Current Biology 24, 473–483, March 3, 2014 ª2014 Elsevier Ltd All rights reserved

http://dx.doi.org/10.1016/j.cub.2014.01.013

Article A Tarantula-Venom Peptide Antagonizes the TRPA1 Nociceptor Ion Channel by Binding to the S1–S4 Gating Domain Junhong Gui,1,2,3,9 Boyi Liu,4,9 Guan Cao,1,2,3 Andrew M. Lipchik,5 Minervo Perez,5 Zoltan Dekan,6 Mehdi Mobli,7 Norelle L. Daly,8 Paul F. Alewood,6 Laurie L. Parker,5 Glenn F. King,6 Yufeng Zhou,1,10 Sven-Eric Jordt,4 and Michael N. Nitabach1,2,3,* 1Department of Cellular and Molecular Physiology 2Department of Genetics 3Program in Cellular Neuroscience, Neurodegeneration, and Repair 4Department of Pharmacology Yale School of Medicine, New Haven, CT 06520, USA 5Department of Medicinal Chemistry and Molecular Pharmacology, College of Pharmacy, Purdue University, West Lafayette, IN 47907, USA 6Institute for Molecular Bioscience 7Centre for Advanced Imaging The University of Queensland, Brisbane, QLD 4072, Australia 8Centre for Biodiscovery and Molecular Development of Therapeutics, Queensland Tropical Health Alliance, James Cook University, Cairns, QLD 4870, Australia

Summary Background: The venoms of predators have been an excellent source of diverse highly specific peptides targeting ion channels. Here we describe the first known peptide antagonist of the nociceptor ion channel transient receptor potential ankyrin 1 (TRPA1). Results: We constructed a recombinant cDNA library encoding w100 diverse GPI-anchored peptide toxins (t-toxins) derived from spider venoms and screened this library by coexpression in Xenopus oocytes with TRPA1. This screen resulted in identification of protoxin-I (ProTx-I), a 35-residue peptide from the venom of the Peruvian green-velvet tarantula, Thrixopelma pruriens, as the first known high-affinity peptide TRPA1 antagonist. ProTx-I was previously identified as an antagonist of voltage-gated sodium (NaV) channels. We constructed a t-toxin library of ProTx-I alanine-scanning mutants and screened this library against NaV1.2 and TRPA1. This revealed distinct partially overlapping surfaces of ProTx-I by which it binds to these two ion channels. Importantly, this mutagenesis yielded two novel ProTx-I variants that are only active against either TRPA1or NaV1.2. By testing its activity against chimeric channels, we identified the extracellular loops of the TRPA1 S1–S4 gating domain as the ProTx-I binding site. Conclusions: These studies establish our approach, which we term ‘‘toxineering,’’ as a generally applicable method for isolation of novel ion channel modifiers and design of ion channel modifiers with altered specificity. They also suggest that ProTx-I will be a valuable pharmacological reagent for addressing biophysical mechanisms of TRPA1 gating and the

9These

authors contributed equally to this work address: Department of Physiology, University of Pennsylvania, Philadelphia, PA 19104, USA *Correspondence: [email protected]

10Present

physiology of TRPA1 function in nociceptors, as well as for potential clinical application in the context of pain and inflammation. Introduction Transient receptor potential (TRP) channels are diverse sixtransmembrane domain (6-TM) cation channels that are related to, and share, the transmembrane topology and functional domain structure of the tetrameric voltage-gated ion channels. The first four transmembrane domains (S1–S4) form the gating domain, while the S5 and S6 segments span the re-entrant pore loop to form the selectivity filter and ion conducting pore [1, 2]. TRP channels are involved in almost all aspects of sensory transduction, including vision, olfaction, mechanosensation, thermosensation, and physiological/pathological pain sensing [3]. Several dozen TRP channels have been identified and are classified into six subfamilies, including the TRP ankyrin family (TRPA), with varying degrees of sequence similarity and permeability to cations [2]. TRPA1 is the only member of the ankyrin subfamily so far identified in mammals, and it contains at least 14 ankyrin repeats in the N-terminal intracellular domain [4]. TRPA1 is activated by a variety of chemical compounds, such as allyl isothiocyanate in mustard oil (MO), allicin in garlic and onion [5, 6], and synthetic drugs such as clotrimazole [7], chlorpromazine [8], and clioquinol [9]. TRPA1 can also be activated by endogenous metabolic products and oxidative stress-derived substances and sensitized through G proteincoupled receptors (GPCRs) [10], cyclic adenosine monophosphate/protein kinase A (cAMP/PKA) [11], and protease activated receptor 2 (PAR2) [12] signaling pathways. In mice, TRPA1 is highly expressed in unmyelinated and thinly myelinated sensory neurons of dorsal root ganglion (DRG), nodose ganglion (NG), and trigeminal ganglion (TG) neurons [4, 13], and expression in the urethra, urinary bladder, prostate gland, and arteries has also been observed [14–18]. TRPA1 is coexpressed with TRPV1 in many small-diameter nociceptor sensory neurons that contain both substance P and calcitonin gene-related peptide [4]. TRPA1 knockout mice exhibit pronounced deficits in bradykinin-evoked nociceptor excitation and pain hypersensitivity [10]. TRPA1 is upregulated in inflammatory injury and neuropathic pain, and perturbation of TRPA1 activity can protect against cold-induced hyperalgesia [19–24]. The central role for TRPA1 in pain sensation is starkly highlighted by a gain-of-function TRPA1 mutation in human patients suffering from familial episodic pain syndrome, in which patients suffer debilitating pain on fasting and physical stress [25]. The importance of TRPA1 in pain signaling makes it a potential target for the treatment of pathological pain [26]. While some small-molecule TRPA1 antagonists have been identified [26, 27], no peptide antagonists of TRPA1 have been reported. The venoms of predators, such as spiders, scorpions, cone snails, sea anemones, and snakes, have been an excellent source of peptide diversity for drug discovery and as pharmacological tools for elucidating the structure, function, and physiological properties of ion channels [28, 29]. For example,

