Structure and Physiology

The cardiac Na⁺ channel complex is composed of a primary α and multiple ancillary β subunits. The approximately 2000-amino-acid α subunit contains the channel’s ion-conducting pore and controls the channel selectivity for Na⁺ ions and voltage-dependent gating machinery. This subunit contains all the drug and toxin interaction sites identified to date. The α subunit (Na_v1.5), encoded by the SCN5A gene, consists of four internally homologous domains (I to IV) that are connected to each other by cytoplasmic linkers (Fig. 2-1). Each domain consists of six membrane-spanning segments (S1 to S6), connected to each other by alternating intracellular and extracellular peptide loops. The four domains are arranged in a fourfold circular symmetry to form the channel. The extracellular loops between S5 and S6 (termed the P segments) have a unique primary structure in each domain (Fig. 2-2). The P segments curve back into the membrane to form an ion-conducting central pore whose structural constituents determine the selectivity and conductance properties of the Na⁺ channel.²

FIGURE 2-1 The sodium channel macromolecular complex. See text for discussion.

(From Boussy T, Paparella G, de Asmundis C, et al: Genetic basis of ventricular arrhythmias. Heart Fail Clin 6:249-266, 2010.)

FIGURE 2-2 Transmembrane organization of sodium channel subunits. The primary structures of the subunits of the voltage-gated ion channels are illustrated as transmembrane folding diagrams. Cylinders represent probable α-helical segments: S1 to S3, blue; S4, green; S5, orange; S6, purple; outer pore loop, shaded orange area. Bold lines represent the polypeptide chains of each subunit, with length approximately proportional to the number of amino acid residues in the brain sodium channel subtypes. The extracellular domains of the β₁ and β₂ subunits are shown as immunoglobulin-like folds. Ψ shows sites of probable N-linked glycosylation. P represents sites of demonstrated protein phosphorylation by protein kinase A (red circles) and protein kinase C (red diamonds); h in the blue circle signifies an inactivation particle in the inactivation gate loop; the empty blue circles represent sites implicated in forming the inactivation gate receptor. The structure of the extracellular domain of the β subunits is illustrated as an immunoglobulin-like fold based on amino acid sequence homology to the myelin P0 protein. Sites of binding of α and β scorpion toxins (α-ScTx, β-ScTx) and a site of interaction between α and β₁ subunits also are shown. H₃N and NH₃ = ammonia.

(From Caterall WA, Maier SK: Voltage-gated sodium channels and electrical excitability of the heart. In Zipes DP, Jalife J, editors: Cardiac electrophysiology: from cell to bedside, ed 5, Philadelphia, 2009, Saunders, pp 9-17.)

Four auxiliary β subunits (Na_vβ1 to Na_vβ4, encoded by the genes SCN1B to SCN4B, respectively) have been identified; each is a glycoprotein with a single membrane-spanning segment. The β₁ subunit likely plays a role in modulation of the gating properties and level of expression of the Na⁺ channel.²

Na⁺ channels are the typical example of voltage-gated ion channels. Na⁺ channels switch among three functional states: deactivated (closed), activated (open), and inactivated (closed), depending on the membrane potential (E_m). These channel states control Na⁺ ion permeability through the channel into the cardiomyocyte. Na⁺ channel activation allows Na⁺ ion influx into the cell, and inactivation blocks entry of Na⁺ ions.²

On excitation of the cardiomyocyte by electrical stimuli from adjacent cells, its resting E_m (approximately −85 mV) depolarizes. The positively charged S4 segment of each domain of the α subunit functions as the sensor of the transmembrane voltage; these segments are believed to undergo rapid structural conformational changes in response to membrane depolarization, thus leading to channel opening (activation) from its resting (closed) state and enabling a large and rapid influx of Na⁺ (inward Na⁺ current [I_Na]) during the rapid upstroke (phase 0) of the action potential in atrial, ventricular, and Purkinje cardiomyocytes.³

Normally, activation of Na⁺ channels is transient; fast inactivation (closing of the pore) starts simultaneously with activation, but because inactivation is slightly delayed relative to activation, the channels remain transiently open to conduct I_Na during phase 0 of the action potential before it closes. Each Na⁺ channel opens very briefly (less than 1 millisecond) during phase 0 of the action potential; collectively, activation of the channel lasts a few milliseconds and is followed by fast inactivation.

Na⁺ channel inactivation comprises different conformational states, including fast, intermediate, and slow inactivation. Fast inactivation is at least partly mediated by rapid occlusion of the inner mouth of the pore by the cytoplasmic interdomain linker between domains III and IV of the α subunit, which has a triplet of hydrophobic residues that likely functions as a hinged “latch” that limits or restricts Na⁺ ion pass through the pore. The carboxyl terminus (C-terminus) also plays an important role in the control of Na⁺ channel inactivation and stabilizing the channels in the inactivated state by interacting with the loop linking domains III and IV. Importantly, although most Na⁺ channels open before inactivating, some actually inactivate without ever opening (a process known as closed-state inactivation).^1,4

Once inactivated, Na⁺ channels do not conduct any more current and cannot be reactivated (reopened) until after recovery from inactivation. The recovery of the Na⁺ channel to reopen is voltage dependent. Channel inactivation is removed when the E_m of the cell repolarizes during phase 4 of the action potential. Membrane repolarization is facilitated by the fast inactivation of the Na⁺ channels (limiting the inward current) and is augmented by activation of voltage-gated K⁺ channels (allowing the outward current). The recovery of channels from inactivation is also time dependent; Na⁺ channels typically activate within 0.2 to 0.3 milliseconds and inactivate completely within 2 to 5 milliseconds.

Following recovery, Na⁺ channels enter a closed state that represents a nonconducting conformation, which allows the channels to be activated again during the next action potential. The fraction of channels available for opening varies from almost 100% at −90 mV and 50% at −75 mV to almost 0% at +40 mV. Consequently, highly polarized (−80 to −90 mV) cell membranes can be depolarized rapidly by stimuli because more Na⁺ channels reopen, whereas partially depolarized cells with potentials close to threshold −70 mV generate a much slower upstroke because of the inactivation of a proportion of Na⁺ channels. Given that Na⁺ channels are major determinants of conduction velocity, this velocity generally slows at a reduced E_m.

Na⁺ channel activation, inactivation, and recovery from inactivation occur within a few milliseconds. At the end of phase 1 of the action potential, more than 99% of Na⁺ channels transit from an open (activated) state to an inactivated state. However, a very few Na⁺ channels are not inactivated and may reactivate (reopen) during action potential phase 3. The small current produced by these channels (less than 1% of the peak I_Na) is called the “window” current because it arises when the sarcolemma reaches a potential that is depolarized sufficiently to reactivate some channels, but not enough to cause complete inactivation. The voltage range for the window current is very restricted and narrow in healthy hearts, thus granting it a small role during the cardiac action potential.

In addition to these rapid gating transitions, Na⁺ channels are also susceptible to slower inactivating processes (slow inactivation) if the membrane remains depolarized for a longer time. These slower events can contribute to the availability of active channels under various physiological conditions. Whereas fast-inactivated Na⁺ channels recover rapidly (within 10 milliseconds) during the hyperpolarized interval between stimuli, slow inactivation requires much longer recovery times (ranging from hundreds of milliseconds to many seconds). The molecular movements leading to slow inactivation are less well understood. The P segments seem to play a key role in slow inactivation.⁴

Some Na⁺ channels occasionally show alternative gating modes consisting of isolated brief openings occurring after variable and prolonged latencies and bursts of openings during which the channel opens repetitively for hundreds of milliseconds. The isolated brief openings are the result of the occasional return from the inactivated state. The bursts of openings are the result of occasional failure of inactivation.¹ Prolonged opening or reopening of some Na⁺ channels during phases 2 and 3 can result in a small late I_Na (I_NaL). Despite its minor contribution in healthy hearts, I_NaL can potentially play an important role in diseased hearts.³

Function

Na⁺ channels play a pivotal role in the initiation, propagation, and maintenance of the normal cardiac rhythm. The I_Na determines excitability and conduction in atrial, His-Purkinje system (HPS), and ventricular myocardium. On membrane depolarization, the voltage-gated Na⁺ channels respond within a millisecond by opening, thus leading to the very rapid depolarization of the cardiac cell membrane (phase 0 of the action potential), reflected by the fast (within tenths of a microsecond) subsequent opening of Na⁺ channels triggering the excitation-contraction coupling. Na⁺ entry during phase 0 of the action potential also modulates intracellular Na⁺ levels and, through Na⁺-Ca²⁺ exchange, intracellular Ca²⁺ concentration and cell contraction.

The cardiac Na⁺ channel also plays a crucial role in the propagation of action potentials throughout the atrium, HPS, and ventricles. The opening of Na⁺ channels in the atria underlies the P wave on the ECG, and in the ventricles I_Na underlies the QRS complex and enables a synchronous ventricular contraction. Because the upstroke of the electrical potential primarily determines the speed of conduction between adjacent cells, Na⁺ channels are present in abundance in tissues where speed is of importance. Cardiac Purkinje cells contain up to 1 million Na⁺ channels, a finding that illustrates the importance of rapid conductance in the heart.²

Na⁺ channels also make a contribution in the plateau phase (phase 2) and help determine the duration of the action potential. After phase 0 of the action potential, I_Na decreases to less than 1% of its peak value over the next several milliseconds because of voltage-dependent inactivation. This persistent or “late” inward I_Na (I_NaL), along with the L-type Ca²⁺ current (I_CaL), helps maintain the action potential plateau.⁵

Furthermore, Na⁺ channel inactivation is very important as it prevents cells from being prematurely reexcited because of the unavailability of the voltage-gated Na⁺ channels. With repolarization, the Na⁺ channel normally recovers rapidly from inactivation (within 10 milliseconds) and is ready to open again. Hence, Na⁺ channels help to determine the frequency of action potential firing. To a lesser extent, cardiac Na⁺ channels are also present in the sinus node and the atrioventricular node (AVN), where they contribute to pacemaker activity.

