Molecular cloning, expression and characterization of three short chain α-neurotoxins from the venom of sea snake—Hydrophiinae Hydrophis cyanocinctus Daudin

Molecular cloning, expression and characterization of three short chain α-neurotoxins from the venom of sea snake—Hydrophiinae Hydrophis cyanocinctus Daudin

Toxicon 42 (2003) 753–761 www.elsevier.com/locate/toxicon Molecular cloning, expression and characterization of three short chain a-neurotoxins from ...

405KB Sizes 4 Downloads 66 Views

Toxicon 42 (2003) 753–761 www.elsevier.com/locate/toxicon

Molecular cloning, expression and characterization of three short chain a-neurotoxins from the venom of sea snake—Hydrophiinae Hydrophis cyanocinctus Daudin Li-Sheng Peng, Xiao-Fen Zhong, Yu-Shan Huang, Yuan Zhang, Sui-Lan Zheng, Jian-Wen Wei, Wen-Yan Wu, An-Long Xu* The Open Laboratory for Marine Functional Genomics of State High-Tech Development, Department of Biochemistry, College of Life Science, Sun Yat-Sen University, Guangzhou 510275, People’s Republic of China Received 26 June 2003; revised 14 September 2003; accepted 3 October 2003

Abstract Three different genes named sn311, sn316 and sn285 were discovered by large-scale randomly sequencing the high quality cDNA library of the venom glands from Hydrophiinae Hydrophis cyanocinctus Daudin. Sequence analysis showed that these three genes encoded three different short chain a-neurotoxins of 81 amino acids, which contained a signal peptide of 21 amino acids and followed by a mature peptide of 60 amino acids. Amino acid comparison reveals that mature peptides of sn311 and sn316 are highly homologous, with the only variance at position 46, which is Lys46 and Ser46, respectively. Whereas the mature peptide of sn285 lacks the most conserved amino acids in short chain a-neurotoxins, Asp31 and Arg33. The coding sequences of three neurotoxins were cloned into a thioredoxin (TRX) fusion expression vector (pTRX) and expressed as soluble recombinant fusion proteins in E. coli. After purification, approximately 10 mg/l recombinant proteins with the purity up to 95% were obtained. These three recombinant proteins are designated as rSN311, rSN316 and rSN285, they have a molecular weight of 6.963, 6.920 and 6.756 kDa, respectively, which are similar to those predicted from amino acid sequences. LD50 values of rSN311, rSN316 and rSN285 are 0.0827, 0.095, and 0.0647 mg/kg to mice, respectively. Studies on effects of these recombinant proteins on neuromuscular transmission were carried out, and results indicate that they all can produce prompt blockade of neuromuscular transmission, but display distinct biological activity characteristic individually. The results from UV-circular dichroism (CD) spectra indicate that they share similar secondary structure compared to other identified aneurotoxins, and no significant structural differences in these recombinant proteins are observed. q 2003 Elsevier Ltd. All rights reserved. Keywords: Hydrophiinae Hydrophis cyanocinctus Daudin; Short chain a-neurotoxins; Fusion expression; LD50; Blockade; Neuromuscular transmission

1. Introduction Most sea snakes are highly poisonous and the lethal toxicities of venom from sea snakes glands are much higher and even more stronger than those from cobras and crotalids, yet molecular compositions of sea snake venoms * Corresponding author. Tel.: þ86-20-84113655; fax: þ 86-2084038377. E-mail address: [email protected] (A.-L. Xu). 0041-0101/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.toxicon.2003.10.002

are more simple than those of land snakes (Chikahisa, 1998). Snake venom contains small toxic proteins among which the main lethal factors are some basic proteins named as postsynaptic neurotoxins (also called as a-neurotoxins). a-neurotoxins can block nerve transmission in the postsynaptic membranes of skeletal muscles by specifically binding to nicotinic acetylcholine receptors on postsynaptic motor end-plate in vertebrates or electric organs in fishes, thus subsequently influence physiological processes commonly elicited by nicotinic acetylcholine binding to motor