Current Biology Vol 24 No 5 474

hanatoxin, a 35-residue peptide from Chilean rose tarantula venom, has been invaluable for probing the structure and function of K+ channels [30, 31]. a-Bungarotoxin, a 74-residue peptide from elapid snakes, has been an essential tool in the study of nicotinic acetylcholine receptors [32]. A peptide toxin from cone snail targeting CaV2.2 channels has already been developed into a Food and Drug Administration (FDA)-approved treatment for intractable pain in human patients [33]. A key issue in ion channel toxin biology has been the mining of this great pharmacological diversity—e.g., spiders alone are estimated to collectively contain millions of distinct peptide toxins in their venoms [28]—to identify toxins with high affinity for particular ion channel targets. Here we report the development and validation of a novel peptide toxin screening platform utilizing the previously described ‘‘tethered-toxin’’ (t-toxin) recombinant expression method [34, 35] and its use to identify the first known peptide antagonist of TRPA1. We constructed a cDNA library containing w100 t-toxins derived from published and unpublished spider toxin sequences (Table S1 available online), each encoded in glycosyl phosphatidyl inositol (GPI)-anchored membrane-tethered form. By screening this library through functional coexpression with cloned ion channel subunits, we identified protoxin-I (ProTx-I), a 35-residue peptide from the venom of the Peruvian greenvelvet tarantula, Thrixopelma pruriens, previously identified as an antagonist of voltage-gated sodium (NaV) channels [36, 37], as a high-affinity TRPA1 antagonist. We found that ProTx-I inhibits both NaV and TRPA1 by binding to the S1–S4 gating domains of these distantly related 6-TM channels. These studies establish screening of t-toxin libraries of native and mutated toxins, which we term ‘‘toxineering,’’ as a generally applicable method for isolation of novel ion channel modifiers and for design of ion channel modifiers with altered target selectivity. They also suggest that ProTx-I will be a valuable pharmacological reagent for addressing the biophysical mechanisms of TRPA1 gating and the physiology and pathophysiology of TRPA1 function in nociceptors, as well as for potential clinical application in the context of pain and inflammation. Results Screen of a Recombinant Membrane-Tethered Spider Toxin Library for TRPA1 Antagonists It has recently been demonstrated that peptide ion channel toxins from venomous predators can be heterologously expressed as fusion proteins covalently tethered to GPI anchors inserted in the extracellular leaflet of the plasma membrane [34, 35]. Various cysteine-rich toxins from snakes, cone snails, and spiders have been expressed in t-toxin form, where they exhibit identical pharmacological specificity to the native toxins [34, 35]. We generated a t-toxin library of approximately 100 cysteine-rich toxin sequences, a combination of both previously published and unpublished sequences from an ongoing large-scale spider venom-gland transcriptomics effort at the University of Queensland. Toxin sequences were cloned into a t-toxin backbone plasmid comprising a trypsin secretory signal sequence, glycine-asparagine repeat linker with embedded c-Myc epitope tag, and GPI targeting sequence from the mammalian Lynx1 peptide, a modulator of the nicotinic acetylcholine receptor (nAChR) (Figure 1A) [34, 35]. The plasmid contains an upstream bacteriophage RNA polymerase promoter for cRNA synthesis. cRNAs of each of the t-toxins were synthesized in vitro and randomly

combined into 14 pools of six or seven toxins each. Xenopus oocytes were coinjected with each t-toxin pool and TRPA1 cRNA (Figure 1B). One pool of toxins significantly suppressed the inward TRPA1 current induced by 100 mM mustard oil (MO) (Figure S1A). Each t-toxin in the positive pool was then tested individually by coinjection with TRPA1 cRNA (Figure S1B). This enabled identification of ProTx-I, a spider toxin previously shown to block several different voltage-gated ion channels [36–38], as a TRPA1 antagonist (Figure 1C). Soluble ProTx-I Is a High-Affinity TRPA1 Antagonist In order to confirm that the observed activity of t-ProTx-I against TRPA1 is not an artifactual consequence of its GPI membrane-tethered configuration, we measured the activity of chemically synthesized soluble ProTx-I against TRPA1. We expressed TRPA1 in HEK293 cells and measured MOinduced currents with perforated whole-cell patch-clamp electrophysiology. Inhibitory activity was defined as Itoxin / IRR, where Itoxin is the current inhibited by bath-applied ProTx-I and IRR is the current inhibited by ruthenium red (RR), a nonspecific TRP channel pore blocker. As shown in Figures 2A and 2B, 1 mM soluble ProTx-I inhibits MO-induced currents by 63%. Dose-response analysis of TRPA1 antagonism by soluble ProTx-1 reveals maximum inhibition of 90.9% 6 2.3%, and IC50 of 389 6 77 nM (Figure 2C). The binding of ProTx-I to TRPA1 is reversible, as inhibition is completely relieved by washout (Figure 2B). Antagonism of TRPA1 by soluble ProTx-I was further confirmed by imaging of Ca2+ influx, as shown in Figure S3A. We also tested the effect of soluble ProTx-I on TRPV1, a thermosensitive and chemosensitive TRP channel that plays an important role in pain signaling [39]. ProTx-I (1 mM) has no significant effect on TRPV1 currents (ANOVA, p = 0.39; Figure S3B). These results confirm that recombinant expression as a t-toxin faithfully recapitulates the pharmacological activity of native ProTx-I. MO, like endogenous tissue stress and injury signals such as 4-hydroxy-nonenal (produced during oxidative stress), hydrogen peroxide, hydroxyl radicals, and hypochlorite (produced by activated neutrophils), activates TRPA1 via covalent modification of the cysteine-rich intracellular C terminus [40–44]. We also tested whether ProTx-I inhibits TRPA1 currents activated by allosteric modulators that interact with transmembrane domains, and not the intracellular C terminus, such as menthol and carvacrol [41, 45, 46]. ProTx-I inhibits the Ca2+ influx through TRPA1 activated with either menthol or carvacrol (Figures S3C and S3D). We also confirmed that ProTx-I inhibits Ca2+ influx through native TRPA1 in cultured mouse dorsal root ganglion nociceptors (Figure S3E). The relatively modest magnitude of the effect on mouse nociceptor TRPA1 is explained by the weaker potency of ProTx-I against cloned mouse TRPA1 versus human TRPA1, with 3 mM ProTx-I having a smaller effect on mouse TRPA1 than 1 mM ProTx-I on human TRPA1 (Figure S3F). t-ProTx-I Inhibits Mammalian and Insect Voltage-Gated Na+ Channels ProTx-I inhibits some subtypes of voltage-gated Na+, Ca2+, and K+ channels [38, 47]. To test whether membrane-tethered t-ProTx-I behaves similarly, we coexpressed t-ProTx-I with NaV channels or with voltage-independent inward-rectifier Kir4.1, which lacks a voltage-sensor domain (VSD) and only possesses the S5–S6 pore-forming domain. When coexpressed with NaV1.2, t-ProTx-I inhibits w60% of the inward Na+ current (Figures 3A and 3B). When coexpressed with the

