INTRODUCTION

Voltage-gated sodium (Nav) channels are essential for the rapid depolarization phase of action potential and play a key role in the electrical signaling in most excitable cells. Structurally, Nav channels are composed of one α subunit and one or more β subunits. The α subunit contains two functionally distinct structural entities, namely, the voltage-sensing domains (VSDs) and the ion-conducting pore domain (Catterall, 2012b, 2014). The β subunits, which bind to α subunit covalently or non-covalently, modulate membrane trafficking, voltage dependence, and channel gating kinetics (Catterall, 2012b, 2014). In mammals, Nav channels have nine known α members distributed in different excitable tissues. Specifically, Nav1.1, Nav1.2, Nav1.3, and Nav1.6 are the primary sodium channels in central nervous system (CNS), Nav1.4 is primarily expressed in skeletal muscle, Nav1.5 is mainly expressed in heart, and Nav1.7, Nav1.8, and Nav1.9 are mainly distributed in peripheral nervous system (Plummer and Meisler, 1999; Goldin, 2001; Catterall et al., 2005).

All α subunits share nearly identical structure topology—a canonical voltage-gated ion channel fold with four homologous repeats, each containing six transmembrane segments S1–S6. Specifically, S5–S6 segments form the pore domain that conducts selective sodium filtering, while S1–S4 segments constitute the voltage-sensing domain that controls voltage-dependent gating (Catterall, 2000). The voltage sensors in the VSDs are featured by a number of positively charged amino acids (arginine or lysine) located at every third position in the S4 segment. Upon membrane depolarization, movements of these charged residues in the S4 segment are coupled to the opening of the pore domain and the subsequent influx of sodium ions across cell membrane. The pore domain is structurally organized with a four-fold pseudo-symmetry. The pore (P) loops, which are supported by the P1 helix (corresponding to the P helix in potassium channel) and P2 helix between S5 and S6 segments in each repeat, constitute the selectivity filter (SF) (Corry and Thomas, 2012). Four amino acid residues (aspartate, glutamate, lysine, and alanine, DEKA, in repeats I, II, III, and IV, respectively) in the P loops are crucial for sodium selectivity. Mutating these residues to glutamates confers calcium selectivity, suggesting that the side chains of these amino acids are likely to interact directly with the sodium ions to determine ion selectivity (Heinemann et al., 1992; Sun et al., 1997).

Nav channels inactivate rapidly. A cluster of hydrophobic amino acids (isoleucine, phenylalanine, methionine, and threonine), namely the IFMT motif, located in the cytosolic regions of domain III and domain IV, are required for rapid inactivation. This is demonstrated by the fact that rapid inactivation could be achieved by titrating small peptides containing the IFMT motif (Vassilev et al., 1988; West et al., 1992).

Sodium channelopathies are a group of diseases caused by defective Nav channels, either, in most cases, of congenital nature or acquired nature (Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10) (George, 2005; Catterall, 2012a; Kim, 2014). For example, Nav1.1 is primarily expressed in the soma of neuronal cells in the CNS, and mutations of Nav1.1 cause GEFS+2 (generalized epilepsy with febrile seizures plus 2) (Catterall et al., 2010). Moreover, mutations of Nav1.1 are also the main causes of EIEE6 (epileptic encephalopathy, early infantile, 6) and ICEGTC (intractable childhood epilepsy with generalized tonic-clonic seizures) (Escayg and Goldin, 2010). Nav1.5 is the major sodium channel expressed in heart. Nav1.5 mutations may lead to various cardiac diseases such as LQT3 (long QT syndrome 3), BRGDA1 (Brugada syndrome 1), and SSS1 (sick sinus syndrome 1) (Olson et al., 2005; Song and Shou, 2012; Veerman et al., 2015). Nav1.7 is preferentially expressed in the sympathetic neurons, olfactory epithelium, and dorsal root ganglion sensory neurons, and plays a cardinal role in pain transmission (Djouhri et al., 2003; Dib-Hajj et al., 2013). Gain-of-function mutations of Nav1.7 are implicated in two distinct paroxysmal pain syndromes—IEM (primary erythermalgia) and PEPD (paroxysmal extreme pain disorder), while loss-of-function mutations of Nav1.7 inflict people with CIP (indifference to pain, congenital, autosomal recessive) (Lampert et al., 2010; Dib-Hajj et al., 2013). In all, Nav channel mutations play a central role in the pathophysiology of sodium channelopathies. Pharmacologic modulation of Nav channels may thereby represent a viable therapeutic approach for the treatment of many neurological disorders such as epilepsy, arrhythmia, and pain.

