An SCN1B Variant Affects Both Cardiac-Type (NaV1.5) and Brain-Type (NaV1.1) Sodium Currents and Contributes to Complex Concomitant Brain and Cardiac Disorders

Voltage-gated sodium (NaV) channels are transmembrane proteins that initiate and propagate neuronal and cardiac action potentials. NaV channel β subunits have been widely studied due to their modulatory role. Mice null for Scn1b, which encodes NaV β1 and β1b subunits, have defects in neuronal development and excitability, spontaneous generalized seizures, cardiac arrhythmias, and early mortality. A mutation in exon 3 of SCN1B, c.308A>T leading to β1_p.D103V and β1b_p.D103V, was previously found in a patient with a history of proarrhythmic conditions with progressive atrial standstill as well as cognitive and motor deficits accompanying structural brain abnormalities. We investigated whether β1 or β1b subunits carrying this mutation affect NaV1.5 and/or NaV1.1 currents using a whole cell patch-clamp technique in tsA201 cells. We observed a decrease in sodium current density in cells co-expressing NaV1.5 or NaV1.1 and β1D103V compared to β1WT. Interestingly, β1bD103V did not affect NaV1.1 sodium current density but induced a positive shift in the voltage dependence of inactivation and a faster recovery from inactivation compared to β1bWT. The β1bD103V isoform did not affect NaV1.5 current properties. Although the SCN1B_c.308A>T mutation may not be the sole cause of the patient’s symptoms, we observed a clear loss of function in both cardiac and brain sodium channels. Our results suggest that the mutant β1 and β1b subunits play a fundamental role in the observed electrical dysfunction.


INTRODUCTION
Voltage-gated sodium (Na V ) channels are transmembrane proteins that initiate and propagate neuronal and cardiac action potentials (Catterall, 2000;Catterall et al., 2005). These channels have an α subunit composed of four domains (DI-DIV), each with six transmembrane segments (S1-S6). Na V α subunits are associated with two or more auxiliary β subunits (Kaplan et al., 2016). Na V β subunits modulate sodium channel biophysical properties in excitable and nonexcitable tissues and function as cell adhesion molecules, which are critical for extracellular/intracellular communication (Chen, 2004;Watanabe et al., 2008;Patino et al., 2011).
Five Na V β subunits have been identified in mammals: β1, β2, β3, β4 (encoded by SCN1B, SCN2B, SCN3B, and SCN4B, respectively), and β1b, a β1 splice variant. All β subunits with the exception of β1b are composed of an extracellular N-terminal immunoglobulin-like (Ig) domain (ECD), a transmembrane region, and an intracellular C-terminal domain (Brackenbury and Isom, 2011). Similar to the other β subunits, β1b is composed of an N-terminal region encoded by exons 1-3 of SCN1B. However, instead of a transmembrane domain, β1b has a different C-terminal region caused by partial retention of intron 3, which contains a stop codon (Hu et al., 2012). Apart from its structural differences, β1b is the only extracellularly secreted β subunit (Patino et al., 2011;O'Malley and Isom, 2016). However, it is predicted to have the same cellular interactions as the β1 subunit because it shares the same ECD .
In a series of experiments using Na V β subunit chimeras, McCormick et al. (1999) demonstrated that the ECD is necessary and sufficient for sodium channel modulation (McCormick et al., 1999). Mutations in the ECD of the β1 subunit have been associated with brain diseases, such as epileptic encephalopathy and genetic (or generalized) epilepsy with febrile seizures plus (Wallace et al., 1998;Audenaert et al., 2003;Scheffer et al., 2007;Brackenbury and Isom, 2011;Darras et al., 2019). Some of these mutations have been proven to disrupt the interaction between Na V 1.1 and the β1 subunit, impairing excitability or reducing cell-cell interactions (Meadows et al., 2002;Chen, 2004;Spampanato, 2004;Patino and Isom, 2010). Although Na V α subunit alone is sufficient to form a conducting channel, β1 subunits enhance protein expression and modulate gating and voltage dependency (McCormick et al., 1999;Meadows et al., 2002;Yu and Catterall, 2003;Olesen et al., 2012). Scn1b-null mice have defective neuronal development and excitability, present spontaneous generalized seizures and cardiac arrhythmias, and die by postnatal day 21 (Chen, 2004;Patino and Isom, 2010;Brackenbury and Isom, 2011;Lin et al., 2015).
The regulatory effect of the β1 WT subunit on Na V 1.5 has been widely studied. SCN1B mutations are reported to cause diseases such as Brugada syndrome (Watanabe et al., 2008;Peeters et al., 2015), long QT syndrome (Riuró et al., 2014), and atrial fibrillation (Olesen et al., 2012), demonstrating the critical effect of β1 on cardiac excitability. In addition, Lopez-Santiago et al. (2007) showed that Scn1b-null mice display prolonged QT and RR intervals. In addition, mutations in β1b have been associated with cardiac arrhythmias, such as Brugada syndrome, long QT syndrome, cardiac conduction disease, and SIDS (Watanabe et al., 2008;Hu et al., 2012;Riuró et al., 2014).
