The Intercalated Disc: A Molecular Network That Integrates Electrical Coupling, Intercellular Adhesion, and Cell Excitability

Published on 20/06/2015 by admin

Filed under Cardiovascular

Last modified 20/06/2015

Print this page

This article have been viewed 1525 times

The Intercalated Disc

A Molecular Network That Integrates Electrical Coupling, Intercellular Adhesion, and Cell Excitability

Marina Cerrone, Esperanza Agullo-Pascual and Mario Delmar

Chapter Outline

Historical Perspective

The heartbeat results from the added output of millions of cells that contract in synchrony. To achieve this function, complex molecular networks work in concert, with exquisite temporal precision. The accurate timing of the molecular events demands a comparable precision on the location of each molecule within the cell. Indeed, molecular networks organize within well-confined microdomains, where physical proximity allows for prompt and efficient interaction. In turn, loss of molecular organization in the nanoscale can be a core component in the pathophysiology of disease.

This chapter focuses on the intercalated disc, a region of specialization formed at the end-end site of contact between cardiac myocytes. When first observed through light microscopy (in 1866), the intercalated disc was considered “a cementing material” at cardiac cell boundaries. However, the scientific community at the time was divided on whether cardiac cells were separate from each other or fused into a single syncytium. The latter hypothesis was in fact favored by most during the early twentieth century. The advent of electron microscopy eventually settled this debate. The studies of Sjostrand and Andersson ¹ and others showed that the intercalated disc consisted of a double membrane, flanked by the termination of myofibrils in dense material. Their observations led Muir² to conclude that “the discs represent the junctions between neighboring cardiac muscle cells.” He wrote that “there is no valid evidence to contest the statement that the intercalated discs are specialized regions of cellular adhesion.” Since then, and as a result of the pioneering electrophysiology experiments of Weidmann³ and other ultrastructural observations,⁴ the intercalated disc has been recognized as an area of specialization that provides a physical continuum between cardiac cells through mechanical junctions (desmosomes, adherens junctions) and intercellular channels (gap junctions).

The availability of immunofluorescence microscopy allowed the demonstration that other molecular complexes, not detectable by electron microscopy, are also present in the intercalated disc. Of particular relevance to this chapter is the fact that channel protein complexes involved in both depolarization and repolarization localize preferentially to the intercalated disc. This physical proximity allows for a key functional consequence; molecules traditionally defined as junctional, such as connexin43 (Cx43) and plakophilin-2 (PKP2), actually regulate the function of ion channels responsible for the action potential. In turn, molecule accessories to ion channels are also relevant for cell adhesion and gap junction function.5–7 These data support the notion that the intercalated disc is not just the site of residence of independent junctional and nonjunctional complexes that are oblivious to the presence and function of the others. It is, rather, the home of a protein interacting network (an interactome) where molecules multitask to achieve jointly, intimately related functions: the entry and exit of charge into the cell, the transfer of charge between cells, and the anchoring of cells to each other, which provides a mechanically stable environment critical to ion channel function.

The following sections contain an update of current knowledge on the composition of selected molecular complexes of the intercalated disc, their interactions, and the possible mechanisms by which dysfunction of intercalated disc molecules may lead to arrhythmia disease. This discussion converges with other investigators to challenge the notions that: (1) connexins are only involved in the formation of gap junctions, (2) sodium channels are only important for single cell excitability, (3) desmosomal molecules are only relevant to cell adhesion, and (4) it is only through modifications of those functions that these proteins participate in the genesis of lethal cardiac arrhythmias, or are potentially valuable as targets for antiarrhythmic therapy.

Intercalated Disc Proteins in Inherited and Acquired Diseases

The function of intercalated disc components is relevant not only to normal physiology, but also to the understanding of disease. It is not the purpose of this chapter to review clinical aspects of arrhythmias, but it seems worth mentioning at the outset selected examples where novel findings regarding intercalated disc biology can provide insight into arrhythmia mechanisms.

