Cardiac ion channels

Long QT syndrome

Type 3 congenital LQTS (LQT3), which accounts for approximately 5% to 10% of congenital LQTS cases, is caused by gain-of-function mutations in the SCN5A gene, which encodes the α subunit of the Na ⁺ channel (Na _v 1.5), SCN5A . More than 200 mutations have been identified in SCN5A , with most being missense mutations mainly clustered in Na _v 1.5 regions that are involved in fast inactivation (i.e., S4 segment of domain IV, the domain III–domain IV linker, and the cytoplasmic loops between the S4 and S5 segments of domain III and domain IV), or in regions that stabilize fast inactivation (e.g., the C-terminus).

Several mechanisms have been identified to underlie ionic effects of SCN5A mutations in LQT3 ( eFigs. 3.1 and 3.2 ). Most SCN5A mutations cause a gain of function through disruption of fast inactivation, thus allowing repeated reopening during sustained depolarization and resulting in an abnormal, small, but functionally important sustained (or persistent) noninactivating Na ⁺ current (I _sus ) during action potential plateau. Because the general membrane conductance is small during the action potential plateau, the presence of a persistent inward I _Na , even of small amplitude, can potentially have a major impact on the plateau duration and can be sufficient to prolong repolarization and QT interval. QT prolongation and the risk of developing arrhythmia are more pronounced at slow heart rates, when the action potential duration is longer, thereby allowing more I _Na to enter the cell.

Other less common mechanisms of SCN5A mutations to cause LQT3 include increased window current, which results from delayed inactivation of mutant Na ⁺ channels, occurring at more positive potentials and widening the voltage range during which the Na ⁺ channel may reactivate without inactivation. Additionally, some mutations cause slower inactivation, which allows longer channel openings and causes a slowly inactivating I _Na . This current is I _NaL and is to be distinguished from I _sus (which does not inactivate). Comparable to I _sus , both the window current and I _NaL exert their effects during phases 2 and 3 of the action potential, in which normally no or very small I _Na is present. Other mutations induce prolonged action potential duration by enhancing recovery from inactivation, an effect that leads to larger peak I _Na by increasing the fraction of channels available for activation (because of faster recovery) during subsequent depolarizations. Finally, some mutations can cause increased expression of mutant Na _v 1.5 through enhanced mRNA translation or protein trafficking to the sarcolemma, decreased protein degradation, or altered modulation by β subunits and regulatory proteins. These effects lead to larger I _Na density during phase 0 of the action potential. Importantly, one single SCN5A mutation can potentially cause several changes in the expression and/or gating properties of the resulting Na ⁺ channels.

Regardless of the mechanism, increased Na ⁺ current (I _sus , window current, I _NaL , or peak I _Na ) upsets the balance between depolarizing and repolarizing currents in favor of depolarization. The resulting delay in the repolarization process triggers early afterdepolarizations (EADs) by reactivation of the L-type Ca ²⁺ channel during phase 2 or 3 of the action potential, especially in Purkinje fiber myocytes, in which action potential durations are intrinsically longer. Compared with other LQT subtypes, patients with LQT3 are particularly at risk for sudden cardiac death (SCD), and cardiac arrest is often the first clinical event

LQT9 is caused by gain-of-function mutations on the CAV3 gene, which encodes caveolin-3, a plasma membrane scaffolding protein that interacts with Na _v 1.5 and plays a role in compartmentalization and regulation of channel function. Mutations in CAV3 induce kinetic alterations of the Na _v 1.5 current that result in persistent Na ⁺ current (I _sus ) and have been reported in cases of sudden infant death syndrome (SIDS).

LQT10 is caused by loss-of-function mutations on the SCN4B gene, which encodes the β subunit (Na _v β4) of the Na _v 1.5 channel. These mutations likely cause a shift in the inactivation of the I _Na toward more positive potentials, resulting in increased window currents at an E _m corresponding to the phase 3 of the action potential.

LQT12 is caused by mutations on the SNTA1 gene, which encodes α1 syntrophin, a cytoplasmic adaptor protein that enables the interaction among Na _v 1.5, nitric oxide synthase, and the sarcolemmal Ca ²⁺ adenosine triphosphatase (ATPase) complex that appears to regulate ion channel function. By disrupting the interaction between Na _v 1.5 and the sarcolemmal Ca ²⁺ ATPase complex, SNTA1 mutations cause increased Na _v 1.5 nitrosylation with consequent reduction of channel inactivation and enhanced I _sus densities.

Brugada syndrome

The Brugada syndrome is an autosomal dominant inherited channelopathy characterized by an elevated ST segment or J wave appearing in the right precordial leads. This syndrome is associated with a high incidence of SCD secondary to a rapid polymorphic ventricular tachycardia (VT) or ventricular fibrillation (VF). Approximately 65% of mutations identified in the SCN5A gene are associated with the Brugada syndrome phenotype (Brugada syndrome type 1), and they account for approximately 11% to 28% of cases of Brugada syndrome. So far, more than 300 Brugada syndrome–associated loss-of-function (i.e., reduced peak I _Na ) mutations have been described in SCN5A . Some of these mutations result in loss of function secondary to impaired channel trafficking to the cell membrane (i.e., reduced expression of functional Na ⁺ channels), disrupted ion conductance (i.e., expression of nonfunctional Na ⁺ channels), or altered gating function. Altered gating properties comprise delayed activation (i.e., activation at more positive potentials), earlier inactivation (i.e., inactivation at more negative potentials), faster inactivation, and enhanced slow inactivation.