We investigated a mutation located in exon 3 of SCN1B [Chr19:35524503 A>T (hg19): NM_001037.4 c.308A>T] identified in a patient with a history of proarrhythmic conditions with progressive atrial standstill, cognitive and motor deficits, and structural brain abnormalities. The patient was part of a sequencing study by Eldomery et al. (2017) that proposed a dual diagnosis for this case. As a result, this β1 variant (β1 D103V ) (dbSNP ID: 1057519457) is indexed as pathogenic in the ClinVar database 1 . However, this variant has not been functionally characterized.
The mutation leads to substitution of an aspartic acid for a valine at position 103 of the protein (p.D103V). Exon 3 is part of the Ig-like loop of the ECD, which is common between β1 and β1b. Therefore, this substitution is present in both β1 and β1b subunit isoforms at a highly conserved position. Thus, this mutation likely affects the function of both β1 and β1b proteins, and could play a role in the patient's pathophysiologic phenotype. This work characterizes the effects of mutant β1 and β1b subunits on the biophysical properties of both cardiac (Na V 1.5) and neuronal (Na V 1.1) sodium channels.

Expression Vectors and Site-Directed Mutagenesis
The pCMV vector harboring SCN1A was a generous gift from Dr. Alfred George Jr. (Vanderbilt University Division of Genetic Medicine, Nashville, TN, United States). The pcDNA3 vector harboring the complementary DNA (cDNA) of human SCN5A was a generous gift from Dr. Matteo Vatta (Baylor College of Medicine, Houston, TX, United States).
The inserts were sequenced to verify presence of the desired mutation and absence of unwanted variations.

Electrophysiological Studies
Sodium currents from cells displaying green fluorescence were studied at room temperature using the patch clamp whole cell configuration. The bath solution contained (in mmol/l): 140 NaCl, 3 KCl, 10 N-2-hydroxyethylpiperazine-N -2-ethanesulfonic acid (HEPES), 1.8 CaCl 2 and 1.2 MgCl 2 , pH 7.4 (NaOH). The pipette solution contained (in mmol/l): 130 CsCl, 1 ethylene glycol-bis(2-amino-ethylether)-N,N,N ,Ntetra-acetic acid (EGTA), 10 HEPES, 10 NaCl and 2 ATP-Mg 2+ , pH 7.2 (CsOH). Osmolality of bath and pipette solutions was adjusted with glucose to 325 and 308 mOsm, respectively. Pipettes were pulled from borosilicate glass capillaries (Sutter Instrument, Novato, CA, United States) with a Narishige PC-10 puller (Narishige International LTD, London, United Kingdom), and their resistance ranged 2-3.5 M when filled with pipette solution. Voltage clamp experiments were conducted with an Axopatch 200B amplifier and pClamp10.2/Digidata 1440A acquisition system (Axon Instruments, Molecular Devices, Sunnyvale, CA, United States) at a sampling rate of 20 kHz. Leak subtraction was not used. OriginPro8 software was used for data analysis and statistics (OriginLab Corporation, Northampton, MA, United States). Data was filtered at 5 kHz. Series resistance compensation of 80-90% was used when necessary. To permit current stabilization, recordings were performed at least 5 min after entering into the whole cell configuration. Membrane potentials were not corrected for junction potentials that arose between the bath and pipette solution.
Sodium currents were studied with voltage-clamp step protocols. To determine the current-voltage relationship and voltage dependence of activation, cells were held at -120 mV and currents were elicited with depolarizing pulses of 50 ms from -80 to 80 mV in 5 mV increments. The voltage dependence of inactivation was determined by applying 50-ms prepulses from -140 to 5 mV in 5 mV increments, followed by a -20 mV test pulse. Recovery from inactivation was studied with a twopulse protocol consisting of a first -20 mV pulse of 50 ms from a holding potential of -120 mV, followed by a -120 mV interpulse of varying duration (1-40 ms), and a second -20 mV pulse of 50 ms. Current density was obtained by normalizing the current at each potential by cell capacitance. Activation and steady-state inactivation data were fitted to a Boltzmann equation of the form G = G max /(1+exp(V 1/2 -V)/k) and I = I max /(1+exp(V-V 1/2 )/k), respectively, where G is conductance, Gmax is maximum conductance, V 1/2 is voltage at which half of channels are activated or inactivated, V is membrane potential, k is slope factor, I is peak current amplitude, and I max is maximum current amplitude. Recovery from inactivation data were fitted to a mono-exponential function to obtain the time constant.