754

L.-S. Peng et al. / Toxicon 42 (2003) 753–761

end-plates such as muscle paralysis and respiratory failure (Endo and Tamiya, 1993). On the basis of the numbers of amino acid residues and disulfide bridges, a-neurotoxins can be classified into two subgroups: short chain and long chain a-neurotoxins (Endo and Tamiya, 1993). Short chain a-neurotoxins consist of 60 – 62 amino acids with four disulfide bonds in common positions whereas most of the long chaina-neurotoxins contain 70 – 74 amino acids with five disulfide bridges. Long chain a-neurotoxins have an extra C-terminal tail with 5– 9 amino acids and an extra disulfide bridge which now are essential determinants for long chain a-neurotoxins interaction with neuronal nicotinic acetylcholine receptors (Servent et al., 1997). Although the primary structures of short chain and long chain aneurotoxins are diverse, they share a common structure of three-finger conformation with a hydrophobic core formed by four disulfide bridges, and three loops I, II, III are protruded from this core (Tsetlin, 1999). As for short chain a-neurotoxins, loop I consists of amino acid residues from position 4 to 16, loop II contains amino acid residues from 25 to 40, and loop III contains amino acid residues from 44 to 53 (Afifiyan et al., 1998). All three loops involve in the neurotoxin interaction with the receptors, but interaction occurs mainly on the surface of loops II and II (Tremeau et al., 1995). Previous studies on the neurotoxins binding to receptors on the basis of site-directed mutagenesis and chemical modification demonstrated that at least 10 residues in Erabutoxin a played a critical role in binding the receptors. Among these 10 residues, Lys27, Trp29, Asp31, Arg33, Glu38and Lys47 are conserved both in short chain and long chain a-neurotoxins, whereas Ser8, Gln7, Gln10 and Ile36 are more variable (Tremeau et al., 1995; Pillet et al., 1993). Any variance of these functionally important residues may cause 10 to 100-fold affinity decrease significantly (Ducancel et al., 1996). Compared to long chain a-neurotoxins, short chain a-neurotoxins display higher homology to each other, they have many natural isoforms, which might serve as an important tool for the digestion of different food preys (Gong et al., 1999). Sea snake Hydrophiidae can be subdivided into Laticauda, Aipysurus and Hydrophis (Rasmussen, 1997). Hydrophiinae Hydrophis cyanocinctus Daudin is one of the Hydrophis, which widely exists in the warm drainage area of South China Sea and Indian Ocean. Snake neurotoxins obtained by chemical extraction from venoms are with high cost and low purity. Since 1995, when gene encoding toxic protein of Dendroaspis angusticeps was cloned and expressed (Smith et al., 1995), many attempts have been made about the gene engineering production of snake neurotoxins or other toxic proteins (Afifiyan et al., 1998; Gong et al., 1999), but few was focused on the sea snake neurotoxins engineering. We now report on three short chain a-neurotoxins which are discovered from large scale sequencing of the cDNA library from the venoms of Hydrophiinae H. cyanocinctus Daudin. They are cloned, and expressed as soluble proteins in E. coli. Functional studies

on these purified recombinant neurotoxins are carried out, interestingly, biological results together with amino acid sequences comparison suggest that many amino residues in a-neurotoxins could affect the biological activities, other than those residues previously found to be functionally important.

2. Material and methods 2.1. Hydrophiinae Hydrophis cyanocinctus Daudin Sea snake Hydrophiinae H. cyanocinctus Daudin was collected from Beihai, Guangxi Autonomous Region of Chuang, People’s Republic of China. Venom glands were immediately dissected and homogenized in liquid nitrogen and stored at 2 70 8C after collection. 2.2. Construction of cDNA library Total RNA was extracted from homogenized venom glands from H. cyanocinctus Daudin using one-stepextraction method (Sambrook et al., 1989). mRNA extraction was carried out using mRNA Purification Kit (Amersham Pharmacia). cDNA library was constructed using TimeSaver cDNA Synthesis Kit (Amersham Pharmacia) according to manufacturer’s protocol. 2.3. Sequence analysis Double stranded DNA sequencing was carried out with the ABI PRISMw BigDyee Terminator Cycle Sequencing Ready Reaction Kit (PE Biosystems) using DNA automated sequencer (PE model 377) with universal T7, SP6, M13 forward and reverse primers. The clones were sequenced completely on both strands, more than three different cDNA clones of the same gene were sequenced to ensure the accuracy of gene sequences, considering possible error introduced by cDNA synthesis and DNA sequencing. Searches for the nucleotide sequence homology were performed using the program BLAST provided by Genbank. Software CLUSTALW was used for nucleotide sequence and amino acid sequence alignment (http://www.ebi.au.uk/ clustalw). Amino acid sequence prediction and protein molecular weight calculation were performed by using DNATOOL5.1. Protein secondary structure was predicted by using the Predictproteins Server (http://cubic.bioc. columbia.edu/predictprotein). 2.4. Construction of fusion expression plasmids Three cDNA clones encoding three different short chain a-neurotoxins were obtained by randomly sequencing the cDNA library. A pair of primers were designed according to the 30 and 50 sequences of the three genes encoding

L.-S. Peng et al. / Toxicon 42 (2003) 753–761

755

the mature peptide, the forward primer was: 50 -CGG GGT ACC GAC GAC GAC GAT AAA ACA TGT TGC AAC CAA CAG TC-30 , and the reverse primer was: 50 -CGC GGA TCC TTA ATT GTT GCA TTC GTT TGT ATG TC-30 . Restriction site Kpn I and protease 3C cleavage site were incorporated in the forward 50 primer, Bam H I was also incorporated in the reverse 30 primer, which allowed PCR products to be subcloned into vectors at the two restriction sites. PCR products were then digested with Kpn I and Bam H I and ligated to Bluescript M13 (2)/Kpn I þ Bam H I. Sequences of subclones were confirmed by DNA sequencing. Later three gene fragments that encoded the mature peptide of three short chain a-neurotoxins were obtained by double digesting the plasmids containing these genes, and cloned at sites of Kpn I and Not I of the expression vector pTRX. PCR amplification and gene cloning were performed using methods previously described (Sambrook et al., 1989).

