The Membrane Protein KCNQ1 Potassium Ion Channel: Functional Diversity and Current Structural Insights

Abstract

Background:

Ion channels play crucial roles in cellular biology, physiology, and communication including sensory perception. Voltage-gated potassium (Kv) channels execute their function by sensor activation, pore-coupling, and pore opening leading to K⁺ conductance.

Scope of review:

This review focuses on a voltage-gated K⁺ ion channel KCNQ1 (Kv 7.1). Firstly, discussing its positioning in the human ion chanome, and the role of KCNQ1 in the multitude of cellular processes. Next, we discuss the overall channel architecture and current structural insights on KCNQ1. Finally, the gating mechanism involving members of the KCNE family and its interaction with non-KCNE partners.

Major conclusions:

KCNQ1 executes its important physiological functions via interacting with KCNE1 and non-KCNE1 proteins/molecules: calmodulin, PIP₂, PKA. Although, KCNQ1 has been studied in great detail, several aspects of the channel structure and function still remain unexplored. This review emphasizes the structural and biophysical studies of KCNQ1, its interaction with KCNE1 and non-KCNE1 proteins and focuses on several seminal findings showing the role of VSD and the pore domain in the channel activation and gating properties.

General significance:

KCNQ1 mutations can result in channel defects and lead to several diseases including atrial fibrillation and long QT syndrome. Therefore, a thorough structure-function understanding of this channel complex is essential to understand its role in both normal and disease biology. Moreover, unraveling the molecular mechanisms underlying the regulation of this channel complex will help to find therapeutic strategies for several diseases.

Graphical abstract

1. Introduction

Ion channels are a ubiquitous class of pore-forming proteins that facilitate the passage of ions across the lipid bilayer in multicellular organisms. Myriad crucial physiological and biological functions are modulated by ion channels, including neurotransmitter release, muscle contraction, electrolyte balance, hormone secretion, sensory transduction, and cognition [1–5].

The classification of ion channels is based on the nature of the gating stimulus, which could be transmembrane voltage potential, chemical ligands, light, temperature, secondary messengers, or mechanical. Yu and Catterall (2004), proposed the term “Chanome” comprising all ion channel proteins, similar to “kinome” for the protein kinase superfamily¹ (Figure 1 A). The chanome was further sub-grouped into a VGL-chanome: the superfamily of voltage-gated-like ion channels, and a LG-chanome: ligand-gated ion channels. Ion channels that execute passive ion flux driven by an electrochemical gradient and are regulated by binding of small molecule ligands are called ligand-gated ion channels (LGICs) [1–3]. The gating process of LGIC conductance involves the binding of specific ligands and subsequent conformational changes leading to the opening of the channel. LGICs play important roles in the somatic neuromuscular junction and fast synaptic transmission and include glycine receptors, P2X receptors, the nicotinic acetylcholine receptor (nAchR), ionotropic glutamate (NMDA, AMPA, and Kainate receptors), 5-HT₃, GABA, IP₃ receptor, acid-sensing (proton-gated) ion channels (ASICs), epithelial sodium channels (ENaC), and zinc-activated channels (ZAC) [1,4–7]. Voltage-gated ion channels are regulated by changes in the transmembrane voltage potential and classified as „voltage-gated-like (VGL) ion channel chanome‟, which includes the voltage-gated potassium ion channel superfamily [1,3–6] (Kv superfamily, Figure 1B). VGLs consists of distinct structural domains with transmembrane helices S1–S4, comprising the voltage sensor domain (VSD), and S5–S6, forming the pore domain (PD) [7]. The general gating mechanism of VGLs involves movement of the S4 transmembrane helix of VSD towards the extracellular region. The voltage sensor in S4 helix is lined by several basic amino acid residues that sense the change in membrane potential. The S4 helix then translates toward the extracellular milieu, resulting in structural changes that are propagated to the PD via the S4–S5 linker and/ by the P-loop region between S5 and S6, which contains the selectivity filter [7,8].

Figure 1.

(A) Ion-channel chanome depicting voltage-gated ion channel superfamily with different sub-classes. (B) Floral diagram depiction of human potassium ion channel (Kv) superfamily and the relative position of KCNQ1 (Kv 7.1) in the channel superfamily.

