SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix

KCNQ1 (Kv7.1 or KvLQT1) encodes the alpha-subunit of a voltage-gated potassium channel found in tissues including heart, brain, epithelia and smooth muscle. Tissue-specific characteristics of KCNQ1 current are diverse, due to modification by ancillary subunits. In heart, KCNQ1 associates with KCNE1 (MinK), producing a slowly activating voltage-dependent channel. In epithelia, KCNQ1 co-assembles with KCNE3 (Mirp2) producing a constitutively open channel. Chromanol 293B is a selective KCNQ1 blocker. We studied drug binding and frequency dependence of 293B on KCNQ1 and ancillary subunits expressed in Xenopus oocytes. Ancillary subunits altered 293B potency up to 100-fold (IC50 for KCNQ1 = 65.4 ± 1.7 μm; KCNQ1/KCNE1 = 15.1 ± 3.3 μm; KCNQ1/KCNE3 = 0.54 ± 0.18 μm). Block of KCNQ1 and KCNQ1/KCNE3 was time independent, but 293B altered KCNQ1/KCNE1 activation. We therefore studied frequency-dependent block of KCNQ1/KCNE1. Repetitive rapid stimulation increased KCNQ1/KCNE1 current biphasically, and 293B abolished the slow component. KCNQ1/KCNE3[V72T] activates slowly with a KCNQ1/KCNE1-like phenotype, but retains the high affinity binding of KCNQ1/KCNE3, demonstrating that subunit-mediated changes in gating can be dissociated from subunit-mediated changes in affinity. This study demonstrates the KCNQ1 pharmacology is significantly altered by ancillary subunits. The response of KCNQ1 to specific blockers will therefore be critically dependent on the electrical stimulation pattern of the target organ. Furthermore, the dissociation between gating and overall affinity suggests that mutations in ancillary subunits can potentially strongly alter drug sensitivity without obvious functional changes in gating behaviour, giving rise to unexpected side-effects such as a predisposition to acquired long QT syndrome.

KCNQ1 was first identified in the heart, where it was called KvLQT1, reflecting the association between mutations in this gene and the inherited increased risk for cardiac arrhythmias and sudden death (Wang et al. 1996). Mutations in KCNQ1 result in abnormal cardiac repolarization, leading to prolongation of the action potential, i.e. long QT syndrome, which is associated with an increase in morbidity and mortality (Wang et al. 1996). Subsequent to its identification in the heart, KCNQ1 (Kv7.1 or KvLQT1) has been identified in a wide range of diverse tissue types, including the heart, neuronal tissue, kidney, colonic crypt cells, pituitary, the stria vascularis and the vestibular dark cells of the cochlea, stomach, small intestine, liver, thymus, exocrine pancreas, prostate, skeletal muscle, airway epithelia, ovarian tissue, testis, uterus, and placenta (Brahmajothi et al. 1996; Wang et al. 1996; Vetter et al. 1996; Yang et al. 1997; Chouabe et al. 1997; Neyroud et al. 1997; Gould & Pfeifer, 1998; Jentsch, 2000; Wulfsen et al. 2000; Schroeder et al. 2000; Demolombe et al. 2001; Mason et al. 2002; Tsevi et al. 2005; Lan et al. 2005). The KCNQ1 channel is therefore a key component in a variety of tissues, and plays a variety of roles within these tissues.

The electrical excitability of these tissues varies widely, as do the characteristics of the ionic currents associated with the KCNQ1 subunit identified in the native cells. Some tissues that contain KCNQ1 exhibit high electrical excitability, but some other KCNQ1-containing tissues, such as epithelia, exhibit little electrical excitability. Although tetramers of the alpha-subunit of the KCNQ1 ion channel can independently form a rapidly activating voltage-dependent potassium channel, when KCNQ1 is co-expressed with ancillary subunits, the behaviour of the channel is dramatically transformed, and can even become voltage insensitive (Sanguinetti et al. 1996; Melman et al. 2001).

Although the details of the interactions between the KCNQ1 channel and KCNE subunits are not fully understood, KCNE subunits clearly have an effect on the mechanism of gating. KCNE3 (MiRP2) appears to ‘lock’ the channel in an open position, resulting in a constitutively open channel (Melman et al. 2001). KCNE1 (MinK) appears to prevent the KCNQ1 from entering the inactivated state which is seen at positive potentials with pure KCNQ1 channels (Seebohm et al. 2005). In both cases, the presence of the subunit has greatly reduced or eliminated a particular gating transition.

The ability of KCNE subunits to alter gating behaviour and alter the allowable physical conformations of the channel can result in changes in the drug binding ability of the channels. Drug binding is strongly dependent upon channel conformational states, and frequently on the open state (Wang et al. 2003; Seebohm et al. 2003a; Bett & Rasmusson, 2004). In the same way that mutations that affect gating can alter drug binding (Wang et al. 1997; Bett & Rasmusson, 2004), ancillary subunits that alter the gating characteristics and conformational states of a channel can also alter drug–channel interactions (Bett et al. 2006). In this study, we used site directed mutagenesis to investigate the correspondence between subunit-mediated changes in the gating behaviour of KCNQ1 and changes in drug potency.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix

Mature female Xenopus laevis (Xenopus Express, FL, USA) were cared for by standards approved by the Institutional Animal Care and Use Committee of the University at Buffalo, SUNY. Frogs were anaesthetized by immersion in 1 g l−1 tricaine solution (Sigma). Oocytes were removed by partial ovariectomy and digested by placing them in a collagenase-containing Ca2+-free OR2 solution (mm): 82.5 NaCl, 2 KCl, 1 MgCl2, 5 Hepes; pH 7.4, 1 mg ml−1 collagenase, type I, Sigma). Frogs were humanely killed following final collection of oocytes. The oocytes were gently shaken for 1.5–2 h, with the enzyme solution refreshed at 1 h. Defolliculated oocytes (stage V–VI) were injected with up to 50 ng mRNA for KCNQ1 using the Nanoject microinjection system (Drummond Scientific Co., PA, USA). Oocytes were co-injected with KCNE1 or KCNE3, where noted, in a 1 : 1 ratio. KCNE1 cDNA was isolated by reverse transcription from human whole heart RNA (Clonetech, Santa Rosa, CA, USA), and cloned into PGEM-HE5 for expression in Xenopus oocytes (GenBank Accession No. NP_000210). Human KCNQ1 cDNA (GenBank Accession No. P51787) was a kind gift from Drs Mark Keating and Michael Sanguinetti. KCNE3 (NP_005463) was a kind gift from Dr Thomas McDonald. The drug 293B (C15H20N2O4S, IUPAC name: N-(6-cyano-3-hydroxy-2,2-dimethyl-chroman-4-yl)-N-methyl-ethanesulphonamide)) was a kind gift from Dr Rainer Greger. The V72T point mutation to KCNE3 was made using the Stratagene Quickchange site-directed mutagenesis kit (Stratagene, TX, USA). A schematic diagram of the KCNQ1 channel, the KCNE ancillary subunits, and the position of the V72T mutation is shown in Fig. 1.

