Inheritable Potassium Channel Diseases

Long QT Syndrome as a K ⁺ Channel Disease

Four subtypes of LQTS (LQT1, LQT2, LQT7, and LQT13) are linked to mutations in genes encoding the pore-forming α-subunits of cardiac K ⁺ channels, whereas three subtypes (LQT5, LQT6, and LQT11) are linked to mutations in genes encoding one of the regulatory subunits of cardiac K ⁺ channels. Until recently, mutations in KCNQ1 (LQT1), KCNH2 (LQT2), KCNJ2 (LQT7; also known as Andersen syndrome ), KCNE1 (LQT5), and KCNE2 (LQT6) were regarded as causative in the etiology of LQTS. Nevertheless, recent reappraisal of genes reported to cause LQTS by three independent gene curation teams of international and multicentered experts, using a standardized and evidence-based framework to score the level of evidence for each gene, classified 9 of the 17 genes as having limited or disputed evidence as LQTS-causing genes. Only the potassium channel genes KCNQ1 (LQT1) and KCNH2 (LQT2) and the sodium channel gene SCN5A (LQT3) were curated as definitive genes for LQTS. Based on these results, it is recommended that genetic variants in some of the LQTS-related genes, including KCNE1, KCNE2, KCNJ2, KCNJ5, and AKAP9, should not be used for clinical decision making before new and sufficient genetic evidence has become available.

Most mutations in genes related to LQTS involve single nucleotide substitutions in the coding regions (exons) of genes that alter a codon and lead to the replacement of one amino acid by a different one (missense mutations; approximately two-thirds of all mutations) or the creation of an early (premature) stop codon resulting in the formation of a truncated protein (nonsense mutations). Single nucleotide substitutions in the noncoding regions (introns) also occur and may result in altered gene transcripts. Specific intronic nucleotide sequences at the intron/exon (acceptor site) and exon/intron (donor site) boundaries are crucial for the process whereby introns are excised from gene transcripts to create mature protein-encoding mRNAs (i.e., the splicing process). Mutations within these highly conserved intronic regions may cause aberrant splicing and lead to the deletion of entire exons (or exon parts) or inclusion of entire introns (or intron parts) in the mature mRNA, thereby often altering the open reading frame of translation and generating a new sequence of amino acids in the final product (i.e., frameshift). Mutations also involve insertion or deletion of one or more nucleotides, which may lead to a shift in the open reading frame, or (when a multiple of three nucleotides are inserted or deleted) to the addition or removal of one or more amino acids in the final product, without affecting the reading frame.

Pathophysiology of Long QT Syndrome

Experimental investigation of mutated K ⁺ channels (or a normal channel with a mutated regulatory subunit) in heterologous expression systems, such as Xenopus oocytes or mammalian cell lines, has unequivocally demonstrated that LQTS-related mutations delay repolarization by reducing the outward K ⁺ current through the affected channel (loss of function). Importantly, the loss-of-function effects of LQTS-related mutations in KCNQ1 and KCNH2 (causing LQT1 and LQT2, respectively) have also been shown in the much more native environment of cardiomyocytes derived from patient-specific pluripotent stem cells from the members of families affected with these diseases. In general, mutations cause a loss of function by decreasing the number of functional channels in the sarcolemma (lower expression) or by disrupting the extent or speed of channel opening and closing (altered gating). Because α-subunits of cardiac K ⁺ channels assemble as dimers or tetramers to form functional ion channels, it is of functional importance whether mutated channel proteins possess the ability to assemble with normal channel proteins in heterozygous mutation carriers where both normal and mutated channels coexist (a normal allele is inherited from one parent and a mutant allele is inherited from the other parent). When mutated subunits assemble with normal subunits, they can disturb the sarcolemmal expression or gating of the normal channel subunits (i.e., a dominant-negative effect). In this case, the overall loss of the K ⁺ current will exceed 50%. In contrast, when mutant subunits do not participate in tetramer assembly, (maximally) 50% reduction of the K ⁺ current is anticipated (i.e., haploinsufficiency). As expected, dominant-negative LQTS–linked mutations have a worse clinical outcome than mutations that cause haploinsufficiency. In addition, LQTS patients with compound mutations (around 8% of all genotyped patients) in the same gene, but particularly in different LQTS-related genes, are also shown to display a more severe phenotype than single mutation carriers. These clinical data indicate that the extent of decrease in cardiac K ⁺ currents may determine disease severity in LQTS.

