Skip main navigation
×

SAP97 and Dystrophin Macromolecular Complexes Determine Two Pools of Cardiac Sodium Channels Nav1.5 in Cardiomyocytes

Originally publishedhttps://doi.org/10.1161/CIRCRESAHA.110.228312Circulation Research. 2011;108:294–304

Abstract

Rationale:

The cardiac sodium channel Nav1.5 plays a key role in excitability and conduction. The 3 last residues of Nav1.5 (Ser-Ile-Val) constitute a PDZ-domain binding motif that interacts with the syntrophin–dystrophin complex. As dystrophin is absent at the intercalated discs, Nav1.5 could potentially interact with other, yet unknown, proteins at this site.

Objective:

The aim of this study was to determine whether Nav1.5 is part of distinct regulatory complexes at lateral membranes and intercalated discs.

Methods and Results:

Immunostaining experiments demonstrated that Nav1.5 localizes at lateral membranes of cardiomyocytes with dystrophin and syntrophin. Optical measurements on isolated dystrophin-deficient mdx hearts revealed significantly reduced conduction velocity, accompanied by strong reduction of Nav1.5 at lateral membranes of mdx cardiomyocytes. Pull-down experiments revealed that the MAGUK protein SAP97 also interacts with the SIV motif of Nav1.5, an interaction specific for SAP97 as no pull-down could be detected with other cardiac MAGUK proteins (PSD95 or ZO-1). Furthermore, immunostainings showed that Nav1.5 and SAP97 are both localized at intercalated discs. Silencing of SAP97 expression in HEK293 and rat cardiomyocytes resulted in reduced sodium current (INa) measured by patch-clamp. The INa generated by Nav1.5 channels lacking the SIV motif was also reduced. Finally, surface expression of Nav1.5 was decreased in silenced cells, as well as in cells transfected with SIV-truncated channels.

Conclusions:

These data support a model with at least 2 coexisting pools of Nav1.5 channels in cardiomyocytes: one targeted at lateral membranes by the syntrophin-dystrophin complex, and one at intercalated discs by SAP97.

The cardiac sodium channel Nav1.5 initiates the cardiac action potential, thus playing a key role in cardiac excitability and impulse propagation. The physiological importance of this channel is illustrated by numerous cardiac pathologies caused by hundreds of mutations identified in SCN5A, the gene encoding Nav1.5.1 The Nav1.5 channel is composed of one 220-kDa α-subunit that constitutes a functional channel, and 30-kDa β-subunits. In addition to these accessory β-subunits, several proteins have been shown to regulate and interact with Nav1.5.1,2 In most cases, the physiological relevance of these interactions is poorly understood, mainly because of a lack of appropriate animal models. Many of the interacting proteins bind to the C terminus of Nav1.5, where several protein–protein interaction motifs are located.1,2 We have shown that the ubiquitin–protein ligase Nedd4-2 binds the PY motif of Nav1.5 and reduces the sodium current (INa) in HEK293 cells by promoting its internalization.3 We have also demonstrated that Nav1.5 associates with the dystrophin–syntrophin multiprotein complex (DMC) in cardiac cells.4 In dystrophin-deficient mice (mdx5cv), INa and total Nav1.5 protein expression are reduced. The association between Nav1.5 and dystrophin occurs via the syntrophin family of adaptor proteins, 3 of which are expressed in the heart.4 The PDZ domain of syntrophin binds to the last 3 C-terminal residues of Nav1.5 (Ser-Ile-Val, SIV), a PDZ-domain binding motif that mediates the interaction with dystrophin and other proteins of the DMC.2 Interestingly, dystrophin has been shown to be absent from the intercalated discs of human6 and rat7 cardiac cells, suggesting that the Nav1.5 channels present at these locations may interact with other regulatory or anchoring proteins.

Proteins of the membrane associated guanylate kinase (MAGUK) family are characterized by numerous protein–protein interaction domains, including PDZ domains.8 They are involved in the function and localization of many ion channels in neurons and epithelial cells, mostly at cell to cell junctions, but little is known about their function in the heart.9 SAP97 (synapse associated protein) and ZO-1 (zonula occludens) are the main MAGUK proteins expressed in cardiomyocytes.9 PSD95 (postsynaptic density) is also expressed in the human heart, but not in the mouse heart.9,10 SAP97 regulates the targeting and localization of cardiac potassium channels, such as Kir2.x11 and Kv1.5 in myocytes.10 El-Haou et al12 recently demonstrated that SAP97 interacts with Kv4.2 and Kv4.3 channels via their PDZ-domain binding motif. They showed that suppression of SAP97 expression in rat myocytes decreases the Kv4.x-mediated current, whereas its overexpression increases it. In neurons, PSD95, which is closely related to SAP97, plays a role in clustering and anchoring ion channels to the postsynaptic plasma membrane.13 Because PSD95 is also present in the human heart and SAP97 has been shown to interact with potassium channels in cardiac cells, we hypothesized that one of these cardiac MAGUK proteins may also associate with Nav1.5.

In this study, we provide evidence for the coexistence of at least 2 pools of Nav1.5 channels in cardiomyocytes, one located at the lateral membrane with the DMC, and the other localizing with SAP97 at the intercalated discs. In sections and isolated cardiomyocytes of dystrophin-deficient mice, we observed a drastic reduction of the pool of Nav1.5 channels located at the lateral membrane. This loss could underlie the significant reduction of conduction velocity observed in these dystrophin-deficient hearts using optical measurements. Finally, we also found that the surface expression of Nav1.5 in HEK293 cells is regulated by its interaction with SAP97.

Methods

An expanded Methods section is available in the Online Data Supplement at .

Cell Preparation, Transfection, and Infection

HEK293 and Sk-Hep cells were cultured as previously described.3

Silencing and Lentiviruses

HEK293 cells stably expressing Nav1.5 were transfected using Lipofectamine (Invitrogen) as detailed in the Online Data Supplement. HEK293 cells were infected with VSV-G pseudo-typed lentiviruses.14 Adult rat cardiomyocytes were transfected with silencing plasmids, as previously published.12

Transfection

For patch-clamp experiments, HEK293 cells were transiently transfected using calcium phosphate with the different constructs mentioned in the text. Details are given in the Online Data Supplement.