Western Blot
Cells were plated and transfected as above. Forty-eight hours after transfection, cells were washed three times with Dulbecco's phosphate-buffered saline (DPBS) and scraped in Triton X-100 lysis buffer containing 1% Triton X-100, 50 mM Tris/HCl pH 7.4, 150 mM NaCl, 1 mM EDTA and cOmplete protease inhibitor cocktail (Roche, Madrid, Spain). Lysates were obtained after 1 h rotating at 4 • C, and insoluble materials were removed by centrifugation. Proteins were quantified using a Pierce BCA protein assay kit (Thermo Scientific, Rockford, IL, United States) and resolved along with a protein marker (PageRuler Plus prestained protein ladder, Thermo Scientific) in 4-15% Mini-PROTEAN TGX Stain-Free precast gels (Bio-Rad Laboratories, Hercules, CA, United States). These gels include a trihalo compound that react with tryptophan residues in proteins, and allow rapid fluorescent detection of proteins in gels or on membranes without staining. Proteins were transferred to PVDF membranes (GE Healthcare Life Sciences, Chicago, IL, United States) overnight at 4 • C. Gels were developed by exposure to UV light before and after protein transfer to the membrane. Membranes were probed with either a rabbit anti-human Na V 1.5 antibody (Alomone Labs) at a 1:1000 dilution for 1 h at room temperature, or a rabbit anti-human Na V 1.1 antibody (Alomone Labs, Jerusalem, Israel) at a 1:100 dilution overnight at 4 • C. Membranes were further incubated with a secondary horseradish peroxidase-conjugated anti-rabbit antibody (Thermo Scientific) at a 1:10000 dilution for 1 h at room temperature. Chemiluminescent signal was obtained with Pierce ECL western blotting substrate (Thermo Scientific) and detected using standard X-ray films. Expression of Na V 1.1 and Na V 1.5 was quantified from digitized X-ray film images using ImageJ software (National Institutes of Health, 2 ). Membrane intensity values for each sample were normalized by total lane density before protein transfer (Taylor et al., 2013).

Cell Surface Protein Biotinylation
Cell surface protein biotinilation was performed following the protocol from Tarradas et al. (2013). Briefly, cells were plated FIGURE 1 | The protein region that includes the mutation is highly conserved. Sequence alignments of the region shared by β1 and β1b proteins containing the D103V mutation were performed using Uniprot. The position of the last amino acid of each sequence, and the reference for each protein according to Uniprot are indicated on the left side of each sequence. (A) Sequence alignment of human voltage-gated sodium channel β-subunit family members. (B) Sequence alignment of human voltage-gated sodium channel β1-subunit of different species. and transfected as described above. Twenty-four hours after transfection, cells were washed twice with DPBS supplemented with 0.9 mM CaCl 2 and 0.49 mM MgCl 2 (DPBS + ). Membrane proteins were biotinylated by incubating cells with 1 mg/ml of EZ-Link Sulfo-NHS-SS-Biotin (Thermo Scientific) in DPBS + for 30 min at 4 • C. Cells were then washed three times in DPBS + with 100 mM glycine, and scrapped in Triton X-100 lysis buffer [1% Triton X-100, 50 mM Tris/HCl pH 7.4, 150 mM NaCl, 1 mM EDTA and Complete Protease Inhibitor Cocktail (Roche, Madrid, Spain)]. Lysates were obtained after 1 h rotating at 4 • C. Insoluble materials were removed by centrifugation. 10% of the supernatants were kept at −80 • C (input samples), the rest (pull-down samples) were incubated with Ultralink Immobilized NeutrAvidin resin (Pierce, Thermo Scientific) overnight at 4 • C. The resin was precipitated and washed with Triton X-100 lysis buffer, then in saline solution (5 mM EDTA, 350 mM NaCl and 0.1% TX-100 in DPBS + ) and finally in 50 mM Tris/HCl pH 7.4, 150 mM NaCl and 1 mM EDTA. Input and pull-down samples were resuspended in SDS-PAGE loading buffer and heated for 5 min at 95 • C. Proteins were resolved in 7.5% acrylamide gels using TGX Stain-Free FastCast Acrylamide Kit (Bio-Rad Laboratories) and transferred overnight at 4 • C to PVDF membranes (Millipore, Billerica, MA, United States). Protein bands using Stain-Free gels were visualized by exposure to UV light before electroblotting to PVDF membranes. Membranes were probed with a rabbit anti-human Na V 1.5 antibody (Alomone Labs) at a 1:1000 dilution either for 1 h at room temperature or overnight at 4 • C. After washing, the membrane was further incubated with a secondary horseradish peroxidase-conjugated antibody (Thermo Scientific, Rockford, IL, United States) at a dilution of 1:10000 for 1 h at room temperature. Chemiluminescent signal was obtained with Clarity Western ECL Substrate (Bio-Rad Laboratories) and detected using the ChemiDoc Imaging System (Bio-Rad Laboratories). Expression of Na V 1.5 was quantified using ImageJ software (National Institutes of Health, 3 ). Membrane intensity values for each sample were normalized by total lane density before protein transfer (Taylor et al., 2013).

Statistical Analisys
OriginPro 2019 software was used for statistical analysis. Results are presented as means ± standard error. Statistical comparisons 3 https://imagej.nih.gov/ij were performed using one-way ANOVA with a post hoc Tukey test. Differences were considered statistically significant at p < 0.05.

An SCN1B Mutation Found by Whole Exome Analysis Is Linked to Cardiac and Brain Dysfunction
The proband, currently an 8-year-old male, was part of a whole exome sequencing study by Eldomery et al. (2017). This study revealed that the proband was heterozygous for the SCN1B missense mutation c.308A>T. This mutation results in an amino acid change from a hydrophilic aspartic acid to a hydrophobic valine at position 103 (p.D103V) of the sodium channel auxiliary β subunit. D103 is probably important to channel function, as it is highly conserved among all sodium channel β subunits, except for β4, and is also conserved among different species (Figures 1A,B). Eldomery et al. indexed the β1 subunit variant D103V as pathogenic (Eldomery et al., 2017).
