Solution structure of IsTX A male scorpion toxin from Opisthacanthus madagascariensis (Ischnuridae) Nahoko Yamaji 1 , Li Dai 1 , Kenji Sugase 1 , Marta Andriantsiferana 2 , Terumi Nakajima 1 and Takashi Iwashita 1 1 Suntory Institute for Bioorganic Research, Mishima-Gun, Osaka, Japan; 2 Faculty of Science, University of Antananarivo, Madagascar The novel sex-specific potassium channel inhibitor IsTX, a 41-residue peptide, was i solated from the venom of male Opisthacanthus madagascariensis.Two-dimensionalNMR techniques revealed t hat the structure of IsTX contains a cysteine-stabilized a/b-fold. IsTX is classified, based on its sequential and structural similarity, in the scorpion short toxin family a-KTx6. The a-KTx6 family contains a single a-he lix and t wo b-strands connected by four disulfide brid- ges and binds to voltage-gated K + channels and apamin- sensitive Ca 2+ -activated K + channels. The three-dimen- sional structure of IsTX is similar to that of Heterometrus spinifer toxin (HsTX1). HsTX1 blocks the Kv1.3 channel a t picomolar concentrations, whereas IsTX has much lower affinities (10 000-fold). To investigate the structure–activity relationship, the geometry of sidechains and electrostatic surface potential maps were compared with HsTX1. As a result of the comparison of the primary structures, Lys27 of IsTX was conserved at the same position in HsTX1. The analogous Lys23 of HsTX1, the most critical residue for binding to potassium channels, binds to the channel pore. However, IsTX has fewer basic residues to interact with acidic channel surfaces than HsTX1. MALDI-TOF MS analysis clear ly indicated t hat IsTX was found in male scorpion venom, but not in female. This is the first report that scorpion venom contains sex-specific compounds. Keywords: cysteine-stabilized a/b-fold; NMR; potass ium channels; s corpion toxin; sex-specific toxin. The venom o f Opisthacanthus madagascariensis scorpions from the I schnuridae fam ily collec ted in Isalo (Madagascar) was investigated by MALDI-TOF MS. Many highly toxic venom components from the Buthidae family of scorpions have been reported [1–4]. However, there have been few reports on the venom of scorpions from the Ischnu ridae family, due to its lower toxicity [5,6]. The resultant mass spectrum differed among males and females. The bioassay of each HPLC fraction of venom revealed that a number of peptides showed either weak or no toxic effects with the exception of IsTX. Further more, male venom was shown to contain IsTX, whereas female venom did not. This is the first report of the sex-related scorpion venom component, IsTX [7,8]. IsTX is a 41-residue toxin and a member of the a-K scorpion toxin family that functions to block K + channels. The a-K scorpion toxins, comprising 23–43 amino acids, have a/b-fold structures stabilized with three or four disulfide bridges. These toxins have b een divided into 12 subfamilies (a-KTx1–12) based on the alignments of cysteine and highly conserved residues [9,10]. IsTX belongs to the four disulfide-bridged a-KTx6 subfamily composed of maurotoxin (MTX) [11–13], HsTX1 [14–16] and Pand- inus imperator toxin 1 (Pi1) [ 17](Fig. 1). These a-KTx6 toxins selectively act on voltage-gated and Ca 2+ -dependent K + channels [18–20]. All the members of a-KTx6 subfamily contain four disulfide bridges. HsTX1 and Pi1 show the same disulfide bridge pattern (C1-C5, C2-C6, C3-C7 and C4-C8) [14,17], however, only MTX shows a nonstandard pattern (C1-C5, C2-C6, C3-C4 and C7-C8) [11]. Even though MTX has a different disulfide bridge pattern from HsTX1 and Pi1, it possesses very similar global fold with other toxins. The structure of IsTX is composed of a single a-helix and triple b-strand connected by four cysteine-disulfide bridges. The amino acid sequence of IsTX is 50% similar to HsTX1 and MTX and 43% similar to Pi1. These findings indicate that IsTX is highly similar to the a-KTx6 s ubfamily. Although IsTX has a highly sequential and structural similarity to HsTX1, the activity of IsTX toward Kv1.3 is 10 000 times lower than that o f HsTX1. The reason for this difference may be elucidated by c omparing the structures of IsTX with HsTX1. Herein, the structure in solution and the interaction between IsTX and the voltage-gated K + channel has been investigated. The study is comprised of (a) characterization Correspondence to N. Yamaji, Suntory Institute for Bioorganic Research, 1-1-1 Wakayamadai, Shimamoto-Cho, Mishima-Gun, Osaka 618–8503, Japan. Fax: +81 75 962 2115; Tel.: +81 75 962 6142, E-mail: yamaji@sunbor.or.jp Abbreviations: DQF-COSY, double-quantum-filtered COSY; HSQC, heteronuclear single quantum coherence; TsTX-Ka, Tityus serrulatus toxin; HsTX1, Heterometrus spinifer toxin 1; IsTX, Ischnuridae toxin; MTX, maurotoxin; Pi1, Pandinus imperator toxin 1; RMSD, root mean square deviation; Kv, voltage-gated potassium channel; ShB, shaker K + channel; SKCa, small-conductance Ca 2+ -activated K + channel. Database: The coordinate for the 20 lo west co nformers IsTX has been submitted to the RCSB Protein Data Bank database under the accession number 1WMT. (Received 14 June 200 4, revised 26 July 2004, accepted 4 August 20 04) Eur. J. Biochem. 