Protein Extraction, Pull-Down, and Western Blot

These procedures were performed as previously reported with HEK293 and Sk-Hep cells.4 See Online Data Supplement for other tissues.

Biotinylation Assay

Transfected HEK293 cells were biotinylated as detailed in the Online Data Supplement. Biotinylated proteins were then analyzed by Western blot.

Immunohistochemistry of Rat and Mouse Ventricular Sections

Immunostainings were performed on heart cryosections as described in the Online Data Supplement.

Isolation and Immunocytochemistry of Mouse Ventricular Myocytes

See the Online Data Supplement for a detailed description.

Electrophysiology

Patch-clamp recordings were carried out in the whole-cell configuration at room temperature using solutions previously described.4 See the Online Data Supplement for details.

Conduction Velocity Measurements on Isolated Mouse Hearts

Optical measurements of conduction velocity (CV) were performed on isolated wild-type and dystrophin-deficient mdx mouse hearts as described in detail in the Online Data Supplement.

Statistical Analysis

Data are represented as mean values±SEM, unless otherwise indicated.

Results

Nav1.5 Is Localized With Dystrophin and Syntrophins Only at the Lateral Membrane

We recently demonstrated that Nav1.5 associates with the DMC in cardiac cells.4 We also observed that sodium current (INa) and Nav1.5 protein expression is reduced in dystrophin-deficient (mdx5cv) mice compared to wild type (WT).4 To further study the importance of this multiprotein complex, we analyzed the distribution of Nav1.5 channels, as well as that of dystrophin and syntrophin in rat ventricular sections (Figure 1A and 1B). Immunostaining experiments showed that Nav1.5 is distributed along the entire membrane of rat cardiomyocytes (Figure 1A and 1B, left). In contrast, we observed that Nav1.5 and dystrophin are in the same membrane compartment exclusively at the lateral membrane (Figure 1A, right), and not at the intercalated discs, where dystrophin is absent (white arrowheads, Figure 1A, middle and right). Similar results were obtained using an antibody that recognized all syntrophin isoforms (Figure 1B). Syntrophins are similarly in close proximity to Nav1.5 exclusively at the lateral membrane (Figure 1B, white arrowheads). In mouse ventricular sections, similar findings were obtained with Nav1.5, dystrophin, and syntrophin (Online Figure I). These results suggest that Nav1.5 could be part of at least 2 different protein complexes: with dystrophin at the lateral membrane and with another partner at the intercalated discs.

Figure 1.

Figure 1. Localization of Nav1.5, dystrophin and syntrophin in rat ventricular sections.A, Left, Nav1.5 (green) is present both at the lateral membranes and the intercalated discs. Middle, Dystrophin (red) is only localized at the lateral membranes. Right, Merge of the 2 images showing close localization of Nav1.5 and dystrophin at the lateral membranes. White arrowheads show the absence of dystrophin at the intercalated discs. Inset, Magnification of a portion of cell. B, Left, Nav1.5 (green) is present both at the lateral membranes and the intercalated discs. Middle, syntrophin (red) is only at the lateral membranes. Right, Merge of the 2 images showing close localization of Nav1.5 and syntrophin at the lateral membranes. The white arrowheads show the absence of syntrophin at the intercalated discs; inset: magnification of a portion of cell.

Nav1.5 Is Decreased at the Lateral Membranes of Dystrophin-Deficient Myocytes

Immunohistochemistry experiments were performed using WT and mdx5cv mouse ventricular sections (Figure 2A). A strong reduction of Nav1.5 staining was observed at the lateral membrane of mdx5cv mouse cardiomyocytes (Figure 2A, bottom) compared to WT mice (Figure 2A, top). Similarly, immunocytochemistry performed on isolated mouse cardiomyocytes confirmed that Nav1.5 is distributed along the entire cell membrane in these cells, and that lateral Nav1.5 staining in dystrophin-deficient cardiomyocytes is strongly reduced when compared to WT (Figure 2B). These results suggest that the decrease in Nav1.5 protein and INa previously seen in dystrophin-deficient mice4 is the consequence of a loss of Nav1.5 channels at the lateral membrane of cardiomyocytes.

Figure 2.

Figure 2. Loss of Nav1.5 lateral membrane staining in dystrophin-deficient cardiomyocytes.A, Nav1.5 (green) in WT (top) and mdx5cv(bottom) ventricular sections. Note the loss of lateral membrane Nav1.5 staining in the mdx5cv sections; intercalated disc staining remains almost unaffected. B, Cardiomyocytes labeled for Nav1.5 (green, left) and dystrophin (red, right). Note the absence of dystrophin staining at the intercalated discs in WT cardiomyocytes (cf, insets with magnifications). Mdx cardiomyocytes were identified by the negative dystrophin staining (bottom, right). In isolated mdx cardiomyocytes, Nav1.5 staining was similarly markedly reduced at the lateral membrane.

Impulse Propagation Is Significantly Slowed in Hearts of mdx Mice

Reduced INa and Nav1.5 protein in dystrophin-deficient hearts may contribute to the previously observed prolongation of the QRS complex duration observed in ECG recordings of mdx5cv mice.4 To investigate whether conduction velocity (CV) is reduced in dystrophin-deficient hearts, we performed optical measurements in isolated Langendorff-perfused WT and mdx hearts stained with the fluorescent voltage-reporter dye Di-4-ANNEPS. Conduction spread of the paced (cycle length 100 ms) left ventricle free wall was recorded following equilibration and confirmation of physiological spontaneous heart rates (500 to 600/min, by volumetric ECG). Maximal CV occurred in the same direction in WT and mdx hearts along the longitudinal (apico-aortal) heart axis. Isochrone maps from WT hearts showed regular anisotropic conduction spread around a point-like pacing stimulus originating from the free LV wall center (Figure 3A, left). Accordingly, CV analysis by ellipsis fitting confirmed longitudinal and transversal velocities consistent with rapid ellipsoid conduction spread and physiological anisotropy (Figure 3A, right) (see online supplemental methods for detailed description). However, conduction spread of mdx hearts resulted in tighter isochrone maps indicating conduction slowing in all directions (Figure 3B, left). In agreement with this observation ellipsis fitting analysis confirmed slowing of conduction spread in the transversal axis resulting in reduced conduction velocities (Figure 3B, right). Figure 3C shows that in contrast to WT hearts, dystrophin-deficient hearts showed on average a reduction in transversal CV by 22.2% from 0.45±0.08 m/sec to 0.35±0.07 m/sec (WT n=5, mdx n=8; P<0.05) and in longitudinal CV by 12.7% from 0.71±0.14 m/sec to 0.62±0.11 m/sec (WT n=5, mdx n=8; P=NS).