In addition to the SCN1B variant, inherited from the child's father, genetic analysis revealed that the proband was heterozygous for two pathogenic variants in POLR1C: c.614delG (p.G205Afs * 49), for which his mother was heterozygous, and c.88C>T (p.P30S), for which his father was heterozygous.
The clinical picture of the patient is complex and includes cardiac and neurological impairment. Arrhythmogenic features presented early in the neonatal stage. These included bradycardia from variable atrioventricular (AV) conduction disturbance along with evidence of a borderline prolongation of QTc interval, which appeared to be secondary to intraventricular conduction delay. This was treated with propranolol. Ventricular dysfunction was observed during the neonatal period but normalized at 18 months of age with medical management. Follow-up evaluations showed poorly discernible atrial activity and intraventricular conduction delays on surface electrocardiograms (ECG). The child developed recalcitrant tachycardia at 7 years of age, identified as cavo-tricuspid isthmus (CTI)-dependent macro-reentrant atrial tachycardia (cycle length ∼420 ms) with 1:1 AV conduction as well as 2:1 AV conduction during intracardiac electrophysiology study. At baseline, he was variably in junctional rhythm with brief periods of atrially mediated rhythm. Normal AV conduction Frontiers in Cell and Developmental Biology | www.frontiersin.org FIGURE 3 | β1 WT and β1b WT modulate Na V 1.5 current. Biophysical properties of sodium currents measured from HEK-293T cells expressing Na V 1.5 alone, or co-expressed with either β1 WT or β1b WT . Whole cell currents were elicited by depolarizing potentials as shown in the insets. Circles are used to depict data for Na V 1.5 alone, triangles for Na V 1.5+β1 WT , and squares for Na V 1.5+β1b WT . Solid lines in panels (C) and (D) represent the fitted curves. Values are expressed as mean ± SEM. The studies with the β1 subunit are represented on the left side of the figure, and the studies with the β1b subunit on the right side. (A) Representative whole-cell Na + current traces from HEK-293T cells expressing Na V 1.5 alone (top left and top right), Na V 1.5+β1 WT (bottom left), and Na V 1.5+β1b WT (bottom right).
(B) Mean current-voltage relationship. I Na amplitude was normalized by the cell capacitance to obtain I Na density values. (C) I Na steady-state voltage dependence of activation and inactivation plots. (D) Recovery from inactivation curves. Data is presented as Mean ± SE. I Na = sodium current; n = number of cells; k = slope factor; V 1/2 = voltage for half-maximal activation or steady-state inactivation; τ = time constant. *vs Na V 1.5. Significantly different, p-value < 0.05. Data is presented as Mean ± SE. I Na = sodium current; n = number of cells; k = slope factor; V 1/2 = voltage for half-maximal activation or steady-state inactivation; τ = time constant. * vs Na V 1.5. Significantly different, p-value < 0.05.
was seen at baseline, with atrial pacing that worsened to 2:1 AV conduction on isoproterenol. Three-dimensional voltage mapping at baseline rhythm revealed extensive scarring of the right atrium and coronary sinus. Resting surface electrocardiogram from index patient (Figure 2A) shows poorly discernible atrial activity with low amplitude P wave, intraventricular conduction delay (QRS duration 114 ms) as well as prolonged QTc secondary to QRS abnormality (uncorrected QTc 477 ms). Intracardiac three-dimensional voltage map from index patient is contrasted against a healthy age-matched peer ( Figure 2B). The patient underwent placement of a transvenous dual-chamber pacemaker. He was programmed in the VVI pacing mode due to progressive atrial undersensing and unreliable atrial capture. By the 2 year of life, the patient also had global developmental delay and increased tone in the lower extremities. Neurological examination at age five demonstrated evidence of significant cognitive deficits (smiled socially, followed simple commands, gave one-word answers to questions), persistent large-amplitude horizontal nystagmus with primary gaze, normal tone in the upper extremities with mixed spasticity and dystonia in the lower extremities (spastic catches and intermittent dystonic extension and inward rotation of the lower extremities when excited). He demonstrated symmetrically brisk deep tendon reflexes with Babinski and Rossolimo signs, and significant dysmetria on the finger-to-nose test as well as axial ataxia affecting the neck and trunk. He sat with support and bore weight with assistance, with scissoring of the legs, and did not ambulate. Brain MRI revealed extensive polymicrogyria involving the bilateral cerebral hemispheres, extensive T2 hyperintense signal throughout the supratentorial white matter with less in the cerebellar white matter and dorsal brainstem, and markedly delayed myelination.
