Some myotoxic or neurotoxic PLA2s (phospholipases A2) from pit viper venoms contain characteristic N6 substitutions. Our survey of the venoms of more than ten pit viper genera revealed that N6-PLA2s exist only in limited Asian pit vipers of two genera, Protobothrops and Gloydius, and exist as either monomers or the basic subunits of heterodimers in some New World pit vipers. For the newly identified N6-PLA2s, the neuromuscular blocking activities were assayed with the chick biventer cervicis neuromuscular tissue, whereas the increased serum creatine kinase level assessed their myotoxicities. The purified N6-PLA2s from Protobothrops mangshanensis and Gloydius intermedius saxatilis were found to be presynaptic neurotoxins. In contrast, all N6-PLA2s from the venoms of Sistrurus miliarius strackeri, S. m. barbouri, Crotalus viridis viridis, C. lepidus lepidus, Cerrophidion godmani and Bothreichis schlegelii were myotoxins without neurotoxicity even in the presence of crotoxin A. Crotoxin-like complexes were for the first time purified from the venoms of Sitrurus catenatus tergeminus, C. mitchelli mitchelli, C. horridus atricaudatus, C. basiliscus and C. durissus cumanensis. The cDNAs encoding six novel N6-PLA2s and subunits of the crotoxin-like complex from S. c. tergeminus were cloned and fully sequenced. Phylogeny analysis showed that two structural subtypes of N6-PLA2s with either F24 or S24 substitution have been evolved in parallel, possibly descended respectively from species related to present-day Protobothrops and Gloydius. Calmodulin binds all the N6-PLA2s but crotoxin A may inhibit its binding to crotoxin B and to other neurotoxic N6-PLA2s. Structure–activity relationships at various regions of the PLA2 molecules were extensively discussed.

INTRODUCTION

Owing to the frequent snakebites by the Asian pit vipers and the American rattlesnakes, studies of the related venoms are medically important. On the basis of the snakebite symptom, two types of rattlesnake venoms appear to exist today, one of them being neurotoxic and highly lethal and the other is haemorrhagic and myotoxic but less lethal [1]. As shown previously, potent neurotoxins from rattlesnake venoms (e.g. crotoxin and mojave toxin) are heterodimeric complexes of PLA2s (phospholipases A2; EC 3.1.1.4) [2,3]. CA (crotoxin A) and CB (crotoxin B), the acidic and basic subunits of crotoxin, are present in the venom of most South American rattlers. Immunochemical studies have been performed to detect this type of neurotoxin in many of the rattlers, using the antiserum against crotoxin or mojave toxin [4,5]. It has been shown that a CA-like subunit is absent in non-neurotoxic venoms of rattlers [6], whereas myotoxins from the pit viper venoms are usually basic monomeric PLA2s with either D49 or K49 substitution [7].

Rattlesnakes (Crotalus and Sistrurus) and other pit vipers of the New World are probably the descendants of some Asian pit vipers [8]. The group II PLA2s in pit viper venoms have been diversified into distinct pharmacological subtypes, and play different roles such as platelet aggregation inhibitor, anticoagulant, neurotoxin or myotoxin [9,10]. One of the subtypes is the myotoxic/neurotoxic D49-PLA2 with N6 substitution (hereafter designated as N6-PLA2). The known members of this venom PLA2 subtype include the basic subunits of crotoxin and mojave toxin, agkistrodotoxin from Gloydius halys brevicaudus (formerly Agkistrodon halys Pallas) [11], trimucrotoxin from Protobothrops mucrosquamatus [12], Tf-PLA-N from P. flavoviridis [13] and the myotoxin from Crotalus viridis viridis (designated as Cvv-N6) [14,15]. In the present study, we purified and sequenced six novel N6-PLA2s from the New World pit viper venoms, and identified several N6-PLA2 neurotoxins from the Asian pit viper venoms. ESI–MS (electrospray ionization–mass spectrometry), N-terminal sequencing and cloning facilitated the structure determination of these myotoxic or neurotoxic N6-PLA2s and their functions were characterized using experimental animals. The evolutionary relationships between these N6-PLA2s were investigated by phylogenetic analysis based on their sequences. Furthermore, the binding specificities of various N6-PLA2s to calmodulin (one of the receptor candidates) [16] have been tested.

EXPERIMENTAL

Venoms and other materials

Freeze-dried venom samples of Gloydius intermedius saxatilis (formerly Agkistrodon intermedius saxatilis), Cerrophidion godmani, Bothreichis schlegelii, Sistrurus miliarius barbouri, Crotalus lepidus lepidus, C. basiliscus, Lachesis muta muta and Atropoides nummifer were purchased from Miami Serpentarium laboratory (FL, U.S.A.). P. flavoviridis, P. tokarensis, P. elegans and P. mucrosquamatus were from Latoxan (Valence, France). Venoms of C. durissus cumanensis, C. d. culminatus, C. horridus horridus, C. mitchelli mitchelli, C. vegrandis, C. molossus molossus and C. v. viridis were purchased from Kentucky Reptile Zoo (Slade, KY, U.S.A.). The venoms of P. mangshanensis and C. v. oreganus were gifts from Professor Y.-Y. Shu (Kuangxi Medical University, People's Republic of China) and Professor S. P. Mackessy (Northern Colorado State University, U.S.A.) respectively. The venom of C. horridus (neurotoxic or type A, from South Carolina) was a gift from the Kentucky Reptile Zoo. C. atrox and C. adamanatus venoms were from Sigma (St. Louis, MO, U.S.A.). Live specimens of B. schlegelii, Ce. godmani, C. v. viridis, Sitrurus catenatus tergeminus and S. miliarius strackeri were purchased from Glades Herp. (Fort Myers, FL, U.S.A.). The venoms were extracted 2 days before the snake was killed and the venom glands were removed immediately for RNA extraction.

The mRNA preparation and the cDNA synthesis kits were purchased from Stratagene. Restriction enzymes and other enzymes were purchased from Promega. Synthetic DPPC (L-dipalmitoyl phosphatidylcholine) was purchased from Avanti Polar Lipids (Alabaster, AL, U.S.A.). Other chemicals were of reagent grade.

Purification and characterization of purified PLA2s

Venom (10–15 mg) was dissolved and fractionated on an FPLC system with a Superdex 75 column (HR10/30; Amersham Biosciences) in 0.1 M ammonium acetate (pH 6.5) at room temperature (25 °C). After freeze drying, the PLA2-containing fractions were further purified by reverse-phase HPLC using a C8 column (4.5 mm×250 mm, 10 μ, Vydac, Hesperia, CA, U.S.A.) equilibrated with 0.07% aqueous trifluoroacetic acid (solvent A), and eluted with a 25–45% linear gradient of CH3CN containing 0.07% trifluoroacetic acid (solvent B). The purified PLA2s were dried in a vacuum-centrifuge device (Labconco, Kansas city, MO, U.S.A.).

The amino acid sequences of purified PLA2s were determined by an automated amino acid sequencer (Model 477A; PE Applied Biosystems, Foster City, CA, U.S.A.). The molecular mass of each PLA2 was determined by ESI–MS (Sciex mass analyser, API100; PerkinElmer) [17].

Cloning and sequencing

The venom gland mRNA and the cDNA were prepared as described previously [15,17]. To amplify cDNA of PLA2s using the venom gland cDNA as a template, PCR [18] was conducted using SuperTaq DNA Polymerase with a pair of mixed-base oligonucleotide primers (sense primer 1: TCTGGATTSAGGAGGATGAGG and antisense primer 2: GCCTGCAGRACTTAGCA), which were designed according to the highly conserved cDNA regions of the group-II PLA2s from snake venoms [12,15]. In addition, an antisense primer 3, AAYCTGYTBCARTTYAAYAAAATGATCAAG, was designed according to the N-terminal sequences of the N6-PLA2s from Ce. godmani and S. m. strackeri, and was used along with primer 1 in PCR.

After treatment with polynucleotide kinase, the amplified DNA fragment was inserted into the pGEM-T easy vector (Promega Biotech, Madison, WI, U.S.A.). It was transformed into Escherichia coli strain JM 109. White transformants were picked up and cDNA clones were selected. The DNA sequencing System model 373A and the Taq-Dye-Dioxy terminator cycle sequencing kit (Applied Biosystems, Foster City, CA, U.S.A.) were used to determine the cDNA sequences by dideoxynucleotide method [19].