271, 3855–3864 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04322.x of IsTX, (b) determination of the three-dimensional structure of IsTX by NMR spectroscopy, (c) calculation of the surface potential map of IsTX. A comparison of the data with HsTX1 revealed the proposed functional sites for scorpion toxin binding to potassium channels. Materials and methods Scorpions (O. madagascariensis) w ere collected in Isalo (Madagascar) and maintained in our laboratory. The crude venom was collected by electrical stimulation of the scorpion telson and extracted using water. The supernatant was stored in the freezer under )70 °C. MALDI-TOF mass spectrometry The matrix for the MALDI-TOF MS analysis was prepared as follows: a-cyanohydroxycinnamic acid was dissolved in 1 : 1 (v/v) acetonitrile/H 2 O supplemented with 0.1% (v/v) trifluoroacetic acid in order to obtain a saturated solution. Samples m ixed with the matrix solution were placed on a sample plate and allowed to dry for 5 min. MS analysis was performed on a Voyager Elite MALDI-TOF mass spectrometer (Applied B iosystems; Fragmingham, MA, USA) in positive mode. Crud e venom was recorded in linear mode. Peptide purification The venom supernatant was loaded directly onto to an RP-C 18 HPLC column (Tosoh, TSK gel, O DS 120T, 250 · 4.6 mm, 5 lm particle size). The temperature was controlled at 40 °C. The flow rate was set at 1 mLÆmin )1 and a linear gradient f rom 0–64% (v/v) acetonitrile/H 2 O with 0.1% (v/v) trifluoroacetic acid was run for 60 min. The detector absorbance was set at 220 nm. Crude fraction 17, which contain ed the peptide IsTX, was collected and lyophilized. It was further purified by RP-C 18 HPLC (Tosoh, TSK gel, ODS 120T, 250 · 4.6 mm, 5 lm particle size) using a 60 min linear gradient from 0–32% (v/v) a cetonitrile/H 2 O w ith 0 .1% (v/v) trifluoroacetic acid at a flow rate of 1 mLÆmin )1 . The analysis was monitored at 220 nm and the fraction containing the p urified peptide w as collected and lyophilized. Primary structure analysis About 20 pmol o f purified peptide was reduced and alkylated using tributyl phosphine and 4-vinyl-pyridine according to the standard protocol [21]. The pyridylethyl- ated product was desalted by C 18 HPLC (Tosoh, TSK gel, ODS 80T, 0.25 · 150 mm) and detected using a MALDI- TOF MS detector. A 60 min linear gradient from 0–64% (v/v) acetonitrile/H 2 O with 0.1% (v/v) trifluoroacetic acid at a flow rate of 1 mLÆmin )1 was run. Half of the pyridylethylated peptide was app lied to an automated Edman degradation gas-phase sequencer (PPSQ-10; Shim- adzu, Kyoto, Japan) and the remaining half was digested using V8 proteases at 37 °C for 12 h. The digestion products were checked with MALDI MS and separated by RP-C 18 HPLC with the same conditions as described above. Four fractions were collected and analyzed by a combination of M S ladder s equencing and automated Edman degradation sequencing. cDNA cloning The total RNA from one scorpion venom gland w as prepared using TRIzol Regent (Invitrogen) according to the standard protocol. The cDNA sequences encoding the IsPT precursor were determined using 3¢/5¢ RACE experiments. The specific primers designed according to the known amino acid sequence were: 3¢ RACE: 3SP1, 5¢-GTXCA YACXAAYATHCCXTG-3¢;3SP2,5¢-AAYATHCCXT GYMGXGGXAC-3¢;3SP3,5¢-GAYTGYTAYGARC CXTGYGA-3¢.5¢ RACE: 5 SP1, 5¢-GACAGAAATTA CATTTTCGTAGCGT; 5SP2, 5¢-CCATGGACAATTG TTGTAGCAGTT-3¢;5SP3,5¢-TGCCTATTCATACAT TTTGCCCT-3¢. The PCR products were cloned into the pCR2.1 vector and the DNA sequence was analyzed using an ABI Prism 310 automated sequen cer (Perkin-E lmer, CA, USA). Fig. 1. Sequence alignment o f I sT X with selected toxins fro m a-KTx6 s ubfamily and d isulfide bridge patterns. The amino acid sequences of HsTX1 (H. spinnifer), Pi1 (P. imperator)andMTX(S. maurus) were aligned according to the eight half-cysteine resid ues. Gaps (–) have been int rodu ced to maximize the alignment. *, C-term inal carboxyamidated e xtremity. The t wo disulfide bridge patterns observed in HsTX1/Pi1 and M TX are shown below the sequences. 3856 N. Yamaji et al. (Eur. J. Biochem. 271) Ó FEBS 2004 NMR experiments and structural calculations Synthetic IsTX was purchased from Peptide Institute, Inc. (Osaka, J apan), and identified by HPLC and activity measurements. For NMR experiments, synthetic IsTX was dissolved in 0.5 mL of H 2 O/D 2 O [90/10 (v/v)] containing 50 l M NaN 2 , the sample concentration was 4.7 m M , and the pH was 3.91 in order to detect amide signals. All NMR spectra were recorded on a Bruker DMX-750 spectrometer (Bruker Biospin, Germany). The temperature was s et to 298 K. Chemical shifts were r eferenced to internal 3-(trimethylsilyl)[2,2,3,3- 2 H 4 ] propionate (TSP). 2D DQF- COSY [22], NOESY [23,24], TOCSY [25] and HSQC experiments were c onducted. The NOESY spectra were acquired with 50, 100 and 200 ms mixing times and water suppression was achieved using the WATERGATE sequence [26]. The TOCSY spectrum was recorded using a MLEV-17 pulse sequence with a 71 ms mixing time. The 2D spectra were recorded using time-proportional phase Incrementation (TPPI) for quadrature detection in the F1 dimension. Spectra were recorded at 512 points for t1 and 2048 points for t2. A 1 H, 13 C HSQC spe ctrum was recorded in natural abundan ce using the e cho-antiecho scheme [27,28]. The time domain data w ere p rocessed u sing XWIN - NMR 2.5 program (Bruker Biospin). Proton signal assignments were achieved using the stand- ard strategy described by Wu ¨ thrich [29] with the graphical software ANSIG .3.3 [30]. The DQF-COSY and Clean- TOCSY spectra gave the spin system fingerprint of the peptide. The s pin systems were then sequentially connected using the NOESY spectra. NOE volumes were converted into distance restraints classified as s trong (1.8–2.7 A ˚ ), medium (1.8–3.3 A ˚ ), weak (1.8–5.0 A ˚ ), and v ery weak (1.8–6.0 A ˚ ). Backbone amide proton temperature coefficients were calculated from the NOESY spectra recorded at four different temperatures from 288–318 K [31,32]. Hydrogen- deuterium exchange experiments were carried out by lyophilizing the H 2 O sample, redissoloved in 250 lLof 99.96% (v/v) D 2 O, and running 3 h TOCSY experiments. Slowly exchanging amide protons were interpreted as hydrogen-bond donors. Distance restraints for the identified hydrogen bonds were included in the subsequent structure calculations. The 3 J NH-Ha coupling constants were estima- ted from the DQF-COSY spectra using either the DECO program (Bruker Biospin) or directly measured from a 1D spectrum. The dihedral angles estimated from the 3 J NH-Ha values were used as / angle constraints within