To date, no functional studies of the SCN1B mutation D103V have been published. Because this SCN1B variant was flagged as the best candidate for electrical dysfunction observed in the patient, we comprehensively analyzed the effects of this variant on heart and brain sodium currents. Na V 1.5 Current Properties Are Modified by β1 WT and β1b WT Subunits Modulation of sodium channels by β subunits has been reported by different groups (Brackenbury and Isom, 2011;O'Malley and Isom, 2016). Our goal was to determine whether mutant β1 (β1 D103V ) and β1b (β1b D103V ) subunits affected Na V 1.5 and Na V 1.1 sodium currents (I Na ). We first performed functional characterization of the sodium current in HEK-293T cells expressing Na V 1.5 alone or co-expressed with either β1 WT (Na V 1.5+β1 WT ) or β1b WT (Na V 1.5+β1b WT ) subunits ( Figure 3A). Co-expression of Na V 1.5 with either β1 WT or β1b WT increased peak I Na density (41.2 and 42.8%, respectively) compared to Na V 1.5 alone ( Figure 3B and Tables 1, 2). Also, the β1 WT subunit induced a negative shift of 5 mV in the voltage dependence of activation compared to Na V 1.5 alone ( Figure 3C left and Table 1). No changes were observed in voltage dependence of activation upon expression of Na V 1.5+β1b WT (Figure 3C right and Table 2). Voltage dependence of inactivation was not altered by either β1 WT or β1b WT subunits. FIGURE 4 | β1 D103V modifies the gating properties of Na V 1.5 channel. Biophysical properties of sodium currents measured from HEK-293T cells expressing Na V 1.5 alone, or co-expressed with either β1 WT or β1 D103V . Whole cell currents were elicited by depolarizing potentials as shown in the insets. Open triangles are used to depict data for Na V 1.5+β1 WT and filled triangles for Na V 1.5+β1 D103V . Solid lines in panels (C) and (D) represent the fitted curves. Values are expressed as mean ± SEM. (A) Representative whole-cell Na + current traces from HEK-293T cells expressing Na V 1.5+β1 WT (top), and Na V 1.5+β1 D103V (bottom). (B) Mean current-voltage relationship. I Na amplitude was normalized by the cell capacitance to obtain I Na density values. (C) I Na steady-state voltage dependence of activation and inactivation plots. (D) Recovery from inactivation curves. (E) Western blot detection of Na V 1.5 (left) and corresponding total protein stain-free gel (right), from HEK-293T cells transfected with either Na V 1.5, Na V 1.5+β1 WT , or Na V 1.5+β1 D103V or non-transfected cells (NT) (n = 2). (F) Western blot detection of Na V 1.5 after cell surface biotinylation (left) and corresponding total protein stain-free gel (right), from HEK-293T cells transfected with either Na V 1.5, Na V 1.5+β1 WT , or Na V 1.5+β1 D103V or non-transfected cells (NT). (G) Bar graph depicts the relative protein expression normalized by the Na V 1.5 α subunit of total protein (left, n = 2) and biotinylated protein (right, n = 4) from HEK-293T cells transfected with either Na V 1.5+β1 WT or Na V 1.5+β1 D103V . Both visible bands from the biotinylated samples were used for quantification, and the ratio was obtained by normalizing each condition with its respective input.
Frontiers in Cell and Developmental Biology | www.frontiersin.org Data is presented as Mean ± SE. I Na = sodium current; n = number of cells; k = slope factor; V 1/2 = voltage for half-maximal activation or steady-state inactivation; τ = time constant. * vs Na V 1.5+β1 WT . Significantly different, p-value < 0.05.
FIGURE 5 | Na V 1.5 current properties are not modified by β1b D103V . Biophysical properties of sodium currents measured from HEK-293T cells expressing Na V 1.5 co-expressed with either β1b WT or β1b D103V . Whole cell currents were elicited by depolarizing potentials as shown in the insets. Open squares are used to depict data for Na V 1.5+β1 WT and filled squares for Na V 1.5+β1 D103V . Solid lines in panels (C) and ( Both Na V 1.5+β1 WT and Na V 1.5+β1b WT currents displayed faster recovery from inactivation compared to Na V 1.5 alone ( Figure 3D and Tables 1, 2).
The Mutant β1 D103V Subunit Decreases Na V 1.5 Sodium Current Density To determine whether β1 D103V affected the cardiac Na V 1.5 sodium current, we co-expressed Na V 1.5 with either the β1 WT or β1 D103V subunit ( Figure 4A). Na V 1.5+β1 D103V decreased sodium current density by 45.7% compared to Na V 1.5+β1 WT (Figure 4B). β1 D103V induced a 3.37-mV positive shift of the voltage dependence of activation compared to β1 WT . The voltage dependence of steady state inactivation was similar between Na V 1.5+β1 WT and Na V 1.5+β1 D103V (Figure 4C). Likewise, the recovery from inactivation time constants were similar when I Na was measured in both experimental conditions ( Figure 4D and Table 3). Data is presented as Mean ± SE. I Na = sodium current; n = number of cells; k = slope factor; V 1/2 = voltage for half-maximal activation or steady-state inactivation; τ = time constant. * vs Na V 1.5+β1b WT . Significantly different, p-value < 0.05.
FIGURE 6 | β1 WT and β1b WT modulate Na V 1.1 current. Biophysical properties of sodium currents measured from HEK-293T cells expressing Na V 1.1 alone, or co-expressed with either β1 WT or β1b WT . Whole cell currents were elicited by depolarizing potentials as shown in the insets. Circles are used to depict data for Na V 1.1 alone, triangles for Na V 1.1+β1 WT , and squares for Na V 1.1+β1b WT . Solid lines in panels (C) and (  Data is presented as Mean ± SE. I Na = sodium current; n = number of cells; k = slope factor; V 1/2 = voltage for half-maximal activation or steady-state inactivation; τ = time constant. * vs Na V 1.1. Significantly different, p-value < 0.05. Notably, we did not detect any sodium current in 34 of 75 cells transfected with Na V 1.5+β1 D103V . However, only 11 of 81 cells expressing Na V 1.5+β1 WT had no detectable I Na . Nevertheless, β1 D103V mutation did not affect protein expression. We observed approximately the same amount of total Na V 1.5 in all three conditions (Na V 1.5 alone, Na V 1.5+β1 WT , and Na V 1.5+β1 D103V ; Figures 4E,G). Moreover, the amount of Na V 1.5 in the plasma membrane was unaffected by the β1 mutation. To determine this we performed immunoblotting analysis of biotinylated surface membrane proteins. Figures 4F,G show that the relative amount of plasma membrane Na V 1.5+β1 WT and Na V 1.5+β1 D103V channels was similar.
