As described in previous studies, the venom of the Grammostola rosea spider was collected by electrical stimulation. The collected venom was lyophilized and stored at −80°C. Subsequently, the venom was dissolved in 0.1% trifluoroacetic acid (TFA) in double-distilled water to a final concentration of 10 mg/ml immediately before being subjected to reversed-phase high-performance liquid chromatography (RP-HPLC) purification. First, reverse-phase HPLC purification was performed using a water HPLC system (Waters Alliance, 2695 HPLC system) with an Ultimate^® XB-C18 column (10 × 250 mm, 5 μm, Welch Materials Inc., Shanghai, China) with a flow rate of 3 ml/min and a gradient of 10–55% A for more than 45 min (solvent A: 0.1% trifluoroacetic acid in acetonitrile, solvent B: 0.1% trifluoroacetic acid in water). The absorbance was measured at 215 nm. The fractions were collected, lyophilized, and then stored at –20°C until the next subdivision. Next, the target fraction containing Gr4b was subjected to a second round of RP-HPLC (Waters Alliance, 2695 HPLC system) using an analytic XB-C18 column, (300 Å, 4.6 mm × 250 mm, Welch Materials Inc., Shanghai, China) with a linear increasing acetonitrile gradient (acetonitrile at an increasing rate of 0.5% per minute and a flow rate of 1 ml/min) to obtain the purified Gr4b. The molecular weight of the peptide was confirmed by matrix assisted laser desorption/ionization-time of-flight mass-spectrometry (MALDI-TOF-TOF MS) spectrometry (AB SCIEX TOF/TOFTM 5800 system, Applied Biosystems, United States). The N-terminal amino acid sequence of the peptide was determined by automated Edman degradation in a PPSQ-53A protein sequencer (Shimadzu Corporation, Kyoto, Japan).

Rat Na_v1.4, human Na_v1.7 and rat Na_v1.8 cDNA clones were kindly gift from Dr. Theodore Cummins (Stark Neurosciences Research Institute, Indiana University School of Medicine, Indianapolis, IN, United States) and were subcloned into the pCMV or pCDNA3.1-blank vectors. Human Na_v1.9 was subcloned into the pEGFP-N1 vector. The C-terminal of hNa_v1.9 was linked to GFP to construct a fusion protein channel (hNa_v1.9-eGFP), which was as described in our previous studies (Zhou et al., 2017). Mutations were made using the site-directed mutation method or recombination-based cloning using GenBuilder™ Cloning Kit (GenScript, United States). Primers presented in Supplementary Tables 1–4. All mutations were verified by DNA sequencing.

ND7/23 and HEK293T cells were maintained at 37°C in a humidified 5% CO₂ incubator in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum, 100 μg/ml streptomycin, 100 U/ml penicillin, and 2 mM L-glutamine. The cells were trypsinized, diluted with 1 ml of culture medium, and seeded at a 1:5 ratio in 35 mm Petri dishes for culture. When grown to 80–90% confluence, the ND7/23 cells were transfected with hNa_v1.9-GFP or hNa_v1.9-GFP mutants using the X-tremeGENE HP DNA Transfection Reagent (Roche, Basel, Switzerland) according to the manufacturer’s instructions. ND7/23 cells were used for hNa_v1.9-GFP chimeric channel expression and the conditions were as previously described (Zhou et al., 2017), the beta subunits (β1 and β3) are endogenously expressed in the cell lines used to study Na_v1.9 (Rogers et al., 2016). Transfections for the other plasmids were performed using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, United States) following the manufacturer’s instructions. Na_v channel plasmid (4 µg) plus 0.5 µg pEGFP-N1 (except for hNa_v1.9) plasmid were co-transfected into HEK293T or ND7/23 (rNa_v1.8 and hNa_v1.9 only) cells. The cells were seeded onto several 3.5 cm dishes at a 1:10 ratio at 4–6 h after transfection. Cells with green fluorescent protein (GFP) were selected for whole-cell patch-clamp analysis at 24–36 h post-transfection.

