is the low molecular mass compounds having the molecular mass of less than 1,500 Da They are less active biologically and are likely to be enzyme cofactors (Bieber 1979) Some of these families are selected and discussed in the subsequent literature review

1.3.1.1 Phospholipases A 2 (PLA 2 )

Phospholipases are esterolytic enzymes that hydrolyze 3-sn-phosphoglycerides

According to the sites of hydrolysis, they are classified as phospholipase A1, A2, B, C and

D (Kini 1997) Snake venoms are one of the richest sources of secretory phospholipases Most of the snake venom phospholipases are PLA2 as theyhydrolyze the sn-2 ester bond

of 3-sn-phosphoglycerides, releasing lysophospholipids and fatty acids (Kini 1997)

Generally, snake PLA2 enzyme is a single chain polypeptide of approximately 118 to 130 amino acid residues with high cysteine content (seven disulfide bonds)(Scott 1997)

Chapter I Literature Review

Snake venom PLA2 enzymes can be divided into classes I and II Class I enzymes are abundant in Elapidae and Hydrophidae snake venoms, whereas class II proteins are mainly isolated from Crotalidae and Viperidae venoms Class I can be further classified into classes IA and IB enzymes, based on the presence or absence of the pancreatic loop

In the region 52 to 65 (bovine pancreatic PLA2 sequence numbering) (Dufton and Hider

1983; Renetseder et al 1985), class I proteins display an insertion of two to three amino

acid residues (the ‘elapid’ loop), which is extended by a further five amino acid residues

in the case of mammalian pancreatic PLA2s (the ‘pancreatic’ loop) This loop is absent in class II PLA2 The position of one of the seven disulfide bonds is also different between class I and II PLA2s Class I PLA2s have the Cys11-Cys77 disulfide bridge which is absent in class II But class II PLA2s possess an alternative disulfide bridge between Cys51-Cys133 at the C-terminal extension (Dufton and Hider 1983)

So far, the protein and cDNA sequences of over 280 snake PLA2 enzymes have

been determined (Danse et al 1997; Tan et al 2003) These sequences indicate that

snake PLA2 contain multiple isoenzymes Gene sequences determined further demonstrate that these isoenzymes are from different but closely related PLA2 genes likely to have evolved from the physiological PLA2 (Kordis and Gubensek 1996;

Nakashima et al 1993) Generally, the primary sequence similarity among snake venom

PLA2 isoenzymes can reach ~40 to 99 % Furthermore, they also share high similarities in their secondary structures and overall foldings (Figure 1.2) (Scott 1997)

Interestingly, unlike mammalian PLA2 enzymes which are only involved in catalysis, snake venom PLA2 isoenzymes are able to induce wide arrays of

pharmacological actions including presynaptic and postsynaptic neurotoxicity (Strong et

al 1976), myotoxicity (Gopalakrishnakone et al 1984; Ponraj and Gopalakrishnakone

1995), cardiotoxicity (Lee et al 1977), hemolytic (Condrea et al 1981), anticoagulant effect (Verheij et al 1980), antiplatelet (Chen and Chen 1989), hypotension (Huang 1984), internal hemorrhage (Vishwanath et al 1987), organ or tissue damage and edema (Vishwanath et al 1987, 1988) The high affinity interaction between PLA2 isoenzymes with their acceptor(s)/receptor(s) is likely due to the complementarity of the contact surfaces in terms of the ionic charges, hydrophobicity and van der Waals force (Kini 2003) Hence, snake PLA2 isoenzymes are able to induce a wide spectrum of pharmacological effects, by the mechanisms either dependent on or independent of their catalytic activity, upon binding to the targets (Kini 2003) Among these pharmacological actions, only neurotoxic, myotoxic and anticoagulant effects have been well-studied, thus providing a great challenge to protein chemists to solve the complex puzzle in the structure-function relationships and mechanisms of action (Kini 2003)

1.3.1.2 Snake venom L-amino acid oxidases

L-amino acid oxidase (EC1.4.3.2) (LAAO) is a flavoenzyme that catalyses the L-amino acid substrate to an α-keto acid along with the production of ammonia and hydrogen peroxide LAAOs are found in many different organisms, such as snakes, bacteria, fungi and plants Snake venom L-amino acid oxidases (SV-LAAOs) represent the best studied

Chapter I Literature Review

Figure 1.2 Snake Phospholipases A2 isoenzymes Three-dimensional structures of snake PLA2s, class I (A) and class II (B) Arrow indicates the extra pancreatic or elapid loop

region which is only present in class I enzymes Figure is generously given and reprinted

with the permission from Prof R Manjunatha Kini

members of this protein family They are non-covalently bound homodimeric glycoproteins of around 110 to 150 kDa with FAD-binding motif (Du and Clemetson 2002) SV-LAAOs are widely distributed in the family of Viperidae, Crotalidae and

Elapidae (Du and Clemetson 2002; Pawelek et al 2000) Sequence alignment of

SV-LAAOs is shown in Figure 1.3A

Extensive studies have been carried out on SV-LAAOs only in the past ten years Studies have shown that SV-LAAOs elicit wide arrays of pharmacological actions For

example, SV-LAAOs from Crotalus adamanteus and Crotalus atrox can associate

specifically with mammalian endothelial cells (Suhr and Kim 1996) and can either induce

(Ahn et al 1997; Ali et al 2000; Li et al 1994) or inhibit platelet aggregation (Sakurai et

al 2001; Suhr and Kim 1996; Takatsuka et al 2001; Tan and Swaminathan 1992).It was found that SV-LAAOs can also induce apoptosis in mammalian endothelial cells, possibly through the productionof highly localized concentrations of hydrogen peroxide

(Abe et al 1998; Ali et al 2000; Souza et al 1999; Suhr and Kim 1996, 1999; Torii et al 1997) Two SV-LAAOs isolated from Pseudechis australis were identified to exhibit potent anti-bacterial effect against gram-positive and gram-negative bacteria (Stiles et al

