The tarantula toxin β/δ-TRTX-Pre1a highlights the importance of the S1-S2 voltage-sensor region for sodium channel subtype selectivity

Abstract

Voltage-gated sodium (NaV) channels are essential for the transmission of pain signals in humans making them prime targets for the development of new analgesics. Spider venoms are a rich source of peptide modulators useful to study ion channel structure and function. Here we describe β/δ-TRTX-Pre1a, a 35-residue tarantula peptide that selectively interacts with neuronal NaV channels inhibiting peak current of hNaV1.1, rNaV1.2, hNaV1.6, and hNaV1.7 while concurrently inhibiting fast inactivation of hNaV1.1 and rNaV1.3. The DII and DIV S3-S4 loops of NaV channel voltage sensors are important for the interaction of Pre1a with NaV channels but cannot account for its unique subtype selectivity. Through analysis of the binding regions we ascertained that the variability of the S1-S2 loops between NaV channels contributes substantially to the selectivity profile observed for Pre1a, particularly with regards to fast inactivation. A serine residue on the DIV S2 helix was found to be sufficient to explain Pre1a’s potent and selective inhibitory effect on the fast inactivation process of NaV1.1 and 1.3. This work highlights that interactions with both S1-S2 and S3-S4 of NaV channels may be necessary for functional modulation, and that targeting the diverse S1-S2 region within voltage-sensing domains provides an avenue to develop subtype selective tools.

Introduction

Voltage-gated sodium channels (NaVs) are membrane proteins with four homologous domains (DI-DIV), each composed of six transmembrane segments (S1-S6) assembled into a single, functional α-subunit (~260 kDa). Each domain has two functionally distinct regions; the S1-S4 segments comprise the voltage-sensing domain (VSD) whereas the S5-S6 helices and extracellular ‘P-loop’ form the selectivity filter and ion-conducting pore. There are nine mammalian α-subunits (NaV1.1 to NaV1.9) with greater than 64% sequence identity between isoforms 1.1 to 1.7, whereas 1.8 and 1.9 share ~50–60% identity with other members1, 2. The NaV channel family contains important therapeutic targets for local anaesthetics, anti-arrhythmics, analgesics, and anti-epileptics3, 4. For example, the peripherally expressed neuronal NaV1.7 has been identified as a potential target for the treatment of chronic pain, largely from studies of human channelopathies5. Mutations of SCN9A that result in NaV1.7 gain-of-function underlie paroxysmal extreme pain disorder (PEPD) and primary erythromelalgia6, 7, whereas mutations that result in the loss of NaV1.7 function underlie ‘congenital insensitivity to pain’ (CIP)8. The neuronal isoforms NaV1.3, NaV1.8 and NaV1.9 have also been implicated in various forms of chronic and acute pain9,10,11,12, whereas recent evidence suggests that NaV1.1 also plays a role in mechanical pain transmission13. However, the similarity between NaV channel α-subunits poses a challenge to discover and develop molecules that can selectively modify the function of each subtype.

Natural toxins have been instrumental in defining NaV channel subtypes and many of the known functionally relevant binding sites on the channel1. In recent decades, venom peptides have proven to be an invaluable source of novel, selective, and potent modulators of ion channels, leading to an increasing interest in these molecules as pharmacological tools and potential therapeutic leads14. In this respect, spiders possibly represent one of the richest sources of novel voltage-gated channel modulators15 that not only target channel α-subunits, but also complexes containing the accessory β-subunits16. A majority of currently characterised NaV channel modulating spider venom peptides share a common structural motif of the inhibitory cysteine knot (ICK) and interact with one of four VSDs at partially defined receptor sites17, 18.

These VSD modulating peptides have been described to modify channel gating in three distinct ways depending on the domain they target and the effect on that domain17. Domain IV of the NaV channel α-subunit uniquely controls inactivation of the channel19, thus spider peptides that interact with and hinder the normal function of VSDIV slow or inhibit fast inactivation of the channel and may result in persistent current20. Conversely, spider venom peptides that interact with DI, DII, and DIII VSD modulate channel activation21. Most of those discovered to date cause a shift in the voltage-dependence of activation in the depolarising direction and inhibit activation22, 23, whereas some such as β-HXTX-Mg1a (Magi 5) cause a hyperpolarising shift, thereby facilitating activation24. The VSDs of different voltage-gated ion channels exhibit higher sequence variability than the highly conserved pore region, thus offering the opportunity for subtype-selective interactions between NaV subtypes as well as the greater voltage-gated ion channel superfamily. Indeed, certain families of spider venom peptides demonstrate remarkable subtype selectivity profiles13, 25.

A deeper understanding of the molecular basis of the interactions between spider venom peptides with NaV channels is helping efforts to design a new generation of pharmacological tools and potential therapeutic lead molecules26,27,28. To this end, we identified β/δ-TRTX-Pre1a (Pre1a) from the venom of the tarantula Psalmopoeus reduncus in a screen for NaV1.7 inhibitors. Pre1a exhibited a unique and complex pharmacological profile across neuronal NaV channel subtypes where it preferentially inhibits fast inactivation of NaV1.3, inhibits activation of NaV1.2, NaV1.6, and NaV1.7, and inhibits both activation and fast inactivation of NaV1.1, with no effect on NaV1.4 or NaV1.5 at sub-micromolar concentrations. This functional profile is dictated by classical interactions with the DII and DIV S3-S4 loops of the NaV channel VSDs. However, our evidence points to the S1-S2 loops as critical for imparting the isoform selectivity demonstrated by Pre1a. Pre1a thus represents a valuable tool to study the subtle differences in DII and DIV interaction sites between members of the NaV channel family.

Results

Isolation and sequence of TRTX-Pre1a

In a small screen of 14 spider venoms, Psalmopoeus reduncus venom at 1:1000 dilution (~200 μg/ml) consistently and potently inhibited human (h) NaV1.7 expressed in Xenopus laevis oocytes. Activity-guided fractionation resulted in the identification of a major component exhibiting inhibitory activity at hNaV1.7 expressed in oocytes (Fig. 1a). The purified peptide eluted with an unusual, leading minor peak that upon mass analysis revealed the same M + H+ of 4227.5 as the principal peak (Fig. 1b) and was found to inhibit both the of peak current of hNaV1.7 and the inactivation of rat (r) NaV1.3 (Fig. 1c). The mass of the reduced/alkylated peptide was 588 Da higher than the native peptide (6 × 98, the MW of maleimide), indicating the presence of three disulfide bonds (data not shown). Edman analysis of the reduced/alkylated peptide resulted in complete sequence determination of a novel 35 amino acid peptide with high sequence similarity to several NaV modulating spider venom peptides (Fig. 1d), determined by a search of non-redundant protein database (NCBI). Sequences were aligned and percent similarity calculated manually. Taking into account disulfide bonds, the observed monoisotopic M + H+ (4227.5) matched that of the theoretical M + H+ with a free acid C-terminus (4226.9, Expasy-PeptideMass) that was confirmed using MSMS on tryptic digest fragments (data not shown). The peptide was named TRTX-Pre1a (Pre1a) according to the current rational nomenclature29.

Figure 1
figure1

Identification and sequence of TRTX-Pre1a. (a) RP-HPLC chromatogram of crude venom from P. reduncus indicating the fraction (F18) responsible for robust inhibition of hNaV1.7 expressed in Xenopus oocytes (inset). (b) Final analytical RP-HPLC purification step of F18 (TRTX-Pre1a) (inset: MALDI-TOF MS spectrum showing M + H+ of 4227.5). (c) Activity of pure, native Pre1a on rNaV1.3 and hNaV1.7 exressed in Xenopus oocytes demonstrating inhibition of inactivation and peak current, respectively. (d) Sequence alignment of TRTX-Pre1a with NaV modulating Theraphotoxins13, 20, 22, 41, 47,48,49,50, 55, 65,66,67,68. Percent similarity was calculated comparing the number of identical (dark gray) and similar (light gray) amino acids.

In order to carry out functional characterisations, we produced Pre1a using Boc solid-phase peptide synthesis. The folded synthetic peptide co-eluted with the peptide isolated from the native source (Supp. Fig. 1). Analytical RP-HPLC analysis of synthetic Pre1a revealed that the unusual non-symmetrical nature of the eluting peptide is an inherent property of the peptide and suggested that Pre1a is structurally heterogeneous (Fig. 2). This structural heterogeneity was confirmed by collecting the body and trailing portions of the major peak as individual fractions then re-injecting under the same conditions, whereupon the same non-symmetrical chromatographic profile was observed (Fig. 2 insets). This suggests a conformational (chemical) exchange process that is slow on the HPLC time scale.

Figure 2
figure2

Pre1a shows conformational flexibility under RP-HPLC conditions. Analytical RP-HPLC of pure synthetic Pre1a shows the presence of multiple conformers in and acetonitrile/water mixture at room temperature. Insets 1 and 2, demonstrate identical elution profiles for reinjection of two fractions (highlighted and numbered) taken from the major peak, discounting the presence of impurities.

