Contribution of Kv7.4/Kv7.5 Heteromers to Intrinsic and Calcitonin Gene-Related Peptide–Induced Cerebral ReactivitySignificance
Objective—Middle cerebral artery (MCA) diameter is regulated by inherent myogenic activity and the effect of potent vasodilators such as calcitonin gene-related peptide (CGRP). Previous studies showed that MCAs express KCNQ1, 4, and 5 potassium channel genes, and the expression products (Kv7 channels) participate in the myogenic control of MCA diameter. The present study investigated the contribution of Kv7.4 and Kv7.5 isoforms to myogenic and CGRP regulation of MCA diameter and determined whether they were affected in hypertensive animals.
Approach and Results—Isometric tension recordings performed on MCA from normotensive rats produced CGRP vasodilations that were inhibited by the pan-Kv7 channel blocker linopirdine (P<0.01) and after transfection of arteries with siRNA against KCNQ4 (P<0.01) but not KCNQ5. However, isobaric myography revealed that myogenic constriction in response to increases in intravascular pressure (20–80 mm Hg) was affected by both KCNQ4 and KCNQ5 siRNA. Proximity ligation assay signals were equally abundant for Kv7.4/Kv7.4 or Kv7.4/Kv7.5 antibody combinations but minimal for Kv7.5/Kv7.5 antibodies or Kv7.4/7.1 combinations. In contrast to systemic arteries, Kv7 function and Kv7.4 abundance in MCA were not altered in hypertensive rats.
Conclusions—This study reveals, for the first time to our knowledge, that in cerebral arteries, Kv7.4 and Kv7.5 proteins exist predominantly as a functional heterotetramer, which regulates intrinsic myogenicity and vasodilation attributed to CGRP. Surprisingly, unlike systemic arteries, Kv7 activity in MCAs is not affected by the development of hypertension, and CGRP-mediated vasodilation is well maintained. As such, cerebrovascular Kv7 channels could be amenable for therapeutic targeting in conditions such as cerebral vasospasm.
The control of cerebral artery diameter is a complicated process, but crucial for maintaining the delivery of oxygen and nutrients to neurones. Dysregulation of cerebral artery diameter, such as during vasospasm associated with subarachnoid hemorrhage, leads to neuronal hypoxia or even cellular death. As such, it is important to define the molecular mechanisms that regulate cerebral arteries so that disease genesis can be determined and new therapeutic mechanisms explored. Work emanating from our laboratory and others has led to KCNQ-encoded voltage-dependent potassium channels (Kv7) being recognized as important regulators of smooth muscle contractility.1–4 We have previously demonstrated that resistance cerebral arteries express KCNQ1, 4, and 5, and Kv7.2 to Kv7.5 activators, such as S-1 or retigabine, are effective suppressors of the inherent myogenic response in the middle cerebral arteries (MCAs)5 and relax preconstricted basilar arteries,6,7 but there is no information on the functional contribution of individual Kv7 isoforms or what constitutes a functional channel. Cerebral blood flow is also regulated by strong vasodilatory effects of neuropeptides, such as calcitonin gene-related peptide (CGRP) released from perivascular nerves.8,9 The activation of ATP-sensitive potassium channels (KATP) underlies a portion, but not all, of vasodilation associated with CGRP receptor stimulation.10,11 Recently, we showed that β-adrenoceptor–mediated vasodilation of renal arteries was inhibited by pan-Kv7 blockers, such as XE991 or linopirdine, as well as by knockdown of KCNQ4 by targeted siRNA oligonucleotides.2 Because β-adrenoceptors and CGRP receptors are both Gs-coupled receptors, it is possible that the activation of Kv7 channels may also contribute to CGRP-induced relaxations of MCA. Importantly, recent evidence demonstrated that Kv7.4 channels in systemic arteries, such as the mesenteric and renal, are severely compromised in hypertension.2,4 A similar situation in cerebral circulation would lead to a predilection toward vasospasm and may underlie ischemic stroke. Here, we used a combination of molecular and functional approaches to define the interaction of different Kv7 channels in MCA and to ascertain the functional impact of individual Kv7 isoforms in the intrinsic and CGRP-mediated regulation of arterial diameter.
