Accelerated evolution of Trimeresurus flavoviridis venom gland ...

Proc. Natl. Acad. Sci. USA Vol. 90, pp. 5964-5968, July 1993 Evolution

Accelerated evolution of Trimeresurus flavoviridis venom gland phospholipase A2 isozymes KIN-ICHI NAKASHIMA*, TOMOHISA OGAWA*, NAOKO ODA*, MASAHIRA HATTORIt, YOSHIYUKI SAKAKIt, HIROSHI KIHARAt, AND MOTONORI OHNO*¶ *Department of Chemistry, Faculty of Science, Kyushu University, Fukuoka 812, Japan; tInstitute of Medical Science, University of Tokyo, Tokyo 108, Japan; and tBiotechnology Research Laboratories, Takara Shuzo Company, Shiga 520-21, Japan Communicated by Christian B. Anfinsen, March 24, 1993

PLA2 isozyme genes have evolved so as to bring about accelerated amino acid substitutions in the protein-coding regions except for the signal peptide-coding domains. Such accelerated substitutions appear to be adaptive and consistent with the fact that T. flavoviridis venom contains PLA2 species with diverse physiological activities.

Six Trimeresurus flavoviridis (Habu snake) ABSTRACT venom gland phospholipase A2 (PLA2) isozyme genes were found to consist of four exons and three introns and to encode proteins of 138 amino acid residues, including the signal sequence of 16 amino acid residues. Comparison of the nucleotide sequences showed that the introns are much more homologous than the protein-coding regions of exons except for the signal peptide-coding region of the first exon. The numbers of nucleotide substitutions per site (KN) for introns are approximately one-fourth of the numbers of nucleotide substitutions per synonymous site (Ks) for the protein-coding regions, indicating that the introns are unsually conserved. The absence of an apparent functional role for the introns suggests that the protein-coding regions, except for the signal peptide-coding domains, have evolved at greater substitution rates than introns. The fact that the numbers of nucleotide substitutions per nonsynonymous site (KA) are close to or larger than Ks values for relevant pairs of genes revealed that Darwinian-type accelerated substitutions have occurred in the protein-coding regions or exons. This is compatible with the presence of PLA2 species with diverse physiological activities in the venom.

EXPERIMENTAL PROCEDURES Materials. Restriction endonucleases and other enzymes were obtained from Takara Shuzo (Kyoto). [a-32P]dCTP (3000 Ci/mmol; 1 Ci = 37 GBq), [-t32P]ATP (5000 Ci/mmol), and [a-[35S]thio]dATP (1000 Ci/mmol) were from Amersham. Specific oligonucleotide probes and primers were synthesized on an Applied Biosystems model 380A DNA synthesizer. Other reagents were of reagent grade. Construction of a Genomnic DNA Library. Genomic DNAs of T. flavoviridis (Tokunoshima Island, Kagoshima Prefecture, Japan) were extracted from liver (10) and digested partially with Sau3A1. A fraction consisting of 10- to 20-kb fragments obtained by sucrose density gradient centrifugation was incorporated into EMBL3 vector (Stratagene) (1.1 x 106 primary recombinants). The DNA library was amplified once in Escherichia coli strain LE392. Screening of the DNA Library. The genomic DNA library was screened by the plaque hybridization method (11) using nearly full-length T. flavoviridis venom gland [Asp49]PLA2 cDNA (8) as a probe. Hybridization was carried out at 65°C for 12 hr in a mixture of 6x standard saline citrate (SSC), 1Ox Denhardt's solution, 1% SDS, and denatured herring sperm DNA at 100 ,g/ml. The filters were washed in 6x SSC, 2x SSC, and 0.1x SSC at 65°C twice for 30 min. Labeling of DNAs and Oligonucleotides. DNAs employed for screening or blot hybridization were labeled with [a-32P]dCTP by the random priming method (12). Oligonucleotides used for probe or for primer extension analysis were labeled at the 5'-hydroxyl end with [y.32P]ATP and T4 polynucleotide kinase (11). Subcloning and DNA Sequencing Analysis. The cloned DNAs encoding [Asp49]PLA2 isozymes in phage vectors were digested with EcoRI and then subcloned into pUC13 at the EcoRI site. The two cloned DNAs encoding [Lys49]PLA2s (basic proteins I and II) in phage vectors were digested with Xba I/Sal I and BamHI/Sal I, respectively,

