Current views on scorpion toxins specific for K-channels

Much of our knowledge on K⁺-channels was elucidated using specific peptide ligands isolated from a number of venomous organisms. Recently, this field received a strong support and increased interest due to the solution of the three-dimensional structure of a couple of K⁺-channels. At the same time, several new subfamilies of specific toxins for K⁺-channels were isolated from scorpion venoms, enhancing the availability and diversity of such useful molecular tools. It opened new lines of research for the better understanding of K⁺-channel biophysics and pharmacology. In this review, we listed 120 amino acid sequences of peptides isolated from scorpion venoms. They were demonstrated or assumed to be specific for K⁺-channels. These sequences were aligned and used to generate a rooted phylogenetic tree. The evolutionary tree indicates that several clusters of divergent peptides show preference for specific subtypes of channels. The three-dimensional structures of representative examples of these peptides were drawn and analysed concerning the molecular fitness of their interactions with the channel targets. Four different interacting modes were identified to exist between scorpion toxins and the various subtypes of K⁺-channels.     

The effect of the venom of the yellow Iranian scorpion Odontobuthus doriae on skeletal muscle preparations in vitro

The yellow Iranian scorpion Odontobuthus doriae can cause fatal envenoming, but its mechanism of action is unclear. One of the reported manifestations of envenoming is moderate to severe involuntary tremor of skeletal muscle. In order to understand better the mechanism of action of this venom on skeletal muscle function, we examined the effects of the venom in vitro on chick biventer cervicis (CBC) and mouse hemidiaphragm (MHD) nerve muscle preparations. O. doriae venom (0.3–10 μg/ml) initially increased and then decreased twitch height. The venom also caused contracture in both preparations. In mouse triangularis sterni preparations, used for all intracellular recording techniques, the venom enhanced the release of acetylcholine and induced repetitive firing of nerve action potentials and endplate potentials in response to single-shock stimulation. With extracellular recording techniques, scorpion venom (1 μg/ml) was found to cause changes to the perineural waveform associated with nerve terminal action potentials consistent with effects on Na⁺ and K⁺ currents. The main facilitatory effects of O. doriae venom are likely to be due to toxins that affect Na⁺ channels in nerve–muscle preparations similar to most Old World scorpion venoms, but blocking effects on K⁺ channels are also possible. Such effects could lead to initial enhancement of transmitter release that could underlie the muscle tremors seen in victims. Toxins acting on Na⁺ and K+ currents have been isolated from the venom [Jalali, A., Bosmans, F., Amininasab, M., Clynen, E., Cuypers, E., Zaremirakabadi, A., Sarbolouki, M.N., Schoofs, L., Vatanpour, H., Tytgat, J., 2005. OD1, the first toxin isolated from the venom of the scorpion Odontobuthus doriae active on voltage-gated Na⁺ channels. FEBS Lett. 579, 4181–4186; Abdel-Mottaleb, Y., Clynen, E., Jalali, A., Bosmans, F., Vatanpour, H., Schoofs, L., Tytgat, J., 2006. The first potassium channel toxin from the venom of the Iranian scorpion Odontobuthus doriae. FEBS Lett. 580, 6254–6258]; however, the muscle paralysis seen at higher concentrations of venom may be due to additional, as yet uncharacterised, components of the venom.     

Toxin determinants required for interaction with voltage-gated K channels

Ion channel-acting toxins are mainly short peptides generally present in minute amounts in the venoms of diverse animal species such as scorpions, snakes, spiders, marine cone snails and sea anemones. Interestingly, these peptides have evolved over time on the basis of clearly distinct architectural motifs present throughout the animal kingdom, but display convergent molecular determinants and functional homologies. As a consequence of this conservation of some key determinants, it has also been evidenced that toxin targets display some common evolutionary origins. Indeed, these peptides often target ion channels and ligand-gated receptors, though other interacting molecules such as enzymes have been further evidenced. In this review, we provide an overview of some selected peptides from various animal species that act on specific K⁺ conducting voltage-gated ion channels. In particular, we emphasize our global analysis on the structural determinants of these molecules that are required for the recognition of a particular ion channel pore structure, a property that should be correlated to the blocking efficacy of the K⁺ efflux out of the cell during channel opening. A better understanding of these molecular determinants is valuable to better specify and derive useful peptide pharmacological properties.     