Regulation

The regulatory proteins interacting with Na_v1.5 may be classified as follows: (1) anchoring-adaptor proteins (e.g., ankyrin-G, syntrophin proteins, multicopy suppressor of gsp1 [MOG1]), which play roles only in trafficking and targeting the channel protein in specific membrane compartments; (2) enzymes interacting with and modifying the channel structure (post-translational modifications), such as protein kinases or ubiquitin ligases; and (3) proteins modulating the biophysical properties of Na_v1.5 on binding (e.g., caveolin-3, calmodulin, glycerol 3-phosphate dehydrogenase 1–like [G3PD1L], telethonin, Plakophilin-2).⁵ Coexpression of Na_v1.5 with its β subunits induces acceleration in the recovery from inactivation and enhancement of I_Na amplitude.

The cardiac Na⁺ channels are subject to phosphorylation and dephosphorylation by kinases or phosphatases. The intracellular linker between domains I and II contains eight consensus sites for cyclic adenosine monophosphate (cAMP)–dependent protein kinase A (PKA) phosphorylation. cAMP-dependent PKA and G protein stimulatory α subunit (Gsα) modulate the function of expressed cardiac Na⁺ channels on β-adrenergic stimulation and enhance I_Na.¹

In contrast, activation of α-adrenergic stimulating protein kinase C (PKC) results in the reduction of I_Na. The effect of PKC is largely attributable to phosphorylation of a highly conserved serine in the linker between domains III and IV. PKC reduces the maximal conductance of the channels and alters gating. Na⁺ channels exhibit a hyperpolarizing shift in the steady-state availability curve, suggesting an enhancement of inactivation from closed states.

All subunits of the Na⁺ channel are modified by glycosylation. The β₁ and β₂ subunits are heavily glycosylated, with up to 40% of the mass being carbohydrate. In contrast, the cardiac α subunit is only 5% sugar by weight. Sialic acid is a prominent component of the N-linked carbohydrate of the Na⁺ channel. The addition of such a highly charged carbohydrate has predictable effects on the voltage dependence of gating through alteration of the surface charge of the channel protein.¹

Pharmacology

Na⁺ channels are the targets for the action of class I antiarrhythmic drugs. Na⁺ channel blockers bind to a specific receptor within the channel’s pore. The binding blocks ion movement through the pore and stabilizes the inactivated state of Na⁺ channels. Blockade of Na⁺ channels tends to decrease tissue excitability and conduction velocity (by attenuating peak I_Na) and can shorten action potential duration (by attenuating late I_Na).^1,6

One important component in the action of antiarrhythmic drugs is a voltage-dependent change in the affinity of the drug-binding site (i.e., the channel is a modulated receptor). Additionally, restricted access to binding sites can contribute to drug action, a phenomenon that has been called the guarded receptor model. Open and inactivated channels are more susceptible to block than resting channels, likely because of a difference in binding affinity or state-dependent access to the binding site. Consequently, binding of antiarrhythmic drug occurs primarily during the action potential (known as use-dependent block), and the block dissipates after repolarization (i.e., in the interval between action potentials). When the time interval between depolarizations is insufficient for block to recover before the next depolarization occurs (secondary to either abbreviation of the interval between action potentials during fast heart rates or slow kinetics of the unbinding of the Na⁺ channel blocker), block of Na⁺ channels accumulates (resulting in an increased number of blocked channels and enhanced blockade).⁶ A drug with rapid kinetics produces less channel block with the subsequent depolarization than does a drug with slower recovery. Use-dependent block is important for the action of antiarrhythmic drugs because it allows strong drug effects during fast heart rates associated with tachyarrhythmias but limits Na⁺ channel block during normal heart rates. Importantly, drug recovery kinetics can potentially be slowed by pathophysiological conditions such as membrane depolarization, ischemia, and acidosis.¹ This property is known as use-dependence and is seen most frequently with the class IC agents, less frequently with the class IA drugs, and rarely with the class IB agents.

Class I antiarrhythmic drugs can be classified into three groups according to rates of drug binding to and dissociation from the channel receptor. Class IC drugs (flecainide and propafenone) block both the open and inactivated state (which is induced by depolarization) Na⁺ channels and have the slowest kinetics of unbinding during diastole. The results are prolongation of conduction at normal heart rates and a further increase in the effect at more rapid rate (use-dependence).

The class IB agents (lidocaine, mexiletine, and tocainide) block both open and inactivated Na⁺ channels and dissociate from the channel more rapidly than do other class I drugs. As a consequence, class IB drugs exhibit minimal or no effects on the Na⁺ channels in normal tissue but cause significant conduction slowing in depolarized tissue, especially at faster depolarization rates. Furthermore, class IB drugs are less effective in the atrium, where the action potential duration is so short that the Na⁺ channel is in the inactivated state only briefly compared with the relatively long diastolic recovery times; thus, accumulation of block is less likely to result from the rapid recovery of block.

Class IA drugs (quinidine, procainamide, and disopyramide) exhibit open state block, have intermediate effects on Na⁺ channels, and generally only cause significant prolongation of conduction in cardiac tissue at rapid heart rates. Because the open state block is dominant and recovery from block is slow, these drugs are effective in both the atrium (where action potential duration is short) and the ventricle (where the action potential duration is long).

The late I_Na (I_NaL) also can be a target for blockade. Several drugs exhibit relative selectivity for block of late I_NaL over peak I_Na, including mexiletine, flecainide, lidocaine, amiodarone, and ranolazine.

Importantly, class IA drugs also have moderate K⁺ channel blocking activity (which tends to slow the rate of repolarization and prolong the action potential duration) and anticholinergic activity, and they tend to depress myocardial contractility. At slower heart rates, when use-dependent blockade of I_Na is not significant, K⁺ channel blockade becomes predominant (reverse use-dependence), leading to prolongation of the action potential duration and QT interval and increased automaticity. Flecainide and propafenone also have K⁺ channel blocking activity and can increase the action potential duration in ventricular myocytes. Propafenone has significant β-adrenergic blocking activity.

Inherited Channelopathies

Mutations in genes that encode various subunits of the cardiac Na⁺ channel or proteins involved in regulation of the inward I_Na have been linked to several types of electrical disorders (Table 2-1). Depending on the mutation, the consequence is either a gain of channel function (with consequent prolongation of action potential duration because more positive ions accumulate in the cell) or an overall loss of channel function that influences the initial depolarizing phase of the action potential (with consequent decrease in cardiac excitability and electrical conduction velocity). It is noteworthy that a single mutation can cause different phenotypes or combinations thereof.²

Acquired Diseases

In heart failure, peak I_Na is reduced (likely secondary to reduced SCN5A expression), whereas I_NaL is increased (likely because of increased phosphorylation of Na⁺ channels). Na_v1.5 expression is reduced in the surviving myocytes in the border zone of the myocardial infarct (MI). Importantly, Na⁺ channel blockers can increase the risk for SCD in patients with ischemic heart disease, possibly by facilitating the initiation of reentrant excitation waves. Additionally, I_NaL increases during myocardial ischemia, explaining why I_NaL inhibition may be an effective therapy for chronic stable angina. Na_v1.5 expression is reduced in response to persistent atrial tachyarrhythmias as part of the “electrical remodeling” process, leading to attenuation of I_Na.³

Furthermore, mutations in SCN5A can predispose affected individuals to acquired LQTS induced by a variety of drugs such as antihistamines or antibiotics. These mutations result in changes in channel activity that exert a significant impact on action potential duration only when combined with drug-induced alteration of other channels.

Structure and Physiology

Cardiac K⁺ channels are membrane-spanning proteins that allow the passive movement of K⁺ ions across the cell membrane along its electrochemical gradient. The ion-conducting or pore-forming subunit is generally referred to as the α subunit. The tripeptide sequence glycine-tyrosine-glycine GYG is common to the pore of all K⁺ channels and constitutes the signature motif for determining K⁺ ion selectivity. A gating mechanism controls switching between open-conducting and closed-nonconducting states.^1,15

K⁺ channels represent the most diverse class of cardiac ion channels (Fig. 2-3). Cardiac K⁺ currents can be categorized as voltage-gated (K_v) and ligand-gated channels. In K_v channels, pore opening is coupled to the movement of a voltage sensor within the membrane electric field, and they include the rapidly activating and inactivating transient outward current (I_to); the ultrarapid (I_Kur), rapid (I_Kr), and slow (I_Ks) components of the delayed rectifier current; and the inward rectifier current (I_K1). In contrast, pore opening in ligand-gated channels is coupled to the binding of an organic molecule, including channels activated by a decrease in the intracellular concentration of adenosine triphosphate (K_ATP) or activated by acetylcholine (K_ACh). Other classes of K⁺ channels respond to different stimuli, including changes in intracellular Ca²⁺ concentration and G proteins.¹⁵

FIGURE 2-3 Molecular compositions of cardiac potassium (K⁺) channels. Amino termini (N) and carboxyl termini (C) are indicated. A, Voltage-gated (K_v), inward rectifier (Kir), and two-pore α subunits are integral membrane proteins with six, two, and four membrane-spanning domains, respectively. B, These α subunits assemble as tetramers or dimers to form K⁺-selective pores.

