Synthesis, 1H NMR structure, and activity of a three-disulfide-bridged maurotoxin analog designed to restore the consensus motif of scorpion toxins.

Maurotoxin (MTX) is a 34-residue toxin that has been isolated from the venom of the chactidae scorpion Scorpio maurus palmatus. The toxin displays an exceptionally wide range of pharmacological activity since it binds onto small conductance Ca(2+)-activated K(+) channels and also blocks Kv channels (Shaker, Kv1.2 and Kv1.3). MTX possesses 53-68% sequence identity with HsTx1 and Pi1, two other K(+) channel short chain scorpion toxins cross-linked by four disulfide bridges. These three toxins differ from other K(+)/Cl(-)/Na(+) channel scorpion toxins cross-linked by either three or four disulfide bridges by the presence of an extra half-cystine residue in the middle of a consensus sequence generally associated with the formation of an alpha/beta scaffold (an alpha-helix connected to an antiparallel beta-sheet by two disulfide bridges). Because MTX exhibits an uncommon disulfide bridge organization among known scorpion toxins (C1-C5, C2-C6, C3-C4, and C7-C8 instead of C1-C4, C2-C5, and C3-C6 for three-disulfide-bridged toxins or C1-C5, C2-C6, C3-C7, and C4-C8 for four-disulfide-bridged toxins), we designed and chemically synthesized an MTX analog with three instead of four disulfide bridges ([Abu(19),Abu(34)]MTX) and in which the entire consensus motif of scorpion toxins was restored by the substitution of the two half-cystines in positions 19 and 34 (corresponding to C4 and C8) by two isosteric alpha-aminobutyrate (Abu) derivatives. The three-dimensional structure of [Abu(19), Abu(34)]MTX in solution was solved by (1)H NMR. This analog adopts the alpha/beta scaffold with now conventional half-cystine pairings connecting C1-C5, C2-C6, and C3-C7 (with C4 and C8 replaced by Abu derivatives). This novel arrangement in half-cystine pairings that concerns the last disulfide bridge results mainly in a reorientation of the alpha-helix regarding the beta-sheet structure. In vivo, [Abu(19),Abu(34)]MTX remains lethal in mice as assessed by intracerebroventricular injection of the peptide (LD(50) value of 0. 25 microg/mouse). The structural variations are also accompanied by changes in the pharmacological selectivity of the peptide, suggesting that the organization pattern of disulfide bridges should affect the three-dimensional presentation of certain key residues critical to the blockage of K(+) channel subtypes.

MTX 1 is a toxin isolated from the venom of the chactidae scorpion Scorpio maurus palmatus (1). It is a basic, C-terminal amidated 34-mer peptide cross-linked by four disulfide bridges. The solid phase technique has been used to obtain synthetic maurotoxin (sMTX) and it was found that both the natural and synthetic MTXs are equally lethal to mice by intracerebroventricular inoculation (LD 50 of 80 ng/mouse). sMTX has been shown to be active in the nanomolar range on both voltagegated K ϩ channels (Shaker B, Kv1.1, Kv1.2, and Kv1.3) and onto rat brain apamin-sensitive small-conductance Ca 2ϩ -activated K ϩ channels (SK) (1). The solution structure of sMTX has been solved by 1 H nuclear magnetic resonance technique (2). The three-dimensional structure shows that the toxin contains a bent helix from residues 6 to 16 connected by a loop to a two-stranded antiparallel ␤-sheet (residues 23-26 and 28 -31), a conformation grossly similar to those of other scorpion toxins.