Heritable mutations in the genes that encode ion channels can result in channel dysfunction or loss of function, causing diseases collectively termed as channelopathies [4–6, 9–13]. Aberrant loss of function gain of function, or dysfunction can result in diseases like epilepsy, asthma, cystic fibrosis, myasthenia gravis, cancer, cardiac arrhythmia, diabetes, deafness, migraine, blindness, hypertension, and multiple sclerosis [4–6, 9–13]. Ion channels have been successfully targeted with several classes of drugs. For example, volatile anesthetic drugs primarily block LGICs. Gamma-aminobutyric acid analogs (gabapentinoids) are used to treat epilepsy and anxiety disorder. Calcium channel blockers (CCBs) are used to treat hypertension patients, whereas, sodium channel blockers are primarily used to treat epilepsy [4–6, 9–13]. I_KS channel blockers are a promising class of antiarrhythmic drugs. Chromanol 293B is an I_KS channel blocker responsible for the prolongation of cardiac action potential. Lerche et al. (2000) showed the inner pore domain of KCNQ1 to be a molecular target of Chromanol 293B. They also identified amino acid mutations V307L and I337V, located in the evolutionary conserved pore loop (PD) region and the S6 transmembrane domain of KCNQ1 almost completely inhibited Chromanol 293B effect [14]. Same effect was observed with mutant KCNQ1/E1 channel. However, mutations involving KCNE1 in the KCNQ1/E1 complex, and its effect on inhibition by Chromanol 293B revealed no direct involvement of KCNE1 subunit as demonstrated by I_KS current measurements. It is asserted that KCNE1 acts allosterically to alter the conformation of KCNQ1to modulate the channel properties [14]. Similar findings were reported by Xu et al., where KCNE1 induced backbone displacement in the selectivity filter region (T312) of KCNQ1. However, no direct contact between KCNE1 and KCNQ1 PD was observed indicating allosteric modulation of the KCNQ1 selectivity filter in the presence of KCNE1 [13].

Several drugs elicit cardiotoxic effects by binding to the KCNQ1/E1 complex, blocking the conduction of K⁺ ions and disturbing the generation of I_KS currents [15]. Fluoxetine, an antidepressant drug has been shown to induce LQTS at high serum levels [15]. It acts by blocking I_Kr and also inhibiting the protein trafficking of I_Kr channel, thereby further reducing I_Kr. Fluoxetine and its derivative norfluoxetine also inhibit KCNQ1/E1 currents leading to QT prolongation. In the presence of K422T mutation in KCNQ1, QT prolongation lasted for 154 ms. Celecoxib, a non-specific K_v blocker is known to block KCNQ1 when coexpressed with KCNE1 subunits [16]. However, previous studies show that celecoxib acts as a KCNQ5 channel opener [17]. The mechanism underlying this opposite regulation remains elusive.

Channel proteins and their functional complexes have been studied using various biophysical techniques [18–27]. These include structural studies that have been instrumental in illuminating the functional properties of different ion channels and their channelopathic landscapes. This review is focused on KCNQ1, a voltage-gated potassium ion channel that has been of much interest over the past three decades because of its direct role in the cardiovascular diseases such as atrial fibrillation (AF) and long QT syndrome (LQTS) [10–13]. In this review, we provide a comprehensive and up-to-date understanding of the structure and function of KCNQ1 and also comment on future direction of this research field.

Functional versatility of KCNQ1 and ‘KCNQ1 channelopathy’

Biological membranes, as well as membrane proteins play crucial roles in normal and disease biology and therefore have been studied in the great detail [28–30]. Ion channels are a distinct class of membrane proteins regulating key physiological processes in the human body (Figure 2 A, B). KCNQ1 is expressed ubiquitously [31–33] and is the pore-forming subunit of the slow delayed rectifier current (I_KS) channel complex [31–47] (Figure 2 A). The functional role of KCNQ1 is thoroughly studied both in the heart [34–37] and in the inner ear [10,13,38]. KCNQ1 forms a complex with the β-subunit KCNE1 (MinK) in the cardiac I_KS channel to regulate the cardiac action potential duration [34–37]. The outward K⁺ current generated by KCNQ1/KCNE1 complex in the I_KS channel in heart, is one of the repolarizing K⁺ currents that contribute to the cessation of the cardiac action potential. Defective cardiac I_KS channels (originating from dysfunctional KCNQ1 protein) cause pathological changes in the action potential duration of the heart, leading to severe cardiac arrhythmias and long QT syndrome (due to a prolonged cardiac action potential duration) [10–13].

