- Email: [email protected]
Expression and Mutagenesis Studies of Cobrotoxin from Taiwan Cobra
Biochemical and Biophysical Research Communications 263, 652– 656 (1999) Article ID bbrc.1999.1418, available online at http://www.idealibrary.com on
Expression and Mutagenesis Studies of Cobrotoxin from Taiwan Cobra Long-sen Chang,* ,1 Ku-chung Chen,† Bin-nan Wu,† Shu-kai Lin,* Pei-fung Wu,† Yi-ren Hong,† and Chen-chung Yang‡ ,1 *Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung 804, Taiwan; †Department of Biochemistry and Pharmacology, Kaohsiung Medical College, Kaohsiung 807, Taiwan; and ‡Department of Life Sciences, National Tsing Hua University, Hsinchu 300, Taiwan, Republic of China
Received July 27, 1999
The cDNA encoding cobrotoxin was constructed from the cellular RNA isolated from the venom glands of Naja naja atra (Taiwan cobra). The cDNA was subcloned into the expression vector pET20b(1) and transformed into BL21(DE3) Escherichia coli strain. Expressed cobrotoxin was isolated from inclusion bodies of E. coli and subjected to refolding into its folded structure. The refolded cobrotoxin was purified by high-performance liquid chromatography and exhibited a neurotoxicity in inhibiting acetylcholineinduced muscle contractions. Recombinant cobrotoxin showed a tendency to isomerize its disulfide bonds as that observed with native cobrotoxin. An appreciable decrease in the rate of isomerization reaction was observed when Glu-38 was replaced with Gln-38 or Lys-47 was replaced with Glu-47 or Gln-47. These results reflect that the element in controlling the disulfide isomerization of cobrotoxin is closely associated with the charged side chains in the cobrotoxin molecule. © 1999 Academic Press
Cobrotoxin is a neurotoxic protein isolated from the venom of Taiwan cobra (Naja naja atra). It is a small, basic protein consisting of a single polypeptide chain of 62 amino acids, cross-linked by four disulfide bonds (1, 2). The four disulfide linkages are Cys3-Cys24, Cys17Cys41, Cys43-Cys54, and Cys55-Cys60, respectively (3). Recent studies show that the disulfide bonds at the C-terminal region of cobrotoxin exhibit a tendency to exchange with each other (4). Two cobrotoxin isomers (cobrotoxin II and cobrotoxin III) have been isolated from the refolding mixtures of cobrotoxin as well as from Taiwan cobra venom. The disulfide linkages at the C-terminus of cobrotoxin II and cobrotoxin III are identified as Cys43-Cys55 & Cys54-Cys60 and Cys431
To whom correspondence may be addressed. Fax: 886-7-5253696. E-mail: [email protected]. 0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Cys-60 & Cys54-Cys55, respectively. Moreover, the isomerization reaction is irreversible. Chemical modification studies indicate that disulfide isomerization of cobrotoxin is, in part driven by the positively charged Lys residues at positions 26, 27, and 47 of the toxin molecule, and decelerated by the negatively charged groups of cobrotoxin (5). These reflect that Lys residues and carboxyl groups constitute the charged environment by which the isomerization reaction is controlled. To clearly elucidate the mechanism of basic and acidic residues in controlling the disulfide isomerization reaction, the best way is to clone the cobrotoxin cDNA and subject it to site-directed mutagenesis. Thus expression and mutagenesis studies of cobrotoxin are carried out in the present study. MATERIALS AND METHODS Preparation of mRNA from venom glands. Cellular RNA was isolated from the snake (Naja naja atra) venom glands which were stored in liquid nitrogen immediately after sacrificed. Two deep frozen glands from one snake were homogenized to extract RNA by a guanidinium isothiocyanate/phenol/chloroform isolation kit (Stratagene Ltd., USA). PCR amplification and cloning. Two oligonucleotide primers of sense and antisense orientations based on the signal peptide and 39noncoding region of cobrotoxin genomic DNA (6) with the forward sequence, 59-ATGAAAACTCTGCTGCTGACCTTGCTG-39 and the reverse one 59-GGATGGTCCTTGATGGATGAGAGC-39 were synthesized. RT-PCR was carried out with 100 ml reaction buffer containing 100 mM Tris–HCl (pH 8.3), 1 mM dNTP, 1 mM antisense primer and 200 ng RNA template. In the reverse transcription, the cDNA was started with rTth reverse transcriptase (5 U) and 2 ml of 10 mM MnCl 2 at 70°C for 15 min, and stopped the reaction by placing the tube on ice until needed. A 8 ml chelating buffer containing 50% glycerol (v/v), 100 mM Tris–HCl (pH 8.3), 1 M KCl and 7.5 mM EGTA/0.5% Tween 20 was added to the reaction. After addition of 8 ml of 25 mM MgCl 2 and 1 mM sense primer, the amplification was proceeded on a thermocycler 94°C/45°C/72°C 1 min each, for a total 30 cycles. The PCR products were cloned into pCRII vector according to the TA-cloning procedures (Invitrogen, San Diego, CA).