(Modified from Oudit GY, Backx PH: Voltage-regulated potassium channels. In Zipes DP, Jalife J, editors: Cardiac electrophysiology: from cell to bedside, ed 5, Philadelphia, 2009, Saunders, pp 29-42.)

On the basis of the primary amino acid sequence of the α subunit, K⁺ channels have been classified into three major families (Table 2-2):

Each voltage-gated K⁺ channel (K_v family) is formed by the coassembly of 4 identical (homotetramers) or a combination of 4 different (from the same subfamily, heterotetramers) α subunits. A total of 38 genes has been cloned and assigned to 12 subfamilies of voltage-gated K channels (K_v1 to K_v12) on the basis of sequence similarities. Each α subunit contains one domain consisting of 6 membrane-spanning segments (S1 to S6), connected to each other by alternating intracellular and extracellular peptide loops (similar to 1 of the 4 domains of voltage-gated Na⁺ and Ca²⁺ channels), with both the amino terminus (N-terminus) and the C-terminus located on the intracellular side of the membrane. The central ion-conducting pore region is formed by the S5 and S6 segments and the S5-S6 linker (P segment); the S5-S6 linker is responsible for K⁺ ion selectivity. The S4 segment serves as the voltage sensor.^1,15

The α subunits of K_v channels can generate voltage dependent K⁺ current when expressed in heterologous systems. However, the assembly of a functional tetramer can occur only in the presence of multiple auxiliary units (see Table 2-2). In many cases, auxiliary subunits coassociate with the α subunits and likely modulate cell surface expression, gating kinetics, and drug sensitivity of the α subunit complex. Most K⁺ channel β subunits assemble with α subunits and give rise to an α₄β₄ complex. K⁺ channel β subunits represent a diverse molecular group, which includes cytoplasmic proteins (K_vβ1 to K_vβ3, KChIP, and KChAP) that interact with the intracellular domains of K_v channels; single transmembrane spanning proteins (e.g., minK and minK-related proteins [MiRPs]) encoded by the KCNE gene family; and large ATP-binding cassette (ABC) transport-related proteins (e.g., the sulfonylurea receptors [SURs]).^1,16

The diversity of K⁺ currents in native tissues exceeds the number of K⁺ channel genes identified. The explanations for this diversity include alternative splicing of gene products, post-translational modification, and heterologous assembly of β subunits within the same family and assembly with accessory β subunits that modulate channel properties.

As with voltage-dependent Na⁺ and Ca²⁺ channels, K_v channels typically fluctuate among distinct conformational states because of molecular movements in response to voltage changes across the cell membrane (voltage-dependent gating). The K_v channel activates (opens) on membrane depolarization, thus allowing the rapid passage of K⁺ ions across the sarcolemma. After opening, the channel undergoes conformational transition in a time-dependent manner to a stable nonconducting (inactivated) state. Inactivated channels are incapable of reopening, even if the transmembrane voltage is favorable, unless they “recover” from inactivation (i.e., enter the closed state) on membrane repolarization. Closed channels are nonconducting but can be activated on membrane depolarization.¹⁷

Three mechanistically distinct types of K_v channel inactivation that are associated with distinct molecular domains have been identified: N-type, C-type, and U-type. N-type (“ball and chain”) inactivation involves physical occlusion of the intracellular mouth of the channel pore through binding of a small group of amino acids (“inactivation ball tethered to a chain”) at the extreme N-terminus. In contrast, C-type inactivation involves conformational changes in the external mouth of the pore. C-type inactivation exists in almost all K⁺ channels and may reflect a slow constriction of the pore. This inactivation process is thought to be voltage independent, coupled to channel opening, and is usually slower than N-type inactivation. Recovery from C-type inactivation is relatively slow and weakly voltage dependent. Importantly, the rate of C-type inactivation and recovery can be strongly influenced by other factors, such as N-type inactivation, drug binding, and changes in extracellular K⁺ concentration. These interactions render C-type inactivation an important biophysical process in regulating repetitive electrical activity and determining certain physiological properties such as refractoriness, drug binding, and sensitivity to extracellular K⁺.¹⁷

In addition to N-type and C-type inactivation, some K_v channels also show another type of inactivation (U-type), which exhibits a U-shaped voltage dependence with prolonged stimulation rates. Those channels appear to exhibit preferential inactivation from partially activated closed states, rapid and strongly voltage-dependent recovery from inactivation and, in some channel types, accelerated inactivation with elevation of extracellular K⁺. The exact conformational changes underlying U-type inactivation remain unclear. Importantly, there is extreme diversity in the kinetic and potentially molecular properties of K_v channel inactivation, particularly of C-type inactivation.¹⁷

Function

K⁺ channels are a diverse and ubiquitous group of membrane proteins that regulate K⁺ ion flow across the cell membrane on the electrochemical gradient and regulate the resting E_m, the frequency of pacemaker cells, and the shape and duration of the cardiac action potential. Because the concentration of K⁺ ions outside the cell membrane is approximately 25-fold lower than that in the intracellular fluid, on activation, the opening of K⁺ channels generates an outward current resulting from the efflux of positively charged ions that offers a mechanism to counteract, dampen, or restrict the depolarization front (phases 1 through 4 of the action potential) triggered by an influx of cations (Na⁺ and Ca²⁺).^1,3,15,16

The variation in the level of expression of K⁺ channels that participate in the genesis of the cardiac action potential explains the regional differences of the configuration and duration of cardiac action potentials from sinus node and atrial to ventricular myocytes and across the myocardial wall (endocardium, midmyocardium, and epicardium). Moreover, the expression and properties of K⁺ channels are not static; heart rate, neurohumoral state, pharmacological agents, cardiovascular diseases (cardiac hypertrophy and failure, MI) and arrhythmias (e.g., AF) can influence those properties, and they underlie the change in action potential configuration in response to variation in heart rate and various physiological and pathological conditions.^1,15,16

Transient Outward Potassium Current (I_to)

Ultrarapidly Activating Delayed Outward Rectifying Current (I_Kur)

Rapidly Activating Delayed Outward Rectifying Current (I_Kr)

Slowly Activating Delayed Outward Rectifying Current (I_Ks)

Inward Rectifying Current (I_K1)

Acetylcholine-Activated Potassium Current (I_KACh)

ATP-Dependent Potassium Current (I_KATP)

Two-Pore Potassium Channels (K_2P)

Structure and Physiology

In cardiac muscle, two types of voltage-dependent Ca²⁺ channels, the L-type and the T-type, transport Ca²⁺ into the cells. The L-type channel (L for long-lasting, because of its slow kinetics of current decay as compared with Na⁺ channels) is found in all cardiac cell types. The T-type channel (T for tiny and transient) is found principally in pacemaker, atrial, and Purkinje cells. The term Ca²⁺ channels is used to refer to the L-type channel.^46–48

Cardiac L-type Ca²⁺ channels are composed of four polypeptide subunits (α_1C, β, α₂, delta) and form a heterotetrameric complex. The α_1C subunit (Ca_v1.2, encoded by the CACNAIC gene) has a structure similar to that of the Na⁺ channel: four homologous domains (I to IV), each consisting of six transmembrane segments (S1 to S6). The S5 and S6 segments and the membrane-associated pore loop (P loop) between them form the central pore through which ions flow down their electrochemical gradient. The P loop contains four negatively charged glutamate residues (EEEE) that are required for the Ca²⁺ selectivity of the channel. S4 in each homologous domain contains a highly conserved positively charged residue (arginine or lysine) at every third or fourth position. This segment serves as the voltage sensor for gating.^46,47,49

The α_1C subunit is the main and largest subunit, and it determines most of the channel characteristics because it harbors the ion-selective pore, voltage sensor, gating apparatus, and binding sites for channel-modulating drugs and is autoregulatory. To form a functional L-type Ca²⁺ channel, the α_1C subunit coassembles with auxiliary subunits in a 1:1:1 ratio: the β subunit, the α₂ subunit, and the delta subunit.^46,47,49

The β subunit (Ca_vβ, encoded by the CACNB gene) is entirely intracellular and is tightly bound to a highly conserved motif in the cytoplasmic linker between domains I and II of the α_1C subunit. Coexpression of β subunits modulates the biophysical properties of the α_1C subunit. The β subunit has a prominent role in channel expression, trafficking, regulation, and facilitation. Sites possible for phosphorylation by various protein kinases (PKA, PKC, protein kinase G [PKG]) have been identified in these subunits. The β subunits are also involved in channel regulation by β-adrenergic stimulation and in response to the changes of the pH of the cell. In addition, the β subunit increases Ca²⁺ current amplitude, accelerates the kinetics of Ca²⁺ channel activation, and alters pharmacological properties of the channel.^46,47

The α₂ and delta subunits are encoded by the same gene (CACNA2D); the mature forms of these subunits are derived by post-translational proteolytic cleavage, but they remain associated through a disulfide bond. The α₂ subunit is completely extracellular, whereas the delta subunit has a single membrane-spanning segment with a very short intracellular part that anchors the α₂-delta subunit complex to the α_1C subunit. The α₂-delta subunit complex has less influence on channel function than the β subunit. The α₂-delta subunit slightly increases Ca²⁺ current amplitude and accelerates channel inactivation and can change the properties of Ca²⁺ channel activation. It can also affect channel density trafficking.^46,47

L-type Ca²⁺ channels are characterized by a large single channel conductance. The channels are closed at the resting potential, but they activate on depolarization to potentials positive to –40 mV. I_CaL peaks at 0 to + 10 mV, and tends to reverse at +60 to +70 mV, following a bell-shaped current-voltage relationship.

