Epilepsy-Related Voltage-Gated Sodium Channelopathies: A Review

Epilepsy is a disease characterized by abnormal brain activity and a predisposition to generate epileptic seizures, leading to neurobiological, cognitive, psychological, social, and economic impacts for the patient. There are several known causes for epilepsy; one of them is the malfunction of ion channels, resulting from mutations. Voltage-gated sodium channels (NaV) play an essential role in the generation and propagation of action potential, and malfunction caused by mutations can induce irregular neuronal activity. That said, several genetic variations in NaV channels have been described and associated with epilepsy. These mutations can affect channel kinetics, modifying channel activation, inactivation, recovery from inactivation, and/or the current window. Among the NaV subtypes related to epilepsy, NaV1.1 is doubtless the most relevant, with more than 1500 mutations described. Truncation and missense mutations are the most observed alterations. In addition, several studies have already related mutated NaV channels with the electrophysiological functioning of the channel, aiming to correlate with the epilepsy phenotype. The present review provides an overview of studies on epilepsy-associated mutated human NaV1.1, NaV1.2, NaV1.3, NaV1.6, and NaV1.7.


INTRODUCTION
Epilepsy is a disease known worldwide, affecting around 70 million people in the world (Thijs et al., 2019). It has been considered a disease and no longer a disorder or a family of disorders since 2014 by International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE) (Falco-Walter et al., 2018). Epilepsy is conceptually defined as a disease in which an individual has at least two unprovoked or reflex seizures in a period greater than 24 h apart, one unprovoked or reflex seizure and a probability of having another seizure similar to the general recurrence risk after two unprovoked seizures (greater than or equal to 60%) over the next ten years or an epilepsy syndrome (Fisher et al., 2014).
When abnormal brain activity begins in one or more identified regions, epilepsy is called focal, whereas, when it occurs in both hemispheres with a wide distribution, it is called generalized. Finally, when it cannot be classified as either focal or generalized, it is called unknown (Devinsky et al., 2018).
Epilepsy can affect anyone, regardless of gender, age, and income levels (Saxena and Li, 2017). Understanding the etiology of epilepsy is crucial for clinical management of patients and for conducting neurobiological research that will direct future therapies (Thomas and Berkovic, 2014). The ILAE Task Force has defined six etiologic categories; they are not hierarchical and more than one might often apply (structural, genetic, infectious, metabolic, immune, and unknown) (Falco-Walter et al., 2018).
Among those genetically caused, it is possible to identify several epilepsy-related genes (Lindy et al., 2018). For example, voltage-gated potassium channel, voltage-gated calcium channel and voltage-gated chloride channel genes, GABA receptors, nicotinic acetylcholine receptors, polymerase (DNA) Gamma genes and voltage-gated sodium channel genes (Deng et al., 2014).
Voltage-gated sodium channels (NaV) can be found mainly in the central nervous system (CNS), peripheral nervous systems (PNS), skeletal, and cardiac muscles (Huang et al., 2017). NaVs are distributed throughout the body and play an important role in the generation and propagation of action potential (Wang et al., 2017b). Structurally, NaVs are composed by an a subunit organized in four homologous ligated domains (DI-DIV), each domain composed by six transmembrane segments (S1-S6), and one or more b subunits associated by non-covalent interactions or disulfide bond (Abdelsayed and Sokolov, 2013;Gilchrist et al., 2013;Catterall, 2017;Bouza and Isom, 2018;Jiang et al., 2020). The domains of an a subunit present a high degree of conservation with each other, presenting the region known as the voltage sensor domains (VSD) located in transmembranes S1-S4, especially S4 helix, which contains positively charged residues, and the pore-forming (PM) domain located in S5-S6 segments, structuring a four VSD around a central pore (Ahern et al., 2016).