Most SCN5A mutations are missense mutations, whereby a single amino acid is replaced by a different amino acid. Missense mutations commonly alter the gating properties of mutant channels. Because virtually all reported SCN5A mutation carriers are heterozygous, mutant channels with altered gating may cause up to 50% reduction of I _Na . Different SCN5A mutations can cause different degrees of I _Na reduction and therefore different degrees of severity of the clinical phenotype of Brugada syndrome.

In addition to SCN5A mutations, reduction in I _Na can be caused by mutations in SCN1B (encoding the β1 and β1b subunits of the Na ⁺ channel), SCN2B (encoding the β2 subunit), and SCN3B (encoding the β3 subunit), resulting in the clinical phenotype of the Brugada syndrome. Recently, SCN10A (which encodes Na _v 1.8, a neuronal Na ⁺ channel, which appears to play a role in the heart) was identified as a major susceptibility gene for Brugada syndrome (identified in 16.7% of probands). Loss of function mutations in SCN10A leads to significant reduction in I _Na .

Furthermore, mutations in GPD1L (which encodes the protein glycerol-3-phosphate dehydrogenase 1-like [G3PD1L] protein) affect the trafficking of the cardiac Na ⁺ channel to the cell surface, resulting in reduction of I _Na and Brugada syndrome. The Brugada phenotype associated with GPD1L mutations is characterized by progressive conduction disease, low sensitivity to procainamide, and a relatively good prognosis.

Mutations in several other genes have been reported to cause reduction in I _Na and lead to the Brugada phenotype, including HEY2 (encoding the transcriptional factor HEY2), FGF12 (encoding for a fibroblast growth factor homologous factor-1, which exerts modulatory effects on cardiac Na ⁺ and Ca ²⁺ channels), PKP2 (encoding the desmosomal protein plakophilin-2, a known susceptibility gene for arrhythmogenic right ventricular cardiomyopathy [ARVC]), RANGRF (encoding MOG1, a protein known to modulate the Na ⁺ channel), and SLMAP (encoding the sarcolemmal membrane–associated protein, SLMAP, a component of T-tubules and sarcoplasmic reticulum).

Early repolarization syndrome

Loss-of-function mutations in the α1 subunit of NaV1.5 ( SCN5A ) and NaV1.8 ( SCN10A ) have been reported in patients with early repolarization syndrome.

Familial progressive cardiac conduction disease

Loss-of-function SCN5A mutations have been linked to familial forms of progressive cardiac conduction disease (referred to as hereditary Lenègre disease, primary cardiac conduction system disease, and familial atrioventricular [AV] block). This disease is characterized by slowing of electrical conduction through the atria, AVN, His bundle, Purkinje fibers, and ventricles, accompanied by an age-related degenerative process and fibrosis of the cardiac conduction system, in the absence of structural or systemic disease. It is often reflected by varying degrees of AV block and bundle branch block. Whether the age-dependent fibrosis of the conduction system is a primary degenerative process in progressive cardiac conduction disease or a physiological process that is accelerated by I _Na reduction remains to be established. A single loss-of-function SCN5A mutation can cause isolated progressive cardiac conduction disease or can be combined with the Brugada syndrome (overlap syndrome). Loss-of-function mutations in SCN1B also have been identified in patients with progressive cardiac conduction disease who carried no mutation in SCN5A .

Congenital sick sinus syndrome

Although I _Na does not play a prominent role in sinus node activity, mutations in SCN5A have been linked to sick sinus syndrome, manifesting as sinus bradycardia, sinus arrest, sinoatrial block, or a combination of these conditions, which can progress to atrial inexcitability (atrial standstill). Loss-of-function SCN5A mutations result in reduced peak I _Na density, hyperpolarizing shifts in the voltage dependence of steady-state channel availability, and slow recovery from inactivation. These effects likely cause reduced automaticity, decreased excitability, and conduction slowing or block of impulses generated in the sinus node to the surrounding atrial tissue. Sinus node dysfunction can also manifest concomitantly with other phenotypes that are linked to SCN5A loss-of-function mutations such as Brugada syndrome and progressive cardiac conduction disorders.

Familial atrial fibrillation

Loss-of-function mutations, gain-of-function mutations, and common polymorphisms on the SCN5A gene have been identified in some cases of atrial fibrillation (AF) occurring in young patients with structurally normal hearts. It is speculated that I _Na reduction can predispose to AF by slowing the electrical conduction velocity and thereby facilitating reentry. On the other hand, gain-of-function mutations can potentially predispose to AF by increasing atrial excitability. AF can occur in patients with other phenotypes of Na ⁺ channelopathies, including LQT3, Brugada syndrome, dilated cardiomyopathy, and sinus node dysfunction. Furthermore, mutations in the SCN1B gene (encoding the β1 subunit of the Na ⁺ channel) and the SCN2B gene (encoding the β2 subunit of the Na ⁺ channel) have been identified in patients with AF, many of whom displayed ECG patterns suggestive of the Brugada syndrome.

Dilated cardiomyopathy

Some cases of familial dilated cardiomyopathy have been linked to SCN5A mutations. Dilated cardiomyopathy–linked SCN5A mutations cause diverse loss-of-function and gain-of-function changes in the gating properties, but how such changes evoke contractile dysfunction is not understood. It is speculated that SCN5A mutations disrupt the interactions between the mutant Na ⁺ channels and intracellular (or extracellular) proteins that are essential for normal cardiomyocyte structure and architecture. Notably, dilated cardiomyopathy with SCN5A mutations often displays atrial or ventricular arrhythmias (including AF, VT, and VF), sinus node dysfunction, AV block, and intraventricular conduction delay.

