J Mol Evol (1995) 41:174-179

Iou.. oMOLECULAR [EVOLUTION © Springer-VerlagNew YorkInc. 1995

The Molecular Cloning of a Phospholipase A 2 from Bothropsjararacussu Snake Venom: Evolution of Venom Group II Phospholipase A2's May Imply Gene Duplications Ana M. Moura-da-Silva, 1-3 Mark J.I. Paine, 1'2 Marcelo R.V. Diniz, 1'2'4 R. David G. Theakston, 2 Julian M. Crampton I a Wolfson Unit of Molecular Genetics, Liverpool School of Tropical Medicine, Liverpool L3 5QA, UK 2 Venom Research Unit, Liverpool School of Tropical Medicine, Liverpool L3 5QA, UK 3 Instituto Butantan, S~o Paulo, Brazil 4 Funda~go Ezequiel Dias, Belo Horizonte, Brazil Received: 10 June 1994 ! Accepted: 15 August 1994

Abstract. The sequence coding for a snake venom phospholipase A 2 (PLA2), BJUPLA 2, has been cloned from a Bothrops jararacussu venom gland cDNA library. The cDNA sequence predicts a precursor containing a 16-residue signal peptide followed by a molecule of 122 amino acid residues with a strong sequence similarity to group II snake venom PLA 2,s. A striking feature of the cDNA is the high sequence conservation of the 5' and 3' untranslated regions in cDNAs coding for PLA2's from a number of viper species. The greatest sequence variation was observed between the regions coding for the mature proteins, with most substitutions occurring in nonsynonymous sites. The phylogenetic tree constructed by alignment of the amino acid sequence of BJUPLA 2 with group II PLA2's in general groups them according to current taxonomical divisions and/or functional activity. It also suggests that gene duplications may have occurred at a number of different points during the evolution of snake venom group II PLA 2' s. Key words: Snake venom - - Bothrops - - Phospholipase - - Myotoxin - - Evolution - - cDNA - - Gene duplication - - Phylogeny

The nucleotide sequence reported in this paper has been submitted to the GenBank/EMBL Data Bank with accession number X76289. Correspondence to: A.M. Moura-da-Silva

Introduction Venomous snakes are classified into four families, Elapidae, which includes cobras and kraits, Hydrophiidae, sea snakes, Viperidae, the vipers, and Colubridae, of which only a few representatives are venomous (Harding and Welch 1980). The Viperidae is divided into two subfamilies, the true vipers (subfamily Viperinae) and the pit vipers (subfamily Crotalinae). The toxic activity of the venoms of these snakes reflects their taxonomy; for example, Elapidae and Hydrophiidae venoms are generally neurotoxic whereas Viperidae venoms are more complex, inducing mainly local tissue damage, hemorrhage, and often also coagulopathy (Warrell 1983). Phospholipase A2's (PLAe's) constitute a family of homologous enzymes which are widespread in snake venoms and in tissues of other animals including man; the best characterized are the secretory pancreatic and venom PLA2's. Recently, nonsecretory PLA2's have been described as being associated with local and systemic inflammatory processes and tissue injury in animals (Wery et al. 1991). Heinrikson et al. (1977), comparing the structure of snake venom PLA 2's, classified them into two distinct groups according to the organization of the cysteines. Elapidae and Hydrophiidae venoms contain group I PLA2's which resemble the primary structure of mammalian pancreatic PLA 2. In the Viperidae, they have the same cysteine organization of nonpancreatic mammalian PLA 2' s, and are classified as