652
Vol. 263, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Cloning and expression of cobrotoxin. Synthetic oligonucleotides were designed to produce a 249-bp amplified DNA fragment spanning the open reading frame of cobrotoxin. Primer 1 introduced a 59 EcoRV site and an in frame initiating Met codon preceding Leu-1 of cobrotoxin,
EcoRV 59-GATATCATGCTGGAATGTCACAACCAACAATCATC-39. AspIleMetLeuGluCysHisAsnGlnGlnSer Primer 2 was the reverse primer for RT-PCR amplification of cobrotoxin. The PCR products were cloned into pCR II vector. The inserted DNA fragment was cut with EcoRV and EcoRI and ligated into the large fragment of EcoRV/EcoRI-cut pET20b(1). The entire sequence was confirmed by dideoxynucleotide sequencing. The resulting plasmid pET-CbTx was transformed into E. coli strain BL21(DE3). Transformants were selected on LB-agar plates supplements with 50 mg/ml ampicillin. For induction of gene expression, E. coli BL21(DE3) cells containing pET-CbTx were grown at 37°C in LB medium containing 50 mg/ml ampicillin. After reaching an OD 550 5 1.0, isopropyl-b-D-thiogalactoside (IPTG) was added to a final concentration of 0.2 mM. The culture was induced for a period of up to 4 h. The cells were harvested and lysed by ultrasonication. The recombinant cobrotoxin was found to appear exclusively in the inclusion bodies of E. coli. Inclusion bodies recovered from 1 liter of bacterial culture were thoroughly washed with 10 mM Tris–1 mM EDTA (pH 8.0) containing 0.5% Triton X-100, and sulfonated with disodium 2-nitro-5-(sulfothio)-benzoate in 8 M urea, 0.3 M Na 2SO 3, pH 8.5 (7, 8). The protein was precipitated by 1% acetic acid and dissolved in 10 ml of 50 mM sodium borate (pH 8.5) containing 5 mM EDTA, 8 M urea, and 4 mM reduced and 2 mM oxidized glutathione. Refolding was performed by a four-fold dilution with the same buffer without urea. After standing for 1 day at room temperature, the refolded cobrotoxin was further purified by HPLC. Oligonucleotide-directed mutagenesis. The inserted DNA in pETCbTx was cut with XbaI and EcoRI and ligated into the large fragment of XbaI/EcoRI-cut M13mp19. The entire sequence was confirmed by dideoxynucleotide sequencing. Oligonucleotide-directed mutagenesis was carried out by using a Muta-Gene kit (Bio-Rad) with synthetic oligonucleotides. The oligonucleotide primers E21, K27, E38, K47, and E51 with sequence 59-CAATTGGTCTKCCCCCCTGAACAAC-39, 59CACGCCAACGCTSTTTATAGCAATTG-39, 59-CCACATCCCCTCTKGGTTCTATATCC-39, 59-CAATGCCGTTCTSCACTGAAGGGCAAC-39 and 59-AACAGTTAATTTKAATGCCGTTCTTC-39 were synthesized for preparing the mutants E21K, E21Q, K27E, K27Q, E38K, E38Q, K47E, K47Q, E51K, and E51Q, respectively. The mutations were confirmed by sequencing the gene on M13mp19 using Sequanase kit. The XbaI/EcoRI fragment of the mutated clone was cloned into a pET20b(1) expression vector. Other tests. DNA sequencing, native gel electrophoresis, SDS– polyacrylamide gel electrophoresis (PAGE), immunoblot analysis, analysis of neurotoxicity and disulfide isomerization reaction were performed in essentially the same manner as previously described (4, 5, 8 –10).
RESULTS AND DISCUSSION Expression of recombinant cobrotoxin. The genomic DNA encoding cobrotoxin precursor had been determined in our laboratory previously (6). To clone the cDNA encoding cobrotoxin, two primers were designed from the regions at the beginning of the signal peptide and 39-noncoding region of cobrotoxin genomic DNA, respectively. PCR amplification of the venom gland cDNA mixtures with the designed primers achieved the isolation of a PCR fragment estimated to be about
FIG. 1. Nucleotide and deduced protein sequences of the precursors of cobrotoxin and cobrotoxin homolog. The nucleotide sequence of 306 base pairs is shown above the amino acid sequence of 83 residues including a signal peptide of 21-amino acid residues. The mature cobrotoxin of 62 residues starts at Leu. The cDNA sequence is essentially the same as that deposited in EMBL, GeneBank and DDBJ nucleotide sequence databases under the accession number U42582. The base-substitutions, base-deletions and amino acid substitutions observed between cobrotoxin homolog and cobrotoxin are shown under the sequence of cobrotoxin. The Asp-31 is substituted by Tyr-31 in the cobrotoxin homolog. Moreover, instead of two amino acid residues, Asp-Arg, at positions 58 and 59 of cobrotoxin, there is only a Gly residue in the cobrotoxin homolog. Thus, the deduced protein sequence of cobrotoxin homolog comprises 61 amino acid residues. The cDNA sequence of cobrotoxin homolog with 303 base pairs has been deposited in EMBL, GeneBank and DDBJ nucleotide sequence databases under Accession No. AJ239050.
