Clinical and electrophysiological features of SCN8A variants causing episodic or chronic ataxia

Abstract

Background: Variants in SCN8A are associated with a spectrum of epilepsies and neurodevelopmental disorders. Ataxia as a predominant symptom of SCN8A variation has not been well studied. We set out to investigate disease mechanisms and genotype-phenotype correlations of SCN8A-related ataxia.

Methods: We collected genetic and electro-clinical data of ten individuals from nine unrelated families carrying novel SCN8A variants associated with chronic progressive or episodic ataxia. Electrophysiological characterizations of these variants were performed in ND7/23 cells and cultured neurons.

Findings: Variants associated with chronic progressive ataxia either decreased Na⁺ current densities and shifted activation curves towards more depolarized potentials (p.Asn995Asp, p.Lys1498Glu and p.Trp1266Cys) or resulted in a premature stop codon (p.Trp937Ter). Three variants (p.Arg847Gln and biallelic p.Arg191Trp/p.Asp1525Tyr) were associated with episodic ataxia causing loss-of-function by decreasing Na⁺ current densities or a hyperpolarizing shift of the inactivation curve. Two additional episodic ataxia-associated variants caused mixed gain- and loss-of function effects in ND7/23 cells and were further examined in primary murine hippocampal neuronal cultures. Neuronal firing in excitatory neurons was increased by p.Arg1629His, but decreased by p.Glu1201Lys. Neuronal firing in inhibitory neurons was decreased for both variants. No functional effect was observed for p.Arg1913Trp. In four individuals, treatment with sodium channel blockers exacerbated symptoms.

Interpretation: We identified episodic or chronic ataxia as predominant phenotypes caused by variants in SCN8A. Genotype-phenotype correlations revealed a more pronounced loss-of-function effect for variants causing chronic ataxia. Sodium channel blockers should be avoided under these conditions.

Funding: BMBF, DFG, the Italian Ministry of Health, University of Tuebingen.

Keywords: Chronic ataxia; Episodic ataxia; Loss-of-function; Patch-clamp; SCN8A.

Fig. 1

SCN8A variants associated with episodic or chronic ataxia. (a) Pedigrees of study participants. Empty symbols show unaffected individuals. For each individual, genetic testing results are shown where available: Alleles are shown as mutant (m), wildtype (+), or status unknown (blank). In individual 6, ovum donation (OD) resulted in dizygotic twins. Individuals 7 and 8 are monozygotic twins affected by biallelic variants. (b) Localization of variants in human Na_v1.6 channel. (c) Variants and their surrounding amino acids are highly evolutionarily conserved except p.Arg1913Trp.

Fig. 2

Functional studies of SCN8A wild-type (WT) and variants in ND7/23 cell. (a) Representative traces of Na⁺ current for WT and variants associated with chronic ataxia. (b) Peak Na⁺ currents normalised by cell capacitances were plotted versus voltage. (c) Voltage-dependent steady state activation and inactivation curves. Lines represent Boltzmann functions fit to the data points. (d) Time course of recovery from fast inactivation at −100 mV. (e) Voltage-dependence of the time constant of fast inactivation τ_h. (f) Representative traces of Na⁺ current for WT and variants associated with episodic ataxia. (g) Peak Na⁺ currents normalised by cell capacitances were plotted versus voltage. (h) Voltage-dependent steady state activation and inactivation curves. Lines represent Boltzmann function as in D. (i) Time course of recovery from fast inactivation at −100 mV. (j) Voltage-dependence of the time constant of fast inactivation τ_h. All data are shown as means ± 95% confidence interval. Numbers of recorded cells and statistical analysis are provided in Table 2.

Fig. 3

Intrinsic neuronal and firing properties of cultured hippocampal excitatory and inhibitory neurons transfected with SCN8A WT or mutant channels in absence of TTX. (a) Transfected hippocampal neurons indicated by CaMKII- or Dlx-GFP (green) were stained with a monoclonal anti-GAD67 antibody (red) and a monoclonal anti-CaMKII antibody (cyan). Scale bar: 20 μm. (b) Representative firing traces of evoked action potentials (APs) recorded in hippocampal excitatory neurons transfected with WT, p.Glu1201Lys (E1201K) or p.Arg1629His (R1629H), respectively. (c) Number of APs plotted versus injected current. WT, n = 16; E1201K, n = 12; R1629H, n = 16. (d) Resting membrane potential and (e) Input resistance of transfected excitatory neurons. (f) Threshold of first evoked AP. (g) Half-width of single AP. (h) Representative firing traces of evoked APs recorded in transfected hippocampal inhibitory neurons. (i) Number of APs plotted versus injected current. WT, n = 23; E1201K, n = 20; R1629H, n = 22. (j) Resting membrane potential and (k) Input resistance of transfected inhibitory neurons. (l) Threshold of first evoked AP. (m) Half-width of single AP. All data are presented as means ± 95% confidence interval. Detailed statistical analysis is provided in Table 3.

Fig. 4

Intrinsic neuronal and firing properties of cultured hippocampal excitatory and inhibitory neurons transfected with SCN8A WT or mutant channels in presence of TTX. (a) Representative firing traces of evoked action potentials (APs) recorded in hippocampal excitatory neurons transfected with WT, p.Glu1201Lys (E1201K) or p.Arg1629His (R1629H) respectively. (b) Number of APs plotted versus injected current. WT, n = 13; E1201K, n = 16; R1629H, n = 12. (c) Area under the curve (AUC) for the input–output relationships. E1201K decreased and R1629H increased AUC compared with WT. (d) Peak Na⁺ current amplitudes, (e) Resting membrane potential and (f) Input resistance of transfected excitatory neurons. (g) Representative traces of evoked single AP for WT and mutant channels recorded in transfected excitatory neurons. E1201K and R1629H decreased single AP amplitude (i), E1201K decreased threshold (h), and R1629H increased half-width(j) of single APs. (k) Representative firing traces of evoked APs recorded in hippocampal inhibitory neurons transfected with WT, E1201K and R1629H, respectively. (l) Number of APs plotted versus injected current. WT, n = 22; E1201K, n = 11; R1629H, n = 19. (m) Both variants decreased AUC of the input–output relationships compared with WT. (n) E1201K decreased peak Na⁺ current amplitudes, which is consistent with the increased input resistance (p) observed in transfected inhibitory neurons. (o) Resting membrane potential of inhibitory neurons transfected with WT or mutant channels. (q) Representative firing traces of evoked single action potential for WT and R1629H recorded in transfected inhibitory neurons. R1629H didn't affect AP threshold(r), but decreased AP amplitude(s) and increased AP half-width (t). All data are presented as means ± 95% confidence interval. One-way ANOVA with Dunnett's post hoc test or ANOVA on ranks with Dunn's post hoc test were performed. Detailed statistical analysis is provided in Table 4.