The Protective Action of Tetrodotoxin and (±)-Kavain on Anaerobic Glycolysis, ATP Content and Intracellular Na and Ca of Anoxic Brain Vesicles

Because recent reports point to Na⁺ channel blockers as protective agents directed against anoxia-induced neuronal damage including protection of anaerobic glycolysis, the influences of tetrodotoxin (TTX) and (±)-kavain on anoxic rat brain vesicles were investigated with respect to lactate synthesis, vesicular ATP content and cytosolic free Na⁺ and Ca²⁺ ([Na⁺]_i, [Ca²⁺]_i), both of the latter determined fluorometrically employing SBFI and FURA-2, respectively. After anoxia, basal lactate production was increased from 2.9 to 9.8 nmol lactate/min/mg protein. Although lactate synthesis seemed to be stable for at least 45 min of anoxia, as deduced from the linearity of lactate production, the ATP content declined continuously with a half life (τ ) af 14.5 min, indicating that anaerobic glycolysis was insufficient to cover the energy demand of anoxic vesicles. Correspondingly, [Na⁺]_i and [Ca²⁺]_i increased persistently after anoxia by 22.1 mmol/l Na⁺ and 274.9 nmol/l Ca²⁺, determined 6.3 min after onset. An additional stimulation of vesicles with veratridine accelerated the drop of ATP (τ = 5.1 min) and provoked a massive Na⁺ overload, which levelled off to 119 mmol/l Na⁺ within a few minutes. Concomitantly, [Ca²⁺]_i increased linearly with a rate of 355 nmol Ca²⁺/l/min. Despite the massive perturbation of ion homeostasis, lactate production was unaffected during the first 8 min of veratridine stimulation. However, complete inhibition of lactate synthesis took place 30 min after veratridine was added. The Na⁺ channel blockers TTX and (±)-kavain, if applied before anoxia, preserved vesicular ATP content, diminished anoxia-induced increases in [Na⁺]_i and [Ca²⁺]_i and prevented both the veratridine-induced increases of [Na⁺]_i and [Ca²⁺]_i and the inhibition of lactate production. The data indicate a considerable Na⁺ influx via voltage-dependent Na⁺ channels during anoxia, which speeds up the decline in ATP and provokes an increase in [Ca²⁺]_i. A massive Na⁺ and Ca²⁺ overload induced by veratridine failed to influence lactate synthesis directly, but initiated its inhibition. © 1997 Elsevier Science Ltd. All rights reserved.     

Pharmacological properties of contraction caused by sodium removal in muscle strips isolated from canine coronary artery

Contractions produced by Na⁺ removal were studied in muscle strips isolated from canine coronary artery. In the presence of 20 mM K⁺ and 0.5 mM Ca⁺, rapid contractions were observed repeatedly on complete replacement of NaCl with sucrose. This contraction in the absence of Na⁺ (0-Na) was not affected by phentolamine but was strongly inhibited by verapamil. Ouabain slowly potentiated the 0-Na contraction and markedly reduced the inhibition due to verapamil. The 0-Na contraction was dependent on external Ca⁺ both with and without ouabain. Bepridil had effects very similar to those of verapamil. Amiloride and excess Mg²⁺ reduced the 0-Na contraction and the degree of their inhibition was similar after ouabain treatment. The decrease in verapamil susceptibility could suggest that the 0-Na contraction has verapamil-sensitive and -insensitive components. The former is probably due to Ca²⁺ influx through voltage-dependent channels and the latter to Ca²⁺ influx through an Na⁺-Ca²⁺ exchange process. Ouabain is considered to increase the contribution of Na⁺-Ca²⁺ exchange to the 0-Na contraction. Mg²⁺ may inhibit both verapamil-sensitive and -insensitive pathways. Amiloride probably exerts its inhibitory effect on the contractile machinery.     