Structural studies of recombinant Pre1a

Preliminary homonuclear NMR analysis of synthetic Pre1a confirmed the presence of conformational heterogeneity, prompting us to isotope label the peptide (using recombinant expression in E. coli, Supp. Fig. 2) to enable less ambiguous heteronculear NMR analysis. Backbone resonance assignments (1HN, 15N, 13Cα, 13Cβ, 13C′) for rPre1a were completed by analysis of the 3D HNCACB, CBCA(CO)NH, and HNCO spectra, and the side chain 1H and 13C chemical shifts from the 3D H(CC) (CO)NH-TOCSY and (H)CC(CO)NH-TOCSY spectra, respectively. Together we achieved 90% completion of all proton assignments. However, due to the structural heterogeneity, insufficient unique constraints were derived from the NOESY spectra to enable the structure calculations using CYANA30 to produce a adequately converged high-resolution structure. However, the chemical shifts determined for the backbone resonances allowed us to predict the secondary structure of Pre1a using the program TALOS+ 31, which suggests that residues around C3–L4 and Y21–K22 are likely to form β-strands (Supp. Fig. 3a).

Furthermore, the NMR data allowed us to confidently determine the disulfide connectivity of the peptide via unambiguous NOEs between sidechain Cβ filtered Hβ protons. NOEs could be observed between the side chain protons of C3 and C18, C10 and C23 and C17 and C30 in the 13C-NOESY (Supp. Fig. 3b), consistent with the formation of the common ICK-motif (C1–C4, C2–C5, C3–C6), which appears to be the dominant structural scaffold among NaV channel modulating spider venom peptides characterised to date.

The 3D NMR data also allowed us to characterise the conformational heterogeneity of Pre1a, first observed on RP-HPLC. The presence of several peptide conformations resulted in multiple backbone 1HN/15N chemical shifts for residues D2–R9 in the 15N-HSQC spectrum, exemplified by G5 (Fig. 3a). These residues form Loop 1 and, with the exception of S8 and R9 (which have duplicate peaks), all have visible triplicate peaks with decreasing intensities (for examples of NMR experiments on related but structurally rigid peptides see Fig. 2 in Lau et al.32 and Fig. S3 in Klint et al.33). Additionally, a homology model of Pre1a was constructed using the structure of HwTx-IV as a template (PDB 2M4X)34 using SWISS-MODEL35. The Pre1a model shows a highly dense packing of five aromatic residues with W6 and F7 at the tip of Loop 1 falling between W29 and Y32/Y21 on one face of the peptide (Fig. 3b). We hypothesise that this overcrowding of bulky aromatic side chains results in the structural flexibility of Pre1a as seen in both aqueous (NMR) and hydrophobic (RP-HPLC) conditions.

Figure 3
figure3

Pre1a shows conformational flexibility in aqueous conditions. (a) 2D 1H-15N-HSQC of recombinantly produced Pre1a. The chemical shifts of resonances for residues in loop 1 (D2–R9, underlined) show multiple peaks indicating the presence of three conformations of the peptide (A = Major, B = middle, C = minor, highlighted for G5 in the inset). sc = side chain NH resonances for N, R and W residues. (b) Two views of a homology model of Pre1a (based on the NMR structure of HwTxIV, PDB: 2M4X) illustrating the relative positions of W6 and F7 at the tip of Loop 1 and W29, Y32 and Y21. Loop 1 residues (that show multiple peaks in the HSQC above) are in red, disulfide bonds are in yellow. The right panel highlights the position of W6 and F7 at the tip of Loop 1 and the positions of C3 and R9, which may act as a hinge region for movement of the loop.

Pre1a affects neuronal NaV channels in a subtype-dependent manner

Pre1a produced by either chemical synthesis or recombinant expression was equipotent at inhibiting the peak current of hNaV1.7 (Supp. Fig. 4a). β-subunits can have profound effects on venom peptide interactions with NaV channels36, however Pre1a’s inhibitory effect on hNaV1.7 was not modified by the accessory β1 subunit (Supp. Fig. 4b). To determine the subtype selectivity profile of recombinant Pre1a it was tested on five different NaV channel α-subunits expressed in X. laevis oocytes (Fig. 4a). Pre1a concentration-dependently inhibited the peak inward current of rNaV1.2 and hNaV1.7 with IC50 values of 189.6 nM (pIC50 6.72 ± 0.06) and 114.0 nM (pIC50 6.94 ± 0.06), respectively (Fig. 4b), and weakly inhibited rNaV1.3 with an IC50 of 8.0 μM (pIC50 5.10 ± 0.05). Similar to its effect on NaV1.3, 1 μM Pre1a only weakly inhibited the peak current amplitude of rNaV1.4 and hNaV1.5 by 16.47 ± 0.05% and 8.60 ± 0.05%, respectively (Fig. 4b).

Figure 4
figure4

Pre1a preferentially inhibits neuronal NaV channels. (a) Representative NaV currents recorded from Xenopus oocytes (α-subunit alone) before addition of 1 µM Pre1a (black) and after reaching steady state inhibition (red). Late current was assessed at 50 ms (100 ms for NaV1.3) from the peak current, as highlighted by the grey box on rNaV1.3. (b) Concentration-effect curves for peak current inhibition by Pre1a for rNaV1.2 (IC50 189.6 nM; n ≥ 5), rNaV1.3 (IC50 8.0 μM; n ≥ 8) and hNaV1.7 (IC50 114.0 nM; n ≥ 8), with single point 1 µM concentrations for rNaV1.4 (n = 6) and hNaV1.5 (n = 7). (c) Concentration-response curve for late current inhibition of inactivation by Pre1a of rNaV1.3 (EC50 45.0 nM). (d) Concentration-dependent effects of Pre1a on the rate of inactivation (τ) for rNaV1.3, with 1 µM demonstrating a significant slowing of inactivation (p < 0.001; n ≥ 5; ANOVA with Dunnet’s test). (e) Representative current traces of hNaV1.1, hNaV1.6, and hNaV1.7 expressed in HEK cells co-expressed with NaVβ1, in the absence (black) and presence (red) of varying Pre1a concentrations. (f) Concentration-response curves for peak current inhibition by Pre1a for hNaV1.1 (IC50 57.1 nM; n ≥ 6), hNaV1.6 (IC50 221.6 nM; n ≥ 6), and hNaV1.7 (IC50 15.0 nM; n ≥ 9) expressed in HEK cells. (g) Concentration-response curves for Pre1a effects on late current (measured at 10 ms from peak) of hNaV1.1 (EC50 41.4 nM), with no measurable effect on hNaV1.6 or hNaV1.7. (h) Concentration-dependent effects of Pre1a on the rate of inactivation (τ) for hNaV1.1, with 30 nM and 300 nM demonstrating a significant slowing of inactivation (p < 0.001; n ≥ 5; ANOVA with Dunnet’s test).

Pre1a demonstrated inhibitory effects on rNaV1.3 fast inactivation (Fig. 4a), thus these effects on NaV1.2, 1.3 and 1.7 were examined in detail. We analysed the rate of fast inactivation (defined as tau, τinact, or the time to reach 36.8% of a single exponential fit of the current decay) and the degree of persistent sodium current measured at 50 ms of depolarisation (NaV1.2 and 1.7) and 100 ms for rNaV1.3 due to the longer time required to reach complete inactivation in the oocyte expression system without β1. Consistent with results observed during assay-guided fractionation, Pre1a potently inhibited the inactivation of rNaV1.3 with an EC50 of 45.5 nM (pEC50 7.34 ± 0.27) (Fig. 4a and c), a selectivity of more than 150-fold for inhibiting rNaV1.3 inactivation over activation. In addition to inhibiting the extent of rNaV1.3 inactivation (as indicated by the level of persistent current), Pre1a also decreased the rate of current inactivation, reflected as a concentration-dependent increase in τinact (Fig. 4d). At 1 μM, a concentration that almost completely inhibited rNaV1.2 and hNaV1.7 currents, Pre1a had no significant effect on the persistent current or τinact of either of these channels. There was no apparent effect of 1 μM Pre1a on the inactivation of either rNaV1.4 or hNaV1.5 (Fig. 4a).

Spider venom peptides similar to Pre1a have been shown to target NaV1.1 and NaV1.622, 28, thus we assessed the activity of Pre1a on hNaV1.1 and hNaV1.6 stably expressed in HEK cells using the Qpatch16 automated patch-clamp platform (Fig. 4e). To compare possible discrepancies in potency between the two different systems, HEK-hNaV1.7 cells were tested in parallel. Pre1a inhibited hNaV1.1, hNaV1.6, and hNaV1.7 with IC50s of 57.1 nM (pIC50 7.24 ± 0.07), 221.6 nM (pIC50 6.65 ± 0.06), and 15.0 nM (pIC50 7.82 ± 0.04), respectively (Fig. 4f). An approximate 7-fold increase in potency was determined for the inhibition of hNaV1.7 peak current as analysed on the QPatch as compared to the IC50 obtained for this channel in oocytes. A difference in potency of some molecules or with some receptors is not uncommon when comparing two-electrode voltage clamp of oocytes to data obtained using whole-cell patch clamping of mammalian cells37,38,39.

