The Tarantula Toxin Psalmotoxin 1 Inhibits Acid-sensing Ion Channel (ASIC) 1a by Increasing Its Apparent H+ Affinity
http://www.100md.com
《普通生理学杂志》
1 Department of Physiology II, University of Tübingen, 72076 Tübingen, Germany
2 Interfakultres Institut für Biochemie, University of Tübingen, 72074 Tübingen, Germany
3 Department of Physiology II, University of Würzburg, 97070 Würzburg, Germany
Acid-sensing ion channels (ASICs) are ion channels activated by extracellular protons. They are involved in higher brain functions and perception of pain, taste, and mechanical stimuli. Homomeric ASIC1a is potently inhibited by the tarantula toxin psalmotoxin 1. The mechanism of this inhibition is unknown. Here we show that psalmotoxin 1 inhibits ASIC1a by a unique mechanism: the toxin increases the apparent affinity for H+ of ASIC1a. Since ASIC1a is activated by H+ concentrations that are only slightly larger than the resting H+ concentration, this increase in H+ affinity is sufficient to shift ASIC1a channels into the desensitized state. As activation of ASIC1a has recently been linked to neurodegeneration associated with stroke, our results suggest chronic desensitization of ASIC1a by a slight increase of its H+ affinity as a possible way of therapeutic intervention in stroke.
Abbreviations used in this paper: ASIC, acid-sensing ion channel; PcTx1, psalmotoxin 1.
INTRODUCTION
H+ is chemically the simplest transmitter. H+ receptor channels have recently been characterized that open upon an increase in the extracellular concentration of H+ and desensitize in the continuous presence of H+ (Waldmann et al., 1997; Waldmann and Lazdunski, 1998). The apparent affinity of these ion channels for H+, 0.3 μM, and their activation and desensitization kinetics is comparable with other ligand-gated channels (Waldmann and Lazdunski, 1998; Bssler et al., 2001), establishing H+ as the genuine ligand for these channels, the acid-sensing ion channels (ASICs). However, in contrast to other ligands, resting concentrations of H+ are close to the threshold concentration needed to activate ASICs. Consequently, slightly increased H+ concentrations, for example during metabolic acidosis, chronically desensitize ASICs (Benson et al., 1999; Alvarez De La Rosa et al., 2002; Babini et al., 2002). This steady-state desensitization of ASICs, induced by ligand concentrations above threshold, is analogous to the steady-state inactivation of voltage-gated channels by continuous depolarization above the threshold potential.
Venomous animals, such as spiders, scorpions, and sea anemones, contain a rich diversity of protein toxins that interact with different classes of ion channels. The main group of toxins that interact with ligand-gated ion channels are the conotoxins produced by cone snails (Terlau and Olivera, 2004). They inhibit channels gated by acetylcholine (McIntosh et al., 1999), by serotonin (England et al., 1998), and by glutamate (Hammerland et al., 1992), either by competitive or noncompetitive antagonism with the ligand. Toxins that interact with voltage-gated ion channels can be divided in two groups based on the mechanism of interaction. One group produces inhibition by physically occluding the ion pore (Catterall, 1980; Hille, 2001). The other group comprises gating modifiers that interact with the voltage sensor, shifting the voltage dependence of the channels. Some of them (mainly spider toxins) shift the voltage dependence to more positive potentials, inhibiting the channels (McDonough et al., 1997a,b; Swartz and MacKinnon, 1997a; Chuang et al., 1998), whereas others (-toxins of scorpions) shift the voltage dependence to more negative potentials, leading to increased channel opening at resting membrane potentials (Cestele et al., 1998; Hille, 2001). Still other toxins (-toxins of scorpions) cause a slowing of inactivation of Na+ channels (Catterall, 1979; Wang and Strichartz, 1985).
Recently, a novel protein toxin from a tarantula, psalmotoxin 1 (PcTx1), has been isolated that inhibits H+-gated ASIC1a (Escoubas et al., 2000) by an unknown mechanism. Interestingly, PcTx1 is structurally unrelated to conotoxins but related to gating modifier toxins of voltage-gated ion channels (Escoubas et al., 2003). Here we show that PcTx1 has a unique mechanism of inhibition: it increases the apparent affinity for H+ of ASIC1a, leading to chronic desensitization at the resting pH of 7.4.
MATERIALS AND METHODS
Electrophysiology
The inhibition of ASIC1a by PcTx1 was investigated by expressing homomeric ASIC1a in Xenopus laevis oocytes. Capped ASIC1a cRNA was synthesized by SP6 RNA polymerase from linearized cDNA, using the mMessage mMachine kit (Ambion). Stage V–VI Xenopus oocytes were injected with 0.01 ng cRNA and kept in OR-2 medium (concentrations in mM: 82.5 NaCl, 2.5 KCl, 1.0 Na2HPO4, 5.0 HEPES, 1.0 MgCl2, 1.0 CaCl2, and 0.5 g/l PVP; pH 7.3) for 2–4 d. Whole cell currents were recorded at 0.1 or 1 kHz and filtered at 20 Hz with a TurboTec 03X amplifier (npi electronic) using an automated, pump-driven solution exchange system together with the oocyte testing carousel controlled by the interface OTC-20 (npi electronic). Data acquisition and solution exchange were managed using the software CellWorks 5.1.1 (npi electronic). Bath solution contained (in mM) 140 NaCl, 1.8 CaCl2, 1.0 MgCl2, 10 HEPES. For the acidic test solutions, HEPES was replaced by MES buffer. Conditioning solution with pH between 6.45 and 7.0 was buffered with 5 mM HEPES/5 mM MES. Solutions containing 0.1 mM Ca2+ were supplemented with 0.1 mM flufenamic acid to block the large conductance induced in Xenopus oocytes by divalent-free extracellular solutions. Holding potential was –70 mV if not otherwise indicated.
Dose–response curves were fit to the Hill equation I = a + (Imax – a)/(1 + (EC50/[L])n), where Imax is the maximal current, a is the residual current, [L] is the concentration of the ligand (PcTx1 or H+), EC50 is the concentration at which half-maximal response occurs, and n is the Hill coefficient. Before fitting, currents from each measurement were normalized to the maximal value measured. Results are reported either, in the text, as mean ± SD or, for the figures and amplitude histograms, as mean ± SEM. They represent the mean of n individual measurements on different oocytes. Statistical analysis was done with the unpaired t-test. Psalmopoeus cambridgei venom was from SpiderPharm.
Toxin Synthesis and Refolding
The synthesis of native psalmotoxin 1 was performed using the Fmoc/But and maximal temporary protection strategy on a Syro-MultiSyntec peptide synthesizer. The chemical procedure used 0.05 mmol of Fmoc-Thr(OtBu)-2-chlorotrityl resin (0.50 mmol/g), an eightfold excess of each amino acid and 2-(1H-benzotriazole-1-yl)1,1,3,3-tetramethyluronium tetrafluoroborat/1-hydroxybenzotriazole (TBTU/HOBt) activation. Deprotection (2 h) and cleavage (100 mg peptide + resin) were achieved using 5 ml of a mixture of trifluoroacetic acid/thioanisole/ethandithiole (90/8/2, vol/vol/vol). The acidic mixture was then precipitated three times with diethylether, dissolved in 10% aqueous acetic acid and freeze dried. The crude toxin was purified by RP-HPLC on a C18 semi-preparative column (10 x 150 mm; Nucleosil) using a 40-min gradient of acetonitrile in 0.055% trifluoroacetic acid (10–80% B in 40 min, where B is 80% acetonitrile/H2O/0.05% trifluoroacetic acid).
Oxidation of the reduced toxin was achieved by dissolving the purified peptide with 2 M acetic acid, and diluted to a peptide concentration of 0.015 mM in the presence of reduced/oxidized glutathione (molar ratio of peptide/GSH/GSSG was 1:100:10) and 2 M guanidine hydrochloride. The solution was adjusted to pH 8.0 with aqueous NH4OH and stirred slowly at 4°C for 3 d. The folding reaction was monitored by analytical HPLC. The solution was concentrated using a C18 SepPak (Waters) cartridge and finally lyophilized. Purification of the oxidized product was achieved first by chromatography on a C8 column using the system above and yielding a purity of 90%. Finally the product was highly purified on a C18 column using a 60-min gradient, resulting in a purity of 95%. The quality of the product was confirmed by analytical HPLC, matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-MS), and electrospray ionization mass spectrometry (ESI-MS), giving the correct mass of oxidized product.
