Regulation of Nav1.5 by β-Subunits
The β subunit family consists of four different proteins β1–4 encoded by four genes,
SCN1B–SCN4B, respectively, with β1 alternatively spliced into two isoforms, β1A and β1B
[190]. The β-subunits, most likely β1-subunits, assemble with Na
v1.5 at the endoplasmic reticulum and influence its maturation and trafficking to the plasma membrane
[97][191]. Alpha-beta subunits assembly is either covalent (β2 or β4) or non-covalent (β1 or β3)
[192]. Particularly, β 4-Na
v1.5 covalent association is ensured by an extracellular cysteine–cysteine single disulfide bond
[193][194], while β2 does not form a disulfide linkage at this position with Na
v1.5 as recently specified
[5], whereas β1 and β3 non-covalently interact with Na
v1.5 through the channels DIV and DIII voltage gating domain respectively
[195].
Despite the structural similarities between β2/β4 on one hand and β1/β3 on the other hand, their expression differs from one cellular sub-domain to another. Inside the cardiomyocyte, β3 are expressed at the T-tubules and β4 at the ID, while β1 and β2 are found at both locations
[186][196][197]. Zimmer et al. have suggested that, unlike β2, β1 associates to Na
v1.5 early at the ER, and both α and β1 subunits are trafficked together to their final destination at the cell membrane
[198]. Subsequent studies revealed that β1-subunits enhance the α-subunits dimerization and promote the dominant-negative effect of trafficking defective mutants
[199]. β2 has been reported to promote surface localization of Na
v1.5
[200]. Importantly, β3 subunits have been demonstrated to bind to Na
v1.5 in multiple sites and promote the formation of α subunit oligomers, including trimers
[201]. However, β4 has been reported as a modulator of Na
v1.5 kinetic and gating properties by increasing
INa [202]. Taken together, these findings are consistent with the idea that the distinct sodium channel β subunits provide support for the pore-forming subunit, facilitate the trafficking of the mature channel to the different membrane domains, and modulate the gating function of Na
v1.5 by increasing the
INa [203][204][205][206][207].
The Nav1.5 and the Intercalated Disc Interactome
As suggested by the Delmar research team, several evidence point to the fact that the ID is not a hub of proteins playing independent functions within the cardiomyocyte, but rather a network of molecules interacting together in order to fulfill a specific function (AP propagation, cell-to-cell coupling, cardiac excitability, etc.) that cannot be accomplished if this “interactome” is impaired
[208]. As a component of the ID proteins, Na
v1.5 has been demonstrated to be in the heart of this interactome by physically and functionally associating to several proteins belonging to this macromolecular complex.
In this context, it is currently well known that Na
v1.5 targeted to the ID are “tagged” with synapse-associated protein 97 (SAP97), a scaffolding MAGUK ((membrane-associated guanylate kinase) protein that is abundantly expressed in human and rat ventricular myocardium
[209]. SAP97 has been introduced as the determinant of the Na
v1.5 ID pool as it plays an important role in targeting Na
v1.5 along with Kir2.1 to this cell membrane domain
[209][210]. Both channels were structurally evidenced to co-assemble to SAP97 by their C-terminal domains
[209][211]. For Na
v1.5, it is assumed that the last three amino-acids (serine–isoleucine–valine or SIV motif) of the C-terminal region form a PDZ (postsynaptic density protein (PSD95), Drosophila disc large tumor suppressor (Dlg1), and zonula occludens-1 protein (zo-1) domain binding motif) that interacts with the syntrophin–dystrophin complex at the cardiomyocyte LM and PDZ domains of SAP97 at the ID
[189]. In the absence of the PDZ-domain-binding motif of Na
v1.5 or SAP97, Na
v1.5 expression at the cell surface decreased, thus leading to a reduction in the cardiac
INa in vitro
[212]. However, a subsequent study by the same team demonstrated that in vivo ablation of SAP97 did not change Na
v1.5 localization and function, but it did decrease the cardiac potassium currents
[213]. The authors of these studies justified this discrepancy by the fact that SAP97 silencing in vitro is induced in adult cardiomyocytes while in vivo, it is a constitutive ablation present early in development, which may impact protein expression and interactions.