the range of )90° and )40° for 3 J NH-Ha < 5.5 Hz, between )160° and )80° for 3 J NH-Ha > 8 Hz [33]. Structural calculations were performed on IsTX using X - PLOR - NIH 2.0.5 [34–36] with 679 NOE-based distance restraints, which contain 199 sequential, 302 medium range and 178 long-range restraints. Starting structures were an extended strand conformation, and were performed using conjugate-gradient minimization. The disulfide bonds were included as pseudo-NOE restraints. In the first stage, the extended strand structures were subjected to 10 ps and 1000 steps of t orsion-angle molecular dynamics at 50 000 K. The structures were then s ubjected t o a 15 ps and 1500 steps slow- cooling t orsion angle molecular dynamics stage in which the temperature was reduced from 50 000 K to 298 K over 250 steps. Finally, the structures were subjected to 200 steps of conjugated-gradient minimization. The following scale fac- tors were used during the three steps: van der Waals, 0.1, 1 and 1; NOE, 150, 150 and 75; dihedral restraints, 100, 200 and 400 and R amachandran restraints, 0.002, 1.0 and 1.0. The initial runs for structure calculations were performed without hydrogen bond restraints, and the obtained struc- ture was examined. Fourteen protons from amide groups were identified to form hydrogen bonds according to the amide proton temperature coefficients and hydrogen-deu- terium exchange experiments. If the hydrogen bond restraint was in agreement with amide temperature coeffi- cients (> )4.6 p.p.b.ÆK )1 ) and exchange in D 2 Odata (signal present after 24 h of exchange at 298 K) and if an oxygen atom was within 2.8 A ˚ of an amide proton, the amide proton was identified a s the acceptor of the hydrogen bond. These restraints were then used in the following stage of structure calculations. The structures were checked for violations of geometric a nd experimental restraints and atom overlapping using AQUA 3.2 and PROCHECK - NMR 3.4 [37]. Finally, a set of 20 conformers was selected based on the least distance and dihedral angle restraint violations (deviations). The coordinate for the 20 lowest conformers of IsTX has been deposited in the RCSB Protein Data Bank (http://www.rcsb.org), accession code 1WMT. Results Purification and primary structure determination Scorpion (O. madagascariensis) venom contained several peptides with molecular masses between 4 and 5 kDa, whichwereassumedtobeshort-chainneurotoxins(Fig.2). However, when each C 18 HPLC fraction was tested for toxicity to crickets, only fraction 17 showed an apparent paralysis e ffect o n crickets (Fig. 3). This fraction was further purified and two components were collected. One was a peptide with an M r of 3147 and no toxicity effects, and the other w as a peptide with an M r of 4819 that caused p aralysis to crickets. This toxic peptide was isolated from a scorpion collected in Isalo (Madagascar) and was named IsTX. Due t o the limited amount o f natural peptide, the primary structure was determined using several techniques including MS/MS analysis, ladder sequencing, Edman degradation s equencing, and subsequently confirmed b y cDNA deduction. The p yridylethylated peptide was sequenced using an automated Edman degradation method and t he N-terminal 15 step was determined as shown in Fig. 4. The V8 digested products of the pyridylethylated peptide were monitored b y LC/MS. Four main fragments were detected in the LC/MS spectra and further separated by RP-HPLC. The amino acid sequences of the four fragments were determined using Edman and ladder sequencing. The complete sequence of IsTX was determined by combining the results obtained above and further confirmed by cDNA sequencing. Molecular biological analysis The precursor of IsTX was deduced from the c DNA sequence. Based on the known amino acid sequences, Ó FEBS 2004 Solution structure of IsTX (Eur. J. Biochem. 271) 3857 specific primers were designated as described above. Molecular cloning of the cDNA from the venom gland mRNA was performed using 5¢/3¢ RACE. The cDNA was completed by overlapping two fragments amplified by 3¢ and 5 ¢ RACE. O ver 10 c lones were obtained with a polyadenylation signal (AATAAA) in the 3¢ untranslated region and a poly(A) tail at the 3¢ end. The open reading frame of IsTX encodes a precursor of 63 residues, contain- ing a signal peptide of 21 amino acid residues and a m ature peptide of 41 residues followed by an additional basic Arg residue that was removed during carboxyl-processing. The deduced mature peptide sequences were completely consis- tent with results of amino acid sequencing (Fig. 5). Chemical synthesis The peptide was synthesized by Peptide Institute, Inc. (Osaka, J apan) using an automated s olid-phase peptide synthesizer 433-A (Applied Biosystems) based on the Fmoc- strategy. Disulfide bridges were oxidized by exposure to air. The identity of synthetic and natural peptides was con- firmed by MS analysis, coinjection experiments on RP-HPLC and capillary electrophoresis (CE). Structure quality NOE correlations of 50 ms 2D NOESY spectrum were used for sequential assignments. The Ha protons of Met29 and Asn38 were not observed due to the o verlap with the water resonance signal. The amide protons of Cys23 and Arg25 were sh ifted upfield relative to their positions in random coil structures, and the Ala24 amide proton w as shifted downfield. T his observation suggests that t hese protons were in close proximity to an aromatic ring. The structure of IsTX was calculated using 679 NOE distance constraints. The 20 accepted structures, which had no NOE violations (£ 0.4 A ˚ )areshowninFig.6A.The average RMSD was 0.385 A ˚ for backbone atoms and 0.975 A ˚ for all heavy atoms (without the C-terminal residues), which were poorly determined (Table 1). A percentage of 52.1% of all residues occupy the most favorable regions of the Ramachandran plot and 41.7% lie in additionally allowed regions. Fig. 3. Purification