1991) The exact mechanism of action(s) and physiological role(s) of SV-LAAOs in the venom is still poorly understood

X-ray crystallographic structure of SV-LAAO from Calloselasma rhodostoma

(PDB: 1F8R) was solved (Figure 1.3B) (Pawelek et al 2000) SV-LAAO is a dimer and

each subunit consists of three domains, namely FAD-binding domain, substrate binding

Chapter I Literature Review

domain and helical domain The structure of SV-LAAO resembles the reported structure

of porcine DAAO (D-amino acid oxidase) particularly within the FAD-binding domain

(Pawelek et al 2000) Furthermore, the structure provides information about two

glycosylation sites at positions Asn172 and Asn361 which appear to be important for its

catalytic activity (Torii et al 2000) The glycans identified from experimental results

were either of bis-sialylated, biantennary and/or core-fucosylated dodecasaccharide

(Geyer et al 2001) which can be accommodated in the crystal structure obtained

1.3.1.3 Three-finger toxins

This is a non-enzymatic protein family found only in the venoms of elapids (cobras, kraits and mambas), hydrophids (sea snakes) and colubrids (mangrove snakes) The common structural feature of all three-finger toxins is the three β-stranded loops that emerge from the small, hydrophobic central core which is cross-linked by four conserved disulfide bridges Because of the 3-D structures appearance, this family of proteins is called the three-finger toxin family (Endo and Tamiya 1991; Kini 2002; Menez 1998; Servent and Menez 2001; Tsetlin 1999) The high cysteine content makes the toxins stable and robust once isolated

Generally, all three-finger toxins are basic proteins of 60 to 74 amino acid residues with 4 or 5 disulfide bridges They can be grouped according to their lengths Short-chain three-finger toxins, for example, erabutoxin-b (Figure 1.4A) and toxin-α

* 20 * 40 * 60

C rhodostoma : A DRNPLEECFRETDYEEFLEIAKNGLTATSNPKRVVIVGAGMAGLSAAYVLAGAGHQVTVL C adamanteus : A DRNPLEECFRETDYEEFLEIAKNGLTATSNPKRVVIVGAGMAGLSAAYVLAGAGHQVTVL T stejnegeri : A DRNPLEECFRETDYEEFLEIARNGLKATSNPKRVVIVGAGMSGLSAAYVLAGTGHEVTVL B moojeni : A DRNPLEECFRETDYEEFLEIAKNGLSTTSNPKRVVIVGAGMSGLSAAYVLANAGHQVTVL B jararacussu : A DRNPLEECFRETDYEEFLEIAKNGLSTTSNPKRVVIVGAGMSGLSAAYVLANAGHQVTVL

* 80 * 100 * 120

C rhodostoma : EASERVGGRVRTYR KKDWYANLGPMRLPTKHRIVREYIKKFDLKLNEFSQENENAWYFIK C adamanteus : EASERVGGRVRTYR KKDWYANLGPMRLPTKHRIVREYIKKFDLKLNEFSQENENAWYFIK T stejnegeri : EASERAGGRVRTYRNDEEGWYANLGPMRLPEKHRIVREYIRKFNLQLNEFSQENDNAWHFVK B moojeni : EASERAGGRVKTYRNEKEGWYANLGPMRLPEKHRIVREYIRKFDLQLNEFSQENENAWYFIK B jararacussu : EASERAGGQVKTYRNEKEGWYANLGPMRLPEKHRIVREYIRKFG QLNEFSQENENAWYFIK

* 140 * 160 * 180

C rhodostoma : NIRKRVREVKNNPGLLEYPVKPSEEGKSAAQLYVESLR VVEELRSTNCKYILDKYDTYSTK C adamanteus : NIRKRVREVKNNPGLLEYPVKPSEEGKSAAQLYVESLR VVEELRSTNCKYILDKYDTYSTK T stejnegeri : NIRKT GEVKKDPGVLKYPVKPSEEGKSAEQLYEESLREVEKELKRTNCSYILNKYDTYSTK B moojeni : NIRKRVGEVNKDPGVLEYPVKPSEVGKSAGQLYEESLQ AVEELRRTNCSYMLNKYDTYSTK B jararacussu : NIRKRVGEVNKDPGVLDYPVKPSEVGKSAGQLYEESLQ AVEELRRTNCSYMLNKYDTYSTK

* 200 * 220 * 240

C rhodostoma : EYLLKEGNLSPGAVDMIGDLLNEDSGYYVSFIESLKHDDIFGYEKRFDEIVGGMDQLPTSMY C adamanteus : EYLLKEGNLSPGAVDMIGDLLNEDSGYYVSFIESLKHDDIFGYEKRFDEIVGGMDQLPTSMY T stejnegeri : EYLIKEGNLSPGAVDMIGDLMNEDAGYYVSFIESMKHDDIFAYEKRFDEIVDGMDKLPTSMY B moojeni : EYLLKEGNLSPGAVDMIGDLLNEDSGYYVSFIESLKHDDIFAYEKRFDEIVGGMDKLPTSMY B jararacussu : EYLLKEGNLSPGAVDMIGDLLNEDSGYYVSFIESLKHDDIFAYEKRFDEIVGGMDKLPTSMY

* 260 * 280 * 300 *

C rhodostoma : EAIKEKVQ HFNARVIEIQQNDREATVTYQTSANEMSSVTADYVIVCTTSRAARRIKFEPPL C adamanteus : EAIKEKVQ HFNARVIEIQQNDREATVTYQTSANEMSSVTADYVIVCTTSRAARRIKFEPPL T stejnegeri : RAIEEKVH FNAQVIKIQK AEE TVTYQTPEKDTSFVTADYVIVCTTSGAARRIKFEPPL B moojeni : QAIQEKVH NARVIKIQQDVKEVTVTYQTSEKETLSVTADYVIVCTTSRAARRIKFEPPL B jararacussu : QAIQEKVH NARVIKIQQDVKEVTVTYQTSEKETLSVTADYVIVCTTSRAARRIKFEPPL