300 bp (data not shown). The DNA fragments were then subcloned by a TA-cloning kit. More than 20 clones were selected for nucleotide sequencing. Three of the selected clones had the same cDNA sequence, which deduced protein sequence corresponded to that of cobrotoxin (Fig. 1). The deduced protein sequence is the same as the previous published sequence determined by protein sequencing technique (1–3). However, one novel cDNA clone encoding a cobrotoxin homolog was also identified. Its deduced protein sequence comprises 61 amino acid residues. Instead of two amino acid residues, Asp-Arg, at positions 58 and 59 of cobrotoxin, there is only a Gly residue in the cobrotoxin homolog. Moreover, a Tyr-31 appears in the cobrotoxin homolog resulted from one base substitution. To subclone the cDNA into the expression vector, a new primer was designed to create NdeI site in the beginning of the nucleotide sequences for encoding
653
Vol. 263, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
amino acid sequence. The primer sequences 59CATATGCTGGAATGTCACAACCAACAATCATC-39 (the underline indicated the NdeI site) was synthesized for subcloning cobrotoxin cDNA into expression vector pT7-7. Noticeably, the resulting expressed protein should contain only a Met residue preceding the first amino acid of cobrotoxin. The PCR procedure was performed with pfu DNA polymerase. The amplified DNA was inserted into pCRII vector, then subcloned into the expression vector pT7-7 by digestion with NdeI/ HindIII. Unfortunately, no appreciable amount of recombinant protein was produced by inducing BL21(DE-3) containing plasmid pT7-7-CbTx with IPTG. Thus, another primer 59-GATATCATGCTGGAATGTCACAACCAACAATCATC-39 (the underline indicated the EcoRV site) was synthesized for subcloning cobrotoxin cDNA into pET20b(1) expression vector. Since the cobrotoxin cDNA was subcloned into the large fragment of EcoRV/EcoRI-cut pET20b(1). The resulting protein should have a fused N-terminal peptide with the amino acid sequence MKYLLPTAAAGLLLLAAQPAMAMDIM. The expression of recombinant cobrotoxin was observed with BL21(DE-3) containing pET-CbTx plasmid. However, the expressed protein exclusively appeared in the inclusion bodies of E. coli. Thus refolding of the expressed protein was carried out according to the procedure described previously (7, 8). As shown in Fig. 2, the refolded protein
FIG. 2. Purification of recombinant cobrotoxin. The refolded proteins were applied on a SynChropak RP-P column equilibrated with 0.1% TFA and eluted with a linear gradient of 15–50% acetonitrile for 70 min. The flow rate was 0.8 ml/min and the eluate was monitored at 235 nm. The arrow indicates the refolded cobrotoxin. (Inset) SDS–PAGE (lanes 1–3) and immunoblot (lanes 4 – 6) analysis of recombinant cobrotoxin. Lanes 1 and 4, molecular markers (prestained SDS–PAGE standards, BIO-RAD, Hercules, CA): phosphorylase (107,000), BSA (76,000), ovalbumin (52,000), carbonic anhydrase (36,800), soybean trypsin inhibitor (27,200) and lysozyme (19,000); lanes 2 and 5, native cobrotoxin; lanes 3 and 6, recombinant cobrotoxin.
FIG. 3. Typical recordings of the inhibitory effects of native cobrotoxin and recombinant cobrotoxin on acetylcholine-induced maximal contractions in isolated frog rectus abdominis muscle. The added concentrations of acetylcholine and toxins (A, native cobrotoxin; B, recombinant cobrotoxin) are 10 and 0.5 mM, respectively.
was purified by HPLC on a SynChropak RP-P column. The recombinant protein had an apparent molecular weight lower than native cobrotoxin as revealed by SDS–PAGE (inset of Fig. 2). Since previous study revealed that cobrotoxin exhibited an anomalous mobility in SDS gel (11), our result suggested that the fused N-terminal peptide should eliminate this phenomenon. Moreover, the recombinant cobrotoxin had an immunoreactivity toward anti-cobrotoxin antibodies as evidenced by immunoblot analysis. As shown in Fig. 3, recombinant cobrotoxin as well as native cobrotoxin effectively inhibited acetylcholine-induced muscle contractions. It was evident that recombinant toxin had a neurotoxic activity as did native cobrotoxin. However, the degree of inhibition caused by the addition of recombinant cobrotoxin was approximately 50% of that observed with the addition of cobrotoxin. Since the toxin itself is devoid of Met, the Met residue preceding the Leu-1 of cobrotoxin serves as a CNBr-cleavage site. The cleavage products were further purified by HPLC on a SynChropak RP-P column (data not shown). However, the isolated toxin fraction was not homogeneous as revealed by N-terminal sequence determination. The toxin fraction had a short N-terminal amino acid extension relative to venom-derived cobrotoxin. It was owing to that CNBr-cleavage did not exclusively occur at the Met preceding the first amino acid of the recombinant toxin, but occurred at other Met residues of the fused peptide. Thus, the recombinant protein was sub-
654
Vol. 263, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Electrophoresis analyses of the isomerization of recombinant cobrotoxin and mutants. The isomerization reaction was carried out by dissolving proteins in 0.1 M Tris (pH 8.7) for 3 days. Samples were desalted through a PD-10 column (Pharmacia Biotech Ltd.) equilibrated with 0.1 M acetic acid. After lyophilization, the samples were analyzed by native gel electrophoresis. Lane 1, recombinant cobrotoxin; lane 2, recombinant cobrotoxin at day 3; lane 3, E38Q; lane 4, E38Q at day 3; lane 5, K47Q; lane 6, K47Q at day 3; lane 7, K47E; lane 8, K47E at day 3. The arrows indicate the isomers of recombinant cobrotoxin and mutants. (Inset) Disulfide isomerization of native cobrotoxin. Lane 1, cobrotoxin; lane 2, cobrotoxin at day 3.