The affinity of synthetic PcTx1 to ASIC1a was determined as inhibition of ASIC1a current by application of PcTx1 for 125 s in the conditioning period at pH 7.4. Under these conditions, Kd was 3.7 nM and the Hill coefficient of 1.65, which is in good agreement with previous published data (Escoubas et al., 2000), demonstrating the purity of our preparation. Complete block was achieved with a concentration of 30 nM PcTx1; this concentration was, therefore, used for most of the experiments described in this study. It should be emphasized, however, that the apparent affinity of ASIC1a to PxTc1, as determined above, is pH dependent. Solutions containing PcTx1 were supplemented with 0.05% BSA to avoid absorption by the tubing.
RESULTS
Inhibition By PcTx1 Is Sensitive to the Conditioning pH
PcTx1 Inhibits ASIC1a Current By Causing a Shift of the Steady-state Desensitization Curve to Lower H+ Concentrations
Since the conditioning pH was crucial for the inhibition of ASIC1a currents by PcTx1 (Fig. 1), we asked whether the toxin affected the steady-state desensitization of ASIC1a. To address this question, we determined the pH dependence of steady-state desensitization for ASIC1a in the absence and in the presence of PcTx1 (30 nM, Kd = 3.7 nM; Fig. 2 A) during the conditioning period. Similar to a previous report (Babini et al., 2002), in the absence of PcTx1, ASIC1a showed half-maximal desensitization at a pH (pH50) of 7.19 ± 0.01 (n = 8). In strong contrast, in the presence of PcTx1, pH50 of desensitization was 7.46 ± 0.02 (Fig. 2, n = 5). Thus, PcTx1 caused a significant (P < 0.01) shift of the steady-state desensitization curve of ASIC1a by 0.27 pH units to lower H+ concentrations. Since steady-state desensitization curves of ASIC1a are very steep (transition from almost 100% to almost 0% available channels in just 0.2 pH units, Fig. 2 C) the PcTx1-induced shift is sufficient to transfer almost all the channels into the desensitized state at pH 7.4, rendering them unavailable for activation. However, at a conditioning pH of 7.9, the PcTx1-induced shift would not be sufficient for desensitization of ASIC1a channels. Thus, the observed shift of the steady-state desensitization curve by PcTx1 can fully account for the inhibition of ASIC1a by PcTx1. These results demonstrate that PcTx1 inhibits ASIC1a currents by drastically shifting the steady-state desensitization curve.
The Kinetics of PcTx1 Inhibition Is Slow
We determined the rate of onset of PcTx1 inhibition by applying 30 nM PcTx1 for a varying period of time at pH 7.4 and assessing inhibition of ASIC1a channels by activation with pH 6.0. As shown in Fig. 3, complete inhibition by PcTx1 was achieved only after application for 150 s The time course of the inhibition could be well described by a single exponential function with a time constant of 52 s. Since PcTx1 induces steady-state desensitization and steady-state desensitization of ASIC1a needs several tens of seconds for completion (Babini et al., 2002), the slow onset of inhibition by PcTx1 is at least in part due to the slow onset of steady-state desensitization and does not directly reflect slow binding of PcTx1. In agreement with this interpretation, application of PcTx1 can, under certain conditions, directly open ASIC1a channels with a time course on the order of 2 s (see Fig. 8).
We determined the rate of recovery from PcTx1 inhibition by applying 30 nM PcTx1 for 150 s to achieve complete inhibition. After washout of PcTx1, available ASIC1a channels were assessed every 70 s by application of pH 6.0. As illustrated in Fig. 3, ASIC1a slowly recovered from the PcTx1 inhibition and reached maximal current amplitude only 350 s after washout of PcTx1. Fit to a mono-exponential function revealed a time constant of 125 s (Koff = 8.0 x 10–3 s–1). Since recovery of ASIC1a from desensitization is complete within a few seconds (Babini et al., 2002), slow recovery from PcTx1 inhibition cannot be explained by slow recovery from desensitization, suggesting slow unbinding of PcTx1.
PcTx1 Increases the Apparent Affinity for H+ of ASIC1a
The shift of the steady-state desensitization curve to lower H+ concentrations can be explained by an increased apparent affinity for H+. To further address the H+ affinity of ASIC1a in the presence of PcTx1, we examined the pH dependence of ASIC1a activation in the absence and in the presence of PcTx1. We used a conditioning pH of 7.9 in order to avoid inhibition by PcTx1 (Fig. 2) and varied the pH to activate ASIC1a channels. PcTx1 was applied only during the conditioning period. Similar to previous results (Babini et al., 2002), in the absence of PcTx1, pH50 of activation was 6.56 ± 0.04 (n = 7). In contrast, in the presence of PcTx1 (30 nM), pH50 of activation was 6.66 ± 0.04 (Fig. 4, n = 7). Thus, PcTx1 also slightly, but significantly (P < 0.01), shifted the activation curves to lower H+ concentrations, confirming that PcTx1 increases the apparent H+ affinity of ASIC1a. This shift of the activation curve by PcTx1 directly demonstrates that PcTx1 bound to ASIC1a at pH 7.9 and that channels with PcTx1 bound can open. In contrast to the apparent affinity to H+, the time constant of desensitization (des) of ASIC1a currents was not significantly affected by PcTx1 (at pH 6.0, des = 4.0 ± 0.7 s, n = 7, without PcTx1, and des = 4.6 ± 1.5 s, n = 7, with PcTx1, P = 0.4).
PcTx1 Competes with Ca2+ in Binding To ASIC1a Channels
We previously reported that Ca2+ stabilizes the closed state of ASIC1a and that it competes with H+ in the binding to ASIC1a (Babini et al., 2002). An increased apparent affinity for H+ can therefore be explained by a reduced binding of Ca2+. Therefore, we next addressed the role of Ca2+ in the inhibition by PcTx1. First, we determined the steady-state desensitization curve in the presence and in the absence of PcTx1, with either a low concentration (0.1 mM) or a high concentration (10 mM) of Ca2+ in the conditioning solution. Fig. 5 shows that with 0.1 mM conditioning Ca2+, PcTx1 shifted the steady-state desensitization curve by 0.42 pH units to lower H+ concentrations (pH50 was 7.36 ± 0.01, n = 7, in the absence of PcTx1, and 7.78 ± 0.01, n = 6, in the presence of PcTx1), while with 10 mM conditioning Ca2+, PcTx1 shifted the steady-state desensitization curve by only 0.16 pH units (pH50 was 6.99 ± 0.02, n = 8, in the absence of PcTx1, and 7.15 ± 0.04, n = 6, in the presence of PcTx1). As shown above (Fig. 2), with a physiological Ca2+ concentration (1.8 mM), PcTx1 shifted the steady-state desensitization curve by 0.27 pH units, an intermediate value to 0.42 and 0.16. Thus, the degree of the PcTx1-induced shift of the steady-state desensitization curve decreased with increasing Ca2+ concentrations, suggesting competition between Ca2+ and PcTx1.
We next addressed the question of whether Ca2+ and PcTx1 compete directly for binding to the channel. To do so, we used conditioning pH 8.0 and activated channels with pH 6.7. As can be deduced from Figs. 2 and 4, under these conditions, PcTx1 will not inhibit ASIC1a but rather potentiate ASIC1a currents, due to the increased affinity for H+. We used this protocol with low and high concentrations of Ca2+ (0.1 and 10 mM) in the conditioning solution. At 0.1 mM Ca2+, potentiation of the ASIC1a current amplitude was significantly (P < 0.01) stronger than at 10 mM Ca2+ (Fig. 6; ratio of peak current amplitudes with and without PcTx1 was 2.2 ± 0.4 with 0.1 mM Ca2+, n = 6, and 1.5 ± 0.2 with 10 mM Ca2+, n = 6). Since the off rate of Ca2+ is large, the different concentrations of Ca2+ in the conditioning solution per se do not change the current amplitude (Babini et al., 2002). Since moreover during the activation by pH 6.7 the toxin was no longer applied, this result shows that during the conditioning period, more toxin bound to the channel at low Ca2+ concentrations than at high Ca2+ concentrations. Thus, Ca2+ competes with PcTx1 by preventing PcTx1 from binding to the channel.
Ca2+ and PcTx1 could compete by binding to the same site or to two different sites on the channel. Binding to the same site would predict that the unbinding of PcTx1 should be independent from the Ca2+ concentration, whereas binding to two different sites would predict that the unbinding of PcTx1 should be facilitated at higher Ca2+ concentrations. To address this point, we monitored recovery from PcTx1 inhibition using conditioning solutions containing either 10 mM Ca2+ or 0.1 mM Ca2+. Fig. 7 shows that the recovery from PcTx1 inhibition became faster with increasing Ca2+ concentrations (time constant = 441 ± 141 s with 0.1 mM Ca2+, n = 7, and = 118 ± 38 s with 10 mM Ca2+, n = 6, P < 0.01). This result is in agreement with the idea that Ca2+ promotes the unbinding of PcTx1 and that Ca2+ binds to a different site than does PcTx1. However, there are other possible explanations for this result and it does not prove allosteric binding.