Whole-cell current recordings were performed using an EPC-10 USB patch-clamp amplifier operated by Patch Master software (HEKA Elektronik, Lambrecht, Germany). The recording pipettes were fabricated from borosilicate glass capillaries using a two-step vertical microelectrode PC-10 puller (Narishige Group, Tokyo, Japan), and the pipette resistance was controlled to be 2.0–3.0 MΩ. Voltage-clamp recordings were acquired with Patch Master software 2 × 73 (HEKA Elektronik) 4 min after establishing whole-cell configuration, and the currents elicited were sampled at 20 kHz and filtered at 5 kHz. After breaking in, the serial resistance was controlled below 5 MΩ, the voltage error was minimized by using 80% serial resistance compensation, and the compensation speed value was 10 µs. For recording Na_v channel currents, the external solution contained (mM): 150 NaCl, 2 KCl, 1.5 CaCl2, 1 MgCl2 and 10 HEPES (pH 7.4, adjusted with NaOH); the pipette solution contained (in mM): 35 NaCl, 105 CsF, 10 EGTA, 10 HEPES (pH 7.4, adjusted with CsOH). The osmotic pressure of the intracellular fluids and extracellular fluids is adjusted to 300–320 mOsm with sucrose. Before use, Gr4b was dissolved in ddH₂O to make a 250 μM stock solution at −20°C. TTX was dissolved in DMSO to make a 1 mM stock solution. TTX was added to bath solution to a final concentration of 1 µM when used to inhibit TTX-sensitive (TTX-S) Na_v channels. Unless otherwise indicated, all chemicals were products of Sigma-Aldrich (St. Louis, MO, United States). For electrophysiology experiments, the stock solution of Gr4b was diluted with fresh bath solution to a concentration of tenfold of the interested concentration, 30 µl of the concentrated peptide was diluted into the recording chamber (containing 270 µl bath solution) far from the recording pipet (the recording cell), and was mixed by repeatedly pipetting to achieve the specified final concentration.

Data were analyzed using the PatchMaster v2x73 (HEKA Elektronik, Lambrecht, Germany), Igor Pro 6 (Wave Metrics, Lake Oswego, OR, United States), Office Excel 2010 (Microsoft Corporation, WA, United States), and GraphPad Prism 7 (GraphPad Software Inc., CA, United States). All data points are shown as mean ± standard error of the mean (SEM), and n was presented as the number of separate experimental cells. The Boltzmann function was used to fit steady-state activation and deactivation curves. Concentration-response curves were fitted using the Hill equation. One-way ANOVA was used to assess the difference between multiple groups. Significant levels were set at p < 0.05.

The venom of the spider Grammostola rosea contains several classes of peptide toxins that target ion channels (Figure 1A). For example, GpTx1 and PaurTx3 are potent inhibitors of the Na_v1.7 channel (Murray et al., 2015; Chen et al., 2020), HaTx1 and VsTx1 significantly inhibit the currents of the K_v2.1 channel (Chen et al., 2012; Bemporad et al., 2006), and GsMTx2/4 blocks mechanosensitive ion channels (Oswald et al., 2002). Due to poor heterologous expression in mammal cells, pharmacological studies of the Na_v1.9 channel have lagged. Previously, we succeeded in achieving functional expression of the Na_v1.9 channel in heterologous cells (Zhou et al., 2017). Using this system, we first identified a spider peptide toxin that activates the Na_v1.9 channel and produces pain in mice (Zhou et al., 2020). In order to identify more specific and novel peptide toxins for Na_v1.9, we used patch-clamp recording to screen animal peptide toxins for the ability to affect the Na_v1.9 channel. Crude venom was fractionated by RP-HPLC, as shown in Figure 1B. By screening the panel, venom fractions with significant Na_v1.9 regulation activity were identified; a fraction potently inhibited the fast inactivation of Na_v1.9 (Figure 1B). This fraction was further purified by RP-HPLC, and approximately 8 μg of purified peptide toxin was obtained from 1 mg crude spider venom (Figure 1C). The purity of this peak was confirmed by MALDI–TOF MS analysis which revealed a peptide toxin with a molecular weight of 4,348.6 Da, which was consistent with the calculated molecular mass (4,348.01 Da). (Figure 1D). N-terminal Edman sequencing and the venom gland transcriptome cDNA data determined a novel 37-residue peptide toxin named Gr4b (rational nomenclature: δ-theraphotoxin Gr4b) (Kimura et al., 2012), as shown in Figure 1E. Sequencing alignment showed that Gr4b shares highly sequence similarity with Family 2 Na_v-targeting spider toxins (NaSpTx), which comprise 42–44 residues and contained six residues and form a conserved cysteine pattern-inhibitor cystine knot (ICK) motif (Klint et al., 2012) (Figure 1E). NaSpTx Family 2 toxins have various ion channel activities that inhibit K_v, Ca_v, and Na_v channels (Klint et al., 2012). Interestingly, the members of the NaSpTx Family 2 Toxins JZTX-XI and Df1a display dual modulatory effects on specific Na_v channels, simultaneously inhibiting peak current and slowing fast inactivation (Liao et al., 2006; Klint et al., 2012; Tang et al., 2014; Cardoso et al., 2017).