Von-Jagow, 1987), followed by staining with Coomassie blue and destaining.

2.5. Expression and purification of recombinant short chain neurotoxins

Lethal toxicity and phrenic nerve-diaphragm muscle preparation were performed as previously described (Xu et al., 1994). Briefly, Kunming mice (20 ^ 2 g) were divided into five groups with males and females each in one group. Lyophilized protein samples were diluted with phosphate buffer to different test concentration, and 0.2 ml of diluted samples were injected intraperitioneally into mice with 0.1 ml of saline buffer as negative control. Mice were observed for the next 48 h after injection, only mice died within 1 h were counted for lethal toxicity assay. The LD50 values were calculated according to Bliss method. Sprague-Dawley (SD) rat was used for phrenic nerve– diaphragm muscle test. Adult SD rat was killed by cervical vertebrate dislocation, and diaphragm muscle was immediately isolated with conterminal phrenic nerve and kept in Krebs buffer (120 mM NaCl, 4.8 mM KCl, 1.2 mM MgSO4·7H2O, 1.2 mM KH2PO4, 2.5 mM CaCl2 and 5.8 mM glucose). The Krebs buffer was maintained at 35 8C and the mix air of 95% oxygen and 5% carbon dioxide was continuously saturated into buffer. Phrenic nerve was stimulated at 0.5 Hz with pulse of 0.5 ms duration and supermaximal voltage of 5 V, muscle contraction was recorded via a LMB-2B transducer and recorder. Lyophilized protein samples were dissolved with 50 mM PB, 150 mM NaCl pH7.0 to a working concentration of 1 mM. D-Tubocurarine (d-TC) was added to block neuromuscular transmission, and exogenous acetylcholine, ACh (50 mM) was also added to serve as a competitive inhibitor.

Bacterial host used for expression of recombinant fusion proteins was E. coli strain BL21 (DE3). Bacterial transfected with expression plasmids were grown at 37 8C to middle log phase in LB broth supplemented with 100 mg/ml ampicillin, then induced with IPTG and glucose to a final concentration at 1 mM and 0.2% (v/v), respectively, then cells were further incubated at 19 8C for 10 h. Cell pellets were harvested by centrifugation and resuspended in phosphate buffer saline (50 mM PB, 500 mM NaCl, pH7.8). Cell suspension was sonicated on ice for six cycles totally, and each cycle consisted of 99 separate burst of 4 s with interval of 4 s. After sonication, greater than 90% of the expressed recombinant fusion proteins were obtained as soluble proteins in the supernatant as detected by SDS-PAGE. The extracts were discarded after centrifugation at 6000 g for 30 min, then supernatant fluid was applied to a previously equilibrated Ni2þ Chelating Sepharose Fast Flow column. The fusion proteins were eluted with the buffer containing 50 mM PB, 500 mM NaCl, 500 mM imidazole, pH6.0, and protein concentration was estimated by UV absorption at 280 nm. Fusion protein fractions were collected and isolated by applying to a Sephadex G-25 with cleavage buffer containing 50 mM Tris, 150 mM NaCl, 1 mM EDTA, pH8.0. Eluted proteins were cleaved at 4 8C overnight, in the presence of 1/400 (v/v) precision protease 3C, and DTT was also added to a final concentration at 1 mM. After cleavage, sample proteins were subjected to an anion exchange CM Sepharose Fast Flow column equilibrated with cleavage buffer, then eluted with the buffer containing 50 mM PB, 300 mM NaCl, pH7.0. Proteins were later applied to a Sephadex G-50 for further purification and removal of DTT from the cleavage buffer. Purified proteins were lyophilized for 48 h on BETA1-8K (CHRIST). Tris/Tricine SDS-PAGE was carried out on a Bio-Rad system to detect the proteins (Schagger and

2.6. Secondary structure determination Circular dichroism (CD) spectra were used to determine the secondary structure of purified recombinant short chain a-neurotoxins. CD analysis was performed on a J710 Spectropolarimeter (Japan Jasco) at 20 8C in the far UV region (190 – 250 nm) in a 1 mm length light path quartz. Protein samples were dissolved in 0.1 M phosphate buffer at pH7.0, and the concentration was maintained at 4.55 mM. Molecular weight of purified proteins was determined by MALDI-TOF mass spectrometry, measurements were carried out on a REFLEX III (Germany Bruker) and Sinapinic was used as a matrix. 2.7. Bioassays

3. Results 3.1. Sequence analysis A high quality cDNA library of the venom glands from H. cyanocinctus Daudin was successfully constructed using

