Investigating the Patterns in SCN8A Mutations Linked to Early-Onset Seizures

Abstract

Voltage-gated sodium ion channels play a vital role in neuron function, which becomes evident when variants in these genes disrupt their function. Mutations in SCN8A have only recently been linked to an epilepsy phenotype characterized by early-onset of seizures, delayed development, and intellectual disability. Using patient data from published papers (n = 75) as well as an online support group (n = 61), we were able to identify several patterns in the pathogenic mutations and compare them to a group of healthy individuals from the Exome Aggregation Consortium (ExAC) (n = 960). Most of the variants are missense, with one reported nonsense mutation in an individual with ataxia instead of epilepsy. The average age of onset from 85 individuals combined from the published papers and the online support group was 4.45 months. Notably, the ages of onset have a bimodal distribution instead of normal, with one peak within the first month of life and a second peak at 4 months of age. Pathogenic mutations are more likely to appear in the transmembrane regions of the protein encoded by SCN8A (Nav1.6) (Proportion Test: p-value = 9.1 x 10⁻⁵) while non-pathogenic variants were more likely to appear in the connecting loops of the protein (Proportion Test: p-value<2.2 x 10⁻¹⁶). This is most likely due to the transmembrane regions having a greater functional role than the loops. When we further separate the mutations into functional parts of the protein, we generally see that areas with more pathogenic mutations have fewer non-pathogenic variants, and vice versa. For example, the inactivation gate of the protein has more pathogenic variants (Exact Binomial Test: p-value = 3.52 x 10⁻⁶) while the N-terminus has more non-pathogenic variants (Exact Binomial Test: p-value = 0.0128). This suggests that areas with more pathogenic variants are less tolerable to variation while those with more non-pathogenic variants are more tolerable. Interestingly, there are some areas of the protein with very few pathogenic and non-pathogenic variants, such as the pore regions of the protein. This is likely due to the vital functional nature of the pore, and any variants in that region lead to lethality. Further analyses are needed to determine if there are correlations between categories of pathogenic mutations and the phenotypes of the patients.