Ion channels play a key role in electrical signalling in the nervous system, muscle contraction and maintenance of ion concentration gradients throughout the body. Voltage gated sodium (Nav) channels allow influx of Sodium ion (Na+) for depolarisation and voltage gated potassium (Kv) channels allow efflux of Potassium ion (K+) for repolarisation. Uniquely, cardiac myocytes require a much longer period of depolarisation and therefore slower repolarisation, allowing adequate diastolic filling time, in order to maintain proper heart function (Osteen, Sampson and Kass, 2010). In the heart, KCNQ1 coassembles with the ? subunit KNCE1 which is the potassium channel responsible for the major cardiac repolarisation (Iks). Channelopatheis in KCNQ1 shows the important role it plays. There are over 100 mutations in KCNQ1 that result in atrial fibrillation, prolonged QT interval, syncopes, and even sudden death (Osteen, Sampson and Kass, 2010).
KCNQ1 is one of 40 Kv channel ? subunits of the human proteome. (Abbott, 2014). It can form a complex with a wider variety of ? subunits than most other Kv channel ? subunits. KCNQ1 can form a K+ channel as a homotetramer like most voltage depenadant ion channels. KCNQ1 can form a complex of four ? subunits and two ? subunits with either KCNE1, KCNE2, KCNE3 or KCNE4. Each KCNE subunit fine tunes the function of the KCNQ1 when associated in a channel. KCNQ1, unlike many other voltage dependant ion channels, is open only when the membrane is depolarised. In the heart, four KCNQ1 proteins coassemble with either two or four of the ? subunit KNCE1 to form the K+ channel which is responsible for the slow delayed rectifier Potassium current. However, there is still some debate as to whether four or two ? subunits are required in each complex.
KCNQ1 is a 676 amino acid protein that forms the ? subunit of voltage gated potassium channels. This is above the median human protein size of 347 amino acids and moderately sized for a voltage gated ion channel protein (Brocchieri, 2005). KCNQ1 has six membrane spanning regions (S1-S6) and both the C and N terminus are intracellular as shown in figure 1.
(Smith et al. 2007)
Figure 1. The primary amino acid sequence for S1-S6 domains of KCNQ1 with mutation sites linked to channelopathies. The channel is in an open state.
KNCQ1 is a voltage gated ion channel and therefore is able to react to a change in voltage. This is achieved by a voltage sensing region which consists of S1 to S4. Each section plays a role in sensing voltage change and ensuring the channel is opened. Although S4 is the S1 and S2 have acidic residues that may protect S4 as it migrates deeper into the membrane during depolarisation. The movement of S1 and S2 is and their proximity to S4 is shown in figure 2.
In the open state S140 and V140 are in close proximity to S4 so likely have van der Waals interactions that stabilise the open state. S4 has a number of basic residues that allows it to act as a voltage sensor. Membrane depolarisation causes S4 to move further into the membrane, this movement is passed onto the pore forming region by the S4-S5 linker. (Bezanilla and Perozo, 2003). This conformational change opens the pore and allows efflux of K+, causing repolarisation. Xu et al. (2008) successfully demonstrated that KCNQ1 I145 (S1) interacted with KCNE1. This indicates that the method by which KCNE1 could possibly affect the characteristics of the channel may be though modifying the S1 interaction with S4.
(Smith et al., 2007)
Figure 2. Locations of KCNQ1 gain-of-function mutation.S140G and V141M are both on S1, I274V is on S5 and A300T is in the pore helix between S5 and S6.
Green – S1Blue – S5 and S6View from membrane plane.
The selectivity of the pore region for K+ is achieved by carbonyl oxygen atoms of glycine residues in the middle of the pore region, between S5 and S6. This motif mimics the hydration shell of aqueous K+. The mimic hydration shell is too small for a hydrated Na+ and too large for a dehydrated Na+. The pore selects for K+: Na+ at a ratio of about 10,000:1 (Richards 2017).
Both intracellular termini have binding sites and interactions that affect the activity of the channel. The C terminus has an assembly domain and an A-kinase anchoring protein binding (AKAP) site. Yotiao, an AKAP, associates with the AKAP site on the C terminus using a leucine zipper. Yotiao then recruits PKA (protein kinase A) and other proteins which, following activation, phosphorylate the PKA phosphorylation site on the N terminus at S27 which increases the activation kinetics and therefore Iks current (Hamilton and Devor, 2016). PKA is cAMP dependant, therefore increasing the concentration of cAMP and PKA can increase these factors. An increase in cAMP is achieved through ?-adrenergic receptor activation via the sympathetic nervous system, commonly called the ‘fight or flight’ response. As KCNQ1 function can be increased by the ‘fight or flight’ response, symptoms of channelopathies may be worsened or triggered by stress, excitement or exercise.