Enzymic activity and toxicity assay

PLA2 concentration was determined by measuring the absorbance at 280 nm, assuming a molar absorption coefficient of 1.5 at 1.0 mg/ml protein. PLA2 activity was measured using the micellar substrates of DPPC (3 mM) mixed with an equal molar of deoxycholate or 2-fold concentration of Triton X-100 on a pH-stat apparatus (Radiometer RTS 822, Copenhagen, Denmark). The initial reaction rate was followed for 5 min and corrected for non-enzymic spontaneous rate.

For neurotoxicity assay, chick biventer cervicis nerve-muscle preparations [20] were isolated from 4–10-day-old chicks. The tissue was suspended in an organ bath filled with 10 ml of modified Tyrode solution with 2 mM CaCl2 at 37 °C and aerated with 95% O2 plus 5% CO2. Contractions of the muscle were elicited by stimulation of the nerve with supramaximal rectangular pulses of 0.05–0.1 ms duration on a Grass polygraph [12].

The myotoxic effect of N6-PLA2s was studied by measuring the serum creatine kinase concentrations after intramuscular injection. Blood (20–40 μl) was drawn by a glass capillary pipette containing EDTA from the retro-orbital sinus of mouse eye 3 h after intramuscular injection of the purified PLA2 [21]. To assay the serum level of the kinase, a reagent kit from Sigma (Cat. no. 520) was used.

Phylogenetic analysis of crotalid basic D49 PLA2s

In addition to the full amino acid sequences of six N6-PLA2s solved in the present study, other N6-PLA2s were selected by BLASTp search (National Center for Biotechnology Information). The amino acid sequence alignment was made by the PILEUP program. Cladograms were constructed based on these sequences by neighbour-joining algorithm using the PHYLIP program [22], and the degree of confidence for the internal linage was determined by bootstrap methods [23].

Expression of recombinant GST (glutathione S-transferase)–calmodulin

The open reading frame of the cDNA of rat calmodulin (accession no. NM-017326) was amplified by PCR, using the primer pair that contains the flanking restriction sites of BamHI and XhoI respectively, i.e. BamHI-sense primer GGATCMTGGCTGACCAACTGACTG and XhoI-antisense primer CTCGASTACTTCGCTGTCATCATTTG. Marathon-ready cDNAs of rat brain (ClonTech) were used as templates in the PCR to generate full-length cDNA. The PCR products were cleaved with the restriction enzymes, purified and then ligated to pGEX-4T-1 expression vector for the production of GST fusion proteins. The resultant plasmid was transformed into E. coli strain BL21 after sequence verification. The recombinant GST–calmodulin was purified on the glutathione–Sepharose according to the manufacturer's instructions (Amersham Biosciences).

Binding to calmodulin

To detect the binding between various N6-PLA2s and calmodulin [24], two approaches were adopted. One used calmodulin-immobilized Sepharose 4B (Amersham Biosciences) to pull down the toxins directly; the other used glutathione–Sepharose 4B to pull down the reaction mixture of GST–calmodulin and the PLA2s. The binding buffer used was 10 mM Tris/HCl (pH 8.0) containing 0.15 M NaCl and 2 mM CaCl2. The final concentrations of PLA2s and GST–calmodulin applied in the assay were 1.0 and 1.5 μM respectively, whereas the affinity resin (50% slurry) was 20 μl. The binding reaction was conducted at 4 °C for 2 h with constant mixing, and the matrix was washed three times each with 1 ml of binding buffer, by centrifugation at 1500 g for 2 min. The pellet was then boiled in SDS sample buffer; the bound material was resolved by SDS/PAGE (12% gel) followed by Coomassie Blue staining.

RESULTS

Purification and identification of the N6-PLA2s

Gel filtration was used to isolate the PLA2-containing fractions from crude venom of more than 20 species. The 11–14 and 22–27 kDa fractions of the venoms (Figure 1) presumably contained monomeric and dimeric PLA2s respectively. The proteins were further purified by HPLC (Figure 2). The molecular mass of each purified protein peak was determined by ESI–MS and each protein of 13.9±0.4 kDa was subjected to N-terminal sequencing and enzyme assay. Our survey revealed that N6-PLA2s are absent from the venoms of the Old World pit viper genera Calloselasma, Hypnale, Tropidolaemus, Ovophis and Trimeresurus (sensu stricto) and those of the New World Agkistrodon, Lachesis, Atropoides and Bothrops.

Representative gel filtration profiles of pit viper venoms

Figure 1
Representative gel filtration profiles of pit viper venoms

Dissolved crude venom was loaded on to a Superdex 75 (HR10/30) column and eluted with 0.1 M ammonium acetate (pH 6.5) at a flow rate of 1.0 ml/min. Fractions containing the dimeric (upper three panels) or the monomeric (lower three panels) PLA2s were pooled (shown by —).

Figure 1
Representative gel filtration profiles of pit viper venoms

Dissolved crude venom was loaded on to a Superdex 75 (HR10/30) column and eluted with 0.1 M ammonium acetate (pH 6.5) at a flow rate of 1.0 ml/min. Fractions containing the dimeric (upper three panels) or the monomeric (lower three panels) PLA2s were pooled (shown by —).

Purification of N6-PLA2s by reversed-phase HPLC

Figure 2
Purification of N6-PLA2s by reversed-phase HPLC

Each PLA2-containing fraction from Figure 1 was freeze-dried, re-dissolved and further purified on a Vydac-C8 HPLC column with a gradient of solvent B (acetonitrile with 0.07% trifluoroacetic acid). *, N6-PLA2 peak; A, CA-like peak.

Figure 2
Purification of N6-PLA2s by reversed-phase HPLC

Each PLA2-containing fraction from Figure 1 was freeze-dried, re-dissolved and further purified on a Vydac-C8 HPLC column with a gradient of solvent B (acetonitrile with 0.07% trifluoroacetic acid). *, N6-PLA2 peak; A, CA-like peak.

We found that N6-PLA2s are present only in the venom of the Asian pit vipers under two genera, Protobothrops and Gloydius. N6-PLA2s with identical masses and sequences could be isolated from the venoms of P. tokarensis (i.e. Pto-N6) and P. flavoviridis (i.e. Tf-PLA-N) and from those of P. mangshanensis and P. mucrosquamatus (Table 1). However, we did not find N6-PLA2 in P. elegans venom. Table 2 lists the molecular masses and N-terminal sequences of the purified N6-PLA2s from the venoms of the New World pit vipers. Crotoxin-like heterodimers were found in the venoms of C. d. terrificus, C. scutulatus scutulatus, C. d. cumanensis, C. vegrandis, C. m. mitchelli, C. basiliscus basiliscus, C. horridus (type A) and S. c. tergeminus. On the other hand, monomeric N6-PLA2s were identified in the venoms of B. schlegelii, Ce. godmani, C. l. lepidus and C. v. viridis, and were abbreviated as Bs-N6, Ceg-N6 etc. respectively (Tables 2 and 4).

Table 1
List of N6-PLA2s purified from the Asian pit viper venoms

Molecular masses of PLA2s were determined by ESI–MS and the N-terminal sequences were determined by the amino acid sequencer. Residues identical with the first line were shown by dots, special substitutions are shown in bold.

Toxin or PLA2Venom speciesMass (kDa)N-terminal sequences 1–23
Trimucrotoxin P. mucrosquamatus 13.902 NLLQFNKMIKIMTKKNAIPFYSS 
Mangshantoxin P. mangshanensis 13.902 ······················· 
Tf-PLA-N P. flavoviridis 14.033 ················GF···T· 
Pto-N6 P. tokarensis 14.033 ················GF···T· 
DAV-N6 D. acutus 14.000 H············R···F···T· 
Agkistrodotoxin G. b. brevicaudus 13.938 ··········EE·G·······AF 
Intermexin G. i. saxatilis 14.145 ··········VE·G·······AF 
Toxin or PLA2Venom speciesMass (kDa)N-terminal sequences 1–23
Trimucrotoxin P. mucrosquamatus 13.902 NLLQFNKMIKIMTKKNAIPFYSS 
Mangshantoxin P. mangshanensis 13.902 ······················· 
Tf-PLA-N P. flavoviridis 14.033 ················GF···T· 
Pto-N6 P. tokarensis 14.033 ················GF···T· 
DAV-N6 D. acutus 14.000 H············R···F···T· 
Agkistrodotoxin G. b. brevicaudus 13.938 ··········EE·G·······AF 
Intermexin G. i. saxatilis 14.145 ··········VE·G·······AF 
Table 2
List of N6-PLA2s purified from the New World pit viper venoms

Other details are the same as given in the legend of Table 1.