175 group II PLA2's. Several toxic activities are associated with both group I and group II PLA2's, including neurotoxic, myotoxic, hemolytic, anticoagulant, and edemainducing activities (Kini and Iwanaga 1986). The most lethal toxic activity associated with group II PLA2's is the presynaptic blocking of acetylcholine release from the nerve terminals resulting in inhibition of neuromuscular transmission (Kini and Evans 1989). A class of PLA2-1ike molecules with the same cysteine organization as group II PLA 2' s, but differing in a Lys substitution of the active site residue, D-49, is also found in Viperidae venoms (Maraganore and Heinrikson 1986). The PLA 2 activity is very low or absent in these molecules, which are often associated with myotoxic activity (Cintra et al. 1993). Basic PLA2's and PLA2-1ike molecules (K-49) are present in some Bothrops (Viperidae, Crotalinae) venoms and are primarily associated with myotoxic activity (Moura-da-Silva et al. 1991). However, despite high concentrations of basic PLA 2 in some Bothrops venoms, neurotoxicity has never been reported as being clinically significant in Bothrops envenoming, suggesting that PLA 2 neurotoxic activity has been lost in this genus. We report here the molecular cloning of a Bothrops jararacussu PLA 2, the structural similarities of the putative protein with other Bothrops myotoxins and Viperidae neurotoxins, together with a phylogeny of venom group II PLA2's. These data suggest that gene duplications may have occurred during the evolution of this gene family, leading to slightly different molecules with distinct toxic activities. Materials and Methods Extraction of mRNA and Construction of the cDNA Library. A specimen of B. jararacussu (Instituto Butantan, Silo Paulo, Brazil) was killed by decapitation 4 days after milking. The venom glands were immediately removed and frozen at -80°C. The extraction of mRNA and cDNA library construction were performed as described by Paine et al. (1992). Briefly, double-stranded cDNA was synthesized from 5-10 gg poly(A)+ RNA template, ligated to EcoRI-XhoI-digested )~ ZAPII DNA, and packaged. The number of independent phages was greater than 106. The cDNA library was screened using an antisense 32-mix 17-mer synthetic oligonucleotide, myo-1 [5'-GC(C/T) TT(G/A) TC(G/A) CA(C/T) TC(A/G) CA - 3'], corresponding to the amino acid region Cys87-Ala92, which is a highly conserved region in group II PLA 2 molecules. The oligonucleotide probe was labeled at the Y-end with 732p-ATP and T4 polynucleotide kinase. Replicas of filters containing the plated library phages were prehybridized at 42°C for 2 h in hybridization solution (1% bovine serum albumin, 1 mM EDTA, 0.5 M sodium phosphate, pH 7.2 and 7% sodium dodecyl sulphate). Hybridization was carried out at 42°C overnight in the hybridization solution containing the 5' end-labeled oligonucleotide myo-1 (>10 s cpm/gg). Following hybridization, filters were washed with 0.1x SSC (15 mM sodium chloride, 1.5 mM sodium citrate, pH 7.0), 0.1% sodium dndecyl sulphate twice at room temperature and for 20 min at 42°C. Filters were autoradiographed at -70°C with pre-flashed Fuji RX X-ray fihn. Sequencing of cDNA Clones. Phage inserts were excised from the ;~ ZAPII vector and recircularized in the presence of R408 helper phage

to form pBluescript. DNA double-stranded sequencing was performed using the dideoxy chain termination method (Sanger et al. 1977) and a USB Sequenase Kit. Oligonucleotide primers to internal sequences w e r e used to obtain overlapping sequence information.