In addition to the effect of Pre1a on NaV1.1 peak current, it also equipotently inhibited NaV1.1 fast inactivation with an EC50 of 41.4 nM (pEC50 7.38 ± 0.51) (Fig. 4g). Beyond 1 µM, inhibition of peak current prevented the analysis of inactivation as the current was fully inhibited. Concentrations up to 1 μM Pre1a on hNaV1.6 or up to 300 nM on hNaV1.7 had no effect on inactivation (Fig. 4g). Together, these results demonstrate that Pre1a has a strong preference for modulating neuronal NaV channel isoforms over the skeletal muscle (NaV1.4) and the cardiac (NaV1.5) isoforms. Using the IC50 for inhibition of NaV1.7 peak current as a reference point (to allow comparison across platforms) results in a relative rank order potency for NaV modulation of rNaV1.3inact (0.4) > hNaV1.7act (1) > rNaV1.2act (1.7) > hNaV1.1inact (2.7) > hNaV1.1act (3.8) > hNaV1.6act (15) > rNaV1.3act (70). Thus the effect of Pre1a on NaV1.3 inactivation has the highest relative potency for all NaV activity (peak current or inactivation) on the channels tested.

Pre1a modulates NaV channel function by inhibiting channel gating

The mechanism of action of Pre1a was studied by determining the voltage of half maximal effect (V 1/2 ) of channel gating for rNaV1.2, rNaV1.3 and hNaV1.7 expressed in oocytes (Fig. 5). Pre1a (1 µM) significantly shifted the V 1/2 of rNaV1.2 and hNaV1.7 activation by +14.9 mV and +13.8 mV, respectively (Fig. 5a), whilst having no effect on the V 1/2 of activation of rNaV1.3 peak current (Fig. 5b). Pre1a had no effect on the V 1/2 of inactivation of rNaV1.2 or hNaV1.7 (Fig. 5a) or rNaV1.3, however, it did prevent the current of NaV1.3 from fully inactivating (Fig. 5b). In contrast, 1 μM Pre1a shifted the V 1/2 of activation of the rNaV1.3 late current by +9.6 mV (Fig. 5c).

Figure 5
figure5

Pre1a (1 μM) affects the voltage-dependence of activation of rNaV1.2 and hNaV1.7 and the steady-state inactivation (SSIN) of rNaV1.3. (a) 1 µM Pre1a causes a depolarizing shift in the V 1/2 of activation of rNaV1.2 (control V 1/2  = −19.73 ± 0.85; Pre1a V 1/2  = −4.88 ± 2.06, n = 5) and hNaV1.7 (control V 1/2  = −22.19 ± 0.01, Pre1a V 1/2  = −8.37 ± 0.02, n = 6). 1 µM Pre1a had no significant effect on V 1/2 inactivation of rNaV1.2 (control V 1/2  = −45.23 ± 0.37; Pre1a V 1/2  = −47.92 ± 0.58) and hNaV1.7 (control V 1/2  = −38.42 ± 0.61; Pre1a V 1/2  = −39.97 ± 0.82). (b) Voltage-dependence of activation and SSIN of rNaV1.3 in the absence and presence of 1 µM Pre1a (n = 11). Pre1a had no significant effect on voltage-dependence of peak current activation (control V 1/2  = −15.13 ± 0.65; Pre1a V 1/2  = −14.23 ± 0.52), or SSIN (control V 1/2  = −18.41 ± 0.52; Pre1a V 1/2  = −19.8 ± 1.26), other than preventing the current from fully inactivating at positive potentials. (c) Pre1a (1 μM) caused a strong positive shift in the voltage-dependence of activation for rNaV1.3 late current (analysed at 100 ms) (control V 1/2  = −20.64 ± 0.54; Pre1a V 1/2  = −11.14 ± 0.58).

Depolarising shifts in the activation of rNaV1.2 and hNaV1.7 and of the rNaV1.3 late current are consistent with inhibition of channel gating via interactions with the voltage-sensing domains of repeat II and IV, respectively. Therefore, Pre1a is clearly a gating modifier that interacts with NaV channels in a subtype-dependent manner. As Pre1a inhibits both NaV channel activation and fast inactivation, we propose using the prefix β/δ according to the proposed nomenclature for spider venom peptides29.

Pre1a interacts with the S3-S4 linkers of hNaV1.7 DII and DIV

The extracellular S3-S4 linker region has been demonstrated to play a key role as a binding determinant for spider venom peptides to voltage-sensing domains18, 40. Given that Pre1a potently affects the activation of hNaV1.7, we wanted to determine if this inhibition was mediated via interactions with a specific voltage-sensor domain. To this end, we used the approach of Bosmans et al., whereby the S3-S4 linker region of the KV2.1 channel was substituted with corresponding linker region of hNaV1.7 DI to DIV18, 41. The effect of Pre1a on the resultant K+ current was then tested for native KV2.1 and each of the domain chimaeras. Pre1a (1 μM) had no effect on wild-type KV2.1 or the DI and DIII chimaeras, however it inhibited outward current carried by the NaV1.7 DII and DIV/KV2.1 chimaeras by 44.0 ± 4.6% and 27.1 ± 3.8% (n = 6), respectively (Fig. 6a and b). Several of the residues crucial for the interaction of HwTxIV with the DII S3-S4 region of NaV1.7 as determine by Xiao et al.23 are also present in DIV, but not in the S3-S4 linker of NaV1.7 DI, DIII or Kv2.1 (Fig. 6c), which supports our observation of weaker effect of Pre1a on the DIV chimaera than the DII chimaera. Interaction with DII is completely consistent with Pre1a’s inhibition of NaV1.7 activation. Interestingly, the interaction with the NaV1.7 DIV/KV2.1 chimaera suggests that Pre1a should have an effect on the inactivation of hNaV1.7, as was noted for hNaV1.1 and rNaV1.3. However, we observed no effects of Pre1a on the inactivation of NaV1.7 at up to 1 μM in oocytes or 300 nM in HEK cells. These results suggest that an interaction with the S3-S4 linker alone is not sufficient to result in a potent functional effect on DIV, consistent with the effects of the spider peptide δ-Hm1a on NaV1.113.

Figure 6
figure6

Pre1a can interact with the DII and DIV S3-S4 linker of hNaV1.7. (a) Representative traces showing the effect of Pre1a (1 µM) on KV2.1 and chimaeras of KV2.1 containing S3-S4 linker region from each domain of hNaV1.7. (b) Normalised peak-current inhibition by Pre1a (1 µM) for each KV2.1/hNaV1.7 chimaera (n = 6). Chimaeras of KV2.1 with the hNaV1.7 DII and DIV had peak current inhibited after addition of 1 µM Pre1a by 44.0 ± 4.6% and 27.1 ± 3.8%, respectively. (c) Alignment of KV2.1/NaV1.7 chimaera S3-S4 regions. Grey highlight indicates the residues determined by Xiao et al.23 to be key for HwTxIV functional effects on hNaV1.7.

Pre1a requires interactions with S3-S4 and S1-S2 regions to produce a functional effect at DIV

The results above show that Pre1a has potent and subtype-dependent effect on NaV channel gating. It potently inhibits the activation (DII effect) of hNaV1.1, rNaV1.7 and rNaV1.2, but not rNaV1.3, whereas it selectively inhibits the inactivation (DIV effect) of NaV1.1 and 1.3, but not 1.2 or 1.7. The striking differences in effects on NaV1.1, 1.2 and 1.3 are of particular significance for understanding the molecular basis of Pre1a’s subtype selectivity as these three channels have essentially identical S3-S4 linker regions in both DII and DIV (Fig. 7a) (note that NaV1.1 DII S3-S4 has an M to V substitution at the top of the S3 helix). Thus, this region cannot account for the substantial differences in NaV channel subtype selectivity observed for Pre1a. The S1-S2 linker region has previously been shown to contribute to the interaction of HwTxIV with NaV1.7 via E753 a the top of S123. This residue is highly conserved in all NaV channels, thus it cannot account for the subtype selectivity that we observed. Figure 7a shows that there is sufficient sequence variation between NaV1.1, 1.2, 1.7 and 1.3 in the first half of S2 of DIV (indicated as 1, 2 and 3) to possibly explain our data. Previous mutagenesis studies on DII23 and DIV13 have shown that the divergent residues in region 1 have little effect on the activity of HwTxIV and δ-Hm1a, respectively. As region 3 of helix S2 is identical in both NaV1.1 and NaV1.3, but not 1.2 and 1.7, we assessed the potential role of these two residues (Ser/Arg) in the ability of Pre1a to inhibit inactivation. Using a chimaera-based approach we found that replacing the S3-S4 linker of rNaV1.4 (Pre1a insensitive) with corresponding linker from hNaV1.1 (Pre1a sensitive) was not sufficient to allow inhibition of inactivation by Pre1a (1 μM) (Fig. 7b,c and d). Using the rNaV1.4(1.1:S3-S4) background, introduction of Ser1379 (S1574 in hNaV1.1), but not Arg1380 from NaV1.1 S1-S2 was sufficient for Pre1a to have a functional effect on both the extent (Fig. 7c) and rate (Fig. 7d) of fast inactivation.

