Deafness-associated genes KCNQ1 (also associated with heart diseases) and KCNQ4 (only associated with hearing loss) encode the homotetrameric voltage-gated potassium ion channels Kv7.1 and Kv7.4, respectively. To date, over 700 KCNQ1 and over 70 KCNQ4 variants have been identified in patients. The vast majority of these variants are inherited dominantly, and their pathogenicity is often explained by dominant-negative inhibition or haploinsufficiency.
1. Introduction
The human genome encodes 40 voltage-dependent potassium ion (K
+) channels that belong to 12 subfamilies (Kv1–Kv12). The voltage-dependent activities of Kv channels play critical roles in controlling the electrophysiological properties of cells and maintaining the ion homeostasis. All Kv channels possess six transmembrane (TM) segments (S1–S6). The first four TM segments (S1–S4) constitute the voltage-sensing domain, while the last two, which flank a channel pore loop (P-loop), constitute the pore-forming domain (S5-PH-S6). Tetramerization is essential for completing the K
+-selective channel pore in all Kv channel types. Two Kv channels, Kv7.1 and Kv7.4, are essential for normal operation of the inner ear
[1][2].
Kv7.1 is encoded by
KCNQ1. It is abundantly expressed in the heart and is crucial for normal repolarization of cardiomyocytes. Mutations in Kv7.1 underlie two forms of the long QT syndrome (LQTS), i.e., the Romano–Ward syndrome (RWS)
[3] and Jervell and Lange–Nielsen syndrome (JLNS)
[4]. Cardiac symptoms in JLNS are typically more severe than those in RWS, and most patients with JLNS also suffer from congenital hearing loss
[4]. RWS is inherited dominantly, with some exceptions
[5][6], whereas JLNS is inherited recessively. In the cochlea, Kv7.1 is expressed in the marginal cells of the stria vascularis (SV)
[4]. Kv7.1 is thought to mediate the secretion of K
+ into the endolymph and the establishment of the endocochlear potential (EP)
[7].
Kv7.4 is encoded by
KCNQ4. In the cochlea, two types of sound-sensing cells, inner hair cells (IHCs) and outer hair cells (OHCs), are housed in the organ of Corti, and their apical surfaces exposed to the endolymph. Kv7.4 is abundantly expressed in OHCs, but it is also expressed in IHCs and the spiral ganglion neurons (SGNs)
[8][9]. The K
+ conductance mediated by Kv7.4 contributes to the establishment of a normal resting membrane potential and is crucial for repolarization of the cells after sound-elicited cell depolarization. The large Kv7.4-mediated conductance in OHCs also contributes to the reduction of the membrane time constant so that the receptor-potential-induced mechanical response of OHCs, i.e., electromotility
[10][11], can respond to sound stimuli at high frequencies
[12]. K
+ that flows into OHCs from the high-K
+-containing endolymph via the stereocilia is thought to be extruded by Kv7.4 from the base of OHCs to the perilymph. This Kv7.4-mediated extrusion of K
+ is believed to be crucial for maintaining the intracellular ionic homeostasis and, thus, for OHC maintenance. Mutations in Kv7.4 are responsible for dominantly inherited progressive nonsyndromic hearing loss, DFNA2A
[13].
2. The Pathologies of JLNS and DFNA2A Hearing Loss
Hearing loss in JLNS patients is congenital, bilateral, and profound (OMIN: 220400). To date, 768
KCNQ1 variants have been reported, ~5% of which are JLNS-associated (the Human Gene Mutation Database, HGMD)
[14]. The ionic composition of endolymph is unique among extracellular fluids in that it contains high K
+ (~160 mM) but low Na
+ (~1 mM) and Ca
2+ (~20 µM)
[15]. The high K
+ in endolymph is due to the secretion of K
+ from the marginal cells of the SV, which is powered by ATP-dependent Na
+/K
+ pumps, a Na
+/K
+/2Cl
− cotransporter, and other ion transporters. K
+ transport from the marginal cells to the endolymph is electrogenic, resulting in positive EP (+80 to +100 mV)
[16]. This very positive EP is a main driving force for sensory mechanotransduction by both hair cells (conversion of sound-induced mechanical vibrations sensed by stereocilia into changes in the receptor potential). Kv7.1, together with its ancillary protein, KCNE1, is thought to mediate K
+ transport from the marginal cells to the endolymph. Thus, JLNS-causing
KCNQ1 variants are presumed to affect EP by impairing the K
+ conductance of Kv7.1. In fact,
Kcnq1−/− mice lose EP and show atrophy of the SV, reduction of the endolymphatic compartment, and degeneration of the organ of Corti and SGNs
[17][18][19][20].