In addition, the Na
v1.5-SAP97-Kir2.1 complex has been demonstrated to reach the ID through the microtubule highway
[103][209][210][214]. Although the exact mechanism by which Na
v1.5 is targeted to the ID is not yet fully discovered, part of it is already elucidated. A few years ago, Agullo-Pascual et al. proved for the first time that the microtubule plus-end tracking protein “end-binding 1” (EB1) is captured to the IDs by connexin 43 (cx43), which facilitates the cargo delivery, including Na
v1.5
[215]. These findings are consistent with Marchal and co-workers’ recent study in which they have further proved that EB1 modulates Na
v1.5 trafficking to the IDs and that loss of EB1 function leads to reduced
INa and conduction slowing
[216]. Moreover, EB1 has been previously demonstrated to bind directly to CLASP2 (cytoplasmic linker associated protein 2) and form a complex at the microtubule plus-end, promoting thus microtubule polymerization and stabilization
[217]. Interestingly, inhibiting the GSK3β (glycogen synthase kinase 3β)-mediated phosphorylation of CLASP2 enhanced the EB1–CLASP2 interaction, which in turn led to an increased Na
v1.5 delivery at the ID of cardiomyocytes and an increased
INa [216]. Furthermore, Rhett et al. have shown that in addition to its known localization at the gap junction where it interacts with zonula occludens-1 (ZO-1)
[218][219], Cx43 also co-localizes with ZO-1 in the zone surrounding the gap junction, conventionally termed as perinexus and that Cx43 but not ZO-1 interact with Na
v1.5 at this zone in physiological conditions
[220]. In vivo and in vitro assays show that Na
v1.5 expression and function are reduced as a result of Cx43 expression/function decrease, thus giving more evidence that Cx43 is required for a proper Na
v1.5 function at the ID
[221].
Importantly, Na
v1.5 and Cx43 interaction at the perinexus is thought to be mediated by scaffolding proteins SAP97 and Ankyrin G (AnkG) as their interaction has been reported
[212][222]. In the cardiovascular system, ankyrins are critical components of ion channels and transporter signaling complexes, and their dysfunction has been linked with abnormal ion channel and transporter membrane organization and fatal human arrhythmias
[223]. Although both ankyrin-B (AnkB, encoded by
ANK2) and ankyrin-G (
ANK3) have been found to be expressed in the myocardium, only ankyrin-G has been shown to interact with Na
v1.5
[224]. Specifically, AnkG is necessary for normal expression of Na
v1.5 and acts as a coordinating signaling center, functionally coupling Na
v1.5 gating with upstream kinase and phosphatase enzymes and downstream cytoskeletal proteins
[80][225]. AnkG is primarily expressed at the ID membrane and T tubules, where it co-localizes with Na
v1.5
[112]. In vitro, it has been demonstrated that AnkG binds to Na
v1.5 and that AnkG downregulation impaired the subcellular localization of Na
v1.5 and reduced the
INa current amplitude
[226][227]. In vivo, Makara and his collaborators have demonstrated that AnkG plays an indispensable role in directing Na
v1.5 and its regulatory protein CaMKII to the ID
[225][228]. Mutational studies have further confirmed that disrupting the binding of AnkG to Na
v1.5 impairs AnkG dependent targeting of the Na
+ channel to the ID leading thus to a reduction in
INa density and cardiac arrhythmias
[224][225][229]. A recent study performed by Yang et al. has demonstrated that AnkG, but not AnkB, are expressed at the IDs and that masking Na
v1.5 binding sites in AnkG using competitive peptides caused a decrease in sodium channel current (
INa) and targeting defects of the Na
+ channels to the ID, but not to LM
[184]. However, a more recent study by Cavus and collaborators specified that only canonical AnkG isoforms have this regulatory effect on Na
v1.5 and that noncanonical (giant) AnkG isoforms mediated electrical dysfunction is independent of Na
v1.5
[230].
