The concept that the heart beats as a functional syncytium composed of discrete muscle cells electrically coupled by gap junctions (GJs) is a foundational theory of modern cardiology (Kléber and Rudy, 2004). Moreover, as GJ remodeling occurs in multiple cardiac pathologies (Jongsma and Wilders, 2000; Stroemlund et al., 2015), it is widely held that disrupted GJ coupling is mechanistically central to arrhythmogenesis in heart disease (De Vuyst et al., 2011; Palatinus et al., 2012). Whilst a GJ-based paradigm for cardiac electrical coupling has been in place for over 50 years, a small group of mathematical biologists has proposed that cardiac conduction in health and disease may involve ephaptic mechanisms (Lin and Keener, 2010; Mann and Sperelakis, 1979; Mori et al., 2008; Pertsov and Medvinskiĭ, 1979). Ephaptic conduction is conceived as involving the intercellular transmission of action potentials via ion accumulation/depletion transients occurring within narrow extracellular clefts between closely apposed myocytes. Further interest in this hypothesis has been stoked by provocative findings such as conduction still occurring in mice with knockout of the principal ventricular GJ protein, connexin43 (Gja1/Cx43) (Gutstein et al., 2001) and in humans with dominant negative mutations of GJA1, the gene encoding Cx43 (Shibayama et al., 2005).

Although the ephaptic hypothesis has remained controversial, owing to a lack of direct experimental evidence, theoretical studies have identified two key elements of a hypothetical structural unit that might support this alternative mechanism of conduction (Mori et al., 2008; Hichri et al., 2018; Kucera et al., 2002; Lin and Keener, 2013): 1) Close proximity (<30 nm) between the membranes of adjacent myocytes and 2) a high density of sodium channels at such points of close inter-membrane apposition. These prerequisites for the formation of a cardiac ephapse pointed to its likely location in the intercalated disk (ID) where cardiac sodium channels are concentrated (Maier et al., 2004; Petitprez et al., 2011). Here, we report that α (Scn5a/Na_V1.5) and β (Scn1b/β1) subunits of cardiac sodium channels preferentially localize at the perinexus, a nanodomain at the edge of GJs, (Rhett et al., 2011) where the membranes of apposed cells routinely approach each other closely enough for ephapse formation (Veeraraghavan et al., 2015). Additionally, we demonstrate that adhesion mediated by the extracellular immunoglobulin (Ig) domain of β1 generates close proximity between the membranes of adjacent cells within perinexal nanodomains, likely enabling these to function as ephapses. Importantly, we identify dynamic changes in perinexal ultrastructure, secondary to compromised β1-mediated adhesion, as a novel mechanism for arrhythmogenic conduction defects.

In this study, we tested the hypothesis that Na_V1.5-rich nanodomains within the ID may enable ephaptic coupling between cardiac myocytes, and that the cell adhesion function of β1 may be important to this phenomenon. We present an array of experimental evidence detailing the nanoscale location, structural properties, and makeup of Na_V1.5-rich ID nanodomains and demonstrate that inhibition of β1-mediated adhesion disrupts these nanodomains, with proarrhythmic consequences.

Sodium channel subunit proteins Nav1.5 and β1 differentially localize within ventricular ID nanodomains

The cardiac voltage-gated sodium channel is a trimer comprised of a pore-forming alpha subunit (Na_V1.5) and two non-pore-forming β subunits, the latter possessing an extracellular Ig domain, which facilitates adhesive interactions (Namadurai et al., 2015; O'Malley and Isom, 2015). To determine the distribution of the cardiac sodium channel complex we raised, characterized, and rigorously validated rabbit polyclonal antibodies against Na_V1.5 and β1. Both antibodies displayed intense immunolabeling of N-cadherin (N-cad) -positive IDs (Figure 1A, Figure 1—figure supplement 1) in laser scanning confocal microscope (LSCM) images of guinea pig (GP) ventricular tissue and single bands at the expected molecular weights in western blots (Figure 1—figure supplement 1; detailed results from validation studies are provided in supplementary material). Also, in dot blot experiments, both antibodies displayed selective affinity for their corresponding epitopes, with no evidence of cross-reactivity to each other’s epitopes (Figure 1—figure supplement 2).

Figure 1 with 3 supplements see all

Download asset Open asset

While the observation of ID-enrichment of Na_V1.5 (Figure 1A, Figure 1—figure supplement 1A) is consistent with previous studies (Maier et al., 2004; Petitprez et al., 2011), en face views of IDs reconstructed from LSCM optical sections revealed a previously unreported aspect of β1 organization (Figure 1B). In these end-on views of IDs, β1 displayed a lattice-like distribution with specific immunolocalization within N-cad-free interplicate sectors of the ID (Figure 1B, Figure 1—figure supplement 3C). Interplicate regions of the ID are well-characterized as being enriched in Cx43 GJs (Severs, 2000). In line with this, punctate Cx43 immunosignal corresponding to GJ demonstrated close association with the strands comprising the β1 lattice (Figure 1—figure supplement 3B).

The LSCM observations were confirmed by super-resolution STochastic Optical Reconstruction Microscopy (STORM) at sub-diffraction resolution (20 nm lateral, 40 nm axial) (Figure 1C–F). A significant proportion of immunolocalized Na_V1.5 molecules were organized into clusters preferentially localized adjacent to clusters of Cx43 molecules (i.e. GJs; Figure 1C), while a population of Na_V1.5 molecules was also observed co-distributing with N-cad (Figure 1D). In contrast to Na_V1.5, β1 molecules organized into a lattice-like distribution with β1 strands punctuated by tight side-by-side association with Cx43 clusters (Figure 1E), but extending almost exclusively through N-cad-free interplicate zones of the ID (Figure 1F).

STORM-RLA indicates Na_V1.5 distributes between two pools within the intercalated disk

To quantitatively assess the overlapping, but distinct distributions of β1 and Na_V1.5 relative to Cx43 (GJs) and N-cad (plicate zone/adherens junctions) within the ID, we used STORM-based relative localization analysis (STORM-RLA; Figure 2) (Veeraraghavan and Gourdie, 2016). Briefly, relative localization of co-labeled proteins is quantitatively assessed by detection of clusters of localized molecules, and measurement of overlap, and closest distances between clusters.

Figure 2

Download asset Open asset

In accordance with visual assessment (Figure 1), STORM-RLA indicated that nearly half of ID-localized Na_V1.5 clusters (48.1%) were located adjacent Cx43 GJ (Figure 2A,C), a sub-population we term the perinexal pool. This pool included 31% of Na_V1.5 clusters that did not overlap Cx43 clusters, and an additional 17.1% which tangentially overlapped Cx43 clusters (Figure 2A,B,C). A second, plicate pool of Na_V1.5, localized to N-cad-rich plicate ID regions, was found to contain 29% of ID-localized Na_V1.5 (Figure 2A,B,D), consistent with previous reports from us (Veeraraghavan and Gourdie, 2016) and the Delmar group (Leo-Macias et al., 2016). Closely paralleling the Na_V1.5 case, nearly half (48.3%) of ID-localized β1 clusters were also identified within perinexal nanodomains (Figure 2A,E). However, only 7.6% β1 clusters were co-located with N-cad clusters (Figure 2A,F), suggesting that an overwhelming 92.4% of ID-localized β1 clusters are located in N-cad-free, GJ-rich interplicate ID regions. In summary, the pore-forming subunit Na_V1.5 is divided between two significant pools within the ID, a perinexal pool, and a plicate pool; however, the β1 subunit is mainly localized in perinexal, and other interplicate ID sites.

βadp1 – a rationally designed inhibitor of β1-mediated adhesion

The extracellular Ig domain of β1 facilitates adhesive interactions, including trans homophillic interaction with β1 molecules on neighboring cells (Calhoun and Isom, 2014). In order to assess this intercellular adhesion function, as well as to create an assay model for identification of β1 adhesion antagonists, we quantified intercellular junctional resistance changes in cells heterologously overexpressing β1 (1610 β1OX) using electric cell-substrate impedance sensing (ECIS). This technique measures the electrical impedance offered by a monolayer of cells to current flow between electrodes located above, and below the monolayer, and previous studies demonstrate that the resistance of the extracellular cleft at cell-cell junctions (junctional resistance) is well-reflected by measurements at low frequencies (62.5–4000 Hz) (Moy et al., 2000; Tiruppathi et al., 1992). Junctional resistance was markedly higher (>3 fold) in 1610 β1OX cells, compared to cells without β1 (1610 Parental; Figure 3A), consistent with the expression of β1 increasing levels of adhesion between the cells. Immunosignals for tight junction proteins and Ca²⁺-dependent cadherins were below detectable levels, and did not change with β1 expression (data not shown), suggesting that these proteins were unlikely to account for adhesion observed between 1610 cells.