The half-cystine pairings of sMTX were identified by enzyme proteolysis and found to be Cys 3 -Cys 24 , Cys 9 -Cys 29 , Cys 13 -Cys 19 , and Cys 31 -Cys 34 , consistent with experimental data obtained by Edman sequencing of the natural MTX (3). The structural and pharmacological features of MTX (less than 40 residues, 4 disulfide bridges, and binding onto K ϩ channels) suggest that MTX belongs to a new class of natural K ϩ channel blockers structurally intermediate between the Na ϩ (60 -70 residues and 4 disulfide bridges) and K ϩ channel scorpion toxin families (less than 40 residues and 3 disulfide bridges) (4). This class also includes Pi1 and HsTx1 from the venoms of the scorpions Pandinus imperator (5) and Heterometrus spinnifer (6), respectively. These K ϩ channel-acting toxins share from 53 to 68% sequence identity with MTX but display slightly different pharmacological selectivities. For instance, Pi1 is inactive on rat Kv1.1 and Kv1.3 channels, whereas HsTx1 is inactive on apamin-sensitive SK channels in contrast to MTX (6). One interesting feature of MTX is that it has a unique disulfide bridge pattern among known scorpion toxins, including toxins from its own class such as Pi1 and HsTx1. In particular, it differs, first, from the classical three-disulfide-bridged toxins active on K ϩ channels by the presence of an extra disulfide bridge and, second, from all three-or four-disulfide-bridged toxins by a loss of the half-cystine pairings observed in classical three-or four-disulfide-bridged toxins (C1-C4, C2-C5, and C3-C6 pairings, or C1-C5, C2-C6, C3-C7, and C4-C8 pairings, respectively). The change concerns the two last disulfide bridges, and the pairings observed in MTX are of the type C1-C5, C2-C6, C3-C4, and C7-C8 instead.
It has been proposed that the ␣/␤ scaffold of scorpion toxins is determined by a consensus sequence of the type [ .] for either three-or four-disulfide-bridged toxins, with two additional half-cystines located outside the motif in the latter case (2,7,8). From the ␣/␤ scaffold arises a great functional diversity of these scorpion toxins. The detailed structural basis for this variability is still poorly apprehended. We noticed that this consensus sequence is, however, altered in the case of MTX since the first halfcystine residue located at the N terminus is absent, whereas a new one is inserted in the central part of the motif (position Cys 19 , Fig. 1). Despite this change in the proposed consensus sequence, it was unexpectedly found that MTX still adopts the ␣/␤ scaffold (2). However, interesting differences result from this new disulfide bridge organization. In particular, it was found that, in this novel disulfide bridge pattern, the ␣-helix is connected by two disulfide bridges (Cys 9 -Cys 29 and Cys 13 -Cys 19 ) to two different strands of the ␤-sheet instead of connecting the ␣-helix to the same strand. These novel half-cystine pairings may nevertheless result in discrete conformational changes and/or positioning of certain residues of the toxin that could impact the pharmacological selectivity of MTX. To test this hypothesis, a three-disulfide-bridged structural analog of MTX was designed to fully restore the consensus sequence of the scorpion toxins. Accordingly, Cys 19 and Cys 34 that do not belong to the consensus sequence were replaced by isosteric ␣-aminobutyrate derivatives. Here, we report the chemical synthesis of [Abu 19 ,Abu 34 ]MTX, its half-cystine pairings, its threedimensional structure in solution, and its biological and pharmacological activities. Our data indicate that there is indeed a rearrangement of the disulfide bridges in the MTX analog without a drastic alteration of the ␣/␤ scaffold, but with a marked reorientation of the ␣-helix regarding the ␤-sheet (angle of 50°) that may be sufficient to affect the pharmacological activity of the peptide.

EXPERIMENTAL PROCEDURES
Materials-N ␣ -Fluorenylmethyloxycarbonyl (Fmoc)-L-amino acids, Fmoc-amide resin, and reagents used for peptide synthesis were obtained from Perkin-Elmer. Solvents were analytical grade products from SDS. Enzymes (trypsin and chymotrypsin) were obtained from Roche Molecular Biochemicals.