Figure 2.

(A) Physiological roles of KCNQ1 in human body (Figure design inspired from Liin et al.; ref. 33). (B) mutated or dysfunctional KCNQ1 is associated with various diseases.

The role of the KCNQ1/KCNE1 complex has been thoroughly studied in ear biology [10,13,38]. Inner ear endolymph K⁺ homeostasis is maintained by the K⁺ flux through I_KS channel (KCNQ1/KCNE1 complex) generating an endocochlear potential [10,13,38]. Dysfunction of I_KS channel results in decreased hearing or congenital deafness [13]. Mutations in KCNQ1 or KCNE1 are associated with two forms of Long QT syndrome [39]. Romano Ward syndrome, an autosomal dominant disorder and Jervell and Lange-Nielson syndrome, inherited as an autosomal recessive trait. The latter being less common but with severe sensorineural hearing loss (SNHL) [39]. The pathophysiology of this hearing impairment along with Long QT syndrome in Jervell and Lange-Nielson syndrome remains unclear. Besides heart and ear, KCNQ1 has been reported to form a channel complex with either KCNE1 or other members of KCNE family (E1–E5) members in the pancreas, kidney, colon and intestine, stomach, thyroid gland and airways [40–48] (Figure 2 A). Similar to the ear, the major function of the KCNQ1/KCNE complexes in these organs, is to contribute towards K⁺ homeostasis. For example, the K CNQ1/KCNE3 complex also plays a significant role in the maintenance of the proper membrane potential for transport of ions across the epithelium, for example, Cl⁻ and gastric acid [42,45,47,48] (Figure 2 A). In the intestine, the KCNQ1/KCNE3 complex regulates Cl⁻ ion secretion in a cAMP dependent manner. This is achieved by constitutively activating the open state of the KCNQ1/KCNE3 channel complex, that aids in the transepithelial transport of Cl⁻ ions via basolateral recycling of K⁺ ions and by increasing the driving force generated by K⁺ efflux for Cl⁻ secretion [47]. KCNQ1 affects the endocrinological functions by regulating the physiology of thyroid gland. KCNQ1/KCNE2 form a constitutively active K⁺ channel in the thyrocytes required for normal thyroid hormone biosynthesis [49]. The brain expresses KCNQ1 where it contributes towards neuronal excitability [11,12]. Major reported disease groups from dysfunctional KCNQ1 are long and short QT syndrome, Romano Ward syndrome, Gestational diabetes, Familial atrial fibrillation, and Jervell and Lange-Nielsen syndrome (Figure 2 B). Bartos et al. identified an R231H mutation in KCNQ1 that is responsible for interfamilial early onset of Atrial Fibrillation (AF) [50]. R231H variants showed increased KCNQ1 current (IKCNQ1) resulting in shortened atrial action potential (AP) duration. R231H was also found to be resistant towards PKA stimulation, suggesting a potential role of PKA in regulating IKCNQ1 in these variants [50]. Based on the reported expression of KCNQ1–KCNE complexes in various body tissues with different functions, the multi-functional aspects of this channel protein in normal and disease physiology is obvious.

2. KCNQ1 channel architecture and structural insights

The functional unit of KCNQ1 consists of a tetramer with each monomer containing a short N-terminal S0 helix, followed by six transmembrane (S1–S6) helices and four intracellular helices in the C-terminal domain [51–53] (Figure 3). Similar to canonical VGLs, helices S5 and S6 ofKCNQ1 line the interior of the complex in the tetramer forming the pore domain (PD). The PD includes the signature GYGD sequence and assists in K⁺ conductance across the channel [51–56]. The topology of the VSD is similar to other membrane-spanning 4 -helical bundle proteins found in nature [57–59]. However, several key structural features of KCNQ1 present a sharp contrast with not only with other members of the Kv superfamily but with all of the other VGLs characterized so far, making it unique with respect to voltage dependence and gating properties. Sun et al. (2017) pin-pointed two key structural differences in the VSD region of KCNQ1 (Q1-VSD) from Xenopus laevis (which shares 78% sequence identity with human KCNQ1) and other VSDs of Kv channels; 1) a nine amino-acid α-helical insertion in the S2–S3 linker region, and 2) a straight S3 helix throughout its length due to the lack of a conserved proline residue present in most Kv channels but absent in KCNQ1 [53]. KCNQ1 presents a unique functional feature when compared to other VGLs, where it can switch from the voltage dependent state to the constitutively active state. The voltage sensor domain (Q1-VSD) can lose its voltage dependence upon association with members of the KCNE family making it constitutively active [35].