Figure 7
figure7

The S3-S4 linker alone accounts for the subtype selectivity of Pre1a. (a) Alignment of DII and DIV extracellular linkers S1-S2 and S3-S4 for NaV channel isoforms used in this study. Grey shading indicates identity to rNaV1.3 for both domains, residue colouring indicates; blue = basic/positive, red = acidic/negative, green = polar, black = hydrophobic. Helices are defined based on the structures of rabbit CaV1.1 (PDB: 5GJV) for DII, and the NaVAb/hNaV1.7 chimaera (PDB: 5EK0) for DIV. *Indicates residues mutated in DII by Xiao et al.23, red highlight indicates importance for HwTxIV interaction. (b) Representative traces showing the effect of 1 µM Pre1a on hNaV1.1 and chimaeras of rNaV1.4 containing the S3-S4 linker region of Nav1.1 DIV, and additional NaV1.4 to 1.1 point mutants in the adjacent S1-S2 linker (schematics illustrating the chimaera constructions are shown below the respective current trace). (c) Normalised effect of Pre1a on the late current of channels in B (n = 5–6). (d) Effect of Pre1a (1 μM) on the Tau of current inactivation (determined from single exponential fit) for channels shown in 7B (n = 5–6). #P < 0.05 Wilcoxon paired t-test.

Effects of Pre1a on SH-SY5Y human neuroblastoma cells and rat dorsal root ganglion neurons

As reported previously, NaV1.2, NaV1.3 and NaV1.7 (but not NaV1.1 or NaV1.6) are robustly expressed in the human neuroblastoma cell line, SH-SY5Y 42. Therefore, synthetic Pre1a was tested on SH-SY5Y cells using manual patch clamp to assess the effects on native human NaV channel currents. As shown in Fig. 8, 300 nM Pre1a inhibited the rapid peak Na+ current in a manner consistent with previously observed effects on rNaV1.2 and hNaV1.7, while concurrently slowing fast inactivation as was observed with rNaV1.3. This suggests that endogenously expressed hNaV1.3 exhibits similar sensitivity to Pre1a as compared to the rat isoform expressed in oocytes.

Figure 8
figure8

Pre1a inhibits native human and rat NaV currents. (a) Application of 300 nM sPre1a on the human neuroblastoma cell line, SH-SY5Y, inhibits both peak current and fast inactivation, consistent with a NaV1.3 and NaV1.7 effect. (b) sPre1a (300 nM) applied to DRG neurons from sham and nerve injured rats resulted in a similar effect to that seen with SH-SY5Y cells. (c) Representative traces showing the effect of Pre1a (300 nM) on SH-SY5Y cell and DRG neuron from sham animal (control = blue; 300 nM sPre1a = red).

Additionally, small DRG neurons were excised from both sham and peripheral nerve ligated (PNL) adult rats, 7 days post operation. In both sets of DRG neurons, 300 nM synthetic Pre1a inhibited the peak Na+ current and slowed fast inactivation in a manner comparable to SH-SY5Y cells (Fig. 8). NaV1.3 expression is barely detectable in Sprague-Dawley rats at 30 days post-natal43 and it has been shown that this channel is upregulated and highly expressed in sensory neurons after 7–9 days post axotomy of peripheral, but not central axons44, 45. Therefore, minimal effects on fast inactivation of neurons from sham rats were expected, but a substantial effect on the fast inactivation of DRG neurons from post-PNL rats was expected but not observed (Fig. 8b). This result was surprising, as developmental and post-injury regulation of NaV1.3 would suggest this channel should not play a significant role in conductance for adult DRG neurons isolated from sham animals. We showed that Pre1a (at ~30 nM) also inhibits fast inactivation of NaV1.1 but not NaV1.6, thus the persistent current observed with Sham rat DRG neurons in the presence of Pre1a is likely due to effects on NaV1.1, which is present in adult neurons and has been reported to not undergo regulatory changes post-axotomy13, 46.

Discussion

We have identified and characterised β/δ-TRTX-Pre1a, a tarantula venom peptide that potently modulates neuronal NaV channels and possesses a unique structural and pharmacological profile. Pre1a exhibits over 70% sequence similarity with several other spider venom peptides (e.g. PaurTx3, CcoTx1, GpTx1 and Ccy1b22, 41, 47); however, the toxin is unique among these peptides in its ability to potently and selectively inhibit fast inactivation of NaV1.1 and NaV1.3. Furthermore, Pre1a shows a marked structural heterogeneity/flexibility in both aqueous and organic solvents, which may be related to its unique activity profile when compared to its closest sequence relatives. Thus, Pre1a represents a valuable research tool to help determine the molecular basis for spider venom peptide interactions with multiple modulatory sites in sodium channels, particularly in respect to effects on fast inactivation.

NaV channel fast inactivation is controlled by the movement of the DIV voltage-sensing domain17. Although several spider venom peptides (such as JzTxI and JzTxII) have previously been shown to preferentially inhibit NaV channel fast inactivation, they are either relatively non-selective48 or preferentially target NaV1.549, 50. Two other spider peptides (ProTx-II and JzTxIV) have been found to inhibit both peak current and fast inactivation of NaV channels20, 25, much like Pre1a. However, all these peptides share less than 50% similarity with Pre1a (See Fig. 1d) and have very different subtype selectivity profiles suggesting that Pre1a can help define the basis for subtype selective interactions with DIV. Indeed, through a series of functional experiments using channel chimaeras and point mutants, we demonstrated that a serine residue in the S2 helix of DIV that is unique to hNaV1.1 (S1574) and rNaV1.3 (S1510) is responsible for the selective modulation of fast inactivation by Pre1a. Our results are consistent with several reports suggesting that this serine residue is a major determinant for the impressively subtype selective effects of both small molecule drugs and another spider venom peptide not related to Pre1a. ICA-121431 inhibits hNaV1.3 and hNaV1.1 peak current with ~1000-fold selectivity and functions by trapping the DIV VSD in the activated state (stabilising fast inactivation, rather than inhibiting it)51, whereas δ-TRTX-Hm1a (Hm1a) was recently characterised as a selective inhibitor of inactivation of NaV1.113, having been previously discovered as a low-affinity KV4.2 inhibitor52. Furthermore, a co-crystal structure of a NaV1.7 VSD IV chimaera with the small molecule NaV channel inhibitor GX-936 has been solved revealing the detailed molecular interactions responsible for its selectivity53. GX-936 has the same mechanism of action as ICA-121431 thus arrests the VSD in the inactivated state as opposed to the resting state targeted by Pre1a and Hm1a. Remarkably, and despite the different state stabilised by these peptides and small molecules, they all rely on the same region in S2 (S1574 in hNaV1.1 and neighbouring residues) for their functional selectivity. Together, these results illustrate that the S2 helix is an important locus for subtype selective modulation of NaV channel fast inactivation and that the key residues involved appear to be accessible to ligands in both the resting and inactivated states.

Although Hm1a and Pre1a share the ability to inhibit NaV1.1 fast inactivation, they differ greatly in primary sequence (only 17% similarity when excluding the structurally critical Cys residues). Interestingly, the small amount of similarity that does exist between them is a hydrophobic patch clustered in loop 1 of the peptide (Fig. 1d). Co-incidentally this is the region of Pre1a that is structurally mobile, a characteristic that has not been reported for venom peptides most closely related to Pre1a and that seem to predominantly inhibit channel activation22, 41, 47. From extensive structure-activity studies carried out by several pharmaceutical companies on spider venom peptides closely related to Pre1a (i.e. GpTx1, CcoTx1 and HwTxIV26,27,28, 34), the aromatic residues at the tip of Loop 1 are key residues for the interaction of these peptides with NaV channels, in particular for their inhibitory effects on channel activation. Whether the flexibility of loop 1 in Pre1a has any role in its ability to potently interact with both DII and/or DIV VSDs (i.e., resulting in two conformations of the same pharmacophore residues) remains to be elucidated.

The interactions of spider venom peptides that inhibit NaV channel activation have been the focus of extensive chimaera and scanning mutagenesis studies to characterise the binding site on the channel. Bosmans et al. demonstrated that the S3-S4 linker region of any of DI to DIII can be a primary binding determinant for a number of venom peptides by transplanting this region in to a toxin insensitive background and transferring functional peptide binding18. Subsequently, a motif, which spans the DII S1-S2 and S3-S4 regions, that is critical for the inhibitory activity of HwTx-IV on hNaV1.7 was identified and consists of the residues E753 in S1, and E811, L814, D816, E818 in S3-S423 (See red asterisks in Fig. 7a). This has since been functionally demonstrated for two other spider peptides with substantial similarity to Pre1a54, 55 as well as suggested structurally (in solution using NMR) for the interaction of VSTX1 and the VSD of KVAp32. Our data showing a greater effect of Pre1a on the DII Kv2.1/hNav1.7 S3-S4 chimaera than that containing the DIV S3-S4 linker is in accord with the these previous studies. However, interaction with the residues identified in S1 and S3-S4 does not explain the substantial differences in Pre1a’s ability to inhibit NaV channel activation and despite the studies carried out to date, little insight into the basis of NaV channel subtype selectivity has been gained in regards to DII. Pre1a most potently inhibited the activation of NaV1.7 and NaV1.2 with ~70-fold selectivity over NaV1.3. The latter two channels have identical sequences in the S3-S4 linker, thus binding to this region cannot account for the observed selectivity. Interestingly hNaV1.1 and rNaV1.2 only differ by only two residues in the S2 helix, a conservative Thr (1.3) to Ser (1.2) substitution (at the position corresponding to the Ser1574 in Nav1.1 DIV) and a less conservative Gln (1.3) to Glu (1.2) substitution at the top of S2. This strongly suggests that residues in one or both of these positions make major contributions to the interaction of Pre1a with NaV VSD (driven primarily by interactions with S3-S4) that together result in functional inhibition of the channel. Thus, similar to what we showed for the molecular basis of sub-type selective inhibition of inactivation, subtype selective interactions at DII are also likely determined by the variability in the S2 helix.