The Mutant β1b D103V Isoform Does Not Modify Na V 1.5 Properties We finally evaluated the effects of β1b D103V on Na V 1.5 sodium current properties. The β1b D103V mutation did not modify I Na density compared to Na V 1.5+β1b WT . Likewise, voltage dependence of activation, steady state inactivation, and recovery from inactivation were not altered by mutation compared to the wildtype subunit ( Figure 5 and Table 4).

The β1 Subunit Increases Na V 1.1 Sodium Current Density
Because the proband also experienced brain pathologies, we investigated whether the D130V mutation in β1 and β1b isoforms also affected the most predominant brain-type sodium current: Na V 1.1. We performed functional characterization of the sodium current in HEK-293T cells expressing Na V 1.1 alone or coexpressed with β1 WT or β1b WT subunits.
We measured macroscopic sodium currents (I Na ) at varying potentials from these transfected cells (Figure 6A). There was a 45.9% increase of peak I Na density when Na V 1.1 was co-expressed with β1 WT compared to Na V 1.1 alone. However, co-expression of Na V 1.1 with β1b WT did not significantly increase I Na density compared to Na V 1.1 alone ( Figure 6B and Table 5). Neither β1 WT nor β1b WT caused any significant changes in the voltage dependence of activation (Table 5). Steady-state inactivation of the Na V 1.1 current was not altered by β1 WT . On the contrary, co-expression of the β1b WT isoform caused a 7.21 mV negative shift of the voltage dependence of inactivation compared to Na V 1.1 alone ( Figure 6C and Table 5). β1 WT markedly decreased the recovery from inactivation time in comparison to Na V 1.1 alone. However, this decrease was not observed when Na V 1.1 was co-expressed with β1b WT (Figure 6D and Table 5).
The β1 D103V Subunit Decreases Na V 1.1 Current Density To determine whether the β1 mutation identified in the patient had an effect on the neuronal Na V 1.1 sodium current, we performed functional characterization in HEK-293T cells expressing Na V 1.1+β1 D103V compared to those expressing Na V 1.1+β1 WT . Peak current density measured from cells expressing Na V 1.1+β1 D103V was 66.9% smaller than that in Na V 1.1+β1 WT -expressing cells (Figures 7A,B and Table 6). Voltage dependence of activation and steady-state inactivation were not altered by β1 D103V (Figure 7C and Table 6). Recovery from inactivation time constants were similar in both experimental conditions ( Figure 7D and Table 6).
Similar to Na V 1.5, we did not detect any current in 23 out of 59 cells expressing Na V 1.1+β1 D103V . However, when cells were transfected with Na V 1.1+β1 WT , only 4 of 56 cells showed no current. Western blot analysis showed no significant differences in Na V 1.1 expression between the three conditions ( Figure 7E).
β1 Mutant Isoform b (β1b D103V ) Modifies Na V 1.1 Properties Because the D103V mutation is located in a region shared by both β1 and β1b subunits, we assessed the effect of β1b D103V on Na V 1.1 current properties. We co-expressed Na V 1.1 with either β1b WT or β1b D103V subunits ( Figure 8A). Co-expression of β1b D103V with Na V 1.1 did not change I Na density compared to Na V 1.1+β1b WT (Figure 8B and Table 7). Likewise, the β1b D103V subunit did not significantly change the voltage dependence of activation compared to β1b WT . However, β1b D103V caused a 6.04-mV right shift of the voltage dependence of steady-state inactivation when compared to the wildtype subunit ( Figure 8C). This change resulted in a positive shift of the window current of Na V 1.1+β1b D103V compared to Na V 1.1+β1b WT (Figure 8D and Table 7). β1b D103V significantly reduced the recovery from inactivation time constant compared to Na V 1.1+β1b WT (Figure 8E and Table 7).