The Na_v1.9 current was evoked to −40 mV by a 100-ms depolarization potential from a holding potential of −120 mV in ND7/23 cells. A concentration of 250 nM Gr4b slowed the fast inactivation of Na_v1.9, leading to a large sustained current (Figure 2A). The half-maximum effective concentration (EC₅₀) of Gr4b was determined to be 25 ± 1.0 nM (Figure 2B). Next, we evaluated the effect of Gr4b on a range of Na_v channels expressed in HEK293T or ND7/23 cells. Interestingly, Gr4b inhibited Na_v1.4 and Na_v1.7 channels in HEK293T cells, with preference for Na_v1.7 (Figures 2C–E). As shown in Figures 2D,E, 250 nM Gr4b completely inhibited Na_v1.7 currents with a half-maximum inhibition concentration (IC₅₀) value of 27 ± 3.0 nM. It had substantially weaker effects on Na_v1.4, with an IC₅₀ value of 411 ± 184 nM (Figures 2C,E). However, application of up to 2.5 μM Gr4b had no effect on Na_v1.8 subtype in ND7/23 cells (Figure 2F). Taken together, these results suggest that Gr4b has different actions on different Na_v channel subtypes.

A major effect of gating modifier toxins on Na_v channels is modification of the voltage dependence of channel activation and inactivation. Gr4b plays the role of a gating modifier toxin to slow fast inactivation of the Na_v1.9 channel. To clarify the mode of action of Gr4b, we used the saturation concentration (250 nM) to analyze the effects of Gr4b on the voltage dependence of activation and inactivation properties of the Na_v1.9 channel. As shown in Figure 3A, Gr4b inhibited the fast inactivation currents at all tested voltages, but did not change the threshold of the initial activation voltage or the reversal potential of the Na_v1.9 current (Figure 3B). However, the peak of the current was significantly shifted by +10 mV (Figure 3B). In addition, Gr4b shifted the voltage dependence of the activation curve to a more positive potential by approximately 12.5 mV (Control: −51.3 ± 2.7 mV, Gr4b: −38.8 ± 2.9 mV, n = 8, p < 0.0001) (Figure 3C and Table 1). Furthermore, a remarkable change was observed in the slopes of the curves from 6.3 ± 0.3 mV in the control to 11.2 ± 0.4 mV in the presence of Gr4b (Figure 3C and Table 1, p < 0.0001), indicating that toxin binding might affect the cooperativity of the four voltage sensors of the Na_v1.9 channel. As shown in Figure 3C and Table 1, the steady-state inactivation curve was significantly shifted to a positive direction by approximately 37.1 mV in the presence of 250 nM Gr4b (Control: −53.7 ± 2.2 mV, Gr4b: −16.6 ± 2.7 mV, n = 6, p < 0.0001), whereas the slope of the curve was not changed (Control: 10.0 ± 0.8 mV, Gr4b: 11.3 ± 0.7 mV, n = 6). We also found that Gr4b introduced a non-inactivated component in the steady-state inactivation curve around the test potential. The β1 subunit is known to modulate the kinetics of fast inactivation (Vijayaragavan et al., 2001; Vijayaragavan et al., 2004). Because Gr4b affects inactivation kinetics, we tested whether overexpression of the β1 with Na_v1.9 altered the effect of the toxin. As shown in Figures 3D,E and Table1, Gr4b significantly inhibited the fast inactivation currents of Na_v1.9 co-expression with β1 in ND7/23 cells, and shifted the kinetics of fast inactivation and