- Email: [email protected]
The isolation and characterization of a peptide that alters sodium channels from Buthus martensii Karsch
Toxicon 38 (2000) 645±660 www.elsevier.com/locate/toxicon
The isolation and characterization of a peptide that alters sodium channels from Buthus martensii Karsch Richard Hahin a,*, Ziyi Chen a, Giridher Reddy b b
a Biological Sciences, Northern Illinois University, DeKalb, IL 60115, USA Department of Pediatrics, Biological Sciences Division, University of Chicago, Chicago, Illinois, USA
Received 26 November 1998; accepted 14 July 1999
Abstract The peptides were puri®ed using gel ®ltration, ion exchange, FPLC, and HPLC chromatography and found to greatly prolong action potentials at nanomolar concentrations when applied to frog and mouse nerves. The N-terminal primary amino acid sequence of one of the peptides, BMK 16(5), was determined. The ®rst 23 amino acids of BMK 16(5) were found to be: VKDGYIADDRNCPYFCGRNAYYD. The two cysteine residues in the sequence appeared as Edman sequence cycle blanks; however, they were assigned to be cysteines due to sequence similarity to other peptide toxins that bind to sodium channels and identi®cation of the presence of cysteines obtained from single time point amino acid analysis. The MW of BMK 16(5) was determined by a Perkin±Elmer API 300 LC/MS/MS to be 3695. The amino acid residues of BMK 16(5) show strong similarity with the ®rst 23 amino acid residues of a number of scorpion alpha neurotoxins. Unlike these neurotoxins, BMK 16(5) possesses a proline residue at position 13 which will likely make it fold in a unique way so as to bind to and alter sodium channels. # 1999 Elsevier Science Ltd. All rights reserved.
1. Introduction A number of peptides that speci®cally bind to and alter sodium channels have * Corresponding author. Fax: +1-815-753-0461. 0041-0101/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 1 - 0 1 0 1 ( 9 9 ) 0 0 1 8 0 - 4
646
R. Hahin et al. / Toxicon 38 (2000) 645±660
been previously described in Martin-Eauclaire and Courard (1995). The ®rst peptide activator known to bind to sodium channels and alter the sodium channel conductance was described by Miranda et al. (1970). Peptides that have been isolated from scorpion venoms that bind to and alter vertebrate sodium channel inactivation have been designated as alpha scorpion toxins. Alpha scorpion toxins bind speci®cally to particular sites on the sodium channel. Distinct receptor binding sites have been predominately characterized using vertebrate excitable preparations (Catterall, 1980, 1986). The alpha scorpion toxins that have been shown to act to lengthen sodium channel inactivation have been shown to bind to a region on the sodium channel designated as receptor site 3 in the rat brain (Catterall and Beress, 1978; Courard et al., 1978; Rogers et al., 1996). Alpha scorpion toxins typically possess a molecular weight in the range of 6000±7500 and contain eight cysteine residues that participate in forming four disul®de bonds. In this report we describe the isolation and puri®cation of a novel peptide from the venom of the scorpion Buthus martensii Karsch. This peptide, which is quite small in size (M.W. 3695) acts to alter sodium channels so that action potentials are greatly prolonged. The peptide also shows sequence similarities to parts of several alpha scorpion toxin peptides that have been shown to bind to and alter sodium channel inactivation. 2. Materials and methods 2.1. Source of venom Lyophilized crude venom was obtained from the People's Republic of China. 2.2. Solubilization of crude venom Crude venom (100 mg) was dissolved in 1 ml of 20 mM ammonium acetate (pH 4.7) buer at 48C for 40 min. The venom solution was centrifuged at 20,000 G for 20 min to remove the insoluble material. The supernatant containing the solubilized venom was then puri®ed using gel ®ltration and ion exchange column chromatography. 2.3. Gel ®ltration, ion exchange, and reversed phase HPLC chromatography Solubilized crude venom was applied to a 1.6 100 cm column containing Sephadex G50 equilibrated with a 20 mM ammonium acetate buer (pH 4.7) at T = 208C. The buer ¯ow rate was held at 6 ml per hour. One ml of solubilized venom (100 mg/ml) was applied to the column and the eluted material was collected in 2 ml fractions. The optical absorbance of the eluent was measured at 280 nm. Fractions were separated and assayed for their ability to lengthen action potentials (APs) recorded using the sucrose gap method. Fractions displaying
R. Hahin et al. / Toxicon 38 (2000) 645±660
647
biological activity were pooled and concentrated using an ultra®ltration membrane (Amicon YM 1) with a 1000 Da retention cut o to produce solutions containing biologically active material. Concentrated biologically active fractions were loaded onto a 2.5 50 cm column containing CM Sephadex C-50 which was equilibrated with 20 mM ammonium acetate buer (pH 4.7) at 208C. Fractions were eluted and collected with a sequence of 3 linear sodium chloride gradients using a ¯ow rate of 50 ml/h. Twenty absorbance peaks were observed and 8 contained material which acted to lengthen APs in frog and mouse nerve. Fractions from peak 16 were pooled and concentrated and applied to a 5 50 mm Mono-S HR5/5 (Pharmacia Biotech Inc., Piscataway, NJ) strong cation exchange resin column equilibrated with 20 mM phosphate buer. A sequence of linear sodium chloride gradients in 20 mM phosphate buer (pH 7.0) were used to elute peptides from the column. The ¯ow rate was 1 ml/min. Absorbance was detected at 280 nm. Fractions were collected, assayed for activity, and further puri®ed using a C18 reversed phased HPLC column (4.6 250 mm). The ¯ow rate was 1 ml/min and peptides were detected at 216 nm. 2.4. Polyacrylamide gel electrophoresis Polyacrylamide gel electrophoresis was performed using the following protocol: acrylamide (30%) and Bis acrylamide (0.8%) and K-acetate were polymerized by ammonium per sulfate and N,N,N ',N '-tetramethylethylenediame (TEMED). A 5% stacking gel was applied above a 15% acrylamide resolving gel. The electrode buer was K-acetate (pH 4.3). Slab gels were stained in 0.25% Coomasie blue for 40 min and destained in 10% acetic acid/30% methanol. 2.5. Molecular weight determination of puri®ed peptides Initial estimates of molecular weight were obtained with a Superdex Peptide HR 10/30 (Pharmacia Biotech Inc., Piscataway, NJ) gel ®ltration column displaying optimal resolution within the molecular weight range of 100±7000. Peptides and standard proteins were applied to the column using a 50 mM pH 7.0 phosphate buer and eluted at a ¯ow rate of 0.5 ml/min. Catalase (MW 220,000) which is totally excluded from the gel was used to determine the void volume of the column. To estimate the molecular weights, a standard curve plotting the partition coecient vs the log molecular weight of a number of molecular weights standards was used. A PE (Perkin±Elmer) SCIEX API 300 LC/MS/MS mass spectrometer at the University of Chicago was used to more accurately and precisely determine the MW of one of the peptides (BMK 16(5)). A major spectral peak occurred at a MW of 3695. Associated with the major peak at various charge states were peaks at 1848 at +2, 1232 at +3, and 1923 at +4. A minor spectral peak occurred at 2112, which is likely a degraded product of the original peptide. Associated with the 2112 peak were peaks at 1056 at +2 and 699 at +3. Another peak occurred at 7392, which is likely a dimer of the 3695 molecule.