Furthermore, AnkG is thought to mediate the interaction between Cx43 and PKP2, thus connecting desmosomal proteins with the molecular complex that captures the microtubule plus-end at the ID, thus allowing for delivery of Na
v1.5
[215][227][231]. This is consistent with the fact that loss of desmosomal integrity impacts cardiac conduction and leads to cardiac arrhythmias
[231][232][233]. Accordingly, loss of Plakophilin-2 (PKP2), a crucial component of the cardiac desmosome, has been demonstrated to decrease
INa in cardiac myocytes
[234]. Similarly, loss of PKP2 expression in HL1 cells and in induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) from a patient with PKP2 deficiency reduced
INa amplitude
[232][235]. Likewise, Rizzo et al. have demonstrated that desmoglein-2 (Dsg2), another desmosome protein, physically interacts with Na
v1.5 at the ID of mouse cardiomyocytes in vivo
[236]. They showed that mice models over-expressing a desmoglein-2 mutation present a wider intercellular space at the level of the ID, longer ventricular activation time, lower conduction velocity, lower upstroke velocity, and lower
INa amplitude compared to wild type. Although no evidence of direct interaction between Desmoplakin (DSP) and Na
v1.5 has been reported, RNAi-based Desmoplakin silencing in vitro resulted in a reduction in Na
v1.5 expression at the ID of cardiomyocytes, an abnormal sub-cellular distribution of Cx43 and Na
v1.5,
INa decay, and slowed conduction velocity suggesting that DSP regulates Na
v1.5
[237].
Similarly, AnkG is established as an adaptor protein that organizes, transports, and anchors Na
v1.5 to the actin/spectrin cytoskeleton
[238][239][240]. In fact, the AnkyrinG-Na
v1.5 complex is believed to connect with the actin/α-spectrin cytoskeleton through CaMKII-β
IV-spectrin interaction where the latter acts as a CaMKII-anchoring protein and thereby orchestrating the whole macromolecular complex; however, no evidence of direct interaction between Na
v1.5 and β
IV-spectrin has been found yet
[226]. On the other hand, β
IV-spectrin is assumed to control the CaMKII-dependent regulation of Na
v1.5 at the ID, and loss of β
IV-spectrin/CaMKII interaction precludes CaMKII-dependent phosphorylation of Na
v1.5 at Serine 571 in the DI–DII linker and abolishes the stress-induced activation of the pathogenic
INa,L [241][242].
Remme’s team
[243] has recently demonstrated that ID Na
v1.5 physically interacts with coxsackie and adenovirus receptor (CAR), a single-pass transmembrane cell adhesion molecule (CAM)
[244]. Furthermore, they have demonstrated that CAR haploinsufficiency decreased
INa amplitude at the ID, which in turn reduced sodium channel availability at this cell membrane compartment. Na
v1.5–CAR interaction is only beginning to be understood, and thus, mechanisms underlying this interaction are still to be studied.
Over the last decades, several controversial studies emerged regarding the sodium channel α-α-subunits interaction and dimerization. However, Clatot and co-workers settled this controversy by demonstrating for the first time that trafficking-defective Na
v1.5 exerts a dominant-negative effect on non-defective ones through α-α-subunits physical interaction at their N-terminal regions, precluding thus their cell surface expression
[245]. Building on these findings, the team further evidenced that cardiac sodium channel α-subunits assemble as dimers with coupled gating and that this dimerization is mediated through an interaction site found within the DI-II linker of Na
v1.5, between amino acids 493 and 517
[246]. Curiously, earlier studies have shown that 14-3-3 protein, a member of highly conserved cytosolic acidic proteins, physically interacts with the DI-II linker of Na
v1.5 (between amino acid 417 and 467) at the ID and that this interaction facilitates the dimerization of cardiac sodium channels
[247]. Strikingly, Clatot et al. identified a second 14-3-3 protein-Na
v1.5 interaction site between amino acid 517–555 and demonstrated that co-operative gating behavior but not dimerization of α-subunits is dependent on 14-3-3-Na
v1.5 interaction
[246].
