Studies on the Molecular Evolution of Phospholipase A2 lsozymes of

九州大学学術情報リポジトリ

Kyushu University Institutional Repository

ハブ属毒腺ホスホリパーゼA_2アイソザイム遺伝子の 分子進化の研究

千々岩, 崇仁

九州大学理学研究科化学専攻

https://doi.org/10.11501/3134910

出版情報：Kyushu University, 1997, 博士（理学）, 課程博士 バージョン：

権利関係：

{

Studies on the Molecular Evolution of Phospholipase A2 lsozymes of

Trimeresurus flavoviridis

Takahito Chijiwa

Department of Chemistry Faculty of Science Kyushu University

1998

CHAPTER 1

CHAPTER 2

CHAPTER 3

CHAPTER 4

CHAPTER 5

CHAPTER 6

CONTENTS

Introduction

Direct Molecular Evidence for the Geographical Variation of Phospholipase A2 Isozymes

from the Venom of Okinawa

PAGE 1

Trimeresurus flavoviridis --- 9

Composition of Phospholipase A2 Isozymes in the Venom of Amami-Oshima, Tokunoshima, and Okinawa Trimeresurus flavoviridis

---

--- 35

Comparative Study of the Protein Compositions and Nucleotide Sequences of Amami-Oshima Trimeresurus flavoviridis venom Phospholipase A2 --- 49

Structural Change of the Genes of Lys-49- Phospholipase A2 Isozymes in the Genome of Okinawa Timeresurus flavoviridis. ---

--- 62

Discussion --- 77

ACKNOWLEDGEMENTS --- 81

REFERENCES --- 82

ABBREVIATION

ASPC, �-arachidonoyl-y-stearoyl-L-a-phosphatidylcholine BAP, basic Asp-49-PLA2 (Okinawa)

BP-I, Lys-49-PLA2 (Asp59) BP-II, Lys-49-PLA2 (Asn59) CAM, S -carboxyamidomethyl CF, carboxyfluorescein IEF, isoelectric focusing

PCR, polymerase chaine reaction p/, isoelectric point

PLA2, phospholipase A2

PLA2, representative phospholipase A2 from the venom ofTokunoshima and Amami­

Oshima T. flavoviridis

PL-X', basic Asp-49-PLA2 (fokunoshima) PTH, phenylthiohydantoin

RP-HPLC, reverse phased high performance liquid chromatography RT, reverse transcription

SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel erectrophoresis SSCP, single stranded conformational polymorphism

TBP, TATA box-binding protein

TFV (T. flavoviridis), Trimeresurus flavoviridis

UTR, untransrated resion

CHAPTER 1

Introduction

Phospholipase A2 (PLA2, EC 3.1.1.4) catalyzes the hydrolysis of the 2-acyl ester bond of 3-sn-phosphoglycerides with the requirement of Ca2+. PLA2s are rich in a wide variety of resources such as mammalian pancreas and animal venoms. In particular, venom PLA2s function as a strong toxins to induce various pathological symptoms. Their contribution to digestion of prey has also been noted (Kini and Evans, 1989 and references therein). More than I 00 primary structures of PLA2s have been determined and most of these PLA2s were well-characterized. Based on the structural features, PLA2s have been classified into two large categories, namely group I and group II (Heinrikson eta!., 1977).

Group I PLA2s have a disulfide brige connecting half cystines (Cys II and

Cys69) and also a characteristic insertion of three amino acids at positions 54-56,

called elapid loop (numbered according to the aligned numbering of PLA2 enzymes

from various sources). They were found in elapinae and hydrophinae venoms, and

mammalian pancreatic juice. Group II PLA2s, on the other hands, do not have such

a loop, but have a C-terminal extention with about six amino acids terminated by a

half cystine linking to Cys50. They were found in viperinae and crotalinae venoms,

mammalian platelets (Kramer et al., 1989), and rheumatoid synovial fluid

(Seilhamer et al., 1989). Despite such small structual differences, these PLA2s

possess quite similar three-dimensional protein structures, and their structures of

catalytic center including calcium-binding site are highly homologous (Renetseder et

al., 1985; Scott" et al., 1992; Arni and Ward, 1996). Recently, a number of novel

types of PLA2 enzymes have been found, for example, from bee, lizard, rat, and

-

mouse, and those are classified all togetger into at least nine subclasses (Keith et al. , 1981� Chen etal., 1994; Balboa etal., 1996).

As mentioned above, the snake venom PLA2s are involved in both group I and group II (Dennis, 1984). Their catalytic sites are constructed in common by the diad (His48 and Asp99) with supportive residues of Tyr52 and Tyr73 (Renetseder et al., 1985). One of the most important residus for thier catalytic activities is the amino acid at position 49, which construct a part of the calcium binding site in the helix-loop-helix motif (Tanaka et al., 1987). PLA2s containing aspartic acid at this position are called Asp-49-PLA2. Another group called Lys-49-PLA2 has lysine at the position. The lipolytic activities of Asp-49-PLA2 and Lys-49-PLA2 are greatly different. The potencies of Lys-49-PLA2s for micellar substrates were reported to be only 0.2-1.8% as a compared to that of Asp-49-PLA2s (Liu et al., 1990;

Maraganore et al., 1984; Yoshizumi et al., 1990). In contrast, Lys-49-PLA2 are almost equally potent with Asp-49-PLA2 when assayed for lipid bilyers. In general, snake venom PLA2s induce very diverse pathological symptoms such as neurotoxicity (Vital and Ex cell, 1971; Lee et al., 1982; Mancheva et al., 1984 ), cardiotoxicity (Dhillon et al., 1987), hemolytic activity (Shukula and Hanahan, 1981 ), myotoxicity (Lomonte and Gutierrez, 1989; Kihara et al., 1992), anticoagulant activity (V ^{erheiji et al.,} 1980; Kini and Evans, 1987), and edema­

inducing activity (Yishwanath et al., 1987; Chi u eta/., 1989). It has been noted that pathological activities are not only related to their enzymatic activities but also to their primary structures. This means that PLA2s owe their activities to the distribution of charged and hydrophobic residues in the protein molecule (Tsai et al., 1987).

Several PLA2 tsozymes were isolated from the venom of Trimeresurus flavoviridis (T. flavoviridis, habu snake, crotalinae) and their primary structures were determined. A PLA2 isozyme called Asp-49-PLA2 (or PLA2) is a major

2

-

component of T.flavoviridis venom and has high lipolytic activity (lshimaru et al.,

1980; Tanaka et al., 1986). Its isoelectric point was estimated to be 7.9, and the physiological activity was considered to be hemolytic.