648
R. Hahin et al. / Toxicon 38 (2000) 645±660
2.6. Cysteine determination A single time point amino acid analysis was performed. Acid hydrolysis was performed on the sample using 6 N double distilled constant boiling HCl for 20 h at 1258C in a nitrogen purged sealed container. Presence of cysteines was con®rmed. 2.7. Amino acid sequence determination The sequence of the peptide was carried out using a pulsed liquid protein ABI 494 Procise Edman N-terminus sequencer at the University of Chicago. Blanks in this sequence, which could be interpreted as either cysteines or tryptophans, were tentatively assigned to be cysteines due to similarity with other peptides and the presence of cysteine found by single time point amino acid analysis experiments performed using a Beckman amino acid analyzer. Cysteines so identi®ed are indicated in the sequence as C. Table 1 shows the pmole yields obtained for each Edman degradation cycle. 2.8. Biological activity assay Fractions to be tested were dissolved in either Ringer's solution (for frog Table 1 Yield obtained for each Edmandegredation cycle Residue number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 17 18 19 20 21 22 23
Amino acid ID
Yield (pmol)
Val Lys Asp Gly Tyr Ile Ala Asp Asp Arg Asn
30.92 25.45 23.80 26.84 27.21 23.15 24.17 15.67 21.93 10.15 14.21
Pro Tyr
11.54 12.66
Gly Arg Asn Ala Tyr Tyr Asp
10.44 6.79 7.50 8.77 7.16 7.84 3.80
R. Hahin et al. / Toxicon 38 (2000) 645±660
649
nerves) or Tyrode solution (for mouse nerves). Compound actions potentials (APs) were elicited using a technique adapted from Hahin and Strichartz (1981) and brie¯y described below. A sciatic nerve was placed in the chamber so that the nerve spanned 2 compartments; the proximal end (placed in the intracellular compartment) was bathed in isosmotic KCl, and the middle (placed in the
Fig. 1. Absorbance pro®le of G-50 gel ®ltration puri®cation step 1. The ordinate represents absorbance units. The abscissa represents the fraction number; each fraction collected represents 2 ml. Arrows signify absorbance peaks that contained peptides that acted to alter sodium (Na) channels. Absorbance peaks 2 and 3 were biologically active in modifying the APs in mouse and frog nerves; peaks 2 and 3 were comprised of fractions 29±37 and 38±48, respectively. Biological activity was de®ned to be the ability of a fraction to lengthen the action potential (AP) when applied to a frog or a mouse sciatic nerve. The two insets depict two experiments designed to show the biological action of pooled fractions obtained from peaks 2 and 3 on frog nerves. APs were recorded sequentially in time prior to and after the addition of an aliquot of pooled fraction 2 (PF2) and pooled fraction 3 (PF3) to the Ringer's solution bathing the nerves. The ®rst AP on the left was recorded prior to the addition of PF2/PF3, and each subsequent AP was recorded sequentially in time after the addition of PF2/PF3 at 5 min intervals in time; each AP was shifted rightward a constant amount to represent the passage of time (5 min interval).
650
R. Hahin et al. / Toxicon 38 (2000) 645±660
extracellular compartment) was bathed in either control solution (Ringer's or Tyrode's) or control solution plus the fraction to be tested, and was separated from KCl by a partition through which 0.22 molar sucrose ¯owed. Sucrose ¯ow was maintained constant (ca 3±8 ml/min) throughout the recording period. APs were elicited supermaximally stimulating the distal end of the nerve periodically with a suction electrode. A brief (100 ms) constant voltage pulse was used to elicit APs. APs were dierentially recorded using Ag/AgCl electrodes by obtaining the potential in the intracellular compartment relative to the potential in the extracellular compartment. APs were recorded on photographs from repetitively applied traces on a ZC6524 digital storage oscilloscope (Hitachi Denshi Ltd, Japan). 3. Results 3.1. Fractionation of scorpion venom Venom from the scorpion Buthus martensii Karsch was fractionated initially with a Sephadex G50 gel ®ltration column. Fig. 1 shows the results of adding 1 ml of solubilized venom (100 mg/ml) to the column. Four absorbance peaks were obtained and two of them (denoted by arrows) contained biologically active material that acted on the sodium channels in frog and mouse nerve to lengthen action potentials (APs). Fractions in absorbance peaks 2 and 3 were separately pooled to form pooled fractions designated as PF2 and PF3. The two insets in Fig. 1 depict two experiments designed to show the biological action of PF2 and PF3 on frog nerves. APs were recorded sequentially in time prior to and after the addition of an aliquot of PF2/PF3 to the Ringer solution bathing the nerves. The ®rst AP on the left was recorded prior to the addition of PF2 (left inset) or PF3
Fig. 2. CM Sephadex C-50 chromatogram. The ordinate represents absorbance units. The abscissa represents the fraction number; each fraction collected represents 8 ml. Arrows signify absorbance peaks that contained peptides that acted to alter Na channels. 20 absorbance peaks were observed; 9 peaks contained fractions with peptides which acted to alter APs. The biologically active peaks were 9± 14, and 16±18.
R. Hahin et al. / Toxicon 38 (2000) 645±660
651
Fig. 3. FPLC cation exchange chromatography of fraction 16. Fraction 16 from the CM-Sephadex C50 was puri®ed by FPLC rechromatography. The column used was a Mono-S HR 5/5 cation exchange column. Buer A was 20 mM phosphate while buer B was phosphate containing 1 M NaCl. The gradients used were: 0±6% B in 5 min, 6±8% B in 2 min, 8% B for 1.5 min, 8±20% B in 13 min, 20± 100% B in 1.5 min, 100% B in 1 min, and 100±0% B in 2 min. The ¯ow rate was 1 ml/min. Peptides were detected at 280 nm. The arrows indicate the fractions that were biologically active. The insets at the top of the ®gure represent the progressive change in the AP produced by fraction 16 (3) (left inset) applied at 177 nM, fraction 16 (4) (middle inset) applied at 117 nM, and fraction 16 (5) (right inset) applied at 480 nM. The ®rst AP recorded in each inset was obtained prior to the application of a fraction, while each other successively recorded AP exhibited progressive shape changes as time proceeded following exposure of the nerve to each fraction. Each AP was shifted rightward a constant amount to indicate the passage of time (5 min intervals).
652
R. Hahin et al. / Toxicon 38 (2000) 645±660
Fig. 4. Molecular weight estimates of BMK peptide fractions. Shown on the left hand side of Fig. 4 are three chromatograms obtained from a Superdex Peptide HR-10/30 (Pharmacia Inc., Piscataway, NJ) gel ®ltration column displaying optimal resolution within the molecular weight range of 100±7000. On the right side of the ®gure at the top is a chromatogram run on the Superdex Peptide column using standard proteins. Peptide fractions and standard proteins were applied to the column using a 50 mM pH 7.0 phosphate buer and eluted at a ¯ow rate of 0.5 ml/min. The chromatograms on the left represent: fraction 16 (5) (top), fraction 16 (4) (middle), and at the bottom fraction 16 (3). Also shown as an inset to the chromatogram of fraction 16 (5) is a polyacrylamide gel electrophoresis pro®le of the crude venom (left lane) and fraction 16 (5) in the right lane. The standard proteins used to obtain a standard curve to estimate the molecular weight were catalase (M.W. 220,000), RNAase (M.W.13,700), alpha cobra toxin (M.W. 7820), Bombesin (M.W. 1619), and glycine (M.W. 75). At the bottom right of Fig. 4 is a standard curve plotting the log M.W. vs the partition coecient (Kav).