Figure 3.

Topological depiction of KCNQ1 in a membrane environment.

Several basic amino acid residues line the S4 helix of Q1-VSD [53,56,60]. Upon membrane depolarization, these positively charged residues sense the change in membrane potential causing the S4 helix to move towards the extracellular region [53,56,60] (Figure 4A). This is accomplished by electrostatic interactions between basic residues residing in the S4 region and acidic residues called “gating charge transfer residues” within the S2 and S3 helical regions [53,61].

Figure 4.

(A) Membrane topology of KCNQ1-Voltage Sensor Domain (Q1-VSD) showing the movement of S4 helix during VSD activation. (B) Important role played by KCNQ1–S6 and S4 helices and their interaction with KCNE1 and KCNE3 (depited as E1 and E3 (yellow) in 4B and 4C respectively) proteins in the opening/closing (gating mechanism) of the channel complex.

A recent cryo-EM structure of KCNQ1 by Sun et al. [53] revealed the structural details underlying the decoupled-depolarized Q1-VSD conformation. While the pore of the channel in this structure is closed, the voltage sensor domain is in the fully activated state. They reported crucial positioning of F157 residue with respect to the gating charges that cap the charge transfer center in the depolarized conformation of Q1-VSD. Another distinct feature, suggesting the uncoupled VSD-PD conformation in the case of KCNQ1, is a structural perturbation that involves shift of a loop by ~10.5 Å near residues 234–236 along with the S4–S5 linker facing away from the S6 helix compared to Kv1.2–2.1. The loop region connecting the S5–S6 in KCNQ1 is lined by negatively charged amino-acids while S6 lines the inner pore of the channel. KCNQ1 does not have a conserved PXP motif (a salient feature of Kv channels) but it is replaced by a functionally significant PAG motif, which is involved in channel activation [60,62]. The soluble C-terminal region consists of four helices (HA-HD). Helices HA and HB interact with Calmodulin (CaM), an important player in the channel function and disease biology [53]. Helices HC and HD are thought to assist in the tetramerization of the channel [53].

One of the essential components of a functional KCNQ1 complex in the heart is its association with the KCNE1 subunit (minK or β) (Figure 5 A). KCNE family members (KCNE1–5) are single-pass transmembrane proteins which associate with KCNQ1 to form the channel complex. Previous studies suggest that up to 4 KCNE subunits could associate with KCNQ1 tetramer [63]. However, the optimal I_KS is generated when there is 4:2 ratio between KCNQ1-KCNE1 [64]. Cysteine cross-linking studies have shown that KCNE1 aligns itself in the cleft between the two Q1-VSDs in the channel complex [65,66] (Figure 5 B). Association with KCNE1 slows down the activation kinetics of I_KS which is crucial during the repolarization phase of cardiac action potential [34,67].

Figure 5.

(A) KCNQ1-KCNE1 channel complex, where each subunit of KCNQ1 is labeled as 1–4 and S1–S6 helices of KCNQ1 are shown in different colors. (B) Top cartoon view of the complex showing VSD in the cleft between KCNQ1 and KCNE1 (shown as E1 in red) monomers. The pore domain is surrounded by blue colored S5–S6 helices.

Hasani et al. used molecular dynamic (MD) simulations and compared un-complexed KCNQ1 and KCNQ1/KCNE1 with respect to K⁺ ion transport across the membrane [68]. They reported four binding sites for K⁺ ions in the selectivity filter region (TIGYG) of KCNQ1 protein that direct its passage through the filter. The presence of KCNE1 causes constriction of the pore, slowing down the passage of K⁺ ions without causing significant changes to the selectivity filter region, thereby slowing down the activation kinetics of KCNQ1, as reported by Xu et al [69]. As opposed to previous studies, Xu et al. found that side chain interactions predominantly dictated the interactions between KCNQ1 and KCNE1 in the extracellular region, which impacted the voltage dependent activation of the channel [69]. This change could be attributed to the inward movement of S6 helical region in the presence of KCNE1.

