Addition of K22 Converts Spider Venom Peptide Pme2a from an Activator to an Inhibitor of NaV1.7

Spider venom is a novel source of disulfide-rich peptides with potent and selective activity at voltage-gated sodium channels (NaV). Here, we describe the discovery of μ-theraphotoxin-Pme1a and μ/δ-theraphotoxin-Pme2a, two novel peptides from the venom of the Gooty Ornamental tarantula Poecilotheria metallica that modulate NaV channels. Pme1a is a 35 residue peptide that inhibits NaV1.7 peak current (IC50 334 ± 114 nM) and shifts the voltage dependence of activation to more depolarised membrane potentials (V1/2 activation: Δ = +11.6 mV). Pme2a is a 33 residue peptide that delays fast inactivation and inhibits NaV1.7 peak current (EC50 > 10 μM). Synthesis of a [+22K]Pme2a analogue increased potency at NaV1.7 (IC50 5.6 ± 1.1 μM) and removed the effect of the native peptide on fast inactivation, indicating that a lysine at position 22 (Pme2a numbering) is important for inhibitory activity. Results from this study may be used to guide the rational design of spider venom-derived peptides with improved potency and selectivity at NaV channels in the future.


Introduction
Voltage-gated sodium channels (Na V ) are pore-forming transmembrane proteins that regulate the influx of Na + ions across excitable cell membranes, making them essential for the initiation and propagation of action potentials. In humans, nine different subtypes have been described (Na V 1.1-1.9), each with unique biophysical properties and tissue-specific expression profiles [1]. Several subtypes, including Na V 1.7 and Na V 1.8, are highly expressed in peripheral sensory neurons and are therefore critical for somatosensation and nociception. Na V 1.7 and Na V 1.8 are almost exclusively expressed in the peripheral nervous system, with preferential expression in small-diameter unmyelinated nociceptive or "pain-sensing" neurons [2][3][4]. In humans, loss-of-function mutations in SCN9A, the gene encoding Na V 1.7, leads to congenital insensitivity to pain, while several gain-of-function mutations in SCN9A and SCN10A (the gene encoding Na V 1.8) are associated with painful peripheral neuropathies [5][6][7]. This is consistent with studies in rodents, whereby knockout of Scn9a or Scn10a leads to deficits in mechanical, thermal and inflammatory pain [8,9], making both Na V 1.7 and Na V 1.8 promising therapeutic targets of interest for the treatment of pain. While many small-molecule Na V inhibitors are used in the clinic, non-selective activity over other Na V subtypes, including the central isoforms Na V 1.1 and Na V 1.2 and the cardiac specific isoform Na V 1.5 [10,11], limits their widespread use as analgesics. Therefore, there has been a push to develop novel Na V inhibitors with improved subtype selectivity that target the less conserved voltage-sensing domains of the channel [12]. One source of novel Na V modulators is venoms from spiders, cone snails and scorpions, from which many peptides with exquisite Na V subtype selectivity have been described [13,14]. However, compared to tetrodotoxin (TTX)-sensitive Na V subtypes, relatively few venom-derived peptides with sub-micromolar potency at Na V 1.8 have been characterised [15].
Therefore, the aim of this study was to use a high-throughput screen at Na V 1.8 to identify novel bioactives from spider venom. Here, we describe the isolation and characterisation of two novel peptides from the Gooty Ornamental tarantula Poecilotheria metallica that modulate sodium channels.