Materials and Methods
Materials and Methods are available in the online-only Supplement.
Initial studies used a pharmacological approach to tease out a functional role for KCNQ1, 4, and 5 that are dominantly expressed in the MCA.5 The application of a selective Kv7.1 channel blocker, HMR1556 (10 μmol/L),12 had no effect on MCA tone, whereas the pan-Kv7 channel blocker, linopirdine, evoked robust contractions of MCAs under similar conditions (Figure 1A). In MCAs precontracted with 0.1 μmol/L U46619, the application of Kv7.2 to Kv7.5 channel activators, retigabine and S-1, caused relaxation in a concentration-dependent manner with the latter being more potent (Figure 1B). In contrast, the Kv7.1-selective activator, R-L3, was an ineffective relaxant of MCAs (Figure 1B) at concentrations that produced relaxations of other arteries.12 Cumulative application of 5-HT concentrations evoked robust contractions of MCAs (Figure 1C). The inhibition of Kv7 channels by linopirdine (1 μmol/L; n=4) caused a moderate enhancement in 5-HT responses. Conversely, S-1 (10 μmol/L; n=4) significantly attenuated 5-HT contractions. These data confirm that Kv7 channels are important regulators of cerebral artery tone and reveal that Kv7.1 channels have no functional impact in these arteries.
Kv7 and KATP Blockers Inhibit CGRP-Evoked Dilation
CGRP is a potent dilator of cerebral arteries through the stimulation of CGRP receptors and activation of Gs-linked cellular signaling. Based on our previous observations,2 we investigated whether Kv7 channels were involved in CGRP-mediated relaxations. Cumulative addition of CGRP (10–9–10–7 mol/L) caused the relaxation of precontracted MCAs with a pEC50 of 8.0±0.2 mol/L (n=5; see the Table). Although HMR1556 (10 μmol/L) had no effect on CGRP-induced dilations (data not shown), the blockade of all Kv7 channels by linopirdine (10 μmol/L) markedly reduced Emax dilation by ≈60% (n=6; Figure 2A; see the Table). The KATP blocker, glibenclamide (10 μmol/L; n=4), attenuated the maximum dilation to CGRP by ≈30% (Figure 2C; see the Table), and the subsequent addition of linopirdine abolished CGRP responses (Figure 2C; Table).
Because the inhibition of Kv7 channels by linopirdine has procontractile effects, it is possible that under these conditions CGRP is unable to relax the vessels (ie, physiological antagonism). To control for this, CGRP responses were conducted in the presence of 4-AP at 1 mmol/L, a concentration known to block several Kv channels but not Kv7. The application of 1 mmol/L 4-AP caused some contraction but did not affect CGRP-evoked relaxations (Figure 2B; Table). Moreover, nicardipine still relaxed the precontracted MCAs in the presence of linopirdine (n=4). These results demonstrate that Kv7 channels account for a considerable portion of relaxation attributed to CGRP in MCAs, whereas KATP underlie the remaining component.
CGRP-Induced Dilation Depends on Kv7.4, But Not Kv7.5 Channel Expression
Based on a lack of effect with Kv7.1-specific modulators, we consider Kv7.4 and Kv7.5 to be the functionally relevant isoforms in MCA. Because there are no specific blockers of Kv7.2 to Kv7.5 isoforms, we used a molecular silencing approach with siRNA targeted against KCNQ4 or 5 to dissect out the role of each subunit. Because of poor specificity of KCNQ5 antibodies tested previously (data not shown), QPCR was used to determine the knockdown of siRNA in cerebral arteries. Figure 3A shows that KCNQ4 expression was reduced in arteries transfected with KCNQ4-targetted siRNA compared with arteries with KCNQ5-targetted siRNA (n=3). This treatment also resulted in ≈40% decrease in Kv7.4 protein (Figure I in the online-only Data Supplement). Similarly, KCNQ5 expression was relatively less in arteries with KCNQ5-targetted siRNA compared with KCNQ4 siRNA (n=3, Figure 3B).