R. Hahin et al. / Toxicon 38 (2000) 645±660
653
(right inset), and each subsequent AP was recorded sequentially in time after the addition of PF2/PF3 at 5 min intervals in time; each AP was shifted rightward a constant amount to represent the passage of time (5 min interval). Eluent fractions representing PF2 and PF3 were pooled and concentrated using an ultra®ltration membrane with a 1000 Da cuto to produce a solution that was applied to an ion exchange column. Fig. 2 shows the results of adding concentrated PF2 and PF3 from the Sephadex G50 column to a weak cation exchange column (CM Sephadex C-50). PF2 and PF3 were combined since the two peaks could not be isolated and separated conveniently. Fractions were eluted and collected during the application of a sequence of linear sodium chloride gradients to the column. Twenty absorbance peaks were observed and nine (designated by arrows) contained material which acted to lengthen APs in frog and mouse nerve. Each of the fractions displaying biological activity showed diering amounts of activity. Fraction 16 was the most potent and was further puri®ed. Pooled fractions from peak 16 of the CM Sephadex C-50 column were applied to a Mono S strong cation exchange column using FPLC. A sequence of linear sodium chloride gradients in 20 mM phosphate buer (pH 7.0) were used to elute peptides from the column. The results of the application of the gradient pro®le are shown in Fig. 3. The third, fourth and ®fth absorbance peaks (of 7) contained peptides that modi®ed (arrows designate the active peaks) APs. Peaks 3±5 derive from single peptides that were designated as BMK 16(3), BMK 16(4), and BMK 16(5). The insets (top) show the eects on nerve APs of aliquots of BMK 16(3), BMK 16(4), and BMK 16(5) to the Ringer solution bathing a frog nerve. Each peptide acted to greatly prolong the AP compared to its initial control value (®rst AP in each inset) following a continued exposure to the peptides. A Superdex peptide HR10/30 gel ®ltration column displaying optimal resolution within the molecular weight range of 100±7000 was used to estimate the M.W. of the biologically active peptides. The top right hand side of Fig. 4 shows a chromatogram displaying the absorbance pro®le of a number of MW standards applied to the Superdex peptide column. On the left hand side of Fig. 4 are displayed three other chromatograms run on the three peptides (BMK 16(5), BMK 16(4) and BMK 16(3)) that were puri®ed. Each of the peptides appears as a single peak on each chromatogram. To estimate the MWs, a standard curve plotting the partition coecient vs the log M.W. was used. Fig. 4 (bottom right) shows the standard curve that was used to estimate the MW. Using this method, the M.W.s of BMK 16(3), BMK 16(4) BMK 16(5) were determined to be 4120, 5150, and 2471, respectively. Peptide BMK 16(5) was characterized more fully than the other two peptides, and its M.W. was determined more accurately using a PE SCIEX API 300 LC/MS/MS mass spectrometer. Using this method, the MW of BMK 16(5) was determined to be 3695. Fig. 4 (top left) also shows (inset) the results of polyacrylamide gel electrophoresis of the crude venom (left lane) and BMK 16(5) which appeared as a single band, suggesting it had been puri®ed to homogeneity. In order to ensure that each peptide was puri®ed to homogeneity, samples of
654
R. Hahin et al. / Toxicon 38 (2000) 645±660
R. Hahin et al. / Toxicon 38 (2000) 645±660
655
Fig. 5 (continued)
each peptide were applied to a reversed phase C18 HPLC column. Fig. 5 shows the absorbance pro®les for each of the peptides. Fig. 5A shows the absorbance pro®le of BMK fraction 16(3) from FPLC applied to the column. A single peak representing peptide BMK 16(3) was observed. Similarly, in Fig. 5B and C, single peaks appear in the HPLC chromatograms indicating that both peptides had been puri®ed to homogeneity. Since the yield for peptides BMK 16(3) and BMK 16(4) was insucient to provide enough material for amino acid sequence determination, they were not studied further at this time.
Fig. 5. HPLC puri®cation of BMK 16(3), 16(4), 16(5). Fig. 5 shows chromatograms obtained from the application of BMK 16(3) (Fig. 5A), BMK 16(4) (Fig. 5B), and BMK 16(5) (Fig. 5C) to a reversed phase C18 (4.6 250 mm) HPLC column that had been equilibrated with 0.1% tri¯uoroacetic acid (TFA) in water. The column was eluted using a linear gradient from solution A (0.1% TFA in water) to 60% B (0.1% TFA in acetonitrile) in 60 min at a ¯ow rate of 1 ml/min. In each chromatogram the absorbance at 216 nM is plotted as a function of the retention time.
656
R. Hahin et al. / Toxicon 38 (2000) 645±660
Fig. 6. BMK peptides act on sodium channels. Shown above are three graphs illustrating the action of BMK 16(3) (top), BMK 16(4) (middle), and BMK 16(5) (bottom) on whole sciatic nerves bathed in TEA/Cd2+ Ringer's solution. TEA/Cd2+ Ringer's solution contained 100 mM CdCl2 and 10 mM TEACl in Ringer's solution. The ®rst AP in each graph was recorded in TEA/Cd2+ Ringer's solution prior to the application of the peptide. The arrow designates the application of the peptide with each concentration applied. Each successive AP recorded after the application of the peptide was recorded every 5 min. A rightward displacement of the AP re¯ects the passage of 5 min intervals of time.
3.2. Chemical characterization of puri®ed BMK 16(5) Since HPLC chromatograms of BMK 16(5) revealed a single peak and polyacrylamide gel electrophoresis of BMK 16(5) yielded a single band, BMK
R. Hahin et al. / Toxicon 38 (2000) 645±660
657
16(5) was subjected to automated Edmund degradation. The amino acid sequence of the ®rst 23 residues was determined to be: VKDGYIADDRNCPYFCGRNAYYD 3.3. BMK peptides alter sodium channels In order to determine whether the BMK peptides act to alter sodium channels to exert their prolongation eects on action potentials, the following experiments were performed. Whole sciatic nerves were dissected-free and mounted onto the sucrose-gap chamber and bathed via the central pool in Ringer's solution. Ringer's solution was then removed and replaced with 10 mM TEACl Ringer solution that also contained 100 mM CdCl2. TEA acts to block K channels while Cd acts to block calcium channels. The BMK peptides were then added to TEA CdCl2 Ringer. Fig. 6 shows the eect of BMK peptides 16(3), 16(4) and 16(5) on nerves exposed to TEA/Cd Ringer. In each case the BMK peptide acted to alter the size of the action potential and its width in the presence of the blockers TEA and Cd. BMK 16(3) and 16(5) dramatically increased the width of the AP, while BMK 16(4) had little eect on the width of the AP. Each of the peptides acted to attenuate the height of the AP. BMK 16(4) acted principally to attenuate the AP, while the other two peptides acted principally to lengthen the APs duration. All three peptides attenuate the AP to some extent. The attenuation and prolongations of the AP that were seen were not completely reversed upon return to TEA/Cd Ringer.
Fig. 7. Peptide similarity to BMK 16(5). Shown above are the ®rst 23 amino acids of peptide BMK 16(5) and the complete sequences of seven other peptides. The amino acid sequence of each peptide is represented using the standard one letter code for each amino acid. On the right are shown the sequence identity between BMK 16(5) and the ®rst 23 amino acids of each peptide.