Figure 3 with 2 supplements see all

Download asset Open asset

Our strategy for preparing a rationally designed antagonist of β1-mediated adhesion was based on an approach successfully applied to two other adhesion proteins of the Ig domain family that share homologies with β1: N-cad (Williams et al., 2000a; Williams et al., 2000b) and desmoglein-1 (Schlipp et al., 2014). This method involves synthesizing a peptide mimicking the adhesion sequence within the Ig loop domain. The location of the adhesion receptor in sodium channel β Ig domain has been characterized in previous structure and function studies (Brackenbury and Isom, 2011; Malhotra et al., 2000; Patino et al., 2009). Based on the recently resolved crystal structure of β3, and structural data from cryo-electron microscopy studies of β1 in the electric eel, we generated a molecular homology model of β1 using SWISS-MODEL (Figure 3B) (Namadurai et al., 2014). We then identified an amino acid sequence (1FVKILRYENEVLQLEEDERF20: βadp1) within the β1 Ig domain homologous to the desmoglein-1 Ig loop adhesion sequence - a peptide mimetic of which acts as a competitive inhibitor of desmosomal adhesion (Harrison et al., 2016). To assess the interaction of βadp1 with β1, and thus its potential ability to act as a competitive inhibitor of β1 adhesion, we allowed the peptide to bind freely to the β1 homology model in silico, determining that it associated with the adhesion surface of the Ig loop (Figure 3C). MM-GBSA refined docking of βadp1 using Maestro displayed a low energy conformation (ΔG binding −44.47) at this location. The predicted binding pose revealed stabilization of this conformation via a network of intramolecular hydrogen bonds, and electrostatic interactions between the C-terminal end of βadp1 and polar residues within the β1 adhesion domain, as well as accommodation of hydrophobic residues of βadp1 within pockets along the β1 adhesion surface. Sequences of similar length from the Ig loop outside of the adhesion domain, as well as a randomized βadp1 sequence (βadp1-scr), did not show propensity to interact with the adhesion surface of the β1 extracellular domain. Substituting the basic arginine (R) at position 19 of βadp1 with acidic aspartic acid (D) yielded a sequence (βadp1-R85D) with significantly abrogated binding to the β1 model in silico (ΔG binding −18.05). The arginine at this position of βadp1 corresponds to R85 on full length β1 and is a mutational hotspot associated with diseases of electrical excitation in the heart and brain of humans (Calhoun and Isom, 2014; Thomas et al., 2007).

We next assessed the ability of βadp1 to abrogate β1 adhesion using the 1610 β1OX model. βadp1 significantly, persistently and dose-dependently decreased junctional resistance in confluent monolayers of 1610 β1OX cells, but not in 1610 parental cells (Figure 3D), which express negligible levels of β1. These results suggest that βadp1 selectively inhibited the intercellular adhesion function of β1 in this in vitro model. Similarly, βadp1 efficiently and dose-dependently inhibited the formation of intercellular interactions between sub-confluent 1610 β1OX cells forming into monolayers (Figure 3E), doing so at lower concentrations of peptide than necessary for established confluent monolayers of β1 over-expressing cells. Neither the treatment of 1610 β1OX cells with βadp1-R85D (Figure 3—figure supplement 1), nor the treatment 1610 parental cells with βadp1 (Figure 3D,E) affected the establishment of cell-cell contacts, as assessed by ECIS. βadp1 showed no evidence of toxic effects on 1610 β1OX cells, or the mouse ventricular myocyte cell line H9C2 even at concentrations twice that used elsewhere in this study (Figure 3—figure supplement 2), suggesting that the peptide’s action was unlikely to result from cellular toxicity.

Inhibition of β1 adhesion results in de-adhesion of β1-enriched perinexal nanodomains

LSCM and STORM results described above demonstrate that perinexal ID regions show high enrichment for β1. Perfusion of βadp1 into Langendorff-perfused GP hearts had profound yet highly selective effects on the ultrastructure of these regions (Figure 4). A confocal image of a section from a GP ventricle perfused with biotinylated βadp1 (βadp1-b; 100 µM) in Figure 4A shows punctate βadp1-b signal (red) adjacent to Cx43 immunosignal (green), consistent with the peptide localizing in perinexal regions of the ID. Representative images (Figure 4B) obtained by transmission electron microscopy (TEM) illustrate close apposition between perinexal membranes in control tyrode-perfused GP ventricles and markedly wider spacing between perinexal membranes in βadp1-treated hearts, reflecting that βadp1 treatment induced a de-adhesion of the β1-enriched membranes of the perinexus.

Figure 4 with 1 supplement see all

Download asset Open asset

Overall, quantification of TEM images indicated that βadp1 increased inter-membrane distance in dose-dependent fashion within perinexi, but importantly, not at other ID locations distal from GJs (Figure 4C). In contrast, a scrambled control peptide (βadp1-scr; 10 µM) did not significantly affect inter-membrane spacing at perinexal or non-perinexal ID sites. These results indicate that βadp1 selectively, and dose-dependently, widened β1-enriched perinexi – a dehiscence of these nanodomains that would be consistent with the specific antagonism of β1-mediated adhesion within the ID by the inhibitory peptide. To further investigate the effects of loss of β1 function on perinexal ultrastructure, we examined ventricles from Scn1b -/- (β1-null) mice. β1-null ventricles displayed striking evidence of perinexal de-adhesion, with > 4 fold greater perinexal inter-membrane distances compared to wild-type (WT) littermates, while inter-membrane distance at non-perinexal ID sites was not different between the two genotypes (Figure 4—figure supplement 1). These results are consistent with the effects of βadp1 in GP ventricles.

Inhibition of β1 adhesion selectively reduces GJ-associated sodium currents

In previous studies, we showed that acute interstitial edema increased perinexal membrane spacing at these Na_V1.5-enriched nanodomains and precipitated anisotropic conduction slowing (Veeraraghavan et al., 2015) – an effect on cardiac conduction that could only be explained by a computer model incorporating both ephaptic and electrotonic (i.e. GJ-based) coupling, and not by one incorporating electrotonic coupling alone. The level of perinexal de-adhesion caused by βadp1 was more marked than that prompted by experimentally induced edema. Thus, the aforementioned results suggested the potential of βadp1 as a tool for selectively probing the contribution of β1, and by extension, the hypothesized ephaptic mechanism, to action potential (AP) propagation between myocytes. As a first step to assess the effects of βadp1 on intercellular conduction of electrical impulse, we undertook voltage and current clamp studies on single myocytes isolated from GP hearts to determine whether the peptide affected AP and/or sodium channel function properties independent of whether cells where attached to one another. Neither βadp1, nor a scrambled control peptide (βadp1-scr), measurably altered the sodium current (I_Na; Figure 5A–D) or the duration or morphology of action potentials (APs; Figure 5E–F), suggesting that neither peptide affected the function of the Na⁺ channel complex, or other ion channels, in isolated, non-contacting cells.

Figure 5

Download asset Open asset

Next, we investigated whether selective targeting of the β1 adhesion domain affected Na⁺ channel activity in cells forming adherent contacts with each other. The rationale for our approach was as follows: The ephaptic coupling hypothesis implies that Na_V1.5 channels on one cell’s membrane can trans-activate Na_V1.5 channels on another cell membrane provided the two cells are separated by a sufficiently narrow extracellular cleft (Mori et al., 2008; Lin and Keener, 2014). Indeed, β1-mediated adhesion between cells may be critical to such a phenomenon by generating the close membrane apposition required for ephapse formation. Thus, modulating β1-mediated adhesion may alter the current from ID-localized Na_V1.5 in the presence of cell-to-cell contacts.

To test this hypothesis, we quantified I_Na from cell-to-cell contact sites using scanning ion conductance microscopy (SICM) –guided smart patch clamp (SPC) in Cx43-EGFP-expressing neonatal rat ventricular myocyte (NRVM) cultures (Figure 6). Confocal immunolabeling of NRVMs revealed close association of β1 signals with Cx43 GJs at cell-to-cell contact sites (Figure 6A, Figure 6—figure supplement 1), consistent with observations from GP ventricles (Figure 1, Figure 1—figure supplement 3). Under control conditions, local I_Na density at junctional sites proximal to Cx43-EGFP fluorescence, quantified by SPC, was significantly greater than that measured at non-junctional sites (Figure 6—figure supplement 2B), in line with previous reports of higher I_Na density at the ID relative to the lateral membrane (Lin et al., 2011). Treatment with βadp1, but not βadp1-scr, significantly reduced I_Na density at junctional sites (Figure 6B–E, Figure 6—figure supplement 2C,D). Currents at membrane sites distal from Cx43-GFP loci were not affected (Figure 6—figure supplement 2E). Importantly, and as was observed in isolated adult myocytes (Figure 5D), neither peptide altered whole cell I_Na density (Figure 6F). These observations indicated that whilst disruption of β1-mediated adhesion caused localized changes in Cx43-GJ-adjacent I_Na current density, global I_Na density across the entire cell remained unaffected – results consistent with β1 maintaining the high density of trans-activating sodium channels at the perinexus necessary for the hypothesized ephaptic conduction mechanism.