In conclusion, we have identified a novel spider venom peptide and demonstrated how its so far unique activity on the NaV channel family can be used to gain insight into the molecular basis of subtype selectivity. It should be noted that few venom peptide modulators of NaV channels have been studied in the same detail as reported here, thus the structural heterogeneity and binding site promiscuity of Pre1a may indeed be a more common property than we realise. Pre1a has great value as a research tool for exploration of the selectivity profile of activation and inactivation on different NaV channel isoforms, as well as exploring the consequences and effects of conformational flexibility to activity in this family of channel modulators.

Methods

Venom peptide purification

Psalmopoeus reduncus venom was purchased from SpiderPharm (Yarnell). Venom was fractionated using a C18 218TP54 column (4.6 × 250 mm, 5 μm, Grace Discovery Sciences) with solvent A (H2O, 0.05% TFA) and B (90% acetonitrile, 0.045% TFA) over a gradient of solvents (15–40% B in 36 min, 40–100% B in 12.5 min at 1 ml/min). Fractions were collected, dried and assayed for activity against hNaV1.7 expressed in Xenopus laevis oocytes. The active fractions were further separated on a PromixMP column (4.6 × 250 mm, Sielc) with a gradient of 10–35% B in 35 min at 1 ml/min, then a Prosphere C4 column (3.0 × 150 mm, Grace Discovery Sciences) with a gradient of 20–45% B at 0.75 ml/min.

Peptides were analysed using MALDI-TOF mass spectrometry (Applied Biosystems 4700 Proteomics Bioanalyser) in reflector mode using α-cyano-4-hydroxy-cinnamic acid (CHCA, 5 mg/ml in 60:40 solvent B:A). Reduction/alkylation and sequencing of approximately 3 μg of the peptide was carried out by N-terminal Edman degradation.

Peptide synthesis

Synthetic Pre1a was assembled manually using Boc SPPS chemistry as described previously56. The side-chain protecting groups chosen were Asn (Xan), Arg (Tos), Asp (OcHex), Cys (4-MeBzl), Lys (ClZ), Ser (Bzl), Trp (CHO) and Tyr (BrZ). The crude reduced peptide was purified using preparative reversed-phase chromatography (Vydac C18 218TP1022), using a gradient of 0–80% B over 80 min). The peptide was oxidised at a concentration of 0.02 mM in either aqueous 0.33 M NH4OAc/0.5 M GnHCI or 2 M NH4OH/0.1 M NH4OAc at pH 7.8, 4 °C in the presence of both reduced and oxidised glutathione (peptide:GSH:GSSG, 1:100:10, molar ratio). Oxidised peptides were purified using preparative RP-HPLC.

Bacterial recombinant production of Pre1a

Pre1a was produced recombinantly using an E. coli periplasmic expression system as described previously57. Briefly, a synthetic gene encoding Pre1a was codon optimised for bacterial expression and subcloned into a pLicC-His6-MBP periplasmic expression vector, where the peptide is expressed as a C-terminal fusion to His6-tagged maltose binding protein (MBP) separated by a tobacco etch virus (TEV) protease cleavage site, leaving an additional N-terminal Ser after cleavage. Fusion proteins were expressed in E. coli strain BL21(λDE3) and isolated from cell lysates using Ni-NTA Superflow resin (Qiagen). The His6-MBP tag was removed from the fusion protein using TEV protease, and recombinant Pre1a purified using RP-HPLC.

For production of uniformly 13C/15N-labelled Pre1a, cultures were grown in minimal medium supplemented with 13C6-glucose and 15NH4Cl as the sole carbon and nitrogen sources, respectively. In order to facilitate comparisons between synthetic and recombinant Pre1a, residue numbers for the native toxin are used throughout the text even though the recombinant toxin contains an additional N-terminal serine residue that is a vestige of the TEV cleavage site.

Two-electrode voltage clamp analysis on NaV channel-expressing oocytes

The preparation and injection of Xenopus laevis oocytes and NaV channel recordings using two-electrode voltage-clamp (TEVC) were carried out as described previously58. Capped cRNA encoding rat NaV1.2, NaV1.3 and NaV1.4, and human NaV1.5 and NaV1.7 were injected at 20–40 ng of cRNA/oocyte and kept at 17–18 °C in ND96 solution containing (in mM) 96 NaCl, 2 KCl, 1 CaCl2, 2 MgCl2, 5 HEPES, 5 pyruvic acid, 50 µg/ml gentamicin (pH 7.4), and horse serum (2.5%). Membrane currents were recorded 2–5 days after injection. Oocytes were clamped at −90 mV and currents were elicited by a 50 ms depolarising step to −10 or 0 mV every 10 s for concentration response curves. Both peak and late current (50 ms post peak) were analysed. The concentration response curves were fitted to the Hill equation (GraphPad Prism 6). Current-voltage (I-V) relationships and voltage-dependence of fast inactivation were determined using a family of 100 ms conditioning pulses from −100 mV to +50 mV (+60 mV for rNaV1.3) in 5 mV steps, followed by a test depolarisation step to 0 mV with a 20 s sweep interval from a holding potential of −90 mV. To determine the voltage-dependence of activation non-normalised peak current was used from the I-V family to calculate channel conductance using the equation: G(V) = I/(VV rev ), in which I, V, and V rev represent inward NaV current, test potential, and reversal potential, respectively. The half-activation potential (V 1/2 ) was determined using a Boltzmann fit (GraphPad Prism 6). All data points are shown as the pIC50 ± S.E.M., as well as the IC50. Replicates (n) represent separate experimental oocytes. Pre1a was resuspended in ND96 solution containing 0.1% bovine serum albumin for all activity analysis. Construction and recording of the KV2.1chimaeras containing hNaV1.7 DI-DIV S3-S4 linker regions was completed as described previously18, 41. Design and construction of the hNaV1.1/rNaV1.4 DIV chimaeras is detailed in Osteen et al.13. Not all isoforms of NaVs from the same organism were available at the time of experimentation. Even though different species (rat and human) were used, careful analysis of the data in relation to the sequence differences of the isoforms used, still yields valuable insights in to the molecular basis of peptide:channel interaction.

NMR data acquisition and structural analysis

13C/15N-labelled peptide in 20 mM potassium phosphate buffer, pH 5 containing 5% D2O was filtered using a low-protein-binding Ultrafree-MC centrifugal filter (0.22 µm pore size; Millipore, MA, USA), then 300 μL of ~300 μM was added to a susceptibility-matched 5 mm outer-diameter microtube (Shigemi Inc.). NMR spectra were acquired at 298 K on a 900 MHz Bruker AVANCE II+ spectrometer (Bruker BioSpin) equipped with a cryogenically cooled probe. Data used for resonance assignment were acquired using non-uniform sampling (NUS); sampling schedules that approximated the rate of signal decay along the various indirect dimensions were generated using sched3D59. The decay rates used were 1 Hz for all constant-time 15N dimensions, 30 Hz for all 13C dimensions, and 15 Hz for the semi-constant indirect 1H dimension of the H(CC) (CO)NH/(H)CC(CO)NH-TOCSY experiments. 13C- and 15N-edited HSQC-NOESY experiments were acquired using linear sampling. Separate experiments were acquired for the aliphatic and aromatic regions of the 13C dimension.

NUS data were processed using the Rowland NMR toolkit (www.rowland.org/rnmrtk/toolkit.html); maximum entropy parameters were selected automatically as described60. NMR spectra were analysed and assigned using the program XEASY61 or SPARKY62. 1HN, 15N, 13Cα, 13Cβ, and 13C′ resonance assignments were obtained from analysis of amide-proton strips in 3D HNCACB, CBCA(CO)NH, and HNCO spectra. Side-chain 1H and 13C chemical shifts were obtained primarily from 3D H(CC) (CO)NH-TOCSY and (H)CC(CO)NH-TOCSY spectra, respectively. The remaining side-chain assignments were derived from 3D 15N- and 13C-edited NOESY-HSQC spectra. The program TALOS+ 31 was used to predict the secondary structure of the peptide. Disulfide connectives were determined from NOESY patterns59. During the automated NOESY assignment/structure calculation process, CYANA assigned 90.1% of all NOESY cross-peaks (829 of 920).