Data Analysis. The sequences used in this paper were obtained by scanning GenBank, Swiss-Prot and PIR data bases using the BLAST program (Altschul et al. 1990). The following sequences were analyzed, and respective accession numbers are shown: Bothrops jararacussu, BJUPLA 2 (this paper) and bothropstoxins I and II (Cintra et al. 1993); B. asper, myotoxins II (spP24605) and III (spP20474); Crotalus atrox, PLA 2 (spP00624); C. adamanteus, PLA 2 (spP00623); C. durissus terrificus, crotoxin (spP08878, spP24027); C. scutulatus scutulatus, Mojave toxin (spP18998, spP23559); Trimeresurus flavoviridis, BP2 (spP20381), PLX' (gb D01239), and PL-X (spP06860); T. gramineus, PLA2-II (pirA44179); T. mucrosquamatus, PLA 2 (pitS15133); Agkistrodon halys blomhoffii, basic PA2-I (spP04417) and acidic PA2-II (spP20249); A. h. pallas, agkistrotoxin (spP14421); A. piscivorus piscivorus, APP K-49 (spP04361); Vipera ammodytes ammodytes, ammodytoxins A (spP00626), B (spP14424), and C (spPll407), ammodytins L (spP17935) and Is (gbX56878), and vipoxin (spP14420); K russeUiformosensis, RV-4 (spQ02471) and RV-7 (spP31100); V. berus berus, PLA2 (spP31854); Cerastes cerastes cerastes, PEA2 (spP21789); Bitis caudalis, caudoxin (spP00622); B. nasicornis, PLA 2 (spP00621); and B. gabonica, PLA 2 (spP00620). Phylogenies were constructed using the Clustal V program (Higgins et al. 1992). Progressive alignments used the multiple alignment algorithms described by Higgins and Sharp (1989), with a fixed gap penalty of 10 and Dayhoff PAM 250 protein weight matrix (Dayhoff et al. 1978). The phylogenetic trees were generated from the above alignments using a neighbor-joining method (Saitou and Nei 1987) and rooted using the group I PLA 2 sequence of bungarotoxin A1 chain from Bungarus multicinctus venom (spP00617) as the outgroup. The degree of error was calculated for each branch by bootstraping (Felsenstein 1985) and values below the 95% confidence limits are indicated in the tree.

Results and Discussion

cDNA Cloning and Sequencing A B. jararacussu venom gland cDNA library was screened with an oligonucleotide probe (myo-1) designed to hybridize to a conserved region of venom group II PLA 2 molecules. Positive clones were detected with a frequency of 2%. The clone BJU-13, with the largest insert (approximately 800 bp), was amplified and the cDNA sequence was determined (Fig. 1). The sequence is 678 nucleotides long, excluding the poly (A) tail, comprising an extensive 5' untranslated region consisting of 157 bp, an open reading frame of 414 nucleotides that predicts a precursor for a PLA 2 of 138 amino acids, and a 3' untranslated region of 106 bp. Analysis of the primary structure of the polypeptide precursor indicates the presence of a well-defined signal sequence of 16 amino acids (underlined), followed by the mature protein comprising 122 amino acids which we have called BJUPLA 2. The predicted protein contains 14 cysteine residues in identical positions to those in the disulphide pattern found in group II PLA2's.

176 att t c c c c t g c c a c g g c t t c t c c t t c t g a t c c t t g c c t a c a g g t