PcTx1 Directly Opens ASIC1a Channels
The basic features of ASIC gating can be described by a kinetic scheme where H+ binds to the closed state of the channel and closed channels with H+ bound can either desensitize or open. H+-bound closed, open, and desensitized states would be cyclically connected (see DISCUSSION). Since PcTx1 shifts channels into the desensitized state, this kinetic scheme predicts that upon PcTx1 binding, some ASIC1a channels briefly open before desensitizing. At low H+ concentrations, the opening of ASIC1a channels by PcTx1 would not lead to a macroscopic current because the rate of opening of individual channels is much smaller than the rate of desensitization and the open state is never significantly populated. Therefore, to address if PcTx1 is indeed able to open ASIC1a channels, we choose conditions under which ASIC1a channels more readily open, such as increasing the conditioning H+ concentration or decreasing the Ca2+ concentration. These conditions should increase the rate of opening of ASIC1a channels. First, we choose a conditioning pH of 7.35 (Fig. 2 C), the lowest pH value where steady-state desensitization of ASIC1a is negligible. Switching from 1.8 mM Ca2+ to 0.1 mM Ca2+ did not open ASIC1a channels (Fig. 8). In contrast, simultaneously switching from 1.8 mM to 0.1 mM Ca2+ and applying PcTx1 (100 nM) induced a transient inward current, which was sensitive to 100 μM amiloride (Fig. 8). Similarly, simultaneously lowering the pH from 7.4 to 7.1 and applying PcTx1 (100 nM) induced a transient amiloride-sensitive inward current, whereas just lowering the pH from 7.4 to 7.1 did not induce any current. These data allow us to watch ASIC1a channels that bound PcTx1 on their way from the closed to the desensitized state, illustrating that PcTx1 inhibits ASIC1a by increasing its apparent affinity to H+.
The rising phase of the PcTx1-induced current could be fit to a single exponential function with a time constant of 1.8 ± 0.3 s (0.1 Ca2+, n = 5) and 2.8 ± 0.8 s (pH 7.1, n = 5), demonstrating that under these conditions, binding of PcTx1 to ASIC1a happened on the order of seconds rather than tens of seconds as suggested by the slow onset of inhibition.
DISCUSSION
In this study, we have shown that the spider toxin PcTx1 inhibits ASIC1a by increasing its apparent affinity for its natural ligand, H+. Evidence for the increased apparent affinity is the shift of the steady-state desensitization curve and the H+ activation curve to lower H+ concentrations. A shift of the steady-state desensitization curve by 0.27 pH units corresponds to an approximately twofold increase in the apparent affinity for H+. This relatively modest increase in affinity can account for inhibition only in conjunction with the extraordinarily steep steady-state desensitization curves and the H+ affinity of ASIC1a being very close to the resting concentration of H+. Resting concentrations of H+ are 40–50 nM (pH 7.4–pH 7.3), and 65 nM of H+ (pH 7.2) will already desensitize 50% of ASIC1a channels at equilibrium (Fig. 2). Hence a considerable fraction of ASIC1a channels will be already desensitized under resting conditions and the twofold increase of the apparent H+ affinity will shift most channels into the desensitized state and can fully account for block by PcTx1. This result also shows that ASIC1a is an extremely sensitive receptor for H+, with an apparent affinity for its ligand as high as it can be without leading to significant channel desensitization. It should be emphasized that the PcTx1-induced shift of the steady-state desensitization curve is concentration dependent; that means that lower PcTx1 concentrations will induce a smaller shift and higher concentrations will induce a larger shift.
Different mechanisms could explain the apparent increase in H+ affinity of ASIC1a by PcTx1. Assuming the above-mentioned kinetic scheme:
PcTx1 could either modify the energetics of the H+ binding step, increasing the affinity of ASIC1a for H+ and thus shifting the equilibrium between the closed state without H+ (C) and the closed state with H+ bound (C·H+) toward the bound state. Or it could modify the energetics of the gating step, shifting the equilibrium between the bound closed state (C·H+) and the open state (O·H+) toward the open state, analogous to the interaction proposed for the tarantula toxin hanatoxin and the voltage-dependent K+ channel Kv2.1 (Swartz and MacKinnon, 1997a). And finally, PcTx1 could have a higher affinity to the open and/or the desensitized state than to the closed state, which would be an example for the modulated receptor concept that has been proposed for the interaction of local anesthetics with voltage-gated Na+ channels (Hille, 1977) and of dihydropyridines with L-type Ca2+ channels (Bean, 1984). Our results do not allow us to unequivocally decide between these three possibilities.
On and off rates for PcTx1 were strikingly slow. We determined an off rate Koff of 8.0 x 10–3 s–1. For the same pH, we determined an apparent Kd of 3.7 x 10–9 M (see MATERIALS AND METHODS). Assuming that Kd = Koff/Kon, one can calculate an apparent on rate Kon of 2.2 x 106 M–1 s–1, which is smaller than expected from a simply diffusion-limited process. Similarly slow on and off rates are characteristic for toxins that are structurally related to PcTx1 (Swartz and MacKinnon, 1995; Oswald et al., 2002). Recently, it has been proposed that these slow rates may be due to a membrane-access mechanism for the toxins (Lee and MacKinnon, 2004; Suchyna et al., 2004). According to this hypothesis, these toxins partition into the membrane, leading to their accumulation in high concentrations around the channel (Lee and MacKinnon, 2004; Suchyna et al., 2004). Thus, binding energy between toxin and channel need not be as high as would be expected for an interaction with nanomolar affinity, and slow inhibition kinetics could be due to a slow rate of membrane partitioning. The highly similar structure of PcTx1 to gating modifier toxins (Escoubas et al., 2003) suggests that PcTx1 also partitions into the membrane, providing a clue to the slow kinetics and implying that the toxin-binding site on ASIC1a is close to or within the membrane-spanning parts of the channel protein.
We also show that Ca2+ competes with PcTx1, probably by inhibiting binding of PcTx1 to the channel. Previously, we and others have shown that Ca2+ is crucial for the activation of ASICs (Babini et al., 2002; Immke and McCleskey, 2003; Paukert et al., 2004). Ca2+ competes with H+, reducing the apparent affinity for H+ (Babini et al., 2002; Immke and McCleskey, 2003). Thus, at least part of the effect of PcTx1 on ASIC1a can be explained in terms of a competition with Ca2+. It is likely that the Ca2+ binding site on the channel contains negatively charged side chains of amino acids and that protonation of these side chains leads to Ca2+ unbinding. Hence the Ca2+ binding site may be identical to the "H+ receptor" of ASICs. Moreover, gating modifier toxins that cause a shift of the voltage dependence of voltage-gated channels interact directly with the voltage sensor (Swartz and MacKinnon, 1997b; Ruta and MacKinnon, 2004). Thus, it is attractive to speculate that PcTx1 interacts with the Ca2+/H+ binding site of ASIC1a. In agreement with this hypothesis is the clustering of several positively charged amino acids in the putative interaction domain of PcTx1 (Escoubas et al., 2003) that are likely to interact with negatively charged amino acids on the channel protein. However, ASIC1a can open with PcTx1 bound, whereas Ca2+ probably has to unbind for channel opening (Immke and McCleskey, 2003; Paukert et al., 2004), arguing against an identical binding site for Ca2+ and PcTx1. Moreover and as pointed out previously, the Ca2+ dependence of the off rate of PcTx1 (Fig. 7) suggests that Ca2+ binds to a different site than PcTx1. Finally, it cannot be excluded that Ca2+ exerts its effects via screening of surface charges; the competition between PcTx1 and Ca2+ could then be much similar to the competition between dihydropyridines (the Ca2+ antagonists) and Ca2+ on heart Ca2+ channels (Kass and Krafte, 1987). Irrespective of the relation to the Ca2+ binding site, we anticipate that identification of the PcTX1 binding site will increase our understanding of activation of ASIC1a by H+.
Within the ASIC family, increased apparent ligand affinity may be an inhibitory mechanism that applies to other toxin/channel interactions than only PcTx1/ASIC1a. Recently, a protein toxin from a sea anemone, APETx2, has been identified that inhibits ASIC3 (Diochot et al., 2004); APETx2 does not inhibit ASIC1a. The interaction of ASIC3 with APETx2 is characterized by inhibition during steady state (at pH 7.4), nanomolar affinity, slow kinetics, and unchanged single channel conductance (Diochot et al., 2004). These characteristics suggest that APETx2 may inhibit ASIC3 by an increased apparent H+ affinity.