Effects of [Na] and [Ca] on the responses to milrinone in rat cardiac preparations

In the present investigation we have studied the influence of changing the [Ca²⁺] and [Na⁺] on the cardiac responses to milrinone in various preparations of rat heart. Milrinone (5 × 10⁻⁵ to 8 × 10⁻⁴ M) produced a dose-dependent positive chronotropic effect on right atrium and a positive inotropic effect on left atrium and papillary muscle of the rat. A decrease in [Ca²⁺] (from 2.2 to 1.1 mM) or an increase in [Na⁺] (from 120 to 60 mM) increased the milrinone-induced inotropic effect in left atrium and papillary muscle. However, in right atrium the chronotropic effect of milrinone was significantly decreased under these conditions. Opposite changes to milrinone-induced responses were observed when [Ca²⁺] was increased (to 3.3 mM) or when the [Na⁺] was decreased to 60 mM. Nifedipine (3 × 10⁻³ M), a selective Ca²⁺ channel blocker, significantly inhibited the chronotropic response to milrinone in right atrium. However, the inotropic response to milrinone was found to be significantly greater in the presence of nifedipine. A veratridine-induced positive inotropic effect in the left atrium was also significantly increased in the presence of nifedipine. Tetrodotoxin (TTX, 1 × 10⁻⁶ M), a fast sodium channel blocker, significantly reduced the inotropic response to milrinone in left atrium and papillary muscle. A milrinone-induced dose-dependent increase in the baseline tension was observed in the right atrium which was abolished in low [Ca²⁺] and significantly increased in high [Ca²⁺]. Our data suggest the possibility that milrinone increases Ca²⁺ influx in the right atrium to cause the chronotropic effect. Milrinone also may possess an action like veratridine, involving an increased influx of Na⁺ through fast Na⁺ channels in left atrium and papillary muscle, and this action is possibly involved in the positive inotropic effect.     

Identification of a mammalian target of κM-conotoxin RIIIK

Despite the great variability of the conus peptides characterized until now only relatively few have been identified that interact with K⁺ channels. κM-conotoxin RIIIK (κM-RIIIK) is a 24 amino acid peptide from Conus radiatus, which is structurally similar to μ-conotoxin GIIIA, a peptide known to block specifically skeletal muscle Na⁺ channels. Recently, it has been shown that κM-RIIIK does not interact with Na⁺ channels, but inhibits Shaker potassium channels expressed in Xenopus oocytes. It was demonstrated that κM-RIIIK binds to the pore region of Shaker channels and a teleost homologue of the Shaker channel TSha1 was identified as a high affinity target of the toxin. In contrast the mammalian Shaker-homologues Kv1.1, Kv1.3, Kv1.4 are not affected by the toxin. In this study the activity of κM-RIIIK on other mammalian Kv1 K⁺ channels expressed in Xenopus oocytes was investigated. We demonstrate that κM-conotoxin RIIIK up to 5 μM exhibits no significant effect on Kv1.5 and Kv1.6 mediated currents, but the human Kv1.2 K⁺ channel is blocked by this peptide. The binding of κM-RIIIK to Kv1.2 channels is state dependent with an IC₅₀ for the closed state of about 200 nM and for the open state of about 400 nM at a test potential of 0 mV. κM-conotoxin RIIIK is the first conotoxin described to block human Kv1.2 potassium channels.     