t at c c t t g a c t t a c a a

acgtt~gtttagtgaccgttctaaggaceattttccagacttcaccaacggaggcgatta

120

acggggtctgctcattcccaggtctggattcgggaggATGAGGACTCTCTGGATAATGGC M

R

T

L

W

I

M

180 A

TGTGTTGCTGGTGGGCGTCGAGGGGGACCTGTGGCAATTCGGGCAGATGATCCTGAAAGA V

L

L

V

G

V

E

G

D

L

W

Q

F

G

Q

M

I

L

K

G

K

L

P

F

P

Y

Y

T

T

Y

G

C

Y

C

G

W

G

G

Q

P

K

D

A

T

D

R

C

C

F

V

H

D

C

C

Y

L

T

N

C

K

P

K

T

D

R

Y

S

Y

S

R

E

N

G

G

I

C

G

E

G

T

P

C

E

K

Q

I

C

E

C

D

K

A

V

C

F

R

E

N

L

R

T

Y

K

K

R

Y

M

A

Y

P

CGTTCTTTGCAAGAAGCCTGCAGAGAAATGCTAAgtctctgcaggccgggaaaaaaccct V

L

C

K

K

P

A

E

K

C

*

88

48O A

GGCAGTCTGCTTCAGAGAGAATCTGCGCACGTACAAGAAAAGATATATGGCGTATCCGGA A

68

420 V

TATCATCTGCGGAGAGGGCACCCCATGTGAGAAACAGATTTGTGAGTGCGACAAGGCCGC I

48

360

AAAACTGACCAACTGCAAACCCAAAACGGACCGCTACTCCTACAGCCGGGAGAATGGGGT K

28

3OO G

CCAAGGCCAGCCAAAGGACGCCACTGACCGCTGCTGCTTTGTGCACGACTGCTGTTACGG Q

8

240 E

~ACGGGGAAATTACCTTTTCCCTACTATACCACTTACGGCTGCTACTGTGGCTGGGGAGG T

6O

I08

54O D

128

6OO 138

caaattatacaaagtagttgtgttattctgaa5gcaatactgagtaataa=~acaggtgcta

660

gctttgcactcaattggcaaaaaaaaaaaaaa

693

Similarity with Other PLA 2 Myotoxins and Neurotoxins Following the data base searches, several snake venom PLA2's were found to exhibit strong homology with BJUPLA 2. The strongest similarity detected was with the myotoxic PLA 2 isolated from B. asper venom, myotoxin III. These two sequences are 90% identical and possess 7% conservative substitutions. Amino acid residues involved in the catalytic network and regions correlated to myotoxic activity are conserved in both proteins. BJUPLA 2 is probably one of the basic myotoxins recently described in B. jararacussu venom (Moura-daSilva et al. 1991). In fact, BJUPLA 2 is probably bothropstoxin II (Cintra et al. 1993) because it is also a myotoxic phospholipase A 2 and the first 33 amino acids of bothropstoxin II are the same as BJUPLA 2 except for an F substtufion with W in bothropstoxin 1] at position 5. Sequence comparison of BJUPLA 2 with other group II myotoxic or neurotoxic PLA2's (Fig. 2) revealed a higher level of similarity with the ammodytoxins isolated from the venom of the true viper, Vipera ammodytes ammodytes, than with other PLA2's isolated from ven-

Fig. 1. The cDNA sequence and deduced amino acid sequence of BJUPLA 2. The cDNA coding region is shown in uppercase and the 5' and 3' untranslated regions in lowercase letters. The sequence of the predicted mature protein is in bold letters, the signal peptide is underlined, and the putative polyadenylation sequence is double underlined.

oms of pit vipers, including B. jararacussu. The close relationship between BJUPLA 2 and ammodytoxins is also reflected in the hydropathy profiles of these molecules. Identical profiles were observed when BJUPLA 2 and ammodytoxin A sequences were analyzed but not when BJUPLA 2 was compared with the basic chains of crotoxin and Mojave toxin or with the K-49 Bothrops myotoxins (not shown).

Evolutionary Relationships of Group II PLA e Sequences Imply Gene Duplications The high similarity between BJUPLA 2 and ammodytoxins is particularly interesting considering the taxonomic and geographical distance between B. jararacussu and V. a. ammodytes. Evolutionary studies of snake venom PLA 2 sequences have been carried out by Dufton and Hider (1983) and Kostetsky et al. (1991). The dendrograms they obtained suggest very distinct groupings for group I and group II venom PLAz's. However, these authors generally used a single sequence from each spe-

177 1

I0

20

3C

40

50

60

70

80

90

i00

II0

122

B. jagaFacu$$u

BJUPLAz

DLWQFGQMI LKET~KLPF~YYTTYGCYCGWGGQGQPKDATDRC¢FVHDCCYGKLTNCKFK-DRYSYSRE;~GVI "CGEGTPCEKQICECDKAAAVCFRENLRTYKKRYMA-YPDVLCKKPAEKC

B. asper

Y y o t o x i n [11

S.I-.AK...-..KR .......................................