Peptide Gating-Domain Antagonist of TRPA1 475

A

B

C

D

Figure 1. Recombinant Membrane-Tethered ICK Toxin Library Screen (A) t-toxins are chimeric fusion proteins with an N-terminal secretory signal sequence, toxin sequence, hydrophilic linker incorporating a c-Myc epitope tag, and a C-terminal GPI membrane-anchor targeting sequence. t-toxins are secreted, but remain covalently linked to the plasma membrane via GPI anchors, where they can bind to target channels that are present on the same cell. (B) Approximately 100 spider toxin sequences were cloned into a plasmid vector backbone containing the other elements of the t-toxin. cRNAs encoding each t-toxin were transcribed in vitro, combined randomly in pools, and then coinjected into Xenopus laevis oocytes with cRNA encoding human TRPA1. MO-induced currents were measured by two-electrode voltage clamp. (C) Average TRPA1 currents induced by 100 mM MO when coexpressed with t-d-ACTX-Hv1a, negative-control toxin that inhibits NaV channel inactivation, (‘‘control’’) or t-ProTx-I, normalized to the average for control. Error bars indicate the mean 6 SEM. n is indicated. (D) Representative examples of recordings from oocytes coexpressing TRPA1 with either t-ProTx-I or negative-control t-d-ACTX-Hv1a toxin as summarized in (C). See also Figure S1 and Table S1.

Drosophila para NaV channel, t-ProTx-I inhibits inward Na+ current completely (Figures 3C and 3D). This suggests that ProTx-I has higher affinity for insect than mammalian Na+ channels, presumably because this toxin has been tuned during the course of spider-venom evolution to target the voltage-gated channels of insect prey. Consistent with the potency of t-ProTx-I at inhibiting para currents, bath-applied soluble ProTx-I completely silences action potential firing in a Drosophila whole-brain electrophysiological preparation (Figure S2). In contrast, t-ProTx-I has no effect on kinetics or amplitude of inward-rectifier K+ current (Figures 3E and 3F). This leads to the hypothesis that ProTx-I binds to the S1–S4 gating domain that is common to ion channels with six TM domains (TRP channels and voltage-gated channels), but

lacking in the inward-rectifier K+ channels that only possess the two pore-spanning TM domains. Voltage- and Time-Dependent Unbinding of ProTx-I from Voltage-Gated Na+ Channel Binding of a-scorpion toxins and ProTx-II, another cysteinerich toxin from the Peruvian green-velvet tarantula, to the VSD of NaV channels can be reversed by sustained membrane depolarization [48–50]. This supports a model in which the toxins dissociate more rapidly from the channel in the activated state than in the closed state, thereby stabilizing the closed conformation [51]. To test whether ProTx-I inhibits NaV channels by a similar mechanism, we imposed depolarizing prepulses (+100 mV) of varying duration, followed by 80 ms

Current Biology Vol 24 No 5 476

A

B

C

Figure 2. ProTx-1 Is a High-Affinity TRPA1 Antagonist (A) Representative current-voltage curves for inhibition by 1 mM ProTx-1 of MO-activated (50 mM) voltage-ramp currents in TRPA1-expressing HEK293 cells measured with perforated-patch whole-cell voltage clamp. (B) ProTx-1 (1 mM) suppresses w60% inward (Vh = 280 mV) and outward (Vh = 80 mV) MO-activated TRPA1 currents. (C) Concentration-response curve for ProTx-1 inhibition of MO-activated inward TRPA1 current for cells held at 280 mV. Percent inhibition of the MOinduced current is normalized to that of inhibition by 10 mM ruthenium red (RR), a nonspecific TRP channel blocker, on the same cell; IC50 = 389 6 77 nM. Each point represents the mean 6 SEM with n = 3–5 cells. See also Figure S3.

at the hyperpolarized holding potential (2100 mV) to allow recovery from fast inactivation, and then a test pulse to +10 mV (Figure 4A). As shown in Figures 4B and 4C, depolarizing prepulses cause unbinding of t-ProTx-I from para coexpressed in Xenopus oocytes, with the amplitude of the unblocked current increasing with the duration of the prepulse. Bath-applied ProTx-I (200 nM) exhibits identical unbinding kinetics, with complete reversal by a 1 s depolarizing prepulse (Figures 4D and 4E). These results indicate that ProTx-I blocks voltage-gated ion channel currents by dissociating more slowly from, and thereby stabilizing, the closed conformation of the activation voltage gate. Moreover, they establish that GPI tethering has no effect on the mechanism of channel binding by ProTx-I. Identification of Channel Binding Surfaces of ProTx-I by Alanine-Scanning of t-ProTx-I In order to identify the surface of ProTx-I that mediates its binding to voltage-gated channels and TRPA1, we generated a library of alanine-scanning mutants of t-ProTx-I, with each non-Cys and non-Ala residue mutated individually to Ala. This alanine-scanning approach has previously been used with chemically synthesized and recombinantly expressed peptide toxins to identify target binding surfaces [52–55]. We screened this library against NaV1.2 and TRPA1 by coexpression in Xenopus oocytes and measurement with twoelectrode voltage clamp of voltage- or MO-induced currents. Detailed results of these recordings with statistical comparisons are shown in Figure S4. Importantly, none of the alanine mutants exhibit reduced t-ProTx-I surface expression, thus ruling out expression or trafficking defects as causes of altered activity (Figure S5). In order to visualize the target binding surfaces, we mapped each residue whose mutation significantly reduces t-ProTx-I inhibitory activity against TRPA1 or NaV1.2 to the high-resolution structure of ProTx-I that we determined using nuclear magnetic resonance spectroscopy (Figure 5). Each of the side chains whose mutation to alanine reduces inhibitory activity against either channel are located on the toxin surface, except for Leu19, which is internal, suggesting that the L19A mutation disrupts ProTx-I folding. The

TRPA1 binding surface (pharmacophore) comprises five side chains, whereas ten residues make up the pharmacophore for NaV1.2. Three residues are in common between the two binding surfaces: Ser22, Trp30, and Phe34. Importantly, mutation of particular residues leads to t-ProTx-I variants that are specific inhibitors of either TRPA1 or NaV1.2, but not the other, such as t-ProTx-I (W5A) and t-ProTx-I (S22A) (Figure 5D). Of the six charged residues of ProTx-I, three participate in channel binding—Arg3, Arg23, and Asp 31—but only with NaV1.2. The involvement of both hydrophobic and charged side chains in the interaction of ProTx-I with NaV1.2 and the larger binding surface are consistent with the substantially higher affinity of ProTx-I for voltage-gated channels (IC50 of 30–90 nM [36]) compared with TRPA1 (389 nM; Figure 2). The substantial overlap of the ProTx-I surfaces that bind to TRPA1 and NaV1.2—three of the five TRPA1-binding side chains are shared between the two pharmacophores—is compatible with the hypothesis that ProTx-I binds to TRPA1 by a similar S1–S4 gating mechanism as it does to voltage-gated channels. ProTx-I Binding to TRPA1 Involves the S1–S2 Extracellular Loop ProTx-I has been previously shown to bind to chimeric KV2.1 containing VSD II or VSD IV from rat NaV1.2 [47]. In combination with the overlapping ProTx-I binding surfaces for NaV1.2 and TRPA1 (Figure 5) and the mechanism of NaV1.2 inhibition (Figure 4), this suggests that ProTx-I could inhibit TRPA1 by binding to the extracellular face of its S1–S4 gating domain. In order to test this hypothesis, we attempted to confer greater sensitivity to ProTx-I on a less-sensitive 6-TM channel by transplant of the TRPA1 S1–S2 extracellular loop. As shown in Figure 6, both soluble and membrane-tethered ProTx-I only slightly inhibit the inward currents of NaChBac bacterial NaV channel expressed in oocytes [56]. In contrast, chimeric NaChBac with the S1–S2 linker of TRPA1 is substantially inhibited by ProTx-I, either membrane tethered or soluble (Figure 6). Similar to NaChBac expressed in Xenopus oocytes, bath-applied soluble ProTx-I suppresses the currents of the