320 * 340 * 360 *

C rhodostoma : PPKKAHALRSVHYRSGTKIFLTCTKKFWEDDGIHGGKSTTDLPSRFIYYPNHNFTSGVGVII C adamanteus : PPKKAHALRSVHYRSGTKIFLTCTKKFWEDDGIHGGKSTTDLPSRFIYYPNHNFTSGVGVII T stejnegeri : P KKAHALRSVHYRSGTKIFLTCTKKFWEDEGIHGGKSTTDLPSRFIYYPNHNFTSGVGVII B moojeni : PPKKAHALRSVHYRSGTKIFLTCTKKFWEDDGIHGGKSTTDLPSRFIYYPNHNFPNGVGVII B jararacussu : PPKKAHALRSVHYRSGTKIFLTCTKKFWEDDGIHGGKSTTDLPSRFIYYPNHNFPNGVGVII

380 * 400 * 420 *

C rhodostoma : AYGIGDDANFFQALDFKDCADIVINDLSLIHELPKEDIQTFC PSMIQRWSLDKYAMGGITT C adamanteus : AYGIGDDANFFQALDFKDCADIVINDLSLIHELPKEDIQTFC PSMIQRWSLDKYAMGGITT T stejnegeri : AYGIGDDANFFQALDLKDCGDIVINDLSLIHQLPREEIQTFC PSMIQKWSLDKYAMGGITT B moojeni : AYGIGDDANYFQALDFEDCGDIVINDLSLIHQLPKEEIQAIC PSMIQRWSLDKYAMGGITT B jararacussu : AYGIGDDANYFEALDFEDCGDIVINDLSLIHQLPKEEIQAIC PSMIQRWSLDKYAMGGITT

440 * 460 * 480 *

C rhodostoma : FTPYQFQHFSEALTAPFKRIYFAGEYTAQFHGWIDSTIKSGLTAARDVNRASENPSGIHLSN C adamanteus : FTPYQFQHFSEALTAPFKRIYFAGEYTAQFHGWIDSTIKSGLTAARDVNRASENPSGIHLSN T stejnegeri : FTPYQFQHFSEALTSHVDRIYFAGEYTAHAHGWIDSSIKSGLTAARDVNRASENPSGIHLSN B moojeni : FTPYQFQHFSEALTAPVDRIYFAGEYTAQAHGWIDSTIKW -

B jararacussu : FTPYQFQHFSEALTAPVDRIYFAGEYTAQAHGWIASTIKSGPEGLDVNRASE -

500

C rhodostoma : DNEF

C adamanteus : DNEF

T stejnegeri : DNEL

B moojeni : -

B jararacussu : -

A

Chapter I Literature Review

B

Figure 1.3 Snake venom L-amino acid oxidases (A) Sequence alignment of

SV-LAAOs Sequences used are: Calloselasma rhodostoma (sp: O93364), Bothrops

moojeni (gb: 39841346), Bothrops jararacussu (gb: 39841344), Trimeresurus stejnegeri

(gb: 33355627) and Crotalus adamanteus (gb: 3426324) The conserved and identical

residues are highlighted (B) Three-dimensional structure of apoxin I, an L-amino acid

oxidase functional dimer obtained from X-ray crystallography of the enzyme isolated

from Calloselasma rhodostoma (PDB: 1F8R) (Pawelek et al 2000) FADs are labeled

pink and positions of the oligosaccharide residues are indicated

from Naja nigricollis, consist of 60 to 62 amino acid residues and are cross-linked by

four conserved disulfide bonds They have cysteine connectivity of 1-3, 2-4, 5-6 and 7-8

On the other hand, long-chain three-finger toxins, for example, α-bungarotoxin and bungarotoxin (Figure 1.4B and 1.4D), have 66 to 74 amino acids and possess five disulfide bridges with the fifth disulfide bridge located in the second loop (loop II) (Endo and Tamiya 1991; Mebs 1989) Long-chain neurotoxin and κ-bungarotoxin contain cysteine connectivity of 1-3, 2-4, 5-6, 7-8 and 9-10 (Fry 1999)

κ-Recently, another group of poorly characterized three-finger toxins was isolated exclusively from the Elapidae and Colubridae families They are called the non-conventional three-finger toxins The unique features that distinguish them from the classical three-finger toxins are the length of the polypeptide chain and the location of disulfide bridges Non-conventional three-finger toxins contain 65 to 67 residues and five disulfide bridges The fifth disulfide bridge is located in the N-terminal loop (loop I) in contrast to the classical long-chain three-finger toxins having their disulfide bonds at loop

II (Nirthanan et al 2003) One such example is condoxin as shown in Figure 1.4E

Although all ‘sibling’ three-finger toxins adopt the common three-finger fold in their tertiary structures, they bind to different receptors/acceptors and hence exhibit a wide variety of biological effects (Dufton and Harvey 1989; Kini 2002) Members of this family include Erabutoxin-a, which antagonize muscle (α1) nicotinic acetylcholine receptors (nAChR) (Figure 1.4A); cardiotoxins (cytotoxins), that exert their toxicity by forming pores in cell membranes (Figure 1.4C); κ-bungarotoxins, which recognize