Table 1 Structural mapping of disease-related mutations identified in human Nav1.7
Full size table
Table 2 Structural mapping of disease-related mutations identified in human Nav1.1
Full size table
Table 3 Structural mapping of disease-related mutations identified in human Nav1.2
Full size table
Table 4 Structural mapping of disease-related mutations identified in human Nav1.3
Full size table
Table 5 Structural mapping of disease-related mutations identified in human Nav1.4
Full size table
Table 6 Structural mapping of disease-related mutations identified in human Nav1.5
Full size table
Table 7 Structural mapping of disease-related mutations identified in human Nav1.6
Full size table
Table 8 Structural mapping of disease-related mutations identified in human Nav1.8
Full size table
Table 9 Structural mapping of disease-related mutations identified in human Nav1.9
Full size table
Table 10 Summary of sodium channelopathies
Full size table

Despite significant advancement in the understanding of Nav channel functions and their relevance to diseases, structural characterization of mammalian Nav channels at atomic level has been challenging, partly due to the substantial technical hurdles in producing mammalian Nav channel proteins in sufficient amount with acceptable purity. The two published bacterial Nav channel crystal structures, NavAb (Payandeh et al., 2011) and NavRh (Zhang et al., 2012), in their full-length have greatly improved our understanding of how those channels conduct and select sodium ions on a structural basis. This is further enhanced by the recently published cryo-electron microscopy (cryo-EM) structure of the rabbit voltage-gated calcium (Cav) channel Cav1.1 (Wu et al., 2015; Wu et al., 2016), which, given the significant similarities between Cav and Nav channels, provides an excellent base model for studying the structure and function of the mammalian Nav channels in lieu of the elusive Nav channel structure (Wu et al., 2015; Wu et al., 2016). In this Resource article, we have built a structure model of the human sodium channel Nav1.7 based on the Cav1.1 cryo-EM structure (PDB code: 5GJV). Disease-related mutations of various Nav channels are systematically mapped onto this Nav1.7 structural model. As expected, most mutations are located in the VSDs and the pore domain, which corroborate the functional disturbance associated with the various conditions. The human Nav1.7 structure model may also provide a useful tool for the structure-based design of drugs that are able to therapeutically target the Nav channels.

STRUCTURE MODEL OF HUMAN SODIUM CHANNEL Nav1.7

Homology models of the mammalian Nav channels have been previously constructed based on the crystal structures of the eukaryotic potassium channels or the prokaryotic sodium channels (Tikhonov and Zhorov, 2012; Yang et al., 2012). However, the relevance of such models has been in question, since the eukaryotic sodium channels are known to be heterotetrameric while the prokaryotic sodium channels and the potassium channels are of homotetrameric nature.

We sought to build a homology-based structural model for human Nav1.7 because of the tremendous interest in drug development targeting this channel. The sequence identity and similarity between human Nav1.7 and rabbit Cav1.1 are 21 and 35%, respectively (Please refer to the online Supplementary Fig. 2 of Wu et al., 2016). Importantly, the key amino acids within the VSDs and the pore domains are highly conserved (Wu et al., 2015; Wu et al., 2016). The cryo-EM structure of rabbit Cav1.1 was then used as the template for homology modeling of human Nav1.7. The primary sequence of human Nav1.7 was aligned with rabbit Cav1.1 in MOE with manual adjustment when necessary. The structure model of human Nav1.7 was created with the Homology Model module in MOE using the GB/VI scoring function with AMBER12:EHT force field (MOE, 2016).