image

Figure 1. Topological cartoon of the putative structure of KCNQ1 channels and the KCNE subunit KCNQ1 channels are thought to have six transmembrane segments, with a charged S4 segment responsible for voltage sensing. The region between S5 and S6 forms the H5 loop which plays a role in sensitivity and gating. The KCNE subunits are proteins with a single transmembrane-spanning domain and a cytosolic C-terminal. The details of the interaction between the KCNE ancillary subunits and the KCNQ1 channel are suggested to involve multiple regions of the channel and subunit. The amino acids in the aligned putative transmembrane segments of KCNE1 and KCNE3 are shown. The valine and positionally analogous threonine relevant to the KCNE3[V72T] mutation are indicated in bold.

Download figure to PowerPoint

Oocytes were voltage-clamped in a whole-cell mode configuration using a two-microelectrode oocyte clamp amplifier (CA-1B, Dagan Corp., Minneapolis, MN, USA), and currents were recorded at room temperature. Microelectrodes with resistances of 0.5–1.5 MΩ (when filled with 3 m KCl) were fabricated from 1.5 mm o.d. borosilicate glass tubing (TW150-4, World Precision Instruments) using a two-stage puller (Kopf Instruments, CA, USA) and filled with 3 m KCl. The control extracellular solution (2 mm K+) contained (mm): 96 NaCl, 2 KCl, 1 MgCl2, 1.8 CaCl2, 10 Hepes, pH 7.4. Voltage clamp protocols used are described as appropriate in the text, and unless otherwise stated, raw two-electrode voltage clamp data traces were not leak or capacitance subtracted.

Data were digitized and analysed using pCLAMP 6.0–9.2 (Axon Instruments). Further analysis was performed using Clampfit 9.2 (Axon Instruments), Excel (Microsoft Corp.) and Origin (Microcal Software Inc., MA, USA). Data were filtered at 2 kHz. Data are shown as mean ±s.e.m. Confidence levels were calculated using Student's t test and ANOVA.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix

Channels formed from the KCNQ1 (Kv7.1 or KvLQT1) alpha-subunit are voltage-gated K+ channels with a putative structure of six transmembrane-spanning segments, including the charged voltage-sensing S4 segment. Figure 1 shows a schematic representation of the KCNQ1 channel and the KCNE subunits, and the position of the KCNE1 valine to threonine point mutation used in this study.

The electrophysiological characteristics of the KCNQ1 alpha-subunit are dramatically altered by the presence of KCNE ancillary subunits. Figure 2 shows representative traces obtained from two electrode voltage clamp experiments on KCNQ1 channels, KCNQ1/KCNE1 channels, and KCNQ1/KCNE3 channels. A 2 s depolarizing pulse was applied from the holding potential of −80 mV to a range of voltages between −90 and +60 mV in 10 mV steps. This was followed by a 1.5 s pulse to −60 mV. The oocyte was then returned to the holding potential of −80 mV, and allowed to rest for 20 s before the next depolarizing pulse was applied. Figure 2 shows that the presence of ancillary subunits substantially altered the kinetics of the KCNQ1 current, as has been previously noted (Schroeder et al. 2000). Figure 2A shows that KCNQ1 channels, when expressed alone, exhibit rapid activation. At positive potentials, there was slight inactivation of the current. The presence of mild inactivation can be detected in the P2 pulse to −60 mV, where the tail current clearly increases before deactivation occurs, i.e. recovery from inactivation is readily observable. Co-expression of KCNQ1 channels with the ancillary subunit KCNE1 (KvLQT1/MinK) is thought to be the molecular basis of the slowly activating delayed rectifier current found in the heart, IKs (Sanguinetti et al. 1996). Figure 2B shows that the presence of the KCNE1 subunit alters the kinetics of gating. KCNQ1/KCNE1 channels have very slow activation, with a sigmoidal onset. The channel continues to activate slowly, and activation is not complete, even at the end of a 3 s pulse. The very slow increase in current may reflect the channel entering a second open state (Pusch et al. 2001). KCNQ1/KCNE1 channels do not display inactivation, even at very positive potentials. On repolarization to −60 mV, KCNQ1/KCNE1 channels deactivate.

image

Figure 2. Representative traces from two-electrode voltage-clamp recordings from heterologously expressed currents in Xenopus oocytes A standard two-pulse protocol was used. The first pulse (2 s) was from the holding potential of −80 mV to a voltage between −90 and +50 mV in 10 mV steps. The second pulse (1.5 s) was to −60 mV. Traces are from: A, KCNQ1 channels; B, KCNQ1 channels co-expressed with KCNE1; C. KCNQ1 channels co-expressed with KCNE3.

Download figure to PowerPoint

KCNQ1 channels are co-expressed with the KCNE3 subunit in epithelial and other cell types. Figure 2C shows that the presence of the KCNE3 subunit affects the basic voltage-dependent gating of the channel, rendering KCNQ1/KCNE3 channels constitutively open. When the transmembrane voltage is changed, there is an instantaneous change in the current due to the change in driving force, but no change in the number of open channels. On repolarization to −60 mV, there is no deactivation of the channel.