PLA2 or subunitVenom speciesMass (kDa)N-terminal sequences 1–23
CB1 C. d. terrificus 14.183±0.003 HLLQFNKMIKFETRKNAIPFYAF 
 C. basiliscus   
 C. vegrandis  
 C. s. scutulatus   
 C. d. cumanensis   
CB2 C. d. terrificus 14.244±0.002 S······················ 
 C. d. cumanensis   
 C. m. mitchelli   
CB3 C. d. terrificus (Paraguay) 14.038 ·············G········· 
Canebraxin B C. h. atricaudatus 14.154 ······················· 
Sistruxin B S. c. tergeminus 14.122 N············NA········ 
Monomeric C. v. viridis 14.200 N·········M··K···F····S 
myotoxin C. l. lepidus 13.861 N·········IM·K·······TS 
 S. m. barbouri 13.948 N·········IM·K·······TS 
 S. m. strackeri 13.893 N·········IM·K·····S·TS 
 Ce. godmani 14.012 N·········IM·K···V···TS 
 B. schlegelii 13.785 N·········IM····G··Y·SS 
PLA2 or subunitVenom speciesMass (kDa)N-terminal sequences 1–23
CB1 C. d. terrificus 14.183±0.003 HLLQFNKMIKFETRKNAIPFYAF 
 C. basiliscus   
 C. vegrandis  
 C. s. scutulatus   
 C. d. cumanensis   
CB2 C. d. terrificus 14.244±0.002 S······················ 
 C. d. cumanensis   
 C. m. mitchelli   
CB3 C. d. terrificus (Paraguay) 14.038 ·············G········· 
Canebraxin B C. h. atricaudatus 14.154 ······················· 
Sistruxin B S. c. tergeminus 14.122 N············NA········ 
Monomeric C. v. viridis 14.200 N·········M··K···F····S 
myotoxin C. l. lepidus 13.861 N·········IM·K·······TS 
 S. m. barbouri 13.948 N·········IM·K·······TS 
 S. m. strackeri 13.893 N·········IM·K·····S·TS 
 Ce. godmani 14.012 N·········IM·K···V···TS 
 B. schlegelii 13.785 N·········IM····G··Y·SS 

CA and CB were purified from C. d. terrificus and related rattlesnakes (Figure 3A). Sistruxins A and B as well as a major acidic PLA2 (i.e. E6) were purified from the venom of an individual S. c. tergeminus (Figure 3B). Moreover, the reconstituted sistruxin complex was found to be as potent as the native heterodimer purified by monoQ column (results not shown).

Purification of crotoxin and sistruxin subunits and neurotoxicity assay of the sistruxin complex

Figure 3
Purification of crotoxin and sistruxin subunits and neurotoxicity assay of the sistruxin complex

(A) Gel filtration of the crude venom by a Superdex 75 column. The neurotoxin fractions (shown by —) were pooled. (B) CA and CB subunits of crotoxin and SA and SB subunits of sistruxin were purified by reversed-phase HPLC on a Vydac-C8 column. The elution conditions were the same as in Figures 1 and 2. (C) Neurotoxicity assay of the complex of sistruxins A and B (0.1 μg/ml) using chick biventer cervicis neuromuscular tissue. Acetylcholine (Ach) was added to a final concentration of 10 μg/ml.

Figure 3
Purification of crotoxin and sistruxin subunits and neurotoxicity assay of the sistruxin complex

(A) Gel filtration of the crude venom by a Superdex 75 column. The neurotoxin fractions (shown by —) were pooled. (B) CA and CB subunits of crotoxin and SA and SB subunits of sistruxin were purified by reversed-phase HPLC on a Vydac-C8 column. The elution conditions were the same as in Figures 1 and 2. (C) Neurotoxicity assay of the complex of sistruxins A and B (0.1 μg/ml) using chick biventer cervicis neuromuscular tissue. Acetylcholine (Ach) was added to a final concentration of 10 μg/ml.

Assay and functional characterization

The enzymic activities of the purified N6-PLA2s towards cationic and zwitter-ionic micellar substrates were studied (Tables 3 and 4). All the S24-N6-PLA2s have similar catalytic activities, which are lower than those of various F24-N6-PLA2s. The activities of S24-N6-PLA2s were especially low when Triton X-100 micelles were the substrates. Effects of CA on the catalytic activities were also investigated.

Table 3
Enzymic activities of F24-N6PLA2s in the presence and absence of CA and their protein contents in the venom

Enzymes were assayed in the presence of 10 mM CaCl2 at 37 °C, pH 7.4. The concentration of added CA was eight times higher than that of N6PLA2.

Specific activity (μmol·mg−1·min−1)
PLA2 or subunitContent (%, w/w)DPPC+deoxycholateDPPC+Triton X-100
Intermexin 13 392±15 222±6 
Intermexin+CA – 359±15 106±2 
CB1 of various rattlesnakes 4–14 652±32 226±24 
CB2 of C. m. mitchelli 30 430±4 123±6 
CB3 of C. d. terrificus 20 672±44 231±20 
Crotoxin complex 35–50 388±25 58±2 
Canebraxin B 480±13 195±16 
Canebraxin complex 11 355±15 63±2 
Sistruxin B 385±38 192±6 
Specific activity (μmol·mg−1·min−1)
PLA2 or subunitContent (%, w/w)DPPC+deoxycholateDPPC+Triton X-100
Intermexin 13 392±15 222±6 
Intermexin+CA – 359±15 106±2 
CB1 of various rattlesnakes 4–14 652±32 226±24 
CB2 of C. m. mitchelli 30 430±4 123±6 
CB3 of C. d. terrificus 20 672±44 231±20 
Crotoxin complex 35–50 388±25 58±2 
Canebraxin B 480±13 195±16 
Canebraxin complex 11 355±15 63±2 
Sistruxin B 385±38 192±6 
Table 4
Enzymic activities of S24-N6PLA2s in the presence and absence of CA and their protein contents in the venom

Legend is same as in Table 3.

Specific activity (μmol·mg−1·min−1)
PLA2 toxin or subunitContent (%, w/w)DPPC+deoxycholateDPPC+Triton X-100
Trimucrotoxin 10 300±15 97±1 
Trimucrotoxin+CA – 280±25 45±4 
Mangshantoxin 521±15 155±4 
Mangshantoxin+CA – 468±11 15±2 
Cvv-N6 1–4 280±8 88±4 
Cvv-N6+CA – 246±1 86±5 
Cll-N6 294±8 7±1 
Cll-N6+CA – 279±18 6±1 
Ceg-N6 251±15 <1 
Ceg-N6+CA – 248±17 <1 
Bs-N6 1–2 208±7 116±2 
Smb-N6 324±13 <1 
Specific activity (μmol·mg−1·min−1)
PLA2 toxin or subunitContent (%, w/w)DPPC+deoxycholateDPPC+Triton X-100
Trimucrotoxin 10 300±15 97±1 
Trimucrotoxin+CA – 280±25 45±4 
Mangshantoxin 521±15 155±4 
Mangshantoxin+CA – 468±11 15±2 
Cvv-N6 1–4 280±8 88±4 
Cvv-N6+CA – 246±1 86±5 
Cll-N6 294±8 7±1 
Cll-N6+CA – 279±18 6±1 
Ceg-N6 251±15 <1 
Ceg-N6+CA – 248±17 <1 
Bs-N6 1–2 208±7 116±2 
Smb-N6 324±13 <1 

Chick biventer cervicis neuromuscular tissue has been shown to be a very sensitive target for crotoxin in electrophysiological test [25,26] and therefore was used to study the neurotoxicity of the newly found sistruxin (Figure 3C) and other N6-PLA2s. Neurotoxicity of the CB subunit was more than 30 times weaker than the intact crotoxin-like complex towards the chick tissue (Table 5).

Table 5
Neurotoxicity of N6-PLA2 towards the chick biventer cervicis tissue

The numbers in parentheses represent the number of experiments.