Understanding the PcTx1/ASIC1a interaction may be of therapeutic value. Very recently, it has been shown that activation of ASIC1a during the acidosis accompanying brain ischemia leads to significant Ca2+ influx that contributes to ischemic cell death in the brain (Xiong et al., 2004). Inhibition of ASIC1a by PcTx1 significantly decreased neuronal death during ischemia (Xiong et al., 2004). Therefore, our results suggest that a modest increase in the apparent affinity of ASIC1a by pharmacological intervention could desensitize ASIC1a, being of potential benefit in patients at risk of stroke.
ACKNOWLEDGMENTS
We thank Drs. M. Pusch and M. Paukert for comments on the manuscript.
This work was supported by a grant of the Deutsche Forschungsgemeinschaft to S. Gründer (GR1771/3-3).
Olaf S. Andersen served as editor.
Submitted: 20 April 2005
Accepted: 26 May 2005
REFERENCES
Alvarez De La Rosa, D., P. Zhang, D. Shao, F. White, and C.M. Canessa. 2002. Functional implications of the localization and activity of acid-sensitive channels in rat peripheral nervous system. Proc. Natl. Acad. Sci. USA. 99:2326–2331.
Babini, E., M. Paukert, H.S. Geisler, and S. Gründer. 2002. Alternative splicing and interaction with di- and polyvalent cations control the dynamic range of acid-sensing ion channel 1 (ASIC1). J. Biol. Chem. 277:41597–41603.
Bean, B.P. 1984. Nitrendipine block of cardiac calcium channels: high-affinity binding to the inactivated state. Proc. Natl. Acad. Sci. USA. 81:6388–6392.
Benson, C.J., S.P. Eckert, and E.W. McCleskey. 1999. Acid-evoked currents in cardiac sensory neurons: a possible mediator of myocardial ischemic sensation. Circ. Res. 84:921–928.
Bssler, E.L., T.J. Ngo-Anh, H.S. Geisler, J.P. Ruppersberg, and S. Gründer. 2001. Molecular and functional characterization of acid-sensing ion channel (ASIC) 1b. J. Biol. Chem. 276:33782–33787.
Catterall, W.A. 1979. Binding of scorpion toxin to receptor sites associated with sodium channels in frog muscle. Correlation of voltage-dependent binding with activation. J. Gen. Physiol. 74:375–391.
Catterall, W.A. 1980. Neurotoxins that act on voltage-sensitive sodium channels in excitable membranes. Annu. Rev. Pharmacol. Toxicol. 20:15–43.
Cestele, S., Y. Qu, J.C. Rogers, H. Rochat, T. Scheuer, and W.A. Catterall. 1998. Voltage sensor-trapping: enhanced activation of sodium channels by -scorpion toxin bound to the S3-S4 loop in domain II. Neuron. 21:919–931.
Chuang, R.S., H. Jaffe, L. Cribbs, E. Perez-Reyes, and K.J. Swartz. 1998. Inhibition of T-type voltage-gated calcium channels by a new scorpion toxin. Nat. Neurosci. 1:668–674.
Diochot, S., A. Baron, L.D. Rash, E. Deval, P. Escoubas, S. Scarzello, M. Salinas, and M. Lazdunski. 2004. A new sea anemone peptide, APETx2, inhibits ASIC3, a major acid-sensitive channel in sensory neurons. EMBO J. 23:1516–1525.
England, L.J., J. Imperial, R. Jacobsen, A.G. Craig, J. Gulyas, M. Akhtar, J. Rivier, D. Julius, and B.M. Olivera. 1998. Inactivation of a serotonin-gated ion channel by a polypeptide toxin from marine snails. Science. 281:575–578.
Escoubas, P., J.R. De Weille, A. Lecoq, S. Diochot, R. Waldmann, G. Champigny, D. Moinier, A. Menez, and M. Lazdunski. 2000. Isolation of a tarantula toxin specific for a class of proton-gated Na+ channels. J. Biol. Chem. 275:25116–25121.
Escoubas, P., C. Bernard, G. Lambeau, M. Lazdunski, and H. Darbon. 2003. Recombinant production and solution structure of PcTx1, the specific peptide inhibitor of ASIC1a proton-gated cation channels. Protein Sci. 12:1332–1343.
Hammerland, L.G., B.M. Olivera, and D. Yoshikami. 1992. Conantokin-G selectively inhibits N-methyl-D-aspartate-induced currents in Xenopus oocytes injected with mouse brain mRNA. Eur. J. Pharmacol. 226:239–244.
Hille, B. 1977. Local anesthetics: hydrophilic and hydrophobic pathways for the drug–receptor reaction. J. Gen. Physiol. 69:497–515.
Hille, B. 2001. Ion Channels of Excitable Membranes. Sinauer Associates, Sunderland, MA. 722 pp.
Immke, D.C., and E.W. McCleskey. 2003. Protons open acid-sensing ion channels by catalyzing relief of Ca2+ blockade. Neuron. 37:75–84.
Kass, R.S., and D.S. Krafte. 1987. Negative surface charge density near heart calcium channels. Relevance to block by dihydropyridines. J. Gen. Physiol. 89:629–644.
Lee, S.Y., and R. MacKinnon. 2004. A membrane-access mechanism of ion channel inhibition by voltage sensor toxins from spider venom. Nature. 430:232–235.
McDonough, S.I., R.A. Lampe, R.A. Keith, and B.P. Bean. 1997a. Voltage-dependent inhibition of N- and P-type calcium channels by the peptide toxin omega-grammotoxin-SIA. Mol. Pharmacol. 52:1095–1104.
McDonough, S.I., I.M. Mintz, and B.P. Bean. 1997b. Alteration of P-type calcium channel gating by the spider toxin -Aga-IVA. Biophys. J. 72:2117–2128.
McIntosh, J.M., A.D. Santos, and B.M. Olivera. 1999. Conus peptides targeted to specific nicotinic acetylcholine receptor subtypes. Annu. Rev. Biochem. 68:59–88.
Oswald, R.E., T.M. Suchyna, R. McFeeters, P. Gottlieb, and F. Sachs. 2002. Solution structure of peptide toxins that block mechanosensitive ion channels. J. Biol. Chem. 277:34443–34450.
Paukert, M., E. Babini, M. Pusch, and S. Gründer. 2004. Identification of the Ca2+ blocking site of acid-sensing ion channel (ASIC) 1: implications for channel gating. J. Gen. Physiol. 124:383–394.
Ruta, V., and R. MacKinnon. 2004. Localization of the voltage-sensor toxin receptor on KvAP. Biochemistry. 43:10071–10079.
Suchyna, T.M., S.E. Tape, R.E. Koeppe II, O.S. Andersen, F. Sachs, and P.A. Gottlieb. 2004. Bilayer-dependent inhibition of mechanosensitive channels by neuroactive peptide enantiomers. Nature. 430:235–240.
Swartz, K.J., and R. MacKinnon. 1995. An inhibitor of the Kv2.1 potassium channel isolated from the venom of a Chilean tarantula. Neuron. 15:941–949.
Swartz, K.J., and R. MacKinnon. 1997a. Hanatoxin modifies the gating of a voltage-dependent K+ channel through multiple binding sites. Neuron. 18:665–673.
Swartz, K.J., and R. MacKinnon. 1997b. Mapping the receptor site for hanatoxin, a gating modifier of voltage-dependent K+ channels. Neuron. 18:675–682.
Terlau, H., and B.M. Olivera. 2004. Conus venoms: a rich source of novel ion channel-targeted peptides. Physiol. Rev. 84:41–68.
Waldmann, R., and M. Lazdunski. 1998. H+-gated cation channels: neuronal acid sensors in the NaC/DEG family of ion channels. Curr. Opin. Neurobiol. 8:418–424.
Waldmann, R., G. Champigny, F. Bassilana, C. Heurteaux, and M. Lazdunski. 1997. A proton-gated cation channel involved in acid-sensing. Nature. 386:173–177.
Wang, G.K., and G. Strichartz. 1985. Kinetic analysis of the action of Leiurus scorpion -toxin on ionic currents in myelinated nerve. J. Gen. Physiol. 86:739–762.