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+. Aspartate-49 (numbered according to the aligned numbering of PLA2 enzymes from various sources) constitutes a part of the Ca2+ binding site (1, 2). The PLA2s are classified into two groups, [Asp49]PLA2 with high activity and [Lys49]PLA2 with extremely low activity (3). We isolated an [Asp49]PLA2 (4, 5) and two [Lys49]PLA2s called basic proteins I and 11 (6, 7) from Trimeresurus flavoviridis (Habu snake) venom. These PLA2s consist of 122 amino acid residues and are structurally homologous to one another. Especially, basic protein I is identical to basic protein II except that aspartate at position 58 of the former is replaced by asparagine in the latter. As a step toward understanding the structure-function relationships of PLA2s, cDNAs encoding [Asp49]PLA2 (8), basic proteins I and II, [Thr37]PLA2, and PLX'-PLA2 (9) were isolated from a T. flavoviridis venom gland cDNA library and sequenced. Comparison of the nucleotide sequences ofthese cDNAs revealed that the 5' and 3' untranslated regions (UTRs) are much more homologous than the protein-coding regions and that the base substitution rates at the first, second, and third positions of codons are similar in the coding regions (9). Such findings are of great interest from the viewpoint of molecular evolution. To gain further insight into this evolutionary phenomenon, we have isolated and sequenced six T. flavoviridis PLA2 isozyme genes. 11 Each gene consisted of four exons and three introns with a common structure. Comparison and analysis of their nucleotide sequences indicated that T. flavoviridis

Abbreviations: PLA2, phospholipase A2; UTR, untranslated region;

KN, number of nucleotide substitutions per site in the noncoding

regions; Ks, number of nucleotide substitutions per synonymous site; KA, number of nucleotide substitutions per nonsynonymous site. lTo whom reprint requests should be addressed at: Laboratory of Biochemistry, Department of Chemistry, Faculty of Science, Kyushu University, Fukuoka 812, Japan. 'The nucleotide sequences reported in this paper have been deposited in the GenBank data base (accession nos. D01235-D01238, D13383, and D13384).

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

5964

RESULTS AND DISCUSSION Genomic Southern and Dot Blot Analyses. Genomic Southern blot analysis of T. flavoviridis liver and venom gland pgPLA B

S

E H B ES HB E

"II

E B

pgPLA 2 E

B

S

A1II

a (2.7 kb)

I kb

I

E

E

,-i

b (2.4 kb)

a (2.7 kb)