We determined the effect of the chromanol 293B on KCNQ1 channels, and how this interaction was modulated by the presence of ancillary subunits. 293B is an open pore blocker which is selective for KCNQ1 channels (Bleich et al. 1997). Figure 3 shows representative traces from two-electrode voltage-clamp experiments showing the effect of 293B on the channels. A 3 s depolarizing pulse was applied from the holding potential of −90 mV to a range of voltages between −90 and +40 mV. This was followed by a 1.5 s pulse to −60 mV. The oocyte was then returned to the holding potential of −90 mV, and allowed to rest for 30 s before the next depolarizing pulse was applied. Figure 3 shows representative traces in control and the presence of drug. KCNQ1 was the least responsive to chromanol 293B, with 10 μm only slightly reducing the magnitude of the current, and having no effect on the time course of channel gating. Figure 3A therefore shows the effect of 50 μm 293B on KCNQ1. KCNQ1/KCNE1 channels were more sensitive to chromanol. Application of 10 μm 293B not only reduced the current, but also altered the apparent time course of channel activation, but did not affect the threshold for voltage-dependent activation. This could be due to direct modification of gating, such as the inhibition of the slowly developing component of block. In theory, it could also occur by a coincidence of the development of a slow open channel block that exactly matches and cancels the slow component of activation. As discussed below, we favour the former explanation. The slowly developing component of activation was apparently reduced by 10 μm 293B, and the apparent time course of activation appeared to have reached a steady state at the end of a 3 s pulse. KCNQ1/KCNE3 currents were most sensitive to 293B, which substantially reduced the current in a time-independent manner. Data from the effect of 293B on multiple oocytes are summarized in Fig. 4.

image

Figure 3. Effect of chromanol 293B on currents Representative traces showing currents elicited by a 3 s depolarizing pulse from the holding potential of −90 mV to potentials between −80 to +40 mV, followed by a 500 ms pulse to −60 mV, then repolarization to the holding potential. A, KCNQ1. B, KCNQ1 (same oocyte as A) with 50 μm 293B. C, KCNQ1/KCNE1. D, KCNQ1/KCNE1 (same oocyte as C) with 50 μm 293B. E, KCNQ1/KCNE1. F, KCNQ1/KCNE1 (same oocyte as E) with 10 μm 293B. G, KCNQ1/KCNE3. H, KCNQ1/KCNE3 (same oocyte as G) with 1 μm 293B.

Download figure to PowerPoint

image

Figure 4. Dose–response curve for the effect of 293B on KCNQ1 channels with ancillary subunits Peak current was measured at the end of a 3 s pulse to +40 mV. Data are fitted by a Hill equation: I= 1 − ([293B]/([293B]+ IC50)). The calculated IC50 values are: KCNQ1, 67.0 ± 1.6 μm (▪); KCNQ1/KCNE1, 16.1 ± 1.8 μm (•); KCNQ1/KCNE3, 0.72 ± 0.02 μm (▴). For each point, n= 3 − 8.

Download figure to PowerPoint

The degree of block of peak current (i.e. as measured at the end of the 3 s pulse) by 293B was strongly dependent on the presence of subunits. Application of 50 μm 293B reduced peak KCNQ1 current to 56.9 ± 5.4% of control (n= 8). KCNQ1/KCNE1 channels exhibited a greater sensitivity to 293B, with 50 μm reducing peak current to only 20.7 ± 2.0% of control (n= 4). Similarly, application of 10 μm 293B only reduced peak KCNQ1 current to 87.9 ± 4.0% of control (n= 4), but reduced peak KCNQ1/KCNE1 current to 64.5 ± 1.2% of control (n= 5). By contrast, it took only 1 μm 293B to reduce peak KCNQ1/KCNE3 current to 41.4 ± 6.1% of control (n= 5). Figure 4 shows a dose–response relationship for the current elicited at the end of a 3 s depolarizing step to +40 mV for KCNQ1, KCNQ1/KCNE1 and KCNQ1/KCNE3 channels in the presence of chromanol 293B. The presence of either subunit shifted the dose–response curve to the left, and increased the apparent affinity of the channel for the drug. The presence of the KCNE subunits significantly altered the IC50 value for 293B, as determined by fitting the data with a Hill equation. The IC50 for KCNQ1 alone (67.0 ± 1.6 μm) was significantly larger than the IC50 for KCNQ1/KCNE1 (16.1 ± 1.8 μm), which was in turn was significantly larger than the IC50 for KCNQ1/KCNE3 (0.72 ± 0.02 μm). The presence of ancillary subunits can therefore result in up to ∼100-fold change in potency for the drug.

Chromanol 293B is an open channel blocker, and the opening of KCNQ1 channels is time dependent. We therefore examined the time dependence of drug binding to KCNQ1 channels, and whether it was affected by the presence of ancillary subunits. Time-dependent changes in the fraction of current block during a step depolarization from −90 to +40 mV were calculated by determining the ratio of the current elicited during a 3 s pulse in the presence of drug to a 3 s control pulse, i.e. Iratio= 1 −Idrug/Icontrol, throughout the pulse duration. Figure 5A shows average current ratios recorded during a 3 s depolarization from −90 to +40 mV. When KCNQ1 channels were depolarized in the presence of 50 μm 293B, there was a rapid onset of 293B block, which is consistent with 293B being an open channel blocker. After the initial onset, the degree of block changed little over the 3 s depolarization. After the rapid onset of block as the channel opens, the initial KCNQ1 channel block was 40.0 ± 6.2% of control. This was similar to the final percentage of block, which was 43.5 ± 5.3% of control (n= 8). When KCNQ1/KCNE3 channels were depolarized in the presence of 1 μm 293B the block was apparent from the very beginning of the pulse, i.e. there was no rapid initial onset of block (Fig. 5C). This is consistent with the notion that KCNQ1/KCNE3 channels are constitutively open. The degree of block at the beginning of the depolarizing pulse was 61.6 ± 5.8% which is not significantly different from the block of 59.0 ± 5.2% of control observed at the end of the 3 s pulse (n= 5).

image

Figure 5. Time dependence of 293B block The fraction of current block during a 3 s depolarization from −90 to +40 mV are shown. Fraction of block is calculated as 1 −Idrug/Icontrol. Continuous lines are averaged traces (n= 4 or 5) and grey lines represent s.e.m. A, KCNQ1 with 50 μm 293B. B, KCNQ1/KCNE1 with 50 μm 293B. C, KCNQ1/KCNE3 with 1 μm 293B. D, E and F are the same data as shown in A, B and C, but with an expanded timescale and Y axis to enable changes in the first few hundred milliseconds to be seen clearly. G, bar charts to show the difference in percentage block of current at the beginning (filled bar) and end (open bar) of the 3 s pulse for KCNQ1, KCNQ1/KCNE1 and KCNQ1/KCNE3 currents. Only the KCNQ1/KCNE1 current was significantly different at the beginning of the pulse compared with the end of the pulse.