A novel scorpion toxin blocking small conductance Ca activated K channel

Small conductance calcium activated potassium channels (SK) are crucial in the regulation of cell firing frequency in the nervous system and other tissues. In the present work, a novel SK channel blocker, designated BmSKTx1, was purified from the scorpion Buthus martensi Karsh venom. The sequence of the N-terminal 22 amino acid residues was determined by Edman degradation. Using this sequence information, the full-length cDNA and genomic gene of BmSKTx1 were cloned and sequenced. By these analyses, BmSKTx1 was found to be a peptide composed of 31 amino acid residues with three disulfide bonds. It shared little sequence homology with other known scorpion -KTxs but showed close relationship with SK channel blockers in the phylogenetic tree. According to the previous nomenclature, BmSKTx1 was classified as -KTx14.1. We examined the effects of BmSKTx1 on different ion channels of rat adrenal chromaffin cells (RACC) and locust dorsal unpaired median (DUM) neurons. BmSKTx1 selectively inhibited apamin-sensitive SK currents in RACC with K_d of 0.72 μM and Hill coefficient of 2.2. And it had no effect on Na⁺, Ca²⁺, Kv, and BK currents in DUM neuron, indicating that BmSKTx1 was a selective SK toxin.     

New analysis of the toxic compounds from the Androctonus mauretanicus mauretanicus scorpion venom

Scorpion venoms are very complex mixtures of molecules, most of which are peptides displaying different kinds of biological activity. Indeed, these peptides specifically bind to a variety of pharmacological targets, in particular ionic channels located in prey tissues, resulting in neurotoxic effects. Toxins modulating Na+, K+, Ca2+ and Cl(-) currents have been described in scorpion venoms. In this work, we have used several specific antibodies raised against the most lethal scorpion toxins already described to screen the Moroccan scorpion Androctonus mauretanicus mauretanicus venom in order to characterize new compounds. This immunological screening was also implemented by toxicity tests in mice and with mass spectrometry study, providing new informations on the molecular composition of this venom. In fine, we were able to determine the molecular masses of 70-80 different compounds. According to the immunological data obtained, many toxins cross-react with three sera raised against the most lethal alpha-toxins found in North African scorpion venoms, but not at all with those raised against the main beta-toxins from South and North American venoms. Some of the previously described toxins from Androctonus mauretanicus mauretanicus venom could thus be detected by combining immunological tests, toxicity in mice and molecular masses. Among these toxins, one of them, which showed a mild cross-reaction with the serum raised against AaH I (a highly potent toxin from the venom of Androctonus australis), was identified as Amm III and fully sequenced.     

The facilitatory actions of snake venom phospholipase A(2) neurotoxins at the neuromuscular junction are not mediated through voltage-gated K(+) channels.

Electrophysiological investigations have previously suggested that phospholipase A₂ (PLA₂) neurotoxins from snake venoms increase the release of acetylcholine (Ach) at the neuromuscular junction by blocking voltage-gated K⁺ channels in motor nerve terminals.

We have tested some of the most potent presynaptically-acting neurotoxins from snake venoms, namely β-bungarotoxin (BuTx), taipoxin, notexin, crotoxin, ammodytoxin C and A (Amotx C & A), for effects on several types of cloned voltage-gated K⁺ channels (mKv1.1, rKv1.2, mKv1.3, hKv1.5 and mKv3.1) stably expressed in mammalian cell lines. By use of the whole-cell configuration of the patch clamp recording technique and concentrations of toxins greater than those required to affect acetylcholine release, these neurotoxins have been shown not to block any of these voltage-gated K⁺ channels. In addition, internal perfusion of the neurotoxins (100 μg/ml) into mouse B82 fibroblast cells that expressed rKv1.2 channels also did not substantially depress K⁺ currents. The results of this study suggest that the mechanism by which these neurotoxins increase the release of acetylcholine at the neuromuscular junction is not related to the direct blockage of voltage-activated Kv1.1, Kv1.2, Kv1.3, Kv1.5 and Kv3.1 K⁺ channels.     