658
R. Hahin et al. / Toxicon 38 (2000) 645±660
4. Discussion 4.1. Comparison of results Shown in Fig. 7 are one letter codes representing the ®rst 23 amino acids of neurotoxin BMK 16(5) and seven other neurotoxins obtained from other scorpion venoms. The peptides are displayed so that each of the cysteines are correspondingly lined up to facilitate comparison. The cysteines in BMK 16(5) have been presumptively identi®ed because of their sequence similarity and the presence of cysteine from a single time point amino acid analysis performed on a Beckman amino acid analyzer. The degree of sequence identity with the BMK 16(5) peptide is shown to the right of each peptide. Each of these toxins like BMK 16(5) acts to alter sodium channels. The toxins showing the greatest degree of similarity derived from Buthus eupeus and Buthus occitanus tunetanus. Unlike all the other neurotoxic peptides shown, BMK 16(5) was substantially smaller. Despite its small size, BMK 16(5) still possesses the ability to bind to and alter sodium channels from mouse and frog nerve. 4.2. BMK 16(5) is a small peptide that retains the ability to alter APs Despite the fact that BMK 16(5) is smaller than all the other peptides, the peptide still retains the ability to alter APs in frog and mouse nerves. An application of 480 nM BMK 16(5) to a frog nerve causes a large prolongation of the AP (Fig. 3; inset) within 10 min that continued following sustained exposure to the peptide, which suggests the peptide acts to alter Na channels. BMK 16(3) and 16(4), larger peptides, added at concentrations of 0.73 mg/ml (ca 177 nM) and 0.6 mg/ml (ca 117 nM) respectively, act to also greatly prolong APs in frog nerve. These larger peptides more potently and eectively lengthened the AP than BMK 16(5). BMK 16(5) and BMK 16(4) also attenuated (Fig. 3) APs during time following their application to a nerve, unlike BMK 16(3) that acted solely to lengthen the AP. This suggests that BMK 16(5)/BMK 16(4) may act to block Na channels in addition to slowing Na channel inactivation. Alternatively, BMK 16(5)/BMK 16(4) may act to alter fast or slow steady state inactivation to produce what appears to be a blocking eect. To support this idea, Fig. 6 shows that BMK 16(5) acts to greatly prolong APs in the presence of TEA and Cd. Since TEA and Cd act to block K and Ca channels, respectively, the greatly prolonged APs observed in the presence of BMK 16(5) suggest that the BMK peptide acts to alter Na channels to exert its eect. 100 mM CdCl2 will act to block some Na channels and shift the voltage dependence of their opening to some extent, however, the amount of Cd present will be insucient to block a substantial amount of Na channels. Therefore, the attenuation of the AP produced in the presence of BMK 16(5) suggests that the peptide also attenuates Na currents. Voltage-clamp and patch clamp experiments will provide clues to the mechanism of action of these peptides. BMK 16(5) may be a proteolysis product during the puri®cation procedure. At
R. Hahin et al. / Toxicon 38 (2000) 645±660
659
this juncture it is not clear whether or not BMK 16(5) is actually present in the crude venom itself. The peptide bond Asp±Pro is labile and undergoes hydrolysis in mild acidic solutions and the twenty-third residue of BMK 16(5) is an Asp residue. The MTX2 toxin from the mamba snake, Dendroaspis angusticepa, is rapidly and eciently cleaved in acidic solutions at an Asp±Pro bond (Segalas et al., 1995). Since this is a possibility, BMK 16(5) may be a truncated version of a larger peptide found in the venom. If this is so, the small sized BMK 16(5) appears to retain its ability to alter sodium channels. BMK 16(5) possesses a proline at residue 13, which is not found in any of the other scorpion toxins shown in Fig. 7. Prolines have been previously shown to be associated with short and long (>10 residues) loops (Martin et al., 1995) and shown to be a key residue in a structural motif that de®ne beta turns (Kessler et al., 1994; Muller et al., 1993). The presence of the proline in BMK 16(5) may provide a clue to how the small peptide folds so as to retain its ability to bind to Na channels and act to alter Na channels so as to greatly prolong APs.
Acknowledgements We would like to kindly acknowledge Professor John Mitchell for carefuly reading this manuscript and providing helpful comments.
References Catterall, W.A., 1980. Neurotoxins that act on voltage-sensitive sodium channel in excitable membranes. Annu. Rev. Pharmacol. Toxicol. 20, 15±43. Catterall, W.A., 1986. Molecular properties of voltage sensitive sodium channels. Annu. Rev. Biochem. 55, 953±985. Catterall, W.A., Beress, L., 1978. Sea anemone toxin and scorpion toxin share a common receptor site associated with the action potential sodium ionophore. J. Biol. Chem. 253, 7393±7396. Courard, F., Rochat, H., Lissitzky, S., 1978. Binding of scorpion and sea anemone neurotoxins to a common site related to the action potential sodium+ ionophore in neuroblastoma cells. Biochem. Biophys. Res. Commun. 83, 1525±1530. Fontecilla-Camps, J.C., Almassy, R.J., Suddath, F.L., Watt, D.D., Bugg, C.E., 1980. Three-dimensional structure of a protein from scorpion venom: A new structural class of neurotoxins. Proc. Natl Acad. Sci. 77, 6496±6500. Hahin, R., Strichartz, G., 1981. Eects of duterium oxide on the rate and dissociation constants for saxotoxin and tetrototoxin action. J. Gen. Physiol. 78, 113±139. Kessler, H., Matter, H., Gemmecker, G., Diehl, H.J., Isernia, C., Mronga, S., 1994. The pseudo-beta Iturn. A new structural motif with cis peptide bond in cyclic hexapeptides. Int. J. Pepp. Protein Res. 43, 47±61. Kopeyan, C., Martinez, G., Rochat, H., 1978. Amino acid sequence of neurotoxin V from the scorpion Leiurus quinquestriatus quinquestriatusrm. FEBS Lett. 89, 54±58. Martin, M.F., Rochat, H., 1984. Puri®cation of 13 toxins active on mice from the venom of the north African scorpion Buthus occitanus tumetanus. Toxicon 22, 279±291. Martin, A.C., Toda, K., Stirk, H.J., Thornton, J.M., 1995. Long loops in proteins. Protein Eng. 8, 1093±1101.
660
R. Hahin et al. / Toxicon 38 (2000) 645±660
Martin-Eauclaire, M.S., Courard, S., 1995. Scorpion Neurotoxins: Eects and Mechanisms. In: Chang, L.W., Dyer, R.S. (Eds.), Handbook of Neurotoxicology. Dekker, New York, pp. 683±716. Miranda, F., Kopeyan, C., Rochat, H., Rochat, C., Lissitzky, S., 1970. Puri®cation of animal neurotoxins. Isolation and characterization of 11 neurotoxins from the venoms of the scorpion Androctonus australis Hector, Buthus occitanus tunetanus, and Leiurus quinquestriatus quinquestriatus. European J. Biochem. 16, 514±523. Muller, G., Gurrath, N., Kurz, M., Kessler, H., 1993. Beta VI turns in peptides and proteins, a model peptide mimicry. Proteins 15, 235±251. Rochat, H., Rochat, C., Sampieri, F., Miranda, F., Lissitzky, S., 1972. The amino acid sequence of neurotoxin II of Androctonus australis Hector. Eur. J. Biochem. 28, 381±388. Rogers, J.C., Qu, Y., Tanada, T.M., Scheuer, T., Catterall, W.A., 1996. Molecular determinants of high anity binding of alpha scorpion toxin and sea anemone toxin in the S3-S4 extracellular loop in domain 4IV of the sodium channel alpha subunit. J. Biol. Chem. 271, 15950±15962. Rosso, J.P., Rochat, H., 1985. Characterization of 10 proteins from the venom of the Moroccan scorpion Androctonus mauretanicus mauretanicus, six of which are toxic to the mouse. Toxicon 23, 113± 125. Segalas, I., Thai, R., Menez, R., Zita, Z., 1995. A particularly labile Asp±Pro bond in the green mamba muscarinic toxin MTX2. Eect of protein con®rmation on the rate of cleavage. FEBS Lett. 371, 171±175. Vargas, O., Gregoire, J., Martin, M-F., Bechis, G., Rochat, H., 1982. Toxicon 20, 79±79. Volkova, T.M., Garsia, A.F., Telezhinskaia, I.M., Potapenko, N.A., Grishin, E.V., 1985. Neurotoxins from the venom of the central Asian scorpion Buthus eupeus. Bioorg. Khim. 11, 1445±1456.