Hearing loss in DFNA2A patients is typically progressive and more prominent at higher frequencies (middle and low frequencies are also affected later in life) (OMIN: 600101). The severity of DFNA2A hearing loss and the rate of progression vary among
KCNQ4 variants. To date, 76
KCNQ4 variants have been reported in HGMD. Given the multiple aforementioned roles of Kv7.4 in OHCs, it is anticipated that DFNA2A-causing
KCNQ4 variants impair normal cochlear operation by primarily affecting OHCs.
Kcnq4−/−,
Kcnq4G286S/+, and
Kcnq4G286S/G286S (p.G285S in humans) mice recapitulate progressive hearing loss that is, indeed, accompanied by OHC dysfunction and degeneration
[21]. Degeneration of IHCs and SGNs at later postnatal ages was also found
[22].
The presence of multiple JLNS-/DFNA2A-associated variants found in KCNQ1/KCNQ4 and identification of hearing phenotypes in Kcnq1 and Kcnq4 mouse models compellingly establish the essentiality of these Kv7 channels in hearing.
3. The Pathological Roles of Kv7.1 and Kv7.4 Variants and Ongoing Pharmacological Strategies
A dominant-negative inhibition (inhibition of the Kv7 channel function or membrane targeting by having a mutated subunit in the tetramer complex) often underlies the observed dominant inheritance of Kv7.1 and Kv7.4 variants associated with RWS and DFNA2A. Since tetramerization of Kv7 channels is mediated by the C-terminal cytosolic helices C and D
[23][24][25][26][27][28][29][30][31], haploinsufficiency is presumed to account for the dominant inheritance of frame-shifted or nonsense Kv7 variants that lack the C-terminal tetramerization region. In any case, the functional consequence of Kv7 variants would be the reduction of overall K
+ channel activity. Hence, pharmacological augmentation of reduced residual Kv7 channel activity by channel openers is considered clinically effective and actively pursued
[30][31][32][33][34][35][36][37][38][39][40][41][42][43].
4. Observations against a Haploinsufficiency-Based Pathological Mechanism
Kv7.1 variants associated with JLNS are inherited recessively. In other words, individuals heterozygous for JLNS-associated Kv7.1 variants suffer from neither LQTS nor hearing loss. This suggests that one functional KCNQ1 allele is sufficient for maintaining normal cardiac and auditory functions, thus arguing against a haploinsufficiency-based pathological mechanism. It is perplexing that, among truncated Kv7.1 variants lacking the C-terminal tetramerization region, some are inherited dominantly, while the others are inherited recessively.
Multiple variants truncating the C-terminal tetramerization region are also identified in
KCNQ4. If haploinsufficiency accounted for dominant inheritance of these truncated Kv7.4 variants, they should all result in a similar and a relatively mild DFNA2A phenotype compared to those that exert a dominant-negative inhibitory effect on the wild-type Kv7.4 (Kv7.4
WT) subunit. The slowly progressive hearing loss found in patients with heterozygous Kv7.4
Q71Sfs (c.211delC)
[44] is in line with such a view. However, Kv7.4
W242X (c.725G>A), which completely lacks the channel pore-forming transmembrane domain and the following C-terminal tetramerization region, is associated with severe to profound hearing loss
[45]. The presence of recessively inherited truncated Kv7.4 variant, Kv7.4
A349Pfs (c.1044_1051del8)
[46], is also incompatible with a haploinsufficiency-based pathological mechanism. Incidentally,
Kcnq4+/− heterozygous mice do not suffer from hearing loss
[21], suggesting that one functional
KCNQ4 allele is sufficient for maintaining normal auditory function, at least in mice.
5. Identification of Cell Death-Inducing Cytotoxicity in Truncated Kv7.1 and Kv7.4 Variants
In order to elucidate the pathogenic roles of truncated Kv7 variants, researchers first tried to confirm the absence of K
+ channel activity for three deafness-associated Kv7.4 variants that completely lack the C-terminal tetramerization region, Kv7.4
Q71Sfs, Kv7.4
W242X, and Kv7.4
A349Pfs, in HEK293T-based stable cell lines. Unexpectedly, researchers encountered great difficulties in establishing stable cell lines that were to constitutively express these variants, especially for Kv7.4
W242X. After confirming the reproducibility of this issue, researchers suspected inherent cytotoxicity in these truncated variants. By performing CellTox Green Cytotoxicity and RealTime-Glo Annexin V Apoptosis assays using doxycycline-inducible stable cell lines, researchers found that these three Kv7.4 variants are indeed cytotoxic and induce cell death to various degrees in a doxycycline-dosage-dependent manner
[47]. The cell-death-inducing cytotoxicity of these Kv7.4 variants was further confirmed in an HEI-OC1 cell line that was derived from the murine inner ear
[47]. Researchers also found that none of these truncated Kv7.4 variants were functional by themselves (when singly expressed); nor were they capable of forming a heteromer with Kv7.4
WT [47], refuting the possibility that Kv7.4
W242X and Kv7.4
A349Pfs physically interact with Kv7.4
WT to exert either an inhibitory (dominant-negative) or cooperative effect.