V. a. armnodytes

Ammodytoxin A

S . L E . . M . . . G . . . . N.LTS.SF . . . . . .

B. jararacussu

Bozhropstoxin I

S.F'L.K...Q .... N.A
Y . . K . . . K . . . G . N . . K . . . . . . W.OKT.V...NN..L.-ZL . . . . . .

V.I.L....G..MK.RY~LKP-F...-.DP.

B, asper

Myotoxi n I I

S.F£L.K...Q . . . . N.&
Y . . K . . . K . . . G . N . . K . . . . . . W.,]KT.V...NNS. L £ L . . . . . .

V . I . L . . . . N..N.K.RYYLKFLF...-ADA.

C. d. ~err~flcus

Crotoxin

S.LQFNK..KF..R. NAV. F.AF . . . . . . . . . . .

C. s. scutulatus

Mojave t o x i n

H.LQFNK..KF..R.NAI.F.AF . . . . . . . . .

S. . . . . . . . . . . . .

V..K.T .................

N.PD.S . . . . . .

KS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

-...L .........

K . H . . . . A . V . . K . . S . . N R . . . . . R . . . I . . . K . . K . . N ' Y I . R N - . . . F . . . . ES...

R. . . . . . . . . . . . . . . . . . .

AK.NT.W.I.R.. L
RV..E.L.RS.S...NE..F-...SR.RE.S.T.

R.~ . . . . . . . . . . . . . . . . . . .

AK.NT.W.I.P..L
RV..E.L.RS.S...YG..F-...SR.RG.S.T.

Fig. 2. Comparison of the predicted amino acid sequence of BJUPLA 2 with sequences of myotoxic and neurotoxic PLA2's from Viperidae venoms. Amino acid residues identical to those of BJUPLA z are indicated by dots and gaps by hyphens. The numbering refers to the BJUPLA 2 sequence.

cies and few data on group II PLA2s, making it difficult to understand the relationship between group II PLA2s. Davidson and Denis (1990) have reported a more complex phylogenetic tree using 16 group II PLA 2 sequences from venoms of different species. In this later alignment, the authors suggest that group II PLA2s are randomly mixed together, irrespective of their taxonomic or geographical distribution, through great mutational flexibility and/or the release of selective pressure operating on these genes. This hypothesis, however, seems inadequate considering the importance of some of these toxins in the overall toxicity of venom and the interactions with specific target cells which are necessary to develop the toxic effect. Thus, we decided to analyze the structural relationships of venom group II PLA2's by constructing a molecular phylogeny with more than one sequence from each species. The evolutionary tree derived from the alignment of 29 venom group II PLA2s (Fig. 3) demonstrates two primary divergent groups. Sequences from the subfamily Viperinae occur in the first group with a further subdivision between sequences of the Bitis or Vipera genera. The second major group comprises the sequences from venoms of the subfamily Crotalinae. The sole exception within these groupings is the presence of V. a. ammodytes ammodytoxins within the Crotalinae cluster, which is closely associated with the Bothrops myotoxins. There are two possible explanations for this: first, since ammodytoxins A, B, and L are present in the venom of the same species as ammodytin 12 (which is located in the Viperinae group) a gene duplication may have occurred before the divergence of Crotalinae/Viperinae. One copy evolved in the subfamily Viperinae and the other is more fully represented in Crotalinae spp. and V. a. ammodytes, coding for ammodytoxins. Second, a gene duplication may have occurred after divergence of the Viperinae and the gene coding for ammodytoxins could represent a convergent evolution with Bothrops myotoxins. In this respect, both are single-chained toxins with identical hydrophatic profiles, suggesting that the structural configurations of the Bothrops myotoxins would be favorable for interactions with membranes without the necessity of a chaperon chain which is usually present together with PLA 2 chains in other neurotoxins.