The depletion of Kv7.4 channels using KCNQ4 siRNA reduced relaxation to the Kv7.2 to Kv7.5 activator S-1 (0.3–10 μmol/L; n=11; Figure 3C) and the Kv7.2/Kv7.4 activator ML213 (1 μmol/L; Figure III in the online Data Supplement). Treatment with KCNQ4 siRNA attenuated the relaxation induced by CGRP (Figure 3D) but did not alter relaxation to the NO donor sodium nitroprusside or nicardipine (Figure III in the online Data Supplement). Surprisingly, the depletion of Kv7.5 channels had no significant effect on relaxations produced by the Kv7 activator S-1 or CGRP compared with scrambled controls (Figure 3E and 3F). In pressurized MCA, KCNQ5 siRNA also had no significant effect on vasodilation produced by 10 μmol/L S-1 (88±2.0% compared with 85.5±4.5% for vessels incubated with scrambled siRNA; n=4; Figure IIB in the online-only Data Supplement), whereas KCNQ4 siRNA reduced the response markedly (38.5±3.0% vasodilation; n=4). Thus, dilations produced by Kv7 channel activators (S-1) or indirect activation by CGRP-induced signaling pathways are highly dependent on Kv7.4 but not Kv7.5 levels.
Effect of Isoform-Specific Knockdown on Myogenic Constriction
Increasing intraluminal pressure produced an active reduction in the vessel diameter in freshly dissected arteries and vessels transfected with scrambled siRNA. The inhibition of Kv7 channels by linopirdine (1 μmol/L) produced constriction at all pressures up to 70 mm Hg in freshly isolated arteries5 (Figure IV in the online-only Data Supplement). The depletion of Kv7.4 channels also produced greater constriction across the pressure range compared with the vessels incubated with scrambled siRNA (Figure 4A, 4B, and 4D), whereas the downregulation of Kv7.5 channels by targeted siRNA markedly increased active myogenic constriction at nearly all pressures, including the physiological range (40–80 mm Hg; n=4; Figure 4A, 4C, and 4E).
Proximity Ligation Assay Detection of Kv7.4/7.5 Coassembly
Proximity ligation assay (PLA) was used to determine the preponderance of individual Kv7.4 and Kv7.5 channels compared with Kv7.4/7.5 heteromers in the MCA. Figure 5A shows that combinations of 2 different Kv7.4 antibodies resulted in a large number of PLA puncta (≈40 per cell per optical slice; Figure 5A and 5B), indicative of closely associated Kv7.4 proteins. The puncta localization reveals the position of Kv7.4/7.4 protein interaction to be at intracellular sites as well as at a plasma membrane–like localization (Figure 5A). A similar localization and number of PLA signals were generated in smooth muscle cells interrogated with 3 different combinations of Kv7.4 and Kv7.5 antibodies (Figure 5A and 5B). However, when 2 Kv7.5 antibodies were combined, very few PLA signals (<20 per cell per optical slice; Figure 5A and 5B) were observed, suggesting that Kv7.5 is unlikely to exist as a homotetramer. Similarly, combining Kv7.4 and Kv7.1 antibodies did not produce significant PLA signals, highlighting that these proteins do not exist in the same cellular microdomain (Figure 5A and 5B). The Kv7.4 and Kv7.5 antibodies, applied alone as controls, did not result in any red fluorescent puncta (Figure 5B).