EH

5965

DNAs digested by restriction enzymes gave several bands for each digest when the Acc I-Pst I fragment (the coding region) of T. flavoviridis [Asp49]PLA2 cDNA was used as hybridization probe (data not shown). Both venom gland and liver DNAs gave the identical profile. Dot blot hybridization of liver DNAs with [Asp49]PLA2 cDNA indicated that the numbers of genes encoding PLA2s highly (90%%) and somewhat (>60%) homologous to [Asp49]PLA2 are 4-8 and 16-32 per haploid genome, respectively, when the genome size of T. flavoviridis is assumed to be equal to that of human (data not shown). These results suggested that PLA2 isozyme genes form a multigene family. Isolation and Characterization of Genes Encoding PLA2 Isozymes. Genomic DNAs extracted from T.flavoviridis liver were partially digested with Sau3A1 and the 10- to 20-kb fragments were inserted into EMBL3 vector. To isolate as many genes encoding PLA2 isozymes as possible, the genomic DNA library thus obtained was screened with nearly full-length T. flavoviridis [Asp49]PLA2 cDNA (8), which contains the 5' and 3' UTRs highly conserved for T. flavoviridis PLA2 isozyme cDNAs (9). One hundred two positive clones were isolated from =2 x 106 plaques. Eleven clones selected on the basis of signal intensity were digested with EcoRI, electrophoresed in a 1.2% agarose gel, and blotted with the AcC I-Pst I fragment (the coding region) of [Asp49]PLA2 cDNA. The two clones named pgPLA 1 and pgPLA 2, which were 14 and 15.6 kb long, respectively, gave both 2.7- and 2.4-kb fragments which hybridized with this probe. The restriction maps of pgPLA 1 and pgPLA 2 are shown in Fig. 1. The restriction sites for BamHI, EcoRI, HindIII, and Sac I were completely identical between them. The four fragments that hybridized with the Acc I-Pst I fragment (the coding region) of [Asp49]PLA2 cDNA were designated pgPLA la (2.7 kb), pgPLA lb (2.4 kb), pgPLA 2a (2.7 kb), and pgPLA 2b (2.4 kb). The first two and last two fragments are arranged in tandem, respectively, as shown in Fig. 1. Two of the remaining 100 clones hybridized with the EcoRI-Pst I fragment (the coding region) of [Lys49]PLA2 cDNAs (9) but not with the Acc I-Pst I fragment (the coding region) of [Asp49]PLA2 cDNA. One of them was digested with Xba I/Sal I and electrophoresed in a 1.2% agarose gel. The 2.7-kb fragment hybridized with OG-I, a heptadecanucleotide probe for basic protein I, and was named BP-I. The other was digested with BamHI/Sal I and electrophoresed. The 3.1-kb fragment hybridized with OG-II, a heptadecanucleotide probe for basic protein II, and was named BP-II. Nucleotide Sequences of Six PLA2 Isozyme Genes. The nucleotide sequences of pgPLA la, pgPLA lb, pgPLA 2a, pgPLA 2b, BP-I, and BP-II were determined by the dideoxy chain-termination method after subcloning into pUC13 or pUC19. The protein-coding regions of six genes were deter-

and then subcloned into pUC19 at the corresponding sites. The nucleotide sequences were determined by the dideoxy chain-termination method (13) using denatured plasmid as template (14). Blot Hybridization Experiments. Southern blot analyses were conducted according to the method reported previously (11). T. flavoviridis liver and venom gland genomic DNAs were digested with Xba I, Pst I, HindIII, EcoRI, or BamHI; electrophoresed in 1.2% agarose gels; transferred to nylon membranes; and then hybridized with the Acc I-Pst I fragment (the coding region) of [Asp49]PLA2 cDNA. Hybridization and washings were conducted in the same manner as described above. Dot blot analysis of phage DNAs for isolation of genes encoding basic proteins I and II in which only the amino acid residue at position 58 is substituted [Asp (GAC)/Asn (AAC)] was carried out as described by Fucharoen et al. (15), with minor modifications. The antisense oligodeoxynucleotide probes employed were OG-I (5'-TTTGGGGTCGCAGCCGG-3') for basic protein I and OG-II (5'-TTTGGGGTTGCAGCCGG-3') for basic protein II. Primer Extension Analysis. Total cellular RNAs from T. flavoviridis venom gland were fractionated according to Chirgwin et al. (16). The 5' end of [Asp49]PLA2 mRNA was defined by primer extension analysis (11) with reverse transcriptase and an antisense 20-mer oligonucleotide primer (5'-ATCATATTCTCGAATTGCCA-3', nucleotide positions 502-521) complementary to the nucleotides coding for Trp3Ile9 of [Asp49IPLA2. Data Analysis. The DNASIS package developed by Hitachi Software Engineering was used for analysis of DNA sequences, and the GENAS system (17) at Kyushu University Computer Center was employed to search the EMBL/ GenBank genetic sequence data bank for sequence homology and to analyze the secondary structures of precursor RNAs for introns according to the method of Zuker and Steigler (18). For each of the pairwise comparisons of the nucleotide sequences of six T. flavoviridis PLA2 isozyme genes, the numbers of nucleotide substitutions per site (KN) for the noncoding regions and the numbers of nucleotide substitutions per synonymous site (Ks) and per nonsynonymous site (KA) for the protein-coding regions were computed according to the method of Miyata and Yasunaga (19) with correlations for multiple substitutions (20). A synonymous site is a site of a codon at which base substitution causes no amino acid change. A nonsynonymous site is a site of a codon at which base substitution causes an amino acid change.