Chemical Synthesis and Physicochemical Characterization of [Abu 19 ,Abu 34 ]MTX-The [Abu 19 ,Abu 34 ]MTX was obtained by the solid phase technique (9) using a peptide synthesizer (model 433A, Applied Biosystems Inc.). Peptide chains were assembled stepwise on 0.25 meq of Fmoc-amide resin (0.65 meq of amino group/g) using 1 mmol of (Fmoc) amino acid derivatives (10). The side chain-protecting groups used for trifunctional residues were: trityl for Cys, Asn, and Gln; tertbutyl for Ser, Thr, Tyr, and Asp; pentamethylchroman for Arg, and tert-butyloxycarbonyl for Lys. The Fmoc-amino acid derivatives were coupled (20 min) as their hydroxybenzotriazole active esters in Nmethylpyrrolidone (4-fold excess). The peptide resin (about 2.2 g) was treated for 2.5 h at room temperature with a mixture of trifluoroacetic acid/H 2 O/thioanisole/ethanedithiol (88:5:5:2, v/v) in the presence of crystalline phenol (2.25 g). After filtration of the mixture, the peptide was precipitated and washed by adding cold diethyloxide. The crude peptide was pelleted by centrifugation (3,000 ϫ g; 10 min) and the supernatant was discarded. The reduced peptide was then dissolved at 2 mM in 0.2 M Tris-HCl buffer, pH 8.3, and stirred under air to allow folding (48 h, room temperature). The target product, [Abu 19 ,Abu 34 ]-MTX, was purified by reversed phase high pressure liquid chromatography (HPLC) (Perkin-Elmer, C 18 Aquapore ODS 20 m, 250 ϫ 10 mm) by means of a 60-min linear gradient of 0.08% (v/v) trifluoroacetic acid and 0 -35% acetonitrile in 0.1% (v/v) trifluoroacetic acid/H 2 O at a flow rate of 5 ml/min ( ϭ 230 nm). The homogeneity and identity of [Abu 19 ,Abu 34 ]MTX was assessed by: (i) analytical C 18 reversed phase HPLC, (ii) amino acid analysis after acidolysis, and (iii) mass determination by matrix-assisted laser desorption ionization-time of flight mass spectrometry.
Assignment of Half-cystine Pairings of [Abu 19 ,Abu 34 ]MTX by Enzyme-based Cleavage and Edman Sequencing Analysis-The [Abu 19 ,Abu 34 ]MTX (800 g) was incubated with a mixture of trypsin and chymotrypsin at 10% (w/w) in 0.2 M Tris-HCl, pH 7.4, for 12 h at 37°C. The peptide fragments were then purified by reversed phase HPLC (Vydac, C 18 5-m column, 4 ϫ 150 mm) with a 60-min linear gradient of 0.08% (v/v) trifluoroacetic acid, 0 -60% acetonitrile in 0.1% (v/v) trifluoroacetic acid/H 2 O at a flow rate of 1 ml/min ( ϭ 230 nm), and freeze-dried prior to their analyses. The peptide fragments were hydrolyzed by acidolysis (6 N HCl/phenol), and their amino acid content was analyzed (Beckman, System 6300 amino acid analyzer). The peptides were further characterized by mass spectrometry analysis (RP-DE Voyager, Perseptive Biosystems), and Edman sequencing using a gas phase microsequencer (Applied Biosystems 470A). In standard HPLC conditions for analyzing phenylthiohydantoin-amino acid derivatives, diphenylthiohydantoin-cystine elutes at a retention time of 9.8 min.
Three ), pH ϭ 3 uncorrected for isotope effects. Proton two-dimensional NMR spectra was first routinely recorded at 300 K. All the data were collected on a Bruker DRX 500. Clean total correlation spectra (TOCSY) (11,12) were acquired with a spin lock of 80 ms. Phase sensitive two-dimensional nuclear Overhauser effect (NOE) spectra (NOESY) (13,14) with watergate (15) composite were acquired using the time proportional phase increment method with mixing time of 100 ms. The solvent-OH resonance was suppressed either by low power irradiation during the relaxation delay and, for NOESY spectra, during the mixing time, or using a watergate 3-9-19 pulse train (15) using a gradient at the magic angle obtained by applying simultaneously x, y, and z gradients prior to detection.
For determination of amide proton exchange rates, the peptide was lyophilized twice and solubilized in 100% D 2 O. Immediately after solubilization, a series of NOESY spectra with a mixing time of 80 ms were recorded at 283 K, the first one during a time period of 1 h (1024 complex points with 256 experiments), followed by spectra of 10 h (1024 complex points with 512 experiments).
All the data were processed using the Bruker software XWINNMR, running on a Silicon Graphic INDY R4000 workstation. The matrices were transformed with a zero filling to the next power of two in the acquisition dimension, and to 1024 points in the other. The signal was multiplied by a shifted sine-bell window, in both dimensions prior to Fourier transform, and a fifth-order polynomial base-line correction was applied. Spectra had finally a 12 ppm width with a digital resolution of 2.93 Hz/point in the 2 dimension and 5.85 Hz/point in the 1 dimension. The spectral analysis and structure calculation were performed as described previously (2).