Download figure to PowerPoint

KCNQ1/KCNE1 channels have a slow time-dependent component to activation, which continues throughout a 3 s depolarization. When KCNQ1/KCNE1 channels are depolarized in the presence of 50 μm 293B there was rapid onset of block to 39.5 ± 7.7% of control (n= 4). For the first 100 ms of the depolarization there was mild relief of block. This was followed by a slow increase in block over the remainder of the depolarizing pulse. At the end of the 3 s pulse the degree of block was 79.3 ± 5.4% of control (n= 4) (Fig. 5B). The degree of block therefore changed by a factor of 2 over a 3 s period (P < 0.01).

The significant time dependence of chromanol 293B block on KCNQ1/KCNE1 channels raises interesting questions about how the interaction of the drug with KCNQ1 channels varies with stimulation frequency (Seebohm et al. 2001a). We therefore subjected KCNQ1 and KCNQ1/KCNE1 channels to different pacing protocols after a period of 30 s without stimulation. Figure 6 shows results from using a 500 ms depolarizing pulse applied from −90 to +50 mV with either a 5 s inter-pulse interval or a 500 ms inter-pulse interval. KCNQ1 channels were tested in the presence of 50 μm 293B and KCNQ1/KCNE1 channels were tested in the presence of 10 μm 293B as these two concentrations of drug result in a similar fraction of current block for the respective currents. Figure 6A shows that with a 5 s inter-pulse interval there was little change in the peak outward current elicited by repetitive depolarization, measured at the end of the 500 ms pulse, in the presence of 293B for either KCNQ1 or KCNQ1/KCNE1 channels. In addition, when a 3 s pulse was applied from −90 mV to +50 mV with a 3 s inter-pulse interval there was little change in the ratio of peak outward current with and without drug (data not shown). However, when the inter-pulse interval was reduced to 500 ms, there was a clear difference in the use dependence of 293B on the different channels (Fig. 6B). KCNQ1 channels showed no use dependence in the presence of 50 μm chromanol 293B with a 500 ms inter-pulse interval. Conversely, there was a clear and rapid use dependence to block of KCNQ1/KCNE1 channels by 10 μm 293B when stimulated at 1 Hz. The maximum current at the end of the first 500 ms depolarizing pulse in the presence of 10 μm 293B was 68.7 ± 5.0% of control (n= 4). However, the maximum current at the end of the 20th depolarization applied with a frequency of 1 Hz was only 43.0 ± 1.7% of control (n= 4). This shows that 293B interacts with KCNQ1/KCNE1 in a use-dependent manner.

image

Figure 6. Stimulation frequency dependence of chromanol 293B A, a 500 ms depolarizing pulse from −90 mV to +50 mV was applied with a 5 s inter-stimulus interval to KCNQ1 channels in the presence of 50 μm 293B and KCNQ1 co-expressed with KCNE1 in the presence of 10 μm 293B. B, a 500 ms depolarizing pulse from −90 mV to 50 mV was applied with a 500 ms inter-stimulus interval to KCNQ1 channels with 10 μm 293B and KCNQ1/KCNE1 with 50 μm 293B. Fractional current is Idrug(t= x)/Icontrol(t= x).

Download figure to PowerPoint

Rapid pacing of KCNQ1/KCNE1 channels, in the absence of drug, results in a slowly developing increase in the peak magnitude of the outward current. Figure 7 shows that the maximum current elicited by a 500 ms depolarizing pulse from −90 to +50 mV applied with a 500 ms inter-pulse interval, i.e. stimulation with a frequency of 1 Hz, increases with respect to the current elicited by the first depolarization after 30 s rest. The magnitude of the normalized current increased from 1 to 2.8 ± 0.1 from the 1st to the 30th pulse (n= 4). This means the magnitude of the current is nearly tripled in 30 s of stimulation at 1 Hz. The increase in current magnitude is well described by two exponentials, with τfast= 1.07 ± 0.04 s and τslow= 7.66 ± 0.43 s (n= 4). The fast time constant is dominant, with a ratio of the amplitudes Afast/Aslow= 4.78 ± 0.16 (n= 4). In the presence of 10 μm 293B, the maximum KCNQ1/KCNE1 current is still increased by repetitive stimulation, but the degree of potentiation is reduced. The time course of the potentiation is also altered, and the increase in current is no longer a bi-exponential process, but instead is well fitted with a single exponential: τ293B= 0.87 ± 0.16 s (n= 4). This monoexponential increase is not significantly different from the fast component of potentiation measured in the absence of 293B. The use dependence of 293B block of KCNQ1/KCNE1 channels seen in Fig. 6 therefore appears to result from the elimination of the slow component of potentiation which is usually seen on repetitive stimulation of the channels. Figure 7C shows representative current traces from the 1st and 20th depolarizations of KCNQ1/KCNE1 subjected to the 500 ms depolarization from −90 to +50 mV with a 500 ms inter-pulse interval, and the 1st and 20th sweeps of KCNQ1/KCNE1 with 10 μm 293B. The current elicited by the first sweep shows the typical slow activation of KCNQ1/KCNE1, with the current slowly increasing during the 500 ms depolarization. There is a clear time dependence to the development of block by 293B.

image

Figure 7. Potentiation of KCNQ1/KCNE1 current with repetitive stimulation A, peak current elicited by a depolarizing step from −90 mV to +50 mV, normalized to the first pulse following 30 s rest is plotted against time. B, repetitive rapid stimulation of KCNQ1/KCNE1 increases the magnitude of the current in a bi-exponential manner (τfast= 1.07 ± 0.04, τslow= 7.66 ± 0.43 s). In the presence of 293B the potentiation is a mono-exponential process (τ293B= 0.87 ± 0.16 s) which is not significantly different from the fast component of the current potentiation in the absence of drug (n= 4, P > 0.2). C, raw traces showing the current from the 1st and 20th sweeps of 1 Hz stimulation. D, ratio of KCNQ1/KCNE1 current with and without 10 μm 293B for the 20th sweep. Also shown are data obtained from similar experiments in the presence of 50 μm 293B. There is relatively little time dependence of drug block during the 20th sweep with either 10 or 50 μm 293B.