The isozymes of Asp-49-PLA2 having Asp49 have been isolated and

characterized. Those were named PLA-A and PLA-B (Yamaguchi et a!.,

unpublished data). PLA-A is the isoform of PLA-B, having �-aspartate at positions 70-71 (Asn-Giy). Isoelectric points of PLA-A and PLA-B are 8.4 and 8.6, respectively. The primary structure of PLA-B is highly similar to that of TFV PL-X (p/ = 9.2) (Kini et al., 1986), and PLA-B and TFV PL-X elicited an edema­

inducing activity. PLA-A was almost inactive. These Asp-49-PLA2 isozymes are rich in basic amino acids and are considerably more basic than ordinary Asp-49- PLA2. This is because they are called basic Asp-49-PLA2.

Two distinct Lys-49-PLA2s were also found in the venom ofT. flavoviridis.

These isozymes were named basic protein I (BP-I) and basic protein II (BP-11), respectively. Their isoelectric points are extremely high; i.e., pi= 10.2 for BP-I and p/ = 10.4 for BP-11. They usually emerge as the last two peaks in the cation­

exchange column chromatography. The difference between BP-I and BP-II is only one residue at the postion of 58, aspartic acid and aspargine for BP-I and BP-II, respectively. Although both BP-I and BP-II exhibited myotoxicity in mice (Kihara

eta!., 1992), BP-II is almost equally potent to ordinary Asp-49-PLA2 (PLA2) when assayed for lipid bilayer (Shimohigashi eta!., 1995). BP-I was less active (ca. 50

o/o) than BP-II and PLA2. Collectively, at least five PLA2 isozymes coexist in the venom ofT. jlavoviridis.

Also, five different cDNAs encoding T. flavoviridis PLA2 isozymes have been isolated from venom gland eDNA library. Those correspond to Asp-49-PLA2 (Oda eta!., 1990), two Lys-49-PLA2 (BP-I and BP-11), Thr-38-PLA2, and PL-X' (Ogawa eta!., 1992). Thr-38-PLA2 is a PLA2 isozyme in which Lys-38 of Asp-

3

-

49-PLA2 is replaced by Thr. This Thr-38-PLA2 has not been isolated yet from the venom. PLX' is grouped into basic Asp-49-PLA2, and very much similar to PLA- 8 (93o/o identity). PL-X' is also similar to TFV PL-X (Kini et al., I 986), with substitutions of nine amino acids. When the nucleotide sequences of cDNAs for these five PLA2 isozymes were compared, it was found that the 5' and 3' untranslated regions (UTRs) are conserved much more than the protein-coding regions (Nakashima

al., 1993). For instance, the sequence homologies were 98o/o for the 5' UTR,s, 67o/o for the protein-coding resions, and 89% for the 3 ' UTRs. Such a great sequence homology in the UTRs had never been reported. To investigate further such a genetic phenomenon, Nakashima et al. isolated and sequenced T. flavoviridis PLA2 isozyme genes ( 1993) (Fig. I). Each gene was found to have the same structural feature, namely the construction of four exons and three i ntrons.

The fact that the UTRs like introns are conserved more than the protein­

coding resions suggested that the genes have evolved so as to bring about accelerated amino acid substitutions in the protein coding regions. Nakashima et al.

( 1995) also analyzed the genes ofT. gramineus venom PLA2 isozymes, and found almost the same pheneomenon. Moreover, in terms of nucleotide divergence in the noncording resions including introns, Nakashima et al. compared T. flavoviridis and T. g"ramineus venom gland PLA2 isozyme genes with the genes of TATA-

binding proteins (fBP) of each species, which are assumed to be a general and nonvenomous genes and have evolved under neutrality (Kimura, 1983; Nei, 1987).

It was suggested that the introns of venom gland PLA2 isozyme genes have evolved at the rate similar to that of the introns of TBP genes in both species. It is likely that PLA2 isozymes ·evolved to gain diverse physiological activities resulting from such accelerated amino acid substitution. This was called 'Accelerated Evolution'. From the reported results (Ohta

al., 1991 ), it was speculated that there would be some

4

-

mechanisms such as gene duplication and gene conversion for these phenomena.

These mechanisms are known for evolution of multi gene families.

Although the analyses of genes of T. flavoviridis and T. gramineus have clarified the accelerated substitutions of amino acid residues of PLA2 isozymes between species, it is an essential requisite to analyze the substitution rate between isozymes in one particular species. Amami-Oshima, Tokunoshima, and Okinawa islands are known to be the representative inhabitats of T. flavoviridis. These islands are believed to be isolated from each other about 300,000 years ago, and are separated by sea about several tens kilometers away (Fig. 2, Kamiya eta/., I 984).

After this separation, it is assumed that the original constituent of snake venom PLA2 isozymes has changed more or less. Although the constitution of PLA2 isozymes ofTokunoshima T. flavoviridis venom has been well-elucidated already, that of the venom of Okinawa T. flavoviridis had been less clarified. Previously, Sadahiro eta/. (I 980) suggested the lack of the hemorrhagic activity of the venom of Okinawa T. flavoviridis. Unlike Tokunoshima T. flavoviridis, Okinawa T.

flavoviridis exhibited very weak hemorrha-ge activity.

Recent chromatographic studies by Hokama et a/. ( 1988) have also suggested that the protein composition in the venom of Okinawa T. flavoviridis differs considerably from those of Amami-Oshima and Tokunoshima T.

flavoviritlis. Such a geographical variation of venom composition within snake species has recently been noted by several different investigations (Iha eta/., 1995;

Tsai eta/., 1996). Daltry et al. suggested that, although the same snake species inhabits around Southeast Asia connected by land, the compositions of venomous proteins have been changed to adapt to their prey (I 996). To gain a further insight into the relationships between such geographical variation of the venom proteins and the evolutionary phenomenon, it would be significant to compare the PLA2 isozyme compositions and sequences between the venoms of the same snake

5

-

species in different islands. The wholesome information to clarify the process of the accelerated evolution would be expected from these analyses and the major effort in the present study has been focused on the analyses of PLA2 isozymes in the venom of Okinawa T. flavoviridis.

6

-.J

Asp-49-PLA2 Thr-38-PLA2 Basic protein I Basic proten II

PLX'-PLA2 Homology

5' UT --y- coding region

• _98.3% -1 .... �·---^----

67.6% • • _89.0% ___.

A-

A-

A-

A-

A-

Fig.

flavoviridis

Dark lines indicate the substituted bases when Asp-49-PLA2 eDNA is taken as a reference.

l

-

Kagoshima Prefecture

• Tokunoshima Okinawa

Kyushu

0 (! Tanegashima Yakushima

0 50 km

Fig. 2. The map of south Kyushu district.