Figure 6 with 2 supplements see all

Download asset Open asset

Inhibition of β1 adhesion induces proarrhythmic conduction slowing

To probe whether loss of β1 adhesion, and the consequent perinexal dehiscence and reduction in GJ-associated I_Na, affected cardiac conduction, we assessed the electrophysiologic impact of βadp1 treatment in Langendorff-perfused GP hearts. Representative electrocardiogram (ECG) traces in Figure 7A of intrinsic activity demonstrate prolongation of the QRS complex and the QT interval in the presence of βadp1 (orange trace) compared to control (black trace). Overall, βadp1, but not βadp1-scr, significantly prolonged both the QRS complex (Figure 7B) and the QT interval (Figure 7C), indicating that βadp1 impaired conduction. The QRS prolongation elicited by βadp1 occurred in a dose-dependent manner (Figure 7D). Further, βadp1 prolonged both the QRS complex and the QT interval to similar extents, suggesting that the ST segment and therefore, AP duration, was not prolonged.

Figure 7 with 2 supplements see all

Download asset Open asset

Representative optical APs in Figure 7E demonstrate similar AP duration during control (black) and βadp1 (orange), consistent with ECG findings. Representative optical isochrone maps in Figure 7F show elliptical spread of activation from sites of unipolar epicardial pacing (white symbols). The maps also reveal crowding of isochrones, particularly along the transverse axis of propagation, in the presence of βadp1 (bottom) relative to control (top). This indicates preferential slowing of transverse conduction in the presence of βadp1. Summary plots in Figure 7G demonstrate that even low concentrations (10 µM) of βadp1 significantly slowed transverse conduction, thereby, increasing anisotropy, whereas βadp1-scr did not measurably affect conduction. In accordance with QRS prolongation, βadp1 reduced CV in a dose-dependent manner (Figure 7H): All doses tested (1–100 µM) preferentially decreased transverse conduction and increased anisotropy, while the higher concentrations (50 and 100 µM) also slowed longitudinal conduction. Importantly, the anisotropic conduction slowing induced by βadp1 proved proarrhythmic in a dose-dependent manner (Figure 7A, bottom trace; Figure 7—figure supplement 1): spontaneous ventricular tachycardias (VTs) were observed in the presence of βadp1 at doses (in µM) of 10 (1/3 hearts, p=0.2 vs. vehicle), 50 (2/3 hearts, p<0.05 vs. vehicle) and 100 (3/4 hearts, p<0.05 vs. vehicle). In contrast, no arrhythmias were observed in the presence of vehicle (DMSO; 0/3 hearts), or βadp1-scr (0/3 hearts).

In order to confirm that the impact of βadp1 reflected effects on AP conduction between cardiomyocytes, and was not due to effects on non-myocyte cells and tissues, we performed additional experiments assessing excitation spread in monolayers of human-induced pluripotent stem-cell-derived cardiomyocytes (iPSC-CMs). Immuno-confocal microscopy revealed enrichment of both Na_V1.5 and β1 adjacent to Cx43 at cell-to-cell contacts (Figure 7—figure supplement 2A,B), consistent with results in adult GP ventricles (Figures 1 and 2) and NRVMs (Figure 6A). In optical mapping experiments, treatment with βadp1, but not βadp1-scr, increased activation delay between equally spaced sites (Figure 7—figure supplement 2C) and slowed conduction in a dose-dependent manner (Figure 7—figure supplement 2D). Taken together with the results from GP ventricles, these data suggest that selective inhibition of β1-mediated adhesion slows conduction between cardiac myocytes in a cell-specific and pro-arrhythmic manner.

In this study, we provide evidence that sodium channel complexes concentrated at the edge of GJs (i.e. in the perinexus) could provide a structural basis for ephaptic conduction in the heart (Figure 8). Importantly, we find that the extracellular adhesion domain of the sodium channel β1 subunit may be critical for the generation of the close (<30 nm) cell-cell interactions necessary for this mechanism to operate. Selective inhibition of β1-mediated adhesion profoundly disrupts the structure of GJ-adjacent perinexi, causing them to dehisce/de-adhere, and swell beyond limits that would theoretically support ephaptic conduction, with attendant loss of perinexal sodium currents, anisotropic conduction slowing and arrhythmias. The clinical translational significance of perinexal dehiscence is further emphasized by our recent report that it is associated with the occurrence of atrial fibrillation in human patients (Raisch et al., 2018). Interestingly, the physical extent of the β1 extracellular domain predicted by our molecular homology model (Figure 3B) is 7–15 nm, depending on relaxation state of the domain. Thus, the span of two β1 extracellular domains on apposed cell membranes in trans homophilic interaction concurs with the EM-based measurements of average perinexal width (15–30 nm) in untreated hearts that we report here, and elsewhere (Veeraraghavan et al., 2015).

Figure 8

Download asset Open asset

Ephaptic coupling has been observed in neural (Anastassiou et al., 2011; Bokil et al., 2001; Chan et al., 1988; Faber and Korn, 1989; Jefferys, 1995; Su et al., 2012; Voronin, 2000), and other (Klaassen et al., 2012; Young, 2007) tissues, and has long been hypothesized to occur in the heart (Mori et al., 2008; Hichri et al., 2018; Kucera et al., 2002; Lin and Keener, 2013; Sperelakis, 2002). Early work on ephaptic coupling in the heart was motivated by the fact that GJs, which are abundant in mammalian hearts, are over two orders of magnitude less frequent in other chordates, especially birds (Martínez-Palomo and Mendez, 1971; Shibata and Yamamoto, 1979). Mathematical modeling suggests that this level of GJ coupling is unlikely to reliably support the beat-to-beat conduction of AP long-term, thus indicating that an alternate paradigm for AP conduction may operate in bird hearts. Nevertheless, the investigation of ephaptic conduction has remained largely theoretical and the hypothesis has drawn considerable skepticism over the years. The present study identifies a potential structural unit within the ID located at the edge of GJs – the ephapse – that may provide a mechanistic basis for ongoing studies of this long-theorized mechanism of cardiac AP conduction.

Subcellular distribution (and thus, nano-structural context) has emerged as an important modulator of Na_V1.5 function (Hichri et al., 2018; Petitprez et al., 2011; Lin et al., 2011; Chen-Izu et al., 2015; Eichel et al., 2016; Hund and Mohler, 2014; Makara et al., 2014): Na_V1.5 has been reported to be divided into lateral and ID-localized pools based on its sarcolemmal localization, differences in scaffolding partners, and current properties. Interestingly, Hichri and colleagues recently demonstrated that Na_V1.5 behavior is influenced by channel clustering, and localization facing narrow extracellular clefts (Hichri et al., 2018), raising the possibility of Na_V1.5 pools playing different, ultra-structure dependent roles. Here, using super-resolution STORM-RLA (Veeraraghavan and Gourdie, 2016), we further categorize ID-localized Na_V1.5 into a ‘perinexal pool’ adjacent to Cx43 and a ‘plicate pool’ that co-distributes with N-cad. The perinexal pool is in agreement with our demonstration of Na_V1.5 enrichment proximal to Cx43 GJs (Veeraraghavan et al., 2015; Veeraraghavan and Gourdie, 2016; Rhett et al., 2012; Veeraraghavan et al., 2016), while the plicate pool is consistent with the results of Leo-Macias and colleagues who reported Na_V1.5 co-distributing with N-cadherin-positive adherens junctions within the ID (Leo-Macias et al., 2016). Whilst adjacent cell membranes are very closely apposed (5–15 nm) within the perinexus (Veeraraghavan et al., 2015), they are further apart (63 nm) in the vicinity of adhesion complexes within plicate ID regions (Leo-Macías et al., 2015). Given that theoretical studies indicate that membrane spacing of ≤30 nm would be necessary to support ephaptic coupling (Mori et al., 2008; Greer-Short et al., 2017), the plicate Na_V1.5 pool is unlikely to play a role in ephaptic coupling. Nonetheless, whilst closely approximated, trans-activating sodium channels at GJ perinexi may act to ‘trigger’ a local depolarization of interplicate membrane, the function of the nearby pools of plicate Na_V1.5 channels could be to provide cis-activated excitatory current, contributing to the spread of depolarization through the downstream myocyte.