jected to study its isomerization reaction without removal of the N-terminal fused peptide. Effect of amino acid substitutions on the isomerization reaction of cobrotoxin. Dideoxynucleotide sequencing analysis on the mutated clones indicated that all the mutants were successfully prepared by MutaGene kit (data not shown). However, only 3 (E38Q, K47E, and K47Q) of 10 mutants were successfully expressed, and these mutated proteins were isolated from the inclusion bodies of E. coli. Alternatively, no detectable amount of mutated proteins, E21K, E21Q, K27E, K27Q, E38K, E51K, and E51Q, appeared in the supernatant or inclusion bodies of E. coli. Refolding reaction of mutants E38Q, K47E and K47Q was carried out under the same conditions for recombinant cobrotoxin. The resulting products were further purified by HPLC on a SynChropak RP-P column. Although recombinant protein and its mutants were homogeneous in molecular weight as revealed by SDS–PAGE analysis (data not shown), the results of native gel electrophoresis showed that the preparations of recombinant cobrotoxin and mutants comprised a number of species with different molecular shape and/or charged properties (Fig. 4). Since the occurrence of random pairings for 8 Cys residues is expected to result in 105 kinds of disulfide isomers, our finding could be interpreted as that a number of disulfide isomers appeared in these preparations. Nevertheless, a major folded product was obtained from the preparations of recombinant cobrotoxin and mutant K47Q as evidenced by the result of native gel analysis (Fig. 4). As shown in Fig. 4, the isomerization reaction of recombinant cobrotoxin occurred spontaneously while dissolving in 0.1 M
Tris buffer (pH 8.7). Isomerization reaction resulted in the production of isomer with a slower electrophoretic mobility. This is essentially the same as that observed for cobrotoxin (inset of Fig. 4). Isomerization reaction was also noted with mutants E38Q, K47E, and K47Q. An appreciable decrease in the rate of isomerization reaction was observed for these mutants. This result was in line with previous results showing that modification on Lys residues at positions 26, 27, and 47 of cobrotoxin decelerated its isomerization reaction (5). Although NMR structure of cobrotoxin shows that Lys-27 and Glu-38 are close in spatial position (12), the action of positively charged Lys-27 was not dominated by replacing Glu-38 with Gln-38 as evidenced by the finding that a decrease rather than an increase in the rate of disulfide isomerization was observed for mutant E38Q. It is likely to reflect that the negatively charged Glu-38 is not one of carboxyl groups by which the rate of isomerization reaction was decelerated (5). In the present study, cobrotoxin was successfully expressed in E. coli. Refolding reaction makes the resulting recombinant cobrotoxin exhibit a neurotoxicity as that observed with native cobrotoxin. Although the mutated clones are easily prepared by oligonucleotidedirected mutagenesis, it is worthy to note that one base substitution in the nucleotide sequence could result in completely eliminating the expression of recombinant protein. Moreover, our result shows the possibility that the preparation of recombinant protein contains a number of isomers with different disulfide pairings. Therefore, the properties of recombinant proteins need carefully evaluate in terms of their disulfide linkages. Alternatively, the finding that the content of disulfide isomers in the preparation of recombinant cobrotoxin and its mutants is not the same, reflects that one amino acid substitution may alter the refolded pathway and the resulting refolded species. ACKNOWLEDGMENTS This work was supported by Grants NSC 89-2316-B110-001 (to L. S. Chang) and NSC 88-2311-B007-002 (to C. C. Yang) from the National Science Council, ROC.
REFERENCES
655
1. Yang, C. C., Yang, H. J., and Huang, J. S. (1969) Biochim. Biophys. Acta 188, 65–77. 2. Yang, C. C., Yang, H. J., and Chiu, R. H. C. (1970) Biochim. Biophys. Acta 214, 355–363. 3. Yang, C. C. (1967) Biochim. Biophys. Acta 133, 346 –355. 4. Chang, L. S., Lin, S. R., and Chang, C. C. (1998) Arch. Biochem. Biophys. 354, 1– 8. 5. Chang, L. S., Lin, S. R., Chang, C. C., and Yang, C. C. (1998) Arch. Biochem. Biophys. 358, 164 –170. 6. Chang, L. S., Lin, J., Chou, Y. C., and Hong, E. (1997) Biochem. Biophys. Res. Commun. 239, 756 –762. 7. Thannhauser, T. W., and Scheraga, H. A. (1985) Biochemistry 24, 7681–7688.
Vol. 263, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
8. Chang, L. S., Wu, P. F., and Chang, C. C. (1996) Biochem. Biophys. Res. Commun. 225, 990 –996. 9. Chang, L. S., Lin, J., Wu, P. F., Chng, C. C., and Hong, E. (1997) Biochem. Biophys. Res. Commun. 230, 192–195. 10. Chang, L. S., Chou, Y. C., Lin, S. R., Wu, B. N., Lin, J., Hong, E., Sun, Y. J., and Hsiao, C. D. (1997) J. Biochem. 122, 1252–1259.
11. Chang, L. S., Chou, L., and Chang, C. C. (1996) Biochem. Mol. Biol. Int. 40, 1227–1232. 12. Yu, C., Bhaskaran, R., Chung L. C., and Yang, C. C. (1992) in Recent Advances in Molecular and Biochemical Research on Proteins (Wei, Y. H., Chen, C. S., and Su, J. C., Eds.), pp. 209 –218, World Scientific, Singapore.