Figure 3.

Figure 3. Left ventricle conduction slowing in mdx hearts. Paced activation of (A, left) representative WT heart characterized by averaged 0.5-ms contour isochrone map showing normal elliptic conduction spread and expected anisotropic behavior versus (B, left) mdx heart showing “crowding” of averaged 0.5-ms isochrone contours, indicating slowed activation of the LV substrate. Bars, 1 mm each. Ap indicates apex; L, left; R, right; Ao, aorta; point of pacing stimulus is located in the isochrone map center; V, peripheral position of pacing electrode with the electrode stimulation pulse occurring at the isochrone center. A and B, Right, Ellipsis fitting analysis of conduction velocity from the same WT and mdx hearts shown on left: filled symbols are determined by individual ellipsis fit indicated in red color on left side. Changes in CV in the mdx heart are indicated by a reduced slope in conduction spread by corresponding data points (filled symbols); data points deviating from the linear trend attributable to proximity to pace site or organ borders were excluded from the analysis (open symbols). C, Bar graphs summarizing average CV and anisotropic ratio from ellipsis fitting analysis. D, Bar graphs summarizing average CV and anisotropic ratio from averaged longitudinal and transversal vector analysis. Data in C and D are represented as means±SD. *P<0.05; ns, not significant as indicated.

Additionally, CV has been analyzed by a second method using averaged CV vectors along the apparent longitudinal and transversal heart axis (see online supplementary methods for detailed description). Figure 3D confirms that in contrast to WT hearts, dystrophin-deficient hearts showed a longitudinal 28.4% reduction in CV (WT n=5, mdx n=8; P<0.005). In addition, transversal CV was significantly reduced by 22.4% in mdx compared to WT hearts (WT n=5, mdx n=8; P<0.05; Figure 3D). The combined reduction of transversal and longitudinal CVs resulted in a 7.3% reduction in the anisotropic ratio (WT n=5, mdx n=8; P=NS; Figure 3D) when analyzed by averaged CV vectors and by 10.7% when analyzed by the ellipsis fitting method (WT n=5, mdx n=8; P=NS; Figure 3C). Thus, under quasi-physiological conditions mimicking the resting mouse heart rate of 600/min the mdx left ventricle showed significantly slower excitation spread both in the longitudinal and transversal directions, confirming the slowed impulse propagation suggested by the widening of QRS complex. As distribution and expression of the main ventricular connexin, Cx43, is unaltered in mdx hearts (see Online Figure II and elsewhere4), this may suggest a selective role of Nav1.5 channels located at the lateral membrane for impulse propagation in cardiac muscle.

The Carboxyl Terminus of Nav1.5 Interacts With SAP97

To identify proteins that could specifically interact with the fractions of Nav1.5 at the intercalated discs, we studied proteins of the MAGUK family. We performed pull-down experiments using GST-fusion proteins comprising the last 66 amino acids of the C terminus (C-ter) of Nav1.5, as well as fusion proteins lacking the 3 last SIV residues (ΔSIV, Figure 4A). Western blot experiments performed using an anti-PSD95 family antibody, known to recognize several cardiac MAGUKs (Figure 4B, Online Figure III), revealed that Nav1.5 C-ter WT precipitated proteins containing PDZ domains (2 bands at ≈140 kDa, the molecular weight of SAP97) from mouse ventricles and human atrium. This interaction is dependent on the PDZ-domain binding motif of Nav1.5, as no protein was observed with the truncated construct. Similar results were obtained using a SAP97 specific antibody (Figure 4C, Online Figure III). The similarity between these 2 results (Figure 4B and 4C) suggests that SAP97 is the predominant MAGUK interacting with Nav1.5. To assess if this interaction was specific for SAP97, we tested a specific PSD95 antibody (Figure 4D). No interaction was seen with either fusion protein (Nav1.5 C-ter WT or ΔSIV). We also verified whether Nav1.5 C terminus could interact with ZO-1, a cardiac MAGUK protein known to interact with connexins. We generated a fusion protein containing the last amino acids of connexin-45 (Cx45) (Figure 4A), which has previously been shown to bind ZO-1.15 Pull-down experiments performed using SK-Hep cell lysates, which express ZO-1 endogenously, confirmed that Cx45 C-ter precipitated ZO-1, but that Nav1.5 C-ter (Figure 4E) did not. Combined, these results support a specific interaction between the PDZ-domain binding motif of Nav1.5 and SAP97 in mouse ventricles and human atria.

Figure 4.

Figure 4. SAP97, but not PSD95, binds to WT Nav1.5 fusion protein in both mouse and human cardiac tissues.A, Schematic representation of the fusion proteins used to perform pull-down assays shown in B through E. B, Western blots of pulled-down fractions performed on mouse ventricular and human atrial lysates using a monoclonal anti-PSD95-family antibody; 100 μg of lysates were loaded in the input lane. C, Same experiment as in panel B, using a monoclonal anti-SAP97 antibody. D, Same experiment as in B on human atria, using a monoclonal anti-PSD95 antibody. E, Western blots of pulled-down fractions performed on Sk-Hep cells using the polyclonal anti–ZO-1 antibody. The bottom images are Ponceau stainings indicating that equal amounts of fusion proteins were used. See the Online Data Supplement for full-size blots of Figure 3B and 3C.

Localization of SAP97, Nav1.5, Cx43, and Syntrophin in Rat Cardiomyocytes

To analyze the distribution of Nav1.5 and SAP97 in native cardiac tissue, we performed immunostainings using Nav1.5, SAP97 and Cx43 antibodies on rat heart sections. As presented in Figure 5A, SAP97 is predominantly expressed at the intercalated discs (Figure 5A, middle) where Nav1.5 is also observed (white arrowheads, Figure 5A right). We also performed coimmunostainings of SAP97 and Cx43, showing that both proteins are located at the intercalated discs (white arrowheads, Figure 5B, right).

Figure 5.