Venom Collection
Venom of female P. metallica spiders was extracted by weak electrical stimulation as previously described [17] and dried and stored at −20 • C, before being pooled and redissolved in milliQ water for further analysis.
2.3. Membrane Potential Assay at Na V 1.8 CHO cells stably expressing hNa V 1.8/β3 were plated 48 h before the assay on 384-well black-walled imaging plates coated with CellBIND (Corning, MA, USA) at a density of 10,000-15,000 cells/well and loaded with red membrane potential dye (Molecular Devices, Sunnyvale, CA, USA) plus TTX (1 µM) diluted in physiological salt solution (PSS; 140 mM NaCl, 11.5 mM glucose, 5.9 mM KCl, 1.4 mM MgCl 2 , 1.2 mM NaH 2 PO 4 , 5 mM NaHCO 3 , 1.8 mM CaCl 2 , 10 mM HEPES) and incubated for 30 min at 37 • C. Crude dried venom (10 µg/well) was diluted in PSS with 0.1% bovine serum albumin (BSA) and added using the FLIPR TETRA (Molecular Devices) and incubated for 5 min before activation of Na V 1.8 by addition of deltamethrin (150 µM). Changes in membrane potential were assessed using the FLIPR TETRA (excitation, 515-545 nm; emission, 565-625 nm) every 2 s for 25 min after adding the agonist. To quantify the activity of crude venom at Na V 1.8, the area under the curve (AUC) after the addition of deltamethrin was computed using ScreenWorks (Molecular Devices, Version 3.2.0.14).
Venom fractions were assessed for activity at hNa V 1.8 using the FLIPR TETRA membrane potential assay as described above. The active fraction was further fractionated to near-purity and peptide masses were determined using matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry (MS) using a Model 4700 Proteomics Analyser (Applied Biosystems, Foster City, CA, USA) with α-cyano-4-hydroxycinnamic acid (7 mg/mL in 50% ACN + 5% formic acid in H 2 O) as the matrix. Peptide sequences were determined by Edman degradation performed by the Australian Proteome Analysis Facility (Macquarie University, NSW, Australia). Sequence ambiguity was clarified by MS/MS sequencing.

Calcium Responses in TRPV1-HEK Cells
HEK cells stably expressing rTRPV1 were plated 48 h before the assay on 384-well black-walled imaging plates coated with poly d-lysine (ViewPlate-384, Perkin Elmer, Victoria, Australia) at a density of 10,000-15,000 cells/well and loaded with Calcium 4 no-wash dye (Molecular Devices) diluted in PSS and incubated for 30 min at 37 • C. Capsaicin and Pme1a were diluted in PSS with 0.1% bovine serum albumin (BSA) at the concentrations stated and added using the FLIPR TETRA . Changes in fluorescence were assessed using the FLIPR TETRA (excitation 470-495 nm, emission 515-575 nm) every 1 s for 300 s after the addition of compounds. Fluorescence responses were quantified by the maximum increase in fluorescence calculated using ScreenWorks 3.2.0.14.

Electrophysiology
Whole-cell patch-clamp experiments in Na V 1.7-HEK cells and Na V 1.8-CHO cells were performed on a QPatch-16 automated electrophysiology platform (Sophion Bioscience, Ballerup, Denmark) as previously described [18]. The extracellular solution contained in mM: NaCl 140, KCl 4, CaCl 2 2, MgCl 2 1, HEPES 10 and glucose 10; pH 7.4; osmolarity 305 mOsm. The intracellular solution contained in mM: CsF 140, EGTA/CsOH 1/5, HEPES 10 and NaCl 10; pH 7.3 with CsOH; osmolarity 320 mOsm. Concentration-response curves were acquired using a holding potential of −90 mV and a 50 ms pulse to −20 mV (for Na V 1.7) or +10 mV (for Na V 1.8) every 20 s (0.05 Hz). Peptides were diluted in extracellular solution with 0.1% BSA and each peptide concentration was incubated for 5 min. Peak current was normalized to buffer control. The time constant of fast inactivation (τ) was computed by fitting the current decay traces with a single exponential function using the QPatch Assay Software 5.6 (Sophion). I-V curves were obtained with a holding potential of −90 mV followed by a series of 500 ms step pulses that ranged from −110 to +55 mV in 5 mV increments (repetition interval 5 s) before and after 5 min incubation or Pme1a (1 µM). Conductance-voltage curves were obtained by calculating the conductance (G) at each voltage (V) using the equation G = I/(V − V rev ), where V rev is the reversal potential and were fitted with a Boltzmann equation.

Data Analysis
Data were plotted and analysed using GraphPad Prism, version 8.2.0. For concentration-response curves, a four-parameter Hill equation with variable Hill coefficient was fitted to the data. Data are presented as the mean ± SEM.