Nav1.5 and the Lateral Membrane’s Interactome
Dystrophin is known to indirectly mediate Na
v1.5 expression at the LM through binding to Syntrophin adapter protein which physically associates to the PDZ domain-binding motif at the C-terminal region of Na
v1.5
[3][212][248][249][250]. Interestingly, Matamoros et al. demonstrated that α1-syntrophin also interacts with the N-terminal region of Na
v1.5 through an “internal” PDZ-like binding domain localized at this region which acts as “chaperone-like” domain that increases Na
v1.5 density at the LM and
INa [251]. The same mechanism has been validated for Kir2.1 and Kir2.2 that were demonstrated to reciprocally interact with Na
v1.5 channels and modulate each other’s trafficking and expression
[251][252].
Interestingly, Na
v1.5 has been demonstrated to interact with CASK (calcium/calmodulin-dependent serine kinase), a member of the MAGUK protein family
[253]. In several ways, CASK is considered an unconventional Na
v1.5 regulator since it is the only MAGUK protein that is lateral membrane-specific and also the only Na
v1.5 interacting protein that exerts a repressive effect on the functional expression of Na
v1.5, most likely by preventing its early trafficking to the LM. In this regard, CASK has been demonstrated to decrease
INa when the former is over-expressed and to increase
INa when CASK is inhibited in vivo and in vitro
[253].
In addition, Na
v1.5 has been evidenced to interact with members of the Z-line scaffolding protein complex, such as α-actinin-2 and telethonin. While α-actinin-2 is currently known to physically interact with Na
v1.5 through the channel DIII–DIV linker
[254], the telethonin interaction site on Na
v1.5 has not yet been identified
[71]. α-actinin-2 is thought to positively regulate Na
v1.5 by increasing its cell surface expression, most likely through promoting its anchoring to the contact zones between T-tubules and Z-lines and connecting the channel to the actin cytoskeleton network
[254]. However, scarce information is available regarding the mechanism of Na
v1.5 regulation by telethonin, although physical interaction between TCAP and Na
v1.5 was evidenced by co-immunoprecipitation methods and mutations in the telethonin coding gene (
TCAP) has been found to alter the channel-gating properties of Na
v1.5 in patients with abnormal gut motility and Brugada syndrome
[255][256].
Moreover, the role of fibroblast growth factor homologous factors (FHFs), a subset of the fibroblast growth factor (FGF) family
[257], has been well elucidated modulating the neuron voltage-gated sodium channels
[258]. However, their role in controlling cardiac sodium channel function is still poorly understood and subject to debate. In this respect, fibroblast growth factor homologous factor 1B (FHF1B), also known as FGF12B, has been reported to regulate the biophysical properties and kinetics of Na
v1.5 through its physical interaction (amino acids 1773–1832) with the Na
v1.5 C terminal region
[259]. Both in vitro data show that FHF1B interacts with Na
v1.5, and this interaction results in hyperpolarizing shift in steady-state inactivation of this channel
[259]. However, the opposite effect has also been reported where a depolarizing shift in the V1/2 of steady-state inactivation has been attributed to the FHF1B-Na
v1.5 interaction
[260]. Furthermore, FGF13 (FHF2), which is the major FHFs in adult mouse hearts, has been identified as a Na
v1.5 interacting protein
[260]. In the cardiomyocyte, FHF2 co-localizes with distinct Na
v1.5 pools, i.e., the LM and ID suggesting an important role for FHF2 modulating Na
v1.5 cell surface expression and function
[261]. Like FGF12B, FGF13 physically binds to Na
v1.5 through the channel’s C terminus region. In vivo, FGF13 knockdown altered Na
v1.5 function resulting in a decreased
INa current density, reduced Na
v1.5 channel availability, slowed Na
v1.5, and reduced
INa current recovery from inactivation
[260]. This effect of FGF13 is isoform-specific
[262]. FHFs have also been implicated in voltage-gated sodium channel trafficking control. In this context, FGF14 has been reported as a modulator of Na
v1.5 current densities in neurons and in the heart by impairing their biophysical properties or by controlling channel trafficking and cell surface expression in vitro
[263].
