Voltage-gated sodium (NaV) channels are transmembrane proteins essential for the rapid influx of sodium ions, initiating and propagating action potentials in excitable cells. Mammals express at least nine NaV isoforms with tissue-specific expression patterns. NaV1.1, NaV1.2, NaV1.3, and NaV1.6, encoded by SCN1A, SCN2A, SCN3A, and SCN8A genes respectively, are highly expressed in the central nervous system (CNS). Pathogenic variants in these channels are linked to neurological disorders including epilepsy, migraine, and neuropathic pain. NaV1.3 plays a crucial role in fetal neuronal development, and mutations are associated with focal epilepsies and polymicrogyria. Furthermore, NaV1.3 is highly re-expressed in injured peripheral sensory neurons, contributing to neuropathic pain hyperexcitability. Therefore, NaV1.3 is a significant therapeutic target for anti-epileptic drugs and analgesics. NaV channels consist of a large pore-forming α-subunit and regulatory β-subunits. The α-subunit comprises four domains (I-IV), each with six transmembrane segments (S1-S6). S1-S4 segments form the voltage-sensor domain (VSD), while S5, S6, and the pore loop constitute the pore module (PM). The intracellular ends of the four S6 helices form the activation gate, and the DIII-DIV loop forms the inactivation gate. Four β-subunits (β1-β4) modulate channel kinetics and cell surface expression. Recent structural studies have revealed conserved structural features of eukaryotic NaV channels and identified multiple natural toxins and clinical drugs targeting various receptor sites, highlighting complex regulatory mechanisms. Most local anesthetics and anti-arrhythmic drugs block sodium conductance by binding to the central pore. However, the high similarity of pore regions across different NaV isoforms limits the development of isoform-selective drugs. Thus, the discovery of new drug binding sites and isoform-selective drugs is crucial to minimize off-target side effects. Natural polypeptide toxins and synthetic drugs modulate NaV channel function by binding to at least six distinct receptor sites. Site-2 neurotoxins, a group of alkaloids with diverse chemical structures, include aconitine, veratridine, grayanotoxin, and batrachotoxin. Bulleyaconitine A (BLA), extracted from Aconitum bulleyanum, is an aconitine analogue used in China to treat chronic pain and rheumatoid arthritis, but its cardiac arrhythmia and hyperexcitability side effects limit its use. Site-2 neurotoxins modify voltage-dependent activation, inactivation, and ion selectivity, shifting voltage-dependent activation to more negative potentials. Aconitine analogues, including BLA, reduce peak current amplitude. Mutagenesis studies suggest that site-2 toxins bind to overlapping but distinct sites within the central pore. Synthetic aryl sulfonamide derivatives, such as ICA121431, selectively inhibit NaV channels with nanomolar potency. ICA121431 selectively inhibits NaV1.3 and NaV1.1, with significantly higher potency than other isoforms. Electrophysiology studies and the crystal structure of a chimeric sodium channel bound to aryl sulfonamide antagonist GX936 indicated that these antagonists bind to a site within the VSD. However, the molecular mechanisms underlying ICA's specific recognition of and inhibition of NaV1.3 remain unclear.

The literature extensively documents the critical roles of voltage-gated sodium channels (NaVs) in neuronal excitability and their involvement in various neurological disorders. Studies have identified specific NaV isoforms, including NaV1.1, NaV1.2, NaV1.3, and NaV1.6, as key players in the central nervous system (CNS). Mutations in these channels have been linked to a range of conditions, such as epilepsy, migraine, and neuropathic pain. Significant attention has focused on NaV1.3, given its involvement in neuronal development and its upregulation in injured sensory neurons, leading to chronic pain. This has spurred research into developing isoform-selective drugs targeting NaV1.3 for therapeutic intervention. Previous studies explored the mechanisms of action of various toxins and drugs targeting NaV channels. The binding sites of numerous toxins, particularly site-2 neurotoxins like batrachotoxin, veratridine, and aconitine, have been investigated using mutagenesis and electrophysiological techniques. These studies indicated that site-2 neurotoxins bind within the central pore, affecting channel gating and ion selectivity. However, the precise binding modes and the structural details remained unclear. Recently, several groups have identified aryl sulfonamide derivatives as potent and isoform-selective inhibitors of NaV channels. The selectivity of these compounds suggested that they may bind to sites outside the central pore, offering a novel approach for developing targeted therapies. The precise binding locations and mechanisms, however, require further investigation.