of IsTX. (A) The crude venom of scorpion was separated using RP-C 18 HPLC (250 · 4.6 mm) with a 60 min linear gradient from 0 to 64% acetonitrile/H 2 O with 0.1% trifluoroacetic acid. The flow rate was set at 1 mLÆmin )1 and the absorbance was monitored at 220 nm. F raction 17 was collected, as it showed a paralysis effect on crickets. (B) C 18 HPLC fraction 17 was further separated with a 60 min linear gradient from 0 to 32% acetonitrile/ H 2 O with 0.1% trifluoroacetic acid. Two main components including IsTX were collected. Fig. 4. Amino acid seq uence determination of IsTX. The reduced and pyridylethylated peptide was digested with V8 proteases a nd sequenced using Edman and ladder sequencing. The amino acid sequences determined by ladder sequencing are underlined. Fig. 2. MALDI-TOF MS spectrum of crude venom of two individual Scorpion (O. madaga- scariensis) in the mass range m/z 500–10000. The top spectrum is th at of a male scorpion and the bottom is of a female. Peptide IsTX was only found in the m ale. 3858 N. Yamaji et al. (Eur. J. Biochem. 271) Ó FEBS 2004 Overall structure description The s olution structure of IsTX comprises an a-helix running from residue Thr10 to Lys21 and three b-strands from His2 to Pro6, Cys27 to Met29 and His32 to Asn34. The two strands of the b-sheet are connected by a type II b-turn formed by residues Asn30 and Arg31 (Fig. 7). In order to further characterize the overall structure of the IsTX toxin, deviations from the random coil position of the Ca chemical shifts were analyzed using the chemical shift index method [38,39]. The Ca chemical shifts exhibit upfield shifts with respect to random coil values in the helical conformation and downfield shifts in the b-strand e xtended conformation. The results are in reasonably good agreement with what is expected based on other NMR parameters. Secondary structure NOE c orrelations and Ca-chemical shift values were used to identify elements of secondary structure in IsTX. Th e s trong HN-HN i,i+1 NOE correlations together with small 3 J NH-Ha coupling constants s uggested the presence o f an a-helix conformation extending from Thr10 to Lys21. A proline insertion i n t he middle of t he helix was observed t o cause d istortion from t he regular pattern. Medium-range Ha-HN i,i+3 NOE correlations between Glu15 and Lys21 Fig. 5. The cDNA and deduced amino acid sequences of I sTX. Amino acids are denoted by the on e-letter symbols. The c DN A encodes a 21 residue signal peptide (underlined), a mature 41 residue peptide (bold) and an additional Arg at the C-terminal end (boxed). The stop codon is indicated by ÔendÕ.The polyadenylation signal (aataaa) is italicized. Fig. 6. Solution structures of IsTX and HsTX1. (A) Stereoview of the 20 final stru ctures of IsTX superimposed over th e bac kbone atom s of the well- defined region (residues 2–39). (B) Ribbon diagram of the backbone peptide folding of IsTX illustrating the single a-helix and triple-stranded b-sheet. (C) Ribbon diagram of HsTX1 illustrating the single a-helix and double-stranded b-sheet (PDB code 1QUZ). Table 1. Structural statistics for the20lowestenergystructures. Measurement Value Ramachandran analysis (residues 2–40) a Residues in most favoured regions (%) 52.1 Residues in additional allowed regions (%) 41.7 Residues in generously allowed regions (%) 6.0 Residues in disallowed regions (%) 0.1 RMSDs between 20 conformers (residues 2–39) b Backbone (A ˚ ) (N,Ca,C,O) 0.385 All heavy atoms (A ˚ ) 0.975 Distance restraints Intraresidue (|i–j| ¼ 0) 199 Sequential (|iu ˆ j| ¼ 1) 189 Medium-range (2 £ |i–j| £ 4) 113 Long-range (|i–j| > 4) 178 Total 679 Deviations from idealized covalent geometry Bonds (A ˚ ) 0.0065 (± 0.0007) Angles (°) 0.719 (± 0.026) Impropers (°) 0.619 (± 0.032) Total energies (kcalÆmol )1 ) 322.32 a PROCHECK - NMR was used to calculate these values. b None of these structures exhibited distance violations > 0.4 A ˚ . Ó FEBS 2004 Solution structure of IsTX (Eur. J. Biochem. 271) 3859 and HN-HN i,i+2 NOE correlations between Thr10 and Glu15 clearly identify residues Thr10-Lys21 as a-helical. Strong Ha-HN i,i+1 connectivities f rom His2 t o Pro6, Cys27 to Met29 and His32 to Cys34 together with large 3 J NH-Ha coupling constants indicate a b-strand conforma- tion. The last two stretches are connected by the Met29 to His32 fragment, which shows a HN-HN i,i+1 connectivity pattern typically assigned to a t ight turn. This t urn was defined as a type II b-turn. An Ha-HN i,i+1 interaction between Met29 and Asn30 was not observed, because a Ha proton of Met29 overlapped with t he water signal. A number of long-range HN-HN i,j [Thr3:Cys33, Lys27:Asn34], Ha-HN i,j [His2:Cys35, Cys28:Asn34], HN- Ha i,j [Asn4:His32, Thr3:Asn34, Lys27:Cys35, Met2 9:Cys33] and Ha-Ha i,j [His2:Asn34] NOE connectivities suggest that the structure is a triple-stranded b-sheet. Disulfide bridge pattern The disulfide bridge pattern of IsTX was established by NMR and modeling u sing data o btained from XPLOR - NIH . The disulfide bridge pattern of IsTX was expected to conserve those in a-KTX6 family. The structural calculations were considered: (a) with a standard disulfide bridge pattern (C1-C5, C2-C6, C3-C7 and C4-C8); (b) with MTX t ype disulfide pattern (C1-C5, C2-C6, C3-C 4 and C 7-C8); and (c) without disulfide bridge restraints. The total energy of the best 20 structures was compared after 1 00 str uctures were calculated for each disulfide bridge patterns. The total energy with MTX pattern restraint (b) 8969.6 kcal was higher than the standard pattern restraint (a) 6446.4 kcal. The number of N OE violations in 100 structures was seven for the standard pattern and 335 for the MTX pattern. In addition, the NOE correlations of the Hb-Hb protons between C3-C7 and C4-C8 were less well resolved, but present. The fact confirmed t hat IsTX has a s tandard disulfide bridge pattern; however, there were no NOE correlations between C1-C5 and C2-C6. Finally, calculations with a standard pattern r estraint (a) and without disulfide bridges converged to the same structure. Hydrogen bonds The amide proton exchange rates a nd the value of the amide proton temperature coefficients were used for the identification of hydrogen bonds [40]. Values of the coefficients ( Dr HN /DT) larger than )4.6 p.p.b.