Although I_CaL is normally activated during phase 0 by the regenerative depolarization caused by the fast I_Na, I_CaL is much smaller than the peak I_Na. In addition, the amplitude of I_CaL is not maximal near the action potential peak because of the time-dependent nature of I_CaL activation, as well as the low driving force (E_m – reversal potential of a cardiac Ca²⁺ channel [E_Ca]) for I_CaL.⁵⁰

The decay of I_CaL during depolarization (i.e., time-dependent inactivation) is very slow and depends on two mechanisms: voltage-dependent inactivation and Ca²⁺-dependent inactivation. These two mechanisms control Ca²⁺ influx into cardiomyocytes and hence regulate signal transduction to sarcoplasmic reticulum Ca²⁺ channels (ryanodine receptor 2 [RyR2]) and ensure normal contraction and relaxation of the heart.^46–48

Fast Ca²⁺-dependent inactivation serves as a negative feedback for Ca²⁺ to limit further Ca²⁺ entry via L-type Ca²⁺ channels. The slow voltage-dependent inactivation (induced by membrane depolarization) prevents a premature rise in I_CaL when intracellular Ca²⁺ concentration decreases and Ca²⁺-dependent inactivation terminates during maintained depolarization. Although still under dispute, the relative contribution of Ca²⁺-dependent inactivation to total inactivation of I_CaL appears to be greater at negative potentials when voltage-dependent inactivation, which typically exhibits a U-shaped availability curve, is weak. After β-adrenergic stimulation, Ca²⁺-dependent inactivation becomes the main inactivation mechanism as a result of a slowing down of voltage-dependent inactivation.^46–48

The Ca²⁺-dependent inactivation mechanism depends primarily on Ca²⁺ released from the sarcoplasmic reticulum. The Ca²⁺-binding protein calmodulin functions as a critical sensor mediating Ca²⁺-induced inactivation of L-type Ca²⁺ channels. Calmodulin binds to two α_1C subunit amino acid sequences (called domains L and K). When local intracellular Ca²⁺ concentration increases (secondary to influx Ca²⁺ via the L-type Ca²⁺ channel, as well as Ca²⁺-induced Ca²⁺ release from the sarcoplasmic reticulum), more Ca²⁺ ions bind to calmodulin, which harbors four Ca²⁺-binding sites. When saturated with Ca²⁺, conformational change of both calmodulin and α_1C subunit leads to blockage of the channel pore.

Voltage steady-state activation and inactivation are sigmoidal, with an activation range over –40 to +10 mV (with a half-activation potential near –15 mV) and a half-inactivation potential near –35 mV. However, a relief of inactivation for voltages positive to 0 mV leads to a U-shaped voltage curve for steady-state inactivation. Overlap of the steady-state voltage-dependent inactivation and activation relations defines a window current near the action potential plateau, within which transitions from closed and open states can occur that may participate in action potential repolarization and may play a major role in the initiation of EADs.^47,48

After inactivation, the transition of Ca²⁺ channels from the inactivated to the closed resting state (i.e., recovery from inactivation [reactivation, restoration, or repriming]) is also Ca²⁺ and voltage dependent. Reduction of intracellular Ca²⁺ concentration in the immediate vicinity of the channel allows recovery from Ca²⁺-dependent inactivation. Acceleration of Ca²⁺ channel reactivation, as may occur secondary to reuptake of Ca²⁺ by the sarcoplasmic reticulum during prolonged depolarization, can result in the recovery from Ca²⁺-dependent inactivation and enable secondary depolarization. This leads to instability of the cell E_m during repolarization and may be the basis for the EADs that are capable of initiating torsades de pointes.^46,48

Voltage-dependent recovery of I_CaL from inactivation between action potentials is slow at a low (depolarized) E_m, and it becomes very fast as the action potential repolarization is nearly complete. As a consequence, I_CaL declines in response to repetitive stimulation at a partially depolarized E_m between pulses (resulting from Ca²⁺ channel incomplete recovery from inactivation), and a negative staircase of contractility is observed.

In contrast, at normal resting potentials, recovery of I_CaL from inactivation is fast, and I_CaL may increase progressively during repetitive stimulation. This positive staircase or rate-dependent potentiation of contractility is Ca²⁺ dependent and likely is the result of diminished Ca²⁺-dependent inactivation at frequencies with less sarcoplasmic reticulum Ca²⁺ release. Additionally, similar to Ca²⁺-dependent inactivation, Ca²⁺-dependent facilitation requires high-affinity binding of calmodulin to the C-terminal tail of the Ca_v1.2 channel and may be facilitated by calmodulin kinase II–dependent phosphorylation. Calmodulin kinase II is a Ca²⁺/calmodulin-dependent serine/threonine kinase that is activated by low intracellular Ca²⁺ concentration. The facilitatory effect of Ca²⁺ entry on subsequent I_CaL is distinct from, but coexistent with, Ca²⁺-dependent inactivation.^46–48

Function

I_CaL is activated by membrane depolarization. It is largely responsible for the action potential plateau and is a major determinant of the duration of the plateau phase and hence of action potential duration and refractoriness. I_CaL also links membrane depolarization to myocardial contraction and constitutes the dominant factor in mediating positive inotropy in all types of cardiac tissue. Additionally, I_CaL is responsible for the upstroke (phase 0) of slow response action potentials (in pacemaking cardiomyocytes and regions of depressed resting E_m) and contributes to physiological frequency regulation in the sinus node.^46–48

L-type Ca²⁺ channels are the principal portal of entry of Ca²⁺ into the cells during depolarization. Ca²⁺ influx during the action potential plateau triggers more massive Ca²⁺ release (Ca²⁺ transients) from the sarcoplasmic reticulum into the cytosol via activation of Ca²⁺-release channels (e.g., RyR2). This amplifying process, termed Ca²⁺-induced Ca²⁺ release (CICR), causes a rapid increase in intracellular Ca²⁺ concentration (from approximately 100 nM to approximately 1 µM) to a level required for optimal binding of Ca²⁺ to troponin C and induction of contraction. Most of the L-type Ca²⁺ channels in the adult myocyte are localized in the transverse tubules (T tubules) facing the sarcoplasmic reticulum junction and the RyR2, organized as a “complex” that ensures coordinated Ca²⁺ release during excitation-contraction coupling.^46,49

Cytosolic Ca²⁺ concentration decreases during diastole: contraction is followed by Ca²⁺ release from troponin C and its reuptake by the sarcoplasmic reticulum via activation of the sarcoplasmic reticulum Ca²⁺-ATPase Ca²⁺ pump, in addition to extrusion across the sarcolemma via the Na⁺-Ca²⁺ exchanger. Intracellular Ca²⁺-dependent inactivation limits Ca²⁺ influx during action potential.^46,47,49

For maintenance of intracellular Ca²⁺ homeostasis and balanced cardiac activity, Ca²⁺ influx into cytoplasm via L-type Ca²⁺ channels has to be terminated. This is achieved by Ca²⁺-dependent inactivation of L-type Ca²⁺ channels. This inactivation serves as a negative feedback mechanism for regulating Ca²⁺ entry into the cell and as a physiological safety mechanism against a harmful Ca²⁺ overload in the cell, which can cause both arrhythmias and cell death. Ca²⁺-dependent inactivation is also a major determinant of action potential duration, and it ensures that contraction and relaxation cycles of the heart muscle fiber are coordinated. A failure to deactivate I_CaL completely may possibly be an essential mechanism underlying EADs caused by suppression of the K⁺ delayed rectifier currents.⁵⁰

Inhibition of α_1C subunit binding to calmodulin eliminates Ca²⁺-dependent inactivation, thus promoting Ca²⁺-dependent facilitation, which contributes to a force-frequency relationship in the heart.^46,47,49

Regulation

Phosphorylation of the pore-forming α_1C subunits by different kinases is one of the most important pathways to change the activity of the L-type Ca²⁺ channel. Phosphorylation by PKA is the main mechanism of Ca²⁺ channel activation, because it increases the probability and duration of the open state of the channels and consequently increases I_CaL.^46,47,49

Several different agonists (e.g., catecholamines, glucagon, histamine, serotonin) can activate PKA-mediated phosphorylation and activation of the L-type Ca²⁺ channel via an intracellular signaling cascade. Once one of these agonists binds to its receptor, receptor stimulation activates guanosine triphosphate (GTP)–binding protein (G_s), which activates adenylyl cyclase, which, in turn, mediates the conversion of ATP into cAMP. The increased cAMP levels stimulate cAMP-dependent PKA phosphorylation of the α_1C subunit of L-type Ca²⁺ channels and result in an increase in I_CaL amplitude and a shift in activation to a more negative E_m. cAMP is degraded by cAMP phosphodiesterases, and the signaling cascade is then suppressed limiting cAMP-dependent phosphorylation; in addition, the signaling cascade is terminated by serine/threonine phosphatases that remove a phosphate group from kinase-phosphorylated proteins.^46,47

The suppression of adenylyl cyclase activity is one of the most common pathways to interrupt PKA-dependent Ca²⁺ channel stimulation. Adenylyl cyclase is usually suppressed (and cAMP synthesis is blocked) by activation of G_i proteins. Stimulation of various G_i protein–coupled receptors (e.g., M₂ muscarinic receptors, adenosine A₁ receptors, opiates, and atrial natriuretic peptides) does not change basal I_CaL in most cases, but reduces I_CaL increased via stimulation of β-adrenergic receptors. Activation of phosphodiesterases is another way to reduce PKA-dependent channel phosphorylation. Phosphodiesterases hydrolyze cAMP and cyclic guanosine monophosphate (cGMP) and decrease their intracellular concentrations.^46,47,49