8

--

CHAPTER 2

Direct Molecular Evidence for the Geographical Variation of Phospholipase A2 Isozymes from the Venom of Okinawa

Trimeresurus flavoviridis

Habu snakes (Trimeresurus jlavoviridis) inhabit the islands of the most southern Japan such as Amami-Oshima, Tokunoshima, and Okinawa. Recent chromatographic studies by Hokama et al. ( 1988) have suggested that the protein composition in the venom of Okinawa T. flavoviridis differs from those of Amami and Tokunoshima T. flavoviridis considerably. Such a geographical variation of venom composition within snake species has recently been noted by several different investigations (Iha et al., 1995; Daltry et al., 1996; Tsai et al., 1996). As to phospholipase Az (PLAz) isozymes of T.flavoviridis, no attention has been paid for the places to collect the snakes in most of the investigations. In particular, those in Okinawa T. flavoviridis venom have not been investigated well, and thus in the present study, PLA2 isozymes from the venom of Okinawa T. flavoviridis were analyzed in detail.

Phospholipase Az is one of maJor toxic proteins

the venom of T . flavoviridis. PLAz catalyzes the hydrolysis of the 2-acyl ester bond of 3-sn­

phosphoglycerides to produce fatty acids and 3-sn-lysophosphoglycerides. In the venom ofT. flavoviridis, many kinds of PLA2 isozymes coexist and some of their amino acid sequences have been determined in our lavoratory (Liu et al., 1990;

Yoshizumi et al., 1990; Oda et al., 199 I). All of these PLAzs classified into group II were further· grouped into the two different types; i.e., the one consisting of aspartate at the position 49, and the other containing lysine at the position. The representative Asp-49-PLA2 is called as PLA2 and there are two Lys-49-PLA2s

9

named BP-I and BP-11. Asp-49, which binds to Ca2+, is essential for the catalytic activity of Asp-49-PLA2, while the role of Lys49 in the lypilytic activity has not been clarified yet. Asp-49-PLA2 similar to PLA2 is the major component among PLA2 isozymes in almost the all kinds of snakes. The presence of Lys-49-PLA2s has recently been proven for several kinds of snakes, and the myonecrotic activity has been pronounced as their charachteristic toxic nature. Although these two types of PLA2 coexist in the venoms of Amami-Oshima and Tokunoshima T.

jlavoviridis, their lipolytic activities are very much different from each other. For instance, Lys-49-PLA2s were extremely weaker (0.1-1 Oo/o) than Asp-49-PLA2s for micellar substrates, despite the fact that they were almost equally potent when assayed for lipid bilayers.

Recently, another type of Asp-49-PLA2 isozymes has been cloned from the venom gland of T.flavoviridis living in Tokunoshima island (Ogawa eta!., 1992), and two isomers of this Asp-49-PLA2 were isolated and sequenced (Yamaguchi

al., unpublished data). They were found to be expressed with a considerably high

amount in the elude venom. These Asp-49-PLA2s are rich in basic amino acids and one of isozymes elicited a strong edema-inducing activity. They are named basic Asp-49-PLA2 because of their basic molecular nature. Collectively, the venoms of Amami-Oshima and Tokunoshima T. jlavoviridis consist of at least five PLA2 isozymes. Since the venom protein composition of Okinwa T. flavoviridis had been considerably different from those of Amami-Oshima and Tokunoshima T.

flavoviridis, the author intended to isolate major PLA2 isozymes of Okinawa T .

jlavoviridis and charachterize their chemical structures and biological activities.

Here, a direct molecular evidence for the geographical variation of T. flavovirdis

venom PLA2 isbzymes is presented.

10

---

Materials and Methods

Materials

Achromobacter protease I I EC 3.4.21.50 I was purchased from Wako (Osaka), and chymotrypsin I EC 3.4.2 I. I I was obtained from Sigma (St Louis, MO, USA). The chemicals used for gas-phase sequencing were supplied from Perkin Ermer Japan (Urayasu, Chiba). B-arachidonoyl-y-stearoyl-L-a-phosphatidylcholine (ASPC) was obtained from Sigma, and 5,6-carboxyfluorescein purchased from Eastman Kodak Co. (Rochester, NY, USA) was recrystallized from methanol-water.

Restriction endonucleases and other enzymes were obtained from Takara Shuzo (Kyoto). la-32PidCTP (3000 Ci I mmol) and la-35SjdATP (1000 Ci I mmol) were from Amersham (Buckinghamshaire, UK). AIJ other chemicals were of the best grade available.

Purification

Crude T. flavoviridis venom (5 g) was applied on a Sephadex G-1 00 column. A fraction showing PLA2 activity was then subjected to a CM-cellulose (Whatman CM 52) column chromatography under the conditions described in the legend for Fig. 3. The peak II emerged at the most basic side, which exhibited a weak but distinct PLA2 activity, was further subjected to reversed-phased high­

performance liquid chromatography (RP-HPLC) under the conditions described in the legend for Fig. 4.

Polyacrylamide gel electrophoresis

Sodium· dodecyl sulfate-polyacrylamide gel erectrophoresis (SDS-PAG E) was caried out according to the method of Laemmli ( 1970) on I 2.5 o/o poly­

acrylamide slab gel. The electrophoresis was continued for 3 h at 30 rnA and then

I I

protein was stained either with Coomassie brilliant blue R-250 (Nacalai Tseque, Kyoto) or with silver oxide (Wako). Polyacrylamide gel isoelectric focusing (IEF) was conducted in the pH range of 8 - 11 with Biolyte (Bio-Rad Japan, Tokyo) according to the instruction manual for a mini-IEF kit (Bio-Rad).

Amino acid analysis

Proteins and peptides were hydroryzed with 5.7 M HCI at I l OOC for 24 h.

Amino acid analysis was carried out on a Hitachi 835 amino acid analyzer by the method of Spackman eta!. ( I 985).

S -C arboxamidomethylation of phospholipase A2

The protein (200 jig) was reduced with 4 mM dithiothreitol in 0.4 M NH4HC03 ( 100 Jil) containing 8 M urea. According to the method of Stone et al.

( 1989), the reaction mixture was treated with I 00 mM iodoacetamide ( l 0 !J.I) to

convert freshly generated cisteine residues into S -carboxamidomethyl (CAM)­

cysteine. After purification on RP-HPLC, CAM-PLA2 was employed as a starting material for the sequence determination analyses.

Enzymatic cleavarages of CAM-PLA2

CAM-PLA2 (100 }J-g) was digested with chymotrypsin (1 o/o by weight) or Achromobacter protease I (I% ) described by Stone et al. ( 1989). The digests were fractionated by RP-HPLC (Wakosil 5C 18-200, 4.6 x I 50 mm). HPLC was conducted on a Gilson model 302 pump equipped with a Gilson model l I 1 B UV detecter and an M & S model 200 I recorder. Separation of enzymatically digested fragments was ·done on a Hitachi 638-30 liquid chromatograph equipped with a Hitachi 638-0410 UV detecter and a Hitachi 056 recorder.