E

_s~

Proc. Natl. Acad. Sci. USA 90 (1993)

Evolution: Nakashima et al.

E H B ES HB E

1'

|

E B

LA

,-{.: :;:+s

E

i.8.

E

It.,ti.

b (2.4 kb)

FIG. 1. Structures of phage clones (pgPLA 1 and pgPLA 2) encoding T. flavoviridis [Asp49]PLA2 isozymes. Restriction sites: B, BamHI; E, EcoRI; H, HindIII; S, Sac I. EcoRI sites have not been detected in the stippled regions. The fragments that hybridized with the Acc I-Pst I fragment (the coding region) of [Asp49] PLA2 cDNA are shown by stippled boxes. These fragments (a and b) were sequenced.


Evolution: Nakashima et aL

5966

mined by matching their sequences with those of [Asp49IPLA2 cDNA (8) and [Lys49]PLA2 cDNAs (9). The transcription initiation sites were determined for pgPLA la and pgPLA 2a by primer extension analysis using total venom gland RNAs as template (data not shown). The sites were identical for both genes and assigned to the adenosine residue located 204 nucleotides upstream from the translation initiation codon (ATG) (21) (positions 205-207), and the TATA-like sequence (22) (CATAAA) was found 34 nucleotides upsteam from the transcription initiation site. Because of their sequence similarities, it is assumed that the transcription initiation sites of these sLx isozyme genes are identical. Each gene spanned about 1.9 kb, contained four exons and three introns, and encoded a protein of 138 amino acid residues, including the signal sequence of 16 amino acid residues (Fig. 2). The nucleotide sequence for combined exons of pgPLA 2a was in accord with that of T. flavoviridis [Asp49]PLA2 cDNA (8), although the latter is 144 nucleotides shorter in the 5' UTR. Five base substitutions were noted between pgPLA la and pgPLA 2a. Although they carry the same mature protein, there is only one amino acid substitution [Glu (GAG)/Asp (GAT)] at the -2 position of the signal-peptide sequence. Five base substitutions were also found between pgPLA lb and pgPLA 2b but all are located in introns. Both pgPLA lb and pgPLA 2b encoded an isozyme of [Asp49]PLA2 (Fig. 2). The genes BP-I and BP-II encoded basic proteins I and II, respectively. There were seven base substitutions and a deletion (or insertion) of four consecutive nucleotides (intron 3) between BP-I and BP-II. Polymorphisms were observed for such nucleotide substitutions between pgPLA la and pgPLA 2a and between pgPLA lb and pgPLA 2b among individuals of T. flavoviridis (data not shown). The gene encoding human pancreatic [Asp49]PLA2 also contains four exons (23). However, the gene coding for human rheumatoid synovial fluid (nonpancreatic) [Asp491PLA2 was reported to have five exons (24). The highly conserved GT (donor site)/AG (acceptor site) structure commonly noted for introns (25) was seen for all the genes determined here. When the nucleotide sequences of four T. flavoviridis [Asp49]PLA2 genes and human pancreatic (23) and nonpancreatic (24) PLA2 genes were compared by taking account of their amino acid sequences, it became evident that the exon-intron junction sites are completely conserved among these genes.