The approach we took to developing a competitive inhibitor of β1 was based on a methodology successfully applied by independent groups to two other Ig domain adhesion molecules that share homologies with β1: N-cad (Williams et al., 2000a), and desmoglein (Schlipp et al., 2014). This strategy involved generating a peptide mimetic of the extracellular adhesion sequence of the β1 Ig domain. In ECIS experiments, β1 overexpression conferred strong adhesion to 1610 cells, which was selectively and dose-dependently inhibited by βadp1 treatment. When perfused into GP ventricles, βadp1 prompted dose-dependent increases in inter-membrane distance, selectively within perinexi. Similar effects were not observed in hearts treated with a scrambled control peptide, βadp1-scr. The conspicuous de-adhesion of perinexal membranes prompted by βadp1 in GP hearts, and by germline loss of Scn1b/β1 in mice was selective to perinexal nanodomains, with inter-membrane spacing at other ID locations (e.g. within adherens junctions and desmosomes) remaining unaffected in both GPs and mice. These results, along with the demonstration that βadp1 treatment had no effects on whole-cell I_Na, but results in localized loss of I_Na adjacent to Cx43-GFP GJs, attest to the selectivity of the novel inhibitor of β1-mediated adhesion. Further, these findings suggest that the increases in extracellular cleft width adjacent to Cx43 GJ induced by inhibition of β1-mediated adhesion likely compromises trans-activation of Na⁺ channels within these nanodomains. Ongoing studies would usefully determine the basis of this localized down-regulation of channel activity. Crucially, the fact that βadp1 treatment of myocyte monolayers elicits these localized changes in I_Na within GJ-adjacent nanodomains, without alteration of whole-cell I_Na, indicates the possibility of local remodeling of Na⁺ channel complexes from perinexal nanodomains.

Examining the functional impact of βadp1 treatment on Langendorff-perfused GP ventricles, we observed prolongation of the QRS complex, pointing to conduction slowing. The QT interval was also prolonged by a similar degree as the QRS complex, indicating no change in ST segment duration. This is consistent with the absence of APD changes following βadp1 treatment in both isolated myocytes and in intact ventricles. The contrast between these results, and those reported in mice lacking β1 (β1-null), where QT prolongation was associated with prolonged AP duration (Lopez-Santiago et al., 2007), likely stem from the differences between acutely and selectively inhibiting β1-mediated adhesion (as in the present study), versus chronic genetic ablation of β1 expression. As yet, conduction velocity measurements have not been reported from the ventricles of β1-nulls, which in light of the data reported herein would be a useful addition to our understanding of cardiac conduction mechanisms.

In keeping with the observed QRS prolongation, βadp1 treatment dose-dependently slowed transverse conduction in intact ventricles, increasing the anisotropy of conduction, and precipitating spontaneous arrhythmias. We have previously demonstrated in silico that anisotropic conduction slowing secondary to a reduction of I_Na at the ID results from compromised ephaptic coupling (Veeraraghavan et al., 2015): Briefly, a wavefront traveling transverse to the fiber axis encounters more cell-cell junctions per unit distance than one traveling longitudinally. Therefore, compromised cell-cell coupling via perinexal nanodomains would have a greater impact on transverse conduction than on longitudinal conduction. Thus, the anisotropic nature of conduction slowing caused by βadp1 is consistent with disruption of ephaptic communication (Plonsey and Barr, 2007), and identifies a hitherto unanticipated role for β1-mediated adhesion in the heart. Taken together, these data lend further support to the hypothesis that β1 adhesion mediates close apposition between Na_V1.5-rich membranes within the perinexus, facilitating ephaptic coupling between cardiac myocytes (Figure 8). In this context, the fact that βadp1 mimics a sequence in β1 (i.e. its Ig domain adhesion sequence) that is a ‘hot spot’ for disease-causing mutations, including mutations associated with atrial arrhythmia (R85H) (Watanabe et al., 2009), conduction disease (R85H) (Watanabe et al., 2009), epilepsy (R85H, R85C) (Scheffer et al., 2007), Brugada syndrome (E87Q) (Hasdemir et al., 2015), and febrile epilepsy (I70_E74del) (Audenaert et al., 2003) lends further support to the importance of β1-mediated adhesion to electrical propagation in the heart, and beyond. Interestingly, a single amino acid substitution of R to D at the R85 ‘hot spot’ within βadp1 significantly abrogated the cell adhesion-inhibitory effects of the peptide. These data suggest that disease-causing mutations associated with this locus may mediate their effects via perinexal disruption.

In summary, we provide experimental evidence suggesting that ephaptic conduction may play a role in the heart, identifying hitherto unanticipated assignments for both Na_V1.5 and β1 in this phenomenon. Although limitations of current technology prevent direct interrogation of electrophysiology at the nanoscale, which would provide direct proof of the ephaptic mechanism, our results strongly suggest a role for it in normal cardiac physiology. Thus, it has not escaped our notice that modulation of β1-mediated adhesion could represent a novel clinical target for the amelioration of cardiac arrhythmias. Increased spacing of perinexal membranes, whether resulting from inhibition of β1-mediated adhesion, or acute interstitial edema, is strongly associated with ventricular arrhythmias. Given that humans with atrial fibrillation also demonstrate perinexal dehiscence (Raisch et al., 2018), these data suggest that drugs that act to stabilize and/or enhance β1-mediated adhesion within IDs could have anti-arrhythmic potential.

The raw data generated in this study are available via Dryad (doi:10.5061/dryad.10351qn). The raw STORM movies are available on request from the corresponding author due to their large size.

1. 1. Veeraraghavan R
   2. Hoeker G
   3. Alvarez-Laviada A
   4. Hoagland D
   5. Wan X
   6. King D
   7. Sanchez-Alonso J
   8. Chen C
   9. Jourdan J
   10. Isom L
   11. Deschenes I
   12. Smyth J
   13. Gorelik J
   14. Poelzing S
   15. Gourdie R

   (2018) Data from: The Adhesion Function of the Sodium Channel Beta Subunit (β1) Contributes to Cardiac Action Potential Propagation

   Available at Dryad Digital Repository under a CC0 Public Domain Dedication.

   https://dx.doi.org/10.5061/dryad.10351qn

1. 1. Anastassiou CA
   2. Perin R
   3. Markram H
   4. Koch C

   (2011) Ephaptic coupling of cortical neurons

   Nature Neuroscience 14:217–223.

   https://doi.org/10.1038/nn.2727

   • PubMed
   • Google Scholar
2. 1. Audenaert D
   2. Claes L
   3. Ceulemans B
   4. Löfgren A
   5. Van Broeckhoven C
   6. De Jonghe P

   (2003) A deletion in SCN1B is associated with febrile seizures and early-onset absence epilepsy

   Neurology 61:854–856.

   https://doi.org/10.1212/01.WNL.0000080362.55784.1C

   • PubMed
   • Google Scholar
3. 1. Bayly PV
   2. KenKnight BH
   3. Rogers JM
   4. Hillsley RE
   5. Ideker RE
   6. Smith WM

   (1998) Estimation of conduction velocity vector fields from epicardial mapping data

   IEEE Transactions on Biomedical Engineering 45:563–571.

   https://doi.org/10.1109/10.668746

   • PubMed
   • Google Scholar
4. 1. Bhargava A
   2. Lin X
   3. Novak P
   4. Mehta K
   5. Korchev Y
   6. Delmar M
   7. Gorelik J

   (2013) Super-resolution scanning patch clamp reveals clustering of functional ion channels in adult ventricular myocyte

   Circulation Research 112:1112–1120.

   https://doi.org/10.1161/CIRCRESAHA.111.300445

   • PubMed
   • Google Scholar
5. 1. Bokil H
   2. Laaris N
   3. Blinder K
   4. Ennis M
   5. Keller A

   (2001) Ephaptic interactions in the mammalian olfactory system

   The Journal of Neuroscience 21:RC173.

   https://doi.org/10.1523/JNEUROSCI.21-20-j0004.2001

   • PubMed
   • Google Scholar
6. 1. Brackenbury WJ
   2. Isom LL

   (2011) Na channel β subunits: overachievers of the ion channel family

   Frontiers in Pharmacology 2:53.

   https://doi.org/10.3389/fphar.2011.00053

   • PubMed
   • Google Scholar
7. 1. Calhoun JD
   2. Isom LL

   (2014) The role of non-pore-forming β subunits in physiology and pathophysiology of voltage-gated sodium channels