Chapter I Literature Review

neuronal (α3β4) nicotinic receptors (Figure 1.4D); fasciculins, that inhibit acetylcholinesterase (Figure 1.4F); muscarinic toxins, with selectivity towards distinct types of muscarinic receptors (Figure 1.4G); FS2 toxin, that block the L-type calcium channels (Figure 1.4H); and dendroaspin, which are antagonists of various cell-adhesion processes (Figure 1.4I) (Kini 2002) In many instances, the functional sites of interaction between these pharmacologically diverse toxins and their molecular targets have been successfully identified using a combination of theoretical and experimental approaches These functional motifs include the cytolytic, anti-platelet aggregation and anagelsic sites (Kini 2002) (Figure 1.5)

1.3.1.4 Snake venom nerve growth factors

Nerve growth factor (NGF) was the first neurotrophic factor to be identified as a protein

that supports neuronal maintenance and survival (Cohen et al 1954) Its activity was described in two sarcoma tissues, 37 and 180 (Cohen et al 1954); and later in snake venom when an active ‘component’ promoting fibre outgrowth in spinal ganglia in vitro

was identified from the crude venom of Agkistrodon piscivorus (Cohen and

Levi-Montalcini 1956) Subsequent to that, reports on isolation and characterization of sNGFs were mainly from the venom of Viperidae, Crotalidae and Elapidae families (Hogue-

Angeletti 1970; Hogue-Angeletti and Bradshaw 1979)

Figure 1.4 Three-dimensional structural similarities among three-finger toxins from

snake venoms (A) Erabutoxin-a (1QKD); (B) α-Bungarotoxin (2ABX); (C) Cardiotoxin V4 (1CDT); (D) κ-Bungarotoxin (1KBA), inset, dimer; (E) Candoxin (1JGK); (F) Fasciculin-2 (1FAS); (G) Muscarinic toxin MT-2 (1FF4); (H) FS2 toxin (1TFS); and (I)

Dendroaspin (1DRS) These β-sheeted loops are numbered right to left as loop I, II and III, respectively Although these toxins share similar three-finger fold, they are differ from each other in their biological activities Figure was adapted from Kini (2002)

Chapter I Literature Review

Figure 1.5 Functional sites of three-finger toxins Various functional sites identified in

three-finger toxins are indicated as red segments Figure is reprinted with permission

from Prof R Manjunatha Kini

Antiplatelet site

Antiplatelet site

sNGFs can further be divided into four major groups based on their physical properties (Kostiza and Meier 1996):

• Group I: sNGFs isolated from Agkistrodon piscivorus, Crotalus adamanteus,

Bothrops jararaca, Naja nigricollos, Naja melanoleuca and Vipera russelli They

have a molecular mass of ~25,000 Da and are made up from two identical subunits similar to the mouse β-NGF (Hogue-Angeletti and Bradshaw 1979)

• Group II: sNGFs isolated mainly from Bothrops atrox, Agkistrodon rhodostoma,

Vipera ammodytes and Vipera russelli They differ from the first group only by

the presence of carbohydrate moiety Thus the homodimer generally has a higher molecular mass of ~35,000 Da (Hogue-Angeletti and Bradshaw 1979)

• Group III: sNGF isolated exclusively from Bitis arientans It comprises of a homodimer that is covalently linked by disulfide bonds (Smith et al 1992)

• Group IV: sNGF is a heterodimer containing γ- and β-subunits The γ-subunit was

found to be the serine proteinase with arginyl esterase activity (Perez-Polo et al

have further shown that sNGFs are devoid of direct toxic effects (Hogue-Angeletti and Bradshaw 1979) To date, PC-12 cells have been established as a standard bioassay that is

Chapter I Literature Review

* 20 * 40 * 60 Chinese cobra : EDHPVHNLGEHSVCDSVSAWV - TKTTATDIKGNTVTVMENVNLDNKVYKQYFFETKCKNP

Monocled cobra : EDHPVHNLGEHSVCDSVSAWV - TKTTATDIKGNTVTVMENVNLDNKVYKQYFFETKCKNP

Indian cobra : EDHPVHNLGEHSVCDSVSAWV - TKTTATDIKGNTVTVMENVNLDNKVYKQYFFETKCKNP

Mouse beta NGF : - GE F SVCDSVS V WV GD KTTATDIKG KE VTVL AE VNINN S VFRQYFFETKCR AS

* 80 * 100 *

Chinese cobra : NPEPSGCRGIDSSHWNSYCTETDTFIKALTMEGNQASWRFIRI E TACVCVITKK K GN

Monocled cobra : NPEPSGCRGIDSSHWNSYCTETDTFIKALTMEGNQASWRFIRI E TACVCVITKK K GN

Indian cobra : NPEPSGCRGIDSSHWNSYCTETDTFIKALTMEGNQASWRFIRI D TACVCVITKK T GN

Mouse beta NGF : NP VE SGCRGIDS K HWNSYCT T H TFVKALT TDEK QA A WRFIRI D TACVCVLSRK A

Figure 1.6 Snake venom nerve growth factors Sequence alignment of sNGFs with

mouse β-nerve growth factor Sequences used are: chinese cobra (gb: 7438538), monocled cobra (gb: 11275218), Indian cobra (gb: 7428567) and mouse β-nerve growth factor (PDB: 1BET) The identical and conserved residues are highlighted sNGFs shares

~72 % similarity with mouse β-NGF The numbering of residues corresponds to that of sNGFs

widely used to detect fractions with NGF activity after chromatographic procedure (Greene and Rukenstein 1989)

The physiological role(s) of sNGFs in the venom is still not clear It was speculated that some neurotrophic-like molecules, such as the sNGFs, are utilized as a carrier for neurotoxins to gain access into the central or peripheral nervous system and subsequently exert their pharmacological effects (Levi-Montalcini 1987) It was also hypothesized that prior contact of sNGFs to basophil cells may enhance the degranulation process upon CVF or PLA2 treatment (Kostiza and Meier 1996) There is no available three-dimensional structure for sNGFs so far Thus this may provide an opportunity for structural biologists to solve its three-dimensional structure and compare with that from the mouse β-NGF (PDB: 1BET)