Initial studies on the relation between desmosome integrity and cardiac electrophysiology were propelled by the finding that most familial cases of arrhythmogenic right ventricular cardiomyopathy (ARVC) in which a genetic link has been found associate with mutations in genes coding for desmosomal molecules.⁸ The latter brought forth the question of how a molecule considered purely relevant to cell adhesion altered the electrical behavior of the heart. The associations between desmosomal proteins and ion channel function (particularly the sodium current I_Na) are extensively reviewed in this chapter. (Recent publications refer to this disease as “arrhythmogenic cardiomyopathy,” to note the occurrence of left ventricular involvement.⁸)

A number of studies have demonstrated remodeling of gap junction proteins (in particular Cx43) in a number of inherited and acquired arrhythmia-related diseases.9–13 Recently, the Fishman laboratory provided evidence that aberrant posttranslational phosphorylation of Cx43 could be the common pathway leading to pathologic gap junction remodeling and arrhythmias.^10,14 Although there is a relationship between Cx43 remodeling and arrhythmias, it is unclear whether these arrhythmias are exclusively consequent to changes in gap junction formation.¹⁵ Recent data show that reduced Cx43 expression alters sodium and potassium current function, and these changes could become part of the arrhythmogenic substrate.^16–18

Disruption of the voltage-gated sodium channel complex is considered an important molecular substrate for arrhythmogenesis. Extensive reviews on the relation between mutations in proteins of the voltage-gated sodium channel (VGSC) complex and arrhythmias are available.¹⁹ Of particular interest in the context of this review is the observation that haploinsufficiency of a desmosomal protein, or overexpression of a mutant desmosomal protein, leads to I_Na deficit and increased arrhythmia susceptibility.^20,21 The latter recalls the previously formulated concept that ARVC and Brugada syndrome (a channelopathy caused by mutations on genes of the VGSC complex) share common features²² and that, as Corrado et al.²³ stated, there is the possibility of “an overlap in clinical manifestation and mechanisms of ventricular arrhythmias between patients with ARVC and Brugada syndrome.”

Structural Features of the Intercalated Disc

In its classic definition, the intercalated disc is composed of three electron-dense structures: adherens junctions, desmosomes, and gap junctions (Figure 22-1). Additional reviews on the characteristics of these structures can be found elsewhere.^24,25 These structures are described briefly in the following sections, and the structural and molecular definitions of the intercalated disc are expanded to include the area composita, the intercellular space, and the nonjunctional ion channel complexes.

Figure 22-1 Transmission electron micrographs of intercalated disc in adult murine heart. Tissue prepared by high-pressure freezing and freeze substitution, thus greatly improving structural preservation. A, Low magnification to demonstrate the entire length of the intercalated disc as it meanders between two cells. B and C, Proximity (and contact in C, yellow arrow) between mitochondria, gap junctions, and desmosomes. D, Example of close proximity between a gap junction and a desmosome. E, Compiled three-dimensional tomographic electron microscopy reconstruction of the intercalated disc. Notice the extensive vesicular activity between the two cells. D, Desmosomes; GJ, gap junction; M, mitochondria. (Reproduced from Delmar M, Liang FX: Connexin43 and the regulation of intercalated disc function. Heart Rhythm 9:835–838, 2012.)

Adherens Junctions

Adherens junctions are specialized structures essential for the mechanical coupling between neighboring cells. The three morphologically different forms of adherens junctions are puncta adherentia, zonula adherens, and fascia adherens, with the last name corresponding to the morphology found in the cardiac intercalated disc.²⁶ Cell-cell mechanical anchoring occurs at two crucial points: the extracellular space, within which cadherins tightly bind to each other, and the intracellular space, within which the cytoplasmic end of cadherin is indirectly attached to the actin cytoskeleton. The association between cadherin and the cytoskeleton involves at least two molecular “hinges”; cadherin binds to β-catenin and plakoglobin, and both molecules in turn bind to α-catenin (among others), the latter being in direct contact with actin. This is only a simplified description, because other interactions are likely to occur.²⁷ This string of intermolecular interactions provides mechanical continuity between cells, allowing for the mechanical work of individual myocytes to integrate into the pumping function of the heart.