activation to positive potential, similar to that of the effect of toxin on Na_v1.9 expression in ND7/23 cells. The predicted window currents of the Na_v1.9 channel were obviously improved in the presence of 250 nM Gr4b (Figure 3C). Indeed, 250 nM Gr4b robustly increased the peak of the ramp current of Na_v1.9 currents in ND7/23 cells by 42.40% (Control: −404.0 ± 116.3 pA, Gr4b: −701.1 ± 236.6 pA, n = 5, p < 0.05) (Figure 3F and Table 2). Consistent with the effect of Gr4b on activation, the peak of the ramp current was observably shifted by 21.9 mV (Control: −47.4 ± 1.5 mV, Gr4b: −25.6 ± 3.4 mV, n = 5, p < 0.001), potentially increasing Na⁺ influx (Figure 3F and Table 2). Based on these findings, Gr4b is clearly a gating modifier that alters the voltage dependence of activation and inactivation of the Na_v1.9 channel.

We further investigated the binding kinetics of toxin on the Na_v1.9 channel. As shown in Figure 4A, the time course for 250 nM Gr4b inducing the inhibited inactivation currents of Na_v1.9 was characterized by a slow onset of action, with a τ_on value of 42.8 ± 3.5 s (n = 3). Inhibition of fast inactivation by Gr4b slowed the reversible recovery upon perfusion bath solution washing, with a recovery of approximately 54.39% of the control current within 2.5 min (Figure 4A). These results indicate that Gr4b has a strong affinity for the Na_v1.9 channel.

Gating modifier toxins typically affect channel gating by regulating the voltage-sensor of the channel. In turn, the affinity of the toxins to the channel is also regulated by the stimulus voltage, e.g., the spider toxin ProTx-II inhibits the current of the Na_v1.7 channel with a significant voltage dependence (Xiao et al., 2010). We previously reported that binding of the gating modifier toxins HNTX-III, HpTx1, and HWTX-IV was reversed by prolonged strong depolarizations that activate the voltage sensor (Xiao et al., 2008; Xiao et al., 2010; Zhou et al., 2020; Liu et al., 2013). Therefore, we examined whether prolonged strong depolarizations could reverse the inhibitory effect of Gr4b on Na_v1.9 using the protocol described in Figure 4B. As shown in Figure 4B, a progressively longer strong depolarization (up to 80 mV) led to an increase in the fraction of sodium current recovered from inhibition by Gr4b. A depolarization time lasting 1,028 ms resulted in complete recovery of the slowed inactivated current. These data indicate that Gr4b dissociated from Na_v1.9 in response to prolonged strong depolarizations. Furthermore, as shown in Figure 4C, the degree of dissociation is positively related to the depolarization time and the depolarization voltage, i.e., increased depolarization time and potential were correlated with increased dissociation. From the time course of dissociation of the toxin after strong depolarizations in the presence of 250 nM Gr4b, the dissociation time constant was fitted with a single exponential function and calculated to be 105.6 ± 36.3 ms, 240.0 ± 26.9 ms, and 473.5 ± 90.8 ms for 80 mV, 40 mV, and 20 mV (n = 4 each), respectively. These results suggest that the rate of Gr4b dissociation is voltage-dependent and that stronger depolarization is correlated with a higher rate of dissociation.