A delicate balance between inward and outward ion currents determines the repolarization of ventricular myocytes. Substantial differences in the expression levels of K ⁺ channels among cardiomyocytes in different cardiac regions create a spatial dispersion of repolarization within the healthy ventricular myocardium. In LQTS, K ⁺ current reduction leads to prolongation of the AP plateau phase (reflected as QT interval prolongation on the ECG), and this allows for recovery from inactivation and reactivation of L-type Ca ²⁺ channels, which produces early afterdepolarizations (EADs). EADs, together with an accentuated spatial dispersion of repolarization, underlie the substrate and the trigger for the development of torsades de pointes in LQTS. There is consensus that EADs initiate torsades de pointes, and EADs are probably the initiating event in LQT2, the second most prevalent type of LQTS. Nevertheless, in LQT1, the most common type of LQTS, where arrhythmic events usually and predictably start at higher heart rates, the arrhythmogenic mechanism might be different. Delayed afterdepolarizations (DADs), spontaneous APs during phase 4 of the cardiac AP, might play a role in this condition. DADs may also trigger ventricular arrhythmias in LQT7 (see further).

Long QT Syndrome Type 1

To date, over 300 mutations in KCNQ1 have been linked to LQT1. Most mutations are missense mutations (around 70%), followed by frameshift mutations (around 10%), splice-site mutations (around 10%), nonsense mutations (around 5%), and in-frame deletions or insertions (around 5%), and large genomic rearrangements (i.e., copy number variants), leading to the complete deletion of one or more exons. ^, LQT1 mutations in K _V 7.1 are mostly located in the transmembrane segments (around 60%), the intracellular C-terminus (around 30%), and the intracellular N-terminus (around 10%). A fraction of these mutations have been investigated in experimental settings, and these studies have revealed an array of molecular mechanisms that underlie I _Ks loss of function ( Fig. 50.1 ). These mechanisms include (1) defective K _V 7.1 protein synthesis, (2) defective trafficking of mutated K _V 7.1 proteins to the sarcolemma and their retention in the endoplasmic reticulum, (3) impaired ability of mutated proteins to coassemble into tetrameric channels, (4) altered biophysical properties, (5) disrupted interaction with regulatory proteins, and (6) defective endosomal recycling. ^, Of note, these mechanisms are not mutually exclusive. Loss-of-function alterations in biophysical properties of mutated I _Ks channels involve a slower rate of channel activation, shift of the voltage-dependence of activation toward more depolarized membrane potentials (indicating later channel activation), and accelerated deactivation (indicating faster channel closing). Disrupted interaction of mutated K _V 7.1 proteins with the key regulatory protein AKAP9 has been demonstrated to reduce K _V 7.1 phosphorylation by PKA upon β-adrenergic stimulation. Disrupted interaction with phosphatidylinositol-4,5-bisphosphate (PIP2) results in reduced PKC-mediated phosphorylation of the K _V 7.1 proteins. PIP2, a sarcolemmal lipid, increases I _Ks activity through phosphorylation. I _Ks is also increased by the stress hormone cortisol through the action of the serum- and glucocorticoid-inducible kinase 1 (SGK1). Cortisol upregulates SGK1 and thereby facilitates endosomal recycling of K _V 7.1 channels and stimulates their insertion into the sarcolemma. Mutations in K _V 7.1 can disturb this process and cause further I _Ks reduction upon stimulation of SGK1 by cortisol.

Consistent with the molecular signaling pathways, LQT1-related cardiac events occur often during exercise (in particular swimming) and psychological stress, when the adrenergic tone and plasma levels of cortisol are increased. In addition, the QT interval fails to shorten appropriately, and might even lengthen (paradoxic QT response), in LQT1 patients after an increase in heart rate (immediately at standing from supine position or at peak exercise and during the recovery phase of a treadmill exercise testing). In healthy subjects, QT intervals shorten at faster heart rates, enabling QTc to remain within normal limits with decreasing R-R intervals. Moreover, QTc duration lengthening is also observed in LQT1 patients in response to the IV infusion of epinephrine. These clinical features indicate loss of an adequate compensatory response of I _Ks to β-adrenergic stimulation and stress hormones, most probably because of reduced phosphorylation and disrupted endosomal recycling of mutated K _V 7.1 channels. As expected, antiadrenergic therapies such as β-blockers and left stellate ganglion ablation have great efficacy in LQT1 patients. β-blockers have been shown to diminish the QTc changes during exercise or standing and significantly reduce the rate of cardiac events. ^,

Molecular genetics may be useful not only for diagnostic purposes but also for risk stratification in LQT1. In a multicenter study of 600 LQT1 patients, significantly higher rates of cardiac events were found in patients with mutations located in transmembrane segments or with mutations that exerted dominant-negative effects on normal I _Ks channel subunits. Another large multicenter study associated mutations in highly conserved amino acid residues in K _v 7.1 with a significant risk for cardiac events. Moreover, a study of 387 LQT1 patients correlated clinical phenotype with changes in biophysical properties of K _V 7.1 channels caused by different KCNQ1 mutations. In particular, a slower rate of channel activation was associated with an increased risk for events in LQT1. In all these studies, the effects of the mutations were independent of traditional risk factors (i.e., QTc, female gender, and β-blocker therapy). Mutation type and location may also be used to predict whether a KCNQ1 mutation is pathogenic or an innocuous rare variant. This is clinically relevant because large overlap of QTc values exists between patients with LQTS and healthy subjects. Non-missense mutations, regardless of location, have an estimated predictive value of more than 99% to be pathogenic, whereas missense mutations have a high predictive value when located in the transmembrane segments, the pore loop, and the C terminus of K _V 7.1 proteins. LQT1 patients with a mutation in the cytoplasmic loops are at higher risk for lethal arrhythmias than patients with a mutation in other regions of K _V 7.1, probably because of a pronounced reduction in channel activation upon β-adrenergic stimulation.