The physiological functions of cardiac L-type Ca²⁺ channels are under control of catecholamines of circulating and neurohumoral origin. The effects of adrenergic stimulation are exerted by phosphorylation of the L-type Ca²⁺ channel subunits by PKA, PKC, and PKG. β₁-Adrenergic receptors couple exclusively to the G_s protein, thus producing a widespread increase in cAMP levels in the cell, whereas β₂-adrenergic receptors couple to both G_s and G_i, thus producing a more localized activation of L-type Ca²⁺ channels.^46,49

The effect of PKC-mediated phosphorylation on I_CaL can be highly diverse. PKC can either increase or decrease I_CaL. Activation of G_q subunits by G_q protein–coupled receptors (e.g., α-adrenergic receptors, endothelin, angiotensin II, and muscarinic receptors) stimulates phospholipase C, which hydrolyzes PIP₂ to inositol 1,4,5-triphosphate (InsP₃) and diacylglycerol (DAG). DAG activates PKC, which, in turn, phosphorylates L-type Ca²⁺ channels. The mechanism of the effect of PKC on the activity of cardiac L-type Ca²⁺ channels is not exactly known. PKC phosphorylates the N-terminus of the α_1C subunit, and the effect on the channel can be either stimulating or suppressive.^46,47

Activation of soluble guanylate cyclase (primarily by nitric oxide) results in the conversion of GTP into cGMP. cGMP activates PKG, which phosphorylates the α_1C subunit of the L-type Ca²⁺ channel, with a resulting inhibition of I_CaL. Besides direct phosphorylation of the L-type Ca²⁺ channel, it is also possible that PKG activates a protein phosphatase, which dephosphorylates the channel, or that cGMP activates phosphodiesterase 2, which reduces cAMP levels. Thus, stimulation of I_CaL by PKA is inhibited. However, besides an inhibition of I_CaL, stimulatory effects of the PKG pathway have been shown.^46,47,49

I_CaL is blocked by several cations (e.g., Mg²⁺, nickel [Ni²⁺], zinc [Zn²⁺]) and drugs (dihydropyridines, phenylalkylamines, benzothiazepines). In addition, coexpression of the β-subunit increases I_CaL amplitude, accelerates the kinetics of Ca²⁺ channel activation, and alters pharmacological properties of the channel.^47,49

Pharmacology

Cardiac L-type Ca²⁺ channels are the targets for the interaction with class IV antiarrhythmic drugs. The three classes of organic Ca²⁺ channel blockers include dihydropyridines (e.g., nifedipine, nicardipine, amlodipine, felodipine), phenylalkylamines (verapamil), and benzothiazepines (diltiazem). Each drug type binds specifically to separate binding sites on the channel’s α_1C subunit. The combined use of Ca²⁺ channel blockers can enhance or weaken the block effect because, at least in part, of the different binding sites for those drugs. It is noteworthy that increased extracellular Ca²⁺ concentrations inhibit the binding of phenylalkylamines and dihydropyridines to their receptors on the Ca²⁺ channel.

Verapamil and diltiazem preferentially block open and inactivated states of the channel. The more frequently the Ca²⁺ channel opens, the better is the penetration of the drug to the binding site; hence, these drugs cause use-dependent block of conduction in cells with Ca²⁺-dependent action potentials such as those in the sinus node and AVN. This explains their preferential effect on nodal tissue in paroxysmal supraventricular tachycardia.^46,47,49

The dihydropyridines block open Ca²⁺ channels. However, the lack of use-dependence and the presence of voltage sensitivity of the dihydropyridines with regard to their binding explain their vascular selectivity. The kinetics of recovery from block is sufficiently fast that these drugs produce no significant cardiac effect but effectively block the smooth muscle Ca²⁺ channel because of its low resting potential.^46,47,49

Inherited Channelopathies

Acquired Diseases

Abnormalities in Ca²⁺ currents or intracellular Ca²⁺ transients, or both, in acquired diseases may induce both arrhythmia and contractile dysfunction. In AF, Ca_v1.2 mRNA and protein levels are downregulated, resulting in I_CaL reduction, which contributes to action potential shortening.³

In heart failure, the membrane density of I_CaL channels is reduced. However, channel phosphorylation is increased, leading to reduced response to phosphorylating interventions and causing increased single channel open probability that compensates for the reductions in channel density. Despite unchanged I_CaL, sarcoplasmic reticulum Ca²⁺ transients are smaller and slower in heart failure, and they cause contractile dysfunction.^3,24,49

Structure and Physiology

The cardiac T-type Ca²⁺ channel (originally called low-voltage–activated channels) is composed of a single α subunit. Ca_v3.1 (α_1G subunit encoded by CACNA1G) and Ca_v3.2 (α_1H subunit encoded by CACNA1H) isoforms are major candidates for the cardiac T-type Ca²⁺ channel. It is not yet clear whether auxiliary subunits exist for native T-type Ca²⁺ channels. The structure of the α_1H and α_1G subunits is similar to that involved in the L-type Ca²⁺ channels.^46,48

T-type Ca²⁺ channels can be distinguished from L-type Ca²⁺ channels on the basis of their distinctive gating and conductance properties. Compared with L-type Ca²⁺ channels, T-type Ca²⁺ channels have a smaller conductance and transient openings, and they open at the significantly more negative E_m that overlaps the pacemaker potentials of sinus node cells. The threshold for activation of I_CaT is –70 to –60 mV, and I_CaT is fully activated at –30 to –10 mV at physiological Ca²⁺ concentration. Membrane depolarization also causes inactivation of I_CaT. The inactivation threshold is near –90 mV, with half-maximal inactivation of –60 mV. In contrast to L-type Ca²⁺ channels, T-type Ca²⁺ channels do not inactivate in a Ca²⁺-dependent manner. The activation and steady-state inactivation overlap near the activation threshold (–60 to –30 mV), thus providing a constant inward current (a window current). This window component may help in facilitating the slow diastolic depolarization in sinus node cells and contribute to automaticity. Unlike L-type channels, T-type Ca²⁺ channels are relatively insensitive to dihydropyridines.^46,48

Function

Ca_v3 channels conduct the T-type Ca²⁺ current (I_CaT), which is important in a wide variety of physiological functions, including neuronal firing, hormone secretion, smooth muscle contraction, cell proliferation of some cardiac tissues, and myoblast fusion. In the heart, T-type channels are abundant in sinus node pacemaker cells and Purkinje fibers of many species and are important for maintenance of pacemaker activity by setting the frequency of action potential firing.^46,48

T-type Ca²⁺ channels are functionally expressed in embryonic hearts, but they are almost undetectable or markedly reduced in postnatal ventricular myocytes, although some reports described substantial amplitude of I_CaT. In the adult heart, the largest I_CaT densities are seen in pacemaker cells located in the conduction system.

Because T-type Ca²⁺ channels are most prevalent in the conduction system in the adult heart and the activation range of I_CaT overlaps the pacemaker potential, it has been suggested that T-type Ca²⁺ channels play a role in generating pacemaker depolarization and contribute to automaticity. However, experimental evidence indicates that I_CaT is not a primary pacemaker current, but it can modify depolarization frequency only slightly. Although organic T-type Ca²⁺ channel blockers (mibefradil) result in marked decrease in firing frequency of sinus node cells in clinical studies, the possibility that these blockers affect other ionic currents, including I_CaL, cannot be entirely excluded. In fact, it is has not yet been determined whether I_CaT exists functionally in atrial, ventricular, and sinoatrial node cells in the human heart. Further studies are necessary to clarify whether T-type Ca²⁺ channels contribute to the automaticity of the human heart.^48,51

Acquired Diseases

Interestingly, the T- type Ca²⁺ channels are re-expressed in atrial and ventricular myocytes in animal models under various pathological conditions such as cardiac hypertrophy, MI, and heart failure. These findings reflect a reversion to a fetal or neonatal pattern of gene expression, and I_CaT contributes to abnormal electrical activity and excitation-contraction coupling. It is possible that similar remodeling occurs in the hypertrophied human heart; however, to date, T-type Ca²⁺ channels have not been detected in normal or diseased human myocardial cells.^46,48,51

Furthermore, experimental evidence suggests that T-type Ca²⁺ channels may be of functional importance in arrhythmogenesis in cardiomyocytes in pulmonary veins, which can initiate paroxysmal AF. I_CaT may directly and indirectly participate in pacemaker depolarization in sinoatrial and other regions of the heart, and this mechanism may become more important in failing hearts.^46,48

Structure and Physiology

Channels responsible for the pacemaker current (I_f; also called the funny current because it displays unusual gating properties) are named hyperpolarization-activated cyclic nucleotide-gated (HCN) channels. HCN channels are members of the voltage-gated cation channel superfamily and, based on sequence homology, are most closely related to the cyclic nucleotide-gated (CNG) channel and ether-a-go-go (EAG) K⁺ channel families. Four α-subunit isoforms are described (HCN1 to HCN4, encoded by HCN1 to HCN4 genes), which are preferentially expressed in sinus and AVN myocytes and Purkinje fibers. HCN isoforms differ in the extent of voltage-dependent gating and sensitivity to cAMP, and they have different relative rates of activation and deactivation, with HCN1 the fastest, HCN4 the slowest, and HCN2 and HCN3 intermediate. HCN4 is the isoform primarily expressed in the sinus node, AVN, and ventricular conducting system, but low levels of HCN1 and HCN2 have also been reported.

It is likely that the HCN channel is formed by the coassembly of four either identical (homotetramers) or nonidentical (heterotetramers) α subunits that create an ion-conducting pore. Each α subunit comprises six transmembrane segments (S1 to S6), with a voltage sensor domain in the S4 segment and a pore-forming region between S5 and S6 carrying the GYG triplet signature of K⁺-permeable channels. Their intracellular C-terminus contains cyclic nucleotide-binding domains, which enable direct cAMP binding. A potential auxiliary subunit of HCN channels is MiRP1 (encoded by KCNE2).