Kv7 Function Is Not Diminished in MCA From Spontaneously Hypertensive Rats
Having established that Kv7 channels are key determinants of the myogenic response and are essential components of CGRP-mediated relaxations, we determined whether the Kv7 activity was compromised in MCAs from spontaneously hypertensive rats (SHRs) similar to previous work.2,4,13 Consistent with our previous observations,4 responses to S-1 (3–10 μmol/L) were severely attenuated (P<0.0001) in mesenteric arteries from SHRs (n=7–11; Figure 6A). However, in MCA from the same SHRs, the response to S-1 was identical to that seen in MCAs from normotensive rats (Figure 6A; n=10 and 6, respectively). CGRP-induced changes in isometric tension or vasodilation in pressurized arteries were also not different in MCAs from normotensive rats or SHRs (n=5; Figure V in the online-only Data Supplement) and were similarly attenuated by 10 μmol/L linopirdine (n=5 each). At a molecular level, there was no difference in the abundance of Kv7.4 protein in the MCA from SHRs or normotensive rats (Figure 6B and 6C). In contrast, mesenteric arteries from the same SHRs exhibited appreciable reduction in Kv7.4 protein (Figure 6B and 6D) similar to previous work4 but no change in the level of KCNQ1 to 5 transcripts (consistent with Khanamiri et al).13
The present study provides new insight into the role of specific Kv7 isoforms in the regulation of cerebral artery diameter. It reveals that although Kv7.1 is present in MCAs,5 the Kv7.1-selective activator, R-L3, had no functional effect at concentrations that relax precontracted systemic and pulmonary arteries,12 and Kv7.1-selective blockers had no effect on MCA diameter per se or on CGRP-induced relaxation. This suggests that the functionally relevant isoforms in the MCA are Kv7.4 and Kv7.5, which, as our optical detection studies reveal, exist predominantly as a Kv7.4/7.5 heteromer. Moreover, we establish for the first time that CGRP-driven dilation of the MCA requires functional Kv7 channels with a strong reliance of Kv7.4 levels. Finally, the present study shows that the activity of Kv7 channel activators is preserved in the MCA from SHRs, suggesting that targeting these channels may be a viable target in the treatment of ischemic cerebral vasospasm. This is a key observation, because previous studies have shown that Kv1, Kv2, and BKCa channels are downregulated in cerebral arteries from hypertensive animals,14–16 so, Kv7 activators such as S-1 can overcome MCA constricted by Kv2 blockade.5
The preserved Kv7 activity in the MCA of SHRs is in stark contrast to the aorta, renal, mesenteric, and coronary arteries2,4,13 (and present study) where responses to various Kv7.2 to Kv7.5 activators are impaired. In the aorta, this is associated with a reduction in KCNQ4 transcripts and Kv7.4 abundance, whereas in other arteries, a decrease in Kv7.4 protein is observed without a change in transcription2,4,13 (and present study). Interestingly, responses to retigabine and 2 novel Kv7 activators are no different in the gracillus artery from normotensive rats and SHRs,17 although Kv7.4 levels were not determined. Trafficking of ion channel proteins to plama membranes and the regulation of channel activity once inserted are very complicated processes. For Kv7.4 proteins, very little is known about their regulation at the transcriptional and post-transcriptional levels in cells, especially vascular smooth muscle cells. Consequently, we have no insight into why Kv7.4 protein levels are protected in MCA but not in renal or mesenteric arteries. The contrasting observations in the present work will inform future studies aimed at understanding these important vascular regulators.
Kv7.4 depletion and the application of the Kv7 blocker, linopirdine, attenuated β-adrenoceptor–mediated relaxation of renal artery,2 adenosine-mediated relaxation of coronary artery,13 and the anticontractile effect of perivascular adipose.17,18 We now reveal that CGRP-mediated relaxation of cerebral arteries is highly dependent on Kv7 activity and Kv7.4, in particular. CGRP produces relaxations through stimulation of the CGRP receptor, which is an amalgamation of the calcitonin receptor protein and receptor activity–modifying protein 1,19,20 and is positively coupled to adenylate cyclase, leading to a rise in cAMP and usually recruitment of protein kinase A. Understanding the molecular mechanisms involved in CGRP-induced cerebral vasodilation is crucial because the loss of CGRP-mediated relaxation contributes to or aggravates spastic constriction.21,22 Alternatively, excessive CGRP activity after release from sensory nerves is strongly implicated in cerebral hyperdilation and increased cellular permeability that is manifest in migraine.23 Previous studies have suggested that the activation of KATP may be involved, although blocking these channels by glibenclamide has minor effects on CGRP responses11,24,25 (and present study). We now reveal that the inhibition of Kv7 channels with linopirdine or depletion of Kv7.4 by targeted siRNA had a marked effect on CGRP-induced relaxations. Moreover, a combination of glibenclamide and linopirdine abrogated the CGRP response. Consequently, the activation of Kv7.4 channels by Gs-linked receptor agonists may be a persistant feature of the vasculature, although the signaling mechanisms linking receptor activation and enhancement of Kv7.4 activity remain to be delineated.