Sequences possessing similar structure/function also tend to be clustered together. Well-defined clusters comprising functionally related toxins were observed. For example, all K-49 PLA2-1ike molecules were clustered together, representing the most diverged branch from the root (group I PLA2) close to Bothrops myotoxins. This suggests a more recent divergence of this gene, which has not yet been reported in the genus Crotalus. Presynaptic neurotoxins and the nontoxic or acidic PLA2's are also clustered in two distinct branches of the tree. Whether gene duplications have occurred generating copies coding for each type of toxin is speculative and requires further investigation. However, a gene duplication may have occurred, generating the acidic and basic chains of presynaptic neurotoxins (Bouchier et al. 1991). Gene duplications are also likely to have occurred in the position generating the K-49 PLA2-1ike analogues. Venoms from five specimens of B. jararacussu were independently analyzed to investigate the chromatographic distribution of their basic proteins as previously described (Moura-da-Silva et al. 1991). At least two fractions with PLA 2 activity were found for each sampie, one corresponding to bothopstoxin I (K-49) and the other probably related to BJUPLA 2 (unpublished resuits). Interestingly, our data is not in agreement with the previous report of Davidson and Dennis (1990). When we first plotted a dendrogram in order to align the sequences, the distribution was very similar to the one reported by these authors. However, when the relative distances were analyzed, the distribution apparently conflicted with that reported by Davidson and Dennis (1990). These differences could be explained by the different methodologies used to generate the alignments and the phylogenetic trees. However, our data does agree with the phylogenetic distribution of pit vipers (Battstrom 1964) deduced from the bones of the snake's skull. Sequences from Agkistrodon spp. are preferentially clustered with those from Crotalus spp., and the sequences of the Bothrops spp. are closer to those of Trimeresurus spp. The separation of the sequences belonging to Viperinae spp. and Crotalinae spp. is very likely to be real, and clustering according to function of the toxins is also likely to have occurred.

178 B. caudalis, caudoxin B. gabonica, PLA~ B. nasicornis, PLA2

I 4~60 t54

V. b.

bet'I/s, PLA2

V. a_ aralodytes, Vipoxin V. r. forgosensis, RV-4 Y. a. atmx~¥tes, Ammodytin12 V, r. RV-T

fonf~osensis,

~3 79

[_

A. C. C. C.

h. blorahoffYl, acidic PLA~ d. terrificus, CrotoMn A s. seutulatus, Mojave toxin A adamanteus, PLA~ C, atrox, PLA2 T. 9raarineus, PLA2-11

A. h. paIlasf Agkistrotoxin C. 0. terrificus, Crotoxin B C. s. scutulatus, MojBve toxin B

. i 66

A. h. blotohoFfii, PLA2 T. flavoviridis, IF PL-X B. jararacussu, BJUPLA z B. asper, Myctoxin Ill

40

1

F--

1

L__

~43 40

If. a- an//iod.ytes,Ammecb/toxinA V. a. oto~¥tes, AmmodytoxinB V. a. af~1~oc~ytes, AmmoclytinL

~

--

A. B. B. T. T.

piscivorus, APP-K49 asper, Myotoxin II (K-49) jararacussu, BothropstoMn I (K-49) fTavoviridis, BP2 (K-49) mucrosqua~atus, PLA2 (K-49)

1

lOg divergence

Fig. 3. Phylogenetic tree of the snake group II PLAz's protein sequences. All amino acid positions of the predicted mature proteins have been included in the analysis using the CLUSTAL V program. The horizontal scale is proportional to the calculated distance (percentage divergence). The numbers indicated at nodes represent the percentage confidence limits by bootstraping. Only those values which were below 95% are indicated.