Figure 5. Localization of Nav1.5, SAP97, and Cx43 in rat heart sections.A, Left, Nav1.5 (green) both at the lateral membranes and the intercalated discs; middle: SAP97 (red) only at the intercalated discs. Right, Merge of the 2 images showing close localization of Nav1.5 and SAP97 at the intercalated discs (white arrowheads, and magnification in the inset). B, Left, Cx43 (green) only at the intercalated discs. Middle, SAP97 (red) only at the intercalated discs. Right, Merge of the 2 images showing punctuate staining of Cx43 and SAP97 at the intercalated discs (white arrowheads and magnification in the inset).

Silencing of SAP97 in HEK293 Cells, Cardiomyocytes, and Nav1.5 Channels Lacking the SIV-Domain Reduces INa

The results presented in Figure 4 indicate that Nav1.5 and SAP97 interact via their SIV-motif and PDZ domain, respectively. Electrophysiological studies were performed to characterize the role of this interaction. First, patch-clamp experiments using HEK293 cells transiently transfected with Nav1.5 and SAP97 were carried out. Unexpectedly, expression of SAP97 did not modify INa (not shown). This result could be an indication that endogenous SAP97 is expressed at a saturating level in HEK293 cells. We subsequently reduced the expression of SAP97 using short hairpin (sh)RNA. When HEK293 cells expressing Nav1.5 (HEK293-Nav1.5) were transfected with SAP97 silencing plasmids, INa was reduced by 55±4% (Figure 6B). However, the transfection technique was not efficient enough to silence SAP97 expression in the majority of the cells. Therefore, we developed a lentivirus-based strategy to deliver shRNA. Western-blots performed with HEK293 cells infected with shRNA-containing lentiviruses revealed that SAP97 expression was drastically reduced (−83%, Figure 6A) when compared to nonsilenced cells infected with scrambled shRNA. In HEK293-Nav1.5 cells silenced with lentiviruses, we observed a reduction of INa that was comparable to the results obtained when using the silencing plasmid (not shown). Next, we studied the consequence of silencing SAP97 on endogenous INa in rat atrial myocytes maintained in short term culture. This procedure has been previously shown to efficiently suppress endogenous cardiac SAP97.12 After 3 to 4 days of culture, current recordings at -20 mV using green-fluorescent atrial myocytes (≈20% of the cells) revealed a much smaller INa in myocytes transfected with shSAP97 than those with scrambled shRNA (−66±8%, Figure 6C). SAP97-silenced or scrambled HEK293 cells were then transiently transfected with Nav1.5 WT or ΔSIV. The current generated by Nav1.5 ΔSIV channels was reduced by 57±9% compared to WT channels in control cells expressing the scrambled shRNA (Figure 6D). In silenced cells (shSAP97), WT and ΔSIV INa were comparable (49±11% and 57±7% of control values, respectively, Figure 6D). Voltage dependence of activation and inactivation was not significantly different between the four conditions (see Online Figure V). Because β1 subunits are known to associate with Nav1.5,16 we performed similar experiments using HEK293 cells transfected with Nav1.5 WT or ΔSIV and β1. The results were similar whether the β1 subunit was present or not (Online Figure VIA), without affecting the voltage dependence of activation and inactivation (Online Figure VIB). We finally investigated whether the reduced INa could be rescued by another protein bearing a PDZ binding domain. As shown in Online Figure VII, overexpression of α1-syntrophin (previously shown to interact with Nav1.54) in silenced cells did not increase INa.

Figure 6.

Figure 6. Effect of SAP97 silencing in HEK293 cells and rat adult myocytes.A, Western blot of scrambled or SAP97-silenced (shSAP97) HEK293 cell lysates and quantification of the level of protein expression. ***P<0.001. B,INa density of HEK293 cells stably expressing Nav1.5 (control: 605.6±83.9 pA/pF) and after SAP97 silencing (shSAP97: 271.8±27.1 pA/pF). ***P<0.001 (inset: representative currents). C,INa density of scrambled or silenced (ShSAP97) rat adult myocytes (scrambled: 229.3±41.0, shSAP97: 78.9±18.9 pA/pF). **P<0.01. D, INa density of scrambled or silenced (ShSAP97) HEK293 cells, transiently transfected with WT or ΔSIV Nav1.5 channels (inset: representative currents); *P<0.05, **P<0.01 (1-way ANOVA and Bonferroni post test compared to the “scrambled+WT” condition). The number of cells is written in the corresponding columns.

Silencing of SAP97 in Cultured Rat Atrial Myocytes Reorganizes Nav1.5 Expression

We also investigated whether silencing of SAP97 (Ad shSAP97) alters the organization of Nav1.5 in cardiomyocytes, using a model of cultured atrial myocytes at confluence stage when cells reestablish highly organized cell-cell contacts as previously shown.17 Immunostainings on short-term cultured rat atrial myocytes indicate that the total expression of Nav1.5 channels in myocytes transduced with shSAP97 is reduced when compared to cells overexpressing SAP97 (Ad SAP97), and that the localization of the channels at the plasma membrane is reorganized (Online Figure IVA). Costainings of Nav1.5 and syntrophin on rat cardiomyocytes confirm that these 2 proteins are both localized at the cell surface but not at cell-cell contacts where syntrophin is absent (Online Figure IV, B, top). In SAP97 silenced myocytes, syntrophin expression remains unchanged at the level of non–cell-to-cell contacts (Online Figure IVB, bottom).

Expression of Nav1.5 Channels Lacking the SIV-Domain Is Reduced at the Cell Surface of HEK293 Cells and SAP97-Silenced Cells

Electrophysiological data indicate that INa is dependent on the presence of SAP97 as well as on the SIV motif of Nav1.5. We examined whether the reduced current amplitude was attributable to decreased expression of the channel at the cell surface. Surface biotinylation assays performed on HEK293 cells that had been transiently transfected with Nav1.5 WT or ΔSIV (Figure 7A and 7B) revealed that cell membrane expression of truncated channels was reduced by 27±3% compared to WT channels, without affecting total Nav1.5 protein expression. Similar experiments using scramble-infected HEK293 cells yielded comparable results (35±6% reduction; Figure 7C and 7D). When SAP97-silenced HEK293 cells were transfected with WT or ΔSIV channels, the number of channels expressed at the cell membrane was further reduced (56±5% and 56±6%; Figure 7C and 7D). These results suggest that the decreased INa measured in SAP97-silenced cells or those with truncated channels may be attributable to reduced cell surface expression of Nav1.5, supporting the idea that SAP97 regulates cell membrane density of Nav1.5. Data presented in Online Figure VIII show that this regulation is independent of the ubiquitin-protein ligase Nedd4-2, which we have previously shown to regulate the number of Nav1.5 channels at the cell surface by interacting with a PY-motif found at the C terminus of the channel.3

Figure 7.