DISCUSSION
We present functional characterization of an SCN1B missense variant (c.308A>T; p.D103V) affecting both β1 and β1b regulatory subunits. This mutation was found in a newborn with heart and brain pathologies. The mutation was previously reported as pathogenic by Eldomery et al. (2017), potentially FIGURE 7 | β1 D103V significantly decreases peak I Na on Na V 1.1. Biophysical properties of sodium currents measured from HEK-293T cells expressing Na V 1.1 co-expressed with either β1 WT or β1 D103V . Whole cell currents were elicited by depolarizing potentials as shown in the insets. Open triangles are used to depict data for Na V 1.1+β1 WT and filled triangles for Na V 1.1+β1 D103V . Solid lines in panels (C) and (D) represent the fitted curves. Values are expressed as mean ± SEM. (A) Representative whole-cell Na + current traces from HEK-293T cells expressing Na V 1.1+β1 WT (top), and Na V 1.1+β1 D103V (bottom). (B) Mean current-voltage relationship. I Na amplitude was normalized by the cell capacitance to obtain I Na density values. (C) I Na steady-state voltage dependence of activation and inactivation plots. (D) Recovery from inactivation curves. (E) Western blot detection of Na V 1.1 (left) and corresponding total protein stain-free gel (middle), from HEK-293T cells transfected with either Na V 1.1, Na V 1.1+β1 WT , or Na V 1.1+β1 D103V or non-transfected cells (NT). Bar graph on the right depicts the relative protein expression normalized by the Na V 1.1 alone (n = 2). Data is presented as Mean ± SE. I Na = sodium current; n = number of cells; k = slope factor; V 1/2 = voltage for half-maximal activation or steady-state inactivation; τ = time constant. * vs Na V 1.1+β1 WT . Significantly different, p-value < 0.05.
causing cardiomyopathies and intellectual disability among other phenotypes. The goal of our study was to determine whether the mutant β1 D103V and β1b D103V subunits affected Na V 1.5 and Na V 1.1 sodium currents.
Effects of β1 D103V but Not β1b D103V Impair Normal Na V 1.5 Function HEK cells expressing Na V 1.5+β1 WT showed increased I Na , consistent with Watanabe et al. (2008) and Qu et al. (1995). Also, we observed a shift of the voltage dependence of activation toward hyperpolarizing potentials in agreement with Ko et al. (2005) and Yuan et al. (2014). Lastly, we found a reduced recovery from inactivation time constant, consistent with Fahmi et al. (2001). Mutations in the β1 subunit have been implicated in various cardiac arrhythmias, such as Brugada syndrome, atrial fibrillation, sudden infant death syndrome (SIDS), long QT syndrome, and cardiac conduction defect (Audenaert et al., 2003;Watanabe et al., 2008;Tan et al., 2010;Ricci et al., 2014;O'Malley and Isom, 2016). We show that the β1 D103V subunit does not increase I Na as β1 WT does. In addition, the mutant subunit shifted the voltage dependence of activation toward more positive potentials, suggesting a physical interaction between β1 D103V and the alpha subunit. This interaction, however, appears not to disrupt Na V 1.5 trafficking to the plasma membrane. Our data does not support this interpretation as relative plasma membrane protein levels are similar in both β1 WT and β1 D103V conditions. Thus the mutant β1 is likely to exert its effect over Na V 1.5 is directly on the channel electrical properties. Taking into account that the β1 subunit is normally present in cardiac cells, this difference between β1 WT and β1 D103V would be considered a loss-of-function of the channel, consistent with the progressive atrial standstill and cardiac conduction disorder observed in the patient. It is necessary to point out that the patient is heterozygous for the β1 mutation. Thus, whether the mutation has a dominant negative effect remains to be determined. Further experiments in either patient-specific iPS derived cardiomyocytes or heterocygous heterologous expression would be needed to further explore this possibility.
The effect of the β1b WT subunit on Na V 1.5 current has not been well-studied. However, some groups have reported presence of the β1b WT subunit in human and rat adult hearts, and expression levels of β1b in the atria and ventricle are greater than those of β1 (Kazen-Gillespie et al., 2000;Yuan et al., 2014). We observed an increase in Na V 1.5 current density in the presence of β1b WT , consistent with previous works, as well as a decrease in the recovery from inactivation time constant. Although Na V 1.5 current was diminished by the presence of the β1b D103V subunit, acceleration in the recovery from inactivation was still observed with β1b D103V . This suggests that the interaction between Na V 1.5 and the mutant β1b subunit is still present.
Effects of β1 D103V on Na V 1.1 Properties Could Impair Normal Neuronal Activity Our results for neuronal Na V 1.1 current show that β1 WT increases Na V 1.1 I Na density, in agreement with findings from Isom et al. (1992). In addition, we detected faster recovery from inactivation, as previously reported by Aman et al. (2009) and Barela (2006). Further, we show that the mutant β1 subunit strongly decreases Na V 1.1 sodium current density compared to the wildtype β1 subunit, which represents a loss-of-function of the channel, in agreement with Meadows et al. (2002). β1b D103V Has a Modulatory Effect Over Na V 1.1 Function The modulatory effect of the β1b subunit on Na V 1.1 is poorly studied and controversial. Some groups propose that β1b expression predominates in embryonic development and early life and is thus essential for brain development (Kazen-Gillespie et al., 2000;Patino et al., 2011;O'Malley and Isom, 2016). Patino et al. (2011) studied an epilepsy-related mutation that only affects the β1b subunit. Although their co-immunoprecipitation studies did not detect an association between β1b and Na V 1.1 or Na V 1.3 channels, they showed that β1b modulates the Na V 1.3 sodium current in heterologous systems. Our study shows that the β1b WT subunit modulates voltage dependence of inactivation by shifting the curve to more negative potentials.