Automated patch clamp analyses on NaV channel-expressing mammalian cells

For patch clamp analysis, HEK293 cells stably co-expressing the hNaV1.1, hNaV1.6, or hNaV1.7 α-subunit with the NaVβ1-subunit (SB Drug Discovery) were cultured following manufacturer guidelines. Cells were removed from culture at 70% confluency using Detachin (Genlantis) and resuspended to 5 × 106 cells/mL in Ex-Cell ACF CHO Medium (Life Technologies) supplemented with 25 mM HEPES (Sigma-Aldrich) and 1 × Glutamax (Life Technologies) before being transferred to the Q-Patch Q-Stirrer and allowed to recover for 30 min before assay.

The external solution contained (in mM): 140 NaCl, 4 KCl, 2 CaCl2, 1 MgCl2, 10 HEPES, 20 TEA-Cl, 10 glucose, pH 7.4 (with NaOH) and adjusted to 315 mOsm 0.05% BSA was added to prevent adsorptive loss of peptide. The intracellular solution contained (in mM): 140 CsF, 1/5 EGTA/CsOH, 10 HEPES and 10 NaCl, pH 7.4 (with CsOH) and adjusted to 320 mOsm. Whole-cell patch-clamp experiments were performed at room temperature on a QPatch-16 automated electrophysiology platform (Biolin Scientific) using 16-channel planar patch chip plates (Q-Plates) with a patch hole diameter of 1 µm and resistance of 2 ± 0.1 MΩ. Cell positioning and sealing parameters were set as follows: positioning pressure −60 mbar, minimum seal resistance 0.1 GΩ, holding potential −90 mV and holding pressure −20 mbar. Whole-cell currents were filtered at 5 kHz (8-pole Bessel) and digitised at 25 kHz. A P/6 online leak-subtraction protocol was used with non-leak-subtracted currents acquired in parallel. Cells were maintained with holding potentials of −80 (hNaV1.1) or −100 mV (hNaV1.6 and hNaV1.7) between protocols, respectively and depolarised to −10 mV with a sweep interval of 20 s to allow complete recovery from inactivation between sweeps. Pre1a concentrations were incubated for a fixed time of 5–10 min to allow steady-state activity, monitored by the I-T plot.

Patch clamp analyses of SH-SY5Y cells and rat DRG neurons

Whole-cell patch clamp electrophysiology was carried out on SH-SY5Y cells as described previously42 or from small (<25 μm diameter) acutely isolated adult male rat DRG neurons as described previously63. DRG neurons were taken from rats that had undergone either partial ligation of the left sciatic nerve (PNL) to induce a state of neuropathic pain (defined as development of significant mechanical allodynia seven days after surgery) or sham operated rats64. SH-SY5Y cells were used within 24–72 h, and DRG neurons were used within 6 hours of plating. Only cells with minimal or no processes were used. Whole-cell patch-clamp recordings were performed at room temperature with fire-polished patch electrodes prepared from borosilicate glass (SDR Clinical Technology). Electrode resistance was 3.5–5 MΩ when filled with an internal solution containing (composition in mM): 120 CsCl, 5 MgATP, 5 NaCl, 2 CaCl2, 20 HEPES, 10 EGTA, pH 7.3 and adjusted to 283–286 mOsm. Cells were continuously bathed in a HEPES-buffered physiological saline (HBS; composition in mM): 155 NaCl, 2.5 KCl, 1.8 CaCl2, 1.2 MgCl2, 10 HEPES, 10 glucose, pH 7.4 (adjusted with NaOH) and adjusted to 328–331 mOsm. Currents were recorded using a HEKA EPC-9 amplifier and PULSE software (v8.8, HEKA Elektronik, Lambrecht/Pfalz, Germany). Data was filtered at 4 kHz and sampled at 20 kHz. Series resistance was compensated by 80%. Capacitance transients were compensated and leak subtraction was performed with a P/8 online protocol. Drugs were applied to cells using a gravity-fed superfusion system (250 μm diameter) positioned directly above the cell resulting in rapid solution exchange (<500 ms).

Ethics statement

All animal experiments complied with the Australian code of practice for the care and use of animals for scientific purposes, (8th Ed. 2013) and great care was taken to minimise animal suffering at all times. Experiments involving rats were approved by the University of Sydney (Approval number K00/1-2009/3/4940) or Royal North Shore Hospital/University of Technology Animal Ethics Committee (Approval number 0411-067A). Experiments using X. laevis were approved by The University of Queensland Animal Ethics Committee (Approval Number: QBI/059/13/ARC/NHMRC). Oocytes were obtained via recovery surgery performed under tricaine methanesulfonate (MS-222) anaesthesia (animals bathed in 1.3 mg/ml solution). Minimum time between surgeries on the same animal was six months.

References

  1. 1.

    Catterall, W. A., Goldin, A. L. & Waxman, S. G. Voltage-gated sodium channels, introduction, http://www.guidetopharmacology.org/GRAC/FamilyIntroductionForward?familyId=82. (2016).

  2. 2.

    Vetter, I. et al. NaV1.7 as a pain target - from gene to pharmacology. Pharmacol. Ther. In press, doi:10.1016/j.pharmthera.2016.11.015 (2016).

  3. 3.

    Gilchrist, J. et al. NaV1.1 modulation by a novel triazole compound attenuates epileptic seizures in rodents. ACS Chem. Biol. 9, 1204–1212, doi:10.1021/cb500108p (2014).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  4. 4.

    Wingerd, J. S., Vetter, I. & Lewis, R. J. In Therapeutic Targets 63–122 (John Wiley & Sons, Inc., 2012).

  5. 5.

    Dib-Hajj, S. D., Yang, Y., Black, J. A. & Waxman, S. G. The NaV1.7 sodium channel: from molecule to man. Nature Reviews Neuroscience 14, 49–62, doi:10.1038/nrn3404 (2012).

    Article  PubMed  Google Scholar 

  6. 6.

    Yang, Y. et al. Mutations in SCN9A, encoding a sodium channel alpha subunit, in patients with primary erythermalgia. J Med Genet 41, 171–174, doi:10.1136/jmg.2003.012153 (2004).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  7. 7.

    Fertleman, C. R. et al. SCN9A mutations in paroxysmal extreme pain disorder: allelic variants underlie distinct channel defects and phenotypes. Neuron 52, 767–774, doi:10.1016/j.neuron.2006.10.006 (2006).

    CAS  Article  PubMed  Google Scholar 

  8. 8.

    Cox, J. J. et al. An SCN9A channelopathy causes congenital inability to experience pain. Nature 444, 894–898, doi:10.1038/nature05413 (2006).

    ADS  CAS  Article  PubMed  Google Scholar 

  9. 9.

    Amaya, F. et al. The voltage-gated sodium channel NaV1.9 is an effector of peripheral inflammatory pain hypersensitivity. J. Neurosci. 26, 12852–12860, doi:10.1523/JNEUROSCI.4015-06.2006 (2006).

    CAS  Article  PubMed  Google Scholar 

  10. 10.

    Hains, B. C., Saab, C. Y. & Waxman, S. G. Changes in electrophysiological properties and sodium channel NaV1.3 expression in thalamic neurons after spinal cord injury. Brain 128, 2359–2371, doi:10.1093/brain/awh623 (2005).

    Article  PubMed  Google Scholar 

  11. 11.

    Kim, C. H., Oh, Y., Chung, J. M. & Chung, K. The changes in expression of three subtypes of TTX sensitive sodium channels in sensory neurons after spinal nerve ligation. Brain Res. Mol. Brain Res. 95, 153–161, doi:10.1016/S0169-328X(01)00226-1 (2001).

    CAS  Article  PubMed  Google Scholar 

  12. 12.

    Leo, S., D’Hooge, R. & Meert, T. Exploring the role of nociceptor-specific sodium channels in pain transmission using Nav1.8 and Nav1.9 knockout mice. Behav. Brain. Res. 208, 149–157, doi:10.1016/j.bbr.2009.11.023 (2010).

    CAS  Article  PubMed  Google Scholar 

  13. 13.

    Osteen, J. D. et al. Selective spider toxins reveal a role for the Nav1.1 channel in mechanical pain. Nature 534, 494–499, doi:10.1038/nature17976. (2016).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  14. 14.

    Vetter, I. et al. Venomics: a new paradigm for natural products-based drug discovery. Amino Acids 40, 15–28, doi:10.1007/s00726-010-0516-4 (2011).

    CAS  Article  PubMed  Google Scholar 

  15. 15.

    Saez, N. J. et al. Spider-Venom Peptides as Therapeutics. Toxins 2, 2851–2871, doi:10.3390/toxins2122851 (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  16. 16.

    Das, S., Gilchrist, J., Bosmans, F. & Van Petegem, F. Binary architecture of the Nav1.2-beta2 signaling complex. Elife 5, doi:10.7554/eLife.10960 (2016).

  17. 17.