The human Nav1.7 model structure resembles the structure of rCav1.1 in general (Fig. 1A). However, the model exhibits pronounced differences from the calcium channel and bacterial sodium channels particularly in selectivity filter. The SF of Nav1.7 consists of four different amino acid residues DEKA (Fig. 1B). In contrast, the Cav1.1 SF is constituted by four repeated essential glutamic acids, EEEE, while NavAb and NavRh contain TLESWS or TLSSWE in each protomer, respectively. This human Nav1.7 structure model represents the first one-chain sodium channel model with asymmetric repeats and is expected to shed new light on the mammalian sodium channel functions.

Figure 1
figure 1

Homology model structure of human Nav1.7 sodium channel. (A) Intra-membrane view and extracellular view of the structure model of Nav1.7. The four domains are colored green, light blue, cyan, and gray for domain I, II, III, and IV, respectively. (B) The pore domain of Nav1.7 structure model. The S5–S6 segments of Nav1.7 are shown and the four selectivity filter amino acids are shown as sticks (left). A close-up view of the four SF residues, D361 in domain I, E927 in domain II, K1406 in domain III, and A1698 in domain IV (right)

Full size image

MAPPING OF DISEASE-RELEVANT MUTATIONS ONTO THE Nav1.7 STRUCTURE MODEL

Human Nav1.7 sodium channel is preferentially expressed in the sensory neurons of dorsal root ganglia and sympathetic ganglia neurons, particularly within the nociceptors, which is essential for perceiving pain (Djouhri et al., 2003; Dib-Hajj et al., 2013). To date, about 60 mutations of Nav1.7 have been found to cause human pain syndromes including IEM, PEPD, CIP, SFN (small fiber neuropathy), DS (Dravet syndrome), and FEB (febrile seizure) (Fig. 2 and Table 1). We mapped all the reported Nav1.7 mutations onto this Nav1.7 structure model (Fig. 2). Nineteen out of 22 IEM mutations are located in the highly conserved regions of VSDs and the pore domain except for the Q10R, P610T, and G616R mutations (Fig. 2). Electrophysiology study showed that IEM mutations cause a prominent shift of the activation voltage toward a more negative region or delay deactivation, which results in neuron hyperexcitability (Choi et al., 2006; Lampert et al., 2006; Choi et al., 2009; Lampert et al., 2010). For example, mutation of A1643 within the S5 segment of domain IV to glycine (A1643G) generates a significant hyperpolarizing shift (Yang et al., 2016). Our structural analysis shows that only two IEM mutations F216S and L834R are located in the S4 positively charged segment that is directly responsible for transmembrane voltage sensing and channel activation. How other IEM mutations influence voltage sensing and channel functions is yet to be elucidated.

Figure 2
figure 2

Amino acid locations of Nav1.7 disease-related mutations on the Nav1.7 structure model. (A) The topology of human Nav1.7 sodium channel. Cylinders represent the transmembrane segments, which are colored in gray except that the S4 voltage-sensing segments are colored in yellow. The lines represent the soluble regions between the transmembrane segments or the N/C-terminus. The two P helices between S5 and S6 segments are shown in cylinders. Mutations of Nav1.7 are discriminately mapped on the topology scheme of Nav1.7 by different colors, namely, IEM (blue), PEPD (red), CIP (cyan), DS (purple), SFN (green), and FEB (pink). (B) Intra-membrane view and intracellular views of the Nav1.7 structure model. Mapping of disease-related mutations onto the Nav1.7 structure model is highlighted by different colors. Summary of Nav1.7 mutations is shown in different gray boxes