Peptide Gating-Domain Antagonist of TRPA1 477

A

B

C

D

E

F

Figure 3. t-ProTx-I Specifically Inhibits 6-TM Ion Channels (A) When coexpressed with NaV1.2, t-ProTx-I inhibits peak depolarization-induced inward currents by w65% compared with t-PLTX-II, a negative-control toxin specific for CaV channels. Error bars indicate the mean 6 SEM. (B) Representative recording of inward currents induced by a series of increasing depolarizations from an oocyte coexpressing NaV1.2 either with t-PLTX-II negative-control toxin or t-ProTx-I. (C) t-ProTx-I completely inhibits depolarization-induced inward currents of the Drosophila para NaV channel. Error bars indicate the mean 6 SEM. (D) Representative recording of inward currents from oocytes coexpressing para with t-PLTX-II or t-ProTx-I. (E) Current-voltage relationship showing that t-ProTx-I has no effect on currents flowing through Kir4.1, an inward-rectifier K+ channel. Points indicate the mean 6 SEM. (F) Representative recording of currents induced by a series of increasing depolarizations of oocytes coexpressing Kir4.1 with either t-ProTx-I or t-d-ACTXHv1a, a negative-control toxin that inhibits NaV inactivation. See also Figure S2.

chimera expressed in HEK293 cells substantially more than wild-type NaChBac (Figures 6E and 6F). Interestingly, the S1–S2 loop chimera NaChBac channel expressed in oocytes appears more sensitive to ProTx-I than when expressed in HEK293 cells, as 200 nM ProTx-I suppresses oocyte currents more than 3 mM ProTx-I suppresses HEK293 currents. This greater antagonism of ProTx-I for the S1–S2 loop chimera NaChBac supports the hypothesis that it inhibits TRPA1 by the same mechanism as voltage-gated channels: slower dissociation and stabilization of the closed conformation of the S1–S4 gating domain.

Discussion We have developed and validated a generally applicable platform for toxin discovery and functional analysis and have identified ProTx-I as the first known peptide TRPA1 antagonist by using this novel recombinant t-toxin library screening approach. ProTx-I will be a valuable pharmacological reagent for addressing the biophysical mechanisms of TRPA1 gating and the physiology and pathophysiology of TRPA1 function in nociceptors, as well as for potential clinical application in the context of pain and inflammation. We also generated and

Current Biology Vol 24 No 5 478

A

B

D

C

E

Figure 4. ProTx-I and t-ProTx-I Bind to and Stabilize Voltage-Gated Channel Activation Gate (A) Voltage-clamp protocol starts with depolarizing prepulse step to +100 mV of varying duration ranging from 0 to 1,000 ms, 80 ms return to holding potential of 2100 mV to allow recovery from inactivation, followed by a depolarizing test pulse to 210 mV to activate the channel. (B) Inward currents recorded from a representative oocyte coexpressing para and t-ProTx-I increase in amplitude as the duration of the depolarizing prepulse increases. (C) Bar graph shows the dependence of peak current during the test pulse on the duration of the prepulse. Test pulse currents are normalized to that for the 1,000 ms prepulse. Error bars indicate the mean 6 SEM. (D) Inward currents as in (B), but for an oocyte expressing only para and in the presence of w200 nM bath-applied soluble ProTx-I. (E) Bar graph as in (C) for oocytes expressing only para and in the presence of w200 nM bath-applied soluble ProTx-I. Error bars indicate the mean 6 SEM.

screened a t-toxin library of alanine-scanning mutants of ProTx-I and thereby identified the partially overlapping surfaces of ProTx-I by which it binds to TRPA1 and NaV1.2 (Figure 5). Importantly, this mutagenesis has yielded novel ProTx-I variants that are only active against either TRPA1 (W5A) or NaV1.2 (S22A) (Figure 5D). On the basis of these identified binding surfaces, further directed mutagenesis and screening of tethered ProTx-I libraries should be a feasible means to obtain variants with even higher affinity and specificity for TRPA1. Comparison of the NaV1.2 binding surface of ProTx-I that we identified here with the NaV1.5 binding surface of ProTx-II previously identified by chemical synthesis of alanine mutants [52] reveals that hydrophobic and electrostatic interactions are important for both ProTx-I and ProTx-II binding to voltage-gated Na+ channels. All tryptophans in both toxins contribute to interactions with NaV channels (Figure S6). For

example, Trp5, Trp27, and Trp30, whose mutation to alanine completely abolishes ProTx-I inhibition of NaV1.2, are conserved in ProTx-II, and their mutation to alanine reduces affinity of ProTx-II for NaV1.5 [52]. Interestingly, with the exception of Trp27, all residues that are important for ProTx-I binding toTRPA1 are not conserved in ProTx-II, which may explain why ProTx-I, but not ProTx-II, binds to and inhibits TRPA1 (t-ProTxII was in the same screening pool as t-ProTx-I and when subscreened on its own had no activity against TRPA1), despite the fact that both toxins inhibit NaV channels [36, 37, 52]. We also compared our ProTx-I alanine-scanning results with those obtained for the related toxin SGTx1 obtained by chemical synthesis of alanine mutants. SGTx1, isolated from the venom of the African featherleg baboon spider Stromatopelma calceatum griseipes, inhibits the KV2.1 voltage-gated K+ channel by binding to residues in the S3b and S4 helices of the VSD [55]. ProTx-I and SGTx1 have roughly 50% sequence identity