N6-PLA2 toxinDose (μg/ml)Time of 90% inhibition of the twitch (min)
Trimucrotoxin 2.0 38±2 (n=4) 
 1.0 76±13 (n=5) 
 0.3 106±2 (n=3) 
Mangshantoxin 1.0 77±3 (n=2) 
Agkistrodotoxin 0.3 75±3 (n=2) 
Intermexin 2.0 80±5 (n=2) 
 1.0 150±5 (n=3) 
CB subunit 2.0 80±20 (n=2 for each venom species) 
 1.0 170±30 (n=2 for each venom species) 
Canebraxin B 2.0 63±3 (n=2) 
 1.0 163±1 (n=2) 
Sistruxin B 0.3 230 (n=1) 
Crotoxin complex 0.10 65±2 (n=2) 
 0.05 95±7 (n=4) 
 0.03 150±10 (n=2) 
Canebraxin complex 0.1 83±7 (n=3) 
Sistruxin complex 0.3 50±3 (n=2) 
 0.1 76±4 (n=2) 
Cvv-N6±CA 4.0 >240 (n=2) 
N6-PLA2 toxinDose (μg/ml)Time of 90% inhibition of the twitch (min)
Trimucrotoxin 2.0 38±2 (n=4) 
 1.0 76±13 (n=5) 
 0.3 106±2 (n=3) 
Mangshantoxin 1.0 77±3 (n=2) 
Agkistrodotoxin 0.3 75±3 (n=2) 
Intermexin 2.0 80±5 (n=2) 
 1.0 150±5 (n=3) 
CB subunit 2.0 80±20 (n=2 for each venom species) 
 1.0 170±30 (n=2 for each venom species) 
Canebraxin B 2.0 63±3 (n=2) 
 1.0 163±1 (n=2) 
Sistruxin B 0.3 230 (n=1) 
Crotoxin complex 0.10 65±2 (n=2) 
 0.05 95±7 (n=4) 
 0.03 150±10 (n=2) 
Canebraxin complex 0.1 83±7 (n=3) 
Sistruxin complex 0.3 50±3 (n=2) 
 0.1 76±4 (n=2) 
Cvv-N6±CA 4.0 >240 (n=2) 

Myotoxicity of the N6-PLA2 towards mice was studied. Table 6 listed the serum level of creatine kinase after intramuscular injection of N6-PLA2s into the extensor digitorum longus muscle of mouse legs.

Table 6
Comparison of myotoxic effects of N6-PLA2s

The creatine kinase levels in mouse serum were measured 2 or 3 h after intramuscular injection of 16 μg of venom PLA2 in sterile PBS. Definition of unit followed that of the Sigma kit and values are the means±S.D. for duplicate experiments.

Creatine kinase (units/ml)
PLA22 h3 h
Trimucrotoxin 4.73±0.01 5.35±0.20 
Intermexin 1.91±1.11 2.70±0.95 
Cvv-N6 5.83±1.66 7.45±0.98 
PBS (control) 0.26±0.21 0.38±0.01 
Creatine kinase (units/ml)
PLA22 h3 h
Trimucrotoxin 4.73±0.01 5.35±0.20 
Intermexin 1.91±1.11 2.70±0.95 
Cvv-N6 5.83±1.66 7.45±0.98 
PBS (control) 0.26±0.21 0.38±0.01 

Cloning and complete sequencing of PLA2s

Using specifically designed PCR primer, cDNAs of many PLA2s could be amplified and cloned. The success of this approach further confirms that the venom PLA2 genes of various pit vipers share highly conserved 5′ and 3′ non-coding regions [13]. The complete amino acid sequences of these N6-PLA2s were predicted from the cDNA sequences and matched against the mass and N-terminal sequences of the purified PLA2s (Tables 1 and 2). Alignment of all the N6-PLA2 sequences obtained from the present study and previous works [1113,2730] shows that their sequence identities are approx. 73–94% (Figure 4A).

Alignment of the amino acid sequences of (A) N6-PLA2s or CB-like subunits and (B) the precursor of sistruxin A and related acidic PLA2s

Figure 4
Alignment of the amino acid sequences of (A) N6-PLA2s or CB-like subunits and (B) the precursor of sistruxin A and related acidic PLA2s

Novel sequences deduced in the present study are shown in boldface. Single-letter codes are used, and the numbering system followed that commonly used [30]. Residues identical with those in the top line are denoted with dots and gaps are marked with hyphens. Neurotoxins or their subunits are denoted with asterisks, and special substitutions possibly related to the neurotoxicity of trimucrotoxin are shown in boldface. (B) GenBank® accession numbers for sequences 5–9 are AAR11860, P20249, P20476, Q9I8F8 and P00624 respectively; regions processed away during the maturation of CA-like subunits were blocked. Substitutions possibly important for the chaperone role of CA are marked with asterisks.

Figure 4
Alignment of the amino acid sequences of (A) N6-PLA2s or CB-like subunits and (B) the precursor of sistruxin A and related acidic PLA2s

Novel sequences deduced in the present study are shown in boldface. Single-letter codes are used, and the numbering system followed that commonly used [30]. Residues identical with those in the top line are denoted with dots and gaps are marked with hyphens. Neurotoxins or their subunits are denoted with asterisks, and special substitutions possibly related to the neurotoxicity of trimucrotoxin are shown in boldface. (B) GenBank® accession numbers for sequences 5–9 are AAR11860, P20249, P20476, Q9I8F8 and P00624 respectively; regions processed away during the maturation of CA-like subunits were blocked. Substitutions possibly important for the chaperone role of CA are marked with asterisks.

In addition, the precursor for sistruxin A (the acidic subunit of sistruxin) was also cloned from the venom gland of S. c. tergeminus. Its deduced sequence was aligned with the precursors of other CA-like proteins, as well as other homologous acidic E6-PLA2s obtained by BLASTp (Figure 4B). The sequence of mature sistruxin A differs from that of CA in only five semiconserved amino acids, and appears to have similar functions and specificities. On the other hand, residues 1–32 of the major PLA2 of S. c. tergeminus are NLIQFETLILKVAKKSGMFSYSAYGCYCGWGG/, which differs from that of the acidic PLA2s of Gloydius blomhoffi brevicaudus [28] by only three substitutions.

Phylogenetic tree of the N6-PLA2s and G6-PLA2s

A phylogenetic tree based on the protein sequences of all the N6-PLA2s and the G6 or A6-myotoxic PLA2s [3134] from pit viper venoms is shown in Figure 5 The W6D49-PLA2 of C. rhodostoma venom [17] was used as out-group. The tree shows the clustering of all the N6-PLA2s, separated from the G6D49-PLA2s and the existence of two lineages of N6-PLA2s, denoted as F24 and S24 respectively. The bootstrap values higher than 50 at each node support the robustness of the cladogram.

Molecular phylogeny of basic crotalid D49 PLA2s

Figure 5
Molecular phylogeny of basic crotalid D49 PLA2s

Data used include all the sequences listed in Figure 4(A) and the G6D49-PLA2 sequences from four species. The PLA2, GenBank® accession numbers and the venom species are agkistrodotoxin (P14421), CB (P07517, P24027), trimucrotoxin (×77645), DAV-N6 (×77649) from Deinagkistrodon acutus, Tf-PLA-N (AB102728) from P. flavoviridis, Bpi-G6D49 (P58464) from B. pirajai, Bas-A6D49 (P20474) from B. asper, Bj-D1G6D49 (AY185201), Bj-S1G6D49 (AY145836) from B. jararacussu, and Bi-G6D49 (AF490535) from B. insularis. Sequence data of Sct-N6, sistruxin B, Sms-N6, Cvv-N6, Ceg-N6 and Bs-N6 from Figure 4(A) are shown in boldface.

Figure 5
Molecular phylogeny of basic crotalid D49 PLA2s

Data used include all the sequences listed in Figure 4(A) and the G6D49-PLA2 sequences from four species. The PLA2, GenBank® accession numbers and the venom species are agkistrodotoxin (P14421), CB (P07517, P24027), trimucrotoxin (×77645), DAV-N6 (×77649) from Deinagkistrodon acutus, Tf-PLA-N (AB102728) from P. flavoviridis, Bpi-G6D49 (P58464) from B. pirajai, Bas-A6D49 (P20474) from B. asper, Bj-D1G6D49 (AY185201), Bj-S1G6D49 (AY145836) from B. jararacussu, and Bi-G6D49 (AF490535) from B. insularis. Sequence data of Sct-N6, sistruxin B, Sms-N6, Cvv-N6, Ceg-N6 and Bs-N6 from Figure 4(A) are shown in boldface.