Xiong, Z.G., X.M. Zhu, X.P. Chu, M. Minami, J. Hey, W.L. Wei, J.F. MacDonald, J.A. Wemmie, M.P. Price, M.J. Welsh, and R.P. Simon. 2004. Neuroprotection in ischemia: blocking calcium-permeable acid-sensing ion channels. Cell. 118:687–698.(Xuanmao Chen,, Hubert Kal)
2 Interfakultres Institut für Biochemie, University of Tübingen, 72074 Tübingen, Germany
3 Department of Physiology II, University of Würzburg, 97070 Würzburg, Germany
Acid-sensing ion channels (ASICs) are ion channels activated by extracellular protons. They are involved in higher brain functions and perception of pain, taste, and mechanical stimuli. Homomeric ASIC1a is potently inhibited by the tarantula toxin psalmotoxin 1. The mechanism of this inhibition is unknown. Here we show that psalmotoxin 1 inhibits ASIC1a by a unique mechanism: the toxin increases the apparent affinity for H+ of ASIC1a. Since ASIC1a is activated by H+ concentrations that are only slightly larger than the resting H+ concentration, this increase in H+ affinity is sufficient to shift ASIC1a channels into the desensitized state. As activation of ASIC1a has recently been linked to neurodegeneration associated with stroke, our results suggest chronic desensitization of ASIC1a by a slight increase of its H+ affinity as a possible way of therapeutic intervention in stroke.
Abbreviations used in this paper: ASIC, acid-sensing ion channel; PcTx1, psalmotoxin 1.
INTRODUCTION
H+ is chemically the simplest transmitter. H+ receptor channels have recently been characterized that open upon an increase in the extracellular concentration of H+ and desensitize in the continuous presence of H+ (Waldmann et al., 1997; Waldmann and Lazdunski, 1998). The apparent affinity of these ion channels for H+, 0.3 μM, and their activation and desensitization kinetics is comparable with other ligand-gated channels (Waldmann and Lazdunski, 1998; Bssler et al., 2001), establishing H+ as the genuine ligand for these channels, the acid-sensing ion channels (ASICs). However, in contrast to other ligands, resting concentrations of H+ are close to the threshold concentration needed to activate ASICs. Consequently, slightly increased H+ concentrations, for example during metabolic acidosis, chronically desensitize ASICs (Benson et al., 1999; Alvarez De La Rosa et al., 2002; Babini et al., 2002). This steady-state desensitization of ASICs, induced by ligand concentrations above threshold, is analogous to the steady-state inactivation of voltage-gated channels by continuous depolarization above the threshold potential.
Venomous animals, such as spiders, scorpions, and sea anemones, contain a rich diversity of protein toxins that interact with different classes of ion channels. The main group of toxins that interact with ligand-gated ion channels are the conotoxins produced by cone snails (Terlau and Olivera, 2004). They inhibit channels gated by acetylcholine (McIntosh et al., 1999), by serotonin (England et al., 1998), and by glutamate (Hammerland et al., 1992), either by competitive or noncompetitive antagonism with the ligand. Toxins that interact with voltage-gated ion channels can be divided in two groups based on the mechanism of interaction. One group produces inhibition by physically occluding the ion pore (Catterall, 1980; Hille, 2001). The other group comprises gating modifiers that interact with the voltage sensor, shifting the voltage dependence of the channels. Some of them (mainly spider toxins) shift the voltage dependence to more positive potentials, inhibiting the channels (McDonough et al., 1997a,b; Swartz and MacKinnon, 1997a; Chuang et al., 1998), whereas others (-toxins of scorpions) shift the voltage dependence to more negative potentials, leading to increased channel opening at resting membrane potentials (Cestele et al., 1998; Hille, 2001). Still other toxins (-toxins of scorpions) cause a slowing of inactivation of Na+ channels (Catterall, 1979; Wang and Strichartz, 1985).
Recently, a novel protein toxin from a tarantula, psalmotoxin 1 (PcTx1), has been isolated that inhibits H+-gated ASIC1a (Escoubas et al., 2000) by an unknown mechanism. Interestingly, PcTx1 is structurally unrelated to conotoxins but related to gating modifier toxins of voltage-gated ion channels (Escoubas et al., 2003). Here we show that PcTx1 has a unique mechanism of inhibition: it increases the apparent affinity for H+ of ASIC1a, leading to chronic desensitization at the resting pH of 7.4.
MATERIALS AND METHODS
Electrophysiology
The inhibition of ASIC1a by PcTx1 was investigated by expressing homomeric ASIC1a in Xenopus laevis oocytes. Capped ASIC1a cRNA was synthesized by SP6 RNA polymerase from linearized cDNA, using the mMessage mMachine kit (Ambion). Stage V–VI Xenopus oocytes were injected with 0.01 ng cRNA and kept in OR-2 medium (concentrations in mM: 82.5 NaCl, 2.5 KCl, 1.0 Na2HPO4, 5.0 HEPES, 1.0 MgCl2, 1.0 CaCl2, and 0.5 g/l PVP; pH 7.3) for 2–4 d. Whole cell currents were recorded at 0.1 or 1 kHz and filtered at 20 Hz with a TurboTec 03X amplifier (npi electronic) using an automated, pump-driven solution exchange system together with the oocyte testing carousel controlled by the interface OTC-20 (npi electronic). Data acquisition and solution exchange were managed using the software CellWorks 5.1.1 (npi electronic). Bath solution contained (in mM) 140 NaCl, 1.8 CaCl2, 1.0 MgCl2, 10 HEPES. For the acidic test solutions, HEPES was replaced by MES buffer. Conditioning solution with pH between 6.45 and 7.0 was buffered with 5 mM HEPES/5 mM MES. Solutions containing 0.1 mM Ca2+ were supplemented with 0.1 mM flufenamic acid to block the large conductance induced in Xenopus oocytes by divalent-free extracellular solutions. Holding potential was –70 mV if not otherwise indicated.
Dose–response curves were fit to the Hill equation I = a + (Imax – a)/(1 + (EC50/[L])n), where Imax is the maximal current, a is the residual current, [L] is the concentration of the ligand (PcTx1 or H+), EC50 is the concentration at which half-maximal response occurs, and n is the Hill coefficient. Before fitting, currents from each measurement were normalized to the maximal value measured. Results are reported either, in the text, as mean ± SD or, for the figures and amplitude histograms, as mean ± SEM. They represent the mean of n individual measurements on different oocytes. Statistical analysis was done with the unpaired t-test. Psalmopoeus cambridgei venom was from SpiderPharm.
Toxin Synthesis and Refolding
The synthesis of native psalmotoxin 1 was performed using the Fmoc/But and maximal temporary protection strategy on a Syro-MultiSyntec peptide synthesizer. The chemical procedure used 0.05 mmol of Fmoc-Thr(OtBu)-2-chlorotrityl resin (0.50 mmol/g), an eightfold excess of each amino acid and 2-(1H-benzotriazole-1-yl)1,1,3,3-tetramethyluronium tetrafluoroborat/1-hydroxybenzotriazole (TBTU/HOBt) activation. Deprotection (2 h) and cleavage (100 mg peptide + resin) were achieved using 5 ml of a mixture of trifluoroacetic acid/thioanisole/ethandithiole (90/8/2, vol/vol/vol). The acidic mixture was then precipitated three times with diethylether, dissolved in 10% aqueous acetic acid and freeze dried. The crude toxin was purified by RP-HPLC on a C18 semi-preparative column (10 x 150 mm; Nucleosil) using a 40-min gradient of acetonitrile in 0.055% trifluoroacetic acid (10–80% B in 40 min, where B is 80% acetonitrile/H2O/0.05% trifluoroacetic acid).
Oxidation of the reduced toxin was achieved by dissolving the purified peptide with 2 M acetic acid, and diluted to a peptide concentration of 0.015 mM in the presence of reduced/oxidized glutathione (molar ratio of peptide/GSH/GSSG was 1:100:10) and 2 M guanidine hydrochloride. The solution was adjusted to pH 8.0 with aqueous NH4OH and stirred slowly at 4°C for 3 d. The folding reaction was monitored by analytical HPLC. The solution was concentrated using a C18 SepPak (Waters) cartridge and finally lyophilized. Purification of the oxidized product was achieved first by chromatography on a C8 column using the system above and yielding a purity of 90%. Finally the product was highly purified on a C18 column using a 60-min gradient, resulting in a purity of 95%. The quality of the product was confirmed by analytical HPLC, matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-MS), and electrospray ionization mass spectrometry (ESI-MS), giving the correct mass of oxidized product.