The isolation and characterization of a peptide that alters sodium channels from Buthus martensii Karsch Richard Hahin a,*, Ziyi Chen a, Giridher Reddy b b
a Biological Sciences, Northern Illinois University, DeKalb, IL 60115, USA Department of Pediatrics, Biological Sciences Division, University of Chicago, Chicago, Illinois, USA
Received 26 November 1998; accepted 14 July 1999
Abstract The peptides were puri®ed using gel ®ltration, ion exchange, FPLC, and HPLC chromatography and found to greatly prolong action potentials at nanomolar concentrations when applied to frog and mouse nerves. The N-terminal primary amino acid sequence of one of the peptides, BMK 16(5), was determined. The ®rst 23 amino acids of BMK 16(5) were found to be: VKDGYIADDRNCPYFCGRNAYYD. The two cysteine residues in the sequence appeared as Edman sequence cycle blanks; however, they were assigned to be cysteines due to sequence similarity to other peptide toxins that bind to sodium channels and identi®cation of the presence of cysteines obtained from single time point amino acid analysis. The MW of BMK 16(5) was determined by a Perkin±Elmer API 300 LC/MS/MS to be 3695. The amino acid residues of BMK 16(5) show strong similarity with the ®rst 23 amino acid residues of a number of scorpion alpha neurotoxins. Unlike these neurotoxins, BMK 16(5) possesses a proline residue at position 13 which will likely make it fold in a unique way so as to bind to and alter sodium channels. # 1999 Elsevier Science Ltd. All rights reserved.
1. Introduction A number of peptides that speci®cally bind to and alter sodium channels have * Corresponding author. Fax: +1-815-753-0461. 0041-0101/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 1 - 0 1 0 1 ( 9 9 ) 0 0 1 8 0 - 4
646
R. Hahin et al. / Toxicon 38 (2000) 645±660
been previously described in Martin-Eauclaire and Courard (1995). The ®rst peptide activator known to bind to sodium channels and alter the sodium channel conductance was described by Miranda et al. (1970). Peptides that have been isolated from scorpion venoms that bind to and alter vertebrate sodium channel inactivation have been designated as alpha scorpion toxins. Alpha scorpion toxins bind speci®cally to particular sites on the sodium channel. Distinct receptor binding sites have been predominately characterized using vertebrate excitable preparations (Catterall, 1980, 1986). The alpha scorpion toxins that have been shown to act to lengthen sodium channel inactivation have been shown to bind to a region on the sodium channel designated as receptor site 3 in the rat brain (Catterall and Beress, 1978; Courard et al., 1978; Rogers et al., 1996). Alpha scorpion toxins typically possess a molecular weight in the range of 6000±7500 and contain eight cysteine residues that participate in forming four disul®de bonds. In this report we describe the isolation and puri®cation of a novel peptide from the venom of the scorpion Buthus martensii Karsch. This peptide, which is quite small in size (M.W. 3695) acts to alter sodium channels so that action potentials are greatly prolonged. The peptide also shows sequence similarities to parts of several alpha scorpion toxin peptides that have been shown to bind to and alter sodium channel inactivation. 2. Materials and methods 2.1. Source of venom Lyophilized crude venom was obtained from the People's Republic of China. 2.2. Solubilization of crude venom Crude venom (100 mg) was dissolved in 1 ml of 20 mM ammonium acetate (pH 4.7) buer at 48C for 40 min. The venom solution was centrifuged at 20,000 G for 20 min to remove the insoluble material. The supernatant containing the solubilized venom was then puri®ed using gel ®ltration and ion exchange column chromatography. 2.3. Gel ®ltration, ion exchange, and reversed phase HPLC chromatography Solubilized crude venom was applied to a 1.6 100 cm column containing Sephadex G50 equilibrated with a 20 mM ammonium acetate buer (pH 4.7) at T = 208C. The buer ¯ow rate was held at 6 ml per hour. One ml of solubilized venom (100 mg/ml) was applied to the column and the eluted material was collected in 2 ml fractions. The optical absorbance of the eluent was measured at 280 nm. Fractions were separated and assayed for their ability to lengthen action potentials (APs) recorded using the sucrose gap method. Fractions displaying
R. Hahin et al. / Toxicon 38 (2000) 645±660
647
biological activity were pooled and concentrated using an ultra®ltration membrane (Amicon YM 1) with a 1000 Da retention cut o to produce solutions containing biologically active material. Concentrated biologically active fractions were loaded onto a 2.5 50 cm column containing CM Sephadex C-50 which was equilibrated with 20 mM ammonium acetate buer (pH 4.7) at 208C. Fractions were eluted and collected with a sequence of 3 linear sodium chloride gradients using a ¯ow rate of 50 ml/h. Twenty absorbance peaks were observed and 8 contained material which acted to lengthen APs in frog and mouse nerve. Fractions from peak 16 were pooled and concentrated and applied to a 5 50 mm Mono-S HR5/5 (Pharmacia Biotech Inc., Piscataway, NJ) strong cation exchange resin column equilibrated with 20 mM phosphate buer. A sequence of linear sodium chloride gradients in 20 mM phosphate buer (pH 7.0) were used to elute peptides from the column. The ¯ow rate was 1 ml/min. Absorbance was detected at 280 nm. Fractions were collected, assayed for activity, and further puri®ed using a C18 reversed phased HPLC column (4.6 250 mm). The ¯ow rate was 1 ml/min and peptides were detected at 216 nm. 2.4. Polyacrylamide gel electrophoresis Polyacrylamide gel electrophoresis was performed using the following protocol: acrylamide (30%) and Bis acrylamide (0.8%) and K-acetate were polymerized by ammonium per sulfate and N,N,N ',N '-tetramethylethylenediame (TEMED). A 5% stacking gel was applied above a 15% acrylamide resolving gel. The electrode buer was K-acetate (pH 4.3). Slab gels were stained in 0.25% Coomasie blue for 40 min and destained in 10% acetic acid/30% methanol. 2.5. Molecular weight determination of puri®ed peptides Initial estimates of molecular weight were obtained with a Superdex Peptide HR 10/30 (Pharmacia Biotech Inc., Piscataway, NJ) gel ®ltration column displaying optimal resolution within the molecular weight range of 100±7000. Peptides and standard proteins were applied to the column using a 50 mM pH 7.0 phosphate buer and eluted at a ¯ow rate of 0.5 ml/min. Catalase (MW 220,000) which is totally excluded from the gel was used to determine the void volume of the column. To estimate the molecular weights, a standard curve plotting the partition coecient vs the log molecular weight of a number of molecular weights standards was used. A PE (Perkin±Elmer) SCIEX API 300 LC/MS/MS mass spectrometer at the University of Chicago was used to more accurately and precisely determine the MW of one of the peptides (BMK 16(5)). A major spectral peak occurred at a MW of 3695. Associated with the major peak at various charge states were peaks at 1848 at +2, 1232 at +3, and 1923 at +4. A minor spectral peak occurred at 2112, which is likely a degraded product of the original peptide. Associated with the 2112 peak were peaks at 1056 at +2 and 699 at +3. Another peak occurred at 7392, which is likely a dimer of the 3695 molecule.