Binding domain of oligomycin on Na(+),K(+)-ATPase

Oligomycin inhibits Na⁺,K⁺-ATPase activity by stabilizing the Na⁺ occlusion but not the K⁺ occlusion. To locate the binding domain of oligomycin on Na⁺,K⁺-ATPase, the tryptic-digestion profile of Na⁺,K⁺-ATPase was compared with the profile of Na⁺ occlusion within the digested Na⁺,K⁺-ATPase in the presence of oligomycin. The Na⁺ occlusion profile is responsible for the digestion profile of the -subunit, which is the catalytic subunit of the ATPase. The effect of oligomycin on chimeric Ca²⁺-ATPase activity was examined. The chimera used, in which the 163 N-terminal amino acids of chicken sarcoplasmic/endoplasmic reticulum Ca²⁺-ATPase 1 were replaced with the 200 N-terminal amino acids of the chicken Na⁺,K⁺-ATPase 1-subunit, partially retains the Na⁺-dependent characteristics of Na⁺,K⁺-ATPase, because the chimeric Ca²⁺-ATPase activity is activated by Na⁺ but inhibited by ouabain, a specific inhibitor of Na⁺,K⁺-ATPase (Ishii, T., Lemas, M.V., Takeyasu, K., 1994, Proc. Natl. Acad. Sci. U. S. A., 91, 6103–6107). Oligomycin depressed the activation by Na⁺ of the chimeric Ca²⁺-ATPase activity. These findings suggest that the 200 N-terminal amino acids of the Na⁺,K⁺-ATPase -subunit include a binding domain for oligomycin.     

Isolation of a toxin from Centruroides infamatus infamatus Koch scorpion venom that modifies Na permeability on chick dorsal root ganglion cells

A novel toxin was isolated and characterized from the venom of the Mexican scorpion Centruroides infamatus infamatus. It has an apparent mol. wt of 7600, compatible with the presence of 66 amino acid residues per molecule. The N-terminal amino acid sequence was determined (up to residue 48) and showed approximately 95% similarity with toxins from other Mexican scorpions of the genus Centruroides. Experiments conducted with chick dorsal root ganglion cells showed that toxin 1 is a Na⁺ channel effector, causing a decrease in the peak Na⁺ permeability, similar to decreases observed for typical β-scorpion toxins.     

Na,K-ATPase activity is selectively increased in thalamus in thiamine deficiency prior to the appearance of neurological symptoms

The relationship between progression of neurological status and the activities of both Na⁺,K⁺- and Mg²⁺-dependent-ATPase (adenosine 5′-triphosphate phosphohydrolase) was investigated in brain regions of pyrithiamine-induced thiamine deficient rats. Thalamic Na⁺,K⁺-ATPase activity was selectively increased by 200% (P < 0.01) prior to the appearance of symptoms of thiamine deficiency and normalized in symptomatic rats. This selective transitory activation precludes a mediation by brain soluble fraction Na⁺,K⁺-ATPase modifiers as does the unaltered distribution in regional high-affinity [³H]ouabain binding densities observed throughout the time-course used in these experiments. Na⁺,K⁺-ATPase maintains cellular ionic gradients and has been implicated in neurotransmitter uptake and release mechanisms. The fact that the increased thalamic Na⁺,K⁺-ATPase activity coincides with the early alterations in serotonin metabolism observed in similarly treated animals and the concomitantly early increase in glucose utilization previously observed in the thalamus of thiamine-deficient rats is discussed.     

Inhibition of aftercontractions and phasic calcium release by yohimbine in ferret papillary muscle

Phasic release of calcium from the sarcoplasmic reticulum occurs in all mammalian cardiac preparations when the intracellular calcium concentration is sufficiently high. The phasic calcium release is often sufficient to trigger electrophysiological responses and aftercontractions. These can be detrimental to normal cardiac function. We induced phasic calcium release in ferret papillary muscles loaded with the calcium indicator aequorin. Development of phasic calcium release was associated with an increase in resting and peak [Ca²⁺]_i. Inhibiting sodium channels with yohimbine reduced resting [Ca²⁺]_i and prevented phasic calcium release. We propose a mechasism where by reduced [Na⁺]_i, and the subsequent increased efflux of calcium via sodium/calcium exchange reduced [Ca²⁺]_i.     