   Handbook of Experimental Pharmacology 221:51–89.

   https://doi.org/10.1007/978-3-642-41588-3_4

   • PubMed
   • Google Scholar
8. 1. Chan CY
   2. Hounsgaard J
   3. Nicholson C

   (1988) Effects of electric fields on transmembrane potential and excitability of turtle cerebellar purkinje cells in vitro

   The Journal of Physiology 402:751–771.

   https://doi.org/10.1113/jphysiol.1988.sp017232

   • PubMed
   • Google Scholar
9. 1. Chen-Izu Y
   2. Shaw RM
   3. Pitt GS
   4. Yarov-Yarovoy V
   5. Sack JT
   6. Abriel H
   7. Aldrich RW
   8. Belardinelli L
   9. Cannell MB
   10. Catterall WA
   11. Chazin WJ
   12. Chiamvimonvat N
   13. Deschenes I
   14. Grandi E
   15. Hund TJ
   16. Izu LT
   17. Maier LS
   18. Maltsev VA
   19. Marionneau C
   20. Mohler PJ
   21. Rajamani S
   22. Rasmusson RL
   23. Sobie EA
   24. Clancy CE
   25. Bers DM

   (2015) Na⁺ channel function, regulation, structure, trafficking and sequestration

   The Journal of Physiology 593:1347–1360.

   https://doi.org/10.1113/jphysiol.2014.281428

   • PubMed
   • Google Scholar
10. Book

    1. Churchman LS
    2. Spudich JA

    (2013) Single-Molecule High-Resolution Colocalization (SHREC)

    In: Roberts G. C. K, editors. Encyclopedia of Biophysics. Berlin Heidelberg: Springer. pp. 2336–2337.

    https://doi.org/10.1007/978-3-642-16712-6_515

    • Google Scholar
11. 1. De Vuyst E
    2. Boengler K
    3. Antoons G
    4. Sipido KR
    5. Schulz R
    6. Leybaert L

    (2011) Pharmacological modulation of connexin-formed channels in cardiac pathophysiology

    British Journal of Pharmacology 163:469–483.

    https://doi.org/10.1111/j.1476-5381.2011.01244.x

    • PubMed
    • Google Scholar
12. 1. Deschout H
    2. Cella Zanacchi F
    3. Mlodzianoski M
    4. Diaspro A
    5. Bewersdorf J
    6. Hess ST
    7. Braeckmans K

    (2014) Precisely and accurately localizing single emitters in fluorescence microscopy

    Nature Methods 11:253–266.

    https://doi.org/10.1038/nmeth.2843

    • PubMed
    • Google Scholar
13. 1. Eichel CA
    2. Beuriot A
    3. Chevalier MY
    4. Rougier JS
    5. Louault F
    6. Dilanian G
    7. Amour J
    8. Coulombe A
    9. Abriel H
    10. Hatem SN
    11. Balse E

    (2016) Lateral membrane-specific MAGUK CASK down-regulates NaV1.5 Channel in cardiac myocytes

    Circulation Research 119:544–556.

    https://doi.org/10.1161/CIRCRESAHA.116.309254

    • PubMed
    • Google Scholar
14. 1. Faber DS
    2. Korn H

    (1989) Electrical field effects: their relevance in central neural networks

    Physiological Reviews 69:821–863.

    https://doi.org/10.1152/physrev.1989.69.3.821

    • PubMed
    • Google Scholar
15. 1. Girouard SD
    2. Laurita KR
    3. Rosenbaum DS

    (1996a) Unique properties of cardiac action potentials recorded with voltage-sensitive dyes

    Journal of Cardiovascular Electrophysiology 7:1024–1038.

    https://doi.org/10.1111/j.1540-8167.1996.tb00478.x

    • PubMed
    • Google Scholar
16. 1. Girouard SD
    2. Pastore JM
    3. Laurita KR
    4. Gregory KW
    5. Rosenbaum DS

    (1996b) Optical mapping in a New Guinea pig model of ventricular tachycardia reveals mechanisms for multiple wavelengths in a single reentrant circuit

    Circulation 93:603–613.

    https://doi.org/10.1161/01.CIR.93.3.603

    • PubMed
    • Google Scholar
17. 1. Gorelik J
    2. Shevchuk A
    3. Ramalho M
    4. Elliott M
    5. Lei C
    6. Higgins CF
    7. Lab MJ
    8. Klenerman D
    9. Krauzewicz N
    10. Korchev Y

    (2002) Scanning surface confocal microscopy for simultaneous topographical and fluorescence imaging: application to single virus-like particle entry into a cell

    PNAS 99:16018–16023.

    https://doi.org/10.1073/pnas.252458399

    • PubMed
    • Google Scholar
18. 1. Greer-Short A
    2. George SA
    3. Poelzing S
    4. Weinberg SH

    (2017) Revealing the concealed nature of Long-QT type 3 syndrome

    Circulation: Arrhythmia and Electrophysiology 10:e004400.

    https://doi.org/10.1161/CIRCEP.116.004400

    • PubMed
    • Google Scholar
19. 1. Gutstein DE
    2. Morley GE
    3. Tamaddon H
    4. Vaidya D
    5. Schneider MD
    6. Chen J
    7. Chien KR
    8. Stuhlmann H
    9. Fishman GI

    (2001) Conduction slowing and sudden arrhythmic death in mice with cardiac-restricted inactivation of connexin43

    Circulation Research 88:333–339.

    https://doi.org/10.1161/01.RES.88.3.333

    • PubMed
    • Google Scholar
20. 1. Hama H
    2. Kurokawa H
    3. Kawano H
    4. Ando R
    5. Shimogori T
    6. Noda H
    7. Fukami K
    8. Sakaue-Sawano A
    9. Miyawaki A

    (2011) Scale: a chemical approach for fluorescence imaging and reconstruction of transparent mouse brain

    Nature Neuroscience 14:1481–1488.

    https://doi.org/10.1038/nn.2928

    • PubMed
    • Google Scholar
21. 1. Harrison OJ
    2. Brasch J
    3. Lasso G
    4. Katsamba PS
    5. Ahlsen G
    6. Honig B
    7. Shapiro L

    (2016) Structural basis of adhesive binding by desmocollins and desmogleins

    PNAS 113:7160–7165.

    https://doi.org/10.1073/pnas.1606272113

    • PubMed
    • Google Scholar
22. 1. Hasdemir C
    2. Payzin S
    3. Kocabas U
    4. Sahin H
    5. Yildirim N
    6. Alp A
    7. Aydin M
    8. Pfeiffer R
    9. Burashnikov E
    10. Wu Y
    11. Antzelevitch C

    (2015) High prevalence of concealed brugada syndrome in patients with atrioventricular nodal reentrant tachycardia

    Heart Rhythm 12:1584–1594.

    https://doi.org/10.1016/j.hrthm.2015.03.015

    • PubMed
    • Google Scholar
23. 1. Hichri E
    2. Abriel H
    3. Kucera JP

    (2018) Distribution of cardiac sodium channels in clusters potentiates ephaptic interactions in the intercalated disc

    The Journal of Physiology 596:563–589.

    https://doi.org/10.1113/JP275351

    • PubMed
    • Google Scholar
24. 1. Hoshi M
    2. Du XX
    3. Shinlapawittayatorn K
    4. Liu H
    5. Chai S
    6. Wan X
    7. Ficker E
    8. Deschênes I

    (2014) Brugada syndrome disease phenotype explained in apparently benign sodium channel mutations

    Circulation: Cardiovascular Genetics 7:123–131.

    https://doi.org/10.1161/CIRCGENETICS.113.000292

    • PubMed
    • Google Scholar
25. 1. Hund TJ
    2. Mohler PJ

    (2014) Nav channel complex heterogeneity: new targets for the treatment of arrhythmia?