Many ion channels would not function without a necessary cofactor: Phosphatidylinositol 4,5-bisphosphate (PIP2). PIP2 is also essential for the Iks. As PIP2 expression is reduced, KCNQ1 activity reduces and PIP2 depletion prevents the channel opening. (Li et al., 2011). This was discovered by expressing KCNQ1 alone and KCNQ1 with KCNE1 in Xenopus oocytes and using patch clamp and two electrode voltage clamp techniques. The activity of the ion channels was recorded at various PIP2 levels. The KCNQ1 channel requires PIP2 to operate; however, levels of PIP2 in the cell membrane are too low to activate all channels if all are KCNQ1 homotetramers. These homotetramers alone do not produce the Iks required for heart function. The KCNQ1 KCNE1 channel complex, however, is 100 times more sensitive to PIP2 than a KCNQ1 homotetramer. This increased sensitivity enables the characteristic Iks curve at physiological levels of PIP2.
Atrial fibrillation (AF) is an aberrant heart rhythm that causes a fast or irregular heartbeat. A possible cause of AF is the reduction of action potential duration the refractory period in atrial myocytes; this would also shorten the refractory period. (Chen et al., 2003). This is caused by a slowing of Iks to such a degree that cardiac myocytes do not fully return to resting state before another action potential. The effect of this is shown in figure 3 the ECG of a heart with AF have no distinct P or T waves, which are atrial depolarisation and ventricular repolarization respectively. The slowed Iks leaves the atrial myocytes more excitable and results in the uncoordinated triggering of action potentials giving the characteristic ‘flutter’ between S and Q.
Figure 3. Typical ECG trace for a healthy heart (top) and a heart with atrial fibrillations (bottom). Note that there are no distinct P and T waves in the ECG with atrial fibrillations.
S140G is one of four a gain of function mutations in Kv 7.1 that cause AF (Chen et al., 2003; Smith et al., 2007). S140G is a nonsense mutation that causes the replacement of serine with glycine at the 140th amino acid in the protein primary structure (this mutation is labelled in yellow on figure 1). Glycine is a smaller amino acid than serine and has only a hydrogen as a side chain. Serine is a polar amino acid as it has a hydroxymethyl group. Serine, in an protein, is capable of forming an ST turn. This structural motif consists of a hydrogen bond between the OH group of the serine and the NH group of the main chain two amino acids away. This structure, if present, would be disrupted by the S140G mutation.
S140G mutation present in KCNQ1 results in a slower rate of activation and lack of inactivation at physiological conditions (Barhanin et al., 1996). Xenopus oocytes have been used to model KCNQ1 mutations. In these models, S140G KCNQ1 exhibited a peak rate of depolarisation greater than four times that of wild type (WT) KCNQ1 under an action potential voltage clamp (Restier, Cheng and Sanguinetti, 2008). Other S140 mutants were expressed in the oocytes and, for most mutations, the result was a complete loss of interaction between KCNQ1 and KCNE1 and loss of function of KCNQ1. (Restier, Cheng and Sanguinetti, 2008). This indicates that S140 plays some role in the interaction between the ? and ? subunits and also a role in the KCNQ1 homotetramer. This seems to be supported by the work of Xu et al., (2008) who proved that KCNE1 interacts with KCNQ1 only 4 amino acids from S140. The proximity of S140 to KCNE1 in the quaternary structure is shown in figure 4.
Smith et al. (2007).
Figure 4. A KCNQ1, KCNE1 heterotetramer channel from the extracellular view in an open state(A) and one KCNQ1 subunit (B) both have amino acids that interact with KCNE1 labelled and van der Waals shells shown.
KCNE1 has been proven to interact with KCNQ1 N terminus and some functional interaction are cause by either the N terminus or membrane spanning region (MSR) of KCNE1. this is supported by Restier, Cheng and Sanguinetti’s work in which loss of interaction between ? and ? subunits was seen in some S140 mutants.
KCNQ1 can form a number of different specific K+ currents. This is made possible by its ability to interact with a number of other subunits. KCNQ1 is a very adaptable voltage gated ion channel and understanding its channelopathies is imperative for understanding a broad range of diseases and channelopathies ranging from signalling in the nervous system, cardiac myocytes and epithelial cell function.