The affinity of synthetic PcTx1 to ASIC1a was determined as inhibition of ASIC1a current by application of PcTx1 for 125 s in the conditioning period at pH 7.4. Under these conditions, Kd was 3.7 nM and the Hill coefficient of 1.65, which is in good agreement with previous published data (Escoubas et al., 2000), demonstrating the purity of our preparation. Complete block was achieved with a concentration of 30 nM PcTx1; this concentration was, therefore, used for most of the experiments described in this study. It should be emphasized, however, that the apparent affinity of ASIC1a to PxTc1, as determined above, is pH dependent. Solutions containing PcTx1 were supplemented with 0.05% BSA to avoid absorption by the tubing.
RESULTS
Inhibition By PcTx1 Is Sensitive to the Conditioning pH
PcTx1 Inhibits ASIC1a Current By Causing a Shift of the Steady-state Desensitization Curve to Lower H+ Concentrations
Since the conditioning pH was crucial for the inhibition of ASIC1a currents by PcTx1 (Fig. 1), we asked whether the toxin affected the steady-state desensitization of ASIC1a. To address this question, we determined the pH dependence of steady-state desensitization for ASIC1a in the absence and in the presence of PcTx1 (30 nM, Kd = 3.7 nM; Fig. 2 A) during the conditioning period. Similar to a previous report (Babini et al., 2002), in the absence of PcTx1, ASIC1a showed half-maximal desensitization at a pH (pH50) of 7.19 ± 0.01 (n = 8). In strong contrast, in the presence of PcTx1, pH50 of desensitization was 7.46 ± 0.02 (Fig. 2, n = 5). Thus, PcTx1 caused a significant (P < 0.01) shift of the steady-state desensitization curve of ASIC1a by 0.27 pH units to lower H+ concentrations. Since steady-state desensitization curves of ASIC1a are very steep (transition from almost 100% to almost 0% available channels in just 0.2 pH units, Fig. 2 C) the PcTx1-induced shift is sufficient to transfer almost all the channels into the desensitized state at pH 7.4, rendering them unavailable for activation. However, at a conditioning pH of 7.9, the PcTx1-induced shift would not be sufficient for desensitization of ASIC1a channels. Thus, the observed shift of the steady-state desensitization curve by PcTx1 can fully account for the inhibition of ASIC1a by PcTx1. These results demonstrate that PcTx1 inhibits ASIC1a currents by drastically shifting the steady-state desensitization curve.
The Kinetics of PcTx1 Inhibition Is Slow
We determined the rate of onset of PcTx1 inhibition by applying 30 nM PcTx1 for a varying period of time at pH 7.4 and assessing inhibition of ASIC1a channels by activation with pH 6.0. As shown in Fig. 3, complete inhibition by PcTx1 was achieved only after application for 150 s The time course of the inhibition could be well described by a single exponential function with a time constant of 52 s. Since PcTx1 induces steady-state desensitization and steady-state desensitization of ASIC1a needs several tens of seconds for completion (Babini et al., 2002), the slow onset of inhibition by PcTx1 is at least in part due to the slow onset of steady-state desensitization and does not directly reflect slow binding of PcTx1. In agreement with this interpretation, application of PcTx1 can, under certain conditions, directly open ASIC1a channels with a time course on the order of 2 s (see Fig. 8).
We determined the rate of recovery from PcTx1 inhibition by applying 30 nM PcTx1 for 150 s to achieve complete inhibition. After washout of PcTx1, available ASIC1a channels were assessed every 70 s by application of pH 6.0. As illustrated in Fig. 3, ASIC1a slowly recovered from the PcTx1 inhibition and reached maximal current amplitude only 350 s after washout of PcTx1. Fit to a mono-exponential function revealed a time constant of 125 s (Koff = 8.0 x 10–3 s–1). Since recovery of ASIC1a from desensitization is complete within a few seconds (Babini et al., 2002), slow recovery from PcTx1 inhibition cannot be explained by slow recovery from desensitization, suggesting slow unbinding of PcTx1.
PcTx1 Increases the Apparent Affinity for H+ of ASIC1a
The shift of the steady-state desensitization curve to lower H+ concentrations can be explained by an increased apparent affinity for H+. To further address the H+ affinity of ASIC1a in the presence of PcTx1, we examined the pH dependence of ASIC1a activation in the absence and in the presence of PcTx1. We used a conditioning pH of 7.9 in order to avoid inhibition by PcTx1 (Fig. 2) and varied the pH to activate ASIC1a channels. PcTx1 was applied only during the conditioning period. Similar to previous results (Babini et al., 2002), in the absence of PcTx1, pH50 of activation was 6.56 ± 0.04 (n = 7). In contrast, in the presence of PcTx1 (30 nM), pH50 of activation was 6.66 ± 0.04 (Fig. 4, n = 7). Thus, PcTx1 also slightly, but significantly (P < 0.01), shifted the activation curves to lower H+ concentrations, confirming that PcTx1 increases the apparent H+ affinity of ASIC1a. This shift of the activation curve by PcTx1 directly demonstrates that PcTx1 bound to ASIC1a at pH 7.9 and that channels with PcTx1 bound can open. In contrast to the apparent affinity to H+, the time constant of desensitization (des) of ASIC1a currents was not significantly affected by PcTx1 (at pH 6.0, des = 4.0 ± 0.7 s, n = 7, without PcTx1, and des = 4.6 ± 1.5 s, n = 7, with PcTx1, P = 0.4).
PcTx1 Competes with Ca2+ in Binding To ASIC1a Channels
We previously reported that Ca2+ stabilizes the closed state of ASIC1a and that it competes with H+ in the binding to ASIC1a (Babini et al., 2002). An increased apparent affinity for H+ can therefore be explained by a reduced binding of Ca2+. Therefore, we next addressed the role of Ca2+ in the inhibition by PcTx1. First, we determined the steady-state desensitization curve in the presence and in the absence of PcTx1, with either a low concentration (0.1 mM) or a high concentration (10 mM) of Ca2+ in the conditioning solution. Fig. 5 shows that with 0.1 mM conditioning Ca2+, PcTx1 shifted the steady-state desensitization curve by 0.42 pH units to lower H+ concentrations (pH50 was 7.36 ± 0.01, n = 7, in the absence of PcTx1, and 7.78 ± 0.01, n = 6, in the presence of PcTx1), while with 10 mM conditioning Ca2+, PcTx1 shifted the steady-state desensitization curve by only 0.16 pH units (pH50 was 6.99 ± 0.02, n = 8, in the absence of PcTx1, and 7.15 ± 0.04, n = 6, in the presence of PcTx1). As shown above (Fig. 2), with a physiological Ca2+ concentration (1.8 mM), PcTx1 shifted the steady-state desensitization curve by 0.27 pH units, an intermediate value to 0.42 and 0.16. Thus, the degree of the PcTx1-induced shift of the steady-state desensitization curve decreased with increasing Ca2+ concentrations, suggesting competition between Ca2+ and PcTx1.
We next addressed the question of whether Ca2+ and PcTx1 compete directly for binding to the channel. To do so, we used conditioning pH 8.0 and activated channels with pH 6.7. As can be deduced from Figs. 2 and 4, under these conditions, PcTx1 will not inhibit ASIC1a but rather potentiate ASIC1a currents, due to the increased affinity for H+. We used this protocol with low and high concentrations of Ca2+ (0.1 and 10 mM) in the conditioning solution. At 0.1 mM Ca2+, potentiation of the ASIC1a current amplitude was significantly (P < 0.01) stronger than at 10 mM Ca2+ (Fig. 6; ratio of peak current amplitudes with and without PcTx1 was 2.2 ± 0.4 with 0.1 mM Ca2+, n = 6, and 1.5 ± 0.2 with 10 mM Ca2+, n = 6). Since the off rate of Ca2+ is large, the different concentrations of Ca2+ in the conditioning solution per se do not change the current amplitude (Babini et al., 2002). Since moreover during the activation by pH 6.7 the toxin was no longer applied, this result shows that during the conditioning period, more toxin bound to the channel at low Ca2+ concentrations than at high Ca2+ concentrations. Thus, Ca2+ competes with PcTx1 by preventing PcTx1 from binding to the channel.
Ca2+ and PcTx1 could compete by binding to the same site or to two different sites on the channel. Binding to the same site would predict that the unbinding of PcTx1 should be independent from the Ca2+ concentration, whereas binding to two different sites would predict that the unbinding of PcTx1 should be facilitated at higher Ca2+ concentrations. To address this point, we monitored recovery from PcTx1 inhibition using conditioning solutions containing either 10 mM Ca2+ or 0.1 mM Ca2+. Fig. 7 shows that the recovery from PcTx1 inhibition became faster with increasing Ca2+ concentrations (time constant = 441 ± 141 s with 0.1 mM Ca2+, n = 7, and = 118 ± 38 s with 10 mM Ca2+, n = 6, P < 0.01). This result is in agreement with the idea that Ca2+ promotes the unbinding of PcTx1 and that Ca2+ binds to a different site than does PcTx1. However, there are other possible explanations for this result and it does not prove allosteric binding.