The S4 helix of DI, DII, and DIII domains moves faster than the S4 helix of DIV during membrane depolarization, and this asynchronous movement is an essential feature in the steady activation voltage-dependent process, which provokes movement of S4-S5 intracellular links followed by the displacement of the S6 segments to initiate Na + influx Oelstrom et al., 2014). The movement of the S4 helix of DIV initiates the process of fast inactivation, since the movement of the voltage sensor in domain DIV is associated with the displacement of an intracellular loop between DIII and DIV within an IFM (isoleucine, phenylalanine, and methionine) motif that binds intracellular to PM and terminate Na + influx Clairfeuille et al., 2019). A second type of reversible inactivation occurs after repetitive or prolonged stimulation and results in steady-state inactivation whose asymmetric movement of S6 segments collapses the pore (Payandeh et al., 2012;Zhang et al., 2012;Gamal El-Din et al., 2013;Silva and Goldstein, 2013;Ghovanloo et al., 2016). Consequently, electrophysiological changes such as increased current density, shifting steady-state activation, and inactivation to negative and positive values, respectively, enhanced persistent current, accelerated recovery from inactivation, and delayed fast inactivation can cause gain-of-function (GoF) in the channel. Also, decreased current density, positive shift in steady-state activation, negative shift in steadystate inactivation, and slower recovery from inactivation can cause loss-of-function (LoF) (Mantegazza et al., 2005;Liao et al., 2010;Lossin et al., 2012;Catterall, 2014b;Vanoye et al., 2014;Wagnon et al., 2017;Yang et al., 2018;Zaman et al., 2018;Wengert et al., 2019;Zhang S. et al., 2020).
NaV channels rank amongst the 2% most conserved proteins in the human genome, with an extremely low rate of coding variation, accounting for nearly 5% of known epileptic encephalopathies (Petrovski et al., 2013;Mercimek-Mahmutoglu et al., 2015;Lek et al., 2016;Heyne et al., 2019). Pathogenic mutated residues are situated in the highly evolutionarily conserved portions of the channel: transmembrane segments, intracellular inactivation gate loop, and the proximal 2/3 of the C-terminal domain (Blanchard et al., 2015;Wagnon and Meisler, 2015). The final 1/3 portion of the C-terminal and cytoplasmic interdomain loops 1 and 2 are less conserved (Denis et al., 2019). The proximal 2/3 of the C-terminal are involved in the interaction of several binding sites for proteins and accessory molecules, like beta subunits b1 and b3, fibroblast growth factors (molecules implicated in neural development), calmodulin (regulatory protein in neuronal function and hyperexcitability) and G protein (Bähler and Rhoads, 2002;Spampanato, 2004;Wittmack et al., 2004;Laezza et al., 2009;Yang et al., 2010). Moreover, the C-terminal has been shown to interact with the inactivated channel via ionic interaction between its positively charged residues and negatively charged residues at the inactivation gate. A shift in any of the charges can brake electrostatic interaction and affect normal channel inactivation (Nguyen and Goldin, 2010;Shen et al., 2017;Johnson et al., 2018).
The N-terminal region seems to play a more important role on protein trafficking than on channel activity. This domain interacts with the light chain of microtubule-associated protein MAP1B, facilitating the traffic of the NaV channel to the neuronal cell surface Blanchard et al., 2015). In addition, mutation in the N-terminal leads to protein retention in the endoplasmic reticulum (Sharkey et al., 2009).
Newer genomic approaches, especially next generation sequencing (NGS), improve the rate and reduce the costs associated with genetic epilepsy diagnosis, since traditional cytogenetic and microarray-based tests are lengthy, expensive, and diagnostic yield is incredibly low (Veeramah et al., 2013;Allen et al., 2016;Sands and Choi, 2017;Orsini et al., 2018). The use of gene panels and whole-exome sequencing (WES) provides a powerful tool to change the paradigm of genetic epilepsy diagnosis (Ng et al., 2010;Clark et al., 2018). These techniques have been widely used to elucidate suspected inherited neurological diseases in the last years, contributing to dramatically increase the number of patients diagnosed with genetic epilepsy. Both mendelian and de novo genetic epilepsy can be detected with these methods, but doubtless, de novo mutations are the most prevalent mutations related to epilepsyrelated voltage-gated sodium channel mutations.