ÆK )1 are good indicators of hydrogen b onds. T he amide proton temperature coefficients are summarized in Fig. 8. The backbone amide protons of IsTX have highly variable temperature coefficients, ranging from )10.2 to 3.6 p.p.b.ÆK )1 . Temperature coefficient values become more positive as NHs are shifted upfield relative to their random coil value and more negative as NHs are shifted downfield [31]. In this case, as the residues (e.g. Cys23: 2.5 p.p.b.ÆK )1 and Arg25: 2.9 p.p.b.ÆK )1 ), which are close to aromatic rings and their amide protons are shifted upfield from random coil values (Cys: 8.23 p.p.m., Arg: 8.24 p.p.m), they s how positive coefficients. In o rder to determine hyd rogen bon ds with higher confidence t han using t he temperature coefficients alone, the following criteria were used to identify them: ( a) the temperature coefficient of the amide proton summarized in Fig. 8 was larger than )4.6 p.p.b.ÆK )1 ; (b) the amide proton was slowly exchanging; (c) a single potential acceptor atom was found within a radius of 2.8 A ˚ ,a60° angle with the donor atom in over 80% of the 20 structures w as calculated without hydrogen-bond restraints. Fig. 7. NMR data summary of Is TX. Secondary structure, a summary of the NOE connectivities and 3 J NHCa coupling constants are shown. The sequential NOEs, extracted from NOESY with mixing times of 200 m s and classified as very weak, weak, medium and strong, are represented by the t hickness of the bars. 3 J NHCa coupling constants are indicated by › (> 8 Hz) and fl (< 5.5 Hz). In the secondary structure, large unfilled a rrows show b-sheets and the coil i ndicates an a-helix. Fig. 8. Graph of the amide proton temper ature coefficients in IsTX. Symbols refer to the values of the temperature coefficients as follows: slowly exchangeable hydrogen bond-forming proton, j; r apidly exchangeable nonhydrogen bond-forming proton; h. The c ontinuo us line corresponds to a value of )4.6 p.p.b.ÆK )1 . 3860 N. Yamaji et al. (Eur. J. Biochem. 271) Ó FEBS 2004 Structural comparisons with HsTX1 IsTX shows high sequence similarity with HsTX1. The overall structure of IsTX i s similar to t hose of o ther members of the a/b-fold scorpion toxin family including HsTX1 (Fig. 6B,C). All of these toxins contain a single a-helix connected to an antiparallel b-sheet. The disulfide bridge pattern of IsTX, C1-C5, C2-C6, C3-C7 and C4-C8, is thesameasthatofHsTX1.IsTXhasdifferentC-terminal and N-terminal regions. The terminal regions of IsTX are longer than those of HsTX1 and there is one more b-st rand at N-terminus and two bulky residues at C-terminus, Pro40 and Trp41. Electrostatic surface potential The e lectrostatic potential distribution was calculated using MOLMOL 2 K .1 [41]. The atomic radius was a van der Waals radius, and a full Coulombic calculation was performed usingatleasta10A ˚ boundary extending beyond the longest axis of the protein. The internal protein dielectric constant was 2.0, the water solvent dielectric constant 80.0, the water radius 1.4 A ˚ , the salt concentration 150 m M ,andthesalt radius 2.0 A ˚ . Atomic charges and the protonation state of the amino acid residues were derived using the Ôpdb-chargeÕ MOLMOL macro prog ram. Discussion Most K + channel toxins were i solated from B uthidae, whereas MTX [11–13], HsTX1 [14–16] and Pi1 [17] were from Scorpionidae and IsTX was from Ischnuridae. An evolutionary correlation among these peptides h as been suggested. When comparing the amino acid sequence of IsTX with those of o ther scorpion toxins, IsTX s howed a 50% sequence s imilarity to H sTX1 and MTX and a 43% similarity to Pi1. These toxins belong to a unique subfamily (a-KTx6) of K + channel toxins. This subfamily consists of a cysteine-stabilized a/b-fold that comprises an a-helix con- nected to a double stranded b-sheet. Because of the high similarity of the sequence with these toxins and similar a/b- fold, IsTX is thus thought to belong to the same subfamily as MTX, HsTX1 and Pi1. The a-KTx6 subfamily has specific activity for the shaker related volta ge gat ed K + channels [10]. MTX shows activities on both voltage-gated K + channels and apamin- sensitive Ca 2+ -activated K + channels (Kv1.1, 37 n M ; Kv1.2, 0.8 n M ; Kv1.3, 150 n M ;ShB,2n M ;SKCa,5n M ) [42,43]. On the other hand, HsTX1 shows high activity on voltage-gated K + channels (Kv1.3, 12 p M ) but is inactive on rat brain apamin-sensitive SK chan nels. Pi1, however, is inactive on Kv1.1 a nd Kv1.3 c hannels at micromolar concentrations (Kv1.2, 0.44 n M ; ShB, 23 n M ) [44,45]. IsTX w as found to have a t ertiary structure similar to that of HsTX1. There are a number of structural similarities between IsTX and HsTX1. However, the functional prop- erties of these toxins are clearly different. HsTX1 blocks the Kv1.3 channel with higher affinity, whereas IsTX binds to this channel with lower affinity (Kv1.3, 1.59 l M ; Kv1.1, 12.7 l M ). In order to elucidate the elements critical for binding to Kv1.3 channel, it is important to compare the structural features, critical residues, sidechain locations and the electrostatic surface potential between IsTX and HsTX1 (Figs 9 and 10). Recently, the crystal structure of the full-length voltage- gated K + channel w as determined [46,47]. In order to explain the affinity and selectivity for K + channels, several binding models were proposed [15,48–51]. The mutagenesis studies of TsTX-Ka (a- KTx family) indicate that Lys27 is the most critical residue involved in blocking K + channels [52]. The affinity measurement for