Binding of N6-PLA2s to calmodulin and the effect of CA

Although it was shown that highly neurotoxic PLA2s may bind to calmodulin with high affinity (Kd=10−8–10−9 M) [24], no direct correlation between the toxic potency of PLA2s and their binding affinity for calmodulin was observed in previous experiments. Results of our pull-down assay showed that in vitro calmodulin binds strongly to all the N6-PLA2s in the absence of CA (Figure 6). However, the binding of calmodulin to the intact complex (e.g. crotoxin and canebraxin) was much weaker. Thus the binding is not specific for the neurotoxic PLA2s. We found that CA may specifically protect neurotoxic N6-PLA2s, including CB, sistruxin B and trimucrotoxin, from binding calmodulin; however, no CA protection was found with the myotoxic Cvv-N6 (Figure 6C) and other non-neurotoxic N6-PLA2s (results not shown).

Calmodulin binding of N6-PLA2s and the effect of CA

Figure 6
Calmodulin binding of N6-PLA2s and the effect of CA

(A) Pull-down assay of purified N6-PLA2s (1.5 μM) with calmodulin–Sepharose gel. (B) Pulldown assay by incubating glutathione resin with GST–calmodulin (1.0 μM) and the PLA2 (1.5 μM). Lane M, protein standards; lane N, control without PLA2; lane 1, intact crotoxin; lane 2, CB; lane 3, intact canebraxin; lane 4, canebraxin B; lanes 5–8, Cvv-N6, Smb-N6, Ceg-N6 and Bs-N6 respectively. (C) Pull-down assay with calmodulin–Sepharose, in the presence (lanes 2, 4, 6 and 8) and absence (lanes 1, 3, 5 and 7) of CA. Lanes 1 and 2, CB; lanes 3 and 4, Cvv-N6; lanes 5 and 6, canebraxin B; lanes 7 and 8, trimucrotoxin.

Figure 6
Calmodulin binding of N6-PLA2s and the effect of CA

(A) Pull-down assay of purified N6-PLA2s (1.5 μM) with calmodulin–Sepharose gel. (B) Pulldown assay by incubating glutathione resin with GST–calmodulin (1.0 μM) and the PLA2 (1.5 μM). Lane M, protein standards; lane N, control without PLA2; lane 1, intact crotoxin; lane 2, CB; lane 3, intact canebraxin; lane 4, canebraxin B; lanes 5–8, Cvv-N6, Smb-N6, Ceg-N6 and Bs-N6 respectively. (C) Pull-down assay with calmodulin–Sepharose, in the presence (lanes 2, 4, 6 and 8) and absence (lanes 1, 3, 5 and 7) of CA. Lanes 1 and 2, CB; lanes 3 and 4, Cvv-N6; lanes 5 and 6, canebraxin B; lanes 7 and 8, trimucrotoxin.

DISCUSSION

Phylogeography and evolution of venom N6-PLA2s

Originally, the name of crotoxins has been designated to the neurotoxic PLA2 complex from the venoms of the South American rattlesnakes including C. d. terrificus, C. d. durissus, C. d. cumanensis [3,4], C. d. cascavella and C. d. collilineatus [35], whereas mojave toxin was named for the neurotoxin in the North American Mojave rattlesnake (C. s. scutulatus) [6,7] (Table 2). As shown in various reports, crotoxin homologues are also present in the venom of several other rattlesnakes, e.g. C. v. concolor, C. vegrandis [36] and canebrake or timber rattlesnake (C. horridus) [37]. Depending on the content of these neurotoxins, the venoms show high or moderate lethal potency with LD50 values of 0.2–0.6 μg/g, which are approx. ten times lower than non-neurotoxic rattlers [1,2]. We here surveyed the presence of N6-PLA2s in all pit viper venoms available, and purified for the first time the crotoxin-like toxins in the venoms of S. c. tergeminus, C. basiliscus and C. m. mitchelli (Table 2).

Two isoforms of CB with molecular masses of 14.180 kDa (CB1) and 14.244 kDa (CB2) were previously identified in the pooled venom of C. d. terrificus [3,38]. We found that CB isoforms purified from various Crotalus venoms are highly conserved. Apparently, CB1 and CB2 were expressed in the venom differentially, and C. m. mitchelli venom expressed only CB2 (Table 2), which possibly explains the lower lethality of this species towards mice when compared with C. durissus containing more CB1 [2,3]. Some other CB variants (e.g. CB3; Table 2) were, however, identified in special geographic samples of C. d. terrificus [39,40]. The basic subunit of canebraxin (from neurotoxic population of C. horridus) has its N-terminal sequence identical with that of CB1 but has a distinct mass of 14.154 kDa. Although the toxicities of sistruxin and canebraxin are as potent as crotoxin (Table 5), their contents in the venom are relatively low (5–12%). Thus envenomings by Pigmy rattlers or Timber rattlers usually do not show neurotoxic symptoms like those by Mojave or neotropical rattlers, which contain crotoxin up to 50% of the total venom content [35,39].

Apparently, the sequences of N6-PLA2 subunits of crotoxins are similar to the sequence of agkistrodotoxin from the Asian Gloydius, and the sequences of CA are similar to those of the E6-PLA2s from Gloydius venoms (Figure 4B). In contrast, the sequences of the monomeric N6-PLA2s from the New World pit vipers are closely related to that of the trimucrotoxin. Thus ancient pit vipers related to the present-day Gloydius and Protobothrops possibly migrated to the New World and evolved into various rattlesnakes. The evolutionary relationships of two lineages of N6-PLA2s (i.e. F24 versus S24) are also supported by the cladogram (Figure 5). Most of the crotalid species express only one of the subtypes, but both subtypes of N6-PLA2s have been cloned from S. c. tergeminus, although only the F24 subtype is expressed in detectable quantity. The origin of S. c. tergeminus remains puzzling as long as the Asiatic pit viper containing both N6-PLA2 subtypes has not been discovered.

CA does not recognize non-neurotoxic N6-PLA2s

CA has been known to form a complex with CB with high affinity (Kd=10−8–10−9 M) [38], and prevent non-specific binding of CB [2527]. CA may increase the toxicity of CB more than 30-fold and those of trimucrotoxin and agkistrodotoxin severalfold towards mouse or chick tissue [12,27]. CA also inhibited the in vitro enzymic activities of CB isoforms, and the inhibition was weak in the presence of deoxycholate because the anionic detergent interfered with the binding between CA and CB [41]. However, the inhibition was prominent with the Triton X-100 micellar substrate (Table 3). Notably, the esterase activities of the basic subunits of canebraxin and sistruxin were inhibited by the CA-like subunit significantly, as also reported for the case of crotoxin complex [40]. However, CA apparently binds poorly to the monomeric Ceg-N6, Cll-N6 and Cvv-N6, since CA at 8-fold molar excess neither inhibited on the enzymic activities of these PLA2s (Table 4), nor did it increase their neurotoxicities towards chick biventer cervicis tissue. Since these N6 myotoxins do not form a complex with CA, it may be concluded that they lack a binding site for CA. The critical evolution of the CA precursor from an E6-PLA2 (Figure 4B) may be a turning point leading to great potentiality of the neurotoxicity of N6-PLA2 in some rattlesnake venoms [6].

Structure–function relationships

All of the N6-PLA2 toxins are basic enzymes and contain 122 amino acid residues and seven conserved disulphide bonds. Their sequences have >80% identity (Figure 4A). The three-dimensional structure model of Cvv-N6 has a carbon backbone almost identical with that of trimucrotoxin (results not shown). Notably, high doses of the neurotoxic trimucrotoxin showed myotoxicity [42]; CB alone also showed myotoxic, oedema-inducing and anticoagulant activities [43]. Thus some neurotoxic sites on the N6-PLA2 molecules possibly overlap with that of the myotoxic sites.

N6-PLA2s may be classified into two structural subtypes (i.e. S24- and F24-N6-PLA2s) by careful sequence alignment and comparison (Figure 4A) and the molecular phylogeny (Figure 5). Easily distinguishable at the N-terminal regions, N1, I11, K14 and T/S23 are conserved in the S24 subtype, whereas H/S1, F11, R14 and A23 are conserved in the F24 subtype (Figure 4A). Moreover, the two subtypes probably show different surface epitopes and are immunologically distinct. Screening of many rattlesnake venoms with anti-CB antibodies showed positive results for only approx. one-third of them [4,6]. It was also shown that all the C. v. viridis venoms collected from the seven States contained N6-PLA2s [15], but the venoms failed to react to the antibodies against crotoxin or mojave toxin [4]. Thus Cvv-N6 (a S24-N6-PLA2) and CB (a F24-N6-PLA2) apparently have different epitopes.