12

--

Sequence analysis

The amino acid sequences of the native protein and the peptide fragments of CAM-PLA2 were analyzed on an Applied Biosystems 470A gas-phase sequencer equipped with a model 120A phenylthiohydantoin (fYfH) anlayzer for the on-line detection of PTH-amino acids.

Designation of peptides

All peptides were designated by letters and numbers. Letters indicate the type of cleavages: C, chymotrypsin; ^K, Achromobacter protease I. Peptide fragments obtained by RP-HPLC were numbered by Arabic numerals in the order of their elution.

Cloning and sequencing of BAP's eDNA Construction of eDNA library

Total RNAs were extracted from T. jlavoviridis venom gland, and mRNAs were purified by using Dyanabeads mRNA purification kit (Dynal, Lake Succes, NY, USA). Blunt-ended eDNA molecules were synthesized using the eDNA synthesis system plus (Amersham, Buckinghamshire, England). EcoRI cohesive termini were attached to the blunt-ended eDNA using Amersham's eDNA rapid adaptor hgation module. The A.MOSEJox vector arms supplied with this module have already been digested with EcoRI, and are therefore ready to accept any EcoRI-ended eDNA. The A.MOSEiox vector arms and cDNAs were ligated to generate recombinant DNA molecules. The cell extracts required to package the recombinant A.MOSEiox molecules into infectious phage particles are supplied with Amersham's A.-DNA in vitro packaging module. The A.MOSEiox vector features a system for automatic subcloning in vivo.

13

Screening of eDNA library

T. flavoviridis venom gland eDNA library obtained was screend by the plaque hybridization method as described by Sambrook (1989) using full-length eDNA encording T. flavoviridis basic Asp-49-PLA2 named PL-X' (Ogawa

al., 1992) as a probe.

DNA sequencing analysis

The nucleotide sequence was determined by the dideoxy chane-termination method (Sanger

al. , 1977) using denatured plasmid as a template (Hattori and Sakaki, 1986). Primers for A.MOSElox were the T7 gene 10 primer and SP6 primer.

Determination of phospholipase A2 activities Egg-yolk assay

The PLA2 activity was routinely determined using egg-yolk emulsion as a substrate on a Radiometer RTS-5 titration assembly (pH 8.0 and 37°C) as described by Ishimaru

al. (1980). Substrates were prepared as micelles mixed with deoxycholate, and enzymatically released fatty acids were titrated with 10 mM NaOH. The activity was assessed by calculating the rate of alkaline uptake against fatty acids liberated.

Preparation of phopholipid liposomes

ASPC (15 mg) dissolved in chloroform (1.5 ml) was exposed to gas blow in a conical glass vessel. Dried lipid was allowed to stand under vaccum overnight and then vortex-mixed for 30 min at OOC in a mixture (1.5 ml) of 10 mM Tris-HCI buffer (pH 8.0). The resulting suspension was sonicated for 10 min at OOC using a Tomy Seiko Model UR-200P ultrasonic disrupter and allowed to stand for 5 min at

14

0°C. This was repeated six times, and the resulting mixture of uni-and multilamellar vesicles was subjected to gel filtration on a column (0.8 x 20 em) of Sepharose 4B . Elution was carried out with 10 mM Tris-HCl buffer (pH 8.0) containig 100 mM NaCI. Small unilameller vesicles was collected and utilized for hydrolysis experiments as described below. The concentrations of phospholipids were detennined by using Phospholipid test-Wako (Wako) principally according to the procedure by Allen

al. (1940).

Preparation of {3-arachidonoyl-y-stearoyl-L-a-phosphatidylcholine mixed micelle

ASPC (1.7 mg) dissolved in chloroform (170 Jll) was exposed to nitrgen gas blow in a conical glass vessel. Dried lipid was allowed to stand under overnight and then vortex-mixed for 30 min in a mixture of 10 mM Tris-HCl buffer (pH 8.0) containing 100 mM NaCl (1379 Jll), 1 mM Triton X-100 (7 Jll), and 10 mM CaCl2 (14 Jll). The resulting suspension was sonicated for 10 min at 0°C. This was repeated three times and the suspension was used as a substrate.

Hydrolysis of {3-arachidonoyl- y-stearoyl- L-a-phosphatidylcholine

The hydrolysis of ASPC substrates were carried out as previously reported (Shimohi·gashi

al., 1995). ASPC contains �-arachidonoyl ester at the C2- position of phosphoglycerides, PLA2 isozymes cleave the C2 ester bond, and release arachidonic acid. Briefly, enzyme solution ( 10 mM final concentration) was added to a suspension of ASPC vesicles in 10 mM Tris-HCl (pH 8.0) buffer with 10 mM CaCb and 100 mM NaCl. The reaction mixture was incubated at 37°C for 60 min and thert evaporated. The residue was dissolved in ethanol (150 Jll), and aliquot (100 Jll) was analyzed by HPLC to assess the amount of arachidonic acid liberated. The retention time was determined using authentic arachidonic acid

15

(Sigma). HPLC was carried out on a Hitachi 655A-11liquid chromatograph with a reversed-phase Wakosil 5C-200 column (0.46 x 15 em) (Wako, Osaka).

In the assay for monomers, to a solution of ASPC (260 J.Lg) in a mixture of diethyl ether and ethanol (95:5 v I v) (0.2 ml) was added a solution (20 ml) of PLA2s (10 J.LM final concentration) dissolved in 10 mM Tris-HCl (pH 8.0) containing 10 mM CaC12. The reaction mixture was incubated for 3 h and treated as described above. In the assay for mixed micelles, enzyme soltuion ( 1.0 J.LM final concentration) was added to a solution of ASPC mixed micelles. The reaction mixture was incubated at 37°C for 3 h and treated as described above.

Preparation of {3-arachidonoyl- y-stearoyl-L- a-phosphatidylcholine liposome containing carboxyfluorescein

ASPC (15 mg) dissolved in chloroform (1.5 rnl) was exposed to gas blow ⁱⁿ

a conical glass vessel. Dried lipid was allowed to stand under vaccum overnight and then vortex-mixed for 30 min at OoC in a mixture (1.5 ml) of 10 mM _Tris-HCl

buffer (pH 8.0) containing 100 mM carboxyfluorescein (CF). The resulting suspension was treated as described above.