Gene therapy is promising as an effective approach to treat genetic diseases. Personalized epilepsy therapies are in development and have shown promising results, ranging from antisense oligonucleotides and small peptides to modulation of gene expression through epigenetics (Riban et al., 2009;Tan et al., 2017;Stoke Therapeutics, 2018;Perucca and Perucca, 2019). Even eating habits may be related to an improvement in the patient's clinical condition. Ketogenic diet has been described as an effective treatment in epilepsy . Moreover, the combination of traditional antiepileptic drugs with new compounds displayed a synergic and improved efficacy, since these molecules do not compete for the same interaction site (Bialer et al., 2018). Each specific epilepsy-related NaV isoform will be presented and discussed in detail in the following sections.

NaV1.1
The SCN1A gene encodes for the a subunit NaV1.1, and is allocated at the 2q24.3 chromosome between 165,984,641 and 166,149,161 base pairs, same gene cluster of SCN2A-SCN3A genes, being the most frequent target of mutation in genetic epilepsy syndromes (OMIM#182389) (Malo et al., 1991;Malo et al., 1994;Catterall et al., 2010). NaV1.1 is widely expressed in the CNS, predominant in inhibitory GABAergic interneurons, regulating neuronal excitability, and the reduction of its activity is one of the factors that cause epileptic diseases due to imbalance between inhibition and excitation (Yu et al., 2006;Verret et al., 2012;Tai et al., 2014;Rubinstein et al., 2015).
In the SCN1A mutation database (http://www.caae.org.cn/ gzneurosci/scn1adatabase/data), among 1727 mutations described for the SCN1A gene, 1528 are related to epileptic diseases (Table 1 and for the full description of mutations in the SCN1A gene, see Supplementary Table S1). Among the epilepsy-related mutations, 945 are related to severe myoclonic epilepsy of infancy (SMEI), 263 are related to severe myoclonic epilepsy (SME), 151 are related to severe myoclonic epilepsy borderline (SMEB), 18 are related to partial epilepsy (PE), 31 are related to partial epilepsy and febrile seizures plus (PEFS +), 8 are related to generalized epilepsy (GE), and 55 are related to generalized epilepsy with febrile seizures plus (GEFS +).
Mutations in the NaV1.1 channel are described in almost all regions of the protein and may cause GoF or LoF Meng et al., 2015). Among the 52 mutations in SCN1A related to epilepsy with functional studies, 35 mutations (67.30%) exclusively display characteristics of LoF, 6 mutations (11.53%) display characteristics unique to GoF, and 11 mutations (21,15%) display characteristics of GoF+LoF, whereas, in GoF+LoF mutations, the main characteristic that gives GoF features is enhanced persistent current, present in 10 out of the 11 GoF+LoF mutations listed (Tables 1 and S1).
Due to the role of the NaV1.1 channels in the regulation of electrical excitability by the inhibitory interneurons, prescription of AEDs non-selective sodium channel blockers (SCB) for SMEI or GEFS + syndromes is contraindicated, for it may aggravate crises due to the enhanced suppress status of the NaV1.1 channels (Catterall, 2014a;Shi et al., 2016;Knupp and Wirrell, 2018;Ziobro et al., 2018). The first-line drug-based therapy for SCN1A epilepsy diseases is the enhancement of postsynaptic GABAergic transmission with allosteric activation of GABA A receptors as target by Clobazam and/or an increase in GABA concentration in synaptic cleft resulting from increased GABA production and decreased GABA degradation as target by Valproic acid (Catterall, 2014a;Hammer et al., 2016;Knupp and Wirrell, 2018;Musto et al., 2020). Antisense nucleotides (ASO) therapy to increase mRNA of SCN1A for NaV1.1 channel expression in normal levels is a promising strategy for genetic disorders involving haploinsufficiency (Hsiao et al., 2016;Stoke Therapeutics, 2018). Drug-resistant Dravet syndrome cases may thrive on alternative therapeutic strategies based on ketogenic diets (Nabbout et al., 2011;Wu et al., 2018). A recent study with 20 patients with medically intractable Dravet syndrome caused by missense, non-sense, insertion, deletions and splicing mutations presents efficacy during three months of treatment in 17 patients, decreasing seizure frequency in more than 50% . Besisdes that, Epidiolex is an FDA approved CBD-based drug approved in June 2018 for the treatment of severe forms of epilepsy, as Dravet and Lennox-Gastaut syndromes (U.S. Food and Drug Administration [website]., 2018). Clinical trials using CBD in DS and LGS shown reduced frequency of seizures in monthly average (Lattanzi et al., 2020;Morano et al., 2020). Voltage-gated sodium channel are inhibit by CBD in low micromolar concentrations, IC 50 between 1.9 and 3.8 mM, NaV1.4 and NaV1.1 being the most sensitive channels to CBD, 1.9 and 2.0 mM respectively, probably the mechanism of action is reducing channel availability due shift to more hyperpolarized potential in steady-state inactivation (Ghovanloo et al., 2019).