756

L.-S. Peng et al. / Toxicon 42 (2003) 753–761

Fig. 1. (A) Nucleotide sequences and deduced amino acid sequences of sn311, sn316 and sn285, The amino acid sequences of the three cDNAs are shown below the nucleotide sequences, only the differences of amino acids are indicated. Underlines indicate the sequences that encode a signal peptide of 21 amino acids. (B) Alignment of the deduced amino acid sequences of the short chain neurotoxins SN311, SN316 and SN285 from H. cyanocinctus Daudin with the sequences of other short chain neurotoxins reported. These sequences, which were all sea snake short chain neurotoxins, are as follows: Astrotia stokesii toxin a (Maeda and Tamiya, 1978); Acalyptophis peronii neurotoxin major (Mori and Tu, 1988); H. cyanocinctus Short neurotoxin 1 (Hydrophitoxin A) (Liu and Blackwell, 1974); Hydrophiinae H. cyanocinctus Daudin short chain neurotoxin SN311, Hydrophilnae, Lapemis hardwickii Gray short chain neurotoxin SN36, SN160, SN12 (Zhong et al., 2001); Enhydrina schistose toxin 5 (Fryklund et al., 1972); Hydrophiinae H. cyanocinctus Daudin short neurotoxin SN316; Aipysurus laevis toxin a (Maeda and Tamiya, 1976); H. lapemoides short neurotoxin 1 (Neurotoxin A) (Tamiya et al., 1983); Hydrophiinae H. cyanocinctus Daudin short neurotoxin SN285; Laticauda laticaudata toxin c (Tamiya et al., 1983); Laticauda laticaudata toxin a0 (Tamiya et al., 1983); Laticauda laticaudata toxin b (Tamiya et al., 1983); Laticauda colubrine toxin II (Tamiya et al., 1983); Laticauda colubrine toxin d (Tamiya et al., 1983); Laticauda semifasciata Neurotoxin b (Endo et al., 1986); Laticauda semifasciata Erabutoxin a (Nastopoulos et al., 1998). The identical residues, similar residues are indicated by ‘*’, ‘:’ and ’.’, respectively.

L.-S. Peng et al. / Toxicon 42 (2003) 753–761

757

Fig. 1 (continued )

5 mg intact mRNA (data not shown). Three full-length cDNAs encoding three short chain a-neurotoxins, named sn311, sn316 and sn285, were discovered by randomly sequencing the cDNA library. Based on the nucleotide sequence and amino acid sequence analyses, these three cDNAs share very high similarities with identity up to 90%, they all contain an open reading frame (ORF) of 243 bp, which encodes three different precursor proteins of 81 amino acids (Fig. 1). Protein analysis using CLUSTALW showed that they all shared the same signal peptide of 21 amino acids, followed by different mature peptides of 60 amino acids. The signal peptide matches with those of short chain a-neurotoxins from other snake species. The mature proteins are designated as SN311, SN316 and SN285, respectively. SN311 and SN316 share a high similarity with only one residue variance occurred at position 46, which is Lys46 in SN311 and Ser46 in SN316. Whereas SN285 displays lower similarity to the other two mature proteins, and varies from position 28 – 31 and 33 – 35 compared to the other two. SN285 lacks the most conserved amino residues Asp31 and Arg33, which have been identified to be functionally important to neurotoxins binding to receptors reported for other similar a-neurotoxins (Tremeau et al., 1995). Alignment analysis revealed that these newly found short chain a-neurotoxins possessed the typical primary structure of short chain a-neurotoxins, and belonged to the three-finger family. The PI values of SN311 SN316 and SN285 were deduced to be 8.225, 8.025 and 8.025, respectively, according to the amino acid sequences.

3.2. Expression and purification of recombinant proteins Comparative analysis of the determined cDNA sequences of these three a-neurotoxins showed that the nucleotide sequences of the 50 and 30 portion of the coding regions were highly conserved, thus two primers were designed from these conserved regions to amplify the sequences encoding the mature peptides. Kpn I was incorporated in the 50 primer and Bam H I in the 30 primer, so as to allow PCR products to be subcloned into vector Bluescript M13 (2). In addition, a protease 3C cleavage site was incorporated in the downstream Kpn I site of 50 primer, which facilitated the obtaining of purified foreign proteins by removal of TRX partner and 6 £ HIS tag from fusion proteins. The PCR products were estimated to be about 225 bp and the nucleotide sequences were further confirmed by DNA sequencing and enzyme digestion. The correct plasmids containing the cloned PCR products were double digested by Kpn I and Not I and cloned into expression vector pTRX. Then recombinant proteins were expressed as fusion proteins in E. coli strain BL21 (DE3). Fusion expression vector pTRX contains a 6 £ HIS tag facilitating the purification of fusion proteins on a Ni2þ Chelating Sepharose column, and partner TRX were coexpressed with foreign proteins, so as to help the proteins correctly folding and expressing as soluble proteins in E. coli. Thus the expressed fusion proteins comprised a 6 £ HIS tag, a TRX partner and a protease 3C cleavage site, followed by target proteins. Since the amino acid