Desmosomes

The desmosome (macula adherens) appears as two parallel tripartite plaques containing an intercellular gap of approximately 30 nm bisected by a distinct line, parallel to the apposed cell membranes (see Figure 22-1).²⁸ Desmosomes contribute to mechanical continuity between cardiac cells. Whereas adherens junctions link the actin cytoskeleton of adjacent cells, desmosomes provide continuity to the intermediate filament network (mainly desmin, in the case of heart).^28,29 In the extracellular space, desmosomal cadherins (desmocollins and desmogleins) bind tightly to each other. In the intracellular space, the intermediate filaments bind to desmoplakin. The interaction between desmoplakin and the desmosomal cadherin can be in some cases direct, but it mostly occurs through their association with plakophilin and plakoglobin.^28,29 The topologic organization of desmosomal molecules was studied by North et al.³⁰ using quantitative immunogold electron microscopy. More recently, Al-Amoudi et al.³¹ solved the three-dimensional molecular structure of the desmosomal plaque. Overall, structural and biochemical evidence combined show that desmoplakin binds to plakophilin through their N-terminal domains,^28,32 whereas desmoplakin binds to the intermediate filament by way of its C-terminal domain,^28,31 yielding a highly organized structure.

Hatsell and Cowin ²⁹ once described the desmosome as “a system as staid and solid as the queen’s corsets.”²⁹ With that analogy, it is easy to imagine the loss of containment that would follow in its absence. Mice deficient in plakoglobin, desmoplakin, or plakophilin-2 (PKP2), die during embryonic development as a result of severe myocardial rupture.^33–35 More relevant from the point of view of clinical cardiology, ARVC in humans has been linked to mutations in desmosomal proteins.³⁶ The relation between various complexes of the intercalated disc and ARVC is discussed later in this chapter and in other review articles.^24,37

The Area Composita

Recent immunoelectron microscopy studies revealed the presence of a structure with mixed features of desmosomes and adherens junctions, dubbed the area composita.³⁸ This structure is found only in the heart of higher vertebrate species including mouse and humans.^39,40 The combination of components of the desmosomes and the adherens junctions allows anchorage of actin and desmin filaments to the same point, perhaps providing additional strength and flexibility to the muscle cell. Knockdown of PKP2 in neonatal cardiomyocytes leads to remodeling of the area composita.⁴¹ Additional studies suggest a role for α-catenin in the maintenance of these hybrid junctions.⁴² Interestingly, knockdown of α-catenin leads to PKP2 decrease only at the area composita and not at the desmosomes, suggesting that the molecular composition of the area composita, and its regulation, can be independent from that of other junctions.⁴² The area composita may represent a physical space where mechanical junction proteins interact with ion channel complexes.

Transcriptional Regulation by Mechanical Junction Proteins

Catenins are a part of both adherens junctions and desmosomes, thus participating in cell adhesion; however, these proteins also act as transcriptional activators. A prominent example is the participation of β-catenin in canonical Wnt signaling.⁴³ Binding of Wnt to its Frizzled receptor leads to an increase in levels of cytosolic β-catenin and a consequent translocation of the protein to the nucleus, where it binds to Tcf/Lef complex and promotes the expression of various genes such as c-Myc or c-Fos. In the heart, the Wnt signaling and the activation of genes by β-catenin/Tcf/Lef has been associated with the regulation of physiologic and pathologic growth of the cardiomyocytes.⁴⁴ Plakoglobin, another protein of the armadillo family, shows high homology with β-catenin and has been associated with the Wnt signaling. Different studies have shown that plakoglobin interacts and competes with β-catenin at multiple levels, acting as an antagonist of the Wnt/β-catenin signaling.^45–47 The fact that desmosomal proteins are involved in the regulation of the Tcf/Lef complex has been invoked as a possible mechanism for the fibrofatty infiltration common in hearts affected with ARVC.^48–50

Gap Junctions

In 1958, Sjostrand et al.⁴ described an area of specialization in the cardiac intercalated disc composed of “three dark lines with two intervening less dense lines”. This structure, which was similar to the one previously identified in the giant axon of the crayfish, was named the “longitudinal connexion” by these investigators. Years later, Revel coined the term gap junctions, thus emphasizing two key features: a gap between the cells and a junction between them.