Analysis of Nudeotfde Substitutions in PLA2 Isozyme Genes Suggests Accelerated Evolution. The nucleotide sequences and the corresponding amino acid sequences for the first intron and the second exon of six PLA2 isozyme genes are shown in Fig. 3. It is easily recognized that the first intron is much more homologous than the second exon. This is also true of other introns and the protein-coding regions of other exons except for the signal sequence-coding domain of the first exon. For example, sectional homologies between pgPLA la and BP-I are 98.5% for the 5' UTR and 97.5% for the signal peptide-coding domain (first exon); 93.8% (first intron), 67.7% (second exon), 92.8% (second intron), 82.1% (third exon), 96.9%o (third intron), and 75.2% for the proteincoding region; and 91.5% for the 3' UTR (fourth exon). It is generally known that the evolution rates of introns are much greater than those of the protein-coding regions (or exons) (20), so that the structures of T. flavoviridis PLA2 isozyme genes seem to be anomalous. The evolutionary significance involved in the nucleotide sequences of T. flavoviridis PLA2 isozyme genes was analyzed by computing KN, KS, and KA values for all the relevant pairs of PLA2 isozyme genes (19, 20). Table 1 shows the values for three pairs of the genes. The data reveal several characteristics. First, as in the 5' and 3' noncoding regions, KN values for introns are approximately one-fourth of Ks values for all the pairs of the genes compared. This indicates that the introns are unusually conserved as compared to the protein-coding regions. The high homology of introns may suggest that there are some functionally important constraints in the introns. However, the regions corresponding to introns of precursor RNAs involved no significant secondary structure when analyzed by the GENAS system (17) according to the method of Zuker and Steigler (18) (data not shown). Thus, it may be reasonable to consider that the introns have no significant functional role. The absence of an apparent role for the introns suggested that the proteincoding regions of the exons have evolved at greater substitution rates than the introns. Second, KA/KS values are much greater than those reported for other isoprotein genes (20, 26). Although synonymous sites are known to be much more variable than nonsynonymous sites because of much less functional constraint in the former (20, 26), this is not the case in the protein-coding regions of T. flavoviridis PLA2 isozyme genes. The KA/KS values of the coding regions for pairs of pgPLA la and pgPLA lb or pgPLA 2b and of pgPLA la and BP-I or BP-II are close to 1, and those for pairs of pgPLA lb 0.5kb

A

Exon

488 60

a b C

> 1297 1397

(I(! 1657

_

I 190

235

MRTLWIMAVLLVGVDGGLWQFENMIIKVVKKSGILSYSAYGCCYCGWGGRGKPKDATDRCCFVHDCCYGKVTGCNPKLGKYTY SWNNGDI VCEGDG-PC-KEVCECDRAAMICFRDN LDTYDRNKYWRYPASNCQEDSEPC MRTLW IMAVLLVGVKGHLMQFENMI KKVTGRSG IWWYGSYGCYCGKGGEGRPQDP SDRCCFVHDCCYGVGCDPKDDFY I YSSENGD IVCGDDD- LCKKEVCECDKAAAI CFRDNMDTY -QNKYWFY PASNCKEESEPC

MRTLWIMAVLLLGVDGSLVQLWKMI 1

-16

-1

12ETGKEAAKNYGLYGCNCGVGRRGKPKDATDSCCYVHKCCYKKVTGCDPKMDSYSYSWKNKAIVCGEKNPPCLKQVCECDKAVAICLRENLGTY-NKKYTIYPKPFCKK-ADTC

124

FIG. 2. Schematic representation ofa common structure of six T.flavoviridis PLA2 isozyme genes-pgPLA la, pgPLA 2a, pgPLA lb, pgPLA 2b, BP-I, and BP-II-and the amino acid sequences corresponding to the coding regions of pgPLA 2a (line a), pgPLA lb (line b), and BP-I (line c). Four exons are indicated by boxes and the UTRs are hatched. The nucleotide position numbers represent those for pgPLA la and pgPLA 2a. Four coding regions and their corresponding amino acid sequences are correlated by lines. The residues involved in the Ca2+ binding site are marked by v, those in catalytic site by v, and the residue at position 58 by *.



pgPLA la pgPLA 2a pgPLA lb pgPLA 2b BP-I

245 GTGAGTGAAGCAAACTTAAAAAA 245 .................................................................... G....T.. 245 .G. 245 245

G.

BP-II

245

G.

313 313

T

5967

AGGTGXAccIc_

TGC

................................................................................

313

....

313

....

312

....

T

T

AA

T .T.

AA.

312.

G.

T.

G.