758

L.-S. Peng et al. / Toxicon 42 (2003) 753–761

sequences of three mature proteins were homologous to each other, the same procedure was used to purify these target proteins. Bacteria containing these genes were induced in log phase by addition of IPTG. Different induction conditions were tested for optimum fusion protein expression, it was revealed that proteins obtained under lower temperature and lower IPTG concentration inductions were better folded and had a more soluble form. So the induction were carried out by adding IPTG to a final concentration of 0.1 mM and the bacteria was further incubated at 19 8C for 10 h. The fusion proteins had a similar molecular weight of 20 kDa observed by Tris/Tricine SDS-PAGE, which agreed with that predicted from sequences. Thin layer scan showed that expressed fusion proteins constituted approximately 20% of total bacteria proteins. Bacteria cells were lysed by sonication, more than 90% of the fusion proteins were recovered in the supernatant, soluble fusion proteins were primarily purified by applying supernatant to a Ni2þ Chelating Sepharose column, pooled by elution with imidazole, TRX partner and 6 £ HIS tag were removed from the fusion proteins by digesting with protease 3C at 4 8C overnight. Two proteins of approximately 13.6 and 6.8 kDa were presented in the Tris/Tricine SDS-PAGE gel, which were in consistent with the molecular weights of TRX and target proteins. The result showed that the majority of the fusion proteins were properly cut into target proteins and TRX. Further purification was performed by applying cleaved proteins to a CM Sepharose column, followed by a G-50 Sephadex column. Ultimately, greater than 10 mg recombinant short chain a-neurotoxins were obtained from the primary 1L culture with the purity up to 95% (as detected by Tris/Tricine SDS-PAGE, shown in Fig. 2). The three purified proteins were designated as rSN311, rSN316 and rSN285, respectively.

Fig. 2. Tris/Tricine SDS-PAGE analysis of rSN311. rSN316 and rSN285; Lane 1 total proteins of BL21 (DE3)-pETTRXSN285; Lane 2 total soluble proteins of BL21 (DE3)-pETTRX-rSN285; Lane 3 the purified fusion protein of TRX-rSN285; Lane 4 the products of the fusion protein TRX-rSN285 digested by PreScission protease 3C; Lane 5 purified recombinant short chain neurotoxin rSN285; Lane 6 purified recombinant short chain neurotoxin rSN311; Lane 7 purified recombinant short chain neurotoxin rSN316; Lane 8 prestained protein marker (NEB).

Fig. 3. Far-UV circular dichroism spectra of three recombinant aneurotoxins rSN311, rSN316 and rSN285 carried out at 20 8C. CD measurements were performed on a Jasco J710 Spectropolarimeter in the far UV region (190– 250 nm) in a 1 mm length light path quartz. Protein samples were dissolved in 0.1 M phosphate buffer at pH7.0, and the concentration was maintained at 4.55 mM.

3.3. Secondary structure determination of rSN311, rSN316 and rSN285 As shown in Fig. 3, rSN311, rSN316 and rSN285 displayed similar secondary structure pattern, and prominent triple-stranded b-sheets were observed with little or no a-helical, which was in accordance with the known aneurotoxins structure conformation (Tsetlin, 1999). No significant differences in the CD spectra of the three purified proteins rSN311, rSN316 and rSN285 were discovered. The molecular weights of rSN311, rSN316 and rSN285 were 6.963, 6.920 and 6.756 kDa, respectively, determined by MALDI-TOF, which were close to those calculated molecular masses based on the amino acid sequences. 3.4. Bioassays Lethal toxicity was performed on Kunming mice by intraperitioneal injection, the LD50 values of rSN311, rSN316 and rSN285 were 0.0827, 0.095 and 0.0647 mg/kg, respectively. The LD50 results showed that three recombinant a-neurotoxins had similar lethal toxicities to mice, but the toxicity of rSN285 to mice was slightly higher than the other two. The LD50 values were in correspondence with that of other a-neurotoxins previously described (Atassi, 1995). Postsynaptic neurotoxins could block nerve transmission by specially binding to the nicotinic acetylcholine receptors (nAChRs), but the binding to receptors do not cause the opening of ion channels, thus muscle contraction cannot be observed during this blocking process. Three purified proteins rSN311, rSN316 and rSN285 all produced prompt blockade of neuromuscular transmission within 1 h. Preliminarily d-TC was used to test the normal activity of nAChRs in SD rats, and common muscle contraction response was recovered after d-TC was washed out. As shown in Fig. 4A, B and C, results revealed that rSN311, rSN316 and rSN285 all could completely block muscle contraction evoked either by stimulating the associated nerve

L.-S. Peng et al. / Toxicon 42 (2003) 753–761

759

Fig. 4. Effects of rSN311, rSN316 and rSN285 on the SD rat phrenic nerve–muscle preparation, (A) 1x mM rSN311; (B) 1 mM rSN316; (C) 1 mM rSN285; ACh: 50 mM. Muscle in the figure represents direct diaphragm muscle stimulation.