Download figure to PowerPoint

The current trace from the 20th sweep shows that the channel is not fully deactivated during the inter-pulse interval. When the cell is depolarized to +50 mV there is immediately a large current flow as the channel is already partially activated. The current continues to increase during the 500 ms depolarization, but with a very different time course to the 1st sweep. Although there is a large change in the magnitude of the currents produced by the 20th sweeps with and without 293B, there is little difference in the time-dependent change in current during the depolarization. Figure 7D shows an averaged ratio of current, Iratio= 1 −Idrug/Icontrol, i.e. the fraction of current block, during the 500 ms depolarization for the 20th sweeps with and without 10 μm 293B. The fraction of block during the 20th sweep is essentially time independent. In other words, the time dependence of drug block seen during the 1st depolarization (Fig. 7C) is essentially eliminated by the 20th depolarization.

A triplet of amino acids on the KCNE subunit play a critical role in the resulting behaviour of the channel formed on co-expression with KCNQ1 (Melman et al. 2002). In particular, the mutation of the valine at position 72 of KCNE3 introduces voltage dependence to the KCNQ1/KCNE3[V72T] channel (Melman et al. 2002). We studied how this mutation altered the modulatory effects of ancillary subunits on drug binding and frequency dependence to examine if activation gating was tightly linked to apparent drug affinity.

Figure 8 shows the current elicited when a standard two-pulse depolarization protocol is applied to KCNQ1/KCNE3[V72T]. Activation is clearly voltage dependent, and on repolarization to −60 mV deactivation is rapid. The presence of 10 μm 293B is sufficient to abolish most of the current. Addition of 1 μm 293B reduced current to 37.4 ± 5.8% (n= 6), which is not significantly different from the block of KCNQ1/KCNE3 channels (41.4 ± 6.1%, n= 5). Figure 8D shows an average of the degree of block of KCNQ1/KCNE3[V72T] current during a step depolarization from −90 to +40 mV. There is a large increase in block during the pulse, as it rises from 33.7 ± 6.7% to 64.5 ± 5.3% (n= 6). Clearly, apparent affinity for 293B is unchanged, despite the introduction of time-dependent activation gating.

image

Figure 8. KCNQ1/KCNE3[V72T] A, representative current traces recorded from KCNQ1/KCNE3[V72T] channels using a two-pulse protocol. The first pulse was 3 s depolarization to potentials between −80 and +40 mV from the holding potential of −90 mV. This was followed by a 500 ms step to −60 mV before returning to the holding potential. B, representative trace showing the effect of 1 μm 293B (same oocyte and scale as A). C, the effect of 1 μm 293B on KCNQ1/KCNE1 current is not significantly different from the effect of 1 μm 293B on KCNQ1/KCNE3[V72T] (n= 6, P > 0.6). D, during a 3 s depolarization from −90 mV to +40 mV in the presence of 1 μm 293B, there is a gradual increase in the degree of block KCNQ1/KCNE3[V72T] current. Black trace is average fraction of block from 6 experiments, grey lines indicate s.e.m.

Download figure to PowerPoint

We studied the use dependence of 293B block on KCNQ1/KCNE3[V72T]. We applied a 500 ms pulse from −90 to +50 mV at a frequency of 1 Hz, in the presence and absence of 1 μm 293B. Figure 9A shows that there is no use dependence to the block of KCNQ1/KCNE3[V72T] by 293B. Similar to KCNQ1/KCNE1 channels, rapid pacing of KCNQ1/KCNE3[V72T] results in an increase in the magnitude of the peak outward current (Fig. 9B). However, unlike KCNQ1/KCNE1 channels (cf. Fig. 7A) this potentiation is not affected by the presence of 293B. The magnitude of the peak outward current increase is well fitted by a single exponential (Fig. 9C). In addition to being insensitive to the presence of drug, the time constant of current potentiation is indistinguishable from the fast component of the KCNQ1/KCNE1 channel, or the single component of KCNQ1/KCNE1 channel potentiation in the presence of 293B block (ANOVA, P > 0.6). Figure 9D is signal-averaged data showing the development of block in a 3 s pulse from −90 to +40 mV for both KCNQ1/KCNE3[V72T] and KCNQ1/KCNE1. For each channel, the fraction of block was normalized to the peak outward current at the end of the 3 s pulse, in order to enable easier comparison of the time course of block development. Clearly, development of block in KCNQ1/KCNE3[V72T] and KCNQ1/KCNE1 channels follow similar time courses, after the first 500 ms.

image

Figure 9. Rate dependence of KCNQ1/KCNE3[V72T] A, a 500 ms depolarizing pulse from −90 to +50 mV was applied every 500 ms in the presence and absence of 1 μm 293B to test for rate dependence of block. The degree of block did not significantly change during the pulse of 80 rapid depolarizations. B, despite the lack of rate-dependent block, the current is still potentiated by rapid stimulation. C, the increase in current is well fitted with a single exponential curve, regardless of the presence of 1 μm 293B. The two rate constants, τControl= 1.11 ± 0.18 s and τ293B= 1.45 ± 0.72 s, were not significantly different (P > 0.5, n= 3). D, signal averaged ratios showing the development of block during a 3 s depolarizing pulse from −90 to +40 mV when KCNQ1/KCNE1 is exposed to 50 μm 293B and KCNQ1/KCNE3[V72T] is exposed to 1 μm 293B (values chosen to give the similar fraction of block in each channel). The fraction of block is normalized to the total block at the end of a 3 s pulse. The time course of development of the late (> 500 ms) block is similar for KCNQ1/KCNE1 and KCNQ1/KCNE3[V72T].