NaV1.2 channels are expressed in the excitatory neurons; therefore, GoF mutations are related to epilepsy because it causes neuronal hyperexcitability. On the other hand, LoF mutations are related to autism and intellectual disability phenotype (Ben-Shalom et al., 2017). Nevertheless, some studies have already related loss of function to epilepsy, as described by Lossin and coworkers (2012) with R1312T mutation (Lossin et al., 2012).     Severe epilepsy Epileptic seizure with pale, atonic periods followed by a spasm-like out-throwing of both arms Predominantly right-sided epileptiform activity (EEG) (Davidsson et al., 2008) *Non genetic origin mutations reported: Mutations described through clinical diagnosis, but the mutation type (Mendelian or de novo) were not reported, mainly due to the lack of parents to perform genotyping and difficulty in contacting the family. Normally, LoF SCN2A gene mutations for epilepsy are related to late-onset epilepsy; however, the mechanism of action is unclear (Mason et al., 2019). In some cases, NaV1.2 seizures are not controlled not even by various antiepileptic drugs, as with the patient described by Syrbe and colleagues (2016). The proband, even after being treated with oxcarbazepine (OXC), valproic acid, topiramate, sulthiame, phenytoin, among other drugs, kept on having seizures (Syrbe et al., 2016). Furthermore, the SCB drugs can assist the patient during the treatment as described by Gorman and King (2017). The patient had seizures controlled after administration of phenytoin (Gorman and King, 2017). In addition, Musto et al. (2020) cite benefits treatments using SCB such as carbamazepine, mexiletine, oxcarbazepine, phenytoin, lidocaine, and lamotrigine for patients with early onset epilepsies (Musto et al., 2020). Besides, Peters and colleagues studied a substance commercially used as an antianginal drug (human heart) called ranolazine that has been shown to affect NaV1.2 channels, reducing macroscopic currents and delaying the recovery of fast and slow inactivation of the NaV1.2 channel, consequently with more future studies ranolazine could be a efficacious therapy for epilepsy (Peters et al., 2013).
Drugs can be important to modulate channel kinetics for both GoF and LoF, but some precautions must be observed. For example, the degree of conservation between subtypes, such as NaV1.2 and other sodium channels as NaV1.5 and the excessive decrease in channel function or the excessive increase in function obtained by the drug (Sanders et al., 2018).
Organizations like the FamilieSCN2A Foundation (www. scn2a.org) might be essential in the search for new treatments. Understanding the genotype-phenotype of gain and loss of function is essential because science-patient relationship may be helpful in the search for new therapies (Sanders et al., 2018).