648
R. Hahin et al. / Toxicon 38 (2000) 645±660
2.6. Cysteine determination A single time point amino acid analysis was performed. Acid hydrolysis was performed on the sample using 6 N double distilled constant boiling HCl for 20 h at 1258C in a nitrogen purged sealed container. Presence of cysteines was con®rmed. 2.7. Amino acid sequence determination The sequence of the peptide was carried out using a pulsed liquid protein ABI 494 Procise Edman N-terminus sequencer at the University of Chicago. Blanks in this sequence, which could be interpreted as either cysteines or tryptophans, were tentatively assigned to be cysteines due to similarity with other peptides and the presence of cysteine found by single time point amino acid analysis experiments performed using a Beckman amino acid analyzer. Cysteines so identi®ed are indicated in the sequence as C. Table 1 shows the pmole yields obtained for each Edman degradation cycle. 2.8. Biological activity assay Fractions to be tested were dissolved in either Ringer's solution (for frog Table 1 Yield obtained for each Edmandegredation cycle Residue number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 17 18 19 20 21 22 23
Amino acid ID
Yield (pmol)
Val Lys Asp Gly Tyr Ile Ala Asp Asp Arg Asn
30.92 25.45 23.80 26.84 27.21 23.15 24.17 15.67 21.93 10.15 14.21
Pro Tyr
11.54 12.66
Gly Arg Asn Ala Tyr Tyr Asp
10.44 6.79 7.50 8.77 7.16 7.84 3.80
R. Hahin et al. / Toxicon 38 (2000) 645±660
649
nerves) or Tyrode solution (for mouse nerves). Compound actions potentials (APs) were elicited using a technique adapted from Hahin and Strichartz (1981) and brie¯y described below. A sciatic nerve was placed in the chamber so that the nerve spanned 2 compartments; the proximal end (placed in the intracellular compartment) was bathed in isosmotic KCl, and the middle (placed in the
Fig. 1. Absorbance pro®le of G-50 gel ®ltration puri®cation step 1. The ordinate represents absorbance units. The abscissa represents the fraction number; each fraction collected represents 2 ml. Arrows signify absorbance peaks that contained peptides that acted to alter sodium (Na) channels. Absorbance peaks 2 and 3 were biologically active in modifying the APs in mouse and frog nerves; peaks 2 and 3 were comprised of fractions 29±37 and 38±48, respectively. Biological activity was de®ned to be the ability of a fraction to lengthen the action potential (AP) when applied to a frog or a mouse sciatic nerve. The two insets depict two experiments designed to show the biological action of pooled fractions obtained from peaks 2 and 3 on frog nerves. APs were recorded sequentially in time prior to and after the addition of an aliquot of pooled fraction 2 (PF2) and pooled fraction 3 (PF3) to the Ringer's solution bathing the nerves. The ®rst AP on the left was recorded prior to the addition of PF2/PF3, and each subsequent AP was recorded sequentially in time after the addition of PF2/PF3 at 5 min intervals in time; each AP was shifted rightward a constant amount to represent the passage of time (5 min interval).
650
R. Hahin et al. / Toxicon 38 (2000) 645±660
extracellular compartment) was bathed in either control solution (Ringer's or Tyrode's) or control solution plus the fraction to be tested, and was separated from KCl by a partition through which 0.22 molar sucrose ¯owed. Sucrose ¯ow was maintained constant (ca 3±8 ml/min) throughout the recording period. APs were elicited supermaximally stimulating the distal end of the nerve periodically with a suction electrode. A brief (100 ms) constant voltage pulse was used to elicit APs. APs were dierentially recorded using Ag/AgCl electrodes by obtaining the potential in the intracellular compartment relative to the potential in the extracellular compartment. APs were recorded on photographs from repetitively applied traces on a ZC6524 digital storage oscilloscope (Hitachi Denshi Ltd, Japan). 3. Results 3.1. Fractionation of scorpion venom Venom from the scorpion Buthus martensii Karsch was fractionated initially with a Sephadex G50 gel ®ltration column. Fig. 1 shows the results of adding 1 ml of solubilized venom (100 mg/ml) to the column. Four absorbance peaks were obtained and two of them (denoted by arrows) contained biologically active material that acted on the sodium channels in frog and mouse nerve to lengthen action potentials (APs). Fractions in absorbance peaks 2 and 3 were separately pooled to form pooled fractions designated as PF2 and PF3. The two insets in Fig. 1 depict two experiments designed to show the biological action of PF2 and PF3 on frog nerves. APs were recorded sequentially in time prior to and after the addition of an aliquot of PF2/PF3 to the Ringer solution bathing the nerves. The ®rst AP on the left was recorded prior to the addition of PF2 (left inset) or PF3
Fig. 2. CM Sephadex C-50 chromatogram. The ordinate represents absorbance units. The abscissa represents the fraction number; each fraction collected represents 8 ml. Arrows signify absorbance peaks that contained peptides that acted to alter Na channels. 20 absorbance peaks were observed; 9 peaks contained fractions with peptides which acted to alter APs. The biologically active peaks were 9± 14, and 16±18.
R. Hahin et al. / Toxicon 38 (2000) 645±660
651
Fig. 3. FPLC cation exchange chromatography of fraction 16. Fraction 16 from the CM-Sephadex C50 was puri®ed by FPLC rechromatography. The column used was a Mono-S HR 5/5 cation exchange column. Buer A was 20 mM phosphate while buer B was phosphate containing 1 M NaCl. The gradients used were: 0±6% B in 5 min, 6±8% B in 2 min, 8% B for 1.5 min, 8±20% B in 13 min, 20± 100% B in 1.5 min, 100% B in 1 min, and 100±0% B in 2 min. The ¯ow rate was 1 ml/min. Peptides were detected at 280 nm. The arrows indicate the fractions that were biologically active. The insets at the top of the ®gure represent the progressive change in the AP produced by fraction 16 (3) (left inset) applied at 177 nM, fraction 16 (4) (middle inset) applied at 117 nM, and fraction 16 (5) (right inset) applied at 480 nM. The ®rst AP recorded in each inset was obtained prior to the application of a fraction, while each other successively recorded AP exhibited progressive shape changes as time proceeded following exposure of the nerve to each fraction. Each AP was shifted rightward a constant amount to indicate the passage of time (5 min intervals).
652
R. Hahin et al. / Toxicon 38 (2000) 645±660
Fig. 4. Molecular weight estimates of BMK peptide fractions. Shown on the left hand side of Fig. 4 are three chromatograms obtained from a Superdex Peptide HR-10/30 (Pharmacia Inc., Piscataway, NJ) gel ®ltration column displaying optimal resolution within the molecular weight range of 100±7000. On the right side of the ®gure at the top is a chromatogram run on the Superdex Peptide column using standard proteins. Peptide fractions and standard proteins were applied to the column using a 50 mM pH 7.0 phosphate buer and eluted at a ¯ow rate of 0.5 ml/min. The chromatograms on the left represent: fraction 16 (5) (top), fraction 16 (4) (middle), and at the bottom fraction 16 (3). Also shown as an inset to the chromatogram of fraction 16 (5) is a polyacrylamide gel electrophoresis pro®le of the crude venom (left lane) and fraction 16 (5) in the right lane. The standard proteins used to obtain a standard curve to estimate the molecular weight were catalase (M.W. 220,000), RNAase (M.W.13,700), alpha cobra toxin (M.W. 7820), Bombesin (M.W. 1619), and glycine (M.W. 75). At the bottom right of Fig. 4 is a standard curve plotting the log M.W. vs the partition coecient (Kav).