Full size image

The PEPD mutations are mostly characterized (nine out of 11) within the S4 segment, S4-S5 linker region, and the cytosolic regions of domain III and domain IV of Nav1.7 except for R185H and R1007C (Fig. 2A and Table 1). Specifically, I1472T, F1473V, and T1475I are within the IFMT motif (Fig. 2A), indicating that they may disturb channel inactivation. Indeed, IFMT mutations usually impair fast inactivation with consequently persistent currents (Fertleman et al., 2006). The V1309D, V1309F, and V1310F mutations are located in the S4-S5 linker region of domain III and they have been shown to cause moderate destabilization of fast inactivation (Jarecki et al., 2008). The G1618R mutation, located within the S4 segment of domain IV, impairs inactivation and retains a persistent current compared to the wild-type (WT) channel (Choi et al., 2011), while another domain IV S4 segment mutation, L1623P, significantly increases ramp current and shortens recovery time from inactivation (Suter et al., 2015). Moreover, electrophysiology study showed that M1638K mutation (within the S5 segment of domain IV) generates faster recovery from inactivation than the WT channel, producing greater currents and reducing the threshold with increased number of action potentials (Fertleman et al., 2006; Dib-Hajj et al., 2008). Another PEPD mutation, A1643E, also located in the S5 segment of domain IV, impedes channel full inactivation, which results in persistent inward currents (Estacion et al., 2008).

The CIP patients, characterized by lack of nociceptive perception, are mostly inflicted by Nav1.7 nonsense mutations, which result in premature protein truncations and inability to produce functional sodium channels. Only three mutations of Nav1.7, namely R907Q, A1247E, and W1786R, have been reported to be associated with CIP (Fig. 2 and Table 1). Diseases such as DS, SFN, and FEB are also known to be caused by Nav1.7 mutations (Fig. 2 and Table 1). For example, all eight SFN mutations have been characterized. Specifically, I228M, I731K, I750V, and M1543I mutations impair slow inactivation, D623N impedes slow and fast inactivation, while R185H, M943L, and V1002L mutations enhance resurgent currents (Faber et al., 2012a). On the other hand, Nav1.7 mutations that are associated with DS (nine mutations) and FEB (six mutations) have not been well characterized.

MAPPING OF OTHER HUMAN SODIUM CHANNEL DISEASE-RELATED MUTATIONS ONTO THE Nav1.7 STRUCTURE MODEL

Members of the human Nav channel family share high sequence similarity and mutations of these Nav channels are known to cause a vast variety of channelopathies. In order to better understand the role of those mutations in disturbing normal channel functions on a structural level, we mapped the disease-related mutations of other human Nav channels onto the Nav1.7 structure model based on the sequence alignment reported in Wu et al., 2016 (Fig. 3).

Figure 3
figure 3

Mapping of Nav channel disease-related mutations onto the Nav1.7 structure model. The Nav1.7 channel is shown in cartoon from the intra-membrane view. The Cα atoms of the disease-related amino acids are shown in spheres. Mapped mutations from nine Nav sodium channels to the Nav1.7 structure model are differentiated by distinct colors, Nav1.1 (A, blue), Nav1.2 (B, cyan), Nav1.3 (C, magenta), Nav1.4 (D, purple blue), Nav1.5 (E, pale cyan), Nav1.6 (F, orange), Nav1.7 (G, red), Nav1.8 (H, green), and Nav1.9 (I, salmon)

Full size image

Among all the nine Nav channels, Nav1.1 and Nav1.5 have the largest numbers of reported mutations (more than 400 each) (Fig. 3A and 3E), while Nav1.3, Nav1.8, and Nav1.9 have the least numbers (less than 10 each) (Fig. 3C, 3H, and 3I). Notably, mutations in Nav1.1, Nav1.2, Nav1.3, and Nav1.6 mainly cause epilepsies; those in Nav1.4 are related to myopathies; in Nav1.5 result in cardiac channelopathies; and in Nav1.7, Nav1.8, and Nav1.9 are associated with pain-related diseases (Fig. 3 and Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10). Mapping of all Nav channel mutations onto the Nav1.7 structure model revealed that more than 80% of mutations are located in the VSDs and pore domains (Fig. 4A and 4B). Notably, disease-causing mutations are somewhat equally distributed in all four Nav channel domains, which account for more than 20 sodium channelopathies (Fig. 4C). Furthermore, mutations are also distributed in various regions of the pore domains, suggesting that they may disturb Nav channel functions by altering sodium ion selectivity and conductivity (Fig. 4D).