A CellTox Green Cytotoxicity assay was also performed for HEK293T-based doxycycline-inducible stable cell lines expressing truncated Kv7.1 variants lacking the C-terminal tetramerization domain, Kv7.1
E261X, Kv7.1
W305X, Kv7.1
Q530X, and Kv7.1
Q531X, among which solely Kv7.1
Q530X was JLNS-associated
[48]. Large cell-death-inducing cytotoxicity was found in Kv7.1
E261X and Kv7.1
W305X, while small cytotoxicity was found in Kv7.1
Q530X and Kv7.1
Q531X. A previous study showed that Kv7.1
Q530X does not exert a dominant-negative inhibitory effect on wild-type Kv7.1 (Kv7.1
WT)
[5][49]. Consistently, another study showed that Kv7.1
Q530X does not bind to Kv7.1
WT [50]. It is probable that the other three Kv7.1 variants, which are truncated similarly (Kv7.1
Q531X) or are shorter (Kv7.1
E261X and Kv7.1W
305X) compared to Kv7.1
Q530X, are also incapable of physically interacting with Kv7.1
WT.
Collectively, these observations affirm the inadequacy of a pathological mechanism that is based on dominant-negative inhibition or haploinsufficiency alone for explaining the mode of inheritance of some Kv7 variants.
6. Updates on the Pathological Mechanisms of JLNS and DFNA2A
A dominant-negative inhibition-based pathological mechanism has been experimentally validated for many dominantly inherited Kv7.1 and Kv7.4 variants. Researchers confirmed that two missense Kv7.4 variants, Kv7.4
G285S and Kv7.4
P291L, can interact with Kv7.4
WT, and that their cytotoxicity is WT-like
[47]. Thus, the study does not challenge the dominant-negative inhibition-based pathological mechanism (at least for these two Kv7.4 missense variants). However, to the best of my knowledge, a haploinsufficiency-based pathological mechanism has never been firmly demonstrated for any Kv7 variants. As mentioned above, a haploinsufficiency-based pathological mechanism is incompatible with several observations and, thus, not compelling.
The cell-death-inducing cytotoxicity identified in the recent study
[47] provides a straightforward explanation for the observed dominant inheritance of truncated Kv7 variants lacking the ability to exert a dominant-negative inhibitory effect. It can also explain why some Kv7.1 variants that are associated with severe cardiac phenotypes are not always co-associated with proportionally longer corrected QT (QTc) intervals
[51][52][53][54]. For example, a sixteen-year-old proband with Kv7.1
W305X died suddenly, and two other affected members experienced syncopal episodes, although the resting QTc interval of patients with this Kv7.1 variant was found to be normal to borderline
[54]. The normal-like QTc interval seems reasonable because it is unlikely that Kv7.1
W305X exerts a dominant-negative inhibitory effect on Kv7.1
WT and because one functional
KCNQ1 allele is sufficient for supporting normal cardiac function (see above). It is conceivable that the severe cardiac phenotypes found in patients with Kv7.1
W305X are ascribed to the large cell-death-inducing cytotoxicity found in this variant
[47]. The Kv7.1
WT-like small cytotoxicity found in recessively inherited Kv7.1
Q530X among four truncated Kv7.1 variants tested
[47] is compatible with such a view.
The vast majority of Kv7.1 variants are inherited dominantly and are associated with LQTS (RWS), but not with hearing loss. Homozygous or compound heterozygous Kv7.1 mutations that are not associated with hearing loss are also reported
[5][6]. These observations suggest that the cardiac function is more sensitive to the reduction of Kv7.1 function compared to the auditory function. Thus, as it has been presumed, JLNS hearing loss is probably caused by drastic or complete loss of Kv7.1-mediated K
+ channel activity. Pathological contributions of cytotoxicity of Kv7.1 variants are likely minimal in JLNS, given the observed recessive inheritance of JLNS variants.