    Circulation 130:132–134.

    https://doi.org/10.1161/CIRCULATIONAHA.114.010867

    • PubMed
    • Google Scholar
26. 1. Isom LL
    2. Scheuer T
    3. Brownstein AB
    4. Ragsdale DS
    5. Murphy BJ
    6. Catterall WA

    (1995) Functional co-expression of the beta 1 and type IIA alpha subunits of sodium channels in a mammalian cell line

    Journal of Biological Chemistry 270:3306–3312.

    https://doi.org/10.1074/jbc.270.7.3306

    • PubMed
    • Google Scholar
27. 1. Isom LL

    (2002) The role of sodium channels in cell adhesion

    Frontiers in Bioscience 7:d12–23.

    https://doi.org/10.2741/A765

    • Google Scholar
28. 1. Jefferys JG

    (1995) Nonsynaptic modulation of neuronal activity in the brain: electric currents and extracellular ions

    Physiological Reviews 75:689–723.

    https://doi.org/10.1152/physrev.1995.75.4.689

    • PubMed
    • Google Scholar
29. 1. Jongsma HJ
    2. Wilders R

    (2000) Gap junctions in cardiovascular disease

    Circulation Research 86:1193–1197.

    https://doi.org/10.1161/01.RES.86.12.1193

    • PubMed
    • Google Scholar
30. 1. Juette MF
    2. Gould TJ
    3. Lessard MD
    4. Mlodzianoski MJ
    5. Nagpure BS
    6. Bennett BT
    7. Hess ST
    8. Bewersdorf J

    (2008) Three-dimensional sub-100 nm resolution fluorescence microscopy of thick samples

    Nature Methods 5:527–529.

    https://doi.org/10.1038/nmeth.1211

    • PubMed
    • Google Scholar
31. 1. Klaassen LJ
    2. Fahrenfort I
    3. Kamermans M

    (2012) Connexin hemichannel mediated ephaptic inhibition in the retina

    Brain Research 1487:25–38.

    https://doi.org/10.1016/j.brainres.2012.04.059

    • PubMed
    • Google Scholar
32. 1. Kléber AG
    2. Rudy Y

    (2004) Basic mechanisms of cardiac impulse propagation and associated arrhythmias

    Physiological Reviews 84:431–488.

    https://doi.org/10.1152/physrev.00025.2003

    • PubMed
    • Google Scholar
33. 1. Kucera JP
    2. Rohr S
    3. Rudy Y

    (2002) Localization of sodium channels in intercalated disks modulates cardiac conduction

    Circulation Research 91:1176–1182.

    https://doi.org/10.1161/01.RES.0000046237.54156.0A

    • PubMed
    • Google Scholar
34. 1. Leo-Macías A
    2. Liang FX
    3. Delmar M

    (2015) Ultrastructure of the intercellular space in adult murine ventricle revealed by quantitative tomographic electron microscopy

    Cardiovascular Research 107:442–452.

    https://doi.org/10.1093/cvr/cvv182

    • PubMed
    • Google Scholar
35. 1. Leo-Macias A
    2. Agullo-Pascual E
    3. Sanchez-Alonso JL
    4. Keegan S
    5. Lin X
    6. Arcos T
    7. Feng-Xia-Liang
    8. Korchev YE
    9. Gorelik J
    10. Fenyö D
    11. Rothenberg E
    12. Rothenberg E
    13. Delmar M

    (2016) Nanoscale visualization of functional adhesion/excitability nodes at the intercalated disc

    Nature Communications 7:10342.

    https://doi.org/10.1038/ncomms10342

    • PubMed
    • Google Scholar
36. 1. Lin J
    2. Keener JP

    (2010) Modeling electrical activity of myocardial cells incorporating the effects of ephaptic coupling

    PNAS 107:20935–20940.

    https://doi.org/10.1073/pnas.1010154107

    • PubMed
    • Google Scholar
37. 1. Lin J
    2. Keener JP

    (2013) Ephaptic coupling in cardiac myocytes

    IEEE Transactions on Biomedical Engineering 60:576–582.

    https://doi.org/10.1109/TBME.2012.2226720

    • PubMed
    • Google Scholar
38. 1. Lin J
    2. Keener JP

    (2014) Microdomain effects on transverse cardiac propagation

    Biophysical Journal 106:925–931.

    https://doi.org/10.1016/j.bpj.2013.11.1117

    • PubMed
    • Google Scholar
39. 1. Lin X
    2. Liu N
    3. Lu J
    4. Zhang J
    5. Anumonwo JM
    6. Isom LL
    7. Fishman GI
    8. Delmar M

    (2011) Subcellular heterogeneity of sodium current properties in adult cardiac ventricular myocytes

    Heart Rhythm 8:1923–1930.

    https://doi.org/10.1016/j.hrthm.2011.07.016

    • PubMed
    • Google Scholar
40. 1. Lopez-Santiago LF
    2. Meadows LS
    3. Ernst SJ
    4. Chen C
    5. Malhotra JD
    6. McEwen DP
    7. Speelman A
    8. Noebels JL
    9. Maier SK
    10. Lopatin AN
    11. Isom LL

    (2007) Sodium channel Scn1b null mice exhibit prolonged QT and RR intervals

    Journal of Molecular and Cellular Cardiology 43:636–647.

    https://doi.org/10.1016/j.yjmcc.2007.07.062

    • PubMed
    • Google Scholar
41. 1. Maier SK
    2. Westenbroek RE
    3. McCormick KA
    4. Curtis R
    5. Scheuer T
    6. Catterall WA

    (2004) Distinct subcellular localization of different sodium channel alpha and beta subunits in single ventricular myocytes from mouse heart

    Circulation 109:1421–1427.

    https://doi.org/10.1161/01.CIR.0000121421.61896.24

    • PubMed
    • Google Scholar
42. 1. Makara MA
    2. Curran J
    3. Little SC
    4. Musa H
    5. Polina I
    6. Smith SA
    7. Wright PJ
    8. Unudurthi SD
    9. Snyder J
    10. Bennett V
    11. Hund TJ
    12. Mohler PJ

    (2014) Ankyrin-G coordinates intercalated disc signaling platform to regulate cardiac excitability in vivo

    Circulation Research 115:929–938.

    https://doi.org/10.1161/CIRCRESAHA.115.305154

    • PubMed
    • Google Scholar
43. 1. Malhotra JD
    2. Kazen-Gillespie K
    3. Hortsch M
    4. Isom LL

    (2000) Sodium channel beta subunits mediate homophilic cell adhesion and recruit ankyrin to points of cell-cell contact

    Journal of Biological Chemistry 275:11383–11388.

    https://doi.org/10.1074/jbc.275.15.11383

    • PubMed
    • Google Scholar
44. 1. Mann JE
    2. Sperelakis N

    (1979) Further development of a model for electrical transmission between myocardial cells not connected by low-resistance pathways

    Journal of Electrocardiology 12:23–33.

    https://doi.org/10.1016/S0022-0736(79)80041-2

    • PubMed
    • Google Scholar
45. 1. Martínez-Palomo A
    2. Mendez R

    (1971) Presence of gap junctions between cardiac cells in the heart of nonmammalian species

    Journal of Ultrastructure Research 37:592–600.

    https://doi.org/10.1016/S0022-5320(71)80027-8

    • PubMed
    • Google Scholar
46. 1. Mlodzianoski MJ
    2. Juette MF
    3. Beane GL
    4. Bewersdorf J

    (2009) Experimental characterization of 3D localization techniques for particle-tracking and super-resolution microscopy

    Optics Express 17:8264–8277.

    https://doi.org/10.1364/OE.17.008264

    • PubMed
    • Google Scholar
47. 1. Mori Y
    2. Fishman GI
    3. Peskin CS

    (2008) Ephaptic conduction in a cardiac strand model with 3D electrodiffusion

    PNAS 105:6463–6468.

    https://doi.org/10.1073/pnas.0801089105

    • PubMed
    • Google Scholar
48. 1. Moy AB
    2. Winter M
    3. Kamath A
    4. Blackwell K
    5. Reyes G
    6. Giaever I
    7. Keese C
    8. Shasby DM

    (2000) Histamine alters endothelial barrier function at cell-cell and cell-matrix sites

    American Journal of Physiology-Lung Cellular and Molecular Physiology 278:L888–L898.

    https://doi.org/10.1152/ajplung.2000.278.5.L888

    • PubMed
    • Google Scholar
49. 1. Namadurai S
    2. Balasuriya D
    3. Rajappa R
    4. Wiemhöfer M
    5. Stott K
    6. Klingauf J
    7. Edwardson JM
    8. Chirgadze DY
    9. Jackson AP

    (2014) Crystal structure and molecular imaging of the nav channel β3 subunit indicates a trimeric assembly

    Journal of Biological Chemistry 289:10797–10811.

    https://doi.org/10.1074/jbc.M113.527994

    • PubMed
    • Google Scholar
50. 1. Namadurai S
    2. Yereddi NR
    3. Cusdin FS
    4. Huang CL
    5. Chirgadze DY
    6. Jackson AP

    (2015) A new look at sodium channel β subunits

    Open Biology 5:140192.

    https://doi.org/10.1098/rsob.140192

    • PubMed
    • Google Scholar
51. 1. Nassal DM
    2. Wan X
    3. Liu H
    4. Deschênes I

    (2016) Myocardial KChIP2 expression in guinea pig resolves an expanded electrophysiologic role