Molecular Mechanics Calculations-Steric energy calculations were aimed at determining the most energetically favored half-cystine pairings of [Abu 19 ,Abu 34 ]MTX. These calculations were based on the threedimensional structure of MTX obtained from the Protein Data Bank (2). Half-cystines in positions 19 and 34 were substituted by Abu derivatives, and full minimization was performed for each sterically possible half-cystine pairing arrangements (12 found possible out of 15, the disulfide bridge 29 -31 being excluded from the calculations because of steric impossibility; see Table II for the various combinations used). Minimizations were achieved using the molecular modeling program Insight II (Molecular Simulations Inc.), the Discover-based minimization, and the CVFF force field. The mathematical method used for minimizations was the gradient conjugate.
Neurotoxic Activity of [Abu 19 ,Abu 34 ]MTX in Mice-The peptide was tested in vivo for toxicity by determining the LD 50 after intracerebroventricular injection into 20-g C57/BL6 mice. Groups of six mice per dose were injected with 5 l of [Abu 19 ,Abu 34 ]MTX solution containing 0.1% (w/v) bovine serum albumin and 0.9% (w/v) sodium chloride.
Binding Assay of 125 I-Apamin and Competition by [Abu 19 ,Abu 34 ]MTX onto Rat Brain Synaptosomes-Rat brain synaptosomes (P2 fraction) were prepared as described by Gray and Whittaker (16). The protein content was determined by a modified Lowry method. 125 I-Apamin (2,000 Ci/mmol) was obtained according to Seagar et al. (17). Aliquots of 50 l of 0.1 nM 125 I-apamin were added to 400 l of synaptosome suspension (0.4 mg of protein/ml). Samples were incubated for 1 h at 4°C with 50 l of one of a series of concentrations of [Abu 19 ,Abu 34 ]MTX (10 Ϫ5 to 10 Ϫ13 M) in 500 l final volume. The incubation buffer was 25 mM Tris-HCl, 10 mM KCl, pH 7.2. The samples were centrifuged, and the resulting pellets were washed three times in 1 ml of the same buffer. Bound radioactivity was determined by counting (Packard Crystal II). The values expressed are the means of triplicate experiments Ϯ S.D. Nonspecific binding, less than 10% of the total binding, was determined in the presence of an excess (10 nM) of unlabeled apamin.
Oocyte Preparation and Electrophysiological Recordings-Stages V and VI Xenopus laevis oocytes were prepared for cRNA injection and electrophysiological recordings as described (18). Briefly, oocytes were prepared by removing the follicular cell layer by enzymatic treatment with 2 mg/ml collagenase IA (Sigma) in classical Barth's medium (in mM: 88 NaCl, 1 KCl, 0.82 MgSO 4 , 0.33 Ca(NO 3 ) 2 , 0.41 CaCl 2 , 2.4 NaHCO 3 , 15 HEPES, pH 7.4 with NaOH). The plasmids were cut with SmaI (Shaker B), NotI (rat Kv1.1), XbaI (rat Kv1.2), and EcoRI (rat Kv1.3). The linearized plasmids were transcribed by means of a T7 or SP6 (mMessage mMachine kit, Ambion). The cRNA were stored frozen in H 2 O at Ϫ80°C at 1 g/l. The cells were micro-injected 2 days latter with 50 nl of cRNA (0.2 g/l Shaker B, rat Kv1.1, rat Kv1.2, or rat Kv1.3 channels). To favor channel expression, cells were incubated at 16°C into a defined nutrient oocyte medium (19) 2-6 days before current recordings. Standard two-microelectrode techniques were used at room temperature (18 -23°C) to record oocyte currents. Both current and voltage electrodes were filled with 140 mM KCl and had resistances comprising between 0.5 and 1 megaohm. The bath solution was clamped to 0 mV, which served as reference potential. Currents were recorded using a voltage-clamp amplifier (GeneClamp 500, Axon Instruments, Foster City, CA) interfaced with a 16-bit AD/DA converter (Digidata 1200A, Axon Instruments) for acquisition and voltage protocol application. Voltage pulses were delivered every 15 s from a holding potential of Ϫ80 mV. Current records were sampled at 10 kHz and low pass-filtered at 2 kHz using an eight-pole Bessel filter and stored on computer for subsequent analysis. The extracellular recording solution contained (in mM): 88 NaCl, 10 KCl, 2 MgCl 2 , 0.5 CaCl 2 , 0.5 niflumic acid, 5 HEPES, pH 7.4 (NaOH). Leak and capacitative currents were subtracted on-line by a P/4 protocol. Residual capacitative artifacts were blanked for display purposes. Toxin solutions were perfused in the recording chamber at a flow rate of 2 ml/min using a ValveBank4 apparatus (Automate Scientific Inc.). 