Gap junctions form intercellular channels that provide a low-resistance pathway for direct cell-to-cell passage of electrical charge between cardiac myocytes. Each gap junction channel is composed of two hexameric structures called connexons that dock across the extracellular space and form a permeable pore isolated from the extracellular space. Each connexon results from oligomerization of an integral membrane protein, connexin. The most abundant connexin isotype in the heart, brain and other tissues is the 43-kD protein, connexin43 (Cx43).

The importance of Cx43 in the propagation of the cardiac action potential is well established. If Cx43 channels are not present, normal propagation is disrupted and lethal arrhythmias can ensue.51,52 Gap junction remodeling has been studied for various inherited and acquired diseases.^{10,11,14,53–55} The implications of connexin remodeling to electrophysiology are discussed later in this chapter.

Intercellular Space

The size of the space separating two cardiac cells at the intercalated disc changes depending on the proximity to the various structures, as well as the vesicular activity between the two cells (see Figure 22-1). It is a common view that the intercellular space is not relevant for electrophysiology. This view, however, is changing. Mathematical modeling studies^56,57 and experimental evidence⁵⁸ support the idea that the intercellular space is critical to propagation via an electric field mechanism.⁵⁹ This model will be discussed later. Of note, increased size of the intercellular space has been reported in animal models of ARVC.^20,21,60,61

Ion Channel Complexes That Reside at the Intercalated Disc

Voltage-Gated Sodium Channel Complex

In 1996, Cohen ⁶² showed that cardiac sodium channel proteins were preferentially localized at the intercalated disc,⁶² although they are also present over the cell surface, following a striated pattern.⁶³ Recent data emphasize how the VGSC interacts also with scaffolding, anchoring, and adhesions proteins, which regulate its function.

α-Subunit Na_V1.5

The cardiac, pore-forming, α-subunit Na_V1.5 contains four homologous transmembrane domains DI-DIV, linked by intracellular loops (IDI-II, IDII-III, IDIII-IV). Each domain is formed by the six transmembrane segments S1 to S6, and is involved in the voltage-dependent activation of the channel. The channel pore conducting Na⁺ is lined by the S6 segment and the S5-S6 pore loops in each domain.¹⁹ The inactivation gate is a complex formed by the DIII-IV loop and the C-terminus.⁶⁴ This channel is defined as TTX-resistant, in contrast with other Nav channels. Of note, TTX-sensitive channel proteins, and currents, have also been found in cardiac myocytes, with the current localized primarily to the midsection of the cell.^63,65

The SCN5A gene, on chromosome 3p21, codes for the Na_V1.5 subunit. Mutations resulting in increased I_Na are associated with long QT syndrome type 3. Mutations that cause a reduction in I_Na are responsible for a series of different diseases, such as Brugada syndrome, progressive cardiac conduction defect, sick sinus syndrome, and a form of inherited atrial fibrillation. Some mutations are associated with a clinical spectrum encompassing more than one of those phenotypes and can manifest differently among carriers, even within the same family.¹⁹ Interestingly, some SCN5A mutations have been linked to a form of dilated cardiomyopathy with a high incidence of atrial and ventricular arrhythmias.⁶⁶ It is not yet clear how mutations in the sodium channel could lead to structural damage of the myocardium. Based on the crosstalk between intercalated disc structures described in this chapter, we are tempted to speculate that the integrity of the sodium channel complex is also relevant to intercellular adhesion strength.^20,21,67