the K + channel using site-directed mutage nesis at L ys27 indicated that Arg substitution reduced the affinity by about 25-fold, and Ala and Glu substitutions reduced the affinity by > 1000-fold. Several reports have shown that the Lys27 residue occludes the pore of the K + channel [ 2]. Lys27 of IsTX, which inserts into the channel pore, is located at the same position as Lys23 of HsTX1. This position is considered critical for the Kv1.3 c hannel blocking activity and forms a hydrogen bond with the Tyr395 carbonyl oxygen atoms of t he human Kv1.3 channel. Furthermore, several other residues (Met29, Asn30 and Arg31) surrounding Lys27 for IsTX are also conserved and located in similar positions to those of HsTX1 (Met25, Asn26 and Arg27) surrounding Lys23 (Fig. 10A). Docking studies of HsT X1 on t he Kv1.3 channel revealed that the Lys23, Met25 and Asn26 residues, Fig. 9. Electrostatic potential surfaces of IsTX (A) and HsTX1 (B). The charge was assigned to normally ionizable residues (Asp, Glut, Lys and Arg). Negatively charged regions are shown i n red and positively charged regions in blue. The figure was generated using MOLMOL . Ó FEBS 2004 Solution structure of IsTX (Eur. J. Biochem. 271) 3861 found in m ost a-KTx6 f unctional sites, a re critical fo r binding [15]. Figure 9 shows the surface electrostatic potential distri- bution of IsTX and HsTX1. HsTX1 has large basic residues (Arg4, Lys7, Arg14, Lys23, Lys28, Lys30 and Lys33) while IsTX has fewer basic residues (Arg8, Arg25, Lys27 and Arg31). The extracellular surface of the entry pore of the Kv1.3 channel bears a large negative electrostatic potential, which is centrosymmetric around the central pore [49]. On the other hand, the surfaces of peptides have positive electro- static potentials. The positively c harged residues i n the peptides interact electrostatically with the acidic re sidues in the channel. The electrostatic potential on the surface of the peptides shows that charge anisotropy is the driving force for the association of peptides to the Kv1.3 channel. The channel has fourfold sy mmetry with the homotetramer [46]. The positively charged residues o f HsTX1, Arg4, Lys7, Lys28 and Arg33, are uniformly distributed around Lys23 on the surface. IsTX has only four positively charged residues, thus, the positive potential is biased. T hese differences are indicative of the fact that IsTX is less potent than HsTX1. Furthermore, IsTX has two bulky residues (Pro40 and Trp41) at the C-terminus. These bulky residues are unfavorable for the interaction with the Kv1.3 channel pore region. This is the first report that scorpion v enom contains sexually linked toxins and that only male scorpions have IsTX in their venom. In spider venoms, some male toxins have been reported (e.g. d-atracotoxin-Ar1a [7], d-missu- lenatoxin-Mb1a [8]). The MALDI-TOF MS clearly eluci- dated the difference between the two genders (Fig. 1). A sexually related sting in some species of scorpions was reported previously [53]. The male scorpion uses its stinger to puncture the female scorpion’s body for 3–20 min or more. It is difficult to determine whether envenomation occurs during this process. This behavior occurs early on and then sporadically later in the mating promenade. If envenomation occurs prior to mating, the sexual sting may drug the female and thus function to subdue her normal aggressive behavior. Whether the sexual sting occurs in scorpion (O. madagascariensis) and whether IsTX plays a role in this biological process is currently under investiga- tion. In summary, the three-dimensional structure of a newly discovered scorpion toxin from O. madagascariensis was determined. The sequence shows that it is a member of a-KTx6 subfamily, which is a voltage-gated potassium channel b locking peptide. IsTX was suggested to have a high affinity to the Kv1.3, as the tertiary structure is similar to HsTX1. HsTX1 has a high sequence similarity to IsTX. However, IsTX is 10 000-fold less potent than HsTX1 in binding to the Kv1.3 channel. These results suggest that the basic amino residues are crucial to binding to voltage-gated potassium channels. Furthermore, two bulky residues at the C-terminus of IsTX may prevent binding. Mutation experiments of IsTX in w hich two bulky C-terminal residues are cleaved w ould demonstrate this proposed prevention. Acknowledgements We are grateful to Drs Jan Tytgat and Hideki Nishio for binding assay and c hemical synthesis of IsTX. This work was supported by a grant from the Research f or the F uture P rogram from the Japan Society for the Promotion of Science (JSPS). References 1. Smith, C., Phillips, M. & Miller, C. (1986) Purification of char- ybdotoxin, a specific inhibitor o f the high-conductance Ca 2+ - activated K + channel. J. Biol. Chem. 261, 14607–14613. 2. Park, C S. & M iller, C. (1992) Mapping function t o structure in a channel-blockin g peptide: electrostatic mutants o f charyb do toxin. Biochemistry 31, 7749–7755. 3. Stampe, P., Kolmakova-Partensky, L. & Miller, C. (1994) Intimations o f K + channel structure from a complete functional map of the molec ular surface of charybdotoxin. Biochemistry 33, 443–450. 4. Blanc, E ., Romi-Lebrun, R., Bornet, O ., Nakajima, T. & Darbo n, H. ( 1998) Solution structure of two n ew toxins fro m the veno m of the chinese scorpion Buthus martensi karsch blockers of the potassium channels. Biochemistry 37, 12412–12418. 5. Dai, L., Yasuda, A., Naoki, H., Corzo, G., Andriantsiferana, M. & Nakajima, T. (2001) IsCT, a novel cytotoxic l inear peptide from scorpion O pisthacanthus madagascariensis. Biochem. Biophys. Res. Commun. 286, 820–825. 6. Dai, L., Corzo, G., Naoki, H., Andriantsiferana, M. & Nakajima, T. (2002) Purification, structure-function analysis, a nd molecular characterization of novel linear peptides from scorpion Opistha- canthus madagascariensis. Biochem. Biophys. Res. Commun. 