    PLoS One 11:e0146561.

    https://doi.org/10.1371/journal.pone.0146561

    • PubMed
    • Google Scholar
52. 1. O'Malley HA
    2. Isom LL

    (2015) Sodium channel β subunits: emerging targets in channelopathies

    Annual Review of Physiology 77:481–504.

    https://doi.org/10.1146/annurev-physiol-021014-071846

    • PubMed
    • Google Scholar
53. 1. Palatinus JA
    2. Rhett JM
    3. Gourdie RG

    (2012) The connexin43 carboxyl terminus and cardiac gap junction organization

    Biochimica Et Biophysica Acta (BBA) - Biomembranes 1818:1831–1843.

    https://doi.org/10.1016/j.bbamem.2011.08.006

    • PubMed
    • Google Scholar
54. 1. Patino GA
    2. Claes LR
    3. Lopez-Santiago LF
    4. Slat EA
    5. Dondeti RS
    6. Chen C
    7. O'Malley HA
    8. Gray CB
    9. Miyazaki H
    10. Nukina N
    11. Oyama F
    12. De Jonghe P
    13. Isom LL

    (2009) A functional null mutation of SCN1B in a patient with dravet syndrome

    Journal of Neuroscience 29:10764–10778.

    https://doi.org/10.1523/JNEUROSCI.2475-09.2009

    • PubMed
    • Google Scholar
55. 1. Pertsov AM
    2. Medvinskiĭ AB

    (1979)

    Effect of specialized contacts on fiber interaction during the spread of excitation in smooth muscle and myocardial tissues]

    Biofizika 24:293–298.

    • PubMed
    • Google Scholar
56. 1. Petitprez S
    2. Zmoos AF
    3. Ogrodnik J
    4. Balse E
    5. Raad N
    6. El-Haou S
    7. Albesa M
    8. Bittihn P
    9. Luther S
    10. Lehnart SE
    11. Hatem SN
    12. Coulombe A
    13. Abriel H

    (2011) SAP97 and dystrophin macromolecular complexes determine two pools of cardiac sodium channels Nav1.5 in cardiomyocytes

    Circulation Research 108:294–304.

    https://doi.org/10.1161/CIRCRESAHA.110.228312

    • PubMed
    • Google Scholar
57. Book

    1. Plonsey R
    2. Barr RC

    (2007)

    Bioelectricity: A Quantitative Approach

    Springer.

    • Google Scholar
58. 1. Poelzing S
    2. Veeraraghavan R

    (2007) Heterogeneous ventricular chamber response to hypokalemia and inward rectifier potassium channel blockade underlies bifurcated T wave in guinea pig

    American Journal of Physiology-Heart and Circulatory Physiology 292:H3043–H3051.

    https://doi.org/10.1152/ajpheart.01312.2006

    • PubMed
    • Google Scholar
59. 1. Raisch TB
    2. Yanoff MS
    3. Larsen TR
    4. Farooqui MA
    5. King DR
    6. Veeraraghavan R
    7. Gourdie RG
    8. Baker JW
    9. Arnold WS
    10. AlMahameed ST
    11. Poelzing S

    (2018) Intercalated disk extracellular nanodomain expansion in patients with atrial fibrillation

    Frontiers in Physiology 9:398.

    https://doi.org/10.3389/fphys.2018.00398

    • PubMed
    • Google Scholar
60. 1. Rhett JM
    2. Jourdan J
    3. Gourdie RG

    (2011) Connexin 43 connexon to gap junction transition is regulated by zonula occludens-1

    Molecular Biology of the Cell 22:1516–1528.

    https://doi.org/10.1091/mbc.e10-06-0548

    • PubMed
    • Google Scholar
61. 1. Rhett JM
    2. Ongstad EL
    3. Jourdan J
    4. Gourdie RG

    (2012) Cx43 associates with na(v)1.5 in the cardiomyocyte perinexus

    The Journal of Membrane Biology 245:411–422.

    https://doi.org/10.1007/s00232-012-9465-z

    • PubMed
    • Google Scholar
62. 1. Scheffer IE
    2. Harkin LA
    3. Grinton BE
    4. Dibbens LM
    5. Turner SJ
    6. Zielinski MA
    7. Xu R
    8. Jackson G
    9. Adams J
    10. Connellan M
    11. Petrou S
    12. Wellard RM
    13. Briellmann RS
    14. Wallace RH
    15. Mulley JC
    16. Berkovic SF

    (2007) Temporal lobe epilepsy and GEFS+ phenotypes associated with SCN1B mutations

    Brain 130:100–109.

    https://doi.org/10.1093/brain/awl272

    • PubMed
    • Google Scholar
63. 1. Schlipp A
    2. Schinner C
    3. Spindler V
    4. Vielmuth F
    5. Gehmlich K
    6. Syrris P
    7. Mckenna WJ
    8. Dendorfer A
    9. Hartlieb E
    10. Waschke J

    (2014) Desmoglein-2 interaction is crucial for cardiomyocyte cohesion and function

    Cardiovascular Research 104:245–257.

    https://doi.org/10.1093/cvr/cvu206

    • PubMed
    • Google Scholar
64. 1. Severs NJ

    (2000) The cardiac muscle cell

    BioEssays 22:188–199.

    https://doi.org/10.1002/(SICI)1521-1878(200002)22:2<188::AID-BIES10>3.0.CO;2-T

    • PubMed
    • Google Scholar
65. 1. Shibata Y
    2. Yamamoto T

    (1979) Freeze-fractrue studies of gap junctions in vertebrate cardiac muscle cells

    Journal of Ultrastructure Research 67:79–88.

    https://doi.org/10.1016/S0022-5320(79)80020-9

    • PubMed
    • Google Scholar
66. 1. Shibayama J
    2. Paznekas W
    3. Seki A
    4. Taffet S
    5. Jabs EW
    6. Delmar M
    7. Musa H

    (2005) Functional characterization of connexin43 mutations found in patients with oculodentodigital dysplasia

    Circulation Research 96:e83–e91.

    https://doi.org/10.1161/01.RES.0000168369.79972.d2

    • PubMed
    • Google Scholar
67. 1. Sperelakis N

    (2002) An electric field mechanism for transmission of excitation between myocardial cells

    Circulation Research 91:985–987.

    https://doi.org/10.1161/01.RES.0000045656.34731.6D

    • PubMed
    • Google Scholar
68. 1. Stroemlund LW
    2. Jensen CF
    3. Qvortrup K
    4. Delmar M
    5. Nielsen MS

    (2015) Gap junctions - guards of excitability

    Biochemical Society Transactions 43:508–512.

    https://doi.org/10.1042/BST20150059

    • PubMed
    • Google Scholar
69. 1. Su CY
    2. Menuz K
    3. Reisert J
    4. Carlson JR

    (2012) Non-synaptic inhibition between grouped neurons in an olfactory circuit

    Nature 492:66–71.

    https://doi.org/10.1038/nature11712

    • PubMed
    • Google Scholar
70. 1. Thomas EA
    2. Xu R
    3. Petrou S

    (2007) Computational analysis of the R85C and R85H epilepsy mutations in na+ channel beta1 subunits

    Neuroscience 147:1034–1046.

    https://doi.org/10.1016/j.neuroscience.2007.05.010

    • PubMed
    • Google Scholar
71. 1. Tiruppathi C
    2. Malik AB
    3. Del Vecchio PJ
    4. Keese CR
    5. Giaever I

    (1992) Electrical method for detection of endothelial cell shape change in real time: assessment of endothelial barrier function

    PNAS 89:7919–7923.

    https://doi.org/10.1073/pnas.89.17.7919

    • PubMed
    • Google Scholar
72. 1. Veeraraghavan R
    2. Gourdie RG

    (2016) Stochastic optical reconstruction microscopy-based relative localization analysis (STORM-RLA) for quantitative nanoscale assessment of spatial protein organization

    Molecular Biology of the Cell 27:3583–3590.

    https://doi.org/10.1091/mbc.e16-02-0125

    • PubMed
    • Google Scholar
73. 1. Veeraraghavan R
    2. Lin J
    3. Hoeker GS
    4. Keener JP
    5. Gourdie RG
    6. Poelzing S

    (2015) Sodium channels in the Cx43 gap junction perinexus may constitute a cardiac ephapse: an experimental and modeling study

    Pflügers Archiv - European Journal of Physiology 467:2093–2105.

    https://doi.org/10.1007/s00424-014-1675-z

    • Google Scholar
74. 1. Veeraraghavan R
    2. Lin J
    3. Keener JP
    4. Gourdie R
    5. Poelzing S

    (2016) Potassium channels in the Cx43 gap junction perinexus modulate ephaptic coupling: an experimental and modeling study