In the present study, Kv7.1 blockers had no effect on resting tension, basal diameter at 80 mm Hg, or CGRP-mediated responses, whereas the pan-Kv7 blocker, linopirdine, affected all these physiological responses. These observations and the lack of effect of the Kv7.1 activator, R-L3, identify Kv7.4 and Kv7.5 channels as important regulators of cerebral blood flow. Our PLA studies allowed us to investigate the structural features of Kv7 channels in the MCA similar to previous studies on Kv2.1 and Kv9.3 association in the same artery.26 With this technique, hybridization of the PLA probes occurs when proteins are <40 nm apart.27 Because combinations of Kv7.5 antibodies resulted in considerably less PLA signal, the most parsimonious conclusion is that the number of channels containing ≥2 Kv7.5 proteins is low. In contrast, incubation with 2 Kv7.4 antibodies or different combinations of Kv7.4/Kv7.5 antibodies produced a high number of PLA signals per cell and of similar magnitude in Kv7.4/7.4 and Kv7.4/7.5 combinations. The ability to generate signal amplification reveals that Kv7.4 proteins reside sufficiently close enough to other Kv7.4 and Kv7.5 proteins. This seems to be a specific relationship because no PLA puncta were detected when antibodies against Kv7.1 and Kv7.4, which do not form heteromultimers,28 were combined. Moreover, the equivalence of PLA signal numbers generated with 2 Kv7.4 antibody or 3 different Kv7.4/Kv7.5 antibody combinations strongly suggests that Kv7.4/7.5 heteromers predominate, consistent with recent findings for mesenteric arteries.29 Extending this point further, it is likely that, because 2 Kv7.5 antibodies generate so few puncta, a Kv7.4/7.5 heteromer cannot contain 2 Kv7.5 contributions to the heteromer resulting in a likely Kv7.4/Kv7.5 stoichiometry of 3:1. It is important to recognize that the PLA probes used here do not exclusively detect subunits coassembled within single-channel complexes. Rather, the Duolink PLUS and MINUS oligonucleotides were of sufficient length to detect proteins within adjacent channel complexes. It is similarly worth noting that, in addition to subunit expression, the detection of channel expression levels is also sensitive to the binding affinity of each primary antibody for each subunit. In this study, we have endeavored to mitigate this limitation by using several different sets of antibodies (raised in different species), which all reveal a similar Kv7.4/Kv7.5 architecture.
In addition to the quantification data generated by PLA signals, the puncta provide exact spatial localization of the antibody recognition event.30 Here we report that Kv7.4/Kv7.4 and Kv7.4/Kv7.5 antibody combinations reveal an intracellular as well as plasma membrane–like localization. Previously, we have shown Kv7.4 concentrated at the plasma membrane with minor cytoplasmic localization.5 However, Kv7.4/Kv7.5 heteromers’ localization could be dynamic, with movement of proteins to and from the membrane. It is also important to note that, because the PLA method is based on equilibrium reactions and enzymatic steps, only a fraction of the interacting molecules is detected. Further studies are required using specific cellular markers together with PLA technology to ascertain the precise location of these protein interactions as well as the frequency of such events.
The preponderance of Kv7.4/Kv7.5 heteromers in the MCA explains why the depletion of Kv7.5 has little impact on CGRP- or S-1–mediated responses. Moreover, it explains why Kv7.4 knockdown mimics the effect of linopirdine on MCA diameter across a range of intraluminal pressures, whereas Kv7.5 depletion produced an augmented response at set pressures. The inability of linopirdine and KCNQ4 siRNA-treated vessels to produce discernable myogenic activity could be due to vessels being constricted and thus unable to respond to pressure changes. However, the diameter maintained in MCA transfected with KCNQ4 siRNA (≈140 μm) is not maximal, which allows the possibility for further constriction. At the moment, we do not know how a rise in cAMP after CGRP receptor activation enhances Kv7.4 activity, and we do not know how the presence of Kv7.5 affects the properties of heteromultimeric channels such that they are more important for the regulation of membrane potential and contractile response to intraluminal pressure elevation. These questions will be the focus of future studies.