3. KCNQ1 voltage dependent gating mechanism

Similar to other Kv channels, KCNQ1 follows a sequential, voltage dependent three step activation, switching from the initial resting state to an intermediate state followed by a fully activated channel [70,71]. We recently reported the topology of the isolated Q1-VSD in the lipid bilayer in its intermediate state [72]. KCNQ1 is unique in the fact that the pore remains open in both the intermediate as well as in the activated state of the Q1-VSD. However, both states demonstrate different gating and permeation properties indicating that the PD-VSD interactions play a role in determining the channel conformation and K⁺ transport.

The S4 helix of Q1-VSD plays an important role in the activation of the voltage sensor. This is assisted by several conserved arginine residues localized in S4 [73]. KCNQ1 differs from other canonical Kv channels in that it has fewer basic amino-acid residues lining the S4 helix. Previous studies have shown that Q1-VSD activation is a two-step process involving interactions between E160 in S2 and S4 arginines [74]. Charge reversal mutations revealed that while the E160R mutation caused loss of current in voltage-clamp fluorometry (VCF) experiments, charge reversal at R231(R231E) or R237 (R237E) partially restored it indicating the arrest of VSD in the states where these residues interact. It was also observed that the mutant pairs E160R/R231E and E160R/R237E led to constitutively open states of the channel [56]. Mutations in these critical S4 arginines alter the voltage dependence of channel activation. The effect of KCNE1 in these states is discussed in the later section. For example, R231A and R228E mutations causes the channel to become constitutively active/open [73,75]. Electrophysiology based patch-clamp experiments have shown that arginine R228 and R237 in S4 interact with E160, located towards the extracellular end of S2. R228 forms electrostatic interactions with E160 in the resting state whereas R237 pairs up with the same in the activated state of Q1-VSD, indicating the outward motion of S4 during Q1-VSD activation [74] (Figure 4A). An interesting feature of KCNQ1 is that its activation is both voltage dependent and mediated by a membrane lipid phosphatidylinositol 4,5-bisphosphate (PIP₂). Depletion of PIP₂ from the inner leaflet of the membrane causes suppression of KCNQ1 channel activity [75,76]. Zhang et al. studied the interaction of PIP2 with the open and closed states of KCNQ2 (Kv7.2) suggesting that PIP2 interacts with the S4–S5 linker region in the activated state and plays a role in gating mechanism, whereas in the closed state this region of contact lies in the S3–S3 loop of KCNQ2 [77]. In KCNQ1, basic residues in the S2–S3 and S4–S5 and the C-terminal region were identified as PIP2 binding sites with a higher PIP2 sensitivity in the presence of KCNE1 [77]. Voltage clamp fluorometry showed that PIP₂ is required for the coupling of VSD-PD and in the presence of lipid phosphatase, CiVSP, the voltage sensor remains active/depolarized while the pore is closed [75,76,78]. PIP₂ depletion causes disruption of the VSD-PD coupling and channel adopts an “uncoupled” state or is said to be “decoupled” [53,60]. The role of PIP2 as a coupling component is also supported by MD simulations [79]. In the resting state, PIP2 is located in the vicinity of S4 towards the cytosolic end of VSD. This prohibits the interaction of PIP2 with the amino acid residues on S6 helix of KCNQ1, thereby favoring a closed/resting state of the channel. In the active state, PIP2 forms salt bridges with the residues on S6 helix, shifting it further away from the positively charged residues on S4 favoring the active/open state of the channel [79]. Thus, PIP2 is suggested to favor the resting/closed and activated/open states of KCNQ1 [79]. The detailed structure of KCNQ1 in the decoupled state was recently published by Sun et al [53]. Based on the charge distribution and interaction of negatively charged PIP₂ with the positively lined pockets between the voltage sensor and the pore forming region, two potential PIP₂ binding regions were identified: One is the interface between Q1-VSD and PD (involving cytosolic end of S4 and S4–S5 linker), and the other region involves HB helix of the soluble C-terminal domain [60]. PIP₂ has been shown to stabilize the open states of other Kv channels like hERG and K_ATP [80,81] and is thought to mediate similar effects in the KCNQ1-KCNE1 channel complex. Studies involving KCNQ1, KCNQ2, and hERG channels have shown that PIP2 migration towards the S2–S3 linker region regulates the deactivation kinetics of these channels [80,81]. The underlying mechanism of PIP2 migration in the case of hERG and KCNQ1/2 channels suggest that PIP2 interacts with the multiple binding sites on both these channels leading to acceleration of deactivation process [82].