Peptide Gating-Domain Antagonist of TRPA1 479

Figure 5. NaV1.2 and TRPA1 Binding Surfaces of ProTx-I Mapped by t-ProTx-I Alanine Scanning (A and B) A library of t-ProTx-I alanine-scanning mutants was screened by coexpression with TRPA1 or NaV1.2 in Xenopus oocytes, with MO- or voltageinduced currents measured by two-electrode voltage clamp. Amino acids exhibiting significantly decreased inhibition of TRPA1 (A) or NaV1.2 (B) when mutated to alanine are colored yellow on the ProTx-I structure. Structures are shown in two orientations, rotated by 180 . (C) Binding surfaces for both TRPA1 and NaV1.2 are mapped onto the ProTx-I structure. Orange indicates amino residues involved in binding toTRPA1, red indicates amino acid residues involved in binding to NaV1.2, and green indicates amino acid residues involved in binding to both ion channels. (D) Representative two-electrode voltage-clamp recordings of Xenopus oocytes coexpressing either NaV1.2 or TRPA1 with negative-control t-toxin (t-PLTX for NaV1.2 and t-d-ACTX-Hv1a for TRPA1), wild-type t-ProTx-I, or the indicated alanine mutants of t-ProTx-I. t-ProTx-I (W5A) is strongly active against TRPA1, but has very little activity against NaV1.2. Conversely, t-ProTx-I (S22A) is strongly active against NaV1.2, but has very little activity against TRPA1. See also Figures S4–S6 and Table S5.

Current Biology Vol 24 No 5 480

A

B

C

D

E

F

Figure 6. ProTx-I Binding to TRPA1 Involves the S1–S2 Extracellular Loop (A) When coexpressed in Xenopus oocytes with a chimera of NaChBac containing the extracellular S1–S2 loop of TRPA1, t-ProTx-I inhibits peak depolarization-induced inward currents by w80% compared with t-L19A, a mutant ProTx-I that fails to inhibit either voltage-gated sodium channels or TRPA1. In contrast, t-ProTx-I has no significant effect on inward currents of wild-type (WT) NaChBac expressed in oocytes. Error bars indicate the mean 6 SEM. (B) Representative recording of inward currents from an oocyte coexpressing either WT or chimeric NaChBac with either t-ProTx-I or t-L19A. (C) Bath-applied ProTx-I (w200 nM) suppresses w50% of peak depolarization-induced inward currents of the chimeric NaChBac channel, but only w20% of the WT NaChBac. Error bars indicate the mean 6 SEM. (D) Representative recordings of inward currents for an oocyte expressing either WT or chimeric NaChBac with or without the presence of bath-applied soluble ProTx-I. (E) ProTx-I (3 mM) inhibits peak depolarization-induced inward currents of chimeric NaChBac by w30%, but only w10% of WT NaChBac expressed in HEK293 cells. Error bars indicate the mean 6 SEM. (F) Representative recordings of the effects of 3 mM ProTx-I on NaChBac WT and chimeric NaChBac inward currents from HEK293 cells.

Peptide Gating-Domain Antagonist of TRPA1 481

(Figure S6). Leu19 in ProTx-I, which is buried inside the toxin as shown in Figure 5 and is possibly involved in ProTx-I folding, is conserved in SGTx1. In the solution structure of SGTx1, the aliphatic side chain of Leu19 is also largely buried inside the toxin, resulting in misfolding of the L19A mutant [55]. As shown in Figure 5, the NaV1.2 binding surface of ProTx-I involves a hydrophobic protrusion comprising the side chains of Trp5, Trp30, and Val29, surrounded by nonhydrophobic residues except Trp27 and Phe34, including Arg3, Ser22, Arg23, Asp31, and Gly32. The topological disposition of these residues is quite similar to those in the KV2.1 binding surface of SGTx1 [55]. These comparisons to related toxin-channel binding interactions strongly suggest that screening of alanine-scanning t-toxin libraries faithfully reveals the target binding surfaces of native soluble toxins. It has been reported that ProTx-I shifts NaV channel activation to more depolarized potentials, but has no effects on inactivation [36, 37]. More recently, ProTx-I was shown to bind to chimeric KV2.1 containing VSD II or VSD IV from rat NaV1.2 [47]. Here, we found that ProTx-I impedes NaV channel activation by binding to and stabilizing the voltage sensor in its closed conformation (Figure 4). Taken together, these data suggest that ProTx-I inhibits NaV1.2 by binding to one of the VSDs, most likely the extracellular surfaces of VSDs in channel domains II or IV. Furthermore, the partial overlap of the TRPA1 and NaV1.2 binding surfaces of ProTx-I and the lack of inhibition of 2-TM inward-rectifer Kir4.1 suggest that ProTx-I binds to TRPA1 by a similar S1–S4 gating mechanism as it does to voltage-gated channels. Consistent with this hypothesis, ProTx-I inhibits chimeric NaChBac bacterial channel with the transplanted S1–S2 loop of TRPA1 more potently than wildtype NaChBac (Figure 6). However, the extent of ProTx-I inhibition of the chimeric NaChBac is less than of TRPA1 itself, which suggests that ProTx-I binding to TRPA1 also involves other parts of the S1–S4 extracellular surfaces (several other NaChBac chimeras were also generated but failed to produce stable currents). Regardless, these results indicate a common conformational gating mechanism between TRP and voltagegated channels and establish ProTx-I as a valuable tool for further detailed biophysical analysis of TRPA1 gating mechanisms. Compared to traditional biochemical fractionation approaches, there are several advantages of the toxineering strategy. First, this method allows screening of large numbers of toxins in equimolar fractions, unlike the highly skewed abundances of toxins in native venoms [57]. Less-abundant toxins may be present in too small a quantity in native venom to detect their activity, or their activity may be masked by more abundant ones. Second, toxineering avoids the need for obtaining native venom or chemically synthesized and in vitro folded toxins. Although various cysteine-rich peptide toxins have been successfully generated in large quantities using either chemical synthesis or recombinant expression in bacteria (e.g., [58, 59]), appropriate conditions for synthesis, expression, and folding of each toxin have to be determined individually due to distinct chemical and structural characteristics, which is expensive and time consuming. Generation of complex t-toxin libraries for screening requires nothing more than diverse toxin sequences—which are now being generated in huge quantities as a result of spider venom-gland transcriptomics efforts [60]— and standard gene synthesis methods. This approach is thus amenable to high-throughput scale up. Third, the toxineering approach enables peptide toxin engineering and screening without the need for chemical synthesis and in vitro folding of