Unusual Pattern of Sequence Variation in Group II PLA 2 cDNAs

Table 1 shows the frequency of substitutions which have occurred in each structural domain of group II PLA 2 cDNAs from three different species of snake in comparison with the BJUPLA 2 eDNA sequence. The cDNAs coding for the signal peptides and 3' and 5' untranslated regions are highly conserved. Variation is highest in the domains coding for the mature proteins, where most of the substitutions occurred in nonsynonymous sites, thus accounting for the observed amino acid substitutions. This can be explained as an accelerated rate of evolution following a possible gene duplication, as may have occurred in genes coding for hemoglobin chains (Czelusniak et al. 1982). Despite the variability in the coding region, the expected conservation of regions coding for the active site and Ca a+ binding loop is apparent. The interesting point here is that the catalytic activity of the molecule is not always essential for toxicity (Rosenberg 1986), suggesting that the variation of the other regions of the molecules already correlated to their toxic activities (Kini and Ywanaga, 1986) could be the domains under selection.

The high similarity of the 5' and 3' untranslated regions in relation to the coding regions has already been reported within the same species (Bouchier et al. 1991). Unusual conservation of the intron sequences of different PLA 2 genes from the same species of snake has also recently been reported (Nakashima et al. 1993; John et al. 1994). Our data extends these observations to include higher sequence conservation of 3' and 5' untranslated regions in relation to the coding regions between species. This similarity therefore introduces a new aspect to the evolution of the genes coding for venom components which has yet to be evaluated. This study suggests that gene duplication, followed by high levels of mutations in domains coding for the mature protein, has occurred during the evolution of group II PLA 2,s. However, alignments of a greater number of sequences will be necessary to predict more precisely where gene duplication may have occurred and how these duplications relate to the toxic activity of the proteins coded for by each of the resulting gene copies. Acknowledgments. This research was funded by the Medical Research Council and the Wellcome Trust. A.M.M.S. and M.R.V.D, are sponsored by CNPq (Conselho Nacional de Desenvolvimento Cienti-

179 Table 1. Nucleotide substitutions per 100 sites in different domains of cDNAs coding for PLA2's from V. a. ammodytes, C. d. terrificus, and T. flavoviridis compared with BJUPLA 2 cDNA from B. jararacussu Mature protein Species/toxin

Noncoding

Signal peptide

V. a. ammodytes, ammodytoxin A C. d. terrificus, crotoxin B T. flavoviridis, PLX

5.1 a 5.5 5.9

13 6.4 6.4

Total

Synonymous

Nonsynonymous

22.4 20.2 19.9

7.7 2.7 4.6

14.7 17.5 15.3

a Only 16 nucleotides of the 5' untranslated region of ammodytoxin A were considered. The remaining 66 upstream bases belong to an extra exon that is apparently present only in true vipers and show no significant similarity with pit viper cDNAs

fico e Tecnologico). J.M.C. is a Wellcome Trust senior research fellow in basic biomedical sciences. The authors thank Dr. Fatima Furtado, Seq~o de Venenos, Instituto Butantan, for kindly providing the venom glands.

References Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local alignment search tools. J Mol Biol 215:403-410 Battstrom BH (1964) Evolution of pit vipers. Trans San Diego Soc Nat Hist 13:185-268 Bouchier C, Boulain JC, Bon C, Menez A (1991) Analysis of cDNA encoding the two subunits of crotoxin, a phospholipase A 2 neurotoxin from rattlesnake venom: the acidic non enzymatic subnnit derives from a phospholipase A2-1ike precursor. Biochim Biophys Acta 1088:401-408 Cintra ACO, Marangoni S, Oliveira B, Giglio JR (1993) Bothropstoxin I: amino acid sequence and function. J Prot Chem 12:57-64 Czelnsniak J, Goodman M, Hewett-Emmett D, Weiss ML, Venta PJ, Tashian RE (1982) Phylogenetic origins and adaptative evolution of avian and mammalian haemoglobin genes. Nature 298:297-300 Davidson FF, Dennis EA (1990) Evolutionary relationships and implications for the regulation of phospholipase A 2 from snake venom to human secreted forms. J Mol Evol 31:228-238 Dayhoff MO, Schartz RM, Orcutt BC (1978) Atlas of protein sequence and structure, vol 5, supplement 3. NBRF, Washington DC, p 345 Dufton MJ, Hider RC (1983) Classification of phospholipases A 2 according to sequence: evolutionary and pharmacological implications. Eur J Biochem 137:545-551 Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783-791 Harding KA, Welch RG (1980) Venomous snakes of the world--a check list. Pergamon Press Heinrikson RL, Krueger ET, Keim PS (1977) Amino acid sequences of phospholipase A 2 from the venom of Crotalus adamanteus. A new classification of phospholipases A 2 based upon structural determinants. J Biol Chem 252:4913-4921 Higgins DG, Sharp PM (1989) Fast and sensitive multiple sequence alignments on a microcomputer. Comput Appl Biosci 5:151-153 Higgins DG, Bleasby AJ, Fuchs R (1992) Clustal V: improved software for multiple sequence alignment. Comput Appl Biosci 2:189-191

John TR, Smith LA, Kaiser II (1994) Genomic sequences encoding the acidic and basic subunits of mojave toxin: unusually high sequence identity in non-coding regions. Gene 139:229-234 Kini RM, Evans HJ (1989) A model to explain the pharmacological effects of snake venom phospholipase A2's. Toxicon 27:613-636 Kini RM, Ywanaga S (1986) Structure-function relationships of phospholipases II: charge density distribution and the myotoxicity of presynaptically neurotoxic phospholipases. Toxicon 34:895-906 Kostetsky PV, Arkhipova SF, Vladimirova RR (1991) Conservative and variable regions of homologous snake phospholipases A2 sequences: prediction of the taxon-specific peptide structure. J Prot Chem 10:593-601 Maraganore JM, Heinrikson RL (1986) The lysine-49 phospholipase A 2 from the venom of Agkistrodon piscivorus piscivorus. J Biol Chem 261:4797-4804 Moura-da-Silva AM, Desmond H, Laing G, Theakston RDG (1991) Isolation and comparison of myotoxins isolated from venoms of nine different species of Bothrops snakes. Toxicon 29:713-723 Nakashima KI, Ogawa T, Oda N, Hattori M, Sakaki Y, Kihara H, Ohno M (1993) Accelerated evolution of Trimeresurus flavoviridis venom gland phospholipase A 2 isozymes. Proc Natl Acad Sci USA 90:5964-5968 Paine MJI, Desmond HP, Theakston RDG, Crampton JM (1992) Purification, cloning and molecular characterization of a high molecular weight haemorrhagic metalloprotease, jararhagin, from Bothtops jararaca venom. J Biol Chem 267:22869-22876 Rosenberg P (1986) The relationships between enzymatic activity and pharmacological properties of phospholipases in natural poisons. In: Harris JB (ed) Natural toxins, animal, plant and microbial. Clarendon Press, Oxford, pp 129-174 Saiton N, Nei M (1987) The neighbour-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4:406--425 Sanger F, Nicklen S, Coulson AR (1977) DNA sequencing with chain terminator inhibitors. Proc Natl Acad Sci USA 74:5463-5467 Warrell DA (1983) Injuries, envenoming, poisoning, and allergic reactions caused by animals. In: Weatherall DJ, Ledingham JGG (eds) Oxford textbook of medicine, vol I. Oxford University Press, Oxford, pp 6.35-6.45 Wery JP, Schevitz RW, Clawson DK, Bobbitt JL, Dow ER, Gamboa G, Goodson T, Hermann RB, Krarner RM, McClure DB, Mihelich ED, Putnam JE, Sharp JD, Stark DH, Teater C, Warrick MW, Jones ND (1991) Structure of recombinant human rheumatoid arthritic synovial fluid phospholipase A 2 at 2.2 A resolution. Nature 352:79-82