4. The KCNQ1-KCNE1 Channel Complex

As previously stated, association of KCNQ1 with KCNE1 is crucial for the channel functioning. KCNE1 slows down the activation/deactivation kinetics of KCNQ1 and modulates the channel conductance critical for repolarization phase of cardiac action potential [31,63,83]. KCNE1 is a single transmembrane protein of 129 amino-acid residues with implications in channel biology and disease conditions [60]. Several studies have suggested that KCNE1 interacts with the PD of KCNQ1 [84–86]. Molecular modeling and studies involving the rat Kv1.2 channel structure showed unoccupied gaps between the VSDs, strengthening the claim that KCNE1 sits in the cleft between the adjacent VSDs and the PD [87,88] (Figure 5B). Cysteine crosslinking experiments involving the extracellular regions of KCNE1 and S1, S4 and S6 of KCNQ1 indicated state dependent interactions in these regions [83] (Figure 4B, 5B). Atrial Fibrillation (AF) related mutants in the S1 helix, S140G and V141M modulated the gating properties in the presence of KCNE1 [83]. Residues F57, T58, and L59 in the transmembrane helix of KCNE1 affected the voltage dependent channel modulation by interacting with residues S338, F339, F340 and A341 in the S6 region of KCNQ1 when mutated in either of the interacting partners [84,89,90]. These results suggest the involvement of S6 in KCNQ1-KCNE1 interaction and function. Cysteine accessibility experiments demonstrated a positive shift in the voltage dependence for T224C mutant in the presence of KCNE1 [91]. Mutations F232A in S4 and F279A in S5 helix showed changes in Q1-VSD movement and channel function in the presence of KCNE [92]. Similar experiments involving A226C in the S4 region slowed down the modification rate of methanethiosulfonate (MTS) reagent in the presence of KCNE1 indicating the stabilization of S4 in the closed state of Q1-VSD [93]. However, reports suggesting the association of KCNE1 altering the voltage dependence of the VSD and not the movement of S4 itself led to the conclusion that the observed changes affected the overall VSD equilibrium [94]. In the presence of KCNE1, the charge reversal mutations reported in the previous section showed interesting features. E160R/R231E currents were inhibited while the cell surface expression remained unaltered suggesting, 1) KCNE1 suppressed the intermediate-open state favoring an activated-open conformation, and 2) the above was a functional consequence on the intermediate-open state [56].

As we pointed out earlier, interaction with KCNE1 alters the overall topology of VSD-PD and has the potential to change the landscape of inter-subunit interaction as well. Based on these findings, it is obvious that KNCE1 modulates the overall functionality of KCNQ1 channel in several ways: by modulating VSD-PD coupling to prefer active-open state over the intermediate-open state, altering the ion selectivity and, increasing the current amplitude through enhanced VSD-PD association in the active-open state [56]. Despite the fragmented attempts to provide site specific interactions of KCNQ1 in the presence of KCNE1, voltage sensor movement and overall channel modulation, it is largely unclear as to how this 129 amino acid peptide regulates the Q1-VSD movement and modulates KCNQ1 activation kinetics on such a wide scale.

5. KCNQ1 and non-KCNE1 partners

While, the KCNQ1 activity modulation by KCNE1 has been studied to a great length, several other proteins (both membrane and cytosolic) and lipid molecules can directly affect the overall functionality of KCNQ1 [31,53,60,95]. One of these non-KCNE1 proteins is KCNE3. In an earlier study, it was suggested that in contrast to KCNE1, which affects both the VSD and gating of KCNQ1, KCNE3 only affects the VSD of the channel complex. The overall effect of KCNE3 is mediated through voltage sensor-to-gate coupling [96] (Figure 4C). Although, KCNE1 and KCNE3 display differences between their mode of action towards KCNQ1 (Figure 4B and 4C), their action is primarily controlled by the residues located in their TM segments. KCNQ1 forms a complex with KCNE3 and plays an important role in the recycling of potassium ions essential for transepithelial chloride ion secretion [97,98]. Transmembrane region of KCNE3 interacts with the open state of KCNQ1 and has been suggested to play an important role in the kinetics of initial binding and KCNQ1-KCNE3 complex formation. KCNE3 stabilizes the VSD region of KCNQ1 in its active state in a voltage dependent manner [99]. Electrophysiology based current measurements have shown that KCNE3 increases the KCNQ1-dependent current density by ~ 10-fold and leads to a constitutively active channel [99]. However, it is still not clear how KCNE3 affects the VSD but not gating of KCNQ1. Several studies have tried to explain the interaction mechanism between KCNQ1 and KNCE1 or, KCNE3 with the anticipation of discovering the interaction-function aspects. For example, Wrobel et al. have described the interactions between KCNQ1 and the N-terminal, the TM segment, and the C-terminal regions for both KCNE1 and KCNE3 (Figure 4B and 4C) [100]. Molecular modeling-based approach suggested that the extracellular end of the transmembrane region of KCNE3 sits in the cleft formed by extracellular end of S1 helix in one subunit and S5/P-loop/helix of the adjacent subunit of KCNQ1 [99]. Furthermore, it has also been proposed that either the pore region or the VSD of KCNQ1 channel could be directly interacting with the TM of KCNEs or, the C-terminus of the both of proteins may be interacting [65,66,84,85,93]. Even though we have several clues about the physical interaction regions/sites involved between KCNQ1 and KCNE1 or, KCNE3, it is still not clear how these interactions modulate the functionality of the overall channel complex. Therefore, this aspect of KCNQ1-KCNE3 biology still needs further exploration.