large numbers of variant peptides. Indeed, based as it is on site-directed mutagenesis of cDNA, it would be straightforward to generate large-scale mutant t-toxin libraries and screen them for variants with improved affinity and/or specificity for desired targets. Toxineering thus provides a functional screening alternative to the recently reported purely bindingbased phage display method [61]. Experimental Procedures t-Toxin Design and Library Construction All t-toxin sequences were generated by replacement of the sequence encoding mammalian Lynx1, a toxin-like nicotinic acetylcholine receptor modulator, by the cDNAs of toxins in frame between the secretion signal and the lynx1 hydrophobic sequence for GPI attachment [34, 35]. A flexible linker containing a glycine-asparagine repeat was inserted between the toxin and the hydrophobic sequence for GPI attachment, and a c-Myc epitope tag was introduced in the middle of the linker. t-toxin cDNAs were cloned into the pCS2+ plasmid vector for in vitro transcription of t-toxin cRNA. Loop NaChBac Chimera Design A chimeric NaChBac channel was generated by substitution of the NaChBac S1–S2 extracellular loop (ETYPRIYADHKWL, residues 43–55, accession number NP_242367 in PubMed) with the S1–S2 extracellular loop of TRPA1 (KPGMAFNSTGIINETSDHSEILDTTNSYLIKT, residues 741–772, accession number O75762 in UniProt). Two-Electrode Voltage Clamp The gene constructs for cloning the Drosophila voltage-gated Na+ channel a subunit (para) and its auxiliary b subunit (tipE), pGH19-13-5 para [62] and pGH tipE, were from M. Williamson (Rothamsted Research). Capped cRNAs were prepared by restriction enzyme linearization, followed by in vitro transcription reaction with SP6 or T7 RNA polymerase (mMessenger mMachine kit, Ambion). Concentrations of cRNAs were measured by NanoDrop 2000 (Thermo Scientific), and all toxin cRNAs were diluted to 1 mg/ml. cRNAs of toxins and ion channels of interest were mixed at 1:1 ratio (v/v) and injected into the oocytes. For details of standard two-electrode voltage-clamp protocol, see the Supplemental Experimental Procedures. HEK293 Cell Culture and Electrophysiology Human embryonic kidney (HEK) 293 cells (ATCC, CRL-1573) were cultured in Dulbecco’s modified Eagle’s medium (Lonza) supplemented with 10% fetal bovine serum (Lonza), 2 mM L-glutamine, 100 units/ml penicillin, and 100 mg/ml streptomycin. Cells were transfected by lipofectamine (Invitrogen). For details of the standard whole-cell voltage-clamp protocol, see the Supplemental Experimental Procedures. Statistical Analysis All quantitative results are presented as mean 6 SEM. Statistical analyses were performed using unpaired Student’s t test or ANOVA with paired comparisons with significance being concluded for p < 0.05. Supplemental Information Supplemental Information includes Supplemental Experimental Procedures, six figures, and two tables and can be found with this article online at http://dx.doi.org/10.1016/j.cub.2014.01.013. Acknowledgments The authors thank Aiwei Sui for preparing the TRPA1 plasmid and mouse DRG neurons, Nathaniel Heintz for providing the GPI-tether construct, and L. Williamson for para and tipE clones. Work in the laboratory of M.N.N. is funded in part by NIH grants R01NS055035, R01NS056443, R21NS058330, and R01GM098931. Work in the laboratory of S.-E.J. is funded in part by NIH grants R01ES015056 and U01ES015674. G.F.K. acknowledges funding from the Australian Research Council (DP1095728). Received: October 1, 2013 Revised: November 22, 2013 Accepted: January 8, 2014 Published: February 13, 2014

Current Biology Vol 24 No 5 482

References 1. Wu, L.J., Sweet, T.B., and Clapham, D.E. (2010). International Union of Basic and Clinical Pharmacology. LXXVI. Current progress in the mammalian TRP ion channel family. Pharmacol. Rev. 62, 381–404. 2. Montell, C. (2005). The TRP superfamily of cation channels. Sci. STKE 2005, re3. 3. Nilius, B. (2007). TRP channels in disease. Biochim. Biophys. Acta 1772, 805–812. 4. Story, G.M., Peier, A.M., Reeve, A.J., Eid, S.R., Mosbacher, J., Hricik, T.R., Earley, T.J., Hergarden, A.C., Andersson, D.A., Hwang, S.W., et al. (2003). ANKTM1, a TRP-like channel expressed in nociceptive neurons, is activated by cold temperatures. Cell 112, 819–829. 5. Macpherson, L.J., Geierstanger, B.H., Viswanath, V., Bandell, M., Eid, S.R., Hwang, S., and Patapoutian, A. (2005). The pungency of garlic: activation of TRPA1 and TRPV1 in response to allicin. Curr. Biol. 15, 929–934. 6. Bautista, D.M., Movahed, P., Hinman, A., Axelsson, H.E., Sterner, O., Ho¨gesta¨tt, E.D., Julius, D., Jordt, S.E., and Zygmunt, P.M. (2005). Pungent products from garlic activate the sensory ion channel TRPA1. Proc. Natl. Acad. Sci. USA 102, 12248–12252. 7. Meseguer, V., Karashima, Y., Talavera, K., D’Hoedt, D., DonovanRodrı´guez, T., Viana, F., Nilius, B., and Voets, T. (2008). Transient receptor potential channels in sensory neurons are targets of the antimycotic agent clotrimazole. J. Neurosci. 28, 576–586. 8. Hill, K., and Schaefer, M. (2007). TRPA1 is differentially modulated by the amphipathic molecules trinitrophenol and chlorpromazine. J. Biol. Chem. 282, 7145–7153. 9. Andersson, D.A., Gentry, C., Moss, S., and Bevan, S. (2009). Clioquinol and pyrithione activate TRPA1 by increasing intracellular Zn2+. Proc. Natl. Acad. Sci. USA 106, 8374–8379. 10. Bautista, D.M., Jordt, S.E., Nikai, T., Tsuruda, P.R., Read, A.J., Poblete, J., Yamoah, E.N., Basbaum, A.I., and Julius, D. (2006). TRPA1 mediates the inflammatory actions of environmental irritants and proalgesic agents. Cell 124, 1269–1282. 11. Wang, S., Dai, Y., Fukuoka, T., Yamanaka, H., Kobayashi, K., Obata, K., Cui, X., Tominaga, M., and Noguchi, K. (2008). Phospholipase C and protein kinase A mediate bradykinin sensitization of TRPA1: a molecular mechanism of inflammatory pain. Brain 131, 1241–1251. 12. Dai, Y., Wang, S., Tominaga, M., Yamamoto, S., Fukuoka, T., Higashi, T., Kobayashi, K., Obata, K., Yamanaka, H., and Noguchi, K. (2007). Sensitization of TRPA1 by PAR2 contributes to the sensation of inflammatory pain. J. Clin. Invest. 117, 1979–1987. 13. Nagata, K., Duggan, A., Kumar, G., and Garcı´a-An˜overos, J. (2005). Nociceptor and hair cell transducer properties of TRPA1, a channel for pain and hearing. J. Neurosci. 25, 4052–4061. 14. Du, S., Araki, I., Kobayashi, H., Zakoji, H., Sawada, N., and Takeda, M. (2008). Differential expression profile of cold (TRPA1) and cool (TRPM8) receptors in human urogenital organs. Urology 72, 450–455. 15. Streng, T., Axelsson, H.E., Hedlund, P., Andersson, D.A., Jordt, S.E., Bevan, S., Andersson, K.E., Ho¨gesta¨tt, E.D., and Zygmunt, P.M. (2008). Distribution and function of the hydrogen sulfide-sensitive TRPA1 ion channel in rat urinary bladder. Eur. Urol. 53, 391–399. 16. Earley, S., Gonzales, A.L., and Crnich, R. (2009). Endothelium-dependent cerebral artery dilation mediated by TRPA1 and Ca2+-Activated K+ channels. Circ. Res. 104, 987–994. 17. Gratzke, C., Weinhold, P., Reich, O., Seitz, M., Schlenker, B., Stief, C.G., Andersson, K.E., and Hedlund, P. (2010). Transient receptor potential A1 and cannabinoid receptor activity in human normal and hyperplastic prostate: relation to nerves and interstitial cells. Eur. Urol. 57, 902–910. 18. Andrade, E.L., Forner, S., Bento, A.F., Leite, D.F., Dias, M.A., Leal, P.C., Koepp, J., and Calixto, J.B. (2011). TRPA1 receptor modulation attenuates bladder overactivity induced by spinal cord injury. Am. J. Physiol. Renal Physiol. 300, F1223–F1234. 19. Obata, K., Katsura, H., Mizushima, T., Yamanaka, H., Kobayashi, K., Dai, Y., Fukuoka, T., Tokunaga, A., Tominaga, M., and Noguchi, K. (2005). TRPA1 induced in sensory neurons contributes to cold hyperalgesia after inflammation and nerve injury. J. Clin. Invest. 115, 2393–2401. 20. da Costa, D.S., Meotti, F.C., Andrade, E.L., Leal, P.C., Motta, E.M., and Calixto, J.B. (2010). The involvement of the transient receptor potential A1 (TRPA1) in the maintenance of mechanical and cold hyperalgesia in persistent inflammation. Pain 148, 431–437.