1.3.1.5 Helveprins/ CRISPs proteins

Cysteine-rich secretory proteins (CRISPs) are single chain proteins with molecular masses ranging from 20 to 30 kDa CRISPs family possesses 16 strictly conserved cysteines, all of them are engaged in disulfide bonds Remarkably, 10 out of 16 cysteine

residues are clustered at the C-terminal end of the proteins (Osipov et al 2005; Yamazaki

et al 2003) CRISPs family was originally described in the male rodent reproductive tract

(Cameo and Blaquier 1976; Kasahara et al 1989) Subsequently, proteins that belong to this family have been identified in all mammals studied, such as horses and humans (Guo

et al 2005) Most mammalian CRISPs members can be grouped into three classes,

namely CRISP-1, CRISP-2 and CRISP-3, which have functions that are related to

sperm-Chapter I Literature Review

egg fusion, binding of spermatocytes to Seritoli cells and innate host defence,

respectively (Guo et al 2005) Interestingly, CRISPs are also found in the venom of

reptiles

The first reported reptilian CRISPs was identified from the skin secretion of the

Mexican beaded lizard Heloderma horridum horridum (Mochca-Morales et al 1990)

This protein was named helothermine Later on, CRISPs were also identified from all venomous snake families inhabiting in different continents, and thus they belong to a new family of snake venom proteins (Yamazaki and Morita 2004) Based on the primary structural similarity to helothermine, this new family was named helveprins (Helothermine-related Venom Proteins) Some of the examples are trifflin from

Trimeresurus flavoviridis (Viperidae, Asia), ablomin from Agkistrodon blomhoffi

(Viperidae, Asia), pseudechetoxin from Pseudechis australis (Elapidae, Australia), ophanin from Ophiophagus hannah (Elapidae, Asia) and tigrin from Rhabdophis tigrinus

tigrinus (Colubridae, Asia)

Helothermine from the lizard venom was shown to modulate the activity of a variety of ion channels, including voltage-gated calcium channels, potassium channels

and ryanodine receptors (Mochca-Morales et al 1990; Morrissette et al 1995; Nobile et

al 1996; Nobile et al 1994) Helveprins, such as ablomin, latisemin, ophanin, triflin and

piscivorin, were found to inhibit the contraction of smooth muscle induced by high concentration of K+ but not caffeine Thus these toxins are L-type Ca2+ channel blockers

(Yamazaki et al 2002b) Futhermore, PsTx and pseudecin were identified as

CNG-channel blockers as both the toxins appear to act on the olfactory and retinal CNG

channel (Brown et al 1999; Yamazaki et al 2002a) Sequence alignment of helveprins is

shown in Figure 1.7A

Crystal structures of stecrisp and natrin, isolated from Trimeresurus stejnegeri and Naja atra, have been solved recently (Guo et al 2005; Wang et al 2005) Both these

structures contain two well-defined independent domains, namely pathogenesis-related proteins of group 1 domain (PR-1) and cysteine-rich domain (CRD) Both the domains are connected by a hinge The 16 conserved cysteine residues form eight-paired disulfide bridges, with three in PR-1 domain, two in the hinge region and three in CRD domain as shown in Figure 1.7B PR-1 domain was proposed to be the catalytic domain for Tex31,

another CRISPs protein isolated from cone snail (Milne et al 2003) Hence the

amino-terminal moiety is suggested to be responsible for the proteolytic activity in some of the CRISPs members Another domain, CRD, shows no obvious sequence similarity, except for its conservation in the cysteine pattern to Bgk and Shk However, it is intriguing to find that CRD shares similar fold with Bgk and Shk, which are the K+ channel-blocking

toxins (Guo et al 2005) This provides evidence for helveprins/ CRISPs proteins exhibit

various receptor-related activities In this regard, CRD is likely to be the interacting domain with different receptor(s) or acceptor(s) that governs diverse functions of

helveprins/ CRISPs members (Guo et al 2005) The actual binding epitopes with the

target(s) are yet to be identified

Chapter I Literature Review

* 20 * 40 * 60

* 80 * 100 * 120

* 140 * 160 * 180

* 200 * 220

A

B

Figure 1.7 Helveprins/ CRISPs (A) Sequence alignment of helveprins Proteins

sequences used are: helothermine (gb: 2500711), PsTx (gb: 48428844), ablomin (gb:48428846), stepcrisp (gb: 37694046), natrin (gb:32492059) and ophanin (gb: 48428838) The identical and conserved residues are highlighted in black The numbering

of residues corresponds to that of helothermine (B) Three-dimensional structure of

stepcrisp (PDB: 1RC9) The PR-1 domain, hinge region and CRD domain are labeled

1.3.1.6 Sarafotoxins

Sarafotoxin (SRTX) is derived from the Hebrew name of Atractaspis engaddensis, Saraf

Ein Gedi, from which these peptides were first isolated and characterized (Kochva et al

1982; Takasaki et al 1988) SRTXs constitute a family of isopeptides that are structurally

and functionally related to mammalian endothelins (ETs) SRTXs contain 21 to 24 residues with two conserved disulfide bridges between Cys1-Cys15; and between Cys3-Cys11 Since the identification of SRTXs in 1982, a number of isopeptides have been

isolated from the venoms of Atractaspis engaddensis and Atractaspidae-related species

Unlike ETs in which each isoform emerges from distinct precursors, SRTX isopeptides are simultaneously encoded by a polycistronic precursor The rosary organization of SRTXs cDNA allows for the minimum use of transcriptional and translational machinery

to produce multiple copies of SRTXs (Figure 1.8A) This probably explains the high

abundance of SRTXs in Atractaspis engaddensis venom in constrast to pico-molar amounts of ETs in mammalian tissues (Joseph et al 2004) Thus this rare regulation

mechanism appears to be an economical and effective way for the production of SRTXs

