Rationale:
Variants in
SCN8A, which encodes the brain voltage-gated sodium channel Na
V1.6, are associated with epileptic encephalopathy and other neurodevelopmental disorders. Similar to other human sodium channels,
SCN8A undergoes developmentally-regulated alternative splicing that generates different Na
V1.6 splice isoforms. Alternative usage of two mutually exclusive versions of exon 5 encoding a portion of the first voltage-sensing domain gives rise to splice isoforms that are expressed predominantly during early development (exon 5N, neonatal Na
V1.6) or later (exon 5A, adult Na
V1.6). When assessing channel function, we found that the isoform used for expression is important for determining variant pathogenicity, even for some variants located in constitutive exons.
Methods:
We determined the functional properties of a series of
SCN8A variants of uncertain significance (VUS) that were restricted to either the adult or neonatal alternative exon 5. Variants were expressed in a derivative of neuronal ND7/23 cells that lack endogenous sodium current. Electrophysiological measurements were made using automated patch clamp recording.
Results:
Three
de novo epilepsy-associated
SCN8A VUS (V211A, R223G, I231T) affecting residues in alternative exon 5A were investigated in the adult-expressed Na
V1.6 splice isoform. All three variants exhibited wild-type (WT) levels of peak current density, but altered voltage-dependence of activation and inactivation consistent with being pathogenic. A fourth engineered variant (Y371C) that renders the channel resistant to tetrodotoxin (TTX), which is commonly used to distinguish recombinant Na
V1.6 from endogenous sodium channels in neurons, has significant effects on current density and voltage-dependence in adult Na
V1.6. We also investigated the functional consequences of two novel VUS (T144S, S217P) discovered in a child with early-onset epileptic encephalopathy. Both variants in this case were on the same
SCN8A allele and one (S217P) was present in alternative exon 5N. Functional studies of the combined genotype in neonatal Na
V1.6 revealed gain-of-function, whereas studies of the T144S variant in adult Na
V1.6 showed WT-like properties. These data implicated the S217P as the pathogenic variant and provided proof-of-concept that induced splice-switching of exon 5N to exon 5A might alleviate the condition.
Conclusions:
Investigating the functional consequences of
SCN8A variants should consider the relevant splice isoform. In addition, an engineered variant designed to render the channel TTX resistant affects function of the adult Na
V1.6 isoform significantly suggesting that using this strategy to study pathogenic variants may give misleading results.
Funding:
This work was funded by NIH grant NS108874