656
Expression and Mutagenesis Studies of Cobrotoxin from Taiwan Cobra Long-sen Chang,* ,1 Ku-chung Chen,† Bin-nan Wu,† Shu-kai Lin,* Pei-fung Wu,† Yi-ren Hong,† and Chen-chung Yang‡ ,1 *Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung 804, Taiwan; †Department of Biochemistry and Pharmacology, Kaohsiung Medical College, Kaohsiung 807, Taiwan; and ‡Department of Life Sciences, National Tsing Hua University, Hsinchu 300, Taiwan, Republic of China
Received July 27, 1999
The cDNA encoding cobrotoxin was constructed from the cellular RNA isolated from the venom glands of Naja naja atra (Taiwan cobra). The cDNA was subcloned into the expression vector pET20b(1) and transformed into BL21(DE3) Escherichia coli strain. Expressed cobrotoxin was isolated from inclusion bodies of E. coli and subjected to refolding into its folded structure. The refolded cobrotoxin was purified by high-performance liquid chromatography and exhibited a neurotoxicity in inhibiting acetylcholineinduced muscle contractions. Recombinant cobrotoxin showed a tendency to isomerize its disulfide bonds as that observed with native cobrotoxin. An appreciable decrease in the rate of isomerization reaction was observed when Glu-38 was replaced with Gln-38 or Lys-47 was replaced with Glu-47 or Gln-47. These results reflect that the element in controlling the disulfide isomerization of cobrotoxin is closely associated with the charged side chains in the cobrotoxin molecule. © 1999 Academic Press
Cobrotoxin is a neurotoxic protein isolated from the venom of Taiwan cobra (Naja naja atra). It is a small, basic protein consisting of a single polypeptide chain of 62 amino acids, cross-linked by four disulfide bonds (1, 2). The four disulfide linkages are Cys3-Cys24, Cys17Cys41, Cys43-Cys54, and Cys55-Cys60, respectively (3). Recent studies show that the disulfide bonds at the C-terminal region of cobrotoxin exhibit a tendency to exchange with each other (4). Two cobrotoxin isomers (cobrotoxin II and cobrotoxin III) have been isolated from the refolding mixtures of cobrotoxin as well as from Taiwan cobra venom. The disulfide linkages at the C-terminus of cobrotoxin II and cobrotoxin III are identified as Cys43-Cys55 & Cys54-Cys60 and Cys431
To whom correspondence may be addressed. Fax: 886-7-5253696. E-mail: [email protected]. 0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Cys-60 & Cys54-Cys55, respectively. Moreover, the isomerization reaction is irreversible. Chemical modification studies indicate that disulfide isomerization of cobrotoxin is, in part driven by the positively charged Lys residues at positions 26, 27, and 47 of the toxin molecule, and decelerated by the negatively charged groups of cobrotoxin (5). These reflect that Lys residues and carboxyl groups constitute the charged environment by which the isomerization reaction is controlled. To clearly elucidate the mechanism of basic and acidic residues in controlling the disulfide isomerization reaction, the best way is to clone the cobrotoxin cDNA and subject it to site-directed mutagenesis. Thus expression and mutagenesis studies of cobrotoxin are carried out in the present study. MATERIALS AND METHODS Preparation of mRNA from venom glands. Cellular RNA was isolated from the snake (Naja naja atra) venom glands which were stored in liquid nitrogen immediately after sacrificed. Two deep frozen glands from one snake were homogenized to extract RNA by a guanidinium isothiocyanate/phenol/chloroform isolation kit (Stratagene Ltd., USA). PCR amplification and cloning. Two oligonucleotide primers of sense and antisense orientations based on the signal peptide and 39noncoding region of cobrotoxin genomic DNA (6) with the forward sequence, 59-ATGAAAACTCTGCTGCTGACCTTGCTG-39 and the reverse one 59-GGATGGTCCTTGATGGATGAGAGC-39 were synthesized. RT-PCR was carried out with 100 ml reaction buffer containing 100 mM Tris–HCl (pH 8.3), 1 mM dNTP, 1 mM antisense primer and 200 ng RNA template. In the reverse transcription, the cDNA was started with rTth reverse transcriptase (5 U) and 2 ml of 10 mM MnCl 2 at 70°C for 15 min, and stopped the reaction by placing the tube on ice until needed. A 8 ml chelating buffer containing 50% glycerol (v/v), 100 mM Tris–HCl (pH 8.3), 1 M KCl and 7.5 mM EGTA/0.5% Tween 20 was added to the reaction. After addition of 8 ml of 25 mM MgCl 2 and 1 mM sense primer, the amplification was proceeded on a thermocycler 94°C/45°C/72°C 1 min each, for a total 30 cycles. The PCR products were cloned into pCRII vector according to the TA-cloning procedures (Invitrogen, San Diego, CA).
