Mapping Structural Distribution and Gating-Property Impacts of Disease-Associated Missense Mutations in Voltage-Gated Sodium Channels

Abstract

Thousands of voltage-gated sodium (Nav) channel variants contribute to a variety of disorders, including epilepsy, autism, cardiac arrhythmia, and pain disorders. Yet variant effects of more mutations remain unclear. The conventional gain-of-function (GoF) or loss-of-function (LoF) classifications is frequently employed to interpret of variant effects on function and guide precision therapy for sodium channelopathies. Our study challenges this binary classification by analyzing 525 mutations associated with 34 diseases across 366 electrophysiology studies, revealing that diseases with similar phenotypic effects can stem from unique molecular mechanisms. Our results show a high biophysical agreement (86%) between homologous disease-associated variants in different Na_v genes, significantly surpassing the 60% phenotype (GoF_o/LoF_o) agreement among homologous mutants, suggesting the need for more nuanced disease categorization and treatment based on specific gating-property changes. Using UniProt data, we mapped over 2,400 disease-associated missense variants across nine human Nav channels and identified three clusters of mutation hotspots. Our findings indicate that mutations near the selectivity filter generally diminish the maximal current amplitude, while those in the fast inactivation region lean towards a depolarizing shift in half-inactivation voltage in steady-state activation, and mutations in the activation gate commonly enhance persistent current. In contrast to mutations in the PD, those within the VSD exhibit diverse impacts and subtle preferences on channel activity. This study shows great potential to enhance prediction accuracy for variant effects based on the structural context, laying the groundwork for targeted drug design in precision medicine.

Figure 1:

The gating properties and functional transitions of Na_v channels. (A) Gating properties are listed with their GoF_GP (green) or LoF_GP (red) effects. These properties comprised maximal current amplitude $\left({\text{I}}_{\max }\right)$, half-activation voltage in steady-state activation (V_{1/2 Act}), half-inactivation voltage in steady-state fast inactivation (V_{1/2 Inact}), recovery rate $\left({\tau }_{\text{rec}}\right)$, persistent current $\left({\mathrm{I}}_{\text{P}}\right)$, and gating pore current (or $\omega$ current, ${\text{I}}_{\omega }$). (B) The gating properties with their relevant transitions in the functional cycle of Na_v channels. The Na_v structure has four similar subunits (I to IV), and each subunit comprises six transmembrane helices (S1-S6). The first four TMs (S1 to S4) form a voltage-sensing domain (VSD), and the TMs S5 and S6 contribute to the pore domain (PD). Sensing the membrane depolarization, VSDs undergo resting-to-activated structural transition. Then the channel inactivates mediated by allosteric blocking of IFMT motif when depolarization is prolonged to a certain timescale. Thirdly, the repolarization of membrane potential allows recovery from the fast inactivation to the resting state. The gating properties are labeled with their relevant functional step.

Figure 2:

The workflow in the study for collecting electrophysiology and genetic data. (A) The workflow to search and extract electrophysiology research articles that study the mutational effects on gating properties. A total of 854 articles from Scopus were reviewed, and a rigorous selection process identified 366 articles relevant to this study. From these articles, 525 unique mutations with gating properties were identified and selected as the core data for further analysis and investigations. (B) Data extraction steps to retrieve disease and variant data from UniProt. Initially, more than 36 million mutations in UniProt were filtered, resulting in a refined dataset of 2.4K non-cancerous pathogenic missense mutations in Na_v.

Figure 3:

The classification of Na_v associated diseases based on overall mutational effect or impacts on gating properties. (A) 34 sodium channelopathies are grouped based on a binary GOF_o/LOF_o classification according to previous literature (Table S2). Diseases are colored in green for GoF_o phenotypes, red for LoF_o phenotypes, and yellow for diseases with mixed overall effect (MIX_o). (B) Gating-property impacts of 536 mutations are mapped into their associated 34 diseases. The diseases are also clustered based on the similarity of the gating-property impacts of their associated mutations. The GoF_GP/LoF_GP preference index is depicted colorimetrically with dark green representing highly consistent GoF_GP effect, dark red for highly consistent LoF_GP effect, and black no such a gating-property data available. The percentage (%) of mutations affecting a certain gating property within a specific phenotype is shown in each grid.

Figure 4.

Preferences of gating-property impacts for disease-associated mutations in different structural segments. The variant effects on six gating properties of 525 mutations from 366 papers are mapped into seven major structural segments(A), different selections in VSDs (B), and PD (C). HS stands for the mutation hotspots of corresponding structural segments. The percentage (%) of mutations affecting a certain gating property within a specific structural segment are shown in each grid. GoF_GP/LoF_GP preference index is depicted colorimetrically with dark green representing highly consistent GoF_GP effect, dark red for highly consistent LoF_GP effect, and black for no such gating-property data available. (D) Shows the selection for selectivity filter (SF), activation gate (AG), fast inactivation region (FIR), as well as the upper and lower part for PD and VSD. The disease-associated mutations are represented in blue (upper) and red (lower) spheres.

Figure 5:

Mapping the mutation hotspots in Na_v channels. The annotated disease-associated mutations from UniProt are mapped to equivalent positions of MSA of the 9 human Na_v channels and the linear protein structure of the channel. The number of phenotypes (upper panel), the number of mutations (middle panel), and the number of proteins with mutation at the same position (lower panel) are used to determine the mutation hotspots. The data for pathogenic missense mutations are shown in blue bars and all information related to all missense (including pathogenic, benign, and uncertain) variants are shown in orange. Hotspots in the N terminal, C terminal, intracellular, and extracellular loops of the proteins are shown in this figure with residue IDs from Na_v1.5.

Figure 6:

The mutation hotspots in VSDs. (Left) The hotspots showing in the MSA of Nav channels for 4 VSDs. All mutation hotspots in VSDs are labeled with the residue IDs in Na_v1.5. (Right) Mapping the hotspots in the structure of VSDs. Residues are shown in licorice and colored according to Taylor color scheme.

Figure 7:

The mutation hotspots in the pore domain. (Left) The hotspots in PD showing in the MSA of Na_v channels. (Right) Mapping of the hotspots in the PD of the Na_v1.5 structure from the sideview (top) and bottom view (middle). The mutation hotspots at the intracellular loops are shown in the right bottom image.

Figure 8:

Mapping the phenotypes based on the gating-property impacts of associated mutations. Phenotype clustering of missense mutation in Na_v based on the similarity of gating-property impacts. Black cells represent no data for the corresponding disease segment. The heatmap is colored based on the percentage (%) of mutations affecting a certain gating property within a specific phenotype. Diseases are colored in green (GoF_o), red (LoF_o), yellow (MiX_o), and black for undetermined phenotypes.