21. Petrus, M., Peier, A.M., Bandell, M., Hwang, S.W., Huynh, T., Olney, N., Jegla, T., and Patapoutian, A. (2007). A role of TRPA1 in mechanical hyperalgesia is revealed by pharmacological inhibition. Mol. Pain 3, 40. 22. Patapoutian, A., Tate, S., and Woolf, C.J. (2009). Transient receptor potential channels: targeting pain at the source. Nat. Rev. Drug Discov. 8, 55–68. 23. Eid, S.R., Crown, E.D., Moore, E.L., Liang, H.A., Choong, K.C., Dima, S., Henze, D.A., Kane, S.A., and Urban, M.O. (2008). HC-030031, a TRPA1 selective antagonist, attenuates inflammatory- and neuropathyinduced mechanical hypersensitivity. Mol. Pain 4, 48. 24. Chen, J., Joshi, S.K., DiDomenico, S., Perner, R.J., Mikusa, J.P., Gauvin, D.M., Segreti, J.A., Han, P., Zhang, X.F., Niforatos, W., et al. (2011). Selective blockade of TRPA1 channel attenuates pathological pain without altering noxious cold sensation or body temperature regulation. Pain 152, 1165–1172. 25. Kremeyer, B., Lopera, F., Cox, J.J., Momin, A., Rugiero, F., Marsh, S., Woods, C.G., Jones, N.G., Paterson, K.J., Fricker, F.R., et al. (2010). A gain-of-function mutation in TRPA1 causes familial episodic pain syndrome. Neuron 66, 671–680. 26. Andrade, E.L., Meotti, F.C., and Calixto, J.B. (2012). TRPA1 antagonists as potential analgesic drugs. Pharmacol. Ther. 133, 189–204. 27. Viana, F., and Ferrer-Montiel, A. (2009). TRPA1 modulators in preclinical development. Expert Opin. Ther. Pat. 19, 1787–1799. 28. King, G.F. (2011). Venoms as a platform for human drugs: translating toxins into therapeutics. Expert Opin. Biol. Ther. 11, 1469–1484. 29. Dutertre, S., and Lewis, R.J. (2010). Use of venom peptides to probe ion channel structure and function. J. Biol. Chem. 285, 13315–13320. 30. Swartz, K.J., and MacKinnon, R. (1997). Hanatoxin modifies the gating of a voltage-dependent K+ channel through multiple binding sites. Neuron 18, 665–673. 31. Swartz, K.J., and MacKinnon, R. (1997). Mapping the receptor site for hanatoxin, a gating modifier of voltage-dependent K+ channels. Neuron 18, 675–682. 32. Tsetlin, V., Utkin, Y., and Kasheverov, I. (2009). Polypeptide and peptide toxins, magnifying lenses for binding sites in nicotinic acetylcholine receptors. Biochem. Pharmacol. 78, 720–731. 33. Snutch, T.P. (2005). Targeting chronic and neuropathic pain: the N-type calcium channel comes of age. NeuroRx 2, 662–670. 34. Iban˜ez-Tallon, I., Wen, H., Miwa, J.M., Xing, J., Tekinay, A.B., Ono, F., Brehm, P., and Heintz, N. (2004). Tethering naturally occurring peptide toxins for cell-autonomous modulation of ion channels and receptors in vivo. Neuron 43, 305–311. 35. Wu, Y., Cao, G., Pavlicek, B., Luo, X., and Nitabach, M.N. (2008). Phase coupling of a circadian neuropeptide with rest/activity rhythms detected using a membrane-tethered spider toxin. PLoS Biol. 6, e273. 36. Middleton, R.E., Warren, V.A., Kraus, R.L., Hwang, J.C., Liu, C.J., Dai, G., Brochu, R.M., Kohler, M.G., Gao, Y.D., Garsky, V.M., et al. (2002). Two tarantula peptides inhibit activation of multiple sodium channels. Biochemistry 41, 14734–14747. 37. Priest, B.T., Blumenthal, K.M., Smith, J.J., Warren, V.A., and Smith, M.M. (2007). ProTx-I and ProTx-II: gating modifiers of voltage-gated sodium channels. Toxicon 49, 194–201. 38. Ohkubo, T., Yamazaki, J., and Kitamura, K. (2010). Tarantula toxin ProTx-I differentiates between human T-type voltage-gated Ca2+ Channels Cav3.1 and Cav3.2. J. Pharmacol. Sci. 112, 452–458. 39. Moran, M.M., McAlexander, M.A., Bı´ro´, T., and Szallasi, A. (2011). Transient receptor potential channels as therapeutic targets. Nat. Rev. Drug Discov. 10, 601–620. 40. Andersson, D.A., Gentry, C., Moss, S., and Bevan, S. (2008). Transient receptor potential A1 is a sensory receptor for multiple products of oxidative stress. J. Neurosci. 28, 2485–2494. 41. Bessac, B.F., Sivula, M., von Hehn, C.A., Escalera, J., Cohn, L., and Jordt, S.E. (2008). TRPA1 is a major oxidant sensor in murine airway sensory neurons. J. Clin. Invest. 118, 1899–1910. 42. Hinman, A., Chuang, H.H., Bautista, D.M., and Julius, D. (2006). TRP channel activation by reversible covalent modification. Proc. Natl. Acad. Sci. USA 103, 19564–19568. 43. Macpherson, L.J., Dubin, A.E., Evans, M.J., Marr, F., Schultz, P.G., Cravatt, B.F., and Patapoutian, A. (2007). Noxious compounds activate TRPA1 ion channels through covalent modification of cysteines. Nature 445, 541–545. 44. Trevisani, M., Siemens, J., Materazzi, S., Bautista, D.M., Nassini, R., Campi, B., Imamachi, N., Andre`, E., Patacchini, R., Cottrell, G.S., et al. (2007). 4-Hydroxynonenal, an endogenous aldehyde, causes pain and

