The working hypothesis for investigating the role of KCNQ-encoded K+ channels (Kv7) in regulating smooth muscle was described in the Introduction (see Figure 1). The assumption is that, like neurons, Kv7 channels are open at rest and the K+ efflux through these channels contributes to the formation of the resting membrane potential. Blockade of these channels will lead to reduced hyperpolarization and an increase in the open probability of VDCCs (Figure 1). If the influx of Ca2+ through VDCCs is sufficient to overpower the inherent homeostatic mechanisms (i.e. Ca-ATPases, Na/Ca exchangers), then a contraction or enhanced response to a low concentration of a vasoconstrictor will occur (see Figure 1). Pharmacological dissection of a functional role relies on good probes, ideally with high selectivity for a certain target ie ion channel. For Kv7 channels the most common research tools are XE991 (10,10-bis(4-pyridinylmethyl)-9(10H)-anthracenone), which inhibits Kv7.1–7.4 with IC50 values ∼1–5 µmol·L−1 (see Yeung et al., 2007) and Kv7.5 with an IC50∼60 µmol·L−1 (Jensen et al., 2005; Yeung et al., (2008a) or the less potent linopirdine (3,3-bis(4-pyridinylmethyl)-1-phenylindolin-2-one, DUP996; Dupuis et al., 2002). Any pharmacological experiment is undertaken with a consideration of the necessary caveats ie selectivity. To date XE991 has been shown to have no effect on Kv1.2, Kv4.3 or ether-a-go-go (EAG)-encoded channels at concentrations below 100 µmol·L−1 (Wang et al., 2000), reduces Kv2.1 channels by only 20% at 100 µmol·L−1 (Wladyka and Kunze, 2006) and inhibits heterologously expressed ERG channels with an IC50 of 107 µmol·L−1 (Elmedyb et al., 2007). Consequently, at concentrations that inhibit Kv7 channels XE991 appears to be selective. However, it is possible that XE991 may have an, as yet unreported, effect on other ion channels in the micromolar range where Kv7 block occurs. There are also other compounds, such as chromanol 293B, HMR 1556 or L-768-673, which preferentially block Kv7.1 containing channels (see Seebohm et al., 2003; Bett et al., 2006; Dong et al., 2006; Lerche et al., 2007). In addition to Kv7 channel blockers there are a few compounds, namely retigabine, flupirtine and the acrylamide S-1 that enhance Kv7 channel activity (see Miceli et al., 2008 for overview). These agents provide an alternative mechanism to probe for a functional role for Kv7 channels. Moreover, because they only activate the so-called neuronal Kv7 channels (encoded by KCNQ2–5, Schenzer et al., 2005; Wuttke et al., 2005) they provide a tool to discriminate Kv7.1-mediated function from that generated by Kv7.2–7.5 channels. In fact, the acrylamide (S)-1 appears to have greater efficacy activating Kv7.4 or 7.5 rather than Kv7.2 or 7.3 (Bentzen et al., 2006).
The electrophysiological effects of XE991 and linopirdine have been assessed in studies on myocytes from mouse portal vein (Ohya et al., 2003; Yeung and Greenwood, 2005), rat mesenteric artery (Mackie et al., 2008) and A7r5-cultured aorta cells (Brueggemann et al., 2007). In the mouse portal vein depolarizing voltage steps evoked a time-dependent current superimposed on a linear, time-independent current. Application of 5 mmol·L−1 4-aminopyridine effectively blocked the time-dependent current but had minimal effect on the linear current. In comparison, XE991 (1–30 µmol·L−1) was a more effective inhibitor of the linear current rather than the time-dependent component. The XE991-sensitive current was well sustained, showing no decrease in current amplitude during a 45 s depolarizing pulse (Yeung and Greenwood, 2005). XE991 (30 µmol·L−1) also inhibited completely the augmented holding current at −60 mV produced by bathing the cells in an external solution containing 36 mmol·L−1 KCl (Yeung and Greenwood, 2005). Mackie et al. (2008) employed a different protocol where they held at −4 mV to allow inactivation of the majority of Kv channels and calculated current activation in the absence and presence of linopirdine. Using this protocol these authors calculated that the Kv7 channels in rat mesenteric arteries were 50% activated at −34 mV. These findings are consistent with a role for Kv7 channels in setting the resting membrane potential. This was confirmed by the observation that blockade of Kv7 channels by either XE991 or linopirdine produced a marked depolarization in murine portal vein (Yeung and Greenwood, 2005), rat mesenteric artery (Mackie et al., 2008) and A7r5-cultured myocytes (Brueggemann et al., 2007). The vascular effects of Kv7 channel enhancers have only been reported in murine portal vein myocytes, rat mesenteric arteries and A7r5 cells (Brueggemann et al., 2007; Mackie et al., 2008; Yeung et al., 2008b). In mouse portal vein myocytes retigabine and flupirtine augmented currents at potentials between −60 and 0 mV, suppressed spontaneous membrane depolarizations and elicited a membrane hyperpolarization of ∼12 mV (Yeung et al., 2008b). However, at more depolarized potentials the Kv7 activators inhibited the voltage-dependent K+ currents. The inhibitory and stimulatory effects of retigabine were abrogated by prior application with XE991 consistent with retigabine having a bi-modal effect on vascular Kv7 currents. In A7r5 cells flupirtine (10 µmol·L−1) augmented K+ currents markedly and reversed completely the membrane depolarization produced by 25 pmol·L−1 vasopressin (Brueggemann et al., 2007). Taken together, it is clear that KCNQ currents are essential components of delayed rectifier K+ currents in vascular smooth muscle cells. Further studies with gene-silencing technology might help to confirm major contributors to vessel KCNQ channels: KCNQ1a/b, KCNQ4 and/or KCNQ5. Dysfunction of KCNQ channels might be associated with development of vascular disorders such as hypertension as reported for other Kv channels.