Fluoremeteric analysis of leaked carboxyfluorescein

To a mixture if 10 mM Tris-HCl buffer (pH 8.0) containing 100 mM NaCl in quartz cuvette was added liposomes ( 4.5 x 10-5 M of lipid concentration) encapsulated 100 mM carboxyfluorescein (CF). Leakage was induced by adding PLA2s dissolved in 10 mM Tris-HCl buffer (pH 8.0) to the vesicle suspension (1

mM) directry in the cuvette. Excitation was set at 4 70 nm and emission was recorded at 515.nm on a JASCO FP-550A spectrofluorometer. The fluorescein intensity was traced in the presence, of Ca2+ (1 0 mM), and the complete dye release

16

--

was achieved by disrupting the vesicles with 1 Oo/o Triton X-1 00 sol uti on (I 0 �I;

final concentration 0.1% v/v). The results are expressed as o/o CF Leakage

l 00 (F- Fo) I (Ft- Fo)

where Fo is the initial fluorescence intensity before adding PLA2s, F the fluo­

rescence induced by PLA2s at different times, and Ft the total fluorescence after addition of Triton X-1 00.

Edema-inducing activity

50mg of PLA2s were injected into the left leg of mouse. As a reference, Ringer's solution was injected into the right leg. After 6.5 h, both legs dissected were weighed and compared.

Results

Purification of phospholipas� A2

After gel filtration of elude venom of Okinawa T. flavoviridis on a S ephadex G-1 00 column, the fraction containing PLA2 activity was applied onto a CM-

cellulose column at pH 6.8 (Fukagawa eta!., 1993). As shown in Fig. 3, two major pe'aks (peak I and II) were found to exhibit PLA2 activity. Each peak pooled was subjected to RP-HPLC for further purification (Fig. 4). Proteins isolated showed a single band on both SDS-PAGE and IEF (pH 8-10), and the estimated molecular weight were approximately 15,000. The isoelectric points were determined to be 7.9 and 8.8 for PLA2s from peak I and peak II, respectively (Fig.

5). The amino·acid analysis showed that peak I PLA2 possesses almost the same amino acid composition as PLA2 from the venom of Amami-Oshima and Tokunoshima T. jlavoviridis. When the mixture of peak I PLA2 and Amami-

17

0 co

<{ (\J

Q) u c ro .n

3

3

�

.n ro

0 co

<( 1 (\J

co 0

<( (\J 3

0

0

/

Amami-os _ym ^a .J,.

Plp _activity // ++

7

0

0

5

5.

c

g_

<

�

3 3

0

::r 0 X

_._ 0 (.0

50 100 150 200 250 300 350 400

Fraction Number

7

Tokunos�irfG

/ Q

.J,.

BLA2 _activity

g_

/ � _�:

/

3 3

0

0 ::r X _._

0 (.0

50 100 150 200 250 300 350 400

Fraction Number

/

Okinawa

I, II :

PLA2 _activity / / /

7

0

0

5

5.

c

g_

<

�

3 3

0 ::r

0

�

(..)

0 LL--�--�--�----L---�--���--� 0

0 50 100 150 200 250 300 350 400

Fraction Number

Fig. 3. CM-cellulose cation exchange chromatography of the proteins from Sephadex G-100 column of Amami-Oshima, Tokunoshima, and Okinawa T. flavoviridis. The proteins were charged on a CM52 _column (3.2

I 00 em) equilibrated with 0.02 M ammonium acetate (pH 6.8) (0.02 to 0.6 M) _{over 2.0} liters. The flowrate was 18.0 ml/h and fractions of 5.0ml were collected.

18

�

<'I

<

\0

�---�

48

Peak I Peak II

� � ⁴⁶

:;1 ₈₀

"

,. /

,. /

,. ; /

,. /

,. /

I I \,, ,

� �

0 00

<'I

<t: 44 co

,...-.,.

� '--"

� co

"

,. ,.. J I �

,; ,.. /

�

,.. /

/

�

/

I \ -i ₄₂

"' -f 40

/ /

!;.

I �---�---�

30 40

30 45 60 0 15

Retention (tmin) Retention (tmin)

..

Fig. 4. Elution profile of peak I and peak II proteins from CM-cellulose cation exchange chromatography on reversed-phase HPLC. The samples were eluted through HPLC with TSK gel ODS-120T column (0.46

25 em) eluted with 0.1% trifluoroacetic acid (A)-acetonitrile containing 20% A (8). Elution was done by a linear concentration gradient of solvent 8 as follows: 20% for initial 15min, 20-80% for middle 45 min, and 60-90% for last 5 min (for peak I). 35-47% for initial 60min, and 60-90% for last 5 min (for peak II). Flow rate was 0.5 ml/min.

'

A SDS-PAGE B IEF

kDa pi

97.4�

66.2�

45.0� 9.3�

31.0�

21.5� �:��

14.4� 8.2�

Fig. 5. SDS-PAGE and polyacrylamide isoelectric focusing of peak I and peak II proteins.

20

Oshima and Tokunoshima Asp-49-PLA2 were applied onto a CM-cellulose column, they were co-eluted as a single peak under the several different elution conditions (data not shown). Furthermore, when these proteins were treated with Achromobacter proteinase ^I, the resulting peptide mixtures from respective digestion showed the exactly the same profiles of peptide mapping on RP-HPLC.

All these results suggested that peak ^I PLA2 is the same protein as Amami-Oshima or Tokushima Asp-49-PLA2. On the other hand, peak ^II PLA2 eluted at the most basic portion was found to have an amino acid composition very much similar to that of basic Asp-49-PLA2 from Tokunoshima T. flavoviridis venom. This basic Asp-49-PLA2 of Okinawa T.flavoviridis was thus designated as BAP.

Sequence analysis of phospholipas A2 by protein sequencing

The primary structure of BAP (peak ^II PLAz) was first detennined by sequencing the protein and the digests of ^S -caboxiamidomethylated (CAM)-BAP.

CAM-BAP was digested with Achromobacter protease ^I and the digest was fractionated by RP-HPLC (data not shown). The whole sequence except for the residues 31-37 was detennined by sequence analyses of isolated peptide fragments designated as Kl "'Kl5. TheN-terminal sequence of K8 was identical to that of the native protein (Fig. 7). For overlaps, CAM-protein was further treated with chymotrypsin to separate the appropriate peptides. The fragment C9 was found to overlap between fragments K 12 and K7, covering the residues 31-37. Thus, 122 amino acid residues were eventually determined as shown in Fig. 7. Although fragments CS and C 15 overlapped K-peptides, no other fragments were identified to cover the interruptions. For confirmation of the sequence determined, the nucleo-tide seqoence of eDNA was analyzed as described below.

For peak ^I PLA2, only the N-terminal sequence of natural protein was determined. The residue 1-20 was found to completely coincident with the N-

21

N N

1; peak I (Okinawa T. flavoviridis)

1 5 10 15 20

GLWQFENMIIKVVKKSGILS

2; PLA2 (Asp-49-PLA2, Tokunoshima T. flavoviridis ) GLWQFENMI IKVVKKSGILS

Fig. 6. Comparison of the N-terminal amino acid sequences of Asp-49-PLA2 (Tokunoshima) and peak I protein (Okinawa). The sample number 1 is the result of the present work. The sample number 2 is the result from the work of Tanaka

al. (1986).