NaV1.3
SCN3A is a gene that encodes for type 3 voltage-gated Na + channel a subunit, the NaV1.3, located on human chromosome 2q24, in a cluster with SCN1A and SCN2A (Holland et al., 2008). NaV1.3 is expressed predominantly in the CNS during embryonic and neonatal development, being extremely low or sometimes undetectable in postnatal individuals. Subsequently, during infancy, it is gradually replaced by increased expression of the NaV1.1 isoform (Felts et al., 1997;Whitaker et al., 2000;Cheah et al., 2013;Zaman et al., 2018). On the other hand, studies regarding nervous system injury and neuropathic pain showed an increasing presence of NaV1.3 channels in affected tissues, suggesting a pivotal hole of these transmembrane proteins in these processes and diseases (Hains et al., 2003;Waxman and Hains, 2006;Black et al., 2008). For the reasons mentioned above, in the last decades, NaV1.3-associated pathogenesis has been restricted to pain. Recently, a genetic linkage between NaV1.3 mutated variants and epilepsy has been suggested, especially in cryptogenic epilepsy cases (OMIM#182391). K354Q was the first described NaV1.3 epilepsy-related mutation that revealed harmful electrophysiological alterations (Holland et al., 2008;Estacion et al., 2010). In fact, mutations can change many functional characteristics of NaV1.3 affecting biophysical properties differently; however, these changes result predominantly in neuronal hyper-responsiveness (Table 3) (Cummins and Waxman, 1997;Chen et al., 2000;Cummins et al., 2001;Sun et al., 2007). Previous reports correlate heterozygous variants in SCN3A in association with moderate forms of epilepsy, while homozygosis is related with severe cognitive damage and premature mortality, resulting in a broad range of epileptic phenotypes (Estacion and Waxman, 2013;Vanoye et al., 2014;Lamar et al., 2017).
Different hereditary mutations on NaV1.3 have been reported to date in patients with epilepsy. In general, the biophysical characterization of these mutations reveals GoF, only one mutation (N302S) is related with LoF (Chen et al., 2015), but both GoF and LoF may lead to an increased seizure susceptibility (Lamar et al., 2017).
Moreover, several de novo mutations in SCN3A have been described in the last three years, related with severe infantile neurological dysfunctions and cognitive impairments. These mutations may alter the functionality of NaV1.3 channels, neurons organization, migration, and proliferation during the embryonic development . Epileptic encephalopathy and polymicrogyria are the main features related with these pathogenic variants, and, so far, polymicrogyria was not reported in other channelopathies, being an exclusive characteristic of SCN3A mutants (Inuzuka et al., 2019).
There is a lack of clinical data on SCN3A-related epilepsies, especially regarding treatment and the use of specific medication. However, in vitro studies reported that mutations related with GoF effect respond favorably to treatment using SCB, like phenytoin, carbamazepine, lacosamide, and topiramate (Sun et al., 2007;Sheets et al., 2008;Colombo et al., 2013;Zaman et al., 2018). The anticonvulsant valproic acid represents a novel and promising epigenetic therapeutic approach (Tan et al., 2017). The compound modulates the SCN3A gene through methylation, downregulating the expression of NaV1.3 and, consequently, decreasing biophysical alterations in the channel.

NaV1.6
The SCN8A gene encodes for type 8 voltage-gated Na + channel a subunit, the NaV1.6, located in chromosome 12q13.13. The first case of SCN8A pathogenic variant associated with epilepsy was reported eight years ago (Veeramah et al., 2012). Thereafter, due to advances in genome sequencing technology, especially the WES, the number of epilepsy diagnosis associated with NaV1.6 mutations has increased significantly (OMIM #600702), with more than 300 patients diagnosed with SCN8A epilepsy mutations and nearly 200 different putative spots of mutations described, totaling over 100 published reports ( Table 4). A website developed especially to present SCN8A epilepsy and related diseases (www.scn8a.net) was created to provide information to families, clinicians, and researchers, gathering news and recent publications on the subject in a private forum for family interaction, to answer questions, strengthening the ties between the community and the researchers. NaV1.6 is expressed