Kv7 channels and contractility. Figure 1 describes the working model underlying the functional impact of Kv7 channels in smooth muscle cells. Consequently, blockade of these channels should generate contraction of quiescent vessels and increased contractile activity in spontaneously active tissues. Conversely, enhancement of Kv7 channel activity should lead to membrane hyperpolarization with a concomitant reduction of the contractile response. An effect of Kv7 channel blockers on smooth muscle contraction was observed initially in whole mouse portal veins (Yeung and Greenwood, 2005). This blood vessel generates spontaneous, phasic contractile activity manifest as bursts of contractions ∼5 s apart (see Yeung and Greenwood, 2005). XE991 (10 µmol·L−1) and linopirdine increased the frequency and amplitude of contractions considerably, an effect also seen in the presence of 4-AP at a concentration that would block most non-KCNQ Kv channels, as well as glibenclamide and paxilline to block ATP-sensitive and Ca-activated K+ channels (Yeung and Greenwood, 2005). Interestingly, chromanol 293B at a concentration that would block Kv7.1 channels (see Bett et al., 2006) had little effect on portal vein contractions. This initial work was followed by a study by Joshi et al. (2006) showing that XE991 and linopirdine were effective spasmogens of rat pulmonary artery segments but not mesenteric arteries. The contractions of pulmonary artery were completely reversed by reducing the entry of Ca2+ through voltage-dependent channels either directly, with a dihydropyridine, or indirectly by hyperpolarizing the membrane potential with the ATP-sensitive K+ channel opener cromakalim. Subsequently, Yeung et al. (2007) performed an exhaustive analysis of the functional effect of Kv7 channel blockers in a variety of murine arterial preparations. They reported that XE991 and linopirdine contracted segments of murine aorta (thoracic and abdominal), carotid artery and femoral artery. Contraction of mesenteric artery was also observed especially if a small degree of pre-tone was supplied by a very low concentration of phenylepherine. Mackie et al. (2008) also saw a reduction of rat mesenteric artery diameter using perfusion myography. The variable effect of XE991 on the mesenteric artery seen by Joshi et al. (2006), Yeung et al. (2007) and Mackie et al. (2008) likely reflects the need for a depolarizing drive to generate contraction that is supplied either by a low concentration of vasoconstrictor or by perfusion pressure. Interestingly, Mackie et al. (2008) showed that linopirdine increased the mean arterial pressure and mesenteric vascular resistance in vivo. Mackie et al. (2008) also postulated that the contraction of mesenteric arteries evoked by a low concentration of vasopressin was mediated by an inhibition of Kv7 channels, reminiscent of the effect of muscarinic agonists on neuronal activity (e.g. Brown and Adams, 1980). However, Brueggemann et al. (2007) showed that Kv7 activator flupirtine was able to reverse the membrane depolarization produced by vasopressin, which would not be expected if the vasoconstrictor had blocked the Kv7 channels (see Yeung et al., 2008b). Moreover, it is worth considering that in many vessels the constrictor effect of XE991 is augmented by a low concentration of vasoconstrictor such as phenylepherine, 5-HT or angiotensin II (Yeung et al., 2007). Future experiments will explore this interesting mechanism further.