Because Gr4b dissociation from Na_v1.9 is voltage-dependent, we hypothesized that the toxin may be bound to the VSD of the channel. Neurotoxins that act on Na_v channels can target six different sites in the channels, with site 3 (DIV VSD) being the hotspot for spider peptide toxins to inhibit fast inactivation (Stevens et al., 2011). One well-characterized example of a Na_v1.9 modulating peptide is HpTx1 from spider venom, which inhibits fast inactivation of the Na_v1.9 channel by binding to the DIV S3–S4 linker (Zhou et al., 2020). In the present study, the underlying mechanism of action of Gr4b on Na_v1.9 channels was similar to that described for HpTx1. To identify the critical region of Na_v1.9 for toxin-induced inhibition of fast inactivation, several chimeric channels were constructed. Since Na_v1.8 is resistant to Gr4b, a chimera strategy was used to screen the critical modules (VSD) responsible for the toxin’s ability to reduce fast inactivation of Na_v1.9. Firstly, we made the chimera Na_v1.9/1.8 DIV VSD, in which the DIV VSD (DIV S1–S4) of Na_v1.9 was replaced with the corresponding domain of Na_v1.8. As shown in Figure 5B, compared with the wildtype (WT) channel, we observed that a Gr4b concentration of 250 nM reduced the efficacy of the Na_v1.9/1.8 DIV VSD chimeric channel (Figures 5B,F). But the steady-state activation curve was significantly shifted to a positive direction by approximately 8 mV in the presence of 250 nM Gr4b (Control: −45.8 ± 3.4 mV, Gr4b: −37.8 ± 3.5 mV, n = 3, p < 0.05) and the slope of the curve was also significantly changed (Control: 6.9 ± 0.6 mV, Gr4b: 10.7 ± 1.4 mV, n = 3, p < 0.05) (Table 3), that consistent with the effect of Gr4b on WT channel. Nevertheless, it still produced large inhibition of fast inactivation of the chimera channel at 2.5 μM, similar to that of WT-Na_v1.9. However, at this concentration, the toxin did not seem to affect Na_v1.8. Therefore, these results indicate that the channel DIV VSD might be involved in the Gr4b–Na_v1.9 interaction to inhibit fast inactivation and that additional binding areas may exist.

The DIII VSD of Na_v channels have been also shown to modulate channel inactivation (Hsu et al., 2017). Thus, we hypothesized that DIII VSD also is critical for Gr4b inhibition of fast inactivation of Na_v1.9. Thus we made the two chimeras Na_v1.9/1.8 DIII VSD and Na_v1.9/1.8 DIII & DIV VSD. The results showed that replacing these regions of Na_v1.9 abolished the effects of Gr4b on the channel. Even at high concentrations (2.5 μM), the chimeric channels were almost unaffected by Gr4b (Figures 5C,D,F). However, the EC₅₀ of Gr4b was determined to be 1.5 ± 0.25 μM in Na_v1.9/1.8 DIV VSD channel (Figure 5E). These results suggest that the DIII VSD of the Na_v1.9 channel plays a key role in Gr4b-mediated inhibition of their fast inactivation currents. Furthermore, we observed that replacing DIII S3-S4 linker region of Na_v1.9 abolished the effects of Gr4b on the channel also (Figures 6A,B,J). To further elucidate the mechanism underlying Gr4b binding to Na_v1.9, amino acid residues in the S3–S4 linker of DIII were replaced by the alanine residues (Figure 6A). The results indicated that three component residues (namely N1139, L1143, and S1145) were critical for Gr4b binding to DIII S3-S4. Mutations in N1139 and L1143 significantly reduced sensitivity to Gr4b (Figures 6B–J and Table 4). In addition, a mutation of Na_v1.9 (S1145A) enhanced sensitivity to Gr4b (Figures 6I,J and Table 4). Taken together, these results suggest that Gr4b preferentially binds to the DIII VSD and has additional interactions with the DIV VSD, while two residues (N1139 and L1143) in the DIII S3-S4 linker may affect the interaction of Na_v1.9 with Gr4b.