(Ducancel et al 1993; Hayashi et al 2004; Takasaki et al 1992)

SRTXs are potent vasoconstrictors that generally act on cardiac muscles and brain

in vertebrates This was based on the previous observations that SRTX-b binds specifically to atrial membrane preparations with maximum binding capacity of 110 fmol

per mg peptide and a dissociation constant (KD) of 3 to 5 nM (Kloog et al 1988) Other

binding experiments demonstrated that 125I-SRTX-b recognizes sites in rat cerebellum

(KD= 3.5 nM) and cerebral cortex (KD= 0.3 nM) (Ambar et al 1988) Furthermore, a

Chapter I Literature Review

recent study showed that SRTXs are able to activate type C and D phospholipases through specific receptor(s) in phosphoinositide signal transduction pathway (Ducancel 2002)

The major sequence variations among the isopeptides are mainly located between

Cys3 and Cys11 (Figure 1.8B) (Sakurai et al 1992) NMR structure of SRTX-b

demonstrates that SRTXs adopt a cysteine-stabilized α-helical motif consist of an extended structure from the first four N-terminal amino acid residues, a turn at residues 5 and 8, followed by an α-helical structure from Lys9 to Cys15, and finally a random coil

conformation at the C-terminal tail as shown in Figure 1.8C (Atkins et al 1995) It is

proposed that the observed differences in biological activity and toxicity of various SRTX isopeptides depend on the residues in the turn region (from Cys3 to Cys11)

1.3.1.7 Bradykinin-potentiating peptides

Bradykinin-potentiating peptides (BPPs) are reported only from Viperid and Crotalid

venoms so far These include Bothrops jararaca (Ferreira 1965; Ferreira and Rocha e Silva 1962; Ferreira and Rocha e Silva 1965), Bothrops neuwiedi (Ferreira et al 1998),

Agkistrodon blomhoffi (Kato and Suzuki 1962; Kato et al 1973), Agkistrodon halys pallas (Cheng-Wu et al 1982), Agkistrodon piscivorus piscivorus (Ferreira et al 1995)

and Crotalus atrox (Politi et al 1985) Generally, as many as five to nine BPP isoforms

have been identified from an individual venom Snake venom BPPs consist of 9 to 13 residues They share the following characteristics, such as having a pyroglutamyl group

SRTX-b

SRTX-a SRTX-c

SRTX-a1

SRTX-e SRTX-a1

SRTX-a1

A

Chapter I Literature Review

* 20 SRTX-a : CSCKDMTDKECLNFCHQDVIW

SRTX-a1 : CSCKDMSDKECLNFCHQDVIW

Figure 1.8 Sarafotoxins (A) Nucleotide and deduced amino acid sequences of

full-length cDNA encoding for SRTX from Atractaspis engaddensis The names of the

isopeptides are indicated at right (B) Sequence alignment of SRTX isopeptides from

Atractaspis engaddensis Intercysteine loops are indicated (C) Three-dimensional

structure of SRTX-b The structure was obtained from NMR spectroscopy (PDB: 1SRB) The two disulfide bridges are indicated as dotted line, extended structures are in white, the turns are in blue and the α-helical segment is in light purple Three-dimensional structure is adapted from Ducancel (2002)

attached to the N-terminal, a high content of proline residues and a tripeptide Ile-Pro-Pro

at the C-terminal (Ferreira et al 1995) Interestingly, results from molecular biology

experiments demonstrated that isopeptides of BPP and natriuretic peptide are synthesized

from the same precursor as shown in Figure 1.9 (Hayashi et al 2003; Higuchi et al 1999; Joseph et al 2004; Murayama et al 1997), providing another example of an ‘economic’

way for producing multiple copies of toxins using the same precursor gene

Snake BPPs are known to be potent hypotensive peptides acting on smooth muscles They can exert this biological activity through two different pathways Firstly, BPPs induce potentiation of bradykinin (BK) Induction of BK causes vasodilation, increases capillary permeability and subsequently followed by a fall in systemic blood pressure Secondly, BPPs can also cause inhibition of angiotensin-converting enzyme

(ACE) both in vitro and in vivo (Ferreira et al 1992; Ferreira et al 1970; Ondetti et al 1971) Reduced level of ACE will cause hypotensive effect (Linz et al 1995) Over the

years, several analogue molecules for BPPs have been synthesized and examined for the biological activities This has finally led to the development of captopril, a potent hypotensive compound, which is widely used today (Cushman and Ondetti 1991;

Cushman et al 1979)

Chapter I Literature Review

Figure 1.9 Bradykinin-potentiating peptides Sequence alignment showing the snake

BPPs and C-type natriuretic peptides are derived from the same precursor molecules

Figure is adapted from Joseph et al 2004

1.3.1.8 Natriuretic peptides

The natriuretic peptide (NP) family consists of peptides that are involved in various physiologically processes controlling natriuresis, diuresis, blood pressure, homeostasis

and inhibition of aldosterone secretion (Chen and Burnett 2000; Cho et al 1999; Matsuo

2001) Generally, NPs have a highly conserved 17-residue, forming cyclic or ring

structure by a cysteine disulfide bridge (Joseph et al 2004) The first isolated NP, atrial

natriuretic peptide (ANP), is primarily synthesized in the heart Subsequent to that, brain natriuretic peptide (BNP) was isolated from porcine brain; and C-type natriuretic peptide (CNP) was found in the central nervous system and vascular endothelial cells Among

the three types of NPs, CNP lacks the C-terminal extension (Amininasab et al 2004)

NPs mediate their function through binding to membrane-bound guanylate

cyclase receptors that are found at high concentrations in the target organs (Chinkers et al 1989; Koller et al 1991) Three types of NPs receptors have been identified so far They

are the NP receptors A, B and C (NPR-A, NPR-B and NPR-C) NPR-A is activated by ANP, but it can also bind to BNP with lower affinity NPR-B is preferentially activated

by CNP, whereas NPR-C binds all three types of receptors with equal affinity but lacks the guanylate cyclase activity (Pandey 2005)