or by addition of exogenous ACh (50 mM), but could not reduce the response evoked by direct muscle stimulation. rSN311, rSN316 and rSN285 displayed distinct physiological functions of blocking neuromuscular transmission in the experiments. Neuromuscular block evoked by rSN311 resulted in prompt completely blockage and muscle did not restore even after continuously washing with Krebs buffer for 7 h, this indicated that rSN311 rapidly bound nAChRs, causing the interruption of nerve transmission and the blocked process was irreversible. Whereas completely blockage was also observed 30 min later after addition of rSN316 to the buffer, when toxins were washed out, muscle response was slowly recovered, and finally reached 50% of the control after washing for 6 h. Results showed that dissociation of rSN316 from nAChRs was slow, but neuromuscular transmission blockade was reversible, although it was only partly recovered. Similar pattern was observed with rSN285, complete block occurred at 60 min after addition of toxins. When toxins were washed out, muscle response rehabilitated rapidly in 10 min, reached 50% of control after washing for 1 h. This revealed that rSN285 bound to nAChRs slowly and dissociated rapidly. All the experiments had repeated for at least three times, similar results were achieved.

4. Disussion In this report, we have discovered three new genes designated as sn311, sn316 and sn285 by large-scale

randomly sequencing the cDNA library of the venom glands from Hydrophiinae H. cyanocinctus Daudin. Approximately 500 cDNA clones were sequenced, among them, 269 cDNA clones were found to encode toxins, including Phospholipase A and a-neurotoxins. These three cDNAs account 33.5% of cDNA library clones, suggesting that short chain a-neurotoxins were main components of toxin proteins from the sea snake venom. They all encode short chain a-neurotoxins with mature peptides of 60 amino acids, the encoded proteins of sn311 and sn316 are highly homologous with the only one variance of 46th residue, which is Lys46 and Ser46, respectively. However, the mature protein of sn285 does not have Asp31 and Arg33, two conserved amino acid residues of the common short chain a-neurotoxins. They are defined to be natural a-neurotoxins isomers, their residue variances all contribute to the one or two changes in nucleotides. Nucleotide substitutions can occur at any position of codons for variable amino acids, AAA in sn311 changes to AGC in sn316 causing residue replacement of Lys with Ser. In sn285, codons encoded the 31st and 33rd residues change from CGT and ACT to CAT and TCT, leading to the replacement of Asp31 and Arg33 with Gly31 and His33, thus non-synonymous nucleotide substitutions that cause amino residues changes happen more frequently here. It is demonstrated that non-coding regions of different neurotoxins are highly conserved compared to coding region, and only one nucleotide change is found to result in coded amino residue variance in coding region (Gong et al., 2000). A similar observation has been

760

L.-S. Peng et al. / Toxicon 42 (2003) 753–761

made in our report. Short chain a-neurotoxins contain a considerable amount of isomers, six isomers of same gene were even found in the venom of same snake, other isomers are also observed in Serine protease (Deshimaru et al., 1996), Phospholipase A (Ogawa et al., 1995), Cardiotoxins (Chang et al., 1997) from other species of snakes. One of the main functions of snake toxin is to digest preys (Davidson and Dennis, 1990), toxin isomers may contribute to different preys in nature. It was also pointed out that the diversity of snake toxin composition in habitat might correlate with diversity of preys under the adaptive pressure, presence of various short chain a-neurotoxins isomers is due to changes in their preys (Daltry et al., 1996). It was assumed that a kind of accelerated evolution was undertaken to adapt to the diversity of preys (Gong et al., 1999). Thus it is possible that a similar accelerated evolution pattern has been adopted in all genes that encoded the short chain neurotoxins of Hydrophiinae H. cyanocinctus Daudin. The LD50 results of purified recombinant neurotoxins rSN311, rSN316 and rSN285 show that their lethal toxicities are similar, the values are all within the range of those of other identified a-neurotoxins, with a typical LD50 value of mice between 50 and 150 mg/kg (Atassi, 1995). Also rSN311, rSN316 and rSN285 exerted strong blockade of neuromuscular transmission by interaction with nAChRs, and addition of exogenous ACh or direct stimulation of the nerve could not restore the block process. Among them, binding of rSN311 to nAChRs is irreversible, whereas binding of rSN316 and rSN285 to receptors is reversible. rSN285 slowly binds to receptor and rapidly dissociates from it, rSN316 quickly interacts with receptors and slowly dissociates. The binding patterns of these three recombinant proteins to the receptors showed interesting differences. Previous studies have revealed that at least 10 amino acids are functionally important for neurotoxins such as Gln7, Ser8, Gln10, Lys27, Trp29, Asp31, Arg33, Ile36, Glu38 and Lys47. Among these, Lys27, Trp29, Asp31, Arg33, Glu38and Lys47 are conserved both in short chain and long chain aneurotoxins (Tremeau et al., 1995; Pillet et al., 1993). It was proposed that these conserved residues formed a receptor interaction core consisted of basic and aromatic residues in neurotoxins, any individual aberrance of binding affinity to receptors among different neurotoxins was due to various specific residues appeared in the functional core (Antil et al., 1999). In our report, the only variance occurs between rSN311 and rSN316 is the 46th residue, which is near the functionally important residue Lys47 (Ducancel et al., 1996), but this amino acid difference does not cause significant toxicity decrease. In addition, amino acids comparison showed that rSN285 lacked the common functionally important residues Asp31 and Arg33, but the replacement of Gly31 and His33 do not cause any changes to the biological activity. Sequence analysis also indicates that other than these two conserved amino residues, changes in other amino acids nearby also occur likewise. But CD results reveal that the secondary structure of rSN285 is