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix

Our data show that the presence of ancillary beta-subunits has a dramatic effect on the gating and pharmacology of KCNQ1 channels. Channels formed from only the core alpha-subunit of KCNQ1 exhibit rapid activation following a depolarizing pulse. At positive potentials there is some inactivation which is particularly noticeable in the tail currents. KCNQ1 tail currents obtained on stepping back to more hyperpolarized potentials from a depolarizing pulse from KCNQ1 show a small increase in current before deactivation proceeds (Tristani-Firouzi & Sanguinetti, 1998). Both the KCNE1 and KCNE3 subunits inhibit this inactivation transition, which is thought to operate via a mechanism similar to classic C-type inactivation (Seebohm et al. 2001b). In a variety of other channels, inhibition of C-type inactivation results in a general decrease in drug binding affinity (Baukrowitz & Yellen, 1995; Wang et al. 2003; Bett & Rasmusson, 2004). In contrast, disruption of inactivation in KCNQ1 channels by site-directed mutagenesis does not appear to alter drug binding (Seebohm et al. 2003a). Similarly, our results with chromanol 293B show increased, rather than decreased, potency when inactivation is inhibited by ancillary subunits.

The two subunits appear to have oppositely directed effects on the activation process. KCNQ1/KCNE1 activation is delayed, and channel opening is slow. In contrast, KCNQ1/KCNE3 co-expression results in a constitutively open channel (see Fig. 10). However, both KCNQ1/KCNE1 and KCNQ1/KCNE3 have higher apparent affinity for 293B than KCNQ1 alone, which suggests that modulation of activation/deactivation is not the sole determinant of modulation of 293B potency by ancillary subunits.

image

Figure 10. Models of KCNQ1 kinetic behaviour A, Seebohm et al. (2003b) proposed this kinetic scheme to describe the complex biphasic gating behaviour in KCNQ1 channels in the absence of ancillary subunits. B, KCNE1 slows activation and increases the sigmoidal delay, suggesting the presence of multiple closed states. With KCNE1 there is no inactivation. Our results suggest that the two open states may have higher affinity than the KCNQ1 open states for chromanol, which gives rise to the time-dependent aspects of block. C, KCNE3 appears to lock the channel in an open state, inhibiting transitions to either the closed or inactivated states. This open state has a very high affinity for chromanol. However, the high affinity site appears unrelated to the portion of the molecule which confers changes of activation gating.

Download figure to PowerPoint

The presence of KCNE1 and KCNE3 subunits can either result in the stabilization of an existing open state, or formation of a completely new open state of KCNQ1. The KCNQ1 specific blocker, 293B, can block KCNQ1, KCNQ1/KCNE1 and KCNQ1/KCNE3 channels, which indicates that the binding site is at least partially preserved in all three channel–subunit combinations. However, the potency, time and stimulation frequency dependence of the drug for the channel is markedly altered in the three different channel–subunit complexes. The subunits must therefore, in some way, alter the binding site. There are several ways in which the presence of subunits could alter the apparent drug affinity as measured by the ability of drug to reduce current magnitude. The subunits could (a) alter the accessibility of the site for drug binding (e.g. by occluding the open pore), (b) modulate the drug binding site itself (e.g. by changing the residues lining the pore), or (c) act via some combination of these two effects.

KCNQ1/KCNE3 channels are constitutively open, which suggests that the KCNE3 subunit locks the channel in an open position. No time- or voltage-dependent component of chromanol binding was observed in KCNQ1/KCNE3 channels, which is consistent with the lack of time- and voltage-dependent gating behaviour of the KCNQ1/KCNE3 combination. Our data suggest that chromanol has the highest apparent affinity for the constitutively open KCNQ1/KCNE3 channel. This suggests that chromanol may act as an open channel blocker, or at least have a preference for the open state of the channel. The presence of an open channel blocking mechanism is further reinforced by the time-dependent development of block observed for the KCNQ1/KCNE1. Although the time course of development of KCNQ1/KCNE1 block occurs relatively slowly following channel opening, it has some other properties which suggest a more complicated mechanism. The activation time course of KCNQ1 alone has a fast early time course of activation, and reaches a stable steady state. In contrast, the activation time course of the KCNQ1/KCNE1 channel is relatively complex, with sigmoidal activation, and a second very slowly developing component. The initial changes in closed pre-activated states during the activation process may contribute to a mild relief of block by 293B through interactions which are not understood (Fig. 5B). However, chromanol appears to inhibit the development of the more slowly developing component of KCNQ1/KCNE1, or blocks it completely relative to earlier open steps in the activation of the channel. The result suggests that 293B has differing levels of interaction with different open states of the channel.

Seebohm et al. (2003b) proposed that the complex activation time courses of KCNQ1 and KCNQ1/KCNE1 are a result of these channels having multiple open states. These open states are kinetically distinct, but both presumably present an open pore with differing stability. One explanation for our results is that these states may be pharmacologically distinct (i.e. they can be distinguished by their differing affinities for drugs). While the binding site for the drug is presumably the open pore, different residues may line the pore in each putative open state. If so, each open state may have a different affinity for drugs such as 293B. The balance between the open states may depend on the presence of associated subunits. Thus, the differing apparent affinity for the open pore with different subunits may represent the effects the subunits have on the balance between these different open states. This case represents a re-balancing of drug binding states which are identical to those open channel states seen in the KCNQ1 channel. Conversely, the subunits may force the channel to adopt a completely new open state or states. The initial rapid block of KCNQ1/KCNE1 is similar to the time-independent block of KCNQ1 channels. However, block of KCNQ1/KCNE1 increases during the depolarization, possibly inhibiting the channel from entering a second, very slowly developing, open state.

