Structure-function and rational design of a spider toxin Ssp1a at human voltage-gated sodium channel subtypes

FIGURE 1. Structure-function of Ssp1a. (A) Primary structure of native Ssp1a and recombinant Ssp1a (rSsp1a) showing the disulfide bond connectivity. (B) Fold potency difference of rSsp1a alanine mutants vs. wild type (WT) rSsp1a at hNaV1.2, hNaV1.3 and hNaV1.7. Data were presented as means ± SEM, with n = 3–12. The W5A, F6A, W24A, K25A, Y26A, W28A, R30A did not complete the dose-response curve at hNaV1.2, hNaV1.3, and hNaV1.7 at the maximum concentration tested. Similarly, Y20A, P27A, Y31A and L33A at hNaV1.3 also demonstrated incomplete dose-response. The IC50 values from the incomplete dose-response for these mutants were used to plot the panel for better comparison of rSsp1a residues key to hNaV1.2, hNaV1.3, and hNaV1.7 activity. The readers are requested to follow Supplementary Table S1 for relative IC50 values and Supplementary Figures S1 for dose-response curves. (C) Active residues in the surface of rSsp1a (PDB: 7SKC) Dongol et al., 2021), GpTx-1 (modeled on engineered GPTX-1, PDB: 6MK5) (Murray et al., 2015), HwTx-IV (PDB: 2M4X) Revell et al., 2013) and m3-HwTx-IV (PDB: 5T3M) (Wisedchaisri et al., 2021) were compared to identify which alanine mutation reduced the toxin potency at…

FIGURE 2. Docking pose of m3-HwTx-IV and rSsp1a. (A) Docking orientation of m3-HwTx-IV at NaVAb/NaV1.7 VS2A captured in cryo-EM and (B–D) docking orientation of rSsp1a at hNaV1.2, hNaV1.3, and hNaV1.7 obtained using HADDOCK are shown with side and top views. The red surface indicates the activity surface of the corresponding toxin, with dark red indicating the Trp residue from the WCK/R motif. The cyan colour indicates non-identical residues among hNaV1.2, hNaV1.3, and hNaV1.7. The recently published NaVAb/NaV1.7 VS2A structure (Wisedchaisri et al., 2021) provided the resting structure of DII that served as the hNaV1.7 DII used to model the hNaV1.2 and hNaV1.3 DII resting state.

FIGURE 3. Molecular docking of rSsp1a. Molecular docking of rSsp1a at (A) hNaV1.2 DII (B) hNaV1.3 DII and (C) hNaV1.3 DII illustrating electrostatic and hydrophobic interactions. The resting conformation of hNaV1.2 and hNaV1.3 DII was modelled on the resting state structure of NaVAb/NaV1.7-VS2A chimera (PDB: 7K48) (Wisedchaisri et al., 2021). HADDOCK2.2 Easy interface (Van Zundert et al., 2016) was used for docking and the results were visualized using Pymol 2.4.1 (Schrodinger, 2018). The green and orange dots around the residues corresponds to the van der Waals radii. (D) Dose-response of rSsp1a at wild type (WT) and F823G-mNaV1.7. Data presented as means ± SEM, with number of experiments (n) indicated in the legend.

FIGURE 4. Dose-response of rSsp1a mutations. (A–C) Dose-response curve of selected rSsp1a analogues compared with rSsp1a at hNaV1.2, hNaV1.3, and hNaV1.7 obtained using the whole-cell automated patch clamp electrophysiology (QPatch 16X) platform. Data presented as means ± SEM, with number of experiments (n) indicated in the legend.

FIGURE 5. Molecular docking of rSsp1a single mutants. Visualization of molecular interactions of (A) N14D-rSsp1a, (B) P27R-rSsp1a and (C) D32Y-rSsp1a compared to rSsp1a at hNaV1.2. Consistent with the activity data, rSsp1a and N14D-rSsp1a docked similarly at hNaV1.2 while P27R-rSsp1a formed additional ionic interactions through the substituted R27 and E785 and Y781 (main chain). In addition, Y26–E779 engaged the S1–S2 loop to further strengthen the toxin–channel complex. Compared to rSsp1a, D32Y-rSsp1a formed a stronger interaction with hNaV1.2, engaging both the S1–S2 and S3–S4 loops with ionic interactions, and consistent with the activity data.

FIGURE 6. Potency and selectivity optimization of rSsp1a. (A) Potency fold-difference of selected rSsp1a analogues presented against the potency of wild type rSsp1a at hNaV1.2, hNaV1.3, and hNaV1.7. Data were presented as means ± SEM (n = 3–12, except at hNaV1.2 for E18F n = 2). (B–D) Dose-response curve of rSsp1a double mutants compared with rSsp1a at hNaV1.2, hNaV1.3 and hNaV1.7, obtained using the whole-cell automated patch clamp electrophysiology (QPatch 16X) platform. Data presented as means ± SEM, with number of experiments (n) indicated in the legend.

FIGURE 7. Molecular docking of rSsp1a double mutants. (A) D1G-D32Y-rSsp1a docked differently at hNaV1.2 compared to rSsp1a (Supplementary Figures S9A) to form stronger ionic bonds, engaging both the S1–S2 and S3–S4 loops, as well as a possible π–π interaction between Y32–Y781, which was consistent with the activity data. (B) D1G-D32Y-rSsp1a docked at hNa V 1.3 similar to docking to hNa V 1.2, but in a different orientation compared to rSsp1a (Supplementary Figures S9B), which enabled stronger interactions that engaged both the S1–S2 and S3–S4 loops and a possible Y32–Y782 π–π interaction. (C) At hNa V 1.7, both rSsp1a (Supplementary Figures S9C) and D1G-D32Y-rSsp1a docked with a similar orientation, forming new interactions including W5–D816 and Y32–H755. (D) At hNa V 1.2, S7R-E18K-rSsp1a docked similar to rSsp1a (Supplementary Figures S9D) with R7 in S7R-E18K-rSsp1a forming an ionic bond with E785 from the S1–S2 loop. In addition, a π–π interaction between W24–Y781 in a parallel displaced geometry was noted. These extra interactions engaging the S1–S2 loop are predicted to form a tighter toxin–channel complex, thus improving the toxin potency. (E) S7R-E18K-rSsp1a docked at hNaV1.3 similar…