Sudden infant death syndrome

Gain-of-function mutations in SCN5A may be the most prevalent genetic cause of SIDS. SCN5A mutations in SIDS commonly increase I _sus . Less frequently, loss-of-function mutations in SCN5A or CAV3 and gain-of-function mutations in GPD1-L have also been found in infants with SIDS. However, it is possible that in these patients SIDS represents a malignant form of LQT3 or Brugada syndrome that manifests during infancy.

Overlap syndrome

A single SCN5A mutation can result in multiple clinical phenotypes and rhythm disturbances within the same family, a phenomenon now referred to as “cardiac sodium channel overlap syndrome.” Not surprisingly, loss-of-function SCN5A mutations have often been associated with overlapping phenotypes of Brugada syndrome, sinus node dysfunction, and progressive cardiac conduction disorders, which all share similar underlying mechanisms that implicate I _Na reduction. More surprisingly, some SCN5A mutations are associated with both BrS1 (I _Na loss-of-function) and LQT3 (I _Na gain-of-function). Carriers of these mutants present with BrS1, LQT3, or mixed ECGs with both ST-segment abnormalities and QT elongation. These mutations are likely associated with altered gating properties in a manner that results in both reduction of the peak I _Na and augmentation of the persistent I _Na . Furthermore, it is likely that genetic background and clinical and environmental factors play a role in the variable disease expressivity and severity.

Structure and physiology

Cardiac I _to channels are macromolecular protein complexes, comprising four pore-forming K _v α subunits and a variety of K _v channel accessory (β) subunits (see Fig. 3.3 ). Two major types of I _to have been characterized: (1) I _to1 generated by voltage-dependent, Ca ²⁺ -independent K _v channels; and (2) I _to2 generated by Ca ²⁺ -activated Cl ^– channels. In human atrial and ventricular myocytes, the presence of I _to2 has not been clearly demonstrated.

I _to1 (which is referred to as I _to ) displays two phenotypes with distinct recovery kinetics: a rapid or fast I _to (I _to,fast or I _to,f ) phenotype and a slower phenotype (I _to,slow or I _to,s ). The transient nature of Ito is secondary to its rapid activation (with time constants of less than 10 milliseconds for both I _to,f and I _to,s ) and rapid inactivation (25–80 milliseconds for I _to,f and 80–200 milliseconds for I _to,s ). However, whereas I _to,f recovers rapidly from inactivation (60–100 milliseconds), I _to,s recovers slowly (with time constants on the order of seconds).

K _v channels mediating I _to,s are formed by the coassembly of four α subunits from the K _v 1.x subfamily (primarily K _v 1.4, and possibly K _v 1.7), whereas those mediating I _to,f are formed by the coassembly of four α subunits from the K _v 4.x subfamily (primarily K _v 4.3, and possibly K _v 4.2) ( eTable 3.2 ). Among the various accessory subunits identified, a crucial role has been definitively demonstrated only for KChIP2, and potentially for MiRP2.

Function

I _to is a prominent repolarizing current; it partially repolarizes the membrane, shapes the rapid (phase 1) repolarization of the action potential, and sets the height of the initial plateau (phase 2). Thus, the activity of I _to channels influences the activation of voltage-gated L-type Ca ²⁺ channels and the balance of inward and outward currents during the plateau (mainly the L-type Ca ²⁺ current [I _CaL ] and the delayed rectifier K ⁺ currents), thereby mediating the duration and the amplitude of phase 2.

The density of I _to varies across the myocardial wall and in different regions of the heart. In human ventricles, I _to densities are much higher in the epicardium and midmyocardium than in the endocardium. Furthermore, I _to,f and I _to,s are differentially expressed in the myocardium, thus contributing to regional heterogeneities in action potential waveforms. I _to,f is the principal subtype expressed in human atrium. The markedly higher densities of I _to,f , together with the expression of the ultrarapid delayed rectifier K ⁺ current, accelerate the early phase of repolarization and lead to lower plateau potentials and shorter action potentials in atrial as compared with ventricular cells. Due to its slow recovery kinetics, I _to,s plays a limited role in repolarization compared to I _to,f , especially at faster heart rates.

Although both I _to,f and I _to,s are expressed in the ventricle, I _to,f is more prominent in the epicardium and midmyocardium (putative M cells) than in the endocardium, whereas I _to,s is mainly present in the endocardium and Purkinje fiber cells. These regional differences are responsible for the shorter duration and the prominent phase 1 notch and the “spike-and-dome” morphology of epicardial and midmyocardial compared with endocardial action potentials. A prominent I _to -mediated action potential notch in ventricular epicardium but not endocardium produces a transmural voltage gradient during early ventricular repolarization that registers as a J wave or J point elevation on the ECG. I _to densities are also reportedly higher in right than in left (midmyocardial and epicardial) ventricular myocytes, consistent with the more pronounced spike-and-dome morphology of right, compared with left, ventricular action potentials, particularly in the epicardium.

Furthermore, variations in cardiac repolarization associated with I _to regional differences strongly influence intracellular Ca ²⁺ transient by modulating Ca ²⁺ entry via I _CaL and Na ⁺ -Ca ²⁺ exchanger, thereby regulating excitation-contraction coupling and regional modulation of myocardial contractility and hence synchronizing the timing of force generation between different ventricular regions and enhancing mechanical efficiency.

Regulation

I _to channels are subject to α- and β-adrenergic regulation. α-Adrenergic stimulation reduces I _to ; concomitant β-adrenergic stimulation appears to counteract the α-adrenergic effect, at least in part. The effects of α- and β-adrenergic stimulation are exerted by phosphorylation of the K _v 1.4, K _v 4.2, and K _v 4.3 α-subunits by PKA as well as PKC. Calmodulin-dependent kinase II, on the other hand, has been shown to be involved in enhancement of I _to . Adrenergic stimulation is also an important determinant of transient outward channel downregulation in cardiac disease. Chronic α-adrenergic stimulation and angiotensin II reduce I _to channel expression, which explains channel downregulation in many types of chronic heart disease.

