Molecular and cellular context influences SCN8A variant function

Pathogenic variants in SCN8A, which encodes the voltage-gated sodium (NaV) channel NaV1.6, associate with neurodevelopmental disorders, including developmental and epileptic encephalopathy. Previous approaches to determine SCN8A variant function may be confounded by use of a neonatally expressed, alternatively spliced isoform of NaV1.6 (NaV1.6N) and engineered mutations rendering the channel tetrodotoxin (TTX) resistant. We investigated the impact of SCN8A alternative splicing on variant function by comparing the functional attributes of 15 variants expressed in 2 developmentally regulated splice isoforms (NaV1.6N, NaV1.6A). We employed automated patch clamp recording to enhance throughput, and developed a neuronal cell line (ND7/LoNav) with low levels of endogenous NaV current to obviate the need for TTX-resistance mutations. Expression of NaV1.6N or NaV1.6A in ND7/LoNav cells generated NaV currents with small, but significant, differences in voltage dependence of activation and inactivation. TTX-resistant versions of both isoforms exhibited significant functional differences compared with the corresponding WT channels. We demonstrated that many of the 15 disease-associated variants studied exhibited isoform-dependent functional effects, and that many of the studied SCN8A variants exhibited functional properties that were not easily classified as either gain- or loss-of-function. Our work illustrates the value of considering molecular and cellular context when investigating SCN8A variants.


Fig. S5 .
Fig. S5.Window current determined for NaV1.6 variants.(A) Average deviation from WT NaV1.6 for window current area.All individual data points are plotted as open symbols and mean values are shown as larger filled symbols (n = 14-69).Error bars represent 95% CI.Data from NaV1.6A or NaV1.6N are indicated as red or blue symbols, respectively.Values to the right or left of the vertical dashed line (average normalized WT value) indicate larger (gain-of-function) or smaller (loss of function) window current, respectively.(B) Averaged Boltzmann fit lines for activation and steady-state inactivation curves of representative variants with smaller (I231T) or larger (Q417P) window current.(C) Volcano plot of mean values highlighting variants significantly different (P<0.01,horizontal dashed line) from WT. Symbols to the right of the vertical dashed line represent larger window current (gain-of-function).Only one variant (I231T) exhibited significantly smaller window current.Black symbols represent variants with no significant difference from WT. Quantitative data with statistical comparisons are provided in Supplemental Dataset S2 (NaV1.6N)and Supplemental Dataset S3 (NaV1.6A).

Fig. S6 .
Fig. S6.Recovery from inactivation determined for NaV1.6 variants.(A,B) Averaged time constants for recovery from inactivation displayed as fold-difference from WT channels recorded in parallel.Time constants (fast component plotted in panel A, slower component plotted in panel B) were determined by fitting the time course of recovery from inactivation to a double exponential function.All individual data points are plotted as open symbols and mean values are shown as larger filled symbols (n = 12-105 per variant).Error bars represent 95% confidence intervals.Data from NaV1.6A or NaV1.6N are indicated as red or blue symbols, respectively.Values to the right or left of the vertical dashed line (normalized WT value) represent larger (slower recovery) or smaller (faster recovery) time constants, respectively.(C,D) Volcano plots highlighting variants with significantly different (P <0.01, horizontal dotted line) fast component (C) and slower component (D) time constants of recovery from inactivation.Symbols to the left of the vertical dotted line denote slower recovery time course (lossof-function), while symbols to right indicate faster recovery time course (gain-of-function).Black symbols represent variants with no significant difference from WT. Quantitative data with statistical comparisons are provided in Supplemental Dataset S2 (NaV1.6N)and Supplemental Dataset S3 (NaV1.6A).

Fig. S7 .
Fig. S7.Use-dependent rundown of NaV1.6 variants.(A) Averaged use-dependent channel rundown at 20 Hz measured for NaV1.6 variants displayed as folddifference from WT channels recorded in parallel.All individual data points are plotted as open symbols and mean values are shown as larger filled symbols (n = 26-153 per variant).Error bars represent 95% confidence intervals.Data from NaV1.6A or NaV1.6N are indicated as red or blue symbols, respectively.Values to the left of the vertical dashed line (normalized WT value) represent greater rundown than WT.No variants exhibited lesser degree of rundown than WT.(B) Averaged currents normalized to first sweep amplitude measured for 30 sweeps at 20 Hz for select variants expressed in NaV1.6A.(C) Volcano plot of mean values highlighting variants with significantly different (P<0.01,horizontal dashed line) usedependent rundown from WT. Symbols to the left of the vertical dashed line represent greater rundown (loss-of-function and there were no variants with lesser rundown.Black symbols represent variants with no significant difference from WT. Quantitative data with statistical comparisons are provided in Supplemental Dataset S2 (NaV1.6N)and Supplemental Dataset S3 (NaV1.6A).

Fig. S8 .
Fig. S8.Comparison of functional properties among SCN8A variants.Radar plots depicting biophysical properties among variants compared to the WT channel.Individual radar plots represent different properties and each individual variant is represented as points (mean values normalized to WT) along each spoke.Red lines connecting each point represent data from NaV1.6A, blue lines represent data from NaV1.6N, and black lines indicate isoform-matched WT values.The scale indicating the magnitude of difference for each biophysical property is shown within the radar plot on the Y371C spoke, except for individual variants and specific properties (e.g., Window Current for Q417P and R1617L; Persistent Current for R1617L and N1768D; and Ramp Current for R1617L and N1768D).Persistent Current and Ramp Current were not determined (nd) for some variants.Quantitative data with statistical comparisons are provided in Supplemental Dataset S2 (NaV1.6N)and Supplemental Dataset S3 (NaV1.6A).

Fig. S9 .
Fig. S9.Heat map illustrating functional properties of NaV1.6 variants.Summary of individual functional properties for variants expressed in either NaV1.6A or NaV1.6N are illustrated as a heat map.Properties with gain-of-function are shaded as blue, and loss-of-function as red.Gray shaded boxes are properties that were not determined, and uncolored boxes indicate WT-like values.Only properties that reached the threshold for statistical significance (P<0.01) are color-coded.The intensity of shading reflects the degree of difference with WT.

Table S2 -Summary of major functional effects of SCN8A variants in this study
(differences between splice isoforms are in bold).