652
Vol. 263, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Cloning and expression of cobrotoxin. Synthetic oligonucleotides were designed to produce a 249-bp amplified DNA fragment spanning the open reading frame of cobrotoxin. Primer 1 introduced a 59 EcoRV site and an in frame initiating Met codon preceding Leu-1 of cobrotoxin,
EcoRV 59-GATATCATGCTGGAATGTCACAACCAACAATCATC-39. AspIleMetLeuGluCysHisAsnGlnGlnSer Primer 2 was the reverse primer for RT-PCR amplification of cobrotoxin. The PCR products were cloned into pCR II vector. The inserted DNA fragment was cut with EcoRV and EcoRI and ligated into the large fragment of EcoRV/EcoRI-cut pET20b(1). The entire sequence was confirmed by dideoxynucleotide sequencing. The resulting plasmid pET-CbTx was transformed into E. coli strain BL21(DE3). Transformants were selected on LB-agar plates supplements with 50 mg/ml ampicillin. For induction of gene expression, E. coli BL21(DE3) cells containing pET-CbTx were grown at 37°C in LB medium containing 50 mg/ml ampicillin. After reaching an OD 550 5 1.0, isopropyl-b-D-thiogalactoside (IPTG) was added to a final concentration of 0.2 mM. The culture was induced for a period of up to 4 h. The cells were harvested and lysed by ultrasonication. The recombinant cobrotoxin was found to appear exclusively in the inclusion bodies of E. coli. Inclusion bodies recovered from 1 liter of bacterial culture were thoroughly washed with 10 mM Tris–1 mM EDTA (pH 8.0) containing 0.5% Triton X-100, and sulfonated with disodium 2-nitro-5-(sulfothio)-benzoate in 8 M urea, 0.3 M Na 2SO 3, pH 8.5 (7, 8). The protein was precipitated by 1% acetic acid and dissolved in 10 ml of 50 mM sodium borate (pH 8.5) containing 5 mM EDTA, 8 M urea, and 4 mM reduced and 2 mM oxidized glutathione. Refolding was performed by a four-fold dilution with the same buffer without urea. After standing for 1 day at room temperature, the refolded cobrotoxin was further purified by HPLC. Oligonucleotide-directed mutagenesis. The inserted DNA in pETCbTx was cut with XbaI and EcoRI and ligated into the large fragment of XbaI/EcoRI-cut M13mp19. The entire sequence was confirmed by dideoxynucleotide sequencing. Oligonucleotide-directed mutagenesis was carried out by using a Muta-Gene kit (Bio-Rad) with synthetic oligonucleotides. The oligonucleotide primers E21, K27, E38, K47, and E51 with sequence 59-CAATTGGTCTKCCCCCCTGAACAAC-39, 59CACGCCAACGCTSTTTATAGCAATTG-39, 59-CCACATCCCCTCTKGGTTCTATATCC-39, 59-CAATGCCGTTCTSCACTGAAGGGCAAC-39 and 59-AACAGTTAATTTKAATGCCGTTCTTC-39 were synthesized for preparing the mutants E21K, E21Q, K27E, K27Q, E38K, E38Q, K47E, K47Q, E51K, and E51Q, respectively. The mutations were confirmed by sequencing the gene on M13mp19 using Sequanase kit. The XbaI/EcoRI fragment of the mutated clone was cloned into a pET20b(1) expression vector. Other tests. DNA sequencing, native gel electrophoresis, SDS– polyacrylamide gel electrophoresis (PAGE), immunoblot analysis, analysis of neurotoxicity and disulfide isomerization reaction were performed in essentially the same manner as previously described (4, 5, 8 –10).
RESULTS AND DISCUSSION Expression of recombinant cobrotoxin. The genomic DNA encoding cobrotoxin precursor had been determined in our laboratory previously (6). To clone the cDNA encoding cobrotoxin, two primers were designed from the regions at the beginning of the signal peptide and 39-noncoding region of cobrotoxin genomic DNA, respectively. PCR amplification of the venom gland cDNA mixtures with the designed primers achieved the isolation of a PCR fragment estimated to be about
FIG. 1. Nucleotide and deduced protein sequences of the precursors of cobrotoxin and cobrotoxin homolog. The nucleotide sequence of 306 base pairs is shown above the amino acid sequence of 83 residues including a signal peptide of 21-amino acid residues. The mature cobrotoxin of 62 residues starts at Leu. The cDNA sequence is essentially the same as that deposited in EMBL, GeneBank and DDBJ nucleotide sequence databases under the accession number U42582. The base-substitutions, base-deletions and amino acid substitutions observed between cobrotoxin homolog and cobrotoxin are shown under the sequence of cobrotoxin. The Asp-31 is substituted by Tyr-31 in the cobrotoxin homolog. Moreover, instead of two amino acid residues, Asp-Arg, at positions 58 and 59 of cobrotoxin, there is only a Gly residue in the cobrotoxin homolog. Thus, the deduced protein sequence of cobrotoxin homolog comprises 61 amino acid residues. The cDNA sequence of cobrotoxin homolog with 303 base pairs has been deposited in EMBL, GeneBank and DDBJ nucleotide sequence databases under Accession No. AJ239050.