KChIP2, when coexpressed with K _v 4.3, increases surface channel density and current amplitude, slows channel inactivation, and markedly accelerates the recovery from inactivation. In the ventricle, KChIP2 mRNA is 25-fold more abundant in the epicardium than in the endocardium. This gradient parallels the gradient in I _to expression, whereas K _v 4.3 mRNA is expressed at equal levels across the ventricular wall. Thus, transcriptional regulation of the KChIP2 gene is the primary determinant of I _to expression in the ventricular wall.

Observations suggest that MiRP2 is required for the physiological functioning of human I _to,f channels and that gain-of-function mutations in MiRP2 predispose to Brugada syndrome through augmentation of I _to,f .

I _to is strongly rate dependent. I _to fails to recover from previous inactivation at very fast heart rates; hence, tachycardia is associated with reduction of I _to , which can be manifest as a decrease in the magnitude of the J wave on the surface ECG. Hence, abrupt changes in rate and pauses have important consequences for the early repolarization of the membrane.

I _to can be enhanced by aging, low sympathetic activity, high parasympathetic activity, bradycardia, hypothermia, and drugs. Estrogen suppresses the expression of the K _v 4.3 channel and results in reduced I _to and a shallow phase 1 notch.

Phase 1 notch of the action potential modulates the kinetics of slower activating ion currents and consequently the later phases of the action potential. Initial enhancement of phase 1 notch promotes phase 2 dome and delays repolarization, presumably by delaying the peak of I _CaL . However, further enhancement of phase 1 notch prevents the rising of phase 2 dome and abbreviates action potential duration, presumably by deactivation or voltage modulation that reduces I _CaL . Thus, progressive deepening of phase 1 notch can cause initial enhancement followed by sudden disappearance of phase 2 dome and corresponding prolongation followed by abbreviation of action potential duration. On the other hand, modulators that decrease I _to lead to a shift of the plateau phase into the positive range of potentials, thus increasing the activation of the delayed rectifier currents, promoting faster repolarization, and reducing the electrochemical driving force for Ca ²⁺ and hence I _CaL . Phase 1 notch also affects the function of the Na ⁺ -Ca ²⁺ exchanger and subsequently intracellular Ca ²⁺ handling and Na ⁺ channel function.

Pharmacology

Quinidine, 4-aminopyridine, flecainide, and propafenone produce an open channel blockade and accelerate I _to inactivation. Quinidine, but not flecainide or propafenone, produces a frequency-dependent block of I _to that results from a slow rate of drug dissociation from the channel. Quinidine has a relatively strong I _to blocking effect, whereas flecainide mildly blocks I _to .

I _to blockers can potentially prolong the action potential duration in the atrium and in ischemic ventricular myocardium. However, because the net effects of I _to blockade on repolarization depend on secondary changes in other currents, the reduction of I _to density can result in a shortening of the ventricular action potential. Moreover, heterogeneous ventricular distribution of I _to can cause marked dispersion of repolarization across the ventricular wall that, when accompanied by prominent conduction delays related to Na ⁺ channel blockade, results in extrasystolic activity through a phase 2 reentrant mechanisms.

Currently, no cardioselective and channel-specific I _to openers or blockers are available for clinical use. Development of an I _to -selective drug is expected to be beneficial in patients with primary abnormality in the I _to or in other channels, such as the Brugada syndrome, in which heterogeneity in the expression of I _to between epicardium and endocardium in the right ventricle (RV) results in the substrate responsible for reentry and ventricular arrhythmias.

Inherited channelopathies

Gain-of-function mutations in KCND3 (which encodes the α-subunit of the I _to channel [K _v 4.3]) and KCNE3 (which encodes the auxiliary β subunit [MiRP2]) result in an increase in I _to density and cause Brugada syndrome. Furthermore, gain-of-function mutations in SCN1B (which encodes the auxiliary β1 subunit of the Na ⁺ channel), in addition to reducing I _Na , can also increase I _to . Mutations in SEMA3A (which encodes semaphorin) were also implicated in Brugada syndrome by increasing I _to .

Additionally, gain-of-function mutations in KCND3 and KCNE3 have been linked to familial AF. KCNE3 mutations were found to increase I _to,f and were postulated to cause AF by shortening action potential duration and facilitating atrial reentrant excitation waves ( eTable 3.3 ).

A genome-wide haplotype-sharing study associated a haplotype on chromosome 7, harboring DPP6, with idiopathic VF in three distantly related families. Overexpression of DPP6 , which encodes dipeptidyl-peptidase 6, a putative component of the I _to channel complex, was proposed as the likely pathogenetic mechanism. DPP6 significantly alters the inactivation kinetics of both K _v 4.2 and K _v 4.3 and promotes expression of these α subunits in the cell membrane.