I_f is a mixed Na⁺-K⁺ current, with a threefold higher selectivity for Na⁺ than for K⁺. Despite the GYG amino acid motif, HCN channels are more permeable Na⁺ than K⁺ ions. Unlike most voltage-gated channels, which are activated on membrane depolarization, HCN channels are activated on hyperpolarization. The HCN channel activates slowly on hyperpolarization (at voltages lower than approximately −40 to −45 mV) and inactivate slowly in a voltage-independent manner on depolarization. The speed of channel opening is strongly dependent on E_m and is faster at more negative potentials. I_f conducts an inward current during phases 3 and 4 of the action potential and may underlie slow membrane depolarization in cells with pacemaker activity (i.e., cells with I_f and little or no I_K1).⁵²

Function

I_f is a major player in both generation of spontaneous activity and rate control of cardiac pacemaker cells, and it is sometimes referred to as the pacemaker current.

The I_f channels are deactivated during the action potential upstroke and the initial plateau phase of repolarization, but they begin to activate at the end of the action potential as repolarization brings the E_m to levels more negative than approximately −40 to −50 mV, and they are fully activated at approximately −100 mV. Once activated, I_f depolarizes the membrane back toward a level at which the Ca²⁺ current activates to initiate the action potential.¹ In its range of activation, which quite properly comprises the voltage range of diastolic depolarization in sinus node cells (approximately −40 to −65 mV), the current is inward, and its reversal occurs at approximately −10 to −20 mV. At the end of the repolarization phase of an action potential, because I_f activation occurs in the background of a decaying outward (K⁺ time-dependent) current, current flow quickly shifts from outward to inward, giving rise to a sudden reversal of voltage change (from repolarizing to depolarizing) at the maximum diastolic potential. Hence, I_f first opposes and then stops the repolarization process (at the maximum diastolic potential) and finally initiates the diastolic depolarization.^52–55

The I_f contribution terminates when, in the late part of diastolic depolarization, Ca²⁺-dependent processes take over, and the threshold for L-type Ca²⁺ current activation and action potential firing is reached. Although deactivation of I_f at depolarized voltages is rapid, complete switch off of the current occurs only during the very early fraction of the action potential, which provides a brief time interval during which I_f carries an outward current at positive voltages.⁵²

I_f is not only involved in principal rhythm generation but also plays a key role in heart rate regulation. The degree of activation of I_f determines, at the end of an action potential, the steepness of phase 4 depolarization and hence the frequency of action potential firing. Additionally, I_f represents a basic physiological mechanism mediating autonomic regulation of heart rate. I_f is regulated by intracellular cAMP and is thus activated and inhibited by β-adrenergic and muscarinic M₂ receptor stimulation, respectively.^52,56

However, given the complexity of the cellular processes involved in rhythmic activity, exact quantification of the extent to which I_f and other mechanisms contribute to pacemaking is still a debated issue.⁵²

Regulation

The voltage dependence of I_f activation is regulated by cAMP direct binding to the cyclic nucleotide-binding domain in the HCN channel and not via phosphorylation-dependent activation mechanisms. Direct interaction of cAMP to the channel shifts the activation curve to more depolarized voltages and strongly accelerates channel activation kinetics. Sympathetic stimulation activates I_f and hence accelerates heart rate via β-adrenoceptor-triggered cAMP production, whereas low-level vagal stimulation lowers heart rate via inhibition of cAMP synthesis and an ensuing inhibition of I_f activity. High vagal tone most likely lowers the heart rate mainly via the activation of I_KACh.⁵⁶

HCN channels are inhibited by increased intracellular acidity (e.g., during myocardial ischemia). Protons shift the activation of I_f to more hyperpolarized potentials and slow pacemaker activity.

Pharmacology

Given the key role of HCN channels in cardiac pacemaking, I_f has become a pharmacological target for the development of novel and more specific heart rate–reducing agents in patients with ischemic heart disease. Whereas current heart rate–lowering drugs adversely affect cardiac contractility, selective I_f inhibition is believed to lower heart rate without impairing contractility. In the past, several agents inhibiting cardiac I_f were developed. Early drugs identified as pure bradycardic agents include zatebradine and cilobradine, which are derived from the L-type Ca²⁺ channel blocker verapamil. More recently, ivabradine was introduced into clinical use as the first therapeutic I_f blocker for the treatment of chronic stable angina. The principal action of all these substances is to reduce the frequency of pacemaker potentials in the sinus node by inducing a reduction of the diastolic depolarization slope. Ivabradine blocks HCN4 and HCN1 channels by accessing the channels from their intracellular side and by exerting a use- and current-dependent block. Interestingly, ivabradine acts as open channel blocker in HCN4 (as in sinus nodal I_f), whereas block of HCN1 requires channels either to be closed or in a transitional state between an open and closed configuration.^52,56 With the exception of ivabradine, other HCN channel blockers are not specific enough for sinus nodal (mainly HCN4-mediated) I_f; they also block neuronal HCN channels (I_h current) in several regions of the nervous system, and this has prevented their clinical utility.⁵⁶

Clonidine, an α₂-adrenergic agonist, was shown to block sinus nodal I_f. Clonidine produces a shift in the voltage dependence of the channel by 10 to 20 mV to more hyperpolarizing potentials.⁵⁶

Inherited Channelopathies

Heterozygous HCN4 mutations have been identified in individuals with sinus bradycardia and chronotropic incompetence. Severe bradycardia, QT prolongation, and torsades de pointes have been described in another family. HCN mutations slow channel activation kinetics or, when located in cyclic nucleotide-binding domains, abolish sensitivity of HCN channels to cAMP, thus reducing I_f and the speed of diastolic depolarization.

Acquired Diseases

HCN2/HCN4 expression is upregulated in the atria of patients with AF and in ventricular tissues in cardiac hypertrophy and congestive heart failure. This response may contribute to the arrhythmias observed in these disease states. Enhancement of I_f in these pathological conditions can potentially initiate arrhythmia by triggering spontaneous excitation of nonpacemaker cardiomyocytes.

Structure and Physiology

The Ca²⁺ release channel is a macromolecular complex, formed by the cardiac ryanodine receptor isoform (RyR2, encoded by the RYR2 gene) homotetramer and certain proteins localized on both the cytosolic and the luminal side of the sarcoplasmic reticulum membrane. The cardiac RyR2, by far the largest protein of the complex, operates as a Ca²⁺-conducting channel. RyR2 channels are approximately 10 times larger than voltage-gated Ca²⁺ and Na⁺ channels.^57,58

Each RyR2 monomer contains a transmembrane domain, the pore-forming region that is composed of an even, but still undetermined number (likely six to eight) of transmembrane segments. This domain encompasses only approximately 10% of the protein clustered at the C-terminus, but it has a critical functional role because it contains sequences that control RyR2 localization and oligomerization and is sufficient to form a functional Ca²⁺ release channel. The remaining 90% of the protein at the N-terminus comprises an enormous cytoplasmic domain that serves as a cytosolic scaffold that interacts with regulatory molecules (including Ca²⁺, ATP) and proteins (including FKBP12.6, calmodulin). On the luminal (sarcoplasmic reticulum) side, RyR2 forms a part of a large quaternary complex with calsequestrin (CASQ2), triadin, and junctin. Together these four proteins form the core of the Ca²⁺ release channel complex.^57–60

Cardiac RyR2 functions as a ligand-activated ion channel that activates (opens) on Ca²⁺ binding. However, the exact structural determinants of RyR gating are as yet unknown. RyR2 is normally closed at low cytosolic diastolic Ca²⁺ concentrations (approximately 100 to 200 nM). At submicromolar cytosolic Ca²⁺ concentrations, Ca²⁺ binds to high-affinity binding sites on RyR2 and thus increases the open probability of the channel (two Ca²⁺ ions are required to open the RyR2 channel) and allows Ca²⁺ release from the sarcoplasmic reticulum into the cytosol.^50,59,60

The precise juxtaposition of the sarcolemmal specialized invaginations (known as T tubules) and sarcoplasmic reticulum forms specific junctional microdomains, creating a 10- to 12-nM gap, known as the dyadic cleft. RyR2s are assembled in a paracrystalline lattice in each dyad, containing 80 to 260 channels, where the RyR2 cytoplasmic region resides, and its transmembrane region spans the sarcoplasmic reticulum membrane to immerse the luminal portion into the sarcoplasmic reticulum Ca²⁺ store. Each array of RyR2s is faced by 10 to 25 L-type Ca²⁺ channels in the sarcolemmal T tubule. Hence, each dyad constitutes a local Ca²⁺ signaling complex, or couplon, whereby these proteins are coordinately regulated via the changing concentrations of Ca²⁺, Na⁺, and K⁺ within the dyadic cleft.^50,59,60

After approximately 10 milliseconds of RyR2 channel opening, Ca²⁺ release from the sarcoplasmic reticulum terminates, and the Ca²⁺ spark signal starts to decay, mostly owing to diffusion of Ca²⁺ away from its source. RyR2 channel activity is maximal at cytosolic Ca²⁺ concentrations of approximately 10 µM. Elevating cytosolic Ca²⁺ concentrations beyond this point leads to a reduction in the open probability of the channel, possibly because of Ca²⁺ binding to low-affinity inhibitory binding sites on the RyR2 channel.⁶⁰

Inactivation of Ca²⁺ release is not well understood. It is likely mediated by Ca²⁺-induced inactivation of RyR2, thus extinguishing of CICR by stochastic attrition, or by Ca²⁺ depletion of the sarcoplasmic reticulum, or both.