293, 1514–1522. Fig. 10. Comparison of structures of IsTX and HsTX1. Selected side- chains are shown for IsTX (yellow) and HsTX1 (dark green) and labeled for IsTX (black) and HsTX1 (gree n). Conserved residues (Met29, Asn30, Arg31) surrounding Lys27 (A) and nonconserved residues and bulky C-terminal residues of IsTX (B). 3862 N. Yamaji et al. (Eur. J. Biochem. 271) Ó FEBS 2004 7. Alewood, D., Birinyi-Strachan, L.C., Pallaghy, P.K., Norton, R.S., Nicholson, G.M. & Alewood, P.F. (2003) Synthesis and characterization of d-atracotoxin-Ar1a, the lethal neuroto xin from venom of the Sydney funnel-web spider (Atrax robustus). Bio- chemistry 42, 12933–12940. 8. Gunning, S.J., Chong, Y., Khalife, A.A., Hains, P.G., Broady, K.W. & Nicholson, G.M. ( 2003) Isolation of delta-mis- sulenatoxin-Mb1a, the major vertebrate-active spider delta-toxin from the venom of Missulena bradleyi (Actinopodidae). FEBS Lett. 554, 211–218. 9. Tytgat, J., George Chandy, K., Garcia, M.L. , Gutman , G.A. , Martin-Eauclaire, M F., van der Walt, J.J. & Possani, L.D. (1999) A unified nomenclature for short-chain peptides isolated from scorpion venoms: a-KTx molecular subfamilies. Trends Pharmacol. Sci. 20, 444–447. 10. Rodriguez de la Vega, R.C., Merino, E., Becerril, B. & Possani, L.D. (2003) Novel interactions between K + channels and scor- pion toxins. Trends Pharmacol. Sci. 24, 222–227. 11. Kharrat, R., Mansuelle, P., Sampieri, F., Crest, M., Oughideni, R.,VanRietschoten,J.,Martin-Eauclaire,M F.,Rochat,H.&El Ayeb, M. (1997) Maurotoxin, a four disulfide bridge toxin from Scorpio maurus venom: purification, structure an d action on po- tassium channels. FEB S Lett. 406, 284–290. 12. Blanc, E., Sabatier, J M., Kharrat, R., Meunier, S., El Ayeb, M., Van Rietschoten, J. & Darbon, H. (1997) Solution structure of maurotoxin, a scorpion toxin from Scorpio maurus, with high affinity for voltage-gated potassium channels. Proteins 29,321– 333. 13. Fajloun, Z., Mosbah, A., Carlier, E., Mansuelle, P., Sandoz, G., Fathallah,M.,diLuccio,E.,Devaux,C.,Rochat,H.,Darbon,H., De Waard, M. & Sabatier, J.M. (2000) Maurotoxin versus Pi1/ HsTX1 toxins: toward ne w insights in the und erstanding of their distinct disulfide bridge patterns. J. Biol. Chem. 275, 39394–39402. 14. Lebrun, B., Romi-Lebrun, R., M artin-Eauclaire, M F., Yasuda, A., Ishiguro, M., Oyama, Y., Pongs, O. & Nakajima, T. (1997) A four-disulphide-bridged toxin, with high a ffinity towards voltage- gated K + channels, isolated from Heterometrus spinnifer (Scor- pionidae) venom. Biochem. J. 328, 321–327. 15. Savarin, P., Romi-Lebrun, R., Zinn-Justin, S., Lebrun, B., Nakajima, T., Gilquin, B. & Menez, A. (1999) Structural and functional consequences of the presence of a fourth disulfide bridge in the scorpion short toxins: Solution structure of the potassium channel inhibitor HsTX1. Protein Sci. 8, 2672–2685. 16. Savarin, P., Zinn-Justin, S. & Gilquin, B. (2001) Variability in automated assignment of NOESY spectra and three-dimensional structure dete rmination: a test case on three small disulfide-bon- ded proteins. J. Biomol. NMR 19, 49–62. 17. Delepierre, M., Proc hnicka-Chalufour, A. & Possani, L.D. ( 1997) A novel potassium channel blocking t oxin from the scorpion Pandinus imperator:a 1 H NMR analysis using a nano-NMR probe. Biochemistry 36, 2649–2658. 18. Tenenholz, T.C., Rogowski, R.S., Collins, J.H., Blaustein, M.P. & Weber, D.J. (1997) Solution structure for Pandinus toxin K-a (PiTX-Ka), a se lective blocker of A-type potassium channels. Biochemistry 36, 2763–2771. 19. Delepierre, M., Prochnicka-Chalufour, A., Boisbouvier, J. & Possani, L.D. (1999) Pi7, an orphan peptide from the scorpion Pandinus imperator:a 1 H-NMR analysis using a nano-NMR probe. Biochemistry 38, 16756–16765. 20. Guijarro, J.I., M’Barek, S., G omez-Lagunas, F., Garnier, D., Rochat,H.,Sabatier,J M.,Possani,L.D.&Delepierre,M.(2003) Solution structure of Pi4, a sh ort four-disulfide-bridged scorpion toxin specific of potassium channels. Prot. Sci. 12, 1844–1854. 21. Katsumi, A., Tuley, E .A., Bodo ´ ., I. & Sadler, J.E. ( 2000) Locali- zation of disulfide bonds in the cystine knot d omain of human von Willebrand Factor. J. Biol. Chem. 275, 25585–25594. 22. Rance, M.S., Sørensen, O.W., Bodenhausen, G., Wagner, G., Ernst, R.R. & Wu ¨ thrich, K. (1983) Improved spectral resolution in COSY 1 H NMR sp ectra of proteins via dou ble qu antum filtering. Biochem. Biophys. Res. Commun. 117, 479–485. 23. Kumar, A., Ernst, R.R. & Wu ¨ thrich, K. (1980) A two-dimen- sional nuclea r Ov erhaus er en hanc ement ( 2D N OE) ex perim ent for the elucidation of complete proton-proton cross-relaxation net- works in biological macromolecules. Bioc hem. Biophys. R es. Commun. 95, 1–6. 24. Bodenhausen, G., Kogler, H. & Ernst, R.R. (1984) Selection of coherence-transfer pathw ays in NMR p ulse experimen ts. J. Magn. Reson. 58, 370–388. 25. Bax, A. & Davis, D.G. (1985) MLEV-17-based two-dimensional homonuclear magnetization transfer spectroscopy. J. Magn. Reson. 65, 355–360. 26. Piotto, M., Saudek, V. & Sklenar, V. (1992) Gradient-tailored excitation for s ingle quantum spectroscopy of aqueous s olutions. J. Biomol. NMR 2, 661–665. 27. Palmer, A.G., Cavanagh, J., Wright, P.E. & Rance, M. (1991) Sensitivity improvement in proton-detected two-dimensional het- eronuclear correlation NMR spectroscopy. J. Magn. Reson. 93, 151–170. 28. Kay, L.E., Keifer, P. & Saarinen, T. (1992) Pure absorption gradient enhanced heteronuclear single quantum correlation spectroscopy with improved sensitivity. J. Am. Chem. Soc. 114, 10663–10665. 29. Wu ¨ thrich, K. (1986) NMR of Protein and Nucleic Acids.John Wiley & Sons, Inc, New York. 30. Kraulis, P.J. (1989) ANSIG: a program for the assignment of protein 1 H 2D NMR spectra by interactive computer graphics. J. Magn. Reson. 84, 627–633. 31. Andersen, N.H., Neidigh, J.W., Harris, S.M., Lee, G.M., Liu, Z. & Tong, H. (1997) Extracting information from the tem- perature gradients of polypeptide NH chemical shifts. 