While β1b D103V does not have any effects on Na V 1.1 current density, we found that it shifts the voltage dependence of inactivation in the depolarizing direction, thus causing the channel to be available for activation in a larger voltage range. In addition, β1b D103V strongly reduces the recovery from inactivation time constant compared to β1b WT . Thus, mutant channels are ready to be activated earlier than those without mutation. Assuming that β1b is normally present and interacts with Na V 1.1 in native tissue, both effects would contribute to gain-of-function of the channel.
β1 subunit mutations that induce a gain-of-function of Na V 1.1 have also been reported to cause brain dysfunction (Meadows et al., 2002;Kruger et al., 2016). Thus, it is possible that β1b D103V causes neuron hyperexcitability, destabilizing normal neuronal behavior of the brain. Also, given the importance of this subunit to brain development (O'Malley and Isom, 2016), this mutation FIGURE 8 | β1b D103V modifies the gating properties of Na V 1.1 channel. Open squares are used to depict data for Na V 1.1+β1b WT and filled squares for Na V 1.1+β1b D103V . Solid lines represent the fitted curves. Values are expressed as mean ± SEM. Sodium currents were obtained with voltage clamp step protocols as shown in the insets. (A) Representative whole-cell Na + current traces from HEK-293T cells expressing Na V 1.1+β1b WT (top), and Na V 1.1+β1b D103V (bottom).
(B) Mean current-voltage relationship. I Na amplitude was normalized by the cell capacitance to obtain I Na density values. (C) I Na steady-state voltage dependence of activation (right) and inactivation (left). (D) Window region bounded by the steady state activation and inactivation voltage dependent curves (Na V 1.1+β1b WT solid gray and Na V 1.1+β1b D103V diagonal pattern). (E) Recovery from inactivation curves. Data is presented as Mean ± SE. I Na = sodium current; n = number of cells; k = slope factor; V 1/2 = voltage for half-maximal activation or steady-state inactivation; τ = time constant. * vs Na V 1.1+β1b WT . Significantly different, p-value < 0.05. may impair brain formation at embryonic stages, thus provoking the brain phenotype at an early age, and may be involved in the patient's polymicrogyria.
Our results showed that the modulatory effects of β1 D103V and β1b D103V on Na V 1.1 I Na are different, even opposing each other. Considering that β1b is predominantly expressed during embryonic development, we expect that the gain-of-function effect of β1b D103V is more important during that stage. A lossof-function caused by β1 D103V would be more important later in development. Since the p.D103V mutation affects both β1 and β1b isoforms, it could differentially compromise neuronal electrical activity during development.
In conclusion, our results strongly suggest that the SCN1B_c.308A>T mutation contributes to the patient's phenotype. We show that the β1 D103V mutant channels cause a loss-of-function of the cardiac-type sodium current, which could explain the clinical presentation of progressive atrial standstill, intra-atrial reentrant tachycardia, and cardiac conduction disorder in the child.
We surmise that the SCN1B variant could contribute to the patient's brain phenotype at two different stages. During development and early life, when the β1b isoform is predominant, β1b D103V leads to a gain-of-function of the channel. During adulthood, when the β1 subunit is predominantly expressed, β1 D103V produces loss-of-function of the sodium current. Both effects could contribute to the cognitive and motor deficits observed in the patient.
This family was part of a sequencing re-analysis project from which the SCN1B variant along with variants in POLR1C were flagged as the best candidates that potentially contribute to phenotypes manifested by both the proband and his elder sister. She had a fetal diagnosis of bradycardia and subsequent postnatal finding of complete AV block, leading to early neonatal death from multi-organ failure (Eldomery et al., 2017).
The proband's father is heterozygous for the c.308A>T variant in SCN1B and for the c.88C>T variant in POLR1C, and his mother is heterozygous for the c.614delG variant in POLR1C. Both mother and father are asymptomatic. Incomplete penetrance is a common feature of channelopathies, and the idea of monogenetic disease has changed in recent years (Symonds and Zuberi, 2018). However, neither the SCN1B variant nor POLR1C variants found in the child could be the sole cause of his complex clinical picture.
POLR1C has been associated with autosomal recessive hypomyelinating leukodystrophy (Thiffault et al., 2015). This gene also has been described to cause recessive Treacher Collins syndrome 3 (OMIM #248390). Our proband did not show clinical manifestations of either of these conditions (Eldomery et al., 2017). However, the patient exhibited several neurological phenotypes consistent with POLR1C mutations, including hypomyelination, ataxia, and nystagmus, but the patient's other clinical features are not described in POLR1Crelated cases. Thus, it is possible that electrical disturbances caused by the SCN1B variant become more severe in the context of POLR1C mutations. Also, three-dimensional voltage mapping at baseline rhythm revealed extensive scarring of the right atrium and coronary sinus. Thus, electrical dysfunction potentially caused by the mutant β1 subunit could be aggravated by a damaged tissue substrate.
In summary, although the overall pathophysiology of the patient is complex, the Na V 1.5 loss of function caused by the mutant β1 subunit D103V largely explains the clinical manifestations related to the patient's heart dysfunction. In addition, loss of function of Na V 1.1 caused by the β1 and β1b mutant subunits might aggravate a brain condition caused by the combination of the two POLR1C mutations.

DATA AVAILABILITY STATEMENT
The datasets generated for this study are available on request to the corresponding author.

ETHICS STATEMENT
Written informed consent was obtained from the individual(s), and minor(s)' legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.