0.1% bovine serum albumin was added to the recording and perfusion solutions to prevent [Abu 19 ,Abu 34 ]MTX loss to the plastic chamber and tubules and nonspecific binding onto the cell. Data analysis was performed using pCLAMP 6.0.3 software (Axon Instruments, Foster City, CA). The results are presented as mean Ϯ S.E. Fig. 1 illustrates the consensus sequence of the scorpion toxins independent of their length, primary structure, and pharmacological selectivity (7,8). In three-disulfide-bridged toxins, the half-cystine pairings are so far of the type C1-C4, C2-C5, and C3-C6. In short chain four-disulfide-bridged toxins active on K ϩ channels, one of the two additional half-cystine residues is located inside the consensus motif, between C3 and C4, whereas the other one is placed after the C-terminal end of the motif (after C6). Of note, the two additional half-cystine residues found in long-chain four-disulfide-bridged scorpion toxins active on Na ϩ channels are located differently, i.e. at Nand C-terminal regions, outside the motif. In the K ϩ channel toxins, the addition of two cysteines affects the disulfide bridge organization of the toxin. Two possibilities are observed so far in this toxin family: (i) toxins such as Pi1 and HsTx1 keep the general overlapping type of half-cystine pairings (see Fig. 1), whereas (ii) MTX presents partially rearranged half-cystine pairings. In particular, the two last bridges are not of an overlapping type in MTX but instead C3 is connected with the extra central half-cystine residue (Cys in position 19 in MTX amino acid sequence) and C6 with the outer C-terminal halfcystine residue (Cys in position 34). To assess the effect of the disulfide bridge rearrangement induced by the presence of these two extra half-cystine residues regarding toxin threedimensional structure and pharmacological activity, we designed and chemically synthesized an analog of MTX in which Cys 19 and Cys 34 were replaced by isosteric ␣-aminobutyrate derivatives (Abu 19 and Abu 34 , which are unable to associate covalently). In this analog, the two half-cystine residues were not engaged in the formation of the same disulfide bridge, and it can therefore be expected that there will be (i) a reduction in the number of disulfide bridges from four to three and (ii) a concomitant rearrangement in pairings of the remaining six half-cystine residues. These mutations aim to restore the entire consensus sequence and half-cystine pairings of three-disulfide-bridged scorpion toxins.

RESULTS AND DISCUSSION
Stepwise assembly of [Abu 19 ,Abu 34 ]MTX was achieved by means of Fmoc/t-butyl chemistry (10). The yield of assembly was 80%. The profiles of elution by C 18 reversed phase HPLC of the crude reduced peptide after final acidolysis are shown in Fig. 2A (left panel). The crude peptide was folded/oxidized by 48 h exposure to air ( Fig. 2A, middle panel), and purified to homogeneity by C 18 reversed phase HPLC ( Fig. 2A, right panel). The amino acid ratios of [Abu 19 ,Abu 34 ]MTX were in agreement with the deduced values (see Table I  the two half-cystines). The most stable conformation (lowest steric energy) obtained was the one that can be found in scorpion toxins cross-linked by three disulfide bridges characterized so far, i.e. C1-C4, C2-C5, and C3-C6 (Table II).