300 bp (data not shown). The DNA fragments were then subcloned by a TA-cloning kit. More than 20 clones were selected for nucleotide sequencing. Three of the selected clones had the same cDNA sequence, which deduced protein sequence corresponded to that of cobrotoxin (Fig. 1). The deduced protein sequence is the same as the previous published sequence determined by protein sequencing technique (1–3). However, one novel cDNA clone encoding a cobrotoxin homolog was also identified. Its deduced protein sequence comprises 61 amino acid residues. Instead of two amino acid residues, Asp-Arg, at positions 58 and 59 of cobrotoxin, there is only a Gly residue in the cobrotoxin homolog. Moreover, a Tyr-31 appears in the cobrotoxin homolog resulted from one base substitution. To subclone the cDNA into the expression vector, a new primer was designed to create NdeI site in the beginning of the nucleotide sequences for encoding
653
Vol. 263, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
amino acid sequence. The primer sequences 59CATATGCTGGAATGTCACAACCAACAATCATC-39 (the underline indicated the NdeI site) was synthesized for subcloning cobrotoxin cDNA into expression vector pT7-7. Noticeably, the resulting expressed protein should contain only a Met residue preceding the first amino acid of cobrotoxin. The PCR procedure was performed with pfu DNA polymerase. The amplified DNA was inserted into pCRII vector, then subcloned into the expression vector pT7-7 by digestion with NdeI/ HindIII. Unfortunately, no appreciable amount of recombinant protein was produced by inducing BL21(DE-3) containing plasmid pT7-7-CbTx with IPTG. Thus, another primer 59-GATATCATGCTGGAATGTCACAACCAACAATCATC-39 (the underline indicated the EcoRV site) was synthesized for subcloning cobrotoxin cDNA into pET20b(1) expression vector. Since the cobrotoxin cDNA was subcloned into the large fragment of EcoRV/EcoRI-cut pET20b(1). The resulting protein should have a fused N-terminal peptide with the amino acid sequence MKYLLPTAAAGLLLLAAQPAMAMDIM. The expression of recombinant cobrotoxin was observed with BL21(DE-3) containing pET-CbTx plasmid. However, the expressed protein exclusively appeared in the inclusion bodies of E. coli. Thus refolding of the expressed protein was carried out according to the procedure described previously (7, 8). As shown in Fig. 2, the refolded protein
FIG. 2. Purification of recombinant cobrotoxin. The refolded proteins were applied on a SynChropak RP-P column equilibrated with 0.1% TFA and eluted with a linear gradient of 15–50% acetonitrile for 70 min. The flow rate was 0.8 ml/min and the eluate was monitored at 235 nm. The arrow indicates the refolded cobrotoxin. (Inset) SDS–PAGE (lanes 1–3) and immunoblot (lanes 4 – 6) analysis of recombinant cobrotoxin. Lanes 1 and 4, molecular markers (prestained SDS–PAGE standards, BIO-RAD, Hercules, CA): phosphorylase (107,000), BSA (76,000), ovalbumin (52,000), carbonic anhydrase (36,800), soybean trypsin inhibitor (27,200) and lysozyme (19,000); lanes 2 and 5, native cobrotoxin; lanes 3 and 6, recombinant cobrotoxin.
FIG. 3. Typical recordings of the inhibitory effects of native cobrotoxin and recombinant cobrotoxin on acetylcholine-induced maximal contractions in isolated frog rectus abdominis muscle. The added concentrations of acetylcholine and toxins (A, native cobrotoxin; B, recombinant cobrotoxin) are 10 and 0.5 mM, respectively.
was purified by HPLC on a SynChropak RP-P column. The recombinant protein had an apparent molecular weight lower than native cobrotoxin as revealed by SDS–PAGE (inset of Fig. 2). Since previous study revealed that cobrotoxin exhibited an anomalous mobility in SDS gel (11), our result suggested that the fused N-terminal peptide should eliminate this phenomenon. Moreover, the recombinant cobrotoxin had an immunoreactivity toward anti-cobrotoxin antibodies as evidenced by immunoblot analysis. As shown in Fig. 3, recombinant cobrotoxin as well as native cobrotoxin effectively inhibited acetylcholine-induced muscle contractions. It was evident that recombinant toxin had a neurotoxic activity as did native cobrotoxin. However, the degree of inhibition caused by the addition of recombinant cobrotoxin was approximately 50% of that observed with the addition of cobrotoxin. Since the toxin itself is devoid of Met, the Met residue preceding the Leu-1 of cobrotoxin serves as a CNBr-cleavage site. The cleavage products were further purified by HPLC on a SynChropak RP-P column (data not shown). However, the isolated toxin fraction was not homogeneous as revealed by N-terminal sequence determination. The toxin fraction had a short N-terminal amino acid extension relative to venom-derived cobrotoxin. It was owing to that CNBr-cleavage did not exclusively occur at the Met preceding the first amino acid of the recombinant toxin, but occurred at other Met residues of the fused peptide. Thus, the recombinant protein was sub-
654
Vol. 263, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Electrophoresis analyses of the isomerization of recombinant cobrotoxin and mutants. The isomerization reaction was carried out by dissolving proteins in 0.1 M Tris (pH 8.7) for 3 days. Samples were desalted through a PD-10 column (Pharmacia Biotech Ltd.) equilibrated with 0.1 M acetic acid. After lyophilization, the samples were analyzed by native gel electrophoresis. Lane 1, recombinant cobrotoxin; lane 2, recombinant cobrotoxin at day 3; lane 3, E38Q; lane 4, E38Q at day 3; lane 5, K47Q; lane 6, K47Q at day 3; lane 7, K47E; lane 8, K47E at day 3. The arrows indicate the isomers of recombinant cobrotoxin and mutants. (Inset) Disulfide isomerization of native cobrotoxin. Lane 1, cobrotoxin; lane 2, cobrotoxin at day 3.