since prenatal, during fetal development (Plummer et al., 1997). Shortly after birth, expression begins to increase, reaching maximum levels during the first years of life. This           Sprissler et al., 2017;Lindy et al., 2018;Zaman et al., 2019;Schreiber et al., 2020) R1872Q C-Terminus Missense Enhanced persistent current Increase peak current density Shift steady-state activation to more negative values Shift steady-state inactivation to more positive values (Larsen et al., 2015;Horvath et al., 2016;Hussain et al., 2016;Arafat et al., 2017;Atanasoska et al., 2018;Lindy et al., 2018) R1872W C-Terminus Missense Enhanced persistent current Increased peak current density Shift steady-state activation to more negative values Shi ft steady-state i nacti vati on to more posi ti ve val ues (Ohba et al., 2014;Larsen et al., 2015;Takahashi et al., 2015;Gardella et al., 2018;Denis et al., 2019;Kim et al., 2019;Zaman et al., 2019) N1877S C-Terminus Missense NR (Anand et al., 2016;Parrini et al., 2017;Wang et al., 2017a;Lindy et al., 2018;Costain et al., 2019;Epifanio et al., 2019;Jain et al., 2019;Ranza et al., 2020   channel is widely expressed in the nodes of Ranvier of myelinated axons and in the distal part of the axon initial segments (AIS), although they are also ubiquitously present throughout the central and peripheral nervous systems, in both excitatory and inhibitory neurons (Caldwell et al., 2000;Oliva et al., 2012). For these reasons, NaV1.6 is one of the most common subtype of voltage-gated sodium channels found in the central nervous system (Caldwell et al., 2000). In humans, the distal AIS is the specialized membrane region in neurons where action potentials are triggered. Overexpression of Nav1.6 in the AIS has been shown to cause an increase in spontaneous and repetitive firing (Hu et al., 2009;Sun et al., 2013), a possible explanation for why SCN8A mutations in epilepsy patients are predominantly GoF and affect the action potential threshold. On the other hand, the functional importance of Nav1.6 in inhibitory interneurons is not clear yet, but evidence indicates a role for Nav1.6 in establishing synaptic inhibition in the thalamic network (Makinson et al., 2017), supporting the LoF results caused by missense mutations in the mature protein.
Sudden unexpected death in epilepsy (SUDEP) has also been linked to SCN8A mutations, described as the most common cause of death in epilepsy patients. Reports have suggested that patients with SCN8A-related epilepsy have increased risk of SUDEP, ranging from 1% to 10% (Hammer et al., 2016;Wang et al., 2017a;Gardella et al., 2018;Johannesen et al., 2018). One possible correlation of SUDEP with SCN8A-related epilepsy is the presence of NaV1.6 in heart muscles and tissues, being broadly expressed within ventricular myocytes (Maier et al., 2002). Single mutations may affect heart function, causing failure of the cardiorespiratory system and, consequently, death (Haufe et al., 2005;Noujaim et al., 2012).
Most recently, few cases of SCN8A-related epilepsies with "milder" phenotype were associated with benign familial infantile seizures-5 (BFIS5; OMIM #617080) (Anand et al., 2016;Gardella et al., 2016;Han et al., 2017). An increase in new described variants made some mutation patterns visible. Wagnon and co-workers observed numerous cases of the same epiletogenic mutation, and suggested that CpG dinucleotides are mutation hotspots that, through enzymatic processing and epigenetic methylation, can convert cytosine to thymine, such as arginine residues 1617 and 1872 (Wagnon and Meisler, 2015). The prominent number of new variant cases in Arg850 indicates this residue as a new hotspot, since the arginine codon holds a CpG dinucleotide. In addition to these mutation hotspots, residues I763, I1327, G1475, A1650, and N1877 do not present CpG dinucleotides in their codon; however, they can be considered recurrent mutations in view of its high repetition cases in literature ( Table 4).
The mutation at position c.-8A>G produces a pathogenic variant, despite not being inside the gene, or promoter regions, *Non genetic origin mutations reported: Mutations described through clinical diagnosis, but the mutation type (Mendelian or de novo) were not reported, mainly due to the lack of parents to perform genotyping and difficulty in contacting the family. Not Reported (NR); Domain (D); Segment (S).

Menezes et al.