Although all S24-N6-PLA2s from the New World pit vipers are devoid of any neurotoxicity, S24 should not be directly responsible for the lack of neurotoxicity since trimucrotoxin containing S24 is 2–3-fold more neurotoxic when compared with CB or sistruxin B-containing F24 (Table 5). However, the side-chain of F24 has been proven to be important for the toxicity of ammodytoxins A and C [44], and modification of Y22 by tetranitromethane greatly reduces the specific binding of crotoxin to synaptosomal vesicles [45]. In addition, the residues 6–8 have been shown to be essential for the neurotoxicity of trimucrotoxin by site-directed mutagenesis [46] and the S24F mutant of trimucrotoxin exhibited lower toxicity compared with the wild-type (results not shown). Thus the N-terminal region is important for the neurotoxicities of N6-PLA2s.

A natural variant of agkistrodotoxin with two substitutions (V56P and N90W) causes a decrease in lethality (on mice) by approx. 7-fold [29]. These two mutations in agkistrodotoxin make it more similar to sistruxin B (85% identities; Figure 4A), which in the absence of sistruxin A is a very weak neurotoxin. Crystallographic structure of agkistrodotoxin also indicated that regions 53–70 and 85–91 are special in this neurotoxic PLA2 [47]. Both regions and the residues 90–108 of CB have been suggested to be involved in toxicity and the interaction between CB and CA, based on the monoclonal antibody mapping methods [48]. However, the modification of residues W70 and W90 only partially reduced its lethality and myotoxicity [43].

Notably, neurotoxic S24-N6-PLA2s from the Old World Protobothrops show up to 90% sequence identity with many of the non-neurotoxic S24-N6-PLA2s from the New World pit vipers (Figure 4A). Distinct substitutions Q34, G53, I81, L106 and H108 in trimucrotoxin are possibly related to its binding to presynaptic sites [12] and with CA (Figure 6C). It has been suggested that the basic and hydrophobic residues of CB may be involved in the myotoxicity [7]. Thus different sets of basic and hydrophobic residues appear to be involved in their distinct functions, e.g. the myotoxins contain R34, K93, K115 and R116, whereas the neurotoxins contain K86, R107, R or H108 as well as hydrophobic F11, F24, Q34, G53, W70, W90 and Y or N115 (Figure 4A). The contribution of these residues to the toxicity remains to be confirmed by site-directed mutagenesis and binding assay. Recently, an in vitro R34Q mutation in a basic PLA2 from a Chinese pit viper venom was found to reduce its haemolytic activity [49], suggesting that R34 in the group II PLA2 is favourable for membrane binding, whereas Q34 may be more favourable for binding to specific neuronal proteins or receptors.

On the basis of the sequence comparison of CA-like acidic PLA2s (Figure 4B), we may predict that the unique acidic residues E47, D79 and E124 as well as neutral residues W36, A53 and G129 of CA may be involved in its recognition and binding with basic CB-like PLA2s. It is also interesting to note that unique basic residues at positions 23, 76 and 119–122 in preCA or presistruxin A (Figure 4B) may be crucial for protease hydrolysis during the post-translational processing of CA.

Binding with calmodulin

Several binding proteins of crotoxin on the neuronal or presynaptic membrane of mammalian brain have been characterized [16,50], but the mechanism by which crotoxin blocks neurotransmitter release remains obscure [51]. Calmodulin, a ubiquitous and cytosolic protein, binds a large range of proteins [16], probably via its hydrophobic and acidic domains [52,53]. The role of calmodulin in the process of neurotoxicity remains questionable. It has been shown that a chimaeric ammodytin L (non-neurotoxic) containing the C-terminal 26 residues of ammodytoxin (a neurotoxin) could bind calmodulin better but did not become neurotoxic. Moreover, the substantial increase in the binding affinity for calmodulin observed by introducing YIRN (positions 115–119 of ammodytoxin) into RVV-PL-VIII also failed to increase the toxicity of the latter [24,54]. In fact, RVV-PL-VIII and most of the non-neurotoxic N6-PLA2s contain KK or KR at positions 115–116, whereas CB has YG. Thus the positively charged 115–116 residues probably contribute to the myotoxicity rather than neurotoxicity of the PLA2s. Other specific binding proteins of crotoxin on the presynaptic membrane [50] related to exocytosis should be investigated.

The finding that CA partially protects monomeric neurotoxins such as trimucrotoxin from binding to calmodulin (Figure 6C) is in agreement with previous reports that addition of CA increased the neuromuscular blocking effects of trimucrotoxin [12] and agkistrodotoxin [27], but not Cvv-N6 (Table 5). This is also consistent with the result of calmodulin pull-down assay that CA could not bind or protect Cvv-N6 from calmodulin binding. Thus CA has a neurotoxin-recognizing specificity, that calmodulin does not have. In practice, this assay of protection from binding with calmodulin may be a convenient test for the binding of various N6-PLA2s with CA. Since CA is released after the binding of crotoxin to presynaptic acceptors [16,26,50], CA possibly binds to similar surface residues of N6-PLA2 as the presynaptic acceptors.

Abbreviations

     
  • CA

    crotoxin A

  •  
  • CB

    crotoxin B

  •  
  • DPPC

    L-dipalmitoyl phosphatidylcholine

  •  
  • ESI–MS

    electrospray ionization–mass spectrometry

  •  
  • GST

    glutathione S-transferase

  •  
  • PLA2

    phospholipase A2

  •  
  • for brevity

    the one-letter system for amino acids has been used, N6, e.g. means Asn-6

We thank Professor Y.-Y. Shu for the gift of the venom of P. mangshanensis, Professor J. R. Parmelee (Simpson College, IA, U.S.A.) for the specimen of S. c. tergeminus, the Kentucky Zoo for the gift of the venom of C. horridus (neurotoxic type) and Mr K. W. Tsai and Mr C. C. Chuang who have helped with the three-dimensional modelling of PLA2. We are grateful to Dr O. F. May for the myotoxicity assay. This research has been supported by grants from the Administration of Education (grant no. 89-B-FA01-1-4) and National Science Council (Taiwan, ROC).

References

References
1
Glenn
 
J. L.
Straight
 
R. C.
 
Tu
 
A. T.
 
The rattlesnake and their venom yield and lethal activity
Rattlesnake Venoms: their Actions and Treatments
1982
New York
Marcel Dekker
(pg. 
3
-
120
)
2
Bieber
 
A. L.
Mills
 
J. P.
Ziolkowski
 
C.
Harris
 
J.
 
Rattlesnake neurotoxins: biochemical and biological aspects
J. Toxicol. Toxin Rev.
1990
, vol. 
16
 (pg. 
33
-
52
)
3
Faure
 
G.
Harvey
 
A. L.
Thomson
 
E.
Saliou
 
B.
Radvanyi
 
F.
Bon
 
C.
 
Comparison of crotoxin isoforms reveals that stability of the complex plays a major role in its pharmacological action
Eur. J. Biochem.
1993
, vol. 
214
 (pg. 
491
-
496
)
4
Weinstein
 
S. A.
Minton
 
S. A.
Wilde
 
C. E.
 
The distribution among ophidian venoms of a toxin isolated from the venom of the Mojave rattlesnake (Crotalus scutulatus scutulatus)
Toxicon
1985
, vol. 
23
 (pg. 
825
-
844
)
5
Rael
 
E. D.
Lieb
 
C. S.
Maddux
 
N.
Varela-Ramirez
 
A.
Perez
 
J.
 