This study employed cryo-electron microscopy (cryo-EM) to determine the high-resolution structures of the human NaV1.3/β1/β2 complex in the presence of both BLA and ICA121431. The researchers first functionally characterized the human NaV1.3/β1/β2 complex expressed in HEK293 cells using whole-cell voltage-clamp recordings. This established the channel's voltage-dependent activation and inactivation properties, confirming its functionality. Subsequently, the NaV1.3/β1/β2 complex was expressed in HEK293F cells at a large scale and purified to homogeneity using detergent-based extraction and affinity chromatography. Importantly, BLA or ICA121431 was included throughout the purification process to ensure the complex formation. Cryo-EM samples were prepared by applying the purified protein complexes to glow-discharged holey copper grids, plunge-frozen in liquid ethane, and imaged using a Titan Krios transmission electron microscope. Image processing, including motion correction, particle picking, 2D and 3D classification, refinement, and CTF correction, was performed using RELION 3.0, cisTEM, and cryoSPARC software packages. The resulting high-resolution cryo-EM density maps allowed for detailed model building of the NaV1.3/β1/β2-BLA and NaV1.3/β1/β2-ICA complexes. The ion conductance pathway was analyzed using HOLE software. Furthermore, electrophysiological recordings were conducted in HEK293T cells to determine the effects of BLA and ICA121431 on NaV1.3 current amplitude, kinetics, and state-dependent inhibition. Specifically, the expression and purification process involved the use of modified pEG BacMam vectors expressing NaV1.3 (with mCherry and Twin-Strep tags), β1, and β2 subunits in insect cells (Sf9) and human cells (HEK293F). The membrane fraction was isolated, and the protein complex was extracted using n-Dodecyl-β-D-maltoside (DDM) and cholesteryl hemisuccinate (CHS). Purification was achieved using Streptactin beads, followed by size-exclusion chromatography. For the NaV1.3/β1/β2-ICA complex, 50 µM ICA121431 was added during elution and size-exclusion chromatography. For the NaV1.3/β1/β2-BLA complex, BLA was maintained at various concentrations throughout the entire purification process. The final samples were analyzed via cryo-EM and the data processing yielded maps with an overall resolution of 3.3 Å and 3.4 Å for Nav1.3/β1/β2-BLA and NaV1.3/β1/β2-ICA, respectively. The 3D reconstructions facilitated the building and refinement of the atomic models of the complexes. Electrophysiology experiments utilized whole-cell voltage clamp recordings to analyze the effects of BLA and ICA on Nav1.3 function in transfected HEK293T cells under various voltage protocols.

The cryo-EM structures revealed distinct binding modes for BLA and ICA121431 on NaV1.3. BLA bound within the central cavity near the domain I-II fenestration, a region previously implicated as the site-2 neurotoxin binding site. This binding resulted in partial blockage of the ion conduction pathway and an expansion of the pore-lining helices. Electrophysiological data showed that BLA reduced the peak current amplitude in a use-dependent manner (preferentially acting on the open channel state) and elicited a small, persistent current, consistent with its mixed activation and inhibition effects on NaV1.3. The interaction of BLA with the channel involved numerous hydrogen bonds and van der Waals interactions with residues from both domain I and II pore modules. The binding site showed high conservation amongst other NaV isoforms, explaining the low isoform selectivity of BLA. ICA121431, in contrast, preferentially bound to the activated domain IV voltage sensor (VSDIV), not the pore region like BLA. This binding stabilized the inactive state of the channel by strengthening the interaction between the inactivation gate IFM motif and its receptor site. Electrophysiological experiments demonstrated that ICA121431 showed state-dependent inhibition, potently inhibiting NaV1.3 in the inactivated state with an IC50 of 95.5 ± 9.3 nM. Key residues involved in ICA binding were identified as the small side-chains of G1603 on the S3 helix and A1626 on the S4 helix, which create space for the thiazole headgroup, and R1560 and S1559, which interact with the diphenyl tail of ICA. Comparison with the Nav1.7-GX936 structure highlighted the key determinants for isoform selectivity, identifying S1559 and R1560 of NaV1.3 as crucial residues mediating ICA's selective inhibition of NaV1.3 and NaV1.1. Comparison of VSDIV conformations in NaV1.3-ICA, resting-state NavAb, and deactivated NaV1.5-LqhIII complexes confirmed ICA's preferential binding to activated VSDIV.

This study provides high-resolution structural insights into the distinct mechanisms of action of BLA and ICA121431 on the human NaV1.3 channel. The findings resolve the long-standing question of BLA's mixed activation and inhibition effects, demonstrating that the bulky molecule physically blocks the ion path while simultaneously stabilizing the open conformation of the channel. This dual mechanism explains the use-dependent nature of its inhibition. The structural analysis also clarifies the isoform selectivity of ICA121431, revealing that the selective binding to the activated VSDIV is driven by key interactions with the gating charges (R2-R4) and unique structural features in NaV1.3 and NaV1.1, which are not conserved in other isoforms. The allosteric mechanism of ICA inhibition is revealed by its stabilization of the activated VSDIV conformation, leading to strengthened IFM motif binding and channel inactivation. This highlights the potential of targeting voltage-sensor domains for developing isoform-selective NaV channel modulators. This work has implications for developing new therapeutic strategies for various neurological disorders, focusing on isoform-specific targeting for reduced side effects.

This study provides the first high-resolution structures of human NaV1.3 in complex with both a site-2 neurotoxin (BLA) and an isoform-selective antagonist (ICA121431). The structures reveal distinct binding modes and mechanisms of action, explaining the mixed activation/inhibition effects of BLA and the selective inhibition by ICA121431. These findings advance our understanding of NaV1.3 modulation and offer valuable insights for developing more selective and effective therapies targeting NaV channels in various neurological conditions. Future studies could explore the development of novel drugs that exploit these distinct binding sites for improved therapeutic outcomes.

The study primarily focuses on a specific splice variant of NaV1.3, and the results may not be fully generalizable to all NaV1.3 isoforms. The structural studies were performed in detergent micelles, which might not entirely represent the native membrane environment. The cryo-EM density for the β2 subunit was relatively weak, suggesting some conformational flexibility not fully captured in the structures. Further investigations are needed to comprehensively understand the role of other subunits and the impact of membrane environment on drug-channel interactions. The use-dependent inhibition observed with BLA might not be fully captured in the static nature of cryo-EM structural analysis. Detailed kinetic studies integrating structural data could enhance understanding of the dynamic interactions.