Importantly, the normally functioning I _to channels play an important role in the electrophysiological consequences of the ionic current abnormalities in the Brugada and J wave syndromes. Heterogeneity in the distribution of I _to channels across the myocardial wall, being more prominent in ventricular epicardium than endocardium, and particularly in the RV, results in the shorter duration and the prominent phase 1 notch and the spike-and-dome morphology of the epicardial action potential as compared with the endocardium. The resultant transmural voltage gradient during the early phases (phases 1 and 2) of the action potential is thought to be responsible for the inscription of the J wave on the surface ECG. An increase in net repolarizing current, secondary to either a decrease in the inward currents (I _Na and I _CaL ) or an increase in the outward K ⁺ currents (I _to , I _Kr , I _Ks , I _KACh , I _KATP ), or both, can accentuate the action potential notch and lead to augmentation of the J wave or the appearance of ST segment elevation on the surface ECG. An outward shift of currents that extends beyond the action potential notch not only accentuates the J wave but also can lead to partial or complete loss of the dome of the action potential, thus resulting in a protracted transmural voltage gradient that manifests as greater ST segment elevation and gives rise to J wave syndromes. The type of the ion current affected and its regional distribution in the ventricles determine the particular phenotype (including the Brugada syndrome, early repolarization syndrome, hypothermia-induced ST segment elevation, and MI-induced ST segment elevation). The degree of accentuation of the action potential notch leading to loss of the dome depends on the magnitude of I _to . These changes are more prominent in regions of the myocardium exhibiting a relatively large I _to , such as the RV epicardium; this explains the appearance of coved ST segment elevation, characteristic of Brugada syndrome, in the right precordial ECG leads.

In this context, factors that influence the kinetics of I _to or the other repolarization currents can modify the manifestation of the J wave on the ECG. Na ⁺ channel blockers (procainamide, pilsicainide, propafenone, flecainide, and disopyramide), which reduce the inward I _Na , can accentuate the J wave and ST segment elevation in patients with concealed J wave syndromes. Quinidine, which inhibits both I _to and I _Na , reduces the magnitude of the J wave and normalizes ST segment elevation. Additionally, acceleration of the heart rate, which is associated with reduction of I _to (because of slow recovery of I _to from inactivation), results in a decrease in the magnitude of the J wave. Male predominance can potentially result from larger epicardial I _to density versus I _to in women.

The increased transmural heterogeneity of ventricular repolarization (i.e., dispersion of repolarization between epicardium and endocardium), which is responsible for J point elevation and early repolarization pattern on the surface ECG, is also responsible for the increased vulnerability to ventricular tachyarrhythmias. A significant outward shift in current can cause partial or complete loss of the dome of the action potential in regions where I _to is prominent (epicardium), with the consequent loss of activation of I _CaL . The dome of the action potential then can propagate from regions where it is preserved (midmyocardium endocardium) to regions where it is lost (epicardium), thus giving rise to phase 2 reentry, which can generate premature ventricular complexes that in turn can initiate polymorphic VT or VF.

Acquired diseases

An alteration in the expression and distribution of I _to is observed in various pathophysiological conditions ( eFig. 3.3 ). Adrenergic effects seem to be involved in at least some of these I _to -regulating processes during heart disease.

In general, myocardial ischemia, MI, dilated cardiomyopathy, and end-stage heart failure cause downregulation of I _to . In fact, I _to downregulation is the most consistent ionic current change in the failing heart. The reduction in I _to results in attenuation of early repolarization (phase 1) and affects the level of plateau (phase 2) of the action potential and other currents involved in delayed repolarization (phase 3), with resulting prolongation and increased heterogeneity of action potential duration. The prominent epicardial I _to contributes to the selective electrical depression of the epicardium. This process leads to the development of a marked dispersion of repolarization between normal and abnormal epicardium and between epicardium and endocardium, which provides the substrate for reentrant arrhythmias and may underlie the increased predisposition to ventricular arrhythmias and SCD in patients with heart failure and ischemic heart disease. Additionally, downregulation of I _to in advanced heart failure likely slows the time course of force generation, thereby contributing to reduced myocardial performance.

On the other hand, compensatory ventricular hypertrophy preceding heart failure is associated with upregulation of I _to . The prolongation of the action potential, concomitant with an increase in I _to , presumably results from the more negative level of the plateau with less I _CaL inactivation and probably less delayed rectifier activation. In contrast, progression of hypertrophy to heart failure is associated with a clear reduction in I _to .

Chronic AF reduces atrial I _to density and K _v 4.3 mRNA levels. Hypothyroidism reduces the expression of KCND2 (K _v 4.2) genes. Additionally, I _to may also be reduced and contribute to QT interval prolongation in diabetes. Importantly, with some delay, insulin therapy partially restores I _to , possibly by enhancing K _v 4.3 expression.

Structure and physiology

The ion-conducting pore of I _Kur channels is formed by four K _v 1.5 α subunits (encoded by KCNA5 ), whereas the ancillary β subunits K _v β1.2, K _v β1.3, and K _v β2.1 control channel trafficking and plasma membrane integration as well as activation and inactivation kinetics.

I _Kur activates rapidly on depolarization in the plateau range and displays outward rectification, but it inactivates very slowly during the time course of the action potential. Inactivation accelerates when K _v 1.5 is coexpressed with its β subunits.

Function

I _Kur is detected only in human atria and not in ventricular or nodal cells, so it is the predominant delayed rectifier current responsible for human atrial repolarization and is a basis for the much shorter duration of the action potential in the atrium.

Regulation

β-Adrenergic stimulation enhances I _Kur , whereas α-adrenergic stimulation inhibits it, effects likely mediated by PKA and PKC, respectively. Membrane depolarization and elevated extracellular K ⁺ concentrations reduce K _v 1.5 expression. Additionally, cAMP, mechanical stretch, hyperthyroidism, and dexamethasone increase K _v 1.5 expression, whereas extracellular acidosis, phenylephrine, and hypothyroidism decrease it.

Coexpression of K _v β1.3 with K _v 1.5 induces a fast inactivation and a hyperpolarizing shift in the activation curve (i.e., K _v β1.3 subunit converts K _v 1.5 from a delayed rectifier with a modest degree of slow inactivation to a channel with both fast and slow components of inactivation). Data suggest that K _v β1.2 and K _v β1.3 subunit modification of K _v 1.5 currents requires phosphorylation by PKC or a related kinase.