T.

T. T .G.

393 ATGTTTTGCTGTAGTCGTTAAGcGGGACTGCCAGCATCTGCCAC A 393 ................................................................................ 393 .C... G. 393 .C.... G.T. 392 T.C.... G..T 392 T.C.... G..T

G

.TC

..

CICTGTGGCAATTrGAG'ATATGCATTAAAGTGGTAGAMAAGATATAC2T alGluGlyGlyLeuTrpGlnPheGluAsnMetIleIleLysValValLysLysSerGlyIleLeu 473 ............................................ T * * * * * A5p * * * * * * * * * * C. A 473. AG AC.. CA... AT T.77W * Lys * His * Met * * * * * * Lys * * ThrGlyArg * * * Trp 473 C. A CA ; XAT.A.. a .G.... T.T1W * Lys * * ThrGlyArg * * * Trp ^ Lys * His * Met * . T. A... .G.T. A A.C. A.C. GCTAAA 472 ..C.C. GG PheGlnGluThrGly * GluAlaAlaLys Asp * Ser * Val * LeuTrpLys 472 ..C ..C. GAA.C.GCTAAA A..G.... T.AC. C * Asp * Ser * Val LsuTrpLys * * PheGlnGluThrGly * GluAlaAlaLys

473 TlcGCTcrTTTACAGTCGA

..............................

.....

..

....

.. ..

...

...

.

.

..

553 TACAGTGCTACGGATGCTAGCTCCCAGGAGGCAAGCCAACCCACCGACCG

SerTyrSerAlaTyrGlyCysTyrCysGlyTrpGlyGlyArgGlyLysProLysAspNlaThrAspAr 553 553

....................................................................

553

552

552

AA Lys

.G.... G.CT.

Trp

*

.G ....

GlySer

*

Asn *

AAC

G. CTTA

*

AAC

...

...

*

*

*

*

G.

Asn * GlyLeu *

A..

*

A Lys

G.C

GlySer * CTEA GlyLeu *

Trp

*

*

Asn

*

A.

Asn

Glu

A..

....

*Glu

CC....

* Arg *

C..

C.

Gln

Gln * ProSer* .T

..G

Val * Arg T....AG aVal Arg

*

*

*

..G..

*

*

C ...GM

CG...

* Arg *

GT.

* ProSer *

AG

A.

*

7C....

*

*

*

*

T.T

*

*

T.

*

A.

Se A.

Se

FIG. 3. Nucleotide sequences of the first introns and the second exons of T. flavoviridis PLA2 isozyme genes pgPLA la, pgPLA 2a, pgPLA lb, pgPLA 2b, BP-I, and BP-II.

and BP-I or BP-II are about 1.8. Such high degrees of substitutions in nonsynonymous sites suggest that T. flavoviridis PLA2 isozyme genes have evolved so as to bring about accelerated amino acid substitutions. The high homology of the signal sequence-coding domain in the first exon can be ascribed to the common role of the signal peptides in membrane penetration. Diverse Physiological Activities of T. flavovidis PLA2 Isozymes. In terms of lipolytic activity, basic proteins I and II

([Lys49IPLA2s), the products derived from the BP-I and BP-II

genes,

respectively,

are

only 1.5-1.7%

as

active

as

[Asp49]PLA2 (6, 7). Low lipolytic activity of basic proteins I and II is thought to be due in a large part to the presence of lysine in place of the Asp49 residue of [Asp49]PLA2. The Asp49 residue is located in the Ca2+ binding site of [Asp49]PLA2 (1, 2). Lys49 in basic proteins I and II is unable to chelate Ca2+ by itself. However, basic proteins I and II were twice as active as [Asp49]PLA2 in necrosis-inducing