To now formerly establish the half-cystine pairings of oxidized [Abu 19 ,Abu 34 ]MTX, the analog was proteolyzed by a mixture of trypsin and chymotrypsin and the resulting peptide fragments were purified by HPLC. Amino acid analysis, mass spectrometry, and Edman sequencing techniques were used to identify the sequence of the peptides connected to each other by a disulfide bond. The results of the enzyme treatment are summarized in Fig. 2B. The half-cystine pairings were thereby mapped as Cys 3 -Cys 24 , Cys 9 -Cys 29 , and Cys 13 -Cys 31 (Fig. 2C). These findings demonstrate that, as expected by introducing the two point mutations in positions 19 and 34 of MTX, the standard overlapping disulfide bridge pattern found in natural three-disulfide-bridged scorpion toxins was restored. In the [Abu 19 ,Abu 34 ]MTX analog, the two first disulfide bridges (Cys 3 -Cys 24 and Cys 9 -Cys 29 ) are identical to those found in MTX, whereas the last one is a recombination between Cys 13 and Cys 31 that initially belonged to two different disulfide bridges. Interestingly, all three disulfide bridges formed in [Abu 19 ,Abu 34 ]MTX are in an identical pattern to the three first disulfide bridges of both Pi1 and HsTx1, confirming that the disulfide bridge pattern adopted by [Abu 19 ,Abu 34 ]MTX was indeed energetically favored as suggested by the molecular mechanics calculations. The formation of a disulfide bridge between Cys 13 and Cys 31 may appear as a surprising result since, in the solution structure of MTX, the sulfurs of Cys 13 and Cys 31 are distant by 5.4 Å. The fact that, in the [Abu 19 ,Abu 34 ]MTX variant, these sulfurs approach within disulfide bonding distance (2 Å) may be due to the substitutions of the half-cystines 19 and 34 by the isosteric Abu derivatives themselves. Indeed, the potential differences in properties (hydrophobicity) between the side chains of Abu and Cys residues are likely to be the real cause of this structural rearrangement. Alternatively, it cannot be ruled out that disulfide bridges may not just simply stabilize pre-folded structures, as generally admitted, but could also affect the final three-dimensional structure of the molecule to a small extent, a possibility that is, however, difficult to address experimentally. Next, we determined by 1 H NMR whether the disulfide bridge rearrangement observed in [Abu 19 ,Abu 34 ]MTX induces significant changes in the three-dimensional structure of the peptide. Spin system assignment was carried out according to the 2-steps method. First, the spin systems were identified by their scalar connectivities. Cross-peaks between HN and H␣ were identified by examination of the DQF-COSY spectrum (Fig. 3A). Spin systems were identified on the basis of DQF-COSY, and the TOCSY spectrum was used to correlate these side-chain spin systems with the HN-H␣ cross-peaks. In a second assignment step, the spin systems were connected in sequence by the virtue of H␣/HN, HN/HN, and H␤/HN connectivities. These together with medium range connectivities are summarized in Fig. 3B. The calculated structures of [Abu 19 ,Abu 34 ]MTX converged into a single solution. The root mean square deviation value (1.1 Å for backbone atoms and 2.2 Å for all heavy atoms) of the 25 structures demonstrates the presence of the conventional ␣/␤ scaffold: a two-stranded ␤-sheet connected to an ␣-helix by two disulfide bridges (Fig. 4). The precise location of these secondary structures has been obtained by PROCHECK-NMR, which indicates the presence of an ␣-helix running from residues 9 to 17 (two turns and a half-helix) and a ␤-sheet made of two stretches from residues 22 to 25 and 28 to 31. The second and the third strands are connected by an undefined ␤-turn formed by the residues at positions 26 and 27. The main difference of the three-disulfidebridged [Abu 19 ,Abu 34 ]MTX as compared with the four-disulfide-bridged MTX is the angle between the axis of the ␣-helix  and the axis of the ␤-sheet. This angle is 50°in MTX, whereas it is close to 0°in [Abu 19 ,Abu 34 ]MTX. This structural difference could be explained by the fact that in MTX, the ␣-helix is connected by two disulfide bridges (Cys 9 -Cys 29 and Cys 13 -Cys 19 ) to each strand of the ␤-sheet, which forces the ␣-helix to orient with an angle of 50°as compared with the ␤-sheet. In contrast, in [Abu 19 ,Abu 34 ]MTX, the two disulfide bridges (conserved Cys 9 -Cys 29 and novel Cys 13 -Cys 31 ) connect the ␣-helix to the same strand (running from residues 28 to 31) of the ␤-sheet. This novel pattern of bridging now constrains the two secondary structures to the same axis orientation (Fig. 4). The relative orientation of the ␣-helix with regard to the main axis of the ␤-sheet measured on all available three-disulfide-bridged scorpion toxins from the Protein Data Bank files ranges from 0 to 70°independently of the pharmacological selectivity and sensitivity of these toxins. Therefore, this structural characteristic cannot be used to anticipate the pharmacological profile of any given toxin.