Hemorrhagic and mojave toxins in the venoms of the offspring of two Mojave rattlesnakes (Crotalus scutulatus scutulatus)
Comp. Biochem. Physiol.
1993
, vol. 
106B
 (pg. 
595
-
600
)
6
Wooldridge
 
B. J.
Pineda
 
G.
Banuelas-Ornelas
 
J. J.
Dagda
 
R. K.
Gasanov
 
S. E.
Rael
 
E. D.
Lieb
 
C. S.
 
Mojave rattlesnakes (Crotalus scutulatus scutulatus) lacking the acidic subunit DNA sequence lack Mojave toxin in their venom
Comp. Biochem. Physiol.
2001
, vol. 
130B
 (pg. 
169
-
179
)
7
Gutiérrez
 
J. M.
Ownby
 
C. L.
 
Skeletal muscle degeneration induced by venom phospholipases A2: insights into the mechanisms of local and systemic myotoxicity
Toxicon
2003
, vol. 
42
 (pg. 
915
-
931
)
8
Parkinson
 
C. L.
 
Molecular systematics and biogeographical history of pit vipers as determined by mitochondria ribosomal DNA sequences
Copeia
1999
, vol. 
1999
 (pg. 
576
-
586
)
9
Danse
 
J. M.
Gasparini
 
S.
Menez
 
A.
 
Kini
 
R. M.
 
Molecular biology of snake venom phospholipase A2
Venom Phospholipase A2 Enzyme: Structure, Function and Mechanism
1997
U.K.
Wiley
(pg. 
29
-
71
)
10
Tsai
 
I. H.
 
Phospholipase A2 of Asian snake venoms
J. Toxicol. Toxin Rev.
1997
, vol. 
16
 (pg. 
79
-
114
)
11
Kondo
 
K.
Zhang
 
J.
Xu
 
K.
Kagamiyama
 
H.
 
Amino acid sequence of a presynaptic neurotoxin, agkistrodotoxin, from the venom of Agkistrodon halys Pallas
J. Biochem. (Tokyo)
1989
, vol. 
105
 (pg. 
196
-
203
)
12
Tsai
 
I. H.
Lu
 
P. J.
Wang
 
Y. M.
Ho
 
C. L.
Liaw
 
L. L.
 
Molecular cloning and characterization of a neurotoxic phospholipase A2 from the venom of Taiwan habu (Trimeresurus mucrosquamatus)
Biochem. J.
1995
, vol. 
311
 (pg. 
895
-
900
)
13
Chijiwa
 
T.
Hamai
 
S.
Tsubouchi
 
S.
Ogawa
 
T.
Deshimaru
 
M.
Oda-Ueda
 
N.
Hattori
 
S.
Kihara
 
H.
Tsunasawa
 
S.
Ohno
 
M.
 
Interisland mutation of a novel phospholipase A2 from Trimeresurus flavoviridis venom and evolution of Crotalinae group II phospholipases A2
J. Mol. Evol.
2003
, vol. 
57
 (pg. 
546
-
554
)
14
Ownby
 
C. L.
Colberg
 
T. R.
White
 
S. P.
 
Isolation, characterization and crystallization of a phospholipase A2 myotoxin from the venom of the prairie rattlesnake (Crotalus viridis viridis)
Toxicon
1997
, vol. 
35
 (pg. 
111
-
124
)
15
Tsai
 
I. H.
Wang
 
Y. M.
Chen
 
Y. H.
Tu
 
A. T.
 
Geographic variations, cloning and functional analyses of the venom acidic phospholipase A2 of Crotalus viridis viridis
Arch. Biochem. Biophys.
2003
, vol. 
411
 (pg. 
289
-
296
)
16
Faure
 
G.
Copic
 
A.
Porrier
 
S. L.
Gubenšek
 
F.
Bon
 
C.
Krizaj
 
I.
 
Crotoxin acceptor protein isolated from Torpedo electric organ: binding properties to crotoxin by surface plasmon resonance
Toxicon
2003
, vol. 
41
 (pg. 
509
-
517
)
17
Tsai
 
I. H.
Wang
 
Y. M.
Au
 
L. C.
Ko
 
T. P.
Chen
 
Y. H.
Chu
 
Y. F.
 
Phospholipases A2 from Calloselasma rhodostoma venom gland. Cloning and sequencing of 10 of the cDNAs, three-dimensional modelling and chemical modification of the major isozyme
Eur. J. Biochem.
2000
, vol. 
267
 (pg. 
6684
-
6691
)
18
Mullis
 
K. B.
Faloona
 
F.
 
Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction
Methods Enzymol.
1987
, vol. 
155
 (pg. 
335
-
350
)
19
Maniatis
 
T.
Fritsch
 
E. F.
Sambrook
 
J.
 
Molecular Cloning: A Laboratory Manual, 2nd edn
1989
Plainview, NY
Cold Spring Harbor Laboratory Press
20
Ginsborg
 
B. L.
Warriner
 
J.
 
The isolated chick biventer cervicis nerve-muscle preparation
Br. J. Pharmacol.
1960
, vol. 
15
 (pg. 
410
-
441
)
21
Melo
 
P. A.
Ownby
 
C. L.
 
Different sensitivity of fast- and slow-twitch muscles to some snake venoms and myotoxins
Toxicon
1996
, vol. 
34
 (pg. 
653
-
669
)
22
Felsenstein
 
J.
 
PHYLIP: the PHYLogeny Inference Package, version 3.573
Computer program distributed by the Department of Genetics
1992
Washington, DC
University of Washington, Seattle
23
Felsenstein
 
J.
 
Confidence limits on phylogenies: an approach using the bootstrap
Evolution
1985
, vol. 
39
 (pg. 
783
-
791
)
24
Prijatelj
 
P.
Sribar
 
J.
Ivanovski
 
G.
Križaj
 
I.
Gubenšek
 
F.
Pungerčar
 
J.
 
Identification of a novel binding site for calmodulin in ammodytoxin A, a neurotoxic group IIA phospholipase A2
Eur. J. Biochem.
2003
, vol. 
270
 (pg. 
3018
-
3025
)
25
Chang
 
C. C.
Su
 
M. J.
 
A study on the interaction of crotapotin with crotoxin phospholipase A2, notexin and other presynaptic neurotoxins
Br. J. Pharmacol.
1981
, vol. 
73
 (pg. 
495
-
503
)
26
Chang
 
C. C.
Su
 
M. J.
 
The mechanism of potentiation of the presynaptic effect of phospholipase A2 by the crotapotin component of the crotoxin complex
Toxicon
1981
, vol. 
16
 (pg. 
402
-
405
)
27
Choumet
 
V.
Saliou
 
B.
Fideler
 
L.
Chen
 
Y. C.
Gubenšek
 
F.
Bon
 
C.
Delot
 
E.
 
Snake-venom phospholipase A2 neurotoxins, potentiation of a single-chain neurotoxin by the chaperon subunit of two-component neurotoxin
Eur. J. Biochem.
1993
, vol. 
211
 (pg. 
57
-
62
)
28
Wang
 
Y. M.
Wang
 
J. H.
Tsai
 
I. H.
 
Molecular cloning and deduced primary structures of acidic and basic phospholipases A2 from the venom of Deinagkistrodon acutus
Toxicon
1996
, vol. 
34
 (pg. 
1191
-
1196
)
29
Ji
 
Y.
Lu
 
L.
Xu
 
K.
Hattori
 
H.
Terakawa
 
S.
 
Purification and partial amino acid sequences of a new presynaptic toxin and a cytotoxin from venom of pit viper Agkistrodon blomhoffii brevicaudus
Sci. China
1995
, vol. 
38B
 (pg. 
169
-
178
)
30
Renetseder
 
R.
Brunie
 
S.
Dijkstra
 
B. W.
Drenth
 
J.
Sigler
 
P. B.
 
A comparison of the crystal structures of phospholipase A2 from bovine pancreas and Crotalus atrox venom
J. Biol. Chem.
1985
, vol. 
260
 (pg. 
11627
-
11634
)
31
Andriao-Escarso
 
S. H.
Soares
 
A. M.
Fontes
 
M. R.
Fuly
 
A. L.
Correa
 
F. M.
Rosa
 
J. C.
Greene
 
L. J.
Giglio
 
J. R.
 
Structural and functional characterization of an acidic platelet aggregation inhibitor and hypotensive phospholipase A2 from Bothrops jararacussu snake venom
Biochem. Pharmacol.
2002
, vol. 
64
 (pg. 
723
-
732
)
32
Bonfim
 
V. L.
Toyama
 
M. H.
Novello
 
J. C.
Hyslop
 
S.
Oliveira
 
C. R.
Rodrigues-Simioni
 
L.
Marangoni
 
S.
 