The venoms of snakes also contain structural and functional equivalents of

mammalian NPs (Joseph et al 2004) In 1992, the first venom NP from the green mamba snake Dendroaspis augusticeps was identified, purified and characterized It was named

Dendroaspis natriuretic peptide (DNP) DNP has biological properties similar to both

Chapter I Literature Review

ANPs and CNPs (Schweitz et al 1992) Other venom NPs have also been reported, such

as NPs from the South American coral snake Micrurus corallinus, Pseudechis australis and Trimesures gramineus (Ho et al 1997) Venom NPs differ significantly from

mammalian NPs at the N- and/ or C-terminal regions These segments vary in lengths and share no sequence similarities Sequence alignment of snake DNP in comparison with mammalian ANP, BNP and CNP is shown in Figure 1.10 Due to the potent hypotensive and vasorelaxant properties, efforts have been made to understand the structure-function relationship of these peptides for the development of orally active peptide analogues with ANP mimetic activity (Chin and Lip 2001; Giuseppe 2001)

1.3.1.9 Waprins

Waprins is a new family of snake venom proteins with a molecular mass ranging from 5

to 6 kDa It was originally identified from the venom of Naja nigricolis (Torres et al

2003) Because of its sequence similarity to WAPs (Whey Acidic Proteins), this new family of snake venom proteins was named Waprins (WAP related proteins) The novel protein isolated was thus named nawaprin (Naja waprin) (sp: P60589) Nawaprin has a molecular mass of 5288.50± 0.08 as assessed by ESI/MS and consisted of 51-amino acid residues Another member of waprins, omwaprin (sp: P83768), was also identified and

purified from Oxyuranus microlepidotus So far, only one single waprin member has

been identified and isolated from individual snake venom Although all the cysteine residues are conserved in these proteins, the intercysteine segments are distinctly different, thus implying that the molecular surfaces of the waprin members are also

* 20 * 40 * 60 DNP : -EVKYDP CFG H KIDRI NHV SGLGC PSL R DPRPNAPSTSA - MNP : LAKEALGDG CFG Q RIDRI CNV SGMGC NHV R TDPAPTALARIIPFSRPVRKESRAALDRMQQPG PNP : GENEPPKKKAPDG CFG H KIDRI GSH SGLGC NKF K PGH - ANP : -SLRRSS CFG G RMDRI GAQ SGLGC NSF R Y - BNP : SPKMVQGSG CFG R KMDRI SSS SGLGC KVL R RH - CNP : -GLSKG CFG L KLDRI GSM SGLGC -

* 20 * 40 * 60 DNP : -EVKYDP CFG H KIDRI NHV SGLGC PSL R DPRPNAPSTSA - MNP : LAKEALGDG CFG Q RIDRI CNV SGMGC NHV R TDPAPTALARIIPFSRPVRKESRAALDRMQQPG PNP : GENEPPKKKAPDG CFG H KIDRI GSH SGLGC NKF K PGH - ANP : -SLRRSS CFG G RMDRI GAQ SGLGC NSF R Y - BNP : SPKMVQGSG CFG R KMDRI SSS SGLGC KVL R RH - CNP : -GLSKG CFG L KLDRI GSM SGLGC -

Figure 1.10 Natriuretic peptides Sequence comparison of snake venom natriuretic

peptides, namely DNP (Dendroaspis angusticeps), MNP (Micrurus corallinus) and PNP (Pseudocerastes persicus) with mammalian ANP, BNP and CNP Conserved and

identical residues are shaded

Chapter I Literature Review

different These proteins may interact differently with the molecular target(s) and hence

could display different biological properties (Dileep et al personal communications)

Recently, the three-dimensional structure of nawaprin was determined by NMR

spectroscopy (Torres et al 2003) Nawaprin has a relatively flat and disc-like structure

It is characterized by its spiral backbone configuration that forms the outer and inner circular segments The two circular segments are then held together by four disulfide bridges as shown in Figure 1.11 As this is a new family of snake venom proteins with unknown biological activities, it would be interesting to elucidate the biological properties in order to better understand the structure-function relationships of the novel proteins belonging to this family

1.3.2 Molecular evolution of snake venom proteins

In general, it is interesting to note that all the snake venom proteins reported so far are highly similar to non-venomous body proteins, for example SRTXs/ ETs, CVF / complement C3, helveprins/ CRISPs, three-finger toxins/ CD59 and others (Kini personal communications) Thus, it has been postulated that snake toxins arise from recruitment events of genes from various body protein families during evolution (Kini personal communications) The basis for the multiple and independent recruitment events of certain protein families for use as toxins are unclear However, it is hypothesized that the chosen protein families are likely to be favored in the snake’s adaptive evolution, particularly for its feeding habits Secondly, the chosen protein families must be

Nawaprin: NE -K G CPDMSMP PPLGICKTLC N DSGCPNVQKCCKNGCGFMTCTTPVP

Omwaprin: KDRPKK G CP P PQKP-CVKEC N DS-CPGQQKCCNYGC-KDECRDPIFVG

Nawaprin: NE -K G CPDMSMP PPLGICKTLC N DSGCPNVQKCCKNGCGFMTCTTPVP

Omwaprin: KDRPKK G CP P PQKP-CVKEC N DS-CPGQQKCCNYGC-KDECRDPIFVG

A

B

Figure 1.11 Waprins (A) Amino acid sequence alignment of members of waprins

family, nawaprin (sp: P60589) and omwaprin (sp: P83768) The complete amino acid sequence was determined by Edman degradation sequencing Intercysteine loops are

indicated (B) Nawaprin structure obtained from NMR spectroscopy (PDB: 1UDK) The

four disulfide bridges are highlighted in yellow

Chapter I Literature Review

beneficial or ‘economical’ for use as stable molecular scaffolds to incorporate functional motifs on the surface to exert various pharmacological actions (Fry 2005)