QT interval duration and risk for cardiac events in LQT1 may not only be determined by mutation type and location but also by the copresence of common variants in KCNQ1, KCNE1 (encoding MinK), AKAP9 (encoding Yotiao), and NOS1AP. In particular, single nucleotide polymorphisms (SNPs) in the nitric oxide 1 adaptor protein gene ( NOS1AP ) have been identified as strong modulators of phenotype in LQT1 (and other LQTS types). First, genome-wide association studies associated SNPs in NOS1AP with QT interval in the general population. Next, a role for NOS1AP SNPs was found in sudden cardiac death in the general population. In 2009, a family-based association analysis linked SNPs in NOS1AP with QT interval prolongation and a risk for cardiac arrest and sudden death in a large South African cohort of LQT1 patients with the A341V mutation. NOS1AP encodes CAPON, an accessory protein of the neural nitric oxide synthase, which controls intracellular nitric oxide production. Functional studies indicate that CAPON fastens cardiac repolarization by inhibiting L-type Ca ²⁺ channels. This provides a rationale for the association of SNPs in NOS1AP with QT interval duration. SNPs in the 3′ untranslated region (3′UTR) of KCNQ1 have also been shown to modulate phenotype in LQT. In a study in two independent LQT1 cohorts from the Netherlands and the United States, SNPs in KCNQ1 ’s 3′UTR were associated with QTc duration and symptomatology in an allele-specific manner. Patients with the derived SNP variants on their mutated KCNQ1 allele had shorter QTc and fewer symptoms, whereas patients with these variants on their normal KCNQ1 allele had longer QTc and more symptoms. Experimental studies showed that the expression of KCNQ1’s 3′UTR with the derived SNP variants was lower than the expression of the 3′UTR with the ancestral SNP variants. The 3′UTR play a crucial regulatory role in gene expression by controlling stability and translation of mRNAs. SNPs in this region were suggested to affect this function of the 3′UTR, thereby altering gene expression in an allele-specific manner. If true, this is expected to be especially relevant when one allele contains a pathogenic mutation. Nevertheless, these findings could not be reproduced in 3 LQTS cohorts with a KCNQ1 founder mutation and thus await replication in larger LQT1 cohorts before their clinical use is proposed.

Long QT Syndrome Type 5

The genetic evidence that variants in KCNE1 can cause LQTS is limited. Nevertheless, emerging data suggest that (common) variants in KCNE1 may lead to mild QT prolongation or a predisposition to acquired LQTS. ^, The KCNE -encoded MinK exerts its regulatory effects on K _V 7.1 via interactions between the transmembrane segments and the C-termini of the two proteins. In general, interactions between transmembrane segments are believed to be essential for normal channel activation, whereas interaction between the C-termini regulate channel assembly and channel deactivation. Most KCNE1 variants associated with LQTS are believed to impair the ability of MinK to modulate gating properties of K _V 7.1. Other mechanisms include defective trafficking of MinK proteins (and thereby K _V 7.1 proteins), impaired channel assembly, and reduced sensitivity to PIP2. Residues in the C-terminus of MinK have been identified as key determinants for sensitivity of K _V 7.1 to PIP2, and LQTS-linked variations in these residues have been shown to reduce PIP2-mediated phosphorylation of K _V 7.1 proteins by PKC. A SNP located in the C-terminus of MinK (D85N) has been associated with QT interval duration in the general population and has been shown to be more prevalent in (genotype-negative) LQTS patients. Heterologous expression studies revealed significant loss-of-function effects of D85N on KCNQ1 -encoded and KCNH 2-encoded currents. Therefore D85N is suggested to cause a mild form of LQTS (known as “LQT-lite”) and act as a modifier of phenotype in LQT1 and LQT2 patients.

Long QT Syndrome Type 11

AKAP9 (Yotiao) has an N-terminal and a C-terminal binding domain that interacts with the C-terminus of K _v 7.1. So far, only two variants in AKAP9 have been identified in single subjects with LQTS. One of these variants has been shown to disrupt, but not eliminate, the interaction between AKAP9 and K _V 7.1, leading to reduced cyclic adenosine monophosphate (cAMP)-mediated phosphorylation of K _V 7.1 by PKA during β-adrenergic stimulation. This is speculated to result in less I _Ks enhancement during sympathetic activity. Although the role of AKAP9 in the etiology of LQTS has been challenged because of a lack of substantial evidence, genetic variations in AKAP9 have been proposed to act as modifiers of phenotype in LQT1.