Peptide Gating-Domain Antagonist of TRPA1 483

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56. 57.

58.

59.

60.

61.

62.

neurogenic inflammation through activation of the irritant receptor TRPA1. Proc. Natl. Acad. Sci. USA 104, 13519–13524. Escalera, J., von Hehn, C.A., Bessac, B.F., Sivula, M., and Jordt, S.E. (2008). TRPA1 mediates the noxious effects of natural sesquiterpene deterrents. J. Biol. Chem. 283, 24136–24144. Xiao, B., Dubin, A.E., Bursulaya, B., Viswanath, V., Jegla, T.J., and Patapoutian, A. (2008). Identification of transmembrane domain 5 as a critical molecular determinant of menthol sensitivity in mammalian TRPA1 channels. J. Neurosci. 28, 9640–9651. Bosmans, F., Martin-Eauclaire, M.F., and Swartz, K.J. (2008). Deconstructing voltage sensor function and pharmacology in sodium channels. Nature 456, 202–208. Sokolov, S., Kraus, R.L., Scheuer, T., and Catterall, W.A. (2008). Inhibition of sodium channel gating by trapping the domain II voltage sensor with protoxin II. Mol. Pharmacol. 73, 1020–1028. Rogers, J.C., Qu, Y., Tanada, T.N., Scheuer, T., and Catterall, W.A. (1996). Molecular determinants of high affinity binding of alpha-scorpion toxin and sea anemone toxin in the S3-S4 extracellular loop in domain IV of the Na+ channel alpha subunit. J. Biol. Chem. 271, 15950–15962. Catterall, W.A. (1977). Membrane potential-dependent binding of scorpion toxin to the action potential Na+ ionophore. Studies with a toxin derivative prepared by lactoperoxidase-catalyzed iodination. J. Biol. Chem. 252, 8660–8668. Phillips, L.R., Milescu, M., Li-Smerin, Y., Mindell, J.A., Kim, J.I., and Swartz, K.J. (2005). Voltage-sensor activation with a tarantula toxin as cargo. Nature 436, 857–860. Smith, J.J., Cummins, T.R., Alphy, S., and Blumenthal, K.M. (2007). Molecular interactions of the gating modifier toxin ProTx-II with NaV 1.5: implied existence of a novel toxin binding site coupled to activation. J. Biol. Chem. 282, 12687–12697. Tedford, H.W., Gilles, N., Me´nez, A., Doering, C.J., Zamponi, G.W., and King, G.F. (2004). Scanning mutagenesis of omega-atracotoxin-Hv1a reveals a spatially restricted epitope that confers selective activity against insect calcium channels. J. Biol. Chem. 279, 44133–44140. Maggio, F., and King, G.F. (2002). Scanning mutagenesis of a Janusfaced atracotoxin reveals a bipartite surface patch that is essential for neurotoxic function. J. Biol. Chem. 277, 22806–22813. Wang, J.M., Roh, S.H., Kim, S., Lee, C.W., Kim, J.I., and Swartz, K.J. (2004). Molecular surface of tarantula toxins interacting with voltage sensors in K(v) channels. J. Gen. Physiol. 123, 455–467. Ren, D., Navarro, B., Xu, H., Yue, L., Shi, Q., and Clapham, D.E. (2001). A prokaryotic voltage-gated sodium channel. Science 294, 2372–2375. Chen, J., Deng, M., He, Q., Meng, E., Jiang, L., Liao, Z., Rong, M., and Liang, S. (2008). Molecular diversity and evolution of cystine knot toxins of the tarantula Chilobrachys jingzhao. Cell. Mol. Life Sci. 65, 2431– 2444. Fletcher, J.I., Smith, R., O’Donoghue, S.I., Nilges, M., Connor, M., Howden, M.E., Christie, M.J., and King, G.F. (1997). The structure of a novel insecticidal neurotoxin, omega-atracotoxin-HV1, from the venom of an Australian funnel web spider. Nat. Struct. Biol. 4, 559–566. Wang, X., Connor, M., Smith, R., Maciejewski, M.W., Howden, M.E., Nicholson, G.M., Christie, M.J., and King, G.F. (2000). Discovery and characterization of a family of insecticidal neurotoxins with a rare vicinal disulfide bridge. Nat. Struct. Biol. 7, 505–513. Chen, J., Zhao, L., Jiang, L., Meng, E., Zhang, Y., Xiong, X., and Liang, S. (2008). Transcriptome analysis revealed novel possible venom components and cellular processes of the tarantula Chilobrachys jingzhao venom gland. Toxicon 52, 794–806. Takacs, Z., Toups, M., Kollewe, A., Johnson, E., Cuello, L.G., Driessens, G., Biancalana, M., Koide, A., Ponte, C.G., Perozo, E., et al. (2009). A designer ligand specific for Kv1.3 channels from a scorpion neurotoxin-based library. Proc. Natl. Acad. Sci. USA 106, 22211–22216. Warmke, J.W., Reenan, R.A., Wang, P., Qian, S., Arena, J.P., Wang, J., Wunderler, D., Liu, K., Kaczorowski, G.J., Van der Ploeg, L.H., et al. (1997). Functional expression of Drosophila para sodium channels. Modulation by the membrane protein TipE and toxin pharmacology. J. Gen. Physiol. 110, 119–133.