PcTx1 Directly Opens ASIC1a Channels
The basic features of ASIC gating can be described by a kinetic scheme where H+ binds to the closed state of the channel and closed channels with H+ bound can either desensitize or open. H+-bound closed, open, and desensitized states would be cyclically connected (see DISCUSSION). Since PcTx1 shifts channels into the desensitized state, this kinetic scheme predicts that upon PcTx1 binding, some ASIC1a channels briefly open before desensitizing. At low H+ concentrations, the opening of ASIC1a channels by PcTx1 would not lead to a macroscopic current because the rate of opening of individual channels is much smaller than the rate of desensitization and the open state is never significantly populated. Therefore, to address if PcTx1 is indeed able to open ASIC1a channels, we choose conditions under which ASIC1a channels more readily open, such as increasing the conditioning H+ concentration or decreasing the Ca2+ concentration. These conditions should increase the rate of opening of ASIC1a channels. First, we choose a conditioning pH of 7.35 (Fig. 2 C), the lowest pH value where steady-state desensitization of ASIC1a is negligible. Switching from 1.8 mM Ca2+ to 0.1 mM Ca2+ did not open ASIC1a channels (Fig. 8). In contrast, simultaneously switching from 1.8 mM to 0.1 mM Ca2+ and applying PcTx1 (100 nM) induced a transient inward current, which was sensitive to 100 μM amiloride (Fig. 8). Similarly, simultaneously lowering the pH from 7.4 to 7.1 and applying PcTx1 (100 nM) induced a transient amiloride-sensitive inward current, whereas just lowering the pH from 7.4 to 7.1 did not induce any current. These data allow us to watch ASIC1a channels that bound PcTx1 on their way from the closed to the desensitized state, illustrating that PcTx1 inhibits ASIC1a by increasing its apparent affinity to H+.
The rising phase of the PcTx1-induced current could be fit to a single exponential function with a time constant of 1.8 ± 0.3 s (0.1 Ca2+, n = 5) and 2.8 ± 0.8 s (pH 7.1, n = 5), demonstrating that under these conditions, binding of PcTx1 to ASIC1a happened on the order of seconds rather than tens of seconds as suggested by the slow onset of inhibition.
DISCUSSION
In this study, we have shown that the spider toxin PcTx1 inhibits ASIC1a by increasing its apparent affinity for its natural ligand, H+. Evidence for the increased apparent affinity is the shift of the steady-state desensitization curve and the H+ activation curve to lower H+ concentrations. A shift of the steady-state desensitization curve by 0.27 pH units corresponds to an approximately twofold increase in the apparent affinity for H+. This relatively modest increase in affinity can account for inhibition only in conjunction with the extraordinarily steep steady-state desensitization curves and the H+ affinity of ASIC1a being very close to the resting concentration of H+. Resting concentrations of H+ are 40–50 nM (pH 7.4–pH 7.3), and 65 nM of H+ (pH 7.2) will already desensitize 50% of ASIC1a channels at equilibrium (Fig. 2). Hence a considerable fraction of ASIC1a channels will be already desensitized under resting conditions and the twofold increase of the apparent H+ affinity will shift most channels into the desensitized state and can fully account for block by PcTx1. This result also shows that ASIC1a is an extremely sensitive receptor for H+, with an apparent affinity for its ligand as high as it can be without leading to significant channel desensitization. It should be emphasized that the PcTx1-induced shift of the steady-state desensitization curve is concentration dependent; that means that lower PcTx1 concentrations will induce a smaller shift and higher concentrations will induce a larger shift.
Different mechanisms could explain the apparent increase in H+ affinity of ASIC1a by PcTx1. Assuming the above-mentioned kinetic scheme:
PcTx1 could either modify the energetics of the H+ binding step, increasing the affinity of ASIC1a for H+ and thus shifting the equilibrium between the closed state without H+ (C) and the closed state with H+ bound (C·H+) toward the bound state. Or it could modify the energetics of the gating step, shifting the equilibrium between the bound closed state (C·H+) and the open state (O·H+) toward the open state, analogous to the interaction proposed for the tarantula toxin hanatoxin and the voltage-dependent K+ channel Kv2.1 (Swartz and MacKinnon, 1997a). And finally, PcTx1 could have a higher affinity to the open and/or the desensitized state than to the closed state, which would be an example for the modulated receptor concept that has been proposed for the interaction of local anesthetics with voltage-gated Na+ channels (Hille, 1977) and of dihydropyridines with L-type Ca2+ channels (Bean, 1984). Our results do not allow us to unequivocally decide between these three possibilities.
On and off rates for PcTx1 were strikingly slow. We determined an off rate Koff of 8.0 x 10–3 s–1. For the same pH, we determined an apparent Kd of 3.7 x 10–9 M (see MATERIALS AND METHODS). Assuming that Kd = Koff/Kon, one can calculate an apparent on rate Kon of 2.2 x 106 M–1 s–1, which is smaller than expected from a simply diffusion-limited process. Similarly slow on and off rates are characteristic for toxins that are structurally related to PcTx1 (Swartz and MacKinnon, 1995; Oswald et al., 2002). Recently, it has been proposed that these slow rates may be due to a membrane-access mechanism for the toxins (Lee and MacKinnon, 2004; Suchyna et al., 2004). According to this hypothesis, these toxins partition into the membrane, leading to their accumulation in high concentrations around the channel (Lee and MacKinnon, 2004; Suchyna et al., 2004). Thus, binding energy between toxin and channel need not be as high as would be expected for an interaction with nanomolar affinity, and slow inhibition kinetics could be due to a slow rate of membrane partitioning. The highly similar structure of PcTx1 to gating modifier toxins (Escoubas et al., 2003) suggests that PcTx1 also partitions into the membrane, providing a clue to the slow kinetics and implying that the toxin-binding site on ASIC1a is close to or within the membrane-spanning parts of the channel protein.
We also show that Ca2+ competes with PcTx1, probably by inhibiting binding of PcTx1 to the channel. Previously, we and others have shown that Ca2+ is crucial for the activation of ASICs (Babini et al., 2002; Immke and McCleskey, 2003; Paukert et al., 2004). Ca2+ competes with H+, reducing the apparent affinity for H+ (Babini et al., 2002; Immke and McCleskey, 2003). Thus, at least part of the effect of PcTx1 on ASIC1a can be explained in terms of a competition with Ca2+. It is likely that the Ca2+ binding site on the channel contains negatively charged side chains of amino acids and that protonation of these side chains leads to Ca2+ unbinding. Hence the Ca2+ binding site may be identical to the "H+ receptor" of ASICs. Moreover, gating modifier toxins that cause a shift of the voltage dependence of voltage-gated channels interact directly with the voltage sensor (Swartz and MacKinnon, 1997b; Ruta and MacKinnon, 2004). Thus, it is attractive to speculate that PcTx1 interacts with the Ca2+/H+ binding site of ASIC1a. In agreement with this hypothesis is the clustering of several positively charged amino acids in the putative interaction domain of PcTx1 (Escoubas et al., 2003) that are likely to interact with negatively charged amino acids on the channel protein. However, ASIC1a can open with PcTx1 bound, whereas Ca2+ probably has to unbind for channel opening (Immke and McCleskey, 2003; Paukert et al., 2004), arguing against an identical binding site for Ca2+ and PcTx1. Moreover and as pointed out previously, the Ca2+ dependence of the off rate of PcTx1 (Fig. 7) suggests that Ca2+ binds to a different site than PcTx1. Finally, it cannot be excluded that Ca2+ exerts its effects via screening of surface charges; the competition between PcTx1 and Ca2+ could then be much similar to the competition between dihydropyridines (the Ca2+ antagonists) and Ca2+ on heart Ca2+ channels (Kass and Krafte, 1987). Irrespective of the relation to the Ca2+ binding site, we anticipate that identification of the PcTX1 binding site will increase our understanding of activation of ASIC1a by H+.
Within the ASIC family, increased apparent ligand affinity may be an inhibitory mechanism that applies to other toxin/channel interactions than only PcTx1/ASIC1a. Recently, a protein toxin from a sea anemone, APETx2, has been identified that inhibits ASIC3 (Diochot et al., 2004); APETx2 does not inhibit ASIC1a. The interaction of ASIC3 with APETx2 is characterized by inhibition during steady state (at pH 7.4), nanomolar affinity, slow kinetics, and unchanged single channel conductance (Diochot et al., 2004). These characteristics suggest that APETx2 may inhibit ASIC3 by an increased apparent H+ affinity.