    Ahern, C. A., Payandeh, J., Bosmans, F. & Chanda, B. The hitchhiker’s guide to the voltage-gated sodium channel galaxy. The Journal of General Physiology 147, 1–24, doi:10.1085/jgp.201511492 (2016).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  18. 18.

    Bosmans, F., Martin-Eauclaire, M. F. & Swartz, K. J. Deconstructing voltage sensor function and pharmacology in sodium channels. Nature 456, 202–208, doi:10.1038/nature07473 (2008).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  19. 19.

    Capes, D. L., Goldschen-Ohm, M. P., Arcisio-Miranda, M., Bezanilla, F. & Chanda, B. Domain IV voltage-sensor movement is both sufficient and rate limiting for fast inactivation in sodium channels. The Journal of General Physiology 142, 101–112, doi:10.1085/jgp.201310998 (2013).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  20. 20.

    Wang, M. et al. JZTX-IV, a unique acidic sodium channel toxin isolated from the spider Chilobrachys jingzhao. Toxicon 52, 871–880, doi:10.1016/j.toxicon.2008.08.018 (2008).

    CAS  Article  PubMed  Google Scholar 

  21. 21.

    Bosmans, F. & Swartz, K. J. Targeting voltage sensors in sodium channels with spider toxins. Trends in Pharmacological Sciences 31, 175–182, doi:10.1016/j.tips.2009.12.007 (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  22. 22.

    Bosmans, F. et al. Four novel tarantula toxins as selective modulators of voltage-gated sodium channel subtypes. Mol Pharmacol 69, 419–429, doi:10.1124/mol.105.015941 (2006).

    CAS  Article  PubMed  Google Scholar 

  23. 23.

    Xiao, Y., Jackson, J. O., Liang, S. & Cummins, T. R. Common molecular determinants of tarantula Huwentoxin-IV inhibition of Na+ channel voltage sensors in domains II and IV. Journal of Biological Chemistry 286, 27301–27310, doi:10.1074/jbc.M111.246876 (2011).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  24. 24.

    Corzo, G. & Escoubas, P. Pharmacologically active spider peptide toxins. CMLS, Cell. Mol. Life Sci. 60, 2409–2426, doi:10.1007/s00018-003-3108-6 (2003).

    CAS  Article  PubMed  Google Scholar 

  25. 25.

    Xiao, Y., Blumenthal, K., Jackson, J. O., Liang, S. & Cummins, T. R. The Tarantula Toxins ProTx-II and Huwentoxin-IV Differentially Interact with Human Nav1.7 Voltage Sensors to Inhibit Channel Activation and Inactivation. Molecular Pharmacology 78, 1124–1134, doi:10.1124/mol.110.066332 (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  26. 26.

    Murray, J. K. et al. Single Residue Substitutions That Confer Voltage-Gated Sodium Ion Channel Subtype Selectivity in the NaV1.7 Inhibitory Peptide GpTx-1. J. Med. Chem. 59, 2704–2717, doi:10.1021/acs.jmedchem.5b01947 (2016).

    CAS  Article  PubMed  Google Scholar 

  27. 27.

    Revell, J. D. et al. Potency optimization of Huwentoxin-IV on hNav1.7: A neurotoxin TTX-S sodium-channel antagonist from the venom of the Chinese bird-eating spider Selenocosmia huwena. Peptides 44, 40–46, doi:10.1016/j.peptides.2013.03.011 (2013).

    CAS  Article  PubMed  Google Scholar 

  28. 28.

    Shcherbatko, A. et al. Engineering Highly Potent and Selective Microproteins Against Nav1.7 Sodium Channel for Treatment of Pain. Journal of Biological Chemistry 291, 13974–13986, doi:10.1074/jbc.M116.725978 (2016).

    CAS  Article  PubMed  Google Scholar 

  29. 29.

    King, G. F., Gentz, M. C., Escoubas, P. & Nicholson, G. M. A rational nomenclature for naming peptide toxins from spiders and other venomous animals. Toxicon 52, 264–276, doi:10.1016/j.toxicon.2008.05.020 (2008).

    CAS  Article  PubMed  Google Scholar 

  30. 30.

    Güntert, P. In Protein NMR Techniques Vol. 278 Methods in Molecular Biology™ (ed. A. Kristina Downing) Ch. 17, 353–378 (Humana Press, 2004).

  31. 31.

    Shen, Y., Delaglio, F., Cornilescu, G. & Bax, A. TALOS+: a hybrid method for predicting protein backbone torsion angles from NMR chemical shifts. Journal of biomolecular NMR 44, 213–223, doi:10.1007/s10858-009-9333-z (2009).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  32. 32.

    Lau, C. H., King, G. F. & Mobli, M. Molecular basis of the interaction between gating modifier spider toxins and the voltage sensor of voltage-gated ion channels. Sci. Rep 6, 34333, doi:10.1038/srep34333 (2016).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  33. 33.

    Klint, J. K., Chin, Y. K. & Mobli, M. Rational Engineering Defines a Molecular Switch That Is Essential for Activity of Spider-Venom Peptides against the Analgesics Target NaV1.7. Mol. Pharmacol. 88, 1002–1010, doi:10.1124/mol.115.100784 (2015).

    CAS  Article  PubMed  Google Scholar 

  34. 34.

    Minassian, N. A. et al. Analysis of the Structural and Molecular Basis of Voltage-sensitive Sodium Channel Inhibition by the Spider Toxin Huwentoxin-IV (μ-TRTX-Hh2a). Journal of Biological Chemistry 288, 22707–22720, doi:10.1074/jbc.M113.461392 (2013).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  35. 35.

    Arnold, K., Bordoli, L., Kopp, J. & Schwede, T. The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22, 195–201, doi:10.1093/bioinformatics/bti770 (2006).

    CAS  Article  PubMed  Google Scholar 

  36. 36.

    Gilchrist, J., Das, S., Van Petegem, F. & Bosmans, F. Crystallographic insights into sodium-channel modulation by the beta4 subunit. Proc. Natl. Acad. Sci. USA 110, E5016–E5024, doi:10.1073/pnas.1314557110 (2013).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  37. 37.

    He, B. & Soderlund, D. M. Differential state-dependent modification of rat Nav1.6 sodium channels expressed in human embryonic kidney (HEK293) cells by the pyrethroid insecticides tefluthrin and deltamethrin. Toxicology and Applied Pharmacology 257, 377–387, doi:10.1016/j.taap.2011.09.021 (2011).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  38. 38.

    Goldin, A. L. In Expression and Analysis of Recombinant Ion Channels: From Structural Studies to Pharmacological Screening (eds Derek J. Trezise, Jeffrey J. Clare) 1–25 (Wiley-VCH Verlag, Weinheim, 2006).

  39. 39.

    Witchel, H. J., Milnes, J. T., Mitcheson, J. S. & Hancox, J. C. Troubleshooting problems with in vitro screening of drugs for QT interval prolongation using HERG K+ channels expressed in mammalian cell lines and Xenopus oocytes. Journal of Pharmacological and Toxicological Methods 48, 65–80, doi:10.1016/S1056-8719(03)00041-8 (2002).

    CAS  Article  PubMed  Google Scholar 

  40. 40.

    Li-Smerin, Y. & Swartz, K. J. Localization and molecular determinants of the Hanatoxin receptors on the voltage-sensing domains of a K(+) channel. J. Gen. Physiol. 115, 673–684, doi:10.1085/jgp.115.6.673 (2000).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  41. 41.

    Klint, J. K. et al. Seven novel modulators of the analgesic target NaV1.7 uncovered using a high-throughput venom-based discovery approach. British Journal of Pharmacology 172, 2445–2458, doi:10.1111/bph.13081 (2015).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  42. 42.

    Vetter, I. et al. Characterisation of Nav types endogenously expressed in human SH-SY5Y neuroblastoma cells. Biochemical Pharmacology 83, 1562–1571, doi:10.1016/j.bcp.2012.02.022 (2012).

    CAS  Article  PubMed  Google Scholar 

  43. 43.

    Felts, P. A., Yokoyama, S., Dib-Hajj, S., Black, J. A. & Waxman, S. G. Sodium channel alpha-subunit mRNAs I, II, III, NaG, Na6 and hNE (PN1): different expression patterns in developing rat nervous system. Brain Res Mol Brain Res 45, 71–82, doi:10.1016/S0169-328X(96)00241-0 (1997).

    CAS  Article  PubMed  Google Scholar 

  44. 44.

    Waxman, S. G., Kocsis, J. D. & Black, J. A. Type III sodium channel mRNA is expressed in embryonic but not adult spinal sensory neurons, and is reexpressed following axotomy. J Neurophysiol 72, 466–470 (1994).

    CAS  PubMed  PubMed Central  Google Scholar 

  45. 45.

    Black, J. A. et al. Upregulation of a Silent Sodium Channel After Peripheral, but not Central, Nerve Injury in DRG Neurons. Journal of Neurophysiology 82, 2776–2785 (1999).

    CAS  PubMed  Google Scholar 

  46. 46.

    Wang, W., Gu, J., Li, Y.-Q. & Tao, Y.-X. Are voltage-gated sodium channels on the dorsal root ganglion involved in the development of neuropathic pain? Molecular Pain 7, 16, doi:10.1186/1744-8069-7-16 (2011).