Furthermore, calmodulin (CaM), a ubiquitous Ca2
+-sensing protein, has been reported to interact with Na
v1.5 N- and C-terminal regions
[264][265][266][267] and the DIII–IV linker
[144][265]. This interaction has been demonstrated to enhance slow inactivation and modulate Na
v1.5 gating
[266], while disruption of CaM binding to Na
v1.5 decreases channel activity and enhances the propensity for persistent Na
+ current, all resulting from a switch in the Na
V inactivation mechanism
[267]. Na
v1.5–CaM interaction has been further studied in a mutational context related to cardiac sodium channelopathies (See
Section 4).
Finally, dipeptidyl peptidase-like protein-10 (DPP10), previously reported as a modulator of K
v4.3-current kinetics
[268], has recently emerged as a new regulator of Na
v1.5
[269]. In vivo, DPP10 has been reported to modulate Na
v1.5 current kinetics as well by altering voltage dependence of Na
+ current and upstroke velocity of the action potential
[269].
The Caveolar Nav1.5
Cardiac sodium channels have also been localized to cardiomyocyte caveolae, which are specialized subsarcolemmal membrane compartments enriched in lipids and play a crucial role in vesicular trafficking and protein targeting to the cell surface
[270][271]. Caveolar Na
v1.5 is exposed to a very rich macromolecular complex encompassing fatty acids, ion channels (pacemaker channels, potassium channels, calcium channels, etc.), and signaling complexes (G-protein-coupled receptors, protein kinases, etc.). This microenvironment has been reported to regulate Na
v1.5 function and membrane expression in a multilayers fashion
[271].
The first layer is related to the biochemical properties of caveolae itself as a specialized lipid raft rich in fatty acids. In this regard, previous reports demonstrated that Na
v1.5 is blocked by polyunsaturated fatty acids (PUFAs), suggesting that interaction of Na
v1.5 with the caveolar lipids that also include PUFAs might have the same effect
[272][273]. Nonetheless, the mechanism by which caveolar lipid rafts regulate Na
v1.5 is not yet fully understood.
The second layer of caveolar Na
v1.5 regulation is mediated by caveolins which are the major proteins of caveolae
[271]. This mechanism was first reported by the Shibata group, which demonstrated that in addition to the indirect β-adrenergic regulation of Na
v1.5, which is PKA-dependent, stimulation of the β-adrenergic pathway in the presence of a PKA inhibitor, activates G-protein (Gsα) cascade, which in turn leads to a rapid increase of
INa [270]. A subsequent study by the same group suggested that caveolar Na
v1.5 channels are stored at caveolae invaginations and that PKA-independent Gsα-dependant stimulation of the β-adrenergic pathway leads to the opening of caveolae, the exposition of Na
v1.5 channels to the extracellular environment, which in turn increase
INa [274]. This mechanism has been completely neutralized by anti-caveolin 3 antibodies dialyzed into the myocytes suggesting that caveolar Na
v1.5 function is dependent on the Gsα-Caveolin 3 (Cav3) interaction
[274]. Although Na
v1.5 has been confirmed to interact with caveolin 3 in rodent and human cardiomyocytes
[270][275], it is not yet clear if this interaction is direct or indirect. Several reports suggested that Cav3 modulates Na
v1.5 function indirectly through inhibiting the nNOS, which is a part of the Na
v1.5-SNTA1-PMCA4b macromolecular complex
[275][276].