Figure 7. Effect of the absence of the SIV motif and of SAP97 silencing on Nav1.5 channel expression at the plasma membrane.A, Western blot of total protein lysates (left, INPUT) and of biotinylated plasma membrane fraction (right) of HEK293 cells transfected with Nav1.5 WT or ΔSIV channels showing a decrease of plasma membrane expression of truncated channels. B, Quantification of Nav1.5 channel expression at the plasma membrane. ***P<0.001; mean of 3 distinct experiments. C, Western blot of total protein lysates (left, INPUT) and of biotinylated plasma membrane fraction (right) of HEK293 cells transfected with Nav1.5 WT or ΔSIV channels in scrambled or SAP97-silenced (shSAP97) HEK293 cells. A decrease in the expression, at the plasma membrane, of truncated channels and in SAP97-silenced cells is observed. The middle lane (SAP97) shows the reduction of SAP97 expression in silenced cells. Bottom, Absence of actin signal in biotinylated fractions. D, Quantification of Nav1.5 channel plasma membrane expression compared to total lysates. ***P<0.001; means of 5 distinct experiments.

Discussion

The major findings of the present study are: (1) Nav1.5 is part of at least 2 distinct multiprotein complexes in cardiomyocytes, one localized at the lateral membrane with dystrophin and syntrophin proteins, and the other involving the MAGUK protein SAP97 at the intercalated discs; (2) specific regulation of lateral membrane localization and density of Nav1.5 is dependent on dystrophin expression, the absence of which may underlie conduction slowing observed in dystrophin-deficient hearts; and (3) the absence of the SIV motif of Nav1.5 or depletion of SAP97 results in reduced channel expression at the cell surface, and a subsequent decrease of INa in HEK293 cells and adult cardiac myocytes.

Distribution of Nav1.5 Channels in Cardiomyocytes

Studies investigating the localization of Nav1.5 at the cell membrane of cardiac cells have revealed conflicting results. Although many articles have described Nav1.5 stainings at the intercalated discs and lateral membrane of rat,18,19 mouse,20 dog,21 and human22 cardiomyocytes, other studies have shown almost exclusive expression of the channel at the intercalated discs.2325 These discrepancies may be attributable to the alternative use of either heart sections or isolated myocytes, where a reorganization of proteins at the plasma membrane can occur. In a recent study,26 macropatch recordings of the 2 subcellular domains were performed, and it was concluded that the Nav1.5-mediated current is almost nonexistent at the lateral membrane. The expression, or lack thereof, of Nav1.5 channels at the lateral membranes likely has functional consequences on impulse propagation, as suggested by modeling studies.25 In the present work, we demonstrate that Nav1.5 channels are present all along the plasma membrane of both rat and mouse cardiomyocytes. This finding is supported by the reduction of lateral Nav1.5 expression in dystrophin-deficient myocytes, with no significant alteration in the expression at the intercalated discs. This observation is consistent with our previous study,4 in which we observed a ≈50% decrease of total Nav1.5 protein expression in the absence of dystrophin, with a concomitant INa reduction. Taken together, these observations strongly support the notion that a significant fraction of functional Nav1.5 channels are located outside of the intercalated discs. These lateral membrane channels are likely to play a role in conduction, because their reduction alters ventricular conduction, which is reflected in a widening of the QRS complex in ECG recordings from mdx5cv mice where the hearts are otherwise free from structural remodeling.4 Confirming this concept, we show here for the first time, using noninvasive optical measurements of conduction velocity from the left ventricle free wall, and by 2 different analysis methods, that impulse propagation is significantly slowed in the transversal direction in dystrophin-deficient hearts that have reduced expression Nav1.5 at the lateral membrane. These findings are in line with those of Baba and coworkers,21 who showed that dog cardiomyocytes isolated from peri-infarcted zones displayed reduced INa density, associated with specific loss of the Nav1.5 staining at the lateral membranes only. As the Nav1.5 staining remained unchanged at intercalated discs, the reduction of INa was proposed to be the cause of altered impulse propagation in the diseased tissue.21 These observations are also in agreement with those of a recent study27 reporting reduced excitability, slowed conduction velocity, and altered Nav1.5 expression in hearts of mdx mice. However, these investigators partly attributed these results to a reorganization of Cx43. Similarly to Sanford and coworkers, we could not confirm this finding in the present study, and found no remodeling of Cx43.28

Interaction Between Nav1.5 and SAP97 at Intercalated Discs

The carboxyl terminus of Nav1.5 has several protein–protein interaction domains that associate with numerous partners to regulate channel activity.2 As aforementioned, we have reported that the last 3 residues of Nav1.5 (SIV) interacts with the DMC, as well as with the protein tyrosine phosphatase PTPH1.29 Other groups have reported that ion channels bearing a PDZ-domain binding motif interact with members of the MAGUK family in various tissues.8 Several potassium channels of the Kir and Kv families display PDZ-domain binding motifs that interact with the MAGUK proteins PSD95 or SAP97.8,12 Our pull-down experiments strongly support the existence of an interaction between Nav1.5 and SAP97 in cardiac tissue via the Nav1.5 SIV motif. These results are supported by a recent study,30 that suggests that Nav1.5 binds the first PDZ domain of SAP97. In our study, we also describe a close localization of SAP97 and Nav1.5 at intercalated discs. Immunostainings confirm data from Godreau et al9 who found an enrichment of SAP97 at this subcellular site. These results suggest that SAP97 is involved in targeting or anchoring of Nav1.5 channels to the intercalated discs of cardiomyocytes. Furthermore, SAP97 expression is unaltered in hearts of mdx mice (see Online Figure IX), indicating that the pool of Nav1.5 channels regulated at their PDZ domain-binding motif in the intercalated discs could be independent of the one complexed to syntrophin and DMC at lateral membranes of cardiomyocytes.