We next tested whether the structural rearrangement induced by the novel disulfide bridge pattern could result in changes in the pharmacological activity of the synthetic peptide. Intracerebroventricular injections of [Abu 19 ,Abu 34 ]MTX still produced a lethal effect in mice with an LD 50 value of 0.25 g/mouse. This effect is thus 3 times less potent than the one observed with the four-disulfide-bridged MTX (1). However, the symptoms induced by the injection of the toxin analog closely resemble those of MTX itself (and other K ϩ channel scorpion toxins), suggesting that K ϩ channels are still likely the molecular targets of [Abu 19 ,Abu 34 ]MTX. Because of the high relatedness of MTX and [Abu 19 ,Abu 34 ]MTX primary structures, we compared their effects on the binding of 125 I-apamin onto rat brain synaptosomes and on the K ϩ currents induced by the expression of Shaker B, rat Kv1.1, Kv1.2, and Kv1.3 into Xenopus oocytes. We first tested the ability of [Abu 19 ,Abu 34 ]MTX to compete with 125 I-apamin for binding onto rat brain synaptosomes. Fig. 5 shows an [Abu 19 ,Abu 34 ]MTX-induced, concentration-dependent, inhibition of 125 I-apamin binding with a half-effect of 100 nM. In comparison, unlabeled apamin produced a complete inhibition (IC 100 ) of 125 I-apamin binding at a much lower concentration (100 pM). The disulfide bridge rearrangement produced by the Cys 19 and Cys 34 substitutions decreased the affinity of the peptide for SK-type channels by approximately 9-fold since an IC 50 of 11 nM was found for sMTX, in agreement with previous data (1)  three-disulfide-bridged scorpion toxins. For instance, SK channel-acting toxins compete with 125 I-apamin for binding onto rat brain synaptosomes at concentrations ranging from 20 pM (P05) to over 1 M (P01). Similarly, for Shaker B current inhibition, IC 50 values reported for other toxins range from 0.16 nM (agitoxin) to 160 nM (noxiustoxin) (20). The only exception to the rule appears to be Kv1.3, which is much less sensitive to [Abu 19 ,Abu 34 ]MTX than to other three-disulfide-bridged toxins. The IC 50 values so far reported range from 4 pM (agitoxin 2) to 1.7 nM (agitoxin 1), which is at least 2 orders of magnitude less than [Abu 19 ,Abu 34 ]MTX. The wide range of toxin IC 50 s reported for any given channel and the difference in affinity observed between three-disulfide-bridged toxins and [Abu 19 ,Abu 34 ]MTX for Kv1.3 suggest that the toxin primary structure is crucial to channel sensitivity and does not solely rely on the disulfide bridge framework.
Overall, it can be concluded that changes in the disulfide bridge organization do not significantly alter the pharmacological targets of the ligand but can be employed as an unique mean to change its selectivity. Unexpectedly, the conformational alterations in toxin structure that accompany the disulfide bridge reorganization do not systematically translate into decreased affinities of the peptide for its K ϩ channel target(s). These data suggest that the disulfide bridge reorganization can be used as a novel approach to tentatively improve the inter-action surface of the ligand for its receptor site. For the channels in which reductions in ligand affinities were observed (i.e. Kv1.2), there must be a three-dimensional repositioning of key residues critical to the blockage of current. This could be associated with the novel orientation observed between the ␣-helix and the ␤-sheet structures, suggesting that residues belonging to both the secondary structures could be involved in the recognition of the binding site. In contrast, for channels in which no or little change in affinity was observed (i.e. Shaker B), the key residues implicated in current blockade may exclusively involve amino acid residues of the ␤-sheet structure, shown to be involved in Kv-type recognition (21). These conclusions should be reinforced by a structure-activity relationship study based on an alanine-scanning approach that ultimately will reveal the intimate differences in the binding surface of MTX for interaction with Shaker B and rat Kv1.2 and Kv1.3 channels. A combination of amino acid substitution and disulfide bridge reorganization used in parallel may reveal itself a powerful approach to the specific increase in toxin selectivity.