Figure 4
figure 4

Mutations that cause sodium channelopathies are plotted on the Nav1.7 sodium channel model. (A) The amino acid residues related with sodium channelopathies are mapped on the Nav1.7 structure model. All mutated residues are shown in spheres and colored for Nav1.1 (blue), Nav1.2 (cyan), Nav1.3 (magenta), Nav1.4 (purple blue), Nav1.5 (pale cyan), Nav1.6 (orange), Nav1.7 (red), Nav1.8 (green), and Nav1.9 (salmon). (B) The distribution of sodium channelopathy-related mutations on the transmembrane regions of the Nav1.7 structure model. Mutations of the VSDs and the pore domain are shown from the intra-membrane and intracellular views. (C) The mutation distributions for the four domains. S1–S6 segments are shown in cylindrical helices. (D) Mapping mutations to the pore domain in four different views

Full size image

Nav1.2 mutations are largely associated with various epilepsy diseases, including BFIS3 (seizures, benign familial infantile 3), EIEE11 (epileptic encephalopathy, early infantile, 11), and DS (Fig. 3B and Table 3). More than 30 Nav1.2 mutations have been discovered and some of them are now functionally characterized. Interestingly, electrophysiological studies showed that Nav1.2 mutations can either be loss-of-function (R1319Q and L1330F) or gain-of-function (M252V, V261M, L1563V, and Y1579C) (Misra et al., 2008; Liao et al., 2010; Lauxmann et al., 2013). It is noted that BFIS3 mutations in Nav1.2 create less pronounced changes in the activation and inactivation potentials than the EIEE11 mutations (Shi et al., 2012).

Only six missense mutations of Nav1.3 have so far been identified in patients with cryptogenic partial epilepsy (Fig. 3C and Table 4). Five of them, namely K354Q, R357Q, D815N, E1160K, and M1372V, have been characterized, all of which are gain-of-function mutations, consistent with the neuronal hyperexcitability phenotype (Estacion et al., 2010; Vanoye et al., 2014).

Nav1.4 is essential for controlling the muscle action potential and consequently crucial for skeletal muscle contraction. Mutations of Nav1.4 are related with various neuromuscular disorders including PMC (paramyotonia congenita of von Eulenburg), HOKPP2 (periodic paralysis hypokalemic 2), HYPP (periodic paralysis hyperkalemic), NKPP (periodic paralysis normokalemic), MYOSCN4A (myotonia SCN4A-related), and CMS16 (myasthenic syndrome, congenital, 16) (Fig. 3D and Table 5). Different disease-causing mutations alter the Nav1.4 channel function through distinct mechanisms. For example, CMS16 mutations R104H, P382T, and C1209F completely abolish the Nav1.4 channel’s ability to conduct sodium ion, while other mutations such as M203K, R225W, and D1069N cause reduced action potential amplitude, leading to impaired channel function (Zaharieva et al., 2016). Compared to the WT channel, a CMS16 voltage sensor mutant R1457H requires longer hyperpolarization to recover which results in increased fast inactivation (Arnold et al., 2015). On the other hand, a HOKPP2 mutation R1135H (the third arginine in the domain III voltage sensor) exhibits increased depolarization, suggesting that R1135H mutation be gain-of-function (Groome et al., 2014). A MYOSCN4A mutation I582V shows a hyperpolarizing shift of 6 mV, indicating the nature of this mutation be also gain-of-function (Corrochano et al., 2014).

Nav1.6 is one of the sodium channels expressed in human brain and mutations of Nav1.6 cause EIEE13 (epileptic encephalopathy, early infantile, 13) (Fig. 3F and Table 7). More than 40 Nav1.6 mutations have been discovered since 2012 (Fig. 3F and Table 7), and seven of them have been studied in the functional assays. Specifically, five Nav1.6 mutations, namely T767I, N984K, T1716I, N1768D, and R1872W/R1872Q/R1872L, are characterized as gain-of-function, which cause hyperpolarizing shift of inactivation voltage or increased persistent current (Veeramah et al., 2012; Estacion et al., 2014; Wagnon et al., 2016), while the other two mutations, R223G and G1451S, are loss-of-function (de Kovel et al., 2014; Blanchard et al., 2015).