Epilepsy-Related Nav Channelopathies: A Review transcriptional and translational sites. This mutation was detected in an untranslated region outside of the Kozak consensus sequence . Its role in SCN8A-related epilepsy is still unclear; however, it may change RNA stability, modulate transcriptional factors and promoters, modify the initiation of translation, or work as an enhancer or silencer in the splicing pattern. For all the reasons mentioned above, Nav1.6 variants are predominantly harmful, and the same mutation can lead to different phenotypes, hampering the correlation of genotypes with phenotypes (Blanchard et al., 2015). SCN8A mutations can be both GoF and LoF, which will likely require different approaches and targets. Even in patients with the same SCN8A mutation, the response to the same drug treatment can differ. Surprisingly, most SCN8A-related epilepsies respond favorably to channel blockers. Phenytoin and lacosamide are SBCs widely used in SCN8A mutations with GoF effect, while carbamazepine exhibited positive seizure control in a patient with NaV 1.6 mutation and LoF effect. (Blanchard et al., 2015;Wagnon and Meisler, 2015;Hammer et al., 2016;Perucca and Perucca, 2019). Phenytoin demonstrated effectiveness in decreasing seizure episodes in several patients with SCN8A-related epilepsies, however, side effects during prolonged use are very common (Boerma et al., 2016;Braakman et al., 2017). A recent study of a DS model using zebrafish demonstrated the use of the channel blocking compound MV1312, which is 5-6 fold selectivity of NaV1.6 over NaV1.1-1.7, reduced burst movement phenotype and the number of epileptiform events, activity similar to that described with the use of a selective NaV1.1 activator AA43279 (Weuring et al., 2020). Selective Nav1.6 blockers may represent a new therapeutic strategy for DS patients. In addition, two precise and promising drugs have been described recently: XEN901 and GS967. XEN901 is an arylsulfonamide highly selective and potent NaV1.6 inhibitor that binds specifically in voltage sensor domain IV, avoiding recovery from inactivation. GS967 is a NaV1.6 modulator that inhibits the persistent sodium current and exhibits a protective effect (Baker et al., 2018;Bialer et al., 2018).

NaV1.7
The SCN9A gene encodes for the NaV1.7 channel, located in chromosome 2q24 (Yang et al., 2018). NaV1.7 is expressed preferably in the PNS, but it is also expressed in the CNS (Cen et al., 2017). Consequently, mutations in this channel are generally related to pain disorders (Young, 2007;Han et al., 2009;Doty, 2010;Rush et al., 2018); however, current studies have described a correlation between epilepsy and this channel (OMIM #603415).
Pain disorder mutations with GoF are related with diseases such as erythromelalgia (EMI), small-fiber neuropathy (SFN) and paroxysmal extreme pain disorder (PEPD), and mutations with LoF are related with congenital insensitivity to pain (CIP) (Cen et al., 2017). Epilepsy studies such as Zhang S. et al. (2020) showed mutations with GoF phenotype: W1150R, N641Y, and K655R mutations ( Table 5). Being that, after treatment with OXC (120 µmol/L), N641Y and K655R reduced sodium current and decreased the opening time of the channel, while W1150R did not alter that (Zhang S. et al., 2020). However, in a study conducted by Yang et al. (2018), one of the patients presented generalized tonic-clonic Faster recovery from inactivation More susceptible to clonic and tonic seizures induced by electrical stimulation (mice) Enhanced persistent current (Singh et al., 2009;Zhang S. et al., 2020) I1901fs

C-terminal
Frameshift Epilepsy Generalized tonic-clonic seizure (Yang et al., 2018) Non genetic origin mutations reported* K655R seizure with fever, treated with sodium valproic acid, and a LoF mutation I1901fs was observed (Yang et al., 2018) (Table 5). Variants of NaV1.7 have been related with febrile seizure or GEFS+ (Cen et al., 2017;Zhang S. et al., 2020) and even as asymptomatic (Singh et al., 2009). However, SCN9A can act as a putative modifier of NaV1.1 gene; consequently, it can elevate the severity of patients' phenotype (Guerrini et al., 2010;Parihar and Ganesh, 2013). Some NaV1.7 mutations could probably contribute to generate a genetic susceptibility to a known epilepsy disease called Dravet syndrome, in a multifactorial way, as a modifier gene (Singh et al., 2009;Doty, 2010;Mulley et al., 2013;Cen et al., 2017;Zhang T. et al., 2020). That said, some rare cases of DS found in patients can be understood (Mulley et al., 2013). For example, even parents with mild phenotype had children with severe cases (Guerrini et al., 2010).

CONCLUSION AND FUTURE PERSPECTIVES
The past two decades have enabled remarkable progress in understanding monogenic epilepsies. NaV-related epilepsies are diseases of phenotypic heterogeneity, since sodium channels are found in both the CNS and the PNS, but with different expression ranges. The lack of a clear genotype-phenotype correlation to help guide patient counseling and management by healthcare professionals makes it very complex, and often expensive, to determine a correct diagnosis. Consequently, identify the monogenic mutation in individual patients with epilepsy is important not only for diagnosis and prognosis, but also for a correct treatment approach Reif et al., 2017).