The activity of KCNQ1 is also modulated by polyunsaturated fatty acids (PUFA). Liin et al. showed that PUFA analogs act as I_KS channel activators and play a role in cardiac exitability [101,102]. They proposed that PUFA and PUFA analogs incorporate their acyl chains into the lipid bilayer between the Q1-VSD, close to KCNE1 and bind to KCNQ1 via the lipid head group in the proximity of the positively charged amino acids of Q1-VSD [101–103]. The channel is activated as a result of this electrostatic interaction. The presence of KCNE1 leads to the protonation of PUFA, thereby exerting an opposite effect [103]. The negatively charged amino acid residues in the loop region connecting S5 to the pore helix play an important role in KCNE1 induced protonation [103] leading to the hypothesis that KCNE1 plays an indirect role in promoting PUFA protonation by inducing conformational changes in KCNQ1. These structural rearrangement result in the movement of the S5-p-helix loop closer to the PUFA binding site [103]. The PUFA analog, N-arachidonoyl taurine (N-AT) is shown to interact with the PD of KCNQ1 following electrostatic interaction with K326 in the S6 region [101]. N-AT increases the channel conductance by pulling K326 away from the pore region causing structural perturbations in the selectivity filter region of KCNQ1. PUFA analogs such as N-AT also interact with the positively charged S4 arginines in Q1-VSD and alter the voltage dependence by shifting it towards more negative voltages causing the channel to open [101].

As described earlier in this review, KCNQ1 plays various physiological roles in the heart as well as epithelial tissues. The multifunctional KCNQ1 is enabled by its interaction with several non-KCNEs proteins and cytosolic molecules acting as direct modulators of KCNQ1 functions. One of these modulators is protein kinase A (PKA). The β-adrenergic receptor stimulation plays an important role in KCNQ1-KCNE1 current generation during a high-stress, fast heart rate condition [104]. The β1-adrenergic receptor activation directly affects the intracellular level of cAMP leading to PKA activation, resulting in phosphorylation of the N-terminal region of KCNQ1. This leads to faster channel activation and, ultimately, shortening of the cardiac action potential. One of the proteins which plays important role in the anchoring of PKA to KCNQ1-KCNE1 complex is Yotiao or A-kinase anchoring protein (AKAP). It forms a macromolecular complex consisting of channel proteins, PKA, phosphodiesterase PDE4D3, protein phosphatase PP1, and adenylyl cyclase (AC9) in the heart [105,106,107]. Yotiao facilitates AC9- KCNQ1 complex formation in the heart following β-adrenergic stimulation [105,106,108]. Yotiao physically interacts with the residues in the cytosolic C-terminal domain of KCNQ1 and modulates the channel activity by affecting the phosphorylation/dephosphorylation events at S27 of the N-terminal end of KCNQ1 [106–108]. S27 phosphorylation does not require KCNE1. However, in the absence of KCNE1, this PKA-mediated phosphorylation of S27 does not result in hyperpolarization and channel activation, suggesting that KCNE1 plays a role in transducing the phosphorylation cascade to modulated channel function [109]. As described for membrane receptor proteins where phosphorylation plays an important role in activation/deactivation kinetics and disease manifestations [110,111], further understanding of the mechanism of PKA mediated post-translational modification of KCNQ1–KCNE1 activity will be important for understanding long QT mutations.