β-Subunits of the Sodium Channel

The Na_V β-subunit family consists of four proteins: β1-4, coded by genes SCN1B to SCN4B, respectively. These are single-span transmembrane proteins oriented with the amino terminus facing the extracellular space. The extracellular domain presents a conserved immunoglobulin domain, homologous to the one in cell adhesion molecules.⁶⁸ The carboxyl terminus associates with cytoskeletal and scaffolding proteins. β1 and β2 are localized at the T-tubules–Z lines and at the intercalated discs in rat cardiac myocytes. β3 colocalizes with β1 at the T-tubules and β4 colocalizes with β2 at the intercalated disc. β1 and β2 associate with ankyrin-G and ankyrin-B in both brain and heart, and their interaction is critical for channel surface expression and modulates the channel function in vivo. β1 and β2 associate with N-cadherin and with Cx43 at the intercalated disc.⁶⁹ Altogether, β-subunits have a key role in the interactions between the VGSC multiple proteins at the intercalated disc, including those relevant for cell adhesion and for electrical coupling between cells.⁶⁹ Furthermore, work from the Isom lab has demonstrated that null mice for the β1 subunit show a significant increase in SCN5A mRNA in cardiac myocytes.⁷⁰ Further research is needed to elucidate the role of β-subunits in transcription regulation in the heart.⁷⁰ Overall, the data show that, as in the case of catenin and plakoglobin, β-subunits can have distal, contact-dependent effects, including regulation of gene transcription, with consequences to the function and structure of the heart.

Subcellular Heterogeneity of Voltage-Gated Sodium Channel

Recent studies have shown that not all sodium channels on the surface of the cardiac myocyte are equal. Instead, the molecular composition and the function of a sodium channel are different depending on whether Na_V1.5 is localized to the intercalated disc or to the midsection of the cell. Petiprez et al.⁷¹ described two separate pools of VGSC in ventricular cardiomyocytes. One of these two subpopulations localizes at the lateral membrane of the myocytes, where Na_V1.5 interacts with dystrophin and syntrophin; the second subpopulation of VGSC localizes at the ID, where Na_V1.5 interacts with the MAGUK proteins SAP97 and ZO1, as well as with PKP2.⁷² Lin et al. recently demonstrated that this subcellular distribution correlates with differences in function.⁶⁵ Using cell-attached macropatch recordings, these authors showed that the magnitude of I_Na is larger at the intercalated disc than in the midsection of the cell. Furthermore, TTX-resistant channels in the midsection showed a significant negative shift in the steady-state inactivation curve, suggesting that these channels are mostly in an inactivated state at the normal resting potential, and that the burden of excitation is on the channels at the intercalated disc. These authors also demonstrated that the amplitude of current is larger if cells remain paired, strongly suggesting that cell adhesion preserves sodium channel function. The interaction between sodium channels and proteins of the intercalated disc is discussed extensively later.

Potassium Channels at the Intercalated Disc

In 1995, Mays et al.⁷³ described for the first time that the potassium channel protein K_V1.5 localizes to the intercalated disc of adult myocardial cells. Later studies showed that K_V1.5 associates with SAP97.⁷⁴ Interestingly, SAP-97 also associates with Na_V1.5,⁷¹ making this protein a candidate for mutual regulation of both depolarization and repolarization channels, as is the case of Na_V1.5 and Kir2.1.⁷⁵ Additional studies have shown that the function of K_V1.5 depends on the expression of N-cadherin,⁷⁶ an interesting parallelism to the interaction between PKP2 (a desmosomal molecule), and Na_V1.5.⁷² Cheng et al.⁷⁶ also showed that cortactin is required for the N-cadherin–dependent regulation of K_V1.5. Of note, mice deficient in kcne2, an ancillary potassium channel subunit, display an impaired ventricular repolarization because of inhibition in the trafficking to the membrane of K_V1.5.⁷⁷