Susceptibility to specific treatments may be different depending on the disease's features, diverging even in patients who share the same phenotype and/or mutation (Weber et al., 2014). The use of innovative tools that facilitate and prevent diagnostic delay in patients with epilepsy of unknown etiology onset is crucial. WES has proved to be a valuable tool to circumvent the lack of an accurate and fast diagnosis to epilepsies caused by monogenic mutation, and also cheapen and drastically anticipate diagnosis. This genetic diagnostic tool may reduce traditional investigation costs by 55 to 70%, besides avoiding further pre-surgical evaluation and epilepsy surgery (Kothur et al., 2018;Oates et al., 2018). In addition to the financial impact, it can anticipate diagnosis from nearly 3.5 years to 21 days, optimizing management and health care support (Oates et al., 2018).
Effective and safe drugs for the treatment of monogenic epilepsy are still an unmet clinical need. The drugs currently available in the pharmaceutical market are only palliative methods for a temporary control of the disease symptoms, and few patients will benefit from the existing pharmacotherapy, since a great number of patients treated with antiepileptic channel blockers showed no improvement in clinical conditions. Also, most treated patients exhibited manifold side effects, and the prolonged use of these medications proved to be harmful (Boerma et al., 2016;Braakman et al., 2017). Several examples of novel and promising candidate compounds to be used in personalized medicine, such as precision therapies, have been suggested. A previously study demonstrated that CBD at 1mM inhibit preferably resurgent currents than transient current in Nav1.6 WT and also inhibit peak resurgent current in Nav1.6 mutant N1768D, with less effect in current density and without alters voltage dependence of activation (Patel et al., 2016) Possibly the modulation of CBD over mutations in SCN8A that promotes a phenotype with increased resurgent currents would cause a reduction in the causative excitability of epileptic seizures. CBD also showed its ability to preferential inhibit resurgent currents in the NaV1.2 channel (Mason and Cummins, 2020). Due the role of Nav1.2 and Nav1.6 in excitatory neurons, preferentially inhibition in resurgent currents by CBD could possibly reduce the excitability in that subset of neurons and decrease the frequency of seizures by a change in threshold of activation and repetitive fire (Lewis and Raman, 2014). Peptides derived from scorpion and spider venom are well known modulator tools in neuroscience and showed specific capacity to regulate most NaV subtypes related with monogenic epilepsy, unlike the available promiscuous drugs that generally interact with any NaV channel isoform (Schiavon et al., 2006;Israel et al., 2018;Richards et al., 2018;Tibery et al., 2019;Zhang et al., 2019). Bioengineering tools, like antisense oligonucleotides capable to regulate NaV1.1 channels expression, and the peptide Hm1, that modulates the function of this subtype of sodium channel, are some innovative treatment examples (Richards et al., 2018;Stoke Therapeutics, 2018).
However, there is still a long path toward the development of efficacious treatments for NaV-related epilepsies. Recent studies offered a better understanding of the complexity of the phenotypic and genetic spectrum, which has only just begun to be elucidated. Biomolecular diagnostic tools will drastically reduce the developmental and cognitive effects caused by misdiagnosis and late diagnosis, and maybe, in the upcoming years, the treatment for inherited NaV-related epilepsies will be conducted ideally in utero, during the prenatal stage. Moreover, further functional studies, with greater cohorts of patients, represent an urgent medical need for a better understanding of the correlations between genotype and clinical symptoms, as well as the different NaV-related epilepsies mechanisms. These studies will improve clinical efficacy and promote safety diagnostic strategies, as well as develop prognosis prediction in the near future.

AUTHOR CONTRIBUTIONS
All authors made an intellectual and direct contribution for this article and approved it for publication.

ACKNOWLEDGMENTS
CNPq, CAPES, and the Molecular Biology postgraduate program of the University of Brasilia. LM received scholarships from CNPq and DT from CAPES. EFS was supported by CNPq.