jected to study its isomerization reaction without removal of the N-terminal fused peptide. Effect of amino acid substitutions on the isomerization reaction of cobrotoxin. Dideoxynucleotide sequencing analysis on the mutated clones indicated that all the mutants were successfully prepared by MutaGene kit (data not shown). However, only 3 (E38Q, K47E, and K47Q) of 10 mutants were successfully expressed, and these mutated proteins were isolated from the inclusion bodies of E. coli. Alternatively, no detectable amount of mutated proteins, E21K, E21Q, K27E, K27Q, E38K, E51K, and E51Q, appeared in the supernatant or inclusion bodies of E. coli. Refolding reaction of mutants E38Q, K47E and K47Q was carried out under the same conditions for recombinant cobrotoxin. The resulting products were further purified by HPLC on a SynChropak RP-P column. Although recombinant protein and its mutants were homogeneous in molecular weight as revealed by SDS–PAGE analysis (data not shown), the results of native gel electrophoresis showed that the preparations of recombinant cobrotoxin and mutants comprised a number of species with different molecular shape and/or charged properties (Fig. 4). Since the occurrence of random pairings for 8 Cys residues is expected to result in 105 kinds of disulfide isomers, our finding could be interpreted as that a number of disulfide isomers appeared in these preparations. Nevertheless, a major folded product was obtained from the preparations of recombinant cobrotoxin and mutant K47Q as evidenced by the result of native gel analysis (Fig. 4). As shown in Fig. 4, the isomerization reaction of recombinant cobrotoxin occurred spontaneously while dissolving in 0.1 M
Tris buffer (pH 8.7). Isomerization reaction resulted in the production of isomer with a slower electrophoretic mobility. This is essentially the same as that observed for cobrotoxin (inset of Fig. 4). Isomerization reaction was also noted with mutants E38Q, K47E, and K47Q. An appreciable decrease in the rate of isomerization reaction was observed for these mutants. This result was in line with previous results showing that modification on Lys residues at positions 26, 27, and 47 of cobrotoxin decelerated its isomerization reaction (5). Although NMR structure of cobrotoxin shows that Lys-27 and Glu-38 are close in spatial position (12), the action of positively charged Lys-27 was not dominated by replacing Glu-38 with Gln-38 as evidenced by the finding that a decrease rather than an increase in the rate of disulfide isomerization was observed for mutant E38Q. It is likely to reflect that the negatively charged Glu-38 is not one of carboxyl groups by which the rate of isomerization reaction was decelerated (5). In the present study, cobrotoxin was successfully expressed in E. coli. Refolding reaction makes the resulting recombinant cobrotoxin exhibit a neurotoxicity as that observed with native cobrotoxin. Although the mutated clones are easily prepared by oligonucleotidedirected mutagenesis, it is worthy to note that one base substitution in the nucleotide sequence could result in completely eliminating the expression of recombinant protein. Moreover, our result shows the possibility that the preparation of recombinant protein contains a number of isomers with different disulfide pairings. Therefore, the properties of recombinant proteins need carefully evaluate in terms of their disulfide linkages. Alternatively, the finding that the content of disulfide isomers in the preparation of recombinant cobrotoxin and its mutants is not the same, reflects that one amino acid substitution may alter the refolded pathway and the resulting refolded species. ACKNOWLEDGMENTS This work was supported by Grants NSC 89-2316-B110-001 (to L. S. Chang) and NSC 88-2311-B007-002 (to C. C. Yang) from the National Science Council, ROC.
REFERENCES
655
1. Yang, C. C., Yang, H. J., and Huang, J. S. (1969) Biochim. Biophys. Acta 188, 65–77. 2. Yang, C. C., Yang, H. J., and Chiu, R. H. C. (1970) Biochim. Biophys. Acta 214, 355–363. 3. Yang, C. C. (1967) Biochim. Biophys. Acta 133, 346 –355. 4. Chang, L. S., Lin, S. R., and Chang, C. C. (1998) Arch. Biochem. Biophys. 354, 1– 8. 5. Chang, L. S., Lin, S. R., Chang, C. C., and Yang, C. C. (1998) Arch. Biochem. Biophys. 358, 164 –170. 6. Chang, L. S., Lin, J., Chou, Y. C., and Hong, E. (1997) Biochem. Biophys. Res. Commun. 239, 756 –762. 7. Thannhauser, T. W., and Scheraga, H. A. (1985) Biochemistry 24, 7681–7688.
Vol. 263, No. 3, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
8. Chang, L. S., Wu, P. F., and Chang, C. C. (1996) Biochem. Biophys. Res. Commun. 225, 990 –996. 9. Chang, L. S., Lin, J., Wu, P. F., Chng, C. C., and Hong, E. (1997) Biochem. Biophys. Res. Commun. 230, 192–195. 10. Chang, L. S., Chou, Y. C., Lin, S. R., Wu, B. N., Lin, J., Hong, E., Sun, Y. J., and Hsiao, C. D. (1997) J. Biochem. 122, 1252–1259.
11. Chang, L. S., Chou, L., and Chang, C. C. (1996) Biochem. Mol. Biol. Int. 40, 1227–1232. 12. Yu, C., Bhaskaran, R., Chung L. C., and Yang, C. C. (1992) in Recent Advances in Molecular and Biochemical Research on Proteins (Wei, Y. H., Chen, C. S., and Su, J. C., Eds.), pp. 209 –218, World Scientific, Singapore.
656