R. Hahin et al. / Toxicon 38 (2000) 645±660
653
(right inset), and each subsequent AP was recorded sequentially in time after the addition of PF2/PF3 at 5 min intervals in time; each AP was shifted rightward a constant amount to represent the passage of time (5 min interval). Eluent fractions representing PF2 and PF3 were pooled and concentrated using an ultra®ltration membrane with a 1000 Da cuto to produce a solution that was applied to an ion exchange column. Fig. 2 shows the results of adding concentrated PF2 and PF3 from the Sephadex G50 column to a weak cation exchange column (CM Sephadex C-50). PF2 and PF3 were combined since the two peaks could not be isolated and separated conveniently. Fractions were eluted and collected during the application of a sequence of linear sodium chloride gradients to the column. Twenty absorbance peaks were observed and nine (designated by arrows) contained material which acted to lengthen APs in frog and mouse nerve. Each of the fractions displaying biological activity showed diering amounts of activity. Fraction 16 was the most potent and was further puri®ed. Pooled fractions from peak 16 of the CM Sephadex C-50 column were applied to a Mono S strong cation exchange column using FPLC. A sequence of linear sodium chloride gradients in 20 mM phosphate buer (pH 7.0) were used to elute peptides from the column. The results of the application of the gradient pro®le are shown in Fig. 3. The third, fourth and ®fth absorbance peaks (of 7) contained peptides that modi®ed (arrows designate the active peaks) APs. Peaks 3±5 derive from single peptides that were designated as BMK 16(3), BMK 16(4), and BMK 16(5). The insets (top) show the eects on nerve APs of aliquots of BMK 16(3), BMK 16(4), and BMK 16(5) to the Ringer solution bathing a frog nerve. Each peptide acted to greatly prolong the AP compared to its initial control value (®rst AP in each inset) following a continued exposure to the peptides. A Superdex peptide HR10/30 gel ®ltration column displaying optimal resolution within the molecular weight range of 100±7000 was used to estimate the M.W. of the biologically active peptides. The top right hand side of Fig. 4 shows a chromatogram displaying the absorbance pro®le of a number of MW standards applied to the Superdex peptide column. On the left hand side of Fig. 4 are displayed three other chromatograms run on the three peptides (BMK 16(5), BMK 16(4) and BMK 16(3)) that were puri®ed. Each of the peptides appears as a single peak on each chromatogram. To estimate the MWs, a standard curve plotting the partition coecient vs the log M.W. was used. Fig. 4 (bottom right) shows the standard curve that was used to estimate the MW. Using this method, the M.W.s of BMK 16(3), BMK 16(4) BMK 16(5) were determined to be 4120, 5150, and 2471, respectively. Peptide BMK 16(5) was characterized more fully than the other two peptides, and its M.W. was determined more accurately using a PE SCIEX API 300 LC/MS/MS mass spectrometer. Using this method, the MW of BMK 16(5) was determined to be 3695. Fig. 4 (top left) also shows (inset) the results of polyacrylamide gel electrophoresis of the crude venom (left lane) and BMK 16(5) which appeared as a single band, suggesting it had been puri®ed to homogeneity. In order to ensure that each peptide was puri®ed to homogeneity, samples of
654
R. Hahin et al. / Toxicon 38 (2000) 645±660
R. Hahin et al. / Toxicon 38 (2000) 645±660
655
Fig. 5 (continued)
each peptide were applied to a reversed phase C18 HPLC column. Fig. 5 shows the absorbance pro®les for each of the peptides. Fig. 5A shows the absorbance pro®le of BMK fraction 16(3) from FPLC applied to the column. A single peak representing peptide BMK 16(3) was observed. Similarly, in Fig. 5B and C, single peaks appear in the HPLC chromatograms indicating that both peptides had been puri®ed to homogeneity. Since the yield for peptides BMK 16(3) and BMK 16(4) was insucient to provide enough material for amino acid sequence determination, they were not studied further at this time.
Fig. 5. HPLC puri®cation of BMK 16(3), 16(4), 16(5). Fig. 5 shows chromatograms obtained from the application of BMK 16(3) (Fig. 5A), BMK 16(4) (Fig. 5B), and BMK 16(5) (Fig. 5C) to a reversed phase C18 (4.6 250 mm) HPLC column that had been equilibrated with 0.1% tri¯uoroacetic acid (TFA) in water. The column was eluted using a linear gradient from solution A (0.1% TFA in water) to 60% B (0.1% TFA in acetonitrile) in 60 min at a ¯ow rate of 1 ml/min. In each chromatogram the absorbance at 216 nM is plotted as a function of the retention time.
656
R. Hahin et al. / Toxicon 38 (2000) 645±660
Fig. 6. BMK peptides act on sodium channels. Shown above are three graphs illustrating the action of BMK 16(3) (top), BMK 16(4) (middle), and BMK 16(5) (bottom) on whole sciatic nerves bathed in TEA/Cd2+ Ringer's solution. TEA/Cd2+ Ringer's solution contained 100 mM CdCl2 and 10 mM TEACl in Ringer's solution. The ®rst AP in each graph was recorded in TEA/Cd2+ Ringer's solution prior to the application of the peptide. The arrow designates the application of the peptide with each concentration applied. Each successive AP recorded after the application of the peptide was recorded every 5 min. A rightward displacement of the AP re¯ects the passage of 5 min intervals of time.
3.2. Chemical characterization of puri®ed BMK 16(5) Since HPLC chromatograms of BMK 16(5) revealed a single peak and polyacrylamide gel electrophoresis of BMK 16(5) yielded a single band, BMK
R. Hahin et al. / Toxicon 38 (2000) 645±660
657
16(5) was subjected to automated Edmund degradation. The amino acid sequence of the ®rst 23 residues was determined to be: VKDGYIADDRNCPYFCGRNAYYD 3.3. BMK peptides alter sodium channels In order to determine whether the BMK peptides act to alter sodium channels to exert their prolongation eects on action potentials, the following experiments were performed. Whole sciatic nerves were dissected-free and mounted onto the sucrose-gap chamber and bathed via the central pool in Ringer's solution. Ringer's solution was then removed and replaced with 10 mM TEACl Ringer solution that also contained 100 mM CdCl2. TEA acts to block K channels while Cd acts to block calcium channels. The BMK peptides were then added to TEA CdCl2 Ringer. Fig. 6 shows the eect of BMK peptides 16(3), 16(4) and 16(5) on nerves exposed to TEA/Cd Ringer. In each case the BMK peptide acted to alter the size of the action potential and its width in the presence of the blockers TEA and Cd. BMK 16(3) and 16(5) dramatically increased the width of the AP, while BMK 16(4) had little eect on the width of the AP. Each of the peptides acted to attenuate the height of the AP. BMK 16(4) acted principally to attenuate the AP, while the other two peptides acted principally to lengthen the APs duration. All three peptides attenuate the AP to some extent. The attenuation and prolongations of the AP that were seen were not completely reversed upon return to TEA/Cd Ringer.
Fig. 7. Peptide similarity to BMK 16(5). Shown above are the ®rst 23 amino acids of peptide BMK 16(5) and the complete sequences of seven other peptides. The amino acid sequence of each peptide is represented using the standard one letter code for each amino acid. On the right are shown the sequence identity between BMK 16(5) and the ®rst 23 amino acids of each peptide.