Definition of the alpha-KTx15 subfamily

Three novel scorpion toxins, Aa1 from Androctonus australis, BmTX3 from Buthus martensi and AmmTX3 from Androctonus mauretanicus were shown able to selectively block A-type K⁺ currents in cerebellum granular cells or cultured striatum neurons from rat brain. In electrophysiology experiments, the transient A-current completely disappeared when 1 μM of the toxins was applied to the external solution whereas the sustained K⁺ current was unaffected.

The three toxins shared high sequence homologies (more than 94%) and constituted a new ‘short-chain’ scorpion toxin subfamily: -KTx₁₅. Monoiododerivative of ¹²⁵I-sBmTX3 specifically bound to rat brain synaptosomes. Under equilibrium binding conditions, maximum binding was 14 fmol/mg of protein and the dissociation constant (K_d) was 0.21 nM. This K_d value was confirmed by kinetic experiments (k_on=6.0×10⁶ M⁻¹ s⁻¹ and k_off=6.0×10⁻⁴ s⁻¹). Competitions with AmmTX3 and Aa1 with ¹²⁵I-sBmTX3 bound to its receptor on rat brain synaptosomes showed that they fully inhibited the ¹²⁵I-sBmTX3 binding (K_i values of 20 and 44 pM, respectively), demonstrating unambiguously that the three molecules shared the same target in rat brain. A panel of toxins described as specific ligands for different K⁺, Na⁺ and Ca²⁺ channels were not able to displace ¹²⁵I-sBmTX3 from its binding site. Thus, ¹²⁵I-sBmTX3 is a new ligand for a still unidentified target in rat brain. In autoradiography, the distribution of ¹²⁵I-sBmTX3 binding sites in the adult rat brain indicated a high density of ¹²⁵I-sBmTX3 receptors in the striatum, hippocampus, superior colliculus, and cerebellum.     

Novel interactions between K+ channels and scorpion toxins

K(+) channels are macromolecules embedded in biological membranes, where they play a key role in cellular excitability and signal transduction pathways. Knowledge of their structure should help improve our understanding of their function and lead to the design of therapeutic compounds. Most pharmacological and structural characteristics of these channels have been elucidated by using high-affinity channel blockers isolated from scorpion venoms. Recent data on the three-dimensional structures of K(+) channels and novel scorpion toxins suggest a variety of novel interacting modes of these channels and toxins, which should help increase our understanding of the K(+) channel structure-function relationship.     

Action of the Na ionophore monensin on vascular smooth muscle of guinea-pig aorta

The effects of the Na⁺ ionophore monensin on contractile responses were investigated in guinea-pig aorta in normal and high K⁺ solutions. In normal K⁺ (5.4 mM) solution, monensin (2 × 10⁻⁵ M) produced a rapid increase in tension followed by slow relaxation. This contraction was markedly inhibited by phentolamine (10⁻⁵ M) or prazosin (10⁻⁶ M) and was accompanied by an increase in tritium efflux from tissue preloaded with [³H]norepinephrine. In the presence of phentolamine, monensin (1–2 × 10⁻⁵ M) or ouabain (1−2 × 10⁻⁵ M) caused only a small and slowly developing contraction. Simultaneous application of these agents caused a more rapid and greater contraction. Either monensin or ouabain gradually increased cellular Na⁺ and decreased cellular K⁺ content. When monensin was applied simultaneously with ouabain, there was a rapid increase in cellular Na⁺ and loss of cellular K⁺. In high K⁺ (65.4 mM) solution, monensin (10⁻⁶ M) slightly reduced the increased tension level but when external glucose was omitted monensin markedly inhibited the contraction. A significant decrease in tissue ATP content was observed only when monensin was applied in glucose-free solution. Similarly, hypoxia (N₂ bubbling) markedly inhibited the high K⁺ contraction and decreased the tissue ATP content only in the absence of glucose. These results suggest that monensin produces a neurogenic contraction due to the release of endogenous catecholamines and also produces a myogenic contraction by a decrease in transmembrane Na⁺ and K⁺ gradients when the Na⁺ and -K⁺ pump is inhibited by ouabain, and that monensin inhibits aerobic energy metabolism of vascular smooth muscle.