Physiological Relevance of the Two Regulating Mechanisms

In addition to the DMC, ankyrin-G has also been shown to interact with and regulate the trafficking of Nav1.5 at the cell membrane of cardiomyocytes.31,32 Ankyrin-G is primarily expressed at the intercalated discs, with another population at T-tubules; it is therefore unlikely that it is responsible for the specific targeting of Nav1.5 to the 2 compartments observed in our study. Whether the roles of syntrophin-dystrophin and ankyrin-G in trafficking, targeting, and anchoring of Nav1.5 to the lateral membrane are complementary or partly redundant remains to be elucidated. It may be speculated that ankyrin-G is involved in forward trafficking and correct targeting of the channel, and that PDZ-domain proteins anchor it in specific membrane compartments. The exact molecular mechanisms underlying the loss of Nav1.5 expression at the lateral membrane in the absence of dystrophin remains still unclear. Cheng and coworkers33 recently described several mutations in the gene encoding α1-syntrophin in patients with sudden infant death syndrome. Despite these results, clearly demonstrating the importance of α1-syntrophin in cardiomyocytes function, our data suggest that this protein complex is restricted to the lateral membrane. Sato and coworkers34 recently demonstrated that INa is regulated by the protein plakophilin-2, which is enriched at the intercalated discs.35 Because plakophilin-2 does not contain a PDZ domain, this suggests that Nav1.5 could be differentially regulated by multiple proteins at the intercalated discs including SAP97 and plakophilin-2, similarly to the lateral membrane where it interacts with the DMC and ankyrin-G, within complexes that may or may not be distinct (Figure 8). Deciphering the specific and distinct roles of the different proteins in trafficking and/or anchoring of cardiac ion channels including Nav1.5 to different cellular compartments will be the goal of future experiments.

Figure 8.

Figure 8. Schematic presentation of Nav1.5 channels of the lateral membrane (A) and the intercalated discs (B). Note that the concept of 2 pools of Nav1.5 channels is most likely an oversimplification. Because ankyrin-G is predominantly found in the T-tubules where little syntrophin-dystrophin is present, a third T-tubular compartment may be proposed.

Limitations of the Study

This study demonstrates the presence of at least 2 different populations of Nav1.5 channels in cardiac cells. However the use of the mdx mouse as a disease model does not allow final conclusions about the physiological function of the 2 different Nav1.5 compartments. The optical mapping data indicate conduction slowing in all directions, and because the exact fiber orientation of the given cardiac substrate is not known, we assume that fiber orientation is similar in WT and mdx hearts at young ages in agreement with previous studies. In addition, our data do not show significant reduction of Nav1.5 expression at the intercalated disc in mdx hearts based on confocal immunofluorescence data. Nevertheless, potential changes of the mdx membrane ultrastructure at this complex site could underlie alterations in intercalated disc excitability that could not be established specifically here because of resolution limitations of confocal light microscopy. Furthermore, subtle, as yet unrecognized intercellular histological changes in mdx hearts could also contribute to the conduction slowing observed in optical mapping experiments and are subject to the same microscopy diffraction limitations. On the other hand, the present data suggests that SAP97 is predominantly localized in intercalated discs, and that SAP97 protein expression is not different in mdx mouse hearts. However, because immunostaining of SAP97 in mouse heart failed to produce specific staining despite repeated attempts with different antibodies and tissue fixations, possible reorganization of SAP97 cannot be excluded. Finally, potential upregulation or remodeling of connexins other than Cx43 was not addressed in this study.

Conclusions

The present study provides evidence that a fraction of Nav1.5 channels is regulated by the DMC at the lateral myocyte membrane, and a different pool interacts with SAP97 at the intercalated discs of cardiomyocytes. Although the functional relevance of these distinct localizations of the main cardiac sodium channel has yet to be clarified, this study establishes that lateral sodium channels have a role in cardiac impulse propagation. Generation of a cardiac-specific SAP97 knock-out mouse models will likely elucidate the specific roles of this protein in cardiac function. Finally, similar to previous examples of proteins interacting with Nav1.5,2 genetic variants of SAP97 may be linked to pathological cardiac phenotypes.

Non-standard Abbreviations and Acronyms

Cx

connexin

CV

conduction velocity

DMC

dystrophin multiprotein complex

MAGUK

membrane-associated guanylate kinase

PSD95

Postsynaptic density 95

SAP97

synapse-associated protein 97

shRNA

short hairpin RNA

WT

wild type

ZO

zonula occludens

Acknowledgments

We are grateful to Maria C. Essers for expert technical assistance.

Sources of Funding

The research leading to these results has received funding from the European Community's Seventh Framework Programme FP7/2007-2013 under grant agreement No. HEALTH-F2-2009-241526, EUTrigTreat (to H.A., S.L., and S.E.L.). Other support was obtained from grants of the Swiss National Science Foundation to HA (310030_120707), Swiss Heart Foundation, Association Francaise contre les Myopathies (grant 14305), ANR-08-GENO-006-01 (project R08147DS), and by the International Max Planck Research School (IMPRS) for Physics of Biological and Complex Systems (PhD student support to N.R.).

Disclosures

None.

Footnotes

In November 2010, the average time from submission to first decision for all original research papers submitted to Circulation Research was 13.2 days.

*Both authors contributed equally to this work.