RyR2 open probability increases by elevation of sarcoplasmic reticulum Ca²⁺ concentration. When levels of Ca²⁺ in the sarcoplasmic reticulum reach a critical threshold, spontaneous Ca²⁺ release (spillover) can occur even in the presence of normal channels (store overload–induced Ca²⁺ release [SOICR]). Ca²⁺ concentration in the sarcoplasmic reticulum is physiologically increased as an effect of adrenergic (sympathetic) stimulation.

Function

The RyR2 channels are an essential component of the excitation-contraction coupling and act as sentinels to the large sarcoplasmic reticulum Ca²⁺ store. Excitation-contraction coupling describes the physiological process of converting an electrical stimulus (action potential) to a mechanical response (muscle contraction). The contraction of a cardiac myocyte is governed primarily by intracellular Ca²⁺ concentration (see Fig. 1-3). Ca²⁺ enters the cell during the plateau phase of the action potential through L-type Ca²⁺ channels that line the sarcolemmal T tubules. However, the rise in intracellular Ca²⁺ is small and not sufficient to induce contraction. Nonetheless, the small amount of Ca²⁺ entering the cell via I_CaL triggers a rapid mobilization of Ca²⁺ from the sarcoplasmic reticulum into the cytosol by opening the RyR2 channels in the CICR process. Approximately 75% of Ca²⁺ present in the cytoplasm during contraction is released from the sarcoplasmic reticulum. The close proximity of the RyR2 to the T tubule enables each L-type Ca²⁺ channel to activate 4 to 6 RyR2s and generate a Ca²⁺ spark. Ca²⁺ influx via I_CaL simultaneously activates approximately 10,000 to 20,000 couplons in each ventricular myocyte with every action potential. Such sophisticated coordination in opening and closing is required to ensure that Ca²⁺ release occurs during the systolic phase of the cardiac cycle and functional silence during diastole.^50,59,60

Luminal Ca²⁺-dependent deactivation of RyR2 is a process by which the decline in sarcoplasmic reticulum Ca²⁺ that follows sarcoplasmic reticulum Ca²⁺ release renders RyR2s functionally inactive. This results in termination of CICR and induction of a refractory state that suppresses Ca²⁺ release during the diastolic phase, which is an important determinant of the mechanical refractoriness required for efficient relaxation and refilling of the heart.^50,58

Regulation

Many proteins interact directly and indirectly with the N-terminal cytoplasmic domain of RyR2 including FK506-binding protein (calstabin2 or FKBP12.6), PKA, Ca²⁺-calmodulin-dependent kinase II (CaMKII), phosphodiesterase 4D3, calmodulin, protein phosphatases 1 and 2A, and sorcin. CASQ2, junctin, and triadin bind with the luminal (sarcoplasmic reticulum) C-terminus of RyR2.^58–60

Pharmacology

RyR2s are targets of multiple experimental drugs. However, thus far, no compounds in clinical use are known to target RyR2 directly.

The plant alkaloid ryanodine binds the RyR2 channel with high affinity in a Ca²⁺-dependent and use-dependent fashion, thus making it an important tool for biochemical characterization of the channel. Two ryanodine-binding sites, a high-affinity site and a low-affinity one, have been described at the C-terminus of RyR2. At the high-affinity site, ryanodine induces long-lasting channel openings at a subconductance state, whereas high concentrations block the channel.

Caffeine, in high concentrations (5 to 20 mM), increases the RyR2 sensitivity to Ca²⁺ and ATP and results in increased RyR2 mean open time and open probability. Caffeine is used experimentally to measure sarcoplasmic reticulum Ca²⁺ content indirectly because its application causes emptying of the sarcoplasmic reticulum Ca²⁺ store.

JTV-519, also known as K201, is a benzothiazepine derivative (an analogue of diltiazem) and an L-type Ca²⁺ channel blocker and stabilizer. JTV-519 can increase the binding affinity of calstabin2 for RyR2, thus stabilizing the closed conformational state of the RyR2 channel and hence preventing diastolic sarcoplasmic reticulum Ca²⁺ leak. JTV-519 can potentially offer an antiarrhythmic benefit in a variety of pathological conditions that lead to destabilization of the RyR2 channel’s closed state, such as RyR2 mutations, hyperphosphorylation of the RyR2 during heart failure, and sarcoplasmic reticulum Ca²⁺ overload.

Among class I antiarrhythmic drugs, only flecainide and propafenone were found to inhibit RyR2 channels (by inducing brief closures of open RyR2 to subconductance states), suppress arrhythmogenic Ca²⁺ sparks, and prevent CPVT in experimental studies. The potency of RyR2 channel inhibition rather than Na⁺ channel blockade appears to determine the efficacy of class I agents for the prevention of CPVT. Flecainide has been demonstrated to prevent lethal ventricular arrhythmias in patients with familial CPVT.^63,64

Several toxins (e.g., the scorpion toxins imperatoxin A and imperatoxin I), some anticancer drugs (e.g., doxorubicin), and some immunosuppressants (e.g., rapamycin) can potentially cause cardiac adverse events likely related to effects on the gating kinetics of RyR2 channels.

Inherited Channelopathies

CPVT is caused by mutations in genes that encode for key Ca²⁺ regulatory proteins. Two genetic variants of CPVT have been described: an autosomal-dominant trait (CPVT1; most common) caused by mutations in the cardiac RyR2 gene, and a recessive form (CPVT2; rare) associated with homozygous mutations in the CASQ2 gene (CASQ2).

Approximately 50% to 70% of patients with CPVT harbor RyR2 mutations. More than 70 RyR2 mutations linked to CPVT have been identified. CPVT mutant RyR2 typically shows gain-of-function defects following channel activation by PKA phosphorylation (in response to β-adrenergic stimulation or caffeine), resulting in uncontrolled Ca²⁺ release from the sarcoplasmic reticulum during electrical diastole. The exaggerated spontaneous Ca²⁺ release from the sarcoplasmic reticulum facilitates the development of DADs and triggered arrhythmias.^58,59

The molecular mechanisms by which RyR2 mutations alter the physiological properties and function of RyR2 are not completely defined. It has been suggested that CPVT mutations in RyR2 reduce the binding affinity of RyR2 for the regulatory protein FKBP12.6 (calstabin2) that stabilizes the closed conformational state of the RyR2 channel, thus enabling the channel to close completely during diastole (at low intracellular Ca²⁺ concentrations), preventing aberrant Ca²⁺ leakage from the sarcoplasmic reticulum, and ensuring muscle relaxation. PKA phosphorylation (induced by β-adrenergic stimulation) of the mutant channels results in further worsening of the binding affinity of FKBP12.6 to the mutant RyR2 and increases the probability of an open state at diastolic Ca²⁺ concentrations. As a consequence, the mutant RyR2 channel fails to close completely during diastole, with a resulting diastolic Ca²⁺ leak from the sarcoplasmic reticulum during stress or exercise.^58–60

An alternative hypothesis is that CPVT mutations in RyR2 sensitize the channel to luminal (sarcoplasmic reticulum) Ca²⁺ such that under baseline conditions, when sarcoplasmic reticulum load is normal, there is no Ca²⁺ leak. Under β-adrenergic (sympathetic) stimulation, sarcoplasmic reticulum Ca²⁺ concentration becomes elevated above the reduced threshold, causing Ca²⁺ to leak out of the sarcoplasmic reticulum. A third hypothesis for RyR2-related CPVT is that mutations in RyR2 impair the intermolecular interactions between discrete RyR2 domains necessary for proper folding of the channel and self-regulation of channel gating.^58,60

So far, seven CASQ2 mutations linked to CPVT have been reported. Although some of these mutations are thought to compromise CASQ2 synthesis and result in reduced expression or complete absence of CASQ2 in the heart, other mutations seem to cause expression of defective CASQ2 proteins with abnormal regulation of cellular Ca²⁺ homeostasis. CASQ2 mutations result in disruption of the control of RyR2s by luminal Ca²⁺ required for effective termination of sarcoplasmic reticulum Ca²⁺ release and prevention of spontaneous Ca²⁺ release during diastole, thus leading to diminished Ca²⁺ signaling refractoriness and generation of arrhythmogenic spontaneous Ca²⁺ releases.^58,59,65

Importantly, in the setting of digitalis poisoning, the abnormal RyR2 behavior leading to spontaneous Ca²⁺ release and DADs is secondary to the elevation of the sarcoplasmic reticulum Ca²⁺ content (SOICR). In CPVT, on the other hand, spontaneous Ca²⁺ release and DADs can occur without Ca²⁺ overload. Mutations in RyR2 or CASQ2 lead to defective Ca²⁺ signaling lowering of the sarcoplasmic reticulum Ca²⁺ threshold for spontaneous Ca²⁺ release to less than the normal baseline level (perceived Ca²⁺ overload).⁵⁸

Missense mutations in RyR2 also been linked to a form of arrhythmogenic cardiomyopathy (ARVD-2) characterized by exercise-induced polymorphic VT that does not appear to have a reentrant mechanism and occurs in the absence of significant structural abnormalities. Patients do not develop characteristic features of ARVD on the 12-lead ECG or signal-averaged ECG, and global RV function remains unaffected. ARVD-2 shows a closer resemblance to familial CPVT in both etiology and phenotype; its inclusion under the umbrella term of ARVD remains controversial.^66–68