    Pflügers Archiv - European Journal of Physiology 468:1651–1661.

    https://doi.org/10.1007/s00424-016-1861-2

    • Google Scholar
75. 1. Veeraraghavan R
    2. Poelzing S

    (2008) Mechanisms underlying increased right ventricular conduction sensitivity to flecainide challenge

    Cardiovascular Research 77:749–756.

    https://doi.org/10.1093/cvr/cvm090

    • PubMed
    • Google Scholar
76. 1. Vite A
    2. Radice GL

    (2014) N-cadherin/catenin complex as a master regulator of intercalated disc function

    Cell Communication & Adhesion 21:169–179.

    https://doi.org/10.3109/15419061.2014.908853

    • PubMed
    • Google Scholar
77. 1. Voronin LL

    (2000) Intrasynaptic ephaptic feedback in central synapses

    Neuroscience and Behavioral Physiology 30:575–585.

    https://doi.org/10.1007/BF02462618

    • PubMed
    • Google Scholar
78. 1. Wan X
    2. Laurita KR
    3. Pruvot EJ
    4. Rosenbaum DS

    (2005) Molecular correlates of repolarization alternans in cardiac myocytes

    Journal of Molecular and Cellular Cardiology 39:419–428.

    https://doi.org/10.1016/j.yjmcc.2005.06.004

    • PubMed
    • Google Scholar
79. 1. Watanabe H
    2. Darbar D
    3. Kaiser DW
    4. Jiramongkolchai K
    5. Chopra S
    6. Donahue BS
    7. Kannankeril PJ
    8. Roden DM

    (2009) Mutations in sodium channel β1- and β2-subunits associated with atrial fibrillation

    Circulation: Arrhythmia and Electrophysiology 2:268–275.

    https://doi.org/10.1161/CIRCEP.108.779181

    • PubMed
    • Google Scholar
80. 1. Williams E
    2. Williams G
    3. Gour BJ
    4. Blaschuk OW
    5. Doherty P

    (2000a) A novel family of cyclic peptide antagonists suggests that N-cadherin specificity is determined by amino acids that flank the HAV motif

    Journal of Biological Chemistry 275:4007–4012.

    https://doi.org/10.1074/jbc.275.6.4007

    • PubMed
    • Google Scholar
81. 1. Williams EJ
    2. Williams G
    3. Gour B
    4. Blaschuk O
    5. Doherty P

    (2000b) INP, a novel N-cadherin antagonist targeted to the amino acids that flank the HAV motif

    Molecular and Cellular Neuroscience 15:456–464.

    https://doi.org/10.1006/mcne.2000.0847

    • PubMed
    • Google Scholar
82. 1. Williams G
    2. Williams EJ
    3. Doherty P

    (2002) Dimeric versions of two short N-cadherin binding motifs (HAVDI and INPISG) function as N-cadherin agonists

    Journal of Biological Chemistry 277:4361–4367.

    https://doi.org/10.1074/jbc.M109185200

    • PubMed
    • Google Scholar
83. 1. Young RC

    (2007) Myometrium, and uterine contractions

    Annals of the New York Academy of Sciences 1101:72–84.

    https://doi.org/10.1196/annals.1389.038

    • PubMed
    • Google Scholar

Author details

1. Rengasayee Veeraraghavan

   1. Virginia Tech Carilion Research Institute, Virginia Polytechnic University, Roanoke, United States
   2. School of Medicine, Virginia Polytechnic University, Roanoke, United States

   Contribution

   Conceptualization, Software, Formal analysis, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   For correspondence

   saiv@vt.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8364-2222

2. Gregory S Hoeker

   1. Virginia Tech Carilion Research Institute, Virginia Polytechnic University, Roanoke, United States
   2. School of Medicine, Virginia Polytechnic University, Roanoke, United States

   Contribution

   Formal analysis, Investigation, Methodology, Writing—review and editing

   For correspondence

   ghoeker@vtc.vt.edu

   Competing interests

   No competing interests declared

3. Anita Alvarez-Laviada

   Department of Myocardial Function, Imperial College London, London, United Kingdom

   Contribution

   Formal analysis, Investigation, Methodology, Writing—review and editing

   For correspondence

   a.alvarez-laviada@imperial.ac.uk

   Competing interests

   No competing interests declared

4. Daniel Hoagland

   1. Virginia Tech Carilion Research Institute, Virginia Polytechnic University, Roanoke, United States
   2. School of Medicine, Virginia Polytechnic University, Roanoke, United States

   Contribution

   Investigation, Visualization

   For correspondence

   dhoag@vtc.vt.edu

   Competing interests

   No competing interests declared

5. Xiaoping Wan

   Heart and Vascular Research Center, MetroHealth Medical Center, Department of Medicine, Case Western Reserve University, Cleveland, United States

   Contribution

   Formal analysis, Investigation, Methodology

   For correspondence

   xwan@metrohealth.org

   Competing interests

   No competing interests declared

6. D Ryan King

   1. Virginia Tech Carilion Research Institute, Virginia Polytechnic University, Roanoke, United States
   2. School of Medicine, Virginia Polytechnic University, Roanoke, United States
   3. Graduate Program in Translational Biology, Medicine and Health, Virginia Tech, Virginia, United States

   Contribution

   Investigation, Writing—review and editing

   For correspondence

   ryanking@vt.edu

   Competing interests

   No competing interests declared

7. Jose Sanchez-Alonso

   Department of Myocardial Function, Imperial College London, London, United Kingdom

   Contribution

   Investigation, Writing—review and editing

   For correspondence

   j.sanchez-alonso-mardones@imperial.ac.uk

   Competing interests

   No competing interests declared

8. Chunling Chen

   Department of Pharmacology, University of Michigan Medical School, Ann Arbor, United States

   Contribution

   Resources, Methodology

   For correspondence

   chunling@med.umich.edu

   Competing interests

   No competing interests declared

9. Jane Jourdan

   1. Virginia Tech Carilion Research Institute, Virginia Polytechnic University, Roanoke, United States
   2. School of Medicine, Virginia Polytechnic University, Roanoke, United States

   Contribution

   Investigation, Methodology

   For correspondence

   jjane48@vtc.vt.edu

   Competing interests

   No competing interests declared

10. Lori L Isom

    Department of Pharmacology, University of Michigan Medical School, Ann Arbor, United States

    Contribution

    Resources, Funding acquisition, Methodology

    For correspondence

    lisom@umich.edu

    Competing interests

    No competing interests declared

11. Isabelle Deschenes

    1. Heart and Vascular Research Center, MetroHealth Medical Center, Department of Medicine, Case Western Reserve University, Cleveland, United States
    2. Department of Physiology and Biophysics, Case Western Reserve University, Cleveland, Unites States

    Contribution

    Resources, Investigation, Methodology

    For correspondence

    ideschenes@metrohealth.org

    Competing interests

    No competing interests declared

12. James W Smith

    1. Virginia Tech Carilion Research Institute, Virginia Polytechnic University, Roanoke, United States
    2. School of Medicine, Virginia Polytechnic University, Roanoke, United States
    3. Department of Biological Sciences, College of Science, Blacksburg, United States

    Contribution

    Funding acquisition, Investigation, Methodology

    Competing interests

    No competing interests declared

13. Julia Gorelik

    Department of Myocardial Function, Imperial College London, London, United Kingdom

    Contribution

    Resources, Investigation, Methodology

    For correspondence

    j.gorelik@imperial.ac.uk

    Competing interests

    No competing interests declared

14. Steven Poelzing

    1. Virginia Tech Carilion Research Institute, Virginia Polytechnic University, Roanoke, United States
    2. School of Medicine, Virginia Polytechnic University, Roanoke, United States
    3. Department of Biomedical Engineering and Mechanics, Virginia Polytechnic University, Blacksburg, United States

    Contribution

    Conceptualization, Resources, Funding acquisition, Investigation, Methodology, Writing—review and editing

    For correspondence

    gourdier@vtc.vt.edu

    Competing interests

    No competing interests declared

15. Robert G Gourdie

    1. Virginia Tech Carilion Research Institute, Virginia Polytechnic University, Roanoke, United States
    2. School of Medicine, Virginia Polytechnic University, Roanoke, United States
    3. Department of Biomedical Engineering and Mechanics, Virginia Polytechnic University, Blacksburg, United States

    Contribution

    Conceptualization, Resources, Supervision, Funding acquisition, Investigation, Writing—original draft, Project administration, Writing—review and editing

    For correspondence

    gourdier@vtc.vt.edu

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0001-6021-0796

• 3,380

  views

• 467

  downloads

• 100

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.