Correspondence to Dr Hugues Abriel, MD, PhD,
University of Bern, Department of Clinical Research, Murtenstrasse, 35, 3010 Bern, Switzerland
. E-mail

References

  • 1. Zimmer T, Surber R. SCN5A channelopathies - an update on mutations and mechanisms. Prog Biophys Mol Biol. 2008; 98:120–136.CrossrefMedlineGoogle Scholar
  • 2. Abriel H. Cardiac sodium channel Nav1.5 and interacting proteins: physiology and pathophysiology. J Mol Cell Cardiol. 2010; 48:2–11.CrossrefMedlineGoogle Scholar
  • 3. Rougier J-S, van Bemmelen MX, Bruce MC, Jespersen T, Gavillet B, Apotheloz F, Cordonier S, Staub O, Rotin D, Abriel H. Molecular determinants of voltage-gated sodium channel regulation by the Nedd4/Nedd4-like proteins. Am J Physiol Cell Physiol. 2005; 288:C692–C701.CrossrefMedlineGoogle Scholar
  • 4. Gavillet B, Rougier JS, Domenighetti AA, Behar R, Boixel C, Ruchat P, Lehr HA, Pedrazzini T, Abriel H. Cardiac sodium channel Nav1.5 is regulated by a multiprotein complex composed of syntrophins and dystrophin. Circ Res. 2006; 99:407–414.LinkGoogle Scholar
  • 5. Blake DJ, Weir A, Newey SE, Davies KE. Function and genetics of dystrophin and dystrophin-related proteins in muscle. Physiol Rev. 2002; 82:291–329.CrossrefMedlineGoogle Scholar
  • 6. Kaprielian RR, Stevenson S, Rothery SM, Cullen MJ, Severs NJ. Distinct patterns of dystrophin organization in myocyte sarcolemma and transverse tubules of normal and diseased human myocardium. Circulation. 2000; 101:2586–2594.CrossrefMedlineGoogle Scholar
  • 7. Stevenson SA, Cullen MJ, Rothery S, Coppen SR, Severs NJ. High-resolution en-face visualization of the cardiomyocyte plasma membrane reveals distinctive distributions of spectrin and dystrophin. Eur J Cell Biol. 2005; 84:961–971.CrossrefMedlineGoogle Scholar
  • 8. Funke L, Dakoji S, Bredt DS. Membrane-associated guanylate kinases regulate adhesion and plasticity at cell junctions. Annu Rev Biochem. 2005; 74:219–245.CrossrefMedlineGoogle Scholar
  • 9. Godreau D, Vranckx R, Maguy A, Rucker-Martin C, Goyenvalle C, Abdelshafy S, Tessier S, Couetil JP, Hatem SN. Expression, regulation and role of the MAGUK protein SAP-97 in human atrial myocardium. Cardiovasc Res. 2002; 56:433–442.CrossrefMedlineGoogle Scholar
  • 10. Godreau D, Vranckx R, Maguy A, Goyenvalle C, Hatem SN. Different isoforms of synapse-associated protein, SAP97, are expressed in the heart and have distinct effects on the voltage-gated K+ channel Kv1.5. J Biol Chem. 2003; 278:47046–47052.CrossrefMedlineGoogle Scholar
  • 11. Leonoudakis D, Conti LR, Anderson S, Radeke CM, McGuire LMM, Adams ME, Froehner SC, Yates JR, Vandenberg CA. Protein trafficking and anchoring complexes revealed by proteomic analysis of inward rectifier potassium channel (Kir2.x) associated proteins. J Biol Chem. 2004; 279:22331–22346.CrossrefMedlineGoogle Scholar
  • 12. El-Haou S, Balse E, Neyroud N, Dilanian G, Gavillet B, Abriel H, Coulombe A, Jeromin A, Hatem SN. Kv4 potassium channels form a tripartite complex with the anchoring protein SAP97 and CaMKII in cardiac myocytes. Circ Res. 2009; 104:758–769.LinkGoogle Scholar
  • 13. Lai HC, Jan LY. The distribution and targeting of neuronal voltage-gated ion channels. Nat Rev Neurosci. 2006; 7:548–562.CrossrefMedlineGoogle Scholar
  • 14. Appert-Collin A, Cotecchia S, Nenniger-Tosato M, Pedrazzini T, Diviani D. The A-kinase anchoring protein (AKAP)-Lbc-signaling complex mediates alpha1 adrenergic receptor-induced cardiomyocyte hypertrophy. Proc Natl Acad Sci U S A. 2007; 104:10140–10145.CrossrefMedlineGoogle Scholar
  • 15. Kausalya PJ, Reichert M, Hunziker W. Connexin45 directly binds to ZO-1 and localizes to the tight junction region in epithelial MDCK cells. FEBS Lett. 2001; 505:92–96.CrossrefMedlineGoogle Scholar
  • 16. Malhotra JD, Chen C, Rivolta I, Abriel H, Malhotra R, Mattei LN, Brosius FC, Kass RS, Isom LL. Characterization of sodium channel alpha- and beta-subunits in rat and mouse cardiac myocytes. Circulation. 2001; 103:1303–1310.LinkGoogle Scholar
  • 17. Abi-Char J, El-Haou S, Balse E, Neyroud N, Vranckx R, Coulombe A, Hatem SN. The anchoring protein SAP97 retains Kv1.5 channels in the plasma membrane of cardiac myocytes. Am J Physiol Heart Circ Physiol. 2008; 294:H1851–H1861.CrossrefMedlineGoogle Scholar
  • 18. Scriven DR, Dan P, Moore ED. Distribution of proteins implicated in excitation-contraction coupling in rat ventricular myocytes. Biophys J. 2000; 79:2682–2691.CrossrefMedlineGoogle Scholar
  • 19. Yarbrough TL, Lu T, Lee HC, Shibata EF. Localization of cardiac sodium channels in caveolin-rich membrane domains: regulation of sodium current amplitude. Circ Res. 2002; 90:443–449.CrossrefMedlineGoogle Scholar
  • 20. Haufe V, Camacho JA, Dumaine R, Gunther B, Bollensdorff C, Segond von Banchet G, Benndorf K, Zimmer T. Expression pattern of neuronal and skeletal muscle voltage-gated Na+ channels in the developing mouse heart. J Physiol. 2005; 564(Pt 3):683–696.CrossrefMedlineGoogle Scholar
  • 21. Baba S, Dun W, Cabo C, Boyden PA. Remodeling in cells from different regions of the reentrant circuit during ventricular tachycardia. Circulation. 2005; 112:2386–2396.LinkGoogle Scholar
  • 22. Chandler NJ, Greener ID, Tellez JO, Inada S, Musa H, Molenaar P, DiFrancesco D, Baruscotti M, Longhi R, Anderson RH, Billeter R, Sharma V, Sigg DC, Boyett MR, Dobrzynski H. Molecular architecture of the human sinus node: insights into the function of the cardiac pacemaker. Circulation. 2009; 119:1562–1575.LinkGoogle Scholar
  • 23. Maier SK, Westenbroek RE, Schenkman KA, Feigl EO, Scheuer T, Catterall WA. An unexpected role for brain-type sodium channels in coupling of cell surface depolarization to contraction in the heart. Proc Natl Acad Sci U S A. 2002; 99:4073–4078.CrossrefMedlineGoogle Scholar
  • 24. Maier SKG, Westenbroek RE, McCormick KA, Curtis R, Scheuer T, Catterall WA. Distinct subcellular localization of different sodium channel {alpha} and {beta} subunits in single ventricular myocytes from mouse heart. Circulation. 2004; 109:1421–1427.LinkGoogle Scholar
  • 25. Kucera JP, Rohr S, Rudy Y. Localization of sodium channels in intercalated disks modulates cardiac conduction. Circ Res. 2002; 91:1176–1182.LinkGoogle Scholar
  • 26. Verkerk AO, van Ginneken ACG, van Veen TAB, Tan HL. Effects of heart failure on brain-type Na+ channels in rabbit ventricular myocytes. Europace. 2007; 9:571–577.CrossrefMedlineGoogle Scholar
  • 27. Colussi C, Berni R, Rosati J, Straino S, Vitale S, Spallotta F, Baruffi S, Bocchi L, Delucchi F, Rossi S, Savi M, Rotili D, Quaini F, Macchi E, Stilli D, Musso E, Mai A, Gaetano C, Capogrossi MC. The histone deacetylase inhibitor suberoylanilide hydroxamic acid reduces cardiac arrhythmias in dystrophic mice. Cardiovasc Res. 2010; 87:73–82.CrossrefMedlineGoogle Scholar
  • 28. Sanford JL, Edwards JD, Mays TA, Gong B, Merriam AP, Rafael-Fortney JA. Claudin-5 localizes to the lateral membranes of cardiomyocytes and is altered in utrophin/dystrophin-deficient cardiomyopathic mice. J Mol Cell Cardiol. 2005; 38:323–332.CrossrefMedlineGoogle Scholar
  • 29. Jespersen T, Gavillet B, van Bemmelen MX, Cordonier S, Thomas MA, Staub O, Abriel H. Cardiac sodium channel Nav1.5 interacts with and is regulated by the protein tyrosine phosphatase PTPH1. Biochem Biophys Res Commun. 2006; 353:1456–1463.Google Scholar
  • 30. Stiffler MA, Chen JR, Grantcharova VP, Lei Y, Fuchs D, Allen JE, Zaslavskaia LA, MacBeath G. PDZ domain binding selectivity is optimized across the mouse proteome. Science. 2007; 317:364–369.CrossrefMedlineGoogle Scholar
  • 31. Lowe JS, Palygin O, Bhasin N, Hund TJ, Boyden PA, Shibata E, Anderson ME, Mohler PJ. Voltage-gated Nav channel targeting in the heart requires an ankyrin-G dependent cellular pathway. J Cell Biol. 2008; 180:173–186.CrossrefMedlineGoogle Scholar
  • 32. Mohler PJ, Splawski I, Napolitano C, Bottelli G, Sharpe L, Timothy K, Priori SG, Keating MT, Bennett V. A cardiac arrhythmia syndrome caused by loss of ankyrin-B function. Proc Natl Acad Sci U S A. 2004; 101:9137–9142.CrossrefMedlineGoogle Scholar
  • 33. Cheng J, Van Norstrand DW, Medeiros-Domingo A, Valdivia C, Tan BH, Ye B, Kroboth S, Vatta M, Tester DJ, January CT, Makielski JC, Ackerman MJ. {alpha}1-Syntrophin mutations identified in sudden infant death syndrome cause an increase in late cardiac sodium current. Circ Arrhythmia Electrophysiol. 2009; 2:667–676.LinkGoogle Scholar
  • 34. Sato PY, Musa H, Coombs W, Guerrero-Serna G, Patino GA, Taffet SM, Isom LL, Delmar M. Loss of plakophilin-2 expression leads to decreased sodium current and slower conduction velocity in cultured cardiac myocytes. Circ Res. 2009; 105:523–526.LinkGoogle Scholar
  • 35. Oxford EM, Everitt M, Coombs W, Fox PR, Kraus M, Gelzer AR, Saffitz J, Taffet SM, Moise NS, Delmar M. Molecular composition of the intercalated disc in a spontaneous canine animal model of arrhythmogenic right ventricular dysplasia/cardiomyopathy. Heart Rhythm. 2007; 4:1196–1205.CrossrefMedlineGoogle Scholar