similar to other a-neurotoxins, the secondary structure of protein is predominantly organized into b-sheets, no significant changes are observed. We found that changes of several nearby amino acids along with the previously identified important residues (Asp31 and Arg33) did not change the function and binding affinity of rSN285, suggesting no conformation changes for the protein functional binding core, thus rSN285 still maintains high toxicity. A new group of neurotoxins named Pt-sntxs from P. Textilis was identified (Gong et al., 1999). These short chain neurotoxins lack of some conserved residues which are found to be functionally important in other identified short chain a-neurotoxins, such as Gln7, Ser8, Gln10, Lys27 and Arg33. Amino acid deletions were also observed in primary structure, which makes them the shortest aneurotoxins of all the identified neurotoxins with only 57 – 58 amino acids. The LD50 values of these new neurotoxins were distinctly lower than those of other identified aneurotoxins. Thus, it was pointed out that the lower functional activities could be caused by many amino residues changes in primary structure and the changes in the conformation of the mature proteins (Gong et al., 1999). Overall, functional activities of short chain a-neurotoxins could be affected by many amino residues other than only one, which could create a combined effect on the binding of the toxins to the receptors. Further structural analysis based on NMR and X-ray crystallography and receptor binding assays will shed more light on understanding the structurefunction relationship for neurotoxins.

Acknowledgements This work was supported by State Hi-Tech Development Project (No. 2001AA62-4120) and (No. 2001AA62-6010) of the Ministry of Science and Technology of China and key project (No. 69935620) of National Natural Science Foundation of China.

References Afifiyan, F., Armugam, A., Gopalakrishnakone, P., Tan, N.H., Tan, C.H., Jeyaseelan, K., 1998. Four new postsynaptic neurotoxins from Naja naja sputatrix venom: cDNA cloning, protein expression, and phylogenetic analysis. Toxicon 36 (12), 1871–1885. Antil, S., Servent, D., Menez, A., 1999. Variability among the sites by which curaremimetic toxins bind to torpedo acetylcholine receptor, as revealed by identification of the functional residues of alpha-cobratoxin. J. Biol. Chem. 274 (49), 34851–34858. Atassi, M.Z., 1995. Postsynaptic-neurotoxin-acetylcholine receptor interaction and the binding sites on the two molecules. In: Tu, A.T., (Ed.), Handbook of Natural Toxins 5: Reptile Venoms and Toxins, Marcel Dekker, New York, pp. 53–83.

L.-S. Peng et al. / Toxicon 42 (2003) 753–761 Chang, L.S., Lin, J., Chou, Y.C., Hong, E., 1997. Genomic structures of cardiotoxin 4 and cobratoxin from Naja naja atra (Taiwan cobra). Biochem. Biophys. Res. 239 (3), 756–762. Chikahisa, T., 1998. The toxinology of sea snake venoms. J. Toxicol. Toxin Rev. 17 (3), 361– 372. Daltry, J.C., Wuster, W., Thorpe, R.S., 1996. Diet and snake venom evolution. Nature 379, 537–540. Davidson, F.F., Dennis, E.A., 1990. Evolutionary relationships and implications for the regulation of phospholipase A2 from snake venom to human secreted forms. J. Mol. Evol. 31 (3), 228–238. Deshimaru, M., Ogawa, T., Nakashima, K., Nobuhisa, I., Chijiwa, T., Shimohigashi, Y., Fukumaki, Y., Niwa, M., Yamashina, I., Hattori, S., Ohno, M., 1996. Accelerated evolution of crotalinae snake venom gland serine proteases. FEBS Lett. 397 (1), 83– 88. Ducancel, F., Merienne, K., Fromen-Romano, C., Tremeau, O., Pillet, L., Drevet, P., Zinn-Justin, S., Boulain, J.C., Menez, A., 1996. Mimicry between receptors and antibodies. Identification of snake toxin determinants recognized by the acetylcholine receptor and an acetylcholine receptor-mimicking monoclonal antibody. J. Biol. Chem. 271 (49), 31345–31353. Endo, T., Tamiya, N., 1993. Structure–function relationship of postsynaptic neurotoxins from snake venoms. In: Harvey, A.L., (Ed.), Snake Toxins, Pergamon press, New York, NY, pp. 165–222. Endo, T., Nakanishi, M., Furukawa, S., Joubert, F.J., Tamiya, N., Hayashi, K., 1986. Stopped-flow fluorescence studies on binding kinetics of neurotoxins with acetylcholine receptor. Biochemistry 25 (2), 395–404. Fryklund, L., Eaker, D., Karlsson, E., 1972. Amino acid sequences of the two principal neurotoxins of Enhydrina schistosa venom. Biochemistry 11 (24), 4633–4640. Gong, N.L., Armugam, A., Jeyaseelan, K., 1999. Postsynaptic short-chain neurotoxins from Pseudonaja textilis. cDNA cloning, expression and protein characterization. Eur. J. Biochem. 265 (3), 982–989. Gong, N.L., Armugam, A., Jeyaseelan, K., 2000. Molecular cloning, characterization and evolution of the gene encoding a new group of short-chain alpha-neurotoxins in an Australian elapid, Pseudonaja textilis. FEBS Lett. 473 (3), 303–310. Liu, C.S., Blackwell, R.Q., 1974. Hydrophitoxin b from Hydrophis cyanocinctus venom. Toxicon 12 (5), 543 –546. Maeda, N., Tamiya, N., 1976. Isolation, properties and amino acid sequences of three neurotoxins from the venom of a sea snake, Aipysurus laevis. Biochem. J. 153 (1), 79 –87. Maeda, N., Tamiya, N., 1978. Three neurotoxins from the venom of a sea snake Astrotia stokesii, including two long-chain neurotoxic proteins with amidated C-termini. Biochem. J. 175 (2), 507–517. Mori, N., Tu, A.T., 1988. Isolation and primary structure of the major toxin from sea snake, Acalyptophis peronii, venom. Arch. Biochem. Biophys. 260 (1), 10–17.