How can the subunits produce distinct binding conformations with different affinities for chromanol? Many lipophilic drugs bind in the open intracellular pore of potassium channels. The major regions involved in drug binding have been identified as residing on the intracellular half of S6 (Yeola et al. 1996; Zhou et al. 2001). Melman et al. (2004) demonstrated that KCNE subunits strongly interact with the pore, and particularly with S6, to regulate gating activity. In Kv1 channels, the end of the short helical segment just before the GYG selectivity filter, has also been implicated as being important in drug binding. In Kv1 channels this segment has the sequence TMTT, with the M being the critical residue for interaction with quinidine (Tamkun et al. 1994). The analogous region in KCNQ1 channels has the sequence TVTT. Significantly, Seebohm et al. (2005) demonstrated that this valine is critical in producing the gating phenotype of KCNQ1 channels and KCNQ1/KCNE1 channels. The orientation of this valine (V310) is modulated by interactions with both S5 and S6. These and other structural experiments all indicate that open channel block at the intracellular pore should be strongly modulated by KCNE subunit expression (Panaghie et al. 2006).

In general, the activation gate of potassium channels appears to involve opening of the inner mouth of the pore, i.e. the lower half of the S6 domain (Seebohm et al. 2006). The fact that KCNE subunits interact with this region and strongly modify activation gating is consistent with this mechanism. Since binding of hydrophobic drugs in this region is ubiquitous, the strong effect of KCNE subunits on chromanol potency also seems consistent with this mechanism. However, our results with the mutant channel KCNQ1/KCNE3[V72T] may appear to contradict this close relationship between modification of the activation gate and drug binding. The apparent affinity of the open state of KCNQ1/KCNE3[V72T] for 293B is the same as for KCNQ1/KCNE3, even though there has been a dramatic alteration in the activation gating kinetics. However, these data can be reconciled by considering the possibility of multiple open states, e.g. KCNQ1/KCNE3[V72T] may have an open state similar to one of the KCNQ1/KCNE1 open states. There are several lines of evidence to support this idea. First, rapid stimulation potentiates the KCNQ1/KCNE3[V72T] current, with a time course identical to that of the 293B-insenstive open state of KCNQ1/KCNE1. In addition, the open state of KCNQ1/KCNE3[V72T] is insensitive to the presence of 293B, similar to the 293B-insenstive open state of KCNQ1/KCNE1. Finally, development of 293B block of KCNQ1/KCNE3[V72T] channels during a step depolarization has a similar time course to the development of 293B block of KCNQ1/KCNE1, except for an initial unbinding of drug in KCNQ1/KCNE1 (the mild relief of block seen in Fig. 5B). The V72T mutation alters activation gating, but does not change the apparent affinity for the drug. Importantly, even though the apparent affinity is not changed, the mutation does alter the time-dependent properties of drug binding. This is consistent with the residue at position 72 influencing the transitions of the activation gate and accessibility to the binding site. However, other residues in the KCNE subunit must be involved in holding the binding site for chromanol in higher and lower affinity conformations.