658
R. Hahin et al. / Toxicon 38 (2000) 645±660
4. Discussion 4.1. Comparison of results Shown in Fig. 7 are one letter codes representing the ®rst 23 amino acids of neurotoxin BMK 16(5) and seven other neurotoxins obtained from other scorpion venoms. The peptides are displayed so that each of the cysteines are correspondingly lined up to facilitate comparison. The cysteines in BMK 16(5) have been presumptively identi®ed because of their sequence similarity and the presence of cysteine from a single time point amino acid analysis performed on a Beckman amino acid analyzer. The degree of sequence identity with the BMK 16(5) peptide is shown to the right of each peptide. Each of these toxins like BMK 16(5) acts to alter sodium channels. The toxins showing the greatest degree of similarity derived from Buthus eupeus and Buthus occitanus tunetanus. Unlike all the other neurotoxic peptides shown, BMK 16(5) was substantially smaller. Despite its small size, BMK 16(5) still possesses the ability to bind to and alter sodium channels from mouse and frog nerve. 4.2. BMK 16(5) is a small peptide that retains the ability to alter APs Despite the fact that BMK 16(5) is smaller than all the other peptides, the peptide still retains the ability to alter APs in frog and mouse nerves. An application of 480 nM BMK 16(5) to a frog nerve causes a large prolongation of the AP (Fig. 3; inset) within 10 min that continued following sustained exposure to the peptide, which suggests the peptide acts to alter Na channels. BMK 16(3) and 16(4), larger peptides, added at concentrations of 0.73 mg/ml (ca 177 nM) and 0.6 mg/ml (ca 117 nM) respectively, act to also greatly prolong APs in frog nerve. These larger peptides more potently and eectively lengthened the AP than BMK 16(5). BMK 16(5) and BMK 16(4) also attenuated (Fig. 3) APs during time following their application to a nerve, unlike BMK 16(3) that acted solely to lengthen the AP. This suggests that BMK 16(5)/BMK 16(4) may act to block Na channels in addition to slowing Na channel inactivation. Alternatively, BMK 16(5)/BMK 16(4) may act to alter fast or slow steady state inactivation to produce what appears to be a blocking eect. To support this idea, Fig. 6 shows that BMK 16(5) acts to greatly prolong APs in the presence of TEA and Cd. Since TEA and Cd act to block K and Ca channels, respectively, the greatly prolonged APs observed in the presence of BMK 16(5) suggest that the BMK peptide acts to alter Na channels to exert its eect. 100 mM CdCl2 will act to block some Na channels and shift the voltage dependence of their opening to some extent, however, the amount of Cd present will be insucient to block a substantial amount of Na channels. Therefore, the attenuation of the AP produced in the presence of BMK 16(5) suggests that the peptide also attenuates Na currents. Voltage-clamp and patch clamp experiments will provide clues to the mechanism of action of these peptides. BMK 16(5) may be a proteolysis product during the puri®cation procedure. At
R. Hahin et al. / Toxicon 38 (2000) 645±660
659
this juncture it is not clear whether or not BMK 16(5) is actually present in the crude venom itself. The peptide bond Asp±Pro is labile and undergoes hydrolysis in mild acidic solutions and the twenty-third residue of BMK 16(5) is an Asp residue. The MTX2 toxin from the mamba snake, Dendroaspis angusticepa, is rapidly and eciently cleaved in acidic solutions at an Asp±Pro bond (Segalas et al., 1995). Since this is a possibility, BMK 16(5) may be a truncated version of a larger peptide found in the venom. If this is so, the small sized BMK 16(5) appears to retain its ability to alter sodium channels. BMK 16(5) possesses a proline at residue 13, which is not found in any of the other scorpion toxins shown in Fig. 7. Prolines have been previously shown to be associated with short and long (>10 residues) loops (Martin et al., 1995) and shown to be a key residue in a structural motif that de®ne beta turns (Kessler et al., 1994; Muller et al., 1993). The presence of the proline in BMK 16(5) may provide a clue to how the small peptide folds so as to retain its ability to bind to Na channels and act to alter Na channels so as to greatly prolong APs.
Acknowledgements We would like to kindly acknowledge Professor John Mitchell for carefuly reading this manuscript and providing helpful comments.
References Catterall, W.A., 1980. Neurotoxins that act on voltage-sensitive sodium channel in excitable membranes. Annu. Rev. Pharmacol. Toxicol. 20, 15±43. Catterall, W.A., 1986. Molecular properties of voltage sensitive sodium channels. Annu. Rev. Biochem. 55, 953±985. Catterall, W.A., Beress, L., 1978. Sea anemone toxin and scorpion toxin share a common receptor site associated with the action potential sodium ionophore. J. Biol. Chem. 253, 7393±7396. Courard, F., Rochat, H., Lissitzky, S., 1978. Binding of scorpion and sea anemone neurotoxins to a common site related to the action potential sodium+ ionophore in neuroblastoma cells. Biochem. Biophys. Res. Commun. 83, 1525±1530. Fontecilla-Camps, J.C., Almassy, R.J., Suddath, F.L., Watt, D.D., Bugg, C.E., 1980. Three-dimensional structure of a protein from scorpion venom: A new structural class of neurotoxins. Proc. Natl Acad. Sci. 77, 6496±6500. Hahin, R., Strichartz, G., 1981. Eects of duterium oxide on the rate and dissociation constants for saxotoxin and tetrototoxin action. J. Gen. Physiol. 78, 113±139. Kessler, H., Matter, H., Gemmecker, G., Diehl, H.J., Isernia, C., Mronga, S., 1994. The pseudo-beta Iturn. A new structural motif with cis peptide bond in cyclic hexapeptides. Int. J. Pepp. Protein Res. 43, 47±61. Kopeyan, C., Martinez, G., Rochat, H., 1978. Amino acid sequence of neurotoxin V from the scorpion Leiurus quinquestriatus quinquestriatusrm. FEBS Lett. 89, 54±58. Martin, M.F., Rochat, H., 1984. Puri®cation of 13 toxins active on mice from the venom of the north African scorpion Buthus occitanus tumetanus. Toxicon 22, 279±291. Martin, A.C., Toda, K., Stirk, H.J., Thornton, J.M., 1995. Long loops in proteins. Protein Eng. 8, 1093±1101.
660
R. Hahin et al. / Toxicon 38 (2000) 645±660
Martin-Eauclaire, M.S., Courard, S., 1995. Scorpion Neurotoxins: Eects and Mechanisms. In: Chang, L.W., Dyer, R.S. (Eds.), Handbook of Neurotoxicology. Dekker, New York, pp. 683±716. Miranda, F., Kopeyan, C., Rochat, H., Rochat, C., Lissitzky, S., 1970. Puri®cation of animal neurotoxins. Isolation and characterization of 11 neurotoxins from the venoms of the scorpion Androctonus australis Hector, Buthus occitanus tunetanus, and Leiurus quinquestriatus quinquestriatus. European J. Biochem. 16, 514±523. Muller, G., Gurrath, N., Kurz, M., Kessler, H., 1993. Beta VI turns in peptides and proteins, a model peptide mimicry. Proteins 15, 235±251. Rochat, H., Rochat, C., Sampieri, F., Miranda, F., Lissitzky, S., 1972. The amino acid sequence of neurotoxin II of Androctonus australis Hector. Eur. J. Biochem. 28, 381±388. Rogers, J.C., Qu, Y., Tanada, T.M., Scheuer, T., Catterall, W.A., 1996. Molecular determinants of high anity binding of alpha scorpion toxin and sea anemone toxin in the S3-S4 extracellular loop in domain 4IV of the sodium channel alpha subunit. J. Biol. Chem. 271, 15950±15962. Rosso, J.P., Rochat, H., 1985. Characterization of 10 proteins from the venom of the Moroccan scorpion Androctonus mauretanicus mauretanicus, six of which are toxic to the mouse. Toxicon 23, 113± 125. Segalas, I., Thai, R., Menez, R., Zita, Z., 1995. A particularly labile Asp±Pro bond in the green mamba muscarinic toxin MTX2. Eect of protein con®rmation on the rate of cleavage. FEBS Lett. 371, 171±175. Vargas, O., Gregoire, J., Martin, M-F., Bechis, G., Rochat, H., 1982. Toxicon 20, 79±79. Volkova, T.M., Garsia, A.F., Telezhinskaia, I.M., Potapenko, N.A., Grishin, E.V., 1985. Neurotoxins from the venom of the central Asian scorpion Buthus eupeus. Bioorg. Khim. 11, 1445±1456.