Table 1. KNIKs and KA/Ks values between pairs of T. flavoviridis PLA2 isozyme genes pgPLA la vs. BP-I pgPLA lb vs. BP-I pgPLA la vs. pgPLA lb Ks Ks KN/KS KN KNIKS KN KNIKS KN Ks. 5' flanking region 0.051 0.224 0.228 0.035 0.253 0.138 0.029 0.183 0.158 5' UTR 0.020 0.224 0.015 0.253 0.137 0.089 0.059 0.025 0.183 Intron 1 0.042 0.224 0.186 0.064 0.253 0.253 0.069 0.183 0.377 0.224 0.058 0.317 Intron 2 0.084 0.375 0.076 0.253 0.300 0.183 0.175 0.224 0.253 0.126 0.032 0.183 Intron 3 0.032 0.143 0.032 0.054 0.183 0.295 AMl introns 0.063 0.224 0.281 0.063 0.253 0.249 3' UTR 0.355 0.059 0.224 0.263 0.059 0.253 0.233 0.065 0.183 0.328 0.224 0.023 0.253 0.091 0.060 0.183 0.060 0.268 3' flanking region 0.273 0.217 0.050 0.183 0.224 0.055 0.253 All noncoding regions 0.057 0.254 Ks Ks Ks KA KAJKs KA KAIKS KA KA/Ks 0.341 0.039 0.113 0.000 0.039 0.000 0.000 0.116 Coding region of exon 1 1.84 0.525 0.286 0.441 0.379 1.16 0.654 0.296 0.452 Exon 2 1.84 0.251 0.136 1.07 0.081 2.05 0.218 0.203 0.166 Exon 3 0.166 1.68 0.279 0.279 0.316 0.855 0.101 0.174 0.579 Coding region of exon 4 0.183 1.75 1.17 0.321 0.752 0.2% 0.253 0.168 0.224 All coding regions The Ks values used for estimation of KN/Ks for the noncoding regions are those for the complete coding regions of the corresponding pairs.

5%8



activity (27). This was ascribed to greater depolarization effects for muscle cell membranes of basic proteins I and II than [Asp4]PLA2 (27). In the guinea pig ileum contraction assay, [Asp49]PLA2 and basic protein II elicited strong contraction at doses 10 times less than that for basic protein I (H. Matsumoto, Y. Shimohigashi, and M.O., unpublished results). Ability of proteins to contract ileum appears to depend on whether the residue at position 58 is charged or noncharged (28). [Asp49]PLA2 and basic protein II have asparagine whereas basic protein I has aspartate at this position. It is assumed that the residue at position 58 is involved in the site that specifically interacts with organized phospholipid matrices. These observations suggest that T. flavoviridis PLA2 isozymes can manifest diverse physiological activities. The presence of such PLA2 isozymes in venom must have a strong selective advantage to disrupt the physiological integrity of animals for catching prey or for defense against predators. Thus, it could be inferred that the nucleotide substitutions in T. flavoviridis PLA2 genes to acquire PLA2 species effective for such purposes have occurred by positive Darwinian selection. Presumption. Accelerated amino acid substitutions have been reported to occur in the active-site regions of serine protease inhibitors (29). It is known for the major histocompatibility complex multigene family that the putative antigenrecognition sites, but not other parts, have evolved via positive Darwinian selection (30). In contrast, nonsynonymous substitutions are spread over the protein-coding region in T. flavoviridis PLA2 isozyme genes except for the region coding for the signal peptide. Serine protease inhibitors (29), major histocompatibility complex gene products (30), and the venom PLA2s described here are regarded as proteins which interact functionally with diverse foreign substances. It may be assumed that such proteins each manifesting a particular biochemical and physiological activity tend to evolve under adaptive pressure. We thank Drs. T. Miyata and H. Hayashida (Department of Biophysics, Faculty of Science, Kyoto University) for computing KN, Ks, and KA values. This work was supported in part by Grants-in-Aid for Scientific Research (nos. 03453165, 03554021, and 9018) from the Ministry of Education, Science, and Culture of Japan. 1. Dijkstra, B. W., Kalk, K. H., Hol, W. G. J. & Drenth, J. (1981) J. Mol. Biol. 147, 97-123. 2. Renetseder, R., Brunie, S., Dijkstra, B. W., Drenth, J. & Sigler, P. B. (1985) J. Biol. Chem. 260, 11627-11634. 3. Maraganore, J. M., Merutka, G., Cho, W., Welches, W., Kezdy, F. J. & Heinrikson, R. L. (1984) J. Biol. Chem. 259, 13839-13843.