761

Nastopoulos, V., Kanellopoulos, P.N., Tsernoglou, D., 1998. Structure of dimeric and monomeric erabutoxin a refined at 1.5 A resolution. Acta Crystallogr. D Biol. Crystallogr. 54 (Pt 5), 964 –974. Ogawa, T., Kitajima, M., Nakashima, K., Sakaki, Y., Ohno, M., 1995. Molecular evolution of group II phospholipases A2. J. Mol. Evol. 41 (6), 867–877. Pillet, L., Tremeau, O., Ducancel, F., Drevet, P., Zinn-Justin, S., Pinkasfeld, S., Boulain, J.C., Menez, A., 1993. Genetic engineering of snake toxins. Role of invariant residues in the structural and functional properties of a curaremimetic toxin, as probed by sitedirected mutagenesis. J. Biol. Chem. 268 (2), 909–916. Rasmussen, A.R., 1997. Systematics of sea snakes: a critical review. In: Thorpe, R.S., Wu¨ster, W., Malhotra, A. (Eds.), Venomous snakes-ecology, evolution and snakebite, Symp. Zool. Soc. Lond., Clarendon Press, Oxford, pp. 15–30. Sambrook, J., Fristch, E.F., Maniatis, T., 1989. Isolation of mRNA in eukaryotic cell. In: Sambrook, J., Fristch, E.F., Maniatis, T. (Eds.), Molecular Cloning: A Laboratory Manual, Cold Spring Harbor, New York, pp. 345–356. Schagger, H., Von-Jagow, G., 1987. Tricine–sodium dodecylsulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368 –379. Servent, D., Winckler-Dietrich, V., Hu, H.Y., Kessler, P., Drevet, P., Bertrand, D., Menez, A., 1997. Only snake curaremimetic toxins with a fifth disulfide bond have high affinity for the neuronal alpha7 nicotinic receptor. J. Biol. Chem. 272 (39), 24279–24286. Smith, L.A., Olson, M.A., Lafaye, P.J., Dolly, J.O., 1995. Cloning and expression of mamba toxins. Toxicon 33 (4), 459 –474. Tamiya, N., Maeda, N., Cogger, H.G., 1983. Neurotoxins from the venoms of the sea snakes Hydrophis ornatus and Hydrophis lapemoides. Biochem. J. 213 (1), 31–38. Tsetlin, V., 1999. Snake venom a-neurotoxins and other threefinger proteins. Eur. J. Biochem. 264, 281–286. Tremeau, O., Lemaire, C., Drevet, P., Pinkasfeld, S., Ducancel, F., Boulain, J.C., Menez, A., 1995. Genetic engineering of snake toxins. The functional site of Erabutoxin a, as delineated by site-directed mutagenesis, includes variant residues. J. Biol. Chem. 270 (16), 9362 – 9369. Xu, S.Y., Bian, R.L., Chen, X., 1994. Methods in Pharmacology, second ed., People Hygiene Press, Beijing. Zhong, X.F., Peng, L.S., Wu, W.Y., Wei, J.W., Yang, H., Yang, Y.Z., XU, A.L., 2001. Identification and functional characterization of three postsynaptic short-chain neurotoxins from Hydrophiinae, Lapemis hardwickii Gray. Acta. Biochemica. Biophysica. Sinica. 33 (4), 457– 462.