Five Nav1.8 mutations are associated with SFN, a condition that is clinically characterized by autonomic dysfunction and burning pain in the distal extremities (Fig. 3H and Table 8). Electrophysiology study has shown that Nav1.8 mutations, specifically L554P, A1304T, G1662S, and I1706V, accelerate inactivation recovery and enhance activation, which result in hyperexcitability (Faber et al., 2012b; Huang et al., 2013; Han et al., 2014). However, another SFN Nav1.8 mutation M650K causes reduced excitability of C fibers (Kist et al., 2016).

FEPS3 (episodic pain syndrome, familial, 3) and HSAN7 (neuropathy, hereditary sensory and autonomic, 7) are thought to be caused by the nine missense gain-of-function mutations of Nav1.9 (Fig. 3I and Table 9). Specifically, compared to the WT channel, R225C and A808G mutations induce hyperexcitability of the DRG neurons (Zhang et al., 2013), G699R enhances activation (Han et al., 2015), L811P significantly increases current density (Leipold et al., 2013), L1158P enhances spontaneous firing (Huang et al., 2014), and V1184A alters the channel voltage dependence that results in channel opening in response to hyperpolarized potentials (Leipold et al., 2015).

DISEASE-RELATED MUTATIONS IN SODIUM CHANNELS Nav1.1 AND Nav1.5

Mutations of Nav1.1 are associated with several neurological disorders including GEFS+2, EIEE6, ICEGTC, FHM3 (migraine, familial hemiplegic, 3), and FEB3A (febrile seizures, familial, 3A) (Table 2 and Table 10). More than 400 mutations of Nav1.1 have been identified, approximately 10% account for GEFS+2 while 80% for EIEE6 (Fig. 5A and Table 2). By mapping the Nav1.1-related mutations to the Nav1.7 structure model, we identified that most mutations are located in the VSDs and the pore domain (Fig. 5A). For example, mutations of the four positively charged residues, R1639G, R1642S, R1645Q, and R1648C, are present in the domain IV S4 segment (Table 2), suggesting that these EIEE6 mutations may alter the voltage sensing behavior of the channel. In addition, it is noteworthy that Nav1.1 mutations can be either loss-of-function or gain-of-function (Catterall et al., 2010; Escayg and Goldin, 2010). For example, two GEFS+2 mutations W1204R and R1648H increase the level of persistent current through gain-of-function (Lossin et al., 2002), while the loss-of-function M145T mutation in FEB3A decreases 60% of the current density (Mantegazza et al., 2005).

Figure 5
figure 5

Distributions of the missense mutations in Nav1.1 and Nav1.5. (A) Distributions of Nav1.1 missense mutations on the Nav1.7 model structure. More than 400 mutations are mapped. Mutations from five Nav1.1-related diseases are shown from intra-membrane, intracellular, and extracellular views. The Nav1.7 model is shown in cylindrical helices and colored by GEFS+2 in red, EIEE6 in blue, ICEGTC in cyan, FHM3 in green, and FEB3A in yellow. (B) Distributions of Nav1.5 related-disease mutations on the Nav1.7 structure model. Mutations from Nav1.5 related diseases are shown from intra-membrane, intracellular, and extracellular views. Different diseases are colored in green for PFHB1A, blue for LQT3, red for BRGDA1, cyan for SSS1, and magenta for VF1, SIDS, ATRST1, CMD1E, ATFB10, and MEPPC