The cytosolic Ca²⁺-binding protein and Ca²⁺ signaling mediator, Calmodulin (CaM), plays important role in KCNQ1 channel assembly. CaM assists KCNQ1 in channel assembly to form functional tetramers [112,113]. In fact, when CaM does not interact with KCNQ1 due to mutations in the channel protein, it prevents tetramerization of KCNQ1, which is directly related with aberrant channel function and long QT syndrome. Furthermore, the cytosolic Ca²⁺ concentration affects the KCNQ1 activation. Physiologically reduced intracellular Ca²⁺ causes an inactivation of homomeric KCNQ1 [112]. CaM binds to the C-terminal of KCNQ1 and is essential for channel folding and assembly. Calmodulin binding results in a dimeric coiled-coil structure further forming a dimer of the dimer leading to tetramerization and a functional KCNQ1 [112,113]. The involvement of the C-terminal in KCNQ1 is unique because most of other Kv channels show this tetramerization mediated through their N-terminal region. The cryo-EM structure of the KCNQ1/CaM complex revealed two contact sites, 1) where the HA-HB helices of the soluble C-terminal segment of KCNQ1 insert in the middle of a clam-shell like structure formed by the N and C lobes of CaM, and, 2) between the S2–S3 loop of Q1-VSD (in the transmembrane region) and the third EF hand motif of CaM [53]. This second contact site along with the involvement of the fourth EF motif of CaM in the first contact site was not observed in the crystal structure reported earlier [114]. CaM interaction with the above two regions in KCNQ1 can be a useful link in unraveling the VGD-PD coupling mechanism. Recent studies by Chang et al. highlighted the role of an Apo/Cam clamp conformation of Cam C-lobe calcium dependent switch [115]. Apo/Cam clamp conformation, a common feature among all Kv7 isoforms, promotes the opening of Kv7.2-Kv7.5 channels but inhibits Kv7.1 (KCNQ1) activation. Binding of calcium to the C-lobe activates a CaM C-lobe causing its transition from semi-open to an open conformation, anchoring the N-lobe of CaM and perturbing the interactions with helix A leading to channel inhibition in Kv7.2- Kv7.5 while facilitating activation of Kv7.1 (KCNQ1) [115].

6. Conclusion and Future Directions

KCNQ1, one of the potassium ion channel family members, displays a multifunctional role and is important for various cellular processes/functions, for example cell volume homeostasis, electrical signaling in cardiac myocytes, glucose homeostasis, and gastric acid secretion. Aberrant KCNQ1 channel activity directly leads to disease phenotype and one of the prominent examples is long QT syndrome. KCNQ1 activity essentially requires its interaction with KCNE family member proteins, most prominently KCNE1. KCNQ1/KCNE1 interaction is further modulated by several other cellular proteins/molecules: calmodulin, PIP₂, PKA, pH, and post translational modification (phosphorylation) of KCNQ1 itself. Due to these diverse interactions, KCNQ1 could perform several functions in the cellular physiology. As described in detail in this review, structural and biophysical studies of KCNQ1 have provided us several seminal findings showing the role of VSD and helix S5–S6 roles in the channel opening and gating. Even if we now know a significant amount regarding the roles for specific residues in KCNQ1 function and about its interactions with KCNE proteins, there are still many avenues of KCNQ1 structural-functional biology which needs to be explored in detail in future studies. Undoubtedly, future regulatory mechanistic studies of KCNQ1 function and its regulation will not only add to the knowledge base regarding KCNQ1 channel physiology under healthy and heart disease conditions, but will also illuminate its role in epithelial tissues, endocrinology, ear function and other cellular homeostasis functions.

Highlights:

• KCNQ1 is related to several diseases, atrial fibrillation, long QT syndrome

• KCNQ1 is ubiquitously expressed, complexes with members of KCNE family

• Non-KCNE partners like PUFA, PIP2, CaM, PKA cause structure/functional changes

Abbreviations:

LGIC

ligand-gated ion channel

VGL

voltage-gated like

nAchR

nicotinic acetylcholine receptor

PIP₂

phosphatidylinositol 4,5-bisphosphate

I_KS

slow delayed rectifier current

PKA

protein kinase A

PUFA

polyunsaturated fatty acid

SNHL

sensorineural hearing loss

AF

Atrial Fibrillation