Isolation and enzymatic characterization of a basic phospholipase A2 from Bothrops jararacussu snake venom
J. Protein Chem.
2001
, vol. 
20
 (pg. 
239
-
245
)
33
Soares
 
A. M.
Andriao-Escarso
 
S. H.
Bortoleto
 
R. K.
Rodrigues-Simioni
 
L.
Arni
 
R. K.
Ward
 
R. J.
Gutierrez
 
J. M.
Giglio
 
J. R.
 
Dissociation of enzymatic and pharmacological properties of piratoxins-I and -III, two myotoxic phospholipases A2 from Bothrops pirajai snake venom
Arch. Biochem. Biophys.
2001
, vol. 
387
 (pg. 
188
-
196
)
34
Lizano
 
S.
Lambeau
 
G.
Lazdunski
 
M.
 
Cloning and cDNA sequence analysis of Lys49 and Asp49 basic phospholipase A2 myotoxin isoforms from Bothrops asper
Int. J. Biochem. Cell Biol.
2001
, vol. 
33
 (pg. 
127
-
132
)
35
Santoro
 
M. L.
Sousa-e-Silva
 
M. C.
Goncalves
 
L. R.
Almeida-Santos
 
S. M.
Cardoso
 
D. F.
Laporta-Ferreira
 
I. L.
Saiki
 
M.
Peres
 
C. A.
Sano-Martins
 
I. S.
 
Comparison of the biological activities in venoms from three subspecies of the South American rattlesnake (Crotalus durissus terrificus, C. durissus cascavella and C. durissus collilineatus)
Comp. Biochem. Physiol.
1999
, vol. 
122C
 (pg. 
61
-
73
)
36
Kaiser
 
I. I.
Aird
 
S. D.
 
A crotoxin homolog from the venom of the uracoan rattlesnake (Crotalus vegrandis)
Toxicon
1987
, vol. 
25
 (pg. 
1113
-
1120
)
37
Glenn
 
J. L.
Straight
 
R. C.
Wolt
 
T. B.
 
Regional variation in the presence of canebrake toxin in Crotalus horridus venom
Comp. Biochem. Physiol.
1994
, vol. 
107C
 (pg. 
337
-
346
)
38
Faure
 
G.
Choumet
 
V.
Bouchier
 
C.
Camoin
 
L.
Guillaume
 
J. L.
Monegier
 
B.
Vuilhorgne
 
M.
Bon
 
C.
 
The origin of the diversity of crotoxin isoforms in the venom of Crotalus durissus terrificus
Eur. J. Biochem.
1994
, vol. 
223
 (pg. 
161
-
164
)
39
de Oliveira
 
D. G.
Toyama
 
M. H.
Martins
 
A. M.
Havt
 
A.
Nobre
 
A. C.
Marangoni
 
S.
Camara
 
P. R.
Antunes
 
E.
de Nucci
 
G.
Beliam
 
L. O.
, et al 
Structural and biological characterization of a crotapotin isoform isolated from Crotalus durissus cascavella venom
Toxicon
2003
, vol. 
42
 (pg. 
53
-
62
)
40
Ponce-Soto
 
L. A.
Toyama
 
M. H.
Hyslop
 
S.
Novello
 
J. C.
Marangoni
 
S.
 
Isolation and preliminary enzymatic characterization of a novel PLA2 from Crotalus durissus collilineatus venom
J. Protein Chem.
2002
, vol. 
21
 (pg. 
131
-
136
)
41
Radvanyi
 
F.
Saliou
 
B.
Lembezat
 
M. P.
Bon
 
C.
 
Binding of crotoxin, a presynaptic phospholipase A2 neurotoxin, to negatively charged phospholipid vesicles
J. Neurochem.
1989
, vol. 
53
 (pg. 
1252
-
1260
)
42
Chen
 
L. N.
Liu
 
C. S.
Chang
 
C. C.
 
Isolation and characterization of a toxic phospholipase A2 from the venom of the Taiwan habu (Trimeresurus mucrosquamatus)
Biotechnol. Appl. Biochem.
1994
, vol. 
19
 (pg. 
61
-
73
)
43
Soares
 
A. M.
Mancin
 
A. C.
Cecchini
 
A. L.
Arantes
 
E. C.
Franca
 
S. C.
Gutierrez
 
J. M.
Giglio
 
J. R.
 
Effects of chemical modifications of crotoxin B, the phospholipase A2 subunit of crotoxin from Crotalus durissus terrificus snake venom, on its enzymatic and pharmacological activities
Int. J. Biochem. Cell Biol.
2001
, vol. 
33
 (pg. 
877
-
888
)
44
Petan
 
T.
Križaj
 
I.
Gubenšek
 
F.
Pungerčar
 
J.
 
Phenylalanine-24 in the N-terminal region of ammodytoxins is important for both enzymic activity and presynaptic toxicity
Biochem. J.
2002
, vol. 
363
 (pg. 
353
-
358
)
45
Tzeng
 
M. C.
Yen
 
C. H.
Hseu
 
M. J.
Dupureur
 
C. M.
Tsai
 
M. D.
 
Conversion of bovine pancreatic phospholipase A2 at a single site into a competitor of neurotoxic phospholipase A2 by site-directed mutagenesis
J. Biol. Chem.
1995
, vol. 
270
 (pg. 
2120
-
2123
)
46
Tsai
 
I. H.
Wang
 
Y. M.
 
Effect of site directed mutagenesis on the activity of recombinant trimucrotoxin, a neurotoxic phospholipase from Trimeresurus mucrosquamatus venom
Toxicon
1998
, vol. 
36
 (pg. 
1591
-
1598
)
47
Tang
 
L.
Zhou
 
Y. C.
Lin
 
Z. J.
 
Crystal structure of agkistrodotoxin, a phospholipase A2-type presynaptic neurotoxin from Agkistrodon halys Pallas
J. Mol. Biol.
1998
, vol. 
282
 (pg. 
1
-
11
)
48
Choumet
 
V.
Lafaye
 
P.
Demangel
 
C.
Bon
 
C.
Mazie
 
J. C.
 
Molecular mimicry between a monoclonal antibody and one subunit of crotoxin, a heterodimeric phospholipase A2 neurotoxin
Biol. Chem.
1999
, vol. 
380
 (pg. 
561
-
568
)
49
Jiao
 
H. M.
Jin
 
Q.
Zhao
 
J. J.
Feng
 
B.
Wu
 
X. F.
Zhou
 
Y. C.
 
The hemolytic site of the basic phospholipase A2 from Agkistrodon halys pallas
Acta Biochim. Biophys. Sinica (China)
2002
, vol. 
34
 (pg. 
383
-
387
)
50
Hseu
 
M. J.
Yen
 
C. H.
Tzeng
 
M. C.
 
Crocalbin: a new calcium-binding protein that is also a binding protein for crotoxin, a neurotoxic phospholipase A2
FEBS Lett.
2000
, vol. 
445
 (pg. 
440
-
444
)
51
Kattah
 
L. R.
Ferraz
 
V.
Matos Santoro
 
M.
Ribeiro da Silva Camargos
 
E.
Ribeiro Diniz
 
C.
De Lima
 
M. E.
 
Analysis of fatty acids released by crotoxin in rat brain synaptosomes
Toxicon
2002
, vol. 
40
 (pg. 
43
-
49
)
52
Gellman
 
S. H.
 
On the role of methionine residues in the sequence-independent recognition of nonpolar protein surfaces
Biochemistry
1991
, vol. 
30
 (pg. 
6633
-
6636
)
53
Chattopadhyaya
 
R.
Meador
 
W. E.
Means
 
A. R.
Quiocho
 
F. A.
 
Calmodulin structure determined at 1.7 Å resolution
J. Mol. Biol.
1992
, vol. 
228
 (pg. 
1177
-
1192
)
54
Prijatelj
 
P.
Križaj
 
I.
Gubenšek
 
F.
Pungerčar
 
J.
 
The C-terminal region of ammodytoxins is important but not sufficient for neurotoxicity
Eur. J. Biochem.
2002
, vol. 
269
 (pg. 
5759
-
5764
)

Author notes

The nucleotide sequence data reported here have been submitted to the DDBJ, EMBL, GenBank® and GSDB Nucleotide Sequence Databases with accession numbers AF403138 for Cvv-N6, AY355166–AY355170 for Sms-N6, Ceg-N6, Bs-N6, Sct-N6 and sistruxin B respectively and AY508693 for sistruxin A precursor.