1.4 CONTRIBUTIONS OF SNAKE VENOM RESEARCH

With the efforts of many scientists from the field of protein chemistry, molecular biology, pharmacology, physiology, immunology, herpetology and structural biology, exciting progress has been made in snake venom research in the recent decades Hundreds of snake venom peptides and proteins have been identified, purified and characterized both functionally and structurally These studies have not only contributed to the understanding of snake venom toxicity and clinical management of snakebite, but also provided numerous opportunities for basic research and for use in clinical applications

1.4.1 Source of enzymes

Snake venoms have been identified as the richest source of enzymes among venomous animals (Tan and Ponnudurai 1992) Many purified snake venom enzymes are sold commercially

• L-amino acid oxidase, for example, has been widely used for identifying optical isomers of amino acids and for preparing α-keto acid;

• phosphodiesterase, is utilized for structural studies of nucleic acids and dinucleotide coenzymes (Laskowski 1966);

• phospholipase, is employed for lipid research (Van Deenen and De Haas 1966); and

• proteinases, such as the metalloproteinase (ecarin) is used as a prothrombin activator and serine proteinases (ancrod, atroxin, Reptilase®, crotalase, thrombocytin, Protac®, batroxobin, noscarin) are also available commercially for use as anticoagulants (Iwanaga and Suzuki 1979; Tu 1996)

1.4.2 Tools for basic research

Venomous snakes have evolved a vast array of toxins for prey capture and defense These peptides are directed against a wide variety of pharmacological targets, thus making them

an invaluable source of ligands for studying the properties of these targets

α-bungarotoxin isolated from Bungarus multicinctus, for example, is one of the most

famous three-finger toxins studied so far In contrast to curare which can be easily removed, α-bungarotoxin is specifically and firmly bound to the receptors in the nerve terminal membrane This had finally led to the complete elucidation of the structure, function and mechanism of the nicotinic acetylcholine receptor (nAChR) (Chang 1979; Mebs 1989)

1.4.3 Therapeutic and medical diagnostic uses

Use of snake venoms to cure a variety of diseases was widely recommended in ancient scripts, such as those from India (Mebs 1989) To date, many of the proteins found in the venoms have been developed for therapeutic and diagnostic applications The basis for the uses is mainly due to high specificity and potency of snake venom proteins against specific molecular targets of preys Thus, the uses of venom proteins are generally associated with limited potentialside effects (Olivera et al 2002). Therapeutic uses of

Chapter I Literature Review

snake venoms include applications as antihypertensive agents, anti-stroke medication, anticoagulants, anti-haemorrhagic agents, anti-angiogenic agents and neurotropic agents

1.4.3.1 Therapeutic uses

Captopril is an excellent example of antihypertensive agents that has found widespread clinical use The development of this novel drug was made in 1967 when pentapeptide

BPP5a derived from snake venom BPPs of the Brazilian snake Bothrops jararaca, were

purified and studied Following this, the nonapeptide BPP9a (also known as teprotide) was developed and shown to be an excellent antihypertensive in both animal and human models via angiotensin-converting enzyme inhibition However, it had the drawback of

requiring intravenous admistration (Cushman et al 1979; Cushman et al 1980) With

increased knowledge and understanding of the chemical and enzymatic properties of angiotensin-converting enzyme from the inhibition mechanism by teprotide, it finally led

to the development of captoril Captoril is a simple non-peptide molecule that would

interact with great affinity at the active site of this enzyme (Cushman et al 1987)

Ancrod (Arvin), a thrombin-like serine proteinase isolated from Calloselasma

rhodostoma (Malayan pit viper) venom, has been investigated extensively for its

anticoagulant action In vivo experiments indicate that it decreases or eliminates

fibrinogen, hence preventing thrombus formation resulting in decreased blood viscosity (Bell 1997) These unique properties are extensively used in the treatment of a patient who has suffered from myocardial infarction and in the prevention of thromboembolic

disease (Kornalik 1991; Tu 1996) Similarly, batroxobin from the venom of Bothrops

atrox moojeni, is able to cleave fibrinogen (Bell 1997; Kornalik 1991) and is widely used

in the early stages of stroke (Samsa et al 2002)

Snake venoms also contain proteins with anti-haemorrhagic properties These include textilinins 1 and 2 (txlns 1 and 2) which show 45 and 43 % identities,

respectively, with aprotinin, a plasmin inhibitor (Masci et al 2000) Txln 1 and 2 are

novel antifibrinolytic serine protease inhibitors isolated from common brown snake

(Pseudonaja textilis) venom (Filippovich 2002) Recombinant textilins are now in

preclinical development as novel anti-bleeding drugs for use in open-heart surgery, obviating the need for reliance on aprotinin which is a bovine product with possible risk

of transmissible diseases (Filippovich 2002; Lewis and Garcia 2003)

In the recent years, snake venom disintegrins, such as the triflavin, accutin,

salmosin and contortrostatin (CN), are found to exhibit anti-angiogenic activity (Kang et

al 1999; Sheu et al 1997; Yeh et al 1998; Zhou et al 2000a; Zhou et al 2000b) Unlike

other disintegrins which are homomeric, CN is a 13.5 kDa homodimeric disulfide-linked protein Because of its homodimeric arrangement, CN possesses two Arg-Gly-Asp sites, which appear to be important for its unique interaction with integrins on tumor cells and angiogenic vascular endothelial cells CN administration not only effectively limits tumor

growth and angiogenesis but also severely curtails tumor metastasis (Swenson et al 2004; Zhou et al 2000a; Zhou et al 2000b)