Pharmacology

I _Kur is relatively insensitive to class III antiarrhythmic drugs of the methanesulfonanilide group, but it is highly sensitive to 4-aminopyridine. Selective inhibition of I _Kur by 4-aminopyridine prolongs the human atrial action potential duration.

Because I _Kur is atrium specific, K _v 1.5 channels are a promising target for the development of new, safer antiarrhythmic drugs to prevent AF without risk of inducing QT prolongation and ventricular arrhythmias. However, although several selective I _Kur blockers have been evaluated, the clinical value of “pure” I _Kur blockers for antiarrhythmic therapy in AF remains unproven. Because K _v 1.5 is downregulated in AF, the beneficial effect of I _Kur blockade becomes less certain. Furthermore, because K _v 1.5 is also expressed in other organs (e.g., brain), discovery of drugs that selectively inhibit cardiac K _v 1.5 channels remains necessary.

Vernakalant, an I _Kur blocker, has been shown effective for the acute termination of AF and was approved in Europe. The efficacy of vernakalant in acute cardioversion of AF is likely due to multiple-channel block. Vernakalant inhibits two atrial-specific K ⁺ channels (I _Kur and I _KACh ) at low concentrations and is able to block I _to and I _Na at higher concentrations.

Although selective I _Kur block would be expected to specifically prolong atrial action potential duration, such effects were not observed in healthy subjects at physiological atrial rates. Simulated inhibition of I _Kur was shown to increase plateau height, leading to additional activation of I _Kr and no net change in action potential duration.

Physiologically, rapid activation of I _Kur in the positive potential range following the action potential upstroke can offset depolarizing I _CaL and hence lead to the less positive plateau phase in atrial compared with ventricular cardiomyocytes. Conversely, block of I _Kur produces a more pronounced spike-and-dome configuration and therefore shifts the potential into a more positive range in which I _CaL activation enhances systolic Ca ²⁺ influx during a free-running action potential. Such an indirect effect on I _CaL should be shared by all I _Kur blockers and is expected to result in a positive atrial inotropic effect.

Inherited channelopathies

KCNA5 mutations have been reported in individuals with familial AF. Heterologous expression of these mutations revealed complete I _Kur loss of function ( eTable 3.3 ). Absence of I _Kur can excessively prolong atrial action potential duration with an enhanced risk of EADs that can trigger or maintain AF.

Acquired diseases

A prolonged period of rapid activation of the atria as occurs during AF leads to a decrease in I _Kur . Additionally, I _Kur may be affected in myocardial ischemia ( eFig. 3.3 ). Decreased K _v 1.5 mRNA levels were reported for the epicardial border zone of infarcted hearts. Moreover, ischemic damage disrupted the normal location of K _v 1.5 in the intercalated discs.

Structure and physiology

I _Kr is formed by coassembly of four pore-forming α subunits (K _v 11.1, encoded by KCNH2 ) and β subunits (MiRP1, encoded by KCNE2 ). KCNH2 is also known as the human ether-a-go-go-related gene ( HERG ), named so because the mutation in the Drosophila fruit fly caused it to shake like a go-go dancer. The current generated by HERG channels shows unusual voltage dependence. In contrast to I _Ks , I _Kr activates relatively rapidly (on the order of 10s of milliseconds) on membrane depolarization. Activation of I _Kr occurs with steep voltage dependence and reaches half-maximum activation at membrane voltage of approximately –20 mV. The magnitude of I _Kr increases as a function of E _m up to approximately 0 mV, but it declines with stronger depolarization (higher than 0 mV), resulting in a negative slope conductance of the current-voltage relationship. During repolarization of the action potential, I _Kr rapidly recovers from inactivation, thus causing the current to peak at –40 mV. The amplitude of the tail current on repolarization exceeds that of the current during the depolarizing pulse.

The unusual voltage dependence of I _Kr results from a fast, voltage-dependent C/P-type inactivation process, which limits outward K ⁺ flow at positive voltages. The large tail current on repolarization from positive voltages results from the rapid recovery of inactivated channels into a conducting state. On repolarization, HERG channels deactivate (close) via a slow, voltage-independent process (in contrast to the voltage-dependent inactivation process).

Unlike most K _v channels, HERG channels exhibit inward rectification. Rectification describes the property of an ion channel to allow currents preferentially to flow in one direction or limit currents from flowing in the other direction. In other words, conductivity of channels carrying such currents is not constant but is altered at a different E _m . A channel that is inwardly rectifying is one that passes current (positive charge) more easily into the cell. This property is critical for limiting outward K ⁺ conductance during the plateau phase of the cardiac action potential. Unlike typical Kir channels, in which rectification derives from blockade of the channel pore by intracellular polyamines (see later discussion), the mechanism of HERG inward rectification is a very rapid inactivation that develops at far more negative potentials (–85 mV) than channel activation (–20 mV).

The inactivation of HERG channels resembles the C/P-type inactivation of other K _v channels in its sensitivity to extracellular cations (including K ⁺ and Na ⁺ ) and tetraethyl ammonium (TEA), and to mutations in the P segment. However, the gating behavior is distinctive. First, channel inactivation is much faster than voltage-dependent activation, thus resulting in its characteristic rectification. Second, HERG inactivation displays intrinsic voltage dependence. Similar to the classic C/P-type inactivation, raising the concentration of extracellular K ⁺ slows HERG channel C/P-type inactivation, an effect that appears to result from occupancy of the pore selectivity filter by K ⁺ .