    PubMed  PubMed Central  Google Scholar 

  47. 47.

    Murray, J. K. et al. Engineering Potent and Selective Analogues of GpTx-1, a Tarantula Venom Peptide Antagonist of the NaV1.7 Sodium Channel. Journal of Medicinal Chemistry 58, 2299–2314, doi:10.1021/jm501765v (2015).

    CAS  Article  PubMed  Google Scholar 

  48. 48.

    Tao, H. et al. Molecular determinant for the tarantula toxin Jingzhaotoxin-I slowing the fast inactivation of voltage-gated sodium channels. Toxicon 111, 13–21, doi:10.1016/j.toxicon.2015.12.009 (2016).

    CAS  Article  PubMed  Google Scholar 

  49. 49.

    Wang, M. et al. Jingzhaotoxin-II, a novel tarantula toxin preferentially targets rat cardiac sodium channel. Biochemical Pharmacology 76, 1716–1727, doi:10.1016/j.bcp.2008.09.008 (2008).

    CAS  Article  PubMed  Google Scholar 

  50. 50.

    Tang, C. et al. The tarantula toxin jingzhaotoxin-XI (κ-theraphotoxin-Cj1a) regulates the activation and inactivation of the voltage-gated sodium channel Nav1.5. Toxicon 92, 6–13, doi:10.1016/j.toxicon.2014.09.002 (2014).

    CAS  Article  PubMed  Google Scholar 

  51. 51.

    McCormack, K. et al. Voltage sensor interaction site for selective small molecule inhibitors of voltage-gated sodium channels. Proceedings of the National Academy of Sciences 110, E2724–E2732, doi:10.1073/pnas.1220844110 (2013).

    ADS  CAS  Article  Google Scholar 

  52. 52.

    Escoubas, P., Diochot, S., Célérier, M.-L., Nakajima, T. & Lazdunski, M. Novel Tarantula Toxins for Subtypes of Voltage-Dependent Potassium Channels in the Kv2 and Kv4 Subfamilies. Molecular Pharmacology 62, 48–57, doi:10.1124/mol.62.1.48 (2002).

    CAS  Article  PubMed  Google Scholar 

  53. 53.

    Ahuja, S. et al. Structural basis of Nav1.7 inhibition by an isoform-selective small-molecule antagonist. Science 350, aac5464, doi:10.1126/science.aac5464 (2015).

  54. 54.

    Cai, T. et al. Mapping the interaction site for the tarantula toxin hainantoxin-IV (β-TRTX-Hn2a) in the voltage sensor module of domain II of voltage-gated sodium channels. Peptides 68, 148–156, doi:10.1016/j.peptides.2014.09.005 (2015).

    CAS  Article  PubMed  Google Scholar 

  55. 55.

    Liu, Z. et al. Structure and Function of Hainantoxin-III, a Selective Antagonist of Neuronal Tetrodotoxin-sensitive Voltage-gated Sodium Channels Isolated from the Chinese Bird Spider Ornithoctonus hainana. Journal of Biological Chemistry 288, 20392–20403, doi:10.1074/jbc.M112.426627 (2013).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  56. 56.

    Klint, J. K. et al. Isolation, synthesis and characterization of ω-TRTX-Cc1a, a novel tarantula venom peptide that selectively targets L-type CaV channels. Biochemical Pharmacology 89, 276–286, doi:10.1016/j.bcp.2014.02.008 (2014).

    ADS  CAS  Article  PubMed  Google Scholar 

  57. 57.

    Klint, J. K. et al. Production of Recombinant Disulfide-Rich Venom Peptides for Structural and Functional Analysis via Expression in the Periplasm of E. coli. PLoS One 8, e63865, doi:10.1371/journal.pone.0063865 (2013).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  58. 58.

    Chow, C., Cristofori-Armstrong, B., Undheim, E., King, G. & Rash, L. Three Peptide Modulators of the Human Voltage-Gated Sodium Channel 1.7, an Important Analgesic Target, from the Venom of an Australian Tarantula. Toxins 7, 2494–513, doi:10.3390/toxins7072494 (2015).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  59. 59.

    Mobli, M., Stern, A. S., Bermel, W., King, G. F. & Hoch, J. C. A non-uniformly sampled 4D HCC(CO)NH-TOCSY experiment processed using maximum entropy for rapid protein sidechain assignment. Journal of Magnetic Resonance 204, 160–164, doi:10.1016/j.jmr.2010.02.012 (2010).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  60. 60.

    Mobli, M., Maciejewski, M. W., Gryk, M. R. & Hoch, J. C. An automated tool for maximum entropy reconstruction of biomolecular NMR spectra. Nat Methods 4, 467–468, doi:10.1038/nmeth0607-467 (2007).

    CAS  Article  PubMed  Google Scholar 

  61. 61.

    Bartels, C., Xia, T. H., Billeter, M., Guntert, P. & Wuthrich, K. The program XEASY for computer-supported NMR spectral analysis of biological macromolecules. Journal of biomolecular NMR 6, 1–10, doi:10.1007/bf00417486 (1995).

    CAS  Article  PubMed  Google Scholar 

  62. 62.

    Goddard, T. D. & Kneller, D. G. SPARKY 3, NMR Assignment and Integration Software (University of California, San Francisco) https://www.cgl.ucsf.edu/home/sparky/ (2008).

  63. 63.

    Murali, S. S., Napier, I. A., Rycroft, B. K. & Christie, M. J. Opioid-related (ORL1) receptors are enriched in a subpopulation of sensory neurons and prolonged activation produces no functional loss of surface N-type calcium channels. The Journal of Physiology 590, 1655–1667, doi:10.1113/jphysiol.2012.228429 (2012).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  64. 64.

    Seltzer, Z. e., Dubner, R. & Shir, Y. A novel behavioral model of neuropathic pain disorders produced in rats by partial sciatic nerve injury. Pain 43, 205–218, doi:10.1016/0304-3959(90)91074-S (1990).

    CAS  Article  PubMed  Google Scholar 

  65. 65.

    Meir, A., Cherki, R. S., Kolb, E., Langut, Y. & Bajayo, N. Novel peptides isolated from spider venom, and uses thereof. (2011).

  66. 66.

    Middleton, R. E. et al. Two tarantula peptides inhibit activation of multiple sodium channels. Biochemistry 41, 14734–14747, doi:10.1021/bi026546a (2002).

    CAS  Article  PubMed  Google Scholar 

  67. 67.

    Peng, K., Shu, Q., Liu, Z. & Liang, S. Function and Solution Structure of Huwentoxin-IV, a Potent Neuronal Tetrodotoxin (TTX)-sensitive Sodium Channel Antagonist from Chinese Bird Spider Selenocosmia huwena. Journal of Biological Chemistry 277, 47564–47571, doi:10.1074/jbc.M204063200 (2002).

    CAS  Article  PubMed  Google Scholar 

  68. 68.

    Xiao, Y.-C. & Liang, S.-P. Purification and characterization of Hainantoxin-V, a tetrodotoxin-sensitive sodium channel inhibitor from the venom of the spider Selenocosmia hainana. Toxicon 41, 643–650, doi:10.1016/S0041-0101(02)00280-5 (2003).

    CAS  Article  PubMed  Google Scholar 

Download references

Acknowledgements

We thank A. Goldin (University of California) for the rNav1.2 and rNav1.3 clones, and R. Kass (Columbia University) for the hNav1.5 clone. This work was supported by the National Health and Medical Research Council of Australia (NHMRC: APP1067940 to L.D.R.; Program Grant 569927 to R.J.L., P.F.A., M.J.C., D.J.A.), the Australian Research Council (Future Fellowship FT110100925 to M.M.), a Ruth Kirschstein NIH predoctoral fellowship (F31NS084646 to J.G.) and the National Institutes of Health (R01NS091352 to F.B.).

Author information

Affiliations

Authors

Contributions

Discovery & sequence characterisation (L.D.R.); Synthesis/folding (C.A.U., T.D., P.F.A.); Recombinant production (J.S.W., L.D.R., M.M.); Structural heterogeneity (C.A.U., J.S.W.); NMR studies (Y.C., M.M., L.D.R., J.S.W.); Oocyte assays (J.S.W., L.D.R., B.C.-A., J.G., F.B., D.J.A.); QPatch assays (J.S.W.); SH-SY5Y cells (C.M., M.J.C.); DRG neurons (S.S.M., C.W.V., M.J.C.). Manuscript writing/editing. All authors. Project conception & oversight (LDR).

Corresponding author

Correspondence to Lachlan D. Rash.

Ethics declarations

Competing Interests

The authors declare that they have no competing interests.

Additional information

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Wingerd, J.S., Mozar, C.A., Ussing, C.A. et al. The tarantula toxin β/δ-TRTX-Pre1a highlights the importance of the S1-S2 voltage-sensor region for sodium channel subtype selectivity. Sci Rep 7, 974 (2017). https://doi.org/10.1038/s41598-017-01129-0

Download citation

Further reading

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.