'

�

.,

5 10 15 20 25 30

His leu Leu Gln Phe Arg Lys Met Ile Lys Lys Met Thr Gly Lys Glu Pro Ile Val Ser Tyr Ala Phe Tyr Gly Cys Tyr Cys Gly Lys Gly

t---K 8 _t----K 4 -i K12 ---1 ----,----,----,----,----,----,----,----,----,----,----,----,----,----,----,----,----,�----,----,----,----, ----,----,----,----,----,----,----,----,

�---c 5---� _r---c 9 _--- ---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT---,-,---"TT---"TT---"TT ---"TT ---"T7 ---"T7 ---"T7 ---"T7 ---"T7 ---"T7

35 40 45 50 55 60

Gly Arg Gly Lys Pro Lys Asp Ala Thr Asp Arg Cys Cys Phe Val His Asp Cys Cys Tyr Glu Lys Val Thr Gly Cys Asp Pro Lys Trp Asp

�---K 7 K13---

----,----,----,----,---,.---,---,.---,----,----,---,.---,----,----,----,----,----,----,----,----,----,----,----,----,----,

9---�

---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT

65 70 75

r--- --C15-------,

---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT---"TT

80 85 90

Tyr Tyr Thr Tyr Ser Leu Glu Asn Gly Asp Ile Val Cys Gly Gly Asp Asn Pro Cys Thr Lys Val Val Cys Glu Cys Asp Lys Ala Ala Ala

---Kl3 K 6 r---K g_---

----.,----,----,----,----,----,----,----,----, ---,.---,----,----,---,.---,----,----,----,----,----, ----,----,----,----,----,----,----,----,----,----,----,

95 100 105 110 115 120

Ile Cys Phe Arg Asp Asn Leu Lys Thr Tyr Lys Lys Arg Tyr Met Thr Phe Pro Asp Ile Phe Cys Thr Asp Pro Thr Glu lys Cys

---K 9 _---t f----K 2 ----i r--- KlS ---1

----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----., ----.,

Fig. 7. Amino acid sequence analysis of peak II protein from Okinawa T. flavoviridis. The initial letter K means the peptide fragments by Achromobacter protease I digestion. The initial letter C means the fragments by chymotrypsin digestion.

�

terminal sequence of Amami-Oshima and Tokunoshima Asp-49-PLA2 (Fig. 6).

The complete sequence was determined by analyzing the eDNA nucleotide sequence.

Sequence analysis of cDNAs of phospholipase A2 isozymes

The eDNA library of T. fravoviridis venom gland from Okinawa was prepared and screened by the PCR-derived probe specific for PL-X' (Ogawa

al . , 1992). PL-X' posesses the amino acid sequence very similar to BAP clarified above. About 20 positive clones were obtained from 6

I os plaques by washings high stringency (0.1

sse, 0.1 o/o SDS, 65°C). The nucleotide sequence of 710 bp was determined. This contained 5'- and 3'-UTRs, the signal peptide domain, and the mature protein coding region. The amino acid sequence of deduced from the eDNA sequence was completely in coincident with that of BAP protein determined above (Fig. 9). Thus, the full sequence of 122 amino acid residues of BAP was established as showen in Fig. 7. Based on this sequence information, the molecular weight was calculated to be about 14,700.

The amino acid sequence of BAP was compared with those of other PLA2s (Fig. 10). It was found that the sequences of BAP and PLA-8 from Tokunoshima T. fravoviridis were very much similar to each other with only three amino acid

differences. In PLA-8 the residues serine (position 68) and glutamic acid (positions 76 and 84) replace, respectively, leucine, glycine, and valine in BAP.

In the case of the eDNA of the peak I protein (Asp-49-PLA2), the same eDNA library was screened by the full length eDNA (680 bp) of Asp-49-PLA2 of Tokunoshima T. flavoviridis (Nakashima

al., 1993). About 50 positive clones were obtained from I

1 Q6 plaques by washings of high stringency. The nucleotide sequence of 741 bp was determined, and this contained 5'- and 3'­

UTRs, the signal peptide domain, and the mature protein coding region. The amino

24

18

1 17

gagggagcctgacaggt

gtgaatagccac atcgttgccattttcccctgcccggcttctccttctgatccttgccta

78 137

caggttatccttgacttacaaccgtttgtttagtgactgttctaaggacc attttccaga

13 8 197

cttttcacc agcggaggcgattaacggggtctgtccattcccaggtctggattcgggagg

198 257

ATGAGGACTCTTTGGATAATGGCCGTGTTGCTGGTGGGCGTCGAGGGGGGCCTGTGGCAA M R T L W I M A V L L V G V E G G L W Q

258 317

TTCGAGAATATGATCATTAAAGTGGTGAAGAAAAGCGGTATACTTTCGTACAGTGCTTAC F E N M I I K V V K K S G

S Y S A Y

318 377

GGATGCTACTGCGGCTGGGGGGGCCGAGGCAAGCCAAAGGACGCCACCGACCGCTGCTGT G C Y C G W G G R G K P K D A T D R C C

378 437

TTTGTGCACGACTGCTGTTACGGAAAAGTGACCGGCTGCAACCCCAAACTGGGCAAGTAC F V

D C C Y G K V T G C N P K L G K Y

438 497

ACCTACAGCTGGAACAACGGGGATATCGTCTGCGAAGGGGACGGCCCGTGCAAGGAGGTT T Y S W N N G D I V C E G D G P C K E V

498 557

TGTGAGTGCGACAGGGCCGCAGCAATTTGCTTCCGAGACAATCTGGACACATACGACAGA C E C D R A A A I C F R D N L D T Y D R

558 611 617

AACAAATATTGGCGGTACCCGGCCTCTAACTGCCAGGAGGATTCAGAACCATGCtaagtc N K Y W R Y P A S N C Q E D S E P C

618 677

tctgcaggccaggaaaaactcctcaaattacatc aatcgtagttcgtcgttactctattt

678 737

ttttgaatgcaatactgagtaataaacaggtgccagctttgcactcaaaaaaaaaaaaaa

Fig. 8. eDNA sequence and predicted amino acid sequence of peak I protein from Okinawa T. flavoviridis venom (Asp-49-PLA2)· The corresponding amino acid residues are indicated directly in the next line-. The nucleotides coding translated region is written in uppercase letters. The signal peptide region is underlined.