Our experiments demonstrate that KCNE co-expression does indeed strongly modulate open channel block by drugs. In this sense, chromanol serves as a pharmacological probe of the ability of KCNE subunits to change pore structure and provides confirmation of the structural hypotheses developed in the preceding biophysical studies based on kinetic analysis of site directed mutations. In physiological terms, this study demonstrates that the pharmacology of KCNQ1 channels is strongly dependent on the expression of ancillary subunits. This subunit-specific pharmacology suggests that drugs that are targeted to KCNQ in specific organs can be made more specific by considering subunit characteristics. Furthermore, the kinetic differences in gating and the corresponding differences in drug use-dependence suggest that the response of KCNQ1 channels to a drug will depend on the patterns of electrical activity of a particular organ or cell type, providing the opportunity for additional specificity of action. Understanding these molecular properties will help in the development of organ-specific pharmacological therapies, and the reduction of adverse side-effects associated with current pharmacological interventions.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix
  • Baukrowitz T & Yellen G (1995). Modulation of K+ current by frequency and external [K+]: a tale of two inactivation mechanisms. Neuron 15, 951960.
  • Bett GC, Morales MJ, Strauss HC & Rasmusson RL (2006). KChIP2b modulates the affinity and use-dependent block of Kv4.3 by nifedipine. Biochem Biophys Res Commun 340, 11671177.
  • Bett GC & Rasmusson RL (2004). Inactivation and recovery in Kv1.4 K+ channels: lipophilic interactions at the intracellular mouth of the pore. J Physiol 556, 109120.
  • Bleich M, Briel M, Busch AE, Lang HJ, Gerlach U, Gogelein H, Greger R & Kunzelmann K (1997). KVLQT channels are inhibited by the K+ channel blocker 293B. Pflugers Arch 434, 499501.
  • Brahmajothi MV, Morales MJ, Liu S, Rasmusson RL, Campbell DL & Strauss HC (1996). In situ hybridization reveals extensive diversity of K+ channel mRNA in isolated ferret cardiac myocytes. Circulation Res 78, 10831089.
  • Chouabe C, Neyroud N, Guicheney P, Lazdunski M, Romey G & Barhanin J (1997). Properties of KvLQT1 K+ channel mutations in Romano-Ward and Jervell and Lange-Nielsen inherited cardiac arrhythmias. EMBO J 16, 54725479.
  • Demolombe S, Franco D, De Boer P, Kuperschmidt S, Roden D, Pereon Y, Jarry A, Moorman AF & Escande D (2001). Differential expression of KvLQT1 and its regulator IsK in mouse epithelia. Am J Physiol Cell Physiol 280, C359C372.
  • Gould TD & Pfeifer K (1998). Imprinting of mouse Kvlqt1 is developmentally regulated. Hum Mol Genet 7, 483487.
  • Jentsch TJ (2000). Neuronal KCNQ potassium channels: physiology and role in disease. Nat Rev Neurosci 1, 2130.
  • Lan WZ, Abbas H, Lemay AM, Briggs MM & Hill CE (2005). Electrophysiological and molecular identification of hepatocellular volume-activated K+ channels. Biochim Biophys Acta 1668, 223233.
  • Mason DE, Mitchell KE, Li Y, Finley MR & Freeman LC (2002). Molecular basis of voltage-dependent potassium currents in porcine granulosa cells. Mol Pharmacol 61, 201213.
  • Melman YF, De La Domenech A, LS & McDonald TV (2001). Structural determinants of KvLQT1 control by the KCNE family of proteins. J Biol Chem 276, 64396444.
  • Melman YF, Krumerman A & McDonald TV (2002). A single transmembrane site in the KCNE-encoded proteins controls the specificity of KvLQT1 channel gating. J Biol Chem 277, 2518725194.
  • Melman YF, Um SY, Krumerman A, Kagan A & McDonald TV (2004). KCNE1 binds to the KCNQ1 pore to regulate potassium channel activity. Neuron 42, 927937.
  • Neyroud N, Tesson F, Denjoy I, Leibovici M, Donger C, Barhanin J, Faure S, Gary F, Coumel P, Petit C, Schwartz K & Guicheney P (1997). A novel mutation in the potassium channel gene KVLQT1 causes the Jervell and Lange-Nielsen cardioauditory syndrome. Nat Genet 15, 186189.
  • Panaghie G, Tai KK & Abbott GW (2006). Interaction of KCNE subunits with the KCNQ1 K+ channel pore. J Physiol 570, 455467.
  • Pusch M, Ferrera L & Friedrich T (2001). Two open states and rate-limiting gating steps revealed by intracellular Na+ block of human KCNQ1 and KCNQ1/KCNE1 K+ channels. J Physiol 533, 135143.
  • Sanguinetti MC, Curran ME, Zou A, Shen J, Spector PS, Atkinson DL & Keating MT (1996). Coassembly of KVLQT1 and minK (IsK) proteins to form cardiac IKs potassium channel. Nature 384, 8083.
  • Schroeder BC, Waldegger S, Fehr S, Bleich M, Warth R, Greger R & Jentsch TJ (2000). A constitutively open potassium channel formed by KCNQ1 and KCNE3. Nature 403, 196199.
  • Seebohm G, Chen J, Strutz N, Culberson C, Lerche C & Sanguinetti MC (2003a). Molecular determinants of KCNQ1 channel block by a benzodiazepine. Mol Pharmacol 64, 7077.
  • Seebohm G, Lerche C, Pusch M, Steinmeyer K, Bruggemann A & Busch AE (2001a). A kinetic study on the stereospecific inhibition of KCNQ1 and IKs by the chromanol 293B. Br J Pharmacol 134, 16471654.
  • Seebohm G, Sanguinetti MC & Pusch M (2003b). Tight coupling of rubidium conductance and inactivation in human KCNQ1 potassium channels. J Physiol 552, 369378.
  • Seebohm G, Scherer CR, Busch AE & Lerche C (2001b). Identification of specific pore residues mediating KCNQ1 inactivation. A novel mechanism for long QT syndrome. J Biol Chem 276, 1360013605.
  • Seebohm G, Strutz-Seebohm N, Ureche ON, Baltaev R, Lampert A, Kornichuk G, Kamiya K, Wuttke TV, Lerche H, Sanguinetti MC & Lang F (2006). Differential roles of S6 domain hinges in the gating of KCNQ potassium channels. Biophys J 90, 22352244.
  • Seebohm G, Westenskow P, Lang F & Sanguinetti MC (2005). Mutation of colocalized residues of the pore helix and transmembrane segments S5 and S6 disrupt deactivation and modify inactivation of KCNQ1 K+ channels. J Physiol 563, 359368.
  • Tamkun MM, Knittle TJ, Deal KK, House MH, Roberds SL, Po S, Bennett PB, George AL Jr & Snyders DJ (1994). Molecular physiology of voltage-gated potassium and sodium channels: Ion channel diversity within the cardiovascular system. In Ion Channels in the Cardiovascular System, ed. SpoonerPM, BrownAM, CatterallWA, KaczorowskiGJ & StraussHC, pp. 287316. Futura Publishing Co., Inc., Armonk , NY .
  • Tristani-Firouzi M & Sanguinetti MC (1998). Voltage-dependent inactivation of the human K+ channel KvLQT1 is eliminated by association with minimal K+ channel (minK) subunits. J Physiol 510, 3745.
  • Tsevi I, Vicente R, Grande M, Lopez-Iglesias C, Figueras A, Capella G, Condom E & Felipe A (2005). KCNQ1/KCNE1 channels during germ-cell differentiation in the rat: expression associated with testis pathologies. J Cell Physiol 202, 400410.
  • Vetter DE, Mann JR, Wangemann P, Liu J, McLaughlin KJ, Lesage F, Marcus DC, Lazdunski M, Heinemann SF & Barhanin J (1996). Inner ear defects induced by null mutation of the isk gene. Neuron 17, 12511264.
  • Wang Q, Curran ME, Splawski I, Burn TC, Millholland JM, VanRaay TJ et al. (1996). Positional cloning of a novel potassium channel gene: KVLQT1 mutations cause cardiac arrhythmias. Nature Genet 12, 1723.
  • Wang S, Morales MJ, Liu S, Strauss HC & Rasmusson RL (1997). Modulation of HERG affinity for E-4031 by [K+]o and C-type inactivation. FEBS Lett 417, 4347.
  • Wang S, Morales MJ, Qu YJ, Bett GC, Strauss HC & Rasmusson RL (2003). Kv1.4 channel block by quinidine: evidence for a drug-induced allosteric effect. J Physiol 546, 387401.
  • Wulfsen I, Hauber HP, Schiemann D, Bauer CK & Schwarz JR (2000). Expression of mRNA for voltage-dependent and inward-rectifying K channels in GH3/B6 cells and rat pituitary. J Neuroendocrinol 12, 263272.
  • Yang WP, Levesque PC, Little WA, Conder ML, Shalaby FY & Blanar MA (1997). KvLQT1, a voltage-gated potassium channel responsible for human cardiac arrhythmias. Proc Natl Acad Sci U S A 94, 40174021.
  • Yeola SW, Rich TC, Uebele VN, Tamkun MM & Snyders DJ (1996). Molecular analysis of a binding site for quinidine in a human cardiac delayed rectifier K+ channel. Role of S6 in antiarrhythmic drug binding. Circulation Res 78, 11051114.
  • Zhou M, Morais-Cabral JH, Mann S & MacKinnon R (2001). Potassium channel receptor site for the inactivation gate and quaternary amine inhibitors. Nature 411, 657661.

Appendix

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Appendix

Acknowledgements

This work was supported in part by National Heart, Lung, and Blood Institute grant HL-62465, a grant from the Oishei Foundation, and local and national scientist development grants from the American Heart Association.