Novelty and Significance

What Is Known?

  • Cardiac sodium channel Nav1.5 plays an essential role in action potential initiation and impulse propagation.

  • Hundreds of mutations in the gene encoding Nav1.5, SCN5A, have been found in patients with various cardiac disorders such as congenital long QT syndrome, Brugada syndrome, and dilated cardiomyopathy.

  • Many regulatory proteins have been found to interact with Nav1.5 and form macromolecular complexes.

What New Information Does This Article Contribute?

  • Cardiac sodium channels are parts of at least 2 distinct macromolecular complexes in cardiac cells: one localized at lateral membranes with the dystrophin complex and one at the intercalated discs.

  • Absence of dystrophin leads to a specific downregulation of lateral Nav1.5 channels and impulse propagation slowing.

  • The scaffolding protein SAP97, which is predominantly found at the intercalated discs, interacts with Nav1.5 and regulates its membrane density, hence forming another macromolecular complex.

Previously, we have demonstrated that Nav1.5 forms a macromolecular complex with dystrophin and syntrophin proteins. However because dystrophin is absent from the intercalated discs, we investigated the composition of the complex in this compartment. We found that Nav1.5 channels are a parts of distinct macromolecular complexes in cardiomyocytes, one localized at the lateral membrane with the dystrophin/syntrophin complex, and the other involving the scaffolding protein SAP97 at the intercalated discs and that specific regulation of lateral membrane localization and density of Nav1.5 is dependent on dystrophin, the absence of which underlie conduction slowing observed in dystrophin-deficient hearts. We report that depletion of SAP97 reduces channel expression at the cell surface and decreases sodium current. These findings suggest that there is not “one” sodium channel in cardiac cells, but multiple Nav1.5 channels that are differentially regulated in the different membrane compartments. These differences may underlie the variability of the cardiac phenotypes observed with genetic dysfunction of the cardiac sodium channel and could impart specific roles to different macromolecular complexes in normal and diseased heart.