Acquired Diseases

RyR2 dysfunction is a key factor leading to arrhythmias in heart failure. Chronic β-adrenergic stimulation results in PKA hyperphosphorylation of RyR2. This process causes the dissociation of the channel-stabilizing protein calstabin and leads to diastolic Ca²⁺ leak from the sarcoplasmic reticulum and the generation of spontaneous Ca²⁺ waves, which can be maintained despite a reduced Ca²⁺ gradient, thus underlying DAD-induced triggered arrhythmias in heart failure.^24,69,70

Structure and Physiology

Cardiomyocytes make contact with each other via multiple intercalated discs, which mediate the transmission of force, electrical continuity, and chemical communication between adjacent cells. Three types of specialized junctions exist in the intercalated disc: (1) the fascia adherens, (2) the macula adherens (desmosome), and (3) the gap junction (nexus). The fascia adherens is an anchoring site for myofibrils, facilitating the transmission of mechanical energy between neighboring cells. The desmosomes link to the cytoskeleton of adjacent cells to provide strong localized adhesion sites that resist shearing forces generated during contraction. Gap junctions are assemblies of intercellular channels that provide electrical continuity and chemical communication between adjacent cells.⁷¹

In addition to the end-to-end and side-to-side gap junctions localized at the intercalated discs, lateral (side-to-side) gap junctions can exist in nondisc lateral membranes of cardiomyocytes, but they are much less common, occurring more in atrial than ventricular myocardium.⁷¹

Each gap junction channel is constructed of two hemichannels (connexons) aligned head-to-head in mirror symmetry across a narrow extracellular gap, one provided by each of the adjoining cells. Each connexon is composed of six integral membrane proteins called connexins (Cx) hexagonally arranged around the pore. Each connexin consists of four membrane-spanning domains (M1 to M4), two extracellular loops (E1, E2), one intracellular loop, and cytoplasmic N- and C-termini. The extracellular loops mediate the docking of the two hemichannels.^71,72

Up to 24 different connexin types have been identified. They are named after their theoretical molecular weight in daltons. In the heart, Cx40, Cx43, and Cx45 are most important for action potential propagation. Although each connexin exhibits a distinct tissue distribution, most cardiomyocytes express more than one connexin isoform. Cx43 is by far the most abundant and is expressed between atrial and ventricular myocytes and distal parts of the Purkinje system. Cx40 is mainly expressed in the atrial myocytes, AVN, and HPS. Cx45 appears to be primarily expressed in nodal tissue (the sinus and compact AVNs), and more weakly in the atrium, His bundle, bundle branches, and Purkinje fibers.⁷¹

Connexons can be composed by the oligomerization of a single connexin type (homomeric) or of different types (heteromeric). In addition, the gap junction channel as a whole may be formed of two matching hemichannels (homotypic) or nonmatching hemichannels (heterotypic).⁷¹

Cardiac connexins exhibit distinctive biophysical properties; hence, the connexin composition of a gap junction channel determines its unitary conductance, voltage sensitivity, and ion selectivity. Cx40 gap junctions express the largest conductance, and Cx45 expresses the smallest. Both Cx40 and Cx45 are highly cation selective, and their conductance is voltage dependent. Cx43 has an intermediate conductance and is nonselective.

The individual gap junction channels allow exchange of nutrients, metabolites, ions (e.g., Na⁺, Cl⁻, K⁺, Ca²⁺) and small molecules (e.g., cAMP, cGMP, inositol triphosphate [IP₃]) with molecular weights up to approximately 1000 Da.⁷¹

Function

Gap junctions maintain direct cell-to-cell communication in the heart by providing biochemical and low-resistance electrical coupling between adjacent cardiomyocytes. Thus, gap junctions are responsible for myocardial electrical current flow propagation from one cardiac cell to another and are crucial in myocardial synchronization and heart function. Gap junctions also provide biochemical coupling, which allows intercellular movement of second-messenger substances (e.g., ATP, cyclic nucleotides, and IP₃) and hence enables coordinated responses of the myocardial syncytium to physiological stimuli.

The role of gap junction channels in action potential propagation and conduction velocity in cardiac tissue depends primarily on static factors of the channels (e.g., the number of channels, channel conductance, and voltage sensitivity) and dynamic factors (e.g., channel gating kinetics), as well as on properties of the propagated action potential, structural aspects of the cell geometry, and tissue architecture. Tissue-specific connexin expression and gap junction spatial distribution, as well as the variation in the structural composition of gap junction channels, allow for a greater versatility of gap junction physiological features and for disparate conduction properties in cardiac tissue. The myocytes of the sinus and AVNs are equipped with small, sparse, dispersed gap junctions containing Cx45, a connexin that forms low conductance channels; this feature underlies the relatively poor intercellular coupling in nodal tissues, a property that is linked to slowing of conduction. In contrast, ventricular muscle expresses predominantly Cx43 and Cx45, which have larger conductance. Atrial muscle and Purkinje fibers express all three cardiac connexins.

Under physiological conditions, a given cardiomyocyte in the adult working myocardium is electrically coupled to an average of approximately 11 adjacent cells, with gap junctions predominantly localized at the intercalated discs at the ends of the rod-shaped cells. Lateral (side-to-side) gap junctions in nondisc lateral membranes of cardiomyocytes are much less abundant and occur more often in atrial than ventricular tissues.⁷¹ Thus, intercellular current flow occurs primarily at the cell termini, although propagation can occur both longitudinally and transversely. This particular subcellular distribution of gap junctions underlies uniform anisotropic impulse propagation throughout the myocardium, whereby conduction in the direction parallel to the long axis of the myocardial fiber bundles is approximately 3 to 5 times more rapid than that in the transverse direction. This property is attributable principally to the lower resistivity of myocardium provided by the gap junctions in the longitudinal versus the transverse direction.^73–75

Regulation

Gap junction channels have a voltage-dependent gating mechanism, depending primarily on transjunctional voltage (i.e., the potential difference between the cytoplasm of the two adjacent cells). At rest, when the junctional voltage is zero, the channels usually are open. During the course of a propagated action potential, the channels tend to close in a voltage- and time-dependent manner. Gap junction channel gating can also be altered by specific changes in intracellular ions and by post-translational modifications (“loop” gating). Cytosolic Na⁺ and Ca²⁺ overload, acidosis, and reduced ATP levels decrease gap junction channel function. Unlike the voltage gate, which closes rapidly and incompletely, the chemical gate closes slowly and completely.⁷⁶

The regulation of gap junction trafficking, assembly and disassembly, and degradation is likely to be critical in the control of intercellular communication. Phosphorylation also appears to play a key role in channel gating that determines channel conductance and has been implicated in the regulation of the connexin “life cycle” at several stages. Gap junction coupling also is regulated by certain endogenous mediators (e.g., acetylcholine, norepinephrine, and angiotensin), likely via phosphorylation-mediated mechanisms. Importantly, channels composed of different connexins possess different properties and are susceptible to different regulation.^71,76

Pharmacology

Several agents have been found to decrease gap junction channel coupling, including long-chain alcohols (e.g., heptanol, octanol), fatty acids (myristoleic acid, decanoic acid, palmitoleic acid), general anesthetics (e.g., halothane, isoflurane), carbachol (an acetylcholine analogue), α-adrenergic agonists (phenylephrine), angiotensin, insulin, insulin-like growth factor, and the nonsteroidal agents fenamates (e.g., flufenamic acid, meclofenamic acid). The mechanism of action of those agents remain largely unknown.

Antimalarial drugs, particularly quinine and quinine derivatives such as mefloquine, can reduce gap junction channel conductance, and their effects seem to be connexin subtype specific. In addition, it has been suggested that the cardiac glycosides strophanthidin, ouabain, and digitoxin decrease intercellular coupling.⁷⁶

Experimental evidence suggests that increasing gap junction conductance can potentially confer an antiarrhythmic effect. Some compounds, including antiarrhythmic peptides and their derivatives (e.g., AAP10, ZP123, rotigaptide) can upregulate Cx43 via modulation either synthesis or degradation and enhance gap junctional communication and were found to reduce conduction slowing and prevent AF and ischemic reentrant VT in various cell and animal models. These peptides presumably act by modulating Cx43 phosphorylation.^77,78

Acquired Diseases

Modification of cell-to-cell coupling occurs in numerous pathophysiological settings (e.g., myocardial ischemia, ventricular hypertrophy, cardiomyopathy) as a consequence of acute changes in the average conductance of gap junctions secondary to ischemia, hypoxia, acidification, or an increase in intracellular Ca2⁺, or it can be produced by changes in expression or cellular distribution patterns of gap junctions.^74,79–81

Remodeling includes a decrease in the number of gap junction channels resulting from the interruption of communication between cells by fibrosis and downregulation of Cx43 formation or of trafficking to the intercalated disc. Additionally, gap junctions can become more prominent along lateral membranes of myocytes (so-called structural remodeling). Influences of Cx43 lateralization on impulse propagation have not yet been well defined.⁸²

Alterations in distribution and function of cardiac gap junctions are associated with conduction delay or block. Inactivation of gap junctions decreases transverse conduction velocity to a greater degree than longitudinal conduction, thus resulting in exaggeration of anisotropy and providing a substrate for reentrant activity and increased susceptibility to arrhythmias.⁷⁴

AF is associated with abnormal expression and distribution of atrial Cx40, which can potentially lead to inhomogeneous electrical coupling and abnormal impulse formation and conduction and thereby provide the substrate for atrial arrhythmias. Furthermore, a rare single nucleotide polymorphism in the atrial-specific Cx40 gene has been found to increase the risk of idiopathic AF.⁸³

Importantly, there is a high redundancy in connexin expression in the heart with regard to conduction of electrical impulse. It has been shown that a 50% reduction in Cx43 does not alter ventricular impulse conduction. Cx43 expression must decrease by 90% to affect conduction, but even then conduction velocity is reduced only by 20%.