1. The importance of conformational averaging. J. Am. Chem. S oc. 119, 8547–8561. 32. Cierpicki, T., Zhoukov, I., Byrd, R.A. & Otlewski, J. (2002) Hydrogen bonds in human ubiquitin reflected in temperature coefficients of amide protons. J. Magn. Reson. 157, 178–180. 33. Pardi, A., Billeter, M. & C¸ thrich, K. (1984) Calibration of the angular dependence of the amide proton-C a proton coupling constants, 3 J HN a, in a globular protein. Use of 3 J HN a for i denti- fication of h elical se condary s tructu re. J. Mol. Biol. 180, 7 41–751. 34. Clore, G .M . & Gronenborn, A.M. ( 1998) New methods of structure refinement for macromolecular s tructure determination by NMR. Proc. Natl Acad. Sci. USA 95 , 5891–5898. 35. Schwieters, C.D., Kuszewski, J.J., Tjandra, N. & Clore, G.M. (2003) The Xplor-NIH NMR molecular structure determination package. J. Magn. Reson. 160, 66–74. 36. Brunger, A.T. (1992) X-PLOR, Version 3.1 Manual. Yale University Press, New Haven, CT, USA. 37. Lakowski, R.A., Rullmann, J.A., MacArthur, M.W., Kaptein, R. & Thornton, J.M. (1996) AQUA and PROCHECK-NMR: pro- grams for checking the quality of protein structu res solved by NMR. J. Biomol. NMR 8, 477–486. 38. Spera, S. & B ax, A. (1991) Empirical correlation between protein backbone conformation and Ca and Cb 13 C nuclear magnetic resonance chemical shifts. J. Am. Chem. Soc. 113, 5490–5492. 39. Kuszewski, J., Qin, J., Gronenborn, A.M. & Clore, G.M. (1995) The impact of direct refinement against 13 C a and 13 C b chemical shifts o n protein structure determination by NMR. J. Magn. Reson. Series B 106, 92–96. 40. Cierpicki, T. & Otlewski, J. (2000) Determination of a high pre- cision structure of a novel protein, Linum usitatissimum trypsin inhibitor (LUTI), using computer-aided assignment of NOESY. J. Mol. Biol. 302, 1179–1192. Ó FEBS 2004 Solution structure of IsTX (Eur. J. Biochem. 271) 3863 41.Koradi,R.,Billeter,M.&Wu ¨ thrich, K. (1996) MOLMOL: a program for display and analysis of macromolecular structures. J. Mol. Graph. 14, 51–55. 42. Kharrat, R., M abrouk , K., Crest, M., Darbon, H., Oughideni, R., Martin-Eauclair, M F., Jacquet, G., El Ayeb, M., Van Riet- schoten, J., Rochat, H. & Sabatier, J.M. (1996) Chemical synthesis and charaterization of maurotoxin, a short scorpion toxin with four disulfide bridges that acts on K + channels. Eur. J. Biochem. 242, 491–498. 43. Kharrat, R., Mansuelle, P., Sampieri, F., Crest, M., Oughideni, R., Van Rietschoten, J., Martin-Eauclair, M F., Rochat, H. & El Ayeb, M. (1997) Maurotoxin, a f our disulfide bridge toxin fro m Scorpio maurus venom: purification, structure and action o potassium channels. FEBS Lett. 406, 284–290. 44. Gomez-Langunas, F ., Olamendi-Portugal, T. & Possani, L.D. (1997) Block of ShakerB K + channels by Pi1, a novel class of scorpion toxin. FEBS Lett. 400, 197–200. 45. Fajloun, Z., Carlier, E., Lecomte, C., Geib, S., Di Luccio, E., Bichet, D., Mabrouk, K., Rochat, H., De Waard, M. & Sabatier, J.M. (2000) Chemical synthesis and characterization of Pi1, a scorpion toxin from Pandinus imperator active on K + channels. Eur. J. Biochem. 267, 5149–5155. 46. Jiang, Y., Lee, A., Chen, J., Ruta, V., Cadene, M., Chait, B.T. & Mackinnon, R. (2003) X-ray structure of a voltage-dependent K + channel. Nature 423, 33–41. 47. Jiang, Y., R uta, V., Chen, J., Lee, A. & Mackinnon, R . (2003) The principle of gating c harge movement in a voltage-de pendent K + channel. Nature 423, 42–48. 48. Mackinnon,R.,Cohen,S.L.,Kuo,A.,Lee,A.&Chait,B.T. (1998) Structural conservation in prokaryotic and eukaryotic potassium channels. Science 280, 106–109. 49. Fu,W.,Cui,M.,Briggs,J.M.,Huang,X.,Xiong,B.,Zhang,Y., Luo,X.,Shen,J.,Ji,R.,Jiang,H.&Chen,K.(2002)Brownian dynamics simulations of the recognition of the scorpion toxin maurotoxin with the voltage-gated potassium ion channels. Biophys. J. 83 , 2370–2385. 50. Zhang, M., K orolkova, Y.V., Liu, J., Jiang, M., Grishin, E.V. & Tseng, G N. (2003) BeKm-1 is a H ERG-sp ecific toxin that shares the structure with ChTx but th e mech anism of action with ErgTx1. Biopys. J. 84, 3022–3036. 51. M’Barek, S., Mosbah, A., Sandoz, G., Fajloun, Z., Olamendi- Portugal, T., Rochat, H., Sampieri, F., Guijarro, J.I., Mansuelle, P.,Delepierre,M.,DeWaard,M.&Sabatier,J M.(2003) Synthesis and characterization of Pi4, a scorpion toxin from Pandinus imperator that acts on K + channels. Eur. J. Biochem. 270, 3583–3592. 52. Ellis, K.C., Tenenholz, T.C., Jerng, H., Hayhurst, M., Dudlak, C.S., gilly, W.F., Blaustein, M.P. & Weber, D.J. (2001) Interaction of a toxin from the scorpion Tityus serrulatus with a cloned K + channel from squid (sqKv1A). Biochemistry 40, 5 942–5953. 53. Francke, O.F. (1979) Observations on the reproductive biology and life h istory of Megacormus gertschi Diaz (Scorpiones: Chactidae: Megacorminae). J. Arachnol. 7, 223–230. Supplementary material The following material is available from http://www. blackwellpublishing.com/products/journals/sup pmat/EJB/ EJB4322/EJB4322sm.htm Table S1. Proton chemical shifts of IsTX. Table S2. Ca and Cb chemical shifts of IsTX. Fig. S1. NOESY spectrum recorded w ith 50 m s mixing time at 298K. Fig. S2. Co-injection o f natural and s ynthetic IsTX in RP-HPLC. Fig. S3. Induced shifts of Ca carbons of IsTX. 3864 N. Yamaji et al. (Eur. J. Biochem. 271) Ó FEBS 2004 . Solution structure of IsTX A male scorpion toxin from Opisthacanthus madagascariensis (Ischnuridae) Nahoko Yamaji 1 , Li Dai 1 , Kenji Sugase 1 , Marta Andriantsiferana 2 , Terumi Nakajima 1 and. Nakajima 1 and Takashi Iwashita 1 1 Suntory Institute for Bioorganic Research, Mishima-Gun, Osaka, Japan; 2 Faculty of Science, University of Antananarivo, Madagascar The novel sex-specific potassium channel. obtained with a polyadenylation signal (AATAAA) in the 3¢ untranslated region and a poly (A) tail at the 3¢ end. The open reading frame of IsTX encodes a precursor of 63 residues, contain- ing a