4. Ishimaru, K., Kihara, H. & Ohno, M. (1980) J. Biochem. (Tokyo) 88, 443-451. 5. Tanaka, S., Mohri, H., Kihara, H. & Ohno, M. (1986) J. Biochem. (Tokyo) 99, 281-289. 6. Yoshizumi, K., Liu, S.-Y., Miyata, T., Saita, S., Ohno, M., Iwanaga, S. & Kihara, H. (1990) Toxicon 28, 43-54. 7. Liu, S.-Y., Yoshizumi, K., Oda, N., Ohno, M., Tokunaga, F., Iwanaga, S. & Kihara, H. (1990) J. Biochem. (Tokyo) 107, 400-408. 8. Oda, N., Ogawa, T., Ohno, M., Sasaki, H., Sakaki, Y. & Kihara, H. (1990) J. Biochem. (Tokyo) 108, 816-821. 9. Ogawa, T., Oda, N., Nakashima, K., Sasaki, H., Hattori, M., Sakaki, Y., Kihara, H. & Ohno, M. (1992) Proc. Natl. Acad. Sci. USA 89, 8557-8561. 10. Blin, N. & Stafford, D. W. (1976) Nucleic Acids Res. 3, 2303-2307. 11. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual (Cold Spring Harbor Lab. Press, Plainview, NY). 12. Feinberg, A. & Vogelstein, B. (1983) Anal. Biochem. 132, 6-13. 13. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. 14. Hattori, M. & Sakaki, Y. (1986) Anal. Biochem. 152, 232-238. 15. Fucharoen, S., Fucharoen, G., Fucharoen, P. & Fukumaki, Y. (1989) J. Biol. Chem. 264, 7780-7783. 16. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J. & Rutter, W. J. (1979) Biochemistry 15, 5294-5299. 17. Kuhara, S., Matsumoto, F., Futamura, S., Fujita, A., Shimohara, T., Takagi, T. & Sakaki, Y. (1984) Nucleic Acids Res. 12,

88-99. 18. Zuker, M. & Steigler, P. (1981) Nucleic Acids Res. 9, 133-148. 19. Miyata, T. & Yasunaga, T. (1980) J. Mol. Evol. 16, 23-36. 20. Kimura, M. (1983) The Neutral Theory of Molecular Evolution (Cambridge Univ. Press, Cambridge, U.K.). 21. Kozak, M. (1984) Nucleic Acids Res. 12, 857-872. 22. Coden, J., Wasylyk, B., Buchwalder, A., Sassone-Corsi, P., Kedinger, C. & Chambon, P. (1980) Science 209, 1406-1414. 23. Seilhamer, J. J., Randall, T. L., Yamanaka, M. & Johnson, L. K. (1986) DNA 5, 512-527. 24. Seilhamer, J. J., Prouzanski, W., Vadas, P., Plant, S., Miller, J. A., Kloss, J. & Johnson, L. K. (1989) J. Biol. Chem. 264, 5335-5338. 25. Breathnach, R., Benoist, C., O'Hara, K., Gannon, F. & Chambon, P. (1978) Proc. Natl. Acad. Sci. USA 75, 4953-4857. 26. Nei, M. (1987) Molecular Evolutionary Genetics (Columbia Univ. Press, Irvington-on-Hudson, NY). 27. Kihara, H., Uchikawa, R., Hattori, S. & Ohno, M. (1992) Biochem. Int. 28, 895-903. 28. Fukagawa, T., Matsumoto, H., Shimohigashi, Y., Ogawa, T., Oda, N., Chang, C.-C. & Ohno, M. (1992) Toxicon 30, 13311334. 29. Hill, R. E. & Hastie, N. D. (1987) Nature (London) 326, 96-99. 30. Hughes, A. L. & Nei, M. (1989) Proc. Natl. Acad. Sci. USA 86,

958-962.