While the pan-Kv7 channel blockers, XE991 and linopirdine contract blood vessels no Kv7.1-selective blocker (i.e. chromanol 293B, HMR 1556 or L-768-673) has been shown to be spasmogenic (see Yeung et al., 2007). This suggests that while KCNQ1 is expressed in all blood vessels the expression product has little impact on the resting membrane potential in these cells. This proposal is supported by the ability of retigabine to relax precontracted murine aortae, carotid artery, femoral artery and mesenteric artery (Yeung et al., 2007). This agent does not affect Kv7.1 channels but augments the activity of Kv7.2–7.5 by shifting the voltage-dependence of activation to more negative potentials through an interaction with a key tryptophan residue in the S5 transmembrane domain (Schenzer et al., 2005; Wuttke et al., 2005). Retigabine relaxed aortic segments precontracted by phenylepherine or 4-aminopyridine to the same extent as nicardipine but had no effect on contractions evoked by blockade of Kv7 channels by XE991. Moreover, the vasorelaxant effects of retigabine were reversed by application of XE991 but not 4-AP. These data are consistent with retigabine working solely on Kv7 channels. Similar effects were observed with flupirtine as well as meclofenamic acid, a cyclooxygenase inhibitor that also activates Kv7 channels (Peretz et al., 2005). Flupirtine also lowers mean arterial pressure in rats in vivo (Mackie et al., 2008). These findings consolidate the view that Kv7 channels other than Kv7.1 are major players in defining vascular reactivity. It also lends credence to the proposal that calling these KCNQ genes ‘neuronal’ may need to be refined.
KCNQ expression and function in non-vascular smooth muscle.
Early electrophysiological studies showed that acetylcholine increased the excitability of toad and guinea-pig stomach partially through suppression of a resting K+ conductance (e.g. Sims et al., 1985; Lammel et al., 1991), similar to the situation in central neurons (Brown and Adams, 1980). Logically if the M-channels in neurons are encoded by KCNQ genes then similar resting K+ conductances in gastric smooth muscle would have a similar molecular identity. However, in contrast to the research on KCNQ genes in vascular smooth muscle there is a dearth of data in non-vascular cell types. A number of groups have shown KCNQ1/KCNE3 expression in the epithelial cells of the gastrointestinal tract (Schroeder et al., 2000b; Dedek and Waldegger, 2001), but none investigated expression beyond these cells in the smooth muscle layers. However, Ohya et al. (2002a) showed that KCNQ1, KCNQ3, KCNE1 and KCNE2 transcripts were expressed in rat gastric antral smooth muscle layers. Due to the limitation of rat KCNQ and KCNE homologues in the DNA database at the time, this study only examined KCNQ1–3 and KCNE1–2 expression, however recently message for KCNQ4, KCNQ5 and KCNE4 were detected in murine proximal and distal colon (Jepps et al., 2007). Message for the truncated KCNQ5 was also detected in murine oesophagus, colon and ileum (Yeung et al., 2008a). In contrast to Ohya et al. (2002a) where no functional experiments were performed, Jepps et al. (2007) showed that XE991 increased the spontaneous contractions of the murine colon. More recently, KCNQ and KCNE expression was detected in murine myometrial smooth muscles throughout the oestrus cycles (McCallum et al., 2008). Real-time PCR and Western blot analyses showed the most readily detected message throughout the oestrus cycle was for KCNQ1, KCNQ5 and KCNE4. However, there was no indication about whether the KCNQ5 expressed was the shorter variant described in vascular smooth muscle (Yeung et al., 2008a). The level of expression for all genes was remarkably consistent over the oestrus cycle except for KCNQ1, KCNQ5 and KCNE1, which were up-regulated in metestrous. XE991 augmented and retigabine decreased the spontaneous contractility at all parts of the oestrus cycle. Again, chromanol 293B had no effect on contractility, even at metestrous when KCNQ1 and KCNE1, whose expression products co-associate to produce a channel with a higher sensitivity to chromanol, were raised. These data provided new insight into the control of uterine activity and presented a possible new therapeutic target for disorders due to uterine dysfunction (e.g. preterm labour).
Several patents granted to American Home Products, Wyeth and ICAgen describe KCNQ activators for the use of disorders in urinary bladder such as overactive bladder, bladder incontinence, bladder spasms and bladder outflow obstruction. KCNQ channels expression has been detected in the bladder, with KCNQ4, KCNQ5 and KCNE4 predominant in murine urinary bladder (S. Ohya, unpubl. obs.). Moreover, retigabine completely inhibited acetic acid-induced micturition, decreased baseline and maximal bladder pressures and increased voided and infused volumes (Streng et al., 2004; Wickenden et al., 2004; Argentirei and Butera, 2006). In guinea-pig bladder detrusor strips XE991 and linopirdine increased and flupirtine decreased spontaneous contractility (Carson and McCloskey, 2007). Interestingly, the same group observed XE991-sensitive currents in the interstitial cells that coexist with the smooth muscle cells and influence their activity (Anderson and McCloskey, 2007). This raises the intriguing possibility that KCNQ-encoded channels located in these pacemaker cells as well as the smooth muscle cells may regulate smooth muscle function.