Function

I _Kr presents the principal repolarizing current at the end of the plateau phase in most cardiac cells and plays an important role in governing the cardiac action potential duration and refractoriness. I _Kr is differentially expressed, with high levels in left atrial and in ventricular endocardium.

I _Kr activates relatively fast on membrane depolarization and allows outward diffusion of K ⁺ ions in accordance with its electrochemical gradient, but voltage-dependent inactivation thereafter is very fast. Hence, only limited numbers of channels remain in the open state, while a considerable fraction resides in the nonconducting inactivated state. The fast voltage-dependent inactivation limits outward current through the channel at positive voltages and thus helps maintain the plateau phase that controls contraction and prevents premature excitation. However, as the voltage becomes less positive at the end of the plateau phase of repolarization, the channels recover rapidly from inactivation, thus leading to a progressive increase in I _Kr amplitudes during action potential phases 2 and 3, with maximal outward current occurring before the final rapid declining phase of the action potential. The large resurgent I _Kr outward current repolarizes the membrane potential. Next, the channel deactivates (closes) slowly. The resulting large and transient outward current adds considerably to the ongoing repolarization and makes HERG especially suitable for robust control of the repolarization phase.

Regulation

β-Adrenergic stimulation and elevation of intracellular cAMP levels enhance I _Kr amplitude both through PKA-mediated effects and by direct interaction with the protein. α-Adrenergic stimulation is inhibitory. Coexpression of HERG with its β subunit ( KCNE1 or KCNE2 ) accentuates the cAMP-induced voltage shift. The net result of these effects is a reduction in I _Kr .

Extracellular Na ⁺ potently inhibits I _Kr by binding to an outer pore site, and it also speeds recovery from inactivation. The inhibitory effect of Na ⁺ is potently relieved by physiological levels of extracellular K ⁺ . Competition with external K ⁺ for a binding site near the external pore explains the finding that elevation of extracellular K ⁺ concentration paradoxically enhances I _Kr despite the decrease in the electrochemical driving force. Hypokalemia causes prolongation of the action potential duration as a result of reduced K ⁺ conductance. Low extracellular K ⁺ levels accelerate fast inactivation of the HERG channel and increase channel blockade by extracellular Na ⁺ .

Pharmacology

I _Kr is the target of class III antiarrhythmic drugs of the methanesulfonanilide group (almokalant, dofetilide, d-sotalol, E-4031, ibutilide, and MK-499). These drugs produce a voltage- and use-dependent block, shorten open times in a manner consistent with open channel block, and exhibit low affinity for closed and inactivated states. I _Kr blockers prolong atrial and ventricular action potential duration (and the QT interval) and refractoriness in the absence of significant changes in conduction velocity. Although selective I _Kr blockers exhibit antiarrhythmic properties against reentrant arrhythmias, they are probably not effective against triggered activity or increased automaticity.

Selective I _Kr blockers have several disadvantages. These drugs tend to prolong the action potential duration in the Purkinje and midmyocardial cells more than in the subepicardial or subendocardial cells, thus resulting in increased dispersion of repolarization across the ventricular wall and, as a consequence, increased arrhythmogenesis. Moreover, the effects of these drugs increase with decreasing heart rate. This reverse frequency-dependent nature of I _Kr blockers can potentially result in excessive prolongation of the QT interval during bradycardia, potentially precipitating torsades de pointes, whereas this prolongation is much less marked or even absent following β-adrenergic stimulation or during sustained tachycardia. This phenomenon limits the efficacy of these drugs in terminating tachyarrhythmias while maximizing the risk of torsades de pointes during slow heart rates, such as during sinus rhythm after termination of AF. Reverse use-dependence has been attributed, at least in part, to the incomplete deactivation (accumulation) of I _Ks during fast heart rates that leads to a progressive increase in current amplitude, which counteracts the action potential prolongation effects of I _Kr blockers.

Azimilide blocks I _Kr , I _Ks , and I _{Ca L} , whereas amiodarone exhibits a complex mechanism of action because it blocks I _Na , I _Ca , I _Kr , I _Ks , I _to , and I _KATP . Quinidine, a class IA agent, also blocks I _Kr at concentrations lower than those required to block I _Ks , I _to , and I _K1 .

Furthermore, HERG channels display an unusual susceptibility to blockade by a variety of drugs compared with other voltage-gated K ⁺ channels. Increasing numbers of drugs with diverse chemical structures (including some antihistaminics, antipsychotics, and antibiotics) decrease I _Kr by depressing HERG channel gating, delay ventricular repolarization, prolong the QT interval (acquired LQTS), and induce torsades de pointes. In fact, almost all drugs that cause acquired LQTS target HERG channels, likely because of unique structural properties rendering this channel unusually susceptible to a wide range of different drugs. Compared with other cardiac K ⁺ channels, the HERG channel has a large, funnel-like vestibule that allows many small molecules to enter and block the channel. The more spacious inner cavity results from a lack of the S6 helix bending Pro-X-Pro sequence, which presumably facilitates access of drugs to the pore region from the intracellular side of the channel to block the channel current. Additionally, the HERG channel contains two aromatic residues located in the S6 domain facing the channel vestibule (not present in most other K ⁺ channels) that provide high-affinity binding sites for a wide range of structurally diverse compounds. The accessory β subunit (MiRP1, KCNE2 ) also determines the drug sensitivity. Interaction of these compounds with the channel’s pore causes functional alteration of its biophysical properties or occlusion of the permeation pathway.

One novel mechanism for acquired LQTS involves compounds interfering with HERG trafficking (i.e., moving the HERG protein from the endoplasmic reticulum to the cell membrane), rather than direct pore blocking. These compounds include arsenic trioxide, pentamidine, probucol (a cholesterol-lowering therapeutic compound), and cardiac glycosides.