It is worth stressing that our findings with the siRNA are not attributable to the nonspecific effects of transfection or culturing process because we observed no significant difference in either passive dilation or active responses to 60 mmol/L KCl and the voltage-gated calcium channel blocker, nicardipine. Moreover, we observed very different effects between the specific sets of targeted oligonucleotides at the same concentration and guanine–cytosine content. Of course, there are caveats to our findings. We have no insight into how Kv7 channels interplay with Kv1 and Kv2 channels that are also known to shape the myogenic response in these arteries,16,26,31–33 and there is always the possibility for off-target effects although these have been largely controlled for with different experimental protocols. The present study has advanced our understanding on the impact of Kv7 channels in the regulation of MCA diameter and the contribution of different isoforms to cerebral perfusion. We have identified that Kv7.4 activation underlies a major portion of CGRP-induced relaxations and highlighted that Kv7 channels remain responsive in hypertension, opening up the possibility that targeting these channels is a real therapeutic option for cerebrovascular diseases.
We thank all the staff at St George’s Biological Research Facility and Imaging Unit.
Sources of Funding
P.S.C. and J.B.S. were funded by British Heart Foundation grants awarded to Iain Greenwood (PG/09/104 and PG/12/63/29824). T.A.J. was funded by a BBSRC-CASE studentship (BB/G016321/1) in association with NeuroSearch A/S. G.C. was funded by Medical Research Council grant (MR/K019074/1). H.-L.Z. was supported by a Fellowship from Alberta Innovates-Health Solutions.
Current affiliation for P.S.C.: Victor Chang Cardiac Research Institute, Sydney, Australia.
Current affiliation for T.A.J.: Department of Biomedical Sciences, University of Copenhagen, Copenhagen, Denmark.
The online-only Data Supplement is available with this article athttp://atvb.ahajournals.org/lookup/suppl/doi:10.1161/ATVBAHA.114.303405/-/DC1.
- Nonstandard Abbreviations and Acronyms
- calcitonin gene-related peptide
- ATP-sensitive potassium channels
- voltage-dependent potassium channel
- middle cerebral artery
- proximity ligation assay
- spontaneously hypertensive rat
- Received October 15, 2013.
- Accepted February 10, 2014.
- © 2014 American Heart Association, Inc.
- Brueggemann LI,
- Mackie AR,
- Martin JL,
- Cribbs LL,
- Byron KL
- Jepps TA,
- Chadha PS,
- Davis AJ,
- Harhun MI,
- Cockerill GW,
- Olesen SP,
- Hansen RS,
- Greenwood IA
- Brain SD,
- Grant AD
- Edwards RM,
- Stack EJ,
- Trizna W
- Amberg GC,
- Rossow CF,
- Navedo MF,
- Santana LF
- Amberg GC,
- Santana LF
- Tobin AA,
- Joseph BK,
- Al-Kindi HN,
- Albarwani S,
- Madden JA,
- Nemetz LT,
- Rusch NJ,
- Rhee SW
- Poyner DR,
- Sexton PM,
- Marshall I,
- Smith DM,
- Quirion R,
- Born W,
- Muff R,
- Fischer JA,
- Foord SM
- Kitazono T,
- Heistad DD,
- Faraci FM
- Kleppisch T,
- Nelson MT
- Schwake M,
- Jentsch TJ,
- Friedrich T
- Brueggemann LI,
- Mackie AR,
- Cribbs LL,
- Freda J,
- Tripathi A,
- Majetschak M,
- Byron KL
- Chen TT,
- Luykenaar KD,
- Walsh EJ,
- Walsh MP,
- Cole WC
- Plane F,
- Johnson R,
- Kerr P,
- Wiehler W,
- Thorneloe K,
- Ishii K,
- Chen T,
- Cole W
KCNQ-encoded Kv channels (Kv7) regulate arterial tone and contribute to relaxations produced by various Gs-linked receptor agonists. This study reveals that in middle cerebral arteries (MCAs), the activation of Kv7.4 also underlies the major component of vasodilation to calcitonin gene-related peptide as well as the pressure-induced myogenic response. Moreover, we provide structural insight into the architecture of arterial Kv7, which is likely to predominate as a Kv7.4/7.5 heteromer, and show that Kv7 activity is preserved in MCA from hypertensive rats in contrast to systemic arteries. Consequently, Kv7.4 channels are an effective therapeutic target in cerebrovascular disease.