Understanding the PcTx1/ASIC1a interaction may be of therapeutic value. Very recently, it has been shown that activation of ASIC1a during the acidosis accompanying brain ischemia leads to significant Ca2+ influx that contributes to ischemic cell death in the brain (Xiong et al., 2004). Inhibition of ASIC1a by PcTx1 significantly decreased neuronal death during ischemia (Xiong et al., 2004). Therefore, our results suggest that a modest increase in the apparent affinity of ASIC1a by pharmacological intervention could desensitize ASIC1a, being of potential benefit in patients at risk of stroke.
ACKNOWLEDGMENTS
We thank Drs. M. Pusch and M. Paukert for comments on the manuscript.
This work was supported by a grant of the Deutsche Forschungsgemeinschaft to S. Gründer (GR1771/3-3).
Olaf S. Andersen served as editor.
Submitted: 20 April 2005
Accepted: 26 May 2005
REFERENCES
Alvarez De La Rosa, D., P. Zhang, D. Shao, F. White, and C.M. Canessa. 2002. Functional implications of the localization and activity of acid-sensitive channels in rat peripheral nervous system. Proc. Natl. Acad. Sci. USA. 99:2326–2331.
Babini, E., M. Paukert, H.S. Geisler, and S. Gründer. 2002. Alternative splicing and interaction with di- and polyvalent cations control the dynamic range of acid-sensing ion channel 1 (ASIC1). J. Biol. Chem. 277:41597–41603.
Bean, B.P. 1984. Nitrendipine block of cardiac calcium channels: high-affinity binding to the inactivated state. Proc. Natl. Acad. Sci. USA. 81:6388–6392.
Benson, C.J., S.P. Eckert, and E.W. McCleskey. 1999. Acid-evoked currents in cardiac sensory neurons: a possible mediator of myocardial ischemic sensation. Circ. Res. 84:921–928.
Bssler, E.L., T.J. Ngo-Anh, H.S. Geisler, J.P. Ruppersberg, and S. Gründer. 2001. Molecular and functional characterization of acid-sensing ion channel (ASIC) 1b. J. Biol. Chem. 276:33782–33787.
Catterall, W.A. 1979. Binding of scorpion toxin to receptor sites associated with sodium channels in frog muscle. Correlation of voltage-dependent binding with activation. J. Gen. Physiol. 74:375–391.
Catterall, W.A. 1980. Neurotoxins that act on voltage-sensitive sodium channels in excitable membranes. Annu. Rev. Pharmacol. Toxicol. 20:15–43.
Cestele, S., Y. Qu, J.C. Rogers, H. Rochat, T. Scheuer, and W.A. Catterall. 1998. Voltage sensor-trapping: enhanced activation of sodium channels by -scorpion toxin bound to the S3-S4 loop in domain II. Neuron. 21:919–931.
Chuang, R.S., H. Jaffe, L. Cribbs, E. Perez-Reyes, and K.J. Swartz. 1998. Inhibition of T-type voltage-gated calcium channels by a new scorpion toxin. Nat. Neurosci. 1:668–674.
Diochot, S., A. Baron, L.D. Rash, E. Deval, P. Escoubas, S. Scarzello, M. Salinas, and M. Lazdunski. 2004. A new sea anemone peptide, APETx2, inhibits ASIC3, a major acid-sensitive channel in sensory neurons. EMBO J. 23:1516–1525.
England, L.J., J. Imperial, R. Jacobsen, A.G. Craig, J. Gulyas, M. Akhtar, J. Rivier, D. Julius, and B.M. Olivera. 1998. Inactivation of a serotonin-gated ion channel by a polypeptide toxin from marine snails. Science. 281:575–578.
Escoubas, P., J.R. De Weille, A. Lecoq, S. Diochot, R. Waldmann, G. Champigny, D. Moinier, A. Menez, and M. Lazdunski. 2000. Isolation of a tarantula toxin specific for a class of proton-gated Na+ channels. J. Biol. Chem. 275:25116–25121.
Escoubas, P., C. Bernard, G. Lambeau, M. Lazdunski, and H. Darbon. 2003. Recombinant production and solution structure of PcTx1, the specific peptide inhibitor of ASIC1a proton-gated cation channels. Protein Sci. 12:1332–1343.
Hammerland, L.G., B.M. Olivera, and D. Yoshikami. 1992. Conantokin-G selectively inhibits N-methyl-D-aspartate-induced currents in Xenopus oocytes injected with mouse brain mRNA. Eur. J. Pharmacol. 226:239–244.
Hille, B. 1977. Local anesthetics: hydrophilic and hydrophobic pathways for the drug–receptor reaction. J. Gen. Physiol. 69:497–515.
Hille, B. 2001. Ion Channels of Excitable Membranes. Sinauer Associates, Sunderland, MA. 722 pp.
Immke, D.C., and E.W. McCleskey. 2003. Protons open acid-sensing ion channels by catalyzing relief of Ca2+ blockade. Neuron. 37:75–84.
Kass, R.S., and D.S. Krafte. 1987. Negative surface charge density near heart calcium channels. Relevance to block by dihydropyridines. J. Gen. Physiol. 89:629–644.
Lee, S.Y., and R. MacKinnon. 2004. A membrane-access mechanism of ion channel inhibition by voltage sensor toxins from spider venom. Nature. 430:232–235.
McDonough, S.I., R.A. Lampe, R.A. Keith, and B.P. Bean. 1997a. Voltage-dependent inhibition of N- and P-type calcium channels by the peptide toxin omega-grammotoxin-SIA. Mol. Pharmacol. 52:1095–1104.
McDonough, S.I., I.M. Mintz, and B.P. Bean. 1997b. Alteration of P-type calcium channel gating by the spider toxin -Aga-IVA. Biophys. J. 72:2117–2128.
McIntosh, J.M., A.D. Santos, and B.M. Olivera. 1999. Conus peptides targeted to specific nicotinic acetylcholine receptor subtypes. Annu. Rev. Biochem. 68:59–88.
Oswald, R.E., T.M. Suchyna, R. McFeeters, P. Gottlieb, and F. Sachs. 2002. Solution structure of peptide toxins that block mechanosensitive ion channels. J. Biol. Chem. 277:34443–34450.
Paukert, M., E. Babini, M. Pusch, and S. Gründer. 2004. Identification of the Ca2+ blocking site of acid-sensing ion channel (ASIC) 1: implications for channel gating. J. Gen. Physiol. 124:383–394.
Ruta, V., and R. MacKinnon. 2004. Localization of the voltage-sensor toxin receptor on KvAP. Biochemistry. 43:10071–10079.
Suchyna, T.M., S.E. Tape, R.E. Koeppe II, O.S. Andersen, F. Sachs, and P.A. Gottlieb. 2004. Bilayer-dependent inhibition of mechanosensitive channels by neuroactive peptide enantiomers. Nature. 430:235–240.
Swartz, K.J., and R. MacKinnon. 1995. An inhibitor of the Kv2.1 potassium channel isolated from the venom of a Chilean tarantula. Neuron. 15:941–949.
Swartz, K.J., and R. MacKinnon. 1997a. Hanatoxin modifies the gating of a voltage-dependent K+ channel through multiple binding sites. Neuron. 18:665–673.
Swartz, K.J., and R. MacKinnon. 1997b. Mapping the receptor site for hanatoxin, a gating modifier of voltage-dependent K+ channels. Neuron. 18:675–682.
Terlau, H., and B.M. Olivera. 2004. Conus venoms: a rich source of novel ion channel-targeted peptides. Physiol. Rev. 84:41–68.
Waldmann, R., and M. Lazdunski. 1998. H+-gated cation channels: neuronal acid sensors in the NaC/DEG family of ion channels. Curr. Opin. Neurobiol. 8:418–424.
Waldmann, R., G. Champigny, F. Bassilana, C. Heurteaux, and M. Lazdunski. 1997. A proton-gated cation channel involved in acid-sensing. Nature. 386:173–177.
Wang, G.K., and G. Strichartz. 1985. Kinetic analysis of the action of Leiurus scorpion -toxin on ionic currents in myelinated nerve. J. Gen. Physiol. 86:739–762.
Xiong, Z.G., X.M. Zhu, X.P. Chu, M. Minami, J. Hey, W.L. Wei, J.F. MacDonald, J.A. Wemmie, M.P. Price, M.J. Welsh, and R.P. Simon. 2004. Neuroprotection in ischemia: blocking calcium-permeable acid-sensing ion channels. Cell. 118:687–698.(Xuanmao Chen,, Hubert Kal)