Isolation of the Novel Spider Venom
Peptides µ-TRTX-Pme1a and µ/δ-TRTX-Pme2a from P. metallica Crude venom isolated from P. metallica ( Figure 1A) inhibited deltamethrin-induced membrane potential changes in CHO cells stably expressing hNa V 1.8, with activity guided fractionation isolating this activity to a single peak eluting at~35% solvent B ( Figure 1A). Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) indicated that this fraction was dominated by two masses (M + H) + of 3911.6 m/z and 3808.5 m/z ( Figure 1B). N-terminal sequencing revealed two novel 35 and 33 residue peptides that we named µ-TRTX-Pme1a (hereafter Pme1a) and µ/δ-TRTX-Pme2a (hereafter Pme2a) based on the rational nomenclature for peptide toxins ( Figure 1C) [19]. The calculated masses and observed masses differed by -1 Da, indicating both Pme1a and Pme2a have an amidated C-terminus. Synthetic Pme1a and Pme2a with C-terminal amidation were used for all further experiments. current was normalized to buffer control. The time constant of fast inactivation (τ) was computed by fitting the current decay traces with a single exponential function using the QPatch Assay Software 5.6 (Sophion). I-V curves were obtained with a holding potential of −90 mV followed by a series of 500 ms step pulses that ranged from −110 to +55 mV in 5 mV increments (repetition interval 5 s) before and after 5 min incubation or Pme1a (1 μM). Conductance-voltage curves were obtained by calculating the conductance (G) at each voltage (V) using the equation where Vrev is the reversal potential and were fitted with a Boltzmann equation.

Data Analysis
Data were plotted and analysed using GraphPad Prism, version 8.2.0. For concentrationresponse curves, a four-parameter Hill equation with variable Hill coefficient was fitted to the data. Data are presented as the mean ± SEM.

Isolation of the Novel Spider Venom Peptides μ-TRTX-Pme1a and μ/δ-TRTX-Pme2a from P. metallica
Crude venom isolated from P. metallica ( Figure 1A) inhibited deltamethrin-induced membrane potential changes in CHO cells stably expressing hNaV1.8, with activity guided fractionation isolating this activity to a single peak eluting at ~35% solvent B ( Figure 1A). Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) indicated that this fraction was dominated by two masses (M + H) + of 3911.6 m/z and 3808.5 m/z ( Figure 1B). N-terminal sequencing revealed two novel 35 and 33 residue peptides that we named μ-TRTX-Pme1a (hereafter Pme1a) and μ/δ-TRTX-Pme2a (hereafter Pme2a) based on the rational nomenclature for peptide toxins ( Figure 1C) [19]. The calculated masses and observed masses differed by -1 Da, indicating both Pme1a and Pme2a have an amidated C-terminus. Synthetic Pme1a and Pme2a with C-terminal amidation were used for all further experiments.

Pharmacological Activity of Pme1a
Alignment of Pme1a to peptide sequences from the Universal Protein Resource (www.uniprot.org) revealed that the peptide shares high sequence homology (56%-80%) with the vanillotoxins (Figure 2), a family of peptides from Psalmopoeus cambridgei that activate TRPV1 with EC 50 s ranging between 0.32 and 12 µM [20]. We therefore tested the activity of Pme1a on TRPV1 heterologously expressed in HEK293 cells. In comparison to the TRPV1 activator capsaicin (EC 50 83 nM), Pme1a (up to 30 µM) had no activity on TRPV1, suggesting that amino acid residue(s) crucial for TRPV1 activity are missing from this peptide ( Figure 3A). As Pme1a was discovered using a Na V channel screen, we next characterised its activity on Na V channels using automated whole-cell patch-clamp electrophysiology.

Pharmacological Activity of Pme1a
Alignment of Pme1a to peptide sequences from the Universal Protein Resource (www.uniprot.org) revealed that the peptide shares high sequence homology (56%-80%) with the vanillotoxins (Figure 2), a family of peptides from Psalmopoeus cambridgei that activate TRPV1 with EC50s ranging between 0.32 and 12 μM [20]. We therefore tested the activity of Pme1a on TRPV1 heterologously expressed in HEK293 cells. In comparison to the TRPV1 activator capsaicin (EC50 83 nM), Pme1a (up to 30 μM) had no activity on TRPV1, suggesting that amino acid residue(s) crucial for TRPV1 activity are missing from this peptide ( Figure 3A). As Pme1a was discovered using a NaV channel screen, we next characterised its activity on NaV channels using automated whole-cell patchclamp electrophysiology. Figure 2. Sequence alignment of Pme1a and Pme2a. Sequence alignment of Pme1a and Pme2a with % sequence identity (to Pme1a) to selected mature spider peptides with a known target from Universal Protein Resource (www.uniprot.org) and two sequences identified from the venom gland of the related species P. formosa with unknown activity. Cysteine residues are shown in bold and * indicates amidated C-terminus. Pme1a and Pme2a align to spider peptides belonging to NaSpTx2 family 2.
Despite being isolated as an inhibitor of NaV1.8 using a fluorescent screen, Pme1a only had modest activity on Nav1.8, inhibiting 21 ± 2 % of the peak current at a concentration of 10 μM ( Figure  3B). In contrast, Pme1a was more potent at NaV1.7, inhibiting peak current with an IC50 of 334 ± 114 nM ( Figure 3C,D). Activity at other NaV channel subtypes is not unexpected, given the high-sequence homology between NaV subtypes [12]. To examine the mechanism of channel block, we next assessed the effect of Pme1a on the voltage-current relationship at NaV1.7 ( Figure 3E). Pme1a (1 μM) shifted the voltage dependence of activation at NaV1.7 to more depolarised membrane potentials (V1/2 activation: Δ = +11.6 mV), confirming it binds to the voltage-sensing domain(s) to modify channel gating ( Figure 3F).