Full size image

Nav1.5 is the primary sodium channel in the heart and is essential for the cardiac action potential initiation. More than 400 Nav1.5 mutations have been discovered and they are implicated in a wide variety of cardiac diseases—including PFHB1A (progressive familial heart block 1A), LQT3, BRGDA1, SSS1, VF1 (familial paroxysmal ventricular fibrillation 1), SIDS (sudden infant death syndrome), ATRST1 (atrial standstill 1), CMD1E (cardiomyopathy, dilated 1E), ATFB10 (atrial fibrillation, familial, 10), and MEPPC (multifocal ectopic Purkinje-related premature contractions) (Fig. 5B and Table 6). By mapping all the Nav1.5 mutations onto the Nav1.7 structure model, it shows that most mutations are located in the transmembrane regions of the channel, suggesting that these mutations might disturb voltage sensing or sodium conduction (Fig. 5B). Furthermore, about 50% of the Nav1.5 mutations account for BRGDA1, while 30% for LQT3. Similar to the case of Nav1.1, mutations in Nav1.5 can be either loss-of-function or gain-of-function. For example, loss-of-function mutations are associated with BRGDA1, CMD1E, SSS1, and ATFB10 (Tan et al., 2001; Smits et al., 2005; Makiyama et al., 2008; Laurent et al., 2012), while gain-of-function mutations of Nav1.5 are responsible for LQT3 (Remme et al., 2006), CMD1E, and ATFB10 (Olson et al., 2005), and most recently MEPPC (Swan et al., 2014).

CONCLUDING REMARKS

The Nav family of sodium channels are important drug targets for the pharmaceutical industry. However, no atomic structure of any mammalian Nav channels is currently available, preventing the establishment of an in-depth structure-function relationship for this important group of sodium channels and application of structure-based approach to rationally design compounds that are able to modulate the functions of those Nav channels in a disease relevant manner. Using the recently published cryo-EM structure of a rabbit Cav channel Cav1.1, we established an atomic level heterotetrameric structure model for the human Nav channel Nav1.7. Disease-related mutations of Nav1.7 and other members of the Nav family, which are largely responsible for many neurological disorders like epilepsies, pains, and myopathies, are mapped onto the structure model. Taken together the available functional data, we attempted to establish a rudimentary structure-function relationship for human Nav1.7 and other members of the Nav channel family. It is noticeable that sodium channelopathies can be attributed to both loss-of-function and gain-of-function mutations.

However, we must realize that the current Nav1.7 structural model has its limitation and the atomic resolution mammalian Nav channel structure is urgently needed. In recent years, cryo-EM technology is becoming a mainstream technology for structural biology, which is able to potentially overcome the significant technical hurdles in producing challenging proteins such as mammalian Nav channels in sufficient quality and the necessity of crystallization for structural elucidation. Detailed mechanisms of how the Nav channels sense voltage changes and conduct sodium ions can only be answered when such atomic resolution structures become available. We hope the Nav1.7 structure model presented here is a temporary surrogate to help understand the Nav channel functions, particularly those relevant to the various neurological diseases, at atomic level, and contributes to the structure-based rational design of the next generation Nav channel modulators.

SUMMARY OF DISEASE-RELATED MUTATIONS FOR SODIUM CHANNELS

Most of the Nav channel disease-related mutations are extracted from the UNIPROT websites:

http://www.uniprot.org/uniprot/P35498 (Nav1.1);

http://www.uniprot.org/uniprot/Q99250 (Nav1.2);

http://www.uniprot.org/uniprot/P35499 (Nav1.4);

http://www.uniprot.org/uniprot/Q14524 (Nav1.5);

http://www.uniprot.org/uniprot/Q9UQD0 (Nav1.6);

http://www.uniprot.org/uniprot/Q15858 (Nav1.7);

http://www.uniprot.org/uniprot/Q9Y5Y9 (Nav1.8);

http://www.uniprot.org/uniprot/Q9UI33 (Nav1.9).

In the UNIPROT websites, there are no mutations described for Nav1.3. During literatures searching, we found that six mutations of Nav1.3 are associated with cryptogenic partial epilepsy. Except for the present mutations in the UNIPROT websites, we found additional mutations of Nav channels in literatures. All mutations are summarized in Tables 1, 2, 3, 4, 5, 6, 7, 8, 9. However, we recognize that our summary may not contain all Nav channel disease-related mutations owing to abundant literatures reporting Nav channel disease-related mutations and increasing volume of work describing new findings.