25

8

1 7

gacaggt 67 gtgaatagccacatcgttgccattttcccctgcctggcttctccttctgatccttgctta

68 127

caggttatccttgacttacaaccgtttgtttagtgaccattctaaggaccattttccaga

128 187

cttttcaccagcggaggcgattaacggggtctgtccattcccaggtctggattcgggagg

188 247

ATGAGGACTCTTTGGATAATGGCCGTGTTGCTGGTGGGCGTTGAGGGGCACCTGCTGCAA M R T L W I M A V L L V G V E G

L L Q

248 307

TTCAGGAAGATGATCAAGAAAATGACGGGAAAAGAGCCTATTGTGTCCTACGCCTTTTAC F R K M I K K M T G K E P I V S Y A F Y

308 367

GGATGCTACTGCGGCAAGGGGGGCCGAGGCAAGCCAAAGGACGCCACCGACCGCTGCTGC G C Y C G K G G R G K P K D A T D R C C

368 427

TTTGTGCACGACTGCTGTTACGAAAAAGTGACCGGCTGCGACCCCAAATGGGACTACTAC F V

D C C Y E K V T G C D P K W D Y Y

428 487

ACCTACAGCTTGGAGAACGGGGATATCGTCTGCGGAGGGGACAACCCGTGCACGAAGGTG T Y S L E N G D I V C G G D N P C T K V

488 547

GTTTGTGAGTGCGACAAGGCAGCGGCAATCTGCTTCCGAGACAATCTGAAGACGTACAAG V C E C D K A A A I C F R D N L K T Y K

548 601 607

AAAAGATATATGACTTTCCCGGACATTTTTTGCACGGACCCTACAGAGAAATGCtaagtc K R Y M T F P D I F C T D P T E K C

608 667

tctgcaggccgggaaaaccctctcaaattacacaatcgtagttgtgttactctattattc

668 710

tgaatgcaatactgagtaataaacaggtgccagctttgcactc

Fig. 9. eDNA sequence and predicted amino acid sequence of peak II

protein from Okinawa T. flavoviridis venom (basic Asp-49-PLA2 ). The corresponding amino acid sequences are indicated directly in the next line. The nucleotides coding translated region is written in uppercase letters. The signal peptide region is underlined.

26

N -...)

10 20 30 40 50

1. basic Asp-49-PLA2 (BAP; Okinawa) HLLQFRKMIKKMTGKEPIVSYAFYGCYCGKGGRGKPKDATDRCCFVHDCC 2. PLA-B (Tokunoshima)

3. common Asp-49-PLA2 (Tokunoshima) G

W

·EN· ·I· VVK

SG

L

SA

W

4. BP-II (Lys-49-PLA2 Tokunoshima) S

V

LW

FQE

AAKN

GL

N

V

R

S

Y

K

60 70 80 90 100 110 120

1. YEKVTGCDPKWDYYTYSLENGDIVCGGDNP-CTKVVCECDKAAAICFRDNLKTY-KKRYMTFPDIFCTDPTEKC 2.

S

E

E

3.

G

N

LGK

WN

E

G

E

R

D

DRNK

WRY

ASN

QEDS

P

4.

K

N

M

S

S

·WK

KA

EK

P

L

Q

V

L

E · · G· ·N ·

TIY

KP

KKA-DT

Fig. 10. Amino acid sequences of PLA2s from T. flavoviridis venom. The dots represent the positions identical to basic Asp-49-PLA2 (Okinawa) in amino acid sequences. The sample number 1 is the result in this work.

The sample number 2 is the result from the work (Yamaguchi

al., unpublished). The sample number 3 is the result

from the work (Oda

al., 1 990). The sample number 4 is the result from the work (Yoshizumi

al .

1 990).

acid sequence of deduced from the eDNA sequence was completely agreed with that of Tokunoshima T. flavoviridis PLA2 (Fig. 8).

Enzymatic activities of phospholipase A2

Enzymatic activity of BAP and Asp-49-PLA2 were evaluated for several different kinds of lipids. For egg-yolk emulsion, BAP was weaker (about 10%) than Asp-49-PLA2, which showed the specific activity of 360 unit I mg. On the other hand, when assayed for micelles of ASPC mixed with Triton X-1 00, BAP was almost as potent as Asp-49-PLA2 (Fig. 12). Isolated PLA2 isozymes were also tested for hydrolysis of bilayer liposome of ASPC. BAP was again as active as Asp-49-PLA2 (Fig. 11 ). These results clearly indicate that BAP has a preference to cleave �-arachidonate better than other fatty acids.

To monitor the interaction of PLAzs with membranes, dye-leakage experiments were carried out at the low lipid concentrations using ASPC liposomes encapusulating C.F. (Shimohigashi

al., 1995). Figure 13 ·shows the time dependent CF leakages (o/o) induced by three PLA2 isozymes, Asp-49-PLA2, BAP, and PLA-8, of the same concentration. Basic Asp-49 PLA2 isozymes were almost as potent as Asp-49-PLA2.

28

100

� 80 c

>

60

u

Q..)

-� � 40

20

0

AA A ^Ll A���

0

50

A

0

100 150

Time (min)

0

A

Asp-49-PLA2

�

BAP 0 PLA-8

200

Fig. 11. Time-dependent release of arachidonic acid from �-arachidonoyl phosphatidylcholine liposomes by PLA2s. PLA2 concentrarion was 1.0

1 o-5 M (final). The concentration of {3-arachidonoyl phosphatidylcholine (ASPC) substrate was 1.5 mM. The reaction buffer was I 0 mM Tris-HCl (pH 8.0) with 10 mM CaC12 and 100 mM NaCJ. The reaction was analyzed by HPLC with TSK gel ODS-120T column (0.46

25 em), eluted with 0.1 o/o trifluoroacetic acid (A)-acetonitrile containing 20o/o A (B).

Elution was done by a linear the concentration gradient of solvent B as follows: 50o/o for initial 5 min, 50-85o/o for middle 20 min, and 85 - 90o/o for last 30 min. Flow rate was

1 ml I min.

29

]()()

� 80

·;;:

0 60 u

.� �

40

Q)

20

A

A

l\ [J

[J

0 50

�

100 150

Time (min)

A

Asp-49-PLA2

l\

BAP

D

PLA-8

200

Fig. 12. Time-dependent release of arachidonic acid from

�-arachidonoyl phosphatidylcholine mixed miceUe by PLA2s. PLA2 concentrarion was 1.0 x I o-5 M (final). The substrates were prepared mixing 1.5 mM �-arachidonoyl phosphatidylcholine (ASPC) and 1.0 mM Triton X-1 00. The reaction buffer was 10 mM Tris-HCI (pH 8.0) with I 0 mM CaC12 and 100 mM NaCL The reaction was analyzed by HPLC with TSK gel ODS-120T column (0.46 x 25 em), eluted with 0.1% trifluoroacetic acid (A)-acetonitrile containing 20% A (B). Elution was done by a linear concentration gradient of solvent B as follows: 50% for initial 5min, 50-85% for middle 20 min, and 85-90% for last 30 min.