Pharmacological Activity of Pme2a
Alignment of Pme2a to peptide sequences from the Universal Protein Resource (www.uniprot.org) revealed that the peptide shares sequence homology (<65%) to Pme1a and other peptides with activity at NaV channels ( Figure 2). Similar to Pme1a, Pme2a had no activity at TRPV1 (data not shown), and only modest activity at Nav1.8, with the highest concentration tested (10 μM) inhibiting 22 ± 4 % of peak current ( Figure 4A). Pme2a also had modest activity at NaV1.7, concentration-dependently delaying-fast inactivation, albeit not potently, with an estimated EC50 > 10 μM ( Figure 4B,C). Due to low potency at NaV1.7, no further pharmacological characterisation was performed.
Sequence alignment revealed that loop 4 of Pme2a is one amino acid residue shorter compared to others in the family, and peptides with activity on TTX-sensitive NaV channels generally contain a positively charged amino acid at the equivalent position, most often a lysine (Figure 2). We therefore synthesised a [+22K]Pme2a mutant and assessed how the addition of a lysine affected activity at NaV1.7.
[+22K]Pme2a concentration-dependently inhibited NaV1.7 peak current, without the delay in fast activation seen with the native peptide, with a more potent IC50 of 5.6 ± 1.1 μM (Figure 4D,E). Despite being isolated as an inhibitor of Na V 1.8 using a fluorescent screen, Pme1a only had modest activity on Nav1.8, inhibiting 21 ± 2 % of the peak current at a concentration of 10 µM ( Figure 3B). In contrast, Pme1a was more potent at Na V 1.7, inhibiting peak current with an IC 50 of 334 ± 114 nM ( Figure 3C,D). Activity at other Na V channel subtypes is not unexpected, given the high-sequence homology between Na V subtypes [12]. To examine the mechanism of channel block, we next assessed the effect of Pme1a on the voltage-current relationship at Na V 1.7 ( Figure 3E). Pme1a (1 µM) shifted the voltage dependence of activation at Na V 1.7 to more depolarised membrane potentials (V 1/2 activation: ∆ = +11.6 mV), confirming it binds to the voltage-sensing domain(s) to modify channel gating ( Figure 3F).

Pharmacological Activity of Pme2a
Alignment of Pme2a to peptide sequences from the Universal Protein Resource (www.uniprot.org) revealed that the peptide shares sequence homology (<65%) to Pme1a and other peptides with activity at Na V channels (Figure 2). Similar to Pme1a, Pme2a had no activity at TRPV1 (data not shown), and only modest activity at Nav1.8, with the highest concentration tested (10 µM) inhibiting 22 ± 4 % of peak current ( Figure 4A). Pme2a also had modest activity at Na V 1.7, concentration-dependently delaying-fast inactivation, albeit not potently, with an estimated EC 50 > 10 µM (Figure 4B,C). Due to low potency at Na V 1.7, no further pharmacological characterisation was performed.