Some drugs (almokalant, norpropoxyphene, azimilide, candesartan, and E3174, the active metabolite of losartan) can enhance I _Kr . Flufenamic acid and niflumic acid also increase I _Kr by accelerating channel opening. These observations open the possibility of developing new I _Kr openers for the treatment of patients with congenital (LQT2) or drug-induced LQTS.

Inherited channelopathies

Acquired diseases

MI can result in reduction in K _v 11.1 mRNA levels and I _Kr with consequent prolongation of the action potential duration ( eFig. 3.3 ). Conversely, I _Kr density increases in subendocardial Purkinje cells in the infarcted heart at 48 hours, which can potentially increase the proarrhythmic effects of I _Kr blockers in patients with MI. Additionally, during acute ischemia, I _Kr is increased and action potential duration is shortened. Such changes can be arrhythmogenic. I _Kr is unchanged in patients with chronic AF and is homogeneously distributed in failing canine hearts.

ATP, derived from either glycolysis or oxidative phosphorylation, is critical for HERG channel function. Both hyperglycemia and hypoglycemia depress I _Kr and can cause QT prolongation and ventricular arrhythmias. In diabetes, K _v 11.1 levels are downregulated, leading to reduction in I _Kr and contributing to QT interval prolongation. Importantly, insulin therapy restores I _Kr function and shortens QT intervals.

Unlike with most other K ⁺ currents, I _Kr amplitude increases on elevation of extracellular K ⁺ concentrations and decreases after removal of extracellular K ⁺ . Elevation of extracellular K ⁺ concentration reduces C/P-type inactivation and increases the single channel conductance of HERG channels. This explains why the action potential durations are shorter at higher extracellular K ⁺ concentrations and longer at low concentrations, and it clarifies the associations among hypokalemia, action potential duration prolongation, and induction of torsades de pointes in patients treated with I _Kr blockers. In contrast, modest elevations of extracellular K ⁺ concentrations using K ⁺ supplements and spironolactone in patients given I _Kr blockers or with LQT2 significantly shorten the QT interval and may prevent torsades de pointes. Moreover, the antiarrhythmic actions of I _Kr blockers can be reversed during ischemia, which is frequently accompanied by elevations of the extracellular K ⁺ concentrations in the narrow intercellular spaces and by catecholamine surges that occur with exercise or other activities associated with fast heart rates.

Structure and physiology

I _Ks is formed by coassembly of four pore-forming α subunits (K _v 7.1, also known as K _v LQT1, encoded by the KCNQ1 gene) and β subunits (minK, encoded by the KCNE1 gene). I _Ks is a K ⁺ -selective current that activates very slowly in response to membrane depolarization to potentials greater than –30 mV and reaches half-maximum activation close to +20 mV. I _Ks has a linear current-voltage relationship; its time course of activation is extremely slow, slower than any other known K ⁺ current; and steady-state amplitude is achieved only with extremely long membrane depolarization.

Inactivation of KCNQ1 channels is half maximal at –18 mV. At its maximum, inactivation reduces fully activated current by approximately 35%. In addition, unlike inactivation of other K _v channels, the onset of I _Ks inactivation occurs after a delay (a delay of approximately 75 milliseconds at +40 mV). In contrast, when inactivation is induced after transient recovery of channels to open states, the onset of inactivation is 10 times faster. The molecular mechanism of KCNQ1 channel inactivation is unknown, but in contrast to a classical C/P-type inactivation, KCNQ1 inactivation is independent of extracellular K ⁺ concentration.

Function

I _Ks contributes to human atrial and ventricular repolarization, particularly during action potentials of long duration. I _Ks gradually increases during the plateau phase of the action potential because its activation is delayed and very slow. As a consequence, the contribution of I _Ks to the net repolarizing current is greatest late in the cardiac action potential plateau phase. I _Ks is expressed in all cell types, but it is reduced in midmyocardial cells. The midmyocardial cells have the longest action potential duration across the myocardial wall.

I _Ks plays an important role in determining the rate-dependent shortening of the cardiac action potential and QT interval. As heart rate increases, I _Ks increases because channel deactivation is slow and incomplete during the shortened diastole. This allows I _Ks channels to accumulate in the open state during rapid heart rates and contribute to the faster rate of repolarization.

Importantly, I _Ks is functionally upregulated when other repolarizing currents (e.g., I _Kr ) are reduced, potentially serving as a safeguard against loss of repolarizing power. As such, several redundant mechanisms contribute to repolarization constituting the repolarization reserve, in which I _Ks plays an important role.

Regulation

I _Ks is markedly enhanced by β-adrenergic stimulation through channel phosphorylation by PKA (requiring A-kinase anchoring protein 9 [AKAP9, also known as Yotiao]) and PKC (requiring minK). This produces a rate-dependent shortening of the action potential duration such as seen during exercise-induced sinus tachycardia. I _Ks is also modulated by α-adrenergic receptors through the PKC pathway. Lowering extracellular K ⁺ and Ca ²⁺ concentrations increases I _Ks .

Coexpression of K _v 7.1 with minK regulates α subunit trafficking and behavior and results in a sevenfold increase in I _Ks magnitude, marked slowing of the time course of activation, and removal (or significant slowing) of inactivation of K _v 7.1 channels.

As noted, I _Kr and I _Ks are functionally linked; when I _Kr is reduced, the action potential is prolonged, causing I _Ks activation to increase to prevent excess repolarization delay. Hence, the duration of the action potential is very tightly tuned via I _Ks and I _Kr regulation.