Flow rate was I ml I min.

30

100

�

!

� �

�

�

�

^�

^a

Q)

80 •

ro Ol) 0

� ro

Q)

60

...l

�

� 40

Asp-49-PLA2

6.

BAP 20

o

PLA-8

0

0 2 4 6 8 10

Time (min)

Fig. 13. Time-dependent carboxyfluorescein release from

�-arachidonoyl phosphatidylcholine liposomes by ^PLA 2s.

PLA2 concentrarion was 1

0

_{I o-5 M (final). 4.5}

_{I o-5 �­}

arachidonoyl phosphatidylcholine (ASPC) liposomes containing carboxyfluorescein (CF) were prepared. The reaction buffer was I 0 mM Tris-HCI (pH 8.0) _{with I} 0 _mM

CaCI2 and I 00 mM NaCI. The reaction was analyzed by spectrofluorometer. Exicitation was set at 470 nm and emission was recorded at 515 nm. Details were noted in Materials and Method.

3 I

Discussion

Two PLA2 1sozymes were purified from the venom of Okinawa ^T.

flavovirisdis. One is the ordinary Asp-49-PLAz whose amino acid sequence was completely in coincident with Asp-49-PLAz (PLA2) isolated from the venoms of Tokunoshima and Amami-Oshima ^T. flavoviridis. The other was also Asp-49- PLAz, but with cosiderable basic nature (p/

8.8). This basic Asp-49-PLA2 named BAP is similar to PL-X (Kini ^eta!., 1986) and PL-X' (Ogawa ^et al., 1992).

Although PL-X seemed to be isolated from the venom of ^T. flavoviridis snakes collected in Okinawa, its exact origin is still unclear. PL-X' is a eDNA-deduced protein which had been expected to be found in the venom of Tokunoshima ^T.

flavoviridis. Indeed, two structurally related basic Asp-49-PLA2 named PLA-A and PLA-8 were isolated and purified from the venom of Tokunoshima ^T.

flavoviridis, and their primary structures were determined (unpublished data). The amino acid sequence of BAP disagrees with that of PL-X at seven different positions, whereas it differs from that of PL-X' only at two positions. Since the sequence of BAP was confirmed from the analyses for both protein and eDNA, it is unlikely that the sequence discrepancy between PL-X and BAP is due to the misread amino acid residues of BAP.

Since the protein corresponding to PL-X' have been isolated from the venom of Tokunoshima ^T. flavoviridis, basic Asp-49-PLA2s are indeed the enzymes commonly expressed in the venom of ^T. jlavoviridis. The amount of basic Asp-49-PLA2s is considerably large in the venom of ^T. flavoviridis from all the three islands, suggesting that these Asp-49-PLA2s are indispensable for the toxic activity. Since the basic Asp-49-PLAz protein isolated from Tokunoshima ^T.

flavoviridis exhibited a strong edema-inducing activity, it is likely these basic Asp- 49-PLA2 are responsible for this particular physiological function in the snake venom.

32

When the lipolytic activities of two PLAzs isolated from Okinawa T.

jlavoviridis were examined using various types of phospholipid substrates, these enzymes were found to possess different substrate specificities. For egg-yolk emulsion, BAP was much weaker than Asp-49-PLAz, exhibiting only 10% activity.

Whilst, for mixed micelles and bilayer vesicles of ASPC, BAP was almost as potent as Asp-49-PLAz. Such a substrate preference of BAP to phospholipids containing

�-arachidonate is very much similar to that of Lys-49-PLAz BP-II isolated from Tokunoshima and Amami-Oshima T. flavoviridis. This implies that the basic molecular nature of BAP and BP-11 might be a determinant for substrate selection, presumably by a favorable interaction with �-arachidonate of ASPC. It should be noted that this preference is definitely related the activity in inducing the edema.

BAP was very similar to PL-X' and thus PLA-B. The amino acid sequences of PLA-B and BAP differ only at three positions, namely (Ser, Leu), (Giu, Gly), and (Glu, Val) at the positions of68 , 76, and 84, respectively. If they are in a similar three-dimensional structure as found for Tokunoshima Asp-49- PLAz (PLA2), these amino acid changes might occur in BAP in the N-terminal (Leu-68) and C-terminal (Glu-76) ends of �-wing and the N-terminal boundary of Helix 4 (Glu-84). These structural changes, however, appear not to affect much the lipolytic activities of BAP, although a small change in the hydrolytic rate was observed between these basic BAP and PLA-8 Asp-49-PLAz.

Several examples of a residual change which induces a change in activity have been reported, the resulting activity change is usually much larger than that found between BAP and PLA-B. For instance, in the case of Lys-49-PLAz isozymes (BP-I and BP-11) present in Tokunoshima and Amami-Oshima T.

jlavoviridis verrom, BP-II was 3 - 4 times active than BP-I towards mixed micelles or liposome bilayers (Shimohigashi

al., 1995). They have only one residual difference at position 58. Asp for BP-I is replaced by Asn in BP-Il, and these

33

residues exist in the middle of third loop which is said to construct a surface recognition site. BP-I and BP-II appear to have completely different substrate preferences (unpublished data).

A striking difference in the chromatographic profiles of the venoms of T.

flavoviridis from three islands, Amami-Oshima, Tokunoshima, and Okinawa (Fig.

3), is the lack of the most basic isozymes of Lys-49-PLA2 in the venom of Okinawa T.flavoviridis. There are five PLA2 isozymes in the venoms of Tokunoshima and Amami-Oshima T. flavoviridis; i.e., Asp-49-PLA2 (PLA2), basic Asp-49-PLA2s (PLA-A and PLA-B), and Lys-49-PLA2s (BP-I and BP-II). However, no peaks

corresponding to BP-I and BP-II were eluted from the venom of Okinawa T.

jlavoviridis (Fig. 3). Apparently, this difference in the PLA2 isozyme constitutions between the venoms of Okinawa and Tokunoshima (or Amami-Oshima) T.

flavoviridis is definitely important, since these snakes belong to the same species.

The direct cause of this phenomenon appears to be due to the fact that the snakes were isolated from each other in three different islands.

Nakashima et al., (1993) suggested that PLA2 isozymes have evolved via accelerated evolution. After the isolation to three islands, T. flavoviridis snakes might have evolved independently in each island, resulting in change in venom constitutions. The isolation of islands occured more than two hundred-thousands years ago (Kamiya, 1984). This might be enough to bring about the 3 residual changes between BAP and PLA-8. It is also clear that something happened to the Lys-49-PLA2 isozyme genes during this period.

34