Discussion
Here, we describe the isolation and NaV activity of two novel spider venom-derived peptides, which are the first peptides to be isolated and functionally characterised from the venom of the species P. metallica, as well as the entire Poecilotheria genus [21]. A transcriptome and proteome from the venom gland of the related species P. formosa identified two related, but not identical peptide sequences (Pf29 and Pf32) [22], indicating that the entire Poecilotheria genus is likely to be a rich source of novel NaV modulators ( Figure 2). Indeed, crude venom from several species of the Poecilotheria genus, including P. metallica, has previously been shown to modulate NaV1.7 [23], consistent with activity of Pme1a and Pme2a at both NaV1.7 and NaV1.8 described here. The two peptides isolated here were only moderately potent at NaV1.7 and NaV1.8, indicating that there may still be additional peptides in the crude venom with more potent activity.
Specifically, both Pme1a and Pme2a align with NaSpTx family 2, which is a large family of spider venom peptides that have promiscuous activity on NaV, KV and CaV channels [24]. Interestingly, this family also contains members that have high selectivity for NaV channels of therapeutic interest, including the NaV1.7 inhibitor μ-TRTX-Pn3a and the NaV1.1 activator δ-TRTX-Hm1a. However, little is known about the structure-activity relationships that define the pharmacology of this family [25,26]. Here, we describe the importance of a lysine (K) at position "22" (using Pme2a numbering) for NaV channel activity. The presence of the positively charged amino acid residue lysine is often important for facilitating membrane interactions in venom-based peptides [18,27]. Addition of K22 altered the pharmacology of Pme2a, converting the native peptide from being a NaV1.7 channel activator to a more potent NaV1.7 channel inhibitor. This is a significant finding, indicating that the amino acid residue(s) present at or around this site can dictate whether family 2 peptides function as NaV activators or inhibitors. It is also possible that extending loop 4 of Pme2a by one amino acid residue contributed to the change in activity by slightly altering the overall structure of the peptide. Future studies elucidating the full NaV selectivity of Pme1a and Pme2a will further define which amino acid residues are important for NaV channel potency and subtype selectivity.
In conclusion, we have identified two novel spider venom-derived peptides that modulate NaV channels. Results from this study provide structure-activity relationship information that may guide the future rational design of spider venom-derived peptides with improved activity and selectivity at NaV channels.  Sequence alignment revealed that loop 4 of Pme2a is one amino acid residue shorter compared to others in the family, and peptides with activity on TTX-sensitive Na V channels generally contain a positively charged amino acid at the equivalent position, most often a lysine ( Figure 2). We therefore synthesised a [+22K]Pme2a mutant and assessed how the addition of a lysine affected activity at Na V 1.7.
[+22K]Pme2a concentration-dependently inhibited Na V 1.7 peak current, without the delay in fast activation seen with the native peptide, with a more potent IC 50 of 5.6 ± 1.1 µM (Figure 4D,E).

Discussion
Here, we describe the isolation and Na V activity of two novel spider venom-derived peptides, which are the first peptides to be isolated and functionally characterised from the venom of the species P. metallica, as well as the entire Poecilotheria genus [21]. A transcriptome and proteome from the venom gland of the related species P. formosa identified two related, but not identical peptide sequences (Pf29 and Pf32) [22], indicating that the entire Poecilotheria genus is likely to be a rich source of novel Na V modulators ( Figure 2). Indeed, crude venom from several species of the Poecilotheria genus, including P. metallica, has previously been shown to modulate Na V 1.7 [23], consistent with activity of Pme1a and Pme2a at both Na V 1.7 and Na V 1.8 described here. The two peptides isolated here were only moderately potent at Na V 1.7 and Na V 1.8, indicating that there may still be additional peptides in the crude venom with more potent activity.
Specifically, both Pme1a and Pme2a align with NaSpTx family 2, which is a large family of spider venom peptides that have promiscuous activity on Na V , K V and Ca V channels [24]. Interestingly, this family also contains members that have high selectivity for Na V channels of therapeutic interest, including the Na V 1.7 inhibitor µ-TRTX-Pn3a and the Na V 1.1 activator δ-TRTX-Hm1a. However, little is known about the structure-activity relationships that define the pharmacology of this family [25,26]. Here, we describe the importance of a lysine (K) at position "22" (using Pme2a numbering) for Na V channel activity. The presence of the positively charged amino acid residue lysine is often important for facilitating membrane interactions in venom-based peptides [18,27]. Addition of K22 altered the pharmacology of Pme2a, converting the native peptide from being a Na V 1.7 channel activator to a more potent Na V 1.7 channel inhibitor. This is a significant finding, indicating that the amino acid residue(s) present at or around this site can dictate whether family 2 peptides function as Na V activators or inhibitors. It is also possible that extending loop 4 of Pme2a by one amino acid residue contributed to the change in activity by slightly altering the overall structure of the peptide. Future studies elucidating the full Na V selectivity of Pme1a and Pme2a will further define which amino acid residues are important for Na V channel potency and subtype selectivity.
In conclusion, we have identified two novel spider venom-derived peptides that modulate Na V channels. Results from this study provide structure-activity relationship information that may guide the future rational design of spider venom-derived peptides with improved activity and selectivity at Na V channels.